diff --git "a/checkpoint-200245/trainer_state.json" "b/checkpoint-200245/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-200245/trainer_state.json"
@@ -0,0 +1,140201 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 200245,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 4.993882493944918e-05,
+      "grad_norm": 0.09688733518123627,
+      "learning_rate": 6.000000000000001e-07,
+      "loss": 10.7967,
+      "step": 10
+    },
+    {
+      "epoch": 9.987764987889836e-05,
+      "grad_norm": 0.09242203086614609,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 10.7959,
+      "step": 20
+    },
+    {
+      "epoch": 0.00014981647481834753,
+      "grad_norm": 0.09640923142433167,
+      "learning_rate": 1.8e-06,
+      "loss": 10.7964,
+      "step": 30
+    },
+    {
+      "epoch": 0.0001997552997577967,
+      "grad_norm": 0.09338584542274475,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 10.7963,
+      "step": 40
+    },
+    {
+      "epoch": 0.00024969412469724587,
+      "grad_norm": 0.09739788621664047,
+      "learning_rate": 3e-06,
+      "loss": 10.796,
+      "step": 50
+    },
+    {
+      "epoch": 0.00029963294963669505,
+      "grad_norm": 0.09315921366214752,
+      "learning_rate": 3.6e-06,
+      "loss": 10.7954,
+      "step": 60
+    },
+    {
+      "epoch": 0.00034957177457614424,
+      "grad_norm": 0.09130595624446869,
+      "learning_rate": 4.2000000000000004e-06,
+      "loss": 10.7956,
+      "step": 70
+    },
+    {
+      "epoch": 0.0003995105995155934,
+      "grad_norm": 0.09694769978523254,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 10.7961,
+      "step": 80
+    },
+    {
+      "epoch": 0.00044944942445504255,
+      "grad_norm": 0.09814093261957169,
+      "learning_rate": 5.4e-06,
+      "loss": 10.7956,
+      "step": 90
+    },
+    {
+      "epoch": 0.0004993882493944917,
+      "grad_norm": 0.0940699502825737,
+      "learning_rate": 6e-06,
+      "loss": 10.7957,
+      "step": 100
+    },
+    {
+      "epoch": 0.000549327074333941,
+      "grad_norm": 0.09281250834465027,
+      "learning_rate": 6.6e-06,
+      "loss": 10.7957,
+      "step": 110
+    },
+    {
+      "epoch": 0.0005992658992733901,
+      "grad_norm": 0.09708160907030106,
+      "learning_rate": 7.2e-06,
+      "loss": 10.7961,
+      "step": 120
+    },
+    {
+      "epoch": 0.0006492047242128392,
+      "grad_norm": 0.09620024263858795,
+      "learning_rate": 7.8e-06,
+      "loss": 10.7957,
+      "step": 130
+    },
+    {
+      "epoch": 0.0006991435491522885,
+      "grad_norm": 0.09342040121555328,
+      "learning_rate": 8.400000000000001e-06,
+      "loss": 10.7958,
+      "step": 140
+    },
+    {
+      "epoch": 0.0007490823740917376,
+      "grad_norm": 0.09522400051355362,
+      "learning_rate": 9e-06,
+      "loss": 10.7951,
+      "step": 150
+    },
+    {
+      "epoch": 0.0007990211990311868,
+      "grad_norm": 0.09366803616285324,
+      "learning_rate": 9.600000000000001e-06,
+      "loss": 10.7963,
+      "step": 160
+    },
+    {
+      "epoch": 0.000848960023970636,
+      "grad_norm": 0.10423287004232407,
+      "learning_rate": 1.02e-05,
+      "loss": 10.796,
+      "step": 170
+    },
+    {
+      "epoch": 0.0008988988489100851,
+      "grad_norm": 0.09598524123430252,
+      "learning_rate": 1.08e-05,
+      "loss": 10.7951,
+      "step": 180
+    },
+    {
+      "epoch": 0.0009488376738495343,
+      "grad_norm": 0.09360150247812271,
+      "learning_rate": 1.1400000000000001e-05,
+      "loss": 10.795,
+      "step": 190
+    },
+    {
+      "epoch": 0.0009987764987889835,
+      "grad_norm": 0.09207968413829803,
+      "learning_rate": 1.2e-05,
+      "loss": 10.7947,
+      "step": 200
+    },
+    {
+      "epoch": 0.0010487153237284327,
+      "grad_norm": 0.09178674221038818,
+      "learning_rate": 1.26e-05,
+      "loss": 10.7944,
+      "step": 210
+    },
+    {
+      "epoch": 0.001098654148667882,
+      "grad_norm": 0.09687130898237228,
+      "learning_rate": 1.32e-05,
+      "loss": 10.7945,
+      "step": 220
+    },
+    {
+      "epoch": 0.001148592973607331,
+      "grad_norm": 0.08996697515249252,
+      "learning_rate": 1.3800000000000002e-05,
+      "loss": 10.7939,
+      "step": 230
+    },
+    {
+      "epoch": 0.0011985317985467802,
+      "grad_norm": 0.09358258545398712,
+      "learning_rate": 1.44e-05,
+      "loss": 10.7943,
+      "step": 240
+    },
+    {
+      "epoch": 0.0012484706234862294,
+      "grad_norm": 0.09620951116085052,
+      "learning_rate": 1.5e-05,
+      "loss": 12.5216,
+      "step": 250
+    },
+    {
+      "epoch": 0.0012984094484256785,
+      "grad_norm": 12312.0439453125,
+      "learning_rate": 1.56e-05,
+      "loss": 15.3101,
+      "step": 260
+    },
+    {
+      "epoch": 0.0013483482733651277,
+      "grad_norm": 4373971.5,
+      "learning_rate": 1.62e-05,
+      "loss": 11.4047,
+      "step": 270
+    },
+    {
+      "epoch": 0.001398287098304577,
+      "grad_norm": 0.09267476201057434,
+      "learning_rate": 1.6800000000000002e-05,
+      "loss": 13.7008,
+      "step": 280
+    },
+    {
+      "epoch": 0.001448225923244026,
+      "grad_norm": 0.09616652131080627,
+      "learning_rate": 1.74e-05,
+      "loss": 10.7943,
+      "step": 290
+    },
+    {
+      "epoch": 0.0014981647481834752,
+      "grad_norm": 0.09366879612207413,
+      "learning_rate": 1.8e-05,
+      "loss": 10.7936,
+      "step": 300
+    },
+    {
+      "epoch": 0.0015481035731229244,
+      "grad_norm": 0.10064131766557693,
+      "learning_rate": 1.86e-05,
+      "loss": 10.7939,
+      "step": 310
+    },
+    {
+      "epoch": 0.0015980423980623737,
+      "grad_norm": 0.09544051438570023,
+      "learning_rate": 1.9200000000000003e-05,
+      "loss": 10.7934,
+      "step": 320
+    },
+    {
+      "epoch": 0.0016479812230018227,
+      "grad_norm": 0.09638150781393051,
+      "learning_rate": 1.98e-05,
+      "loss": 10.7934,
+      "step": 330
+    },
+    {
+      "epoch": 0.001697920047941272,
+      "grad_norm": 0.09321698546409607,
+      "learning_rate": 2.04e-05,
+      "loss": 10.793,
+      "step": 340
+    },
+    {
+      "epoch": 0.0017478588728807212,
+      "grad_norm": 0.09490924328565598,
+      "learning_rate": 2.1e-05,
+      "loss": 10.7924,
+      "step": 350
+    },
+    {
+      "epoch": 0.0017977976978201702,
+      "grad_norm": 0.09169238060712814,
+      "learning_rate": 2.16e-05,
+      "loss": 10.7926,
+      "step": 360
+    },
+    {
+      "epoch": 0.0018477365227596194,
+      "grad_norm": 0.09536827355623245,
+      "learning_rate": 2.22e-05,
+      "loss": 10.7916,
+      "step": 370
+    },
+    {
+      "epoch": 0.0018976753476990687,
+      "grad_norm": 0.0935748964548111,
+      "learning_rate": 2.2800000000000002e-05,
+      "loss": 10.7922,
+      "step": 380
+    },
+    {
+      "epoch": 0.001947614172638518,
+      "grad_norm": 0.09191051870584488,
+      "learning_rate": 2.3400000000000003e-05,
+      "loss": 10.7917,
+      "step": 390
+    },
+    {
+      "epoch": 0.001997552997577967,
+      "grad_norm": 0.09062574058771133,
+      "learning_rate": 2.4e-05,
+      "loss": 10.7913,
+      "step": 400
+    },
+    {
+      "epoch": 0.002047491822517416,
+      "grad_norm": 0.09686335176229477,
+      "learning_rate": 2.4599999999999998e-05,
+      "loss": 10.7911,
+      "step": 410
+    },
+    {
+      "epoch": 0.0020974306474568654,
+      "grad_norm": 0.09458820521831512,
+      "learning_rate": 2.52e-05,
+      "loss": 10.7904,
+      "step": 420
+    },
+    {
+      "epoch": 0.0021473694723963144,
+      "grad_norm": 0.09392788261175156,
+      "learning_rate": 2.58e-05,
+      "loss": 10.79,
+      "step": 430
+    },
+    {
+      "epoch": 0.002197308297335764,
+      "grad_norm": 0.09329679608345032,
+      "learning_rate": 2.64e-05,
+      "loss": 10.7903,
+      "step": 440
+    },
+    {
+      "epoch": 0.002247247122275213,
+      "grad_norm": 0.09585896879434586,
+      "learning_rate": 2.7000000000000002e-05,
+      "loss": 10.7897,
+      "step": 450
+    },
+    {
+      "epoch": 0.002297185947214662,
+      "grad_norm": 0.0919712483882904,
+      "learning_rate": 2.7600000000000003e-05,
+      "loss": 10.7897,
+      "step": 460
+    },
+    {
+      "epoch": 0.0023471247721541114,
+      "grad_norm": 0.09430967271327972,
+      "learning_rate": 2.8199999999999998e-05,
+      "loss": 10.7888,
+      "step": 470
+    },
+    {
+      "epoch": 0.0023970635970935604,
+      "grad_norm": 0.09424952417612076,
+      "learning_rate": 2.88e-05,
+      "loss": 10.7897,
+      "step": 480
+    },
+    {
+      "epoch": 0.0024470024220330094,
+      "grad_norm": 0.09542497992515564,
+      "learning_rate": 2.94e-05,
+      "loss": 10.7901,
+      "step": 490
+    },
+    {
+      "epoch": 0.002496941246972459,
+      "grad_norm": 0.09828152507543564,
+      "learning_rate": 3e-05,
+      "loss": 11.0404,
+      "step": 500
+    },
+    {
+      "epoch": 0.002546880071911908,
+      "grad_norm": 0.09262736141681671,
+      "learning_rate": 2.9998498085058448e-05,
+      "loss": 10.7884,
+      "step": 510
+    },
+    {
+      "epoch": 0.002596818896851357,
+      "grad_norm": 0.0968865230679512,
+      "learning_rate": 2.99969961701169e-05,
+      "loss": 10.7878,
+      "step": 520
+    },
+    {
+      "epoch": 0.0026467577217908064,
+      "grad_norm": 0.0930342897772789,
+      "learning_rate": 2.9995494255175348e-05,
+      "loss": 10.7871,
+      "step": 530
+    },
+    {
+      "epoch": 0.0026966965467302554,
+      "grad_norm": 0.0979408249258995,
+      "learning_rate": 2.99939923402338e-05,
+      "loss": 10.7879,
+      "step": 540
+    },
+    {
+      "epoch": 0.0027466353716697044,
+      "grad_norm": 0.09784050285816193,
+      "learning_rate": 2.9992490425292248e-05,
+      "loss": 10.7864,
+      "step": 550
+    },
+    {
+      "epoch": 0.002796574196609154,
+      "grad_norm": 0.0974036231637001,
+      "learning_rate": 2.9990988510350695e-05,
+      "loss": 10.786,
+      "step": 560
+    },
+    {
+      "epoch": 0.002846513021548603,
+      "grad_norm": 0.09453345090150833,
+      "learning_rate": 2.998948659540915e-05,
+      "loss": 10.786,
+      "step": 570
+    },
+    {
+      "epoch": 0.002896451846488052,
+      "grad_norm": 0.09815464913845062,
+      "learning_rate": 2.9987984680467595e-05,
+      "loss": 10.7855,
+      "step": 580
+    },
+    {
+      "epoch": 0.0029463906714275014,
+      "grad_norm": 0.0883261188864708,
+      "learning_rate": 2.998648276552605e-05,
+      "loss": 10.7854,
+      "step": 590
+    },
+    {
+      "epoch": 0.0029963294963669504,
+      "grad_norm": 0.09347826987504959,
+      "learning_rate": 2.9984980850584496e-05,
+      "loss": 10.785,
+      "step": 600
+    },
+    {
+      "epoch": 0.0030462683213064,
+      "grad_norm": 0.09312307834625244,
+      "learning_rate": 2.9983478935642943e-05,
+      "loss": 10.7847,
+      "step": 610
+    },
+    {
+      "epoch": 0.003096207146245849,
+      "grad_norm": 0.09311032295227051,
+      "learning_rate": 2.9981977020701396e-05,
+      "loss": 10.785,
+      "step": 620
+    },
+    {
+      "epoch": 0.003146145971185298,
+      "grad_norm": 0.09509855508804321,
+      "learning_rate": 2.9980475105759843e-05,
+      "loss": 10.7844,
+      "step": 630
+    },
+    {
+      "epoch": 0.0031960847961247474,
+      "grad_norm": 0.09063342213630676,
+      "learning_rate": 2.9978973190818296e-05,
+      "loss": 10.7836,
+      "step": 640
+    },
+    {
+      "epoch": 0.0032460236210641964,
+      "grad_norm": 0.095396488904953,
+      "learning_rate": 2.9977471275876743e-05,
+      "loss": 10.7838,
+      "step": 650
+    },
+    {
+      "epoch": 0.0032959624460036454,
+      "grad_norm": 0.09543643146753311,
+      "learning_rate": 2.997596936093519e-05,
+      "loss": 10.7835,
+      "step": 660
+    },
+    {
+      "epoch": 0.003345901270943095,
+      "grad_norm": 11308.7890625,
+      "learning_rate": 2.9974467445993644e-05,
+      "loss": 11.5043,
+      "step": 670
+    },
+    {
+      "epoch": 0.003395840095882544,
+      "grad_norm": 0.09934733808040619,
+      "learning_rate": 2.997296553105209e-05,
+      "loss": 17.0433,
+      "step": 680
+    },
+    {
+      "epoch": 0.003445778920821993,
+      "grad_norm": 0.09415256977081299,
+      "learning_rate": 2.9971463616110544e-05,
+      "loss": 10.9281,
+      "step": 690
+    },
+    {
+      "epoch": 0.0034957177457614424,
+      "grad_norm": 0.09561091661453247,
+      "learning_rate": 2.996996170116899e-05,
+      "loss": 12.0749,
+      "step": 700
+    },
+    {
+      "epoch": 0.0035456565707008914,
+      "grad_norm": 308366.0625,
+      "learning_rate": 2.996845978622744e-05,
+      "loss": 65.9464,
+      "step": 710
+    },
+    {
+      "epoch": 0.0035955953956403404,
+      "grad_norm": 36107.77734375,
+      "learning_rate": 2.996695787128589e-05,
+      "loss": 87.2685,
+      "step": 720
+    },
+    {
+      "epoch": 0.00364553422057979,
+      "grad_norm": 0.09354925155639648,
+      "learning_rate": 2.9965455956344338e-05,
+      "loss": 10.9749,
+      "step": 730
+    },
+    {
+      "epoch": 0.003695473045519239,
+      "grad_norm": 0.09494537115097046,
+      "learning_rate": 2.996395404140279e-05,
+      "loss": 10.7808,
+      "step": 740
+    },
+    {
+      "epoch": 0.003745411870458688,
+      "grad_norm": 0.10302335023880005,
+      "learning_rate": 2.9962452126461238e-05,
+      "loss": 13.2444,
+      "step": 750
+    },
+    {
+      "epoch": 0.0037953506953981374,
+      "grad_norm": 0.09412359446287155,
+      "learning_rate": 2.996095021151969e-05,
+      "loss": 11.111,
+      "step": 760
+    },
+    {
+      "epoch": 0.0038452895203375864,
+      "grad_norm": 0.10018070042133331,
+      "learning_rate": 2.995944829657814e-05,
+      "loss": 10.7808,
+      "step": 770
+    },
+    {
+      "epoch": 0.003895228345277036,
+      "grad_norm": 0.09139321744441986,
+      "learning_rate": 2.9957946381636585e-05,
+      "loss": 10.7802,
+      "step": 780
+    },
+    {
+      "epoch": 0.003945167170216484,
+      "grad_norm": 0.09905160218477249,
+      "learning_rate": 2.995644446669504e-05,
+      "loss": 10.7798,
+      "step": 790
+    },
+    {
+      "epoch": 0.003995105995155934,
+      "grad_norm": 0.09502265602350235,
+      "learning_rate": 2.9954942551753486e-05,
+      "loss": 10.7806,
+      "step": 800
+    },
+    {
+      "epoch": 0.004045044820095383,
+      "grad_norm": 0.09315773844718933,
+      "learning_rate": 2.9953440636811936e-05,
+      "loss": 10.7789,
+      "step": 810
+    },
+    {
+      "epoch": 0.004094983645034832,
+      "grad_norm": 0.09650130569934845,
+      "learning_rate": 2.9951938721870386e-05,
+      "loss": 10.7788,
+      "step": 820
+    },
+    {
+      "epoch": 0.004144922469974281,
+      "grad_norm": 0.09284522384405136,
+      "learning_rate": 2.9950436806928833e-05,
+      "loss": 10.7786,
+      "step": 830
+    },
+    {
+      "epoch": 0.004194861294913731,
+      "grad_norm": 0.09788000583648682,
+      "learning_rate": 2.9948934891987286e-05,
+      "loss": 10.7783,
+      "step": 840
+    },
+    {
+      "epoch": 0.004244800119853179,
+      "grad_norm": 0.09668333828449249,
+      "learning_rate": 2.9947432977045733e-05,
+      "loss": 10.7779,
+      "step": 850
+    },
+    {
+      "epoch": 0.004294738944792629,
+      "grad_norm": 0.09662320464849472,
+      "learning_rate": 2.9945931062104183e-05,
+      "loss": 10.7769,
+      "step": 860
+    },
+    {
+      "epoch": 0.004344677769732078,
+      "grad_norm": 0.09260370582342148,
+      "learning_rate": 2.9944429147162634e-05,
+      "loss": 10.7772,
+      "step": 870
+    },
+    {
+      "epoch": 0.004394616594671528,
+      "grad_norm": 0.09133943170309067,
+      "learning_rate": 2.994292723222108e-05,
+      "loss": 10.7774,
+      "step": 880
+    },
+    {
+      "epoch": 0.004444555419610976,
+      "grad_norm": 0.09417100250720978,
+      "learning_rate": 2.9941425317279534e-05,
+      "loss": 10.7768,
+      "step": 890
+    },
+    {
+      "epoch": 0.004494494244550426,
+      "grad_norm": 0.09618920087814331,
+      "learning_rate": 2.993992340233798e-05,
+      "loss": 10.776,
+      "step": 900
+    },
+    {
+      "epoch": 0.004544433069489875,
+      "grad_norm": 0.09578631818294525,
+      "learning_rate": 2.993842148739643e-05,
+      "loss": 10.7759,
+      "step": 910
+    },
+    {
+      "epoch": 0.004594371894429324,
+      "grad_norm": 0.09697435051202774,
+      "learning_rate": 2.993691957245488e-05,
+      "loss": 10.7744,
+      "step": 920
+    },
+    {
+      "epoch": 0.004644310719368773,
+      "grad_norm": 0.09273876994848251,
+      "learning_rate": 2.9935417657513328e-05,
+      "loss": 10.7755,
+      "step": 930
+    },
+    {
+      "epoch": 0.004694249544308223,
+      "grad_norm": 0.0970180332660675,
+      "learning_rate": 2.993391574257178e-05,
+      "loss": 10.7751,
+      "step": 940
+    },
+    {
+      "epoch": 0.004744188369247671,
+      "grad_norm": 0.0976632609963417,
+      "learning_rate": 2.9932413827630228e-05,
+      "loss": 10.7745,
+      "step": 950
+    },
+    {
+      "epoch": 0.004794127194187121,
+      "grad_norm": 0.09234705567359924,
+      "learning_rate": 2.993091191268868e-05,
+      "loss": 10.7746,
+      "step": 960
+    },
+    {
+      "epoch": 0.00484406601912657,
+      "grad_norm": 0.09593399614095688,
+      "learning_rate": 2.992940999774713e-05,
+      "loss": 10.7742,
+      "step": 970
+    },
+    {
+      "epoch": 0.004894004844066019,
+      "grad_norm": 0.09624140709638596,
+      "learning_rate": 2.9927908082805575e-05,
+      "loss": 10.7732,
+      "step": 980
+    },
+    {
+      "epoch": 0.004943943669005468,
+      "grad_norm": 0.09578274935483932,
+      "learning_rate": 2.992640616786403e-05,
+      "loss": 10.7728,
+      "step": 990
+    },
+    {
+      "epoch": 0.004993882493944918,
+      "grad_norm": 0.10335574299097061,
+      "learning_rate": 2.9924904252922476e-05,
+      "loss": 10.7731,
+      "step": 1000
+    },
+    {
+      "epoch": 0.005043821318884366,
+      "grad_norm": 0.09355924278497696,
+      "learning_rate": 2.9923402337980926e-05,
+      "loss": 10.7723,
+      "step": 1010
+    },
+    {
+      "epoch": 0.005093760143823816,
+      "grad_norm": 0.09353188425302505,
+      "learning_rate": 2.9921900423039376e-05,
+      "loss": 10.773,
+      "step": 1020
+    },
+    {
+      "epoch": 0.005143698968763265,
+      "grad_norm": 0.09189878404140472,
+      "learning_rate": 2.9920398508097826e-05,
+      "loss": 10.7719,
+      "step": 1030
+    },
+    {
+      "epoch": 0.005193637793702714,
+      "grad_norm": 0.09627330303192139,
+      "learning_rate": 2.9918896593156277e-05,
+      "loss": 10.7721,
+      "step": 1040
+    },
+    {
+      "epoch": 0.005243576618642163,
+      "grad_norm": 0.09078870713710785,
+      "learning_rate": 2.9917394678214723e-05,
+      "loss": 10.7711,
+      "step": 1050
+    },
+    {
+      "epoch": 0.005293515443581613,
+      "grad_norm": 0.09166533499956131,
+      "learning_rate": 2.9915892763273173e-05,
+      "loss": 10.7711,
+      "step": 1060
+    },
+    {
+      "epoch": 0.005343454268521061,
+      "grad_norm": 0.0925443097949028,
+      "learning_rate": 2.9914390848331624e-05,
+      "loss": 10.771,
+      "step": 1070
+    },
+    {
+      "epoch": 0.005393393093460511,
+      "grad_norm": 0.09320785850286484,
+      "learning_rate": 2.9912888933390074e-05,
+      "loss": 10.7703,
+      "step": 1080
+    },
+    {
+      "epoch": 0.00544333191839996,
+      "grad_norm": 0.09314128011465073,
+      "learning_rate": 2.9911387018448524e-05,
+      "loss": 10.7707,
+      "step": 1090
+    },
+    {
+      "epoch": 0.005493270743339409,
+      "grad_norm": 0.09605088829994202,
+      "learning_rate": 2.990988510350697e-05,
+      "loss": 10.7705,
+      "step": 1100
+    },
+    {
+      "epoch": 0.005543209568278858,
+      "grad_norm": 0.08931813389062881,
+      "learning_rate": 2.990838318856542e-05,
+      "loss": 10.7689,
+      "step": 1110
+    },
+    {
+      "epoch": 0.005593148393218308,
+      "grad_norm": 0.09207310527563095,
+      "learning_rate": 2.990688127362387e-05,
+      "loss": 10.7691,
+      "step": 1120
+    },
+    {
+      "epoch": 0.005643087218157756,
+      "grad_norm": 0.09695383906364441,
+      "learning_rate": 2.990537935868232e-05,
+      "loss": 10.7687,
+      "step": 1130
+    },
+    {
+      "epoch": 0.005693026043097206,
+      "grad_norm": 0.09819135814905167,
+      "learning_rate": 2.990387744374077e-05,
+      "loss": 10.7679,
+      "step": 1140
+    },
+    {
+      "epoch": 0.005742964868036655,
+      "grad_norm": 0.09095334261655807,
+      "learning_rate": 2.990237552879922e-05,
+      "loss": 10.7685,
+      "step": 1150
+    },
+    {
+      "epoch": 0.005792903692976104,
+      "grad_norm": 0.08971603959798813,
+      "learning_rate": 2.990087361385767e-05,
+      "loss": 10.7678,
+      "step": 1160
+    },
+    {
+      "epoch": 0.005842842517915553,
+      "grad_norm": 0.0930616483092308,
+      "learning_rate": 2.989937169891612e-05,
+      "loss": 10.7679,
+      "step": 1170
+    },
+    {
+      "epoch": 0.005892781342855003,
+      "grad_norm": 0.09826360642910004,
+      "learning_rate": 2.989786978397457e-05,
+      "loss": 10.767,
+      "step": 1180
+    },
+    {
+      "epoch": 0.005942720167794451,
+      "grad_norm": 0.09297499060630798,
+      "learning_rate": 2.989636786903302e-05,
+      "loss": 10.7671,
+      "step": 1190
+    },
+    {
+      "epoch": 0.005992658992733901,
+      "grad_norm": 0.09357722103595734,
+      "learning_rate": 2.9894865954091466e-05,
+      "loss": 10.7661,
+      "step": 1200
+    },
+    {
+      "epoch": 0.00604259781767335,
+      "grad_norm": 0.09486209601163864,
+      "learning_rate": 2.9893364039149916e-05,
+      "loss": 10.7658,
+      "step": 1210
+    },
+    {
+      "epoch": 0.0060925366426128,
+      "grad_norm": 0.09341169148683548,
+      "learning_rate": 2.9891862124208366e-05,
+      "loss": 10.7663,
+      "step": 1220
+    },
+    {
+      "epoch": 0.006142475467552248,
+      "grad_norm": 0.09570655226707458,
+      "learning_rate": 2.9890360209266816e-05,
+      "loss": 10.7658,
+      "step": 1230
+    },
+    {
+      "epoch": 0.006192414292491698,
+      "grad_norm": 0.09547128528356552,
+      "learning_rate": 2.9888858294325267e-05,
+      "loss": 10.764,
+      "step": 1240
+    },
+    {
+      "epoch": 0.006242353117431147,
+      "grad_norm": 0.08968691527843475,
+      "learning_rate": 2.9887356379383713e-05,
+      "loss": 10.7649,
+      "step": 1250
+    },
+    {
+      "epoch": 0.006292291942370596,
+      "grad_norm": 0.09444384276866913,
+      "learning_rate": 2.9885854464442163e-05,
+      "loss": 10.7647,
+      "step": 1260
+    },
+    {
+      "epoch": 0.006342230767310045,
+      "grad_norm": 0.0966639518737793,
+      "learning_rate": 2.9884352549500614e-05,
+      "loss": 10.764,
+      "step": 1270
+    },
+    {
+      "epoch": 0.006392169592249495,
+      "grad_norm": 0.09153666347265244,
+      "learning_rate": 2.9882850634559064e-05,
+      "loss": 10.7646,
+      "step": 1280
+    },
+    {
+      "epoch": 0.006442108417188943,
+      "grad_norm": 0.09613633900880814,
+      "learning_rate": 2.9881348719617514e-05,
+      "loss": 10.763,
+      "step": 1290
+    },
+    {
+      "epoch": 0.006492047242128393,
+      "grad_norm": 0.09370370209217072,
+      "learning_rate": 2.987984680467596e-05,
+      "loss": 10.7634,
+      "step": 1300
+    },
+    {
+      "epoch": 0.006541986067067842,
+      "grad_norm": 0.09263869374990463,
+      "learning_rate": 2.987834488973441e-05,
+      "loss": 10.7635,
+      "step": 1310
+    },
+    {
+      "epoch": 0.006591924892007291,
+      "grad_norm": 0.09146302193403244,
+      "learning_rate": 2.987684297479286e-05,
+      "loss": 10.7622,
+      "step": 1320
+    },
+    {
+      "epoch": 0.00664186371694674,
+      "grad_norm": 0.09400025755167007,
+      "learning_rate": 2.987534105985131e-05,
+      "loss": 10.7627,
+      "step": 1330
+    },
+    {
+      "epoch": 0.00669180254188619,
+      "grad_norm": 0.09636477380990982,
+      "learning_rate": 2.987383914490976e-05,
+      "loss": 10.762,
+      "step": 1340
+    },
+    {
+      "epoch": 0.006741741366825638,
+      "grad_norm": 0.08832681179046631,
+      "learning_rate": 2.9872337229968212e-05,
+      "loss": 10.7615,
+      "step": 1350
+    },
+    {
+      "epoch": 0.006791680191765088,
+      "grad_norm": 0.10007186979055405,
+      "learning_rate": 2.987083531502666e-05,
+      "loss": 10.7612,
+      "step": 1360
+    },
+    {
+      "epoch": 0.006841619016704537,
+      "grad_norm": 0.09131281077861786,
+      "learning_rate": 2.986933340008511e-05,
+      "loss": 10.761,
+      "step": 1370
+    },
+    {
+      "epoch": 0.006891557841643986,
+      "grad_norm": 0.0982031375169754,
+      "learning_rate": 2.986783148514356e-05,
+      "loss": 10.7597,
+      "step": 1380
+    },
+    {
+      "epoch": 0.006941496666583435,
+      "grad_norm": 0.10071565955877304,
+      "learning_rate": 2.986632957020201e-05,
+      "loss": 10.7601,
+      "step": 1390
+    },
+    {
+      "epoch": 0.006991435491522885,
+      "grad_norm": 0.09393281489610672,
+      "learning_rate": 2.986482765526046e-05,
+      "loss": 10.7598,
+      "step": 1400
+    },
+    {
+      "epoch": 0.007041374316462333,
+      "grad_norm": 0.09201990813016891,
+      "learning_rate": 2.9863325740318906e-05,
+      "loss": 10.7588,
+      "step": 1410
+    },
+    {
+      "epoch": 0.007091313141401783,
+      "grad_norm": 0.09910114109516144,
+      "learning_rate": 2.9861823825377356e-05,
+      "loss": 10.7597,
+      "step": 1420
+    },
+    {
+      "epoch": 0.007141251966341232,
+      "grad_norm": 0.09161806851625443,
+      "learning_rate": 2.9860321910435806e-05,
+      "loss": 10.7587,
+      "step": 1430
+    },
+    {
+      "epoch": 0.007191190791280681,
+      "grad_norm": 0.09380662441253662,
+      "learning_rate": 2.9858819995494257e-05,
+      "loss": 10.7589,
+      "step": 1440
+    },
+    {
+      "epoch": 0.00724112961622013,
+      "grad_norm": 0.09290491044521332,
+      "learning_rate": 2.9857318080552707e-05,
+      "loss": 10.7591,
+      "step": 1450
+    },
+    {
+      "epoch": 0.00729106844115958,
+      "grad_norm": 0.09262260794639587,
+      "learning_rate": 2.9855816165611154e-05,
+      "loss": 10.7581,
+      "step": 1460
+    },
+    {
+      "epoch": 0.007341007266099028,
+      "grad_norm": 698.4036865234375,
+      "learning_rate": 2.9854314250669604e-05,
+      "loss": 10.758,
+      "step": 1470
+    },
+    {
+      "epoch": 0.007390946091038478,
+      "grad_norm": 0.09542746841907501,
+      "learning_rate": 2.9852812335728054e-05,
+      "loss": 12.0635,
+      "step": 1480
+    },
+    {
+      "epoch": 0.007440884915977927,
+      "grad_norm": 0.09986045211553574,
+      "learning_rate": 2.9851310420786504e-05,
+      "loss": 10.7564,
+      "step": 1490
+    },
+    {
+      "epoch": 0.007490823740917376,
+      "grad_norm": 0.09409373998641968,
+      "learning_rate": 2.9849808505844954e-05,
+      "loss": 10.7568,
+      "step": 1500
+    },
+    {
+      "epoch": 0.007540762565856825,
+      "grad_norm": 0.092925526201725,
+      "learning_rate": 2.98483065909034e-05,
+      "loss": 10.7565,
+      "step": 1510
+    },
+    {
+      "epoch": 0.007590701390796275,
+      "grad_norm": 0.0927615836262703,
+      "learning_rate": 2.984680467596185e-05,
+      "loss": 10.7556,
+      "step": 1520
+    },
+    {
+      "epoch": 0.007640640215735723,
+      "grad_norm": 0.10118461400270462,
+      "learning_rate": 2.98453027610203e-05,
+      "loss": 10.7564,
+      "step": 1530
+    },
+    {
+      "epoch": 0.007690579040675173,
+      "grad_norm": 0.09674585610628128,
+      "learning_rate": 2.984380084607875e-05,
+      "loss": 10.7554,
+      "step": 1540
+    },
+    {
+      "epoch": 0.007740517865614622,
+      "grad_norm": 0.09508625417947769,
+      "learning_rate": 2.9842298931137202e-05,
+      "loss": 10.7555,
+      "step": 1550
+    },
+    {
+      "epoch": 0.007790456690554072,
+      "grad_norm": 0.0912444218993187,
+      "learning_rate": 2.984079701619565e-05,
+      "loss": 10.7547,
+      "step": 1560
+    },
+    {
+      "epoch": 0.007840395515493521,
+      "grad_norm": 0.09958932548761368,
+      "learning_rate": 2.98392951012541e-05,
+      "loss": 10.7539,
+      "step": 1570
+    },
+    {
+      "epoch": 0.007890334340432969,
+      "grad_norm": 0.0945759117603302,
+      "learning_rate": 2.983779318631255e-05,
+      "loss": 10.7541,
+      "step": 1580
+    },
+    {
+      "epoch": 0.007940273165372418,
+      "grad_norm": 0.09051518142223358,
+      "learning_rate": 2.9836291271371e-05,
+      "loss": 10.7533,
+      "step": 1590
+    },
+    {
+      "epoch": 0.007990211990311868,
+      "grad_norm": 0.09735298156738281,
+      "learning_rate": 2.983478935642945e-05,
+      "loss": 10.753,
+      "step": 1600
+    },
+    {
+      "epoch": 0.008040150815251317,
+      "grad_norm": 0.09062410145998001,
+      "learning_rate": 2.9833287441487896e-05,
+      "loss": 10.7529,
+      "step": 1610
+    },
+    {
+      "epoch": 0.008090089640190767,
+      "grad_norm": 0.091838039457798,
+      "learning_rate": 2.9831785526546346e-05,
+      "loss": 10.7524,
+      "step": 1620
+    },
+    {
+      "epoch": 0.008140028465130216,
+      "grad_norm": 0.0958361029624939,
+      "learning_rate": 2.9830283611604796e-05,
+      "loss": 10.7529,
+      "step": 1630
+    },
+    {
+      "epoch": 0.008189967290069664,
+      "grad_norm": 0.09625451266765594,
+      "learning_rate": 2.9828781696663247e-05,
+      "loss": 10.7514,
+      "step": 1640
+    },
+    {
+      "epoch": 0.008239906115009113,
+      "grad_norm": 0.0954735055565834,
+      "learning_rate": 2.9827279781721697e-05,
+      "loss": 10.7526,
+      "step": 1650
+    },
+    {
+      "epoch": 0.008289844939948563,
+      "grad_norm": 0.09449174255132675,
+      "learning_rate": 2.9825777866780144e-05,
+      "loss": 10.7519,
+      "step": 1660
+    },
+    {
+      "epoch": 0.008339783764888012,
+      "grad_norm": 0.08916933834552765,
+      "learning_rate": 2.9824275951838597e-05,
+      "loss": 10.7517,
+      "step": 1670
+    },
+    {
+      "epoch": 0.008389722589827462,
+      "grad_norm": 0.0955243706703186,
+      "learning_rate": 2.9822774036897044e-05,
+      "loss": 10.7513,
+      "step": 1680
+    },
+    {
+      "epoch": 0.008439661414766911,
+      "grad_norm": 0.0943785309791565,
+      "learning_rate": 2.9821272121955494e-05,
+      "loss": 10.7504,
+      "step": 1690
+    },
+    {
+      "epoch": 0.008489600239706359,
+      "grad_norm": 0.09509387612342834,
+      "learning_rate": 2.9819770207013944e-05,
+      "loss": 10.7505,
+      "step": 1700
+    },
+    {
+      "epoch": 0.008539539064645808,
+      "grad_norm": 0.09808336943387985,
+      "learning_rate": 2.981826829207239e-05,
+      "loss": 10.7496,
+      "step": 1710
+    },
+    {
+      "epoch": 0.008589477889585258,
+      "grad_norm": 0.0928177759051323,
+      "learning_rate": 2.9816766377130845e-05,
+      "loss": 10.7501,
+      "step": 1720
+    },
+    {
+      "epoch": 0.008639416714524707,
+      "grad_norm": 0.09282363951206207,
+      "learning_rate": 2.981526446218929e-05,
+      "loss": 10.7484,
+      "step": 1730
+    },
+    {
+      "epoch": 0.008689355539464157,
+      "grad_norm": 0.09041490405797958,
+      "learning_rate": 2.981376254724774e-05,
+      "loss": 10.7492,
+      "step": 1740
+    },
+    {
+      "epoch": 0.008739294364403606,
+      "grad_norm": 0.09924125671386719,
+      "learning_rate": 2.9812260632306192e-05,
+      "loss": 10.7486,
+      "step": 1750
+    },
+    {
+      "epoch": 0.008789233189343056,
+      "grad_norm": 0.09147768467664719,
+      "learning_rate": 2.981075871736464e-05,
+      "loss": 10.7488,
+      "step": 1760
+    },
+    {
+      "epoch": 0.008839172014282503,
+      "grad_norm": 0.09582570195198059,
+      "learning_rate": 2.9809256802423092e-05,
+      "loss": 10.7485,
+      "step": 1770
+    },
+    {
+      "epoch": 0.008889110839221953,
+      "grad_norm": 0.09464222192764282,
+      "learning_rate": 2.980775488748154e-05,
+      "loss": 10.7473,
+      "step": 1780
+    },
+    {
+      "epoch": 0.008939049664161402,
+      "grad_norm": 0.09430133551359177,
+      "learning_rate": 2.980625297253999e-05,
+      "loss": 10.7484,
+      "step": 1790
+    },
+    {
+      "epoch": 0.008988988489100852,
+      "grad_norm": 0.09573784470558167,
+      "learning_rate": 2.980475105759844e-05,
+      "loss": 10.7467,
+      "step": 1800
+    },
+    {
+      "epoch": 0.009038927314040301,
+      "grad_norm": 0.09440425038337708,
+      "learning_rate": 2.9803249142656886e-05,
+      "loss": 10.7473,
+      "step": 1810
+    },
+    {
+      "epoch": 0.00908886613897975,
+      "grad_norm": 0.098334401845932,
+      "learning_rate": 2.980174722771534e-05,
+      "loss": 10.7465,
+      "step": 1820
+    },
+    {
+      "epoch": 0.009138804963919198,
+      "grad_norm": 0.095046766102314,
+      "learning_rate": 2.9800245312773786e-05,
+      "loss": 10.7463,
+      "step": 1830
+    },
+    {
+      "epoch": 0.009188743788858648,
+      "grad_norm": 0.09811218827962875,
+      "learning_rate": 2.9798743397832237e-05,
+      "loss": 10.7457,
+      "step": 1840
+    },
+    {
+      "epoch": 0.009238682613798097,
+      "grad_norm": 0.09796342253684998,
+      "learning_rate": 2.9797241482890687e-05,
+      "loss": 10.7463,
+      "step": 1850
+    },
+    {
+      "epoch": 0.009288621438737547,
+      "grad_norm": 0.09163960069417953,
+      "learning_rate": 2.9795739567949134e-05,
+      "loss": 10.7455,
+      "step": 1860
+    },
+    {
+      "epoch": 0.009338560263676996,
+      "grad_norm": 0.09211383759975433,
+      "learning_rate": 2.9794237653007587e-05,
+      "loss": 10.7452,
+      "step": 1870
+    },
+    {
+      "epoch": 0.009388499088616446,
+      "grad_norm": 0.1004691869020462,
+      "learning_rate": 2.9792735738066034e-05,
+      "loss": 10.7444,
+      "step": 1880
+    },
+    {
+      "epoch": 0.009438437913555893,
+      "grad_norm": 0.09318966418504715,
+      "learning_rate": 2.9791233823124484e-05,
+      "loss": 10.744,
+      "step": 1890
+    },
+    {
+      "epoch": 0.009488376738495343,
+      "grad_norm": 0.09770355373620987,
+      "learning_rate": 2.9789731908182934e-05,
+      "loss": 10.7435,
+      "step": 1900
+    },
+    {
+      "epoch": 0.009538315563434792,
+      "grad_norm": 0.09747663885354996,
+      "learning_rate": 2.978822999324138e-05,
+      "loss": 10.7435,
+      "step": 1910
+    },
+    {
+      "epoch": 0.009588254388374242,
+      "grad_norm": 0.0939902812242508,
+      "learning_rate": 2.9786728078299835e-05,
+      "loss": 10.7433,
+      "step": 1920
+    },
+    {
+      "epoch": 0.009638193213313691,
+      "grad_norm": 0.09048355370759964,
+      "learning_rate": 2.978522616335828e-05,
+      "loss": 10.7432,
+      "step": 1930
+    },
+    {
+      "epoch": 0.00968813203825314,
+      "grad_norm": 0.09295009821653366,
+      "learning_rate": 2.978372424841673e-05,
+      "loss": 10.743,
+      "step": 1940
+    },
+    {
+      "epoch": 0.009738070863192588,
+      "grad_norm": 0.09584958851337433,
+      "learning_rate": 2.9782222333475182e-05,
+      "loss": 10.7431,
+      "step": 1950
+    },
+    {
+      "epoch": 0.009788009688132038,
+      "grad_norm": 0.09165474027395248,
+      "learning_rate": 2.978072041853363e-05,
+      "loss": 10.7419,
+      "step": 1960
+    },
+    {
+      "epoch": 0.009837948513071487,
+      "grad_norm": 0.0949409231543541,
+      "learning_rate": 2.9779218503592082e-05,
+      "loss": 10.7418,
+      "step": 1970
+    },
+    {
+      "epoch": 0.009887887338010937,
+      "grad_norm": 0.09680484980344772,
+      "learning_rate": 2.977771658865053e-05,
+      "loss": 10.7418,
+      "step": 1980
+    },
+    {
+      "epoch": 0.009937826162950386,
+      "grad_norm": 0.10440608114004135,
+      "learning_rate": 2.9776214673708982e-05,
+      "loss": 10.7418,
+      "step": 1990
+    },
+    {
+      "epoch": 0.009987764987889836,
+      "grad_norm": 0.09684424102306366,
+      "learning_rate": 2.977471275876743e-05,
+      "loss": 10.741,
+      "step": 2000
+    },
+    {
+      "epoch": 0.010037703812829283,
+      "grad_norm": 0.09417276084423065,
+      "learning_rate": 2.9773210843825876e-05,
+      "loss": 10.7413,
+      "step": 2010
+    },
+    {
+      "epoch": 0.010087642637768733,
+      "grad_norm": 0.09294068813323975,
+      "learning_rate": 2.977170892888433e-05,
+      "loss": 10.74,
+      "step": 2020
+    },
+    {
+      "epoch": 0.010137581462708182,
+      "grad_norm": 0.09408864378929138,
+      "learning_rate": 2.9770207013942776e-05,
+      "loss": 10.7388,
+      "step": 2030
+    },
+    {
+      "epoch": 0.010187520287647632,
+      "grad_norm": 0.09455525130033493,
+      "learning_rate": 2.976870509900123e-05,
+      "loss": 10.7401,
+      "step": 2040
+    },
+    {
+      "epoch": 0.010237459112587081,
+      "grad_norm": 0.09249377250671387,
+      "learning_rate": 2.9767203184059677e-05,
+      "loss": 10.7395,
+      "step": 2050
+    },
+    {
+      "epoch": 0.01028739793752653,
+      "grad_norm": 0.09711704403162003,
+      "learning_rate": 2.9765701269118124e-05,
+      "loss": 10.7387,
+      "step": 2060
+    },
+    {
+      "epoch": 0.01033733676246598,
+      "grad_norm": 0.09462075680494308,
+      "learning_rate": 2.9764199354176577e-05,
+      "loss": 10.7389,
+      "step": 2070
+    },
+    {
+      "epoch": 0.010387275587405428,
+      "grad_norm": 0.09138345718383789,
+      "learning_rate": 2.9762697439235024e-05,
+      "loss": 10.7381,
+      "step": 2080
+    },
+    {
+      "epoch": 0.010437214412344877,
+      "grad_norm": 0.0979338064789772,
+      "learning_rate": 2.9761195524293477e-05,
+      "loss": 10.7373,
+      "step": 2090
+    },
+    {
+      "epoch": 0.010487153237284327,
+      "grad_norm": 428.1760559082031,
+      "learning_rate": 2.9759693609351924e-05,
+      "loss": 10.7368,
+      "step": 2100
+    },
+    {
+      "epoch": 0.010537092062223776,
+      "grad_norm": 0.09210320562124252,
+      "learning_rate": 2.975819169441037e-05,
+      "loss": 11.136,
+      "step": 2110
+    },
+    {
+      "epoch": 0.010587030887163226,
+      "grad_norm": 0.09596727043390274,
+      "learning_rate": 2.9756689779468825e-05,
+      "loss": 12.8231,
+      "step": 2120
+    },
+    {
+      "epoch": 0.010636969712102675,
+      "grad_norm": 0.10231532901525497,
+      "learning_rate": 2.975518786452727e-05,
+      "loss": 10.8774,
+      "step": 2130
+    },
+    {
+      "epoch": 0.010686908537042123,
+      "grad_norm": 0.09213022887706757,
+      "learning_rate": 2.9753685949585725e-05,
+      "loss": 10.7361,
+      "step": 2140
+    },
+    {
+      "epoch": 0.010736847361981572,
+      "grad_norm": 0.0937543660402298,
+      "learning_rate": 2.9752184034644172e-05,
+      "loss": 10.7364,
+      "step": 2150
+    },
+    {
+      "epoch": 0.010786786186921022,
+      "grad_norm": 0.09011025726795197,
+      "learning_rate": 2.975068211970262e-05,
+      "loss": 10.7361,
+      "step": 2160
+    },
+    {
+      "epoch": 0.010836725011860471,
+      "grad_norm": 0.09514773637056351,
+      "learning_rate": 2.9749180204761072e-05,
+      "loss": 10.7348,
+      "step": 2170
+    },
+    {
+      "epoch": 0.01088666383679992,
+      "grad_norm": 0.0912405401468277,
+      "learning_rate": 2.974767828981952e-05,
+      "loss": 10.7349,
+      "step": 2180
+    },
+    {
+      "epoch": 0.01093660266173937,
+      "grad_norm": 0.10206656157970428,
+      "learning_rate": 2.9746176374877972e-05,
+      "loss": 10.7357,
+      "step": 2190
+    },
+    {
+      "epoch": 0.010986541486678818,
+      "grad_norm": 0.10607360303401947,
+      "learning_rate": 2.974467445993642e-05,
+      "loss": 10.7347,
+      "step": 2200
+    },
+    {
+      "epoch": 0.011036480311618267,
+      "grad_norm": 0.09714030474424362,
+      "learning_rate": 2.9743172544994866e-05,
+      "loss": 10.7333,
+      "step": 2210
+    },
+    {
+      "epoch": 0.011086419136557717,
+      "grad_norm": 0.09556932747364044,
+      "learning_rate": 2.974167063005332e-05,
+      "loss": 10.7338,
+      "step": 2220
+    },
+    {
+      "epoch": 0.011136357961497166,
+      "grad_norm": 0.09247563034296036,
+      "learning_rate": 2.9740168715111766e-05,
+      "loss": 10.7334,
+      "step": 2230
+    },
+    {
+      "epoch": 0.011186296786436616,
+      "grad_norm": 0.08850020170211792,
+      "learning_rate": 2.973866680017022e-05,
+      "loss": 10.7334,
+      "step": 2240
+    },
+    {
+      "epoch": 0.011236235611376065,
+      "grad_norm": 0.09381990134716034,
+      "learning_rate": 2.9737164885228667e-05,
+      "loss": 10.7324,
+      "step": 2250
+    },
+    {
+      "epoch": 0.011286174436315513,
+      "grad_norm": 0.09170781821012497,
+      "learning_rate": 2.9735662970287114e-05,
+      "loss": 10.7327,
+      "step": 2260
+    },
+    {
+      "epoch": 0.011336113261254962,
+      "grad_norm": 0.09936770796775818,
+      "learning_rate": 2.9734161055345567e-05,
+      "loss": 10.7321,
+      "step": 2270
+    },
+    {
+      "epoch": 0.011386052086194412,
+      "grad_norm": 0.09585506469011307,
+      "learning_rate": 2.9732659140404014e-05,
+      "loss": 10.7322,
+      "step": 2280
+    },
+    {
+      "epoch": 0.011435990911133861,
+      "grad_norm": 0.0958356037735939,
+      "learning_rate": 2.9731157225462467e-05,
+      "loss": 10.7323,
+      "step": 2290
+    },
+    {
+      "epoch": 0.01148592973607331,
+      "grad_norm": 0.09194158017635345,
+      "learning_rate": 2.9729655310520914e-05,
+      "loss": 10.7318,
+      "step": 2300
+    },
+    {
+      "epoch": 0.01153586856101276,
+      "grad_norm": 0.09054262191057205,
+      "learning_rate": 2.9728153395579364e-05,
+      "loss": 10.7313,
+      "step": 2310
+    },
+    {
+      "epoch": 0.011585807385952208,
+      "grad_norm": 0.09786096960306168,
+      "learning_rate": 2.9726651480637815e-05,
+      "loss": 10.7306,
+      "step": 2320
+    },
+    {
+      "epoch": 0.011635746210891657,
+      "grad_norm": 0.09685620665550232,
+      "learning_rate": 2.972514956569626e-05,
+      "loss": 10.731,
+      "step": 2330
+    },
+    {
+      "epoch": 0.011685685035831107,
+      "grad_norm": 0.09197534620761871,
+      "learning_rate": 2.9723647650754715e-05,
+      "loss": 10.7293,
+      "step": 2340
+    },
+    {
+      "epoch": 0.011735623860770556,
+      "grad_norm": 0.09799861162900925,
+      "learning_rate": 2.9722145735813162e-05,
+      "loss": 10.7295,
+      "step": 2350
+    },
+    {
+      "epoch": 0.011785562685710006,
+      "grad_norm": 0.09268220514059067,
+      "learning_rate": 2.9720643820871612e-05,
+      "loss": 10.7298,
+      "step": 2360
+    },
+    {
+      "epoch": 0.011835501510649455,
+      "grad_norm": 0.09487041085958481,
+      "learning_rate": 2.9719141905930062e-05,
+      "loss": 10.7286,
+      "step": 2370
+    },
+    {
+      "epoch": 0.011885440335588903,
+      "grad_norm": 0.09200552850961685,
+      "learning_rate": 2.971763999098851e-05,
+      "loss": 10.7287,
+      "step": 2380
+    },
+    {
+      "epoch": 0.011935379160528352,
+      "grad_norm": 0.09358713030815125,
+      "learning_rate": 2.9716138076046962e-05,
+      "loss": 10.7276,
+      "step": 2390
+    },
+    {
+      "epoch": 0.011985317985467802,
+      "grad_norm": 0.09549114853143692,
+      "learning_rate": 2.971463616110541e-05,
+      "loss": 10.7284,
+      "step": 2400
+    },
+    {
+      "epoch": 0.012035256810407251,
+      "grad_norm": 0.09079451113939285,
+      "learning_rate": 2.971313424616386e-05,
+      "loss": 10.7274,
+      "step": 2410
+    },
+    {
+      "epoch": 0.0120851956353467,
+      "grad_norm": 0.09617462754249573,
+      "learning_rate": 2.971163233122231e-05,
+      "loss": 10.7274,
+      "step": 2420
+    },
+    {
+      "epoch": 0.01213513446028615,
+      "grad_norm": 0.09577368944883347,
+      "learning_rate": 2.9710130416280756e-05,
+      "loss": 10.7275,
+      "step": 2430
+    },
+    {
+      "epoch": 0.0121850732852256,
+      "grad_norm": 0.09303394705057144,
+      "learning_rate": 2.970862850133921e-05,
+      "loss": 10.7257,
+      "step": 2440
+    },
+    {
+      "epoch": 0.012235012110165047,
+      "grad_norm": 0.09505324810743332,
+      "learning_rate": 2.9707126586397657e-05,
+      "loss": 10.726,
+      "step": 2450
+    },
+    {
+      "epoch": 0.012284950935104497,
+      "grad_norm": 0.0933798998594284,
+      "learning_rate": 2.9705624671456107e-05,
+      "loss": 10.7261,
+      "step": 2460
+    },
+    {
+      "epoch": 0.012334889760043946,
+      "grad_norm": 0.09893235564231873,
+      "learning_rate": 2.9704122756514557e-05,
+      "loss": 10.7264,
+      "step": 2470
+    },
+    {
+      "epoch": 0.012384828584983396,
+      "grad_norm": 0.08968345075845718,
+      "learning_rate": 2.9702620841573004e-05,
+      "loss": 10.7259,
+      "step": 2480
+    },
+    {
+      "epoch": 0.012434767409922845,
+      "grad_norm": 0.09330223500728607,
+      "learning_rate": 2.9701118926631458e-05,
+      "loss": 10.7266,
+      "step": 2490
+    },
+    {
+      "epoch": 0.012484706234862294,
+      "grad_norm": 0.10579249262809753,
+      "learning_rate": 2.9699617011689904e-05,
+      "loss": 10.7249,
+      "step": 2500
+    },
+    {
+      "epoch": 0.012534645059801742,
+      "grad_norm": 0.0928342342376709,
+      "learning_rate": 2.9698115096748354e-05,
+      "loss": 10.7238,
+      "step": 2510
+    },
+    {
+      "epoch": 0.012584583884741192,
+      "grad_norm": 0.09858880937099457,
+      "learning_rate": 2.9696613181806805e-05,
+      "loss": 10.7247,
+      "step": 2520
+    },
+    {
+      "epoch": 0.012634522709680641,
+      "grad_norm": 0.09675079584121704,
+      "learning_rate": 2.969511126686525e-05,
+      "loss": 10.7236,
+      "step": 2530
+    },
+    {
+      "epoch": 0.01268446153462009,
+      "grad_norm": 0.09549619257450104,
+      "learning_rate": 2.9693609351923705e-05,
+      "loss": 10.7238,
+      "step": 2540
+    },
+    {
+      "epoch": 0.01273440035955954,
+      "grad_norm": 0.09920501708984375,
+      "learning_rate": 2.9692107436982152e-05,
+      "loss": 10.7236,
+      "step": 2550
+    },
+    {
+      "epoch": 0.01278433918449899,
+      "grad_norm": 0.09226509928703308,
+      "learning_rate": 2.9690605522040602e-05,
+      "loss": 10.7233,
+      "step": 2560
+    },
+    {
+      "epoch": 0.012834278009438437,
+      "grad_norm": 0.09456513077020645,
+      "learning_rate": 2.9689103607099052e-05,
+      "loss": 10.7237,
+      "step": 2570
+    },
+    {
+      "epoch": 0.012884216834377887,
+      "grad_norm": 0.09502000361680984,
+      "learning_rate": 2.96876016921575e-05,
+      "loss": 10.7231,
+      "step": 2580
+    },
+    {
+      "epoch": 0.012934155659317336,
+      "grad_norm": 0.09463300555944443,
+      "learning_rate": 2.9686099777215953e-05,
+      "loss": 10.722,
+      "step": 2590
+    },
+    {
+      "epoch": 0.012984094484256786,
+      "grad_norm": 0.09156177937984467,
+      "learning_rate": 2.96845978622744e-05,
+      "loss": 10.722,
+      "step": 2600
+    },
+    {
+      "epoch": 0.013034033309196235,
+      "grad_norm": 0.09233105927705765,
+      "learning_rate": 2.968309594733285e-05,
+      "loss": 10.7217,
+      "step": 2610
+    },
+    {
+      "epoch": 0.013083972134135684,
+      "grad_norm": 0.0961083248257637,
+      "learning_rate": 2.96815940323913e-05,
+      "loss": 10.7208,
+      "step": 2620
+    },
+    {
+      "epoch": 0.013133910959075132,
+      "grad_norm": 0.0997881218791008,
+      "learning_rate": 2.968009211744975e-05,
+      "loss": 10.7205,
+      "step": 2630
+    },
+    {
+      "epoch": 0.013183849784014582,
+      "grad_norm": 0.09116201847791672,
+      "learning_rate": 2.96785902025082e-05,
+      "loss": 10.72,
+      "step": 2640
+    },
+    {
+      "epoch": 0.013233788608954031,
+      "grad_norm": 0.0893169716000557,
+      "learning_rate": 2.9677088287566647e-05,
+      "loss": 10.7206,
+      "step": 2650
+    },
+    {
+      "epoch": 0.01328372743389348,
+      "grad_norm": 0.09925134479999542,
+      "learning_rate": 2.9675586372625097e-05,
+      "loss": 31.1372,
+      "step": 2660
+    },
+    {
+      "epoch": 0.01333366625883293,
+      "grad_norm": 0.09098075330257416,
+      "learning_rate": 2.9674084457683547e-05,
+      "loss": 10.7191,
+      "step": 2670
+    },
+    {
+      "epoch": 0.01338360508377238,
+      "grad_norm": 0.09981559216976166,
+      "learning_rate": 2.9672582542741997e-05,
+      "loss": 18.3812,
+      "step": 2680
+    },
+    {
+      "epoch": 0.013433543908711827,
+      "grad_norm": 0.10291223227977753,
+      "learning_rate": 2.9671080627800448e-05,
+      "loss": 10.7193,
+      "step": 2690
+    },
+    {
+      "epoch": 0.013483482733651277,
+      "grad_norm": 0.09718188643455505,
+      "learning_rate": 2.9669578712858894e-05,
+      "loss": 10.789,
+      "step": 2700
+    },
+    {
+      "epoch": 0.013533421558590726,
+      "grad_norm": 0.1061011552810669,
+      "learning_rate": 2.9668076797917344e-05,
+      "loss": 10.7188,
+      "step": 2710
+    },
+    {
+      "epoch": 0.013583360383530176,
+      "grad_norm": 65854.7578125,
+      "learning_rate": 2.9666574882975795e-05,
+      "loss": 11.0129,
+      "step": 2720
+    },
+    {
+      "epoch": 0.013633299208469625,
+      "grad_norm": 743.2807006835938,
+      "learning_rate": 2.9665072968034245e-05,
+      "loss": 10.7305,
+      "step": 2730
+    },
+    {
+      "epoch": 0.013683238033409074,
+      "grad_norm": 0.09718185663223267,
+      "learning_rate": 2.9663571053092695e-05,
+      "loss": 11.708,
+      "step": 2740
+    },
+    {
+      "epoch": 0.013733176858348524,
+      "grad_norm": 0.09672287851572037,
+      "learning_rate": 2.9662069138151142e-05,
+      "loss": 10.7177,
+      "step": 2750
+    },
+    {
+      "epoch": 0.013783115683287972,
+      "grad_norm": 0.093637615442276,
+      "learning_rate": 2.9660567223209595e-05,
+      "loss": 10.8841,
+      "step": 2760
+    },
+    {
+      "epoch": 0.013833054508227421,
+      "grad_norm": 0.10001079738140106,
+      "learning_rate": 2.9659065308268042e-05,
+      "loss": 10.8078,
+      "step": 2770
+    },
+    {
+      "epoch": 0.01388299333316687,
+      "grad_norm": 0.09189160168170929,
+      "learning_rate": 2.9657563393326492e-05,
+      "loss": 11.6597,
+      "step": 2780
+    },
+    {
+      "epoch": 0.01393293215810632,
+      "grad_norm": 9842.095703125,
+      "learning_rate": 2.9656061478384943e-05,
+      "loss": 13.5412,
+      "step": 2790
+    },
+    {
+      "epoch": 0.01398287098304577,
+      "grad_norm": 0.09831225126981735,
+      "learning_rate": 2.965455956344339e-05,
+      "loss": 10.7168,
+      "step": 2800
+    },
+    {
+      "epoch": 0.014032809807985219,
+      "grad_norm": 0.09885016083717346,
+      "learning_rate": 2.9653057648501843e-05,
+      "loss": 10.7171,
+      "step": 2810
+    },
+    {
+      "epoch": 0.014082748632924667,
+      "grad_norm": 0.09697610884904861,
+      "learning_rate": 2.965155573356029e-05,
+      "loss": 10.7158,
+      "step": 2820
+    },
+    {
+      "epoch": 0.014132687457864116,
+      "grad_norm": 0.09410693496465683,
+      "learning_rate": 2.965005381861874e-05,
+      "loss": 10.7153,
+      "step": 2830
+    },
+    {
+      "epoch": 0.014182626282803566,
+      "grad_norm": 0.09180265665054321,
+      "learning_rate": 2.964855190367719e-05,
+      "loss": 10.7161,
+      "step": 2840
+    },
+    {
+      "epoch": 0.014232565107743015,
+      "grad_norm": 0.10008688271045685,
+      "learning_rate": 2.9647049988735637e-05,
+      "loss": 10.7156,
+      "step": 2850
+    },
+    {
+      "epoch": 0.014282503932682464,
+      "grad_norm": 0.09007236361503601,
+      "learning_rate": 2.964554807379409e-05,
+      "loss": 10.7152,
+      "step": 2860
+    },
+    {
+      "epoch": 0.014332442757621914,
+      "grad_norm": 0.09538665413856506,
+      "learning_rate": 2.9644046158852537e-05,
+      "loss": 10.7139,
+      "step": 2870
+    },
+    {
+      "epoch": 0.014382381582561362,
+      "grad_norm": 0.09514635801315308,
+      "learning_rate": 2.9642544243910987e-05,
+      "loss": 10.7136,
+      "step": 2880
+    },
+    {
+      "epoch": 0.014432320407500811,
+      "grad_norm": 0.09610595554113388,
+      "learning_rate": 2.9641042328969438e-05,
+      "loss": 10.7138,
+      "step": 2890
+    },
+    {
+      "epoch": 0.01448225923244026,
+      "grad_norm": 0.09452015161514282,
+      "learning_rate": 2.9639540414027884e-05,
+      "loss": 10.7137,
+      "step": 2900
+    },
+    {
+      "epoch": 0.01453219805737971,
+      "grad_norm": 0.09626973420381546,
+      "learning_rate": 2.9638038499086338e-05,
+      "loss": 10.7121,
+      "step": 2910
+    },
+    {
+      "epoch": 0.01458213688231916,
+      "grad_norm": 0.0968800038099289,
+      "learning_rate": 2.9636536584144785e-05,
+      "loss": 10.7127,
+      "step": 2920
+    },
+    {
+      "epoch": 0.014632075707258609,
+      "grad_norm": 0.09562117606401443,
+      "learning_rate": 2.9635034669203235e-05,
+      "loss": 10.7123,
+      "step": 2930
+    },
+    {
+      "epoch": 0.014682014532198057,
+      "grad_norm": 0.09600751847028732,
+      "learning_rate": 2.9633532754261685e-05,
+      "loss": 10.7121,
+      "step": 2940
+    },
+    {
+      "epoch": 0.014731953357137506,
+      "grad_norm": 0.09042752534151077,
+      "learning_rate": 2.9632030839320135e-05,
+      "loss": 10.7112,
+      "step": 2950
+    },
+    {
+      "epoch": 0.014781892182076956,
+      "grad_norm": 0.09635340422391891,
+      "learning_rate": 2.9630528924378585e-05,
+      "loss": 10.7109,
+      "step": 2960
+    },
+    {
+      "epoch": 0.014831831007016405,
+      "grad_norm": 0.09323673695325851,
+      "learning_rate": 2.9629027009437032e-05,
+      "loss": 10.7115,
+      "step": 2970
+    },
+    {
+      "epoch": 0.014881769831955854,
+      "grad_norm": 0.09715006500482559,
+      "learning_rate": 2.9627525094495482e-05,
+      "loss": 10.7108,
+      "step": 2980
+    },
+    {
+      "epoch": 0.014931708656895304,
+      "grad_norm": 0.09540782868862152,
+      "learning_rate": 2.9626023179553933e-05,
+      "loss": 10.7107,
+      "step": 2990
+    },
+    {
+      "epoch": 0.014981647481834752,
+      "grad_norm": 0.09565411508083344,
+      "learning_rate": 2.9624521264612383e-05,
+      "loss": 10.7098,
+      "step": 3000
+    },
+    {
+      "epoch": 0.015031586306774201,
+      "grad_norm": 0.09608177095651627,
+      "learning_rate": 2.9623019349670833e-05,
+      "loss": 10.7095,
+      "step": 3010
+    },
+    {
+      "epoch": 0.01508152513171365,
+      "grad_norm": 0.09020178765058517,
+      "learning_rate": 2.962151743472928e-05,
+      "loss": 10.7091,
+      "step": 3020
+    },
+    {
+      "epoch": 0.0151314639566531,
+      "grad_norm": 0.09340853989124298,
+      "learning_rate": 2.962001551978773e-05,
+      "loss": 10.7083,
+      "step": 3030
+    },
+    {
+      "epoch": 0.01518140278159255,
+      "grad_norm": 0.09446629136800766,
+      "learning_rate": 2.961851360484618e-05,
+      "loss": 10.7087,
+      "step": 3040
+    },
+    {
+      "epoch": 0.015231341606531999,
+      "grad_norm": 0.0978376492857933,
+      "learning_rate": 2.961701168990463e-05,
+      "loss": 10.7081,
+      "step": 3050
+    },
+    {
+      "epoch": 0.015281280431471447,
+      "grad_norm": 0.09334233403205872,
+      "learning_rate": 2.961550977496308e-05,
+      "loss": 10.709,
+      "step": 3060
+    },
+    {
+      "epoch": 0.015331219256410896,
+      "grad_norm": 0.098800890147686,
+      "learning_rate": 2.9614007860021527e-05,
+      "loss": 10.7079,
+      "step": 3070
+    },
+    {
+      "epoch": 0.015381158081350346,
+      "grad_norm": 0.09233418852090836,
+      "learning_rate": 2.9612505945079977e-05,
+      "loss": 10.7071,
+      "step": 3080
+    },
+    {
+      "epoch": 0.015431096906289795,
+      "grad_norm": 0.0931033119559288,
+      "learning_rate": 2.9611004030138428e-05,
+      "loss": 10.7077,
+      "step": 3090
+    },
+    {
+      "epoch": 0.015481035731229244,
+      "grad_norm": 0.08983912318944931,
+      "learning_rate": 2.9609502115196878e-05,
+      "loss": 10.7069,
+      "step": 3100
+    },
+    {
+      "epoch": 0.015530974556168694,
+      "grad_norm": 0.09569084644317627,
+      "learning_rate": 2.9608000200255328e-05,
+      "loss": 10.7061,
+      "step": 3110
+    },
+    {
+      "epoch": 0.015580913381108143,
+      "grad_norm": 0.09349966049194336,
+      "learning_rate": 2.9606498285313775e-05,
+      "loss": 10.7061,
+      "step": 3120
+    },
+    {
+      "epoch": 0.015630852206047593,
+      "grad_norm": 0.09354298561811447,
+      "learning_rate": 2.9604996370372225e-05,
+      "loss": 10.7061,
+      "step": 3130
+    },
+    {
+      "epoch": 0.015680791030987042,
+      "grad_norm": 0.09295152127742767,
+      "learning_rate": 2.9603494455430675e-05,
+      "loss": 10.7059,
+      "step": 3140
+    },
+    {
+      "epoch": 0.01573072985592649,
+      "grad_norm": 0.09449790418148041,
+      "learning_rate": 2.9601992540489125e-05,
+      "loss": 10.7059,
+      "step": 3150
+    },
+    {
+      "epoch": 0.015780668680865938,
+      "grad_norm": 0.0972711518406868,
+      "learning_rate": 2.9600490625547575e-05,
+      "loss": 10.7049,
+      "step": 3160
+    },
+    {
+      "epoch": 0.015830607505805387,
+      "grad_norm": 0.09353794902563095,
+      "learning_rate": 2.9598988710606022e-05,
+      "loss": 10.7047,
+      "step": 3170
+    },
+    {
+      "epoch": 0.015880546330744837,
+      "grad_norm": 0.09542354941368103,
+      "learning_rate": 2.9597486795664472e-05,
+      "loss": 10.7048,
+      "step": 3180
+    },
+    {
+      "epoch": 0.015930485155684286,
+      "grad_norm": 0.09505563974380493,
+      "learning_rate": 2.9595984880722923e-05,
+      "loss": 10.7042,
+      "step": 3190
+    },
+    {
+      "epoch": 0.015980423980623736,
+      "grad_norm": 0.09707723557949066,
+      "learning_rate": 2.9594482965781373e-05,
+      "loss": 10.7039,
+      "step": 3200
+    },
+    {
+      "epoch": 0.016030362805563185,
+      "grad_norm": 0.09463075548410416,
+      "learning_rate": 2.9592981050839823e-05,
+      "loss": 10.7039,
+      "step": 3210
+    },
+    {
+      "epoch": 0.016080301630502634,
+      "grad_norm": 0.09552239626646042,
+      "learning_rate": 2.959147913589827e-05,
+      "loss": 10.7023,
+      "step": 3220
+    },
+    {
+      "epoch": 0.016130240455442084,
+      "grad_norm": 0.0979151502251625,
+      "learning_rate": 2.958997722095672e-05,
+      "loss": 10.7026,
+      "step": 3230
+    },
+    {
+      "epoch": 0.016180179280381533,
+      "grad_norm": 0.0954020619392395,
+      "learning_rate": 2.958847530601517e-05,
+      "loss": 10.702,
+      "step": 3240
+    },
+    {
+      "epoch": 0.016230118105320983,
+      "grad_norm": 0.09606318175792694,
+      "learning_rate": 2.958697339107362e-05,
+      "loss": 10.702,
+      "step": 3250
+    },
+    {
+      "epoch": 0.016280056930260432,
+      "grad_norm": 0.09881041944026947,
+      "learning_rate": 2.958547147613207e-05,
+      "loss": 10.7023,
+      "step": 3260
+    },
+    {
+      "epoch": 0.01632999575519988,
+      "grad_norm": 0.09216438978910446,
+      "learning_rate": 2.958396956119052e-05,
+      "loss": 10.7006,
+      "step": 3270
+    },
+    {
+      "epoch": 0.016379934580139328,
+      "grad_norm": 0.09119243174791336,
+      "learning_rate": 2.9582467646248967e-05,
+      "loss": 10.7009,
+      "step": 3280
+    },
+    {
+      "epoch": 0.016429873405078777,
+      "grad_norm": 0.09380321204662323,
+      "learning_rate": 2.9580965731307418e-05,
+      "loss": 10.7007,
+      "step": 3290
+    },
+    {
+      "epoch": 0.016479812230018227,
+      "grad_norm": 0.0956210270524025,
+      "learning_rate": 2.9579463816365868e-05,
+      "loss": 10.7,
+      "step": 3300
+    },
+    {
+      "epoch": 0.016529751054957676,
+      "grad_norm": 0.0973009392619133,
+      "learning_rate": 2.9577961901424318e-05,
+      "loss": 10.6998,
+      "step": 3310
+    },
+    {
+      "epoch": 0.016579689879897126,
+      "grad_norm": 0.09392161667346954,
+      "learning_rate": 2.9576459986482768e-05,
+      "loss": 10.7002,
+      "step": 3320
+    },
+    {
+      "epoch": 0.016629628704836575,
+      "grad_norm": 0.09009403735399246,
+      "learning_rate": 2.9574958071541215e-05,
+      "loss": 10.7001,
+      "step": 3330
+    },
+    {
+      "epoch": 0.016679567529776024,
+      "grad_norm": 0.09653884917497635,
+      "learning_rate": 2.9573456156599665e-05,
+      "loss": 10.6986,
+      "step": 3340
+    },
+    {
+      "epoch": 0.016729506354715474,
+      "grad_norm": 0.09588049352169037,
+      "learning_rate": 2.9571954241658115e-05,
+      "loss": 10.6994,
+      "step": 3350
+    },
+    {
+      "epoch": 0.016779445179654923,
+      "grad_norm": 0.09776197373867035,
+      "learning_rate": 2.9570452326716565e-05,
+      "loss": 10.6991,
+      "step": 3360
+    },
+    {
+      "epoch": 0.016829384004594373,
+      "grad_norm": 0.09142053127288818,
+      "learning_rate": 2.9568950411775016e-05,
+      "loss": 10.6981,
+      "step": 3370
+    },
+    {
+      "epoch": 0.016879322829533822,
+      "grad_norm": 0.09754452109336853,
+      "learning_rate": 2.9567448496833462e-05,
+      "loss": 10.6978,
+      "step": 3380
+    },
+    {
+      "epoch": 0.01692926165447327,
+      "grad_norm": 0.09043591469526291,
+      "learning_rate": 2.9565946581891913e-05,
+      "loss": 10.6968,
+      "step": 3390
+    },
+    {
+      "epoch": 0.016979200479412718,
+      "grad_norm": 0.09461095184087753,
+      "learning_rate": 2.9564444666950363e-05,
+      "loss": 10.6971,
+      "step": 3400
+    },
+    {
+      "epoch": 0.017029139304352167,
+      "grad_norm": 0.10008252412080765,
+      "learning_rate": 2.9562942752008813e-05,
+      "loss": 10.6968,
+      "step": 3410
+    },
+    {
+      "epoch": 0.017079078129291617,
+      "grad_norm": 0.09465460479259491,
+      "learning_rate": 2.9561440837067263e-05,
+      "loss": 10.6963,
+      "step": 3420
+    },
+    {
+      "epoch": 0.017129016954231066,
+      "grad_norm": 0.09323673695325851,
+      "learning_rate": 2.955993892212571e-05,
+      "loss": 10.6959,
+      "step": 3430
+    },
+    {
+      "epoch": 0.017178955779170516,
+      "grad_norm": 0.09518439322710037,
+      "learning_rate": 2.955843700718416e-05,
+      "loss": 10.6961,
+      "step": 3440
+    },
+    {
+      "epoch": 0.017228894604109965,
+      "grad_norm": 0.09833721816539764,
+      "learning_rate": 2.955693509224261e-05,
+      "loss": 10.6954,
+      "step": 3450
+    },
+    {
+      "epoch": 0.017278833429049414,
+      "grad_norm": 0.09518902748823166,
+      "learning_rate": 2.955543317730106e-05,
+      "loss": 10.6944,
+      "step": 3460
+    },
+    {
+      "epoch": 0.017328772253988864,
+      "grad_norm": 0.10006536543369293,
+      "learning_rate": 2.955393126235951e-05,
+      "loss": 10.6956,
+      "step": 3470
+    },
+    {
+      "epoch": 0.017378711078928313,
+      "grad_norm": 0.08888301998376846,
+      "learning_rate": 2.9552429347417957e-05,
+      "loss": 10.6955,
+      "step": 3480
+    },
+    {
+      "epoch": 0.017428649903867763,
+      "grad_norm": 0.09507909417152405,
+      "learning_rate": 2.9550927432476408e-05,
+      "loss": 10.695,
+      "step": 3490
+    },
+    {
+      "epoch": 0.017478588728807212,
+      "grad_norm": 0.09857828170061111,
+      "learning_rate": 2.9549425517534858e-05,
+      "loss": 10.6974,
+      "step": 3500
+    },
+    {
+      "epoch": 0.01752852755374666,
+      "grad_norm": 0.09427110850811005,
+      "learning_rate": 2.9547923602593308e-05,
+      "loss": 10.6932,
+      "step": 3510
+    },
+    {
+      "epoch": 0.01757846637868611,
+      "grad_norm": 0.08865749090909958,
+      "learning_rate": 2.9546421687651758e-05,
+      "loss": 10.6936,
+      "step": 3520
+    },
+    {
+      "epoch": 0.017628405203625557,
+      "grad_norm": 0.09774231165647507,
+      "learning_rate": 2.9544919772710205e-05,
+      "loss": 10.6936,
+      "step": 3530
+    },
+    {
+      "epoch": 0.017678344028565007,
+      "grad_norm": 0.10259677469730377,
+      "learning_rate": 2.9543417857768655e-05,
+      "loss": 10.6929,
+      "step": 3540
+    },
+    {
+      "epoch": 0.017728282853504456,
+      "grad_norm": 0.09243781864643097,
+      "learning_rate": 2.9541915942827105e-05,
+      "loss": 10.6924,
+      "step": 3550
+    },
+    {
+      "epoch": 0.017778221678443906,
+      "grad_norm": 0.08847206085920334,
+      "learning_rate": 2.9540414027885555e-05,
+      "loss": 10.6917,
+      "step": 3560
+    },
+    {
+      "epoch": 0.017828160503383355,
+      "grad_norm": 0.09328203648328781,
+      "learning_rate": 2.9538912112944006e-05,
+      "loss": 10.692,
+      "step": 3570
+    },
+    {
+      "epoch": 0.017878099328322804,
+      "grad_norm": 0.09655408561229706,
+      "learning_rate": 2.9537410198002452e-05,
+      "loss": 10.6916,
+      "step": 3580
+    },
+    {
+      "epoch": 0.017928038153262254,
+      "grad_norm": 0.09533986449241638,
+      "learning_rate": 2.9535908283060903e-05,
+      "loss": 10.6912,
+      "step": 3590
+    },
+    {
+      "epoch": 0.017977976978201703,
+      "grad_norm": 0.09806837886571884,
+      "learning_rate": 2.9534406368119353e-05,
+      "loss": 10.6905,
+      "step": 3600
+    },
+    {
+      "epoch": 0.018027915803141153,
+      "grad_norm": 0.09532346576452255,
+      "learning_rate": 2.9532904453177803e-05,
+      "loss": 10.6909,
+      "step": 3610
+    },
+    {
+      "epoch": 0.018077854628080602,
+      "grad_norm": 0.09362402558326721,
+      "learning_rate": 2.9531402538236253e-05,
+      "loss": 10.6905,
+      "step": 3620
+    },
+    {
+      "epoch": 0.01812779345302005,
+      "grad_norm": 0.09821821749210358,
+      "learning_rate": 2.95299006232947e-05,
+      "loss": 10.6909,
+      "step": 3630
+    },
+    {
+      "epoch": 0.0181777322779595,
+      "grad_norm": 0.09857241809368134,
+      "learning_rate": 2.9528398708353153e-05,
+      "loss": 10.6896,
+      "step": 3640
+    },
+    {
+      "epoch": 0.018227671102898947,
+      "grad_norm": 0.0939745232462883,
+      "learning_rate": 2.95268967934116e-05,
+      "loss": 10.6893,
+      "step": 3650
+    },
+    {
+      "epoch": 0.018277609927838397,
+      "grad_norm": 0.09561046957969666,
+      "learning_rate": 2.952539487847005e-05,
+      "loss": 10.6889,
+      "step": 3660
+    },
+    {
+      "epoch": 0.018327548752777846,
+      "grad_norm": 0.09186465293169022,
+      "learning_rate": 2.95238929635285e-05,
+      "loss": 10.6892,
+      "step": 3670
+    },
+    {
+      "epoch": 0.018377487577717296,
+      "grad_norm": 0.1008329913020134,
+      "learning_rate": 2.9522391048586947e-05,
+      "loss": 10.6882,
+      "step": 3680
+    },
+    {
+      "epoch": 0.018427426402656745,
+      "grad_norm": 0.09180609881877899,
+      "learning_rate": 2.95208891336454e-05,
+      "loss": 10.6886,
+      "step": 3690
+    },
+    {
+      "epoch": 0.018477365227596194,
+      "grad_norm": 0.09365356713533401,
+      "learning_rate": 2.9519387218703848e-05,
+      "loss": 10.6877,
+      "step": 3700
+    },
+    {
+      "epoch": 0.018527304052535644,
+      "grad_norm": 0.09670042246580124,
+      "learning_rate": 2.9517885303762298e-05,
+      "loss": 10.6865,
+      "step": 3710
+    },
+    {
+      "epoch": 0.018577242877475093,
+      "grad_norm": 0.1001424565911293,
+      "learning_rate": 2.9516383388820748e-05,
+      "loss": 10.6872,
+      "step": 3720
+    },
+    {
+      "epoch": 0.018627181702414543,
+      "grad_norm": 0.0908576175570488,
+      "learning_rate": 2.9514881473879195e-05,
+      "loss": 10.6879,
+      "step": 3730
+    },
+    {
+      "epoch": 0.018677120527353992,
+      "grad_norm": 0.0932522565126419,
+      "learning_rate": 2.951337955893765e-05,
+      "loss": 10.6877,
+      "step": 3740
+    },
+    {
+      "epoch": 0.01872705935229344,
+      "grad_norm": 0.0923110842704773,
+      "learning_rate": 2.9511877643996095e-05,
+      "loss": 10.6855,
+      "step": 3750
+    },
+    {
+      "epoch": 0.01877699817723289,
+      "grad_norm": 0.09654710441827774,
+      "learning_rate": 2.9510375729054545e-05,
+      "loss": 10.6861,
+      "step": 3760
+    },
+    {
+      "epoch": 0.01882693700217234,
+      "grad_norm": 0.09733393788337708,
+      "learning_rate": 2.9508873814112996e-05,
+      "loss": 10.6865,
+      "step": 3770
+    },
+    {
+      "epoch": 0.018876875827111787,
+      "grad_norm": 0.09079433232545853,
+      "learning_rate": 2.9507371899171442e-05,
+      "loss": 10.6857,
+      "step": 3780
+    },
+    {
+      "epoch": 0.018926814652051236,
+      "grad_norm": 0.09449440240859985,
+      "learning_rate": 2.9505869984229896e-05,
+      "loss": 10.6846,
+      "step": 3790
+    },
+    {
+      "epoch": 0.018976753476990686,
+      "grad_norm": 0.0952952429652214,
+      "learning_rate": 2.9504368069288343e-05,
+      "loss": 10.6846,
+      "step": 3800
+    },
+    {
+      "epoch": 0.019026692301930135,
+      "grad_norm": 0.09154447913169861,
+      "learning_rate": 2.9502866154346793e-05,
+      "loss": 10.6848,
+      "step": 3810
+    },
+    {
+      "epoch": 0.019076631126869584,
+      "grad_norm": 0.09125672280788422,
+      "learning_rate": 2.9501364239405243e-05,
+      "loss": 10.6841,
+      "step": 3820
+    },
+    {
+      "epoch": 0.019126569951809034,
+      "grad_norm": 0.09873983263969421,
+      "learning_rate": 2.949986232446369e-05,
+      "loss": 10.6839,
+      "step": 3830
+    },
+    {
+      "epoch": 0.019176508776748483,
+      "grad_norm": 0.09625153988599777,
+      "learning_rate": 2.9498360409522143e-05,
+      "loss": 10.6842,
+      "step": 3840
+    },
+    {
+      "epoch": 0.019226447601687933,
+      "grad_norm": 0.09592974185943604,
+      "learning_rate": 2.949685849458059e-05,
+      "loss": 10.6827,
+      "step": 3850
+    },
+    {
+      "epoch": 0.019276386426627382,
+      "grad_norm": 0.10091127455234528,
+      "learning_rate": 2.949535657963904e-05,
+      "loss": 10.684,
+      "step": 3860
+    },
+    {
+      "epoch": 0.01932632525156683,
+      "grad_norm": 0.09118268638849258,
+      "learning_rate": 2.949385466469749e-05,
+      "loss": 10.6836,
+      "step": 3870
+    },
+    {
+      "epoch": 0.01937626407650628,
+      "grad_norm": 0.09394898265600204,
+      "learning_rate": 2.9492352749755937e-05,
+      "loss": 10.6824,
+      "step": 3880
+    },
+    {
+      "epoch": 0.01942620290144573,
+      "grad_norm": 0.08936263620853424,
+      "learning_rate": 2.949085083481439e-05,
+      "loss": 10.6821,
+      "step": 3890
+    },
+    {
+      "epoch": 0.019476141726385177,
+      "grad_norm": 0.09612365067005157,
+      "learning_rate": 2.9489348919872838e-05,
+      "loss": 10.6827,
+      "step": 3900
+    },
+    {
+      "epoch": 0.019526080551324626,
+      "grad_norm": 0.09695883840322495,
+      "learning_rate": 2.9487847004931288e-05,
+      "loss": 10.6824,
+      "step": 3910
+    },
+    {
+      "epoch": 0.019576019376264076,
+      "grad_norm": 0.0948113203048706,
+      "learning_rate": 2.9486345089989738e-05,
+      "loss": 10.6813,
+      "step": 3920
+    },
+    {
+      "epoch": 0.019625958201203525,
+      "grad_norm": 0.09577417373657227,
+      "learning_rate": 2.9484843175048185e-05,
+      "loss": 10.6822,
+      "step": 3930
+    },
+    {
+      "epoch": 0.019675897026142974,
+      "grad_norm": 0.09592152386903763,
+      "learning_rate": 2.948334126010664e-05,
+      "loss": 10.6808,
+      "step": 3940
+    },
+    {
+      "epoch": 0.019725835851082424,
+      "grad_norm": 0.09353554248809814,
+      "learning_rate": 2.9481839345165085e-05,
+      "loss": 10.68,
+      "step": 3950
+    },
+    {
+      "epoch": 0.019775774676021873,
+      "grad_norm": 0.09733685851097107,
+      "learning_rate": 2.948033743022354e-05,
+      "loss": 10.6801,
+      "step": 3960
+    },
+    {
+      "epoch": 0.019825713500961323,
+      "grad_norm": 0.09616942703723907,
+      "learning_rate": 2.9478835515281986e-05,
+      "loss": 10.6782,
+      "step": 3970
+    },
+    {
+      "epoch": 0.019875652325900772,
+      "grad_norm": 0.091241754591465,
+      "learning_rate": 2.9477333600340432e-05,
+      "loss": 10.6792,
+      "step": 3980
+    },
+    {
+      "epoch": 0.01992559115084022,
+      "grad_norm": 0.09594473242759705,
+      "learning_rate": 2.9475831685398886e-05,
+      "loss": 10.6788,
+      "step": 3990
+    },
+    {
+      "epoch": 0.01997552997577967,
+      "grad_norm": 0.09622420370578766,
+      "learning_rate": 2.9474329770457333e-05,
+      "loss": 10.679,
+      "step": 4000
+    },
+    {
+      "epoch": 0.02002546880071912,
+      "grad_norm": 0.09455501288175583,
+      "learning_rate": 2.9472827855515786e-05,
+      "loss": 10.6778,
+      "step": 4010
+    },
+    {
+      "epoch": 0.020075407625658567,
+      "grad_norm": 0.08896750956773758,
+      "learning_rate": 2.9471325940574233e-05,
+      "loss": 10.679,
+      "step": 4020
+    },
+    {
+      "epoch": 0.020125346450598016,
+      "grad_norm": 0.09691905975341797,
+      "learning_rate": 2.946982402563268e-05,
+      "loss": 10.6772,
+      "step": 4030
+    },
+    {
+      "epoch": 0.020175285275537466,
+      "grad_norm": 0.09754662215709686,
+      "learning_rate": 2.9468322110691134e-05,
+      "loss": 10.6766,
+      "step": 4040
+    },
+    {
+      "epoch": 0.020225224100476915,
+      "grad_norm": 0.09269405156373978,
+      "learning_rate": 2.946682019574958e-05,
+      "loss": 10.6776,
+      "step": 4050
+    },
+    {
+      "epoch": 0.020275162925416364,
+      "grad_norm": 0.08767396956682205,
+      "learning_rate": 2.9465318280808034e-05,
+      "loss": 10.6769,
+      "step": 4060
+    },
+    {
+      "epoch": 0.020325101750355814,
+      "grad_norm": 0.09328394383192062,
+      "learning_rate": 2.946381636586648e-05,
+      "loss": 10.6763,
+      "step": 4070
+    },
+    {
+      "epoch": 0.020375040575295263,
+      "grad_norm": 0.09484425187110901,
+      "learning_rate": 2.9462314450924927e-05,
+      "loss": 10.6768,
+      "step": 4080
+    },
+    {
+      "epoch": 0.020424979400234713,
+      "grad_norm": 0.09627150744199753,
+      "learning_rate": 2.946081253598338e-05,
+      "loss": 10.6757,
+      "step": 4090
+    },
+    {
+      "epoch": 0.020474918225174162,
+      "grad_norm": 0.093561552464962,
+      "learning_rate": 2.9459310621041828e-05,
+      "loss": 10.6754,
+      "step": 4100
+    },
+    {
+      "epoch": 0.02052485705011361,
+      "grad_norm": 0.0967199057340622,
+      "learning_rate": 2.945780870610028e-05,
+      "loss": 10.6758,
+      "step": 4110
+    },
+    {
+      "epoch": 0.02057479587505306,
+      "grad_norm": 0.09370432049036026,
+      "learning_rate": 2.9456306791158728e-05,
+      "loss": 10.6746,
+      "step": 4120
+    },
+    {
+      "epoch": 0.02062473469999251,
+      "grad_norm": 0.10350590944290161,
+      "learning_rate": 2.9454804876217175e-05,
+      "loss": 10.6754,
+      "step": 4130
+    },
+    {
+      "epoch": 0.02067467352493196,
+      "grad_norm": 0.09343244880437851,
+      "learning_rate": 2.945330296127563e-05,
+      "loss": 10.6739,
+      "step": 4140
+    },
+    {
+      "epoch": 0.020724612349871406,
+      "grad_norm": 0.09434296935796738,
+      "learning_rate": 2.9451801046334075e-05,
+      "loss": 10.6729,
+      "step": 4150
+    },
+    {
+      "epoch": 0.020774551174810856,
+      "grad_norm": 0.09708379209041595,
+      "learning_rate": 2.945029913139253e-05,
+      "loss": 10.6736,
+      "step": 4160
+    },
+    {
+      "epoch": 0.020824489999750305,
+      "grad_norm": 0.0962655320763588,
+      "learning_rate": 2.9448797216450976e-05,
+      "loss": 10.6728,
+      "step": 4170
+    },
+    {
+      "epoch": 0.020874428824689754,
+      "grad_norm": 0.10440271347761154,
+      "learning_rate": 2.9447295301509422e-05,
+      "loss": 10.6742,
+      "step": 4180
+    },
+    {
+      "epoch": 0.020924367649629204,
+      "grad_norm": 0.10102055221796036,
+      "learning_rate": 2.9445793386567876e-05,
+      "loss": 10.6736,
+      "step": 4190
+    },
+    {
+      "epoch": 0.020974306474568653,
+      "grad_norm": 0.10021764785051346,
+      "learning_rate": 2.9444291471626323e-05,
+      "loss": 10.6728,
+      "step": 4200
+    },
+    {
+      "epoch": 0.021024245299508103,
+      "grad_norm": 0.09241849929094315,
+      "learning_rate": 2.9442789556684776e-05,
+      "loss": 10.6724,
+      "step": 4210
+    },
+    {
+      "epoch": 0.021074184124447552,
+      "grad_norm": 0.10048867762088776,
+      "learning_rate": 2.9441287641743223e-05,
+      "loss": 10.6714,
+      "step": 4220
+    },
+    {
+      "epoch": 0.021124122949387,
+      "grad_norm": 0.09459193050861359,
+      "learning_rate": 2.943978572680167e-05,
+      "loss": 10.6714,
+      "step": 4230
+    },
+    {
+      "epoch": 0.02117406177432645,
+      "grad_norm": 0.09565653651952744,
+      "learning_rate": 2.9438283811860124e-05,
+      "loss": 10.6698,
+      "step": 4240
+    },
+    {
+      "epoch": 0.0212240005992659,
+      "grad_norm": 0.09747759997844696,
+      "learning_rate": 2.943678189691857e-05,
+      "loss": 10.6708,
+      "step": 4250
+    },
+    {
+      "epoch": 0.02127393942420535,
+      "grad_norm": 0.09385592490434647,
+      "learning_rate": 2.9435279981977024e-05,
+      "loss": 10.6712,
+      "step": 4260
+    },
+    {
+      "epoch": 0.021323878249144796,
+      "grad_norm": 0.09437979012727737,
+      "learning_rate": 2.943377806703547e-05,
+      "loss": 10.672,
+      "step": 4270
+    },
+    {
+      "epoch": 0.021373817074084245,
+      "grad_norm": 0.09702524542808533,
+      "learning_rate": 2.943227615209392e-05,
+      "loss": 10.6702,
+      "step": 4280
+    },
+    {
+      "epoch": 0.021423755899023695,
+      "grad_norm": 0.09239441156387329,
+      "learning_rate": 2.943077423715237e-05,
+      "loss": 10.6703,
+      "step": 4290
+    },
+    {
+      "epoch": 0.021473694723963144,
+      "grad_norm": 0.09539250284433365,
+      "learning_rate": 2.9429272322210818e-05,
+      "loss": 10.6698,
+      "step": 4300
+    },
+    {
+      "epoch": 0.021523633548902594,
+      "grad_norm": 0.09849133342504501,
+      "learning_rate": 2.942777040726927e-05,
+      "loss": 10.6696,
+      "step": 4310
+    },
+    {
+      "epoch": 0.021573572373842043,
+      "grad_norm": 0.09273874014616013,
+      "learning_rate": 2.9426268492327718e-05,
+      "loss": 10.6689,
+      "step": 4320
+    },
+    {
+      "epoch": 0.021623511198781493,
+      "grad_norm": 0.09297140687704086,
+      "learning_rate": 2.942476657738617e-05,
+      "loss": 10.6676,
+      "step": 4330
+    },
+    {
+      "epoch": 0.021673450023720942,
+      "grad_norm": 0.09432685375213623,
+      "learning_rate": 2.942326466244462e-05,
+      "loss": 10.6677,
+      "step": 4340
+    },
+    {
+      "epoch": 0.02172338884866039,
+      "grad_norm": 3660.670654296875,
+      "learning_rate": 2.9421762747503065e-05,
+      "loss": 10.674,
+      "step": 4350
+    },
+    {
+      "epoch": 0.02177332767359984,
+      "grad_norm": 0.09602420032024384,
+      "learning_rate": 2.942026083256152e-05,
+      "loss": 16.9632,
+      "step": 4360
+    },
+    {
+      "epoch": 0.02182326649853929,
+      "grad_norm": 0.09601210057735443,
+      "learning_rate": 2.9418758917619966e-05,
+      "loss": 10.6676,
+      "step": 4370
+    },
+    {
+      "epoch": 0.02187320532347874,
+      "grad_norm": 0.08743233233690262,
+      "learning_rate": 2.9417257002678416e-05,
+      "loss": 10.6673,
+      "step": 4380
+    },
+    {
+      "epoch": 0.021923144148418186,
+      "grad_norm": 0.10627735406160355,
+      "learning_rate": 2.9415755087736866e-05,
+      "loss": 10.8175,
+      "step": 4390
+    },
+    {
+      "epoch": 0.021973082973357635,
+      "grad_norm": 0.09581291675567627,
+      "learning_rate": 2.9414253172795313e-05,
+      "loss": 10.6657,
+      "step": 4400
+    },
+    {
+      "epoch": 0.022023021798297085,
+      "grad_norm": 0.09985919296741486,
+      "learning_rate": 2.9412751257853766e-05,
+      "loss": 10.6651,
+      "step": 4410
+    },
+    {
+      "epoch": 0.022072960623236534,
+      "grad_norm": 0.09798286855220795,
+      "learning_rate": 2.9411249342912213e-05,
+      "loss": 10.6658,
+      "step": 4420
+    },
+    {
+      "epoch": 0.022122899448175984,
+      "grad_norm": 0.09146228432655334,
+      "learning_rate": 2.9409747427970663e-05,
+      "loss": 10.6667,
+      "step": 4430
+    },
+    {
+      "epoch": 0.022172838273115433,
+      "grad_norm": 0.10088124871253967,
+      "learning_rate": 2.9408245513029114e-05,
+      "loss": 10.6659,
+      "step": 4440
+    },
+    {
+      "epoch": 0.022222777098054883,
+      "grad_norm": 0.09959380328655243,
+      "learning_rate": 2.940674359808756e-05,
+      "loss": 10.666,
+      "step": 4450
+    },
+    {
+      "epoch": 0.022272715922994332,
+      "grad_norm": 0.09625295549631119,
+      "learning_rate": 2.9405241683146014e-05,
+      "loss": 10.6653,
+      "step": 4460
+    },
+    {
+      "epoch": 0.02232265474793378,
+      "grad_norm": 0.09174207597970963,
+      "learning_rate": 2.940373976820446e-05,
+      "loss": 10.6652,
+      "step": 4470
+    },
+    {
+      "epoch": 0.02237259357287323,
+      "grad_norm": 0.09853401780128479,
+      "learning_rate": 2.940223785326291e-05,
+      "loss": 10.6626,
+      "step": 4480
+    },
+    {
+      "epoch": 0.02242253239781268,
+      "grad_norm": 0.09681205451488495,
+      "learning_rate": 2.940073593832136e-05,
+      "loss": 10.6643,
+      "step": 4490
+    },
+    {
+      "epoch": 0.02247247122275213,
+      "grad_norm": 0.0938250795006752,
+      "learning_rate": 2.9399234023379808e-05,
+      "loss": 10.6627,
+      "step": 4500
+    },
+    {
+      "epoch": 0.02252241004769158,
+      "grad_norm": 0.09339260309934616,
+      "learning_rate": 2.939773210843826e-05,
+      "loss": 10.6642,
+      "step": 4510
+    },
+    {
+      "epoch": 0.022572348872631025,
+      "grad_norm": 0.0960143581032753,
+      "learning_rate": 2.9396230193496708e-05,
+      "loss": 10.6622,
+      "step": 4520
+    },
+    {
+      "epoch": 0.022622287697570475,
+      "grad_norm": 0.09255195409059525,
+      "learning_rate": 2.939472827855516e-05,
+      "loss": 10.6628,
+      "step": 4530
+    },
+    {
+      "epoch": 0.022672226522509924,
+      "grad_norm": 0.0997820645570755,
+      "learning_rate": 2.939322636361361e-05,
+      "loss": 10.661,
+      "step": 4540
+    },
+    {
+      "epoch": 0.022722165347449374,
+      "grad_norm": 0.09495710581541061,
+      "learning_rate": 2.9391724448672055e-05,
+      "loss": 10.6627,
+      "step": 4550
+    },
+    {
+      "epoch": 0.022772104172388823,
+      "grad_norm": 0.09391613304615021,
+      "learning_rate": 2.939022253373051e-05,
+      "loss": 10.6619,
+      "step": 4560
+    },
+    {
+      "epoch": 0.022822042997328273,
+      "grad_norm": 0.09510276466608047,
+      "learning_rate": 2.9388720618788956e-05,
+      "loss": 10.6613,
+      "step": 4570
+    },
+    {
+      "epoch": 0.022871981822267722,
+      "grad_norm": 0.09492931514978409,
+      "learning_rate": 2.9387218703847406e-05,
+      "loss": 10.6618,
+      "step": 4580
+    },
+    {
+      "epoch": 0.02292192064720717,
+      "grad_norm": 0.0956352949142456,
+      "learning_rate": 2.9385716788905856e-05,
+      "loss": 10.6594,
+      "step": 4590
+    },
+    {
+      "epoch": 0.02297185947214662,
+      "grad_norm": 0.09539540112018585,
+      "learning_rate": 2.9384214873964306e-05,
+      "loss": 10.6604,
+      "step": 4600
+    },
+    {
+      "epoch": 0.02302179829708607,
+      "grad_norm": 0.09625688195228577,
+      "learning_rate": 2.9382712959022756e-05,
+      "loss": 10.6606,
+      "step": 4610
+    },
+    {
+      "epoch": 0.02307173712202552,
+      "grad_norm": 0.09467311948537827,
+      "learning_rate": 2.9381211044081203e-05,
+      "loss": 10.6607,
+      "step": 4620
+    },
+    {
+      "epoch": 0.02312167594696497,
+      "grad_norm": 0.10024222731590271,
+      "learning_rate": 2.9379709129139653e-05,
+      "loss": 10.6599,
+      "step": 4630
+    },
+    {
+      "epoch": 0.023171614771904415,
+      "grad_norm": 0.09454458951950073,
+      "learning_rate": 2.9378207214198104e-05,
+      "loss": 10.6598,
+      "step": 4640
+    },
+    {
+      "epoch": 0.023221553596843865,
+      "grad_norm": 0.10061488300561905,
+      "learning_rate": 2.9376705299256554e-05,
+      "loss": 10.6583,
+      "step": 4650
+    },
+    {
+      "epoch": 0.023271492421783314,
+      "grad_norm": 0.09349951148033142,
+      "learning_rate": 2.9375203384315004e-05,
+      "loss": 10.6581,
+      "step": 4660
+    },
+    {
+      "epoch": 0.023321431246722764,
+      "grad_norm": 0.09272319078445435,
+      "learning_rate": 2.937370146937345e-05,
+      "loss": 10.6588,
+      "step": 4670
+    },
+    {
+      "epoch": 0.023371370071662213,
+      "grad_norm": 0.09603145718574524,
+      "learning_rate": 2.93721995544319e-05,
+      "loss": 10.6568,
+      "step": 4680
+    },
+    {
+      "epoch": 0.023421308896601663,
+      "grad_norm": 0.09523800760507584,
+      "learning_rate": 2.937069763949035e-05,
+      "loss": 10.6573,
+      "step": 4690
+    },
+    {
+      "epoch": 0.023471247721541112,
+      "grad_norm": 0.09391947090625763,
+      "learning_rate": 2.93691957245488e-05,
+      "loss": 10.6567,
+      "step": 4700
+    },
+    {
+      "epoch": 0.02352118654648056,
+      "grad_norm": 0.09547725319862366,
+      "learning_rate": 2.936769380960725e-05,
+      "loss": 10.656,
+      "step": 4710
+    },
+    {
+      "epoch": 0.02357112537142001,
+      "grad_norm": 0.09450241923332214,
+      "learning_rate": 2.9366191894665698e-05,
+      "loss": 10.6568,
+      "step": 4720
+    },
+    {
+      "epoch": 0.02362106419635946,
+      "grad_norm": 0.09558774530887604,
+      "learning_rate": 2.936468997972415e-05,
+      "loss": 10.6559,
+      "step": 4730
+    },
+    {
+      "epoch": 0.02367100302129891,
+      "grad_norm": 0.09340886026620865,
+      "learning_rate": 2.93631880647826e-05,
+      "loss": 10.6566,
+      "step": 4740
+    },
+    {
+      "epoch": 0.02372094184623836,
+      "grad_norm": 0.09513580799102783,
+      "learning_rate": 2.936168614984105e-05,
+      "loss": 10.6557,
+      "step": 4750
+    },
+    {
+      "epoch": 0.023770880671177805,
+      "grad_norm": 0.09321212023496628,
+      "learning_rate": 2.93601842348995e-05,
+      "loss": 10.6559,
+      "step": 4760
+    },
+    {
+      "epoch": 0.023820819496117255,
+      "grad_norm": 0.10149511694908142,
+      "learning_rate": 2.9358682319957946e-05,
+      "loss": 10.6545,
+      "step": 4770
+    },
+    {
+      "epoch": 0.023870758321056704,
+      "grad_norm": 0.09594745188951492,
+      "learning_rate": 2.9357180405016396e-05,
+      "loss": 10.6542,
+      "step": 4780
+    },
+    {
+      "epoch": 0.023920697145996154,
+      "grad_norm": 0.09818349033594131,
+      "learning_rate": 2.9355678490074846e-05,
+      "loss": 10.6542,
+      "step": 4790
+    },
+    {
+      "epoch": 0.023970635970935603,
+      "grad_norm": 0.09233896434307098,
+      "learning_rate": 2.9354176575133296e-05,
+      "loss": 10.654,
+      "step": 4800
+    },
+    {
+      "epoch": 0.024020574795875053,
+      "grad_norm": 0.09461528062820435,
+      "learning_rate": 2.9352674660191746e-05,
+      "loss": 10.6537,
+      "step": 4810
+    },
+    {
+      "epoch": 0.024070513620814502,
+      "grad_norm": 0.10067877173423767,
+      "learning_rate": 2.9351172745250193e-05,
+      "loss": 10.6535,
+      "step": 4820
+    },
+    {
+      "epoch": 0.02412045244575395,
+      "grad_norm": 0.08958308398723602,
+      "learning_rate": 2.9349670830308643e-05,
+      "loss": 10.6532,
+      "step": 4830
+    },
+    {
+      "epoch": 0.0241703912706934,
+      "grad_norm": 0.09203724563121796,
+      "learning_rate": 2.9348168915367094e-05,
+      "loss": 10.652,
+      "step": 4840
+    },
+    {
+      "epoch": 0.02422033009563285,
+      "grad_norm": 0.11031246185302734,
+      "learning_rate": 2.9346667000425544e-05,
+      "loss": 10.6523,
+      "step": 4850
+    },
+    {
+      "epoch": 0.0242702689205723,
+      "grad_norm": 0.09454483538866043,
+      "learning_rate": 2.9345165085483994e-05,
+      "loss": 10.6523,
+      "step": 4860
+    },
+    {
+      "epoch": 0.02432020774551175,
+      "grad_norm": 0.0910816341638565,
+      "learning_rate": 2.934366317054244e-05,
+      "loss": 10.6521,
+      "step": 4870
+    },
+    {
+      "epoch": 0.0243701465704512,
+      "grad_norm": 0.09349098056554794,
+      "learning_rate": 2.934216125560089e-05,
+      "loss": 10.6514,
+      "step": 4880
+    },
+    {
+      "epoch": 0.024420085395390645,
+      "grad_norm": 0.0930526852607727,
+      "learning_rate": 2.934065934065934e-05,
+      "loss": 10.6526,
+      "step": 4890
+    },
+    {
+      "epoch": 0.024470024220330094,
+      "grad_norm": 0.09221765398979187,
+      "learning_rate": 2.933915742571779e-05,
+      "loss": 10.652,
+      "step": 4900
+    },
+    {
+      "epoch": 0.024519963045269544,
+      "grad_norm": 0.08790691196918488,
+      "learning_rate": 2.933765551077624e-05,
+      "loss": 10.6504,
+      "step": 4910
+    },
+    {
+      "epoch": 0.024569901870208993,
+      "grad_norm": 0.09350787848234177,
+      "learning_rate": 2.933615359583469e-05,
+      "loss": 10.6491,
+      "step": 4920
+    },
+    {
+      "epoch": 0.024619840695148443,
+      "grad_norm": 0.10474139451980591,
+      "learning_rate": 2.933465168089314e-05,
+      "loss": 10.651,
+      "step": 4930
+    },
+    {
+      "epoch": 0.024669779520087892,
+      "grad_norm": 0.09355979412794113,
+      "learning_rate": 2.933314976595159e-05,
+      "loss": 10.6504,
+      "step": 4940
+    },
+    {
+      "epoch": 0.02471971834502734,
+      "grad_norm": 0.09748471528291702,
+      "learning_rate": 2.933164785101004e-05,
+      "loss": 10.6502,
+      "step": 4950
+    },
+    {
+      "epoch": 0.02476965716996679,
+      "grad_norm": 0.09395548701286316,
+      "learning_rate": 2.933014593606849e-05,
+      "loss": 10.6488,
+      "step": 4960
+    },
+    {
+      "epoch": 0.02481959599490624,
+      "grad_norm": 0.09552334249019623,
+      "learning_rate": 2.932864402112694e-05,
+      "loss": 10.6497,
+      "step": 4970
+    },
+    {
+      "epoch": 0.02486953481984569,
+      "grad_norm": 0.0992099866271019,
+      "learning_rate": 2.9327142106185386e-05,
+      "loss": 10.6502,
+      "step": 4980
+    },
+    {
+      "epoch": 0.02491947364478514,
+      "grad_norm": 0.09515640139579773,
+      "learning_rate": 2.9325640191243836e-05,
+      "loss": 10.6497,
+      "step": 4990
+    },
+    {
+      "epoch": 0.02496941246972459,
+      "grad_norm": 0.09259581565856934,
+      "learning_rate": 2.9324138276302286e-05,
+      "loss": 10.648,
+      "step": 5000
+    },
+    {
+      "epoch": 0.025019351294664035,
+      "grad_norm": 0.09485942125320435,
+      "learning_rate": 2.9322636361360736e-05,
+      "loss": 10.6474,
+      "step": 5010
+    },
+    {
+      "epoch": 0.025069290119603484,
+      "grad_norm": 0.09455141425132751,
+      "learning_rate": 2.9321134446419187e-05,
+      "loss": 10.6474,
+      "step": 5020
+    },
+    {
+      "epoch": 0.025119228944542934,
+      "grad_norm": 0.09555475413799286,
+      "learning_rate": 2.9319632531477633e-05,
+      "loss": 10.6471,
+      "step": 5030
+    },
+    {
+      "epoch": 0.025169167769482383,
+      "grad_norm": 0.09917616844177246,
+      "learning_rate": 2.9318130616536084e-05,
+      "loss": 10.646,
+      "step": 5040
+    },
+    {
+      "epoch": 0.025219106594421833,
+      "grad_norm": 0.09640449285507202,
+      "learning_rate": 2.9316628701594534e-05,
+      "loss": 10.6467,
+      "step": 5050
+    },
+    {
+      "epoch": 0.025269045419361282,
+      "grad_norm": 0.09314526617527008,
+      "learning_rate": 2.9315126786652984e-05,
+      "loss": 10.6474,
+      "step": 5060
+    },
+    {
+      "epoch": 0.02531898424430073,
+      "grad_norm": 0.09635156393051147,
+      "learning_rate": 2.9313624871711434e-05,
+      "loss": 10.6464,
+      "step": 5070
+    },
+    {
+      "epoch": 0.02536892306924018,
+      "grad_norm": 0.09240912646055222,
+      "learning_rate": 2.931212295676988e-05,
+      "loss": 10.6449,
+      "step": 5080
+    },
+    {
+      "epoch": 0.02541886189417963,
+      "grad_norm": 0.09308675676584244,
+      "learning_rate": 2.931062104182833e-05,
+      "loss": 10.6455,
+      "step": 5090
+    },
+    {
+      "epoch": 0.02546880071911908,
+      "grad_norm": 0.10017043352127075,
+      "learning_rate": 2.930911912688678e-05,
+      "loss": 10.6449,
+      "step": 5100
+    },
+    {
+      "epoch": 0.02551873954405853,
+      "grad_norm": 0.1008182093501091,
+      "learning_rate": 2.930761721194523e-05,
+      "loss": 10.6452,
+      "step": 5110
+    },
+    {
+      "epoch": 0.02556867836899798,
+      "grad_norm": 0.09373480826616287,
+      "learning_rate": 2.930611529700368e-05,
+      "loss": 10.6442,
+      "step": 5120
+    },
+    {
+      "epoch": 0.025618617193937425,
+      "grad_norm": 0.0935322567820549,
+      "learning_rate": 2.930461338206213e-05,
+      "loss": 10.6445,
+      "step": 5130
+    },
+    {
+      "epoch": 0.025668556018876874,
+      "grad_norm": 0.08992544561624527,
+      "learning_rate": 2.930311146712058e-05,
+      "loss": 10.6437,
+      "step": 5140
+    },
+    {
+      "epoch": 0.025718494843816324,
+      "grad_norm": 0.09083853662014008,
+      "learning_rate": 2.930160955217903e-05,
+      "loss": 10.6428,
+      "step": 5150
+    },
+    {
+      "epoch": 0.025768433668755773,
+      "grad_norm": 0.09427718818187714,
+      "learning_rate": 2.930010763723748e-05,
+      "loss": 10.642,
+      "step": 5160
+    },
+    {
+      "epoch": 0.025818372493695223,
+      "grad_norm": 0.09764672070741653,
+      "learning_rate": 2.929860572229593e-05,
+      "loss": 10.6418,
+      "step": 5170
+    },
+    {
+      "epoch": 0.025868311318634672,
+      "grad_norm": 0.09804851561784744,
+      "learning_rate": 2.9297103807354376e-05,
+      "loss": 10.6429,
+      "step": 5180
+    },
+    {
+      "epoch": 0.02591825014357412,
+      "grad_norm": 0.095367930829525,
+      "learning_rate": 2.9295601892412826e-05,
+      "loss": 10.6437,
+      "step": 5190
+    },
+    {
+      "epoch": 0.02596818896851357,
+      "grad_norm": 0.09083083271980286,
+      "learning_rate": 2.9294099977471276e-05,
+      "loss": 10.6429,
+      "step": 5200
+    },
+    {
+      "epoch": 0.02601812779345302,
+      "grad_norm": 0.09228214621543884,
+      "learning_rate": 2.9292598062529726e-05,
+      "loss": 10.6408,
+      "step": 5210
+    },
+    {
+      "epoch": 0.02606806661839247,
+      "grad_norm": 0.09770570695400238,
+      "learning_rate": 2.9291096147588177e-05,
+      "loss": 10.6422,
+      "step": 5220
+    },
+    {
+      "epoch": 0.02611800544333192,
+      "grad_norm": 0.0948592945933342,
+      "learning_rate": 2.9289594232646623e-05,
+      "loss": 10.6408,
+      "step": 5230
+    },
+    {
+      "epoch": 0.02616794426827137,
+      "grad_norm": 0.09406052529811859,
+      "learning_rate": 2.9288092317705077e-05,
+      "loss": 10.6412,
+      "step": 5240
+    },
+    {
+      "epoch": 0.02621788309321082,
+      "grad_norm": 0.10018353909254074,
+      "learning_rate": 2.9286590402763524e-05,
+      "loss": 10.6396,
+      "step": 5250
+    },
+    {
+      "epoch": 0.026267821918150264,
+      "grad_norm": 0.08951601386070251,
+      "learning_rate": 2.9285088487821974e-05,
+      "loss": 10.642,
+      "step": 5260
+    },
+    {
+      "epoch": 0.026317760743089714,
+      "grad_norm": 0.10409606993198395,
+      "learning_rate": 2.9283586572880424e-05,
+      "loss": 10.6401,
+      "step": 5270
+    },
+    {
+      "epoch": 0.026367699568029163,
+      "grad_norm": 0.09470842033624649,
+      "learning_rate": 2.928208465793887e-05,
+      "loss": 10.6407,
+      "step": 5280
+    },
+    {
+      "epoch": 0.026417638392968613,
+      "grad_norm": 0.09823372960090637,
+      "learning_rate": 2.9280582742997324e-05,
+      "loss": 10.6392,
+      "step": 5290
+    },
+    {
+      "epoch": 0.026467577217908062,
+      "grad_norm": 0.10075034201145172,
+      "learning_rate": 2.927908082805577e-05,
+      "loss": 10.6392,
+      "step": 5300
+    },
+    {
+      "epoch": 0.02651751604284751,
+      "grad_norm": 0.09643511474132538,
+      "learning_rate": 2.927757891311422e-05,
+      "loss": 10.6394,
+      "step": 5310
+    },
+    {
+      "epoch": 0.02656745486778696,
+      "grad_norm": 0.09283813089132309,
+      "learning_rate": 2.927607699817267e-05,
+      "loss": 10.6378,
+      "step": 5320
+    },
+    {
+      "epoch": 0.02661739369272641,
+      "grad_norm": 0.09951145201921463,
+      "learning_rate": 2.927457508323112e-05,
+      "loss": 10.6374,
+      "step": 5330
+    },
+    {
+      "epoch": 0.02666733251766586,
+      "grad_norm": 0.0960063636302948,
+      "learning_rate": 2.9273073168289572e-05,
+      "loss": 10.6374,
+      "step": 5340
+    },
+    {
+      "epoch": 0.02671727134260531,
+      "grad_norm": 0.09506009519100189,
+      "learning_rate": 2.927157125334802e-05,
+      "loss": 10.6369,
+      "step": 5350
+    },
+    {
+      "epoch": 0.02676721016754476,
+      "grad_norm": 0.09488440304994583,
+      "learning_rate": 2.927006933840647e-05,
+      "loss": 10.6387,
+      "step": 5360
+    },
+    {
+      "epoch": 0.02681714899248421,
+      "grad_norm": 0.09644606709480286,
+      "learning_rate": 2.926856742346492e-05,
+      "loss": 10.6359,
+      "step": 5370
+    },
+    {
+      "epoch": 0.026867087817423654,
+      "grad_norm": 0.09125857800245285,
+      "learning_rate": 2.9267065508523366e-05,
+      "loss": 10.6361,
+      "step": 5380
+    },
+    {
+      "epoch": 0.026917026642363104,
+      "grad_norm": 0.09280921518802643,
+      "learning_rate": 2.926556359358182e-05,
+      "loss": 10.6364,
+      "step": 5390
+    },
+    {
+      "epoch": 0.026966965467302553,
+      "grad_norm": 0.09589015692472458,
+      "learning_rate": 2.9264061678640266e-05,
+      "loss": 10.6376,
+      "step": 5400
+    },
+    {
+      "epoch": 0.027016904292242003,
+      "grad_norm": 0.0969088077545166,
+      "learning_rate": 2.9262559763698716e-05,
+      "loss": 10.6358,
+      "step": 5410
+    },
+    {
+      "epoch": 0.027066843117181452,
+      "grad_norm": 0.09490292519330978,
+      "learning_rate": 2.9261057848757167e-05,
+      "loss": 10.6357,
+      "step": 5420
+    },
+    {
+      "epoch": 0.0271167819421209,
+      "grad_norm": 0.09488876163959503,
+      "learning_rate": 2.9259555933815613e-05,
+      "loss": 10.6352,
+      "step": 5430
+    },
+    {
+      "epoch": 0.02716672076706035,
+      "grad_norm": 0.09537888318300247,
+      "learning_rate": 2.9258054018874067e-05,
+      "loss": 10.6348,
+      "step": 5440
+    },
+    {
+      "epoch": 0.0272166595919998,
+      "grad_norm": 0.0909004881978035,
+      "learning_rate": 2.9256552103932514e-05,
+      "loss": 10.6343,
+      "step": 5450
+    },
+    {
+      "epoch": 0.02726659841693925,
+      "grad_norm": 0.0989050418138504,
+      "learning_rate": 2.9255050188990964e-05,
+      "loss": 10.6339,
+      "step": 5460
+    },
+    {
+      "epoch": 0.0273165372418787,
+      "grad_norm": 0.09519681334495544,
+      "learning_rate": 2.9253548274049414e-05,
+      "loss": 10.6333,
+      "step": 5470
+    },
+    {
+      "epoch": 0.02736647606681815,
+      "grad_norm": 0.09233660250902176,
+      "learning_rate": 2.925204635910786e-05,
+      "loss": 10.6336,
+      "step": 5480
+    },
+    {
+      "epoch": 0.0274164148917576,
+      "grad_norm": 0.09617756307125092,
+      "learning_rate": 2.9250544444166314e-05,
+      "loss": 10.6345,
+      "step": 5490
+    },
+    {
+      "epoch": 0.027466353716697048,
+      "grad_norm": 0.09878206998109818,
+      "learning_rate": 2.924904252922476e-05,
+      "loss": 10.6326,
+      "step": 5500
+    },
+    {
+      "epoch": 0.027516292541636494,
+      "grad_norm": 0.10213974118232727,
+      "learning_rate": 2.924754061428321e-05,
+      "loss": 10.6315,
+      "step": 5510
+    },
+    {
+      "epoch": 0.027566231366575943,
+      "grad_norm": 0.09810013324022293,
+      "learning_rate": 2.924603869934166e-05,
+      "loss": 10.6324,
+      "step": 5520
+    },
+    {
+      "epoch": 0.027616170191515393,
+      "grad_norm": 0.0984368622303009,
+      "learning_rate": 2.924453678440011e-05,
+      "loss": 10.6327,
+      "step": 5530
+    },
+    {
+      "epoch": 0.027666109016454842,
+      "grad_norm": 0.094744011759758,
+      "learning_rate": 2.9243034869458562e-05,
+      "loss": 10.6325,
+      "step": 5540
+    },
+    {
+      "epoch": 0.02771604784139429,
+      "grad_norm": 0.09556036442518234,
+      "learning_rate": 2.924153295451701e-05,
+      "loss": 10.6324,
+      "step": 5550
+    },
+    {
+      "epoch": 0.02776598666633374,
+      "grad_norm": 0.09775315970182419,
+      "learning_rate": 2.9240031039575462e-05,
+      "loss": 10.6308,
+      "step": 5560
+    },
+    {
+      "epoch": 0.02781592549127319,
+      "grad_norm": 0.09518972039222717,
+      "learning_rate": 2.923852912463391e-05,
+      "loss": 10.6312,
+      "step": 5570
+    },
+    {
+      "epoch": 0.02786586431621264,
+      "grad_norm": 0.10233526676893234,
+      "learning_rate": 2.9237027209692356e-05,
+      "loss": 10.631,
+      "step": 5580
+    },
+    {
+      "epoch": 0.02791580314115209,
+      "grad_norm": 0.09543079882860184,
+      "learning_rate": 2.923552529475081e-05,
+      "loss": 10.6295,
+      "step": 5590
+    },
+    {
+      "epoch": 0.02796574196609154,
+      "grad_norm": 0.08888249844312668,
+      "learning_rate": 2.9234023379809256e-05,
+      "loss": 10.6316,
+      "step": 5600
+    },
+    {
+      "epoch": 0.02801568079103099,
+      "grad_norm": 0.09589431434869766,
+      "learning_rate": 2.923252146486771e-05,
+      "loss": 10.6294,
+      "step": 5610
+    },
+    {
+      "epoch": 0.028065619615970438,
+      "grad_norm": 0.09286142885684967,
+      "learning_rate": 2.9231019549926157e-05,
+      "loss": 10.6302,
+      "step": 5620
+    },
+    {
+      "epoch": 0.028115558440909884,
+      "grad_norm": 0.09479831159114838,
+      "learning_rate": 2.9229517634984603e-05,
+      "loss": 10.629,
+      "step": 5630
+    },
+    {
+      "epoch": 0.028165497265849333,
+      "grad_norm": 0.09502566605806351,
+      "learning_rate": 2.9228015720043057e-05,
+      "loss": 10.6301,
+      "step": 5640
+    },
+    {
+      "epoch": 0.028215436090788783,
+      "grad_norm": 0.0879800021648407,
+      "learning_rate": 2.9226513805101504e-05,
+      "loss": 10.6291,
+      "step": 5650
+    },
+    {
+      "epoch": 0.028265374915728232,
+      "grad_norm": 0.09621425718069077,
+      "learning_rate": 2.9225011890159957e-05,
+      "loss": 10.6279,
+      "step": 5660
+    },
+    {
+      "epoch": 0.02831531374066768,
+      "grad_norm": 0.0952867642045021,
+      "learning_rate": 2.9223509975218404e-05,
+      "loss": 10.629,
+      "step": 5670
+    },
+    {
+      "epoch": 0.02836525256560713,
+      "grad_norm": 0.09745540469884872,
+      "learning_rate": 2.922200806027685e-05,
+      "loss": 10.6263,
+      "step": 5680
+    },
+    {
+      "epoch": 0.02841519139054658,
+      "grad_norm": 0.10703998804092407,
+      "learning_rate": 2.9220506145335305e-05,
+      "loss": 10.6267,
+      "step": 5690
+    },
+    {
+      "epoch": 0.02846513021548603,
+      "grad_norm": 0.09526613354682922,
+      "learning_rate": 2.921900423039375e-05,
+      "loss": 10.6286,
+      "step": 5700
+    },
+    {
+      "epoch": 0.02851506904042548,
+      "grad_norm": 0.09355132281780243,
+      "learning_rate": 2.9217502315452205e-05,
+      "loss": 10.6258,
+      "step": 5710
+    },
+    {
+      "epoch": 0.02856500786536493,
+      "grad_norm": 0.09581465274095535,
+      "learning_rate": 2.921600040051065e-05,
+      "loss": 10.6266,
+      "step": 5720
+    },
+    {
+      "epoch": 0.02861494669030438,
+      "grad_norm": 0.09699860215187073,
+      "learning_rate": 2.92144984855691e-05,
+      "loss": 10.6275,
+      "step": 5730
+    },
+    {
+      "epoch": 0.028664885515243828,
+      "grad_norm": 0.09118512272834778,
+      "learning_rate": 2.9212996570627552e-05,
+      "loss": 10.6246,
+      "step": 5740
+    },
+    {
+      "epoch": 0.028714824340183274,
+      "grad_norm": 0.09603618830442429,
+      "learning_rate": 2.9211494655686e-05,
+      "loss": 10.6266,
+      "step": 5750
+    },
+    {
+      "epoch": 0.028764763165122723,
+      "grad_norm": 0.09384205937385559,
+      "learning_rate": 2.9209992740744452e-05,
+      "loss": 10.6246,
+      "step": 5760
+    },
+    {
+      "epoch": 0.028814701990062173,
+      "grad_norm": 0.09352291375398636,
+      "learning_rate": 2.92084908258029e-05,
+      "loss": 10.6245,
+      "step": 5770
+    },
+    {
+      "epoch": 0.028864640815001622,
+      "grad_norm": 0.09608279168605804,
+      "learning_rate": 2.9206988910861346e-05,
+      "loss": 10.6247,
+      "step": 5780
+    },
+    {
+      "epoch": 0.02891457963994107,
+      "grad_norm": 0.09599608927965164,
+      "learning_rate": 2.92054869959198e-05,
+      "loss": 10.6243,
+      "step": 5790
+    },
+    {
+      "epoch": 0.02896451846488052,
+      "grad_norm": 0.10410738736391068,
+      "learning_rate": 2.9203985080978246e-05,
+      "loss": 10.6244,
+      "step": 5800
+    },
+    {
+      "epoch": 0.02901445728981997,
+      "grad_norm": 0.10078850388526917,
+      "learning_rate": 2.92024831660367e-05,
+      "loss": 10.6234,
+      "step": 5810
+    },
+    {
+      "epoch": 0.02906439611475942,
+      "grad_norm": 0.09706875681877136,
+      "learning_rate": 2.9200981251095147e-05,
+      "loss": 10.6225,
+      "step": 5820
+    },
+    {
+      "epoch": 0.02911433493969887,
+      "grad_norm": 0.09206659346818924,
+      "learning_rate": 2.9199479336153593e-05,
+      "loss": 10.6223,
+      "step": 5830
+    },
+    {
+      "epoch": 0.02916427376463832,
+      "grad_norm": 0.09456048905849457,
+      "learning_rate": 2.9197977421212047e-05,
+      "loss": 10.6215,
+      "step": 5840
+    },
+    {
+      "epoch": 0.02921421258957777,
+      "grad_norm": 0.09819670021533966,
+      "learning_rate": 2.9196475506270494e-05,
+      "loss": 10.6221,
+      "step": 5850
+    },
+    {
+      "epoch": 0.029264151414517218,
+      "grad_norm": 0.09750770032405853,
+      "learning_rate": 2.9194973591328947e-05,
+      "loss": 10.6216,
+      "step": 5860
+    },
+    {
+      "epoch": 0.029314090239456667,
+      "grad_norm": 0.09041731804609299,
+      "learning_rate": 2.9193471676387394e-05,
+      "loss": 10.6213,
+      "step": 5870
+    },
+    {
+      "epoch": 0.029364029064396113,
+      "grad_norm": 0.09223254024982452,
+      "learning_rate": 2.9191969761445844e-05,
+      "loss": 10.6214,
+      "step": 5880
+    },
+    {
+      "epoch": 0.029413967889335563,
+      "grad_norm": 0.09175052493810654,
+      "learning_rate": 2.9190467846504295e-05,
+      "loss": 10.62,
+      "step": 5890
+    },
+    {
+      "epoch": 0.029463906714275012,
+      "grad_norm": 0.0979679524898529,
+      "learning_rate": 2.918896593156274e-05,
+      "loss": 10.6207,
+      "step": 5900
+    },
+    {
+      "epoch": 0.02951384553921446,
+      "grad_norm": 0.0917288288474083,
+      "learning_rate": 2.9187464016621195e-05,
+      "loss": 10.6207,
+      "step": 5910
+    },
+    {
+      "epoch": 0.02956378436415391,
+      "grad_norm": 0.09289643913507462,
+      "learning_rate": 2.918596210167964e-05,
+      "loss": 10.6209,
+      "step": 5920
+    },
+    {
+      "epoch": 0.02961372318909336,
+      "grad_norm": 0.09195461124181747,
+      "learning_rate": 2.9184460186738092e-05,
+      "loss": 10.6197,
+      "step": 5930
+    },
+    {
+      "epoch": 0.02966366201403281,
+      "grad_norm": 0.09683648496866226,
+      "learning_rate": 2.9182958271796542e-05,
+      "loss": 10.6192,
+      "step": 5940
+    },
+    {
+      "epoch": 0.02971360083897226,
+      "grad_norm": 0.08719411492347717,
+      "learning_rate": 2.918145635685499e-05,
+      "loss": 10.619,
+      "step": 5950
+    },
+    {
+      "epoch": 0.02976353966391171,
+      "grad_norm": 0.0998111441731453,
+      "learning_rate": 2.9179954441913442e-05,
+      "loss": 10.6193,
+      "step": 5960
+    },
+    {
+      "epoch": 0.02981347848885116,
+      "grad_norm": 0.09359613060951233,
+      "learning_rate": 2.917845252697189e-05,
+      "loss": 10.618,
+      "step": 5970
+    },
+    {
+      "epoch": 0.029863417313790608,
+      "grad_norm": 0.09427807480096817,
+      "learning_rate": 2.917695061203034e-05,
+      "loss": 10.6174,
+      "step": 5980
+    },
+    {
+      "epoch": 0.029913356138730057,
+      "grad_norm": 0.09695295989513397,
+      "learning_rate": 2.917544869708879e-05,
+      "loss": 10.6173,
+      "step": 5990
+    },
+    {
+      "epoch": 0.029963294963669503,
+      "grad_norm": 0.10022547841072083,
+      "learning_rate": 2.9173946782147236e-05,
+      "loss": 10.6184,
+      "step": 6000
+    },
+    {
+      "epoch": 0.030013233788608953,
+      "grad_norm": 0.09429972618818283,
+      "learning_rate": 2.917244486720569e-05,
+      "loss": 10.6166,
+      "step": 6010
+    },
+    {
+      "epoch": 0.030063172613548402,
+      "grad_norm": 0.09150966256856918,
+      "learning_rate": 2.9170942952264137e-05,
+      "loss": 10.6176,
+      "step": 6020
+    },
+    {
+      "epoch": 0.03011311143848785,
+      "grad_norm": 0.09376371651887894,
+      "learning_rate": 2.9169441037322587e-05,
+      "loss": 10.6167,
+      "step": 6030
+    },
+    {
+      "epoch": 0.0301630502634273,
+      "grad_norm": 0.09400255978107452,
+      "learning_rate": 2.9167939122381037e-05,
+      "loss": 10.6147,
+      "step": 6040
+    },
+    {
+      "epoch": 0.03021298908836675,
+      "grad_norm": 0.10719768702983856,
+      "learning_rate": 2.9166437207439484e-05,
+      "loss": 10.6168,
+      "step": 6050
+    },
+    {
+      "epoch": 0.0302629279133062,
+      "grad_norm": 0.0984666720032692,
+      "learning_rate": 2.9164935292497937e-05,
+      "loss": 10.6148,
+      "step": 6060
+    },
+    {
+      "epoch": 0.03031286673824565,
+      "grad_norm": 0.09107008576393127,
+      "learning_rate": 2.9163433377556384e-05,
+      "loss": 10.6148,
+      "step": 6070
+    },
+    {
+      "epoch": 0.0303628055631851,
+      "grad_norm": 0.09866014868021011,
+      "learning_rate": 2.9161931462614834e-05,
+      "loss": 10.6146,
+      "step": 6080
+    },
+    {
+      "epoch": 0.03041274438812455,
+      "grad_norm": 0.0949750542640686,
+      "learning_rate": 2.9160429547673285e-05,
+      "loss": 10.6148,
+      "step": 6090
+    },
+    {
+      "epoch": 0.030462683213063998,
+      "grad_norm": 0.09634987264871597,
+      "learning_rate": 2.915892763273173e-05,
+      "loss": 10.6139,
+      "step": 6100
+    },
+    {
+      "epoch": 0.030512622038003447,
+      "grad_norm": 0.09872616082429886,
+      "learning_rate": 2.9157425717790185e-05,
+      "loss": 10.615,
+      "step": 6110
+    },
+    {
+      "epoch": 0.030562560862942893,
+      "grad_norm": 0.09585590660572052,
+      "learning_rate": 2.915592380284863e-05,
+      "loss": 10.6138,
+      "step": 6120
+    },
+    {
+      "epoch": 0.030612499687882343,
+      "grad_norm": 0.09388266503810883,
+      "learning_rate": 2.9154421887907082e-05,
+      "loss": 10.614,
+      "step": 6130
+    },
+    {
+      "epoch": 0.030662438512821792,
+      "grad_norm": 0.10067202895879745,
+      "learning_rate": 2.9152919972965532e-05,
+      "loss": 10.6142,
+      "step": 6140
+    },
+    {
+      "epoch": 0.03071237733776124,
+      "grad_norm": 0.09900005161762238,
+      "learning_rate": 2.915141805802398e-05,
+      "loss": 10.6135,
+      "step": 6150
+    },
+    {
+      "epoch": 0.03076231616270069,
+      "grad_norm": 0.09707643836736679,
+      "learning_rate": 2.9149916143082432e-05,
+      "loss": 10.6129,
+      "step": 6160
+    },
+    {
+      "epoch": 0.03081225498764014,
+      "grad_norm": 0.09792651236057281,
+      "learning_rate": 2.914841422814088e-05,
+      "loss": 10.6117,
+      "step": 6170
+    },
+    {
+      "epoch": 0.03086219381257959,
+      "grad_norm": 0.09310603886842728,
+      "learning_rate": 2.914691231319933e-05,
+      "loss": 10.6116,
+      "step": 6180
+    },
+    {
+      "epoch": 0.03091213263751904,
+      "grad_norm": 0.09561890363693237,
+      "learning_rate": 2.914541039825778e-05,
+      "loss": 10.6131,
+      "step": 6190
+    },
+    {
+      "epoch": 0.03096207146245849,
+      "grad_norm": 0.09028071165084839,
+      "learning_rate": 2.914390848331623e-05,
+      "loss": 10.6135,
+      "step": 6200
+    },
+    {
+      "epoch": 0.03101201028739794,
+      "grad_norm": 0.09210582822561264,
+      "learning_rate": 2.914240656837468e-05,
+      "loss": 10.6127,
+      "step": 6210
+    },
+    {
+      "epoch": 0.031061949112337388,
+      "grad_norm": 0.09679680317640305,
+      "learning_rate": 2.9140904653433127e-05,
+      "loss": 10.611,
+      "step": 6220
+    },
+    {
+      "epoch": 0.031111887937276837,
+      "grad_norm": 0.09834344685077667,
+      "learning_rate": 2.9139402738491577e-05,
+      "loss": 10.6099,
+      "step": 6230
+    },
+    {
+      "epoch": 0.031161826762216287,
+      "grad_norm": 0.09306725114583969,
+      "learning_rate": 2.9137900823550027e-05,
+      "loss": 10.6113,
+      "step": 6240
+    },
+    {
+      "epoch": 0.031211765587155733,
+      "grad_norm": 0.09439186751842499,
+      "learning_rate": 2.9136398908608477e-05,
+      "loss": 10.6104,
+      "step": 6250
+    },
+    {
+      "epoch": 0.031261704412095186,
+      "grad_norm": 0.10274600237607956,
+      "learning_rate": 2.9134896993666927e-05,
+      "loss": 10.6088,
+      "step": 6260
+    },
+    {
+      "epoch": 0.031311643237034635,
+      "grad_norm": 0.10353299975395203,
+      "learning_rate": 2.9133395078725374e-05,
+      "loss": 10.6096,
+      "step": 6270
+    },
+    {
+      "epoch": 0.031361582061974085,
+      "grad_norm": 0.09838922321796417,
+      "learning_rate": 2.9131893163783824e-05,
+      "loss": 10.6104,
+      "step": 6280
+    },
+    {
+      "epoch": 0.031411520886913534,
+      "grad_norm": 0.0952482670545578,
+      "learning_rate": 2.9130391248842275e-05,
+      "loss": 10.6079,
+      "step": 6290
+    },
+    {
+      "epoch": 0.03146145971185298,
+      "grad_norm": 0.0897747352719307,
+      "learning_rate": 2.9128889333900725e-05,
+      "loss": 10.6092,
+      "step": 6300
+    },
+    {
+      "epoch": 0.031511398536792426,
+      "grad_norm": 0.08722511678934097,
+      "learning_rate": 2.9127387418959175e-05,
+      "loss": 10.611,
+      "step": 6310
+    },
+    {
+      "epoch": 0.031561337361731875,
+      "grad_norm": 0.09219997376203537,
+      "learning_rate": 2.9125885504017622e-05,
+      "loss": 10.6096,
+      "step": 6320
+    },
+    {
+      "epoch": 0.031611276186671325,
+      "grad_norm": 0.08804293721914291,
+      "learning_rate": 2.9124383589076072e-05,
+      "loss": 10.6068,
+      "step": 6330
+    },
+    {
+      "epoch": 0.031661215011610774,
+      "grad_norm": 0.09116089344024658,
+      "learning_rate": 2.9122881674134522e-05,
+      "loss": 10.6073,
+      "step": 6340
+    },
+    {
+      "epoch": 0.031711153836550224,
+      "grad_norm": 0.10001252591609955,
+      "learning_rate": 2.9121379759192972e-05,
+      "loss": 10.6077,
+      "step": 6350
+    },
+    {
+      "epoch": 0.03176109266148967,
+      "grad_norm": 0.0931568518280983,
+      "learning_rate": 2.9119877844251422e-05,
+      "loss": 10.6072,
+      "step": 6360
+    },
+    {
+      "epoch": 0.03181103148642912,
+      "grad_norm": 0.09584708511829376,
+      "learning_rate": 2.911837592930987e-05,
+      "loss": 10.6061,
+      "step": 6370
+    },
+    {
+      "epoch": 0.03186097031136857,
+      "grad_norm": 0.09385517984628677,
+      "learning_rate": 2.911687401436832e-05,
+      "loss": 10.6075,
+      "step": 6380
+    },
+    {
+      "epoch": 0.03191090913630802,
+      "grad_norm": 0.09303200244903564,
+      "learning_rate": 2.911537209942677e-05,
+      "loss": 10.607,
+      "step": 6390
+    },
+    {
+      "epoch": 0.03196084796124747,
+      "grad_norm": 0.09322622418403625,
+      "learning_rate": 2.911387018448522e-05,
+      "loss": 10.6056,
+      "step": 6400
+    },
+    {
+      "epoch": 0.03201078678618692,
+      "grad_norm": 0.09804125130176544,
+      "learning_rate": 2.911236826954367e-05,
+      "loss": 10.6035,
+      "step": 6410
+    },
+    {
+      "epoch": 0.03206072561112637,
+      "grad_norm": 0.09895630925893784,
+      "learning_rate": 2.9110866354602117e-05,
+      "loss": 10.6053,
+      "step": 6420
+    },
+    {
+      "epoch": 0.03211066443606582,
+      "grad_norm": 0.09591073542833328,
+      "learning_rate": 2.9109364439660567e-05,
+      "loss": 10.6044,
+      "step": 6430
+    },
+    {
+      "epoch": 0.03216060326100527,
+      "grad_norm": 0.08683682233095169,
+      "learning_rate": 2.9107862524719017e-05,
+      "loss": 10.6046,
+      "step": 6440
+    },
+    {
+      "epoch": 0.03221054208594472,
+      "grad_norm": 0.09420903772115707,
+      "learning_rate": 2.9106360609777467e-05,
+      "loss": 10.6029,
+      "step": 6450
+    },
+    {
+      "epoch": 0.03226048091088417,
+      "grad_norm": 0.09442789852619171,
+      "learning_rate": 2.9104858694835917e-05,
+      "loss": 10.6037,
+      "step": 6460
+    },
+    {
+      "epoch": 0.03231041973582362,
+      "grad_norm": 0.09980372339487076,
+      "learning_rate": 2.9103356779894364e-05,
+      "loss": 10.6028,
+      "step": 6470
+    },
+    {
+      "epoch": 0.03236035856076307,
+      "grad_norm": 0.0942731723189354,
+      "learning_rate": 2.9101854864952814e-05,
+      "loss": 10.603,
+      "step": 6480
+    },
+    {
+      "epoch": 0.032410297385702516,
+      "grad_norm": 0.09348214417695999,
+      "learning_rate": 2.9100352950011265e-05,
+      "loss": 10.6036,
+      "step": 6490
+    },
+    {
+      "epoch": 0.032460236210641966,
+      "grad_norm": 0.09484933316707611,
+      "learning_rate": 2.9098851035069715e-05,
+      "loss": 10.6022,
+      "step": 6500
+    },
+    {
+      "epoch": 0.032510175035581415,
+      "grad_norm": 0.09451691806316376,
+      "learning_rate": 2.9097349120128165e-05,
+      "loss": 10.6036,
+      "step": 6510
+    },
+    {
+      "epoch": 0.032560113860520865,
+      "grad_norm": 0.0937938317656517,
+      "learning_rate": 2.9095847205186612e-05,
+      "loss": 10.6008,
+      "step": 6520
+    },
+    {
+      "epoch": 0.032610052685460314,
+      "grad_norm": 0.09523924440145493,
+      "learning_rate": 2.9094345290245062e-05,
+      "loss": 10.601,
+      "step": 6530
+    },
+    {
+      "epoch": 0.03265999151039976,
+      "grad_norm": 0.09146000444889069,
+      "learning_rate": 2.9092843375303512e-05,
+      "loss": 10.601,
+      "step": 6540
+    },
+    {
+      "epoch": 0.03270993033533921,
+      "grad_norm": 0.09606587141752243,
+      "learning_rate": 2.9091341460361962e-05,
+      "loss": 10.5993,
+      "step": 6550
+    },
+    {
+      "epoch": 0.032759869160278655,
+      "grad_norm": 0.09281494468450546,
+      "learning_rate": 2.9089839545420412e-05,
+      "loss": 10.6012,
+      "step": 6560
+    },
+    {
+      "epoch": 0.032809807985218105,
+      "grad_norm": 0.09290701895952225,
+      "learning_rate": 2.9088337630478863e-05,
+      "loss": 10.6008,
+      "step": 6570
+    },
+    {
+      "epoch": 0.032859746810157554,
+      "grad_norm": 0.10137023776769638,
+      "learning_rate": 2.908683571553731e-05,
+      "loss": 10.6012,
+      "step": 6580
+    },
+    {
+      "epoch": 0.032909685635097004,
+      "grad_norm": 0.09537509083747864,
+      "learning_rate": 2.908533380059576e-05,
+      "loss": 10.599,
+      "step": 6590
+    },
+    {
+      "epoch": 0.03295962446003645,
+      "grad_norm": 0.0944240391254425,
+      "learning_rate": 2.908383188565421e-05,
+      "loss": 10.6006,
+      "step": 6600
+    },
+    {
+      "epoch": 0.0330095632849759,
+      "grad_norm": 0.09300459921360016,
+      "learning_rate": 2.908232997071266e-05,
+      "loss": 10.6005,
+      "step": 6610
+    },
+    {
+      "epoch": 0.03305950210991535,
+      "grad_norm": 0.09325224906206131,
+      "learning_rate": 2.908082805577111e-05,
+      "loss": 10.5999,
+      "step": 6620
+    },
+    {
+      "epoch": 0.0331094409348548,
+      "grad_norm": 0.09946674108505249,
+      "learning_rate": 2.9079326140829557e-05,
+      "loss": 10.597,
+      "step": 6630
+    },
+    {
+      "epoch": 0.03315937975979425,
+      "grad_norm": 0.09831472486257553,
+      "learning_rate": 2.9077824225888007e-05,
+      "loss": 10.5973,
+      "step": 6640
+    },
+    {
+      "epoch": 0.0332093185847337,
+      "grad_norm": 0.09227102994918823,
+      "learning_rate": 2.9076322310946457e-05,
+      "loss": 10.5992,
+      "step": 6650
+    },
+    {
+      "epoch": 0.03325925740967315,
+      "grad_norm": 0.094718337059021,
+      "learning_rate": 2.9074820396004907e-05,
+      "loss": 10.5978,
+      "step": 6660
+    },
+    {
+      "epoch": 0.0333091962346126,
+      "grad_norm": 0.09251998364925385,
+      "learning_rate": 2.9073318481063358e-05,
+      "loss": 10.5985,
+      "step": 6670
+    },
+    {
+      "epoch": 0.03335913505955205,
+      "grad_norm": 0.0993405282497406,
+      "learning_rate": 2.9071816566121804e-05,
+      "loss": 10.599,
+      "step": 6680
+    },
+    {
+      "epoch": 0.0334090738844915,
+      "grad_norm": 0.09692016988992691,
+      "learning_rate": 2.9070314651180255e-05,
+      "loss": 10.5975,
+      "step": 6690
+    },
+    {
+      "epoch": 0.03345901270943095,
+      "grad_norm": 0.0930512323975563,
+      "learning_rate": 2.9068812736238705e-05,
+      "loss": 10.5965,
+      "step": 6700
+    },
+    {
+      "epoch": 0.0335089515343704,
+      "grad_norm": 0.09981938451528549,
+      "learning_rate": 2.9067310821297155e-05,
+      "loss": 10.5969,
+      "step": 6710
+    },
+    {
+      "epoch": 0.03355889035930985,
+      "grad_norm": 0.09343113005161285,
+      "learning_rate": 2.9065808906355605e-05,
+      "loss": 10.5961,
+      "step": 6720
+    },
+    {
+      "epoch": 0.033608829184249296,
+      "grad_norm": 0.09517756849527359,
+      "learning_rate": 2.9064306991414052e-05,
+      "loss": 10.5965,
+      "step": 6730
+    },
+    {
+      "epoch": 0.033658768009188746,
+      "grad_norm": 0.090797059237957,
+      "learning_rate": 2.9062805076472502e-05,
+      "loss": 10.5962,
+      "step": 6740
+    },
+    {
+      "epoch": 0.033708706834128195,
+      "grad_norm": 0.09465719759464264,
+      "learning_rate": 2.9061303161530952e-05,
+      "loss": 10.5961,
+      "step": 6750
+    },
+    {
+      "epoch": 0.033758645659067645,
+      "grad_norm": 0.09611517190933228,
+      "learning_rate": 2.9059801246589402e-05,
+      "loss": 10.5927,
+      "step": 6760
+    },
+    {
+      "epoch": 0.033808584484007094,
+      "grad_norm": 0.10133732110261917,
+      "learning_rate": 2.9058299331647853e-05,
+      "loss": 10.5945,
+      "step": 6770
+    },
+    {
+      "epoch": 0.03385852330894654,
+      "grad_norm": 0.09177147597074509,
+      "learning_rate": 2.90567974167063e-05,
+      "loss": 10.5955,
+      "step": 6780
+    },
+    {
+      "epoch": 0.03390846213388599,
+      "grad_norm": 0.09687118232250214,
+      "learning_rate": 2.905529550176475e-05,
+      "loss": 10.5931,
+      "step": 6790
+    },
+    {
+      "epoch": 0.033958400958825435,
+      "grad_norm": 0.09665414690971375,
+      "learning_rate": 2.90537935868232e-05,
+      "loss": 10.5954,
+      "step": 6800
+    },
+    {
+      "epoch": 0.034008339783764885,
+      "grad_norm": 0.09662660956382751,
+      "learning_rate": 2.905229167188165e-05,
+      "loss": 10.5924,
+      "step": 6810
+    },
+    {
+      "epoch": 0.034058278608704334,
+      "grad_norm": 0.09507598727941513,
+      "learning_rate": 2.90507897569401e-05,
+      "loss": 10.5926,
+      "step": 6820
+    },
+    {
+      "epoch": 0.034108217433643784,
+      "grad_norm": 0.09256746619939804,
+      "learning_rate": 2.9049287841998547e-05,
+      "loss": 10.5938,
+      "step": 6830
+    },
+    {
+      "epoch": 0.03415815625858323,
+      "grad_norm": 0.09468525648117065,
+      "learning_rate": 2.9047785927056997e-05,
+      "loss": 10.5927,
+      "step": 6840
+    },
+    {
+      "epoch": 0.03420809508352268,
+      "grad_norm": 0.09714210778474808,
+      "learning_rate": 2.9046284012115447e-05,
+      "loss": 10.5932,
+      "step": 6850
+    },
+    {
+      "epoch": 0.03425803390846213,
+      "grad_norm": 0.09986428171396255,
+      "learning_rate": 2.9044782097173897e-05,
+      "loss": 10.5918,
+      "step": 6860
+    },
+    {
+      "epoch": 0.03430797273340158,
+      "grad_norm": 0.09874702244997025,
+      "learning_rate": 2.9043280182232348e-05,
+      "loss": 10.5937,
+      "step": 6870
+    },
+    {
+      "epoch": 0.03435791155834103,
+      "grad_norm": 0.09810672700405121,
+      "learning_rate": 2.9041778267290794e-05,
+      "loss": 10.591,
+      "step": 6880
+    },
+    {
+      "epoch": 0.03440785038328048,
+      "grad_norm": 0.09897971153259277,
+      "learning_rate": 2.9040276352349248e-05,
+      "loss": 10.5932,
+      "step": 6890
+    },
+    {
+      "epoch": 0.03445778920821993,
+      "grad_norm": 0.09357500821352005,
+      "learning_rate": 2.9038774437407695e-05,
+      "loss": 10.5916,
+      "step": 6900
+    },
+    {
+      "epoch": 0.03450772803315938,
+      "grad_norm": 0.09426924586296082,
+      "learning_rate": 2.9037272522466145e-05,
+      "loss": 10.5915,
+      "step": 6910
+    },
+    {
+      "epoch": 0.03455766685809883,
+      "grad_norm": 0.09487421810626984,
+      "learning_rate": 2.9035770607524595e-05,
+      "loss": 10.5889,
+      "step": 6920
+    },
+    {
+      "epoch": 0.03460760568303828,
+      "grad_norm": 0.08992663025856018,
+      "learning_rate": 2.9034268692583042e-05,
+      "loss": 10.5891,
+      "step": 6930
+    },
+    {
+      "epoch": 0.03465754450797773,
+      "grad_norm": 0.0984937772154808,
+      "learning_rate": 2.9032766777641495e-05,
+      "loss": 10.5908,
+      "step": 6940
+    },
+    {
+      "epoch": 0.03470748333291718,
+      "grad_norm": 0.08962874859571457,
+      "learning_rate": 2.9031264862699942e-05,
+      "loss": 10.59,
+      "step": 6950
+    },
+    {
+      "epoch": 0.03475742215785663,
+      "grad_norm": 0.09817676246166229,
+      "learning_rate": 2.9029762947758392e-05,
+      "loss": 10.5901,
+      "step": 6960
+    },
+    {
+      "epoch": 0.034807360982796076,
+      "grad_norm": 0.09869422018527985,
+      "learning_rate": 2.9028261032816843e-05,
+      "loss": 10.589,
+      "step": 6970
+    },
+    {
+      "epoch": 0.034857299807735526,
+      "grad_norm": 0.09235729277133942,
+      "learning_rate": 2.9026759117875293e-05,
+      "loss": 10.588,
+      "step": 6980
+    },
+    {
+      "epoch": 0.034907238632674975,
+      "grad_norm": 0.0891089141368866,
+      "learning_rate": 2.9025257202933743e-05,
+      "loss": 16.5832,
+      "step": 6990
+    },
+    {
+      "epoch": 0.034957177457614425,
+      "grad_norm": 0.0962633341550827,
+      "learning_rate": 2.902375528799219e-05,
+      "loss": 22.1721,
+      "step": 7000
+    },
+    {
+      "epoch": 0.035007116282553874,
+      "grad_norm": 0.09975840896368027,
+      "learning_rate": 2.902225337305064e-05,
+      "loss": 10.604,
+      "step": 7010
+    },
+    {
+      "epoch": 0.03505705510749332,
+      "grad_norm": 0.09406078606843948,
+      "learning_rate": 2.902075145810909e-05,
+      "loss": 12.82,
+      "step": 7020
+    },
+    {
+      "epoch": 0.03510699393243277,
+      "grad_norm": 0.09256692975759506,
+      "learning_rate": 2.901924954316754e-05,
+      "loss": 10.5856,
+      "step": 7030
+    },
+    {
+      "epoch": 0.03515693275737222,
+      "grad_norm": 0.09028827399015427,
+      "learning_rate": 2.901774762822599e-05,
+      "loss": 10.5869,
+      "step": 7040
+    },
+    {
+      "epoch": 0.035206871582311665,
+      "grad_norm": 0.0957457572221756,
+      "learning_rate": 2.9016245713284437e-05,
+      "loss": 10.5873,
+      "step": 7050
+    },
+    {
+      "epoch": 0.035256810407251114,
+      "grad_norm": 0.09316723793745041,
+      "learning_rate": 2.9014743798342887e-05,
+      "loss": 10.5875,
+      "step": 7060
+    },
+    {
+      "epoch": 0.035306749232190564,
+      "grad_norm": 0.09424072504043579,
+      "learning_rate": 2.9013241883401338e-05,
+      "loss": 10.5852,
+      "step": 7070
+    },
+    {
+      "epoch": 0.03535668805713001,
+      "grad_norm": 0.09654801338911057,
+      "learning_rate": 2.9011739968459788e-05,
+      "loss": 12.5143,
+      "step": 7080
+    },
+    {
+      "epoch": 0.03540662688206946,
+      "grad_norm": 0.10188226401805878,
+      "learning_rate": 2.9010238053518238e-05,
+      "loss": 10.5845,
+      "step": 7090
+    },
+    {
+      "epoch": 0.03545656570700891,
+      "grad_norm": 0.0997641459107399,
+      "learning_rate": 2.9008736138576685e-05,
+      "loss": 12.043,
+      "step": 7100
+    },
+    {
+      "epoch": 0.03550650453194836,
+      "grad_norm": 0.09502726793289185,
+      "learning_rate": 2.9007234223635135e-05,
+      "loss": 10.5843,
+      "step": 7110
+    },
+    {
+      "epoch": 0.03555644335688781,
+      "grad_norm": 0.09691105782985687,
+      "learning_rate": 2.9005732308693585e-05,
+      "loss": 10.5821,
+      "step": 7120
+    },
+    {
+      "epoch": 0.03560638218182726,
+      "grad_norm": 0.09250783175230026,
+      "learning_rate": 2.9004230393752035e-05,
+      "loss": 10.5865,
+      "step": 7130
+    },
+    {
+      "epoch": 0.03565632100676671,
+      "grad_norm": 0.09487821161746979,
+      "learning_rate": 2.9002728478810486e-05,
+      "loss": 10.5849,
+      "step": 7140
+    },
+    {
+      "epoch": 0.03570625983170616,
+      "grad_norm": 3425.232666015625,
+      "learning_rate": 2.9001226563868932e-05,
+      "loss": 10.598,
+      "step": 7150
+    },
+    {
+      "epoch": 0.03575619865664561,
+      "grad_norm": 0.10075180232524872,
+      "learning_rate": 2.8999724648927382e-05,
+      "loss": 10.5865,
+      "step": 7160
+    },
+    {
+      "epoch": 0.03580613748158506,
+      "grad_norm": 0.09755649417638779,
+      "learning_rate": 2.8998222733985833e-05,
+      "loss": 10.5843,
+      "step": 7170
+    },
+    {
+      "epoch": 0.03585607630652451,
+      "grad_norm": 0.09459280222654343,
+      "learning_rate": 2.8996720819044283e-05,
+      "loss": 10.5831,
+      "step": 7180
+    },
+    {
+      "epoch": 0.03590601513146396,
+      "grad_norm": 0.09937462210655212,
+      "learning_rate": 2.8995218904102733e-05,
+      "loss": 10.5837,
+      "step": 7190
+    },
+    {
+      "epoch": 0.03595595395640341,
+      "grad_norm": 0.09443425387144089,
+      "learning_rate": 2.899371698916118e-05,
+      "loss": 10.5812,
+      "step": 7200
+    },
+    {
+      "epoch": 0.036005892781342856,
+      "grad_norm": 0.0933324322104454,
+      "learning_rate": 2.8992215074219633e-05,
+      "loss": 10.5829,
+      "step": 7210
+    },
+    {
+      "epoch": 0.036055831606282306,
+      "grad_norm": 0.09167325496673584,
+      "learning_rate": 2.899071315927808e-05,
+      "loss": 10.5823,
+      "step": 7220
+    },
+    {
+      "epoch": 0.036105770431221755,
+      "grad_norm": 0.09948615729808807,
+      "learning_rate": 2.898921124433653e-05,
+      "loss": 10.5817,
+      "step": 7230
+    },
+    {
+      "epoch": 0.036155709256161205,
+      "grad_norm": 0.09661941975355148,
+      "learning_rate": 2.898770932939498e-05,
+      "loss": 10.58,
+      "step": 7240
+    },
+    {
+      "epoch": 0.036205648081100654,
+      "grad_norm": 0.09533500671386719,
+      "learning_rate": 2.8986207414453427e-05,
+      "loss": 10.5822,
+      "step": 7250
+    },
+    {
+      "epoch": 0.0362555869060401,
+      "grad_norm": 0.09911171346902847,
+      "learning_rate": 2.898470549951188e-05,
+      "loss": 10.5812,
+      "step": 7260
+    },
+    {
+      "epoch": 0.03630552573097955,
+      "grad_norm": 0.09520751982927322,
+      "learning_rate": 2.8983203584570328e-05,
+      "loss": 10.5781,
+      "step": 7270
+    },
+    {
+      "epoch": 0.036355464555919,
+      "grad_norm": 0.09380774945020676,
+      "learning_rate": 2.8981701669628778e-05,
+      "loss": 10.5798,
+      "step": 7280
+    },
+    {
+      "epoch": 0.03640540338085845,
+      "grad_norm": 0.09551588445901871,
+      "learning_rate": 2.8980199754687228e-05,
+      "loss": 10.5789,
+      "step": 7290
+    },
+    {
+      "epoch": 0.036455342205797894,
+      "grad_norm": 0.09562008827924728,
+      "learning_rate": 2.8978697839745675e-05,
+      "loss": 10.5788,
+      "step": 7300
+    },
+    {
+      "epoch": 0.036505281030737344,
+      "grad_norm": 0.09640656411647797,
+      "learning_rate": 2.897719592480413e-05,
+      "loss": 10.5793,
+      "step": 7310
+    },
+    {
+      "epoch": 0.03655521985567679,
+      "grad_norm": 0.09420552849769592,
+      "learning_rate": 2.8975694009862575e-05,
+      "loss": 10.5776,
+      "step": 7320
+    },
+    {
+      "epoch": 0.03660515868061624,
+      "grad_norm": 0.09690862894058228,
+      "learning_rate": 2.8974192094921025e-05,
+      "loss": 10.577,
+      "step": 7330
+    },
+    {
+      "epoch": 0.03665509750555569,
+      "grad_norm": 0.09202788025140762,
+      "learning_rate": 2.8972690179979476e-05,
+      "loss": 10.5785,
+      "step": 7340
+    },
+    {
+      "epoch": 0.03670503633049514,
+      "grad_norm": 0.08765114843845367,
+      "learning_rate": 2.8971188265037922e-05,
+      "loss": 10.5781,
+      "step": 7350
+    },
+    {
+      "epoch": 0.03675497515543459,
+      "grad_norm": 0.09262295067310333,
+      "learning_rate": 2.8969686350096376e-05,
+      "loss": 10.5772,
+      "step": 7360
+    },
+    {
+      "epoch": 0.03680491398037404,
+      "grad_norm": 0.09478264302015305,
+      "learning_rate": 2.8968184435154823e-05,
+      "loss": 10.5783,
+      "step": 7370
+    },
+    {
+      "epoch": 0.03685485280531349,
+      "grad_norm": 0.09246762841939926,
+      "learning_rate": 2.8966682520213273e-05,
+      "loss": 10.5775,
+      "step": 7380
+    },
+    {
+      "epoch": 0.03690479163025294,
+      "grad_norm": 0.0920056626200676,
+      "learning_rate": 2.8965180605271723e-05,
+      "loss": 10.5771,
+      "step": 7390
+    },
+    {
+      "epoch": 0.03695473045519239,
+      "grad_norm": 0.09541013091802597,
+      "learning_rate": 2.896367869033017e-05,
+      "loss": 10.5757,
+      "step": 7400
+    },
+    {
+      "epoch": 0.03700466928013184,
+      "grad_norm": 0.0953700914978981,
+      "learning_rate": 2.8962176775388623e-05,
+      "loss": 10.5765,
+      "step": 7410
+    },
+    {
+      "epoch": 0.03705460810507129,
+      "grad_norm": 0.09888526797294617,
+      "learning_rate": 2.896067486044707e-05,
+      "loss": 10.5757,
+      "step": 7420
+    },
+    {
+      "epoch": 0.03710454693001074,
+      "grad_norm": 0.09751054644584656,
+      "learning_rate": 2.895917294550552e-05,
+      "loss": 10.5733,
+      "step": 7430
+    },
+    {
+      "epoch": 0.03715448575495019,
+      "grad_norm": 0.095875583589077,
+      "learning_rate": 2.895767103056397e-05,
+      "loss": 10.5762,
+      "step": 7440
+    },
+    {
+      "epoch": 0.037204424579889636,
+      "grad_norm": 0.0908399224281311,
+      "learning_rate": 2.8956169115622417e-05,
+      "loss": 10.5755,
+      "step": 7450
+    },
+    {
+      "epoch": 0.037254363404829086,
+      "grad_norm": 0.09418118000030518,
+      "learning_rate": 2.895466720068087e-05,
+      "loss": 10.5738,
+      "step": 7460
+    },
+    {
+      "epoch": 0.037304302229768535,
+      "grad_norm": 0.09714969247579575,
+      "learning_rate": 2.8953165285739318e-05,
+      "loss": 10.5742,
+      "step": 7470
+    },
+    {
+      "epoch": 0.037354241054707985,
+      "grad_norm": 0.09826507419347763,
+      "learning_rate": 2.8951663370797768e-05,
+      "loss": 10.5731,
+      "step": 7480
+    },
+    {
+      "epoch": 0.037404179879647434,
+      "grad_norm": 0.0948546826839447,
+      "learning_rate": 2.8950161455856218e-05,
+      "loss": 10.5723,
+      "step": 7490
+    },
+    {
+      "epoch": 0.03745411870458688,
+      "grad_norm": 0.09687359631061554,
+      "learning_rate": 2.8948659540914665e-05,
+      "loss": 10.5733,
+      "step": 7500
+    },
+    {
+      "epoch": 0.03750405752952633,
+      "grad_norm": 0.10149165242910385,
+      "learning_rate": 2.894715762597312e-05,
+      "loss": 10.5732,
+      "step": 7510
+    },
+    {
+      "epoch": 0.03755399635446578,
+      "grad_norm": 0.09481348842382431,
+      "learning_rate": 2.8945655711031565e-05,
+      "loss": 10.5732,
+      "step": 7520
+    },
+    {
+      "epoch": 0.03760393517940523,
+      "grad_norm": 0.09526333957910538,
+      "learning_rate": 2.894415379609002e-05,
+      "loss": 10.5736,
+      "step": 7530
+    },
+    {
+      "epoch": 0.03765387400434468,
+      "grad_norm": 0.09095079451799393,
+      "learning_rate": 2.8942651881148466e-05,
+      "loss": 10.5709,
+      "step": 7540
+    },
+    {
+      "epoch": 0.037703812829284124,
+      "grad_norm": 0.09532736986875534,
+      "learning_rate": 2.8941149966206912e-05,
+      "loss": 10.5712,
+      "step": 7550
+    },
+    {
+      "epoch": 0.03775375165422357,
+      "grad_norm": 0.09308824688196182,
+      "learning_rate": 2.8939648051265366e-05,
+      "loss": 10.5711,
+      "step": 7560
+    },
+    {
+      "epoch": 0.03780369047916302,
+      "grad_norm": 0.0935669094324112,
+      "learning_rate": 2.8938146136323813e-05,
+      "loss": 10.5715,
+      "step": 7570
+    },
+    {
+      "epoch": 0.03785362930410247,
+      "grad_norm": 0.09036646783351898,
+      "learning_rate": 2.8936644221382266e-05,
+      "loss": 10.5704,
+      "step": 7580
+    },
+    {
+      "epoch": 0.03790356812904192,
+      "grad_norm": 0.09510631859302521,
+      "learning_rate": 2.8935142306440713e-05,
+      "loss": 10.5709,
+      "step": 7590
+    },
+    {
+      "epoch": 0.03795350695398137,
+      "grad_norm": 0.09474785625934601,
+      "learning_rate": 2.893364039149916e-05,
+      "loss": 10.571,
+      "step": 7600
+    },
+    {
+      "epoch": 0.03800344577892082,
+      "grad_norm": 0.09734784811735153,
+      "learning_rate": 2.8932138476557613e-05,
+      "loss": 10.5693,
+      "step": 7610
+    },
+    {
+      "epoch": 0.03805338460386027,
+      "grad_norm": 0.10142108052968979,
+      "learning_rate": 2.893063656161606e-05,
+      "loss": 10.5701,
+      "step": 7620
+    },
+    {
+      "epoch": 0.03810332342879972,
+      "grad_norm": 0.09453453868627548,
+      "learning_rate": 2.8929134646674514e-05,
+      "loss": 10.5697,
+      "step": 7630
+    },
+    {
+      "epoch": 0.03815326225373917,
+      "grad_norm": 0.09844522923231125,
+      "learning_rate": 2.892763273173296e-05,
+      "loss": 10.5688,
+      "step": 7640
+    },
+    {
+      "epoch": 0.03820320107867862,
+      "grad_norm": 0.10033461451530457,
+      "learning_rate": 2.8926130816791407e-05,
+      "loss": 10.5689,
+      "step": 7650
+    },
+    {
+      "epoch": 0.03825313990361807,
+      "grad_norm": 0.10299541801214218,
+      "learning_rate": 2.892462890184986e-05,
+      "loss": 10.5693,
+      "step": 7660
+    },
+    {
+      "epoch": 0.03830307872855752,
+      "grad_norm": 0.09430062025785446,
+      "learning_rate": 2.8923126986908308e-05,
+      "loss": 10.5672,
+      "step": 7670
+    },
+    {
+      "epoch": 0.03835301755349697,
+      "grad_norm": 0.10111190378665924,
+      "learning_rate": 2.892162507196676e-05,
+      "loss": 10.5674,
+      "step": 7680
+    },
+    {
+      "epoch": 0.038402956378436416,
+      "grad_norm": 0.09100598096847534,
+      "learning_rate": 2.8920123157025208e-05,
+      "loss": 10.5685,
+      "step": 7690
+    },
+    {
+      "epoch": 0.038452895203375866,
+      "grad_norm": 0.09059002995491028,
+      "learning_rate": 2.8918621242083655e-05,
+      "loss": 10.5676,
+      "step": 7700
+    },
+    {
+      "epoch": 0.038502834028315315,
+      "grad_norm": 0.09399350732564926,
+      "learning_rate": 2.891711932714211e-05,
+      "loss": 10.568,
+      "step": 7710
+    },
+    {
+      "epoch": 0.038552772853254764,
+      "grad_norm": 0.09312798827886581,
+      "learning_rate": 2.8915617412200555e-05,
+      "loss": 10.5672,
+      "step": 7720
+    },
+    {
+      "epoch": 0.038602711678194214,
+      "grad_norm": 0.0949626937508583,
+      "learning_rate": 2.891411549725901e-05,
+      "loss": 10.5663,
+      "step": 7730
+    },
+    {
+      "epoch": 0.03865265050313366,
+      "grad_norm": 0.08603295683860779,
+      "learning_rate": 2.8912613582317456e-05,
+      "loss": 10.5665,
+      "step": 7740
+    },
+    {
+      "epoch": 0.03870258932807311,
+      "grad_norm": 0.09113872051239014,
+      "learning_rate": 2.8911111667375902e-05,
+      "loss": 10.567,
+      "step": 7750
+    },
+    {
+      "epoch": 0.03875252815301256,
+      "grad_norm": 0.09675870090723038,
+      "learning_rate": 2.8909609752434356e-05,
+      "loss": 10.5649,
+      "step": 7760
+    },
+    {
+      "epoch": 0.03880246697795201,
+      "grad_norm": 0.09332741051912308,
+      "learning_rate": 2.8908107837492803e-05,
+      "loss": 10.5656,
+      "step": 7770
+    },
+    {
+      "epoch": 0.03885240580289146,
+      "grad_norm": 0.09331821650266647,
+      "learning_rate": 2.8906605922551256e-05,
+      "loss": 10.5659,
+      "step": 7780
+    },
+    {
+      "epoch": 0.038902344627830904,
+      "grad_norm": 0.09395716339349747,
+      "learning_rate": 2.8905104007609703e-05,
+      "loss": 10.5656,
+      "step": 7790
+    },
+    {
+      "epoch": 0.03895228345277035,
+      "grad_norm": 0.09877283871173859,
+      "learning_rate": 2.890360209266815e-05,
+      "loss": 10.5653,
+      "step": 7800
+    },
+    {
+      "epoch": 0.0390022222777098,
+      "grad_norm": 0.09474792331457138,
+      "learning_rate": 2.8902100177726603e-05,
+      "loss": 10.5637,
+      "step": 7810
+    },
+    {
+      "epoch": 0.03905216110264925,
+      "grad_norm": 0.09844552725553513,
+      "learning_rate": 2.890059826278505e-05,
+      "loss": 10.5633,
+      "step": 7820
+    },
+    {
+      "epoch": 0.0391020999275887,
+      "grad_norm": 0.09523854404687881,
+      "learning_rate": 2.8899096347843504e-05,
+      "loss": 10.5637,
+      "step": 7830
+    },
+    {
+      "epoch": 0.03915203875252815,
+      "grad_norm": 0.09699324518442154,
+      "learning_rate": 2.889759443290195e-05,
+      "loss": 10.5631,
+      "step": 7840
+    },
+    {
+      "epoch": 0.0392019775774676,
+      "grad_norm": 0.09227604418992996,
+      "learning_rate": 2.88960925179604e-05,
+      "loss": 10.5623,
+      "step": 7850
+    },
+    {
+      "epoch": 0.03925191640240705,
+      "grad_norm": 0.0924190804362297,
+      "learning_rate": 2.889459060301885e-05,
+      "loss": 10.5623,
+      "step": 7860
+    },
+    {
+      "epoch": 0.0393018552273465,
+      "grad_norm": 0.10132279247045517,
+      "learning_rate": 2.8893088688077298e-05,
+      "loss": 10.5635,
+      "step": 7870
+    },
+    {
+      "epoch": 0.03935179405228595,
+      "grad_norm": 0.09170647710561752,
+      "learning_rate": 2.889158677313575e-05,
+      "loss": 10.5616,
+      "step": 7880
+    },
+    {
+      "epoch": 0.0394017328772254,
+      "grad_norm": 0.09620120376348495,
+      "learning_rate": 2.8890084858194198e-05,
+      "loss": 10.5616,
+      "step": 7890
+    },
+    {
+      "epoch": 0.03945167170216485,
+      "grad_norm": 0.09894801676273346,
+      "learning_rate": 2.8888582943252648e-05,
+      "loss": 10.5605,
+      "step": 7900
+    },
+    {
+      "epoch": 0.0395016105271043,
+      "grad_norm": 0.09626492857933044,
+      "learning_rate": 2.88870810283111e-05,
+      "loss": 10.5602,
+      "step": 7910
+    },
+    {
+      "epoch": 0.03955154935204375,
+      "grad_norm": 0.09040050953626633,
+      "learning_rate": 2.8885579113369545e-05,
+      "loss": 10.5595,
+      "step": 7920
+    },
+    {
+      "epoch": 0.039601488176983196,
+      "grad_norm": 0.09787009656429291,
+      "learning_rate": 2.8884077198428e-05,
+      "loss": 10.5598,
+      "step": 7930
+    },
+    {
+      "epoch": 0.039651427001922646,
+      "grad_norm": 0.09217660874128342,
+      "learning_rate": 2.8882575283486446e-05,
+      "loss": 10.5596,
+      "step": 7940
+    },
+    {
+      "epoch": 0.039701365826862095,
+      "grad_norm": 0.09433452039957047,
+      "learning_rate": 2.8881073368544896e-05,
+      "loss": 10.56,
+      "step": 7950
+    },
+    {
+      "epoch": 0.039751304651801544,
+      "grad_norm": 0.09549769759178162,
+      "learning_rate": 2.8879571453603346e-05,
+      "loss": 10.5576,
+      "step": 7960
+    },
+    {
+      "epoch": 0.039801243476740994,
+      "grad_norm": 0.09538113325834274,
+      "learning_rate": 2.8878069538661793e-05,
+      "loss": 10.56,
+      "step": 7970
+    },
+    {
+      "epoch": 0.03985118230168044,
+      "grad_norm": 0.09417406469583511,
+      "learning_rate": 2.8876567623720246e-05,
+      "loss": 10.5598,
+      "step": 7980
+    },
+    {
+      "epoch": 0.03990112112661989,
+      "grad_norm": 0.09565738588571548,
+      "learning_rate": 2.8875065708778693e-05,
+      "loss": 10.5599,
+      "step": 7990
+    },
+    {
+      "epoch": 0.03995105995155934,
+      "grad_norm": 0.09348537772893906,
+      "learning_rate": 2.8873563793837143e-05,
+      "loss": 10.5605,
+      "step": 8000
+    },
+    {
+      "epoch": 0.04000099877649879,
+      "grad_norm": 0.0924496129155159,
+      "learning_rate": 2.8872061878895593e-05,
+      "loss": 10.5592,
+      "step": 8010
+    },
+    {
+      "epoch": 0.04005093760143824,
+      "grad_norm": 0.09314726293087006,
+      "learning_rate": 2.887055996395404e-05,
+      "loss": 10.5568,
+      "step": 8020
+    },
+    {
+      "epoch": 0.04010087642637769,
+      "grad_norm": 0.09801434725522995,
+      "learning_rate": 2.8869058049012494e-05,
+      "loss": 10.558,
+      "step": 8030
+    },
+    {
+      "epoch": 0.04015081525131713,
+      "grad_norm": 0.09441787749528885,
+      "learning_rate": 2.886755613407094e-05,
+      "loss": 10.5562,
+      "step": 8040
+    },
+    {
+      "epoch": 0.04020075407625658,
+      "grad_norm": 0.09265992045402527,
+      "learning_rate": 2.886605421912939e-05,
+      "loss": 10.5588,
+      "step": 8050
+    },
+    {
+      "epoch": 0.04025069290119603,
+      "grad_norm": 0.09449957311153412,
+      "learning_rate": 2.886455230418784e-05,
+      "loss": 10.5578,
+      "step": 8060
+    },
+    {
+      "epoch": 0.04030063172613548,
+      "grad_norm": 0.09760826081037521,
+      "learning_rate": 2.8863050389246288e-05,
+      "loss": 10.5543,
+      "step": 8070
+    },
+    {
+      "epoch": 0.04035057055107493,
+      "grad_norm": 0.09547002613544464,
+      "learning_rate": 2.886154847430474e-05,
+      "loss": 10.5575,
+      "step": 8080
+    },
+    {
+      "epoch": 0.04040050937601438,
+      "grad_norm": 0.09818068146705627,
+      "learning_rate": 2.8860046559363188e-05,
+      "loss": 10.5569,
+      "step": 8090
+    },
+    {
+      "epoch": 0.04045044820095383,
+      "grad_norm": 0.09117206931114197,
+      "learning_rate": 2.8858544644421638e-05,
+      "loss": 10.5529,
+      "step": 8100
+    },
+    {
+      "epoch": 0.04050038702589328,
+      "grad_norm": 0.10111954063177109,
+      "learning_rate": 2.885704272948009e-05,
+      "loss": 10.5541,
+      "step": 8110
+    },
+    {
+      "epoch": 0.04055032585083273,
+      "grad_norm": 0.09272032976150513,
+      "learning_rate": 2.8855540814538535e-05,
+      "loss": 10.5528,
+      "step": 8120
+    },
+    {
+      "epoch": 0.04060026467577218,
+      "grad_norm": 0.09511199593544006,
+      "learning_rate": 2.885403889959699e-05,
+      "loss": 10.5561,
+      "step": 8130
+    },
+    {
+      "epoch": 0.04065020350071163,
+      "grad_norm": 0.09244706481695175,
+      "learning_rate": 2.8852536984655436e-05,
+      "loss": 10.5553,
+      "step": 8140
+    },
+    {
+      "epoch": 0.04070014232565108,
+      "grad_norm": 0.09505173563957214,
+      "learning_rate": 2.8851035069713886e-05,
+      "loss": 10.5548,
+      "step": 8150
+    },
+    {
+      "epoch": 0.04075008115059053,
+      "grad_norm": 0.09145089983940125,
+      "learning_rate": 2.8849533154772336e-05,
+      "loss": 10.5544,
+      "step": 8160
+    },
+    {
+      "epoch": 0.040800019975529976,
+      "grad_norm": 0.09063593298196793,
+      "learning_rate": 2.8848031239830786e-05,
+      "loss": 10.5549,
+      "step": 8170
+    },
+    {
+      "epoch": 0.040849958800469426,
+      "grad_norm": 0.09323187917470932,
+      "learning_rate": 2.8846529324889236e-05,
+      "loss": 10.5522,
+      "step": 8180
+    },
+    {
+      "epoch": 0.040899897625408875,
+      "grad_norm": 0.09502808004617691,
+      "learning_rate": 2.8845027409947683e-05,
+      "loss": 10.5537,
+      "step": 8190
+    },
+    {
+      "epoch": 0.040949836450348324,
+      "grad_norm": 0.10582366585731506,
+      "learning_rate": 2.8843525495006133e-05,
+      "loss": 10.5538,
+      "step": 8200
+    },
+    {
+      "epoch": 0.040999775275287774,
+      "grad_norm": 0.09141421318054199,
+      "learning_rate": 2.8842023580064583e-05,
+      "loss": 10.5523,
+      "step": 8210
+    },
+    {
+      "epoch": 0.04104971410022722,
+      "grad_norm": 0.10094629228115082,
+      "learning_rate": 2.8840521665123034e-05,
+      "loss": 10.5521,
+      "step": 8220
+    },
+    {
+      "epoch": 0.04109965292516667,
+      "grad_norm": 0.08866948634386063,
+      "learning_rate": 2.8839019750181484e-05,
+      "loss": 10.5503,
+      "step": 8230
+    },
+    {
+      "epoch": 0.04114959175010612,
+      "grad_norm": 0.09741657972335815,
+      "learning_rate": 2.883751783523993e-05,
+      "loss": 10.5508,
+      "step": 8240
+    },
+    {
+      "epoch": 0.04119953057504557,
+      "grad_norm": 0.09136499464511871,
+      "learning_rate": 2.883601592029838e-05,
+      "loss": 10.5507,
+      "step": 8250
+    },
+    {
+      "epoch": 0.04124946939998502,
+      "grad_norm": 0.09198024123907089,
+      "learning_rate": 2.883451400535683e-05,
+      "loss": 10.5514,
+      "step": 8260
+    },
+    {
+      "epoch": 0.04129940822492447,
+      "grad_norm": 0.09833436459302902,
+      "learning_rate": 2.883301209041528e-05,
+      "loss": 10.5492,
+      "step": 8270
+    },
+    {
+      "epoch": 0.04134934704986392,
+      "grad_norm": 0.09501312673091888,
+      "learning_rate": 2.883151017547373e-05,
+      "loss": 10.5505,
+      "step": 8280
+    },
+    {
+      "epoch": 0.04139928587480336,
+      "grad_norm": 0.08871445059776306,
+      "learning_rate": 2.8830008260532178e-05,
+      "loss": 10.5488,
+      "step": 8290
+    },
+    {
+      "epoch": 0.04144922469974281,
+      "grad_norm": 0.09673582017421722,
+      "learning_rate": 2.8828506345590628e-05,
+      "loss": 10.55,
+      "step": 8300
+    },
+    {
+      "epoch": 0.04149916352468226,
+      "grad_norm": 0.09481798112392426,
+      "learning_rate": 2.882700443064908e-05,
+      "loss": 10.5475,
+      "step": 8310
+    },
+    {
+      "epoch": 0.04154910234962171,
+      "grad_norm": 0.09376543760299683,
+      "learning_rate": 2.882550251570753e-05,
+      "loss": 10.5485,
+      "step": 8320
+    },
+    {
+      "epoch": 0.04159904117456116,
+      "grad_norm": 0.09397636353969574,
+      "learning_rate": 2.882400060076598e-05,
+      "loss": 10.5472,
+      "step": 8330
+    },
+    {
+      "epoch": 0.04164897999950061,
+      "grad_norm": 0.09375918656587601,
+      "learning_rate": 2.8822498685824426e-05,
+      "loss": 10.5469,
+      "step": 8340
+    },
+    {
+      "epoch": 0.04169891882444006,
+      "grad_norm": 0.09085606038570404,
+      "learning_rate": 2.8820996770882876e-05,
+      "loss": 10.5496,
+      "step": 8350
+    },
+    {
+      "epoch": 0.04174885764937951,
+      "grad_norm": 0.0982041284441948,
+      "learning_rate": 2.8819494855941326e-05,
+      "loss": 10.5468,
+      "step": 8360
+    },
+    {
+      "epoch": 0.04179879647431896,
+      "grad_norm": 0.09510507434606552,
+      "learning_rate": 2.8817992940999776e-05,
+      "loss": 10.5458,
+      "step": 8370
+    },
+    {
+      "epoch": 0.04184873529925841,
+      "grad_norm": 0.09728366136550903,
+      "learning_rate": 2.8816491026058226e-05,
+      "loss": 10.5469,
+      "step": 8380
+    },
+    {
+      "epoch": 0.04189867412419786,
+      "grad_norm": 0.10339502990245819,
+      "learning_rate": 2.8814989111116673e-05,
+      "loss": 10.5475,
+      "step": 8390
+    },
+    {
+      "epoch": 0.04194861294913731,
+      "grad_norm": 0.09696449339389801,
+      "learning_rate": 2.8813487196175123e-05,
+      "loss": 10.5457,
+      "step": 8400
+    },
+    {
+      "epoch": 0.041998551774076756,
+      "grad_norm": 0.09432040899991989,
+      "learning_rate": 2.8811985281233573e-05,
+      "loss": 10.5464,
+      "step": 8410
+    },
+    {
+      "epoch": 0.042048490599016206,
+      "grad_norm": 0.09533160924911499,
+      "learning_rate": 2.8810483366292024e-05,
+      "loss": 10.5458,
+      "step": 8420
+    },
+    {
+      "epoch": 0.042098429423955655,
+      "grad_norm": 0.09491016715765,
+      "learning_rate": 2.8808981451350474e-05,
+      "loss": 10.546,
+      "step": 8430
+    },
+    {
+      "epoch": 0.042148368248895104,
+      "grad_norm": 0.0958162397146225,
+      "learning_rate": 2.880747953640892e-05,
+      "loss": 10.5436,
+      "step": 8440
+    },
+    {
+      "epoch": 0.042198307073834554,
+      "grad_norm": 0.10142121464014053,
+      "learning_rate": 2.880597762146737e-05,
+      "loss": 10.5453,
+      "step": 8450
+    },
+    {
+      "epoch": 0.042248245898774,
+      "grad_norm": 0.09626547992229462,
+      "learning_rate": 2.880447570652582e-05,
+      "loss": 10.5443,
+      "step": 8460
+    },
+    {
+      "epoch": 0.04229818472371345,
+      "grad_norm": 0.09666489064693451,
+      "learning_rate": 2.880297379158427e-05,
+      "loss": 10.5445,
+      "step": 8470
+    },
+    {
+      "epoch": 0.0423481235486529,
+      "grad_norm": 0.0963977724313736,
+      "learning_rate": 2.880147187664272e-05,
+      "loss": 10.545,
+      "step": 8480
+    },
+    {
+      "epoch": 0.04239806237359235,
+      "grad_norm": 0.09442082047462463,
+      "learning_rate": 2.879996996170117e-05,
+      "loss": 10.5455,
+      "step": 8490
+    },
+    {
+      "epoch": 0.0424480011985318,
+      "grad_norm": 0.09085852652788162,
+      "learning_rate": 2.8798468046759618e-05,
+      "loss": 10.5439,
+      "step": 8500
+    },
+    {
+      "epoch": 0.04249794002347125,
+      "grad_norm": 0.0952218547463417,
+      "learning_rate": 2.879696613181807e-05,
+      "loss": 10.5411,
+      "step": 8510
+    },
+    {
+      "epoch": 0.0425478788484107,
+      "grad_norm": 0.09868868440389633,
+      "learning_rate": 2.879546421687652e-05,
+      "loss": 10.5424,
+      "step": 8520
+    },
+    {
+      "epoch": 0.04259781767335014,
+      "grad_norm": 0.09193531423807144,
+      "learning_rate": 2.879396230193497e-05,
+      "loss": 10.5434,
+      "step": 8530
+    },
+    {
+      "epoch": 0.04264775649828959,
+      "grad_norm": 0.09298193454742432,
+      "learning_rate": 2.879246038699342e-05,
+      "loss": 10.5418,
+      "step": 8540
+    },
+    {
+      "epoch": 0.04269769532322904,
+      "grad_norm": 0.09158916771411896,
+      "learning_rate": 2.8790958472051866e-05,
+      "loss": 10.5422,
+      "step": 8550
+    },
+    {
+      "epoch": 0.04274763414816849,
+      "grad_norm": 0.09237027168273926,
+      "learning_rate": 2.8789456557110316e-05,
+      "loss": 10.5407,
+      "step": 8560
+    },
+    {
+      "epoch": 0.04279757297310794,
+      "grad_norm": 0.1070277988910675,
+      "learning_rate": 2.8787954642168766e-05,
+      "loss": 10.5417,
+      "step": 8570
+    },
+    {
+      "epoch": 0.04284751179804739,
+      "grad_norm": 0.0924166664481163,
+      "learning_rate": 2.8786452727227216e-05,
+      "loss": 10.5391,
+      "step": 8580
+    },
+    {
+      "epoch": 0.04289745062298684,
+      "grad_norm": 0.09547102451324463,
+      "learning_rate": 2.8784950812285667e-05,
+      "loss": 10.5386,
+      "step": 8590
+    },
+    {
+      "epoch": 0.04294738944792629,
+      "grad_norm": 0.09805493801832199,
+      "learning_rate": 2.8783448897344113e-05,
+      "loss": 10.5404,
+      "step": 8600
+    },
+    {
+      "epoch": 0.04299732827286574,
+      "grad_norm": 0.09734487533569336,
+      "learning_rate": 2.8781946982402563e-05,
+      "loss": 10.5402,
+      "step": 8610
+    },
+    {
+      "epoch": 0.04304726709780519,
+      "grad_norm": 0.09356056153774261,
+      "learning_rate": 2.8780445067461014e-05,
+      "loss": 10.5394,
+      "step": 8620
+    },
+    {
+      "epoch": 0.04309720592274464,
+      "grad_norm": 0.09781930595636368,
+      "learning_rate": 2.8778943152519464e-05,
+      "loss": 10.5389,
+      "step": 8630
+    },
+    {
+      "epoch": 0.04314714474768409,
+      "grad_norm": 0.0970260426402092,
+      "learning_rate": 2.8777441237577914e-05,
+      "loss": 10.5402,
+      "step": 8640
+    },
+    {
+      "epoch": 0.043197083572623536,
+      "grad_norm": 0.09671913832426071,
+      "learning_rate": 2.877593932263636e-05,
+      "loss": 10.5387,
+      "step": 8650
+    },
+    {
+      "epoch": 0.043247022397562986,
+      "grad_norm": 0.0935794860124588,
+      "learning_rate": 2.877443740769481e-05,
+      "loss": 10.5378,
+      "step": 8660
+    },
+    {
+      "epoch": 0.043296961222502435,
+      "grad_norm": 0.0933162197470665,
+      "learning_rate": 2.877293549275326e-05,
+      "loss": 10.5399,
+      "step": 8670
+    },
+    {
+      "epoch": 0.043346900047441884,
+      "grad_norm": 0.09513531625270844,
+      "learning_rate": 2.877143357781171e-05,
+      "loss": 10.5379,
+      "step": 8680
+    },
+    {
+      "epoch": 0.043396838872381334,
+      "grad_norm": 0.10161089897155762,
+      "learning_rate": 2.876993166287016e-05,
+      "loss": 10.5386,
+      "step": 8690
+    },
+    {
+      "epoch": 0.04344677769732078,
+      "grad_norm": 0.09620557725429535,
+      "learning_rate": 2.8768429747928608e-05,
+      "loss": 10.5381,
+      "step": 8700
+    },
+    {
+      "epoch": 0.04349671652226023,
+      "grad_norm": 0.09468070417642593,
+      "learning_rate": 2.876692783298706e-05,
+      "loss": 10.5371,
+      "step": 8710
+    },
+    {
+      "epoch": 0.04354665534719968,
+      "grad_norm": 0.09541548788547516,
+      "learning_rate": 2.876542591804551e-05,
+      "loss": 10.5374,
+      "step": 8720
+    },
+    {
+      "epoch": 0.04359659417213913,
+      "grad_norm": 0.09611213952302933,
+      "learning_rate": 2.876392400310396e-05,
+      "loss": 10.5382,
+      "step": 8730
+    },
+    {
+      "epoch": 0.04364653299707858,
+      "grad_norm": 0.08991603553295135,
+      "learning_rate": 2.876242208816241e-05,
+      "loss": 10.536,
+      "step": 8740
+    },
+    {
+      "epoch": 0.04369647182201803,
+      "grad_norm": 0.09493478387594223,
+      "learning_rate": 2.8760920173220856e-05,
+      "loss": 10.5358,
+      "step": 8750
+    },
+    {
+      "epoch": 0.04374641064695748,
+      "grad_norm": 0.09500102698802948,
+      "learning_rate": 2.8759418258279306e-05,
+      "loss": 10.5345,
+      "step": 8760
+    },
+    {
+      "epoch": 0.04379634947189693,
+      "grad_norm": 0.09735292196273804,
+      "learning_rate": 2.8757916343337756e-05,
+      "loss": 10.5329,
+      "step": 8770
+    },
+    {
+      "epoch": 0.04384628829683637,
+      "grad_norm": 0.09669680148363113,
+      "learning_rate": 2.8756414428396206e-05,
+      "loss": 10.5363,
+      "step": 8780
+    },
+    {
+      "epoch": 0.04389622712177582,
+      "grad_norm": 0.09384785592556,
+      "learning_rate": 2.8754912513454657e-05,
+      "loss": 10.5342,
+      "step": 8790
+    },
+    {
+      "epoch": 0.04394616594671527,
+      "grad_norm": 0.09973820298910141,
+      "learning_rate": 2.8753410598513103e-05,
+      "loss": 10.5346,
+      "step": 8800
+    },
+    {
+      "epoch": 0.04399610477165472,
+      "grad_norm": 0.09043479710817337,
+      "learning_rate": 2.8751908683571557e-05,
+      "loss": 10.5345,
+      "step": 8810
+    },
+    {
+      "epoch": 0.04404604359659417,
+      "grad_norm": 0.09604961425065994,
+      "learning_rate": 2.8750406768630004e-05,
+      "loss": 10.5338,
+      "step": 8820
+    },
+    {
+      "epoch": 0.04409598242153362,
+      "grad_norm": 0.09182944893836975,
+      "learning_rate": 2.8748904853688454e-05,
+      "loss": 10.5345,
+      "step": 8830
+    },
+    {
+      "epoch": 0.04414592124647307,
+      "grad_norm": 0.09448856860399246,
+      "learning_rate": 2.8747402938746904e-05,
+      "loss": 10.5337,
+      "step": 8840
+    },
+    {
+      "epoch": 0.04419586007141252,
+      "grad_norm": 0.09963658452033997,
+      "learning_rate": 2.874590102380535e-05,
+      "loss": 10.5313,
+      "step": 8850
+    },
+    {
+      "epoch": 0.04424579889635197,
+      "grad_norm": 0.09832189232110977,
+      "learning_rate": 2.8744399108863804e-05,
+      "loss": 10.5341,
+      "step": 8860
+    },
+    {
+      "epoch": 0.04429573772129142,
+      "grad_norm": 0.0896415188908577,
+      "learning_rate": 2.874289719392225e-05,
+      "loss": 10.5319,
+      "step": 8870
+    },
+    {
+      "epoch": 0.04434567654623087,
+      "grad_norm": 0.09824098646640778,
+      "learning_rate": 2.87413952789807e-05,
+      "loss": 10.5328,
+      "step": 8880
+    },
+    {
+      "epoch": 0.044395615371170316,
+      "grad_norm": 0.09288741648197174,
+      "learning_rate": 2.873989336403915e-05,
+      "loss": 10.5329,
+      "step": 8890
+    },
+    {
+      "epoch": 0.044445554196109766,
+      "grad_norm": 0.09383495151996613,
+      "learning_rate": 2.87383914490976e-05,
+      "loss": 10.5323,
+      "step": 8900
+    },
+    {
+      "epoch": 0.044495493021049215,
+      "grad_norm": 0.09476141631603241,
+      "learning_rate": 2.8736889534156052e-05,
+      "loss": 10.5316,
+      "step": 8910
+    },
+    {
+      "epoch": 0.044545431845988664,
+      "grad_norm": 0.0921645388007164,
+      "learning_rate": 2.87353876192145e-05,
+      "loss": 10.5313,
+      "step": 8920
+    },
+    {
+      "epoch": 0.044595370670928114,
+      "grad_norm": 0.09352537989616394,
+      "learning_rate": 2.873388570427295e-05,
+      "loss": 10.5309,
+      "step": 8930
+    },
+    {
+      "epoch": 0.04464530949586756,
+      "grad_norm": 0.08910872787237167,
+      "learning_rate": 2.87323837893314e-05,
+      "loss": 10.5303,
+      "step": 8940
+    },
+    {
+      "epoch": 0.04469524832080701,
+      "grad_norm": 0.09727400541305542,
+      "learning_rate": 2.8730881874389846e-05,
+      "loss": 10.5306,
+      "step": 8950
+    },
+    {
+      "epoch": 0.04474518714574646,
+      "grad_norm": 0.09605511277914047,
+      "learning_rate": 2.87293799594483e-05,
+      "loss": 10.5287,
+      "step": 8960
+    },
+    {
+      "epoch": 0.04479512597068591,
+      "grad_norm": 0.0909186601638794,
+      "learning_rate": 2.8727878044506746e-05,
+      "loss": 10.5279,
+      "step": 8970
+    },
+    {
+      "epoch": 0.04484506479562536,
+      "grad_norm": 0.0998016744852066,
+      "learning_rate": 2.8726376129565196e-05,
+      "loss": 10.5307,
+      "step": 8980
+    },
+    {
+      "epoch": 0.04489500362056481,
+      "grad_norm": 0.09300219267606735,
+      "learning_rate": 2.8724874214623647e-05,
+      "loss": 10.5304,
+      "step": 8990
+    },
+    {
+      "epoch": 0.04494494244550426,
+      "grad_norm": 0.09390337765216827,
+      "learning_rate": 2.8723372299682093e-05,
+      "loss": 10.5274,
+      "step": 9000
+    },
+    {
+      "epoch": 0.04499488127044371,
+      "grad_norm": 0.09505102783441544,
+      "learning_rate": 2.8721870384740547e-05,
+      "loss": 10.529,
+      "step": 9010
+    },
+    {
+      "epoch": 0.04504482009538316,
+      "grad_norm": 0.09317026287317276,
+      "learning_rate": 2.8720368469798994e-05,
+      "loss": 10.5281,
+      "step": 9020
+    },
+    {
+      "epoch": 0.0450947589203226,
+      "grad_norm": 0.09734094887971878,
+      "learning_rate": 2.8718866554857444e-05,
+      "loss": 10.5281,
+      "step": 9030
+    },
+    {
+      "epoch": 0.04514469774526205,
+      "grad_norm": 0.09334099292755127,
+      "learning_rate": 2.8717364639915894e-05,
+      "loss": 10.5263,
+      "step": 9040
+    },
+    {
+      "epoch": 0.0451946365702015,
+      "grad_norm": 0.0930711030960083,
+      "learning_rate": 2.871586272497434e-05,
+      "loss": 10.526,
+      "step": 9050
+    },
+    {
+      "epoch": 0.04524457539514095,
+      "grad_norm": 0.09420018643140793,
+      "learning_rate": 2.8714360810032794e-05,
+      "loss": 10.527,
+      "step": 9060
+    },
+    {
+      "epoch": 0.0452945142200804,
+      "grad_norm": 0.0920024886727333,
+      "learning_rate": 2.871285889509124e-05,
+      "loss": 10.5282,
+      "step": 9070
+    },
+    {
+      "epoch": 0.04534445304501985,
+      "grad_norm": 0.1000896766781807,
+      "learning_rate": 2.871135698014969e-05,
+      "loss": 10.5276,
+      "step": 9080
+    },
+    {
+      "epoch": 0.0453943918699593,
+      "grad_norm": 0.09131770581007004,
+      "learning_rate": 2.870985506520814e-05,
+      "loss": 10.5263,
+      "step": 9090
+    },
+    {
+      "epoch": 0.04544433069489875,
+      "grad_norm": 0.09568016231060028,
+      "learning_rate": 2.870835315026659e-05,
+      "loss": 10.5241,
+      "step": 9100
+    },
+    {
+      "epoch": 0.0454942695198382,
+      "grad_norm": 0.1060924157500267,
+      "learning_rate": 2.8706851235325042e-05,
+      "loss": 10.5241,
+      "step": 9110
+    },
+    {
+      "epoch": 0.04554420834477765,
+      "grad_norm": 0.10000130534172058,
+      "learning_rate": 2.870534932038349e-05,
+      "loss": 10.5277,
+      "step": 9120
+    },
+    {
+      "epoch": 0.045594147169717096,
+      "grad_norm": 0.09147665649652481,
+      "learning_rate": 2.8703847405441942e-05,
+      "loss": 10.5265,
+      "step": 9130
+    },
+    {
+      "epoch": 0.045644085994656546,
+      "grad_norm": 0.09924129396677017,
+      "learning_rate": 2.870234549050039e-05,
+      "loss": 10.5251,
+      "step": 9140
+    },
+    {
+      "epoch": 0.045694024819595995,
+      "grad_norm": 0.09299519658088684,
+      "learning_rate": 2.8700843575558836e-05,
+      "loss": 10.5242,
+      "step": 9150
+    },
+    {
+      "epoch": 0.045743963644535444,
+      "grad_norm": 0.09638933092355728,
+      "learning_rate": 2.869934166061729e-05,
+      "loss": 10.5243,
+      "step": 9160
+    },
+    {
+      "epoch": 0.045793902469474894,
+      "grad_norm": 0.10021718591451645,
+      "learning_rate": 2.8697839745675736e-05,
+      "loss": 10.5254,
+      "step": 9170
+    },
+    {
+      "epoch": 0.04584384129441434,
+      "grad_norm": 0.09364977478981018,
+      "learning_rate": 2.869633783073419e-05,
+      "loss": 10.524,
+      "step": 9180
+    },
+    {
+      "epoch": 0.04589378011935379,
+      "grad_norm": 0.09493537992238998,
+      "learning_rate": 2.8694835915792637e-05,
+      "loss": 10.524,
+      "step": 9190
+    },
+    {
+      "epoch": 0.04594371894429324,
+      "grad_norm": 0.09197970479726791,
+      "learning_rate": 2.8693334000851083e-05,
+      "loss": 10.5239,
+      "step": 9200
+    },
+    {
+      "epoch": 0.04599365776923269,
+      "grad_norm": 0.09489205479621887,
+      "learning_rate": 2.8691832085909537e-05,
+      "loss": 10.5234,
+      "step": 9210
+    },
+    {
+      "epoch": 0.04604359659417214,
+      "grad_norm": 0.09727194160223007,
+      "learning_rate": 2.8690330170967984e-05,
+      "loss": 10.5224,
+      "step": 9220
+    },
+    {
+      "epoch": 0.04609353541911159,
+      "grad_norm": 0.09218050539493561,
+      "learning_rate": 2.8688828256026437e-05,
+      "loss": 10.5209,
+      "step": 9230
+    },
+    {
+      "epoch": 0.04614347424405104,
+      "grad_norm": 0.08994846791028976,
+      "learning_rate": 2.8687326341084884e-05,
+      "loss": 10.5225,
+      "step": 9240
+    },
+    {
+      "epoch": 0.04619341306899049,
+      "grad_norm": 0.09504954516887665,
+      "learning_rate": 2.868582442614333e-05,
+      "loss": 10.5213,
+      "step": 9250
+    },
+    {
+      "epoch": 0.04624335189392994,
+      "grad_norm": 0.0972573459148407,
+      "learning_rate": 2.8684322511201784e-05,
+      "loss": 10.522,
+      "step": 9260
+    },
+    {
+      "epoch": 0.04629329071886939,
+      "grad_norm": 0.09305988252162933,
+      "learning_rate": 2.868282059626023e-05,
+      "loss": 10.5211,
+      "step": 9270
+    },
+    {
+      "epoch": 0.04634322954380883,
+      "grad_norm": 0.09756896644830704,
+      "learning_rate": 2.8681318681318685e-05,
+      "loss": 10.5218,
+      "step": 9280
+    },
+    {
+      "epoch": 0.04639316836874828,
+      "grad_norm": 0.09515320509672165,
+      "learning_rate": 2.867981676637713e-05,
+      "loss": 10.5219,
+      "step": 9290
+    },
+    {
+      "epoch": 0.04644310719368773,
+      "grad_norm": 0.1009177565574646,
+      "learning_rate": 2.867831485143558e-05,
+      "loss": 10.5203,
+      "step": 9300
+    },
+    {
+      "epoch": 0.04649304601862718,
+      "grad_norm": 0.09644551575183868,
+      "learning_rate": 2.8676812936494032e-05,
+      "loss": 10.5188,
+      "step": 9310
+    },
+    {
+      "epoch": 0.04654298484356663,
+      "grad_norm": 0.09817861765623093,
+      "learning_rate": 2.867531102155248e-05,
+      "loss": 10.5195,
+      "step": 9320
+    },
+    {
+      "epoch": 0.04659292366850608,
+      "grad_norm": 0.09904427081346512,
+      "learning_rate": 2.8673809106610932e-05,
+      "loss": 10.5177,
+      "step": 9330
+    },
+    {
+      "epoch": 0.04664286249344553,
+      "grad_norm": 0.09583260118961334,
+      "learning_rate": 2.867230719166938e-05,
+      "loss": 10.5209,
+      "step": 9340
+    },
+    {
+      "epoch": 0.04669280131838498,
+      "grad_norm": 0.09900553524494171,
+      "learning_rate": 2.8670805276727826e-05,
+      "loss": 10.5185,
+      "step": 9350
+    },
+    {
+      "epoch": 0.04674274014332443,
+      "grad_norm": 0.09579353779554367,
+      "learning_rate": 2.866930336178628e-05,
+      "loss": 10.5189,
+      "step": 9360
+    },
+    {
+      "epoch": 0.046792678968263876,
+      "grad_norm": 0.09370560199022293,
+      "learning_rate": 2.8667801446844726e-05,
+      "loss": 10.5196,
+      "step": 9370
+    },
+    {
+      "epoch": 0.046842617793203326,
+      "grad_norm": 0.09705270081758499,
+      "learning_rate": 2.866629953190318e-05,
+      "loss": 10.5178,
+      "step": 9380
+    },
+    {
+      "epoch": 0.046892556618142775,
+      "grad_norm": 0.09159563481807709,
+      "learning_rate": 2.8664797616961627e-05,
+      "loss": 10.5194,
+      "step": 9390
+    },
+    {
+      "epoch": 0.046942495443082224,
+      "grad_norm": 0.09416472911834717,
+      "learning_rate": 2.8663295702020073e-05,
+      "loss": 10.5183,
+      "step": 9400
+    },
+    {
+      "epoch": 0.046992434268021674,
+      "grad_norm": 0.09269614517688751,
+      "learning_rate": 2.8661793787078527e-05,
+      "loss": 10.5174,
+      "step": 9410
+    },
+    {
+      "epoch": 0.04704237309296112,
+      "grad_norm": 0.09696542471647263,
+      "learning_rate": 2.8660291872136974e-05,
+      "loss": 10.5164,
+      "step": 9420
+    },
+    {
+      "epoch": 0.04709231191790057,
+      "grad_norm": 0.09807953983545303,
+      "learning_rate": 2.8658789957195427e-05,
+      "loss": 10.5161,
+      "step": 9430
+    },
+    {
+      "epoch": 0.04714225074284002,
+      "grad_norm": 0.09456553310155869,
+      "learning_rate": 2.8657288042253874e-05,
+      "loss": 10.5161,
+      "step": 9440
+    },
+    {
+      "epoch": 0.04719218956777947,
+      "grad_norm": 0.09571606665849686,
+      "learning_rate": 2.865578612731232e-05,
+      "loss": 10.5163,
+      "step": 9450
+    },
+    {
+      "epoch": 0.04724212839271892,
+      "grad_norm": 0.09193666279315948,
+      "learning_rate": 2.8654284212370774e-05,
+      "loss": 10.5135,
+      "step": 9460
+    },
+    {
+      "epoch": 0.04729206721765837,
+      "grad_norm": 0.09799723327159882,
+      "learning_rate": 2.865278229742922e-05,
+      "loss": 10.515,
+      "step": 9470
+    },
+    {
+      "epoch": 0.04734200604259782,
+      "grad_norm": 0.09496576339006424,
+      "learning_rate": 2.8651280382487675e-05,
+      "loss": 10.5143,
+      "step": 9480
+    },
+    {
+      "epoch": 0.04739194486753727,
+      "grad_norm": 0.09969381988048553,
+      "learning_rate": 2.864977846754612e-05,
+      "loss": 10.5162,
+      "step": 9490
+    },
+    {
+      "epoch": 0.04744188369247672,
+      "grad_norm": 0.08909635990858078,
+      "learning_rate": 2.8648276552604572e-05,
+      "loss": 10.5178,
+      "step": 9500
+    },
+    {
+      "epoch": 0.04749182251741617,
+      "grad_norm": 0.10057427734136581,
+      "learning_rate": 2.8646774637663022e-05,
+      "loss": 10.5147,
+      "step": 9510
+    },
+    {
+      "epoch": 0.04754176134235561,
+      "grad_norm": 0.09253738075494766,
+      "learning_rate": 2.864527272272147e-05,
+      "loss": 10.5167,
+      "step": 9520
+    },
+    {
+      "epoch": 0.04759170016729506,
+      "grad_norm": 0.09274590760469437,
+      "learning_rate": 2.8643770807779922e-05,
+      "loss": 10.9066,
+      "step": 9530
+    },
+    {
+      "epoch": 0.04764163899223451,
+      "grad_norm": 0.09520485997200012,
+      "learning_rate": 2.864226889283837e-05,
+      "loss": 10.5115,
+      "step": 9540
+    },
+    {
+      "epoch": 0.04769157781717396,
+      "grad_norm": 0.09519509971141815,
+      "learning_rate": 2.864076697789682e-05,
+      "loss": 10.5245,
+      "step": 9550
+    },
+    {
+      "epoch": 0.04774151664211341,
+      "grad_norm": 0.09528204053640366,
+      "learning_rate": 2.863926506295527e-05,
+      "loss": 10.5133,
+      "step": 9560
+    },
+    {
+      "epoch": 0.04779145546705286,
+      "grad_norm": 0.09493537992238998,
+      "learning_rate": 2.8637763148013716e-05,
+      "loss": 10.5144,
+      "step": 9570
+    },
+    {
+      "epoch": 0.04784139429199231,
+      "grad_norm": 0.09186563640832901,
+      "learning_rate": 2.863626123307217e-05,
+      "loss": 10.5186,
+      "step": 9580
+    },
+    {
+      "epoch": 0.04789133311693176,
+      "grad_norm": 0.09734216332435608,
+      "learning_rate": 2.8634759318130617e-05,
+      "loss": 10.5126,
+      "step": 9590
+    },
+    {
+      "epoch": 0.04794127194187121,
+      "grad_norm": 0.09098707884550095,
+      "learning_rate": 2.8633257403189067e-05,
+      "loss": 11.7738,
+      "step": 9600
+    },
+    {
+      "epoch": 0.047991210766810656,
+      "grad_norm": 0.09297781437635422,
+      "learning_rate": 2.8631755488247517e-05,
+      "loss": 10.6659,
+      "step": 9610
+    },
+    {
+      "epoch": 0.048041149591750106,
+      "grad_norm": 0.09503110498189926,
+      "learning_rate": 2.8630253573305964e-05,
+      "loss": 10.5111,
+      "step": 9620
+    },
+    {
+      "epoch": 0.048091088416689555,
+      "grad_norm": 0.09521497786045074,
+      "learning_rate": 2.8628751658364417e-05,
+      "loss": 10.5093,
+      "step": 9630
+    },
+    {
+      "epoch": 0.048141027241629004,
+      "grad_norm": 0.09482381492853165,
+      "learning_rate": 2.8627249743422864e-05,
+      "loss": 10.5128,
+      "step": 9640
+    },
+    {
+      "epoch": 0.048190966066568454,
+      "grad_norm": 0.09641779214143753,
+      "learning_rate": 2.8625747828481314e-05,
+      "loss": 10.51,
+      "step": 9650
+    },
+    {
+      "epoch": 0.0482409048915079,
+      "grad_norm": 0.10081770271062851,
+      "learning_rate": 2.8624245913539764e-05,
+      "loss": 10.511,
+      "step": 9660
+    },
+    {
+      "epoch": 0.04829084371644735,
+      "grad_norm": 0.09441135823726654,
+      "learning_rate": 2.862274399859821e-05,
+      "loss": 10.5093,
+      "step": 9670
+    },
+    {
+      "epoch": 0.0483407825413868,
+      "grad_norm": 0.09851568937301636,
+      "learning_rate": 2.8621242083656665e-05,
+      "loss": 10.5099,
+      "step": 9680
+    },
+    {
+      "epoch": 0.04839072136632625,
+      "grad_norm": 0.09381608664989471,
+      "learning_rate": 2.861974016871511e-05,
+      "loss": 10.5099,
+      "step": 9690
+    },
+    {
+      "epoch": 0.0484406601912657,
+      "grad_norm": 0.0943373441696167,
+      "learning_rate": 2.8618238253773562e-05,
+      "loss": 10.5086,
+      "step": 9700
+    },
+    {
+      "epoch": 0.04849059901620515,
+      "grad_norm": 0.09664776921272278,
+      "learning_rate": 2.8616736338832012e-05,
+      "loss": 10.51,
+      "step": 9710
+    },
+    {
+      "epoch": 0.0485405378411446,
+      "grad_norm": 0.09082762151956558,
+      "learning_rate": 2.861523442389046e-05,
+      "loss": 10.509,
+      "step": 9720
+    },
+    {
+      "epoch": 0.04859047666608405,
+      "grad_norm": 0.1018294245004654,
+      "learning_rate": 2.8613732508948912e-05,
+      "loss": 10.5095,
+      "step": 9730
+    },
+    {
+      "epoch": 0.0486404154910235,
+      "grad_norm": 0.09788943082094193,
+      "learning_rate": 2.861223059400736e-05,
+      "loss": 10.506,
+      "step": 9740
+    },
+    {
+      "epoch": 0.04869035431596295,
+      "grad_norm": 0.0945865660905838,
+      "learning_rate": 2.861072867906581e-05,
+      "loss": 10.5045,
+      "step": 9750
+    },
+    {
+      "epoch": 0.0487402931409024,
+      "grad_norm": 0.09245441854000092,
+      "learning_rate": 2.860922676412426e-05,
+      "loss": 10.5076,
+      "step": 9760
+    },
+    {
+      "epoch": 0.04879023196584184,
+      "grad_norm": 0.09692434966564178,
+      "learning_rate": 2.8607724849182706e-05,
+      "loss": 10.5044,
+      "step": 9770
+    },
+    {
+      "epoch": 0.04884017079078129,
+      "grad_norm": 0.10064822435379028,
+      "learning_rate": 2.860622293424116e-05,
+      "loss": 10.5054,
+      "step": 9780
+    },
+    {
+      "epoch": 0.04889010961572074,
+      "grad_norm": 0.09322358667850494,
+      "learning_rate": 2.8604721019299607e-05,
+      "loss": 10.5048,
+      "step": 9790
+    },
+    {
+      "epoch": 0.04894004844066019,
+      "grad_norm": 0.09551732242107391,
+      "learning_rate": 2.8603219104358057e-05,
+      "loss": 10.5049,
+      "step": 9800
+    },
+    {
+      "epoch": 0.04898998726559964,
+      "grad_norm": 0.09352359920740128,
+      "learning_rate": 2.8601717189416507e-05,
+      "loss": 10.5055,
+      "step": 9810
+    },
+    {
+      "epoch": 0.04903992609053909,
+      "grad_norm": 0.09425562620162964,
+      "learning_rate": 2.8600215274474957e-05,
+      "loss": 10.5046,
+      "step": 9820
+    },
+    {
+      "epoch": 0.04908986491547854,
+      "grad_norm": 0.09174810349941254,
+      "learning_rate": 2.8598713359533407e-05,
+      "loss": 10.5061,
+      "step": 9830
+    },
+    {
+      "epoch": 0.04913980374041799,
+      "grad_norm": 0.0945669412612915,
+      "learning_rate": 2.8597211444591854e-05,
+      "loss": 10.5056,
+      "step": 9840
+    },
+    {
+      "epoch": 0.049189742565357436,
+      "grad_norm": 0.1002650186419487,
+      "learning_rate": 2.8595709529650304e-05,
+      "loss": 10.5034,
+      "step": 9850
+    },
+    {
+      "epoch": 0.049239681390296886,
+      "grad_norm": 0.09847228229045868,
+      "learning_rate": 2.8594207614708754e-05,
+      "loss": 10.503,
+      "step": 9860
+    },
+    {
+      "epoch": 0.049289620215236335,
+      "grad_norm": 0.10454359650611877,
+      "learning_rate": 2.8592705699767205e-05,
+      "loss": 10.504,
+      "step": 9870
+    },
+    {
+      "epoch": 0.049339559040175784,
+      "grad_norm": 0.09625834971666336,
+      "learning_rate": 2.8591203784825655e-05,
+      "loss": 10.5023,
+      "step": 9880
+    },
+    {
+      "epoch": 0.049389497865115234,
+      "grad_norm": 0.09287148714065552,
+      "learning_rate": 2.85897018698841e-05,
+      "loss": 10.5059,
+      "step": 9890
+    },
+    {
+      "epoch": 0.04943943669005468,
+      "grad_norm": 0.09204880893230438,
+      "learning_rate": 2.8588199954942552e-05,
+      "loss": 10.5014,
+      "step": 9900
+    },
+    {
+      "epoch": 0.04948937551499413,
+      "grad_norm": 0.09391821175813675,
+      "learning_rate": 2.8586698040001002e-05,
+      "loss": 10.5023,
+      "step": 9910
+    },
+    {
+      "epoch": 0.04953931433993358,
+      "grad_norm": 0.09617898613214493,
+      "learning_rate": 2.8585196125059452e-05,
+      "loss": 10.5032,
+      "step": 9920
+    },
+    {
+      "epoch": 0.04958925316487303,
+      "grad_norm": 0.09573251754045486,
+      "learning_rate": 2.8583694210117902e-05,
+      "loss": 10.5024,
+      "step": 9930
+    },
+    {
+      "epoch": 0.04963919198981248,
+      "grad_norm": 0.09664706140756607,
+      "learning_rate": 2.858219229517635e-05,
+      "loss": 10.5018,
+      "step": 9940
+    },
+    {
+      "epoch": 0.04968913081475193,
+      "grad_norm": 0.09707572311162949,
+      "learning_rate": 2.85806903802348e-05,
+      "loss": 10.5026,
+      "step": 9950
+    },
+    {
+      "epoch": 0.04973906963969138,
+      "grad_norm": 0.09420596808195114,
+      "learning_rate": 2.857918846529325e-05,
+      "loss": 10.5004,
+      "step": 9960
+    },
+    {
+      "epoch": 0.04978900846463083,
+      "grad_norm": 0.09333644807338715,
+      "learning_rate": 2.85776865503517e-05,
+      "loss": 10.4996,
+      "step": 9970
+    },
+    {
+      "epoch": 0.04983894728957028,
+      "grad_norm": 0.09561195969581604,
+      "learning_rate": 2.857618463541015e-05,
+      "loss": 10.5005,
+      "step": 9980
+    },
+    {
+      "epoch": 0.04988888611450973,
+      "grad_norm": 0.09285534173250198,
+      "learning_rate": 2.8574682720468597e-05,
+      "loss": 10.5016,
+      "step": 9990
+    },
+    {
+      "epoch": 0.04993882493944918,
+      "grad_norm": 0.09716643393039703,
+      "learning_rate": 2.8573180805527047e-05,
+      "loss": 10.4969,
+      "step": 10000
+    },
+    {
+      "epoch": 0.04998876376438863,
+      "grad_norm": 0.09244739264249802,
+      "learning_rate": 2.8571678890585497e-05,
+      "loss": 10.4983,
+      "step": 10010
+    },
+    {
+      "epoch": 0.05003870258932807,
+      "grad_norm": 0.099833182990551,
+      "learning_rate": 2.8570176975643947e-05,
+      "loss": 10.5007,
+      "step": 10020
+    },
+    {
+      "epoch": 0.05008864141426752,
+      "grad_norm": 0.09650156646966934,
+      "learning_rate": 2.8568675060702397e-05,
+      "loss": 10.4993,
+      "step": 10030
+    },
+    {
+      "epoch": 0.05013858023920697,
+      "grad_norm": 0.10166086256504059,
+      "learning_rate": 2.8567173145760844e-05,
+      "loss": 10.4981,
+      "step": 10040
+    },
+    {
+      "epoch": 0.05018851906414642,
+      "grad_norm": 0.09234780818223953,
+      "learning_rate": 2.8565671230819294e-05,
+      "loss": 10.4969,
+      "step": 10050
+    },
+    {
+      "epoch": 0.05023845788908587,
+      "grad_norm": 0.0958356112241745,
+      "learning_rate": 2.8564169315877744e-05,
+      "loss": 10.4996,
+      "step": 10060
+    },
+    {
+      "epoch": 0.05028839671402532,
+      "grad_norm": 0.09565756469964981,
+      "learning_rate": 2.8562667400936195e-05,
+      "loss": 10.4971,
+      "step": 10070
+    },
+    {
+      "epoch": 0.05033833553896477,
+      "grad_norm": 0.09427642822265625,
+      "learning_rate": 2.8561165485994645e-05,
+      "loss": 10.4966,
+      "step": 10080
+    },
+    {
+      "epoch": 0.050388274363904216,
+      "grad_norm": 0.0930957943201065,
+      "learning_rate": 2.855966357105309e-05,
+      "loss": 10.4987,
+      "step": 10090
+    },
+    {
+      "epoch": 0.050438213188843666,
+      "grad_norm": 0.09489066898822784,
+      "learning_rate": 2.8558161656111542e-05,
+      "loss": 10.496,
+      "step": 10100
+    },
+    {
+      "epoch": 0.050488152013783115,
+      "grad_norm": 0.09817587584257126,
+      "learning_rate": 2.8556659741169992e-05,
+      "loss": 10.4956,
+      "step": 10110
+    },
+    {
+      "epoch": 0.050538090838722564,
+      "grad_norm": 0.09142786264419556,
+      "learning_rate": 2.8555157826228442e-05,
+      "loss": 10.4974,
+      "step": 10120
+    },
+    {
+      "epoch": 0.050588029663662014,
+      "grad_norm": 0.09748794138431549,
+      "learning_rate": 2.8553655911286892e-05,
+      "loss": 10.4959,
+      "step": 10130
+    },
+    {
+      "epoch": 0.05063796848860146,
+      "grad_norm": 0.09116542339324951,
+      "learning_rate": 2.8552153996345343e-05,
+      "loss": 10.498,
+      "step": 10140
+    },
+    {
+      "epoch": 0.05068790731354091,
+      "grad_norm": 0.09313584864139557,
+      "learning_rate": 2.855065208140379e-05,
+      "loss": 10.4932,
+      "step": 10150
+    },
+    {
+      "epoch": 0.05073784613848036,
+      "grad_norm": 0.09887810051441193,
+      "learning_rate": 2.854915016646224e-05,
+      "loss": 10.4945,
+      "step": 10160
+    },
+    {
+      "epoch": 0.05078778496341981,
+      "grad_norm": 0.09741490334272385,
+      "learning_rate": 2.854764825152069e-05,
+      "loss": 10.4943,
+      "step": 10170
+    },
+    {
+      "epoch": 0.05083772378835926,
+      "grad_norm": 0.09374571591615677,
+      "learning_rate": 2.854614633657914e-05,
+      "loss": 10.4943,
+      "step": 10180
+    },
+    {
+      "epoch": 0.05088766261329871,
+      "grad_norm": 0.09434989094734192,
+      "learning_rate": 2.854464442163759e-05,
+      "loss": 10.4963,
+      "step": 10190
+    },
+    {
+      "epoch": 0.05093760143823816,
+      "grad_norm": 0.09508496522903442,
+      "learning_rate": 2.8543142506696037e-05,
+      "loss": 10.4925,
+      "step": 10200
+    },
+    {
+      "epoch": 0.05098754026317761,
+      "grad_norm": 0.09928717464208603,
+      "learning_rate": 2.8541640591754487e-05,
+      "loss": 10.4933,
+      "step": 10210
+    },
+    {
+      "epoch": 0.05103747908811706,
+      "grad_norm": 0.09969381988048553,
+      "learning_rate": 2.8540138676812937e-05,
+      "loss": 10.4932,
+      "step": 10220
+    },
+    {
+      "epoch": 0.05108741791305651,
+      "grad_norm": 0.0988709032535553,
+      "learning_rate": 2.8538636761871387e-05,
+      "loss": 10.4942,
+      "step": 10230
+    },
+    {
+      "epoch": 0.05113735673799596,
+      "grad_norm": 0.09207888692617416,
+      "learning_rate": 2.8537134846929838e-05,
+      "loss": 10.492,
+      "step": 10240
+    },
+    {
+      "epoch": 0.05118729556293541,
+      "grad_norm": 0.09073358029127121,
+      "learning_rate": 2.8535632931988284e-05,
+      "loss": 10.494,
+      "step": 10250
+    },
+    {
+      "epoch": 0.05123723438787485,
+      "grad_norm": 0.09680426120758057,
+      "learning_rate": 2.8534131017046734e-05,
+      "loss": 10.4916,
+      "step": 10260
+    },
+    {
+      "epoch": 0.0512871732128143,
+      "grad_norm": 0.09444583207368851,
+      "learning_rate": 2.8532629102105185e-05,
+      "loss": 10.4934,
+      "step": 10270
+    },
+    {
+      "epoch": 0.05133711203775375,
+      "grad_norm": 0.09547161310911179,
+      "learning_rate": 2.8531127187163635e-05,
+      "loss": 10.4909,
+      "step": 10280
+    },
+    {
+      "epoch": 0.0513870508626932,
+      "grad_norm": 0.09873054176568985,
+      "learning_rate": 2.8529625272222085e-05,
+      "loss": 10.4924,
+      "step": 10290
+    },
+    {
+      "epoch": 0.05143698968763265,
+      "grad_norm": 0.09690874069929123,
+      "learning_rate": 2.8528123357280532e-05,
+      "loss": 10.4925,
+      "step": 10300
+    },
+    {
+      "epoch": 0.0514869285125721,
+      "grad_norm": 0.09330197423696518,
+      "learning_rate": 2.8526621442338982e-05,
+      "loss": 10.4923,
+      "step": 10310
+    },
+    {
+      "epoch": 0.05153686733751155,
+      "grad_norm": 0.10089316219091415,
+      "learning_rate": 2.8525119527397432e-05,
+      "loss": 10.4924,
+      "step": 10320
+    },
+    {
+      "epoch": 0.051586806162450996,
+      "grad_norm": 0.0995856523513794,
+      "learning_rate": 2.8523617612455882e-05,
+      "loss": 10.4916,
+      "step": 10330
+    },
+    {
+      "epoch": 0.051636744987390445,
+      "grad_norm": 0.09683484584093094,
+      "learning_rate": 2.8522115697514333e-05,
+      "loss": 10.4886,
+      "step": 10340
+    },
+    {
+      "epoch": 0.051686683812329895,
+      "grad_norm": 0.09331785142421722,
+      "learning_rate": 2.852061378257278e-05,
+      "loss": 10.4894,
+      "step": 10350
+    },
+    {
+      "epoch": 0.051736622637269344,
+      "grad_norm": 0.09782253205776215,
+      "learning_rate": 2.851911186763123e-05,
+      "loss": 10.4884,
+      "step": 10360
+    },
+    {
+      "epoch": 0.051786561462208794,
+      "grad_norm": 0.09803865849971771,
+      "learning_rate": 2.851760995268968e-05,
+      "loss": 10.4905,
+      "step": 10370
+    },
+    {
+      "epoch": 0.05183650028714824,
+      "grad_norm": 0.09339244663715363,
+      "learning_rate": 2.851610803774813e-05,
+      "loss": 10.4909,
+      "step": 10380
+    },
+    {
+      "epoch": 0.05188643911208769,
+      "grad_norm": 0.09537281095981598,
+      "learning_rate": 2.851460612280658e-05,
+      "loss": 10.4881,
+      "step": 10390
+    },
+    {
+      "epoch": 0.05193637793702714,
+      "grad_norm": 0.09507238864898682,
+      "learning_rate": 2.8513104207865027e-05,
+      "loss": 10.4862,
+      "step": 10400
+    },
+    {
+      "epoch": 0.05198631676196659,
+      "grad_norm": 0.09395944327116013,
+      "learning_rate": 2.8511602292923477e-05,
+      "loss": 10.486,
+      "step": 10410
+    },
+    {
+      "epoch": 0.05203625558690604,
+      "grad_norm": 0.09412617236375809,
+      "learning_rate": 2.8510100377981927e-05,
+      "loss": 10.4888,
+      "step": 10420
+    },
+    {
+      "epoch": 0.05208619441184549,
+      "grad_norm": 0.10089904814958572,
+      "learning_rate": 2.8508598463040377e-05,
+      "loss": 10.488,
+      "step": 10430
+    },
+    {
+      "epoch": 0.05213613323678494,
+      "grad_norm": 0.09662104398012161,
+      "learning_rate": 2.8507096548098828e-05,
+      "loss": 10.4879,
+      "step": 10440
+    },
+    {
+      "epoch": 0.05218607206172439,
+      "grad_norm": 0.09902186691761017,
+      "learning_rate": 2.8505594633157274e-05,
+      "loss": 10.4877,
+      "step": 10450
+    },
+    {
+      "epoch": 0.05223601088666384,
+      "grad_norm": 0.09871252626180649,
+      "learning_rate": 2.8504092718215728e-05,
+      "loss": 10.489,
+      "step": 10460
+    },
+    {
+      "epoch": 0.05228594971160329,
+      "grad_norm": 0.09363856911659241,
+      "learning_rate": 2.8502590803274175e-05,
+      "loss": 10.4855,
+      "step": 10470
+    },
+    {
+      "epoch": 0.05233588853654274,
+      "grad_norm": 0.0946008712053299,
+      "learning_rate": 2.8501088888332625e-05,
+      "loss": 10.4855,
+      "step": 10480
+    },
+    {
+      "epoch": 0.05238582736148219,
+      "grad_norm": 0.09601443260908127,
+      "learning_rate": 2.8499586973391075e-05,
+      "loss": 10.4831,
+      "step": 10490
+    },
+    {
+      "epoch": 0.05243576618642164,
+      "grad_norm": 0.09569672495126724,
+      "learning_rate": 2.8498085058449522e-05,
+      "loss": 10.4833,
+      "step": 10500
+    },
+    {
+      "epoch": 0.05248570501136108,
+      "grad_norm": 0.09212665259838104,
+      "learning_rate": 2.8496583143507975e-05,
+      "loss": 10.4858,
+      "step": 10510
+    },
+    {
+      "epoch": 0.05253564383630053,
+      "grad_norm": 0.09606766700744629,
+      "learning_rate": 2.8495081228566422e-05,
+      "loss": 10.4837,
+      "step": 10520
+    },
+    {
+      "epoch": 0.05258558266123998,
+      "grad_norm": 0.09896976500749588,
+      "learning_rate": 2.8493579313624872e-05,
+      "loss": 10.4854,
+      "step": 10530
+    },
+    {
+      "epoch": 0.05263552148617943,
+      "grad_norm": 0.0996663048863411,
+      "learning_rate": 2.8492077398683323e-05,
+      "loss": 10.4849,
+      "step": 10540
+    },
+    {
+      "epoch": 0.05268546031111888,
+      "grad_norm": 0.09384378045797348,
+      "learning_rate": 2.849057548374177e-05,
+      "loss": 10.4849,
+      "step": 10550
+    },
+    {
+      "epoch": 0.05273539913605833,
+      "grad_norm": 0.09996800869703293,
+      "learning_rate": 2.8489073568800223e-05,
+      "loss": 10.4818,
+      "step": 10560
+    },
+    {
+      "epoch": 0.052785337960997776,
+      "grad_norm": 0.09955079853534698,
+      "learning_rate": 2.848757165385867e-05,
+      "loss": 10.4836,
+      "step": 10570
+    },
+    {
+      "epoch": 0.052835276785937225,
+      "grad_norm": 0.0904826894402504,
+      "learning_rate": 2.848606973891712e-05,
+      "loss": 10.483,
+      "step": 10580
+    },
+    {
+      "epoch": 0.052885215610876675,
+      "grad_norm": 0.0925852507352829,
+      "learning_rate": 2.848456782397557e-05,
+      "loss": 10.4826,
+      "step": 10590
+    },
+    {
+      "epoch": 0.052935154435816124,
+      "grad_norm": 0.09814697504043579,
+      "learning_rate": 2.8483065909034017e-05,
+      "loss": 10.483,
+      "step": 10600
+    },
+    {
+      "epoch": 0.052985093260755574,
+      "grad_norm": 0.09643968194723129,
+      "learning_rate": 2.848156399409247e-05,
+      "loss": 10.4833,
+      "step": 10610
+    },
+    {
+      "epoch": 0.05303503208569502,
+      "grad_norm": 0.09589682519435883,
+      "learning_rate": 2.8480062079150917e-05,
+      "loss": 10.4814,
+      "step": 10620
+    },
+    {
+      "epoch": 0.05308497091063447,
+      "grad_norm": 0.09177594631910324,
+      "learning_rate": 2.8478560164209367e-05,
+      "loss": 10.4801,
+      "step": 10630
+    },
+    {
+      "epoch": 0.05313490973557392,
+      "grad_norm": 0.08999846875667572,
+      "learning_rate": 2.8477058249267818e-05,
+      "loss": 10.4834,
+      "step": 10640
+    },
+    {
+      "epoch": 0.05318484856051337,
+      "grad_norm": 0.09998583048582077,
+      "learning_rate": 2.8475556334326264e-05,
+      "loss": 10.4802,
+      "step": 10650
+    },
+    {
+      "epoch": 0.05323478738545282,
+      "grad_norm": 0.0951545462012291,
+      "learning_rate": 2.8474054419384718e-05,
+      "loss": 10.4797,
+      "step": 10660
+    },
+    {
+      "epoch": 0.05328472621039227,
+      "grad_norm": 0.09384284913539886,
+      "learning_rate": 2.8472552504443165e-05,
+      "loss": 10.4819,
+      "step": 10670
+    },
+    {
+      "epoch": 0.05333466503533172,
+      "grad_norm": 0.09465983510017395,
+      "learning_rate": 2.8471050589501615e-05,
+      "loss": 10.4797,
+      "step": 10680
+    },
+    {
+      "epoch": 0.05338460386027117,
+      "grad_norm": 0.09355910867452621,
+      "learning_rate": 2.8469548674560065e-05,
+      "loss": 10.4793,
+      "step": 10690
+    },
+    {
+      "epoch": 0.05343454268521062,
+      "grad_norm": 0.10254157334566116,
+      "learning_rate": 2.8468046759618512e-05,
+      "loss": 10.481,
+      "step": 10700
+    },
+    {
+      "epoch": 0.05348448151015007,
+      "grad_norm": 0.09623821824789047,
+      "learning_rate": 2.8466544844676965e-05,
+      "loss": 10.4798,
+      "step": 10710
+    },
+    {
+      "epoch": 0.05353442033508952,
+      "grad_norm": 0.0956728383898735,
+      "learning_rate": 2.8465042929735412e-05,
+      "loss": 10.4775,
+      "step": 10720
+    },
+    {
+      "epoch": 0.05358435916002897,
+      "grad_norm": 0.1043584942817688,
+      "learning_rate": 2.8463541014793862e-05,
+      "loss": 10.4797,
+      "step": 10730
+    },
+    {
+      "epoch": 0.05363429798496842,
+      "grad_norm": 0.09671546518802643,
+      "learning_rate": 2.8462039099852313e-05,
+      "loss": 10.4768,
+      "step": 10740
+    },
+    {
+      "epoch": 0.053684236809907866,
+      "grad_norm": 0.0880408063530922,
+      "learning_rate": 2.846053718491076e-05,
+      "loss": 10.4796,
+      "step": 10750
+    },
+    {
+      "epoch": 0.05373417563484731,
+      "grad_norm": 0.0982058197259903,
+      "learning_rate": 2.8459035269969213e-05,
+      "loss": 10.4792,
+      "step": 10760
+    },
+    {
+      "epoch": 0.05378411445978676,
+      "grad_norm": 0.09881541877985,
+      "learning_rate": 2.845753335502766e-05,
+      "loss": 10.4785,
+      "step": 10770
+    },
+    {
+      "epoch": 0.05383405328472621,
+      "grad_norm": 0.10053160041570663,
+      "learning_rate": 2.8456031440086113e-05,
+      "loss": 10.4742,
+      "step": 10780
+    },
+    {
+      "epoch": 0.05388399210966566,
+      "grad_norm": 0.0990302637219429,
+      "learning_rate": 2.845452952514456e-05,
+      "loss": 10.4763,
+      "step": 10790
+    },
+    {
+      "epoch": 0.05393393093460511,
+      "grad_norm": 0.09606090933084488,
+      "learning_rate": 2.8453027610203007e-05,
+      "loss": 10.478,
+      "step": 10800
+    },
+    {
+      "epoch": 0.053983869759544556,
+      "grad_norm": 0.09462147206068039,
+      "learning_rate": 2.845152569526146e-05,
+      "loss": 10.4758,
+      "step": 10810
+    },
+    {
+      "epoch": 0.054033808584484005,
+      "grad_norm": 0.09580814093351364,
+      "learning_rate": 2.8450023780319907e-05,
+      "loss": 10.4743,
+      "step": 10820
+    },
+    {
+      "epoch": 0.054083747409423455,
+      "grad_norm": 0.09848089516162872,
+      "learning_rate": 2.844852186537836e-05,
+      "loss": 10.4749,
+      "step": 10830
+    },
+    {
+      "epoch": 0.054133686234362904,
+      "grad_norm": 0.09434827417135239,
+      "learning_rate": 2.8447019950436808e-05,
+      "loss": 10.4777,
+      "step": 10840
+    },
+    {
+      "epoch": 0.054183625059302354,
+      "grad_norm": 0.09295500814914703,
+      "learning_rate": 2.8445518035495254e-05,
+      "loss": 10.4744,
+      "step": 10850
+    },
+    {
+      "epoch": 0.0542335638842418,
+      "grad_norm": 0.0959131121635437,
+      "learning_rate": 2.8444016120553708e-05,
+      "loss": 10.4755,
+      "step": 10860
+    },
+    {
+      "epoch": 0.05428350270918125,
+      "grad_norm": 0.09555883705615997,
+      "learning_rate": 2.8442514205612155e-05,
+      "loss": 10.4737,
+      "step": 10870
+    },
+    {
+      "epoch": 0.0543334415341207,
+      "grad_norm": 0.09721937030553818,
+      "learning_rate": 2.8441012290670608e-05,
+      "loss": 10.4752,
+      "step": 10880
+    },
+    {
+      "epoch": 0.05438338035906015,
+      "grad_norm": 0.09354712814092636,
+      "learning_rate": 2.8439510375729055e-05,
+      "loss": 10.4721,
+      "step": 10890
+    },
+    {
+      "epoch": 0.0544333191839996,
+      "grad_norm": 0.09647790342569351,
+      "learning_rate": 2.8438008460787502e-05,
+      "loss": 10.4745,
+      "step": 10900
+    },
+    {
+      "epoch": 0.05448325800893905,
+      "grad_norm": 0.10081273317337036,
+      "learning_rate": 2.8436506545845955e-05,
+      "loss": 10.4725,
+      "step": 10910
+    },
+    {
+      "epoch": 0.0545331968338785,
+      "grad_norm": 0.09475501626729965,
+      "learning_rate": 2.8435004630904402e-05,
+      "loss": 10.4723,
+      "step": 10920
+    },
+    {
+      "epoch": 0.05458313565881795,
+      "grad_norm": 0.09256067872047424,
+      "learning_rate": 2.8433502715962856e-05,
+      "loss": 10.4717,
+      "step": 10930
+    },
+    {
+      "epoch": 0.0546330744837574,
+      "grad_norm": 0.09541960060596466,
+      "learning_rate": 2.8432000801021303e-05,
+      "loss": 10.4727,
+      "step": 10940
+    },
+    {
+      "epoch": 0.05468301330869685,
+      "grad_norm": 0.09717325121164322,
+      "learning_rate": 2.843049888607975e-05,
+      "loss": 10.4725,
+      "step": 10950
+    },
+    {
+      "epoch": 0.0547329521336363,
+      "grad_norm": 0.09780695289373398,
+      "learning_rate": 2.8428996971138203e-05,
+      "loss": 10.4726,
+      "step": 10960
+    },
+    {
+      "epoch": 0.05478289095857575,
+      "grad_norm": 0.09682448208332062,
+      "learning_rate": 2.842749505619665e-05,
+      "loss": 10.4728,
+      "step": 10970
+    },
+    {
+      "epoch": 0.0548328297835152,
+      "grad_norm": 0.0957455188035965,
+      "learning_rate": 2.8425993141255103e-05,
+      "loss": 10.4719,
+      "step": 10980
+    },
+    {
+      "epoch": 0.054882768608454646,
+      "grad_norm": 0.0956406518816948,
+      "learning_rate": 2.842449122631355e-05,
+      "loss": 10.4714,
+      "step": 10990
+    },
+    {
+      "epoch": 0.054932707433394096,
+      "grad_norm": 0.0934106856584549,
+      "learning_rate": 2.8422989311371997e-05,
+      "loss": 10.4694,
+      "step": 11000
+    },
+    {
+      "epoch": 0.05498264625833354,
+      "grad_norm": 0.09818681329488754,
+      "learning_rate": 2.842148739643045e-05,
+      "loss": 10.4721,
+      "step": 11010
+    },
+    {
+      "epoch": 0.05503258508327299,
+      "grad_norm": 0.09489887952804565,
+      "learning_rate": 2.8419985481488897e-05,
+      "loss": 10.4708,
+      "step": 11020
+    },
+    {
+      "epoch": 0.05508252390821244,
+      "grad_norm": 0.09038610011339188,
+      "learning_rate": 2.841848356654735e-05,
+      "loss": 10.4731,
+      "step": 11030
+    },
+    {
+      "epoch": 0.05513246273315189,
+      "grad_norm": 0.0944511890411377,
+      "learning_rate": 2.8416981651605798e-05,
+      "loss": 10.4708,
+      "step": 11040
+    },
+    {
+      "epoch": 0.055182401558091336,
+      "grad_norm": 0.09606794267892838,
+      "learning_rate": 2.8415479736664244e-05,
+      "loss": 10.4689,
+      "step": 11050
+    },
+    {
+      "epoch": 0.055232340383030785,
+      "grad_norm": 0.09804631024599075,
+      "learning_rate": 2.8413977821722698e-05,
+      "loss": 10.4698,
+      "step": 11060
+    },
+    {
+      "epoch": 0.055282279207970235,
+      "grad_norm": 0.10113038122653961,
+      "learning_rate": 2.8412475906781145e-05,
+      "loss": 10.4686,
+      "step": 11070
+    },
+    {
+      "epoch": 0.055332218032909684,
+      "grad_norm": 0.10174177587032318,
+      "learning_rate": 2.8410973991839598e-05,
+      "loss": 10.4694,
+      "step": 11080
+    },
+    {
+      "epoch": 0.055382156857849134,
+      "grad_norm": 0.09673687815666199,
+      "learning_rate": 2.8409472076898045e-05,
+      "loss": 10.4677,
+      "step": 11090
+    },
+    {
+      "epoch": 0.05543209568278858,
+      "grad_norm": 0.09577104449272156,
+      "learning_rate": 2.8407970161956495e-05,
+      "loss": 10.4673,
+      "step": 11100
+    },
+    {
+      "epoch": 0.05548203450772803,
+      "grad_norm": 0.09360501170158386,
+      "learning_rate": 2.8406468247014945e-05,
+      "loss": 10.47,
+      "step": 11110
+    },
+    {
+      "epoch": 0.05553197333266748,
+      "grad_norm": 0.09364285320043564,
+      "learning_rate": 2.8404966332073392e-05,
+      "loss": 10.4676,
+      "step": 11120
+    },
+    {
+      "epoch": 0.05558191215760693,
+      "grad_norm": 0.093747578561306,
+      "learning_rate": 2.8403464417131846e-05,
+      "loss": 10.4669,
+      "step": 11130
+    },
+    {
+      "epoch": 0.05563185098254638,
+      "grad_norm": 0.0995292142033577,
+      "learning_rate": 2.8401962502190293e-05,
+      "loss": 10.4677,
+      "step": 11140
+    },
+    {
+      "epoch": 0.05568178980748583,
+      "grad_norm": 0.0943877249956131,
+      "learning_rate": 2.8400460587248743e-05,
+      "loss": 10.4691,
+      "step": 11150
+    },
+    {
+      "epoch": 0.05573172863242528,
+      "grad_norm": 0.09354741126298904,
+      "learning_rate": 2.8398958672307193e-05,
+      "loss": 10.4657,
+      "step": 11160
+    },
+    {
+      "epoch": 0.05578166745736473,
+      "grad_norm": 0.09523334354162216,
+      "learning_rate": 2.839745675736564e-05,
+      "loss": 10.4658,
+      "step": 11170
+    },
+    {
+      "epoch": 0.05583160628230418,
+      "grad_norm": 0.09460029006004333,
+      "learning_rate": 2.8395954842424093e-05,
+      "loss": 10.4631,
+      "step": 11180
+    },
+    {
+      "epoch": 0.05588154510724363,
+      "grad_norm": 0.09529443830251694,
+      "learning_rate": 2.839445292748254e-05,
+      "loss": 10.4652,
+      "step": 11190
+    },
+    {
+      "epoch": 0.05593148393218308,
+      "grad_norm": 0.09970243275165558,
+      "learning_rate": 2.839295101254099e-05,
+      "loss": 10.4644,
+      "step": 11200
+    },
+    {
+      "epoch": 0.05598142275712253,
+      "grad_norm": 0.09669343382120132,
+      "learning_rate": 2.839144909759944e-05,
+      "loss": 10.465,
+      "step": 11210
+    },
+    {
+      "epoch": 0.05603136158206198,
+      "grad_norm": 0.0948340967297554,
+      "learning_rate": 2.8389947182657887e-05,
+      "loss": 10.4656,
+      "step": 11220
+    },
+    {
+      "epoch": 0.056081300407001426,
+      "grad_norm": 0.09210893511772156,
+      "learning_rate": 2.838844526771634e-05,
+      "loss": 10.4647,
+      "step": 11230
+    },
+    {
+      "epoch": 0.056131239231940876,
+      "grad_norm": 0.09912335872650146,
+      "learning_rate": 2.8386943352774788e-05,
+      "loss": 10.4635,
+      "step": 11240
+    },
+    {
+      "epoch": 0.05618117805688032,
+      "grad_norm": 0.09609910845756531,
+      "learning_rate": 2.8385441437833238e-05,
+      "loss": 10.4647,
+      "step": 11250
+    },
+    {
+      "epoch": 0.05623111688181977,
+      "grad_norm": 0.09431801736354828,
+      "learning_rate": 2.8383939522891688e-05,
+      "loss": 10.4616,
+      "step": 11260
+    },
+    {
+      "epoch": 0.05628105570675922,
+      "grad_norm": 0.09194690734148026,
+      "learning_rate": 2.8382437607950135e-05,
+      "loss": 10.4663,
+      "step": 11270
+    },
+    {
+      "epoch": 0.05633099453169867,
+      "grad_norm": 0.09139011800289154,
+      "learning_rate": 2.8380935693008588e-05,
+      "loss": 10.462,
+      "step": 11280
+    },
+    {
+      "epoch": 0.056380933356638116,
+      "grad_norm": 0.09335776418447495,
+      "learning_rate": 2.8379433778067035e-05,
+      "loss": 10.4619,
+      "step": 11290
+    },
+    {
+      "epoch": 0.056430872181577565,
+      "grad_norm": 0.0932675376534462,
+      "learning_rate": 2.8377931863125485e-05,
+      "loss": 10.4618,
+      "step": 11300
+    },
+    {
+      "epoch": 0.056480811006517015,
+      "grad_norm": 0.09276323765516281,
+      "learning_rate": 2.8376429948183935e-05,
+      "loss": 10.4618,
+      "step": 11310
+    },
+    {
+      "epoch": 0.056530749831456464,
+      "grad_norm": 0.09329287707805634,
+      "learning_rate": 2.8374928033242382e-05,
+      "loss": 10.4613,
+      "step": 11320
+    },
+    {
+      "epoch": 0.056580688656395914,
+      "grad_norm": 0.09470265358686447,
+      "learning_rate": 2.8373426118300836e-05,
+      "loss": 10.4618,
+      "step": 11330
+    },
+    {
+      "epoch": 0.05663062748133536,
+      "grad_norm": 0.09163019806146622,
+      "learning_rate": 2.8371924203359283e-05,
+      "loss": 10.4631,
+      "step": 11340
+    },
+    {
+      "epoch": 0.05668056630627481,
+      "grad_norm": 0.09264067560434341,
+      "learning_rate": 2.8370422288417733e-05,
+      "loss": 10.4619,
+      "step": 11350
+    },
+    {
+      "epoch": 0.05673050513121426,
+      "grad_norm": 0.09891056269407272,
+      "learning_rate": 2.8368920373476183e-05,
+      "loss": 10.459,
+      "step": 11360
+    },
+    {
+      "epoch": 0.05678044395615371,
+      "grad_norm": 0.09387898445129395,
+      "learning_rate": 2.836741845853463e-05,
+      "loss": 10.4621,
+      "step": 11370
+    },
+    {
+      "epoch": 0.05683038278109316,
+      "grad_norm": 0.09117914736270905,
+      "learning_rate": 2.8365916543593083e-05,
+      "loss": 10.4607,
+      "step": 11380
+    },
+    {
+      "epoch": 0.05688032160603261,
+      "grad_norm": 0.09382776916027069,
+      "learning_rate": 2.836441462865153e-05,
+      "loss": 10.4601,
+      "step": 11390
+    },
+    {
+      "epoch": 0.05693026043097206,
+      "grad_norm": 0.09707043319940567,
+      "learning_rate": 2.836291271370998e-05,
+      "loss": 10.4598,
+      "step": 11400
+    },
+    {
+      "epoch": 0.05698019925591151,
+      "grad_norm": 0.09478152543306351,
+      "learning_rate": 2.836141079876843e-05,
+      "loss": 10.4594,
+      "step": 11410
+    },
+    {
+      "epoch": 0.05703013808085096,
+      "grad_norm": 0.09193360805511475,
+      "learning_rate": 2.835990888382688e-05,
+      "loss": 10.4586,
+      "step": 11420
+    },
+    {
+      "epoch": 0.05708007690579041,
+      "grad_norm": 0.09122911840677261,
+      "learning_rate": 2.835840696888533e-05,
+      "loss": 10.4587,
+      "step": 11430
+    },
+    {
+      "epoch": 0.05713001573072986,
+      "grad_norm": 0.09300627559423447,
+      "learning_rate": 2.8356905053943778e-05,
+      "loss": 10.4607,
+      "step": 11440
+    },
+    {
+      "epoch": 0.05717995455566931,
+      "grad_norm": 0.09430982172489166,
+      "learning_rate": 2.8355403139002228e-05,
+      "loss": 10.4581,
+      "step": 11450
+    },
+    {
+      "epoch": 0.05722989338060876,
+      "grad_norm": 0.1006307452917099,
+      "learning_rate": 2.8353901224060678e-05,
+      "loss": 10.4581,
+      "step": 11460
+    },
+    {
+      "epoch": 0.057279832205548206,
+      "grad_norm": 0.09385743737220764,
+      "learning_rate": 2.8352399309119128e-05,
+      "loss": 10.4586,
+      "step": 11470
+    },
+    {
+      "epoch": 0.057329771030487656,
+      "grad_norm": 0.09824958443641663,
+      "learning_rate": 2.835089739417758e-05,
+      "loss": 10.4587,
+      "step": 11480
+    },
+    {
+      "epoch": 0.057379709855427105,
+      "grad_norm": 0.0924108624458313,
+      "learning_rate": 2.8349395479236025e-05,
+      "loss": 10.4568,
+      "step": 11490
+    },
+    {
+      "epoch": 0.05742964868036655,
+      "grad_norm": 0.09391859918832779,
+      "learning_rate": 2.834789356429448e-05,
+      "loss": 10.4601,
+      "step": 11500
+    },
+    {
+      "epoch": 0.057479587505306,
+      "grad_norm": 0.0961366668343544,
+      "learning_rate": 2.8346391649352925e-05,
+      "loss": 10.4558,
+      "step": 11510
+    },
+    {
+      "epoch": 0.05752952633024545,
+      "grad_norm": 0.0933762714266777,
+      "learning_rate": 2.8344889734411376e-05,
+      "loss": 10.4547,
+      "step": 11520
+    },
+    {
+      "epoch": 0.057579465155184896,
+      "grad_norm": 0.09340479224920273,
+      "learning_rate": 2.8343387819469826e-05,
+      "loss": 10.4567,
+      "step": 11530
+    },
+    {
+      "epoch": 0.057629403980124345,
+      "grad_norm": 0.09362339973449707,
+      "learning_rate": 2.8341885904528273e-05,
+      "loss": 10.4587,
+      "step": 11540
+    },
+    {
+      "epoch": 0.057679342805063795,
+      "grad_norm": 0.09768813848495483,
+      "learning_rate": 2.8340383989586726e-05,
+      "loss": 10.4558,
+      "step": 11550
+    },
+    {
+      "epoch": 0.057729281630003244,
+      "grad_norm": 0.09130477160215378,
+      "learning_rate": 2.8338882074645173e-05,
+      "loss": 10.4552,
+      "step": 11560
+    },
+    {
+      "epoch": 0.057779220454942694,
+      "grad_norm": 0.0935993641614914,
+      "learning_rate": 2.8337380159703623e-05,
+      "loss": 10.4522,
+      "step": 11570
+    },
+    {
+      "epoch": 0.05782915927988214,
+      "grad_norm": 0.09301052987575531,
+      "learning_rate": 2.8335878244762073e-05,
+      "loss": 10.4558,
+      "step": 11580
+    },
+    {
+      "epoch": 0.05787909810482159,
+      "grad_norm": 0.09924156963825226,
+      "learning_rate": 2.833437632982052e-05,
+      "loss": 10.4555,
+      "step": 11590
+    },
+    {
+      "epoch": 0.05792903692976104,
+      "grad_norm": 0.09675630927085876,
+      "learning_rate": 2.8332874414878974e-05,
+      "loss": 10.4523,
+      "step": 11600
+    },
+    {
+      "epoch": 0.05797897575470049,
+      "grad_norm": 0.09440577030181885,
+      "learning_rate": 2.833137249993742e-05,
+      "loss": 10.4536,
+      "step": 11610
+    },
+    {
+      "epoch": 0.05802891457963994,
+      "grad_norm": 0.09317108243703842,
+      "learning_rate": 2.832987058499587e-05,
+      "loss": 10.4548,
+      "step": 11620
+    },
+    {
+      "epoch": 0.05807885340457939,
+      "grad_norm": 0.09359714388847351,
+      "learning_rate": 2.832836867005432e-05,
+      "loss": 10.4534,
+      "step": 11630
+    },
+    {
+      "epoch": 0.05812879222951884,
+      "grad_norm": 0.09872293472290039,
+      "learning_rate": 2.8326866755112768e-05,
+      "loss": 10.4551,
+      "step": 11640
+    },
+    {
+      "epoch": 0.05817873105445829,
+      "grad_norm": 0.0959540382027626,
+      "learning_rate": 2.832536484017122e-05,
+      "loss": 10.4505,
+      "step": 11650
+    },
+    {
+      "epoch": 0.05822866987939774,
+      "grad_norm": 0.09325166791677475,
+      "learning_rate": 2.8323862925229668e-05,
+      "loss": 10.4514,
+      "step": 11660
+    },
+    {
+      "epoch": 0.05827860870433719,
+      "grad_norm": 0.0949905589222908,
+      "learning_rate": 2.8322361010288118e-05,
+      "loss": 10.4537,
+      "step": 11670
+    },
+    {
+      "epoch": 0.05832854752927664,
+      "grad_norm": 0.09572743624448776,
+      "learning_rate": 2.832085909534657e-05,
+      "loss": 10.4515,
+      "step": 11680
+    },
+    {
+      "epoch": 0.05837848635421609,
+      "grad_norm": 0.09633267670869827,
+      "learning_rate": 2.8319357180405015e-05,
+      "loss": 10.4507,
+      "step": 11690
+    },
+    {
+      "epoch": 0.05842842517915554,
+      "grad_norm": 0.09619545936584473,
+      "learning_rate": 2.831785526546347e-05,
+      "loss": 10.4505,
+      "step": 11700
+    },
+    {
+      "epoch": 0.058478364004094986,
+      "grad_norm": 0.0950017124414444,
+      "learning_rate": 2.8316353350521915e-05,
+      "loss": 10.4486,
+      "step": 11710
+    },
+    {
+      "epoch": 0.058528302829034436,
+      "grad_norm": 0.09573563188314438,
+      "learning_rate": 2.8314851435580366e-05,
+      "loss": 10.45,
+      "step": 11720
+    },
+    {
+      "epoch": 0.058578241653973885,
+      "grad_norm": 0.09022394567728043,
+      "learning_rate": 2.8313349520638816e-05,
+      "loss": 10.4487,
+      "step": 11730
+    },
+    {
+      "epoch": 0.058628180478913335,
+      "grad_norm": 0.09762822836637497,
+      "learning_rate": 2.8311847605697266e-05,
+      "loss": 10.451,
+      "step": 11740
+    },
+    {
+      "epoch": 0.05867811930385278,
+      "grad_norm": 0.10274951159954071,
+      "learning_rate": 2.8310345690755716e-05,
+      "loss": 10.4502,
+      "step": 11750
+    },
+    {
+      "epoch": 0.05872805812879223,
+      "grad_norm": 0.09342820197343826,
+      "learning_rate": 2.8308843775814163e-05,
+      "loss": 10.4512,
+      "step": 11760
+    },
+    {
+      "epoch": 0.058777996953731676,
+      "grad_norm": 0.09088432788848877,
+      "learning_rate": 2.8307341860872613e-05,
+      "loss": 10.4489,
+      "step": 11770
+    },
+    {
+      "epoch": 0.058827935778671125,
+      "grad_norm": 0.09308407455682755,
+      "learning_rate": 2.8305839945931063e-05,
+      "loss": 10.4484,
+      "step": 11780
+    },
+    {
+      "epoch": 0.058877874603610575,
+      "grad_norm": 0.09679020196199417,
+      "learning_rate": 2.8304338030989514e-05,
+      "loss": 10.4507,
+      "step": 11790
+    },
+    {
+      "epoch": 0.058927813428550024,
+      "grad_norm": 0.09807083010673523,
+      "learning_rate": 2.8302836116047964e-05,
+      "loss": 10.4473,
+      "step": 11800
+    },
+    {
+      "epoch": 0.058977752253489474,
+      "grad_norm": 0.09622442722320557,
+      "learning_rate": 2.830133420110641e-05,
+      "loss": 10.4472,
+      "step": 11810
+    },
+    {
+      "epoch": 0.05902769107842892,
+      "grad_norm": 0.09451454877853394,
+      "learning_rate": 2.829983228616486e-05,
+      "loss": 10.4465,
+      "step": 11820
+    },
+    {
+      "epoch": 0.05907762990336837,
+      "grad_norm": 0.10234472155570984,
+      "learning_rate": 2.829833037122331e-05,
+      "loss": 10.4481,
+      "step": 11830
+    },
+    {
+      "epoch": 0.05912756872830782,
+      "grad_norm": 0.09240438789129257,
+      "learning_rate": 2.829682845628176e-05,
+      "loss": 10.4469,
+      "step": 11840
+    },
+    {
+      "epoch": 0.05917750755324727,
+      "grad_norm": 0.09405849874019623,
+      "learning_rate": 2.829532654134021e-05,
+      "loss": 10.4481,
+      "step": 11850
+    },
+    {
+      "epoch": 0.05922744637818672,
+      "grad_norm": 0.08946629613637924,
+      "learning_rate": 2.8293824626398658e-05,
+      "loss": 10.4498,
+      "step": 11860
+    },
+    {
+      "epoch": 0.05927738520312617,
+      "grad_norm": 0.09469353407621384,
+      "learning_rate": 2.8292322711457108e-05,
+      "loss": 10.4447,
+      "step": 11870
+    },
+    {
+      "epoch": 0.05932732402806562,
+      "grad_norm": 0.09458359330892563,
+      "learning_rate": 2.829082079651556e-05,
+      "loss": 10.4443,
+      "step": 11880
+    },
+    {
+      "epoch": 0.05937726285300507,
+      "grad_norm": 0.08880256861448288,
+      "learning_rate": 2.828931888157401e-05,
+      "loss": 10.4459,
+      "step": 11890
+    },
+    {
+      "epoch": 0.05942720167794452,
+      "grad_norm": 0.09645318239927292,
+      "learning_rate": 2.828781696663246e-05,
+      "loss": 10.4466,
+      "step": 11900
+    },
+    {
+      "epoch": 0.05947714050288397,
+      "grad_norm": 0.0943111777305603,
+      "learning_rate": 2.8286315051690905e-05,
+      "loss": 10.4462,
+      "step": 11910
+    },
+    {
+      "epoch": 0.05952707932782342,
+      "grad_norm": 0.09864623099565506,
+      "learning_rate": 2.8284813136749356e-05,
+      "loss": 10.4449,
+      "step": 11920
+    },
+    {
+      "epoch": 0.05957701815276287,
+      "grad_norm": 0.09035992622375488,
+      "learning_rate": 2.8283311221807806e-05,
+      "loss": 10.4435,
+      "step": 11930
+    },
+    {
+      "epoch": 0.05962695697770232,
+      "grad_norm": 0.09477786719799042,
+      "learning_rate": 2.8281809306866256e-05,
+      "loss": 10.4472,
+      "step": 11940
+    },
+    {
+      "epoch": 0.059676895802641766,
+      "grad_norm": 0.09425050765275955,
+      "learning_rate": 2.8280307391924706e-05,
+      "loss": 10.4472,
+      "step": 11950
+    },
+    {
+      "epoch": 0.059726834627581216,
+      "grad_norm": 0.0916745588183403,
+      "learning_rate": 2.8278805476983153e-05,
+      "loss": 10.4452,
+      "step": 11960
+    },
+    {
+      "epoch": 0.059776773452520665,
+      "grad_norm": 0.09402048587799072,
+      "learning_rate": 2.8277303562041603e-05,
+      "loss": 10.4401,
+      "step": 11970
+    },
+    {
+      "epoch": 0.059826712277460115,
+      "grad_norm": 0.09692014008760452,
+      "learning_rate": 2.8275801647100053e-05,
+      "loss": 10.444,
+      "step": 11980
+    },
+    {
+      "epoch": 0.05987665110239956,
+      "grad_norm": 0.09546181559562683,
+      "learning_rate": 2.8274299732158504e-05,
+      "loss": 10.4423,
+      "step": 11990
+    },
+    {
+      "epoch": 0.059926589927339007,
+      "grad_norm": 0.09214867651462555,
+      "learning_rate": 2.8272797817216954e-05,
+      "loss": 10.4401,
+      "step": 12000
+    },
+    {
+      "epoch": 0.059976528752278456,
+      "grad_norm": 0.09537333995103836,
+      "learning_rate": 2.82712959022754e-05,
+      "loss": 10.4422,
+      "step": 12010
+    },
+    {
+      "epoch": 0.060026467577217905,
+      "grad_norm": 0.0944821685552597,
+      "learning_rate": 2.826979398733385e-05,
+      "loss": 10.4448,
+      "step": 12020
+    },
+    {
+      "epoch": 0.060076406402157355,
+      "grad_norm": 0.08920750021934509,
+      "learning_rate": 2.82682920723923e-05,
+      "loss": 10.4432,
+      "step": 12030
+    },
+    {
+      "epoch": 0.060126345227096804,
+      "grad_norm": 0.09650328755378723,
+      "learning_rate": 2.826679015745075e-05,
+      "loss": 10.4419,
+      "step": 12040
+    },
+    {
+      "epoch": 0.060176284052036254,
+      "grad_norm": 0.09876326471567154,
+      "learning_rate": 2.82652882425092e-05,
+      "loss": 10.439,
+      "step": 12050
+    },
+    {
+      "epoch": 0.0602262228769757,
+      "grad_norm": 0.09975486248731613,
+      "learning_rate": 2.826378632756765e-05,
+      "loss": 10.439,
+      "step": 12060
+    },
+    {
+      "epoch": 0.06027616170191515,
+      "grad_norm": 0.09765785932540894,
+      "learning_rate": 2.8262284412626098e-05,
+      "loss": 10.4414,
+      "step": 12070
+    },
+    {
+      "epoch": 0.0603261005268546,
+      "grad_norm": 0.09784495830535889,
+      "learning_rate": 2.826078249768455e-05,
+      "loss": 10.4388,
+      "step": 12080
+    },
+    {
+      "epoch": 0.06037603935179405,
+      "grad_norm": 0.09981703758239746,
+      "learning_rate": 2.8259280582743e-05,
+      "loss": 10.4412,
+      "step": 12090
+    },
+    {
+      "epoch": 0.0604259781767335,
+      "grad_norm": 0.09435340017080307,
+      "learning_rate": 2.825777866780145e-05,
+      "loss": 10.439,
+      "step": 12100
+    },
+    {
+      "epoch": 0.06047591700167295,
+      "grad_norm": 0.10208732634782791,
+      "learning_rate": 2.82562767528599e-05,
+      "loss": 10.4376,
+      "step": 12110
+    },
+    {
+      "epoch": 0.0605258558266124,
+      "grad_norm": 0.09435919672250748,
+      "learning_rate": 2.8254774837918346e-05,
+      "loss": 10.44,
+      "step": 12120
+    },
+    {
+      "epoch": 0.06057579465155185,
+      "grad_norm": 0.09108801931142807,
+      "learning_rate": 2.8253272922976796e-05,
+      "loss": 10.4375,
+      "step": 12130
+    },
+    {
+      "epoch": 0.0606257334764913,
+      "grad_norm": 0.0965215414762497,
+      "learning_rate": 2.8251771008035246e-05,
+      "loss": 10.439,
+      "step": 12140
+    },
+    {
+      "epoch": 0.06067567230143075,
+      "grad_norm": 0.09519945085048676,
+      "learning_rate": 2.8250269093093696e-05,
+      "loss": 10.4363,
+      "step": 12150
+    },
+    {
+      "epoch": 0.0607256111263702,
+      "grad_norm": 0.09803465753793716,
+      "learning_rate": 2.8248767178152146e-05,
+      "loss": 10.4383,
+      "step": 12160
+    },
+    {
+      "epoch": 0.06077554995130965,
+      "grad_norm": 0.09530342370271683,
+      "learning_rate": 2.8247265263210593e-05,
+      "loss": 10.438,
+      "step": 12170
+    },
+    {
+      "epoch": 0.0608254887762491,
+      "grad_norm": 0.09839750826358795,
+      "learning_rate": 2.8245763348269043e-05,
+      "loss": 10.4355,
+      "step": 12180
+    },
+    {
+      "epoch": 0.060875427601188546,
+      "grad_norm": 0.09567271173000336,
+      "learning_rate": 2.8244261433327494e-05,
+      "loss": 10.4368,
+      "step": 12190
+    },
+    {
+      "epoch": 0.060925366426127996,
+      "grad_norm": 0.09974398463964462,
+      "learning_rate": 2.8242759518385944e-05,
+      "loss": 10.4357,
+      "step": 12200
+    },
+    {
+      "epoch": 0.060975305251067445,
+      "grad_norm": 0.09306304156780243,
+      "learning_rate": 2.8241257603444394e-05,
+      "loss": 10.4359,
+      "step": 12210
+    },
+    {
+      "epoch": 0.061025244076006895,
+      "grad_norm": 0.09435778856277466,
+      "learning_rate": 2.823975568850284e-05,
+      "loss": 10.4371,
+      "step": 12220
+    },
+    {
+      "epoch": 0.061075182900946344,
+      "grad_norm": 0.10123757272958755,
+      "learning_rate": 2.823825377356129e-05,
+      "loss": 10.4379,
+      "step": 12230
+    },
+    {
+      "epoch": 0.061125121725885787,
+      "grad_norm": 0.09351859986782074,
+      "learning_rate": 2.823675185861974e-05,
+      "loss": 10.434,
+      "step": 12240
+    },
+    {
+      "epoch": 0.061175060550825236,
+      "grad_norm": 0.09679336100816727,
+      "learning_rate": 2.823524994367819e-05,
+      "loss": 10.4362,
+      "step": 12250
+    },
+    {
+      "epoch": 0.061224999375764685,
+      "grad_norm": 0.09072909504175186,
+      "learning_rate": 2.823374802873664e-05,
+      "loss": 10.4377,
+      "step": 12260
+    },
+    {
+      "epoch": 0.061274938200704135,
+      "grad_norm": 0.08954602479934692,
+      "learning_rate": 2.8232246113795088e-05,
+      "loss": 10.4366,
+      "step": 12270
+    },
+    {
+      "epoch": 0.061324877025643584,
+      "grad_norm": 0.0955163910984993,
+      "learning_rate": 2.823074419885354e-05,
+      "loss": 10.4351,
+      "step": 12280
+    },
+    {
+      "epoch": 0.061374815850583034,
+      "grad_norm": 0.09503049403429031,
+      "learning_rate": 2.822924228391199e-05,
+      "loss": 10.4351,
+      "step": 12290
+    },
+    {
+      "epoch": 0.06142475467552248,
+      "grad_norm": 0.10008996725082397,
+      "learning_rate": 2.822774036897044e-05,
+      "loss": 10.4331,
+      "step": 12300
+    },
+    {
+      "epoch": 0.06147469350046193,
+      "grad_norm": 0.1014408990740776,
+      "learning_rate": 2.822623845402889e-05,
+      "loss": 10.4347,
+      "step": 12310
+    },
+    {
+      "epoch": 0.06152463232540138,
+      "grad_norm": 0.09109607338905334,
+      "learning_rate": 2.8224736539087336e-05,
+      "loss": 10.4333,
+      "step": 12320
+    },
+    {
+      "epoch": 0.06157457115034083,
+      "grad_norm": 0.09781739115715027,
+      "learning_rate": 2.8223234624145786e-05,
+      "loss": 10.4332,
+      "step": 12330
+    },
+    {
+      "epoch": 0.06162450997528028,
+      "grad_norm": 0.09548245370388031,
+      "learning_rate": 2.8221732709204236e-05,
+      "loss": 10.4314,
+      "step": 12340
+    },
+    {
+      "epoch": 0.06167444880021973,
+      "grad_norm": 0.09480303525924683,
+      "learning_rate": 2.8220230794262686e-05,
+      "loss": 10.43,
+      "step": 12350
+    },
+    {
+      "epoch": 0.06172438762515918,
+      "grad_norm": 0.09702970832586288,
+      "learning_rate": 2.8218728879321136e-05,
+      "loss": 10.433,
+      "step": 12360
+    },
+    {
+      "epoch": 0.06177432645009863,
+      "grad_norm": 0.09529626369476318,
+      "learning_rate": 2.8217226964379583e-05,
+      "loss": 10.4315,
+      "step": 12370
+    },
+    {
+      "epoch": 0.06182426527503808,
+      "grad_norm": 0.09766676276922226,
+      "learning_rate": 2.8215725049438033e-05,
+      "loss": 10.4315,
+      "step": 12380
+    },
+    {
+      "epoch": 0.06187420409997753,
+      "grad_norm": 0.09520022571086884,
+      "learning_rate": 2.8214223134496484e-05,
+      "loss": 10.4313,
+      "step": 12390
+    },
+    {
+      "epoch": 0.06192414292491698,
+      "grad_norm": 0.09235453605651855,
+      "learning_rate": 2.8212721219554934e-05,
+      "loss": 10.4308,
+      "step": 12400
+    },
+    {
+      "epoch": 0.06197408174985643,
+      "grad_norm": 0.09697452187538147,
+      "learning_rate": 2.8211219304613384e-05,
+      "loss": 10.4296,
+      "step": 12410
+    },
+    {
+      "epoch": 0.06202402057479588,
+      "grad_norm": 0.09309589117765427,
+      "learning_rate": 2.820971738967183e-05,
+      "loss": 10.4306,
+      "step": 12420
+    },
+    {
+      "epoch": 0.062073959399735326,
+      "grad_norm": 0.09192889928817749,
+      "learning_rate": 2.8208215474730284e-05,
+      "loss": 10.4305,
+      "step": 12430
+    },
+    {
+      "epoch": 0.062123898224674776,
+      "grad_norm": 0.09932176768779755,
+      "learning_rate": 2.820671355978873e-05,
+      "loss": 10.4302,
+      "step": 12440
+    },
+    {
+      "epoch": 0.062173837049614225,
+      "grad_norm": 0.09811878204345703,
+      "learning_rate": 2.820521164484718e-05,
+      "loss": 10.429,
+      "step": 12450
+    },
+    {
+      "epoch": 0.062223775874553675,
+      "grad_norm": 0.09283413738012314,
+      "learning_rate": 2.820370972990563e-05,
+      "loss": 10.4304,
+      "step": 12460
+    },
+    {
+      "epoch": 0.062273714699493124,
+      "grad_norm": 0.09722374379634857,
+      "learning_rate": 2.8202207814964078e-05,
+      "loss": 10.429,
+      "step": 12470
+    },
+    {
+      "epoch": 0.06232365352443257,
+      "grad_norm": 0.09965924918651581,
+      "learning_rate": 2.8200705900022532e-05,
+      "loss": 10.4291,
+      "step": 12480
+    },
+    {
+      "epoch": 0.062373592349372016,
+      "grad_norm": 0.09636944532394409,
+      "learning_rate": 2.819920398508098e-05,
+      "loss": 10.4304,
+      "step": 12490
+    },
+    {
+      "epoch": 0.062423531174311465,
+      "grad_norm": 0.09442856162786484,
+      "learning_rate": 2.819770207013943e-05,
+      "loss": 10.4282,
+      "step": 12500
+    },
+    {
+      "epoch": 0.062473469999250915,
+      "grad_norm": 0.09579756110906601,
+      "learning_rate": 2.819620015519788e-05,
+      "loss": 10.4286,
+      "step": 12510
+    },
+    {
+      "epoch": 0.06252340882419037,
+      "grad_norm": 0.09269638359546661,
+      "learning_rate": 2.8194698240256326e-05,
+      "loss": 10.4284,
+      "step": 12520
+    },
+    {
+      "epoch": 0.06257334764912982,
+      "grad_norm": 0.0913536548614502,
+      "learning_rate": 2.819319632531478e-05,
+      "loss": 10.4262,
+      "step": 12530
+    },
+    {
+      "epoch": 0.06262328647406927,
+      "grad_norm": 0.09674961864948273,
+      "learning_rate": 2.8191694410373226e-05,
+      "loss": 10.4301,
+      "step": 12540
+    },
+    {
+      "epoch": 0.06267322529900872,
+      "grad_norm": 0.09084776788949966,
+      "learning_rate": 2.8190192495431676e-05,
+      "loss": 10.4284,
+      "step": 12550
+    },
+    {
+      "epoch": 0.06272316412394817,
+      "grad_norm": 0.09603569656610489,
+      "learning_rate": 2.8188690580490126e-05,
+      "loss": 10.4262,
+      "step": 12560
+    },
+    {
+      "epoch": 0.06277310294888762,
+      "grad_norm": 0.100166916847229,
+      "learning_rate": 2.8187188665548573e-05,
+      "loss": 10.4253,
+      "step": 12570
+    },
+    {
+      "epoch": 0.06282304177382707,
+      "grad_norm": 0.09294571727514267,
+      "learning_rate": 2.8185686750607027e-05,
+      "loss": 10.4256,
+      "step": 12580
+    },
+    {
+      "epoch": 0.06287298059876652,
+      "grad_norm": 0.09238424152135849,
+      "learning_rate": 2.8184184835665474e-05,
+      "loss": 10.427,
+      "step": 12590
+    },
+    {
+      "epoch": 0.06292291942370597,
+      "grad_norm": 0.09537918120622635,
+      "learning_rate": 2.8182682920723924e-05,
+      "loss": 10.425,
+      "step": 12600
+    },
+    {
+      "epoch": 0.0629728582486454,
+      "grad_norm": 0.09167756885290146,
+      "learning_rate": 2.8181181005782374e-05,
+      "loss": 10.4268,
+      "step": 12610
+    },
+    {
+      "epoch": 0.06302279707358485,
+      "grad_norm": 0.09255441278219223,
+      "learning_rate": 2.817967909084082e-05,
+      "loss": 10.4283,
+      "step": 12620
+    },
+    {
+      "epoch": 0.0630727358985243,
+      "grad_norm": 0.09267347306013107,
+      "learning_rate": 2.8178177175899274e-05,
+      "loss": 10.4236,
+      "step": 12630
+    },
+    {
+      "epoch": 0.06312267472346375,
+      "grad_norm": 0.09058328717947006,
+      "learning_rate": 2.817667526095772e-05,
+      "loss": 10.4257,
+      "step": 12640
+    },
+    {
+      "epoch": 0.0631726135484032,
+      "grad_norm": 0.09093445539474487,
+      "learning_rate": 2.817517334601617e-05,
+      "loss": 10.4247,
+      "step": 12650
+    },
+    {
+      "epoch": 0.06322255237334265,
+      "grad_norm": 0.09584364295005798,
+      "learning_rate": 2.817367143107462e-05,
+      "loss": 10.4212,
+      "step": 12660
+    },
+    {
+      "epoch": 0.0632724911982821,
+      "grad_norm": 0.09648899734020233,
+      "learning_rate": 2.8172169516133068e-05,
+      "loss": 10.4236,
+      "step": 12670
+    },
+    {
+      "epoch": 0.06332243002322155,
+      "grad_norm": 0.10392139106988907,
+      "learning_rate": 2.8170667601191522e-05,
+      "loss": 10.4231,
+      "step": 12680
+    },
+    {
+      "epoch": 0.063372368848161,
+      "grad_norm": 0.09030050039291382,
+      "learning_rate": 2.816916568624997e-05,
+      "loss": 10.424,
+      "step": 12690
+    },
+    {
+      "epoch": 0.06342230767310045,
+      "grad_norm": 0.0974743589758873,
+      "learning_rate": 2.816766377130842e-05,
+      "loss": 10.4226,
+      "step": 12700
+    },
+    {
+      "epoch": 0.0634722464980399,
+      "grad_norm": 0.09758103638887405,
+      "learning_rate": 2.816616185636687e-05,
+      "loss": 10.4236,
+      "step": 12710
+    },
+    {
+      "epoch": 0.06352218532297935,
+      "grad_norm": 0.09443262219429016,
+      "learning_rate": 2.8164659941425316e-05,
+      "loss": 10.4204,
+      "step": 12720
+    },
+    {
+      "epoch": 0.0635721241479188,
+      "grad_norm": 0.09165245294570923,
+      "learning_rate": 2.816315802648377e-05,
+      "loss": 10.4208,
+      "step": 12730
+    },
+    {
+      "epoch": 0.06362206297285825,
+      "grad_norm": 0.09566860646009445,
+      "learning_rate": 2.8161656111542216e-05,
+      "loss": 10.4215,
+      "step": 12740
+    },
+    {
+      "epoch": 0.0636720017977977,
+      "grad_norm": 0.0996147021651268,
+      "learning_rate": 2.816015419660067e-05,
+      "loss": 10.423,
+      "step": 12750
+    },
+    {
+      "epoch": 0.06372194062273714,
+      "grad_norm": 0.09040533006191254,
+      "learning_rate": 2.8158652281659116e-05,
+      "loss": 10.422,
+      "step": 12760
+    },
+    {
+      "epoch": 0.0637718794476766,
+      "grad_norm": 0.09722060710191727,
+      "learning_rate": 2.8157150366717563e-05,
+      "loss": 10.4204,
+      "step": 12770
+    },
+    {
+      "epoch": 0.06382181827261604,
+      "grad_norm": 0.09629692882299423,
+      "learning_rate": 2.8155648451776017e-05,
+      "loss": 10.4208,
+      "step": 12780
+    },
+    {
+      "epoch": 0.06387175709755549,
+      "grad_norm": 0.08845196664333344,
+      "learning_rate": 2.8154146536834464e-05,
+      "loss": 10.4187,
+      "step": 12790
+    },
+    {
+      "epoch": 0.06392169592249494,
+      "grad_norm": 0.0945558026432991,
+      "learning_rate": 2.8152644621892917e-05,
+      "loss": 10.4237,
+      "step": 12800
+    },
+    {
+      "epoch": 0.06397163474743439,
+      "grad_norm": 0.09783702343702316,
+      "learning_rate": 2.8151142706951364e-05,
+      "loss": 10.4185,
+      "step": 12810
+    },
+    {
+      "epoch": 0.06402157357237384,
+      "grad_norm": 0.09721019119024277,
+      "learning_rate": 2.814964079200981e-05,
+      "loss": 10.4202,
+      "step": 12820
+    },
+    {
+      "epoch": 0.06407151239731329,
+      "grad_norm": 0.09728706628084183,
+      "learning_rate": 2.8148138877068264e-05,
+      "loss": 10.4211,
+      "step": 12830
+    },
+    {
+      "epoch": 0.06412145122225274,
+      "grad_norm": 0.0896829217672348,
+      "learning_rate": 2.814663696212671e-05,
+      "loss": 10.4196,
+      "step": 12840
+    },
+    {
+      "epoch": 0.06417139004719219,
+      "grad_norm": 0.09445061534643173,
+      "learning_rate": 2.8145135047185165e-05,
+      "loss": 10.4193,
+      "step": 12850
+    },
+    {
+      "epoch": 0.06422132887213164,
+      "grad_norm": 0.0961589440703392,
+      "learning_rate": 2.814363313224361e-05,
+      "loss": 10.4175,
+      "step": 12860
+    },
+    {
+      "epoch": 0.06427126769707109,
+      "grad_norm": 0.09850755333900452,
+      "learning_rate": 2.8142131217302058e-05,
+      "loss": 10.4181,
+      "step": 12870
+    },
+    {
+      "epoch": 0.06432120652201054,
+      "grad_norm": 0.093666210770607,
+      "learning_rate": 2.8140629302360512e-05,
+      "loss": 10.419,
+      "step": 12880
+    },
+    {
+      "epoch": 0.06437114534694999,
+      "grad_norm": 0.09768545627593994,
+      "learning_rate": 2.813912738741896e-05,
+      "loss": 10.417,
+      "step": 12890
+    },
+    {
+      "epoch": 0.06442108417188944,
+      "grad_norm": 0.0899498239159584,
+      "learning_rate": 2.8137625472477412e-05,
+      "loss": 10.4175,
+      "step": 12900
+    },
+    {
+      "epoch": 0.06447102299682889,
+      "grad_norm": 0.09297134727239609,
+      "learning_rate": 2.813612355753586e-05,
+      "loss": 10.418,
+      "step": 12910
+    },
+    {
+      "epoch": 0.06452096182176834,
+      "grad_norm": 0.09266901016235352,
+      "learning_rate": 2.8134621642594306e-05,
+      "loss": 10.4167,
+      "step": 12920
+    },
+    {
+      "epoch": 0.06457090064670779,
+      "grad_norm": 0.09560717642307281,
+      "learning_rate": 2.813311972765276e-05,
+      "loss": 10.4158,
+      "step": 12930
+    },
+    {
+      "epoch": 0.06462083947164723,
+      "grad_norm": 0.09199609607458115,
+      "learning_rate": 2.8131617812711206e-05,
+      "loss": 10.4151,
+      "step": 12940
+    },
+    {
+      "epoch": 0.06467077829658668,
+      "grad_norm": 0.08880805224180222,
+      "learning_rate": 2.813011589776966e-05,
+      "loss": 10.4152,
+      "step": 12950
+    },
+    {
+      "epoch": 0.06472071712152613,
+      "grad_norm": 0.09574782103300095,
+      "learning_rate": 2.8128613982828106e-05,
+      "loss": 10.4178,
+      "step": 12960
+    },
+    {
+      "epoch": 0.06477065594646558,
+      "grad_norm": 0.09296387434005737,
+      "learning_rate": 2.8127112067886553e-05,
+      "loss": 10.415,
+      "step": 12970
+    },
+    {
+      "epoch": 0.06482059477140503,
+      "grad_norm": 0.09417938441038132,
+      "learning_rate": 2.8125610152945007e-05,
+      "loss": 10.4148,
+      "step": 12980
+    },
+    {
+      "epoch": 0.06487053359634448,
+      "grad_norm": 0.09101897478103638,
+      "learning_rate": 2.8124108238003454e-05,
+      "loss": 10.4149,
+      "step": 12990
+    },
+    {
+      "epoch": 0.06492047242128393,
+      "grad_norm": 0.09890080988407135,
+      "learning_rate": 2.8122606323061907e-05,
+      "loss": 10.4157,
+      "step": 13000
+    },
+    {
+      "epoch": 0.06497041124622338,
+      "grad_norm": 0.09360602498054504,
+      "learning_rate": 2.8121104408120354e-05,
+      "loss": 10.4141,
+      "step": 13010
+    },
+    {
+      "epoch": 0.06502035007116283,
+      "grad_norm": 0.10021226108074188,
+      "learning_rate": 2.81196024931788e-05,
+      "loss": 10.4117,
+      "step": 13020
+    },
+    {
+      "epoch": 0.06507028889610228,
+      "grad_norm": 0.09726858139038086,
+      "learning_rate": 2.8118100578237254e-05,
+      "loss": 10.4117,
+      "step": 13030
+    },
+    {
+      "epoch": 0.06512022772104173,
+      "grad_norm": 0.09342258423566818,
+      "learning_rate": 2.81165986632957e-05,
+      "loss": 10.4155,
+      "step": 13040
+    },
+    {
+      "epoch": 0.06517016654598118,
+      "grad_norm": 0.09800725430250168,
+      "learning_rate": 2.8115096748354155e-05,
+      "loss": 10.4146,
+      "step": 13050
+    },
+    {
+      "epoch": 0.06522010537092063,
+      "grad_norm": 0.09531977027654648,
+      "learning_rate": 2.81135948334126e-05,
+      "loss": 10.4122,
+      "step": 13060
+    },
+    {
+      "epoch": 0.06527004419586008,
+      "grad_norm": 0.09499124437570572,
+      "learning_rate": 2.811209291847105e-05,
+      "loss": 10.4126,
+      "step": 13070
+    },
+    {
+      "epoch": 0.06531998302079953,
+      "grad_norm": 0.09590071439743042,
+      "learning_rate": 2.8110591003529502e-05,
+      "loss": 10.4127,
+      "step": 13080
+    },
+    {
+      "epoch": 0.06536992184573898,
+      "grad_norm": 0.09835195541381836,
+      "learning_rate": 2.810908908858795e-05,
+      "loss": 10.4109,
+      "step": 13090
+    },
+    {
+      "epoch": 0.06541986067067843,
+      "grad_norm": 0.09182047098875046,
+      "learning_rate": 2.8107587173646402e-05,
+      "loss": 10.4132,
+      "step": 13100
+    },
+    {
+      "epoch": 0.06546979949561786,
+      "grad_norm": 0.09572348743677139,
+      "learning_rate": 2.810608525870485e-05,
+      "loss": 10.4111,
+      "step": 13110
+    },
+    {
+      "epoch": 0.06551973832055731,
+      "grad_norm": 0.09192756563425064,
+      "learning_rate": 2.81045833437633e-05,
+      "loss": 10.4094,
+      "step": 13120
+    },
+    {
+      "epoch": 0.06556967714549676,
+      "grad_norm": 0.09450101852416992,
+      "learning_rate": 2.810308142882175e-05,
+      "loss": 10.4124,
+      "step": 13130
+    },
+    {
+      "epoch": 0.06561961597043621,
+      "grad_norm": 0.09333983063697815,
+      "learning_rate": 2.8101579513880196e-05,
+      "loss": 10.4101,
+      "step": 13140
+    },
+    {
+      "epoch": 0.06566955479537566,
+      "grad_norm": 0.09769093245267868,
+      "learning_rate": 2.810007759893865e-05,
+      "loss": 10.4109,
+      "step": 13150
+    },
+    {
+      "epoch": 0.06571949362031511,
+      "grad_norm": 0.09517354518175125,
+      "learning_rate": 2.8098575683997096e-05,
+      "loss": 10.4101,
+      "step": 13160
+    },
+    {
+      "epoch": 0.06576943244525456,
+      "grad_norm": 0.09823792427778244,
+      "learning_rate": 2.8097073769055547e-05,
+      "loss": 10.4095,
+      "step": 13170
+    },
+    {
+      "epoch": 0.06581937127019401,
+      "grad_norm": 0.09689707309007645,
+      "learning_rate": 2.8095571854113997e-05,
+      "loss": 10.4104,
+      "step": 13180
+    },
+    {
+      "epoch": 0.06586931009513346,
+      "grad_norm": 0.09225734323263168,
+      "learning_rate": 2.8094069939172444e-05,
+      "loss": 10.4067,
+      "step": 13190
+    },
+    {
+      "epoch": 0.0659192489200729,
+      "grad_norm": 0.10383229702711105,
+      "learning_rate": 2.8092568024230897e-05,
+      "loss": 10.41,
+      "step": 13200
+    },
+    {
+      "epoch": 0.06596918774501236,
+      "grad_norm": 0.09270164370536804,
+      "learning_rate": 2.8091066109289344e-05,
+      "loss": 10.4071,
+      "step": 13210
+    },
+    {
+      "epoch": 0.0660191265699518,
+      "grad_norm": 0.09225078672170639,
+      "learning_rate": 2.8089564194347794e-05,
+      "loss": 10.4071,
+      "step": 13220
+    },
+    {
+      "epoch": 0.06606906539489125,
+      "grad_norm": 0.09302686154842377,
+      "learning_rate": 2.8088062279406244e-05,
+      "loss": 10.4091,
+      "step": 13230
+    },
+    {
+      "epoch": 0.0661190042198307,
+      "grad_norm": 0.09300468862056732,
+      "learning_rate": 2.808656036446469e-05,
+      "loss": 10.4079,
+      "step": 13240
+    },
+    {
+      "epoch": 0.06616894304477015,
+      "grad_norm": 0.10051462054252625,
+      "learning_rate": 2.8085058449523145e-05,
+      "loss": 10.4058,
+      "step": 13250
+    },
+    {
+      "epoch": 0.0662188818697096,
+      "grad_norm": 0.10076721012592316,
+      "learning_rate": 2.808355653458159e-05,
+      "loss": 10.4052,
+      "step": 13260
+    },
+    {
+      "epoch": 0.06626882069464905,
+      "grad_norm": 0.09321360290050507,
+      "learning_rate": 2.808205461964004e-05,
+      "loss": 10.4091,
+      "step": 13270
+    },
+    {
+      "epoch": 0.0663187595195885,
+      "grad_norm": 0.09066988527774811,
+      "learning_rate": 2.8080552704698492e-05,
+      "loss": 10.4054,
+      "step": 13280
+    },
+    {
+      "epoch": 0.06636869834452795,
+      "grad_norm": 0.09279420971870422,
+      "learning_rate": 2.807905078975694e-05,
+      "loss": 10.4051,
+      "step": 13290
+    },
+    {
+      "epoch": 0.0664186371694674,
+      "grad_norm": 0.09202752262353897,
+      "learning_rate": 2.8077548874815392e-05,
+      "loss": 10.4091,
+      "step": 13300
+    },
+    {
+      "epoch": 0.06646857599440685,
+      "grad_norm": 0.09188825637102127,
+      "learning_rate": 2.807604695987384e-05,
+      "loss": 10.4076,
+      "step": 13310
+    },
+    {
+      "epoch": 0.0665185148193463,
+      "grad_norm": 0.0955401360988617,
+      "learning_rate": 2.807454504493229e-05,
+      "loss": 10.406,
+      "step": 13320
+    },
+    {
+      "epoch": 0.06656845364428575,
+      "grad_norm": 0.09799564629793167,
+      "learning_rate": 2.807304312999074e-05,
+      "loss": 10.4049,
+      "step": 13330
+    },
+    {
+      "epoch": 0.0666183924692252,
+      "grad_norm": 0.09366542100906372,
+      "learning_rate": 2.8071541215049186e-05,
+      "loss": 10.4073,
+      "step": 13340
+    },
+    {
+      "epoch": 0.06666833129416465,
+      "grad_norm": 0.08726493269205093,
+      "learning_rate": 2.807003930010764e-05,
+      "loss": 10.4078,
+      "step": 13350
+    },
+    {
+      "epoch": 0.0667182701191041,
+      "grad_norm": 0.09408682584762573,
+      "learning_rate": 2.8068537385166086e-05,
+      "loss": 10.4018,
+      "step": 13360
+    },
+    {
+      "epoch": 0.06676820894404355,
+      "grad_norm": 0.1020817831158638,
+      "learning_rate": 2.8067035470224537e-05,
+      "loss": 10.4036,
+      "step": 13370
+    },
+    {
+      "epoch": 0.066818147768983,
+      "grad_norm": 0.0937168076634407,
+      "learning_rate": 2.8065533555282987e-05,
+      "loss": 10.405,
+      "step": 13380
+    },
+    {
+      "epoch": 0.06686808659392245,
+      "grad_norm": 0.08735665678977966,
+      "learning_rate": 2.8064031640341437e-05,
+      "loss": 10.4038,
+      "step": 13390
+    },
+    {
+      "epoch": 0.0669180254188619,
+      "grad_norm": 0.09232959151268005,
+      "learning_rate": 2.8062529725399887e-05,
+      "loss": 10.4007,
+      "step": 13400
+    },
+    {
+      "epoch": 0.06696796424380135,
+      "grad_norm": 0.09150397777557373,
+      "learning_rate": 2.8061027810458334e-05,
+      "loss": 10.4059,
+      "step": 13410
+    },
+    {
+      "epoch": 0.0670179030687408,
+      "grad_norm": 0.09197428822517395,
+      "learning_rate": 2.8059525895516784e-05,
+      "loss": 10.4031,
+      "step": 13420
+    },
+    {
+      "epoch": 0.06706784189368024,
+      "grad_norm": 0.09242592751979828,
+      "learning_rate": 2.8058023980575234e-05,
+      "loss": 10.402,
+      "step": 13430
+    },
+    {
+      "epoch": 0.0671177807186197,
+      "grad_norm": 0.09760529547929764,
+      "learning_rate": 2.8056522065633685e-05,
+      "loss": 10.402,
+      "step": 13440
+    },
+    {
+      "epoch": 0.06716771954355914,
+      "grad_norm": 0.09522455185651779,
+      "learning_rate": 2.8055020150692135e-05,
+      "loss": 10.4012,
+      "step": 13450
+    },
+    {
+      "epoch": 0.06721765836849859,
+      "grad_norm": 0.093946672976017,
+      "learning_rate": 2.805351823575058e-05,
+      "loss": 10.4044,
+      "step": 13460
+    },
+    {
+      "epoch": 0.06726759719343804,
+      "grad_norm": 0.09539136290550232,
+      "learning_rate": 2.805201632080903e-05,
+      "loss": 10.4038,
+      "step": 13470
+    },
+    {
+      "epoch": 0.06731753601837749,
+      "grad_norm": 0.10141439735889435,
+      "learning_rate": 2.8050514405867482e-05,
+      "loss": 10.4001,
+      "step": 13480
+    },
+    {
+      "epoch": 0.06736747484331694,
+      "grad_norm": 0.094613216817379,
+      "learning_rate": 2.8049012490925932e-05,
+      "loss": 10.4008,
+      "step": 13490
+    },
+    {
+      "epoch": 0.06741741366825639,
+      "grad_norm": 0.09187793731689453,
+      "learning_rate": 2.8047510575984382e-05,
+      "loss": 10.4017,
+      "step": 13500
+    },
+    {
+      "epoch": 0.06746735249319584,
+      "grad_norm": 0.10149643570184708,
+      "learning_rate": 2.804600866104283e-05,
+      "loss": 10.4028,
+      "step": 13510
+    },
+    {
+      "epoch": 0.06751729131813529,
+      "grad_norm": 0.09373277425765991,
+      "learning_rate": 2.804450674610128e-05,
+      "loss": 10.3996,
+      "step": 13520
+    },
+    {
+      "epoch": 0.06756723014307474,
+      "grad_norm": 0.09465789794921875,
+      "learning_rate": 2.804300483115973e-05,
+      "loss": 10.3994,
+      "step": 13530
+    },
+    {
+      "epoch": 0.06761716896801419,
+      "grad_norm": 0.08999823033809662,
+      "learning_rate": 2.804150291621818e-05,
+      "loss": 10.4001,
+      "step": 13540
+    },
+    {
+      "epoch": 0.06766710779295364,
+      "grad_norm": 0.09075751155614853,
+      "learning_rate": 2.804000100127663e-05,
+      "loss": 10.4001,
+      "step": 13550
+    },
+    {
+      "epoch": 0.06771704661789309,
+      "grad_norm": 0.09290343523025513,
+      "learning_rate": 2.8038499086335076e-05,
+      "loss": 10.3973,
+      "step": 13560
+    },
+    {
+      "epoch": 0.06776698544283254,
+      "grad_norm": 0.09069476276636124,
+      "learning_rate": 2.8036997171393527e-05,
+      "loss": 10.3977,
+      "step": 13570
+    },
+    {
+      "epoch": 0.06781692426777199,
+      "grad_norm": 0.09564889967441559,
+      "learning_rate": 2.8035495256451977e-05,
+      "loss": 10.4,
+      "step": 13580
+    },
+    {
+      "epoch": 0.06786686309271144,
+      "grad_norm": 0.09241501986980438,
+      "learning_rate": 2.8033993341510427e-05,
+      "loss": 10.3973,
+      "step": 13590
+    },
+    {
+      "epoch": 0.06791680191765087,
+      "grad_norm": 0.09421627223491669,
+      "learning_rate": 2.8032491426568877e-05,
+      "loss": 10.3985,
+      "step": 13600
+    },
+    {
+      "epoch": 0.06796674074259032,
+      "grad_norm": 0.0919337123632431,
+      "learning_rate": 2.8030989511627324e-05,
+      "loss": 10.3968,
+      "step": 13610
+    },
+    {
+      "epoch": 0.06801667956752977,
+      "grad_norm": 0.09475470334291458,
+      "learning_rate": 2.8029487596685774e-05,
+      "loss": 10.3991,
+      "step": 13620
+    },
+    {
+      "epoch": 0.06806661839246922,
+      "grad_norm": 0.09623359888792038,
+      "learning_rate": 2.8027985681744224e-05,
+      "loss": 10.3965,
+      "step": 13630
+    },
+    {
+      "epoch": 0.06811655721740867,
+      "grad_norm": 0.09497568756341934,
+      "learning_rate": 2.8026483766802675e-05,
+      "loss": 10.3988,
+      "step": 13640
+    },
+    {
+      "epoch": 0.06816649604234812,
+      "grad_norm": 0.09501062333583832,
+      "learning_rate": 2.8024981851861125e-05,
+      "loss": 10.3959,
+      "step": 13650
+    },
+    {
+      "epoch": 0.06821643486728757,
+      "grad_norm": 0.09278145432472229,
+      "learning_rate": 2.802347993691957e-05,
+      "loss": 10.3948,
+      "step": 13660
+    },
+    {
+      "epoch": 0.06826637369222702,
+      "grad_norm": 0.09373094141483307,
+      "learning_rate": 2.802197802197802e-05,
+      "loss": 10.395,
+      "step": 13670
+    },
+    {
+      "epoch": 0.06831631251716647,
+      "grad_norm": 0.09346432983875275,
+      "learning_rate": 2.8020476107036472e-05,
+      "loss": 10.396,
+      "step": 13680
+    },
+    {
+      "epoch": 0.06836625134210592,
+      "grad_norm": 0.09329260885715485,
+      "learning_rate": 2.8018974192094922e-05,
+      "loss": 10.3962,
+      "step": 13690
+    },
+    {
+      "epoch": 0.06841619016704537,
+      "grad_norm": 0.09555350244045258,
+      "learning_rate": 2.8017472277153372e-05,
+      "loss": 10.3954,
+      "step": 13700
+    },
+    {
+      "epoch": 0.06846612899198481,
+      "grad_norm": 0.09405256807804108,
+      "learning_rate": 2.8015970362211822e-05,
+      "loss": 10.3936,
+      "step": 13710
+    },
+    {
+      "epoch": 0.06851606781692426,
+      "grad_norm": 0.09874436259269714,
+      "learning_rate": 2.801446844727027e-05,
+      "loss": 10.3941,
+      "step": 13720
+    },
+    {
+      "epoch": 0.06856600664186371,
+      "grad_norm": 0.0999235138297081,
+      "learning_rate": 2.801296653232872e-05,
+      "loss": 10.3911,
+      "step": 13730
+    },
+    {
+      "epoch": 0.06861594546680316,
+      "grad_norm": 0.09541259706020355,
+      "learning_rate": 2.801146461738717e-05,
+      "loss": 10.3937,
+      "step": 13740
+    },
+    {
+      "epoch": 0.06866588429174261,
+      "grad_norm": 0.09852302819490433,
+      "learning_rate": 2.800996270244562e-05,
+      "loss": 10.3953,
+      "step": 13750
+    },
+    {
+      "epoch": 0.06871582311668206,
+      "grad_norm": 0.09277693927288055,
+      "learning_rate": 2.800846078750407e-05,
+      "loss": 10.3929,
+      "step": 13760
+    },
+    {
+      "epoch": 0.06876576194162151,
+      "grad_norm": 0.09327711910009384,
+      "learning_rate": 2.8006958872562517e-05,
+      "loss": 10.3907,
+      "step": 13770
+    },
+    {
+      "epoch": 0.06881570076656096,
+      "grad_norm": 0.08925262838602066,
+      "learning_rate": 2.8005456957620967e-05,
+      "loss": 10.3949,
+      "step": 13780
+    },
+    {
+      "epoch": 0.06886563959150041,
+      "grad_norm": 0.0973198264837265,
+      "learning_rate": 2.8003955042679417e-05,
+      "loss": 10.3892,
+      "step": 13790
+    },
+    {
+      "epoch": 0.06891557841643986,
+      "grad_norm": 0.09705457836389542,
+      "learning_rate": 2.8002453127737867e-05,
+      "loss": 10.3911,
+      "step": 13800
+    },
+    {
+      "epoch": 0.06896551724137931,
+      "grad_norm": 0.09261533617973328,
+      "learning_rate": 2.8000951212796317e-05,
+      "loss": 10.3915,
+      "step": 13810
+    },
+    {
+      "epoch": 0.06901545606631876,
+      "grad_norm": 0.09760327637195587,
+      "learning_rate": 2.7999449297854764e-05,
+      "loss": 10.3938,
+      "step": 13820
+    },
+    {
+      "epoch": 0.06906539489125821,
+      "grad_norm": 0.09349595010280609,
+      "learning_rate": 2.7997947382913214e-05,
+      "loss": 10.3904,
+      "step": 13830
+    },
+    {
+      "epoch": 0.06911533371619766,
+      "grad_norm": 0.0936850905418396,
+      "learning_rate": 2.7996445467971665e-05,
+      "loss": 10.3909,
+      "step": 13840
+    },
+    {
+      "epoch": 0.06916527254113711,
+      "grad_norm": 0.09198077023029327,
+      "learning_rate": 2.7994943553030115e-05,
+      "loss": 10.3907,
+      "step": 13850
+    },
+    {
+      "epoch": 0.06921521136607656,
+      "grad_norm": 0.09293072670698166,
+      "learning_rate": 2.7993441638088565e-05,
+      "loss": 10.3906,
+      "step": 13860
+    },
+    {
+      "epoch": 0.069265150191016,
+      "grad_norm": 0.10718231648206711,
+      "learning_rate": 2.7991939723147012e-05,
+      "loss": 10.388,
+      "step": 13870
+    },
+    {
+      "epoch": 0.06931508901595546,
+      "grad_norm": 0.09268306940793991,
+      "learning_rate": 2.7990437808205462e-05,
+      "loss": 10.3907,
+      "step": 13880
+    },
+    {
+      "epoch": 0.0693650278408949,
+      "grad_norm": 0.09039410948753357,
+      "learning_rate": 2.7988935893263912e-05,
+      "loss": 10.3901,
+      "step": 13890
+    },
+    {
+      "epoch": 0.06941496666583435,
+      "grad_norm": 0.09920943528413773,
+      "learning_rate": 2.7987433978322362e-05,
+      "loss": 10.3878,
+      "step": 13900
+    },
+    {
+      "epoch": 0.0694649054907738,
+      "grad_norm": 0.09512588381767273,
+      "learning_rate": 2.7985932063380812e-05,
+      "loss": 10.3883,
+      "step": 13910
+    },
+    {
+      "epoch": 0.06951484431571325,
+      "grad_norm": 0.0949581190943718,
+      "learning_rate": 2.798443014843926e-05,
+      "loss": 10.388,
+      "step": 13920
+    },
+    {
+      "epoch": 0.0695647831406527,
+      "grad_norm": 0.09540549665689468,
+      "learning_rate": 2.798292823349771e-05,
+      "loss": 10.3886,
+      "step": 13930
+    },
+    {
+      "epoch": 0.06961472196559215,
+      "grad_norm": 0.10153383016586304,
+      "learning_rate": 2.798142631855616e-05,
+      "loss": 10.3875,
+      "step": 13940
+    },
+    {
+      "epoch": 0.0696646607905316,
+      "grad_norm": 0.09448625147342682,
+      "learning_rate": 2.797992440361461e-05,
+      "loss": 10.3908,
+      "step": 13950
+    },
+    {
+      "epoch": 0.06971459961547105,
+      "grad_norm": 0.09156591445207596,
+      "learning_rate": 2.797842248867306e-05,
+      "loss": 10.3874,
+      "step": 13960
+    },
+    {
+      "epoch": 0.0697645384404105,
+      "grad_norm": 0.10002894699573517,
+      "learning_rate": 2.7976920573731507e-05,
+      "loss": 10.3878,
+      "step": 13970
+    },
+    {
+      "epoch": 0.06981447726534995,
+      "grad_norm": 0.1005663126707077,
+      "learning_rate": 2.7975418658789957e-05,
+      "loss": 10.3849,
+      "step": 13980
+    },
+    {
+      "epoch": 0.0698644160902894,
+      "grad_norm": 0.09588824957609177,
+      "learning_rate": 2.7973916743848407e-05,
+      "loss": 10.3876,
+      "step": 13990
+    },
+    {
+      "epoch": 0.06991435491522885,
+      "grad_norm": 0.10445947200059891,
+      "learning_rate": 2.7972414828906857e-05,
+      "loss": 10.3845,
+      "step": 14000
+    },
+    {
+      "epoch": 0.0699642937401683,
+      "grad_norm": 0.09666018187999725,
+      "learning_rate": 2.7970912913965307e-05,
+      "loss": 10.3877,
+      "step": 14010
+    },
+    {
+      "epoch": 0.07001423256510775,
+      "grad_norm": 0.09730076789855957,
+      "learning_rate": 2.7969410999023754e-05,
+      "loss": 10.3862,
+      "step": 14020
+    },
+    {
+      "epoch": 0.0700641713900472,
+      "grad_norm": 0.09373754262924194,
+      "learning_rate": 2.7967909084082208e-05,
+      "loss": 10.3866,
+      "step": 14030
+    },
+    {
+      "epoch": 0.07011411021498665,
+      "grad_norm": 0.09593051671981812,
+      "learning_rate": 2.7966407169140655e-05,
+      "loss": 10.3842,
+      "step": 14040
+    },
+    {
+      "epoch": 0.0701640490399261,
+      "grad_norm": 0.09878380596637726,
+      "learning_rate": 2.7964905254199105e-05,
+      "loss": 10.3858,
+      "step": 14050
+    },
+    {
+      "epoch": 0.07021398786486555,
+      "grad_norm": 0.09511502087116241,
+      "learning_rate": 2.7963403339257555e-05,
+      "loss": 10.3866,
+      "step": 14060
+    },
+    {
+      "epoch": 0.070263926689805,
+      "grad_norm": 0.09328655153512955,
+      "learning_rate": 2.7961901424316002e-05,
+      "loss": 10.3844,
+      "step": 14070
+    },
+    {
+      "epoch": 0.07031386551474444,
+      "grad_norm": 0.09099029004573822,
+      "learning_rate": 2.7960399509374455e-05,
+      "loss": 10.3844,
+      "step": 14080
+    },
+    {
+      "epoch": 0.0703638043396839,
+      "grad_norm": 0.09044034779071808,
+      "learning_rate": 2.7958897594432902e-05,
+      "loss": 10.383,
+      "step": 14090
+    },
+    {
+      "epoch": 0.07041374316462333,
+      "grad_norm": 0.09568076580762863,
+      "learning_rate": 2.7957395679491352e-05,
+      "loss": 10.3804,
+      "step": 14100
+    },
+    {
+      "epoch": 0.07046368198956278,
+      "grad_norm": 0.09366782754659653,
+      "learning_rate": 2.7955893764549802e-05,
+      "loss": 10.3835,
+      "step": 14110
+    },
+    {
+      "epoch": 0.07051362081450223,
+      "grad_norm": 0.0898858904838562,
+      "learning_rate": 2.795439184960825e-05,
+      "loss": 10.3813,
+      "step": 14120
+    },
+    {
+      "epoch": 0.07056355963944168,
+      "grad_norm": 0.09581518173217773,
+      "learning_rate": 2.7952889934666703e-05,
+      "loss": 10.3861,
+      "step": 14130
+    },
+    {
+      "epoch": 0.07061349846438113,
+      "grad_norm": 0.09709804505109787,
+      "learning_rate": 2.795138801972515e-05,
+      "loss": 10.3827,
+      "step": 14140
+    },
+    {
+      "epoch": 0.07066343728932058,
+      "grad_norm": 0.09556079655885696,
+      "learning_rate": 2.79498861047836e-05,
+      "loss": 10.3806,
+      "step": 14150
+    },
+    {
+      "epoch": 0.07071337611426003,
+      "grad_norm": 0.09610017389059067,
+      "learning_rate": 2.794838418984205e-05,
+      "loss": 10.3825,
+      "step": 14160
+    },
+    {
+      "epoch": 0.07076331493919948,
+      "grad_norm": 0.094112828373909,
+      "learning_rate": 2.7946882274900497e-05,
+      "loss": 10.3794,
+      "step": 14170
+    },
+    {
+      "epoch": 0.07081325376413893,
+      "grad_norm": 0.09225358068943024,
+      "learning_rate": 2.794538035995895e-05,
+      "loss": 10.3818,
+      "step": 14180
+    },
+    {
+      "epoch": 0.07086319258907837,
+      "grad_norm": 0.09522094577550888,
+      "learning_rate": 2.7943878445017397e-05,
+      "loss": 10.3811,
+      "step": 14190
+    },
+    {
+      "epoch": 0.07091313141401782,
+      "grad_norm": 0.09772941470146179,
+      "learning_rate": 2.7942376530075847e-05,
+      "loss": 10.3835,
+      "step": 14200
+    },
+    {
+      "epoch": 0.07096307023895727,
+      "grad_norm": 0.09984889626502991,
+      "learning_rate": 2.7940874615134297e-05,
+      "loss": 10.3794,
+      "step": 14210
+    },
+    {
+      "epoch": 0.07101300906389672,
+      "grad_norm": 0.09417369961738586,
+      "learning_rate": 2.7939372700192744e-05,
+      "loss": 10.3824,
+      "step": 14220
+    },
+    {
+      "epoch": 0.07106294788883617,
+      "grad_norm": 0.09816978126764297,
+      "learning_rate": 2.7937870785251198e-05,
+      "loss": 10.3796,
+      "step": 14230
+    },
+    {
+      "epoch": 0.07111288671377562,
+      "grad_norm": 0.10052361339330673,
+      "learning_rate": 2.7936368870309645e-05,
+      "loss": 10.3792,
+      "step": 14240
+    },
+    {
+      "epoch": 0.07116282553871507,
+      "grad_norm": 0.09660011529922485,
+      "learning_rate": 2.7934866955368095e-05,
+      "loss": 10.38,
+      "step": 14250
+    },
+    {
+      "epoch": 0.07121276436365452,
+      "grad_norm": 0.09375877678394318,
+      "learning_rate": 2.7933365040426545e-05,
+      "loss": 10.3788,
+      "step": 14260
+    },
+    {
+      "epoch": 0.07126270318859397,
+      "grad_norm": 0.09189482033252716,
+      "learning_rate": 2.7931863125484992e-05,
+      "loss": 10.3798,
+      "step": 14270
+    },
+    {
+      "epoch": 0.07131264201353342,
+      "grad_norm": 0.09980373829603195,
+      "learning_rate": 2.7930361210543445e-05,
+      "loss": 10.381,
+      "step": 14280
+    },
+    {
+      "epoch": 0.07136258083847287,
+      "grad_norm": 0.08954178541898727,
+      "learning_rate": 2.7928859295601892e-05,
+      "loss": 10.3781,
+      "step": 14290
+    },
+    {
+      "epoch": 0.07141251966341232,
+      "grad_norm": 0.08987992256879807,
+      "learning_rate": 2.7927357380660342e-05,
+      "loss": 10.3801,
+      "step": 14300
+    },
+    {
+      "epoch": 0.07146245848835177,
+      "grad_norm": 0.0954456776380539,
+      "learning_rate": 2.7925855465718792e-05,
+      "loss": 10.3775,
+      "step": 14310
+    },
+    {
+      "epoch": 0.07151239731329122,
+      "grad_norm": 0.09449037164449692,
+      "learning_rate": 2.792435355077724e-05,
+      "loss": 10.3813,
+      "step": 14320
+    },
+    {
+      "epoch": 0.07156233613823067,
+      "grad_norm": 0.09394067525863647,
+      "learning_rate": 2.7922851635835693e-05,
+      "loss": 10.3778,
+      "step": 14330
+    },
+    {
+      "epoch": 0.07161227496317012,
+      "grad_norm": 0.0954408049583435,
+      "learning_rate": 2.792134972089414e-05,
+      "loss": 10.3772,
+      "step": 14340
+    },
+    {
+      "epoch": 0.07166221378810957,
+      "grad_norm": 0.09376630187034607,
+      "learning_rate": 2.7919847805952593e-05,
+      "loss": 10.3763,
+      "step": 14350
+    },
+    {
+      "epoch": 0.07171215261304902,
+      "grad_norm": 0.0952349603176117,
+      "learning_rate": 2.791834589101104e-05,
+      "loss": 10.3765,
+      "step": 14360
+    },
+    {
+      "epoch": 0.07176209143798847,
+      "grad_norm": 0.10054927319288254,
+      "learning_rate": 2.7916843976069487e-05,
+      "loss": 10.3729,
+      "step": 14370
+    },
+    {
+      "epoch": 0.07181203026292791,
+      "grad_norm": 0.09942678362131119,
+      "learning_rate": 2.791534206112794e-05,
+      "loss": 10.3755,
+      "step": 14380
+    },
+    {
+      "epoch": 0.07186196908786736,
+      "grad_norm": 0.09826812148094177,
+      "learning_rate": 2.7913840146186387e-05,
+      "loss": 10.3753,
+      "step": 14390
+    },
+    {
+      "epoch": 0.07191190791280681,
+      "grad_norm": 0.09203541278839111,
+      "learning_rate": 2.791233823124484e-05,
+      "loss": 10.3753,
+      "step": 14400
+    },
+    {
+      "epoch": 0.07196184673774626,
+      "grad_norm": 0.09987717121839523,
+      "learning_rate": 2.7910836316303287e-05,
+      "loss": 10.3732,
+      "step": 14410
+    },
+    {
+      "epoch": 0.07201178556268571,
+      "grad_norm": 0.09702242910861969,
+      "learning_rate": 2.7909334401361734e-05,
+      "loss": 10.3748,
+      "step": 14420
+    },
+    {
+      "epoch": 0.07206172438762516,
+      "grad_norm": 0.09924949705600739,
+      "learning_rate": 2.7907832486420188e-05,
+      "loss": 10.3758,
+      "step": 14430
+    },
+    {
+      "epoch": 0.07211166321256461,
+      "grad_norm": 0.10308583825826645,
+      "learning_rate": 2.7906330571478635e-05,
+      "loss": 10.3698,
+      "step": 14440
+    },
+    {
+      "epoch": 0.07216160203750406,
+      "grad_norm": 0.09504105895757675,
+      "learning_rate": 2.7904828656537088e-05,
+      "loss": 10.3751,
+      "step": 14450
+    },
+    {
+      "epoch": 0.07221154086244351,
+      "grad_norm": 0.0910039022564888,
+      "learning_rate": 2.7903326741595535e-05,
+      "loss": 10.3762,
+      "step": 14460
+    },
+    {
+      "epoch": 0.07226147968738296,
+      "grad_norm": 0.09338394552469254,
+      "learning_rate": 2.7901824826653982e-05,
+      "loss": 10.3741,
+      "step": 14470
+    },
+    {
+      "epoch": 0.07231141851232241,
+      "grad_norm": 0.095545195043087,
+      "learning_rate": 2.7900322911712435e-05,
+      "loss": 10.3728,
+      "step": 14480
+    },
+    {
+      "epoch": 0.07236135733726186,
+      "grad_norm": 0.0895906612277031,
+      "learning_rate": 2.7898820996770882e-05,
+      "loss": 10.3754,
+      "step": 14490
+    },
+    {
+      "epoch": 0.07241129616220131,
+      "grad_norm": 0.09767705202102661,
+      "learning_rate": 2.7897319081829336e-05,
+      "loss": 10.3734,
+      "step": 14500
+    },
+    {
+      "epoch": 0.07246123498714076,
+      "grad_norm": 0.09386033564805984,
+      "learning_rate": 2.7895817166887782e-05,
+      "loss": 10.3728,
+      "step": 14510
+    },
+    {
+      "epoch": 0.0725111738120802,
+      "grad_norm": 0.09241237491369247,
+      "learning_rate": 2.789431525194623e-05,
+      "loss": 10.3727,
+      "step": 14520
+    },
+    {
+      "epoch": 0.07256111263701966,
+      "grad_norm": 0.09265079349279404,
+      "learning_rate": 2.7892813337004683e-05,
+      "loss": 10.3714,
+      "step": 14530
+    },
+    {
+      "epoch": 0.0726110514619591,
+      "grad_norm": 0.09971288591623306,
+      "learning_rate": 2.789131142206313e-05,
+      "loss": 10.3708,
+      "step": 14540
+    },
+    {
+      "epoch": 0.07266099028689856,
+      "grad_norm": 0.09628092497587204,
+      "learning_rate": 2.7889809507121583e-05,
+      "loss": 10.3714,
+      "step": 14550
+    },
+    {
+      "epoch": 0.072710929111838,
+      "grad_norm": 0.09946174919605255,
+      "learning_rate": 2.788830759218003e-05,
+      "loss": 10.3706,
+      "step": 14560
+    },
+    {
+      "epoch": 0.07276086793677745,
+      "grad_norm": 0.09489379823207855,
+      "learning_rate": 2.7886805677238477e-05,
+      "loss": 10.3715,
+      "step": 14570
+    },
+    {
+      "epoch": 0.0728108067617169,
+      "grad_norm": 0.09554824233055115,
+      "learning_rate": 2.788530376229693e-05,
+      "loss": 10.3715,
+      "step": 14580
+    },
+    {
+      "epoch": 0.07286074558665634,
+      "grad_norm": 0.09551879018545151,
+      "learning_rate": 2.7883801847355377e-05,
+      "loss": 10.3715,
+      "step": 14590
+    },
+    {
+      "epoch": 0.07291068441159579,
+      "grad_norm": 0.10136498510837555,
+      "learning_rate": 2.788229993241383e-05,
+      "loss": 10.371,
+      "step": 14600
+    },
+    {
+      "epoch": 0.07296062323653524,
+      "grad_norm": 0.09273940324783325,
+      "learning_rate": 2.7880798017472277e-05,
+      "loss": 10.3697,
+      "step": 14610
+    },
+    {
+      "epoch": 0.07301056206147469,
+      "grad_norm": 0.09797700494527817,
+      "learning_rate": 2.7879296102530724e-05,
+      "loss": 10.369,
+      "step": 14620
+    },
+    {
+      "epoch": 0.07306050088641414,
+      "grad_norm": 0.0974142774939537,
+      "learning_rate": 2.7877794187589178e-05,
+      "loss": 10.3689,
+      "step": 14630
+    },
+    {
+      "epoch": 0.07311043971135359,
+      "grad_norm": 0.10973493754863739,
+      "learning_rate": 2.7876292272647625e-05,
+      "loss": 10.3659,
+      "step": 14640
+    },
+    {
+      "epoch": 0.07316037853629304,
+      "grad_norm": 0.0951044112443924,
+      "learning_rate": 2.7874790357706078e-05,
+      "loss": 10.3671,
+      "step": 14650
+    },
+    {
+      "epoch": 0.07321031736123249,
+      "grad_norm": 0.09662732481956482,
+      "learning_rate": 2.7873288442764525e-05,
+      "loss": 10.369,
+      "step": 14660
+    },
+    {
+      "epoch": 0.07326025618617193,
+      "grad_norm": 0.09085254371166229,
+      "learning_rate": 2.7871786527822975e-05,
+      "loss": 10.3666,
+      "step": 14670
+    },
+    {
+      "epoch": 0.07331019501111138,
+      "grad_norm": 0.09421254694461823,
+      "learning_rate": 2.7870284612881425e-05,
+      "loss": 10.3648,
+      "step": 14680
+    },
+    {
+      "epoch": 0.07336013383605083,
+      "grad_norm": 0.09277543425559998,
+      "learning_rate": 2.7868782697939872e-05,
+      "loss": 10.3655,
+      "step": 14690
+    },
+    {
+      "epoch": 0.07341007266099028,
+      "grad_norm": 0.09704422950744629,
+      "learning_rate": 2.7867280782998326e-05,
+      "loss": 10.3657,
+      "step": 14700
+    },
+    {
+      "epoch": 0.07346001148592973,
+      "grad_norm": 0.09107200056314468,
+      "learning_rate": 2.7865778868056772e-05,
+      "loss": 10.3688,
+      "step": 14710
+    },
+    {
+      "epoch": 0.07350995031086918,
+      "grad_norm": 0.08991669118404388,
+      "learning_rate": 2.7864276953115223e-05,
+      "loss": 10.3676,
+      "step": 14720
+    },
+    {
+      "epoch": 0.07355988913580863,
+      "grad_norm": 0.09315147995948792,
+      "learning_rate": 2.7862775038173673e-05,
+      "loss": 10.3692,
+      "step": 14730
+    },
+    {
+      "epoch": 0.07360982796074808,
+      "grad_norm": 0.09433738887310028,
+      "learning_rate": 2.786127312323212e-05,
+      "loss": 10.364,
+      "step": 14740
+    },
+    {
+      "epoch": 0.07365976678568753,
+      "grad_norm": 0.0911378338932991,
+      "learning_rate": 2.7859771208290573e-05,
+      "loss": 10.3655,
+      "step": 14750
+    },
+    {
+      "epoch": 0.07370970561062698,
+      "grad_norm": 0.09321308135986328,
+      "learning_rate": 2.785826929334902e-05,
+      "loss": 10.3665,
+      "step": 14760
+    },
+    {
+      "epoch": 0.07375964443556643,
+      "grad_norm": 0.09273282438516617,
+      "learning_rate": 2.785676737840747e-05,
+      "loss": 10.3633,
+      "step": 14770
+    },
+    {
+      "epoch": 0.07380958326050588,
+      "grad_norm": 0.0901220440864563,
+      "learning_rate": 2.785526546346592e-05,
+      "loss": 10.3613,
+      "step": 14780
+    },
+    {
+      "epoch": 0.07385952208544533,
+      "grad_norm": 0.09423750638961792,
+      "learning_rate": 2.7853763548524367e-05,
+      "loss": 10.3643,
+      "step": 14790
+    },
+    {
+      "epoch": 0.07390946091038478,
+      "grad_norm": 0.09342154115438461,
+      "learning_rate": 2.785226163358282e-05,
+      "loss": 10.366,
+      "step": 14800
+    },
+    {
+      "epoch": 0.07395939973532423,
+      "grad_norm": 0.09549295902252197,
+      "learning_rate": 2.7850759718641267e-05,
+      "loss": 10.3637,
+      "step": 14810
+    },
+    {
+      "epoch": 0.07400933856026368,
+      "grad_norm": 0.09613852202892303,
+      "learning_rate": 2.7849257803699718e-05,
+      "loss": 10.3668,
+      "step": 14820
+    },
+    {
+      "epoch": 0.07405927738520313,
+      "grad_norm": 0.09182946383953094,
+      "learning_rate": 2.7847755888758168e-05,
+      "loss": 10.3647,
+      "step": 14830
+    },
+    {
+      "epoch": 0.07410921621014258,
+      "grad_norm": 0.08932546526193619,
+      "learning_rate": 2.7846253973816615e-05,
+      "loss": 10.3626,
+      "step": 14840
+    },
+    {
+      "epoch": 0.07415915503508202,
+      "grad_norm": 0.09887928515672684,
+      "learning_rate": 2.7844752058875068e-05,
+      "loss": 10.3609,
+      "step": 14850
+    },
+    {
+      "epoch": 0.07420909386002147,
+      "grad_norm": 0.09238041192293167,
+      "learning_rate": 2.7843250143933515e-05,
+      "loss": 10.3625,
+      "step": 14860
+    },
+    {
+      "epoch": 0.07425903268496092,
+      "grad_norm": 0.08984861522912979,
+      "learning_rate": 2.7841748228991965e-05,
+      "loss": 10.3626,
+      "step": 14870
+    },
+    {
+      "epoch": 0.07430897150990037,
+      "grad_norm": 0.09524226188659668,
+      "learning_rate": 2.7840246314050415e-05,
+      "loss": 10.3622,
+      "step": 14880
+    },
+    {
+      "epoch": 0.07435891033483982,
+      "grad_norm": 0.09836632758378983,
+      "learning_rate": 2.7838744399108862e-05,
+      "loss": 10.3668,
+      "step": 14890
+    },
+    {
+      "epoch": 0.07440884915977927,
+      "grad_norm": 0.08989029377698898,
+      "learning_rate": 2.7837242484167316e-05,
+      "loss": 10.364,
+      "step": 14900
+    },
+    {
+      "epoch": 0.07445878798471872,
+      "grad_norm": 0.09499987214803696,
+      "learning_rate": 2.7835740569225762e-05,
+      "loss": 10.3642,
+      "step": 14910
+    },
+    {
+      "epoch": 0.07450872680965817,
+      "grad_norm": 0.09698102623224258,
+      "learning_rate": 2.7834238654284213e-05,
+      "loss": 10.3599,
+      "step": 14920
+    },
+    {
+      "epoch": 0.07455866563459762,
+      "grad_norm": 0.09677815437316895,
+      "learning_rate": 2.7832736739342663e-05,
+      "loss": 10.3597,
+      "step": 14930
+    },
+    {
+      "epoch": 0.07460860445953707,
+      "grad_norm": 0.09893867373466492,
+      "learning_rate": 2.783123482440111e-05,
+      "loss": 10.358,
+      "step": 14940
+    },
+    {
+      "epoch": 0.07465854328447652,
+      "grad_norm": 0.09698786586523056,
+      "learning_rate": 2.7829732909459563e-05,
+      "loss": 10.3576,
+      "step": 14950
+    },
+    {
+      "epoch": 0.07470848210941597,
+      "grad_norm": 0.08816642314195633,
+      "learning_rate": 2.782823099451801e-05,
+      "loss": 10.3598,
+      "step": 14960
+    },
+    {
+      "epoch": 0.07475842093435542,
+      "grad_norm": 0.09455441683530807,
+      "learning_rate": 2.782672907957646e-05,
+      "loss": 10.3586,
+      "step": 14970
+    },
+    {
+      "epoch": 0.07480835975929487,
+      "grad_norm": 0.09937630593776703,
+      "learning_rate": 2.782522716463491e-05,
+      "loss": 10.3571,
+      "step": 14980
+    },
+    {
+      "epoch": 0.07485829858423432,
+      "grad_norm": 0.09834357351064682,
+      "learning_rate": 2.782372524969336e-05,
+      "loss": 10.357,
+      "step": 14990
+    },
+    {
+      "epoch": 0.07490823740917377,
+      "grad_norm": 0.09351138025522232,
+      "learning_rate": 2.782222333475181e-05,
+      "loss": 10.3599,
+      "step": 15000
+    },
+    {
+      "epoch": 0.07495817623411322,
+      "grad_norm": 0.09158463031053543,
+      "learning_rate": 2.7820721419810257e-05,
+      "loss": 10.3608,
+      "step": 15010
+    },
+    {
+      "epoch": 0.07500811505905267,
+      "grad_norm": 0.09933115541934967,
+      "learning_rate": 2.7819219504868708e-05,
+      "loss": 10.3587,
+      "step": 15020
+    },
+    {
+      "epoch": 0.07505805388399212,
+      "grad_norm": 0.09445559233427048,
+      "learning_rate": 2.7817717589927158e-05,
+      "loss": 10.356,
+      "step": 15030
+    },
+    {
+      "epoch": 0.07510799270893156,
+      "grad_norm": 0.0958174541592598,
+      "learning_rate": 2.7816215674985608e-05,
+      "loss": 10.3601,
+      "step": 15040
+    },
+    {
+      "epoch": 0.07515793153387101,
+      "grad_norm": 0.09285058826208115,
+      "learning_rate": 2.7814713760044058e-05,
+      "loss": 10.3572,
+      "step": 15050
+    },
+    {
+      "epoch": 0.07520787035881046,
+      "grad_norm": 0.09183011949062347,
+      "learning_rate": 2.7813211845102505e-05,
+      "loss": 10.3567,
+      "step": 15060
+    },
+    {
+      "epoch": 0.07525780918374991,
+      "grad_norm": 0.09525202959775925,
+      "learning_rate": 2.7811709930160955e-05,
+      "loss": 10.3597,
+      "step": 15070
+    },
+    {
+      "epoch": 0.07530774800868936,
+      "grad_norm": 0.09245378524065018,
+      "learning_rate": 2.7810208015219405e-05,
+      "loss": 10.3559,
+      "step": 15080
+    },
+    {
+      "epoch": 0.0753576868336288,
+      "grad_norm": 0.09502734988927841,
+      "learning_rate": 2.7808706100277856e-05,
+      "loss": 10.3569,
+      "step": 15090
+    },
+    {
+      "epoch": 0.07540762565856825,
+      "grad_norm": 0.09782931953668594,
+      "learning_rate": 2.7807204185336306e-05,
+      "loss": 10.3561,
+      "step": 15100
+    },
+    {
+      "epoch": 0.0754575644835077,
+      "grad_norm": 0.09502676874399185,
+      "learning_rate": 2.7805702270394752e-05,
+      "loss": 10.3553,
+      "step": 15110
+    },
+    {
+      "epoch": 0.07550750330844715,
+      "grad_norm": 0.09578611701726913,
+      "learning_rate": 2.7804200355453203e-05,
+      "loss": 10.3571,
+      "step": 15120
+    },
+    {
+      "epoch": 0.0755574421333866,
+      "grad_norm": 0.0919770896434784,
+      "learning_rate": 2.7802698440511653e-05,
+      "loss": 10.3541,
+      "step": 15130
+    },
+    {
+      "epoch": 0.07560738095832605,
+      "grad_norm": 0.09363967925310135,
+      "learning_rate": 2.7801196525570103e-05,
+      "loss": 10.3542,
+      "step": 15140
+    },
+    {
+      "epoch": 0.0756573197832655,
+      "grad_norm": 0.09395650029182434,
+      "learning_rate": 2.7799694610628553e-05,
+      "loss": 10.3552,
+      "step": 15150
+    },
+    {
+      "epoch": 0.07570725860820494,
+      "grad_norm": 0.09580358117818832,
+      "learning_rate": 2.7798192695687e-05,
+      "loss": 10.3551,
+      "step": 15160
+    },
+    {
+      "epoch": 0.0757571974331444,
+      "grad_norm": 0.09569013118743896,
+      "learning_rate": 2.779669078074545e-05,
+      "loss": 10.3505,
+      "step": 15170
+    },
+    {
+      "epoch": 0.07580713625808384,
+      "grad_norm": 0.09753712266683578,
+      "learning_rate": 2.77951888658039e-05,
+      "loss": 10.3539,
+      "step": 15180
+    },
+    {
+      "epoch": 0.07585707508302329,
+      "grad_norm": 0.09192202240228653,
+      "learning_rate": 2.779368695086235e-05,
+      "loss": 10.354,
+      "step": 15190
+    },
+    {
+      "epoch": 0.07590701390796274,
+      "grad_norm": 0.09564004093408585,
+      "learning_rate": 2.77921850359208e-05,
+      "loss": 10.3533,
+      "step": 15200
+    },
+    {
+      "epoch": 0.07595695273290219,
+      "grad_norm": 0.08988484740257263,
+      "learning_rate": 2.7790683120979248e-05,
+      "loss": 10.3559,
+      "step": 15210
+    },
+    {
+      "epoch": 0.07600689155784164,
+      "grad_norm": 0.09135802835226059,
+      "learning_rate": 2.7789181206037698e-05,
+      "loss": 10.355,
+      "step": 15220
+    },
+    {
+      "epoch": 0.07605683038278109,
+      "grad_norm": 0.10438718646764755,
+      "learning_rate": 2.7787679291096148e-05,
+      "loss": 10.3533,
+      "step": 15230
+    },
+    {
+      "epoch": 0.07610676920772054,
+      "grad_norm": 0.0949469655752182,
+      "learning_rate": 2.7786177376154598e-05,
+      "loss": 10.3526,
+      "step": 15240
+    },
+    {
+      "epoch": 0.07615670803265999,
+      "grad_norm": 0.09750612825155258,
+      "learning_rate": 2.7784675461213048e-05,
+      "loss": 10.3543,
+      "step": 15250
+    },
+    {
+      "epoch": 0.07620664685759944,
+      "grad_norm": 0.09662304818630219,
+      "learning_rate": 2.7783173546271495e-05,
+      "loss": 10.3492,
+      "step": 15260
+    },
+    {
+      "epoch": 0.07625658568253889,
+      "grad_norm": 0.09173522889614105,
+      "learning_rate": 2.7781671631329945e-05,
+      "loss": 10.3505,
+      "step": 15270
+    },
+    {
+      "epoch": 0.07630652450747834,
+      "grad_norm": 0.09761576354503632,
+      "learning_rate": 2.7780169716388395e-05,
+      "loss": 10.35,
+      "step": 15280
+    },
+    {
+      "epoch": 0.07635646333241779,
+      "grad_norm": 0.091497503221035,
+      "learning_rate": 2.7778667801446846e-05,
+      "loss": 10.353,
+      "step": 15290
+    },
+    {
+      "epoch": 0.07640640215735724,
+      "grad_norm": 0.0867045447230339,
+      "learning_rate": 2.7777165886505296e-05,
+      "loss": 10.3486,
+      "step": 15300
+    },
+    {
+      "epoch": 0.07645634098229669,
+      "grad_norm": 0.09160704165697098,
+      "learning_rate": 2.7775663971563743e-05,
+      "loss": 10.3531,
+      "step": 15310
+    },
+    {
+      "epoch": 0.07650627980723614,
+      "grad_norm": 0.09530753642320633,
+      "learning_rate": 2.7774162056622193e-05,
+      "loss": 10.3502,
+      "step": 15320
+    },
+    {
+      "epoch": 0.07655621863217558,
+      "grad_norm": 0.09408596903085709,
+      "learning_rate": 2.7772660141680643e-05,
+      "loss": 10.3514,
+      "step": 15330
+    },
+    {
+      "epoch": 0.07660615745711503,
+      "grad_norm": 0.10033749043941498,
+      "learning_rate": 2.7771158226739093e-05,
+      "loss": 10.3467,
+      "step": 15340
+    },
+    {
+      "epoch": 0.07665609628205448,
+      "grad_norm": 0.08937489241361618,
+      "learning_rate": 2.7769656311797543e-05,
+      "loss": 10.3505,
+      "step": 15350
+    },
+    {
+      "epoch": 0.07670603510699393,
+      "grad_norm": 0.09332039952278137,
+      "learning_rate": 2.7768154396855993e-05,
+      "loss": 10.349,
+      "step": 15360
+    },
+    {
+      "epoch": 0.07675597393193338,
+      "grad_norm": 0.10080984979867935,
+      "learning_rate": 2.776665248191444e-05,
+      "loss": 10.3472,
+      "step": 15370
+    },
+    {
+      "epoch": 0.07680591275687283,
+      "grad_norm": 0.09363698959350586,
+      "learning_rate": 2.776515056697289e-05,
+      "loss": 10.3491,
+      "step": 15380
+    },
+    {
+      "epoch": 0.07685585158181228,
+      "grad_norm": 0.09421728551387787,
+      "learning_rate": 2.776364865203134e-05,
+      "loss": 10.3457,
+      "step": 15390
+    },
+    {
+      "epoch": 0.07690579040675173,
+      "grad_norm": 0.09391890466213226,
+      "learning_rate": 2.776214673708979e-05,
+      "loss": 10.3481,
+      "step": 15400
+    },
+    {
+      "epoch": 0.07695572923169118,
+      "grad_norm": 0.09497596323490143,
+      "learning_rate": 2.776064482214824e-05,
+      "loss": 10.3486,
+      "step": 15410
+    },
+    {
+      "epoch": 0.07700566805663063,
+      "grad_norm": 0.09952494502067566,
+      "learning_rate": 2.7759142907206688e-05,
+      "loss": 10.3471,
+      "step": 15420
+    },
+    {
+      "epoch": 0.07705560688157008,
+      "grad_norm": 0.09322735667228699,
+      "learning_rate": 2.7757640992265138e-05,
+      "loss": 10.346,
+      "step": 15430
+    },
+    {
+      "epoch": 0.07710554570650953,
+      "grad_norm": 0.09606785327196121,
+      "learning_rate": 2.7756139077323588e-05,
+      "loss": 10.3519,
+      "step": 15440
+    },
+    {
+      "epoch": 0.07715548453144898,
+      "grad_norm": 0.09380047023296356,
+      "learning_rate": 2.7754637162382038e-05,
+      "loss": 10.3466,
+      "step": 15450
+    },
+    {
+      "epoch": 0.07720542335638843,
+      "grad_norm": 0.09241063892841339,
+      "learning_rate": 2.775313524744049e-05,
+      "loss": 10.3461,
+      "step": 15460
+    },
+    {
+      "epoch": 0.07725536218132788,
+      "grad_norm": 0.09774025529623032,
+      "learning_rate": 2.7751633332498935e-05,
+      "loss": 10.3434,
+      "step": 15470
+    },
+    {
+      "epoch": 0.07730530100626733,
+      "grad_norm": 0.09376423805952072,
+      "learning_rate": 2.7750131417557385e-05,
+      "loss": 10.3487,
+      "step": 15480
+    },
+    {
+      "epoch": 0.07735523983120678,
+      "grad_norm": 0.08905985951423645,
+      "learning_rate": 2.7748629502615836e-05,
+      "loss": 10.3459,
+      "step": 15490
+    },
+    {
+      "epoch": 0.07740517865614623,
+      "grad_norm": 0.09236487746238708,
+      "learning_rate": 2.7747127587674286e-05,
+      "loss": 10.3436,
+      "step": 15500
+    },
+    {
+      "epoch": 0.07745511748108568,
+      "grad_norm": 0.0898410752415657,
+      "learning_rate": 2.7745625672732736e-05,
+      "loss": 10.3419,
+      "step": 15510
+    },
+    {
+      "epoch": 0.07750505630602512,
+      "grad_norm": 0.09169359505176544,
+      "learning_rate": 2.7744123757791183e-05,
+      "loss": 10.3459,
+      "step": 15520
+    },
+    {
+      "epoch": 0.07755499513096457,
+      "grad_norm": 0.09277211874723434,
+      "learning_rate": 2.7742621842849633e-05,
+      "loss": 10.3462,
+      "step": 15530
+    },
+    {
+      "epoch": 0.07760493395590402,
+      "grad_norm": 0.09616000950336456,
+      "learning_rate": 2.7741119927908083e-05,
+      "loss": 10.3427,
+      "step": 15540
+    },
+    {
+      "epoch": 0.07765487278084347,
+      "grad_norm": 0.09537960588932037,
+      "learning_rate": 2.7739618012966533e-05,
+      "loss": 10.3464,
+      "step": 15550
+    },
+    {
+      "epoch": 0.07770481160578292,
+      "grad_norm": 0.09606116265058517,
+      "learning_rate": 2.7738116098024983e-05,
+      "loss": 10.348,
+      "step": 15560
+    },
+    {
+      "epoch": 0.07775475043072237,
+      "grad_norm": 0.10531776398420334,
+      "learning_rate": 2.773661418308343e-05,
+      "loss": 10.344,
+      "step": 15570
+    },
+    {
+      "epoch": 0.07780468925566181,
+      "grad_norm": 0.09302783012390137,
+      "learning_rate": 2.773511226814188e-05,
+      "loss": 10.3427,
+      "step": 15580
+    },
+    {
+      "epoch": 0.07785462808060126,
+      "grad_norm": 0.09965275973081589,
+      "learning_rate": 2.773361035320033e-05,
+      "loss": 10.3477,
+      "step": 15590
+    },
+    {
+      "epoch": 0.0779045669055407,
+      "grad_norm": 0.0914713442325592,
+      "learning_rate": 2.773210843825878e-05,
+      "loss": 10.343,
+      "step": 15600
+    },
+    {
+      "epoch": 0.07795450573048016,
+      "grad_norm": 0.09361792355775833,
+      "learning_rate": 2.773060652331723e-05,
+      "loss": 10.3476,
+      "step": 15610
+    },
+    {
+      "epoch": 0.0780044445554196,
+      "grad_norm": 0.09312635660171509,
+      "learning_rate": 2.7729104608375678e-05,
+      "loss": 10.3434,
+      "step": 15620
+    },
+    {
+      "epoch": 0.07805438338035905,
+      "grad_norm": 0.09331481158733368,
+      "learning_rate": 2.7727602693434128e-05,
+      "loss": 10.3433,
+      "step": 15630
+    },
+    {
+      "epoch": 0.0781043222052985,
+      "grad_norm": 0.0952400341629982,
+      "learning_rate": 2.7726100778492578e-05,
+      "loss": 10.343,
+      "step": 15640
+    },
+    {
+      "epoch": 0.07815426103023795,
+      "grad_norm": 0.09251481294631958,
+      "learning_rate": 2.7724598863551028e-05,
+      "loss": 10.3406,
+      "step": 15650
+    },
+    {
+      "epoch": 0.0782041998551774,
+      "grad_norm": 0.09364163130521774,
+      "learning_rate": 2.772309694860948e-05,
+      "loss": 10.3429,
+      "step": 15660
+    },
+    {
+      "epoch": 0.07825413868011685,
+      "grad_norm": 0.09640027582645416,
+      "learning_rate": 2.772159503366793e-05,
+      "loss": 10.3399,
+      "step": 15670
+    },
+    {
+      "epoch": 0.0783040775050563,
+      "grad_norm": 0.09361618012189865,
+      "learning_rate": 2.772009311872638e-05,
+      "loss": 10.3391,
+      "step": 15680
+    },
+    {
+      "epoch": 0.07835401632999575,
+      "grad_norm": 0.09318644553422928,
+      "learning_rate": 2.7718591203784826e-05,
+      "loss": 10.3398,
+      "step": 15690
+    },
+    {
+      "epoch": 0.0784039551549352,
+      "grad_norm": 0.09669560939073563,
+      "learning_rate": 2.7717089288843276e-05,
+      "loss": 10.3404,
+      "step": 15700
+    },
+    {
+      "epoch": 0.07845389397987465,
+      "grad_norm": 0.09504924714565277,
+      "learning_rate": 2.7715587373901726e-05,
+      "loss": 10.3368,
+      "step": 15710
+    },
+    {
+      "epoch": 0.0785038328048141,
+      "grad_norm": 0.09268130362033844,
+      "learning_rate": 2.7714085458960176e-05,
+      "loss": 10.3395,
+      "step": 15720
+    },
+    {
+      "epoch": 0.07855377162975355,
+      "grad_norm": 0.09593498706817627,
+      "learning_rate": 2.7712583544018626e-05,
+      "loss": 10.3378,
+      "step": 15730
+    },
+    {
+      "epoch": 0.078603710454693,
+      "grad_norm": 0.09858926385641098,
+      "learning_rate": 2.7711081629077073e-05,
+      "loss": 10.3395,
+      "step": 15740
+    },
+    {
+      "epoch": 0.07865364927963245,
+      "grad_norm": 0.09996471554040909,
+      "learning_rate": 2.7709579714135523e-05,
+      "loss": 10.3401,
+      "step": 15750
+    },
+    {
+      "epoch": 0.0787035881045719,
+      "grad_norm": 0.09353151172399521,
+      "learning_rate": 2.7708077799193973e-05,
+      "loss": 10.3392,
+      "step": 15760
+    },
+    {
+      "epoch": 0.07875352692951135,
+      "grad_norm": 0.0970112755894661,
+      "learning_rate": 2.7706575884252424e-05,
+      "loss": 10.337,
+      "step": 15770
+    },
+    {
+      "epoch": 0.0788034657544508,
+      "grad_norm": 0.09143032133579254,
+      "learning_rate": 2.7705073969310874e-05,
+      "loss": 10.3367,
+      "step": 15780
+    },
+    {
+      "epoch": 0.07885340457939025,
+      "grad_norm": 0.09837393462657928,
+      "learning_rate": 2.770357205436932e-05,
+      "loss": 10.3368,
+      "step": 15790
+    },
+    {
+      "epoch": 0.0789033434043297,
+      "grad_norm": 0.0949447974562645,
+      "learning_rate": 2.770207013942777e-05,
+      "loss": 10.3381,
+      "step": 15800
+    },
+    {
+      "epoch": 0.07895328222926914,
+      "grad_norm": 0.09616595506668091,
+      "learning_rate": 2.770056822448622e-05,
+      "loss": 10.3356,
+      "step": 15810
+    },
+    {
+      "epoch": 0.0790032210542086,
+      "grad_norm": 0.0987720862030983,
+      "learning_rate": 2.769906630954467e-05,
+      "loss": 10.3361,
+      "step": 15820
+    },
+    {
+      "epoch": 0.07905315987914804,
+      "grad_norm": 0.09198544174432755,
+      "learning_rate": 2.769756439460312e-05,
+      "loss": 10.3322,
+      "step": 15830
+    },
+    {
+      "epoch": 0.0791030987040875,
+      "grad_norm": 0.09431766718626022,
+      "learning_rate": 2.7696062479661568e-05,
+      "loss": 10.3391,
+      "step": 15840
+    },
+    {
+      "epoch": 0.07915303752902694,
+      "grad_norm": 0.08981557935476303,
+      "learning_rate": 2.7694560564720018e-05,
+      "loss": 10.3416,
+      "step": 15850
+    },
+    {
+      "epoch": 0.07920297635396639,
+      "grad_norm": 0.09577486664056778,
+      "learning_rate": 2.769305864977847e-05,
+      "loss": 10.3368,
+      "step": 15860
+    },
+    {
+      "epoch": 0.07925291517890584,
+      "grad_norm": 0.09910629689693451,
+      "learning_rate": 2.769155673483692e-05,
+      "loss": 10.3372,
+      "step": 15870
+    },
+    {
+      "epoch": 0.07930285400384529,
+      "grad_norm": 0.0959116742014885,
+      "learning_rate": 2.769005481989537e-05,
+      "loss": 10.332,
+      "step": 15880
+    },
+    {
+      "epoch": 0.07935279282878474,
+      "grad_norm": 0.09615825861692429,
+      "learning_rate": 2.7688552904953816e-05,
+      "loss": 10.3353,
+      "step": 15890
+    },
+    {
+      "epoch": 0.07940273165372419,
+      "grad_norm": 0.09139242023229599,
+      "learning_rate": 2.7687050990012266e-05,
+      "loss": 10.3368,
+      "step": 15900
+    },
+    {
+      "epoch": 0.07945267047866364,
+      "grad_norm": 0.09158849716186523,
+      "learning_rate": 2.7685549075070716e-05,
+      "loss": 10.3365,
+      "step": 15910
+    },
+    {
+      "epoch": 0.07950260930360309,
+      "grad_norm": 0.09999959915876389,
+      "learning_rate": 2.7684047160129166e-05,
+      "loss": 10.3314,
+      "step": 15920
+    },
+    {
+      "epoch": 0.07955254812854254,
+      "grad_norm": 0.09477412700653076,
+      "learning_rate": 2.7682545245187616e-05,
+      "loss": 10.3342,
+      "step": 15930
+    },
+    {
+      "epoch": 0.07960248695348199,
+      "grad_norm": 0.09604287892580032,
+      "learning_rate": 2.7681043330246063e-05,
+      "loss": 10.3322,
+      "step": 15940
+    },
+    {
+      "epoch": 0.07965242577842144,
+      "grad_norm": 0.09351278841495514,
+      "learning_rate": 2.7679541415304513e-05,
+      "loss": 10.3344,
+      "step": 15950
+    },
+    {
+      "epoch": 0.07970236460336089,
+      "grad_norm": 0.09625541418790817,
+      "learning_rate": 2.7678039500362963e-05,
+      "loss": 10.3328,
+      "step": 15960
+    },
+    {
+      "epoch": 0.07975230342830034,
+      "grad_norm": 0.09256501495838165,
+      "learning_rate": 2.7676537585421414e-05,
+      "loss": 10.3316,
+      "step": 15970
+    },
+    {
+      "epoch": 0.07980224225323979,
+      "grad_norm": 0.09308816492557526,
+      "learning_rate": 2.7675035670479864e-05,
+      "loss": 10.3311,
+      "step": 15980
+    },
+    {
+      "epoch": 0.07985218107817924,
+      "grad_norm": 0.09923449158668518,
+      "learning_rate": 2.767353375553831e-05,
+      "loss": 10.3336,
+      "step": 15990
+    },
+    {
+      "epoch": 0.07990211990311868,
+      "grad_norm": 0.09855549037456512,
+      "learning_rate": 2.7672031840596764e-05,
+      "loss": 10.3271,
+      "step": 16000
+    },
+    {
+      "epoch": 0.07995205872805813,
+      "grad_norm": 0.09425193816423416,
+      "learning_rate": 2.767052992565521e-05,
+      "loss": 10.3321,
+      "step": 16010
+    },
+    {
+      "epoch": 0.08000199755299758,
+      "grad_norm": 0.09234164655208588,
+      "learning_rate": 2.766902801071366e-05,
+      "loss": 10.334,
+      "step": 16020
+    },
+    {
+      "epoch": 0.08005193637793703,
+      "grad_norm": 0.09213145077228546,
+      "learning_rate": 2.766752609577211e-05,
+      "loss": 10.3343,
+      "step": 16030
+    },
+    {
+      "epoch": 0.08010187520287648,
+      "grad_norm": 0.09849076718091965,
+      "learning_rate": 2.7666024180830558e-05,
+      "loss": 10.3308,
+      "step": 16040
+    },
+    {
+      "epoch": 0.08015181402781593,
+      "grad_norm": 0.09698436409235,
+      "learning_rate": 2.766452226588901e-05,
+      "loss": 10.3301,
+      "step": 16050
+    },
+    {
+      "epoch": 0.08020175285275538,
+      "grad_norm": 0.10218020528554916,
+      "learning_rate": 2.766302035094746e-05,
+      "loss": 10.3282,
+      "step": 16060
+    },
+    {
+      "epoch": 0.08025169167769482,
+      "grad_norm": 0.0932408943772316,
+      "learning_rate": 2.766151843600591e-05,
+      "loss": 10.3309,
+      "step": 16070
+    },
+    {
+      "epoch": 0.08030163050263427,
+      "grad_norm": 0.09290464967489243,
+      "learning_rate": 2.766001652106436e-05,
+      "loss": 10.3298,
+      "step": 16080
+    },
+    {
+      "epoch": 0.08035156932757372,
+      "grad_norm": 0.09502555429935455,
+      "learning_rate": 2.7658514606122806e-05,
+      "loss": 10.3309,
+      "step": 16090
+    },
+    {
+      "epoch": 0.08040150815251317,
+      "grad_norm": 0.09941313415765762,
+      "learning_rate": 2.765701269118126e-05,
+      "loss": 10.3306,
+      "step": 16100
+    },
+    {
+      "epoch": 0.08045144697745261,
+      "grad_norm": 0.09274566918611526,
+      "learning_rate": 2.7655510776239706e-05,
+      "loss": 10.3321,
+      "step": 16110
+    },
+    {
+      "epoch": 0.08050138580239206,
+      "grad_norm": 0.094278983771801,
+      "learning_rate": 2.7654008861298156e-05,
+      "loss": 10.3285,
+      "step": 16120
+    },
+    {
+      "epoch": 0.08055132462733151,
+      "grad_norm": 0.09539202600717545,
+      "learning_rate": 2.7652506946356606e-05,
+      "loss": 10.3332,
+      "step": 16130
+    },
+    {
+      "epoch": 0.08060126345227096,
+      "grad_norm": 0.09607014805078506,
+      "learning_rate": 2.7651005031415053e-05,
+      "loss": 10.3289,
+      "step": 16140
+    },
+    {
+      "epoch": 0.08065120227721041,
+      "grad_norm": 0.09501274675130844,
+      "learning_rate": 2.7649503116473507e-05,
+      "loss": 10.3272,
+      "step": 16150
+    },
+    {
+      "epoch": 0.08070114110214986,
+      "grad_norm": 0.10182977467775345,
+      "learning_rate": 2.7648001201531953e-05,
+      "loss": 10.3256,
+      "step": 16160
+    },
+    {
+      "epoch": 0.08075107992708931,
+      "grad_norm": 0.10245219618082047,
+      "learning_rate": 2.7646499286590404e-05,
+      "loss": 10.3296,
+      "step": 16170
+    },
+    {
+      "epoch": 0.08080101875202876,
+      "grad_norm": 0.09179327636957169,
+      "learning_rate": 2.7644997371648854e-05,
+      "loss": 10.3269,
+      "step": 16180
+    },
+    {
+      "epoch": 0.08085095757696821,
+      "grad_norm": 0.08983848243951797,
+      "learning_rate": 2.76434954567073e-05,
+      "loss": 10.3286,
+      "step": 16190
+    },
+    {
+      "epoch": 0.08090089640190766,
+      "grad_norm": 0.09037093073129654,
+      "learning_rate": 2.7641993541765754e-05,
+      "loss": 10.3275,
+      "step": 16200
+    },
+    {
+      "epoch": 0.08095083522684711,
+      "grad_norm": 0.09235038608312607,
+      "learning_rate": 2.76404916268242e-05,
+      "loss": 10.3267,
+      "step": 16210
+    },
+    {
+      "epoch": 0.08100077405178656,
+      "grad_norm": 0.09570086747407913,
+      "learning_rate": 2.763898971188265e-05,
+      "loss": 10.3259,
+      "step": 16220
+    },
+    {
+      "epoch": 0.08105071287672601,
+      "grad_norm": 0.10826196521520615,
+      "learning_rate": 2.76374877969411e-05,
+      "loss": 10.3268,
+      "step": 16230
+    },
+    {
+      "epoch": 0.08110065170166546,
+      "grad_norm": 0.09768252819776535,
+      "learning_rate": 2.7635985881999548e-05,
+      "loss": 10.3236,
+      "step": 16240
+    },
+    {
+      "epoch": 0.08115059052660491,
+      "grad_norm": 0.09747245907783508,
+      "learning_rate": 2.7634483967058e-05,
+      "loss": 10.3286,
+      "step": 16250
+    },
+    {
+      "epoch": 0.08120052935154436,
+      "grad_norm": 0.0909043624997139,
+      "learning_rate": 2.763298205211645e-05,
+      "loss": 10.3277,
+      "step": 16260
+    },
+    {
+      "epoch": 0.0812504681764838,
+      "grad_norm": 0.09629534184932709,
+      "learning_rate": 2.76314801371749e-05,
+      "loss": 10.3245,
+      "step": 16270
+    },
+    {
+      "epoch": 0.08130040700142326,
+      "grad_norm": 0.08753444999456406,
+      "learning_rate": 2.762997822223335e-05,
+      "loss": 10.3248,
+      "step": 16280
+    },
+    {
+      "epoch": 0.0813503458263627,
+      "grad_norm": 0.09728357940912247,
+      "learning_rate": 2.7628476307291796e-05,
+      "loss": 10.3241,
+      "step": 16290
+    },
+    {
+      "epoch": 0.08140028465130215,
+      "grad_norm": 0.09314234554767609,
+      "learning_rate": 2.762697439235025e-05,
+      "loss": 10.3236,
+      "step": 16300
+    },
+    {
+      "epoch": 0.0814502234762416,
+      "grad_norm": 0.10189732909202576,
+      "learning_rate": 2.7625472477408696e-05,
+      "loss": 10.323,
+      "step": 16310
+    },
+    {
+      "epoch": 0.08150016230118105,
+      "grad_norm": 0.09725593775510788,
+      "learning_rate": 2.762397056246715e-05,
+      "loss": 10.3217,
+      "step": 16320
+    },
+    {
+      "epoch": 0.0815501011261205,
+      "grad_norm": 0.09180109947919846,
+      "learning_rate": 2.7622468647525596e-05,
+      "loss": 10.3236,
+      "step": 16330
+    },
+    {
+      "epoch": 0.08160003995105995,
+      "grad_norm": 0.09535349905490875,
+      "learning_rate": 2.7620966732584043e-05,
+      "loss": 10.3231,
+      "step": 16340
+    },
+    {
+      "epoch": 0.0816499787759994,
+      "grad_norm": 0.09873170405626297,
+      "learning_rate": 2.7619464817642497e-05,
+      "loss": 10.3229,
+      "step": 16350
+    },
+    {
+      "epoch": 0.08169991760093885,
+      "grad_norm": 0.09569983929395676,
+      "learning_rate": 2.7617962902700943e-05,
+      "loss": 10.3197,
+      "step": 16360
+    },
+    {
+      "epoch": 0.0817498564258783,
+      "grad_norm": 0.0878266841173172,
+      "learning_rate": 2.7616460987759397e-05,
+      "loss": 10.3213,
+      "step": 16370
+    },
+    {
+      "epoch": 0.08179979525081775,
+      "grad_norm": 0.09356684237718582,
+      "learning_rate": 2.7614959072817844e-05,
+      "loss": 10.3215,
+      "step": 16380
+    },
+    {
+      "epoch": 0.0818497340757572,
+      "grad_norm": 0.09417729079723358,
+      "learning_rate": 2.761345715787629e-05,
+      "loss": 10.3196,
+      "step": 16390
+    },
+    {
+      "epoch": 0.08189967290069665,
+      "grad_norm": 0.09582322090864182,
+      "learning_rate": 2.7611955242934744e-05,
+      "loss": 10.3188,
+      "step": 16400
+    },
+    {
+      "epoch": 0.0819496117256361,
+      "grad_norm": 0.09269662201404572,
+      "learning_rate": 2.761045332799319e-05,
+      "loss": 10.3197,
+      "step": 16410
+    },
+    {
+      "epoch": 0.08199955055057555,
+      "grad_norm": 0.09405113756656647,
+      "learning_rate": 2.7608951413051645e-05,
+      "loss": 10.3219,
+      "step": 16420
+    },
+    {
+      "epoch": 0.082049489375515,
+      "grad_norm": 0.09437499940395355,
+      "learning_rate": 2.760744949811009e-05,
+      "loss": 10.3209,
+      "step": 16430
+    },
+    {
+      "epoch": 0.08209942820045445,
+      "grad_norm": 0.09764499962329865,
+      "learning_rate": 2.7605947583168538e-05,
+      "loss": 10.3186,
+      "step": 16440
+    },
+    {
+      "epoch": 0.0821493670253939,
+      "grad_norm": 0.09104657918214798,
+      "learning_rate": 2.7604445668226992e-05,
+      "loss": 10.3198,
+      "step": 16450
+    },
+    {
+      "epoch": 0.08219930585033335,
+      "grad_norm": 0.09402047097682953,
+      "learning_rate": 2.760294375328544e-05,
+      "loss": 10.3226,
+      "step": 16460
+    },
+    {
+      "epoch": 0.0822492446752728,
+      "grad_norm": 0.09690111875534058,
+      "learning_rate": 2.7601441838343892e-05,
+      "loss": 10.3231,
+      "step": 16470
+    },
+    {
+      "epoch": 0.08229918350021224,
+      "grad_norm": 0.09647475183010101,
+      "learning_rate": 2.759993992340234e-05,
+      "loss": 10.318,
+      "step": 16480
+    },
+    {
+      "epoch": 0.0823491223251517,
+      "grad_norm": 0.09171656519174576,
+      "learning_rate": 2.7598438008460786e-05,
+      "loss": 10.3201,
+      "step": 16490
+    },
+    {
+      "epoch": 0.08239906115009114,
+      "grad_norm": 0.09819825738668442,
+      "learning_rate": 2.759693609351924e-05,
+      "loss": 10.3202,
+      "step": 16500
+    },
+    {
+      "epoch": 0.08244899997503059,
+      "grad_norm": 0.09369684010744095,
+      "learning_rate": 2.7595434178577686e-05,
+      "loss": 10.32,
+      "step": 16510
+    },
+    {
+      "epoch": 0.08249893879997004,
+      "grad_norm": 0.0983380377292633,
+      "learning_rate": 2.759393226363614e-05,
+      "loss": 10.3198,
+      "step": 16520
+    },
+    {
+      "epoch": 0.08254887762490949,
+      "grad_norm": 0.0950574278831482,
+      "learning_rate": 2.7592430348694586e-05,
+      "loss": 10.3151,
+      "step": 16530
+    },
+    {
+      "epoch": 0.08259881644984894,
+      "grad_norm": 0.09450704604387283,
+      "learning_rate": 2.7590928433753033e-05,
+      "loss": 10.3157,
+      "step": 16540
+    },
+    {
+      "epoch": 0.08264875527478839,
+      "grad_norm": 0.09773395955562592,
+      "learning_rate": 2.7589426518811487e-05,
+      "loss": 10.3158,
+      "step": 16550
+    },
+    {
+      "epoch": 0.08269869409972784,
+      "grad_norm": 0.09340257197618484,
+      "learning_rate": 2.7587924603869933e-05,
+      "loss": 10.318,
+      "step": 16560
+    },
+    {
+      "epoch": 0.08274863292466728,
+      "grad_norm": 0.09730247408151627,
+      "learning_rate": 2.7586422688928387e-05,
+      "loss": 10.3164,
+      "step": 16570
+    },
+    {
+      "epoch": 0.08279857174960673,
+      "grad_norm": 0.09569399803876877,
+      "learning_rate": 2.7584920773986834e-05,
+      "loss": 10.3131,
+      "step": 16580
+    },
+    {
+      "epoch": 0.08284851057454617,
+      "grad_norm": 0.09237007796764374,
+      "learning_rate": 2.758341885904528e-05,
+      "loss": 10.3144,
+      "step": 16590
+    },
+    {
+      "epoch": 0.08289844939948562,
+      "grad_norm": 0.09663589298725128,
+      "learning_rate": 2.7581916944103734e-05,
+      "loss": 10.3138,
+      "step": 16600
+    },
+    {
+      "epoch": 0.08294838822442507,
+      "grad_norm": 0.09514245390892029,
+      "learning_rate": 2.758041502916218e-05,
+      "loss": 10.3155,
+      "step": 16610
+    },
+    {
+      "epoch": 0.08299832704936452,
+      "grad_norm": 0.09728768467903137,
+      "learning_rate": 2.7578913114220635e-05,
+      "loss": 10.3166,
+      "step": 16620
+    },
+    {
+      "epoch": 0.08304826587430397,
+      "grad_norm": 0.10066792368888855,
+      "learning_rate": 2.757741119927908e-05,
+      "loss": 10.3149,
+      "step": 16630
+    },
+    {
+      "epoch": 0.08309820469924342,
+      "grad_norm": 0.09562237560749054,
+      "learning_rate": 2.757590928433753e-05,
+      "loss": 10.3161,
+      "step": 16640
+    },
+    {
+      "epoch": 0.08314814352418287,
+      "grad_norm": 0.09799202531576157,
+      "learning_rate": 2.7574407369395982e-05,
+      "loss": 10.3133,
+      "step": 16650
+    },
+    {
+      "epoch": 0.08319808234912232,
+      "grad_norm": 0.09569067507982254,
+      "learning_rate": 2.757290545445443e-05,
+      "loss": 10.314,
+      "step": 16660
+    },
+    {
+      "epoch": 0.08324802117406177,
+      "grad_norm": 0.09926281124353409,
+      "learning_rate": 2.7571403539512882e-05,
+      "loss": 10.3098,
+      "step": 16670
+    },
+    {
+      "epoch": 0.08329795999900122,
+      "grad_norm": 0.09658607095479965,
+      "learning_rate": 2.756990162457133e-05,
+      "loss": 10.3153,
+      "step": 16680
+    },
+    {
+      "epoch": 0.08334789882394067,
+      "grad_norm": 0.09640868008136749,
+      "learning_rate": 2.756839970962978e-05,
+      "loss": 10.3109,
+      "step": 16690
+    },
+    {
+      "epoch": 0.08339783764888012,
+      "grad_norm": 0.0960702896118164,
+      "learning_rate": 2.756689779468823e-05,
+      "loss": 10.3135,
+      "step": 16700
+    },
+    {
+      "epoch": 0.08344777647381957,
+      "grad_norm": 0.10049708187580109,
+      "learning_rate": 2.7565395879746676e-05,
+      "loss": 10.3111,
+      "step": 16710
+    },
+    {
+      "epoch": 0.08349771529875902,
+      "grad_norm": 0.1001792773604393,
+      "learning_rate": 2.756389396480513e-05,
+      "loss": 10.3115,
+      "step": 16720
+    },
+    {
+      "epoch": 0.08354765412369847,
+      "grad_norm": 0.09567177295684814,
+      "learning_rate": 2.7562392049863576e-05,
+      "loss": 10.3118,
+      "step": 16730
+    },
+    {
+      "epoch": 0.08359759294863792,
+      "grad_norm": 0.09801898151636124,
+      "learning_rate": 2.7560890134922027e-05,
+      "loss": 43.1471,
+      "step": 16740
+    },
+    {
+      "epoch": 0.08364753177357737,
+      "grad_norm": 0.09383946657180786,
+      "learning_rate": 2.7559388219980477e-05,
+      "loss": 14.7706,
+      "step": 16750
+    },
+    {
+      "epoch": 0.08369747059851682,
+      "grad_norm": 0.10004527121782303,
+      "learning_rate": 2.7557886305038924e-05,
+      "loss": 10.3106,
+      "step": 16760
+    },
+    {
+      "epoch": 0.08374740942345626,
+      "grad_norm": 0.10205840319395065,
+      "learning_rate": 2.7556384390097377e-05,
+      "loss": 10.7248,
+      "step": 16770
+    },
+    {
+      "epoch": 0.08379734824839571,
+      "grad_norm": 0.10261974483728409,
+      "learning_rate": 2.7554882475155824e-05,
+      "loss": 10.3106,
+      "step": 16780
+    },
+    {
+      "epoch": 0.08384728707333516,
+      "grad_norm": 0.09215449541807175,
+      "learning_rate": 2.7553380560214274e-05,
+      "loss": 10.3098,
+      "step": 16790
+    },
+    {
+      "epoch": 0.08389722589827461,
+      "grad_norm": 0.09583956003189087,
+      "learning_rate": 2.7551878645272724e-05,
+      "loss": 10.3098,
+      "step": 16800
+    },
+    {
+      "epoch": 0.08394716472321406,
+      "grad_norm": 0.09968181699514389,
+      "learning_rate": 2.755037673033117e-05,
+      "loss": 10.3098,
+      "step": 16810
+    },
+    {
+      "epoch": 0.08399710354815351,
+      "grad_norm": 0.09686349332332611,
+      "learning_rate": 2.7548874815389625e-05,
+      "loss": 10.3073,
+      "step": 16820
+    },
+    {
+      "epoch": 0.08404704237309296,
+      "grad_norm": 0.09120567888021469,
+      "learning_rate": 2.754737290044807e-05,
+      "loss": 10.3103,
+      "step": 16830
+    },
+    {
+      "epoch": 0.08409698119803241,
+      "grad_norm": 0.0937332808971405,
+      "learning_rate": 2.754587098550652e-05,
+      "loss": 10.3085,
+      "step": 16840
+    },
+    {
+      "epoch": 0.08414692002297186,
+      "grad_norm": 0.091478131711483,
+      "learning_rate": 2.7544369070564972e-05,
+      "loss": 10.3071,
+      "step": 16850
+    },
+    {
+      "epoch": 0.08419685884791131,
+      "grad_norm": 0.09270336478948593,
+      "learning_rate": 2.754286715562342e-05,
+      "loss": 10.3107,
+      "step": 16860
+    },
+    {
+      "epoch": 0.08424679767285076,
+      "grad_norm": 0.09209531545639038,
+      "learning_rate": 2.7541365240681872e-05,
+      "loss": 10.3128,
+      "step": 16870
+    },
+    {
+      "epoch": 0.08429673649779021,
+      "grad_norm": 0.09706257283687592,
+      "learning_rate": 2.753986332574032e-05,
+      "loss": 10.3076,
+      "step": 16880
+    },
+    {
+      "epoch": 0.08434667532272966,
+      "grad_norm": 0.09445540606975555,
+      "learning_rate": 2.753836141079877e-05,
+      "loss": 10.3094,
+      "step": 16890
+    },
+    {
+      "epoch": 0.08439661414766911,
+      "grad_norm": 0.09534412622451782,
+      "learning_rate": 2.753685949585722e-05,
+      "loss": 10.311,
+      "step": 16900
+    },
+    {
+      "epoch": 0.08444655297260856,
+      "grad_norm": 0.09859678149223328,
+      "learning_rate": 2.7535357580915666e-05,
+      "loss": 10.3073,
+      "step": 16910
+    },
+    {
+      "epoch": 0.084496491797548,
+      "grad_norm": 0.09540943056344986,
+      "learning_rate": 2.753385566597412e-05,
+      "loss": 10.3114,
+      "step": 16920
+    },
+    {
+      "epoch": 0.08454643062248746,
+      "grad_norm": 0.09608754515647888,
+      "learning_rate": 2.7532353751032566e-05,
+      "loss": 10.3052,
+      "step": 16930
+    },
+    {
+      "epoch": 0.0845963694474269,
+      "grad_norm": 0.09397275745868683,
+      "learning_rate": 2.7530851836091017e-05,
+      "loss": 10.3078,
+      "step": 16940
+    },
+    {
+      "epoch": 0.08464630827236636,
+      "grad_norm": 0.09713399410247803,
+      "learning_rate": 2.7529349921149467e-05,
+      "loss": 10.3083,
+      "step": 16950
+    },
+    {
+      "epoch": 0.0846962470973058,
+      "grad_norm": 0.09662938863039017,
+      "learning_rate": 2.7527848006207917e-05,
+      "loss": 10.304,
+      "step": 16960
+    },
+    {
+      "epoch": 0.08474618592224525,
+      "grad_norm": 0.09605155885219574,
+      "learning_rate": 2.7526346091266367e-05,
+      "loss": 10.3048,
+      "step": 16970
+    },
+    {
+      "epoch": 0.0847961247471847,
+      "grad_norm": 0.09501200169324875,
+      "learning_rate": 2.7524844176324814e-05,
+      "loss": 10.3018,
+      "step": 16980
+    },
+    {
+      "epoch": 0.08484606357212415,
+      "grad_norm": 0.0933113843202591,
+      "learning_rate": 2.7523342261383264e-05,
+      "loss": 10.3059,
+      "step": 16990
+    },
+    {
+      "epoch": 0.0848960023970636,
+      "grad_norm": 0.09310649335384369,
+      "learning_rate": 2.7521840346441714e-05,
+      "loss": 10.3055,
+      "step": 17000
+    },
+    {
+      "epoch": 0.08494594122200305,
+      "grad_norm": 0.09754558652639389,
+      "learning_rate": 2.7520338431500164e-05,
+      "loss": 10.3086,
+      "step": 17010
+    },
+    {
+      "epoch": 0.0849958800469425,
+      "grad_norm": 0.08995413780212402,
+      "learning_rate": 2.7518836516558615e-05,
+      "loss": 10.3065,
+      "step": 17020
+    },
+    {
+      "epoch": 0.08504581887188195,
+      "grad_norm": 0.09209508448839188,
+      "learning_rate": 2.751733460161706e-05,
+      "loss": 10.3058,
+      "step": 17030
+    },
+    {
+      "epoch": 0.0850957576968214,
+      "grad_norm": 0.0900413915514946,
+      "learning_rate": 2.751583268667551e-05,
+      "loss": 10.3033,
+      "step": 17040
+    },
+    {
+      "epoch": 0.08514569652176085,
+      "grad_norm": 0.10113733261823654,
+      "learning_rate": 2.7514330771733962e-05,
+      "loss": 10.3027,
+      "step": 17050
+    },
+    {
+      "epoch": 0.08519563534670029,
+      "grad_norm": 0.10684780031442642,
+      "learning_rate": 2.7512828856792412e-05,
+      "loss": 10.3006,
+      "step": 17060
+    },
+    {
+      "epoch": 0.08524557417163973,
+      "grad_norm": 0.09294813871383667,
+      "learning_rate": 2.7511326941850862e-05,
+      "loss": 10.3065,
+      "step": 17070
+    },
+    {
+      "epoch": 0.08529551299657918,
+      "grad_norm": 0.09303073585033417,
+      "learning_rate": 2.750982502690931e-05,
+      "loss": 10.3031,
+      "step": 17080
+    },
+    {
+      "epoch": 0.08534545182151863,
+      "grad_norm": 0.09494984894990921,
+      "learning_rate": 2.750832311196776e-05,
+      "loss": 10.3067,
+      "step": 17090
+    },
+    {
+      "epoch": 0.08539539064645808,
+      "grad_norm": 0.09534880518913269,
+      "learning_rate": 2.750682119702621e-05,
+      "loss": 10.302,
+      "step": 17100
+    },
+    {
+      "epoch": 0.08544532947139753,
+      "grad_norm": 0.09925422072410583,
+      "learning_rate": 2.750531928208466e-05,
+      "loss": 10.3026,
+      "step": 17110
+    },
+    {
+      "epoch": 0.08549526829633698,
+      "grad_norm": 0.09157177805900574,
+      "learning_rate": 2.750381736714311e-05,
+      "loss": 10.3023,
+      "step": 17120
+    },
+    {
+      "epoch": 0.08554520712127643,
+      "grad_norm": 0.09446504712104797,
+      "learning_rate": 2.7502315452201556e-05,
+      "loss": 10.3023,
+      "step": 17130
+    },
+    {
+      "epoch": 0.08559514594621588,
+      "grad_norm": 0.08893059939146042,
+      "learning_rate": 2.7500813537260007e-05,
+      "loss": 10.3017,
+      "step": 17140
+    },
+    {
+      "epoch": 0.08564508477115533,
+      "grad_norm": 0.09873572736978531,
+      "learning_rate": 2.7499311622318457e-05,
+      "loss": 10.3041,
+      "step": 17150
+    },
+    {
+      "epoch": 0.08569502359609478,
+      "grad_norm": 0.10007481276988983,
+      "learning_rate": 2.7497809707376907e-05,
+      "loss": 10.2991,
+      "step": 17160
+    },
+    {
+      "epoch": 0.08574496242103423,
+      "grad_norm": 0.0909745916724205,
+      "learning_rate": 2.7496307792435357e-05,
+      "loss": 12.4831,
+      "step": 17170
+    },
+    {
+      "epoch": 0.08579490124597368,
+      "grad_norm": 0.09608662128448486,
+      "learning_rate": 2.7494805877493804e-05,
+      "loss": 10.3047,
+      "step": 17180
+    },
+    {
+      "epoch": 0.08584484007091313,
+      "grad_norm": 0.0924767255783081,
+      "learning_rate": 2.7493303962552254e-05,
+      "loss": 10.2995,
+      "step": 17190
+    },
+    {
+      "epoch": 0.08589477889585258,
+      "grad_norm": 0.09744284301996231,
+      "learning_rate": 2.7491802047610704e-05,
+      "loss": 10.3003,
+      "step": 17200
+    },
+    {
+      "epoch": 0.08594471772079203,
+      "grad_norm": 0.09493273496627808,
+      "learning_rate": 2.7490300132669154e-05,
+      "loss": 10.3025,
+      "step": 17210
+    },
+    {
+      "epoch": 0.08599465654573148,
+      "grad_norm": 0.10000015050172806,
+      "learning_rate": 2.7488798217727605e-05,
+      "loss": 10.3011,
+      "step": 17220
+    },
+    {
+      "epoch": 0.08604459537067093,
+      "grad_norm": 0.09295312315225601,
+      "learning_rate": 2.748729630278605e-05,
+      "loss": 10.3004,
+      "step": 17230
+    },
+    {
+      "epoch": 0.08609453419561038,
+      "grad_norm": 0.10081910341978073,
+      "learning_rate": 2.74857943878445e-05,
+      "loss": 10.2935,
+      "step": 17240
+    },
+    {
+      "epoch": 0.08614447302054982,
+      "grad_norm": 0.09449261426925659,
+      "learning_rate": 2.7484292472902952e-05,
+      "loss": 10.2981,
+      "step": 17250
+    },
+    {
+      "epoch": 0.08619441184548927,
+      "grad_norm": 0.09793805330991745,
+      "learning_rate": 2.7482790557961402e-05,
+      "loss": 10.2947,
+      "step": 17260
+    },
+    {
+      "epoch": 0.08624435067042872,
+      "grad_norm": 0.10022538155317307,
+      "learning_rate": 2.7481288643019852e-05,
+      "loss": 10.2971,
+      "step": 17270
+    },
+    {
+      "epoch": 0.08629428949536817,
+      "grad_norm": 0.0923459604382515,
+      "learning_rate": 2.7479786728078302e-05,
+      "loss": 10.2982,
+      "step": 17280
+    },
+    {
+      "epoch": 0.08634422832030762,
+      "grad_norm": 0.09417003393173218,
+      "learning_rate": 2.747828481313675e-05,
+      "loss": 10.2989,
+      "step": 17290
+    },
+    {
+      "epoch": 0.08639416714524707,
+      "grad_norm": 0.09343262761831284,
+      "learning_rate": 2.74767828981952e-05,
+      "loss": 10.2968,
+      "step": 17300
+    },
+    {
+      "epoch": 0.08644410597018652,
+      "grad_norm": 0.09520164877176285,
+      "learning_rate": 2.747528098325365e-05,
+      "loss": 10.2959,
+      "step": 17310
+    },
+    {
+      "epoch": 0.08649404479512597,
+      "grad_norm": 0.09455534815788269,
+      "learning_rate": 2.74737790683121e-05,
+      "loss": 10.2957,
+      "step": 17320
+    },
+    {
+      "epoch": 0.08654398362006542,
+      "grad_norm": 0.09446435421705246,
+      "learning_rate": 2.747227715337055e-05,
+      "loss": 10.2937,
+      "step": 17330
+    },
+    {
+      "epoch": 0.08659392244500487,
+      "grad_norm": 0.09533445537090302,
+      "learning_rate": 2.7470775238428997e-05,
+      "loss": 10.2936,
+      "step": 17340
+    },
+    {
+      "epoch": 0.08664386126994432,
+      "grad_norm": 0.09248590469360352,
+      "learning_rate": 2.7469273323487447e-05,
+      "loss": 10.2965,
+      "step": 17350
+    },
+    {
+      "epoch": 0.08669380009488377,
+      "grad_norm": 0.0972449779510498,
+      "learning_rate": 2.7467771408545897e-05,
+      "loss": 10.2934,
+      "step": 17360
+    },
+    {
+      "epoch": 0.08674373891982322,
+      "grad_norm": 0.08786053210496902,
+      "learning_rate": 2.7466269493604347e-05,
+      "loss": 10.2956,
+      "step": 17370
+    },
+    {
+      "epoch": 0.08679367774476267,
+      "grad_norm": 0.10034439712762833,
+      "learning_rate": 2.7464767578662797e-05,
+      "loss": 10.2939,
+      "step": 17380
+    },
+    {
+      "epoch": 0.08684361656970212,
+      "grad_norm": 0.10132072865962982,
+      "learning_rate": 2.7463265663721244e-05,
+      "loss": 10.2934,
+      "step": 17390
+    },
+    {
+      "epoch": 0.08689355539464157,
+      "grad_norm": 0.093051977455616,
+      "learning_rate": 2.7461763748779694e-05,
+      "loss": 10.2921,
+      "step": 17400
+    },
+    {
+      "epoch": 0.08694349421958102,
+      "grad_norm": 0.09505248069763184,
+      "learning_rate": 2.7460261833838144e-05,
+      "loss": 10.2927,
+      "step": 17410
+    },
+    {
+      "epoch": 0.08699343304452047,
+      "grad_norm": 0.09846062958240509,
+      "learning_rate": 2.7458759918896595e-05,
+      "loss": 10.2926,
+      "step": 17420
+    },
+    {
+      "epoch": 0.08704337186945992,
+      "grad_norm": 0.10331861674785614,
+      "learning_rate": 2.7457258003955045e-05,
+      "loss": 10.2922,
+      "step": 17430
+    },
+    {
+      "epoch": 0.08709331069439936,
+      "grad_norm": 0.0902588963508606,
+      "learning_rate": 2.745575608901349e-05,
+      "loss": 10.2916,
+      "step": 17440
+    },
+    {
+      "epoch": 0.08714324951933881,
+      "grad_norm": 0.09760275483131409,
+      "learning_rate": 2.7454254174071942e-05,
+      "loss": 10.294,
+      "step": 17450
+    },
+    {
+      "epoch": 0.08719318834427826,
+      "grad_norm": 0.09371180832386017,
+      "learning_rate": 2.7452752259130392e-05,
+      "loss": 10.2919,
+      "step": 17460
+    },
+    {
+      "epoch": 0.08724312716921771,
+      "grad_norm": 0.09257213771343231,
+      "learning_rate": 2.7451250344188842e-05,
+      "loss": 10.2953,
+      "step": 17470
+    },
+    {
+      "epoch": 0.08729306599415716,
+      "grad_norm": 0.09264030307531357,
+      "learning_rate": 2.7449748429247292e-05,
+      "loss": 10.2946,
+      "step": 17480
+    },
+    {
+      "epoch": 0.08734300481909661,
+      "grad_norm": 0.09427089989185333,
+      "learning_rate": 2.744824651430574e-05,
+      "loss": 10.2918,
+      "step": 17490
+    },
+    {
+      "epoch": 0.08739294364403606,
+      "grad_norm": 0.09720293432474136,
+      "learning_rate": 2.744674459936419e-05,
+      "loss": 10.2937,
+      "step": 17500
+    },
+    {
+      "epoch": 0.08744288246897551,
+      "grad_norm": 0.09387167543172836,
+      "learning_rate": 2.744524268442264e-05,
+      "loss": 10.2891,
+      "step": 17510
+    },
+    {
+      "epoch": 0.08749282129391496,
+      "grad_norm": 0.09527067095041275,
+      "learning_rate": 2.744374076948109e-05,
+      "loss": 10.2929,
+      "step": 17520
+    },
+    {
+      "epoch": 0.08754276011885441,
+      "grad_norm": 0.09935829043388367,
+      "learning_rate": 2.744223885453954e-05,
+      "loss": 10.2936,
+      "step": 17530
+    },
+    {
+      "epoch": 0.08759269894379386,
+      "grad_norm": 0.0906045064330101,
+      "learning_rate": 2.7440736939597987e-05,
+      "loss": 10.2928,
+      "step": 17540
+    },
+    {
+      "epoch": 0.08764263776873331,
+      "grad_norm": 0.09333600848913193,
+      "learning_rate": 2.7439235024656437e-05,
+      "loss": 10.2918,
+      "step": 17550
+    },
+    {
+      "epoch": 0.08769257659367274,
+      "grad_norm": 0.09050074964761734,
+      "learning_rate": 2.7437733109714887e-05,
+      "loss": 10.2893,
+      "step": 17560
+    },
+    {
+      "epoch": 0.0877425154186122,
+      "grad_norm": 0.0935349315404892,
+      "learning_rate": 2.7436231194773337e-05,
+      "loss": 10.2924,
+      "step": 17570
+    },
+    {
+      "epoch": 0.08779245424355164,
+      "grad_norm": 0.09864573180675507,
+      "learning_rate": 2.7434729279831787e-05,
+      "loss": 10.2915,
+      "step": 17580
+    },
+    {
+      "epoch": 0.08784239306849109,
+      "grad_norm": 0.09615197032690048,
+      "learning_rate": 2.7433227364890234e-05,
+      "loss": 10.2898,
+      "step": 17590
+    },
+    {
+      "epoch": 0.08789233189343054,
+      "grad_norm": 0.09623583406209946,
+      "learning_rate": 2.7431725449948688e-05,
+      "loss": 10.2904,
+      "step": 17600
+    },
+    {
+      "epoch": 0.08794227071836999,
+      "grad_norm": 0.0986003652215004,
+      "learning_rate": 2.7430223535007134e-05,
+      "loss": 10.2876,
+      "step": 17610
+    },
+    {
+      "epoch": 0.08799220954330944,
+      "grad_norm": 0.09806670993566513,
+      "learning_rate": 2.7428721620065585e-05,
+      "loss": 10.2888,
+      "step": 17620
+    },
+    {
+      "epoch": 0.08804214836824889,
+      "grad_norm": 0.09443091601133347,
+      "learning_rate": 2.7427219705124035e-05,
+      "loss": 10.2883,
+      "step": 17630
+    },
+    {
+      "epoch": 0.08809208719318834,
+      "grad_norm": 0.0967443585395813,
+      "learning_rate": 2.742571779018248e-05,
+      "loss": 10.2869,
+      "step": 17640
+    },
+    {
+      "epoch": 0.08814202601812779,
+      "grad_norm": 0.09863567352294922,
+      "learning_rate": 2.7424215875240935e-05,
+      "loss": 10.2899,
+      "step": 17650
+    },
+    {
+      "epoch": 0.08819196484306724,
+      "grad_norm": 0.09256859123706818,
+      "learning_rate": 2.7422713960299382e-05,
+      "loss": 10.2843,
+      "step": 17660
+    },
+    {
+      "epoch": 0.08824190366800669,
+      "grad_norm": 0.09759136289358139,
+      "learning_rate": 2.7421212045357832e-05,
+      "loss": 10.2869,
+      "step": 17670
+    },
+    {
+      "epoch": 0.08829184249294614,
+      "grad_norm": 0.09880513697862625,
+      "learning_rate": 2.7419710130416282e-05,
+      "loss": 10.2891,
+      "step": 17680
+    },
+    {
+      "epoch": 0.08834178131788559,
+      "grad_norm": 0.09688255190849304,
+      "learning_rate": 2.741820821547473e-05,
+      "loss": 10.2857,
+      "step": 17690
+    },
+    {
+      "epoch": 0.08839172014282504,
+      "grad_norm": 0.0934658944606781,
+      "learning_rate": 2.7416706300533183e-05,
+      "loss": 10.2866,
+      "step": 17700
+    },
+    {
+      "epoch": 0.08844165896776449,
+      "grad_norm": 0.09990676492452621,
+      "learning_rate": 2.741520438559163e-05,
+      "loss": 10.2842,
+      "step": 17710
+    },
+    {
+      "epoch": 0.08849159779270394,
+      "grad_norm": 0.09611719101667404,
+      "learning_rate": 2.741370247065008e-05,
+      "loss": 10.2883,
+      "step": 17720
+    },
+    {
+      "epoch": 0.08854153661764338,
+      "grad_norm": 0.09569796919822693,
+      "learning_rate": 2.741220055570853e-05,
+      "loss": 10.2871,
+      "step": 17730
+    },
+    {
+      "epoch": 0.08859147544258283,
+      "grad_norm": 0.09596303105354309,
+      "learning_rate": 2.7410698640766977e-05,
+      "loss": 10.2823,
+      "step": 17740
+    },
+    {
+      "epoch": 0.08864141426752228,
+      "grad_norm": 0.0938568115234375,
+      "learning_rate": 2.740919672582543e-05,
+      "loss": 10.2833,
+      "step": 17750
+    },
+    {
+      "epoch": 0.08869135309246173,
+      "grad_norm": 0.09523828327655792,
+      "learning_rate": 2.7407694810883877e-05,
+      "loss": 10.2872,
+      "step": 17760
+    },
+    {
+      "epoch": 0.08874129191740118,
+      "grad_norm": 0.09327805787324905,
+      "learning_rate": 2.7406192895942327e-05,
+      "loss": 10.2872,
+      "step": 17770
+    },
+    {
+      "epoch": 0.08879123074234063,
+      "grad_norm": 0.09197764843702316,
+      "learning_rate": 2.7404690981000777e-05,
+      "loss": 10.2836,
+      "step": 17780
+    },
+    {
+      "epoch": 0.08884116956728008,
+      "grad_norm": 0.08873019367456436,
+      "learning_rate": 2.7403189066059224e-05,
+      "loss": 10.2816,
+      "step": 17790
+    },
+    {
+      "epoch": 0.08889110839221953,
+      "grad_norm": 0.0956963300704956,
+      "learning_rate": 2.7401687151117678e-05,
+      "loss": 10.2801,
+      "step": 17800
+    },
+    {
+      "epoch": 0.08894104721715898,
+      "grad_norm": 0.09462413936853409,
+      "learning_rate": 2.7400185236176124e-05,
+      "loss": 10.2806,
+      "step": 17810
+    },
+    {
+      "epoch": 0.08899098604209843,
+      "grad_norm": 0.09391899406909943,
+      "learning_rate": 2.7398683321234575e-05,
+      "loss": 10.2877,
+      "step": 17820
+    },
+    {
+      "epoch": 0.08904092486703788,
+      "grad_norm": 0.09864463657140732,
+      "learning_rate": 2.7397181406293025e-05,
+      "loss": 10.2851,
+      "step": 17830
+    },
+    {
+      "epoch": 0.08909086369197733,
+      "grad_norm": 0.09195532649755478,
+      "learning_rate": 2.739567949135147e-05,
+      "loss": 10.2851,
+      "step": 17840
+    },
+    {
+      "epoch": 0.08914080251691678,
+      "grad_norm": 0.09261298924684525,
+      "learning_rate": 2.7394177576409925e-05,
+      "loss": 10.2853,
+      "step": 17850
+    },
+    {
+      "epoch": 0.08919074134185623,
+      "grad_norm": 0.09226420521736145,
+      "learning_rate": 2.7392675661468372e-05,
+      "loss": 10.2833,
+      "step": 17860
+    },
+    {
+      "epoch": 0.08924068016679568,
+      "grad_norm": 0.09687785059213638,
+      "learning_rate": 2.7391173746526822e-05,
+      "loss": 10.2819,
+      "step": 17870
+    },
+    {
+      "epoch": 0.08929061899173513,
+      "grad_norm": 0.09745992720127106,
+      "learning_rate": 2.7389671831585272e-05,
+      "loss": 10.285,
+      "step": 17880
+    },
+    {
+      "epoch": 0.08934055781667458,
+      "grad_norm": 0.10478182882070541,
+      "learning_rate": 2.738816991664372e-05,
+      "loss": 10.2759,
+      "step": 17890
+    },
+    {
+      "epoch": 0.08939049664161403,
+      "grad_norm": 0.0948200598359108,
+      "learning_rate": 2.7386668001702173e-05,
+      "loss": 10.2803,
+      "step": 17900
+    },
+    {
+      "epoch": 0.08944043546655348,
+      "grad_norm": 0.09617318212985992,
+      "learning_rate": 2.738516608676062e-05,
+      "loss": 10.2789,
+      "step": 17910
+    },
+    {
+      "epoch": 0.08949037429149292,
+      "grad_norm": 0.09210513532161713,
+      "learning_rate": 2.7383664171819073e-05,
+      "loss": 10.2788,
+      "step": 17920
+    },
+    {
+      "epoch": 0.08954031311643237,
+      "grad_norm": 0.09981706738471985,
+      "learning_rate": 2.738216225687752e-05,
+      "loss": 10.2767,
+      "step": 17930
+    },
+    {
+      "epoch": 0.08959025194137182,
+      "grad_norm": 0.1067042201757431,
+      "learning_rate": 2.7380660341935967e-05,
+      "loss": 10.2798,
+      "step": 17940
+    },
+    {
+      "epoch": 0.08964019076631127,
+      "grad_norm": 0.0890965685248375,
+      "learning_rate": 2.737915842699442e-05,
+      "loss": 10.2807,
+      "step": 17950
+    },
+    {
+      "epoch": 0.08969012959125072,
+      "grad_norm": 0.09566470980644226,
+      "learning_rate": 2.7377656512052867e-05,
+      "loss": 10.2792,
+      "step": 17960
+    },
+    {
+      "epoch": 0.08974006841619017,
+      "grad_norm": 0.09969697147607803,
+      "learning_rate": 2.737615459711132e-05,
+      "loss": 10.2778,
+      "step": 17970
+    },
+    {
+      "epoch": 0.08979000724112962,
+      "grad_norm": 0.09558671712875366,
+      "learning_rate": 2.7374652682169767e-05,
+      "loss": 10.2777,
+      "step": 17980
+    },
+    {
+      "epoch": 0.08983994606606907,
+      "grad_norm": 0.09214569628238678,
+      "learning_rate": 2.7373150767228214e-05,
+      "loss": 10.2794,
+      "step": 17990
+    },
+    {
+      "epoch": 0.08988988489100852,
+      "grad_norm": 0.09558131545782089,
+      "learning_rate": 2.7371648852286668e-05,
+      "loss": 10.279,
+      "step": 18000
+    },
+    {
+      "epoch": 0.08993982371594797,
+      "grad_norm": 0.09858100861310959,
+      "learning_rate": 2.7370146937345114e-05,
+      "loss": 10.2783,
+      "step": 18010
+    },
+    {
+      "epoch": 0.08998976254088742,
+      "grad_norm": 0.09311861544847488,
+      "learning_rate": 2.7368645022403568e-05,
+      "loss": 10.28,
+      "step": 18020
+    },
+    {
+      "epoch": 0.09003970136582687,
+      "grad_norm": 0.09338545799255371,
+      "learning_rate": 2.7367143107462015e-05,
+      "loss": 10.2755,
+      "step": 18030
+    },
+    {
+      "epoch": 0.09008964019076632,
+      "grad_norm": 0.09420605003833771,
+      "learning_rate": 2.736564119252046e-05,
+      "loss": 10.2778,
+      "step": 18040
+    },
+    {
+      "epoch": 0.09013957901570575,
+      "grad_norm": 0.09855175018310547,
+      "learning_rate": 2.7364139277578915e-05,
+      "loss": 10.2763,
+      "step": 18050
+    },
+    {
+      "epoch": 0.0901895178406452,
+      "grad_norm": 0.09705475717782974,
+      "learning_rate": 2.7362637362637362e-05,
+      "loss": 10.2781,
+      "step": 18060
+    },
+    {
+      "epoch": 0.09023945666558465,
+      "grad_norm": 0.09603998064994812,
+      "learning_rate": 2.7361135447695816e-05,
+      "loss": 10.277,
+      "step": 18070
+    },
+    {
+      "epoch": 0.0902893954905241,
+      "grad_norm": 0.09418918937444687,
+      "learning_rate": 2.7359633532754262e-05,
+      "loss": 10.274,
+      "step": 18080
+    },
+    {
+      "epoch": 0.09033933431546355,
+      "grad_norm": 0.09442311525344849,
+      "learning_rate": 2.735813161781271e-05,
+      "loss": 10.2778,
+      "step": 18090
+    },
+    {
+      "epoch": 0.090389273140403,
+      "grad_norm": 0.0943770557641983,
+      "learning_rate": 2.7356629702871163e-05,
+      "loss": 10.273,
+      "step": 18100
+    },
+    {
+      "epoch": 0.09043921196534245,
+      "grad_norm": 0.09986016154289246,
+      "learning_rate": 2.735512778792961e-05,
+      "loss": 10.2782,
+      "step": 18110
+    },
+    {
+      "epoch": 0.0904891507902819,
+      "grad_norm": 0.09035969525575638,
+      "learning_rate": 2.7353625872988063e-05,
+      "loss": 10.2772,
+      "step": 18120
+    },
+    {
+      "epoch": 0.09053908961522135,
+      "grad_norm": 0.09936581552028656,
+      "learning_rate": 2.735212395804651e-05,
+      "loss": 10.2784,
+      "step": 18130
+    },
+    {
+      "epoch": 0.0905890284401608,
+      "grad_norm": 0.09473800659179688,
+      "learning_rate": 2.7350622043104957e-05,
+      "loss": 10.2698,
+      "step": 18140
+    },
+    {
+      "epoch": 0.09063896726510025,
+      "grad_norm": 0.09298662096261978,
+      "learning_rate": 2.734912012816341e-05,
+      "loss": 10.2763,
+      "step": 18150
+    },
+    {
+      "epoch": 0.0906889060900397,
+      "grad_norm": 0.09476146847009659,
+      "learning_rate": 2.7347618213221857e-05,
+      "loss": 10.2771,
+      "step": 18160
+    },
+    {
+      "epoch": 0.09073884491497915,
+      "grad_norm": 0.09148777276277542,
+      "learning_rate": 2.734611629828031e-05,
+      "loss": 10.2752,
+      "step": 18170
+    },
+    {
+      "epoch": 0.0907887837399186,
+      "grad_norm": 0.09469624608755112,
+      "learning_rate": 2.7344614383338757e-05,
+      "loss": 10.27,
+      "step": 18180
+    },
+    {
+      "epoch": 0.09083872256485805,
+      "grad_norm": 0.09769047796726227,
+      "learning_rate": 2.7343112468397204e-05,
+      "loss": 10.2733,
+      "step": 18190
+    },
+    {
+      "epoch": 0.0908886613897975,
+      "grad_norm": 0.09332720935344696,
+      "learning_rate": 2.7341610553455658e-05,
+      "loss": 10.2736,
+      "step": 18200
+    },
+    {
+      "epoch": 0.09093860021473694,
+      "grad_norm": 0.09613288938999176,
+      "learning_rate": 2.7340108638514105e-05,
+      "loss": 10.2693,
+      "step": 18210
+    },
+    {
+      "epoch": 0.0909885390396764,
+      "grad_norm": 0.09733350574970245,
+      "learning_rate": 2.7338606723572558e-05,
+      "loss": 10.2727,
+      "step": 18220
+    },
+    {
+      "epoch": 0.09103847786461584,
+      "grad_norm": 0.09478358179330826,
+      "learning_rate": 2.7337104808631005e-05,
+      "loss": 10.2741,
+      "step": 18230
+    },
+    {
+      "epoch": 0.0910884166895553,
+      "grad_norm": 0.0969361811876297,
+      "learning_rate": 2.733560289368945e-05,
+      "loss": 10.2716,
+      "step": 18240
+    },
+    {
+      "epoch": 0.09113835551449474,
+      "grad_norm": 0.09295036643743515,
+      "learning_rate": 2.7334100978747905e-05,
+      "loss": 10.2748,
+      "step": 18250
+    },
+    {
+      "epoch": 0.09118829433943419,
+      "grad_norm": 0.09057419002056122,
+      "learning_rate": 2.7332599063806352e-05,
+      "loss": 10.2733,
+      "step": 18260
+    },
+    {
+      "epoch": 0.09123823316437364,
+      "grad_norm": 0.09219224005937576,
+      "learning_rate": 2.7331097148864806e-05,
+      "loss": 10.2736,
+      "step": 18270
+    },
+    {
+      "epoch": 0.09128817198931309,
+      "grad_norm": 0.09382915496826172,
+      "learning_rate": 2.7329595233923252e-05,
+      "loss": 10.2732,
+      "step": 18280
+    },
+    {
+      "epoch": 0.09133811081425254,
+      "grad_norm": 0.09011480212211609,
+      "learning_rate": 2.7328093318981703e-05,
+      "loss": 10.2728,
+      "step": 18290
+    },
+    {
+      "epoch": 0.09138804963919199,
+      "grad_norm": 0.09689683467149734,
+      "learning_rate": 2.7326591404040153e-05,
+      "loss": 10.2674,
+      "step": 18300
+    },
+    {
+      "epoch": 0.09143798846413144,
+      "grad_norm": 0.09588810056447983,
+      "learning_rate": 2.73250894890986e-05,
+      "loss": 10.2684,
+      "step": 18310
+    },
+    {
+      "epoch": 0.09148792728907089,
+      "grad_norm": 0.09385774284601212,
+      "learning_rate": 2.7323587574157053e-05,
+      "loss": 10.2706,
+      "step": 18320
+    },
+    {
+      "epoch": 0.09153786611401034,
+      "grad_norm": 0.09056130051612854,
+      "learning_rate": 2.73220856592155e-05,
+      "loss": 10.27,
+      "step": 18330
+    },
+    {
+      "epoch": 0.09158780493894979,
+      "grad_norm": 0.09509393572807312,
+      "learning_rate": 2.732058374427395e-05,
+      "loss": 10.2709,
+      "step": 18340
+    },
+    {
+      "epoch": 0.09163774376388924,
+      "grad_norm": 0.09376955777406693,
+      "learning_rate": 2.73190818293324e-05,
+      "loss": 10.2682,
+      "step": 18350
+    },
+    {
+      "epoch": 0.09168768258882869,
+      "grad_norm": 0.10004284977912903,
+      "learning_rate": 2.7317579914390847e-05,
+      "loss": 10.2671,
+      "step": 18360
+    },
+    {
+      "epoch": 0.09173762141376814,
+      "grad_norm": 0.09222781658172607,
+      "learning_rate": 2.73160779994493e-05,
+      "loss": 10.2699,
+      "step": 18370
+    },
+    {
+      "epoch": 0.09178756023870759,
+      "grad_norm": 0.09528329968452454,
+      "learning_rate": 2.7314576084507747e-05,
+      "loss": 10.2711,
+      "step": 18380
+    },
+    {
+      "epoch": 0.09183749906364704,
+      "grad_norm": 0.09617341309785843,
+      "learning_rate": 2.7313074169566198e-05,
+      "loss": 10.2668,
+      "step": 18390
+    },
+    {
+      "epoch": 0.09188743788858648,
+      "grad_norm": 0.09537350386381149,
+      "learning_rate": 2.7311572254624648e-05,
+      "loss": 10.2692,
+      "step": 18400
+    },
+    {
+      "epoch": 0.09193737671352593,
+      "grad_norm": 0.09477705508470535,
+      "learning_rate": 2.7310070339683095e-05,
+      "loss": 10.2695,
+      "step": 18410
+    },
+    {
+      "epoch": 0.09198731553846538,
+      "grad_norm": 0.0950855240225792,
+      "learning_rate": 2.7308568424741548e-05,
+      "loss": 10.2652,
+      "step": 18420
+    },
+    {
+      "epoch": 0.09203725436340483,
+      "grad_norm": 0.09190298616886139,
+      "learning_rate": 2.7307066509799995e-05,
+      "loss": 10.2666,
+      "step": 18430
+    },
+    {
+      "epoch": 0.09208719318834428,
+      "grad_norm": 0.09800129383802414,
+      "learning_rate": 2.7305564594858445e-05,
+      "loss": 10.2699,
+      "step": 18440
+    },
+    {
+      "epoch": 0.09213713201328373,
+      "grad_norm": 0.09911572933197021,
+      "learning_rate": 2.7304062679916895e-05,
+      "loss": 10.2665,
+      "step": 18450
+    },
+    {
+      "epoch": 0.09218707083822318,
+      "grad_norm": 0.09081263840198517,
+      "learning_rate": 2.7302560764975342e-05,
+      "loss": 10.2648,
+      "step": 18460
+    },
+    {
+      "epoch": 0.09223700966316263,
+      "grad_norm": 0.09170430898666382,
+      "learning_rate": 2.7301058850033796e-05,
+      "loss": 10.2636,
+      "step": 18470
+    },
+    {
+      "epoch": 0.09228694848810208,
+      "grad_norm": 0.09938862174749374,
+      "learning_rate": 2.7299556935092242e-05,
+      "loss": 10.2678,
+      "step": 18480
+    },
+    {
+      "epoch": 0.09233688731304153,
+      "grad_norm": 0.09364289045333862,
+      "learning_rate": 2.7298055020150693e-05,
+      "loss": 10.2656,
+      "step": 18490
+    },
+    {
+      "epoch": 0.09238682613798098,
+      "grad_norm": 0.0970052108168602,
+      "learning_rate": 2.7296553105209143e-05,
+      "loss": 10.2641,
+      "step": 18500
+    },
+    {
+      "epoch": 0.09243676496292043,
+      "grad_norm": 0.09379066526889801,
+      "learning_rate": 2.729505119026759e-05,
+      "loss": 10.2681,
+      "step": 18510
+    },
+    {
+      "epoch": 0.09248670378785988,
+      "grad_norm": 0.0943591296672821,
+      "learning_rate": 2.7293549275326043e-05,
+      "loss": 10.2699,
+      "step": 18520
+    },
+    {
+      "epoch": 0.09253664261279933,
+      "grad_norm": 0.09379053860902786,
+      "learning_rate": 2.729204736038449e-05,
+      "loss": 10.2647,
+      "step": 18530
+    },
+    {
+      "epoch": 0.09258658143773878,
+      "grad_norm": 0.10091900825500488,
+      "learning_rate": 2.729054544544294e-05,
+      "loss": 10.2604,
+      "step": 18540
+    },
+    {
+      "epoch": 0.09263652026267821,
+      "grad_norm": 0.09774335473775864,
+      "learning_rate": 2.728904353050139e-05,
+      "loss": 10.263,
+      "step": 18550
+    },
+    {
+      "epoch": 0.09268645908761766,
+      "grad_norm": 0.0980936661362648,
+      "learning_rate": 2.7287541615559837e-05,
+      "loss": 10.2615,
+      "step": 18560
+    },
+    {
+      "epoch": 0.09273639791255711,
+      "grad_norm": 0.09376488626003265,
+      "learning_rate": 2.728603970061829e-05,
+      "loss": 10.2634,
+      "step": 18570
+    },
+    {
+      "epoch": 0.09278633673749656,
+      "grad_norm": 0.10236544907093048,
+      "learning_rate": 2.7284537785676737e-05,
+      "loss": 10.2677,
+      "step": 18580
+    },
+    {
+      "epoch": 0.09283627556243601,
+      "grad_norm": 0.09201527386903763,
+      "learning_rate": 2.7283035870735188e-05,
+      "loss": 10.2606,
+      "step": 18590
+    },
+    {
+      "epoch": 0.09288621438737546,
+      "grad_norm": 0.09707942605018616,
+      "learning_rate": 2.7281533955793638e-05,
+      "loss": 10.2669,
+      "step": 18600
+    },
+    {
+      "epoch": 0.09293615321231491,
+      "grad_norm": 0.09611304849386215,
+      "learning_rate": 2.7280032040852088e-05,
+      "loss": 10.2644,
+      "step": 18610
+    },
+    {
+      "epoch": 0.09298609203725436,
+      "grad_norm": 0.09938688576221466,
+      "learning_rate": 2.7278530125910538e-05,
+      "loss": 10.262,
+      "step": 18620
+    },
+    {
+      "epoch": 0.09303603086219381,
+      "grad_norm": 0.10115701705217361,
+      "learning_rate": 2.7277028210968985e-05,
+      "loss": 10.26,
+      "step": 18630
+    },
+    {
+      "epoch": 0.09308596968713326,
+      "grad_norm": 0.0930614024400711,
+      "learning_rate": 2.7275526296027435e-05,
+      "loss": 10.2623,
+      "step": 18640
+    },
+    {
+      "epoch": 0.09313590851207271,
+      "grad_norm": 0.09945568442344666,
+      "learning_rate": 2.7274024381085885e-05,
+      "loss": 10.2623,
+      "step": 18650
+    },
+    {
+      "epoch": 0.09318584733701216,
+      "grad_norm": 0.09213350713253021,
+      "learning_rate": 2.7272522466144335e-05,
+      "loss": 10.2614,
+      "step": 18660
+    },
+    {
+      "epoch": 0.0932357861619516,
+      "grad_norm": 0.09411916881799698,
+      "learning_rate": 2.7271020551202786e-05,
+      "loss": 10.2594,
+      "step": 18670
+    },
+    {
+      "epoch": 0.09328572498689106,
+      "grad_norm": 0.09325142949819565,
+      "learning_rate": 2.7269518636261232e-05,
+      "loss": 10.2569,
+      "step": 18680
+    },
+    {
+      "epoch": 0.0933356638118305,
+      "grad_norm": 0.09338999539613724,
+      "learning_rate": 2.7268016721319683e-05,
+      "loss": 10.2605,
+      "step": 18690
+    },
+    {
+      "epoch": 0.09338560263676995,
+      "grad_norm": 0.09095557034015656,
+      "learning_rate": 2.7266514806378133e-05,
+      "loss": 10.2629,
+      "step": 18700
+    },
+    {
+      "epoch": 0.0934355414617094,
+      "grad_norm": 0.09274104982614517,
+      "learning_rate": 2.7265012891436583e-05,
+      "loss": 10.2582,
+      "step": 18710
+    },
+    {
+      "epoch": 0.09348548028664885,
+      "grad_norm": 0.09546738117933273,
+      "learning_rate": 2.7263510976495033e-05,
+      "loss": 10.2595,
+      "step": 18720
+    },
+    {
+      "epoch": 0.0935354191115883,
+      "grad_norm": 0.0984828770160675,
+      "learning_rate": 2.726200906155348e-05,
+      "loss": 10.2569,
+      "step": 18730
+    },
+    {
+      "epoch": 0.09358535793652775,
+      "grad_norm": 0.09252218902111053,
+      "learning_rate": 2.726050714661193e-05,
+      "loss": 10.2585,
+      "step": 18740
+    },
+    {
+      "epoch": 0.0936352967614672,
+      "grad_norm": 0.09975884854793549,
+      "learning_rate": 2.725900523167038e-05,
+      "loss": 10.2581,
+      "step": 18750
+    },
+    {
+      "epoch": 0.09368523558640665,
+      "grad_norm": 0.08888278901576996,
+      "learning_rate": 2.725750331672883e-05,
+      "loss": 10.257,
+      "step": 18760
+    },
+    {
+      "epoch": 0.0937351744113461,
+      "grad_norm": 0.09609197080135345,
+      "learning_rate": 2.725600140178728e-05,
+      "loss": 10.2589,
+      "step": 18770
+    },
+    {
+      "epoch": 0.09378511323628555,
+      "grad_norm": 0.09457485377788544,
+      "learning_rate": 2.7254499486845727e-05,
+      "loss": 10.2558,
+      "step": 18780
+    },
+    {
+      "epoch": 0.093835052061225,
+      "grad_norm": 0.09543312340974808,
+      "learning_rate": 2.7252997571904178e-05,
+      "loss": 10.2583,
+      "step": 18790
+    },
+    {
+      "epoch": 0.09388499088616445,
+      "grad_norm": 0.08900411427021027,
+      "learning_rate": 2.7251495656962628e-05,
+      "loss": 10.2572,
+      "step": 18800
+    },
+    {
+      "epoch": 0.0939349297111039,
+      "grad_norm": 0.09842579811811447,
+      "learning_rate": 2.7249993742021078e-05,
+      "loss": 10.2532,
+      "step": 18810
+    },
+    {
+      "epoch": 0.09398486853604335,
+      "grad_norm": 0.0965215340256691,
+      "learning_rate": 2.7248491827079528e-05,
+      "loss": 10.2548,
+      "step": 18820
+    },
+    {
+      "epoch": 0.0940348073609828,
+      "grad_norm": 0.09778942912817001,
+      "learning_rate": 2.7246989912137975e-05,
+      "loss": 10.2543,
+      "step": 18830
+    },
+    {
+      "epoch": 0.09408474618592225,
+      "grad_norm": 0.09559378027915955,
+      "learning_rate": 2.7245487997196425e-05,
+      "loss": 10.2552,
+      "step": 18840
+    },
+    {
+      "epoch": 0.0941346850108617,
+      "grad_norm": 0.09030424058437347,
+      "learning_rate": 2.7243986082254875e-05,
+      "loss": 10.2562,
+      "step": 18850
+    },
+    {
+      "epoch": 0.09418462383580115,
+      "grad_norm": 0.09515605866909027,
+      "learning_rate": 2.7242484167313325e-05,
+      "loss": 10.2521,
+      "step": 18860
+    },
+    {
+      "epoch": 0.0942345626607406,
+      "grad_norm": 0.09284040331840515,
+      "learning_rate": 2.7240982252371776e-05,
+      "loss": 10.2552,
+      "step": 18870
+    },
+    {
+      "epoch": 0.09428450148568004,
+      "grad_norm": 0.09655296802520752,
+      "learning_rate": 2.7239480337430222e-05,
+      "loss": 10.2526,
+      "step": 18880
+    },
+    {
+      "epoch": 0.0943344403106195,
+      "grad_norm": 0.09709782898426056,
+      "learning_rate": 2.7237978422488673e-05,
+      "loss": 10.2554,
+      "step": 18890
+    },
+    {
+      "epoch": 0.09438437913555894,
+      "grad_norm": 0.09283418953418732,
+      "learning_rate": 2.7236476507547123e-05,
+      "loss": 10.2538,
+      "step": 18900
+    },
+    {
+      "epoch": 0.09443431796049839,
+      "grad_norm": 0.09396416693925858,
+      "learning_rate": 2.7234974592605573e-05,
+      "loss": 10.2539,
+      "step": 18910
+    },
+    {
+      "epoch": 0.09448425678543784,
+      "grad_norm": 0.1044858917593956,
+      "learning_rate": 2.7233472677664023e-05,
+      "loss": 10.2573,
+      "step": 18920
+    },
+    {
+      "epoch": 0.09453419561037729,
+      "grad_norm": 0.09172073006629944,
+      "learning_rate": 2.7231970762722473e-05,
+      "loss": 10.2548,
+      "step": 18930
+    },
+    {
+      "epoch": 0.09458413443531674,
+      "grad_norm": 0.09052781015634537,
+      "learning_rate": 2.723046884778092e-05,
+      "loss": 10.2535,
+      "step": 18940
+    },
+    {
+      "epoch": 0.09463407326025619,
+      "grad_norm": 0.09258178621530533,
+      "learning_rate": 2.722896693283937e-05,
+      "loss": 10.2505,
+      "step": 18950
+    },
+    {
+      "epoch": 0.09468401208519564,
+      "grad_norm": 0.09820058196783066,
+      "learning_rate": 2.722746501789782e-05,
+      "loss": 10.2543,
+      "step": 18960
+    },
+    {
+      "epoch": 0.09473395091013509,
+      "grad_norm": 0.09973189234733582,
+      "learning_rate": 2.722596310295627e-05,
+      "loss": 10.2533,
+      "step": 18970
+    },
+    {
+      "epoch": 0.09478388973507454,
+      "grad_norm": 0.09172434359788895,
+      "learning_rate": 2.722446118801472e-05,
+      "loss": 10.2502,
+      "step": 18980
+    },
+    {
+      "epoch": 0.09483382856001399,
+      "grad_norm": 0.0971081554889679,
+      "learning_rate": 2.7222959273073168e-05,
+      "loss": 10.2485,
+      "step": 18990
+    },
+    {
+      "epoch": 0.09488376738495344,
+      "grad_norm": 0.09259314090013504,
+      "learning_rate": 2.7221457358131618e-05,
+      "loss": 10.2516,
+      "step": 19000
+    },
+    {
+      "epoch": 0.09493370620989289,
+      "grad_norm": 0.10186239331960678,
+      "learning_rate": 2.7219955443190068e-05,
+      "loss": 10.2502,
+      "step": 19010
+    },
+    {
+      "epoch": 0.09498364503483234,
+      "grad_norm": 0.09782648831605911,
+      "learning_rate": 2.7218453528248518e-05,
+      "loss": 10.254,
+      "step": 19020
+    },
+    {
+      "epoch": 0.09503358385977179,
+      "grad_norm": 0.09522926807403564,
+      "learning_rate": 2.7216951613306968e-05,
+      "loss": 10.2542,
+      "step": 19030
+    },
+    {
+      "epoch": 0.09508352268471122,
+      "grad_norm": 0.09580672532320023,
+      "learning_rate": 2.7215449698365415e-05,
+      "loss": 10.2509,
+      "step": 19040
+    },
+    {
+      "epoch": 0.09513346150965067,
+      "grad_norm": 0.10172213613986969,
+      "learning_rate": 2.7213947783423865e-05,
+      "loss": 10.2519,
+      "step": 19050
+    },
+    {
+      "epoch": 0.09518340033459012,
+      "grad_norm": 0.09953866899013519,
+      "learning_rate": 2.7212445868482315e-05,
+      "loss": 10.2514,
+      "step": 19060
+    },
+    {
+      "epoch": 0.09523333915952957,
+      "grad_norm": 0.09473235160112381,
+      "learning_rate": 2.7210943953540766e-05,
+      "loss": 10.2493,
+      "step": 19070
+    },
+    {
+      "epoch": 0.09528327798446902,
+      "grad_norm": 0.09612436592578888,
+      "learning_rate": 2.7209442038599216e-05,
+      "loss": 10.2495,
+      "step": 19080
+    },
+    {
+      "epoch": 0.09533321680940847,
+      "grad_norm": 0.09395024180412292,
+      "learning_rate": 2.7207940123657663e-05,
+      "loss": 10.2494,
+      "step": 19090
+    },
+    {
+      "epoch": 0.09538315563434792,
+      "grad_norm": 0.0924721285700798,
+      "learning_rate": 2.7206438208716113e-05,
+      "loss": 10.2493,
+      "step": 19100
+    },
+    {
+      "epoch": 0.09543309445928737,
+      "grad_norm": 0.09010235965251923,
+      "learning_rate": 2.7204936293774563e-05,
+      "loss": 10.2508,
+      "step": 19110
+    },
+    {
+      "epoch": 0.09548303328422682,
+      "grad_norm": 0.09913980215787888,
+      "learning_rate": 2.7203434378833013e-05,
+      "loss": 10.2485,
+      "step": 19120
+    },
+    {
+      "epoch": 0.09553297210916627,
+      "grad_norm": 0.09284950792789459,
+      "learning_rate": 2.7201932463891463e-05,
+      "loss": 10.2497,
+      "step": 19130
+    },
+    {
+      "epoch": 0.09558291093410572,
+      "grad_norm": 0.09394965320825577,
+      "learning_rate": 2.720043054894991e-05,
+      "loss": 10.2504,
+      "step": 19140
+    },
+    {
+      "epoch": 0.09563284975904517,
+      "grad_norm": 0.09422177821397781,
+      "learning_rate": 2.719892863400836e-05,
+      "loss": 10.2443,
+      "step": 19150
+    },
+    {
+      "epoch": 0.09568278858398462,
+      "grad_norm": 0.09094090759754181,
+      "learning_rate": 2.719742671906681e-05,
+      "loss": 10.2472,
+      "step": 19160
+    },
+    {
+      "epoch": 0.09573272740892406,
+      "grad_norm": 0.09635195881128311,
+      "learning_rate": 2.719592480412526e-05,
+      "loss": 10.2485,
+      "step": 19170
+    },
+    {
+      "epoch": 0.09578266623386351,
+      "grad_norm": 0.09993939101696014,
+      "learning_rate": 2.719442288918371e-05,
+      "loss": 10.2454,
+      "step": 19180
+    },
+    {
+      "epoch": 0.09583260505880296,
+      "grad_norm": 0.09371164441108704,
+      "learning_rate": 2.7192920974242158e-05,
+      "loss": 10.2504,
+      "step": 19190
+    },
+    {
+      "epoch": 0.09588254388374241,
+      "grad_norm": 0.09679362177848816,
+      "learning_rate": 2.7191419059300608e-05,
+      "loss": 10.2465,
+      "step": 19200
+    },
+    {
+      "epoch": 0.09593248270868186,
+      "grad_norm": 0.08799337595701218,
+      "learning_rate": 2.7189917144359058e-05,
+      "loss": 10.2454,
+      "step": 19210
+    },
+    {
+      "epoch": 0.09598242153362131,
+      "grad_norm": 0.09332183748483658,
+      "learning_rate": 2.7188415229417508e-05,
+      "loss": 10.2492,
+      "step": 19220
+    },
+    {
+      "epoch": 0.09603236035856076,
+      "grad_norm": 0.09035744518041611,
+      "learning_rate": 2.718691331447596e-05,
+      "loss": 10.2473,
+      "step": 19230
+    },
+    {
+      "epoch": 0.09608229918350021,
+      "grad_norm": 0.09255316108465195,
+      "learning_rate": 2.7185411399534405e-05,
+      "loss": 10.2487,
+      "step": 19240
+    },
+    {
+      "epoch": 0.09613223800843966,
+      "grad_norm": 0.0899682343006134,
+      "learning_rate": 2.718390948459286e-05,
+      "loss": 10.2411,
+      "step": 19250
+    },
+    {
+      "epoch": 0.09618217683337911,
+      "grad_norm": 0.09768253564834595,
+      "learning_rate": 2.7182407569651305e-05,
+      "loss": 10.2458,
+      "step": 19260
+    },
+    {
+      "epoch": 0.09623211565831856,
+      "grad_norm": 0.09567224979400635,
+      "learning_rate": 2.7180905654709756e-05,
+      "loss": 10.2468,
+      "step": 19270
+    },
+    {
+      "epoch": 0.09628205448325801,
+      "grad_norm": 0.09296666830778122,
+      "learning_rate": 2.7179403739768206e-05,
+      "loss": 10.2441,
+      "step": 19280
+    },
+    {
+      "epoch": 0.09633199330819746,
+      "grad_norm": 0.09625297039747238,
+      "learning_rate": 2.7177901824826653e-05,
+      "loss": 10.2453,
+      "step": 19290
+    },
+    {
+      "epoch": 0.09638193213313691,
+      "grad_norm": 0.09271930158138275,
+      "learning_rate": 2.7176399909885106e-05,
+      "loss": 10.2413,
+      "step": 19300
+    },
+    {
+      "epoch": 0.09643187095807636,
+      "grad_norm": 0.09876267611980438,
+      "learning_rate": 2.7174897994943553e-05,
+      "loss": 10.2448,
+      "step": 19310
+    },
+    {
+      "epoch": 0.0964818097830158,
+      "grad_norm": 0.09215805679559708,
+      "learning_rate": 2.7173396080002003e-05,
+      "loss": 10.2435,
+      "step": 19320
+    },
+    {
+      "epoch": 0.09653174860795526,
+      "grad_norm": 0.09160521626472473,
+      "learning_rate": 2.7171894165060453e-05,
+      "loss": 10.2456,
+      "step": 19330
+    },
+    {
+      "epoch": 0.0965816874328947,
+      "grad_norm": 0.09767121821641922,
+      "learning_rate": 2.71703922501189e-05,
+      "loss": 10.2404,
+      "step": 19340
+    },
+    {
+      "epoch": 0.09663162625783416,
+      "grad_norm": 0.09696942567825317,
+      "learning_rate": 2.7168890335177354e-05,
+      "loss": 10.2442,
+      "step": 19350
+    },
+    {
+      "epoch": 0.0966815650827736,
+      "grad_norm": 0.09589537233114243,
+      "learning_rate": 2.71673884202358e-05,
+      "loss": 10.2434,
+      "step": 19360
+    },
+    {
+      "epoch": 0.09673150390771305,
+      "grad_norm": 0.09620683640241623,
+      "learning_rate": 2.716588650529425e-05,
+      "loss": 10.2424,
+      "step": 19370
+    },
+    {
+      "epoch": 0.0967814427326525,
+      "grad_norm": 0.08876433968544006,
+      "learning_rate": 2.71643845903527e-05,
+      "loss": 10.2431,
+      "step": 19380
+    },
+    {
+      "epoch": 0.09683138155759195,
+      "grad_norm": 0.094940684735775,
+      "learning_rate": 2.7162882675411148e-05,
+      "loss": 10.2415,
+      "step": 19390
+    },
+    {
+      "epoch": 0.0968813203825314,
+      "grad_norm": 0.09419713914394379,
+      "learning_rate": 2.71613807604696e-05,
+      "loss": 10.2437,
+      "step": 19400
+    },
+    {
+      "epoch": 0.09693125920747085,
+      "grad_norm": 0.09718811511993408,
+      "learning_rate": 2.7159878845528048e-05,
+      "loss": 10.2393,
+      "step": 19410
+    },
+    {
+      "epoch": 0.0969811980324103,
+      "grad_norm": 0.09436991810798645,
+      "learning_rate": 2.7158376930586498e-05,
+      "loss": 10.2408,
+      "step": 19420
+    },
+    {
+      "epoch": 0.09703113685734975,
+      "grad_norm": 0.09804237633943558,
+      "learning_rate": 2.715687501564495e-05,
+      "loss": 10.24,
+      "step": 19430
+    },
+    {
+      "epoch": 0.0970810756822892,
+      "grad_norm": 0.09041360020637512,
+      "learning_rate": 2.7155373100703395e-05,
+      "loss": 10.2429,
+      "step": 19440
+    },
+    {
+      "epoch": 0.09713101450722865,
+      "grad_norm": 0.09413240849971771,
+      "learning_rate": 2.715387118576185e-05,
+      "loss": 10.2405,
+      "step": 19450
+    },
+    {
+      "epoch": 0.0971809533321681,
+      "grad_norm": 0.09756959229707718,
+      "learning_rate": 2.7152369270820295e-05,
+      "loss": 10.4918,
+      "step": 19460
+    },
+    {
+      "epoch": 0.09723089215710755,
+      "grad_norm": 0.08898182958364487,
+      "learning_rate": 2.7150867355878746e-05,
+      "loss": 10.2411,
+      "step": 19470
+    },
+    {
+      "epoch": 0.097280830982047,
+      "grad_norm": 0.09335238486528397,
+      "learning_rate": 2.7149365440937196e-05,
+      "loss": 10.2348,
+      "step": 19480
+    },
+    {
+      "epoch": 0.09733076980698645,
+      "grad_norm": 0.09306208789348602,
+      "learning_rate": 2.7147863525995643e-05,
+      "loss": 10.2427,
+      "step": 19490
+    },
+    {
+      "epoch": 0.0973807086319259,
+      "grad_norm": 0.09693284332752228,
+      "learning_rate": 2.7146361611054096e-05,
+      "loss": 10.2377,
+      "step": 19500
+    },
+    {
+      "epoch": 0.09743064745686535,
+      "grad_norm": 0.10119107365608215,
+      "learning_rate": 2.7144859696112543e-05,
+      "loss": 10.2366,
+      "step": 19510
+    },
+    {
+      "epoch": 0.0974805862818048,
+      "grad_norm": 0.10089292377233505,
+      "learning_rate": 2.7143357781170993e-05,
+      "loss": 10.2385,
+      "step": 19520
+    },
+    {
+      "epoch": 0.09753052510674425,
+      "grad_norm": 0.09562947601079941,
+      "learning_rate": 2.7141855866229443e-05,
+      "loss": 10.2378,
+      "step": 19530
+    },
+    {
+      "epoch": 0.09758046393168368,
+      "grad_norm": 0.09517926722764969,
+      "learning_rate": 2.714035395128789e-05,
+      "loss": 10.237,
+      "step": 19540
+    },
+    {
+      "epoch": 0.09763040275662313,
+      "grad_norm": 0.0980682447552681,
+      "learning_rate": 2.7138852036346344e-05,
+      "loss": 10.2385,
+      "step": 19550
+    },
+    {
+      "epoch": 0.09768034158156258,
+      "grad_norm": 0.09545796364545822,
+      "learning_rate": 2.713735012140479e-05,
+      "loss": 10.2375,
+      "step": 19560
+    },
+    {
+      "epoch": 0.09773028040650203,
+      "grad_norm": 0.09550634771585464,
+      "learning_rate": 2.7135848206463244e-05,
+      "loss": 10.2384,
+      "step": 19570
+    },
+    {
+      "epoch": 0.09778021923144148,
+      "grad_norm": 0.09427410364151001,
+      "learning_rate": 2.713434629152169e-05,
+      "loss": 10.2344,
+      "step": 19580
+    },
+    {
+      "epoch": 0.09783015805638093,
+      "grad_norm": 0.09462705999612808,
+      "learning_rate": 2.7132844376580138e-05,
+      "loss": 10.2342,
+      "step": 19590
+    },
+    {
+      "epoch": 0.09788009688132038,
+      "grad_norm": 0.1018369272351265,
+      "learning_rate": 2.713134246163859e-05,
+      "loss": 10.2329,
+      "step": 19600
+    },
+    {
+      "epoch": 0.09793003570625983,
+      "grad_norm": 0.09698480367660522,
+      "learning_rate": 2.7129840546697038e-05,
+      "loss": 10.2358,
+      "step": 19610
+    },
+    {
+      "epoch": 0.09797997453119928,
+      "grad_norm": 0.09769893437623978,
+      "learning_rate": 2.712833863175549e-05,
+      "loss": 10.2382,
+      "step": 19620
+    },
+    {
+      "epoch": 0.09802991335613873,
+      "grad_norm": 0.09582178294658661,
+      "learning_rate": 2.712683671681394e-05,
+      "loss": 10.2342,
+      "step": 19630
+    },
+    {
+      "epoch": 0.09807985218107818,
+      "grad_norm": 0.09619859606027603,
+      "learning_rate": 2.7125334801872385e-05,
+      "loss": 10.2356,
+      "step": 19640
+    },
+    {
+      "epoch": 0.09812979100601762,
+      "grad_norm": 0.09638599306344986,
+      "learning_rate": 2.712383288693084e-05,
+      "loss": 10.2338,
+      "step": 19650
+    },
+    {
+      "epoch": 0.09817972983095707,
+      "grad_norm": 0.10089550167322159,
+      "learning_rate": 2.7122330971989285e-05,
+      "loss": 10.2301,
+      "step": 19660
+    },
+    {
+      "epoch": 0.09822966865589652,
+      "grad_norm": 0.09721601009368896,
+      "learning_rate": 2.712082905704774e-05,
+      "loss": 10.2364,
+      "step": 19670
+    },
+    {
+      "epoch": 0.09827960748083597,
+      "grad_norm": 0.09357859939336777,
+      "learning_rate": 2.7119327142106186e-05,
+      "loss": 10.236,
+      "step": 19680
+    },
+    {
+      "epoch": 0.09832954630577542,
+      "grad_norm": 0.09338711202144623,
+      "learning_rate": 2.7117825227164633e-05,
+      "loss": 10.2298,
+      "step": 19690
+    },
+    {
+      "epoch": 0.09837948513071487,
+      "grad_norm": 0.09414970874786377,
+      "learning_rate": 2.7116323312223086e-05,
+      "loss": 10.2383,
+      "step": 19700
+    },
+    {
+      "epoch": 0.09842942395565432,
+      "grad_norm": 0.10141325742006302,
+      "learning_rate": 2.7114821397281533e-05,
+      "loss": 10.2353,
+      "step": 19710
+    },
+    {
+      "epoch": 0.09847936278059377,
+      "grad_norm": 0.09696213901042938,
+      "learning_rate": 2.7113319482339987e-05,
+      "loss": 10.2326,
+      "step": 19720
+    },
+    {
+      "epoch": 0.09852930160553322,
+      "grad_norm": 0.09888730943202972,
+      "learning_rate": 2.7111817567398433e-05,
+      "loss": 10.2354,
+      "step": 19730
+    },
+    {
+      "epoch": 0.09857924043047267,
+      "grad_norm": 0.089784175157547,
+      "learning_rate": 2.711031565245688e-05,
+      "loss": 10.2333,
+      "step": 19740
+    },
+    {
+      "epoch": 0.09862917925541212,
+      "grad_norm": 0.09449610859155655,
+      "learning_rate": 2.7108813737515334e-05,
+      "loss": 10.233,
+      "step": 19750
+    },
+    {
+      "epoch": 0.09867911808035157,
+      "grad_norm": 0.09268537908792496,
+      "learning_rate": 2.710731182257378e-05,
+      "loss": 10.2273,
+      "step": 19760
+    },
+    {
+      "epoch": 0.09872905690529102,
+      "grad_norm": 0.09909249097108841,
+      "learning_rate": 2.7105809907632234e-05,
+      "loss": 10.2293,
+      "step": 19770
+    },
+    {
+      "epoch": 0.09877899573023047,
+      "grad_norm": 0.10036230832338333,
+      "learning_rate": 2.710430799269068e-05,
+      "loss": 10.2342,
+      "step": 19780
+    },
+    {
+      "epoch": 0.09882893455516992,
+      "grad_norm": 0.10188131779432297,
+      "learning_rate": 2.7102806077749128e-05,
+      "loss": 10.2328,
+      "step": 19790
+    },
+    {
+      "epoch": 0.09887887338010937,
+      "grad_norm": 0.09562569111585617,
+      "learning_rate": 2.710130416280758e-05,
+      "loss": 10.2278,
+      "step": 19800
+    },
+    {
+      "epoch": 0.09892881220504882,
+      "grad_norm": 0.09543158859014511,
+      "learning_rate": 2.7099802247866028e-05,
+      "loss": 10.2257,
+      "step": 19810
+    },
+    {
+      "epoch": 0.09897875102998827,
+      "grad_norm": 0.09362302720546722,
+      "learning_rate": 2.709830033292448e-05,
+      "loss": 10.2302,
+      "step": 19820
+    },
+    {
+      "epoch": 0.09902868985492772,
+      "grad_norm": 0.09583684056997299,
+      "learning_rate": 2.709679841798293e-05,
+      "loss": 10.2322,
+      "step": 19830
+    },
+    {
+      "epoch": 0.09907862867986716,
+      "grad_norm": 0.09406079351902008,
+      "learning_rate": 2.7095296503041375e-05,
+      "loss": 10.2291,
+      "step": 19840
+    },
+    {
+      "epoch": 0.09912856750480661,
+      "grad_norm": 0.09514211863279343,
+      "learning_rate": 2.709379458809983e-05,
+      "loss": 10.2303,
+      "step": 19850
+    },
+    {
+      "epoch": 0.09917850632974606,
+      "grad_norm": 0.09583709388971329,
+      "learning_rate": 2.7092292673158276e-05,
+      "loss": 10.2299,
+      "step": 19860
+    },
+    {
+      "epoch": 0.09922844515468551,
+      "grad_norm": 0.09366525709629059,
+      "learning_rate": 2.709079075821673e-05,
+      "loss": 10.2287,
+      "step": 19870
+    },
+    {
+      "epoch": 0.09927838397962496,
+      "grad_norm": 0.09371297061443329,
+      "learning_rate": 2.7089288843275176e-05,
+      "loss": 10.2276,
+      "step": 19880
+    },
+    {
+      "epoch": 0.09932832280456441,
+      "grad_norm": 0.09627794474363327,
+      "learning_rate": 2.7087786928333626e-05,
+      "loss": 10.2293,
+      "step": 19890
+    },
+    {
+      "epoch": 0.09937826162950386,
+      "grad_norm": 0.09279222786426544,
+      "learning_rate": 2.7086285013392076e-05,
+      "loss": 10.2283,
+      "step": 19900
+    },
+    {
+      "epoch": 0.09942820045444331,
+      "grad_norm": 0.0960431769490242,
+      "learning_rate": 2.7084783098450523e-05,
+      "loss": 10.2295,
+      "step": 19910
+    },
+    {
+      "epoch": 0.09947813927938276,
+      "grad_norm": 0.0974978655576706,
+      "learning_rate": 2.7083281183508977e-05,
+      "loss": 10.2283,
+      "step": 19920
+    },
+    {
+      "epoch": 0.09952807810432221,
+      "grad_norm": 0.09452254325151443,
+      "learning_rate": 2.7081779268567423e-05,
+      "loss": 10.2259,
+      "step": 19930
+    },
+    {
+      "epoch": 0.09957801692926166,
+      "grad_norm": 0.09121208637952805,
+      "learning_rate": 2.7080277353625874e-05,
+      "loss": 10.2277,
+      "step": 19940
+    },
+    {
+      "epoch": 0.09962795575420111,
+      "grad_norm": 0.09757377207279205,
+      "learning_rate": 2.7078775438684324e-05,
+      "loss": 10.2261,
+      "step": 19950
+    },
+    {
+      "epoch": 0.09967789457914056,
+      "grad_norm": 0.09297657012939453,
+      "learning_rate": 2.707727352374277e-05,
+      "loss": 10.2242,
+      "step": 19960
+    },
+    {
+      "epoch": 0.09972783340408001,
+      "grad_norm": 0.09815360605716705,
+      "learning_rate": 2.7075771608801224e-05,
+      "loss": 10.2245,
+      "step": 19970
+    },
+    {
+      "epoch": 0.09977777222901946,
+      "grad_norm": 0.09438446164131165,
+      "learning_rate": 2.707426969385967e-05,
+      "loss": 10.229,
+      "step": 19980
+    },
+    {
+      "epoch": 0.0998277110539589,
+      "grad_norm": 0.09827311336994171,
+      "learning_rate": 2.707276777891812e-05,
+      "loss": 10.2242,
+      "step": 19990
+    },
+    {
+      "epoch": 0.09987764987889836,
+      "grad_norm": 0.09929145127534866,
+      "learning_rate": 2.707126586397657e-05,
+      "loss": 10.2251,
+      "step": 20000
+    },
+    {
+      "epoch": 0.0999275887038378,
+      "grad_norm": 0.09331025183200836,
+      "learning_rate": 2.7069763949035018e-05,
+      "loss": 10.2282,
+      "step": 20010
+    },
+    {
+      "epoch": 0.09997752752877725,
+      "grad_norm": 0.09668101370334625,
+      "learning_rate": 2.706826203409347e-05,
+      "loss": 10.2285,
+      "step": 20020
+    },
+    {
+      "epoch": 0.10002746635371669,
+      "grad_norm": 0.09241931140422821,
+      "learning_rate": 2.706676011915192e-05,
+      "loss": 10.2261,
+      "step": 20030
+    },
+    {
+      "epoch": 0.10007740517865614,
+      "grad_norm": 0.09374601393938065,
+      "learning_rate": 2.706525820421037e-05,
+      "loss": 10.2251,
+      "step": 20040
+    },
+    {
+      "epoch": 0.10012734400359559,
+      "grad_norm": 0.09970963001251221,
+      "learning_rate": 2.706375628926882e-05,
+      "loss": 10.2242,
+      "step": 20050
+    },
+    {
+      "epoch": 0.10017728282853504,
+      "grad_norm": 0.10114763677120209,
+      "learning_rate": 2.7062254374327266e-05,
+      "loss": 10.2234,
+      "step": 20060
+    },
+    {
+      "epoch": 0.10022722165347449,
+      "grad_norm": 0.09584357589483261,
+      "learning_rate": 2.706075245938572e-05,
+      "loss": 10.223,
+      "step": 20070
+    },
+    {
+      "epoch": 0.10027716047841394,
+      "grad_norm": 0.09605442732572556,
+      "learning_rate": 2.7059250544444166e-05,
+      "loss": 10.2236,
+      "step": 20080
+    },
+    {
+      "epoch": 0.10032709930335339,
+      "grad_norm": 0.09641440957784653,
+      "learning_rate": 2.7057748629502616e-05,
+      "loss": 10.2236,
+      "step": 20090
+    },
+    {
+      "epoch": 0.10037703812829284,
+      "grad_norm": 0.09594506770372391,
+      "learning_rate": 2.7056246714561066e-05,
+      "loss": 10.2208,
+      "step": 20100
+    },
+    {
+      "epoch": 0.10042697695323229,
+      "grad_norm": 0.09294579178094864,
+      "learning_rate": 2.7054744799619513e-05,
+      "loss": 10.2232,
+      "step": 20110
+    },
+    {
+      "epoch": 0.10047691577817174,
+      "grad_norm": 0.09371384978294373,
+      "learning_rate": 2.7053242884677967e-05,
+      "loss": 10.2246,
+      "step": 20120
+    },
+    {
+      "epoch": 0.10052685460311118,
+      "grad_norm": 0.0971645787358284,
+      "learning_rate": 2.7051740969736413e-05,
+      "loss": 10.2199,
+      "step": 20130
+    },
+    {
+      "epoch": 0.10057679342805063,
+      "grad_norm": 0.09300613403320312,
+      "learning_rate": 2.7050239054794864e-05,
+      "loss": 10.2229,
+      "step": 20140
+    },
+    {
+      "epoch": 0.10062673225299008,
+      "grad_norm": 0.09003175050020218,
+      "learning_rate": 2.7048737139853314e-05,
+      "loss": 10.2235,
+      "step": 20150
+    },
+    {
+      "epoch": 0.10067667107792953,
+      "grad_norm": 0.09512855857610703,
+      "learning_rate": 2.704723522491176e-05,
+      "loss": 10.2204,
+      "step": 20160
+    },
+    {
+      "epoch": 0.10072660990286898,
+      "grad_norm": 0.10006555169820786,
+      "learning_rate": 2.7045733309970214e-05,
+      "loss": 10.2225,
+      "step": 20170
+    },
+    {
+      "epoch": 0.10077654872780843,
+      "grad_norm": 0.09965218603610992,
+      "learning_rate": 2.704423139502866e-05,
+      "loss": 10.2212,
+      "step": 20180
+    },
+    {
+      "epoch": 0.10082648755274788,
+      "grad_norm": 0.0935189425945282,
+      "learning_rate": 2.7042729480087114e-05,
+      "loss": 10.2217,
+      "step": 20190
+    },
+    {
+      "epoch": 0.10087642637768733,
+      "grad_norm": 0.09258928894996643,
+      "learning_rate": 2.704122756514556e-05,
+      "loss": 10.2222,
+      "step": 20200
+    },
+    {
+      "epoch": 0.10092636520262678,
+      "grad_norm": 0.0927526131272316,
+      "learning_rate": 2.703972565020401e-05,
+      "loss": 10.2206,
+      "step": 20210
+    },
+    {
+      "epoch": 0.10097630402756623,
+      "grad_norm": 0.09727808088064194,
+      "learning_rate": 2.703822373526246e-05,
+      "loss": 10.2186,
+      "step": 20220
+    },
+    {
+      "epoch": 0.10102624285250568,
+      "grad_norm": 0.09847782552242279,
+      "learning_rate": 2.703672182032091e-05,
+      "loss": 10.2159,
+      "step": 20230
+    },
+    {
+      "epoch": 0.10107618167744513,
+      "grad_norm": 0.10164419561624527,
+      "learning_rate": 2.7035219905379362e-05,
+      "loss": 10.2183,
+      "step": 20240
+    },
+    {
+      "epoch": 0.10112612050238458,
+      "grad_norm": 0.0917501449584961,
+      "learning_rate": 2.703371799043781e-05,
+      "loss": 10.2188,
+      "step": 20250
+    },
+    {
+      "epoch": 0.10117605932732403,
+      "grad_norm": 0.09016897529363632,
+      "learning_rate": 2.703221607549626e-05,
+      "loss": 10.2207,
+      "step": 20260
+    },
+    {
+      "epoch": 0.10122599815226348,
+      "grad_norm": 0.09659720957279205,
+      "learning_rate": 2.703071416055471e-05,
+      "loss": 10.2198,
+      "step": 20270
+    },
+    {
+      "epoch": 0.10127593697720293,
+      "grad_norm": 0.09625518321990967,
+      "learning_rate": 2.7029212245613156e-05,
+      "loss": 10.2209,
+      "step": 20280
+    },
+    {
+      "epoch": 0.10132587580214238,
+      "grad_norm": 0.09196675568819046,
+      "learning_rate": 2.702771033067161e-05,
+      "loss": 10.2194,
+      "step": 20290
+    },
+    {
+      "epoch": 0.10137581462708183,
+      "grad_norm": 0.09483517706394196,
+      "learning_rate": 2.7026208415730056e-05,
+      "loss": 10.217,
+      "step": 20300
+    },
+    {
+      "epoch": 0.10142575345202128,
+      "grad_norm": 0.09146818518638611,
+      "learning_rate": 2.7024706500788506e-05,
+      "loss": 10.2151,
+      "step": 20310
+    },
+    {
+      "epoch": 0.10147569227696072,
+      "grad_norm": 0.09970217198133469,
+      "learning_rate": 2.7023204585846957e-05,
+      "loss": 10.2176,
+      "step": 20320
+    },
+    {
+      "epoch": 0.10152563110190017,
+      "grad_norm": 0.10319959372282028,
+      "learning_rate": 2.7021702670905403e-05,
+      "loss": 10.2165,
+      "step": 20330
+    },
+    {
+      "epoch": 0.10157556992683962,
+      "grad_norm": 0.09527314454317093,
+      "learning_rate": 2.7020200755963857e-05,
+      "loss": 10.2176,
+      "step": 20340
+    },
+    {
+      "epoch": 0.10162550875177907,
+      "grad_norm": 0.09798305481672287,
+      "learning_rate": 2.7018698841022304e-05,
+      "loss": 10.2151,
+      "step": 20350
+    },
+    {
+      "epoch": 0.10167544757671852,
+      "grad_norm": 0.10353653877973557,
+      "learning_rate": 2.7017196926080754e-05,
+      "loss": 10.2197,
+      "step": 20360
+    },
+    {
+      "epoch": 0.10172538640165797,
+      "grad_norm": 0.09512107819318771,
+      "learning_rate": 2.7015695011139204e-05,
+      "loss": 10.2186,
+      "step": 20370
+    },
+    {
+      "epoch": 0.10177532522659742,
+      "grad_norm": 0.09513498097658157,
+      "learning_rate": 2.701419309619765e-05,
+      "loss": 10.2179,
+      "step": 20380
+    },
+    {
+      "epoch": 0.10182526405153687,
+      "grad_norm": 0.09943837672472,
+      "learning_rate": 2.7012691181256104e-05,
+      "loss": 10.2123,
+      "step": 20390
+    },
+    {
+      "epoch": 0.10187520287647632,
+      "grad_norm": 0.09323469549417496,
+      "learning_rate": 2.701118926631455e-05,
+      "loss": 10.2183,
+      "step": 20400
+    },
+    {
+      "epoch": 0.10192514170141577,
+      "grad_norm": 0.09345997124910355,
+      "learning_rate": 2.7009687351373e-05,
+      "loss": 10.2135,
+      "step": 20410
+    },
+    {
+      "epoch": 0.10197508052635522,
+      "grad_norm": 0.10201733559370041,
+      "learning_rate": 2.700818543643145e-05,
+      "loss": 10.2155,
+      "step": 20420
+    },
+    {
+      "epoch": 0.10202501935129467,
+      "grad_norm": 0.08883191645145416,
+      "learning_rate": 2.70066835214899e-05,
+      "loss": 10.2125,
+      "step": 20430
+    },
+    {
+      "epoch": 0.10207495817623412,
+      "grad_norm": 0.09999528527259827,
+      "learning_rate": 2.7005181606548352e-05,
+      "loss": 10.2138,
+      "step": 20440
+    },
+    {
+      "epoch": 0.10212489700117357,
+      "grad_norm": 0.09363902360200882,
+      "learning_rate": 2.70036796916068e-05,
+      "loss": 10.2105,
+      "step": 20450
+    },
+    {
+      "epoch": 0.10217483582611302,
+      "grad_norm": 0.1012590229511261,
+      "learning_rate": 2.700217777666525e-05,
+      "loss": 10.2133,
+      "step": 20460
+    },
+    {
+      "epoch": 0.10222477465105247,
+      "grad_norm": 0.09568557888269424,
+      "learning_rate": 2.70006758617237e-05,
+      "loss": 10.2142,
+      "step": 20470
+    },
+    {
+      "epoch": 0.10227471347599192,
+      "grad_norm": 0.10000815987586975,
+      "learning_rate": 2.6999173946782146e-05,
+      "loss": 10.2139,
+      "step": 20480
+    },
+    {
+      "epoch": 0.10232465230093137,
+      "grad_norm": 0.09673640131950378,
+      "learning_rate": 2.69976720318406e-05,
+      "loss": 10.2101,
+      "step": 20490
+    },
+    {
+      "epoch": 0.10237459112587081,
+      "grad_norm": 0.09184068441390991,
+      "learning_rate": 2.6996170116899046e-05,
+      "loss": 10.2125,
+      "step": 20500
+    },
+    {
+      "epoch": 0.10242452995081026,
+      "grad_norm": 0.09533202648162842,
+      "learning_rate": 2.6994668201957496e-05,
+      "loss": 10.2116,
+      "step": 20510
+    },
+    {
+      "epoch": 0.1024744687757497,
+      "grad_norm": 0.09686296433210373,
+      "learning_rate": 2.6993166287015947e-05,
+      "loss": 10.213,
+      "step": 20520
+    },
+    {
+      "epoch": 0.10252440760068915,
+      "grad_norm": 0.09219254553318024,
+      "learning_rate": 2.6991664372074397e-05,
+      "loss": 10.2101,
+      "step": 20530
+    },
+    {
+      "epoch": 0.1025743464256286,
+      "grad_norm": 0.0953301191329956,
+      "learning_rate": 2.6990162457132847e-05,
+      "loss": 10.2123,
+      "step": 20540
+    },
+    {
+      "epoch": 0.10262428525056805,
+      "grad_norm": 0.09723500162363052,
+      "learning_rate": 2.6988660542191294e-05,
+      "loss": 10.2132,
+      "step": 20550
+    },
+    {
+      "epoch": 0.1026742240755075,
+      "grad_norm": 0.09278090298175812,
+      "learning_rate": 2.6987158627249744e-05,
+      "loss": 10.2107,
+      "step": 20560
+    },
+    {
+      "epoch": 0.10272416290044695,
+      "grad_norm": 0.09172572195529938,
+      "learning_rate": 2.6985656712308194e-05,
+      "loss": 10.2093,
+      "step": 20570
+    },
+    {
+      "epoch": 0.1027741017253864,
+      "grad_norm": 0.09108912199735641,
+      "learning_rate": 2.6984154797366644e-05,
+      "loss": 10.2067,
+      "step": 20580
+    },
+    {
+      "epoch": 0.10282404055032585,
+      "grad_norm": 0.09847711771726608,
+      "learning_rate": 2.6982652882425094e-05,
+      "loss": 10.211,
+      "step": 20590
+    },
+    {
+      "epoch": 0.1028739793752653,
+      "grad_norm": 0.08993741869926453,
+      "learning_rate": 2.698115096748354e-05,
+      "loss": 10.2101,
+      "step": 20600
+    },
+    {
+      "epoch": 0.10292391820020474,
+      "grad_norm": 0.0974576473236084,
+      "learning_rate": 2.697964905254199e-05,
+      "loss": 10.2092,
+      "step": 20610
+    },
+    {
+      "epoch": 0.1029738570251442,
+      "grad_norm": 0.09782136976718903,
+      "learning_rate": 2.697814713760044e-05,
+      "loss": 10.206,
+      "step": 20620
+    },
+    {
+      "epoch": 0.10302379585008364,
+      "grad_norm": 0.09769613295793533,
+      "learning_rate": 2.6976645222658892e-05,
+      "loss": 10.2108,
+      "step": 20630
+    },
+    {
+      "epoch": 0.1030737346750231,
+      "grad_norm": 0.09461203962564468,
+      "learning_rate": 2.6975143307717342e-05,
+      "loss": 10.2053,
+      "step": 20640
+    },
+    {
+      "epoch": 0.10312367349996254,
+      "grad_norm": 0.09983788430690765,
+      "learning_rate": 2.697364139277579e-05,
+      "loss": 10.2081,
+      "step": 20650
+    },
+    {
+      "epoch": 0.10317361232490199,
+      "grad_norm": 0.09139873832464218,
+      "learning_rate": 2.697213947783424e-05,
+      "loss": 10.2104,
+      "step": 20660
+    },
+    {
+      "epoch": 0.10322355114984144,
+      "grad_norm": 0.10040886700153351,
+      "learning_rate": 2.697063756289269e-05,
+      "loss": 10.2067,
+      "step": 20670
+    },
+    {
+      "epoch": 0.10327348997478089,
+      "grad_norm": 0.0953102856874466,
+      "learning_rate": 2.696913564795114e-05,
+      "loss": 10.2053,
+      "step": 20680
+    },
+    {
+      "epoch": 0.10332342879972034,
+      "grad_norm": 0.0996142104268074,
+      "learning_rate": 2.696763373300959e-05,
+      "loss": 10.2083,
+      "step": 20690
+    },
+    {
+      "epoch": 0.10337336762465979,
+      "grad_norm": 0.09398601204156876,
+      "learning_rate": 2.6966131818068036e-05,
+      "loss": 10.2055,
+      "step": 20700
+    },
+    {
+      "epoch": 0.10342330644959924,
+      "grad_norm": 0.09323547780513763,
+      "learning_rate": 2.6964629903126486e-05,
+      "loss": 10.2101,
+      "step": 20710
+    },
+    {
+      "epoch": 0.10347324527453869,
+      "grad_norm": 0.09476854652166367,
+      "learning_rate": 2.6963127988184937e-05,
+      "loss": 10.2063,
+      "step": 20720
+    },
+    {
+      "epoch": 0.10352318409947814,
+      "grad_norm": 0.08883500099182129,
+      "learning_rate": 2.6961626073243387e-05,
+      "loss": 10.2072,
+      "step": 20730
+    },
+    {
+      "epoch": 0.10357312292441759,
+      "grad_norm": 0.09644431620836258,
+      "learning_rate": 2.6960124158301837e-05,
+      "loss": 10.2066,
+      "step": 20740
+    },
+    {
+      "epoch": 0.10362306174935704,
+      "grad_norm": 0.09586118906736374,
+      "learning_rate": 2.6958622243360284e-05,
+      "loss": 10.2093,
+      "step": 20750
+    },
+    {
+      "epoch": 0.10367300057429649,
+      "grad_norm": 0.09478766471147537,
+      "learning_rate": 2.6957120328418734e-05,
+      "loss": 10.2054,
+      "step": 20760
+    },
+    {
+      "epoch": 0.10372293939923594,
+      "grad_norm": 0.09579043835401535,
+      "learning_rate": 2.6955618413477184e-05,
+      "loss": 10.207,
+      "step": 20770
+    },
+    {
+      "epoch": 0.10377287822417539,
+      "grad_norm": 0.08978202193975449,
+      "learning_rate": 2.6954116498535634e-05,
+      "loss": 10.2019,
+      "step": 20780
+    },
+    {
+      "epoch": 0.10382281704911483,
+      "grad_norm": 0.09112431108951569,
+      "learning_rate": 2.6952614583594084e-05,
+      "loss": 10.2072,
+      "step": 20790
+    },
+    {
+      "epoch": 0.10387275587405428,
+      "grad_norm": 0.09767849743366241,
+      "learning_rate": 2.695111266865253e-05,
+      "loss": 10.2056,
+      "step": 20800
+    },
+    {
+      "epoch": 0.10392269469899373,
+      "grad_norm": 0.09825193881988525,
+      "learning_rate": 2.694961075371098e-05,
+      "loss": 10.2068,
+      "step": 20810
+    },
+    {
+      "epoch": 0.10397263352393318,
+      "grad_norm": 0.09760141372680664,
+      "learning_rate": 2.694810883876943e-05,
+      "loss": 10.2078,
+      "step": 20820
+    },
+    {
+      "epoch": 0.10402257234887263,
+      "grad_norm": 0.09558884054422379,
+      "learning_rate": 2.6946606923827882e-05,
+      "loss": 10.2056,
+      "step": 20830
+    },
+    {
+      "epoch": 0.10407251117381208,
+      "grad_norm": 0.09403236955404282,
+      "learning_rate": 2.6945105008886332e-05,
+      "loss": 10.2005,
+      "step": 20840
+    },
+    {
+      "epoch": 0.10412244999875153,
+      "grad_norm": 0.09680522978305817,
+      "learning_rate": 2.6943603093944782e-05,
+      "loss": 10.2052,
+      "step": 20850
+    },
+    {
+      "epoch": 0.10417238882369098,
+      "grad_norm": 0.09517577290534973,
+      "learning_rate": 2.694210117900323e-05,
+      "loss": 10.2047,
+      "step": 20860
+    },
+    {
+      "epoch": 0.10422232764863043,
+      "grad_norm": 0.09840694069862366,
+      "learning_rate": 2.694059926406168e-05,
+      "loss": 10.2048,
+      "step": 20870
+    },
+    {
+      "epoch": 0.10427226647356988,
+      "grad_norm": 0.09037858247756958,
+      "learning_rate": 2.693909734912013e-05,
+      "loss": 10.1986,
+      "step": 20880
+    },
+    {
+      "epoch": 0.10432220529850933,
+      "grad_norm": 0.09096881747245789,
+      "learning_rate": 2.693759543417858e-05,
+      "loss": 10.204,
+      "step": 20890
+    },
+    {
+      "epoch": 0.10437214412344878,
+      "grad_norm": 0.0983474999666214,
+      "learning_rate": 2.693609351923703e-05,
+      "loss": 10.2116,
+      "step": 20900
+    },
+    {
+      "epoch": 0.10442208294838823,
+      "grad_norm": 0.08912882208824158,
+      "learning_rate": 2.6934591604295476e-05,
+      "loss": 10.2036,
+      "step": 20910
+    },
+    {
+      "epoch": 0.10447202177332768,
+      "grad_norm": 122905.046875,
+      "learning_rate": 2.6933089689353927e-05,
+      "loss": 11.4131,
+      "step": 20920
+    },
+    {
+      "epoch": 0.10452196059826713,
+      "grad_norm": 0.09772419929504395,
+      "learning_rate": 2.6931587774412377e-05,
+      "loss": 38.3405,
+      "step": 20930
+    },
+    {
+      "epoch": 0.10457189942320658,
+      "grad_norm": 0.09270507097244263,
+      "learning_rate": 2.6930085859470827e-05,
+      "loss": 10.579,
+      "step": 20940
+    },
+    {
+      "epoch": 0.10462183824814603,
+      "grad_norm": 0.10307878255844116,
+      "learning_rate": 2.6928583944529277e-05,
+      "loss": 10.1982,
+      "step": 20950
+    },
+    {
+      "epoch": 0.10467177707308548,
+      "grad_norm": 0.10254894942045212,
+      "learning_rate": 2.6927082029587724e-05,
+      "loss": 10.2039,
+      "step": 20960
+    },
+    {
+      "epoch": 0.10472171589802493,
+      "grad_norm": 0.0985196977853775,
+      "learning_rate": 2.6925580114646174e-05,
+      "loss": 10.2041,
+      "step": 20970
+    },
+    {
+      "epoch": 0.10477165472296437,
+      "grad_norm": 0.0948171392083168,
+      "learning_rate": 2.6924078199704624e-05,
+      "loss": 10.2012,
+      "step": 20980
+    },
+    {
+      "epoch": 0.10482159354790382,
+      "grad_norm": 0.09431871771812439,
+      "learning_rate": 2.6922576284763075e-05,
+      "loss": 10.2013,
+      "step": 20990
+    },
+    {
+      "epoch": 0.10487153237284327,
+      "grad_norm": 0.09847448766231537,
+      "learning_rate": 2.6921074369821525e-05,
+      "loss": 10.1986,
+      "step": 21000
+    },
+    {
+      "epoch": 0.10492147119778272,
+      "grad_norm": 0.09803617000579834,
+      "learning_rate": 2.691957245487997e-05,
+      "loss": 10.1999,
+      "step": 21010
+    },
+    {
+      "epoch": 0.10497141002272216,
+      "grad_norm": 0.09164676815271378,
+      "learning_rate": 2.691807053993842e-05,
+      "loss": 10.1964,
+      "step": 21020
+    },
+    {
+      "epoch": 0.10502134884766161,
+      "grad_norm": 0.09696748852729797,
+      "learning_rate": 2.6916568624996872e-05,
+      "loss": 10.1992,
+      "step": 21030
+    },
+    {
+      "epoch": 0.10507128767260106,
+      "grad_norm": 0.090817391872406,
+      "learning_rate": 2.6915066710055322e-05,
+      "loss": 10.199,
+      "step": 21040
+    },
+    {
+      "epoch": 0.10512122649754051,
+      "grad_norm": 0.09760018438100815,
+      "learning_rate": 2.6913564795113772e-05,
+      "loss": 10.2002,
+      "step": 21050
+    },
+    {
+      "epoch": 0.10517116532247996,
+      "grad_norm": 0.09421108663082123,
+      "learning_rate": 2.691206288017222e-05,
+      "loss": 10.1982,
+      "step": 21060
+    },
+    {
+      "epoch": 0.1052211041474194,
+      "grad_norm": 0.09282322973012924,
+      "learning_rate": 2.691056096523067e-05,
+      "loss": 10.1998,
+      "step": 21070
+    },
+    {
+      "epoch": 0.10527104297235886,
+      "grad_norm": 0.09494070708751678,
+      "learning_rate": 2.690905905028912e-05,
+      "loss": 10.1958,
+      "step": 21080
+    },
+    {
+      "epoch": 0.1053209817972983,
+      "grad_norm": 0.0971512645483017,
+      "learning_rate": 2.690755713534757e-05,
+      "loss": 10.2041,
+      "step": 21090
+    },
+    {
+      "epoch": 0.10537092062223775,
+      "grad_norm": 0.09018143266439438,
+      "learning_rate": 2.690605522040602e-05,
+      "loss": 10.1996,
+      "step": 21100
+    },
+    {
+      "epoch": 0.1054208594471772,
+      "grad_norm": 0.09023455530405045,
+      "learning_rate": 2.6904553305464466e-05,
+      "loss": 10.2006,
+      "step": 21110
+    },
+    {
+      "epoch": 0.10547079827211665,
+      "grad_norm": 0.09394627809524536,
+      "learning_rate": 2.6903051390522917e-05,
+      "loss": 10.1966,
+      "step": 21120
+    },
+    {
+      "epoch": 0.1055207370970561,
+      "grad_norm": 0.09698754549026489,
+      "learning_rate": 2.6901549475581367e-05,
+      "loss": 10.1971,
+      "step": 21130
+    },
+    {
+      "epoch": 0.10557067592199555,
+      "grad_norm": 0.09606551378965378,
+      "learning_rate": 2.6900047560639817e-05,
+      "loss": 10.1975,
+      "step": 21140
+    },
+    {
+      "epoch": 0.105620614746935,
+      "grad_norm": 0.0962950810790062,
+      "learning_rate": 2.6898545645698267e-05,
+      "loss": 10.194,
+      "step": 21150
+    },
+    {
+      "epoch": 0.10567055357187445,
+      "grad_norm": 0.09697403013706207,
+      "learning_rate": 2.6897043730756714e-05,
+      "loss": 10.1934,
+      "step": 21160
+    },
+    {
+      "epoch": 0.1057204923968139,
+      "grad_norm": 0.09771595895290375,
+      "learning_rate": 2.6895541815815164e-05,
+      "loss": 10.197,
+      "step": 21170
+    },
+    {
+      "epoch": 0.10577043122175335,
+      "grad_norm": 0.09692160040140152,
+      "learning_rate": 2.6894039900873614e-05,
+      "loss": 10.196,
+      "step": 21180
+    },
+    {
+      "epoch": 0.1058203700466928,
+      "grad_norm": 0.09422864764928818,
+      "learning_rate": 2.6892537985932065e-05,
+      "loss": 10.1967,
+      "step": 21190
+    },
+    {
+      "epoch": 0.10587030887163225,
+      "grad_norm": 0.09830430895090103,
+      "learning_rate": 2.6891036070990515e-05,
+      "loss": 10.1954,
+      "step": 21200
+    },
+    {
+      "epoch": 0.1059202476965717,
+      "grad_norm": 0.09837440401315689,
+      "learning_rate": 2.688953415604896e-05,
+      "loss": 10.1955,
+      "step": 21210
+    },
+    {
+      "epoch": 0.10597018652151115,
+      "grad_norm": 0.09582128375768661,
+      "learning_rate": 2.6888032241107415e-05,
+      "loss": 10.1957,
+      "step": 21220
+    },
+    {
+      "epoch": 0.1060201253464506,
+      "grad_norm": 0.09249364584684372,
+      "learning_rate": 2.6886530326165862e-05,
+      "loss": 10.1941,
+      "step": 21230
+    },
+    {
+      "epoch": 0.10607006417139005,
+      "grad_norm": 0.09359084814786911,
+      "learning_rate": 2.6885028411224312e-05,
+      "loss": 10.1938,
+      "step": 21240
+    },
+    {
+      "epoch": 0.1061200029963295,
+      "grad_norm": 0.09493474662303925,
+      "learning_rate": 2.6883526496282762e-05,
+      "loss": 10.1944,
+      "step": 21250
+    },
+    {
+      "epoch": 0.10616994182126895,
+      "grad_norm": 0.09470515698194504,
+      "learning_rate": 2.688202458134121e-05,
+      "loss": 10.1947,
+      "step": 21260
+    },
+    {
+      "epoch": 0.1062198806462084,
+      "grad_norm": 0.09477045387029648,
+      "learning_rate": 2.6880522666399663e-05,
+      "loss": 10.1968,
+      "step": 21270
+    },
+    {
+      "epoch": 0.10626981947114784,
+      "grad_norm": 0.0963263064622879,
+      "learning_rate": 2.687902075145811e-05,
+      "loss": 10.1905,
+      "step": 21280
+    },
+    {
+      "epoch": 0.1063197582960873,
+      "grad_norm": 0.08664928376674652,
+      "learning_rate": 2.687751883651656e-05,
+      "loss": 10.1931,
+      "step": 21290
+    },
+    {
+      "epoch": 0.10636969712102674,
+      "grad_norm": 0.09358591586351395,
+      "learning_rate": 2.687601692157501e-05,
+      "loss": 10.1923,
+      "step": 21300
+    },
+    {
+      "epoch": 0.10641963594596619,
+      "grad_norm": 0.09352956712245941,
+      "learning_rate": 2.6874515006633457e-05,
+      "loss": 10.1936,
+      "step": 21310
+    },
+    {
+      "epoch": 0.10646957477090564,
+      "grad_norm": 0.09512390941381454,
+      "learning_rate": 2.687301309169191e-05,
+      "loss": 10.1936,
+      "step": 21320
+    },
+    {
+      "epoch": 0.10651951359584509,
+      "grad_norm": 0.09855396300554276,
+      "learning_rate": 2.6871511176750357e-05,
+      "loss": 10.1925,
+      "step": 21330
+    },
+    {
+      "epoch": 0.10656945242078454,
+      "grad_norm": 0.09649109095335007,
+      "learning_rate": 2.6870009261808807e-05,
+      "loss": 10.189,
+      "step": 21340
+    },
+    {
+      "epoch": 0.10661939124572399,
+      "grad_norm": 0.09623111039400101,
+      "learning_rate": 2.6868507346867257e-05,
+      "loss": 10.1903,
+      "step": 21350
+    },
+    {
+      "epoch": 0.10666933007066344,
+      "grad_norm": 0.1012498065829277,
+      "learning_rate": 2.6867005431925704e-05,
+      "loss": 10.1914,
+      "step": 21360
+    },
+    {
+      "epoch": 0.10671926889560289,
+      "grad_norm": 0.10435005277395248,
+      "learning_rate": 2.6865503516984158e-05,
+      "loss": 10.1913,
+      "step": 21370
+    },
+    {
+      "epoch": 0.10676920772054234,
+      "grad_norm": 0.09784939885139465,
+      "learning_rate": 2.6864001602042604e-05,
+      "loss": 10.1911,
+      "step": 21380
+    },
+    {
+      "epoch": 0.10681914654548179,
+      "grad_norm": 0.09625697135925293,
+      "learning_rate": 2.6862499687101055e-05,
+      "loss": 10.1886,
+      "step": 21390
+    },
+    {
+      "epoch": 0.10686908537042124,
+      "grad_norm": 0.09613115340471268,
+      "learning_rate": 2.6860997772159505e-05,
+      "loss": 10.1896,
+      "step": 21400
+    },
+    {
+      "epoch": 0.10691902419536069,
+      "grad_norm": 0.0873713418841362,
+      "learning_rate": 2.685949585721795e-05,
+      "loss": 10.188,
+      "step": 21410
+    },
+    {
+      "epoch": 0.10696896302030014,
+      "grad_norm": 0.09823717176914215,
+      "learning_rate": 2.6857993942276405e-05,
+      "loss": 10.1895,
+      "step": 21420
+    },
+    {
+      "epoch": 0.10701890184523959,
+      "grad_norm": 0.09435474872589111,
+      "learning_rate": 2.6856492027334852e-05,
+      "loss": 10.1872,
+      "step": 21430
+    },
+    {
+      "epoch": 0.10706884067017904,
+      "grad_norm": 0.09932689368724823,
+      "learning_rate": 2.6854990112393302e-05,
+      "loss": 10.189,
+      "step": 21440
+    },
+    {
+      "epoch": 0.10711877949511849,
+      "grad_norm": 0.1061321571469307,
+      "learning_rate": 2.6853488197451752e-05,
+      "loss": 10.1871,
+      "step": 21450
+    },
+    {
+      "epoch": 0.10716871832005793,
+      "grad_norm": 0.08677263557910919,
+      "learning_rate": 2.68519862825102e-05,
+      "loss": 10.1935,
+      "step": 21460
+    },
+    {
+      "epoch": 0.10721865714499738,
+      "grad_norm": 0.09986065328121185,
+      "learning_rate": 2.6850484367568653e-05,
+      "loss": 10.1886,
+      "step": 21470
+    },
+    {
+      "epoch": 0.10726859596993683,
+      "grad_norm": 0.09638167917728424,
+      "learning_rate": 2.68489824526271e-05,
+      "loss": 10.1887,
+      "step": 21480
+    },
+    {
+      "epoch": 0.10731853479487628,
+      "grad_norm": 0.09618506580591202,
+      "learning_rate": 2.684748053768555e-05,
+      "loss": 10.1871,
+      "step": 21490
+    },
+    {
+      "epoch": 0.10736847361981573,
+      "grad_norm": 0.09868462383747101,
+      "learning_rate": 2.6845978622744e-05,
+      "loss": 10.1872,
+      "step": 21500
+    },
+    {
+      "epoch": 0.10741841244475517,
+      "grad_norm": 0.08971121162176132,
+      "learning_rate": 2.6844476707802447e-05,
+      "loss": 10.1863,
+      "step": 21510
+    },
+    {
+      "epoch": 0.10746835126969462,
+      "grad_norm": 0.10220789909362793,
+      "learning_rate": 2.68429747928609e-05,
+      "loss": 10.1839,
+      "step": 21520
+    },
+    {
+      "epoch": 0.10751829009463407,
+      "grad_norm": 0.09101521223783493,
+      "learning_rate": 2.6841472877919347e-05,
+      "loss": 10.1825,
+      "step": 21530
+    },
+    {
+      "epoch": 0.10756822891957352,
+      "grad_norm": 0.09986945986747742,
+      "learning_rate": 2.68399709629778e-05,
+      "loss": 10.1873,
+      "step": 21540
+    },
+    {
+      "epoch": 0.10761816774451297,
+      "grad_norm": 0.09732682257890701,
+      "learning_rate": 2.6838469048036247e-05,
+      "loss": 10.1841,
+      "step": 21550
+    },
+    {
+      "epoch": 0.10766810656945242,
+      "grad_norm": 0.09686817973852158,
+      "learning_rate": 2.6836967133094694e-05,
+      "loss": 10.1836,
+      "step": 21560
+    },
+    {
+      "epoch": 0.10771804539439186,
+      "grad_norm": 0.09880684316158295,
+      "learning_rate": 2.6835465218153148e-05,
+      "loss": 10.1874,
+      "step": 21570
+    },
+    {
+      "epoch": 0.10776798421933131,
+      "grad_norm": 0.09467620402574539,
+      "learning_rate": 2.6833963303211594e-05,
+      "loss": 10.1826,
+      "step": 21580
+    },
+    {
+      "epoch": 0.10781792304427076,
+      "grad_norm": 0.0975777879357338,
+      "learning_rate": 2.6832461388270048e-05,
+      "loss": 10.1839,
+      "step": 21590
+    },
+    {
+      "epoch": 0.10786786186921021,
+      "grad_norm": 0.09700115025043488,
+      "learning_rate": 2.6830959473328495e-05,
+      "loss": 10.1848,
+      "step": 21600
+    },
+    {
+      "epoch": 0.10791780069414966,
+      "grad_norm": 0.0938863530755043,
+      "learning_rate": 2.682945755838694e-05,
+      "loss": 10.1856,
+      "step": 21610
+    },
+    {
+      "epoch": 0.10796773951908911,
+      "grad_norm": 0.09892918914556503,
+      "learning_rate": 2.6827955643445395e-05,
+      "loss": 10.1825,
+      "step": 21620
+    },
+    {
+      "epoch": 0.10801767834402856,
+      "grad_norm": 0.08988924324512482,
+      "learning_rate": 2.6826453728503842e-05,
+      "loss": 10.1828,
+      "step": 21630
+    },
+    {
+      "epoch": 0.10806761716896801,
+      "grad_norm": 0.09243562072515488,
+      "learning_rate": 2.6824951813562295e-05,
+      "loss": 10.1822,
+      "step": 21640
+    },
+    {
+      "epoch": 0.10811755599390746,
+      "grad_norm": 0.09508119523525238,
+      "learning_rate": 2.6823449898620742e-05,
+      "loss": 10.1855,
+      "step": 21650
+    },
+    {
+      "epoch": 0.10816749481884691,
+      "grad_norm": 0.09622988849878311,
+      "learning_rate": 2.682194798367919e-05,
+      "loss": 10.1833,
+      "step": 21660
+    },
+    {
+      "epoch": 0.10821743364378636,
+      "grad_norm": 0.09975172579288483,
+      "learning_rate": 2.6820446068737643e-05,
+      "loss": 10.1856,
+      "step": 21670
+    },
+    {
+      "epoch": 0.10826737246872581,
+      "grad_norm": 0.10020721703767776,
+      "learning_rate": 2.681894415379609e-05,
+      "loss": 10.1846,
+      "step": 21680
+    },
+    {
+      "epoch": 0.10831731129366526,
+      "grad_norm": 0.091154083609581,
+      "learning_rate": 2.6817442238854543e-05,
+      "loss": 10.1915,
+      "step": 21690
+    },
+    {
+      "epoch": 0.10836725011860471,
+      "grad_norm": 0.09663428366184235,
+      "learning_rate": 2.681594032391299e-05,
+      "loss": 10.1844,
+      "step": 21700
+    },
+    {
+      "epoch": 0.10841718894354416,
+      "grad_norm": 0.09310304373502731,
+      "learning_rate": 2.6814438408971437e-05,
+      "loss": 10.1833,
+      "step": 21710
+    },
+    {
+      "epoch": 0.1084671277684836,
+      "grad_norm": 0.09270327538251877,
+      "learning_rate": 2.681293649402989e-05,
+      "loss": 10.1821,
+      "step": 21720
+    },
+    {
+      "epoch": 0.10851706659342306,
+      "grad_norm": 0.09398692846298218,
+      "learning_rate": 2.6811434579088337e-05,
+      "loss": 10.1829,
+      "step": 21730
+    },
+    {
+      "epoch": 0.1085670054183625,
+      "grad_norm": 0.10346375405788422,
+      "learning_rate": 2.680993266414679e-05,
+      "loss": 10.1817,
+      "step": 21740
+    },
+    {
+      "epoch": 0.10861694424330195,
+      "grad_norm": 0.09369218349456787,
+      "learning_rate": 2.6808430749205237e-05,
+      "loss": 10.1809,
+      "step": 21750
+    },
+    {
+      "epoch": 0.1086668830682414,
+      "grad_norm": 0.09752266108989716,
+      "learning_rate": 2.6806928834263684e-05,
+      "loss": 10.1832,
+      "step": 21760
+    },
+    {
+      "epoch": 0.10871682189318085,
+      "grad_norm": 0.09794414043426514,
+      "learning_rate": 2.6805426919322138e-05,
+      "loss": 10.1814,
+      "step": 21770
+    },
+    {
+      "epoch": 0.1087667607181203,
+      "grad_norm": 0.09713537245988846,
+      "learning_rate": 2.6803925004380584e-05,
+      "loss": 10.1832,
+      "step": 21780
+    },
+    {
+      "epoch": 0.10881669954305975,
+      "grad_norm": 0.09311962127685547,
+      "learning_rate": 2.6802423089439038e-05,
+      "loss": 10.1797,
+      "step": 21790
+    },
+    {
+      "epoch": 0.1088666383679992,
+      "grad_norm": 0.095693439245224,
+      "learning_rate": 2.6800921174497485e-05,
+      "loss": 10.1804,
+      "step": 21800
+    },
+    {
+      "epoch": 0.10891657719293865,
+      "grad_norm": 0.0987323671579361,
+      "learning_rate": 2.679941925955593e-05,
+      "loss": 10.1776,
+      "step": 21810
+    },
+    {
+      "epoch": 0.1089665160178781,
+      "grad_norm": 0.09942522644996643,
+      "learning_rate": 2.6797917344614385e-05,
+      "loss": 10.1785,
+      "step": 21820
+    },
+    {
+      "epoch": 0.10901645484281755,
+      "grad_norm": 0.09082325547933578,
+      "learning_rate": 2.6796415429672832e-05,
+      "loss": 10.1772,
+      "step": 21830
+    },
+    {
+      "epoch": 0.109066393667757,
+      "grad_norm": 0.08826711773872375,
+      "learning_rate": 2.6794913514731285e-05,
+      "loss": 10.1822,
+      "step": 21840
+    },
+    {
+      "epoch": 0.10911633249269645,
+      "grad_norm": 0.09767315536737442,
+      "learning_rate": 2.6793411599789732e-05,
+      "loss": 10.1764,
+      "step": 21850
+    },
+    {
+      "epoch": 0.1091662713176359,
+      "grad_norm": 0.0958879366517067,
+      "learning_rate": 2.6791909684848182e-05,
+      "loss": 10.1805,
+      "step": 21860
+    },
+    {
+      "epoch": 0.10921621014257535,
+      "grad_norm": 0.09533664584159851,
+      "learning_rate": 2.6790407769906633e-05,
+      "loss": 10.1824,
+      "step": 21870
+    },
+    {
+      "epoch": 0.1092661489675148,
+      "grad_norm": 0.09033321589231491,
+      "learning_rate": 2.678890585496508e-05,
+      "loss": 10.1783,
+      "step": 21880
+    },
+    {
+      "epoch": 0.10931608779245425,
+      "grad_norm": 0.09859579056501389,
+      "learning_rate": 2.6787403940023533e-05,
+      "loss": 10.1763,
+      "step": 21890
+    },
+    {
+      "epoch": 0.1093660266173937,
+      "grad_norm": 0.09806695580482483,
+      "learning_rate": 2.678590202508198e-05,
+      "loss": 10.1808,
+      "step": 21900
+    },
+    {
+      "epoch": 0.10941596544233315,
+      "grad_norm": 0.09141387790441513,
+      "learning_rate": 2.678440011014043e-05,
+      "loss": 10.1787,
+      "step": 21910
+    },
+    {
+      "epoch": 0.1094659042672726,
+      "grad_norm": 0.09609120339155197,
+      "learning_rate": 2.678289819519888e-05,
+      "loss": 10.1731,
+      "step": 21920
+    },
+    {
+      "epoch": 0.10951584309221205,
+      "grad_norm": 0.09060762822628021,
+      "learning_rate": 2.6781396280257327e-05,
+      "loss": 10.1776,
+      "step": 21930
+    },
+    {
+      "epoch": 0.1095657819171515,
+      "grad_norm": 0.09392700344324112,
+      "learning_rate": 2.677989436531578e-05,
+      "loss": 10.1774,
+      "step": 21940
+    },
+    {
+      "epoch": 0.10961572074209094,
+      "grad_norm": 0.09288621693849564,
+      "learning_rate": 2.6778392450374227e-05,
+      "loss": 10.1799,
+      "step": 21950
+    },
+    {
+      "epoch": 0.1096656595670304,
+      "grad_norm": 0.09960731863975525,
+      "learning_rate": 2.6776890535432677e-05,
+      "loss": 10.1738,
+      "step": 21960
+    },
+    {
+      "epoch": 0.10971559839196984,
+      "grad_norm": 0.09052469581365585,
+      "learning_rate": 2.6775388620491128e-05,
+      "loss": 10.1785,
+      "step": 21970
+    },
+    {
+      "epoch": 0.10976553721690929,
+      "grad_norm": 0.09761349111795425,
+      "learning_rate": 2.6773886705549574e-05,
+      "loss": 10.1752,
+      "step": 21980
+    },
+    {
+      "epoch": 0.10981547604184874,
+      "grad_norm": 0.09489648789167404,
+      "learning_rate": 2.6772384790608028e-05,
+      "loss": 10.1732,
+      "step": 21990
+    },
+    {
+      "epoch": 0.10986541486678819,
+      "grad_norm": 0.0943918228149414,
+      "learning_rate": 2.6770882875666475e-05,
+      "loss": 10.1755,
+      "step": 22000
+    },
+    {
+      "epoch": 0.10991535369172763,
+      "grad_norm": 0.09196553379297256,
+      "learning_rate": 2.6769380960724925e-05,
+      "loss": 10.1725,
+      "step": 22010
+    },
+    {
+      "epoch": 0.10996529251666708,
+      "grad_norm": 0.09334488213062286,
+      "learning_rate": 2.6767879045783375e-05,
+      "loss": 10.1756,
+      "step": 22020
+    },
+    {
+      "epoch": 0.11001523134160653,
+      "grad_norm": 0.10055167973041534,
+      "learning_rate": 2.6766377130841822e-05,
+      "loss": 10.1705,
+      "step": 22030
+    },
+    {
+      "epoch": 0.11006517016654598,
+      "grad_norm": 0.09944339841604233,
+      "learning_rate": 2.6764875215900275e-05,
+      "loss": 10.1801,
+      "step": 22040
+    },
+    {
+      "epoch": 0.11011510899148542,
+      "grad_norm": 0.10031431168317795,
+      "learning_rate": 2.6763373300958722e-05,
+      "loss": 10.1704,
+      "step": 22050
+    },
+    {
+      "epoch": 0.11016504781642487,
+      "grad_norm": 0.09707958251237869,
+      "learning_rate": 2.6761871386017172e-05,
+      "loss": 10.1752,
+      "step": 22060
+    },
+    {
+      "epoch": 0.11021498664136432,
+      "grad_norm": 0.09501238167285919,
+      "learning_rate": 2.6760369471075623e-05,
+      "loss": 10.1698,
+      "step": 22070
+    },
+    {
+      "epoch": 0.11026492546630377,
+      "grad_norm": 0.09578476846218109,
+      "learning_rate": 2.675886755613407e-05,
+      "loss": 10.175,
+      "step": 22080
+    },
+    {
+      "epoch": 0.11031486429124322,
+      "grad_norm": 0.09270567446947098,
+      "learning_rate": 2.6757365641192523e-05,
+      "loss": 10.1732,
+      "step": 22090
+    },
+    {
+      "epoch": 0.11036480311618267,
+      "grad_norm": 0.09588302671909332,
+      "learning_rate": 2.675586372625097e-05,
+      "loss": 10.1726,
+      "step": 22100
+    },
+    {
+      "epoch": 0.11041474194112212,
+      "grad_norm": 0.09637803584337234,
+      "learning_rate": 2.675436181130942e-05,
+      "loss": 10.1763,
+      "step": 22110
+    },
+    {
+      "epoch": 0.11046468076606157,
+      "grad_norm": 0.09648361802101135,
+      "learning_rate": 2.675285989636787e-05,
+      "loss": 10.1731,
+      "step": 22120
+    },
+    {
+      "epoch": 0.11051461959100102,
+      "grad_norm": 0.0957055613398552,
+      "learning_rate": 2.6751357981426317e-05,
+      "loss": 10.1696,
+      "step": 22130
+    },
+    {
+      "epoch": 0.11056455841594047,
+      "grad_norm": 0.09542543441057205,
+      "learning_rate": 2.674985606648477e-05,
+      "loss": 10.1725,
+      "step": 22140
+    },
+    {
+      "epoch": 0.11061449724087992,
+      "grad_norm": 0.09192683547735214,
+      "learning_rate": 2.6748354151543217e-05,
+      "loss": 10.1725,
+      "step": 22150
+    },
+    {
+      "epoch": 0.11066443606581937,
+      "grad_norm": 0.09353506565093994,
+      "learning_rate": 2.6746852236601667e-05,
+      "loss": 10.1663,
+      "step": 22160
+    },
+    {
+      "epoch": 0.11071437489075882,
+      "grad_norm": 0.09622788429260254,
+      "learning_rate": 2.6745350321660118e-05,
+      "loss": 10.1658,
+      "step": 22170
+    },
+    {
+      "epoch": 0.11076431371569827,
+      "grad_norm": 0.0987735167145729,
+      "learning_rate": 2.6743848406718568e-05,
+      "loss": 10.1678,
+      "step": 22180
+    },
+    {
+      "epoch": 0.11081425254063772,
+      "grad_norm": 0.09520336985588074,
+      "learning_rate": 2.6742346491777018e-05,
+      "loss": 10.1679,
+      "step": 22190
+    },
+    {
+      "epoch": 0.11086419136557717,
+      "grad_norm": 0.0966307744383812,
+      "learning_rate": 2.6740844576835465e-05,
+      "loss": 10.1683,
+      "step": 22200
+    },
+    {
+      "epoch": 0.11091413019051662,
+      "grad_norm": 0.10376681387424469,
+      "learning_rate": 2.6739342661893915e-05,
+      "loss": 10.1714,
+      "step": 22210
+    },
+    {
+      "epoch": 0.11096406901545607,
+      "grad_norm": 0.09316543489694595,
+      "learning_rate": 2.6737840746952365e-05,
+      "loss": 10.169,
+      "step": 22220
+    },
+    {
+      "epoch": 0.11101400784039551,
+      "grad_norm": 0.09309463202953339,
+      "learning_rate": 2.6736338832010815e-05,
+      "loss": 10.1702,
+      "step": 22230
+    },
+    {
+      "epoch": 0.11106394666533496,
+      "grad_norm": 0.09572753310203552,
+      "learning_rate": 2.6734836917069265e-05,
+      "loss": 10.1718,
+      "step": 22240
+    },
+    {
+      "epoch": 0.11111388549027441,
+      "grad_norm": 0.0948120579123497,
+      "learning_rate": 2.6733335002127712e-05,
+      "loss": 10.1707,
+      "step": 22250
+    },
+    {
+      "epoch": 0.11116382431521386,
+      "grad_norm": 0.09530732035636902,
+      "learning_rate": 2.6731833087186162e-05,
+      "loss": 10.1678,
+      "step": 22260
+    },
+    {
+      "epoch": 0.11121376314015331,
+      "grad_norm": 0.10655371099710464,
+      "learning_rate": 2.6730331172244613e-05,
+      "loss": 10.1669,
+      "step": 22270
+    },
+    {
+      "epoch": 0.11126370196509276,
+      "grad_norm": 0.09528721123933792,
+      "learning_rate": 2.6728829257303063e-05,
+      "loss": 10.1656,
+      "step": 22280
+    },
+    {
+      "epoch": 0.11131364079003221,
+      "grad_norm": 0.09272367507219315,
+      "learning_rate": 2.6727327342361513e-05,
+      "loss": 10.1669,
+      "step": 22290
+    },
+    {
+      "epoch": 0.11136357961497166,
+      "grad_norm": 0.0940340906381607,
+      "learning_rate": 2.672582542741996e-05,
+      "loss": 10.1666,
+      "step": 22300
+    },
+    {
+      "epoch": 0.11141351843991111,
+      "grad_norm": 0.09455035626888275,
+      "learning_rate": 2.672432351247841e-05,
+      "loss": 10.1686,
+      "step": 22310
+    },
+    {
+      "epoch": 0.11146345726485056,
+      "grad_norm": 0.09895509481430054,
+      "learning_rate": 2.672282159753686e-05,
+      "loss": 10.1666,
+      "step": 22320
+    },
+    {
+      "epoch": 0.11151339608979001,
+      "grad_norm": 0.09388074278831482,
+      "learning_rate": 2.672131968259531e-05,
+      "loss": 10.1655,
+      "step": 22330
+    },
+    {
+      "epoch": 0.11156333491472946,
+      "grad_norm": 0.09425579011440277,
+      "learning_rate": 2.671981776765376e-05,
+      "loss": 10.1626,
+      "step": 22340
+    },
+    {
+      "epoch": 0.11161327373966891,
+      "grad_norm": 0.09784049540758133,
+      "learning_rate": 2.6718315852712207e-05,
+      "loss": 10.1689,
+      "step": 22350
+    },
+    {
+      "epoch": 0.11166321256460836,
+      "grad_norm": 0.09686186909675598,
+      "learning_rate": 2.6716813937770657e-05,
+      "loss": 10.1677,
+      "step": 22360
+    },
+    {
+      "epoch": 0.11171315138954781,
+      "grad_norm": 0.09419489651918411,
+      "learning_rate": 2.6715312022829108e-05,
+      "loss": 10.1638,
+      "step": 22370
+    },
+    {
+      "epoch": 0.11176309021448726,
+      "grad_norm": 0.09300322085618973,
+      "learning_rate": 2.6713810107887558e-05,
+      "loss": 10.1697,
+      "step": 22380
+    },
+    {
+      "epoch": 0.1118130290394267,
+      "grad_norm": 0.09139790385961533,
+      "learning_rate": 2.6712308192946008e-05,
+      "loss": 10.165,
+      "step": 22390
+    },
+    {
+      "epoch": 0.11186296786436616,
+      "grad_norm": 0.09501411765813828,
+      "learning_rate": 2.6710806278004455e-05,
+      "loss": 10.1646,
+      "step": 22400
+    },
+    {
+      "epoch": 0.1119129066893056,
+      "grad_norm": 0.09894577413797379,
+      "learning_rate": 2.6709304363062905e-05,
+      "loss": 10.1673,
+      "step": 22410
+    },
+    {
+      "epoch": 0.11196284551424505,
+      "grad_norm": 0.09686759859323502,
+      "learning_rate": 2.6707802448121355e-05,
+      "loss": 10.1675,
+      "step": 22420
+    },
+    {
+      "epoch": 0.1120127843391845,
+      "grad_norm": 0.09430637210607529,
+      "learning_rate": 2.6706300533179805e-05,
+      "loss": 10.165,
+      "step": 22430
+    },
+    {
+      "epoch": 0.11206272316412395,
+      "grad_norm": 0.09339676797389984,
+      "learning_rate": 2.6704798618238256e-05,
+      "loss": 10.1635,
+      "step": 22440
+    },
+    {
+      "epoch": 0.1121126619890634,
+      "grad_norm": 0.09374126046895981,
+      "learning_rate": 2.6703296703296702e-05,
+      "loss": 10.1578,
+      "step": 22450
+    },
+    {
+      "epoch": 0.11216260081400285,
+      "grad_norm": 0.09082947671413422,
+      "learning_rate": 2.6701794788355152e-05,
+      "loss": 10.1673,
+      "step": 22460
+    },
+    {
+      "epoch": 0.1122125396389423,
+      "grad_norm": 0.0888797864317894,
+      "learning_rate": 2.6700292873413603e-05,
+      "loss": 10.1637,
+      "step": 22470
+    },
+    {
+      "epoch": 0.11226247846388175,
+      "grad_norm": 0.09704219549894333,
+      "learning_rate": 2.6698790958472053e-05,
+      "loss": 10.1671,
+      "step": 22480
+    },
+    {
+      "epoch": 0.1123124172888212,
+      "grad_norm": 0.09523064643144608,
+      "learning_rate": 2.6697289043530503e-05,
+      "loss": 10.1643,
+      "step": 22490
+    },
+    {
+      "epoch": 0.11236235611376064,
+      "grad_norm": 0.09382279962301254,
+      "learning_rate": 2.6695787128588953e-05,
+      "loss": 10.1594,
+      "step": 22500
+    },
+    {
+      "epoch": 0.11241229493870009,
+      "grad_norm": 0.09759511053562164,
+      "learning_rate": 2.66942852136474e-05,
+      "loss": 10.1612,
+      "step": 22510
+    },
+    {
+      "epoch": 0.11246223376363954,
+      "grad_norm": 0.09149253368377686,
+      "learning_rate": 2.669278329870585e-05,
+      "loss": 10.1641,
+      "step": 22520
+    },
+    {
+      "epoch": 0.11251217258857898,
+      "grad_norm": 0.09213311225175858,
+      "learning_rate": 2.66912813837643e-05,
+      "loss": 10.1583,
+      "step": 22530
+    },
+    {
+      "epoch": 0.11256211141351843,
+      "grad_norm": 0.09670817106962204,
+      "learning_rate": 2.668977946882275e-05,
+      "loss": 10.1616,
+      "step": 22540
+    },
+    {
+      "epoch": 0.11261205023845788,
+      "grad_norm": 0.09670945256948471,
+      "learning_rate": 2.66882775538812e-05,
+      "loss": 10.1624,
+      "step": 22550
+    },
+    {
+      "epoch": 0.11266198906339733,
+      "grad_norm": 0.09598565846681595,
+      "learning_rate": 2.6686775638939647e-05,
+      "loss": 10.1624,
+      "step": 22560
+    },
+    {
+      "epoch": 0.11271192788833678,
+      "grad_norm": 0.0956323966383934,
+      "learning_rate": 2.6685273723998098e-05,
+      "loss": 10.1591,
+      "step": 22570
+    },
+    {
+      "epoch": 0.11276186671327623,
+      "grad_norm": 0.09476076811552048,
+      "learning_rate": 2.6683771809056548e-05,
+      "loss": 10.1595,
+      "step": 22580
+    },
+    {
+      "epoch": 0.11281180553821568,
+      "grad_norm": 0.09837251901626587,
+      "learning_rate": 2.6682269894114998e-05,
+      "loss": 10.1638,
+      "step": 22590
+    },
+    {
+      "epoch": 0.11286174436315513,
+      "grad_norm": 0.09544532746076584,
+      "learning_rate": 2.6680767979173448e-05,
+      "loss": 10.1615,
+      "step": 22600
+    },
+    {
+      "epoch": 0.11291168318809458,
+      "grad_norm": 0.09193170815706253,
+      "learning_rate": 2.6679266064231895e-05,
+      "loss": 10.162,
+      "step": 22610
+    },
+    {
+      "epoch": 0.11296162201303403,
+      "grad_norm": 0.09794174134731293,
+      "learning_rate": 2.6677764149290345e-05,
+      "loss": 10.1608,
+      "step": 22620
+    },
+    {
+      "epoch": 0.11301156083797348,
+      "grad_norm": 0.08994992822408676,
+      "learning_rate": 2.6676262234348795e-05,
+      "loss": 10.1568,
+      "step": 22630
+    },
+    {
+      "epoch": 0.11306149966291293,
+      "grad_norm": 0.09660772234201431,
+      "learning_rate": 2.6674760319407246e-05,
+      "loss": 10.166,
+      "step": 22640
+    },
+    {
+      "epoch": 0.11311143848785238,
+      "grad_norm": 0.09649979323148727,
+      "learning_rate": 2.6673258404465696e-05,
+      "loss": 10.1625,
+      "step": 22650
+    },
+    {
+      "epoch": 0.11316137731279183,
+      "grad_norm": 0.0917321965098381,
+      "learning_rate": 2.6671756489524142e-05,
+      "loss": 10.1596,
+      "step": 22660
+    },
+    {
+      "epoch": 0.11321131613773128,
+      "grad_norm": 0.0942411795258522,
+      "learning_rate": 2.6670254574582593e-05,
+      "loss": 10.1607,
+      "step": 22670
+    },
+    {
+      "epoch": 0.11326125496267073,
+      "grad_norm": 0.0937599241733551,
+      "learning_rate": 2.6668752659641043e-05,
+      "loss": 10.1558,
+      "step": 22680
+    },
+    {
+      "epoch": 0.11331119378761018,
+      "grad_norm": 0.09375976026058197,
+      "learning_rate": 2.6667250744699493e-05,
+      "loss": 10.1595,
+      "step": 22690
+    },
+    {
+      "epoch": 0.11336113261254963,
+      "grad_norm": 0.09384457021951675,
+      "learning_rate": 2.6665748829757943e-05,
+      "loss": 10.1543,
+      "step": 22700
+    },
+    {
+      "epoch": 0.11341107143748907,
+      "grad_norm": 0.09696808457374573,
+      "learning_rate": 2.666424691481639e-05,
+      "loss": 10.1586,
+      "step": 22710
+    },
+    {
+      "epoch": 0.11346101026242852,
+      "grad_norm": 0.09339691698551178,
+      "learning_rate": 2.666274499987484e-05,
+      "loss": 10.1598,
+      "step": 22720
+    },
+    {
+      "epoch": 0.11351094908736797,
+      "grad_norm": 0.09268683195114136,
+      "learning_rate": 2.666124308493329e-05,
+      "loss": 10.1516,
+      "step": 22730
+    },
+    {
+      "epoch": 0.11356088791230742,
+      "grad_norm": 0.10044291615486145,
+      "learning_rate": 2.665974116999174e-05,
+      "loss": 10.1582,
+      "step": 22740
+    },
+    {
+      "epoch": 0.11361082673724687,
+      "grad_norm": 0.0959634929895401,
+      "learning_rate": 2.665823925505019e-05,
+      "loss": 10.1546,
+      "step": 22750
+    },
+    {
+      "epoch": 0.11366076556218632,
+      "grad_norm": 0.09456854313611984,
+      "learning_rate": 2.6656737340108637e-05,
+      "loss": 10.154,
+      "step": 22760
+    },
+    {
+      "epoch": 0.11371070438712577,
+      "grad_norm": 0.10183733701705933,
+      "learning_rate": 2.6655235425167088e-05,
+      "loss": 10.1546,
+      "step": 22770
+    },
+    {
+      "epoch": 0.11376064321206522,
+      "grad_norm": 0.09504352509975433,
+      "learning_rate": 2.6653733510225538e-05,
+      "loss": 10.1579,
+      "step": 22780
+    },
+    {
+      "epoch": 0.11381058203700467,
+      "grad_norm": 0.09749595075845718,
+      "learning_rate": 2.6652231595283988e-05,
+      "loss": 10.1589,
+      "step": 22790
+    },
+    {
+      "epoch": 0.11386052086194412,
+      "grad_norm": 0.09672117978334427,
+      "learning_rate": 2.6650729680342438e-05,
+      "loss": 10.156,
+      "step": 22800
+    },
+    {
+      "epoch": 0.11391045968688357,
+      "grad_norm": 0.10440251231193542,
+      "learning_rate": 2.6649227765400885e-05,
+      "loss": 10.1534,
+      "step": 22810
+    },
+    {
+      "epoch": 0.11396039851182302,
+      "grad_norm": 0.09660029411315918,
+      "learning_rate": 2.664772585045934e-05,
+      "loss": 10.1507,
+      "step": 22820
+    },
+    {
+      "epoch": 0.11401033733676247,
+      "grad_norm": 0.09805593639612198,
+      "learning_rate": 2.6646223935517785e-05,
+      "loss": 10.1509,
+      "step": 22830
+    },
+    {
+      "epoch": 0.11406027616170192,
+      "grad_norm": 0.09688452631235123,
+      "learning_rate": 2.6644722020576236e-05,
+      "loss": 10.1586,
+      "step": 22840
+    },
+    {
+      "epoch": 0.11411021498664137,
+      "grad_norm": 0.09908785670995712,
+      "learning_rate": 2.6643220105634686e-05,
+      "loss": 10.1529,
+      "step": 22850
+    },
+    {
+      "epoch": 0.11416015381158082,
+      "grad_norm": 0.0970318540930748,
+      "learning_rate": 2.6641718190693133e-05,
+      "loss": 10.15,
+      "step": 22860
+    },
+    {
+      "epoch": 0.11421009263652027,
+      "grad_norm": 0.0938095897436142,
+      "learning_rate": 2.6640216275751586e-05,
+      "loss": 10.1534,
+      "step": 22870
+    },
+    {
+      "epoch": 0.11426003146145972,
+      "grad_norm": 0.09250219166278839,
+      "learning_rate": 2.6638714360810033e-05,
+      "loss": 10.1529,
+      "step": 22880
+    },
+    {
+      "epoch": 0.11430997028639917,
+      "grad_norm": 0.09546291828155518,
+      "learning_rate": 2.6637212445868483e-05,
+      "loss": 10.1523,
+      "step": 22890
+    },
+    {
+      "epoch": 0.11435990911133861,
+      "grad_norm": 0.09982664883136749,
+      "learning_rate": 2.6635710530926933e-05,
+      "loss": 10.1498,
+      "step": 22900
+    },
+    {
+      "epoch": 0.11440984793627806,
+      "grad_norm": 0.0915086641907692,
+      "learning_rate": 2.663420861598538e-05,
+      "loss": 10.1501,
+      "step": 22910
+    },
+    {
+      "epoch": 0.11445978676121751,
+      "grad_norm": 0.09142784029245377,
+      "learning_rate": 2.6632706701043834e-05,
+      "loss": 10.1543,
+      "step": 22920
+    },
+    {
+      "epoch": 0.11450972558615696,
+      "grad_norm": 0.09327565878629684,
+      "learning_rate": 2.663120478610228e-05,
+      "loss": 10.1495,
+      "step": 22930
+    },
+    {
+      "epoch": 0.11455966441109641,
+      "grad_norm": 0.10108178108930588,
+      "learning_rate": 2.662970287116073e-05,
+      "loss": 10.15,
+      "step": 22940
+    },
+    {
+      "epoch": 0.11460960323603586,
+      "grad_norm": 0.10374581068754196,
+      "learning_rate": 2.662820095621918e-05,
+      "loss": 10.1521,
+      "step": 22950
+    },
+    {
+      "epoch": 0.11465954206097531,
+      "grad_norm": 0.09310846030712128,
+      "learning_rate": 2.6626699041277628e-05,
+      "loss": 10.1512,
+      "step": 22960
+    },
+    {
+      "epoch": 0.11470948088591476,
+      "grad_norm": 0.093265101313591,
+      "learning_rate": 2.662519712633608e-05,
+      "loss": 10.1533,
+      "step": 22970
+    },
+    {
+      "epoch": 0.11475941971085421,
+      "grad_norm": 0.09261230379343033,
+      "learning_rate": 2.6623695211394528e-05,
+      "loss": 10.1498,
+      "step": 22980
+    },
+    {
+      "epoch": 0.11480935853579366,
+      "grad_norm": 0.09824974089860916,
+      "learning_rate": 2.6622193296452978e-05,
+      "loss": 10.1488,
+      "step": 22990
+    },
+    {
+      "epoch": 0.1148592973607331,
+      "grad_norm": 0.09499943256378174,
+      "learning_rate": 2.6620691381511428e-05,
+      "loss": 10.1487,
+      "step": 23000
+    },
+    {
+      "epoch": 0.11490923618567254,
+      "grad_norm": 0.09166286885738373,
+      "learning_rate": 2.6619189466569875e-05,
+      "loss": 10.1469,
+      "step": 23010
+    },
+    {
+      "epoch": 0.114959175010612,
+      "grad_norm": 0.09660668671131134,
+      "learning_rate": 2.661768755162833e-05,
+      "loss": 10.1496,
+      "step": 23020
+    },
+    {
+      "epoch": 0.11500911383555144,
+      "grad_norm": 0.09428876638412476,
+      "learning_rate": 2.6616185636686775e-05,
+      "loss": 10.1469,
+      "step": 23030
+    },
+    {
+      "epoch": 0.1150590526604909,
+      "grad_norm": 0.09304223209619522,
+      "learning_rate": 2.6614683721745226e-05,
+      "loss": 10.1458,
+      "step": 23040
+    },
+    {
+      "epoch": 0.11510899148543034,
+      "grad_norm": 0.09473152458667755,
+      "learning_rate": 2.6613181806803676e-05,
+      "loss": 10.1446,
+      "step": 23050
+    },
+    {
+      "epoch": 0.11515893031036979,
+      "grad_norm": 0.09486373513936996,
+      "learning_rate": 2.6611679891862123e-05,
+      "loss": 10.1459,
+      "step": 23060
+    },
+    {
+      "epoch": 0.11520886913530924,
+      "grad_norm": 0.0942765548825264,
+      "learning_rate": 2.6610177976920576e-05,
+      "loss": 10.1487,
+      "step": 23070
+    },
+    {
+      "epoch": 0.11525880796024869,
+      "grad_norm": 0.09221851825714111,
+      "learning_rate": 2.6608676061979023e-05,
+      "loss": 10.1516,
+      "step": 23080
+    },
+    {
+      "epoch": 0.11530874678518814,
+      "grad_norm": 0.09266392886638641,
+      "learning_rate": 2.6607174147037473e-05,
+      "loss": 10.1478,
+      "step": 23090
+    },
+    {
+      "epoch": 0.11535868561012759,
+      "grad_norm": 0.09701654314994812,
+      "learning_rate": 2.6605672232095923e-05,
+      "loss": 10.146,
+      "step": 23100
+    },
+    {
+      "epoch": 0.11540862443506704,
+      "grad_norm": 0.09978927671909332,
+      "learning_rate": 2.660417031715437e-05,
+      "loss": 10.1477,
+      "step": 23110
+    },
+    {
+      "epoch": 0.11545856326000649,
+      "grad_norm": 0.09633700549602509,
+      "learning_rate": 2.6602668402212824e-05,
+      "loss": 10.1477,
+      "step": 23120
+    },
+    {
+      "epoch": 0.11550850208494594,
+      "grad_norm": 0.09246396273374557,
+      "learning_rate": 2.660116648727127e-05,
+      "loss": 10.1476,
+      "step": 23130
+    },
+    {
+      "epoch": 0.11555844090988539,
+      "grad_norm": 0.09979529678821564,
+      "learning_rate": 2.6599664572329724e-05,
+      "loss": 10.1439,
+      "step": 23140
+    },
+    {
+      "epoch": 0.11560837973482484,
+      "grad_norm": 0.09765896201133728,
+      "learning_rate": 2.659816265738817e-05,
+      "loss": 10.1442,
+      "step": 23150
+    },
+    {
+      "epoch": 0.11565831855976429,
+      "grad_norm": 0.09836973249912262,
+      "learning_rate": 2.6596660742446618e-05,
+      "loss": 10.1468,
+      "step": 23160
+    },
+    {
+      "epoch": 0.11570825738470374,
+      "grad_norm": 0.09357340633869171,
+      "learning_rate": 2.659515882750507e-05,
+      "loss": 10.1475,
+      "step": 23170
+    },
+    {
+      "epoch": 0.11575819620964319,
+      "grad_norm": 0.09507403522729874,
+      "learning_rate": 2.6593656912563518e-05,
+      "loss": 10.1419,
+      "step": 23180
+    },
+    {
+      "epoch": 0.11580813503458263,
+      "grad_norm": 0.09896954894065857,
+      "learning_rate": 2.659215499762197e-05,
+      "loss": 10.1431,
+      "step": 23190
+    },
+    {
+      "epoch": 0.11585807385952208,
+      "grad_norm": 0.09166592359542847,
+      "learning_rate": 2.6590653082680418e-05,
+      "loss": 10.144,
+      "step": 23200
+    },
+    {
+      "epoch": 0.11590801268446153,
+      "grad_norm": 0.09727290272712708,
+      "learning_rate": 2.6589151167738865e-05,
+      "loss": 10.1475,
+      "step": 23210
+    },
+    {
+      "epoch": 0.11595795150940098,
+      "grad_norm": 0.09719046205282211,
+      "learning_rate": 2.658764925279732e-05,
+      "loss": 10.1424,
+      "step": 23220
+    },
+    {
+      "epoch": 0.11600789033434043,
+      "grad_norm": 0.09605339914560318,
+      "learning_rate": 2.6586147337855765e-05,
+      "loss": 10.1377,
+      "step": 23230
+    },
+    {
+      "epoch": 0.11605782915927988,
+      "grad_norm": 0.10000360757112503,
+      "learning_rate": 2.658464542291422e-05,
+      "loss": 10.1372,
+      "step": 23240
+    },
+    {
+      "epoch": 0.11610776798421933,
+      "grad_norm": 0.09418469667434692,
+      "learning_rate": 2.6583143507972666e-05,
+      "loss": 10.1447,
+      "step": 23250
+    },
+    {
+      "epoch": 0.11615770680915878,
+      "grad_norm": 0.09615140408277512,
+      "learning_rate": 2.6581641593031113e-05,
+      "loss": 10.1425,
+      "step": 23260
+    },
+    {
+      "epoch": 0.11620764563409823,
+      "grad_norm": 0.0987091064453125,
+      "learning_rate": 2.6580139678089566e-05,
+      "loss": 10.1446,
+      "step": 23270
+    },
+    {
+      "epoch": 0.11625758445903768,
+      "grad_norm": 0.09312444925308228,
+      "learning_rate": 2.6578637763148013e-05,
+      "loss": 10.1424,
+      "step": 23280
+    },
+    {
+      "epoch": 0.11630752328397713,
+      "grad_norm": 0.09860526770353317,
+      "learning_rate": 2.6577135848206466e-05,
+      "loss": 10.143,
+      "step": 23290
+    },
+    {
+      "epoch": 0.11635746210891658,
+      "grad_norm": 0.0991264283657074,
+      "learning_rate": 2.6575633933264913e-05,
+      "loss": 10.1415,
+      "step": 23300
+    },
+    {
+      "epoch": 0.11640740093385603,
+      "grad_norm": 0.0935107171535492,
+      "learning_rate": 2.657413201832336e-05,
+      "loss": 10.1431,
+      "step": 23310
+    },
+    {
+      "epoch": 0.11645733975879548,
+      "grad_norm": 0.09420657157897949,
+      "learning_rate": 2.6572630103381814e-05,
+      "loss": 10.1419,
+      "step": 23320
+    },
+    {
+      "epoch": 0.11650727858373493,
+      "grad_norm": 0.09580549597740173,
+      "learning_rate": 2.657112818844026e-05,
+      "loss": 10.1444,
+      "step": 23330
+    },
+    {
+      "epoch": 0.11655721740867438,
+      "grad_norm": 0.09820415079593658,
+      "learning_rate": 2.6569626273498714e-05,
+      "loss": 10.1408,
+      "step": 23340
+    },
+    {
+      "epoch": 0.11660715623361383,
+      "grad_norm": 0.0958128347992897,
+      "learning_rate": 2.656812435855716e-05,
+      "loss": 10.143,
+      "step": 23350
+    },
+    {
+      "epoch": 0.11665709505855328,
+      "grad_norm": 0.09436112642288208,
+      "learning_rate": 2.6566622443615608e-05,
+      "loss": 10.1437,
+      "step": 23360
+    },
+    {
+      "epoch": 0.11670703388349273,
+      "grad_norm": 0.09518658369779587,
+      "learning_rate": 2.656512052867406e-05,
+      "loss": 10.1386,
+      "step": 23370
+    },
+    {
+      "epoch": 0.11675697270843217,
+      "grad_norm": 0.08935555815696716,
+      "learning_rate": 2.6563618613732508e-05,
+      "loss": 10.1418,
+      "step": 23380
+    },
+    {
+      "epoch": 0.11680691153337162,
+      "grad_norm": 0.09481310844421387,
+      "learning_rate": 2.656211669879096e-05,
+      "loss": 10.1424,
+      "step": 23390
+    },
+    {
+      "epoch": 0.11685685035831107,
+      "grad_norm": 0.09546245634555817,
+      "learning_rate": 2.6560614783849408e-05,
+      "loss": 10.1411,
+      "step": 23400
+    },
+    {
+      "epoch": 0.11690678918325052,
+      "grad_norm": 0.09654063731431961,
+      "learning_rate": 2.6559112868907855e-05,
+      "loss": 10.1385,
+      "step": 23410
+    },
+    {
+      "epoch": 0.11695672800818997,
+      "grad_norm": 0.09841468185186386,
+      "learning_rate": 2.655761095396631e-05,
+      "loss": 10.1373,
+      "step": 23420
+    },
+    {
+      "epoch": 0.11700666683312942,
+      "grad_norm": 0.09300010651350021,
+      "learning_rate": 2.6556109039024755e-05,
+      "loss": 10.1345,
+      "step": 23430
+    },
+    {
+      "epoch": 0.11705660565806887,
+      "grad_norm": 0.09334631264209747,
+      "learning_rate": 2.655460712408321e-05,
+      "loss": 10.1359,
+      "step": 23440
+    },
+    {
+      "epoch": 0.11710654448300832,
+      "grad_norm": 0.09312853217124939,
+      "learning_rate": 2.6553105209141656e-05,
+      "loss": 10.1425,
+      "step": 23450
+    },
+    {
+      "epoch": 0.11715648330794777,
+      "grad_norm": 0.0956452414393425,
+      "learning_rate": 2.6551603294200106e-05,
+      "loss": 10.1425,
+      "step": 23460
+    },
+    {
+      "epoch": 0.11720642213288722,
+      "grad_norm": 0.0932190865278244,
+      "learning_rate": 2.6550101379258556e-05,
+      "loss": 10.1389,
+      "step": 23470
+    },
+    {
+      "epoch": 0.11725636095782667,
+      "grad_norm": 0.09214978665113449,
+      "learning_rate": 2.6548599464317003e-05,
+      "loss": 10.1381,
+      "step": 23480
+    },
+    {
+      "epoch": 0.1173062997827661,
+      "grad_norm": 0.09422241896390915,
+      "learning_rate": 2.6547097549375456e-05,
+      "loss": 10.137,
+      "step": 23490
+    },
+    {
+      "epoch": 0.11735623860770555,
+      "grad_norm": 0.09438539296388626,
+      "learning_rate": 2.6545595634433903e-05,
+      "loss": 10.1369,
+      "step": 23500
+    },
+    {
+      "epoch": 0.117406177432645,
+      "grad_norm": 0.09662014991044998,
+      "learning_rate": 2.6544093719492353e-05,
+      "loss": 10.1367,
+      "step": 23510
+    },
+    {
+      "epoch": 0.11745611625758445,
+      "grad_norm": 0.08834582567214966,
+      "learning_rate": 2.6542591804550804e-05,
+      "loss": 10.1387,
+      "step": 23520
+    },
+    {
+      "epoch": 0.1175060550825239,
+      "grad_norm": 0.09564156085252762,
+      "learning_rate": 2.654108988960925e-05,
+      "loss": 10.1417,
+      "step": 23530
+    },
+    {
+      "epoch": 0.11755599390746335,
+      "grad_norm": 0.09563954174518585,
+      "learning_rate": 2.6539587974667704e-05,
+      "loss": 10.1339,
+      "step": 23540
+    },
+    {
+      "epoch": 0.1176059327324028,
+      "grad_norm": 0.0935792475938797,
+      "learning_rate": 2.653808605972615e-05,
+      "loss": 10.1307,
+      "step": 23550
+    },
+    {
+      "epoch": 0.11765587155734225,
+      "grad_norm": 0.10603775829076767,
+      "learning_rate": 2.65365841447846e-05,
+      "loss": 10.1355,
+      "step": 23560
+    },
+    {
+      "epoch": 0.1177058103822817,
+      "grad_norm": 0.09258319437503815,
+      "learning_rate": 2.653508222984305e-05,
+      "loss": 10.1355,
+      "step": 23570
+    },
+    {
+      "epoch": 0.11775574920722115,
+      "grad_norm": 0.09184292703866959,
+      "learning_rate": 2.6533580314901498e-05,
+      "loss": 10.1333,
+      "step": 23580
+    },
+    {
+      "epoch": 0.1178056880321606,
+      "grad_norm": 0.09775898605585098,
+      "learning_rate": 2.653207839995995e-05,
+      "loss": 10.137,
+      "step": 23590
+    },
+    {
+      "epoch": 0.11785562685710005,
+      "grad_norm": 0.0923045426607132,
+      "learning_rate": 2.6530576485018398e-05,
+      "loss": 10.1308,
+      "step": 23600
+    },
+    {
+      "epoch": 0.1179055656820395,
+      "grad_norm": 0.09056799858808517,
+      "learning_rate": 2.652907457007685e-05,
+      "loss": 10.1359,
+      "step": 23610
+    },
+    {
+      "epoch": 0.11795550450697895,
+      "grad_norm": 0.09372033178806305,
+      "learning_rate": 2.65275726551353e-05,
+      "loss": 10.1317,
+      "step": 23620
+    },
+    {
+      "epoch": 0.1180054433319184,
+      "grad_norm": 0.09503508359193802,
+      "learning_rate": 2.6526070740193745e-05,
+      "loss": 10.1331,
+      "step": 23630
+    },
+    {
+      "epoch": 0.11805538215685785,
+      "grad_norm": 0.09287771582603455,
+      "learning_rate": 2.65245688252522e-05,
+      "loss": 10.1362,
+      "step": 23640
+    },
+    {
+      "epoch": 0.1181053209817973,
+      "grad_norm": 0.09275928884744644,
+      "learning_rate": 2.6523066910310646e-05,
+      "loss": 10.1329,
+      "step": 23650
+    },
+    {
+      "epoch": 0.11815525980673675,
+      "grad_norm": 0.09057994186878204,
+      "learning_rate": 2.6521564995369096e-05,
+      "loss": 10.1317,
+      "step": 23660
+    },
+    {
+      "epoch": 0.1182051986316762,
+      "grad_norm": 0.09248855710029602,
+      "learning_rate": 2.6520063080427546e-05,
+      "loss": 10.1321,
+      "step": 23670
+    },
+    {
+      "epoch": 0.11825513745661564,
+      "grad_norm": 0.09153629094362259,
+      "learning_rate": 2.6518561165485993e-05,
+      "loss": 10.1315,
+      "step": 23680
+    },
+    {
+      "epoch": 0.1183050762815551,
+      "grad_norm": 0.08953340351581573,
+      "learning_rate": 2.6517059250544446e-05,
+      "loss": 10.1348,
+      "step": 23690
+    },
+    {
+      "epoch": 0.11835501510649454,
+      "grad_norm": 0.09959051012992859,
+      "learning_rate": 2.6515557335602893e-05,
+      "loss": 10.1346,
+      "step": 23700
+    },
+    {
+      "epoch": 0.11840495393143399,
+      "grad_norm": 0.09226525574922562,
+      "learning_rate": 2.6514055420661343e-05,
+      "loss": 10.1304,
+      "step": 23710
+    },
+    {
+      "epoch": 0.11845489275637344,
+      "grad_norm": 0.09568504989147186,
+      "learning_rate": 2.6512553505719794e-05,
+      "loss": 10.1275,
+      "step": 23720
+    },
+    {
+      "epoch": 0.11850483158131289,
+      "grad_norm": 0.0921492800116539,
+      "learning_rate": 2.651105159077824e-05,
+      "loss": 10.1283,
+      "step": 23730
+    },
+    {
+      "epoch": 0.11855477040625234,
+      "grad_norm": 0.09444059431552887,
+      "learning_rate": 2.6509549675836694e-05,
+      "loss": 10.1336,
+      "step": 23740
+    },
+    {
+      "epoch": 0.11860470923119179,
+      "grad_norm": 0.10168434679508209,
+      "learning_rate": 2.650804776089514e-05,
+      "loss": 10.1297,
+      "step": 23750
+    },
+    {
+      "epoch": 0.11865464805613124,
+      "grad_norm": 0.09380350261926651,
+      "learning_rate": 2.650654584595359e-05,
+      "loss": 10.1278,
+      "step": 23760
+    },
+    {
+      "epoch": 0.11870458688107069,
+      "grad_norm": 0.10053762793540955,
+      "learning_rate": 2.650504393101204e-05,
+      "loss": 10.1349,
+      "step": 23770
+    },
+    {
+      "epoch": 0.11875452570601014,
+      "grad_norm": 0.09257914870977402,
+      "learning_rate": 2.650354201607049e-05,
+      "loss": 10.1311,
+      "step": 23780
+    },
+    {
+      "epoch": 0.11880446453094959,
+      "grad_norm": 0.09091336280107498,
+      "learning_rate": 2.650204010112894e-05,
+      "loss": 10.1314,
+      "step": 23790
+    },
+    {
+      "epoch": 0.11885440335588904,
+      "grad_norm": 0.09331763535737991,
+      "learning_rate": 2.6500538186187388e-05,
+      "loss": 10.1299,
+      "step": 23800
+    },
+    {
+      "epoch": 0.11890434218082849,
+      "grad_norm": 0.09806916862726212,
+      "learning_rate": 2.649903627124584e-05,
+      "loss": 10.1285,
+      "step": 23810
+    },
+    {
+      "epoch": 0.11895428100576794,
+      "grad_norm": 0.09568050503730774,
+      "learning_rate": 2.649753435630429e-05,
+      "loss": 10.1295,
+      "step": 23820
+    },
+    {
+      "epoch": 0.11900421983070739,
+      "grad_norm": 0.09739750623703003,
+      "learning_rate": 2.649603244136274e-05,
+      "loss": 10.1288,
+      "step": 23830
+    },
+    {
+      "epoch": 0.11905415865564684,
+      "grad_norm": 0.0919250100851059,
+      "learning_rate": 2.649453052642119e-05,
+      "loss": 10.1282,
+      "step": 23840
+    },
+    {
+      "epoch": 0.11910409748058629,
+      "grad_norm": 0.09634441137313843,
+      "learning_rate": 2.6493028611479636e-05,
+      "loss": 10.1263,
+      "step": 23850
+    },
+    {
+      "epoch": 0.11915403630552573,
+      "grad_norm": 0.09239911288022995,
+      "learning_rate": 2.6491526696538086e-05,
+      "loss": 10.1264,
+      "step": 23860
+    },
+    {
+      "epoch": 0.11920397513046518,
+      "grad_norm": 0.09922140091657639,
+      "learning_rate": 2.6490024781596536e-05,
+      "loss": 10.1257,
+      "step": 23870
+    },
+    {
+      "epoch": 0.11925391395540463,
+      "grad_norm": 0.09263952821493149,
+      "learning_rate": 2.6488522866654986e-05,
+      "loss": 10.127,
+      "step": 23880
+    },
+    {
+      "epoch": 0.11930385278034408,
+      "grad_norm": 0.09615729004144669,
+      "learning_rate": 2.6487020951713437e-05,
+      "loss": 10.125,
+      "step": 23890
+    },
+    {
+      "epoch": 0.11935379160528353,
+      "grad_norm": 0.09644229710102081,
+      "learning_rate": 2.6485519036771883e-05,
+      "loss": 10.1238,
+      "step": 23900
+    },
+    {
+      "epoch": 0.11940373043022298,
+      "grad_norm": 0.09358090162277222,
+      "learning_rate": 2.6484017121830333e-05,
+      "loss": 10.1288,
+      "step": 23910
+    },
+    {
+      "epoch": 0.11945366925516243,
+      "grad_norm": 0.0884949341416359,
+      "learning_rate": 2.6482515206888784e-05,
+      "loss": 10.1262,
+      "step": 23920
+    },
+    {
+      "epoch": 0.11950360808010188,
+      "grad_norm": 0.09220154583454132,
+      "learning_rate": 2.6481013291947234e-05,
+      "loss": 10.1266,
+      "step": 23930
+    },
+    {
+      "epoch": 0.11955354690504133,
+      "grad_norm": 0.09416556358337402,
+      "learning_rate": 2.6479511377005684e-05,
+      "loss": 10.1293,
+      "step": 23940
+    },
+    {
+      "epoch": 0.11960348572998078,
+      "grad_norm": 0.09228892624378204,
+      "learning_rate": 2.647800946206413e-05,
+      "loss": 10.1261,
+      "step": 23950
+    },
+    {
+      "epoch": 0.11965342455492023,
+      "grad_norm": 0.09278880804777145,
+      "learning_rate": 2.647650754712258e-05,
+      "loss": 10.1252,
+      "step": 23960
+    },
+    {
+      "epoch": 0.11970336337985968,
+      "grad_norm": 0.09540235996246338,
+      "learning_rate": 2.647500563218103e-05,
+      "loss": 10.1248,
+      "step": 23970
+    },
+    {
+      "epoch": 0.11975330220479911,
+      "grad_norm": 0.09458044171333313,
+      "learning_rate": 2.647350371723948e-05,
+      "loss": 10.1272,
+      "step": 23980
+    },
+    {
+      "epoch": 0.11980324102973856,
+      "grad_norm": 0.09298811852931976,
+      "learning_rate": 2.647200180229793e-05,
+      "loss": 10.1196,
+      "step": 23990
+    },
+    {
+      "epoch": 0.11985317985467801,
+      "grad_norm": 0.09707672894001007,
+      "learning_rate": 2.6470499887356378e-05,
+      "loss": 10.1218,
+      "step": 24000
+    },
+    {
+      "epoch": 0.11990311867961746,
+      "grad_norm": 0.09375189244747162,
+      "learning_rate": 2.646899797241483e-05,
+      "loss": 10.1282,
+      "step": 24010
+    },
+    {
+      "epoch": 0.11995305750455691,
+      "grad_norm": 0.1018557995557785,
+      "learning_rate": 2.646749605747328e-05,
+      "loss": 10.1204,
+      "step": 24020
+    },
+    {
+      "epoch": 0.12000299632949636,
+      "grad_norm": 0.09262654185295105,
+      "learning_rate": 2.646599414253173e-05,
+      "loss": 10.1235,
+      "step": 24030
+    },
+    {
+      "epoch": 0.12005293515443581,
+      "grad_norm": 0.09324292093515396,
+      "learning_rate": 2.646449222759018e-05,
+      "loss": 10.1253,
+      "step": 24040
+    },
+    {
+      "epoch": 0.12010287397937526,
+      "grad_norm": 0.09518759697675705,
+      "learning_rate": 2.6462990312648626e-05,
+      "loss": 10.1231,
+      "step": 24050
+    },
+    {
+      "epoch": 0.12015281280431471,
+      "grad_norm": 0.09306249022483826,
+      "learning_rate": 2.6461488397707076e-05,
+      "loss": 10.126,
+      "step": 24060
+    },
+    {
+      "epoch": 0.12020275162925416,
+      "grad_norm": 0.09253344684839249,
+      "learning_rate": 2.6459986482765526e-05,
+      "loss": 10.1173,
+      "step": 24070
+    },
+    {
+      "epoch": 0.12025269045419361,
+      "grad_norm": 0.09460706263780594,
+      "learning_rate": 2.6458484567823976e-05,
+      "loss": 10.1269,
+      "step": 24080
+    },
+    {
+      "epoch": 0.12030262927913306,
+      "grad_norm": 0.08938644081354141,
+      "learning_rate": 2.6456982652882427e-05,
+      "loss": 10.1225,
+      "step": 24090
+    },
+    {
+      "epoch": 0.12035256810407251,
+      "grad_norm": 0.09312314540147781,
+      "learning_rate": 2.6455480737940873e-05,
+      "loss": 10.122,
+      "step": 24100
+    },
+    {
+      "epoch": 0.12040250692901196,
+      "grad_norm": 0.09598222374916077,
+      "learning_rate": 2.6453978822999323e-05,
+      "loss": 10.1195,
+      "step": 24110
+    },
+    {
+      "epoch": 0.1204524457539514,
+      "grad_norm": 0.09335832297801971,
+      "learning_rate": 2.6452476908057774e-05,
+      "loss": 10.1204,
+      "step": 24120
+    },
+    {
+      "epoch": 0.12050238457889086,
+      "grad_norm": 0.09654422104358673,
+      "learning_rate": 2.6450974993116224e-05,
+      "loss": 10.1206,
+      "step": 24130
+    },
+    {
+      "epoch": 0.1205523234038303,
+      "grad_norm": 0.0975603386759758,
+      "learning_rate": 2.6449473078174674e-05,
+      "loss": 10.1211,
+      "step": 24140
+    },
+    {
+      "epoch": 0.12060226222876975,
+      "grad_norm": 0.09276950359344482,
+      "learning_rate": 2.6447971163233124e-05,
+      "loss": 10.1188,
+      "step": 24150
+    },
+    {
+      "epoch": 0.1206522010537092,
+      "grad_norm": 0.0958276093006134,
+      "learning_rate": 2.644646924829157e-05,
+      "loss": 10.1241,
+      "step": 24160
+    },
+    {
+      "epoch": 0.12070213987864865,
+      "grad_norm": 0.09583685547113419,
+      "learning_rate": 2.644496733335002e-05,
+      "loss": 10.1209,
+      "step": 24170
+    },
+    {
+      "epoch": 0.1207520787035881,
+      "grad_norm": 0.09288566559553146,
+      "learning_rate": 2.644346541840847e-05,
+      "loss": 10.1184,
+      "step": 24180
+    },
+    {
+      "epoch": 0.12080201752852755,
+      "grad_norm": 0.0899089053273201,
+      "learning_rate": 2.644196350346692e-05,
+      "loss": 10.1188,
+      "step": 24190
+    },
+    {
+      "epoch": 0.120851956353467,
+      "grad_norm": 0.09436280280351639,
+      "learning_rate": 2.6440461588525372e-05,
+      "loss": 10.12,
+      "step": 24200
+    },
+    {
+      "epoch": 0.12090189517840645,
+      "grad_norm": 0.09277939051389694,
+      "learning_rate": 2.643895967358382e-05,
+      "loss": 10.1183,
+      "step": 24210
+    },
+    {
+      "epoch": 0.1209518340033459,
+      "grad_norm": 0.08972795307636261,
+      "learning_rate": 2.643745775864227e-05,
+      "loss": 10.1201,
+      "step": 24220
+    },
+    {
+      "epoch": 0.12100177282828535,
+      "grad_norm": 0.09541310369968414,
+      "learning_rate": 2.643595584370072e-05,
+      "loss": 10.1192,
+      "step": 24230
+    },
+    {
+      "epoch": 0.1210517116532248,
+      "grad_norm": 0.08959456533193588,
+      "learning_rate": 2.643445392875917e-05,
+      "loss": 10.1224,
+      "step": 24240
+    },
+    {
+      "epoch": 0.12110165047816425,
+      "grad_norm": 0.09786181151866913,
+      "learning_rate": 2.643295201381762e-05,
+      "loss": 10.117,
+      "step": 24250
+    },
+    {
+      "epoch": 0.1211515893031037,
+      "grad_norm": 0.09291178733110428,
+      "learning_rate": 2.6431450098876066e-05,
+      "loss": 10.1203,
+      "step": 24260
+    },
+    {
+      "epoch": 0.12120152812804315,
+      "grad_norm": 0.10062135756015778,
+      "learning_rate": 2.6429948183934516e-05,
+      "loss": 10.1179,
+      "step": 24270
+    },
+    {
+      "epoch": 0.1212514669529826,
+      "grad_norm": 0.09641305357217789,
+      "learning_rate": 2.6428446268992966e-05,
+      "loss": 10.1198,
+      "step": 24280
+    },
+    {
+      "epoch": 0.12130140577792205,
+      "grad_norm": 0.09644697606563568,
+      "learning_rate": 2.6426944354051417e-05,
+      "loss": 10.113,
+      "step": 24290
+    },
+    {
+      "epoch": 0.1213513446028615,
+      "grad_norm": 0.0892978310585022,
+      "learning_rate": 2.6425442439109867e-05,
+      "loss": 10.118,
+      "step": 24300
+    },
+    {
+      "epoch": 0.12140128342780095,
+      "grad_norm": 0.0910135954618454,
+      "learning_rate": 2.6423940524168313e-05,
+      "loss": 10.1194,
+      "step": 24310
+    },
+    {
+      "epoch": 0.1214512222527404,
+      "grad_norm": 0.0961785763502121,
+      "learning_rate": 2.6422438609226764e-05,
+      "loss": 10.116,
+      "step": 24320
+    },
+    {
+      "epoch": 0.12150116107767985,
+      "grad_norm": 0.09364509582519531,
+      "learning_rate": 2.6420936694285214e-05,
+      "loss": 10.111,
+      "step": 24330
+    },
+    {
+      "epoch": 0.1215510999026193,
+      "grad_norm": 0.09308196604251862,
+      "learning_rate": 2.6419434779343664e-05,
+      "loss": 10.1159,
+      "step": 24340
+    },
+    {
+      "epoch": 0.12160103872755874,
+      "grad_norm": 0.09322807192802429,
+      "learning_rate": 2.6417932864402114e-05,
+      "loss": 10.1159,
+      "step": 24350
+    },
+    {
+      "epoch": 0.1216509775524982,
+      "grad_norm": 0.09460312128067017,
+      "learning_rate": 2.641643094946056e-05,
+      "loss": 10.1212,
+      "step": 24360
+    },
+    {
+      "epoch": 0.12170091637743764,
+      "grad_norm": 0.0933726355433464,
+      "learning_rate": 2.641492903451901e-05,
+      "loss": 10.1137,
+      "step": 24370
+    },
+    {
+      "epoch": 0.12175085520237709,
+      "grad_norm": 0.09838582575321198,
+      "learning_rate": 2.641342711957746e-05,
+      "loss": 10.1168,
+      "step": 24380
+    },
+    {
+      "epoch": 0.12180079402731654,
+      "grad_norm": 0.09491819143295288,
+      "learning_rate": 2.641192520463591e-05,
+      "loss": 10.1143,
+      "step": 24390
+    },
+    {
+      "epoch": 0.12185073285225599,
+      "grad_norm": 0.0954071581363678,
+      "learning_rate": 2.6410423289694362e-05,
+      "loss": 10.1135,
+      "step": 24400
+    },
+    {
+      "epoch": 0.12190067167719544,
+      "grad_norm": 0.09948980063199997,
+      "learning_rate": 2.6408921374752812e-05,
+      "loss": 10.1176,
+      "step": 24410
+    },
+    {
+      "epoch": 0.12195061050213489,
+      "grad_norm": 0.09514214843511581,
+      "learning_rate": 2.640741945981126e-05,
+      "loss": 10.1187,
+      "step": 24420
+    },
+    {
+      "epoch": 0.12200054932707434,
+      "grad_norm": 0.09480323642492294,
+      "learning_rate": 2.640591754486971e-05,
+      "loss": 10.1122,
+      "step": 24430
+    },
+    {
+      "epoch": 0.12205048815201379,
+      "grad_norm": 0.09891213476657867,
+      "learning_rate": 2.640441562992816e-05,
+      "loss": 10.1107,
+      "step": 24440
+    },
+    {
+      "epoch": 0.12210042697695324,
+      "grad_norm": 0.09468024224042892,
+      "learning_rate": 2.640291371498661e-05,
+      "loss": 10.1156,
+      "step": 24450
+    },
+    {
+      "epoch": 0.12215036580189269,
+      "grad_norm": 0.08860956132411957,
+      "learning_rate": 2.640141180004506e-05,
+      "loss": 10.1124,
+      "step": 24460
+    },
+    {
+      "epoch": 0.12220030462683214,
+      "grad_norm": 0.09128830581903458,
+      "learning_rate": 2.639990988510351e-05,
+      "loss": 10.1173,
+      "step": 24470
+    },
+    {
+      "epoch": 0.12225024345177157,
+      "grad_norm": 0.09434867650270462,
+      "learning_rate": 2.6398407970161956e-05,
+      "loss": 10.1109,
+      "step": 24480
+    },
+    {
+      "epoch": 0.12230018227671102,
+      "grad_norm": 0.08857893943786621,
+      "learning_rate": 2.6396906055220407e-05,
+      "loss": 10.1099,
+      "step": 24490
+    },
+    {
+      "epoch": 0.12235012110165047,
+      "grad_norm": 0.09494607895612717,
+      "learning_rate": 2.6395404140278857e-05,
+      "loss": 10.1118,
+      "step": 24500
+    },
+    {
+      "epoch": 0.12240005992658992,
+      "grad_norm": 0.09442712366580963,
+      "learning_rate": 2.6393902225337307e-05,
+      "loss": 10.1108,
+      "step": 24510
+    },
+    {
+      "epoch": 0.12244999875152937,
+      "grad_norm": 0.0935530737042427,
+      "learning_rate": 2.6392400310395757e-05,
+      "loss": 10.1098,
+      "step": 24520
+    },
+    {
+      "epoch": 0.12249993757646882,
+      "grad_norm": 0.09096985310316086,
+      "learning_rate": 2.6390898395454204e-05,
+      "loss": 10.1132,
+      "step": 24530
+    },
+    {
+      "epoch": 0.12254987640140827,
+      "grad_norm": 0.10358789563179016,
+      "learning_rate": 2.6389396480512654e-05,
+      "loss": 10.1113,
+      "step": 24540
+    },
+    {
+      "epoch": 0.12259981522634772,
+      "grad_norm": 0.10106457769870758,
+      "learning_rate": 2.6387894565571104e-05,
+      "loss": 10.1086,
+      "step": 24550
+    },
+    {
+      "epoch": 0.12264975405128717,
+      "grad_norm": 0.10155883431434631,
+      "learning_rate": 2.6386392650629554e-05,
+      "loss": 10.1106,
+      "step": 24560
+    },
+    {
+      "epoch": 0.12269969287622662,
+      "grad_norm": 0.09161602705717087,
+      "learning_rate": 2.6384890735688005e-05,
+      "loss": 10.1158,
+      "step": 24570
+    },
+    {
+      "epoch": 0.12274963170116607,
+      "grad_norm": 0.09335344284772873,
+      "learning_rate": 2.638338882074645e-05,
+      "loss": 10.1075,
+      "step": 24580
+    },
+    {
+      "epoch": 0.12279957052610552,
+      "grad_norm": 0.0948810800909996,
+      "learning_rate": 2.63818869058049e-05,
+      "loss": 10.1117,
+      "step": 24590
+    },
+    {
+      "epoch": 0.12284950935104497,
+      "grad_norm": 0.09355373680591583,
+      "learning_rate": 2.6380384990863352e-05,
+      "loss": 10.1119,
+      "step": 24600
+    },
+    {
+      "epoch": 0.12289944817598442,
+      "grad_norm": 0.09632193297147751,
+      "learning_rate": 2.6378883075921802e-05,
+      "loss": 10.109,
+      "step": 24610
+    },
+    {
+      "epoch": 0.12294938700092387,
+      "grad_norm": 0.09129033237695694,
+      "learning_rate": 2.6377381160980252e-05,
+      "loss": 10.1067,
+      "step": 24620
+    },
+    {
+      "epoch": 0.12299932582586331,
+      "grad_norm": 0.09343235939741135,
+      "learning_rate": 2.63758792460387e-05,
+      "loss": 10.11,
+      "step": 24630
+    },
+    {
+      "epoch": 0.12304926465080276,
+      "grad_norm": 0.0997580885887146,
+      "learning_rate": 2.637437733109715e-05,
+      "loss": 10.1113,
+      "step": 24640
+    },
+    {
+      "epoch": 0.12309920347574221,
+      "grad_norm": 0.09580541402101517,
+      "learning_rate": 2.63728754161556e-05,
+      "loss": 10.1108,
+      "step": 24650
+    },
+    {
+      "epoch": 0.12314914230068166,
+      "grad_norm": 0.0917908325791359,
+      "learning_rate": 2.637137350121405e-05,
+      "loss": 10.1129,
+      "step": 24660
+    },
+    {
+      "epoch": 0.12319908112562111,
+      "grad_norm": 0.09625443816184998,
+      "learning_rate": 2.63698715862725e-05,
+      "loss": 10.1065,
+      "step": 24670
+    },
+    {
+      "epoch": 0.12324901995056056,
+      "grad_norm": 0.0961778461933136,
+      "learning_rate": 2.6368369671330946e-05,
+      "loss": 10.1107,
+      "step": 24680
+    },
+    {
+      "epoch": 0.12329895877550001,
+      "grad_norm": 0.09377621114253998,
+      "learning_rate": 2.6366867756389397e-05,
+      "loss": 10.1132,
+      "step": 24690
+    },
+    {
+      "epoch": 0.12334889760043946,
+      "grad_norm": 0.09123267978429794,
+      "learning_rate": 2.6365365841447847e-05,
+      "loss": 10.1092,
+      "step": 24700
+    },
+    {
+      "epoch": 0.12339883642537891,
+      "grad_norm": 0.09666462987661362,
+      "learning_rate": 2.6363863926506297e-05,
+      "loss": 10.106,
+      "step": 24710
+    },
+    {
+      "epoch": 0.12344877525031836,
+      "grad_norm": 0.09382550418376923,
+      "learning_rate": 2.6362362011564747e-05,
+      "loss": 10.1071,
+      "step": 24720
+    },
+    {
+      "epoch": 0.12349871407525781,
+      "grad_norm": 0.10108333826065063,
+      "learning_rate": 2.6360860096623194e-05,
+      "loss": 10.1022,
+      "step": 24730
+    },
+    {
+      "epoch": 0.12354865290019726,
+      "grad_norm": 0.089703269302845,
+      "learning_rate": 2.6359358181681644e-05,
+      "loss": 10.1107,
+      "step": 24740
+    },
+    {
+      "epoch": 0.12359859172513671,
+      "grad_norm": 0.09390883892774582,
+      "learning_rate": 2.6357856266740094e-05,
+      "loss": 10.1051,
+      "step": 24750
+    },
+    {
+      "epoch": 0.12364853055007616,
+      "grad_norm": 0.0946129560470581,
+      "learning_rate": 2.6356354351798544e-05,
+      "loss": 10.1041,
+      "step": 24760
+    },
+    {
+      "epoch": 0.12369846937501561,
+      "grad_norm": 0.09596185386180878,
+      "learning_rate": 2.6354852436856995e-05,
+      "loss": 10.1073,
+      "step": 24770
+    },
+    {
+      "epoch": 0.12374840819995506,
+      "grad_norm": 0.09378567337989807,
+      "learning_rate": 2.635335052191544e-05,
+      "loss": 10.1047,
+      "step": 24780
+    },
+    {
+      "epoch": 0.1237983470248945,
+      "grad_norm": 0.09474289417266846,
+      "learning_rate": 2.6351848606973895e-05,
+      "loss": 10.1043,
+      "step": 24790
+    },
+    {
+      "epoch": 0.12384828584983396,
+      "grad_norm": 0.09176385402679443,
+      "learning_rate": 2.6350346692032342e-05,
+      "loss": 10.1048,
+      "step": 24800
+    },
+    {
+      "epoch": 0.1238982246747734,
+      "grad_norm": 0.0884096696972847,
+      "learning_rate": 2.6348844777090792e-05,
+      "loss": 10.1021,
+      "step": 24810
+    },
+    {
+      "epoch": 0.12394816349971285,
+      "grad_norm": 0.09617920964956284,
+      "learning_rate": 2.6347342862149242e-05,
+      "loss": 10.1041,
+      "step": 24820
+    },
+    {
+      "epoch": 0.1239981023246523,
+      "grad_norm": 0.09283927828073502,
+      "learning_rate": 2.634584094720769e-05,
+      "loss": 10.1026,
+      "step": 24830
+    },
+    {
+      "epoch": 0.12404804114959175,
+      "grad_norm": 0.09213951230049133,
+      "learning_rate": 2.6344339032266142e-05,
+      "loss": 10.1019,
+      "step": 24840
+    },
+    {
+      "epoch": 0.1240979799745312,
+      "grad_norm": 0.09610681235790253,
+      "learning_rate": 2.634283711732459e-05,
+      "loss": 10.1075,
+      "step": 24850
+    },
+    {
+      "epoch": 0.12414791879947065,
+      "grad_norm": 0.10006497800350189,
+      "learning_rate": 2.634133520238304e-05,
+      "loss": 10.103,
+      "step": 24860
+    },
+    {
+      "epoch": 0.1241978576244101,
+      "grad_norm": 0.09113096445798874,
+      "learning_rate": 2.633983328744149e-05,
+      "loss": 10.1074,
+      "step": 24870
+    },
+    {
+      "epoch": 0.12424779644934955,
+      "grad_norm": 0.10478292405605316,
+      "learning_rate": 2.6338331372499936e-05,
+      "loss": 10.0992,
+      "step": 24880
+    },
+    {
+      "epoch": 0.124297735274289,
+      "grad_norm": 0.09672685712575912,
+      "learning_rate": 2.633682945755839e-05,
+      "loss": 10.1028,
+      "step": 24890
+    },
+    {
+      "epoch": 0.12434767409922845,
+      "grad_norm": 0.0986834466457367,
+      "learning_rate": 2.6335327542616837e-05,
+      "loss": 10.0989,
+      "step": 24900
+    },
+    {
+      "epoch": 0.1243976129241679,
+      "grad_norm": 0.10003640502691269,
+      "learning_rate": 2.6333825627675287e-05,
+      "loss": 10.1044,
+      "step": 24910
+    },
+    {
+      "epoch": 0.12444755174910735,
+      "grad_norm": 0.09739897400140762,
+      "learning_rate": 2.6332323712733737e-05,
+      "loss": 10.1055,
+      "step": 24920
+    },
+    {
+      "epoch": 0.1244974905740468,
+      "grad_norm": 0.09403092414140701,
+      "learning_rate": 2.6330821797792184e-05,
+      "loss": 10.1052,
+      "step": 24930
+    },
+    {
+      "epoch": 0.12454742939898625,
+      "grad_norm": 0.0988364890217781,
+      "learning_rate": 2.6329319882850637e-05,
+      "loss": 10.1025,
+      "step": 24940
+    },
+    {
+      "epoch": 0.1245973682239257,
+      "grad_norm": 0.09991788864135742,
+      "learning_rate": 2.6327817967909084e-05,
+      "loss": 10.1009,
+      "step": 24950
+    },
+    {
+      "epoch": 0.12464730704886515,
+      "grad_norm": 0.09546318650245667,
+      "learning_rate": 2.6326316052967534e-05,
+      "loss": 10.0996,
+      "step": 24960
+    },
+    {
+      "epoch": 0.12469724587380458,
+      "grad_norm": 0.09446793794631958,
+      "learning_rate": 2.6324814138025985e-05,
+      "loss": 10.1004,
+      "step": 24970
+    },
+    {
+      "epoch": 0.12474718469874403,
+      "grad_norm": 0.09391462057828903,
+      "learning_rate": 2.632331222308443e-05,
+      "loss": 10.1005,
+      "step": 24980
+    },
+    {
+      "epoch": 0.12479712352368348,
+      "grad_norm": 0.08938247710466385,
+      "learning_rate": 2.6321810308142885e-05,
+      "loss": 10.0976,
+      "step": 24990
+    },
+    {
+      "epoch": 0.12484706234862293,
+      "grad_norm": 0.0972147136926651,
+      "learning_rate": 2.6320308393201332e-05,
+      "loss": 10.0994,
+      "step": 25000
+    },
+    {
+      "epoch": 0.12489700117356238,
+      "grad_norm": 0.09196930378675461,
+      "learning_rate": 2.6318806478259782e-05,
+      "loss": 10.0997,
+      "step": 25010
+    },
+    {
+      "epoch": 0.12494693999850183,
+      "grad_norm": 0.10306332260370255,
+      "learning_rate": 2.6317304563318232e-05,
+      "loss": 10.0984,
+      "step": 25020
+    },
+    {
+      "epoch": 0.12499687882344128,
+      "grad_norm": 0.09357580542564392,
+      "learning_rate": 2.631580264837668e-05,
+      "loss": 10.1022,
+      "step": 25030
+    },
+    {
+      "epoch": 0.12504681764838074,
+      "grad_norm": 0.09815985709428787,
+      "learning_rate": 2.6314300733435132e-05,
+      "loss": 10.0966,
+      "step": 25040
+    },
+    {
+      "epoch": 0.1250967564733202,
+      "grad_norm": 0.09139269590377808,
+      "learning_rate": 2.631279881849358e-05,
+      "loss": 10.0957,
+      "step": 25050
+    },
+    {
+      "epoch": 0.12514669529825964,
+      "grad_norm": 0.10160529613494873,
+      "learning_rate": 2.631129690355203e-05,
+      "loss": 10.098,
+      "step": 25060
+    },
+    {
+      "epoch": 0.1251966341231991,
+      "grad_norm": 0.0978497862815857,
+      "learning_rate": 2.630979498861048e-05,
+      "loss": 10.1,
+      "step": 25070
+    },
+    {
+      "epoch": 0.12524657294813854,
+      "grad_norm": 0.09045752882957458,
+      "learning_rate": 2.6308293073668926e-05,
+      "loss": 10.1027,
+      "step": 25080
+    },
+    {
+      "epoch": 0.125296511773078,
+      "grad_norm": 0.08881735056638718,
+      "learning_rate": 2.630679115872738e-05,
+      "loss": 10.0955,
+      "step": 25090
+    },
+    {
+      "epoch": 0.12534645059801744,
+      "grad_norm": 0.0939692035317421,
+      "learning_rate": 2.6305289243785827e-05,
+      "loss": 10.101,
+      "step": 25100
+    },
+    {
+      "epoch": 0.1253963894229569,
+      "grad_norm": 0.09511993080377579,
+      "learning_rate": 2.630378732884428e-05,
+      "loss": 10.0969,
+      "step": 25110
+    },
+    {
+      "epoch": 0.12544632824789634,
+      "grad_norm": 0.1041262075304985,
+      "learning_rate": 2.6302285413902727e-05,
+      "loss": 10.0912,
+      "step": 25120
+    },
+    {
+      "epoch": 0.1254962670728358,
+      "grad_norm": 0.09648961573839188,
+      "learning_rate": 2.6300783498961174e-05,
+      "loss": 10.0957,
+      "step": 25130
+    },
+    {
+      "epoch": 0.12554620589777524,
+      "grad_norm": 0.0973757728934288,
+      "learning_rate": 2.6299281584019627e-05,
+      "loss": 10.1005,
+      "step": 25140
+    },
+    {
+      "epoch": 0.1255961447227147,
+      "grad_norm": 0.08688199520111084,
+      "learning_rate": 2.6297779669078074e-05,
+      "loss": 10.0966,
+      "step": 25150
+    },
+    {
+      "epoch": 0.12564608354765414,
+      "grad_norm": 0.09321999549865723,
+      "learning_rate": 2.6296277754136528e-05,
+      "loss": 10.0924,
+      "step": 25160
+    },
+    {
+      "epoch": 0.12569602237259359,
+      "grad_norm": 0.09925235062837601,
+      "learning_rate": 2.6294775839194975e-05,
+      "loss": 10.0918,
+      "step": 25170
+    },
+    {
+      "epoch": 0.12574596119753303,
+      "grad_norm": 0.096058689057827,
+      "learning_rate": 2.629327392425342e-05,
+      "loss": 10.0918,
+      "step": 25180
+    },
+    {
+      "epoch": 0.12579590002247248,
+      "grad_norm": 0.09322872012853622,
+      "learning_rate": 2.6291772009311875e-05,
+      "loss": 10.0941,
+      "step": 25190
+    },
+    {
+      "epoch": 0.12584583884741193,
+      "grad_norm": 0.09035428613424301,
+      "learning_rate": 2.6290270094370322e-05,
+      "loss": 10.09,
+      "step": 25200
+    },
+    {
+      "epoch": 0.12589577767235138,
+      "grad_norm": 0.09396304190158844,
+      "learning_rate": 2.6288768179428775e-05,
+      "loss": 10.0977,
+      "step": 25210
+    },
+    {
+      "epoch": 0.1259457164972908,
+      "grad_norm": 0.09849023818969727,
+      "learning_rate": 2.6287266264487222e-05,
+      "loss": 10.0954,
+      "step": 25220
+    },
+    {
+      "epoch": 0.12599565532223025,
+      "grad_norm": 0.09524643421173096,
+      "learning_rate": 2.628576434954567e-05,
+      "loss": 10.0929,
+      "step": 25230
+    },
+    {
+      "epoch": 0.1260455941471697,
+      "grad_norm": 0.09641117602586746,
+      "learning_rate": 2.6284262434604122e-05,
+      "loss": 10.093,
+      "step": 25240
+    },
+    {
+      "epoch": 0.12609553297210915,
+      "grad_norm": 0.0981585755944252,
+      "learning_rate": 2.628276051966257e-05,
+      "loss": 10.0965,
+      "step": 25250
+    },
+    {
+      "epoch": 0.1261454717970486,
+      "grad_norm": 0.09640777111053467,
+      "learning_rate": 2.6281258604721023e-05,
+      "loss": 10.0929,
+      "step": 25260
+    },
+    {
+      "epoch": 0.12619541062198805,
+      "grad_norm": 0.0920519009232521,
+      "learning_rate": 2.627975668977947e-05,
+      "loss": 10.0947,
+      "step": 25270
+    },
+    {
+      "epoch": 0.1262453494469275,
+      "grad_norm": 0.09729231894016266,
+      "learning_rate": 2.6278254774837916e-05,
+      "loss": 10.0888,
+      "step": 25280
+    },
+    {
+      "epoch": 0.12629528827186695,
+      "grad_norm": 0.09746874123811722,
+      "learning_rate": 2.627675285989637e-05,
+      "loss": 10.0952,
+      "step": 25290
+    },
+    {
+      "epoch": 0.1263452270968064,
+      "grad_norm": 0.09574176371097565,
+      "learning_rate": 2.6275250944954817e-05,
+      "loss": 10.088,
+      "step": 25300
+    },
+    {
+      "epoch": 0.12639516592174585,
+      "grad_norm": 0.09351847320795059,
+      "learning_rate": 2.627374903001327e-05,
+      "loss": 10.0934,
+      "step": 25310
+    },
+    {
+      "epoch": 0.1264451047466853,
+      "grad_norm": 0.09245570003986359,
+      "learning_rate": 2.6272247115071717e-05,
+      "loss": 10.0963,
+      "step": 25320
+    },
+    {
+      "epoch": 0.12649504357162475,
+      "grad_norm": 0.0946178138256073,
+      "learning_rate": 2.6270745200130164e-05,
+      "loss": 10.0888,
+      "step": 25330
+    },
+    {
+      "epoch": 0.1265449823965642,
+      "grad_norm": 0.09379423409700394,
+      "learning_rate": 2.6269243285188617e-05,
+      "loss": 10.088,
+      "step": 25340
+    },
+    {
+      "epoch": 0.12659492122150365,
+      "grad_norm": 0.09336696565151215,
+      "learning_rate": 2.6267741370247064e-05,
+      "loss": 10.0913,
+      "step": 25350
+    },
+    {
+      "epoch": 0.1266448600464431,
+      "grad_norm": 0.09480159729719162,
+      "learning_rate": 2.6266239455305518e-05,
+      "loss": 10.0915,
+      "step": 25360
+    },
+    {
+      "epoch": 0.12669479887138255,
+      "grad_norm": 0.09639475494623184,
+      "learning_rate": 2.6264737540363965e-05,
+      "loss": 10.0904,
+      "step": 25370
+    },
+    {
+      "epoch": 0.126744737696322,
+      "grad_norm": 0.09714771062135696,
+      "learning_rate": 2.626323562542241e-05,
+      "loss": 10.0899,
+      "step": 25380
+    },
+    {
+      "epoch": 0.12679467652126145,
+      "grad_norm": 0.09559177607297897,
+      "learning_rate": 2.6261733710480865e-05,
+      "loss": 10.0885,
+      "step": 25390
+    },
+    {
+      "epoch": 0.1268446153462009,
+      "grad_norm": 0.09046468883752823,
+      "learning_rate": 2.6260231795539312e-05,
+      "loss": 10.0975,
+      "step": 25400
+    },
+    {
+      "epoch": 0.12689455417114034,
+      "grad_norm": 0.0968378558754921,
+      "learning_rate": 2.6258729880597765e-05,
+      "loss": 10.0879,
+      "step": 25410
+    },
+    {
+      "epoch": 0.1269444929960798,
+      "grad_norm": 0.09533467143774033,
+      "learning_rate": 2.6257227965656212e-05,
+      "loss": 10.0857,
+      "step": 25420
+    },
+    {
+      "epoch": 0.12699443182101924,
+      "grad_norm": 0.0917276069521904,
+      "learning_rate": 2.6255726050714662e-05,
+      "loss": 10.0927,
+      "step": 25430
+    },
+    {
+      "epoch": 0.1270443706459587,
+      "grad_norm": 0.096129409968853,
+      "learning_rate": 2.6254224135773113e-05,
+      "loss": 10.0886,
+      "step": 25440
+    },
+    {
+      "epoch": 0.12709430947089814,
+      "grad_norm": 0.09592961519956589,
+      "learning_rate": 2.625272222083156e-05,
+      "loss": 10.0869,
+      "step": 25450
+    },
+    {
+      "epoch": 0.1271442482958376,
+      "grad_norm": 0.1041722372174263,
+      "learning_rate": 2.6251220305890013e-05,
+      "loss": 10.0925,
+      "step": 25460
+    },
+    {
+      "epoch": 0.12719418712077704,
+      "grad_norm": 0.09002278000116348,
+      "learning_rate": 2.624971839094846e-05,
+      "loss": 10.0886,
+      "step": 25470
+    },
+    {
+      "epoch": 0.1272441259457165,
+      "grad_norm": 0.09098079055547714,
+      "learning_rate": 2.624821647600691e-05,
+      "loss": 10.0871,
+      "step": 25480
+    },
+    {
+      "epoch": 0.12729406477065594,
+      "grad_norm": 0.09563304483890533,
+      "learning_rate": 2.624671456106536e-05,
+      "loss": 10.0872,
+      "step": 25490
+    },
+    {
+      "epoch": 0.1273440035955954,
+      "grad_norm": 0.09505288302898407,
+      "learning_rate": 2.6245212646123807e-05,
+      "loss": 10.0871,
+      "step": 25500
+    },
+    {
+      "epoch": 0.12739394242053484,
+      "grad_norm": 0.09651105850934982,
+      "learning_rate": 2.624371073118226e-05,
+      "loss": 10.0902,
+      "step": 25510
+    },
+    {
+      "epoch": 0.1274438812454743,
+      "grad_norm": 0.09501788765192032,
+      "learning_rate": 2.6242208816240707e-05,
+      "loss": 10.0867,
+      "step": 25520
+    },
+    {
+      "epoch": 0.12749382007041374,
+      "grad_norm": 0.09667658805847168,
+      "learning_rate": 2.6240706901299157e-05,
+      "loss": 10.0871,
+      "step": 25530
+    },
+    {
+      "epoch": 0.1275437588953532,
+      "grad_norm": 0.09683676809072495,
+      "learning_rate": 2.6239204986357608e-05,
+      "loss": 10.0856,
+      "step": 25540
+    },
+    {
+      "epoch": 0.12759369772029264,
+      "grad_norm": 0.09272299706935883,
+      "learning_rate": 2.6237703071416054e-05,
+      "loss": 10.0899,
+      "step": 25550
+    },
+    {
+      "epoch": 0.1276436365452321,
+      "grad_norm": 0.08883216977119446,
+      "learning_rate": 2.6236201156474508e-05,
+      "loss": 10.0848,
+      "step": 25560
+    },
+    {
+      "epoch": 0.12769357537017154,
+      "grad_norm": 0.0918259397149086,
+      "learning_rate": 2.6234699241532955e-05,
+      "loss": 10.0828,
+      "step": 25570
+    },
+    {
+      "epoch": 0.12774351419511099,
+      "grad_norm": 0.0947398915886879,
+      "learning_rate": 2.6233197326591405e-05,
+      "loss": 10.0866,
+      "step": 25580
+    },
+    {
+      "epoch": 0.12779345302005043,
+      "grad_norm": 0.09740963578224182,
+      "learning_rate": 2.6231695411649855e-05,
+      "loss": 10.0829,
+      "step": 25590
+    },
+    {
+      "epoch": 0.12784339184498988,
+      "grad_norm": 0.09066878259181976,
+      "learning_rate": 2.6230193496708302e-05,
+      "loss": 10.0876,
+      "step": 25600
+    },
+    {
+      "epoch": 0.12789333066992933,
+      "grad_norm": 0.09469142556190491,
+      "learning_rate": 2.6228691581766755e-05,
+      "loss": 10.0785,
+      "step": 25610
+    },
+    {
+      "epoch": 0.12794326949486878,
+      "grad_norm": 0.09159637987613678,
+      "learning_rate": 2.6227189666825202e-05,
+      "loss": 10.0811,
+      "step": 25620
+    },
+    {
+      "epoch": 0.12799320831980823,
+      "grad_norm": 0.09090546518564224,
+      "learning_rate": 2.6225687751883652e-05,
+      "loss": 10.0842,
+      "step": 25630
+    },
+    {
+      "epoch": 0.12804314714474768,
+      "grad_norm": 0.0922180786728859,
+      "learning_rate": 2.6224185836942103e-05,
+      "loss": 10.0828,
+      "step": 25640
+    },
+    {
+      "epoch": 0.12809308596968713,
+      "grad_norm": 0.09305445849895477,
+      "learning_rate": 2.622268392200055e-05,
+      "loss": 10.0872,
+      "step": 25650
+    },
+    {
+      "epoch": 0.12814302479462658,
+      "grad_norm": 0.09471817314624786,
+      "learning_rate": 2.6221182007059003e-05,
+      "loss": 10.0857,
+      "step": 25660
+    },
+    {
+      "epoch": 0.12819296361956603,
+      "grad_norm": 0.09348314255475998,
+      "learning_rate": 2.621968009211745e-05,
+      "loss": 10.0884,
+      "step": 25670
+    },
+    {
+      "epoch": 0.12824290244450548,
+      "grad_norm": 0.09191960841417313,
+      "learning_rate": 2.62181781771759e-05,
+      "loss": 10.0895,
+      "step": 25680
+    },
+    {
+      "epoch": 0.12829284126944493,
+      "grad_norm": 0.09789571166038513,
+      "learning_rate": 2.621667626223435e-05,
+      "loss": 10.0818,
+      "step": 25690
+    },
+    {
+      "epoch": 0.12834278009438438,
+      "grad_norm": 0.09516225755214691,
+      "learning_rate": 2.6215174347292797e-05,
+      "loss": 10.0831,
+      "step": 25700
+    },
+    {
+      "epoch": 0.12839271891932383,
+      "grad_norm": 0.09300491213798523,
+      "learning_rate": 2.621367243235125e-05,
+      "loss": 10.0844,
+      "step": 25710
+    },
+    {
+      "epoch": 0.12844265774426328,
+      "grad_norm": 0.09599104523658752,
+      "learning_rate": 2.6212170517409697e-05,
+      "loss": 10.0853,
+      "step": 25720
+    },
+    {
+      "epoch": 0.12849259656920273,
+      "grad_norm": 0.09136704355478287,
+      "learning_rate": 2.6210668602468147e-05,
+      "loss": 10.0873,
+      "step": 25730
+    },
+    {
+      "epoch": 0.12854253539414218,
+      "grad_norm": 0.09239959716796875,
+      "learning_rate": 2.6209166687526598e-05,
+      "loss": 10.0794,
+      "step": 25740
+    },
+    {
+      "epoch": 0.12859247421908163,
+      "grad_norm": 0.09590020775794983,
+      "learning_rate": 2.6207664772585048e-05,
+      "loss": 10.0835,
+      "step": 25750
+    },
+    {
+      "epoch": 0.12864241304402108,
+      "grad_norm": 0.09230869263410568,
+      "learning_rate": 2.6206162857643498e-05,
+      "loss": 10.0833,
+      "step": 25760
+    },
+    {
+      "epoch": 0.12869235186896053,
+      "grad_norm": 0.0960649698972702,
+      "learning_rate": 2.6204660942701945e-05,
+      "loss": 10.0822,
+      "step": 25770
+    },
+    {
+      "epoch": 0.12874229069389997,
+      "grad_norm": 0.09332367777824402,
+      "learning_rate": 2.6203159027760395e-05,
+      "loss": 10.0848,
+      "step": 25780
+    },
+    {
+      "epoch": 0.12879222951883942,
+      "grad_norm": 0.09813748300075531,
+      "learning_rate": 2.6201657112818845e-05,
+      "loss": 10.0807,
+      "step": 25790
+    },
+    {
+      "epoch": 0.12884216834377887,
+      "grad_norm": 0.09281601011753082,
+      "learning_rate": 2.6200155197877295e-05,
+      "loss": 10.0862,
+      "step": 25800
+    },
+    {
+      "epoch": 0.12889210716871832,
+      "grad_norm": 0.09574627131223679,
+      "learning_rate": 2.6198653282935745e-05,
+      "loss": 10.0784,
+      "step": 25810
+    },
+    {
+      "epoch": 0.12894204599365777,
+      "grad_norm": 0.0939335897564888,
+      "learning_rate": 2.6197151367994192e-05,
+      "loss": 10.0865,
+      "step": 25820
+    },
+    {
+      "epoch": 0.12899198481859722,
+      "grad_norm": 0.0975639671087265,
+      "learning_rate": 2.6195649453052642e-05,
+      "loss": 10.0783,
+      "step": 25830
+    },
+    {
+      "epoch": 0.12904192364353667,
+      "grad_norm": 0.0939013734459877,
+      "learning_rate": 2.6194147538111093e-05,
+      "loss": 10.0837,
+      "step": 25840
+    },
+    {
+      "epoch": 0.12909186246847612,
+      "grad_norm": 0.09316052496433258,
+      "learning_rate": 2.6192645623169543e-05,
+      "loss": 10.0781,
+      "step": 25850
+    },
+    {
+      "epoch": 0.12914180129341557,
+      "grad_norm": 0.09889581054449081,
+      "learning_rate": 2.6191143708227993e-05,
+      "loss": 10.0742,
+      "step": 25860
+    },
+    {
+      "epoch": 0.12919174011835502,
+      "grad_norm": 0.09303954988718033,
+      "learning_rate": 2.618964179328644e-05,
+      "loss": 10.0812,
+      "step": 25870
+    },
+    {
+      "epoch": 0.12924167894329447,
+      "grad_norm": 0.09554919600486755,
+      "learning_rate": 2.618813987834489e-05,
+      "loss": 10.0706,
+      "step": 25880
+    },
+    {
+      "epoch": 0.12929161776823392,
+      "grad_norm": 0.09778593480587006,
+      "learning_rate": 2.618663796340334e-05,
+      "loss": 10.0771,
+      "step": 25890
+    },
+    {
+      "epoch": 0.12934155659317337,
+      "grad_norm": 0.09445203095674515,
+      "learning_rate": 2.618513604846179e-05,
+      "loss": 10.0793,
+      "step": 25900
+    },
+    {
+      "epoch": 0.12939149541811282,
+      "grad_norm": 0.09079139679670334,
+      "learning_rate": 2.618363413352024e-05,
+      "loss": 10.0755,
+      "step": 25910
+    },
+    {
+      "epoch": 0.12944143424305227,
+      "grad_norm": 5018.11767578125,
+      "learning_rate": 2.6182132218578687e-05,
+      "loss": 10.1051,
+      "step": 25920
+    },
+    {
+      "epoch": 0.12949137306799172,
+      "grad_norm": 0.09410411864519119,
+      "learning_rate": 2.6180630303637137e-05,
+      "loss": 10.0779,
+      "step": 25930
+    },
+    {
+      "epoch": 0.12954131189293117,
+      "grad_norm": 0.09051043540239334,
+      "learning_rate": 2.6179128388695588e-05,
+      "loss": 10.0751,
+      "step": 25940
+    },
+    {
+      "epoch": 0.12959125071787062,
+      "grad_norm": 0.09015491604804993,
+      "learning_rate": 2.6177626473754038e-05,
+      "loss": 10.0803,
+      "step": 25950
+    },
+    {
+      "epoch": 0.12964118954281006,
+      "grad_norm": 0.09758006781339645,
+      "learning_rate": 2.6176124558812488e-05,
+      "loss": 10.0734,
+      "step": 25960
+    },
+    {
+      "epoch": 0.12969112836774951,
+      "grad_norm": 0.0914853885769844,
+      "learning_rate": 2.6174622643870935e-05,
+      "loss": 10.0797,
+      "step": 25970
+    },
+    {
+      "epoch": 0.12974106719268896,
+      "grad_norm": 0.09737791121006012,
+      "learning_rate": 2.6173120728929385e-05,
+      "loss": 10.0749,
+      "step": 25980
+    },
+    {
+      "epoch": 0.1297910060176284,
+      "grad_norm": 0.09865185618400574,
+      "learning_rate": 2.6171618813987835e-05,
+      "loss": 10.0764,
+      "step": 25990
+    },
+    {
+      "epoch": 0.12984094484256786,
+      "grad_norm": 0.0985475480556488,
+      "learning_rate": 2.6170116899046285e-05,
+      "loss": 10.075,
+      "step": 26000
+    },
+    {
+      "epoch": 0.1298908836675073,
+      "grad_norm": 0.09121818840503693,
+      "learning_rate": 2.6168614984104735e-05,
+      "loss": 10.0797,
+      "step": 26010
+    },
+    {
+      "epoch": 0.12994082249244676,
+      "grad_norm": 0.10556277632713318,
+      "learning_rate": 2.6167113069163182e-05,
+      "loss": 10.0783,
+      "step": 26020
+    },
+    {
+      "epoch": 0.1299907613173862,
+      "grad_norm": 0.0959169790148735,
+      "learning_rate": 2.6165611154221632e-05,
+      "loss": 10.0753,
+      "step": 26030
+    },
+    {
+      "epoch": 0.13004070014232566,
+      "grad_norm": 0.09328239411115646,
+      "learning_rate": 2.6164109239280083e-05,
+      "loss": 10.0748,
+      "step": 26040
+    },
+    {
+      "epoch": 0.1300906389672651,
+      "grad_norm": 0.08964414894580841,
+      "learning_rate": 2.6162607324338533e-05,
+      "loss": 10.075,
+      "step": 26050
+    },
+    {
+      "epoch": 0.13014057779220456,
+      "grad_norm": 0.1068107932806015,
+      "learning_rate": 2.6161105409396983e-05,
+      "loss": 10.0722,
+      "step": 26060
+    },
+    {
+      "epoch": 0.130190516617144,
+      "grad_norm": 0.09689830243587494,
+      "learning_rate": 2.6159603494455433e-05,
+      "loss": 10.0727,
+      "step": 26070
+    },
+    {
+      "epoch": 0.13024045544208346,
+      "grad_norm": 0.09589334577322006,
+      "learning_rate": 2.615810157951388e-05,
+      "loss": 10.0663,
+      "step": 26080
+    },
+    {
+      "epoch": 0.1302903942670229,
+      "grad_norm": 0.09473288059234619,
+      "learning_rate": 2.615659966457233e-05,
+      "loss": 10.0675,
+      "step": 26090
+    },
+    {
+      "epoch": 0.13034033309196236,
+      "grad_norm": 0.09076739102602005,
+      "learning_rate": 2.615509774963078e-05,
+      "loss": 10.0772,
+      "step": 26100
+    },
+    {
+      "epoch": 0.1303902719169018,
+      "grad_norm": 0.09733912348747253,
+      "learning_rate": 2.615359583468923e-05,
+      "loss": 10.0746,
+      "step": 26110
+    },
+    {
+      "epoch": 0.13044021074184126,
+      "grad_norm": 0.10011307895183563,
+      "learning_rate": 2.615209391974768e-05,
+      "loss": 10.0657,
+      "step": 26120
+    },
+    {
+      "epoch": 0.1304901495667807,
+      "grad_norm": 0.09807174652814865,
+      "learning_rate": 2.6150592004806127e-05,
+      "loss": 10.0688,
+      "step": 26130
+    },
+    {
+      "epoch": 0.13054008839172015,
+      "grad_norm": 0.092112235724926,
+      "learning_rate": 2.6149090089864578e-05,
+      "loss": 10.0723,
+      "step": 26140
+    },
+    {
+      "epoch": 0.1305900272166596,
+      "grad_norm": 0.09255208820104599,
+      "learning_rate": 2.6147588174923028e-05,
+      "loss": 10.0682,
+      "step": 26150
+    },
+    {
+      "epoch": 0.13063996604159905,
+      "grad_norm": 0.09178891032934189,
+      "learning_rate": 2.6146086259981478e-05,
+      "loss": 10.0694,
+      "step": 26160
+    },
+    {
+      "epoch": 0.1306899048665385,
+      "grad_norm": 0.09472060203552246,
+      "learning_rate": 2.6144584345039928e-05,
+      "loss": 10.0669,
+      "step": 26170
+    },
+    {
+      "epoch": 0.13073984369147795,
+      "grad_norm": 0.08864433318376541,
+      "learning_rate": 2.6143082430098375e-05,
+      "loss": 10.0716,
+      "step": 26180
+    },
+    {
+      "epoch": 0.1307897825164174,
+      "grad_norm": 0.09347887337207794,
+      "learning_rate": 2.6141580515156825e-05,
+      "loss": 10.0755,
+      "step": 26190
+    },
+    {
+      "epoch": 0.13083972134135685,
+      "grad_norm": 0.09428052604198456,
+      "learning_rate": 2.6140078600215275e-05,
+      "loss": 10.0727,
+      "step": 26200
+    },
+    {
+      "epoch": 0.13088966016629627,
+      "grad_norm": 0.09582565724849701,
+      "learning_rate": 2.6138576685273725e-05,
+      "loss": 10.0665,
+      "step": 26210
+    },
+    {
+      "epoch": 0.13093959899123572,
+      "grad_norm": 0.09516657888889313,
+      "learning_rate": 2.6137074770332176e-05,
+      "loss": 10.0645,
+      "step": 26220
+    },
+    {
+      "epoch": 0.13098953781617517,
+      "grad_norm": 0.10089933127164841,
+      "learning_rate": 2.6135572855390622e-05,
+      "loss": 10.0687,
+      "step": 26230
+    },
+    {
+      "epoch": 0.13103947664111462,
+      "grad_norm": 0.09348998963832855,
+      "learning_rate": 2.6134070940449073e-05,
+      "loss": 10.0749,
+      "step": 26240
+    },
+    {
+      "epoch": 0.13108941546605407,
+      "grad_norm": 0.09638164937496185,
+      "learning_rate": 2.6132569025507523e-05,
+      "loss": 10.0666,
+      "step": 26250
+    },
+    {
+      "epoch": 0.13113935429099352,
+      "grad_norm": 0.08938256651163101,
+      "learning_rate": 2.6131067110565973e-05,
+      "loss": 10.0736,
+      "step": 26260
+    },
+    {
+      "epoch": 0.13118929311593297,
+      "grad_norm": 0.09375530481338501,
+      "learning_rate": 2.6129565195624423e-05,
+      "loss": 10.0723,
+      "step": 26270
+    },
+    {
+      "epoch": 0.13123923194087242,
+      "grad_norm": 0.09841980785131454,
+      "learning_rate": 2.612806328068287e-05,
+      "loss": 10.0654,
+      "step": 26280
+    },
+    {
+      "epoch": 0.13128917076581187,
+      "grad_norm": 0.0966048538684845,
+      "learning_rate": 2.612656136574132e-05,
+      "loss": 10.0719,
+      "step": 26290
+    },
+    {
+      "epoch": 0.13133910959075132,
+      "grad_norm": 0.09200683981180191,
+      "learning_rate": 2.612505945079977e-05,
+      "loss": 10.072,
+      "step": 26300
+    },
+    {
+      "epoch": 0.13138904841569077,
+      "grad_norm": 0.09424161165952682,
+      "learning_rate": 2.612355753585822e-05,
+      "loss": 10.067,
+      "step": 26310
+    },
+    {
+      "epoch": 0.13143898724063022,
+      "grad_norm": 0.09710439294576645,
+      "learning_rate": 2.612205562091667e-05,
+      "loss": 10.0681,
+      "step": 26320
+    },
+    {
+      "epoch": 0.13148892606556967,
+      "grad_norm": 0.09168367087841034,
+      "learning_rate": 2.6120553705975117e-05,
+      "loss": 10.0653,
+      "step": 26330
+    },
+    {
+      "epoch": 0.13153886489050912,
+      "grad_norm": 0.09984847903251648,
+      "learning_rate": 2.6119051791033568e-05,
+      "loss": 10.0658,
+      "step": 26340
+    },
+    {
+      "epoch": 0.13158880371544857,
+      "grad_norm": 0.09909839183092117,
+      "learning_rate": 2.6117549876092018e-05,
+      "loss": 10.0639,
+      "step": 26350
+    },
+    {
+      "epoch": 0.13163874254038802,
+      "grad_norm": 0.09866788238286972,
+      "learning_rate": 2.6116047961150468e-05,
+      "loss": 10.0642,
+      "step": 26360
+    },
+    {
+      "epoch": 0.13168868136532746,
+      "grad_norm": 0.0965615063905716,
+      "learning_rate": 2.6114546046208918e-05,
+      "loss": 10.0713,
+      "step": 26370
+    },
+    {
+      "epoch": 0.13173862019026691,
+      "grad_norm": 0.09519165009260178,
+      "learning_rate": 2.6113044131267365e-05,
+      "loss": 10.0636,
+      "step": 26380
+    },
+    {
+      "epoch": 0.13178855901520636,
+      "grad_norm": 0.10046073794364929,
+      "learning_rate": 2.611154221632582e-05,
+      "loss": 10.0641,
+      "step": 26390
+    },
+    {
+      "epoch": 0.1318384978401458,
+      "grad_norm": 0.09428393840789795,
+      "learning_rate": 2.6110040301384265e-05,
+      "loss": 10.0645,
+      "step": 26400
+    },
+    {
+      "epoch": 0.13188843666508526,
+      "grad_norm": 0.09596385806798935,
+      "learning_rate": 2.6108538386442715e-05,
+      "loss": 10.0841,
+      "step": 26410
+    },
+    {
+      "epoch": 0.1319383754900247,
+      "grad_norm": 0.0943203791975975,
+      "learning_rate": 2.6107036471501166e-05,
+      "loss": 10.0695,
+      "step": 26420
+    },
+    {
+      "epoch": 0.13198831431496416,
+      "grad_norm": 0.0989902913570404,
+      "learning_rate": 2.6105534556559612e-05,
+      "loss": 10.0657,
+      "step": 26430
+    },
+    {
+      "epoch": 0.1320382531399036,
+      "grad_norm": 0.0922727957367897,
+      "learning_rate": 2.6104032641618066e-05,
+      "loss": 10.5972,
+      "step": 26440
+    },
+    {
+      "epoch": 0.13208819196484306,
+      "grad_norm": 0.09232750535011292,
+      "learning_rate": 2.6102530726676513e-05,
+      "loss": 10.063,
+      "step": 26450
+    },
+    {
+      "epoch": 0.1321381307897825,
+      "grad_norm": 0.09481442719697952,
+      "learning_rate": 2.6101028811734963e-05,
+      "loss": 10.0732,
+      "step": 26460
+    },
+    {
+      "epoch": 0.13218806961472196,
+      "grad_norm": 0.09335809201002121,
+      "learning_rate": 2.6099526896793413e-05,
+      "loss": 10.0669,
+      "step": 26470
+    },
+    {
+      "epoch": 0.1322380084396614,
+      "grad_norm": 0.10007120668888092,
+      "learning_rate": 2.609802498185186e-05,
+      "loss": 10.0633,
+      "step": 26480
+    },
+    {
+      "epoch": 0.13228794726460086,
+      "grad_norm": 0.09182542562484741,
+      "learning_rate": 2.6096523066910313e-05,
+      "loss": 10.0626,
+      "step": 26490
+    },
+    {
+      "epoch": 0.1323378860895403,
+      "grad_norm": 0.08880601823329926,
+      "learning_rate": 2.609502115196876e-05,
+      "loss": 10.0639,
+      "step": 26500
+    },
+    {
+      "epoch": 0.13238782491447976,
+      "grad_norm": 0.09500660002231598,
+      "learning_rate": 2.609351923702721e-05,
+      "loss": 10.0594,
+      "step": 26510
+    },
+    {
+      "epoch": 0.1324377637394192,
+      "grad_norm": 0.09459840506315231,
+      "learning_rate": 2.609201732208566e-05,
+      "loss": 10.0689,
+      "step": 26520
+    },
+    {
+      "epoch": 0.13248770256435866,
+      "grad_norm": 0.09772436320781708,
+      "learning_rate": 2.6090515407144107e-05,
+      "loss": 10.0646,
+      "step": 26530
+    },
+    {
+      "epoch": 0.1325376413892981,
+      "grad_norm": 0.09241725504398346,
+      "learning_rate": 2.608901349220256e-05,
+      "loss": 10.0626,
+      "step": 26540
+    },
+    {
+      "epoch": 0.13258758021423755,
+      "grad_norm": 0.09605549275875092,
+      "learning_rate": 2.6087511577261008e-05,
+      "loss": 10.062,
+      "step": 26550
+    },
+    {
+      "epoch": 0.132637519039177,
+      "grad_norm": 0.08990667015314102,
+      "learning_rate": 2.6086009662319458e-05,
+      "loss": 10.0607,
+      "step": 26560
+    },
+    {
+      "epoch": 0.13268745786411645,
+      "grad_norm": 0.09539429098367691,
+      "learning_rate": 2.6084507747377908e-05,
+      "loss": 10.0607,
+      "step": 26570
+    },
+    {
+      "epoch": 0.1327373966890559,
+      "grad_norm": 0.09879159927368164,
+      "learning_rate": 2.6083005832436355e-05,
+      "loss": 10.059,
+      "step": 26580
+    },
+    {
+      "epoch": 0.13278733551399535,
+      "grad_norm": 0.09552617371082306,
+      "learning_rate": 2.608150391749481e-05,
+      "loss": 10.0612,
+      "step": 26590
+    },
+    {
+      "epoch": 0.1328372743389348,
+      "grad_norm": 0.09348921477794647,
+      "learning_rate": 2.6080002002553255e-05,
+      "loss": 10.062,
+      "step": 26600
+    },
+    {
+      "epoch": 0.13288721316387425,
+      "grad_norm": 0.0956648588180542,
+      "learning_rate": 2.6078500087611705e-05,
+      "loss": 10.0591,
+      "step": 26610
+    },
+    {
+      "epoch": 0.1329371519888137,
+      "grad_norm": 0.09512264281511307,
+      "learning_rate": 2.6076998172670156e-05,
+      "loss": 10.0588,
+      "step": 26620
+    },
+    {
+      "epoch": 0.13298709081375315,
+      "grad_norm": 0.09354131668806076,
+      "learning_rate": 2.6075496257728602e-05,
+      "loss": 10.0635,
+      "step": 26630
+    },
+    {
+      "epoch": 0.1330370296386926,
+      "grad_norm": 0.09954062849283218,
+      "learning_rate": 2.6073994342787056e-05,
+      "loss": 10.0639,
+      "step": 26640
+    },
+    {
+      "epoch": 0.13308696846363205,
+      "grad_norm": 0.09436924755573273,
+      "learning_rate": 2.6072492427845503e-05,
+      "loss": 10.0608,
+      "step": 26650
+    },
+    {
+      "epoch": 0.1331369072885715,
+      "grad_norm": 0.09259260445833206,
+      "learning_rate": 2.6070990512903953e-05,
+      "loss": 10.0609,
+      "step": 26660
+    },
+    {
+      "epoch": 0.13318684611351095,
+      "grad_norm": 0.09337363392114639,
+      "learning_rate": 2.6069488597962403e-05,
+      "loss": 10.0629,
+      "step": 26670
+    },
+    {
+      "epoch": 0.1332367849384504,
+      "grad_norm": 0.09187287837266922,
+      "learning_rate": 2.606798668302085e-05,
+      "loss": 10.0571,
+      "step": 26680
+    },
+    {
+      "epoch": 0.13328672376338985,
+      "grad_norm": 0.09519325196743011,
+      "learning_rate": 2.6066484768079303e-05,
+      "loss": 10.063,
+      "step": 26690
+    },
+    {
+      "epoch": 0.1333366625883293,
+      "grad_norm": 0.09201279282569885,
+      "learning_rate": 2.606498285313775e-05,
+      "loss": 10.0609,
+      "step": 26700
+    },
+    {
+      "epoch": 0.13338660141326875,
+      "grad_norm": 0.09440702199935913,
+      "learning_rate": 2.6063480938196204e-05,
+      "loss": 10.0543,
+      "step": 26710
+    },
+    {
+      "epoch": 0.1334365402382082,
+      "grad_norm": 0.09421040117740631,
+      "learning_rate": 2.606197902325465e-05,
+      "loss": 10.0605,
+      "step": 26720
+    },
+    {
+      "epoch": 0.13348647906314764,
+      "grad_norm": 0.09064962714910507,
+      "learning_rate": 2.6060477108313097e-05,
+      "loss": 10.0651,
+      "step": 26730
+    },
+    {
+      "epoch": 0.1335364178880871,
+      "grad_norm": 0.10387849062681198,
+      "learning_rate": 2.605897519337155e-05,
+      "loss": 10.0539,
+      "step": 26740
+    },
+    {
+      "epoch": 0.13358635671302654,
+      "grad_norm": 0.09637302160263062,
+      "learning_rate": 2.6057473278429998e-05,
+      "loss": 10.0546,
+      "step": 26750
+    },
+    {
+      "epoch": 0.133636295537966,
+      "grad_norm": 0.09184058010578156,
+      "learning_rate": 2.605597136348845e-05,
+      "loss": 10.0554,
+      "step": 26760
+    },
+    {
+      "epoch": 0.13368623436290544,
+      "grad_norm": 0.09594525396823883,
+      "learning_rate": 2.6054469448546898e-05,
+      "loss": 10.0525,
+      "step": 26770
+    },
+    {
+      "epoch": 0.1337361731878449,
+      "grad_norm": 0.09645157307386398,
+      "learning_rate": 2.6052967533605345e-05,
+      "loss": 10.0595,
+      "step": 26780
+    },
+    {
+      "epoch": 0.13378611201278434,
+      "grad_norm": 0.09152582287788391,
+      "learning_rate": 2.60514656186638e-05,
+      "loss": 10.0652,
+      "step": 26790
+    },
+    {
+      "epoch": 0.1338360508377238,
+      "grad_norm": 0.09593851864337921,
+      "learning_rate": 2.6049963703722245e-05,
+      "loss": 10.0514,
+      "step": 26800
+    },
+    {
+      "epoch": 0.13388598966266324,
+      "grad_norm": 0.09910731017589569,
+      "learning_rate": 2.60484617887807e-05,
+      "loss": 10.0577,
+      "step": 26810
+    },
+    {
+      "epoch": 0.1339359284876027,
+      "grad_norm": 0.09356366842985153,
+      "learning_rate": 2.6046959873839146e-05,
+      "loss": 10.0535,
+      "step": 26820
+    },
+    {
+      "epoch": 0.13398586731254214,
+      "grad_norm": 0.09344334155321121,
+      "learning_rate": 2.6045457958897592e-05,
+      "loss": 10.0619,
+      "step": 26830
+    },
+    {
+      "epoch": 0.1340358061374816,
+      "grad_norm": 0.09502086043357849,
+      "learning_rate": 2.6043956043956046e-05,
+      "loss": 10.0514,
+      "step": 26840
+    },
+    {
+      "epoch": 0.13408574496242104,
+      "grad_norm": 0.0968647301197052,
+      "learning_rate": 2.6042454129014493e-05,
+      "loss": 10.0599,
+      "step": 26850
+    },
+    {
+      "epoch": 0.1341356837873605,
+      "grad_norm": 0.09345351159572601,
+      "learning_rate": 2.6040952214072946e-05,
+      "loss": 10.0544,
+      "step": 26860
+    },
+    {
+      "epoch": 0.13418562261229994,
+      "grad_norm": 0.09822539240121841,
+      "learning_rate": 2.6039450299131393e-05,
+      "loss": 10.0519,
+      "step": 26870
+    },
+    {
+      "epoch": 0.1342355614372394,
+      "grad_norm": 0.09218066930770874,
+      "learning_rate": 2.603794838418984e-05,
+      "loss": 10.0573,
+      "step": 26880
+    },
+    {
+      "epoch": 0.13428550026217884,
+      "grad_norm": 0.09610606729984283,
+      "learning_rate": 2.6036446469248293e-05,
+      "loss": 10.0547,
+      "step": 26890
+    },
+    {
+      "epoch": 0.13433543908711829,
+      "grad_norm": 0.09326432645320892,
+      "learning_rate": 2.603494455430674e-05,
+      "loss": 10.0527,
+      "step": 26900
+    },
+    {
+      "epoch": 0.13438537791205774,
+      "grad_norm": 0.0933103859424591,
+      "learning_rate": 2.6033442639365194e-05,
+      "loss": 10.0503,
+      "step": 26910
+    },
+    {
+      "epoch": 0.13443531673699718,
+      "grad_norm": 0.09903916716575623,
+      "learning_rate": 2.603194072442364e-05,
+      "loss": 10.0514,
+      "step": 26920
+    },
+    {
+      "epoch": 0.13448525556193663,
+      "grad_norm": 0.10167834907770157,
+      "learning_rate": 2.6030438809482087e-05,
+      "loss": 10.0525,
+      "step": 26930
+    },
+    {
+      "epoch": 0.13453519438687608,
+      "grad_norm": 0.09627428650856018,
+      "learning_rate": 2.602893689454054e-05,
+      "loss": 10.0544,
+      "step": 26940
+    },
+    {
+      "epoch": 0.13458513321181553,
+      "grad_norm": 0.09547838568687439,
+      "learning_rate": 2.6027434979598988e-05,
+      "loss": 10.0523,
+      "step": 26950
+    },
+    {
+      "epoch": 0.13463507203675498,
+      "grad_norm": 0.0934177041053772,
+      "learning_rate": 2.602593306465744e-05,
+      "loss": 10.0577,
+      "step": 26960
+    },
+    {
+      "epoch": 0.13468501086169443,
+      "grad_norm": 0.09201879799365997,
+      "learning_rate": 2.6024431149715888e-05,
+      "loss": 10.0561,
+      "step": 26970
+    },
+    {
+      "epoch": 0.13473494968663388,
+      "grad_norm": 0.10229359567165375,
+      "learning_rate": 2.6022929234774335e-05,
+      "loss": 10.0537,
+      "step": 26980
+    },
+    {
+      "epoch": 0.13478488851157333,
+      "grad_norm": 0.0951157957315445,
+      "learning_rate": 2.602142731983279e-05,
+      "loss": 10.0489,
+      "step": 26990
+    },
+    {
+      "epoch": 0.13483482733651278,
+      "grad_norm": 0.09461617469787598,
+      "learning_rate": 2.6019925404891235e-05,
+      "loss": 10.0504,
+      "step": 27000
+    },
+    {
+      "epoch": 0.13488476616145223,
+      "grad_norm": 0.09404211491346359,
+      "learning_rate": 2.601842348994969e-05,
+      "loss": 10.05,
+      "step": 27010
+    },
+    {
+      "epoch": 0.13493470498639168,
+      "grad_norm": 0.09388327598571777,
+      "learning_rate": 2.6016921575008136e-05,
+      "loss": 10.0507,
+      "step": 27020
+    },
+    {
+      "epoch": 0.13498464381133113,
+      "grad_norm": 0.08963081240653992,
+      "learning_rate": 2.6015419660066582e-05,
+      "loss": 10.0509,
+      "step": 27030
+    },
+    {
+      "epoch": 0.13503458263627058,
+      "grad_norm": 0.09284106642007828,
+      "learning_rate": 2.6013917745125036e-05,
+      "loss": 10.0477,
+      "step": 27040
+    },
+    {
+      "epoch": 0.13508452146121003,
+      "grad_norm": 0.09641976654529572,
+      "learning_rate": 2.6012415830183483e-05,
+      "loss": 10.0484,
+      "step": 27050
+    },
+    {
+      "epoch": 0.13513446028614948,
+      "grad_norm": 0.09596869349479675,
+      "learning_rate": 2.6010913915241936e-05,
+      "loss": 10.0532,
+      "step": 27060
+    },
+    {
+      "epoch": 0.13518439911108893,
+      "grad_norm": 0.09437578171491623,
+      "learning_rate": 2.6009412000300383e-05,
+      "loss": 10.0535,
+      "step": 27070
+    },
+    {
+      "epoch": 0.13523433793602838,
+      "grad_norm": 0.09599731862545013,
+      "learning_rate": 2.6007910085358833e-05,
+      "loss": 10.0469,
+      "step": 27080
+    },
+    {
+      "epoch": 0.13528427676096783,
+      "grad_norm": 0.09609924256801605,
+      "learning_rate": 2.6006408170417284e-05,
+      "loss": 10.0484,
+      "step": 27090
+    },
+    {
+      "epoch": 0.13533421558590727,
+      "grad_norm": 0.09731438755989075,
+      "learning_rate": 2.600490625547573e-05,
+      "loss": 10.0502,
+      "step": 27100
+    },
+    {
+      "epoch": 0.13538415441084672,
+      "grad_norm": 0.09791937470436096,
+      "learning_rate": 2.6003404340534184e-05,
+      "loss": 10.0518,
+      "step": 27110
+    },
+    {
+      "epoch": 0.13543409323578617,
+      "grad_norm": 0.10084173083305359,
+      "learning_rate": 2.600190242559263e-05,
+      "loss": 10.0464,
+      "step": 27120
+    },
+    {
+      "epoch": 0.13548403206072562,
+      "grad_norm": 0.09352555871009827,
+      "learning_rate": 2.600040051065108e-05,
+      "loss": 10.0546,
+      "step": 27130
+    },
+    {
+      "epoch": 0.13553397088566507,
+      "grad_norm": 0.09164499491453171,
+      "learning_rate": 2.599889859570953e-05,
+      "loss": 10.05,
+      "step": 27140
+    },
+    {
+      "epoch": 0.13558390971060452,
+      "grad_norm": 0.09976156800985336,
+      "learning_rate": 2.5997396680767978e-05,
+      "loss": 10.0509,
+      "step": 27150
+    },
+    {
+      "epoch": 0.13563384853554397,
+      "grad_norm": 0.09886104613542557,
+      "learning_rate": 2.599589476582643e-05,
+      "loss": 10.0528,
+      "step": 27160
+    },
+    {
+      "epoch": 0.13568378736048342,
+      "grad_norm": 0.09629525989294052,
+      "learning_rate": 2.5994392850884878e-05,
+      "loss": 10.0424,
+      "step": 27170
+    },
+    {
+      "epoch": 0.13573372618542287,
+      "grad_norm": 0.09974990785121918,
+      "learning_rate": 2.599289093594333e-05,
+      "loss": 10.0463,
+      "step": 27180
+    },
+    {
+      "epoch": 0.13578366501036232,
+      "grad_norm": 0.09218470752239227,
+      "learning_rate": 2.599138902100178e-05,
+      "loss": 10.0519,
+      "step": 27190
+    },
+    {
+      "epoch": 0.13583360383530174,
+      "grad_norm": 0.09512577950954437,
+      "learning_rate": 2.5989887106060225e-05,
+      "loss": 10.0465,
+      "step": 27200
+    },
+    {
+      "epoch": 0.1358835426602412,
+      "grad_norm": 0.09620299190282822,
+      "learning_rate": 2.598838519111868e-05,
+      "loss": 10.0529,
+      "step": 27210
+    },
+    {
+      "epoch": 0.13593348148518064,
+      "grad_norm": 0.09048976749181747,
+      "learning_rate": 2.5986883276177126e-05,
+      "loss": 10.0434,
+      "step": 27220
+    },
+    {
+      "epoch": 0.1359834203101201,
+      "grad_norm": 0.09220554679632187,
+      "learning_rate": 2.5985381361235576e-05,
+      "loss": 10.049,
+      "step": 27230
+    },
+    {
+      "epoch": 0.13603335913505954,
+      "grad_norm": 0.0914304107427597,
+      "learning_rate": 2.5983879446294026e-05,
+      "loss": 10.0438,
+      "step": 27240
+    },
+    {
+      "epoch": 0.136083297959999,
+      "grad_norm": 0.09436330199241638,
+      "learning_rate": 2.5982377531352473e-05,
+      "loss": 10.047,
+      "step": 27250
+    },
+    {
+      "epoch": 0.13613323678493844,
+      "grad_norm": 0.09602109342813492,
+      "learning_rate": 2.5980875616410926e-05,
+      "loss": 10.0436,
+      "step": 27260
+    },
+    {
+      "epoch": 0.1361831756098779,
+      "grad_norm": 0.09520605206489563,
+      "learning_rate": 2.5979373701469373e-05,
+      "loss": 10.0481,
+      "step": 27270
+    },
+    {
+      "epoch": 0.13623311443481734,
+      "grad_norm": 0.09378485381603241,
+      "learning_rate": 2.5977871786527823e-05,
+      "loss": 10.049,
+      "step": 27280
+    },
+    {
+      "epoch": 0.1362830532597568,
+      "grad_norm": 0.08965323120355606,
+      "learning_rate": 2.5976369871586274e-05,
+      "loss": 10.0505,
+      "step": 27290
+    },
+    {
+      "epoch": 0.13633299208469624,
+      "grad_norm": 0.09475590288639069,
+      "learning_rate": 2.597486795664472e-05,
+      "loss": 10.0487,
+      "step": 27300
+    },
+    {
+      "epoch": 0.13638293090963569,
+      "grad_norm": 0.09472818672657013,
+      "learning_rate": 2.5973366041703174e-05,
+      "loss": 10.0429,
+      "step": 27310
+    },
+    {
+      "epoch": 0.13643286973457514,
+      "grad_norm": 0.09863143414258957,
+      "learning_rate": 2.597186412676162e-05,
+      "loss": 10.0474,
+      "step": 27320
+    },
+    {
+      "epoch": 0.13648280855951458,
+      "grad_norm": 0.0966014415025711,
+      "learning_rate": 2.597036221182007e-05,
+      "loss": 10.0459,
+      "step": 27330
+    },
+    {
+      "epoch": 0.13653274738445403,
+      "grad_norm": 0.09497446566820145,
+      "learning_rate": 2.596886029687852e-05,
+      "loss": 10.0439,
+      "step": 27340
+    },
+    {
+      "epoch": 0.13658268620939348,
+      "grad_norm": 0.09334766864776611,
+      "learning_rate": 2.5967358381936968e-05,
+      "loss": 10.0404,
+      "step": 27350
+    },
+    {
+      "epoch": 0.13663262503433293,
+      "grad_norm": 0.09661349654197693,
+      "learning_rate": 2.596585646699542e-05,
+      "loss": 10.0469,
+      "step": 27360
+    },
+    {
+      "epoch": 0.13668256385927238,
+      "grad_norm": 0.0978384017944336,
+      "learning_rate": 2.5964354552053868e-05,
+      "loss": 10.0472,
+      "step": 27370
+    },
+    {
+      "epoch": 0.13673250268421183,
+      "grad_norm": 0.09664426743984222,
+      "learning_rate": 2.596285263711232e-05,
+      "loss": 10.0426,
+      "step": 27380
+    },
+    {
+      "epoch": 0.13678244150915128,
+      "grad_norm": 0.09428177773952484,
+      "learning_rate": 2.596135072217077e-05,
+      "loss": 10.0432,
+      "step": 27390
+    },
+    {
+      "epoch": 0.13683238033409073,
+      "grad_norm": 0.09939315170049667,
+      "learning_rate": 2.595984880722922e-05,
+      "loss": 10.0388,
+      "step": 27400
+    },
+    {
+      "epoch": 0.13688231915903018,
+      "grad_norm": 0.09356003999710083,
+      "learning_rate": 2.595834689228767e-05,
+      "loss": 10.0395,
+      "step": 27410
+    },
+    {
+      "epoch": 0.13693225798396963,
+      "grad_norm": 0.09405787289142609,
+      "learning_rate": 2.5956844977346116e-05,
+      "loss": 10.0424,
+      "step": 27420
+    },
+    {
+      "epoch": 0.13698219680890908,
+      "grad_norm": 0.09802021831274033,
+      "learning_rate": 2.5955343062404566e-05,
+      "loss": 10.0428,
+      "step": 27430
+    },
+    {
+      "epoch": 0.13703213563384853,
+      "grad_norm": 0.10091257095336914,
+      "learning_rate": 2.5953841147463016e-05,
+      "loss": 10.0415,
+      "step": 27440
+    },
+    {
+      "epoch": 0.13708207445878798,
+      "grad_norm": 0.09429014474153519,
+      "learning_rate": 2.5952339232521466e-05,
+      "loss": 10.0421,
+      "step": 27450
+    },
+    {
+      "epoch": 0.13713201328372743,
+      "grad_norm": 0.09676757454872131,
+      "learning_rate": 2.5950837317579916e-05,
+      "loss": 10.0448,
+      "step": 27460
+    },
+    {
+      "epoch": 0.13718195210866688,
+      "grad_norm": 0.09487222135066986,
+      "learning_rate": 2.5949335402638363e-05,
+      "loss": 10.0386,
+      "step": 27470
+    },
+    {
+      "epoch": 0.13723189093360633,
+      "grad_norm": 0.09765653312206268,
+      "learning_rate": 2.5947833487696813e-05,
+      "loss": 10.0413,
+      "step": 27480
+    },
+    {
+      "epoch": 0.13728182975854578,
+      "grad_norm": 0.09836971014738083,
+      "learning_rate": 2.5946331572755264e-05,
+      "loss": 10.0376,
+      "step": 27490
+    },
+    {
+      "epoch": 0.13733176858348523,
+      "grad_norm": 0.09549671411514282,
+      "learning_rate": 2.5944829657813714e-05,
+      "loss": 10.0385,
+      "step": 27500
+    },
+    {
+      "epoch": 0.13738170740842467,
+      "grad_norm": 0.08956001698970795,
+      "learning_rate": 2.5943327742872164e-05,
+      "loss": 10.0366,
+      "step": 27510
+    },
+    {
+      "epoch": 0.13743164623336412,
+      "grad_norm": 0.09554048627614975,
+      "learning_rate": 2.594182582793061e-05,
+      "loss": 10.0428,
+      "step": 27520
+    },
+    {
+      "epoch": 0.13748158505830357,
+      "grad_norm": 0.10129525512456894,
+      "learning_rate": 2.594032391298906e-05,
+      "loss": 10.0432,
+      "step": 27530
+    },
+    {
+      "epoch": 0.13753152388324302,
+      "grad_norm": 0.09291373938322067,
+      "learning_rate": 2.593882199804751e-05,
+      "loss": 10.0404,
+      "step": 27540
+    },
+    {
+      "epoch": 0.13758146270818247,
+      "grad_norm": 0.09440364688634872,
+      "learning_rate": 2.593732008310596e-05,
+      "loss": 10.0399,
+      "step": 27550
+    },
+    {
+      "epoch": 0.13763140153312192,
+      "grad_norm": 0.09612912684679031,
+      "learning_rate": 2.593581816816441e-05,
+      "loss": 10.0394,
+      "step": 27560
+    },
+    {
+      "epoch": 0.13768134035806137,
+      "grad_norm": 0.09454838931560516,
+      "learning_rate": 2.5934316253222858e-05,
+      "loss": 10.0397,
+      "step": 27570
+    },
+    {
+      "epoch": 0.13773127918300082,
+      "grad_norm": 0.10840394347906113,
+      "learning_rate": 2.593281433828131e-05,
+      "loss": 10.0346,
+      "step": 27580
+    },
+    {
+      "epoch": 0.13778121800794027,
+      "grad_norm": 0.09229467809200287,
+      "learning_rate": 2.593131242333976e-05,
+      "loss": 10.0374,
+      "step": 27590
+    },
+    {
+      "epoch": 0.13783115683287972,
+      "grad_norm": 0.09308604896068573,
+      "learning_rate": 2.592981050839821e-05,
+      "loss": 10.0424,
+      "step": 27600
+    },
+    {
+      "epoch": 0.13788109565781917,
+      "grad_norm": 0.09472574293613434,
+      "learning_rate": 2.592830859345666e-05,
+      "loss": 10.0377,
+      "step": 27610
+    },
+    {
+      "epoch": 0.13793103448275862,
+      "grad_norm": 0.09195975959300995,
+      "learning_rate": 2.5926806678515106e-05,
+      "loss": 10.0403,
+      "step": 27620
+    },
+    {
+      "epoch": 0.13798097330769807,
+      "grad_norm": 0.09835057705640793,
+      "learning_rate": 2.5925304763573556e-05,
+      "loss": 10.0352,
+      "step": 27630
+    },
+    {
+      "epoch": 0.13803091213263752,
+      "grad_norm": 0.10062667727470398,
+      "learning_rate": 2.5923802848632006e-05,
+      "loss": 10.0336,
+      "step": 27640
+    },
+    {
+      "epoch": 0.13808085095757697,
+      "grad_norm": 0.08686701953411102,
+      "learning_rate": 2.5922300933690456e-05,
+      "loss": 10.0383,
+      "step": 27650
+    },
+    {
+      "epoch": 0.13813078978251642,
+      "grad_norm": 0.09707959741353989,
+      "learning_rate": 2.5920799018748906e-05,
+      "loss": 10.0345,
+      "step": 27660
+    },
+    {
+      "epoch": 0.13818072860745587,
+      "grad_norm": 0.09171800315380096,
+      "learning_rate": 2.5919297103807353e-05,
+      "loss": 10.0336,
+      "step": 27670
+    },
+    {
+      "epoch": 0.13823066743239532,
+      "grad_norm": 0.0923093855381012,
+      "learning_rate": 2.5917795188865803e-05,
+      "loss": 10.0352,
+      "step": 27680
+    },
+    {
+      "epoch": 0.13828060625733476,
+      "grad_norm": 0.0912717804312706,
+      "learning_rate": 2.5916293273924254e-05,
+      "loss": 10.0339,
+      "step": 27690
+    },
+    {
+      "epoch": 0.13833054508227421,
+      "grad_norm": 0.09480732679367065,
+      "learning_rate": 2.5914791358982704e-05,
+      "loss": 10.0389,
+      "step": 27700
+    },
+    {
+      "epoch": 0.13838048390721366,
+      "grad_norm": 0.09554393589496613,
+      "learning_rate": 2.5913289444041154e-05,
+      "loss": 10.0352,
+      "step": 27710
+    },
+    {
+      "epoch": 0.1384304227321531,
+      "grad_norm": 0.09049583226442337,
+      "learning_rate": 2.5911787529099604e-05,
+      "loss": 10.032,
+      "step": 27720
+    },
+    {
+      "epoch": 0.13848036155709256,
+      "grad_norm": 0.09413400292396545,
+      "learning_rate": 2.591028561415805e-05,
+      "loss": 10.0393,
+      "step": 27730
+    },
+    {
+      "epoch": 0.138530300382032,
+      "grad_norm": 0.09317301958799362,
+      "learning_rate": 2.59087836992165e-05,
+      "loss": 10.0328,
+      "step": 27740
+    },
+    {
+      "epoch": 0.13858023920697146,
+      "grad_norm": 0.09751195460557938,
+      "learning_rate": 2.590728178427495e-05,
+      "loss": 10.0299,
+      "step": 27750
+    },
+    {
+      "epoch": 0.1386301780319109,
+      "grad_norm": 0.09604477882385254,
+      "learning_rate": 2.59057798693334e-05,
+      "loss": 10.0348,
+      "step": 27760
+    },
+    {
+      "epoch": 0.13868011685685036,
+      "grad_norm": 0.09635339677333832,
+      "learning_rate": 2.590427795439185e-05,
+      "loss": 10.0346,
+      "step": 27770
+    },
+    {
+      "epoch": 0.1387300556817898,
+      "grad_norm": 0.10057932138442993,
+      "learning_rate": 2.59027760394503e-05,
+      "loss": 10.0319,
+      "step": 27780
+    },
+    {
+      "epoch": 0.13877999450672926,
+      "grad_norm": 0.09397371113300323,
+      "learning_rate": 2.590127412450875e-05,
+      "loss": 10.0331,
+      "step": 27790
+    },
+    {
+      "epoch": 0.1388299333316687,
+      "grad_norm": 0.09505651146173477,
+      "learning_rate": 2.58997722095672e-05,
+      "loss": 10.0293,
+      "step": 27800
+    },
+    {
+      "epoch": 0.13887987215660816,
+      "grad_norm": 0.09163686633110046,
+      "learning_rate": 2.589827029462565e-05,
+      "loss": 10.0265,
+      "step": 27810
+    },
+    {
+      "epoch": 0.1389298109815476,
+      "grad_norm": 0.09803500026464462,
+      "learning_rate": 2.58967683796841e-05,
+      "loss": 10.0362,
+      "step": 27820
+    },
+    {
+      "epoch": 0.13897974980648706,
+      "grad_norm": 0.1022925078868866,
+      "learning_rate": 2.5895266464742546e-05,
+      "loss": 10.033,
+      "step": 27830
+    },
+    {
+      "epoch": 0.1390296886314265,
+      "grad_norm": 0.09815297275781631,
+      "learning_rate": 2.5893764549800996e-05,
+      "loss": 10.0308,
+      "step": 27840
+    },
+    {
+      "epoch": 0.13907962745636596,
+      "grad_norm": 0.09278080612421036,
+      "learning_rate": 2.5892262634859446e-05,
+      "loss": 10.0297,
+      "step": 27850
+    },
+    {
+      "epoch": 0.1391295662813054,
+      "grad_norm": 0.09531033784151077,
+      "learning_rate": 2.5890760719917896e-05,
+      "loss": 10.0336,
+      "step": 27860
+    },
+    {
+      "epoch": 0.13917950510624486,
+      "grad_norm": 0.1017637774348259,
+      "learning_rate": 2.5889258804976347e-05,
+      "loss": 10.0323,
+      "step": 27870
+    },
+    {
+      "epoch": 0.1392294439311843,
+      "grad_norm": 0.08980844169855118,
+      "learning_rate": 2.5887756890034793e-05,
+      "loss": 10.0351,
+      "step": 27880
+    },
+    {
+      "epoch": 0.13927938275612375,
+      "grad_norm": 0.09667633473873138,
+      "learning_rate": 2.5886254975093244e-05,
+      "loss": 10.0344,
+      "step": 27890
+    },
+    {
+      "epoch": 0.1393293215810632,
+      "grad_norm": 0.09221838414669037,
+      "learning_rate": 2.5884753060151694e-05,
+      "loss": 10.0287,
+      "step": 27900
+    },
+    {
+      "epoch": 0.13937926040600265,
+      "grad_norm": 0.09094412624835968,
+      "learning_rate": 2.5883251145210144e-05,
+      "loss": 10.0363,
+      "step": 27910
+    },
+    {
+      "epoch": 0.1394291992309421,
+      "grad_norm": 0.09816347807645798,
+      "learning_rate": 2.5881749230268594e-05,
+      "loss": 10.0317,
+      "step": 27920
+    },
+    {
+      "epoch": 0.13947913805588155,
+      "grad_norm": 0.09702466428279877,
+      "learning_rate": 2.588024731532704e-05,
+      "loss": 10.0289,
+      "step": 27930
+    },
+    {
+      "epoch": 0.139529076880821,
+      "grad_norm": 0.09658083319664001,
+      "learning_rate": 2.587874540038549e-05,
+      "loss": 10.0304,
+      "step": 27940
+    },
+    {
+      "epoch": 0.13957901570576045,
+      "grad_norm": 0.09620265662670135,
+      "learning_rate": 2.587724348544394e-05,
+      "loss": 10.0229,
+      "step": 27950
+    },
+    {
+      "epoch": 0.1396289545306999,
+      "grad_norm": 0.09803494811058044,
+      "learning_rate": 2.587574157050239e-05,
+      "loss": 10.0264,
+      "step": 27960
+    },
+    {
+      "epoch": 0.13967889335563935,
+      "grad_norm": 0.09247355163097382,
+      "learning_rate": 2.587423965556084e-05,
+      "loss": 10.0331,
+      "step": 27970
+    },
+    {
+      "epoch": 0.1397288321805788,
+      "grad_norm": 0.09678816050291061,
+      "learning_rate": 2.587273774061929e-05,
+      "loss": 10.0242,
+      "step": 27980
+    },
+    {
+      "epoch": 0.13977877100551825,
+      "grad_norm": 0.0913391187787056,
+      "learning_rate": 2.587123582567774e-05,
+      "loss": 10.0266,
+      "step": 27990
+    },
+    {
+      "epoch": 0.1398287098304577,
+      "grad_norm": 0.0930318757891655,
+      "learning_rate": 2.586973391073619e-05,
+      "loss": 10.0247,
+      "step": 28000
+    },
+    {
+      "epoch": 0.13987864865539715,
+      "grad_norm": 0.09507840126752853,
+      "learning_rate": 2.586823199579464e-05,
+      "loss": 10.0282,
+      "step": 28010
+    },
+    {
+      "epoch": 0.1399285874803366,
+      "grad_norm": 0.09485821425914764,
+      "learning_rate": 2.586673008085309e-05,
+      "loss": 10.0286,
+      "step": 28020
+    },
+    {
+      "epoch": 0.13997852630527605,
+      "grad_norm": 0.09455297142267227,
+      "learning_rate": 2.5865228165911536e-05,
+      "loss": 10.0309,
+      "step": 28030
+    },
+    {
+      "epoch": 0.1400284651302155,
+      "grad_norm": 0.0947793647646904,
+      "learning_rate": 2.586372625096999e-05,
+      "loss": 10.027,
+      "step": 28040
+    },
+    {
+      "epoch": 0.14007840395515495,
+      "grad_norm": 0.09246467053890228,
+      "learning_rate": 2.5862224336028436e-05,
+      "loss": 10.0264,
+      "step": 28050
+    },
+    {
+      "epoch": 0.1401283427800944,
+      "grad_norm": 0.09456957876682281,
+      "learning_rate": 2.5860722421086886e-05,
+      "loss": 10.0264,
+      "step": 28060
+    },
+    {
+      "epoch": 0.14017828160503384,
+      "grad_norm": 0.09399653971195221,
+      "learning_rate": 2.5859220506145337e-05,
+      "loss": 10.0288,
+      "step": 28070
+    },
+    {
+      "epoch": 0.1402282204299733,
+      "grad_norm": 0.09784771502017975,
+      "learning_rate": 2.5857718591203783e-05,
+      "loss": 10.0218,
+      "step": 28080
+    },
+    {
+      "epoch": 0.14027815925491274,
+      "grad_norm": 0.09228380024433136,
+      "learning_rate": 2.5856216676262237e-05,
+      "loss": 10.0276,
+      "step": 28090
+    },
+    {
+      "epoch": 0.1403280980798522,
+      "grad_norm": 0.09873951226472855,
+      "learning_rate": 2.5854714761320684e-05,
+      "loss": 10.0257,
+      "step": 28100
+    },
+    {
+      "epoch": 0.14037803690479164,
+      "grad_norm": 0.09930938482284546,
+      "learning_rate": 2.5853212846379134e-05,
+      "loss": 10.0304,
+      "step": 28110
+    },
+    {
+      "epoch": 0.1404279757297311,
+      "grad_norm": 0.09772654622793198,
+      "learning_rate": 2.5851710931437584e-05,
+      "loss": 10.0231,
+      "step": 28120
+    },
+    {
+      "epoch": 0.14047791455467054,
+      "grad_norm": 0.08995553106069565,
+      "learning_rate": 2.585020901649603e-05,
+      "loss": 10.0225,
+      "step": 28130
+    },
+    {
+      "epoch": 0.14052785337961,
+      "grad_norm": 0.09744606912136078,
+      "learning_rate": 2.5848707101554484e-05,
+      "loss": 10.0267,
+      "step": 28140
+    },
+    {
+      "epoch": 0.14057779220454944,
+      "grad_norm": 0.09355470538139343,
+      "learning_rate": 2.584720518661293e-05,
+      "loss": 10.0261,
+      "step": 28150
+    },
+    {
+      "epoch": 0.1406277310294889,
+      "grad_norm": 0.097645103931427,
+      "learning_rate": 2.584570327167138e-05,
+      "loss": 10.0217,
+      "step": 28160
+    },
+    {
+      "epoch": 0.14067766985442834,
+      "grad_norm": 0.09685888141393661,
+      "learning_rate": 2.584420135672983e-05,
+      "loss": 10.0217,
+      "step": 28170
+    },
+    {
+      "epoch": 0.1407276086793678,
+      "grad_norm": 0.09667977690696716,
+      "learning_rate": 2.584269944178828e-05,
+      "loss": 10.0277,
+      "step": 28180
+    },
+    {
+      "epoch": 0.1407775475043072,
+      "grad_norm": 0.09020134806632996,
+      "learning_rate": 2.5841197526846732e-05,
+      "loss": 10.0203,
+      "step": 28190
+    },
+    {
+      "epoch": 0.14082748632924666,
+      "grad_norm": 0.09370194375514984,
+      "learning_rate": 2.583969561190518e-05,
+      "loss": 10.0225,
+      "step": 28200
+    },
+    {
+      "epoch": 0.1408774251541861,
+      "grad_norm": 0.09775953739881516,
+      "learning_rate": 2.583819369696363e-05,
+      "loss": 10.0206,
+      "step": 28210
+    },
+    {
+      "epoch": 0.14092736397912556,
+      "grad_norm": 0.09635622054338455,
+      "learning_rate": 2.583669178202208e-05,
+      "loss": 10.0231,
+      "step": 28220
+    },
+    {
+      "epoch": 0.140977302804065,
+      "grad_norm": 0.09073793143033981,
+      "learning_rate": 2.5835189867080526e-05,
+      "loss": 10.0269,
+      "step": 28230
+    },
+    {
+      "epoch": 0.14102724162900446,
+      "grad_norm": 0.10052619129419327,
+      "learning_rate": 2.583368795213898e-05,
+      "loss": 10.0258,
+      "step": 28240
+    },
+    {
+      "epoch": 0.1410771804539439,
+      "grad_norm": 0.09381050616502762,
+      "learning_rate": 2.5832186037197426e-05,
+      "loss": 10.0208,
+      "step": 28250
+    },
+    {
+      "epoch": 0.14112711927888336,
+      "grad_norm": 0.10019423812627792,
+      "learning_rate": 2.5830684122255876e-05,
+      "loss": 10.0195,
+      "step": 28260
+    },
+    {
+      "epoch": 0.1411770581038228,
+      "grad_norm": 0.09306546300649643,
+      "learning_rate": 2.5829182207314327e-05,
+      "loss": 10.0191,
+      "step": 28270
+    },
+    {
+      "epoch": 0.14122699692876226,
+      "grad_norm": 0.09635606408119202,
+      "learning_rate": 2.5827680292372773e-05,
+      "loss": 10.0268,
+      "step": 28280
+    },
+    {
+      "epoch": 0.1412769357537017,
+      "grad_norm": 0.10159686952829361,
+      "learning_rate": 2.5826178377431227e-05,
+      "loss": 10.0214,
+      "step": 28290
+    },
+    {
+      "epoch": 0.14132687457864115,
+      "grad_norm": 0.09453240036964417,
+      "learning_rate": 2.5824676462489674e-05,
+      "loss": 10.0215,
+      "step": 28300
+    },
+    {
+      "epoch": 0.1413768134035806,
+      "grad_norm": 0.09013237059116364,
+      "learning_rate": 2.5823174547548124e-05,
+      "loss": 10.024,
+      "step": 28310
+    },
+    {
+      "epoch": 0.14142675222852005,
+      "grad_norm": 0.0938793197274208,
+      "learning_rate": 2.5821672632606574e-05,
+      "loss": 10.0176,
+      "step": 28320
+    },
+    {
+      "epoch": 0.1414766910534595,
+      "grad_norm": 0.09485707432031631,
+      "learning_rate": 2.582017071766502e-05,
+      "loss": 10.0171,
+      "step": 28330
+    },
+    {
+      "epoch": 0.14152662987839895,
+      "grad_norm": 0.10150688141584396,
+      "learning_rate": 2.5818668802723474e-05,
+      "loss": 10.019,
+      "step": 28340
+    },
+    {
+      "epoch": 0.1415765687033384,
+      "grad_norm": 0.09411424398422241,
+      "learning_rate": 2.581716688778192e-05,
+      "loss": 10.0231,
+      "step": 28350
+    },
+    {
+      "epoch": 0.14162650752827785,
+      "grad_norm": 0.09432416409254074,
+      "learning_rate": 2.5815664972840375e-05,
+      "loss": 10.0187,
+      "step": 28360
+    },
+    {
+      "epoch": 0.1416764463532173,
+      "grad_norm": 0.09197516739368439,
+      "learning_rate": 2.581416305789882e-05,
+      "loss": 10.0196,
+      "step": 28370
+    },
+    {
+      "epoch": 0.14172638517815675,
+      "grad_norm": 0.09398947656154633,
+      "learning_rate": 2.581266114295727e-05,
+      "loss": 10.0159,
+      "step": 28380
+    },
+    {
+      "epoch": 0.1417763240030962,
+      "grad_norm": 0.10007443279027939,
+      "learning_rate": 2.5811159228015722e-05,
+      "loss": 10.0165,
+      "step": 28390
+    },
+    {
+      "epoch": 0.14182626282803565,
+      "grad_norm": 0.09710991382598877,
+      "learning_rate": 2.580965731307417e-05,
+      "loss": 10.0245,
+      "step": 28400
+    },
+    {
+      "epoch": 0.1418762016529751,
+      "grad_norm": 0.09509746730327606,
+      "learning_rate": 2.5808155398132622e-05,
+      "loss": 10.0178,
+      "step": 28410
+    },
+    {
+      "epoch": 0.14192614047791455,
+      "grad_norm": 0.09114133566617966,
+      "learning_rate": 2.580665348319107e-05,
+      "loss": 10.0192,
+      "step": 28420
+    },
+    {
+      "epoch": 0.141976079302854,
+      "grad_norm": 0.0989239513874054,
+      "learning_rate": 2.5805151568249516e-05,
+      "loss": 10.017,
+      "step": 28430
+    },
+    {
+      "epoch": 0.14202601812779345,
+      "grad_norm": 0.0925961434841156,
+      "learning_rate": 2.580364965330797e-05,
+      "loss": 10.0186,
+      "step": 28440
+    },
+    {
+      "epoch": 0.1420759569527329,
+      "grad_norm": 0.09893997758626938,
+      "learning_rate": 2.5802147738366416e-05,
+      "loss": 10.0155,
+      "step": 28450
+    },
+    {
+      "epoch": 0.14212589577767235,
+      "grad_norm": 0.0943673700094223,
+      "learning_rate": 2.580064582342487e-05,
+      "loss": 10.0159,
+      "step": 28460
+    },
+    {
+      "epoch": 0.1421758346026118,
+      "grad_norm": 0.09693938493728638,
+      "learning_rate": 2.5799143908483317e-05,
+      "loss": 10.0115,
+      "step": 28470
+    },
+    {
+      "epoch": 0.14222577342755124,
+      "grad_norm": 0.09356331825256348,
+      "learning_rate": 2.5797641993541763e-05,
+      "loss": 10.014,
+      "step": 28480
+    },
+    {
+      "epoch": 0.1422757122524907,
+      "grad_norm": 0.09694939106702805,
+      "learning_rate": 2.5796140078600217e-05,
+      "loss": 10.0138,
+      "step": 28490
+    },
+    {
+      "epoch": 0.14232565107743014,
+      "grad_norm": 0.09721967577934265,
+      "learning_rate": 2.5794638163658664e-05,
+      "loss": 10.0084,
+      "step": 28500
+    },
+    {
+      "epoch": 0.1423755899023696,
+      "grad_norm": 0.09602759033441544,
+      "learning_rate": 2.5793136248717117e-05,
+      "loss": 10.015,
+      "step": 28510
+    },
+    {
+      "epoch": 0.14242552872730904,
+      "grad_norm": 0.09390842169523239,
+      "learning_rate": 2.5791634333775564e-05,
+      "loss": 10.0172,
+      "step": 28520
+    },
+    {
+      "epoch": 0.1424754675522485,
+      "grad_norm": 0.09624374657869339,
+      "learning_rate": 2.579013241883401e-05,
+      "loss": 10.0161,
+      "step": 28530
+    },
+    {
+      "epoch": 0.14252540637718794,
+      "grad_norm": 0.09962889552116394,
+      "learning_rate": 2.5788630503892465e-05,
+      "loss": 10.0149,
+      "step": 28540
+    },
+    {
+      "epoch": 0.1425753452021274,
+      "grad_norm": 0.09729303419589996,
+      "learning_rate": 2.578712858895091e-05,
+      "loss": 10.0157,
+      "step": 28550
+    },
+    {
+      "epoch": 0.14262528402706684,
+      "grad_norm": 0.09288518875837326,
+      "learning_rate": 2.5785626674009365e-05,
+      "loss": 10.0173,
+      "step": 28560
+    },
+    {
+      "epoch": 0.1426752228520063,
+      "grad_norm": 0.09366150200366974,
+      "learning_rate": 2.578412475906781e-05,
+      "loss": 10.0122,
+      "step": 28570
+    },
+    {
+      "epoch": 0.14272516167694574,
+      "grad_norm": 0.0944962128996849,
+      "learning_rate": 2.5782622844126262e-05,
+      "loss": 10.0112,
+      "step": 28580
+    },
+    {
+      "epoch": 0.1427751005018852,
+      "grad_norm": 0.09178702533245087,
+      "learning_rate": 2.5781120929184712e-05,
+      "loss": 10.0179,
+      "step": 28590
+    },
+    {
+      "epoch": 0.14282503932682464,
+      "grad_norm": 0.0905400812625885,
+      "learning_rate": 2.577961901424316e-05,
+      "loss": 10.0134,
+      "step": 28600
+    },
+    {
+      "epoch": 0.1428749781517641,
+      "grad_norm": 0.09418443590402603,
+      "learning_rate": 2.5778117099301612e-05,
+      "loss": 10.0168,
+      "step": 28610
+    },
+    {
+      "epoch": 0.14292491697670354,
+      "grad_norm": 0.0970693901181221,
+      "learning_rate": 2.577661518436006e-05,
+      "loss": 10.0162,
+      "step": 28620
+    },
+    {
+      "epoch": 0.14297485580164299,
+      "grad_norm": 0.09255459159612656,
+      "learning_rate": 2.577511326941851e-05,
+      "loss": 10.0129,
+      "step": 28630
+    },
+    {
+      "epoch": 0.14302479462658244,
+      "grad_norm": 0.10080841928720474,
+      "learning_rate": 2.577361135447696e-05,
+      "loss": 10.0144,
+      "step": 28640
+    },
+    {
+      "epoch": 0.14307473345152188,
+      "grad_norm": 0.09627335518598557,
+      "learning_rate": 2.5772109439535406e-05,
+      "loss": 10.0095,
+      "step": 28650
+    },
+    {
+      "epoch": 0.14312467227646133,
+      "grad_norm": 0.09631509333848953,
+      "learning_rate": 2.577060752459386e-05,
+      "loss": 10.0167,
+      "step": 28660
+    },
+    {
+      "epoch": 0.14317461110140078,
+      "grad_norm": 0.09696370363235474,
+      "learning_rate": 2.5769105609652307e-05,
+      "loss": 10.0145,
+      "step": 28670
+    },
+    {
+      "epoch": 0.14322454992634023,
+      "grad_norm": 0.08783876895904541,
+      "learning_rate": 2.5767603694710757e-05,
+      "loss": 10.0139,
+      "step": 28680
+    },
+    {
+      "epoch": 0.14327448875127968,
+      "grad_norm": 0.09430178999900818,
+      "learning_rate": 2.5766101779769207e-05,
+      "loss": 10.0151,
+      "step": 28690
+    },
+    {
+      "epoch": 0.14332442757621913,
+      "grad_norm": 0.09581022709608078,
+      "learning_rate": 2.5764599864827654e-05,
+      "loss": 10.0074,
+      "step": 28700
+    },
+    {
+      "epoch": 0.14337436640115858,
+      "grad_norm": 0.10260827839374542,
+      "learning_rate": 2.5763097949886107e-05,
+      "loss": 10.0112,
+      "step": 28710
+    },
+    {
+      "epoch": 0.14342430522609803,
+      "grad_norm": 0.09252434223890305,
+      "learning_rate": 2.5761596034944554e-05,
+      "loss": 10.0146,
+      "step": 28720
+    },
+    {
+      "epoch": 0.14347424405103748,
+      "grad_norm": 0.08901256322860718,
+      "learning_rate": 2.5760094120003004e-05,
+      "loss": 10.0066,
+      "step": 28730
+    },
+    {
+      "epoch": 0.14352418287597693,
+      "grad_norm": 0.09571991860866547,
+      "learning_rate": 2.5758592205061455e-05,
+      "loss": 10.0124,
+      "step": 28740
+    },
+    {
+      "epoch": 0.14357412170091638,
+      "grad_norm": 0.09751693159341812,
+      "learning_rate": 2.57570902901199e-05,
+      "loss": 10.0108,
+      "step": 28750
+    },
+    {
+      "epoch": 0.14362406052585583,
+      "grad_norm": 0.09803242981433868,
+      "learning_rate": 2.5755588375178355e-05,
+      "loss": 10.0063,
+      "step": 28760
+    },
+    {
+      "epoch": 0.14367399935079528,
+      "grad_norm": 0.0984642431139946,
+      "learning_rate": 2.57540864602368e-05,
+      "loss": 10.0058,
+      "step": 28770
+    },
+    {
+      "epoch": 0.14372393817573473,
+      "grad_norm": 0.0973120778799057,
+      "learning_rate": 2.5752584545295252e-05,
+      "loss": 10.0062,
+      "step": 28780
+    },
+    {
+      "epoch": 0.14377387700067418,
+      "grad_norm": 0.09399261325597763,
+      "learning_rate": 2.5751082630353702e-05,
+      "loss": 10.0063,
+      "step": 28790
+    },
+    {
+      "epoch": 0.14382381582561363,
+      "grad_norm": 0.09200508892536163,
+      "learning_rate": 2.574958071541215e-05,
+      "loss": 10.0079,
+      "step": 28800
+    },
+    {
+      "epoch": 0.14387375465055308,
+      "grad_norm": 0.09684068709611893,
+      "learning_rate": 2.5748078800470602e-05,
+      "loss": 10.004,
+      "step": 28810
+    },
+    {
+      "epoch": 0.14392369347549253,
+      "grad_norm": 0.09978806972503662,
+      "learning_rate": 2.574657688552905e-05,
+      "loss": 10.0058,
+      "step": 28820
+    },
+    {
+      "epoch": 0.14397363230043198,
+      "grad_norm": 0.09788274019956589,
+      "learning_rate": 2.57450749705875e-05,
+      "loss": 10.0059,
+      "step": 28830
+    },
+    {
+      "epoch": 0.14402357112537142,
+      "grad_norm": 0.09136934578418732,
+      "learning_rate": 2.574357305564595e-05,
+      "loss": 10.0068,
+      "step": 28840
+    },
+    {
+      "epoch": 0.14407350995031087,
+      "grad_norm": 0.09080319106578827,
+      "learning_rate": 2.5742071140704396e-05,
+      "loss": 10.0057,
+      "step": 28850
+    },
+    {
+      "epoch": 0.14412344877525032,
+      "grad_norm": 0.09714391082525253,
+      "learning_rate": 2.574056922576285e-05,
+      "loss": 10.0084,
+      "step": 28860
+    },
+    {
+      "epoch": 0.14417338760018977,
+      "grad_norm": 0.1014237254858017,
+      "learning_rate": 2.5739067310821297e-05,
+      "loss": 10.0073,
+      "step": 28870
+    },
+    {
+      "epoch": 0.14422332642512922,
+      "grad_norm": 0.09329380840063095,
+      "learning_rate": 2.573756539587975e-05,
+      "loss": 10.0075,
+      "step": 28880
+    },
+    {
+      "epoch": 0.14427326525006867,
+      "grad_norm": 0.09168153256177902,
+      "learning_rate": 2.5736063480938197e-05,
+      "loss": 10.0039,
+      "step": 28890
+    },
+    {
+      "epoch": 0.14432320407500812,
+      "grad_norm": 0.09341901540756226,
+      "learning_rate": 2.5734561565996644e-05,
+      "loss": 10.0078,
+      "step": 28900
+    },
+    {
+      "epoch": 0.14437314289994757,
+      "grad_norm": 0.08910293877124786,
+      "learning_rate": 2.5733059651055097e-05,
+      "loss": 10.0063,
+      "step": 28910
+    },
+    {
+      "epoch": 0.14442308172488702,
+      "grad_norm": 0.09554349631071091,
+      "learning_rate": 2.5731557736113544e-05,
+      "loss": 10.0101,
+      "step": 28920
+    },
+    {
+      "epoch": 0.14447302054982647,
+      "grad_norm": 0.09408532083034515,
+      "learning_rate": 2.5730055821171998e-05,
+      "loss": 10.0043,
+      "step": 28930
+    },
+    {
+      "epoch": 0.14452295937476592,
+      "grad_norm": 0.09786105901002884,
+      "learning_rate": 2.5728553906230445e-05,
+      "loss": 10.0024,
+      "step": 28940
+    },
+    {
+      "epoch": 0.14457289819970537,
+      "grad_norm": 0.09660477936267853,
+      "learning_rate": 2.572705199128889e-05,
+      "loss": 10.0064,
+      "step": 28950
+    },
+    {
+      "epoch": 0.14462283702464482,
+      "grad_norm": 0.088541179895401,
+      "learning_rate": 2.5725550076347345e-05,
+      "loss": 10.0054,
+      "step": 28960
+    },
+    {
+      "epoch": 0.14467277584958427,
+      "grad_norm": 0.09211627393960953,
+      "learning_rate": 2.572404816140579e-05,
+      "loss": 10.0047,
+      "step": 28970
+    },
+    {
+      "epoch": 0.14472271467452372,
+      "grad_norm": 0.10103971511125565,
+      "learning_rate": 2.5722546246464245e-05,
+      "loss": 10.0037,
+      "step": 28980
+    },
+    {
+      "epoch": 0.14477265349946317,
+      "grad_norm": 0.09156627953052521,
+      "learning_rate": 2.5721044331522692e-05,
+      "loss": 10.0016,
+      "step": 28990
+    },
+    {
+      "epoch": 0.14482259232440262,
+      "grad_norm": 0.08952153474092484,
+      "learning_rate": 2.5719542416581142e-05,
+      "loss": 10.0043,
+      "step": 29000
+    },
+    {
+      "epoch": 0.14487253114934207,
+      "grad_norm": 0.09983877092599869,
+      "learning_rate": 2.5718040501639592e-05,
+      "loss": 10.0048,
+      "step": 29010
+    },
+    {
+      "epoch": 0.14492246997428151,
+      "grad_norm": 0.09366819262504578,
+      "learning_rate": 2.571653858669804e-05,
+      "loss": 10.0025,
+      "step": 29020
+    },
+    {
+      "epoch": 0.14497240879922096,
+      "grad_norm": 0.09434466063976288,
+      "learning_rate": 2.5715036671756493e-05,
+      "loss": 10.0053,
+      "step": 29030
+    },
+    {
+      "epoch": 0.1450223476241604,
+      "grad_norm": 0.09528578817844391,
+      "learning_rate": 2.571353475681494e-05,
+      "loss": 10.0025,
+      "step": 29040
+    },
+    {
+      "epoch": 0.14507228644909986,
+      "grad_norm": 0.09328389912843704,
+      "learning_rate": 2.571203284187339e-05,
+      "loss": 9.9988,
+      "step": 29050
+    },
+    {
+      "epoch": 0.1451222252740393,
+      "grad_norm": 0.09483915567398071,
+      "learning_rate": 2.571053092693184e-05,
+      "loss": 10.0042,
+      "step": 29060
+    },
+    {
+      "epoch": 0.14517216409897876,
+      "grad_norm": 0.09249535948038101,
+      "learning_rate": 2.5709029011990287e-05,
+      "loss": 10.0043,
+      "step": 29070
+    },
+    {
+      "epoch": 0.1452221029239182,
+      "grad_norm": 0.09545839577913284,
+      "learning_rate": 2.570752709704874e-05,
+      "loss": 10.0014,
+      "step": 29080
+    },
+    {
+      "epoch": 0.14527204174885766,
+      "grad_norm": 0.0961858257651329,
+      "learning_rate": 2.5706025182107187e-05,
+      "loss": 10.0063,
+      "step": 29090
+    },
+    {
+      "epoch": 0.1453219805737971,
+      "grad_norm": 0.09866221994161606,
+      "learning_rate": 2.5704523267165637e-05,
+      "loss": 9.9957,
+      "step": 29100
+    },
+    {
+      "epoch": 0.14537191939873656,
+      "grad_norm": 0.0952538251876831,
+      "learning_rate": 2.5703021352224087e-05,
+      "loss": 9.9998,
+      "step": 29110
+    },
+    {
+      "epoch": 0.145421858223676,
+      "grad_norm": 0.09659268707036972,
+      "learning_rate": 2.5701519437282534e-05,
+      "loss": 10.0082,
+      "step": 29120
+    },
+    {
+      "epoch": 0.14547179704861546,
+      "grad_norm": 0.09689272195100784,
+      "learning_rate": 2.5700017522340988e-05,
+      "loss": 10.0042,
+      "step": 29130
+    },
+    {
+      "epoch": 0.1455217358735549,
+      "grad_norm": 0.09408452361822128,
+      "learning_rate": 2.5698515607399435e-05,
+      "loss": 9.9958,
+      "step": 29140
+    },
+    {
+      "epoch": 0.14557167469849436,
+      "grad_norm": 0.09306766837835312,
+      "learning_rate": 2.5697013692457885e-05,
+      "loss": 9.9998,
+      "step": 29150
+    },
+    {
+      "epoch": 0.1456216135234338,
+      "grad_norm": 0.10271524637937546,
+      "learning_rate": 2.5695511777516335e-05,
+      "loss": 9.9997,
+      "step": 29160
+    },
+    {
+      "epoch": 0.14567155234837326,
+      "grad_norm": 0.09216585755348206,
+      "learning_rate": 2.5694009862574782e-05,
+      "loss": 9.9944,
+      "step": 29170
+    },
+    {
+      "epoch": 0.14572149117331268,
+      "grad_norm": 0.09553508460521698,
+      "learning_rate": 2.5692507947633235e-05,
+      "loss": 10.0022,
+      "step": 29180
+    },
+    {
+      "epoch": 0.14577142999825213,
+      "grad_norm": 0.09927748143672943,
+      "learning_rate": 2.5691006032691682e-05,
+      "loss": 9.9974,
+      "step": 29190
+    },
+    {
+      "epoch": 0.14582136882319158,
+      "grad_norm": 0.09433567523956299,
+      "learning_rate": 2.5689504117750132e-05,
+      "loss": 10.0014,
+      "step": 29200
+    },
+    {
+      "epoch": 0.14587130764813103,
+      "grad_norm": 0.0930803120136261,
+      "learning_rate": 2.5688002202808582e-05,
+      "loss": 10.0037,
+      "step": 29210
+    },
+    {
+      "epoch": 0.14592124647307048,
+      "grad_norm": 0.09351327270269394,
+      "learning_rate": 2.568650028786703e-05,
+      "loss": 9.9983,
+      "step": 29220
+    },
+    {
+      "epoch": 0.14597118529800993,
+      "grad_norm": 0.09411612898111343,
+      "learning_rate": 2.5684998372925483e-05,
+      "loss": 10.002,
+      "step": 29230
+    },
+    {
+      "epoch": 0.14602112412294938,
+      "grad_norm": 0.09385941177606583,
+      "learning_rate": 2.568349645798393e-05,
+      "loss": 10.0001,
+      "step": 29240
+    },
+    {
+      "epoch": 0.14607106294788882,
+      "grad_norm": 0.09528261423110962,
+      "learning_rate": 2.568199454304238e-05,
+      "loss": 9.9901,
+      "step": 29250
+    },
+    {
+      "epoch": 0.14612100177282827,
+      "grad_norm": 0.09263186156749725,
+      "learning_rate": 2.568049262810083e-05,
+      "loss": 10.0004,
+      "step": 29260
+    },
+    {
+      "epoch": 0.14617094059776772,
+      "grad_norm": 0.09151321649551392,
+      "learning_rate": 2.5678990713159277e-05,
+      "loss": 10.0003,
+      "step": 29270
+    },
+    {
+      "epoch": 0.14622087942270717,
+      "grad_norm": 0.09789709746837616,
+      "learning_rate": 2.567748879821773e-05,
+      "loss": 9.9961,
+      "step": 29280
+    },
+    {
+      "epoch": 0.14627081824764662,
+      "grad_norm": 0.0894283652305603,
+      "learning_rate": 2.5675986883276177e-05,
+      "loss": 9.9954,
+      "step": 29290
+    },
+    {
+      "epoch": 0.14632075707258607,
+      "grad_norm": 0.10029701143503189,
+      "learning_rate": 2.5674484968334627e-05,
+      "loss": 9.9962,
+      "step": 29300
+    },
+    {
+      "epoch": 0.14637069589752552,
+      "grad_norm": 0.09673379361629486,
+      "learning_rate": 2.5672983053393077e-05,
+      "loss": 9.9969,
+      "step": 29310
+    },
+    {
+      "epoch": 0.14642063472246497,
+      "grad_norm": 0.09456515312194824,
+      "learning_rate": 2.5671481138451528e-05,
+      "loss": 9.9946,
+      "step": 29320
+    },
+    {
+      "epoch": 0.14647057354740442,
+      "grad_norm": 0.09132513403892517,
+      "learning_rate": 2.5669979223509978e-05,
+      "loss": 9.994,
+      "step": 29330
+    },
+    {
+      "epoch": 0.14652051237234387,
+      "grad_norm": 0.09338773041963577,
+      "learning_rate": 2.5668477308568425e-05,
+      "loss": 9.9982,
+      "step": 29340
+    },
+    {
+      "epoch": 0.14657045119728332,
+      "grad_norm": 0.0984477698802948,
+      "learning_rate": 2.5666975393626875e-05,
+      "loss": 9.9914,
+      "step": 29350
+    },
+    {
+      "epoch": 0.14662039002222277,
+      "grad_norm": 0.09212279319763184,
+      "learning_rate": 2.5665473478685325e-05,
+      "loss": 9.9958,
+      "step": 29360
+    },
+    {
+      "epoch": 0.14667032884716222,
+      "grad_norm": 0.10177190601825714,
+      "learning_rate": 2.5663971563743775e-05,
+      "loss": 9.9991,
+      "step": 29370
+    },
+    {
+      "epoch": 0.14672026767210167,
+      "grad_norm": 0.09410154819488525,
+      "learning_rate": 2.5662469648802225e-05,
+      "loss": 9.9934,
+      "step": 29380
+    },
+    {
+      "epoch": 0.14677020649704112,
+      "grad_norm": 0.09799119085073471,
+      "learning_rate": 2.5660967733860672e-05,
+      "loss": 9.9874,
+      "step": 29390
+    },
+    {
+      "epoch": 0.14682014532198057,
+      "grad_norm": 0.10114170610904694,
+      "learning_rate": 2.5659465818919122e-05,
+      "loss": 9.9944,
+      "step": 29400
+    },
+    {
+      "epoch": 0.14687008414692002,
+      "grad_norm": 0.09417644143104553,
+      "learning_rate": 2.5657963903977572e-05,
+      "loss": 9.9959,
+      "step": 29410
+    },
+    {
+      "epoch": 0.14692002297185947,
+      "grad_norm": 0.09454021602869034,
+      "learning_rate": 2.5656461989036023e-05,
+      "loss": 9.9909,
+      "step": 29420
+    },
+    {
+      "epoch": 0.14696996179679891,
+      "grad_norm": 0.09609093517065048,
+      "learning_rate": 2.5654960074094473e-05,
+      "loss": 9.9893,
+      "step": 29430
+    },
+    {
+      "epoch": 0.14701990062173836,
+      "grad_norm": 0.09451326727867126,
+      "learning_rate": 2.565345815915292e-05,
+      "loss": 9.9925,
+      "step": 29440
+    },
+    {
+      "epoch": 0.1470698394466778,
+      "grad_norm": 0.09044196456670761,
+      "learning_rate": 2.565195624421137e-05,
+      "loss": 9.9939,
+      "step": 29450
+    },
+    {
+      "epoch": 0.14711977827161726,
+      "grad_norm": 0.10042398422956467,
+      "learning_rate": 2.565045432926982e-05,
+      "loss": 9.9976,
+      "step": 29460
+    },
+    {
+      "epoch": 0.1471697170965567,
+      "grad_norm": 743.2418823242188,
+      "learning_rate": 2.564895241432827e-05,
+      "loss": 9.9953,
+      "step": 29470
+    },
+    {
+      "epoch": 0.14721965592149616,
+      "grad_norm": 0.10025718063116074,
+      "learning_rate": 2.564745049938672e-05,
+      "loss": 10.5647,
+      "step": 29480
+    },
+    {
+      "epoch": 0.1472695947464356,
+      "grad_norm": 0.0930456891655922,
+      "learning_rate": 2.5645948584445167e-05,
+      "loss": 9.9927,
+      "step": 29490
+    },
+    {
+      "epoch": 0.14731953357137506,
+      "grad_norm": 0.09810245782136917,
+      "learning_rate": 2.5644446669503617e-05,
+      "loss": 9.9906,
+      "step": 29500
+    },
+    {
+      "epoch": 0.1473694723963145,
+      "grad_norm": 0.0949150025844574,
+      "learning_rate": 2.5642944754562067e-05,
+      "loss": 11.4577,
+      "step": 29510
+    },
+    {
+      "epoch": 0.14741941122125396,
+      "grad_norm": 0.09857135266065598,
+      "learning_rate": 2.5641442839620518e-05,
+      "loss": 9.9955,
+      "step": 29520
+    },
+    {
+      "epoch": 0.1474693500461934,
+      "grad_norm": 0.09233613312244415,
+      "learning_rate": 2.5639940924678968e-05,
+      "loss": 9.9922,
+      "step": 29530
+    },
+    {
+      "epoch": 0.14751928887113286,
+      "grad_norm": 0.0993812084197998,
+      "learning_rate": 2.5638439009737415e-05,
+      "loss": 9.9864,
+      "step": 29540
+    },
+    {
+      "epoch": 0.1475692276960723,
+      "grad_norm": 0.09710720926523209,
+      "learning_rate": 2.5636937094795865e-05,
+      "loss": 9.9892,
+      "step": 29550
+    },
+    {
+      "epoch": 0.14761916652101176,
+      "grad_norm": 0.09210072457790375,
+      "learning_rate": 2.5635435179854315e-05,
+      "loss": 9.9928,
+      "step": 29560
+    },
+    {
+      "epoch": 0.1476691053459512,
+      "grad_norm": 0.0917966440320015,
+      "learning_rate": 2.5633933264912765e-05,
+      "loss": 9.9889,
+      "step": 29570
+    },
+    {
+      "epoch": 0.14771904417089066,
+      "grad_norm": 0.09141924977302551,
+      "learning_rate": 2.5632431349971215e-05,
+      "loss": 9.9931,
+      "step": 29580
+    },
+    {
+      "epoch": 0.1477689829958301,
+      "grad_norm": 0.09177865833044052,
+      "learning_rate": 2.5630929435029662e-05,
+      "loss": 9.9971,
+      "step": 29590
+    },
+    {
+      "epoch": 0.14781892182076956,
+      "grad_norm": 0.09628382325172424,
+      "learning_rate": 2.5629427520088112e-05,
+      "loss": 9.991,
+      "step": 29600
+    },
+    {
+      "epoch": 0.147868860645709,
+      "grad_norm": 0.10353731364011765,
+      "learning_rate": 2.5627925605146562e-05,
+      "loss": 9.9869,
+      "step": 29610
+    },
+    {
+      "epoch": 0.14791879947064845,
+      "grad_norm": 0.08936607092618942,
+      "learning_rate": 2.5626423690205013e-05,
+      "loss": 9.991,
+      "step": 29620
+    },
+    {
+      "epoch": 0.1479687382955879,
+      "grad_norm": 0.09186360985040665,
+      "learning_rate": 2.5624921775263463e-05,
+      "loss": 9.9913,
+      "step": 29630
+    },
+    {
+      "epoch": 0.14801867712052735,
+      "grad_norm": 0.09984660148620605,
+      "learning_rate": 2.5623419860321913e-05,
+      "loss": 9.9862,
+      "step": 29640
+    },
+    {
+      "epoch": 0.1480686159454668,
+      "grad_norm": 0.0906200036406517,
+      "learning_rate": 2.562191794538036e-05,
+      "loss": 9.9897,
+      "step": 29650
+    },
+    {
+      "epoch": 0.14811855477040625,
+      "grad_norm": 0.1004503145813942,
+      "learning_rate": 2.562041603043881e-05,
+      "loss": 9.9894,
+      "step": 29660
+    },
+    {
+      "epoch": 0.1481684935953457,
+      "grad_norm": 0.09145411103963852,
+      "learning_rate": 2.561891411549726e-05,
+      "loss": 9.9877,
+      "step": 29670
+    },
+    {
+      "epoch": 0.14821843242028515,
+      "grad_norm": 0.09083334356546402,
+      "learning_rate": 2.561741220055571e-05,
+      "loss": 9.9825,
+      "step": 29680
+    },
+    {
+      "epoch": 0.1482683712452246,
+      "grad_norm": 0.1027732640504837,
+      "learning_rate": 2.561591028561416e-05,
+      "loss": 9.988,
+      "step": 29690
+    },
+    {
+      "epoch": 0.14831831007016405,
+      "grad_norm": 0.09426605701446533,
+      "learning_rate": 2.5614408370672607e-05,
+      "loss": 9.986,
+      "step": 29700
+    },
+    {
+      "epoch": 0.1483682488951035,
+      "grad_norm": 0.09159605205059052,
+      "learning_rate": 2.5612906455731057e-05,
+      "loss": 9.9925,
+      "step": 29710
+    },
+    {
+      "epoch": 0.14841818772004295,
+      "grad_norm": 0.09627106785774231,
+      "learning_rate": 2.5611404540789508e-05,
+      "loss": 9.9841,
+      "step": 29720
+    },
+    {
+      "epoch": 0.1484681265449824,
+      "grad_norm": 0.09180369973182678,
+      "learning_rate": 2.5609902625847958e-05,
+      "loss": 9.992,
+      "step": 29730
+    },
+    {
+      "epoch": 0.14851806536992185,
+      "grad_norm": 0.093332439661026,
+      "learning_rate": 2.5608400710906408e-05,
+      "loss": 9.9877,
+      "step": 29740
+    },
+    {
+      "epoch": 0.1485680041948613,
+      "grad_norm": 0.09420640766620636,
+      "learning_rate": 2.5606898795964855e-05,
+      "loss": 9.9862,
+      "step": 29750
+    },
+    {
+      "epoch": 0.14861794301980075,
+      "grad_norm": 0.08834953606128693,
+      "learning_rate": 2.5605396881023305e-05,
+      "loss": 9.9859,
+      "step": 29760
+    },
+    {
+      "epoch": 0.1486678818447402,
+      "grad_norm": 0.09357737004756927,
+      "learning_rate": 2.5603894966081755e-05,
+      "loss": 9.9881,
+      "step": 29770
+    },
+    {
+      "epoch": 0.14871782066967965,
+      "grad_norm": 0.09611498564481735,
+      "learning_rate": 2.5602393051140205e-05,
+      "loss": 9.9835,
+      "step": 29780
+    },
+    {
+      "epoch": 0.1487677594946191,
+      "grad_norm": 0.09604179859161377,
+      "learning_rate": 2.5600891136198655e-05,
+      "loss": 9.9854,
+      "step": 29790
+    },
+    {
+      "epoch": 0.14881769831955854,
+      "grad_norm": 0.0894085168838501,
+      "learning_rate": 2.5599389221257102e-05,
+      "loss": 9.9843,
+      "step": 29800
+    },
+    {
+      "epoch": 0.148867637144498,
+      "grad_norm": 0.0987863540649414,
+      "learning_rate": 2.5597887306315552e-05,
+      "loss": 9.9861,
+      "step": 29810
+    },
+    {
+      "epoch": 0.14891757596943744,
+      "grad_norm": 0.09360215812921524,
+      "learning_rate": 2.5596385391374003e-05,
+      "loss": 9.979,
+      "step": 29820
+    },
+    {
+      "epoch": 0.1489675147943769,
+      "grad_norm": 0.09655702859163284,
+      "learning_rate": 2.5594883476432453e-05,
+      "loss": 9.9865,
+      "step": 29830
+    },
+    {
+      "epoch": 0.14901745361931634,
+      "grad_norm": 0.09509619325399399,
+      "learning_rate": 2.5593381561490903e-05,
+      "loss": 9.982,
+      "step": 29840
+    },
+    {
+      "epoch": 0.1490673924442558,
+      "grad_norm": 0.09575647860765457,
+      "learning_rate": 2.559187964654935e-05,
+      "loss": 9.9806,
+      "step": 29850
+    },
+    {
+      "epoch": 0.14911733126919524,
+      "grad_norm": 0.09219710528850555,
+      "learning_rate": 2.55903777316078e-05,
+      "loss": 9.9857,
+      "step": 29860
+    },
+    {
+      "epoch": 0.1491672700941347,
+      "grad_norm": 0.09096106886863708,
+      "learning_rate": 2.558887581666625e-05,
+      "loss": 9.9862,
+      "step": 29870
+    },
+    {
+      "epoch": 0.14921720891907414,
+      "grad_norm": 0.09332380443811417,
+      "learning_rate": 2.55873739017247e-05,
+      "loss": 9.9837,
+      "step": 29880
+    },
+    {
+      "epoch": 0.1492671477440136,
+      "grad_norm": 0.09098540246486664,
+      "learning_rate": 2.558587198678315e-05,
+      "loss": 9.9893,
+      "step": 29890
+    },
+    {
+      "epoch": 0.14931708656895304,
+      "grad_norm": 0.09403214603662491,
+      "learning_rate": 2.5584370071841597e-05,
+      "loss": 9.9855,
+      "step": 29900
+    },
+    {
+      "epoch": 0.1493670253938925,
+      "grad_norm": 0.10132903605699539,
+      "learning_rate": 2.5582868156900047e-05,
+      "loss": 9.9836,
+      "step": 29910
+    },
+    {
+      "epoch": 0.14941696421883194,
+      "grad_norm": 0.09413734078407288,
+      "learning_rate": 2.5581366241958498e-05,
+      "loss": 9.9824,
+      "step": 29920
+    },
+    {
+      "epoch": 0.1494669030437714,
+      "grad_norm": 0.09860097616910934,
+      "learning_rate": 2.5579864327016948e-05,
+      "loss": 9.9844,
+      "step": 29930
+    },
+    {
+      "epoch": 0.14951684186871084,
+      "grad_norm": 0.09427344053983688,
+      "learning_rate": 2.5578362412075398e-05,
+      "loss": 9.9842,
+      "step": 29940
+    },
+    {
+      "epoch": 0.1495667806936503,
+      "grad_norm": 0.09929951280355453,
+      "learning_rate": 2.5576860497133845e-05,
+      "loss": 9.9852,
+      "step": 29950
+    },
+    {
+      "epoch": 0.14961671951858974,
+      "grad_norm": 0.08913746476173401,
+      "learning_rate": 2.5575358582192295e-05,
+      "loss": 9.988,
+      "step": 29960
+    },
+    {
+      "epoch": 0.14966665834352919,
+      "grad_norm": 0.09487251192331314,
+      "learning_rate": 2.5573856667250745e-05,
+      "loss": 9.9882,
+      "step": 29970
+    },
+    {
+      "epoch": 0.14971659716846863,
+      "grad_norm": 0.0945911556482315,
+      "learning_rate": 2.5572354752309195e-05,
+      "loss": 9.9848,
+      "step": 29980
+    },
+    {
+      "epoch": 0.14976653599340808,
+      "grad_norm": 0.08827967196702957,
+      "learning_rate": 2.5570852837367645e-05,
+      "loss": 9.9878,
+      "step": 29990
+    },
+    {
+      "epoch": 0.14981647481834753,
+      "grad_norm": 0.09094835072755814,
+      "learning_rate": 2.5569350922426092e-05,
+      "loss": 9.9809,
+      "step": 30000
+    },
+    {
+      "epoch": 0.14986641364328698,
+      "grad_norm": 0.0922902449965477,
+      "learning_rate": 2.5567849007484546e-05,
+      "loss": 9.9868,
+      "step": 30010
+    },
+    {
+      "epoch": 0.14991635246822643,
+      "grad_norm": 0.09045281261205673,
+      "learning_rate": 2.5566347092542993e-05,
+      "loss": 9.9841,
+      "step": 30020
+    },
+    {
+      "epoch": 0.14996629129316588,
+      "grad_norm": 0.09216750413179398,
+      "learning_rate": 2.5564845177601443e-05,
+      "loss": 9.9797,
+      "step": 30030
+    },
+    {
+      "epoch": 0.15001623011810533,
+      "grad_norm": 0.09659641981124878,
+      "learning_rate": 2.5563343262659893e-05,
+      "loss": 9.9785,
+      "step": 30040
+    },
+    {
+      "epoch": 0.15006616894304478,
+      "grad_norm": 0.09455203264951706,
+      "learning_rate": 2.556184134771834e-05,
+      "loss": 9.9815,
+      "step": 30050
+    },
+    {
+      "epoch": 0.15011610776798423,
+      "grad_norm": 0.089305080473423,
+      "learning_rate": 2.5560339432776793e-05,
+      "loss": 9.9792,
+      "step": 30060
+    },
+    {
+      "epoch": 0.15016604659292368,
+      "grad_norm": 0.09118703007698059,
+      "learning_rate": 2.555883751783524e-05,
+      "loss": 9.9802,
+      "step": 30070
+    },
+    {
+      "epoch": 0.15021598541786313,
+      "grad_norm": 0.09706766903400421,
+      "learning_rate": 2.555733560289369e-05,
+      "loss": 9.9764,
+      "step": 30080
+    },
+    {
+      "epoch": 0.15026592424280258,
+      "grad_norm": 0.09246896207332611,
+      "learning_rate": 2.555583368795214e-05,
+      "loss": 9.981,
+      "step": 30090
+    },
+    {
+      "epoch": 0.15031586306774203,
+      "grad_norm": 0.09420756250619888,
+      "learning_rate": 2.5554331773010587e-05,
+      "loss": 9.9817,
+      "step": 30100
+    },
+    {
+      "epoch": 0.15036580189268148,
+      "grad_norm": 0.0959128737449646,
+      "learning_rate": 2.555282985806904e-05,
+      "loss": 9.9834,
+      "step": 30110
+    },
+    {
+      "epoch": 0.15041574071762093,
+      "grad_norm": 0.09950920939445496,
+      "learning_rate": 2.5551327943127488e-05,
+      "loss": 9.979,
+      "step": 30120
+    },
+    {
+      "epoch": 0.15046567954256038,
+      "grad_norm": 0.09626981616020203,
+      "learning_rate": 2.5549826028185938e-05,
+      "loss": 9.974,
+      "step": 30130
+    },
+    {
+      "epoch": 0.15051561836749983,
+      "grad_norm": 0.09594815224409103,
+      "learning_rate": 2.5548324113244388e-05,
+      "loss": 9.9792,
+      "step": 30140
+    },
+    {
+      "epoch": 0.15056555719243928,
+      "grad_norm": 0.10035426914691925,
+      "learning_rate": 2.5546822198302835e-05,
+      "loss": 9.976,
+      "step": 30150
+    },
+    {
+      "epoch": 0.15061549601737872,
+      "grad_norm": 0.09873773902654648,
+      "learning_rate": 2.554532028336129e-05,
+      "loss": 9.9735,
+      "step": 30160
+    },
+    {
+      "epoch": 0.15066543484231815,
+      "grad_norm": 0.09737424552440643,
+      "learning_rate": 2.5543818368419735e-05,
+      "loss": 9.9808,
+      "step": 30170
+    },
+    {
+      "epoch": 0.1507153736672576,
+      "grad_norm": 0.09332314133644104,
+      "learning_rate": 2.5542316453478185e-05,
+      "loss": 9.9723,
+      "step": 30180
+    },
+    {
+      "epoch": 0.15076531249219705,
+      "grad_norm": 0.09951841086149216,
+      "learning_rate": 2.5540814538536636e-05,
+      "loss": 9.9775,
+      "step": 30190
+    },
+    {
+      "epoch": 0.1508152513171365,
+      "grad_norm": 0.09631986916065216,
+      "learning_rate": 2.5539312623595082e-05,
+      "loss": 9.9765,
+      "step": 30200
+    },
+    {
+      "epoch": 0.15086519014207594,
+      "grad_norm": 0.08957190811634064,
+      "learning_rate": 2.5537810708653536e-05,
+      "loss": 9.985,
+      "step": 30210
+    },
+    {
+      "epoch": 0.1509151289670154,
+      "grad_norm": 0.09298118948936462,
+      "learning_rate": 2.5536308793711983e-05,
+      "loss": 9.9777,
+      "step": 30220
+    },
+    {
+      "epoch": 0.15096506779195484,
+      "grad_norm": 0.09950188547372818,
+      "learning_rate": 2.5534806878770433e-05,
+      "loss": 9.9758,
+      "step": 30230
+    },
+    {
+      "epoch": 0.1510150066168943,
+      "grad_norm": 0.09786109626293182,
+      "learning_rate": 2.5533304963828883e-05,
+      "loss": 9.9709,
+      "step": 30240
+    },
+    {
+      "epoch": 0.15106494544183374,
+      "grad_norm": 0.09600091725587845,
+      "learning_rate": 2.553180304888733e-05,
+      "loss": 9.9772,
+      "step": 30250
+    },
+    {
+      "epoch": 0.1511148842667732,
+      "grad_norm": 0.0918637216091156,
+      "learning_rate": 2.5530301133945783e-05,
+      "loss": 9.9781,
+      "step": 30260
+    },
+    {
+      "epoch": 0.15116482309171264,
+      "grad_norm": 0.09260197728872299,
+      "learning_rate": 2.552879921900423e-05,
+      "loss": 9.9701,
+      "step": 30270
+    },
+    {
+      "epoch": 0.1512147619166521,
+      "grad_norm": 0.0909905806183815,
+      "learning_rate": 2.552729730406268e-05,
+      "loss": 9.9747,
+      "step": 30280
+    },
+    {
+      "epoch": 0.15126470074159154,
+      "grad_norm": 0.09207629412412643,
+      "learning_rate": 2.552579538912113e-05,
+      "loss": 9.9713,
+      "step": 30290
+    },
+    {
+      "epoch": 0.151314639566531,
+      "grad_norm": 0.09557502716779709,
+      "learning_rate": 2.5524293474179577e-05,
+      "loss": 9.9844,
+      "step": 30300
+    },
+    {
+      "epoch": 0.15136457839147044,
+      "grad_norm": 0.0928947702050209,
+      "learning_rate": 2.552279155923803e-05,
+      "loss": 9.9739,
+      "step": 30310
+    },
+    {
+      "epoch": 0.1514145172164099,
+      "grad_norm": 0.09521695226430893,
+      "learning_rate": 2.5521289644296478e-05,
+      "loss": 9.9733,
+      "step": 30320
+    },
+    {
+      "epoch": 0.15146445604134934,
+      "grad_norm": 0.0963277593255043,
+      "learning_rate": 2.551978772935493e-05,
+      "loss": 9.969,
+      "step": 30330
+    },
+    {
+      "epoch": 0.1515143948662888,
+      "grad_norm": 0.10125837475061417,
+      "learning_rate": 2.5518285814413378e-05,
+      "loss": 9.9739,
+      "step": 30340
+    },
+    {
+      "epoch": 0.15156433369122824,
+      "grad_norm": 0.09548031538724899,
+      "learning_rate": 2.5516783899471825e-05,
+      "loss": 9.9702,
+      "step": 30350
+    },
+    {
+      "epoch": 0.1516142725161677,
+      "grad_norm": 0.09445524960756302,
+      "learning_rate": 2.551528198453028e-05,
+      "loss": 9.9767,
+      "step": 30360
+    },
+    {
+      "epoch": 0.15166421134110714,
+      "grad_norm": 0.09280666708946228,
+      "learning_rate": 2.5513780069588725e-05,
+      "loss": 9.9779,
+      "step": 30370
+    },
+    {
+      "epoch": 0.15171415016604659,
+      "grad_norm": 0.09591789543628693,
+      "learning_rate": 2.551227815464718e-05,
+      "loss": 9.9722,
+      "step": 30380
+    },
+    {
+      "epoch": 0.15176408899098603,
+      "grad_norm": 0.09540422260761261,
+      "learning_rate": 2.5510776239705626e-05,
+      "loss": 9.9732,
+      "step": 30390
+    },
+    {
+      "epoch": 0.15181402781592548,
+      "grad_norm": 0.0966714397072792,
+      "learning_rate": 2.5509274324764072e-05,
+      "loss": 9.9744,
+      "step": 30400
+    },
+    {
+      "epoch": 0.15186396664086493,
+      "grad_norm": 0.09428321570158005,
+      "learning_rate": 2.5507772409822526e-05,
+      "loss": 9.9751,
+      "step": 30410
+    },
+    {
+      "epoch": 0.15191390546580438,
+      "grad_norm": 0.09702390432357788,
+      "learning_rate": 2.5506270494880973e-05,
+      "loss": 9.9724,
+      "step": 30420
+    },
+    {
+      "epoch": 0.15196384429074383,
+      "grad_norm": 0.09882203489542007,
+      "learning_rate": 2.5504768579939426e-05,
+      "loss": 9.9698,
+      "step": 30430
+    },
+    {
+      "epoch": 0.15201378311568328,
+      "grad_norm": 0.09188614040613174,
+      "learning_rate": 2.5503266664997873e-05,
+      "loss": 9.9745,
+      "step": 30440
+    },
+    {
+      "epoch": 0.15206372194062273,
+      "grad_norm": 0.09487085789442062,
+      "learning_rate": 2.550176475005632e-05,
+      "loss": 9.9672,
+      "step": 30450
+    },
+    {
+      "epoch": 0.15211366076556218,
+      "grad_norm": 0.09500377625226974,
+      "learning_rate": 2.5500262835114773e-05,
+      "loss": 9.9702,
+      "step": 30460
+    },
+    {
+      "epoch": 0.15216359959050163,
+      "grad_norm": 0.09745945781469345,
+      "learning_rate": 2.549876092017322e-05,
+      "loss": 9.9739,
+      "step": 30470
+    },
+    {
+      "epoch": 0.15221353841544108,
+      "grad_norm": 0.0949743390083313,
+      "learning_rate": 2.5497259005231674e-05,
+      "loss": 9.9646,
+      "step": 30480
+    },
+    {
+      "epoch": 0.15226347724038053,
+      "grad_norm": 0.09280611574649811,
+      "learning_rate": 2.549575709029012e-05,
+      "loss": 9.967,
+      "step": 30490
+    },
+    {
+      "epoch": 0.15231341606531998,
+      "grad_norm": 0.09759785979986191,
+      "learning_rate": 2.5494255175348567e-05,
+      "loss": 9.9691,
+      "step": 30500
+    },
+    {
+      "epoch": 0.15236335489025943,
+      "grad_norm": 0.10251536220312119,
+      "learning_rate": 2.549275326040702e-05,
+      "loss": 9.968,
+      "step": 30510
+    },
+    {
+      "epoch": 0.15241329371519888,
+      "grad_norm": 0.09339610487222672,
+      "learning_rate": 2.5491251345465468e-05,
+      "loss": 9.9709,
+      "step": 30520
+    },
+    {
+      "epoch": 0.15246323254013833,
+      "grad_norm": 0.09631470590829849,
+      "learning_rate": 2.548974943052392e-05,
+      "loss": 9.9649,
+      "step": 30530
+    },
+    {
+      "epoch": 0.15251317136507778,
+      "grad_norm": 0.09462681412696838,
+      "learning_rate": 2.5488247515582368e-05,
+      "loss": 9.9656,
+      "step": 30540
+    },
+    {
+      "epoch": 0.15256311019001723,
+      "grad_norm": 0.09926269203424454,
+      "learning_rate": 2.5486745600640815e-05,
+      "loss": 9.9666,
+      "step": 30550
+    },
+    {
+      "epoch": 0.15261304901495668,
+      "grad_norm": 0.09245786815881729,
+      "learning_rate": 2.548524368569927e-05,
+      "loss": 9.9648,
+      "step": 30560
+    },
+    {
+      "epoch": 0.15266298783989612,
+      "grad_norm": 0.09163583815097809,
+      "learning_rate": 2.5483741770757715e-05,
+      "loss": 9.9667,
+      "step": 30570
+    },
+    {
+      "epoch": 0.15271292666483557,
+      "grad_norm": 0.0959586575627327,
+      "learning_rate": 2.548223985581617e-05,
+      "loss": 9.9713,
+      "step": 30580
+    },
+    {
+      "epoch": 0.15276286548977502,
+      "grad_norm": 0.09638317674398422,
+      "learning_rate": 2.5480737940874616e-05,
+      "loss": 9.9625,
+      "step": 30590
+    },
+    {
+      "epoch": 0.15281280431471447,
+      "grad_norm": 0.0959702804684639,
+      "learning_rate": 2.5479236025933062e-05,
+      "loss": 9.9676,
+      "step": 30600
+    },
+    {
+      "epoch": 0.15286274313965392,
+      "grad_norm": 0.09670331329107285,
+      "learning_rate": 2.5477734110991516e-05,
+      "loss": 9.9673,
+      "step": 30610
+    },
+    {
+      "epoch": 0.15291268196459337,
+      "grad_norm": 0.09731889516115189,
+      "learning_rate": 2.5476232196049963e-05,
+      "loss": 9.9669,
+      "step": 30620
+    },
+    {
+      "epoch": 0.15296262078953282,
+      "grad_norm": 0.09183140099048615,
+      "learning_rate": 2.5474730281108416e-05,
+      "loss": 9.9672,
+      "step": 30630
+    },
+    {
+      "epoch": 0.15301255961447227,
+      "grad_norm": 0.09476214647293091,
+      "learning_rate": 2.5473228366166863e-05,
+      "loss": 9.9662,
+      "step": 30640
+    },
+    {
+      "epoch": 0.15306249843941172,
+      "grad_norm": 0.09765252470970154,
+      "learning_rate": 2.5471726451225313e-05,
+      "loss": 9.9693,
+      "step": 30650
+    },
+    {
+      "epoch": 0.15311243726435117,
+      "grad_norm": 0.0965784415602684,
+      "learning_rate": 2.5470224536283763e-05,
+      "loss": 9.9612,
+      "step": 30660
+    },
+    {
+      "epoch": 0.15316237608929062,
+      "grad_norm": 0.0938754677772522,
+      "learning_rate": 2.546872262134221e-05,
+      "loss": 9.9638,
+      "step": 30670
+    },
+    {
+      "epoch": 0.15321231491423007,
+      "grad_norm": 0.09392806887626648,
+      "learning_rate": 2.5467220706400664e-05,
+      "loss": 9.9641,
+      "step": 30680
+    },
+    {
+      "epoch": 0.15326225373916952,
+      "grad_norm": 0.09800152480602264,
+      "learning_rate": 2.546571879145911e-05,
+      "loss": 9.9662,
+      "step": 30690
+    },
+    {
+      "epoch": 0.15331219256410897,
+      "grad_norm": 0.09477967023849487,
+      "learning_rate": 2.546421687651756e-05,
+      "loss": 9.9667,
+      "step": 30700
+    },
+    {
+      "epoch": 0.15336213138904842,
+      "grad_norm": 0.09264533221721649,
+      "learning_rate": 2.546271496157601e-05,
+      "loss": 9.966,
+      "step": 30710
+    },
+    {
+      "epoch": 0.15341207021398787,
+      "grad_norm": 0.09792211651802063,
+      "learning_rate": 2.5461213046634458e-05,
+      "loss": 9.9652,
+      "step": 30720
+    },
+    {
+      "epoch": 0.15346200903892732,
+      "grad_norm": 0.09196563065052032,
+      "learning_rate": 2.545971113169291e-05,
+      "loss": 9.9634,
+      "step": 30730
+    },
+    {
+      "epoch": 0.15351194786386677,
+      "grad_norm": 0.09375791996717453,
+      "learning_rate": 2.5458209216751358e-05,
+      "loss": 9.9615,
+      "step": 30740
+    },
+    {
+      "epoch": 0.15356188668880622,
+      "grad_norm": 0.10490723699331284,
+      "learning_rate": 2.5456707301809808e-05,
+      "loss": 9.9654,
+      "step": 30750
+    },
+    {
+      "epoch": 0.15361182551374566,
+      "grad_norm": 0.08919631689786911,
+      "learning_rate": 2.545520538686826e-05,
+      "loss": 9.9659,
+      "step": 30760
+    },
+    {
+      "epoch": 0.15366176433868511,
+      "grad_norm": 0.09268412739038467,
+      "learning_rate": 2.5453703471926705e-05,
+      "loss": 9.9595,
+      "step": 30770
+    },
+    {
+      "epoch": 0.15371170316362456,
+      "grad_norm": 0.09661845862865448,
+      "learning_rate": 2.545220155698516e-05,
+      "loss": 9.9571,
+      "step": 30780
+    },
+    {
+      "epoch": 0.153761641988564,
+      "grad_norm": 0.0946783795952797,
+      "learning_rate": 2.5450699642043606e-05,
+      "loss": 9.9624,
+      "step": 30790
+    },
+    {
+      "epoch": 0.15381158081350346,
+      "grad_norm": 0.08772801607847214,
+      "learning_rate": 2.5449197727102056e-05,
+      "loss": 9.9651,
+      "step": 30800
+    },
+    {
+      "epoch": 0.1538615196384429,
+      "grad_norm": 0.09275208413600922,
+      "learning_rate": 2.5447695812160506e-05,
+      "loss": 9.9595,
+      "step": 30810
+    },
+    {
+      "epoch": 0.15391145846338236,
+      "grad_norm": 0.09553563594818115,
+      "learning_rate": 2.5446193897218953e-05,
+      "loss": 9.9569,
+      "step": 30820
+    },
+    {
+      "epoch": 0.1539613972883218,
+      "grad_norm": 0.09462607651948929,
+      "learning_rate": 2.5444691982277406e-05,
+      "loss": 9.9615,
+      "step": 30830
+    },
+    {
+      "epoch": 0.15401133611326126,
+      "grad_norm": 0.09698895364999771,
+      "learning_rate": 2.5443190067335853e-05,
+      "loss": 9.9661,
+      "step": 30840
+    },
+    {
+      "epoch": 0.1540612749382007,
+      "grad_norm": 0.09504923969507217,
+      "learning_rate": 2.5441688152394303e-05,
+      "loss": 9.9681,
+      "step": 30850
+    },
+    {
+      "epoch": 0.15411121376314016,
+      "grad_norm": 0.10031065344810486,
+      "learning_rate": 2.5440186237452753e-05,
+      "loss": 9.9588,
+      "step": 30860
+    },
+    {
+      "epoch": 0.1541611525880796,
+      "grad_norm": 0.09581837058067322,
+      "learning_rate": 2.54386843225112e-05,
+      "loss": 9.959,
+      "step": 30870
+    },
+    {
+      "epoch": 0.15421109141301906,
+      "grad_norm": 0.09290026128292084,
+      "learning_rate": 2.5437182407569654e-05,
+      "loss": 9.9649,
+      "step": 30880
+    },
+    {
+      "epoch": 0.1542610302379585,
+      "grad_norm": 0.0961495116353035,
+      "learning_rate": 2.54356804926281e-05,
+      "loss": 9.9566,
+      "step": 30890
+    },
+    {
+      "epoch": 0.15431096906289796,
+      "grad_norm": 0.09705748409032822,
+      "learning_rate": 2.543417857768655e-05,
+      "loss": 9.9632,
+      "step": 30900
+    },
+    {
+      "epoch": 0.1543609078878374,
+      "grad_norm": 0.09860498458147049,
+      "learning_rate": 2.5432676662745e-05,
+      "loss": 9.957,
+      "step": 30910
+    },
+    {
+      "epoch": 0.15441084671277686,
+      "grad_norm": 0.09649617969989777,
+      "learning_rate": 2.5431174747803448e-05,
+      "loss": 9.9609,
+      "step": 30920
+    },
+    {
+      "epoch": 0.1544607855377163,
+      "grad_norm": 0.09514536708593369,
+      "learning_rate": 2.54296728328619e-05,
+      "loss": 9.9607,
+      "step": 30930
+    },
+    {
+      "epoch": 0.15451072436265575,
+      "grad_norm": 0.09036850929260254,
+      "learning_rate": 2.5428170917920348e-05,
+      "loss": 9.9562,
+      "step": 30940
+    },
+    {
+      "epoch": 0.1545606631875952,
+      "grad_norm": 0.09239833056926727,
+      "learning_rate": 2.5426669002978798e-05,
+      "loss": 9.9597,
+      "step": 30950
+    },
+    {
+      "epoch": 0.15461060201253465,
+      "grad_norm": 0.09781666100025177,
+      "learning_rate": 2.542516708803725e-05,
+      "loss": 9.9674,
+      "step": 30960
+    },
+    {
+      "epoch": 0.1546605408374741,
+      "grad_norm": 0.09337397664785385,
+      "learning_rate": 2.54236651730957e-05,
+      "loss": 9.9614,
+      "step": 30970
+    },
+    {
+      "epoch": 0.15471047966241355,
+      "grad_norm": 0.09049335867166519,
+      "learning_rate": 2.542216325815415e-05,
+      "loss": 9.9614,
+      "step": 30980
+    },
+    {
+      "epoch": 0.154760418487353,
+      "grad_norm": 0.09918399900197983,
+      "learning_rate": 2.5420661343212596e-05,
+      "loss": 9.9559,
+      "step": 30990
+    },
+    {
+      "epoch": 0.15481035731229245,
+      "grad_norm": 0.09200762212276459,
+      "learning_rate": 2.5419159428271046e-05,
+      "loss": 9.9508,
+      "step": 31000
+    },
+    {
+      "epoch": 0.1548602961372319,
+      "grad_norm": 0.09439615160226822,
+      "learning_rate": 2.5417657513329496e-05,
+      "loss": 9.9625,
+      "step": 31010
+    },
+    {
+      "epoch": 0.15491023496217135,
+      "grad_norm": 0.09249376505613327,
+      "learning_rate": 2.5416155598387946e-05,
+      "loss": 9.9598,
+      "step": 31020
+    },
+    {
+      "epoch": 0.1549601737871108,
+      "grad_norm": 0.09385478496551514,
+      "learning_rate": 2.5414653683446396e-05,
+      "loss": 9.9589,
+      "step": 31030
+    },
+    {
+      "epoch": 0.15501011261205025,
+      "grad_norm": 0.0966297909617424,
+      "learning_rate": 2.5413151768504843e-05,
+      "loss": 9.9642,
+      "step": 31040
+    },
+    {
+      "epoch": 0.1550600514369897,
+      "grad_norm": 0.09643752872943878,
+      "learning_rate": 2.5411649853563293e-05,
+      "loss": 9.9559,
+      "step": 31050
+    },
+    {
+      "epoch": 0.15510999026192915,
+      "grad_norm": 0.09509705007076263,
+      "learning_rate": 2.5410147938621743e-05,
+      "loss": 9.9555,
+      "step": 31060
+    },
+    {
+      "epoch": 0.1551599290868686,
+      "grad_norm": 0.09453869611024857,
+      "learning_rate": 2.5408646023680194e-05,
+      "loss": 9.9591,
+      "step": 31070
+    },
+    {
+      "epoch": 0.15520986791180805,
+      "grad_norm": 0.09096340090036392,
+      "learning_rate": 2.5407144108738644e-05,
+      "loss": 9.9628,
+      "step": 31080
+    },
+    {
+      "epoch": 0.1552598067367475,
+      "grad_norm": 0.09323681145906448,
+      "learning_rate": 2.540564219379709e-05,
+      "loss": 9.959,
+      "step": 31090
+    },
+    {
+      "epoch": 0.15530974556168695,
+      "grad_norm": 0.09170509874820709,
+      "learning_rate": 2.540414027885554e-05,
+      "loss": 9.9538,
+      "step": 31100
+    },
+    {
+      "epoch": 0.1553596843866264,
+      "grad_norm": 0.09655750542879105,
+      "learning_rate": 2.540263836391399e-05,
+      "loss": 9.9561,
+      "step": 31110
+    },
+    {
+      "epoch": 0.15540962321156584,
+      "grad_norm": 0.09387839585542679,
+      "learning_rate": 2.540113644897244e-05,
+      "loss": 9.9571,
+      "step": 31120
+    },
+    {
+      "epoch": 0.1554595620365053,
+      "grad_norm": 0.09941628575325012,
+      "learning_rate": 2.539963453403089e-05,
+      "loss": 9.9566,
+      "step": 31130
+    },
+    {
+      "epoch": 0.15550950086144474,
+      "grad_norm": 0.09646540135145187,
+      "learning_rate": 2.5398132619089338e-05,
+      "loss": 9.9543,
+      "step": 31140
+    },
+    {
+      "epoch": 0.15555943968638417,
+      "grad_norm": 0.10054589062929153,
+      "learning_rate": 2.5396630704147788e-05,
+      "loss": 9.9528,
+      "step": 31150
+    },
+    {
+      "epoch": 0.15560937851132361,
+      "grad_norm": 0.09326030313968658,
+      "learning_rate": 2.539512878920624e-05,
+      "loss": 9.962,
+      "step": 31160
+    },
+    {
+      "epoch": 0.15565931733626306,
+      "grad_norm": 0.08991136401891708,
+      "learning_rate": 2.539362687426469e-05,
+      "loss": 9.9522,
+      "step": 31170
+    },
+    {
+      "epoch": 0.15570925616120251,
+      "grad_norm": 0.09380258619785309,
+      "learning_rate": 2.539212495932314e-05,
+      "loss": 9.9584,
+      "step": 31180
+    },
+    {
+      "epoch": 0.15575919498614196,
+      "grad_norm": 0.09635492414236069,
+      "learning_rate": 2.5390623044381586e-05,
+      "loss": 9.9486,
+      "step": 31190
+    },
+    {
+      "epoch": 0.1558091338110814,
+      "grad_norm": 0.09452281892299652,
+      "learning_rate": 2.5389121129440036e-05,
+      "loss": 9.9505,
+      "step": 31200
+    },
+    {
+      "epoch": 0.15585907263602086,
+      "grad_norm": 0.09984589368104935,
+      "learning_rate": 2.5387619214498486e-05,
+      "loss": 9.9507,
+      "step": 31210
+    },
+    {
+      "epoch": 0.1559090114609603,
+      "grad_norm": 0.09607294946908951,
+      "learning_rate": 2.5386117299556936e-05,
+      "loss": 9.9526,
+      "step": 31220
+    },
+    {
+      "epoch": 0.15595895028589976,
+      "grad_norm": 0.0886487290263176,
+      "learning_rate": 2.5384615384615386e-05,
+      "loss": 9.9552,
+      "step": 31230
+    },
+    {
+      "epoch": 0.1560088891108392,
+      "grad_norm": 0.09027554839849472,
+      "learning_rate": 2.5383113469673833e-05,
+      "loss": 9.958,
+      "step": 31240
+    },
+    {
+      "epoch": 0.15605882793577866,
+      "grad_norm": 0.09098269045352936,
+      "learning_rate": 2.5381611554732283e-05,
+      "loss": 9.9465,
+      "step": 31250
+    },
+    {
+      "epoch": 0.1561087667607181,
+      "grad_norm": 0.10451994836330414,
+      "learning_rate": 2.5380109639790733e-05,
+      "loss": 9.9555,
+      "step": 31260
+    },
+    {
+      "epoch": 0.15615870558565756,
+      "grad_norm": 0.09637783467769623,
+      "learning_rate": 2.5378607724849184e-05,
+      "loss": 9.9493,
+      "step": 31270
+    },
+    {
+      "epoch": 0.156208644410597,
+      "grad_norm": 0.09026754647493362,
+      "learning_rate": 2.5377105809907634e-05,
+      "loss": 9.9526,
+      "step": 31280
+    },
+    {
+      "epoch": 0.15625858323553646,
+      "grad_norm": 0.09922222793102264,
+      "learning_rate": 2.5375603894966084e-05,
+      "loss": 9.9455,
+      "step": 31290
+    },
+    {
+      "epoch": 0.1563085220604759,
+      "grad_norm": 0.09385956078767776,
+      "learning_rate": 2.537410198002453e-05,
+      "loss": 9.9496,
+      "step": 31300
+    },
+    {
+      "epoch": 0.15635846088541536,
+      "grad_norm": 0.09471774101257324,
+      "learning_rate": 2.537260006508298e-05,
+      "loss": 9.9536,
+      "step": 31310
+    },
+    {
+      "epoch": 0.1564083997103548,
+      "grad_norm": 0.09749903529882431,
+      "learning_rate": 2.537109815014143e-05,
+      "loss": 9.9501,
+      "step": 31320
+    },
+    {
+      "epoch": 0.15645833853529426,
+      "grad_norm": 0.09855715185403824,
+      "learning_rate": 2.536959623519988e-05,
+      "loss": 9.9486,
+      "step": 31330
+    },
+    {
+      "epoch": 0.1565082773602337,
+      "grad_norm": 0.0999755784869194,
+      "learning_rate": 2.536809432025833e-05,
+      "loss": 9.9482,
+      "step": 31340
+    },
+    {
+      "epoch": 0.15655821618517315,
+      "grad_norm": 0.10329025983810425,
+      "learning_rate": 2.5366592405316778e-05,
+      "loss": 9.9535,
+      "step": 31350
+    },
+    {
+      "epoch": 0.1566081550101126,
+      "grad_norm": 0.09251412004232407,
+      "learning_rate": 2.536509049037523e-05,
+      "loss": 9.9509,
+      "step": 31360
+    },
+    {
+      "epoch": 0.15665809383505205,
+      "grad_norm": 0.09849464893341064,
+      "learning_rate": 2.536358857543368e-05,
+      "loss": 9.9476,
+      "step": 31370
+    },
+    {
+      "epoch": 0.1567080326599915,
+      "grad_norm": 0.09341175109148026,
+      "learning_rate": 2.536208666049213e-05,
+      "loss": 9.9519,
+      "step": 31380
+    },
+    {
+      "epoch": 0.15675797148493095,
+      "grad_norm": 0.09484104812145233,
+      "learning_rate": 2.536058474555058e-05,
+      "loss": 9.9555,
+      "step": 31390
+    },
+    {
+      "epoch": 0.1568079103098704,
+      "grad_norm": 0.09805377572774887,
+      "learning_rate": 2.5359082830609026e-05,
+      "loss": 9.948,
+      "step": 31400
+    },
+    {
+      "epoch": 0.15685784913480985,
+      "grad_norm": 0.09305404126644135,
+      "learning_rate": 2.5357580915667476e-05,
+      "loss": 9.95,
+      "step": 31410
+    },
+    {
+      "epoch": 0.1569077879597493,
+      "grad_norm": 0.09033924341201782,
+      "learning_rate": 2.5356079000725926e-05,
+      "loss": 9.949,
+      "step": 31420
+    },
+    {
+      "epoch": 0.15695772678468875,
+      "grad_norm": 0.094557024538517,
+      "learning_rate": 2.5354577085784376e-05,
+      "loss": 9.9521,
+      "step": 31430
+    },
+    {
+      "epoch": 0.1570076656096282,
+      "grad_norm": 0.09300687164068222,
+      "learning_rate": 2.5353075170842826e-05,
+      "loss": 9.945,
+      "step": 31440
+    },
+    {
+      "epoch": 0.15705760443456765,
+      "grad_norm": 0.09191596508026123,
+      "learning_rate": 2.5351573255901273e-05,
+      "loss": 9.948,
+      "step": 31450
+    },
+    {
+      "epoch": 0.1571075432595071,
+      "grad_norm": 0.0944577232003212,
+      "learning_rate": 2.5350071340959723e-05,
+      "loss": 9.9515,
+      "step": 31460
+    },
+    {
+      "epoch": 0.15715748208444655,
+      "grad_norm": 0.09393108636140823,
+      "learning_rate": 2.5348569426018174e-05,
+      "loss": 9.9459,
+      "step": 31470
+    },
+    {
+      "epoch": 0.157207420909386,
+      "grad_norm": 0.09408476948738098,
+      "learning_rate": 2.5347067511076624e-05,
+      "loss": 9.9547,
+      "step": 31480
+    },
+    {
+      "epoch": 0.15725735973432545,
+      "grad_norm": 0.09239377081394196,
+      "learning_rate": 2.5345565596135074e-05,
+      "loss": 9.9479,
+      "step": 31490
+    },
+    {
+      "epoch": 0.1573072985592649,
+      "grad_norm": 0.09746348112821579,
+      "learning_rate": 2.534406368119352e-05,
+      "loss": 9.9445,
+      "step": 31500
+    },
+    {
+      "epoch": 0.15735723738420435,
+      "grad_norm": 0.09604940563440323,
+      "learning_rate": 2.534256176625197e-05,
+      "loss": 9.9443,
+      "step": 31510
+    },
+    {
+      "epoch": 0.1574071762091438,
+      "grad_norm": 0.09521540254354477,
+      "learning_rate": 2.534105985131042e-05,
+      "loss": 9.9499,
+      "step": 31520
+    },
+    {
+      "epoch": 0.15745711503408324,
+      "grad_norm": 0.09604847431182861,
+      "learning_rate": 2.533955793636887e-05,
+      "loss": 9.9507,
+      "step": 31530
+    },
+    {
+      "epoch": 0.1575070538590227,
+      "grad_norm": 0.09625346958637238,
+      "learning_rate": 2.533805602142732e-05,
+      "loss": 9.9499,
+      "step": 31540
+    },
+    {
+      "epoch": 0.15755699268396214,
+      "grad_norm": 0.09137043356895447,
+      "learning_rate": 2.5336554106485768e-05,
+      "loss": 9.9387,
+      "step": 31550
+    },
+    {
+      "epoch": 0.1576069315089016,
+      "grad_norm": 0.09646125137805939,
+      "learning_rate": 2.533505219154422e-05,
+      "loss": 9.9431,
+      "step": 31560
+    },
+    {
+      "epoch": 0.15765687033384104,
+      "grad_norm": 0.0926939845085144,
+      "learning_rate": 2.533355027660267e-05,
+      "loss": 9.9423,
+      "step": 31570
+    },
+    {
+      "epoch": 0.1577068091587805,
+      "grad_norm": 0.09652858227491379,
+      "learning_rate": 2.533204836166112e-05,
+      "loss": 9.9464,
+      "step": 31580
+    },
+    {
+      "epoch": 0.15775674798371994,
+      "grad_norm": 0.09520108252763748,
+      "learning_rate": 2.533054644671957e-05,
+      "loss": 9.946,
+      "step": 31590
+    },
+    {
+      "epoch": 0.1578066868086594,
+      "grad_norm": 0.0973607674241066,
+      "learning_rate": 2.5329044531778016e-05,
+      "loss": 9.94,
+      "step": 31600
+    },
+    {
+      "epoch": 0.15785662563359884,
+      "grad_norm": 0.09493687748908997,
+      "learning_rate": 2.532754261683647e-05,
+      "loss": 9.944,
+      "step": 31610
+    },
+    {
+      "epoch": 0.1579065644585383,
+      "grad_norm": 0.09770392626523972,
+      "learning_rate": 2.5326040701894916e-05,
+      "loss": 9.9393,
+      "step": 31620
+    },
+    {
+      "epoch": 0.15795650328347774,
+      "grad_norm": 0.09622366726398468,
+      "learning_rate": 2.5324538786953366e-05,
+      "loss": 9.9456,
+      "step": 31630
+    },
+    {
+      "epoch": 0.1580064421084172,
+      "grad_norm": 0.09520258009433746,
+      "learning_rate": 2.5323036872011817e-05,
+      "loss": 9.9408,
+      "step": 31640
+    },
+    {
+      "epoch": 0.15805638093335664,
+      "grad_norm": 0.09456585347652435,
+      "learning_rate": 2.5321534957070263e-05,
+      "loss": 9.9386,
+      "step": 31650
+    },
+    {
+      "epoch": 0.1581063197582961,
+      "grad_norm": 0.09494321048259735,
+      "learning_rate": 2.5320033042128717e-05,
+      "loss": 9.9401,
+      "step": 31660
+    },
+    {
+      "epoch": 0.15815625858323554,
+      "grad_norm": 0.09259496629238129,
+      "learning_rate": 2.5318531127187164e-05,
+      "loss": 9.9453,
+      "step": 31670
+    },
+    {
+      "epoch": 0.158206197408175,
+      "grad_norm": 0.0940915197134018,
+      "learning_rate": 2.5317029212245614e-05,
+      "loss": 9.95,
+      "step": 31680
+    },
+    {
+      "epoch": 0.15825613623311444,
+      "grad_norm": 0.09100774675607681,
+      "learning_rate": 2.5315527297304064e-05,
+      "loss": 9.9416,
+      "step": 31690
+    },
+    {
+      "epoch": 0.15830607505805389,
+      "grad_norm": 0.09511730074882507,
+      "learning_rate": 2.531402538236251e-05,
+      "loss": 9.9399,
+      "step": 31700
+    },
+    {
+      "epoch": 0.15835601388299334,
+      "grad_norm": 0.0985988974571228,
+      "learning_rate": 2.5312523467420964e-05,
+      "loss": 11.4693,
+      "step": 31710
+    },
+    {
+      "epoch": 0.15840595270793278,
+      "grad_norm": 0.09434106200933456,
+      "learning_rate": 2.531102155247941e-05,
+      "loss": 20.0531,
+      "step": 31720
+    },
+    {
+      "epoch": 0.15845589153287223,
+      "grad_norm": 0.0996764674782753,
+      "learning_rate": 2.530951963753786e-05,
+      "loss": 9.9376,
+      "step": 31730
+    },
+    {
+      "epoch": 0.15850583035781168,
+      "grad_norm": 0.09661601483821869,
+      "learning_rate": 2.530801772259631e-05,
+      "loss": 13.1668,
+      "step": 31740
+    },
+    {
+      "epoch": 0.15855576918275113,
+      "grad_norm": 0.09104971587657928,
+      "learning_rate": 2.530651580765476e-05,
+      "loss": 10.18,
+      "step": 31750
+    },
+    {
+      "epoch": 0.15860570800769058,
+      "grad_norm": 0.09955008327960968,
+      "learning_rate": 2.5305013892713212e-05,
+      "loss": 9.9397,
+      "step": 31760
+    },
+    {
+      "epoch": 0.15865564683263003,
+      "grad_norm": 0.09806697070598602,
+      "learning_rate": 2.530351197777166e-05,
+      "loss": 9.9451,
+      "step": 31770
+    },
+    {
+      "epoch": 0.15870558565756948,
+      "grad_norm": 0.09614180773496628,
+      "learning_rate": 2.530201006283011e-05,
+      "loss": 9.9402,
+      "step": 31780
+    },
+    {
+      "epoch": 0.15875552448250893,
+      "grad_norm": 0.09188991039991379,
+      "learning_rate": 2.530050814788856e-05,
+      "loss": 9.9389,
+      "step": 31790
+    },
+    {
+      "epoch": 0.15880546330744838,
+      "grad_norm": 0.0986422747373581,
+      "learning_rate": 2.5299006232947006e-05,
+      "loss": 9.9408,
+      "step": 31800
+    },
+    {
+      "epoch": 0.15885540213238783,
+      "grad_norm": 0.09337294101715088,
+      "learning_rate": 2.529750431800546e-05,
+      "loss": 9.9416,
+      "step": 31810
+    },
+    {
+      "epoch": 0.15890534095732728,
+      "grad_norm": 0.09738942235708237,
+      "learning_rate": 2.5296002403063906e-05,
+      "loss": 9.9414,
+      "step": 31820
+    },
+    {
+      "epoch": 0.15895527978226673,
+      "grad_norm": 0.09532233327627182,
+      "learning_rate": 2.5294500488122356e-05,
+      "loss": 9.9396,
+      "step": 31830
+    },
+    {
+      "epoch": 0.15900521860720618,
+      "grad_norm": 0.09321427345275879,
+      "learning_rate": 2.5292998573180807e-05,
+      "loss": 9.9449,
+      "step": 31840
+    },
+    {
+      "epoch": 0.15905515743214563,
+      "grad_norm": 0.09735007584095001,
+      "learning_rate": 2.5291496658239253e-05,
+      "loss": 9.9388,
+      "step": 31850
+    },
+    {
+      "epoch": 0.15910509625708508,
+      "grad_norm": 0.09496832638978958,
+      "learning_rate": 2.5289994743297707e-05,
+      "loss": 9.9346,
+      "step": 31860
+    },
+    {
+      "epoch": 0.15915503508202453,
+      "grad_norm": 0.09069550782442093,
+      "learning_rate": 2.5288492828356154e-05,
+      "loss": 9.9392,
+      "step": 31870
+    },
+    {
+      "epoch": 0.15920497390696398,
+      "grad_norm": 0.09822098165750504,
+      "learning_rate": 2.5286990913414604e-05,
+      "loss": 9.9366,
+      "step": 31880
+    },
+    {
+      "epoch": 0.15925491273190343,
+      "grad_norm": 0.09519645571708679,
+      "learning_rate": 2.5285488998473054e-05,
+      "loss": 9.9355,
+      "step": 31890
+    },
+    {
+      "epoch": 0.15930485155684287,
+      "grad_norm": 0.09621410071849823,
+      "learning_rate": 2.52839870835315e-05,
+      "loss": 9.9392,
+      "step": 31900
+    },
+    {
+      "epoch": 0.15935479038178232,
+      "grad_norm": 0.0956474095582962,
+      "learning_rate": 2.5282485168589954e-05,
+      "loss": 9.9361,
+      "step": 31910
+    },
+    {
+      "epoch": 0.15940472920672177,
+      "grad_norm": 0.09148707240819931,
+      "learning_rate": 2.52809832536484e-05,
+      "loss": 9.9364,
+      "step": 31920
+    },
+    {
+      "epoch": 0.15945466803166122,
+      "grad_norm": 0.09263239800930023,
+      "learning_rate": 2.5279481338706855e-05,
+      "loss": 9.9372,
+      "step": 31930
+    },
+    {
+      "epoch": 0.15950460685660067,
+      "grad_norm": 0.09924571961164474,
+      "learning_rate": 2.52779794237653e-05,
+      "loss": 9.9365,
+      "step": 31940
+    },
+    {
+      "epoch": 0.15955454568154012,
+      "grad_norm": 0.09445744752883911,
+      "learning_rate": 2.527647750882375e-05,
+      "loss": 9.9395,
+      "step": 31950
+    },
+    {
+      "epoch": 0.15960448450647957,
+      "grad_norm": 0.09523255378007889,
+      "learning_rate": 2.5274975593882202e-05,
+      "loss": 9.9367,
+      "step": 31960
+    },
+    {
+      "epoch": 0.15965442333141902,
+      "grad_norm": 0.09308159351348877,
+      "learning_rate": 2.527347367894065e-05,
+      "loss": 9.9337,
+      "step": 31970
+    },
+    {
+      "epoch": 0.15970436215635847,
+      "grad_norm": 0.09193790704011917,
+      "learning_rate": 2.5271971763999102e-05,
+      "loss": 9.9315,
+      "step": 31980
+    },
+    {
+      "epoch": 0.15975430098129792,
+      "grad_norm": 0.09079098701477051,
+      "learning_rate": 2.527046984905755e-05,
+      "loss": 9.942,
+      "step": 31990
+    },
+    {
+      "epoch": 0.15980423980623737,
+      "grad_norm": 0.09207965433597565,
+      "learning_rate": 2.5268967934115996e-05,
+      "loss": 9.9348,
+      "step": 32000
+    },
+    {
+      "epoch": 0.15985417863117682,
+      "grad_norm": 0.09450741112232208,
+      "learning_rate": 2.526746601917445e-05,
+      "loss": 9.938,
+      "step": 32010
+    },
+    {
+      "epoch": 0.15990411745611627,
+      "grad_norm": 0.08837747573852539,
+      "learning_rate": 2.5265964104232896e-05,
+      "loss": 9.9355,
+      "step": 32020
+    },
+    {
+      "epoch": 0.15995405628105572,
+      "grad_norm": 0.09545224159955978,
+      "learning_rate": 2.526446218929135e-05,
+      "loss": 9.9378,
+      "step": 32030
+    },
+    {
+      "epoch": 0.16000399510599517,
+      "grad_norm": 0.09526308625936508,
+      "learning_rate": 2.5262960274349797e-05,
+      "loss": 9.9396,
+      "step": 32040
+    },
+    {
+      "epoch": 0.16005393393093462,
+      "grad_norm": 0.09556493163108826,
+      "learning_rate": 2.5261458359408243e-05,
+      "loss": 9.929,
+      "step": 32050
+    },
+    {
+      "epoch": 0.16010387275587407,
+      "grad_norm": 0.09006509184837341,
+      "learning_rate": 2.5259956444466697e-05,
+      "loss": 9.9293,
+      "step": 32060
+    },
+    {
+      "epoch": 0.16015381158081352,
+      "grad_norm": 0.09654060006141663,
+      "learning_rate": 2.5258454529525144e-05,
+      "loss": 9.9321,
+      "step": 32070
+    },
+    {
+      "epoch": 0.16020375040575296,
+      "grad_norm": 0.09235110133886337,
+      "learning_rate": 2.5256952614583597e-05,
+      "loss": 9.9315,
+      "step": 32080
+    },
+    {
+      "epoch": 0.16025368923069241,
+      "grad_norm": 0.09377003461122513,
+      "learning_rate": 2.5255450699642044e-05,
+      "loss": 9.9327,
+      "step": 32090
+    },
+    {
+      "epoch": 0.16030362805563186,
+      "grad_norm": 0.0920942947268486,
+      "learning_rate": 2.525394878470049e-05,
+      "loss": 9.93,
+      "step": 32100
+    },
+    {
+      "epoch": 0.1603535668805713,
+      "grad_norm": 0.09704404324293137,
+      "learning_rate": 2.5252446869758944e-05,
+      "loss": 9.9293,
+      "step": 32110
+    },
+    {
+      "epoch": 0.16040350570551076,
+      "grad_norm": 0.09537354111671448,
+      "learning_rate": 2.525094495481739e-05,
+      "loss": 9.9361,
+      "step": 32120
+    },
+    {
+      "epoch": 0.1604534445304502,
+      "grad_norm": 0.09306449443101883,
+      "learning_rate": 2.5249443039875845e-05,
+      "loss": 9.9365,
+      "step": 32130
+    },
+    {
+      "epoch": 0.16050338335538963,
+      "grad_norm": 0.09331690520048141,
+      "learning_rate": 2.524794112493429e-05,
+      "loss": 9.9333,
+      "step": 32140
+    },
+    {
+      "epoch": 0.16055332218032908,
+      "grad_norm": 0.09822415560483932,
+      "learning_rate": 2.524643920999274e-05,
+      "loss": 9.9411,
+      "step": 32150
+    },
+    {
+      "epoch": 0.16060326100526853,
+      "grad_norm": 0.0978788435459137,
+      "learning_rate": 2.5244937295051192e-05,
+      "loss": 9.9302,
+      "step": 32160
+    },
+    {
+      "epoch": 0.16065319983020798,
+      "grad_norm": 0.09318576008081436,
+      "learning_rate": 2.524343538010964e-05,
+      "loss": 9.9322,
+      "step": 32170
+    },
+    {
+      "epoch": 0.16070313865514743,
+      "grad_norm": 0.09461916238069534,
+      "learning_rate": 2.5241933465168092e-05,
+      "loss": 9.9289,
+      "step": 32180
+    },
+    {
+      "epoch": 0.16075307748008688,
+      "grad_norm": 0.08854075521230698,
+      "learning_rate": 2.524043155022654e-05,
+      "loss": 9.9322,
+      "step": 32190
+    },
+    {
+      "epoch": 0.16080301630502633,
+      "grad_norm": 0.09415885806083679,
+      "learning_rate": 2.5238929635284986e-05,
+      "loss": 9.9298,
+      "step": 32200
+    },
+    {
+      "epoch": 0.16085295512996578,
+      "grad_norm": 0.09000342339277267,
+      "learning_rate": 2.523742772034344e-05,
+      "loss": 9.9248,
+      "step": 32210
+    },
+    {
+      "epoch": 0.16090289395490523,
+      "grad_norm": 0.09261956810951233,
+      "learning_rate": 2.5235925805401886e-05,
+      "loss": 9.9282,
+      "step": 32220
+    },
+    {
+      "epoch": 0.16095283277984468,
+      "grad_norm": 0.09217025339603424,
+      "learning_rate": 2.523442389046034e-05,
+      "loss": 9.9289,
+      "step": 32230
+    },
+    {
+      "epoch": 0.16100277160478413,
+      "grad_norm": 0.09619983285665512,
+      "learning_rate": 2.5232921975518787e-05,
+      "loss": 9.9366,
+      "step": 32240
+    },
+    {
+      "epoch": 0.16105271042972358,
+      "grad_norm": 0.0968618094921112,
+      "learning_rate": 2.5231420060577237e-05,
+      "loss": 9.9347,
+      "step": 32250
+    },
+    {
+      "epoch": 0.16110264925466303,
+      "grad_norm": 0.09648020565509796,
+      "learning_rate": 2.5229918145635687e-05,
+      "loss": 9.929,
+      "step": 32260
+    },
+    {
+      "epoch": 0.16115258807960248,
+      "grad_norm": 0.09538908302783966,
+      "learning_rate": 2.5228416230694134e-05,
+      "loss": 9.9289,
+      "step": 32270
+    },
+    {
+      "epoch": 0.16120252690454193,
+      "grad_norm": 0.09745009243488312,
+      "learning_rate": 2.5226914315752587e-05,
+      "loss": 9.9325,
+      "step": 32280
+    },
+    {
+      "epoch": 0.16125246572948138,
+      "grad_norm": 0.09166844934225082,
+      "learning_rate": 2.5225412400811034e-05,
+      "loss": 9.9295,
+      "step": 32290
+    },
+    {
+      "epoch": 0.16130240455442083,
+      "grad_norm": 0.09489347785711288,
+      "learning_rate": 2.5223910485869484e-05,
+      "loss": 9.9295,
+      "step": 32300
+    },
+    {
+      "epoch": 0.16135234337936027,
+      "grad_norm": 0.09680917114019394,
+      "learning_rate": 2.5222408570927934e-05,
+      "loss": 9.926,
+      "step": 32310
+    },
+    {
+      "epoch": 0.16140228220429972,
+      "grad_norm": 0.08917848020792007,
+      "learning_rate": 2.522090665598638e-05,
+      "loss": 9.9332,
+      "step": 32320
+    },
+    {
+      "epoch": 0.16145222102923917,
+      "grad_norm": 0.0952332615852356,
+      "learning_rate": 2.5219404741044835e-05,
+      "loss": 9.922,
+      "step": 32330
+    },
+    {
+      "epoch": 0.16150215985417862,
+      "grad_norm": 0.095283642411232,
+      "learning_rate": 2.521790282610328e-05,
+      "loss": 9.9313,
+      "step": 32340
+    },
+    {
+      "epoch": 0.16155209867911807,
+      "grad_norm": 0.09335513412952423,
+      "learning_rate": 2.5216400911161732e-05,
+      "loss": 9.9285,
+      "step": 32350
+    },
+    {
+      "epoch": 0.16160203750405752,
+      "grad_norm": 0.09106207638978958,
+      "learning_rate": 2.5214898996220182e-05,
+      "loss": 9.9263,
+      "step": 32360
+    },
+    {
+      "epoch": 0.16165197632899697,
+      "grad_norm": 0.08792679011821747,
+      "learning_rate": 2.521339708127863e-05,
+      "loss": 9.9225,
+      "step": 32370
+    },
+    {
+      "epoch": 0.16170191515393642,
+      "grad_norm": 0.08863825350999832,
+      "learning_rate": 2.5211895166337082e-05,
+      "loss": 9.9298,
+      "step": 32380
+    },
+    {
+      "epoch": 0.16175185397887587,
+      "grad_norm": 0.08740777522325516,
+      "learning_rate": 2.521039325139553e-05,
+      "loss": 9.9241,
+      "step": 32390
+    },
+    {
+      "epoch": 0.16180179280381532,
+      "grad_norm": 0.10181976854801178,
+      "learning_rate": 2.520889133645398e-05,
+      "loss": 9.9275,
+      "step": 32400
+    },
+    {
+      "epoch": 0.16185173162875477,
+      "grad_norm": 0.09686804562807083,
+      "learning_rate": 2.520738942151243e-05,
+      "loss": 9.9219,
+      "step": 32410
+    },
+    {
+      "epoch": 0.16190167045369422,
+      "grad_norm": 0.09239942580461502,
+      "learning_rate": 2.5205887506570876e-05,
+      "loss": 9.9238,
+      "step": 32420
+    },
+    {
+      "epoch": 0.16195160927863367,
+      "grad_norm": 0.09450032562017441,
+      "learning_rate": 2.520438559162933e-05,
+      "loss": 9.9273,
+      "step": 32430
+    },
+    {
+      "epoch": 0.16200154810357312,
+      "grad_norm": 0.09564870595932007,
+      "learning_rate": 2.5202883676687777e-05,
+      "loss": 9.919,
+      "step": 32440
+    },
+    {
+      "epoch": 0.16205148692851257,
+      "grad_norm": 0.09656788408756256,
+      "learning_rate": 2.5201381761746227e-05,
+      "loss": 9.9207,
+      "step": 32450
+    },
+    {
+      "epoch": 0.16210142575345202,
+      "grad_norm": 0.09183277189731598,
+      "learning_rate": 2.5199879846804677e-05,
+      "loss": 9.928,
+      "step": 32460
+    },
+    {
+      "epoch": 0.16215136457839147,
+      "grad_norm": 0.09268153458833694,
+      "learning_rate": 2.5198377931863124e-05,
+      "loss": 9.9258,
+      "step": 32470
+    },
+    {
+      "epoch": 0.16220130340333092,
+      "grad_norm": 0.09812065958976746,
+      "learning_rate": 2.5196876016921577e-05,
+      "loss": 9.9289,
+      "step": 32480
+    },
+    {
+      "epoch": 0.16225124222827036,
+      "grad_norm": 0.09496765583753586,
+      "learning_rate": 2.5195374101980024e-05,
+      "loss": 9.9278,
+      "step": 32490
+    },
+    {
+      "epoch": 0.16230118105320981,
+      "grad_norm": 0.09568601101636887,
+      "learning_rate": 2.5193872187038474e-05,
+      "loss": 9.9279,
+      "step": 32500
+    },
+    {
+      "epoch": 0.16235111987814926,
+      "grad_norm": 0.0902891531586647,
+      "learning_rate": 2.5192370272096924e-05,
+      "loss": 9.9261,
+      "step": 32510
+    },
+    {
+      "epoch": 0.1624010587030887,
+      "grad_norm": 0.09505823254585266,
+      "learning_rate": 2.519086835715537e-05,
+      "loss": 9.921,
+      "step": 32520
+    },
+    {
+      "epoch": 0.16245099752802816,
+      "grad_norm": 0.09195969253778458,
+      "learning_rate": 2.5189366442213825e-05,
+      "loss": 9.925,
+      "step": 32530
+    },
+    {
+      "epoch": 0.1625009363529676,
+      "grad_norm": 0.09071581065654755,
+      "learning_rate": 2.518786452727227e-05,
+      "loss": 9.9256,
+      "step": 32540
+    },
+    {
+      "epoch": 0.16255087517790706,
+      "grad_norm": 0.0933861956000328,
+      "learning_rate": 2.5186362612330722e-05,
+      "loss": 9.9248,
+      "step": 32550
+    },
+    {
+      "epoch": 0.1626008140028465,
+      "grad_norm": 0.09781572967767715,
+      "learning_rate": 2.5184860697389172e-05,
+      "loss": 9.9215,
+      "step": 32560
+    },
+    {
+      "epoch": 0.16265075282778596,
+      "grad_norm": 0.09224456548690796,
+      "learning_rate": 2.5183358782447622e-05,
+      "loss": 9.925,
+      "step": 32570
+    },
+    {
+      "epoch": 0.1627006916527254,
+      "grad_norm": 0.09544172137975693,
+      "learning_rate": 2.5181856867506072e-05,
+      "loss": 9.9167,
+      "step": 32580
+    },
+    {
+      "epoch": 0.16275063047766486,
+      "grad_norm": 0.09155009686946869,
+      "learning_rate": 2.518035495256452e-05,
+      "loss": 9.9184,
+      "step": 32590
+    },
+    {
+      "epoch": 0.1628005693026043,
+      "grad_norm": 0.10186931490898132,
+      "learning_rate": 2.517885303762297e-05,
+      "loss": 9.9262,
+      "step": 32600
+    },
+    {
+      "epoch": 0.16285050812754376,
+      "grad_norm": 0.09243344515562057,
+      "learning_rate": 2.517735112268142e-05,
+      "loss": 9.9216,
+      "step": 32610
+    },
+    {
+      "epoch": 0.1629004469524832,
+      "grad_norm": 0.08900917321443558,
+      "learning_rate": 2.517584920773987e-05,
+      "loss": 9.9274,
+      "step": 32620
+    },
+    {
+      "epoch": 0.16295038577742266,
+      "grad_norm": 0.09651615470647812,
+      "learning_rate": 2.517434729279832e-05,
+      "loss": 9.9232,
+      "step": 32630
+    },
+    {
+      "epoch": 0.1630003246023621,
+      "grad_norm": 0.09579268842935562,
+      "learning_rate": 2.5172845377856767e-05,
+      "loss": 9.9149,
+      "step": 32640
+    },
+    {
+      "epoch": 0.16305026342730156,
+      "grad_norm": 0.09731176495552063,
+      "learning_rate": 2.5171343462915217e-05,
+      "loss": 9.9206,
+      "step": 32650
+    },
+    {
+      "epoch": 0.163100202252241,
+      "grad_norm": 0.10048079490661621,
+      "learning_rate": 2.5169841547973667e-05,
+      "loss": 9.9182,
+      "step": 32660
+    },
+    {
+      "epoch": 0.16315014107718045,
+      "grad_norm": 0.0903536006808281,
+      "learning_rate": 2.5168339633032117e-05,
+      "loss": 9.9161,
+      "step": 32670
+    },
+    {
+      "epoch": 0.1632000799021199,
+      "grad_norm": 0.096148781478405,
+      "learning_rate": 2.5166837718090567e-05,
+      "loss": 9.9185,
+      "step": 32680
+    },
+    {
+      "epoch": 0.16325001872705935,
+      "grad_norm": 0.09729062020778656,
+      "learning_rate": 2.5165335803149014e-05,
+      "loss": 9.9234,
+      "step": 32690
+    },
+    {
+      "epoch": 0.1632999575519988,
+      "grad_norm": 0.09655334800481796,
+      "learning_rate": 2.5163833888207464e-05,
+      "loss": 9.9228,
+      "step": 32700
+    },
+    {
+      "epoch": 0.16334989637693825,
+      "grad_norm": 0.09325189888477325,
+      "learning_rate": 2.5162331973265914e-05,
+      "loss": 9.9221,
+      "step": 32710
+    },
+    {
+      "epoch": 0.1633998352018777,
+      "grad_norm": 0.09188661724328995,
+      "learning_rate": 2.5160830058324365e-05,
+      "loss": 9.9184,
+      "step": 32720
+    },
+    {
+      "epoch": 0.16344977402681715,
+      "grad_norm": 0.08944937586784363,
+      "learning_rate": 2.5159328143382815e-05,
+      "loss": 9.9146,
+      "step": 32730
+    },
+    {
+      "epoch": 0.1634997128517566,
+      "grad_norm": 0.09515637904405594,
+      "learning_rate": 2.515782622844126e-05,
+      "loss": 9.9245,
+      "step": 32740
+    },
+    {
+      "epoch": 0.16354965167669605,
+      "grad_norm": 0.09636104851961136,
+      "learning_rate": 2.5156324313499712e-05,
+      "loss": 9.9186,
+      "step": 32750
+    },
+    {
+      "epoch": 0.1635995905016355,
+      "grad_norm": 0.09423839300870895,
+      "learning_rate": 2.5154822398558162e-05,
+      "loss": 9.9193,
+      "step": 32760
+    },
+    {
+      "epoch": 0.16364952932657495,
+      "grad_norm": 0.09561903774738312,
+      "learning_rate": 2.5153320483616612e-05,
+      "loss": 9.9095,
+      "step": 32770
+    },
+    {
+      "epoch": 0.1636994681515144,
+      "grad_norm": 0.0979316234588623,
+      "learning_rate": 2.5151818568675062e-05,
+      "loss": 9.9165,
+      "step": 32780
+    },
+    {
+      "epoch": 0.16374940697645385,
+      "grad_norm": 0.09158878028392792,
+      "learning_rate": 2.515031665373351e-05,
+      "loss": 9.9188,
+      "step": 32790
+    },
+    {
+      "epoch": 0.1637993458013933,
+      "grad_norm": 0.09471914172172546,
+      "learning_rate": 2.514881473879196e-05,
+      "loss": 9.9175,
+      "step": 32800
+    },
+    {
+      "epoch": 0.16384928462633275,
+      "grad_norm": 0.08805624395608902,
+      "learning_rate": 2.514731282385041e-05,
+      "loss": 9.9205,
+      "step": 32810
+    },
+    {
+      "epoch": 0.1638992234512722,
+      "grad_norm": 0.09481382369995117,
+      "learning_rate": 2.514581090890886e-05,
+      "loss": 9.9162,
+      "step": 32820
+    },
+    {
+      "epoch": 0.16394916227621165,
+      "grad_norm": 0.09198730438947678,
+      "learning_rate": 2.514430899396731e-05,
+      "loss": 9.9177,
+      "step": 32830
+    },
+    {
+      "epoch": 0.1639991011011511,
+      "grad_norm": 0.09711703658103943,
+      "learning_rate": 2.5142807079025757e-05,
+      "loss": 9.9141,
+      "step": 32840
+    },
+    {
+      "epoch": 0.16404903992609055,
+      "grad_norm": 0.09685643017292023,
+      "learning_rate": 2.5141305164084207e-05,
+      "loss": 9.9108,
+      "step": 32850
+    },
+    {
+      "epoch": 0.16409897875103,
+      "grad_norm": 0.09421391785144806,
+      "learning_rate": 2.5139803249142657e-05,
+      "loss": 9.9192,
+      "step": 32860
+    },
+    {
+      "epoch": 0.16414891757596944,
+      "grad_norm": 0.09753596037626266,
+      "learning_rate": 2.5138301334201107e-05,
+      "loss": 9.9174,
+      "step": 32870
+    },
+    {
+      "epoch": 0.1641988564009089,
+      "grad_norm": 0.09626904875040054,
+      "learning_rate": 2.5136799419259557e-05,
+      "loss": 9.9172,
+      "step": 32880
+    },
+    {
+      "epoch": 0.16424879522584834,
+      "grad_norm": 0.0984250158071518,
+      "learning_rate": 2.5135297504318004e-05,
+      "loss": 9.9117,
+      "step": 32890
+    },
+    {
+      "epoch": 0.1642987340507878,
+      "grad_norm": 0.09708181023597717,
+      "learning_rate": 2.5133795589376454e-05,
+      "loss": 9.9127,
+      "step": 32900
+    },
+    {
+      "epoch": 0.16434867287572724,
+      "grad_norm": 0.09856320172548294,
+      "learning_rate": 2.5132293674434904e-05,
+      "loss": 9.9167,
+      "step": 32910
+    },
+    {
+      "epoch": 0.1643986117006667,
+      "grad_norm": 0.09314543008804321,
+      "learning_rate": 2.5130791759493355e-05,
+      "loss": 9.9153,
+      "step": 32920
+    },
+    {
+      "epoch": 0.16444855052560614,
+      "grad_norm": 0.09600864350795746,
+      "learning_rate": 2.5129289844551805e-05,
+      "loss": 9.9172,
+      "step": 32930
+    },
+    {
+      "epoch": 0.1644984893505456,
+      "grad_norm": 0.09382302314043045,
+      "learning_rate": 2.5127787929610255e-05,
+      "loss": 9.9149,
+      "step": 32940
+    },
+    {
+      "epoch": 0.16454842817548504,
+      "grad_norm": 0.09613067656755447,
+      "learning_rate": 2.5126286014668702e-05,
+      "loss": 9.9112,
+      "step": 32950
+    },
+    {
+      "epoch": 0.1645983670004245,
+      "grad_norm": 0.09294107556343079,
+      "learning_rate": 2.5124784099727152e-05,
+      "loss": 9.9157,
+      "step": 32960
+    },
+    {
+      "epoch": 0.16464830582536394,
+      "grad_norm": 0.09056387096643448,
+      "learning_rate": 2.5123282184785602e-05,
+      "loss": 9.9145,
+      "step": 32970
+    },
+    {
+      "epoch": 0.1646982446503034,
+      "grad_norm": 0.09541086107492447,
+      "learning_rate": 2.5121780269844052e-05,
+      "loss": 9.9074,
+      "step": 32980
+    },
+    {
+      "epoch": 0.16474818347524284,
+      "grad_norm": 0.09487113356590271,
+      "learning_rate": 2.5120278354902502e-05,
+      "loss": 9.9122,
+      "step": 32990
+    },
+    {
+      "epoch": 0.1647981223001823,
+      "grad_norm": 0.0954999253153801,
+      "learning_rate": 2.511877643996095e-05,
+      "loss": 9.9143,
+      "step": 33000
+    },
+    {
+      "epoch": 0.16484806112512174,
+      "grad_norm": 0.09315845370292664,
+      "learning_rate": 2.51172745250194e-05,
+      "loss": 9.9189,
+      "step": 33010
+    },
+    {
+      "epoch": 0.16489799995006119,
+      "grad_norm": 0.09187664091587067,
+      "learning_rate": 2.511577261007785e-05,
+      "loss": 9.9106,
+      "step": 33020
+    },
+    {
+      "epoch": 0.16494793877500064,
+      "grad_norm": 0.09694460779428482,
+      "learning_rate": 2.51142706951363e-05,
+      "loss": 9.9149,
+      "step": 33030
+    },
+    {
+      "epoch": 0.16499787759994008,
+      "grad_norm": 0.08832021057605743,
+      "learning_rate": 2.511276878019475e-05,
+      "loss": 9.911,
+      "step": 33040
+    },
+    {
+      "epoch": 0.16504781642487953,
+      "grad_norm": 0.09793966263532639,
+      "learning_rate": 2.5111266865253197e-05,
+      "loss": 9.9152,
+      "step": 33050
+    },
+    {
+      "epoch": 0.16509775524981898,
+      "grad_norm": 0.09660662710666656,
+      "learning_rate": 2.5109764950311647e-05,
+      "loss": 9.9142,
+      "step": 33060
+    },
+    {
+      "epoch": 0.16514769407475843,
+      "grad_norm": 0.09192895144224167,
+      "learning_rate": 2.5108263035370097e-05,
+      "loss": 9.9064,
+      "step": 33070
+    },
+    {
+      "epoch": 0.16519763289969788,
+      "grad_norm": 0.09510926902294159,
+      "learning_rate": 2.5106761120428547e-05,
+      "loss": 9.9095,
+      "step": 33080
+    },
+    {
+      "epoch": 0.16524757172463733,
+      "grad_norm": 0.09554640203714371,
+      "learning_rate": 2.5105259205486998e-05,
+      "loss": 9.9071,
+      "step": 33090
+    },
+    {
+      "epoch": 0.16529751054957678,
+      "grad_norm": 0.0876171737909317,
+      "learning_rate": 2.5103757290545448e-05,
+      "loss": 9.9134,
+      "step": 33100
+    },
+    {
+      "epoch": 0.16534744937451623,
+      "grad_norm": 0.09115521609783173,
+      "learning_rate": 2.5102255375603894e-05,
+      "loss": 9.9153,
+      "step": 33110
+    },
+    {
+      "epoch": 0.16539738819945568,
+      "grad_norm": 0.09953109920024872,
+      "learning_rate": 2.5100753460662345e-05,
+      "loss": 9.9073,
+      "step": 33120
+    },
+    {
+      "epoch": 0.1654473270243951,
+      "grad_norm": 0.09405183792114258,
+      "learning_rate": 2.5099251545720795e-05,
+      "loss": 9.906,
+      "step": 33130
+    },
+    {
+      "epoch": 0.16549726584933455,
+      "grad_norm": 0.09410317987203598,
+      "learning_rate": 2.5097749630779245e-05,
+      "loss": 9.9176,
+      "step": 33140
+    },
+    {
+      "epoch": 0.165547204674274,
+      "grad_norm": 0.09550243616104126,
+      "learning_rate": 2.5096247715837695e-05,
+      "loss": 9.9054,
+      "step": 33150
+    },
+    {
+      "epoch": 0.16559714349921345,
+      "grad_norm": 0.09511136263608932,
+      "learning_rate": 2.5094745800896142e-05,
+      "loss": 9.917,
+      "step": 33160
+    },
+    {
+      "epoch": 0.1656470823241529,
+      "grad_norm": 0.09608213603496552,
+      "learning_rate": 2.5093243885954592e-05,
+      "loss": 9.9164,
+      "step": 33170
+    },
+    {
+      "epoch": 0.16569702114909235,
+      "grad_norm": 0.0949927344918251,
+      "learning_rate": 2.5091741971013042e-05,
+      "loss": 9.9107,
+      "step": 33180
+    },
+    {
+      "epoch": 0.1657469599740318,
+      "grad_norm": 0.09686841070652008,
+      "learning_rate": 2.5090240056071493e-05,
+      "loss": 9.9089,
+      "step": 33190
+    },
+    {
+      "epoch": 0.16579689879897125,
+      "grad_norm": 0.09343890100717545,
+      "learning_rate": 2.5088738141129943e-05,
+      "loss": 9.9128,
+      "step": 33200
+    },
+    {
+      "epoch": 0.1658468376239107,
+      "grad_norm": 0.09721093624830246,
+      "learning_rate": 2.508723622618839e-05,
+      "loss": 9.9105,
+      "step": 33210
+    },
+    {
+      "epoch": 0.16589677644885015,
+      "grad_norm": 0.0927630066871643,
+      "learning_rate": 2.508573431124684e-05,
+      "loss": 9.9052,
+      "step": 33220
+    },
+    {
+      "epoch": 0.1659467152737896,
+      "grad_norm": 0.09124623984098434,
+      "learning_rate": 2.508423239630529e-05,
+      "loss": 9.9141,
+      "step": 33230
+    },
+    {
+      "epoch": 0.16599665409872905,
+      "grad_norm": 0.10382476449012756,
+      "learning_rate": 2.508273048136374e-05,
+      "loss": 9.9072,
+      "step": 33240
+    },
+    {
+      "epoch": 0.1660465929236685,
+      "grad_norm": 0.09035416692495346,
+      "learning_rate": 2.508122856642219e-05,
+      "loss": 9.9108,
+      "step": 33250
+    },
+    {
+      "epoch": 0.16609653174860795,
+      "grad_norm": 0.0920075997710228,
+      "learning_rate": 2.507972665148064e-05,
+      "loss": 9.9083,
+      "step": 33260
+    },
+    {
+      "epoch": 0.1661464705735474,
+      "grad_norm": 0.09175487607717514,
+      "learning_rate": 2.5078224736539087e-05,
+      "loss": 9.9091,
+      "step": 33270
+    },
+    {
+      "epoch": 0.16619640939848684,
+      "grad_norm": 0.09441912174224854,
+      "learning_rate": 2.5076722821597537e-05,
+      "loss": 9.9066,
+      "step": 33280
+    },
+    {
+      "epoch": 0.1662463482234263,
+      "grad_norm": 0.09477470815181732,
+      "learning_rate": 2.5075220906655988e-05,
+      "loss": 9.9087,
+      "step": 33290
+    },
+    {
+      "epoch": 0.16629628704836574,
+      "grad_norm": 0.08985955268144608,
+      "learning_rate": 2.5073718991714438e-05,
+      "loss": 9.9046,
+      "step": 33300
+    },
+    {
+      "epoch": 0.1663462258733052,
+      "grad_norm": 0.0965331420302391,
+      "learning_rate": 2.5072217076772888e-05,
+      "loss": 9.9071,
+      "step": 33310
+    },
+    {
+      "epoch": 0.16639616469824464,
+      "grad_norm": 0.0963759496808052,
+      "learning_rate": 2.5070715161831335e-05,
+      "loss": 9.9057,
+      "step": 33320
+    },
+    {
+      "epoch": 0.1664461035231841,
+      "grad_norm": 0.097702257335186,
+      "learning_rate": 2.5069213246889785e-05,
+      "loss": 9.9115,
+      "step": 33330
+    },
+    {
+      "epoch": 0.16649604234812354,
+      "grad_norm": 0.09388408064842224,
+      "learning_rate": 2.5067711331948235e-05,
+      "loss": 9.9045,
+      "step": 33340
+    },
+    {
+      "epoch": 0.166545981173063,
+      "grad_norm": 0.09655765444040298,
+      "learning_rate": 2.5066209417006685e-05,
+      "loss": 9.9053,
+      "step": 33350
+    },
+    {
+      "epoch": 0.16659591999800244,
+      "grad_norm": 0.09741511940956116,
+      "learning_rate": 2.5064707502065135e-05,
+      "loss": 9.9012,
+      "step": 33360
+    },
+    {
+      "epoch": 0.1666458588229419,
+      "grad_norm": 0.09620445966720581,
+      "learning_rate": 2.5063205587123582e-05,
+      "loss": 9.8977,
+      "step": 33370
+    },
+    {
+      "epoch": 0.16669579764788134,
+      "grad_norm": 0.09696647524833679,
+      "learning_rate": 2.5061703672182032e-05,
+      "loss": 9.9067,
+      "step": 33380
+    },
+    {
+      "epoch": 0.1667457364728208,
+      "grad_norm": 0.09665276110172272,
+      "learning_rate": 2.5060201757240483e-05,
+      "loss": 9.8977,
+      "step": 33390
+    },
+    {
+      "epoch": 0.16679567529776024,
+      "grad_norm": 0.09218195825815201,
+      "learning_rate": 2.5058699842298933e-05,
+      "loss": 9.9092,
+      "step": 33400
+    },
+    {
+      "epoch": 0.1668456141226997,
+      "grad_norm": 0.09188750386238098,
+      "learning_rate": 2.5057197927357383e-05,
+      "loss": 9.9054,
+      "step": 33410
+    },
+    {
+      "epoch": 0.16689555294763914,
+      "grad_norm": 0.09387121349573135,
+      "learning_rate": 2.505569601241583e-05,
+      "loss": 9.9054,
+      "step": 33420
+    },
+    {
+      "epoch": 0.16694549177257859,
+      "grad_norm": 0.09394029527902603,
+      "learning_rate": 2.505419409747428e-05,
+      "loss": 9.8994,
+      "step": 33430
+    },
+    {
+      "epoch": 0.16699543059751804,
+      "grad_norm": 0.09561529755592346,
+      "learning_rate": 2.505269218253273e-05,
+      "loss": 9.9005,
+      "step": 33440
+    },
+    {
+      "epoch": 0.16704536942245748,
+      "grad_norm": 0.09472378343343735,
+      "learning_rate": 2.505119026759118e-05,
+      "loss": 9.8989,
+      "step": 33450
+    },
+    {
+      "epoch": 0.16709530824739693,
+      "grad_norm": 0.09782572835683823,
+      "learning_rate": 2.504968835264963e-05,
+      "loss": 9.8989,
+      "step": 33460
+    },
+    {
+      "epoch": 0.16714524707233638,
+      "grad_norm": 0.0978260487318039,
+      "learning_rate": 2.5048186437708077e-05,
+      "loss": 9.9044,
+      "step": 33470
+    },
+    {
+      "epoch": 0.16719518589727583,
+      "grad_norm": 0.09694740176200867,
+      "learning_rate": 2.5046684522766527e-05,
+      "loss": 9.9096,
+      "step": 33480
+    },
+    {
+      "epoch": 0.16724512472221528,
+      "grad_norm": 0.0943945124745369,
+      "learning_rate": 2.5045182607824978e-05,
+      "loss": 9.9032,
+      "step": 33490
+    },
+    {
+      "epoch": 0.16729506354715473,
+      "grad_norm": 0.09394306689500809,
+      "learning_rate": 2.5043680692883428e-05,
+      "loss": 9.8996,
+      "step": 33500
+    },
+    {
+      "epoch": 0.16734500237209418,
+      "grad_norm": 0.09680566936731339,
+      "learning_rate": 2.5042178777941878e-05,
+      "loss": 9.8961,
+      "step": 33510
+    },
+    {
+      "epoch": 0.16739494119703363,
+      "grad_norm": 0.09609083831310272,
+      "learning_rate": 2.5040676863000325e-05,
+      "loss": 9.9013,
+      "step": 33520
+    },
+    {
+      "epoch": 0.16744488002197308,
+      "grad_norm": 0.09596184641122818,
+      "learning_rate": 2.5039174948058775e-05,
+      "loss": 9.9,
+      "step": 33530
+    },
+    {
+      "epoch": 0.16749481884691253,
+      "grad_norm": 0.09482643753290176,
+      "learning_rate": 2.5037673033117225e-05,
+      "loss": 9.8908,
+      "step": 33540
+    },
+    {
+      "epoch": 0.16754475767185198,
+      "grad_norm": 0.09858851134777069,
+      "learning_rate": 2.5036171118175675e-05,
+      "loss": 9.8957,
+      "step": 33550
+    },
+    {
+      "epoch": 0.16759469649679143,
+      "grad_norm": 0.09529074281454086,
+      "learning_rate": 2.5034669203234125e-05,
+      "loss": 9.8975,
+      "step": 33560
+    },
+    {
+      "epoch": 0.16764463532173088,
+      "grad_norm": 0.09430502355098724,
+      "learning_rate": 2.5033167288292572e-05,
+      "loss": 9.9093,
+      "step": 33570
+    },
+    {
+      "epoch": 0.16769457414667033,
+      "grad_norm": 0.09423274546861649,
+      "learning_rate": 2.5031665373351026e-05,
+      "loss": 9.9008,
+      "step": 33580
+    },
+    {
+      "epoch": 0.16774451297160978,
+      "grad_norm": 0.09213889390230179,
+      "learning_rate": 2.5030163458409473e-05,
+      "loss": 9.8997,
+      "step": 33590
+    },
+    {
+      "epoch": 0.16779445179654923,
+      "grad_norm": 0.10176771134138107,
+      "learning_rate": 2.5028661543467923e-05,
+      "loss": 9.8974,
+      "step": 33600
+    },
+    {
+      "epoch": 0.16784439062148868,
+      "grad_norm": 0.0979059562087059,
+      "learning_rate": 2.5027159628526373e-05,
+      "loss": 9.8964,
+      "step": 33610
+    },
+    {
+      "epoch": 0.16789432944642813,
+      "grad_norm": 0.09319134056568146,
+      "learning_rate": 2.502565771358482e-05,
+      "loss": 9.8955,
+      "step": 33620
+    },
+    {
+      "epoch": 0.16794426827136757,
+      "grad_norm": 0.09434917569160461,
+      "learning_rate": 2.5024155798643273e-05,
+      "loss": 9.9026,
+      "step": 33630
+    },
+    {
+      "epoch": 0.16799420709630702,
+      "grad_norm": 0.09023196995258331,
+      "learning_rate": 2.502265388370172e-05,
+      "loss": 9.8935,
+      "step": 33640
+    },
+    {
+      "epoch": 0.16804414592124647,
+      "grad_norm": 0.08715837448835373,
+      "learning_rate": 2.502115196876017e-05,
+      "loss": 9.8996,
+      "step": 33650
+    },
+    {
+      "epoch": 0.16809408474618592,
+      "grad_norm": 0.09682469815015793,
+      "learning_rate": 2.501965005381862e-05,
+      "loss": 9.9018,
+      "step": 33660
+    },
+    {
+      "epoch": 0.16814402357112537,
+      "grad_norm": 0.0928163006901741,
+      "learning_rate": 2.5018148138877067e-05,
+      "loss": 9.8917,
+      "step": 33670
+    },
+    {
+      "epoch": 0.16819396239606482,
+      "grad_norm": 0.10032060742378235,
+      "learning_rate": 2.501664622393552e-05,
+      "loss": 9.8959,
+      "step": 33680
+    },
+    {
+      "epoch": 0.16824390122100427,
+      "grad_norm": 0.0933447927236557,
+      "learning_rate": 2.5015144308993968e-05,
+      "loss": 9.9003,
+      "step": 33690
+    },
+    {
+      "epoch": 0.16829384004594372,
+      "grad_norm": 0.09343995898962021,
+      "learning_rate": 2.5013642394052418e-05,
+      "loss": 9.897,
+      "step": 33700
+    },
+    {
+      "epoch": 0.16834377887088317,
+      "grad_norm": 0.09390633553266525,
+      "learning_rate": 2.5012140479110868e-05,
+      "loss": 9.8975,
+      "step": 33710
+    },
+    {
+      "epoch": 0.16839371769582262,
+      "grad_norm": 0.10121657699346542,
+      "learning_rate": 2.5010638564169315e-05,
+      "loss": 9.9001,
+      "step": 33720
+    },
+    {
+      "epoch": 0.16844365652076207,
+      "grad_norm": 0.09688078612089157,
+      "learning_rate": 2.5009136649227768e-05,
+      "loss": 9.8943,
+      "step": 33730
+    },
+    {
+      "epoch": 0.16849359534570152,
+      "grad_norm": 0.09601810574531555,
+      "learning_rate": 2.5007634734286215e-05,
+      "loss": 9.8948,
+      "step": 33740
+    },
+    {
+      "epoch": 0.16854353417064097,
+      "grad_norm": 0.09208814799785614,
+      "learning_rate": 2.5006132819344665e-05,
+      "loss": 9.8938,
+      "step": 33750
+    },
+    {
+      "epoch": 0.16859347299558042,
+      "grad_norm": 0.10235219448804855,
+      "learning_rate": 2.5004630904403115e-05,
+      "loss": 9.8951,
+      "step": 33760
+    },
+    {
+      "epoch": 0.16864341182051987,
+      "grad_norm": 0.09691403061151505,
+      "learning_rate": 2.5003128989461562e-05,
+      "loss": 9.8948,
+      "step": 33770
+    },
+    {
+      "epoch": 0.16869335064545932,
+      "grad_norm": 0.0951026976108551,
+      "learning_rate": 2.5001627074520016e-05,
+      "loss": 9.9,
+      "step": 33780
+    },
+    {
+      "epoch": 0.16874328947039877,
+      "grad_norm": 0.08896276354789734,
+      "learning_rate": 2.5000125159578463e-05,
+      "loss": 9.8944,
+      "step": 33790
+    },
+    {
+      "epoch": 0.16879322829533822,
+      "grad_norm": 0.09217982739210129,
+      "learning_rate": 2.4998623244636913e-05,
+      "loss": 9.8922,
+      "step": 33800
+    },
+    {
+      "epoch": 0.16884316712027767,
+      "grad_norm": 0.09872495383024216,
+      "learning_rate": 2.4997121329695363e-05,
+      "loss": 9.8956,
+      "step": 33810
+    },
+    {
+      "epoch": 0.16889310594521711,
+      "grad_norm": 0.09138542413711548,
+      "learning_rate": 2.499561941475381e-05,
+      "loss": 9.8934,
+      "step": 33820
+    },
+    {
+      "epoch": 0.16894304477015656,
+      "grad_norm": 0.09778387099504471,
+      "learning_rate": 2.4994117499812263e-05,
+      "loss": 9.8991,
+      "step": 33830
+    },
+    {
+      "epoch": 0.168992983595096,
+      "grad_norm": 0.0923265740275383,
+      "learning_rate": 2.499261558487071e-05,
+      "loss": 9.8911,
+      "step": 33840
+    },
+    {
+      "epoch": 0.16904292242003546,
+      "grad_norm": 0.09650073945522308,
+      "learning_rate": 2.499111366992916e-05,
+      "loss": 9.8927,
+      "step": 33850
+    },
+    {
+      "epoch": 0.1690928612449749,
+      "grad_norm": 0.09646689891815186,
+      "learning_rate": 2.498961175498761e-05,
+      "loss": 9.8933,
+      "step": 33860
+    },
+    {
+      "epoch": 0.16914280006991436,
+      "grad_norm": 0.09530924260616302,
+      "learning_rate": 2.4988109840046057e-05,
+      "loss": 9.893,
+      "step": 33870
+    },
+    {
+      "epoch": 0.1691927388948538,
+      "grad_norm": 0.09329960495233536,
+      "learning_rate": 2.498660792510451e-05,
+      "loss": 9.8912,
+      "step": 33880
+    },
+    {
+      "epoch": 0.16924267771979326,
+      "grad_norm": 0.09914696216583252,
+      "learning_rate": 2.4985106010162958e-05,
+      "loss": 9.8876,
+      "step": 33890
+    },
+    {
+      "epoch": 0.1692926165447327,
+      "grad_norm": 0.09866105765104294,
+      "learning_rate": 2.498360409522141e-05,
+      "loss": 9.9011,
+      "step": 33900
+    },
+    {
+      "epoch": 0.16934255536967216,
+      "grad_norm": 0.09826882928609848,
+      "learning_rate": 2.4982102180279858e-05,
+      "loss": 9.8992,
+      "step": 33910
+    },
+    {
+      "epoch": 0.1693924941946116,
+      "grad_norm": 0.09396472573280334,
+      "learning_rate": 2.4980600265338305e-05,
+      "loss": 9.8908,
+      "step": 33920
+    },
+    {
+      "epoch": 0.16944243301955106,
+      "grad_norm": 0.08894440531730652,
+      "learning_rate": 2.4979098350396758e-05,
+      "loss": 9.8922,
+      "step": 33930
+    },
+    {
+      "epoch": 0.1694923718444905,
+      "grad_norm": 0.09588942676782608,
+      "learning_rate": 2.4977596435455205e-05,
+      "loss": 9.8917,
+      "step": 33940
+    },
+    {
+      "epoch": 0.16954231066942996,
+      "grad_norm": 0.09317868947982788,
+      "learning_rate": 2.497609452051366e-05,
+      "loss": 9.8941,
+      "step": 33950
+    },
+    {
+      "epoch": 0.1695922494943694,
+      "grad_norm": 0.0976976752281189,
+      "learning_rate": 2.4974592605572105e-05,
+      "loss": 9.8955,
+      "step": 33960
+    },
+    {
+      "epoch": 0.16964218831930886,
+      "grad_norm": 0.10638713836669922,
+      "learning_rate": 2.4973090690630552e-05,
+      "loss": 9.8906,
+      "step": 33970
+    },
+    {
+      "epoch": 0.1696921271442483,
+      "grad_norm": 0.0989893451333046,
+      "learning_rate": 2.4971588775689006e-05,
+      "loss": 9.8929,
+      "step": 33980
+    },
+    {
+      "epoch": 0.16974206596918776,
+      "grad_norm": 0.09523206949234009,
+      "learning_rate": 2.4970086860747453e-05,
+      "loss": 9.8959,
+      "step": 33990
+    },
+    {
+      "epoch": 0.1697920047941272,
+      "grad_norm": 0.08967464417219162,
+      "learning_rate": 2.4968584945805906e-05,
+      "loss": 9.8947,
+      "step": 34000
+    },
+    {
+      "epoch": 0.16984194361906665,
+      "grad_norm": 0.09731745719909668,
+      "learning_rate": 2.4967083030864353e-05,
+      "loss": 9.8914,
+      "step": 34010
+    },
+    {
+      "epoch": 0.1698918824440061,
+      "grad_norm": 0.09333444386720657,
+      "learning_rate": 2.49655811159228e-05,
+      "loss": 9.897,
+      "step": 34020
+    },
+    {
+      "epoch": 0.16994182126894555,
+      "grad_norm": 0.09605353325605392,
+      "learning_rate": 2.4964079200981253e-05,
+      "loss": 9.8829,
+      "step": 34030
+    },
+    {
+      "epoch": 0.169991760093885,
+      "grad_norm": 0.08907973766326904,
+      "learning_rate": 2.49625772860397e-05,
+      "loss": 9.8886,
+      "step": 34040
+    },
+    {
+      "epoch": 0.17004169891882445,
+      "grad_norm": 0.09453407675027847,
+      "learning_rate": 2.4961075371098154e-05,
+      "loss": 9.8884,
+      "step": 34050
+    },
+    {
+      "epoch": 0.1700916377437639,
+      "grad_norm": 0.09984616935253143,
+      "learning_rate": 2.49595734561566e-05,
+      "loss": 9.899,
+      "step": 34060
+    },
+    {
+      "epoch": 0.17014157656870335,
+      "grad_norm": 0.0900527685880661,
+      "learning_rate": 2.4958071541215047e-05,
+      "loss": 9.8932,
+      "step": 34070
+    },
+    {
+      "epoch": 0.1701915153936428,
+      "grad_norm": 0.09352007508277893,
+      "learning_rate": 2.49565696262735e-05,
+      "loss": 9.8887,
+      "step": 34080
+    },
+    {
+      "epoch": 0.17024145421858225,
+      "grad_norm": 0.0973382219672203,
+      "learning_rate": 2.4955067711331948e-05,
+      "loss": 9.8908,
+      "step": 34090
+    },
+    {
+      "epoch": 0.1702913930435217,
+      "grad_norm": 0.09201547503471375,
+      "learning_rate": 2.49535657963904e-05,
+      "loss": 9.8875,
+      "step": 34100
+    },
+    {
+      "epoch": 0.17034133186846115,
+      "grad_norm": 0.09478182345628738,
+      "learning_rate": 2.4952063881448848e-05,
+      "loss": 9.8797,
+      "step": 34110
+    },
+    {
+      "epoch": 0.17039127069340057,
+      "grad_norm": 0.09311734884977341,
+      "learning_rate": 2.4950561966507295e-05,
+      "loss": 9.8855,
+      "step": 34120
+    },
+    {
+      "epoch": 0.17044120951834002,
+      "grad_norm": 0.09643889218568802,
+      "learning_rate": 2.4949060051565748e-05,
+      "loss": 9.8851,
+      "step": 34130
+    },
+    {
+      "epoch": 0.17049114834327947,
+      "grad_norm": 0.09145352989435196,
+      "learning_rate": 2.4947558136624195e-05,
+      "loss": 9.8896,
+      "step": 34140
+    },
+    {
+      "epoch": 0.17054108716821892,
+      "grad_norm": 0.0948040559887886,
+      "learning_rate": 2.494605622168265e-05,
+      "loss": 9.8845,
+      "step": 34150
+    },
+    {
+      "epoch": 0.17059102599315837,
+      "grad_norm": 0.09217320382595062,
+      "learning_rate": 2.4944554306741095e-05,
+      "loss": 9.8888,
+      "step": 34160
+    },
+    {
+      "epoch": 0.17064096481809782,
+      "grad_norm": 0.09132317453622818,
+      "learning_rate": 2.4943052391799542e-05,
+      "loss": 9.8871,
+      "step": 34170
+    },
+    {
+      "epoch": 0.17069090364303727,
+      "grad_norm": 0.0995708778500557,
+      "learning_rate": 2.4941550476857996e-05,
+      "loss": 9.8957,
+      "step": 34180
+    },
+    {
+      "epoch": 0.17074084246797672,
+      "grad_norm": 0.09466378390789032,
+      "learning_rate": 2.4940048561916443e-05,
+      "loss": 9.8955,
+      "step": 34190
+    },
+    {
+      "epoch": 0.17079078129291617,
+      "grad_norm": 0.09082814306020737,
+      "learning_rate": 2.4938546646974896e-05,
+      "loss": 9.8832,
+      "step": 34200
+    },
+    {
+      "epoch": 0.17084072011785562,
+      "grad_norm": 0.09607654809951782,
+      "learning_rate": 2.4937044732033343e-05,
+      "loss": 9.887,
+      "step": 34210
+    },
+    {
+      "epoch": 0.17089065894279507,
+      "grad_norm": 0.09264031052589417,
+      "learning_rate": 2.4935542817091793e-05,
+      "loss": 9.8864,
+      "step": 34220
+    },
+    {
+      "epoch": 0.17094059776773451,
+      "grad_norm": 0.09279338270425797,
+      "learning_rate": 2.4934040902150243e-05,
+      "loss": 9.888,
+      "step": 34230
+    },
+    {
+      "epoch": 0.17099053659267396,
+      "grad_norm": 0.09093698859214783,
+      "learning_rate": 2.493253898720869e-05,
+      "loss": 9.8898,
+      "step": 34240
+    },
+    {
+      "epoch": 0.1710404754176134,
+      "grad_norm": 0.09663654863834381,
+      "learning_rate": 2.4931037072267144e-05,
+      "loss": 9.8794,
+      "step": 34250
+    },
+    {
+      "epoch": 0.17109041424255286,
+      "grad_norm": 0.09918060153722763,
+      "learning_rate": 2.492953515732559e-05,
+      "loss": 9.8779,
+      "step": 34260
+    },
+    {
+      "epoch": 0.1711403530674923,
+      "grad_norm": 0.09901434928178787,
+      "learning_rate": 2.492803324238404e-05,
+      "loss": 9.8849,
+      "step": 34270
+    },
+    {
+      "epoch": 0.17119029189243176,
+      "grad_norm": 0.09233778715133667,
+      "learning_rate": 2.492653132744249e-05,
+      "loss": 9.8817,
+      "step": 34280
+    },
+    {
+      "epoch": 0.1712402307173712,
+      "grad_norm": 0.08926848322153091,
+      "learning_rate": 2.4925029412500938e-05,
+      "loss": 9.8866,
+      "step": 34290
+    },
+    {
+      "epoch": 0.17129016954231066,
+      "grad_norm": 0.0978291779756546,
+      "learning_rate": 2.492352749755939e-05,
+      "loss": 9.8874,
+      "step": 34300
+    },
+    {
+      "epoch": 0.1713401083672501,
+      "grad_norm": 0.09829317033290863,
+      "learning_rate": 2.4922025582617838e-05,
+      "loss": 9.8843,
+      "step": 34310
+    },
+    {
+      "epoch": 0.17139004719218956,
+      "grad_norm": 0.08857450634241104,
+      "learning_rate": 2.4920523667676288e-05,
+      "loss": 9.8929,
+      "step": 34320
+    },
+    {
+      "epoch": 0.171439986017129,
+      "grad_norm": 0.10009986907243729,
+      "learning_rate": 2.4919021752734738e-05,
+      "loss": 9.8932,
+      "step": 34330
+    },
+    {
+      "epoch": 0.17148992484206846,
+      "grad_norm": 0.09351924806833267,
+      "learning_rate": 2.4917519837793185e-05,
+      "loss": 9.8878,
+      "step": 34340
+    },
+    {
+      "epoch": 0.1715398636670079,
+      "grad_norm": 0.09353633224964142,
+      "learning_rate": 2.491601792285164e-05,
+      "loss": 9.8878,
+      "step": 34350
+    },
+    {
+      "epoch": 0.17158980249194736,
+      "grad_norm": 0.09166029840707779,
+      "learning_rate": 2.4914516007910085e-05,
+      "loss": 9.8803,
+      "step": 34360
+    },
+    {
+      "epoch": 0.1716397413168868,
+      "grad_norm": 0.09764494746923447,
+      "learning_rate": 2.4913014092968536e-05,
+      "loss": 9.8855,
+      "step": 34370
+    },
+    {
+      "epoch": 0.17168968014182626,
+      "grad_norm": 0.0947176069021225,
+      "learning_rate": 2.4911512178026986e-05,
+      "loss": 9.8796,
+      "step": 34380
+    },
+    {
+      "epoch": 0.1717396189667657,
+      "grad_norm": 0.09326879680156708,
+      "learning_rate": 2.4910010263085433e-05,
+      "loss": 9.8775,
+      "step": 34390
+    },
+    {
+      "epoch": 0.17178955779170516,
+      "grad_norm": 0.0948229506611824,
+      "learning_rate": 2.4908508348143886e-05,
+      "loss": 9.8862,
+      "step": 34400
+    },
+    {
+      "epoch": 0.1718394966166446,
+      "grad_norm": 0.09219582378864288,
+      "learning_rate": 2.4907006433202333e-05,
+      "loss": 9.8888,
+      "step": 34410
+    },
+    {
+      "epoch": 0.17188943544158405,
+      "grad_norm": 0.0936218723654747,
+      "learning_rate": 2.4905504518260783e-05,
+      "loss": 9.8747,
+      "step": 34420
+    },
+    {
+      "epoch": 0.1719393742665235,
+      "grad_norm": 0.09321941435337067,
+      "learning_rate": 2.4904002603319233e-05,
+      "loss": 9.891,
+      "step": 34430
+    },
+    {
+      "epoch": 0.17198931309146295,
+      "grad_norm": 0.10023363679647446,
+      "learning_rate": 2.490250068837768e-05,
+      "loss": 9.8808,
+      "step": 34440
+    },
+    {
+      "epoch": 0.1720392519164024,
+      "grad_norm": 0.09513888508081436,
+      "learning_rate": 2.4900998773436134e-05,
+      "loss": 9.882,
+      "step": 34450
+    },
+    {
+      "epoch": 0.17208919074134185,
+      "grad_norm": 0.09558258205652237,
+      "learning_rate": 2.489949685849458e-05,
+      "loss": 9.8793,
+      "step": 34460
+    },
+    {
+      "epoch": 0.1721391295662813,
+      "grad_norm": 0.10544897615909576,
+      "learning_rate": 2.489799494355303e-05,
+      "loss": 9.8799,
+      "step": 34470
+    },
+    {
+      "epoch": 0.17218906839122075,
+      "grad_norm": 0.09681084007024765,
+      "learning_rate": 2.489649302861148e-05,
+      "loss": 9.879,
+      "step": 34480
+    },
+    {
+      "epoch": 0.1722390072161602,
+      "grad_norm": 0.10312927514314651,
+      "learning_rate": 2.4894991113669928e-05,
+      "loss": 9.8859,
+      "step": 34490
+    },
+    {
+      "epoch": 0.17228894604109965,
+      "grad_norm": 0.09519031643867493,
+      "learning_rate": 2.489348919872838e-05,
+      "loss": 9.8728,
+      "step": 34500
+    },
+    {
+      "epoch": 0.1723388848660391,
+      "grad_norm": 0.09442928433418274,
+      "learning_rate": 2.4891987283786828e-05,
+      "loss": 9.8753,
+      "step": 34510
+    },
+    {
+      "epoch": 0.17238882369097855,
+      "grad_norm": 0.09661301970481873,
+      "learning_rate": 2.4890485368845278e-05,
+      "loss": 9.8788,
+      "step": 34520
+    },
+    {
+      "epoch": 0.172438762515918,
+      "grad_norm": 0.0956299677491188,
+      "learning_rate": 2.488898345390373e-05,
+      "loss": 9.8807,
+      "step": 34530
+    },
+    {
+      "epoch": 0.17248870134085745,
+      "grad_norm": 0.097622811794281,
+      "learning_rate": 2.488748153896218e-05,
+      "loss": 9.8781,
+      "step": 34540
+    },
+    {
+      "epoch": 0.1725386401657969,
+      "grad_norm": 0.0975886806845665,
+      "learning_rate": 2.488597962402063e-05,
+      "loss": 9.8766,
+      "step": 34550
+    },
+    {
+      "epoch": 0.17258857899073635,
+      "grad_norm": 0.09032001346349716,
+      "learning_rate": 2.4884477709079075e-05,
+      "loss": 9.8834,
+      "step": 34560
+    },
+    {
+      "epoch": 0.1726385178156758,
+      "grad_norm": 0.09692685306072235,
+      "learning_rate": 2.4882975794137526e-05,
+      "loss": 9.8768,
+      "step": 34570
+    },
+    {
+      "epoch": 0.17268845664061525,
+      "grad_norm": 0.09569180756807327,
+      "learning_rate": 2.4881473879195976e-05,
+      "loss": 9.8756,
+      "step": 34580
+    },
+    {
+      "epoch": 0.1727383954655547,
+      "grad_norm": 0.0921887531876564,
+      "learning_rate": 2.4879971964254426e-05,
+      "loss": 9.8815,
+      "step": 34590
+    },
+    {
+      "epoch": 0.17278833429049414,
+      "grad_norm": 0.09386293590068817,
+      "learning_rate": 2.4878470049312876e-05,
+      "loss": 9.8763,
+      "step": 34600
+    },
+    {
+      "epoch": 0.1728382731154336,
+      "grad_norm": 0.09377525746822357,
+      "learning_rate": 2.4876968134371323e-05,
+      "loss": 9.8792,
+      "step": 34610
+    },
+    {
+      "epoch": 0.17288821194037304,
+      "grad_norm": 0.09777195751667023,
+      "learning_rate": 2.4875466219429773e-05,
+      "loss": 9.8779,
+      "step": 34620
+    },
+    {
+      "epoch": 0.1729381507653125,
+      "grad_norm": 0.08953432738780975,
+      "learning_rate": 2.4873964304488223e-05,
+      "loss": 9.8801,
+      "step": 34630
+    },
+    {
+      "epoch": 0.17298808959025194,
+      "grad_norm": 0.09429240971803665,
+      "learning_rate": 2.4872462389546674e-05,
+      "loss": 9.8756,
+      "step": 34640
+    },
+    {
+      "epoch": 0.1730380284151914,
+      "grad_norm": 0.09474718570709229,
+      "learning_rate": 2.4870960474605124e-05,
+      "loss": 9.8751,
+      "step": 34650
+    },
+    {
+      "epoch": 0.17308796724013084,
+      "grad_norm": 0.09288160502910614,
+      "learning_rate": 2.486945855966357e-05,
+      "loss": 9.8807,
+      "step": 34660
+    },
+    {
+      "epoch": 0.1731379060650703,
+      "grad_norm": 0.09404736012220383,
+      "learning_rate": 2.486795664472202e-05,
+      "loss": 9.8768,
+      "step": 34670
+    },
+    {
+      "epoch": 0.17318784489000974,
+      "grad_norm": 0.0975610539317131,
+      "learning_rate": 2.486645472978047e-05,
+      "loss": 9.8753,
+      "step": 34680
+    },
+    {
+      "epoch": 0.1732377837149492,
+      "grad_norm": 0.0982472151517868,
+      "learning_rate": 2.486495281483892e-05,
+      "loss": 9.8792,
+      "step": 34690
+    },
+    {
+      "epoch": 0.17328772253988864,
+      "grad_norm": 0.09330155700445175,
+      "learning_rate": 2.486345089989737e-05,
+      "loss": 9.8726,
+      "step": 34700
+    },
+    {
+      "epoch": 0.1733376613648281,
+      "grad_norm": 0.09651447087526321,
+      "learning_rate": 2.4861948984955818e-05,
+      "loss": 9.8716,
+      "step": 34710
+    },
+    {
+      "epoch": 0.17338760018976754,
+      "grad_norm": 0.09619554132223129,
+      "learning_rate": 2.4860447070014268e-05,
+      "loss": 9.8736,
+      "step": 34720
+    },
+    {
+      "epoch": 0.173437539014707,
+      "grad_norm": 0.09537581354379654,
+      "learning_rate": 2.485894515507272e-05,
+      "loss": 9.8743,
+      "step": 34730
+    },
+    {
+      "epoch": 0.17348747783964644,
+      "grad_norm": 0.09322098642587662,
+      "learning_rate": 2.485744324013117e-05,
+      "loss": 9.8832,
+      "step": 34740
+    },
+    {
+      "epoch": 0.1735374166645859,
+      "grad_norm": 0.09934062510728836,
+      "learning_rate": 2.485594132518962e-05,
+      "loss": 9.8735,
+      "step": 34750
+    },
+    {
+      "epoch": 0.17358735548952534,
+      "grad_norm": 0.09587859362363815,
+      "learning_rate": 2.4854439410248065e-05,
+      "loss": 9.8625,
+      "step": 34760
+    },
+    {
+      "epoch": 0.17363729431446479,
+      "grad_norm": 0.09331879019737244,
+      "learning_rate": 2.4852937495306516e-05,
+      "loss": 9.8709,
+      "step": 34770
+    },
+    {
+      "epoch": 0.17368723313940423,
+      "grad_norm": 0.0957440435886383,
+      "learning_rate": 2.4851435580364966e-05,
+      "loss": 9.8665,
+      "step": 34780
+    },
+    {
+      "epoch": 0.17373717196434368,
+      "grad_norm": 0.10028447955846786,
+      "learning_rate": 2.4849933665423416e-05,
+      "loss": 9.8696,
+      "step": 34790
+    },
+    {
+      "epoch": 0.17378711078928313,
+      "grad_norm": 0.09930728375911713,
+      "learning_rate": 2.4848431750481866e-05,
+      "loss": 9.8731,
+      "step": 34800
+    },
+    {
+      "epoch": 0.17383704961422258,
+      "grad_norm": 0.09186220914125443,
+      "learning_rate": 2.4846929835540313e-05,
+      "loss": 9.8791,
+      "step": 34810
+    },
+    {
+      "epoch": 0.17388698843916203,
+      "grad_norm": 0.09665971249341965,
+      "learning_rate": 2.4845427920598763e-05,
+      "loss": 9.8828,
+      "step": 34820
+    },
+    {
+      "epoch": 0.17393692726410148,
+      "grad_norm": 0.09465493261814117,
+      "learning_rate": 2.4843926005657213e-05,
+      "loss": 9.8729,
+      "step": 34830
+    },
+    {
+      "epoch": 0.17398686608904093,
+      "grad_norm": 0.09141222387552261,
+      "learning_rate": 2.4842424090715664e-05,
+      "loss": 9.8718,
+      "step": 34840
+    },
+    {
+      "epoch": 0.17403680491398038,
+      "grad_norm": 0.09826017171144485,
+      "learning_rate": 2.4840922175774114e-05,
+      "loss": 9.8765,
+      "step": 34850
+    },
+    {
+      "epoch": 0.17408674373891983,
+      "grad_norm": 0.10129930078983307,
+      "learning_rate": 2.4839420260832564e-05,
+      "loss": 9.8716,
+      "step": 34860
+    },
+    {
+      "epoch": 0.17413668256385928,
+      "grad_norm": 0.09397407621145248,
+      "learning_rate": 2.483791834589101e-05,
+      "loss": 9.875,
+      "step": 34870
+    },
+    {
+      "epoch": 0.17418662138879873,
+      "grad_norm": 0.09884011745452881,
+      "learning_rate": 2.483641643094946e-05,
+      "loss": 9.8708,
+      "step": 34880
+    },
+    {
+      "epoch": 0.17423656021373818,
+      "grad_norm": 0.09467148035764694,
+      "learning_rate": 2.483491451600791e-05,
+      "loss": 9.8709,
+      "step": 34890
+    },
+    {
+      "epoch": 0.17428649903867763,
+      "grad_norm": 0.0914902612566948,
+      "learning_rate": 2.483341260106636e-05,
+      "loss": 9.8711,
+      "step": 34900
+    },
+    {
+      "epoch": 0.17433643786361708,
+      "grad_norm": 0.0950956791639328,
+      "learning_rate": 2.483191068612481e-05,
+      "loss": 9.8675,
+      "step": 34910
+    },
+    {
+      "epoch": 0.17438637668855653,
+      "grad_norm": 0.09284093976020813,
+      "learning_rate": 2.4830408771183258e-05,
+      "loss": 9.8682,
+      "step": 34920
+    },
+    {
+      "epoch": 0.17443631551349598,
+      "grad_norm": 0.09593179076910019,
+      "learning_rate": 2.482890685624171e-05,
+      "loss": 9.8714,
+      "step": 34930
+    },
+    {
+      "epoch": 0.17448625433843543,
+      "grad_norm": 0.09419502317905426,
+      "learning_rate": 2.482740494130016e-05,
+      "loss": 9.8694,
+      "step": 34940
+    },
+    {
+      "epoch": 0.17453619316337488,
+      "grad_norm": 0.09543459862470627,
+      "learning_rate": 2.482590302635861e-05,
+      "loss": 9.8715,
+      "step": 34950
+    },
+    {
+      "epoch": 0.17458613198831432,
+      "grad_norm": 0.09066810458898544,
+      "learning_rate": 2.482440111141706e-05,
+      "loss": 9.8734,
+      "step": 34960
+    },
+    {
+      "epoch": 0.17463607081325377,
+      "grad_norm": 0.09701773524284363,
+      "learning_rate": 2.4822899196475506e-05,
+      "loss": 9.8711,
+      "step": 34970
+    },
+    {
+      "epoch": 0.17468600963819322,
+      "grad_norm": 0.09097738564014435,
+      "learning_rate": 2.4821397281533956e-05,
+      "loss": 9.875,
+      "step": 34980
+    },
+    {
+      "epoch": 0.17473594846313267,
+      "grad_norm": 0.0955294743180275,
+      "learning_rate": 2.4819895366592406e-05,
+      "loss": 9.8632,
+      "step": 34990
+    },
+    {
+      "epoch": 0.17478588728807212,
+      "grad_norm": 0.09581121802330017,
+      "learning_rate": 2.4818393451650856e-05,
+      "loss": 9.8698,
+      "step": 35000
+    },
+    {
+      "epoch": 0.17483582611301157,
+      "grad_norm": 0.09599103778600693,
+      "learning_rate": 2.4816891536709306e-05,
+      "loss": 9.867,
+      "step": 35010
+    },
+    {
+      "epoch": 0.17488576493795102,
+      "grad_norm": 0.1030481606721878,
+      "learning_rate": 2.4815389621767753e-05,
+      "loss": 9.8662,
+      "step": 35020
+    },
+    {
+      "epoch": 0.17493570376289047,
+      "grad_norm": 0.09832312166690826,
+      "learning_rate": 2.4813887706826203e-05,
+      "loss": 9.8675,
+      "step": 35030
+    },
+    {
+      "epoch": 0.17498564258782992,
+      "grad_norm": 0.09342101961374283,
+      "learning_rate": 2.4812385791884654e-05,
+      "loss": 9.8692,
+      "step": 35040
+    },
+    {
+      "epoch": 0.17503558141276937,
+      "grad_norm": 0.09640762209892273,
+      "learning_rate": 2.4810883876943104e-05,
+      "loss": 9.8705,
+      "step": 35050
+    },
+    {
+      "epoch": 0.17508552023770882,
+      "grad_norm": 0.09944464266300201,
+      "learning_rate": 2.4809381962001554e-05,
+      "loss": 9.8669,
+      "step": 35060
+    },
+    {
+      "epoch": 0.17513545906264827,
+      "grad_norm": 0.09033916890621185,
+      "learning_rate": 2.480788004706e-05,
+      "loss": 9.8649,
+      "step": 35070
+    },
+    {
+      "epoch": 0.17518539788758772,
+      "grad_norm": 0.09908484667539597,
+      "learning_rate": 2.480637813211845e-05,
+      "loss": 9.8719,
+      "step": 35080
+    },
+    {
+      "epoch": 0.17523533671252717,
+      "grad_norm": 0.09301396459341049,
+      "learning_rate": 2.48048762171769e-05,
+      "loss": 9.8616,
+      "step": 35090
+    },
+    {
+      "epoch": 0.17528527553746662,
+      "grad_norm": 0.09278962016105652,
+      "learning_rate": 2.480337430223535e-05,
+      "loss": 9.8608,
+      "step": 35100
+    },
+    {
+      "epoch": 0.17533521436240604,
+      "grad_norm": 0.09971888363361359,
+      "learning_rate": 2.48018723872938e-05,
+      "loss": 9.8672,
+      "step": 35110
+    },
+    {
+      "epoch": 0.1753851531873455,
+      "grad_norm": 0.09398912638425827,
+      "learning_rate": 2.4800370472352248e-05,
+      "loss": 9.8683,
+      "step": 35120
+    },
+    {
+      "epoch": 0.17543509201228494,
+      "grad_norm": 0.09367619454860687,
+      "learning_rate": 2.47988685574107e-05,
+      "loss": 9.8646,
+      "step": 35130
+    },
+    {
+      "epoch": 0.1754850308372244,
+      "grad_norm": 0.08890117704868317,
+      "learning_rate": 2.479736664246915e-05,
+      "loss": 9.8611,
+      "step": 35140
+    },
+    {
+      "epoch": 0.17553496966216384,
+      "grad_norm": 0.09400789439678192,
+      "learning_rate": 2.47958647275276e-05,
+      "loss": 9.8675,
+      "step": 35150
+    },
+    {
+      "epoch": 0.1755849084871033,
+      "grad_norm": 0.0941472202539444,
+      "learning_rate": 2.479436281258605e-05,
+      "loss": 9.8662,
+      "step": 35160
+    },
+    {
+      "epoch": 0.17563484731204274,
+      "grad_norm": 0.0907822698354721,
+      "learning_rate": 2.4792860897644496e-05,
+      "loss": 9.8724,
+      "step": 35170
+    },
+    {
+      "epoch": 0.17568478613698219,
+      "grad_norm": 0.09430518746376038,
+      "learning_rate": 2.479135898270295e-05,
+      "loss": 9.8677,
+      "step": 35180
+    },
+    {
+      "epoch": 0.17573472496192163,
+      "grad_norm": 0.09625846147537231,
+      "learning_rate": 2.4789857067761396e-05,
+      "loss": 9.86,
+      "step": 35190
+    },
+    {
+      "epoch": 0.17578466378686108,
+      "grad_norm": 0.09888491779565811,
+      "learning_rate": 2.4788355152819846e-05,
+      "loss": 9.864,
+      "step": 35200
+    },
+    {
+      "epoch": 0.17583460261180053,
+      "grad_norm": 0.10357237607240677,
+      "learning_rate": 2.4786853237878296e-05,
+      "loss": 9.8568,
+      "step": 35210
+    },
+    {
+      "epoch": 0.17588454143673998,
+      "grad_norm": 0.09656186401844025,
+      "learning_rate": 2.4785351322936743e-05,
+      "loss": 9.8683,
+      "step": 35220
+    },
+    {
+      "epoch": 0.17593448026167943,
+      "grad_norm": 0.09330909699201584,
+      "learning_rate": 2.4783849407995197e-05,
+      "loss": 9.8613,
+      "step": 35230
+    },
+    {
+      "epoch": 0.17598441908661888,
+      "grad_norm": 0.09761014580726624,
+      "learning_rate": 2.4782347493053644e-05,
+      "loss": 9.864,
+      "step": 35240
+    },
+    {
+      "epoch": 0.17603435791155833,
+      "grad_norm": 0.09538024663925171,
+      "learning_rate": 2.4780845578112094e-05,
+      "loss": 9.8589,
+      "step": 35250
+    },
+    {
+      "epoch": 0.17608429673649778,
+      "grad_norm": 0.09242220222949982,
+      "learning_rate": 2.4779343663170544e-05,
+      "loss": 9.8593,
+      "step": 35260
+    },
+    {
+      "epoch": 0.17613423556143723,
+      "grad_norm": 0.09523601830005646,
+      "learning_rate": 2.477784174822899e-05,
+      "loss": 9.8591,
+      "step": 35270
+    },
+    {
+      "epoch": 0.17618417438637668,
+      "grad_norm": 0.09717884659767151,
+      "learning_rate": 2.4776339833287444e-05,
+      "loss": 9.8658,
+      "step": 35280
+    },
+    {
+      "epoch": 0.17623411321131613,
+      "grad_norm": 0.11009235680103302,
+      "learning_rate": 2.477483791834589e-05,
+      "loss": 9.8577,
+      "step": 35290
+    },
+    {
+      "epoch": 0.17628405203625558,
+      "grad_norm": 0.09665638953447342,
+      "learning_rate": 2.477333600340434e-05,
+      "loss": 9.8682,
+      "step": 35300
+    },
+    {
+      "epoch": 0.17633399086119503,
+      "grad_norm": 0.09212061762809753,
+      "learning_rate": 2.477183408846279e-05,
+      "loss": 9.8629,
+      "step": 35310
+    },
+    {
+      "epoch": 0.17638392968613448,
+      "grad_norm": 0.09347554296255112,
+      "learning_rate": 2.4770332173521238e-05,
+      "loss": 9.8615,
+      "step": 35320
+    },
+    {
+      "epoch": 0.17643386851107393,
+      "grad_norm": 0.09957797080278397,
+      "learning_rate": 2.4768830258579692e-05,
+      "loss": 9.8613,
+      "step": 35330
+    },
+    {
+      "epoch": 0.17648380733601338,
+      "grad_norm": 0.0971374586224556,
+      "learning_rate": 2.476732834363814e-05,
+      "loss": 9.861,
+      "step": 35340
+    },
+    {
+      "epoch": 0.17653374616095283,
+      "grad_norm": 0.09239890426397324,
+      "learning_rate": 2.476582642869659e-05,
+      "loss": 9.865,
+      "step": 35350
+    },
+    {
+      "epoch": 0.17658368498589228,
+      "grad_norm": 0.09785747528076172,
+      "learning_rate": 2.476432451375504e-05,
+      "loss": 9.8611,
+      "step": 35360
+    },
+    {
+      "epoch": 0.17663362381083172,
+      "grad_norm": 0.09622731059789658,
+      "learning_rate": 2.4762822598813486e-05,
+      "loss": 9.8552,
+      "step": 35370
+    },
+    {
+      "epoch": 0.17668356263577117,
+      "grad_norm": 0.09199772030115128,
+      "learning_rate": 2.476132068387194e-05,
+      "loss": 9.8599,
+      "step": 35380
+    },
+    {
+      "epoch": 0.17673350146071062,
+      "grad_norm": 0.09622587263584137,
+      "learning_rate": 2.4759818768930386e-05,
+      "loss": 9.8641,
+      "step": 35390
+    },
+    {
+      "epoch": 0.17678344028565007,
+      "grad_norm": 0.09253069758415222,
+      "learning_rate": 2.4758316853988836e-05,
+      "loss": 9.8557,
+      "step": 35400
+    },
+    {
+      "epoch": 0.17683337911058952,
+      "grad_norm": 0.09692025184631348,
+      "learning_rate": 2.4756814939047286e-05,
+      "loss": 9.8591,
+      "step": 35410
+    },
+    {
+      "epoch": 0.17688331793552897,
+      "grad_norm": 0.0917033851146698,
+      "learning_rate": 2.4755313024105733e-05,
+      "loss": 9.8558,
+      "step": 35420
+    },
+    {
+      "epoch": 0.17693325676046842,
+      "grad_norm": 0.09357017278671265,
+      "learning_rate": 2.4753811109164187e-05,
+      "loss": 9.8624,
+      "step": 35430
+    },
+    {
+      "epoch": 0.17698319558540787,
+      "grad_norm": 0.09529455006122589,
+      "learning_rate": 2.4752309194222634e-05,
+      "loss": 9.8557,
+      "step": 35440
+    },
+    {
+      "epoch": 0.17703313441034732,
+      "grad_norm": 0.0990082174539566,
+      "learning_rate": 2.4750807279281084e-05,
+      "loss": 9.8607,
+      "step": 35450
+    },
+    {
+      "epoch": 0.17708307323528677,
+      "grad_norm": 0.09450554102659225,
+      "learning_rate": 2.4749305364339534e-05,
+      "loss": 9.8506,
+      "step": 35460
+    },
+    {
+      "epoch": 0.17713301206022622,
+      "grad_norm": 0.09906260669231415,
+      "learning_rate": 2.474780344939798e-05,
+      "loss": 9.8575,
+      "step": 35470
+    },
+    {
+      "epoch": 0.17718295088516567,
+      "grad_norm": 0.09663223475217819,
+      "learning_rate": 2.4746301534456434e-05,
+      "loss": 9.8683,
+      "step": 35480
+    },
+    {
+      "epoch": 0.17723288971010512,
+      "grad_norm": 0.0887630507349968,
+      "learning_rate": 2.474479961951488e-05,
+      "loss": 9.8639,
+      "step": 35490
+    },
+    {
+      "epoch": 0.17728282853504457,
+      "grad_norm": 0.09145206212997437,
+      "learning_rate": 2.474329770457333e-05,
+      "loss": 9.8622,
+      "step": 35500
+    },
+    {
+      "epoch": 0.17733276735998402,
+      "grad_norm": 0.09705221652984619,
+      "learning_rate": 2.474179578963178e-05,
+      "loss": 9.8569,
+      "step": 35510
+    },
+    {
+      "epoch": 0.17738270618492347,
+      "grad_norm": 0.09884809702634811,
+      "learning_rate": 2.4740293874690228e-05,
+      "loss": 9.8575,
+      "step": 35520
+    },
+    {
+      "epoch": 0.17743264500986292,
+      "grad_norm": 0.09495537728071213,
+      "learning_rate": 2.4738791959748682e-05,
+      "loss": 9.8571,
+      "step": 35530
+    },
+    {
+      "epoch": 0.17748258383480237,
+      "grad_norm": 0.0902133509516716,
+      "learning_rate": 2.473729004480713e-05,
+      "loss": 9.8606,
+      "step": 35540
+    },
+    {
+      "epoch": 0.17753252265974181,
+      "grad_norm": 0.09612247347831726,
+      "learning_rate": 2.4735788129865582e-05,
+      "loss": 9.8519,
+      "step": 35550
+    },
+    {
+      "epoch": 0.17758246148468126,
+      "grad_norm": 0.09739767014980316,
+      "learning_rate": 2.473428621492403e-05,
+      "loss": 9.8538,
+      "step": 35560
+    },
+    {
+      "epoch": 0.1776324003096207,
+      "grad_norm": 0.10079686343669891,
+      "learning_rate": 2.4732784299982476e-05,
+      "loss": 9.8571,
+      "step": 35570
+    },
+    {
+      "epoch": 0.17768233913456016,
+      "grad_norm": 0.0968751311302185,
+      "learning_rate": 2.473128238504093e-05,
+      "loss": 9.8485,
+      "step": 35580
+    },
+    {
+      "epoch": 0.1777322779594996,
+      "grad_norm": 0.10343873500823975,
+      "learning_rate": 2.4729780470099376e-05,
+      "loss": 9.8578,
+      "step": 35590
+    },
+    {
+      "epoch": 0.17778221678443906,
+      "grad_norm": 0.09628947824239731,
+      "learning_rate": 2.472827855515783e-05,
+      "loss": 9.8457,
+      "step": 35600
+    },
+    {
+      "epoch": 0.1778321556093785,
+      "grad_norm": 0.08937991410493851,
+      "learning_rate": 2.4726776640216276e-05,
+      "loss": 9.8513,
+      "step": 35610
+    },
+    {
+      "epoch": 0.17788209443431796,
+      "grad_norm": 0.09612028300762177,
+      "learning_rate": 2.4725274725274723e-05,
+      "loss": 9.8552,
+      "step": 35620
+    },
+    {
+      "epoch": 0.1779320332592574,
+      "grad_norm": 0.09348214417695999,
+      "learning_rate": 2.4723772810333177e-05,
+      "loss": 9.8598,
+      "step": 35630
+    },
+    {
+      "epoch": 0.17798197208419686,
+      "grad_norm": 0.09429159015417099,
+      "learning_rate": 2.4722270895391624e-05,
+      "loss": 9.8548,
+      "step": 35640
+    },
+    {
+      "epoch": 0.1780319109091363,
+      "grad_norm": 0.09209638088941574,
+      "learning_rate": 2.4720768980450077e-05,
+      "loss": 9.8631,
+      "step": 35650
+    },
+    {
+      "epoch": 0.17808184973407576,
+      "grad_norm": 0.1002785861492157,
+      "learning_rate": 2.4719267065508524e-05,
+      "loss": 9.845,
+      "step": 35660
+    },
+    {
+      "epoch": 0.1781317885590152,
+      "grad_norm": 0.09212984144687653,
+      "learning_rate": 2.471776515056697e-05,
+      "loss": 9.8511,
+      "step": 35670
+    },
+    {
+      "epoch": 0.17818172738395466,
+      "grad_norm": 0.09603580087423325,
+      "learning_rate": 2.4716263235625424e-05,
+      "loss": 9.8495,
+      "step": 35680
+    },
+    {
+      "epoch": 0.1782316662088941,
+      "grad_norm": 0.0989808738231659,
+      "learning_rate": 2.471476132068387e-05,
+      "loss": 9.8624,
+      "step": 35690
+    },
+    {
+      "epoch": 0.17828160503383356,
+      "grad_norm": 0.09753132611513138,
+      "learning_rate": 2.4713259405742325e-05,
+      "loss": 9.8588,
+      "step": 35700
+    },
+    {
+      "epoch": 0.178331543858773,
+      "grad_norm": 0.09459420293569565,
+      "learning_rate": 2.471175749080077e-05,
+      "loss": 9.8551,
+      "step": 35710
+    },
+    {
+      "epoch": 0.17838148268371246,
+      "grad_norm": 0.09288576990365982,
+      "learning_rate": 2.4710255575859218e-05,
+      "loss": 9.852,
+      "step": 35720
+    },
+    {
+      "epoch": 0.1784314215086519,
+      "grad_norm": 0.09371695667505264,
+      "learning_rate": 2.4708753660917672e-05,
+      "loss": 9.8552,
+      "step": 35730
+    },
+    {
+      "epoch": 0.17848136033359135,
+      "grad_norm": 0.09233945608139038,
+      "learning_rate": 2.470725174597612e-05,
+      "loss": 9.8561,
+      "step": 35740
+    },
+    {
+      "epoch": 0.1785312991585308,
+      "grad_norm": 0.09840219467878342,
+      "learning_rate": 2.4705749831034572e-05,
+      "loss": 9.8552,
+      "step": 35750
+    },
+    {
+      "epoch": 0.17858123798347025,
+      "grad_norm": 0.09583665430545807,
+      "learning_rate": 2.470424791609302e-05,
+      "loss": 9.8496,
+      "step": 35760
+    },
+    {
+      "epoch": 0.1786311768084097,
+      "grad_norm": 0.09573546051979065,
+      "learning_rate": 2.4702746001151466e-05,
+      "loss": 9.8575,
+      "step": 35770
+    },
+    {
+      "epoch": 0.17868111563334915,
+      "grad_norm": 0.09594116359949112,
+      "learning_rate": 2.470124408620992e-05,
+      "loss": 9.8538,
+      "step": 35780
+    },
+    {
+      "epoch": 0.1787310544582886,
+      "grad_norm": 0.09792909026145935,
+      "learning_rate": 2.4699742171268366e-05,
+      "loss": 9.853,
+      "step": 35790
+    },
+    {
+      "epoch": 0.17878099328322805,
+      "grad_norm": 0.09887704998254776,
+      "learning_rate": 2.469824025632682e-05,
+      "loss": 9.8519,
+      "step": 35800
+    },
+    {
+      "epoch": 0.1788309321081675,
+      "grad_norm": 0.09814807772636414,
+      "learning_rate": 2.4696738341385266e-05,
+      "loss": 9.8464,
+      "step": 35810
+    },
+    {
+      "epoch": 0.17888087093310695,
+      "grad_norm": 0.09986128658056259,
+      "learning_rate": 2.4695236426443713e-05,
+      "loss": 9.8488,
+      "step": 35820
+    },
+    {
+      "epoch": 0.1789308097580464,
+      "grad_norm": 0.0871347114443779,
+      "learning_rate": 2.4693734511502167e-05,
+      "loss": 9.8522,
+      "step": 35830
+    },
+    {
+      "epoch": 0.17898074858298585,
+      "grad_norm": 0.09611291438341141,
+      "learning_rate": 2.4692232596560614e-05,
+      "loss": 9.8551,
+      "step": 35840
+    },
+    {
+      "epoch": 0.1790306874079253,
+      "grad_norm": 0.1038837879896164,
+      "learning_rate": 2.4690730681619067e-05,
+      "loss": 9.8544,
+      "step": 35850
+    },
+    {
+      "epoch": 0.17908062623286475,
+      "grad_norm": 0.09933873265981674,
+      "learning_rate": 2.4689228766677514e-05,
+      "loss": 9.8551,
+      "step": 35860
+    },
+    {
+      "epoch": 0.1791305650578042,
+      "grad_norm": 0.0895809531211853,
+      "learning_rate": 2.4687726851735964e-05,
+      "loss": 9.8486,
+      "step": 35870
+    },
+    {
+      "epoch": 0.17918050388274365,
+      "grad_norm": 0.09082268923521042,
+      "learning_rate": 2.4686224936794414e-05,
+      "loss": 9.8563,
+      "step": 35880
+    },
+    {
+      "epoch": 0.1792304427076831,
+      "grad_norm": 0.09877155721187592,
+      "learning_rate": 2.468472302185286e-05,
+      "loss": 9.8526,
+      "step": 35890
+    },
+    {
+      "epoch": 0.17928038153262255,
+      "grad_norm": 0.09575866162776947,
+      "learning_rate": 2.4683221106911315e-05,
+      "loss": 9.857,
+      "step": 35900
+    },
+    {
+      "epoch": 0.179330320357562,
+      "grad_norm": 0.0871095061302185,
+      "learning_rate": 2.468171919196976e-05,
+      "loss": 9.8527,
+      "step": 35910
+    },
+    {
+      "epoch": 0.17938025918250144,
+      "grad_norm": 0.09860642999410629,
+      "learning_rate": 2.468021727702821e-05,
+      "loss": 9.8475,
+      "step": 35920
+    },
+    {
+      "epoch": 0.1794301980074409,
+      "grad_norm": 0.09469640254974365,
+      "learning_rate": 2.4678715362086662e-05,
+      "loss": 9.8499,
+      "step": 35930
+    },
+    {
+      "epoch": 0.17948013683238034,
+      "grad_norm": 0.09654894471168518,
+      "learning_rate": 2.467721344714511e-05,
+      "loss": 9.8446,
+      "step": 35940
+    },
+    {
+      "epoch": 0.1795300756573198,
+      "grad_norm": 0.09052994847297668,
+      "learning_rate": 2.4675711532203562e-05,
+      "loss": 9.8469,
+      "step": 35950
+    },
+    {
+      "epoch": 0.17958001448225924,
+      "grad_norm": 0.09217459708452225,
+      "learning_rate": 2.467420961726201e-05,
+      "loss": 9.8483,
+      "step": 35960
+    },
+    {
+      "epoch": 0.1796299533071987,
+      "grad_norm": 0.09349222481250763,
+      "learning_rate": 2.467270770232046e-05,
+      "loss": 9.8498,
+      "step": 35970
+    },
+    {
+      "epoch": 0.17967989213213814,
+      "grad_norm": 0.09795039892196655,
+      "learning_rate": 2.467120578737891e-05,
+      "loss": 9.848,
+      "step": 35980
+    },
+    {
+      "epoch": 0.1797298309570776,
+      "grad_norm": 0.09592344611883163,
+      "learning_rate": 2.4669703872437356e-05,
+      "loss": 9.8407,
+      "step": 35990
+    },
+    {
+      "epoch": 0.17977976978201704,
+      "grad_norm": 0.09404613822698593,
+      "learning_rate": 2.466820195749581e-05,
+      "loss": 9.8429,
+      "step": 36000
+    },
+    {
+      "epoch": 0.1798297086069565,
+      "grad_norm": 0.09326544404029846,
+      "learning_rate": 2.4666700042554256e-05,
+      "loss": 9.8567,
+      "step": 36010
+    },
+    {
+      "epoch": 0.17987964743189594,
+      "grad_norm": 0.09783726930618286,
+      "learning_rate": 2.4665198127612707e-05,
+      "loss": 9.8545,
+      "step": 36020
+    },
+    {
+      "epoch": 0.1799295862568354,
+      "grad_norm": 0.09461402893066406,
+      "learning_rate": 2.4663696212671157e-05,
+      "loss": 9.8438,
+      "step": 36030
+    },
+    {
+      "epoch": 0.17997952508177484,
+      "grad_norm": 0.09616295248270035,
+      "learning_rate": 2.4662194297729604e-05,
+      "loss": 9.8417,
+      "step": 36040
+    },
+    {
+      "epoch": 0.1800294639067143,
+      "grad_norm": 0.09772490710020065,
+      "learning_rate": 2.4660692382788057e-05,
+      "loss": 9.8518,
+      "step": 36050
+    },
+    {
+      "epoch": 0.18007940273165374,
+      "grad_norm": 0.09420907497406006,
+      "learning_rate": 2.4659190467846504e-05,
+      "loss": 9.8473,
+      "step": 36060
+    },
+    {
+      "epoch": 0.1801293415565932,
+      "grad_norm": 0.09083731472492218,
+      "learning_rate": 2.4657688552904954e-05,
+      "loss": 9.8436,
+      "step": 36070
+    },
+    {
+      "epoch": 0.18017928038153264,
+      "grad_norm": 0.09854559600353241,
+      "learning_rate": 2.4656186637963404e-05,
+      "loss": 9.8468,
+      "step": 36080
+    },
+    {
+      "epoch": 0.18022921920647209,
+      "grad_norm": 0.09707348048686981,
+      "learning_rate": 2.465468472302185e-05,
+      "loss": 9.8503,
+      "step": 36090
+    },
+    {
+      "epoch": 0.1802791580314115,
+      "grad_norm": 0.09600484371185303,
+      "learning_rate": 2.4653182808080305e-05,
+      "loss": 9.8457,
+      "step": 36100
+    },
+    {
+      "epoch": 0.18032909685635096,
+      "grad_norm": 0.0931532233953476,
+      "learning_rate": 2.465168089313875e-05,
+      "loss": 9.8469,
+      "step": 36110
+    },
+    {
+      "epoch": 0.1803790356812904,
+      "grad_norm": 0.09178706258535385,
+      "learning_rate": 2.46501789781972e-05,
+      "loss": 9.8432,
+      "step": 36120
+    },
+    {
+      "epoch": 0.18042897450622986,
+      "grad_norm": 0.09376762807369232,
+      "learning_rate": 2.4648677063255652e-05,
+      "loss": 9.8451,
+      "step": 36130
+    },
+    {
+      "epoch": 0.1804789133311693,
+      "grad_norm": 0.09681384265422821,
+      "learning_rate": 2.46471751483141e-05,
+      "loss": 9.8427,
+      "step": 36140
+    },
+    {
+      "epoch": 0.18052885215610875,
+      "grad_norm": 0.09608221054077148,
+      "learning_rate": 2.4645673233372552e-05,
+      "loss": 9.8498,
+      "step": 36150
+    },
+    {
+      "epoch": 0.1805787909810482,
+      "grad_norm": 0.0937625989317894,
+      "learning_rate": 2.4644171318431e-05,
+      "loss": 9.8502,
+      "step": 36160
+    },
+    {
+      "epoch": 0.18062872980598765,
+      "grad_norm": 0.09719646722078323,
+      "learning_rate": 2.464266940348945e-05,
+      "loss": 9.849,
+      "step": 36170
+    },
+    {
+      "epoch": 0.1806786686309271,
+      "grad_norm": 0.09256922453641891,
+      "learning_rate": 2.46411674885479e-05,
+      "loss": 9.8471,
+      "step": 36180
+    },
+    {
+      "epoch": 0.18072860745586655,
+      "grad_norm": 0.09239103645086288,
+      "learning_rate": 2.463966557360635e-05,
+      "loss": 9.8389,
+      "step": 36190
+    },
+    {
+      "epoch": 0.180778546280806,
+      "grad_norm": 0.09770379960536957,
+      "learning_rate": 2.46381636586648e-05,
+      "loss": 9.8456,
+      "step": 36200
+    },
+    {
+      "epoch": 0.18082848510574545,
+      "grad_norm": 0.09287633001804352,
+      "learning_rate": 2.4636661743723246e-05,
+      "loss": 9.8502,
+      "step": 36210
+    },
+    {
+      "epoch": 0.1808784239306849,
+      "grad_norm": 0.0940169245004654,
+      "learning_rate": 2.4635159828781697e-05,
+      "loss": 9.848,
+      "step": 36220
+    },
+    {
+      "epoch": 0.18092836275562435,
+      "grad_norm": 0.0958079993724823,
+      "learning_rate": 2.4633657913840147e-05,
+      "loss": 9.8366,
+      "step": 36230
+    },
+    {
+      "epoch": 0.1809783015805638,
+      "grad_norm": 0.09197930246591568,
+      "learning_rate": 2.4632155998898597e-05,
+      "loss": 12.8969,
+      "step": 36240
+    },
+    {
+      "epoch": 0.18102824040550325,
+      "grad_norm": 0.08773362636566162,
+      "learning_rate": 2.4630654083957047e-05,
+      "loss": 14.5353,
+      "step": 36250
+    },
+    {
+      "epoch": 0.1810781792304427,
+      "grad_norm": 0.10025262087583542,
+      "learning_rate": 2.4629152169015494e-05,
+      "loss": 9.8401,
+      "step": 36260
+    },
+    {
+      "epoch": 0.18112811805538215,
+      "grad_norm": 0.0945044681429863,
+      "learning_rate": 2.4627650254073944e-05,
+      "loss": 9.8424,
+      "step": 36270
+    },
+    {
+      "epoch": 0.1811780568803216,
+      "grad_norm": 0.0924852043390274,
+      "learning_rate": 2.4626148339132394e-05,
+      "loss": 9.8448,
+      "step": 36280
+    },
+    {
+      "epoch": 0.18122799570526105,
+      "grad_norm": 0.09346658736467361,
+      "learning_rate": 2.4624646424190845e-05,
+      "loss": 9.8408,
+      "step": 36290
+    },
+    {
+      "epoch": 0.1812779345302005,
+      "grad_norm": 0.10162097215652466,
+      "learning_rate": 2.4623144509249295e-05,
+      "loss": 9.8382,
+      "step": 36300
+    },
+    {
+      "epoch": 0.18132787335513995,
+      "grad_norm": 0.093153215944767,
+      "learning_rate": 2.462164259430774e-05,
+      "loss": 9.8415,
+      "step": 36310
+    },
+    {
+      "epoch": 0.1813778121800794,
+      "grad_norm": 0.09772991389036179,
+      "learning_rate": 2.462014067936619e-05,
+      "loss": 9.8463,
+      "step": 36320
+    },
+    {
+      "epoch": 0.18142775100501884,
+      "grad_norm": 0.09410078823566437,
+      "learning_rate": 2.4618638764424642e-05,
+      "loss": 9.8374,
+      "step": 36330
+    },
+    {
+      "epoch": 0.1814776898299583,
+      "grad_norm": 0.09498627483844757,
+      "learning_rate": 2.4617136849483092e-05,
+      "loss": 9.8382,
+      "step": 36340
+    },
+    {
+      "epoch": 0.18152762865489774,
+      "grad_norm": 0.09718122333288193,
+      "learning_rate": 2.4615634934541542e-05,
+      "loss": 9.839,
+      "step": 36350
+    },
+    {
+      "epoch": 0.1815775674798372,
+      "grad_norm": 0.09762446582317352,
+      "learning_rate": 2.461413301959999e-05,
+      "loss": 9.8365,
+      "step": 36360
+    },
+    {
+      "epoch": 0.18162750630477664,
+      "grad_norm": 0.09202492982149124,
+      "learning_rate": 2.461263110465844e-05,
+      "loss": 9.8351,
+      "step": 36370
+    },
+    {
+      "epoch": 0.1816774451297161,
+      "grad_norm": 0.09190283715724945,
+      "learning_rate": 2.461112918971689e-05,
+      "loss": 9.8375,
+      "step": 36380
+    },
+    {
+      "epoch": 0.18172738395465554,
+      "grad_norm": 0.09559813141822815,
+      "learning_rate": 2.460962727477534e-05,
+      "loss": 9.8424,
+      "step": 36390
+    },
+    {
+      "epoch": 0.181777322779595,
+      "grad_norm": 0.09660519659519196,
+      "learning_rate": 2.460812535983379e-05,
+      "loss": 9.8377,
+      "step": 36400
+    },
+    {
+      "epoch": 0.18182726160453444,
+      "grad_norm": 0.0976775512099266,
+      "learning_rate": 2.4606623444892236e-05,
+      "loss": 9.841,
+      "step": 36410
+    },
+    {
+      "epoch": 0.1818772004294739,
+      "grad_norm": 0.10077891498804092,
+      "learning_rate": 2.4605121529950687e-05,
+      "loss": 9.8422,
+      "step": 36420
+    },
+    {
+      "epoch": 0.18192713925441334,
+      "grad_norm": 0.09487045556306839,
+      "learning_rate": 2.4603619615009137e-05,
+      "loss": 9.8388,
+      "step": 36430
+    },
+    {
+      "epoch": 0.1819770780793528,
+      "grad_norm": 0.08921613544225693,
+      "learning_rate": 2.4602117700067587e-05,
+      "loss": 9.8387,
+      "step": 36440
+    },
+    {
+      "epoch": 0.18202701690429224,
+      "grad_norm": 0.09209630638360977,
+      "learning_rate": 2.4600615785126037e-05,
+      "loss": 9.8364,
+      "step": 36450
+    },
+    {
+      "epoch": 0.1820769557292317,
+      "grad_norm": 0.09444855153560638,
+      "learning_rate": 2.4599113870184484e-05,
+      "loss": 9.8348,
+      "step": 36460
+    },
+    {
+      "epoch": 0.18212689455417114,
+      "grad_norm": 0.10311925411224365,
+      "learning_rate": 2.4597611955242934e-05,
+      "loss": 9.8378,
+      "step": 36470
+    },
+    {
+      "epoch": 0.1821768333791106,
+      "grad_norm": 0.09053095430135727,
+      "learning_rate": 2.4596110040301384e-05,
+      "loss": 9.8406,
+      "step": 36480
+    },
+    {
+      "epoch": 0.18222677220405004,
+      "grad_norm": 0.09320933371782303,
+      "learning_rate": 2.4594608125359835e-05,
+      "loss": 9.8384,
+      "step": 36490
+    },
+    {
+      "epoch": 0.18227671102898949,
+      "grad_norm": 0.09924657642841339,
+      "learning_rate": 2.4593106210418285e-05,
+      "loss": 9.8366,
+      "step": 36500
+    },
+    {
+      "epoch": 0.18232664985392893,
+      "grad_norm": 0.09330344945192337,
+      "learning_rate": 2.4591604295476735e-05,
+      "loss": 9.829,
+      "step": 36510
+    },
+    {
+      "epoch": 0.18237658867886838,
+      "grad_norm": 0.0905805230140686,
+      "learning_rate": 2.459010238053518e-05,
+      "loss": 9.8449,
+      "step": 36520
+    },
+    {
+      "epoch": 0.18242652750380783,
+      "grad_norm": 0.09339253604412079,
+      "learning_rate": 2.4588600465593632e-05,
+      "loss": 9.8361,
+      "step": 36530
+    },
+    {
+      "epoch": 0.18247646632874728,
+      "grad_norm": 0.09230770170688629,
+      "learning_rate": 2.4587098550652082e-05,
+      "loss": 9.8377,
+      "step": 36540
+    },
+    {
+      "epoch": 0.18252640515368673,
+      "grad_norm": 0.09087874740362167,
+      "learning_rate": 2.4585596635710532e-05,
+      "loss": 9.8389,
+      "step": 36550
+    },
+    {
+      "epoch": 0.18257634397862618,
+      "grad_norm": 0.09459523111581802,
+      "learning_rate": 2.4584094720768982e-05,
+      "loss": 9.8339,
+      "step": 36560
+    },
+    {
+      "epoch": 0.18262628280356563,
+      "grad_norm": 0.09835857897996902,
+      "learning_rate": 2.458259280582743e-05,
+      "loss": 9.8379,
+      "step": 36570
+    },
+    {
+      "epoch": 0.18267622162850508,
+      "grad_norm": 0.09462913870811462,
+      "learning_rate": 2.458109089088588e-05,
+      "loss": 9.8366,
+      "step": 36580
+    },
+    {
+      "epoch": 0.18272616045344453,
+      "grad_norm": 0.09966031461954117,
+      "learning_rate": 2.457958897594433e-05,
+      "loss": 9.8301,
+      "step": 36590
+    },
+    {
+      "epoch": 0.18277609927838398,
+      "grad_norm": 0.09340931475162506,
+      "learning_rate": 2.457808706100278e-05,
+      "loss": 9.8389,
+      "step": 36600
+    },
+    {
+      "epoch": 0.18282603810332343,
+      "grad_norm": 0.09110315144062042,
+      "learning_rate": 2.457658514606123e-05,
+      "loss": 9.8342,
+      "step": 36610
+    },
+    {
+      "epoch": 0.18287597692826288,
+      "grad_norm": 0.09554065018892288,
+      "learning_rate": 2.4575083231119677e-05,
+      "loss": 9.8408,
+      "step": 36620
+    },
+    {
+      "epoch": 0.18292591575320233,
+      "grad_norm": 0.09898494929075241,
+      "learning_rate": 2.4573581316178127e-05,
+      "loss": 9.8364,
+      "step": 36630
+    },
+    {
+      "epoch": 0.18297585457814178,
+      "grad_norm": 0.09261037409305573,
+      "learning_rate": 2.4572079401236577e-05,
+      "loss": 9.8335,
+      "step": 36640
+    },
+    {
+      "epoch": 0.18302579340308123,
+      "grad_norm": 0.09171412140130997,
+      "learning_rate": 2.4570577486295027e-05,
+      "loss": 9.8286,
+      "step": 36650
+    },
+    {
+      "epoch": 0.18307573222802068,
+      "grad_norm": 0.09472862631082535,
+      "learning_rate": 2.4569075571353477e-05,
+      "loss": 9.8358,
+      "step": 36660
+    },
+    {
+      "epoch": 0.18312567105296013,
+      "grad_norm": 0.09369245171546936,
+      "learning_rate": 2.4567573656411924e-05,
+      "loss": 9.8337,
+      "step": 36670
+    },
+    {
+      "epoch": 0.18317560987789958,
+      "grad_norm": 0.09198103100061417,
+      "learning_rate": 2.4566071741470374e-05,
+      "loss": 9.8384,
+      "step": 36680
+    },
+    {
+      "epoch": 0.18322554870283903,
+      "grad_norm": 0.09761819988489151,
+      "learning_rate": 2.4564569826528825e-05,
+      "loss": 9.833,
+      "step": 36690
+    },
+    {
+      "epoch": 0.18327548752777847,
+      "grad_norm": 0.09576039016246796,
+      "learning_rate": 2.4563067911587275e-05,
+      "loss": 9.8346,
+      "step": 36700
+    },
+    {
+      "epoch": 0.18332542635271792,
+      "grad_norm": 0.10029350966215134,
+      "learning_rate": 2.4561565996645725e-05,
+      "loss": 9.8383,
+      "step": 36710
+    },
+    {
+      "epoch": 0.18337536517765737,
+      "grad_norm": 0.09115588665008545,
+      "learning_rate": 2.456006408170417e-05,
+      "loss": 9.8311,
+      "step": 36720
+    },
+    {
+      "epoch": 0.18342530400259682,
+      "grad_norm": 0.09495572000741959,
+      "learning_rate": 2.4558562166762622e-05,
+      "loss": 9.8296,
+      "step": 36730
+    },
+    {
+      "epoch": 0.18347524282753627,
+      "grad_norm": 0.0938972532749176,
+      "learning_rate": 2.4557060251821072e-05,
+      "loss": 9.8303,
+      "step": 36740
+    },
+    {
+      "epoch": 0.18352518165247572,
+      "grad_norm": 0.0916903167963028,
+      "learning_rate": 2.4555558336879522e-05,
+      "loss": 9.8256,
+      "step": 36750
+    },
+    {
+      "epoch": 0.18357512047741517,
+      "grad_norm": 0.09134495258331299,
+      "learning_rate": 2.4554056421937972e-05,
+      "loss": 9.8269,
+      "step": 36760
+    },
+    {
+      "epoch": 0.18362505930235462,
+      "grad_norm": 0.09667996317148209,
+      "learning_rate": 2.455255450699642e-05,
+      "loss": 9.8308,
+      "step": 36770
+    },
+    {
+      "epoch": 0.18367499812729407,
+      "grad_norm": 0.09833179414272308,
+      "learning_rate": 2.455105259205487e-05,
+      "loss": 9.8352,
+      "step": 36780
+    },
+    {
+      "epoch": 0.18372493695223352,
+      "grad_norm": 0.09507912397384644,
+      "learning_rate": 2.454955067711332e-05,
+      "loss": 9.8326,
+      "step": 36790
+    },
+    {
+      "epoch": 0.18377487577717297,
+      "grad_norm": 0.09790566563606262,
+      "learning_rate": 2.454804876217177e-05,
+      "loss": 9.831,
+      "step": 36800
+    },
+    {
+      "epoch": 0.18382481460211242,
+      "grad_norm": 0.1005285233259201,
+      "learning_rate": 2.454654684723022e-05,
+      "loss": 9.8316,
+      "step": 36810
+    },
+    {
+      "epoch": 0.18387475342705187,
+      "grad_norm": 0.09769100695848465,
+      "learning_rate": 2.4545044932288667e-05,
+      "loss": 9.8273,
+      "step": 36820
+    },
+    {
+      "epoch": 0.18392469225199132,
+      "grad_norm": 0.09412973374128342,
+      "learning_rate": 2.454354301734712e-05,
+      "loss": 9.826,
+      "step": 36830
+    },
+    {
+      "epoch": 0.18397463107693077,
+      "grad_norm": 0.09262508153915405,
+      "learning_rate": 2.4542041102405567e-05,
+      "loss": 9.8291,
+      "step": 36840
+    },
+    {
+      "epoch": 0.18402456990187022,
+      "grad_norm": 0.09512876719236374,
+      "learning_rate": 2.4540539187464017e-05,
+      "loss": 9.8278,
+      "step": 36850
+    },
+    {
+      "epoch": 0.18407450872680967,
+      "grad_norm": 0.0943412259221077,
+      "learning_rate": 2.4539037272522467e-05,
+      "loss": 9.8363,
+      "step": 36860
+    },
+    {
+      "epoch": 0.18412444755174912,
+      "grad_norm": 0.09464114904403687,
+      "learning_rate": 2.4537535357580914e-05,
+      "loss": 9.8294,
+      "step": 36870
+    },
+    {
+      "epoch": 0.18417438637668856,
+      "grad_norm": 0.09377477318048477,
+      "learning_rate": 2.4536033442639368e-05,
+      "loss": 9.8332,
+      "step": 36880
+    },
+    {
+      "epoch": 0.18422432520162801,
+      "grad_norm": 0.09622970223426819,
+      "learning_rate": 2.4534531527697815e-05,
+      "loss": 9.8243,
+      "step": 36890
+    },
+    {
+      "epoch": 0.18427426402656746,
+      "grad_norm": 0.09202242642641068,
+      "learning_rate": 2.4533029612756265e-05,
+      "loss": 9.8243,
+      "step": 36900
+    },
+    {
+      "epoch": 0.1843242028515069,
+      "grad_norm": 0.09235205501317978,
+      "learning_rate": 2.4531527697814715e-05,
+      "loss": 9.8343,
+      "step": 36910
+    },
+    {
+      "epoch": 0.18437414167644636,
+      "grad_norm": 0.09455617517232895,
+      "learning_rate": 2.4530025782873162e-05,
+      "loss": 9.8255,
+      "step": 36920
+    },
+    {
+      "epoch": 0.1844240805013858,
+      "grad_norm": 0.09160789847373962,
+      "learning_rate": 2.4528523867931615e-05,
+      "loss": 9.8267,
+      "step": 36930
+    },
+    {
+      "epoch": 0.18447401932632526,
+      "grad_norm": 0.09022436290979385,
+      "learning_rate": 2.4527021952990062e-05,
+      "loss": 9.8328,
+      "step": 36940
+    },
+    {
+      "epoch": 0.1845239581512647,
+      "grad_norm": 0.09537603706121445,
+      "learning_rate": 2.4525520038048512e-05,
+      "loss": 9.8258,
+      "step": 36950
+    },
+    {
+      "epoch": 0.18457389697620416,
+      "grad_norm": 0.09323366731405258,
+      "learning_rate": 2.4524018123106962e-05,
+      "loss": 9.8279,
+      "step": 36960
+    },
+    {
+      "epoch": 0.1846238358011436,
+      "grad_norm": 0.0941099300980568,
+      "learning_rate": 2.452251620816541e-05,
+      "loss": 9.8292,
+      "step": 36970
+    },
+    {
+      "epoch": 0.18467377462608306,
+      "grad_norm": 0.0980052724480629,
+      "learning_rate": 2.4521014293223863e-05,
+      "loss": 9.8218,
+      "step": 36980
+    },
+    {
+      "epoch": 0.1847237134510225,
+      "grad_norm": 0.09246113896369934,
+      "learning_rate": 2.451951237828231e-05,
+      "loss": 9.8246,
+      "step": 36990
+    },
+    {
+      "epoch": 0.18477365227596196,
+      "grad_norm": 0.09564212709665298,
+      "learning_rate": 2.451801046334076e-05,
+      "loss": 9.8209,
+      "step": 37000
+    },
+    {
+      "epoch": 0.1848235911009014,
+      "grad_norm": 0.09970725327730179,
+      "learning_rate": 2.451650854839921e-05,
+      "loss": 9.8251,
+      "step": 37010
+    },
+    {
+      "epoch": 0.18487352992584086,
+      "grad_norm": 0.09713415801525116,
+      "learning_rate": 2.4515006633457657e-05,
+      "loss": 9.8269,
+      "step": 37020
+    },
+    {
+      "epoch": 0.1849234687507803,
+      "grad_norm": 0.09196287393569946,
+      "learning_rate": 2.451350471851611e-05,
+      "loss": 9.827,
+      "step": 37030
+    },
+    {
+      "epoch": 0.18497340757571976,
+      "grad_norm": 0.09455425292253494,
+      "learning_rate": 2.4512002803574557e-05,
+      "loss": 9.8193,
+      "step": 37040
+    },
+    {
+      "epoch": 0.1850233464006592,
+      "grad_norm": 0.09092764556407928,
+      "learning_rate": 2.4510500888633007e-05,
+      "loss": 9.8283,
+      "step": 37050
+    },
+    {
+      "epoch": 0.18507328522559865,
+      "grad_norm": 0.0928642749786377,
+      "learning_rate": 2.4508998973691457e-05,
+      "loss": 9.8273,
+      "step": 37060
+    },
+    {
+      "epoch": 0.1851232240505381,
+      "grad_norm": 0.09685216099023819,
+      "learning_rate": 2.4507497058749904e-05,
+      "loss": 9.8225,
+      "step": 37070
+    },
+    {
+      "epoch": 0.18517316287547755,
+      "grad_norm": 0.0929524376988411,
+      "learning_rate": 2.4505995143808358e-05,
+      "loss": 9.8339,
+      "step": 37080
+    },
+    {
+      "epoch": 0.18522310170041698,
+      "grad_norm": 0.09492181241512299,
+      "learning_rate": 2.4504493228866805e-05,
+      "loss": 9.8241,
+      "step": 37090
+    },
+    {
+      "epoch": 0.18527304052535642,
+      "grad_norm": 0.09493514150381088,
+      "learning_rate": 2.4502991313925255e-05,
+      "loss": 9.8202,
+      "step": 37100
+    },
+    {
+      "epoch": 0.18532297935029587,
+      "grad_norm": 0.0974758043885231,
+      "learning_rate": 2.4501489398983705e-05,
+      "loss": 9.8215,
+      "step": 37110
+    },
+    {
+      "epoch": 0.18537291817523532,
+      "grad_norm": 0.09460314363241196,
+      "learning_rate": 2.4499987484042152e-05,
+      "loss": 9.8209,
+      "step": 37120
+    },
+    {
+      "epoch": 0.18542285700017477,
+      "grad_norm": 0.10088241845369339,
+      "learning_rate": 2.4498485569100605e-05,
+      "loss": 9.8161,
+      "step": 37130
+    },
+    {
+      "epoch": 0.18547279582511422,
+      "grad_norm": 0.09594068676233292,
+      "learning_rate": 2.4496983654159052e-05,
+      "loss": 9.8158,
+      "step": 37140
+    },
+    {
+      "epoch": 0.18552273465005367,
+      "grad_norm": 0.09308291226625443,
+      "learning_rate": 2.4495481739217506e-05,
+      "loss": 9.8229,
+      "step": 37150
+    },
+    {
+      "epoch": 0.18557267347499312,
+      "grad_norm": 0.09384151548147202,
+      "learning_rate": 2.4493979824275952e-05,
+      "loss": 9.8258,
+      "step": 37160
+    },
+    {
+      "epoch": 0.18562261229993257,
+      "grad_norm": 0.09089542925357819,
+      "learning_rate": 2.44924779093344e-05,
+      "loss": 9.8214,
+      "step": 37170
+    },
+    {
+      "epoch": 0.18567255112487202,
+      "grad_norm": 0.10226479917764664,
+      "learning_rate": 2.4490975994392853e-05,
+      "loss": 9.8194,
+      "step": 37180
+    },
+    {
+      "epoch": 0.18572248994981147,
+      "grad_norm": 0.09530473500490189,
+      "learning_rate": 2.44894740794513e-05,
+      "loss": 9.8214,
+      "step": 37190
+    },
+    {
+      "epoch": 0.18577242877475092,
+      "grad_norm": 0.09461791068315506,
+      "learning_rate": 2.4487972164509753e-05,
+      "loss": 9.8234,
+      "step": 37200
+    },
+    {
+      "epoch": 0.18582236759969037,
+      "grad_norm": 0.09858336299657822,
+      "learning_rate": 2.44864702495682e-05,
+      "loss": 9.8273,
+      "step": 37210
+    },
+    {
+      "epoch": 0.18587230642462982,
+      "grad_norm": 0.0900527834892273,
+      "learning_rate": 2.4484968334626647e-05,
+      "loss": 9.8225,
+      "step": 37220
+    },
+    {
+      "epoch": 0.18592224524956927,
+      "grad_norm": 0.09260288625955582,
+      "learning_rate": 2.44834664196851e-05,
+      "loss": 9.8215,
+      "step": 37230
+    },
+    {
+      "epoch": 0.18597218407450872,
+      "grad_norm": 0.0938689112663269,
+      "learning_rate": 2.4481964504743547e-05,
+      "loss": 9.8189,
+      "step": 37240
+    },
+    {
+      "epoch": 0.18602212289944817,
+      "grad_norm": 0.09121273458003998,
+      "learning_rate": 2.4480462589802e-05,
+      "loss": 9.819,
+      "step": 37250
+    },
+    {
+      "epoch": 0.18607206172438762,
+      "grad_norm": 0.09315800666809082,
+      "learning_rate": 2.4478960674860447e-05,
+      "loss": 9.8215,
+      "step": 37260
+    },
+    {
+      "epoch": 0.18612200054932707,
+      "grad_norm": 0.09155386686325073,
+      "learning_rate": 2.4477458759918898e-05,
+      "loss": 9.8212,
+      "step": 37270
+    },
+    {
+      "epoch": 0.18617193937426652,
+      "grad_norm": 0.09750566631555557,
+      "learning_rate": 2.4475956844977348e-05,
+      "loss": 9.8185,
+      "step": 37280
+    },
+    {
+      "epoch": 0.18622187819920596,
+      "grad_norm": 0.09988453984260559,
+      "learning_rate": 2.4474454930035795e-05,
+      "loss": 9.8174,
+      "step": 37290
+    },
+    {
+      "epoch": 0.18627181702414541,
+      "grad_norm": 0.09393157064914703,
+      "learning_rate": 2.4472953015094248e-05,
+      "loss": 9.8232,
+      "step": 37300
+    },
+    {
+      "epoch": 0.18632175584908486,
+      "grad_norm": 0.09441966563463211,
+      "learning_rate": 2.4471451100152695e-05,
+      "loss": 9.8228,
+      "step": 37310
+    },
+    {
+      "epoch": 0.1863716946740243,
+      "grad_norm": 0.091115802526474,
+      "learning_rate": 2.4469949185211145e-05,
+      "loss": 9.8182,
+      "step": 37320
+    },
+    {
+      "epoch": 0.18642163349896376,
+      "grad_norm": 0.09511063247919083,
+      "learning_rate": 2.4468447270269595e-05,
+      "loss": 9.8202,
+      "step": 37330
+    },
+    {
+      "epoch": 0.1864715723239032,
+      "grad_norm": 0.10219678282737732,
+      "learning_rate": 2.4466945355328042e-05,
+      "loss": 9.8191,
+      "step": 37340
+    },
+    {
+      "epoch": 0.18652151114884266,
+      "grad_norm": 0.09869439899921417,
+      "learning_rate": 2.4465443440386496e-05,
+      "loss": 9.8132,
+      "step": 37350
+    },
+    {
+      "epoch": 0.1865714499737821,
+      "grad_norm": 0.09464725852012634,
+      "learning_rate": 2.4463941525444942e-05,
+      "loss": 9.8219,
+      "step": 37360
+    },
+    {
+      "epoch": 0.18662138879872156,
+      "grad_norm": 0.08857977390289307,
+      "learning_rate": 2.4462439610503393e-05,
+      "loss": 9.8222,
+      "step": 37370
+    },
+    {
+      "epoch": 0.186671327623661,
+      "grad_norm": 0.09524288773536682,
+      "learning_rate": 2.4460937695561843e-05,
+      "loss": 9.8197,
+      "step": 37380
+    },
+    {
+      "epoch": 0.18672126644860046,
+      "grad_norm": 0.09553853422403336,
+      "learning_rate": 2.445943578062029e-05,
+      "loss": 9.8198,
+      "step": 37390
+    },
+    {
+      "epoch": 0.1867712052735399,
+      "grad_norm": 0.10078942030668259,
+      "learning_rate": 2.4457933865678743e-05,
+      "loss": 9.8121,
+      "step": 37400
+    },
+    {
+      "epoch": 0.18682114409847936,
+      "grad_norm": 0.09711633622646332,
+      "learning_rate": 2.445643195073719e-05,
+      "loss": 9.8176,
+      "step": 37410
+    },
+    {
+      "epoch": 0.1868710829234188,
+      "grad_norm": 0.0985088124871254,
+      "learning_rate": 2.445493003579564e-05,
+      "loss": 9.8125,
+      "step": 37420
+    },
+    {
+      "epoch": 0.18692102174835826,
+      "grad_norm": 0.10270990431308746,
+      "learning_rate": 2.445342812085409e-05,
+      "loss": 9.824,
+      "step": 37430
+    },
+    {
+      "epoch": 0.1869709605732977,
+      "grad_norm": 0.09112194925546646,
+      "learning_rate": 2.4451926205912537e-05,
+      "loss": 9.8115,
+      "step": 37440
+    },
+    {
+      "epoch": 0.18702089939823716,
+      "grad_norm": 0.09221579879522324,
+      "learning_rate": 2.445042429097099e-05,
+      "loss": 9.8166,
+      "step": 37450
+    },
+    {
+      "epoch": 0.1870708382231766,
+      "grad_norm": 0.09492717683315277,
+      "learning_rate": 2.4448922376029437e-05,
+      "loss": 9.816,
+      "step": 37460
+    },
+    {
+      "epoch": 0.18712077704811605,
+      "grad_norm": 0.09594427794218063,
+      "learning_rate": 2.4447420461087888e-05,
+      "loss": 9.8106,
+      "step": 37470
+    },
+    {
+      "epoch": 0.1871707158730555,
+      "grad_norm": 0.10151471197605133,
+      "learning_rate": 2.4445918546146338e-05,
+      "loss": 9.8237,
+      "step": 37480
+    },
+    {
+      "epoch": 0.18722065469799495,
+      "grad_norm": 0.09668134897947311,
+      "learning_rate": 2.4444416631204785e-05,
+      "loss": 9.8231,
+      "step": 37490
+    },
+    {
+      "epoch": 0.1872705935229344,
+      "grad_norm": 0.09197778254747391,
+      "learning_rate": 2.4442914716263238e-05,
+      "loss": 9.817,
+      "step": 37500
+    },
+    {
+      "epoch": 0.18732053234787385,
+      "grad_norm": 0.09514784067869186,
+      "learning_rate": 2.4441412801321685e-05,
+      "loss": 9.817,
+      "step": 37510
+    },
+    {
+      "epoch": 0.1873704711728133,
+      "grad_norm": 0.09452585130929947,
+      "learning_rate": 2.4439910886380135e-05,
+      "loss": 9.8115,
+      "step": 37520
+    },
+    {
+      "epoch": 0.18742040999775275,
+      "grad_norm": 0.09140285104513168,
+      "learning_rate": 2.4438408971438585e-05,
+      "loss": 9.8115,
+      "step": 37530
+    },
+    {
+      "epoch": 0.1874703488226922,
+      "grad_norm": 0.10165819525718689,
+      "learning_rate": 2.4436907056497032e-05,
+      "loss": 9.8183,
+      "step": 37540
+    },
+    {
+      "epoch": 0.18752028764763165,
+      "grad_norm": 0.09387430548667908,
+      "learning_rate": 2.4435405141555486e-05,
+      "loss": 9.8187,
+      "step": 37550
+    },
+    {
+      "epoch": 0.1875702264725711,
+      "grad_norm": 0.09394615888595581,
+      "learning_rate": 2.4433903226613932e-05,
+      "loss": 9.8146,
+      "step": 37560
+    },
+    {
+      "epoch": 0.18762016529751055,
+      "grad_norm": 0.09397023171186447,
+      "learning_rate": 2.4432401311672386e-05,
+      "loss": 9.8182,
+      "step": 37570
+    },
+    {
+      "epoch": 0.18767010412245,
+      "grad_norm": 0.09779095649719238,
+      "learning_rate": 2.4430899396730833e-05,
+      "loss": 9.81,
+      "step": 37580
+    },
+    {
+      "epoch": 0.18772004294738945,
+      "grad_norm": 0.09826376289129257,
+      "learning_rate": 2.442939748178928e-05,
+      "loss": 9.8129,
+      "step": 37590
+    },
+    {
+      "epoch": 0.1877699817723289,
+      "grad_norm": 0.09404350817203522,
+      "learning_rate": 2.4427895566847733e-05,
+      "loss": 9.813,
+      "step": 37600
+    },
+    {
+      "epoch": 0.18781992059726835,
+      "grad_norm": 0.09514794498682022,
+      "learning_rate": 2.442639365190618e-05,
+      "loss": 9.8125,
+      "step": 37610
+    },
+    {
+      "epoch": 0.1878698594222078,
+      "grad_norm": 0.09750908613204956,
+      "learning_rate": 2.4424891736964634e-05,
+      "loss": 9.8146,
+      "step": 37620
+    },
+    {
+      "epoch": 0.18791979824714725,
+      "grad_norm": 0.09314355999231339,
+      "learning_rate": 2.442338982202308e-05,
+      "loss": 9.8149,
+      "step": 37630
+    },
+    {
+      "epoch": 0.1879697370720867,
+      "grad_norm": 0.08955375105142593,
+      "learning_rate": 2.4421887907081527e-05,
+      "loss": 9.8092,
+      "step": 37640
+    },
+    {
+      "epoch": 0.18801967589702615,
+      "grad_norm": 0.09085706621408463,
+      "learning_rate": 2.442038599213998e-05,
+      "loss": 9.8129,
+      "step": 37650
+    },
+    {
+      "epoch": 0.1880696147219656,
+      "grad_norm": 0.09261947870254517,
+      "learning_rate": 2.4418884077198427e-05,
+      "loss": 9.8183,
+      "step": 37660
+    },
+    {
+      "epoch": 0.18811955354690504,
+      "grad_norm": 0.09396150708198547,
+      "learning_rate": 2.441738216225688e-05,
+      "loss": 9.8103,
+      "step": 37670
+    },
+    {
+      "epoch": 0.1881694923718445,
+      "grad_norm": 0.10040120780467987,
+      "learning_rate": 2.4415880247315328e-05,
+      "loss": 9.815,
+      "step": 37680
+    },
+    {
+      "epoch": 0.18821943119678394,
+      "grad_norm": 0.09295698255300522,
+      "learning_rate": 2.4414378332373775e-05,
+      "loss": 9.8175,
+      "step": 37690
+    },
+    {
+      "epoch": 0.1882693700217234,
+      "grad_norm": 0.09030512720346451,
+      "learning_rate": 2.4412876417432228e-05,
+      "loss": 9.8094,
+      "step": 37700
+    },
+    {
+      "epoch": 0.18831930884666284,
+      "grad_norm": 0.08929255604743958,
+      "learning_rate": 2.4411374502490675e-05,
+      "loss": 9.8134,
+      "step": 37710
+    },
+    {
+      "epoch": 0.1883692476716023,
+      "grad_norm": 0.09024254232645035,
+      "learning_rate": 2.440987258754913e-05,
+      "loss": 9.816,
+      "step": 37720
+    },
+    {
+      "epoch": 0.18841918649654174,
+      "grad_norm": 0.09843594580888748,
+      "learning_rate": 2.4408370672607575e-05,
+      "loss": 9.8173,
+      "step": 37730
+    },
+    {
+      "epoch": 0.1884691253214812,
+      "grad_norm": 0.09375924617052078,
+      "learning_rate": 2.4406868757666022e-05,
+      "loss": 9.8139,
+      "step": 37740
+    },
+    {
+      "epoch": 0.18851906414642064,
+      "grad_norm": 0.09652825444936752,
+      "learning_rate": 2.4405366842724476e-05,
+      "loss": 9.8105,
+      "step": 37750
+    },
+    {
+      "epoch": 0.1885690029713601,
+      "grad_norm": 0.09221908450126648,
+      "learning_rate": 2.4403864927782922e-05,
+      "loss": 9.8123,
+      "step": 37760
+    },
+    {
+      "epoch": 0.18861894179629954,
+      "grad_norm": 0.0924220085144043,
+      "learning_rate": 2.4402363012841376e-05,
+      "loss": 9.8106,
+      "step": 37770
+    },
+    {
+      "epoch": 0.188668880621239,
+      "grad_norm": 0.09092016518115997,
+      "learning_rate": 2.4400861097899823e-05,
+      "loss": 9.8147,
+      "step": 37780
+    },
+    {
+      "epoch": 0.18871881944617844,
+      "grad_norm": 0.09375354647636414,
+      "learning_rate": 2.4399359182958273e-05,
+      "loss": 9.8052,
+      "step": 37790
+    },
+    {
+      "epoch": 0.1887687582711179,
+      "grad_norm": 0.09391508996486664,
+      "learning_rate": 2.4397857268016723e-05,
+      "loss": 9.8157,
+      "step": 37800
+    },
+    {
+      "epoch": 0.18881869709605734,
+      "grad_norm": 0.0947243720293045,
+      "learning_rate": 2.439635535307517e-05,
+      "loss": 9.808,
+      "step": 37810
+    },
+    {
+      "epoch": 0.18886863592099679,
+      "grad_norm": 0.09377534687519073,
+      "learning_rate": 2.4394853438133624e-05,
+      "loss": 9.8004,
+      "step": 37820
+    },
+    {
+      "epoch": 0.18891857474593624,
+      "grad_norm": 0.09614383429288864,
+      "learning_rate": 2.439335152319207e-05,
+      "loss": 9.8076,
+      "step": 37830
+    },
+    {
+      "epoch": 0.18896851357087568,
+      "grad_norm": 0.0933867022395134,
+      "learning_rate": 2.439184960825052e-05,
+      "loss": 9.8132,
+      "step": 37840
+    },
+    {
+      "epoch": 0.18901845239581513,
+      "grad_norm": 0.0957120954990387,
+      "learning_rate": 2.439034769330897e-05,
+      "loss": 9.8068,
+      "step": 37850
+    },
+    {
+      "epoch": 0.18906839122075458,
+      "grad_norm": 0.09493666887283325,
+      "learning_rate": 2.4388845778367417e-05,
+      "loss": 9.8015,
+      "step": 37860
+    },
+    {
+      "epoch": 0.18911833004569403,
+      "grad_norm": 0.09380702674388885,
+      "learning_rate": 2.438734386342587e-05,
+      "loss": 9.8049,
+      "step": 37870
+    },
+    {
+      "epoch": 0.18916826887063348,
+      "grad_norm": 0.09432664513587952,
+      "learning_rate": 2.4385841948484318e-05,
+      "loss": 9.8008,
+      "step": 37880
+    },
+    {
+      "epoch": 0.18921820769557293,
+      "grad_norm": 0.09340386092662811,
+      "learning_rate": 2.4384340033542768e-05,
+      "loss": 9.8072,
+      "step": 37890
+    },
+    {
+      "epoch": 0.18926814652051238,
+      "grad_norm": 0.09343277662992477,
+      "learning_rate": 2.4382838118601218e-05,
+      "loss": 9.8092,
+      "step": 37900
+    },
+    {
+      "epoch": 0.18931808534545183,
+      "grad_norm": 0.09355462342500687,
+      "learning_rate": 2.4381336203659665e-05,
+      "loss": 9.8062,
+      "step": 37910
+    },
+    {
+      "epoch": 0.18936802417039128,
+      "grad_norm": 0.09270920604467392,
+      "learning_rate": 2.437983428871812e-05,
+      "loss": 9.8113,
+      "step": 37920
+    },
+    {
+      "epoch": 0.18941796299533073,
+      "grad_norm": 0.09088815748691559,
+      "learning_rate": 2.4378332373776565e-05,
+      "loss": 9.8089,
+      "step": 37930
+    },
+    {
+      "epoch": 0.18946790182027018,
+      "grad_norm": 0.09358256310224533,
+      "learning_rate": 2.4376830458835016e-05,
+      "loss": 9.8005,
+      "step": 37940
+    },
+    {
+      "epoch": 0.18951784064520963,
+      "grad_norm": 0.09258565306663513,
+      "learning_rate": 2.4375328543893466e-05,
+      "loss": 9.8029,
+      "step": 37950
+    },
+    {
+      "epoch": 0.18956777947014908,
+      "grad_norm": 0.09664854407310486,
+      "learning_rate": 2.4373826628951912e-05,
+      "loss": 9.8015,
+      "step": 37960
+    },
+    {
+      "epoch": 0.18961771829508853,
+      "grad_norm": 0.09192045032978058,
+      "learning_rate": 2.4372324714010366e-05,
+      "loss": 9.8067,
+      "step": 37970
+    },
+    {
+      "epoch": 0.18966765712002798,
+      "grad_norm": 0.09485509991645813,
+      "learning_rate": 2.4370822799068813e-05,
+      "loss": 9.8016,
+      "step": 37980
+    },
+    {
+      "epoch": 0.18971759594496743,
+      "grad_norm": 0.09736743569374084,
+      "learning_rate": 2.4369320884127263e-05,
+      "loss": 9.8028,
+      "step": 37990
+    },
+    {
+      "epoch": 0.18976753476990688,
+      "grad_norm": 0.09561127424240112,
+      "learning_rate": 2.4367818969185713e-05,
+      "loss": 9.803,
+      "step": 38000
+    },
+    {
+      "epoch": 0.18981747359484633,
+      "grad_norm": 0.0922846645116806,
+      "learning_rate": 2.436631705424416e-05,
+      "loss": 9.8083,
+      "step": 38010
+    },
+    {
+      "epoch": 0.18986741241978577,
+      "grad_norm": 0.09271165728569031,
+      "learning_rate": 2.4364815139302614e-05,
+      "loss": 9.8141,
+      "step": 38020
+    },
+    {
+      "epoch": 0.18991735124472522,
+      "grad_norm": 0.09106654673814774,
+      "learning_rate": 2.436331322436106e-05,
+      "loss": 9.8048,
+      "step": 38030
+    },
+    {
+      "epoch": 0.18996729006966467,
+      "grad_norm": 0.09738942235708237,
+      "learning_rate": 2.436181130941951e-05,
+      "loss": 9.8077,
+      "step": 38040
+    },
+    {
+      "epoch": 0.19001722889460412,
+      "grad_norm": 0.09800420701503754,
+      "learning_rate": 2.436030939447796e-05,
+      "loss": 9.8041,
+      "step": 38050
+    },
+    {
+      "epoch": 0.19006716771954357,
+      "grad_norm": 0.09482547640800476,
+      "learning_rate": 2.4358807479536407e-05,
+      "loss": 9.804,
+      "step": 38060
+    },
+    {
+      "epoch": 0.19011710654448302,
+      "grad_norm": 0.09160934388637543,
+      "learning_rate": 2.435730556459486e-05,
+      "loss": 9.8012,
+      "step": 38070
+    },
+    {
+      "epoch": 0.19016704536942244,
+      "grad_norm": 0.09235961735248566,
+      "learning_rate": 2.4355803649653308e-05,
+      "loss": 9.8039,
+      "step": 38080
+    },
+    {
+      "epoch": 0.1902169841943619,
+      "grad_norm": 0.0927954837679863,
+      "learning_rate": 2.4354301734711758e-05,
+      "loss": 9.805,
+      "step": 38090
+    },
+    {
+      "epoch": 0.19026692301930134,
+      "grad_norm": 0.09588232636451721,
+      "learning_rate": 2.4352799819770208e-05,
+      "loss": 9.8001,
+      "step": 38100
+    },
+    {
+      "epoch": 0.1903168618442408,
+      "grad_norm": 0.09209771454334259,
+      "learning_rate": 2.435129790482866e-05,
+      "loss": 9.7988,
+      "step": 38110
+    },
+    {
+      "epoch": 0.19036680066918024,
+      "grad_norm": 0.09557371586561203,
+      "learning_rate": 2.434979598988711e-05,
+      "loss": 9.7982,
+      "step": 38120
+    },
+    {
+      "epoch": 0.1904167394941197,
+      "grad_norm": 0.08873193711042404,
+      "learning_rate": 2.4348294074945555e-05,
+      "loss": 9.8029,
+      "step": 38130
+    },
+    {
+      "epoch": 0.19046667831905914,
+      "grad_norm": 0.09229250252246857,
+      "learning_rate": 2.4346792160004006e-05,
+      "loss": 9.8071,
+      "step": 38140
+    },
+    {
+      "epoch": 0.1905166171439986,
+      "grad_norm": 0.09603758901357651,
+      "learning_rate": 2.4345290245062456e-05,
+      "loss": 9.8052,
+      "step": 38150
+    },
+    {
+      "epoch": 0.19056655596893804,
+      "grad_norm": 0.09332188963890076,
+      "learning_rate": 2.4343788330120906e-05,
+      "loss": 9.7968,
+      "step": 38160
+    },
+    {
+      "epoch": 0.1906164947938775,
+      "grad_norm": 0.09354902803897858,
+      "learning_rate": 2.4342286415179356e-05,
+      "loss": 9.7974,
+      "step": 38170
+    },
+    {
+      "epoch": 0.19066643361881694,
+      "grad_norm": 0.0928342416882515,
+      "learning_rate": 2.4340784500237803e-05,
+      "loss": 9.7945,
+      "step": 38180
+    },
+    {
+      "epoch": 0.1907163724437564,
+      "grad_norm": 0.0948493480682373,
+      "learning_rate": 2.4339282585296253e-05,
+      "loss": 9.7977,
+      "step": 38190
+    },
+    {
+      "epoch": 0.19076631126869584,
+      "grad_norm": 0.09960921853780746,
+      "learning_rate": 2.4337780670354703e-05,
+      "loss": 9.8052,
+      "step": 38200
+    },
+    {
+      "epoch": 0.1908162500936353,
+      "grad_norm": 0.09617515653371811,
+      "learning_rate": 2.4336278755413153e-05,
+      "loss": 9.7999,
+      "step": 38210
+    },
+    {
+      "epoch": 0.19086618891857474,
+      "grad_norm": 0.09649699926376343,
+      "learning_rate": 2.4334776840471604e-05,
+      "loss": 9.8018,
+      "step": 38220
+    },
+    {
+      "epoch": 0.19091612774351419,
+      "grad_norm": 0.09706174582242966,
+      "learning_rate": 2.433327492553005e-05,
+      "loss": 9.8035,
+      "step": 38230
+    },
+    {
+      "epoch": 0.19096606656845364,
+      "grad_norm": 0.09482137858867645,
+      "learning_rate": 2.43317730105885e-05,
+      "loss": 9.8009,
+      "step": 38240
+    },
+    {
+      "epoch": 0.19101600539339308,
+      "grad_norm": 0.0948929712176323,
+      "learning_rate": 2.433027109564695e-05,
+      "loss": 9.796,
+      "step": 38250
+    },
+    {
+      "epoch": 0.19106594421833253,
+      "grad_norm": 0.09649039804935455,
+      "learning_rate": 2.43287691807054e-05,
+      "loss": 9.8034,
+      "step": 38260
+    },
+    {
+      "epoch": 0.19111588304327198,
+      "grad_norm": 0.09573374688625336,
+      "learning_rate": 2.432726726576385e-05,
+      "loss": 9.7922,
+      "step": 38270
+    },
+    {
+      "epoch": 0.19116582186821143,
+      "grad_norm": 0.10029622167348862,
+      "learning_rate": 2.4325765350822298e-05,
+      "loss": 9.7982,
+      "step": 38280
+    },
+    {
+      "epoch": 0.19121576069315088,
+      "grad_norm": 0.09856070578098297,
+      "learning_rate": 2.4324263435880748e-05,
+      "loss": 9.8087,
+      "step": 38290
+    },
+    {
+      "epoch": 0.19126569951809033,
+      "grad_norm": 0.08724341541528702,
+      "learning_rate": 2.4322761520939198e-05,
+      "loss": 9.7989,
+      "step": 38300
+    },
+    {
+      "epoch": 0.19131563834302978,
+      "grad_norm": 0.09495957940816879,
+      "learning_rate": 2.432125960599765e-05,
+      "loss": 9.7892,
+      "step": 38310
+    },
+    {
+      "epoch": 0.19136557716796923,
+      "grad_norm": 0.09710295498371124,
+      "learning_rate": 2.43197576910561e-05,
+      "loss": 9.8049,
+      "step": 38320
+    },
+    {
+      "epoch": 0.19141551599290868,
+      "grad_norm": 0.0903317853808403,
+      "learning_rate": 2.4318255776114545e-05,
+      "loss": 9.8063,
+      "step": 38330
+    },
+    {
+      "epoch": 0.19146545481784813,
+      "grad_norm": 0.09293761849403381,
+      "learning_rate": 2.4316753861172996e-05,
+      "loss": 9.7994,
+      "step": 38340
+    },
+    {
+      "epoch": 0.19151539364278758,
+      "grad_norm": 0.09713525325059891,
+      "learning_rate": 2.4315251946231446e-05,
+      "loss": 9.7968,
+      "step": 38350
+    },
+    {
+      "epoch": 0.19156533246772703,
+      "grad_norm": 0.09488287568092346,
+      "learning_rate": 2.4313750031289896e-05,
+      "loss": 9.8005,
+      "step": 38360
+    },
+    {
+      "epoch": 0.19161527129266648,
+      "grad_norm": 0.0962325856089592,
+      "learning_rate": 2.4312248116348346e-05,
+      "loss": 9.7941,
+      "step": 38370
+    },
+    {
+      "epoch": 0.19166521011760593,
+      "grad_norm": 0.09315495938062668,
+      "learning_rate": 2.4310746201406793e-05,
+      "loss": 9.7957,
+      "step": 38380
+    },
+    {
+      "epoch": 0.19171514894254538,
+      "grad_norm": 0.09810058772563934,
+      "learning_rate": 2.4309244286465243e-05,
+      "loss": 9.8038,
+      "step": 38390
+    },
+    {
+      "epoch": 0.19176508776748483,
+      "grad_norm": 0.09418797492980957,
+      "learning_rate": 2.4307742371523693e-05,
+      "loss": 9.7966,
+      "step": 38400
+    },
+    {
+      "epoch": 0.19181502659242428,
+      "grad_norm": 0.0971032902598381,
+      "learning_rate": 2.4306240456582143e-05,
+      "loss": 9.7953,
+      "step": 38410
+    },
+    {
+      "epoch": 0.19186496541736373,
+      "grad_norm": 0.09385687112808228,
+      "learning_rate": 2.4304738541640594e-05,
+      "loss": 9.7957,
+      "step": 38420
+    },
+    {
+      "epoch": 0.19191490424230317,
+      "grad_norm": 0.09179292619228363,
+      "learning_rate": 2.430323662669904e-05,
+      "loss": 9.7907,
+      "step": 38430
+    },
+    {
+      "epoch": 0.19196484306724262,
+      "grad_norm": 0.09452084451913834,
+      "learning_rate": 2.430173471175749e-05,
+      "loss": 9.8014,
+      "step": 38440
+    },
+    {
+      "epoch": 0.19201478189218207,
+      "grad_norm": 0.09643220156431198,
+      "learning_rate": 2.430023279681594e-05,
+      "loss": 9.7915,
+      "step": 38450
+    },
+    {
+      "epoch": 0.19206472071712152,
+      "grad_norm": 0.0984674021601677,
+      "learning_rate": 2.429873088187439e-05,
+      "loss": 9.7976,
+      "step": 38460
+    },
+    {
+      "epoch": 0.19211465954206097,
+      "grad_norm": 0.09368205070495605,
+      "learning_rate": 2.429722896693284e-05,
+      "loss": 9.7938,
+      "step": 38470
+    },
+    {
+      "epoch": 0.19216459836700042,
+      "grad_norm": 0.09682975709438324,
+      "learning_rate": 2.429572705199129e-05,
+      "loss": 9.7942,
+      "step": 38480
+    },
+    {
+      "epoch": 0.19221453719193987,
+      "grad_norm": 0.09321840852499008,
+      "learning_rate": 2.4294225137049738e-05,
+      "loss": 9.7997,
+      "step": 38490
+    },
+    {
+      "epoch": 0.19226447601687932,
+      "grad_norm": 0.09486759454011917,
+      "learning_rate": 2.4292723222108188e-05,
+      "loss": 9.7955,
+      "step": 38500
+    },
+    {
+      "epoch": 0.19231441484181877,
+      "grad_norm": 0.09923727810382843,
+      "learning_rate": 2.429122130716664e-05,
+      "loss": 9.7936,
+      "step": 38510
+    },
+    {
+      "epoch": 0.19236435366675822,
+      "grad_norm": 0.09178563952445984,
+      "learning_rate": 2.428971939222509e-05,
+      "loss": 9.7945,
+      "step": 38520
+    },
+    {
+      "epoch": 0.19241429249169767,
+      "grad_norm": 0.09080483019351959,
+      "learning_rate": 2.428821747728354e-05,
+      "loss": 9.7983,
+      "step": 38530
+    },
+    {
+      "epoch": 0.19246423131663712,
+      "grad_norm": 0.09831558167934418,
+      "learning_rate": 2.4286715562341986e-05,
+      "loss": 9.7918,
+      "step": 38540
+    },
+    {
+      "epoch": 0.19251417014157657,
+      "grad_norm": 0.095986008644104,
+      "learning_rate": 2.4285213647400436e-05,
+      "loss": 9.792,
+      "step": 38550
+    },
+    {
+      "epoch": 0.19256410896651602,
+      "grad_norm": 0.10043915361166,
+      "learning_rate": 2.4283711732458886e-05,
+      "loss": 9.7913,
+      "step": 38560
+    },
+    {
+      "epoch": 0.19261404779145547,
+      "grad_norm": 0.09741128236055374,
+      "learning_rate": 2.4282209817517336e-05,
+      "loss": 9.7836,
+      "step": 38570
+    },
+    {
+      "epoch": 0.19266398661639492,
+      "grad_norm": 0.09496723860502243,
+      "learning_rate": 2.4280707902575786e-05,
+      "loss": 9.7869,
+      "step": 38580
+    },
+    {
+      "epoch": 0.19271392544133437,
+      "grad_norm": 0.09604073315858841,
+      "learning_rate": 2.4279205987634233e-05,
+      "loss": 9.7975,
+      "step": 38590
+    },
+    {
+      "epoch": 0.19276386426627382,
+      "grad_norm": 0.09631392359733582,
+      "learning_rate": 2.4277704072692683e-05,
+      "loss": 9.7917,
+      "step": 38600
+    },
+    {
+      "epoch": 0.19281380309121326,
+      "grad_norm": 0.09735754877328873,
+      "learning_rate": 2.4276202157751133e-05,
+      "loss": 9.7898,
+      "step": 38610
+    },
+    {
+      "epoch": 0.19286374191615271,
+      "grad_norm": 0.09502192586660385,
+      "learning_rate": 2.4274700242809584e-05,
+      "loss": 9.7959,
+      "step": 38620
+    },
+    {
+      "epoch": 0.19291368074109216,
+      "grad_norm": 0.09539380669593811,
+      "learning_rate": 2.4273198327868034e-05,
+      "loss": 9.7939,
+      "step": 38630
+    },
+    {
+      "epoch": 0.1929636195660316,
+      "grad_norm": 0.09433183073997498,
+      "learning_rate": 2.427169641292648e-05,
+      "loss": 9.7888,
+      "step": 38640
+    },
+    {
+      "epoch": 0.19301355839097106,
+      "grad_norm": 0.09847880899906158,
+      "learning_rate": 2.427019449798493e-05,
+      "loss": 9.7895,
+      "step": 38650
+    },
+    {
+      "epoch": 0.1930634972159105,
+      "grad_norm": 0.09963309019804001,
+      "learning_rate": 2.426869258304338e-05,
+      "loss": 9.7924,
+      "step": 38660
+    },
+    {
+      "epoch": 0.19311343604084996,
+      "grad_norm": 0.09441536664962769,
+      "learning_rate": 2.426719066810183e-05,
+      "loss": 9.7913,
+      "step": 38670
+    },
+    {
+      "epoch": 0.1931633748657894,
+      "grad_norm": 0.09207624942064285,
+      "learning_rate": 2.426568875316028e-05,
+      "loss": 9.7925,
+      "step": 38680
+    },
+    {
+      "epoch": 0.19321331369072886,
+      "grad_norm": 0.09765110164880753,
+      "learning_rate": 2.4264186838218728e-05,
+      "loss": 9.7963,
+      "step": 38690
+    },
+    {
+      "epoch": 0.1932632525156683,
+      "grad_norm": 0.09711901098489761,
+      "learning_rate": 2.4262684923277178e-05,
+      "loss": 9.788,
+      "step": 38700
+    },
+    {
+      "epoch": 0.19331319134060776,
+      "grad_norm": 0.09210279583930969,
+      "learning_rate": 2.426118300833563e-05,
+      "loss": 9.7852,
+      "step": 38710
+    },
+    {
+      "epoch": 0.1933631301655472,
+      "grad_norm": 0.09786038845777512,
+      "learning_rate": 2.425968109339408e-05,
+      "loss": 9.7915,
+      "step": 38720
+    },
+    {
+      "epoch": 0.19341306899048666,
+      "grad_norm": 0.09457087516784668,
+      "learning_rate": 2.425817917845253e-05,
+      "loss": 9.7967,
+      "step": 38730
+    },
+    {
+      "epoch": 0.1934630078154261,
+      "grad_norm": 0.09489721804857254,
+      "learning_rate": 2.4256677263510976e-05,
+      "loss": 9.7965,
+      "step": 38740
+    },
+    {
+      "epoch": 0.19351294664036556,
+      "grad_norm": 0.09461694210767746,
+      "learning_rate": 2.4255175348569426e-05,
+      "loss": 9.7884,
+      "step": 38750
+    },
+    {
+      "epoch": 0.193562885465305,
+      "grad_norm": 0.10121167451143265,
+      "learning_rate": 2.4253673433627876e-05,
+      "loss": 9.7838,
+      "step": 38760
+    },
+    {
+      "epoch": 0.19361282429024446,
+      "grad_norm": 0.09524286538362503,
+      "learning_rate": 2.4252171518686326e-05,
+      "loss": 9.7871,
+      "step": 38770
+    },
+    {
+      "epoch": 0.1936627631151839,
+      "grad_norm": 0.09742001444101334,
+      "learning_rate": 2.4250669603744776e-05,
+      "loss": 9.7947,
+      "step": 38780
+    },
+    {
+      "epoch": 0.19371270194012336,
+      "grad_norm": 0.09000395238399506,
+      "learning_rate": 2.4249167688803223e-05,
+      "loss": 9.7865,
+      "step": 38790
+    },
+    {
+      "epoch": 0.1937626407650628,
+      "grad_norm": 0.09661939740180969,
+      "learning_rate": 2.4247665773861677e-05,
+      "loss": 9.7887,
+      "step": 38800
+    },
+    {
+      "epoch": 0.19381257959000225,
+      "grad_norm": 0.09397201985120773,
+      "learning_rate": 2.4246163858920123e-05,
+      "loss": 9.7851,
+      "step": 38810
+    },
+    {
+      "epoch": 0.1938625184149417,
+      "grad_norm": 0.0975542664527893,
+      "learning_rate": 2.4244661943978574e-05,
+      "loss": 9.786,
+      "step": 38820
+    },
+    {
+      "epoch": 0.19391245723988115,
+      "grad_norm": 0.0956290140748024,
+      "learning_rate": 2.4243160029037024e-05,
+      "loss": 9.783,
+      "step": 38830
+    },
+    {
+      "epoch": 0.1939623960648206,
+      "grad_norm": 0.09244171530008316,
+      "learning_rate": 2.424165811409547e-05,
+      "loss": 9.7812,
+      "step": 38840
+    },
+    {
+      "epoch": 0.19401233488976005,
+      "grad_norm": 0.10087534040212631,
+      "learning_rate": 2.4240156199153924e-05,
+      "loss": 9.7849,
+      "step": 38850
+    },
+    {
+      "epoch": 0.1940622737146995,
+      "grad_norm": 0.09302230924367905,
+      "learning_rate": 2.423865428421237e-05,
+      "loss": 9.7873,
+      "step": 38860
+    },
+    {
+      "epoch": 0.19411221253963895,
+      "grad_norm": 0.09214049577713013,
+      "learning_rate": 2.423715236927082e-05,
+      "loss": 9.7873,
+      "step": 38870
+    },
+    {
+      "epoch": 0.1941621513645784,
+      "grad_norm": 0.09628362953662872,
+      "learning_rate": 2.423565045432927e-05,
+      "loss": 9.7885,
+      "step": 38880
+    },
+    {
+      "epoch": 0.19421209018951785,
+      "grad_norm": 0.09119916707277298,
+      "learning_rate": 2.4234148539387718e-05,
+      "loss": 9.7847,
+      "step": 38890
+    },
+    {
+      "epoch": 0.1942620290144573,
+      "grad_norm": 0.09237845242023468,
+      "learning_rate": 2.423264662444617e-05,
+      "loss": 9.7874,
+      "step": 38900
+    },
+    {
+      "epoch": 0.19431196783939675,
+      "grad_norm": 0.08875925093889236,
+      "learning_rate": 2.423114470950462e-05,
+      "loss": 9.7925,
+      "step": 38910
+    },
+    {
+      "epoch": 0.1943619066643362,
+      "grad_norm": 0.0932464450597763,
+      "learning_rate": 2.422964279456307e-05,
+      "loss": 9.7909,
+      "step": 38920
+    },
+    {
+      "epoch": 0.19441184548927565,
+      "grad_norm": 0.09021969884634018,
+      "learning_rate": 2.422814087962152e-05,
+      "loss": 9.7863,
+      "step": 38930
+    },
+    {
+      "epoch": 0.1944617843142151,
+      "grad_norm": 0.09389606863260269,
+      "learning_rate": 2.4226638964679966e-05,
+      "loss": 9.7808,
+      "step": 38940
+    },
+    {
+      "epoch": 0.19451172313915455,
+      "grad_norm": 0.09377970546483994,
+      "learning_rate": 2.422513704973842e-05,
+      "loss": 9.7729,
+      "step": 38950
+    },
+    {
+      "epoch": 0.194561661964094,
+      "grad_norm": 0.09550289064645767,
+      "learning_rate": 2.4223635134796866e-05,
+      "loss": 9.7801,
+      "step": 38960
+    },
+    {
+      "epoch": 0.19461160078903345,
+      "grad_norm": 0.0985620841383934,
+      "learning_rate": 2.4222133219855316e-05,
+      "loss": 9.7896,
+      "step": 38970
+    },
+    {
+      "epoch": 0.1946615396139729,
+      "grad_norm": 0.09124714881181717,
+      "learning_rate": 2.4220631304913766e-05,
+      "loss": 9.7848,
+      "step": 38980
+    },
+    {
+      "epoch": 0.19471147843891234,
+      "grad_norm": 0.09538350999355316,
+      "learning_rate": 2.4219129389972213e-05,
+      "loss": 9.7863,
+      "step": 38990
+    },
+    {
+      "epoch": 0.1947614172638518,
+      "grad_norm": 0.09843659400939941,
+      "learning_rate": 2.4217627475030667e-05,
+      "loss": 9.7772,
+      "step": 39000
+    },
+    {
+      "epoch": 0.19481135608879124,
+      "grad_norm": 0.09667788445949554,
+      "learning_rate": 2.4216125560089113e-05,
+      "loss": 9.7902,
+      "step": 39010
+    },
+    {
+      "epoch": 0.1948612949137307,
+      "grad_norm": 0.09687555581331253,
+      "learning_rate": 2.4214623645147564e-05,
+      "loss": 9.7835,
+      "step": 39020
+    },
+    {
+      "epoch": 0.19491123373867014,
+      "grad_norm": 0.09759131073951721,
+      "learning_rate": 2.4213121730206014e-05,
+      "loss": 9.7849,
+      "step": 39030
+    },
+    {
+      "epoch": 0.1949611725636096,
+      "grad_norm": 0.09392527490854263,
+      "learning_rate": 2.421161981526446e-05,
+      "loss": 9.784,
+      "step": 39040
+    },
+    {
+      "epoch": 0.19501111138854904,
+      "grad_norm": 0.09155687689781189,
+      "learning_rate": 2.4210117900322914e-05,
+      "loss": 9.7849,
+      "step": 39050
+    },
+    {
+      "epoch": 0.1950610502134885,
+      "grad_norm": 0.09576915949583054,
+      "learning_rate": 2.420861598538136e-05,
+      "loss": 9.7815,
+      "step": 39060
+    },
+    {
+      "epoch": 0.1951109890384279,
+      "grad_norm": 0.09315671026706696,
+      "learning_rate": 2.420711407043981e-05,
+      "loss": 9.7802,
+      "step": 39070
+    },
+    {
+      "epoch": 0.19516092786336736,
+      "grad_norm": 0.09243351221084595,
+      "learning_rate": 2.420561215549826e-05,
+      "loss": 9.7802,
+      "step": 39080
+    },
+    {
+      "epoch": 0.1952108666883068,
+      "grad_norm": 0.09323740005493164,
+      "learning_rate": 2.4204110240556708e-05,
+      "loss": 9.7835,
+      "step": 39090
+    },
+    {
+      "epoch": 0.19526080551324626,
+      "grad_norm": 0.09786425530910492,
+      "learning_rate": 2.420260832561516e-05,
+      "loss": 9.78,
+      "step": 39100
+    },
+    {
+      "epoch": 0.1953107443381857,
+      "grad_norm": 0.09116844832897186,
+      "learning_rate": 2.420110641067361e-05,
+      "loss": 9.7841,
+      "step": 39110
+    },
+    {
+      "epoch": 0.19536068316312516,
+      "grad_norm": 0.09829219430685043,
+      "learning_rate": 2.4199604495732062e-05,
+      "loss": 9.7839,
+      "step": 39120
+    },
+    {
+      "epoch": 0.1954106219880646,
+      "grad_norm": 0.0964910238981247,
+      "learning_rate": 2.419810258079051e-05,
+      "loss": 9.7754,
+      "step": 39130
+    },
+    {
+      "epoch": 0.19546056081300406,
+      "grad_norm": 0.09510182589292526,
+      "learning_rate": 2.4196600665848956e-05,
+      "loss": 9.779,
+      "step": 39140
+    },
+    {
+      "epoch": 0.1955104996379435,
+      "grad_norm": 0.09461788833141327,
+      "learning_rate": 2.419509875090741e-05,
+      "loss": 9.7828,
+      "step": 39150
+    },
+    {
+      "epoch": 0.19556043846288296,
+      "grad_norm": 0.09442439675331116,
+      "learning_rate": 2.4193596835965856e-05,
+      "loss": 45.9869,
+      "step": 39160
+    },
+    {
+      "epoch": 0.1956103772878224,
+      "grad_norm": 0.09443948417901993,
+      "learning_rate": 2.419209492102431e-05,
+      "loss": 9.7777,
+      "step": 39170
+    },
+    {
+      "epoch": 0.19566031611276186,
+      "grad_norm": 0.08871215581893921,
+      "learning_rate": 2.4190593006082756e-05,
+      "loss": 9.7808,
+      "step": 39180
+    },
+    {
+      "epoch": 0.1957102549377013,
+      "grad_norm": 0.09522920846939087,
+      "learning_rate": 2.4189091091141203e-05,
+      "loss": 9.7887,
+      "step": 39190
+    },
+    {
+      "epoch": 0.19576019376264076,
+      "grad_norm": 0.0978919193148613,
+      "learning_rate": 2.4187589176199657e-05,
+      "loss": 9.7724,
+      "step": 39200
+    },
+    {
+      "epoch": 0.1958101325875802,
+      "grad_norm": 0.09242391586303711,
+      "learning_rate": 2.4186087261258103e-05,
+      "loss": 9.7893,
+      "step": 39210
+    },
+    {
+      "epoch": 0.19586007141251965,
+      "grad_norm": 0.09181750565767288,
+      "learning_rate": 2.4184585346316557e-05,
+      "loss": 9.7748,
+      "step": 39220
+    },
+    {
+      "epoch": 0.1959100102374591,
+      "grad_norm": 0.0938679501414299,
+      "learning_rate": 2.4183083431375004e-05,
+      "loss": 9.7745,
+      "step": 39230
+    },
+    {
+      "epoch": 0.19595994906239855,
+      "grad_norm": 0.09096869826316833,
+      "learning_rate": 2.418158151643345e-05,
+      "loss": 9.7836,
+      "step": 39240
+    },
+    {
+      "epoch": 0.196009887887338,
+      "grad_norm": 0.09524725377559662,
+      "learning_rate": 2.4180079601491904e-05,
+      "loss": 9.787,
+      "step": 39250
+    },
+    {
+      "epoch": 0.19605982671227745,
+      "grad_norm": 0.09219633787870407,
+      "learning_rate": 2.417857768655035e-05,
+      "loss": 9.7777,
+      "step": 39260
+    },
+    {
+      "epoch": 0.1961097655372169,
+      "grad_norm": 0.09788725525140762,
+      "learning_rate": 2.4177075771608805e-05,
+      "loss": 9.7787,
+      "step": 39270
+    },
+    {
+      "epoch": 0.19615970436215635,
+      "grad_norm": 0.09663385152816772,
+      "learning_rate": 2.417557385666725e-05,
+      "loss": 9.7779,
+      "step": 39280
+    },
+    {
+      "epoch": 0.1962096431870958,
+      "grad_norm": 0.10418345034122467,
+      "learning_rate": 2.4174071941725698e-05,
+      "loss": 9.7713,
+      "step": 39290
+    },
+    {
+      "epoch": 0.19625958201203525,
+      "grad_norm": 0.09116250276565552,
+      "learning_rate": 2.417257002678415e-05,
+      "loss": 9.7772,
+      "step": 39300
+    },
+    {
+      "epoch": 0.1963095208369747,
+      "grad_norm": 0.09256359934806824,
+      "learning_rate": 2.41710681118426e-05,
+      "loss": 9.7782,
+      "step": 39310
+    },
+    {
+      "epoch": 0.19635945966191415,
+      "grad_norm": 0.10003110766410828,
+      "learning_rate": 2.4169566196901052e-05,
+      "loss": 9.7717,
+      "step": 39320
+    },
+    {
+      "epoch": 0.1964093984868536,
+      "grad_norm": 0.09774117916822433,
+      "learning_rate": 2.41680642819595e-05,
+      "loss": 9.777,
+      "step": 39330
+    },
+    {
+      "epoch": 0.19645933731179305,
+      "grad_norm": 0.09338162839412689,
+      "learning_rate": 2.4166562367017946e-05,
+      "loss": 9.7811,
+      "step": 39340
+    },
+    {
+      "epoch": 0.1965092761367325,
+      "grad_norm": 0.09204205125570297,
+      "learning_rate": 2.41650604520764e-05,
+      "loss": 9.7827,
+      "step": 39350
+    },
+    {
+      "epoch": 0.19655921496167195,
+      "grad_norm": 0.0937473401427269,
+      "learning_rate": 2.4163558537134846e-05,
+      "loss": 9.7759,
+      "step": 39360
+    },
+    {
+      "epoch": 0.1966091537866114,
+      "grad_norm": 0.09584249556064606,
+      "learning_rate": 2.41620566221933e-05,
+      "loss": 9.7817,
+      "step": 39370
+    },
+    {
+      "epoch": 0.19665909261155085,
+      "grad_norm": 0.09530872106552124,
+      "learning_rate": 2.4160554707251746e-05,
+      "loss": 9.7736,
+      "step": 39380
+    },
+    {
+      "epoch": 0.1967090314364903,
+      "grad_norm": 0.09426391869783401,
+      "learning_rate": 2.4159052792310193e-05,
+      "loss": 9.7796,
+      "step": 39390
+    },
+    {
+      "epoch": 0.19675897026142974,
+      "grad_norm": 0.09266220033168793,
+      "learning_rate": 2.4157550877368647e-05,
+      "loss": 9.7739,
+      "step": 39400
+    },
+    {
+      "epoch": 0.1968089090863692,
+      "grad_norm": 0.0935848280787468,
+      "learning_rate": 2.4156048962427093e-05,
+      "loss": 9.7755,
+      "step": 39410
+    },
+    {
+      "epoch": 0.19685884791130864,
+      "grad_norm": 0.0957341194152832,
+      "learning_rate": 2.4154547047485547e-05,
+      "loss": 9.7699,
+      "step": 39420
+    },
+    {
+      "epoch": 0.1969087867362481,
+      "grad_norm": 0.09587478637695312,
+      "learning_rate": 2.4153045132543994e-05,
+      "loss": 9.7725,
+      "step": 39430
+    },
+    {
+      "epoch": 0.19695872556118754,
+      "grad_norm": 0.09046075493097305,
+      "learning_rate": 2.4151543217602444e-05,
+      "loss": 9.776,
+      "step": 39440
+    },
+    {
+      "epoch": 0.197008664386127,
+      "grad_norm": 0.09086257219314575,
+      "learning_rate": 2.4150041302660894e-05,
+      "loss": 9.7735,
+      "step": 39450
+    },
+    {
+      "epoch": 0.19705860321106644,
+      "grad_norm": 0.10431148111820221,
+      "learning_rate": 2.414853938771934e-05,
+      "loss": 9.773,
+      "step": 39460
+    },
+    {
+      "epoch": 0.1971085420360059,
+      "grad_norm": 0.09601059556007385,
+      "learning_rate": 2.4147037472777795e-05,
+      "loss": 9.779,
+      "step": 39470
+    },
+    {
+      "epoch": 0.19715848086094534,
+      "grad_norm": 0.09748616814613342,
+      "learning_rate": 2.414553555783624e-05,
+      "loss": 9.7766,
+      "step": 39480
+    },
+    {
+      "epoch": 0.1972084196858848,
+      "grad_norm": 0.10166972130537033,
+      "learning_rate": 2.414403364289469e-05,
+      "loss": 9.7738,
+      "step": 39490
+    },
+    {
+      "epoch": 0.19725835851082424,
+      "grad_norm": 0.09647940844297409,
+      "learning_rate": 2.4142531727953142e-05,
+      "loss": 9.7646,
+      "step": 39500
+    },
+    {
+      "epoch": 0.1973082973357637,
+      "grad_norm": 0.09661668539047241,
+      "learning_rate": 2.414102981301159e-05,
+      "loss": 9.7793,
+      "step": 39510
+    },
+    {
+      "epoch": 0.19735823616070314,
+      "grad_norm": 0.09515753388404846,
+      "learning_rate": 2.4139527898070042e-05,
+      "loss": 9.7797,
+      "step": 39520
+    },
+    {
+      "epoch": 0.1974081749856426,
+      "grad_norm": 0.0913529172539711,
+      "learning_rate": 2.413802598312849e-05,
+      "loss": 9.7787,
+      "step": 39530
+    },
+    {
+      "epoch": 0.19745811381058204,
+      "grad_norm": 0.09632540494203568,
+      "learning_rate": 2.413652406818694e-05,
+      "loss": 9.7703,
+      "step": 39540
+    },
+    {
+      "epoch": 0.19750805263552149,
+      "grad_norm": 0.09521202743053436,
+      "learning_rate": 2.413502215324539e-05,
+      "loss": 9.7713,
+      "step": 39550
+    },
+    {
+      "epoch": 0.19755799146046094,
+      "grad_norm": 0.09425238519906998,
+      "learning_rate": 2.4133520238303836e-05,
+      "loss": 9.7705,
+      "step": 39560
+    },
+    {
+      "epoch": 0.19760793028540038,
+      "grad_norm": 0.09330090880393982,
+      "learning_rate": 2.413201832336229e-05,
+      "loss": 9.7747,
+      "step": 39570
+    },
+    {
+      "epoch": 0.19765786911033983,
+      "grad_norm": 0.09047570824623108,
+      "learning_rate": 2.4130516408420736e-05,
+      "loss": 9.7726,
+      "step": 39580
+    },
+    {
+      "epoch": 0.19770780793527928,
+      "grad_norm": 0.09425200521945953,
+      "learning_rate": 2.4129014493479187e-05,
+      "loss": 9.7723,
+      "step": 39590
+    },
+    {
+      "epoch": 0.19775774676021873,
+      "grad_norm": 0.09698259085416794,
+      "learning_rate": 2.4127512578537637e-05,
+      "loss": 9.7848,
+      "step": 39600
+    },
+    {
+      "epoch": 0.19780768558515818,
+      "grad_norm": 0.09237400442361832,
+      "learning_rate": 2.4126010663596083e-05,
+      "loss": 9.774,
+      "step": 39610
+    },
+    {
+      "epoch": 0.19785762441009763,
+      "grad_norm": 0.09415727108716965,
+      "learning_rate": 2.4124508748654537e-05,
+      "loss": 9.7657,
+      "step": 39620
+    },
+    {
+      "epoch": 0.19790756323503708,
+      "grad_norm": 0.09116394817829132,
+      "learning_rate": 2.4123006833712984e-05,
+      "loss": 9.7767,
+      "step": 39630
+    },
+    {
+      "epoch": 0.19795750205997653,
+      "grad_norm": 0.09226060658693314,
+      "learning_rate": 2.4121504918771434e-05,
+      "loss": 9.7691,
+      "step": 39640
+    },
+    {
+      "epoch": 0.19800744088491598,
+      "grad_norm": 0.09217618405818939,
+      "learning_rate": 2.4120003003829884e-05,
+      "loss": 9.7718,
+      "step": 39650
+    },
+    {
+      "epoch": 0.19805737970985543,
+      "grad_norm": 0.09849265962839127,
+      "learning_rate": 2.411850108888833e-05,
+      "loss": 9.7769,
+      "step": 39660
+    },
+    {
+      "epoch": 0.19810731853479488,
+      "grad_norm": 0.09423698484897614,
+      "learning_rate": 2.4116999173946785e-05,
+      "loss": 9.7673,
+      "step": 39670
+    },
+    {
+      "epoch": 0.19815725735973433,
+      "grad_norm": 0.09352479875087738,
+      "learning_rate": 2.411549725900523e-05,
+      "loss": 9.764,
+      "step": 39680
+    },
+    {
+      "epoch": 0.19820719618467378,
+      "grad_norm": 0.09337835013866425,
+      "learning_rate": 2.411399534406368e-05,
+      "loss": 9.7659,
+      "step": 39690
+    },
+    {
+      "epoch": 0.19825713500961323,
+      "grad_norm": 0.08959398418664932,
+      "learning_rate": 2.4112493429122132e-05,
+      "loss": 9.7724,
+      "step": 39700
+    },
+    {
+      "epoch": 0.19830707383455268,
+      "grad_norm": 0.09081783145666122,
+      "learning_rate": 2.411099151418058e-05,
+      "loss": 9.7665,
+      "step": 39710
+    },
+    {
+      "epoch": 0.19835701265949213,
+      "grad_norm": 0.09610147774219513,
+      "learning_rate": 2.4109489599239032e-05,
+      "loss": 9.759,
+      "step": 39720
+    },
+    {
+      "epoch": 0.19840695148443158,
+      "grad_norm": 0.09528447687625885,
+      "learning_rate": 2.410798768429748e-05,
+      "loss": 9.7684,
+      "step": 39730
+    },
+    {
+      "epoch": 0.19845689030937103,
+      "grad_norm": 0.0971222072839737,
+      "learning_rate": 2.410648576935593e-05,
+      "loss": 9.7688,
+      "step": 39740
+    },
+    {
+      "epoch": 0.19850682913431048,
+      "grad_norm": 0.09684096276760101,
+      "learning_rate": 2.410498385441438e-05,
+      "loss": 9.7715,
+      "step": 39750
+    },
+    {
+      "epoch": 0.19855676795924992,
+      "grad_norm": 0.0940563753247261,
+      "learning_rate": 2.410348193947283e-05,
+      "loss": 9.7577,
+      "step": 39760
+    },
+    {
+      "epoch": 0.19860670678418937,
+      "grad_norm": 0.09706791490316391,
+      "learning_rate": 2.410198002453128e-05,
+      "loss": 9.7708,
+      "step": 39770
+    },
+    {
+      "epoch": 0.19865664560912882,
+      "grad_norm": 0.10084111243486404,
+      "learning_rate": 2.4100478109589726e-05,
+      "loss": 9.7658,
+      "step": 39780
+    },
+    {
+      "epoch": 0.19870658443406827,
+      "grad_norm": 0.09000294655561447,
+      "learning_rate": 2.4098976194648177e-05,
+      "loss": 9.7698,
+      "step": 39790
+    },
+    {
+      "epoch": 0.19875652325900772,
+      "grad_norm": 0.09340784698724747,
+      "learning_rate": 2.4097474279706627e-05,
+      "loss": 9.7694,
+      "step": 39800
+    },
+    {
+      "epoch": 0.19880646208394717,
+      "grad_norm": 0.09671244770288467,
+      "learning_rate": 2.4095972364765077e-05,
+      "loss": 9.7608,
+      "step": 39810
+    },
+    {
+      "epoch": 0.19885640090888662,
+      "grad_norm": 0.09407161921262741,
+      "learning_rate": 2.4094470449823527e-05,
+      "loss": 9.7632,
+      "step": 39820
+    },
+    {
+      "epoch": 0.19890633973382607,
+      "grad_norm": 0.09390780329704285,
+      "learning_rate": 2.4092968534881974e-05,
+      "loss": 9.7639,
+      "step": 39830
+    },
+    {
+      "epoch": 0.19895627855876552,
+      "grad_norm": 0.10034262388944626,
+      "learning_rate": 2.4091466619940424e-05,
+      "loss": 9.7688,
+      "step": 39840
+    },
+    {
+      "epoch": 0.19900621738370497,
+      "grad_norm": 0.09268135577440262,
+      "learning_rate": 2.4089964704998874e-05,
+      "loss": 9.7654,
+      "step": 39850
+    },
+    {
+      "epoch": 0.19905615620864442,
+      "grad_norm": 0.09061171859502792,
+      "learning_rate": 2.4088462790057324e-05,
+      "loss": 9.7667,
+      "step": 39860
+    },
+    {
+      "epoch": 0.19910609503358387,
+      "grad_norm": 0.09159281104803085,
+      "learning_rate": 2.4086960875115775e-05,
+      "loss": 9.7727,
+      "step": 39870
+    },
+    {
+      "epoch": 0.19915603385852332,
+      "grad_norm": 0.09168130904436111,
+      "learning_rate": 2.408545896017422e-05,
+      "loss": 9.7752,
+      "step": 39880
+    },
+    {
+      "epoch": 0.19920597268346277,
+      "grad_norm": 0.09341683238744736,
+      "learning_rate": 2.408395704523267e-05,
+      "loss": 9.7705,
+      "step": 39890
+    },
+    {
+      "epoch": 0.19925591150840222,
+      "grad_norm": 0.10817967355251312,
+      "learning_rate": 2.4082455130291122e-05,
+      "loss": 9.7616,
+      "step": 39900
+    },
+    {
+      "epoch": 0.19930585033334167,
+      "grad_norm": 0.09794482588768005,
+      "learning_rate": 2.4080953215349572e-05,
+      "loss": 9.7589,
+      "step": 39910
+    },
+    {
+      "epoch": 0.19935578915828112,
+      "grad_norm": 0.09640230238437653,
+      "learning_rate": 2.4079451300408022e-05,
+      "loss": 9.7614,
+      "step": 39920
+    },
+    {
+      "epoch": 0.19940572798322057,
+      "grad_norm": 0.09363255649805069,
+      "learning_rate": 2.407794938546647e-05,
+      "loss": 9.7684,
+      "step": 39930
+    },
+    {
+      "epoch": 0.19945566680816001,
+      "grad_norm": 0.09409525245428085,
+      "learning_rate": 2.407644747052492e-05,
+      "loss": 9.7558,
+      "step": 39940
+    },
+    {
+      "epoch": 0.19950560563309946,
+      "grad_norm": 0.0943169966340065,
+      "learning_rate": 2.407494555558337e-05,
+      "loss": 9.7663,
+      "step": 39950
+    },
+    {
+      "epoch": 0.1995555444580389,
+      "grad_norm": 0.09789732098579407,
+      "learning_rate": 2.407344364064182e-05,
+      "loss": 9.7623,
+      "step": 39960
+    },
+    {
+      "epoch": 0.19960548328297836,
+      "grad_norm": 0.09631295502185822,
+      "learning_rate": 2.407194172570027e-05,
+      "loss": 9.7692,
+      "step": 39970
+    },
+    {
+      "epoch": 0.1996554221079178,
+      "grad_norm": 0.09948205947875977,
+      "learning_rate": 2.4070439810758716e-05,
+      "loss": 9.7623,
+      "step": 39980
+    },
+    {
+      "epoch": 0.19970536093285726,
+      "grad_norm": 0.0964735895395279,
+      "learning_rate": 2.4068937895817167e-05,
+      "loss": 9.7621,
+      "step": 39990
+    },
+    {
+      "epoch": 0.1997552997577967,
+      "grad_norm": 0.09828507900238037,
+      "learning_rate": 2.4067435980875617e-05,
+      "loss": 9.7603,
+      "step": 40000
+    },
+    {
+      "epoch": 0.19980523858273616,
+      "grad_norm": 0.09244830161333084,
+      "learning_rate": 2.4065934065934067e-05,
+      "loss": 9.7644,
+      "step": 40010
+    },
+    {
+      "epoch": 0.1998551774076756,
+      "grad_norm": 0.09815704822540283,
+      "learning_rate": 2.4064432150992517e-05,
+      "loss": 9.7665,
+      "step": 40020
+    },
+    {
+      "epoch": 0.19990511623261506,
+      "grad_norm": 0.0952480360865593,
+      "learning_rate": 2.4062930236050964e-05,
+      "loss": 9.7668,
+      "step": 40030
+    },
+    {
+      "epoch": 0.1999550550575545,
+      "grad_norm": 0.09597064554691315,
+      "learning_rate": 2.4061428321109414e-05,
+      "loss": 9.7613,
+      "step": 40040
+    },
+    {
+      "epoch": 0.20000499388249393,
+      "grad_norm": 0.09825017303228378,
+      "learning_rate": 2.4059926406167864e-05,
+      "loss": 9.76,
+      "step": 40050
+    },
+    {
+      "epoch": 0.20005493270743338,
+      "grad_norm": 0.09238851070404053,
+      "learning_rate": 2.4058424491226314e-05,
+      "loss": 9.765,
+      "step": 40060
+    },
+    {
+      "epoch": 0.20010487153237283,
+      "grad_norm": 0.09596622735261917,
+      "learning_rate": 2.4056922576284765e-05,
+      "loss": 9.7618,
+      "step": 40070
+    },
+    {
+      "epoch": 0.20015481035731228,
+      "grad_norm": 0.09516778588294983,
+      "learning_rate": 2.4055420661343215e-05,
+      "loss": 9.766,
+      "step": 40080
+    },
+    {
+      "epoch": 0.20020474918225173,
+      "grad_norm": 0.10546738654375076,
+      "learning_rate": 2.405391874640166e-05,
+      "loss": 9.7556,
+      "step": 40090
+    },
+    {
+      "epoch": 0.20025468800719118,
+      "grad_norm": 0.09358225017786026,
+      "learning_rate": 2.4052416831460112e-05,
+      "loss": 9.759,
+      "step": 40100
+    },
+    {
+      "epoch": 0.20030462683213063,
+      "grad_norm": 0.08909553289413452,
+      "learning_rate": 2.4050914916518562e-05,
+      "loss": 9.7602,
+      "step": 40110
+    },
+    {
+      "epoch": 0.20035456565707008,
+      "grad_norm": 0.1016073226928711,
+      "learning_rate": 2.4049413001577012e-05,
+      "loss": 9.7603,
+      "step": 40120
+    },
+    {
+      "epoch": 0.20040450448200953,
+      "grad_norm": 0.09752840548753738,
+      "learning_rate": 2.4047911086635462e-05,
+      "loss": 9.7584,
+      "step": 40130
+    },
+    {
+      "epoch": 0.20045444330694898,
+      "grad_norm": 0.09266415238380432,
+      "learning_rate": 2.404640917169391e-05,
+      "loss": 9.7625,
+      "step": 40140
+    },
+    {
+      "epoch": 0.20050438213188843,
+      "grad_norm": 0.09927397966384888,
+      "learning_rate": 2.404490725675236e-05,
+      "loss": 9.7559,
+      "step": 40150
+    },
+    {
+      "epoch": 0.20055432095682788,
+      "grad_norm": 0.09626518934965134,
+      "learning_rate": 2.404340534181081e-05,
+      "loss": 9.7664,
+      "step": 40160
+    },
+    {
+      "epoch": 0.20060425978176732,
+      "grad_norm": 0.09638325124979019,
+      "learning_rate": 2.404190342686926e-05,
+      "loss": 9.7583,
+      "step": 40170
+    },
+    {
+      "epoch": 0.20065419860670677,
+      "grad_norm": 0.09331472218036652,
+      "learning_rate": 2.404040151192771e-05,
+      "loss": 9.752,
+      "step": 40180
+    },
+    {
+      "epoch": 0.20070413743164622,
+      "grad_norm": 0.09364567697048187,
+      "learning_rate": 2.4038899596986157e-05,
+      "loss": 9.7648,
+      "step": 40190
+    },
+    {
+      "epoch": 0.20075407625658567,
+      "grad_norm": 0.09309536218643188,
+      "learning_rate": 2.4037397682044607e-05,
+      "loss": 9.7596,
+      "step": 40200
+    },
+    {
+      "epoch": 0.20080401508152512,
+      "grad_norm": 0.091143399477005,
+      "learning_rate": 2.4035895767103057e-05,
+      "loss": 9.7572,
+      "step": 40210
+    },
+    {
+      "epoch": 0.20085395390646457,
+      "grad_norm": 0.09167341887950897,
+      "learning_rate": 2.4034393852161507e-05,
+      "loss": 9.7644,
+      "step": 40220
+    },
+    {
+      "epoch": 0.20090389273140402,
+      "grad_norm": 0.09573984146118164,
+      "learning_rate": 2.4032891937219957e-05,
+      "loss": 9.7607,
+      "step": 40230
+    },
+    {
+      "epoch": 0.20095383155634347,
+      "grad_norm": 0.0927746593952179,
+      "learning_rate": 2.4031390022278404e-05,
+      "loss": 9.7661,
+      "step": 40240
+    },
+    {
+      "epoch": 0.20100377038128292,
+      "grad_norm": 0.09519543498754501,
+      "learning_rate": 2.4029888107336854e-05,
+      "loss": 9.758,
+      "step": 40250
+    },
+    {
+      "epoch": 0.20105370920622237,
+      "grad_norm": 0.09138039499521255,
+      "learning_rate": 2.4028386192395304e-05,
+      "loss": 9.7658,
+      "step": 40260
+    },
+    {
+      "epoch": 0.20110364803116182,
+      "grad_norm": 0.0913807824254036,
+      "learning_rate": 2.4026884277453755e-05,
+      "loss": 9.7553,
+      "step": 40270
+    },
+    {
+      "epoch": 0.20115358685610127,
+      "grad_norm": 0.09113917499780655,
+      "learning_rate": 2.4025382362512205e-05,
+      "loss": 9.7586,
+      "step": 40280
+    },
+    {
+      "epoch": 0.20120352568104072,
+      "grad_norm": 0.09379607439041138,
+      "learning_rate": 2.402388044757065e-05,
+      "loss": 9.7621,
+      "step": 40290
+    },
+    {
+      "epoch": 0.20125346450598017,
+      "grad_norm": 0.09333807975053787,
+      "learning_rate": 2.4022378532629102e-05,
+      "loss": 9.7623,
+      "step": 40300
+    },
+    {
+      "epoch": 0.20130340333091962,
+      "grad_norm": 0.09286303073167801,
+      "learning_rate": 2.4020876617687552e-05,
+      "loss": 9.7592,
+      "step": 40310
+    },
+    {
+      "epoch": 0.20135334215585907,
+      "grad_norm": 0.09583623707294464,
+      "learning_rate": 2.4019374702746002e-05,
+      "loss": 9.7556,
+      "step": 40320
+    },
+    {
+      "epoch": 0.20140328098079852,
+      "grad_norm": 0.09831235557794571,
+      "learning_rate": 2.4017872787804452e-05,
+      "loss": 9.7499,
+      "step": 40330
+    },
+    {
+      "epoch": 0.20145321980573797,
+      "grad_norm": 0.09651533514261246,
+      "learning_rate": 2.40163708728629e-05,
+      "loss": 9.7571,
+      "step": 40340
+    },
+    {
+      "epoch": 0.20150315863067741,
+      "grad_norm": 0.0948728695511818,
+      "learning_rate": 2.401486895792135e-05,
+      "loss": 9.7519,
+      "step": 40350
+    },
+    {
+      "epoch": 0.20155309745561686,
+      "grad_norm": 0.09901313483715057,
+      "learning_rate": 2.40133670429798e-05,
+      "loss": 9.7589,
+      "step": 40360
+    },
+    {
+      "epoch": 0.2016030362805563,
+      "grad_norm": 0.09512390196323395,
+      "learning_rate": 2.401186512803825e-05,
+      "loss": 9.7564,
+      "step": 40370
+    },
+    {
+      "epoch": 0.20165297510549576,
+      "grad_norm": 0.08876697719097137,
+      "learning_rate": 2.40103632130967e-05,
+      "loss": 9.7596,
+      "step": 40380
+    },
+    {
+      "epoch": 0.2017029139304352,
+      "grad_norm": 0.09231546521186829,
+      "learning_rate": 2.4008861298155147e-05,
+      "loss": 9.7513,
+      "step": 40390
+    },
+    {
+      "epoch": 0.20175285275537466,
+      "grad_norm": 0.08933798968791962,
+      "learning_rate": 2.40073593832136e-05,
+      "loss": 9.7478,
+      "step": 40400
+    },
+    {
+      "epoch": 0.2018027915803141,
+      "grad_norm": 0.09653082489967346,
+      "learning_rate": 2.4005857468272047e-05,
+      "loss": 9.751,
+      "step": 40410
+    },
+    {
+      "epoch": 0.20185273040525356,
+      "grad_norm": 0.09329882264137268,
+      "learning_rate": 2.4004355553330497e-05,
+      "loss": 9.7608,
+      "step": 40420
+    },
+    {
+      "epoch": 0.201902669230193,
+      "grad_norm": 0.09065824002027512,
+      "learning_rate": 2.4002853638388947e-05,
+      "loss": 9.7584,
+      "step": 40430
+    },
+    {
+      "epoch": 0.20195260805513246,
+      "grad_norm": 0.09478091448545456,
+      "learning_rate": 2.4001351723447394e-05,
+      "loss": 9.7633,
+      "step": 40440
+    },
+    {
+      "epoch": 0.2020025468800719,
+      "grad_norm": 0.09942101687192917,
+      "learning_rate": 2.3999849808505848e-05,
+      "loss": 9.7596,
+      "step": 40450
+    },
+    {
+      "epoch": 0.20205248570501136,
+      "grad_norm": 0.0919698178768158,
+      "learning_rate": 2.3998347893564294e-05,
+      "loss": 9.7657,
+      "step": 40460
+    },
+    {
+      "epoch": 0.2021024245299508,
+      "grad_norm": 0.08957584947347641,
+      "learning_rate": 2.3996845978622745e-05,
+      "loss": 9.7507,
+      "step": 40470
+    },
+    {
+      "epoch": 0.20215236335489026,
+      "grad_norm": 0.09440477192401886,
+      "learning_rate": 2.3995344063681195e-05,
+      "loss": 9.7558,
+      "step": 40480
+    },
+    {
+      "epoch": 0.2022023021798297,
+      "grad_norm": 0.0963861346244812,
+      "learning_rate": 2.399384214873964e-05,
+      "loss": 9.7496,
+      "step": 40490
+    },
+    {
+      "epoch": 0.20225224100476916,
+      "grad_norm": 0.09470155835151672,
+      "learning_rate": 2.3992340233798095e-05,
+      "loss": 9.7483,
+      "step": 40500
+    },
+    {
+      "epoch": 0.2023021798297086,
+      "grad_norm": 0.09395986050367355,
+      "learning_rate": 2.3990838318856542e-05,
+      "loss": 9.7426,
+      "step": 40510
+    },
+    {
+      "epoch": 0.20235211865464806,
+      "grad_norm": 0.09907002747058868,
+      "learning_rate": 2.3989336403914992e-05,
+      "loss": 9.7482,
+      "step": 40520
+    },
+    {
+      "epoch": 0.2024020574795875,
+      "grad_norm": 0.09406791627407074,
+      "learning_rate": 2.3987834488973442e-05,
+      "loss": 9.7551,
+      "step": 40530
+    },
+    {
+      "epoch": 0.20245199630452695,
+      "grad_norm": 0.09641828387975693,
+      "learning_rate": 2.398633257403189e-05,
+      "loss": 9.756,
+      "step": 40540
+    },
+    {
+      "epoch": 0.2025019351294664,
+      "grad_norm": 0.08803495764732361,
+      "learning_rate": 2.3984830659090343e-05,
+      "loss": 9.7517,
+      "step": 40550
+    },
+    {
+      "epoch": 0.20255187395440585,
+      "grad_norm": 0.089680016040802,
+      "learning_rate": 2.398332874414879e-05,
+      "loss": 9.7548,
+      "step": 40560
+    },
+    {
+      "epoch": 0.2026018127793453,
+      "grad_norm": 0.09483745694160461,
+      "learning_rate": 2.398182682920724e-05,
+      "loss": 9.7505,
+      "step": 40570
+    },
+    {
+      "epoch": 0.20265175160428475,
+      "grad_norm": 0.09230205416679382,
+      "learning_rate": 2.398032491426569e-05,
+      "loss": 9.7526,
+      "step": 40580
+    },
+    {
+      "epoch": 0.2027016904292242,
+      "grad_norm": 0.0983130931854248,
+      "learning_rate": 2.3978822999324137e-05,
+      "loss": 9.7463,
+      "step": 40590
+    },
+    {
+      "epoch": 0.20275162925416365,
+      "grad_norm": 0.0941585898399353,
+      "learning_rate": 2.397732108438259e-05,
+      "loss": 9.7575,
+      "step": 40600
+    },
+    {
+      "epoch": 0.2028015680791031,
+      "grad_norm": 0.09447518736124039,
+      "learning_rate": 2.3975819169441037e-05,
+      "loss": 9.7447,
+      "step": 40610
+    },
+    {
+      "epoch": 0.20285150690404255,
+      "grad_norm": 0.09139948338270187,
+      "learning_rate": 2.3974317254499487e-05,
+      "loss": 9.7499,
+      "step": 40620
+    },
+    {
+      "epoch": 0.202901445728982,
+      "grad_norm": 0.09922108799219131,
+      "learning_rate": 2.3972815339557937e-05,
+      "loss": 9.7463,
+      "step": 40630
+    },
+    {
+      "epoch": 0.20295138455392145,
+      "grad_norm": 0.09205441176891327,
+      "learning_rate": 2.3971313424616384e-05,
+      "loss": 9.7516,
+      "step": 40640
+    },
+    {
+      "epoch": 0.2030013233788609,
+      "grad_norm": 0.097032330930233,
+      "learning_rate": 2.3969811509674838e-05,
+      "loss": 9.7507,
+      "step": 40650
+    },
+    {
+      "epoch": 0.20305126220380035,
+      "grad_norm": 0.09273160994052887,
+      "learning_rate": 2.3968309594733284e-05,
+      "loss": 9.7485,
+      "step": 40660
+    },
+    {
+      "epoch": 0.2031012010287398,
+      "grad_norm": 0.09222143143415451,
+      "learning_rate": 2.3966807679791735e-05,
+      "loss": 9.7494,
+      "step": 40670
+    },
+    {
+      "epoch": 0.20315113985367925,
+      "grad_norm": 0.10249687731266022,
+      "learning_rate": 2.3965305764850185e-05,
+      "loss": 9.749,
+      "step": 40680
+    },
+    {
+      "epoch": 0.2032010786786187,
+      "grad_norm": 0.09170138835906982,
+      "learning_rate": 2.396380384990863e-05,
+      "loss": 9.7538,
+      "step": 40690
+    },
+    {
+      "epoch": 0.20325101750355815,
+      "grad_norm": 0.09489130973815918,
+      "learning_rate": 2.3962301934967085e-05,
+      "loss": 9.7539,
+      "step": 40700
+    },
+    {
+      "epoch": 0.2033009563284976,
+      "grad_norm": 0.09625539183616638,
+      "learning_rate": 2.3960800020025532e-05,
+      "loss": 9.7509,
+      "step": 40710
+    },
+    {
+      "epoch": 0.20335089515343704,
+      "grad_norm": 0.09620354324579239,
+      "learning_rate": 2.3959298105083986e-05,
+      "loss": 9.7504,
+      "step": 40720
+    },
+    {
+      "epoch": 0.2034008339783765,
+      "grad_norm": 0.10117631405591965,
+      "learning_rate": 2.3957796190142432e-05,
+      "loss": 9.7386,
+      "step": 40730
+    },
+    {
+      "epoch": 0.20345077280331594,
+      "grad_norm": 0.09158354997634888,
+      "learning_rate": 2.395629427520088e-05,
+      "loss": 9.7471,
+      "step": 40740
+    },
+    {
+      "epoch": 0.2035007116282554,
+      "grad_norm": 0.09191151708364487,
+      "learning_rate": 2.3954792360259333e-05,
+      "loss": 9.7529,
+      "step": 40750
+    },
+    {
+      "epoch": 0.20355065045319484,
+      "grad_norm": 0.09399069845676422,
+      "learning_rate": 2.395329044531778e-05,
+      "loss": 9.745,
+      "step": 40760
+    },
+    {
+      "epoch": 0.2036005892781343,
+      "grad_norm": 0.09574571996927261,
+      "learning_rate": 2.3951788530376233e-05,
+      "loss": 9.7506,
+      "step": 40770
+    },
+    {
+      "epoch": 0.20365052810307374,
+      "grad_norm": 0.09355718642473221,
+      "learning_rate": 2.395028661543468e-05,
+      "loss": 9.7426,
+      "step": 40780
+    },
+    {
+      "epoch": 0.2037004669280132,
+      "grad_norm": 0.09288044273853302,
+      "learning_rate": 2.3948784700493127e-05,
+      "loss": 9.752,
+      "step": 40790
+    },
+    {
+      "epoch": 0.20375040575295264,
+      "grad_norm": 0.09448196738958359,
+      "learning_rate": 2.394728278555158e-05,
+      "loss": 9.7449,
+      "step": 40800
+    },
+    {
+      "epoch": 0.2038003445778921,
+      "grad_norm": 0.09567010402679443,
+      "learning_rate": 2.3945780870610027e-05,
+      "loss": 9.7417,
+      "step": 40810
+    },
+    {
+      "epoch": 0.20385028340283154,
+      "grad_norm": 0.08981973677873611,
+      "learning_rate": 2.394427895566848e-05,
+      "loss": 9.7473,
+      "step": 40820
+    },
+    {
+      "epoch": 0.203900222227771,
+      "grad_norm": 0.09741527587175369,
+      "learning_rate": 2.3942777040726927e-05,
+      "loss": 9.7482,
+      "step": 40830
+    },
+    {
+      "epoch": 0.20395016105271044,
+      "grad_norm": 0.09928309917449951,
+      "learning_rate": 2.3941275125785374e-05,
+      "loss": 9.7452,
+      "step": 40840
+    },
+    {
+      "epoch": 0.2040000998776499,
+      "grad_norm": 0.09600803256034851,
+      "learning_rate": 2.3939773210843828e-05,
+      "loss": 9.7513,
+      "step": 40850
+    },
+    {
+      "epoch": 0.20405003870258934,
+      "grad_norm": 0.09777117520570755,
+      "learning_rate": 2.3938271295902274e-05,
+      "loss": 9.752,
+      "step": 40860
+    },
+    {
+      "epoch": 0.2040999775275288,
+      "grad_norm": 0.09423068910837173,
+      "learning_rate": 2.3936769380960728e-05,
+      "loss": 9.7462,
+      "step": 40870
+    },
+    {
+      "epoch": 0.20414991635246824,
+      "grad_norm": 0.09674641489982605,
+      "learning_rate": 2.3935267466019175e-05,
+      "loss": 9.745,
+      "step": 40880
+    },
+    {
+      "epoch": 0.20419985517740769,
+      "grad_norm": 0.095578134059906,
+      "learning_rate": 2.393376555107762e-05,
+      "loss": 9.7457,
+      "step": 40890
+    },
+    {
+      "epoch": 0.20424979400234713,
+      "grad_norm": 0.09196728467941284,
+      "learning_rate": 2.3932263636136075e-05,
+      "loss": 9.7413,
+      "step": 40900
+    },
+    {
+      "epoch": 0.20429973282728658,
+      "grad_norm": 0.09680663794279099,
+      "learning_rate": 2.3930761721194522e-05,
+      "loss": 9.7394,
+      "step": 40910
+    },
+    {
+      "epoch": 0.20434967165222603,
+      "grad_norm": 0.10243181139230728,
+      "learning_rate": 2.3929259806252976e-05,
+      "loss": 9.7504,
+      "step": 40920
+    },
+    {
+      "epoch": 0.20439961047716548,
+      "grad_norm": 0.09534817934036255,
+      "learning_rate": 2.3927757891311422e-05,
+      "loss": 9.7527,
+      "step": 40930
+    },
+    {
+      "epoch": 0.20444954930210493,
+      "grad_norm": 0.10091328620910645,
+      "learning_rate": 2.392625597636987e-05,
+      "loss": 9.7457,
+      "step": 40940
+    },
+    {
+      "epoch": 0.20449948812704438,
+      "grad_norm": 0.09642726182937622,
+      "learning_rate": 2.3924754061428323e-05,
+      "loss": 9.7483,
+      "step": 40950
+    },
+    {
+      "epoch": 0.20454942695198383,
+      "grad_norm": 0.09273668378591537,
+      "learning_rate": 2.392325214648677e-05,
+      "loss": 9.7487,
+      "step": 40960
+    },
+    {
+      "epoch": 0.20459936577692328,
+      "grad_norm": 0.09491418302059174,
+      "learning_rate": 2.3921750231545223e-05,
+      "loss": 9.7384,
+      "step": 40970
+    },
+    {
+      "epoch": 0.20464930460186273,
+      "grad_norm": 0.09379623830318451,
+      "learning_rate": 2.392024831660367e-05,
+      "loss": 9.7397,
+      "step": 40980
+    },
+    {
+      "epoch": 0.20469924342680218,
+      "grad_norm": 0.09489590674638748,
+      "learning_rate": 2.3918746401662117e-05,
+      "loss": 9.738,
+      "step": 40990
+    },
+    {
+      "epoch": 0.20474918225174163,
+      "grad_norm": 0.09257600456476212,
+      "learning_rate": 2.391724448672057e-05,
+      "loss": 9.743,
+      "step": 41000
+    },
+    {
+      "epoch": 0.20479912107668108,
+      "grad_norm": 0.09949919581413269,
+      "learning_rate": 2.3915742571779017e-05,
+      "loss": 9.7409,
+      "step": 41010
+    },
+    {
+      "epoch": 0.20484905990162053,
+      "grad_norm": 0.09563277661800385,
+      "learning_rate": 2.391424065683747e-05,
+      "loss": 9.738,
+      "step": 41020
+    },
+    {
+      "epoch": 0.20489899872655998,
+      "grad_norm": 0.09452064335346222,
+      "learning_rate": 2.3912738741895917e-05,
+      "loss": 9.736,
+      "step": 41030
+    },
+    {
+      "epoch": 0.2049489375514994,
+      "grad_norm": 0.09839780628681183,
+      "learning_rate": 2.3911236826954368e-05,
+      "loss": 9.7439,
+      "step": 41040
+    },
+    {
+      "epoch": 0.20499887637643885,
+      "grad_norm": 0.0967588871717453,
+      "learning_rate": 2.3909734912012818e-05,
+      "loss": 9.7426,
+      "step": 41050
+    },
+    {
+      "epoch": 0.2050488152013783,
+      "grad_norm": 0.09972970187664032,
+      "learning_rate": 2.3908232997071264e-05,
+      "loss": 9.7462,
+      "step": 41060
+    },
+    {
+      "epoch": 0.20509875402631775,
+      "grad_norm": 0.09793765842914581,
+      "learning_rate": 2.3906731082129718e-05,
+      "loss": 9.7425,
+      "step": 41070
+    },
+    {
+      "epoch": 0.2051486928512572,
+      "grad_norm": 0.09164223074913025,
+      "learning_rate": 2.3905229167188165e-05,
+      "loss": 9.7346,
+      "step": 41080
+    },
+    {
+      "epoch": 0.20519863167619665,
+      "grad_norm": 0.09696740657091141,
+      "learning_rate": 2.3903727252246615e-05,
+      "loss": 9.7373,
+      "step": 41090
+    },
+    {
+      "epoch": 0.2052485705011361,
+      "grad_norm": 0.09416086226701736,
+      "learning_rate": 2.3902225337305065e-05,
+      "loss": 9.74,
+      "step": 41100
+    },
+    {
+      "epoch": 0.20529850932607555,
+      "grad_norm": 0.09924998879432678,
+      "learning_rate": 2.3900723422363512e-05,
+      "loss": 9.7414,
+      "step": 41110
+    },
+    {
+      "epoch": 0.205348448151015,
+      "grad_norm": 0.08897435665130615,
+      "learning_rate": 2.3899221507421966e-05,
+      "loss": 9.7388,
+      "step": 41120
+    },
+    {
+      "epoch": 0.20539838697595444,
+      "grad_norm": 0.09157969057559967,
+      "learning_rate": 2.3897719592480412e-05,
+      "loss": 9.7346,
+      "step": 41130
+    },
+    {
+      "epoch": 0.2054483258008939,
+      "grad_norm": 0.09318001568317413,
+      "learning_rate": 2.3896217677538863e-05,
+      "loss": 9.7399,
+      "step": 41140
+    },
+    {
+      "epoch": 0.20549826462583334,
+      "grad_norm": 0.09765854477882385,
+      "learning_rate": 2.3894715762597313e-05,
+      "loss": 9.7369,
+      "step": 41150
+    },
+    {
+      "epoch": 0.2055482034507728,
+      "grad_norm": 0.09092031419277191,
+      "learning_rate": 2.389321384765576e-05,
+      "loss": 9.7477,
+      "step": 41160
+    },
+    {
+      "epoch": 0.20559814227571224,
+      "grad_norm": 0.09183992445468903,
+      "learning_rate": 2.3891711932714213e-05,
+      "loss": 9.741,
+      "step": 41170
+    },
+    {
+      "epoch": 0.2056480811006517,
+      "grad_norm": 0.0956728532910347,
+      "learning_rate": 2.389021001777266e-05,
+      "loss": 9.7374,
+      "step": 41180
+    },
+    {
+      "epoch": 0.20569801992559114,
+      "grad_norm": 0.09464164078235626,
+      "learning_rate": 2.388870810283111e-05,
+      "loss": 9.7397,
+      "step": 41190
+    },
+    {
+      "epoch": 0.2057479587505306,
+      "grad_norm": 0.09706345945596695,
+      "learning_rate": 2.388720618788956e-05,
+      "loss": 9.7364,
+      "step": 41200
+    },
+    {
+      "epoch": 0.20579789757547004,
+      "grad_norm": 0.09342235326766968,
+      "learning_rate": 2.3885704272948007e-05,
+      "loss": 9.7431,
+      "step": 41210
+    },
+    {
+      "epoch": 0.2058478364004095,
+      "grad_norm": 0.09735559672117233,
+      "learning_rate": 2.388420235800646e-05,
+      "loss": 9.737,
+      "step": 41220
+    },
+    {
+      "epoch": 0.20589777522534894,
+      "grad_norm": 0.0972183495759964,
+      "learning_rate": 2.3882700443064907e-05,
+      "loss": 9.735,
+      "step": 41230
+    },
+    {
+      "epoch": 0.2059477140502884,
+      "grad_norm": 0.09024586528539658,
+      "learning_rate": 2.3881198528123358e-05,
+      "loss": 9.7344,
+      "step": 41240
+    },
+    {
+      "epoch": 0.20599765287522784,
+      "grad_norm": 0.09795460104942322,
+      "learning_rate": 2.3879696613181808e-05,
+      "loss": 9.7426,
+      "step": 41250
+    },
+    {
+      "epoch": 0.2060475917001673,
+      "grad_norm": 0.10017476975917816,
+      "learning_rate": 2.3878194698240255e-05,
+      "loss": 9.7292,
+      "step": 41260
+    },
+    {
+      "epoch": 0.20609753052510674,
+      "grad_norm": 0.09363752603530884,
+      "learning_rate": 2.3876692783298708e-05,
+      "loss": 9.7424,
+      "step": 41270
+    },
+    {
+      "epoch": 0.2061474693500462,
+      "grad_norm": 0.09633088111877441,
+      "learning_rate": 2.3875190868357155e-05,
+      "loss": 9.7319,
+      "step": 41280
+    },
+    {
+      "epoch": 0.20619740817498564,
+      "grad_norm": 0.09266993403434753,
+      "learning_rate": 2.3873688953415605e-05,
+      "loss": 9.732,
+      "step": 41290
+    },
+    {
+      "epoch": 0.20624734699992509,
+      "grad_norm": 0.10170450061559677,
+      "learning_rate": 2.3872187038474055e-05,
+      "loss": 9.7402,
+      "step": 41300
+    },
+    {
+      "epoch": 0.20629728582486453,
+      "grad_norm": 0.093755342066288,
+      "learning_rate": 2.3870685123532502e-05,
+      "loss": 9.7369,
+      "step": 41310
+    },
+    {
+      "epoch": 0.20634722464980398,
+      "grad_norm": 0.09129472076892853,
+      "learning_rate": 2.3869183208590956e-05,
+      "loss": 9.7356,
+      "step": 41320
+    },
+    {
+      "epoch": 0.20639716347474343,
+      "grad_norm": 0.09476242959499359,
+      "learning_rate": 2.3867681293649402e-05,
+      "loss": 9.7394,
+      "step": 41330
+    },
+    {
+      "epoch": 0.20644710229968288,
+      "grad_norm": 0.09445454180240631,
+      "learning_rate": 2.3866179378707853e-05,
+      "loss": 9.7293,
+      "step": 41340
+    },
+    {
+      "epoch": 0.20649704112462233,
+      "grad_norm": 0.09727781265974045,
+      "learning_rate": 2.3864677463766303e-05,
+      "loss": 9.7333,
+      "step": 41350
+    },
+    {
+      "epoch": 0.20654697994956178,
+      "grad_norm": 0.093324214220047,
+      "learning_rate": 2.386317554882475e-05,
+      "loss": 9.7324,
+      "step": 41360
+    },
+    {
+      "epoch": 0.20659691877450123,
+      "grad_norm": 0.09454476833343506,
+      "learning_rate": 2.3861673633883203e-05,
+      "loss": 9.7345,
+      "step": 41370
+    },
+    {
+      "epoch": 0.20664685759944068,
+      "grad_norm": 0.09696672111749649,
+      "learning_rate": 2.386017171894165e-05,
+      "loss": 9.7345,
+      "step": 41380
+    },
+    {
+      "epoch": 0.20669679642438013,
+      "grad_norm": 0.09323279559612274,
+      "learning_rate": 2.38586698040001e-05,
+      "loss": 9.7358,
+      "step": 41390
+    },
+    {
+      "epoch": 0.20674673524931958,
+      "grad_norm": 0.09088480472564697,
+      "learning_rate": 2.385716788905855e-05,
+      "loss": 9.7327,
+      "step": 41400
+    },
+    {
+      "epoch": 0.20679667407425903,
+      "grad_norm": 0.08868908882141113,
+      "learning_rate": 2.3855665974117e-05,
+      "loss": 9.7327,
+      "step": 41410
+    },
+    {
+      "epoch": 0.20684661289919848,
+      "grad_norm": 0.0900205671787262,
+      "learning_rate": 2.385416405917545e-05,
+      "loss": 9.7318,
+      "step": 41420
+    },
+    {
+      "epoch": 0.20689655172413793,
+      "grad_norm": 0.0981634333729744,
+      "learning_rate": 2.3852662144233897e-05,
+      "loss": 9.7366,
+      "step": 41430
+    },
+    {
+      "epoch": 0.20694649054907738,
+      "grad_norm": 0.10074867308139801,
+      "learning_rate": 2.3851160229292348e-05,
+      "loss": 9.729,
+      "step": 41440
+    },
+    {
+      "epoch": 0.20699642937401683,
+      "grad_norm": 0.09191495180130005,
+      "learning_rate": 2.3849658314350798e-05,
+      "loss": 9.7327,
+      "step": 41450
+    },
+    {
+      "epoch": 0.20704636819895628,
+      "grad_norm": 0.09636077284812927,
+      "learning_rate": 2.3848156399409248e-05,
+      "loss": 9.7327,
+      "step": 41460
+    },
+    {
+      "epoch": 0.20709630702389573,
+      "grad_norm": 0.09494554251432419,
+      "learning_rate": 2.3846654484467698e-05,
+      "loss": 9.7348,
+      "step": 41470
+    },
+    {
+      "epoch": 0.20714624584883518,
+      "grad_norm": 0.09379441291093826,
+      "learning_rate": 2.3845152569526145e-05,
+      "loss": 9.7329,
+      "step": 41480
+    },
+    {
+      "epoch": 0.20719618467377462,
+      "grad_norm": 0.09855134785175323,
+      "learning_rate": 2.3843650654584595e-05,
+      "loss": 9.7449,
+      "step": 41490
+    },
+    {
+      "epoch": 0.20724612349871407,
+      "grad_norm": 0.09271756559610367,
+      "learning_rate": 2.3842148739643045e-05,
+      "loss": 9.7368,
+      "step": 41500
+    },
+    {
+      "epoch": 0.20729606232365352,
+      "grad_norm": 0.09612946957349777,
+      "learning_rate": 2.3840646824701495e-05,
+      "loss": 9.7314,
+      "step": 41510
+    },
+    {
+      "epoch": 0.20734600114859297,
+      "grad_norm": 0.09541863203048706,
+      "learning_rate": 2.3839144909759946e-05,
+      "loss": 9.7345,
+      "step": 41520
+    },
+    {
+      "epoch": 0.20739593997353242,
+      "grad_norm": 0.09798285365104675,
+      "learning_rate": 2.3837642994818392e-05,
+      "loss": 9.7369,
+      "step": 41530
+    },
+    {
+      "epoch": 0.20744587879847187,
+      "grad_norm": 0.09581144899129868,
+      "learning_rate": 2.3836141079876843e-05,
+      "loss": 9.7313,
+      "step": 41540
+    },
+    {
+      "epoch": 0.20749581762341132,
+      "grad_norm": 0.09617334604263306,
+      "learning_rate": 2.3834639164935293e-05,
+      "loss": 9.7242,
+      "step": 41550
+    },
+    {
+      "epoch": 0.20754575644835077,
+      "grad_norm": 0.09934971481561661,
+      "learning_rate": 2.3833137249993743e-05,
+      "loss": 9.7257,
+      "step": 41560
+    },
+    {
+      "epoch": 0.20759569527329022,
+      "grad_norm": 0.09321165084838867,
+      "learning_rate": 2.3831635335052193e-05,
+      "loss": 9.7286,
+      "step": 41570
+    },
+    {
+      "epoch": 0.20764563409822967,
+      "grad_norm": 0.09776625037193298,
+      "learning_rate": 2.383013342011064e-05,
+      "loss": 9.7297,
+      "step": 41580
+    },
+    {
+      "epoch": 0.20769557292316912,
+      "grad_norm": 0.09788148105144501,
+      "learning_rate": 2.382863150516909e-05,
+      "loss": 9.7298,
+      "step": 41590
+    },
+    {
+      "epoch": 0.20774551174810857,
+      "grad_norm": 0.09117887914180756,
+      "learning_rate": 2.382712959022754e-05,
+      "loss": 9.7353,
+      "step": 41600
+    },
+    {
+      "epoch": 0.20779545057304802,
+      "grad_norm": 0.09594250470399857,
+      "learning_rate": 2.382562767528599e-05,
+      "loss": 9.7353,
+      "step": 41610
+    },
+    {
+      "epoch": 0.20784538939798747,
+      "grad_norm": 0.08893455564975739,
+      "learning_rate": 2.382412576034444e-05,
+      "loss": 9.7268,
+      "step": 41620
+    },
+    {
+      "epoch": 0.20789532822292692,
+      "grad_norm": 0.0987037941813469,
+      "learning_rate": 2.3822623845402887e-05,
+      "loss": 9.7287,
+      "step": 41630
+    },
+    {
+      "epoch": 0.20794526704786637,
+      "grad_norm": 0.09531259536743164,
+      "learning_rate": 2.3821121930461338e-05,
+      "loss": 9.7335,
+      "step": 41640
+    },
+    {
+      "epoch": 0.20799520587280582,
+      "grad_norm": 0.09097182750701904,
+      "learning_rate": 2.3819620015519788e-05,
+      "loss": 9.7319,
+      "step": 41650
+    },
+    {
+      "epoch": 0.20804514469774527,
+      "grad_norm": 0.09137894958257675,
+      "learning_rate": 2.3818118100578238e-05,
+      "loss": 9.7361,
+      "step": 41660
+    },
+    {
+      "epoch": 0.20809508352268472,
+      "grad_norm": 0.09314315766096115,
+      "learning_rate": 2.3816616185636688e-05,
+      "loss": 9.727,
+      "step": 41670
+    },
+    {
+      "epoch": 0.20814502234762416,
+      "grad_norm": 0.09129998832941055,
+      "learning_rate": 2.3815114270695135e-05,
+      "loss": 9.7235,
+      "step": 41680
+    },
+    {
+      "epoch": 0.20819496117256361,
+      "grad_norm": 0.09588906913995743,
+      "learning_rate": 2.3813612355753585e-05,
+      "loss": 9.7317,
+      "step": 41690
+    },
+    {
+      "epoch": 0.20824489999750306,
+      "grad_norm": 0.089946448802948,
+      "learning_rate": 2.3812110440812035e-05,
+      "loss": 9.7317,
+      "step": 41700
+    },
+    {
+      "epoch": 0.2082948388224425,
+      "grad_norm": 0.09250712394714355,
+      "learning_rate": 2.3810608525870485e-05,
+      "loss": 9.7321,
+      "step": 41710
+    },
+    {
+      "epoch": 0.20834477764738196,
+      "grad_norm": 0.09257816523313522,
+      "learning_rate": 2.3809106610928936e-05,
+      "loss": 9.7294,
+      "step": 41720
+    },
+    {
+      "epoch": 0.2083947164723214,
+      "grad_norm": 0.09143438935279846,
+      "learning_rate": 2.3807604695987386e-05,
+      "loss": 9.731,
+      "step": 41730
+    },
+    {
+      "epoch": 0.20844465529726086,
+      "grad_norm": 0.09791240841150284,
+      "learning_rate": 2.3806102781045833e-05,
+      "loss": 9.7234,
+      "step": 41740
+    },
+    {
+      "epoch": 0.2084945941222003,
+      "grad_norm": 0.10063060373067856,
+      "learning_rate": 2.3804600866104283e-05,
+      "loss": 9.7279,
+      "step": 41750
+    },
+    {
+      "epoch": 0.20854453294713976,
+      "grad_norm": 0.09549122303724289,
+      "learning_rate": 2.3803098951162733e-05,
+      "loss": 9.7233,
+      "step": 41760
+    },
+    {
+      "epoch": 0.2085944717720792,
+      "grad_norm": 0.09345372021198273,
+      "learning_rate": 2.3801597036221183e-05,
+      "loss": 9.7282,
+      "step": 41770
+    },
+    {
+      "epoch": 0.20864441059701866,
+      "grad_norm": 0.0949343666434288,
+      "learning_rate": 2.3800095121279633e-05,
+      "loss": 9.721,
+      "step": 41780
+    },
+    {
+      "epoch": 0.2086943494219581,
+      "grad_norm": 0.09225095063447952,
+      "learning_rate": 2.3798593206338083e-05,
+      "loss": 9.7369,
+      "step": 41790
+    },
+    {
+      "epoch": 0.20874428824689756,
+      "grad_norm": 0.09359176456928253,
+      "learning_rate": 2.379709129139653e-05,
+      "loss": 9.7298,
+      "step": 41800
+    },
+    {
+      "epoch": 0.208794227071837,
+      "grad_norm": 0.09750591963529587,
+      "learning_rate": 2.379558937645498e-05,
+      "loss": 9.7263,
+      "step": 41810
+    },
+    {
+      "epoch": 0.20884416589677646,
+      "grad_norm": 0.0969536155462265,
+      "learning_rate": 2.379408746151343e-05,
+      "loss": 9.7193,
+      "step": 41820
+    },
+    {
+      "epoch": 0.2088941047217159,
+      "grad_norm": 0.09383013099431992,
+      "learning_rate": 2.379258554657188e-05,
+      "loss": 9.7168,
+      "step": 41830
+    },
+    {
+      "epoch": 0.20894404354665536,
+      "grad_norm": 0.09686887264251709,
+      "learning_rate": 2.379108363163033e-05,
+      "loss": 9.7198,
+      "step": 41840
+    },
+    {
+      "epoch": 0.2089939823715948,
+      "grad_norm": 0.09708838164806366,
+      "learning_rate": 2.3789581716688778e-05,
+      "loss": 9.7315,
+      "step": 41850
+    },
+    {
+      "epoch": 0.20904392119653425,
+      "grad_norm": 0.09624648839235306,
+      "learning_rate": 2.3788079801747228e-05,
+      "loss": 9.7148,
+      "step": 41860
+    },
+    {
+      "epoch": 0.2090938600214737,
+      "grad_norm": 0.09800230711698532,
+      "learning_rate": 2.3786577886805678e-05,
+      "loss": 9.7328,
+      "step": 41870
+    },
+    {
+      "epoch": 0.20914379884641315,
+      "grad_norm": 0.09165570139884949,
+      "learning_rate": 2.3785075971864128e-05,
+      "loss": 9.7219,
+      "step": 41880
+    },
+    {
+      "epoch": 0.2091937376713526,
+      "grad_norm": 0.09577257931232452,
+      "learning_rate": 2.378357405692258e-05,
+      "loss": 9.7289,
+      "step": 41890
+    },
+    {
+      "epoch": 0.20924367649629205,
+      "grad_norm": 0.09999013692140579,
+      "learning_rate": 2.3782072141981025e-05,
+      "loss": 9.7178,
+      "step": 41900
+    },
+    {
+      "epoch": 0.2092936153212315,
+      "grad_norm": 0.09558932483196259,
+      "learning_rate": 2.3780570227039475e-05,
+      "loss": 9.7261,
+      "step": 41910
+    },
+    {
+      "epoch": 0.20934355414617095,
+      "grad_norm": 0.09391254931688309,
+      "learning_rate": 2.3779068312097926e-05,
+      "loss": 9.7244,
+      "step": 41920
+    },
+    {
+      "epoch": 0.2093934929711104,
+      "grad_norm": 0.09899161756038666,
+      "learning_rate": 2.3777566397156376e-05,
+      "loss": 9.7294,
+      "step": 41930
+    },
+    {
+      "epoch": 0.20944343179604985,
+      "grad_norm": 0.09575141966342926,
+      "learning_rate": 2.3776064482214826e-05,
+      "loss": 9.72,
+      "step": 41940
+    },
+    {
+      "epoch": 0.2094933706209893,
+      "grad_norm": 0.10070329904556274,
+      "learning_rate": 2.3774562567273273e-05,
+      "loss": 9.7279,
+      "step": 41950
+    },
+    {
+      "epoch": 0.20954330944592875,
+      "grad_norm": 0.09497936069965363,
+      "learning_rate": 2.3773060652331723e-05,
+      "loss": 9.7317,
+      "step": 41960
+    },
+    {
+      "epoch": 0.2095932482708682,
+      "grad_norm": 0.09233599156141281,
+      "learning_rate": 2.3771558737390173e-05,
+      "loss": 9.7265,
+      "step": 41970
+    },
+    {
+      "epoch": 0.20964318709580765,
+      "grad_norm": 0.08788701891899109,
+      "learning_rate": 2.3770056822448623e-05,
+      "loss": 9.7191,
+      "step": 41980
+    },
+    {
+      "epoch": 0.2096931259207471,
+      "grad_norm": 0.09631108492612839,
+      "learning_rate": 2.3768554907507073e-05,
+      "loss": 9.7255,
+      "step": 41990
+    },
+    {
+      "epoch": 0.20974306474568655,
+      "grad_norm": 0.09688840806484222,
+      "learning_rate": 2.376705299256552e-05,
+      "loss": 9.7247,
+      "step": 42000
+    },
+    {
+      "epoch": 0.209793003570626,
+      "grad_norm": 0.09418191015720367,
+      "learning_rate": 2.376555107762397e-05,
+      "loss": 9.7267,
+      "step": 42010
+    },
+    {
+      "epoch": 0.20984294239556545,
+      "grad_norm": 0.0922936499118805,
+      "learning_rate": 2.376404916268242e-05,
+      "loss": 9.7249,
+      "step": 42020
+    },
+    {
+      "epoch": 0.20989288122050487,
+      "grad_norm": 0.09569290280342102,
+      "learning_rate": 2.376254724774087e-05,
+      "loss": 9.723,
+      "step": 42030
+    },
+    {
+      "epoch": 0.20994282004544432,
+      "grad_norm": 0.10295400768518448,
+      "learning_rate": 2.376104533279932e-05,
+      "loss": 9.7236,
+      "step": 42040
+    },
+    {
+      "epoch": 0.20999275887038377,
+      "grad_norm": 0.09752771258354187,
+      "learning_rate": 2.375954341785777e-05,
+      "loss": 9.7179,
+      "step": 42050
+    },
+    {
+      "epoch": 0.21004269769532322,
+      "grad_norm": 0.09303586930036545,
+      "learning_rate": 2.3758041502916218e-05,
+      "loss": 9.7193,
+      "step": 42060
+    },
+    {
+      "epoch": 0.21009263652026267,
+      "grad_norm": 0.10309559851884842,
+      "learning_rate": 2.3756539587974668e-05,
+      "loss": 9.7216,
+      "step": 42070
+    },
+    {
+      "epoch": 0.21014257534520212,
+      "grad_norm": 0.09949391335248947,
+      "learning_rate": 2.375503767303312e-05,
+      "loss": 9.7174,
+      "step": 42080
+    },
+    {
+      "epoch": 0.21019251417014156,
+      "grad_norm": 0.0931687206029892,
+      "learning_rate": 2.375353575809157e-05,
+      "loss": 9.7203,
+      "step": 42090
+    },
+    {
+      "epoch": 0.21024245299508101,
+      "grad_norm": 0.09643073379993439,
+      "learning_rate": 2.375203384315002e-05,
+      "loss": 9.7225,
+      "step": 42100
+    },
+    {
+      "epoch": 0.21029239182002046,
+      "grad_norm": 0.09048034995794296,
+      "learning_rate": 2.3750531928208465e-05,
+      "loss": 9.7207,
+      "step": 42110
+    },
+    {
+      "epoch": 0.2103423306449599,
+      "grad_norm": 0.10079729557037354,
+      "learning_rate": 2.3749030013266916e-05,
+      "loss": 9.7121,
+      "step": 42120
+    },
+    {
+      "epoch": 0.21039226946989936,
+      "grad_norm": 0.09450411796569824,
+      "learning_rate": 2.3747528098325366e-05,
+      "loss": 9.7202,
+      "step": 42130
+    },
+    {
+      "epoch": 0.2104422082948388,
+      "grad_norm": 0.09755988419055939,
+      "learning_rate": 2.3746026183383816e-05,
+      "loss": 9.7198,
+      "step": 42140
+    },
+    {
+      "epoch": 0.21049214711977826,
+      "grad_norm": 0.09404526650905609,
+      "learning_rate": 2.3744524268442266e-05,
+      "loss": 9.7194,
+      "step": 42150
+    },
+    {
+      "epoch": 0.2105420859447177,
+      "grad_norm": 0.09205637127161026,
+      "learning_rate": 2.3743022353500713e-05,
+      "loss": 9.7256,
+      "step": 42160
+    },
+    {
+      "epoch": 0.21059202476965716,
+      "grad_norm": 0.09348385035991669,
+      "learning_rate": 2.3741520438559163e-05,
+      "loss": 9.7176,
+      "step": 42170
+    },
+    {
+      "epoch": 0.2106419635945966,
+      "grad_norm": 0.09955950081348419,
+      "learning_rate": 2.3740018523617613e-05,
+      "loss": 9.7243,
+      "step": 42180
+    },
+    {
+      "epoch": 0.21069190241953606,
+      "grad_norm": 0.09449559450149536,
+      "learning_rate": 2.3738516608676063e-05,
+      "loss": 9.7126,
+      "step": 42190
+    },
+    {
+      "epoch": 0.2107418412444755,
+      "grad_norm": 0.09414367377758026,
+      "learning_rate": 2.3737014693734514e-05,
+      "loss": 9.7162,
+      "step": 42200
+    },
+    {
+      "epoch": 0.21079178006941496,
+      "grad_norm": 0.09569656103849411,
+      "learning_rate": 2.373551277879296e-05,
+      "loss": 9.7146,
+      "step": 42210
+    },
+    {
+      "epoch": 0.2108417188943544,
+      "grad_norm": 0.10219884663820267,
+      "learning_rate": 2.373401086385141e-05,
+      "loss": 9.7196,
+      "step": 42220
+    },
+    {
+      "epoch": 0.21089165771929386,
+      "grad_norm": 0.091194286942482,
+      "learning_rate": 2.373250894890986e-05,
+      "loss": 9.7179,
+      "step": 42230
+    },
+    {
+      "epoch": 0.2109415965442333,
+      "grad_norm": 0.09087838232517242,
+      "learning_rate": 2.373100703396831e-05,
+      "loss": 9.7237,
+      "step": 42240
+    },
+    {
+      "epoch": 0.21099153536917276,
+      "grad_norm": 0.09350999444723129,
+      "learning_rate": 2.372950511902676e-05,
+      "loss": 9.7151,
+      "step": 42250
+    },
+    {
+      "epoch": 0.2110414741941122,
+      "grad_norm": 0.08830499649047852,
+      "learning_rate": 2.3728003204085208e-05,
+      "loss": 9.7138,
+      "step": 42260
+    },
+    {
+      "epoch": 0.21109141301905165,
+      "grad_norm": 0.09413766860961914,
+      "learning_rate": 2.3726501289143658e-05,
+      "loss": 9.7158,
+      "step": 42270
+    },
+    {
+      "epoch": 0.2111413518439911,
+      "grad_norm": 0.09809092432260513,
+      "learning_rate": 2.372499937420211e-05,
+      "loss": 9.7141,
+      "step": 42280
+    },
+    {
+      "epoch": 0.21119129066893055,
+      "grad_norm": 0.09166799485683441,
+      "learning_rate": 2.372349745926056e-05,
+      "loss": 9.7125,
+      "step": 42290
+    },
+    {
+      "epoch": 0.21124122949387,
+      "grad_norm": 0.09736365079879761,
+      "learning_rate": 2.372199554431901e-05,
+      "loss": 9.7256,
+      "step": 42300
+    },
+    {
+      "epoch": 0.21129116831880945,
+      "grad_norm": 0.10044820606708527,
+      "learning_rate": 2.3720493629377455e-05,
+      "loss": 9.7196,
+      "step": 42310
+    },
+    {
+      "epoch": 0.2113411071437489,
+      "grad_norm": 0.09292640537023544,
+      "learning_rate": 2.3718991714435906e-05,
+      "loss": 9.7137,
+      "step": 42320
+    },
+    {
+      "epoch": 0.21139104596868835,
+      "grad_norm": 0.09433227777481079,
+      "learning_rate": 2.3717489799494356e-05,
+      "loss": 9.7101,
+      "step": 42330
+    },
+    {
+      "epoch": 0.2114409847936278,
+      "grad_norm": 0.0979892909526825,
+      "learning_rate": 2.3715987884552806e-05,
+      "loss": 9.7208,
+      "step": 42340
+    },
+    {
+      "epoch": 0.21149092361856725,
+      "grad_norm": 0.09797633439302444,
+      "learning_rate": 2.3714485969611256e-05,
+      "loss": 9.7147,
+      "step": 42350
+    },
+    {
+      "epoch": 0.2115408624435067,
+      "grad_norm": 0.09626011550426483,
+      "learning_rate": 2.3712984054669703e-05,
+      "loss": 9.71,
+      "step": 42360
+    },
+    {
+      "epoch": 0.21159080126844615,
+      "grad_norm": 0.10008444637060165,
+      "learning_rate": 2.3711482139728157e-05,
+      "loss": 9.7133,
+      "step": 42370
+    },
+    {
+      "epoch": 0.2116407400933856,
+      "grad_norm": 0.09693645685911179,
+      "learning_rate": 2.3709980224786603e-05,
+      "loss": 9.7155,
+      "step": 42380
+    },
+    {
+      "epoch": 0.21169067891832505,
+      "grad_norm": 0.09394808113574982,
+      "learning_rate": 2.3708478309845054e-05,
+      "loss": 9.7087,
+      "step": 42390
+    },
+    {
+      "epoch": 0.2117406177432645,
+      "grad_norm": 0.0966886356472969,
+      "learning_rate": 2.3706976394903504e-05,
+      "loss": 9.7124,
+      "step": 42400
+    },
+    {
+      "epoch": 0.21179055656820395,
+      "grad_norm": 0.09753970056772232,
+      "learning_rate": 2.370547447996195e-05,
+      "loss": 9.712,
+      "step": 42410
+    },
+    {
+      "epoch": 0.2118404953931434,
+      "grad_norm": 0.095003142952919,
+      "learning_rate": 2.3703972565020404e-05,
+      "loss": 9.7176,
+      "step": 42420
+    },
+    {
+      "epoch": 0.21189043421808285,
+      "grad_norm": 0.09311114996671677,
+      "learning_rate": 2.370247065007885e-05,
+      "loss": 9.7169,
+      "step": 42430
+    },
+    {
+      "epoch": 0.2119403730430223,
+      "grad_norm": 0.08979105204343796,
+      "learning_rate": 2.37009687351373e-05,
+      "loss": 9.7247,
+      "step": 42440
+    },
+    {
+      "epoch": 0.21199031186796174,
+      "grad_norm": 0.09097933769226074,
+      "learning_rate": 2.369946682019575e-05,
+      "loss": 9.7103,
+      "step": 42450
+    },
+    {
+      "epoch": 0.2120402506929012,
+      "grad_norm": 0.09349437057971954,
+      "learning_rate": 2.3697964905254198e-05,
+      "loss": 9.7147,
+      "step": 42460
+    },
+    {
+      "epoch": 0.21209018951784064,
+      "grad_norm": 0.09845960140228271,
+      "learning_rate": 2.369646299031265e-05,
+      "loss": 9.7105,
+      "step": 42470
+    },
+    {
+      "epoch": 0.2121401283427801,
+      "grad_norm": 0.09305304288864136,
+      "learning_rate": 2.36949610753711e-05,
+      "loss": 9.7152,
+      "step": 42480
+    },
+    {
+      "epoch": 0.21219006716771954,
+      "grad_norm": 0.09439215064048767,
+      "learning_rate": 2.369345916042955e-05,
+      "loss": 9.7083,
+      "step": 42490
+    },
+    {
+      "epoch": 0.212240005992659,
+      "grad_norm": 0.09390342235565186,
+      "learning_rate": 2.3691957245488e-05,
+      "loss": 9.7082,
+      "step": 42500
+    },
+    {
+      "epoch": 0.21228994481759844,
+      "grad_norm": 0.08965789526700974,
+      "learning_rate": 2.3690455330546445e-05,
+      "loss": 9.7106,
+      "step": 42510
+    },
+    {
+      "epoch": 0.2123398836425379,
+      "grad_norm": 0.09593262523412704,
+      "learning_rate": 2.36889534156049e-05,
+      "loss": 9.7163,
+      "step": 42520
+    },
+    {
+      "epoch": 0.21238982246747734,
+      "grad_norm": 0.09541445225477219,
+      "learning_rate": 2.3687451500663346e-05,
+      "loss": 9.7153,
+      "step": 42530
+    },
+    {
+      "epoch": 0.2124397612924168,
+      "grad_norm": 0.09254483878612518,
+      "learning_rate": 2.3685949585721796e-05,
+      "loss": 9.7166,
+      "step": 42540
+    },
+    {
+      "epoch": 0.21248970011735624,
+      "grad_norm": 0.09330740571022034,
+      "learning_rate": 2.3684447670780246e-05,
+      "loss": 9.7049,
+      "step": 42550
+    },
+    {
+      "epoch": 0.2125396389422957,
+      "grad_norm": 0.10075175762176514,
+      "learning_rate": 2.3682945755838693e-05,
+      "loss": 9.7118,
+      "step": 42560
+    },
+    {
+      "epoch": 0.21258957776723514,
+      "grad_norm": 0.09620960056781769,
+      "learning_rate": 2.3681443840897147e-05,
+      "loss": 9.711,
+      "step": 42570
+    },
+    {
+      "epoch": 0.2126395165921746,
+      "grad_norm": 0.09333734214305878,
+      "learning_rate": 2.3679941925955593e-05,
+      "loss": 9.7069,
+      "step": 42580
+    },
+    {
+      "epoch": 0.21268945541711404,
+      "grad_norm": 0.09951247274875641,
+      "learning_rate": 2.3678440011014044e-05,
+      "loss": 9.7065,
+      "step": 42590
+    },
+    {
+      "epoch": 0.2127393942420535,
+      "grad_norm": 0.09395639598369598,
+      "learning_rate": 2.3676938096072494e-05,
+      "loss": 9.7071,
+      "step": 42600
+    },
+    {
+      "epoch": 0.21278933306699294,
+      "grad_norm": 0.10100211948156357,
+      "learning_rate": 2.367543618113094e-05,
+      "loss": 9.7104,
+      "step": 42610
+    },
+    {
+      "epoch": 0.21283927189193239,
+      "grad_norm": 0.09487060457468033,
+      "learning_rate": 2.3673934266189394e-05,
+      "loss": 9.7075,
+      "step": 42620
+    },
+    {
+      "epoch": 0.21288921071687184,
+      "grad_norm": 0.09256965667009354,
+      "learning_rate": 2.367243235124784e-05,
+      "loss": 9.7168,
+      "step": 42630
+    },
+    {
+      "epoch": 0.21293914954181128,
+      "grad_norm": 0.0953928530216217,
+      "learning_rate": 2.367093043630629e-05,
+      "loss": 9.708,
+      "step": 42640
+    },
+    {
+      "epoch": 0.21298908836675073,
+      "grad_norm": 0.09185754507780075,
+      "learning_rate": 2.366942852136474e-05,
+      "loss": 9.7165,
+      "step": 42650
+    },
+    {
+      "epoch": 0.21303902719169018,
+      "grad_norm": 0.09027005732059479,
+      "learning_rate": 2.3667926606423188e-05,
+      "loss": 9.71,
+      "step": 42660
+    },
+    {
+      "epoch": 0.21308896601662963,
+      "grad_norm": 0.0954156145453453,
+      "learning_rate": 2.366642469148164e-05,
+      "loss": 9.709,
+      "step": 42670
+    },
+    {
+      "epoch": 0.21313890484156908,
+      "grad_norm": 0.08864131569862366,
+      "learning_rate": 2.366492277654009e-05,
+      "loss": 9.7154,
+      "step": 42680
+    },
+    {
+      "epoch": 0.21318884366650853,
+      "grad_norm": 0.09630046784877777,
+      "learning_rate": 2.3663420861598542e-05,
+      "loss": 9.712,
+      "step": 42690
+    },
+    {
+      "epoch": 0.21323878249144798,
+      "grad_norm": 0.09661024063825607,
+      "learning_rate": 2.366191894665699e-05,
+      "loss": 9.7077,
+      "step": 42700
+    },
+    {
+      "epoch": 0.21328872131638743,
+      "grad_norm": 0.09176916629076004,
+      "learning_rate": 2.3660417031715436e-05,
+      "loss": 9.7138,
+      "step": 42710
+    },
+    {
+      "epoch": 0.21333866014132688,
+      "grad_norm": 0.09320518374443054,
+      "learning_rate": 2.365891511677389e-05,
+      "loss": 9.712,
+      "step": 42720
+    },
+    {
+      "epoch": 0.21338859896626633,
+      "grad_norm": 0.10201375186443329,
+      "learning_rate": 2.3657413201832336e-05,
+      "loss": 9.7025,
+      "step": 42730
+    },
+    {
+      "epoch": 0.21343853779120578,
+      "grad_norm": 0.09826554358005524,
+      "learning_rate": 2.365591128689079e-05,
+      "loss": 9.7106,
+      "step": 42740
+    },
+    {
+      "epoch": 0.21348847661614523,
+      "grad_norm": 0.09512682259082794,
+      "learning_rate": 2.3654409371949236e-05,
+      "loss": 9.7113,
+      "step": 42750
+    },
+    {
+      "epoch": 0.21353841544108468,
+      "grad_norm": 0.10446712374687195,
+      "learning_rate": 2.3652907457007683e-05,
+      "loss": 9.7097,
+      "step": 42760
+    },
+    {
+      "epoch": 0.21358835426602413,
+      "grad_norm": 0.09374655038118362,
+      "learning_rate": 2.3651405542066137e-05,
+      "loss": 9.7101,
+      "step": 42770
+    },
+    {
+      "epoch": 0.21363829309096358,
+      "grad_norm": 0.10137982666492462,
+      "learning_rate": 2.3649903627124583e-05,
+      "loss": 9.7063,
+      "step": 42780
+    },
+    {
+      "epoch": 0.21368823191590303,
+      "grad_norm": 0.09552907198667526,
+      "learning_rate": 2.3648401712183037e-05,
+      "loss": 9.7035,
+      "step": 42790
+    },
+    {
+      "epoch": 0.21373817074084248,
+      "grad_norm": 0.10118672996759415,
+      "learning_rate": 2.3646899797241484e-05,
+      "loss": 9.7137,
+      "step": 42800
+    },
+    {
+      "epoch": 0.21378810956578193,
+      "grad_norm": 0.09724298119544983,
+      "learning_rate": 2.364539788229993e-05,
+      "loss": 9.7052,
+      "step": 42810
+    },
+    {
+      "epoch": 0.21383804839072137,
+      "grad_norm": 0.09356910735368729,
+      "learning_rate": 2.3643895967358384e-05,
+      "loss": 9.7068,
+      "step": 42820
+    },
+    {
+      "epoch": 0.21388798721566082,
+      "grad_norm": 0.09422975033521652,
+      "learning_rate": 2.364239405241683e-05,
+      "loss": 9.7097,
+      "step": 42830
+    },
+    {
+      "epoch": 0.21393792604060027,
+      "grad_norm": 0.09154810756444931,
+      "learning_rate": 2.3640892137475284e-05,
+      "loss": 9.7082,
+      "step": 42840
+    },
+    {
+      "epoch": 0.21398786486553972,
+      "grad_norm": 0.08905761688947678,
+      "learning_rate": 2.363939022253373e-05,
+      "loss": 9.7014,
+      "step": 42850
+    },
+    {
+      "epoch": 0.21403780369047917,
+      "grad_norm": 0.0964711457490921,
+      "learning_rate": 2.3637888307592178e-05,
+      "loss": 9.7035,
+      "step": 42860
+    },
+    {
+      "epoch": 0.21408774251541862,
+      "grad_norm": 0.09149517863988876,
+      "learning_rate": 2.363638639265063e-05,
+      "loss": 9.7052,
+      "step": 42870
+    },
+    {
+      "epoch": 0.21413768134035807,
+      "grad_norm": 0.09464892745018005,
+      "learning_rate": 2.363488447770908e-05,
+      "loss": 9.7134,
+      "step": 42880
+    },
+    {
+      "epoch": 0.21418762016529752,
+      "grad_norm": 0.09502291679382324,
+      "learning_rate": 2.3633382562767532e-05,
+      "loss": 9.7036,
+      "step": 42890
+    },
+    {
+      "epoch": 0.21423755899023697,
+      "grad_norm": 0.09215651452541351,
+      "learning_rate": 2.363188064782598e-05,
+      "loss": 9.7038,
+      "step": 42900
+    },
+    {
+      "epoch": 0.21428749781517642,
+      "grad_norm": 0.09255354106426239,
+      "learning_rate": 2.3630378732884426e-05,
+      "loss": 9.6993,
+      "step": 42910
+    },
+    {
+      "epoch": 0.21433743664011587,
+      "grad_norm": 0.09328746050596237,
+      "learning_rate": 2.362887681794288e-05,
+      "loss": 9.7027,
+      "step": 42920
+    },
+    {
+      "epoch": 0.21438737546505532,
+      "grad_norm": 0.0980675145983696,
+      "learning_rate": 2.3627374903001326e-05,
+      "loss": 9.7081,
+      "step": 42930
+    },
+    {
+      "epoch": 0.21443731428999477,
+      "grad_norm": 0.09392713755369186,
+      "learning_rate": 2.362587298805978e-05,
+      "loss": 9.7009,
+      "step": 42940
+    },
+    {
+      "epoch": 0.21448725311493422,
+      "grad_norm": 0.09184177964925766,
+      "learning_rate": 2.3624371073118226e-05,
+      "loss": 9.7055,
+      "step": 42950
+    },
+    {
+      "epoch": 0.21453719193987367,
+      "grad_norm": 0.10423033684492111,
+      "learning_rate": 2.3622869158176673e-05,
+      "loss": 9.7083,
+      "step": 42960
+    },
+    {
+      "epoch": 0.21458713076481312,
+      "grad_norm": 0.09163152426481247,
+      "learning_rate": 2.3621367243235127e-05,
+      "loss": 10.5286,
+      "step": 42970
+    },
+    {
+      "epoch": 0.21463706958975257,
+      "grad_norm": 0.09652361273765564,
+      "learning_rate": 2.3619865328293573e-05,
+      "loss": 41.2823,
+      "step": 42980
+    },
+    {
+      "epoch": 0.21468700841469202,
+      "grad_norm": 0.09223783016204834,
+      "learning_rate": 2.3618363413352027e-05,
+      "loss": 9.7045,
+      "step": 42990
+    },
+    {
+      "epoch": 0.21473694723963146,
+      "grad_norm": 0.09330345690250397,
+      "learning_rate": 2.3616861498410474e-05,
+      "loss": 9.7024,
+      "step": 43000
+    },
+    {
+      "epoch": 0.21478688606457091,
+      "grad_norm": 0.09638207405805588,
+      "learning_rate": 2.3615359583468924e-05,
+      "loss": 9.7073,
+      "step": 43010
+    },
+    {
+      "epoch": 0.21483682488951034,
+      "grad_norm": 0.09385758638381958,
+      "learning_rate": 2.3613857668527374e-05,
+      "loss": 9.706,
+      "step": 43020
+    },
+    {
+      "epoch": 0.21488676371444979,
+      "grad_norm": 0.09166792035102844,
+      "learning_rate": 2.361235575358582e-05,
+      "loss": 9.6945,
+      "step": 43030
+    },
+    {
+      "epoch": 0.21493670253938923,
+      "grad_norm": 0.09677714109420776,
+      "learning_rate": 2.3610853838644274e-05,
+      "loss": 9.7021,
+      "step": 43040
+    },
+    {
+      "epoch": 0.21498664136432868,
+      "grad_norm": 0.09546297788619995,
+      "learning_rate": 2.360935192370272e-05,
+      "loss": 9.6937,
+      "step": 43050
+    },
+    {
+      "epoch": 0.21503658018926813,
+      "grad_norm": 0.09486702084541321,
+      "learning_rate": 2.360785000876117e-05,
+      "loss": 9.7021,
+      "step": 43060
+    },
+    {
+      "epoch": 0.21508651901420758,
+      "grad_norm": 0.09048479050397873,
+      "learning_rate": 2.360634809381962e-05,
+      "loss": 9.7101,
+      "step": 43070
+    },
+    {
+      "epoch": 0.21513645783914703,
+      "grad_norm": 0.0977143719792366,
+      "learning_rate": 2.360484617887807e-05,
+      "loss": 9.6991,
+      "step": 43080
+    },
+    {
+      "epoch": 0.21518639666408648,
+      "grad_norm": 0.0984925925731659,
+      "learning_rate": 2.3603344263936522e-05,
+      "loss": 9.6967,
+      "step": 43090
+    },
+    {
+      "epoch": 0.21523633548902593,
+      "grad_norm": 0.09045634418725967,
+      "learning_rate": 2.360184234899497e-05,
+      "loss": 9.6997,
+      "step": 43100
+    },
+    {
+      "epoch": 0.21528627431396538,
+      "grad_norm": 0.09143086522817612,
+      "learning_rate": 2.360034043405342e-05,
+      "loss": 9.7057,
+      "step": 43110
+    },
+    {
+      "epoch": 0.21533621313890483,
+      "grad_norm": 0.09460347890853882,
+      "learning_rate": 2.359883851911187e-05,
+      "loss": 9.6983,
+      "step": 43120
+    },
+    {
+      "epoch": 0.21538615196384428,
+      "grad_norm": 0.09431096911430359,
+      "learning_rate": 2.3597336604170316e-05,
+      "loss": 9.7025,
+      "step": 43130
+    },
+    {
+      "epoch": 0.21543609078878373,
+      "grad_norm": 0.1000736653804779,
+      "learning_rate": 2.359583468922877e-05,
+      "loss": 9.694,
+      "step": 43140
+    },
+    {
+      "epoch": 0.21548602961372318,
+      "grad_norm": 0.09227129071950912,
+      "learning_rate": 2.3594332774287216e-05,
+      "loss": 9.7007,
+      "step": 43150
+    },
+    {
+      "epoch": 0.21553596843866263,
+      "grad_norm": 0.09356602281332016,
+      "learning_rate": 2.3592830859345666e-05,
+      "loss": 9.6967,
+      "step": 43160
+    },
+    {
+      "epoch": 0.21558590726360208,
+      "grad_norm": 0.09989676624536514,
+      "learning_rate": 2.3591328944404117e-05,
+      "loss": 9.7074,
+      "step": 43170
+    },
+    {
+      "epoch": 0.21563584608854153,
+      "grad_norm": 0.0949828028678894,
+      "learning_rate": 2.3589827029462563e-05,
+      "loss": 9.6986,
+      "step": 43180
+    },
+    {
+      "epoch": 0.21568578491348098,
+      "grad_norm": 0.09488669037818909,
+      "learning_rate": 2.3588325114521017e-05,
+      "loss": 9.7024,
+      "step": 43190
+    },
+    {
+      "epoch": 0.21573572373842043,
+      "grad_norm": 0.09980147331953049,
+      "learning_rate": 2.3586823199579464e-05,
+      "loss": 9.6998,
+      "step": 43200
+    },
+    {
+      "epoch": 0.21578566256335988,
+      "grad_norm": 0.09408322721719742,
+      "learning_rate": 2.3585321284637914e-05,
+      "loss": 9.6962,
+      "step": 43210
+    },
+    {
+      "epoch": 0.21583560138829933,
+      "grad_norm": 0.09337528049945831,
+      "learning_rate": 2.3583819369696364e-05,
+      "loss": 9.7064,
+      "step": 43220
+    },
+    {
+      "epoch": 0.21588554021323877,
+      "grad_norm": 0.09422304481267929,
+      "learning_rate": 2.358231745475481e-05,
+      "loss": 9.6981,
+      "step": 43230
+    },
+    {
+      "epoch": 0.21593547903817822,
+      "grad_norm": 0.09392604231834412,
+      "learning_rate": 2.3580815539813264e-05,
+      "loss": 9.702,
+      "step": 43240
+    },
+    {
+      "epoch": 0.21598541786311767,
+      "grad_norm": 0.09568940103054047,
+      "learning_rate": 2.357931362487171e-05,
+      "loss": 9.6984,
+      "step": 43250
+    },
+    {
+      "epoch": 0.21603535668805712,
+      "grad_norm": 0.09230421483516693,
+      "learning_rate": 2.357781170993016e-05,
+      "loss": 9.6961,
+      "step": 43260
+    },
+    {
+      "epoch": 0.21608529551299657,
+      "grad_norm": 0.0997990295290947,
+      "learning_rate": 2.357630979498861e-05,
+      "loss": 9.6982,
+      "step": 43270
+    },
+    {
+      "epoch": 0.21613523433793602,
+      "grad_norm": 0.09223617613315582,
+      "learning_rate": 2.357480788004706e-05,
+      "loss": 9.6975,
+      "step": 43280
+    },
+    {
+      "epoch": 0.21618517316287547,
+      "grad_norm": 0.09891097992658615,
+      "learning_rate": 2.3573305965105512e-05,
+      "loss": 9.6933,
+      "step": 43290
+    },
+    {
+      "epoch": 0.21623511198781492,
+      "grad_norm": 0.09234698116779327,
+      "learning_rate": 2.357180405016396e-05,
+      "loss": 9.6913,
+      "step": 43300
+    },
+    {
+      "epoch": 0.21628505081275437,
+      "grad_norm": 0.0998927652835846,
+      "learning_rate": 2.357030213522241e-05,
+      "loss": 9.6922,
+      "step": 43310
+    },
+    {
+      "epoch": 0.21633498963769382,
+      "grad_norm": 0.09355306625366211,
+      "learning_rate": 2.356880022028086e-05,
+      "loss": 9.6943,
+      "step": 43320
+    },
+    {
+      "epoch": 0.21638492846263327,
+      "grad_norm": 0.09233734756708145,
+      "learning_rate": 2.356729830533931e-05,
+      "loss": 9.6983,
+      "step": 43330
+    },
+    {
+      "epoch": 0.21643486728757272,
+      "grad_norm": 0.09967396408319473,
+      "learning_rate": 2.356579639039776e-05,
+      "loss": 9.6938,
+      "step": 43340
+    },
+    {
+      "epoch": 0.21648480611251217,
+      "grad_norm": 0.09385713189840317,
+      "learning_rate": 2.3564294475456206e-05,
+      "loss": 9.6936,
+      "step": 43350
+    },
+    {
+      "epoch": 0.21653474493745162,
+      "grad_norm": 0.09764021635055542,
+      "learning_rate": 2.3562792560514656e-05,
+      "loss": 9.6984,
+      "step": 43360
+    },
+    {
+      "epoch": 0.21658468376239107,
+      "grad_norm": 0.0969097763299942,
+      "learning_rate": 2.3561290645573107e-05,
+      "loss": 9.6953,
+      "step": 43370
+    },
+    {
+      "epoch": 0.21663462258733052,
+      "grad_norm": 0.09680244326591492,
+      "learning_rate": 2.3559788730631557e-05,
+      "loss": 9.6967,
+      "step": 43380
+    },
+    {
+      "epoch": 0.21668456141226997,
+      "grad_norm": 0.09880103170871735,
+      "learning_rate": 2.3558286815690007e-05,
+      "loss": 9.703,
+      "step": 43390
+    },
+    {
+      "epoch": 0.21673450023720942,
+      "grad_norm": 0.09156890958547592,
+      "learning_rate": 2.3556784900748454e-05,
+      "loss": 9.6993,
+      "step": 43400
+    },
+    {
+      "epoch": 0.21678443906214886,
+      "grad_norm": 0.09816037863492966,
+      "learning_rate": 2.3555282985806904e-05,
+      "loss": 9.6899,
+      "step": 43410
+    },
+    {
+      "epoch": 0.21683437788708831,
+      "grad_norm": 0.09547406435012817,
+      "learning_rate": 2.3553781070865354e-05,
+      "loss": 9.6893,
+      "step": 43420
+    },
+    {
+      "epoch": 0.21688431671202776,
+      "grad_norm": 0.10225875675678253,
+      "learning_rate": 2.3552279155923804e-05,
+      "loss": 9.6887,
+      "step": 43430
+    },
+    {
+      "epoch": 0.2169342555369672,
+      "grad_norm": 0.09828753769397736,
+      "learning_rate": 2.3550777240982254e-05,
+      "loss": 9.6982,
+      "step": 43440
+    },
+    {
+      "epoch": 0.21698419436190666,
+      "grad_norm": 0.09586260467767715,
+      "learning_rate": 2.35492753260407e-05,
+      "loss": 9.6886,
+      "step": 43450
+    },
+    {
+      "epoch": 0.2170341331868461,
+      "grad_norm": 0.09193310141563416,
+      "learning_rate": 2.354777341109915e-05,
+      "loss": 9.7037,
+      "step": 43460
+    },
+    {
+      "epoch": 0.21708407201178556,
+      "grad_norm": 0.09209885448217392,
+      "learning_rate": 2.35462714961576e-05,
+      "loss": 9.692,
+      "step": 43470
+    },
+    {
+      "epoch": 0.217134010836725,
+      "grad_norm": 0.09702476859092712,
+      "learning_rate": 2.3544769581216052e-05,
+      "loss": 9.6904,
+      "step": 43480
+    },
+    {
+      "epoch": 0.21718394966166446,
+      "grad_norm": 0.0922791138291359,
+      "learning_rate": 2.3543267666274502e-05,
+      "loss": 9.6992,
+      "step": 43490
+    },
+    {
+      "epoch": 0.2172338884866039,
+      "grad_norm": 0.09711603820323944,
+      "learning_rate": 2.354176575133295e-05,
+      "loss": 9.6947,
+      "step": 43500
+    },
+    {
+      "epoch": 0.21728382731154336,
+      "grad_norm": 0.0968707948923111,
+      "learning_rate": 2.35402638363914e-05,
+      "loss": 9.6859,
+      "step": 43510
+    },
+    {
+      "epoch": 0.2173337661364828,
+      "grad_norm": 0.09466083347797394,
+      "learning_rate": 2.353876192144985e-05,
+      "loss": 9.6967,
+      "step": 43520
+    },
+    {
+      "epoch": 0.21738370496142226,
+      "grad_norm": 0.09528151154518127,
+      "learning_rate": 2.35372600065083e-05,
+      "loss": 9.6887,
+      "step": 43530
+    },
+    {
+      "epoch": 0.2174336437863617,
+      "grad_norm": 0.09486556053161621,
+      "learning_rate": 2.353575809156675e-05,
+      "loss": 9.688,
+      "step": 43540
+    },
+    {
+      "epoch": 0.21748358261130116,
+      "grad_norm": 0.09180723130702972,
+      "learning_rate": 2.3534256176625196e-05,
+      "loss": 9.6934,
+      "step": 43550
+    },
+    {
+      "epoch": 0.2175335214362406,
+      "grad_norm": 0.09143299609422684,
+      "learning_rate": 2.3532754261683646e-05,
+      "loss": 9.6973,
+      "step": 43560
+    },
+    {
+      "epoch": 0.21758346026118006,
+      "grad_norm": 0.09400318562984467,
+      "learning_rate": 2.3531252346742097e-05,
+      "loss": 9.6916,
+      "step": 43570
+    },
+    {
+      "epoch": 0.2176333990861195,
+      "grad_norm": 0.09261635690927505,
+      "learning_rate": 2.3529750431800547e-05,
+      "loss": 9.701,
+      "step": 43580
+    },
+    {
+      "epoch": 0.21768333791105896,
+      "grad_norm": 0.09949515014886856,
+      "learning_rate": 2.3528248516858997e-05,
+      "loss": 9.6943,
+      "step": 43590
+    },
+    {
+      "epoch": 0.2177332767359984,
+      "grad_norm": 0.09125660359859467,
+      "learning_rate": 2.3526746601917444e-05,
+      "loss": 9.6932,
+      "step": 43600
+    },
+    {
+      "epoch": 0.21778321556093785,
+      "grad_norm": 0.08959287405014038,
+      "learning_rate": 2.3525244686975894e-05,
+      "loss": 9.6986,
+      "step": 43610
+    },
+    {
+      "epoch": 0.2178331543858773,
+      "grad_norm": 0.0992596298456192,
+      "learning_rate": 2.3523742772034344e-05,
+      "loss": 9.6865,
+      "step": 43620
+    },
+    {
+      "epoch": 0.21788309321081675,
+      "grad_norm": 0.09467126429080963,
+      "learning_rate": 2.3522240857092794e-05,
+      "loss": 9.6988,
+      "step": 43630
+    },
+    {
+      "epoch": 0.2179330320357562,
+      "grad_norm": 0.09459062665700912,
+      "learning_rate": 2.3520738942151244e-05,
+      "loss": 9.6991,
+      "step": 43640
+    },
+    {
+      "epoch": 0.21798297086069565,
+      "grad_norm": 0.09231843799352646,
+      "learning_rate": 2.3519237027209695e-05,
+      "loss": 9.689,
+      "step": 43650
+    },
+    {
+      "epoch": 0.2180329096856351,
+      "grad_norm": 0.09330347180366516,
+      "learning_rate": 2.351773511226814e-05,
+      "loss": 9.6886,
+      "step": 43660
+    },
+    {
+      "epoch": 0.21808284851057455,
+      "grad_norm": 0.09178127348423004,
+      "learning_rate": 2.351623319732659e-05,
+      "loss": 9.6855,
+      "step": 43670
+    },
+    {
+      "epoch": 0.218132787335514,
+      "grad_norm": 0.09343292564153671,
+      "learning_rate": 2.3514731282385042e-05,
+      "loss": 9.6845,
+      "step": 43680
+    },
+    {
+      "epoch": 0.21818272616045345,
+      "grad_norm": 0.09548097103834152,
+      "learning_rate": 2.3513229367443492e-05,
+      "loss": 9.6896,
+      "step": 43690
+    },
+    {
+      "epoch": 0.2182326649853929,
+      "grad_norm": 0.0939573422074318,
+      "learning_rate": 2.3511727452501942e-05,
+      "loss": 9.6855,
+      "step": 43700
+    },
+    {
+      "epoch": 0.21828260381033235,
+      "grad_norm": 0.09515061974525452,
+      "learning_rate": 2.351022553756039e-05,
+      "loss": 9.6916,
+      "step": 43710
+    },
+    {
+      "epoch": 0.2183325426352718,
+      "grad_norm": 0.09717291593551636,
+      "learning_rate": 2.350872362261884e-05,
+      "loss": 9.6881,
+      "step": 43720
+    },
+    {
+      "epoch": 0.21838248146021125,
+      "grad_norm": 0.09339723736047745,
+      "learning_rate": 2.350722170767729e-05,
+      "loss": 9.6886,
+      "step": 43730
+    },
+    {
+      "epoch": 0.2184324202851507,
+      "grad_norm": 0.0933520719408989,
+      "learning_rate": 2.350571979273574e-05,
+      "loss": 9.686,
+      "step": 43740
+    },
+    {
+      "epoch": 0.21848235911009015,
+      "grad_norm": 0.09928298741579056,
+      "learning_rate": 2.350421787779419e-05,
+      "loss": 9.6908,
+      "step": 43750
+    },
+    {
+      "epoch": 0.2185322979350296,
+      "grad_norm": 0.09567858278751373,
+      "learning_rate": 2.3502715962852636e-05,
+      "loss": 9.6884,
+      "step": 43760
+    },
+    {
+      "epoch": 0.21858223675996905,
+      "grad_norm": 0.09107061475515366,
+      "learning_rate": 2.3501214047911087e-05,
+      "loss": 9.6938,
+      "step": 43770
+    },
+    {
+      "epoch": 0.2186321755849085,
+      "grad_norm": 0.09627822786569595,
+      "learning_rate": 2.3499712132969537e-05,
+      "loss": 9.6838,
+      "step": 43780
+    },
+    {
+      "epoch": 0.21868211440984794,
+      "grad_norm": 0.09654753655195236,
+      "learning_rate": 2.3498210218027987e-05,
+      "loss": 9.6865,
+      "step": 43790
+    },
+    {
+      "epoch": 0.2187320532347874,
+      "grad_norm": 0.09376217424869537,
+      "learning_rate": 2.3496708303086437e-05,
+      "loss": 9.6885,
+      "step": 43800
+    },
+    {
+      "epoch": 0.21878199205972684,
+      "grad_norm": 0.08824321627616882,
+      "learning_rate": 2.3495206388144884e-05,
+      "loss": 9.6851,
+      "step": 43810
+    },
+    {
+      "epoch": 0.2188319308846663,
+      "grad_norm": 0.09631256759166718,
+      "learning_rate": 2.3493704473203334e-05,
+      "loss": 9.6929,
+      "step": 43820
+    },
+    {
+      "epoch": 0.21888186970960574,
+      "grad_norm": 0.0864676758646965,
+      "learning_rate": 2.3492202558261784e-05,
+      "loss": 9.6841,
+      "step": 43830
+    },
+    {
+      "epoch": 0.2189318085345452,
+      "grad_norm": 0.09546799957752228,
+      "learning_rate": 2.3490700643320235e-05,
+      "loss": 9.691,
+      "step": 43840
+    },
+    {
+      "epoch": 0.21898174735948464,
+      "grad_norm": 0.09540455043315887,
+      "learning_rate": 2.3489198728378685e-05,
+      "loss": 9.6935,
+      "step": 43850
+    },
+    {
+      "epoch": 0.2190316861844241,
+      "grad_norm": 0.0951889380812645,
+      "learning_rate": 2.348769681343713e-05,
+      "loss": 9.6776,
+      "step": 43860
+    },
+    {
+      "epoch": 0.21908162500936354,
+      "grad_norm": 0.09509312361478806,
+      "learning_rate": 2.348619489849558e-05,
+      "loss": 9.6906,
+      "step": 43870
+    },
+    {
+      "epoch": 0.219131563834303,
+      "grad_norm": 0.09178054332733154,
+      "learning_rate": 2.3484692983554032e-05,
+      "loss": 9.6842,
+      "step": 43880
+    },
+    {
+      "epoch": 0.21918150265924244,
+      "grad_norm": 0.09988975524902344,
+      "learning_rate": 2.3483191068612482e-05,
+      "loss": 9.6777,
+      "step": 43890
+    },
+    {
+      "epoch": 0.2192314414841819,
+      "grad_norm": 0.09594890475273132,
+      "learning_rate": 2.3481689153670932e-05,
+      "loss": 9.6763,
+      "step": 43900
+    },
+    {
+      "epoch": 0.21928138030912134,
+      "grad_norm": 0.09186796098947525,
+      "learning_rate": 2.348018723872938e-05,
+      "loss": 9.6859,
+      "step": 43910
+    },
+    {
+      "epoch": 0.2193313191340608,
+      "grad_norm": 0.10070665925741196,
+      "learning_rate": 2.347868532378783e-05,
+      "loss": 9.6894,
+      "step": 43920
+    },
+    {
+      "epoch": 0.21938125795900024,
+      "grad_norm": 0.09850089251995087,
+      "learning_rate": 2.347718340884628e-05,
+      "loss": 9.6834,
+      "step": 43930
+    },
+    {
+      "epoch": 0.21943119678393969,
+      "grad_norm": 0.09155581891536713,
+      "learning_rate": 2.347568149390473e-05,
+      "loss": 9.6808,
+      "step": 43940
+    },
+    {
+      "epoch": 0.21948113560887914,
+      "grad_norm": 0.0951627716422081,
+      "learning_rate": 2.347417957896318e-05,
+      "loss": 9.6904,
+      "step": 43950
+    },
+    {
+      "epoch": 0.21953107443381858,
+      "grad_norm": 0.09084520488977432,
+      "learning_rate": 2.3472677664021626e-05,
+      "loss": 9.6849,
+      "step": 43960
+    },
+    {
+      "epoch": 0.21958101325875803,
+      "grad_norm": 0.09903666377067566,
+      "learning_rate": 2.347117574908008e-05,
+      "loss": 9.6876,
+      "step": 43970
+    },
+    {
+      "epoch": 0.21963095208369748,
+      "grad_norm": 0.09081403911113739,
+      "learning_rate": 2.3469673834138527e-05,
+      "loss": 9.6861,
+      "step": 43980
+    },
+    {
+      "epoch": 0.21968089090863693,
+      "grad_norm": 0.09616552293300629,
+      "learning_rate": 2.3468171919196977e-05,
+      "loss": 9.6771,
+      "step": 43990
+    },
+    {
+      "epoch": 0.21973082973357638,
+      "grad_norm": 0.09797076880931854,
+      "learning_rate": 2.3466670004255427e-05,
+      "loss": 9.6872,
+      "step": 44000
+    },
+    {
+      "epoch": 0.2197807685585158,
+      "grad_norm": 0.09519995003938675,
+      "learning_rate": 2.3465168089313874e-05,
+      "loss": 9.6835,
+      "step": 44010
+    },
+    {
+      "epoch": 0.21983070738345525,
+      "grad_norm": 0.0880652442574501,
+      "learning_rate": 2.3463666174372328e-05,
+      "loss": 9.68,
+      "step": 44020
+    },
+    {
+      "epoch": 0.2198806462083947,
+      "grad_norm": 0.09858622401952744,
+      "learning_rate": 2.3462164259430774e-05,
+      "loss": 9.688,
+      "step": 44030
+    },
+    {
+      "epoch": 0.21993058503333415,
+      "grad_norm": 0.09564345329999924,
+      "learning_rate": 2.3460662344489225e-05,
+      "loss": 9.6826,
+      "step": 44040
+    },
+    {
+      "epoch": 0.2199805238582736,
+      "grad_norm": 0.0934138298034668,
+      "learning_rate": 2.3459160429547675e-05,
+      "loss": 9.6803,
+      "step": 44050
+    },
+    {
+      "epoch": 0.22003046268321305,
+      "grad_norm": 0.09328287839889526,
+      "learning_rate": 2.345765851460612e-05,
+      "loss": 9.6871,
+      "step": 44060
+    },
+    {
+      "epoch": 0.2200804015081525,
+      "grad_norm": 0.09457116574048996,
+      "learning_rate": 2.3456156599664575e-05,
+      "loss": 9.6918,
+      "step": 44070
+    },
+    {
+      "epoch": 0.22013034033309195,
+      "grad_norm": 0.10091323405504227,
+      "learning_rate": 2.3454654684723022e-05,
+      "loss": 9.6813,
+      "step": 44080
+    },
+    {
+      "epoch": 0.2201802791580314,
+      "grad_norm": 0.09307578206062317,
+      "learning_rate": 2.3453152769781472e-05,
+      "loss": 9.6833,
+      "step": 44090
+    },
+    {
+      "epoch": 0.22023021798297085,
+      "grad_norm": 0.09910399466753006,
+      "learning_rate": 2.3451650854839922e-05,
+      "loss": 9.6762,
+      "step": 44100
+    },
+    {
+      "epoch": 0.2202801568079103,
+      "grad_norm": 0.10042911767959595,
+      "learning_rate": 2.345014893989837e-05,
+      "loss": 9.6861,
+      "step": 44110
+    },
+    {
+      "epoch": 0.22033009563284975,
+      "grad_norm": 0.09451974183320999,
+      "learning_rate": 2.3448647024956823e-05,
+      "loss": 9.6838,
+      "step": 44120
+    },
+    {
+      "epoch": 0.2203800344577892,
+      "grad_norm": 0.09194231033325195,
+      "learning_rate": 2.344714511001527e-05,
+      "loss": 9.6873,
+      "step": 44130
+    },
+    {
+      "epoch": 0.22042997328272865,
+      "grad_norm": 0.09405980259180069,
+      "learning_rate": 2.344564319507372e-05,
+      "loss": 9.6875,
+      "step": 44140
+    },
+    {
+      "epoch": 0.2204799121076681,
+      "grad_norm": 0.09576237946748734,
+      "learning_rate": 2.344414128013217e-05,
+      "loss": 9.6794,
+      "step": 44150
+    },
+    {
+      "epoch": 0.22052985093260755,
+      "grad_norm": 0.09872204065322876,
+      "learning_rate": 2.3442639365190616e-05,
+      "loss": 9.6833,
+      "step": 44160
+    },
+    {
+      "epoch": 0.220579789757547,
+      "grad_norm": 0.09933184832334518,
+      "learning_rate": 2.344113745024907e-05,
+      "loss": 9.6823,
+      "step": 44170
+    },
+    {
+      "epoch": 0.22062972858248645,
+      "grad_norm": 0.09750817716121674,
+      "learning_rate": 2.3439635535307517e-05,
+      "loss": 9.6704,
+      "step": 44180
+    },
+    {
+      "epoch": 0.2206796674074259,
+      "grad_norm": 0.09613807499408722,
+      "learning_rate": 2.3438133620365967e-05,
+      "loss": 9.6826,
+      "step": 44190
+    },
+    {
+      "epoch": 0.22072960623236534,
+      "grad_norm": 0.10209544748067856,
+      "learning_rate": 2.3436631705424417e-05,
+      "loss": 9.682,
+      "step": 44200
+    },
+    {
+      "epoch": 0.2207795450573048,
+      "grad_norm": 0.0971246138215065,
+      "learning_rate": 2.3435129790482864e-05,
+      "loss": 9.6785,
+      "step": 44210
+    },
+    {
+      "epoch": 0.22082948388224424,
+      "grad_norm": 0.10060384124517441,
+      "learning_rate": 2.3433627875541318e-05,
+      "loss": 9.6795,
+      "step": 44220
+    },
+    {
+      "epoch": 0.2208794227071837,
+      "grad_norm": 0.09796610474586487,
+      "learning_rate": 2.3432125960599764e-05,
+      "loss": 9.6724,
+      "step": 44230
+    },
+    {
+      "epoch": 0.22092936153212314,
+      "grad_norm": 0.09403125196695328,
+      "learning_rate": 2.3430624045658215e-05,
+      "loss": 9.6751,
+      "step": 44240
+    },
+    {
+      "epoch": 0.2209793003570626,
+      "grad_norm": 0.0945911630988121,
+      "learning_rate": 2.3429122130716665e-05,
+      "loss": 9.6734,
+      "step": 44250
+    },
+    {
+      "epoch": 0.22102923918200204,
+      "grad_norm": 0.09785491228103638,
+      "learning_rate": 2.342762021577511e-05,
+      "loss": 9.678,
+      "step": 44260
+    },
+    {
+      "epoch": 0.2210791780069415,
+      "grad_norm": 0.09269008040428162,
+      "learning_rate": 2.3426118300833565e-05,
+      "loss": 9.682,
+      "step": 44270
+    },
+    {
+      "epoch": 0.22112911683188094,
+      "grad_norm": 0.09193340688943863,
+      "learning_rate": 2.3424616385892012e-05,
+      "loss": 9.6794,
+      "step": 44280
+    },
+    {
+      "epoch": 0.2211790556568204,
+      "grad_norm": 0.09782993048429489,
+      "learning_rate": 2.3423114470950462e-05,
+      "loss": 9.6675,
+      "step": 44290
+    },
+    {
+      "epoch": 0.22122899448175984,
+      "grad_norm": 0.09436007589101791,
+      "learning_rate": 2.3421612556008912e-05,
+      "loss": 9.6741,
+      "step": 44300
+    },
+    {
+      "epoch": 0.2212789333066993,
+      "grad_norm": 0.09482413530349731,
+      "learning_rate": 2.342011064106736e-05,
+      "loss": 9.6703,
+      "step": 44310
+    },
+    {
+      "epoch": 0.22132887213163874,
+      "grad_norm": 0.09317035228013992,
+      "learning_rate": 2.3418608726125813e-05,
+      "loss": 9.6756,
+      "step": 44320
+    },
+    {
+      "epoch": 0.2213788109565782,
+      "grad_norm": 0.09680955857038498,
+      "learning_rate": 2.341710681118426e-05,
+      "loss": 9.6781,
+      "step": 44330
+    },
+    {
+      "epoch": 0.22142874978151764,
+      "grad_norm": 0.09705173224210739,
+      "learning_rate": 2.3415604896242713e-05,
+      "loss": 9.6713,
+      "step": 44340
+    },
+    {
+      "epoch": 0.22147868860645709,
+      "grad_norm": 0.09335210174322128,
+      "learning_rate": 2.341410298130116e-05,
+      "loss": 9.6774,
+      "step": 44350
+    },
+    {
+      "epoch": 0.22152862743139654,
+      "grad_norm": 0.09018994867801666,
+      "learning_rate": 2.3412601066359607e-05,
+      "loss": 9.673,
+      "step": 44360
+    },
+    {
+      "epoch": 0.22157856625633598,
+      "grad_norm": 0.09429071098566055,
+      "learning_rate": 2.341109915141806e-05,
+      "loss": 9.681,
+      "step": 44370
+    },
+    {
+      "epoch": 0.22162850508127543,
+      "grad_norm": 0.09190639108419418,
+      "learning_rate": 2.3409597236476507e-05,
+      "loss": 9.6753,
+      "step": 44380
+    },
+    {
+      "epoch": 0.22167844390621488,
+      "grad_norm": 0.10482806712388992,
+      "learning_rate": 2.340809532153496e-05,
+      "loss": 9.6804,
+      "step": 44390
+    },
+    {
+      "epoch": 0.22172838273115433,
+      "grad_norm": 0.09960668534040451,
+      "learning_rate": 2.3406593406593407e-05,
+      "loss": 9.6752,
+      "step": 44400
+    },
+    {
+      "epoch": 0.22177832155609378,
+      "grad_norm": 0.09253636002540588,
+      "learning_rate": 2.3405091491651854e-05,
+      "loss": 9.6777,
+      "step": 44410
+    },
+    {
+      "epoch": 0.22182826038103323,
+      "grad_norm": 0.09326852858066559,
+      "learning_rate": 2.3403589576710308e-05,
+      "loss": 9.673,
+      "step": 44420
+    },
+    {
+      "epoch": 0.22187819920597268,
+      "grad_norm": 0.0913088470697403,
+      "learning_rate": 2.3402087661768754e-05,
+      "loss": 9.6732,
+      "step": 44430
+    },
+    {
+      "epoch": 0.22192813803091213,
+      "grad_norm": 0.0977054163813591,
+      "learning_rate": 2.3400585746827208e-05,
+      "loss": 9.671,
+      "step": 44440
+    },
+    {
+      "epoch": 0.22197807685585158,
+      "grad_norm": 0.09495818614959717,
+      "learning_rate": 2.3399083831885655e-05,
+      "loss": 9.6742,
+      "step": 44450
+    },
+    {
+      "epoch": 0.22202801568079103,
+      "grad_norm": 0.09410752356052399,
+      "learning_rate": 2.33975819169441e-05,
+      "loss": 9.6658,
+      "step": 44460
+    },
+    {
+      "epoch": 0.22207795450573048,
+      "grad_norm": 0.09136299043893814,
+      "learning_rate": 2.3396080002002555e-05,
+      "loss": 9.6762,
+      "step": 44470
+    },
+    {
+      "epoch": 0.22212789333066993,
+      "grad_norm": 0.09776278585195541,
+      "learning_rate": 2.3394578087061002e-05,
+      "loss": 9.6718,
+      "step": 44480
+    },
+    {
+      "epoch": 0.22217783215560938,
+      "grad_norm": 0.09050683677196503,
+      "learning_rate": 2.3393076172119455e-05,
+      "loss": 9.6827,
+      "step": 44490
+    },
+    {
+      "epoch": 0.22222777098054883,
+      "grad_norm": 0.09264349937438965,
+      "learning_rate": 2.3391574257177902e-05,
+      "loss": 9.6754,
+      "step": 44500
+    },
+    {
+      "epoch": 0.22227770980548828,
+      "grad_norm": 0.09283478558063507,
+      "learning_rate": 2.339007234223635e-05,
+      "loss": 9.6718,
+      "step": 44510
+    },
+    {
+      "epoch": 0.22232764863042773,
+      "grad_norm": 0.09404191374778748,
+      "learning_rate": 2.3388570427294803e-05,
+      "loss": 9.6742,
+      "step": 44520
+    },
+    {
+      "epoch": 0.22237758745536718,
+      "grad_norm": 0.09505549818277359,
+      "learning_rate": 2.338706851235325e-05,
+      "loss": 9.6828,
+      "step": 44530
+    },
+    {
+      "epoch": 0.22242752628030663,
+      "grad_norm": 0.09772829711437225,
+      "learning_rate": 2.3385566597411703e-05,
+      "loss": 9.6717,
+      "step": 44540
+    },
+    {
+      "epoch": 0.22247746510524607,
+      "grad_norm": 0.09645674377679825,
+      "learning_rate": 2.338406468247015e-05,
+      "loss": 9.6705,
+      "step": 44550
+    },
+    {
+      "epoch": 0.22252740393018552,
+      "grad_norm": 0.0913274735212326,
+      "learning_rate": 2.3382562767528597e-05,
+      "loss": 9.6726,
+      "step": 44560
+    },
+    {
+      "epoch": 0.22257734275512497,
+      "grad_norm": 0.09277801215648651,
+      "learning_rate": 2.338106085258705e-05,
+      "loss": 9.6706,
+      "step": 44570
+    },
+    {
+      "epoch": 0.22262728158006442,
+      "grad_norm": 0.09037318825721741,
+      "learning_rate": 2.3379558937645497e-05,
+      "loss": 9.6727,
+      "step": 44580
+    },
+    {
+      "epoch": 0.22267722040500387,
+      "grad_norm": 0.09477484226226807,
+      "learning_rate": 2.337805702270395e-05,
+      "loss": 9.6708,
+      "step": 44590
+    },
+    {
+      "epoch": 0.22272715922994332,
+      "grad_norm": 0.09787924587726593,
+      "learning_rate": 2.3376555107762397e-05,
+      "loss": 9.6842,
+      "step": 44600
+    },
+    {
+      "epoch": 0.22277709805488277,
+      "grad_norm": 0.09797964245080948,
+      "learning_rate": 2.3375053192820844e-05,
+      "loss": 9.6674,
+      "step": 44610
+    },
+    {
+      "epoch": 0.22282703687982222,
+      "grad_norm": 0.09463818371295929,
+      "learning_rate": 2.3373551277879298e-05,
+      "loss": 9.6736,
+      "step": 44620
+    },
+    {
+      "epoch": 0.22287697570476167,
+      "grad_norm": 0.09300590306520462,
+      "learning_rate": 2.3372049362937744e-05,
+      "loss": 9.6651,
+      "step": 44630
+    },
+    {
+      "epoch": 0.22292691452970112,
+      "grad_norm": 0.09273765981197357,
+      "learning_rate": 2.3370547447996198e-05,
+      "loss": 9.6731,
+      "step": 44640
+    },
+    {
+      "epoch": 0.22297685335464057,
+      "grad_norm": 0.09590218961238861,
+      "learning_rate": 2.3369045533054645e-05,
+      "loss": 9.6727,
+      "step": 44650
+    },
+    {
+      "epoch": 0.22302679217958002,
+      "grad_norm": 0.09408383816480637,
+      "learning_rate": 2.3367543618113095e-05,
+      "loss": 9.6691,
+      "step": 44660
+    },
+    {
+      "epoch": 0.22307673100451947,
+      "grad_norm": 0.101545050740242,
+      "learning_rate": 2.3366041703171545e-05,
+      "loss": 9.6705,
+      "step": 44670
+    },
+    {
+      "epoch": 0.22312666982945892,
+      "grad_norm": 0.0954471305012703,
+      "learning_rate": 2.3364539788229992e-05,
+      "loss": 9.6699,
+      "step": 44680
+    },
+    {
+      "epoch": 0.22317660865439837,
+      "grad_norm": 0.0932590514421463,
+      "learning_rate": 2.3363037873288445e-05,
+      "loss": 9.669,
+      "step": 44690
+    },
+    {
+      "epoch": 0.22322654747933782,
+      "grad_norm": 0.09442409873008728,
+      "learning_rate": 2.3361535958346892e-05,
+      "loss": 9.6754,
+      "step": 44700
+    },
+    {
+      "epoch": 0.22327648630427727,
+      "grad_norm": 0.09490916877985,
+      "learning_rate": 2.3360034043405342e-05,
+      "loss": 9.6672,
+      "step": 44710
+    },
+    {
+      "epoch": 0.22332642512921672,
+      "grad_norm": 0.0963781476020813,
+      "learning_rate": 2.3358532128463793e-05,
+      "loss": 9.6703,
+      "step": 44720
+    },
+    {
+      "epoch": 0.22337636395415617,
+      "grad_norm": 0.09573322534561157,
+      "learning_rate": 2.335703021352224e-05,
+      "loss": 9.6784,
+      "step": 44730
+    },
+    {
+      "epoch": 0.22342630277909561,
+      "grad_norm": 0.09953943639993668,
+      "learning_rate": 2.3355528298580693e-05,
+      "loss": 9.6667,
+      "step": 44740
+    },
+    {
+      "epoch": 0.22347624160403506,
+      "grad_norm": 0.0922081395983696,
+      "learning_rate": 2.335402638363914e-05,
+      "loss": 9.6731,
+      "step": 44750
+    },
+    {
+      "epoch": 0.2235261804289745,
+      "grad_norm": 0.09071052074432373,
+      "learning_rate": 2.335252446869759e-05,
+      "loss": 9.6642,
+      "step": 44760
+    },
+    {
+      "epoch": 0.22357611925391396,
+      "grad_norm": 0.09215296804904938,
+      "learning_rate": 2.335102255375604e-05,
+      "loss": 9.667,
+      "step": 44770
+    },
+    {
+      "epoch": 0.2236260580788534,
+      "grad_norm": 0.10084545612335205,
+      "learning_rate": 2.3349520638814487e-05,
+      "loss": 9.6648,
+      "step": 44780
+    },
+    {
+      "epoch": 0.22367599690379286,
+      "grad_norm": 0.09967309981584549,
+      "learning_rate": 2.334801872387294e-05,
+      "loss": 9.659,
+      "step": 44790
+    },
+    {
+      "epoch": 0.2237259357287323,
+      "grad_norm": 0.09629634022712708,
+      "learning_rate": 2.3346516808931387e-05,
+      "loss": 9.6709,
+      "step": 44800
+    },
+    {
+      "epoch": 0.22377587455367176,
+      "grad_norm": 0.09285489469766617,
+      "learning_rate": 2.3345014893989837e-05,
+      "loss": 9.6672,
+      "step": 44810
+    },
+    {
+      "epoch": 0.2238258133786112,
+      "grad_norm": 0.09779229015111923,
+      "learning_rate": 2.3343512979048288e-05,
+      "loss": 9.6669,
+      "step": 44820
+    },
+    {
+      "epoch": 0.22387575220355066,
+      "grad_norm": 0.09266447275876999,
+      "learning_rate": 2.3342011064106734e-05,
+      "loss": 9.6644,
+      "step": 44830
+    },
+    {
+      "epoch": 0.2239256910284901,
+      "grad_norm": 0.09615839272737503,
+      "learning_rate": 2.3340509149165188e-05,
+      "loss": 9.6617,
+      "step": 44840
+    },
+    {
+      "epoch": 0.22397562985342956,
+      "grad_norm": 0.09608431905508041,
+      "learning_rate": 2.3339007234223635e-05,
+      "loss": 9.663,
+      "step": 44850
+    },
+    {
+      "epoch": 0.224025568678369,
+      "grad_norm": 0.09576740115880966,
+      "learning_rate": 2.3337505319282085e-05,
+      "loss": 9.6596,
+      "step": 44860
+    },
+    {
+      "epoch": 0.22407550750330846,
+      "grad_norm": 0.09225677698850632,
+      "learning_rate": 2.3336003404340535e-05,
+      "loss": 9.6597,
+      "step": 44870
+    },
+    {
+      "epoch": 0.2241254463282479,
+      "grad_norm": 0.09069118648767471,
+      "learning_rate": 2.3334501489398982e-05,
+      "loss": 9.6613,
+      "step": 44880
+    },
+    {
+      "epoch": 0.22417538515318736,
+      "grad_norm": 0.10063677281141281,
+      "learning_rate": 2.3332999574457435e-05,
+      "loss": 9.6712,
+      "step": 44890
+    },
+    {
+      "epoch": 0.2242253239781268,
+      "grad_norm": 0.0957692414522171,
+      "learning_rate": 2.3331497659515882e-05,
+      "loss": 9.6627,
+      "step": 44900
+    },
+    {
+      "epoch": 0.22427526280306626,
+      "grad_norm": 0.0920589417219162,
+      "learning_rate": 2.3329995744574332e-05,
+      "loss": 9.6683,
+      "step": 44910
+    },
+    {
+      "epoch": 0.2243252016280057,
+      "grad_norm": 0.09941014647483826,
+      "learning_rate": 2.3328493829632783e-05,
+      "loss": 9.6632,
+      "step": 44920
+    },
+    {
+      "epoch": 0.22437514045294515,
+      "grad_norm": 0.09040158987045288,
+      "learning_rate": 2.332699191469123e-05,
+      "loss": 9.6632,
+      "step": 44930
+    },
+    {
+      "epoch": 0.2244250792778846,
+      "grad_norm": 0.09432369470596313,
+      "learning_rate": 2.3325489999749683e-05,
+      "loss": 9.6624,
+      "step": 44940
+    },
+    {
+      "epoch": 0.22447501810282405,
+      "grad_norm": 0.09329992532730103,
+      "learning_rate": 2.332398808480813e-05,
+      "loss": 9.6735,
+      "step": 44950
+    },
+    {
+      "epoch": 0.2245249569277635,
+      "grad_norm": 0.09206376224756241,
+      "learning_rate": 2.332248616986658e-05,
+      "loss": 9.6687,
+      "step": 44960
+    },
+    {
+      "epoch": 0.22457489575270295,
+      "grad_norm": 0.09406663477420807,
+      "learning_rate": 2.332098425492503e-05,
+      "loss": 9.6687,
+      "step": 44970
+    },
+    {
+      "epoch": 0.2246248345776424,
+      "grad_norm": 0.09512316435575485,
+      "learning_rate": 2.331948233998348e-05,
+      "loss": 9.6635,
+      "step": 44980
+    },
+    {
+      "epoch": 0.22467477340258185,
+      "grad_norm": 0.09398701786994934,
+      "learning_rate": 2.331798042504193e-05,
+      "loss": 9.6649,
+      "step": 44990
+    },
+    {
+      "epoch": 0.22472471222752127,
+      "grad_norm": 0.09211496263742447,
+      "learning_rate": 2.3316478510100377e-05,
+      "loss": 9.6654,
+      "step": 45000
+    },
+    {
+      "epoch": 0.22477465105246072,
+      "grad_norm": 0.09733974188566208,
+      "learning_rate": 2.3314976595158827e-05,
+      "loss": 9.6564,
+      "step": 45010
+    },
+    {
+      "epoch": 0.22482458987740017,
+      "grad_norm": 0.09194665402173996,
+      "learning_rate": 2.3313474680217278e-05,
+      "loss": 9.6596,
+      "step": 45020
+    },
+    {
+      "epoch": 0.22487452870233962,
+      "grad_norm": 0.08959387242794037,
+      "learning_rate": 2.3311972765275728e-05,
+      "loss": 9.6623,
+      "step": 45030
+    },
+    {
+      "epoch": 0.22492446752727907,
+      "grad_norm": 0.09784001857042313,
+      "learning_rate": 2.3310470850334178e-05,
+      "loss": 9.6599,
+      "step": 45040
+    },
+    {
+      "epoch": 0.22497440635221852,
+      "grad_norm": 0.09329735487699509,
+      "learning_rate": 2.3308968935392625e-05,
+      "loss": 9.6656,
+      "step": 45050
+    },
+    {
+      "epoch": 0.22502434517715797,
+      "grad_norm": 0.08951118588447571,
+      "learning_rate": 2.3307467020451075e-05,
+      "loss": 9.6618,
+      "step": 45060
+    },
+    {
+      "epoch": 0.22507428400209742,
+      "grad_norm": 0.09492070972919464,
+      "learning_rate": 2.3305965105509525e-05,
+      "loss": 9.6589,
+      "step": 45070
+    },
+    {
+      "epoch": 0.22512422282703687,
+      "grad_norm": 0.09987737983465195,
+      "learning_rate": 2.3304463190567975e-05,
+      "loss": 9.6615,
+      "step": 45080
+    },
+    {
+      "epoch": 0.22517416165197632,
+      "grad_norm": 0.10113034397363663,
+      "learning_rate": 2.3302961275626425e-05,
+      "loss": 9.6608,
+      "step": 45090
+    },
+    {
+      "epoch": 0.22522410047691577,
+      "grad_norm": 0.09234467148780823,
+      "learning_rate": 2.3301459360684872e-05,
+      "loss": 9.669,
+      "step": 45100
+    },
+    {
+      "epoch": 0.22527403930185522,
+      "grad_norm": 0.09484550356864929,
+      "learning_rate": 2.3299957445743322e-05,
+      "loss": 9.6537,
+      "step": 45110
+    },
+    {
+      "epoch": 0.22532397812679467,
+      "grad_norm": 0.09227019548416138,
+      "learning_rate": 2.3298455530801773e-05,
+      "loss": 9.6642,
+      "step": 45120
+    },
+    {
+      "epoch": 0.22537391695173412,
+      "grad_norm": 0.0964643582701683,
+      "learning_rate": 2.3296953615860223e-05,
+      "loss": 9.6621,
+      "step": 45130
+    },
+    {
+      "epoch": 0.22542385577667357,
+      "grad_norm": 0.09387196600437164,
+      "learning_rate": 2.3295451700918673e-05,
+      "loss": 9.6668,
+      "step": 45140
+    },
+    {
+      "epoch": 0.22547379460161301,
+      "grad_norm": 0.08830858767032623,
+      "learning_rate": 2.329394978597712e-05,
+      "loss": 9.6616,
+      "step": 45150
+    },
+    {
+      "epoch": 0.22552373342655246,
+      "grad_norm": 0.09885809570550919,
+      "learning_rate": 2.329244787103557e-05,
+      "loss": 9.6554,
+      "step": 45160
+    },
+    {
+      "epoch": 0.2255736722514919,
+      "grad_norm": 0.09711761027574539,
+      "learning_rate": 2.329094595609402e-05,
+      "loss": 9.6619,
+      "step": 45170
+    },
+    {
+      "epoch": 0.22562361107643136,
+      "grad_norm": 0.0906030535697937,
+      "learning_rate": 2.328944404115247e-05,
+      "loss": 9.661,
+      "step": 45180
+    },
+    {
+      "epoch": 0.2256735499013708,
+      "grad_norm": 0.09631696343421936,
+      "learning_rate": 2.328794212621092e-05,
+      "loss": 9.6565,
+      "step": 45190
+    },
+    {
+      "epoch": 0.22572348872631026,
+      "grad_norm": 0.09392791241407394,
+      "learning_rate": 2.3286440211269367e-05,
+      "loss": 9.6611,
+      "step": 45200
+    },
+    {
+      "epoch": 0.2257734275512497,
+      "grad_norm": 0.10017602145671844,
+      "learning_rate": 2.3284938296327817e-05,
+      "loss": 9.6605,
+      "step": 45210
+    },
+    {
+      "epoch": 0.22582336637618916,
+      "grad_norm": 0.09598571807146072,
+      "learning_rate": 2.3283436381386268e-05,
+      "loss": 9.664,
+      "step": 45220
+    },
+    {
+      "epoch": 0.2258733052011286,
+      "grad_norm": 0.09706757217645645,
+      "learning_rate": 2.3281934466444718e-05,
+      "loss": 9.6531,
+      "step": 45230
+    },
+    {
+      "epoch": 0.22592324402606806,
+      "grad_norm": 0.09351912885904312,
+      "learning_rate": 2.3280432551503168e-05,
+      "loss": 9.6569,
+      "step": 45240
+    },
+    {
+      "epoch": 0.2259731828510075,
+      "grad_norm": 0.08977209031581879,
+      "learning_rate": 2.3278930636561615e-05,
+      "loss": 9.6535,
+      "step": 45250
+    },
+    {
+      "epoch": 0.22602312167594696,
+      "grad_norm": 0.09070810675621033,
+      "learning_rate": 2.3277428721620065e-05,
+      "loss": 9.6611,
+      "step": 45260
+    },
+    {
+      "epoch": 0.2260730605008864,
+      "grad_norm": 0.09817781299352646,
+      "learning_rate": 2.3275926806678515e-05,
+      "loss": 9.654,
+      "step": 45270
+    },
+    {
+      "epoch": 0.22612299932582586,
+      "grad_norm": 0.09307067096233368,
+      "learning_rate": 2.3274424891736965e-05,
+      "loss": 9.6607,
+      "step": 45280
+    },
+    {
+      "epoch": 0.2261729381507653,
+      "grad_norm": 0.09242681413888931,
+      "learning_rate": 2.3272922976795415e-05,
+      "loss": 9.6639,
+      "step": 45290
+    },
+    {
+      "epoch": 0.22622287697570476,
+      "grad_norm": 0.09181907773017883,
+      "learning_rate": 2.3271421061853866e-05,
+      "loss": 9.653,
+      "step": 45300
+    },
+    {
+      "epoch": 0.2262728158006442,
+      "grad_norm": 0.09517987072467804,
+      "learning_rate": 2.3269919146912312e-05,
+      "loss": 9.6602,
+      "step": 45310
+    },
+    {
+      "epoch": 0.22632275462558366,
+      "grad_norm": 0.0962613895535469,
+      "learning_rate": 2.3268417231970763e-05,
+      "loss": 9.6629,
+      "step": 45320
+    },
+    {
+      "epoch": 0.2263726934505231,
+      "grad_norm": 0.09088754653930664,
+      "learning_rate": 2.3266915317029213e-05,
+      "loss": 9.6521,
+      "step": 45330
+    },
+    {
+      "epoch": 0.22642263227546255,
+      "grad_norm": 0.09162914752960205,
+      "learning_rate": 2.3265413402087663e-05,
+      "loss": 9.6579,
+      "step": 45340
+    },
+    {
+      "epoch": 0.226472571100402,
+      "grad_norm": 0.09387911111116409,
+      "learning_rate": 2.3263911487146113e-05,
+      "loss": 9.6558,
+      "step": 45350
+    },
+    {
+      "epoch": 0.22652250992534145,
+      "grad_norm": 0.08971641957759857,
+      "learning_rate": 2.326240957220456e-05,
+      "loss": 9.6652,
+      "step": 45360
+    },
+    {
+      "epoch": 0.2265724487502809,
+      "grad_norm": 0.090997114777565,
+      "learning_rate": 2.326090765726301e-05,
+      "loss": 9.6571,
+      "step": 45370
+    },
+    {
+      "epoch": 0.22662238757522035,
+      "grad_norm": 0.09356195479631424,
+      "learning_rate": 2.325940574232146e-05,
+      "loss": 9.6564,
+      "step": 45380
+    },
+    {
+      "epoch": 0.2266723264001598,
+      "grad_norm": 0.09340345114469528,
+      "learning_rate": 2.325790382737991e-05,
+      "loss": 9.6616,
+      "step": 45390
+    },
+    {
+      "epoch": 0.22672226522509925,
+      "grad_norm": 0.09831879287958145,
+      "learning_rate": 2.325640191243836e-05,
+      "loss": 9.6553,
+      "step": 45400
+    },
+    {
+      "epoch": 0.2267722040500387,
+      "grad_norm": 0.0913265198469162,
+      "learning_rate": 2.3254899997496807e-05,
+      "loss": 9.6529,
+      "step": 45410
+    },
+    {
+      "epoch": 0.22682214287497815,
+      "grad_norm": 0.09678717702627182,
+      "learning_rate": 2.3253398082555258e-05,
+      "loss": 9.657,
+      "step": 45420
+    },
+    {
+      "epoch": 0.2268720816999176,
+      "grad_norm": 0.0980067104101181,
+      "learning_rate": 2.3251896167613708e-05,
+      "loss": 9.6614,
+      "step": 45430
+    },
+    {
+      "epoch": 0.22692202052485705,
+      "grad_norm": 0.09428383409976959,
+      "learning_rate": 2.3250394252672158e-05,
+      "loss": 9.6596,
+      "step": 45440
+    },
+    {
+      "epoch": 0.2269719593497965,
+      "grad_norm": 0.09209676086902618,
+      "learning_rate": 2.3248892337730608e-05,
+      "loss": 9.6553,
+      "step": 45450
+    },
+    {
+      "epoch": 0.22702189817473595,
+      "grad_norm": 0.09065406024456024,
+      "learning_rate": 2.3247390422789055e-05,
+      "loss": 9.6595,
+      "step": 45460
+    },
+    {
+      "epoch": 0.2270718369996754,
+      "grad_norm": 0.0918533131480217,
+      "learning_rate": 2.3245888507847505e-05,
+      "loss": 9.6571,
+      "step": 45470
+    },
+    {
+      "epoch": 0.22712177582461485,
+      "grad_norm": 0.0916934683918953,
+      "learning_rate": 2.3244386592905955e-05,
+      "loss": 9.667,
+      "step": 45480
+    },
+    {
+      "epoch": 0.2271717146495543,
+      "grad_norm": 0.09149292856454849,
+      "learning_rate": 2.3242884677964406e-05,
+      "loss": 9.6524,
+      "step": 45490
+    },
+    {
+      "epoch": 0.22722165347449375,
+      "grad_norm": 0.09491834789514542,
+      "learning_rate": 2.3241382763022856e-05,
+      "loss": 9.6487,
+      "step": 45500
+    },
+    {
+      "epoch": 0.2272715922994332,
+      "grad_norm": 0.09356304258108139,
+      "learning_rate": 2.3239880848081302e-05,
+      "loss": 9.6529,
+      "step": 45510
+    },
+    {
+      "epoch": 0.22732153112437264,
+      "grad_norm": 0.09925433993339539,
+      "learning_rate": 2.3238378933139753e-05,
+      "loss": 9.6515,
+      "step": 45520
+    },
+    {
+      "epoch": 0.2273714699493121,
+      "grad_norm": 0.09688141942024231,
+      "learning_rate": 2.3236877018198203e-05,
+      "loss": 9.656,
+      "step": 45530
+    },
+    {
+      "epoch": 0.22742140877425154,
+      "grad_norm": 0.08831819891929626,
+      "learning_rate": 2.3235375103256653e-05,
+      "loss": 9.6533,
+      "step": 45540
+    },
+    {
+      "epoch": 0.227471347599191,
+      "grad_norm": 0.0944901704788208,
+      "learning_rate": 2.3233873188315103e-05,
+      "loss": 9.657,
+      "step": 45550
+    },
+    {
+      "epoch": 0.22752128642413044,
+      "grad_norm": 0.0912630707025528,
+      "learning_rate": 2.323237127337355e-05,
+      "loss": 9.6442,
+      "step": 45560
+    },
+    {
+      "epoch": 0.2275712252490699,
+      "grad_norm": 0.09489890933036804,
+      "learning_rate": 2.3230869358432e-05,
+      "loss": 9.6535,
+      "step": 45570
+    },
+    {
+      "epoch": 0.22762116407400934,
+      "grad_norm": 0.09343596547842026,
+      "learning_rate": 2.322936744349045e-05,
+      "loss": 9.6522,
+      "step": 45580
+    },
+    {
+      "epoch": 0.2276711028989488,
+      "grad_norm": 0.0928681418299675,
+      "learning_rate": 2.32278655285489e-05,
+      "loss": 9.6557,
+      "step": 45590
+    },
+    {
+      "epoch": 0.22772104172388824,
+      "grad_norm": 0.09620653092861176,
+      "learning_rate": 2.322636361360735e-05,
+      "loss": 9.6563,
+      "step": 45600
+    },
+    {
+      "epoch": 0.2277709805488277,
+      "grad_norm": 0.09333667159080505,
+      "learning_rate": 2.3224861698665797e-05,
+      "loss": 9.6608,
+      "step": 45610
+    },
+    {
+      "epoch": 0.22782091937376714,
+      "grad_norm": 0.0948384627699852,
+      "learning_rate": 2.322335978372425e-05,
+      "loss": 9.6551,
+      "step": 45620
+    },
+    {
+      "epoch": 0.2278708581987066,
+      "grad_norm": 0.09374862164258957,
+      "learning_rate": 2.3221857868782698e-05,
+      "loss": 9.6495,
+      "step": 45630
+    },
+    {
+      "epoch": 0.22792079702364604,
+      "grad_norm": 0.09943927079439163,
+      "learning_rate": 2.3220355953841148e-05,
+      "loss": 9.657,
+      "step": 45640
+    },
+    {
+      "epoch": 0.2279707358485855,
+      "grad_norm": 0.09763931483030319,
+      "learning_rate": 2.3218854038899598e-05,
+      "loss": 9.6523,
+      "step": 45650
+    },
+    {
+      "epoch": 0.22802067467352494,
+      "grad_norm": 0.09670069813728333,
+      "learning_rate": 2.3217352123958045e-05,
+      "loss": 9.652,
+      "step": 45660
+    },
+    {
+      "epoch": 0.2280706134984644,
+      "grad_norm": 0.09232018142938614,
+      "learning_rate": 2.32158502090165e-05,
+      "loss": 9.657,
+      "step": 45670
+    },
+    {
+      "epoch": 0.22812055232340384,
+      "grad_norm": 0.09801393002271652,
+      "learning_rate": 2.3214348294074945e-05,
+      "loss": 9.6503,
+      "step": 45680
+    },
+    {
+      "epoch": 0.22817049114834329,
+      "grad_norm": 0.09959478676319122,
+      "learning_rate": 2.3212846379133396e-05,
+      "loss": 9.6495,
+      "step": 45690
+    },
+    {
+      "epoch": 0.22822042997328273,
+      "grad_norm": 0.09530535340309143,
+      "learning_rate": 2.3211344464191846e-05,
+      "loss": 9.6492,
+      "step": 45700
+    },
+    {
+      "epoch": 0.22827036879822218,
+      "grad_norm": 0.08995451778173447,
+      "learning_rate": 2.3209842549250292e-05,
+      "loss": 9.6496,
+      "step": 45710
+    },
+    {
+      "epoch": 0.22832030762316163,
+      "grad_norm": 0.09303654730319977,
+      "learning_rate": 2.3208340634308746e-05,
+      "loss": 9.6446,
+      "step": 45720
+    },
+    {
+      "epoch": 0.22837024644810108,
+      "grad_norm": 0.09486031532287598,
+      "learning_rate": 2.3206838719367193e-05,
+      "loss": 9.6531,
+      "step": 45730
+    },
+    {
+      "epoch": 0.22842018527304053,
+      "grad_norm": 0.0937868058681488,
+      "learning_rate": 2.3205336804425643e-05,
+      "loss": 9.65,
+      "step": 45740
+    },
+    {
+      "epoch": 0.22847012409797998,
+      "grad_norm": 0.09792537987232208,
+      "learning_rate": 2.3203834889484093e-05,
+      "loss": 9.651,
+      "step": 45750
+    },
+    {
+      "epoch": 0.22852006292291943,
+      "grad_norm": 0.09761232137680054,
+      "learning_rate": 2.320233297454254e-05,
+      "loss": 9.6487,
+      "step": 45760
+    },
+    {
+      "epoch": 0.22857000174785888,
+      "grad_norm": 0.09970896691083908,
+      "learning_rate": 2.3200831059600994e-05,
+      "loss": 9.6473,
+      "step": 45770
+    },
+    {
+      "epoch": 0.22861994057279833,
+      "grad_norm": 0.096064992249012,
+      "learning_rate": 2.319932914465944e-05,
+      "loss": 9.6548,
+      "step": 45780
+    },
+    {
+      "epoch": 0.22866987939773778,
+      "grad_norm": 0.09543237090110779,
+      "learning_rate": 2.319782722971789e-05,
+      "loss": 9.6464,
+      "step": 45790
+    },
+    {
+      "epoch": 0.22871981822267723,
+      "grad_norm": 0.09486628323793411,
+      "learning_rate": 2.319632531477634e-05,
+      "loss": 9.6529,
+      "step": 45800
+    },
+    {
+      "epoch": 0.22876975704761668,
+      "grad_norm": 0.09722953289747238,
+      "learning_rate": 2.3194823399834788e-05,
+      "loss": 9.6543,
+      "step": 45810
+    },
+    {
+      "epoch": 0.22881969587255613,
+      "grad_norm": 0.09415905922651291,
+      "learning_rate": 2.319332148489324e-05,
+      "loss": 9.6438,
+      "step": 45820
+    },
+    {
+      "epoch": 0.22886963469749558,
+      "grad_norm": 0.08979658782482147,
+      "learning_rate": 2.3191819569951688e-05,
+      "loss": 9.653,
+      "step": 45830
+    },
+    {
+      "epoch": 0.22891957352243503,
+      "grad_norm": 0.09261351078748703,
+      "learning_rate": 2.3190317655010138e-05,
+      "loss": 9.6534,
+      "step": 45840
+    },
+    {
+      "epoch": 0.22896951234737448,
+      "grad_norm": 0.09331527352333069,
+      "learning_rate": 2.3188815740068588e-05,
+      "loss": 9.6481,
+      "step": 45850
+    },
+    {
+      "epoch": 0.22901945117231393,
+      "grad_norm": 0.09803377836942673,
+      "learning_rate": 2.3187313825127035e-05,
+      "loss": 9.6467,
+      "step": 45860
+    },
+    {
+      "epoch": 0.22906938999725338,
+      "grad_norm": 0.09804876148700714,
+      "learning_rate": 2.318581191018549e-05,
+      "loss": 9.6431,
+      "step": 45870
+    },
+    {
+      "epoch": 0.22911932882219282,
+      "grad_norm": 0.09359174221754074,
+      "learning_rate": 2.3184309995243935e-05,
+      "loss": 9.6487,
+      "step": 45880
+    },
+    {
+      "epoch": 0.22916926764713227,
+      "grad_norm": 0.09482156485319138,
+      "learning_rate": 2.3182808080302386e-05,
+      "loss": 9.6482,
+      "step": 45890
+    },
+    {
+      "epoch": 0.22921920647207172,
+      "grad_norm": 0.09959374368190765,
+      "learning_rate": 2.3181306165360836e-05,
+      "loss": 9.6447,
+      "step": 45900
+    },
+    {
+      "epoch": 0.22926914529701117,
+      "grad_norm": 0.09299067407846451,
+      "learning_rate": 2.3179804250419283e-05,
+      "loss": 9.6411,
+      "step": 45910
+    },
+    {
+      "epoch": 0.22931908412195062,
+      "grad_norm": 0.0913899689912796,
+      "learning_rate": 2.3178302335477736e-05,
+      "loss": 9.6476,
+      "step": 45920
+    },
+    {
+      "epoch": 0.22936902294689007,
+      "grad_norm": 0.09546446055173874,
+      "learning_rate": 2.3176800420536183e-05,
+      "loss": 9.6434,
+      "step": 45930
+    },
+    {
+      "epoch": 0.22941896177182952,
+      "grad_norm": 0.09306006133556366,
+      "learning_rate": 2.3175298505594636e-05,
+      "loss": 9.6393,
+      "step": 45940
+    },
+    {
+      "epoch": 0.22946890059676897,
+      "grad_norm": 0.09446984529495239,
+      "learning_rate": 2.3173796590653083e-05,
+      "loss": 9.6481,
+      "step": 45950
+    },
+    {
+      "epoch": 0.22951883942170842,
+      "grad_norm": 0.09467601031064987,
+      "learning_rate": 2.317229467571153e-05,
+      "loss": 9.6485,
+      "step": 45960
+    },
+    {
+      "epoch": 0.22956877824664787,
+      "grad_norm": 0.09725523740053177,
+      "learning_rate": 2.3170792760769984e-05,
+      "loss": 9.6504,
+      "step": 45970
+    },
+    {
+      "epoch": 0.22961871707158732,
+      "grad_norm": 0.09193247556686401,
+      "learning_rate": 2.316929084582843e-05,
+      "loss": 9.6495,
+      "step": 45980
+    },
+    {
+      "epoch": 0.22966865589652674,
+      "grad_norm": 0.09462161362171173,
+      "learning_rate": 2.3167788930886884e-05,
+      "loss": 9.6387,
+      "step": 45990
+    },
+    {
+      "epoch": 0.2297185947214662,
+      "grad_norm": 0.09702228009700775,
+      "learning_rate": 2.316628701594533e-05,
+      "loss": 9.6417,
+      "step": 46000
+    },
+    {
+      "epoch": 0.22976853354640564,
+      "grad_norm": 0.09509331732988358,
+      "learning_rate": 2.316478510100378e-05,
+      "loss": 9.6414,
+      "step": 46010
+    },
+    {
+      "epoch": 0.2298184723713451,
+      "grad_norm": 0.09766579419374466,
+      "learning_rate": 2.316328318606223e-05,
+      "loss": 9.6406,
+      "step": 46020
+    },
+    {
+      "epoch": 0.22986841119628454,
+      "grad_norm": 0.10346722602844238,
+      "learning_rate": 2.3161781271120678e-05,
+      "loss": 9.6464,
+      "step": 46030
+    },
+    {
+      "epoch": 0.229918350021224,
+      "grad_norm": 0.09228979051113129,
+      "learning_rate": 2.316027935617913e-05,
+      "loss": 9.6473,
+      "step": 46040
+    },
+    {
+      "epoch": 0.22996828884616344,
+      "grad_norm": 0.10133613646030426,
+      "learning_rate": 2.3158777441237578e-05,
+      "loss": 9.6507,
+      "step": 46050
+    },
+    {
+      "epoch": 0.2300182276711029,
+      "grad_norm": 0.09652470797300339,
+      "learning_rate": 2.315727552629603e-05,
+      "loss": 9.6408,
+      "step": 46060
+    },
+    {
+      "epoch": 0.23006816649604234,
+      "grad_norm": 0.09428387135267258,
+      "learning_rate": 2.315577361135448e-05,
+      "loss": 9.6406,
+      "step": 46070
+    },
+    {
+      "epoch": 0.2301181053209818,
+      "grad_norm": 0.09347372502088547,
+      "learning_rate": 2.3154271696412925e-05,
+      "loss": 9.6449,
+      "step": 46080
+    },
+    {
+      "epoch": 0.23016804414592124,
+      "grad_norm": 0.09821302443742752,
+      "learning_rate": 2.315276978147138e-05,
+      "loss": 9.6412,
+      "step": 46090
+    },
+    {
+      "epoch": 0.23021798297086069,
+      "grad_norm": 0.09411518275737762,
+      "learning_rate": 2.3151267866529826e-05,
+      "loss": 9.6504,
+      "step": 46100
+    },
+    {
+      "epoch": 0.23026792179580013,
+      "grad_norm": 0.09326085448265076,
+      "learning_rate": 2.3149765951588276e-05,
+      "loss": 9.6424,
+      "step": 46110
+    },
+    {
+      "epoch": 0.23031786062073958,
+      "grad_norm": 0.09685725718736649,
+      "learning_rate": 2.3148264036646726e-05,
+      "loss": 9.6439,
+      "step": 46120
+    },
+    {
+      "epoch": 0.23036779944567903,
+      "grad_norm": 0.09456616640090942,
+      "learning_rate": 2.3146762121705173e-05,
+      "loss": 9.649,
+      "step": 46130
+    },
+    {
+      "epoch": 0.23041773827061848,
+      "grad_norm": 0.08801116794347763,
+      "learning_rate": 2.3145260206763626e-05,
+      "loss": 9.6457,
+      "step": 46140
+    },
+    {
+      "epoch": 0.23046767709555793,
+      "grad_norm": 0.09126181155443192,
+      "learning_rate": 2.3143758291822073e-05,
+      "loss": 9.6452,
+      "step": 46150
+    },
+    {
+      "epoch": 0.23051761592049738,
+      "grad_norm": 0.0932285413146019,
+      "learning_rate": 2.3142256376880523e-05,
+      "loss": 9.6401,
+      "step": 46160
+    },
+    {
+      "epoch": 0.23056755474543683,
+      "grad_norm": 0.10316822677850723,
+      "learning_rate": 2.3140754461938974e-05,
+      "loss": 9.6444,
+      "step": 46170
+    },
+    {
+      "epoch": 0.23061749357037628,
+      "grad_norm": 0.0956098735332489,
+      "learning_rate": 2.313925254699742e-05,
+      "loss": 9.6408,
+      "step": 46180
+    },
+    {
+      "epoch": 0.23066743239531573,
+      "grad_norm": 0.09577090293169022,
+      "learning_rate": 2.3137750632055874e-05,
+      "loss": 9.6406,
+      "step": 46190
+    },
+    {
+      "epoch": 0.23071737122025518,
+      "grad_norm": 0.09496666491031647,
+      "learning_rate": 2.313624871711432e-05,
+      "loss": 9.6419,
+      "step": 46200
+    },
+    {
+      "epoch": 0.23076731004519463,
+      "grad_norm": 0.09271476417779922,
+      "learning_rate": 2.313474680217277e-05,
+      "loss": 9.6484,
+      "step": 46210
+    },
+    {
+      "epoch": 0.23081724887013408,
+      "grad_norm": 0.09295427054166794,
+      "learning_rate": 2.313324488723122e-05,
+      "loss": 9.6385,
+      "step": 46220
+    },
+    {
+      "epoch": 0.23086718769507353,
+      "grad_norm": 0.09634546935558319,
+      "learning_rate": 2.3131742972289668e-05,
+      "loss": 9.6381,
+      "step": 46230
+    },
+    {
+      "epoch": 0.23091712652001298,
+      "grad_norm": 0.09890441596508026,
+      "learning_rate": 2.313024105734812e-05,
+      "loss": 9.6422,
+      "step": 46240
+    },
+    {
+      "epoch": 0.23096706534495243,
+      "grad_norm": 0.0954676941037178,
+      "learning_rate": 2.3128739142406568e-05,
+      "loss": 9.6382,
+      "step": 46250
+    },
+    {
+      "epoch": 0.23101700416989188,
+      "grad_norm": 0.09203564375638962,
+      "learning_rate": 2.312723722746502e-05,
+      "loss": 9.6359,
+      "step": 46260
+    },
+    {
+      "epoch": 0.23106694299483133,
+      "grad_norm": 0.09297526627779007,
+      "learning_rate": 2.312573531252347e-05,
+      "loss": 9.6386,
+      "step": 46270
+    },
+    {
+      "epoch": 0.23111688181977078,
+      "grad_norm": 0.09271712601184845,
+      "learning_rate": 2.3124233397581915e-05,
+      "loss": 9.6389,
+      "step": 46280
+    },
+    {
+      "epoch": 0.23116682064471022,
+      "grad_norm": 0.09680858999490738,
+      "learning_rate": 2.312273148264037e-05,
+      "loss": 9.6354,
+      "step": 46290
+    },
+    {
+      "epoch": 0.23121675946964967,
+      "grad_norm": 0.09343792498111725,
+      "learning_rate": 2.3121229567698816e-05,
+      "loss": 9.6396,
+      "step": 46300
+    },
+    {
+      "epoch": 0.23126669829458912,
+      "grad_norm": 0.10061474144458771,
+      "learning_rate": 2.311972765275727e-05,
+      "loss": 9.6373,
+      "step": 46310
+    },
+    {
+      "epoch": 0.23131663711952857,
+      "grad_norm": 0.09283909946680069,
+      "learning_rate": 2.3118225737815716e-05,
+      "loss": 9.6334,
+      "step": 46320
+    },
+    {
+      "epoch": 0.23136657594446802,
+      "grad_norm": 0.09406852722167969,
+      "learning_rate": 2.3116723822874163e-05,
+      "loss": 9.6364,
+      "step": 46330
+    },
+    {
+      "epoch": 0.23141651476940747,
+      "grad_norm": 0.0919579491019249,
+      "learning_rate": 2.3115221907932616e-05,
+      "loss": 9.6377,
+      "step": 46340
+    },
+    {
+      "epoch": 0.23146645359434692,
+      "grad_norm": 0.09919621050357819,
+      "learning_rate": 2.3113719992991063e-05,
+      "loss": 9.6411,
+      "step": 46350
+    },
+    {
+      "epoch": 0.23151639241928637,
+      "grad_norm": 0.09455198049545288,
+      "learning_rate": 2.3112218078049517e-05,
+      "loss": 9.6333,
+      "step": 46360
+    },
+    {
+      "epoch": 0.23156633124422582,
+      "grad_norm": 0.09368839114904404,
+      "learning_rate": 2.3110716163107964e-05,
+      "loss": 9.6295,
+      "step": 46370
+    },
+    {
+      "epoch": 0.23161627006916527,
+      "grad_norm": 0.09375040233135223,
+      "learning_rate": 2.310921424816641e-05,
+      "loss": 9.6388,
+      "step": 46380
+    },
+    {
+      "epoch": 0.23166620889410472,
+      "grad_norm": 0.09624538570642471,
+      "learning_rate": 2.3107712333224864e-05,
+      "loss": 9.6401,
+      "step": 46390
+    },
+    {
+      "epoch": 0.23171614771904417,
+      "grad_norm": 0.09258920699357986,
+      "learning_rate": 2.310621041828331e-05,
+      "loss": 9.6337,
+      "step": 46400
+    },
+    {
+      "epoch": 0.23176608654398362,
+      "grad_norm": 0.09907475113868713,
+      "learning_rate": 2.3104708503341764e-05,
+      "loss": 9.632,
+      "step": 46410
+    },
+    {
+      "epoch": 0.23181602536892307,
+      "grad_norm": 0.094297394156456,
+      "learning_rate": 2.310320658840021e-05,
+      "loss": 9.6353,
+      "step": 46420
+    },
+    {
+      "epoch": 0.23186596419386252,
+      "grad_norm": 0.09265825897455215,
+      "learning_rate": 2.3101704673458658e-05,
+      "loss": 9.6436,
+      "step": 46430
+    },
+    {
+      "epoch": 0.23191590301880197,
+      "grad_norm": 0.09331150352954865,
+      "learning_rate": 2.310020275851711e-05,
+      "loss": 9.6316,
+      "step": 46440
+    },
+    {
+      "epoch": 0.23196584184374142,
+      "grad_norm": 0.0924522653222084,
+      "learning_rate": 2.3098700843575558e-05,
+      "loss": 9.6313,
+      "step": 46450
+    },
+    {
+      "epoch": 0.23201578066868087,
+      "grad_norm": 0.09283218532800674,
+      "learning_rate": 2.3097198928634012e-05,
+      "loss": 9.6353,
+      "step": 46460
+    },
+    {
+      "epoch": 0.23206571949362031,
+      "grad_norm": 0.09375794231891632,
+      "learning_rate": 2.309569701369246e-05,
+      "loss": 9.6385,
+      "step": 46470
+    },
+    {
+      "epoch": 0.23211565831855976,
+      "grad_norm": 0.0972442626953125,
+      "learning_rate": 2.3094195098750905e-05,
+      "loss": 9.6413,
+      "step": 46480
+    },
+    {
+      "epoch": 0.2321655971434992,
+      "grad_norm": 0.09976401925086975,
+      "learning_rate": 2.309269318380936e-05,
+      "loss": 9.6418,
+      "step": 46490
+    },
+    {
+      "epoch": 0.23221553596843866,
+      "grad_norm": 0.09310724586248398,
+      "learning_rate": 2.3091191268867806e-05,
+      "loss": 9.642,
+      "step": 46500
+    },
+    {
+      "epoch": 0.2322654747933781,
+      "grad_norm": 0.09604420512914658,
+      "learning_rate": 2.308968935392626e-05,
+      "loss": 9.6406,
+      "step": 46510
+    },
+    {
+      "epoch": 0.23231541361831756,
+      "grad_norm": 0.09983832389116287,
+      "learning_rate": 2.3088187438984706e-05,
+      "loss": 9.6365,
+      "step": 46520
+    },
+    {
+      "epoch": 0.232365352443257,
+      "grad_norm": 0.09731082618236542,
+      "learning_rate": 2.3086685524043153e-05,
+      "loss": 9.6396,
+      "step": 46530
+    },
+    {
+      "epoch": 0.23241529126819646,
+      "grad_norm": 0.09260038286447525,
+      "learning_rate": 2.3085183609101606e-05,
+      "loss": 9.6343,
+      "step": 46540
+    },
+    {
+      "epoch": 0.2324652300931359,
+      "grad_norm": 0.09418473392724991,
+      "learning_rate": 2.3083681694160053e-05,
+      "loss": 9.6387,
+      "step": 46550
+    },
+    {
+      "epoch": 0.23251516891807536,
+      "grad_norm": 0.09961055219173431,
+      "learning_rate": 2.3082179779218507e-05,
+      "loss": 9.6389,
+      "step": 46560
+    },
+    {
+      "epoch": 0.2325651077430148,
+      "grad_norm": 0.0885944813489914,
+      "learning_rate": 2.3080677864276954e-05,
+      "loss": 9.6307,
+      "step": 46570
+    },
+    {
+      "epoch": 0.23261504656795426,
+      "grad_norm": 0.0929674431681633,
+      "learning_rate": 2.3079175949335404e-05,
+      "loss": 9.6414,
+      "step": 46580
+    },
+    {
+      "epoch": 0.2326649853928937,
+      "grad_norm": 0.09185917675495148,
+      "learning_rate": 2.3077674034393854e-05,
+      "loss": 9.6348,
+      "step": 46590
+    },
+    {
+      "epoch": 0.23271492421783316,
+      "grad_norm": 0.09937243163585663,
+      "learning_rate": 2.30761721194523e-05,
+      "loss": 9.6294,
+      "step": 46600
+    },
+    {
+      "epoch": 0.2327648630427726,
+      "grad_norm": 0.09624093770980835,
+      "learning_rate": 2.3074670204510754e-05,
+      "loss": 9.6273,
+      "step": 46610
+    },
+    {
+      "epoch": 0.23281480186771206,
+      "grad_norm": 0.09259884804487228,
+      "learning_rate": 2.30731682895692e-05,
+      "loss": 9.6259,
+      "step": 46620
+    },
+    {
+      "epoch": 0.2328647406926515,
+      "grad_norm": 0.09767475724220276,
+      "learning_rate": 2.307166637462765e-05,
+      "loss": 9.6371,
+      "step": 46630
+    },
+    {
+      "epoch": 0.23291467951759096,
+      "grad_norm": 0.09269337356090546,
+      "learning_rate": 2.30701644596861e-05,
+      "loss": 9.6273,
+      "step": 46640
+    },
+    {
+      "epoch": 0.2329646183425304,
+      "grad_norm": 0.09703583270311356,
+      "learning_rate": 2.3068662544744548e-05,
+      "loss": 9.6257,
+      "step": 46650
+    },
+    {
+      "epoch": 0.23301455716746985,
+      "grad_norm": 0.09539482742547989,
+      "learning_rate": 2.3067160629803002e-05,
+      "loss": 9.631,
+      "step": 46660
+    },
+    {
+      "epoch": 0.2330644959924093,
+      "grad_norm": 0.09522981196641922,
+      "learning_rate": 2.306565871486145e-05,
+      "loss": 9.6335,
+      "step": 46670
+    },
+    {
+      "epoch": 0.23311443481734875,
+      "grad_norm": 0.09748247265815735,
+      "learning_rate": 2.30641567999199e-05,
+      "loss": 9.6364,
+      "step": 46680
+    },
+    {
+      "epoch": 0.2331643736422882,
+      "grad_norm": 0.09750822186470032,
+      "learning_rate": 2.306265488497835e-05,
+      "loss": 9.6327,
+      "step": 46690
+    },
+    {
+      "epoch": 0.23321431246722765,
+      "grad_norm": 0.09520017355680466,
+      "learning_rate": 2.3061152970036796e-05,
+      "loss": 9.6306,
+      "step": 46700
+    },
+    {
+      "epoch": 0.2332642512921671,
+      "grad_norm": 0.09403938800096512,
+      "learning_rate": 2.305965105509525e-05,
+      "loss": 9.6276,
+      "step": 46710
+    },
+    {
+      "epoch": 0.23331419011710655,
+      "grad_norm": 0.09592479467391968,
+      "learning_rate": 2.3058149140153696e-05,
+      "loss": 9.6331,
+      "step": 46720
+    },
+    {
+      "epoch": 0.233364128942046,
+      "grad_norm": 0.09444668143987656,
+      "learning_rate": 2.3056647225212146e-05,
+      "loss": 9.6345,
+      "step": 46730
+    },
+    {
+      "epoch": 0.23341406776698545,
+      "grad_norm": 0.09617168456315994,
+      "learning_rate": 2.3055145310270596e-05,
+      "loss": 9.6335,
+      "step": 46740
+    },
+    {
+      "epoch": 0.2334640065919249,
+      "grad_norm": 0.09481420367956161,
+      "learning_rate": 2.3053643395329043e-05,
+      "loss": 9.6353,
+      "step": 46750
+    },
+    {
+      "epoch": 0.23351394541686435,
+      "grad_norm": 0.09595254808664322,
+      "learning_rate": 2.3052141480387497e-05,
+      "loss": 9.6378,
+      "step": 46760
+    },
+    {
+      "epoch": 0.2335638842418038,
+      "grad_norm": 0.09526467323303223,
+      "learning_rate": 2.3050639565445944e-05,
+      "loss": 9.6292,
+      "step": 46770
+    },
+    {
+      "epoch": 0.23361382306674325,
+      "grad_norm": 0.09057611972093582,
+      "learning_rate": 2.3049137650504394e-05,
+      "loss": 9.6269,
+      "step": 46780
+    },
+    {
+      "epoch": 0.2336637618916827,
+      "grad_norm": 0.09435748308897018,
+      "learning_rate": 2.3047635735562844e-05,
+      "loss": 9.638,
+      "step": 46790
+    },
+    {
+      "epoch": 0.23371370071662215,
+      "grad_norm": 0.09341932833194733,
+      "learning_rate": 2.304613382062129e-05,
+      "loss": 9.6302,
+      "step": 46800
+    },
+    {
+      "epoch": 0.2337636395415616,
+      "grad_norm": 0.09780159592628479,
+      "learning_rate": 2.3044631905679744e-05,
+      "loss": 9.6343,
+      "step": 46810
+    },
+    {
+      "epoch": 0.23381357836650105,
+      "grad_norm": 0.09608551114797592,
+      "learning_rate": 2.304312999073819e-05,
+      "loss": 9.6324,
+      "step": 46820
+    },
+    {
+      "epoch": 0.2338635171914405,
+      "grad_norm": 0.10504171252250671,
+      "learning_rate": 2.304162807579664e-05,
+      "loss": 9.6361,
+      "step": 46830
+    },
+    {
+      "epoch": 0.23391345601637994,
+      "grad_norm": 0.09910661727190018,
+      "learning_rate": 2.304012616085509e-05,
+      "loss": 9.6292,
+      "step": 46840
+    },
+    {
+      "epoch": 0.2339633948413194,
+      "grad_norm": 0.10046257078647614,
+      "learning_rate": 2.3038624245913538e-05,
+      "loss": 9.6332,
+      "step": 46850
+    },
+    {
+      "epoch": 0.23401333366625884,
+      "grad_norm": 0.09717267006635666,
+      "learning_rate": 2.3037122330971992e-05,
+      "loss": 9.6163,
+      "step": 46860
+    },
+    {
+      "epoch": 0.2340632724911983,
+      "grad_norm": 0.09229366481304169,
+      "learning_rate": 2.303562041603044e-05,
+      "loss": 9.6261,
+      "step": 46870
+    },
+    {
+      "epoch": 0.23411321131613774,
+      "grad_norm": 0.09404976665973663,
+      "learning_rate": 2.303411850108889e-05,
+      "loss": 9.6323,
+      "step": 46880
+    },
+    {
+      "epoch": 0.2341631501410772,
+      "grad_norm": 0.09583253413438797,
+      "learning_rate": 2.303261658614734e-05,
+      "loss": 9.6264,
+      "step": 46890
+    },
+    {
+      "epoch": 0.23421308896601664,
+      "grad_norm": 0.09781666100025177,
+      "learning_rate": 2.303111467120579e-05,
+      "loss": 9.6268,
+      "step": 46900
+    },
+    {
+      "epoch": 0.2342630277909561,
+      "grad_norm": 0.09339626133441925,
+      "learning_rate": 2.302961275626424e-05,
+      "loss": 9.6277,
+      "step": 46910
+    },
+    {
+      "epoch": 0.23431296661589554,
+      "grad_norm": 0.09981784224510193,
+      "learning_rate": 2.3028110841322686e-05,
+      "loss": 9.6338,
+      "step": 46920
+    },
+    {
+      "epoch": 0.234362905440835,
+      "grad_norm": 0.09911654889583588,
+      "learning_rate": 2.3026608926381136e-05,
+      "loss": 9.6334,
+      "step": 46930
+    },
+    {
+      "epoch": 0.23441284426577444,
+      "grad_norm": 0.09392616897821426,
+      "learning_rate": 2.3025107011439587e-05,
+      "loss": 9.6212,
+      "step": 46940
+    },
+    {
+      "epoch": 0.2344627830907139,
+      "grad_norm": 0.10090331733226776,
+      "learning_rate": 2.3023605096498037e-05,
+      "loss": 9.6232,
+      "step": 46950
+    },
+    {
+      "epoch": 0.23451272191565334,
+      "grad_norm": 0.09081251919269562,
+      "learning_rate": 2.3022103181556487e-05,
+      "loss": 9.6258,
+      "step": 46960
+    },
+    {
+      "epoch": 0.2345626607405928,
+      "grad_norm": 0.09683696180582047,
+      "learning_rate": 2.3020601266614934e-05,
+      "loss": 9.6245,
+      "step": 46970
+    },
+    {
+      "epoch": 0.2346125995655322,
+      "grad_norm": 0.09155721217393875,
+      "learning_rate": 2.3019099351673384e-05,
+      "loss": 9.6267,
+      "step": 46980
+    },
+    {
+      "epoch": 0.23466253839047166,
+      "grad_norm": 0.09573863446712494,
+      "learning_rate": 2.3017597436731834e-05,
+      "loss": 9.6289,
+      "step": 46990
+    },
+    {
+      "epoch": 0.2347124772154111,
+      "grad_norm": 0.09463366866111755,
+      "learning_rate": 2.3016095521790284e-05,
+      "loss": 9.6195,
+      "step": 47000
+    },
+    {
+      "epoch": 0.23476241604035056,
+      "grad_norm": 0.10097618401050568,
+      "learning_rate": 2.3014593606848734e-05,
+      "loss": 9.6101,
+      "step": 47010
+    },
+    {
+      "epoch": 0.23481235486529,
+      "grad_norm": 0.09264184534549713,
+      "learning_rate": 2.301309169190718e-05,
+      "loss": 9.621,
+      "step": 47020
+    },
+    {
+      "epoch": 0.23486229369022946,
+      "grad_norm": 0.09526827186346054,
+      "learning_rate": 2.301158977696563e-05,
+      "loss": 9.6284,
+      "step": 47030
+    },
+    {
+      "epoch": 0.2349122325151689,
+      "grad_norm": 0.09512627869844437,
+      "learning_rate": 2.301008786202408e-05,
+      "loss": 9.6286,
+      "step": 47040
+    },
+    {
+      "epoch": 0.23496217134010836,
+      "grad_norm": 0.09627030789852142,
+      "learning_rate": 2.300858594708253e-05,
+      "loss": 9.6311,
+      "step": 47050
+    },
+    {
+      "epoch": 0.2350121101650478,
+      "grad_norm": 0.08953950554132462,
+      "learning_rate": 2.3007084032140982e-05,
+      "loss": 9.6195,
+      "step": 47060
+    },
+    {
+      "epoch": 0.23506204898998725,
+      "grad_norm": 0.09324587136507034,
+      "learning_rate": 2.300558211719943e-05,
+      "loss": 9.6292,
+      "step": 47070
+    },
+    {
+      "epoch": 0.2351119878149267,
+      "grad_norm": 0.09066016227006912,
+      "learning_rate": 2.300408020225788e-05,
+      "loss": 9.6237,
+      "step": 47080
+    },
+    {
+      "epoch": 0.23516192663986615,
+      "grad_norm": 0.08983506262302399,
+      "learning_rate": 2.300257828731633e-05,
+      "loss": 9.6283,
+      "step": 47090
+    },
+    {
+      "epoch": 0.2352118654648056,
+      "grad_norm": 0.09250736236572266,
+      "learning_rate": 2.300107637237478e-05,
+      "loss": 9.6245,
+      "step": 47100
+    },
+    {
+      "epoch": 0.23526180428974505,
+      "grad_norm": 0.09599629789590836,
+      "learning_rate": 2.299957445743323e-05,
+      "loss": 9.6202,
+      "step": 47110
+    },
+    {
+      "epoch": 0.2353117431146845,
+      "grad_norm": 0.09286179393529892,
+      "learning_rate": 2.2998072542491676e-05,
+      "loss": 9.6168,
+      "step": 47120
+    },
+    {
+      "epoch": 0.23536168193962395,
+      "grad_norm": 0.09755680710077286,
+      "learning_rate": 2.2996570627550126e-05,
+      "loss": 9.6249,
+      "step": 47130
+    },
+    {
+      "epoch": 0.2354116207645634,
+      "grad_norm": 0.09816092252731323,
+      "learning_rate": 2.2995068712608577e-05,
+      "loss": 9.6213,
+      "step": 47140
+    },
+    {
+      "epoch": 0.23546155958950285,
+      "grad_norm": 0.08990190178155899,
+      "learning_rate": 2.2993566797667027e-05,
+      "loss": 9.6237,
+      "step": 47150
+    },
+    {
+      "epoch": 0.2355114984144423,
+      "grad_norm": 0.09264282137155533,
+      "learning_rate": 2.2992064882725477e-05,
+      "loss": 9.6222,
+      "step": 47160
+    },
+    {
+      "epoch": 0.23556143723938175,
+      "grad_norm": 0.0953594446182251,
+      "learning_rate": 2.2990562967783924e-05,
+      "loss": 9.6219,
+      "step": 47170
+    },
+    {
+      "epoch": 0.2356113760643212,
+      "grad_norm": 0.09457553923130035,
+      "learning_rate": 2.2989061052842374e-05,
+      "loss": 9.6167,
+      "step": 47180
+    },
+    {
+      "epoch": 0.23566131488926065,
+      "grad_norm": 0.09653682261705399,
+      "learning_rate": 2.2987559137900824e-05,
+      "loss": 9.6214,
+      "step": 47190
+    },
+    {
+      "epoch": 0.2357112537142001,
+      "grad_norm": 0.09378449618816376,
+      "learning_rate": 2.2986057222959274e-05,
+      "loss": 9.6282,
+      "step": 47200
+    },
+    {
+      "epoch": 0.23576119253913955,
+      "grad_norm": 0.0928981825709343,
+      "learning_rate": 2.2984555308017724e-05,
+      "loss": 9.6326,
+      "step": 47210
+    },
+    {
+      "epoch": 0.235811131364079,
+      "grad_norm": 0.0937650203704834,
+      "learning_rate": 2.298305339307617e-05,
+      "loss": 9.6249,
+      "step": 47220
+    },
+    {
+      "epoch": 0.23586107018901845,
+      "grad_norm": 0.09575259685516357,
+      "learning_rate": 2.298155147813462e-05,
+      "loss": 9.6179,
+      "step": 47230
+    },
+    {
+      "epoch": 0.2359110090139579,
+      "grad_norm": 75396.46875,
+      "learning_rate": 2.298004956319307e-05,
+      "loss": 9.9622,
+      "step": 47240
+    },
+    {
+      "epoch": 0.23596094783889734,
+      "grad_norm": 0.09559078514575958,
+      "learning_rate": 2.2978547648251522e-05,
+      "loss": 9.6256,
+      "step": 47250
+    },
+    {
+      "epoch": 0.2360108866638368,
+      "grad_norm": 0.09144776314496994,
+      "learning_rate": 2.2977045733309972e-05,
+      "loss": 17.3369,
+      "step": 47260
+    },
+    {
+      "epoch": 0.23606082548877624,
+      "grad_norm": 0.09429620951414108,
+      "learning_rate": 2.2975543818368422e-05,
+      "loss": 12.2017,
+      "step": 47270
+    },
+    {
+      "epoch": 0.2361107643137157,
+      "grad_norm": 0.09179428964853287,
+      "learning_rate": 2.297404190342687e-05,
+      "loss": 9.6117,
+      "step": 47280
+    },
+    {
+      "epoch": 0.23616070313865514,
+      "grad_norm": 0.09685272723436356,
+      "learning_rate": 2.297253998848532e-05,
+      "loss": 9.6269,
+      "step": 47290
+    },
+    {
+      "epoch": 0.2362106419635946,
+      "grad_norm": 0.09493184089660645,
+      "learning_rate": 2.297103807354377e-05,
+      "loss": 9.6188,
+      "step": 47300
+    },
+    {
+      "epoch": 0.23626058078853404,
+      "grad_norm": 0.10096865147352219,
+      "learning_rate": 2.296953615860222e-05,
+      "loss": 9.6169,
+      "step": 47310
+    },
+    {
+      "epoch": 0.2363105196134735,
+      "grad_norm": 0.10004733502864838,
+      "learning_rate": 2.296803424366067e-05,
+      "loss": 9.6244,
+      "step": 47320
+    },
+    {
+      "epoch": 0.23636045843841294,
+      "grad_norm": 0.09608258306980133,
+      "learning_rate": 2.2966532328719116e-05,
+      "loss": 9.6154,
+      "step": 47330
+    },
+    {
+      "epoch": 0.2364103972633524,
+      "grad_norm": 0.09426455199718475,
+      "learning_rate": 2.2965030413777567e-05,
+      "loss": 9.6204,
+      "step": 47340
+    },
+    {
+      "epoch": 0.23646033608829184,
+      "grad_norm": 0.09753237664699554,
+      "learning_rate": 2.2963528498836017e-05,
+      "loss": 9.6185,
+      "step": 47350
+    },
+    {
+      "epoch": 0.2365102749132313,
+      "grad_norm": 0.09157487750053406,
+      "learning_rate": 2.2962026583894467e-05,
+      "loss": 9.6327,
+      "step": 47360
+    },
+    {
+      "epoch": 0.23656021373817074,
+      "grad_norm": 0.10023976117372513,
+      "learning_rate": 2.2960524668952917e-05,
+      "loss": 9.6226,
+      "step": 47370
+    },
+    {
+      "epoch": 0.2366101525631102,
+      "grad_norm": 0.09586095809936523,
+      "learning_rate": 2.2959022754011364e-05,
+      "loss": 9.6217,
+      "step": 47380
+    },
+    {
+      "epoch": 0.23666009138804964,
+      "grad_norm": 0.0944489985704422,
+      "learning_rate": 2.2957520839069814e-05,
+      "loss": 9.6157,
+      "step": 47390
+    },
+    {
+      "epoch": 0.2367100302129891,
+      "grad_norm": 0.09229391068220139,
+      "learning_rate": 2.2956018924128264e-05,
+      "loss": 9.6213,
+      "step": 47400
+    },
+    {
+      "epoch": 0.23675996903792854,
+      "grad_norm": 0.09394244849681854,
+      "learning_rate": 2.2954517009186714e-05,
+      "loss": 9.6235,
+      "step": 47410
+    },
+    {
+      "epoch": 0.23680990786286799,
+      "grad_norm": 0.0892910584807396,
+      "learning_rate": 2.2953015094245165e-05,
+      "loss": 9.6202,
+      "step": 47420
+    },
+    {
+      "epoch": 0.23685984668780743,
+      "grad_norm": 0.09789809584617615,
+      "learning_rate": 2.295151317930361e-05,
+      "loss": 9.6224,
+      "step": 47430
+    },
+    {
+      "epoch": 0.23690978551274688,
+      "grad_norm": 0.0950971245765686,
+      "learning_rate": 2.295001126436206e-05,
+      "loss": 9.6107,
+      "step": 47440
+    },
+    {
+      "epoch": 0.23695972433768633,
+      "grad_norm": 0.09064780175685883,
+      "learning_rate": 2.2948509349420512e-05,
+      "loss": 9.6138,
+      "step": 47450
+    },
+    {
+      "epoch": 0.23700966316262578,
+      "grad_norm": 0.09859991818666458,
+      "learning_rate": 2.2947007434478962e-05,
+      "loss": 9.6197,
+      "step": 47460
+    },
+    {
+      "epoch": 0.23705960198756523,
+      "grad_norm": 0.0948246568441391,
+      "learning_rate": 2.2945505519537412e-05,
+      "loss": 9.6189,
+      "step": 47470
+    },
+    {
+      "epoch": 0.23710954081250468,
+      "grad_norm": 0.09848882257938385,
+      "learning_rate": 2.294400360459586e-05,
+      "loss": 9.611,
+      "step": 47480
+    },
+    {
+      "epoch": 0.23715947963744413,
+      "grad_norm": 0.10083547234535217,
+      "learning_rate": 2.294250168965431e-05,
+      "loss": 9.6214,
+      "step": 47490
+    },
+    {
+      "epoch": 0.23720941846238358,
+      "grad_norm": 0.09797154366970062,
+      "learning_rate": 2.294099977471276e-05,
+      "loss": 9.6184,
+      "step": 47500
+    },
+    {
+      "epoch": 0.23725935728732303,
+      "grad_norm": 0.0983353927731514,
+      "learning_rate": 2.293949785977121e-05,
+      "loss": 9.6119,
+      "step": 47510
+    },
+    {
+      "epoch": 0.23730929611226248,
+      "grad_norm": 0.10210006684064865,
+      "learning_rate": 2.293799594482966e-05,
+      "loss": 9.6136,
+      "step": 47520
+    },
+    {
+      "epoch": 0.23735923493720193,
+      "grad_norm": 0.10113011300563812,
+      "learning_rate": 2.2936494029888106e-05,
+      "loss": 9.6137,
+      "step": 47530
+    },
+    {
+      "epoch": 0.23740917376214138,
+      "grad_norm": 0.0893891304731369,
+      "learning_rate": 2.2934992114946557e-05,
+      "loss": 9.6155,
+      "step": 47540
+    },
+    {
+      "epoch": 0.23745911258708083,
+      "grad_norm": 0.09196850657463074,
+      "learning_rate": 2.2933490200005007e-05,
+      "loss": 9.6032,
+      "step": 47550
+    },
+    {
+      "epoch": 0.23750905141202028,
+      "grad_norm": 0.08714939653873444,
+      "learning_rate": 2.2931988285063457e-05,
+      "loss": 9.6155,
+      "step": 47560
+    },
+    {
+      "epoch": 0.23755899023695973,
+      "grad_norm": 0.0882970318198204,
+      "learning_rate": 2.2930486370121907e-05,
+      "loss": 9.6081,
+      "step": 47570
+    },
+    {
+      "epoch": 0.23760892906189918,
+      "grad_norm": 0.09797061234712601,
+      "learning_rate": 2.2928984455180354e-05,
+      "loss": 9.6092,
+      "step": 47580
+    },
+    {
+      "epoch": 0.23765886788683863,
+      "grad_norm": 0.09515173733234406,
+      "learning_rate": 2.2927482540238807e-05,
+      "loss": 9.6153,
+      "step": 47590
+    },
+    {
+      "epoch": 0.23770880671177808,
+      "grad_norm": 0.09541980177164078,
+      "learning_rate": 2.2925980625297254e-05,
+      "loss": 9.6167,
+      "step": 47600
+    },
+    {
+      "epoch": 0.23775874553671753,
+      "grad_norm": 0.09236780554056168,
+      "learning_rate": 2.2924478710355704e-05,
+      "loss": 9.6251,
+      "step": 47610
+    },
+    {
+      "epoch": 0.23780868436165697,
+      "grad_norm": 0.09183859080076218,
+      "learning_rate": 2.2922976795414155e-05,
+      "loss": 9.6104,
+      "step": 47620
+    },
+    {
+      "epoch": 0.23785862318659642,
+      "grad_norm": 0.1002270057797432,
+      "learning_rate": 2.29214748804726e-05,
+      "loss": 9.6142,
+      "step": 47630
+    },
+    {
+      "epoch": 0.23790856201153587,
+      "grad_norm": 0.09763851016759872,
+      "learning_rate": 2.2919972965531055e-05,
+      "loss": 9.6067,
+      "step": 47640
+    },
+    {
+      "epoch": 0.23795850083647532,
+      "grad_norm": 0.09223950654268265,
+      "learning_rate": 2.2918471050589502e-05,
+      "loss": 9.6198,
+      "step": 47650
+    },
+    {
+      "epoch": 0.23800843966141477,
+      "grad_norm": 0.09290538728237152,
+      "learning_rate": 2.2916969135647952e-05,
+      "loss": 9.6121,
+      "step": 47660
+    },
+    {
+      "epoch": 0.23805837848635422,
+      "grad_norm": 0.09840866923332214,
+      "learning_rate": 2.2915467220706402e-05,
+      "loss": 9.6148,
+      "step": 47670
+    },
+    {
+      "epoch": 0.23810831731129367,
+      "grad_norm": 0.0904998853802681,
+      "learning_rate": 2.291396530576485e-05,
+      "loss": 9.6134,
+      "step": 47680
+    },
+    {
+      "epoch": 0.23815825613623312,
+      "grad_norm": 0.09012880176305771,
+      "learning_rate": 2.2912463390823302e-05,
+      "loss": 9.6154,
+      "step": 47690
+    },
+    {
+      "epoch": 0.23820819496117257,
+      "grad_norm": 0.08958422392606735,
+      "learning_rate": 2.291096147588175e-05,
+      "loss": 9.6192,
+      "step": 47700
+    },
+    {
+      "epoch": 0.23825813378611202,
+      "grad_norm": 0.09270545840263367,
+      "learning_rate": 2.29094595609402e-05,
+      "loss": 9.6096,
+      "step": 47710
+    },
+    {
+      "epoch": 0.23830807261105147,
+      "grad_norm": 0.09306894987821579,
+      "learning_rate": 2.290795764599865e-05,
+      "loss": 9.604,
+      "step": 47720
+    },
+    {
+      "epoch": 0.23835801143599092,
+      "grad_norm": 0.1006632074713707,
+      "learning_rate": 2.2906455731057096e-05,
+      "loss": 9.6153,
+      "step": 47730
+    },
+    {
+      "epoch": 0.23840795026093037,
+      "grad_norm": 0.08890628069639206,
+      "learning_rate": 2.290495381611555e-05,
+      "loss": 9.6092,
+      "step": 47740
+    },
+    {
+      "epoch": 0.23845788908586982,
+      "grad_norm": 0.09797194600105286,
+      "learning_rate": 2.2903451901173997e-05,
+      "loss": 9.6124,
+      "step": 47750
+    },
+    {
+      "epoch": 0.23850782791080927,
+      "grad_norm": 0.0914837047457695,
+      "learning_rate": 2.2901949986232447e-05,
+      "loss": 9.6155,
+      "step": 47760
+    },
+    {
+      "epoch": 0.23855776673574872,
+      "grad_norm": 0.09398198127746582,
+      "learning_rate": 2.2900448071290897e-05,
+      "loss": 9.6183,
+      "step": 47770
+    },
+    {
+      "epoch": 0.23860770556068817,
+      "grad_norm": 0.10565483570098877,
+      "learning_rate": 2.2898946156349344e-05,
+      "loss": 9.6146,
+      "step": 47780
+    },
+    {
+      "epoch": 0.23865764438562762,
+      "grad_norm": 0.09536195546388626,
+      "learning_rate": 2.2897444241407797e-05,
+      "loss": 9.6134,
+      "step": 47790
+    },
+    {
+      "epoch": 0.23870758321056706,
+      "grad_norm": 0.0907558798789978,
+      "learning_rate": 2.2895942326466244e-05,
+      "loss": 9.6117,
+      "step": 47800
+    },
+    {
+      "epoch": 0.23875752203550651,
+      "grad_norm": 0.09679032117128372,
+      "learning_rate": 2.2894440411524694e-05,
+      "loss": 9.6129,
+      "step": 47810
+    },
+    {
+      "epoch": 0.23880746086044596,
+      "grad_norm": 0.08943892270326614,
+      "learning_rate": 2.2892938496583145e-05,
+      "loss": 9.618,
+      "step": 47820
+    },
+    {
+      "epoch": 0.2388573996853854,
+      "grad_norm": 0.0918293297290802,
+      "learning_rate": 2.289143658164159e-05,
+      "loss": 9.6098,
+      "step": 47830
+    },
+    {
+      "epoch": 0.23890733851032486,
+      "grad_norm": 0.09785201400518417,
+      "learning_rate": 2.2889934666700045e-05,
+      "loss": 9.6052,
+      "step": 47840
+    },
+    {
+      "epoch": 0.2389572773352643,
+      "grad_norm": 0.09538784623146057,
+      "learning_rate": 2.2888432751758492e-05,
+      "loss": 9.5993,
+      "step": 47850
+    },
+    {
+      "epoch": 0.23900721616020376,
+      "grad_norm": 0.09343953430652618,
+      "learning_rate": 2.2886930836816942e-05,
+      "loss": 9.6063,
+      "step": 47860
+    },
+    {
+      "epoch": 0.2390571549851432,
+      "grad_norm": 0.0926864743232727,
+      "learning_rate": 2.2885428921875392e-05,
+      "loss": 9.6103,
+      "step": 47870
+    },
+    {
+      "epoch": 0.23910709381008266,
+      "grad_norm": 0.09175824373960495,
+      "learning_rate": 2.288392700693384e-05,
+      "loss": 9.6072,
+      "step": 47880
+    },
+    {
+      "epoch": 0.2391570326350221,
+      "grad_norm": 0.09742142260074615,
+      "learning_rate": 2.2882425091992292e-05,
+      "loss": 9.6058,
+      "step": 47890
+    },
+    {
+      "epoch": 0.23920697145996156,
+      "grad_norm": 0.09516909718513489,
+      "learning_rate": 2.288092317705074e-05,
+      "loss": 9.6085,
+      "step": 47900
+    },
+    {
+      "epoch": 0.239256910284901,
+      "grad_norm": 0.09717977792024612,
+      "learning_rate": 2.2879421262109193e-05,
+      "loss": 9.6036,
+      "step": 47910
+    },
+    {
+      "epoch": 0.23930684910984046,
+      "grad_norm": 0.09463629871606827,
+      "learning_rate": 2.287791934716764e-05,
+      "loss": 9.6029,
+      "step": 47920
+    },
+    {
+      "epoch": 0.2393567879347799,
+      "grad_norm": 0.09392336010932922,
+      "learning_rate": 2.2876417432226086e-05,
+      "loss": 9.6177,
+      "step": 47930
+    },
+    {
+      "epoch": 0.23940672675971936,
+      "grad_norm": 0.09007874131202698,
+      "learning_rate": 2.287491551728454e-05,
+      "loss": 9.6079,
+      "step": 47940
+    },
+    {
+      "epoch": 0.2394566655846588,
+      "grad_norm": 0.09397672861814499,
+      "learning_rate": 2.2873413602342987e-05,
+      "loss": 9.6104,
+      "step": 47950
+    },
+    {
+      "epoch": 0.23950660440959823,
+      "grad_norm": 0.08829968422651291,
+      "learning_rate": 2.287191168740144e-05,
+      "loss": 9.6029,
+      "step": 47960
+    },
+    {
+      "epoch": 0.23955654323453768,
+      "grad_norm": 0.09608025848865509,
+      "learning_rate": 2.2870409772459887e-05,
+      "loss": 9.6063,
+      "step": 47970
+    },
+    {
+      "epoch": 0.23960648205947713,
+      "grad_norm": 0.09699834138154984,
+      "learning_rate": 2.2868907857518334e-05,
+      "loss": 9.6119,
+      "step": 47980
+    },
+    {
+      "epoch": 0.23965642088441658,
+      "grad_norm": 0.08946532756090164,
+      "learning_rate": 2.2867405942576787e-05,
+      "loss": 9.6014,
+      "step": 47990
+    },
+    {
+      "epoch": 0.23970635970935603,
+      "grad_norm": 0.0944380983710289,
+      "learning_rate": 2.2865904027635234e-05,
+      "loss": 9.6076,
+      "step": 48000
+    },
+    {
+      "epoch": 0.23975629853429548,
+      "grad_norm": 0.09391778707504272,
+      "learning_rate": 2.2864402112693688e-05,
+      "loss": 9.6034,
+      "step": 48010
+    },
+    {
+      "epoch": 0.23980623735923493,
+      "grad_norm": 0.09647521376609802,
+      "learning_rate": 2.2862900197752135e-05,
+      "loss": 9.6077,
+      "step": 48020
+    },
+    {
+      "epoch": 0.23985617618417437,
+      "grad_norm": 0.09497778117656708,
+      "learning_rate": 2.286139828281058e-05,
+      "loss": 9.6047,
+      "step": 48030
+    },
+    {
+      "epoch": 0.23990611500911382,
+      "grad_norm": 0.10870006680488586,
+      "learning_rate": 2.2859896367869035e-05,
+      "loss": 9.6094,
+      "step": 48040
+    },
+    {
+      "epoch": 0.23995605383405327,
+      "grad_norm": 0.09138607233762741,
+      "learning_rate": 2.2858394452927482e-05,
+      "loss": 9.6093,
+      "step": 48050
+    },
+    {
+      "epoch": 0.24000599265899272,
+      "grad_norm": 0.09472063183784485,
+      "learning_rate": 2.2856892537985935e-05,
+      "loss": 9.6105,
+      "step": 48060
+    },
+    {
+      "epoch": 0.24005593148393217,
+      "grad_norm": 0.09620670229196548,
+      "learning_rate": 2.2855390623044382e-05,
+      "loss": 9.5972,
+      "step": 48070
+    },
+    {
+      "epoch": 0.24010587030887162,
+      "grad_norm": 0.09534239768981934,
+      "learning_rate": 2.285388870810283e-05,
+      "loss": 9.6064,
+      "step": 48080
+    },
+    {
+      "epoch": 0.24015580913381107,
+      "grad_norm": 0.09170860797166824,
+      "learning_rate": 2.2852386793161282e-05,
+      "loss": 9.5999,
+      "step": 48090
+    },
+    {
+      "epoch": 0.24020574795875052,
+      "grad_norm": 0.09515348076820374,
+      "learning_rate": 2.285088487821973e-05,
+      "loss": 9.6172,
+      "step": 48100
+    },
+    {
+      "epoch": 0.24025568678368997,
+      "grad_norm": 0.09469642490148544,
+      "learning_rate": 2.2849382963278183e-05,
+      "loss": 9.6089,
+      "step": 48110
+    },
+    {
+      "epoch": 0.24030562560862942,
+      "grad_norm": 0.09230419993400574,
+      "learning_rate": 2.284788104833663e-05,
+      "loss": 9.6098,
+      "step": 48120
+    },
+    {
+      "epoch": 0.24035556443356887,
+      "grad_norm": 0.09317916631698608,
+      "learning_rate": 2.2846379133395076e-05,
+      "loss": 9.5989,
+      "step": 48130
+    },
+    {
+      "epoch": 0.24040550325850832,
+      "grad_norm": 0.0909963995218277,
+      "learning_rate": 2.284487721845353e-05,
+      "loss": 9.6125,
+      "step": 48140
+    },
+    {
+      "epoch": 0.24045544208344777,
+      "grad_norm": 0.09574473649263382,
+      "learning_rate": 2.2843375303511977e-05,
+      "loss": 9.601,
+      "step": 48150
+    },
+    {
+      "epoch": 0.24050538090838722,
+      "grad_norm": 0.09136365354061127,
+      "learning_rate": 2.284187338857043e-05,
+      "loss": 9.6029,
+      "step": 48160
+    },
+    {
+      "epoch": 0.24055531973332667,
+      "grad_norm": 0.0943564847111702,
+      "learning_rate": 2.2840371473628877e-05,
+      "loss": 9.6088,
+      "step": 48170
+    },
+    {
+      "epoch": 0.24060525855826612,
+      "grad_norm": 0.09551718086004257,
+      "learning_rate": 2.2838869558687324e-05,
+      "loss": 9.6124,
+      "step": 48180
+    },
+    {
+      "epoch": 0.24065519738320557,
+      "grad_norm": 0.09354714304208755,
+      "learning_rate": 2.2837367643745777e-05,
+      "loss": 9.6024,
+      "step": 48190
+    },
+    {
+      "epoch": 0.24070513620814502,
+      "grad_norm": 0.09227943420410156,
+      "learning_rate": 2.2835865728804224e-05,
+      "loss": 9.6047,
+      "step": 48200
+    },
+    {
+      "epoch": 0.24075507503308446,
+      "grad_norm": 0.09799333661794662,
+      "learning_rate": 2.2834363813862678e-05,
+      "loss": 9.6027,
+      "step": 48210
+    },
+    {
+      "epoch": 0.24080501385802391,
+      "grad_norm": 0.09457400441169739,
+      "learning_rate": 2.2832861898921125e-05,
+      "loss": 9.6017,
+      "step": 48220
+    },
+    {
+      "epoch": 0.24085495268296336,
+      "grad_norm": 0.09242776036262512,
+      "learning_rate": 2.2831359983979575e-05,
+      "loss": 9.6013,
+      "step": 48230
+    },
+    {
+      "epoch": 0.2409048915079028,
+      "grad_norm": 0.09647472202777863,
+      "learning_rate": 2.2829858069038025e-05,
+      "loss": 9.5992,
+      "step": 48240
+    },
+    {
+      "epoch": 0.24095483033284226,
+      "grad_norm": 0.09526066482067108,
+      "learning_rate": 2.2828356154096472e-05,
+      "loss": 9.6049,
+      "step": 48250
+    },
+    {
+      "epoch": 0.2410047691577817,
+      "grad_norm": 0.09362996369600296,
+      "learning_rate": 2.2826854239154925e-05,
+      "loss": 9.6031,
+      "step": 48260
+    },
+    {
+      "epoch": 0.24105470798272116,
+      "grad_norm": 0.0959758386015892,
+      "learning_rate": 2.2825352324213372e-05,
+      "loss": 9.6,
+      "step": 48270
+    },
+    {
+      "epoch": 0.2411046468076606,
+      "grad_norm": 0.09466012567281723,
+      "learning_rate": 2.2823850409271822e-05,
+      "loss": 9.6062,
+      "step": 48280
+    },
+    {
+      "epoch": 0.24115458563260006,
+      "grad_norm": 0.08923938125371933,
+      "learning_rate": 2.2822348494330272e-05,
+      "loss": 9.5986,
+      "step": 48290
+    },
+    {
+      "epoch": 0.2412045244575395,
+      "grad_norm": 0.09768188744783401,
+      "learning_rate": 2.282084657938872e-05,
+      "loss": 9.5975,
+      "step": 48300
+    },
+    {
+      "epoch": 0.24125446328247896,
+      "grad_norm": 0.0997987762093544,
+      "learning_rate": 2.2819344664447173e-05,
+      "loss": 9.5978,
+      "step": 48310
+    },
+    {
+      "epoch": 0.2413044021074184,
+      "grad_norm": 0.0951227992773056,
+      "learning_rate": 2.281784274950562e-05,
+      "loss": 9.6114,
+      "step": 48320
+    },
+    {
+      "epoch": 0.24135434093235786,
+      "grad_norm": 0.08892112970352173,
+      "learning_rate": 2.281634083456407e-05,
+      "loss": 9.6082,
+      "step": 48330
+    },
+    {
+      "epoch": 0.2414042797572973,
+      "grad_norm": 0.0971013680100441,
+      "learning_rate": 2.281483891962252e-05,
+      "loss": 9.6065,
+      "step": 48340
+    },
+    {
+      "epoch": 0.24145421858223676,
+      "grad_norm": 0.09072565287351608,
+      "learning_rate": 2.2813337004680967e-05,
+      "loss": 9.5978,
+      "step": 48350
+    },
+    {
+      "epoch": 0.2415041574071762,
+      "grad_norm": 0.08944379538297653,
+      "learning_rate": 2.281183508973942e-05,
+      "loss": 9.5961,
+      "step": 48360
+    },
+    {
+      "epoch": 0.24155409623211566,
+      "grad_norm": 0.09630691260099411,
+      "learning_rate": 2.2810333174797867e-05,
+      "loss": 9.6005,
+      "step": 48370
+    },
+    {
+      "epoch": 0.2416040350570551,
+      "grad_norm": 0.09116937965154648,
+      "learning_rate": 2.2808831259856317e-05,
+      "loss": 9.6018,
+      "step": 48380
+    },
+    {
+      "epoch": 0.24165397388199455,
+      "grad_norm": 0.09356780350208282,
+      "learning_rate": 2.2807329344914768e-05,
+      "loss": 9.6053,
+      "step": 48390
+    },
+    {
+      "epoch": 0.241703912706934,
+      "grad_norm": 0.09499246627092361,
+      "learning_rate": 2.2805827429973214e-05,
+      "loss": 9.6004,
+      "step": 48400
+    },
+    {
+      "epoch": 0.24175385153187345,
+      "grad_norm": 0.09654933959245682,
+      "learning_rate": 2.2804325515031668e-05,
+      "loss": 9.5968,
+      "step": 48410
+    },
+    {
+      "epoch": 0.2418037903568129,
+      "grad_norm": 0.09912944585084915,
+      "learning_rate": 2.2802823600090115e-05,
+      "loss": 9.5996,
+      "step": 48420
+    },
+    {
+      "epoch": 0.24185372918175235,
+      "grad_norm": 0.09671138226985931,
+      "learning_rate": 2.2801321685148565e-05,
+      "loss": 9.6,
+      "step": 48430
+    },
+    {
+      "epoch": 0.2419036680066918,
+      "grad_norm": 0.09061242640018463,
+      "learning_rate": 2.2799819770207015e-05,
+      "loss": 9.6022,
+      "step": 48440
+    },
+    {
+      "epoch": 0.24195360683163125,
+      "grad_norm": 0.09459752589464188,
+      "learning_rate": 2.2798317855265462e-05,
+      "loss": 9.5978,
+      "step": 48450
+    },
+    {
+      "epoch": 0.2420035456565707,
+      "grad_norm": 0.08927514404058456,
+      "learning_rate": 2.2796815940323915e-05,
+      "loss": 9.6072,
+      "step": 48460
+    },
+    {
+      "epoch": 0.24205348448151015,
+      "grad_norm": 0.09183401614427567,
+      "learning_rate": 2.2795314025382362e-05,
+      "loss": 9.5982,
+      "step": 48470
+    },
+    {
+      "epoch": 0.2421034233064496,
+      "grad_norm": 0.09792063385248184,
+      "learning_rate": 2.2793812110440812e-05,
+      "loss": 9.5965,
+      "step": 48480
+    },
+    {
+      "epoch": 0.24215336213138905,
+      "grad_norm": 0.09104972332715988,
+      "learning_rate": 2.2792310195499263e-05,
+      "loss": 9.5929,
+      "step": 48490
+    },
+    {
+      "epoch": 0.2422033009563285,
+      "grad_norm": 0.09637004882097244,
+      "learning_rate": 2.279080828055771e-05,
+      "loss": 9.6063,
+      "step": 48500
+    },
+    {
+      "epoch": 0.24225323978126795,
+      "grad_norm": 0.09254331886768341,
+      "learning_rate": 2.2789306365616163e-05,
+      "loss": 9.6029,
+      "step": 48510
+    },
+    {
+      "epoch": 0.2423031786062074,
+      "grad_norm": 0.09216886013746262,
+      "learning_rate": 2.278780445067461e-05,
+      "loss": 9.5954,
+      "step": 48520
+    },
+    {
+      "epoch": 0.24235311743114685,
+      "grad_norm": 0.09522795677185059,
+      "learning_rate": 2.278630253573306e-05,
+      "loss": 9.6019,
+      "step": 48530
+    },
+    {
+      "epoch": 0.2424030562560863,
+      "grad_norm": 0.09067758917808533,
+      "learning_rate": 2.278480062079151e-05,
+      "loss": 9.5966,
+      "step": 48540
+    },
+    {
+      "epoch": 0.24245299508102575,
+      "grad_norm": 0.10162653774023056,
+      "learning_rate": 2.278329870584996e-05,
+      "loss": 9.5896,
+      "step": 48550
+    },
+    {
+      "epoch": 0.2425029339059652,
+      "grad_norm": 0.09732117503881454,
+      "learning_rate": 2.278179679090841e-05,
+      "loss": 9.5988,
+      "step": 48560
+    },
+    {
+      "epoch": 0.24255287273090465,
+      "grad_norm": 0.09825015068054199,
+      "learning_rate": 2.2780294875966857e-05,
+      "loss": 9.586,
+      "step": 48570
+    },
+    {
+      "epoch": 0.2426028115558441,
+      "grad_norm": 0.09816539287567139,
+      "learning_rate": 2.2778792961025307e-05,
+      "loss": 9.6021,
+      "step": 48580
+    },
+    {
+      "epoch": 0.24265275038078354,
+      "grad_norm": 0.09649891406297684,
+      "learning_rate": 2.2777291046083758e-05,
+      "loss": 9.5972,
+      "step": 48590
+    },
+    {
+      "epoch": 0.242702689205723,
+      "grad_norm": 0.09952486306428909,
+      "learning_rate": 2.2775789131142208e-05,
+      "loss": 9.6009,
+      "step": 48600
+    },
+    {
+      "epoch": 0.24275262803066244,
+      "grad_norm": 0.0953533872961998,
+      "learning_rate": 2.2774287216200658e-05,
+      "loss": 9.602,
+      "step": 48610
+    },
+    {
+      "epoch": 0.2428025668556019,
+      "grad_norm": 0.09693508595228195,
+      "learning_rate": 2.2772785301259105e-05,
+      "loss": 9.5935,
+      "step": 48620
+    },
+    {
+      "epoch": 0.24285250568054134,
+      "grad_norm": 0.08956372737884521,
+      "learning_rate": 2.2771283386317555e-05,
+      "loss": 9.5966,
+      "step": 48630
+    },
+    {
+      "epoch": 0.2429024445054808,
+      "grad_norm": 0.10089200735092163,
+      "learning_rate": 2.2769781471376005e-05,
+      "loss": 9.5974,
+      "step": 48640
+    },
+    {
+      "epoch": 0.24295238333042024,
+      "grad_norm": 0.09470240026712418,
+      "learning_rate": 2.2768279556434455e-05,
+      "loss": 9.5923,
+      "step": 48650
+    },
+    {
+      "epoch": 0.2430023221553597,
+      "grad_norm": 0.08645424246788025,
+      "learning_rate": 2.2766777641492905e-05,
+      "loss": 9.597,
+      "step": 48660
+    },
+    {
+      "epoch": 0.24305226098029914,
+      "grad_norm": 0.09936508536338806,
+      "learning_rate": 2.2765275726551352e-05,
+      "loss": 9.5877,
+      "step": 48670
+    },
+    {
+      "epoch": 0.2431021998052386,
+      "grad_norm": 0.09653277695178986,
+      "learning_rate": 2.2763773811609802e-05,
+      "loss": 9.5927,
+      "step": 48680
+    },
+    {
+      "epoch": 0.24315213863017804,
+      "grad_norm": 0.09702669084072113,
+      "learning_rate": 2.2762271896668253e-05,
+      "loss": 9.5947,
+      "step": 48690
+    },
+    {
+      "epoch": 0.2432020774551175,
+      "grad_norm": 0.09212882816791534,
+      "learning_rate": 2.2760769981726703e-05,
+      "loss": 9.5822,
+      "step": 48700
+    },
+    {
+      "epoch": 0.24325201628005694,
+      "grad_norm": 0.09269695729017258,
+      "learning_rate": 2.2759268066785153e-05,
+      "loss": 9.5927,
+      "step": 48710
+    },
+    {
+      "epoch": 0.2433019551049964,
+      "grad_norm": 0.09830940514802933,
+      "learning_rate": 2.27577661518436e-05,
+      "loss": 9.5933,
+      "step": 48720
+    },
+    {
+      "epoch": 0.24335189392993584,
+      "grad_norm": 0.08732043951749802,
+      "learning_rate": 2.275626423690205e-05,
+      "loss": 9.5939,
+      "step": 48730
+    },
+    {
+      "epoch": 0.24340183275487529,
+      "grad_norm": 0.09308458864688873,
+      "learning_rate": 2.27547623219605e-05,
+      "loss": 9.5926,
+      "step": 48740
+    },
+    {
+      "epoch": 0.24345177157981474,
+      "grad_norm": 0.09478411078453064,
+      "learning_rate": 2.275326040701895e-05,
+      "loss": 9.5954,
+      "step": 48750
+    },
+    {
+      "epoch": 0.24350171040475418,
+      "grad_norm": 0.09362474828958511,
+      "learning_rate": 2.27517584920774e-05,
+      "loss": 9.5937,
+      "step": 48760
+    },
+    {
+      "epoch": 0.24355164922969363,
+      "grad_norm": 0.09029746055603027,
+      "learning_rate": 2.2750256577135847e-05,
+      "loss": 9.5973,
+      "step": 48770
+    },
+    {
+      "epoch": 0.24360158805463308,
+      "grad_norm": 0.10020599514245987,
+      "learning_rate": 2.2748754662194297e-05,
+      "loss": 9.5866,
+      "step": 48780
+    },
+    {
+      "epoch": 0.24365152687957253,
+      "grad_norm": 0.1007223054766655,
+      "learning_rate": 2.2747252747252748e-05,
+      "loss": 9.5955,
+      "step": 48790
+    },
+    {
+      "epoch": 0.24370146570451198,
+      "grad_norm": 0.0981803610920906,
+      "learning_rate": 2.2745750832311198e-05,
+      "loss": 9.5805,
+      "step": 48800
+    },
+    {
+      "epoch": 0.24375140452945143,
+      "grad_norm": 0.09510128200054169,
+      "learning_rate": 2.2744248917369648e-05,
+      "loss": 9.5954,
+      "step": 48810
+    },
+    {
+      "epoch": 0.24380134335439088,
+      "grad_norm": 0.09191947430372238,
+      "learning_rate": 2.2742747002428095e-05,
+      "loss": 9.6015,
+      "step": 48820
+    },
+    {
+      "epoch": 0.24385128217933033,
+      "grad_norm": 0.09448053687810898,
+      "learning_rate": 2.2741245087486545e-05,
+      "loss": 9.5898,
+      "step": 48830
+    },
+    {
+      "epoch": 0.24390122100426978,
+      "grad_norm": 0.09542269259691238,
+      "learning_rate": 2.2739743172544995e-05,
+      "loss": 9.591,
+      "step": 48840
+    },
+    {
+      "epoch": 0.24395115982920923,
+      "grad_norm": 0.09461547434329987,
+      "learning_rate": 2.2738241257603445e-05,
+      "loss": 9.5968,
+      "step": 48850
+    },
+    {
+      "epoch": 0.24400109865414868,
+      "grad_norm": 0.09337666630744934,
+      "learning_rate": 2.2736739342661895e-05,
+      "loss": 9.5898,
+      "step": 48860
+    },
+    {
+      "epoch": 0.24405103747908813,
+      "grad_norm": 0.09441894292831421,
+      "learning_rate": 2.2735237427720346e-05,
+      "loss": 9.5903,
+      "step": 48870
+    },
+    {
+      "epoch": 0.24410097630402758,
+      "grad_norm": 0.0947941318154335,
+      "learning_rate": 2.2733735512778792e-05,
+      "loss": 9.5884,
+      "step": 48880
+    },
+    {
+      "epoch": 0.24415091512896703,
+      "grad_norm": 0.10225453972816467,
+      "learning_rate": 2.2732233597837243e-05,
+      "loss": 9.5906,
+      "step": 48890
+    },
+    {
+      "epoch": 0.24420085395390648,
+      "grad_norm": 0.09397625923156738,
+      "learning_rate": 2.2730731682895693e-05,
+      "loss": 9.5955,
+      "step": 48900
+    },
+    {
+      "epoch": 0.24425079277884593,
+      "grad_norm": 0.0943833664059639,
+      "learning_rate": 2.2729229767954143e-05,
+      "loss": 9.5859,
+      "step": 48910
+    },
+    {
+      "epoch": 0.24430073160378538,
+      "grad_norm": 0.09139799326658249,
+      "learning_rate": 2.2727727853012593e-05,
+      "loss": 9.5942,
+      "step": 48920
+    },
+    {
+      "epoch": 0.24435067042872483,
+      "grad_norm": 0.09776920080184937,
+      "learning_rate": 2.272622593807104e-05,
+      "loss": 9.5919,
+      "step": 48930
+    },
+    {
+      "epoch": 0.24440060925366427,
+      "grad_norm": 0.09115762263536453,
+      "learning_rate": 2.272472402312949e-05,
+      "loss": 9.5915,
+      "step": 48940
+    },
+    {
+      "epoch": 0.2444505480786037,
+      "grad_norm": 0.09906817972660065,
+      "learning_rate": 2.272322210818794e-05,
+      "loss": 9.593,
+      "step": 48950
+    },
+    {
+      "epoch": 0.24450048690354315,
+      "grad_norm": 0.09307815879583359,
+      "learning_rate": 2.272172019324639e-05,
+      "loss": 9.5819,
+      "step": 48960
+    },
+    {
+      "epoch": 0.2445504257284826,
+      "grad_norm": 0.10123563557863235,
+      "learning_rate": 2.272021827830484e-05,
+      "loss": 9.59,
+      "step": 48970
+    },
+    {
+      "epoch": 0.24460036455342204,
+      "grad_norm": 0.09160536527633667,
+      "learning_rate": 2.2718716363363287e-05,
+      "loss": 9.5937,
+      "step": 48980
+    },
+    {
+      "epoch": 0.2446503033783615,
+      "grad_norm": 0.0956059992313385,
+      "learning_rate": 2.2717214448421738e-05,
+      "loss": 9.5891,
+      "step": 48990
+    },
+    {
+      "epoch": 0.24470024220330094,
+      "grad_norm": 0.08904523402452469,
+      "learning_rate": 2.2715712533480188e-05,
+      "loss": 9.5975,
+      "step": 49000
+    },
+    {
+      "epoch": 0.2447501810282404,
+      "grad_norm": 0.09728097915649414,
+      "learning_rate": 2.2714210618538638e-05,
+      "loss": 9.5892,
+      "step": 49010
+    },
+    {
+      "epoch": 0.24480011985317984,
+      "grad_norm": 0.0932324156165123,
+      "learning_rate": 2.2712708703597088e-05,
+      "loss": 9.5944,
+      "step": 49020
+    },
+    {
+      "epoch": 0.2448500586781193,
+      "grad_norm": 0.09640751779079437,
+      "learning_rate": 2.2711206788655535e-05,
+      "loss": 9.5948,
+      "step": 49030
+    },
+    {
+      "epoch": 0.24489999750305874,
+      "grad_norm": 0.09052955359220505,
+      "learning_rate": 2.2709704873713985e-05,
+      "loss": 9.5832,
+      "step": 49040
+    },
+    {
+      "epoch": 0.2449499363279982,
+      "grad_norm": 0.09775557368993759,
+      "learning_rate": 2.2708202958772435e-05,
+      "loss": 9.5852,
+      "step": 49050
+    },
+    {
+      "epoch": 0.24499987515293764,
+      "grad_norm": 0.0953780934214592,
+      "learning_rate": 2.2706701043830885e-05,
+      "loss": 9.5839,
+      "step": 49060
+    },
+    {
+      "epoch": 0.2450498139778771,
+      "grad_norm": 0.09656697511672974,
+      "learning_rate": 2.2705199128889336e-05,
+      "loss": 9.5869,
+      "step": 49070
+    },
+    {
+      "epoch": 0.24509975280281654,
+      "grad_norm": 0.09594185650348663,
+      "learning_rate": 2.2703697213947782e-05,
+      "loss": 9.5924,
+      "step": 49080
+    },
+    {
+      "epoch": 0.245149691627756,
+      "grad_norm": 0.09402401000261307,
+      "learning_rate": 2.2702195299006233e-05,
+      "loss": 9.5812,
+      "step": 49090
+    },
+    {
+      "epoch": 0.24519963045269544,
+      "grad_norm": 0.09970910847187042,
+      "learning_rate": 2.2700693384064683e-05,
+      "loss": 9.5868,
+      "step": 49100
+    },
+    {
+      "epoch": 0.2452495692776349,
+      "grad_norm": 0.09406423568725586,
+      "learning_rate": 2.2699191469123133e-05,
+      "loss": 9.5864,
+      "step": 49110
+    },
+    {
+      "epoch": 0.24529950810257434,
+      "grad_norm": 0.09630411863327026,
+      "learning_rate": 2.2697689554181583e-05,
+      "loss": 9.586,
+      "step": 49120
+    },
+    {
+      "epoch": 0.2453494469275138,
+      "grad_norm": 0.09320022165775299,
+      "learning_rate": 2.269618763924003e-05,
+      "loss": 9.5852,
+      "step": 49130
+    },
+    {
+      "epoch": 0.24539938575245324,
+      "grad_norm": 0.0887378454208374,
+      "learning_rate": 2.269468572429848e-05,
+      "loss": 9.5917,
+      "step": 49140
+    },
+    {
+      "epoch": 0.24544932457739269,
+      "grad_norm": 0.09708340466022491,
+      "learning_rate": 2.269318380935693e-05,
+      "loss": 9.5828,
+      "step": 49150
+    },
+    {
+      "epoch": 0.24549926340233214,
+      "grad_norm": 0.09893514215946198,
+      "learning_rate": 2.269168189441538e-05,
+      "loss": 9.5893,
+      "step": 49160
+    },
+    {
+      "epoch": 0.24554920222727158,
+      "grad_norm": 0.09499365836381912,
+      "learning_rate": 2.269017997947383e-05,
+      "loss": 9.5846,
+      "step": 49170
+    },
+    {
+      "epoch": 0.24559914105221103,
+      "grad_norm": 0.09623656421899796,
+      "learning_rate": 2.2688678064532277e-05,
+      "loss": 9.5843,
+      "step": 49180
+    },
+    {
+      "epoch": 0.24564907987715048,
+      "grad_norm": 0.09261438250541687,
+      "learning_rate": 2.268717614959073e-05,
+      "loss": 9.5892,
+      "step": 49190
+    },
+    {
+      "epoch": 0.24569901870208993,
+      "grad_norm": 0.08927629142999649,
+      "learning_rate": 2.2685674234649178e-05,
+      "loss": 9.5861,
+      "step": 49200
+    },
+    {
+      "epoch": 0.24574895752702938,
+      "grad_norm": 0.08987867087125778,
+      "learning_rate": 2.2684172319707628e-05,
+      "loss": 9.5839,
+      "step": 49210
+    },
+    {
+      "epoch": 0.24579889635196883,
+      "grad_norm": 0.09199423342943192,
+      "learning_rate": 2.2682670404766078e-05,
+      "loss": 9.5921,
+      "step": 49220
+    },
+    {
+      "epoch": 0.24584883517690828,
+      "grad_norm": 0.09492474794387817,
+      "learning_rate": 2.2681168489824525e-05,
+      "loss": 9.5821,
+      "step": 49230
+    },
+    {
+      "epoch": 0.24589877400184773,
+      "grad_norm": 0.09393247216939926,
+      "learning_rate": 2.267966657488298e-05,
+      "loss": 9.5838,
+      "step": 49240
+    },
+    {
+      "epoch": 0.24594871282678718,
+      "grad_norm": 0.09099004417657852,
+      "learning_rate": 2.2678164659941425e-05,
+      "loss": 9.5917,
+      "step": 49250
+    },
+    {
+      "epoch": 0.24599865165172663,
+      "grad_norm": 0.09554600715637207,
+      "learning_rate": 2.2676662744999875e-05,
+      "loss": 9.5849,
+      "step": 49260
+    },
+    {
+      "epoch": 0.24604859047666608,
+      "grad_norm": 0.09146859496831894,
+      "learning_rate": 2.2675160830058326e-05,
+      "loss": 9.5806,
+      "step": 49270
+    },
+    {
+      "epoch": 0.24609852930160553,
+      "grad_norm": 0.09467651695013046,
+      "learning_rate": 2.2673658915116772e-05,
+      "loss": 9.5945,
+      "step": 49280
+    },
+    {
+      "epoch": 0.24614846812654498,
+      "grad_norm": 0.09306614845991135,
+      "learning_rate": 2.2672157000175226e-05,
+      "loss": 9.5833,
+      "step": 49290
+    },
+    {
+      "epoch": 0.24619840695148443,
+      "grad_norm": 0.09221172332763672,
+      "learning_rate": 2.2670655085233673e-05,
+      "loss": 9.576,
+      "step": 49300
+    },
+    {
+      "epoch": 0.24624834577642388,
+      "grad_norm": 0.09125102311372757,
+      "learning_rate": 2.2669153170292123e-05,
+      "loss": 9.5789,
+      "step": 49310
+    },
+    {
+      "epoch": 0.24629828460136333,
+      "grad_norm": 0.09555398672819138,
+      "learning_rate": 2.2667651255350573e-05,
+      "loss": 9.5835,
+      "step": 49320
+    },
+    {
+      "epoch": 0.24634822342630278,
+      "grad_norm": 0.09381726384162903,
+      "learning_rate": 2.266614934040902e-05,
+      "loss": 9.5803,
+      "step": 49330
+    },
+    {
+      "epoch": 0.24639816225124223,
+      "grad_norm": 0.09570520371198654,
+      "learning_rate": 2.2664647425467473e-05,
+      "loss": 9.5808,
+      "step": 49340
+    },
+    {
+      "epoch": 0.24644810107618167,
+      "grad_norm": 0.08948148041963577,
+      "learning_rate": 2.266314551052592e-05,
+      "loss": 9.5837,
+      "step": 49350
+    },
+    {
+      "epoch": 0.24649803990112112,
+      "grad_norm": 0.09439452737569809,
+      "learning_rate": 2.266164359558437e-05,
+      "loss": 9.5819,
+      "step": 49360
+    },
+    {
+      "epoch": 0.24654797872606057,
+      "grad_norm": 0.09754544496536255,
+      "learning_rate": 2.266014168064282e-05,
+      "loss": 9.5824,
+      "step": 49370
+    },
+    {
+      "epoch": 0.24659791755100002,
+      "grad_norm": 0.09239448606967926,
+      "learning_rate": 2.2658639765701267e-05,
+      "loss": 9.5796,
+      "step": 49380
+    },
+    {
+      "epoch": 0.24664785637593947,
+      "grad_norm": 0.09098103642463684,
+      "learning_rate": 2.265713785075972e-05,
+      "loss": 9.5934,
+      "step": 49390
+    },
+    {
+      "epoch": 0.24669779520087892,
+      "grad_norm": 0.09144534915685654,
+      "learning_rate": 2.2655635935818168e-05,
+      "loss": 9.5866,
+      "step": 49400
+    },
+    {
+      "epoch": 0.24674773402581837,
+      "grad_norm": 0.09623269736766815,
+      "learning_rate": 2.2654134020876618e-05,
+      "loss": 9.5856,
+      "step": 49410
+    },
+    {
+      "epoch": 0.24679767285075782,
+      "grad_norm": 0.09903118759393692,
+      "learning_rate": 2.2652632105935068e-05,
+      "loss": 9.5846,
+      "step": 49420
+    },
+    {
+      "epoch": 0.24684761167569727,
+      "grad_norm": 0.09627585858106613,
+      "learning_rate": 2.2651130190993515e-05,
+      "loss": 9.578,
+      "step": 49430
+    },
+    {
+      "epoch": 0.24689755050063672,
+      "grad_norm": 0.09699337929487228,
+      "learning_rate": 2.264962827605197e-05,
+      "loss": 9.5867,
+      "step": 49440
+    },
+    {
+      "epoch": 0.24694748932557617,
+      "grad_norm": 0.08975996822118759,
+      "learning_rate": 2.2648126361110415e-05,
+      "loss": 9.5786,
+      "step": 49450
+    },
+    {
+      "epoch": 0.24699742815051562,
+      "grad_norm": 0.09180519729852676,
+      "learning_rate": 2.2646624446168865e-05,
+      "loss": 9.5793,
+      "step": 49460
+    },
+    {
+      "epoch": 0.24704736697545507,
+      "grad_norm": 0.09850934892892838,
+      "learning_rate": 2.2645122531227316e-05,
+      "loss": 9.581,
+      "step": 49470
+    },
+    {
+      "epoch": 0.24709730580039452,
+      "grad_norm": 0.09343817830085754,
+      "learning_rate": 2.2643620616285762e-05,
+      "loss": 9.5812,
+      "step": 49480
+    },
+    {
+      "epoch": 0.24714724462533397,
+      "grad_norm": 0.09533870965242386,
+      "learning_rate": 2.2642118701344216e-05,
+      "loss": 9.5801,
+      "step": 49490
+    },
+    {
+      "epoch": 0.24719718345027342,
+      "grad_norm": 0.08947062492370605,
+      "learning_rate": 2.2640616786402663e-05,
+      "loss": 9.5813,
+      "step": 49500
+    },
+    {
+      "epoch": 0.24724712227521287,
+      "grad_norm": 0.09306909888982773,
+      "learning_rate": 2.2639114871461116e-05,
+      "loss": 9.5802,
+      "step": 49510
+    },
+    {
+      "epoch": 0.24729706110015232,
+      "grad_norm": 0.09695802628993988,
+      "learning_rate": 2.2637612956519563e-05,
+      "loss": 9.5719,
+      "step": 49520
+    },
+    {
+      "epoch": 0.24734699992509177,
+      "grad_norm": 0.09378542751073837,
+      "learning_rate": 2.263611104157801e-05,
+      "loss": 9.5747,
+      "step": 49530
+    },
+    {
+      "epoch": 0.24739693875003121,
+      "grad_norm": 0.0942843034863472,
+      "learning_rate": 2.2634609126636463e-05,
+      "loss": 9.5746,
+      "step": 49540
+    },
+    {
+      "epoch": 0.24744687757497066,
+      "grad_norm": 0.09266909956932068,
+      "learning_rate": 2.263310721169491e-05,
+      "loss": 9.5729,
+      "step": 49550
+    },
+    {
+      "epoch": 0.2474968163999101,
+      "grad_norm": 0.09565820544958115,
+      "learning_rate": 2.2631605296753364e-05,
+      "loss": 9.5685,
+      "step": 49560
+    },
+    {
+      "epoch": 0.24754675522484956,
+      "grad_norm": 0.09224898368120193,
+      "learning_rate": 2.263010338181181e-05,
+      "loss": 9.586,
+      "step": 49570
+    },
+    {
+      "epoch": 0.247596694049789,
+      "grad_norm": 0.09254671633243561,
+      "learning_rate": 2.2628601466870257e-05,
+      "loss": 9.5773,
+      "step": 49580
+    },
+    {
+      "epoch": 0.24764663287472846,
+      "grad_norm": 0.09858168661594391,
+      "learning_rate": 2.262709955192871e-05,
+      "loss": 9.5796,
+      "step": 49590
+    },
+    {
+      "epoch": 0.2476965716996679,
+      "grad_norm": 0.0942915603518486,
+      "learning_rate": 2.2625597636987158e-05,
+      "loss": 9.5813,
+      "step": 49600
+    },
+    {
+      "epoch": 0.24774651052460736,
+      "grad_norm": 0.09052585065364838,
+      "learning_rate": 2.262409572204561e-05,
+      "loss": 9.5741,
+      "step": 49610
+    },
+    {
+      "epoch": 0.2477964493495468,
+      "grad_norm": 0.09947668015956879,
+      "learning_rate": 2.2622593807104058e-05,
+      "loss": 9.5746,
+      "step": 49620
+    },
+    {
+      "epoch": 0.24784638817448626,
+      "grad_norm": 0.09705556184053421,
+      "learning_rate": 2.2621091892162505e-05,
+      "loss": 9.5732,
+      "step": 49630
+    },
+    {
+      "epoch": 0.2478963269994257,
+      "grad_norm": 0.0956898182630539,
+      "learning_rate": 2.261958997722096e-05,
+      "loss": 9.5757,
+      "step": 49640
+    },
+    {
+      "epoch": 0.24794626582436516,
+      "grad_norm": 0.09402977675199509,
+      "learning_rate": 2.2618088062279405e-05,
+      "loss": 9.5683,
+      "step": 49650
+    },
+    {
+      "epoch": 0.2479962046493046,
+      "grad_norm": 0.09913989156484604,
+      "learning_rate": 2.261658614733786e-05,
+      "loss": 9.5776,
+      "step": 49660
+    },
+    {
+      "epoch": 0.24804614347424406,
+      "grad_norm": 0.09860474616289139,
+      "learning_rate": 2.2615084232396306e-05,
+      "loss": 9.5795,
+      "step": 49670
+    },
+    {
+      "epoch": 0.2480960822991835,
+      "grad_norm": 0.0972980335354805,
+      "learning_rate": 2.2613582317454752e-05,
+      "loss": 9.567,
+      "step": 49680
+    },
+    {
+      "epoch": 0.24814602112412296,
+      "grad_norm": 0.09505248814821243,
+      "learning_rate": 2.2612080402513206e-05,
+      "loss": 9.5719,
+      "step": 49690
+    },
+    {
+      "epoch": 0.2481959599490624,
+      "grad_norm": 0.09183008968830109,
+      "learning_rate": 2.2610578487571653e-05,
+      "loss": 9.5819,
+      "step": 49700
+    },
+    {
+      "epoch": 0.24824589877400186,
+      "grad_norm": 0.09852754324674606,
+      "learning_rate": 2.2609076572630106e-05,
+      "loss": 9.577,
+      "step": 49710
+    },
+    {
+      "epoch": 0.2482958375989413,
+      "grad_norm": 0.09413760155439377,
+      "learning_rate": 2.2607574657688553e-05,
+      "loss": 9.5743,
+      "step": 49720
+    },
+    {
+      "epoch": 0.24834577642388075,
+      "grad_norm": 0.09190001338720322,
+      "learning_rate": 2.2606072742747e-05,
+      "loss": 9.5768,
+      "step": 49730
+    },
+    {
+      "epoch": 0.2483957152488202,
+      "grad_norm": 0.09404946118593216,
+      "learning_rate": 2.2604570827805453e-05,
+      "loss": 9.5756,
+      "step": 49740
+    },
+    {
+      "epoch": 0.24844565407375965,
+      "grad_norm": 0.09731113910675049,
+      "learning_rate": 2.26030689128639e-05,
+      "loss": 9.5738,
+      "step": 49750
+    },
+    {
+      "epoch": 0.2484955928986991,
+      "grad_norm": 0.09584534913301468,
+      "learning_rate": 2.2601566997922354e-05,
+      "loss": 9.5745,
+      "step": 49760
+    },
+    {
+      "epoch": 0.24854553172363855,
+      "grad_norm": 0.09506174921989441,
+      "learning_rate": 2.26000650829808e-05,
+      "loss": 9.5702,
+      "step": 49770
+    },
+    {
+      "epoch": 0.248595470548578,
+      "grad_norm": 0.09768600016832352,
+      "learning_rate": 2.2598563168039247e-05,
+      "loss": 9.5804,
+      "step": 49780
+    },
+    {
+      "epoch": 0.24864540937351745,
+      "grad_norm": 0.09207095205783844,
+      "learning_rate": 2.25970612530977e-05,
+      "loss": 9.569,
+      "step": 49790
+    },
+    {
+      "epoch": 0.2486953481984569,
+      "grad_norm": 0.09260823577642441,
+      "learning_rate": 2.2595559338156148e-05,
+      "loss": 9.5739,
+      "step": 49800
+    },
+    {
+      "epoch": 0.24874528702339635,
+      "grad_norm": 0.09007450938224792,
+      "learning_rate": 2.25940574232146e-05,
+      "loss": 9.5792,
+      "step": 49810
+    },
+    {
+      "epoch": 0.2487952258483358,
+      "grad_norm": 0.10160695761442184,
+      "learning_rate": 2.2592555508273048e-05,
+      "loss": 9.5803,
+      "step": 49820
+    },
+    {
+      "epoch": 0.24884516467327525,
+      "grad_norm": 0.09343059360980988,
+      "learning_rate": 2.25910535933315e-05,
+      "loss": 9.5758,
+      "step": 49830
+    },
+    {
+      "epoch": 0.2488951034982147,
+      "grad_norm": 0.09455156326293945,
+      "learning_rate": 2.258955167838995e-05,
+      "loss": 9.5728,
+      "step": 49840
+    },
+    {
+      "epoch": 0.24894504232315415,
+      "grad_norm": 0.09680726379156113,
+      "learning_rate": 2.2588049763448395e-05,
+      "loss": 9.5752,
+      "step": 49850
+    },
+    {
+      "epoch": 0.2489949811480936,
+      "grad_norm": 0.09316403418779373,
+      "learning_rate": 2.258654784850685e-05,
+      "loss": 9.5631,
+      "step": 49860
+    },
+    {
+      "epoch": 0.24904491997303305,
+      "grad_norm": 0.0938340425491333,
+      "learning_rate": 2.2585045933565296e-05,
+      "loss": 9.5747,
+      "step": 49870
+    },
+    {
+      "epoch": 0.2490948587979725,
+      "grad_norm": 0.094419926404953,
+      "learning_rate": 2.2583544018623746e-05,
+      "loss": 9.5744,
+      "step": 49880
+    },
+    {
+      "epoch": 0.24914479762291195,
+      "grad_norm": 0.09464523941278458,
+      "learning_rate": 2.2582042103682196e-05,
+      "loss": 9.5676,
+      "step": 49890
+    },
+    {
+      "epoch": 0.2491947364478514,
+      "grad_norm": 0.09369165450334549,
+      "learning_rate": 2.2580540188740643e-05,
+      "loss": 9.5752,
+      "step": 49900
+    },
+    {
+      "epoch": 0.24924467527279084,
+      "grad_norm": 0.09760770201683044,
+      "learning_rate": 2.2579038273799096e-05,
+      "loss": 9.5642,
+      "step": 49910
+    },
+    {
+      "epoch": 0.2492946140977303,
+      "grad_norm": 0.09786707162857056,
+      "learning_rate": 2.2577536358857543e-05,
+      "loss": 9.5733,
+      "step": 49920
+    },
+    {
+      "epoch": 0.24934455292266974,
+      "grad_norm": 0.09324178099632263,
+      "learning_rate": 2.2576034443915993e-05,
+      "loss": 9.5683,
+      "step": 49930
+    },
+    {
+      "epoch": 0.24939449174760916,
+      "grad_norm": 0.09547551721334457,
+      "learning_rate": 2.2574532528974444e-05,
+      "loss": 9.5693,
+      "step": 49940
+    },
+    {
+      "epoch": 0.24944443057254861,
+      "grad_norm": 0.09578131884336472,
+      "learning_rate": 2.257303061403289e-05,
+      "loss": 9.5682,
+      "step": 49950
+    },
+    {
+      "epoch": 0.24949436939748806,
+      "grad_norm": 0.09277050942182541,
+      "learning_rate": 2.2571528699091344e-05,
+      "loss": 9.5733,
+      "step": 49960
+    },
+    {
+      "epoch": 0.2495443082224275,
+      "grad_norm": 0.09631931036710739,
+      "learning_rate": 2.257002678414979e-05,
+      "loss": 9.5709,
+      "step": 49970
+    },
+    {
+      "epoch": 0.24959424704736696,
+      "grad_norm": 0.09734215587377548,
+      "learning_rate": 2.256852486920824e-05,
+      "loss": 9.5697,
+      "step": 49980
+    },
+    {
+      "epoch": 0.2496441858723064,
+      "grad_norm": 0.09608133882284164,
+      "learning_rate": 2.256702295426669e-05,
+      "loss": 9.5712,
+      "step": 49990
+    },
+    {
+      "epoch": 0.24969412469724586,
+      "grad_norm": 0.09397196024656296,
+      "learning_rate": 2.2565521039325138e-05,
+      "loss": 9.571,
+      "step": 50000
+    },
+    {
+      "epoch": 0.2497440635221853,
+      "grad_norm": 0.09108542650938034,
+      "learning_rate": 2.256401912438359e-05,
+      "loss": 9.5636,
+      "step": 50010
+    },
+    {
+      "epoch": 0.24979400234712476,
+      "grad_norm": 0.09850990772247314,
+      "learning_rate": 2.2562517209442038e-05,
+      "loss": 9.5599,
+      "step": 50020
+    },
+    {
+      "epoch": 0.2498439411720642,
+      "grad_norm": 0.10261461138725281,
+      "learning_rate": 2.256101529450049e-05,
+      "loss": 9.5668,
+      "step": 50030
+    },
+    {
+      "epoch": 0.24989387999700366,
+      "grad_norm": 0.09602981060743332,
+      "learning_rate": 2.255951337955894e-05,
+      "loss": 9.5784,
+      "step": 50040
+    },
+    {
+      "epoch": 0.2499438188219431,
+      "grad_norm": 0.10035931318998337,
+      "learning_rate": 2.2558011464617385e-05,
+      "loss": 9.5678,
+      "step": 50050
+    },
+    {
+      "epoch": 0.24999375764688256,
+      "grad_norm": 0.09186173975467682,
+      "learning_rate": 2.255650954967584e-05,
+      "loss": 9.5682,
+      "step": 50060
+    },
+    {
+      "epoch": 0.250043696471822,
+      "grad_norm": 0.09531144052743912,
+      "learning_rate": 2.2555007634734286e-05,
+      "loss": 9.5753,
+      "step": 50070
+    },
+    {
+      "epoch": 0.2500936352967615,
+      "grad_norm": 0.0925971195101738,
+      "learning_rate": 2.2553505719792736e-05,
+      "loss": 9.5576,
+      "step": 50080
+    },
+    {
+      "epoch": 0.2501435741217009,
+      "grad_norm": 0.10341840982437134,
+      "learning_rate": 2.2552003804851186e-05,
+      "loss": 9.5649,
+      "step": 50090
+    },
+    {
+      "epoch": 0.2501935129466404,
+      "grad_norm": 0.08949480205774307,
+      "learning_rate": 2.2550501889909633e-05,
+      "loss": 9.5681,
+      "step": 50100
+    },
+    {
+      "epoch": 0.2502434517715798,
+      "grad_norm": 0.0952087789773941,
+      "learning_rate": 2.2548999974968086e-05,
+      "loss": 9.5719,
+      "step": 50110
+    },
+    {
+      "epoch": 0.2502933905965193,
+      "grad_norm": 0.09297273308038712,
+      "learning_rate": 2.2547498060026533e-05,
+      "loss": 9.5602,
+      "step": 50120
+    },
+    {
+      "epoch": 0.2503433294214587,
+      "grad_norm": 0.09558509290218353,
+      "learning_rate": 2.2545996145084983e-05,
+      "loss": 9.5775,
+      "step": 50130
+    },
+    {
+      "epoch": 0.2503932682463982,
+      "grad_norm": 0.09406739473342896,
+      "learning_rate": 2.2544494230143434e-05,
+      "loss": 9.5728,
+      "step": 50140
+    },
+    {
+      "epoch": 0.2504432070713376,
+      "grad_norm": 0.08946365118026733,
+      "learning_rate": 2.254299231520188e-05,
+      "loss": 9.5646,
+      "step": 50150
+    },
+    {
+      "epoch": 0.2504931458962771,
+      "grad_norm": 0.0948345735669136,
+      "learning_rate": 2.2541490400260334e-05,
+      "loss": 9.5688,
+      "step": 50160
+    },
+    {
+      "epoch": 0.2505430847212165,
+      "grad_norm": 0.09642423689365387,
+      "learning_rate": 2.253998848531878e-05,
+      "loss": 9.5646,
+      "step": 50170
+    },
+    {
+      "epoch": 0.250593023546156,
+      "grad_norm": 0.09099261462688446,
+      "learning_rate": 2.253848657037723e-05,
+      "loss": 9.5648,
+      "step": 50180
+    },
+    {
+      "epoch": 0.2506429623710954,
+      "grad_norm": 0.09220701456069946,
+      "learning_rate": 2.253698465543568e-05,
+      "loss": 9.5665,
+      "step": 50190
+    },
+    {
+      "epoch": 0.2506929011960349,
+      "grad_norm": 0.09915616363286972,
+      "learning_rate": 2.253548274049413e-05,
+      "loss": 9.5697,
+      "step": 50200
+    },
+    {
+      "epoch": 0.2507428400209743,
+      "grad_norm": 0.0929388552904129,
+      "learning_rate": 2.253398082555258e-05,
+      "loss": 9.5623,
+      "step": 50210
+    },
+    {
+      "epoch": 0.2507927788459138,
+      "grad_norm": 0.09379485994577408,
+      "learning_rate": 2.2532478910611028e-05,
+      "loss": 9.5631,
+      "step": 50220
+    },
+    {
+      "epoch": 0.2508427176708532,
+      "grad_norm": 0.0981220230460167,
+      "learning_rate": 2.253097699566948e-05,
+      "loss": 9.5638,
+      "step": 50230
+    },
+    {
+      "epoch": 0.2508926564957927,
+      "grad_norm": 0.09549273550510406,
+      "learning_rate": 2.252947508072793e-05,
+      "loss": 9.5607,
+      "step": 50240
+    },
+    {
+      "epoch": 0.2509425953207321,
+      "grad_norm": 0.09716612845659256,
+      "learning_rate": 2.252797316578638e-05,
+      "loss": 9.5685,
+      "step": 50250
+    },
+    {
+      "epoch": 0.2509925341456716,
+      "grad_norm": 0.09582599252462387,
+      "learning_rate": 2.252647125084483e-05,
+      "loss": 9.5682,
+      "step": 50260
+    },
+    {
+      "epoch": 0.251042472970611,
+      "grad_norm": 0.09806785732507706,
+      "learning_rate": 2.2524969335903276e-05,
+      "loss": 9.5639,
+      "step": 50270
+    },
+    {
+      "epoch": 0.2510924117955505,
+      "grad_norm": 0.10009253770112991,
+      "learning_rate": 2.2523467420961726e-05,
+      "loss": 9.5696,
+      "step": 50280
+    },
+    {
+      "epoch": 0.2511423506204899,
+      "grad_norm": 0.10564589500427246,
+      "learning_rate": 2.2521965506020176e-05,
+      "loss": 9.5722,
+      "step": 50290
+    },
+    {
+      "epoch": 0.2511922894454294,
+      "grad_norm": 0.09391903877258301,
+      "learning_rate": 2.2520463591078626e-05,
+      "loss": 9.5661,
+      "step": 50300
+    },
+    {
+      "epoch": 0.2512422282703688,
+      "grad_norm": 0.09433561563491821,
+      "learning_rate": 2.2518961676137076e-05,
+      "loss": 9.5701,
+      "step": 50310
+    },
+    {
+      "epoch": 0.25129216709530827,
+      "grad_norm": 0.09320453554391861,
+      "learning_rate": 2.2517459761195523e-05,
+      "loss": 9.5589,
+      "step": 50320
+    },
+    {
+      "epoch": 0.2513421059202477,
+      "grad_norm": 0.0926797166466713,
+      "learning_rate": 2.2515957846253973e-05,
+      "loss": 9.5641,
+      "step": 50330
+    },
+    {
+      "epoch": 0.25139204474518717,
+      "grad_norm": 0.09965445101261139,
+      "learning_rate": 2.2514455931312424e-05,
+      "loss": 9.572,
+      "step": 50340
+    },
+    {
+      "epoch": 0.2514419835701266,
+      "grad_norm": 0.09860392659902573,
+      "learning_rate": 2.2512954016370874e-05,
+      "loss": 9.569,
+      "step": 50350
+    },
+    {
+      "epoch": 0.25149192239506607,
+      "grad_norm": 0.09547075629234314,
+      "learning_rate": 2.2511452101429324e-05,
+      "loss": 9.5689,
+      "step": 50360
+    },
+    {
+      "epoch": 0.2515418612200055,
+      "grad_norm": 0.09795694053173065,
+      "learning_rate": 2.250995018648777e-05,
+      "loss": 9.5628,
+      "step": 50370
+    },
+    {
+      "epoch": 0.25159180004494497,
+      "grad_norm": 0.09307505935430527,
+      "learning_rate": 2.250844827154622e-05,
+      "loss": 9.5662,
+      "step": 50380
+    },
+    {
+      "epoch": 0.2516417388698844,
+      "grad_norm": 0.10071365535259247,
+      "learning_rate": 2.250694635660467e-05,
+      "loss": 9.569,
+      "step": 50390
+    },
+    {
+      "epoch": 0.25169167769482387,
+      "grad_norm": 0.09713767468929291,
+      "learning_rate": 2.250544444166312e-05,
+      "loss": 9.5577,
+      "step": 50400
+    },
+    {
+      "epoch": 0.2517416165197633,
+      "grad_norm": 0.09521119296550751,
+      "learning_rate": 2.250394252672157e-05,
+      "loss": 9.5673,
+      "step": 50410
+    },
+    {
+      "epoch": 0.25179155534470277,
+      "grad_norm": 0.09349004924297333,
+      "learning_rate": 2.2502440611780018e-05,
+      "loss": 9.5571,
+      "step": 50420
+    },
+    {
+      "epoch": 0.2518414941696422,
+      "grad_norm": 0.0950050801038742,
+      "learning_rate": 2.250093869683847e-05,
+      "loss": 9.5636,
+      "step": 50430
+    },
+    {
+      "epoch": 0.2518914329945816,
+      "grad_norm": 0.09199751913547516,
+      "learning_rate": 2.249943678189692e-05,
+      "loss": 9.5595,
+      "step": 50440
+    },
+    {
+      "epoch": 0.2519413718195211,
+      "grad_norm": 0.09100238978862762,
+      "learning_rate": 2.249793486695537e-05,
+      "loss": 9.5609,
+      "step": 50450
+    },
+    {
+      "epoch": 0.2519913106444605,
+      "grad_norm": 0.09246649593114853,
+      "learning_rate": 2.249643295201382e-05,
+      "loss": 9.5659,
+      "step": 50460
+    },
+    {
+      "epoch": 0.2520412494694,
+      "grad_norm": 0.09442534297704697,
+      "learning_rate": 2.2494931037072266e-05,
+      "loss": 9.5646,
+      "step": 50470
+    },
+    {
+      "epoch": 0.2520911882943394,
+      "grad_norm": 0.08858856558799744,
+      "learning_rate": 2.249342912213072e-05,
+      "loss": 9.5523,
+      "step": 50480
+    },
+    {
+      "epoch": 0.2521411271192789,
+      "grad_norm": 0.09395717829465866,
+      "learning_rate": 2.2491927207189166e-05,
+      "loss": 9.5661,
+      "step": 50490
+    },
+    {
+      "epoch": 0.2521910659442183,
+      "grad_norm": 0.101358562707901,
+      "learning_rate": 2.2490425292247616e-05,
+      "loss": 9.5592,
+      "step": 50500
+    },
+    {
+      "epoch": 0.2522410047691578,
+      "grad_norm": 0.09509819746017456,
+      "learning_rate": 2.2488923377306066e-05,
+      "loss": 9.5605,
+      "step": 50510
+    },
+    {
+      "epoch": 0.2522909435940972,
+      "grad_norm": 0.09509771317243576,
+      "learning_rate": 2.2487421462364517e-05,
+      "loss": 9.561,
+      "step": 50520
+    },
+    {
+      "epoch": 0.2523408824190367,
+      "grad_norm": 0.09046350419521332,
+      "learning_rate": 2.2485919547422967e-05,
+      "loss": 9.5595,
+      "step": 50530
+    },
+    {
+      "epoch": 0.2523908212439761,
+      "grad_norm": 0.09777220338582993,
+      "learning_rate": 2.2484417632481414e-05,
+      "loss": 9.5526,
+      "step": 50540
+    },
+    {
+      "epoch": 0.2524407600689156,
+      "grad_norm": 0.09290176630020142,
+      "learning_rate": 2.2482915717539864e-05,
+      "loss": 9.562,
+      "step": 50550
+    },
+    {
+      "epoch": 0.252490698893855,
+      "grad_norm": 0.0935697928071022,
+      "learning_rate": 2.2481413802598314e-05,
+      "loss": 9.5649,
+      "step": 50560
+    },
+    {
+      "epoch": 0.2525406377187945,
+      "grad_norm": 0.09431051462888718,
+      "learning_rate": 2.2479911887656764e-05,
+      "loss": 9.5558,
+      "step": 50570
+    },
+    {
+      "epoch": 0.2525905765437339,
+      "grad_norm": 0.09408137947320938,
+      "learning_rate": 2.2478409972715214e-05,
+      "loss": 9.5587,
+      "step": 50580
+    },
+    {
+      "epoch": 0.2526405153686734,
+      "grad_norm": 0.0971836969256401,
+      "learning_rate": 2.247690805777366e-05,
+      "loss": 9.5572,
+      "step": 50590
+    },
+    {
+      "epoch": 0.2526904541936128,
+      "grad_norm": 0.09811126440763474,
+      "learning_rate": 2.247540614283211e-05,
+      "loss": 9.5591,
+      "step": 50600
+    },
+    {
+      "epoch": 0.2527403930185523,
+      "grad_norm": 0.09719320386648178,
+      "learning_rate": 2.247390422789056e-05,
+      "loss": 9.5642,
+      "step": 50610
+    },
+    {
+      "epoch": 0.2527903318434917,
+      "grad_norm": 0.09279009699821472,
+      "learning_rate": 2.247240231294901e-05,
+      "loss": 9.554,
+      "step": 50620
+    },
+    {
+      "epoch": 0.2528402706684312,
+      "grad_norm": 0.09156560897827148,
+      "learning_rate": 2.2470900398007462e-05,
+      "loss": 9.5613,
+      "step": 50630
+    },
+    {
+      "epoch": 0.2528902094933706,
+      "grad_norm": 0.08856366574764252,
+      "learning_rate": 2.246939848306591e-05,
+      "loss": 9.5541,
+      "step": 50640
+    },
+    {
+      "epoch": 0.2529401483183101,
+      "grad_norm": 0.09104395657777786,
+      "learning_rate": 2.246789656812436e-05,
+      "loss": 9.5616,
+      "step": 50650
+    },
+    {
+      "epoch": 0.2529900871432495,
+      "grad_norm": 0.09743545204401016,
+      "learning_rate": 2.246639465318281e-05,
+      "loss": 9.5614,
+      "step": 50660
+    },
+    {
+      "epoch": 0.253040025968189,
+      "grad_norm": 0.09513422101736069,
+      "learning_rate": 2.246489273824126e-05,
+      "loss": 9.5536,
+      "step": 50670
+    },
+    {
+      "epoch": 0.2530899647931284,
+      "grad_norm": 0.09526988118886948,
+      "learning_rate": 2.246339082329971e-05,
+      "loss": 9.5638,
+      "step": 50680
+    },
+    {
+      "epoch": 0.2531399036180679,
+      "grad_norm": 0.0918373093008995,
+      "learning_rate": 2.2461888908358156e-05,
+      "loss": 9.5538,
+      "step": 50690
+    },
+    {
+      "epoch": 0.2531898424430073,
+      "grad_norm": 0.09240086376667023,
+      "learning_rate": 2.2460386993416606e-05,
+      "loss": 9.5696,
+      "step": 50700
+    },
+    {
+      "epoch": 0.2532397812679468,
+      "grad_norm": 0.09589746594429016,
+      "learning_rate": 2.2458885078475056e-05,
+      "loss": 9.5527,
+      "step": 50710
+    },
+    {
+      "epoch": 0.2532897200928862,
+      "grad_norm": 0.0918283686041832,
+      "learning_rate": 2.2457383163533507e-05,
+      "loss": 9.5575,
+      "step": 50720
+    },
+    {
+      "epoch": 0.25333965891782567,
+      "grad_norm": 0.09194007515907288,
+      "learning_rate": 2.2455881248591957e-05,
+      "loss": 9.5578,
+      "step": 50730
+    },
+    {
+      "epoch": 0.2533895977427651,
+      "grad_norm": 0.09658579528331757,
+      "learning_rate": 2.2454379333650404e-05,
+      "loss": 9.5613,
+      "step": 50740
+    },
+    {
+      "epoch": 0.25343953656770457,
+      "grad_norm": 0.09084783494472504,
+      "learning_rate": 2.2452877418708854e-05,
+      "loss": 9.5606,
+      "step": 50750
+    },
+    {
+      "epoch": 0.253489475392644,
+      "grad_norm": 0.09604687243700027,
+      "learning_rate": 2.2451375503767304e-05,
+      "loss": 9.5488,
+      "step": 50760
+    },
+    {
+      "epoch": 0.25353941421758347,
+      "grad_norm": 0.08728381246328354,
+      "learning_rate": 2.2449873588825754e-05,
+      "loss": 9.555,
+      "step": 50770
+    },
+    {
+      "epoch": 0.2535893530425229,
+      "grad_norm": 0.09549268335103989,
+      "learning_rate": 2.2448371673884204e-05,
+      "loss": 9.5534,
+      "step": 50780
+    },
+    {
+      "epoch": 0.25363929186746237,
+      "grad_norm": 0.0957789346575737,
+      "learning_rate": 2.244686975894265e-05,
+      "loss": 9.5565,
+      "step": 50790
+    },
+    {
+      "epoch": 0.2536892306924018,
+      "grad_norm": 0.09618537127971649,
+      "learning_rate": 2.24453678440011e-05,
+      "loss": 9.5515,
+      "step": 50800
+    },
+    {
+      "epoch": 0.25373916951734127,
+      "grad_norm": 0.0949334055185318,
+      "learning_rate": 2.244386592905955e-05,
+      "loss": 9.5482,
+      "step": 50810
+    },
+    {
+      "epoch": 0.2537891083422807,
+      "grad_norm": 0.09194260835647583,
+      "learning_rate": 2.2442364014118e-05,
+      "loss": 9.5547,
+      "step": 50820
+    },
+    {
+      "epoch": 0.25383904716722017,
+      "grad_norm": 0.09208869934082031,
+      "learning_rate": 2.2440862099176452e-05,
+      "loss": 9.561,
+      "step": 50830
+    },
+    {
+      "epoch": 0.2538889859921596,
+      "grad_norm": 0.0971684604883194,
+      "learning_rate": 2.2439360184234902e-05,
+      "loss": 9.559,
+      "step": 50840
+    },
+    {
+      "epoch": 0.25393892481709907,
+      "grad_norm": 0.09714303910732269,
+      "learning_rate": 2.243785826929335e-05,
+      "loss": 9.5505,
+      "step": 50850
+    },
+    {
+      "epoch": 0.2539888636420385,
+      "grad_norm": 0.09106447547674179,
+      "learning_rate": 2.24363563543518e-05,
+      "loss": 9.5475,
+      "step": 50860
+    },
+    {
+      "epoch": 0.25403880246697796,
+      "grad_norm": 0.09171400219202042,
+      "learning_rate": 2.243485443941025e-05,
+      "loss": 9.5539,
+      "step": 50870
+    },
+    {
+      "epoch": 0.2540887412919174,
+      "grad_norm": 0.09836923331022263,
+      "learning_rate": 2.24333525244687e-05,
+      "loss": 9.5509,
+      "step": 50880
+    },
+    {
+      "epoch": 0.25413868011685686,
+      "grad_norm": 0.09668246656656265,
+      "learning_rate": 2.243185060952715e-05,
+      "loss": 9.5561,
+      "step": 50890
+    },
+    {
+      "epoch": 0.2541886189417963,
+      "grad_norm": 0.09654278308153152,
+      "learning_rate": 2.2430348694585596e-05,
+      "loss": 9.5565,
+      "step": 50900
+    },
+    {
+      "epoch": 0.25423855776673576,
+      "grad_norm": 0.09737430512905121,
+      "learning_rate": 2.2428846779644046e-05,
+      "loss": 9.5577,
+      "step": 50910
+    },
+    {
+      "epoch": 0.2542884965916752,
+      "grad_norm": 0.09896323084831238,
+      "learning_rate": 2.2427344864702497e-05,
+      "loss": 9.5504,
+      "step": 50920
+    },
+    {
+      "epoch": 0.25433843541661466,
+      "grad_norm": 0.09133461862802505,
+      "learning_rate": 2.2425842949760947e-05,
+      "loss": 9.5474,
+      "step": 50930
+    },
+    {
+      "epoch": 0.2543883742415541,
+      "grad_norm": 0.09348398447036743,
+      "learning_rate": 2.2424341034819397e-05,
+      "loss": 9.5559,
+      "step": 50940
+    },
+    {
+      "epoch": 0.25443831306649356,
+      "grad_norm": 0.09214718639850616,
+      "learning_rate": 2.2422839119877844e-05,
+      "loss": 9.5534,
+      "step": 50950
+    },
+    {
+      "epoch": 0.254488251891433,
+      "grad_norm": 0.09419059008359909,
+      "learning_rate": 2.2421337204936294e-05,
+      "loss": 9.5525,
+      "step": 50960
+    },
+    {
+      "epoch": 0.25453819071637246,
+      "grad_norm": 0.10455143451690674,
+      "learning_rate": 2.2419835289994744e-05,
+      "loss": 9.5567,
+      "step": 50970
+    },
+    {
+      "epoch": 0.2545881295413119,
+      "grad_norm": 0.09650573879480362,
+      "learning_rate": 2.2418333375053194e-05,
+      "loss": 9.5585,
+      "step": 50980
+    },
+    {
+      "epoch": 0.25463806836625136,
+      "grad_norm": 0.09631022065877914,
+      "learning_rate": 2.2416831460111644e-05,
+      "loss": 9.5525,
+      "step": 50990
+    },
+    {
+      "epoch": 0.2546880071911908,
+      "grad_norm": 0.09867773205041885,
+      "learning_rate": 2.241532954517009e-05,
+      "loss": 9.5604,
+      "step": 51000
+    },
+    {
+      "epoch": 0.25473794601613026,
+      "grad_norm": 0.09457637369632721,
+      "learning_rate": 2.241382763022854e-05,
+      "loss": 9.5552,
+      "step": 51010
+    },
+    {
+      "epoch": 0.2547878848410697,
+      "grad_norm": 0.09107904881238937,
+      "learning_rate": 2.241232571528699e-05,
+      "loss": 9.5587,
+      "step": 51020
+    },
+    {
+      "epoch": 0.25483782366600916,
+      "grad_norm": 0.0893888846039772,
+      "learning_rate": 2.2410823800345442e-05,
+      "loss": 9.5496,
+      "step": 51030
+    },
+    {
+      "epoch": 0.2548877624909486,
+      "grad_norm": 0.09376450628042221,
+      "learning_rate": 2.2409321885403892e-05,
+      "loss": 9.5533,
+      "step": 51040
+    },
+    {
+      "epoch": 0.25493770131588805,
+      "grad_norm": 0.09373288601636887,
+      "learning_rate": 2.240781997046234e-05,
+      "loss": 9.5508,
+      "step": 51050
+    },
+    {
+      "epoch": 0.2549876401408275,
+      "grad_norm": 0.08769457787275314,
+      "learning_rate": 2.240631805552079e-05,
+      "loss": 9.5622,
+      "step": 51060
+    },
+    {
+      "epoch": 0.25503757896576695,
+      "grad_norm": 0.09537973999977112,
+      "learning_rate": 2.240481614057924e-05,
+      "loss": 9.5541,
+      "step": 51070
+    },
+    {
+      "epoch": 0.2550875177907064,
+      "grad_norm": 0.09325198084115982,
+      "learning_rate": 2.240331422563769e-05,
+      "loss": 9.5471,
+      "step": 51080
+    },
+    {
+      "epoch": 0.25513745661564585,
+      "grad_norm": 0.08798488229513168,
+      "learning_rate": 2.240181231069614e-05,
+      "loss": 9.5516,
+      "step": 51090
+    },
+    {
+      "epoch": 0.2551873954405853,
+      "grad_norm": 0.09604234993457794,
+      "learning_rate": 2.2400310395754586e-05,
+      "loss": 9.5472,
+      "step": 51100
+    },
+    {
+      "epoch": 0.25523733426552475,
+      "grad_norm": 0.0939023569226265,
+      "learning_rate": 2.2398808480813036e-05,
+      "loss": 9.5409,
+      "step": 51110
+    },
+    {
+      "epoch": 0.2552872730904642,
+      "grad_norm": 0.09348843991756439,
+      "learning_rate": 2.2397306565871487e-05,
+      "loss": 9.5484,
+      "step": 51120
+    },
+    {
+      "epoch": 0.25533721191540365,
+      "grad_norm": 0.0899486169219017,
+      "learning_rate": 2.2395804650929937e-05,
+      "loss": 9.547,
+      "step": 51130
+    },
+    {
+      "epoch": 0.25538715074034307,
+      "grad_norm": 0.0945720225572586,
+      "learning_rate": 2.2394302735988387e-05,
+      "loss": 9.5554,
+      "step": 51140
+    },
+    {
+      "epoch": 0.25543708956528255,
+      "grad_norm": 0.0940125435590744,
+      "learning_rate": 2.2392800821046834e-05,
+      "loss": 9.5527,
+      "step": 51150
+    },
+    {
+      "epoch": 0.25548702839022197,
+      "grad_norm": 0.09849490225315094,
+      "learning_rate": 2.2391298906105287e-05,
+      "loss": 9.5505,
+      "step": 51160
+    },
+    {
+      "epoch": 0.25553696721516145,
+      "grad_norm": 0.08866614103317261,
+      "learning_rate": 2.2389796991163734e-05,
+      "loss": 9.5502,
+      "step": 51170
+    },
+    {
+      "epoch": 0.25558690604010087,
+      "grad_norm": 0.09326382726430893,
+      "learning_rate": 2.2388295076222184e-05,
+      "loss": 9.5456,
+      "step": 51180
+    },
+    {
+      "epoch": 0.25563684486504035,
+      "grad_norm": 0.09560631960630417,
+      "learning_rate": 2.2386793161280634e-05,
+      "loss": 9.544,
+      "step": 51190
+    },
+    {
+      "epoch": 0.25568678368997977,
+      "grad_norm": 0.09622514992952347,
+      "learning_rate": 2.238529124633908e-05,
+      "loss": 9.5561,
+      "step": 51200
+    },
+    {
+      "epoch": 0.25573672251491925,
+      "grad_norm": 0.09464744478464127,
+      "learning_rate": 2.2383789331397535e-05,
+      "loss": 9.5411,
+      "step": 51210
+    },
+    {
+      "epoch": 0.25578666133985867,
+      "grad_norm": 0.09231311827898026,
+      "learning_rate": 2.238228741645598e-05,
+      "loss": 9.5537,
+      "step": 51220
+    },
+    {
+      "epoch": 0.25583660016479814,
+      "grad_norm": 0.09209968149662018,
+      "learning_rate": 2.2380785501514432e-05,
+      "loss": 9.5526,
+      "step": 51230
+    },
+    {
+      "epoch": 0.25588653898973757,
+      "grad_norm": 0.09474463015794754,
+      "learning_rate": 2.2379283586572882e-05,
+      "loss": 9.5419,
+      "step": 51240
+    },
+    {
+      "epoch": 0.25593647781467704,
+      "grad_norm": 0.09012725204229355,
+      "learning_rate": 2.237778167163133e-05,
+      "loss": 9.5522,
+      "step": 51250
+    },
+    {
+      "epoch": 0.25598641663961647,
+      "grad_norm": 0.09033194929361343,
+      "learning_rate": 2.2376279756689782e-05,
+      "loss": 9.5544,
+      "step": 51260
+    },
+    {
+      "epoch": 0.25603635546455594,
+      "grad_norm": 0.0948658287525177,
+      "learning_rate": 2.237477784174823e-05,
+      "loss": 9.546,
+      "step": 51270
+    },
+    {
+      "epoch": 0.25608629428949536,
+      "grad_norm": 0.09429352730512619,
+      "learning_rate": 2.237327592680668e-05,
+      "loss": 9.5531,
+      "step": 51280
+    },
+    {
+      "epoch": 0.25613623311443484,
+      "grad_norm": 0.09330032765865326,
+      "learning_rate": 2.237177401186513e-05,
+      "loss": 9.5449,
+      "step": 51290
+    },
+    {
+      "epoch": 0.25618617193937426,
+      "grad_norm": 0.08929108828306198,
+      "learning_rate": 2.2370272096923576e-05,
+      "loss": 9.551,
+      "step": 51300
+    },
+    {
+      "epoch": 0.25623611076431374,
+      "grad_norm": 0.0942007303237915,
+      "learning_rate": 2.236877018198203e-05,
+      "loss": 9.5506,
+      "step": 51310
+    },
+    {
+      "epoch": 0.25628604958925316,
+      "grad_norm": 0.09233163297176361,
+      "learning_rate": 2.2367268267040477e-05,
+      "loss": 9.5422,
+      "step": 51320
+    },
+    {
+      "epoch": 0.25633598841419264,
+      "grad_norm": 0.09945209324359894,
+      "learning_rate": 2.2365766352098927e-05,
+      "loss": 9.539,
+      "step": 51330
+    },
+    {
+      "epoch": 0.25638592723913206,
+      "grad_norm": 0.09467910975217819,
+      "learning_rate": 2.2364264437157377e-05,
+      "loss": 9.551,
+      "step": 51340
+    },
+    {
+      "epoch": 0.25643586606407154,
+      "grad_norm": 0.09829029440879822,
+      "learning_rate": 2.2362762522215824e-05,
+      "loss": 9.5447,
+      "step": 51350
+    },
+    {
+      "epoch": 0.25648580488901096,
+      "grad_norm": 0.09287157654762268,
+      "learning_rate": 2.2361260607274277e-05,
+      "loss": 9.5401,
+      "step": 51360
+    },
+    {
+      "epoch": 0.25653574371395044,
+      "grad_norm": 0.09296029806137085,
+      "learning_rate": 2.2359758692332724e-05,
+      "loss": 9.5441,
+      "step": 51370
+    },
+    {
+      "epoch": 0.25658568253888986,
+      "grad_norm": 0.09422246366739273,
+      "learning_rate": 2.2358256777391174e-05,
+      "loss": 9.5485,
+      "step": 51380
+    },
+    {
+      "epoch": 0.25663562136382934,
+      "grad_norm": 0.0900813490152359,
+      "learning_rate": 2.2356754862449624e-05,
+      "loss": 9.5456,
+      "step": 51390
+    },
+    {
+      "epoch": 0.25668556018876876,
+      "grad_norm": 0.09812017530202866,
+      "learning_rate": 2.235525294750807e-05,
+      "loss": 9.5497,
+      "step": 51400
+    },
+    {
+      "epoch": 0.25673549901370823,
+      "grad_norm": 0.09745459258556366,
+      "learning_rate": 2.2353751032566525e-05,
+      "loss": 9.5479,
+      "step": 51410
+    },
+    {
+      "epoch": 0.25678543783864766,
+      "grad_norm": 0.09394165873527527,
+      "learning_rate": 2.235224911762497e-05,
+      "loss": 9.5428,
+      "step": 51420
+    },
+    {
+      "epoch": 0.2568353766635871,
+      "grad_norm": 0.09442657977342606,
+      "learning_rate": 2.2350747202683422e-05,
+      "loss": 9.5507,
+      "step": 51430
+    },
+    {
+      "epoch": 0.25688531548852656,
+      "grad_norm": 0.10503870993852615,
+      "learning_rate": 2.2349245287741872e-05,
+      "loss": 9.5499,
+      "step": 51440
+    },
+    {
+      "epoch": 0.256935254313466,
+      "grad_norm": 0.09587554633617401,
+      "learning_rate": 2.234774337280032e-05,
+      "loss": 9.5433,
+      "step": 51450
+    },
+    {
+      "epoch": 0.25698519313840545,
+      "grad_norm": 0.09741827845573425,
+      "learning_rate": 2.2346241457858772e-05,
+      "loss": 9.5378,
+      "step": 51460
+    },
+    {
+      "epoch": 0.2570351319633449,
+      "grad_norm": 0.09930811822414398,
+      "learning_rate": 2.234473954291722e-05,
+      "loss": 9.5404,
+      "step": 51470
+    },
+    {
+      "epoch": 0.25708507078828435,
+      "grad_norm": 0.09290921688079834,
+      "learning_rate": 2.2343237627975673e-05,
+      "loss": 9.5358,
+      "step": 51480
+    },
+    {
+      "epoch": 0.2571350096132238,
+      "grad_norm": 0.09306775778532028,
+      "learning_rate": 2.234173571303412e-05,
+      "loss": 9.5465,
+      "step": 51490
+    },
+    {
+      "epoch": 0.25718494843816325,
+      "grad_norm": 0.09969160705804825,
+      "learning_rate": 2.2340233798092566e-05,
+      "loss": 9.5441,
+      "step": 51500
+    },
+    {
+      "epoch": 0.2572348872631027,
+      "grad_norm": 0.0934043824672699,
+      "learning_rate": 2.233873188315102e-05,
+      "loss": 9.5415,
+      "step": 51510
+    },
+    {
+      "epoch": 0.25728482608804215,
+      "grad_norm": 0.10059630125761032,
+      "learning_rate": 2.2337229968209467e-05,
+      "loss": 9.5438,
+      "step": 51520
+    },
+    {
+      "epoch": 0.2573347649129816,
+      "grad_norm": 0.09175658226013184,
+      "learning_rate": 2.233572805326792e-05,
+      "loss": 9.5439,
+      "step": 51530
+    },
+    {
+      "epoch": 0.25738470373792105,
+      "grad_norm": 0.09227610379457474,
+      "learning_rate": 2.2334226138326367e-05,
+      "loss": 9.545,
+      "step": 51540
+    },
+    {
+      "epoch": 0.25743464256286047,
+      "grad_norm": 0.09750757366418839,
+      "learning_rate": 2.2332724223384814e-05,
+      "loss": 9.5387,
+      "step": 51550
+    },
+    {
+      "epoch": 0.25748458138779995,
+      "grad_norm": 0.09199635684490204,
+      "learning_rate": 2.2331222308443267e-05,
+      "loss": 9.5451,
+      "step": 51560
+    },
+    {
+      "epoch": 0.25753452021273937,
+      "grad_norm": 0.09396953880786896,
+      "learning_rate": 2.2329720393501714e-05,
+      "loss": 9.5419,
+      "step": 51570
+    },
+    {
+      "epoch": 0.25758445903767885,
+      "grad_norm": 0.0948280543088913,
+      "learning_rate": 2.2328218478560168e-05,
+      "loss": 9.543,
+      "step": 51580
+    },
+    {
+      "epoch": 0.25763439786261827,
+      "grad_norm": 0.09518975019454956,
+      "learning_rate": 2.2326716563618615e-05,
+      "loss": 9.5468,
+      "step": 51590
+    },
+    {
+      "epoch": 0.25768433668755775,
+      "grad_norm": 0.0924508273601532,
+      "learning_rate": 2.232521464867706e-05,
+      "loss": 9.5395,
+      "step": 51600
+    },
+    {
+      "epoch": 0.25773427551249717,
+      "grad_norm": 0.09529732912778854,
+      "learning_rate": 2.2323712733735515e-05,
+      "loss": 9.54,
+      "step": 51610
+    },
+    {
+      "epoch": 0.25778421433743665,
+      "grad_norm": 0.09561322629451752,
+      "learning_rate": 2.232221081879396e-05,
+      "loss": 9.5452,
+      "step": 51620
+    },
+    {
+      "epoch": 0.25783415316237607,
+      "grad_norm": 0.0942932739853859,
+      "learning_rate": 2.2320708903852415e-05,
+      "loss": 9.5309,
+      "step": 51630
+    },
+    {
+      "epoch": 0.25788409198731554,
+      "grad_norm": 0.09694642573595047,
+      "learning_rate": 2.2319206988910862e-05,
+      "loss": 9.5398,
+      "step": 51640
+    },
+    {
+      "epoch": 0.25793403081225497,
+      "grad_norm": 0.09523008018732071,
+      "learning_rate": 2.231770507396931e-05,
+      "loss": 9.536,
+      "step": 51650
+    },
+    {
+      "epoch": 0.25798396963719444,
+      "grad_norm": 0.09009785205125809,
+      "learning_rate": 2.2316203159027762e-05,
+      "loss": 9.5383,
+      "step": 51660
+    },
+    {
+      "epoch": 0.25803390846213387,
+      "grad_norm": 0.09198308736085892,
+      "learning_rate": 2.231470124408621e-05,
+      "loss": 9.5422,
+      "step": 51670
+    },
+    {
+      "epoch": 0.25808384728707334,
+      "grad_norm": 0.0955272763967514,
+      "learning_rate": 2.2313199329144663e-05,
+      "loss": 9.5352,
+      "step": 51680
+    },
+    {
+      "epoch": 0.25813378611201276,
+      "grad_norm": 0.0900355726480484,
+      "learning_rate": 2.231169741420311e-05,
+      "loss": 9.5293,
+      "step": 51690
+    },
+    {
+      "epoch": 0.25818372493695224,
+      "grad_norm": 0.0917731374502182,
+      "learning_rate": 2.2310195499261556e-05,
+      "loss": 9.5424,
+      "step": 51700
+    },
+    {
+      "epoch": 0.25823366376189166,
+      "grad_norm": 0.09225501120090485,
+      "learning_rate": 2.230869358432001e-05,
+      "loss": 9.5442,
+      "step": 51710
+    },
+    {
+      "epoch": 0.25828360258683114,
+      "grad_norm": 0.09712720662355423,
+      "learning_rate": 2.2307191669378457e-05,
+      "loss": 9.5388,
+      "step": 51720
+    },
+    {
+      "epoch": 0.25833354141177056,
+      "grad_norm": 0.09137700498104095,
+      "learning_rate": 2.230568975443691e-05,
+      "loss": 9.5402,
+      "step": 51730
+    },
+    {
+      "epoch": 0.25838348023671004,
+      "grad_norm": 0.09159164875745773,
+      "learning_rate": 2.2304187839495357e-05,
+      "loss": 9.5458,
+      "step": 51740
+    },
+    {
+      "epoch": 0.25843341906164946,
+      "grad_norm": 0.09769763052463531,
+      "learning_rate": 2.2302685924553804e-05,
+      "loss": 9.5476,
+      "step": 51750
+    },
+    {
+      "epoch": 0.25848335788658894,
+      "grad_norm": 0.09766492992639542,
+      "learning_rate": 2.2301184009612257e-05,
+      "loss": 9.5393,
+      "step": 51760
+    },
+    {
+      "epoch": 0.25853329671152836,
+      "grad_norm": 0.09677687287330627,
+      "learning_rate": 2.2299682094670704e-05,
+      "loss": 9.534,
+      "step": 51770
+    },
+    {
+      "epoch": 0.25858323553646784,
+      "grad_norm": 0.09470915794372559,
+      "learning_rate": 2.2298180179729158e-05,
+      "loss": 9.5395,
+      "step": 51780
+    },
+    {
+      "epoch": 0.25863317436140726,
+      "grad_norm": 0.08983705192804337,
+      "learning_rate": 2.2296678264787605e-05,
+      "loss": 9.5329,
+      "step": 51790
+    },
+    {
+      "epoch": 0.25868311318634674,
+      "grad_norm": 0.10104523599147797,
+      "learning_rate": 2.2295176349846055e-05,
+      "loss": 9.5324,
+      "step": 51800
+    },
+    {
+      "epoch": 0.25873305201128616,
+      "grad_norm": 0.09755957871675491,
+      "learning_rate": 2.2293674434904505e-05,
+      "loss": 9.5345,
+      "step": 51810
+    },
+    {
+      "epoch": 0.25878299083622563,
+      "grad_norm": 0.09510654956102371,
+      "learning_rate": 2.229217251996295e-05,
+      "loss": 9.5396,
+      "step": 51820
+    },
+    {
+      "epoch": 0.25883292966116506,
+      "grad_norm": 0.10025627911090851,
+      "learning_rate": 2.2290670605021405e-05,
+      "loss": 9.541,
+      "step": 51830
+    },
+    {
+      "epoch": 0.25888286848610453,
+      "grad_norm": 0.0947519913315773,
+      "learning_rate": 2.2289168690079852e-05,
+      "loss": 9.5382,
+      "step": 51840
+    },
+    {
+      "epoch": 0.25893280731104396,
+      "grad_norm": 0.09344780445098877,
+      "learning_rate": 2.2287666775138302e-05,
+      "loss": 9.5367,
+      "step": 51850
+    },
+    {
+      "epoch": 0.25898274613598343,
+      "grad_norm": 0.09429247677326202,
+      "learning_rate": 2.2286164860196752e-05,
+      "loss": 9.535,
+      "step": 51860
+    },
+    {
+      "epoch": 0.25903268496092285,
+      "grad_norm": 0.09451945871114731,
+      "learning_rate": 2.22846629452552e-05,
+      "loss": 9.5271,
+      "step": 51870
+    },
+    {
+      "epoch": 0.25908262378586233,
+      "grad_norm": 0.09508265554904938,
+      "learning_rate": 2.2283161030313653e-05,
+      "loss": 9.5307,
+      "step": 51880
+    },
+    {
+      "epoch": 0.25913256261080175,
+      "grad_norm": 0.09736274182796478,
+      "learning_rate": 2.22816591153721e-05,
+      "loss": 9.5394,
+      "step": 51890
+    },
+    {
+      "epoch": 0.25918250143574123,
+      "grad_norm": 0.09334583580493927,
+      "learning_rate": 2.228015720043055e-05,
+      "loss": 9.5279,
+      "step": 51900
+    },
+    {
+      "epoch": 0.25923244026068065,
+      "grad_norm": 0.0936061292886734,
+      "learning_rate": 2.2278655285489e-05,
+      "loss": 9.5434,
+      "step": 51910
+    },
+    {
+      "epoch": 0.25928237908562013,
+      "grad_norm": 0.09247171133756638,
+      "learning_rate": 2.2277153370547447e-05,
+      "loss": 9.5407,
+      "step": 51920
+    },
+    {
+      "epoch": 0.25933231791055955,
+      "grad_norm": 0.09118923544883728,
+      "learning_rate": 2.22756514556059e-05,
+      "loss": 9.533,
+      "step": 51930
+    },
+    {
+      "epoch": 0.25938225673549903,
+      "grad_norm": 0.09702643752098083,
+      "learning_rate": 2.2274149540664347e-05,
+      "loss": 9.5396,
+      "step": 51940
+    },
+    {
+      "epoch": 0.25943219556043845,
+      "grad_norm": 0.0896030142903328,
+      "learning_rate": 2.2272647625722797e-05,
+      "loss": 9.5384,
+      "step": 51950
+    },
+    {
+      "epoch": 0.2594821343853779,
+      "grad_norm": 0.09771367162466049,
+      "learning_rate": 2.2271145710781247e-05,
+      "loss": 9.531,
+      "step": 51960
+    },
+    {
+      "epoch": 0.25953207321031735,
+      "grad_norm": 0.0948578342795372,
+      "learning_rate": 2.2269643795839694e-05,
+      "loss": 9.5386,
+      "step": 51970
+    },
+    {
+      "epoch": 0.2595820120352568,
+      "grad_norm": 0.10098248720169067,
+      "learning_rate": 2.2268141880898148e-05,
+      "loss": 9.5335,
+      "step": 51980
+    },
+    {
+      "epoch": 0.25963195086019625,
+      "grad_norm": 0.08970336616039276,
+      "learning_rate": 2.2266639965956595e-05,
+      "loss": 9.5386,
+      "step": 51990
+    },
+    {
+      "epoch": 0.2596818896851357,
+      "grad_norm": 0.09408001601696014,
+      "learning_rate": 2.2265138051015045e-05,
+      "loss": 9.5355,
+      "step": 52000
+    },
+    {
+      "epoch": 0.25973182851007515,
+      "grad_norm": 0.08999434113502502,
+      "learning_rate": 2.2263636136073495e-05,
+      "loss": 9.5373,
+      "step": 52010
+    },
+    {
+      "epoch": 0.2597817673350146,
+      "grad_norm": 0.09277333319187164,
+      "learning_rate": 2.226213422113194e-05,
+      "loss": 9.5405,
+      "step": 52020
+    },
+    {
+      "epoch": 0.25983170615995405,
+      "grad_norm": 0.09205369651317596,
+      "learning_rate": 2.2260632306190395e-05,
+      "loss": 9.5254,
+      "step": 52030
+    },
+    {
+      "epoch": 0.2598816449848935,
+      "grad_norm": 0.09553728252649307,
+      "learning_rate": 2.2259130391248842e-05,
+      "loss": 9.5262,
+      "step": 52040
+    },
+    {
+      "epoch": 0.25993158380983294,
+      "grad_norm": 0.08813092112541199,
+      "learning_rate": 2.2257628476307292e-05,
+      "loss": 9.536,
+      "step": 52050
+    },
+    {
+      "epoch": 0.2599815226347724,
+      "grad_norm": 0.09862591326236725,
+      "learning_rate": 2.2256126561365742e-05,
+      "loss": 9.537,
+      "step": 52060
+    },
+    {
+      "epoch": 0.26003146145971184,
+      "grad_norm": 0.09814270585775375,
+      "learning_rate": 2.225462464642419e-05,
+      "loss": 9.535,
+      "step": 52070
+    },
+    {
+      "epoch": 0.2600814002846513,
+      "grad_norm": 0.08804214745759964,
+      "learning_rate": 2.2253122731482643e-05,
+      "loss": 9.5421,
+      "step": 52080
+    },
+    {
+      "epoch": 0.26013133910959074,
+      "grad_norm": 0.09738943725824356,
+      "learning_rate": 2.225162081654109e-05,
+      "loss": 9.5307,
+      "step": 52090
+    },
+    {
+      "epoch": 0.2601812779345302,
+      "grad_norm": 0.09532274305820465,
+      "learning_rate": 2.225011890159954e-05,
+      "loss": 9.541,
+      "step": 52100
+    },
+    {
+      "epoch": 0.26023121675946964,
+      "grad_norm": 0.09409879148006439,
+      "learning_rate": 2.224861698665799e-05,
+      "loss": 9.5364,
+      "step": 52110
+    },
+    {
+      "epoch": 0.2602811555844091,
+      "grad_norm": 0.09654781222343445,
+      "learning_rate": 2.224711507171644e-05,
+      "loss": 9.5443,
+      "step": 52120
+    },
+    {
+      "epoch": 0.26033109440934854,
+      "grad_norm": 0.09525051712989807,
+      "learning_rate": 2.224561315677489e-05,
+      "loss": 9.5404,
+      "step": 52130
+    },
+    {
+      "epoch": 0.260381033234288,
+      "grad_norm": 0.09068022668361664,
+      "learning_rate": 2.2244111241833337e-05,
+      "loss": 9.529,
+      "step": 52140
+    },
+    {
+      "epoch": 0.26043097205922744,
+      "grad_norm": 0.09324584156274796,
+      "learning_rate": 2.2242609326891787e-05,
+      "loss": 9.5402,
+      "step": 52150
+    },
+    {
+      "epoch": 0.2604809108841669,
+      "grad_norm": 0.0990164652466774,
+      "learning_rate": 2.2241107411950237e-05,
+      "loss": 9.5354,
+      "step": 52160
+    },
+    {
+      "epoch": 0.26053084970910634,
+      "grad_norm": 0.09122282266616821,
+      "learning_rate": 2.2239605497008688e-05,
+      "loss": 9.538,
+      "step": 52170
+    },
+    {
+      "epoch": 0.2605807885340458,
+      "grad_norm": 0.09253876656293869,
+      "learning_rate": 2.2238103582067138e-05,
+      "loss": 9.5287,
+      "step": 52180
+    },
+    {
+      "epoch": 0.26063072735898524,
+      "grad_norm": 0.09256315976381302,
+      "learning_rate": 2.2236601667125585e-05,
+      "loss": 9.5346,
+      "step": 52190
+    },
+    {
+      "epoch": 0.2606806661839247,
+      "grad_norm": 0.09512501209974289,
+      "learning_rate": 2.2235099752184035e-05,
+      "loss": 9.533,
+      "step": 52200
+    },
+    {
+      "epoch": 0.26073060500886414,
+      "grad_norm": 0.09339733421802521,
+      "learning_rate": 2.2233597837242485e-05,
+      "loss": 9.5352,
+      "step": 52210
+    },
+    {
+      "epoch": 0.2607805438338036,
+      "grad_norm": 0.10032928735017776,
+      "learning_rate": 2.2232095922300935e-05,
+      "loss": 9.5266,
+      "step": 52220
+    },
+    {
+      "epoch": 0.26083048265874303,
+      "grad_norm": 0.09287399053573608,
+      "learning_rate": 2.2230594007359385e-05,
+      "loss": 9.5329,
+      "step": 52230
+    },
+    {
+      "epoch": 0.2608804214836825,
+      "grad_norm": 0.09904525429010391,
+      "learning_rate": 2.2229092092417832e-05,
+      "loss": 9.534,
+      "step": 52240
+    },
+    {
+      "epoch": 0.26093036030862193,
+      "grad_norm": 0.10111285001039505,
+      "learning_rate": 2.2227590177476282e-05,
+      "loss": 9.5349,
+      "step": 52250
+    },
+    {
+      "epoch": 0.2609802991335614,
+      "grad_norm": 0.10239645838737488,
+      "learning_rate": 2.2226088262534732e-05,
+      "loss": 9.5331,
+      "step": 52260
+    },
+    {
+      "epoch": 0.26103023795850083,
+      "grad_norm": 0.09370391070842743,
+      "learning_rate": 2.2224586347593183e-05,
+      "loss": 9.5251,
+      "step": 52270
+    },
+    {
+      "epoch": 0.2610801767834403,
+      "grad_norm": 0.09311920404434204,
+      "learning_rate": 2.2223084432651633e-05,
+      "loss": 9.5298,
+      "step": 52280
+    },
+    {
+      "epoch": 0.26113011560837973,
+      "grad_norm": 0.0964302271604538,
+      "learning_rate": 2.222158251771008e-05,
+      "loss": 9.5224,
+      "step": 52290
+    },
+    {
+      "epoch": 0.2611800544333192,
+      "grad_norm": 0.09276770055294037,
+      "learning_rate": 2.222008060276853e-05,
+      "loss": 9.5257,
+      "step": 52300
+    },
+    {
+      "epoch": 0.26122999325825863,
+      "grad_norm": 0.09495028853416443,
+      "learning_rate": 2.221857868782698e-05,
+      "loss": 9.5282,
+      "step": 52310
+    },
+    {
+      "epoch": 0.2612799320831981,
+      "grad_norm": 0.09358720481395721,
+      "learning_rate": 2.221707677288543e-05,
+      "loss": 9.5337,
+      "step": 52320
+    },
+    {
+      "epoch": 0.26132987090813753,
+      "grad_norm": 0.09454384446144104,
+      "learning_rate": 2.221557485794388e-05,
+      "loss": 9.5262,
+      "step": 52330
+    },
+    {
+      "epoch": 0.261379809733077,
+      "grad_norm": 0.09213569760322571,
+      "learning_rate": 2.2214072943002327e-05,
+      "loss": 9.5294,
+      "step": 52340
+    },
+    {
+      "epoch": 0.26142974855801643,
+      "grad_norm": 0.08967851847410202,
+      "learning_rate": 2.2212571028060777e-05,
+      "loss": 9.5267,
+      "step": 52350
+    },
+    {
+      "epoch": 0.2614796873829559,
+      "grad_norm": 0.09236017614603043,
+      "learning_rate": 2.2211069113119227e-05,
+      "loss": 9.5257,
+      "step": 52360
+    },
+    {
+      "epoch": 0.2615296262078953,
+      "grad_norm": 0.09722200036048889,
+      "learning_rate": 2.2209567198177678e-05,
+      "loss": 9.526,
+      "step": 52370
+    },
+    {
+      "epoch": 0.2615795650328348,
+      "grad_norm": 0.09535818547010422,
+      "learning_rate": 2.2208065283236128e-05,
+      "loss": 9.5367,
+      "step": 52380
+    },
+    {
+      "epoch": 0.2616295038577742,
+      "grad_norm": 0.09433511644601822,
+      "learning_rate": 2.2206563368294575e-05,
+      "loss": 9.5386,
+      "step": 52390
+    },
+    {
+      "epoch": 0.2616794426827137,
+      "grad_norm": 0.09581523388624191,
+      "learning_rate": 2.2205061453353025e-05,
+      "loss": 9.5319,
+      "step": 52400
+    },
+    {
+      "epoch": 0.2617293815076531,
+      "grad_norm": 0.0966779887676239,
+      "learning_rate": 2.2203559538411475e-05,
+      "loss": 9.5217,
+      "step": 52410
+    },
+    {
+      "epoch": 0.26177932033259255,
+      "grad_norm": 0.09320307523012161,
+      "learning_rate": 2.2202057623469925e-05,
+      "loss": 9.5237,
+      "step": 52420
+    },
+    {
+      "epoch": 0.261829259157532,
+      "grad_norm": 0.09114668518304825,
+      "learning_rate": 2.2200555708528375e-05,
+      "loss": 9.53,
+      "step": 52430
+    },
+    {
+      "epoch": 0.26187919798247145,
+      "grad_norm": 0.09709704667329788,
+      "learning_rate": 2.2199053793586825e-05,
+      "loss": 9.5339,
+      "step": 52440
+    },
+    {
+      "epoch": 0.2619291368074109,
+      "grad_norm": 0.09141276776790619,
+      "learning_rate": 2.2197551878645272e-05,
+      "loss": 9.52,
+      "step": 52450
+    },
+    {
+      "epoch": 0.26197907563235034,
+      "grad_norm": 0.09864918142557144,
+      "learning_rate": 2.2196049963703722e-05,
+      "loss": 9.525,
+      "step": 52460
+    },
+    {
+      "epoch": 0.2620290144572898,
+      "grad_norm": 0.09360401332378387,
+      "learning_rate": 2.2194548048762173e-05,
+      "loss": 9.5368,
+      "step": 52470
+    },
+    {
+      "epoch": 0.26207895328222924,
+      "grad_norm": 0.09955708682537079,
+      "learning_rate": 2.2193046133820623e-05,
+      "loss": 9.5265,
+      "step": 52480
+    },
+    {
+      "epoch": 0.2621288921071687,
+      "grad_norm": 0.09114803373813629,
+      "learning_rate": 2.2191544218879073e-05,
+      "loss": 9.5272,
+      "step": 52490
+    },
+    {
+      "epoch": 0.26217883093210814,
+      "grad_norm": 0.09551452845335007,
+      "learning_rate": 2.219004230393752e-05,
+      "loss": 9.5356,
+      "step": 52500
+    },
+    {
+      "epoch": 0.2622287697570476,
+      "grad_norm": 0.09264279901981354,
+      "learning_rate": 2.218854038899597e-05,
+      "loss": 9.5254,
+      "step": 52510
+    },
+    {
+      "epoch": 0.26227870858198704,
+      "grad_norm": 0.09141562134027481,
+      "learning_rate": 2.218703847405442e-05,
+      "loss": 9.5245,
+      "step": 52520
+    },
+    {
+      "epoch": 0.2623286474069265,
+      "grad_norm": 0.0920877605676651,
+      "learning_rate": 2.218553655911287e-05,
+      "loss": 9.5334,
+      "step": 52530
+    },
+    {
+      "epoch": 0.26237858623186594,
+      "grad_norm": 0.09428731352090836,
+      "learning_rate": 2.218403464417132e-05,
+      "loss": 9.5331,
+      "step": 52540
+    },
+    {
+      "epoch": 0.2624285250568054,
+      "grad_norm": 0.09918078780174255,
+      "learning_rate": 2.2182532729229767e-05,
+      "loss": 9.515,
+      "step": 52550
+    },
+    {
+      "epoch": 0.26247846388174484,
+      "grad_norm": 0.08825034648180008,
+      "learning_rate": 2.2181030814288217e-05,
+      "loss": 9.5283,
+      "step": 52560
+    },
+    {
+      "epoch": 0.2625284027066843,
+      "grad_norm": 0.09419302642345428,
+      "learning_rate": 2.2179528899346668e-05,
+      "loss": 9.5242,
+      "step": 52570
+    },
+    {
+      "epoch": 0.26257834153162374,
+      "grad_norm": 0.09222215414047241,
+      "learning_rate": 2.2178026984405118e-05,
+      "loss": 9.5165,
+      "step": 52580
+    },
+    {
+      "epoch": 0.2626282803565632,
+      "grad_norm": 0.0944628119468689,
+      "learning_rate": 2.2176525069463568e-05,
+      "loss": 9.519,
+      "step": 52590
+    },
+    {
+      "epoch": 0.26267821918150264,
+      "grad_norm": 0.09280452877283096,
+      "learning_rate": 2.2175023154522015e-05,
+      "loss": 9.5191,
+      "step": 52600
+    },
+    {
+      "epoch": 0.2627281580064421,
+      "grad_norm": 0.09857330471277237,
+      "learning_rate": 2.2173521239580465e-05,
+      "loss": 9.5195,
+      "step": 52610
+    },
+    {
+      "epoch": 0.26277809683138154,
+      "grad_norm": 0.10611597448587418,
+      "learning_rate": 2.2172019324638915e-05,
+      "loss": 9.5201,
+      "step": 52620
+    },
+    {
+      "epoch": 0.262828035656321,
+      "grad_norm": 0.09467349946498871,
+      "learning_rate": 2.2170517409697365e-05,
+      "loss": 9.5355,
+      "step": 52630
+    },
+    {
+      "epoch": 0.26287797448126043,
+      "grad_norm": 0.0893920361995697,
+      "learning_rate": 2.2169015494755815e-05,
+      "loss": 9.5229,
+      "step": 52640
+    },
+    {
+      "epoch": 0.2629279133061999,
+      "grad_norm": 0.09524576365947723,
+      "learning_rate": 2.2167513579814262e-05,
+      "loss": 9.5212,
+      "step": 52650
+    },
+    {
+      "epoch": 0.26297785213113933,
+      "grad_norm": 0.09462094306945801,
+      "learning_rate": 2.2166011664872712e-05,
+      "loss": 9.5331,
+      "step": 52660
+    },
+    {
+      "epoch": 0.2630277909560788,
+      "grad_norm": 0.1007109135389328,
+      "learning_rate": 2.2164509749931163e-05,
+      "loss": 9.5255,
+      "step": 52670
+    },
+    {
+      "epoch": 0.26307772978101823,
+      "grad_norm": 0.09446932375431061,
+      "learning_rate": 2.2163007834989613e-05,
+      "loss": 9.5272,
+      "step": 52680
+    },
+    {
+      "epoch": 0.2631276686059577,
+      "grad_norm": 0.08853420615196228,
+      "learning_rate": 2.2161505920048063e-05,
+      "loss": 9.5298,
+      "step": 52690
+    },
+    {
+      "epoch": 0.26317760743089713,
+      "grad_norm": 0.09556553512811661,
+      "learning_rate": 2.216000400510651e-05,
+      "loss": 9.5255,
+      "step": 52700
+    },
+    {
+      "epoch": 0.2632275462558366,
+      "grad_norm": 0.09378375113010406,
+      "learning_rate": 2.215850209016496e-05,
+      "loss": 9.5139,
+      "step": 52710
+    },
+    {
+      "epoch": 0.26327748508077603,
+      "grad_norm": 0.09631894528865814,
+      "learning_rate": 2.215700017522341e-05,
+      "loss": 9.5168,
+      "step": 52720
+    },
+    {
+      "epoch": 0.2633274239057155,
+      "grad_norm": 0.09347805380821228,
+      "learning_rate": 2.215549826028186e-05,
+      "loss": 9.5312,
+      "step": 52730
+    },
+    {
+      "epoch": 0.26337736273065493,
+      "grad_norm": 0.09734681993722916,
+      "learning_rate": 2.215399634534031e-05,
+      "loss": 9.5144,
+      "step": 52740
+    },
+    {
+      "epoch": 0.2634273015555944,
+      "grad_norm": 0.08855658769607544,
+      "learning_rate": 2.2152494430398757e-05,
+      "loss": 9.5155,
+      "step": 52750
+    },
+    {
+      "epoch": 0.26347724038053383,
+      "grad_norm": 0.09233860671520233,
+      "learning_rate": 2.215099251545721e-05,
+      "loss": 9.5132,
+      "step": 52760
+    },
+    {
+      "epoch": 0.2635271792054733,
+      "grad_norm": 0.0949973613023758,
+      "learning_rate": 2.2149490600515658e-05,
+      "loss": 9.5156,
+      "step": 52770
+    },
+    {
+      "epoch": 0.2635771180304127,
+      "grad_norm": 0.10060823708772659,
+      "learning_rate": 2.2147988685574108e-05,
+      "loss": 9.5138,
+      "step": 52780
+    },
+    {
+      "epoch": 0.2636270568553522,
+      "grad_norm": 0.09591439366340637,
+      "learning_rate": 2.2146486770632558e-05,
+      "loss": 9.5292,
+      "step": 52790
+    },
+    {
+      "epoch": 0.2636769956802916,
+      "grad_norm": 0.09454584121704102,
+      "learning_rate": 2.2144984855691005e-05,
+      "loss": 9.5177,
+      "step": 52800
+    },
+    {
+      "epoch": 0.2637269345052311,
+      "grad_norm": 0.09631575644016266,
+      "learning_rate": 2.214348294074946e-05,
+      "loss": 9.5153,
+      "step": 52810
+    },
+    {
+      "epoch": 0.2637768733301705,
+      "grad_norm": 0.09541364014148712,
+      "learning_rate": 2.2141981025807905e-05,
+      "loss": 9.5259,
+      "step": 52820
+    },
+    {
+      "epoch": 0.26382681215511,
+      "grad_norm": 0.09758414328098297,
+      "learning_rate": 2.2140479110866355e-05,
+      "loss": 9.5273,
+      "step": 52830
+    },
+    {
+      "epoch": 0.2638767509800494,
+      "grad_norm": 0.09002380818128586,
+      "learning_rate": 2.2138977195924805e-05,
+      "loss": 9.5144,
+      "step": 52840
+    },
+    {
+      "epoch": 0.2639266898049889,
+      "grad_norm": 0.09718296676874161,
+      "learning_rate": 2.2137475280983252e-05,
+      "loss": 9.515,
+      "step": 52850
+    },
+    {
+      "epoch": 0.2639766286299283,
+      "grad_norm": 0.09190072864294052,
+      "learning_rate": 2.2135973366041706e-05,
+      "loss": 9.5239,
+      "step": 52860
+    },
+    {
+      "epoch": 0.2640265674548678,
+      "grad_norm": 0.0970335304737091,
+      "learning_rate": 2.2134471451100153e-05,
+      "loss": 9.5219,
+      "step": 52870
+    },
+    {
+      "epoch": 0.2640765062798072,
+      "grad_norm": 0.09237813204526901,
+      "learning_rate": 2.2132969536158603e-05,
+      "loss": 9.5324,
+      "step": 52880
+    },
+    {
+      "epoch": 0.2641264451047467,
+      "grad_norm": 0.09487534314393997,
+      "learning_rate": 2.2131467621217053e-05,
+      "loss": 9.5275,
+      "step": 52890
+    },
+    {
+      "epoch": 0.2641763839296861,
+      "grad_norm": 0.09346019476652145,
+      "learning_rate": 2.21299657062755e-05,
+      "loss": 9.521,
+      "step": 52900
+    },
+    {
+      "epoch": 0.2642263227546256,
+      "grad_norm": 0.09257887303829193,
+      "learning_rate": 2.2128463791333953e-05,
+      "loss": 9.5117,
+      "step": 52910
+    },
+    {
+      "epoch": 0.264276261579565,
+      "grad_norm": 0.09638823568820953,
+      "learning_rate": 2.21269618763924e-05,
+      "loss": 9.5236,
+      "step": 52920
+    },
+    {
+      "epoch": 0.2643262004045045,
+      "grad_norm": 0.09802195429801941,
+      "learning_rate": 2.212545996145085e-05,
+      "loss": 9.5187,
+      "step": 52930
+    },
+    {
+      "epoch": 0.2643761392294439,
+      "grad_norm": 0.09367150813341141,
+      "learning_rate": 2.21239580465093e-05,
+      "loss": 9.5275,
+      "step": 52940
+    },
+    {
+      "epoch": 0.2644260780543834,
+      "grad_norm": 0.09576943516731262,
+      "learning_rate": 2.2122456131567747e-05,
+      "loss": 9.5152,
+      "step": 52950
+    },
+    {
+      "epoch": 0.2644760168793228,
+      "grad_norm": 0.10221412777900696,
+      "learning_rate": 2.21209542166262e-05,
+      "loss": 9.5211,
+      "step": 52960
+    },
+    {
+      "epoch": 0.2645259557042623,
+      "grad_norm": 0.08980675041675568,
+      "learning_rate": 2.2119452301684648e-05,
+      "loss": 9.5274,
+      "step": 52970
+    },
+    {
+      "epoch": 0.2645758945292017,
+      "grad_norm": 0.0930297002196312,
+      "learning_rate": 2.2117950386743098e-05,
+      "loss": 9.5114,
+      "step": 52980
+    },
+    {
+      "epoch": 0.2646258333541412,
+      "grad_norm": 0.09479690343141556,
+      "learning_rate": 2.2116448471801548e-05,
+      "loss": 9.5105,
+      "step": 52990
+    },
+    {
+      "epoch": 0.2646757721790806,
+      "grad_norm": 0.0916314572095871,
+      "learning_rate": 2.2114946556859995e-05,
+      "loss": 9.5178,
+      "step": 53000
+    },
+    {
+      "epoch": 0.2647257110040201,
+      "grad_norm": 0.09060303866863251,
+      "learning_rate": 2.211344464191845e-05,
+      "loss": 9.5146,
+      "step": 53010
+    },
+    {
+      "epoch": 0.2647756498289595,
+      "grad_norm": 0.09311266988515854,
+      "learning_rate": 2.2111942726976895e-05,
+      "loss": 9.5126,
+      "step": 53020
+    },
+    {
+      "epoch": 0.264825588653899,
+      "grad_norm": 0.09283456206321716,
+      "learning_rate": 2.2110440812035345e-05,
+      "loss": 9.5175,
+      "step": 53030
+    },
+    {
+      "epoch": 0.2648755274788384,
+      "grad_norm": 0.09691060334444046,
+      "learning_rate": 2.2108938897093796e-05,
+      "loss": 9.5163,
+      "step": 53040
+    },
+    {
+      "epoch": 0.2649254663037779,
+      "grad_norm": 0.09444165974855423,
+      "learning_rate": 2.2107436982152242e-05,
+      "loss": 9.5194,
+      "step": 53050
+    },
+    {
+      "epoch": 0.2649754051287173,
+      "grad_norm": 0.0917515829205513,
+      "learning_rate": 2.2105935067210696e-05,
+      "loss": 9.522,
+      "step": 53060
+    },
+    {
+      "epoch": 0.2650253439536568,
+      "grad_norm": 0.093546062707901,
+      "learning_rate": 2.2104433152269143e-05,
+      "loss": 9.5124,
+      "step": 53070
+    },
+    {
+      "epoch": 0.2650752827785962,
+      "grad_norm": 0.09106338769197464,
+      "learning_rate": 2.2102931237327593e-05,
+      "loss": 9.5194,
+      "step": 53080
+    },
+    {
+      "epoch": 0.2651252216035357,
+      "grad_norm": 0.09387896209955215,
+      "learning_rate": 2.2101429322386043e-05,
+      "loss": 9.5174,
+      "step": 53090
+    },
+    {
+      "epoch": 0.2651751604284751,
+      "grad_norm": 0.09550568461418152,
+      "learning_rate": 2.209992740744449e-05,
+      "loss": 9.5169,
+      "step": 53100
+    },
+    {
+      "epoch": 0.2652250992534146,
+      "grad_norm": 0.09244140982627869,
+      "learning_rate": 2.2098425492502943e-05,
+      "loss": 9.513,
+      "step": 53110
+    },
+    {
+      "epoch": 0.265275038078354,
+      "grad_norm": 0.09430118650197983,
+      "learning_rate": 2.209692357756139e-05,
+      "loss": 9.5206,
+      "step": 53120
+    },
+    {
+      "epoch": 0.2653249769032935,
+      "grad_norm": 0.10021936148405075,
+      "learning_rate": 2.2095421662619844e-05,
+      "loss": 9.5203,
+      "step": 53130
+    },
+    {
+      "epoch": 0.2653749157282329,
+      "grad_norm": 0.09453583508729935,
+      "learning_rate": 2.209391974767829e-05,
+      "loss": 9.5158,
+      "step": 53140
+    },
+    {
+      "epoch": 0.2654248545531724,
+      "grad_norm": 0.09471450746059418,
+      "learning_rate": 2.2092417832736737e-05,
+      "loss": 9.5179,
+      "step": 53150
+    },
+    {
+      "epoch": 0.2654747933781118,
+      "grad_norm": 0.0972927138209343,
+      "learning_rate": 2.209091591779519e-05,
+      "loss": 9.5145,
+      "step": 53160
+    },
+    {
+      "epoch": 0.2655247322030513,
+      "grad_norm": 0.09311800450086594,
+      "learning_rate": 2.2089414002853638e-05,
+      "loss": 9.5106,
+      "step": 53170
+    },
+    {
+      "epoch": 0.2655746710279907,
+      "grad_norm": 0.09145248681306839,
+      "learning_rate": 2.208791208791209e-05,
+      "loss": 9.5156,
+      "step": 53180
+    },
+    {
+      "epoch": 0.2656246098529302,
+      "grad_norm": 0.09487327933311462,
+      "learning_rate": 2.2086410172970538e-05,
+      "loss": 9.5138,
+      "step": 53190
+    },
+    {
+      "epoch": 0.2656745486778696,
+      "grad_norm": 0.09175003319978714,
+      "learning_rate": 2.2084908258028985e-05,
+      "loss": 9.5251,
+      "step": 53200
+    },
+    {
+      "epoch": 0.2657244875028091,
+      "grad_norm": 0.09299546480178833,
+      "learning_rate": 2.208340634308744e-05,
+      "loss": 9.5187,
+      "step": 53210
+    },
+    {
+      "epoch": 0.2657744263277485,
+      "grad_norm": 0.09463711827993393,
+      "learning_rate": 2.2081904428145885e-05,
+      "loss": 9.5196,
+      "step": 53220
+    },
+    {
+      "epoch": 0.265824365152688,
+      "grad_norm": 0.09696915000677109,
+      "learning_rate": 2.208040251320434e-05,
+      "loss": 9.5094,
+      "step": 53230
+    },
+    {
+      "epoch": 0.2658743039776274,
+      "grad_norm": 0.1047021821141243,
+      "learning_rate": 2.2078900598262786e-05,
+      "loss": 9.5022,
+      "step": 53240
+    },
+    {
+      "epoch": 0.2659242428025669,
+      "grad_norm": 0.09355685114860535,
+      "learning_rate": 2.2077398683321232e-05,
+      "loss": 9.5123,
+      "step": 53250
+    },
+    {
+      "epoch": 0.2659741816275063,
+      "grad_norm": 0.09120495617389679,
+      "learning_rate": 2.2075896768379686e-05,
+      "loss": 9.5105,
+      "step": 53260
+    },
+    {
+      "epoch": 0.2660241204524458,
+      "grad_norm": 0.08929228782653809,
+      "learning_rate": 2.2074394853438133e-05,
+      "loss": 9.5146,
+      "step": 53270
+    },
+    {
+      "epoch": 0.2660740592773852,
+      "grad_norm": 0.09182194620370865,
+      "learning_rate": 2.2072892938496586e-05,
+      "loss": 9.5202,
+      "step": 53280
+    },
+    {
+      "epoch": 0.2661239981023247,
+      "grad_norm": 0.0908316969871521,
+      "learning_rate": 2.2071391023555033e-05,
+      "loss": 9.515,
+      "step": 53290
+    },
+    {
+      "epoch": 0.2661739369272641,
+      "grad_norm": 0.09303124248981476,
+      "learning_rate": 2.206988910861348e-05,
+      "loss": 9.5074,
+      "step": 53300
+    },
+    {
+      "epoch": 0.2662238757522036,
+      "grad_norm": 0.09067052602767944,
+      "learning_rate": 2.2068387193671933e-05,
+      "loss": 9.5062,
+      "step": 53310
+    },
+    {
+      "epoch": 0.266273814577143,
+      "grad_norm": 0.09253434091806412,
+      "learning_rate": 2.206688527873038e-05,
+      "loss": 9.5053,
+      "step": 53320
+    },
+    {
+      "epoch": 0.2663237534020825,
+      "grad_norm": 0.09222152829170227,
+      "learning_rate": 2.2065383363788834e-05,
+      "loss": 9.5116,
+      "step": 53330
+    },
+    {
+      "epoch": 0.2663736922270219,
+      "grad_norm": 0.09831732511520386,
+      "learning_rate": 2.206388144884728e-05,
+      "loss": 9.5086,
+      "step": 53340
+    },
+    {
+      "epoch": 0.2664236310519614,
+      "grad_norm": 0.08993234485387802,
+      "learning_rate": 2.2062379533905727e-05,
+      "loss": 9.5213,
+      "step": 53350
+    },
+    {
+      "epoch": 0.2664735698769008,
+      "grad_norm": 0.09357697516679764,
+      "learning_rate": 2.206087761896418e-05,
+      "loss": 9.5129,
+      "step": 53360
+    },
+    {
+      "epoch": 0.2665235087018403,
+      "grad_norm": 0.09632600098848343,
+      "learning_rate": 2.2059375704022628e-05,
+      "loss": 9.5079,
+      "step": 53370
+    },
+    {
+      "epoch": 0.2665734475267797,
+      "grad_norm": 0.09463771432638168,
+      "learning_rate": 2.205787378908108e-05,
+      "loss": 9.5108,
+      "step": 53380
+    },
+    {
+      "epoch": 0.26662338635171917,
+      "grad_norm": 0.09664889425039291,
+      "learning_rate": 2.2056371874139528e-05,
+      "loss": 9.5136,
+      "step": 53390
+    },
+    {
+      "epoch": 0.2666733251766586,
+      "grad_norm": 0.0904850959777832,
+      "learning_rate": 2.2054869959197975e-05,
+      "loss": 9.5074,
+      "step": 53400
+    },
+    {
+      "epoch": 0.266723264001598,
+      "grad_norm": 0.0935087651014328,
+      "learning_rate": 2.205336804425643e-05,
+      "loss": 9.5163,
+      "step": 53410
+    },
+    {
+      "epoch": 0.2667732028265375,
+      "grad_norm": 0.09283655881881714,
+      "learning_rate": 2.2051866129314875e-05,
+      "loss": 9.517,
+      "step": 53420
+    },
+    {
+      "epoch": 0.2668231416514769,
+      "grad_norm": 0.09471392631530762,
+      "learning_rate": 2.205036421437333e-05,
+      "loss": 9.5124,
+      "step": 53430
+    },
+    {
+      "epoch": 0.2668730804764164,
+      "grad_norm": 0.10181424766778946,
+      "learning_rate": 2.2048862299431776e-05,
+      "loss": 9.4963,
+      "step": 53440
+    },
+    {
+      "epoch": 0.2669230193013558,
+      "grad_norm": 0.09600277245044708,
+      "learning_rate": 2.2047360384490226e-05,
+      "loss": 9.5115,
+      "step": 53450
+    },
+    {
+      "epoch": 0.2669729581262953,
+      "grad_norm": 0.10142569988965988,
+      "learning_rate": 2.2045858469548676e-05,
+      "loss": 9.5058,
+      "step": 53460
+    },
+    {
+      "epoch": 0.2670228969512347,
+      "grad_norm": 0.09453926235437393,
+      "learning_rate": 2.2044356554607123e-05,
+      "loss": 9.5034,
+      "step": 53470
+    },
+    {
+      "epoch": 0.2670728357761742,
+      "grad_norm": 0.09854920208454132,
+      "learning_rate": 2.2042854639665576e-05,
+      "loss": 9.5109,
+      "step": 53480
+    },
+    {
+      "epoch": 0.2671227746011136,
+      "grad_norm": 0.09525490552186966,
+      "learning_rate": 2.2041352724724023e-05,
+      "loss": 9.5093,
+      "step": 53490
+    },
+    {
+      "epoch": 0.2671727134260531,
+      "grad_norm": 0.10124125331640244,
+      "learning_rate": 2.2039850809782473e-05,
+      "loss": 9.5103,
+      "step": 53500
+    },
+    {
+      "epoch": 0.2672226522509925,
+      "grad_norm": 0.09206513315439224,
+      "learning_rate": 2.2038348894840923e-05,
+      "loss": 9.5106,
+      "step": 53510
+    },
+    {
+      "epoch": 0.267272591075932,
+      "grad_norm": 0.09180595725774765,
+      "learning_rate": 2.203684697989937e-05,
+      "loss": 9.5057,
+      "step": 53520
+    },
+    {
+      "epoch": 0.2673225299008714,
+      "grad_norm": 0.09658707678318024,
+      "learning_rate": 2.2035345064957824e-05,
+      "loss": 9.5024,
+      "step": 53530
+    },
+    {
+      "epoch": 0.2673724687258109,
+      "grad_norm": 0.08750133216381073,
+      "learning_rate": 2.203384315001627e-05,
+      "loss": 9.5091,
+      "step": 53540
+    },
+    {
+      "epoch": 0.2674224075507503,
+      "grad_norm": 0.09664824604988098,
+      "learning_rate": 2.203234123507472e-05,
+      "loss": 9.5102,
+      "step": 53550
+    },
+    {
+      "epoch": 0.2674723463756898,
+      "grad_norm": 0.09974256902933121,
+      "learning_rate": 2.203083932013317e-05,
+      "loss": 9.5205,
+      "step": 53560
+    },
+    {
+      "epoch": 0.2675222852006292,
+      "grad_norm": 0.08979537338018417,
+      "learning_rate": 2.2029337405191618e-05,
+      "loss": 9.5085,
+      "step": 53570
+    },
+    {
+      "epoch": 0.2675722240255687,
+      "grad_norm": 0.09149536490440369,
+      "learning_rate": 2.202783549025007e-05,
+      "loss": 9.5113,
+      "step": 53580
+    },
+    {
+      "epoch": 0.2676221628505081,
+      "grad_norm": 0.09277680516242981,
+      "learning_rate": 2.2026333575308518e-05,
+      "loss": 9.5072,
+      "step": 53590
+    },
+    {
+      "epoch": 0.2676721016754476,
+      "grad_norm": 0.09466489404439926,
+      "learning_rate": 2.2024831660366968e-05,
+      "loss": 9.5135,
+      "step": 53600
+    },
+    {
+      "epoch": 0.267722040500387,
+      "grad_norm": 0.10129041969776154,
+      "learning_rate": 2.202332974542542e-05,
+      "loss": 9.5097,
+      "step": 53610
+    },
+    {
+      "epoch": 0.2677719793253265,
+      "grad_norm": 0.09101355820894241,
+      "learning_rate": 2.2021827830483865e-05,
+      "loss": 9.5139,
+      "step": 53620
+    },
+    {
+      "epoch": 0.2678219181502659,
+      "grad_norm": 0.09977221488952637,
+      "learning_rate": 2.202032591554232e-05,
+      "loss": 9.5064,
+      "step": 53630
+    },
+    {
+      "epoch": 0.2678718569752054,
+      "grad_norm": 0.09725436568260193,
+      "learning_rate": 2.2018824000600766e-05,
+      "loss": 9.505,
+      "step": 53640
+    },
+    {
+      "epoch": 0.2679217958001448,
+      "grad_norm": 0.09925449639558792,
+      "learning_rate": 2.2017322085659216e-05,
+      "loss": 9.5171,
+      "step": 53650
+    },
+    {
+      "epoch": 0.2679717346250843,
+      "grad_norm": 0.09515231102705002,
+      "learning_rate": 2.2015820170717666e-05,
+      "loss": 9.5013,
+      "step": 53660
+    },
+    {
+      "epoch": 0.2680216734500237,
+      "grad_norm": 0.09482384473085403,
+      "learning_rate": 2.2014318255776113e-05,
+      "loss": 9.4974,
+      "step": 53670
+    },
+    {
+      "epoch": 0.2680716122749632,
+      "grad_norm": 0.09810610860586166,
+      "learning_rate": 2.2012816340834566e-05,
+      "loss": 9.5044,
+      "step": 53680
+    },
+    {
+      "epoch": 0.2681215510999026,
+      "grad_norm": 0.09548275917768478,
+      "learning_rate": 2.2011314425893013e-05,
+      "loss": 9.5023,
+      "step": 53690
+    },
+    {
+      "epoch": 0.2681714899248421,
+      "grad_norm": 0.10010819137096405,
+      "learning_rate": 2.2009812510951463e-05,
+      "loss": 9.5103,
+      "step": 53700
+    },
+    {
+      "epoch": 0.2682214287497815,
+      "grad_norm": 0.09645957499742508,
+      "learning_rate": 2.2008310596009913e-05,
+      "loss": 9.4937,
+      "step": 53710
+    },
+    {
+      "epoch": 0.268271367574721,
+      "grad_norm": 0.09398103505373001,
+      "learning_rate": 2.200680868106836e-05,
+      "loss": 9.496,
+      "step": 53720
+    },
+    {
+      "epoch": 0.2683213063996604,
+      "grad_norm": 0.09286051243543625,
+      "learning_rate": 2.2005306766126814e-05,
+      "loss": 9.5036,
+      "step": 53730
+    },
+    {
+      "epoch": 0.2683712452245999,
+      "grad_norm": 0.09161993116140366,
+      "learning_rate": 2.200380485118526e-05,
+      "loss": 9.5114,
+      "step": 53740
+    },
+    {
+      "epoch": 0.2684211840495393,
+      "grad_norm": 0.09624673426151276,
+      "learning_rate": 2.200230293624371e-05,
+      "loss": 9.5007,
+      "step": 53750
+    },
+    {
+      "epoch": 0.2684711228744788,
+      "grad_norm": 0.08799513429403305,
+      "learning_rate": 2.200080102130216e-05,
+      "loss": 9.5013,
+      "step": 53760
+    },
+    {
+      "epoch": 0.2685210616994182,
+      "grad_norm": 0.09283116459846497,
+      "learning_rate": 2.199929910636061e-05,
+      "loss": 9.5102,
+      "step": 53770
+    },
+    {
+      "epoch": 0.2685710005243577,
+      "grad_norm": 0.09318029880523682,
+      "learning_rate": 2.199779719141906e-05,
+      "loss": 9.509,
+      "step": 53780
+    },
+    {
+      "epoch": 0.2686209393492971,
+      "grad_norm": 0.09814724326133728,
+      "learning_rate": 2.1996295276477508e-05,
+      "loss": 9.502,
+      "step": 53790
+    },
+    {
+      "epoch": 0.26867087817423657,
+      "grad_norm": 0.08983851969242096,
+      "learning_rate": 2.1994793361535958e-05,
+      "loss": 9.504,
+      "step": 53800
+    },
+    {
+      "epoch": 0.268720816999176,
+      "grad_norm": 0.09457878768444061,
+      "learning_rate": 2.199329144659441e-05,
+      "loss": 9.4922,
+      "step": 53810
+    },
+    {
+      "epoch": 0.26877075582411547,
+      "grad_norm": 0.09290546178817749,
+      "learning_rate": 2.199178953165286e-05,
+      "loss": 9.497,
+      "step": 53820
+    },
+    {
+      "epoch": 0.2688206946490549,
+      "grad_norm": 0.09579872339963913,
+      "learning_rate": 2.199028761671131e-05,
+      "loss": 9.5007,
+      "step": 53830
+    },
+    {
+      "epoch": 0.26887063347399437,
+      "grad_norm": 0.09526251256465912,
+      "learning_rate": 2.1988785701769756e-05,
+      "loss": 9.4973,
+      "step": 53840
+    },
+    {
+      "epoch": 0.2689205722989338,
+      "grad_norm": 0.09075421094894409,
+      "learning_rate": 2.1987283786828206e-05,
+      "loss": 9.4985,
+      "step": 53850
+    },
+    {
+      "epoch": 0.26897051112387327,
+      "grad_norm": 0.09878144413232803,
+      "learning_rate": 2.1985781871886656e-05,
+      "loss": 9.4944,
+      "step": 53860
+    },
+    {
+      "epoch": 0.2690204499488127,
+      "grad_norm": 0.092621810734272,
+      "learning_rate": 2.1984279956945106e-05,
+      "loss": 9.4981,
+      "step": 53870
+    },
+    {
+      "epoch": 0.26907038877375217,
+      "grad_norm": 0.0957510843873024,
+      "learning_rate": 2.1982778042003556e-05,
+      "loss": 9.5005,
+      "step": 53880
+    },
+    {
+      "epoch": 0.2691203275986916,
+      "grad_norm": 0.0936402752995491,
+      "learning_rate": 2.1981276127062003e-05,
+      "loss": 9.4948,
+      "step": 53890
+    },
+    {
+      "epoch": 0.26917026642363107,
+      "grad_norm": 0.09587475657463074,
+      "learning_rate": 2.1979774212120453e-05,
+      "loss": 9.5018,
+      "step": 53900
+    },
+    {
+      "epoch": 0.2692202052485705,
+      "grad_norm": 0.09456077963113785,
+      "learning_rate": 2.1978272297178903e-05,
+      "loss": 9.496,
+      "step": 53910
+    },
+    {
+      "epoch": 0.26927014407350996,
+      "grad_norm": 0.10105657577514648,
+      "learning_rate": 2.1976770382237354e-05,
+      "loss": 9.4946,
+      "step": 53920
+    },
+    {
+      "epoch": 0.2693200828984494,
+      "grad_norm": 0.09980333596467972,
+      "learning_rate": 2.1975268467295804e-05,
+      "loss": 9.4955,
+      "step": 53930
+    },
+    {
+      "epoch": 0.26937002172338886,
+      "grad_norm": 0.09290049970149994,
+      "learning_rate": 2.197376655235425e-05,
+      "loss": 9.5054,
+      "step": 53940
+    },
+    {
+      "epoch": 0.2694199605483283,
+      "grad_norm": 0.09242033958435059,
+      "learning_rate": 2.19722646374127e-05,
+      "loss": 9.4997,
+      "step": 53950
+    },
+    {
+      "epoch": 0.26946989937326776,
+      "grad_norm": 0.09912257641553879,
+      "learning_rate": 2.197076272247115e-05,
+      "loss": 9.4969,
+      "step": 53960
+    },
+    {
+      "epoch": 0.2695198381982072,
+      "grad_norm": 0.10186777263879776,
+      "learning_rate": 2.19692608075296e-05,
+      "loss": 9.4975,
+      "step": 53970
+    },
+    {
+      "epoch": 0.26956977702314666,
+      "grad_norm": 0.09263597428798676,
+      "learning_rate": 2.196775889258805e-05,
+      "loss": 9.4902,
+      "step": 53980
+    },
+    {
+      "epoch": 0.2696197158480861,
+      "grad_norm": 0.09329408407211304,
+      "learning_rate": 2.1966256977646498e-05,
+      "loss": 9.4953,
+      "step": 53990
+    },
+    {
+      "epoch": 0.26966965467302556,
+      "grad_norm": 0.09728638082742691,
+      "learning_rate": 2.1964755062704948e-05,
+      "loss": 9.5026,
+      "step": 54000
+    },
+    {
+      "epoch": 0.269719593497965,
+      "grad_norm": 0.0942746102809906,
+      "learning_rate": 2.19632531477634e-05,
+      "loss": 9.5005,
+      "step": 54010
+    },
+    {
+      "epoch": 0.26976953232290446,
+      "grad_norm": 0.08841151744127274,
+      "learning_rate": 2.196175123282185e-05,
+      "loss": 9.501,
+      "step": 54020
+    },
+    {
+      "epoch": 0.2698194711478439,
+      "grad_norm": 0.09076721221208572,
+      "learning_rate": 2.19602493178803e-05,
+      "loss": 9.4855,
+      "step": 54030
+    },
+    {
+      "epoch": 0.26986940997278336,
+      "grad_norm": 0.09586405754089355,
+      "learning_rate": 2.1958747402938746e-05,
+      "loss": 9.5019,
+      "step": 54040
+    },
+    {
+      "epoch": 0.2699193487977228,
+      "grad_norm": 0.09588458389043808,
+      "learning_rate": 2.1957245487997196e-05,
+      "loss": 9.4985,
+      "step": 54050
+    },
+    {
+      "epoch": 0.26996928762266226,
+      "grad_norm": 0.09645280987024307,
+      "learning_rate": 2.1955743573055646e-05,
+      "loss": 9.5033,
+      "step": 54060
+    },
+    {
+      "epoch": 0.2700192264476017,
+      "grad_norm": 0.0946224108338356,
+      "learning_rate": 2.1954241658114096e-05,
+      "loss": 9.507,
+      "step": 54070
+    },
+    {
+      "epoch": 0.27006916527254116,
+      "grad_norm": 0.09653602540493011,
+      "learning_rate": 2.1952739743172546e-05,
+      "loss": 9.5072,
+      "step": 54080
+    },
+    {
+      "epoch": 0.2701191040974806,
+      "grad_norm": 0.09602764248847961,
+      "learning_rate": 2.1951237828230996e-05,
+      "loss": 9.4974,
+      "step": 54090
+    },
+    {
+      "epoch": 0.27016904292242006,
+      "grad_norm": 0.09846575558185577,
+      "learning_rate": 2.1949735913289443e-05,
+      "loss": 9.4918,
+      "step": 54100
+    },
+    {
+      "epoch": 0.2702189817473595,
+      "grad_norm": 0.09469395130872726,
+      "learning_rate": 2.1948233998347893e-05,
+      "loss": 9.4976,
+      "step": 54110
+    },
+    {
+      "epoch": 0.27026892057229895,
+      "grad_norm": 0.09206046909093857,
+      "learning_rate": 2.1946732083406344e-05,
+      "loss": 9.494,
+      "step": 54120
+    },
+    {
+      "epoch": 0.2703188593972384,
+      "grad_norm": 0.10043230652809143,
+      "learning_rate": 2.1945230168464794e-05,
+      "loss": 9.4947,
+      "step": 54130
+    },
+    {
+      "epoch": 0.27036879822217785,
+      "grad_norm": 0.09006008505821228,
+      "learning_rate": 2.1943728253523244e-05,
+      "loss": 9.5057,
+      "step": 54140
+    },
+    {
+      "epoch": 0.2704187370471173,
+      "grad_norm": 0.09778618067502975,
+      "learning_rate": 2.194222633858169e-05,
+      "loss": 9.5036,
+      "step": 54150
+    },
+    {
+      "epoch": 0.27046867587205675,
+      "grad_norm": 0.09430201351642609,
+      "learning_rate": 2.194072442364014e-05,
+      "loss": 9.5066,
+      "step": 54160
+    },
+    {
+      "epoch": 0.2705186146969962,
+      "grad_norm": 0.09432802349328995,
+      "learning_rate": 2.193922250869859e-05,
+      "loss": 9.4916,
+      "step": 54170
+    },
+    {
+      "epoch": 0.27056855352193565,
+      "grad_norm": 0.09633664041757584,
+      "learning_rate": 2.193772059375704e-05,
+      "loss": 9.4962,
+      "step": 54180
+    },
+    {
+      "epoch": 0.2706184923468751,
+      "grad_norm": 0.09365957230329514,
+      "learning_rate": 2.193621867881549e-05,
+      "loss": 9.5004,
+      "step": 54190
+    },
+    {
+      "epoch": 0.27066843117181455,
+      "grad_norm": 0.0957445502281189,
+      "learning_rate": 2.1934716763873938e-05,
+      "loss": 9.5006,
+      "step": 54200
+    },
+    {
+      "epoch": 0.27071836999675397,
+      "grad_norm": 0.09612950682640076,
+      "learning_rate": 2.193321484893239e-05,
+      "loss": 9.4985,
+      "step": 54210
+    },
+    {
+      "epoch": 0.27076830882169345,
+      "grad_norm": 0.09785423427820206,
+      "learning_rate": 2.193171293399084e-05,
+      "loss": 9.4971,
+      "step": 54220
+    },
+    {
+      "epoch": 0.27081824764663287,
+      "grad_norm": 0.09333615750074387,
+      "learning_rate": 2.193021101904929e-05,
+      "loss": 9.4973,
+      "step": 54230
+    },
+    {
+      "epoch": 0.27086818647157235,
+      "grad_norm": 0.10054484754800797,
+      "learning_rate": 2.192870910410774e-05,
+      "loss": 9.5062,
+      "step": 54240
+    },
+    {
+      "epoch": 0.27091812529651177,
+      "grad_norm": 0.09158135205507278,
+      "learning_rate": 2.1927207189166186e-05,
+      "loss": 9.5018,
+      "step": 54250
+    },
+    {
+      "epoch": 0.27096806412145125,
+      "grad_norm": 0.09675116837024689,
+      "learning_rate": 2.1925705274224636e-05,
+      "loss": 9.5022,
+      "step": 54260
+    },
+    {
+      "epoch": 0.27101800294639067,
+      "grad_norm": 0.09301751852035522,
+      "learning_rate": 2.1924203359283086e-05,
+      "loss": 9.4876,
+      "step": 54270
+    },
+    {
+      "epoch": 0.27106794177133015,
+      "grad_norm": 0.09641415625810623,
+      "learning_rate": 2.1922701444341536e-05,
+      "loss": 9.489,
+      "step": 54280
+    },
+    {
+      "epoch": 0.27111788059626957,
+      "grad_norm": 0.09570831805467606,
+      "learning_rate": 2.1921199529399986e-05,
+      "loss": 9.4929,
+      "step": 54290
+    },
+    {
+      "epoch": 0.27116781942120904,
+      "grad_norm": 0.09328965097665787,
+      "learning_rate": 2.1919697614458433e-05,
+      "loss": 9.4963,
+      "step": 54300
+    },
+    {
+      "epoch": 0.27121775824614847,
+      "grad_norm": 0.09763366729021072,
+      "learning_rate": 2.1918195699516883e-05,
+      "loss": 9.4935,
+      "step": 54310
+    },
+    {
+      "epoch": 0.27126769707108794,
+      "grad_norm": 0.097484290599823,
+      "learning_rate": 2.1916693784575334e-05,
+      "loss": 9.4983,
+      "step": 54320
+    },
+    {
+      "epoch": 0.27131763589602736,
+      "grad_norm": 0.09273017942905426,
+      "learning_rate": 2.1915191869633784e-05,
+      "loss": 9.5023,
+      "step": 54330
+    },
+    {
+      "epoch": 0.27136757472096684,
+      "grad_norm": 0.08936069160699844,
+      "learning_rate": 2.1913689954692234e-05,
+      "loss": 9.4918,
+      "step": 54340
+    },
+    {
+      "epoch": 0.27141751354590626,
+      "grad_norm": 0.08930017799139023,
+      "learning_rate": 2.191218803975068e-05,
+      "loss": 9.4937,
+      "step": 54350
+    },
+    {
+      "epoch": 0.27146745237084574,
+      "grad_norm": 0.0880388543009758,
+      "learning_rate": 2.191068612480913e-05,
+      "loss": 9.4962,
+      "step": 54360
+    },
+    {
+      "epoch": 0.27151739119578516,
+      "grad_norm": 0.10103661566972733,
+      "learning_rate": 2.190918420986758e-05,
+      "loss": 9.4915,
+      "step": 54370
+    },
+    {
+      "epoch": 0.27156733002072464,
+      "grad_norm": 0.09336966276168823,
+      "learning_rate": 2.190768229492603e-05,
+      "loss": 9.487,
+      "step": 54380
+    },
+    {
+      "epoch": 0.27161726884566406,
+      "grad_norm": 0.09124721586704254,
+      "learning_rate": 2.190618037998448e-05,
+      "loss": 9.4872,
+      "step": 54390
+    },
+    {
+      "epoch": 0.2716672076706035,
+      "grad_norm": 0.09562219679355621,
+      "learning_rate": 2.1904678465042928e-05,
+      "loss": 9.4953,
+      "step": 54400
+    },
+    {
+      "epoch": 0.27171714649554296,
+      "grad_norm": 0.0949045866727829,
+      "learning_rate": 2.1903176550101382e-05,
+      "loss": 9.505,
+      "step": 54410
+    },
+    {
+      "epoch": 0.2717670853204824,
+      "grad_norm": 0.09756708145141602,
+      "learning_rate": 2.190167463515983e-05,
+      "loss": 9.5009,
+      "step": 54420
+    },
+    {
+      "epoch": 0.27181702414542186,
+      "grad_norm": 0.08797673135995865,
+      "learning_rate": 2.190017272021828e-05,
+      "loss": 9.4954,
+      "step": 54430
+    },
+    {
+      "epoch": 0.2718669629703613,
+      "grad_norm": 0.09841335564851761,
+      "learning_rate": 2.189867080527673e-05,
+      "loss": 9.4911,
+      "step": 54440
+    },
+    {
+      "epoch": 0.27191690179530076,
+      "grad_norm": 0.09503602236509323,
+      "learning_rate": 2.1897168890335176e-05,
+      "loss": 9.5034,
+      "step": 54450
+    },
+    {
+      "epoch": 0.2719668406202402,
+      "grad_norm": 0.09067325294017792,
+      "learning_rate": 2.189566697539363e-05,
+      "loss": 9.4863,
+      "step": 54460
+    },
+    {
+      "epoch": 0.27201677944517966,
+      "grad_norm": 0.100099578499794,
+      "learning_rate": 2.1894165060452076e-05,
+      "loss": 9.5065,
+      "step": 54470
+    },
+    {
+      "epoch": 0.2720667182701191,
+      "grad_norm": 0.09303244203329086,
+      "learning_rate": 2.1892663145510526e-05,
+      "loss": 9.4933,
+      "step": 54480
+    },
+    {
+      "epoch": 0.27211665709505856,
+      "grad_norm": 0.09559056162834167,
+      "learning_rate": 2.1891161230568977e-05,
+      "loss": 9.492,
+      "step": 54490
+    },
+    {
+      "epoch": 0.272166595919998,
+      "grad_norm": 0.09210564941167831,
+      "learning_rate": 2.1889659315627423e-05,
+      "loss": 9.5017,
+      "step": 54500
+    },
+    {
+      "epoch": 0.27221653474493746,
+      "grad_norm": 0.09366162121295929,
+      "learning_rate": 2.1888157400685877e-05,
+      "loss": 9.493,
+      "step": 54510
+    },
+    {
+      "epoch": 0.2722664735698769,
+      "grad_norm": 0.09506445378065109,
+      "learning_rate": 2.1886655485744324e-05,
+      "loss": 9.4811,
+      "step": 54520
+    },
+    {
+      "epoch": 0.27231641239481635,
+      "grad_norm": 0.10406965017318726,
+      "learning_rate": 2.1885153570802774e-05,
+      "loss": 9.4821,
+      "step": 54530
+    },
+    {
+      "epoch": 0.2723663512197558,
+      "grad_norm": 0.08874404430389404,
+      "learning_rate": 2.1883651655861224e-05,
+      "loss": 9.5002,
+      "step": 54540
+    },
+    {
+      "epoch": 0.27241629004469525,
+      "grad_norm": 0.09706084430217743,
+      "learning_rate": 2.188214974091967e-05,
+      "loss": 9.4876,
+      "step": 54550
+    },
+    {
+      "epoch": 0.2724662288696347,
+      "grad_norm": 0.09322942048311234,
+      "learning_rate": 2.1880647825978124e-05,
+      "loss": 9.4887,
+      "step": 54560
+    },
+    {
+      "epoch": 0.27251616769457415,
+      "grad_norm": 0.09634638577699661,
+      "learning_rate": 2.187914591103657e-05,
+      "loss": 9.4933,
+      "step": 54570
+    },
+    {
+      "epoch": 0.2725661065195136,
+      "grad_norm": 0.09559839218854904,
+      "learning_rate": 2.187764399609502e-05,
+      "loss": 9.4836,
+      "step": 54580
+    },
+    {
+      "epoch": 0.27261604534445305,
+      "grad_norm": 0.08715943992137909,
+      "learning_rate": 2.187614208115347e-05,
+      "loss": 9.4903,
+      "step": 54590
+    },
+    {
+      "epoch": 0.2726659841693925,
+      "grad_norm": 0.09356392920017242,
+      "learning_rate": 2.1874640166211918e-05,
+      "loss": 9.4909,
+      "step": 54600
+    },
+    {
+      "epoch": 0.27271592299433195,
+      "grad_norm": 0.09131593257188797,
+      "learning_rate": 2.1873138251270372e-05,
+      "loss": 9.4938,
+      "step": 54610
+    },
+    {
+      "epoch": 0.27276586181927137,
+      "grad_norm": 0.09553367644548416,
+      "learning_rate": 2.187163633632882e-05,
+      "loss": 9.4918,
+      "step": 54620
+    },
+    {
+      "epoch": 0.27281580064421085,
+      "grad_norm": 0.09892634302377701,
+      "learning_rate": 2.187013442138727e-05,
+      "loss": 9.484,
+      "step": 54630
+    },
+    {
+      "epoch": 0.27286573946915027,
+      "grad_norm": 0.09537822008132935,
+      "learning_rate": 2.186863250644572e-05,
+      "loss": 9.4901,
+      "step": 54640
+    },
+    {
+      "epoch": 0.27291567829408975,
+      "grad_norm": 0.09602360427379608,
+      "learning_rate": 2.1867130591504166e-05,
+      "loss": 9.4821,
+      "step": 54650
+    },
+    {
+      "epoch": 0.27296561711902917,
+      "grad_norm": 0.09913076460361481,
+      "learning_rate": 2.186562867656262e-05,
+      "loss": 9.488,
+      "step": 54660
+    },
+    {
+      "epoch": 0.27301555594396865,
+      "grad_norm": 0.09501135349273682,
+      "learning_rate": 2.1864126761621066e-05,
+      "loss": 9.4864,
+      "step": 54670
+    },
+    {
+      "epoch": 0.27306549476890807,
+      "grad_norm": 0.0974724143743515,
+      "learning_rate": 2.1862624846679516e-05,
+      "loss": 9.4942,
+      "step": 54680
+    },
+    {
+      "epoch": 0.27311543359384755,
+      "grad_norm": 0.09023966640233994,
+      "learning_rate": 2.1861122931737967e-05,
+      "loss": 9.4859,
+      "step": 54690
+    },
+    {
+      "epoch": 0.27316537241878697,
+      "grad_norm": 0.09034571051597595,
+      "learning_rate": 2.1859621016796417e-05,
+      "loss": 9.4907,
+      "step": 54700
+    },
+    {
+      "epoch": 0.27321531124372644,
+      "grad_norm": 0.09617186337709427,
+      "learning_rate": 2.1858119101854867e-05,
+      "loss": 9.483,
+      "step": 54710
+    },
+    {
+      "epoch": 0.27326525006866587,
+      "grad_norm": 0.09283477067947388,
+      "learning_rate": 2.1856617186913314e-05,
+      "loss": 9.4723,
+      "step": 54720
+    },
+    {
+      "epoch": 0.27331518889360534,
+      "grad_norm": 0.0881299078464508,
+      "learning_rate": 2.1855115271971767e-05,
+      "loss": 9.4951,
+      "step": 54730
+    },
+    {
+      "epoch": 0.27336512771854476,
+      "grad_norm": 0.09953461587429047,
+      "learning_rate": 2.1853613357030214e-05,
+      "loss": 9.4853,
+      "step": 54740
+    },
+    {
+      "epoch": 0.27341506654348424,
+      "grad_norm": 0.09831681102514267,
+      "learning_rate": 2.1852111442088664e-05,
+      "loss": 9.4849,
+      "step": 54750
+    },
+    {
+      "epoch": 0.27346500536842366,
+      "grad_norm": 0.0924922525882721,
+      "learning_rate": 2.1850609527147114e-05,
+      "loss": 9.4929,
+      "step": 54760
+    },
+    {
+      "epoch": 0.27351494419336314,
+      "grad_norm": 0.09693264961242676,
+      "learning_rate": 2.184910761220556e-05,
+      "loss": 9.4778,
+      "step": 54770
+    },
+    {
+      "epoch": 0.27356488301830256,
+      "grad_norm": 0.0940731093287468,
+      "learning_rate": 2.1847605697264015e-05,
+      "loss": 9.4909,
+      "step": 54780
+    },
+    {
+      "epoch": 0.27361482184324204,
+      "grad_norm": 0.09551883488893509,
+      "learning_rate": 2.184610378232246e-05,
+      "loss": 9.485,
+      "step": 54790
+    },
+    {
+      "epoch": 0.27366476066818146,
+      "grad_norm": 0.09278986603021622,
+      "learning_rate": 2.1844601867380912e-05,
+      "loss": 9.4857,
+      "step": 54800
+    },
+    {
+      "epoch": 0.27371469949312094,
+      "grad_norm": 0.09064249694347382,
+      "learning_rate": 2.1843099952439362e-05,
+      "loss": 9.4826,
+      "step": 54810
+    },
+    {
+      "epoch": 0.27376463831806036,
+      "grad_norm": 0.0937790721654892,
+      "learning_rate": 2.184159803749781e-05,
+      "loss": 9.4923,
+      "step": 54820
+    },
+    {
+      "epoch": 0.27381457714299984,
+      "grad_norm": 0.0964457243680954,
+      "learning_rate": 2.1840096122556262e-05,
+      "loss": 9.4783,
+      "step": 54830
+    },
+    {
+      "epoch": 0.27386451596793926,
+      "grad_norm": 0.08797242492437363,
+      "learning_rate": 2.183859420761471e-05,
+      "loss": 9.481,
+      "step": 54840
+    },
+    {
+      "epoch": 0.27391445479287874,
+      "grad_norm": 0.09727727621793747,
+      "learning_rate": 2.183709229267316e-05,
+      "loss": 9.4801,
+      "step": 54850
+    },
+    {
+      "epoch": 0.27396439361781816,
+      "grad_norm": 0.0918155238032341,
+      "learning_rate": 2.183559037773161e-05,
+      "loss": 9.4905,
+      "step": 54860
+    },
+    {
+      "epoch": 0.27401433244275764,
+      "grad_norm": 0.09824883192777634,
+      "learning_rate": 2.1834088462790056e-05,
+      "loss": 9.4761,
+      "step": 54870
+    },
+    {
+      "epoch": 0.27406427126769706,
+      "grad_norm": 0.09386686235666275,
+      "learning_rate": 2.183258654784851e-05,
+      "loss": 9.4937,
+      "step": 54880
+    },
+    {
+      "epoch": 0.27411421009263653,
+      "grad_norm": 0.09294228255748749,
+      "learning_rate": 2.1831084632906957e-05,
+      "loss": 9.4793,
+      "step": 54890
+    },
+    {
+      "epoch": 0.27416414891757596,
+      "grad_norm": 0.10264691710472107,
+      "learning_rate": 2.1829582717965407e-05,
+      "loss": 9.4796,
+      "step": 54900
+    },
+    {
+      "epoch": 0.27421408774251543,
+      "grad_norm": 0.09530775249004364,
+      "learning_rate": 2.1828080803023857e-05,
+      "loss": 9.479,
+      "step": 54910
+    },
+    {
+      "epoch": 0.27426402656745485,
+      "grad_norm": 0.0969926044344902,
+      "learning_rate": 2.1826578888082304e-05,
+      "loss": 9.4762,
+      "step": 54920
+    },
+    {
+      "epoch": 0.27431396539239433,
+      "grad_norm": 0.0955069363117218,
+      "learning_rate": 2.1825076973140757e-05,
+      "loss": 9.4781,
+      "step": 54930
+    },
+    {
+      "epoch": 0.27436390421733375,
+      "grad_norm": 0.09909159690141678,
+      "learning_rate": 2.1823575058199204e-05,
+      "loss": 9.4861,
+      "step": 54940
+    },
+    {
+      "epoch": 0.27441384304227323,
+      "grad_norm": 0.09283483028411865,
+      "learning_rate": 2.1822073143257654e-05,
+      "loss": 9.4774,
+      "step": 54950
+    },
+    {
+      "epoch": 0.27446378186721265,
+      "grad_norm": 0.09053994715213776,
+      "learning_rate": 2.1820571228316104e-05,
+      "loss": 9.4813,
+      "step": 54960
+    },
+    {
+      "epoch": 0.27451372069215213,
+      "grad_norm": 0.0921054258942604,
+      "learning_rate": 2.181906931337455e-05,
+      "loss": 9.4889,
+      "step": 54970
+    },
+    {
+      "epoch": 0.27456365951709155,
+      "grad_norm": 0.09481295198202133,
+      "learning_rate": 2.1817567398433005e-05,
+      "loss": 9.485,
+      "step": 54980
+    },
+    {
+      "epoch": 0.27461359834203103,
+      "grad_norm": 0.10501012206077576,
+      "learning_rate": 2.181606548349145e-05,
+      "loss": 9.4831,
+      "step": 54990
+    },
+    {
+      "epoch": 0.27466353716697045,
+      "grad_norm": 0.09528646618127823,
+      "learning_rate": 2.1814563568549902e-05,
+      "loss": 9.4872,
+      "step": 55000
+    },
+    {
+      "epoch": 0.27471347599190993,
+      "grad_norm": 0.09487199038267136,
+      "learning_rate": 2.1813061653608352e-05,
+      "loss": 9.4929,
+      "step": 55010
+    },
+    {
+      "epoch": 0.27476341481684935,
+      "grad_norm": 0.09505639970302582,
+      "learning_rate": 2.18115597386668e-05,
+      "loss": 9.4858,
+      "step": 55020
+    },
+    {
+      "epoch": 0.2748133536417888,
+      "grad_norm": 0.0927814468741417,
+      "learning_rate": 2.1810057823725252e-05,
+      "loss": 9.4791,
+      "step": 55030
+    },
+    {
+      "epoch": 0.27486329246672825,
+      "grad_norm": 0.09519551694393158,
+      "learning_rate": 2.18085559087837e-05,
+      "loss": 9.4805,
+      "step": 55040
+    },
+    {
+      "epoch": 0.2749132312916677,
+      "grad_norm": 0.09184595942497253,
+      "learning_rate": 2.1807053993842153e-05,
+      "loss": 9.4717,
+      "step": 55050
+    },
+    {
+      "epoch": 0.27496317011660715,
+      "grad_norm": 0.09657662361860275,
+      "learning_rate": 2.18055520789006e-05,
+      "loss": 9.4875,
+      "step": 55060
+    },
+    {
+      "epoch": 0.2750131089415466,
+      "grad_norm": 0.0927683562040329,
+      "learning_rate": 2.1804050163959046e-05,
+      "loss": 9.4826,
+      "step": 55070
+    },
+    {
+      "epoch": 0.27506304776648605,
+      "grad_norm": 0.09400233626365662,
+      "learning_rate": 2.18025482490175e-05,
+      "loss": 9.4751,
+      "step": 55080
+    },
+    {
+      "epoch": 0.2751129865914255,
+      "grad_norm": 0.09439916908740997,
+      "learning_rate": 2.1801046334075947e-05,
+      "loss": 9.4739,
+      "step": 55090
+    },
+    {
+      "epoch": 0.27516292541636495,
+      "grad_norm": 0.09644033014774323,
+      "learning_rate": 2.17995444191344e-05,
+      "loss": 9.4795,
+      "step": 55100
+    },
+    {
+      "epoch": 0.2752128642413044,
+      "grad_norm": 0.09258447587490082,
+      "learning_rate": 2.1798042504192847e-05,
+      "loss": 9.4774,
+      "step": 55110
+    },
+    {
+      "epoch": 0.27526280306624384,
+      "grad_norm": 0.09144683927297592,
+      "learning_rate": 2.1796540589251294e-05,
+      "loss": 9.4767,
+      "step": 55120
+    },
+    {
+      "epoch": 0.2753127418911833,
+      "grad_norm": 0.09092001616954803,
+      "learning_rate": 2.1795038674309747e-05,
+      "loss": 9.4839,
+      "step": 55130
+    },
+    {
+      "epoch": 0.27536268071612274,
+      "grad_norm": 0.0951116681098938,
+      "learning_rate": 2.1793536759368194e-05,
+      "loss": 9.4774,
+      "step": 55140
+    },
+    {
+      "epoch": 0.2754126195410622,
+      "grad_norm": 0.09364614635705948,
+      "learning_rate": 2.1792034844426648e-05,
+      "loss": 9.4739,
+      "step": 55150
+    },
+    {
+      "epoch": 0.27546255836600164,
+      "grad_norm": 0.09528619796037674,
+      "learning_rate": 2.1790532929485094e-05,
+      "loss": 9.4745,
+      "step": 55160
+    },
+    {
+      "epoch": 0.2755124971909411,
+      "grad_norm": 0.09560272097587585,
+      "learning_rate": 2.178903101454354e-05,
+      "loss": 9.4794,
+      "step": 55170
+    },
+    {
+      "epoch": 0.27556243601588054,
+      "grad_norm": 0.09209144115447998,
+      "learning_rate": 2.1787529099601995e-05,
+      "loss": 9.4788,
+      "step": 55180
+    },
+    {
+      "epoch": 0.27561237484082,
+      "grad_norm": 0.09365902096033096,
+      "learning_rate": 2.178602718466044e-05,
+      "loss": 9.472,
+      "step": 55190
+    },
+    {
+      "epoch": 0.27566231366575944,
+      "grad_norm": 0.09534572809934616,
+      "learning_rate": 2.1784525269718895e-05,
+      "loss": 9.4737,
+      "step": 55200
+    },
+    {
+      "epoch": 0.2757122524906989,
+      "grad_norm": 0.09372628480195999,
+      "learning_rate": 2.1783023354777342e-05,
+      "loss": 9.4766,
+      "step": 55210
+    },
+    {
+      "epoch": 0.27576219131563834,
+      "grad_norm": 0.09953497350215912,
+      "learning_rate": 2.178152143983579e-05,
+      "loss": 9.4827,
+      "step": 55220
+    },
+    {
+      "epoch": 0.2758121301405778,
+      "grad_norm": 0.0936177521944046,
+      "learning_rate": 2.1780019524894242e-05,
+      "loss": 9.4838,
+      "step": 55230
+    },
+    {
+      "epoch": 0.27586206896551724,
+      "grad_norm": 0.0956781804561615,
+      "learning_rate": 2.177851760995269e-05,
+      "loss": 9.4818,
+      "step": 55240
+    },
+    {
+      "epoch": 0.2759120077904567,
+      "grad_norm": 0.09287876635789871,
+      "learning_rate": 2.1777015695011143e-05,
+      "loss": 9.4868,
+      "step": 55250
+    },
+    {
+      "epoch": 0.27596194661539614,
+      "grad_norm": 0.09168554097414017,
+      "learning_rate": 2.177551378006959e-05,
+      "loss": 9.4796,
+      "step": 55260
+    },
+    {
+      "epoch": 0.2760118854403356,
+      "grad_norm": 0.09445545822381973,
+      "learning_rate": 2.1774011865128036e-05,
+      "loss": 9.4829,
+      "step": 55270
+    },
+    {
+      "epoch": 0.27606182426527504,
+      "grad_norm": 0.09545770287513733,
+      "learning_rate": 2.177250995018649e-05,
+      "loss": 9.4689,
+      "step": 55280
+    },
+    {
+      "epoch": 0.2761117630902145,
+      "grad_norm": 0.09333658963441849,
+      "learning_rate": 2.1771008035244937e-05,
+      "loss": 9.4856,
+      "step": 55290
+    },
+    {
+      "epoch": 0.27616170191515393,
+      "grad_norm": 0.09187939018011093,
+      "learning_rate": 2.176950612030339e-05,
+      "loss": 9.4839,
+      "step": 55300
+    },
+    {
+      "epoch": 0.2762116407400934,
+      "grad_norm": 0.09599092602729797,
+      "learning_rate": 2.1768004205361837e-05,
+      "loss": 9.4787,
+      "step": 55310
+    },
+    {
+      "epoch": 0.27626157956503283,
+      "grad_norm": 0.09836266934871674,
+      "learning_rate": 2.1766502290420284e-05,
+      "loss": 9.4733,
+      "step": 55320
+    },
+    {
+      "epoch": 0.2763115183899723,
+      "grad_norm": 0.08872494846582413,
+      "learning_rate": 2.1765000375478737e-05,
+      "loss": 9.4781,
+      "step": 55330
+    },
+    {
+      "epoch": 0.27636145721491173,
+      "grad_norm": 0.09891574829816818,
+      "learning_rate": 2.1763498460537184e-05,
+      "loss": 9.4845,
+      "step": 55340
+    },
+    {
+      "epoch": 0.2764113960398512,
+      "grad_norm": 0.09147507697343826,
+      "learning_rate": 2.1761996545595638e-05,
+      "loss": 9.4784,
+      "step": 55350
+    },
+    {
+      "epoch": 0.27646133486479063,
+      "grad_norm": 0.09677154570817947,
+      "learning_rate": 2.1760494630654084e-05,
+      "loss": 9.4795,
+      "step": 55360
+    },
+    {
+      "epoch": 0.2765112736897301,
+      "grad_norm": 0.09255880117416382,
+      "learning_rate": 2.1758992715712535e-05,
+      "loss": 9.4825,
+      "step": 55370
+    },
+    {
+      "epoch": 0.27656121251466953,
+      "grad_norm": 0.09383849054574966,
+      "learning_rate": 2.1757490800770985e-05,
+      "loss": 9.4684,
+      "step": 55380
+    },
+    {
+      "epoch": 0.27661115133960895,
+      "grad_norm": 0.09055565297603607,
+      "learning_rate": 2.175598888582943e-05,
+      "loss": 9.4714,
+      "step": 55390
+    },
+    {
+      "epoch": 0.27666109016454843,
+      "grad_norm": 0.09910639375448227,
+      "learning_rate": 2.1754486970887885e-05,
+      "loss": 9.4788,
+      "step": 55400
+    },
+    {
+      "epoch": 0.27671102898948785,
+      "grad_norm": 0.09024091809988022,
+      "learning_rate": 2.1752985055946332e-05,
+      "loss": 9.4775,
+      "step": 55410
+    },
+    {
+      "epoch": 0.2767609678144273,
+      "grad_norm": 0.09249179810285568,
+      "learning_rate": 2.1751483141004782e-05,
+      "loss": 9.4655,
+      "step": 55420
+    },
+    {
+      "epoch": 0.27681090663936675,
+      "grad_norm": 0.09385978430509567,
+      "learning_rate": 2.1749981226063232e-05,
+      "loss": 9.4792,
+      "step": 55430
+    },
+    {
+      "epoch": 0.2768608454643062,
+      "grad_norm": 0.09004578739404678,
+      "learning_rate": 2.174847931112168e-05,
+      "loss": 9.4759,
+      "step": 55440
+    },
+    {
+      "epoch": 0.27691078428924565,
+      "grad_norm": 0.09427126497030258,
+      "learning_rate": 2.1746977396180133e-05,
+      "loss": 9.4672,
+      "step": 55450
+    },
+    {
+      "epoch": 0.2769607231141851,
+      "grad_norm": 0.09121444821357727,
+      "learning_rate": 2.174547548123858e-05,
+      "loss": 9.4653,
+      "step": 55460
+    },
+    {
+      "epoch": 0.27701066193912455,
+      "grad_norm": 0.0954887717962265,
+      "learning_rate": 2.174397356629703e-05,
+      "loss": 9.47,
+      "step": 55470
+    },
+    {
+      "epoch": 0.277060600764064,
+      "grad_norm": 0.09227491915225983,
+      "learning_rate": 2.174247165135548e-05,
+      "loss": 9.4773,
+      "step": 55480
+    },
+    {
+      "epoch": 0.27711053958900345,
+      "grad_norm": 0.09483451396226883,
+      "learning_rate": 2.1740969736413927e-05,
+      "loss": 9.4778,
+      "step": 55490
+    },
+    {
+      "epoch": 0.2771604784139429,
+      "grad_norm": 0.09309407323598862,
+      "learning_rate": 2.173946782147238e-05,
+      "loss": 9.4802,
+      "step": 55500
+    },
+    {
+      "epoch": 0.27721041723888235,
+      "grad_norm": 0.09302882105112076,
+      "learning_rate": 2.1737965906530827e-05,
+      "loss": 9.4772,
+      "step": 55510
+    },
+    {
+      "epoch": 0.2772603560638218,
+      "grad_norm": 0.09708143770694733,
+      "learning_rate": 2.1736463991589277e-05,
+      "loss": 9.4642,
+      "step": 55520
+    },
+    {
+      "epoch": 0.27731029488876124,
+      "grad_norm": 0.09857852011919022,
+      "learning_rate": 2.1734962076647727e-05,
+      "loss": 9.4724,
+      "step": 55530
+    },
+    {
+      "epoch": 0.2773602337137007,
+      "grad_norm": 0.08941424638032913,
+      "learning_rate": 2.1733460161706174e-05,
+      "loss": 9.4653,
+      "step": 55540
+    },
+    {
+      "epoch": 0.27741017253864014,
+      "grad_norm": 0.09356237202882767,
+      "learning_rate": 2.1731958246764628e-05,
+      "loss": 9.4737,
+      "step": 55550
+    },
+    {
+      "epoch": 0.2774601113635796,
+      "grad_norm": 0.09721656143665314,
+      "learning_rate": 2.1730456331823074e-05,
+      "loss": 9.4649,
+      "step": 55560
+    },
+    {
+      "epoch": 0.27751005018851904,
+      "grad_norm": 0.0922091156244278,
+      "learning_rate": 2.1728954416881525e-05,
+      "loss": 9.4746,
+      "step": 55570
+    },
+    {
+      "epoch": 0.2775599890134585,
+      "grad_norm": 0.09237723052501678,
+      "learning_rate": 2.1727452501939975e-05,
+      "loss": 9.4742,
+      "step": 55580
+    },
+    {
+      "epoch": 0.27760992783839794,
+      "grad_norm": 0.09540941566228867,
+      "learning_rate": 2.172595058699842e-05,
+      "loss": 9.4836,
+      "step": 55590
+    },
+    {
+      "epoch": 0.2776598666633374,
+      "grad_norm": 0.09385842829942703,
+      "learning_rate": 2.1724448672056875e-05,
+      "loss": 9.4653,
+      "step": 55600
+    },
+    {
+      "epoch": 0.27770980548827684,
+      "grad_norm": 0.09224699437618256,
+      "learning_rate": 2.1722946757115322e-05,
+      "loss": 9.4769,
+      "step": 55610
+    },
+    {
+      "epoch": 0.2777597443132163,
+      "grad_norm": 0.08767042309045792,
+      "learning_rate": 2.1721444842173772e-05,
+      "loss": 9.4674,
+      "step": 55620
+    },
+    {
+      "epoch": 0.27780968313815574,
+      "grad_norm": 0.09636397659778595,
+      "learning_rate": 2.1719942927232222e-05,
+      "loss": 9.4676,
+      "step": 55630
+    },
+    {
+      "epoch": 0.2778596219630952,
+      "grad_norm": 0.09190484136343002,
+      "learning_rate": 2.171844101229067e-05,
+      "loss": 9.4758,
+      "step": 55640
+    },
+    {
+      "epoch": 0.27790956078803464,
+      "grad_norm": 0.09460058063268661,
+      "learning_rate": 2.1716939097349123e-05,
+      "loss": 9.4664,
+      "step": 55650
+    },
+    {
+      "epoch": 0.2779594996129741,
+      "grad_norm": 0.0935700535774231,
+      "learning_rate": 2.171543718240757e-05,
+      "loss": 9.4654,
+      "step": 55660
+    },
+    {
+      "epoch": 0.27800943843791354,
+      "grad_norm": 0.09616656601428986,
+      "learning_rate": 2.171393526746602e-05,
+      "loss": 9.4798,
+      "step": 55670
+    },
+    {
+      "epoch": 0.278059377262853,
+      "grad_norm": 0.09719350188970566,
+      "learning_rate": 2.171243335252447e-05,
+      "loss": 9.4636,
+      "step": 55680
+    },
+    {
+      "epoch": 0.27810931608779244,
+      "grad_norm": 0.09644226729869843,
+      "learning_rate": 2.171093143758292e-05,
+      "loss": 9.471,
+      "step": 55690
+    },
+    {
+      "epoch": 0.2781592549127319,
+      "grad_norm": 0.09436850249767303,
+      "learning_rate": 2.170942952264137e-05,
+      "loss": 9.4831,
+      "step": 55700
+    },
+    {
+      "epoch": 0.27820919373767133,
+      "grad_norm": 0.09048687666654587,
+      "learning_rate": 2.1707927607699817e-05,
+      "loss": 9.4731,
+      "step": 55710
+    },
+    {
+      "epoch": 0.2782591325626108,
+      "grad_norm": 0.10046453028917313,
+      "learning_rate": 2.1706425692758267e-05,
+      "loss": 9.4598,
+      "step": 55720
+    },
+    {
+      "epoch": 0.27830907138755023,
+      "grad_norm": 0.09537816792726517,
+      "learning_rate": 2.1704923777816717e-05,
+      "loss": 9.4707,
+      "step": 55730
+    },
+    {
+      "epoch": 0.2783590102124897,
+      "grad_norm": 0.09480351954698563,
+      "learning_rate": 2.1703421862875167e-05,
+      "loss": 9.478,
+      "step": 55740
+    },
+    {
+      "epoch": 0.27840894903742913,
+      "grad_norm": 0.09742586314678192,
+      "learning_rate": 2.1701919947933618e-05,
+      "loss": 9.4817,
+      "step": 55750
+    },
+    {
+      "epoch": 0.2784588878623686,
+      "grad_norm": 0.1001049056649208,
+      "learning_rate": 2.1700418032992064e-05,
+      "loss": 9.4723,
+      "step": 55760
+    },
+    {
+      "epoch": 0.27850882668730803,
+      "grad_norm": 0.09015222638845444,
+      "learning_rate": 2.1698916118050515e-05,
+      "loss": 9.4722,
+      "step": 55770
+    },
+    {
+      "epoch": 0.2785587655122475,
+      "grad_norm": 0.09560934454202652,
+      "learning_rate": 2.1697414203108965e-05,
+      "loss": 9.4617,
+      "step": 55780
+    },
+    {
+      "epoch": 0.27860870433718693,
+      "grad_norm": 0.0906476080417633,
+      "learning_rate": 2.1695912288167415e-05,
+      "loss": 9.4799,
+      "step": 55790
+    },
+    {
+      "epoch": 0.2786586431621264,
+      "grad_norm": 0.09162773191928864,
+      "learning_rate": 2.1694410373225865e-05,
+      "loss": 9.4484,
+      "step": 55800
+    },
+    {
+      "epoch": 0.27870858198706583,
+      "grad_norm": 0.08697135001420975,
+      "learning_rate": 2.1692908458284312e-05,
+      "loss": 9.4698,
+      "step": 55810
+    },
+    {
+      "epoch": 0.2787585208120053,
+      "grad_norm": 0.09465092420578003,
+      "learning_rate": 2.1691406543342762e-05,
+      "loss": 9.4725,
+      "step": 55820
+    },
+    {
+      "epoch": 0.2788084596369447,
+      "grad_norm": 0.09739962965250015,
+      "learning_rate": 2.1689904628401212e-05,
+      "loss": 9.4746,
+      "step": 55830
+    },
+    {
+      "epoch": 0.2788583984618842,
+      "grad_norm": 0.09323502331972122,
+      "learning_rate": 2.1688402713459662e-05,
+      "loss": 9.4643,
+      "step": 55840
+    },
+    {
+      "epoch": 0.2789083372868236,
+      "grad_norm": 0.09480603039264679,
+      "learning_rate": 2.1686900798518113e-05,
+      "loss": 9.4724,
+      "step": 55850
+    },
+    {
+      "epoch": 0.2789582761117631,
+      "grad_norm": 0.09424508363008499,
+      "learning_rate": 2.168539888357656e-05,
+      "loss": 9.4578,
+      "step": 55860
+    },
+    {
+      "epoch": 0.2790082149367025,
+      "grad_norm": 0.10481972247362137,
+      "learning_rate": 2.168389696863501e-05,
+      "loss": 9.4665,
+      "step": 55870
+    },
+    {
+      "epoch": 0.279058153761642,
+      "grad_norm": 0.09700514376163483,
+      "learning_rate": 2.168239505369346e-05,
+      "loss": 9.4596,
+      "step": 55880
+    },
+    {
+      "epoch": 0.2791080925865814,
+      "grad_norm": 0.09476059675216675,
+      "learning_rate": 2.168089313875191e-05,
+      "loss": 9.4671,
+      "step": 55890
+    },
+    {
+      "epoch": 0.2791580314115209,
+      "grad_norm": 0.09556054323911667,
+      "learning_rate": 2.167939122381036e-05,
+      "loss": 9.4661,
+      "step": 55900
+    },
+    {
+      "epoch": 0.2792079702364603,
+      "grad_norm": 0.09373265504837036,
+      "learning_rate": 2.1677889308868807e-05,
+      "loss": 9.4751,
+      "step": 55910
+    },
+    {
+      "epoch": 0.2792579090613998,
+      "grad_norm": 0.10191362351179123,
+      "learning_rate": 2.1676387393927257e-05,
+      "loss": 9.4675,
+      "step": 55920
+    },
+    {
+      "epoch": 0.2793078478863392,
+      "grad_norm": 0.09132136404514313,
+      "learning_rate": 2.1674885478985707e-05,
+      "loss": 9.4671,
+      "step": 55930
+    },
+    {
+      "epoch": 0.2793577867112787,
+      "grad_norm": 0.08985847234725952,
+      "learning_rate": 2.1673383564044157e-05,
+      "loss": 9.4687,
+      "step": 55940
+    },
+    {
+      "epoch": 0.2794077255362181,
+      "grad_norm": 0.10665588080883026,
+      "learning_rate": 2.1671881649102608e-05,
+      "loss": 9.4769,
+      "step": 55950
+    },
+    {
+      "epoch": 0.2794576643611576,
+      "grad_norm": 0.10030438750982285,
+      "learning_rate": 2.1670379734161054e-05,
+      "loss": 9.4622,
+      "step": 55960
+    },
+    {
+      "epoch": 0.279507603186097,
+      "grad_norm": 0.09504380822181702,
+      "learning_rate": 2.1668877819219505e-05,
+      "loss": 9.468,
+      "step": 55970
+    },
+    {
+      "epoch": 0.2795575420110365,
+      "grad_norm": 0.09228894859552383,
+      "learning_rate": 2.1667375904277955e-05,
+      "loss": 9.4615,
+      "step": 55980
+    },
+    {
+      "epoch": 0.2796074808359759,
+      "grad_norm": 0.09327695518732071,
+      "learning_rate": 2.1665873989336405e-05,
+      "loss": 9.4575,
+      "step": 55990
+    },
+    {
+      "epoch": 0.2796574196609154,
+      "grad_norm": 0.09014037996530533,
+      "learning_rate": 2.1664372074394855e-05,
+      "loss": 9.4682,
+      "step": 56000
+    },
+    {
+      "epoch": 0.2797073584858548,
+      "grad_norm": 0.09762946516275406,
+      "learning_rate": 2.1662870159453302e-05,
+      "loss": 9.466,
+      "step": 56010
+    },
+    {
+      "epoch": 0.2797572973107943,
+      "grad_norm": 0.08341635018587112,
+      "learning_rate": 2.1661368244511752e-05,
+      "loss": 9.4691,
+      "step": 56020
+    },
+    {
+      "epoch": 0.2798072361357337,
+      "grad_norm": 0.09523497521877289,
+      "learning_rate": 2.1659866329570202e-05,
+      "loss": 9.4618,
+      "step": 56030
+    },
+    {
+      "epoch": 0.2798571749606732,
+      "grad_norm": 0.10125718265771866,
+      "learning_rate": 2.1658364414628653e-05,
+      "loss": 9.4623,
+      "step": 56040
+    },
+    {
+      "epoch": 0.2799071137856126,
+      "grad_norm": 0.10129445791244507,
+      "learning_rate": 2.1656862499687103e-05,
+      "loss": 9.458,
+      "step": 56050
+    },
+    {
+      "epoch": 0.2799570526105521,
+      "grad_norm": 0.08802901953458786,
+      "learning_rate": 2.1655360584745553e-05,
+      "loss": 9.465,
+      "step": 56060
+    },
+    {
+      "epoch": 0.2800069914354915,
+      "grad_norm": 0.09462311863899231,
+      "learning_rate": 2.1653858669804e-05,
+      "loss": 9.4684,
+      "step": 56070
+    },
+    {
+      "epoch": 0.280056930260431,
+      "grad_norm": 0.09879063069820404,
+      "learning_rate": 2.165235675486245e-05,
+      "loss": 9.4562,
+      "step": 56080
+    },
+    {
+      "epoch": 0.2801068690853704,
+      "grad_norm": 0.09174858778715134,
+      "learning_rate": 2.16508548399209e-05,
+      "loss": 9.4649,
+      "step": 56090
+    },
+    {
+      "epoch": 0.2801568079103099,
+      "grad_norm": 0.09843375533819199,
+      "learning_rate": 2.164935292497935e-05,
+      "loss": 9.4666,
+      "step": 56100
+    },
+    {
+      "epoch": 0.2802067467352493,
+      "grad_norm": 0.09890784323215485,
+      "learning_rate": 2.16478510100378e-05,
+      "loss": 9.464,
+      "step": 56110
+    },
+    {
+      "epoch": 0.2802566855601888,
+      "grad_norm": 0.09890832751989365,
+      "learning_rate": 2.1646349095096247e-05,
+      "loss": 9.47,
+      "step": 56120
+    },
+    {
+      "epoch": 0.2803066243851282,
+      "grad_norm": 0.09250981360673904,
+      "learning_rate": 2.1644847180154697e-05,
+      "loss": 9.4576,
+      "step": 56130
+    },
+    {
+      "epoch": 0.2803565632100677,
+      "grad_norm": 0.09066881984472275,
+      "learning_rate": 2.1643345265213148e-05,
+      "loss": 9.4633,
+      "step": 56140
+    },
+    {
+      "epoch": 0.2804065020350071,
+      "grad_norm": 0.09215233474969864,
+      "learning_rate": 2.1641843350271598e-05,
+      "loss": 9.4647,
+      "step": 56150
+    },
+    {
+      "epoch": 0.2804564408599466,
+      "grad_norm": 0.09696812182664871,
+      "learning_rate": 2.1640341435330048e-05,
+      "loss": 9.4715,
+      "step": 56160
+    },
+    {
+      "epoch": 0.280506379684886,
+      "grad_norm": 0.09409933537244797,
+      "learning_rate": 2.1638839520388495e-05,
+      "loss": 9.4697,
+      "step": 56170
+    },
+    {
+      "epoch": 0.2805563185098255,
+      "grad_norm": 0.09850755333900452,
+      "learning_rate": 2.1637337605446945e-05,
+      "loss": 9.4676,
+      "step": 56180
+    },
+    {
+      "epoch": 0.2806062573347649,
+      "grad_norm": 0.09551995247602463,
+      "learning_rate": 2.1635835690505395e-05,
+      "loss": 9.4628,
+      "step": 56190
+    },
+    {
+      "epoch": 0.2806561961597044,
+      "grad_norm": 0.08868438005447388,
+      "learning_rate": 2.1634333775563845e-05,
+      "loss": 9.4624,
+      "step": 56200
+    },
+    {
+      "epoch": 0.2807061349846438,
+      "grad_norm": 0.09749383479356766,
+      "learning_rate": 2.1632831860622295e-05,
+      "loss": 9.4605,
+      "step": 56210
+    },
+    {
+      "epoch": 0.2807560738095833,
+      "grad_norm": 0.09395034611225128,
+      "learning_rate": 2.1631329945680742e-05,
+      "loss": 9.4619,
+      "step": 56220
+    },
+    {
+      "epoch": 0.2808060126345227,
+      "grad_norm": 0.09494924545288086,
+      "learning_rate": 2.1629828030739192e-05,
+      "loss": 9.4666,
+      "step": 56230
+    },
+    {
+      "epoch": 0.2808559514594622,
+      "grad_norm": 0.09371868520975113,
+      "learning_rate": 2.1628326115797643e-05,
+      "loss": 9.4607,
+      "step": 56240
+    },
+    {
+      "epoch": 0.2809058902844016,
+      "grad_norm": 0.0938858613371849,
+      "learning_rate": 2.1626824200856093e-05,
+      "loss": 9.4649,
+      "step": 56250
+    },
+    {
+      "epoch": 0.2809558291093411,
+      "grad_norm": 0.09127645194530487,
+      "learning_rate": 2.1625322285914543e-05,
+      "loss": 9.4614,
+      "step": 56260
+    },
+    {
+      "epoch": 0.2810057679342805,
+      "grad_norm": 0.09199647605419159,
+      "learning_rate": 2.162382037097299e-05,
+      "loss": 9.4671,
+      "step": 56270
+    },
+    {
+      "epoch": 0.28105570675922,
+      "grad_norm": 0.09332824498414993,
+      "learning_rate": 2.162231845603144e-05,
+      "loss": 9.4663,
+      "step": 56280
+    },
+    {
+      "epoch": 0.2811056455841594,
+      "grad_norm": 0.09723930805921555,
+      "learning_rate": 2.162081654108989e-05,
+      "loss": 9.459,
+      "step": 56290
+    },
+    {
+      "epoch": 0.2811555844090989,
+      "grad_norm": 0.09130632877349854,
+      "learning_rate": 2.161931462614834e-05,
+      "loss": 9.4603,
+      "step": 56300
+    },
+    {
+      "epoch": 0.2812055232340383,
+      "grad_norm": 0.09806390851736069,
+      "learning_rate": 2.161781271120679e-05,
+      "loss": 9.463,
+      "step": 56310
+    },
+    {
+      "epoch": 0.2812554620589778,
+      "grad_norm": 0.09688523411750793,
+      "learning_rate": 2.1616310796265237e-05,
+      "loss": 9.459,
+      "step": 56320
+    },
+    {
+      "epoch": 0.2813054008839172,
+      "grad_norm": 0.09443571418523788,
+      "learning_rate": 2.1614808881323687e-05,
+      "loss": 9.4629,
+      "step": 56330
+    },
+    {
+      "epoch": 0.2813553397088567,
+      "grad_norm": 0.09263693541288376,
+      "learning_rate": 2.1613306966382138e-05,
+      "loss": 9.4674,
+      "step": 56340
+    },
+    {
+      "epoch": 0.2814052785337961,
+      "grad_norm": 0.08993376046419144,
+      "learning_rate": 2.1611805051440588e-05,
+      "loss": 9.462,
+      "step": 56350
+    },
+    {
+      "epoch": 0.2814552173587356,
+      "grad_norm": 0.09446647763252258,
+      "learning_rate": 2.1610303136499038e-05,
+      "loss": 9.4597,
+      "step": 56360
+    },
+    {
+      "epoch": 0.281505156183675,
+      "grad_norm": 0.09347580373287201,
+      "learning_rate": 2.1608801221557485e-05,
+      "loss": 9.4585,
+      "step": 56370
+    },
+    {
+      "epoch": 0.2815550950086144,
+      "grad_norm": 0.09771845489740372,
+      "learning_rate": 2.1607299306615938e-05,
+      "loss": 9.4685,
+      "step": 56380
+    },
+    {
+      "epoch": 0.2816050338335539,
+      "grad_norm": 0.09477143734693527,
+      "learning_rate": 2.1605797391674385e-05,
+      "loss": 9.4516,
+      "step": 56390
+    },
+    {
+      "epoch": 0.2816549726584933,
+      "grad_norm": 0.09481296688318253,
+      "learning_rate": 2.1604295476732835e-05,
+      "loss": 9.4627,
+      "step": 56400
+    },
+    {
+      "epoch": 0.2817049114834328,
+      "grad_norm": 0.09837774932384491,
+      "learning_rate": 2.1602793561791285e-05,
+      "loss": 9.4678,
+      "step": 56410
+    },
+    {
+      "epoch": 0.2817548503083722,
+      "grad_norm": 0.09405245631933212,
+      "learning_rate": 2.1601291646849732e-05,
+      "loss": 9.4549,
+      "step": 56420
+    },
+    {
+      "epoch": 0.2818047891333117,
+      "grad_norm": 0.09251265972852707,
+      "learning_rate": 2.1599789731908186e-05,
+      "loss": 9.466,
+      "step": 56430
+    },
+    {
+      "epoch": 0.2818547279582511,
+      "grad_norm": 0.0892677903175354,
+      "learning_rate": 2.1598287816966633e-05,
+      "loss": 9.4575,
+      "step": 56440
+    },
+    {
+      "epoch": 0.2819046667831906,
+      "grad_norm": 0.09336142987012863,
+      "learning_rate": 2.1596785902025083e-05,
+      "loss": 9.4523,
+      "step": 56450
+    },
+    {
+      "epoch": 0.28195460560813,
+      "grad_norm": 0.09530828148126602,
+      "learning_rate": 2.1595283987083533e-05,
+      "loss": 9.4588,
+      "step": 56460
+    },
+    {
+      "epoch": 0.2820045444330695,
+      "grad_norm": 0.09306281059980392,
+      "learning_rate": 2.159378207214198e-05,
+      "loss": 9.4528,
+      "step": 56470
+    },
+    {
+      "epoch": 0.2820544832580089,
+      "grad_norm": 0.0939321219921112,
+      "learning_rate": 2.1592280157200433e-05,
+      "loss": 9.4545,
+      "step": 56480
+    },
+    {
+      "epoch": 0.2821044220829484,
+      "grad_norm": 0.1002638041973114,
+      "learning_rate": 2.159077824225888e-05,
+      "loss": 9.4582,
+      "step": 56490
+    },
+    {
+      "epoch": 0.2821543609078878,
+      "grad_norm": 0.09518595784902573,
+      "learning_rate": 2.158927632731733e-05,
+      "loss": 9.4531,
+      "step": 56500
+    },
+    {
+      "epoch": 0.2822042997328273,
+      "grad_norm": 0.0943618193268776,
+      "learning_rate": 2.158777441237578e-05,
+      "loss": 9.4467,
+      "step": 56510
+    },
+    {
+      "epoch": 0.2822542385577667,
+      "grad_norm": 0.09317539632320404,
+      "learning_rate": 2.1586272497434227e-05,
+      "loss": 9.4538,
+      "step": 56520
+    },
+    {
+      "epoch": 0.2823041773827062,
+      "grad_norm": 0.09565674513578415,
+      "learning_rate": 2.158477058249268e-05,
+      "loss": 9.457,
+      "step": 56530
+    },
+    {
+      "epoch": 0.2823541162076456,
+      "grad_norm": 0.09357752650976181,
+      "learning_rate": 2.1583268667551128e-05,
+      "loss": 9.4566,
+      "step": 56540
+    },
+    {
+      "epoch": 0.2824040550325851,
+      "grad_norm": 0.09761801362037659,
+      "learning_rate": 2.1581766752609578e-05,
+      "loss": 9.46,
+      "step": 56550
+    },
+    {
+      "epoch": 0.2824539938575245,
+      "grad_norm": 0.08818177133798599,
+      "learning_rate": 2.1580264837668028e-05,
+      "loss": 9.453,
+      "step": 56560
+    },
+    {
+      "epoch": 0.282503932682464,
+      "grad_norm": 0.09338049590587616,
+      "learning_rate": 2.1578762922726475e-05,
+      "loss": 9.4508,
+      "step": 56570
+    },
+    {
+      "epoch": 0.2825538715074034,
+      "grad_norm": 0.09481517225503922,
+      "learning_rate": 2.1577261007784928e-05,
+      "loss": 9.4517,
+      "step": 56580
+    },
+    {
+      "epoch": 0.2826038103323429,
+      "grad_norm": 0.0910092368721962,
+      "learning_rate": 2.1575759092843375e-05,
+      "loss": 9.4635,
+      "step": 56590
+    },
+    {
+      "epoch": 0.2826537491572823,
+      "grad_norm": 0.0879504382610321,
+      "learning_rate": 2.1574257177901825e-05,
+      "loss": 9.4676,
+      "step": 56600
+    },
+    {
+      "epoch": 0.2827036879822218,
+      "grad_norm": 0.09749528020620346,
+      "learning_rate": 2.1572755262960275e-05,
+      "loss": 9.4548,
+      "step": 56610
+    },
+    {
+      "epoch": 0.2827536268071612,
+      "grad_norm": 0.09335489571094513,
+      "learning_rate": 2.1571253348018722e-05,
+      "loss": 9.459,
+      "step": 56620
+    },
+    {
+      "epoch": 0.2828035656321007,
+      "grad_norm": 0.09115227311849594,
+      "learning_rate": 2.1569751433077176e-05,
+      "loss": 9.4516,
+      "step": 56630
+    },
+    {
+      "epoch": 0.2828535044570401,
+      "grad_norm": 0.09703268110752106,
+      "learning_rate": 2.1568249518135623e-05,
+      "loss": 9.4488,
+      "step": 56640
+    },
+    {
+      "epoch": 0.2829034432819796,
+      "grad_norm": 0.08919961750507355,
+      "learning_rate": 2.1566747603194073e-05,
+      "loss": 9.4534,
+      "step": 56650
+    },
+    {
+      "epoch": 0.282953382106919,
+      "grad_norm": 0.09611677378416061,
+      "learning_rate": 2.1565245688252523e-05,
+      "loss": 9.4558,
+      "step": 56660
+    },
+    {
+      "epoch": 0.2830033209318585,
+      "grad_norm": 0.09488637745380402,
+      "learning_rate": 2.156374377331097e-05,
+      "loss": 9.4519,
+      "step": 56670
+    },
+    {
+      "epoch": 0.2830532597567979,
+      "grad_norm": 0.09528210014104843,
+      "learning_rate": 2.1562241858369423e-05,
+      "loss": 9.4565,
+      "step": 56680
+    },
+    {
+      "epoch": 0.2831031985817374,
+      "grad_norm": 0.0950738713145256,
+      "learning_rate": 2.156073994342787e-05,
+      "loss": 9.4551,
+      "step": 56690
+    },
+    {
+      "epoch": 0.2831531374066768,
+      "grad_norm": 0.09509290009737015,
+      "learning_rate": 2.1559238028486324e-05,
+      "loss": 9.4597,
+      "step": 56700
+    },
+    {
+      "epoch": 0.2832030762316163,
+      "grad_norm": 0.09353553503751755,
+      "learning_rate": 2.155773611354477e-05,
+      "loss": 9.443,
+      "step": 56710
+    },
+    {
+      "epoch": 0.2832530150565557,
+      "grad_norm": 0.09463495761156082,
+      "learning_rate": 2.1556234198603217e-05,
+      "loss": 9.454,
+      "step": 56720
+    },
+    {
+      "epoch": 0.2833029538814952,
+      "grad_norm": 0.0871269553899765,
+      "learning_rate": 2.155473228366167e-05,
+      "loss": 9.4553,
+      "step": 56730
+    },
+    {
+      "epoch": 0.2833528927064346,
+      "grad_norm": 0.10174693167209625,
+      "learning_rate": 2.1553230368720118e-05,
+      "loss": 9.4618,
+      "step": 56740
+    },
+    {
+      "epoch": 0.2834028315313741,
+      "grad_norm": 0.09311925619840622,
+      "learning_rate": 2.155172845377857e-05,
+      "loss": 9.4507,
+      "step": 56750
+    },
+    {
+      "epoch": 0.2834527703563135,
+      "grad_norm": 0.09324943274259567,
+      "learning_rate": 2.1550226538837018e-05,
+      "loss": 9.4522,
+      "step": 56760
+    },
+    {
+      "epoch": 0.283502709181253,
+      "grad_norm": 0.0928729698061943,
+      "learning_rate": 2.1548724623895465e-05,
+      "loss": 9.4504,
+      "step": 56770
+    },
+    {
+      "epoch": 0.2835526480061924,
+      "grad_norm": 0.09641097486019135,
+      "learning_rate": 2.1547222708953918e-05,
+      "loss": 9.4463,
+      "step": 56780
+    },
+    {
+      "epoch": 0.2836025868311319,
+      "grad_norm": 0.09734059870243073,
+      "learning_rate": 2.1545720794012365e-05,
+      "loss": 9.4444,
+      "step": 56790
+    },
+    {
+      "epoch": 0.2836525256560713,
+      "grad_norm": 0.09037147462368011,
+      "learning_rate": 2.154421887907082e-05,
+      "loss": 9.4487,
+      "step": 56800
+    },
+    {
+      "epoch": 0.2837024644810108,
+      "grad_norm": 0.09848001599311829,
+      "learning_rate": 2.1542716964129265e-05,
+      "loss": 9.4588,
+      "step": 56810
+    },
+    {
+      "epoch": 0.2837524033059502,
+      "grad_norm": 0.092610202729702,
+      "learning_rate": 2.1541215049187712e-05,
+      "loss": 9.453,
+      "step": 56820
+    },
+    {
+      "epoch": 0.2838023421308897,
+      "grad_norm": 0.09564469754695892,
+      "learning_rate": 2.1539713134246166e-05,
+      "loss": 9.4589,
+      "step": 56830
+    },
+    {
+      "epoch": 0.2838522809558291,
+      "grad_norm": 0.08878234773874283,
+      "learning_rate": 2.1538211219304613e-05,
+      "loss": 9.4545,
+      "step": 56840
+    },
+    {
+      "epoch": 0.28390221978076857,
+      "grad_norm": 0.09450232982635498,
+      "learning_rate": 2.1536709304363066e-05,
+      "loss": 9.4541,
+      "step": 56850
+    },
+    {
+      "epoch": 0.283952158605708,
+      "grad_norm": 0.09449663758277893,
+      "learning_rate": 2.1535207389421513e-05,
+      "loss": 9.4419,
+      "step": 56860
+    },
+    {
+      "epoch": 0.28400209743064747,
+      "grad_norm": 0.09098829329013824,
+      "learning_rate": 2.153370547447996e-05,
+      "loss": 9.452,
+      "step": 56870
+    },
+    {
+      "epoch": 0.2840520362555869,
+      "grad_norm": 0.09370938688516617,
+      "learning_rate": 2.1532203559538413e-05,
+      "loss": 9.45,
+      "step": 56880
+    },
+    {
+      "epoch": 0.28410197508052637,
+      "grad_norm": 0.09251664578914642,
+      "learning_rate": 2.153070164459686e-05,
+      "loss": 9.453,
+      "step": 56890
+    },
+    {
+      "epoch": 0.2841519139054658,
+      "grad_norm": 0.0982016995549202,
+      "learning_rate": 2.1529199729655314e-05,
+      "loss": 9.4479,
+      "step": 56900
+    },
+    {
+      "epoch": 0.28420185273040527,
+      "grad_norm": 0.0964377224445343,
+      "learning_rate": 2.152769781471376e-05,
+      "loss": 9.4496,
+      "step": 56910
+    },
+    {
+      "epoch": 0.2842517915553447,
+      "grad_norm": 0.09132152795791626,
+      "learning_rate": 2.1526195899772207e-05,
+      "loss": 9.4547,
+      "step": 56920
+    },
+    {
+      "epoch": 0.28430173038028417,
+      "grad_norm": 0.09878633171319962,
+      "learning_rate": 2.152469398483066e-05,
+      "loss": 9.4533,
+      "step": 56930
+    },
+    {
+      "epoch": 0.2843516692052236,
+      "grad_norm": 0.09209530800580978,
+      "learning_rate": 2.1523192069889108e-05,
+      "loss": 9.4452,
+      "step": 56940
+    },
+    {
+      "epoch": 0.28440160803016307,
+      "grad_norm": 0.09105639159679413,
+      "learning_rate": 2.152169015494756e-05,
+      "loss": 9.4402,
+      "step": 56950
+    },
+    {
+      "epoch": 0.2844515468551025,
+      "grad_norm": 0.09028221666812897,
+      "learning_rate": 2.1520188240006008e-05,
+      "loss": 9.4486,
+      "step": 56960
+    },
+    {
+      "epoch": 0.28450148568004197,
+      "grad_norm": 0.09749405086040497,
+      "learning_rate": 2.1518686325064455e-05,
+      "loss": 9.452,
+      "step": 56970
+    },
+    {
+      "epoch": 0.2845514245049814,
+      "grad_norm": 0.0886070504784584,
+      "learning_rate": 2.1517184410122908e-05,
+      "loss": 9.4518,
+      "step": 56980
+    },
+    {
+      "epoch": 0.28460136332992086,
+      "grad_norm": 0.09163407981395721,
+      "learning_rate": 2.1515682495181355e-05,
+      "loss": 9.4533,
+      "step": 56990
+    },
+    {
+      "epoch": 0.2846513021548603,
+      "grad_norm": 0.09811826795339584,
+      "learning_rate": 2.151418058023981e-05,
+      "loss": 9.4422,
+      "step": 57000
+    },
+    {
+      "epoch": 0.28470124097979976,
+      "grad_norm": 0.08839601278305054,
+      "learning_rate": 2.1512678665298255e-05,
+      "loss": 9.4462,
+      "step": 57010
+    },
+    {
+      "epoch": 0.2847511798047392,
+      "grad_norm": 0.09699650853872299,
+      "learning_rate": 2.1511176750356706e-05,
+      "loss": 9.4547,
+      "step": 57020
+    },
+    {
+      "epoch": 0.28480111862967866,
+      "grad_norm": 0.0974980965256691,
+      "learning_rate": 2.1509674835415156e-05,
+      "loss": 9.4466,
+      "step": 57030
+    },
+    {
+      "epoch": 0.2848510574546181,
+      "grad_norm": 0.09446827322244644,
+      "learning_rate": 2.1508172920473603e-05,
+      "loss": 9.4488,
+      "step": 57040
+    },
+    {
+      "epoch": 0.28490099627955756,
+      "grad_norm": 0.09437047690153122,
+      "learning_rate": 2.1506671005532056e-05,
+      "loss": 9.4458,
+      "step": 57050
+    },
+    {
+      "epoch": 0.284950935104497,
+      "grad_norm": 0.09088893234729767,
+      "learning_rate": 2.1505169090590503e-05,
+      "loss": 9.4443,
+      "step": 57060
+    },
+    {
+      "epoch": 0.28500087392943646,
+      "grad_norm": 0.0951780378818512,
+      "learning_rate": 2.1503667175648953e-05,
+      "loss": 9.4552,
+      "step": 57070
+    },
+    {
+      "epoch": 0.2850508127543759,
+      "grad_norm": 0.08948690444231033,
+      "learning_rate": 2.1502165260707403e-05,
+      "loss": 9.4513,
+      "step": 57080
+    },
+    {
+      "epoch": 0.28510075157931536,
+      "grad_norm": 0.09669423848390579,
+      "learning_rate": 2.150066334576585e-05,
+      "loss": 9.4471,
+      "step": 57090
+    },
+    {
+      "epoch": 0.2851506904042548,
+      "grad_norm": 0.09193644672632217,
+      "learning_rate": 2.1499161430824304e-05,
+      "loss": 9.4458,
+      "step": 57100
+    },
+    {
+      "epoch": 0.28520062922919426,
+      "grad_norm": 0.10340452194213867,
+      "learning_rate": 2.149765951588275e-05,
+      "loss": 9.4497,
+      "step": 57110
+    },
+    {
+      "epoch": 0.2852505680541337,
+      "grad_norm": 0.09467227011919022,
+      "learning_rate": 2.14961576009412e-05,
+      "loss": 9.4555,
+      "step": 57120
+    },
+    {
+      "epoch": 0.28530050687907316,
+      "grad_norm": 0.09359690546989441,
+      "learning_rate": 2.149465568599965e-05,
+      "loss": 9.4468,
+      "step": 57130
+    },
+    {
+      "epoch": 0.2853504457040126,
+      "grad_norm": 0.08810552209615707,
+      "learning_rate": 2.1493153771058098e-05,
+      "loss": 9.4406,
+      "step": 57140
+    },
+    {
+      "epoch": 0.28540038452895206,
+      "grad_norm": 0.09267257899045944,
+      "learning_rate": 2.149165185611655e-05,
+      "loss": 9.4496,
+      "step": 57150
+    },
+    {
+      "epoch": 0.2854503233538915,
+      "grad_norm": 0.09369373321533203,
+      "learning_rate": 2.1490149941174998e-05,
+      "loss": 9.4524,
+      "step": 57160
+    },
+    {
+      "epoch": 0.28550026217883095,
+      "grad_norm": 0.09611758589744568,
+      "learning_rate": 2.1488648026233448e-05,
+      "loss": 9.446,
+      "step": 57170
+    },
+    {
+      "epoch": 0.2855502010037704,
+      "grad_norm": 0.09398515522480011,
+      "learning_rate": 2.1487146111291898e-05,
+      "loss": 9.4492,
+      "step": 57180
+    },
+    {
+      "epoch": 0.28560013982870985,
+      "grad_norm": 0.08984824270009995,
+      "learning_rate": 2.1485644196350345e-05,
+      "loss": 9.4438,
+      "step": 57190
+    },
+    {
+      "epoch": 0.2856500786536493,
+      "grad_norm": 0.09207414090633392,
+      "learning_rate": 2.14841422814088e-05,
+      "loss": 9.4522,
+      "step": 57200
+    },
+    {
+      "epoch": 0.28570001747858875,
+      "grad_norm": 0.09611576050519943,
+      "learning_rate": 2.1482640366467245e-05,
+      "loss": 9.4511,
+      "step": 57210
+    },
+    {
+      "epoch": 0.2857499563035282,
+      "grad_norm": 0.09785609692335129,
+      "learning_rate": 2.1481138451525696e-05,
+      "loss": 9.4475,
+      "step": 57220
+    },
+    {
+      "epoch": 0.28579989512846765,
+      "grad_norm": 0.09928670525550842,
+      "learning_rate": 2.1479636536584146e-05,
+      "loss": 9.4436,
+      "step": 57230
+    },
+    {
+      "epoch": 0.2858498339534071,
+      "grad_norm": 0.09629528224468231,
+      "learning_rate": 2.1478134621642593e-05,
+      "loss": 9.4383,
+      "step": 57240
+    },
+    {
+      "epoch": 0.28589977277834655,
+      "grad_norm": 0.09739939123392105,
+      "learning_rate": 2.1476632706701046e-05,
+      "loss": 9.447,
+      "step": 57250
+    },
+    {
+      "epoch": 0.28594971160328597,
+      "grad_norm": 0.09917205572128296,
+      "learning_rate": 2.1475130791759493e-05,
+      "loss": 9.4335,
+      "step": 57260
+    },
+    {
+      "epoch": 0.28599965042822545,
+      "grad_norm": 0.09072194248437881,
+      "learning_rate": 2.1473628876817943e-05,
+      "loss": 9.4544,
+      "step": 57270
+    },
+    {
+      "epoch": 0.28604958925316487,
+      "grad_norm": 0.09361054748296738,
+      "learning_rate": 2.1472126961876393e-05,
+      "loss": 9.4429,
+      "step": 57280
+    },
+    {
+      "epoch": 0.28609952807810435,
+      "grad_norm": 0.09945287555456161,
+      "learning_rate": 2.147062504693484e-05,
+      "loss": 9.4512,
+      "step": 57290
+    },
+    {
+      "epoch": 0.28614946690304377,
+      "grad_norm": 0.09822694212198257,
+      "learning_rate": 2.1469123131993294e-05,
+      "loss": 9.434,
+      "step": 57300
+    },
+    {
+      "epoch": 0.28619940572798325,
+      "grad_norm": 0.09281609952449799,
+      "learning_rate": 2.146762121705174e-05,
+      "loss": 9.4456,
+      "step": 57310
+    },
+    {
+      "epoch": 0.28624934455292267,
+      "grad_norm": 0.09337762743234634,
+      "learning_rate": 2.146611930211019e-05,
+      "loss": 9.4507,
+      "step": 57320
+    },
+    {
+      "epoch": 0.28629928337786215,
+      "grad_norm": 0.09160379320383072,
+      "learning_rate": 2.146461738716864e-05,
+      "loss": 9.4441,
+      "step": 57330
+    },
+    {
+      "epoch": 0.28634922220280157,
+      "grad_norm": 0.09638303518295288,
+      "learning_rate": 2.146311547222709e-05,
+      "loss": 9.45,
+      "step": 57340
+    },
+    {
+      "epoch": 0.28639916102774104,
+      "grad_norm": 0.0943082943558693,
+      "learning_rate": 2.146161355728554e-05,
+      "loss": 9.4481,
+      "step": 57350
+    },
+    {
+      "epoch": 0.28644909985268047,
+      "grad_norm": 0.09442918747663498,
+      "learning_rate": 2.1460111642343988e-05,
+      "loss": 9.4396,
+      "step": 57360
+    },
+    {
+      "epoch": 0.2864990386776199,
+      "grad_norm": 0.10041014105081558,
+      "learning_rate": 2.1458609727402438e-05,
+      "loss": 9.4372,
+      "step": 57370
+    },
+    {
+      "epoch": 0.28654897750255937,
+      "grad_norm": 0.0975830927491188,
+      "learning_rate": 2.145710781246089e-05,
+      "loss": 9.4574,
+      "step": 57380
+    },
+    {
+      "epoch": 0.2865989163274988,
+      "grad_norm": 0.0888514295220375,
+      "learning_rate": 2.145560589751934e-05,
+      "loss": 9.4463,
+      "step": 57390
+    },
+    {
+      "epoch": 0.28664885515243826,
+      "grad_norm": 0.09187906235456467,
+      "learning_rate": 2.145410398257779e-05,
+      "loss": 9.4551,
+      "step": 57400
+    },
+    {
+      "epoch": 0.2866987939773777,
+      "grad_norm": 0.09454265236854553,
+      "learning_rate": 2.1452602067636235e-05,
+      "loss": 9.4473,
+      "step": 57410
+    },
+    {
+      "epoch": 0.28674873280231716,
+      "grad_norm": 0.09518423676490784,
+      "learning_rate": 2.1451100152694686e-05,
+      "loss": 9.4421,
+      "step": 57420
+    },
+    {
+      "epoch": 0.2867986716272566,
+      "grad_norm": 0.09494660049676895,
+      "learning_rate": 2.1449598237753136e-05,
+      "loss": 9.4452,
+      "step": 57430
+    },
+    {
+      "epoch": 0.28684861045219606,
+      "grad_norm": 0.09462330490350723,
+      "learning_rate": 2.1448096322811586e-05,
+      "loss": 9.444,
+      "step": 57440
+    },
+    {
+      "epoch": 0.2868985492771355,
+      "grad_norm": 0.09740941226482391,
+      "learning_rate": 2.1446594407870036e-05,
+      "loss": 9.4363,
+      "step": 57450
+    },
+    {
+      "epoch": 0.28694848810207496,
+      "grad_norm": 0.09297572076320648,
+      "learning_rate": 2.1445092492928483e-05,
+      "loss": 9.4484,
+      "step": 57460
+    },
+    {
+      "epoch": 0.2869984269270144,
+      "grad_norm": 0.09299388527870178,
+      "learning_rate": 2.1443590577986933e-05,
+      "loss": 9.4431,
+      "step": 57470
+    },
+    {
+      "epoch": 0.28704836575195386,
+      "grad_norm": 0.09611527621746063,
+      "learning_rate": 2.1442088663045383e-05,
+      "loss": 9.4422,
+      "step": 57480
+    },
+    {
+      "epoch": 0.2870983045768933,
+      "grad_norm": 0.09383988380432129,
+      "learning_rate": 2.1440586748103833e-05,
+      "loss": 9.4434,
+      "step": 57490
+    },
+    {
+      "epoch": 0.28714824340183276,
+      "grad_norm": 0.09201846271753311,
+      "learning_rate": 2.1439084833162284e-05,
+      "loss": 9.4427,
+      "step": 57500
+    },
+    {
+      "epoch": 0.2871981822267722,
+      "grad_norm": 0.09528767317533493,
+      "learning_rate": 2.143758291822073e-05,
+      "loss": 9.4362,
+      "step": 57510
+    },
+    {
+      "epoch": 0.28724812105171166,
+      "grad_norm": 0.09210474044084549,
+      "learning_rate": 2.143608100327918e-05,
+      "loss": 9.4403,
+      "step": 57520
+    },
+    {
+      "epoch": 0.2872980598766511,
+      "grad_norm": 0.09716343879699707,
+      "learning_rate": 2.143457908833763e-05,
+      "loss": 9.4486,
+      "step": 57530
+    },
+    {
+      "epoch": 0.28734799870159056,
+      "grad_norm": 0.0949615091085434,
+      "learning_rate": 2.143307717339608e-05,
+      "loss": 9.4341,
+      "step": 57540
+    },
+    {
+      "epoch": 0.28739793752653,
+      "grad_norm": 0.09865114837884903,
+      "learning_rate": 2.143157525845453e-05,
+      "loss": 9.4345,
+      "step": 57550
+    },
+    {
+      "epoch": 0.28744787635146946,
+      "grad_norm": 0.09301295876502991,
+      "learning_rate": 2.1430073343512978e-05,
+      "loss": 9.4398,
+      "step": 57560
+    },
+    {
+      "epoch": 0.2874978151764089,
+      "grad_norm": 0.09104243665933609,
+      "learning_rate": 2.1428571428571428e-05,
+      "loss": 9.4445,
+      "step": 57570
+    },
+    {
+      "epoch": 0.28754775400134835,
+      "grad_norm": 0.09278678148984909,
+      "learning_rate": 2.142706951362988e-05,
+      "loss": 9.4513,
+      "step": 57580
+    },
+    {
+      "epoch": 0.2875976928262878,
+      "grad_norm": 0.0937771275639534,
+      "learning_rate": 2.142556759868833e-05,
+      "loss": 9.4393,
+      "step": 57590
+    },
+    {
+      "epoch": 0.28764763165122725,
+      "grad_norm": 0.09479419887065887,
+      "learning_rate": 2.142406568374678e-05,
+      "loss": 9.4373,
+      "step": 57600
+    },
+    {
+      "epoch": 0.2876975704761667,
+      "grad_norm": 0.09332539886236191,
+      "learning_rate": 2.1422563768805225e-05,
+      "loss": 9.443,
+      "step": 57610
+    },
+    {
+      "epoch": 0.28774750930110615,
+      "grad_norm": 0.09496680647134781,
+      "learning_rate": 2.1421061853863676e-05,
+      "loss": 9.4405,
+      "step": 57620
+    },
+    {
+      "epoch": 0.2877974481260456,
+      "grad_norm": 0.09168408811092377,
+      "learning_rate": 2.1419559938922126e-05,
+      "loss": 9.4475,
+      "step": 57630
+    },
+    {
+      "epoch": 0.28784738695098505,
+      "grad_norm": 0.08694005757570267,
+      "learning_rate": 2.1418058023980576e-05,
+      "loss": 9.4491,
+      "step": 57640
+    },
+    {
+      "epoch": 0.2878973257759245,
+      "grad_norm": 0.0956626832485199,
+      "learning_rate": 2.1416556109039026e-05,
+      "loss": 9.4481,
+      "step": 57650
+    },
+    {
+      "epoch": 0.28794726460086395,
+      "grad_norm": 0.09360313415527344,
+      "learning_rate": 2.1415054194097476e-05,
+      "loss": 9.438,
+      "step": 57660
+    },
+    {
+      "epoch": 0.28799720342580337,
+      "grad_norm": 0.09131594747304916,
+      "learning_rate": 2.1413552279155923e-05,
+      "loss": 9.4273,
+      "step": 57670
+    },
+    {
+      "epoch": 0.28804714225074285,
+      "grad_norm": 0.09347652643918991,
+      "learning_rate": 2.1412050364214373e-05,
+      "loss": 9.4416,
+      "step": 57680
+    },
+    {
+      "epoch": 0.28809708107568227,
+      "grad_norm": 0.09475225955247879,
+      "learning_rate": 2.1410548449272824e-05,
+      "loss": 9.431,
+      "step": 57690
+    },
+    {
+      "epoch": 0.28814701990062175,
+      "grad_norm": 0.0879056304693222,
+      "learning_rate": 2.1409046534331274e-05,
+      "loss": 9.4484,
+      "step": 57700
+    },
+    {
+      "epoch": 0.28819695872556117,
+      "grad_norm": 0.090447798371315,
+      "learning_rate": 2.1407544619389724e-05,
+      "loss": 9.4465,
+      "step": 57710
+    },
+    {
+      "epoch": 0.28824689755050065,
+      "grad_norm": 0.09707989543676376,
+      "learning_rate": 2.140604270444817e-05,
+      "loss": 9.4368,
+      "step": 57720
+    },
+    {
+      "epoch": 0.28829683637544007,
+      "grad_norm": 0.09571398794651031,
+      "learning_rate": 2.140454078950662e-05,
+      "loss": 9.4407,
+      "step": 57730
+    },
+    {
+      "epoch": 0.28834677520037955,
+      "grad_norm": 0.09500686079263687,
+      "learning_rate": 2.140303887456507e-05,
+      "loss": 9.4291,
+      "step": 57740
+    },
+    {
+      "epoch": 0.28839671402531897,
+      "grad_norm": 0.09296731650829315,
+      "learning_rate": 2.140153695962352e-05,
+      "loss": 9.4408,
+      "step": 57750
+    },
+    {
+      "epoch": 0.28844665285025844,
+      "grad_norm": 0.0914110615849495,
+      "learning_rate": 2.140003504468197e-05,
+      "loss": 9.4334,
+      "step": 57760
+    },
+    {
+      "epoch": 0.28849659167519787,
+      "grad_norm": 0.09195576608181,
+      "learning_rate": 2.1398533129740418e-05,
+      "loss": 9.4371,
+      "step": 57770
+    },
+    {
+      "epoch": 0.28854653050013734,
+      "grad_norm": 0.09556440263986588,
+      "learning_rate": 2.139703121479887e-05,
+      "loss": 9.4361,
+      "step": 57780
+    },
+    {
+      "epoch": 0.28859646932507677,
+      "grad_norm": 0.09003462642431259,
+      "learning_rate": 2.139552929985732e-05,
+      "loss": 9.4242,
+      "step": 57790
+    },
+    {
+      "epoch": 0.28864640815001624,
+      "grad_norm": 0.09391631186008453,
+      "learning_rate": 2.139402738491577e-05,
+      "loss": 9.4349,
+      "step": 57800
+    },
+    {
+      "epoch": 0.28869634697495566,
+      "grad_norm": 0.09652207046747208,
+      "learning_rate": 2.139252546997422e-05,
+      "loss": 9.4379,
+      "step": 57810
+    },
+    {
+      "epoch": 0.28874628579989514,
+      "grad_norm": 0.09347474575042725,
+      "learning_rate": 2.1391023555032666e-05,
+      "loss": 9.4349,
+      "step": 57820
+    },
+    {
+      "epoch": 0.28879622462483456,
+      "grad_norm": 0.09264619648456573,
+      "learning_rate": 2.1389521640091116e-05,
+      "loss": 9.4356,
+      "step": 57830
+    },
+    {
+      "epoch": 0.28884616344977404,
+      "grad_norm": 0.09425893425941467,
+      "learning_rate": 2.1388019725149566e-05,
+      "loss": 9.4268,
+      "step": 57840
+    },
+    {
+      "epoch": 0.28889610227471346,
+      "grad_norm": 0.09370515495538712,
+      "learning_rate": 2.1386517810208016e-05,
+      "loss": 9.4335,
+      "step": 57850
+    },
+    {
+      "epoch": 0.28894604109965294,
+      "grad_norm": 0.09193460643291473,
+      "learning_rate": 2.1385015895266466e-05,
+      "loss": 9.4294,
+      "step": 57860
+    },
+    {
+      "epoch": 0.28899597992459236,
+      "grad_norm": 0.09023167192935944,
+      "learning_rate": 2.1383513980324913e-05,
+      "loss": 9.4308,
+      "step": 57870
+    },
+    {
+      "epoch": 0.28904591874953184,
+      "grad_norm": 0.09350664168596268,
+      "learning_rate": 2.1382012065383363e-05,
+      "loss": 9.4267,
+      "step": 57880
+    },
+    {
+      "epoch": 0.28909585757447126,
+      "grad_norm": 0.09755279123783112,
+      "learning_rate": 2.1380510150441814e-05,
+      "loss": 9.4382,
+      "step": 57890
+    },
+    {
+      "epoch": 0.28914579639941074,
+      "grad_norm": 0.09248659759759903,
+      "learning_rate": 2.1379008235500264e-05,
+      "loss": 9.4328,
+      "step": 57900
+    },
+    {
+      "epoch": 0.28919573522435016,
+      "grad_norm": 0.09720529615879059,
+      "learning_rate": 2.1377506320558714e-05,
+      "loss": 9.4405,
+      "step": 57910
+    },
+    {
+      "epoch": 0.28924567404928964,
+      "grad_norm": 0.09519968181848526,
+      "learning_rate": 2.137600440561716e-05,
+      "loss": 9.4223,
+      "step": 57920
+    },
+    {
+      "epoch": 0.28929561287422906,
+      "grad_norm": 0.09368513524532318,
+      "learning_rate": 2.137450249067561e-05,
+      "loss": 9.4369,
+      "step": 57930
+    },
+    {
+      "epoch": 0.28934555169916854,
+      "grad_norm": 0.09568150341510773,
+      "learning_rate": 2.137300057573406e-05,
+      "loss": 9.4303,
+      "step": 57940
+    },
+    {
+      "epoch": 0.28939549052410796,
+      "grad_norm": 0.09773694723844528,
+      "learning_rate": 2.137149866079251e-05,
+      "loss": 9.4328,
+      "step": 57950
+    },
+    {
+      "epoch": 0.28944542934904743,
+      "grad_norm": 0.0943894013762474,
+      "learning_rate": 2.136999674585096e-05,
+      "loss": 9.4307,
+      "step": 57960
+    },
+    {
+      "epoch": 0.28949536817398686,
+      "grad_norm": 0.09780600666999817,
+      "learning_rate": 2.1368494830909408e-05,
+      "loss": 9.4274,
+      "step": 57970
+    },
+    {
+      "epoch": 0.28954530699892633,
+      "grad_norm": 0.09167556464672089,
+      "learning_rate": 2.1366992915967862e-05,
+      "loss": 9.4372,
+      "step": 57980
+    },
+    {
+      "epoch": 0.28959524582386575,
+      "grad_norm": 0.08926273137331009,
+      "learning_rate": 2.136549100102631e-05,
+      "loss": 9.4382,
+      "step": 57990
+    },
+    {
+      "epoch": 0.28964518464880523,
+      "grad_norm": 0.09246543794870377,
+      "learning_rate": 2.136398908608476e-05,
+      "loss": 9.4295,
+      "step": 58000
+    },
+    {
+      "epoch": 0.28969512347374465,
+      "grad_norm": 0.10091344267129898,
+      "learning_rate": 2.136248717114321e-05,
+      "loss": 9.4334,
+      "step": 58010
+    },
+    {
+      "epoch": 0.28974506229868413,
+      "grad_norm": 0.10019103437662125,
+      "learning_rate": 2.1360985256201656e-05,
+      "loss": 9.443,
+      "step": 58020
+    },
+    {
+      "epoch": 0.28979500112362355,
+      "grad_norm": 0.09648888558149338,
+      "learning_rate": 2.135948334126011e-05,
+      "loss": 9.4307,
+      "step": 58030
+    },
+    {
+      "epoch": 0.28984493994856303,
+      "grad_norm": 0.09710298478603363,
+      "learning_rate": 2.1357981426318556e-05,
+      "loss": 9.4356,
+      "step": 58040
+    },
+    {
+      "epoch": 0.28989487877350245,
+      "grad_norm": 0.0928901955485344,
+      "learning_rate": 2.1356479511377006e-05,
+      "loss": 9.4327,
+      "step": 58050
+    },
+    {
+      "epoch": 0.28994481759844193,
+      "grad_norm": 0.09765011072158813,
+      "learning_rate": 2.1354977596435456e-05,
+      "loss": 9.4317,
+      "step": 58060
+    },
+    {
+      "epoch": 0.28999475642338135,
+      "grad_norm": 0.08833613991737366,
+      "learning_rate": 2.1353475681493903e-05,
+      "loss": 9.4366,
+      "step": 58070
+    },
+    {
+      "epoch": 0.2900446952483208,
+      "grad_norm": 0.09767667204141617,
+      "learning_rate": 2.1351973766552357e-05,
+      "loss": 9.4237,
+      "step": 58080
+    },
+    {
+      "epoch": 0.29009463407326025,
+      "grad_norm": 0.09400112926959991,
+      "learning_rate": 2.1350471851610804e-05,
+      "loss": 9.4363,
+      "step": 58090
+    },
+    {
+      "epoch": 0.2901445728981997,
+      "grad_norm": 0.09494834393262863,
+      "learning_rate": 2.1348969936669254e-05,
+      "loss": 9.4317,
+      "step": 58100
+    },
+    {
+      "epoch": 0.29019451172313915,
+      "grad_norm": 0.08743976801633835,
+      "learning_rate": 2.1347468021727704e-05,
+      "loss": 9.4312,
+      "step": 58110
+    },
+    {
+      "epoch": 0.2902444505480786,
+      "grad_norm": 0.0986538678407669,
+      "learning_rate": 2.134596610678615e-05,
+      "loss": 9.4404,
+      "step": 58120
+    },
+    {
+      "epoch": 0.29029438937301805,
+      "grad_norm": 0.09177923202514648,
+      "learning_rate": 2.1344464191844604e-05,
+      "loss": 9.4274,
+      "step": 58130
+    },
+    {
+      "epoch": 0.2903443281979575,
+      "grad_norm": 0.09830587357282639,
+      "learning_rate": 2.134296227690305e-05,
+      "loss": 9.4314,
+      "step": 58140
+    },
+    {
+      "epoch": 0.29039426702289695,
+      "grad_norm": 0.09088832139968872,
+      "learning_rate": 2.13414603619615e-05,
+      "loss": 9.4367,
+      "step": 58150
+    },
+    {
+      "epoch": 0.2904442058478364,
+      "grad_norm": 0.0943973958492279,
+      "learning_rate": 2.133995844701995e-05,
+      "loss": 9.4372,
+      "step": 58160
+    },
+    {
+      "epoch": 0.29049414467277584,
+      "grad_norm": 0.09387973695993423,
+      "learning_rate": 2.1338456532078398e-05,
+      "loss": 9.4323,
+      "step": 58170
+    },
+    {
+      "epoch": 0.2905440834977153,
+      "grad_norm": 0.0891706719994545,
+      "learning_rate": 2.1336954617136852e-05,
+      "loss": 9.4364,
+      "step": 58180
+    },
+    {
+      "epoch": 0.29059402232265474,
+      "grad_norm": 0.09315983951091766,
+      "learning_rate": 2.13354527021953e-05,
+      "loss": 9.428,
+      "step": 58190
+    },
+    {
+      "epoch": 0.2906439611475942,
+      "grad_norm": 0.09957679361104965,
+      "learning_rate": 2.133395078725375e-05,
+      "loss": 9.4258,
+      "step": 58200
+    },
+    {
+      "epoch": 0.29069389997253364,
+      "grad_norm": 0.0999840795993805,
+      "learning_rate": 2.13324488723122e-05,
+      "loss": 9.4291,
+      "step": 58210
+    },
+    {
+      "epoch": 0.2907438387974731,
+      "grad_norm": 0.09671057015657425,
+      "learning_rate": 2.1330946957370646e-05,
+      "loss": 9.4328,
+      "step": 58220
+    },
+    {
+      "epoch": 0.29079377762241254,
+      "grad_norm": 0.0913921594619751,
+      "learning_rate": 2.13294450424291e-05,
+      "loss": 9.4308,
+      "step": 58230
+    },
+    {
+      "epoch": 0.290843716447352,
+      "grad_norm": 0.09564314782619476,
+      "learning_rate": 2.1327943127487546e-05,
+      "loss": 9.4269,
+      "step": 58240
+    },
+    {
+      "epoch": 0.29089365527229144,
+      "grad_norm": 0.09441643208265305,
+      "learning_rate": 2.1326441212545996e-05,
+      "loss": 9.4385,
+      "step": 58250
+    },
+    {
+      "epoch": 0.2909435940972309,
+      "grad_norm": 0.0950765311717987,
+      "learning_rate": 2.1324939297604446e-05,
+      "loss": 9.4394,
+      "step": 58260
+    },
+    {
+      "epoch": 0.29099353292217034,
+      "grad_norm": 0.09390682727098465,
+      "learning_rate": 2.1323437382662893e-05,
+      "loss": 9.4312,
+      "step": 58270
+    },
+    {
+      "epoch": 0.2910434717471098,
+      "grad_norm": 0.09104769676923752,
+      "learning_rate": 2.1321935467721347e-05,
+      "loss": 9.4247,
+      "step": 58280
+    },
+    {
+      "epoch": 0.29109341057204924,
+      "grad_norm": 0.10046746581792831,
+      "learning_rate": 2.1320433552779794e-05,
+      "loss": 9.4248,
+      "step": 58290
+    },
+    {
+      "epoch": 0.2911433493969887,
+      "grad_norm": 0.09288429468870163,
+      "learning_rate": 2.1318931637838247e-05,
+      "loss": 9.4221,
+      "step": 58300
+    },
+    {
+      "epoch": 0.29119328822192814,
+      "grad_norm": 0.09124504029750824,
+      "learning_rate": 2.1317429722896694e-05,
+      "loss": 9.4278,
+      "step": 58310
+    },
+    {
+      "epoch": 0.2912432270468676,
+      "grad_norm": 0.09842488169670105,
+      "learning_rate": 2.131592780795514e-05,
+      "loss": 9.4219,
+      "step": 58320
+    },
+    {
+      "epoch": 0.29129316587180704,
+      "grad_norm": 0.09622178226709366,
+      "learning_rate": 2.1314425893013594e-05,
+      "loss": 9.4214,
+      "step": 58330
+    },
+    {
+      "epoch": 0.2913431046967465,
+      "grad_norm": 0.09460314363241196,
+      "learning_rate": 2.131292397807204e-05,
+      "loss": 9.4139,
+      "step": 58340
+    },
+    {
+      "epoch": 0.29139304352168593,
+      "grad_norm": 0.10045454651117325,
+      "learning_rate": 2.1311422063130495e-05,
+      "loss": 9.4236,
+      "step": 58350
+    },
+    {
+      "epoch": 0.29144298234662536,
+      "grad_norm": 0.09059098362922668,
+      "learning_rate": 2.130992014818894e-05,
+      "loss": 9.4255,
+      "step": 58360
+    },
+    {
+      "epoch": 0.29149292117156483,
+      "grad_norm": 0.09203209728002548,
+      "learning_rate": 2.1308418233247388e-05,
+      "loss": 9.4297,
+      "step": 58370
+    },
+    {
+      "epoch": 0.29154285999650426,
+      "grad_norm": 0.08972401171922684,
+      "learning_rate": 2.1306916318305842e-05,
+      "loss": 9.4402,
+      "step": 58380
+    },
+    {
+      "epoch": 0.29159279882144373,
+      "grad_norm": 0.09233536571264267,
+      "learning_rate": 2.130541440336429e-05,
+      "loss": 9.4283,
+      "step": 58390
+    },
+    {
+      "epoch": 0.29164273764638315,
+      "grad_norm": 0.10041224211454391,
+      "learning_rate": 2.1303912488422742e-05,
+      "loss": 9.425,
+      "step": 58400
+    },
+    {
+      "epoch": 0.29169267647132263,
+      "grad_norm": 0.09314385056495667,
+      "learning_rate": 2.130241057348119e-05,
+      "loss": 9.4202,
+      "step": 58410
+    },
+    {
+      "epoch": 0.29174261529626205,
+      "grad_norm": 0.09177546203136444,
+      "learning_rate": 2.1300908658539636e-05,
+      "loss": 9.4319,
+      "step": 58420
+    },
+    {
+      "epoch": 0.29179255412120153,
+      "grad_norm": 0.09403910487890244,
+      "learning_rate": 2.129940674359809e-05,
+      "loss": 9.4367,
+      "step": 58430
+    },
+    {
+      "epoch": 0.29184249294614095,
+      "grad_norm": 0.09180597215890884,
+      "learning_rate": 2.1297904828656536e-05,
+      "loss": 9.4304,
+      "step": 58440
+    },
+    {
+      "epoch": 0.29189243177108043,
+      "grad_norm": 0.09582697600126266,
+      "learning_rate": 2.129640291371499e-05,
+      "loss": 9.4268,
+      "step": 58450
+    },
+    {
+      "epoch": 0.29194237059601985,
+      "grad_norm": 0.09342509508132935,
+      "learning_rate": 2.1294900998773436e-05,
+      "loss": 9.4324,
+      "step": 58460
+    },
+    {
+      "epoch": 0.29199230942095933,
+      "grad_norm": 0.09267037361860275,
+      "learning_rate": 2.1293399083831883e-05,
+      "loss": 9.4224,
+      "step": 58470
+    },
+    {
+      "epoch": 0.29204224824589875,
+      "grad_norm": 0.10199777036905289,
+      "learning_rate": 2.1291897168890337e-05,
+      "loss": 9.4253,
+      "step": 58480
+    },
+    {
+      "epoch": 0.2920921870708382,
+      "grad_norm": 0.08985645323991776,
+      "learning_rate": 2.1290395253948784e-05,
+      "loss": 9.4197,
+      "step": 58490
+    },
+    {
+      "epoch": 0.29214212589577765,
+      "grad_norm": 0.09782855212688446,
+      "learning_rate": 2.1288893339007237e-05,
+      "loss": 9.4231,
+      "step": 58500
+    },
+    {
+      "epoch": 0.2921920647207171,
+      "grad_norm": 0.09603475779294968,
+      "learning_rate": 2.1287391424065684e-05,
+      "loss": 9.4237,
+      "step": 58510
+    },
+    {
+      "epoch": 0.29224200354565655,
+      "grad_norm": 0.0944107323884964,
+      "learning_rate": 2.128588950912413e-05,
+      "loss": 9.4173,
+      "step": 58520
+    },
+    {
+      "epoch": 0.292291942370596,
+      "grad_norm": 0.0947156697511673,
+      "learning_rate": 2.1284387594182584e-05,
+      "loss": 9.4352,
+      "step": 58530
+    },
+    {
+      "epoch": 0.29234188119553545,
+      "grad_norm": 0.093475840985775,
+      "learning_rate": 2.128288567924103e-05,
+      "loss": 9.4267,
+      "step": 58540
+    },
+    {
+      "epoch": 0.2923918200204749,
+      "grad_norm": 0.1016530990600586,
+      "learning_rate": 2.1281383764299485e-05,
+      "loss": 9.4312,
+      "step": 58550
+    },
+    {
+      "epoch": 0.29244175884541435,
+      "grad_norm": 0.09788955748081207,
+      "learning_rate": 2.127988184935793e-05,
+      "loss": 9.4174,
+      "step": 58560
+    },
+    {
+      "epoch": 0.2924916976703538,
+      "grad_norm": 0.09802679717540741,
+      "learning_rate": 2.1278379934416378e-05,
+      "loss": 9.428,
+      "step": 58570
+    },
+    {
+      "epoch": 0.29254163649529324,
+      "grad_norm": 0.09107000380754471,
+      "learning_rate": 2.1276878019474832e-05,
+      "loss": 9.4274,
+      "step": 58580
+    },
+    {
+      "epoch": 0.2925915753202327,
+      "grad_norm": 0.09072667360305786,
+      "learning_rate": 2.127537610453328e-05,
+      "loss": 9.4244,
+      "step": 58590
+    },
+    {
+      "epoch": 0.29264151414517214,
+      "grad_norm": 0.097968690097332,
+      "learning_rate": 2.1273874189591732e-05,
+      "loss": 9.4158,
+      "step": 58600
+    },
+    {
+      "epoch": 0.2926914529701116,
+      "grad_norm": 0.09575551748275757,
+      "learning_rate": 2.127237227465018e-05,
+      "loss": 9.4244,
+      "step": 58610
+    },
+    {
+      "epoch": 0.29274139179505104,
+      "grad_norm": 0.09668948501348495,
+      "learning_rate": 2.127087035970863e-05,
+      "loss": 9.4156,
+      "step": 58620
+    },
+    {
+      "epoch": 0.2927913306199905,
+      "grad_norm": 0.09409730136394501,
+      "learning_rate": 2.126936844476708e-05,
+      "loss": 9.4266,
+      "step": 58630
+    },
+    {
+      "epoch": 0.29284126944492994,
+      "grad_norm": 0.09310100227594376,
+      "learning_rate": 2.1267866529825526e-05,
+      "loss": 9.4158,
+      "step": 58640
+    },
+    {
+      "epoch": 0.2928912082698694,
+      "grad_norm": 0.09272795915603638,
+      "learning_rate": 2.126636461488398e-05,
+      "loss": 9.4275,
+      "step": 58650
+    },
+    {
+      "epoch": 0.29294114709480884,
+      "grad_norm": 0.09329953789710999,
+      "learning_rate": 2.1264862699942426e-05,
+      "loss": 9.425,
+      "step": 58660
+    },
+    {
+      "epoch": 0.2929910859197483,
+      "grad_norm": 0.08982928097248077,
+      "learning_rate": 2.1263360785000877e-05,
+      "loss": 9.4224,
+      "step": 58670
+    },
+    {
+      "epoch": 0.29304102474468774,
+      "grad_norm": 0.09044983983039856,
+      "learning_rate": 2.1261858870059327e-05,
+      "loss": 9.4227,
+      "step": 58680
+    },
+    {
+      "epoch": 0.2930909635696272,
+      "grad_norm": 0.09245163947343826,
+      "learning_rate": 2.1260356955117774e-05,
+      "loss": 9.424,
+      "step": 58690
+    },
+    {
+      "epoch": 0.29314090239456664,
+      "grad_norm": 0.09064756333827972,
+      "learning_rate": 2.1258855040176227e-05,
+      "loss": 9.4198,
+      "step": 58700
+    },
+    {
+      "epoch": 0.2931908412195061,
+      "grad_norm": 0.08946257829666138,
+      "learning_rate": 2.1257353125234674e-05,
+      "loss": 9.4247,
+      "step": 58710
+    },
+    {
+      "epoch": 0.29324078004444554,
+      "grad_norm": 0.09985718131065369,
+      "learning_rate": 2.1255851210293124e-05,
+      "loss": 9.4138,
+      "step": 58720
+    },
+    {
+      "epoch": 0.293290718869385,
+      "grad_norm": 0.0930742546916008,
+      "learning_rate": 2.1254349295351574e-05,
+      "loss": 9.4201,
+      "step": 58730
+    },
+    {
+      "epoch": 0.29334065769432444,
+      "grad_norm": 0.08928178995847702,
+      "learning_rate": 2.125284738041002e-05,
+      "loss": 9.4215,
+      "step": 58740
+    },
+    {
+      "epoch": 0.2933905965192639,
+      "grad_norm": 0.09685919433832169,
+      "learning_rate": 2.1251345465468475e-05,
+      "loss": 9.4265,
+      "step": 58750
+    },
+    {
+      "epoch": 0.29344053534420333,
+      "grad_norm": 0.09459210932254791,
+      "learning_rate": 2.124984355052692e-05,
+      "loss": 9.411,
+      "step": 58760
+    },
+    {
+      "epoch": 0.2934904741691428,
+      "grad_norm": 0.09364203363656998,
+      "learning_rate": 2.124834163558537e-05,
+      "loss": 9.425,
+      "step": 58770
+    },
+    {
+      "epoch": 0.29354041299408223,
+      "grad_norm": 0.09589599817991257,
+      "learning_rate": 2.1246839720643822e-05,
+      "loss": 9.4135,
+      "step": 58780
+    },
+    {
+      "epoch": 0.2935903518190217,
+      "grad_norm": 0.08911266922950745,
+      "learning_rate": 2.124533780570227e-05,
+      "loss": 9.4207,
+      "step": 58790
+    },
+    {
+      "epoch": 0.29364029064396113,
+      "grad_norm": 0.09034483134746552,
+      "learning_rate": 2.1243835890760722e-05,
+      "loss": 9.4197,
+      "step": 58800
+    },
+    {
+      "epoch": 0.2936902294689006,
+      "grad_norm": 0.0975160002708435,
+      "learning_rate": 2.124233397581917e-05,
+      "loss": 9.4134,
+      "step": 58810
+    },
+    {
+      "epoch": 0.29374016829384003,
+      "grad_norm": 0.09520293772220612,
+      "learning_rate": 2.124083206087762e-05,
+      "loss": 9.4159,
+      "step": 58820
+    },
+    {
+      "epoch": 0.2937901071187795,
+      "grad_norm": 0.09131459891796112,
+      "learning_rate": 2.123933014593607e-05,
+      "loss": 9.4216,
+      "step": 58830
+    },
+    {
+      "epoch": 0.29384004594371893,
+      "grad_norm": 0.10248485952615738,
+      "learning_rate": 2.1237828230994516e-05,
+      "loss": 9.4095,
+      "step": 58840
+    },
+    {
+      "epoch": 0.2938899847686584,
+      "grad_norm": 0.09531592577695847,
+      "learning_rate": 2.123632631605297e-05,
+      "loss": 9.4285,
+      "step": 58850
+    },
+    {
+      "epoch": 0.29393992359359783,
+      "grad_norm": 0.09358898550271988,
+      "learning_rate": 2.1234824401111416e-05,
+      "loss": 9.4086,
+      "step": 58860
+    },
+    {
+      "epoch": 0.2939898624185373,
+      "grad_norm": 0.09685607254505157,
+      "learning_rate": 2.1233322486169867e-05,
+      "loss": 9.4227,
+      "step": 58870
+    },
+    {
+      "epoch": 0.29403980124347673,
+      "grad_norm": 0.09395089745521545,
+      "learning_rate": 2.1231820571228317e-05,
+      "loss": 9.4146,
+      "step": 58880
+    },
+    {
+      "epoch": 0.2940897400684162,
+      "grad_norm": 0.0898551344871521,
+      "learning_rate": 2.1230318656286764e-05,
+      "loss": 9.4178,
+      "step": 58890
+    },
+    {
+      "epoch": 0.2941396788933556,
+      "grad_norm": 0.0958523377776146,
+      "learning_rate": 2.1228816741345217e-05,
+      "loss": 9.4159,
+      "step": 58900
+    },
+    {
+      "epoch": 0.2941896177182951,
+      "grad_norm": 0.093168243765831,
+      "learning_rate": 2.1227314826403664e-05,
+      "loss": 9.4083,
+      "step": 58910
+    },
+    {
+      "epoch": 0.2942395565432345,
+      "grad_norm": 0.09986717253923416,
+      "learning_rate": 2.1225812911462114e-05,
+      "loss": 9.4089,
+      "step": 58920
+    },
+    {
+      "epoch": 0.294289495368174,
+      "grad_norm": 0.0952669233083725,
+      "learning_rate": 2.1224310996520564e-05,
+      "loss": 9.4216,
+      "step": 58930
+    },
+    {
+      "epoch": 0.2943394341931134,
+      "grad_norm": 0.0995723307132721,
+      "learning_rate": 2.122280908157901e-05,
+      "loss": 9.4183,
+      "step": 58940
+    },
+    {
+      "epoch": 0.2943893730180529,
+      "grad_norm": 0.09398746490478516,
+      "learning_rate": 2.1221307166637465e-05,
+      "loss": 9.4149,
+      "step": 58950
+    },
+    {
+      "epoch": 0.2944393118429923,
+      "grad_norm": 0.10161332786083221,
+      "learning_rate": 2.121980525169591e-05,
+      "loss": 9.4085,
+      "step": 58960
+    },
+    {
+      "epoch": 0.2944892506679318,
+      "grad_norm": 0.09317755699157715,
+      "learning_rate": 2.121830333675436e-05,
+      "loss": 9.4227,
+      "step": 58970
+    },
+    {
+      "epoch": 0.2945391894928712,
+      "grad_norm": 0.09553561359643936,
+      "learning_rate": 2.1216801421812812e-05,
+      "loss": 9.4188,
+      "step": 58980
+    },
+    {
+      "epoch": 0.2945891283178107,
+      "grad_norm": 0.09423384815454483,
+      "learning_rate": 2.1215299506871262e-05,
+      "loss": 9.4171,
+      "step": 58990
+    },
+    {
+      "epoch": 0.2946390671427501,
+      "grad_norm": 0.09159230440855026,
+      "learning_rate": 2.1213797591929712e-05,
+      "loss": 9.4175,
+      "step": 59000
+    },
+    {
+      "epoch": 0.2946890059676896,
+      "grad_norm": 0.09593381732702255,
+      "learning_rate": 2.121229567698816e-05,
+      "loss": 9.4132,
+      "step": 59010
+    },
+    {
+      "epoch": 0.294738944792629,
+      "grad_norm": 0.09449375420808792,
+      "learning_rate": 2.121079376204661e-05,
+      "loss": 9.422,
+      "step": 59020
+    },
+    {
+      "epoch": 0.2947888836175685,
+      "grad_norm": 0.09151158481836319,
+      "learning_rate": 2.120929184710506e-05,
+      "loss": 9.4252,
+      "step": 59030
+    },
+    {
+      "epoch": 0.2948388224425079,
+      "grad_norm": 0.09154769778251648,
+      "learning_rate": 2.120778993216351e-05,
+      "loss": 9.4157,
+      "step": 59040
+    },
+    {
+      "epoch": 0.2948887612674474,
+      "grad_norm": 0.09065014123916626,
+      "learning_rate": 2.120628801722196e-05,
+      "loss": 9.4179,
+      "step": 59050
+    },
+    {
+      "epoch": 0.2949387000923868,
+      "grad_norm": 0.09222662448883057,
+      "learning_rate": 2.1204786102280406e-05,
+      "loss": 9.4056,
+      "step": 59060
+    },
+    {
+      "epoch": 0.2949886389173263,
+      "grad_norm": 0.09881994873285294,
+      "learning_rate": 2.1203284187338857e-05,
+      "loss": 9.4033,
+      "step": 59070
+    },
+    {
+      "epoch": 0.2950385777422657,
+      "grad_norm": 0.10036629438400269,
+      "learning_rate": 2.1201782272397307e-05,
+      "loss": 9.4071,
+      "step": 59080
+    },
+    {
+      "epoch": 0.2950885165672052,
+      "grad_norm": 0.09214989095926285,
+      "learning_rate": 2.1200280357455757e-05,
+      "loss": 9.4131,
+      "step": 59090
+    },
+    {
+      "epoch": 0.2951384553921446,
+      "grad_norm": 0.09671517461538315,
+      "learning_rate": 2.1198778442514207e-05,
+      "loss": 9.407,
+      "step": 59100
+    },
+    {
+      "epoch": 0.2951883942170841,
+      "grad_norm": 0.09238772839307785,
+      "learning_rate": 2.1197276527572654e-05,
+      "loss": 9.4105,
+      "step": 59110
+    },
+    {
+      "epoch": 0.2952383330420235,
+      "grad_norm": 0.09682830423116684,
+      "learning_rate": 2.1195774612631104e-05,
+      "loss": 9.4144,
+      "step": 59120
+    },
+    {
+      "epoch": 0.295288271866963,
+      "grad_norm": 0.09174001961946487,
+      "learning_rate": 2.1194272697689554e-05,
+      "loss": 9.409,
+      "step": 59130
+    },
+    {
+      "epoch": 0.2953382106919024,
+      "grad_norm": 0.09227842837572098,
+      "learning_rate": 2.1192770782748005e-05,
+      "loss": 9.4068,
+      "step": 59140
+    },
+    {
+      "epoch": 0.2953881495168419,
+      "grad_norm": 0.09464056044816971,
+      "learning_rate": 2.1191268867806455e-05,
+      "loss": 9.4129,
+      "step": 59150
+    },
+    {
+      "epoch": 0.2954380883417813,
+      "grad_norm": 0.09324008971452713,
+      "learning_rate": 2.11897669528649e-05,
+      "loss": 9.4128,
+      "step": 59160
+    },
+    {
+      "epoch": 0.2954880271667208,
+      "grad_norm": 0.093290314078331,
+      "learning_rate": 2.118826503792335e-05,
+      "loss": 9.4178,
+      "step": 59170
+    },
+    {
+      "epoch": 0.2955379659916602,
+      "grad_norm": 0.09760567545890808,
+      "learning_rate": 2.1186763122981802e-05,
+      "loss": 9.4109,
+      "step": 59180
+    },
+    {
+      "epoch": 0.2955879048165997,
+      "grad_norm": 0.09324108064174652,
+      "learning_rate": 2.1185261208040252e-05,
+      "loss": 9.4177,
+      "step": 59190
+    },
+    {
+      "epoch": 0.2956378436415391,
+      "grad_norm": 0.09072353690862656,
+      "learning_rate": 2.1183759293098702e-05,
+      "loss": 9.4191,
+      "step": 59200
+    },
+    {
+      "epoch": 0.2956877824664786,
+      "grad_norm": 0.09248016029596329,
+      "learning_rate": 2.118225737815715e-05,
+      "loss": 9.4159,
+      "step": 59210
+    },
+    {
+      "epoch": 0.295737721291418,
+      "grad_norm": 0.08923139423131943,
+      "learning_rate": 2.1180755463215603e-05,
+      "loss": 9.4086,
+      "step": 59220
+    },
+    {
+      "epoch": 0.2957876601163575,
+      "grad_norm": 0.09490073472261429,
+      "learning_rate": 2.117925354827405e-05,
+      "loss": 9.4123,
+      "step": 59230
+    },
+    {
+      "epoch": 0.2958375989412969,
+      "grad_norm": 0.09313952177762985,
+      "learning_rate": 2.11777516333325e-05,
+      "loss": 9.4169,
+      "step": 59240
+    },
+    {
+      "epoch": 0.2958875377662364,
+      "grad_norm": 0.0937030240893364,
+      "learning_rate": 2.117624971839095e-05,
+      "loss": 9.4101,
+      "step": 59250
+    },
+    {
+      "epoch": 0.2959374765911758,
+      "grad_norm": 0.0976044312119484,
+      "learning_rate": 2.1174747803449396e-05,
+      "loss": 9.4097,
+      "step": 59260
+    },
+    {
+      "epoch": 0.2959874154161153,
+      "grad_norm": 0.09942024946212769,
+      "learning_rate": 2.117324588850785e-05,
+      "loss": 9.4033,
+      "step": 59270
+    },
+    {
+      "epoch": 0.2960373542410547,
+      "grad_norm": 0.09450352191925049,
+      "learning_rate": 2.1171743973566297e-05,
+      "loss": 9.4104,
+      "step": 59280
+    },
+    {
+      "epoch": 0.2960872930659942,
+      "grad_norm": 0.09626131504774094,
+      "learning_rate": 2.1170242058624747e-05,
+      "loss": 9.4153,
+      "step": 59290
+    },
+    {
+      "epoch": 0.2961372318909336,
+      "grad_norm": 0.09533768892288208,
+      "learning_rate": 2.1168740143683197e-05,
+      "loss": 9.4257,
+      "step": 59300
+    },
+    {
+      "epoch": 0.2961871707158731,
+      "grad_norm": 0.09322089701890945,
+      "learning_rate": 2.1167238228741647e-05,
+      "loss": 9.4129,
+      "step": 59310
+    },
+    {
+      "epoch": 0.2962371095408125,
+      "grad_norm": 0.08978445082902908,
+      "learning_rate": 2.1165736313800098e-05,
+      "loss": 9.4065,
+      "step": 59320
+    },
+    {
+      "epoch": 0.296287048365752,
+      "grad_norm": 0.09518822282552719,
+      "learning_rate": 2.1164234398858544e-05,
+      "loss": 9.4149,
+      "step": 59330
+    },
+    {
+      "epoch": 0.2963369871906914,
+      "grad_norm": 0.09288329631090164,
+      "learning_rate": 2.1162732483916995e-05,
+      "loss": 9.4026,
+      "step": 59340
+    },
+    {
+      "epoch": 0.2963869260156308,
+      "grad_norm": 0.0915319174528122,
+      "learning_rate": 2.1161230568975445e-05,
+      "loss": 9.4154,
+      "step": 59350
+    },
+    {
+      "epoch": 0.2964368648405703,
+      "grad_norm": 0.09252842515707016,
+      "learning_rate": 2.1159728654033895e-05,
+      "loss": 9.4147,
+      "step": 59360
+    },
+    {
+      "epoch": 0.2964868036655097,
+      "grad_norm": 0.09411928057670593,
+      "learning_rate": 2.1158226739092345e-05,
+      "loss": 9.4116,
+      "step": 59370
+    },
+    {
+      "epoch": 0.2965367424904492,
+      "grad_norm": 0.09231306612491608,
+      "learning_rate": 2.1156724824150792e-05,
+      "loss": 9.4012,
+      "step": 59380
+    },
+    {
+      "epoch": 0.2965866813153886,
+      "grad_norm": 0.09552402049303055,
+      "learning_rate": 2.1155222909209242e-05,
+      "loss": 9.4092,
+      "step": 59390
+    },
+    {
+      "epoch": 0.2966366201403281,
+      "grad_norm": 0.0954194962978363,
+      "learning_rate": 2.1153720994267692e-05,
+      "loss": 9.4077,
+      "step": 59400
+    },
+    {
+      "epoch": 0.2966865589652675,
+      "grad_norm": 0.09317540377378464,
+      "learning_rate": 2.1152219079326142e-05,
+      "loss": 9.4094,
+      "step": 59410
+    },
+    {
+      "epoch": 0.296736497790207,
+      "grad_norm": 0.09430140256881714,
+      "learning_rate": 2.1150717164384593e-05,
+      "loss": 9.4044,
+      "step": 59420
+    },
+    {
+      "epoch": 0.2967864366151464,
+      "grad_norm": 0.09407494962215424,
+      "learning_rate": 2.114921524944304e-05,
+      "loss": 9.4147,
+      "step": 59430
+    },
+    {
+      "epoch": 0.2968363754400859,
+      "grad_norm": 0.09957774728536606,
+      "learning_rate": 2.114771333450149e-05,
+      "loss": 9.4126,
+      "step": 59440
+    },
+    {
+      "epoch": 0.2968863142650253,
+      "grad_norm": 0.09334376454353333,
+      "learning_rate": 2.114621141955994e-05,
+      "loss": 9.4016,
+      "step": 59450
+    },
+    {
+      "epoch": 0.2969362530899648,
+      "grad_norm": 0.0957721695303917,
+      "learning_rate": 2.114470950461839e-05,
+      "loss": 9.4109,
+      "step": 59460
+    },
+    {
+      "epoch": 0.2969861919149042,
+      "grad_norm": 0.09447581321001053,
+      "learning_rate": 2.114320758967684e-05,
+      "loss": 9.4193,
+      "step": 59470
+    },
+    {
+      "epoch": 0.2970361307398437,
+      "grad_norm": 0.0941816195845604,
+      "learning_rate": 2.1141705674735287e-05,
+      "loss": 9.4047,
+      "step": 59480
+    },
+    {
+      "epoch": 0.2970860695647831,
+      "grad_norm": 0.08960052579641342,
+      "learning_rate": 2.1140203759793737e-05,
+      "loss": 9.4161,
+      "step": 59490
+    },
+    {
+      "epoch": 0.2971360083897226,
+      "grad_norm": 0.09240063279867172,
+      "learning_rate": 2.1138701844852187e-05,
+      "loss": 9.4053,
+      "step": 59500
+    },
+    {
+      "epoch": 0.297185947214662,
+      "grad_norm": 0.08736561983823776,
+      "learning_rate": 2.1137199929910637e-05,
+      "loss": 9.4092,
+      "step": 59510
+    },
+    {
+      "epoch": 0.2972358860396015,
+      "grad_norm": 0.09654782712459564,
+      "learning_rate": 2.1135698014969088e-05,
+      "loss": 9.4072,
+      "step": 59520
+    },
+    {
+      "epoch": 0.2972858248645409,
+      "grad_norm": 0.09687676280736923,
+      "learning_rate": 2.1134196100027534e-05,
+      "loss": 9.409,
+      "step": 59530
+    },
+    {
+      "epoch": 0.2973357636894804,
+      "grad_norm": 0.0934508666396141,
+      "learning_rate": 2.1132694185085985e-05,
+      "loss": 9.4048,
+      "step": 59540
+    },
+    {
+      "epoch": 0.2973857025144198,
+      "grad_norm": 0.09141778200864792,
+      "learning_rate": 2.1131192270144435e-05,
+      "loss": 9.4101,
+      "step": 59550
+    },
+    {
+      "epoch": 0.2974356413393593,
+      "grad_norm": 0.0905558317899704,
+      "learning_rate": 2.1129690355202885e-05,
+      "loss": 9.4071,
+      "step": 59560
+    },
+    {
+      "epoch": 0.2974855801642987,
+      "grad_norm": 0.09105515480041504,
+      "learning_rate": 2.1128188440261335e-05,
+      "loss": 9.4076,
+      "step": 59570
+    },
+    {
+      "epoch": 0.2975355189892382,
+      "grad_norm": 0.09343723952770233,
+      "learning_rate": 2.1126686525319782e-05,
+      "loss": 9.4212,
+      "step": 59580
+    },
+    {
+      "epoch": 0.2975854578141776,
+      "grad_norm": 0.09228665381669998,
+      "learning_rate": 2.1125184610378232e-05,
+      "loss": 9.4151,
+      "step": 59590
+    },
+    {
+      "epoch": 0.2976353966391171,
+      "grad_norm": 0.09571309387683868,
+      "learning_rate": 2.1123682695436682e-05,
+      "loss": 9.3969,
+      "step": 59600
+    },
+    {
+      "epoch": 0.2976853354640565,
+      "grad_norm": 0.09124372899532318,
+      "learning_rate": 2.1122180780495132e-05,
+      "loss": 9.4104,
+      "step": 59610
+    },
+    {
+      "epoch": 0.297735274288996,
+      "grad_norm": 0.09096134454011917,
+      "learning_rate": 2.1120678865553583e-05,
+      "loss": 9.4108,
+      "step": 59620
+    },
+    {
+      "epoch": 0.2977852131139354,
+      "grad_norm": 0.09474842250347137,
+      "learning_rate": 2.1119176950612033e-05,
+      "loss": 9.4011,
+      "step": 59630
+    },
+    {
+      "epoch": 0.2978351519388749,
+      "grad_norm": 0.09214451164007187,
+      "learning_rate": 2.111767503567048e-05,
+      "loss": 9.4138,
+      "step": 59640
+    },
+    {
+      "epoch": 0.2978850907638143,
+      "grad_norm": 0.09250292927026749,
+      "learning_rate": 2.111617312072893e-05,
+      "loss": 9.4054,
+      "step": 59650
+    },
+    {
+      "epoch": 0.2979350295887538,
+      "grad_norm": 0.09642202407121658,
+      "learning_rate": 2.111467120578738e-05,
+      "loss": 9.4088,
+      "step": 59660
+    },
+    {
+      "epoch": 0.2979849684136932,
+      "grad_norm": 0.09199708700180054,
+      "learning_rate": 2.111316929084583e-05,
+      "loss": 9.4239,
+      "step": 59670
+    },
+    {
+      "epoch": 0.2980349072386327,
+      "grad_norm": 0.09029726684093475,
+      "learning_rate": 2.111166737590428e-05,
+      "loss": 9.4053,
+      "step": 59680
+    },
+    {
+      "epoch": 0.2980848460635721,
+      "grad_norm": 0.09503061324357986,
+      "learning_rate": 2.1110165460962727e-05,
+      "loss": 9.4035,
+      "step": 59690
+    },
+    {
+      "epoch": 0.2981347848885116,
+      "grad_norm": 0.09824039041996002,
+      "learning_rate": 2.1108663546021177e-05,
+      "loss": 9.4013,
+      "step": 59700
+    },
+    {
+      "epoch": 0.298184723713451,
+      "grad_norm": 0.09801938384771347,
+      "learning_rate": 2.1107161631079627e-05,
+      "loss": 9.3948,
+      "step": 59710
+    },
+    {
+      "epoch": 0.2982346625383905,
+      "grad_norm": 0.10737218707799911,
+      "learning_rate": 2.1105659716138078e-05,
+      "loss": 9.3953,
+      "step": 59720
+    },
+    {
+      "epoch": 0.2982846013633299,
+      "grad_norm": 0.09229201078414917,
+      "learning_rate": 2.1104157801196528e-05,
+      "loss": 9.4062,
+      "step": 59730
+    },
+    {
+      "epoch": 0.2983345401882694,
+      "grad_norm": 0.08875761926174164,
+      "learning_rate": 2.1102655886254975e-05,
+      "loss": 9.4129,
+      "step": 59740
+    },
+    {
+      "epoch": 0.2983844790132088,
+      "grad_norm": 0.09321201592683792,
+      "learning_rate": 2.1101153971313425e-05,
+      "loss": 9.412,
+      "step": 59750
+    },
+    {
+      "epoch": 0.2984344178381483,
+      "grad_norm": 0.0906151756644249,
+      "learning_rate": 2.1099652056371875e-05,
+      "loss": 9.3997,
+      "step": 59760
+    },
+    {
+      "epoch": 0.2984843566630877,
+      "grad_norm": 0.10253120213747025,
+      "learning_rate": 2.1098150141430325e-05,
+      "loss": 9.397,
+      "step": 59770
+    },
+    {
+      "epoch": 0.2985342954880272,
+      "grad_norm": 0.08926044404506683,
+      "learning_rate": 2.1096648226488775e-05,
+      "loss": 9.4117,
+      "step": 59780
+    },
+    {
+      "epoch": 0.2985842343129666,
+      "grad_norm": 0.09025093168020248,
+      "learning_rate": 2.1095146311547222e-05,
+      "loss": 9.4118,
+      "step": 59790
+    },
+    {
+      "epoch": 0.2986341731379061,
+      "grad_norm": 0.09714997559785843,
+      "learning_rate": 2.1093644396605672e-05,
+      "loss": 9.4087,
+      "step": 59800
+    },
+    {
+      "epoch": 0.2986841119628455,
+      "grad_norm": 0.09052036702632904,
+      "learning_rate": 2.1092142481664122e-05,
+      "loss": 9.4045,
+      "step": 59810
+    },
+    {
+      "epoch": 0.298734050787785,
+      "grad_norm": 0.09755855053663254,
+      "learning_rate": 2.1090640566722573e-05,
+      "loss": 9.398,
+      "step": 59820
+    },
+    {
+      "epoch": 0.2987839896127244,
+      "grad_norm": 0.09479790180921555,
+      "learning_rate": 2.1089138651781023e-05,
+      "loss": 9.404,
+      "step": 59830
+    },
+    {
+      "epoch": 0.2988339284376639,
+      "grad_norm": 0.09147018939256668,
+      "learning_rate": 2.108763673683947e-05,
+      "loss": 9.3966,
+      "step": 59840
+    },
+    {
+      "epoch": 0.2988838672626033,
+      "grad_norm": 0.0920906662940979,
+      "learning_rate": 2.108613482189792e-05,
+      "loss": 9.4109,
+      "step": 59850
+    },
+    {
+      "epoch": 0.2989338060875428,
+      "grad_norm": 0.0949472039937973,
+      "learning_rate": 2.108463290695637e-05,
+      "loss": 9.3958,
+      "step": 59860
+    },
+    {
+      "epoch": 0.2989837449124822,
+      "grad_norm": 0.0973886027932167,
+      "learning_rate": 2.108313099201482e-05,
+      "loss": 9.4021,
+      "step": 59870
+    },
+    {
+      "epoch": 0.2990336837374217,
+      "grad_norm": 0.08694946020841599,
+      "learning_rate": 2.108162907707327e-05,
+      "loss": 9.405,
+      "step": 59880
+    },
+    {
+      "epoch": 0.2990836225623611,
+      "grad_norm": 0.09538460522890091,
+      "learning_rate": 2.1080127162131717e-05,
+      "loss": 9.4057,
+      "step": 59890
+    },
+    {
+      "epoch": 0.2991335613873006,
+      "grad_norm": 0.0949239507317543,
+      "learning_rate": 2.1078625247190167e-05,
+      "loss": 9.4029,
+      "step": 59900
+    },
+    {
+      "epoch": 0.29918350021224,
+      "grad_norm": 0.09560471773147583,
+      "learning_rate": 2.1077123332248617e-05,
+      "loss": 9.3962,
+      "step": 59910
+    },
+    {
+      "epoch": 0.29923343903717947,
+      "grad_norm": 0.09594585001468658,
+      "learning_rate": 2.1075621417307068e-05,
+      "loss": 9.4046,
+      "step": 59920
+    },
+    {
+      "epoch": 0.2992833778621189,
+      "grad_norm": 0.09318960458040237,
+      "learning_rate": 2.1074119502365518e-05,
+      "loss": 9.407,
+      "step": 59930
+    },
+    {
+      "epoch": 0.29933331668705837,
+      "grad_norm": 0.09141054004430771,
+      "learning_rate": 2.1072617587423965e-05,
+      "loss": 9.3944,
+      "step": 59940
+    },
+    {
+      "epoch": 0.2993832555119978,
+      "grad_norm": 0.09421342611312866,
+      "learning_rate": 2.1071115672482418e-05,
+      "loss": 9.4033,
+      "step": 59950
+    },
+    {
+      "epoch": 0.29943319433693727,
+      "grad_norm": 0.09517619758844376,
+      "learning_rate": 2.1069613757540865e-05,
+      "loss": 9.3988,
+      "step": 59960
+    },
+    {
+      "epoch": 0.2994831331618767,
+      "grad_norm": 0.09366179257631302,
+      "learning_rate": 2.1068111842599315e-05,
+      "loss": 9.4144,
+      "step": 59970
+    },
+    {
+      "epoch": 0.29953307198681617,
+      "grad_norm": 0.09368981420993805,
+      "learning_rate": 2.1066609927657765e-05,
+      "loss": 9.3938,
+      "step": 59980
+    },
+    {
+      "epoch": 0.2995830108117556,
+      "grad_norm": 0.09122604131698608,
+      "learning_rate": 2.1065108012716212e-05,
+      "loss": 9.3962,
+      "step": 59990
+    },
+    {
+      "epoch": 0.29963294963669507,
+      "grad_norm": 0.0922602191567421,
+      "learning_rate": 2.1063606097774666e-05,
+      "loss": 9.3997,
+      "step": 60000
+    },
+    {
+      "epoch": 0.2996828884616345,
+      "grad_norm": 0.1016426831483841,
+      "learning_rate": 2.1062104182833112e-05,
+      "loss": 9.3922,
+      "step": 60010
+    },
+    {
+      "epoch": 0.29973282728657397,
+      "grad_norm": 0.08960018306970596,
+      "learning_rate": 2.1060602267891563e-05,
+      "loss": 9.3926,
+      "step": 60020
+    },
+    {
+      "epoch": 0.2997827661115134,
+      "grad_norm": 0.08921145647764206,
+      "learning_rate": 2.1059100352950013e-05,
+      "loss": 9.3912,
+      "step": 60030
+    },
+    {
+      "epoch": 0.29983270493645287,
+      "grad_norm": 0.09575637429952621,
+      "learning_rate": 2.105759843800846e-05,
+      "loss": 9.3966,
+      "step": 60040
+    },
+    {
+      "epoch": 0.2998826437613923,
+      "grad_norm": 0.09875485301017761,
+      "learning_rate": 2.1056096523066913e-05,
+      "loss": 9.3956,
+      "step": 60050
+    },
+    {
+      "epoch": 0.29993258258633176,
+      "grad_norm": 0.09163234382867813,
+      "learning_rate": 2.105459460812536e-05,
+      "loss": 9.3946,
+      "step": 60060
+    },
+    {
+      "epoch": 0.2999825214112712,
+      "grad_norm": 0.0965883880853653,
+      "learning_rate": 2.105309269318381e-05,
+      "loss": 9.3962,
+      "step": 60070
+    },
+    {
+      "epoch": 0.30003246023621066,
+      "grad_norm": 0.09682179242372513,
+      "learning_rate": 2.105159077824226e-05,
+      "loss": 9.4011,
+      "step": 60080
+    },
+    {
+      "epoch": 0.3000823990611501,
+      "grad_norm": 0.09401979297399521,
+      "learning_rate": 2.1050088863300707e-05,
+      "loss": 9.3984,
+      "step": 60090
+    },
+    {
+      "epoch": 0.30013233788608956,
+      "grad_norm": 0.09467914700508118,
+      "learning_rate": 2.104858694835916e-05,
+      "loss": 9.3967,
+      "step": 60100
+    },
+    {
+      "epoch": 0.300182276711029,
+      "grad_norm": 0.09743199497461319,
+      "learning_rate": 2.1047085033417607e-05,
+      "loss": 9.3941,
+      "step": 60110
+    },
+    {
+      "epoch": 0.30023221553596846,
+      "grad_norm": 0.08682700246572495,
+      "learning_rate": 2.1045583118476058e-05,
+      "loss": 9.3914,
+      "step": 60120
+    },
+    {
+      "epoch": 0.3002821543609079,
+      "grad_norm": 0.0950264036655426,
+      "learning_rate": 2.1044081203534508e-05,
+      "loss": 9.395,
+      "step": 60130
+    },
+    {
+      "epoch": 0.30033209318584736,
+      "grad_norm": 0.09148851037025452,
+      "learning_rate": 2.1042579288592955e-05,
+      "loss": 9.3978,
+      "step": 60140
+    },
+    {
+      "epoch": 0.3003820320107868,
+      "grad_norm": 0.09214703738689423,
+      "learning_rate": 2.1041077373651408e-05,
+      "loss": 9.3972,
+      "step": 60150
+    },
+    {
+      "epoch": 0.30043197083572626,
+      "grad_norm": 0.0951230525970459,
+      "learning_rate": 2.1039575458709855e-05,
+      "loss": 9.4042,
+      "step": 60160
+    },
+    {
+      "epoch": 0.3004819096606657,
+      "grad_norm": 0.09221307933330536,
+      "learning_rate": 2.1038073543768305e-05,
+      "loss": 9.4051,
+      "step": 60170
+    },
+    {
+      "epoch": 0.30053184848560516,
+      "grad_norm": 0.09453621506690979,
+      "learning_rate": 2.1036571628826755e-05,
+      "loss": 9.3895,
+      "step": 60180
+    },
+    {
+      "epoch": 0.3005817873105446,
+      "grad_norm": 0.09616675227880478,
+      "learning_rate": 2.1035069713885202e-05,
+      "loss": 9.4036,
+      "step": 60190
+    },
+    {
+      "epoch": 0.30063172613548406,
+      "grad_norm": 0.0845404714345932,
+      "learning_rate": 2.1033567798943656e-05,
+      "loss": 9.4068,
+      "step": 60200
+    },
+    {
+      "epoch": 0.3006816649604235,
+      "grad_norm": 0.092598095536232,
+      "learning_rate": 2.1032065884002102e-05,
+      "loss": 9.396,
+      "step": 60210
+    },
+    {
+      "epoch": 0.30073160378536296,
+      "grad_norm": 0.09395390748977661,
+      "learning_rate": 2.1030563969060553e-05,
+      "loss": 9.4003,
+      "step": 60220
+    },
+    {
+      "epoch": 0.3007815426103024,
+      "grad_norm": 0.09719889611005783,
+      "learning_rate": 2.1029062054119003e-05,
+      "loss": 9.383,
+      "step": 60230
+    },
+    {
+      "epoch": 0.30083148143524185,
+      "grad_norm": 0.10100460797548294,
+      "learning_rate": 2.102756013917745e-05,
+      "loss": 9.398,
+      "step": 60240
+    },
+    {
+      "epoch": 0.3008814202601813,
+      "grad_norm": 0.09339495748281479,
+      "learning_rate": 2.1026058224235903e-05,
+      "loss": 9.3972,
+      "step": 60250
+    },
+    {
+      "epoch": 0.30093135908512075,
+      "grad_norm": 0.095674529671669,
+      "learning_rate": 2.102455630929435e-05,
+      "loss": 9.3979,
+      "step": 60260
+    },
+    {
+      "epoch": 0.3009812979100602,
+      "grad_norm": 0.09552553296089172,
+      "learning_rate": 2.1023054394352804e-05,
+      "loss": 9.4002,
+      "step": 60270
+    },
+    {
+      "epoch": 0.30103123673499965,
+      "grad_norm": 0.0926952138543129,
+      "learning_rate": 2.102155247941125e-05,
+      "loss": 9.4003,
+      "step": 60280
+    },
+    {
+      "epoch": 0.3010811755599391,
+      "grad_norm": 0.09167633205652237,
+      "learning_rate": 2.1020050564469697e-05,
+      "loss": 9.3939,
+      "step": 60290
+    },
+    {
+      "epoch": 0.30113111438487855,
+      "grad_norm": 0.10082550346851349,
+      "learning_rate": 2.101854864952815e-05,
+      "loss": 9.3882,
+      "step": 60300
+    },
+    {
+      "epoch": 0.301181053209818,
+      "grad_norm": 0.09979651123285294,
+      "learning_rate": 2.1017046734586597e-05,
+      "loss": 9.3897,
+      "step": 60310
+    },
+    {
+      "epoch": 0.30123099203475745,
+      "grad_norm": 0.09624478965997696,
+      "learning_rate": 2.101554481964505e-05,
+      "loss": 9.3969,
+      "step": 60320
+    },
+    {
+      "epoch": 0.30128093085969687,
+      "grad_norm": 0.09281089156866074,
+      "learning_rate": 2.1014042904703498e-05,
+      "loss": 9.3837,
+      "step": 60330
+    },
+    {
+      "epoch": 0.3013308696846363,
+      "grad_norm": 0.08858885616064072,
+      "learning_rate": 2.1012540989761945e-05,
+      "loss": 9.3936,
+      "step": 60340
+    },
+    {
+      "epoch": 0.30138080850957577,
+      "grad_norm": 0.09177447855472565,
+      "learning_rate": 2.1011039074820398e-05,
+      "loss": 9.4041,
+      "step": 60350
+    },
+    {
+      "epoch": 0.3014307473345152,
+      "grad_norm": 0.09817986190319061,
+      "learning_rate": 2.1009537159878845e-05,
+      "loss": 9.4026,
+      "step": 60360
+    },
+    {
+      "epoch": 0.30148068615945467,
+      "grad_norm": 0.09388108551502228,
+      "learning_rate": 2.10080352449373e-05,
+      "loss": 9.3916,
+      "step": 60370
+    },
+    {
+      "epoch": 0.3015306249843941,
+      "grad_norm": 0.09480094164609909,
+      "learning_rate": 2.1006533329995745e-05,
+      "loss": 9.3932,
+      "step": 60380
+    },
+    {
+      "epoch": 0.30158056380933357,
+      "grad_norm": 0.09471967816352844,
+      "learning_rate": 2.1005031415054192e-05,
+      "loss": 9.3966,
+      "step": 60390
+    },
+    {
+      "epoch": 0.301630502634273,
+      "grad_norm": 0.0964706614613533,
+      "learning_rate": 2.1003529500112646e-05,
+      "loss": 9.3854,
+      "step": 60400
+    },
+    {
+      "epoch": 0.30168044145921247,
+      "grad_norm": 0.0946371778845787,
+      "learning_rate": 2.1002027585171092e-05,
+      "loss": 9.3927,
+      "step": 60410
+    },
+    {
+      "epoch": 0.3017303802841519,
+      "grad_norm": 0.09347691386938095,
+      "learning_rate": 2.1000525670229546e-05,
+      "loss": 9.4019,
+      "step": 60420
+    },
+    {
+      "epoch": 0.30178031910909137,
+      "grad_norm": 0.09268350899219513,
+      "learning_rate": 2.0999023755287993e-05,
+      "loss": 9.3979,
+      "step": 60430
+    },
+    {
+      "epoch": 0.3018302579340308,
+      "grad_norm": 0.10059063881635666,
+      "learning_rate": 2.099752184034644e-05,
+      "loss": 9.3866,
+      "step": 60440
+    },
+    {
+      "epoch": 0.30188019675897027,
+      "grad_norm": 0.08915655314922333,
+      "learning_rate": 2.0996019925404893e-05,
+      "loss": 9.3971,
+      "step": 60450
+    },
+    {
+      "epoch": 0.3019301355839097,
+      "grad_norm": 0.09569905698299408,
+      "learning_rate": 2.099451801046334e-05,
+      "loss": 9.3943,
+      "step": 60460
+    },
+    {
+      "epoch": 0.30198007440884916,
+      "grad_norm": 0.10031815618276596,
+      "learning_rate": 2.0993016095521794e-05,
+      "loss": 9.3903,
+      "step": 60470
+    },
+    {
+      "epoch": 0.3020300132337886,
+      "grad_norm": 0.09447156637907028,
+      "learning_rate": 2.099151418058024e-05,
+      "loss": 9.3889,
+      "step": 60480
+    },
+    {
+      "epoch": 0.30207995205872806,
+      "grad_norm": 0.09017782658338547,
+      "learning_rate": 2.0990012265638687e-05,
+      "loss": 9.3948,
+      "step": 60490
+    },
+    {
+      "epoch": 0.3021298908836675,
+      "grad_norm": 0.09461423754692078,
+      "learning_rate": 2.098851035069714e-05,
+      "loss": 9.3869,
+      "step": 60500
+    },
+    {
+      "epoch": 0.30217982970860696,
+      "grad_norm": 0.09656091034412384,
+      "learning_rate": 2.0987008435755587e-05,
+      "loss": 9.3937,
+      "step": 60510
+    },
+    {
+      "epoch": 0.3022297685335464,
+      "grad_norm": 0.08991716802120209,
+      "learning_rate": 2.098550652081404e-05,
+      "loss": 9.3899,
+      "step": 60520
+    },
+    {
+      "epoch": 0.30227970735848586,
+      "grad_norm": 0.09503340721130371,
+      "learning_rate": 2.0984004605872488e-05,
+      "loss": 9.3889,
+      "step": 60530
+    },
+    {
+      "epoch": 0.3023296461834253,
+      "grad_norm": 0.08786161243915558,
+      "learning_rate": 2.0982502690930935e-05,
+      "loss": 9.3972,
+      "step": 60540
+    },
+    {
+      "epoch": 0.30237958500836476,
+      "grad_norm": 0.09755858033895493,
+      "learning_rate": 2.0981000775989388e-05,
+      "loss": 9.4025,
+      "step": 60550
+    },
+    {
+      "epoch": 0.3024295238333042,
+      "grad_norm": 0.09725630283355713,
+      "learning_rate": 2.0979498861047835e-05,
+      "loss": 9.3909,
+      "step": 60560
+    },
+    {
+      "epoch": 0.30247946265824366,
+      "grad_norm": 0.08783844858407974,
+      "learning_rate": 2.097799694610629e-05,
+      "loss": 9.3987,
+      "step": 60570
+    },
+    {
+      "epoch": 0.3025294014831831,
+      "grad_norm": 0.09798564016819,
+      "learning_rate": 2.0976495031164735e-05,
+      "loss": 9.4008,
+      "step": 60580
+    },
+    {
+      "epoch": 0.30257934030812256,
+      "grad_norm": 0.09155137091875076,
+      "learning_rate": 2.0974993116223185e-05,
+      "loss": 9.3922,
+      "step": 60590
+    },
+    {
+      "epoch": 0.302629279133062,
+      "grad_norm": 0.09524587541818619,
+      "learning_rate": 2.0973491201281636e-05,
+      "loss": 9.3868,
+      "step": 60600
+    },
+    {
+      "epoch": 0.30267921795800146,
+      "grad_norm": 0.09223838150501251,
+      "learning_rate": 2.0971989286340082e-05,
+      "loss": 9.3943,
+      "step": 60610
+    },
+    {
+      "epoch": 0.3027291567829409,
+      "grad_norm": 0.09258273243904114,
+      "learning_rate": 2.0970487371398536e-05,
+      "loss": 9.382,
+      "step": 60620
+    },
+    {
+      "epoch": 0.30277909560788036,
+      "grad_norm": 0.09514257311820984,
+      "learning_rate": 2.0968985456456983e-05,
+      "loss": 9.3933,
+      "step": 60630
+    },
+    {
+      "epoch": 0.3028290344328198,
+      "grad_norm": 0.08645520359277725,
+      "learning_rate": 2.0967483541515433e-05,
+      "loss": 9.3864,
+      "step": 60640
+    },
+    {
+      "epoch": 0.30287897325775925,
+      "grad_norm": 0.09346000850200653,
+      "learning_rate": 2.0965981626573883e-05,
+      "loss": 9.4062,
+      "step": 60650
+    },
+    {
+      "epoch": 0.3029289120826987,
+      "grad_norm": 0.09467698633670807,
+      "learning_rate": 2.096447971163233e-05,
+      "loss": 9.3819,
+      "step": 60660
+    },
+    {
+      "epoch": 0.30297885090763815,
+      "grad_norm": 0.09375496953725815,
+      "learning_rate": 2.0962977796690784e-05,
+      "loss": 9.3892,
+      "step": 60670
+    },
+    {
+      "epoch": 0.3030287897325776,
+      "grad_norm": 0.09323824197053909,
+      "learning_rate": 2.096147588174923e-05,
+      "loss": 9.3915,
+      "step": 60680
+    },
+    {
+      "epoch": 0.30307872855751705,
+      "grad_norm": 0.0925745964050293,
+      "learning_rate": 2.095997396680768e-05,
+      "loss": 9.3835,
+      "step": 60690
+    },
+    {
+      "epoch": 0.3031286673824565,
+      "grad_norm": 0.09299085289239883,
+      "learning_rate": 2.095847205186613e-05,
+      "loss": 9.3882,
+      "step": 60700
+    },
+    {
+      "epoch": 0.30317860620739595,
+      "grad_norm": 0.09627874940633774,
+      "learning_rate": 2.0956970136924577e-05,
+      "loss": 9.3914,
+      "step": 60710
+    },
+    {
+      "epoch": 0.3032285450323354,
+      "grad_norm": 0.09084891527891159,
+      "learning_rate": 2.095546822198303e-05,
+      "loss": 9.3905,
+      "step": 60720
+    },
+    {
+      "epoch": 0.30327848385727485,
+      "grad_norm": 0.09070669114589691,
+      "learning_rate": 2.0953966307041478e-05,
+      "loss": 9.3851,
+      "step": 60730
+    },
+    {
+      "epoch": 0.30332842268221427,
+      "grad_norm": 0.09068963676691055,
+      "learning_rate": 2.0952464392099928e-05,
+      "loss": 9.3869,
+      "step": 60740
+    },
+    {
+      "epoch": 0.30337836150715375,
+      "grad_norm": 0.09145542979240417,
+      "learning_rate": 2.0950962477158378e-05,
+      "loss": 9.3823,
+      "step": 60750
+    },
+    {
+      "epoch": 0.30342830033209317,
+      "grad_norm": 0.09390787780284882,
+      "learning_rate": 2.0949460562216825e-05,
+      "loss": 9.3873,
+      "step": 60760
+    },
+    {
+      "epoch": 0.30347823915703265,
+      "grad_norm": 0.09399698674678802,
+      "learning_rate": 2.094795864727528e-05,
+      "loss": 9.3884,
+      "step": 60770
+    },
+    {
+      "epoch": 0.30352817798197207,
+      "grad_norm": 0.09260044991970062,
+      "learning_rate": 2.0946456732333725e-05,
+      "loss": 9.3739,
+      "step": 60780
+    },
+    {
+      "epoch": 0.30357811680691155,
+      "grad_norm": 0.09860619902610779,
+      "learning_rate": 2.0944954817392176e-05,
+      "loss": 9.4007,
+      "step": 60790
+    },
+    {
+      "epoch": 0.30362805563185097,
+      "grad_norm": 0.0905844122171402,
+      "learning_rate": 2.0943452902450626e-05,
+      "loss": 9.3961,
+      "step": 60800
+    },
+    {
+      "epoch": 0.30367799445679045,
+      "grad_norm": 0.09324529021978378,
+      "learning_rate": 2.0941950987509072e-05,
+      "loss": 9.3876,
+      "step": 60810
+    },
+    {
+      "epoch": 0.30372793328172987,
+      "grad_norm": 0.08999308943748474,
+      "learning_rate": 2.0940449072567526e-05,
+      "loss": 9.3901,
+      "step": 60820
+    },
+    {
+      "epoch": 0.30377787210666934,
+      "grad_norm": 0.09419707208871841,
+      "learning_rate": 2.0938947157625973e-05,
+      "loss": 9.3841,
+      "step": 60830
+    },
+    {
+      "epoch": 0.30382781093160877,
+      "grad_norm": 0.09293625503778458,
+      "learning_rate": 2.0937445242684423e-05,
+      "loss": 9.3892,
+      "step": 60840
+    },
+    {
+      "epoch": 0.30387774975654824,
+      "grad_norm": 0.09216861426830292,
+      "learning_rate": 2.0935943327742873e-05,
+      "loss": 9.3948,
+      "step": 60850
+    },
+    {
+      "epoch": 0.30392768858148766,
+      "grad_norm": 0.0980692207813263,
+      "learning_rate": 2.093444141280132e-05,
+      "loss": 9.3888,
+      "step": 60860
+    },
+    {
+      "epoch": 0.30397762740642714,
+      "grad_norm": 0.09464273601770401,
+      "learning_rate": 2.0932939497859774e-05,
+      "loss": 9.3793,
+      "step": 60870
+    },
+    {
+      "epoch": 0.30402756623136656,
+      "grad_norm": 0.0960172563791275,
+      "learning_rate": 2.093143758291822e-05,
+      "loss": 9.3839,
+      "step": 60880
+    },
+    {
+      "epoch": 0.30407750505630604,
+      "grad_norm": 0.09181035310029984,
+      "learning_rate": 2.092993566797667e-05,
+      "loss": 9.3835,
+      "step": 60890
+    },
+    {
+      "epoch": 0.30412744388124546,
+      "grad_norm": 0.09236995130777359,
+      "learning_rate": 2.092843375303512e-05,
+      "loss": 9.3809,
+      "step": 60900
+    },
+    {
+      "epoch": 0.30417738270618494,
+      "grad_norm": 0.08640799671411514,
+      "learning_rate": 2.092693183809357e-05,
+      "loss": 9.378,
+      "step": 60910
+    },
+    {
+      "epoch": 0.30422732153112436,
+      "grad_norm": 0.0906277671456337,
+      "learning_rate": 2.092542992315202e-05,
+      "loss": 9.381,
+      "step": 60920
+    },
+    {
+      "epoch": 0.30427726035606384,
+      "grad_norm": 0.10483743995428085,
+      "learning_rate": 2.0923928008210468e-05,
+      "loss": 9.3815,
+      "step": 60930
+    },
+    {
+      "epoch": 0.30432719918100326,
+      "grad_norm": 0.08990181982517242,
+      "learning_rate": 2.0922426093268918e-05,
+      "loss": 9.3943,
+      "step": 60940
+    },
+    {
+      "epoch": 0.30437713800594274,
+      "grad_norm": 0.09138570725917816,
+      "learning_rate": 2.0920924178327368e-05,
+      "loss": 9.3993,
+      "step": 60950
+    },
+    {
+      "epoch": 0.30442707683088216,
+      "grad_norm": 0.0894051194190979,
+      "learning_rate": 2.091942226338582e-05,
+      "loss": 9.3783,
+      "step": 60960
+    },
+    {
+      "epoch": 0.30447701565582164,
+      "grad_norm": 0.09751047194004059,
+      "learning_rate": 2.091792034844427e-05,
+      "loss": 9.3815,
+      "step": 60970
+    },
+    {
+      "epoch": 0.30452695448076106,
+      "grad_norm": 0.09688021242618561,
+      "learning_rate": 2.0916418433502715e-05,
+      "loss": 9.3885,
+      "step": 60980
+    },
+    {
+      "epoch": 0.30457689330570054,
+      "grad_norm": 0.09227102994918823,
+      "learning_rate": 2.0914916518561166e-05,
+      "loss": 9.3901,
+      "step": 60990
+    },
+    {
+      "epoch": 0.30462683213063996,
+      "grad_norm": 0.09256913512945175,
+      "learning_rate": 2.0913414603619616e-05,
+      "loss": 9.3922,
+      "step": 61000
+    },
+    {
+      "epoch": 0.30467677095557943,
+      "grad_norm": 0.08989210426807404,
+      "learning_rate": 2.0911912688678066e-05,
+      "loss": 9.3778,
+      "step": 61010
+    },
+    {
+      "epoch": 0.30472670978051886,
+      "grad_norm": 0.09482806921005249,
+      "learning_rate": 2.0910410773736516e-05,
+      "loss": 9.3865,
+      "step": 61020
+    },
+    {
+      "epoch": 0.30477664860545833,
+      "grad_norm": 0.09472500532865524,
+      "learning_rate": 2.0908908858794963e-05,
+      "loss": 9.3832,
+      "step": 61030
+    },
+    {
+      "epoch": 0.30482658743039776,
+      "grad_norm": 0.09536511451005936,
+      "learning_rate": 2.0907406943853413e-05,
+      "loss": 9.3852,
+      "step": 61040
+    },
+    {
+      "epoch": 0.30487652625533723,
+      "grad_norm": 0.0961252823472023,
+      "learning_rate": 2.0905905028911863e-05,
+      "loss": 9.3877,
+      "step": 61050
+    },
+    {
+      "epoch": 0.30492646508027665,
+      "grad_norm": 0.0982387587428093,
+      "learning_rate": 2.0904403113970313e-05,
+      "loss": 9.3891,
+      "step": 61060
+    },
+    {
+      "epoch": 0.30497640390521613,
+      "grad_norm": 0.09710310399532318,
+      "learning_rate": 2.0902901199028764e-05,
+      "loss": 9.3805,
+      "step": 61070
+    },
+    {
+      "epoch": 0.30502634273015555,
+      "grad_norm": 0.09833624958992004,
+      "learning_rate": 2.090139928408721e-05,
+      "loss": 9.3872,
+      "step": 61080
+    },
+    {
+      "epoch": 0.30507628155509503,
+      "grad_norm": 0.09045345336198807,
+      "learning_rate": 2.089989736914566e-05,
+      "loss": 9.3848,
+      "step": 61090
+    },
+    {
+      "epoch": 0.30512622038003445,
+      "grad_norm": 0.0948140025138855,
+      "learning_rate": 2.089839545420411e-05,
+      "loss": 9.3753,
+      "step": 61100
+    },
+    {
+      "epoch": 0.30517615920497393,
+      "grad_norm": 0.09809758514165878,
+      "learning_rate": 2.089689353926256e-05,
+      "loss": 9.3902,
+      "step": 61110
+    },
+    {
+      "epoch": 0.30522609802991335,
+      "grad_norm": 0.09343090653419495,
+      "learning_rate": 2.089539162432101e-05,
+      "loss": 9.3775,
+      "step": 61120
+    },
+    {
+      "epoch": 0.30527603685485283,
+      "grad_norm": 0.09006762504577637,
+      "learning_rate": 2.0893889709379458e-05,
+      "loss": 9.376,
+      "step": 61130
+    },
+    {
+      "epoch": 0.30532597567979225,
+      "grad_norm": 0.09377524256706238,
+      "learning_rate": 2.0892387794437908e-05,
+      "loss": 9.3931,
+      "step": 61140
+    },
+    {
+      "epoch": 0.3053759145047317,
+      "grad_norm": 0.09025995433330536,
+      "learning_rate": 2.0890885879496358e-05,
+      "loss": 9.3832,
+      "step": 61150
+    },
+    {
+      "epoch": 0.30542585332967115,
+      "grad_norm": 0.09238019585609436,
+      "learning_rate": 2.088938396455481e-05,
+      "loss": 9.3832,
+      "step": 61160
+    },
+    {
+      "epoch": 0.3054757921546106,
+      "grad_norm": 0.0937105119228363,
+      "learning_rate": 2.088788204961326e-05,
+      "loss": 9.3737,
+      "step": 61170
+    },
+    {
+      "epoch": 0.30552573097955005,
+      "grad_norm": 0.10375673323869705,
+      "learning_rate": 2.0886380134671705e-05,
+      "loss": 9.3801,
+      "step": 61180
+    },
+    {
+      "epoch": 0.3055756698044895,
+      "grad_norm": 0.09446416795253754,
+      "learning_rate": 2.0884878219730156e-05,
+      "loss": 9.3759,
+      "step": 61190
+    },
+    {
+      "epoch": 0.30562560862942895,
+      "grad_norm": 0.10232505947351456,
+      "learning_rate": 2.0883376304788606e-05,
+      "loss": 9.3806,
+      "step": 61200
+    },
+    {
+      "epoch": 0.3056755474543684,
+      "grad_norm": 0.09139604866504669,
+      "learning_rate": 2.0881874389847056e-05,
+      "loss": 9.385,
+      "step": 61210
+    },
+    {
+      "epoch": 0.30572548627930785,
+      "grad_norm": 0.09424486756324768,
+      "learning_rate": 2.0880372474905506e-05,
+      "loss": 9.384,
+      "step": 61220
+    },
+    {
+      "epoch": 0.3057754251042473,
+      "grad_norm": 0.10143927484750748,
+      "learning_rate": 2.0878870559963956e-05,
+      "loss": 9.3777,
+      "step": 61230
+    },
+    {
+      "epoch": 0.30582536392918674,
+      "grad_norm": 0.09673741459846497,
+      "learning_rate": 2.0877368645022403e-05,
+      "loss": 9.3755,
+      "step": 61240
+    },
+    {
+      "epoch": 0.3058753027541262,
+      "grad_norm": 0.08992431312799454,
+      "learning_rate": 2.0875866730080853e-05,
+      "loss": 9.3668,
+      "step": 61250
+    },
+    {
+      "epoch": 0.30592524157906564,
+      "grad_norm": 0.0944616049528122,
+      "learning_rate": 2.0874364815139303e-05,
+      "loss": 9.3816,
+      "step": 61260
+    },
+    {
+      "epoch": 0.3059751804040051,
+      "grad_norm": 0.09414882957935333,
+      "learning_rate": 2.0872862900197754e-05,
+      "loss": 9.3888,
+      "step": 61270
+    },
+    {
+      "epoch": 0.30602511922894454,
+      "grad_norm": 0.0981215313076973,
+      "learning_rate": 2.0871360985256204e-05,
+      "loss": 9.3905,
+      "step": 61280
+    },
+    {
+      "epoch": 0.306075058053884,
+      "grad_norm": 0.0904826819896698,
+      "learning_rate": 2.086985907031465e-05,
+      "loss": 9.3693,
+      "step": 61290
+    },
+    {
+      "epoch": 0.30612499687882344,
+      "grad_norm": 0.09661656618118286,
+      "learning_rate": 2.08683571553731e-05,
+      "loss": 9.3803,
+      "step": 61300
+    },
+    {
+      "epoch": 0.30617493570376286,
+      "grad_norm": 0.09500022977590561,
+      "learning_rate": 2.086685524043155e-05,
+      "loss": 9.369,
+      "step": 61310
+    },
+    {
+      "epoch": 0.30622487452870234,
+      "grad_norm": 0.0918661504983902,
+      "learning_rate": 2.086535332549e-05,
+      "loss": 9.3814,
+      "step": 61320
+    },
+    {
+      "epoch": 0.30627481335364176,
+      "grad_norm": 0.09368791431188583,
+      "learning_rate": 2.086385141054845e-05,
+      "loss": 9.3804,
+      "step": 61330
+    },
+    {
+      "epoch": 0.30632475217858124,
+      "grad_norm": 0.09218349307775497,
+      "learning_rate": 2.0862349495606898e-05,
+      "loss": 9.3819,
+      "step": 61340
+    },
+    {
+      "epoch": 0.30637469100352066,
+      "grad_norm": 0.10055679827928543,
+      "learning_rate": 2.0860847580665348e-05,
+      "loss": 9.3934,
+      "step": 61350
+    },
+    {
+      "epoch": 0.30642462982846014,
+      "grad_norm": 0.09349419176578522,
+      "learning_rate": 2.08593456657238e-05,
+      "loss": 9.3864,
+      "step": 61360
+    },
+    {
+      "epoch": 0.30647456865339956,
+      "grad_norm": 0.09518557041883469,
+      "learning_rate": 2.085784375078225e-05,
+      "loss": 9.3871,
+      "step": 61370
+    },
+    {
+      "epoch": 0.30652450747833904,
+      "grad_norm": 0.09079370647668839,
+      "learning_rate": 2.08563418358407e-05,
+      "loss": 9.3785,
+      "step": 61380
+    },
+    {
+      "epoch": 0.30657444630327846,
+      "grad_norm": 0.0915151834487915,
+      "learning_rate": 2.0854839920899146e-05,
+      "loss": 9.3852,
+      "step": 61390
+    },
+    {
+      "epoch": 0.30662438512821794,
+      "grad_norm": 0.09642606973648071,
+      "learning_rate": 2.0853338005957596e-05,
+      "loss": 9.3797,
+      "step": 61400
+    },
+    {
+      "epoch": 0.30667432395315736,
+      "grad_norm": 0.08891963213682175,
+      "learning_rate": 2.0851836091016046e-05,
+      "loss": 9.382,
+      "step": 61410
+    },
+    {
+      "epoch": 0.30672426277809683,
+      "grad_norm": 0.0949094146490097,
+      "learning_rate": 2.0850334176074496e-05,
+      "loss": 9.3788,
+      "step": 61420
+    },
+    {
+      "epoch": 0.30677420160303626,
+      "grad_norm": 0.0930948406457901,
+      "learning_rate": 2.0848832261132946e-05,
+      "loss": 9.3803,
+      "step": 61430
+    },
+    {
+      "epoch": 0.30682414042797573,
+      "grad_norm": 0.09792865067720413,
+      "learning_rate": 2.0847330346191393e-05,
+      "loss": 9.3795,
+      "step": 61440
+    },
+    {
+      "epoch": 0.30687407925291516,
+      "grad_norm": 0.09721236675977707,
+      "learning_rate": 2.0845828431249843e-05,
+      "loss": 9.3869,
+      "step": 61450
+    },
+    {
+      "epoch": 0.30692401807785463,
+      "grad_norm": 0.095427006483078,
+      "learning_rate": 2.0844326516308293e-05,
+      "loss": 9.3857,
+      "step": 61460
+    },
+    {
+      "epoch": 0.30697395690279405,
+      "grad_norm": 0.093628890812397,
+      "learning_rate": 2.0842824601366744e-05,
+      "loss": 9.3756,
+      "step": 61470
+    },
+    {
+      "epoch": 0.30702389572773353,
+      "grad_norm": 0.09637012332677841,
+      "learning_rate": 2.0841322686425194e-05,
+      "loss": 9.3804,
+      "step": 61480
+    },
+    {
+      "epoch": 0.30707383455267295,
+      "grad_norm": 0.09300060570240021,
+      "learning_rate": 2.083982077148364e-05,
+      "loss": 9.3687,
+      "step": 61490
+    },
+    {
+      "epoch": 0.30712377337761243,
+      "grad_norm": 0.08992978185415268,
+      "learning_rate": 2.083831885654209e-05,
+      "loss": 9.382,
+      "step": 61500
+    },
+    {
+      "epoch": 0.30717371220255185,
+      "grad_norm": 0.09597385674715042,
+      "learning_rate": 2.083681694160054e-05,
+      "loss": 9.379,
+      "step": 61510
+    },
+    {
+      "epoch": 0.30722365102749133,
+      "grad_norm": 0.09319521486759186,
+      "learning_rate": 2.083531502665899e-05,
+      "loss": 9.3809,
+      "step": 61520
+    },
+    {
+      "epoch": 0.30727358985243075,
+      "grad_norm": 0.09977719187736511,
+      "learning_rate": 2.083381311171744e-05,
+      "loss": 9.3695,
+      "step": 61530
+    },
+    {
+      "epoch": 0.30732352867737023,
+      "grad_norm": 0.09011849015951157,
+      "learning_rate": 2.0832311196775888e-05,
+      "loss": 9.3751,
+      "step": 61540
+    },
+    {
+      "epoch": 0.30737346750230965,
+      "grad_norm": 0.09430202841758728,
+      "learning_rate": 2.083080928183434e-05,
+      "loss": 9.3763,
+      "step": 61550
+    },
+    {
+      "epoch": 0.3074234063272491,
+      "grad_norm": 0.09825265407562256,
+      "learning_rate": 2.082930736689279e-05,
+      "loss": 9.3643,
+      "step": 61560
+    },
+    {
+      "epoch": 0.30747334515218855,
+      "grad_norm": 0.09236638993024826,
+      "learning_rate": 2.082780545195124e-05,
+      "loss": 9.3792,
+      "step": 61570
+    },
+    {
+      "epoch": 0.307523283977128,
+      "grad_norm": 0.09510213881731033,
+      "learning_rate": 2.082630353700969e-05,
+      "loss": 9.3854,
+      "step": 61580
+    },
+    {
+      "epoch": 0.30757322280206745,
+      "grad_norm": 0.09195613861083984,
+      "learning_rate": 2.0824801622068136e-05,
+      "loss": 9.3655,
+      "step": 61590
+    },
+    {
+      "epoch": 0.3076231616270069,
+      "grad_norm": 0.09247229248285294,
+      "learning_rate": 2.082329970712659e-05,
+      "loss": 9.3756,
+      "step": 61600
+    },
+    {
+      "epoch": 0.30767310045194635,
+      "grad_norm": 0.08966951817274094,
+      "learning_rate": 2.0821797792185036e-05,
+      "loss": 9.379,
+      "step": 61610
+    },
+    {
+      "epoch": 0.3077230392768858,
+      "grad_norm": 0.10270446538925171,
+      "learning_rate": 2.0820295877243486e-05,
+      "loss": 9.3774,
+      "step": 61620
+    },
+    {
+      "epoch": 0.30777297810182525,
+      "grad_norm": 0.09327306598424911,
+      "learning_rate": 2.0818793962301936e-05,
+      "loss": 9.3764,
+      "step": 61630
+    },
+    {
+      "epoch": 0.3078229169267647,
+      "grad_norm": 0.09807731211185455,
+      "learning_rate": 2.0817292047360383e-05,
+      "loss": 9.3884,
+      "step": 61640
+    },
+    {
+      "epoch": 0.30787285575170414,
+      "grad_norm": 0.09317595511674881,
+      "learning_rate": 2.0815790132418837e-05,
+      "loss": 9.3817,
+      "step": 61650
+    },
+    {
+      "epoch": 0.3079227945766436,
+      "grad_norm": 0.09404472261667252,
+      "learning_rate": 2.0814288217477283e-05,
+      "loss": 9.3812,
+      "step": 61660
+    },
+    {
+      "epoch": 0.30797273340158304,
+      "grad_norm": 0.0946296826004982,
+      "learning_rate": 2.0812786302535734e-05,
+      "loss": 9.3784,
+      "step": 61670
+    },
+    {
+      "epoch": 0.3080226722265225,
+      "grad_norm": 0.09800449013710022,
+      "learning_rate": 2.0811284387594184e-05,
+      "loss": 9.3691,
+      "step": 61680
+    },
+    {
+      "epoch": 0.30807261105146194,
+      "grad_norm": 0.09789472073316574,
+      "learning_rate": 2.080978247265263e-05,
+      "loss": 9.3718,
+      "step": 61690
+    },
+    {
+      "epoch": 0.3081225498764014,
+      "grad_norm": 0.0944046601653099,
+      "learning_rate": 2.0808280557711084e-05,
+      "loss": 9.3751,
+      "step": 61700
+    },
+    {
+      "epoch": 0.30817248870134084,
+      "grad_norm": 0.09153825044631958,
+      "learning_rate": 2.080677864276953e-05,
+      "loss": 9.3673,
+      "step": 61710
+    },
+    {
+      "epoch": 0.3082224275262803,
+      "grad_norm": 0.09950811415910721,
+      "learning_rate": 2.080527672782798e-05,
+      "loss": 9.3759,
+      "step": 61720
+    },
+    {
+      "epoch": 0.30827236635121974,
+      "grad_norm": 0.09401055425405502,
+      "learning_rate": 2.080377481288643e-05,
+      "loss": 9.3752,
+      "step": 61730
+    },
+    {
+      "epoch": 0.3083223051761592,
+      "grad_norm": 0.09673509001731873,
+      "learning_rate": 2.0802272897944878e-05,
+      "loss": 9.3674,
+      "step": 61740
+    },
+    {
+      "epoch": 0.30837224400109864,
+      "grad_norm": 0.09277281165122986,
+      "learning_rate": 2.080077098300333e-05,
+      "loss": 9.3649,
+      "step": 61750
+    },
+    {
+      "epoch": 0.3084221828260381,
+      "grad_norm": 0.09555141627788544,
+      "learning_rate": 2.079926906806178e-05,
+      "loss": 9.3666,
+      "step": 61760
+    },
+    {
+      "epoch": 0.30847212165097754,
+      "grad_norm": 0.09273938834667206,
+      "learning_rate": 2.079776715312023e-05,
+      "loss": 9.3702,
+      "step": 61770
+    },
+    {
+      "epoch": 0.308522060475917,
+      "grad_norm": 0.09557077288627625,
+      "learning_rate": 2.079626523817868e-05,
+      "loss": 9.3766,
+      "step": 61780
+    },
+    {
+      "epoch": 0.30857199930085644,
+      "grad_norm": 0.09353851526975632,
+      "learning_rate": 2.0794763323237126e-05,
+      "loss": 9.3753,
+      "step": 61790
+    },
+    {
+      "epoch": 0.3086219381257959,
+      "grad_norm": 0.09252606332302094,
+      "learning_rate": 2.079326140829558e-05,
+      "loss": 9.3661,
+      "step": 61800
+    },
+    {
+      "epoch": 0.30867187695073534,
+      "grad_norm": 0.09201821684837341,
+      "learning_rate": 2.0791759493354026e-05,
+      "loss": 9.3726,
+      "step": 61810
+    },
+    {
+      "epoch": 0.3087218157756748,
+      "grad_norm": 0.09545756131410599,
+      "learning_rate": 2.0790257578412476e-05,
+      "loss": 9.3795,
+      "step": 61820
+    },
+    {
+      "epoch": 0.30877175460061423,
+      "grad_norm": 0.0946696475148201,
+      "learning_rate": 2.0788755663470926e-05,
+      "loss": 9.3637,
+      "step": 61830
+    },
+    {
+      "epoch": 0.3088216934255537,
+      "grad_norm": 0.09465292096138,
+      "learning_rate": 2.0787253748529373e-05,
+      "loss": 9.3803,
+      "step": 61840
+    },
+    {
+      "epoch": 0.30887163225049313,
+      "grad_norm": 0.09846868366003036,
+      "learning_rate": 2.0785751833587827e-05,
+      "loss": 9.3652,
+      "step": 61850
+    },
+    {
+      "epoch": 0.3089215710754326,
+      "grad_norm": 0.09574449062347412,
+      "learning_rate": 2.0784249918646273e-05,
+      "loss": 9.3696,
+      "step": 61860
+    },
+    {
+      "epoch": 0.30897150990037203,
+      "grad_norm": 0.09601958841085434,
+      "learning_rate": 2.0782748003704724e-05,
+      "loss": 9.3682,
+      "step": 61870
+    },
+    {
+      "epoch": 0.3090214487253115,
+      "grad_norm": 0.0923629179596901,
+      "learning_rate": 2.0781246088763174e-05,
+      "loss": 9.3626,
+      "step": 61880
+    },
+    {
+      "epoch": 0.30907138755025093,
+      "grad_norm": 0.09300635010004044,
+      "learning_rate": 2.077974417382162e-05,
+      "loss": 9.3663,
+      "step": 61890
+    },
+    {
+      "epoch": 0.3091213263751904,
+      "grad_norm": 0.09495260566473007,
+      "learning_rate": 2.0778242258880074e-05,
+      "loss": 9.3814,
+      "step": 61900
+    },
+    {
+      "epoch": 0.30917126520012983,
+      "grad_norm": 0.0935489758849144,
+      "learning_rate": 2.077674034393852e-05,
+      "loss": 9.3656,
+      "step": 61910
+    },
+    {
+      "epoch": 0.3092212040250693,
+      "grad_norm": 0.09819766879081726,
+      "learning_rate": 2.0775238428996975e-05,
+      "loss": 9.3711,
+      "step": 61920
+    },
+    {
+      "epoch": 0.30927114285000873,
+      "grad_norm": 0.09066694229841232,
+      "learning_rate": 2.077373651405542e-05,
+      "loss": 9.3818,
+      "step": 61930
+    },
+    {
+      "epoch": 0.3093210816749482,
+      "grad_norm": 0.09707902371883392,
+      "learning_rate": 2.0772234599113868e-05,
+      "loss": 9.3693,
+      "step": 61940
+    },
+    {
+      "epoch": 0.30937102049988763,
+      "grad_norm": 0.09382162988185883,
+      "learning_rate": 2.077073268417232e-05,
+      "loss": 9.3681,
+      "step": 61950
+    },
+    {
+      "epoch": 0.3094209593248271,
+      "grad_norm": 0.09599854052066803,
+      "learning_rate": 2.076923076923077e-05,
+      "loss": 9.3744,
+      "step": 61960
+    },
+    {
+      "epoch": 0.3094708981497665,
+      "grad_norm": 0.09321361035108566,
+      "learning_rate": 2.0767728854289222e-05,
+      "loss": 9.3718,
+      "step": 61970
+    },
+    {
+      "epoch": 0.309520836974706,
+      "grad_norm": 0.09474252909421921,
+      "learning_rate": 2.076622693934767e-05,
+      "loss": 9.3747,
+      "step": 61980
+    },
+    {
+      "epoch": 0.3095707757996454,
+      "grad_norm": 0.09249020367860794,
+      "learning_rate": 2.0764725024406116e-05,
+      "loss": 9.3677,
+      "step": 61990
+    },
+    {
+      "epoch": 0.3096207146245849,
+      "grad_norm": 0.09276629239320755,
+      "learning_rate": 2.076322310946457e-05,
+      "loss": 9.375,
+      "step": 62000
+    },
+    {
+      "epoch": 0.3096706534495243,
+      "grad_norm": 0.09531158208847046,
+      "learning_rate": 2.0761721194523016e-05,
+      "loss": 9.3636,
+      "step": 62010
+    },
+    {
+      "epoch": 0.3097205922744638,
+      "grad_norm": 0.09786634147167206,
+      "learning_rate": 2.076021927958147e-05,
+      "loss": 9.3572,
+      "step": 62020
+    },
+    {
+      "epoch": 0.3097705310994032,
+      "grad_norm": 0.09066082537174225,
+      "learning_rate": 2.0758717364639916e-05,
+      "loss": 9.3621,
+      "step": 62030
+    },
+    {
+      "epoch": 0.3098204699243427,
+      "grad_norm": 0.09768658876419067,
+      "learning_rate": 2.0757215449698363e-05,
+      "loss": 9.3643,
+      "step": 62040
+    },
+    {
+      "epoch": 0.3098704087492821,
+      "grad_norm": 0.09411395341157913,
+      "learning_rate": 2.0755713534756817e-05,
+      "loss": 9.3757,
+      "step": 62050
+    },
+    {
+      "epoch": 0.3099203475742216,
+      "grad_norm": 0.09201698750257492,
+      "learning_rate": 2.0754211619815263e-05,
+      "loss": 9.3652,
+      "step": 62060
+    },
+    {
+      "epoch": 0.309970286399161,
+      "grad_norm": 0.09616990387439728,
+      "learning_rate": 2.0752709704873717e-05,
+      "loss": 9.3675,
+      "step": 62070
+    },
+    {
+      "epoch": 0.3100202252241005,
+      "grad_norm": 0.09069666266441345,
+      "learning_rate": 2.0751207789932164e-05,
+      "loss": 9.3703,
+      "step": 62080
+    },
+    {
+      "epoch": 0.3100701640490399,
+      "grad_norm": 0.09700020402669907,
+      "learning_rate": 2.074970587499061e-05,
+      "loss": 9.3644,
+      "step": 62090
+    },
+    {
+      "epoch": 0.3101201028739794,
+      "grad_norm": 0.10156318545341492,
+      "learning_rate": 2.0748203960049064e-05,
+      "loss": 9.3651,
+      "step": 62100
+    },
+    {
+      "epoch": 0.3101700416989188,
+      "grad_norm": 0.09886864572763443,
+      "learning_rate": 2.074670204510751e-05,
+      "loss": 9.3718,
+      "step": 62110
+    },
+    {
+      "epoch": 0.3102199805238583,
+      "grad_norm": 0.09852635115385056,
+      "learning_rate": 2.0745200130165965e-05,
+      "loss": 9.3617,
+      "step": 62120
+    },
+    {
+      "epoch": 0.3102699193487977,
+      "grad_norm": 0.09089913219213486,
+      "learning_rate": 2.074369821522441e-05,
+      "loss": 9.3574,
+      "step": 62130
+    },
+    {
+      "epoch": 0.3103198581737372,
+      "grad_norm": 0.08944336324930191,
+      "learning_rate": 2.0742196300282858e-05,
+      "loss": 9.3706,
+      "step": 62140
+    },
+    {
+      "epoch": 0.3103697969986766,
+      "grad_norm": 0.09155122190713882,
+      "learning_rate": 2.074069438534131e-05,
+      "loss": 9.3682,
+      "step": 62150
+    },
+    {
+      "epoch": 0.3104197358236161,
+      "grad_norm": 0.09597411006689072,
+      "learning_rate": 2.073919247039976e-05,
+      "loss": 9.3717,
+      "step": 62160
+    },
+    {
+      "epoch": 0.3104696746485555,
+      "grad_norm": 0.09866214543581009,
+      "learning_rate": 2.0737690555458212e-05,
+      "loss": 9.3543,
+      "step": 62170
+    },
+    {
+      "epoch": 0.310519613473495,
+      "grad_norm": 0.09262767434120178,
+      "learning_rate": 2.073618864051666e-05,
+      "loss": 9.3623,
+      "step": 62180
+    },
+    {
+      "epoch": 0.3105695522984344,
+      "grad_norm": 0.08718739449977875,
+      "learning_rate": 2.0734686725575106e-05,
+      "loss": 9.3704,
+      "step": 62190
+    },
+    {
+      "epoch": 0.3106194911233739,
+      "grad_norm": 0.08802207559347153,
+      "learning_rate": 2.073318481063356e-05,
+      "loss": 9.3702,
+      "step": 62200
+    },
+    {
+      "epoch": 0.3106694299483133,
+      "grad_norm": 0.09451784193515778,
+      "learning_rate": 2.0731682895692006e-05,
+      "loss": 9.3645,
+      "step": 62210
+    },
+    {
+      "epoch": 0.3107193687732528,
+      "grad_norm": 0.09479044377803802,
+      "learning_rate": 2.073018098075046e-05,
+      "loss": 9.3502,
+      "step": 62220
+    },
+    {
+      "epoch": 0.3107693075981922,
+      "grad_norm": 0.09419619292020798,
+      "learning_rate": 2.0728679065808906e-05,
+      "loss": 9.3656,
+      "step": 62230
+    },
+    {
+      "epoch": 0.3108192464231317,
+      "grad_norm": 0.09506448358297348,
+      "learning_rate": 2.0727177150867357e-05,
+      "loss": 9.3581,
+      "step": 62240
+    },
+    {
+      "epoch": 0.3108691852480711,
+      "grad_norm": 0.09851628541946411,
+      "learning_rate": 2.0725675235925807e-05,
+      "loss": 9.362,
+      "step": 62250
+    },
+    {
+      "epoch": 0.3109191240730106,
+      "grad_norm": 0.0938975065946579,
+      "learning_rate": 2.0724173320984253e-05,
+      "loss": 9.3594,
+      "step": 62260
+    },
+    {
+      "epoch": 0.31096906289795,
+      "grad_norm": 0.0952107235789299,
+      "learning_rate": 2.0722671406042707e-05,
+      "loss": 9.3598,
+      "step": 62270
+    },
+    {
+      "epoch": 0.3110190017228895,
+      "grad_norm": 0.09428625553846359,
+      "learning_rate": 2.0721169491101154e-05,
+      "loss": 9.3688,
+      "step": 62280
+    },
+    {
+      "epoch": 0.3110689405478289,
+      "grad_norm": 0.09929768741130829,
+      "learning_rate": 2.0719667576159604e-05,
+      "loss": 9.3563,
+      "step": 62290
+    },
+    {
+      "epoch": 0.31111887937276833,
+      "grad_norm": 0.09252603352069855,
+      "learning_rate": 2.0718165661218054e-05,
+      "loss": 9.3555,
+      "step": 62300
+    },
+    {
+      "epoch": 0.3111688181977078,
+      "grad_norm": 0.0918884426355362,
+      "learning_rate": 2.07166637462765e-05,
+      "loss": 9.3604,
+      "step": 62310
+    },
+    {
+      "epoch": 0.31121875702264723,
+      "grad_norm": 0.09412012249231339,
+      "learning_rate": 2.0715161831334955e-05,
+      "loss": 9.36,
+      "step": 62320
+    },
+    {
+      "epoch": 0.3112686958475867,
+      "grad_norm": 0.08864130824804306,
+      "learning_rate": 2.07136599163934e-05,
+      "loss": 9.3741,
+      "step": 62330
+    },
+    {
+      "epoch": 0.31131863467252613,
+      "grad_norm": 0.09902546554803848,
+      "learning_rate": 2.071215800145185e-05,
+      "loss": 9.3707,
+      "step": 62340
+    },
+    {
+      "epoch": 0.3113685734974656,
+      "grad_norm": 0.0945492535829544,
+      "learning_rate": 2.07106560865103e-05,
+      "loss": 9.362,
+      "step": 62350
+    },
+    {
+      "epoch": 0.31141851232240503,
+      "grad_norm": 0.0934128388762474,
+      "learning_rate": 2.070915417156875e-05,
+      "loss": 9.3579,
+      "step": 62360
+    },
+    {
+      "epoch": 0.3114684511473445,
+      "grad_norm": 0.10103520005941391,
+      "learning_rate": 2.0707652256627202e-05,
+      "loss": 9.3705,
+      "step": 62370
+    },
+    {
+      "epoch": 0.3115183899722839,
+      "grad_norm": 0.09728195518255234,
+      "learning_rate": 2.070615034168565e-05,
+      "loss": 9.3674,
+      "step": 62380
+    },
+    {
+      "epoch": 0.3115683287972234,
+      "grad_norm": 0.0965341180562973,
+      "learning_rate": 2.07046484267441e-05,
+      "loss": 9.368,
+      "step": 62390
+    },
+    {
+      "epoch": 0.3116182676221628,
+      "grad_norm": 0.09560953825712204,
+      "learning_rate": 2.070314651180255e-05,
+      "loss": 9.3595,
+      "step": 62400
+    },
+    {
+      "epoch": 0.3116682064471023,
+      "grad_norm": 0.09266313165426254,
+      "learning_rate": 2.0701644596860996e-05,
+      "loss": 9.3655,
+      "step": 62410
+    },
+    {
+      "epoch": 0.3117181452720417,
+      "grad_norm": 0.09788119792938232,
+      "learning_rate": 2.070014268191945e-05,
+      "loss": 9.3602,
+      "step": 62420
+    },
+    {
+      "epoch": 0.3117680840969812,
+      "grad_norm": 0.09870301932096481,
+      "learning_rate": 2.0698640766977896e-05,
+      "loss": 9.3552,
+      "step": 62430
+    },
+    {
+      "epoch": 0.3118180229219206,
+      "grad_norm": 0.09183664619922638,
+      "learning_rate": 2.0697138852036347e-05,
+      "loss": 9.3642,
+      "step": 62440
+    },
+    {
+      "epoch": 0.3118679617468601,
+      "grad_norm": 0.09062953293323517,
+      "learning_rate": 2.0695636937094797e-05,
+      "loss": 9.3599,
+      "step": 62450
+    },
+    {
+      "epoch": 0.3119179005717995,
+      "grad_norm": 0.08905572444200516,
+      "learning_rate": 2.0694135022153243e-05,
+      "loss": 9.3629,
+      "step": 62460
+    },
+    {
+      "epoch": 0.311967839396739,
+      "grad_norm": 0.08906285464763641,
+      "learning_rate": 2.0692633107211697e-05,
+      "loss": 9.3675,
+      "step": 62470
+    },
+    {
+      "epoch": 0.3120177782216784,
+      "grad_norm": 0.09582462906837463,
+      "learning_rate": 2.0691131192270144e-05,
+      "loss": 9.3733,
+      "step": 62480
+    },
+    {
+      "epoch": 0.3120677170466179,
+      "grad_norm": 0.09110390394926071,
+      "learning_rate": 2.0689629277328594e-05,
+      "loss": 9.3662,
+      "step": 62490
+    },
+    {
+      "epoch": 0.3121176558715573,
+      "grad_norm": 0.09354531019926071,
+      "learning_rate": 2.0688127362387044e-05,
+      "loss": 9.3623,
+      "step": 62500
+    },
+    {
+      "epoch": 0.3121675946964968,
+      "grad_norm": 0.0941261276602745,
+      "learning_rate": 2.068662544744549e-05,
+      "loss": 9.3616,
+      "step": 62510
+    },
+    {
+      "epoch": 0.3122175335214362,
+      "grad_norm": 0.09080076217651367,
+      "learning_rate": 2.0685123532503945e-05,
+      "loss": 9.3641,
+      "step": 62520
+    },
+    {
+      "epoch": 0.3122674723463757,
+      "grad_norm": 0.09199047833681107,
+      "learning_rate": 2.068362161756239e-05,
+      "loss": 9.3541,
+      "step": 62530
+    },
+    {
+      "epoch": 0.3123174111713151,
+      "grad_norm": 0.09623630344867706,
+      "learning_rate": 2.068211970262084e-05,
+      "loss": 9.3581,
+      "step": 62540
+    },
+    {
+      "epoch": 0.3123673499962546,
+      "grad_norm": 0.09023137390613556,
+      "learning_rate": 2.0680617787679292e-05,
+      "loss": 9.3669,
+      "step": 62550
+    },
+    {
+      "epoch": 0.312417288821194,
+      "grad_norm": 0.09632597118616104,
+      "learning_rate": 2.0679115872737742e-05,
+      "loss": 9.3582,
+      "step": 62560
+    },
+    {
+      "epoch": 0.3124672276461335,
+      "grad_norm": 0.09074317663908005,
+      "learning_rate": 2.0677613957796192e-05,
+      "loss": 9.3596,
+      "step": 62570
+    },
+    {
+      "epoch": 0.3125171664710729,
+      "grad_norm": 0.09316986799240112,
+      "learning_rate": 2.067611204285464e-05,
+      "loss": 9.3562,
+      "step": 62580
+    },
+    {
+      "epoch": 0.3125671052960124,
+      "grad_norm": 0.09124487638473511,
+      "learning_rate": 2.067461012791309e-05,
+      "loss": 9.3615,
+      "step": 62590
+    },
+    {
+      "epoch": 0.3126170441209518,
+      "grad_norm": 0.09576891362667084,
+      "learning_rate": 2.067310821297154e-05,
+      "loss": 9.3529,
+      "step": 62600
+    },
+    {
+      "epoch": 0.3126669829458913,
+      "grad_norm": 0.09395425021648407,
+      "learning_rate": 2.067160629802999e-05,
+      "loss": 9.3519,
+      "step": 62610
+    },
+    {
+      "epoch": 0.3127169217708307,
+      "grad_norm": 0.09081248939037323,
+      "learning_rate": 2.067010438308844e-05,
+      "loss": 9.3598,
+      "step": 62620
+    },
+    {
+      "epoch": 0.3127668605957702,
+      "grad_norm": 0.09516331553459167,
+      "learning_rate": 2.0668602468146886e-05,
+      "loss": 9.3653,
+      "step": 62630
+    },
+    {
+      "epoch": 0.3128167994207096,
+      "grad_norm": 0.09487758576869965,
+      "learning_rate": 2.0667100553205337e-05,
+      "loss": 9.3517,
+      "step": 62640
+    },
+    {
+      "epoch": 0.3128667382456491,
+      "grad_norm": 0.0971754714846611,
+      "learning_rate": 2.0665598638263787e-05,
+      "loss": 9.3476,
+      "step": 62650
+    },
+    {
+      "epoch": 0.3129166770705885,
+      "grad_norm": 0.09306435286998749,
+      "learning_rate": 2.0664096723322237e-05,
+      "loss": 9.3561,
+      "step": 62660
+    },
+    {
+      "epoch": 0.312966615895528,
+      "grad_norm": 0.0970197319984436,
+      "learning_rate": 2.0662594808380687e-05,
+      "loss": 9.3684,
+      "step": 62670
+    },
+    {
+      "epoch": 0.3130165547204674,
+      "grad_norm": 0.09373274445533752,
+      "learning_rate": 2.0661092893439134e-05,
+      "loss": 9.3663,
+      "step": 62680
+    },
+    {
+      "epoch": 0.3130664935454069,
+      "grad_norm": 0.09603800624608994,
+      "learning_rate": 2.0659590978497584e-05,
+      "loss": 9.3468,
+      "step": 62690
+    },
+    {
+      "epoch": 0.3131164323703463,
+      "grad_norm": 0.09086347371339798,
+      "learning_rate": 2.0658089063556034e-05,
+      "loss": 9.3515,
+      "step": 62700
+    },
+    {
+      "epoch": 0.3131663711952858,
+      "grad_norm": 0.09210800379514694,
+      "learning_rate": 2.0656587148614484e-05,
+      "loss": 9.3627,
+      "step": 62710
+    },
+    {
+      "epoch": 0.3132163100202252,
+      "grad_norm": 0.09706009924411774,
+      "learning_rate": 2.0655085233672935e-05,
+      "loss": 9.3576,
+      "step": 62720
+    },
+    {
+      "epoch": 0.3132662488451647,
+      "grad_norm": 0.09735792875289917,
+      "learning_rate": 2.065358331873138e-05,
+      "loss": 9.3584,
+      "step": 62730
+    },
+    {
+      "epoch": 0.3133161876701041,
+      "grad_norm": 0.08968986570835114,
+      "learning_rate": 2.065208140378983e-05,
+      "loss": 9.3586,
+      "step": 62740
+    },
+    {
+      "epoch": 0.3133661264950436,
+      "grad_norm": 0.09612874686717987,
+      "learning_rate": 2.0650579488848282e-05,
+      "loss": 9.3559,
+      "step": 62750
+    },
+    {
+      "epoch": 0.313416065319983,
+      "grad_norm": 0.10443656146526337,
+      "learning_rate": 2.0649077573906732e-05,
+      "loss": 9.3511,
+      "step": 62760
+    },
+    {
+      "epoch": 0.3134660041449225,
+      "grad_norm": 0.09903521835803986,
+      "learning_rate": 2.0647575658965182e-05,
+      "loss": 9.3614,
+      "step": 62770
+    },
+    {
+      "epoch": 0.3135159429698619,
+      "grad_norm": 0.09267803281545639,
+      "learning_rate": 2.064607374402363e-05,
+      "loss": 9.3444,
+      "step": 62780
+    },
+    {
+      "epoch": 0.3135658817948014,
+      "grad_norm": 0.09199144691228867,
+      "learning_rate": 2.064457182908208e-05,
+      "loss": 9.3585,
+      "step": 62790
+    },
+    {
+      "epoch": 0.3136158206197408,
+      "grad_norm": 0.09724773466587067,
+      "learning_rate": 2.064306991414053e-05,
+      "loss": 9.352,
+      "step": 62800
+    },
+    {
+      "epoch": 0.3136657594446803,
+      "grad_norm": 0.10160032659769058,
+      "learning_rate": 2.064156799919898e-05,
+      "loss": 9.3515,
+      "step": 62810
+    },
+    {
+      "epoch": 0.3137156982696197,
+      "grad_norm": 0.09469520300626755,
+      "learning_rate": 2.064006608425743e-05,
+      "loss": 9.3425,
+      "step": 62820
+    },
+    {
+      "epoch": 0.3137656370945592,
+      "grad_norm": 0.09688717126846313,
+      "learning_rate": 2.0638564169315876e-05,
+      "loss": 9.3486,
+      "step": 62830
+    },
+    {
+      "epoch": 0.3138155759194986,
+      "grad_norm": 0.09648570418357849,
+      "learning_rate": 2.0637062254374327e-05,
+      "loss": 9.365,
+      "step": 62840
+    },
+    {
+      "epoch": 0.3138655147444381,
+      "grad_norm": 0.09986833482980728,
+      "learning_rate": 2.0635560339432777e-05,
+      "loss": 9.3472,
+      "step": 62850
+    },
+    {
+      "epoch": 0.3139154535693775,
+      "grad_norm": 0.09859945625066757,
+      "learning_rate": 2.0634058424491227e-05,
+      "loss": 9.3553,
+      "step": 62860
+    },
+    {
+      "epoch": 0.313965392394317,
+      "grad_norm": 0.09469161182641983,
+      "learning_rate": 2.0632556509549677e-05,
+      "loss": 9.351,
+      "step": 62870
+    },
+    {
+      "epoch": 0.3140153312192564,
+      "grad_norm": 0.10031906515359879,
+      "learning_rate": 2.0631054594608127e-05,
+      "loss": 9.3604,
+      "step": 62880
+    },
+    {
+      "epoch": 0.3140652700441959,
+      "grad_norm": 0.09583910554647446,
+      "learning_rate": 2.0629552679666574e-05,
+      "loss": 9.3615,
+      "step": 62890
+    },
+    {
+      "epoch": 0.3141152088691353,
+      "grad_norm": 0.09656809270381927,
+      "learning_rate": 2.0628050764725024e-05,
+      "loss": 9.3522,
+      "step": 62900
+    },
+    {
+      "epoch": 0.3141651476940748,
+      "grad_norm": 0.08953146636486053,
+      "learning_rate": 2.0626548849783474e-05,
+      "loss": 9.3564,
+      "step": 62910
+    },
+    {
+      "epoch": 0.3142150865190142,
+      "grad_norm": 0.09325104206800461,
+      "learning_rate": 2.0625046934841925e-05,
+      "loss": 9.3426,
+      "step": 62920
+    },
+    {
+      "epoch": 0.3142650253439537,
+      "grad_norm": 0.09127029776573181,
+      "learning_rate": 2.0623545019900375e-05,
+      "loss": 9.3543,
+      "step": 62930
+    },
+    {
+      "epoch": 0.3143149641688931,
+      "grad_norm": 0.09044724702835083,
+      "learning_rate": 2.062204310495882e-05,
+      "loss": 9.3675,
+      "step": 62940
+    },
+    {
+      "epoch": 0.3143649029938326,
+      "grad_norm": 0.0922994613647461,
+      "learning_rate": 2.0620541190017272e-05,
+      "loss": 9.3608,
+      "step": 62950
+    },
+    {
+      "epoch": 0.314414841818772,
+      "grad_norm": 0.09617674350738525,
+      "learning_rate": 2.0619039275075722e-05,
+      "loss": 9.353,
+      "step": 62960
+    },
+    {
+      "epoch": 0.3144647806437115,
+      "grad_norm": 0.09723023325204849,
+      "learning_rate": 2.0617537360134172e-05,
+      "loss": 9.3553,
+      "step": 62970
+    },
+    {
+      "epoch": 0.3145147194686509,
+      "grad_norm": 0.09201812744140625,
+      "learning_rate": 2.0616035445192622e-05,
+      "loss": 9.3563,
+      "step": 62980
+    },
+    {
+      "epoch": 0.31456465829359037,
+      "grad_norm": 0.09177816659212112,
+      "learning_rate": 2.061453353025107e-05,
+      "loss": 9.3471,
+      "step": 62990
+    },
+    {
+      "epoch": 0.3146145971185298,
+      "grad_norm": 0.09503309428691864,
+      "learning_rate": 2.061303161530952e-05,
+      "loss": 9.3563,
+      "step": 63000
+    },
+    {
+      "epoch": 0.31466453594346927,
+      "grad_norm": 0.0928831547498703,
+      "learning_rate": 2.061152970036797e-05,
+      "loss": 9.3518,
+      "step": 63010
+    },
+    {
+      "epoch": 0.3147144747684087,
+      "grad_norm": 0.09431839734315872,
+      "learning_rate": 2.061002778542642e-05,
+      "loss": 9.3627,
+      "step": 63020
+    },
+    {
+      "epoch": 0.31476441359334817,
+      "grad_norm": 0.08945141732692719,
+      "learning_rate": 2.060852587048487e-05,
+      "loss": 9.3505,
+      "step": 63030
+    },
+    {
+      "epoch": 0.3148143524182876,
+      "grad_norm": 0.09035779535770416,
+      "learning_rate": 2.0607023955543317e-05,
+      "loss": 9.3552,
+      "step": 63040
+    },
+    {
+      "epoch": 0.31486429124322707,
+      "grad_norm": 0.09948616474866867,
+      "learning_rate": 2.0605522040601767e-05,
+      "loss": 9.3376,
+      "step": 63050
+    },
+    {
+      "epoch": 0.3149142300681665,
+      "grad_norm": 0.09741248935461044,
+      "learning_rate": 2.0604020125660217e-05,
+      "loss": 9.343,
+      "step": 63060
+    },
+    {
+      "epoch": 0.31496416889310597,
+      "grad_norm": 0.0959203913807869,
+      "learning_rate": 2.0602518210718667e-05,
+      "loss": 9.3452,
+      "step": 63070
+    },
+    {
+      "epoch": 0.3150141077180454,
+      "grad_norm": 0.09166992455720901,
+      "learning_rate": 2.0601016295777117e-05,
+      "loss": 9.3536,
+      "step": 63080
+    },
+    {
+      "epoch": 0.31506404654298487,
+      "grad_norm": 0.09764059633016586,
+      "learning_rate": 2.0599514380835564e-05,
+      "loss": 9.3498,
+      "step": 63090
+    },
+    {
+      "epoch": 0.3151139853679243,
+      "grad_norm": 0.09447203576564789,
+      "learning_rate": 2.0598012465894014e-05,
+      "loss": 9.3514,
+      "step": 63100
+    },
+    {
+      "epoch": 0.31516392419286376,
+      "grad_norm": 0.09235192835330963,
+      "learning_rate": 2.0596510550952464e-05,
+      "loss": 9.3443,
+      "step": 63110
+    },
+    {
+      "epoch": 0.3152138630178032,
+      "grad_norm": 0.09150667488574982,
+      "learning_rate": 2.0595008636010915e-05,
+      "loss": 9.3494,
+      "step": 63120
+    },
+    {
+      "epoch": 0.31526380184274266,
+      "grad_norm": 0.09310675412416458,
+      "learning_rate": 2.0593506721069365e-05,
+      "loss": 9.3463,
+      "step": 63130
+    },
+    {
+      "epoch": 0.3153137406676821,
+      "grad_norm": 0.09127092361450195,
+      "learning_rate": 2.059200480612781e-05,
+      "loss": 9.3564,
+      "step": 63140
+    },
+    {
+      "epoch": 0.31536367949262156,
+      "grad_norm": 0.09105300903320312,
+      "learning_rate": 2.0590502891186262e-05,
+      "loss": 9.3598,
+      "step": 63150
+    },
+    {
+      "epoch": 0.315413618317561,
+      "grad_norm": 0.09039835631847382,
+      "learning_rate": 2.0589000976244712e-05,
+      "loss": 9.3697,
+      "step": 63160
+    },
+    {
+      "epoch": 0.31546355714250046,
+      "grad_norm": 0.09602472931146622,
+      "learning_rate": 2.0587499061303162e-05,
+      "loss": 9.3542,
+      "step": 63170
+    },
+    {
+      "epoch": 0.3155134959674399,
+      "grad_norm": 0.09388317167758942,
+      "learning_rate": 2.0585997146361612e-05,
+      "loss": 9.3465,
+      "step": 63180
+    },
+    {
+      "epoch": 0.31556343479237936,
+      "grad_norm": 0.092608243227005,
+      "learning_rate": 2.058449523142006e-05,
+      "loss": 9.3552,
+      "step": 63190
+    },
+    {
+      "epoch": 0.3156133736173188,
+      "grad_norm": 0.09442637860774994,
+      "learning_rate": 2.0582993316478513e-05,
+      "loss": 9.3467,
+      "step": 63200
+    },
+    {
+      "epoch": 0.31566331244225826,
+      "grad_norm": 0.09088288247585297,
+      "learning_rate": 2.058149140153696e-05,
+      "loss": 9.3525,
+      "step": 63210
+    },
+    {
+      "epoch": 0.3157132512671977,
+      "grad_norm": 0.09530657529830933,
+      "learning_rate": 2.057998948659541e-05,
+      "loss": 9.3462,
+      "step": 63220
+    },
+    {
+      "epoch": 0.31576319009213716,
+      "grad_norm": 0.09368572384119034,
+      "learning_rate": 2.057848757165386e-05,
+      "loss": 9.3582,
+      "step": 63230
+    },
+    {
+      "epoch": 0.3158131289170766,
+      "grad_norm": 0.09014841169118881,
+      "learning_rate": 2.0576985656712307e-05,
+      "loss": 9.3449,
+      "step": 63240
+    },
+    {
+      "epoch": 0.31586306774201606,
+      "grad_norm": 0.09297845512628555,
+      "learning_rate": 2.057548374177076e-05,
+      "loss": 9.3541,
+      "step": 63250
+    },
+    {
+      "epoch": 0.3159130065669555,
+      "grad_norm": 0.09388581663370132,
+      "learning_rate": 2.0573981826829207e-05,
+      "loss": 9.3532,
+      "step": 63260
+    },
+    {
+      "epoch": 0.31596294539189496,
+      "grad_norm": 0.09536658972501755,
+      "learning_rate": 2.0572479911887657e-05,
+      "loss": 9.3481,
+      "step": 63270
+    },
+    {
+      "epoch": 0.3160128842168344,
+      "grad_norm": 0.09043373167514801,
+      "learning_rate": 2.0570977996946107e-05,
+      "loss": 9.3532,
+      "step": 63280
+    },
+    {
+      "epoch": 0.3160628230417738,
+      "grad_norm": 0.09627131372690201,
+      "learning_rate": 2.0569476082004554e-05,
+      "loss": 9.3527,
+      "step": 63290
+    },
+    {
+      "epoch": 0.3161127618667133,
+      "grad_norm": 0.09719482809305191,
+      "learning_rate": 2.0567974167063008e-05,
+      "loss": 9.348,
+      "step": 63300
+    },
+    {
+      "epoch": 0.3161627006916527,
+      "grad_norm": 0.09590967744588852,
+      "learning_rate": 2.0566472252121454e-05,
+      "loss": 9.3409,
+      "step": 63310
+    },
+    {
+      "epoch": 0.3162126395165922,
+      "grad_norm": 0.09232126921415329,
+      "learning_rate": 2.0564970337179905e-05,
+      "loss": 9.3467,
+      "step": 63320
+    },
+    {
+      "epoch": 0.3162625783415316,
+      "grad_norm": 0.09349498897790909,
+      "learning_rate": 2.0563468422238355e-05,
+      "loss": 9.3443,
+      "step": 63330
+    },
+    {
+      "epoch": 0.3163125171664711,
+      "grad_norm": 0.09130989015102386,
+      "learning_rate": 2.05619665072968e-05,
+      "loss": 9.3482,
+      "step": 63340
+    },
+    {
+      "epoch": 0.3163624559914105,
+      "grad_norm": 0.09377502650022507,
+      "learning_rate": 2.0560464592355255e-05,
+      "loss": 9.3469,
+      "step": 63350
+    },
+    {
+      "epoch": 0.31641239481635,
+      "grad_norm": 0.08837824314832687,
+      "learning_rate": 2.0558962677413702e-05,
+      "loss": 9.3549,
+      "step": 63360
+    },
+    {
+      "epoch": 0.3164623336412894,
+      "grad_norm": 0.08911017328500748,
+      "learning_rate": 2.0557460762472152e-05,
+      "loss": 9.3441,
+      "step": 63370
+    },
+    {
+      "epoch": 0.31651227246622887,
+      "grad_norm": 0.09551316499710083,
+      "learning_rate": 2.0555958847530602e-05,
+      "loss": 9.3406,
+      "step": 63380
+    },
+    {
+      "epoch": 0.3165622112911683,
+      "grad_norm": 0.098643459379673,
+      "learning_rate": 2.0554456932589052e-05,
+      "loss": 9.3425,
+      "step": 63390
+    },
+    {
+      "epoch": 0.31661215011610777,
+      "grad_norm": 0.09688541293144226,
+      "learning_rate": 2.0552955017647503e-05,
+      "loss": 9.3504,
+      "step": 63400
+    },
+    {
+      "epoch": 0.3166620889410472,
+      "grad_norm": 0.0948943942785263,
+      "learning_rate": 2.055145310270595e-05,
+      "loss": 9.3492,
+      "step": 63410
+    },
+    {
+      "epoch": 0.31671202776598667,
+      "grad_norm": 0.09211491793394089,
+      "learning_rate": 2.05499511877644e-05,
+      "loss": 9.3537,
+      "step": 63420
+    },
+    {
+      "epoch": 0.3167619665909261,
+      "grad_norm": 0.0964820459485054,
+      "learning_rate": 2.054844927282285e-05,
+      "loss": 9.339,
+      "step": 63430
+    },
+    {
+      "epoch": 0.31681190541586557,
+      "grad_norm": 0.09166949987411499,
+      "learning_rate": 2.05469473578813e-05,
+      "loss": 9.3545,
+      "step": 63440
+    },
+    {
+      "epoch": 0.316861844240805,
+      "grad_norm": 0.10011352598667145,
+      "learning_rate": 2.054544544293975e-05,
+      "loss": 9.3521,
+      "step": 63450
+    },
+    {
+      "epoch": 0.31691178306574447,
+      "grad_norm": 0.0967571884393692,
+      "learning_rate": 2.0543943527998197e-05,
+      "loss": 9.3413,
+      "step": 63460
+    },
+    {
+      "epoch": 0.3169617218906839,
+      "grad_norm": 0.09536176174879074,
+      "learning_rate": 2.0542441613056647e-05,
+      "loss": 9.3398,
+      "step": 63470
+    },
+    {
+      "epoch": 0.31701166071562337,
+      "grad_norm": 0.09382636100053787,
+      "learning_rate": 2.0540939698115097e-05,
+      "loss": 9.3471,
+      "step": 63480
+    },
+    {
+      "epoch": 0.3170615995405628,
+      "grad_norm": 0.09492380917072296,
+      "learning_rate": 2.0539437783173547e-05,
+      "loss": 9.3293,
+      "step": 63490
+    },
+    {
+      "epoch": 0.31711153836550227,
+      "grad_norm": 0.09516463428735733,
+      "learning_rate": 2.0537935868231998e-05,
+      "loss": 9.3534,
+      "step": 63500
+    },
+    {
+      "epoch": 0.3171614771904417,
+      "grad_norm": 0.09900394827127457,
+      "learning_rate": 2.0536433953290444e-05,
+      "loss": 9.3535,
+      "step": 63510
+    },
+    {
+      "epoch": 0.31721141601538116,
+      "grad_norm": 0.09245557337999344,
+      "learning_rate": 2.0534932038348898e-05,
+      "loss": 9.3483,
+      "step": 63520
+    },
+    {
+      "epoch": 0.3172613548403206,
+      "grad_norm": 0.09061167389154434,
+      "learning_rate": 2.0533430123407345e-05,
+      "loss": 9.348,
+      "step": 63530
+    },
+    {
+      "epoch": 0.31731129366526006,
+      "grad_norm": 0.09653984010219574,
+      "learning_rate": 2.0531928208465795e-05,
+      "loss": 9.344,
+      "step": 63540
+    },
+    {
+      "epoch": 0.3173612324901995,
+      "grad_norm": 0.09344928711652756,
+      "learning_rate": 2.0530426293524245e-05,
+      "loss": 9.355,
+      "step": 63550
+    },
+    {
+      "epoch": 0.31741117131513896,
+      "grad_norm": 0.09327616542577744,
+      "learning_rate": 2.0528924378582692e-05,
+      "loss": 9.3527,
+      "step": 63560
+    },
+    {
+      "epoch": 0.3174611101400784,
+      "grad_norm": 0.088762067258358,
+      "learning_rate": 2.0527422463641146e-05,
+      "loss": 9.3329,
+      "step": 63570
+    },
+    {
+      "epoch": 0.31751104896501786,
+      "grad_norm": 0.09158555418252945,
+      "learning_rate": 2.0525920548699592e-05,
+      "loss": 9.3442,
+      "step": 63580
+    },
+    {
+      "epoch": 0.3175609877899573,
+      "grad_norm": 0.0912177637219429,
+      "learning_rate": 2.0524418633758042e-05,
+      "loss": 9.3425,
+      "step": 63590
+    },
+    {
+      "epoch": 0.31761092661489676,
+      "grad_norm": 0.08927998691797256,
+      "learning_rate": 2.0522916718816493e-05,
+      "loss": 9.3505,
+      "step": 63600
+    },
+    {
+      "epoch": 0.3176608654398362,
+      "grad_norm": 0.09646981954574585,
+      "learning_rate": 2.052141480387494e-05,
+      "loss": 9.338,
+      "step": 63610
+    },
+    {
+      "epoch": 0.31771080426477566,
+      "grad_norm": 0.10072862356901169,
+      "learning_rate": 2.0519912888933393e-05,
+      "loss": 9.3302,
+      "step": 63620
+    },
+    {
+      "epoch": 0.3177607430897151,
+      "grad_norm": 0.09051260352134705,
+      "learning_rate": 2.051841097399184e-05,
+      "loss": 9.342,
+      "step": 63630
+    },
+    {
+      "epoch": 0.31781068191465456,
+      "grad_norm": 0.09302885830402374,
+      "learning_rate": 2.051690905905029e-05,
+      "loss": 9.338,
+      "step": 63640
+    },
+    {
+      "epoch": 0.317860620739594,
+      "grad_norm": 0.0928134098649025,
+      "learning_rate": 2.051540714410874e-05,
+      "loss": 9.3399,
+      "step": 63650
+    },
+    {
+      "epoch": 0.31791055956453346,
+      "grad_norm": 0.09679362177848816,
+      "learning_rate": 2.0513905229167187e-05,
+      "loss": 9.3486,
+      "step": 63660
+    },
+    {
+      "epoch": 0.3179604983894729,
+      "grad_norm": 0.09691763669252396,
+      "learning_rate": 2.051240331422564e-05,
+      "loss": 9.3392,
+      "step": 63670
+    },
+    {
+      "epoch": 0.31801043721441236,
+      "grad_norm": 0.09337911009788513,
+      "learning_rate": 2.0510901399284087e-05,
+      "loss": 9.3413,
+      "step": 63680
+    },
+    {
+      "epoch": 0.3180603760393518,
+      "grad_norm": 0.09015453606843948,
+      "learning_rate": 2.0509399484342538e-05,
+      "loss": 9.3384,
+      "step": 63690
+    },
+    {
+      "epoch": 0.31811031486429125,
+      "grad_norm": 0.09561469405889511,
+      "learning_rate": 2.0507897569400988e-05,
+      "loss": 9.3374,
+      "step": 63700
+    },
+    {
+      "epoch": 0.3181602536892307,
+      "grad_norm": 0.09089536219835281,
+      "learning_rate": 2.0506395654459434e-05,
+      "loss": 9.3469,
+      "step": 63710
+    },
+    {
+      "epoch": 0.31821019251417015,
+      "grad_norm": 0.09029261767864227,
+      "learning_rate": 2.0504893739517888e-05,
+      "loss": 9.3534,
+      "step": 63720
+    },
+    {
+      "epoch": 0.3182601313391096,
+      "grad_norm": 0.092425636947155,
+      "learning_rate": 2.0503391824576335e-05,
+      "loss": 9.3394,
+      "step": 63730
+    },
+    {
+      "epoch": 0.31831007016404905,
+      "grad_norm": 0.09339354932308197,
+      "learning_rate": 2.0501889909634785e-05,
+      "loss": 9.3421,
+      "step": 63740
+    },
+    {
+      "epoch": 0.3183600089889885,
+      "grad_norm": 0.09930525720119476,
+      "learning_rate": 2.0500387994693235e-05,
+      "loss": 9.3387,
+      "step": 63750
+    },
+    {
+      "epoch": 0.31840994781392795,
+      "grad_norm": 0.09377606213092804,
+      "learning_rate": 2.0498886079751682e-05,
+      "loss": 9.3461,
+      "step": 63760
+    },
+    {
+      "epoch": 0.3184598866388674,
+      "grad_norm": 0.09875767678022385,
+      "learning_rate": 2.0497384164810136e-05,
+      "loss": 9.3302,
+      "step": 63770
+    },
+    {
+      "epoch": 0.31850982546380685,
+      "grad_norm": 0.09149225801229477,
+      "learning_rate": 2.0495882249868582e-05,
+      "loss": 9.342,
+      "step": 63780
+    },
+    {
+      "epoch": 0.31855976428874627,
+      "grad_norm": 0.0938776284456253,
+      "learning_rate": 2.0494380334927033e-05,
+      "loss": 9.3441,
+      "step": 63790
+    },
+    {
+      "epoch": 0.31860970311368575,
+      "grad_norm": 0.0972360149025917,
+      "learning_rate": 2.0492878419985483e-05,
+      "loss": 9.3438,
+      "step": 63800
+    },
+    {
+      "epoch": 0.31865964193862517,
+      "grad_norm": 0.09202911704778671,
+      "learning_rate": 2.049137650504393e-05,
+      "loss": 9.3384,
+      "step": 63810
+    },
+    {
+      "epoch": 0.31870958076356465,
+      "grad_norm": 0.09489790350198746,
+      "learning_rate": 2.0489874590102383e-05,
+      "loss": 9.3465,
+      "step": 63820
+    },
+    {
+      "epoch": 0.31875951958850407,
+      "grad_norm": 0.09030698239803314,
+      "learning_rate": 2.048837267516083e-05,
+      "loss": 9.3487,
+      "step": 63830
+    },
+    {
+      "epoch": 0.31880945841344355,
+      "grad_norm": 0.09348948299884796,
+      "learning_rate": 2.0486870760219283e-05,
+      "loss": 9.3306,
+      "step": 63840
+    },
+    {
+      "epoch": 0.31885939723838297,
+      "grad_norm": 0.09794354438781738,
+      "learning_rate": 2.048536884527773e-05,
+      "loss": 9.3349,
+      "step": 63850
+    },
+    {
+      "epoch": 0.31890933606332245,
+      "grad_norm": 0.09906934201717377,
+      "learning_rate": 2.0483866930336177e-05,
+      "loss": 9.3347,
+      "step": 63860
+    },
+    {
+      "epoch": 0.31895927488826187,
+      "grad_norm": 0.09116369485855103,
+      "learning_rate": 2.048236501539463e-05,
+      "loss": 9.3346,
+      "step": 63870
+    },
+    {
+      "epoch": 0.31900921371320135,
+      "grad_norm": 0.0981554165482521,
+      "learning_rate": 2.0480863100453077e-05,
+      "loss": 9.3393,
+      "step": 63880
+    },
+    {
+      "epoch": 0.31905915253814077,
+      "grad_norm": 0.09284525364637375,
+      "learning_rate": 2.047936118551153e-05,
+      "loss": 9.3378,
+      "step": 63890
+    },
+    {
+      "epoch": 0.31910909136308024,
+      "grad_norm": 0.08612337708473206,
+      "learning_rate": 2.0477859270569978e-05,
+      "loss": 9.3374,
+      "step": 63900
+    },
+    {
+      "epoch": 0.31915903018801967,
+      "grad_norm": 0.0932999700307846,
+      "learning_rate": 2.0476357355628424e-05,
+      "loss": 9.3239,
+      "step": 63910
+    },
+    {
+      "epoch": 0.31920896901295914,
+      "grad_norm": 0.09308277070522308,
+      "learning_rate": 2.0474855440686878e-05,
+      "loss": 9.3398,
+      "step": 63920
+    },
+    {
+      "epoch": 0.31925890783789856,
+      "grad_norm": 0.09518188238143921,
+      "learning_rate": 2.0473353525745325e-05,
+      "loss": 9.343,
+      "step": 63930
+    },
+    {
+      "epoch": 0.31930884666283804,
+      "grad_norm": 0.09198715537786484,
+      "learning_rate": 2.047185161080378e-05,
+      "loss": 9.333,
+      "step": 63940
+    },
+    {
+      "epoch": 0.31935878548777746,
+      "grad_norm": 0.09532855451107025,
+      "learning_rate": 2.0470349695862225e-05,
+      "loss": 9.3383,
+      "step": 63950
+    },
+    {
+      "epoch": 0.31940872431271694,
+      "grad_norm": 0.09205342829227448,
+      "learning_rate": 2.0468847780920672e-05,
+      "loss": 9.3471,
+      "step": 63960
+    },
+    {
+      "epoch": 0.31945866313765636,
+      "grad_norm": 0.09050062298774719,
+      "learning_rate": 2.0467345865979126e-05,
+      "loss": 9.3426,
+      "step": 63970
+    },
+    {
+      "epoch": 0.31950860196259584,
+      "grad_norm": 0.0922810360789299,
+      "learning_rate": 2.0465843951037572e-05,
+      "loss": 9.3293,
+      "step": 63980
+    },
+    {
+      "epoch": 0.31955854078753526,
+      "grad_norm": 0.09575898200273514,
+      "learning_rate": 2.0464342036096026e-05,
+      "loss": 9.3398,
+      "step": 63990
+    },
+    {
+      "epoch": 0.31960847961247474,
+      "grad_norm": 0.09558529406785965,
+      "learning_rate": 2.0462840121154473e-05,
+      "loss": 9.3395,
+      "step": 64000
+    },
+    {
+      "epoch": 0.31965841843741416,
+      "grad_norm": 0.09229427576065063,
+      "learning_rate": 2.046133820621292e-05,
+      "loss": 9.3204,
+      "step": 64010
+    },
+    {
+      "epoch": 0.31970835726235364,
+      "grad_norm": 0.10348726063966751,
+      "learning_rate": 2.0459836291271373e-05,
+      "loss": 9.3323,
+      "step": 64020
+    },
+    {
+      "epoch": 0.31975829608729306,
+      "grad_norm": 0.09877219051122665,
+      "learning_rate": 2.045833437632982e-05,
+      "loss": 9.3436,
+      "step": 64030
+    },
+    {
+      "epoch": 0.31980823491223254,
+      "grad_norm": 0.09454214572906494,
+      "learning_rate": 2.0456832461388273e-05,
+      "loss": 9.3461,
+      "step": 64040
+    },
+    {
+      "epoch": 0.31985817373717196,
+      "grad_norm": 0.09850972890853882,
+      "learning_rate": 2.045533054644672e-05,
+      "loss": 9.3352,
+      "step": 64050
+    },
+    {
+      "epoch": 0.31990811256211144,
+      "grad_norm": 0.0938955768942833,
+      "learning_rate": 2.0453828631505167e-05,
+      "loss": 9.3256,
+      "step": 64060
+    },
+    {
+      "epoch": 0.31995805138705086,
+      "grad_norm": 0.08874368667602539,
+      "learning_rate": 2.045232671656362e-05,
+      "loss": 9.3407,
+      "step": 64070
+    },
+    {
+      "epoch": 0.32000799021199033,
+      "grad_norm": 0.09031280875205994,
+      "learning_rate": 2.0450824801622067e-05,
+      "loss": 9.3334,
+      "step": 64080
+    },
+    {
+      "epoch": 0.32005792903692976,
+      "grad_norm": 0.09989767521619797,
+      "learning_rate": 2.044932288668052e-05,
+      "loss": 9.3453,
+      "step": 64090
+    },
+    {
+      "epoch": 0.32010786786186923,
+      "grad_norm": 0.09756146371364594,
+      "learning_rate": 2.0447820971738968e-05,
+      "loss": 9.335,
+      "step": 64100
+    },
+    {
+      "epoch": 0.32015780668680865,
+      "grad_norm": 0.09458436816930771,
+      "learning_rate": 2.0446319056797414e-05,
+      "loss": 9.3411,
+      "step": 64110
+    },
+    {
+      "epoch": 0.32020774551174813,
+      "grad_norm": 0.0925472229719162,
+      "learning_rate": 2.0444817141855868e-05,
+      "loss": 9.3411,
+      "step": 64120
+    },
+    {
+      "epoch": 0.32025768433668755,
+      "grad_norm": 0.09772704541683197,
+      "learning_rate": 2.0443315226914315e-05,
+      "loss": 9.3346,
+      "step": 64130
+    },
+    {
+      "epoch": 0.32030762316162703,
+      "grad_norm": 0.09279755502939224,
+      "learning_rate": 2.044181331197277e-05,
+      "loss": 9.3347,
+      "step": 64140
+    },
+    {
+      "epoch": 0.32035756198656645,
+      "grad_norm": 0.09363307803869247,
+      "learning_rate": 2.0440311397031215e-05,
+      "loss": 9.3403,
+      "step": 64150
+    },
+    {
+      "epoch": 0.32040750081150593,
+      "grad_norm": 0.09853540360927582,
+      "learning_rate": 2.0438809482089665e-05,
+      "loss": 9.3262,
+      "step": 64160
+    },
+    {
+      "epoch": 0.32045743963644535,
+      "grad_norm": 0.09178748726844788,
+      "learning_rate": 2.0437307567148116e-05,
+      "loss": 9.3342,
+      "step": 64170
+    },
+    {
+      "epoch": 0.32050737846138483,
+      "grad_norm": 0.09160963445901871,
+      "learning_rate": 2.0435805652206562e-05,
+      "loss": 9.3396,
+      "step": 64180
+    },
+    {
+      "epoch": 0.32055731728632425,
+      "grad_norm": 0.09168528020381927,
+      "learning_rate": 2.0434303737265016e-05,
+      "loss": 9.3464,
+      "step": 64190
+    },
+    {
+      "epoch": 0.3206072561112637,
+      "grad_norm": 0.09851713478565216,
+      "learning_rate": 2.0432801822323463e-05,
+      "loss": 9.3393,
+      "step": 64200
+    },
+    {
+      "epoch": 0.32065719493620315,
+      "grad_norm": 0.09476348012685776,
+      "learning_rate": 2.0431299907381913e-05,
+      "loss": 9.3281,
+      "step": 64210
+    },
+    {
+      "epoch": 0.3207071337611426,
+      "grad_norm": 0.08930637687444687,
+      "learning_rate": 2.0429797992440363e-05,
+      "loss": 9.3371,
+      "step": 64220
+    },
+    {
+      "epoch": 0.32075707258608205,
+      "grad_norm": 0.09610996395349503,
+      "learning_rate": 2.042829607749881e-05,
+      "loss": 9.3344,
+      "step": 64230
+    },
+    {
+      "epoch": 0.3208070114110215,
+      "grad_norm": 0.09887643158435822,
+      "learning_rate": 2.0426794162557263e-05,
+      "loss": 9.3348,
+      "step": 64240
+    },
+    {
+      "epoch": 0.32085695023596095,
+      "grad_norm": 0.09950941801071167,
+      "learning_rate": 2.042529224761571e-05,
+      "loss": 9.3368,
+      "step": 64250
+    },
+    {
+      "epoch": 0.3209068890609004,
+      "grad_norm": 0.09124087542295456,
+      "learning_rate": 2.042379033267416e-05,
+      "loss": 9.3366,
+      "step": 64260
+    },
+    {
+      "epoch": 0.32095682788583985,
+      "grad_norm": 0.08983422070741653,
+      "learning_rate": 2.042228841773261e-05,
+      "loss": 9.3281,
+      "step": 64270
+    },
+    {
+      "epoch": 0.32100676671077927,
+      "grad_norm": 0.09503147006034851,
+      "learning_rate": 2.0420786502791057e-05,
+      "loss": 9.336,
+      "step": 64280
+    },
+    {
+      "epoch": 0.32105670553571874,
+      "grad_norm": 0.09027642011642456,
+      "learning_rate": 2.041928458784951e-05,
+      "loss": 9.3261,
+      "step": 64290
+    },
+    {
+      "epoch": 0.32110664436065817,
+      "grad_norm": 0.09734740853309631,
+      "learning_rate": 2.0417782672907958e-05,
+      "loss": 9.3386,
+      "step": 64300
+    },
+    {
+      "epoch": 0.32115658318559764,
+      "grad_norm": 0.09774748980998993,
+      "learning_rate": 2.0416280757966408e-05,
+      "loss": 9.335,
+      "step": 64310
+    },
+    {
+      "epoch": 0.32120652201053707,
+      "grad_norm": 0.09811116009950638,
+      "learning_rate": 2.0414778843024858e-05,
+      "loss": 9.3404,
+      "step": 64320
+    },
+    {
+      "epoch": 0.32125646083547654,
+      "grad_norm": 0.09968292713165283,
+      "learning_rate": 2.0413276928083305e-05,
+      "loss": 9.3334,
+      "step": 64330
+    },
+    {
+      "epoch": 0.32130639966041596,
+      "grad_norm": 0.09214014559984207,
+      "learning_rate": 2.041177501314176e-05,
+      "loss": 9.3353,
+      "step": 64340
+    },
+    {
+      "epoch": 0.32135633848535544,
+      "grad_norm": 0.09302917867898941,
+      "learning_rate": 2.0410273098200205e-05,
+      "loss": 9.3377,
+      "step": 64350
+    },
+    {
+      "epoch": 0.32140627731029486,
+      "grad_norm": 0.09487059712409973,
+      "learning_rate": 2.0408771183258655e-05,
+      "loss": 9.3373,
+      "step": 64360
+    },
+    {
+      "epoch": 0.32145621613523434,
+      "grad_norm": 0.09410389512777328,
+      "learning_rate": 2.0407269268317106e-05,
+      "loss": 9.3376,
+      "step": 64370
+    },
+    {
+      "epoch": 0.32150615496017376,
+      "grad_norm": 0.09394783526659012,
+      "learning_rate": 2.0405767353375552e-05,
+      "loss": 9.3304,
+      "step": 64380
+    },
+    {
+      "epoch": 0.32155609378511324,
+      "grad_norm": 0.10009703040122986,
+      "learning_rate": 2.0404265438434006e-05,
+      "loss": 9.3396,
+      "step": 64390
+    },
+    {
+      "epoch": 0.32160603261005266,
+      "grad_norm": 0.088577039539814,
+      "learning_rate": 2.0402763523492453e-05,
+      "loss": 9.3351,
+      "step": 64400
+    },
+    {
+      "epoch": 0.32165597143499214,
+      "grad_norm": 0.09477560222148895,
+      "learning_rate": 2.0401261608550903e-05,
+      "loss": 9.3235,
+      "step": 64410
+    },
+    {
+      "epoch": 0.32170591025993156,
+      "grad_norm": 0.09325342625379562,
+      "learning_rate": 2.0399759693609353e-05,
+      "loss": 9.344,
+      "step": 64420
+    },
+    {
+      "epoch": 0.32175584908487104,
+      "grad_norm": 0.09685123711824417,
+      "learning_rate": 2.03982577786678e-05,
+      "loss": 9.3409,
+      "step": 64430
+    },
+    {
+      "epoch": 0.32180578790981046,
+      "grad_norm": 0.09679123014211655,
+      "learning_rate": 2.0396755863726253e-05,
+      "loss": 9.3162,
+      "step": 64440
+    },
+    {
+      "epoch": 0.32185572673474994,
+      "grad_norm": 0.09780653566122055,
+      "learning_rate": 2.03952539487847e-05,
+      "loss": 9.3269,
+      "step": 64450
+    },
+    {
+      "epoch": 0.32190566555968936,
+      "grad_norm": 0.09367544203996658,
+      "learning_rate": 2.039375203384315e-05,
+      "loss": 9.3264,
+      "step": 64460
+    },
+    {
+      "epoch": 0.32195560438462884,
+      "grad_norm": 0.0960097685456276,
+      "learning_rate": 2.03922501189016e-05,
+      "loss": 9.334,
+      "step": 64470
+    },
+    {
+      "epoch": 0.32200554320956826,
+      "grad_norm": 0.09054725617170334,
+      "learning_rate": 2.039074820396005e-05,
+      "loss": 9.3354,
+      "step": 64480
+    },
+    {
+      "epoch": 0.32205548203450773,
+      "grad_norm": 0.09367144107818604,
+      "learning_rate": 2.03892462890185e-05,
+      "loss": 9.3318,
+      "step": 64490
+    },
+    {
+      "epoch": 0.32210542085944716,
+      "grad_norm": 0.09095479547977448,
+      "learning_rate": 2.0387744374076948e-05,
+      "loss": 9.3297,
+      "step": 64500
+    },
+    {
+      "epoch": 0.32215535968438663,
+      "grad_norm": 0.09100858122110367,
+      "learning_rate": 2.0386242459135398e-05,
+      "loss": 9.3333,
+      "step": 64510
+    },
+    {
+      "epoch": 0.32220529850932605,
+      "grad_norm": 0.09426578879356384,
+      "learning_rate": 2.0384740544193848e-05,
+      "loss": 9.3289,
+      "step": 64520
+    },
+    {
+      "epoch": 0.32225523733426553,
+      "grad_norm": 0.08965863287448883,
+      "learning_rate": 2.0383238629252298e-05,
+      "loss": 9.3264,
+      "step": 64530
+    },
+    {
+      "epoch": 0.32230517615920495,
+      "grad_norm": 0.092529296875,
+      "learning_rate": 2.038173671431075e-05,
+      "loss": 9.337,
+      "step": 64540
+    },
+    {
+      "epoch": 0.32235511498414443,
+      "grad_norm": 0.09353473037481308,
+      "learning_rate": 2.0380234799369195e-05,
+      "loss": 9.3338,
+      "step": 64550
+    },
+    {
+      "epoch": 0.32240505380908385,
+      "grad_norm": 0.09257248789072037,
+      "learning_rate": 2.0378732884427645e-05,
+      "loss": 9.3232,
+      "step": 64560
+    },
+    {
+      "epoch": 0.32245499263402333,
+      "grad_norm": 0.09056426584720612,
+      "learning_rate": 2.0377230969486096e-05,
+      "loss": 9.3363,
+      "step": 64570
+    },
+    {
+      "epoch": 0.32250493145896275,
+      "grad_norm": 0.09313210844993591,
+      "learning_rate": 2.0375729054544546e-05,
+      "loss": 9.3281,
+      "step": 64580
+    },
+    {
+      "epoch": 0.32255487028390223,
+      "grad_norm": 0.09300418943166733,
+      "learning_rate": 2.0374227139602996e-05,
+      "loss": 9.3343,
+      "step": 64590
+    },
+    {
+      "epoch": 0.32260480910884165,
+      "grad_norm": 0.09289450198411942,
+      "learning_rate": 2.0372725224661443e-05,
+      "loss": 9.3244,
+      "step": 64600
+    },
+    {
+      "epoch": 0.3226547479337811,
+      "grad_norm": 0.09735582768917084,
+      "learning_rate": 2.0371223309719893e-05,
+      "loss": 9.3233,
+      "step": 64610
+    },
+    {
+      "epoch": 0.32270468675872055,
+      "grad_norm": 0.09457716345787048,
+      "learning_rate": 2.0369721394778343e-05,
+      "loss": 9.3306,
+      "step": 64620
+    },
+    {
+      "epoch": 0.32275462558366,
+      "grad_norm": 0.10149849951267242,
+      "learning_rate": 2.0368219479836793e-05,
+      "loss": 9.339,
+      "step": 64630
+    },
+    {
+      "epoch": 0.32280456440859945,
+      "grad_norm": 0.094261035323143,
+      "learning_rate": 2.0366717564895243e-05,
+      "loss": 9.3271,
+      "step": 64640
+    },
+    {
+      "epoch": 0.3228545032335389,
+      "grad_norm": 0.09714456647634506,
+      "learning_rate": 2.036521564995369e-05,
+      "loss": 9.3295,
+      "step": 64650
+    },
+    {
+      "epoch": 0.32290444205847835,
+      "grad_norm": 0.09245141595602036,
+      "learning_rate": 2.036371373501214e-05,
+      "loss": 9.3433,
+      "step": 64660
+    },
+    {
+      "epoch": 0.3229543808834178,
+      "grad_norm": 0.09247272461652756,
+      "learning_rate": 2.036221182007059e-05,
+      "loss": 9.3184,
+      "step": 64670
+    },
+    {
+      "epoch": 0.32300431970835725,
+      "grad_norm": 0.08987551927566528,
+      "learning_rate": 2.036070990512904e-05,
+      "loss": 9.3249,
+      "step": 64680
+    },
+    {
+      "epoch": 0.3230542585332967,
+      "grad_norm": 0.09094039350748062,
+      "learning_rate": 2.035920799018749e-05,
+      "loss": 9.341,
+      "step": 64690
+    },
+    {
+      "epoch": 0.32310419735823614,
+      "grad_norm": 0.09806418418884277,
+      "learning_rate": 2.0357706075245938e-05,
+      "loss": 9.3295,
+      "step": 64700
+    },
+    {
+      "epoch": 0.3231541361831756,
+      "grad_norm": 0.09336648136377335,
+      "learning_rate": 2.0356204160304388e-05,
+      "loss": 9.3318,
+      "step": 64710
+    },
+    {
+      "epoch": 0.32320407500811504,
+      "grad_norm": 0.09484011679887772,
+      "learning_rate": 2.0354702245362838e-05,
+      "loss": 9.3308,
+      "step": 64720
+    },
+    {
+      "epoch": 0.3232540138330545,
+      "grad_norm": 0.09836547821760178,
+      "learning_rate": 2.0353200330421288e-05,
+      "loss": 9.3284,
+      "step": 64730
+    },
+    {
+      "epoch": 0.32330395265799394,
+      "grad_norm": 0.09364361315965652,
+      "learning_rate": 2.035169841547974e-05,
+      "loss": 9.322,
+      "step": 64740
+    },
+    {
+      "epoch": 0.3233538914829334,
+      "grad_norm": 0.09323278069496155,
+      "learning_rate": 2.0350196500538185e-05,
+      "loss": 9.3337,
+      "step": 64750
+    },
+    {
+      "epoch": 0.32340383030787284,
+      "grad_norm": 0.09357263892889023,
+      "learning_rate": 2.0348694585596635e-05,
+      "loss": 9.3295,
+      "step": 64760
+    },
+    {
+      "epoch": 0.3234537691328123,
+      "grad_norm": 0.09743960201740265,
+      "learning_rate": 2.0347192670655086e-05,
+      "loss": 9.3393,
+      "step": 64770
+    },
+    {
+      "epoch": 0.32350370795775174,
+      "grad_norm": 0.09175863862037659,
+      "learning_rate": 2.0345690755713536e-05,
+      "loss": 9.3253,
+      "step": 64780
+    },
+    {
+      "epoch": 0.3235536467826912,
+      "grad_norm": 0.08689554035663605,
+      "learning_rate": 2.0344188840771986e-05,
+      "loss": 9.3292,
+      "step": 64790
+    },
+    {
+      "epoch": 0.32360358560763064,
+      "grad_norm": 0.09737929701805115,
+      "learning_rate": 2.0342686925830433e-05,
+      "loss": 9.3275,
+      "step": 64800
+    },
+    {
+      "epoch": 0.3236535244325701,
+      "grad_norm": 0.09190335869789124,
+      "learning_rate": 2.0341185010888883e-05,
+      "loss": 9.3309,
+      "step": 64810
+    },
+    {
+      "epoch": 0.32370346325750954,
+      "grad_norm": 0.09484836459159851,
+      "learning_rate": 2.0339683095947333e-05,
+      "loss": 9.3211,
+      "step": 64820
+    },
+    {
+      "epoch": 0.323753402082449,
+      "grad_norm": 0.09082531183958054,
+      "learning_rate": 2.0338181181005783e-05,
+      "loss": 9.324,
+      "step": 64830
+    },
+    {
+      "epoch": 0.32380334090738844,
+      "grad_norm": 0.09505081921815872,
+      "learning_rate": 2.0336679266064233e-05,
+      "loss": 9.3322,
+      "step": 64840
+    },
+    {
+      "epoch": 0.3238532797323279,
+      "grad_norm": 0.09214816242456436,
+      "learning_rate": 2.0335177351122684e-05,
+      "loss": 9.3252,
+      "step": 64850
+    },
+    {
+      "epoch": 0.32390321855726734,
+      "grad_norm": 0.09512375295162201,
+      "learning_rate": 2.033367543618113e-05,
+      "loss": 9.3216,
+      "step": 64860
+    },
+    {
+      "epoch": 0.3239531573822068,
+      "grad_norm": 0.09365373104810715,
+      "learning_rate": 2.033217352123958e-05,
+      "loss": 9.3385,
+      "step": 64870
+    },
+    {
+      "epoch": 0.32400309620714624,
+      "grad_norm": 0.09065309166908264,
+      "learning_rate": 2.033067160629803e-05,
+      "loss": 9.3191,
+      "step": 64880
+    },
+    {
+      "epoch": 0.3240530350320857,
+      "grad_norm": 0.09178464859724045,
+      "learning_rate": 2.032916969135648e-05,
+      "loss": 9.3268,
+      "step": 64890
+    },
+    {
+      "epoch": 0.32410297385702513,
+      "grad_norm": 0.09575120359659195,
+      "learning_rate": 2.032766777641493e-05,
+      "loss": 9.3209,
+      "step": 64900
+    },
+    {
+      "epoch": 0.3241529126819646,
+      "grad_norm": 0.09635672718286514,
+      "learning_rate": 2.0326165861473378e-05,
+      "loss": 9.3288,
+      "step": 64910
+    },
+    {
+      "epoch": 0.32420285150690403,
+      "grad_norm": 0.09563815593719482,
+      "learning_rate": 2.0324663946531828e-05,
+      "loss": 9.3158,
+      "step": 64920
+    },
+    {
+      "epoch": 0.3242527903318435,
+      "grad_norm": 0.09533455222845078,
+      "learning_rate": 2.0323162031590278e-05,
+      "loss": 9.325,
+      "step": 64930
+    },
+    {
+      "epoch": 0.32430272915678293,
+      "grad_norm": 0.09778301417827606,
+      "learning_rate": 2.032166011664873e-05,
+      "loss": 9.3241,
+      "step": 64940
+    },
+    {
+      "epoch": 0.3243526679817224,
+      "grad_norm": 0.09181251376867294,
+      "learning_rate": 2.032015820170718e-05,
+      "loss": 9.3237,
+      "step": 64950
+    },
+    {
+      "epoch": 0.32440260680666183,
+      "grad_norm": 0.09311333298683167,
+      "learning_rate": 2.0318656286765625e-05,
+      "loss": 9.3285,
+      "step": 64960
+    },
+    {
+      "epoch": 0.3244525456316013,
+      "grad_norm": 0.0935024842619896,
+      "learning_rate": 2.0317154371824076e-05,
+      "loss": 9.324,
+      "step": 64970
+    },
+    {
+      "epoch": 0.32450248445654073,
+      "grad_norm": 0.10296680778265,
+      "learning_rate": 2.0315652456882526e-05,
+      "loss": 9.3188,
+      "step": 64980
+    },
+    {
+      "epoch": 0.3245524232814802,
+      "grad_norm": 0.0903451070189476,
+      "learning_rate": 2.0314150541940976e-05,
+      "loss": 9.3293,
+      "step": 64990
+    },
+    {
+      "epoch": 0.32460236210641963,
+      "grad_norm": 0.09111391007900238,
+      "learning_rate": 2.0312648626999426e-05,
+      "loss": 9.3187,
+      "step": 65000
+    },
+    {
+      "epoch": 0.3246523009313591,
+      "grad_norm": 0.09491260349750519,
+      "learning_rate": 2.0311146712057873e-05,
+      "loss": 9.3262,
+      "step": 65010
+    },
+    {
+      "epoch": 0.3247022397562985,
+      "grad_norm": 0.09257805347442627,
+      "learning_rate": 2.0309644797116323e-05,
+      "loss": 9.3281,
+      "step": 65020
+    },
+    {
+      "epoch": 0.324752178581238,
+      "grad_norm": 0.09723227471113205,
+      "learning_rate": 2.0308142882174773e-05,
+      "loss": 9.3198,
+      "step": 65030
+    },
+    {
+      "epoch": 0.3248021174061774,
+      "grad_norm": 0.09565103054046631,
+      "learning_rate": 2.0306640967233223e-05,
+      "loss": 9.3241,
+      "step": 65040
+    },
+    {
+      "epoch": 0.3248520562311169,
+      "grad_norm": 0.09280645102262497,
+      "learning_rate": 2.0305139052291674e-05,
+      "loss": 9.3315,
+      "step": 65050
+    },
+    {
+      "epoch": 0.3249019950560563,
+      "grad_norm": 0.09414808452129364,
+      "learning_rate": 2.030363713735012e-05,
+      "loss": 9.318,
+      "step": 65060
+    },
+    {
+      "epoch": 0.3249519338809958,
+      "grad_norm": 0.09369461238384247,
+      "learning_rate": 2.030213522240857e-05,
+      "loss": 9.3343,
+      "step": 65070
+    },
+    {
+      "epoch": 0.3250018727059352,
+      "grad_norm": 0.09321442246437073,
+      "learning_rate": 2.030063330746702e-05,
+      "loss": 9.3193,
+      "step": 65080
+    },
+    {
+      "epoch": 0.3250518115308747,
+      "grad_norm": 0.0960422232747078,
+      "learning_rate": 2.029913139252547e-05,
+      "loss": 9.3226,
+      "step": 65090
+    },
+    {
+      "epoch": 0.3251017503558141,
+      "grad_norm": 0.0959312692284584,
+      "learning_rate": 2.029762947758392e-05,
+      "loss": 9.3237,
+      "step": 65100
+    },
+    {
+      "epoch": 0.3251516891807536,
+      "grad_norm": 0.09452484548091888,
+      "learning_rate": 2.0296127562642368e-05,
+      "loss": 9.3171,
+      "step": 65110
+    },
+    {
+      "epoch": 0.325201628005693,
+      "grad_norm": 0.09644035995006561,
+      "learning_rate": 2.0294625647700818e-05,
+      "loss": 9.3132,
+      "step": 65120
+    },
+    {
+      "epoch": 0.3252515668306325,
+      "grad_norm": 0.09128149598836899,
+      "learning_rate": 2.029312373275927e-05,
+      "loss": 9.3258,
+      "step": 65130
+    },
+    {
+      "epoch": 0.3253015056555719,
+      "grad_norm": 0.08971234411001205,
+      "learning_rate": 2.029162181781772e-05,
+      "loss": 9.3171,
+      "step": 65140
+    },
+    {
+      "epoch": 0.3253514444805114,
+      "grad_norm": 0.09548742324113846,
+      "learning_rate": 2.029011990287617e-05,
+      "loss": 9.3202,
+      "step": 65150
+    },
+    {
+      "epoch": 0.3254013833054508,
+      "grad_norm": 0.0900539830327034,
+      "learning_rate": 2.0288617987934615e-05,
+      "loss": 9.3234,
+      "step": 65160
+    },
+    {
+      "epoch": 0.3254513221303903,
+      "grad_norm": 0.08711142092943192,
+      "learning_rate": 2.028711607299307e-05,
+      "loss": 9.3191,
+      "step": 65170
+    },
+    {
+      "epoch": 0.3255012609553297,
+      "grad_norm": 0.09119576215744019,
+      "learning_rate": 2.0285614158051516e-05,
+      "loss": 9.3174,
+      "step": 65180
+    },
+    {
+      "epoch": 0.3255511997802692,
+      "grad_norm": 0.09764738380908966,
+      "learning_rate": 2.0284112243109966e-05,
+      "loss": 9.3225,
+      "step": 65190
+    },
+    {
+      "epoch": 0.3256011386052086,
+      "grad_norm": 0.09541153162717819,
+      "learning_rate": 2.0282610328168416e-05,
+      "loss": 9.3139,
+      "step": 65200
+    },
+    {
+      "epoch": 0.3256510774301481,
+      "grad_norm": 0.09369537979364395,
+      "learning_rate": 2.0281108413226863e-05,
+      "loss": 9.325,
+      "step": 65210
+    },
+    {
+      "epoch": 0.3257010162550875,
+      "grad_norm": 0.09462661296129227,
+      "learning_rate": 2.0279606498285317e-05,
+      "loss": 9.3226,
+      "step": 65220
+    },
+    {
+      "epoch": 0.325750955080027,
+      "grad_norm": 0.09155022352933884,
+      "learning_rate": 2.0278104583343763e-05,
+      "loss": 9.3225,
+      "step": 65230
+    },
+    {
+      "epoch": 0.3258008939049664,
+      "grad_norm": 0.09792565554380417,
+      "learning_rate": 2.0276602668402214e-05,
+      "loss": 9.3181,
+      "step": 65240
+    },
+    {
+      "epoch": 0.3258508327299059,
+      "grad_norm": 0.09263359010219574,
+      "learning_rate": 2.0275100753460664e-05,
+      "loss": 9.3194,
+      "step": 65250
+    },
+    {
+      "epoch": 0.3259007715548453,
+      "grad_norm": 0.09698522835969925,
+      "learning_rate": 2.027359883851911e-05,
+      "loss": 9.3212,
+      "step": 65260
+    },
+    {
+      "epoch": 0.32595071037978474,
+      "grad_norm": 0.09113046526908875,
+      "learning_rate": 2.0272096923577564e-05,
+      "loss": 9.3182,
+      "step": 65270
+    },
+    {
+      "epoch": 0.3260006492047242,
+      "grad_norm": 0.10038909316062927,
+      "learning_rate": 2.027059500863601e-05,
+      "loss": 9.3206,
+      "step": 65280
+    },
+    {
+      "epoch": 0.32605058802966363,
+      "grad_norm": 0.09228735417127609,
+      "learning_rate": 2.026909309369446e-05,
+      "loss": 9.3112,
+      "step": 65290
+    },
+    {
+      "epoch": 0.3261005268546031,
+      "grad_norm": 0.09285099059343338,
+      "learning_rate": 2.026759117875291e-05,
+      "loss": 9.319,
+      "step": 65300
+    },
+    {
+      "epoch": 0.32615046567954253,
+      "grad_norm": 0.09246840327978134,
+      "learning_rate": 2.0266089263811358e-05,
+      "loss": 9.3158,
+      "step": 65310
+    },
+    {
+      "epoch": 0.326200404504482,
+      "grad_norm": 0.09260924160480499,
+      "learning_rate": 2.026458734886981e-05,
+      "loss": 9.3126,
+      "step": 65320
+    },
+    {
+      "epoch": 0.32625034332942143,
+      "grad_norm": 0.09276574850082397,
+      "learning_rate": 2.026308543392826e-05,
+      "loss": 9.3142,
+      "step": 65330
+    },
+    {
+      "epoch": 0.3263002821543609,
+      "grad_norm": 0.09407460689544678,
+      "learning_rate": 2.026158351898671e-05,
+      "loss": 9.3089,
+      "step": 65340
+    },
+    {
+      "epoch": 0.32635022097930033,
+      "grad_norm": 0.09039918333292007,
+      "learning_rate": 2.026008160404516e-05,
+      "loss": 9.3121,
+      "step": 65350
+    },
+    {
+      "epoch": 0.3264001598042398,
+      "grad_norm": 0.09433960914611816,
+      "learning_rate": 2.0258579689103605e-05,
+      "loss": 9.3153,
+      "step": 65360
+    },
+    {
+      "epoch": 0.32645009862917923,
+      "grad_norm": 0.0907716453075409,
+      "learning_rate": 2.025707777416206e-05,
+      "loss": 9.313,
+      "step": 65370
+    },
+    {
+      "epoch": 0.3265000374541187,
+      "grad_norm": 0.097490094602108,
+      "learning_rate": 2.0255575859220506e-05,
+      "loss": 9.3096,
+      "step": 65380
+    },
+    {
+      "epoch": 0.32654997627905813,
+      "grad_norm": 0.0970582440495491,
+      "learning_rate": 2.0254073944278956e-05,
+      "loss": 9.3117,
+      "step": 65390
+    },
+    {
+      "epoch": 0.3265999151039976,
+      "grad_norm": 0.09796217083930969,
+      "learning_rate": 2.0252572029337406e-05,
+      "loss": 9.3095,
+      "step": 65400
+    },
+    {
+      "epoch": 0.32664985392893703,
+      "grad_norm": 0.09917185455560684,
+      "learning_rate": 2.0251070114395853e-05,
+      "loss": 9.315,
+      "step": 65410
+    },
+    {
+      "epoch": 0.3266997927538765,
+      "grad_norm": 0.0976761057972908,
+      "learning_rate": 2.0249568199454307e-05,
+      "loss": 9.3177,
+      "step": 65420
+    },
+    {
+      "epoch": 0.3267497315788159,
+      "grad_norm": 0.09032870829105377,
+      "learning_rate": 2.0248066284512753e-05,
+      "loss": 9.3253,
+      "step": 65430
+    },
+    {
+      "epoch": 0.3267996704037554,
+      "grad_norm": 0.0941029042005539,
+      "learning_rate": 2.0246564369571204e-05,
+      "loss": 9.3135,
+      "step": 65440
+    },
+    {
+      "epoch": 0.3268496092286948,
+      "grad_norm": 0.09661585092544556,
+      "learning_rate": 2.0245062454629654e-05,
+      "loss": 9.3097,
+      "step": 65450
+    },
+    {
+      "epoch": 0.3268995480536343,
+      "grad_norm": 0.09369435906410217,
+      "learning_rate": 2.02435605396881e-05,
+      "loss": 9.3211,
+      "step": 65460
+    },
+    {
+      "epoch": 0.3269494868785737,
+      "grad_norm": 0.0993766039609909,
+      "learning_rate": 2.0242058624746554e-05,
+      "loss": 9.3172,
+      "step": 65470
+    },
+    {
+      "epoch": 0.3269994257035132,
+      "grad_norm": 0.0999893844127655,
+      "learning_rate": 2.0240556709805e-05,
+      "loss": 9.3048,
+      "step": 65480
+    },
+    {
+      "epoch": 0.3270493645284526,
+      "grad_norm": 0.09266514331102371,
+      "learning_rate": 2.0239054794863454e-05,
+      "loss": 9.3177,
+      "step": 65490
+    },
+    {
+      "epoch": 0.3270993033533921,
+      "grad_norm": 0.09590865671634674,
+      "learning_rate": 2.02375528799219e-05,
+      "loss": 9.3148,
+      "step": 65500
+    },
+    {
+      "epoch": 0.3271492421783315,
+      "grad_norm": 0.10051383823156357,
+      "learning_rate": 2.0236050964980348e-05,
+      "loss": 9.3207,
+      "step": 65510
+    },
+    {
+      "epoch": 0.327199181003271,
+      "grad_norm": 0.09786202013492584,
+      "learning_rate": 2.02345490500388e-05,
+      "loss": 9.3105,
+      "step": 65520
+    },
+    {
+      "epoch": 0.3272491198282104,
+      "grad_norm": 0.09516331553459167,
+      "learning_rate": 2.023304713509725e-05,
+      "loss": 9.3199,
+      "step": 65530
+    },
+    {
+      "epoch": 0.3272990586531499,
+      "grad_norm": 0.09292834252119064,
+      "learning_rate": 2.0231545220155702e-05,
+      "loss": 9.3057,
+      "step": 65540
+    },
+    {
+      "epoch": 0.3273489974780893,
+      "grad_norm": 0.0924135223031044,
+      "learning_rate": 2.023004330521415e-05,
+      "loss": 9.3168,
+      "step": 65550
+    },
+    {
+      "epoch": 0.3273989363030288,
+      "grad_norm": 0.08832070231437683,
+      "learning_rate": 2.0228541390272595e-05,
+      "loss": 9.3189,
+      "step": 65560
+    },
+    {
+      "epoch": 0.3274488751279682,
+      "grad_norm": 0.09260696917772293,
+      "learning_rate": 2.022703947533105e-05,
+      "loss": 9.3099,
+      "step": 65570
+    },
+    {
+      "epoch": 0.3274988139529077,
+      "grad_norm": 0.09234783053398132,
+      "learning_rate": 2.0225537560389496e-05,
+      "loss": 9.3106,
+      "step": 65580
+    },
+    {
+      "epoch": 0.3275487527778471,
+      "grad_norm": 0.09439610689878464,
+      "learning_rate": 2.022403564544795e-05,
+      "loss": 9.309,
+      "step": 65590
+    },
+    {
+      "epoch": 0.3275986916027866,
+      "grad_norm": 0.09166769683361053,
+      "learning_rate": 2.0222533730506396e-05,
+      "loss": 9.3154,
+      "step": 65600
+    },
+    {
+      "epoch": 0.327648630427726,
+      "grad_norm": 0.09410682320594788,
+      "learning_rate": 2.0221031815564843e-05,
+      "loss": 9.3146,
+      "step": 65610
+    },
+    {
+      "epoch": 0.3276985692526655,
+      "grad_norm": 0.09807700663805008,
+      "learning_rate": 2.0219529900623297e-05,
+      "loss": 9.3173,
+      "step": 65620
+    },
+    {
+      "epoch": 0.3277485080776049,
+      "grad_norm": 0.10019030421972275,
+      "learning_rate": 2.0218027985681743e-05,
+      "loss": 9.3117,
+      "step": 65630
+    },
+    {
+      "epoch": 0.3277984469025444,
+      "grad_norm": 0.09734559059143066,
+      "learning_rate": 2.0216526070740197e-05,
+      "loss": 9.3175,
+      "step": 65640
+    },
+    {
+      "epoch": 0.3278483857274838,
+      "grad_norm": 0.09119731932878494,
+      "learning_rate": 2.0215024155798644e-05,
+      "loss": 9.3123,
+      "step": 65650
+    },
+    {
+      "epoch": 0.3278983245524233,
+      "grad_norm": 0.09404603391885757,
+      "learning_rate": 2.021352224085709e-05,
+      "loss": 9.3134,
+      "step": 65660
+    },
+    {
+      "epoch": 0.3279482633773627,
+      "grad_norm": 0.09394799917936325,
+      "learning_rate": 2.0212020325915544e-05,
+      "loss": 9.3197,
+      "step": 65670
+    },
+    {
+      "epoch": 0.3279982022023022,
+      "grad_norm": 0.0975813940167427,
+      "learning_rate": 2.021051841097399e-05,
+      "loss": 9.3185,
+      "step": 65680
+    },
+    {
+      "epoch": 0.3280481410272416,
+      "grad_norm": 0.09289105236530304,
+      "learning_rate": 2.0209016496032444e-05,
+      "loss": 9.317,
+      "step": 65690
+    },
+    {
+      "epoch": 0.3280980798521811,
+      "grad_norm": 0.08842571079730988,
+      "learning_rate": 2.020751458109089e-05,
+      "loss": 9.3099,
+      "step": 65700
+    },
+    {
+      "epoch": 0.3281480186771205,
+      "grad_norm": 0.09984628111124039,
+      "learning_rate": 2.0206012666149338e-05,
+      "loss": 9.3071,
+      "step": 65710
+    },
+    {
+      "epoch": 0.32819795750206,
+      "grad_norm": 0.09205308556556702,
+      "learning_rate": 2.020451075120779e-05,
+      "loss": 9.3137,
+      "step": 65720
+    },
+    {
+      "epoch": 0.3282478963269994,
+      "grad_norm": 0.09461898356676102,
+      "learning_rate": 2.020300883626624e-05,
+      "loss": 9.3055,
+      "step": 65730
+    },
+    {
+      "epoch": 0.3282978351519389,
+      "grad_norm": 0.09863221645355225,
+      "learning_rate": 2.0201506921324692e-05,
+      "loss": 9.3214,
+      "step": 65740
+    },
+    {
+      "epoch": 0.3283477739768783,
+      "grad_norm": 0.09489452093839645,
+      "learning_rate": 2.020000500638314e-05,
+      "loss": 9.3222,
+      "step": 65750
+    },
+    {
+      "epoch": 0.3283977128018178,
+      "grad_norm": 0.09285447001457214,
+      "learning_rate": 2.0198503091441586e-05,
+      "loss": 9.3073,
+      "step": 65760
+    },
+    {
+      "epoch": 0.3284476516267572,
+      "grad_norm": 0.09722167253494263,
+      "learning_rate": 2.019700117650004e-05,
+      "loss": 9.3122,
+      "step": 65770
+    },
+    {
+      "epoch": 0.3284975904516967,
+      "grad_norm": 0.09583593159914017,
+      "learning_rate": 2.0195499261558486e-05,
+      "loss": 9.3235,
+      "step": 65780
+    },
+    {
+      "epoch": 0.3285475292766361,
+      "grad_norm": 0.09201577305793762,
+      "learning_rate": 2.019399734661694e-05,
+      "loss": 9.3062,
+      "step": 65790
+    },
+    {
+      "epoch": 0.3285974681015756,
+      "grad_norm": 0.09347929805517197,
+      "learning_rate": 2.0192495431675386e-05,
+      "loss": 9.3122,
+      "step": 65800
+    },
+    {
+      "epoch": 0.328647406926515,
+      "grad_norm": 0.0961994007229805,
+      "learning_rate": 2.0190993516733836e-05,
+      "loss": 9.3093,
+      "step": 65810
+    },
+    {
+      "epoch": 0.3286973457514545,
+      "grad_norm": 0.0859488919377327,
+      "learning_rate": 2.0189491601792287e-05,
+      "loss": 9.3174,
+      "step": 65820
+    },
+    {
+      "epoch": 0.3287472845763939,
+      "grad_norm": 0.09572838991880417,
+      "learning_rate": 2.0187989686850733e-05,
+      "loss": 9.3193,
+      "step": 65830
+    },
+    {
+      "epoch": 0.3287972234013334,
+      "grad_norm": 0.09346995502710342,
+      "learning_rate": 2.0186487771909187e-05,
+      "loss": 9.3144,
+      "step": 65840
+    },
+    {
+      "epoch": 0.3288471622262728,
+      "grad_norm": 0.10546184331178665,
+      "learning_rate": 2.0184985856967634e-05,
+      "loss": 9.3125,
+      "step": 65850
+    },
+    {
+      "epoch": 0.3288971010512123,
+      "grad_norm": 0.08991089463233948,
+      "learning_rate": 2.0183483942026084e-05,
+      "loss": 9.3123,
+      "step": 65860
+    },
+    {
+      "epoch": 0.3289470398761517,
+      "grad_norm": 0.09135644137859344,
+      "learning_rate": 2.0181982027084534e-05,
+      "loss": 9.3112,
+      "step": 65870
+    },
+    {
+      "epoch": 0.3289969787010912,
+      "grad_norm": 0.0967700332403183,
+      "learning_rate": 2.018048011214298e-05,
+      "loss": 9.3096,
+      "step": 65880
+    },
+    {
+      "epoch": 0.3290469175260306,
+      "grad_norm": 0.09798181802034378,
+      "learning_rate": 2.0178978197201434e-05,
+      "loss": 9.3139,
+      "step": 65890
+    },
+    {
+      "epoch": 0.3290968563509701,
+      "grad_norm": 0.09936074167490005,
+      "learning_rate": 2.017747628225988e-05,
+      "loss": 9.3105,
+      "step": 65900
+    },
+    {
+      "epoch": 0.3291467951759095,
+      "grad_norm": 0.094666488468647,
+      "learning_rate": 2.017597436731833e-05,
+      "loss": 9.3142,
+      "step": 65910
+    },
+    {
+      "epoch": 0.329196734000849,
+      "grad_norm": 0.09142427891492844,
+      "learning_rate": 2.017447245237678e-05,
+      "loss": 9.3042,
+      "step": 65920
+    },
+    {
+      "epoch": 0.3292466728257884,
+      "grad_norm": 0.09408102929592133,
+      "learning_rate": 2.017297053743523e-05,
+      "loss": 9.3053,
+      "step": 65930
+    },
+    {
+      "epoch": 0.3292966116507279,
+      "grad_norm": 0.09405004233121872,
+      "learning_rate": 2.0171468622493682e-05,
+      "loss": 9.3083,
+      "step": 65940
+    },
+    {
+      "epoch": 0.3293465504756673,
+      "grad_norm": 0.09288471937179565,
+      "learning_rate": 2.016996670755213e-05,
+      "loss": 9.3006,
+      "step": 65950
+    },
+    {
+      "epoch": 0.3293964893006068,
+      "grad_norm": 0.09375298768281937,
+      "learning_rate": 2.016846479261058e-05,
+      "loss": 9.3127,
+      "step": 65960
+    },
+    {
+      "epoch": 0.3294464281255462,
+      "grad_norm": 0.0929841622710228,
+      "learning_rate": 2.016696287766903e-05,
+      "loss": 9.3043,
+      "step": 65970
+    },
+    {
+      "epoch": 0.3294963669504857,
+      "grad_norm": 0.09631754457950592,
+      "learning_rate": 2.0165460962727476e-05,
+      "loss": 9.3143,
+      "step": 65980
+    },
+    {
+      "epoch": 0.3295463057754251,
+      "grad_norm": 0.09023533016443253,
+      "learning_rate": 2.016395904778593e-05,
+      "loss": 9.3028,
+      "step": 65990
+    },
+    {
+      "epoch": 0.3295962446003646,
+      "grad_norm": 0.09742230176925659,
+      "learning_rate": 2.0162457132844376e-05,
+      "loss": 9.308,
+      "step": 66000
+    },
+    {
+      "epoch": 0.329646183425304,
+      "grad_norm": 0.09984151273965836,
+      "learning_rate": 2.0160955217902826e-05,
+      "loss": 9.309,
+      "step": 66010
+    },
+    {
+      "epoch": 0.3296961222502435,
+      "grad_norm": 0.09294534474611282,
+      "learning_rate": 2.0159453302961277e-05,
+      "loss": 9.3217,
+      "step": 66020
+    },
+    {
+      "epoch": 0.3297460610751829,
+      "grad_norm": 0.09389375895261765,
+      "learning_rate": 2.0157951388019723e-05,
+      "loss": 9.3111,
+      "step": 66030
+    },
+    {
+      "epoch": 0.32979599990012237,
+      "grad_norm": 0.09610763192176819,
+      "learning_rate": 2.0156449473078177e-05,
+      "loss": 9.2965,
+      "step": 66040
+    },
+    {
+      "epoch": 0.3298459387250618,
+      "grad_norm": 0.1011497899889946,
+      "learning_rate": 2.0154947558136624e-05,
+      "loss": 9.3115,
+      "step": 66050
+    },
+    {
+      "epoch": 0.32989587755000127,
+      "grad_norm": 0.09551998227834702,
+      "learning_rate": 2.0153445643195074e-05,
+      "loss": 9.312,
+      "step": 66060
+    },
+    {
+      "epoch": 0.3299458163749407,
+      "grad_norm": 0.09386122971773148,
+      "learning_rate": 2.0151943728253524e-05,
+      "loss": 9.3002,
+      "step": 66070
+    },
+    {
+      "epoch": 0.32999575519988017,
+      "grad_norm": 0.08605486899614334,
+      "learning_rate": 2.015044181331197e-05,
+      "loss": 9.3053,
+      "step": 66080
+    },
+    {
+      "epoch": 0.3300456940248196,
+      "grad_norm": 0.09588326513767242,
+      "learning_rate": 2.0148939898370424e-05,
+      "loss": 9.3046,
+      "step": 66090
+    },
+    {
+      "epoch": 0.33009563284975907,
+      "grad_norm": 0.09192916750907898,
+      "learning_rate": 2.014743798342887e-05,
+      "loss": 9.3079,
+      "step": 66100
+    },
+    {
+      "epoch": 0.3301455716746985,
+      "grad_norm": 0.0940568745136261,
+      "learning_rate": 2.014593606848732e-05,
+      "loss": 9.2998,
+      "step": 66110
+    },
+    {
+      "epoch": 0.33019551049963797,
+      "grad_norm": 0.09915883094072342,
+      "learning_rate": 2.014443415354577e-05,
+      "loss": 9.2972,
+      "step": 66120
+    },
+    {
+      "epoch": 0.3302454493245774,
+      "grad_norm": 0.09629667550325394,
+      "learning_rate": 2.0142932238604222e-05,
+      "loss": 9.3091,
+      "step": 66130
+    },
+    {
+      "epoch": 0.33029538814951687,
+      "grad_norm": 0.08921325206756592,
+      "learning_rate": 2.0141430323662672e-05,
+      "loss": 9.3056,
+      "step": 66140
+    },
+    {
+      "epoch": 0.3303453269744563,
+      "grad_norm": 0.09300815314054489,
+      "learning_rate": 2.013992840872112e-05,
+      "loss": 9.298,
+      "step": 66150
+    },
+    {
+      "epoch": 0.33039526579939577,
+      "grad_norm": 0.09541340917348862,
+      "learning_rate": 2.013842649377957e-05,
+      "loss": 9.323,
+      "step": 66160
+    },
+    {
+      "epoch": 0.3304452046243352,
+      "grad_norm": 0.09386754035949707,
+      "learning_rate": 2.013692457883802e-05,
+      "loss": 9.3023,
+      "step": 66170
+    },
+    {
+      "epoch": 0.33049514344927466,
+      "grad_norm": 0.09818427264690399,
+      "learning_rate": 2.013542266389647e-05,
+      "loss": 9.2981,
+      "step": 66180
+    },
+    {
+      "epoch": 0.3305450822742141,
+      "grad_norm": 0.09862353652715683,
+      "learning_rate": 2.013392074895492e-05,
+      "loss": 9.3156,
+      "step": 66190
+    },
+    {
+      "epoch": 0.33059502109915356,
+      "grad_norm": 0.10540249943733215,
+      "learning_rate": 2.0132418834013366e-05,
+      "loss": 9.3077,
+      "step": 66200
+    },
+    {
+      "epoch": 0.330644959924093,
+      "grad_norm": 0.09350631386041641,
+      "learning_rate": 2.0130916919071816e-05,
+      "loss": 9.309,
+      "step": 66210
+    },
+    {
+      "epoch": 0.33069489874903246,
+      "grad_norm": 0.09390132129192352,
+      "learning_rate": 2.0129415004130267e-05,
+      "loss": 9.2976,
+      "step": 66220
+    },
+    {
+      "epoch": 0.3307448375739719,
+      "grad_norm": 0.09093783795833588,
+      "learning_rate": 2.0127913089188717e-05,
+      "loss": 9.3009,
+      "step": 66230
+    },
+    {
+      "epoch": 0.33079477639891136,
+      "grad_norm": 0.09060053527355194,
+      "learning_rate": 2.0126411174247167e-05,
+      "loss": 9.2951,
+      "step": 66240
+    },
+    {
+      "epoch": 0.3308447152238508,
+      "grad_norm": 0.09574903547763824,
+      "learning_rate": 2.0124909259305614e-05,
+      "loss": 9.3095,
+      "step": 66250
+    },
+    {
+      "epoch": 0.3308946540487902,
+      "grad_norm": 0.09056057035923004,
+      "learning_rate": 2.0123407344364064e-05,
+      "loss": 9.3067,
+      "step": 66260
+    },
+    {
+      "epoch": 0.3309445928737297,
+      "grad_norm": 0.09261029213666916,
+      "learning_rate": 2.0121905429422514e-05,
+      "loss": 9.3051,
+      "step": 66270
+    },
+    {
+      "epoch": 0.3309945316986691,
+      "grad_norm": 0.0922454297542572,
+      "learning_rate": 2.0120403514480964e-05,
+      "loss": 9.3012,
+      "step": 66280
+    },
+    {
+      "epoch": 0.3310444705236086,
+      "grad_norm": 0.0930676981806755,
+      "learning_rate": 2.0118901599539414e-05,
+      "loss": 9.2958,
+      "step": 66290
+    },
+    {
+      "epoch": 0.331094409348548,
+      "grad_norm": 0.0963793620467186,
+      "learning_rate": 2.011739968459786e-05,
+      "loss": 9.2974,
+      "step": 66300
+    },
+    {
+      "epoch": 0.3311443481734875,
+      "grad_norm": 0.09505847096443176,
+      "learning_rate": 2.011589776965631e-05,
+      "loss": 9.3042,
+      "step": 66310
+    },
+    {
+      "epoch": 0.3311942869984269,
+      "grad_norm": 0.09292503446340561,
+      "learning_rate": 2.011439585471476e-05,
+      "loss": 9.3068,
+      "step": 66320
+    },
+    {
+      "epoch": 0.3312442258233664,
+      "grad_norm": 0.09125857800245285,
+      "learning_rate": 2.0112893939773212e-05,
+      "loss": 9.3085,
+      "step": 66330
+    },
+    {
+      "epoch": 0.3312941646483058,
+      "grad_norm": 0.09520285576581955,
+      "learning_rate": 2.0111392024831662e-05,
+      "loss": 9.3046,
+      "step": 66340
+    },
+    {
+      "epoch": 0.3313441034732453,
+      "grad_norm": 0.0909276232123375,
+      "learning_rate": 2.010989010989011e-05,
+      "loss": 9.3044,
+      "step": 66350
+    },
+    {
+      "epoch": 0.3313940422981847,
+      "grad_norm": 0.0963204950094223,
+      "learning_rate": 2.010838819494856e-05,
+      "loss": 9.3014,
+      "step": 66360
+    },
+    {
+      "epoch": 0.3314439811231242,
+      "grad_norm": 0.09370825439691544,
+      "learning_rate": 2.010688628000701e-05,
+      "loss": 9.3011,
+      "step": 66370
+    },
+    {
+      "epoch": 0.3314939199480636,
+      "grad_norm": 0.09406444430351257,
+      "learning_rate": 2.010538436506546e-05,
+      "loss": 9.3059,
+      "step": 66380
+    },
+    {
+      "epoch": 0.3315438587730031,
+      "grad_norm": 0.0930875912308693,
+      "learning_rate": 2.010388245012391e-05,
+      "loss": 9.2974,
+      "step": 66390
+    },
+    {
+      "epoch": 0.3315937975979425,
+      "grad_norm": 0.09899143129587173,
+      "learning_rate": 2.0102380535182356e-05,
+      "loss": 9.2958,
+      "step": 66400
+    },
+    {
+      "epoch": 0.331643736422882,
+      "grad_norm": 0.09795042127370834,
+      "learning_rate": 2.0100878620240806e-05,
+      "loss": 9.2952,
+      "step": 66410
+    },
+    {
+      "epoch": 0.3316936752478214,
+      "grad_norm": 0.09661424905061722,
+      "learning_rate": 2.0099376705299257e-05,
+      "loss": 9.3051,
+      "step": 66420
+    },
+    {
+      "epoch": 0.3317436140727609,
+      "grad_norm": 0.09605047106742859,
+      "learning_rate": 2.0097874790357707e-05,
+      "loss": 9.2961,
+      "step": 66430
+    },
+    {
+      "epoch": 0.3317935528977003,
+      "grad_norm": 0.09522906690835953,
+      "learning_rate": 2.0096372875416157e-05,
+      "loss": 9.3023,
+      "step": 66440
+    },
+    {
+      "epoch": 0.33184349172263977,
+      "grad_norm": 0.09131506085395813,
+      "learning_rate": 2.0094870960474607e-05,
+      "loss": 9.2996,
+      "step": 66450
+    },
+    {
+      "epoch": 0.3318934305475792,
+      "grad_norm": 0.09322021901607513,
+      "learning_rate": 2.0093369045533054e-05,
+      "loss": 9.3029,
+      "step": 66460
+    },
+    {
+      "epoch": 0.33194336937251867,
+      "grad_norm": 0.09292212128639221,
+      "learning_rate": 2.0091867130591504e-05,
+      "loss": 9.3136,
+      "step": 66470
+    },
+    {
+      "epoch": 0.3319933081974581,
+      "grad_norm": 0.09753609448671341,
+      "learning_rate": 2.0090365215649954e-05,
+      "loss": 9.2932,
+      "step": 66480
+    },
+    {
+      "epoch": 0.33204324702239757,
+      "grad_norm": 0.09411279857158661,
+      "learning_rate": 2.0088863300708404e-05,
+      "loss": 9.3017,
+      "step": 66490
+    },
+    {
+      "epoch": 0.332093185847337,
+      "grad_norm": 0.09396139532327652,
+      "learning_rate": 2.0087361385766855e-05,
+      "loss": 9.296,
+      "step": 66500
+    },
+    {
+      "epoch": 0.33214312467227647,
+      "grad_norm": 0.089934341609478,
+      "learning_rate": 2.00858594708253e-05,
+      "loss": 9.2877,
+      "step": 66510
+    },
+    {
+      "epoch": 0.3321930634972159,
+      "grad_norm": 0.09298644214868546,
+      "learning_rate": 2.008435755588375e-05,
+      "loss": 9.2991,
+      "step": 66520
+    },
+    {
+      "epoch": 0.33224300232215537,
+      "grad_norm": 0.09051031619310379,
+      "learning_rate": 2.0082855640942202e-05,
+      "loss": 9.2918,
+      "step": 66530
+    },
+    {
+      "epoch": 0.3322929411470948,
+      "grad_norm": 0.0935496836900711,
+      "learning_rate": 2.0081353726000652e-05,
+      "loss": 9.3055,
+      "step": 66540
+    },
+    {
+      "epoch": 0.33234287997203427,
+      "grad_norm": 0.0968925952911377,
+      "learning_rate": 2.0079851811059102e-05,
+      "loss": 9.2846,
+      "step": 66550
+    },
+    {
+      "epoch": 0.3323928187969737,
+      "grad_norm": 0.09733214229345322,
+      "learning_rate": 2.007834989611755e-05,
+      "loss": 9.2999,
+      "step": 66560
+    },
+    {
+      "epoch": 0.33244275762191317,
+      "grad_norm": 0.08981287479400635,
+      "learning_rate": 2.0076847981176e-05,
+      "loss": 9.2988,
+      "step": 66570
+    },
+    {
+      "epoch": 0.3324926964468526,
+      "grad_norm": 0.09502604603767395,
+      "learning_rate": 2.007534606623445e-05,
+      "loss": 9.3041,
+      "step": 66580
+    },
+    {
+      "epoch": 0.33254263527179206,
+      "grad_norm": 0.0965120941400528,
+      "learning_rate": 2.00738441512929e-05,
+      "loss": 9.3026,
+      "step": 66590
+    },
+    {
+      "epoch": 0.3325925740967315,
+      "grad_norm": 0.09275338053703308,
+      "learning_rate": 2.007234223635135e-05,
+      "loss": 9.3102,
+      "step": 66600
+    },
+    {
+      "epoch": 0.33264251292167096,
+      "grad_norm": 0.09039387851953506,
+      "learning_rate": 2.0070840321409796e-05,
+      "loss": 9.3041,
+      "step": 66610
+    },
+    {
+      "epoch": 0.3326924517466104,
+      "grad_norm": 0.09301149100065231,
+      "learning_rate": 2.0069338406468247e-05,
+      "loss": 9.295,
+      "step": 66620
+    },
+    {
+      "epoch": 0.33274239057154986,
+      "grad_norm": 0.09034721553325653,
+      "learning_rate": 2.0067836491526697e-05,
+      "loss": 9.3058,
+      "step": 66630
+    },
+    {
+      "epoch": 0.3327923293964893,
+      "grad_norm": 0.09412179887294769,
+      "learning_rate": 2.0066334576585147e-05,
+      "loss": 9.2925,
+      "step": 66640
+    },
+    {
+      "epoch": 0.33284226822142876,
+      "grad_norm": 0.09281489998102188,
+      "learning_rate": 2.0064832661643597e-05,
+      "loss": 9.303,
+      "step": 66650
+    },
+    {
+      "epoch": 0.3328922070463682,
+      "grad_norm": 0.09423484653234482,
+      "learning_rate": 2.0063330746702044e-05,
+      "loss": 9.3017,
+      "step": 66660
+    },
+    {
+      "epoch": 0.33294214587130766,
+      "grad_norm": 0.0937783271074295,
+      "learning_rate": 2.0061828831760494e-05,
+      "loss": 9.2948,
+      "step": 66670
+    },
+    {
+      "epoch": 0.3329920846962471,
+      "grad_norm": 0.09942499548196793,
+      "learning_rate": 2.0060326916818944e-05,
+      "loss": 9.3057,
+      "step": 66680
+    },
+    {
+      "epoch": 0.33304202352118656,
+      "grad_norm": 0.09352242201566696,
+      "learning_rate": 2.0058825001877394e-05,
+      "loss": 9.3025,
+      "step": 66690
+    },
+    {
+      "epoch": 0.333091962346126,
+      "grad_norm": 0.09306678175926208,
+      "learning_rate": 2.0057323086935845e-05,
+      "loss": 9.2967,
+      "step": 66700
+    },
+    {
+      "epoch": 0.33314190117106546,
+      "grad_norm": 0.09103065729141235,
+      "learning_rate": 2.005582117199429e-05,
+      "loss": 9.2991,
+      "step": 66710
+    },
+    {
+      "epoch": 0.3331918399960049,
+      "grad_norm": 0.0972660630941391,
+      "learning_rate": 2.005431925705274e-05,
+      "loss": 9.304,
+      "step": 66720
+    },
+    {
+      "epoch": 0.33324177882094436,
+      "grad_norm": 0.09878892451524734,
+      "learning_rate": 2.0052817342111192e-05,
+      "loss": 9.2933,
+      "step": 66730
+    },
+    {
+      "epoch": 0.3332917176458838,
+      "grad_norm": 0.08945970237255096,
+      "learning_rate": 2.0051315427169642e-05,
+      "loss": 9.296,
+      "step": 66740
+    },
+    {
+      "epoch": 0.33334165647082326,
+      "grad_norm": 0.0968952402472496,
+      "learning_rate": 2.0049813512228092e-05,
+      "loss": 9.304,
+      "step": 66750
+    },
+    {
+      "epoch": 0.3333915952957627,
+      "grad_norm": 0.09566505998373032,
+      "learning_rate": 2.004831159728654e-05,
+      "loss": 9.2936,
+      "step": 66760
+    },
+    {
+      "epoch": 0.33344153412070215,
+      "grad_norm": 0.09034927189350128,
+      "learning_rate": 2.0046809682344993e-05,
+      "loss": 9.2968,
+      "step": 66770
+    },
+    {
+      "epoch": 0.3334914729456416,
+      "grad_norm": 0.09364010393619537,
+      "learning_rate": 2.004530776740344e-05,
+      "loss": 9.3066,
+      "step": 66780
+    },
+    {
+      "epoch": 0.33354141177058105,
+      "grad_norm": 0.09636852890253067,
+      "learning_rate": 2.004380585246189e-05,
+      "loss": 9.3017,
+      "step": 66790
+    },
+    {
+      "epoch": 0.3335913505955205,
+      "grad_norm": 0.09206737577915192,
+      "learning_rate": 2.004230393752034e-05,
+      "loss": 9.3021,
+      "step": 66800
+    },
+    {
+      "epoch": 0.33364128942045995,
+      "grad_norm": 0.09280619025230408,
+      "learning_rate": 2.0040802022578786e-05,
+      "loss": 9.2885,
+      "step": 66810
+    },
+    {
+      "epoch": 0.3336912282453994,
+      "grad_norm": 0.09028727561235428,
+      "learning_rate": 2.003930010763724e-05,
+      "loss": 9.2951,
+      "step": 66820
+    },
+    {
+      "epoch": 0.33374116707033885,
+      "grad_norm": 0.0951092317700386,
+      "learning_rate": 2.0037798192695687e-05,
+      "loss": 9.305,
+      "step": 66830
+    },
+    {
+      "epoch": 0.3337911058952783,
+      "grad_norm": 0.09845414012670517,
+      "learning_rate": 2.0036296277754137e-05,
+      "loss": 9.2961,
+      "step": 66840
+    },
+    {
+      "epoch": 0.33384104472021775,
+      "grad_norm": 0.09636221081018448,
+      "learning_rate": 2.0034794362812587e-05,
+      "loss": 9.2949,
+      "step": 66850
+    },
+    {
+      "epoch": 0.33389098354515717,
+      "grad_norm": 0.09213186800479889,
+      "learning_rate": 2.0033292447871034e-05,
+      "loss": 9.2933,
+      "step": 66860
+    },
+    {
+      "epoch": 0.33394092237009665,
+      "grad_norm": 0.09357234835624695,
+      "learning_rate": 2.0031790532929488e-05,
+      "loss": 9.3005,
+      "step": 66870
+    },
+    {
+      "epoch": 0.33399086119503607,
+      "grad_norm": 0.0960192084312439,
+      "learning_rate": 2.0030288617987934e-05,
+      "loss": 9.2875,
+      "step": 66880
+    },
+    {
+      "epoch": 0.33404080001997555,
+      "grad_norm": 0.09275079518556595,
+      "learning_rate": 2.0028786703046385e-05,
+      "loss": 9.2984,
+      "step": 66890
+    },
+    {
+      "epoch": 0.33409073884491497,
+      "grad_norm": 0.08964371681213379,
+      "learning_rate": 2.0027284788104835e-05,
+      "loss": 9.301,
+      "step": 66900
+    },
+    {
+      "epoch": 0.33414067766985445,
+      "grad_norm": 0.096300408244133,
+      "learning_rate": 2.002578287316328e-05,
+      "loss": 9.3032,
+      "step": 66910
+    },
+    {
+      "epoch": 0.33419061649479387,
+      "grad_norm": 0.09406279772520065,
+      "learning_rate": 2.0024280958221735e-05,
+      "loss": 9.29,
+      "step": 66920
+    },
+    {
+      "epoch": 0.33424055531973335,
+      "grad_norm": 0.09408601373434067,
+      "learning_rate": 2.0022779043280182e-05,
+      "loss": 9.2842,
+      "step": 66930
+    },
+    {
+      "epoch": 0.33429049414467277,
+      "grad_norm": 0.10076167434453964,
+      "learning_rate": 2.0021277128338632e-05,
+      "loss": 9.2893,
+      "step": 66940
+    },
+    {
+      "epoch": 0.33434043296961224,
+      "grad_norm": 0.09452462196350098,
+      "learning_rate": 2.0019775213397082e-05,
+      "loss": 9.3011,
+      "step": 66950
+    },
+    {
+      "epoch": 0.33439037179455167,
+      "grad_norm": 0.1001209020614624,
+      "learning_rate": 2.001827329845553e-05,
+      "loss": 9.2893,
+      "step": 66960
+    },
+    {
+      "epoch": 0.33444031061949114,
+      "grad_norm": 0.09850946813821793,
+      "learning_rate": 2.0016771383513983e-05,
+      "loss": 9.2799,
+      "step": 66970
+    },
+    {
+      "epoch": 0.33449024944443057,
+      "grad_norm": 0.09696486592292786,
+      "learning_rate": 2.001526946857243e-05,
+      "loss": 9.2929,
+      "step": 66980
+    },
+    {
+      "epoch": 0.33454018826937004,
+      "grad_norm": 0.09150009602308273,
+      "learning_rate": 2.001376755363088e-05,
+      "loss": 9.2956,
+      "step": 66990
+    },
+    {
+      "epoch": 0.33459012709430946,
+      "grad_norm": 0.09880898147821426,
+      "learning_rate": 2.001226563868933e-05,
+      "loss": 9.2894,
+      "step": 67000
+    },
+    {
+      "epoch": 0.33464006591924894,
+      "grad_norm": 0.1000562235713005,
+      "learning_rate": 2.0010763723747776e-05,
+      "loss": 9.2878,
+      "step": 67010
+    },
+    {
+      "epoch": 0.33469000474418836,
+      "grad_norm": 0.09261028468608856,
+      "learning_rate": 2.000926180880623e-05,
+      "loss": 9.3001,
+      "step": 67020
+    },
+    {
+      "epoch": 0.33473994356912784,
+      "grad_norm": 0.09857013821601868,
+      "learning_rate": 2.0007759893864677e-05,
+      "loss": 9.2877,
+      "step": 67030
+    },
+    {
+      "epoch": 0.33478988239406726,
+      "grad_norm": 0.09461692720651627,
+      "learning_rate": 2.0006257978923127e-05,
+      "loss": 9.2916,
+      "step": 67040
+    },
+    {
+      "epoch": 0.33483982121900674,
+      "grad_norm": 0.09334218502044678,
+      "learning_rate": 2.0004756063981577e-05,
+      "loss": 9.2941,
+      "step": 67050
+    },
+    {
+      "epoch": 0.33488976004394616,
+      "grad_norm": 0.09218782186508179,
+      "learning_rate": 2.0003254149040024e-05,
+      "loss": 9.2912,
+      "step": 67060
+    },
+    {
+      "epoch": 0.33493969886888564,
+      "grad_norm": 0.09374010562896729,
+      "learning_rate": 2.0001752234098478e-05,
+      "loss": 9.3061,
+      "step": 67070
+    },
+    {
+      "epoch": 0.33498963769382506,
+      "grad_norm": 0.09382879734039307,
+      "learning_rate": 2.0000250319156924e-05,
+      "loss": 9.2875,
+      "step": 67080
+    },
+    {
+      "epoch": 0.33503957651876454,
+      "grad_norm": 0.0962020605802536,
+      "learning_rate": 1.9998748404215378e-05,
+      "loss": 9.2897,
+      "step": 67090
+    },
+    {
+      "epoch": 0.33508951534370396,
+      "grad_norm": 0.08903557062149048,
+      "learning_rate": 1.9997246489273825e-05,
+      "loss": 9.2978,
+      "step": 67100
+    },
+    {
+      "epoch": 0.33513945416864344,
+      "grad_norm": 0.09756463766098022,
+      "learning_rate": 1.999574457433227e-05,
+      "loss": 9.2837,
+      "step": 67110
+    },
+    {
+      "epoch": 0.33518939299358286,
+      "grad_norm": 0.09464458376169205,
+      "learning_rate": 1.9994242659390725e-05,
+      "loss": 9.2818,
+      "step": 67120
+    },
+    {
+      "epoch": 0.33523933181852233,
+      "grad_norm": 0.09560305625200272,
+      "learning_rate": 1.9992740744449172e-05,
+      "loss": 9.299,
+      "step": 67130
+    },
+    {
+      "epoch": 0.33528927064346176,
+      "grad_norm": 0.09456129372119904,
+      "learning_rate": 1.9991238829507625e-05,
+      "loss": 9.2885,
+      "step": 67140
+    },
+    {
+      "epoch": 0.33533920946840123,
+      "grad_norm": 0.09367375075817108,
+      "learning_rate": 1.9989736914566072e-05,
+      "loss": 9.2979,
+      "step": 67150
+    },
+    {
+      "epoch": 0.33538914829334066,
+      "grad_norm": 0.09220660477876663,
+      "learning_rate": 1.998823499962452e-05,
+      "loss": 9.2871,
+      "step": 67160
+    },
+    {
+      "epoch": 0.33543908711828013,
+      "grad_norm": 0.09274411201477051,
+      "learning_rate": 1.9986733084682973e-05,
+      "loss": 9.2825,
+      "step": 67170
+    },
+    {
+      "epoch": 0.33548902594321955,
+      "grad_norm": 0.09277375042438507,
+      "learning_rate": 1.998523116974142e-05,
+      "loss": 9.2908,
+      "step": 67180
+    },
+    {
+      "epoch": 0.33553896476815903,
+      "grad_norm": 0.09955313801765442,
+      "learning_rate": 1.9983729254799873e-05,
+      "loss": 9.2862,
+      "step": 67190
+    },
+    {
+      "epoch": 0.33558890359309845,
+      "grad_norm": 0.09518168866634369,
+      "learning_rate": 1.998222733985832e-05,
+      "loss": 9.2919,
+      "step": 67200
+    },
+    {
+      "epoch": 0.33563884241803793,
+      "grad_norm": 0.09314264357089996,
+      "learning_rate": 1.9980725424916767e-05,
+      "loss": 9.2893,
+      "step": 67210
+    },
+    {
+      "epoch": 0.33568878124297735,
+      "grad_norm": 0.09429054707288742,
+      "learning_rate": 1.997922350997522e-05,
+      "loss": 9.3004,
+      "step": 67220
+    },
+    {
+      "epoch": 0.33573872006791683,
+      "grad_norm": 0.09523293375968933,
+      "learning_rate": 1.9977721595033667e-05,
+      "loss": 9.3031,
+      "step": 67230
+    },
+    {
+      "epoch": 0.33578865889285625,
+      "grad_norm": 0.09158944338560104,
+      "learning_rate": 1.997621968009212e-05,
+      "loss": 9.2918,
+      "step": 67240
+    },
+    {
+      "epoch": 0.3358385977177957,
+      "grad_norm": 0.09265114367008209,
+      "learning_rate": 1.9974717765150567e-05,
+      "loss": 9.2749,
+      "step": 67250
+    },
+    {
+      "epoch": 0.33588853654273515,
+      "grad_norm": 0.09522463381290436,
+      "learning_rate": 1.9973215850209014e-05,
+      "loss": 9.288,
+      "step": 67260
+    },
+    {
+      "epoch": 0.33593847536767457,
+      "grad_norm": 0.08956436067819595,
+      "learning_rate": 1.9971713935267468e-05,
+      "loss": 9.2902,
+      "step": 67270
+    },
+    {
+      "epoch": 0.33598841419261405,
+      "grad_norm": 0.09145728498697281,
+      "learning_rate": 1.9970212020325914e-05,
+      "loss": 9.2872,
+      "step": 67280
+    },
+    {
+      "epoch": 0.33603835301755347,
+      "grad_norm": 0.093343086540699,
+      "learning_rate": 1.9968710105384368e-05,
+      "loss": 9.2793,
+      "step": 67290
+    },
+    {
+      "epoch": 0.33608829184249295,
+      "grad_norm": 0.08741866797208786,
+      "learning_rate": 1.9967208190442815e-05,
+      "loss": 9.2963,
+      "step": 67300
+    },
+    {
+      "epoch": 0.33613823066743237,
+      "grad_norm": 0.09440305083990097,
+      "learning_rate": 1.996570627550126e-05,
+      "loss": 9.2873,
+      "step": 67310
+    },
+    {
+      "epoch": 0.33618816949237185,
+      "grad_norm": 0.09325043857097626,
+      "learning_rate": 1.9964204360559715e-05,
+      "loss": 9.2893,
+      "step": 67320
+    },
+    {
+      "epoch": 0.33623810831731127,
+      "grad_norm": 0.09446980804204941,
+      "learning_rate": 1.9962702445618162e-05,
+      "loss": 9.2772,
+      "step": 67330
+    },
+    {
+      "epoch": 0.33628804714225075,
+      "grad_norm": 0.09616648405790329,
+      "learning_rate": 1.9961200530676615e-05,
+      "loss": 9.2826,
+      "step": 67340
+    },
+    {
+      "epoch": 0.33633798596719017,
+      "grad_norm": 0.09551723301410675,
+      "learning_rate": 1.9959698615735062e-05,
+      "loss": 9.279,
+      "step": 67350
+    },
+    {
+      "epoch": 0.33638792479212964,
+      "grad_norm": 0.09174014627933502,
+      "learning_rate": 1.995819670079351e-05,
+      "loss": 9.2902,
+      "step": 67360
+    },
+    {
+      "epoch": 0.33643786361706907,
+      "grad_norm": 0.09369547665119171,
+      "learning_rate": 1.9956694785851963e-05,
+      "loss": 9.297,
+      "step": 67370
+    },
+    {
+      "epoch": 0.33648780244200854,
+      "grad_norm": 0.09878925234079361,
+      "learning_rate": 1.995519287091041e-05,
+      "loss": 9.283,
+      "step": 67380
+    },
+    {
+      "epoch": 0.33653774126694797,
+      "grad_norm": 0.09697102010250092,
+      "learning_rate": 1.9953690955968863e-05,
+      "loss": 9.2967,
+      "step": 67390
+    },
+    {
+      "epoch": 0.33658768009188744,
+      "grad_norm": 0.10312024503946304,
+      "learning_rate": 1.995218904102731e-05,
+      "loss": 9.2857,
+      "step": 67400
+    },
+    {
+      "epoch": 0.33663761891682686,
+      "grad_norm": 0.10092073678970337,
+      "learning_rate": 1.995068712608576e-05,
+      "loss": 9.2912,
+      "step": 67410
+    },
+    {
+      "epoch": 0.33668755774176634,
+      "grad_norm": 0.08944955468177795,
+      "learning_rate": 1.994918521114421e-05,
+      "loss": 9.2866,
+      "step": 67420
+    },
+    {
+      "epoch": 0.33673749656670576,
+      "grad_norm": 0.09892292320728302,
+      "learning_rate": 1.9947683296202657e-05,
+      "loss": 9.2795,
+      "step": 67430
+    },
+    {
+      "epoch": 0.33678743539164524,
+      "grad_norm": 0.09553686529397964,
+      "learning_rate": 1.994618138126111e-05,
+      "loss": 9.2821,
+      "step": 67440
+    },
+    {
+      "epoch": 0.33683737421658466,
+      "grad_norm": 0.09259875863790512,
+      "learning_rate": 1.9944679466319557e-05,
+      "loss": 9.2918,
+      "step": 67450
+    },
+    {
+      "epoch": 0.33688731304152414,
+      "grad_norm": 0.09672240912914276,
+      "learning_rate": 1.9943177551378007e-05,
+      "loss": 9.2893,
+      "step": 67460
+    },
+    {
+      "epoch": 0.33693725186646356,
+      "grad_norm": 0.09132887423038483,
+      "learning_rate": 1.9941675636436458e-05,
+      "loss": 9.2791,
+      "step": 67470
+    },
+    {
+      "epoch": 0.33698719069140304,
+      "grad_norm": 0.09487363696098328,
+      "learning_rate": 1.9940173721494904e-05,
+      "loss": 9.2924,
+      "step": 67480
+    },
+    {
+      "epoch": 0.33703712951634246,
+      "grad_norm": 0.09358032792806625,
+      "learning_rate": 1.9938671806553358e-05,
+      "loss": 9.2942,
+      "step": 67490
+    },
+    {
+      "epoch": 0.33708706834128194,
+      "grad_norm": 0.09767814725637436,
+      "learning_rate": 1.9937169891611805e-05,
+      "loss": 9.2936,
+      "step": 67500
+    },
+    {
+      "epoch": 0.33713700716622136,
+      "grad_norm": 0.0938803032040596,
+      "learning_rate": 1.9935667976670255e-05,
+      "loss": 9.2903,
+      "step": 67510
+    },
+    {
+      "epoch": 0.33718694599116084,
+      "grad_norm": 0.09350651502609253,
+      "learning_rate": 1.9934166061728705e-05,
+      "loss": 9.2828,
+      "step": 67520
+    },
+    {
+      "epoch": 0.33723688481610026,
+      "grad_norm": 0.09044313430786133,
+      "learning_rate": 1.9932664146787152e-05,
+      "loss": 9.2797,
+      "step": 67530
+    },
+    {
+      "epoch": 0.33728682364103973,
+      "grad_norm": 0.09183132648468018,
+      "learning_rate": 1.9931162231845605e-05,
+      "loss": 9.2823,
+      "step": 67540
+    },
+    {
+      "epoch": 0.33733676246597916,
+      "grad_norm": 0.09145781397819519,
+      "learning_rate": 1.9929660316904052e-05,
+      "loss": 9.2747,
+      "step": 67550
+    },
+    {
+      "epoch": 0.33738670129091863,
+      "grad_norm": 0.09186349809169769,
+      "learning_rate": 1.9928158401962502e-05,
+      "loss": 9.283,
+      "step": 67560
+    },
+    {
+      "epoch": 0.33743664011585806,
+      "grad_norm": 0.09399545937776566,
+      "learning_rate": 1.9926656487020953e-05,
+      "loss": 9.2948,
+      "step": 67570
+    },
+    {
+      "epoch": 0.33748657894079753,
+      "grad_norm": 0.09273233264684677,
+      "learning_rate": 1.99251545720794e-05,
+      "loss": 9.2942,
+      "step": 67580
+    },
+    {
+      "epoch": 0.33753651776573695,
+      "grad_norm": 0.09469340741634369,
+      "learning_rate": 1.9923652657137853e-05,
+      "loss": 9.2731,
+      "step": 67590
+    },
+    {
+      "epoch": 0.33758645659067643,
+      "grad_norm": 0.0966138169169426,
+      "learning_rate": 1.99221507421963e-05,
+      "loss": 9.2785,
+      "step": 67600
+    },
+    {
+      "epoch": 0.33763639541561585,
+      "grad_norm": 0.08690140396356583,
+      "learning_rate": 1.992064882725475e-05,
+      "loss": 9.2847,
+      "step": 67610
+    },
+    {
+      "epoch": 0.33768633424055533,
+      "grad_norm": 0.09711094200611115,
+      "learning_rate": 1.99191469123132e-05,
+      "loss": 9.2989,
+      "step": 67620
+    },
+    {
+      "epoch": 0.33773627306549475,
+      "grad_norm": 0.09582893550395966,
+      "learning_rate": 1.9917644997371647e-05,
+      "loss": 9.2828,
+      "step": 67630
+    },
+    {
+      "epoch": 0.33778621189043423,
+      "grad_norm": 0.09695571660995483,
+      "learning_rate": 1.99161430824301e-05,
+      "loss": 9.2946,
+      "step": 67640
+    },
+    {
+      "epoch": 0.33783615071537365,
+      "grad_norm": 0.09266763180494308,
+      "learning_rate": 1.9914641167488547e-05,
+      "loss": 9.2909,
+      "step": 67650
+    },
+    {
+      "epoch": 0.33788608954031313,
+      "grad_norm": 0.09419077634811401,
+      "learning_rate": 1.9913139252546997e-05,
+      "loss": 9.2872,
+      "step": 67660
+    },
+    {
+      "epoch": 0.33793602836525255,
+      "grad_norm": 0.09436140954494476,
+      "learning_rate": 1.9911637337605448e-05,
+      "loss": 9.2898,
+      "step": 67670
+    },
+    {
+      "epoch": 0.337985967190192,
+      "grad_norm": 0.09518197923898697,
+      "learning_rate": 1.9910135422663894e-05,
+      "loss": 9.2801,
+      "step": 67680
+    },
+    {
+      "epoch": 0.33803590601513145,
+      "grad_norm": 0.09597449004650116,
+      "learning_rate": 1.9908633507722348e-05,
+      "loss": 9.2772,
+      "step": 67690
+    },
+    {
+      "epoch": 0.3380858448400709,
+      "grad_norm": 0.09348492324352264,
+      "learning_rate": 1.9907131592780795e-05,
+      "loss": 9.2983,
+      "step": 67700
+    },
+    {
+      "epoch": 0.33813578366501035,
+      "grad_norm": 0.09242329746484756,
+      "learning_rate": 1.9905629677839245e-05,
+      "loss": 9.2903,
+      "step": 67710
+    },
+    {
+      "epoch": 0.3381857224899498,
+      "grad_norm": 0.09632088989019394,
+      "learning_rate": 1.9904127762897695e-05,
+      "loss": 9.2947,
+      "step": 67720
+    },
+    {
+      "epoch": 0.33823566131488925,
+      "grad_norm": 0.09371468424797058,
+      "learning_rate": 1.9902625847956142e-05,
+      "loss": 9.2904,
+      "step": 67730
+    },
+    {
+      "epoch": 0.3382856001398287,
+      "grad_norm": 0.09897583723068237,
+      "learning_rate": 1.9901123933014595e-05,
+      "loss": 9.2772,
+      "step": 67740
+    },
+    {
+      "epoch": 0.33833553896476815,
+      "grad_norm": 0.0937107503414154,
+      "learning_rate": 1.9899622018073042e-05,
+      "loss": 9.2801,
+      "step": 67750
+    },
+    {
+      "epoch": 0.3383854777897076,
+      "grad_norm": 0.09183280169963837,
+      "learning_rate": 1.9898120103131492e-05,
+      "loss": 9.2962,
+      "step": 67760
+    },
+    {
+      "epoch": 0.33843541661464704,
+      "grad_norm": 0.09664995968341827,
+      "learning_rate": 1.9896618188189943e-05,
+      "loss": 9.2849,
+      "step": 67770
+    },
+    {
+      "epoch": 0.3384853554395865,
+      "grad_norm": 0.09200311452150345,
+      "learning_rate": 1.9895116273248393e-05,
+      "loss": 9.2824,
+      "step": 67780
+    },
+    {
+      "epoch": 0.33853529426452594,
+      "grad_norm": 0.09540259838104248,
+      "learning_rate": 1.9893614358306843e-05,
+      "loss": 9.2852,
+      "step": 67790
+    },
+    {
+      "epoch": 0.3385852330894654,
+      "grad_norm": 0.09478534013032913,
+      "learning_rate": 1.989211244336529e-05,
+      "loss": 9.2823,
+      "step": 67800
+    },
+    {
+      "epoch": 0.33863517191440484,
+      "grad_norm": 0.09047430753707886,
+      "learning_rate": 1.989061052842374e-05,
+      "loss": 9.2842,
+      "step": 67810
+    },
+    {
+      "epoch": 0.3386851107393443,
+      "grad_norm": 0.09374905377626419,
+      "learning_rate": 1.988910861348219e-05,
+      "loss": 9.2843,
+      "step": 67820
+    },
+    {
+      "epoch": 0.33873504956428374,
+      "grad_norm": 0.08902385830879211,
+      "learning_rate": 1.988760669854064e-05,
+      "loss": 9.2835,
+      "step": 67830
+    },
+    {
+      "epoch": 0.3387849883892232,
+      "grad_norm": 0.09766287356615067,
+      "learning_rate": 1.988610478359909e-05,
+      "loss": 9.2869,
+      "step": 67840
+    },
+    {
+      "epoch": 0.33883492721416264,
+      "grad_norm": 0.0938125029206276,
+      "learning_rate": 1.9884602868657537e-05,
+      "loss": 9.2816,
+      "step": 67850
+    },
+    {
+      "epoch": 0.3388848660391021,
+      "grad_norm": 0.09323128312826157,
+      "learning_rate": 1.9883100953715987e-05,
+      "loss": 9.268,
+      "step": 67860
+    },
+    {
+      "epoch": 0.33893480486404154,
+      "grad_norm": 0.092827707529068,
+      "learning_rate": 1.9881599038774438e-05,
+      "loss": 9.2823,
+      "step": 67870
+    },
+    {
+      "epoch": 0.338984743688981,
+      "grad_norm": 0.09353949874639511,
+      "learning_rate": 1.9880097123832888e-05,
+      "loss": 9.2872,
+      "step": 67880
+    },
+    {
+      "epoch": 0.33903468251392044,
+      "grad_norm": 0.09071552753448486,
+      "learning_rate": 1.9878595208891338e-05,
+      "loss": 9.2822,
+      "step": 67890
+    },
+    {
+      "epoch": 0.3390846213388599,
+      "grad_norm": 0.09243173152208328,
+      "learning_rate": 1.9877093293949785e-05,
+      "loss": 9.2805,
+      "step": 67900
+    },
+    {
+      "epoch": 0.33913456016379934,
+      "grad_norm": 0.09218911081552505,
+      "learning_rate": 1.987559137900824e-05,
+      "loss": 9.2797,
+      "step": 67910
+    },
+    {
+      "epoch": 0.3391844989887388,
+      "grad_norm": 0.09476479142904282,
+      "learning_rate": 1.9874089464066685e-05,
+      "loss": 9.2845,
+      "step": 67920
+    },
+    {
+      "epoch": 0.33923443781367824,
+      "grad_norm": 0.10015514492988586,
+      "learning_rate": 1.9872587549125135e-05,
+      "loss": 9.2723,
+      "step": 67930
+    },
+    {
+      "epoch": 0.3392843766386177,
+      "grad_norm": 0.09330648928880692,
+      "learning_rate": 1.9871085634183585e-05,
+      "loss": 9.2797,
+      "step": 67940
+    },
+    {
+      "epoch": 0.33933431546355713,
+      "grad_norm": 0.0925719290971756,
+      "learning_rate": 1.9869583719242032e-05,
+      "loss": 9.2761,
+      "step": 67950
+    },
+    {
+      "epoch": 0.3393842542884966,
+      "grad_norm": 0.0938037782907486,
+      "learning_rate": 1.9868081804300486e-05,
+      "loss": 9.2795,
+      "step": 67960
+    },
+    {
+      "epoch": 0.33943419311343603,
+      "grad_norm": 0.09363068640232086,
+      "learning_rate": 1.9866579889358933e-05,
+      "loss": 9.2788,
+      "step": 67970
+    },
+    {
+      "epoch": 0.3394841319383755,
+      "grad_norm": 0.09175083041191101,
+      "learning_rate": 1.9865077974417383e-05,
+      "loss": 9.2829,
+      "step": 67980
+    },
+    {
+      "epoch": 0.33953407076331493,
+      "grad_norm": 0.09043523669242859,
+      "learning_rate": 1.9863576059475833e-05,
+      "loss": 9.2802,
+      "step": 67990
+    },
+    {
+      "epoch": 0.3395840095882544,
+      "grad_norm": 0.09502432495355606,
+      "learning_rate": 1.986207414453428e-05,
+      "loss": 9.2757,
+      "step": 68000
+    },
+    {
+      "epoch": 0.33963394841319383,
+      "grad_norm": 0.09515243023633957,
+      "learning_rate": 1.9860572229592733e-05,
+      "loss": 9.2656,
+      "step": 68010
+    },
+    {
+      "epoch": 0.3396838872381333,
+      "grad_norm": 0.0931011438369751,
+      "learning_rate": 1.985907031465118e-05,
+      "loss": 9.271,
+      "step": 68020
+    },
+    {
+      "epoch": 0.33973382606307273,
+      "grad_norm": 0.09343037009239197,
+      "learning_rate": 1.985756839970963e-05,
+      "loss": 9.2821,
+      "step": 68030
+    },
+    {
+      "epoch": 0.3397837648880122,
+      "grad_norm": 0.08982733637094498,
+      "learning_rate": 1.985606648476808e-05,
+      "loss": 9.2779,
+      "step": 68040
+    },
+    {
+      "epoch": 0.33983370371295163,
+      "grad_norm": 0.09892088919878006,
+      "learning_rate": 1.9854564569826527e-05,
+      "loss": 9.2804,
+      "step": 68050
+    },
+    {
+      "epoch": 0.3398836425378911,
+      "grad_norm": 0.09419706463813782,
+      "learning_rate": 1.985306265488498e-05,
+      "loss": 9.2811,
+      "step": 68060
+    },
+    {
+      "epoch": 0.33993358136283053,
+      "grad_norm": 0.09958070516586304,
+      "learning_rate": 1.9851560739943428e-05,
+      "loss": 9.2808,
+      "step": 68070
+    },
+    {
+      "epoch": 0.33998352018777,
+      "grad_norm": 0.09093392640352249,
+      "learning_rate": 1.9850058825001878e-05,
+      "loss": 9.2843,
+      "step": 68080
+    },
+    {
+      "epoch": 0.3400334590127094,
+      "grad_norm": 0.0960288792848587,
+      "learning_rate": 1.9848556910060328e-05,
+      "loss": 9.2756,
+      "step": 68090
+    },
+    {
+      "epoch": 0.3400833978376489,
+      "grad_norm": 0.0980386883020401,
+      "learning_rate": 1.9847054995118778e-05,
+      "loss": 9.2687,
+      "step": 68100
+    },
+    {
+      "epoch": 0.3401333366625883,
+      "grad_norm": 0.09633202850818634,
+      "learning_rate": 1.984555308017723e-05,
+      "loss": 9.2726,
+      "step": 68110
+    },
+    {
+      "epoch": 0.3401832754875278,
+      "grad_norm": 0.09432492405176163,
+      "learning_rate": 1.9844051165235675e-05,
+      "loss": 9.2645,
+      "step": 68120
+    },
+    {
+      "epoch": 0.3402332143124672,
+      "grad_norm": 0.09382835030555725,
+      "learning_rate": 1.9842549250294125e-05,
+      "loss": 9.2777,
+      "step": 68130
+    },
+    {
+      "epoch": 0.3402831531374067,
+      "grad_norm": 0.09988310188055038,
+      "learning_rate": 1.9841047335352575e-05,
+      "loss": 9.2738,
+      "step": 68140
+    },
+    {
+      "epoch": 0.3403330919623461,
+      "grad_norm": 0.09288916736841202,
+      "learning_rate": 1.9839545420411026e-05,
+      "loss": 9.2727,
+      "step": 68150
+    },
+    {
+      "epoch": 0.3403830307872856,
+      "grad_norm": 0.08969771862030029,
+      "learning_rate": 1.9838043505469476e-05,
+      "loss": 9.2757,
+      "step": 68160
+    },
+    {
+      "epoch": 0.340432969612225,
+      "grad_norm": 0.09481577575206757,
+      "learning_rate": 1.9836541590527923e-05,
+      "loss": 9.2677,
+      "step": 68170
+    },
+    {
+      "epoch": 0.3404829084371645,
+      "grad_norm": 0.09572725743055344,
+      "learning_rate": 1.9835039675586373e-05,
+      "loss": 9.2771,
+      "step": 68180
+    },
+    {
+      "epoch": 0.3405328472621039,
+      "grad_norm": 0.09492908418178558,
+      "learning_rate": 1.9833537760644823e-05,
+      "loss": 9.2752,
+      "step": 68190
+    },
+    {
+      "epoch": 0.3405827860870434,
+      "grad_norm": 0.09422598034143448,
+      "learning_rate": 1.9832035845703273e-05,
+      "loss": 9.2728,
+      "step": 68200
+    },
+    {
+      "epoch": 0.3406327249119828,
+      "grad_norm": 0.08828997611999512,
+      "learning_rate": 1.9830533930761723e-05,
+      "loss": 9.2755,
+      "step": 68210
+    },
+    {
+      "epoch": 0.3406826637369223,
+      "grad_norm": 0.09550122916698456,
+      "learning_rate": 1.982903201582017e-05,
+      "loss": 9.2761,
+      "step": 68220
+    },
+    {
+      "epoch": 0.3407326025618617,
+      "grad_norm": 0.09217823296785355,
+      "learning_rate": 1.982753010087862e-05,
+      "loss": 9.28,
+      "step": 68230
+    },
+    {
+      "epoch": 0.34078254138680114,
+      "grad_norm": 0.09574399888515472,
+      "learning_rate": 1.982602818593707e-05,
+      "loss": 9.2811,
+      "step": 68240
+    },
+    {
+      "epoch": 0.3408324802117406,
+      "grad_norm": 0.09653940051794052,
+      "learning_rate": 1.982452627099552e-05,
+      "loss": 9.2771,
+      "step": 68250
+    },
+    {
+      "epoch": 0.34088241903668004,
+      "grad_norm": 0.09815537184476852,
+      "learning_rate": 1.982302435605397e-05,
+      "loss": 9.2731,
+      "step": 68260
+    },
+    {
+      "epoch": 0.3409323578616195,
+      "grad_norm": 0.09170655906200409,
+      "learning_rate": 1.9821522441112418e-05,
+      "loss": 9.2808,
+      "step": 68270
+    },
+    {
+      "epoch": 0.34098229668655894,
+      "grad_norm": 0.09200761467218399,
+      "learning_rate": 1.9820020526170868e-05,
+      "loss": 9.2747,
+      "step": 68280
+    },
+    {
+      "epoch": 0.3410322355114984,
+      "grad_norm": 0.08978271484375,
+      "learning_rate": 1.9818518611229318e-05,
+      "loss": 9.2733,
+      "step": 68290
+    },
+    {
+      "epoch": 0.34108217433643784,
+      "grad_norm": 0.09526646882295609,
+      "learning_rate": 1.9817016696287768e-05,
+      "loss": 9.2776,
+      "step": 68300
+    },
+    {
+      "epoch": 0.3411321131613773,
+      "grad_norm": 0.0968102440237999,
+      "learning_rate": 1.981551478134622e-05,
+      "loss": 9.267,
+      "step": 68310
+    },
+    {
+      "epoch": 0.34118205198631674,
+      "grad_norm": 0.09626856446266174,
+      "learning_rate": 1.9814012866404665e-05,
+      "loss": 9.2769,
+      "step": 68320
+    },
+    {
+      "epoch": 0.3412319908112562,
+      "grad_norm": 0.09758766740560532,
+      "learning_rate": 1.9812510951463115e-05,
+      "loss": 9.2728,
+      "step": 68330
+    },
+    {
+      "epoch": 0.34128192963619564,
+      "grad_norm": 0.09370464831590652,
+      "learning_rate": 1.9811009036521566e-05,
+      "loss": 9.2702,
+      "step": 68340
+    },
+    {
+      "epoch": 0.3413318684611351,
+      "grad_norm": 0.0944569855928421,
+      "learning_rate": 1.9809507121580016e-05,
+      "loss": 9.282,
+      "step": 68350
+    },
+    {
+      "epoch": 0.34138180728607453,
+      "grad_norm": 0.09909389168024063,
+      "learning_rate": 1.9808005206638466e-05,
+      "loss": 9.2794,
+      "step": 68360
+    },
+    {
+      "epoch": 0.341431746111014,
+      "grad_norm": 0.097555011510849,
+      "learning_rate": 1.9806503291696913e-05,
+      "loss": 9.2815,
+      "step": 68370
+    },
+    {
+      "epoch": 0.34148168493595343,
+      "grad_norm": 0.0935775637626648,
+      "learning_rate": 1.9805001376755363e-05,
+      "loss": 9.2732,
+      "step": 68380
+    },
+    {
+      "epoch": 0.3415316237608929,
+      "grad_norm": 0.09548195451498032,
+      "learning_rate": 1.9803499461813813e-05,
+      "loss": 9.2617,
+      "step": 68390
+    },
+    {
+      "epoch": 0.34158156258583233,
+      "grad_norm": 0.09691401571035385,
+      "learning_rate": 1.9801997546872263e-05,
+      "loss": 9.2762,
+      "step": 68400
+    },
+    {
+      "epoch": 0.3416315014107718,
+      "grad_norm": 0.09655098617076874,
+      "learning_rate": 1.9800495631930713e-05,
+      "loss": 9.2838,
+      "step": 68410
+    },
+    {
+      "epoch": 0.34168144023571123,
+      "grad_norm": 0.09758975356817245,
+      "learning_rate": 1.9798993716989164e-05,
+      "loss": 9.2779,
+      "step": 68420
+    },
+    {
+      "epoch": 0.3417313790606507,
+      "grad_norm": 0.0990617647767067,
+      "learning_rate": 1.979749180204761e-05,
+      "loss": 9.2683,
+      "step": 68430
+    },
+    {
+      "epoch": 0.34178131788559013,
+      "grad_norm": 0.09439375251531601,
+      "learning_rate": 1.979598988710606e-05,
+      "loss": 9.2771,
+      "step": 68440
+    },
+    {
+      "epoch": 0.3418312567105296,
+      "grad_norm": 0.09546418488025665,
+      "learning_rate": 1.979448797216451e-05,
+      "loss": 9.2703,
+      "step": 68450
+    },
+    {
+      "epoch": 0.34188119553546903,
+      "grad_norm": 0.09475396573543549,
+      "learning_rate": 1.979298605722296e-05,
+      "loss": 9.2702,
+      "step": 68460
+    },
+    {
+      "epoch": 0.3419311343604085,
+      "grad_norm": 0.08913654834032059,
+      "learning_rate": 1.979148414228141e-05,
+      "loss": 9.2771,
+      "step": 68470
+    },
+    {
+      "epoch": 0.34198107318534793,
+      "grad_norm": 0.09099783003330231,
+      "learning_rate": 1.9789982227339858e-05,
+      "loss": 9.2692,
+      "step": 68480
+    },
+    {
+      "epoch": 0.3420310120102874,
+      "grad_norm": 0.09147406369447708,
+      "learning_rate": 1.9788480312398308e-05,
+      "loss": 9.2736,
+      "step": 68490
+    },
+    {
+      "epoch": 0.3420809508352268,
+      "grad_norm": 0.09418332576751709,
+      "learning_rate": 1.9786978397456758e-05,
+      "loss": 9.2729,
+      "step": 68500
+    },
+    {
+      "epoch": 0.3421308896601663,
+      "grad_norm": 0.09262189269065857,
+      "learning_rate": 1.978547648251521e-05,
+      "loss": 9.2679,
+      "step": 68510
+    },
+    {
+      "epoch": 0.3421808284851057,
+      "grad_norm": 0.10010916739702225,
+      "learning_rate": 1.978397456757366e-05,
+      "loss": 9.2779,
+      "step": 68520
+    },
+    {
+      "epoch": 0.3422307673100452,
+      "grad_norm": 0.09533961862325668,
+      "learning_rate": 1.9782472652632105e-05,
+      "loss": 9.2709,
+      "step": 68530
+    },
+    {
+      "epoch": 0.3422807061349846,
+      "grad_norm": 0.09138842672109604,
+      "learning_rate": 1.9780970737690556e-05,
+      "loss": 9.2689,
+      "step": 68540
+    },
+    {
+      "epoch": 0.3423306449599241,
+      "grad_norm": 0.09413190186023712,
+      "learning_rate": 1.9779468822749006e-05,
+      "loss": 9.2706,
+      "step": 68550
+    },
+    {
+      "epoch": 0.3423805837848635,
+      "grad_norm": 0.09632963687181473,
+      "learning_rate": 1.9777966907807456e-05,
+      "loss": 9.2716,
+      "step": 68560
+    },
+    {
+      "epoch": 0.342430522609803,
+      "grad_norm": 0.09684199839830399,
+      "learning_rate": 1.9776464992865906e-05,
+      "loss": 9.2731,
+      "step": 68570
+    },
+    {
+      "epoch": 0.3424804614347424,
+      "grad_norm": 0.09297710657119751,
+      "learning_rate": 1.9774963077924353e-05,
+      "loss": 9.2758,
+      "step": 68580
+    },
+    {
+      "epoch": 0.3425304002596819,
+      "grad_norm": 0.09406734257936478,
+      "learning_rate": 1.9773461162982803e-05,
+      "loss": 9.2581,
+      "step": 68590
+    },
+    {
+      "epoch": 0.3425803390846213,
+      "grad_norm": 0.09379177540540695,
+      "learning_rate": 1.9771959248041253e-05,
+      "loss": 9.2652,
+      "step": 68600
+    },
+    {
+      "epoch": 0.3426302779095608,
+      "grad_norm": 0.0957665890455246,
+      "learning_rate": 1.9770457333099703e-05,
+      "loss": 9.2713,
+      "step": 68610
+    },
+    {
+      "epoch": 0.3426802167345002,
+      "grad_norm": 0.0952882319688797,
+      "learning_rate": 1.9768955418158154e-05,
+      "loss": 9.2763,
+      "step": 68620
+    },
+    {
+      "epoch": 0.3427301555594397,
+      "grad_norm": 0.09209991991519928,
+      "learning_rate": 1.97674535032166e-05,
+      "loss": 9.2651,
+      "step": 68630
+    },
+    {
+      "epoch": 0.3427800943843791,
+      "grad_norm": 0.09648279845714569,
+      "learning_rate": 1.976595158827505e-05,
+      "loss": 9.2669,
+      "step": 68640
+    },
+    {
+      "epoch": 0.3428300332093186,
+      "grad_norm": 0.09398400038480759,
+      "learning_rate": 1.97644496733335e-05,
+      "loss": 9.2713,
+      "step": 68650
+    },
+    {
+      "epoch": 0.342879972034258,
+      "grad_norm": 0.0925755649805069,
+      "learning_rate": 1.976294775839195e-05,
+      "loss": 9.2797,
+      "step": 68660
+    },
+    {
+      "epoch": 0.3429299108591975,
+      "grad_norm": 0.0941460132598877,
+      "learning_rate": 1.97614458434504e-05,
+      "loss": 9.2715,
+      "step": 68670
+    },
+    {
+      "epoch": 0.3429798496841369,
+      "grad_norm": 0.0912114679813385,
+      "learning_rate": 1.9759943928508848e-05,
+      "loss": 9.2772,
+      "step": 68680
+    },
+    {
+      "epoch": 0.3430297885090764,
+      "grad_norm": 0.09493990242481232,
+      "learning_rate": 1.9758442013567298e-05,
+      "loss": 9.26,
+      "step": 68690
+    },
+    {
+      "epoch": 0.3430797273340158,
+      "grad_norm": 0.09542454034090042,
+      "learning_rate": 1.9756940098625748e-05,
+      "loss": 9.2713,
+      "step": 68700
+    },
+    {
+      "epoch": 0.3431296661589553,
+      "grad_norm": 0.09154530614614487,
+      "learning_rate": 1.97554381836842e-05,
+      "loss": 9.2737,
+      "step": 68710
+    },
+    {
+      "epoch": 0.3431796049838947,
+      "grad_norm": 0.09703396260738373,
+      "learning_rate": 1.975393626874265e-05,
+      "loss": 9.2639,
+      "step": 68720
+    },
+    {
+      "epoch": 0.3432295438088342,
+      "grad_norm": 0.09249304234981537,
+      "learning_rate": 1.9752434353801095e-05,
+      "loss": 9.265,
+      "step": 68730
+    },
+    {
+      "epoch": 0.3432794826337736,
+      "grad_norm": 0.10178221017122269,
+      "learning_rate": 1.975093243885955e-05,
+      "loss": 9.2766,
+      "step": 68740
+    },
+    {
+      "epoch": 0.3433294214587131,
+      "grad_norm": 0.09313027560710907,
+      "learning_rate": 1.9749430523917996e-05,
+      "loss": 9.2733,
+      "step": 68750
+    },
+    {
+      "epoch": 0.3433793602836525,
+      "grad_norm": 0.09608368575572968,
+      "learning_rate": 1.9747928608976446e-05,
+      "loss": 9.2663,
+      "step": 68760
+    },
+    {
+      "epoch": 0.343429299108592,
+      "grad_norm": 0.09877718985080719,
+      "learning_rate": 1.9746426694034896e-05,
+      "loss": 9.2556,
+      "step": 68770
+    },
+    {
+      "epoch": 0.3434792379335314,
+      "grad_norm": 0.09784068167209625,
+      "learning_rate": 1.9744924779093343e-05,
+      "loss": 9.2615,
+      "step": 68780
+    },
+    {
+      "epoch": 0.3435291767584709,
+      "grad_norm": 0.09086261689662933,
+      "learning_rate": 1.9743422864151796e-05,
+      "loss": 9.2652,
+      "step": 68790
+    },
+    {
+      "epoch": 0.3435791155834103,
+      "grad_norm": 0.09742986410856247,
+      "learning_rate": 1.9741920949210243e-05,
+      "loss": 9.2638,
+      "step": 68800
+    },
+    {
+      "epoch": 0.3436290544083498,
+      "grad_norm": 0.09637963771820068,
+      "learning_rate": 1.9740419034268693e-05,
+      "loss": 9.2678,
+      "step": 68810
+    },
+    {
+      "epoch": 0.3436789932332892,
+      "grad_norm": 0.09257320314645767,
+      "learning_rate": 1.9738917119327144e-05,
+      "loss": 9.2625,
+      "step": 68820
+    },
+    {
+      "epoch": 0.3437289320582287,
+      "grad_norm": 0.09262960404157639,
+      "learning_rate": 1.973741520438559e-05,
+      "loss": 9.2747,
+      "step": 68830
+    },
+    {
+      "epoch": 0.3437788708831681,
+      "grad_norm": 0.09434715658426285,
+      "learning_rate": 1.9735913289444044e-05,
+      "loss": 9.2595,
+      "step": 68840
+    },
+    {
+      "epoch": 0.3438288097081076,
+      "grad_norm": 0.09280122816562653,
+      "learning_rate": 1.973441137450249e-05,
+      "loss": 9.2605,
+      "step": 68850
+    },
+    {
+      "epoch": 0.343878748533047,
+      "grad_norm": 0.09355524182319641,
+      "learning_rate": 1.973290945956094e-05,
+      "loss": 9.2708,
+      "step": 68860
+    },
+    {
+      "epoch": 0.3439286873579865,
+      "grad_norm": 0.09285325556993484,
+      "learning_rate": 1.973140754461939e-05,
+      "loss": 9.277,
+      "step": 68870
+    },
+    {
+      "epoch": 0.3439786261829259,
+      "grad_norm": 0.09656701236963272,
+      "learning_rate": 1.9729905629677838e-05,
+      "loss": 9.2666,
+      "step": 68880
+    },
+    {
+      "epoch": 0.3440285650078654,
+      "grad_norm": 0.10322730988264084,
+      "learning_rate": 1.972840371473629e-05,
+      "loss": 9.2636,
+      "step": 68890
+    },
+    {
+      "epoch": 0.3440785038328048,
+      "grad_norm": 0.09647629410028458,
+      "learning_rate": 1.9726901799794738e-05,
+      "loss": 9.276,
+      "step": 68900
+    },
+    {
+      "epoch": 0.3441284426577443,
+      "grad_norm": 0.09318669140338898,
+      "learning_rate": 1.972539988485319e-05,
+      "loss": 9.2602,
+      "step": 68910
+    },
+    {
+      "epoch": 0.3441783814826837,
+      "grad_norm": 0.09435082972049713,
+      "learning_rate": 1.972389796991164e-05,
+      "loss": 9.2701,
+      "step": 68920
+    },
+    {
+      "epoch": 0.3442283203076232,
+      "grad_norm": 0.09164106100797653,
+      "learning_rate": 1.9722396054970085e-05,
+      "loss": 9.2785,
+      "step": 68930
+    },
+    {
+      "epoch": 0.3442782591325626,
+      "grad_norm": 0.08812560886144638,
+      "learning_rate": 1.972089414002854e-05,
+      "loss": 9.2758,
+      "step": 68940
+    },
+    {
+      "epoch": 0.3443281979575021,
+      "grad_norm": 0.0958331823348999,
+      "learning_rate": 1.9719392225086986e-05,
+      "loss": 9.2655,
+      "step": 68950
+    },
+    {
+      "epoch": 0.3443781367824415,
+      "grad_norm": 0.0955391451716423,
+      "learning_rate": 1.9717890310145436e-05,
+      "loss": 9.2702,
+      "step": 68960
+    },
+    {
+      "epoch": 0.344428075607381,
+      "grad_norm": 0.09317716956138611,
+      "learning_rate": 1.9716388395203886e-05,
+      "loss": 9.2806,
+      "step": 68970
+    },
+    {
+      "epoch": 0.3444780144323204,
+      "grad_norm": 0.09601125121116638,
+      "learning_rate": 1.9714886480262333e-05,
+      "loss": 9.2552,
+      "step": 68980
+    },
+    {
+      "epoch": 0.3445279532572599,
+      "grad_norm": 0.08980926871299744,
+      "learning_rate": 1.9713384565320786e-05,
+      "loss": 9.2754,
+      "step": 68990
+    },
+    {
+      "epoch": 0.3445778920821993,
+      "grad_norm": 0.09495431184768677,
+      "learning_rate": 1.9711882650379233e-05,
+      "loss": 9.2638,
+      "step": 69000
+    },
+    {
+      "epoch": 0.3446278309071388,
+      "grad_norm": 0.09350882470607758,
+      "learning_rate": 1.9710380735437683e-05,
+      "loss": 9.2606,
+      "step": 69010
+    },
+    {
+      "epoch": 0.3446777697320782,
+      "grad_norm": 0.08809029310941696,
+      "learning_rate": 1.9708878820496134e-05,
+      "loss": 9.262,
+      "step": 69020
+    },
+    {
+      "epoch": 0.3447277085570177,
+      "grad_norm": 0.09742359071969986,
+      "learning_rate": 1.970737690555458e-05,
+      "loss": 9.2644,
+      "step": 69030
+    },
+    {
+      "epoch": 0.3447776473819571,
+      "grad_norm": 0.09343401342630386,
+      "learning_rate": 1.9705874990613034e-05,
+      "loss": 9.2674,
+      "step": 69040
+    },
+    {
+      "epoch": 0.3448275862068966,
+      "grad_norm": 0.09134664386510849,
+      "learning_rate": 1.970437307567148e-05,
+      "loss": 9.2566,
+      "step": 69050
+    },
+    {
+      "epoch": 0.344877525031836,
+      "grad_norm": 0.09671717882156372,
+      "learning_rate": 1.9702871160729934e-05,
+      "loss": 9.2612,
+      "step": 69060
+    },
+    {
+      "epoch": 0.3449274638567755,
+      "grad_norm": 0.09348636120557785,
+      "learning_rate": 1.970136924578838e-05,
+      "loss": 9.2665,
+      "step": 69070
+    },
+    {
+      "epoch": 0.3449774026817149,
+      "grad_norm": 0.0987660363316536,
+      "learning_rate": 1.9699867330846828e-05,
+      "loss": 9.2633,
+      "step": 69080
+    },
+    {
+      "epoch": 0.3450273415066544,
+      "grad_norm": 0.09916608035564423,
+      "learning_rate": 1.969836541590528e-05,
+      "loss": 9.2581,
+      "step": 69090
+    },
+    {
+      "epoch": 0.3450772803315938,
+      "grad_norm": 0.0879778116941452,
+      "learning_rate": 1.9696863500963728e-05,
+      "loss": 9.258,
+      "step": 69100
+    },
+    {
+      "epoch": 0.34512721915653327,
+      "grad_norm": 0.09248411655426025,
+      "learning_rate": 1.9695361586022182e-05,
+      "loss": 9.262,
+      "step": 69110
+    },
+    {
+      "epoch": 0.3451771579814727,
+      "grad_norm": 0.09013424068689346,
+      "learning_rate": 1.969385967108063e-05,
+      "loss": 9.2611,
+      "step": 69120
+    },
+    {
+      "epoch": 0.34522709680641217,
+      "grad_norm": 0.09179334342479706,
+      "learning_rate": 1.9692357756139075e-05,
+      "loss": 9.2494,
+      "step": 69130
+    },
+    {
+      "epoch": 0.3452770356313516,
+      "grad_norm": 0.09357766062021255,
+      "learning_rate": 1.969085584119753e-05,
+      "loss": 9.2625,
+      "step": 69140
+    },
+    {
+      "epoch": 0.34532697445629107,
+      "grad_norm": 0.09096767753362656,
+      "learning_rate": 1.9689353926255976e-05,
+      "loss": 9.2662,
+      "step": 69150
+    },
+    {
+      "epoch": 0.3453769132812305,
+      "grad_norm": 0.0954696536064148,
+      "learning_rate": 1.968785201131443e-05,
+      "loss": 9.2555,
+      "step": 69160
+    },
+    {
+      "epoch": 0.34542685210616997,
+      "grad_norm": 0.09414954483509064,
+      "learning_rate": 1.9686350096372876e-05,
+      "loss": 9.2556,
+      "step": 69170
+    },
+    {
+      "epoch": 0.3454767909311094,
+      "grad_norm": 0.09444237500429153,
+      "learning_rate": 1.9684848181431323e-05,
+      "loss": 9.2622,
+      "step": 69180
+    },
+    {
+      "epoch": 0.34552672975604887,
+      "grad_norm": 0.09886594861745834,
+      "learning_rate": 1.9683346266489776e-05,
+      "loss": 9.2541,
+      "step": 69190
+    },
+    {
+      "epoch": 0.3455766685809883,
+      "grad_norm": 0.09522037953138351,
+      "learning_rate": 1.9681844351548223e-05,
+      "loss": 9.2655,
+      "step": 69200
+    },
+    {
+      "epoch": 0.34562660740592777,
+      "grad_norm": 0.09733402729034424,
+      "learning_rate": 1.9680342436606677e-05,
+      "loss": 9.2584,
+      "step": 69210
+    },
+    {
+      "epoch": 0.3456765462308672,
+      "grad_norm": 0.09290558099746704,
+      "learning_rate": 1.9678840521665124e-05,
+      "loss": 9.2615,
+      "step": 69220
+    },
+    {
+      "epoch": 0.3457264850558066,
+      "grad_norm": 0.09087418764829636,
+      "learning_rate": 1.967733860672357e-05,
+      "loss": 9.2717,
+      "step": 69230
+    },
+    {
+      "epoch": 0.3457764238807461,
+      "grad_norm": 0.09687705338001251,
+      "learning_rate": 1.9675836691782024e-05,
+      "loss": 9.2599,
+      "step": 69240
+    },
+    {
+      "epoch": 0.3458263627056855,
+      "grad_norm": 0.09234501421451569,
+      "learning_rate": 1.967433477684047e-05,
+      "loss": 9.2586,
+      "step": 69250
+    },
+    {
+      "epoch": 0.345876301530625,
+      "grad_norm": 0.09087526053190231,
+      "learning_rate": 1.9672832861898924e-05,
+      "loss": 9.2582,
+      "step": 69260
+    },
+    {
+      "epoch": 0.3459262403555644,
+      "grad_norm": 0.0971473827958107,
+      "learning_rate": 1.967133094695737e-05,
+      "loss": 9.2627,
+      "step": 69270
+    },
+    {
+      "epoch": 0.3459761791805039,
+      "grad_norm": 0.0920579731464386,
+      "learning_rate": 1.9669829032015818e-05,
+      "loss": 9.2558,
+      "step": 69280
+    },
+    {
+      "epoch": 0.3460261180054433,
+      "grad_norm": 0.09076216071844101,
+      "learning_rate": 1.966832711707427e-05,
+      "loss": 9.2622,
+      "step": 69290
+    },
+    {
+      "epoch": 0.3460760568303828,
+      "grad_norm": 0.09444985538721085,
+      "learning_rate": 1.9666825202132718e-05,
+      "loss": 9.2538,
+      "step": 69300
+    },
+    {
+      "epoch": 0.3461259956553222,
+      "grad_norm": 0.08982237428426743,
+      "learning_rate": 1.9665323287191172e-05,
+      "loss": 9.257,
+      "step": 69310
+    },
+    {
+      "epoch": 0.3461759344802617,
+      "grad_norm": 0.09677401930093765,
+      "learning_rate": 1.966382137224962e-05,
+      "loss": 9.2609,
+      "step": 69320
+    },
+    {
+      "epoch": 0.3462258733052011,
+      "grad_norm": 0.09712191671133041,
+      "learning_rate": 1.9662319457308065e-05,
+      "loss": 9.2516,
+      "step": 69330
+    },
+    {
+      "epoch": 0.3462758121301406,
+      "grad_norm": 0.09230844676494598,
+      "learning_rate": 1.966081754236652e-05,
+      "loss": 9.2593,
+      "step": 69340
+    },
+    {
+      "epoch": 0.34632575095508,
+      "grad_norm": 0.09202716499567032,
+      "learning_rate": 1.9659315627424966e-05,
+      "loss": 9.2567,
+      "step": 69350
+    },
+    {
+      "epoch": 0.3463756897800195,
+      "grad_norm": 0.09389092773199081,
+      "learning_rate": 1.965781371248342e-05,
+      "loss": 9.2622,
+      "step": 69360
+    },
+    {
+      "epoch": 0.3464256286049589,
+      "grad_norm": 0.09576646238565445,
+      "learning_rate": 1.9656311797541866e-05,
+      "loss": 9.267,
+      "step": 69370
+    },
+    {
+      "epoch": 0.3464755674298984,
+      "grad_norm": 0.09165678918361664,
+      "learning_rate": 1.9654809882600316e-05,
+      "loss": 9.2661,
+      "step": 69380
+    },
+    {
+      "epoch": 0.3465255062548378,
+      "grad_norm": 0.09849521517753601,
+      "learning_rate": 1.9653307967658766e-05,
+      "loss": 9.2615,
+      "step": 69390
+    },
+    {
+      "epoch": 0.3465754450797773,
+      "grad_norm": 0.09328168630599976,
+      "learning_rate": 1.9651806052717213e-05,
+      "loss": 9.2594,
+      "step": 69400
+    },
+    {
+      "epoch": 0.3466253839047167,
+      "grad_norm": 0.09389086067676544,
+      "learning_rate": 1.9650304137775667e-05,
+      "loss": 9.2545,
+      "step": 69410
+    },
+    {
+      "epoch": 0.3466753227296562,
+      "grad_norm": 0.09465324133634567,
+      "learning_rate": 1.9648802222834114e-05,
+      "loss": 9.2589,
+      "step": 69420
+    },
+    {
+      "epoch": 0.3467252615545956,
+      "grad_norm": 0.09655381739139557,
+      "learning_rate": 1.9647300307892564e-05,
+      "loss": 9.2549,
+      "step": 69430
+    },
+    {
+      "epoch": 0.3467752003795351,
+      "grad_norm": 0.09776412695646286,
+      "learning_rate": 1.9645798392951014e-05,
+      "loss": 9.2511,
+      "step": 69440
+    },
+    {
+      "epoch": 0.3468251392044745,
+      "grad_norm": 0.09319782257080078,
+      "learning_rate": 1.964429647800946e-05,
+      "loss": 9.2605,
+      "step": 69450
+    },
+    {
+      "epoch": 0.346875078029414,
+      "grad_norm": 0.096680648624897,
+      "learning_rate": 1.9642794563067914e-05,
+      "loss": 9.2572,
+      "step": 69460
+    },
+    {
+      "epoch": 0.3469250168543534,
+      "grad_norm": 0.0916510671377182,
+      "learning_rate": 1.964129264812636e-05,
+      "loss": 9.2463,
+      "step": 69470
+    },
+    {
+      "epoch": 0.3469749556792929,
+      "grad_norm": 0.09772134572267532,
+      "learning_rate": 1.963979073318481e-05,
+      "loss": 9.2529,
+      "step": 69480
+    },
+    {
+      "epoch": 0.3470248945042323,
+      "grad_norm": 0.0930207371711731,
+      "learning_rate": 1.963828881824326e-05,
+      "loss": 9.2567,
+      "step": 69490
+    },
+    {
+      "epoch": 0.3470748333291718,
+      "grad_norm": 0.09252675622701645,
+      "learning_rate": 1.9636786903301708e-05,
+      "loss": 9.2588,
+      "step": 69500
+    },
+    {
+      "epoch": 0.3471247721541112,
+      "grad_norm": 0.09338805079460144,
+      "learning_rate": 1.9635284988360162e-05,
+      "loss": 9.2576,
+      "step": 69510
+    },
+    {
+      "epoch": 0.34717471097905067,
+      "grad_norm": 0.09176600724458694,
+      "learning_rate": 1.963378307341861e-05,
+      "loss": 18.6508,
+      "step": 69520
+    },
+    {
+      "epoch": 0.3472246498039901,
+      "grad_norm": 0.09212823212146759,
+      "learning_rate": 1.963228115847706e-05,
+      "loss": 15.6648,
+      "step": 69530
+    },
+    {
+      "epoch": 0.34727458862892957,
+      "grad_norm": 0.092359758913517,
+      "learning_rate": 1.963077924353551e-05,
+      "loss": 9.2493,
+      "step": 69540
+    },
+    {
+      "epoch": 0.347324527453869,
+      "grad_norm": 0.08757336437702179,
+      "learning_rate": 1.9629277328593956e-05,
+      "loss": 9.2656,
+      "step": 69550
+    },
+    {
+      "epoch": 0.34737446627880847,
+      "grad_norm": 0.08831485360860825,
+      "learning_rate": 1.962777541365241e-05,
+      "loss": 9.2628,
+      "step": 69560
+    },
+    {
+      "epoch": 0.3474244051037479,
+      "grad_norm": 0.09570340067148209,
+      "learning_rate": 1.9626273498710856e-05,
+      "loss": 9.2617,
+      "step": 69570
+    },
+    {
+      "epoch": 0.34747434392868737,
+      "grad_norm": 0.09006953239440918,
+      "learning_rate": 1.9624771583769306e-05,
+      "loss": 9.2511,
+      "step": 69580
+    },
+    {
+      "epoch": 0.3475242827536268,
+      "grad_norm": 0.09478635340929031,
+      "learning_rate": 1.9623269668827756e-05,
+      "loss": 9.268,
+      "step": 69590
+    },
+    {
+      "epoch": 0.34757422157856627,
+      "grad_norm": 0.09337896853685379,
+      "learning_rate": 1.9621767753886203e-05,
+      "loss": 9.2629,
+      "step": 69600
+    },
+    {
+      "epoch": 0.3476241604035057,
+      "grad_norm": 0.09267710894346237,
+      "learning_rate": 1.9620265838944657e-05,
+      "loss": 9.2478,
+      "step": 69610
+    },
+    {
+      "epoch": 0.34767409922844517,
+      "grad_norm": 0.09669796377420425,
+      "learning_rate": 1.9618763924003104e-05,
+      "loss": 9.2465,
+      "step": 69620
+    },
+    {
+      "epoch": 0.3477240380533846,
+      "grad_norm": 0.09430265426635742,
+      "learning_rate": 1.9617262009061554e-05,
+      "loss": 9.2617,
+      "step": 69630
+    },
+    {
+      "epoch": 0.34777397687832406,
+      "grad_norm": 0.09212590754032135,
+      "learning_rate": 1.9615760094120004e-05,
+      "loss": 9.2578,
+      "step": 69640
+    },
+    {
+      "epoch": 0.3478239157032635,
+      "grad_norm": 0.09378217160701752,
+      "learning_rate": 1.961425817917845e-05,
+      "loss": 9.2584,
+      "step": 69650
+    },
+    {
+      "epoch": 0.34787385452820296,
+      "grad_norm": 0.09014896303415298,
+      "learning_rate": 1.9612756264236904e-05,
+      "loss": 9.2596,
+      "step": 69660
+    },
+    {
+      "epoch": 0.3479237933531424,
+      "grad_norm": 0.0964147299528122,
+      "learning_rate": 1.961125434929535e-05,
+      "loss": 9.2503,
+      "step": 69670
+    },
+    {
+      "epoch": 0.34797373217808186,
+      "grad_norm": 0.09427673369646072,
+      "learning_rate": 1.96097524343538e-05,
+      "loss": 9.2688,
+      "step": 69680
+    },
+    {
+      "epoch": 0.3480236710030213,
+      "grad_norm": 0.09578701853752136,
+      "learning_rate": 1.960825051941225e-05,
+      "loss": 9.2582,
+      "step": 69690
+    },
+    {
+      "epoch": 0.34807360982796076,
+      "grad_norm": 0.09862515330314636,
+      "learning_rate": 1.96067486044707e-05,
+      "loss": 9.2589,
+      "step": 69700
+    },
+    {
+      "epoch": 0.3481235486529002,
+      "grad_norm": 0.09679950028657913,
+      "learning_rate": 1.9605246689529152e-05,
+      "loss": 9.2639,
+      "step": 69710
+    },
+    {
+      "epoch": 0.34817348747783966,
+      "grad_norm": 0.09329384565353394,
+      "learning_rate": 1.96037447745876e-05,
+      "loss": 9.2534,
+      "step": 69720
+    },
+    {
+      "epoch": 0.3482234263027791,
+      "grad_norm": 0.09050307422876358,
+      "learning_rate": 1.960224285964605e-05,
+      "loss": 9.258,
+      "step": 69730
+    },
+    {
+      "epoch": 0.34827336512771856,
+      "grad_norm": 0.08931013941764832,
+      "learning_rate": 1.96007409447045e-05,
+      "loss": 9.2506,
+      "step": 69740
+    },
+    {
+      "epoch": 0.348323303952658,
+      "grad_norm": 0.09590967744588852,
+      "learning_rate": 1.959923902976295e-05,
+      "loss": 9.2381,
+      "step": 69750
+    },
+    {
+      "epoch": 0.34837324277759746,
+      "grad_norm": 0.09243065118789673,
+      "learning_rate": 1.95977371148214e-05,
+      "loss": 9.2571,
+      "step": 69760
+    },
+    {
+      "epoch": 0.3484231816025369,
+      "grad_norm": 0.10139338672161102,
+      "learning_rate": 1.9596235199879846e-05,
+      "loss": 9.2542,
+      "step": 69770
+    },
+    {
+      "epoch": 0.34847312042747636,
+      "grad_norm": 0.09506342560052872,
+      "learning_rate": 1.9594733284938296e-05,
+      "loss": 9.2574,
+      "step": 69780
+    },
+    {
+      "epoch": 0.3485230592524158,
+      "grad_norm": 0.09532468020915985,
+      "learning_rate": 1.9593231369996746e-05,
+      "loss": 9.2527,
+      "step": 69790
+    },
+    {
+      "epoch": 0.34857299807735526,
+      "grad_norm": 0.09643257409334183,
+      "learning_rate": 1.9591729455055197e-05,
+      "loss": 9.2545,
+      "step": 69800
+    },
+    {
+      "epoch": 0.3486229369022947,
+      "grad_norm": 0.09859835356473923,
+      "learning_rate": 1.9590227540113647e-05,
+      "loss": 9.2565,
+      "step": 69810
+    },
+    {
+      "epoch": 0.34867287572723416,
+      "grad_norm": 0.09410768747329712,
+      "learning_rate": 1.9588725625172094e-05,
+      "loss": 9.2615,
+      "step": 69820
+    },
+    {
+      "epoch": 0.3487228145521736,
+      "grad_norm": 0.0925586149096489,
+      "learning_rate": 1.9587223710230544e-05,
+      "loss": 9.2492,
+      "step": 69830
+    },
+    {
+      "epoch": 0.34877275337711305,
+      "grad_norm": 0.09222124516963959,
+      "learning_rate": 1.9585721795288994e-05,
+      "loss": 9.2527,
+      "step": 69840
+    },
+    {
+      "epoch": 0.3488226922020525,
+      "grad_norm": 0.08799050003290176,
+      "learning_rate": 1.9584219880347444e-05,
+      "loss": 9.2572,
+      "step": 69850
+    },
+    {
+      "epoch": 0.34887263102699195,
+      "grad_norm": 0.09063519537448883,
+      "learning_rate": 1.9582717965405894e-05,
+      "loss": 9.2635,
+      "step": 69860
+    },
+    {
+      "epoch": 0.3489225698519314,
+      "grad_norm": 0.09138821065425873,
+      "learning_rate": 1.958121605046434e-05,
+      "loss": 9.2414,
+      "step": 69870
+    },
+    {
+      "epoch": 0.34897250867687085,
+      "grad_norm": 0.0935179740190506,
+      "learning_rate": 1.957971413552279e-05,
+      "loss": 9.2509,
+      "step": 69880
+    },
+    {
+      "epoch": 0.3490224475018103,
+      "grad_norm": 0.09351733326911926,
+      "learning_rate": 1.957821222058124e-05,
+      "loss": 9.2526,
+      "step": 69890
+    },
+    {
+      "epoch": 0.34907238632674975,
+      "grad_norm": 0.09050756692886353,
+      "learning_rate": 1.957671030563969e-05,
+      "loss": 9.2616,
+      "step": 69900
+    },
+    {
+      "epoch": 0.3491223251516892,
+      "grad_norm": 0.09149663895368576,
+      "learning_rate": 1.9575208390698142e-05,
+      "loss": 9.2517,
+      "step": 69910
+    },
+    {
+      "epoch": 0.34917226397662865,
+      "grad_norm": 0.09362272918224335,
+      "learning_rate": 1.957370647575659e-05,
+      "loss": 9.2577,
+      "step": 69920
+    },
+    {
+      "epoch": 0.34922220280156807,
+      "grad_norm": 0.09068772941827774,
+      "learning_rate": 1.957220456081504e-05,
+      "loss": 9.2577,
+      "step": 69930
+    },
+    {
+      "epoch": 0.34927214162650755,
+      "grad_norm": 0.08827726542949677,
+      "learning_rate": 1.957070264587349e-05,
+      "loss": 9.2646,
+      "step": 69940
+    },
+    {
+      "epoch": 0.34932208045144697,
+      "grad_norm": 0.09964391589164734,
+      "learning_rate": 1.956920073093194e-05,
+      "loss": 9.2494,
+      "step": 69950
+    },
+    {
+      "epoch": 0.34937201927638645,
+      "grad_norm": 0.09400717914104462,
+      "learning_rate": 1.956769881599039e-05,
+      "loss": 9.2437,
+      "step": 69960
+    },
+    {
+      "epoch": 0.34942195810132587,
+      "grad_norm": 0.0930166095495224,
+      "learning_rate": 1.9566196901048836e-05,
+      "loss": 9.2528,
+      "step": 69970
+    },
+    {
+      "epoch": 0.34947189692626535,
+      "grad_norm": 0.09116925299167633,
+      "learning_rate": 1.9564694986107286e-05,
+      "loss": 9.2635,
+      "step": 69980
+    },
+    {
+      "epoch": 0.34952183575120477,
+      "grad_norm": 0.09678564965724945,
+      "learning_rate": 1.9563193071165737e-05,
+      "loss": 9.2448,
+      "step": 69990
+    },
+    {
+      "epoch": 0.34957177457614425,
+      "grad_norm": 0.09541235119104385,
+      "learning_rate": 1.9561691156224187e-05,
+      "loss": 9.2439,
+      "step": 70000
+    },
+    {
+      "epoch": 0.34962171340108367,
+      "grad_norm": 0.09007764607667923,
+      "learning_rate": 1.9560189241282637e-05,
+      "loss": 9.2601,
+      "step": 70010
+    },
+    {
+      "epoch": 0.34967165222602314,
+      "grad_norm": 0.09365357458591461,
+      "learning_rate": 1.9558687326341087e-05,
+      "loss": 9.2501,
+      "step": 70020
+    },
+    {
+      "epoch": 0.34972159105096257,
+      "grad_norm": 0.09102270007133484,
+      "learning_rate": 1.9557185411399534e-05,
+      "loss": 9.2509,
+      "step": 70030
+    },
+    {
+      "epoch": 0.34977152987590204,
+      "grad_norm": 0.09124330431222916,
+      "learning_rate": 1.9555683496457984e-05,
+      "loss": 9.2493,
+      "step": 70040
+    },
+    {
+      "epoch": 0.34982146870084146,
+      "grad_norm": 0.09593762457370758,
+      "learning_rate": 1.9554181581516434e-05,
+      "loss": 9.2587,
+      "step": 70050
+    },
+    {
+      "epoch": 0.34987140752578094,
+      "grad_norm": 0.09397412836551666,
+      "learning_rate": 1.9552679666574884e-05,
+      "loss": 9.2518,
+      "step": 70060
+    },
+    {
+      "epoch": 0.34992134635072036,
+      "grad_norm": 0.09382015466690063,
+      "learning_rate": 1.9551177751633335e-05,
+      "loss": 9.2479,
+      "step": 70070
+    },
+    {
+      "epoch": 0.34997128517565984,
+      "grad_norm": 0.09169856458902359,
+      "learning_rate": 1.954967583669178e-05,
+      "loss": 9.2492,
+      "step": 70080
+    },
+    {
+      "epoch": 0.35002122400059926,
+      "grad_norm": 0.09524683654308319,
+      "learning_rate": 1.954817392175023e-05,
+      "loss": 9.2446,
+      "step": 70090
+    },
+    {
+      "epoch": 0.35007116282553874,
+      "grad_norm": 0.09151513129472733,
+      "learning_rate": 1.9546672006808682e-05,
+      "loss": 9.26,
+      "step": 70100
+    },
+    {
+      "epoch": 0.35012110165047816,
+      "grad_norm": 0.09831450879573822,
+      "learning_rate": 1.9545170091867132e-05,
+      "loss": 9.2467,
+      "step": 70110
+    },
+    {
+      "epoch": 0.35017104047541764,
+      "grad_norm": 0.08933068811893463,
+      "learning_rate": 1.9543668176925582e-05,
+      "loss": 9.2457,
+      "step": 70120
+    },
+    {
+      "epoch": 0.35022097930035706,
+      "grad_norm": 0.09220155328512192,
+      "learning_rate": 1.954216626198403e-05,
+      "loss": 9.2452,
+      "step": 70130
+    },
+    {
+      "epoch": 0.35027091812529654,
+      "grad_norm": 0.09587565809488297,
+      "learning_rate": 1.954066434704248e-05,
+      "loss": 9.2367,
+      "step": 70140
+    },
+    {
+      "epoch": 0.35032085695023596,
+      "grad_norm": 0.09450395405292511,
+      "learning_rate": 1.953916243210093e-05,
+      "loss": 9.2412,
+      "step": 70150
+    },
+    {
+      "epoch": 0.35037079577517544,
+      "grad_norm": 0.0945536196231842,
+      "learning_rate": 1.953766051715938e-05,
+      "loss": 9.2447,
+      "step": 70160
+    },
+    {
+      "epoch": 0.35042073460011486,
+      "grad_norm": 0.09448517858982086,
+      "learning_rate": 1.953615860221783e-05,
+      "loss": 9.2549,
+      "step": 70170
+    },
+    {
+      "epoch": 0.35047067342505434,
+      "grad_norm": 0.09346494823694229,
+      "learning_rate": 1.9534656687276276e-05,
+      "loss": 9.2633,
+      "step": 70180
+    },
+    {
+      "epoch": 0.35052061224999376,
+      "grad_norm": 0.0924750491976738,
+      "learning_rate": 1.9533154772334727e-05,
+      "loss": 9.243,
+      "step": 70190
+    },
+    {
+      "epoch": 0.35057055107493323,
+      "grad_norm": 0.09105124324560165,
+      "learning_rate": 1.9531652857393177e-05,
+      "loss": 9.2572,
+      "step": 70200
+    },
+    {
+      "epoch": 0.35062048989987266,
+      "grad_norm": 0.09556052833795547,
+      "learning_rate": 1.9530150942451627e-05,
+      "loss": 9.2382,
+      "step": 70210
+    },
+    {
+      "epoch": 0.3506704287248121,
+      "grad_norm": 0.09211760014295578,
+      "learning_rate": 1.9528649027510077e-05,
+      "loss": 9.2363,
+      "step": 70220
+    },
+    {
+      "epoch": 0.35072036754975155,
+      "grad_norm": 0.09628025442361832,
+      "learning_rate": 1.9527147112568524e-05,
+      "loss": 9.249,
+      "step": 70230
+    },
+    {
+      "epoch": 0.350770306374691,
+      "grad_norm": 0.09602992981672287,
+      "learning_rate": 1.9525645197626974e-05,
+      "loss": 9.2451,
+      "step": 70240
+    },
+    {
+      "epoch": 0.35082024519963045,
+      "grad_norm": 0.09380794316530228,
+      "learning_rate": 1.9524143282685424e-05,
+      "loss": 9.2573,
+      "step": 70250
+    },
+    {
+      "epoch": 0.3508701840245699,
+      "grad_norm": 0.09438929706811905,
+      "learning_rate": 1.9522641367743874e-05,
+      "loss": 9.2478,
+      "step": 70260
+    },
+    {
+      "epoch": 0.35092012284950935,
+      "grad_norm": 0.09579429030418396,
+      "learning_rate": 1.9521139452802325e-05,
+      "loss": 9.2498,
+      "step": 70270
+    },
+    {
+      "epoch": 0.3509700616744488,
+      "grad_norm": 0.0962810218334198,
+      "learning_rate": 1.951963753786077e-05,
+      "loss": 9.2395,
+      "step": 70280
+    },
+    {
+      "epoch": 0.35102000049938825,
+      "grad_norm": 0.0956900343298912,
+      "learning_rate": 1.951813562291922e-05,
+      "loss": 9.2403,
+      "step": 70290
+    },
+    {
+      "epoch": 0.3510699393243277,
+      "grad_norm": 0.09246978163719177,
+      "learning_rate": 1.9516633707977672e-05,
+      "loss": 9.2375,
+      "step": 70300
+    },
+    {
+      "epoch": 0.35111987814926715,
+      "grad_norm": 0.09685346484184265,
+      "learning_rate": 1.9515131793036122e-05,
+      "loss": 9.2514,
+      "step": 70310
+    },
+    {
+      "epoch": 0.3511698169742066,
+      "grad_norm": 0.09638645499944687,
+      "learning_rate": 1.9513629878094572e-05,
+      "loss": 9.2439,
+      "step": 70320
+    },
+    {
+      "epoch": 0.35121975579914605,
+      "grad_norm": 0.09618096798658371,
+      "learning_rate": 1.951212796315302e-05,
+      "loss": 9.2504,
+      "step": 70330
+    },
+    {
+      "epoch": 0.35126969462408547,
+      "grad_norm": 0.09635983407497406,
+      "learning_rate": 1.9510626048211472e-05,
+      "loss": 9.2449,
+      "step": 70340
+    },
+    {
+      "epoch": 0.35131963344902495,
+      "grad_norm": 0.09640626609325409,
+      "learning_rate": 1.950912413326992e-05,
+      "loss": 9.2455,
+      "step": 70350
+    },
+    {
+      "epoch": 0.35136957227396437,
+      "grad_norm": 0.08988912403583527,
+      "learning_rate": 1.950762221832837e-05,
+      "loss": 9.2467,
+      "step": 70360
+    },
+    {
+      "epoch": 0.35141951109890385,
+      "grad_norm": 0.09091424196958542,
+      "learning_rate": 1.950612030338682e-05,
+      "loss": 9.2542,
+      "step": 70370
+    },
+    {
+      "epoch": 0.35146944992384327,
+      "grad_norm": 0.09256154298782349,
+      "learning_rate": 1.9504618388445266e-05,
+      "loss": 9.2376,
+      "step": 70380
+    },
+    {
+      "epoch": 0.35151938874878275,
+      "grad_norm": 0.09144950658082962,
+      "learning_rate": 1.950311647350372e-05,
+      "loss": 9.2372,
+      "step": 70390
+    },
+    {
+      "epoch": 0.35156932757372217,
+      "grad_norm": 0.09599640965461731,
+      "learning_rate": 1.9501614558562167e-05,
+      "loss": 9.2434,
+      "step": 70400
+    },
+    {
+      "epoch": 0.35161926639866165,
+      "grad_norm": 0.0928352028131485,
+      "learning_rate": 1.9500112643620617e-05,
+      "loss": 9.253,
+      "step": 70410
+    },
+    {
+      "epoch": 0.35166920522360107,
+      "grad_norm": 0.09690866619348526,
+      "learning_rate": 1.9498610728679067e-05,
+      "loss": 9.2466,
+      "step": 70420
+    },
+    {
+      "epoch": 0.35171914404854054,
+      "grad_norm": 0.09045714139938354,
+      "learning_rate": 1.9497108813737514e-05,
+      "loss": 9.2495,
+      "step": 70430
+    },
+    {
+      "epoch": 0.35176908287347997,
+      "grad_norm": 0.09313403815031052,
+      "learning_rate": 1.9495606898795967e-05,
+      "loss": 9.2456,
+      "step": 70440
+    },
+    {
+      "epoch": 0.35181902169841944,
+      "grad_norm": 0.0966280922293663,
+      "learning_rate": 1.9494104983854414e-05,
+      "loss": 9.2418,
+      "step": 70450
+    },
+    {
+      "epoch": 0.35186896052335886,
+      "grad_norm": 0.09932943433523178,
+      "learning_rate": 1.9492603068912864e-05,
+      "loss": 9.24,
+      "step": 70460
+    },
+    {
+      "epoch": 0.35191889934829834,
+      "grad_norm": 0.10107915848493576,
+      "learning_rate": 1.9491101153971315e-05,
+      "loss": 9.2464,
+      "step": 70470
+    },
+    {
+      "epoch": 0.35196883817323776,
+      "grad_norm": 0.09643235802650452,
+      "learning_rate": 1.948959923902976e-05,
+      "loss": 9.2393,
+      "step": 70480
+    },
+    {
+      "epoch": 0.35201877699817724,
+      "grad_norm": 0.09176507592201233,
+      "learning_rate": 1.9488097324088215e-05,
+      "loss": 9.2509,
+      "step": 70490
+    },
+    {
+      "epoch": 0.35206871582311666,
+      "grad_norm": 0.09726466983556747,
+      "learning_rate": 1.9486595409146662e-05,
+      "loss": 9.235,
+      "step": 70500
+    },
+    {
+      "epoch": 0.35211865464805614,
+      "grad_norm": 0.09814643114805222,
+      "learning_rate": 1.9485093494205112e-05,
+      "loss": 9.2476,
+      "step": 70510
+    },
+    {
+      "epoch": 0.35216859347299556,
+      "grad_norm": 0.09261635690927505,
+      "learning_rate": 1.9483591579263562e-05,
+      "loss": 9.2453,
+      "step": 70520
+    },
+    {
+      "epoch": 0.35221853229793504,
+      "grad_norm": 0.09454991668462753,
+      "learning_rate": 1.948208966432201e-05,
+      "loss": 9.2552,
+      "step": 70530
+    },
+    {
+      "epoch": 0.35226847112287446,
+      "grad_norm": 0.09696483612060547,
+      "learning_rate": 1.9480587749380462e-05,
+      "loss": 9.2484,
+      "step": 70540
+    },
+    {
+      "epoch": 0.35231840994781394,
+      "grad_norm": 0.09152613580226898,
+      "learning_rate": 1.947908583443891e-05,
+      "loss": 9.249,
+      "step": 70550
+    },
+    {
+      "epoch": 0.35236834877275336,
+      "grad_norm": 0.09792204946279526,
+      "learning_rate": 1.947758391949736e-05,
+      "loss": 9.2417,
+      "step": 70560
+    },
+    {
+      "epoch": 0.35241828759769284,
+      "grad_norm": 0.09588468819856644,
+      "learning_rate": 1.947608200455581e-05,
+      "loss": 9.2381,
+      "step": 70570
+    },
+    {
+      "epoch": 0.35246822642263226,
+      "grad_norm": 0.09020346403121948,
+      "learning_rate": 1.9474580089614256e-05,
+      "loss": 9.2366,
+      "step": 70580
+    },
+    {
+      "epoch": 0.35251816524757174,
+      "grad_norm": 0.08940046280622482,
+      "learning_rate": 1.947307817467271e-05,
+      "loss": 9.2436,
+      "step": 70590
+    },
+    {
+      "epoch": 0.35256810407251116,
+      "grad_norm": 0.09213332086801529,
+      "learning_rate": 1.9471576259731157e-05,
+      "loss": 9.2367,
+      "step": 70600
+    },
+    {
+      "epoch": 0.35261804289745063,
+      "grad_norm": 0.09027668833732605,
+      "learning_rate": 1.9470074344789607e-05,
+      "loss": 9.242,
+      "step": 70610
+    },
+    {
+      "epoch": 0.35266798172239006,
+      "grad_norm": 0.0947854295372963,
+      "learning_rate": 1.9468572429848057e-05,
+      "loss": 9.2426,
+      "step": 70620
+    },
+    {
+      "epoch": 0.35271792054732953,
+      "grad_norm": 0.09516201168298721,
+      "learning_rate": 1.9467070514906504e-05,
+      "loss": 9.2549,
+      "step": 70630
+    },
+    {
+      "epoch": 0.35276785937226895,
+      "grad_norm": 0.09099344164133072,
+      "learning_rate": 1.9465568599964957e-05,
+      "loss": 9.2451,
+      "step": 70640
+    },
+    {
+      "epoch": 0.35281779819720843,
+      "grad_norm": 0.09635750204324722,
+      "learning_rate": 1.9464066685023404e-05,
+      "loss": 9.2471,
+      "step": 70650
+    },
+    {
+      "epoch": 0.35286773702214785,
+      "grad_norm": 0.0907939150929451,
+      "learning_rate": 1.9462564770081854e-05,
+      "loss": 9.2428,
+      "step": 70660
+    },
+    {
+      "epoch": 0.35291767584708733,
+      "grad_norm": 0.0959140956401825,
+      "learning_rate": 1.9461062855140305e-05,
+      "loss": 9.244,
+      "step": 70670
+    },
+    {
+      "epoch": 0.35296761467202675,
+      "grad_norm": 0.09400851279497147,
+      "learning_rate": 1.945956094019875e-05,
+      "loss": 9.2519,
+      "step": 70680
+    },
+    {
+      "epoch": 0.35301755349696623,
+      "grad_norm": 0.09195028990507126,
+      "learning_rate": 1.9458059025257205e-05,
+      "loss": 9.244,
+      "step": 70690
+    },
+    {
+      "epoch": 0.35306749232190565,
+      "grad_norm": 0.09665300697088242,
+      "learning_rate": 1.9456557110315652e-05,
+      "loss": 9.2462,
+      "step": 70700
+    },
+    {
+      "epoch": 0.35311743114684513,
+      "grad_norm": 0.09066880494356155,
+      "learning_rate": 1.9455055195374105e-05,
+      "loss": 9.2491,
+      "step": 70710
+    },
+    {
+      "epoch": 0.35316736997178455,
+      "grad_norm": 0.0936349481344223,
+      "learning_rate": 1.9453553280432552e-05,
+      "loss": 9.2481,
+      "step": 70720
+    },
+    {
+      "epoch": 0.353217308796724,
+      "grad_norm": 0.0960550457239151,
+      "learning_rate": 1.9452051365491e-05,
+      "loss": 9.2365,
+      "step": 70730
+    },
+    {
+      "epoch": 0.35326724762166345,
+      "grad_norm": 0.09481688588857651,
+      "learning_rate": 1.9450549450549452e-05,
+      "loss": 9.2371,
+      "step": 70740
+    },
+    {
+      "epoch": 0.3533171864466029,
+      "grad_norm": 0.099585622549057,
+      "learning_rate": 1.94490475356079e-05,
+      "loss": 9.241,
+      "step": 70750
+    },
+    {
+      "epoch": 0.35336712527154235,
+      "grad_norm": 0.09653332084417343,
+      "learning_rate": 1.9447545620666353e-05,
+      "loss": 9.2421,
+      "step": 70760
+    },
+    {
+      "epoch": 0.3534170640964818,
+      "grad_norm": 0.09810932725667953,
+      "learning_rate": 1.94460437057248e-05,
+      "loss": 9.2449,
+      "step": 70770
+    },
+    {
+      "epoch": 0.35346700292142125,
+      "grad_norm": 0.09474793076515198,
+      "learning_rate": 1.9444541790783246e-05,
+      "loss": 9.2413,
+      "step": 70780
+    },
+    {
+      "epoch": 0.3535169417463607,
+      "grad_norm": 0.09423346817493439,
+      "learning_rate": 1.94430398758417e-05,
+      "loss": 9.244,
+      "step": 70790
+    },
+    {
+      "epoch": 0.35356688057130015,
+      "grad_norm": 0.09000688046216965,
+      "learning_rate": 1.9441537960900147e-05,
+      "loss": 9.2306,
+      "step": 70800
+    },
+    {
+      "epoch": 0.3536168193962396,
+      "grad_norm": 0.09328895062208176,
+      "learning_rate": 1.94400360459586e-05,
+      "loss": 9.2454,
+      "step": 70810
+    },
+    {
+      "epoch": 0.35366675822117905,
+      "grad_norm": 0.09733468294143677,
+      "learning_rate": 1.9438534131017047e-05,
+      "loss": 9.2466,
+      "step": 70820
+    },
+    {
+      "epoch": 0.3537166970461185,
+      "grad_norm": 0.09681283682584763,
+      "learning_rate": 1.9437032216075494e-05,
+      "loss": 9.2315,
+      "step": 70830
+    },
+    {
+      "epoch": 0.35376663587105794,
+      "grad_norm": 0.09605410695075989,
+      "learning_rate": 1.9435530301133947e-05,
+      "loss": 9.2368,
+      "step": 70840
+    },
+    {
+      "epoch": 0.3538165746959974,
+      "grad_norm": 0.10018260776996613,
+      "learning_rate": 1.9434028386192394e-05,
+      "loss": 9.2381,
+      "step": 70850
+    },
+    {
+      "epoch": 0.35386651352093684,
+      "grad_norm": 0.09301795065402985,
+      "learning_rate": 1.9432526471250848e-05,
+      "loss": 9.2454,
+      "step": 70860
+    },
+    {
+      "epoch": 0.3539164523458763,
+      "grad_norm": 0.09586118161678314,
+      "learning_rate": 1.9431024556309295e-05,
+      "loss": 9.2435,
+      "step": 70870
+    },
+    {
+      "epoch": 0.35396639117081574,
+      "grad_norm": 0.09101895242929459,
+      "learning_rate": 1.942952264136774e-05,
+      "loss": 9.2297,
+      "step": 70880
+    },
+    {
+      "epoch": 0.3540163299957552,
+      "grad_norm": 0.08979874104261398,
+      "learning_rate": 1.9428020726426195e-05,
+      "loss": 9.2355,
+      "step": 70890
+    },
+    {
+      "epoch": 0.35406626882069464,
+      "grad_norm": 0.09098219126462936,
+      "learning_rate": 1.9426518811484642e-05,
+      "loss": 9.2347,
+      "step": 70900
+    },
+    {
+      "epoch": 0.3541162076456341,
+      "grad_norm": 0.09509828686714172,
+      "learning_rate": 1.9425016896543095e-05,
+      "loss": 9.2406,
+      "step": 70910
+    },
+    {
+      "epoch": 0.35416614647057354,
+      "grad_norm": 0.09496752172708511,
+      "learning_rate": 1.9423514981601542e-05,
+      "loss": 9.246,
+      "step": 70920
+    },
+    {
+      "epoch": 0.354216085295513,
+      "grad_norm": 0.09700390696525574,
+      "learning_rate": 1.942201306665999e-05,
+      "loss": 9.2376,
+      "step": 70930
+    },
+    {
+      "epoch": 0.35426602412045244,
+      "grad_norm": 0.09865306317806244,
+      "learning_rate": 1.9420511151718442e-05,
+      "loss": 9.2343,
+      "step": 70940
+    },
+    {
+      "epoch": 0.3543159629453919,
+      "grad_norm": 0.09420738369226456,
+      "learning_rate": 1.941900923677689e-05,
+      "loss": 9.2405,
+      "step": 70950
+    },
+    {
+      "epoch": 0.35436590177033134,
+      "grad_norm": 0.09997241944074631,
+      "learning_rate": 1.9417507321835343e-05,
+      "loss": 9.2398,
+      "step": 70960
+    },
+    {
+      "epoch": 0.3544158405952708,
+      "grad_norm": 0.09588761627674103,
+      "learning_rate": 1.941600540689379e-05,
+      "loss": 9.2359,
+      "step": 70970
+    },
+    {
+      "epoch": 0.35446577942021024,
+      "grad_norm": 0.09791746735572815,
+      "learning_rate": 1.9414503491952236e-05,
+      "loss": 9.2349,
+      "step": 70980
+    },
+    {
+      "epoch": 0.3545157182451497,
+      "grad_norm": 0.0913577750325203,
+      "learning_rate": 1.941300157701069e-05,
+      "loss": 9.238,
+      "step": 70990
+    },
+    {
+      "epoch": 0.35456565707008914,
+      "grad_norm": 0.08983467519283295,
+      "learning_rate": 1.9411499662069137e-05,
+      "loss": 9.2347,
+      "step": 71000
+    },
+    {
+      "epoch": 0.3546155958950286,
+      "grad_norm": 0.09971988946199417,
+      "learning_rate": 1.940999774712759e-05,
+      "loss": 9.2341,
+      "step": 71010
+    },
+    {
+      "epoch": 0.35466553471996803,
+      "grad_norm": 0.0975378155708313,
+      "learning_rate": 1.9408495832186037e-05,
+      "loss": 9.2304,
+      "step": 71020
+    },
+    {
+      "epoch": 0.3547154735449075,
+      "grad_norm": 0.09106051921844482,
+      "learning_rate": 1.9406993917244487e-05,
+      "loss": 9.2377,
+      "step": 71030
+    },
+    {
+      "epoch": 0.35476541236984693,
+      "grad_norm": 0.09189257025718689,
+      "learning_rate": 1.9405492002302937e-05,
+      "loss": 9.2449,
+      "step": 71040
+    },
+    {
+      "epoch": 0.3548153511947864,
+      "grad_norm": 0.09312552958726883,
+      "learning_rate": 1.9403990087361384e-05,
+      "loss": 9.2376,
+      "step": 71050
+    },
+    {
+      "epoch": 0.35486529001972583,
+      "grad_norm": 0.09094572812318802,
+      "learning_rate": 1.9402488172419838e-05,
+      "loss": 9.2444,
+      "step": 71060
+    },
+    {
+      "epoch": 0.3549152288446653,
+      "grad_norm": 0.09242424368858337,
+      "learning_rate": 1.9400986257478285e-05,
+      "loss": 9.2318,
+      "step": 71070
+    },
+    {
+      "epoch": 0.35496516766960473,
+      "grad_norm": 0.09304199367761612,
+      "learning_rate": 1.9399484342536735e-05,
+      "loss": 9.25,
+      "step": 71080
+    },
+    {
+      "epoch": 0.3550151064945442,
+      "grad_norm": 0.09726741909980774,
+      "learning_rate": 1.9397982427595185e-05,
+      "loss": 9.2376,
+      "step": 71090
+    },
+    {
+      "epoch": 0.35506504531948363,
+      "grad_norm": 0.09090828150510788,
+      "learning_rate": 1.9396480512653632e-05,
+      "loss": 9.2332,
+      "step": 71100
+    },
+    {
+      "epoch": 0.3551149841444231,
+      "grad_norm": 0.09156540036201477,
+      "learning_rate": 1.9394978597712085e-05,
+      "loss": 9.2448,
+      "step": 71110
+    },
+    {
+      "epoch": 0.35516492296936253,
+      "grad_norm": 0.0888853594660759,
+      "learning_rate": 1.9393476682770532e-05,
+      "loss": 9.2353,
+      "step": 71120
+    },
+    {
+      "epoch": 0.355214861794302,
+      "grad_norm": 0.09629367291927338,
+      "learning_rate": 1.9391974767828982e-05,
+      "loss": 9.2288,
+      "step": 71130
+    },
+    {
+      "epoch": 0.3552648006192414,
+      "grad_norm": 0.09092582762241364,
+      "learning_rate": 1.9390472852887432e-05,
+      "loss": 9.2368,
+      "step": 71140
+    },
+    {
+      "epoch": 0.3553147394441809,
+      "grad_norm": 0.09426487237215042,
+      "learning_rate": 1.938897093794588e-05,
+      "loss": 9.2381,
+      "step": 71150
+    },
+    {
+      "epoch": 0.3553646782691203,
+      "grad_norm": 0.09681005775928497,
+      "learning_rate": 1.9387469023004333e-05,
+      "loss": 9.2227,
+      "step": 71160
+    },
+    {
+      "epoch": 0.3554146170940598,
+      "grad_norm": 0.0895727351307869,
+      "learning_rate": 1.938596710806278e-05,
+      "loss": 9.2364,
+      "step": 71170
+    },
+    {
+      "epoch": 0.3554645559189992,
+      "grad_norm": 0.09309788048267365,
+      "learning_rate": 1.938446519312123e-05,
+      "loss": 9.2421,
+      "step": 71180
+    },
+    {
+      "epoch": 0.3555144947439387,
+      "grad_norm": 0.09737984091043472,
+      "learning_rate": 1.938296327817968e-05,
+      "loss": 9.236,
+      "step": 71190
+    },
+    {
+      "epoch": 0.3555644335688781,
+      "grad_norm": 0.09271812438964844,
+      "learning_rate": 1.9381461363238127e-05,
+      "loss": 9.23,
+      "step": 71200
+    },
+    {
+      "epoch": 0.35561437239381755,
+      "grad_norm": 0.09458890557289124,
+      "learning_rate": 1.937995944829658e-05,
+      "loss": 9.2267,
+      "step": 71210
+    },
+    {
+      "epoch": 0.355664311218757,
+      "grad_norm": 0.09335251152515411,
+      "learning_rate": 1.9378457533355027e-05,
+      "loss": 9.2298,
+      "step": 71220
+    },
+    {
+      "epoch": 0.35571425004369644,
+      "grad_norm": 0.09530318528413773,
+      "learning_rate": 1.9376955618413477e-05,
+      "loss": 9.2515,
+      "step": 71230
+    },
+    {
+      "epoch": 0.3557641888686359,
+      "grad_norm": 0.0916392132639885,
+      "learning_rate": 1.9375453703471927e-05,
+      "loss": 9.2341,
+      "step": 71240
+    },
+    {
+      "epoch": 0.35581412769357534,
+      "grad_norm": 0.09264536947011948,
+      "learning_rate": 1.9373951788530374e-05,
+      "loss": 9.2404,
+      "step": 71250
+    },
+    {
+      "epoch": 0.3558640665185148,
+      "grad_norm": 0.09547609835863113,
+      "learning_rate": 1.9372449873588828e-05,
+      "loss": 9.2349,
+      "step": 71260
+    },
+    {
+      "epoch": 0.35591400534345424,
+      "grad_norm": 0.099251389503479,
+      "learning_rate": 1.9370947958647275e-05,
+      "loss": 9.234,
+      "step": 71270
+    },
+    {
+      "epoch": 0.3559639441683937,
+      "grad_norm": 0.08727670460939407,
+      "learning_rate": 1.9369446043705725e-05,
+      "loss": 9.2337,
+      "step": 71280
+    },
+    {
+      "epoch": 0.35601388299333314,
+      "grad_norm": 0.09636933356523514,
+      "learning_rate": 1.9367944128764175e-05,
+      "loss": 9.2463,
+      "step": 71290
+    },
+    {
+      "epoch": 0.3560638218182726,
+      "grad_norm": 0.09374624490737915,
+      "learning_rate": 1.9366442213822622e-05,
+      "loss": 9.2331,
+      "step": 71300
+    },
+    {
+      "epoch": 0.35611376064321204,
+      "grad_norm": 0.08606693148612976,
+      "learning_rate": 1.9364940298881075e-05,
+      "loss": 9.22,
+      "step": 71310
+    },
+    {
+      "epoch": 0.3561636994681515,
+      "grad_norm": 0.09905658662319183,
+      "learning_rate": 1.9363438383939522e-05,
+      "loss": 9.2281,
+      "step": 71320
+    },
+    {
+      "epoch": 0.35621363829309094,
+      "grad_norm": 0.09433192014694214,
+      "learning_rate": 1.9361936468997972e-05,
+      "loss": 9.2424,
+      "step": 71330
+    },
+    {
+      "epoch": 0.3562635771180304,
+      "grad_norm": 0.09173821657896042,
+      "learning_rate": 1.9360434554056423e-05,
+      "loss": 9.2255,
+      "step": 71340
+    },
+    {
+      "epoch": 0.35631351594296984,
+      "grad_norm": 0.09594710916280746,
+      "learning_rate": 1.9358932639114873e-05,
+      "loss": 9.2267,
+      "step": 71350
+    },
+    {
+      "epoch": 0.3563634547679093,
+      "grad_norm": 0.09441184997558594,
+      "learning_rate": 1.9357430724173323e-05,
+      "loss": 9.2384,
+      "step": 71360
+    },
+    {
+      "epoch": 0.35641339359284874,
+      "grad_norm": 0.09391657263040543,
+      "learning_rate": 1.935592880923177e-05,
+      "loss": 9.2407,
+      "step": 71370
+    },
+    {
+      "epoch": 0.3564633324177882,
+      "grad_norm": 0.09173735976219177,
+      "learning_rate": 1.935442689429022e-05,
+      "loss": 9.2305,
+      "step": 71380
+    },
+    {
+      "epoch": 0.35651327124272764,
+      "grad_norm": 0.09266167879104614,
+      "learning_rate": 1.935292497934867e-05,
+      "loss": 9.2273,
+      "step": 71390
+    },
+    {
+      "epoch": 0.3565632100676671,
+      "grad_norm": 0.09733938425779343,
+      "learning_rate": 1.935142306440712e-05,
+      "loss": 9.222,
+      "step": 71400
+    },
+    {
+      "epoch": 0.35661314889260654,
+      "grad_norm": 0.09893626719713211,
+      "learning_rate": 1.934992114946557e-05,
+      "loss": 9.2267,
+      "step": 71410
+    },
+    {
+      "epoch": 0.356663087717546,
+      "grad_norm": 0.08972984552383423,
+      "learning_rate": 1.9348419234524017e-05,
+      "loss": 9.2272,
+      "step": 71420
+    },
+    {
+      "epoch": 0.35671302654248543,
+      "grad_norm": 0.09227081388235092,
+      "learning_rate": 1.9346917319582467e-05,
+      "loss": 9.2393,
+      "step": 71430
+    },
+    {
+      "epoch": 0.3567629653674249,
+      "grad_norm": 0.09158577024936676,
+      "learning_rate": 1.9345415404640918e-05,
+      "loss": 9.2323,
+      "step": 71440
+    },
+    {
+      "epoch": 0.35681290419236433,
+      "grad_norm": 0.09240981191396713,
+      "learning_rate": 1.9343913489699368e-05,
+      "loss": 9.2337,
+      "step": 71450
+    },
+    {
+      "epoch": 0.3568628430173038,
+      "grad_norm": 0.09477624297142029,
+      "learning_rate": 1.9342411574757818e-05,
+      "loss": 9.2299,
+      "step": 71460
+    },
+    {
+      "epoch": 0.35691278184224323,
+      "grad_norm": 0.0945139229297638,
+      "learning_rate": 1.9340909659816265e-05,
+      "loss": 9.2225,
+      "step": 71470
+    },
+    {
+      "epoch": 0.3569627206671827,
+      "grad_norm": 0.09509062021970749,
+      "learning_rate": 1.9339407744874715e-05,
+      "loss": 9.2309,
+      "step": 71480
+    },
+    {
+      "epoch": 0.35701265949212213,
+      "grad_norm": 0.08713449537754059,
+      "learning_rate": 1.9337905829933165e-05,
+      "loss": 9.2212,
+      "step": 71490
+    },
+    {
+      "epoch": 0.3570625983170616,
+      "grad_norm": 0.10225412994623184,
+      "learning_rate": 1.9336403914991615e-05,
+      "loss": 9.2294,
+      "step": 71500
+    },
+    {
+      "epoch": 0.35711253714200103,
+      "grad_norm": 0.09623237699270248,
+      "learning_rate": 1.9334902000050065e-05,
+      "loss": 9.2218,
+      "step": 71510
+    },
+    {
+      "epoch": 0.3571624759669405,
+      "grad_norm": 0.09265895932912827,
+      "learning_rate": 1.9333400085108512e-05,
+      "loss": 9.2148,
+      "step": 71520
+    },
+    {
+      "epoch": 0.35721241479187993,
+      "grad_norm": 0.09351124614477158,
+      "learning_rate": 1.9331898170166962e-05,
+      "loss": 9.2253,
+      "step": 71530
+    },
+    {
+      "epoch": 0.3572623536168194,
+      "grad_norm": 0.09479804337024689,
+      "learning_rate": 1.9330396255225413e-05,
+      "loss": 9.2182,
+      "step": 71540
+    },
+    {
+      "epoch": 0.3573122924417588,
+      "grad_norm": 0.09572158753871918,
+      "learning_rate": 1.9328894340283863e-05,
+      "loss": 9.2224,
+      "step": 71550
+    },
+    {
+      "epoch": 0.3573622312666983,
+      "grad_norm": 0.09059182554483414,
+      "learning_rate": 1.9327392425342313e-05,
+      "loss": 9.2344,
+      "step": 71560
+    },
+    {
+      "epoch": 0.3574121700916377,
+      "grad_norm": 0.09375757724046707,
+      "learning_rate": 1.932589051040076e-05,
+      "loss": 9.2307,
+      "step": 71570
+    },
+    {
+      "epoch": 0.3574621089165772,
+      "grad_norm": 0.08641654253005981,
+      "learning_rate": 1.932438859545921e-05,
+      "loss": 9.2391,
+      "step": 71580
+    },
+    {
+      "epoch": 0.3575120477415166,
+      "grad_norm": 0.09109997749328613,
+      "learning_rate": 1.932288668051766e-05,
+      "loss": 9.2417,
+      "step": 71590
+    },
+    {
+      "epoch": 0.3575619865664561,
+      "grad_norm": 0.1009616032242775,
+      "learning_rate": 1.932138476557611e-05,
+      "loss": 9.2308,
+      "step": 71600
+    },
+    {
+      "epoch": 0.3576119253913955,
+      "grad_norm": 0.09245188534259796,
+      "learning_rate": 1.931988285063456e-05,
+      "loss": 9.2251,
+      "step": 71610
+    },
+    {
+      "epoch": 0.357661864216335,
+      "grad_norm": 0.09447357058525085,
+      "learning_rate": 1.9318380935693007e-05,
+      "loss": 9.2347,
+      "step": 71620
+    },
+    {
+      "epoch": 0.3577118030412744,
+      "grad_norm": 0.0957716852426529,
+      "learning_rate": 1.9316879020751457e-05,
+      "loss": 9.2316,
+      "step": 71630
+    },
+    {
+      "epoch": 0.3577617418662139,
+      "grad_norm": 0.09259764105081558,
+      "learning_rate": 1.9315377105809908e-05,
+      "loss": 9.2325,
+      "step": 71640
+    },
+    {
+      "epoch": 0.3578116806911533,
+      "grad_norm": 0.09912045300006866,
+      "learning_rate": 1.9313875190868358e-05,
+      "loss": 9.2225,
+      "step": 71650
+    },
+    {
+      "epoch": 0.3578616195160928,
+      "grad_norm": 0.09714524447917938,
+      "learning_rate": 1.9312373275926808e-05,
+      "loss": 9.2267,
+      "step": 71660
+    },
+    {
+      "epoch": 0.3579115583410322,
+      "grad_norm": 0.08966837078332901,
+      "learning_rate": 1.9310871360985258e-05,
+      "loss": 9.2311,
+      "step": 71670
+    },
+    {
+      "epoch": 0.3579614971659717,
+      "grad_norm": 0.09709369391202927,
+      "learning_rate": 1.9309369446043705e-05,
+      "loss": 9.2262,
+      "step": 71680
+    },
+    {
+      "epoch": 0.3580114359909111,
+      "grad_norm": 0.0912432074546814,
+      "learning_rate": 1.9307867531102155e-05,
+      "loss": 9.2241,
+      "step": 71690
+    },
+    {
+      "epoch": 0.3580613748158506,
+      "grad_norm": 0.09726667404174805,
+      "learning_rate": 1.9306365616160605e-05,
+      "loss": 9.2349,
+      "step": 71700
+    },
+    {
+      "epoch": 0.35811131364079,
+      "grad_norm": 0.10004935413599014,
+      "learning_rate": 1.9304863701219055e-05,
+      "loss": 9.2206,
+      "step": 71710
+    },
+    {
+      "epoch": 0.3581612524657295,
+      "grad_norm": 0.0955386683344841,
+      "learning_rate": 1.9303361786277506e-05,
+      "loss": 9.2124,
+      "step": 71720
+    },
+    {
+      "epoch": 0.3582111912906689,
+      "grad_norm": 0.09714854508638382,
+      "learning_rate": 1.9301859871335952e-05,
+      "loss": 9.2274,
+      "step": 71730
+    },
+    {
+      "epoch": 0.3582611301156084,
+      "grad_norm": 0.0956362634897232,
+      "learning_rate": 1.9300357956394403e-05,
+      "loss": 9.2244,
+      "step": 71740
+    },
+    {
+      "epoch": 0.3583110689405478,
+      "grad_norm": 0.08816993236541748,
+      "learning_rate": 1.9298856041452853e-05,
+      "loss": 9.2184,
+      "step": 71750
+    },
+    {
+      "epoch": 0.3583610077654873,
+      "grad_norm": 0.09849813580513,
+      "learning_rate": 1.9297354126511303e-05,
+      "loss": 9.2217,
+      "step": 71760
+    },
+    {
+      "epoch": 0.3584109465904267,
+      "grad_norm": 0.0990118458867073,
+      "learning_rate": 1.9295852211569753e-05,
+      "loss": 9.2132,
+      "step": 71770
+    },
+    {
+      "epoch": 0.3584608854153662,
+      "grad_norm": 0.0910298228263855,
+      "learning_rate": 1.92943502966282e-05,
+      "loss": 9.2336,
+      "step": 71780
+    },
+    {
+      "epoch": 0.3585108242403056,
+      "grad_norm": 0.09591134637594223,
+      "learning_rate": 1.929284838168665e-05,
+      "loss": 9.2223,
+      "step": 71790
+    },
+    {
+      "epoch": 0.3585607630652451,
+      "grad_norm": 0.0979471430182457,
+      "learning_rate": 1.92913464667451e-05,
+      "loss": 9.2379,
+      "step": 71800
+    },
+    {
+      "epoch": 0.3586107018901845,
+      "grad_norm": 0.10352909564971924,
+      "learning_rate": 1.928984455180355e-05,
+      "loss": 9.2422,
+      "step": 71810
+    },
+    {
+      "epoch": 0.358660640715124,
+      "grad_norm": 0.09178424626588821,
+      "learning_rate": 1.9288342636862e-05,
+      "loss": 9.2233,
+      "step": 71820
+    },
+    {
+      "epoch": 0.3587105795400634,
+      "grad_norm": 0.09857258945703506,
+      "learning_rate": 1.9286840721920447e-05,
+      "loss": 9.2315,
+      "step": 71830
+    },
+    {
+      "epoch": 0.3587605183650029,
+      "grad_norm": 0.09110614657402039,
+      "learning_rate": 1.9285338806978898e-05,
+      "loss": 9.221,
+      "step": 71840
+    },
+    {
+      "epoch": 0.3588104571899423,
+      "grad_norm": 0.09467658400535583,
+      "learning_rate": 1.9283836892037348e-05,
+      "loss": 9.2234,
+      "step": 71850
+    },
+    {
+      "epoch": 0.3588603960148818,
+      "grad_norm": 0.10426273941993713,
+      "learning_rate": 1.9282334977095798e-05,
+      "loss": 9.2211,
+      "step": 71860
+    },
+    {
+      "epoch": 0.3589103348398212,
+      "grad_norm": 0.09133409708738327,
+      "learning_rate": 1.9280833062154248e-05,
+      "loss": 9.2181,
+      "step": 71870
+    },
+    {
+      "epoch": 0.3589602736647607,
+      "grad_norm": 0.09624101221561432,
+      "learning_rate": 1.9279331147212695e-05,
+      "loss": 9.2161,
+      "step": 71880
+    },
+    {
+      "epoch": 0.3590102124897001,
+      "grad_norm": 0.08982755988836288,
+      "learning_rate": 1.9277829232271145e-05,
+      "loss": 9.2249,
+      "step": 71890
+    },
+    {
+      "epoch": 0.3590601513146396,
+      "grad_norm": 0.09154070168733597,
+      "learning_rate": 1.9276327317329595e-05,
+      "loss": 9.2266,
+      "step": 71900
+    },
+    {
+      "epoch": 0.359110090139579,
+      "grad_norm": 0.09227220714092255,
+      "learning_rate": 1.9274825402388045e-05,
+      "loss": 9.2306,
+      "step": 71910
+    },
+    {
+      "epoch": 0.3591600289645185,
+      "grad_norm": 0.09317903965711594,
+      "learning_rate": 1.9273323487446496e-05,
+      "loss": 9.217,
+      "step": 71920
+    },
+    {
+      "epoch": 0.3592099677894579,
+      "grad_norm": 0.09271214157342911,
+      "learning_rate": 1.9271821572504942e-05,
+      "loss": 9.2249,
+      "step": 71930
+    },
+    {
+      "epoch": 0.3592599066143974,
+      "grad_norm": 0.10173588246107101,
+      "learning_rate": 1.9270319657563393e-05,
+      "loss": 9.2237,
+      "step": 71940
+    },
+    {
+      "epoch": 0.3593098454393368,
+      "grad_norm": 0.09192373603582382,
+      "learning_rate": 1.9268817742621843e-05,
+      "loss": 9.2223,
+      "step": 71950
+    },
+    {
+      "epoch": 0.3593597842642763,
+      "grad_norm": 0.09208924323320389,
+      "learning_rate": 1.9267315827680293e-05,
+      "loss": 9.2111,
+      "step": 71960
+    },
+    {
+      "epoch": 0.3594097230892157,
+      "grad_norm": 0.09781036525964737,
+      "learning_rate": 1.9265813912738743e-05,
+      "loss": 9.2236,
+      "step": 71970
+    },
+    {
+      "epoch": 0.3594596619141552,
+      "grad_norm": 0.09488288313150406,
+      "learning_rate": 1.926431199779719e-05,
+      "loss": 9.2249,
+      "step": 71980
+    },
+    {
+      "epoch": 0.3595096007390946,
+      "grad_norm": 0.08927453309297562,
+      "learning_rate": 1.9262810082855643e-05,
+      "loss": 9.2323,
+      "step": 71990
+    },
+    {
+      "epoch": 0.3595595395640341,
+      "grad_norm": 0.09006758779287338,
+      "learning_rate": 1.926130816791409e-05,
+      "loss": 9.2224,
+      "step": 72000
+    },
+    {
+      "epoch": 0.3596094783889735,
+      "grad_norm": 0.09184703975915909,
+      "learning_rate": 1.925980625297254e-05,
+      "loss": 9.2305,
+      "step": 72010
+    },
+    {
+      "epoch": 0.359659417213913,
+      "grad_norm": 0.0887824296951294,
+      "learning_rate": 1.925830433803099e-05,
+      "loss": 9.2437,
+      "step": 72020
+    },
+    {
+      "epoch": 0.3597093560388524,
+      "grad_norm": 0.09438683092594147,
+      "learning_rate": 1.9256802423089437e-05,
+      "loss": 9.2207,
+      "step": 72030
+    },
+    {
+      "epoch": 0.3597592948637919,
+      "grad_norm": 0.0963621437549591,
+      "learning_rate": 1.925530050814789e-05,
+      "loss": 9.2236,
+      "step": 72040
+    },
+    {
+      "epoch": 0.3598092336887313,
+      "grad_norm": 0.09187499433755875,
+      "learning_rate": 1.9253798593206338e-05,
+      "loss": 9.2143,
+      "step": 72050
+    },
+    {
+      "epoch": 0.3598591725136708,
+      "grad_norm": 0.0932173877954483,
+      "learning_rate": 1.9252296678264788e-05,
+      "loss": 9.228,
+      "step": 72060
+    },
+    {
+      "epoch": 0.3599091113386102,
+      "grad_norm": 0.09906915575265884,
+      "learning_rate": 1.9250794763323238e-05,
+      "loss": 9.2218,
+      "step": 72070
+    },
+    {
+      "epoch": 0.3599590501635497,
+      "grad_norm": 0.09343940764665604,
+      "learning_rate": 1.9249292848381685e-05,
+      "loss": 9.2335,
+      "step": 72080
+    },
+    {
+      "epoch": 0.3600089889884891,
+      "grad_norm": 0.09147544950246811,
+      "learning_rate": 1.924779093344014e-05,
+      "loss": 9.2215,
+      "step": 72090
+    },
+    {
+      "epoch": 0.3600589278134286,
+      "grad_norm": 0.0935380831360817,
+      "learning_rate": 1.9246289018498585e-05,
+      "loss": 9.217,
+      "step": 72100
+    },
+    {
+      "epoch": 0.360108866638368,
+      "grad_norm": 0.09664879739284515,
+      "learning_rate": 1.9244787103557035e-05,
+      "loss": 9.2295,
+      "step": 72110
+    },
+    {
+      "epoch": 0.3601588054633075,
+      "grad_norm": 0.09369748085737228,
+      "learning_rate": 1.9243285188615486e-05,
+      "loss": 9.2209,
+      "step": 72120
+    },
+    {
+      "epoch": 0.3602087442882469,
+      "grad_norm": 0.09158635139465332,
+      "learning_rate": 1.9241783273673936e-05,
+      "loss": 9.2208,
+      "step": 72130
+    },
+    {
+      "epoch": 0.3602586831131864,
+      "grad_norm": 0.08802355080842972,
+      "learning_rate": 1.9240281358732386e-05,
+      "loss": 9.2261,
+      "step": 72140
+    },
+    {
+      "epoch": 0.3603086219381258,
+      "grad_norm": 0.09446180611848831,
+      "learning_rate": 1.9238779443790833e-05,
+      "loss": 9.2287,
+      "step": 72150
+    },
+    {
+      "epoch": 0.36035856076306527,
+      "grad_norm": 0.10025842487812042,
+      "learning_rate": 1.9237277528849283e-05,
+      "loss": 9.2188,
+      "step": 72160
+    },
+    {
+      "epoch": 0.3604084995880047,
+      "grad_norm": 0.09552531689405441,
+      "learning_rate": 1.9235775613907733e-05,
+      "loss": 9.213,
+      "step": 72170
+    },
+    {
+      "epoch": 0.36045843841294417,
+      "grad_norm": 0.09716292470693588,
+      "learning_rate": 1.9234273698966183e-05,
+      "loss": 9.2195,
+      "step": 72180
+    },
+    {
+      "epoch": 0.3605083772378836,
+      "grad_norm": 0.09107545018196106,
+      "learning_rate": 1.9232771784024633e-05,
+      "loss": 9.2349,
+      "step": 72190
+    },
+    {
+      "epoch": 0.360558316062823,
+      "grad_norm": 0.09314329922199249,
+      "learning_rate": 1.923126986908308e-05,
+      "loss": 9.2195,
+      "step": 72200
+    },
+    {
+      "epoch": 0.3606082548877625,
+      "grad_norm": 0.09497959911823273,
+      "learning_rate": 1.922976795414153e-05,
+      "loss": 9.216,
+      "step": 72210
+    },
+    {
+      "epoch": 0.3606581937127019,
+      "grad_norm": 0.09450450539588928,
+      "learning_rate": 1.922826603919998e-05,
+      "loss": 9.2196,
+      "step": 72220
+    },
+    {
+      "epoch": 0.3607081325376414,
+      "grad_norm": 0.09029270708560944,
+      "learning_rate": 1.922676412425843e-05,
+      "loss": 9.2221,
+      "step": 72230
+    },
+    {
+      "epoch": 0.3607580713625808,
+      "grad_norm": 0.09402526915073395,
+      "learning_rate": 1.922526220931688e-05,
+      "loss": 9.2112,
+      "step": 72240
+    },
+    {
+      "epoch": 0.3608080101875203,
+      "grad_norm": 0.09604809433221817,
+      "learning_rate": 1.9223760294375328e-05,
+      "loss": 9.2266,
+      "step": 72250
+    },
+    {
+      "epoch": 0.3608579490124597,
+      "grad_norm": 0.09894632548093796,
+      "learning_rate": 1.9222258379433778e-05,
+      "loss": 9.2271,
+      "step": 72260
+    },
+    {
+      "epoch": 0.3609078878373992,
+      "grad_norm": 0.0926297977566719,
+      "learning_rate": 1.9220756464492228e-05,
+      "loss": 9.219,
+      "step": 72270
+    },
+    {
+      "epoch": 0.3609578266623386,
+      "grad_norm": 0.08637505769729614,
+      "learning_rate": 1.9219254549550678e-05,
+      "loss": 9.2202,
+      "step": 72280
+    },
+    {
+      "epoch": 0.3610077654872781,
+      "grad_norm": 0.09499508887529373,
+      "learning_rate": 1.921775263460913e-05,
+      "loss": 9.2172,
+      "step": 72290
+    },
+    {
+      "epoch": 0.3610577043122175,
+      "grad_norm": 0.09541734308004379,
+      "learning_rate": 1.9216250719667575e-05,
+      "loss": 9.2201,
+      "step": 72300
+    },
+    {
+      "epoch": 0.361107643137157,
+      "grad_norm": 0.09449708461761475,
+      "learning_rate": 1.921474880472603e-05,
+      "loss": 9.2271,
+      "step": 72310
+    },
+    {
+      "epoch": 0.3611575819620964,
+      "grad_norm": 0.1025526151061058,
+      "learning_rate": 1.9213246889784476e-05,
+      "loss": 9.2293,
+      "step": 72320
+    },
+    {
+      "epoch": 0.3612075207870359,
+      "grad_norm": 0.08863754570484161,
+      "learning_rate": 1.9211744974842926e-05,
+      "loss": 9.2345,
+      "step": 72330
+    },
+    {
+      "epoch": 0.3612574596119753,
+      "grad_norm": 0.09427140653133392,
+      "learning_rate": 1.9210243059901376e-05,
+      "loss": 9.2335,
+      "step": 72340
+    },
+    {
+      "epoch": 0.3613073984369148,
+      "grad_norm": 0.09397964179515839,
+      "learning_rate": 1.9208741144959823e-05,
+      "loss": 9.2185,
+      "step": 72350
+    },
+    {
+      "epoch": 0.3613573372618542,
+      "grad_norm": 0.09295367449522018,
+      "learning_rate": 1.9207239230018276e-05,
+      "loss": 9.2255,
+      "step": 72360
+    },
+    {
+      "epoch": 0.3614072760867937,
+      "grad_norm": 0.09041722118854523,
+      "learning_rate": 1.9205737315076723e-05,
+      "loss": 9.2303,
+      "step": 72370
+    },
+    {
+      "epoch": 0.3614572149117331,
+      "grad_norm": 0.08931518346071243,
+      "learning_rate": 1.9204235400135173e-05,
+      "loss": 9.2179,
+      "step": 72380
+    },
+    {
+      "epoch": 0.3615071537366726,
+      "grad_norm": 0.0969778448343277,
+      "learning_rate": 1.9202733485193623e-05,
+      "loss": 9.2101,
+      "step": 72390
+    },
+    {
+      "epoch": 0.361557092561612,
+      "grad_norm": 0.09464286267757416,
+      "learning_rate": 1.920123157025207e-05,
+      "loss": 9.2113,
+      "step": 72400
+    },
+    {
+      "epoch": 0.3616070313865515,
+      "grad_norm": 0.09430237114429474,
+      "learning_rate": 1.9199729655310524e-05,
+      "loss": 9.2103,
+      "step": 72410
+    },
+    {
+      "epoch": 0.3616569702114909,
+      "grad_norm": 0.0944664403796196,
+      "learning_rate": 1.919822774036897e-05,
+      "loss": 9.2215,
+      "step": 72420
+    },
+    {
+      "epoch": 0.3617069090364304,
+      "grad_norm": 0.09468348324298859,
+      "learning_rate": 1.919672582542742e-05,
+      "loss": 9.2222,
+      "step": 72430
+    },
+    {
+      "epoch": 0.3617568478613698,
+      "grad_norm": 0.08907091617584229,
+      "learning_rate": 1.919522391048587e-05,
+      "loss": 9.2238,
+      "step": 72440
+    },
+    {
+      "epoch": 0.3618067866863093,
+      "grad_norm": 0.09387124329805374,
+      "learning_rate": 1.9193721995544318e-05,
+      "loss": 9.2185,
+      "step": 72450
+    },
+    {
+      "epoch": 0.3618567255112487,
+      "grad_norm": 0.09770091623067856,
+      "learning_rate": 1.919222008060277e-05,
+      "loss": 9.2177,
+      "step": 72460
+    },
+    {
+      "epoch": 0.3619066643361882,
+      "grad_norm": 0.09893268346786499,
+      "learning_rate": 1.9190718165661218e-05,
+      "loss": 9.2117,
+      "step": 72470
+    },
+    {
+      "epoch": 0.3619566031611276,
+      "grad_norm": 0.09433605521917343,
+      "learning_rate": 1.9189216250719668e-05,
+      "loss": 9.2256,
+      "step": 72480
+    },
+    {
+      "epoch": 0.3620065419860671,
+      "grad_norm": 0.09570042043924332,
+      "learning_rate": 1.918771433577812e-05,
+      "loss": 9.2137,
+      "step": 72490
+    },
+    {
+      "epoch": 0.3620564808110065,
+      "grad_norm": 0.09163347631692886,
+      "learning_rate": 1.9186212420836565e-05,
+      "loss": 9.2188,
+      "step": 72500
+    },
+    {
+      "epoch": 0.362106419635946,
+      "grad_norm": 0.08918924629688263,
+      "learning_rate": 1.918471050589502e-05,
+      "loss": 9.215,
+      "step": 72510
+    },
+    {
+      "epoch": 0.3621563584608854,
+      "grad_norm": 0.10020863264799118,
+      "learning_rate": 1.9183208590953466e-05,
+      "loss": 9.2027,
+      "step": 72520
+    },
+    {
+      "epoch": 0.3622062972858249,
+      "grad_norm": 0.09962757676839828,
+      "learning_rate": 1.9181706676011916e-05,
+      "loss": 9.2011,
+      "step": 72530
+    },
+    {
+      "epoch": 0.3622562361107643,
+      "grad_norm": 0.09085911512374878,
+      "learning_rate": 1.9180204761070366e-05,
+      "loss": 9.2144,
+      "step": 72540
+    },
+    {
+      "epoch": 0.3623061749357038,
+      "grad_norm": 0.0901787281036377,
+      "learning_rate": 1.9178702846128813e-05,
+      "loss": 9.2207,
+      "step": 72550
+    },
+    {
+      "epoch": 0.3623561137606432,
+      "grad_norm": 0.09482782334089279,
+      "learning_rate": 1.9177200931187266e-05,
+      "loss": 9.2262,
+      "step": 72560
+    },
+    {
+      "epoch": 0.36240605258558267,
+      "grad_norm": 0.09666303545236588,
+      "learning_rate": 1.9175699016245713e-05,
+      "loss": 9.2132,
+      "step": 72570
+    },
+    {
+      "epoch": 0.3624559914105221,
+      "grad_norm": 0.09145186096429825,
+      "learning_rate": 1.9174197101304163e-05,
+      "loss": 9.219,
+      "step": 72580
+    },
+    {
+      "epoch": 0.36250593023546157,
+      "grad_norm": 0.09651076048612595,
+      "learning_rate": 1.9172695186362613e-05,
+      "loss": 9.2178,
+      "step": 72590
+    },
+    {
+      "epoch": 0.362555869060401,
+      "grad_norm": 0.0937245786190033,
+      "learning_rate": 1.917119327142106e-05,
+      "loss": 9.2151,
+      "step": 72600
+    },
+    {
+      "epoch": 0.36260580788534047,
+      "grad_norm": 0.09232796728610992,
+      "learning_rate": 1.9169691356479514e-05,
+      "loss": 9.2195,
+      "step": 72610
+    },
+    {
+      "epoch": 0.3626557467102799,
+      "grad_norm": 0.09451702237129211,
+      "learning_rate": 1.916818944153796e-05,
+      "loss": 9.2251,
+      "step": 72620
+    },
+    {
+      "epoch": 0.36270568553521937,
+      "grad_norm": 0.08963721245527267,
+      "learning_rate": 1.9166687526596414e-05,
+      "loss": 9.2175,
+      "step": 72630
+    },
+    {
+      "epoch": 0.3627556243601588,
+      "grad_norm": 0.09087999910116196,
+      "learning_rate": 1.916518561165486e-05,
+      "loss": 9.2081,
+      "step": 72640
+    },
+    {
+      "epoch": 0.36280556318509827,
+      "grad_norm": 0.09726609289646149,
+      "learning_rate": 1.9163683696713308e-05,
+      "loss": 9.2042,
+      "step": 72650
+    },
+    {
+      "epoch": 0.3628555020100377,
+      "grad_norm": 0.09195300191640854,
+      "learning_rate": 1.916218178177176e-05,
+      "loss": 9.2071,
+      "step": 72660
+    },
+    {
+      "epoch": 0.36290544083497717,
+      "grad_norm": 0.09186240285634995,
+      "learning_rate": 1.9160679866830208e-05,
+      "loss": 9.1994,
+      "step": 72670
+    },
+    {
+      "epoch": 0.3629553796599166,
+      "grad_norm": 0.09509220719337463,
+      "learning_rate": 1.9159177951888662e-05,
+      "loss": 9.2143,
+      "step": 72680
+    },
+    {
+      "epoch": 0.36300531848485607,
+      "grad_norm": 0.10255935788154602,
+      "learning_rate": 1.915767603694711e-05,
+      "loss": 9.2106,
+      "step": 72690
+    },
+    {
+      "epoch": 0.3630552573097955,
+      "grad_norm": 0.09038600325584412,
+      "learning_rate": 1.9156174122005555e-05,
+      "loss": 9.2043,
+      "step": 72700
+    },
+    {
+      "epoch": 0.36310519613473496,
+      "grad_norm": 0.09224878251552582,
+      "learning_rate": 1.915467220706401e-05,
+      "loss": 9.201,
+      "step": 72710
+    },
+    {
+      "epoch": 0.3631551349596744,
+      "grad_norm": 0.0966305211186409,
+      "learning_rate": 1.9153170292122456e-05,
+      "loss": 9.2141,
+      "step": 72720
+    },
+    {
+      "epoch": 0.36320507378461386,
+      "grad_norm": 0.0945696085691452,
+      "learning_rate": 1.915166837718091e-05,
+      "loss": 9.2121,
+      "step": 72730
+    },
+    {
+      "epoch": 0.3632550126095533,
+      "grad_norm": 0.09432712197303772,
+      "learning_rate": 1.9150166462239356e-05,
+      "loss": 9.2173,
+      "step": 72740
+    },
+    {
+      "epoch": 0.36330495143449276,
+      "grad_norm": 0.09079490602016449,
+      "learning_rate": 1.9148664547297803e-05,
+      "loss": 9.2089,
+      "step": 72750
+    },
+    {
+      "epoch": 0.3633548902594322,
+      "grad_norm": 0.09984689205884933,
+      "learning_rate": 1.9147162632356256e-05,
+      "loss": 9.2111,
+      "step": 72760
+    },
+    {
+      "epoch": 0.36340482908437166,
+      "grad_norm": 0.09602593630552292,
+      "learning_rate": 1.9145660717414703e-05,
+      "loss": 9.207,
+      "step": 72770
+    },
+    {
+      "epoch": 0.3634547679093111,
+      "grad_norm": 0.09502044320106506,
+      "learning_rate": 1.9144158802473157e-05,
+      "loss": 9.2066,
+      "step": 72780
+    },
+    {
+      "epoch": 0.36350470673425056,
+      "grad_norm": 0.0925392359495163,
+      "learning_rate": 1.9142656887531603e-05,
+      "loss": 9.2145,
+      "step": 72790
+    },
+    {
+      "epoch": 0.36355464555919,
+      "grad_norm": 0.09208685904741287,
+      "learning_rate": 1.914115497259005e-05,
+      "loss": 9.229,
+      "step": 72800
+    },
+    {
+      "epoch": 0.36360458438412946,
+      "grad_norm": 0.09518954902887344,
+      "learning_rate": 1.9139653057648504e-05,
+      "loss": 9.2098,
+      "step": 72810
+    },
+    {
+      "epoch": 0.3636545232090689,
+      "grad_norm": 0.09336472302675247,
+      "learning_rate": 1.913815114270695e-05,
+      "loss": 9.2115,
+      "step": 72820
+    },
+    {
+      "epoch": 0.36370446203400836,
+      "grad_norm": 0.09499195218086243,
+      "learning_rate": 1.9136649227765404e-05,
+      "loss": 9.2051,
+      "step": 72830
+    },
+    {
+      "epoch": 0.3637544008589478,
+      "grad_norm": 0.09333008527755737,
+      "learning_rate": 1.913514731282385e-05,
+      "loss": 9.2088,
+      "step": 72840
+    },
+    {
+      "epoch": 0.36380433968388726,
+      "grad_norm": 0.0993448942899704,
+      "learning_rate": 1.9133645397882298e-05,
+      "loss": 9.2097,
+      "step": 72850
+    },
+    {
+      "epoch": 0.3638542785088267,
+      "grad_norm": 0.0906200036406517,
+      "learning_rate": 1.913214348294075e-05,
+      "loss": 9.2196,
+      "step": 72860
+    },
+    {
+      "epoch": 0.36390421733376616,
+      "grad_norm": 0.09111246466636658,
+      "learning_rate": 1.9130641567999198e-05,
+      "loss": 9.21,
+      "step": 72870
+    },
+    {
+      "epoch": 0.3639541561587056,
+      "grad_norm": 0.09346730262041092,
+      "learning_rate": 1.9129139653057652e-05,
+      "loss": 9.2001,
+      "step": 72880
+    },
+    {
+      "epoch": 0.36400409498364505,
+      "grad_norm": 0.09085837006568909,
+      "learning_rate": 1.91276377381161e-05,
+      "loss": 9.2215,
+      "step": 72890
+    },
+    {
+      "epoch": 0.3640540338085845,
+      "grad_norm": 0.09417754411697388,
+      "learning_rate": 1.9126135823174545e-05,
+      "loss": 9.2197,
+      "step": 72900
+    },
+    {
+      "epoch": 0.36410397263352395,
+      "grad_norm": 0.09656990319490433,
+      "learning_rate": 1.9124633908233e-05,
+      "loss": 9.2044,
+      "step": 72910
+    },
+    {
+      "epoch": 0.3641539114584634,
+      "grad_norm": 0.0960773453116417,
+      "learning_rate": 1.9123131993291446e-05,
+      "loss": 9.2099,
+      "step": 72920
+    },
+    {
+      "epoch": 0.36420385028340285,
+      "grad_norm": 0.10151112079620361,
+      "learning_rate": 1.91216300783499e-05,
+      "loss": 9.2108,
+      "step": 72930
+    },
+    {
+      "epoch": 0.3642537891083423,
+      "grad_norm": 0.09444685280323029,
+      "learning_rate": 1.9120128163408346e-05,
+      "loss": 9.2017,
+      "step": 72940
+    },
+    {
+      "epoch": 0.36430372793328175,
+      "grad_norm": 0.09510553628206253,
+      "learning_rate": 1.9118626248466796e-05,
+      "loss": 9.2096,
+      "step": 72950
+    },
+    {
+      "epoch": 0.3643536667582212,
+      "grad_norm": 0.09737194329500198,
+      "learning_rate": 1.9117124333525246e-05,
+      "loss": 9.2117,
+      "step": 72960
+    },
+    {
+      "epoch": 0.36440360558316065,
+      "grad_norm": 0.09484457224607468,
+      "learning_rate": 1.9115622418583693e-05,
+      "loss": 9.1955,
+      "step": 72970
+    },
+    {
+      "epoch": 0.36445354440810007,
+      "grad_norm": 0.09121875464916229,
+      "learning_rate": 1.9114120503642147e-05,
+      "loss": 9.2148,
+      "step": 72980
+    },
+    {
+      "epoch": 0.36450348323303955,
+      "grad_norm": 0.09583168476819992,
+      "learning_rate": 1.9112618588700594e-05,
+      "loss": 9.2091,
+      "step": 72990
+    },
+    {
+      "epoch": 0.36455342205797897,
+      "grad_norm": 0.09783457964658737,
+      "learning_rate": 1.9111116673759044e-05,
+      "loss": 9.1992,
+      "step": 73000
+    },
+    {
+      "epoch": 0.36460336088291845,
+      "grad_norm": 0.09332898259162903,
+      "learning_rate": 1.9109614758817494e-05,
+      "loss": 9.2091,
+      "step": 73010
+    },
+    {
+      "epoch": 0.36465329970785787,
+      "grad_norm": 0.09087220579385757,
+      "learning_rate": 1.910811284387594e-05,
+      "loss": 9.2208,
+      "step": 73020
+    },
+    {
+      "epoch": 0.36470323853279735,
+      "grad_norm": 0.09480462223291397,
+      "learning_rate": 1.9106610928934394e-05,
+      "loss": 9.2173,
+      "step": 73030
+    },
+    {
+      "epoch": 0.36475317735773677,
+      "grad_norm": 0.0977669283747673,
+      "learning_rate": 1.910510901399284e-05,
+      "loss": 9.2026,
+      "step": 73040
+    },
+    {
+      "epoch": 0.36480311618267625,
+      "grad_norm": 0.08492804318666458,
+      "learning_rate": 1.910360709905129e-05,
+      "loss": 9.2201,
+      "step": 73050
+    },
+    {
+      "epoch": 0.36485305500761567,
+      "grad_norm": 0.09246183931827545,
+      "learning_rate": 1.910210518410974e-05,
+      "loss": 9.2205,
+      "step": 73060
+    },
+    {
+      "epoch": 0.36490299383255514,
+      "grad_norm": 0.08867323398590088,
+      "learning_rate": 1.9100603269168188e-05,
+      "loss": 9.2047,
+      "step": 73070
+    },
+    {
+      "epoch": 0.36495293265749457,
+      "grad_norm": 0.09597288072109222,
+      "learning_rate": 1.9099101354226642e-05,
+      "loss": 9.2167,
+      "step": 73080
+    },
+    {
+      "epoch": 0.36500287148243404,
+      "grad_norm": 0.0952095165848732,
+      "learning_rate": 1.909759943928509e-05,
+      "loss": 9.2064,
+      "step": 73090
+    },
+    {
+      "epoch": 0.36505281030737347,
+      "grad_norm": 0.08946932852268219,
+      "learning_rate": 1.909609752434354e-05,
+      "loss": 9.2084,
+      "step": 73100
+    },
+    {
+      "epoch": 0.36510274913231294,
+      "grad_norm": 0.09637124091386795,
+      "learning_rate": 1.909459560940199e-05,
+      "loss": 9.2031,
+      "step": 73110
+    },
+    {
+      "epoch": 0.36515268795725236,
+      "grad_norm": 0.09121710807085037,
+      "learning_rate": 1.9093093694460436e-05,
+      "loss": 9.2069,
+      "step": 73120
+    },
+    {
+      "epoch": 0.36520262678219184,
+      "grad_norm": 0.09637744724750519,
+      "learning_rate": 1.909159177951889e-05,
+      "loss": 9.2288,
+      "step": 73130
+    },
+    {
+      "epoch": 0.36525256560713126,
+      "grad_norm": 0.09214238077402115,
+      "learning_rate": 1.9090089864577336e-05,
+      "loss": 9.2124,
+      "step": 73140
+    },
+    {
+      "epoch": 0.36530250443207074,
+      "grad_norm": 0.08832499384880066,
+      "learning_rate": 1.9088587949635786e-05,
+      "loss": 9.2153,
+      "step": 73150
+    },
+    {
+      "epoch": 0.36535244325701016,
+      "grad_norm": 0.09127902239561081,
+      "learning_rate": 1.9087086034694236e-05,
+      "loss": 9.2171,
+      "step": 73160
+    },
+    {
+      "epoch": 0.36540238208194964,
+      "grad_norm": 0.09237400442361832,
+      "learning_rate": 1.9085584119752683e-05,
+      "loss": 9.2115,
+      "step": 73170
+    },
+    {
+      "epoch": 0.36545232090688906,
+      "grad_norm": 0.09848447889089584,
+      "learning_rate": 1.9084082204811137e-05,
+      "loss": 9.2101,
+      "step": 73180
+    },
+    {
+      "epoch": 0.3655022597318285,
+      "grad_norm": 0.09511636197566986,
+      "learning_rate": 1.9082580289869584e-05,
+      "loss": 9.2122,
+      "step": 73190
+    },
+    {
+      "epoch": 0.36555219855676796,
+      "grad_norm": 0.09176871925592422,
+      "learning_rate": 1.9081078374928034e-05,
+      "loss": 9.2076,
+      "step": 73200
+    },
+    {
+      "epoch": 0.3656021373817074,
+      "grad_norm": 0.0967327430844307,
+      "learning_rate": 1.9079576459986484e-05,
+      "loss": 9.213,
+      "step": 73210
+    },
+    {
+      "epoch": 0.36565207620664686,
+      "grad_norm": 0.09490523487329483,
+      "learning_rate": 1.907807454504493e-05,
+      "loss": 9.2131,
+      "step": 73220
+    },
+    {
+      "epoch": 0.3657020150315863,
+      "grad_norm": 0.09480072557926178,
+      "learning_rate": 1.9076572630103384e-05,
+      "loss": 9.2003,
+      "step": 73230
+    },
+    {
+      "epoch": 0.36575195385652576,
+      "grad_norm": 0.08888121694326401,
+      "learning_rate": 1.907507071516183e-05,
+      "loss": 9.1996,
+      "step": 73240
+    },
+    {
+      "epoch": 0.3658018926814652,
+      "grad_norm": 0.09302229434251785,
+      "learning_rate": 1.907356880022028e-05,
+      "loss": 9.2104,
+      "step": 73250
+    },
+    {
+      "epoch": 0.36585183150640466,
+      "grad_norm": 0.09518302977085114,
+      "learning_rate": 1.907206688527873e-05,
+      "loss": 9.1888,
+      "step": 73260
+    },
+    {
+      "epoch": 0.3659017703313441,
+      "grad_norm": 0.09398709237575531,
+      "learning_rate": 1.907056497033718e-05,
+      "loss": 9.2083,
+      "step": 73270
+    },
+    {
+      "epoch": 0.36595170915628356,
+      "grad_norm": 0.09403207898139954,
+      "learning_rate": 1.9069063055395632e-05,
+      "loss": 9.2055,
+      "step": 73280
+    },
+    {
+      "epoch": 0.366001647981223,
+      "grad_norm": 0.09166137129068375,
+      "learning_rate": 1.906756114045408e-05,
+      "loss": 9.2105,
+      "step": 73290
+    },
+    {
+      "epoch": 0.36605158680616245,
+      "grad_norm": 0.08635640889406204,
+      "learning_rate": 1.906605922551253e-05,
+      "loss": 9.2103,
+      "step": 73300
+    },
+    {
+      "epoch": 0.3661015256311019,
+      "grad_norm": 0.09352033585309982,
+      "learning_rate": 1.906455731057098e-05,
+      "loss": 9.2081,
+      "step": 73310
+    },
+    {
+      "epoch": 0.36615146445604135,
+      "grad_norm": 0.09585576504468918,
+      "learning_rate": 1.906305539562943e-05,
+      "loss": 9.2063,
+      "step": 73320
+    },
+    {
+      "epoch": 0.3662014032809808,
+      "grad_norm": 0.10186425596475601,
+      "learning_rate": 1.906155348068788e-05,
+      "loss": 9.2024,
+      "step": 73330
+    },
+    {
+      "epoch": 0.36625134210592025,
+      "grad_norm": 0.08858046680688858,
+      "learning_rate": 1.9060051565746326e-05,
+      "loss": 9.2075,
+      "step": 73340
+    },
+    {
+      "epoch": 0.3663012809308597,
+      "grad_norm": 0.09600647538900375,
+      "learning_rate": 1.9058549650804776e-05,
+      "loss": 9.1993,
+      "step": 73350
+    },
+    {
+      "epoch": 0.36635121975579915,
+      "grad_norm": 0.093485526740551,
+      "learning_rate": 1.9057047735863226e-05,
+      "loss": 9.2022,
+      "step": 73360
+    },
+    {
+      "epoch": 0.3664011585807386,
+      "grad_norm": 0.09988955408334732,
+      "learning_rate": 1.9055545820921677e-05,
+      "loss": 9.2141,
+      "step": 73370
+    },
+    {
+      "epoch": 0.36645109740567805,
+      "grad_norm": 0.10079696774482727,
+      "learning_rate": 1.9054043905980127e-05,
+      "loss": 9.2063,
+      "step": 73380
+    },
+    {
+      "epoch": 0.36650103623061747,
+      "grad_norm": 0.09347499907016754,
+      "learning_rate": 1.9052541991038574e-05,
+      "loss": 9.2044,
+      "step": 73390
+    },
+    {
+      "epoch": 0.36655097505555695,
+      "grad_norm": 0.09521867334842682,
+      "learning_rate": 1.9051040076097024e-05,
+      "loss": 9.2019,
+      "step": 73400
+    },
+    {
+      "epoch": 0.36660091388049637,
+      "grad_norm": 0.09266289323568344,
+      "learning_rate": 1.9049538161155474e-05,
+      "loss": 9.2043,
+      "step": 73410
+    },
+    {
+      "epoch": 0.36665085270543585,
+      "grad_norm": 0.09217309206724167,
+      "learning_rate": 1.9048036246213924e-05,
+      "loss": 9.2132,
+      "step": 73420
+    },
+    {
+      "epoch": 0.36670079153037527,
+      "grad_norm": 0.0919642373919487,
+      "learning_rate": 1.9046534331272374e-05,
+      "loss": 9.2002,
+      "step": 73430
+    },
+    {
+      "epoch": 0.36675073035531475,
+      "grad_norm": 0.0903974175453186,
+      "learning_rate": 1.904503241633082e-05,
+      "loss": 9.2031,
+      "step": 73440
+    },
+    {
+      "epoch": 0.36680066918025417,
+      "grad_norm": 0.09081780910491943,
+      "learning_rate": 1.904353050138927e-05,
+      "loss": 9.2017,
+      "step": 73450
+    },
+    {
+      "epoch": 0.36685060800519365,
+      "grad_norm": 0.09484659135341644,
+      "learning_rate": 1.904202858644772e-05,
+      "loss": 9.2111,
+      "step": 73460
+    },
+    {
+      "epoch": 0.36690054683013307,
+      "grad_norm": 0.09380007535219193,
+      "learning_rate": 1.904052667150617e-05,
+      "loss": 9.2005,
+      "step": 73470
+    },
+    {
+      "epoch": 0.36695048565507254,
+      "grad_norm": 0.09407030791044235,
+      "learning_rate": 1.9039024756564622e-05,
+      "loss": 9.2089,
+      "step": 73480
+    },
+    {
+      "epoch": 0.36700042448001197,
+      "grad_norm": 0.09412670135498047,
+      "learning_rate": 1.903752284162307e-05,
+      "loss": 9.2009,
+      "step": 73490
+    },
+    {
+      "epoch": 0.36705036330495144,
+      "grad_norm": 0.09622817486524582,
+      "learning_rate": 1.903602092668152e-05,
+      "loss": 9.2022,
+      "step": 73500
+    },
+    {
+      "epoch": 0.36710030212989087,
+      "grad_norm": 0.09574905782938004,
+      "learning_rate": 1.903451901173997e-05,
+      "loss": 9.2067,
+      "step": 73510
+    },
+    {
+      "epoch": 0.36715024095483034,
+      "grad_norm": 0.09512118250131607,
+      "learning_rate": 1.903301709679842e-05,
+      "loss": 9.2085,
+      "step": 73520
+    },
+    {
+      "epoch": 0.36720017977976976,
+      "grad_norm": 0.08740708976984024,
+      "learning_rate": 1.903151518185687e-05,
+      "loss": 9.2168,
+      "step": 73530
+    },
+    {
+      "epoch": 0.36725011860470924,
+      "grad_norm": 0.091730996966362,
+      "learning_rate": 1.9030013266915316e-05,
+      "loss": 9.2058,
+      "step": 73540
+    },
+    {
+      "epoch": 0.36730005742964866,
+      "grad_norm": 0.09345594793558121,
+      "learning_rate": 1.9028511351973766e-05,
+      "loss": 9.1933,
+      "step": 73550
+    },
+    {
+      "epoch": 0.36734999625458814,
+      "grad_norm": 0.1008530855178833,
+      "learning_rate": 1.9027009437032216e-05,
+      "loss": 9.195,
+      "step": 73560
+    },
+    {
+      "epoch": 0.36739993507952756,
+      "grad_norm": 0.10389885306358337,
+      "learning_rate": 1.9025507522090667e-05,
+      "loss": 9.2012,
+      "step": 73570
+    },
+    {
+      "epoch": 0.36744987390446704,
+      "grad_norm": 0.09878107160329819,
+      "learning_rate": 1.9024005607149117e-05,
+      "loss": 9.2016,
+      "step": 73580
+    },
+    {
+      "epoch": 0.36749981272940646,
+      "grad_norm": 0.09092985093593597,
+      "learning_rate": 1.9022503692207564e-05,
+      "loss": 9.2007,
+      "step": 73590
+    },
+    {
+      "epoch": 0.36754975155434594,
+      "grad_norm": 0.09142585843801498,
+      "learning_rate": 1.9021001777266014e-05,
+      "loss": 9.2079,
+      "step": 73600
+    },
+    {
+      "epoch": 0.36759969037928536,
+      "grad_norm": 0.09261012822389603,
+      "learning_rate": 1.9019499862324464e-05,
+      "loss": 9.2041,
+      "step": 73610
+    },
+    {
+      "epoch": 0.36764962920422484,
+      "grad_norm": 0.09699272364377975,
+      "learning_rate": 1.9017997947382914e-05,
+      "loss": 9.2008,
+      "step": 73620
+    },
+    {
+      "epoch": 0.36769956802916426,
+      "grad_norm": 0.0889492779970169,
+      "learning_rate": 1.9016496032441364e-05,
+      "loss": 9.1979,
+      "step": 73630
+    },
+    {
+      "epoch": 0.36774950685410374,
+      "grad_norm": 0.09963024407625198,
+      "learning_rate": 1.9014994117499814e-05,
+      "loss": 9.2071,
+      "step": 73640
+    },
+    {
+      "epoch": 0.36779944567904316,
+      "grad_norm": 0.08926968276500702,
+      "learning_rate": 1.901349220255826e-05,
+      "loss": 9.2077,
+      "step": 73650
+    },
+    {
+      "epoch": 0.36784938450398263,
+      "grad_norm": 0.10151121020317078,
+      "learning_rate": 1.901199028761671e-05,
+      "loss": 9.1939,
+      "step": 73660
+    },
+    {
+      "epoch": 0.36789932332892206,
+      "grad_norm": 0.09443957358598709,
+      "learning_rate": 1.901048837267516e-05,
+      "loss": 9.2078,
+      "step": 73670
+    },
+    {
+      "epoch": 0.36794926215386153,
+      "grad_norm": 0.09295155853033066,
+      "learning_rate": 1.9008986457733612e-05,
+      "loss": 9.2032,
+      "step": 73680
+    },
+    {
+      "epoch": 0.36799920097880096,
+      "grad_norm": 0.09365091472864151,
+      "learning_rate": 1.9007484542792062e-05,
+      "loss": 9.1929,
+      "step": 73690
+    },
+    {
+      "epoch": 0.36804913980374043,
+      "grad_norm": 0.09538646787405014,
+      "learning_rate": 1.900598262785051e-05,
+      "loss": 9.1853,
+      "step": 73700
+    },
+    {
+      "epoch": 0.36809907862867985,
+      "grad_norm": 0.09404728561639786,
+      "learning_rate": 1.900448071290896e-05,
+      "loss": 9.2012,
+      "step": 73710
+    },
+    {
+      "epoch": 0.36814901745361933,
+      "grad_norm": 0.09444870799779892,
+      "learning_rate": 1.900297879796741e-05,
+      "loss": 9.1991,
+      "step": 73720
+    },
+    {
+      "epoch": 0.36819895627855875,
+      "grad_norm": 0.09309567511081696,
+      "learning_rate": 1.900147688302586e-05,
+      "loss": 9.1996,
+      "step": 73730
+    },
+    {
+      "epoch": 0.36824889510349823,
+      "grad_norm": 0.09412892907857895,
+      "learning_rate": 1.899997496808431e-05,
+      "loss": 9.1956,
+      "step": 73740
+    },
+    {
+      "epoch": 0.36829883392843765,
+      "grad_norm": 0.09056131541728973,
+      "learning_rate": 1.8998473053142756e-05,
+      "loss": 9.2004,
+      "step": 73750
+    },
+    {
+      "epoch": 0.36834877275337713,
+      "grad_norm": 0.09614895284175873,
+      "learning_rate": 1.8996971138201206e-05,
+      "loss": 9.2077,
+      "step": 73760
+    },
+    {
+      "epoch": 0.36839871157831655,
+      "grad_norm": 0.09568426012992859,
+      "learning_rate": 1.8995469223259657e-05,
+      "loss": 9.196,
+      "step": 73770
+    },
+    {
+      "epoch": 0.36844865040325603,
+      "grad_norm": 0.09588686376810074,
+      "learning_rate": 1.8993967308318107e-05,
+      "loss": 9.1894,
+      "step": 73780
+    },
+    {
+      "epoch": 0.36849858922819545,
+      "grad_norm": 0.09887325763702393,
+      "learning_rate": 1.8992465393376557e-05,
+      "loss": 9.1864,
+      "step": 73790
+    },
+    {
+      "epoch": 0.3685485280531349,
+      "grad_norm": 0.09332797676324844,
+      "learning_rate": 1.8990963478435004e-05,
+      "loss": 9.1968,
+      "step": 73800
+    },
+    {
+      "epoch": 0.36859846687807435,
+      "grad_norm": 0.10205885022878647,
+      "learning_rate": 1.8989461563493454e-05,
+      "loss": 9.1965,
+      "step": 73810
+    },
+    {
+      "epoch": 0.3686484057030138,
+      "grad_norm": 0.09981803596019745,
+      "learning_rate": 1.8987959648551904e-05,
+      "loss": 9.195,
+      "step": 73820
+    },
+    {
+      "epoch": 0.36869834452795325,
+      "grad_norm": 0.09201083332300186,
+      "learning_rate": 1.8986457733610354e-05,
+      "loss": 9.2003,
+      "step": 73830
+    },
+    {
+      "epoch": 0.3687482833528927,
+      "grad_norm": 0.09341354668140411,
+      "learning_rate": 1.8984955818668804e-05,
+      "loss": 9.2033,
+      "step": 73840
+    },
+    {
+      "epoch": 0.36879822217783215,
+      "grad_norm": 0.0915553867816925,
+      "learning_rate": 1.898345390372725e-05,
+      "loss": 9.2004,
+      "step": 73850
+    },
+    {
+      "epoch": 0.3688481610027716,
+      "grad_norm": 0.09335311502218246,
+      "learning_rate": 1.89819519887857e-05,
+      "loss": 9.2035,
+      "step": 73860
+    },
+    {
+      "epoch": 0.36889809982771105,
+      "grad_norm": 0.10016678273677826,
+      "learning_rate": 1.898045007384415e-05,
+      "loss": 9.196,
+      "step": 73870
+    },
+    {
+      "epoch": 0.3689480386526505,
+      "grad_norm": 0.09652851521968842,
+      "learning_rate": 1.8978948158902602e-05,
+      "loss": 9.202,
+      "step": 73880
+    },
+    {
+      "epoch": 0.36899797747758994,
+      "grad_norm": 0.09326925128698349,
+      "learning_rate": 1.8977446243961052e-05,
+      "loss": 9.1939,
+      "step": 73890
+    },
+    {
+      "epoch": 0.3690479163025294,
+      "grad_norm": 0.09441187977790833,
+      "learning_rate": 1.89759443290195e-05,
+      "loss": 9.1978,
+      "step": 73900
+    },
+    {
+      "epoch": 0.36909785512746884,
+      "grad_norm": 0.09244664758443832,
+      "learning_rate": 1.897444241407795e-05,
+      "loss": 9.1851,
+      "step": 73910
+    },
+    {
+      "epoch": 0.3691477939524083,
+      "grad_norm": 0.09576987475156784,
+      "learning_rate": 1.89729404991364e-05,
+      "loss": 9.1974,
+      "step": 73920
+    },
+    {
+      "epoch": 0.36919773277734774,
+      "grad_norm": 0.10163576900959015,
+      "learning_rate": 1.897143858419485e-05,
+      "loss": 9.1942,
+      "step": 73930
+    },
+    {
+      "epoch": 0.3692476716022872,
+      "grad_norm": 0.09293472766876221,
+      "learning_rate": 1.89699366692533e-05,
+      "loss": 9.2004,
+      "step": 73940
+    },
+    {
+      "epoch": 0.36929761042722664,
+      "grad_norm": 0.09498841315507889,
+      "learning_rate": 1.8968434754311746e-05,
+      "loss": 9.2015,
+      "step": 73950
+    },
+    {
+      "epoch": 0.3693475492521661,
+      "grad_norm": 0.0949832871556282,
+      "learning_rate": 1.89669328393702e-05,
+      "loss": 9.1872,
+      "step": 73960
+    },
+    {
+      "epoch": 0.36939748807710554,
+      "grad_norm": 0.09256528317928314,
+      "learning_rate": 1.8965430924428647e-05,
+      "loss": 9.1972,
+      "step": 73970
+    },
+    {
+      "epoch": 0.369447426902045,
+      "grad_norm": 0.0952831581234932,
+      "learning_rate": 1.8963929009487097e-05,
+      "loss": 9.1852,
+      "step": 73980
+    },
+    {
+      "epoch": 0.36949736572698444,
+      "grad_norm": 0.10047879070043564,
+      "learning_rate": 1.8962427094545547e-05,
+      "loss": 9.2014,
+      "step": 73990
+    },
+    {
+      "epoch": 0.3695473045519239,
+      "grad_norm": 0.09470636397600174,
+      "learning_rate": 1.8960925179603994e-05,
+      "loss": 9.1897,
+      "step": 74000
+    },
+    {
+      "epoch": 0.36959724337686334,
+      "grad_norm": 0.08998063206672668,
+      "learning_rate": 1.8959423264662447e-05,
+      "loss": 9.2008,
+      "step": 74010
+    },
+    {
+      "epoch": 0.3696471822018028,
+      "grad_norm": 0.09682891517877579,
+      "learning_rate": 1.8957921349720894e-05,
+      "loss": 9.1968,
+      "step": 74020
+    },
+    {
+      "epoch": 0.36969712102674224,
+      "grad_norm": 0.093925341963768,
+      "learning_rate": 1.8956419434779344e-05,
+      "loss": 9.2174,
+      "step": 74030
+    },
+    {
+      "epoch": 0.3697470598516817,
+      "grad_norm": 0.09512005746364594,
+      "learning_rate": 1.8954917519837794e-05,
+      "loss": 9.194,
+      "step": 74040
+    },
+    {
+      "epoch": 0.36979699867662114,
+      "grad_norm": 0.09474792331457138,
+      "learning_rate": 1.895341560489624e-05,
+      "loss": 9.1955,
+      "step": 74050
+    },
+    {
+      "epoch": 0.3698469375015606,
+      "grad_norm": 0.08794379979372025,
+      "learning_rate": 1.8951913689954695e-05,
+      "loss": 9.1912,
+      "step": 74060
+    },
+    {
+      "epoch": 0.36989687632650003,
+      "grad_norm": 0.09021158516407013,
+      "learning_rate": 1.895041177501314e-05,
+      "loss": 9.1967,
+      "step": 74070
+    },
+    {
+      "epoch": 0.3699468151514395,
+      "grad_norm": 0.09125207364559174,
+      "learning_rate": 1.8948909860071592e-05,
+      "loss": 9.1879,
+      "step": 74080
+    },
+    {
+      "epoch": 0.36999675397637893,
+      "grad_norm": 0.09770797193050385,
+      "learning_rate": 1.8947407945130042e-05,
+      "loss": 9.1984,
+      "step": 74090
+    },
+    {
+      "epoch": 0.3700466928013184,
+      "grad_norm": 0.09228277951478958,
+      "learning_rate": 1.894590603018849e-05,
+      "loss": 9.1979,
+      "step": 74100
+    },
+    {
+      "epoch": 0.37009663162625783,
+      "grad_norm": 0.09251270443201065,
+      "learning_rate": 1.8944404115246942e-05,
+      "loss": 9.199,
+      "step": 74110
+    },
+    {
+      "epoch": 0.3701465704511973,
+      "grad_norm": 0.09587238729000092,
+      "learning_rate": 1.894290220030539e-05,
+      "loss": 9.1933,
+      "step": 74120
+    },
+    {
+      "epoch": 0.37019650927613673,
+      "grad_norm": 0.08899569511413574,
+      "learning_rate": 1.894140028536384e-05,
+      "loss": 9.1968,
+      "step": 74130
+    },
+    {
+      "epoch": 0.3702464481010762,
+      "grad_norm": 0.09335777163505554,
+      "learning_rate": 1.893989837042229e-05,
+      "loss": 9.1918,
+      "step": 74140
+    },
+    {
+      "epoch": 0.37029638692601563,
+      "grad_norm": 0.09356404095888138,
+      "learning_rate": 1.8938396455480736e-05,
+      "loss": 9.1845,
+      "step": 74150
+    },
+    {
+      "epoch": 0.3703463257509551,
+      "grad_norm": 0.09852045774459839,
+      "learning_rate": 1.893689454053919e-05,
+      "loss": 9.1924,
+      "step": 74160
+    },
+    {
+      "epoch": 0.37039626457589453,
+      "grad_norm": 0.09660044312477112,
+      "learning_rate": 1.8935392625597637e-05,
+      "loss": 9.2063,
+      "step": 74170
+    },
+    {
+      "epoch": 0.37044620340083395,
+      "grad_norm": 0.09085432440042496,
+      "learning_rate": 1.8933890710656087e-05,
+      "loss": 9.1791,
+      "step": 74180
+    },
+    {
+      "epoch": 0.37049614222577343,
+      "grad_norm": 0.09493508189916611,
+      "learning_rate": 1.8932388795714537e-05,
+      "loss": 9.1994,
+      "step": 74190
+    },
+    {
+      "epoch": 0.37054608105071285,
+      "grad_norm": 0.09557243436574936,
+      "learning_rate": 1.8930886880772984e-05,
+      "loss": 9.1895,
+      "step": 74200
+    },
+    {
+      "epoch": 0.3705960198756523,
+      "grad_norm": 0.09346624463796616,
+      "learning_rate": 1.8929384965831437e-05,
+      "loss": 9.1871,
+      "step": 74210
+    },
+    {
+      "epoch": 0.37064595870059175,
+      "grad_norm": 0.09579519182443619,
+      "learning_rate": 1.8927883050889884e-05,
+      "loss": 9.1834,
+      "step": 74220
+    },
+    {
+      "epoch": 0.3706958975255312,
+      "grad_norm": 0.10192973166704178,
+      "learning_rate": 1.8926381135948334e-05,
+      "loss": 9.1982,
+      "step": 74230
+    },
+    {
+      "epoch": 0.37074583635047065,
+      "grad_norm": 0.09080871194601059,
+      "learning_rate": 1.8924879221006784e-05,
+      "loss": 9.2006,
+      "step": 74240
+    },
+    {
+      "epoch": 0.3707957751754101,
+      "grad_norm": 0.09512729942798615,
+      "learning_rate": 1.892337730606523e-05,
+      "loss": 9.191,
+      "step": 74250
+    },
+    {
+      "epoch": 0.37084571400034955,
+      "grad_norm": 0.09607906639575958,
+      "learning_rate": 1.8921875391123685e-05,
+      "loss": 9.1898,
+      "step": 74260
+    },
+    {
+      "epoch": 0.370895652825289,
+      "grad_norm": 0.09581431746482849,
+      "learning_rate": 1.892037347618213e-05,
+      "loss": 9.1822,
+      "step": 74270
+    },
+    {
+      "epoch": 0.37094559165022845,
+      "grad_norm": 0.09435351938009262,
+      "learning_rate": 1.8918871561240585e-05,
+      "loss": 9.1853,
+      "step": 74280
+    },
+    {
+      "epoch": 0.3709955304751679,
+      "grad_norm": 0.09405918419361115,
+      "learning_rate": 1.8917369646299032e-05,
+      "loss": 9.1877,
+      "step": 74290
+    },
+    {
+      "epoch": 0.37104546930010734,
+      "grad_norm": 0.09423495829105377,
+      "learning_rate": 1.891586773135748e-05,
+      "loss": 9.2003,
+      "step": 74300
+    },
+    {
+      "epoch": 0.3710954081250468,
+      "grad_norm": 0.0974857285618782,
+      "learning_rate": 1.8914365816415932e-05,
+      "loss": 9.1888,
+      "step": 74310
+    },
+    {
+      "epoch": 0.37114534694998624,
+      "grad_norm": 0.09585097432136536,
+      "learning_rate": 1.891286390147438e-05,
+      "loss": 9.2001,
+      "step": 74320
+    },
+    {
+      "epoch": 0.3711952857749257,
+      "grad_norm": 0.09716727584600449,
+      "learning_rate": 1.8911361986532833e-05,
+      "loss": 9.1998,
+      "step": 74330
+    },
+    {
+      "epoch": 0.37124522459986514,
+      "grad_norm": 0.09428082406520844,
+      "learning_rate": 1.890986007159128e-05,
+      "loss": 9.1946,
+      "step": 74340
+    },
+    {
+      "epoch": 0.3712951634248046,
+      "grad_norm": 0.09271190315485,
+      "learning_rate": 1.8908358156649726e-05,
+      "loss": 9.1989,
+      "step": 74350
+    },
+    {
+      "epoch": 0.37134510224974404,
+      "grad_norm": 0.09324190765619278,
+      "learning_rate": 1.890685624170818e-05,
+      "loss": 9.1995,
+      "step": 74360
+    },
+    {
+      "epoch": 0.3713950410746835,
+      "grad_norm": 0.09561111032962799,
+      "learning_rate": 1.8905354326766627e-05,
+      "loss": 9.201,
+      "step": 74370
+    },
+    {
+      "epoch": 0.37144497989962294,
+      "grad_norm": 0.09302332252264023,
+      "learning_rate": 1.890385241182508e-05,
+      "loss": 9.1906,
+      "step": 74380
+    },
+    {
+      "epoch": 0.3714949187245624,
+      "grad_norm": 0.09225195646286011,
+      "learning_rate": 1.8902350496883527e-05,
+      "loss": 9.1941,
+      "step": 74390
+    },
+    {
+      "epoch": 0.37154485754950184,
+      "grad_norm": 0.09414453059434891,
+      "learning_rate": 1.8900848581941974e-05,
+      "loss": 9.1922,
+      "step": 74400
+    },
+    {
+      "epoch": 0.3715947963744413,
+      "grad_norm": 0.1040930524468422,
+      "learning_rate": 1.8899346667000427e-05,
+      "loss": 9.1835,
+      "step": 74410
+    },
+    {
+      "epoch": 0.37164473519938074,
+      "grad_norm": 0.10101134330034256,
+      "learning_rate": 1.8897844752058874e-05,
+      "loss": 9.1919,
+      "step": 74420
+    },
+    {
+      "epoch": 0.3716946740243202,
+      "grad_norm": 0.09623407572507858,
+      "learning_rate": 1.8896342837117328e-05,
+      "loss": 9.1749,
+      "step": 74430
+    },
+    {
+      "epoch": 0.37174461284925964,
+      "grad_norm": 0.0893046110868454,
+      "learning_rate": 1.8894840922175775e-05,
+      "loss": 9.191,
+      "step": 74440
+    },
+    {
+      "epoch": 0.3717945516741991,
+      "grad_norm": 0.09399503469467163,
+      "learning_rate": 1.889333900723422e-05,
+      "loss": 9.1985,
+      "step": 74450
+    },
+    {
+      "epoch": 0.37184449049913854,
+      "grad_norm": 0.09701071679592133,
+      "learning_rate": 1.8891837092292675e-05,
+      "loss": 9.1864,
+      "step": 74460
+    },
+    {
+      "epoch": 0.371894429324078,
+      "grad_norm": 0.09602676331996918,
+      "learning_rate": 1.889033517735112e-05,
+      "loss": 9.1936,
+      "step": 74470
+    },
+    {
+      "epoch": 0.37194436814901743,
+      "grad_norm": 0.09337245672941208,
+      "learning_rate": 1.8888833262409575e-05,
+      "loss": 9.1728,
+      "step": 74480
+    },
+    {
+      "epoch": 0.3719943069739569,
+      "grad_norm": 0.09136126190423965,
+      "learning_rate": 1.8887331347468022e-05,
+      "loss": 9.1816,
+      "step": 74490
+    },
+    {
+      "epoch": 0.37204424579889633,
+      "grad_norm": 0.09118318557739258,
+      "learning_rate": 1.888582943252647e-05,
+      "loss": 9.1867,
+      "step": 74500
+    },
+    {
+      "epoch": 0.3720941846238358,
+      "grad_norm": 0.09339746087789536,
+      "learning_rate": 1.8884327517584922e-05,
+      "loss": 9.1911,
+      "step": 74510
+    },
+    {
+      "epoch": 0.37214412344877523,
+      "grad_norm": 0.09358755499124527,
+      "learning_rate": 1.888282560264337e-05,
+      "loss": 9.1926,
+      "step": 74520
+    },
+    {
+      "epoch": 0.3721940622737147,
+      "grad_norm": 0.09610449522733688,
+      "learning_rate": 1.8881323687701823e-05,
+      "loss": 9.188,
+      "step": 74530
+    },
+    {
+      "epoch": 0.37224400109865413,
+      "grad_norm": 0.09586481004953384,
+      "learning_rate": 1.887982177276027e-05,
+      "loss": 9.1889,
+      "step": 74540
+    },
+    {
+      "epoch": 0.3722939399235936,
+      "grad_norm": 0.0982070192694664,
+      "learning_rate": 1.8878319857818716e-05,
+      "loss": 9.1799,
+      "step": 74550
+    },
+    {
+      "epoch": 0.37234387874853303,
+      "grad_norm": 0.09355320036411285,
+      "learning_rate": 1.887681794287717e-05,
+      "loss": 9.191,
+      "step": 74560
+    },
+    {
+      "epoch": 0.3723938175734725,
+      "grad_norm": 0.09481959044933319,
+      "learning_rate": 1.8875316027935617e-05,
+      "loss": 9.186,
+      "step": 74570
+    },
+    {
+      "epoch": 0.37244375639841193,
+      "grad_norm": 0.09273957461118698,
+      "learning_rate": 1.887381411299407e-05,
+      "loss": 9.1892,
+      "step": 74580
+    },
+    {
+      "epoch": 0.3724936952233514,
+      "grad_norm": 0.09700888395309448,
+      "learning_rate": 1.8872312198052517e-05,
+      "loss": 9.1935,
+      "step": 74590
+    },
+    {
+      "epoch": 0.37254363404829083,
+      "grad_norm": 0.09876874834299088,
+      "learning_rate": 1.8870810283110967e-05,
+      "loss": 9.1894,
+      "step": 74600
+    },
+    {
+      "epoch": 0.3725935728732303,
+      "grad_norm": 0.09617974609136581,
+      "learning_rate": 1.8869308368169417e-05,
+      "loss": 9.1823,
+      "step": 74610
+    },
+    {
+      "epoch": 0.3726435116981697,
+      "grad_norm": 0.09874873608350754,
+      "learning_rate": 1.8867806453227864e-05,
+      "loss": 9.198,
+      "step": 74620
+    },
+    {
+      "epoch": 0.3726934505231092,
+      "grad_norm": 0.0900963768362999,
+      "learning_rate": 1.8866304538286318e-05,
+      "loss": 9.1833,
+      "step": 74630
+    },
+    {
+      "epoch": 0.3727433893480486,
+      "grad_norm": 0.09320524334907532,
+      "learning_rate": 1.8864802623344765e-05,
+      "loss": 9.1847,
+      "step": 74640
+    },
+    {
+      "epoch": 0.3727933281729881,
+      "grad_norm": 0.09384310245513916,
+      "learning_rate": 1.8863300708403215e-05,
+      "loss": 9.182,
+      "step": 74650
+    },
+    {
+      "epoch": 0.3728432669979275,
+      "grad_norm": 0.09499943256378174,
+      "learning_rate": 1.8861798793461665e-05,
+      "loss": 9.1918,
+      "step": 74660
+    },
+    {
+      "epoch": 0.372893205822867,
+      "grad_norm": 0.09199872612953186,
+      "learning_rate": 1.886029687852011e-05,
+      "loss": 9.181,
+      "step": 74670
+    },
+    {
+      "epoch": 0.3729431446478064,
+      "grad_norm": 0.0918089747428894,
+      "learning_rate": 1.8858794963578565e-05,
+      "loss": 9.1903,
+      "step": 74680
+    },
+    {
+      "epoch": 0.3729930834727459,
+      "grad_norm": 0.09253137558698654,
+      "learning_rate": 1.8857293048637012e-05,
+      "loss": 9.1935,
+      "step": 74690
+    },
+    {
+      "epoch": 0.3730430222976853,
+      "grad_norm": 0.09675570577383041,
+      "learning_rate": 1.8855791133695462e-05,
+      "loss": 9.1777,
+      "step": 74700
+    },
+    {
+      "epoch": 0.3730929611226248,
+      "grad_norm": 0.09218452125787735,
+      "learning_rate": 1.8854289218753912e-05,
+      "loss": 9.1843,
+      "step": 74710
+    },
+    {
+      "epoch": 0.3731428999475642,
+      "grad_norm": 0.09730592370033264,
+      "learning_rate": 1.885278730381236e-05,
+      "loss": 9.1803,
+      "step": 74720
+    },
+    {
+      "epoch": 0.3731928387725037,
+      "grad_norm": 0.09571095556020737,
+      "learning_rate": 1.8851285388870813e-05,
+      "loss": 9.1813,
+      "step": 74730
+    },
+    {
+      "epoch": 0.3732427775974431,
+      "grad_norm": 0.09244304150342941,
+      "learning_rate": 1.884978347392926e-05,
+      "loss": 9.1753,
+      "step": 74740
+    },
+    {
+      "epoch": 0.3732927164223826,
+      "grad_norm": 0.08950365334749222,
+      "learning_rate": 1.884828155898771e-05,
+      "loss": 9.1876,
+      "step": 74750
+    },
+    {
+      "epoch": 0.373342655247322,
+      "grad_norm": 0.0952899232506752,
+      "learning_rate": 1.884677964404616e-05,
+      "loss": 9.1857,
+      "step": 74760
+    },
+    {
+      "epoch": 0.3733925940722615,
+      "grad_norm": 0.0960029885172844,
+      "learning_rate": 1.8845277729104607e-05,
+      "loss": 9.1913,
+      "step": 74770
+    },
+    {
+      "epoch": 0.3734425328972009,
+      "grad_norm": 0.08860934525728226,
+      "learning_rate": 1.884377581416306e-05,
+      "loss": 9.1803,
+      "step": 74780
+    },
+    {
+      "epoch": 0.3734924717221404,
+      "grad_norm": 0.09400645643472672,
+      "learning_rate": 1.8842273899221507e-05,
+      "loss": 9.1803,
+      "step": 74790
+    },
+    {
+      "epoch": 0.3735424105470798,
+      "grad_norm": 0.09583377838134766,
+      "learning_rate": 1.8840771984279957e-05,
+      "loss": 9.1885,
+      "step": 74800
+    },
+    {
+      "epoch": 0.3735923493720193,
+      "grad_norm": 0.09530255943536758,
+      "learning_rate": 1.8839270069338407e-05,
+      "loss": 9.1873,
+      "step": 74810
+    },
+    {
+      "epoch": 0.3736422881969587,
+      "grad_norm": 0.09696003049612045,
+      "learning_rate": 1.8837768154396854e-05,
+      "loss": 9.1831,
+      "step": 74820
+    },
+    {
+      "epoch": 0.3736922270218982,
+      "grad_norm": 0.09576781839132309,
+      "learning_rate": 1.8836266239455308e-05,
+      "loss": 9.1916,
+      "step": 74830
+    },
+    {
+      "epoch": 0.3737421658468376,
+      "grad_norm": 0.08903126418590546,
+      "learning_rate": 1.8834764324513755e-05,
+      "loss": 9.1917,
+      "step": 74840
+    },
+    {
+      "epoch": 0.3737921046717771,
+      "grad_norm": 0.09006597101688385,
+      "learning_rate": 1.8833262409572205e-05,
+      "loss": 9.174,
+      "step": 74850
+    },
+    {
+      "epoch": 0.3738420434967165,
+      "grad_norm": 0.09339811652898788,
+      "learning_rate": 1.8831760494630655e-05,
+      "loss": 9.1918,
+      "step": 74860
+    },
+    {
+      "epoch": 0.373891982321656,
+      "grad_norm": 0.09909702837467194,
+      "learning_rate": 1.88302585796891e-05,
+      "loss": 9.1877,
+      "step": 74870
+    },
+    {
+      "epoch": 0.3739419211465954,
+      "grad_norm": 0.09441126883029938,
+      "learning_rate": 1.8828756664747555e-05,
+      "loss": 9.1836,
+      "step": 74880
+    },
+    {
+      "epoch": 0.3739918599715349,
+      "grad_norm": 0.09800399839878082,
+      "learning_rate": 1.8827254749806002e-05,
+      "loss": 9.1844,
+      "step": 74890
+    },
+    {
+      "epoch": 0.3740417987964743,
+      "grad_norm": 0.10018418729305267,
+      "learning_rate": 1.8825752834864452e-05,
+      "loss": 9.1881,
+      "step": 74900
+    },
+    {
+      "epoch": 0.3740917376214138,
+      "grad_norm": 0.0955013558268547,
+      "learning_rate": 1.8824250919922902e-05,
+      "loss": 9.1919,
+      "step": 74910
+    },
+    {
+      "epoch": 0.3741416764463532,
+      "grad_norm": 0.09473446756601334,
+      "learning_rate": 1.8822749004981353e-05,
+      "loss": 9.191,
+      "step": 74920
+    },
+    {
+      "epoch": 0.3741916152712927,
+      "grad_norm": 0.09358023852109909,
+      "learning_rate": 1.8821247090039803e-05,
+      "loss": 9.1891,
+      "step": 74930
+    },
+    {
+      "epoch": 0.3742415540962321,
+      "grad_norm": 0.09420783072710037,
+      "learning_rate": 1.881974517509825e-05,
+      "loss": 9.1874,
+      "step": 74940
+    },
+    {
+      "epoch": 0.3742914929211716,
+      "grad_norm": 0.09367396682500839,
+      "learning_rate": 1.88182432601567e-05,
+      "loss": 9.1976,
+      "step": 74950
+    },
+    {
+      "epoch": 0.374341431746111,
+      "grad_norm": 0.09339455515146255,
+      "learning_rate": 1.881674134521515e-05,
+      "loss": 9.1843,
+      "step": 74960
+    },
+    {
+      "epoch": 0.3743913705710505,
+      "grad_norm": 0.09701774269342422,
+      "learning_rate": 1.88152394302736e-05,
+      "loss": 9.1973,
+      "step": 74970
+    },
+    {
+      "epoch": 0.3744413093959899,
+      "grad_norm": 0.09740804880857468,
+      "learning_rate": 1.881373751533205e-05,
+      "loss": 9.1823,
+      "step": 74980
+    },
+    {
+      "epoch": 0.3744912482209294,
+      "grad_norm": 0.09477312117815018,
+      "learning_rate": 1.8812235600390497e-05,
+      "loss": 9.1887,
+      "step": 74990
+    },
+    {
+      "epoch": 0.3745411870458688,
+      "grad_norm": 0.09397511929273605,
+      "learning_rate": 1.8810733685448947e-05,
+      "loss": 9.178,
+      "step": 75000
+    },
+    {
+      "epoch": 0.3745911258708083,
+      "grad_norm": 0.093665711581707,
+      "learning_rate": 1.8809231770507397e-05,
+      "loss": 9.1876,
+      "step": 75010
+    },
+    {
+      "epoch": 0.3746410646957477,
+      "grad_norm": 0.09372160583734512,
+      "learning_rate": 1.8807729855565848e-05,
+      "loss": 9.1882,
+      "step": 75020
+    },
+    {
+      "epoch": 0.3746910035206872,
+      "grad_norm": 0.09285315126180649,
+      "learning_rate": 1.8806227940624298e-05,
+      "loss": 9.1869,
+      "step": 75030
+    },
+    {
+      "epoch": 0.3747409423456266,
+      "grad_norm": 0.10530349612236023,
+      "learning_rate": 1.8804726025682745e-05,
+      "loss": 9.1785,
+      "step": 75040
+    },
+    {
+      "epoch": 0.3747908811705661,
+      "grad_norm": 0.09486869722604752,
+      "learning_rate": 1.8803224110741195e-05,
+      "loss": 9.1772,
+      "step": 75050
+    },
+    {
+      "epoch": 0.3748408199955055,
+      "grad_norm": 0.09095942974090576,
+      "learning_rate": 1.8801722195799645e-05,
+      "loss": 9.1886,
+      "step": 75060
+    },
+    {
+      "epoch": 0.374890758820445,
+      "grad_norm": 0.09552841633558273,
+      "learning_rate": 1.8800220280858095e-05,
+      "loss": 9.1746,
+      "step": 75070
+    },
+    {
+      "epoch": 0.3749406976453844,
+      "grad_norm": 0.09439145773649216,
+      "learning_rate": 1.8798718365916545e-05,
+      "loss": 9.1689,
+      "step": 75080
+    },
+    {
+      "epoch": 0.3749906364703239,
+      "grad_norm": 0.0905107706785202,
+      "learning_rate": 1.8797216450974992e-05,
+      "loss": 9.1867,
+      "step": 75090
+    },
+    {
+      "epoch": 0.3750405752952633,
+      "grad_norm": 0.0866839811205864,
+      "learning_rate": 1.8795714536033442e-05,
+      "loss": 9.1908,
+      "step": 75100
+    },
+    {
+      "epoch": 0.3750905141202028,
+      "grad_norm": 0.09543773531913757,
+      "learning_rate": 1.8794212621091892e-05,
+      "loss": 9.1941,
+      "step": 75110
+    },
+    {
+      "epoch": 0.3751404529451422,
+      "grad_norm": 0.09861119091510773,
+      "learning_rate": 1.8792710706150343e-05,
+      "loss": 9.186,
+      "step": 75120
+    },
+    {
+      "epoch": 0.3751903917700817,
+      "grad_norm": 0.09284248948097229,
+      "learning_rate": 1.8791208791208793e-05,
+      "loss": 9.1845,
+      "step": 75130
+    },
+    {
+      "epoch": 0.3752403305950211,
+      "grad_norm": 0.0985037088394165,
+      "learning_rate": 1.878970687626724e-05,
+      "loss": 9.172,
+      "step": 75140
+    },
+    {
+      "epoch": 0.3752902694199606,
+      "grad_norm": 0.09096460044384003,
+      "learning_rate": 1.878820496132569e-05,
+      "loss": 9.1894,
+      "step": 75150
+    },
+    {
+      "epoch": 0.3753402082449,
+      "grad_norm": 0.09545445442199707,
+      "learning_rate": 1.878670304638414e-05,
+      "loss": 9.1824,
+      "step": 75160
+    },
+    {
+      "epoch": 0.3753901470698394,
+      "grad_norm": 0.08448511362075806,
+      "learning_rate": 1.878520113144259e-05,
+      "loss": 9.1783,
+      "step": 75170
+    },
+    {
+      "epoch": 0.3754400858947789,
+      "grad_norm": 0.09297330677509308,
+      "learning_rate": 1.878369921650104e-05,
+      "loss": 9.1779,
+      "step": 75180
+    },
+    {
+      "epoch": 0.3754900247197183,
+      "grad_norm": 0.09360940754413605,
+      "learning_rate": 1.8782197301559487e-05,
+      "loss": 9.1796,
+      "step": 75190
+    },
+    {
+      "epoch": 0.3755399635446578,
+      "grad_norm": 0.09942979365587234,
+      "learning_rate": 1.8780695386617937e-05,
+      "loss": 9.1824,
+      "step": 75200
+    },
+    {
+      "epoch": 0.3755899023695972,
+      "grad_norm": 0.08949796110391617,
+      "learning_rate": 1.8779193471676387e-05,
+      "loss": 9.1709,
+      "step": 75210
+    },
+    {
+      "epoch": 0.3756398411945367,
+      "grad_norm": 0.09671638906002045,
+      "learning_rate": 1.8777691556734838e-05,
+      "loss": 9.1738,
+      "step": 75220
+    },
+    {
+      "epoch": 0.3756897800194761,
+      "grad_norm": 0.09702275693416595,
+      "learning_rate": 1.8776189641793288e-05,
+      "loss": 9.19,
+      "step": 75230
+    },
+    {
+      "epoch": 0.3757397188444156,
+      "grad_norm": 0.09984516352415085,
+      "learning_rate": 1.8774687726851738e-05,
+      "loss": 9.1752,
+      "step": 75240
+    },
+    {
+      "epoch": 0.375789657669355,
+      "grad_norm": 0.08986981213092804,
+      "learning_rate": 1.8773185811910185e-05,
+      "loss": 9.1787,
+      "step": 75250
+    },
+    {
+      "epoch": 0.3758395964942945,
+      "grad_norm": 0.09705835580825806,
+      "learning_rate": 1.8771683896968635e-05,
+      "loss": 9.1834,
+      "step": 75260
+    },
+    {
+      "epoch": 0.3758895353192339,
+      "grad_norm": 0.09416690468788147,
+      "learning_rate": 1.8770181982027085e-05,
+      "loss": 9.1762,
+      "step": 75270
+    },
+    {
+      "epoch": 0.3759394741441734,
+      "grad_norm": 0.09269827604293823,
+      "learning_rate": 1.8768680067085535e-05,
+      "loss": 9.173,
+      "step": 75280
+    },
+    {
+      "epoch": 0.3759894129691128,
+      "grad_norm": 0.08903074264526367,
+      "learning_rate": 1.8767178152143985e-05,
+      "loss": 9.183,
+      "step": 75290
+    },
+    {
+      "epoch": 0.3760393517940523,
+      "grad_norm": 0.09470517188310623,
+      "learning_rate": 1.8765676237202432e-05,
+      "loss": 9.1779,
+      "step": 75300
+    },
+    {
+      "epoch": 0.3760892906189917,
+      "grad_norm": 0.09774582087993622,
+      "learning_rate": 1.8764174322260882e-05,
+      "loss": 9.1639,
+      "step": 75310
+    },
+    {
+      "epoch": 0.3761392294439312,
+      "grad_norm": 0.0924445167183876,
+      "learning_rate": 1.8762672407319333e-05,
+      "loss": 9.1775,
+      "step": 75320
+    },
+    {
+      "epoch": 0.3761891682688706,
+      "grad_norm": 0.09741353988647461,
+      "learning_rate": 1.8761170492377783e-05,
+      "loss": 9.1726,
+      "step": 75330
+    },
+    {
+      "epoch": 0.3762391070938101,
+      "grad_norm": 0.0985332578420639,
+      "learning_rate": 1.8759668577436233e-05,
+      "loss": 9.1825,
+      "step": 75340
+    },
+    {
+      "epoch": 0.3762890459187495,
+      "grad_norm": 0.09728459268808365,
+      "learning_rate": 1.875816666249468e-05,
+      "loss": 9.1818,
+      "step": 75350
+    },
+    {
+      "epoch": 0.376338984743689,
+      "grad_norm": 0.09466975927352905,
+      "learning_rate": 1.875666474755313e-05,
+      "loss": 9.1808,
+      "step": 75360
+    },
+    {
+      "epoch": 0.3763889235686284,
+      "grad_norm": 0.09836606681346893,
+      "learning_rate": 1.875516283261158e-05,
+      "loss": 9.1755,
+      "step": 75370
+    },
+    {
+      "epoch": 0.3764388623935679,
+      "grad_norm": 0.09210119396448135,
+      "learning_rate": 1.875366091767003e-05,
+      "loss": 9.1742,
+      "step": 75380
+    },
+    {
+      "epoch": 0.3764888012185073,
+      "grad_norm": 0.09598752856254578,
+      "learning_rate": 1.875215900272848e-05,
+      "loss": 9.1785,
+      "step": 75390
+    },
+    {
+      "epoch": 0.3765387400434468,
+      "grad_norm": 0.09594704955816269,
+      "learning_rate": 1.8750657087786927e-05,
+      "loss": 9.17,
+      "step": 75400
+    },
+    {
+      "epoch": 0.3765886788683862,
+      "grad_norm": 0.09366414695978165,
+      "learning_rate": 1.8749155172845377e-05,
+      "loss": 9.1825,
+      "step": 75410
+    },
+    {
+      "epoch": 0.3766386176933257,
+      "grad_norm": 0.09399878233671188,
+      "learning_rate": 1.8747653257903828e-05,
+      "loss": 9.1663,
+      "step": 75420
+    },
+    {
+      "epoch": 0.3766885565182651,
+      "grad_norm": 0.09145677834749222,
+      "learning_rate": 1.8746151342962278e-05,
+      "loss": 9.1777,
+      "step": 75430
+    },
+    {
+      "epoch": 0.3767384953432046,
+      "grad_norm": 0.0970328152179718,
+      "learning_rate": 1.8744649428020728e-05,
+      "loss": 9.1692,
+      "step": 75440
+    },
+    {
+      "epoch": 0.376788434168144,
+      "grad_norm": 0.09038615226745605,
+      "learning_rate": 1.8743147513079175e-05,
+      "loss": 9.1891,
+      "step": 75450
+    },
+    {
+      "epoch": 0.3768383729930835,
+      "grad_norm": 0.09302138537168503,
+      "learning_rate": 1.8741645598137625e-05,
+      "loss": 9.1833,
+      "step": 75460
+    },
+    {
+      "epoch": 0.3768883118180229,
+      "grad_norm": 0.09453801065683365,
+      "learning_rate": 1.8740143683196075e-05,
+      "loss": 9.177,
+      "step": 75470
+    },
+    {
+      "epoch": 0.3769382506429624,
+      "grad_norm": 0.09908052533864975,
+      "learning_rate": 1.8738641768254525e-05,
+      "loss": 9.182,
+      "step": 75480
+    },
+    {
+      "epoch": 0.3769881894679018,
+      "grad_norm": 0.08896263688802719,
+      "learning_rate": 1.8737139853312975e-05,
+      "loss": 9.1729,
+      "step": 75490
+    },
+    {
+      "epoch": 0.3770381282928413,
+      "grad_norm": 0.09187068045139313,
+      "learning_rate": 1.8735637938371422e-05,
+      "loss": 9.1645,
+      "step": 75500
+    },
+    {
+      "epoch": 0.3770880671177807,
+      "grad_norm": 0.09145993739366531,
+      "learning_rate": 1.8734136023429872e-05,
+      "loss": 9.1855,
+      "step": 75510
+    },
+    {
+      "epoch": 0.3771380059427202,
+      "grad_norm": 0.09123537689447403,
+      "learning_rate": 1.8732634108488323e-05,
+      "loss": 9.1817,
+      "step": 75520
+    },
+    {
+      "epoch": 0.3771879447676596,
+      "grad_norm": 0.09334246069192886,
+      "learning_rate": 1.8731132193546773e-05,
+      "loss": 9.1796,
+      "step": 75530
+    },
+    {
+      "epoch": 0.3772378835925991,
+      "grad_norm": 0.09891509264707565,
+      "learning_rate": 1.8729630278605223e-05,
+      "loss": 9.1878,
+      "step": 75540
+    },
+    {
+      "epoch": 0.3772878224175385,
+      "grad_norm": 0.0893426239490509,
+      "learning_rate": 1.872812836366367e-05,
+      "loss": 9.1732,
+      "step": 75550
+    },
+    {
+      "epoch": 0.377337761242478,
+      "grad_norm": 0.09921654313802719,
+      "learning_rate": 1.8726626448722123e-05,
+      "loss": 9.1717,
+      "step": 75560
+    },
+    {
+      "epoch": 0.3773877000674174,
+      "grad_norm": 0.0956064984202385,
+      "learning_rate": 1.872512453378057e-05,
+      "loss": 9.1776,
+      "step": 75570
+    },
+    {
+      "epoch": 0.3774376388923569,
+      "grad_norm": 0.09236710518598557,
+      "learning_rate": 1.872362261883902e-05,
+      "loss": 9.1823,
+      "step": 75580
+    },
+    {
+      "epoch": 0.3774875777172963,
+      "grad_norm": 0.09194951504468918,
+      "learning_rate": 1.872212070389747e-05,
+      "loss": 9.1675,
+      "step": 75590
+    },
+    {
+      "epoch": 0.3775375165422358,
+      "grad_norm": 0.09250642359256744,
+      "learning_rate": 1.8720618788955917e-05,
+      "loss": 9.1711,
+      "step": 75600
+    },
+    {
+      "epoch": 0.3775874553671752,
+      "grad_norm": 0.09264052659273148,
+      "learning_rate": 1.871911687401437e-05,
+      "loss": 9.1799,
+      "step": 75610
+    },
+    {
+      "epoch": 0.3776373941921147,
+      "grad_norm": 0.09335634112358093,
+      "learning_rate": 1.8717614959072818e-05,
+      "loss": 9.1752,
+      "step": 75620
+    },
+    {
+      "epoch": 0.3776873330170541,
+      "grad_norm": 0.10500754415988922,
+      "learning_rate": 1.8716113044131268e-05,
+      "loss": 9.1719,
+      "step": 75630
+    },
+    {
+      "epoch": 0.37773727184199357,
+      "grad_norm": 0.09396055340766907,
+      "learning_rate": 1.8714611129189718e-05,
+      "loss": 9.1832,
+      "step": 75640
+    },
+    {
+      "epoch": 0.377787210666933,
+      "grad_norm": 0.09786514937877655,
+      "learning_rate": 1.8713109214248165e-05,
+      "loss": 9.1689,
+      "step": 75650
+    },
+    {
+      "epoch": 0.37783714949187247,
+      "grad_norm": 0.09014048427343369,
+      "learning_rate": 1.871160729930662e-05,
+      "loss": 9.1699,
+      "step": 75660
+    },
+    {
+      "epoch": 0.3778870883168119,
+      "grad_norm": 0.09452909231185913,
+      "learning_rate": 1.8710105384365065e-05,
+      "loss": 9.1733,
+      "step": 75670
+    },
+    {
+      "epoch": 0.37793702714175137,
+      "grad_norm": 0.09486046433448792,
+      "learning_rate": 1.8708603469423515e-05,
+      "loss": 9.1663,
+      "step": 75680
+    },
+    {
+      "epoch": 0.3779869659666908,
+      "grad_norm": 0.09308832138776779,
+      "learning_rate": 1.8707101554481965e-05,
+      "loss": 9.1795,
+      "step": 75690
+    },
+    {
+      "epoch": 0.37803690479163027,
+      "grad_norm": 0.09890872985124588,
+      "learning_rate": 1.8705599639540412e-05,
+      "loss": 9.1685,
+      "step": 75700
+    },
+    {
+      "epoch": 0.3780868436165697,
+      "grad_norm": 0.09047377854585648,
+      "learning_rate": 1.8704097724598866e-05,
+      "loss": 9.1856,
+      "step": 75710
+    },
+    {
+      "epoch": 0.37813678244150917,
+      "grad_norm": 0.09121832996606827,
+      "learning_rate": 1.8702595809657313e-05,
+      "loss": 9.1825,
+      "step": 75720
+    },
+    {
+      "epoch": 0.3781867212664486,
+      "grad_norm": 0.09841986000537872,
+      "learning_rate": 1.8701093894715763e-05,
+      "loss": 9.1776,
+      "step": 75730
+    },
+    {
+      "epoch": 0.37823666009138807,
+      "grad_norm": 0.09331878274679184,
+      "learning_rate": 1.8699591979774213e-05,
+      "loss": 9.1805,
+      "step": 75740
+    },
+    {
+      "epoch": 0.3782865989163275,
+      "grad_norm": 0.09422630816698074,
+      "learning_rate": 1.869809006483266e-05,
+      "loss": 9.1732,
+      "step": 75750
+    },
+    {
+      "epoch": 0.37833653774126697,
+      "grad_norm": 0.09663139283657074,
+      "learning_rate": 1.8696588149891113e-05,
+      "loss": 9.1744,
+      "step": 75760
+    },
+    {
+      "epoch": 0.3783864765662064,
+      "grad_norm": 0.0949576273560524,
+      "learning_rate": 1.869508623494956e-05,
+      "loss": 9.1771,
+      "step": 75770
+    },
+    {
+      "epoch": 0.37843641539114586,
+      "grad_norm": 0.09695305675268173,
+      "learning_rate": 1.869358432000801e-05,
+      "loss": 9.1587,
+      "step": 75780
+    },
+    {
+      "epoch": 0.3784863542160853,
+      "grad_norm": 0.09372842311859131,
+      "learning_rate": 1.869208240506646e-05,
+      "loss": 9.1616,
+      "step": 75790
+    },
+    {
+      "epoch": 0.37853629304102476,
+      "grad_norm": 0.09371675550937653,
+      "learning_rate": 1.8690580490124907e-05,
+      "loss": 9.1689,
+      "step": 75800
+    },
+    {
+      "epoch": 0.3785862318659642,
+      "grad_norm": 0.09579101204872131,
+      "learning_rate": 1.868907857518336e-05,
+      "loss": 9.1754,
+      "step": 75810
+    },
+    {
+      "epoch": 0.37863617069090366,
+      "grad_norm": 0.09259112924337387,
+      "learning_rate": 1.8687576660241808e-05,
+      "loss": 9.1816,
+      "step": 75820
+    },
+    {
+      "epoch": 0.3786861095158431,
+      "grad_norm": 0.09288639575242996,
+      "learning_rate": 1.8686074745300258e-05,
+      "loss": 9.1795,
+      "step": 75830
+    },
+    {
+      "epoch": 0.37873604834078256,
+      "grad_norm": 0.09235890954732895,
+      "learning_rate": 1.8684572830358708e-05,
+      "loss": 9.1722,
+      "step": 75840
+    },
+    {
+      "epoch": 0.378785987165722,
+      "grad_norm": 0.0942944809794426,
+      "learning_rate": 1.8683070915417155e-05,
+      "loss": 9.1728,
+      "step": 75850
+    },
+    {
+      "epoch": 0.37883592599066146,
+      "grad_norm": 0.09688469022512436,
+      "learning_rate": 1.868156900047561e-05,
+      "loss": 9.1777,
+      "step": 75860
+    },
+    {
+      "epoch": 0.3788858648156009,
+      "grad_norm": 0.08886689692735672,
+      "learning_rate": 1.8680067085534055e-05,
+      "loss": 9.1681,
+      "step": 75870
+    },
+    {
+      "epoch": 0.37893580364054036,
+      "grad_norm": 0.09597429633140564,
+      "learning_rate": 1.867856517059251e-05,
+      "loss": 9.1737,
+      "step": 75880
+    },
+    {
+      "epoch": 0.3789857424654798,
+      "grad_norm": 0.08873099833726883,
+      "learning_rate": 1.8677063255650955e-05,
+      "loss": 9.1717,
+      "step": 75890
+    },
+    {
+      "epoch": 0.37903568129041926,
+      "grad_norm": 0.08912356197834015,
+      "learning_rate": 1.8675561340709402e-05,
+      "loss": 9.1843,
+      "step": 75900
+    },
+    {
+      "epoch": 0.3790856201153587,
+      "grad_norm": 0.09434056282043457,
+      "learning_rate": 1.8674059425767856e-05,
+      "loss": 9.1582,
+      "step": 75910
+    },
+    {
+      "epoch": 0.37913555894029816,
+      "grad_norm": 0.08778839558362961,
+      "learning_rate": 1.8672557510826303e-05,
+      "loss": 9.1786,
+      "step": 75920
+    },
+    {
+      "epoch": 0.3791854977652376,
+      "grad_norm": 0.09374192357063293,
+      "learning_rate": 1.8671055595884756e-05,
+      "loss": 9.1732,
+      "step": 75930
+    },
+    {
+      "epoch": 0.37923543659017706,
+      "grad_norm": 0.0933762714266777,
+      "learning_rate": 1.8669553680943203e-05,
+      "loss": 9.1647,
+      "step": 75940
+    },
+    {
+      "epoch": 0.3792853754151165,
+      "grad_norm": 0.09484566748142242,
+      "learning_rate": 1.866805176600165e-05,
+      "loss": 9.1594,
+      "step": 75950
+    },
+    {
+      "epoch": 0.37933531424005595,
+      "grad_norm": 0.09006204456090927,
+      "learning_rate": 1.8666549851060103e-05,
+      "loss": 9.1784,
+      "step": 75960
+    },
+    {
+      "epoch": 0.3793852530649954,
+      "grad_norm": 0.09235017746686935,
+      "learning_rate": 1.866504793611855e-05,
+      "loss": 9.164,
+      "step": 75970
+    },
+    {
+      "epoch": 0.37943519188993485,
+      "grad_norm": 0.09987694025039673,
+      "learning_rate": 1.8663546021177004e-05,
+      "loss": 9.1695,
+      "step": 75980
+    },
+    {
+      "epoch": 0.3794851307148743,
+      "grad_norm": 0.08617845177650452,
+      "learning_rate": 1.866204410623545e-05,
+      "loss": 9.182,
+      "step": 75990
+    },
+    {
+      "epoch": 0.37953506953981375,
+      "grad_norm": 0.08854833245277405,
+      "learning_rate": 1.8660542191293897e-05,
+      "loss": 9.1611,
+      "step": 76000
+    },
+    {
+      "epoch": 0.3795850083647532,
+      "grad_norm": 0.09249111264944077,
+      "learning_rate": 1.865904027635235e-05,
+      "loss": 9.1808,
+      "step": 76010
+    },
+    {
+      "epoch": 0.37963494718969265,
+      "grad_norm": 0.09673107415437698,
+      "learning_rate": 1.8657538361410798e-05,
+      "loss": 9.1706,
+      "step": 76020
+    },
+    {
+      "epoch": 0.3796848860146321,
+      "grad_norm": 0.09875845909118652,
+      "learning_rate": 1.865603644646925e-05,
+      "loss": 9.1723,
+      "step": 76030
+    },
+    {
+      "epoch": 0.37973482483957155,
+      "grad_norm": 0.09128276258707047,
+      "learning_rate": 1.8654534531527698e-05,
+      "loss": 9.1666,
+      "step": 76040
+    },
+    {
+      "epoch": 0.37978476366451097,
+      "grad_norm": 0.09420790523290634,
+      "learning_rate": 1.8653032616586145e-05,
+      "loss": 9.163,
+      "step": 76050
+    },
+    {
+      "epoch": 0.37983470248945045,
+      "grad_norm": 0.09466836601495743,
+      "learning_rate": 1.86515307016446e-05,
+      "loss": 9.1646,
+      "step": 76060
+    },
+    {
+      "epoch": 0.37988464131438987,
+      "grad_norm": 0.10341818630695343,
+      "learning_rate": 1.8650028786703045e-05,
+      "loss": 9.1677,
+      "step": 76070
+    },
+    {
+      "epoch": 0.37993458013932935,
+      "grad_norm": 0.0988142341375351,
+      "learning_rate": 1.86485268717615e-05,
+      "loss": 9.1715,
+      "step": 76080
+    },
+    {
+      "epoch": 0.37998451896426877,
+      "grad_norm": 0.09664300084114075,
+      "learning_rate": 1.8647024956819946e-05,
+      "loss": 9.1588,
+      "step": 76090
+    },
+    {
+      "epoch": 0.38003445778920825,
+      "grad_norm": 0.09185276925563812,
+      "learning_rate": 1.8645523041878392e-05,
+      "loss": 9.1747,
+      "step": 76100
+    },
+    {
+      "epoch": 0.38008439661414767,
+      "grad_norm": 0.09593552350997925,
+      "learning_rate": 1.8644021126936846e-05,
+      "loss": 9.1682,
+      "step": 76110
+    },
+    {
+      "epoch": 0.38013433543908715,
+      "grad_norm": 0.09431128203868866,
+      "learning_rate": 1.8642519211995293e-05,
+      "loss": 9.1772,
+      "step": 76120
+    },
+    {
+      "epoch": 0.38018427426402657,
+      "grad_norm": 0.09099570661783218,
+      "learning_rate": 1.8641017297053746e-05,
+      "loss": 9.1571,
+      "step": 76130
+    },
+    {
+      "epoch": 0.38023421308896604,
+      "grad_norm": 0.08938246220350266,
+      "learning_rate": 1.8639515382112193e-05,
+      "loss": 9.1721,
+      "step": 76140
+    },
+    {
+      "epoch": 0.38028415191390547,
+      "grad_norm": 0.09171757102012634,
+      "learning_rate": 1.863801346717064e-05,
+      "loss": 9.1665,
+      "step": 76150
+    },
+    {
+      "epoch": 0.3803340907388449,
+      "grad_norm": 0.09394662827253342,
+      "learning_rate": 1.8636511552229093e-05,
+      "loss": 9.1786,
+      "step": 76160
+    },
+    {
+      "epoch": 0.38038402956378436,
+      "grad_norm": 0.0892520397901535,
+      "learning_rate": 1.863500963728754e-05,
+      "loss": 9.1615,
+      "step": 76170
+    },
+    {
+      "epoch": 0.3804339683887238,
+      "grad_norm": 0.09431727975606918,
+      "learning_rate": 1.8633507722345994e-05,
+      "loss": 9.1807,
+      "step": 76180
+    },
+    {
+      "epoch": 0.38048390721366326,
+      "grad_norm": 0.09187456965446472,
+      "learning_rate": 1.863200580740444e-05,
+      "loss": 9.1699,
+      "step": 76190
+    },
+    {
+      "epoch": 0.3805338460386027,
+      "grad_norm": 0.09355001896619797,
+      "learning_rate": 1.863050389246289e-05,
+      "loss": 9.1677,
+      "step": 76200
+    },
+    {
+      "epoch": 0.38058378486354216,
+      "grad_norm": 0.09874553978443146,
+      "learning_rate": 1.862900197752134e-05,
+      "loss": 9.1619,
+      "step": 76210
+    },
+    {
+      "epoch": 0.3806337236884816,
+      "grad_norm": 0.0971277505159378,
+      "learning_rate": 1.8627500062579788e-05,
+      "loss": 9.157,
+      "step": 76220
+    },
+    {
+      "epoch": 0.38068366251342106,
+      "grad_norm": 0.09368085116147995,
+      "learning_rate": 1.862599814763824e-05,
+      "loss": 9.1691,
+      "step": 76230
+    },
+    {
+      "epoch": 0.3807336013383605,
+      "grad_norm": 0.08936905115842819,
+      "learning_rate": 1.8624496232696688e-05,
+      "loss": 9.1715,
+      "step": 76240
+    },
+    {
+      "epoch": 0.38078354016329996,
+      "grad_norm": 0.09351339191198349,
+      "learning_rate": 1.8622994317755138e-05,
+      "loss": 9.1657,
+      "step": 76250
+    },
+    {
+      "epoch": 0.3808334789882394,
+      "grad_norm": 0.09461808949708939,
+      "learning_rate": 1.862149240281359e-05,
+      "loss": 9.1624,
+      "step": 76260
+    },
+    {
+      "epoch": 0.38088341781317886,
+      "grad_norm": 0.08949335664510727,
+      "learning_rate": 1.8619990487872035e-05,
+      "loss": 9.1684,
+      "step": 76270
+    },
+    {
+      "epoch": 0.3809333566381183,
+      "grad_norm": 0.08932676911354065,
+      "learning_rate": 1.861848857293049e-05,
+      "loss": 9.1631,
+      "step": 76280
+    },
+    {
+      "epoch": 0.38098329546305776,
+      "grad_norm": 0.09498759359121323,
+      "learning_rate": 1.8616986657988936e-05,
+      "loss": 9.163,
+      "step": 76290
+    },
+    {
+      "epoch": 0.3810332342879972,
+      "grad_norm": 0.09363845735788345,
+      "learning_rate": 1.8615484743047386e-05,
+      "loss": 9.174,
+      "step": 76300
+    },
+    {
+      "epoch": 0.38108317311293666,
+      "grad_norm": 0.08573130518198013,
+      "learning_rate": 1.8613982828105836e-05,
+      "loss": 9.1665,
+      "step": 76310
+    },
+    {
+      "epoch": 0.3811331119378761,
+      "grad_norm": 0.09437863528728485,
+      "learning_rate": 1.8612480913164283e-05,
+      "loss": 9.1553,
+      "step": 76320
+    },
+    {
+      "epoch": 0.38118305076281556,
+      "grad_norm": 0.09514939785003662,
+      "learning_rate": 1.8610978998222736e-05,
+      "loss": 9.161,
+      "step": 76330
+    },
+    {
+      "epoch": 0.381232989587755,
+      "grad_norm": 0.0890861302614212,
+      "learning_rate": 1.8609477083281183e-05,
+      "loss": 9.1585,
+      "step": 76340
+    },
+    {
+      "epoch": 0.38128292841269446,
+      "grad_norm": 0.09901072829961777,
+      "learning_rate": 1.8607975168339633e-05,
+      "loss": 9.1656,
+      "step": 76350
+    },
+    {
+      "epoch": 0.3813328672376339,
+      "grad_norm": 0.09135191142559052,
+      "learning_rate": 1.8606473253398083e-05,
+      "loss": 9.1679,
+      "step": 76360
+    },
+    {
+      "epoch": 0.38138280606257335,
+      "grad_norm": 0.08994290977716446,
+      "learning_rate": 1.860497133845653e-05,
+      "loss": 9.1553,
+      "step": 76370
+    },
+    {
+      "epoch": 0.3814327448875128,
+      "grad_norm": 0.09007209539413452,
+      "learning_rate": 1.8603469423514984e-05,
+      "loss": 9.1617,
+      "step": 76380
+    },
+    {
+      "epoch": 0.38148268371245225,
+      "grad_norm": 0.09258635342121124,
+      "learning_rate": 1.860196750857343e-05,
+      "loss": 9.1664,
+      "step": 76390
+    },
+    {
+      "epoch": 0.3815326225373917,
+      "grad_norm": 0.09930216521024704,
+      "learning_rate": 1.860046559363188e-05,
+      "loss": 9.1606,
+      "step": 76400
+    },
+    {
+      "epoch": 0.38158256136233115,
+      "grad_norm": 0.10146905481815338,
+      "learning_rate": 1.859896367869033e-05,
+      "loss": 9.1603,
+      "step": 76410
+    },
+    {
+      "epoch": 0.3816325001872706,
+      "grad_norm": 0.09039684385061264,
+      "learning_rate": 1.8597461763748778e-05,
+      "loss": 9.1433,
+      "step": 76420
+    },
+    {
+      "epoch": 0.38168243901221005,
+      "grad_norm": 0.09020677208900452,
+      "learning_rate": 1.859595984880723e-05,
+      "loss": 9.1615,
+      "step": 76430
+    },
+    {
+      "epoch": 0.3817323778371495,
+      "grad_norm": 0.0926944836974144,
+      "learning_rate": 1.8594457933865678e-05,
+      "loss": 9.1614,
+      "step": 76440
+    },
+    {
+      "epoch": 0.38178231666208895,
+      "grad_norm": 0.08853311091661453,
+      "learning_rate": 1.8592956018924128e-05,
+      "loss": 9.1708,
+      "step": 76450
+    },
+    {
+      "epoch": 0.38183225548702837,
+      "grad_norm": 0.09793388843536377,
+      "learning_rate": 1.859145410398258e-05,
+      "loss": 9.1545,
+      "step": 76460
+    },
+    {
+      "epoch": 0.38188219431196785,
+      "grad_norm": 0.08943363279104233,
+      "learning_rate": 1.8589952189041025e-05,
+      "loss": 9.1679,
+      "step": 76470
+    },
+    {
+      "epoch": 0.38193213313690727,
+      "grad_norm": 0.09909660369157791,
+      "learning_rate": 1.858845027409948e-05,
+      "loss": 9.1526,
+      "step": 76480
+    },
+    {
+      "epoch": 0.38198207196184675,
+      "grad_norm": 0.09491763263940811,
+      "learning_rate": 1.8586948359157926e-05,
+      "loss": 9.1674,
+      "step": 76490
+    },
+    {
+      "epoch": 0.38203201078678617,
+      "grad_norm": 0.0981326624751091,
+      "learning_rate": 1.8585446444216376e-05,
+      "loss": 9.1596,
+      "step": 76500
+    },
+    {
+      "epoch": 0.38208194961172565,
+      "grad_norm": 0.09292647987604141,
+      "learning_rate": 1.8583944529274826e-05,
+      "loss": 9.1512,
+      "step": 76510
+    },
+    {
+      "epoch": 0.38213188843666507,
+      "grad_norm": 0.0934506356716156,
+      "learning_rate": 1.8582442614333273e-05,
+      "loss": 9.1633,
+      "step": 76520
+    },
+    {
+      "epoch": 0.38218182726160455,
+      "grad_norm": 0.09270615875720978,
+      "learning_rate": 1.8580940699391726e-05,
+      "loss": 9.155,
+      "step": 76530
+    },
+    {
+      "epoch": 0.38223176608654397,
+      "grad_norm": 0.09436081349849701,
+      "learning_rate": 1.8579438784450173e-05,
+      "loss": 9.1638,
+      "step": 76540
+    },
+    {
+      "epoch": 0.38228170491148344,
+      "grad_norm": 0.09928996115922928,
+      "learning_rate": 1.8577936869508623e-05,
+      "loss": 9.1611,
+      "step": 76550
+    },
+    {
+      "epoch": 0.38233164373642287,
+      "grad_norm": 0.09387561678886414,
+      "learning_rate": 1.8576434954567073e-05,
+      "loss": 9.1742,
+      "step": 76560
+    },
+    {
+      "epoch": 0.38238158256136234,
+      "grad_norm": 0.09134583920240402,
+      "learning_rate": 1.8574933039625524e-05,
+      "loss": 9.1692,
+      "step": 76570
+    },
+    {
+      "epoch": 0.38243152138630176,
+      "grad_norm": 0.09804858267307281,
+      "learning_rate": 1.8573431124683974e-05,
+      "loss": 9.1597,
+      "step": 76580
+    },
+    {
+      "epoch": 0.38248146021124124,
+      "grad_norm": 0.09265462309122086,
+      "learning_rate": 1.857192920974242e-05,
+      "loss": 9.1541,
+      "step": 76590
+    },
+    {
+      "epoch": 0.38253139903618066,
+      "grad_norm": 0.0962829664349556,
+      "learning_rate": 1.8570427294800874e-05,
+      "loss": 9.1548,
+      "step": 76600
+    },
+    {
+      "epoch": 0.38258133786112014,
+      "grad_norm": 0.10263549536466599,
+      "learning_rate": 1.856892537985932e-05,
+      "loss": 9.1655,
+      "step": 76610
+    },
+    {
+      "epoch": 0.38263127668605956,
+      "grad_norm": 0.0963212177157402,
+      "learning_rate": 1.856742346491777e-05,
+      "loss": 9.161,
+      "step": 76620
+    },
+    {
+      "epoch": 0.38268121551099904,
+      "grad_norm": 0.08831122517585754,
+      "learning_rate": 1.856592154997622e-05,
+      "loss": 9.1554,
+      "step": 76630
+    },
+    {
+      "epoch": 0.38273115433593846,
+      "grad_norm": 0.0948970839381218,
+      "learning_rate": 1.8564419635034668e-05,
+      "loss": 9.1628,
+      "step": 76640
+    },
+    {
+      "epoch": 0.38278109316087794,
+      "grad_norm": 0.09297934174537659,
+      "learning_rate": 1.856291772009312e-05,
+      "loss": 9.1581,
+      "step": 76650
+    },
+    {
+      "epoch": 0.38283103198581736,
+      "grad_norm": 0.09483528882265091,
+      "learning_rate": 1.856141580515157e-05,
+      "loss": 9.1561,
+      "step": 76660
+    },
+    {
+      "epoch": 0.38288097081075684,
+      "grad_norm": 0.09543635696172714,
+      "learning_rate": 1.855991389021002e-05,
+      "loss": 9.1531,
+      "step": 76670
+    },
+    {
+      "epoch": 0.38293090963569626,
+      "grad_norm": 0.09006334096193314,
+      "learning_rate": 1.855841197526847e-05,
+      "loss": 9.157,
+      "step": 76680
+    },
+    {
+      "epoch": 0.38298084846063574,
+      "grad_norm": 0.09074097126722336,
+      "learning_rate": 1.8556910060326916e-05,
+      "loss": 9.1644,
+      "step": 76690
+    },
+    {
+      "epoch": 0.38303078728557516,
+      "grad_norm": 0.09905076771974564,
+      "learning_rate": 1.855540814538537e-05,
+      "loss": 9.1571,
+      "step": 76700
+    },
+    {
+      "epoch": 0.38308072611051464,
+      "grad_norm": 0.09774630516767502,
+      "learning_rate": 1.8553906230443816e-05,
+      "loss": 9.1457,
+      "step": 76710
+    },
+    {
+      "epoch": 0.38313066493545406,
+      "grad_norm": 0.09291476756334305,
+      "learning_rate": 1.8552404315502266e-05,
+      "loss": 9.144,
+      "step": 76720
+    },
+    {
+      "epoch": 0.38318060376039353,
+      "grad_norm": 0.09428045153617859,
+      "learning_rate": 1.8550902400560716e-05,
+      "loss": 9.1604,
+      "step": 76730
+    },
+    {
+      "epoch": 0.38323054258533296,
+      "grad_norm": 0.09279389679431915,
+      "learning_rate": 1.8549400485619163e-05,
+      "loss": 9.1636,
+      "step": 76740
+    },
+    {
+      "epoch": 0.38328048141027243,
+      "grad_norm": 0.10000916570425034,
+      "learning_rate": 1.8547898570677617e-05,
+      "loss": 9.1772,
+      "step": 76750
+    },
+    {
+      "epoch": 0.38333042023521186,
+      "grad_norm": 0.09311681985855103,
+      "learning_rate": 1.8546396655736063e-05,
+      "loss": 9.1693,
+      "step": 76760
+    },
+    {
+      "epoch": 0.38338035906015133,
+      "grad_norm": 0.09120059013366699,
+      "learning_rate": 1.8544894740794514e-05,
+      "loss": 9.1541,
+      "step": 76770
+    },
+    {
+      "epoch": 0.38343029788509075,
+      "grad_norm": 0.08788907527923584,
+      "learning_rate": 1.8543392825852964e-05,
+      "loss": 9.1603,
+      "step": 76780
+    },
+    {
+      "epoch": 0.38348023671003023,
+      "grad_norm": 0.09764920175075531,
+      "learning_rate": 1.854189091091141e-05,
+      "loss": 9.1641,
+      "step": 76790
+    },
+    {
+      "epoch": 0.38353017553496965,
+      "grad_norm": 0.098087377846241,
+      "learning_rate": 1.8540388995969864e-05,
+      "loss": 9.1472,
+      "step": 76800
+    },
+    {
+      "epoch": 0.38358011435990913,
+      "grad_norm": 0.09712155908346176,
+      "learning_rate": 1.853888708102831e-05,
+      "loss": 9.1688,
+      "step": 76810
+    },
+    {
+      "epoch": 0.38363005318484855,
+      "grad_norm": 0.09204491227865219,
+      "learning_rate": 1.853738516608676e-05,
+      "loss": 9.1757,
+      "step": 76820
+    },
+    {
+      "epoch": 0.38367999200978803,
+      "grad_norm": 0.09970824420452118,
+      "learning_rate": 1.853588325114521e-05,
+      "loss": 9.171,
+      "step": 76830
+    },
+    {
+      "epoch": 0.38372993083472745,
+      "grad_norm": 0.09648123383522034,
+      "learning_rate": 1.8534381336203658e-05,
+      "loss": 9.1693,
+      "step": 76840
+    },
+    {
+      "epoch": 0.38377986965966693,
+      "grad_norm": 0.09382256120443344,
+      "learning_rate": 1.853287942126211e-05,
+      "loss": 9.1641,
+      "step": 76850
+    },
+    {
+      "epoch": 0.38382980848460635,
+      "grad_norm": 0.09341836720705032,
+      "learning_rate": 1.853137750632056e-05,
+      "loss": 9.1648,
+      "step": 76860
+    },
+    {
+      "epoch": 0.3838797473095458,
+      "grad_norm": 0.09189660847187042,
+      "learning_rate": 1.852987559137901e-05,
+      "loss": 9.1553,
+      "step": 76870
+    },
+    {
+      "epoch": 0.38392968613448525,
+      "grad_norm": 0.09186588227748871,
+      "learning_rate": 1.852837367643746e-05,
+      "loss": 9.1624,
+      "step": 76880
+    },
+    {
+      "epoch": 0.3839796249594247,
+      "grad_norm": 0.09395985305309296,
+      "learning_rate": 1.852687176149591e-05,
+      "loss": 9.1741,
+      "step": 76890
+    },
+    {
+      "epoch": 0.38402956378436415,
+      "grad_norm": 0.09236618131399155,
+      "learning_rate": 1.852536984655436e-05,
+      "loss": 9.1473,
+      "step": 76900
+    },
+    {
+      "epoch": 0.3840795026093036,
+      "grad_norm": 0.1052381843328476,
+      "learning_rate": 1.8523867931612806e-05,
+      "loss": 9.1679,
+      "step": 76910
+    },
+    {
+      "epoch": 0.38412944143424305,
+      "grad_norm": 0.09067463874816895,
+      "learning_rate": 1.8522366016671256e-05,
+      "loss": 9.1509,
+      "step": 76920
+    },
+    {
+      "epoch": 0.3841793802591825,
+      "grad_norm": 0.09203337877988815,
+      "learning_rate": 1.8520864101729706e-05,
+      "loss": 9.1688,
+      "step": 76930
+    },
+    {
+      "epoch": 0.38422931908412195,
+      "grad_norm": 0.09451109915971756,
+      "learning_rate": 1.8519362186788156e-05,
+      "loss": 9.1508,
+      "step": 76940
+    },
+    {
+      "epoch": 0.3842792579090614,
+      "grad_norm": 0.09311412274837494,
+      "learning_rate": 1.8517860271846607e-05,
+      "loss": 9.1549,
+      "step": 76950
+    },
+    {
+      "epoch": 0.38432919673400084,
+      "grad_norm": 0.08877348154783249,
+      "learning_rate": 1.8516358356905053e-05,
+      "loss": 9.1582,
+      "step": 76960
+    },
+    {
+      "epoch": 0.3843791355589403,
+      "grad_norm": 0.09304699301719666,
+      "learning_rate": 1.8514856441963504e-05,
+      "loss": 9.1632,
+      "step": 76970
+    },
+    {
+      "epoch": 0.38442907438387974,
+      "grad_norm": 0.09285631775856018,
+      "learning_rate": 1.8513354527021954e-05,
+      "loss": 9.1592,
+      "step": 76980
+    },
+    {
+      "epoch": 0.3844790132088192,
+      "grad_norm": 0.09747754037380219,
+      "learning_rate": 1.8511852612080404e-05,
+      "loss": 9.1499,
+      "step": 76990
+    },
+    {
+      "epoch": 0.38452895203375864,
+      "grad_norm": 0.09608352929353714,
+      "learning_rate": 1.8510350697138854e-05,
+      "loss": 9.1616,
+      "step": 77000
+    },
+    {
+      "epoch": 0.3845788908586981,
+      "grad_norm": 0.09752684831619263,
+      "learning_rate": 1.85088487821973e-05,
+      "loss": 9.1647,
+      "step": 77010
+    },
+    {
+      "epoch": 0.38462882968363754,
+      "grad_norm": 0.09097377210855484,
+      "learning_rate": 1.850734686725575e-05,
+      "loss": 9.1597,
+      "step": 77020
+    },
+    {
+      "epoch": 0.384678768508577,
+      "grad_norm": 0.09708648175001144,
+      "learning_rate": 1.85058449523142e-05,
+      "loss": 9.158,
+      "step": 77030
+    },
+    {
+      "epoch": 0.38472870733351644,
+      "grad_norm": 0.09771420806646347,
+      "learning_rate": 1.850434303737265e-05,
+      "loss": 9.1618,
+      "step": 77040
+    },
+    {
+      "epoch": 0.3847786461584559,
+      "grad_norm": 0.09642086923122406,
+      "learning_rate": 1.85028411224311e-05,
+      "loss": 9.1574,
+      "step": 77050
+    },
+    {
+      "epoch": 0.38482858498339534,
+      "grad_norm": 0.09355588257312775,
+      "learning_rate": 1.850133920748955e-05,
+      "loss": 9.1562,
+      "step": 77060
+    },
+    {
+      "epoch": 0.3848785238083348,
+      "grad_norm": 0.09392425417900085,
+      "learning_rate": 1.8499837292548e-05,
+      "loss": 9.1579,
+      "step": 77070
+    },
+    {
+      "epoch": 0.38492846263327424,
+      "grad_norm": 0.09295708686113358,
+      "learning_rate": 1.849833537760645e-05,
+      "loss": 9.1519,
+      "step": 77080
+    },
+    {
+      "epoch": 0.3849784014582137,
+      "grad_norm": 0.09548468142747879,
+      "learning_rate": 1.84968334626649e-05,
+      "loss": 9.1605,
+      "step": 77090
+    },
+    {
+      "epoch": 0.38502834028315314,
+      "grad_norm": 0.09481420367956161,
+      "learning_rate": 1.849533154772335e-05,
+      "loss": 9.1501,
+      "step": 77100
+    },
+    {
+      "epoch": 0.3850782791080926,
+      "grad_norm": 0.10212825238704681,
+      "learning_rate": 1.8493829632781796e-05,
+      "loss": 9.1495,
+      "step": 77110
+    },
+    {
+      "epoch": 0.38512821793303204,
+      "grad_norm": 0.09605475515127182,
+      "learning_rate": 1.8492327717840246e-05,
+      "loss": 9.1723,
+      "step": 77120
+    },
+    {
+      "epoch": 0.3851781567579715,
+      "grad_norm": 0.09045825153589249,
+      "learning_rate": 1.8490825802898696e-05,
+      "loss": 9.1645,
+      "step": 77130
+    },
+    {
+      "epoch": 0.38522809558291093,
+      "grad_norm": 0.0943480134010315,
+      "learning_rate": 1.8489323887957146e-05,
+      "loss": 9.1471,
+      "step": 77140
+    },
+    {
+      "epoch": 0.38527803440785036,
+      "grad_norm": 0.09214089065790176,
+      "learning_rate": 1.8487821973015597e-05,
+      "loss": 9.1521,
+      "step": 77150
+    },
+    {
+      "epoch": 0.38532797323278983,
+      "grad_norm": 0.09132665395736694,
+      "learning_rate": 1.8486320058074043e-05,
+      "loss": 9.1588,
+      "step": 77160
+    },
+    {
+      "epoch": 0.38537791205772925,
+      "grad_norm": 0.09081696718931198,
+      "learning_rate": 1.8484818143132494e-05,
+      "loss": 9.1676,
+      "step": 77170
+    },
+    {
+      "epoch": 0.38542785088266873,
+      "grad_norm": 0.09519261121749878,
+      "learning_rate": 1.8483316228190944e-05,
+      "loss": 9.1709,
+      "step": 77180
+    },
+    {
+      "epoch": 0.38547778970760815,
+      "grad_norm": 0.090855672955513,
+      "learning_rate": 1.8481814313249394e-05,
+      "loss": 9.1543,
+      "step": 77190
+    },
+    {
+      "epoch": 0.38552772853254763,
+      "grad_norm": 0.09267590194940567,
+      "learning_rate": 1.8480312398307844e-05,
+      "loss": 9.1494,
+      "step": 77200
+    },
+    {
+      "epoch": 0.38557766735748705,
+      "grad_norm": 0.1009821891784668,
+      "learning_rate": 1.8478810483366294e-05,
+      "loss": 9.1579,
+      "step": 77210
+    },
+    {
+      "epoch": 0.38562760618242653,
+      "grad_norm": 0.09545056521892548,
+      "learning_rate": 1.847730856842474e-05,
+      "loss": 9.1533,
+      "step": 77220
+    },
+    {
+      "epoch": 0.38567754500736595,
+      "grad_norm": 0.09436902403831482,
+      "learning_rate": 1.847580665348319e-05,
+      "loss": 9.1525,
+      "step": 77230
+    },
+    {
+      "epoch": 0.38572748383230543,
+      "grad_norm": 0.09197907149791718,
+      "learning_rate": 1.847430473854164e-05,
+      "loss": 9.1414,
+      "step": 77240
+    },
+    {
+      "epoch": 0.38577742265724485,
+      "grad_norm": 0.08811882883310318,
+      "learning_rate": 1.847280282360009e-05,
+      "loss": 9.1626,
+      "step": 77250
+    },
+    {
+      "epoch": 0.38582736148218433,
+      "grad_norm": 0.09159471839666367,
+      "learning_rate": 1.8471300908658542e-05,
+      "loss": 9.1453,
+      "step": 77260
+    },
+    {
+      "epoch": 0.38587730030712375,
+      "grad_norm": 0.08988473564386368,
+      "learning_rate": 1.846979899371699e-05,
+      "loss": 9.1385,
+      "step": 77270
+    },
+    {
+      "epoch": 0.3859272391320632,
+      "grad_norm": 0.09255439788103104,
+      "learning_rate": 1.846829707877544e-05,
+      "loss": 9.1462,
+      "step": 77280
+    },
+    {
+      "epoch": 0.38597717795700265,
+      "grad_norm": 0.09517369419336319,
+      "learning_rate": 1.846679516383389e-05,
+      "loss": 9.1512,
+      "step": 77290
+    },
+    {
+      "epoch": 0.3860271167819421,
+      "grad_norm": 0.09043008089065552,
+      "learning_rate": 1.846529324889234e-05,
+      "loss": 9.1543,
+      "step": 77300
+    },
+    {
+      "epoch": 0.38607705560688155,
+      "grad_norm": 0.0920136496424675,
+      "learning_rate": 1.846379133395079e-05,
+      "loss": 9.1581,
+      "step": 77310
+    },
+    {
+      "epoch": 0.386126994431821,
+      "grad_norm": 0.09350711852312088,
+      "learning_rate": 1.8462289419009236e-05,
+      "loss": 9.1577,
+      "step": 77320
+    },
+    {
+      "epoch": 0.38617693325676045,
+      "grad_norm": 0.09980452060699463,
+      "learning_rate": 1.8460787504067686e-05,
+      "loss": 9.1488,
+      "step": 77330
+    },
+    {
+      "epoch": 0.3862268720816999,
+      "grad_norm": 0.09318140149116516,
+      "learning_rate": 1.8459285589126136e-05,
+      "loss": 9.1622,
+      "step": 77340
+    },
+    {
+      "epoch": 0.38627681090663935,
+      "grad_norm": 0.08900347352027893,
+      "learning_rate": 1.8457783674184587e-05,
+      "loss": 9.1495,
+      "step": 77350
+    },
+    {
+      "epoch": 0.3863267497315788,
+      "grad_norm": 0.09255148470401764,
+      "learning_rate": 1.8456281759243037e-05,
+      "loss": 9.1642,
+      "step": 77360
+    },
+    {
+      "epoch": 0.38637668855651824,
+      "grad_norm": 0.09274660795927048,
+      "learning_rate": 1.8454779844301484e-05,
+      "loss": 9.1568,
+      "step": 77370
+    },
+    {
+      "epoch": 0.3864266273814577,
+      "grad_norm": 0.09435738623142242,
+      "learning_rate": 1.8453277929359934e-05,
+      "loss": 9.1571,
+      "step": 77380
+    },
+    {
+      "epoch": 0.38647656620639714,
+      "grad_norm": 0.09425579011440277,
+      "learning_rate": 1.8451776014418384e-05,
+      "loss": 9.1503,
+      "step": 77390
+    },
+    {
+      "epoch": 0.3865265050313366,
+      "grad_norm": 0.09888068586587906,
+      "learning_rate": 1.8450274099476834e-05,
+      "loss": 9.1483,
+      "step": 77400
+    },
+    {
+      "epoch": 0.38657644385627604,
+      "grad_norm": 0.0868912860751152,
+      "learning_rate": 1.8448772184535284e-05,
+      "loss": 9.1477,
+      "step": 77410
+    },
+    {
+      "epoch": 0.3866263826812155,
+      "grad_norm": 0.09297952800989151,
+      "learning_rate": 1.844727026959373e-05,
+      "loss": 9.1485,
+      "step": 77420
+    },
+    {
+      "epoch": 0.38667632150615494,
+      "grad_norm": 0.09065989404916763,
+      "learning_rate": 1.844576835465218e-05,
+      "loss": 9.1491,
+      "step": 77430
+    },
+    {
+      "epoch": 0.3867262603310944,
+      "grad_norm": 0.0999770238995552,
+      "learning_rate": 1.844426643971063e-05,
+      "loss": 9.1413,
+      "step": 77440
+    },
+    {
+      "epoch": 0.38677619915603384,
+      "grad_norm": 0.09342008084058762,
+      "learning_rate": 1.844276452476908e-05,
+      "loss": 9.1529,
+      "step": 77450
+    },
+    {
+      "epoch": 0.3868261379809733,
+      "grad_norm": 0.09146399050951004,
+      "learning_rate": 1.8441262609827532e-05,
+      "loss": 9.1423,
+      "step": 77460
+    },
+    {
+      "epoch": 0.38687607680591274,
+      "grad_norm": 0.09426107257604599,
+      "learning_rate": 1.843976069488598e-05,
+      "loss": 9.1583,
+      "step": 77470
+    },
+    {
+      "epoch": 0.3869260156308522,
+      "grad_norm": 0.0957350954413414,
+      "learning_rate": 1.843825877994443e-05,
+      "loss": 9.1493,
+      "step": 77480
+    },
+    {
+      "epoch": 0.38697595445579164,
+      "grad_norm": 0.09453484416007996,
+      "learning_rate": 1.843675686500288e-05,
+      "loss": 9.1603,
+      "step": 77490
+    },
+    {
+      "epoch": 0.3870258932807311,
+      "grad_norm": 0.09307164698839188,
+      "learning_rate": 1.843525495006133e-05,
+      "loss": 9.1468,
+      "step": 77500
+    },
+    {
+      "epoch": 0.38707583210567054,
+      "grad_norm": 0.09427579492330551,
+      "learning_rate": 1.843375303511978e-05,
+      "loss": 9.1512,
+      "step": 77510
+    },
+    {
+      "epoch": 0.38712577093061,
+      "grad_norm": 0.09382929652929306,
+      "learning_rate": 1.8432251120178226e-05,
+      "loss": 9.154,
+      "step": 77520
+    },
+    {
+      "epoch": 0.38717570975554944,
+      "grad_norm": 0.0964818224310875,
+      "learning_rate": 1.843074920523668e-05,
+      "loss": 9.1498,
+      "step": 77530
+    },
+    {
+      "epoch": 0.3872256485804889,
+      "grad_norm": 0.09739819914102554,
+      "learning_rate": 1.8429247290295127e-05,
+      "loss": 9.1422,
+      "step": 77540
+    },
+    {
+      "epoch": 0.38727558740542833,
+      "grad_norm": 0.0981798991560936,
+      "learning_rate": 1.8427745375353577e-05,
+      "loss": 9.1577,
+      "step": 77550
+    },
+    {
+      "epoch": 0.3873255262303678,
+      "grad_norm": 0.09459348767995834,
+      "learning_rate": 1.8426243460412027e-05,
+      "loss": 9.1525,
+      "step": 77560
+    },
+    {
+      "epoch": 0.38737546505530723,
+      "grad_norm": 0.08874860405921936,
+      "learning_rate": 1.8424741545470474e-05,
+      "loss": 9.1339,
+      "step": 77570
+    },
+    {
+      "epoch": 0.3874254038802467,
+      "grad_norm": 0.0875902771949768,
+      "learning_rate": 1.8423239630528927e-05,
+      "loss": 9.1537,
+      "step": 77580
+    },
+    {
+      "epoch": 0.38747534270518613,
+      "grad_norm": 0.09034288674592972,
+      "learning_rate": 1.8421737715587374e-05,
+      "loss": 9.1553,
+      "step": 77590
+    },
+    {
+      "epoch": 0.3875252815301256,
+      "grad_norm": 0.09590810537338257,
+      "learning_rate": 1.8420235800645824e-05,
+      "loss": 9.162,
+      "step": 77600
+    },
+    {
+      "epoch": 0.38757522035506503,
+      "grad_norm": 0.0925174131989479,
+      "learning_rate": 1.8418733885704274e-05,
+      "loss": 9.1575,
+      "step": 77610
+    },
+    {
+      "epoch": 0.3876251591800045,
+      "grad_norm": 0.09790638089179993,
+      "learning_rate": 1.841723197076272e-05,
+      "loss": 9.1518,
+      "step": 77620
+    },
+    {
+      "epoch": 0.38767509800494393,
+      "grad_norm": 0.09194985032081604,
+      "learning_rate": 1.8415730055821175e-05,
+      "loss": 9.1374,
+      "step": 77630
+    },
+    {
+      "epoch": 0.3877250368298834,
+      "grad_norm": 0.09064792096614838,
+      "learning_rate": 1.841422814087962e-05,
+      "loss": 9.1472,
+      "step": 77640
+    },
+    {
+      "epoch": 0.38777497565482283,
+      "grad_norm": 0.09329324215650558,
+      "learning_rate": 1.841272622593807e-05,
+      "loss": 9.1595,
+      "step": 77650
+    },
+    {
+      "epoch": 0.3878249144797623,
+      "grad_norm": 0.09056134521961212,
+      "learning_rate": 1.8411224310996522e-05,
+      "loss": 9.1453,
+      "step": 77660
+    },
+    {
+      "epoch": 0.3878748533047017,
+      "grad_norm": 0.08973592519760132,
+      "learning_rate": 1.840972239605497e-05,
+      "loss": 9.1536,
+      "step": 77670
+    },
+    {
+      "epoch": 0.3879247921296412,
+      "grad_norm": 0.09086141735315323,
+      "learning_rate": 1.8408220481113422e-05,
+      "loss": 9.1471,
+      "step": 77680
+    },
+    {
+      "epoch": 0.3879747309545806,
+      "grad_norm": 0.09189939498901367,
+      "learning_rate": 1.840671856617187e-05,
+      "loss": 9.1613,
+      "step": 77690
+    },
+    {
+      "epoch": 0.3880246697795201,
+      "grad_norm": 0.09157737344503403,
+      "learning_rate": 1.840521665123032e-05,
+      "loss": 9.1504,
+      "step": 77700
+    },
+    {
+      "epoch": 0.3880746086044595,
+      "grad_norm": 0.09373574703931808,
+      "learning_rate": 1.840371473628877e-05,
+      "loss": 9.1432,
+      "step": 77710
+    },
+    {
+      "epoch": 0.388124547429399,
+      "grad_norm": 0.09311201423406601,
+      "learning_rate": 1.8402212821347216e-05,
+      "loss": 9.1433,
+      "step": 77720
+    },
+    {
+      "epoch": 0.3881744862543384,
+      "grad_norm": 0.09637656062841415,
+      "learning_rate": 1.840071090640567e-05,
+      "loss": 9.1406,
+      "step": 77730
+    },
+    {
+      "epoch": 0.3882244250792779,
+      "grad_norm": 0.09410546720027924,
+      "learning_rate": 1.8399208991464117e-05,
+      "loss": 9.1557,
+      "step": 77740
+    },
+    {
+      "epoch": 0.3882743639042173,
+      "grad_norm": 0.09236341714859009,
+      "learning_rate": 1.8397707076522567e-05,
+      "loss": 9.151,
+      "step": 77750
+    },
+    {
+      "epoch": 0.3883243027291568,
+      "grad_norm": 0.10012514889240265,
+      "learning_rate": 1.8396205161581017e-05,
+      "loss": 9.1482,
+      "step": 77760
+    },
+    {
+      "epoch": 0.3883742415540962,
+      "grad_norm": 0.09507282823324203,
+      "learning_rate": 1.8394703246639464e-05,
+      "loss": 9.1616,
+      "step": 77770
+    },
+    {
+      "epoch": 0.3884241803790357,
+      "grad_norm": 0.09531018882989883,
+      "learning_rate": 1.8393201331697917e-05,
+      "loss": 9.1508,
+      "step": 77780
+    },
+    {
+      "epoch": 0.3884741192039751,
+      "grad_norm": 0.09432236105203629,
+      "learning_rate": 1.8391699416756364e-05,
+      "loss": 9.1536,
+      "step": 77790
+    },
+    {
+      "epoch": 0.3885240580289146,
+      "grad_norm": 0.09611290693283081,
+      "learning_rate": 1.8390197501814814e-05,
+      "loss": 9.1479,
+      "step": 77800
+    },
+    {
+      "epoch": 0.388573996853854,
+      "grad_norm": 0.09352581202983856,
+      "learning_rate": 1.8388695586873264e-05,
+      "loss": 9.157,
+      "step": 77810
+    },
+    {
+      "epoch": 0.3886239356787935,
+      "grad_norm": 0.09698393195867538,
+      "learning_rate": 1.838719367193171e-05,
+      "loss": 9.1378,
+      "step": 77820
+    },
+    {
+      "epoch": 0.3886738745037329,
+      "grad_norm": 0.09182106703519821,
+      "learning_rate": 1.8385691756990165e-05,
+      "loss": 9.1501,
+      "step": 77830
+    },
+    {
+      "epoch": 0.3887238133286724,
+      "grad_norm": 0.09386096894741058,
+      "learning_rate": 1.838418984204861e-05,
+      "loss": 9.1381,
+      "step": 77840
+    },
+    {
+      "epoch": 0.3887737521536118,
+      "grad_norm": 0.09277167916297913,
+      "learning_rate": 1.8382687927107065e-05,
+      "loss": 9.1395,
+      "step": 77850
+    },
+    {
+      "epoch": 0.3888236909785513,
+      "grad_norm": 0.09407127648591995,
+      "learning_rate": 1.8381186012165512e-05,
+      "loss": 9.1482,
+      "step": 77860
+    },
+    {
+      "epoch": 0.3888736298034907,
+      "grad_norm": 0.09299039095640182,
+      "learning_rate": 1.837968409722396e-05,
+      "loss": 9.1526,
+      "step": 77870
+    },
+    {
+      "epoch": 0.3889235686284302,
+      "grad_norm": 0.09196474403142929,
+      "learning_rate": 1.8378182182282412e-05,
+      "loss": 9.1458,
+      "step": 77880
+    },
+    {
+      "epoch": 0.3889735074533696,
+      "grad_norm": 0.09697432816028595,
+      "learning_rate": 1.837668026734086e-05,
+      "loss": 9.1328,
+      "step": 77890
+    },
+    {
+      "epoch": 0.3890234462783091,
+      "grad_norm": 0.09332502633333206,
+      "learning_rate": 1.8375178352399313e-05,
+      "loss": 9.1421,
+      "step": 77900
+    },
+    {
+      "epoch": 0.3890733851032485,
+      "grad_norm": 0.09647480398416519,
+      "learning_rate": 1.837367643745776e-05,
+      "loss": 9.1494,
+      "step": 77910
+    },
+    {
+      "epoch": 0.389123323928188,
+      "grad_norm": 0.08914009481668472,
+      "learning_rate": 1.8372174522516206e-05,
+      "loss": 9.1537,
+      "step": 77920
+    },
+    {
+      "epoch": 0.3891732627531274,
+      "grad_norm": 0.09153462946414948,
+      "learning_rate": 1.837067260757466e-05,
+      "loss": 9.1461,
+      "step": 77930
+    },
+    {
+      "epoch": 0.3892232015780669,
+      "grad_norm": 0.09255599975585938,
+      "learning_rate": 1.8369170692633107e-05,
+      "loss": 9.1481,
+      "step": 77940
+    },
+    {
+      "epoch": 0.3892731404030063,
+      "grad_norm": 0.10034976899623871,
+      "learning_rate": 1.836766877769156e-05,
+      "loss": 9.1427,
+      "step": 77950
+    },
+    {
+      "epoch": 0.3893230792279458,
+      "grad_norm": 0.09407934546470642,
+      "learning_rate": 1.8366166862750007e-05,
+      "loss": 9.1505,
+      "step": 77960
+    },
+    {
+      "epoch": 0.3893730180528852,
+      "grad_norm": 0.09382222592830658,
+      "learning_rate": 1.8364664947808454e-05,
+      "loss": 9.1392,
+      "step": 77970
+    },
+    {
+      "epoch": 0.3894229568778247,
+      "grad_norm": 0.09767767041921616,
+      "learning_rate": 1.8363163032866907e-05,
+      "loss": 9.1463,
+      "step": 77980
+    },
+    {
+      "epoch": 0.3894728957027641,
+      "grad_norm": 0.0940956324338913,
+      "learning_rate": 1.8361661117925354e-05,
+      "loss": 9.144,
+      "step": 77990
+    },
+    {
+      "epoch": 0.3895228345277036,
+      "grad_norm": 0.09381180256605148,
+      "learning_rate": 1.8360159202983808e-05,
+      "loss": 9.1565,
+      "step": 78000
+    },
+    {
+      "epoch": 0.389572773352643,
+      "grad_norm": 0.09209650754928589,
+      "learning_rate": 1.8358657288042254e-05,
+      "loss": 9.149,
+      "step": 78010
+    },
+    {
+      "epoch": 0.3896227121775825,
+      "grad_norm": 0.09117661416530609,
+      "learning_rate": 1.83571553731007e-05,
+      "loss": 9.1426,
+      "step": 78020
+    },
+    {
+      "epoch": 0.3896726510025219,
+      "grad_norm": 0.0906655564904213,
+      "learning_rate": 1.8355653458159155e-05,
+      "loss": 9.1527,
+      "step": 78030
+    },
+    {
+      "epoch": 0.3897225898274614,
+      "grad_norm": 0.09310529381036758,
+      "learning_rate": 1.83541515432176e-05,
+      "loss": 9.1374,
+      "step": 78040
+    },
+    {
+      "epoch": 0.3897725286524008,
+      "grad_norm": 0.08677937090396881,
+      "learning_rate": 1.8352649628276055e-05,
+      "loss": 9.1483,
+      "step": 78050
+    },
+    {
+      "epoch": 0.3898224674773403,
+      "grad_norm": 0.09317050874233246,
+      "learning_rate": 1.8351147713334502e-05,
+      "loss": 9.1481,
+      "step": 78060
+    },
+    {
+      "epoch": 0.3898724063022797,
+      "grad_norm": 0.09428145736455917,
+      "learning_rate": 1.834964579839295e-05,
+      "loss": 9.1525,
+      "step": 78070
+    },
+    {
+      "epoch": 0.3899223451272192,
+      "grad_norm": 0.09383665770292282,
+      "learning_rate": 1.8348143883451402e-05,
+      "loss": 9.1374,
+      "step": 78080
+    },
+    {
+      "epoch": 0.3899722839521586,
+      "grad_norm": 0.08776134252548218,
+      "learning_rate": 1.834664196850985e-05,
+      "loss": 9.1481,
+      "step": 78090
+    },
+    {
+      "epoch": 0.3900222227770981,
+      "grad_norm": 0.09563107043504715,
+      "learning_rate": 1.8345140053568303e-05,
+      "loss": 9.1403,
+      "step": 78100
+    },
+    {
+      "epoch": 0.3900721616020375,
+      "grad_norm": 0.09589152038097382,
+      "learning_rate": 1.834363813862675e-05,
+      "loss": 9.1492,
+      "step": 78110
+    },
+    {
+      "epoch": 0.390122100426977,
+      "grad_norm": 0.09651998430490494,
+      "learning_rate": 1.8342136223685196e-05,
+      "loss": 9.1421,
+      "step": 78120
+    },
+    {
+      "epoch": 0.3901720392519164,
+      "grad_norm": 0.0984501987695694,
+      "learning_rate": 1.834063430874365e-05,
+      "loss": 9.1399,
+      "step": 78130
+    },
+    {
+      "epoch": 0.3902219780768558,
+      "grad_norm": 0.09631562978029251,
+      "learning_rate": 1.8339132393802097e-05,
+      "loss": 9.1386,
+      "step": 78140
+    },
+    {
+      "epoch": 0.3902719169017953,
+      "grad_norm": 0.09599178284406662,
+      "learning_rate": 1.833763047886055e-05,
+      "loss": 9.1481,
+      "step": 78150
+    },
+    {
+      "epoch": 0.3903218557267347,
+      "grad_norm": 0.0964631512761116,
+      "learning_rate": 1.8336128563918997e-05,
+      "loss": 9.14,
+      "step": 78160
+    },
+    {
+      "epoch": 0.3903717945516742,
+      "grad_norm": 0.09247509390115738,
+      "learning_rate": 1.8334626648977447e-05,
+      "loss": 9.1432,
+      "step": 78170
+    },
+    {
+      "epoch": 0.3904217333766136,
+      "grad_norm": 0.09320325404405594,
+      "learning_rate": 1.8333124734035897e-05,
+      "loss": 9.1439,
+      "step": 78180
+    },
+    {
+      "epoch": 0.3904716722015531,
+      "grad_norm": 0.09139348566532135,
+      "learning_rate": 1.8331622819094344e-05,
+      "loss": 9.156,
+      "step": 78190
+    },
+    {
+      "epoch": 0.3905216110264925,
+      "grad_norm": 0.09385591000318527,
+      "learning_rate": 1.8330120904152798e-05,
+      "loss": 9.1441,
+      "step": 78200
+    },
+    {
+      "epoch": 0.390571549851432,
+      "grad_norm": 0.09762391448020935,
+      "learning_rate": 1.8328618989211244e-05,
+      "loss": 9.1494,
+      "step": 78210
+    },
+    {
+      "epoch": 0.3906214886763714,
+      "grad_norm": 0.09472424536943436,
+      "learning_rate": 1.8327117074269695e-05,
+      "loss": 9.1339,
+      "step": 78220
+    },
+    {
+      "epoch": 0.3906714275013109,
+      "grad_norm": 0.09121805429458618,
+      "learning_rate": 1.8325615159328145e-05,
+      "loss": 9.1461,
+      "step": 78230
+    },
+    {
+      "epoch": 0.3907213663262503,
+      "grad_norm": 0.09944367408752441,
+      "learning_rate": 1.832411324438659e-05,
+      "loss": 9.1434,
+      "step": 78240
+    },
+    {
+      "epoch": 0.3907713051511898,
+      "grad_norm": 0.08815374970436096,
+      "learning_rate": 1.8322611329445045e-05,
+      "loss": 9.138,
+      "step": 78250
+    },
+    {
+      "epoch": 0.3908212439761292,
+      "grad_norm": 0.09055221825838089,
+      "learning_rate": 1.8321109414503492e-05,
+      "loss": 9.1388,
+      "step": 78260
+    },
+    {
+      "epoch": 0.3908711828010687,
+      "grad_norm": 0.09463029354810715,
+      "learning_rate": 1.8319607499561942e-05,
+      "loss": 9.1477,
+      "step": 78270
+    },
+    {
+      "epoch": 0.3909211216260081,
+      "grad_norm": 0.09334297478199005,
+      "learning_rate": 1.8318105584620392e-05,
+      "loss": 9.1358,
+      "step": 78280
+    },
+    {
+      "epoch": 0.3909710604509476,
+      "grad_norm": 0.09212377667427063,
+      "learning_rate": 1.831660366967884e-05,
+      "loss": 9.1363,
+      "step": 78290
+    },
+    {
+      "epoch": 0.391020999275887,
+      "grad_norm": 0.09191081672906876,
+      "learning_rate": 1.8315101754737293e-05,
+      "loss": 9.1443,
+      "step": 78300
+    },
+    {
+      "epoch": 0.3910709381008265,
+      "grad_norm": 0.0968773290514946,
+      "learning_rate": 1.831359983979574e-05,
+      "loss": 9.1507,
+      "step": 78310
+    },
+    {
+      "epoch": 0.3911208769257659,
+      "grad_norm": 0.09789808839559555,
+      "learning_rate": 1.831209792485419e-05,
+      "loss": 9.1365,
+      "step": 78320
+    },
+    {
+      "epoch": 0.3911708157507054,
+      "grad_norm": 0.09304534643888474,
+      "learning_rate": 1.831059600991264e-05,
+      "loss": 9.1417,
+      "step": 78330
+    },
+    {
+      "epoch": 0.3912207545756448,
+      "grad_norm": 0.09302885830402374,
+      "learning_rate": 1.8309094094971087e-05,
+      "loss": 9.1402,
+      "step": 78340
+    },
+    {
+      "epoch": 0.3912706934005843,
+      "grad_norm": 0.10683131963014603,
+      "learning_rate": 1.830759218002954e-05,
+      "loss": 9.1388,
+      "step": 78350
+    },
+    {
+      "epoch": 0.3913206322255237,
+      "grad_norm": 0.09414584934711456,
+      "learning_rate": 1.8306090265087987e-05,
+      "loss": 9.1488,
+      "step": 78360
+    },
+    {
+      "epoch": 0.3913705710504632,
+      "grad_norm": 0.10074611753225327,
+      "learning_rate": 1.8304588350146437e-05,
+      "loss": 9.1423,
+      "step": 78370
+    },
+    {
+      "epoch": 0.3914205098754026,
+      "grad_norm": 0.0921291634440422,
+      "learning_rate": 1.8303086435204887e-05,
+      "loss": 9.1431,
+      "step": 78380
+    },
+    {
+      "epoch": 0.3914704487003421,
+      "grad_norm": 0.09323565661907196,
+      "learning_rate": 1.8301584520263334e-05,
+      "loss": 9.1458,
+      "step": 78390
+    },
+    {
+      "epoch": 0.3915203875252815,
+      "grad_norm": 0.09994424134492874,
+      "learning_rate": 1.8300082605321788e-05,
+      "loss": 9.1407,
+      "step": 78400
+    },
+    {
+      "epoch": 0.391570326350221,
+      "grad_norm": 0.09224946796894073,
+      "learning_rate": 1.8298580690380234e-05,
+      "loss": 9.1374,
+      "step": 78410
+    },
+    {
+      "epoch": 0.3916202651751604,
+      "grad_norm": 0.09452834725379944,
+      "learning_rate": 1.8297078775438685e-05,
+      "loss": 9.145,
+      "step": 78420
+    },
+    {
+      "epoch": 0.3916702040000999,
+      "grad_norm": 0.10009356588125229,
+      "learning_rate": 1.8295576860497135e-05,
+      "loss": 9.1396,
+      "step": 78430
+    },
+    {
+      "epoch": 0.3917201428250393,
+      "grad_norm": 0.0899641364812851,
+      "learning_rate": 1.829407494555558e-05,
+      "loss": 9.1334,
+      "step": 78440
+    },
+    {
+      "epoch": 0.3917700816499788,
+      "grad_norm": 0.09872201830148697,
+      "learning_rate": 1.8292573030614035e-05,
+      "loss": 9.1392,
+      "step": 78450
+    },
+    {
+      "epoch": 0.3918200204749182,
+      "grad_norm": 0.09648354351520538,
+      "learning_rate": 1.8291071115672482e-05,
+      "loss": 9.1291,
+      "step": 78460
+    },
+    {
+      "epoch": 0.3918699592998577,
+      "grad_norm": 0.09189938753843307,
+      "learning_rate": 1.8289569200730932e-05,
+      "loss": 9.1376,
+      "step": 78470
+    },
+    {
+      "epoch": 0.3919198981247971,
+      "grad_norm": 0.0942487046122551,
+      "learning_rate": 1.8288067285789382e-05,
+      "loss": 9.1554,
+      "step": 78480
+    },
+    {
+      "epoch": 0.3919698369497366,
+      "grad_norm": 0.0920250341296196,
+      "learning_rate": 1.8286565370847832e-05,
+      "loss": 9.1335,
+      "step": 78490
+    },
+    {
+      "epoch": 0.392019775774676,
+      "grad_norm": 0.09978582710027695,
+      "learning_rate": 1.8285063455906283e-05,
+      "loss": 9.1339,
+      "step": 78500
+    },
+    {
+      "epoch": 0.3920697145996155,
+      "grad_norm": 0.09699604660272598,
+      "learning_rate": 1.828356154096473e-05,
+      "loss": 9.146,
+      "step": 78510
+    },
+    {
+      "epoch": 0.3921196534245549,
+      "grad_norm": 0.09093842655420303,
+      "learning_rate": 1.828205962602318e-05,
+      "loss": 9.1379,
+      "step": 78520
+    },
+    {
+      "epoch": 0.3921695922494944,
+      "grad_norm": 0.09174724668264389,
+      "learning_rate": 1.828055771108163e-05,
+      "loss": 9.1321,
+      "step": 78530
+    },
+    {
+      "epoch": 0.3922195310744338,
+      "grad_norm": 0.10035578906536102,
+      "learning_rate": 1.827905579614008e-05,
+      "loss": 9.1408,
+      "step": 78540
+    },
+    {
+      "epoch": 0.3922694698993733,
+      "grad_norm": 0.09467300027608871,
+      "learning_rate": 1.827755388119853e-05,
+      "loss": 9.1563,
+      "step": 78550
+    },
+    {
+      "epoch": 0.3923194087243127,
+      "grad_norm": 0.09329366683959961,
+      "learning_rate": 1.8276051966256977e-05,
+      "loss": 9.1288,
+      "step": 78560
+    },
+    {
+      "epoch": 0.3923693475492522,
+      "grad_norm": 0.09315571188926697,
+      "learning_rate": 1.8274550051315427e-05,
+      "loss": 9.1358,
+      "step": 78570
+    },
+    {
+      "epoch": 0.3924192863741916,
+      "grad_norm": 0.09450618177652359,
+      "learning_rate": 1.8273048136373877e-05,
+      "loss": 9.1401,
+      "step": 78580
+    },
+    {
+      "epoch": 0.3924692251991311,
+      "grad_norm": 0.09535399079322815,
+      "learning_rate": 1.8271546221432327e-05,
+      "loss": 9.1415,
+      "step": 78590
+    },
+    {
+      "epoch": 0.3925191640240705,
+      "grad_norm": 0.09569433331489563,
+      "learning_rate": 1.8270044306490778e-05,
+      "loss": 9.145,
+      "step": 78600
+    },
+    {
+      "epoch": 0.39256910284901,
+      "grad_norm": 0.09955570846796036,
+      "learning_rate": 1.8268542391549224e-05,
+      "loss": 9.1322,
+      "step": 78610
+    },
+    {
+      "epoch": 0.3926190416739494,
+      "grad_norm": 0.10329762101173401,
+      "learning_rate": 1.8267040476607675e-05,
+      "loss": 9.127,
+      "step": 78620
+    },
+    {
+      "epoch": 0.3926689804988889,
+      "grad_norm": 0.09589072316884995,
+      "learning_rate": 1.8265538561666125e-05,
+      "loss": 9.1384,
+      "step": 78630
+    },
+    {
+      "epoch": 0.3927189193238283,
+      "grad_norm": 0.09281908720731735,
+      "learning_rate": 1.8264036646724575e-05,
+      "loss": 9.143,
+      "step": 78640
+    },
+    {
+      "epoch": 0.3927688581487678,
+      "grad_norm": 0.08548375964164734,
+      "learning_rate": 1.8262534731783025e-05,
+      "loss": 9.1432,
+      "step": 78650
+    },
+    {
+      "epoch": 0.3928187969737072,
+      "grad_norm": 0.10105335712432861,
+      "learning_rate": 1.8261032816841472e-05,
+      "loss": 9.1397,
+      "step": 78660
+    },
+    {
+      "epoch": 0.3928687357986467,
+      "grad_norm": 0.10008376091718674,
+      "learning_rate": 1.8259530901899922e-05,
+      "loss": 9.1335,
+      "step": 78670
+    },
+    {
+      "epoch": 0.3929186746235861,
+      "grad_norm": 0.09541326016187668,
+      "learning_rate": 1.8258028986958372e-05,
+      "loss": 9.1324,
+      "step": 78680
+    },
+    {
+      "epoch": 0.39296861344852557,
+      "grad_norm": 0.0945446565747261,
+      "learning_rate": 1.8256527072016822e-05,
+      "loss": 9.1329,
+      "step": 78690
+    },
+    {
+      "epoch": 0.393018552273465,
+      "grad_norm": 0.09741361439228058,
+      "learning_rate": 1.8255025157075273e-05,
+      "loss": 9.1452,
+      "step": 78700
+    },
+    {
+      "epoch": 0.39306849109840447,
+      "grad_norm": 0.09333933889865875,
+      "learning_rate": 1.825352324213372e-05,
+      "loss": 9.1364,
+      "step": 78710
+    },
+    {
+      "epoch": 0.3931184299233439,
+      "grad_norm": 0.09559603035449982,
+      "learning_rate": 1.825202132719217e-05,
+      "loss": 9.1332,
+      "step": 78720
+    },
+    {
+      "epoch": 0.39316836874828337,
+      "grad_norm": 0.09066354483366013,
+      "learning_rate": 1.825051941225062e-05,
+      "loss": 9.1269,
+      "step": 78730
+    },
+    {
+      "epoch": 0.3932183075732228,
+      "grad_norm": 0.09495797008275986,
+      "learning_rate": 1.824901749730907e-05,
+      "loss": 9.1402,
+      "step": 78740
+    },
+    {
+      "epoch": 0.39326824639816227,
+      "grad_norm": 0.08994494378566742,
+      "learning_rate": 1.824751558236752e-05,
+      "loss": 9.145,
+      "step": 78750
+    },
+    {
+      "epoch": 0.3933181852231017,
+      "grad_norm": 0.08942119777202606,
+      "learning_rate": 1.8246013667425967e-05,
+      "loss": 9.1393,
+      "step": 78760
+    },
+    {
+      "epoch": 0.39336812404804117,
+      "grad_norm": 0.09385786950588226,
+      "learning_rate": 1.8244511752484417e-05,
+      "loss": 9.141,
+      "step": 78770
+    },
+    {
+      "epoch": 0.3934180628729806,
+      "grad_norm": 0.09488862007856369,
+      "learning_rate": 1.8243009837542867e-05,
+      "loss": 9.1458,
+      "step": 78780
+    },
+    {
+      "epoch": 0.39346800169792007,
+      "grad_norm": 0.0987120270729065,
+      "learning_rate": 1.8241507922601317e-05,
+      "loss": 9.1272,
+      "step": 78790
+    },
+    {
+      "epoch": 0.3935179405228595,
+      "grad_norm": 0.09095241129398346,
+      "learning_rate": 1.8240006007659768e-05,
+      "loss": 9.13,
+      "step": 78800
+    },
+    {
+      "epoch": 0.39356787934779897,
+      "grad_norm": 0.08821319043636322,
+      "learning_rate": 1.8238504092718218e-05,
+      "loss": 9.1371,
+      "step": 78810
+    },
+    {
+      "epoch": 0.3936178181727384,
+      "grad_norm": 0.09783562272787094,
+      "learning_rate": 1.8237002177776665e-05,
+      "loss": 9.1444,
+      "step": 78820
+    },
+    {
+      "epoch": 0.39366775699767786,
+      "grad_norm": 0.09780313819646835,
+      "learning_rate": 1.8235500262835115e-05,
+      "loss": 9.1244,
+      "step": 78830
+    },
+    {
+      "epoch": 0.3937176958226173,
+      "grad_norm": 0.0926724523305893,
+      "learning_rate": 1.8233998347893565e-05,
+      "loss": 9.134,
+      "step": 78840
+    },
+    {
+      "epoch": 0.39376763464755676,
+      "grad_norm": 0.09907516092061996,
+      "learning_rate": 1.8232496432952015e-05,
+      "loss": 9.137,
+      "step": 78850
+    },
+    {
+      "epoch": 0.3938175734724962,
+      "grad_norm": 0.09416226297616959,
+      "learning_rate": 1.8230994518010465e-05,
+      "loss": 9.1347,
+      "step": 78860
+    },
+    {
+      "epoch": 0.39386751229743566,
+      "grad_norm": 0.09291341155767441,
+      "learning_rate": 1.8229492603068912e-05,
+      "loss": 9.1378,
+      "step": 78870
+    },
+    {
+      "epoch": 0.3939174511223751,
+      "grad_norm": 0.09237024933099747,
+      "learning_rate": 1.8227990688127362e-05,
+      "loss": 9.135,
+      "step": 78880
+    },
+    {
+      "epoch": 0.39396738994731456,
+      "grad_norm": 0.09598385542631149,
+      "learning_rate": 1.8226488773185812e-05,
+      "loss": 9.1253,
+      "step": 78890
+    },
+    {
+      "epoch": 0.394017328772254,
+      "grad_norm": 0.09644147008657455,
+      "learning_rate": 1.8224986858244263e-05,
+      "loss": 9.1379,
+      "step": 78900
+    },
+    {
+      "epoch": 0.39406726759719346,
+      "grad_norm": 0.09208100289106369,
+      "learning_rate": 1.8223484943302713e-05,
+      "loss": 9.1481,
+      "step": 78910
+    },
+    {
+      "epoch": 0.3941172064221329,
+      "grad_norm": 0.091837577521801,
+      "learning_rate": 1.822198302836116e-05,
+      "loss": 9.142,
+      "step": 78920
+    },
+    {
+      "epoch": 0.39416714524707236,
+      "grad_norm": 0.09068531543016434,
+      "learning_rate": 1.822048111341961e-05,
+      "loss": 9.1359,
+      "step": 78930
+    },
+    {
+      "epoch": 0.3942170840720118,
+      "grad_norm": 0.09340940415859222,
+      "learning_rate": 1.821897919847806e-05,
+      "loss": 9.1407,
+      "step": 78940
+    },
+    {
+      "epoch": 0.39426702289695126,
+      "grad_norm": 0.09838276356458664,
+      "learning_rate": 1.821747728353651e-05,
+      "loss": 9.1273,
+      "step": 78950
+    },
+    {
+      "epoch": 0.3943169617218907,
+      "grad_norm": 0.09257405251264572,
+      "learning_rate": 1.821597536859496e-05,
+      "loss": 9.1309,
+      "step": 78960
+    },
+    {
+      "epoch": 0.39436690054683016,
+      "grad_norm": 0.09464617073535919,
+      "learning_rate": 1.8214473453653407e-05,
+      "loss": 9.125,
+      "step": 78970
+    },
+    {
+      "epoch": 0.3944168393717696,
+      "grad_norm": 0.09580957889556885,
+      "learning_rate": 1.8212971538711857e-05,
+      "loss": 9.1253,
+      "step": 78980
+    },
+    {
+      "epoch": 0.39446677819670906,
+      "grad_norm": 0.09709136933088303,
+      "learning_rate": 1.8211469623770308e-05,
+      "loss": 9.1327,
+      "step": 78990
+    },
+    {
+      "epoch": 0.3945167170216485,
+      "grad_norm": 0.10155075043439865,
+      "learning_rate": 1.8209967708828758e-05,
+      "loss": 9.1258,
+      "step": 79000
+    },
+    {
+      "epoch": 0.39456665584658795,
+      "grad_norm": 0.08959164470434189,
+      "learning_rate": 1.8208465793887208e-05,
+      "loss": 9.1309,
+      "step": 79010
+    },
+    {
+      "epoch": 0.3946165946715274,
+      "grad_norm": 0.10539384931325912,
+      "learning_rate": 1.8206963878945655e-05,
+      "loss": 9.1265,
+      "step": 79020
+    },
+    {
+      "epoch": 0.39466653349646685,
+      "grad_norm": 0.09576883167028427,
+      "learning_rate": 1.8205461964004105e-05,
+      "loss": 9.1324,
+      "step": 79030
+    },
+    {
+      "epoch": 0.3947164723214063,
+      "grad_norm": 0.09333497285842896,
+      "learning_rate": 1.8203960049062555e-05,
+      "loss": 9.1188,
+      "step": 79040
+    },
+    {
+      "epoch": 0.39476641114634575,
+      "grad_norm": 0.10127103328704834,
+      "learning_rate": 1.8202458134121005e-05,
+      "loss": 9.1255,
+      "step": 79050
+    },
+    {
+      "epoch": 0.3948163499712852,
+      "grad_norm": 0.09161249548196793,
+      "learning_rate": 1.8200956219179455e-05,
+      "loss": 9.1357,
+      "step": 79060
+    },
+    {
+      "epoch": 0.39486628879622465,
+      "grad_norm": 0.08713789284229279,
+      "learning_rate": 1.8199454304237902e-05,
+      "loss": 9.1448,
+      "step": 79070
+    },
+    {
+      "epoch": 0.3949162276211641,
+      "grad_norm": 0.09979274123907089,
+      "learning_rate": 1.8197952389296352e-05,
+      "loss": 9.1245,
+      "step": 79080
+    },
+    {
+      "epoch": 0.39496616644610355,
+      "grad_norm": 0.09479266405105591,
+      "learning_rate": 1.8196450474354803e-05,
+      "loss": 9.1359,
+      "step": 79090
+    },
+    {
+      "epoch": 0.39501610527104297,
+      "grad_norm": 0.0914844274520874,
+      "learning_rate": 1.8194948559413253e-05,
+      "loss": 9.1356,
+      "step": 79100
+    },
+    {
+      "epoch": 0.3950660440959824,
+      "grad_norm": 0.09772706776857376,
+      "learning_rate": 1.8193446644471703e-05,
+      "loss": 9.1253,
+      "step": 79110
+    },
+    {
+      "epoch": 0.39511598292092187,
+      "grad_norm": 0.08755208551883698,
+      "learning_rate": 1.819194472953015e-05,
+      "loss": 9.1353,
+      "step": 79120
+    },
+    {
+      "epoch": 0.3951659217458613,
+      "grad_norm": 0.09900221973657608,
+      "learning_rate": 1.8190442814588603e-05,
+      "loss": 9.1268,
+      "step": 79130
+    },
+    {
+      "epoch": 0.39521586057080077,
+      "grad_norm": 0.09705707430839539,
+      "learning_rate": 1.818894089964705e-05,
+      "loss": 9.1355,
+      "step": 79140
+    },
+    {
+      "epoch": 0.3952657993957402,
+      "grad_norm": 0.09519544243812561,
+      "learning_rate": 1.81874389847055e-05,
+      "loss": 9.1456,
+      "step": 79150
+    },
+    {
+      "epoch": 0.39531573822067967,
+      "grad_norm": 0.09430651366710663,
+      "learning_rate": 1.818593706976395e-05,
+      "loss": 9.1446,
+      "step": 79160
+    },
+    {
+      "epoch": 0.3953656770456191,
+      "grad_norm": 0.0924086719751358,
+      "learning_rate": 1.8184435154822397e-05,
+      "loss": 9.1273,
+      "step": 79170
+    },
+    {
+      "epoch": 0.39541561587055857,
+      "grad_norm": 0.09312140196561813,
+      "learning_rate": 1.818293323988085e-05,
+      "loss": 9.1378,
+      "step": 79180
+    },
+    {
+      "epoch": 0.395465554695498,
+      "grad_norm": 0.08896911144256592,
+      "learning_rate": 1.8181431324939298e-05,
+      "loss": 9.1462,
+      "step": 79190
+    },
+    {
+      "epoch": 0.39551549352043747,
+      "grad_norm": 0.09490480273962021,
+      "learning_rate": 1.8179929409997748e-05,
+      "loss": 9.1304,
+      "step": 79200
+    },
+    {
+      "epoch": 0.3955654323453769,
+      "grad_norm": 0.09386398643255234,
+      "learning_rate": 1.8178427495056198e-05,
+      "loss": 9.1264,
+      "step": 79210
+    },
+    {
+      "epoch": 0.39561537117031637,
+      "grad_norm": 0.09263758361339569,
+      "learning_rate": 1.8176925580114645e-05,
+      "loss": 9.1253,
+      "step": 79220
+    },
+    {
+      "epoch": 0.3956653099952558,
+      "grad_norm": 0.09182369709014893,
+      "learning_rate": 1.8175423665173098e-05,
+      "loss": 9.1266,
+      "step": 79230
+    },
+    {
+      "epoch": 0.39571524882019526,
+      "grad_norm": 0.09554483741521835,
+      "learning_rate": 1.8173921750231545e-05,
+      "loss": 9.1378,
+      "step": 79240
+    },
+    {
+      "epoch": 0.3957651876451347,
+      "grad_norm": 0.09417613595724106,
+      "learning_rate": 1.8172419835289995e-05,
+      "loss": 9.1335,
+      "step": 79250
+    },
+    {
+      "epoch": 0.39581512647007416,
+      "grad_norm": 0.09561773389577866,
+      "learning_rate": 1.8170917920348445e-05,
+      "loss": 9.1292,
+      "step": 79260
+    },
+    {
+      "epoch": 0.3958650652950136,
+      "grad_norm": 0.09833815693855286,
+      "learning_rate": 1.8169416005406892e-05,
+      "loss": 9.1259,
+      "step": 79270
+    },
+    {
+      "epoch": 0.39591500411995306,
+      "grad_norm": 0.09367582947015762,
+      "learning_rate": 1.8167914090465346e-05,
+      "loss": 9.1323,
+      "step": 79280
+    },
+    {
+      "epoch": 0.3959649429448925,
+      "grad_norm": 0.09364238381385803,
+      "learning_rate": 1.8166412175523793e-05,
+      "loss": 9.1206,
+      "step": 79290
+    },
+    {
+      "epoch": 0.39601488176983196,
+      "grad_norm": 0.09396982192993164,
+      "learning_rate": 1.8164910260582243e-05,
+      "loss": 9.1248,
+      "step": 79300
+    },
+    {
+      "epoch": 0.3960648205947714,
+      "grad_norm": 0.09186158329248428,
+      "learning_rate": 1.8163408345640693e-05,
+      "loss": 9.1398,
+      "step": 79310
+    },
+    {
+      "epoch": 0.39611475941971086,
+      "grad_norm": 0.09493670612573624,
+      "learning_rate": 1.816190643069914e-05,
+      "loss": 9.129,
+      "step": 79320
+    },
+    {
+      "epoch": 0.3961646982446503,
+      "grad_norm": 0.0939958244562149,
+      "learning_rate": 1.8160404515757593e-05,
+      "loss": 9.1374,
+      "step": 79330
+    },
+    {
+      "epoch": 0.39621463706958976,
+      "grad_norm": 0.09256591647863388,
+      "learning_rate": 1.815890260081604e-05,
+      "loss": 9.1247,
+      "step": 79340
+    },
+    {
+      "epoch": 0.3962645758945292,
+      "grad_norm": 0.0958244577050209,
+      "learning_rate": 1.815740068587449e-05,
+      "loss": 9.1239,
+      "step": 79350
+    },
+    {
+      "epoch": 0.39631451471946866,
+      "grad_norm": 0.09303276985883713,
+      "learning_rate": 1.815589877093294e-05,
+      "loss": 9.1319,
+      "step": 79360
+    },
+    {
+      "epoch": 0.3963644535444081,
+      "grad_norm": 0.10332775115966797,
+      "learning_rate": 1.8154396855991387e-05,
+      "loss": 9.1269,
+      "step": 79370
+    },
+    {
+      "epoch": 0.39641439236934756,
+      "grad_norm": 0.09653986990451813,
+      "learning_rate": 1.815289494104984e-05,
+      "loss": 9.1214,
+      "step": 79380
+    },
+    {
+      "epoch": 0.396464331194287,
+      "grad_norm": 0.09338979423046112,
+      "learning_rate": 1.8151393026108288e-05,
+      "loss": 9.1155,
+      "step": 79390
+    },
+    {
+      "epoch": 0.39651427001922646,
+      "grad_norm": 0.09556922316551208,
+      "learning_rate": 1.8149891111166738e-05,
+      "loss": 9.1303,
+      "step": 79400
+    },
+    {
+      "epoch": 0.3965642088441659,
+      "grad_norm": 0.09039589762687683,
+      "learning_rate": 1.8148389196225188e-05,
+      "loss": 9.1323,
+      "step": 79410
+    },
+    {
+      "epoch": 0.39661414766910535,
+      "grad_norm": 0.09440034627914429,
+      "learning_rate": 1.8146887281283635e-05,
+      "loss": 9.131,
+      "step": 79420
+    },
+    {
+      "epoch": 0.3966640864940448,
+      "grad_norm": 0.0989178940653801,
+      "learning_rate": 1.8145385366342088e-05,
+      "loss": 9.1244,
+      "step": 79430
+    },
+    {
+      "epoch": 0.39671402531898425,
+      "grad_norm": 0.0976061150431633,
+      "learning_rate": 1.8143883451400535e-05,
+      "loss": 9.1148,
+      "step": 79440
+    },
+    {
+      "epoch": 0.3967639641439237,
+      "grad_norm": 0.09087694436311722,
+      "learning_rate": 1.8142381536458985e-05,
+      "loss": 9.1231,
+      "step": 79450
+    },
+    {
+      "epoch": 0.39681390296886315,
+      "grad_norm": 0.08781003206968307,
+      "learning_rate": 1.8140879621517435e-05,
+      "loss": 9.1166,
+      "step": 79460
+    },
+    {
+      "epoch": 0.3968638417938026,
+      "grad_norm": 0.08944723010063171,
+      "learning_rate": 1.8139377706575882e-05,
+      "loss": 9.1332,
+      "step": 79470
+    },
+    {
+      "epoch": 0.39691378061874205,
+      "grad_norm": 0.09698821604251862,
+      "learning_rate": 1.8137875791634336e-05,
+      "loss": 9.123,
+      "step": 79480
+    },
+    {
+      "epoch": 0.3969637194436815,
+      "grad_norm": 0.09338833391666412,
+      "learning_rate": 1.8136373876692783e-05,
+      "loss": 9.1365,
+      "step": 79490
+    },
+    {
+      "epoch": 0.39701365826862095,
+      "grad_norm": 0.09354853630065918,
+      "learning_rate": 1.8134871961751236e-05,
+      "loss": 9.1218,
+      "step": 79500
+    },
+    {
+      "epoch": 0.39706359709356037,
+      "grad_norm": 0.09762848913669586,
+      "learning_rate": 1.8133370046809683e-05,
+      "loss": 9.1297,
+      "step": 79510
+    },
+    {
+      "epoch": 0.39711353591849985,
+      "grad_norm": 0.0980941578745842,
+      "learning_rate": 1.813186813186813e-05,
+      "loss": 9.1269,
+      "step": 79520
+    },
+    {
+      "epoch": 0.39716347474343927,
+      "grad_norm": 0.09508627653121948,
+      "learning_rate": 1.8130366216926583e-05,
+      "loss": 9.1146,
+      "step": 79530
+    },
+    {
+      "epoch": 0.39721341356837875,
+      "grad_norm": 0.09081006050109863,
+      "learning_rate": 1.812886430198503e-05,
+      "loss": 9.1403,
+      "step": 79540
+    },
+    {
+      "epoch": 0.39726335239331817,
+      "grad_norm": 0.09213867783546448,
+      "learning_rate": 1.8127362387043484e-05,
+      "loss": 9.116,
+      "step": 79550
+    },
+    {
+      "epoch": 0.39731329121825765,
+      "grad_norm": 0.09143271297216415,
+      "learning_rate": 1.812586047210193e-05,
+      "loss": 9.1232,
+      "step": 79560
+    },
+    {
+      "epoch": 0.39736323004319707,
+      "grad_norm": 0.09485948085784912,
+      "learning_rate": 1.8124358557160377e-05,
+      "loss": 9.1199,
+      "step": 79570
+    },
+    {
+      "epoch": 0.39741316886813655,
+      "grad_norm": 0.09463505446910858,
+      "learning_rate": 1.812285664221883e-05,
+      "loss": 9.1312,
+      "step": 79580
+    },
+    {
+      "epoch": 0.39746310769307597,
+      "grad_norm": 0.09464465081691742,
+      "learning_rate": 1.8121354727277278e-05,
+      "loss": 9.1267,
+      "step": 79590
+    },
+    {
+      "epoch": 0.39751304651801544,
+      "grad_norm": 0.09754239022731781,
+      "learning_rate": 1.811985281233573e-05,
+      "loss": 9.1344,
+      "step": 79600
+    },
+    {
+      "epoch": 0.39756298534295487,
+      "grad_norm": 0.09760978072881699,
+      "learning_rate": 1.8118350897394178e-05,
+      "loss": 9.1294,
+      "step": 79610
+    },
+    {
+      "epoch": 0.39761292416789434,
+      "grad_norm": 0.08997610211372375,
+      "learning_rate": 1.8116848982452625e-05,
+      "loss": 9.1341,
+      "step": 79620
+    },
+    {
+      "epoch": 0.39766286299283377,
+      "grad_norm": 0.09305060654878616,
+      "learning_rate": 1.8115347067511078e-05,
+      "loss": 9.1429,
+      "step": 79630
+    },
+    {
+      "epoch": 0.39771280181777324,
+      "grad_norm": 0.0855415090918541,
+      "learning_rate": 1.8113845152569525e-05,
+      "loss": 9.1285,
+      "step": 79640
+    },
+    {
+      "epoch": 0.39776274064271266,
+      "grad_norm": 0.09722486883401871,
+      "learning_rate": 1.811234323762798e-05,
+      "loss": 9.1218,
+      "step": 79650
+    },
+    {
+      "epoch": 0.39781267946765214,
+      "grad_norm": 0.09455563873052597,
+      "learning_rate": 1.8110841322686425e-05,
+      "loss": 9.1177,
+      "step": 79660
+    },
+    {
+      "epoch": 0.39786261829259156,
+      "grad_norm": 0.09245886653661728,
+      "learning_rate": 1.8109339407744872e-05,
+      "loss": 9.1312,
+      "step": 79670
+    },
+    {
+      "epoch": 0.39791255711753104,
+      "grad_norm": 0.09259022772312164,
+      "learning_rate": 1.8107837492803326e-05,
+      "loss": 9.1313,
+      "step": 79680
+    },
+    {
+      "epoch": 0.39796249594247046,
+      "grad_norm": 0.08969958871603012,
+      "learning_rate": 1.8106335577861773e-05,
+      "loss": 9.1278,
+      "step": 79690
+    },
+    {
+      "epoch": 0.39801243476740994,
+      "grad_norm": 0.08772096782922745,
+      "learning_rate": 1.8104833662920226e-05,
+      "loss": 9.132,
+      "step": 79700
+    },
+    {
+      "epoch": 0.39806237359234936,
+      "grad_norm": 0.0951986312866211,
+      "learning_rate": 1.8103331747978673e-05,
+      "loss": 9.1208,
+      "step": 79710
+    },
+    {
+      "epoch": 0.39811231241728884,
+      "grad_norm": 0.09764955192804337,
+      "learning_rate": 1.810182983303712e-05,
+      "loss": 9.1099,
+      "step": 79720
+    },
+    {
+      "epoch": 0.39816225124222826,
+      "grad_norm": 0.09765150398015976,
+      "learning_rate": 1.8100327918095573e-05,
+      "loss": 9.1259,
+      "step": 79730
+    },
+    {
+      "epoch": 0.39821219006716774,
+      "grad_norm": 0.09225156158208847,
+      "learning_rate": 1.809882600315402e-05,
+      "loss": 9.1319,
+      "step": 79740
+    },
+    {
+      "epoch": 0.39826212889210716,
+      "grad_norm": 0.09323499351739883,
+      "learning_rate": 1.8097324088212474e-05,
+      "loss": 9.1295,
+      "step": 79750
+    },
+    {
+      "epoch": 0.39831206771704664,
+      "grad_norm": 0.0989677682518959,
+      "learning_rate": 1.809582217327092e-05,
+      "loss": 9.1328,
+      "step": 79760
+    },
+    {
+      "epoch": 0.39836200654198606,
+      "grad_norm": 0.09290597587823868,
+      "learning_rate": 1.8094320258329367e-05,
+      "loss": 9.1173,
+      "step": 79770
+    },
+    {
+      "epoch": 0.39841194536692554,
+      "grad_norm": 0.09229425340890884,
+      "learning_rate": 1.809281834338782e-05,
+      "loss": 9.1144,
+      "step": 79780
+    },
+    {
+      "epoch": 0.39846188419186496,
+      "grad_norm": 0.09045591205358505,
+      "learning_rate": 1.8091316428446268e-05,
+      "loss": 9.1248,
+      "step": 79790
+    },
+    {
+      "epoch": 0.39851182301680443,
+      "grad_norm": 0.0939536839723587,
+      "learning_rate": 1.808981451350472e-05,
+      "loss": 9.1254,
+      "step": 79800
+    },
+    {
+      "epoch": 0.39856176184174386,
+      "grad_norm": 0.09491851180791855,
+      "learning_rate": 1.8088312598563168e-05,
+      "loss": 9.123,
+      "step": 79810
+    },
+    {
+      "epoch": 0.39861170066668333,
+      "grad_norm": 0.08969860523939133,
+      "learning_rate": 1.8086810683621618e-05,
+      "loss": 9.1183,
+      "step": 79820
+    },
+    {
+      "epoch": 0.39866163949162275,
+      "grad_norm": 0.10173804312944412,
+      "learning_rate": 1.8085308768680068e-05,
+      "loss": 9.1313,
+      "step": 79830
+    },
+    {
+      "epoch": 0.39871157831656223,
+      "grad_norm": 0.0922439694404602,
+      "learning_rate": 1.8083806853738515e-05,
+      "loss": 9.1186,
+      "step": 79840
+    },
+    {
+      "epoch": 0.39876151714150165,
+      "grad_norm": 0.08746393769979477,
+      "learning_rate": 1.808230493879697e-05,
+      "loss": 9.1344,
+      "step": 79850
+    },
+    {
+      "epoch": 0.39881145596644113,
+      "grad_norm": 0.095425546169281,
+      "learning_rate": 1.8080803023855415e-05,
+      "loss": 9.1175,
+      "step": 79860
+    },
+    {
+      "epoch": 0.39886139479138055,
+      "grad_norm": 0.09297501295804977,
+      "learning_rate": 1.8079301108913866e-05,
+      "loss": 9.1252,
+      "step": 79870
+    },
+    {
+      "epoch": 0.39891133361632003,
+      "grad_norm": 0.09445119649171829,
+      "learning_rate": 1.8077799193972316e-05,
+      "loss": 9.1152,
+      "step": 79880
+    },
+    {
+      "epoch": 0.39896127244125945,
+      "grad_norm": 0.08981538563966751,
+      "learning_rate": 1.8076297279030763e-05,
+      "loss": 9.1123,
+      "step": 79890
+    },
+    {
+      "epoch": 0.39901121126619893,
+      "grad_norm": 0.09903477132320404,
+      "learning_rate": 1.8074795364089216e-05,
+      "loss": 9.1247,
+      "step": 79900
+    },
+    {
+      "epoch": 0.39906115009113835,
+      "grad_norm": 0.09373074769973755,
+      "learning_rate": 1.8073293449147663e-05,
+      "loss": 9.1176,
+      "step": 79910
+    },
+    {
+      "epoch": 0.3991110889160778,
+      "grad_norm": 0.0910414308309555,
+      "learning_rate": 1.8071791534206113e-05,
+      "loss": 9.124,
+      "step": 79920
+    },
+    {
+      "epoch": 0.39916102774101725,
+      "grad_norm": 0.09311582893133163,
+      "learning_rate": 1.8070289619264563e-05,
+      "loss": 9.111,
+      "step": 79930
+    },
+    {
+      "epoch": 0.3992109665659567,
+      "grad_norm": 0.09247269481420517,
+      "learning_rate": 1.806878770432301e-05,
+      "loss": 9.1145,
+      "step": 79940
+    },
+    {
+      "epoch": 0.39926090539089615,
+      "grad_norm": 0.09804600477218628,
+      "learning_rate": 1.8067285789381464e-05,
+      "loss": 9.1214,
+      "step": 79950
+    },
+    {
+      "epoch": 0.3993108442158356,
+      "grad_norm": 0.09990818053483963,
+      "learning_rate": 1.806578387443991e-05,
+      "loss": 9.1349,
+      "step": 79960
+    },
+    {
+      "epoch": 0.39936078304077505,
+      "grad_norm": 0.09712336957454681,
+      "learning_rate": 1.806428195949836e-05,
+      "loss": 9.1245,
+      "step": 79970
+    },
+    {
+      "epoch": 0.3994107218657145,
+      "grad_norm": 0.09431684017181396,
+      "learning_rate": 1.806278004455681e-05,
+      "loss": 9.1178,
+      "step": 79980
+    },
+    {
+      "epoch": 0.39946066069065395,
+      "grad_norm": 0.10282929986715317,
+      "learning_rate": 1.8061278129615258e-05,
+      "loss": 9.1093,
+      "step": 79990
+    },
+    {
+      "epoch": 0.3995105995155934,
+      "grad_norm": 0.09497708082199097,
+      "learning_rate": 1.805977621467371e-05,
+      "loss": 9.1184,
+      "step": 80000
+    },
+    {
+      "epoch": 0.39956053834053284,
+      "grad_norm": 0.09546716511249542,
+      "learning_rate": 1.8058274299732158e-05,
+      "loss": 9.1205,
+      "step": 80010
+    },
+    {
+      "epoch": 0.3996104771654723,
+      "grad_norm": 0.0929088369011879,
+      "learning_rate": 1.8056772384790608e-05,
+      "loss": 9.1212,
+      "step": 80020
+    },
+    {
+      "epoch": 0.39966041599041174,
+      "grad_norm": 0.09245330840349197,
+      "learning_rate": 1.8055270469849058e-05,
+      "loss": 9.1023,
+      "step": 80030
+    },
+    {
+      "epoch": 0.3997103548153512,
+      "grad_norm": 0.09106311202049255,
+      "learning_rate": 1.8053768554907505e-05,
+      "loss": 9.1202,
+      "step": 80040
+    },
+    {
+      "epoch": 0.39976029364029064,
+      "grad_norm": 0.10017859935760498,
+      "learning_rate": 1.805226663996596e-05,
+      "loss": 9.126,
+      "step": 80050
+    },
+    {
+      "epoch": 0.3998102324652301,
+      "grad_norm": 0.09317342936992645,
+      "learning_rate": 1.8050764725024405e-05,
+      "loss": 9.1289,
+      "step": 80060
+    },
+    {
+      "epoch": 0.39986017129016954,
+      "grad_norm": 0.0955187976360321,
+      "learning_rate": 1.8049262810082856e-05,
+      "loss": 9.1126,
+      "step": 80070
+    },
+    {
+      "epoch": 0.399910110115109,
+      "grad_norm": 0.09439694136381149,
+      "learning_rate": 1.8047760895141306e-05,
+      "loss": 9.1324,
+      "step": 80080
+    },
+    {
+      "epoch": 0.39996004894004844,
+      "grad_norm": 0.09123414754867554,
+      "learning_rate": 1.8046258980199753e-05,
+      "loss": 9.1342,
+      "step": 80090
+    },
+    {
+      "epoch": 0.40000998776498786,
+      "grad_norm": 0.09292822331190109,
+      "learning_rate": 1.8044757065258206e-05,
+      "loss": 9.111,
+      "step": 80100
+    },
+    {
+      "epoch": 0.40005992658992734,
+      "grad_norm": 0.09875541925430298,
+      "learning_rate": 1.8043255150316653e-05,
+      "loss": 9.1114,
+      "step": 80110
+    },
+    {
+      "epoch": 0.40010986541486676,
+      "grad_norm": 0.09288553893566132,
+      "learning_rate": 1.8041753235375103e-05,
+      "loss": 9.1085,
+      "step": 80120
+    },
+    {
+      "epoch": 0.40015980423980624,
+      "grad_norm": 0.09572578966617584,
+      "learning_rate": 1.8040251320433553e-05,
+      "loss": 9.1151,
+      "step": 80130
+    },
+    {
+      "epoch": 0.40020974306474566,
+      "grad_norm": 0.09946201741695404,
+      "learning_rate": 1.8038749405492003e-05,
+      "loss": 9.1163,
+      "step": 80140
+    },
+    {
+      "epoch": 0.40025968188968514,
+      "grad_norm": 0.09416273236274719,
+      "learning_rate": 1.8037247490550454e-05,
+      "loss": 9.1196,
+      "step": 80150
+    },
+    {
+      "epoch": 0.40030962071462456,
+      "grad_norm": 0.09236134588718414,
+      "learning_rate": 1.80357455756089e-05,
+      "loss": 9.1084,
+      "step": 80160
+    },
+    {
+      "epoch": 0.40035955953956404,
+      "grad_norm": 0.09351241588592529,
+      "learning_rate": 1.803424366066735e-05,
+      "loss": 9.1082,
+      "step": 80170
+    },
+    {
+      "epoch": 0.40040949836450346,
+      "grad_norm": 0.09657840430736542,
+      "learning_rate": 1.80327417457258e-05,
+      "loss": 9.1077,
+      "step": 80180
+    },
+    {
+      "epoch": 0.40045943718944293,
+      "grad_norm": 0.09256424009799957,
+      "learning_rate": 1.803123983078425e-05,
+      "loss": 9.1178,
+      "step": 80190
+    },
+    {
+      "epoch": 0.40050937601438236,
+      "grad_norm": 0.09746617078781128,
+      "learning_rate": 1.80297379158427e-05,
+      "loss": 9.1223,
+      "step": 80200
+    },
+    {
+      "epoch": 0.40055931483932183,
+      "grad_norm": 0.09331115335226059,
+      "learning_rate": 1.8028236000901148e-05,
+      "loss": 9.1215,
+      "step": 80210
+    },
+    {
+      "epoch": 0.40060925366426126,
+      "grad_norm": 0.09966427087783813,
+      "learning_rate": 1.8026734085959598e-05,
+      "loss": 9.1028,
+      "step": 80220
+    },
+    {
+      "epoch": 0.40065919248920073,
+      "grad_norm": 0.09679210931062698,
+      "learning_rate": 1.8025232171018048e-05,
+      "loss": 9.1186,
+      "step": 80230
+    },
+    {
+      "epoch": 0.40070913131414015,
+      "grad_norm": 0.09198209643363953,
+      "learning_rate": 1.80237302560765e-05,
+      "loss": 9.1265,
+      "step": 80240
+    },
+    {
+      "epoch": 0.40075907013907963,
+      "grad_norm": 0.09457174688577652,
+      "learning_rate": 1.802222834113495e-05,
+      "loss": 9.1212,
+      "step": 80250
+    },
+    {
+      "epoch": 0.40080900896401905,
+      "grad_norm": 0.09352682530879974,
+      "learning_rate": 1.8020726426193395e-05,
+      "loss": 9.1148,
+      "step": 80260
+    },
+    {
+      "epoch": 0.40085894778895853,
+      "grad_norm": 0.0890619084239006,
+      "learning_rate": 1.8019224511251846e-05,
+      "loss": 9.126,
+      "step": 80270
+    },
+    {
+      "epoch": 0.40090888661389795,
+      "grad_norm": 0.09067501127719879,
+      "learning_rate": 1.8017722596310296e-05,
+      "loss": 9.1201,
+      "step": 80280
+    },
+    {
+      "epoch": 0.40095882543883743,
+      "grad_norm": 0.09653503447771072,
+      "learning_rate": 1.8016220681368746e-05,
+      "loss": 9.1192,
+      "step": 80290
+    },
+    {
+      "epoch": 0.40100876426377685,
+      "grad_norm": 0.09168026596307755,
+      "learning_rate": 1.8014718766427196e-05,
+      "loss": 9.1156,
+      "step": 80300
+    },
+    {
+      "epoch": 0.40105870308871633,
+      "grad_norm": 0.09609059989452362,
+      "learning_rate": 1.8013216851485643e-05,
+      "loss": 9.1139,
+      "step": 80310
+    },
+    {
+      "epoch": 0.40110864191365575,
+      "grad_norm": 0.09476880729198456,
+      "learning_rate": 1.8011714936544093e-05,
+      "loss": 9.1134,
+      "step": 80320
+    },
+    {
+      "epoch": 0.4011585807385952,
+      "grad_norm": 0.10273495316505432,
+      "learning_rate": 1.8010213021602543e-05,
+      "loss": 9.1195,
+      "step": 80330
+    },
+    {
+      "epoch": 0.40120851956353465,
+      "grad_norm": 0.09645021706819534,
+      "learning_rate": 1.8008711106660993e-05,
+      "loss": 9.1238,
+      "step": 80340
+    },
+    {
+      "epoch": 0.4012584583884741,
+      "grad_norm": 0.09573419392108917,
+      "learning_rate": 1.8007209191719444e-05,
+      "loss": 9.1229,
+      "step": 80350
+    },
+    {
+      "epoch": 0.40130839721341355,
+      "grad_norm": 0.09707140922546387,
+      "learning_rate": 1.800570727677789e-05,
+      "loss": 9.1014,
+      "step": 80360
+    },
+    {
+      "epoch": 0.401358336038353,
+      "grad_norm": 0.08955365419387817,
+      "learning_rate": 1.800420536183634e-05,
+      "loss": 9.1242,
+      "step": 80370
+    },
+    {
+      "epoch": 0.40140827486329245,
+      "grad_norm": 0.09678129851818085,
+      "learning_rate": 1.800270344689479e-05,
+      "loss": 9.1202,
+      "step": 80380
+    },
+    {
+      "epoch": 0.4014582136882319,
+      "grad_norm": 0.09080725908279419,
+      "learning_rate": 1.800120153195324e-05,
+      "loss": 9.1198,
+      "step": 80390
+    },
+    {
+      "epoch": 0.40150815251317135,
+      "grad_norm": 0.09258688241243362,
+      "learning_rate": 1.799969961701169e-05,
+      "loss": 9.1001,
+      "step": 80400
+    },
+    {
+      "epoch": 0.4015580913381108,
+      "grad_norm": 0.09553803503513336,
+      "learning_rate": 1.7998197702070138e-05,
+      "loss": 9.1137,
+      "step": 80410
+    },
+    {
+      "epoch": 0.40160803016305024,
+      "grad_norm": 0.09325069189071655,
+      "learning_rate": 1.7996695787128588e-05,
+      "loss": 9.1155,
+      "step": 80420
+    },
+    {
+      "epoch": 0.4016579689879897,
+      "grad_norm": 0.09590888768434525,
+      "learning_rate": 1.799519387218704e-05,
+      "loss": 9.1142,
+      "step": 80430
+    },
+    {
+      "epoch": 0.40170790781292914,
+      "grad_norm": 0.08667058497667313,
+      "learning_rate": 1.799369195724549e-05,
+      "loss": 9.1148,
+      "step": 80440
+    },
+    {
+      "epoch": 0.4017578466378686,
+      "grad_norm": 0.088908351957798,
+      "learning_rate": 1.799219004230394e-05,
+      "loss": 9.1099,
+      "step": 80450
+    },
+    {
+      "epoch": 0.40180778546280804,
+      "grad_norm": 0.09739526361227036,
+      "learning_rate": 1.799068812736239e-05,
+      "loss": 9.114,
+      "step": 80460
+    },
+    {
+      "epoch": 0.4018577242877475,
+      "grad_norm": 0.09346520155668259,
+      "learning_rate": 1.7989186212420836e-05,
+      "loss": 9.1227,
+      "step": 80470
+    },
+    {
+      "epoch": 0.40190766311268694,
+      "grad_norm": 0.09403660148382187,
+      "learning_rate": 1.7987684297479286e-05,
+      "loss": 9.123,
+      "step": 80480
+    },
+    {
+      "epoch": 0.4019576019376264,
+      "grad_norm": 0.09151037782430649,
+      "learning_rate": 1.7986182382537736e-05,
+      "loss": 9.109,
+      "step": 80490
+    },
+    {
+      "epoch": 0.40200754076256584,
+      "grad_norm": 0.09313507378101349,
+      "learning_rate": 1.7984680467596186e-05,
+      "loss": 9.102,
+      "step": 80500
+    },
+    {
+      "epoch": 0.4020574795875053,
+      "grad_norm": 0.09583766758441925,
+      "learning_rate": 1.7983178552654636e-05,
+      "loss": 9.095,
+      "step": 80510
+    },
+    {
+      "epoch": 0.40210741841244474,
+      "grad_norm": 0.10357358306646347,
+      "learning_rate": 1.7981676637713083e-05,
+      "loss": 9.1297,
+      "step": 80520
+    },
+    {
+      "epoch": 0.4021573572373842,
+      "grad_norm": 0.09157849848270416,
+      "learning_rate": 1.7980174722771533e-05,
+      "loss": 9.0986,
+      "step": 80530
+    },
+    {
+      "epoch": 0.40220729606232364,
+      "grad_norm": 0.0905398353934288,
+      "learning_rate": 1.7978672807829984e-05,
+      "loss": 9.1144,
+      "step": 80540
+    },
+    {
+      "epoch": 0.4022572348872631,
+      "grad_norm": 0.09321150183677673,
+      "learning_rate": 1.7977170892888434e-05,
+      "loss": 9.1098,
+      "step": 80550
+    },
+    {
+      "epoch": 0.40230717371220254,
+      "grad_norm": 0.09091385453939438,
+      "learning_rate": 1.7975668977946884e-05,
+      "loss": 9.1172,
+      "step": 80560
+    },
+    {
+      "epoch": 0.402357112537142,
+      "grad_norm": 0.09568551182746887,
+      "learning_rate": 1.797416706300533e-05,
+      "loss": 9.1055,
+      "step": 80570
+    },
+    {
+      "epoch": 0.40240705136208144,
+      "grad_norm": 0.08929819613695145,
+      "learning_rate": 1.797266514806378e-05,
+      "loss": 9.1045,
+      "step": 80580
+    },
+    {
+      "epoch": 0.4024569901870209,
+      "grad_norm": 0.09343628585338593,
+      "learning_rate": 1.797116323312223e-05,
+      "loss": 9.0965,
+      "step": 80590
+    },
+    {
+      "epoch": 0.40250692901196033,
+      "grad_norm": 0.0918562188744545,
+      "learning_rate": 1.796966131818068e-05,
+      "loss": 9.103,
+      "step": 80600
+    },
+    {
+      "epoch": 0.4025568678368998,
+      "grad_norm": 0.09256646037101746,
+      "learning_rate": 1.796815940323913e-05,
+      "loss": 9.1031,
+      "step": 80610
+    },
+    {
+      "epoch": 0.40260680666183923,
+      "grad_norm": 0.08834106475114822,
+      "learning_rate": 1.7966657488297578e-05,
+      "loss": 9.1066,
+      "step": 80620
+    },
+    {
+      "epoch": 0.4026567454867787,
+      "grad_norm": 0.09646911174058914,
+      "learning_rate": 1.796515557335603e-05,
+      "loss": 9.1038,
+      "step": 80630
+    },
+    {
+      "epoch": 0.40270668431171813,
+      "grad_norm": 0.09727106243371964,
+      "learning_rate": 1.796365365841448e-05,
+      "loss": 9.1046,
+      "step": 80640
+    },
+    {
+      "epoch": 0.4027566231366576,
+      "grad_norm": 0.0980573520064354,
+      "learning_rate": 1.796215174347293e-05,
+      "loss": 9.1105,
+      "step": 80650
+    },
+    {
+      "epoch": 0.40280656196159703,
+      "grad_norm": 0.0933801680803299,
+      "learning_rate": 1.796064982853138e-05,
+      "loss": 9.1102,
+      "step": 80660
+    },
+    {
+      "epoch": 0.4028565007865365,
+      "grad_norm": 0.09309963136911392,
+      "learning_rate": 1.7959147913589826e-05,
+      "loss": 9.1067,
+      "step": 80670
+    },
+    {
+      "epoch": 0.40290643961147593,
+      "grad_norm": 0.09479737281799316,
+      "learning_rate": 1.7957645998648276e-05,
+      "loss": 9.118,
+      "step": 80680
+    },
+    {
+      "epoch": 0.4029563784364154,
+      "grad_norm": 0.08766873925924301,
+      "learning_rate": 1.7956144083706726e-05,
+      "loss": 9.1088,
+      "step": 80690
+    },
+    {
+      "epoch": 0.40300631726135483,
+      "grad_norm": 0.10010071843862534,
+      "learning_rate": 1.7954642168765176e-05,
+      "loss": 9.1077,
+      "step": 80700
+    },
+    {
+      "epoch": 0.4030562560862943,
+      "grad_norm": 0.09243807941675186,
+      "learning_rate": 1.7953140253823626e-05,
+      "loss": 9.1047,
+      "step": 80710
+    },
+    {
+      "epoch": 0.40310619491123373,
+      "grad_norm": 0.0926956832408905,
+      "learning_rate": 1.7951638338882073e-05,
+      "loss": 9.1029,
+      "step": 80720
+    },
+    {
+      "epoch": 0.4031561337361732,
+      "grad_norm": 0.09393865615129471,
+      "learning_rate": 1.7950136423940523e-05,
+      "loss": 9.1034,
+      "step": 80730
+    },
+    {
+      "epoch": 0.4032060725611126,
+      "grad_norm": 0.09327403455972672,
+      "learning_rate": 1.7948634508998974e-05,
+      "loss": 9.1105,
+      "step": 80740
+    },
+    {
+      "epoch": 0.4032560113860521,
+      "grad_norm": 0.09919864684343338,
+      "learning_rate": 1.7947132594057424e-05,
+      "loss": 9.1182,
+      "step": 80750
+    },
+    {
+      "epoch": 0.4033059502109915,
+      "grad_norm": 0.09142395853996277,
+      "learning_rate": 1.7945630679115874e-05,
+      "loss": 9.1112,
+      "step": 80760
+    },
+    {
+      "epoch": 0.403355889035931,
+      "grad_norm": 0.09137079864740372,
+      "learning_rate": 1.794412876417432e-05,
+      "loss": 9.1173,
+      "step": 80770
+    },
+    {
+      "epoch": 0.4034058278608704,
+      "grad_norm": 0.09005033224821091,
+      "learning_rate": 1.7942626849232774e-05,
+      "loss": 9.116,
+      "step": 80780
+    },
+    {
+      "epoch": 0.4034557666858099,
+      "grad_norm": 0.09195708483457565,
+      "learning_rate": 1.794112493429122e-05,
+      "loss": 9.1215,
+      "step": 80790
+    },
+    {
+      "epoch": 0.4035057055107493,
+      "grad_norm": 0.09235309809446335,
+      "learning_rate": 1.793962301934967e-05,
+      "loss": 9.1151,
+      "step": 80800
+    },
+    {
+      "epoch": 0.4035556443356888,
+      "grad_norm": 0.09343268722295761,
+      "learning_rate": 1.793812110440812e-05,
+      "loss": 9.1282,
+      "step": 80810
+    },
+    {
+      "epoch": 0.4036055831606282,
+      "grad_norm": 0.09357087314128876,
+      "learning_rate": 1.793661918946657e-05,
+      "loss": 9.1123,
+      "step": 80820
+    },
+    {
+      "epoch": 0.4036555219855677,
+      "grad_norm": 0.08820904046297073,
+      "learning_rate": 1.7935117274525022e-05,
+      "loss": 9.1184,
+      "step": 80830
+    },
+    {
+      "epoch": 0.4037054608105071,
+      "grad_norm": 0.08932943642139435,
+      "learning_rate": 1.793361535958347e-05,
+      "loss": 9.1211,
+      "step": 80840
+    },
+    {
+      "epoch": 0.4037553996354466,
+      "grad_norm": 0.09997081011533737,
+      "learning_rate": 1.793211344464192e-05,
+      "loss": 9.0972,
+      "step": 80850
+    },
+    {
+      "epoch": 0.403805338460386,
+      "grad_norm": 0.09369473904371262,
+      "learning_rate": 1.793061152970037e-05,
+      "loss": 9.1223,
+      "step": 80860
+    },
+    {
+      "epoch": 0.4038552772853255,
+      "grad_norm": 0.09159389138221741,
+      "learning_rate": 1.792910961475882e-05,
+      "loss": 9.1404,
+      "step": 80870
+    },
+    {
+      "epoch": 0.4039052161102649,
+      "grad_norm": 0.09258721768856049,
+      "learning_rate": 1.792760769981727e-05,
+      "loss": 9.1047,
+      "step": 80880
+    },
+    {
+      "epoch": 0.4039551549352044,
+      "grad_norm": 0.0894956886768341,
+      "learning_rate": 1.7926105784875716e-05,
+      "loss": 9.1137,
+      "step": 80890
+    },
+    {
+      "epoch": 0.4040050937601438,
+      "grad_norm": 0.09068527817726135,
+      "learning_rate": 1.7924603869934166e-05,
+      "loss": 9.1043,
+      "step": 80900
+    },
+    {
+      "epoch": 0.4040550325850833,
+      "grad_norm": 0.09284330904483795,
+      "learning_rate": 1.7923101954992616e-05,
+      "loss": 9.1099,
+      "step": 80910
+    },
+    {
+      "epoch": 0.4041049714100227,
+      "grad_norm": 0.10179001837968826,
+      "learning_rate": 1.7921600040051067e-05,
+      "loss": 9.0992,
+      "step": 80920
+    },
+    {
+      "epoch": 0.4041549102349622,
+      "grad_norm": 0.09713611751794815,
+      "learning_rate": 1.7920098125109517e-05,
+      "loss": 9.1143,
+      "step": 80930
+    },
+    {
+      "epoch": 0.4042048490599016,
+      "grad_norm": 0.10156739503145218,
+      "learning_rate": 1.7918596210167964e-05,
+      "loss": 9.1069,
+      "step": 80940
+    },
+    {
+      "epoch": 0.4042547878848411,
+      "grad_norm": 0.09405851364135742,
+      "learning_rate": 1.7917094295226414e-05,
+      "loss": 9.1072,
+      "step": 80950
+    },
+    {
+      "epoch": 0.4043047267097805,
+      "grad_norm": 0.0946517214179039,
+      "learning_rate": 1.7915592380284864e-05,
+      "loss": 9.1207,
+      "step": 80960
+    },
+    {
+      "epoch": 0.40435466553472,
+      "grad_norm": 0.09372934699058533,
+      "learning_rate": 1.7914090465343314e-05,
+      "loss": 9.1101,
+      "step": 80970
+    },
+    {
+      "epoch": 0.4044046043596594,
+      "grad_norm": 0.08745318651199341,
+      "learning_rate": 1.7912588550401764e-05,
+      "loss": 9.1133,
+      "step": 80980
+    },
+    {
+      "epoch": 0.4044545431845989,
+      "grad_norm": 0.09735770523548126,
+      "learning_rate": 1.791108663546021e-05,
+      "loss": 9.1006,
+      "step": 80990
+    },
+    {
+      "epoch": 0.4045044820095383,
+      "grad_norm": 0.10448341071605682,
+      "learning_rate": 1.790958472051866e-05,
+      "loss": 9.0897,
+      "step": 81000
+    },
+    {
+      "epoch": 0.4045544208344778,
+      "grad_norm": 0.09520471096038818,
+      "learning_rate": 1.790808280557711e-05,
+      "loss": 9.1102,
+      "step": 81010
+    },
+    {
+      "epoch": 0.4046043596594172,
+      "grad_norm": 0.09410350024700165,
+      "learning_rate": 1.790658089063556e-05,
+      "loss": 9.1109,
+      "step": 81020
+    },
+    {
+      "epoch": 0.4046542984843567,
+      "grad_norm": 0.09858789294958115,
+      "learning_rate": 1.7905078975694012e-05,
+      "loss": 9.1101,
+      "step": 81030
+    },
+    {
+      "epoch": 0.4047042373092961,
+      "grad_norm": 0.09166577458381653,
+      "learning_rate": 1.790357706075246e-05,
+      "loss": 9.1214,
+      "step": 81040
+    },
+    {
+      "epoch": 0.4047541761342356,
+      "grad_norm": 0.0967087671160698,
+      "learning_rate": 1.790207514581091e-05,
+      "loss": 9.1012,
+      "step": 81050
+    },
+    {
+      "epoch": 0.404804114959175,
+      "grad_norm": 0.09214182943105698,
+      "learning_rate": 1.790057323086936e-05,
+      "loss": 9.1048,
+      "step": 81060
+    },
+    {
+      "epoch": 0.4048540537841145,
+      "grad_norm": 0.09241543710231781,
+      "learning_rate": 1.789907131592781e-05,
+      "loss": 9.1192,
+      "step": 81070
+    },
+    {
+      "epoch": 0.4049039926090539,
+      "grad_norm": 0.09916938096284866,
+      "learning_rate": 1.789756940098626e-05,
+      "loss": 9.101,
+      "step": 81080
+    },
+    {
+      "epoch": 0.40495393143399333,
+      "grad_norm": 0.09278815984725952,
+      "learning_rate": 1.7896067486044706e-05,
+      "loss": 9.1098,
+      "step": 81090
+    },
+    {
+      "epoch": 0.4050038702589328,
+      "grad_norm": 0.09353374689817429,
+      "learning_rate": 1.789456557110316e-05,
+      "loss": 9.1081,
+      "step": 81100
+    },
+    {
+      "epoch": 0.40505380908387223,
+      "grad_norm": 0.09639580547809601,
+      "learning_rate": 1.7893063656161606e-05,
+      "loss": 9.0991,
+      "step": 81110
+    },
+    {
+      "epoch": 0.4051037479088117,
+      "grad_norm": 0.09381458908319473,
+      "learning_rate": 1.7891561741220057e-05,
+      "loss": 9.1231,
+      "step": 81120
+    },
+    {
+      "epoch": 0.40515368673375113,
+      "grad_norm": 0.09496497362852097,
+      "learning_rate": 1.7890059826278507e-05,
+      "loss": 9.1022,
+      "step": 81130
+    },
+    {
+      "epoch": 0.4052036255586906,
+      "grad_norm": 0.09724625945091248,
+      "learning_rate": 1.7888557911336954e-05,
+      "loss": 9.1116,
+      "step": 81140
+    },
+    {
+      "epoch": 0.40525356438363,
+      "grad_norm": 0.09216088056564331,
+      "learning_rate": 1.7887055996395407e-05,
+      "loss": 9.1035,
+      "step": 81150
+    },
+    {
+      "epoch": 0.4053035032085695,
+      "grad_norm": 0.09402379393577576,
+      "learning_rate": 1.7885554081453854e-05,
+      "loss": 9.1074,
+      "step": 81160
+    },
+    {
+      "epoch": 0.4053534420335089,
+      "grad_norm": 0.09243787080049515,
+      "learning_rate": 1.7884052166512304e-05,
+      "loss": 9.0979,
+      "step": 81170
+    },
+    {
+      "epoch": 0.4054033808584484,
+      "grad_norm": 0.0954827070236206,
+      "learning_rate": 1.7882550251570754e-05,
+      "loss": 9.104,
+      "step": 81180
+    },
+    {
+      "epoch": 0.4054533196833878,
+      "grad_norm": 0.09604278206825256,
+      "learning_rate": 1.78810483366292e-05,
+      "loss": 9.0999,
+      "step": 81190
+    },
+    {
+      "epoch": 0.4055032585083273,
+      "grad_norm": 0.09881725907325745,
+      "learning_rate": 1.7879546421687655e-05,
+      "loss": 9.1049,
+      "step": 81200
+    },
+    {
+      "epoch": 0.4055531973332667,
+      "grad_norm": 0.09090153127908707,
+      "learning_rate": 1.78780445067461e-05,
+      "loss": 9.1028,
+      "step": 81210
+    },
+    {
+      "epoch": 0.4056031361582062,
+      "grad_norm": 0.09422878175973892,
+      "learning_rate": 1.787654259180455e-05,
+      "loss": 9.0912,
+      "step": 81220
+    },
+    {
+      "epoch": 0.4056530749831456,
+      "grad_norm": 0.09033548831939697,
+      "learning_rate": 1.7875040676863002e-05,
+      "loss": 9.1124,
+      "step": 81230
+    },
+    {
+      "epoch": 0.4057030138080851,
+      "grad_norm": 0.08588909357786179,
+      "learning_rate": 1.787353876192145e-05,
+      "loss": 9.0967,
+      "step": 81240
+    },
+    {
+      "epoch": 0.4057529526330245,
+      "grad_norm": 0.0940273180603981,
+      "learning_rate": 1.7872036846979902e-05,
+      "loss": 9.1105,
+      "step": 81250
+    },
+    {
+      "epoch": 0.405802891457964,
+      "grad_norm": 0.0950455591082573,
+      "learning_rate": 1.787053493203835e-05,
+      "loss": 9.1095,
+      "step": 81260
+    },
+    {
+      "epoch": 0.4058528302829034,
+      "grad_norm": 0.09447299689054489,
+      "learning_rate": 1.78690330170968e-05,
+      "loss": 9.1007,
+      "step": 81270
+    },
+    {
+      "epoch": 0.4059027691078429,
+      "grad_norm": 0.09144210070371628,
+      "learning_rate": 1.786753110215525e-05,
+      "loss": 9.1033,
+      "step": 81280
+    },
+    {
+      "epoch": 0.4059527079327823,
+      "grad_norm": 0.08806926012039185,
+      "learning_rate": 1.7866029187213696e-05,
+      "loss": 9.105,
+      "step": 81290
+    },
+    {
+      "epoch": 0.4060026467577218,
+      "grad_norm": 0.09366438537836075,
+      "learning_rate": 1.786452727227215e-05,
+      "loss": 9.0969,
+      "step": 81300
+    },
+    {
+      "epoch": 0.4060525855826612,
+      "grad_norm": 0.0918484702706337,
+      "learning_rate": 1.7863025357330596e-05,
+      "loss": 9.1047,
+      "step": 81310
+    },
+    {
+      "epoch": 0.4061025244076007,
+      "grad_norm": 0.09249420464038849,
+      "learning_rate": 1.7861523442389047e-05,
+      "loss": 9.0934,
+      "step": 81320
+    },
+    {
+      "epoch": 0.4061524632325401,
+      "grad_norm": 0.09651041775941849,
+      "learning_rate": 1.7860021527447497e-05,
+      "loss": 9.1173,
+      "step": 81330
+    },
+    {
+      "epoch": 0.4062024020574796,
+      "grad_norm": 0.09468764811754227,
+      "learning_rate": 1.7858519612505944e-05,
+      "loss": 9.0975,
+      "step": 81340
+    },
+    {
+      "epoch": 0.406252340882419,
+      "grad_norm": 0.09159006178379059,
+      "learning_rate": 1.7857017697564397e-05,
+      "loss": 9.108,
+      "step": 81350
+    },
+    {
+      "epoch": 0.4063022797073585,
+      "grad_norm": 0.08911147713661194,
+      "learning_rate": 1.7855515782622844e-05,
+      "loss": 9.1011,
+      "step": 81360
+    },
+    {
+      "epoch": 0.4063522185322979,
+      "grad_norm": 0.09557855129241943,
+      "learning_rate": 1.7854013867681294e-05,
+      "loss": 9.1114,
+      "step": 81370
+    },
+    {
+      "epoch": 0.4064021573572374,
+      "grad_norm": 0.09145300090312958,
+      "learning_rate": 1.7852511952739744e-05,
+      "loss": 9.0937,
+      "step": 81380
+    },
+    {
+      "epoch": 0.4064520961821768,
+      "grad_norm": 0.09059559553861618,
+      "learning_rate": 1.785101003779819e-05,
+      "loss": 9.1113,
+      "step": 81390
+    },
+    {
+      "epoch": 0.4065020350071163,
+      "grad_norm": 0.09117017686367035,
+      "learning_rate": 1.7849508122856645e-05,
+      "loss": 9.1032,
+      "step": 81400
+    },
+    {
+      "epoch": 0.4065519738320557,
+      "grad_norm": 0.09077339619398117,
+      "learning_rate": 1.784800620791509e-05,
+      "loss": 9.102,
+      "step": 81410
+    },
+    {
+      "epoch": 0.4066019126569952,
+      "grad_norm": 0.0927463248372078,
+      "learning_rate": 1.7846504292973545e-05,
+      "loss": 9.1111,
+      "step": 81420
+    },
+    {
+      "epoch": 0.4066518514819346,
+      "grad_norm": 0.09298672527074814,
+      "learning_rate": 1.7845002378031992e-05,
+      "loss": 9.1062,
+      "step": 81430
+    },
+    {
+      "epoch": 0.4067017903068741,
+      "grad_norm": 0.09268884360790253,
+      "learning_rate": 1.784350046309044e-05,
+      "loss": 9.0985,
+      "step": 81440
+    },
+    {
+      "epoch": 0.4067517291318135,
+      "grad_norm": 0.09739190340042114,
+      "learning_rate": 1.7841998548148892e-05,
+      "loss": 9.1097,
+      "step": 81450
+    },
+    {
+      "epoch": 0.406801667956753,
+      "grad_norm": 0.08956177532672882,
+      "learning_rate": 1.784049663320734e-05,
+      "loss": 9.104,
+      "step": 81460
+    },
+    {
+      "epoch": 0.4068516067816924,
+      "grad_norm": 0.09477641433477402,
+      "learning_rate": 1.7838994718265792e-05,
+      "loss": 9.1018,
+      "step": 81470
+    },
+    {
+      "epoch": 0.4069015456066319,
+      "grad_norm": 0.09133501350879669,
+      "learning_rate": 1.783749280332424e-05,
+      "loss": 9.1138,
+      "step": 81480
+    },
+    {
+      "epoch": 0.4069514844315713,
+      "grad_norm": 0.09000437706708908,
+      "learning_rate": 1.7835990888382686e-05,
+      "loss": 9.1056,
+      "step": 81490
+    },
+    {
+      "epoch": 0.4070014232565108,
+      "grad_norm": 0.09407982975244522,
+      "learning_rate": 1.783448897344114e-05,
+      "loss": 9.1159,
+      "step": 81500
+    },
+    {
+      "epoch": 0.4070513620814502,
+      "grad_norm": 0.09220529347658157,
+      "learning_rate": 1.7832987058499586e-05,
+      "loss": 9.0999,
+      "step": 81510
+    },
+    {
+      "epoch": 0.4071013009063897,
+      "grad_norm": 0.10050280392169952,
+      "learning_rate": 1.783148514355804e-05,
+      "loss": 9.0938,
+      "step": 81520
+    },
+    {
+      "epoch": 0.4071512397313291,
+      "grad_norm": 0.09240719676017761,
+      "learning_rate": 1.7829983228616487e-05,
+      "loss": 9.1099,
+      "step": 81530
+    },
+    {
+      "epoch": 0.4072011785562686,
+      "grad_norm": 0.09365271776914597,
+      "learning_rate": 1.7828481313674934e-05,
+      "loss": 9.105,
+      "step": 81540
+    },
+    {
+      "epoch": 0.407251117381208,
+      "grad_norm": 0.0895438864827156,
+      "learning_rate": 1.7826979398733387e-05,
+      "loss": 9.0933,
+      "step": 81550
+    },
+    {
+      "epoch": 0.4073010562061475,
+      "grad_norm": 0.09172806888818741,
+      "learning_rate": 1.7825477483791834e-05,
+      "loss": 9.0977,
+      "step": 81560
+    },
+    {
+      "epoch": 0.4073509950310869,
+      "grad_norm": 0.09493505954742432,
+      "learning_rate": 1.7823975568850287e-05,
+      "loss": 9.1043,
+      "step": 81570
+    },
+    {
+      "epoch": 0.4074009338560264,
+      "grad_norm": 0.08669907599687576,
+      "learning_rate": 1.7822473653908734e-05,
+      "loss": 9.0986,
+      "step": 81580
+    },
+    {
+      "epoch": 0.4074508726809658,
+      "grad_norm": 0.09328383207321167,
+      "learning_rate": 1.782097173896718e-05,
+      "loss": 9.0989,
+      "step": 81590
+    },
+    {
+      "epoch": 0.4075008115059053,
+      "grad_norm": 0.09149834513664246,
+      "learning_rate": 1.7819469824025635e-05,
+      "loss": 9.114,
+      "step": 81600
+    },
+    {
+      "epoch": 0.4075507503308447,
+      "grad_norm": 0.08791434019804001,
+      "learning_rate": 1.781796790908408e-05,
+      "loss": 9.1065,
+      "step": 81610
+    },
+    {
+      "epoch": 0.4076006891557842,
+      "grad_norm": 0.09626881778240204,
+      "learning_rate": 1.7816465994142535e-05,
+      "loss": 9.0977,
+      "step": 81620
+    },
+    {
+      "epoch": 0.4076506279807236,
+      "grad_norm": 0.08901573717594147,
+      "learning_rate": 1.7814964079200982e-05,
+      "loss": 9.1001,
+      "step": 81630
+    },
+    {
+      "epoch": 0.4077005668056631,
+      "grad_norm": 0.09143085032701492,
+      "learning_rate": 1.781346216425943e-05,
+      "loss": 9.0981,
+      "step": 81640
+    },
+    {
+      "epoch": 0.4077505056306025,
+      "grad_norm": 0.09885165840387344,
+      "learning_rate": 1.7811960249317882e-05,
+      "loss": 9.0968,
+      "step": 81650
+    },
+    {
+      "epoch": 0.407800444455542,
+      "grad_norm": 0.09109603613615036,
+      "learning_rate": 1.781045833437633e-05,
+      "loss": 9.0985,
+      "step": 81660
+    },
+    {
+      "epoch": 0.4078503832804814,
+      "grad_norm": 0.09030817449092865,
+      "learning_rate": 1.7808956419434783e-05,
+      "loss": 9.1018,
+      "step": 81670
+    },
+    {
+      "epoch": 0.4079003221054209,
+      "grad_norm": 0.09494989365339279,
+      "learning_rate": 1.780745450449323e-05,
+      "loss": 9.1095,
+      "step": 81680
+    },
+    {
+      "epoch": 0.4079502609303603,
+      "grad_norm": 0.09289959818124771,
+      "learning_rate": 1.7805952589551676e-05,
+      "loss": 9.0857,
+      "step": 81690
+    },
+    {
+      "epoch": 0.4080001997552998,
+      "grad_norm": 0.09349525719881058,
+      "learning_rate": 1.780445067461013e-05,
+      "loss": 9.097,
+      "step": 81700
+    },
+    {
+      "epoch": 0.4080501385802392,
+      "grad_norm": 0.08992131799459457,
+      "learning_rate": 1.7802948759668576e-05,
+      "loss": 9.0873,
+      "step": 81710
+    },
+    {
+      "epoch": 0.4081000774051787,
+      "grad_norm": 0.09782059490680695,
+      "learning_rate": 1.780144684472703e-05,
+      "loss": 9.0824,
+      "step": 81720
+    },
+    {
+      "epoch": 0.4081500162301181,
+      "grad_norm": 0.09610580652952194,
+      "learning_rate": 1.7799944929785477e-05,
+      "loss": 9.1044,
+      "step": 81730
+    },
+    {
+      "epoch": 0.4081999550550576,
+      "grad_norm": 0.09353209286928177,
+      "learning_rate": 1.7798443014843927e-05,
+      "loss": 9.0979,
+      "step": 81740
+    },
+    {
+      "epoch": 0.408249893879997,
+      "grad_norm": 0.08947628736495972,
+      "learning_rate": 1.7796941099902377e-05,
+      "loss": 9.0948,
+      "step": 81750
+    },
+    {
+      "epoch": 0.40829983270493647,
+      "grad_norm": 0.09303894639015198,
+      "learning_rate": 1.7795439184960824e-05,
+      "loss": 9.0996,
+      "step": 81760
+    },
+    {
+      "epoch": 0.4083497715298759,
+      "grad_norm": 0.09388189017772675,
+      "learning_rate": 1.7793937270019278e-05,
+      "loss": 9.1086,
+      "step": 81770
+    },
+    {
+      "epoch": 0.40839971035481537,
+      "grad_norm": 0.09600455313920975,
+      "learning_rate": 1.7792435355077724e-05,
+      "loss": 9.0987,
+      "step": 81780
+    },
+    {
+      "epoch": 0.4084496491797548,
+      "grad_norm": 0.091989666223526,
+      "learning_rate": 1.7790933440136174e-05,
+      "loss": 9.1066,
+      "step": 81790
+    },
+    {
+      "epoch": 0.40849958800469427,
+      "grad_norm": 0.09617997705936432,
+      "learning_rate": 1.7789431525194625e-05,
+      "loss": 9.1038,
+      "step": 81800
+    },
+    {
+      "epoch": 0.4085495268296337,
+      "grad_norm": 0.09295359253883362,
+      "learning_rate": 1.778792961025307e-05,
+      "loss": 9.0986,
+      "step": 81810
+    },
+    {
+      "epoch": 0.40859946565457317,
+      "grad_norm": 0.09218577295541763,
+      "learning_rate": 1.7786427695311525e-05,
+      "loss": 9.1062,
+      "step": 81820
+    },
+    {
+      "epoch": 0.4086494044795126,
+      "grad_norm": 0.09626112878322601,
+      "learning_rate": 1.7784925780369972e-05,
+      "loss": 9.0865,
+      "step": 81830
+    },
+    {
+      "epoch": 0.40869934330445207,
+      "grad_norm": 0.08941785991191864,
+      "learning_rate": 1.7783423865428422e-05,
+      "loss": 9.0962,
+      "step": 81840
+    },
+    {
+      "epoch": 0.4087492821293915,
+      "grad_norm": 0.0907067134976387,
+      "learning_rate": 1.7781921950486872e-05,
+      "loss": 9.1065,
+      "step": 81850
+    },
+    {
+      "epoch": 0.40879922095433097,
+      "grad_norm": 0.08822061866521835,
+      "learning_rate": 1.778042003554532e-05,
+      "loss": 9.1054,
+      "step": 81860
+    },
+    {
+      "epoch": 0.4088491597792704,
+      "grad_norm": 0.08900550752878189,
+      "learning_rate": 1.7778918120603773e-05,
+      "loss": 9.1029,
+      "step": 81870
+    },
+    {
+      "epoch": 0.40889909860420987,
+      "grad_norm": 0.09885650873184204,
+      "learning_rate": 1.777741620566222e-05,
+      "loss": 9.0903,
+      "step": 81880
+    },
+    {
+      "epoch": 0.4089490374291493,
+      "grad_norm": 0.09264232963323593,
+      "learning_rate": 1.777591429072067e-05,
+      "loss": 9.0968,
+      "step": 81890
+    },
+    {
+      "epoch": 0.40899897625408876,
+      "grad_norm": 0.0919133648276329,
+      "learning_rate": 1.777441237577912e-05,
+      "loss": 9.0921,
+      "step": 81900
+    },
+    {
+      "epoch": 0.4090489150790282,
+      "grad_norm": 0.09178303927183151,
+      "learning_rate": 1.7772910460837566e-05,
+      "loss": 9.0978,
+      "step": 81910
+    },
+    {
+      "epoch": 0.40909885390396766,
+      "grad_norm": 0.09064862877130508,
+      "learning_rate": 1.777140854589602e-05,
+      "loss": 9.1053,
+      "step": 81920
+    },
+    {
+      "epoch": 0.4091487927289071,
+      "grad_norm": 0.0949510857462883,
+      "learning_rate": 1.7769906630954467e-05,
+      "loss": 9.1046,
+      "step": 81930
+    },
+    {
+      "epoch": 0.40919873155384656,
+      "grad_norm": 0.09381455928087234,
+      "learning_rate": 1.7768404716012917e-05,
+      "loss": 9.1019,
+      "step": 81940
+    },
+    {
+      "epoch": 0.409248670378786,
+      "grad_norm": 0.0917896181344986,
+      "learning_rate": 1.7766902801071367e-05,
+      "loss": 9.0914,
+      "step": 81950
+    },
+    {
+      "epoch": 0.40929860920372546,
+      "grad_norm": 0.09546224027872086,
+      "learning_rate": 1.7765400886129814e-05,
+      "loss": 9.0897,
+      "step": 81960
+    },
+    {
+      "epoch": 0.4093485480286649,
+      "grad_norm": 0.09134580194950104,
+      "learning_rate": 1.7763898971188268e-05,
+      "loss": 9.0924,
+      "step": 81970
+    },
+    {
+      "epoch": 0.40939848685360436,
+      "grad_norm": 0.09422563761472702,
+      "learning_rate": 1.7762397056246714e-05,
+      "loss": 9.1046,
+      "step": 81980
+    },
+    {
+      "epoch": 0.4094484256785438,
+      "grad_norm": 0.09437835216522217,
+      "learning_rate": 1.7760895141305164e-05,
+      "loss": 9.1111,
+      "step": 81990
+    },
+    {
+      "epoch": 0.40949836450348326,
+      "grad_norm": 0.09388506412506104,
+      "learning_rate": 1.7759393226363615e-05,
+      "loss": 9.0865,
+      "step": 82000
+    },
+    {
+      "epoch": 0.4095483033284227,
+      "grad_norm": 0.09779687970876694,
+      "learning_rate": 1.775789131142206e-05,
+      "loss": 9.0946,
+      "step": 82010
+    },
+    {
+      "epoch": 0.40959824215336216,
+      "grad_norm": 0.08977722376585007,
+      "learning_rate": 1.7756389396480515e-05,
+      "loss": 9.1029,
+      "step": 82020
+    },
+    {
+      "epoch": 0.4096481809783016,
+      "grad_norm": 0.09358174353837967,
+      "learning_rate": 1.7754887481538962e-05,
+      "loss": 9.0842,
+      "step": 82030
+    },
+    {
+      "epoch": 0.40969811980324106,
+      "grad_norm": 0.09870479255914688,
+      "learning_rate": 1.7753385566597412e-05,
+      "loss": 9.0784,
+      "step": 82040
+    },
+    {
+      "epoch": 0.4097480586281805,
+      "grad_norm": 0.09142067283391953,
+      "learning_rate": 1.7751883651655862e-05,
+      "loss": 9.0896,
+      "step": 82050
+    },
+    {
+      "epoch": 0.40979799745311996,
+      "grad_norm": 0.08930368721485138,
+      "learning_rate": 1.7750381736714312e-05,
+      "loss": 9.1012,
+      "step": 82060
+    },
+    {
+      "epoch": 0.4098479362780594,
+      "grad_norm": 0.09085031598806381,
+      "learning_rate": 1.7748879821772763e-05,
+      "loss": 9.0979,
+      "step": 82070
+    },
+    {
+      "epoch": 0.4098978751029988,
+      "grad_norm": 0.09649605304002762,
+      "learning_rate": 1.774737790683121e-05,
+      "loss": 9.0945,
+      "step": 82080
+    },
+    {
+      "epoch": 0.4099478139279383,
+      "grad_norm": 0.09378213435411453,
+      "learning_rate": 1.774587599188966e-05,
+      "loss": 9.1019,
+      "step": 82090
+    },
+    {
+      "epoch": 0.4099977527528777,
+      "grad_norm": 0.09320469200611115,
+      "learning_rate": 1.774437407694811e-05,
+      "loss": 9.0815,
+      "step": 82100
+    },
+    {
+      "epoch": 0.4100476915778172,
+      "grad_norm": 0.09025924652814865,
+      "learning_rate": 1.774287216200656e-05,
+      "loss": 9.1096,
+      "step": 82110
+    },
+    {
+      "epoch": 0.4100976304027566,
+      "grad_norm": 0.09439290314912796,
+      "learning_rate": 1.774137024706501e-05,
+      "loss": 9.1072,
+      "step": 82120
+    },
+    {
+      "epoch": 0.4101475692276961,
+      "grad_norm": 0.09332076460123062,
+      "learning_rate": 1.7739868332123457e-05,
+      "loss": 9.104,
+      "step": 82130
+    },
+    {
+      "epoch": 0.4101975080526355,
+      "grad_norm": 0.10301212221384048,
+      "learning_rate": 1.7738366417181907e-05,
+      "loss": 9.0906,
+      "step": 82140
+    },
+    {
+      "epoch": 0.410247446877575,
+      "grad_norm": 0.08951026201248169,
+      "learning_rate": 1.7736864502240357e-05,
+      "loss": 9.1012,
+      "step": 82150
+    },
+    {
+      "epoch": 0.4102973857025144,
+      "grad_norm": 0.09387951344251633,
+      "learning_rate": 1.7735362587298807e-05,
+      "loss": 9.1009,
+      "step": 82160
+    },
+    {
+      "epoch": 0.41034732452745387,
+      "grad_norm": 0.0935429260134697,
+      "learning_rate": 1.7733860672357258e-05,
+      "loss": 9.0949,
+      "step": 82170
+    },
+    {
+      "epoch": 0.4103972633523933,
+      "grad_norm": 0.09872562438249588,
+      "learning_rate": 1.7732358757415704e-05,
+      "loss": 9.0956,
+      "step": 82180
+    },
+    {
+      "epoch": 0.41044720217733277,
+      "grad_norm": 0.09391787648200989,
+      "learning_rate": 1.7730856842474155e-05,
+      "loss": 9.0856,
+      "step": 82190
+    },
+    {
+      "epoch": 0.4104971410022722,
+      "grad_norm": 0.09664194285869598,
+      "learning_rate": 1.7729354927532605e-05,
+      "loss": 9.0931,
+      "step": 82200
+    },
+    {
+      "epoch": 0.41054707982721167,
+      "grad_norm": 0.10238590836524963,
+      "learning_rate": 1.7727853012591055e-05,
+      "loss": 9.0853,
+      "step": 82210
+    },
+    {
+      "epoch": 0.4105970186521511,
+      "grad_norm": 0.08652331680059433,
+      "learning_rate": 1.7726351097649505e-05,
+      "loss": 9.1032,
+      "step": 82220
+    },
+    {
+      "epoch": 0.41064695747709057,
+      "grad_norm": 0.09455471485853195,
+      "learning_rate": 1.7724849182707952e-05,
+      "loss": 9.0888,
+      "step": 82230
+    },
+    {
+      "epoch": 0.41069689630203,
+      "grad_norm": 0.09298815578222275,
+      "learning_rate": 1.7723347267766402e-05,
+      "loss": 9.0829,
+      "step": 82240
+    },
+    {
+      "epoch": 0.41074683512696947,
+      "grad_norm": 0.09878503531217575,
+      "learning_rate": 1.7721845352824852e-05,
+      "loss": 9.0886,
+      "step": 82250
+    },
+    {
+      "epoch": 0.4107967739519089,
+      "grad_norm": 0.09556316584348679,
+      "learning_rate": 1.7720343437883302e-05,
+      "loss": 9.0951,
+      "step": 82260
+    },
+    {
+      "epoch": 0.41084671277684837,
+      "grad_norm": 0.094581738114357,
+      "learning_rate": 1.7718841522941753e-05,
+      "loss": 9.0941,
+      "step": 82270
+    },
+    {
+      "epoch": 0.4108966516017878,
+      "grad_norm": 0.09531483799219131,
+      "learning_rate": 1.77173396080002e-05,
+      "loss": 9.0941,
+      "step": 82280
+    },
+    {
+      "epoch": 0.41094659042672727,
+      "grad_norm": 0.09399186819791794,
+      "learning_rate": 1.771583769305865e-05,
+      "loss": 9.0851,
+      "step": 82290
+    },
+    {
+      "epoch": 0.4109965292516667,
+      "grad_norm": 0.09608032554388046,
+      "learning_rate": 1.77143357781171e-05,
+      "loss": 9.0931,
+      "step": 82300
+    },
+    {
+      "epoch": 0.41104646807660616,
+      "grad_norm": 0.0954754576086998,
+      "learning_rate": 1.771283386317555e-05,
+      "loss": 9.1091,
+      "step": 82310
+    },
+    {
+      "epoch": 0.4110964069015456,
+      "grad_norm": 0.08752164989709854,
+      "learning_rate": 1.7711331948234e-05,
+      "loss": 9.0904,
+      "step": 82320
+    },
+    {
+      "epoch": 0.41114634572648506,
+      "grad_norm": 0.09011632949113846,
+      "learning_rate": 1.7709830033292447e-05,
+      "loss": 9.1022,
+      "step": 82330
+    },
+    {
+      "epoch": 0.4111962845514245,
+      "grad_norm": 0.08889009058475494,
+      "learning_rate": 1.7708328118350897e-05,
+      "loss": 9.0813,
+      "step": 82340
+    },
+    {
+      "epoch": 0.41124622337636396,
+      "grad_norm": 0.0934472531080246,
+      "learning_rate": 1.7706826203409347e-05,
+      "loss": 9.084,
+      "step": 82350
+    },
+    {
+      "epoch": 0.4112961622013034,
+      "grad_norm": 0.0992710068821907,
+      "learning_rate": 1.7705324288467797e-05,
+      "loss": 9.1012,
+      "step": 82360
+    },
+    {
+      "epoch": 0.41134610102624286,
+      "grad_norm": 0.0951753556728363,
+      "learning_rate": 1.7703822373526248e-05,
+      "loss": 9.0896,
+      "step": 82370
+    },
+    {
+      "epoch": 0.4113960398511823,
+      "grad_norm": 0.09202361106872559,
+      "learning_rate": 1.7702320458584694e-05,
+      "loss": 9.0929,
+      "step": 82380
+    },
+    {
+      "epoch": 0.41144597867612176,
+      "grad_norm": 0.09348968416452408,
+      "learning_rate": 1.7700818543643145e-05,
+      "loss": 9.0928,
+      "step": 82390
+    },
+    {
+      "epoch": 0.4114959175010612,
+      "grad_norm": 0.0965779572725296,
+      "learning_rate": 1.7699316628701595e-05,
+      "loss": 9.086,
+      "step": 82400
+    },
+    {
+      "epoch": 0.41154585632600066,
+      "grad_norm": 0.09712937474250793,
+      "learning_rate": 1.7697814713760045e-05,
+      "loss": 9.085,
+      "step": 82410
+    },
+    {
+      "epoch": 0.4115957951509401,
+      "grad_norm": 0.09527904540300369,
+      "learning_rate": 1.7696312798818495e-05,
+      "loss": 9.0898,
+      "step": 82420
+    },
+    {
+      "epoch": 0.41164573397587956,
+      "grad_norm": 0.09039422124624252,
+      "learning_rate": 1.7694810883876945e-05,
+      "loss": 9.0958,
+      "step": 82430
+    },
+    {
+      "epoch": 0.411695672800819,
+      "grad_norm": 0.09513916820287704,
+      "learning_rate": 1.7693308968935392e-05,
+      "loss": 9.085,
+      "step": 82440
+    },
+    {
+      "epoch": 0.41174561162575846,
+      "grad_norm": 0.09290063381195068,
+      "learning_rate": 1.7691807053993842e-05,
+      "loss": 9.0877,
+      "step": 82450
+    },
+    {
+      "epoch": 0.4117955504506979,
+      "grad_norm": 0.09197252988815308,
+      "learning_rate": 1.7690305139052292e-05,
+      "loss": 9.0885,
+      "step": 82460
+    },
+    {
+      "epoch": 0.41184548927563736,
+      "grad_norm": 0.0950205847620964,
+      "learning_rate": 1.7688803224110743e-05,
+      "loss": 9.0964,
+      "step": 82470
+    },
+    {
+      "epoch": 0.4118954281005768,
+      "grad_norm": 0.09569496661424637,
+      "learning_rate": 1.7687301309169193e-05,
+      "loss": 9.0962,
+      "step": 82480
+    },
+    {
+      "epoch": 0.41194536692551625,
+      "grad_norm": 0.09163631498813629,
+      "learning_rate": 1.768579939422764e-05,
+      "loss": 9.101,
+      "step": 82490
+    },
+    {
+      "epoch": 0.4119953057504557,
+      "grad_norm": 0.09723185002803802,
+      "learning_rate": 1.768429747928609e-05,
+      "loss": 9.0751,
+      "step": 82500
+    },
+    {
+      "epoch": 0.41204524457539515,
+      "grad_norm": 0.09386081993579865,
+      "learning_rate": 1.768279556434454e-05,
+      "loss": 9.0936,
+      "step": 82510
+    },
+    {
+      "epoch": 0.4120951834003346,
+      "grad_norm": 0.09162143617868423,
+      "learning_rate": 1.768129364940299e-05,
+      "loss": 9.1021,
+      "step": 82520
+    },
+    {
+      "epoch": 0.41214512222527405,
+      "grad_norm": 0.09791895002126694,
+      "learning_rate": 1.767979173446144e-05,
+      "loss": 9.0871,
+      "step": 82530
+    },
+    {
+      "epoch": 0.4121950610502135,
+      "grad_norm": 0.09404843300580978,
+      "learning_rate": 1.7678289819519887e-05,
+      "loss": 9.0875,
+      "step": 82540
+    },
+    {
+      "epoch": 0.41224499987515295,
+      "grad_norm": 0.09097657352685928,
+      "learning_rate": 1.7676787904578337e-05,
+      "loss": 9.0881,
+      "step": 82550
+    },
+    {
+      "epoch": 0.4122949387000924,
+      "grad_norm": 0.090410977602005,
+      "learning_rate": 1.7675285989636787e-05,
+      "loss": 9.0898,
+      "step": 82560
+    },
+    {
+      "epoch": 0.41234487752503185,
+      "grad_norm": 0.09892985969781876,
+      "learning_rate": 1.7673784074695238e-05,
+      "loss": 9.0896,
+      "step": 82570
+    },
+    {
+      "epoch": 0.41239481634997127,
+      "grad_norm": 0.09155625849962234,
+      "learning_rate": 1.7672282159753688e-05,
+      "loss": 9.0865,
+      "step": 82580
+    },
+    {
+      "epoch": 0.41244475517491075,
+      "grad_norm": 0.09150595217943192,
+      "learning_rate": 1.7670780244812135e-05,
+      "loss": 9.0801,
+      "step": 82590
+    },
+    {
+      "epoch": 0.41249469399985017,
+      "grad_norm": 0.09165318310260773,
+      "learning_rate": 1.7669278329870585e-05,
+      "loss": 9.0751,
+      "step": 82600
+    },
+    {
+      "epoch": 0.41254463282478965,
+      "grad_norm": 0.09589849412441254,
+      "learning_rate": 1.7667776414929035e-05,
+      "loss": 9.0826,
+      "step": 82610
+    },
+    {
+      "epoch": 0.41259457164972907,
+      "grad_norm": 0.10229449719190598,
+      "learning_rate": 1.7666274499987485e-05,
+      "loss": 9.0724,
+      "step": 82620
+    },
+    {
+      "epoch": 0.41264451047466855,
+      "grad_norm": 0.09139035642147064,
+      "learning_rate": 1.7664772585045935e-05,
+      "loss": 9.091,
+      "step": 82630
+    },
+    {
+      "epoch": 0.41269444929960797,
+      "grad_norm": 0.09130216389894485,
+      "learning_rate": 1.7663270670104382e-05,
+      "loss": 9.0797,
+      "step": 82640
+    },
+    {
+      "epoch": 0.41274438812454745,
+      "grad_norm": 0.08872245997190475,
+      "learning_rate": 1.7661768755162832e-05,
+      "loss": 9.0946,
+      "step": 82650
+    },
+    {
+      "epoch": 0.41279432694948687,
+      "grad_norm": 0.09483634680509567,
+      "learning_rate": 1.7660266840221282e-05,
+      "loss": 9.0982,
+      "step": 82660
+    },
+    {
+      "epoch": 0.41284426577442634,
+      "grad_norm": 0.10040302574634552,
+      "learning_rate": 1.7658764925279733e-05,
+      "loss": 9.0882,
+      "step": 82670
+    },
+    {
+      "epoch": 0.41289420459936577,
+      "grad_norm": 0.09902846813201904,
+      "learning_rate": 1.7657263010338183e-05,
+      "loss": 9.0838,
+      "step": 82680
+    },
+    {
+      "epoch": 0.41294414342430524,
+      "grad_norm": 0.090021513402462,
+      "learning_rate": 1.765576109539663e-05,
+      "loss": 9.0854,
+      "step": 82690
+    },
+    {
+      "epoch": 0.41299408224924467,
+      "grad_norm": 0.09512148797512054,
+      "learning_rate": 1.765425918045508e-05,
+      "loss": 9.0922,
+      "step": 82700
+    },
+    {
+      "epoch": 0.41304402107418414,
+      "grad_norm": 0.09530332684516907,
+      "learning_rate": 1.765275726551353e-05,
+      "loss": 9.0939,
+      "step": 82710
+    },
+    {
+      "epoch": 0.41309395989912356,
+      "grad_norm": 0.09510703384876251,
+      "learning_rate": 1.765125535057198e-05,
+      "loss": 9.0849,
+      "step": 82720
+    },
+    {
+      "epoch": 0.41314389872406304,
+      "grad_norm": 0.09195013344287872,
+      "learning_rate": 1.764975343563043e-05,
+      "loss": 9.0738,
+      "step": 82730
+    },
+    {
+      "epoch": 0.41319383754900246,
+      "grad_norm": 0.09670493006706238,
+      "learning_rate": 1.7648251520688877e-05,
+      "loss": 9.0946,
+      "step": 82740
+    },
+    {
+      "epoch": 0.41324377637394194,
+      "grad_norm": 0.09696520864963531,
+      "learning_rate": 1.764674960574733e-05,
+      "loss": 9.0757,
+      "step": 82750
+    },
+    {
+      "epoch": 0.41329371519888136,
+      "grad_norm": 0.09648612141609192,
+      "learning_rate": 1.7645247690805777e-05,
+      "loss": 9.0891,
+      "step": 82760
+    },
+    {
+      "epoch": 0.41334365402382084,
+      "grad_norm": 0.09383885562419891,
+      "learning_rate": 1.7643745775864228e-05,
+      "loss": 9.0929,
+      "step": 82770
+    },
+    {
+      "epoch": 0.41339359284876026,
+      "grad_norm": 0.09602250903844833,
+      "learning_rate": 1.7642243860922678e-05,
+      "loss": 9.0904,
+      "step": 82780
+    },
+    {
+      "epoch": 0.41344353167369974,
+      "grad_norm": 0.08740779012441635,
+      "learning_rate": 1.7640741945981125e-05,
+      "loss": 9.0999,
+      "step": 82790
+    },
+    {
+      "epoch": 0.41349347049863916,
+      "grad_norm": 0.10102899372577667,
+      "learning_rate": 1.7639240031039578e-05,
+      "loss": 9.0845,
+      "step": 82800
+    },
+    {
+      "epoch": 0.41354340932357864,
+      "grad_norm": 0.10466084629297256,
+      "learning_rate": 1.7637738116098025e-05,
+      "loss": 9.0695,
+      "step": 82810
+    },
+    {
+      "epoch": 0.41359334814851806,
+      "grad_norm": 0.09442269057035446,
+      "learning_rate": 1.7636236201156475e-05,
+      "loss": 9.0806,
+      "step": 82820
+    },
+    {
+      "epoch": 0.41364328697345754,
+      "grad_norm": 0.09268420934677124,
+      "learning_rate": 1.7634734286214925e-05,
+      "loss": 9.1006,
+      "step": 82830
+    },
+    {
+      "epoch": 0.41369322579839696,
+      "grad_norm": 0.09057000279426575,
+      "learning_rate": 1.7633232371273372e-05,
+      "loss": 9.0809,
+      "step": 82840
+    },
+    {
+      "epoch": 0.41374316462333643,
+      "grad_norm": 0.0951695516705513,
+      "learning_rate": 1.7631730456331826e-05,
+      "loss": 9.0833,
+      "step": 82850
+    },
+    {
+      "epoch": 0.41379310344827586,
+      "grad_norm": 0.09882236272096634,
+      "learning_rate": 1.7630228541390272e-05,
+      "loss": 9.0878,
+      "step": 82860
+    },
+    {
+      "epoch": 0.41384304227321533,
+      "grad_norm": 0.09375642985105515,
+      "learning_rate": 1.7628726626448723e-05,
+      "loss": 9.0873,
+      "step": 82870
+    },
+    {
+      "epoch": 0.41389298109815476,
+      "grad_norm": 0.08775999397039413,
+      "learning_rate": 1.7627224711507173e-05,
+      "loss": 9.1052,
+      "step": 82880
+    },
+    {
+      "epoch": 0.41394291992309423,
+      "grad_norm": 0.09294060617685318,
+      "learning_rate": 1.762572279656562e-05,
+      "loss": 9.0823,
+      "step": 82890
+    },
+    {
+      "epoch": 0.41399285874803365,
+      "grad_norm": 0.10115856677293777,
+      "learning_rate": 1.7624220881624073e-05,
+      "loss": 9.0824,
+      "step": 82900
+    },
+    {
+      "epoch": 0.41404279757297313,
+      "grad_norm": 0.09972229599952698,
+      "learning_rate": 1.762271896668252e-05,
+      "loss": 9.0816,
+      "step": 82910
+    },
+    {
+      "epoch": 0.41409273639791255,
+      "grad_norm": 0.0935303345322609,
+      "learning_rate": 1.762121705174097e-05,
+      "loss": 9.0848,
+      "step": 82920
+    },
+    {
+      "epoch": 0.41414267522285203,
+      "grad_norm": 0.09236697852611542,
+      "learning_rate": 1.761971513679942e-05,
+      "loss": 9.091,
+      "step": 82930
+    },
+    {
+      "epoch": 0.41419261404779145,
+      "grad_norm": 0.08565404266119003,
+      "learning_rate": 1.7618213221857867e-05,
+      "loss": 9.0823,
+      "step": 82940
+    },
+    {
+      "epoch": 0.41424255287273093,
+      "grad_norm": 0.09559847414493561,
+      "learning_rate": 1.761671130691632e-05,
+      "loss": 9.0797,
+      "step": 82950
+    },
+    {
+      "epoch": 0.41429249169767035,
+      "grad_norm": 0.09396830946207047,
+      "learning_rate": 1.7615209391974767e-05,
+      "loss": 9.088,
+      "step": 82960
+    },
+    {
+      "epoch": 0.41434243052260983,
+      "grad_norm": 0.09308158606290817,
+      "learning_rate": 1.7613707477033218e-05,
+      "loss": 9.0753,
+      "step": 82970
+    },
+    {
+      "epoch": 0.41439236934754925,
+      "grad_norm": 0.10932270437479019,
+      "learning_rate": 1.7612205562091668e-05,
+      "loss": 9.0863,
+      "step": 82980
+    },
+    {
+      "epoch": 0.4144423081724887,
+      "grad_norm": 0.09359308332204819,
+      "learning_rate": 1.7610703647150115e-05,
+      "loss": 9.084,
+      "step": 82990
+    },
+    {
+      "epoch": 0.41449224699742815,
+      "grad_norm": 0.09583938866853714,
+      "learning_rate": 1.7609201732208568e-05,
+      "loss": 9.075,
+      "step": 83000
+    },
+    {
+      "epoch": 0.4145421858223676,
+      "grad_norm": 0.09451348334550858,
+      "learning_rate": 1.7607699817267015e-05,
+      "loss": 9.0976,
+      "step": 83010
+    },
+    {
+      "epoch": 0.41459212464730705,
+      "grad_norm": 0.09366924315690994,
+      "learning_rate": 1.7606197902325465e-05,
+      "loss": 9.0789,
+      "step": 83020
+    },
+    {
+      "epoch": 0.4146420634722465,
+      "grad_norm": 0.09320022910833359,
+      "learning_rate": 1.7604695987383915e-05,
+      "loss": 9.0808,
+      "step": 83030
+    },
+    {
+      "epoch": 0.41469200229718595,
+      "grad_norm": 0.09090333431959152,
+      "learning_rate": 1.7603194072442362e-05,
+      "loss": 9.0839,
+      "step": 83040
+    },
+    {
+      "epoch": 0.4147419411221254,
+      "grad_norm": 0.093370221555233,
+      "learning_rate": 1.7601692157500816e-05,
+      "loss": 9.0735,
+      "step": 83050
+    },
+    {
+      "epoch": 0.41479187994706485,
+      "grad_norm": 0.09438198059797287,
+      "learning_rate": 1.7600190242559262e-05,
+      "loss": 9.085,
+      "step": 83060
+    },
+    {
+      "epoch": 0.41484181877200427,
+      "grad_norm": 0.09503324329853058,
+      "learning_rate": 1.7598688327617716e-05,
+      "loss": 9.087,
+      "step": 83070
+    },
+    {
+      "epoch": 0.41489175759694374,
+      "grad_norm": 0.09370449930429459,
+      "learning_rate": 1.7597186412676163e-05,
+      "loss": 9.0811,
+      "step": 83080
+    },
+    {
+      "epoch": 0.41494169642188317,
+      "grad_norm": 0.09288004785776138,
+      "learning_rate": 1.759568449773461e-05,
+      "loss": 9.0904,
+      "step": 83090
+    },
+    {
+      "epoch": 0.41499163524682264,
+      "grad_norm": 0.09713056683540344,
+      "learning_rate": 1.7594182582793063e-05,
+      "loss": 9.0817,
+      "step": 83100
+    },
+    {
+      "epoch": 0.41504157407176206,
+      "grad_norm": 0.09737755358219147,
+      "learning_rate": 1.759268066785151e-05,
+      "loss": 9.071,
+      "step": 83110
+    },
+    {
+      "epoch": 0.41509151289670154,
+      "grad_norm": 0.10059583187103271,
+      "learning_rate": 1.7591178752909963e-05,
+      "loss": 9.0756,
+      "step": 83120
+    },
+    {
+      "epoch": 0.41514145172164096,
+      "grad_norm": 0.0912448912858963,
+      "learning_rate": 1.758967683796841e-05,
+      "loss": 9.0848,
+      "step": 83130
+    },
+    {
+      "epoch": 0.41519139054658044,
+      "grad_norm": 0.09705382585525513,
+      "learning_rate": 1.7588174923026857e-05,
+      "loss": 9.0865,
+      "step": 83140
+    },
+    {
+      "epoch": 0.41524132937151986,
+      "grad_norm": 0.08912192285060883,
+      "learning_rate": 1.758667300808531e-05,
+      "loss": 9.0884,
+      "step": 83150
+    },
+    {
+      "epoch": 0.41529126819645934,
+      "grad_norm": 0.09693977981805801,
+      "learning_rate": 1.7585171093143757e-05,
+      "loss": 9.081,
+      "step": 83160
+    },
+    {
+      "epoch": 0.41534120702139876,
+      "grad_norm": 0.09407562017440796,
+      "learning_rate": 1.758366917820221e-05,
+      "loss": 9.0875,
+      "step": 83170
+    },
+    {
+      "epoch": 0.41539114584633824,
+      "grad_norm": 0.0872037410736084,
+      "learning_rate": 1.7582167263260658e-05,
+      "loss": 9.0934,
+      "step": 83180
+    },
+    {
+      "epoch": 0.41544108467127766,
+      "grad_norm": 0.09205826371908188,
+      "learning_rate": 1.7580665348319105e-05,
+      "loss": 9.0891,
+      "step": 83190
+    },
+    {
+      "epoch": 0.41549102349621714,
+      "grad_norm": 0.09072985500097275,
+      "learning_rate": 1.7579163433377558e-05,
+      "loss": 9.0847,
+      "step": 83200
+    },
+    {
+      "epoch": 0.41554096232115656,
+      "grad_norm": 0.08977384865283966,
+      "learning_rate": 1.7577661518436005e-05,
+      "loss": 9.0795,
+      "step": 83210
+    },
+    {
+      "epoch": 0.41559090114609604,
+      "grad_norm": 0.0946430042386055,
+      "learning_rate": 1.757615960349446e-05,
+      "loss": 9.0804,
+      "step": 83220
+    },
+    {
+      "epoch": 0.41564083997103546,
+      "grad_norm": 0.09836602956056595,
+      "learning_rate": 1.7574657688552905e-05,
+      "loss": 9.0745,
+      "step": 83230
+    },
+    {
+      "epoch": 0.41569077879597494,
+      "grad_norm": 0.09133485704660416,
+      "learning_rate": 1.7573155773611352e-05,
+      "loss": 9.0921,
+      "step": 83240
+    },
+    {
+      "epoch": 0.41574071762091436,
+      "grad_norm": 0.093567855656147,
+      "learning_rate": 1.7571653858669806e-05,
+      "loss": 9.0733,
+      "step": 83250
+    },
+    {
+      "epoch": 0.41579065644585383,
+      "grad_norm": 0.09800833463668823,
+      "learning_rate": 1.7570151943728252e-05,
+      "loss": 9.0626,
+      "step": 83260
+    },
+    {
+      "epoch": 0.41584059527079326,
+      "grad_norm": 0.1007581278681755,
+      "learning_rate": 1.7568650028786706e-05,
+      "loss": 9.0695,
+      "step": 83270
+    },
+    {
+      "epoch": 0.41589053409573273,
+      "grad_norm": 0.09314744919538498,
+      "learning_rate": 1.7567148113845153e-05,
+      "loss": 9.0963,
+      "step": 83280
+    },
+    {
+      "epoch": 0.41594047292067216,
+      "grad_norm": 0.09255973249673843,
+      "learning_rate": 1.75656461989036e-05,
+      "loss": 9.0841,
+      "step": 83290
+    },
+    {
+      "epoch": 0.41599041174561163,
+      "grad_norm": 0.0947415828704834,
+      "learning_rate": 1.7564144283962053e-05,
+      "loss": 9.0759,
+      "step": 83300
+    },
+    {
+      "epoch": 0.41604035057055105,
+      "grad_norm": 0.09376344829797745,
+      "learning_rate": 1.75626423690205e-05,
+      "loss": 9.0787,
+      "step": 83310
+    },
+    {
+      "epoch": 0.41609028939549053,
+      "grad_norm": 0.09048786014318466,
+      "learning_rate": 1.7561140454078954e-05,
+      "loss": 9.0731,
+      "step": 83320
+    },
+    {
+      "epoch": 0.41614022822042995,
+      "grad_norm": 0.09429807960987091,
+      "learning_rate": 1.75596385391374e-05,
+      "loss": 9.0738,
+      "step": 83330
+    },
+    {
+      "epoch": 0.41619016704536943,
+      "grad_norm": 0.0935714915394783,
+      "learning_rate": 1.7558136624195847e-05,
+      "loss": 9.065,
+      "step": 83340
+    },
+    {
+      "epoch": 0.41624010587030885,
+      "grad_norm": 0.09685614705085754,
+      "learning_rate": 1.75566347092543e-05,
+      "loss": 9.0758,
+      "step": 83350
+    },
+    {
+      "epoch": 0.41629004469524833,
+      "grad_norm": 0.09270419180393219,
+      "learning_rate": 1.7555132794312747e-05,
+      "loss": 9.0771,
+      "step": 83360
+    },
+    {
+      "epoch": 0.41633998352018775,
+      "grad_norm": 0.09448056668043137,
+      "learning_rate": 1.75536308793712e-05,
+      "loss": 9.0852,
+      "step": 83370
+    },
+    {
+      "epoch": 0.41638992234512723,
+      "grad_norm": 0.09013166278600693,
+      "learning_rate": 1.7552128964429648e-05,
+      "loss": 9.0791,
+      "step": 83380
+    },
+    {
+      "epoch": 0.41643986117006665,
+      "grad_norm": 0.09527698159217834,
+      "learning_rate": 1.7550627049488098e-05,
+      "loss": 9.0773,
+      "step": 83390
+    },
+    {
+      "epoch": 0.4164897999950061,
+      "grad_norm": 0.09563423693180084,
+      "learning_rate": 1.7549125134546548e-05,
+      "loss": 9.0723,
+      "step": 83400
+    },
+    {
+      "epoch": 0.41653973881994555,
+      "grad_norm": 0.09608939290046692,
+      "learning_rate": 1.7547623219604995e-05,
+      "loss": 9.0768,
+      "step": 83410
+    },
+    {
+      "epoch": 0.416589677644885,
+      "grad_norm": 0.09871455281972885,
+      "learning_rate": 1.754612130466345e-05,
+      "loss": 9.0737,
+      "step": 83420
+    },
+    {
+      "epoch": 0.41663961646982445,
+      "grad_norm": 0.09225298464298248,
+      "learning_rate": 1.7544619389721895e-05,
+      "loss": 9.0819,
+      "step": 83430
+    },
+    {
+      "epoch": 0.4166895552947639,
+      "grad_norm": 0.09640655666589737,
+      "learning_rate": 1.7543117474780345e-05,
+      "loss": 9.0798,
+      "step": 83440
+    },
+    {
+      "epoch": 0.41673949411970335,
+      "grad_norm": 0.09529406577348709,
+      "learning_rate": 1.7541615559838796e-05,
+      "loss": 9.0874,
+      "step": 83450
+    },
+    {
+      "epoch": 0.4167894329446428,
+      "grad_norm": 0.09108598530292511,
+      "learning_rate": 1.7540113644897242e-05,
+      "loss": 9.071,
+      "step": 83460
+    },
+    {
+      "epoch": 0.41683937176958225,
+      "grad_norm": 0.09113084524869919,
+      "learning_rate": 1.7538611729955696e-05,
+      "loss": 9.0852,
+      "step": 83470
+    },
+    {
+      "epoch": 0.4168893105945217,
+      "grad_norm": 0.09776317328214645,
+      "learning_rate": 1.7537109815014143e-05,
+      "loss": 9.0882,
+      "step": 83480
+    },
+    {
+      "epoch": 0.41693924941946114,
+      "grad_norm": 0.09335512667894363,
+      "learning_rate": 1.7535607900072593e-05,
+      "loss": 9.077,
+      "step": 83490
+    },
+    {
+      "epoch": 0.4169891882444006,
+      "grad_norm": 0.09077412635087967,
+      "learning_rate": 1.7534105985131043e-05,
+      "loss": 9.0792,
+      "step": 83500
+    },
+    {
+      "epoch": 0.41703912706934004,
+      "grad_norm": 0.1023026630282402,
+      "learning_rate": 1.753260407018949e-05,
+      "loss": 9.0743,
+      "step": 83510
+    },
+    {
+      "epoch": 0.4170890658942795,
+      "grad_norm": 0.09414954483509064,
+      "learning_rate": 1.7531102155247944e-05,
+      "loss": 9.0811,
+      "step": 83520
+    },
+    {
+      "epoch": 0.41713900471921894,
+      "grad_norm": 0.09618096053600311,
+      "learning_rate": 1.752960024030639e-05,
+      "loss": 9.0919,
+      "step": 83530
+    },
+    {
+      "epoch": 0.4171889435441584,
+      "grad_norm": 0.0886245146393776,
+      "learning_rate": 1.752809832536484e-05,
+      "loss": 9.0966,
+      "step": 83540
+    },
+    {
+      "epoch": 0.41723888236909784,
+      "grad_norm": 0.09386541694402695,
+      "learning_rate": 1.752659641042329e-05,
+      "loss": 9.0889,
+      "step": 83550
+    },
+    {
+      "epoch": 0.4172888211940373,
+      "grad_norm": 0.08851998299360275,
+      "learning_rate": 1.7525094495481737e-05,
+      "loss": 9.0748,
+      "step": 83560
+    },
+    {
+      "epoch": 0.41733876001897674,
+      "grad_norm": 0.1024288460612297,
+      "learning_rate": 1.752359258054019e-05,
+      "loss": 9.0864,
+      "step": 83570
+    },
+    {
+      "epoch": 0.4173886988439162,
+      "grad_norm": 0.09188997000455856,
+      "learning_rate": 1.7522090665598638e-05,
+      "loss": 9.0743,
+      "step": 83580
+    },
+    {
+      "epoch": 0.41743863766885564,
+      "grad_norm": 0.09162962436676025,
+      "learning_rate": 1.7520588750657088e-05,
+      "loss": 9.0847,
+      "step": 83590
+    },
+    {
+      "epoch": 0.4174885764937951,
+      "grad_norm": 0.0950562134385109,
+      "learning_rate": 1.7519086835715538e-05,
+      "loss": 9.0715,
+      "step": 83600
+    },
+    {
+      "epoch": 0.41753851531873454,
+      "grad_norm": 0.09186748415231705,
+      "learning_rate": 1.7517584920773985e-05,
+      "loss": 9.072,
+      "step": 83610
+    },
+    {
+      "epoch": 0.417588454143674,
+      "grad_norm": 0.0918838381767273,
+      "learning_rate": 1.751608300583244e-05,
+      "loss": 9.0721,
+      "step": 83620
+    },
+    {
+      "epoch": 0.41763839296861344,
+      "grad_norm": 0.09041111171245575,
+      "learning_rate": 1.7514581090890885e-05,
+      "loss": 9.0816,
+      "step": 83630
+    },
+    {
+      "epoch": 0.4176883317935529,
+      "grad_norm": 0.08900655061006546,
+      "learning_rate": 1.7513079175949336e-05,
+      "loss": 9.0725,
+      "step": 83640
+    },
+    {
+      "epoch": 0.41773827061849234,
+      "grad_norm": 0.09458507597446442,
+      "learning_rate": 1.7511577261007786e-05,
+      "loss": 9.0837,
+      "step": 83650
+    },
+    {
+      "epoch": 0.4177882094434318,
+      "grad_norm": 0.08990652859210968,
+      "learning_rate": 1.7510075346066232e-05,
+      "loss": 9.0666,
+      "step": 83660
+    },
+    {
+      "epoch": 0.41783814826837123,
+      "grad_norm": 0.09383875876665115,
+      "learning_rate": 1.7508573431124686e-05,
+      "loss": 9.0684,
+      "step": 83670
+    },
+    {
+      "epoch": 0.4178880870933107,
+      "grad_norm": 0.0938841700553894,
+      "learning_rate": 1.7507071516183133e-05,
+      "loss": 9.076,
+      "step": 83680
+    },
+    {
+      "epoch": 0.41793802591825013,
+      "grad_norm": 0.0932869017124176,
+      "learning_rate": 1.7505569601241583e-05,
+      "loss": 9.0731,
+      "step": 83690
+    },
+    {
+      "epoch": 0.4179879647431896,
+      "grad_norm": 0.09501129388809204,
+      "learning_rate": 1.7504067686300033e-05,
+      "loss": 9.0626,
+      "step": 83700
+    },
+    {
+      "epoch": 0.41803790356812903,
+      "grad_norm": 0.09345944225788116,
+      "learning_rate": 1.7502565771358483e-05,
+      "loss": 9.0749,
+      "step": 83710
+    },
+    {
+      "epoch": 0.4180878423930685,
+      "grad_norm": 0.09460506588220596,
+      "learning_rate": 1.7501063856416934e-05,
+      "loss": 9.0764,
+      "step": 83720
+    },
+    {
+      "epoch": 0.41813778121800793,
+      "grad_norm": 0.09743547439575195,
+      "learning_rate": 1.749956194147538e-05,
+      "loss": 9.0757,
+      "step": 83730
+    },
+    {
+      "epoch": 0.4181877200429474,
+      "grad_norm": 0.09816765040159225,
+      "learning_rate": 1.749806002653383e-05,
+      "loss": 9.0767,
+      "step": 83740
+    },
+    {
+      "epoch": 0.41823765886788683,
+      "grad_norm": 0.09210207313299179,
+      "learning_rate": 1.749655811159228e-05,
+      "loss": 9.0721,
+      "step": 83750
+    },
+    {
+      "epoch": 0.4182875976928263,
+      "grad_norm": 0.09752412140369415,
+      "learning_rate": 1.749505619665073e-05,
+      "loss": 9.0679,
+      "step": 83760
+    },
+    {
+      "epoch": 0.41833753651776573,
+      "grad_norm": 0.09351370483636856,
+      "learning_rate": 1.749355428170918e-05,
+      "loss": 9.0862,
+      "step": 83770
+    },
+    {
+      "epoch": 0.4183874753427052,
+      "grad_norm": 0.09450095891952515,
+      "learning_rate": 1.7492052366767628e-05,
+      "loss": 9.0659,
+      "step": 83780
+    },
+    {
+      "epoch": 0.41843741416764463,
+      "grad_norm": 0.09616388380527496,
+      "learning_rate": 1.7490550451826078e-05,
+      "loss": 9.0898,
+      "step": 83790
+    },
+    {
+      "epoch": 0.4184873529925841,
+      "grad_norm": 0.09121173620223999,
+      "learning_rate": 1.7489048536884528e-05,
+      "loss": 9.1012,
+      "step": 83800
+    },
+    {
+      "epoch": 0.4185372918175235,
+      "grad_norm": 0.09462346136569977,
+      "learning_rate": 1.748754662194298e-05,
+      "loss": 9.0969,
+      "step": 83810
+    },
+    {
+      "epoch": 0.418587230642463,
+      "grad_norm": 0.09324286133050919,
+      "learning_rate": 1.748604470700143e-05,
+      "loss": 9.0746,
+      "step": 83820
+    },
+    {
+      "epoch": 0.4186371694674024,
+      "grad_norm": 0.09117257595062256,
+      "learning_rate": 1.7484542792059875e-05,
+      "loss": 9.0724,
+      "step": 83830
+    },
+    {
+      "epoch": 0.4186871082923419,
+      "grad_norm": 0.09084919095039368,
+      "learning_rate": 1.7483040877118326e-05,
+      "loss": 9.0775,
+      "step": 83840
+    },
+    {
+      "epoch": 0.4187370471172813,
+      "grad_norm": 0.09522408992052078,
+      "learning_rate": 1.7481538962176776e-05,
+      "loss": 9.0747,
+      "step": 83850
+    },
+    {
+      "epoch": 0.4187869859422208,
+      "grad_norm": 0.09409532696008682,
+      "learning_rate": 1.7480037047235226e-05,
+      "loss": 9.0663,
+      "step": 83860
+    },
+    {
+      "epoch": 0.4188369247671602,
+      "grad_norm": 0.08830953389406204,
+      "learning_rate": 1.7478535132293676e-05,
+      "loss": 9.0784,
+      "step": 83870
+    },
+    {
+      "epoch": 0.4188868635920997,
+      "grad_norm": 0.0929829478263855,
+      "learning_rate": 1.7477033217352123e-05,
+      "loss": 9.0728,
+      "step": 83880
+    },
+    {
+      "epoch": 0.4189368024170391,
+      "grad_norm": 0.09145011007785797,
+      "learning_rate": 1.7475531302410573e-05,
+      "loss": 9.0703,
+      "step": 83890
+    },
+    {
+      "epoch": 0.4189867412419786,
+      "grad_norm": 0.09224165230989456,
+      "learning_rate": 1.7474029387469023e-05,
+      "loss": 9.077,
+      "step": 83900
+    },
+    {
+      "epoch": 0.419036680066918,
+      "grad_norm": 0.09355958551168442,
+      "learning_rate": 1.7472527472527473e-05,
+      "loss": 9.0668,
+      "step": 83910
+    },
+    {
+      "epoch": 0.4190866188918575,
+      "grad_norm": 0.09556030482053757,
+      "learning_rate": 1.7471025557585924e-05,
+      "loss": 9.0659,
+      "step": 83920
+    },
+    {
+      "epoch": 0.4191365577167969,
+      "grad_norm": 0.09351398795843124,
+      "learning_rate": 1.746952364264437e-05,
+      "loss": 9.0652,
+      "step": 83930
+    },
+    {
+      "epoch": 0.4191864965417364,
+      "grad_norm": 0.09275408834218979,
+      "learning_rate": 1.746802172770282e-05,
+      "loss": 9.0618,
+      "step": 83940
+    },
+    {
+      "epoch": 0.4192364353666758,
+      "grad_norm": 0.0906294509768486,
+      "learning_rate": 1.746651981276127e-05,
+      "loss": 9.0744,
+      "step": 83950
+    },
+    {
+      "epoch": 0.4192863741916153,
+      "grad_norm": 0.09481019526720047,
+      "learning_rate": 1.746501789781972e-05,
+      "loss": 9.0686,
+      "step": 83960
+    },
+    {
+      "epoch": 0.4193363130165547,
+      "grad_norm": 0.0955205038189888,
+      "learning_rate": 1.746351598287817e-05,
+      "loss": 9.0717,
+      "step": 83970
+    },
+    {
+      "epoch": 0.4193862518414942,
+      "grad_norm": 0.09485220909118652,
+      "learning_rate": 1.7462014067936618e-05,
+      "loss": 9.0709,
+      "step": 83980
+    },
+    {
+      "epoch": 0.4194361906664336,
+      "grad_norm": 0.09109359234571457,
+      "learning_rate": 1.7460512152995068e-05,
+      "loss": 9.0645,
+      "step": 83990
+    },
+    {
+      "epoch": 0.4194861294913731,
+      "grad_norm": 0.09005378186702728,
+      "learning_rate": 1.7459010238053518e-05,
+      "loss": 9.0704,
+      "step": 84000
+    },
+    {
+      "epoch": 0.4195360683163125,
+      "grad_norm": 0.09958352148532867,
+      "learning_rate": 1.745750832311197e-05,
+      "loss": 9.0769,
+      "step": 84010
+    },
+    {
+      "epoch": 0.419586007141252,
+      "grad_norm": 0.09399083256721497,
+      "learning_rate": 1.745600640817042e-05,
+      "loss": 9.0654,
+      "step": 84020
+    },
+    {
+      "epoch": 0.4196359459661914,
+      "grad_norm": 0.09548033028841019,
+      "learning_rate": 1.745450449322887e-05,
+      "loss": 9.0775,
+      "step": 84030
+    },
+    {
+      "epoch": 0.4196858847911309,
+      "grad_norm": 0.09409498423337936,
+      "learning_rate": 1.7453002578287316e-05,
+      "loss": 9.0756,
+      "step": 84040
+    },
+    {
+      "epoch": 0.4197358236160703,
+      "grad_norm": 0.09729974716901779,
+      "learning_rate": 1.7451500663345766e-05,
+      "loss": 9.0749,
+      "step": 84050
+    },
+    {
+      "epoch": 0.41978576244100974,
+      "grad_norm": 0.09663716703653336,
+      "learning_rate": 1.7449998748404216e-05,
+      "loss": 9.069,
+      "step": 84060
+    },
+    {
+      "epoch": 0.4198357012659492,
+      "grad_norm": 0.09065845608711243,
+      "learning_rate": 1.7448496833462666e-05,
+      "loss": 9.066,
+      "step": 84070
+    },
+    {
+      "epoch": 0.41988564009088863,
+      "grad_norm": 0.09165765345096588,
+      "learning_rate": 1.7446994918521116e-05,
+      "loss": 9.0866,
+      "step": 84080
+    },
+    {
+      "epoch": 0.4199355789158281,
+      "grad_norm": 0.09514272212982178,
+      "learning_rate": 1.7445493003579563e-05,
+      "loss": 9.07,
+      "step": 84090
+    },
+    {
+      "epoch": 0.41998551774076753,
+      "grad_norm": 0.090760737657547,
+      "learning_rate": 1.7443991088638013e-05,
+      "loss": 9.0838,
+      "step": 84100
+    },
+    {
+      "epoch": 0.420035456565707,
+      "grad_norm": 0.09050191193819046,
+      "learning_rate": 1.7442489173696463e-05,
+      "loss": 9.0707,
+      "step": 84110
+    },
+    {
+      "epoch": 0.42008539539064643,
+      "grad_norm": 0.09378772228956223,
+      "learning_rate": 1.7440987258754914e-05,
+      "loss": 9.0724,
+      "step": 84120
+    },
+    {
+      "epoch": 0.4201353342155859,
+      "grad_norm": 0.09341657161712646,
+      "learning_rate": 1.7439485343813364e-05,
+      "loss": 9.067,
+      "step": 84130
+    },
+    {
+      "epoch": 0.42018527304052533,
+      "grad_norm": 0.09716533869504929,
+      "learning_rate": 1.743798342887181e-05,
+      "loss": 9.0632,
+      "step": 84140
+    },
+    {
+      "epoch": 0.4202352118654648,
+      "grad_norm": 0.09457621723413467,
+      "learning_rate": 1.743648151393026e-05,
+      "loss": 9.0649,
+      "step": 84150
+    },
+    {
+      "epoch": 0.42028515069040423,
+      "grad_norm": 0.0924045741558075,
+      "learning_rate": 1.743497959898871e-05,
+      "loss": 9.0646,
+      "step": 84160
+    },
+    {
+      "epoch": 0.4203350895153437,
+      "grad_norm": 0.09211113303899765,
+      "learning_rate": 1.743347768404716e-05,
+      "loss": 9.0715,
+      "step": 84170
+    },
+    {
+      "epoch": 0.42038502834028313,
+      "grad_norm": 0.09424125403165817,
+      "learning_rate": 1.743197576910561e-05,
+      "loss": 9.069,
+      "step": 84180
+    },
+    {
+      "epoch": 0.4204349671652226,
+      "grad_norm": 0.08973129838705063,
+      "learning_rate": 1.7430473854164058e-05,
+      "loss": 9.0637,
+      "step": 84190
+    },
+    {
+      "epoch": 0.42048490599016203,
+      "grad_norm": 0.09753432869911194,
+      "learning_rate": 1.7428971939222508e-05,
+      "loss": 9.0612,
+      "step": 84200
+    },
+    {
+      "epoch": 0.4205348448151015,
+      "grad_norm": 0.09066448360681534,
+      "learning_rate": 1.742747002428096e-05,
+      "loss": 9.0633,
+      "step": 84210
+    },
+    {
+      "epoch": 0.4205847836400409,
+      "grad_norm": 0.09251467883586884,
+      "learning_rate": 1.742596810933941e-05,
+      "loss": 9.0689,
+      "step": 84220
+    },
+    {
+      "epoch": 0.4206347224649804,
+      "grad_norm": 0.09440715610980988,
+      "learning_rate": 1.742446619439786e-05,
+      "loss": 9.0761,
+      "step": 84230
+    },
+    {
+      "epoch": 0.4206846612899198,
+      "grad_norm": 0.09155986458063126,
+      "learning_rate": 1.7422964279456306e-05,
+      "loss": 9.0703,
+      "step": 84240
+    },
+    {
+      "epoch": 0.4207346001148593,
+      "grad_norm": 0.0903191938996315,
+      "learning_rate": 1.7421462364514756e-05,
+      "loss": 9.0739,
+      "step": 84250
+    },
+    {
+      "epoch": 0.4207845389397987,
+      "grad_norm": 0.09451457858085632,
+      "learning_rate": 1.7419960449573206e-05,
+      "loss": 9.0768,
+      "step": 84260
+    },
+    {
+      "epoch": 0.4208344777647382,
+      "grad_norm": 0.09495178610086441,
+      "learning_rate": 1.7418458534631656e-05,
+      "loss": 9.0598,
+      "step": 84270
+    },
+    {
+      "epoch": 0.4208844165896776,
+      "grad_norm": 0.09594985842704773,
+      "learning_rate": 1.7416956619690106e-05,
+      "loss": 9.0666,
+      "step": 84280
+    },
+    {
+      "epoch": 0.4209343554146171,
+      "grad_norm": 0.09474395960569382,
+      "learning_rate": 1.7415454704748553e-05,
+      "loss": 9.0648,
+      "step": 84290
+    },
+    {
+      "epoch": 0.4209842942395565,
+      "grad_norm": 0.10143403708934784,
+      "learning_rate": 1.7413952789807003e-05,
+      "loss": 9.0528,
+      "step": 84300
+    },
+    {
+      "epoch": 0.421034233064496,
+      "grad_norm": 0.09076518565416336,
+      "learning_rate": 1.7412450874865453e-05,
+      "loss": 9.0731,
+      "step": 84310
+    },
+    {
+      "epoch": 0.4210841718894354,
+      "grad_norm": 0.0921441987156868,
+      "learning_rate": 1.7410948959923904e-05,
+      "loss": 9.067,
+      "step": 84320
+    },
+    {
+      "epoch": 0.4211341107143749,
+      "grad_norm": 0.09396324306726456,
+      "learning_rate": 1.7409447044982354e-05,
+      "loss": 9.0719,
+      "step": 84330
+    },
+    {
+      "epoch": 0.4211840495393143,
+      "grad_norm": 0.0950896143913269,
+      "learning_rate": 1.74079451300408e-05,
+      "loss": 9.0815,
+      "step": 84340
+    },
+    {
+      "epoch": 0.4212339883642538,
+      "grad_norm": 0.08926611393690109,
+      "learning_rate": 1.7406443215099254e-05,
+      "loss": 9.0629,
+      "step": 84350
+    },
+    {
+      "epoch": 0.4212839271891932,
+      "grad_norm": 0.09594748169183731,
+      "learning_rate": 1.74049413001577e-05,
+      "loss": 9.0707,
+      "step": 84360
+    },
+    {
+      "epoch": 0.4213338660141327,
+      "grad_norm": 0.09475325793027878,
+      "learning_rate": 1.740343938521615e-05,
+      "loss": 9.074,
+      "step": 84370
+    },
+    {
+      "epoch": 0.4213838048390721,
+      "grad_norm": 0.09772231429815292,
+      "learning_rate": 1.74019374702746e-05,
+      "loss": 9.0669,
+      "step": 84380
+    },
+    {
+      "epoch": 0.4214337436640116,
+      "grad_norm": 0.09714603424072266,
+      "learning_rate": 1.7400435555333048e-05,
+      "loss": 9.0703,
+      "step": 84390
+    },
+    {
+      "epoch": 0.421483682488951,
+      "grad_norm": 0.09235162287950516,
+      "learning_rate": 1.73989336403915e-05,
+      "loss": 9.059,
+      "step": 84400
+    },
+    {
+      "epoch": 0.4215336213138905,
+      "grad_norm": 0.0921136885881424,
+      "learning_rate": 1.739743172544995e-05,
+      "loss": 9.0542,
+      "step": 84410
+    },
+    {
+      "epoch": 0.4215835601388299,
+      "grad_norm": 0.09008321166038513,
+      "learning_rate": 1.73959298105084e-05,
+      "loss": 9.066,
+      "step": 84420
+    },
+    {
+      "epoch": 0.4216334989637694,
+      "grad_norm": 0.0914497822523117,
+      "learning_rate": 1.739442789556685e-05,
+      "loss": 9.0638,
+      "step": 84430
+    },
+    {
+      "epoch": 0.4216834377887088,
+      "grad_norm": 0.09600444883108139,
+      "learning_rate": 1.7392925980625296e-05,
+      "loss": 9.0733,
+      "step": 84440
+    },
+    {
+      "epoch": 0.4217333766136483,
+      "grad_norm": 0.09309738129377365,
+      "learning_rate": 1.739142406568375e-05,
+      "loss": 9.0743,
+      "step": 84450
+    },
+    {
+      "epoch": 0.4217833154385877,
+      "grad_norm": 0.09112364053726196,
+      "learning_rate": 1.7389922150742196e-05,
+      "loss": 9.0722,
+      "step": 84460
+    },
+    {
+      "epoch": 0.4218332542635272,
+      "grad_norm": 0.09220999479293823,
+      "learning_rate": 1.7388420235800646e-05,
+      "loss": 9.0659,
+      "step": 84470
+    },
+    {
+      "epoch": 0.4218831930884666,
+      "grad_norm": 0.09681863337755203,
+      "learning_rate": 1.7386918320859096e-05,
+      "loss": 9.0701,
+      "step": 84480
+    },
+    {
+      "epoch": 0.4219331319134061,
+      "grad_norm": 0.09077723324298859,
+      "learning_rate": 1.7385416405917543e-05,
+      "loss": 9.0616,
+      "step": 84490
+    },
+    {
+      "epoch": 0.4219830707383455,
+      "grad_norm": 0.09060750901699066,
+      "learning_rate": 1.7383914490975997e-05,
+      "loss": 9.06,
+      "step": 84500
+    },
+    {
+      "epoch": 0.422033009563285,
+      "grad_norm": 0.09467575699090958,
+      "learning_rate": 1.7382412576034443e-05,
+      "loss": 9.0742,
+      "step": 84510
+    },
+    {
+      "epoch": 0.4220829483882244,
+      "grad_norm": 0.09085453301668167,
+      "learning_rate": 1.7380910661092894e-05,
+      "loss": 9.0705,
+      "step": 84520
+    },
+    {
+      "epoch": 0.4221328872131639,
+      "grad_norm": 0.09209267795085907,
+      "learning_rate": 1.7379408746151344e-05,
+      "loss": 9.0544,
+      "step": 84530
+    },
+    {
+      "epoch": 0.4221828260381033,
+      "grad_norm": 0.09571056067943573,
+      "learning_rate": 1.737790683120979e-05,
+      "loss": 9.0732,
+      "step": 84540
+    },
+    {
+      "epoch": 0.4222327648630428,
+      "grad_norm": 0.09486297518014908,
+      "learning_rate": 1.7376404916268244e-05,
+      "loss": 9.0753,
+      "step": 84550
+    },
+    {
+      "epoch": 0.4222827036879822,
+      "grad_norm": 0.09144411236047745,
+      "learning_rate": 1.737490300132669e-05,
+      "loss": 9.0669,
+      "step": 84560
+    },
+    {
+      "epoch": 0.4223326425129217,
+      "grad_norm": 0.09834633022546768,
+      "learning_rate": 1.737340108638514e-05,
+      "loss": 9.072,
+      "step": 84570
+    },
+    {
+      "epoch": 0.4223825813378611,
+      "grad_norm": 0.09790683537721634,
+      "learning_rate": 1.737189917144359e-05,
+      "loss": 9.0566,
+      "step": 84580
+    },
+    {
+      "epoch": 0.4224325201628006,
+      "grad_norm": 0.09472320228815079,
+      "learning_rate": 1.7370397256502038e-05,
+      "loss": 9.0709,
+      "step": 84590
+    },
+    {
+      "epoch": 0.42248245898774,
+      "grad_norm": 0.0928400382399559,
+      "learning_rate": 1.736889534156049e-05,
+      "loss": 9.063,
+      "step": 84600
+    },
+    {
+      "epoch": 0.4225323978126795,
+      "grad_norm": 0.09452667087316513,
+      "learning_rate": 1.736739342661894e-05,
+      "loss": 9.0671,
+      "step": 84610
+    },
+    {
+      "epoch": 0.4225823366376189,
+      "grad_norm": 0.09719450771808624,
+      "learning_rate": 1.736589151167739e-05,
+      "loss": 9.0658,
+      "step": 84620
+    },
+    {
+      "epoch": 0.4226322754625584,
+      "grad_norm": 0.08982671052217484,
+      "learning_rate": 1.736438959673584e-05,
+      "loss": 9.0629,
+      "step": 84630
+    },
+    {
+      "epoch": 0.4226822142874978,
+      "grad_norm": 0.08864292502403259,
+      "learning_rate": 1.7362887681794286e-05,
+      "loss": 9.0618,
+      "step": 84640
+    },
+    {
+      "epoch": 0.4227321531124373,
+      "grad_norm": 0.09744229912757874,
+      "learning_rate": 1.736138576685274e-05,
+      "loss": 9.0557,
+      "step": 84650
+    },
+    {
+      "epoch": 0.4227820919373767,
+      "grad_norm": 0.09893699735403061,
+      "learning_rate": 1.7359883851911186e-05,
+      "loss": 9.067,
+      "step": 84660
+    },
+    {
+      "epoch": 0.4228320307623162,
+      "grad_norm": 0.09269163012504578,
+      "learning_rate": 1.735838193696964e-05,
+      "loss": 9.0676,
+      "step": 84670
+    },
+    {
+      "epoch": 0.4228819695872556,
+      "grad_norm": 0.09532048553228378,
+      "learning_rate": 1.7356880022028086e-05,
+      "loss": 9.0715,
+      "step": 84680
+    },
+    {
+      "epoch": 0.4229319084121951,
+      "grad_norm": 0.09102669358253479,
+      "learning_rate": 1.7355378107086533e-05,
+      "loss": 9.0674,
+      "step": 84690
+    },
+    {
+      "epoch": 0.4229818472371345,
+      "grad_norm": 0.09197783470153809,
+      "learning_rate": 1.7353876192144987e-05,
+      "loss": 9.0677,
+      "step": 84700
+    },
+    {
+      "epoch": 0.423031786062074,
+      "grad_norm": 0.09266051650047302,
+      "learning_rate": 1.7352374277203433e-05,
+      "loss": 9.0652,
+      "step": 84710
+    },
+    {
+      "epoch": 0.4230817248870134,
+      "grad_norm": 0.09375884383916855,
+      "learning_rate": 1.7350872362261887e-05,
+      "loss": 9.0622,
+      "step": 84720
+    },
+    {
+      "epoch": 0.4231316637119529,
+      "grad_norm": 0.09187036752700806,
+      "learning_rate": 1.7349370447320334e-05,
+      "loss": 9.0654,
+      "step": 84730
+    },
+    {
+      "epoch": 0.4231816025368923,
+      "grad_norm": 0.09767387062311172,
+      "learning_rate": 1.734786853237878e-05,
+      "loss": 9.074,
+      "step": 84740
+    },
+    {
+      "epoch": 0.4232315413618318,
+      "grad_norm": 0.09047545492649078,
+      "learning_rate": 1.7346366617437234e-05,
+      "loss": 9.0664,
+      "step": 84750
+    },
+    {
+      "epoch": 0.4232814801867712,
+      "grad_norm": 0.09297401458024979,
+      "learning_rate": 1.734486470249568e-05,
+      "loss": 9.073,
+      "step": 84760
+    },
+    {
+      "epoch": 0.4233314190117107,
+      "grad_norm": 0.0917564183473587,
+      "learning_rate": 1.7343362787554135e-05,
+      "loss": 9.064,
+      "step": 84770
+    },
+    {
+      "epoch": 0.4233813578366501,
+      "grad_norm": 0.09373626857995987,
+      "learning_rate": 1.734186087261258e-05,
+      "loss": 9.0597,
+      "step": 84780
+    },
+    {
+      "epoch": 0.4234312966615896,
+      "grad_norm": 0.09210002422332764,
+      "learning_rate": 1.7340358957671028e-05,
+      "loss": 9.0762,
+      "step": 84790
+    },
+    {
+      "epoch": 0.423481235486529,
+      "grad_norm": 0.09276442974805832,
+      "learning_rate": 1.733885704272948e-05,
+      "loss": 9.0565,
+      "step": 84800
+    },
+    {
+      "epoch": 0.4235311743114685,
+      "grad_norm": 0.09392396360635757,
+      "learning_rate": 1.733735512778793e-05,
+      "loss": 9.0691,
+      "step": 84810
+    },
+    {
+      "epoch": 0.4235811131364079,
+      "grad_norm": 0.09003560990095139,
+      "learning_rate": 1.7335853212846382e-05,
+      "loss": 9.0619,
+      "step": 84820
+    },
+    {
+      "epoch": 0.42363105196134737,
+      "grad_norm": 0.09067775309085846,
+      "learning_rate": 1.733435129790483e-05,
+      "loss": 9.0586,
+      "step": 84830
+    },
+    {
+      "epoch": 0.4236809907862868,
+      "grad_norm": 0.0904303789138794,
+      "learning_rate": 1.7332849382963276e-05,
+      "loss": 9.054,
+      "step": 84840
+    },
+    {
+      "epoch": 0.42373092961122627,
+      "grad_norm": 0.09555737674236298,
+      "learning_rate": 1.733134746802173e-05,
+      "loss": 9.0534,
+      "step": 84850
+    },
+    {
+      "epoch": 0.4237808684361657,
+      "grad_norm": 0.100531667470932,
+      "learning_rate": 1.7329845553080176e-05,
+      "loss": 9.0776,
+      "step": 84860
+    },
+    {
+      "epoch": 0.42383080726110517,
+      "grad_norm": 0.09057488292455673,
+      "learning_rate": 1.732834363813863e-05,
+      "loss": 9.0704,
+      "step": 84870
+    },
+    {
+      "epoch": 0.4238807460860446,
+      "grad_norm": 0.09145992249250412,
+      "learning_rate": 1.7326841723197076e-05,
+      "loss": 9.0489,
+      "step": 84880
+    },
+    {
+      "epoch": 0.42393068491098407,
+      "grad_norm": 0.0870828926563263,
+      "learning_rate": 1.7325339808255523e-05,
+      "loss": 9.0725,
+      "step": 84890
+    },
+    {
+      "epoch": 0.4239806237359235,
+      "grad_norm": 0.09605187177658081,
+      "learning_rate": 1.7323837893313977e-05,
+      "loss": 9.0487,
+      "step": 84900
+    },
+    {
+      "epoch": 0.42403056256086297,
+      "grad_norm": 0.09087910503149033,
+      "learning_rate": 1.7322335978372423e-05,
+      "loss": 9.0577,
+      "step": 84910
+    },
+    {
+      "epoch": 0.4240805013858024,
+      "grad_norm": 0.09988180547952652,
+      "learning_rate": 1.7320834063430877e-05,
+      "loss": 9.0576,
+      "step": 84920
+    },
+    {
+      "epoch": 0.42413044021074187,
+      "grad_norm": 0.09044559299945831,
+      "learning_rate": 1.7319332148489324e-05,
+      "loss": 9.0767,
+      "step": 84930
+    },
+    {
+      "epoch": 0.4241803790356813,
+      "grad_norm": 0.09994645416736603,
+      "learning_rate": 1.731783023354777e-05,
+      "loss": 9.0492,
+      "step": 84940
+    },
+    {
+      "epoch": 0.42423031786062076,
+      "grad_norm": 0.09559715539216995,
+      "learning_rate": 1.7316328318606224e-05,
+      "loss": 9.0664,
+      "step": 84950
+    },
+    {
+      "epoch": 0.4242802566855602,
+      "grad_norm": 0.08689738065004349,
+      "learning_rate": 1.731482640366467e-05,
+      "loss": 9.0722,
+      "step": 84960
+    },
+    {
+      "epoch": 0.42433019551049966,
+      "grad_norm": 0.10374455153942108,
+      "learning_rate": 1.7313324488723125e-05,
+      "loss": 9.0643,
+      "step": 84970
+    },
+    {
+      "epoch": 0.4243801343354391,
+      "grad_norm": 0.0957735925912857,
+      "learning_rate": 1.731182257378157e-05,
+      "loss": 9.0582,
+      "step": 84980
+    },
+    {
+      "epoch": 0.42443007316037856,
+      "grad_norm": 0.09154541045427322,
+      "learning_rate": 1.731032065884002e-05,
+      "loss": 9.06,
+      "step": 84990
+    },
+    {
+      "epoch": 0.424480011985318,
+      "grad_norm": 0.09392927587032318,
+      "learning_rate": 1.730881874389847e-05,
+      "loss": 9.0737,
+      "step": 85000
+    },
+    {
+      "epoch": 0.42452995081025746,
+      "grad_norm": 0.09651625156402588,
+      "learning_rate": 1.730731682895692e-05,
+      "loss": 9.0567,
+      "step": 85010
+    },
+    {
+      "epoch": 0.4245798896351969,
+      "grad_norm": 0.09320859611034393,
+      "learning_rate": 1.7305814914015372e-05,
+      "loss": 9.0656,
+      "step": 85020
+    },
+    {
+      "epoch": 0.42462982846013636,
+      "grad_norm": 0.09693901240825653,
+      "learning_rate": 1.730431299907382e-05,
+      "loss": 9.0709,
+      "step": 85030
+    },
+    {
+      "epoch": 0.4246797672850758,
+      "grad_norm": 0.09250442683696747,
+      "learning_rate": 1.730281108413227e-05,
+      "loss": 9.059,
+      "step": 85040
+    },
+    {
+      "epoch": 0.4247297061100152,
+      "grad_norm": 0.09281265735626221,
+      "learning_rate": 1.730130916919072e-05,
+      "loss": 9.0618,
+      "step": 85050
+    },
+    {
+      "epoch": 0.4247796449349547,
+      "grad_norm": 0.09139450639486313,
+      "learning_rate": 1.7299807254249166e-05,
+      "loss": 9.0574,
+      "step": 85060
+    },
+    {
+      "epoch": 0.4248295837598941,
+      "grad_norm": 0.09575843065977097,
+      "learning_rate": 1.729830533930762e-05,
+      "loss": 9.0575,
+      "step": 85070
+    },
+    {
+      "epoch": 0.4248795225848336,
+      "grad_norm": 0.09058381617069244,
+      "learning_rate": 1.7296803424366066e-05,
+      "loss": 9.0458,
+      "step": 85080
+    },
+    {
+      "epoch": 0.424929461409773,
+      "grad_norm": 0.09589716792106628,
+      "learning_rate": 1.7295301509424516e-05,
+      "loss": 9.0613,
+      "step": 85090
+    },
+    {
+      "epoch": 0.4249794002347125,
+      "grad_norm": 0.09588911384344101,
+      "learning_rate": 1.7293799594482967e-05,
+      "loss": 9.0706,
+      "step": 85100
+    },
+    {
+      "epoch": 0.4250293390596519,
+      "grad_norm": 0.08973708748817444,
+      "learning_rate": 1.7292297679541413e-05,
+      "loss": 9.0601,
+      "step": 85110
+    },
+    {
+      "epoch": 0.4250792778845914,
+      "grad_norm": 0.09021148830652237,
+      "learning_rate": 1.7290795764599867e-05,
+      "loss": 9.0696,
+      "step": 85120
+    },
+    {
+      "epoch": 0.4251292167095308,
+      "grad_norm": 0.08464304357767105,
+      "learning_rate": 1.7289293849658314e-05,
+      "loss": 9.0598,
+      "step": 85130
+    },
+    {
+      "epoch": 0.4251791555344703,
+      "grad_norm": 0.09651916474103928,
+      "learning_rate": 1.7287791934716764e-05,
+      "loss": 9.0664,
+      "step": 85140
+    },
+    {
+      "epoch": 0.4252290943594097,
+      "grad_norm": 0.09342118352651596,
+      "learning_rate": 1.7286290019775214e-05,
+      "loss": 9.0592,
+      "step": 85150
+    },
+    {
+      "epoch": 0.4252790331843492,
+      "grad_norm": 0.09311968088150024,
+      "learning_rate": 1.728478810483366e-05,
+      "loss": 9.0665,
+      "step": 85160
+    },
+    {
+      "epoch": 0.4253289720092886,
+      "grad_norm": 0.08970993012189865,
+      "learning_rate": 1.7283286189892115e-05,
+      "loss": 9.0662,
+      "step": 85170
+    },
+    {
+      "epoch": 0.4253789108342281,
+      "grad_norm": 0.09633956849575043,
+      "learning_rate": 1.728178427495056e-05,
+      "loss": 9.0603,
+      "step": 85180
+    },
+    {
+      "epoch": 0.4254288496591675,
+      "grad_norm": 0.09612761437892914,
+      "learning_rate": 1.728028236000901e-05,
+      "loss": 9.0691,
+      "step": 85190
+    },
+    {
+      "epoch": 0.425478788484107,
+      "grad_norm": 0.0876702144742012,
+      "learning_rate": 1.727878044506746e-05,
+      "loss": 9.0624,
+      "step": 85200
+    },
+    {
+      "epoch": 0.4255287273090464,
+      "grad_norm": 0.09160758554935455,
+      "learning_rate": 1.727727853012591e-05,
+      "loss": 9.0717,
+      "step": 85210
+    },
+    {
+      "epoch": 0.4255786661339859,
+      "grad_norm": 0.09074576944112778,
+      "learning_rate": 1.7275776615184362e-05,
+      "loss": 9.0706,
+      "step": 85220
+    },
+    {
+      "epoch": 0.4256286049589253,
+      "grad_norm": 0.08834607899188995,
+      "learning_rate": 1.727427470024281e-05,
+      "loss": 9.06,
+      "step": 85230
+    },
+    {
+      "epoch": 0.42567854378386477,
+      "grad_norm": 0.09007090330123901,
+      "learning_rate": 1.727277278530126e-05,
+      "loss": 9.0507,
+      "step": 85240
+    },
+    {
+      "epoch": 0.4257284826088042,
+      "grad_norm": 0.09260173887014389,
+      "learning_rate": 1.727127087035971e-05,
+      "loss": 9.0542,
+      "step": 85250
+    },
+    {
+      "epoch": 0.42577842143374367,
+      "grad_norm": 0.0915934219956398,
+      "learning_rate": 1.7269768955418156e-05,
+      "loss": 9.0627,
+      "step": 85260
+    },
+    {
+      "epoch": 0.4258283602586831,
+      "grad_norm": 0.09786804020404816,
+      "learning_rate": 1.726826704047661e-05,
+      "loss": 9.0529,
+      "step": 85270
+    },
+    {
+      "epoch": 0.42587829908362257,
+      "grad_norm": 0.09305689483880997,
+      "learning_rate": 1.7266765125535056e-05,
+      "loss": 9.0447,
+      "step": 85280
+    },
+    {
+      "epoch": 0.425928237908562,
+      "grad_norm": 0.09683950990438461,
+      "learning_rate": 1.7265263210593507e-05,
+      "loss": 9.0657,
+      "step": 85290
+    },
+    {
+      "epoch": 0.42597817673350147,
+      "grad_norm": 0.08442121744155884,
+      "learning_rate": 1.7263761295651957e-05,
+      "loss": 9.065,
+      "step": 85300
+    },
+    {
+      "epoch": 0.4260281155584409,
+      "grad_norm": 0.09468364715576172,
+      "learning_rate": 1.7262259380710403e-05,
+      "loss": 9.0679,
+      "step": 85310
+    },
+    {
+      "epoch": 0.42607805438338037,
+      "grad_norm": 0.09155453741550446,
+      "learning_rate": 1.7260757465768857e-05,
+      "loss": 9.0554,
+      "step": 85320
+    },
+    {
+      "epoch": 0.4261279932083198,
+      "grad_norm": 0.09124631434679031,
+      "learning_rate": 1.7259255550827304e-05,
+      "loss": 9.0581,
+      "step": 85330
+    },
+    {
+      "epoch": 0.42617793203325927,
+      "grad_norm": 0.09624254703521729,
+      "learning_rate": 1.7257753635885757e-05,
+      "loss": 9.0521,
+      "step": 85340
+    },
+    {
+      "epoch": 0.4262278708581987,
+      "grad_norm": 0.09301303327083588,
+      "learning_rate": 1.7256251720944204e-05,
+      "loss": 9.0474,
+      "step": 85350
+    },
+    {
+      "epoch": 0.42627780968313816,
+      "grad_norm": 0.09513042867183685,
+      "learning_rate": 1.7254749806002654e-05,
+      "loss": 9.0544,
+      "step": 85360
+    },
+    {
+      "epoch": 0.4263277485080776,
+      "grad_norm": 0.08930395543575287,
+      "learning_rate": 1.7253247891061105e-05,
+      "loss": 9.0651,
+      "step": 85370
+    },
+    {
+      "epoch": 0.42637768733301706,
+      "grad_norm": 0.09650768339633942,
+      "learning_rate": 1.725174597611955e-05,
+      "loss": 9.057,
+      "step": 85380
+    },
+    {
+      "epoch": 0.4264276261579565,
+      "grad_norm": 0.09203314781188965,
+      "learning_rate": 1.7250244061178005e-05,
+      "loss": 9.0673,
+      "step": 85390
+    },
+    {
+      "epoch": 0.42647756498289596,
+      "grad_norm": 0.09252451360225677,
+      "learning_rate": 1.7248742146236452e-05,
+      "loss": 9.0654,
+      "step": 85400
+    },
+    {
+      "epoch": 0.4265275038078354,
+      "grad_norm": 0.09380394965410233,
+      "learning_rate": 1.7247240231294902e-05,
+      "loss": 9.0483,
+      "step": 85410
+    },
+    {
+      "epoch": 0.42657744263277486,
+      "grad_norm": 0.09272889792919159,
+      "learning_rate": 1.7245738316353352e-05,
+      "loss": 9.0737,
+      "step": 85420
+    },
+    {
+      "epoch": 0.4266273814577143,
+      "grad_norm": 0.09201991558074951,
+      "learning_rate": 1.72442364014118e-05,
+      "loss": 9.0526,
+      "step": 85430
+    },
+    {
+      "epoch": 0.42667732028265376,
+      "grad_norm": 0.09079743921756744,
+      "learning_rate": 1.7242734486470252e-05,
+      "loss": 9.062,
+      "step": 85440
+    },
+    {
+      "epoch": 0.4267272591075932,
+      "grad_norm": 0.09716988354921341,
+      "learning_rate": 1.72412325715287e-05,
+      "loss": 9.0581,
+      "step": 85450
+    },
+    {
+      "epoch": 0.42677719793253266,
+      "grad_norm": 0.10238094627857208,
+      "learning_rate": 1.723973065658715e-05,
+      "loss": 9.0545,
+      "step": 85460
+    },
+    {
+      "epoch": 0.4268271367574721,
+      "grad_norm": 0.09213068336248398,
+      "learning_rate": 1.72382287416456e-05,
+      "loss": 9.0599,
+      "step": 85470
+    },
+    {
+      "epoch": 0.42687707558241156,
+      "grad_norm": 0.0878293439745903,
+      "learning_rate": 1.7236726826704046e-05,
+      "loss": 9.058,
+      "step": 85480
+    },
+    {
+      "epoch": 0.426927014407351,
+      "grad_norm": 0.09897506982088089,
+      "learning_rate": 1.72352249117625e-05,
+      "loss": 9.0604,
+      "step": 85490
+    },
+    {
+      "epoch": 0.42697695323229046,
+      "grad_norm": 0.09527362138032913,
+      "learning_rate": 1.7233722996820947e-05,
+      "loss": 9.0577,
+      "step": 85500
+    },
+    {
+      "epoch": 0.4270268920572299,
+      "grad_norm": 0.09203501790761948,
+      "learning_rate": 1.7232221081879397e-05,
+      "loss": 9.0629,
+      "step": 85510
+    },
+    {
+      "epoch": 0.42707683088216936,
+      "grad_norm": 0.09153282642364502,
+      "learning_rate": 1.7230719166937847e-05,
+      "loss": 9.0658,
+      "step": 85520
+    },
+    {
+      "epoch": 0.4271267697071088,
+      "grad_norm": 0.09082131832838058,
+      "learning_rate": 1.7229217251996294e-05,
+      "loss": 9.0558,
+      "step": 85530
+    },
+    {
+      "epoch": 0.42717670853204825,
+      "grad_norm": 0.09474789351224899,
+      "learning_rate": 1.7227715337054747e-05,
+      "loss": 9.0649,
+      "step": 85540
+    },
+    {
+      "epoch": 0.4272266473569877,
+      "grad_norm": 0.09650855511426926,
+      "learning_rate": 1.7226213422113194e-05,
+      "loss": 9.0518,
+      "step": 85550
+    },
+    {
+      "epoch": 0.42727658618192715,
+      "grad_norm": 0.09041815251111984,
+      "learning_rate": 1.7224711507171644e-05,
+      "loss": 9.0682,
+      "step": 85560
+    },
+    {
+      "epoch": 0.4273265250068666,
+      "grad_norm": 0.09136392921209335,
+      "learning_rate": 1.7223209592230095e-05,
+      "loss": 9.0575,
+      "step": 85570
+    },
+    {
+      "epoch": 0.42737646383180605,
+      "grad_norm": 0.09177730977535248,
+      "learning_rate": 1.722170767728854e-05,
+      "loss": 9.0442,
+      "step": 85580
+    },
+    {
+      "epoch": 0.4274264026567455,
+      "grad_norm": 0.09383508563041687,
+      "learning_rate": 1.7220205762346995e-05,
+      "loss": 9.0745,
+      "step": 85590
+    },
+    {
+      "epoch": 0.42747634148168495,
+      "grad_norm": 0.08896640688180923,
+      "learning_rate": 1.7218703847405442e-05,
+      "loss": 9.0603,
+      "step": 85600
+    },
+    {
+      "epoch": 0.4275262803066244,
+      "grad_norm": 0.0970030426979065,
+      "learning_rate": 1.7217201932463892e-05,
+      "loss": 9.0605,
+      "step": 85610
+    },
+    {
+      "epoch": 0.42757621913156385,
+      "grad_norm": 0.09975878149271011,
+      "learning_rate": 1.7215700017522342e-05,
+      "loss": 9.0549,
+      "step": 85620
+    },
+    {
+      "epoch": 0.42762615795650327,
+      "grad_norm": 0.09379398822784424,
+      "learning_rate": 1.721419810258079e-05,
+      "loss": 9.0637,
+      "step": 85630
+    },
+    {
+      "epoch": 0.42767609678144275,
+      "grad_norm": 0.09030137956142426,
+      "learning_rate": 1.7212696187639242e-05,
+      "loss": 9.0584,
+      "step": 85640
+    },
+    {
+      "epoch": 0.42772603560638217,
+      "grad_norm": 0.09938555955886841,
+      "learning_rate": 1.721119427269769e-05,
+      "loss": 9.0446,
+      "step": 85650
+    },
+    {
+      "epoch": 0.42777597443132165,
+      "grad_norm": 0.09401064366102219,
+      "learning_rate": 1.720969235775614e-05,
+      "loss": 9.0472,
+      "step": 85660
+    },
+    {
+      "epoch": 0.42782591325626107,
+      "grad_norm": 0.10197415202856064,
+      "learning_rate": 1.720819044281459e-05,
+      "loss": 9.0556,
+      "step": 85670
+    },
+    {
+      "epoch": 0.42787585208120055,
+      "grad_norm": 0.09055991470813751,
+      "learning_rate": 1.720668852787304e-05,
+      "loss": 9.0479,
+      "step": 85680
+    },
+    {
+      "epoch": 0.42792579090613997,
+      "grad_norm": 0.0905674397945404,
+      "learning_rate": 1.720518661293149e-05,
+      "loss": 9.0593,
+      "step": 85690
+    },
+    {
+      "epoch": 0.42797572973107945,
+      "grad_norm": 0.09657610207796097,
+      "learning_rate": 1.7203684697989937e-05,
+      "loss": 9.0484,
+      "step": 85700
+    },
+    {
+      "epoch": 0.42802566855601887,
+      "grad_norm": 0.09140429645776749,
+      "learning_rate": 1.7202182783048387e-05,
+      "loss": 9.0471,
+      "step": 85710
+    },
+    {
+      "epoch": 0.42807560738095835,
+      "grad_norm": 0.09539390355348587,
+      "learning_rate": 1.7200680868106837e-05,
+      "loss": 9.0487,
+      "step": 85720
+    },
+    {
+      "epoch": 0.42812554620589777,
+      "grad_norm": 0.09347163140773773,
+      "learning_rate": 1.7199178953165287e-05,
+      "loss": 9.0556,
+      "step": 85730
+    },
+    {
+      "epoch": 0.42817548503083724,
+      "grad_norm": 0.08956640213727951,
+      "learning_rate": 1.7197677038223737e-05,
+      "loss": 9.0475,
+      "step": 85740
+    },
+    {
+      "epoch": 0.42822542385577667,
+      "grad_norm": 0.09470704942941666,
+      "learning_rate": 1.7196175123282184e-05,
+      "loss": 9.0453,
+      "step": 85750
+    },
+    {
+      "epoch": 0.42827536268071614,
+      "grad_norm": 0.0948714017868042,
+      "learning_rate": 1.7194673208340634e-05,
+      "loss": 9.0497,
+      "step": 85760
+    },
+    {
+      "epoch": 0.42832530150565556,
+      "grad_norm": 0.09796865284442902,
+      "learning_rate": 1.7193171293399085e-05,
+      "loss": 9.0603,
+      "step": 85770
+    },
+    {
+      "epoch": 0.42837524033059504,
+      "grad_norm": 0.09614430367946625,
+      "learning_rate": 1.7191669378457535e-05,
+      "loss": 9.0641,
+      "step": 85780
+    },
+    {
+      "epoch": 0.42842517915553446,
+      "grad_norm": 0.09698038548231125,
+      "learning_rate": 1.7190167463515985e-05,
+      "loss": 9.0549,
+      "step": 85790
+    },
+    {
+      "epoch": 0.42847511798047394,
+      "grad_norm": 0.08912073075771332,
+      "learning_rate": 1.7188665548574432e-05,
+      "loss": 9.0442,
+      "step": 85800
+    },
+    {
+      "epoch": 0.42852505680541336,
+      "grad_norm": 0.0899166464805603,
+      "learning_rate": 1.7187163633632882e-05,
+      "loss": 9.04,
+      "step": 85810
+    },
+    {
+      "epoch": 0.42857499563035284,
+      "grad_norm": 0.09650041162967682,
+      "learning_rate": 1.7185661718691332e-05,
+      "loss": 9.0408,
+      "step": 85820
+    },
+    {
+      "epoch": 0.42862493445529226,
+      "grad_norm": 0.09182535111904144,
+      "learning_rate": 1.7184159803749782e-05,
+      "loss": 9.0617,
+      "step": 85830
+    },
+    {
+      "epoch": 0.42867487328023174,
+      "grad_norm": 0.09780148416757584,
+      "learning_rate": 1.7182657888808232e-05,
+      "loss": 9.0625,
+      "step": 85840
+    },
+    {
+      "epoch": 0.42872481210517116,
+      "grad_norm": 0.09503006935119629,
+      "learning_rate": 1.718115597386668e-05,
+      "loss": 9.0631,
+      "step": 85850
+    },
+    {
+      "epoch": 0.42877475093011064,
+      "grad_norm": 0.09269234538078308,
+      "learning_rate": 1.717965405892513e-05,
+      "loss": 9.055,
+      "step": 85860
+    },
+    {
+      "epoch": 0.42882468975505006,
+      "grad_norm": 0.09522657096385956,
+      "learning_rate": 1.717815214398358e-05,
+      "loss": 9.0593,
+      "step": 85870
+    },
+    {
+      "epoch": 0.42887462857998954,
+      "grad_norm": 0.09069998562335968,
+      "learning_rate": 1.717665022904203e-05,
+      "loss": 9.0425,
+      "step": 85880
+    },
+    {
+      "epoch": 0.42892456740492896,
+      "grad_norm": 0.09758871048688889,
+      "learning_rate": 1.717514831410048e-05,
+      "loss": 9.0392,
+      "step": 85890
+    },
+    {
+      "epoch": 0.42897450622986844,
+      "grad_norm": 0.09257560968399048,
+      "learning_rate": 1.7173646399158927e-05,
+      "loss": 9.0638,
+      "step": 85900
+    },
+    {
+      "epoch": 0.42902444505480786,
+      "grad_norm": 0.09189421683549881,
+      "learning_rate": 1.7172144484217377e-05,
+      "loss": 9.0469,
+      "step": 85910
+    },
+    {
+      "epoch": 0.42907438387974733,
+      "grad_norm": 0.09272563457489014,
+      "learning_rate": 1.7170642569275827e-05,
+      "loss": 9.0536,
+      "step": 85920
+    },
+    {
+      "epoch": 0.42912432270468676,
+      "grad_norm": 0.08748695999383926,
+      "learning_rate": 1.7169140654334277e-05,
+      "loss": 9.0641,
+      "step": 85930
+    },
+    {
+      "epoch": 0.42917426152962623,
+      "grad_norm": 0.09138987958431244,
+      "learning_rate": 1.7167638739392727e-05,
+      "loss": 9.057,
+      "step": 85940
+    },
+    {
+      "epoch": 0.42922420035456565,
+      "grad_norm": 0.09145860373973846,
+      "learning_rate": 1.7166136824451174e-05,
+      "loss": 9.0571,
+      "step": 85950
+    },
+    {
+      "epoch": 0.42927413917950513,
+      "grad_norm": 0.09510332345962524,
+      "learning_rate": 1.7164634909509624e-05,
+      "loss": 9.0455,
+      "step": 85960
+    },
+    {
+      "epoch": 0.42932407800444455,
+      "grad_norm": 0.09964827448129654,
+      "learning_rate": 1.7163132994568075e-05,
+      "loss": 9.0313,
+      "step": 85970
+    },
+    {
+      "epoch": 0.42937401682938403,
+      "grad_norm": 0.09246429055929184,
+      "learning_rate": 1.7161631079626525e-05,
+      "loss": 9.0544,
+      "step": 85980
+    },
+    {
+      "epoch": 0.42942395565432345,
+      "grad_norm": 0.10065965354442596,
+      "learning_rate": 1.7160129164684975e-05,
+      "loss": 9.0548,
+      "step": 85990
+    },
+    {
+      "epoch": 0.42947389447926293,
+      "grad_norm": 0.09086819738149643,
+      "learning_rate": 1.7158627249743425e-05,
+      "loss": 9.0661,
+      "step": 86000
+    },
+    {
+      "epoch": 0.42952383330420235,
+      "grad_norm": 0.09284672141075134,
+      "learning_rate": 1.7157125334801872e-05,
+      "loss": 9.0489,
+      "step": 86010
+    },
+    {
+      "epoch": 0.42957377212914183,
+      "grad_norm": 0.09478854387998581,
+      "learning_rate": 1.7155623419860322e-05,
+      "loss": 9.0511,
+      "step": 86020
+    },
+    {
+      "epoch": 0.42962371095408125,
+      "grad_norm": 0.0997512936592102,
+      "learning_rate": 1.7154121504918772e-05,
+      "loss": 9.0386,
+      "step": 86030
+    },
+    {
+      "epoch": 0.42967364977902067,
+      "grad_norm": 0.0923648402094841,
+      "learning_rate": 1.7152619589977222e-05,
+      "loss": 9.0533,
+      "step": 86040
+    },
+    {
+      "epoch": 0.42972358860396015,
+      "grad_norm": 0.09019633382558823,
+      "learning_rate": 1.7151117675035673e-05,
+      "loss": 9.0481,
+      "step": 86050
+    },
+    {
+      "epoch": 0.42977352742889957,
+      "grad_norm": 0.09562560170888901,
+      "learning_rate": 1.714961576009412e-05,
+      "loss": 9.0553,
+      "step": 86060
+    },
+    {
+      "epoch": 0.42982346625383905,
+      "grad_norm": 0.09037364274263382,
+      "learning_rate": 1.714811384515257e-05,
+      "loss": 9.0621,
+      "step": 86070
+    },
+    {
+      "epoch": 0.42987340507877847,
+      "grad_norm": 0.09504220634698868,
+      "learning_rate": 1.714661193021102e-05,
+      "loss": 9.0369,
+      "step": 86080
+    },
+    {
+      "epoch": 0.42992334390371795,
+      "grad_norm": 0.09927166253328323,
+      "learning_rate": 1.714511001526947e-05,
+      "loss": 9.0363,
+      "step": 86090
+    },
+    {
+      "epoch": 0.42997328272865737,
+      "grad_norm": 0.09197903424501419,
+      "learning_rate": 1.714360810032792e-05,
+      "loss": 9.0647,
+      "step": 86100
+    },
+    {
+      "epoch": 0.43002322155359685,
+      "grad_norm": 0.08889306336641312,
+      "learning_rate": 1.7142106185386367e-05,
+      "loss": 9.0551,
+      "step": 86110
+    },
+    {
+      "epoch": 0.43007316037853627,
+      "grad_norm": 0.09508626163005829,
+      "learning_rate": 1.7140604270444817e-05,
+      "loss": 9.0492,
+      "step": 86120
+    },
+    {
+      "epoch": 0.43012309920347574,
+      "grad_norm": 0.0925615131855011,
+      "learning_rate": 1.7139102355503267e-05,
+      "loss": 9.0463,
+      "step": 86130
+    },
+    {
+      "epoch": 0.43017303802841517,
+      "grad_norm": 0.0957641676068306,
+      "learning_rate": 1.7137600440561717e-05,
+      "loss": 9.0549,
+      "step": 86140
+    },
+    {
+      "epoch": 0.43022297685335464,
+      "grad_norm": 0.09168253093957901,
+      "learning_rate": 1.7136098525620168e-05,
+      "loss": 9.0504,
+      "step": 86150
+    },
+    {
+      "epoch": 0.43027291567829407,
+      "grad_norm": 0.09538407623767853,
+      "learning_rate": 1.7134596610678614e-05,
+      "loss": 9.0464,
+      "step": 86160
+    },
+    {
+      "epoch": 0.43032285450323354,
+      "grad_norm": 0.09280803054571152,
+      "learning_rate": 1.7133094695737065e-05,
+      "loss": 9.0527,
+      "step": 86170
+    },
+    {
+      "epoch": 0.43037279332817296,
+      "grad_norm": 0.09550796449184418,
+      "learning_rate": 1.7131592780795515e-05,
+      "loss": 9.0521,
+      "step": 86180
+    },
+    {
+      "epoch": 0.43042273215311244,
+      "grad_norm": 0.09481924027204514,
+      "learning_rate": 1.7130090865853965e-05,
+      "loss": 9.058,
+      "step": 86190
+    },
+    {
+      "epoch": 0.43047267097805186,
+      "grad_norm": 0.0906360000371933,
+      "learning_rate": 1.7128588950912415e-05,
+      "loss": 9.0544,
+      "step": 86200
+    },
+    {
+      "epoch": 0.43052260980299134,
+      "grad_norm": 0.1016702726483345,
+      "learning_rate": 1.7127087035970862e-05,
+      "loss": 9.0553,
+      "step": 86210
+    },
+    {
+      "epoch": 0.43057254862793076,
+      "grad_norm": 0.0934482216835022,
+      "learning_rate": 1.7125585121029312e-05,
+      "loss": 9.0411,
+      "step": 86220
+    },
+    {
+      "epoch": 0.43062248745287024,
+      "grad_norm": 0.09161151945590973,
+      "learning_rate": 1.7124083206087762e-05,
+      "loss": 9.054,
+      "step": 86230
+    },
+    {
+      "epoch": 0.43067242627780966,
+      "grad_norm": 0.09385374188423157,
+      "learning_rate": 1.7122581291146212e-05,
+      "loss": 9.0394,
+      "step": 86240
+    },
+    {
+      "epoch": 0.43072236510274914,
+      "grad_norm": 0.09769625961780548,
+      "learning_rate": 1.7121079376204663e-05,
+      "loss": 9.0444,
+      "step": 86250
+    },
+    {
+      "epoch": 0.43077230392768856,
+      "grad_norm": 0.09559006243944168,
+      "learning_rate": 1.711957746126311e-05,
+      "loss": 9.0386,
+      "step": 86260
+    },
+    {
+      "epoch": 0.43082224275262804,
+      "grad_norm": 0.0927807167172432,
+      "learning_rate": 1.711807554632156e-05,
+      "loss": 9.064,
+      "step": 86270
+    },
+    {
+      "epoch": 0.43087218157756746,
+      "grad_norm": 0.09350989013910294,
+      "learning_rate": 1.711657363138001e-05,
+      "loss": 9.0505,
+      "step": 86280
+    },
+    {
+      "epoch": 0.43092212040250694,
+      "grad_norm": 0.09480515867471695,
+      "learning_rate": 1.711507171643846e-05,
+      "loss": 9.0377,
+      "step": 86290
+    },
+    {
+      "epoch": 0.43097205922744636,
+      "grad_norm": 0.0930512323975563,
+      "learning_rate": 1.711356980149691e-05,
+      "loss": 9.0554,
+      "step": 86300
+    },
+    {
+      "epoch": 0.43102199805238584,
+      "grad_norm": 0.09537147730588913,
+      "learning_rate": 1.7112067886555357e-05,
+      "loss": 9.0552,
+      "step": 86310
+    },
+    {
+      "epoch": 0.43107193687732526,
+      "grad_norm": 0.09353245049715042,
+      "learning_rate": 1.711056597161381e-05,
+      "loss": 9.043,
+      "step": 86320
+    },
+    {
+      "epoch": 0.43112187570226473,
+      "grad_norm": 0.09246620535850525,
+      "learning_rate": 1.7109064056672257e-05,
+      "loss": 9.0528,
+      "step": 86330
+    },
+    {
+      "epoch": 0.43117181452720416,
+      "grad_norm": 0.09530563652515411,
+      "learning_rate": 1.7107562141730707e-05,
+      "loss": 9.0463,
+      "step": 86340
+    },
+    {
+      "epoch": 0.43122175335214363,
+      "grad_norm": 0.09027519822120667,
+      "learning_rate": 1.7106060226789158e-05,
+      "loss": 9.0485,
+      "step": 86350
+    },
+    {
+      "epoch": 0.43127169217708305,
+      "grad_norm": 0.09005729854106903,
+      "learning_rate": 1.7104558311847604e-05,
+      "loss": 9.0492,
+      "step": 86360
+    },
+    {
+      "epoch": 0.43132163100202253,
+      "grad_norm": 0.0943072959780693,
+      "learning_rate": 1.7103056396906058e-05,
+      "loss": 9.054,
+      "step": 86370
+    },
+    {
+      "epoch": 0.43137156982696195,
+      "grad_norm": 0.09449702501296997,
+      "learning_rate": 1.7101554481964505e-05,
+      "loss": 9.0423,
+      "step": 86380
+    },
+    {
+      "epoch": 0.43142150865190143,
+      "grad_norm": 0.09583057463169098,
+      "learning_rate": 1.7100052567022955e-05,
+      "loss": 9.0625,
+      "step": 86390
+    },
+    {
+      "epoch": 0.43147144747684085,
+      "grad_norm": 0.09351655840873718,
+      "learning_rate": 1.7098550652081405e-05,
+      "loss": 9.0511,
+      "step": 86400
+    },
+    {
+      "epoch": 0.43152138630178033,
+      "grad_norm": 0.0870162844657898,
+      "learning_rate": 1.7097048737139852e-05,
+      "loss": 9.0509,
+      "step": 86410
+    },
+    {
+      "epoch": 0.43157132512671975,
+      "grad_norm": 0.09593592584133148,
+      "learning_rate": 1.7095546822198306e-05,
+      "loss": 9.0419,
+      "step": 86420
+    },
+    {
+      "epoch": 0.43162126395165923,
+      "grad_norm": 0.0929575264453888,
+      "learning_rate": 1.7094044907256752e-05,
+      "loss": 9.0523,
+      "step": 86430
+    },
+    {
+      "epoch": 0.43167120277659865,
+      "grad_norm": 0.09491048753261566,
+      "learning_rate": 1.7092542992315202e-05,
+      "loss": 9.0308,
+      "step": 86440
+    },
+    {
+      "epoch": 0.4317211416015381,
+      "grad_norm": 0.09691637009382248,
+      "learning_rate": 1.7091041077373653e-05,
+      "loss": 9.0335,
+      "step": 86450
+    },
+    {
+      "epoch": 0.43177108042647755,
+      "grad_norm": 0.09719613939523697,
+      "learning_rate": 1.70895391624321e-05,
+      "loss": 9.0532,
+      "step": 86460
+    },
+    {
+      "epoch": 0.431821019251417,
+      "grad_norm": 0.08485887944698334,
+      "learning_rate": 1.7088037247490553e-05,
+      "loss": 9.045,
+      "step": 86470
+    },
+    {
+      "epoch": 0.43187095807635645,
+      "grad_norm": 0.09217669814825058,
+      "learning_rate": 1.7086535332549e-05,
+      "loss": 9.0437,
+      "step": 86480
+    },
+    {
+      "epoch": 0.4319208969012959,
+      "grad_norm": 0.09682594239711761,
+      "learning_rate": 1.708503341760745e-05,
+      "loss": 9.0475,
+      "step": 86490
+    },
+    {
+      "epoch": 0.43197083572623535,
+      "grad_norm": 0.09146720916032791,
+      "learning_rate": 1.70835315026659e-05,
+      "loss": 9.0369,
+      "step": 86500
+    },
+    {
+      "epoch": 0.4320207745511748,
+      "grad_norm": 0.09073561429977417,
+      "learning_rate": 1.7082029587724347e-05,
+      "loss": 9.0416,
+      "step": 86510
+    },
+    {
+      "epoch": 0.43207071337611425,
+      "grad_norm": 0.09679694473743439,
+      "learning_rate": 1.70805276727828e-05,
+      "loss": 9.0619,
+      "step": 86520
+    },
+    {
+      "epoch": 0.4321206522010537,
+      "grad_norm": 0.09839240461587906,
+      "learning_rate": 1.7079025757841247e-05,
+      "loss": 9.0497,
+      "step": 86530
+    },
+    {
+      "epoch": 0.43217059102599314,
+      "grad_norm": 0.09931270033121109,
+      "learning_rate": 1.7077523842899697e-05,
+      "loss": 9.055,
+      "step": 86540
+    },
+    {
+      "epoch": 0.4322205298509326,
+      "grad_norm": 0.09096875041723251,
+      "learning_rate": 1.7076021927958148e-05,
+      "loss": 9.0453,
+      "step": 86550
+    },
+    {
+      "epoch": 0.43227046867587204,
+      "grad_norm": 0.08868992328643799,
+      "learning_rate": 1.7074520013016594e-05,
+      "loss": 9.0377,
+      "step": 86560
+    },
+    {
+      "epoch": 0.4323204075008115,
+      "grad_norm": 0.09178798645734787,
+      "learning_rate": 1.7073018098075048e-05,
+      "loss": 9.0494,
+      "step": 86570
+    },
+    {
+      "epoch": 0.43237034632575094,
+      "grad_norm": 0.09545016288757324,
+      "learning_rate": 1.7071516183133495e-05,
+      "loss": 9.0489,
+      "step": 86580
+    },
+    {
+      "epoch": 0.4324202851506904,
+      "grad_norm": 0.0965125560760498,
+      "learning_rate": 1.7070014268191945e-05,
+      "loss": 9.0364,
+      "step": 86590
+    },
+    {
+      "epoch": 0.43247022397562984,
+      "grad_norm": 0.09696787595748901,
+      "learning_rate": 1.7068512353250395e-05,
+      "loss": 9.0447,
+      "step": 86600
+    },
+    {
+      "epoch": 0.4325201628005693,
+      "grad_norm": 0.09214460849761963,
+      "learning_rate": 1.7067010438308842e-05,
+      "loss": 9.0573,
+      "step": 86610
+    },
+    {
+      "epoch": 0.43257010162550874,
+      "grad_norm": 0.09552972763776779,
+      "learning_rate": 1.7065508523367296e-05,
+      "loss": 9.0402,
+      "step": 86620
+    },
+    {
+      "epoch": 0.4326200404504482,
+      "grad_norm": 0.09437466412782669,
+      "learning_rate": 1.7064006608425742e-05,
+      "loss": 9.0492,
+      "step": 86630
+    },
+    {
+      "epoch": 0.43266997927538764,
+      "grad_norm": 0.09180578589439392,
+      "learning_rate": 1.7062504693484196e-05,
+      "loss": 9.0461,
+      "step": 86640
+    },
+    {
+      "epoch": 0.4327199181003271,
+      "grad_norm": 0.0951187014579773,
+      "learning_rate": 1.7061002778542643e-05,
+      "loss": 9.0486,
+      "step": 86650
+    },
+    {
+      "epoch": 0.43276985692526654,
+      "grad_norm": 0.0927475169301033,
+      "learning_rate": 1.705950086360109e-05,
+      "loss": 9.04,
+      "step": 86660
+    },
+    {
+      "epoch": 0.432819795750206,
+      "grad_norm": 0.09508585184812546,
+      "learning_rate": 1.7057998948659543e-05,
+      "loss": 9.0499,
+      "step": 86670
+    },
+    {
+      "epoch": 0.43286973457514544,
+      "grad_norm": 0.09572357684373856,
+      "learning_rate": 1.705649703371799e-05,
+      "loss": 9.0267,
+      "step": 86680
+    },
+    {
+      "epoch": 0.4329196734000849,
+      "grad_norm": 0.09100820124149323,
+      "learning_rate": 1.7054995118776443e-05,
+      "loss": 9.0476,
+      "step": 86690
+    },
+    {
+      "epoch": 0.43296961222502434,
+      "grad_norm": 0.09594081342220306,
+      "learning_rate": 1.705349320383489e-05,
+      "loss": 9.0425,
+      "step": 86700
+    },
+    {
+      "epoch": 0.4330195510499638,
+      "grad_norm": 0.09707988053560257,
+      "learning_rate": 1.7051991288893337e-05,
+      "loss": 9.0399,
+      "step": 86710
+    },
+    {
+      "epoch": 0.43306948987490324,
+      "grad_norm": 0.09649763256311417,
+      "learning_rate": 1.705048937395179e-05,
+      "loss": 9.041,
+      "step": 86720
+    },
+    {
+      "epoch": 0.4331194286998427,
+      "grad_norm": 0.0913090854883194,
+      "learning_rate": 1.7048987459010237e-05,
+      "loss": 9.0297,
+      "step": 86730
+    },
+    {
+      "epoch": 0.43316936752478213,
+      "grad_norm": 0.09227044880390167,
+      "learning_rate": 1.704748554406869e-05,
+      "loss": 9.0434,
+      "step": 86740
+    },
+    {
+      "epoch": 0.4332193063497216,
+      "grad_norm": 0.09608667343854904,
+      "learning_rate": 1.7045983629127138e-05,
+      "loss": 9.0428,
+      "step": 86750
+    },
+    {
+      "epoch": 0.43326924517466103,
+      "grad_norm": 0.09107893705368042,
+      "learning_rate": 1.7044481714185584e-05,
+      "loss": 9.0533,
+      "step": 86760
+    },
+    {
+      "epoch": 0.4333191839996005,
+      "grad_norm": 0.09340152889490128,
+      "learning_rate": 1.7042979799244038e-05,
+      "loss": 9.0453,
+      "step": 86770
+    },
+    {
+      "epoch": 0.43336912282453993,
+      "grad_norm": 0.0952560156583786,
+      "learning_rate": 1.7041477884302485e-05,
+      "loss": 9.0421,
+      "step": 86780
+    },
+    {
+      "epoch": 0.4334190616494794,
+      "grad_norm": 0.09432324022054672,
+      "learning_rate": 1.703997596936094e-05,
+      "loss": 9.0402,
+      "step": 86790
+    },
+    {
+      "epoch": 0.43346900047441883,
+      "grad_norm": 0.09584697335958481,
+      "learning_rate": 1.7038474054419385e-05,
+      "loss": 9.0374,
+      "step": 86800
+    },
+    {
+      "epoch": 0.4335189392993583,
+      "grad_norm": 0.09084445238113403,
+      "learning_rate": 1.7036972139477832e-05,
+      "loss": 9.0376,
+      "step": 86810
+    },
+    {
+      "epoch": 0.43356887812429773,
+      "grad_norm": 0.09363221377134323,
+      "learning_rate": 1.7035470224536286e-05,
+      "loss": 9.0522,
+      "step": 86820
+    },
+    {
+      "epoch": 0.4336188169492372,
+      "grad_norm": 0.09301798045635223,
+      "learning_rate": 1.7033968309594732e-05,
+      "loss": 9.0359,
+      "step": 86830
+    },
+    {
+      "epoch": 0.43366875577417663,
+      "grad_norm": 0.09216799587011337,
+      "learning_rate": 1.7032466394653186e-05,
+      "loss": 9.0435,
+      "step": 86840
+    },
+    {
+      "epoch": 0.4337186945991161,
+      "grad_norm": 0.09435062110424042,
+      "learning_rate": 1.7030964479711633e-05,
+      "loss": 9.0433,
+      "step": 86850
+    },
+    {
+      "epoch": 0.4337686334240555,
+      "grad_norm": 0.08969014137983322,
+      "learning_rate": 1.702946256477008e-05,
+      "loss": 9.0326,
+      "step": 86860
+    },
+    {
+      "epoch": 0.433818572248995,
+      "grad_norm": 0.09770084917545319,
+      "learning_rate": 1.7027960649828533e-05,
+      "loss": 9.0366,
+      "step": 86870
+    },
+    {
+      "epoch": 0.4338685110739344,
+      "grad_norm": 0.0956239253282547,
+      "learning_rate": 1.702645873488698e-05,
+      "loss": 9.0302,
+      "step": 86880
+    },
+    {
+      "epoch": 0.4339184498988739,
+      "grad_norm": 0.09152868390083313,
+      "learning_rate": 1.7024956819945433e-05,
+      "loss": 9.0351,
+      "step": 86890
+    },
+    {
+      "epoch": 0.4339683887238133,
+      "grad_norm": 0.09322716295719147,
+      "learning_rate": 1.702345490500388e-05,
+      "loss": 9.0341,
+      "step": 86900
+    },
+    {
+      "epoch": 0.4340183275487528,
+      "grad_norm": 0.08923067152500153,
+      "learning_rate": 1.7021952990062327e-05,
+      "loss": 9.0475,
+      "step": 86910
+    },
+    {
+      "epoch": 0.4340682663736922,
+      "grad_norm": 0.09851467609405518,
+      "learning_rate": 1.702045107512078e-05,
+      "loss": 9.0337,
+      "step": 86920
+    },
+    {
+      "epoch": 0.4341182051986317,
+      "grad_norm": 0.09365841746330261,
+      "learning_rate": 1.7018949160179227e-05,
+      "loss": 9.0379,
+      "step": 86930
+    },
+    {
+      "epoch": 0.4341681440235711,
+      "grad_norm": 0.09401962906122208,
+      "learning_rate": 1.701744724523768e-05,
+      "loss": 9.053,
+      "step": 86940
+    },
+    {
+      "epoch": 0.4342180828485106,
+      "grad_norm": 0.09199943393468857,
+      "learning_rate": 1.7015945330296128e-05,
+      "loss": 9.0282,
+      "step": 86950
+    },
+    {
+      "epoch": 0.43426802167345,
+      "grad_norm": 0.09017936885356903,
+      "learning_rate": 1.7014443415354578e-05,
+      "loss": 9.0395,
+      "step": 86960
+    },
+    {
+      "epoch": 0.4343179604983895,
+      "grad_norm": 0.09726618975400925,
+      "learning_rate": 1.7012941500413028e-05,
+      "loss": 9.0404,
+      "step": 86970
+    },
+    {
+      "epoch": 0.4343678993233289,
+      "grad_norm": 0.0918731540441513,
+      "learning_rate": 1.7011439585471475e-05,
+      "loss": 9.0201,
+      "step": 86980
+    },
+    {
+      "epoch": 0.4344178381482684,
+      "grad_norm": 0.09042965620756149,
+      "learning_rate": 1.700993767052993e-05,
+      "loss": 9.035,
+      "step": 86990
+    },
+    {
+      "epoch": 0.4344677769732078,
+      "grad_norm": 0.09324657171964645,
+      "learning_rate": 1.7008435755588375e-05,
+      "loss": 9.0361,
+      "step": 87000
+    },
+    {
+      "epoch": 0.4345177157981473,
+      "grad_norm": 0.09195488691329956,
+      "learning_rate": 1.7006933840646825e-05,
+      "loss": 9.049,
+      "step": 87010
+    },
+    {
+      "epoch": 0.4345676546230867,
+      "grad_norm": 0.08934830129146576,
+      "learning_rate": 1.7005431925705276e-05,
+      "loss": 9.0324,
+      "step": 87020
+    },
+    {
+      "epoch": 0.43461759344802614,
+      "grad_norm": 0.0945764034986496,
+      "learning_rate": 1.7003930010763722e-05,
+      "loss": 9.0484,
+      "step": 87030
+    },
+    {
+      "epoch": 0.4346675322729656,
+      "grad_norm": 0.09223344922065735,
+      "learning_rate": 1.7002428095822176e-05,
+      "loss": 9.0433,
+      "step": 87040
+    },
+    {
+      "epoch": 0.43471747109790504,
+      "grad_norm": 0.09764394164085388,
+      "learning_rate": 1.7000926180880623e-05,
+      "loss": 9.0318,
+      "step": 87050
+    },
+    {
+      "epoch": 0.4347674099228445,
+      "grad_norm": 0.08980736136436462,
+      "learning_rate": 1.6999424265939073e-05,
+      "loss": 9.0387,
+      "step": 87060
+    },
+    {
+      "epoch": 0.43481734874778394,
+      "grad_norm": 0.0914120152592659,
+      "learning_rate": 1.6997922350997523e-05,
+      "loss": 9.052,
+      "step": 87070
+    },
+    {
+      "epoch": 0.4348672875727234,
+      "grad_norm": 0.09246059507131577,
+      "learning_rate": 1.699642043605597e-05,
+      "loss": 9.0412,
+      "step": 87080
+    },
+    {
+      "epoch": 0.43491722639766284,
+      "grad_norm": 0.09065555036067963,
+      "learning_rate": 1.6994918521114423e-05,
+      "loss": 9.035,
+      "step": 87090
+    },
+    {
+      "epoch": 0.4349671652226023,
+      "grad_norm": 0.0936904326081276,
+      "learning_rate": 1.699341660617287e-05,
+      "loss": 9.0367,
+      "step": 87100
+    },
+    {
+      "epoch": 0.43501710404754174,
+      "grad_norm": 0.1051855981349945,
+      "learning_rate": 1.699191469123132e-05,
+      "loss": 9.0431,
+      "step": 87110
+    },
+    {
+      "epoch": 0.4350670428724812,
+      "grad_norm": 0.08944005519151688,
+      "learning_rate": 1.699041277628977e-05,
+      "loss": 9.0287,
+      "step": 87120
+    },
+    {
+      "epoch": 0.43511698169742064,
+      "grad_norm": 0.09485218673944473,
+      "learning_rate": 1.6988910861348217e-05,
+      "loss": 9.0493,
+      "step": 87130
+    },
+    {
+      "epoch": 0.4351669205223601,
+      "grad_norm": 0.09217767417430878,
+      "learning_rate": 1.698740894640667e-05,
+      "loss": 9.0362,
+      "step": 87140
+    },
+    {
+      "epoch": 0.43521685934729953,
+      "grad_norm": 0.09442111849784851,
+      "learning_rate": 1.6985907031465118e-05,
+      "loss": 9.0246,
+      "step": 87150
+    },
+    {
+      "epoch": 0.435266798172239,
+      "grad_norm": 0.10553400963544846,
+      "learning_rate": 1.6984405116523568e-05,
+      "loss": 9.0343,
+      "step": 87160
+    },
+    {
+      "epoch": 0.43531673699717843,
+      "grad_norm": 0.08932023495435715,
+      "learning_rate": 1.6982903201582018e-05,
+      "loss": 9.0449,
+      "step": 87170
+    },
+    {
+      "epoch": 0.4353666758221179,
+      "grad_norm": 0.09344980865716934,
+      "learning_rate": 1.6981401286640465e-05,
+      "loss": 9.0279,
+      "step": 87180
+    },
+    {
+      "epoch": 0.43541661464705733,
+      "grad_norm": 0.09347915649414062,
+      "learning_rate": 1.697989937169892e-05,
+      "loss": 9.0341,
+      "step": 87190
+    },
+    {
+      "epoch": 0.4354665534719968,
+      "grad_norm": 0.0941552221775055,
+      "learning_rate": 1.6978397456757365e-05,
+      "loss": 9.0331,
+      "step": 87200
+    },
+    {
+      "epoch": 0.43551649229693623,
+      "grad_norm": 0.09416943788528442,
+      "learning_rate": 1.6976895541815815e-05,
+      "loss": 9.0384,
+      "step": 87210
+    },
+    {
+      "epoch": 0.4355664311218757,
+      "grad_norm": 0.0943121612071991,
+      "learning_rate": 1.6975393626874266e-05,
+      "loss": 9.038,
+      "step": 87220
+    },
+    {
+      "epoch": 0.43561636994681513,
+      "grad_norm": 0.09376168996095657,
+      "learning_rate": 1.6973891711932712e-05,
+      "loss": 9.0261,
+      "step": 87230
+    },
+    {
+      "epoch": 0.4356663087717546,
+      "grad_norm": 0.09052184224128723,
+      "learning_rate": 1.6972389796991166e-05,
+      "loss": 9.0356,
+      "step": 87240
+    },
+    {
+      "epoch": 0.43571624759669403,
+      "grad_norm": 0.0898606926202774,
+      "learning_rate": 1.6970887882049613e-05,
+      "loss": 9.0347,
+      "step": 87250
+    },
+    {
+      "epoch": 0.4357661864216335,
+      "grad_norm": 0.09599337726831436,
+      "learning_rate": 1.6969385967108063e-05,
+      "loss": 9.0272,
+      "step": 87260
+    },
+    {
+      "epoch": 0.4358161252465729,
+      "grad_norm": 0.093189537525177,
+      "learning_rate": 1.6967884052166513e-05,
+      "loss": 9.033,
+      "step": 87270
+    },
+    {
+      "epoch": 0.4358660640715124,
+      "grad_norm": 0.09416911005973816,
+      "learning_rate": 1.6966382137224963e-05,
+      "loss": 9.0371,
+      "step": 87280
+    },
+    {
+      "epoch": 0.4359160028964518,
+      "grad_norm": 0.09415122121572495,
+      "learning_rate": 1.6964880222283413e-05,
+      "loss": 9.0361,
+      "step": 87290
+    },
+    {
+      "epoch": 0.4359659417213913,
+      "grad_norm": 0.09333626925945282,
+      "learning_rate": 1.696337830734186e-05,
+      "loss": 9.0317,
+      "step": 87300
+    },
+    {
+      "epoch": 0.4360158805463307,
+      "grad_norm": 0.09361656755208969,
+      "learning_rate": 1.696187639240031e-05,
+      "loss": 9.0233,
+      "step": 87310
+    },
+    {
+      "epoch": 0.4360658193712702,
+      "grad_norm": 0.0928577408194542,
+      "learning_rate": 1.696037447745876e-05,
+      "loss": 9.044,
+      "step": 87320
+    },
+    {
+      "epoch": 0.4361157581962096,
+      "grad_norm": 0.09233950078487396,
+      "learning_rate": 1.695887256251721e-05,
+      "loss": 9.0395,
+      "step": 87330
+    },
+    {
+      "epoch": 0.4361656970211491,
+      "grad_norm": 0.09364930540323257,
+      "learning_rate": 1.695737064757566e-05,
+      "loss": 9.0438,
+      "step": 87340
+    },
+    {
+      "epoch": 0.4362156358460885,
+      "grad_norm": 0.09141746908426285,
+      "learning_rate": 1.6955868732634108e-05,
+      "loss": 9.0329,
+      "step": 87350
+    },
+    {
+      "epoch": 0.436265574671028,
+      "grad_norm": 0.08941878378391266,
+      "learning_rate": 1.6954366817692558e-05,
+      "loss": 9.0257,
+      "step": 87360
+    },
+    {
+      "epoch": 0.4363155134959674,
+      "grad_norm": 0.09561556577682495,
+      "learning_rate": 1.6952864902751008e-05,
+      "loss": 9.0315,
+      "step": 87370
+    },
+    {
+      "epoch": 0.4363654523209069,
+      "grad_norm": 0.09164825826883316,
+      "learning_rate": 1.6951362987809458e-05,
+      "loss": 9.027,
+      "step": 87380
+    },
+    {
+      "epoch": 0.4364153911458463,
+      "grad_norm": 0.0955611914396286,
+      "learning_rate": 1.694986107286791e-05,
+      "loss": 9.034,
+      "step": 87390
+    },
+    {
+      "epoch": 0.4364653299707858,
+      "grad_norm": 0.09065394103527069,
+      "learning_rate": 1.6948359157926355e-05,
+      "loss": 9.0308,
+      "step": 87400
+    },
+    {
+      "epoch": 0.4365152687957252,
+      "grad_norm": 0.09663858264684677,
+      "learning_rate": 1.6946857242984805e-05,
+      "loss": 9.0187,
+      "step": 87410
+    },
+    {
+      "epoch": 0.4365652076206647,
+      "grad_norm": 0.0890546515583992,
+      "learning_rate": 1.6945355328043256e-05,
+      "loss": 9.0392,
+      "step": 87420
+    },
+    {
+      "epoch": 0.4366151464456041,
+      "grad_norm": 0.09616228938102722,
+      "learning_rate": 1.6943853413101706e-05,
+      "loss": 9.0342,
+      "step": 87430
+    },
+    {
+      "epoch": 0.4366650852705436,
+      "grad_norm": 0.10570753365755081,
+      "learning_rate": 1.6942351498160156e-05,
+      "loss": 9.0383,
+      "step": 87440
+    },
+    {
+      "epoch": 0.436715024095483,
+      "grad_norm": 0.09239917248487473,
+      "learning_rate": 1.6940849583218603e-05,
+      "loss": 9.0355,
+      "step": 87450
+    },
+    {
+      "epoch": 0.4367649629204225,
+      "grad_norm": 0.09370817989110947,
+      "learning_rate": 1.6939347668277053e-05,
+      "loss": 9.0242,
+      "step": 87460
+    },
+    {
+      "epoch": 0.4368149017453619,
+      "grad_norm": 0.09490583091974258,
+      "learning_rate": 1.6937845753335503e-05,
+      "loss": 9.0441,
+      "step": 87470
+    },
+    {
+      "epoch": 0.4368648405703014,
+      "grad_norm": 0.0928240418434143,
+      "learning_rate": 1.6936343838393953e-05,
+      "loss": 9.0362,
+      "step": 87480
+    },
+    {
+      "epoch": 0.4369147793952408,
+      "grad_norm": 0.09075388312339783,
+      "learning_rate": 1.6934841923452403e-05,
+      "loss": 9.0322,
+      "step": 87490
+    },
+    {
+      "epoch": 0.4369647182201803,
+      "grad_norm": 0.09307154268026352,
+      "learning_rate": 1.693334000851085e-05,
+      "loss": 9.0344,
+      "step": 87500
+    },
+    {
+      "epoch": 0.4370146570451197,
+      "grad_norm": 0.0918651670217514,
+      "learning_rate": 1.69318380935693e-05,
+      "loss": 9.0393,
+      "step": 87510
+    },
+    {
+      "epoch": 0.4370645958700592,
+      "grad_norm": 0.0964551717042923,
+      "learning_rate": 1.693033617862775e-05,
+      "loss": 9.0446,
+      "step": 87520
+    },
+    {
+      "epoch": 0.4371145346949986,
+      "grad_norm": 0.09495722502470016,
+      "learning_rate": 1.69288342636862e-05,
+      "loss": 9.0284,
+      "step": 87530
+    },
+    {
+      "epoch": 0.4371644735199381,
+      "grad_norm": 0.09483332186937332,
+      "learning_rate": 1.692733234874465e-05,
+      "loss": 9.0341,
+      "step": 87540
+    },
+    {
+      "epoch": 0.4372144123448775,
+      "grad_norm": 0.08984597027301788,
+      "learning_rate": 1.6925830433803098e-05,
+      "loss": 9.0351,
+      "step": 87550
+    },
+    {
+      "epoch": 0.437264351169817,
+      "grad_norm": 0.09564351290464401,
+      "learning_rate": 1.6924328518861548e-05,
+      "loss": 9.023,
+      "step": 87560
+    },
+    {
+      "epoch": 0.4373142899947564,
+      "grad_norm": 0.09509795159101486,
+      "learning_rate": 1.6922826603919998e-05,
+      "loss": 9.0338,
+      "step": 87570
+    },
+    {
+      "epoch": 0.4373642288196959,
+      "grad_norm": 0.09072297811508179,
+      "learning_rate": 1.6921324688978448e-05,
+      "loss": 9.0442,
+      "step": 87580
+    },
+    {
+      "epoch": 0.4374141676446353,
+      "grad_norm": 0.09715980291366577,
+      "learning_rate": 1.69198227740369e-05,
+      "loss": 9.0252,
+      "step": 87590
+    },
+    {
+      "epoch": 0.4374641064695748,
+      "grad_norm": 0.09227906912565231,
+      "learning_rate": 1.691832085909535e-05,
+      "loss": 9.0273,
+      "step": 87600
+    },
+    {
+      "epoch": 0.4375140452945142,
+      "grad_norm": 0.10533630102872849,
+      "learning_rate": 1.6916818944153795e-05,
+      "loss": 9.0322,
+      "step": 87610
+    },
+    {
+      "epoch": 0.4375639841194537,
+      "grad_norm": 0.09329358488321304,
+      "learning_rate": 1.6915317029212246e-05,
+      "loss": 9.0214,
+      "step": 87620
+    },
+    {
+      "epoch": 0.4376139229443931,
+      "grad_norm": 0.09366413950920105,
+      "learning_rate": 1.6913815114270696e-05,
+      "loss": 9.0308,
+      "step": 87630
+    },
+    {
+      "epoch": 0.4376638617693326,
+      "grad_norm": 0.10361622273921967,
+      "learning_rate": 1.6912313199329146e-05,
+      "loss": 9.0371,
+      "step": 87640
+    },
+    {
+      "epoch": 0.437713800594272,
+      "grad_norm": 0.09569653123617172,
+      "learning_rate": 1.6910811284387596e-05,
+      "loss": 9.0347,
+      "step": 87650
+    },
+    {
+      "epoch": 0.4377637394192115,
+      "grad_norm": 0.09628830850124359,
+      "learning_rate": 1.6909309369446043e-05,
+      "loss": 9.0361,
+      "step": 87660
+    },
+    {
+      "epoch": 0.4378136782441509,
+      "grad_norm": 0.09476719051599503,
+      "learning_rate": 1.6907807454504493e-05,
+      "loss": 9.0287,
+      "step": 87670
+    },
+    {
+      "epoch": 0.4378636170690904,
+      "grad_norm": 0.09486107528209686,
+      "learning_rate": 1.6906305539562943e-05,
+      "loss": 9.0361,
+      "step": 87680
+    },
+    {
+      "epoch": 0.4379135558940298,
+      "grad_norm": 0.08894047141075134,
+      "learning_rate": 1.6904803624621393e-05,
+      "loss": 9.0196,
+      "step": 87690
+    },
+    {
+      "epoch": 0.4379634947189693,
+      "grad_norm": 0.09540939331054688,
+      "learning_rate": 1.6903301709679844e-05,
+      "loss": 9.039,
+      "step": 87700
+    },
+    {
+      "epoch": 0.4380134335439087,
+      "grad_norm": 0.09612215310335159,
+      "learning_rate": 1.690179979473829e-05,
+      "loss": 9.0244,
+      "step": 87710
+    },
+    {
+      "epoch": 0.4380633723688482,
+      "grad_norm": 0.0884467363357544,
+      "learning_rate": 1.690029787979674e-05,
+      "loss": 9.0421,
+      "step": 87720
+    },
+    {
+      "epoch": 0.4381133111937876,
+      "grad_norm": 0.09004423022270203,
+      "learning_rate": 1.689879596485519e-05,
+      "loss": 9.0254,
+      "step": 87730
+    },
+    {
+      "epoch": 0.4381632500187271,
+      "grad_norm": 0.09423902630805969,
+      "learning_rate": 1.689729404991364e-05,
+      "loss": 9.0198,
+      "step": 87740
+    },
+    {
+      "epoch": 0.4382131888436665,
+      "grad_norm": 0.09269099682569504,
+      "learning_rate": 1.689579213497209e-05,
+      "loss": 9.0444,
+      "step": 87750
+    },
+    {
+      "epoch": 0.438263127668606,
+      "grad_norm": 0.0918196514248848,
+      "learning_rate": 1.6894290220030538e-05,
+      "loss": 9.024,
+      "step": 87760
+    },
+    {
+      "epoch": 0.4383130664935454,
+      "grad_norm": 0.09599689394235611,
+      "learning_rate": 1.6892788305088988e-05,
+      "loss": 9.0233,
+      "step": 87770
+    },
+    {
+      "epoch": 0.4383630053184849,
+      "grad_norm": 0.09426075220108032,
+      "learning_rate": 1.6891286390147438e-05,
+      "loss": 9.0151,
+      "step": 87780
+    },
+    {
+      "epoch": 0.4384129441434243,
+      "grad_norm": 0.09202740341424942,
+      "learning_rate": 1.688978447520589e-05,
+      "loss": 9.0304,
+      "step": 87790
+    },
+    {
+      "epoch": 0.4384628829683638,
+      "grad_norm": 0.09728818386793137,
+      "learning_rate": 1.688828256026434e-05,
+      "loss": 9.0325,
+      "step": 87800
+    },
+    {
+      "epoch": 0.4385128217933032,
+      "grad_norm": 0.09795814007520676,
+      "learning_rate": 1.6886780645322785e-05,
+      "loss": 9.0093,
+      "step": 87810
+    },
+    {
+      "epoch": 0.4385627606182427,
+      "grad_norm": 0.0872364342212677,
+      "learning_rate": 1.6885278730381236e-05,
+      "loss": 9.0299,
+      "step": 87820
+    },
+    {
+      "epoch": 0.4386126994431821,
+      "grad_norm": 0.09706996381282806,
+      "learning_rate": 1.6883776815439686e-05,
+      "loss": 9.0341,
+      "step": 87830
+    },
+    {
+      "epoch": 0.4386626382681216,
+      "grad_norm": 0.08879675716161728,
+      "learning_rate": 1.6882274900498136e-05,
+      "loss": 9.0358,
+      "step": 87840
+    },
+    {
+      "epoch": 0.438712577093061,
+      "grad_norm": 0.09184462577104568,
+      "learning_rate": 1.6880772985556586e-05,
+      "loss": 9.0178,
+      "step": 87850
+    },
+    {
+      "epoch": 0.4387625159180005,
+      "grad_norm": 0.09184456616640091,
+      "learning_rate": 1.6879271070615033e-05,
+      "loss": 9.0246,
+      "step": 87860
+    },
+    {
+      "epoch": 0.4388124547429399,
+      "grad_norm": 0.08778972923755646,
+      "learning_rate": 1.6877769155673483e-05,
+      "loss": 9.0291,
+      "step": 87870
+    },
+    {
+      "epoch": 0.43886239356787937,
+      "grad_norm": 0.09503248333930969,
+      "learning_rate": 1.6876267240731933e-05,
+      "loss": 9.0374,
+      "step": 87880
+    },
+    {
+      "epoch": 0.4389123323928188,
+      "grad_norm": 0.08978200703859329,
+      "learning_rate": 1.6874765325790383e-05,
+      "loss": 9.0381,
+      "step": 87890
+    },
+    {
+      "epoch": 0.43896227121775827,
+      "grad_norm": 0.098361074924469,
+      "learning_rate": 1.6873263410848834e-05,
+      "loss": 9.0196,
+      "step": 87900
+    },
+    {
+      "epoch": 0.4390122100426977,
+      "grad_norm": 0.09547007083892822,
+      "learning_rate": 1.687176149590728e-05,
+      "loss": 9.0084,
+      "step": 87910
+    },
+    {
+      "epoch": 0.43906214886763717,
+      "grad_norm": 0.09400343149900436,
+      "learning_rate": 1.6870259580965734e-05,
+      "loss": 9.0175,
+      "step": 87920
+    },
+    {
+      "epoch": 0.4391120876925766,
+      "grad_norm": 0.09383000433444977,
+      "learning_rate": 1.686875766602418e-05,
+      "loss": 9.0114,
+      "step": 87930
+    },
+    {
+      "epoch": 0.43916202651751607,
+      "grad_norm": 0.09463747590780258,
+      "learning_rate": 1.686725575108263e-05,
+      "loss": 9.0246,
+      "step": 87940
+    },
+    {
+      "epoch": 0.4392119653424555,
+      "grad_norm": 0.09992758929729462,
+      "learning_rate": 1.686575383614108e-05,
+      "loss": 9.0191,
+      "step": 87950
+    },
+    {
+      "epoch": 0.43926190416739497,
+      "grad_norm": 0.09506499767303467,
+      "learning_rate": 1.6864251921199528e-05,
+      "loss": 9.0212,
+      "step": 87960
+    },
+    {
+      "epoch": 0.4393118429923344,
+      "grad_norm": 0.09309326857328415,
+      "learning_rate": 1.686275000625798e-05,
+      "loss": 9.0304,
+      "step": 87970
+    },
+    {
+      "epoch": 0.43936178181727387,
+      "grad_norm": 0.09387887269258499,
+      "learning_rate": 1.686124809131643e-05,
+      "loss": 9.0263,
+      "step": 87980
+    },
+    {
+      "epoch": 0.4394117206422133,
+      "grad_norm": 0.09608437120914459,
+      "learning_rate": 1.685974617637488e-05,
+      "loss": 9.0282,
+      "step": 87990
+    },
+    {
+      "epoch": 0.43946165946715277,
+      "grad_norm": 0.10036759823560715,
+      "learning_rate": 1.685824426143333e-05,
+      "loss": 9.0298,
+      "step": 88000
+    },
+    {
+      "epoch": 0.4395115982920922,
+      "grad_norm": 0.09929006546735764,
+      "learning_rate": 1.6856742346491775e-05,
+      "loss": 9.0269,
+      "step": 88010
+    },
+    {
+      "epoch": 0.4395615371170316,
+      "grad_norm": 0.08878950029611588,
+      "learning_rate": 1.685524043155023e-05,
+      "loss": 9.0244,
+      "step": 88020
+    },
+    {
+      "epoch": 0.4396114759419711,
+      "grad_norm": 0.09790655225515366,
+      "learning_rate": 1.6853738516608676e-05,
+      "loss": 9.0289,
+      "step": 88030
+    },
+    {
+      "epoch": 0.4396614147669105,
+      "grad_norm": 0.09261323511600494,
+      "learning_rate": 1.6852236601667126e-05,
+      "loss": 9.032,
+      "step": 88040
+    },
+    {
+      "epoch": 0.43971135359185,
+      "grad_norm": 0.09531623125076294,
+      "learning_rate": 1.6850734686725576e-05,
+      "loss": 9.0355,
+      "step": 88050
+    },
+    {
+      "epoch": 0.4397612924167894,
+      "grad_norm": 0.09662208706140518,
+      "learning_rate": 1.6849232771784023e-05,
+      "loss": 9.0177,
+      "step": 88060
+    },
+    {
+      "epoch": 0.4398112312417289,
+      "grad_norm": 0.09118068218231201,
+      "learning_rate": 1.6847730856842477e-05,
+      "loss": 9.017,
+      "step": 88070
+    },
+    {
+      "epoch": 0.4398611700666683,
+      "grad_norm": 0.09556883573532104,
+      "learning_rate": 1.6846228941900923e-05,
+      "loss": 9.0367,
+      "step": 88080
+    },
+    {
+      "epoch": 0.4399111088916078,
+      "grad_norm": 0.08983524143695831,
+      "learning_rate": 1.6844727026959373e-05,
+      "loss": 9.0285,
+      "step": 88090
+    },
+    {
+      "epoch": 0.4399610477165472,
+      "grad_norm": 0.09559369832277298,
+      "learning_rate": 1.6843225112017824e-05,
+      "loss": 9.0309,
+      "step": 88100
+    },
+    {
+      "epoch": 0.4400109865414867,
+      "grad_norm": 0.09884733706712723,
+      "learning_rate": 1.684172319707627e-05,
+      "loss": 9.0156,
+      "step": 88110
+    },
+    {
+      "epoch": 0.4400609253664261,
+      "grad_norm": 0.08653125166893005,
+      "learning_rate": 1.6840221282134724e-05,
+      "loss": 9.0307,
+      "step": 88120
+    },
+    {
+      "epoch": 0.4401108641913656,
+      "grad_norm": 0.09182146936655045,
+      "learning_rate": 1.683871936719317e-05,
+      "loss": 9.0374,
+      "step": 88130
+    },
+    {
+      "epoch": 0.440160803016305,
+      "grad_norm": 0.091147780418396,
+      "learning_rate": 1.683721745225162e-05,
+      "loss": 9.0369,
+      "step": 88140
+    },
+    {
+      "epoch": 0.4402107418412445,
+      "grad_norm": 0.09663031250238419,
+      "learning_rate": 1.683571553731007e-05,
+      "loss": 9.0242,
+      "step": 88150
+    },
+    {
+      "epoch": 0.4402606806661839,
+      "grad_norm": 0.09547109156847,
+      "learning_rate": 1.6834213622368518e-05,
+      "loss": 9.0246,
+      "step": 88160
+    },
+    {
+      "epoch": 0.4403106194911234,
+      "grad_norm": 0.09579718112945557,
+      "learning_rate": 1.683271170742697e-05,
+      "loss": 9.022,
+      "step": 88170
+    },
+    {
+      "epoch": 0.4403605583160628,
+      "grad_norm": 0.09461940824985504,
+      "learning_rate": 1.683120979248542e-05,
+      "loss": 9.0285,
+      "step": 88180
+    },
+    {
+      "epoch": 0.4404104971410023,
+      "grad_norm": 0.09216568619012833,
+      "learning_rate": 1.682970787754387e-05,
+      "loss": 9.0253,
+      "step": 88190
+    },
+    {
+      "epoch": 0.4404604359659417,
+      "grad_norm": 0.08952445536851883,
+      "learning_rate": 1.682820596260232e-05,
+      "loss": 9.0147,
+      "step": 88200
+    },
+    {
+      "epoch": 0.4405103747908812,
+      "grad_norm": 0.09351477026939392,
+      "learning_rate": 1.6826704047660765e-05,
+      "loss": 9.0297,
+      "step": 88210
+    },
+    {
+      "epoch": 0.4405603136158206,
+      "grad_norm": 0.09434119611978531,
+      "learning_rate": 1.682520213271922e-05,
+      "loss": 9.0314,
+      "step": 88220
+    },
+    {
+      "epoch": 0.4406102524407601,
+      "grad_norm": 0.09438735991716385,
+      "learning_rate": 1.6823700217777666e-05,
+      "loss": 9.0262,
+      "step": 88230
+    },
+    {
+      "epoch": 0.4406601912656995,
+      "grad_norm": 0.08906865119934082,
+      "learning_rate": 1.6822198302836116e-05,
+      "loss": 9.0275,
+      "step": 88240
+    },
+    {
+      "epoch": 0.440710130090639,
+      "grad_norm": 0.09299363940954208,
+      "learning_rate": 1.6820696387894566e-05,
+      "loss": 9.0376,
+      "step": 88250
+    },
+    {
+      "epoch": 0.4407600689155784,
+      "grad_norm": 0.09099581837654114,
+      "learning_rate": 1.6819194472953013e-05,
+      "loss": 9.03,
+      "step": 88260
+    },
+    {
+      "epoch": 0.4408100077405179,
+      "grad_norm": 0.08823831379413605,
+      "learning_rate": 1.6817692558011467e-05,
+      "loss": 9.0356,
+      "step": 88270
+    },
+    {
+      "epoch": 0.4408599465654573,
+      "grad_norm": 0.0898958370089531,
+      "learning_rate": 1.6816190643069913e-05,
+      "loss": 9.0328,
+      "step": 88280
+    },
+    {
+      "epoch": 0.44090988539039677,
+      "grad_norm": 0.09236274659633636,
+      "learning_rate": 1.6814688728128367e-05,
+      "loss": 9.0355,
+      "step": 88290
+    },
+    {
+      "epoch": 0.4409598242153362,
+      "grad_norm": 0.08883727341890335,
+      "learning_rate": 1.6813186813186814e-05,
+      "loss": 9.0354,
+      "step": 88300
+    },
+    {
+      "epoch": 0.44100976304027567,
+      "grad_norm": 0.09699812531471252,
+      "learning_rate": 1.681168489824526e-05,
+      "loss": 9.0269,
+      "step": 88310
+    },
+    {
+      "epoch": 0.4410597018652151,
+      "grad_norm": 0.09533574432134628,
+      "learning_rate": 1.6810182983303714e-05,
+      "loss": 9.0313,
+      "step": 88320
+    },
+    {
+      "epoch": 0.44110964069015457,
+      "grad_norm": 0.09363710880279541,
+      "learning_rate": 1.680868106836216e-05,
+      "loss": 9.0249,
+      "step": 88330
+    },
+    {
+      "epoch": 0.441159579515094,
+      "grad_norm": 0.09056729823350906,
+      "learning_rate": 1.6807179153420614e-05,
+      "loss": 9.013,
+      "step": 88340
+    },
+    {
+      "epoch": 0.44120951834003347,
+      "grad_norm": 0.08941638469696045,
+      "learning_rate": 1.680567723847906e-05,
+      "loss": 9.0277,
+      "step": 88350
+    },
+    {
+      "epoch": 0.4412594571649729,
+      "grad_norm": 0.09786325693130493,
+      "learning_rate": 1.6804175323537508e-05,
+      "loss": 9.0257,
+      "step": 88360
+    },
+    {
+      "epoch": 0.44130939598991237,
+      "grad_norm": 0.09090113639831543,
+      "learning_rate": 1.680267340859596e-05,
+      "loss": 9.0268,
+      "step": 88370
+    },
+    {
+      "epoch": 0.4413593348148518,
+      "grad_norm": 0.09401243180036545,
+      "learning_rate": 1.680117149365441e-05,
+      "loss": 9.0274,
+      "step": 88380
+    },
+    {
+      "epoch": 0.44140927363979127,
+      "grad_norm": 0.09182031452655792,
+      "learning_rate": 1.6799669578712862e-05,
+      "loss": 9.0217,
+      "step": 88390
+    },
+    {
+      "epoch": 0.4414592124647307,
+      "grad_norm": 0.09471350908279419,
+      "learning_rate": 1.679816766377131e-05,
+      "loss": 9.0152,
+      "step": 88400
+    },
+    {
+      "epoch": 0.44150915128967017,
+      "grad_norm": 0.0950457975268364,
+      "learning_rate": 1.6796665748829755e-05,
+      "loss": 9.0367,
+      "step": 88410
+    },
+    {
+      "epoch": 0.4415590901146096,
+      "grad_norm": 0.09334580600261688,
+      "learning_rate": 1.679516383388821e-05,
+      "loss": 9.0324,
+      "step": 88420
+    },
+    {
+      "epoch": 0.44160902893954906,
+      "grad_norm": 0.09345501661300659,
+      "learning_rate": 1.6793661918946656e-05,
+      "loss": 9.0421,
+      "step": 88430
+    },
+    {
+      "epoch": 0.4416589677644885,
+      "grad_norm": 0.09563722461462021,
+      "learning_rate": 1.679216000400511e-05,
+      "loss": 9.0331,
+      "step": 88440
+    },
+    {
+      "epoch": 0.44170890658942796,
+      "grad_norm": 0.09366366267204285,
+      "learning_rate": 1.6790658089063556e-05,
+      "loss": 9.022,
+      "step": 88450
+    },
+    {
+      "epoch": 0.4417588454143674,
+      "grad_norm": 0.09602910280227661,
+      "learning_rate": 1.6789156174122003e-05,
+      "loss": 9.0281,
+      "step": 88460
+    },
+    {
+      "epoch": 0.44180878423930686,
+      "grad_norm": 0.09258101880550385,
+      "learning_rate": 1.6787654259180457e-05,
+      "loss": 9.0255,
+      "step": 88470
+    },
+    {
+      "epoch": 0.4418587230642463,
+      "grad_norm": 0.09126337617635727,
+      "learning_rate": 1.6786152344238903e-05,
+      "loss": 9.0267,
+      "step": 88480
+    },
+    {
+      "epoch": 0.44190866188918576,
+      "grad_norm": 0.0905313566327095,
+      "learning_rate": 1.6784650429297357e-05,
+      "loss": 9.0256,
+      "step": 88490
+    },
+    {
+      "epoch": 0.4419586007141252,
+      "grad_norm": 0.09191538393497467,
+      "learning_rate": 1.6783148514355804e-05,
+      "loss": 9.0323,
+      "step": 88500
+    },
+    {
+      "epoch": 0.44200853953906466,
+      "grad_norm": 0.08896350860595703,
+      "learning_rate": 1.678164659941425e-05,
+      "loss": 9.0267,
+      "step": 88510
+    },
+    {
+      "epoch": 0.4420584783640041,
+      "grad_norm": 0.0963086411356926,
+      "learning_rate": 1.6780144684472704e-05,
+      "loss": 9.0095,
+      "step": 88520
+    },
+    {
+      "epoch": 0.44210841718894356,
+      "grad_norm": 0.09583403170108795,
+      "learning_rate": 1.677864276953115e-05,
+      "loss": 9.0238,
+      "step": 88530
+    },
+    {
+      "epoch": 0.442158356013883,
+      "grad_norm": 0.09216772019863129,
+      "learning_rate": 1.6777140854589604e-05,
+      "loss": 9.0123,
+      "step": 88540
+    },
+    {
+      "epoch": 0.44220829483882246,
+      "grad_norm": 0.09259237349033356,
+      "learning_rate": 1.677563893964805e-05,
+      "loss": 9.024,
+      "step": 88550
+    },
+    {
+      "epoch": 0.4422582336637619,
+      "grad_norm": 0.09866563975811005,
+      "learning_rate": 1.6774137024706498e-05,
+      "loss": 9.0138,
+      "step": 88560
+    },
+    {
+      "epoch": 0.44230817248870136,
+      "grad_norm": 0.09354985505342484,
+      "learning_rate": 1.677263510976495e-05,
+      "loss": 9.026,
+      "step": 88570
+    },
+    {
+      "epoch": 0.4423581113136408,
+      "grad_norm": 0.10093557089567184,
+      "learning_rate": 1.67711331948234e-05,
+      "loss": 9.0229,
+      "step": 88580
+    },
+    {
+      "epoch": 0.44240805013858026,
+      "grad_norm": 0.09272787719964981,
+      "learning_rate": 1.6769631279881852e-05,
+      "loss": 9.0249,
+      "step": 88590
+    },
+    {
+      "epoch": 0.4424579889635197,
+      "grad_norm": 0.09445803612470627,
+      "learning_rate": 1.67681293649403e-05,
+      "loss": 9.0102,
+      "step": 88600
+    },
+    {
+      "epoch": 0.44250792778845915,
+      "grad_norm": 0.09491786360740662,
+      "learning_rate": 1.676662744999875e-05,
+      "loss": 9.0247,
+      "step": 88610
+    },
+    {
+      "epoch": 0.4425578666133986,
+      "grad_norm": 0.09124266356229782,
+      "learning_rate": 1.67651255350572e-05,
+      "loss": 9.0244,
+      "step": 88620
+    },
+    {
+      "epoch": 0.44260780543833805,
+      "grad_norm": 0.09176667034626007,
+      "learning_rate": 1.6763623620115646e-05,
+      "loss": 9.0165,
+      "step": 88630
+    },
+    {
+      "epoch": 0.4426577442632775,
+      "grad_norm": 0.10504433512687683,
+      "learning_rate": 1.67621217051741e-05,
+      "loss": 9.0253,
+      "step": 88640
+    },
+    {
+      "epoch": 0.44270768308821695,
+      "grad_norm": 0.09985138475894928,
+      "learning_rate": 1.6760619790232546e-05,
+      "loss": 9.0125,
+      "step": 88650
+    },
+    {
+      "epoch": 0.4427576219131564,
+      "grad_norm": 0.0883818119764328,
+      "learning_rate": 1.6759117875290996e-05,
+      "loss": 9.0167,
+      "step": 88660
+    },
+    {
+      "epoch": 0.44280756073809585,
+      "grad_norm": 0.08730896562337875,
+      "learning_rate": 1.6757615960349447e-05,
+      "loss": 9.033,
+      "step": 88670
+    },
+    {
+      "epoch": 0.4428574995630353,
+      "grad_norm": 0.08953774720430374,
+      "learning_rate": 1.6756114045407893e-05,
+      "loss": 9.0397,
+      "step": 88680
+    },
+    {
+      "epoch": 0.44290743838797475,
+      "grad_norm": 0.09112643450498581,
+      "learning_rate": 1.6754612130466347e-05,
+      "loss": 9.0137,
+      "step": 88690
+    },
+    {
+      "epoch": 0.44295737721291417,
+      "grad_norm": 0.09895043820142746,
+      "learning_rate": 1.6753110215524794e-05,
+      "loss": 9.0121,
+      "step": 88700
+    },
+    {
+      "epoch": 0.44300731603785365,
+      "grad_norm": 0.09690089523792267,
+      "learning_rate": 1.6751608300583244e-05,
+      "loss": 9.0322,
+      "step": 88710
+    },
+    {
+      "epoch": 0.44305725486279307,
+      "grad_norm": 0.09116695076227188,
+      "learning_rate": 1.6750106385641694e-05,
+      "loss": 9.0266,
+      "step": 88720
+    },
+    {
+      "epoch": 0.44310719368773255,
+      "grad_norm": 0.09873844683170319,
+      "learning_rate": 1.674860447070014e-05,
+      "loss": 9.0144,
+      "step": 88730
+    },
+    {
+      "epoch": 0.44315713251267197,
+      "grad_norm": 0.08854333311319351,
+      "learning_rate": 1.6747102555758594e-05,
+      "loss": 9.0264,
+      "step": 88740
+    },
+    {
+      "epoch": 0.44320707133761145,
+      "grad_norm": 0.09332463145256042,
+      "learning_rate": 1.674560064081704e-05,
+      "loss": 9.0221,
+      "step": 88750
+    },
+    {
+      "epoch": 0.44325701016255087,
+      "grad_norm": 0.0980052724480629,
+      "learning_rate": 1.674409872587549e-05,
+      "loss": 9.0311,
+      "step": 88760
+    },
+    {
+      "epoch": 0.44330694898749035,
+      "grad_norm": 0.09149298816919327,
+      "learning_rate": 1.674259681093394e-05,
+      "loss": 9.0343,
+      "step": 88770
+    },
+    {
+      "epoch": 0.44335688781242977,
+      "grad_norm": 0.0906062051653862,
+      "learning_rate": 1.674109489599239e-05,
+      "loss": 9.0131,
+      "step": 88780
+    },
+    {
+      "epoch": 0.44340682663736924,
+      "grad_norm": 0.09233875572681427,
+      "learning_rate": 1.6739592981050842e-05,
+      "loss": 9.0148,
+      "step": 88790
+    },
+    {
+      "epoch": 0.44345676546230867,
+      "grad_norm": 0.09254756569862366,
+      "learning_rate": 1.673809106610929e-05,
+      "loss": 9.0121,
+      "step": 88800
+    },
+    {
+      "epoch": 0.44350670428724814,
+      "grad_norm": 0.0948517769575119,
+      "learning_rate": 1.673658915116774e-05,
+      "loss": 9.0205,
+      "step": 88810
+    },
+    {
+      "epoch": 0.44355664311218757,
+      "grad_norm": 0.0976976752281189,
+      "learning_rate": 1.673508723622619e-05,
+      "loss": 9.0087,
+      "step": 88820
+    },
+    {
+      "epoch": 0.44360658193712704,
+      "grad_norm": 0.0981476828455925,
+      "learning_rate": 1.6733585321284636e-05,
+      "loss": 9.0068,
+      "step": 88830
+    },
+    {
+      "epoch": 0.44365652076206646,
+      "grad_norm": 0.09623267501592636,
+      "learning_rate": 1.673208340634309e-05,
+      "loss": 9.0195,
+      "step": 88840
+    },
+    {
+      "epoch": 0.44370645958700594,
+      "grad_norm": 0.09452231973409653,
+      "learning_rate": 1.6730581491401536e-05,
+      "loss": 9.016,
+      "step": 88850
+    },
+    {
+      "epoch": 0.44375639841194536,
+      "grad_norm": 0.09390220791101456,
+      "learning_rate": 1.6729079576459986e-05,
+      "loss": 9.0134,
+      "step": 88860
+    },
+    {
+      "epoch": 0.44380633723688484,
+      "grad_norm": 0.09678220003843307,
+      "learning_rate": 1.6727577661518437e-05,
+      "loss": 9.012,
+      "step": 88870
+    },
+    {
+      "epoch": 0.44385627606182426,
+      "grad_norm": 0.09419732540845871,
+      "learning_rate": 1.6726075746576883e-05,
+      "loss": 9.0304,
+      "step": 88880
+    },
+    {
+      "epoch": 0.44390621488676374,
+      "grad_norm": 0.08998965471982956,
+      "learning_rate": 1.6724573831635337e-05,
+      "loss": 9.0247,
+      "step": 88890
+    },
+    {
+      "epoch": 0.44395615371170316,
+      "grad_norm": 0.09407693147659302,
+      "learning_rate": 1.6723071916693784e-05,
+      "loss": 9.0233,
+      "step": 88900
+    },
+    {
+      "epoch": 0.44400609253664264,
+      "grad_norm": 0.094719298183918,
+      "learning_rate": 1.6721570001752234e-05,
+      "loss": 9.0212,
+      "step": 88910
+    },
+    {
+      "epoch": 0.44405603136158206,
+      "grad_norm": 0.09490375965833664,
+      "learning_rate": 1.6720068086810684e-05,
+      "loss": 9.0231,
+      "step": 88920
+    },
+    {
+      "epoch": 0.44410597018652154,
+      "grad_norm": 0.09172754734754562,
+      "learning_rate": 1.6718566171869134e-05,
+      "loss": 9.0039,
+      "step": 88930
+    },
+    {
+      "epoch": 0.44415590901146096,
+      "grad_norm": 0.09175007045269012,
+      "learning_rate": 1.6717064256927584e-05,
+      "loss": 9.0265,
+      "step": 88940
+    },
+    {
+      "epoch": 0.44420584783640044,
+      "grad_norm": 0.0903623029589653,
+      "learning_rate": 1.671556234198603e-05,
+      "loss": 9.0105,
+      "step": 88950
+    },
+    {
+      "epoch": 0.44425578666133986,
+      "grad_norm": 0.09337350726127625,
+      "learning_rate": 1.671406042704448e-05,
+      "loss": 9.0146,
+      "step": 88960
+    },
+    {
+      "epoch": 0.44430572548627933,
+      "grad_norm": 0.0933094471693039,
+      "learning_rate": 1.671255851210293e-05,
+      "loss": 9.0125,
+      "step": 88970
+    },
+    {
+      "epoch": 0.44435566431121876,
+      "grad_norm": 0.0940171629190445,
+      "learning_rate": 1.6711056597161382e-05,
+      "loss": 9.0088,
+      "step": 88980
+    },
+    {
+      "epoch": 0.44440560313615823,
+      "grad_norm": 0.0906391590833664,
+      "learning_rate": 1.6709554682219832e-05,
+      "loss": 9.0209,
+      "step": 88990
+    },
+    {
+      "epoch": 0.44445554196109766,
+      "grad_norm": 0.09381049126386642,
+      "learning_rate": 1.670805276727828e-05,
+      "loss": 9.0177,
+      "step": 89000
+    },
+    {
+      "epoch": 0.4445054807860371,
+      "grad_norm": 0.09514244645833969,
+      "learning_rate": 1.670655085233673e-05,
+      "loss": 9.0153,
+      "step": 89010
+    },
+    {
+      "epoch": 0.44455541961097655,
+      "grad_norm": 0.09447330236434937,
+      "learning_rate": 1.670504893739518e-05,
+      "loss": 9.0064,
+      "step": 89020
+    },
+    {
+      "epoch": 0.444605358435916,
+      "grad_norm": 0.09046784043312073,
+      "learning_rate": 1.670354702245363e-05,
+      "loss": 9.011,
+      "step": 89030
+    },
+    {
+      "epoch": 0.44465529726085545,
+      "grad_norm": 0.09899982064962387,
+      "learning_rate": 1.670204510751208e-05,
+      "loss": 9.0164,
+      "step": 89040
+    },
+    {
+      "epoch": 0.4447052360857949,
+      "grad_norm": 0.0965995118021965,
+      "learning_rate": 1.6700543192570526e-05,
+      "loss": 9.0145,
+      "step": 89050
+    },
+    {
+      "epoch": 0.44475517491073435,
+      "grad_norm": 0.09488395601511002,
+      "learning_rate": 1.6699041277628976e-05,
+      "loss": 9.0108,
+      "step": 89060
+    },
+    {
+      "epoch": 0.4448051137356738,
+      "grad_norm": 0.09230642020702362,
+      "learning_rate": 1.6697539362687427e-05,
+      "loss": 9.0086,
+      "step": 89070
+    },
+    {
+      "epoch": 0.44485505256061325,
+      "grad_norm": 0.10005513578653336,
+      "learning_rate": 1.6696037447745877e-05,
+      "loss": 9.0129,
+      "step": 89080
+    },
+    {
+      "epoch": 0.4449049913855527,
+      "grad_norm": 0.09370043128728867,
+      "learning_rate": 1.6694535532804327e-05,
+      "loss": 9.0219,
+      "step": 89090
+    },
+    {
+      "epoch": 0.44495493021049215,
+      "grad_norm": 0.09334039688110352,
+      "learning_rate": 1.6693033617862774e-05,
+      "loss": 9.0091,
+      "step": 89100
+    },
+    {
+      "epoch": 0.44500486903543157,
+      "grad_norm": 0.09181877970695496,
+      "learning_rate": 1.6691531702921224e-05,
+      "loss": 9.0125,
+      "step": 89110
+    },
+    {
+      "epoch": 0.44505480786037105,
+      "grad_norm": 0.09253057092428207,
+      "learning_rate": 1.6690029787979674e-05,
+      "loss": 9.0284,
+      "step": 89120
+    },
+    {
+      "epoch": 0.44510474668531047,
+      "grad_norm": 0.08906669169664383,
+      "learning_rate": 1.6688527873038124e-05,
+      "loss": 9.0251,
+      "step": 89130
+    },
+    {
+      "epoch": 0.44515468551024995,
+      "grad_norm": 0.09391651302576065,
+      "learning_rate": 1.6687025958096574e-05,
+      "loss": 9.0138,
+      "step": 89140
+    },
+    {
+      "epoch": 0.44520462433518937,
+      "grad_norm": 0.09748826175928116,
+      "learning_rate": 1.668552404315502e-05,
+      "loss": 9.0223,
+      "step": 89150
+    },
+    {
+      "epoch": 0.44525456316012885,
+      "grad_norm": 0.09539515525102615,
+      "learning_rate": 1.668402212821347e-05,
+      "loss": 8.9965,
+      "step": 89160
+    },
+    {
+      "epoch": 0.44530450198506827,
+      "grad_norm": 0.0892728641629219,
+      "learning_rate": 1.668252021327192e-05,
+      "loss": 9.0155,
+      "step": 89170
+    },
+    {
+      "epoch": 0.44535444081000775,
+      "grad_norm": 0.09498701989650726,
+      "learning_rate": 1.6681018298330372e-05,
+      "loss": 9.0113,
+      "step": 89180
+    },
+    {
+      "epoch": 0.44540437963494717,
+      "grad_norm": 0.09873063117265701,
+      "learning_rate": 1.6679516383388822e-05,
+      "loss": 9.0206,
+      "step": 89190
+    },
+    {
+      "epoch": 0.44545431845988664,
+      "grad_norm": 0.093873530626297,
+      "learning_rate": 1.667801446844727e-05,
+      "loss": 9.0141,
+      "step": 89200
+    },
+    {
+      "epoch": 0.44550425728482607,
+      "grad_norm": 0.09323079884052277,
+      "learning_rate": 1.667651255350572e-05,
+      "loss": 9.0133,
+      "step": 89210
+    },
+    {
+      "epoch": 0.44555419610976554,
+      "grad_norm": 0.09667900949716568,
+      "learning_rate": 1.667501063856417e-05,
+      "loss": 9.0206,
+      "step": 89220
+    },
+    {
+      "epoch": 0.44560413493470497,
+      "grad_norm": 0.1016487404704094,
+      "learning_rate": 1.667350872362262e-05,
+      "loss": 9.036,
+      "step": 89230
+    },
+    {
+      "epoch": 0.44565407375964444,
+      "grad_norm": 0.09627226740121841,
+      "learning_rate": 1.667200680868107e-05,
+      "loss": 9.0168,
+      "step": 89240
+    },
+    {
+      "epoch": 0.44570401258458386,
+      "grad_norm": 0.09360161423683167,
+      "learning_rate": 1.667050489373952e-05,
+      "loss": 9.0101,
+      "step": 89250
+    },
+    {
+      "epoch": 0.44575395140952334,
+      "grad_norm": 0.08747625350952148,
+      "learning_rate": 1.6669002978797966e-05,
+      "loss": 9.0054,
+      "step": 89260
+    },
+    {
+      "epoch": 0.44580389023446276,
+      "grad_norm": 0.09256748855113983,
+      "learning_rate": 1.6667501063856417e-05,
+      "loss": 9.0228,
+      "step": 89270
+    },
+    {
+      "epoch": 0.44585382905940224,
+      "grad_norm": 0.09370601922273636,
+      "learning_rate": 1.6665999148914867e-05,
+      "loss": 9.0211,
+      "step": 89280
+    },
+    {
+      "epoch": 0.44590376788434166,
+      "grad_norm": 0.09242367744445801,
+      "learning_rate": 1.6664497233973317e-05,
+      "loss": 9.0083,
+      "step": 89290
+    },
+    {
+      "epoch": 0.44595370670928114,
+      "grad_norm": 0.0926143079996109,
+      "learning_rate": 1.6662995319031767e-05,
+      "loss": 9.0202,
+      "step": 89300
+    },
+    {
+      "epoch": 0.44600364553422056,
+      "grad_norm": 0.0988818109035492,
+      "learning_rate": 1.6661493404090214e-05,
+      "loss": 9.0048,
+      "step": 89310
+    },
+    {
+      "epoch": 0.44605358435916004,
+      "grad_norm": 0.09795771539211273,
+      "learning_rate": 1.6659991489148664e-05,
+      "loss": 9.0216,
+      "step": 89320
+    },
+    {
+      "epoch": 0.44610352318409946,
+      "grad_norm": 0.09289045631885529,
+      "learning_rate": 1.6658489574207114e-05,
+      "loss": 9.0187,
+      "step": 89330
+    },
+    {
+      "epoch": 0.44615346200903894,
+      "grad_norm": 0.09757115691900253,
+      "learning_rate": 1.6656987659265564e-05,
+      "loss": 9.0029,
+      "step": 89340
+    },
+    {
+      "epoch": 0.44620340083397836,
+      "grad_norm": 0.09858351945877075,
+      "learning_rate": 1.6655485744324015e-05,
+      "loss": 9.0203,
+      "step": 89350
+    },
+    {
+      "epoch": 0.44625333965891784,
+      "grad_norm": 0.09712626785039902,
+      "learning_rate": 1.665398382938246e-05,
+      "loss": 9.0095,
+      "step": 89360
+    },
+    {
+      "epoch": 0.44630327848385726,
+      "grad_norm": 0.09400411695241928,
+      "learning_rate": 1.665248191444091e-05,
+      "loss": 9.007,
+      "step": 89370
+    },
+    {
+      "epoch": 0.44635321730879673,
+      "grad_norm": 0.0991569310426712,
+      "learning_rate": 1.6650979999499362e-05,
+      "loss": 9.0118,
+      "step": 89380
+    },
+    {
+      "epoch": 0.44640315613373616,
+      "grad_norm": 0.09559716284275055,
+      "learning_rate": 1.6649478084557812e-05,
+      "loss": 8.9983,
+      "step": 89390
+    },
+    {
+      "epoch": 0.44645309495867563,
+      "grad_norm": 0.09568917006254196,
+      "learning_rate": 1.6647976169616262e-05,
+      "loss": 9.0076,
+      "step": 89400
+    },
+    {
+      "epoch": 0.44650303378361506,
+      "grad_norm": 0.09183740615844727,
+      "learning_rate": 1.664647425467471e-05,
+      "loss": 9.0106,
+      "step": 89410
+    },
+    {
+      "epoch": 0.44655297260855453,
+      "grad_norm": 0.0893789678812027,
+      "learning_rate": 1.664497233973316e-05,
+      "loss": 9.0001,
+      "step": 89420
+    },
+    {
+      "epoch": 0.44660291143349395,
+      "grad_norm": 0.09366993606090546,
+      "learning_rate": 1.664347042479161e-05,
+      "loss": 9.0137,
+      "step": 89430
+    },
+    {
+      "epoch": 0.44665285025843343,
+      "grad_norm": 0.09987689554691315,
+      "learning_rate": 1.664196850985006e-05,
+      "loss": 9.0049,
+      "step": 89440
+    },
+    {
+      "epoch": 0.44670278908337285,
+      "grad_norm": 0.09278554469347,
+      "learning_rate": 1.664046659490851e-05,
+      "loss": 9.0149,
+      "step": 89450
+    },
+    {
+      "epoch": 0.44675272790831233,
+      "grad_norm": 0.08935824036598206,
+      "learning_rate": 1.6638964679966956e-05,
+      "loss": 9.0228,
+      "step": 89460
+    },
+    {
+      "epoch": 0.44680266673325175,
+      "grad_norm": 0.08843576163053513,
+      "learning_rate": 1.6637462765025407e-05,
+      "loss": 9.0136,
+      "step": 89470
+    },
+    {
+      "epoch": 0.44685260555819123,
+      "grad_norm": 0.09229853004217148,
+      "learning_rate": 1.6635960850083857e-05,
+      "loss": 9.0065,
+      "step": 89480
+    },
+    {
+      "epoch": 0.44690254438313065,
+      "grad_norm": 0.09317019581794739,
+      "learning_rate": 1.6634458935142307e-05,
+      "loss": 9.0074,
+      "step": 89490
+    },
+    {
+      "epoch": 0.44695248320807013,
+      "grad_norm": 0.09514672309160233,
+      "learning_rate": 1.6632957020200757e-05,
+      "loss": 9.0009,
+      "step": 89500
+    },
+    {
+      "epoch": 0.44700242203300955,
+      "grad_norm": 0.09481526166200638,
+      "learning_rate": 1.6631455105259207e-05,
+      "loss": 9.0216,
+      "step": 89510
+    },
+    {
+      "epoch": 0.447052360857949,
+      "grad_norm": 0.09262920171022415,
+      "learning_rate": 1.6629953190317654e-05,
+      "loss": 9.0146,
+      "step": 89520
+    },
+    {
+      "epoch": 0.44710229968288845,
+      "grad_norm": 0.09552852809429169,
+      "learning_rate": 1.6628451275376104e-05,
+      "loss": 9.0086,
+      "step": 89530
+    },
+    {
+      "epoch": 0.4471522385078279,
+      "grad_norm": 0.09410271048545837,
+      "learning_rate": 1.6626949360434554e-05,
+      "loss": 9.0166,
+      "step": 89540
+    },
+    {
+      "epoch": 0.44720217733276735,
+      "grad_norm": 0.09333918243646622,
+      "learning_rate": 1.6625447445493005e-05,
+      "loss": 9.0171,
+      "step": 89550
+    },
+    {
+      "epoch": 0.4472521161577068,
+      "grad_norm": 0.09849578142166138,
+      "learning_rate": 1.6623945530551455e-05,
+      "loss": 9.0056,
+      "step": 89560
+    },
+    {
+      "epoch": 0.44730205498264625,
+      "grad_norm": 0.09501996636390686,
+      "learning_rate": 1.6622443615609905e-05,
+      "loss": 9.0066,
+      "step": 89570
+    },
+    {
+      "epoch": 0.4473519938075857,
+      "grad_norm": 0.09807047247886658,
+      "learning_rate": 1.6620941700668352e-05,
+      "loss": 9.0202,
+      "step": 89580
+    },
+    {
+      "epoch": 0.44740193263252515,
+      "grad_norm": 0.09879709780216217,
+      "learning_rate": 1.6619439785726802e-05,
+      "loss": 9.0203,
+      "step": 89590
+    },
+    {
+      "epoch": 0.4474518714574646,
+      "grad_norm": 0.08965755254030228,
+      "learning_rate": 1.6617937870785252e-05,
+      "loss": 9.0089,
+      "step": 89600
+    },
+    {
+      "epoch": 0.44750181028240404,
+      "grad_norm": 0.09854756295681,
+      "learning_rate": 1.6616435955843702e-05,
+      "loss": 9.0021,
+      "step": 89610
+    },
+    {
+      "epoch": 0.4475517491073435,
+      "grad_norm": 0.09752686321735382,
+      "learning_rate": 1.6614934040902153e-05,
+      "loss": 9.0026,
+      "step": 89620
+    },
+    {
+      "epoch": 0.44760168793228294,
+      "grad_norm": 0.09632866829633713,
+      "learning_rate": 1.66134321259606e-05,
+      "loss": 9.0084,
+      "step": 89630
+    },
+    {
+      "epoch": 0.4476516267572224,
+      "grad_norm": 0.09156852960586548,
+      "learning_rate": 1.661193021101905e-05,
+      "loss": 9.0133,
+      "step": 89640
+    },
+    {
+      "epoch": 0.44770156558216184,
+      "grad_norm": 0.08667552471160889,
+      "learning_rate": 1.66104282960775e-05,
+      "loss": 9.0044,
+      "step": 89650
+    },
+    {
+      "epoch": 0.4477515044071013,
+      "grad_norm": 0.09284383803606033,
+      "learning_rate": 1.660892638113595e-05,
+      "loss": 9.0035,
+      "step": 89660
+    },
+    {
+      "epoch": 0.44780144323204074,
+      "grad_norm": 0.09126190096139908,
+      "learning_rate": 1.66074244661944e-05,
+      "loss": 9.014,
+      "step": 89670
+    },
+    {
+      "epoch": 0.4478513820569802,
+      "grad_norm": 0.10024043917655945,
+      "learning_rate": 1.6605922551252847e-05,
+      "loss": 8.9948,
+      "step": 89680
+    },
+    {
+      "epoch": 0.44790132088191964,
+      "grad_norm": 0.09772263467311859,
+      "learning_rate": 1.6604420636311297e-05,
+      "loss": 9.0078,
+      "step": 89690
+    },
+    {
+      "epoch": 0.4479512597068591,
+      "grad_norm": 0.09419326484203339,
+      "learning_rate": 1.6602918721369747e-05,
+      "loss": 9.0066,
+      "step": 89700
+    },
+    {
+      "epoch": 0.44800119853179854,
+      "grad_norm": 0.08692155033349991,
+      "learning_rate": 1.6601416806428197e-05,
+      "loss": 9.0163,
+      "step": 89710
+    },
+    {
+      "epoch": 0.448051137356738,
+      "grad_norm": 0.09567783027887344,
+      "learning_rate": 1.6599914891486648e-05,
+      "loss": 9.0016,
+      "step": 89720
+    },
+    {
+      "epoch": 0.44810107618167744,
+      "grad_norm": 0.0902228057384491,
+      "learning_rate": 1.6598412976545094e-05,
+      "loss": 9.0098,
+      "step": 89730
+    },
+    {
+      "epoch": 0.4481510150066169,
+      "grad_norm": 0.09784096479415894,
+      "learning_rate": 1.6596911061603545e-05,
+      "loss": 9.0028,
+      "step": 89740
+    },
+    {
+      "epoch": 0.44820095383155634,
+      "grad_norm": 0.09646725654602051,
+      "learning_rate": 1.6595409146661995e-05,
+      "loss": 8.9891,
+      "step": 89750
+    },
+    {
+      "epoch": 0.4482508926564958,
+      "grad_norm": 0.09333818405866623,
+      "learning_rate": 1.6593907231720445e-05,
+      "loss": 8.9968,
+      "step": 89760
+    },
+    {
+      "epoch": 0.44830083148143524,
+      "grad_norm": 0.09000226110219955,
+      "learning_rate": 1.6592405316778895e-05,
+      "loss": 9.0099,
+      "step": 89770
+    },
+    {
+      "epoch": 0.4483507703063747,
+      "grad_norm": 0.09583406150341034,
+      "learning_rate": 1.6590903401837342e-05,
+      "loss": 9.0108,
+      "step": 89780
+    },
+    {
+      "epoch": 0.44840070913131413,
+      "grad_norm": 0.09547807276248932,
+      "learning_rate": 1.6589401486895792e-05,
+      "loss": 9.0058,
+      "step": 89790
+    },
+    {
+      "epoch": 0.4484506479562536,
+      "grad_norm": 0.09241116046905518,
+      "learning_rate": 1.6587899571954242e-05,
+      "loss": 8.9989,
+      "step": 89800
+    },
+    {
+      "epoch": 0.44850058678119303,
+      "grad_norm": 0.09252715855836868,
+      "learning_rate": 1.6586397657012692e-05,
+      "loss": 9.0156,
+      "step": 89810
+    },
+    {
+      "epoch": 0.4485505256061325,
+      "grad_norm": 0.09530036151409149,
+      "learning_rate": 1.6584895742071143e-05,
+      "loss": 9.0116,
+      "step": 89820
+    },
+    {
+      "epoch": 0.44860046443107193,
+      "grad_norm": 0.08951985090970993,
+      "learning_rate": 1.658339382712959e-05,
+      "loss": 9.0116,
+      "step": 89830
+    },
+    {
+      "epoch": 0.4486504032560114,
+      "grad_norm": 0.09694526344537735,
+      "learning_rate": 1.658189191218804e-05,
+      "loss": 8.9973,
+      "step": 89840
+    },
+    {
+      "epoch": 0.44870034208095083,
+      "grad_norm": 0.09165141731500626,
+      "learning_rate": 1.658038999724649e-05,
+      "loss": 9.0058,
+      "step": 89850
+    },
+    {
+      "epoch": 0.4487502809058903,
+      "grad_norm": 0.09192651510238647,
+      "learning_rate": 1.657888808230494e-05,
+      "loss": 9.0044,
+      "step": 89860
+    },
+    {
+      "epoch": 0.44880021973082973,
+      "grad_norm": 0.08684073388576508,
+      "learning_rate": 1.657738616736339e-05,
+      "loss": 9.0147,
+      "step": 89870
+    },
+    {
+      "epoch": 0.4488501585557692,
+      "grad_norm": 0.09026368707418442,
+      "learning_rate": 1.6575884252421837e-05,
+      "loss": 8.9914,
+      "step": 89880
+    },
+    {
+      "epoch": 0.44890009738070863,
+      "grad_norm": 0.09688662737607956,
+      "learning_rate": 1.657438233748029e-05,
+      "loss": 9.0052,
+      "step": 89890
+    },
+    {
+      "epoch": 0.4489500362056481,
+      "grad_norm": 0.09581411629915237,
+      "learning_rate": 1.6572880422538737e-05,
+      "loss": 8.9905,
+      "step": 89900
+    },
+    {
+      "epoch": 0.44899997503058753,
+      "grad_norm": 0.1003839373588562,
+      "learning_rate": 1.6571378507597187e-05,
+      "loss": 9.0148,
+      "step": 89910
+    },
+    {
+      "epoch": 0.449049913855527,
+      "grad_norm": 0.09192468225955963,
+      "learning_rate": 1.6569876592655638e-05,
+      "loss": 9.0248,
+      "step": 89920
+    },
+    {
+      "epoch": 0.4490998526804664,
+      "grad_norm": 0.09126332402229309,
+      "learning_rate": 1.6568374677714084e-05,
+      "loss": 9.0122,
+      "step": 89930
+    },
+    {
+      "epoch": 0.4491497915054059,
+      "grad_norm": 0.0944957584142685,
+      "learning_rate": 1.6566872762772538e-05,
+      "loss": 9.0082,
+      "step": 89940
+    },
+    {
+      "epoch": 0.4491997303303453,
+      "grad_norm": 0.0930357500910759,
+      "learning_rate": 1.6565370847830985e-05,
+      "loss": 8.9977,
+      "step": 89950
+    },
+    {
+      "epoch": 0.4492496691552848,
+      "grad_norm": 0.09657001495361328,
+      "learning_rate": 1.6563868932889435e-05,
+      "loss": 9.006,
+      "step": 89960
+    },
+    {
+      "epoch": 0.4492996079802242,
+      "grad_norm": 0.09886710345745087,
+      "learning_rate": 1.6562367017947885e-05,
+      "loss": 9.0056,
+      "step": 89970
+    },
+    {
+      "epoch": 0.4493495468051637,
+      "grad_norm": 0.0890527293086052,
+      "learning_rate": 1.6560865103006332e-05,
+      "loss": 9.0023,
+      "step": 89980
+    },
+    {
+      "epoch": 0.4493994856301031,
+      "grad_norm": 0.0896250382065773,
+      "learning_rate": 1.6559363188064785e-05,
+      "loss": 9.0064,
+      "step": 89990
+    },
+    {
+      "epoch": 0.44944942445504255,
+      "grad_norm": 0.09379280358552933,
+      "learning_rate": 1.6557861273123232e-05,
+      "loss": 9.0103,
+      "step": 90000
+    },
+    {
+      "epoch": 0.449499363279982,
+      "grad_norm": 0.09411850571632385,
+      "learning_rate": 1.6556359358181682e-05,
+      "loss": 9.0098,
+      "step": 90010
+    },
+    {
+      "epoch": 0.44954930210492144,
+      "grad_norm": 0.08976029604673386,
+      "learning_rate": 1.6554857443240133e-05,
+      "loss": 9.0026,
+      "step": 90020
+    },
+    {
+      "epoch": 0.4495992409298609,
+      "grad_norm": 0.0921340063214302,
+      "learning_rate": 1.655335552829858e-05,
+      "loss": 9.0126,
+      "step": 90030
+    },
+    {
+      "epoch": 0.44964917975480034,
+      "grad_norm": 0.09887304157018661,
+      "learning_rate": 1.6551853613357033e-05,
+      "loss": 9.0057,
+      "step": 90040
+    },
+    {
+      "epoch": 0.4496991185797398,
+      "grad_norm": 0.09693904966115952,
+      "learning_rate": 1.655035169841548e-05,
+      "loss": 8.9959,
+      "step": 90050
+    },
+    {
+      "epoch": 0.44974905740467924,
+      "grad_norm": 0.09151757508516312,
+      "learning_rate": 1.654884978347393e-05,
+      "loss": 9.0101,
+      "step": 90060
+    },
+    {
+      "epoch": 0.4497989962296187,
+      "grad_norm": 0.1048135980963707,
+      "learning_rate": 1.654734786853238e-05,
+      "loss": 8.9994,
+      "step": 90070
+    },
+    {
+      "epoch": 0.44984893505455814,
+      "grad_norm": 0.09414626657962799,
+      "learning_rate": 1.6545845953590827e-05,
+      "loss": 9.0041,
+      "step": 90080
+    },
+    {
+      "epoch": 0.4498988738794976,
+      "grad_norm": 0.0916576012969017,
+      "learning_rate": 1.654434403864928e-05,
+      "loss": 8.9947,
+      "step": 90090
+    },
+    {
+      "epoch": 0.44994881270443704,
+      "grad_norm": 0.09090348333120346,
+      "learning_rate": 1.6542842123707727e-05,
+      "loss": 8.9906,
+      "step": 90100
+    },
+    {
+      "epoch": 0.4499987515293765,
+      "grad_norm": 0.0916331559419632,
+      "learning_rate": 1.6541340208766177e-05,
+      "loss": 9.0048,
+      "step": 90110
+    },
+    {
+      "epoch": 0.45004869035431594,
+      "grad_norm": 0.09484575688838959,
+      "learning_rate": 1.6539838293824628e-05,
+      "loss": 9.0064,
+      "step": 90120
+    },
+    {
+      "epoch": 0.4500986291792554,
+      "grad_norm": 0.09016244858503342,
+      "learning_rate": 1.6538336378883074e-05,
+      "loss": 9.0041,
+      "step": 90130
+    },
+    {
+      "epoch": 0.45014856800419484,
+      "grad_norm": 0.09121375530958176,
+      "learning_rate": 1.6536834463941528e-05,
+      "loss": 9.0068,
+      "step": 90140
+    },
+    {
+      "epoch": 0.4501985068291343,
+      "grad_norm": 0.09483055770397186,
+      "learning_rate": 1.6535332548999975e-05,
+      "loss": 9.0051,
+      "step": 90150
+    },
+    {
+      "epoch": 0.45024844565407374,
+      "grad_norm": 0.09125950932502747,
+      "learning_rate": 1.6533830634058425e-05,
+      "loss": 8.9956,
+      "step": 90160
+    },
+    {
+      "epoch": 0.4502983844790132,
+      "grad_norm": 0.09234707057476044,
+      "learning_rate": 1.6532328719116875e-05,
+      "loss": 9.0004,
+      "step": 90170
+    },
+    {
+      "epoch": 0.45034832330395264,
+      "grad_norm": 0.09413158893585205,
+      "learning_rate": 1.6530826804175322e-05,
+      "loss": 8.9987,
+      "step": 90180
+    },
+    {
+      "epoch": 0.4503982621288921,
+      "grad_norm": 0.09710296243429184,
+      "learning_rate": 1.6529324889233775e-05,
+      "loss": 8.9944,
+      "step": 90190
+    },
+    {
+      "epoch": 0.45044820095383153,
+      "grad_norm": 0.0949583351612091,
+      "learning_rate": 1.6527822974292222e-05,
+      "loss": 9.003,
+      "step": 90200
+    },
+    {
+      "epoch": 0.450498139778771,
+      "grad_norm": 0.09892937541007996,
+      "learning_rate": 1.6526321059350676e-05,
+      "loss": 8.9983,
+      "step": 90210
+    },
+    {
+      "epoch": 0.45054807860371043,
+      "grad_norm": 0.09644380211830139,
+      "learning_rate": 1.6524819144409123e-05,
+      "loss": 8.9919,
+      "step": 90220
+    },
+    {
+      "epoch": 0.4505980174286499,
+      "grad_norm": 0.09606988728046417,
+      "learning_rate": 1.652331722946757e-05,
+      "loss": 9.0063,
+      "step": 90230
+    },
+    {
+      "epoch": 0.45064795625358933,
+      "grad_norm": 0.09773648530244827,
+      "learning_rate": 1.6521815314526023e-05,
+      "loss": 8.9835,
+      "step": 90240
+    },
+    {
+      "epoch": 0.4506978950785288,
+      "grad_norm": 0.09979324787855148,
+      "learning_rate": 1.652031339958447e-05,
+      "loss": 8.9963,
+      "step": 90250
+    },
+    {
+      "epoch": 0.45074783390346823,
+      "grad_norm": 0.09351041167974472,
+      "learning_rate": 1.6518811484642923e-05,
+      "loss": 9.0103,
+      "step": 90260
+    },
+    {
+      "epoch": 0.4507977727284077,
+      "grad_norm": 0.09703806042671204,
+      "learning_rate": 1.651730956970137e-05,
+      "loss": 8.9954,
+      "step": 90270
+    },
+    {
+      "epoch": 0.45084771155334713,
+      "grad_norm": 0.09339191764593124,
+      "learning_rate": 1.6515807654759817e-05,
+      "loss": 9.0095,
+      "step": 90280
+    },
+    {
+      "epoch": 0.4508976503782866,
+      "grad_norm": 0.09220651537179947,
+      "learning_rate": 1.651430573981827e-05,
+      "loss": 8.9959,
+      "step": 90290
+    },
+    {
+      "epoch": 0.45094758920322603,
+      "grad_norm": 0.09635385125875473,
+      "learning_rate": 1.6512803824876717e-05,
+      "loss": 8.9845,
+      "step": 90300
+    },
+    {
+      "epoch": 0.4509975280281655,
+      "grad_norm": 0.08985468745231628,
+      "learning_rate": 1.651130190993517e-05,
+      "loss": 9.0037,
+      "step": 90310
+    },
+    {
+      "epoch": 0.45104746685310493,
+      "grad_norm": 0.0972968265414238,
+      "learning_rate": 1.6509799994993618e-05,
+      "loss": 8.9896,
+      "step": 90320
+    },
+    {
+      "epoch": 0.4510974056780444,
+      "grad_norm": 0.09027385711669922,
+      "learning_rate": 1.6508298080052064e-05,
+      "loss": 9.0151,
+      "step": 90330
+    },
+    {
+      "epoch": 0.4511473445029838,
+      "grad_norm": 0.09425677359104156,
+      "learning_rate": 1.6506796165110518e-05,
+      "loss": 9.0061,
+      "step": 90340
+    },
+    {
+      "epoch": 0.4511972833279233,
+      "grad_norm": 0.09738679975271225,
+      "learning_rate": 1.6505294250168965e-05,
+      "loss": 8.9966,
+      "step": 90350
+    },
+    {
+      "epoch": 0.4512472221528627,
+      "grad_norm": 0.08872897177934647,
+      "learning_rate": 1.6503792335227418e-05,
+      "loss": 9.0142,
+      "step": 90360
+    },
+    {
+      "epoch": 0.4512971609778022,
+      "grad_norm": 0.09494879096746445,
+      "learning_rate": 1.6502290420285865e-05,
+      "loss": 9.0005,
+      "step": 90370
+    },
+    {
+      "epoch": 0.4513470998027416,
+      "grad_norm": 0.09621385484933853,
+      "learning_rate": 1.6500788505344312e-05,
+      "loss": 9.0012,
+      "step": 90380
+    },
+    {
+      "epoch": 0.4513970386276811,
+      "grad_norm": 0.09302900731563568,
+      "learning_rate": 1.6499286590402765e-05,
+      "loss": 9.0184,
+      "step": 90390
+    },
+    {
+      "epoch": 0.4514469774526205,
+      "grad_norm": 0.09508734941482544,
+      "learning_rate": 1.6497784675461212e-05,
+      "loss": 9.007,
+      "step": 90400
+    },
+    {
+      "epoch": 0.45149691627756,
+      "grad_norm": 0.09385523945093155,
+      "learning_rate": 1.6496282760519666e-05,
+      "loss": 9.006,
+      "step": 90410
+    },
+    {
+      "epoch": 0.4515468551024994,
+      "grad_norm": 0.09022466838359833,
+      "learning_rate": 1.6494780845578113e-05,
+      "loss": 9.0222,
+      "step": 90420
+    },
+    {
+      "epoch": 0.4515967939274389,
+      "grad_norm": 0.09833023697137833,
+      "learning_rate": 1.649327893063656e-05,
+      "loss": 8.985,
+      "step": 90430
+    },
+    {
+      "epoch": 0.4516467327523783,
+      "grad_norm": 0.09564008563756943,
+      "learning_rate": 1.6491777015695013e-05,
+      "loss": 9.0073,
+      "step": 90440
+    },
+    {
+      "epoch": 0.4516966715773178,
+      "grad_norm": 0.0922297015786171,
+      "learning_rate": 1.649027510075346e-05,
+      "loss": 8.9923,
+      "step": 90450
+    },
+    {
+      "epoch": 0.4517466104022572,
+      "grad_norm": 0.09411637485027313,
+      "learning_rate": 1.6488773185811913e-05,
+      "loss": 9.0043,
+      "step": 90460
+    },
+    {
+      "epoch": 0.4517965492271967,
+      "grad_norm": 0.09667911380529404,
+      "learning_rate": 1.648727127087036e-05,
+      "loss": 8.9967,
+      "step": 90470
+    },
+    {
+      "epoch": 0.4518464880521361,
+      "grad_norm": 0.09421195834875107,
+      "learning_rate": 1.6485769355928807e-05,
+      "loss": 9.0058,
+      "step": 90480
+    },
+    {
+      "epoch": 0.4518964268770756,
+      "grad_norm": 0.09238166362047195,
+      "learning_rate": 1.648426744098726e-05,
+      "loss": 9.0087,
+      "step": 90490
+    },
+    {
+      "epoch": 0.451946365702015,
+      "grad_norm": 0.09417644143104553,
+      "learning_rate": 1.6482765526045707e-05,
+      "loss": 9.0012,
+      "step": 90500
+    },
+    {
+      "epoch": 0.4519963045269545,
+      "grad_norm": 0.09228955209255219,
+      "learning_rate": 1.648126361110416e-05,
+      "loss": 9.0016,
+      "step": 90510
+    },
+    {
+      "epoch": 0.4520462433518939,
+      "grad_norm": 0.09172794967889786,
+      "learning_rate": 1.6479761696162608e-05,
+      "loss": 8.9866,
+      "step": 90520
+    },
+    {
+      "epoch": 0.4520961821768334,
+      "grad_norm": 0.10119758546352386,
+      "learning_rate": 1.6478259781221058e-05,
+      "loss": 9.0031,
+      "step": 90530
+    },
+    {
+      "epoch": 0.4521461210017728,
+      "grad_norm": 0.10038284212350845,
+      "learning_rate": 1.6476757866279508e-05,
+      "loss": 9.0136,
+      "step": 90540
+    },
+    {
+      "epoch": 0.4521960598267123,
+      "grad_norm": 0.09320379048585892,
+      "learning_rate": 1.6475255951337955e-05,
+      "loss": 8.9853,
+      "step": 90550
+    },
+    {
+      "epoch": 0.4522459986516517,
+      "grad_norm": 0.08812331408262253,
+      "learning_rate": 1.647375403639641e-05,
+      "loss": 9.0052,
+      "step": 90560
+    },
+    {
+      "epoch": 0.4522959374765912,
+      "grad_norm": 0.08946395665407181,
+      "learning_rate": 1.6472252121454855e-05,
+      "loss": 9.0011,
+      "step": 90570
+    },
+    {
+      "epoch": 0.4523458763015306,
+      "grad_norm": 0.09697841852903366,
+      "learning_rate": 1.6470750206513305e-05,
+      "loss": 9.004,
+      "step": 90580
+    },
+    {
+      "epoch": 0.4523958151264701,
+      "grad_norm": 0.09299259632825851,
+      "learning_rate": 1.6469248291571755e-05,
+      "loss": 8.9935,
+      "step": 90590
+    },
+    {
+      "epoch": 0.4524457539514095,
+      "grad_norm": 0.09182324260473251,
+      "learning_rate": 1.6467746376630202e-05,
+      "loss": 9.0015,
+      "step": 90600
+    },
+    {
+      "epoch": 0.452495692776349,
+      "grad_norm": 0.08882872015237808,
+      "learning_rate": 1.6466244461688656e-05,
+      "loss": 9.0002,
+      "step": 90610
+    },
+    {
+      "epoch": 0.4525456316012884,
+      "grad_norm": 0.08870082348585129,
+      "learning_rate": 1.6464742546747103e-05,
+      "loss": 8.9936,
+      "step": 90620
+    },
+    {
+      "epoch": 0.4525955704262279,
+      "grad_norm": 0.09134729206562042,
+      "learning_rate": 1.6463240631805553e-05,
+      "loss": 8.9908,
+      "step": 90630
+    },
+    {
+      "epoch": 0.4526455092511673,
+      "grad_norm": 0.09212338924407959,
+      "learning_rate": 1.6461738716864003e-05,
+      "loss": 9.0031,
+      "step": 90640
+    },
+    {
+      "epoch": 0.4526954480761068,
+      "grad_norm": 0.09525448828935623,
+      "learning_rate": 1.646023680192245e-05,
+      "loss": 8.998,
+      "step": 90650
+    },
+    {
+      "epoch": 0.4527453869010462,
+      "grad_norm": 0.09147049486637115,
+      "learning_rate": 1.6458734886980903e-05,
+      "loss": 9.0005,
+      "step": 90660
+    },
+    {
+      "epoch": 0.4527953257259857,
+      "grad_norm": 0.09832822531461716,
+      "learning_rate": 1.645723297203935e-05,
+      "loss": 8.9923,
+      "step": 90670
+    },
+    {
+      "epoch": 0.4528452645509251,
+      "grad_norm": 0.09704606980085373,
+      "learning_rate": 1.64557310570978e-05,
+      "loss": 8.9927,
+      "step": 90680
+    },
+    {
+      "epoch": 0.4528952033758646,
+      "grad_norm": 0.09421838819980621,
+      "learning_rate": 1.645422914215625e-05,
+      "loss": 9.0079,
+      "step": 90690
+    },
+    {
+      "epoch": 0.452945142200804,
+      "grad_norm": 0.09166116267442703,
+      "learning_rate": 1.6452727227214697e-05,
+      "loss": 8.9912,
+      "step": 90700
+    },
+    {
+      "epoch": 0.4529950810257435,
+      "grad_norm": 0.0929771363735199,
+      "learning_rate": 1.645122531227315e-05,
+      "loss": 8.9928,
+      "step": 90710
+    },
+    {
+      "epoch": 0.4530450198506829,
+      "grad_norm": 0.09662994742393494,
+      "learning_rate": 1.6449723397331598e-05,
+      "loss": 8.997,
+      "step": 90720
+    },
+    {
+      "epoch": 0.4530949586756224,
+      "grad_norm": 0.09966935217380524,
+      "learning_rate": 1.6448221482390048e-05,
+      "loss": 8.9901,
+      "step": 90730
+    },
+    {
+      "epoch": 0.4531448975005618,
+      "grad_norm": 0.09319128841161728,
+      "learning_rate": 1.6446719567448498e-05,
+      "loss": 8.9974,
+      "step": 90740
+    },
+    {
+      "epoch": 0.4531948363255013,
+      "grad_norm": 0.09141194820404053,
+      "learning_rate": 1.6445217652506945e-05,
+      "loss": 9.0124,
+      "step": 90750
+    },
+    {
+      "epoch": 0.4532447751504407,
+      "grad_norm": 0.09347580373287201,
+      "learning_rate": 1.64437157375654e-05,
+      "loss": 8.996,
+      "step": 90760
+    },
+    {
+      "epoch": 0.4532947139753802,
+      "grad_norm": 0.09879345446825027,
+      "learning_rate": 1.6442213822623845e-05,
+      "loss": 8.9926,
+      "step": 90770
+    },
+    {
+      "epoch": 0.4533446528003196,
+      "grad_norm": 0.09470192342996597,
+      "learning_rate": 1.6440711907682295e-05,
+      "loss": 9.0014,
+      "step": 90780
+    },
+    {
+      "epoch": 0.4533945916252591,
+      "grad_norm": 0.09187999367713928,
+      "learning_rate": 1.6439209992740745e-05,
+      "loss": 9.0035,
+      "step": 90790
+    },
+    {
+      "epoch": 0.4534445304501985,
+      "grad_norm": 0.0892479196190834,
+      "learning_rate": 1.6437708077799192e-05,
+      "loss": 9.0248,
+      "step": 90800
+    },
+    {
+      "epoch": 0.453494469275138,
+      "grad_norm": 0.09171134233474731,
+      "learning_rate": 1.6436206162857646e-05,
+      "loss": 8.9885,
+      "step": 90810
+    },
+    {
+      "epoch": 0.4535444081000774,
+      "grad_norm": 0.09449323266744614,
+      "learning_rate": 1.6434704247916093e-05,
+      "loss": 9.0067,
+      "step": 90820
+    },
+    {
+      "epoch": 0.4535943469250169,
+      "grad_norm": 0.09131663292646408,
+      "learning_rate": 1.6433202332974543e-05,
+      "loss": 8.9834,
+      "step": 90830
+    },
+    {
+      "epoch": 0.4536442857499563,
+      "grad_norm": 0.0928645208477974,
+      "learning_rate": 1.6431700418032993e-05,
+      "loss": 8.9925,
+      "step": 90840
+    },
+    {
+      "epoch": 0.4536942245748958,
+      "grad_norm": 0.09293372929096222,
+      "learning_rate": 1.6430198503091443e-05,
+      "loss": 8.998,
+      "step": 90850
+    },
+    {
+      "epoch": 0.4537441633998352,
+      "grad_norm": 0.09626687318086624,
+      "learning_rate": 1.6428696588149893e-05,
+      "loss": 9.0024,
+      "step": 90860
+    },
+    {
+      "epoch": 0.4537941022247747,
+      "grad_norm": 0.09633610397577286,
+      "learning_rate": 1.642719467320834e-05,
+      "loss": 8.9846,
+      "step": 90870
+    },
+    {
+      "epoch": 0.4538440410497141,
+      "grad_norm": 0.09213805943727493,
+      "learning_rate": 1.642569275826679e-05,
+      "loss": 8.9934,
+      "step": 90880
+    },
+    {
+      "epoch": 0.4538939798746536,
+      "grad_norm": 0.09418069571256638,
+      "learning_rate": 1.642419084332524e-05,
+      "loss": 8.987,
+      "step": 90890
+    },
+    {
+      "epoch": 0.453943918699593,
+      "grad_norm": 0.09727146476507187,
+      "learning_rate": 1.642268892838369e-05,
+      "loss": 8.9973,
+      "step": 90900
+    },
+    {
+      "epoch": 0.4539938575245325,
+      "grad_norm": 0.09132228046655655,
+      "learning_rate": 1.642118701344214e-05,
+      "loss": 9.0022,
+      "step": 90910
+    },
+    {
+      "epoch": 0.4540437963494719,
+      "grad_norm": 0.09720085561275482,
+      "learning_rate": 1.6419685098500588e-05,
+      "loss": 9.0058,
+      "step": 90920
+    },
+    {
+      "epoch": 0.4540937351744114,
+      "grad_norm": 0.08539389073848724,
+      "learning_rate": 1.6418183183559038e-05,
+      "loss": 9.0059,
+      "step": 90930
+    },
+    {
+      "epoch": 0.4541436739993508,
+      "grad_norm": 0.09037039428949356,
+      "learning_rate": 1.6416681268617488e-05,
+      "loss": 8.9984,
+      "step": 90940
+    },
+    {
+      "epoch": 0.45419361282429027,
+      "grad_norm": 0.09338293969631195,
+      "learning_rate": 1.6415179353675938e-05,
+      "loss": 8.998,
+      "step": 90950
+    },
+    {
+      "epoch": 0.4542435516492297,
+      "grad_norm": 0.09810806810855865,
+      "learning_rate": 1.641367743873439e-05,
+      "loss": 8.9885,
+      "step": 90960
+    },
+    {
+      "epoch": 0.45429349047416917,
+      "grad_norm": 0.09057030826807022,
+      "learning_rate": 1.6412175523792835e-05,
+      "loss": 8.9988,
+      "step": 90970
+    },
+    {
+      "epoch": 0.4543434292991086,
+      "grad_norm": 0.09103839844465256,
+      "learning_rate": 1.6410673608851285e-05,
+      "loss": 8.9834,
+      "step": 90980
+    },
+    {
+      "epoch": 0.454393368124048,
+      "grad_norm": 0.09123621881008148,
+      "learning_rate": 1.6409171693909735e-05,
+      "loss": 8.9904,
+      "step": 90990
+    },
+    {
+      "epoch": 0.4544433069489875,
+      "grad_norm": 0.09989268332719803,
+      "learning_rate": 1.6407669778968186e-05,
+      "loss": 8.9973,
+      "step": 91000
+    },
+    {
+      "epoch": 0.4544932457739269,
+      "grad_norm": 0.09387070685625076,
+      "learning_rate": 1.6406167864026636e-05,
+      "loss": 8.9821,
+      "step": 91010
+    },
+    {
+      "epoch": 0.4545431845988664,
+      "grad_norm": 0.09730041772127151,
+      "learning_rate": 1.6404665949085083e-05,
+      "loss": 8.9905,
+      "step": 91020
+    },
+    {
+      "epoch": 0.4545931234238058,
+      "grad_norm": 0.09198877215385437,
+      "learning_rate": 1.6403164034143533e-05,
+      "loss": 8.9827,
+      "step": 91030
+    },
+    {
+      "epoch": 0.4546430622487453,
+      "grad_norm": 0.10066971182823181,
+      "learning_rate": 1.6401662119201983e-05,
+      "loss": 8.9825,
+      "step": 91040
+    },
+    {
+      "epoch": 0.4546930010736847,
+      "grad_norm": 0.09631640464067459,
+      "learning_rate": 1.6400160204260433e-05,
+      "loss": 8.9979,
+      "step": 91050
+    },
+    {
+      "epoch": 0.4547429398986242,
+      "grad_norm": 0.09197012335062027,
+      "learning_rate": 1.6398658289318883e-05,
+      "loss": 8.9957,
+      "step": 91060
+    },
+    {
+      "epoch": 0.4547928787235636,
+      "grad_norm": 0.09876050055027008,
+      "learning_rate": 1.639715637437733e-05,
+      "loss": 8.9956,
+      "step": 91070
+    },
+    {
+      "epoch": 0.4548428175485031,
+      "grad_norm": 0.09191004186868668,
+      "learning_rate": 1.639565445943578e-05,
+      "loss": 8.9973,
+      "step": 91080
+    },
+    {
+      "epoch": 0.4548927563734425,
+      "grad_norm": 0.09481745958328247,
+      "learning_rate": 1.639415254449423e-05,
+      "loss": 9.0117,
+      "step": 91090
+    },
+    {
+      "epoch": 0.454942695198382,
+      "grad_norm": 0.09424830228090286,
+      "learning_rate": 1.639265062955268e-05,
+      "loss": 9.0045,
+      "step": 91100
+    },
+    {
+      "epoch": 0.4549926340233214,
+      "grad_norm": 0.09049578011035919,
+      "learning_rate": 1.639114871461113e-05,
+      "loss": 9.0088,
+      "step": 91110
+    },
+    {
+      "epoch": 0.4550425728482609,
+      "grad_norm": 0.09442015737295151,
+      "learning_rate": 1.6389646799669578e-05,
+      "loss": 8.9877,
+      "step": 91120
+    },
+    {
+      "epoch": 0.4550925116732003,
+      "grad_norm": 0.09153556078672409,
+      "learning_rate": 1.6388144884728028e-05,
+      "loss": 8.9957,
+      "step": 91130
+    },
+    {
+      "epoch": 0.4551424504981398,
+      "grad_norm": 0.08938729763031006,
+      "learning_rate": 1.6386642969786478e-05,
+      "loss": 8.9926,
+      "step": 91140
+    },
+    {
+      "epoch": 0.4551923893230792,
+      "grad_norm": 0.10056551545858383,
+      "learning_rate": 1.6385141054844928e-05,
+      "loss": 8.9848,
+      "step": 91150
+    },
+    {
+      "epoch": 0.4552423281480187,
+      "grad_norm": 0.09079209715127945,
+      "learning_rate": 1.638363913990338e-05,
+      "loss": 8.9895,
+      "step": 91160
+    },
+    {
+      "epoch": 0.4552922669729581,
+      "grad_norm": 0.08931328356266022,
+      "learning_rate": 1.6382137224961825e-05,
+      "loss": 8.9885,
+      "step": 91170
+    },
+    {
+      "epoch": 0.4553422057978976,
+      "grad_norm": 0.09365954250097275,
+      "learning_rate": 1.6380635310020275e-05,
+      "loss": 8.9991,
+      "step": 91180
+    },
+    {
+      "epoch": 0.455392144622837,
+      "grad_norm": 0.0947716236114502,
+      "learning_rate": 1.6379133395078725e-05,
+      "loss": 8.993,
+      "step": 91190
+    },
+    {
+      "epoch": 0.4554420834477765,
+      "grad_norm": 0.09569630026817322,
+      "learning_rate": 1.6377631480137176e-05,
+      "loss": 8.9796,
+      "step": 91200
+    },
+    {
+      "epoch": 0.4554920222727159,
+      "grad_norm": 0.09704722464084625,
+      "learning_rate": 1.6376129565195626e-05,
+      "loss": 8.9829,
+      "step": 91210
+    },
+    {
+      "epoch": 0.4555419610976554,
+      "grad_norm": 0.0929335281252861,
+      "learning_rate": 1.6374627650254076e-05,
+      "loss": 9.0032,
+      "step": 91220
+    },
+    {
+      "epoch": 0.4555918999225948,
+      "grad_norm": 0.09553791582584381,
+      "learning_rate": 1.6373125735312523e-05,
+      "loss": 8.9954,
+      "step": 91230
+    },
+    {
+      "epoch": 0.4556418387475343,
+      "grad_norm": 0.09323886781930923,
+      "learning_rate": 1.6371623820370973e-05,
+      "loss": 9.0015,
+      "step": 91240
+    },
+    {
+      "epoch": 0.4556917775724737,
+      "grad_norm": 0.09437048435211182,
+      "learning_rate": 1.6370121905429423e-05,
+      "loss": 9.0025,
+      "step": 91250
+    },
+    {
+      "epoch": 0.4557417163974132,
+      "grad_norm": 0.09459199756383896,
+      "learning_rate": 1.6368619990487873e-05,
+      "loss": 8.9955,
+      "step": 91260
+    },
+    {
+      "epoch": 0.4557916552223526,
+      "grad_norm": 0.09053412079811096,
+      "learning_rate": 1.6367118075546324e-05,
+      "loss": 8.9967,
+      "step": 91270
+    },
+    {
+      "epoch": 0.4558415940472921,
+      "grad_norm": 0.0967337116599083,
+      "learning_rate": 1.636561616060477e-05,
+      "loss": 8.9871,
+      "step": 91280
+    },
+    {
+      "epoch": 0.4558915328722315,
+      "grad_norm": 0.09642165154218674,
+      "learning_rate": 1.636411424566322e-05,
+      "loss": 9.0127,
+      "step": 91290
+    },
+    {
+      "epoch": 0.455941471697171,
+      "grad_norm": 0.09764105081558228,
+      "learning_rate": 1.636261233072167e-05,
+      "loss": 9.0031,
+      "step": 91300
+    },
+    {
+      "epoch": 0.4559914105221104,
+      "grad_norm": 0.08854164928197861,
+      "learning_rate": 1.636111041578012e-05,
+      "loss": 8.9874,
+      "step": 91310
+    },
+    {
+      "epoch": 0.4560413493470499,
+      "grad_norm": 0.09614869207143784,
+      "learning_rate": 1.635960850083857e-05,
+      "loss": 8.977,
+      "step": 91320
+    },
+    {
+      "epoch": 0.4560912881719893,
+      "grad_norm": 0.09570986032485962,
+      "learning_rate": 1.6358106585897018e-05,
+      "loss": 8.9986,
+      "step": 91330
+    },
+    {
+      "epoch": 0.4561412269969288,
+      "grad_norm": 0.0936940461397171,
+      "learning_rate": 1.6356604670955468e-05,
+      "loss": 8.988,
+      "step": 91340
+    },
+    {
+      "epoch": 0.4561911658218682,
+      "grad_norm": 0.08797448128461838,
+      "learning_rate": 1.6355102756013918e-05,
+      "loss": 8.998,
+      "step": 91350
+    },
+    {
+      "epoch": 0.45624110464680767,
+      "grad_norm": 0.0959649458527565,
+      "learning_rate": 1.635360084107237e-05,
+      "loss": 8.9901,
+      "step": 91360
+    },
+    {
+      "epoch": 0.4562910434717471,
+      "grad_norm": 0.08933617919683456,
+      "learning_rate": 1.635209892613082e-05,
+      "loss": 9.0037,
+      "step": 91370
+    },
+    {
+      "epoch": 0.45634098229668657,
+      "grad_norm": 0.09029172360897064,
+      "learning_rate": 1.6350597011189265e-05,
+      "loss": 8.9963,
+      "step": 91380
+    },
+    {
+      "epoch": 0.456390921121626,
+      "grad_norm": 0.0927506536245346,
+      "learning_rate": 1.6349095096247716e-05,
+      "loss": 8.9835,
+      "step": 91390
+    },
+    {
+      "epoch": 0.45644085994656547,
+      "grad_norm": 0.08801010996103287,
+      "learning_rate": 1.6347593181306166e-05,
+      "loss": 8.9871,
+      "step": 91400
+    },
+    {
+      "epoch": 0.4564907987715049,
+      "grad_norm": 0.09252002835273743,
+      "learning_rate": 1.6346091266364616e-05,
+      "loss": 9.0016,
+      "step": 91410
+    },
+    {
+      "epoch": 0.45654073759644437,
+      "grad_norm": 0.09484586119651794,
+      "learning_rate": 1.6344589351423066e-05,
+      "loss": 8.9857,
+      "step": 91420
+    },
+    {
+      "epoch": 0.4565906764213838,
+      "grad_norm": 0.10182981193065643,
+      "learning_rate": 1.6343087436481513e-05,
+      "loss": 8.9776,
+      "step": 91430
+    },
+    {
+      "epoch": 0.45664061524632327,
+      "grad_norm": 0.0948307141661644,
+      "learning_rate": 1.6341585521539963e-05,
+      "loss": 8.9847,
+      "step": 91440
+    },
+    {
+      "epoch": 0.4566905540712627,
+      "grad_norm": 0.0894957110285759,
+      "learning_rate": 1.6340083606598413e-05,
+      "loss": 8.9764,
+      "step": 91450
+    },
+    {
+      "epoch": 0.45674049289620217,
+      "grad_norm": 0.09226840734481812,
+      "learning_rate": 1.6338581691656863e-05,
+      "loss": 8.9938,
+      "step": 91460
+    },
+    {
+      "epoch": 0.4567904317211416,
+      "grad_norm": 0.09212039411067963,
+      "learning_rate": 1.6337079776715314e-05,
+      "loss": 8.9952,
+      "step": 91470
+    },
+    {
+      "epoch": 0.45684037054608106,
+      "grad_norm": 0.09596432745456696,
+      "learning_rate": 1.633557786177376e-05,
+      "loss": 8.9927,
+      "step": 91480
+    },
+    {
+      "epoch": 0.4568903093710205,
+      "grad_norm": 0.09584586322307587,
+      "learning_rate": 1.633407594683221e-05,
+      "loss": 8.9862,
+      "step": 91490
+    },
+    {
+      "epoch": 0.45694024819595996,
+      "grad_norm": 0.09555765986442566,
+      "learning_rate": 1.633257403189066e-05,
+      "loss": 8.9894,
+      "step": 91500
+    },
+    {
+      "epoch": 0.4569901870208994,
+      "grad_norm": 0.09596934914588928,
+      "learning_rate": 1.633107211694911e-05,
+      "loss": 8.9971,
+      "step": 91510
+    },
+    {
+      "epoch": 0.45704012584583886,
+      "grad_norm": 0.09511646628379822,
+      "learning_rate": 1.632957020200756e-05,
+      "loss": 8.9951,
+      "step": 91520
+    },
+    {
+      "epoch": 0.4570900646707783,
+      "grad_norm": 0.09373897314071655,
+      "learning_rate": 1.6328068287066008e-05,
+      "loss": 8.9797,
+      "step": 91530
+    },
+    {
+      "epoch": 0.45714000349571776,
+      "grad_norm": 0.0908622294664383,
+      "learning_rate": 1.632656637212446e-05,
+      "loss": 9.0091,
+      "step": 91540
+    },
+    {
+      "epoch": 0.4571899423206572,
+      "grad_norm": 0.09412607550621033,
+      "learning_rate": 1.6325064457182908e-05,
+      "loss": 8.9875,
+      "step": 91550
+    },
+    {
+      "epoch": 0.45723988114559666,
+      "grad_norm": 0.09757382422685623,
+      "learning_rate": 1.632356254224136e-05,
+      "loss": 8.9888,
+      "step": 91560
+    },
+    {
+      "epoch": 0.4572898199705361,
+      "grad_norm": 0.09289320558309555,
+      "learning_rate": 1.632206062729981e-05,
+      "loss": 8.9767,
+      "step": 91570
+    },
+    {
+      "epoch": 0.45733975879547556,
+      "grad_norm": 0.08800926059484482,
+      "learning_rate": 1.6320558712358255e-05,
+      "loss": 8.9755,
+      "step": 91580
+    },
+    {
+      "epoch": 0.457389697620415,
+      "grad_norm": 0.09186160564422607,
+      "learning_rate": 1.631905679741671e-05,
+      "loss": 8.992,
+      "step": 91590
+    },
+    {
+      "epoch": 0.45743963644535446,
+      "grad_norm": 0.09121951460838318,
+      "learning_rate": 1.6317554882475156e-05,
+      "loss": 8.9873,
+      "step": 91600
+    },
+    {
+      "epoch": 0.4574895752702939,
+      "grad_norm": 0.09226646274328232,
+      "learning_rate": 1.6316052967533606e-05,
+      "loss": 8.9873,
+      "step": 91610
+    },
+    {
+      "epoch": 0.45753951409523336,
+      "grad_norm": 0.0960613414645195,
+      "learning_rate": 1.6314551052592056e-05,
+      "loss": 8.9843,
+      "step": 91620
+    },
+    {
+      "epoch": 0.4575894529201728,
+      "grad_norm": 0.10511282831430435,
+      "learning_rate": 1.6313049137650503e-05,
+      "loss": 8.9846,
+      "step": 91630
+    },
+    {
+      "epoch": 0.45763939174511226,
+      "grad_norm": 0.09277866035699844,
+      "learning_rate": 1.6311547222708956e-05,
+      "loss": 8.9834,
+      "step": 91640
+    },
+    {
+      "epoch": 0.4576893305700517,
+      "grad_norm": 0.09160225838422775,
+      "learning_rate": 1.6310045307767403e-05,
+      "loss": 8.9996,
+      "step": 91650
+    },
+    {
+      "epoch": 0.45773926939499116,
+      "grad_norm": 0.08975838869810104,
+      "learning_rate": 1.6308543392825853e-05,
+      "loss": 8.9845,
+      "step": 91660
+    },
+    {
+      "epoch": 0.4577892082199306,
+      "grad_norm": 0.09508371353149414,
+      "learning_rate": 1.6307041477884304e-05,
+      "loss": 8.9995,
+      "step": 91670
+    },
+    {
+      "epoch": 0.45783914704487005,
+      "grad_norm": 0.09741570800542831,
+      "learning_rate": 1.630553956294275e-05,
+      "loss": 8.9987,
+      "step": 91680
+    },
+    {
+      "epoch": 0.4578890858698095,
+      "grad_norm": 0.09434978663921356,
+      "learning_rate": 1.6304037648001204e-05,
+      "loss": 8.9934,
+      "step": 91690
+    },
+    {
+      "epoch": 0.45793902469474895,
+      "grad_norm": 0.0960199385881424,
+      "learning_rate": 1.630253573305965e-05,
+      "loss": 8.9918,
+      "step": 91700
+    },
+    {
+      "epoch": 0.4579889635196884,
+      "grad_norm": 0.09522512555122375,
+      "learning_rate": 1.63010338181181e-05,
+      "loss": 8.9856,
+      "step": 91710
+    },
+    {
+      "epoch": 0.45803890234462785,
+      "grad_norm": 0.09180162101984024,
+      "learning_rate": 1.629953190317655e-05,
+      "loss": 8.986,
+      "step": 91720
+    },
+    {
+      "epoch": 0.4580888411695673,
+      "grad_norm": 0.09206482768058777,
+      "learning_rate": 1.6298029988234998e-05,
+      "loss": 8.9953,
+      "step": 91730
+    },
+    {
+      "epoch": 0.45813877999450675,
+      "grad_norm": 0.09780178964138031,
+      "learning_rate": 1.629652807329345e-05,
+      "loss": 8.9888,
+      "step": 91740
+    },
+    {
+      "epoch": 0.4581887188194462,
+      "grad_norm": 0.09504421800374985,
+      "learning_rate": 1.6295026158351898e-05,
+      "loss": 8.9846,
+      "step": 91750
+    },
+    {
+      "epoch": 0.45823865764438565,
+      "grad_norm": 0.09662793576717377,
+      "learning_rate": 1.629352424341035e-05,
+      "loss": 8.9828,
+      "step": 91760
+    },
+    {
+      "epoch": 0.45828859646932507,
+      "grad_norm": 0.09310472756624222,
+      "learning_rate": 1.62920223284688e-05,
+      "loss": 9.0069,
+      "step": 91770
+    },
+    {
+      "epoch": 0.45833853529426455,
+      "grad_norm": 0.09177019447088242,
+      "learning_rate": 1.6290520413527245e-05,
+      "loss": 8.9781,
+      "step": 91780
+    },
+    {
+      "epoch": 0.45838847411920397,
+      "grad_norm": 0.09365440160036087,
+      "learning_rate": 1.62890184985857e-05,
+      "loss": 8.9753,
+      "step": 91790
+    },
+    {
+      "epoch": 0.45843841294414345,
+      "grad_norm": 0.09410764276981354,
+      "learning_rate": 1.6287516583644146e-05,
+      "loss": 8.977,
+      "step": 91800
+    },
+    {
+      "epoch": 0.45848835176908287,
+      "grad_norm": 0.09689280390739441,
+      "learning_rate": 1.6286014668702596e-05,
+      "loss": 8.9838,
+      "step": 91810
+    },
+    {
+      "epoch": 0.45853829059402235,
+      "grad_norm": 0.09834188967943192,
+      "learning_rate": 1.6284512753761046e-05,
+      "loss": 8.98,
+      "step": 91820
+    },
+    {
+      "epoch": 0.45858822941896177,
+      "grad_norm": 0.10252006351947784,
+      "learning_rate": 1.6283010838819493e-05,
+      "loss": 8.9913,
+      "step": 91830
+    },
+    {
+      "epoch": 0.45863816824390125,
+      "grad_norm": 0.09507527202367783,
+      "learning_rate": 1.6281508923877946e-05,
+      "loss": 8.9938,
+      "step": 91840
+    },
+    {
+      "epoch": 0.45868810706884067,
+      "grad_norm": 0.09752275049686432,
+      "learning_rate": 1.6280007008936393e-05,
+      "loss": 8.9923,
+      "step": 91850
+    },
+    {
+      "epoch": 0.45873804589378014,
+      "grad_norm": 0.10007300972938538,
+      "learning_rate": 1.6278505093994847e-05,
+      "loss": 9.0013,
+      "step": 91860
+    },
+    {
+      "epoch": 0.45878798471871957,
+      "grad_norm": 0.09385237097740173,
+      "learning_rate": 1.6277003179053294e-05,
+      "loss": 8.975,
+      "step": 91870
+    },
+    {
+      "epoch": 0.45883792354365904,
+      "grad_norm": 0.09489066153764725,
+      "learning_rate": 1.627550126411174e-05,
+      "loss": 8.9761,
+      "step": 91880
+    },
+    {
+      "epoch": 0.45888786236859846,
+      "grad_norm": 0.0923139825463295,
+      "learning_rate": 1.6273999349170194e-05,
+      "loss": 8.9836,
+      "step": 91890
+    },
+    {
+      "epoch": 0.45893780119353794,
+      "grad_norm": 0.09415728598833084,
+      "learning_rate": 1.627249743422864e-05,
+      "loss": 8.984,
+      "step": 91900
+    },
+    {
+      "epoch": 0.45898774001847736,
+      "grad_norm": 0.08979801088571548,
+      "learning_rate": 1.6270995519287094e-05,
+      "loss": 8.9757,
+      "step": 91910
+    },
+    {
+      "epoch": 0.45903767884341684,
+      "grad_norm": 0.09274829924106598,
+      "learning_rate": 1.626949360434554e-05,
+      "loss": 8.9781,
+      "step": 91920
+    },
+    {
+      "epoch": 0.45908761766835626,
+      "grad_norm": 0.09117092192173004,
+      "learning_rate": 1.6267991689403988e-05,
+      "loss": 8.9811,
+      "step": 91930
+    },
+    {
+      "epoch": 0.45913755649329574,
+      "grad_norm": 0.09696992486715317,
+      "learning_rate": 1.626648977446244e-05,
+      "loss": 8.9823,
+      "step": 91940
+    },
+    {
+      "epoch": 0.45918749531823516,
+      "grad_norm": 0.09253089129924774,
+      "learning_rate": 1.6264987859520888e-05,
+      "loss": 8.9919,
+      "step": 91950
+    },
+    {
+      "epoch": 0.45923743414317464,
+      "grad_norm": 0.09517562389373779,
+      "learning_rate": 1.6263485944579342e-05,
+      "loss": 9.0024,
+      "step": 91960
+    },
+    {
+      "epoch": 0.45928737296811406,
+      "grad_norm": 0.09681175649166107,
+      "learning_rate": 1.626198402963779e-05,
+      "loss": 8.9724,
+      "step": 91970
+    },
+    {
+      "epoch": 0.4593373117930535,
+      "grad_norm": 0.08983436971902847,
+      "learning_rate": 1.6260482114696235e-05,
+      "loss": 8.9792,
+      "step": 91980
+    },
+    {
+      "epoch": 0.45938725061799296,
+      "grad_norm": 0.0926215648651123,
+      "learning_rate": 1.625898019975469e-05,
+      "loss": 8.9755,
+      "step": 91990
+    },
+    {
+      "epoch": 0.4594371894429324,
+      "grad_norm": 0.09400998055934906,
+      "learning_rate": 1.6257478284813136e-05,
+      "loss": 8.9726,
+      "step": 92000
+    },
+    {
+      "epoch": 0.45948712826787186,
+      "grad_norm": 0.09896674752235413,
+      "learning_rate": 1.625597636987159e-05,
+      "loss": 8.9919,
+      "step": 92010
+    },
+    {
+      "epoch": 0.4595370670928113,
+      "grad_norm": 0.09126316010951996,
+      "learning_rate": 1.6254474454930036e-05,
+      "loss": 8.9908,
+      "step": 92020
+    },
+    {
+      "epoch": 0.45958700591775076,
+      "grad_norm": 0.09881032258272171,
+      "learning_rate": 1.6252972539988483e-05,
+      "loss": 8.9719,
+      "step": 92030
+    },
+    {
+      "epoch": 0.4596369447426902,
+      "grad_norm": 0.09345117956399918,
+      "learning_rate": 1.6251470625046936e-05,
+      "loss": 8.9801,
+      "step": 92040
+    },
+    {
+      "epoch": 0.45968688356762966,
+      "grad_norm": 0.0911770686507225,
+      "learning_rate": 1.6249968710105383e-05,
+      "loss": 8.9831,
+      "step": 92050
+    },
+    {
+      "epoch": 0.4597368223925691,
+      "grad_norm": 0.09339731186628342,
+      "learning_rate": 1.6248466795163837e-05,
+      "loss": 8.9748,
+      "step": 92060
+    },
+    {
+      "epoch": 0.45978676121750855,
+      "grad_norm": 0.09199456870555878,
+      "learning_rate": 1.6246964880222284e-05,
+      "loss": 8.9885,
+      "step": 92070
+    },
+    {
+      "epoch": 0.459836700042448,
+      "grad_norm": 0.09440702199935913,
+      "learning_rate": 1.624546296528073e-05,
+      "loss": 8.9864,
+      "step": 92080
+    },
+    {
+      "epoch": 0.45988663886738745,
+      "grad_norm": 0.09405873715877533,
+      "learning_rate": 1.6243961050339184e-05,
+      "loss": 8.9815,
+      "step": 92090
+    },
+    {
+      "epoch": 0.4599365776923269,
+      "grad_norm": 0.09130971878767014,
+      "learning_rate": 1.624245913539763e-05,
+      "loss": 8.9784,
+      "step": 92100
+    },
+    {
+      "epoch": 0.45998651651726635,
+      "grad_norm": 0.09779450297355652,
+      "learning_rate": 1.6240957220456084e-05,
+      "loss": 8.9849,
+      "step": 92110
+    },
+    {
+      "epoch": 0.4600364553422058,
+      "grad_norm": 0.09931951016187668,
+      "learning_rate": 1.623945530551453e-05,
+      "loss": 8.9855,
+      "step": 92120
+    },
+    {
+      "epoch": 0.46008639416714525,
+      "grad_norm": 0.08796689659357071,
+      "learning_rate": 1.6237953390572978e-05,
+      "loss": 8.9822,
+      "step": 92130
+    },
+    {
+      "epoch": 0.4601363329920847,
+      "grad_norm": 0.09218308329582214,
+      "learning_rate": 1.623645147563143e-05,
+      "loss": 8.9771,
+      "step": 92140
+    },
+    {
+      "epoch": 0.46018627181702415,
+      "grad_norm": 0.09494258463382721,
+      "learning_rate": 1.6234949560689878e-05,
+      "loss": 8.9803,
+      "step": 92150
+    },
+    {
+      "epoch": 0.4602362106419636,
+      "grad_norm": 0.09072136878967285,
+      "learning_rate": 1.6233447645748332e-05,
+      "loss": 8.9806,
+      "step": 92160
+    },
+    {
+      "epoch": 0.46028614946690305,
+      "grad_norm": 0.09422437846660614,
+      "learning_rate": 1.623194573080678e-05,
+      "loss": 8.992,
+      "step": 92170
+    },
+    {
+      "epoch": 0.46033608829184247,
+      "grad_norm": 0.09309976547956467,
+      "learning_rate": 1.623044381586523e-05,
+      "loss": 8.9822,
+      "step": 92180
+    },
+    {
+      "epoch": 0.46038602711678195,
+      "grad_norm": 0.09436167776584625,
+      "learning_rate": 1.622894190092368e-05,
+      "loss": 8.9833,
+      "step": 92190
+    },
+    {
+      "epoch": 0.46043596594172137,
+      "grad_norm": 0.09215915203094482,
+      "learning_rate": 1.6227439985982126e-05,
+      "loss": 8.9867,
+      "step": 92200
+    },
+    {
+      "epoch": 0.46048590476666085,
+      "grad_norm": 0.0927964299917221,
+      "learning_rate": 1.622593807104058e-05,
+      "loss": 8.9965,
+      "step": 92210
+    },
+    {
+      "epoch": 0.46053584359160027,
+      "grad_norm": 0.08990570157766342,
+      "learning_rate": 1.6224436156099026e-05,
+      "loss": 8.9703,
+      "step": 92220
+    },
+    {
+      "epoch": 0.46058578241653975,
+      "grad_norm": 0.09287389367818832,
+      "learning_rate": 1.6222934241157476e-05,
+      "loss": 8.9852,
+      "step": 92230
+    },
+    {
+      "epoch": 0.46063572124147917,
+      "grad_norm": 0.09232201427221298,
+      "learning_rate": 1.6221432326215926e-05,
+      "loss": 8.9978,
+      "step": 92240
+    },
+    {
+      "epoch": 0.46068566006641865,
+      "grad_norm": 0.09181149303913116,
+      "learning_rate": 1.6219930411274373e-05,
+      "loss": 8.9753,
+      "step": 92250
+    },
+    {
+      "epoch": 0.46073559889135807,
+      "grad_norm": 0.08810899406671524,
+      "learning_rate": 1.6218428496332827e-05,
+      "loss": 8.9912,
+      "step": 92260
+    },
+    {
+      "epoch": 0.46078553771629754,
+      "grad_norm": 0.09229025989770889,
+      "learning_rate": 1.6216926581391274e-05,
+      "loss": 8.9822,
+      "step": 92270
+    },
+    {
+      "epoch": 0.46083547654123697,
+      "grad_norm": 0.09612645953893661,
+      "learning_rate": 1.6215424666449724e-05,
+      "loss": 8.9762,
+      "step": 92280
+    },
+    {
+      "epoch": 0.46088541536617644,
+      "grad_norm": 0.09475258737802505,
+      "learning_rate": 1.6213922751508174e-05,
+      "loss": 8.9794,
+      "step": 92290
+    },
+    {
+      "epoch": 0.46093535419111586,
+      "grad_norm": 0.08649539202451706,
+      "learning_rate": 1.621242083656662e-05,
+      "loss": 8.9705,
+      "step": 92300
+    },
+    {
+      "epoch": 0.46098529301605534,
+      "grad_norm": 0.09077353775501251,
+      "learning_rate": 1.6210918921625074e-05,
+      "loss": 8.9873,
+      "step": 92310
+    },
+    {
+      "epoch": 0.46103523184099476,
+      "grad_norm": 0.09602386504411697,
+      "learning_rate": 1.620941700668352e-05,
+      "loss": 8.9859,
+      "step": 92320
+    },
+    {
+      "epoch": 0.46108517066593424,
+      "grad_norm": 0.09134723991155624,
+      "learning_rate": 1.620791509174197e-05,
+      "loss": 8.9876,
+      "step": 92330
+    },
+    {
+      "epoch": 0.46113510949087366,
+      "grad_norm": 0.09485665708780289,
+      "learning_rate": 1.620641317680042e-05,
+      "loss": 8.9749,
+      "step": 92340
+    },
+    {
+      "epoch": 0.46118504831581314,
+      "grad_norm": 0.089450903236866,
+      "learning_rate": 1.6204911261858868e-05,
+      "loss": 8.9815,
+      "step": 92350
+    },
+    {
+      "epoch": 0.46123498714075256,
+      "grad_norm": 0.09187912195920944,
+      "learning_rate": 1.6203409346917322e-05,
+      "loss": 8.9729,
+      "step": 92360
+    },
+    {
+      "epoch": 0.46128492596569204,
+      "grad_norm": 0.09176033735275269,
+      "learning_rate": 1.620190743197577e-05,
+      "loss": 8.9813,
+      "step": 92370
+    },
+    {
+      "epoch": 0.46133486479063146,
+      "grad_norm": 0.09702310711145401,
+      "learning_rate": 1.620040551703422e-05,
+      "loss": 8.9771,
+      "step": 92380
+    },
+    {
+      "epoch": 0.46138480361557094,
+      "grad_norm": 0.09183552861213684,
+      "learning_rate": 1.619890360209267e-05,
+      "loss": 8.9827,
+      "step": 92390
+    },
+    {
+      "epoch": 0.46143474244051036,
+      "grad_norm": 0.09094628691673279,
+      "learning_rate": 1.6197401687151116e-05,
+      "loss": 8.9792,
+      "step": 92400
+    },
+    {
+      "epoch": 0.46148468126544984,
+      "grad_norm": 0.09234099090099335,
+      "learning_rate": 1.619589977220957e-05,
+      "loss": 8.9879,
+      "step": 92410
+    },
+    {
+      "epoch": 0.46153462009038926,
+      "grad_norm": 0.0963030532002449,
+      "learning_rate": 1.6194397857268016e-05,
+      "loss": 8.9812,
+      "step": 92420
+    },
+    {
+      "epoch": 0.46158455891532874,
+      "grad_norm": 0.09307033568620682,
+      "learning_rate": 1.6192895942326466e-05,
+      "loss": 8.9806,
+      "step": 92430
+    },
+    {
+      "epoch": 0.46163449774026816,
+      "grad_norm": 0.08998160809278488,
+      "learning_rate": 1.6191394027384916e-05,
+      "loss": 8.9827,
+      "step": 92440
+    },
+    {
+      "epoch": 0.46168443656520763,
+      "grad_norm": 0.09202240407466888,
+      "learning_rate": 1.6189892112443363e-05,
+      "loss": 8.9737,
+      "step": 92450
+    },
+    {
+      "epoch": 0.46173437539014706,
+      "grad_norm": 0.08661884814500809,
+      "learning_rate": 1.6188390197501817e-05,
+      "loss": 8.9747,
+      "step": 92460
+    },
+    {
+      "epoch": 0.46178431421508653,
+      "grad_norm": 0.08981632441282272,
+      "learning_rate": 1.6186888282560264e-05,
+      "loss": 8.973,
+      "step": 92470
+    },
+    {
+      "epoch": 0.46183425304002595,
+      "grad_norm": 0.09431172162294388,
+      "learning_rate": 1.6185386367618714e-05,
+      "loss": 8.972,
+      "step": 92480
+    },
+    {
+      "epoch": 0.46188419186496543,
+      "grad_norm": 0.09960649162530899,
+      "learning_rate": 1.6183884452677164e-05,
+      "loss": 8.9833,
+      "step": 92490
+    },
+    {
+      "epoch": 0.46193413068990485,
+      "grad_norm": 0.09025320410728455,
+      "learning_rate": 1.6182382537735614e-05,
+      "loss": 8.9745,
+      "step": 92500
+    },
+    {
+      "epoch": 0.46198406951484433,
+      "grad_norm": 0.09079807996749878,
+      "learning_rate": 1.6180880622794064e-05,
+      "loss": 8.9746,
+      "step": 92510
+    },
+    {
+      "epoch": 0.46203400833978375,
+      "grad_norm": 0.09874525666236877,
+      "learning_rate": 1.617937870785251e-05,
+      "loss": 8.9767,
+      "step": 92520
+    },
+    {
+      "epoch": 0.46208394716472323,
+      "grad_norm": 0.09728190302848816,
+      "learning_rate": 1.617787679291096e-05,
+      "loss": 8.9802,
+      "step": 92530
+    },
+    {
+      "epoch": 0.46213388598966265,
+      "grad_norm": 0.09223682433366776,
+      "learning_rate": 1.617637487796941e-05,
+      "loss": 8.9872,
+      "step": 92540
+    },
+    {
+      "epoch": 0.46218382481460213,
+      "grad_norm": 0.09065338224172592,
+      "learning_rate": 1.617487296302786e-05,
+      "loss": 8.9868,
+      "step": 92550
+    },
+    {
+      "epoch": 0.46223376363954155,
+      "grad_norm": 0.09436984360218048,
+      "learning_rate": 1.6173371048086312e-05,
+      "loss": 8.989,
+      "step": 92560
+    },
+    {
+      "epoch": 0.46228370246448103,
+      "grad_norm": 0.09117233753204346,
+      "learning_rate": 1.617186913314476e-05,
+      "loss": 8.9857,
+      "step": 92570
+    },
+    {
+      "epoch": 0.46233364128942045,
+      "grad_norm": 0.09342578798532486,
+      "learning_rate": 1.617036721820321e-05,
+      "loss": 8.9786,
+      "step": 92580
+    },
+    {
+      "epoch": 0.4623835801143599,
+      "grad_norm": 0.09534727782011032,
+      "learning_rate": 1.616886530326166e-05,
+      "loss": 8.9899,
+      "step": 92590
+    },
+    {
+      "epoch": 0.46243351893929935,
+      "grad_norm": 0.09320477396249771,
+      "learning_rate": 1.616736338832011e-05,
+      "loss": 8.9657,
+      "step": 92600
+    },
+    {
+      "epoch": 0.4624834577642388,
+      "grad_norm": 0.09745677560567856,
+      "learning_rate": 1.616586147337856e-05,
+      "loss": 8.9708,
+      "step": 92610
+    },
+    {
+      "epoch": 0.46253339658917825,
+      "grad_norm": 0.09360513091087341,
+      "learning_rate": 1.6164359558437006e-05,
+      "loss": 8.9846,
+      "step": 92620
+    },
+    {
+      "epoch": 0.4625833354141177,
+      "grad_norm": 0.09421921521425247,
+      "learning_rate": 1.6162857643495456e-05,
+      "loss": 8.9829,
+      "step": 92630
+    },
+    {
+      "epoch": 0.46263327423905715,
+      "grad_norm": 0.10035274922847748,
+      "learning_rate": 1.6161355728553906e-05,
+      "loss": 8.9633,
+      "step": 92640
+    },
+    {
+      "epoch": 0.4626832130639966,
+      "grad_norm": 0.09385058283805847,
+      "learning_rate": 1.6159853813612357e-05,
+      "loss": 8.9702,
+      "step": 92650
+    },
+    {
+      "epoch": 0.46273315188893605,
+      "grad_norm": 0.0909324511885643,
+      "learning_rate": 1.6158351898670807e-05,
+      "loss": 8.9831,
+      "step": 92660
+    },
+    {
+      "epoch": 0.4627830907138755,
+      "grad_norm": 0.08683860301971436,
+      "learning_rate": 1.6156849983729254e-05,
+      "loss": 8.9797,
+      "step": 92670
+    },
+    {
+      "epoch": 0.46283302953881494,
+      "grad_norm": 0.09413138777017593,
+      "learning_rate": 1.6155348068787704e-05,
+      "loss": 8.9752,
+      "step": 92680
+    },
+    {
+      "epoch": 0.4628829683637544,
+      "grad_norm": 0.09523726254701614,
+      "learning_rate": 1.6153846153846154e-05,
+      "loss": 8.9746,
+      "step": 92690
+    },
+    {
+      "epoch": 0.46293290718869384,
+      "grad_norm": 0.09190825372934341,
+      "learning_rate": 1.6152344238904604e-05,
+      "loss": 8.9821,
+      "step": 92700
+    },
+    {
+      "epoch": 0.4629828460136333,
+      "grad_norm": 0.09379702806472778,
+      "learning_rate": 1.6150842323963054e-05,
+      "loss": 8.9772,
+      "step": 92710
+    },
+    {
+      "epoch": 0.46303278483857274,
+      "grad_norm": 0.09580748528242111,
+      "learning_rate": 1.61493404090215e-05,
+      "loss": 8.9785,
+      "step": 92720
+    },
+    {
+      "epoch": 0.4630827236635122,
+      "grad_norm": 0.10296031832695007,
+      "learning_rate": 1.614783849407995e-05,
+      "loss": 8.9668,
+      "step": 92730
+    },
+    {
+      "epoch": 0.46313266248845164,
+      "grad_norm": 0.09141591936349869,
+      "learning_rate": 1.61463365791384e-05,
+      "loss": 8.9757,
+      "step": 92740
+    },
+    {
+      "epoch": 0.4631826013133911,
+      "grad_norm": 0.09824127703905106,
+      "learning_rate": 1.614483466419685e-05,
+      "loss": 8.9831,
+      "step": 92750
+    },
+    {
+      "epoch": 0.46323254013833054,
+      "grad_norm": 0.0958201214671135,
+      "learning_rate": 1.6143332749255302e-05,
+      "loss": 8.9645,
+      "step": 92760
+    },
+    {
+      "epoch": 0.46328247896327,
+      "grad_norm": 0.08969037979841232,
+      "learning_rate": 1.614183083431375e-05,
+      "loss": 8.9695,
+      "step": 92770
+    },
+    {
+      "epoch": 0.46333241778820944,
+      "grad_norm": 0.0948222279548645,
+      "learning_rate": 1.61403289193722e-05,
+      "loss": 8.9856,
+      "step": 92780
+    },
+    {
+      "epoch": 0.4633823566131489,
+      "grad_norm": 0.08855247497558594,
+      "learning_rate": 1.613882700443065e-05,
+      "loss": 8.9811,
+      "step": 92790
+    },
+    {
+      "epoch": 0.46343229543808834,
+      "grad_norm": 0.08753466606140137,
+      "learning_rate": 1.61373250894891e-05,
+      "loss": 8.9728,
+      "step": 92800
+    },
+    {
+      "epoch": 0.4634822342630278,
+      "grad_norm": 0.09330339729785919,
+      "learning_rate": 1.613582317454755e-05,
+      "loss": 8.9759,
+      "step": 92810
+    },
+    {
+      "epoch": 0.46353217308796724,
+      "grad_norm": 0.0953167974948883,
+      "learning_rate": 1.6134321259606e-05,
+      "loss": 8.9906,
+      "step": 92820
+    },
+    {
+      "epoch": 0.4635821119129067,
+      "grad_norm": 0.09375966340303421,
+      "learning_rate": 1.6132819344664446e-05,
+      "loss": 8.9694,
+      "step": 92830
+    },
+    {
+      "epoch": 0.46363205073784614,
+      "grad_norm": 0.08928825706243515,
+      "learning_rate": 1.6131317429722897e-05,
+      "loss": 8.9641,
+      "step": 92840
+    },
+    {
+      "epoch": 0.4636819895627856,
+      "grad_norm": 0.09437457472085953,
+      "learning_rate": 1.6129815514781347e-05,
+      "loss": 8.9668,
+      "step": 92850
+    },
+    {
+      "epoch": 0.46373192838772503,
+      "grad_norm": 0.0952828899025917,
+      "learning_rate": 1.6128313599839797e-05,
+      "loss": 8.9885,
+      "step": 92860
+    },
+    {
+      "epoch": 0.4637818672126645,
+      "grad_norm": 0.08954987674951553,
+      "learning_rate": 1.6126811684898247e-05,
+      "loss": 8.9779,
+      "step": 92870
+    },
+    {
+      "epoch": 0.46383180603760393,
+      "grad_norm": 0.095025435090065,
+      "learning_rate": 1.6125309769956694e-05,
+      "loss": 8.978,
+      "step": 92880
+    },
+    {
+      "epoch": 0.4638817448625434,
+      "grad_norm": 0.09305167198181152,
+      "learning_rate": 1.6123807855015144e-05,
+      "loss": 8.975,
+      "step": 92890
+    },
+    {
+      "epoch": 0.46393168368748283,
+      "grad_norm": 0.09365760535001755,
+      "learning_rate": 1.6122305940073594e-05,
+      "loss": 8.9796,
+      "step": 92900
+    },
+    {
+      "epoch": 0.4639816225124223,
+      "grad_norm": 0.0907236710190773,
+      "learning_rate": 1.6120804025132044e-05,
+      "loss": 8.9814,
+      "step": 92910
+    },
+    {
+      "epoch": 0.46403156133736173,
+      "grad_norm": 0.09334593266248703,
+      "learning_rate": 1.6119302110190495e-05,
+      "loss": 8.9935,
+      "step": 92920
+    },
+    {
+      "epoch": 0.4640815001623012,
+      "grad_norm": 0.0942806750535965,
+      "learning_rate": 1.611780019524894e-05,
+      "loss": 8.968,
+      "step": 92930
+    },
+    {
+      "epoch": 0.46413143898724063,
+      "grad_norm": 0.09642312675714493,
+      "learning_rate": 1.611629828030739e-05,
+      "loss": 8.9773,
+      "step": 92940
+    },
+    {
+      "epoch": 0.4641813778121801,
+      "grad_norm": 0.10480822622776031,
+      "learning_rate": 1.611479636536584e-05,
+      "loss": 8.9643,
+      "step": 92950
+    },
+    {
+      "epoch": 0.46423131663711953,
+      "grad_norm": 0.09461807459592819,
+      "learning_rate": 1.6113294450424292e-05,
+      "loss": 8.9774,
+      "step": 92960
+    },
+    {
+      "epoch": 0.46428125546205895,
+      "grad_norm": 0.09156806021928787,
+      "learning_rate": 1.6111792535482742e-05,
+      "loss": 8.983,
+      "step": 92970
+    },
+    {
+      "epoch": 0.4643311942869984,
+      "grad_norm": 0.09091312438249588,
+      "learning_rate": 1.611029062054119e-05,
+      "loss": 8.9742,
+      "step": 92980
+    },
+    {
+      "epoch": 0.46438113311193785,
+      "grad_norm": 0.09511202573776245,
+      "learning_rate": 1.610878870559964e-05,
+      "loss": 8.9784,
+      "step": 92990
+    },
+    {
+      "epoch": 0.4644310719368773,
+      "grad_norm": 0.09062772989273071,
+      "learning_rate": 1.610728679065809e-05,
+      "loss": 8.9831,
+      "step": 93000
+    },
+    {
+      "epoch": 0.46448101076181675,
+      "grad_norm": 0.09874343872070312,
+      "learning_rate": 1.610578487571654e-05,
+      "loss": 8.9679,
+      "step": 93010
+    },
+    {
+      "epoch": 0.4645309495867562,
+      "grad_norm": 0.09453938156366348,
+      "learning_rate": 1.610428296077499e-05,
+      "loss": 8.9649,
+      "step": 93020
+    },
+    {
+      "epoch": 0.46458088841169565,
+      "grad_norm": 0.09480418264865875,
+      "learning_rate": 1.6102781045833436e-05,
+      "loss": 8.9785,
+      "step": 93030
+    },
+    {
+      "epoch": 0.4646308272366351,
+      "grad_norm": 0.08807534724473953,
+      "learning_rate": 1.6101279130891887e-05,
+      "loss": 8.9728,
+      "step": 93040
+    },
+    {
+      "epoch": 0.46468076606157455,
+      "grad_norm": 0.09540516883134842,
+      "learning_rate": 1.6099777215950337e-05,
+      "loss": 8.9793,
+      "step": 93050
+    },
+    {
+      "epoch": 0.464730704886514,
+      "grad_norm": 0.09324967861175537,
+      "learning_rate": 1.6098275301008787e-05,
+      "loss": 8.9792,
+      "step": 93060
+    },
+    {
+      "epoch": 0.46478064371145345,
+      "grad_norm": 0.08924508839845657,
+      "learning_rate": 1.6096773386067237e-05,
+      "loss": 8.9898,
+      "step": 93070
+    },
+    {
+      "epoch": 0.4648305825363929,
+      "grad_norm": 0.09416824579238892,
+      "learning_rate": 1.6095271471125684e-05,
+      "loss": 8.9758,
+      "step": 93080
+    },
+    {
+      "epoch": 0.46488052136133234,
+      "grad_norm": 0.09986349195241928,
+      "learning_rate": 1.6093769556184134e-05,
+      "loss": 8.9774,
+      "step": 93090
+    },
+    {
+      "epoch": 0.4649304601862718,
+      "grad_norm": 0.10064547508955002,
+      "learning_rate": 1.6092267641242584e-05,
+      "loss": 8.9643,
+      "step": 93100
+    },
+    {
+      "epoch": 0.46498039901121124,
+      "grad_norm": 0.09657155722379684,
+      "learning_rate": 1.6090765726301034e-05,
+      "loss": 8.992,
+      "step": 93110
+    },
+    {
+      "epoch": 0.4650303378361507,
+      "grad_norm": 0.09835278242826462,
+      "learning_rate": 1.6089263811359485e-05,
+      "loss": 8.9773,
+      "step": 93120
+    },
+    {
+      "epoch": 0.46508027666109014,
+      "grad_norm": 0.09444111585617065,
+      "learning_rate": 1.608776189641793e-05,
+      "loss": 8.9678,
+      "step": 93130
+    },
+    {
+      "epoch": 0.4651302154860296,
+      "grad_norm": 0.09481803327798843,
+      "learning_rate": 1.6086259981476385e-05,
+      "loss": 8.9745,
+      "step": 93140
+    },
+    {
+      "epoch": 0.46518015431096904,
+      "grad_norm": 0.09177707135677338,
+      "learning_rate": 1.6084758066534832e-05,
+      "loss": 8.973,
+      "step": 93150
+    },
+    {
+      "epoch": 0.4652300931359085,
+      "grad_norm": 0.09027852863073349,
+      "learning_rate": 1.6083256151593282e-05,
+      "loss": 8.9625,
+      "step": 93160
+    },
+    {
+      "epoch": 0.46528003196084794,
+      "grad_norm": 0.0904780849814415,
+      "learning_rate": 1.6081754236651732e-05,
+      "loss": 8.9686,
+      "step": 93170
+    },
+    {
+      "epoch": 0.4653299707857874,
+      "grad_norm": 0.09435466676950455,
+      "learning_rate": 1.608025232171018e-05,
+      "loss": 8.9654,
+      "step": 93180
+    },
+    {
+      "epoch": 0.46537990961072684,
+      "grad_norm": 0.09567007422447205,
+      "learning_rate": 1.6078750406768632e-05,
+      "loss": 8.9825,
+      "step": 93190
+    },
+    {
+      "epoch": 0.4654298484356663,
+      "grad_norm": 0.09451539069414139,
+      "learning_rate": 1.607724849182708e-05,
+      "loss": 8.9788,
+      "step": 93200
+    },
+    {
+      "epoch": 0.46547978726060574,
+      "grad_norm": 0.09207061678171158,
+      "learning_rate": 1.607574657688553e-05,
+      "loss": 8.972,
+      "step": 93210
+    },
+    {
+      "epoch": 0.4655297260855452,
+      "grad_norm": 0.09581438452005386,
+      "learning_rate": 1.607424466194398e-05,
+      "loss": 8.9847,
+      "step": 93220
+    },
+    {
+      "epoch": 0.46557966491048464,
+      "grad_norm": 0.09088779985904694,
+      "learning_rate": 1.6072742747002426e-05,
+      "loss": 8.9714,
+      "step": 93230
+    },
+    {
+      "epoch": 0.4656296037354241,
+      "grad_norm": 0.09318163245916367,
+      "learning_rate": 1.607124083206088e-05,
+      "loss": 8.9726,
+      "step": 93240
+    },
+    {
+      "epoch": 0.46567954256036354,
+      "grad_norm": 0.09258940815925598,
+      "learning_rate": 1.6069738917119327e-05,
+      "loss": 8.9878,
+      "step": 93250
+    },
+    {
+      "epoch": 0.465729481385303,
+      "grad_norm": 0.09311637282371521,
+      "learning_rate": 1.6068237002177777e-05,
+      "loss": 8.9817,
+      "step": 93260
+    },
+    {
+      "epoch": 0.46577942021024243,
+      "grad_norm": 0.09517588466405869,
+      "learning_rate": 1.6066735087236227e-05,
+      "loss": 8.9635,
+      "step": 93270
+    },
+    {
+      "epoch": 0.4658293590351819,
+      "grad_norm": 0.08702480047941208,
+      "learning_rate": 1.6065233172294674e-05,
+      "loss": 8.9762,
+      "step": 93280
+    },
+    {
+      "epoch": 0.46587929786012133,
+      "grad_norm": 0.0963396281003952,
+      "learning_rate": 1.6063731257353127e-05,
+      "loss": 8.967,
+      "step": 93290
+    },
+    {
+      "epoch": 0.4659292366850608,
+      "grad_norm": 0.09384248405694962,
+      "learning_rate": 1.6062229342411574e-05,
+      "loss": 8.9723,
+      "step": 93300
+    },
+    {
+      "epoch": 0.46597917551000023,
+      "grad_norm": 0.1044907197356224,
+      "learning_rate": 1.6060727427470024e-05,
+      "loss": 8.9606,
+      "step": 93310
+    },
+    {
+      "epoch": 0.4660291143349397,
+      "grad_norm": 0.09456614404916763,
+      "learning_rate": 1.6059225512528475e-05,
+      "loss": 8.9805,
+      "step": 93320
+    },
+    {
+      "epoch": 0.46607905315987913,
+      "grad_norm": 0.09772161394357681,
+      "learning_rate": 1.605772359758692e-05,
+      "loss": 8.967,
+      "step": 93330
+    },
+    {
+      "epoch": 0.4661289919848186,
+      "grad_norm": 0.08841816335916519,
+      "learning_rate": 1.6056221682645375e-05,
+      "loss": 8.9762,
+      "step": 93340
+    },
+    {
+      "epoch": 0.46617893080975803,
+      "grad_norm": 0.0914144292473793,
+      "learning_rate": 1.6054719767703822e-05,
+      "loss": 8.978,
+      "step": 93350
+    },
+    {
+      "epoch": 0.4662288696346975,
+      "grad_norm": 0.09596528857946396,
+      "learning_rate": 1.6053217852762272e-05,
+      "loss": 8.956,
+      "step": 93360
+    },
+    {
+      "epoch": 0.46627880845963693,
+      "grad_norm": 0.09019865840673447,
+      "learning_rate": 1.6051715937820722e-05,
+      "loss": 8.965,
+      "step": 93370
+    },
+    {
+      "epoch": 0.4663287472845764,
+      "grad_norm": 0.09683724492788315,
+      "learning_rate": 1.605021402287917e-05,
+      "loss": 8.9587,
+      "step": 93380
+    },
+    {
+      "epoch": 0.4663786861095158,
+      "grad_norm": 0.08984372764825821,
+      "learning_rate": 1.6048712107937622e-05,
+      "loss": 8.9803,
+      "step": 93390
+    },
+    {
+      "epoch": 0.4664286249344553,
+      "grad_norm": 0.09242388606071472,
+      "learning_rate": 1.604721019299607e-05,
+      "loss": 8.991,
+      "step": 93400
+    },
+    {
+      "epoch": 0.4664785637593947,
+      "grad_norm": 0.0912506952881813,
+      "learning_rate": 1.604570827805452e-05,
+      "loss": 8.9865,
+      "step": 93410
+    },
+    {
+      "epoch": 0.4665285025843342,
+      "grad_norm": 0.09702800214290619,
+      "learning_rate": 1.604420636311297e-05,
+      "loss": 8.9536,
+      "step": 93420
+    },
+    {
+      "epoch": 0.4665784414092736,
+      "grad_norm": 0.09083381295204163,
+      "learning_rate": 1.6042704448171416e-05,
+      "loss": 8.9659,
+      "step": 93430
+    },
+    {
+      "epoch": 0.4666283802342131,
+      "grad_norm": 0.09859057515859604,
+      "learning_rate": 1.604120253322987e-05,
+      "loss": 8.9757,
+      "step": 93440
+    },
+    {
+      "epoch": 0.4666783190591525,
+      "grad_norm": 0.09232448786497116,
+      "learning_rate": 1.6039700618288317e-05,
+      "loss": 8.9693,
+      "step": 93450
+    },
+    {
+      "epoch": 0.466728257884092,
+      "grad_norm": 0.09486878663301468,
+      "learning_rate": 1.603819870334677e-05,
+      "loss": 8.9813,
+      "step": 93460
+    },
+    {
+      "epoch": 0.4667781967090314,
+      "grad_norm": 0.09302292764186859,
+      "learning_rate": 1.6036696788405217e-05,
+      "loss": 8.9782,
+      "step": 93470
+    },
+    {
+      "epoch": 0.4668281355339709,
+      "grad_norm": 0.09293092787265778,
+      "learning_rate": 1.6035194873463664e-05,
+      "loss": 8.9629,
+      "step": 93480
+    },
+    {
+      "epoch": 0.4668780743589103,
+      "grad_norm": 0.09174246340990067,
+      "learning_rate": 1.6033692958522117e-05,
+      "loss": 8.9694,
+      "step": 93490
+    },
+    {
+      "epoch": 0.4669280131838498,
+      "grad_norm": 0.09290467202663422,
+      "learning_rate": 1.6032191043580564e-05,
+      "loss": 8.9646,
+      "step": 93500
+    },
+    {
+      "epoch": 0.4669779520087892,
+      "grad_norm": 0.09408622980117798,
+      "learning_rate": 1.6030689128639018e-05,
+      "loss": 8.9694,
+      "step": 93510
+    },
+    {
+      "epoch": 0.4670278908337287,
+      "grad_norm": 0.09096479415893555,
+      "learning_rate": 1.6029187213697465e-05,
+      "loss": 8.966,
+      "step": 93520
+    },
+    {
+      "epoch": 0.4670778296586681,
+      "grad_norm": 0.09430520981550217,
+      "learning_rate": 1.602768529875591e-05,
+      "loss": 8.9614,
+      "step": 93530
+    },
+    {
+      "epoch": 0.4671277684836076,
+      "grad_norm": 0.09501869231462479,
+      "learning_rate": 1.6026183383814365e-05,
+      "loss": 8.954,
+      "step": 93540
+    },
+    {
+      "epoch": 0.467177707308547,
+      "grad_norm": 0.09236142039299011,
+      "learning_rate": 1.6024681468872812e-05,
+      "loss": 8.9749,
+      "step": 93550
+    },
+    {
+      "epoch": 0.4672276461334865,
+      "grad_norm": 0.09646128863096237,
+      "learning_rate": 1.6023179553931265e-05,
+      "loss": 8.985,
+      "step": 93560
+    },
+    {
+      "epoch": 0.4672775849584259,
+      "grad_norm": 0.08994904160499573,
+      "learning_rate": 1.6021677638989712e-05,
+      "loss": 8.9734,
+      "step": 93570
+    },
+    {
+      "epoch": 0.4673275237833654,
+      "grad_norm": 0.0974574014544487,
+      "learning_rate": 1.602017572404816e-05,
+      "loss": 8.9715,
+      "step": 93580
+    },
+    {
+      "epoch": 0.4673774626083048,
+      "grad_norm": 0.09206629544496536,
+      "learning_rate": 1.6018673809106612e-05,
+      "loss": 8.9758,
+      "step": 93590
+    },
+    {
+      "epoch": 0.4674274014332443,
+      "grad_norm": 0.08921888470649719,
+      "learning_rate": 1.601717189416506e-05,
+      "loss": 8.9747,
+      "step": 93600
+    },
+    {
+      "epoch": 0.4674773402581837,
+      "grad_norm": 0.09754341095685959,
+      "learning_rate": 1.6015669979223513e-05,
+      "loss": 8.9635,
+      "step": 93610
+    },
+    {
+      "epoch": 0.4675272790831232,
+      "grad_norm": 0.09427353739738464,
+      "learning_rate": 1.601416806428196e-05,
+      "loss": 8.9758,
+      "step": 93620
+    },
+    {
+      "epoch": 0.4675772179080626,
+      "grad_norm": 0.09223822504281998,
+      "learning_rate": 1.6012666149340406e-05,
+      "loss": 8.9795,
+      "step": 93630
+    },
+    {
+      "epoch": 0.4676271567330021,
+      "grad_norm": 0.09685010462999344,
+      "learning_rate": 1.601116423439886e-05,
+      "loss": 8.9705,
+      "step": 93640
+    },
+    {
+      "epoch": 0.4676770955579415,
+      "grad_norm": 0.09444907307624817,
+      "learning_rate": 1.6009662319457307e-05,
+      "loss": 8.9722,
+      "step": 93650
+    },
+    {
+      "epoch": 0.467727034382881,
+      "grad_norm": 0.09394410252571106,
+      "learning_rate": 1.600816040451576e-05,
+      "loss": 8.9806,
+      "step": 93660
+    },
+    {
+      "epoch": 0.4677769732078204,
+      "grad_norm": 0.09267649054527283,
+      "learning_rate": 1.6006658489574207e-05,
+      "loss": 8.9722,
+      "step": 93670
+    },
+    {
+      "epoch": 0.4678269120327599,
+      "grad_norm": 0.08718396723270416,
+      "learning_rate": 1.6005156574632654e-05,
+      "loss": 8.9693,
+      "step": 93680
+    },
+    {
+      "epoch": 0.4678768508576993,
+      "grad_norm": 0.08936282247304916,
+      "learning_rate": 1.6003654659691107e-05,
+      "loss": 8.9726,
+      "step": 93690
+    },
+    {
+      "epoch": 0.4679267896826388,
+      "grad_norm": 0.09415874630212784,
+      "learning_rate": 1.6002152744749554e-05,
+      "loss": 8.9622,
+      "step": 93700
+    },
+    {
+      "epoch": 0.4679767285075782,
+      "grad_norm": 0.09426844865083694,
+      "learning_rate": 1.6000650829808008e-05,
+      "loss": 8.9738,
+      "step": 93710
+    },
+    {
+      "epoch": 0.4680266673325177,
+      "grad_norm": 0.09191013872623444,
+      "learning_rate": 1.5999148914866455e-05,
+      "loss": 8.9705,
+      "step": 93720
+    },
+    {
+      "epoch": 0.4680766061574571,
+      "grad_norm": 0.09426682442426682,
+      "learning_rate": 1.5997646999924905e-05,
+      "loss": 8.9628,
+      "step": 93730
+    },
+    {
+      "epoch": 0.4681265449823966,
+      "grad_norm": 0.08494877815246582,
+      "learning_rate": 1.5996145084983355e-05,
+      "loss": 8.9671,
+      "step": 93740
+    },
+    {
+      "epoch": 0.468176483807336,
+      "grad_norm": 0.09612907469272614,
+      "learning_rate": 1.5994643170041802e-05,
+      "loss": 8.9614,
+      "step": 93750
+    },
+    {
+      "epoch": 0.4682264226322755,
+      "grad_norm": 0.09567809104919434,
+      "learning_rate": 1.5993141255100255e-05,
+      "loss": 8.9663,
+      "step": 93760
+    },
+    {
+      "epoch": 0.4682763614572149,
+      "grad_norm": 0.09354501962661743,
+      "learning_rate": 1.5991639340158702e-05,
+      "loss": 8.9696,
+      "step": 93770
+    },
+    {
+      "epoch": 0.4683263002821544,
+      "grad_norm": 0.09527242928743362,
+      "learning_rate": 1.5990137425217152e-05,
+      "loss": 8.976,
+      "step": 93780
+    },
+    {
+      "epoch": 0.4683762391070938,
+      "grad_norm": 0.09733609855175018,
+      "learning_rate": 1.5988635510275602e-05,
+      "loss": 8.9744,
+      "step": 93790
+    },
+    {
+      "epoch": 0.4684261779320333,
+      "grad_norm": 0.10153631865978241,
+      "learning_rate": 1.598713359533405e-05,
+      "loss": 8.9545,
+      "step": 93800
+    },
+    {
+      "epoch": 0.4684761167569727,
+      "grad_norm": 0.09621601551771164,
+      "learning_rate": 1.5985631680392503e-05,
+      "loss": 8.992,
+      "step": 93810
+    },
+    {
+      "epoch": 0.4685260555819122,
+      "grad_norm": 0.09544295817613602,
+      "learning_rate": 1.598412976545095e-05,
+      "loss": 8.9597,
+      "step": 93820
+    },
+    {
+      "epoch": 0.4685759944068516,
+      "grad_norm": 0.09087269753217697,
+      "learning_rate": 1.59826278505094e-05,
+      "loss": 8.9515,
+      "step": 93830
+    },
+    {
+      "epoch": 0.4686259332317911,
+      "grad_norm": 0.09846234321594238,
+      "learning_rate": 1.598112593556785e-05,
+      "loss": 8.9714,
+      "step": 93840
+    },
+    {
+      "epoch": 0.4686758720567305,
+      "grad_norm": 0.09184664487838745,
+      "learning_rate": 1.5979624020626297e-05,
+      "loss": 8.9678,
+      "step": 93850
+    },
+    {
+      "epoch": 0.46872581088167,
+      "grad_norm": 0.0921328142285347,
+      "learning_rate": 1.597812210568475e-05,
+      "loss": 8.9638,
+      "step": 93860
+    },
+    {
+      "epoch": 0.4687757497066094,
+      "grad_norm": 0.09225039184093475,
+      "learning_rate": 1.5976620190743197e-05,
+      "loss": 8.9748,
+      "step": 93870
+    },
+    {
+      "epoch": 0.4688256885315489,
+      "grad_norm": 0.09426619857549667,
+      "learning_rate": 1.5975118275801647e-05,
+      "loss": 8.957,
+      "step": 93880
+    },
+    {
+      "epoch": 0.4688756273564883,
+      "grad_norm": 0.08906444162130356,
+      "learning_rate": 1.5973616360860097e-05,
+      "loss": 8.9666,
+      "step": 93890
+    },
+    {
+      "epoch": 0.4689255661814278,
+      "grad_norm": 0.09076976776123047,
+      "learning_rate": 1.5972114445918544e-05,
+      "loss": 8.9806,
+      "step": 93900
+    },
+    {
+      "epoch": 0.4689755050063672,
+      "grad_norm": 0.09307584911584854,
+      "learning_rate": 1.5970612530976998e-05,
+      "loss": 8.9683,
+      "step": 93910
+    },
+    {
+      "epoch": 0.4690254438313067,
+      "grad_norm": 0.09109794348478317,
+      "learning_rate": 1.5969110616035445e-05,
+      "loss": 8.9724,
+      "step": 93920
+    },
+    {
+      "epoch": 0.4690753826562461,
+      "grad_norm": 0.09649874269962311,
+      "learning_rate": 1.5967608701093895e-05,
+      "loss": 8.9707,
+      "step": 93930
+    },
+    {
+      "epoch": 0.4691253214811856,
+      "grad_norm": 0.09002482146024704,
+      "learning_rate": 1.5966106786152345e-05,
+      "loss": 8.9689,
+      "step": 93940
+    },
+    {
+      "epoch": 0.469175260306125,
+      "grad_norm": 0.0938982143998146,
+      "learning_rate": 1.5964604871210792e-05,
+      "loss": 8.9652,
+      "step": 93950
+    },
+    {
+      "epoch": 0.4692251991310644,
+      "grad_norm": 0.0954759269952774,
+      "learning_rate": 1.5963102956269245e-05,
+      "loss": 8.9681,
+      "step": 93960
+    },
+    {
+      "epoch": 0.4692751379560039,
+      "grad_norm": 0.09431935846805573,
+      "learning_rate": 1.5961601041327692e-05,
+      "loss": 8.9643,
+      "step": 93970
+    },
+    {
+      "epoch": 0.4693250767809433,
+      "grad_norm": 0.10007733106613159,
+      "learning_rate": 1.5960099126386142e-05,
+      "loss": 8.9655,
+      "step": 93980
+    },
+    {
+      "epoch": 0.4693750156058828,
+      "grad_norm": 0.09566391259431839,
+      "learning_rate": 1.5958597211444592e-05,
+      "loss": 8.9486,
+      "step": 93990
+    },
+    {
+      "epoch": 0.4694249544308222,
+      "grad_norm": 0.08696243166923523,
+      "learning_rate": 1.595709529650304e-05,
+      "loss": 8.9544,
+      "step": 94000
+    },
+    {
+      "epoch": 0.4694748932557617,
+      "grad_norm": 0.0935448408126831,
+      "learning_rate": 1.5955593381561493e-05,
+      "loss": 8.9584,
+      "step": 94010
+    },
+    {
+      "epoch": 0.4695248320807011,
+      "grad_norm": 0.09453287720680237,
+      "learning_rate": 1.595409146661994e-05,
+      "loss": 8.9592,
+      "step": 94020
+    },
+    {
+      "epoch": 0.4695747709056406,
+      "grad_norm": 0.0924343541264534,
+      "learning_rate": 1.5952589551678393e-05,
+      "loss": 8.9641,
+      "step": 94030
+    },
+    {
+      "epoch": 0.46962470973058,
+      "grad_norm": 0.09673213213682175,
+      "learning_rate": 1.595108763673684e-05,
+      "loss": 8.9499,
+      "step": 94040
+    },
+    {
+      "epoch": 0.4696746485555195,
+      "grad_norm": 0.08831555396318436,
+      "learning_rate": 1.5949585721795287e-05,
+      "loss": 8.9571,
+      "step": 94050
+    },
+    {
+      "epoch": 0.4697245873804589,
+      "grad_norm": 0.09080936759710312,
+      "learning_rate": 1.594808380685374e-05,
+      "loss": 8.9697,
+      "step": 94060
+    },
+    {
+      "epoch": 0.4697745262053984,
+      "grad_norm": 0.09151973575353622,
+      "learning_rate": 1.5946581891912187e-05,
+      "loss": 8.9607,
+      "step": 94070
+    },
+    {
+      "epoch": 0.4698244650303378,
+      "grad_norm": 0.09453890472650528,
+      "learning_rate": 1.594507997697064e-05,
+      "loss": 8.9754,
+      "step": 94080
+    },
+    {
+      "epoch": 0.4698744038552773,
+      "grad_norm": 0.09043656289577484,
+      "learning_rate": 1.5943578062029087e-05,
+      "loss": 8.9796,
+      "step": 94090
+    },
+    {
+      "epoch": 0.4699243426802167,
+      "grad_norm": 0.09902900457382202,
+      "learning_rate": 1.5942076147087534e-05,
+      "loss": 8.9608,
+      "step": 94100
+    },
+    {
+      "epoch": 0.4699742815051562,
+      "grad_norm": 0.09724241495132446,
+      "learning_rate": 1.5940574232145988e-05,
+      "loss": 8.9604,
+      "step": 94110
+    },
+    {
+      "epoch": 0.4700242203300956,
+      "grad_norm": 0.09389027208089828,
+      "learning_rate": 1.5939072317204435e-05,
+      "loss": 8.9627,
+      "step": 94120
+    },
+    {
+      "epoch": 0.4700741591550351,
+      "grad_norm": 0.08884572237730026,
+      "learning_rate": 1.5937570402262888e-05,
+      "loss": 8.9623,
+      "step": 94130
+    },
+    {
+      "epoch": 0.4701240979799745,
+      "grad_norm": 0.09251714497804642,
+      "learning_rate": 1.5936068487321335e-05,
+      "loss": 8.9787,
+      "step": 94140
+    },
+    {
+      "epoch": 0.470174036804914,
+      "grad_norm": 0.09012773633003235,
+      "learning_rate": 1.5934566572379785e-05,
+      "loss": 8.9716,
+      "step": 94150
+    },
+    {
+      "epoch": 0.4702239756298534,
+      "grad_norm": 0.09131991118192673,
+      "learning_rate": 1.5933064657438235e-05,
+      "loss": 8.9629,
+      "step": 94160
+    },
+    {
+      "epoch": 0.4702739144547929,
+      "grad_norm": 0.09043431282043457,
+      "learning_rate": 1.5931562742496682e-05,
+      "loss": 8.9654,
+      "step": 94170
+    },
+    {
+      "epoch": 0.4703238532797323,
+      "grad_norm": 0.09283377230167389,
+      "learning_rate": 1.5930060827555136e-05,
+      "loss": 8.9507,
+      "step": 94180
+    },
+    {
+      "epoch": 0.4703737921046718,
+      "grad_norm": 0.09180030971765518,
+      "learning_rate": 1.5928558912613582e-05,
+      "loss": 8.9708,
+      "step": 94190
+    },
+    {
+      "epoch": 0.4704237309296112,
+      "grad_norm": 0.08900587260723114,
+      "learning_rate": 1.5927056997672033e-05,
+      "loss": 8.9548,
+      "step": 94200
+    },
+    {
+      "epoch": 0.4704736697545507,
+      "grad_norm": 0.10260471701622009,
+      "learning_rate": 1.5925555082730483e-05,
+      "loss": 8.9641,
+      "step": 94210
+    },
+    {
+      "epoch": 0.4705236085794901,
+      "grad_norm": 0.09451324492692947,
+      "learning_rate": 1.592405316778893e-05,
+      "loss": 8.9697,
+      "step": 94220
+    },
+    {
+      "epoch": 0.4705735474044296,
+      "grad_norm": 0.1005653440952301,
+      "learning_rate": 1.5922551252847383e-05,
+      "loss": 8.9598,
+      "step": 94230
+    },
+    {
+      "epoch": 0.470623486229369,
+      "grad_norm": 0.09918899834156036,
+      "learning_rate": 1.592104933790583e-05,
+      "loss": 8.9575,
+      "step": 94240
+    },
+    {
+      "epoch": 0.4706734250543085,
+      "grad_norm": 0.09747923910617828,
+      "learning_rate": 1.591954742296428e-05,
+      "loss": 8.9644,
+      "step": 94250
+    },
+    {
+      "epoch": 0.4707233638792479,
+      "grad_norm": 0.09829793125391006,
+      "learning_rate": 1.591804550802273e-05,
+      "loss": 8.9704,
+      "step": 94260
+    },
+    {
+      "epoch": 0.4707733027041874,
+      "grad_norm": 0.10006602108478546,
+      "learning_rate": 1.5916543593081177e-05,
+      "loss": 8.9625,
+      "step": 94270
+    },
+    {
+      "epoch": 0.4708232415291268,
+      "grad_norm": 0.09519977867603302,
+      "learning_rate": 1.591504167813963e-05,
+      "loss": 8.9572,
+      "step": 94280
+    },
+    {
+      "epoch": 0.4708731803540663,
+      "grad_norm": 0.09414579719305038,
+      "learning_rate": 1.5913539763198077e-05,
+      "loss": 8.9618,
+      "step": 94290
+    },
+    {
+      "epoch": 0.4709231191790057,
+      "grad_norm": 0.09214229881763458,
+      "learning_rate": 1.5912037848256528e-05,
+      "loss": 8.968,
+      "step": 94300
+    },
+    {
+      "epoch": 0.4709730580039452,
+      "grad_norm": 0.09208973497152328,
+      "learning_rate": 1.5910535933314978e-05,
+      "loss": 8.9711,
+      "step": 94310
+    },
+    {
+      "epoch": 0.4710229968288846,
+      "grad_norm": 0.09273689985275269,
+      "learning_rate": 1.5909034018373425e-05,
+      "loss": 8.9477,
+      "step": 94320
+    },
+    {
+      "epoch": 0.4710729356538241,
+      "grad_norm": 0.09345366805791855,
+      "learning_rate": 1.5907532103431878e-05,
+      "loss": 8.9459,
+      "step": 94330
+    },
+    {
+      "epoch": 0.4711228744787635,
+      "grad_norm": 0.09309998899698257,
+      "learning_rate": 1.5906030188490325e-05,
+      "loss": 8.9669,
+      "step": 94340
+    },
+    {
+      "epoch": 0.471172813303703,
+      "grad_norm": 0.0877714604139328,
+      "learning_rate": 1.5904528273548775e-05,
+      "loss": 8.967,
+      "step": 94350
+    },
+    {
+      "epoch": 0.4712227521286424,
+      "grad_norm": 0.09679785370826721,
+      "learning_rate": 1.5903026358607225e-05,
+      "loss": 8.9599,
+      "step": 94360
+    },
+    {
+      "epoch": 0.4712726909535819,
+      "grad_norm": 0.10014106333255768,
+      "learning_rate": 1.5901524443665672e-05,
+      "loss": 8.9567,
+      "step": 94370
+    },
+    {
+      "epoch": 0.4713226297785213,
+      "grad_norm": 0.09142360091209412,
+      "learning_rate": 1.5900022528724126e-05,
+      "loss": 8.9641,
+      "step": 94380
+    },
+    {
+      "epoch": 0.4713725686034608,
+      "grad_norm": 0.09642203897237778,
+      "learning_rate": 1.5898520613782573e-05,
+      "loss": 8.9582,
+      "step": 94390
+    },
+    {
+      "epoch": 0.4714225074284002,
+      "grad_norm": 0.0904778316617012,
+      "learning_rate": 1.5897018698841023e-05,
+      "loss": 8.9714,
+      "step": 94400
+    },
+    {
+      "epoch": 0.47147244625333967,
+      "grad_norm": 0.09670837968587875,
+      "learning_rate": 1.5895516783899473e-05,
+      "loss": 8.9466,
+      "step": 94410
+    },
+    {
+      "epoch": 0.4715223850782791,
+      "grad_norm": 0.09427211433649063,
+      "learning_rate": 1.589401486895792e-05,
+      "loss": 8.9493,
+      "step": 94420
+    },
+    {
+      "epoch": 0.47157232390321857,
+      "grad_norm": 0.09509798139333725,
+      "learning_rate": 1.5892512954016373e-05,
+      "loss": 8.9536,
+      "step": 94430
+    },
+    {
+      "epoch": 0.471622262728158,
+      "grad_norm": 0.09237992763519287,
+      "learning_rate": 1.589101103907482e-05,
+      "loss": 8.9698,
+      "step": 94440
+    },
+    {
+      "epoch": 0.47167220155309747,
+      "grad_norm": 0.09039132297039032,
+      "learning_rate": 1.588950912413327e-05,
+      "loss": 8.96,
+      "step": 94450
+    },
+    {
+      "epoch": 0.4717221403780369,
+      "grad_norm": 0.09362545609474182,
+      "learning_rate": 1.588800720919172e-05,
+      "loss": 8.9552,
+      "step": 94460
+    },
+    {
+      "epoch": 0.47177207920297637,
+      "grad_norm": 0.09487809240818024,
+      "learning_rate": 1.588650529425017e-05,
+      "loss": 8.9529,
+      "step": 94470
+    },
+    {
+      "epoch": 0.4718220180279158,
+      "grad_norm": 0.09452398866415024,
+      "learning_rate": 1.588500337930862e-05,
+      "loss": 8.9601,
+      "step": 94480
+    },
+    {
+      "epoch": 0.47187195685285527,
+      "grad_norm": 0.08807900547981262,
+      "learning_rate": 1.5883501464367068e-05,
+      "loss": 8.9602,
+      "step": 94490
+    },
+    {
+      "epoch": 0.4719218956777947,
+      "grad_norm": 0.09054072201251984,
+      "learning_rate": 1.5881999549425518e-05,
+      "loss": 8.9705,
+      "step": 94500
+    },
+    {
+      "epoch": 0.47197183450273417,
+      "grad_norm": 0.0914943590760231,
+      "learning_rate": 1.5880497634483968e-05,
+      "loss": 8.9667,
+      "step": 94510
+    },
+    {
+      "epoch": 0.4720217733276736,
+      "grad_norm": 0.09846024960279465,
+      "learning_rate": 1.5878995719542418e-05,
+      "loss": 8.964,
+      "step": 94520
+    },
+    {
+      "epoch": 0.47207171215261307,
+      "grad_norm": 0.08973458409309387,
+      "learning_rate": 1.5877493804600868e-05,
+      "loss": 8.9691,
+      "step": 94530
+    },
+    {
+      "epoch": 0.4721216509775525,
+      "grad_norm": 0.09380858391523361,
+      "learning_rate": 1.5875991889659315e-05,
+      "loss": 8.9595,
+      "step": 94540
+    },
+    {
+      "epoch": 0.47217158980249196,
+      "grad_norm": 0.09617038071155548,
+      "learning_rate": 1.5874489974717765e-05,
+      "loss": 8.9691,
+      "step": 94550
+    },
+    {
+      "epoch": 0.4722215286274314,
+      "grad_norm": 0.09885700792074203,
+      "learning_rate": 1.5872988059776215e-05,
+      "loss": 8.9462,
+      "step": 94560
+    },
+    {
+      "epoch": 0.47227146745237086,
+      "grad_norm": 0.08958373963832855,
+      "learning_rate": 1.5871486144834666e-05,
+      "loss": 8.9666,
+      "step": 94570
+    },
+    {
+      "epoch": 0.4723214062773103,
+      "grad_norm": 0.09378182142972946,
+      "learning_rate": 1.5869984229893116e-05,
+      "loss": 8.9554,
+      "step": 94580
+    },
+    {
+      "epoch": 0.47237134510224976,
+      "grad_norm": 0.09632568061351776,
+      "learning_rate": 1.5868482314951563e-05,
+      "loss": 8.9563,
+      "step": 94590
+    },
+    {
+      "epoch": 0.4724212839271892,
+      "grad_norm": 0.09778321534395218,
+      "learning_rate": 1.5866980400010013e-05,
+      "loss": 8.9532,
+      "step": 94600
+    },
+    {
+      "epoch": 0.47247122275212866,
+      "grad_norm": 0.08909459412097931,
+      "learning_rate": 1.5865478485068463e-05,
+      "loss": 8.9759,
+      "step": 94610
+    },
+    {
+      "epoch": 0.4725211615770681,
+      "grad_norm": 0.09531407058238983,
+      "learning_rate": 1.5863976570126913e-05,
+      "loss": 8.9577,
+      "step": 94620
+    },
+    {
+      "epoch": 0.47257110040200756,
+      "grad_norm": 0.09269222617149353,
+      "learning_rate": 1.5862474655185363e-05,
+      "loss": 8.959,
+      "step": 94630
+    },
+    {
+      "epoch": 0.472621039226947,
+      "grad_norm": 0.09362973272800446,
+      "learning_rate": 1.586097274024381e-05,
+      "loss": 8.953,
+      "step": 94640
+    },
+    {
+      "epoch": 0.47267097805188646,
+      "grad_norm": 0.09257864207029343,
+      "learning_rate": 1.585947082530226e-05,
+      "loss": 8.9646,
+      "step": 94650
+    },
+    {
+      "epoch": 0.4727209168768259,
+      "grad_norm": 0.10312341898679733,
+      "learning_rate": 1.585796891036071e-05,
+      "loss": 8.9662,
+      "step": 94660
+    },
+    {
+      "epoch": 0.47277085570176536,
+      "grad_norm": 0.09162463247776031,
+      "learning_rate": 1.585646699541916e-05,
+      "loss": 8.9528,
+      "step": 94670
+    },
+    {
+      "epoch": 0.4728207945267048,
+      "grad_norm": 0.09410829097032547,
+      "learning_rate": 1.585496508047761e-05,
+      "loss": 8.9487,
+      "step": 94680
+    },
+    {
+      "epoch": 0.47287073335164426,
+      "grad_norm": 0.09216289222240448,
+      "learning_rate": 1.5853463165536058e-05,
+      "loss": 8.9574,
+      "step": 94690
+    },
+    {
+      "epoch": 0.4729206721765837,
+      "grad_norm": 0.09374202787876129,
+      "learning_rate": 1.5851961250594508e-05,
+      "loss": 8.9629,
+      "step": 94700
+    },
+    {
+      "epoch": 0.47297061100152316,
+      "grad_norm": 0.09870696812868118,
+      "learning_rate": 1.5850459335652958e-05,
+      "loss": 8.9696,
+      "step": 94710
+    },
+    {
+      "epoch": 0.4730205498264626,
+      "grad_norm": 0.0899372324347496,
+      "learning_rate": 1.5848957420711408e-05,
+      "loss": 8.9589,
+      "step": 94720
+    },
+    {
+      "epoch": 0.47307048865140205,
+      "grad_norm": 0.0908857062458992,
+      "learning_rate": 1.5847455505769858e-05,
+      "loss": 8.9605,
+      "step": 94730
+    },
+    {
+      "epoch": 0.4731204274763415,
+      "grad_norm": 0.08634936064481735,
+      "learning_rate": 1.5845953590828305e-05,
+      "loss": 8.9659,
+      "step": 94740
+    },
+    {
+      "epoch": 0.47317036630128095,
+      "grad_norm": 0.09862346947193146,
+      "learning_rate": 1.5844451675886755e-05,
+      "loss": 8.9518,
+      "step": 94750
+    },
+    {
+      "epoch": 0.4732203051262204,
+      "grad_norm": 0.09859306365251541,
+      "learning_rate": 1.5842949760945205e-05,
+      "loss": 8.9509,
+      "step": 94760
+    },
+    {
+      "epoch": 0.47327024395115985,
+      "grad_norm": 0.09249970316886902,
+      "learning_rate": 1.5841447846003656e-05,
+      "loss": 8.9593,
+      "step": 94770
+    },
+    {
+      "epoch": 0.4733201827760993,
+      "grad_norm": 0.09957744181156158,
+      "learning_rate": 1.5839945931062106e-05,
+      "loss": 8.9597,
+      "step": 94780
+    },
+    {
+      "epoch": 0.47337012160103875,
+      "grad_norm": 0.0913504958152771,
+      "learning_rate": 1.5838444016120556e-05,
+      "loss": 8.968,
+      "step": 94790
+    },
+    {
+      "epoch": 0.4734200604259782,
+      "grad_norm": 0.09559644013643265,
+      "learning_rate": 1.5836942101179003e-05,
+      "loss": 8.9594,
+      "step": 94800
+    },
+    {
+      "epoch": 0.47346999925091765,
+      "grad_norm": 0.09303098171949387,
+      "learning_rate": 1.5835440186237453e-05,
+      "loss": 8.9535,
+      "step": 94810
+    },
+    {
+      "epoch": 0.47351993807585707,
+      "grad_norm": 0.09096693992614746,
+      "learning_rate": 1.5833938271295903e-05,
+      "loss": 8.9503,
+      "step": 94820
+    },
+    {
+      "epoch": 0.47356987690079655,
+      "grad_norm": 0.09463432431221008,
+      "learning_rate": 1.5832436356354353e-05,
+      "loss": 8.9561,
+      "step": 94830
+    },
+    {
+      "epoch": 0.47361981572573597,
+      "grad_norm": 0.09306497126817703,
+      "learning_rate": 1.5830934441412803e-05,
+      "loss": 8.9495,
+      "step": 94840
+    },
+    {
+      "epoch": 0.47366975455067545,
+      "grad_norm": 0.09330885857343674,
+      "learning_rate": 1.582943252647125e-05,
+      "loss": 8.9607,
+      "step": 94850
+    },
+    {
+      "epoch": 0.47371969337561487,
+      "grad_norm": 0.0966872051358223,
+      "learning_rate": 1.58279306115297e-05,
+      "loss": 8.9557,
+      "step": 94860
+    },
+    {
+      "epoch": 0.47376963220055435,
+      "grad_norm": 0.08871348202228546,
+      "learning_rate": 1.582642869658815e-05,
+      "loss": 8.9677,
+      "step": 94870
+    },
+    {
+      "epoch": 0.47381957102549377,
+      "grad_norm": 0.090989850461483,
+      "learning_rate": 1.58249267816466e-05,
+      "loss": 8.9661,
+      "step": 94880
+    },
+    {
+      "epoch": 0.47386950985043325,
+      "grad_norm": 0.0886576771736145,
+      "learning_rate": 1.582342486670505e-05,
+      "loss": 8.9559,
+      "step": 94890
+    },
+    {
+      "epoch": 0.47391944867537267,
+      "grad_norm": 0.09537599235773087,
+      "learning_rate": 1.5821922951763498e-05,
+      "loss": 8.9496,
+      "step": 94900
+    },
+    {
+      "epoch": 0.47396938750031214,
+      "grad_norm": 0.08645906299352646,
+      "learning_rate": 1.5820421036821948e-05,
+      "loss": 8.9599,
+      "step": 94910
+    },
+    {
+      "epoch": 0.47401932632525157,
+      "grad_norm": 0.0953529104590416,
+      "learning_rate": 1.5818919121880398e-05,
+      "loss": 8.9579,
+      "step": 94920
+    },
+    {
+      "epoch": 0.47406926515019104,
+      "grad_norm": 0.10060861706733704,
+      "learning_rate": 1.5817417206938848e-05,
+      "loss": 8.9599,
+      "step": 94930
+    },
+    {
+      "epoch": 0.47411920397513047,
+      "grad_norm": 0.09143946319818497,
+      "learning_rate": 1.58159152919973e-05,
+      "loss": 8.9482,
+      "step": 94940
+    },
+    {
+      "epoch": 0.4741691428000699,
+      "grad_norm": 0.09554131329059601,
+      "learning_rate": 1.5814413377055745e-05,
+      "loss": 8.9516,
+      "step": 94950
+    },
+    {
+      "epoch": 0.47421908162500936,
+      "grad_norm": 0.09513738751411438,
+      "learning_rate": 1.5812911462114195e-05,
+      "loss": 8.9709,
+      "step": 94960
+    },
+    {
+      "epoch": 0.4742690204499488,
+      "grad_norm": 0.09006723016500473,
+      "learning_rate": 1.5811409547172646e-05,
+      "loss": 8.9641,
+      "step": 94970
+    },
+    {
+      "epoch": 0.47431895927488826,
+      "grad_norm": 0.08737257868051529,
+      "learning_rate": 1.5809907632231096e-05,
+      "loss": 8.9553,
+      "step": 94980
+    },
+    {
+      "epoch": 0.4743688980998277,
+      "grad_norm": 0.09388820081949234,
+      "learning_rate": 1.5808405717289546e-05,
+      "loss": 8.9634,
+      "step": 94990
+    },
+    {
+      "epoch": 0.47441883692476716,
+      "grad_norm": 0.0908634141087532,
+      "learning_rate": 1.5806903802347993e-05,
+      "loss": 8.954,
+      "step": 95000
+    },
+    {
+      "epoch": 0.4744687757497066,
+      "grad_norm": 0.09147128462791443,
+      "learning_rate": 1.5805401887406443e-05,
+      "loss": 8.9601,
+      "step": 95010
+    },
+    {
+      "epoch": 0.47451871457464606,
+      "grad_norm": 0.09336107224225998,
+      "learning_rate": 1.5803899972464893e-05,
+      "loss": 8.9496,
+      "step": 95020
+    },
+    {
+      "epoch": 0.4745686533995855,
+      "grad_norm": 0.09616447985172272,
+      "learning_rate": 1.5802398057523343e-05,
+      "loss": 8.9433,
+      "step": 95030
+    },
+    {
+      "epoch": 0.47461859222452496,
+      "grad_norm": 0.09238088130950928,
+      "learning_rate": 1.5800896142581793e-05,
+      "loss": 8.9607,
+      "step": 95040
+    },
+    {
+      "epoch": 0.4746685310494644,
+      "grad_norm": 0.09115064889192581,
+      "learning_rate": 1.579939422764024e-05,
+      "loss": 8.9525,
+      "step": 95050
+    },
+    {
+      "epoch": 0.47471846987440386,
+      "grad_norm": 0.09081438928842545,
+      "learning_rate": 1.579789231269869e-05,
+      "loss": 8.9477,
+      "step": 95060
+    },
+    {
+      "epoch": 0.4747684086993433,
+      "grad_norm": 0.09441505372524261,
+      "learning_rate": 1.579639039775714e-05,
+      "loss": 8.9469,
+      "step": 95070
+    },
+    {
+      "epoch": 0.47481834752428276,
+      "grad_norm": 0.0935746282339096,
+      "learning_rate": 1.579488848281559e-05,
+      "loss": 8.978,
+      "step": 95080
+    },
+    {
+      "epoch": 0.4748682863492222,
+      "grad_norm": 0.09515486657619476,
+      "learning_rate": 1.579338656787404e-05,
+      "loss": 8.9698,
+      "step": 95090
+    },
+    {
+      "epoch": 0.47491822517416166,
+      "grad_norm": 0.10014467686414719,
+      "learning_rate": 1.5791884652932488e-05,
+      "loss": 8.9462,
+      "step": 95100
+    },
+    {
+      "epoch": 0.4749681639991011,
+      "grad_norm": 0.09181597828865051,
+      "learning_rate": 1.579038273799094e-05,
+      "loss": 8.9549,
+      "step": 95110
+    },
+    {
+      "epoch": 0.47501810282404056,
+      "grad_norm": 0.09579970687627792,
+      "learning_rate": 1.5788880823049388e-05,
+      "loss": 8.9546,
+      "step": 95120
+    },
+    {
+      "epoch": 0.47506804164898,
+      "grad_norm": 0.0962633490562439,
+      "learning_rate": 1.5787378908107838e-05,
+      "loss": 8.9442,
+      "step": 95130
+    },
+    {
+      "epoch": 0.47511798047391945,
+      "grad_norm": 0.09618624299764633,
+      "learning_rate": 1.578587699316629e-05,
+      "loss": 8.9584,
+      "step": 95140
+    },
+    {
+      "epoch": 0.4751679192988589,
+      "grad_norm": 0.08641847223043442,
+      "learning_rate": 1.5784375078224735e-05,
+      "loss": 8.9636,
+      "step": 95150
+    },
+    {
+      "epoch": 0.47521785812379835,
+      "grad_norm": 0.0992530807852745,
+      "learning_rate": 1.578287316328319e-05,
+      "loss": 8.9572,
+      "step": 95160
+    },
+    {
+      "epoch": 0.4752677969487378,
+      "grad_norm": 0.09517235308885574,
+      "learning_rate": 1.5781371248341636e-05,
+      "loss": 8.9476,
+      "step": 95170
+    },
+    {
+      "epoch": 0.47531773577367725,
+      "grad_norm": 0.09363456070423126,
+      "learning_rate": 1.5779869333400086e-05,
+      "loss": 8.9558,
+      "step": 95180
+    },
+    {
+      "epoch": 0.4753676745986167,
+      "grad_norm": 0.0897691547870636,
+      "learning_rate": 1.5778367418458536e-05,
+      "loss": 8.9506,
+      "step": 95190
+    },
+    {
+      "epoch": 0.47541761342355615,
+      "grad_norm": 0.09347530454397202,
+      "learning_rate": 1.5776865503516983e-05,
+      "loss": 8.9519,
+      "step": 95200
+    },
+    {
+      "epoch": 0.4754675522484956,
+      "grad_norm": 0.09927235543727875,
+      "learning_rate": 1.5775363588575436e-05,
+      "loss": 8.9348,
+      "step": 95210
+    },
+    {
+      "epoch": 0.47551749107343505,
+      "grad_norm": 0.09206566959619522,
+      "learning_rate": 1.5773861673633883e-05,
+      "loss": 8.9514,
+      "step": 95220
+    },
+    {
+      "epoch": 0.47556742989837447,
+      "grad_norm": 0.09086683392524719,
+      "learning_rate": 1.5772359758692333e-05,
+      "loss": 8.9393,
+      "step": 95230
+    },
+    {
+      "epoch": 0.47561736872331395,
+      "grad_norm": 0.0883910059928894,
+      "learning_rate": 1.5770857843750783e-05,
+      "loss": 8.9475,
+      "step": 95240
+    },
+    {
+      "epoch": 0.47566730754825337,
+      "grad_norm": 0.09318321943283081,
+      "learning_rate": 1.576935592880923e-05,
+      "loss": 8.9376,
+      "step": 95250
+    },
+    {
+      "epoch": 0.47571724637319285,
+      "grad_norm": 0.09458793699741364,
+      "learning_rate": 1.5767854013867684e-05,
+      "loss": 8.9374,
+      "step": 95260
+    },
+    {
+      "epoch": 0.47576718519813227,
+      "grad_norm": 0.09189481288194656,
+      "learning_rate": 1.576635209892613e-05,
+      "loss": 8.946,
+      "step": 95270
+    },
+    {
+      "epoch": 0.47581712402307175,
+      "grad_norm": 0.09554219245910645,
+      "learning_rate": 1.576485018398458e-05,
+      "loss": 8.9526,
+      "step": 95280
+    },
+    {
+      "epoch": 0.47586706284801117,
+      "grad_norm": 0.09450355917215347,
+      "learning_rate": 1.576334826904303e-05,
+      "loss": 8.957,
+      "step": 95290
+    },
+    {
+      "epoch": 0.47591700167295065,
+      "grad_norm": 0.08791901171207428,
+      "learning_rate": 1.5761846354101478e-05,
+      "loss": 8.9516,
+      "step": 95300
+    },
+    {
+      "epoch": 0.47596694049789007,
+      "grad_norm": 0.08911871910095215,
+      "learning_rate": 1.576034443915993e-05,
+      "loss": 8.9511,
+      "step": 95310
+    },
+    {
+      "epoch": 0.47601687932282954,
+      "grad_norm": 0.09467475861310959,
+      "learning_rate": 1.5758842524218378e-05,
+      "loss": 8.9439,
+      "step": 95320
+    },
+    {
+      "epoch": 0.47606681814776897,
+      "grad_norm": 0.09240403026342392,
+      "learning_rate": 1.5757340609276828e-05,
+      "loss": 8.9443,
+      "step": 95330
+    },
+    {
+      "epoch": 0.47611675697270844,
+      "grad_norm": 0.09284543991088867,
+      "learning_rate": 1.575583869433528e-05,
+      "loss": 8.9489,
+      "step": 95340
+    },
+    {
+      "epoch": 0.47616669579764787,
+      "grad_norm": 0.0898863673210144,
+      "learning_rate": 1.5754336779393725e-05,
+      "loss": 8.9601,
+      "step": 95350
+    },
+    {
+      "epoch": 0.47621663462258734,
+      "grad_norm": 0.09170831739902496,
+      "learning_rate": 1.575283486445218e-05,
+      "loss": 8.9568,
+      "step": 95360
+    },
+    {
+      "epoch": 0.47626657344752676,
+      "grad_norm": 0.0939890593290329,
+      "learning_rate": 1.5751332949510626e-05,
+      "loss": 8.9392,
+      "step": 95370
+    },
+    {
+      "epoch": 0.47631651227246624,
+      "grad_norm": 0.08675019443035126,
+      "learning_rate": 1.5749831034569076e-05,
+      "loss": 8.9485,
+      "step": 95380
+    },
+    {
+      "epoch": 0.47636645109740566,
+      "grad_norm": 0.08740539848804474,
+      "learning_rate": 1.5748329119627526e-05,
+      "loss": 8.9461,
+      "step": 95390
+    },
+    {
+      "epoch": 0.47641638992234514,
+      "grad_norm": 0.08891706168651581,
+      "learning_rate": 1.5746827204685973e-05,
+      "loss": 8.9638,
+      "step": 95400
+    },
+    {
+      "epoch": 0.47646632874728456,
+      "grad_norm": 0.09209989011287689,
+      "learning_rate": 1.5745325289744426e-05,
+      "loss": 8.9543,
+      "step": 95410
+    },
+    {
+      "epoch": 0.47651626757222404,
+      "grad_norm": 0.09303548187017441,
+      "learning_rate": 1.5743823374802873e-05,
+      "loss": 8.9472,
+      "step": 95420
+    },
+    {
+      "epoch": 0.47656620639716346,
+      "grad_norm": 0.09217501431703568,
+      "learning_rate": 1.5742321459861327e-05,
+      "loss": 8.9547,
+      "step": 95430
+    },
+    {
+      "epoch": 0.47661614522210294,
+      "grad_norm": 0.08850811421871185,
+      "learning_rate": 1.5740819544919773e-05,
+      "loss": 8.9523,
+      "step": 95440
+    },
+    {
+      "epoch": 0.47666608404704236,
+      "grad_norm": 0.09346584975719452,
+      "learning_rate": 1.573931762997822e-05,
+      "loss": 8.9549,
+      "step": 95450
+    },
+    {
+      "epoch": 0.47671602287198184,
+      "grad_norm": 0.09290006756782532,
+      "learning_rate": 1.5737815715036674e-05,
+      "loss": 8.9536,
+      "step": 95460
+    },
+    {
+      "epoch": 0.47676596169692126,
+      "grad_norm": 0.09277894347906113,
+      "learning_rate": 1.573631380009512e-05,
+      "loss": 8.9525,
+      "step": 95470
+    },
+    {
+      "epoch": 0.47681590052186074,
+      "grad_norm": 0.09250402450561523,
+      "learning_rate": 1.5734811885153574e-05,
+      "loss": 8.9305,
+      "step": 95480
+    },
+    {
+      "epoch": 0.47686583934680016,
+      "grad_norm": 0.08638934046030045,
+      "learning_rate": 1.573330997021202e-05,
+      "loss": 8.949,
+      "step": 95490
+    },
+    {
+      "epoch": 0.47691577817173963,
+      "grad_norm": 0.10017383098602295,
+      "learning_rate": 1.5731808055270468e-05,
+      "loss": 8.9447,
+      "step": 95500
+    },
+    {
+      "epoch": 0.47696571699667906,
+      "grad_norm": 0.09218341112136841,
+      "learning_rate": 1.573030614032892e-05,
+      "loss": 8.9477,
+      "step": 95510
+    },
+    {
+      "epoch": 0.47701565582161853,
+      "grad_norm": 0.09426803141832352,
+      "learning_rate": 1.5728804225387368e-05,
+      "loss": 8.9439,
+      "step": 95520
+    },
+    {
+      "epoch": 0.47706559464655796,
+      "grad_norm": 0.08649339526891708,
+      "learning_rate": 1.572730231044582e-05,
+      "loss": 8.9515,
+      "step": 95530
+    },
+    {
+      "epoch": 0.47711553347149743,
+      "grad_norm": 0.09477028250694275,
+      "learning_rate": 1.572580039550427e-05,
+      "loss": 8.958,
+      "step": 95540
+    },
+    {
+      "epoch": 0.47716547229643685,
+      "grad_norm": 0.09427261352539062,
+      "learning_rate": 1.5724298480562715e-05,
+      "loss": 8.9478,
+      "step": 95550
+    },
+    {
+      "epoch": 0.47721541112137633,
+      "grad_norm": 0.09864936023950577,
+      "learning_rate": 1.572279656562117e-05,
+      "loss": 8.9301,
+      "step": 95560
+    },
+    {
+      "epoch": 0.47726534994631575,
+      "grad_norm": 0.09512222558259964,
+      "learning_rate": 1.5721294650679616e-05,
+      "loss": 8.9521,
+      "step": 95570
+    },
+    {
+      "epoch": 0.47731528877125523,
+      "grad_norm": 0.09052599221467972,
+      "learning_rate": 1.571979273573807e-05,
+      "loss": 8.949,
+      "step": 95580
+    },
+    {
+      "epoch": 0.47736522759619465,
+      "grad_norm": 0.09813720732927322,
+      "learning_rate": 1.5718290820796516e-05,
+      "loss": 8.9293,
+      "step": 95590
+    },
+    {
+      "epoch": 0.47741516642113413,
+      "grad_norm": 0.09309673309326172,
+      "learning_rate": 1.5716788905854963e-05,
+      "loss": 8.9349,
+      "step": 95600
+    },
+    {
+      "epoch": 0.47746510524607355,
+      "grad_norm": 0.08683782070875168,
+      "learning_rate": 1.5715286990913416e-05,
+      "loss": 8.9493,
+      "step": 95610
+    },
+    {
+      "epoch": 0.47751504407101303,
+      "grad_norm": 0.09030024707317352,
+      "learning_rate": 1.5713785075971863e-05,
+      "loss": 8.9459,
+      "step": 95620
+    },
+    {
+      "epoch": 0.47756498289595245,
+      "grad_norm": 0.09140963852405548,
+      "learning_rate": 1.5712283161030317e-05,
+      "loss": 8.9456,
+      "step": 95630
+    },
+    {
+      "epoch": 0.4776149217208919,
+      "grad_norm": 0.09844744205474854,
+      "learning_rate": 1.5710781246088763e-05,
+      "loss": 8.9562,
+      "step": 95640
+    },
+    {
+      "epoch": 0.47766486054583135,
+      "grad_norm": 0.09246145933866501,
+      "learning_rate": 1.570927933114721e-05,
+      "loss": 8.9507,
+      "step": 95650
+    },
+    {
+      "epoch": 0.4777147993707708,
+      "grad_norm": 0.09520356357097626,
+      "learning_rate": 1.5707777416205664e-05,
+      "loss": 8.9419,
+      "step": 95660
+    },
+    {
+      "epoch": 0.47776473819571025,
+      "grad_norm": 0.09316331148147583,
+      "learning_rate": 1.570627550126411e-05,
+      "loss": 8.9404,
+      "step": 95670
+    },
+    {
+      "epoch": 0.4778146770206497,
+      "grad_norm": 0.09663616865873337,
+      "learning_rate": 1.5704773586322564e-05,
+      "loss": 8.9278,
+      "step": 95680
+    },
+    {
+      "epoch": 0.47786461584558915,
+      "grad_norm": 0.09337840229272842,
+      "learning_rate": 1.570327167138101e-05,
+      "loss": 8.9441,
+      "step": 95690
+    },
+    {
+      "epoch": 0.4779145546705286,
+      "grad_norm": 0.0952988788485527,
+      "learning_rate": 1.5701769756439458e-05,
+      "loss": 8.9642,
+      "step": 95700
+    },
+    {
+      "epoch": 0.47796449349546805,
+      "grad_norm": 0.09496844559907913,
+      "learning_rate": 1.570026784149791e-05,
+      "loss": 8.9474,
+      "step": 95710
+    },
+    {
+      "epoch": 0.4780144323204075,
+      "grad_norm": 0.09619852900505066,
+      "learning_rate": 1.5698765926556358e-05,
+      "loss": 8.9422,
+      "step": 95720
+    },
+    {
+      "epoch": 0.47806437114534694,
+      "grad_norm": 0.10100951045751572,
+      "learning_rate": 1.5697264011614812e-05,
+      "loss": 8.9483,
+      "step": 95730
+    },
+    {
+      "epoch": 0.4781143099702864,
+      "grad_norm": 0.08543970435857773,
+      "learning_rate": 1.569576209667326e-05,
+      "loss": 8.9605,
+      "step": 95740
+    },
+    {
+      "epoch": 0.47816424879522584,
+      "grad_norm": 0.09065885096788406,
+      "learning_rate": 1.569426018173171e-05,
+      "loss": 8.9482,
+      "step": 95750
+    },
+    {
+      "epoch": 0.4782141876201653,
+      "grad_norm": 0.09514321386814117,
+      "learning_rate": 1.569275826679016e-05,
+      "loss": 8.9535,
+      "step": 95760
+    },
+    {
+      "epoch": 0.47826412644510474,
+      "grad_norm": 0.08945092558860779,
+      "learning_rate": 1.5691256351848606e-05,
+      "loss": 8.9462,
+      "step": 95770
+    },
+    {
+      "epoch": 0.4783140652700442,
+      "grad_norm": 0.09275572746992111,
+      "learning_rate": 1.568975443690706e-05,
+      "loss": 8.9564,
+      "step": 95780
+    },
+    {
+      "epoch": 0.47836400409498364,
+      "grad_norm": 0.09393645077943802,
+      "learning_rate": 1.5688252521965506e-05,
+      "loss": 8.9396,
+      "step": 95790
+    },
+    {
+      "epoch": 0.4784139429199231,
+      "grad_norm": 0.08892881125211716,
+      "learning_rate": 1.5686750607023956e-05,
+      "loss": 8.9488,
+      "step": 95800
+    },
+    {
+      "epoch": 0.47846388174486254,
+      "grad_norm": 0.09092603623867035,
+      "learning_rate": 1.5685248692082406e-05,
+      "loss": 8.9466,
+      "step": 95810
+    },
+    {
+      "epoch": 0.478513820569802,
+      "grad_norm": 0.09448976814746857,
+      "learning_rate": 1.5683746777140853e-05,
+      "loss": 8.9469,
+      "step": 95820
+    },
+    {
+      "epoch": 0.47856375939474144,
+      "grad_norm": 0.0917850136756897,
+      "learning_rate": 1.5682244862199307e-05,
+      "loss": 8.9422,
+      "step": 95830
+    },
+    {
+      "epoch": 0.4786136982196809,
+      "grad_norm": 0.09401079267263412,
+      "learning_rate": 1.5680742947257754e-05,
+      "loss": 8.9421,
+      "step": 95840
+    },
+    {
+      "epoch": 0.47866363704462034,
+      "grad_norm": 0.09111663699150085,
+      "learning_rate": 1.5679241032316204e-05,
+      "loss": 8.9468,
+      "step": 95850
+    },
+    {
+      "epoch": 0.4787135758695598,
+      "grad_norm": 0.09453728049993515,
+      "learning_rate": 1.5677739117374654e-05,
+      "loss": 8.962,
+      "step": 95860
+    },
+    {
+      "epoch": 0.47876351469449924,
+      "grad_norm": 0.09403866529464722,
+      "learning_rate": 1.56762372024331e-05,
+      "loss": 8.9499,
+      "step": 95870
+    },
+    {
+      "epoch": 0.4788134535194387,
+      "grad_norm": 0.09163626283407211,
+      "learning_rate": 1.5674735287491554e-05,
+      "loss": 8.9528,
+      "step": 95880
+    },
+    {
+      "epoch": 0.47886339234437814,
+      "grad_norm": 0.09393519908189774,
+      "learning_rate": 1.567323337255e-05,
+      "loss": 8.9393,
+      "step": 95890
+    },
+    {
+      "epoch": 0.4789133311693176,
+      "grad_norm": 0.09879069030284882,
+      "learning_rate": 1.567173145760845e-05,
+      "loss": 8.9523,
+      "step": 95900
+    },
+    {
+      "epoch": 0.47896326999425703,
+      "grad_norm": 0.09257882833480835,
+      "learning_rate": 1.56702295426669e-05,
+      "loss": 8.9472,
+      "step": 95910
+    },
+    {
+      "epoch": 0.47901320881919646,
+      "grad_norm": 0.09598603844642639,
+      "learning_rate": 1.5668727627725348e-05,
+      "loss": 8.9522,
+      "step": 95920
+    },
+    {
+      "epoch": 0.47906314764413593,
+      "grad_norm": 0.09990739077329636,
+      "learning_rate": 1.5667225712783802e-05,
+      "loss": 8.9547,
+      "step": 95930
+    },
+    {
+      "epoch": 0.47911308646907536,
+      "grad_norm": 0.09603408724069595,
+      "learning_rate": 1.566572379784225e-05,
+      "loss": 8.9491,
+      "step": 95940
+    },
+    {
+      "epoch": 0.47916302529401483,
+      "grad_norm": 0.09766265004873276,
+      "learning_rate": 1.56642218829007e-05,
+      "loss": 8.9592,
+      "step": 95950
+    },
+    {
+      "epoch": 0.47921296411895425,
+      "grad_norm": 0.08897819370031357,
+      "learning_rate": 1.566271996795915e-05,
+      "loss": 8.9378,
+      "step": 95960
+    },
+    {
+      "epoch": 0.47926290294389373,
+      "grad_norm": 0.09036344289779663,
+      "learning_rate": 1.5661218053017596e-05,
+      "loss": 8.9585,
+      "step": 95970
+    },
+    {
+      "epoch": 0.47931284176883315,
+      "grad_norm": 0.0929788202047348,
+      "learning_rate": 1.565971613807605e-05,
+      "loss": 8.941,
+      "step": 95980
+    },
+    {
+      "epoch": 0.47936278059377263,
+      "grad_norm": 0.08704034984111786,
+      "learning_rate": 1.5658214223134496e-05,
+      "loss": 8.9409,
+      "step": 95990
+    },
+    {
+      "epoch": 0.47941271941871205,
+      "grad_norm": 0.08739351481199265,
+      "learning_rate": 1.5656712308192946e-05,
+      "loss": 8.9567,
+      "step": 96000
+    },
+    {
+      "epoch": 0.47946265824365153,
+      "grad_norm": 0.0888032615184784,
+      "learning_rate": 1.5655210393251396e-05,
+      "loss": 8.95,
+      "step": 96010
+    },
+    {
+      "epoch": 0.47951259706859095,
+      "grad_norm": 0.09704812616109848,
+      "learning_rate": 1.5653708478309843e-05,
+      "loss": 8.9473,
+      "step": 96020
+    },
+    {
+      "epoch": 0.47956253589353043,
+      "grad_norm": 0.0939079076051712,
+      "learning_rate": 1.5652206563368297e-05,
+      "loss": 8.939,
+      "step": 96030
+    },
+    {
+      "epoch": 0.47961247471846985,
+      "grad_norm": 0.09814861416816711,
+      "learning_rate": 1.5650704648426744e-05,
+      "loss": 8.9571,
+      "step": 96040
+    },
+    {
+      "epoch": 0.4796624135434093,
+      "grad_norm": 0.09225606173276901,
+      "learning_rate": 1.5649202733485194e-05,
+      "loss": 8.9364,
+      "step": 96050
+    },
+    {
+      "epoch": 0.47971235236834875,
+      "grad_norm": 0.08847086131572723,
+      "learning_rate": 1.5647700818543644e-05,
+      "loss": 8.9359,
+      "step": 96060
+    },
+    {
+      "epoch": 0.4797622911932882,
+      "grad_norm": 0.09193941950798035,
+      "learning_rate": 1.5646198903602094e-05,
+      "loss": 8.9464,
+      "step": 96070
+    },
+    {
+      "epoch": 0.47981223001822765,
+      "grad_norm": 0.09449310600757599,
+      "learning_rate": 1.5644696988660544e-05,
+      "loss": 8.9496,
+      "step": 96080
+    },
+    {
+      "epoch": 0.4798621688431671,
+      "grad_norm": 0.08825480937957764,
+      "learning_rate": 1.564319507371899e-05,
+      "loss": 8.9294,
+      "step": 96090
+    },
+    {
+      "epoch": 0.47991210766810655,
+      "grad_norm": 0.10034892708063126,
+      "learning_rate": 1.564169315877744e-05,
+      "loss": 8.9352,
+      "step": 96100
+    },
+    {
+      "epoch": 0.479962046493046,
+      "grad_norm": 0.08834687620401382,
+      "learning_rate": 1.564019124383589e-05,
+      "loss": 8.9406,
+      "step": 96110
+    },
+    {
+      "epoch": 0.48001198531798545,
+      "grad_norm": 0.09365418553352356,
+      "learning_rate": 1.563868932889434e-05,
+      "loss": 8.942,
+      "step": 96120
+    },
+    {
+      "epoch": 0.4800619241429249,
+      "grad_norm": 0.09072919934988022,
+      "learning_rate": 1.5637187413952792e-05,
+      "loss": 8.9519,
+      "step": 96130
+    },
+    {
+      "epoch": 0.48011186296786434,
+      "grad_norm": 0.09022905677556992,
+      "learning_rate": 1.563568549901124e-05,
+      "loss": 8.9463,
+      "step": 96140
+    },
+    {
+      "epoch": 0.4801618017928038,
+      "grad_norm": 0.09697787463665009,
+      "learning_rate": 1.563418358406969e-05,
+      "loss": 8.9448,
+      "step": 96150
+    },
+    {
+      "epoch": 0.48021174061774324,
+      "grad_norm": 0.09243085980415344,
+      "learning_rate": 1.563268166912814e-05,
+      "loss": 8.9423,
+      "step": 96160
+    },
+    {
+      "epoch": 0.4802616794426827,
+      "grad_norm": 0.08860752731561661,
+      "learning_rate": 1.563117975418659e-05,
+      "loss": 8.9517,
+      "step": 96170
+    },
+    {
+      "epoch": 0.48031161826762214,
+      "grad_norm": 0.09040497988462448,
+      "learning_rate": 1.562967783924504e-05,
+      "loss": 8.9399,
+      "step": 96180
+    },
+    {
+      "epoch": 0.4803615570925616,
+      "grad_norm": 0.09311524033546448,
+      "learning_rate": 1.5628175924303486e-05,
+      "loss": 8.929,
+      "step": 96190
+    },
+    {
+      "epoch": 0.48041149591750104,
+      "grad_norm": 0.08853588253259659,
+      "learning_rate": 1.5626674009361936e-05,
+      "loss": 8.9327,
+      "step": 96200
+    },
+    {
+      "epoch": 0.4804614347424405,
+      "grad_norm": 0.09366882592439651,
+      "learning_rate": 1.5625172094420386e-05,
+      "loss": 8.9437,
+      "step": 96210
+    },
+    {
+      "epoch": 0.48051137356737994,
+      "grad_norm": 0.09654311090707779,
+      "learning_rate": 1.5623670179478837e-05,
+      "loss": 8.9322,
+      "step": 96220
+    },
+    {
+      "epoch": 0.4805613123923194,
+      "grad_norm": 0.09593677520751953,
+      "learning_rate": 1.5622168264537287e-05,
+      "loss": 8.9382,
+      "step": 96230
+    },
+    {
+      "epoch": 0.48061125121725884,
+      "grad_norm": 0.09233121573925018,
+      "learning_rate": 1.5620666349595734e-05,
+      "loss": 8.9288,
+      "step": 96240
+    },
+    {
+      "epoch": 0.4806611900421983,
+      "grad_norm": 0.0946589857339859,
+      "learning_rate": 1.5619164434654184e-05,
+      "loss": 8.9519,
+      "step": 96250
+    },
+    {
+      "epoch": 0.48071112886713774,
+      "grad_norm": 0.09263359010219574,
+      "learning_rate": 1.5617662519712634e-05,
+      "loss": 8.9297,
+      "step": 96260
+    },
+    {
+      "epoch": 0.4807610676920772,
+      "grad_norm": 0.09187857806682587,
+      "learning_rate": 1.5616160604771084e-05,
+      "loss": 8.9482,
+      "step": 96270
+    },
+    {
+      "epoch": 0.48081100651701664,
+      "grad_norm": 0.09805098176002502,
+      "learning_rate": 1.5614658689829534e-05,
+      "loss": 8.9409,
+      "step": 96280
+    },
+    {
+      "epoch": 0.4808609453419561,
+      "grad_norm": 0.09072684496641159,
+      "learning_rate": 1.561315677488798e-05,
+      "loss": 8.9378,
+      "step": 96290
+    },
+    {
+      "epoch": 0.48091088416689554,
+      "grad_norm": 0.0932784229516983,
+      "learning_rate": 1.561165485994643e-05,
+      "loss": 8.9335,
+      "step": 96300
+    },
+    {
+      "epoch": 0.480960822991835,
+      "grad_norm": 0.09350146353244781,
+      "learning_rate": 1.561015294500488e-05,
+      "loss": 8.9404,
+      "step": 96310
+    },
+    {
+      "epoch": 0.48101076181677443,
+      "grad_norm": 0.09393427520990372,
+      "learning_rate": 1.560865103006333e-05,
+      "loss": 8.937,
+      "step": 96320
+    },
+    {
+      "epoch": 0.4810607006417139,
+      "grad_norm": 0.09077678620815277,
+      "learning_rate": 1.5607149115121782e-05,
+      "loss": 8.9297,
+      "step": 96330
+    },
+    {
+      "epoch": 0.48111063946665333,
+      "grad_norm": 0.08859416097402573,
+      "learning_rate": 1.560564720018023e-05,
+      "loss": 8.9511,
+      "step": 96340
+    },
+    {
+      "epoch": 0.4811605782915928,
+      "grad_norm": 0.09834136813879013,
+      "learning_rate": 1.560414528523868e-05,
+      "loss": 8.9462,
+      "step": 96350
+    },
+    {
+      "epoch": 0.48121051711653223,
+      "grad_norm": 0.09708964824676514,
+      "learning_rate": 1.560264337029713e-05,
+      "loss": 8.9358,
+      "step": 96360
+    },
+    {
+      "epoch": 0.4812604559414717,
+      "grad_norm": 0.09646966308355331,
+      "learning_rate": 1.560114145535558e-05,
+      "loss": 8.9394,
+      "step": 96370
+    },
+    {
+      "epoch": 0.48131039476641113,
+      "grad_norm": 0.09074628353118896,
+      "learning_rate": 1.559963954041403e-05,
+      "loss": 8.9441,
+      "step": 96380
+    },
+    {
+      "epoch": 0.4813603335913506,
+      "grad_norm": 0.09401838481426239,
+      "learning_rate": 1.559813762547248e-05,
+      "loss": 8.9365,
+      "step": 96390
+    },
+    {
+      "epoch": 0.48141027241629003,
+      "grad_norm": 0.09368253499269485,
+      "learning_rate": 1.5596635710530926e-05,
+      "loss": 8.9477,
+      "step": 96400
+    },
+    {
+      "epoch": 0.4814602112412295,
+      "grad_norm": 0.09199754893779755,
+      "learning_rate": 1.5595133795589376e-05,
+      "loss": 8.9274,
+      "step": 96410
+    },
+    {
+      "epoch": 0.48151015006616893,
+      "grad_norm": 0.09040980041027069,
+      "learning_rate": 1.5593631880647827e-05,
+      "loss": 8.9531,
+      "step": 96420
+    },
+    {
+      "epoch": 0.4815600888911084,
+      "grad_norm": 0.09219523519277573,
+      "learning_rate": 1.5592129965706277e-05,
+      "loss": 8.9384,
+      "step": 96430
+    },
+    {
+      "epoch": 0.48161002771604783,
+      "grad_norm": 0.09335026144981384,
+      "learning_rate": 1.5590628050764727e-05,
+      "loss": 8.9469,
+      "step": 96440
+    },
+    {
+      "epoch": 0.4816599665409873,
+      "grad_norm": 0.09123118221759796,
+      "learning_rate": 1.5589126135823174e-05,
+      "loss": 8.9421,
+      "step": 96450
+    },
+    {
+      "epoch": 0.4817099053659267,
+      "grad_norm": 0.09461282938718796,
+      "learning_rate": 1.5587624220881624e-05,
+      "loss": 8.9259,
+      "step": 96460
+    },
+    {
+      "epoch": 0.4817598441908662,
+      "grad_norm": 0.09215594083070755,
+      "learning_rate": 1.5586122305940074e-05,
+      "loss": 8.9334,
+      "step": 96470
+    },
+    {
+      "epoch": 0.4818097830158056,
+      "grad_norm": 0.09032759815454483,
+      "learning_rate": 1.5584620390998524e-05,
+      "loss": 8.946,
+      "step": 96480
+    },
+    {
+      "epoch": 0.4818597218407451,
+      "grad_norm": 0.09267091751098633,
+      "learning_rate": 1.5583118476056974e-05,
+      "loss": 8.9375,
+      "step": 96490
+    },
+    {
+      "epoch": 0.4819096606656845,
+      "grad_norm": 0.09548508375883102,
+      "learning_rate": 1.558161656111542e-05,
+      "loss": 8.9397,
+      "step": 96500
+    },
+    {
+      "epoch": 0.481959599490624,
+      "grad_norm": 0.09211142361164093,
+      "learning_rate": 1.558011464617387e-05,
+      "loss": 8.943,
+      "step": 96510
+    },
+    {
+      "epoch": 0.4820095383155634,
+      "grad_norm": 0.09218604117631912,
+      "learning_rate": 1.557861273123232e-05,
+      "loss": 8.9385,
+      "step": 96520
+    },
+    {
+      "epoch": 0.4820594771405029,
+      "grad_norm": 0.09151312708854675,
+      "learning_rate": 1.5577110816290772e-05,
+      "loss": 8.9425,
+      "step": 96530
+    },
+    {
+      "epoch": 0.4821094159654423,
+      "grad_norm": 0.09401395171880722,
+      "learning_rate": 1.5575608901349222e-05,
+      "loss": 8.9411,
+      "step": 96540
+    },
+    {
+      "epoch": 0.4821593547903818,
+      "grad_norm": 0.09752167761325836,
+      "learning_rate": 1.557410698640767e-05,
+      "loss": 8.954,
+      "step": 96550
+    },
+    {
+      "epoch": 0.4822092936153212,
+      "grad_norm": 0.0942562073469162,
+      "learning_rate": 1.557260507146612e-05,
+      "loss": 8.9508,
+      "step": 96560
+    },
+    {
+      "epoch": 0.4822592324402607,
+      "grad_norm": 0.09495299309492111,
+      "learning_rate": 1.557110315652457e-05,
+      "loss": 8.9436,
+      "step": 96570
+    },
+    {
+      "epoch": 0.4823091712652001,
+      "grad_norm": 0.08862724155187607,
+      "learning_rate": 1.556960124158302e-05,
+      "loss": 8.9357,
+      "step": 96580
+    },
+    {
+      "epoch": 0.4823591100901396,
+      "grad_norm": 0.09534230828285217,
+      "learning_rate": 1.556809932664147e-05,
+      "loss": 8.938,
+      "step": 96590
+    },
+    {
+      "epoch": 0.482409048915079,
+      "grad_norm": 0.08949341624975204,
+      "learning_rate": 1.5566597411699916e-05,
+      "loss": 8.9301,
+      "step": 96600
+    },
+    {
+      "epoch": 0.4824589877400185,
+      "grad_norm": 0.08901667594909668,
+      "learning_rate": 1.5565095496758366e-05,
+      "loss": 8.9381,
+      "step": 96610
+    },
+    {
+      "epoch": 0.4825089265649579,
+      "grad_norm": 0.09051734954118729,
+      "learning_rate": 1.5563593581816817e-05,
+      "loss": 8.9413,
+      "step": 96620
+    },
+    {
+      "epoch": 0.4825588653898974,
+      "grad_norm": 0.09591720253229141,
+      "learning_rate": 1.5562091666875267e-05,
+      "loss": 8.9389,
+      "step": 96630
+    },
+    {
+      "epoch": 0.4826088042148368,
+      "grad_norm": 0.09589020907878876,
+      "learning_rate": 1.5560589751933717e-05,
+      "loss": 8.9373,
+      "step": 96640
+    },
+    {
+      "epoch": 0.4826587430397763,
+      "grad_norm": 0.0880039632320404,
+      "learning_rate": 1.5559087836992164e-05,
+      "loss": 8.9322,
+      "step": 96650
+    },
+    {
+      "epoch": 0.4827086818647157,
+      "grad_norm": 0.09445958584547043,
+      "learning_rate": 1.5557585922050614e-05,
+      "loss": 8.9557,
+      "step": 96660
+    },
+    {
+      "epoch": 0.4827586206896552,
+      "grad_norm": 0.09995859116315842,
+      "learning_rate": 1.5556084007109064e-05,
+      "loss": 8.9236,
+      "step": 96670
+    },
+    {
+      "epoch": 0.4828085595145946,
+      "grad_norm": 0.09632241725921631,
+      "learning_rate": 1.5554582092167514e-05,
+      "loss": 8.9337,
+      "step": 96680
+    },
+    {
+      "epoch": 0.4828584983395341,
+      "grad_norm": 0.09016299247741699,
+      "learning_rate": 1.5553080177225964e-05,
+      "loss": 8.9381,
+      "step": 96690
+    },
+    {
+      "epoch": 0.4829084371644735,
+      "grad_norm": 0.08898003399372101,
+      "learning_rate": 1.555157826228441e-05,
+      "loss": 8.9435,
+      "step": 96700
+    },
+    {
+      "epoch": 0.482958375989413,
+      "grad_norm": 0.09666922688484192,
+      "learning_rate": 1.5550076347342865e-05,
+      "loss": 8.9445,
+      "step": 96710
+    },
+    {
+      "epoch": 0.4830083148143524,
+      "grad_norm": 0.09168431162834167,
+      "learning_rate": 1.554857443240131e-05,
+      "loss": 8.9312,
+      "step": 96720
+    },
+    {
+      "epoch": 0.4830582536392919,
+      "grad_norm": 0.09085845202207565,
+      "learning_rate": 1.5547072517459762e-05,
+      "loss": 8.9415,
+      "step": 96730
+    },
+    {
+      "epoch": 0.4831081924642313,
+      "grad_norm": 0.09779160469770432,
+      "learning_rate": 1.5545570602518212e-05,
+      "loss": 8.9521,
+      "step": 96740
+    },
+    {
+      "epoch": 0.4831581312891708,
+      "grad_norm": 0.09065205603837967,
+      "learning_rate": 1.554406868757666e-05,
+      "loss": 8.9444,
+      "step": 96750
+    },
+    {
+      "epoch": 0.4832080701141102,
+      "grad_norm": 0.0941033810377121,
+      "learning_rate": 1.5542566772635112e-05,
+      "loss": 8.9362,
+      "step": 96760
+    },
+    {
+      "epoch": 0.4832580089390497,
+      "grad_norm": 0.09128523617982864,
+      "learning_rate": 1.554106485769356e-05,
+      "loss": 8.9498,
+      "step": 96770
+    },
+    {
+      "epoch": 0.4833079477639891,
+      "grad_norm": 0.09456328302621841,
+      "learning_rate": 1.553956294275201e-05,
+      "loss": 8.9385,
+      "step": 96780
+    },
+    {
+      "epoch": 0.4833578865889286,
+      "grad_norm": 0.090219646692276,
+      "learning_rate": 1.553806102781046e-05,
+      "loss": 8.9453,
+      "step": 96790
+    },
+    {
+      "epoch": 0.483407825413868,
+      "grad_norm": 0.09211535006761551,
+      "learning_rate": 1.5536559112868906e-05,
+      "loss": 8.957,
+      "step": 96800
+    },
+    {
+      "epoch": 0.4834577642388075,
+      "grad_norm": 0.09447281807661057,
+      "learning_rate": 1.553505719792736e-05,
+      "loss": 8.9422,
+      "step": 96810
+    },
+    {
+      "epoch": 0.4835077030637469,
+      "grad_norm": 0.09818176925182343,
+      "learning_rate": 1.5533555282985807e-05,
+      "loss": 8.9414,
+      "step": 96820
+    },
+    {
+      "epoch": 0.4835576418886864,
+      "grad_norm": 0.0923561379313469,
+      "learning_rate": 1.5532053368044257e-05,
+      "loss": 8.9497,
+      "step": 96830
+    },
+    {
+      "epoch": 0.4836075807136258,
+      "grad_norm": 0.09403201192617416,
+      "learning_rate": 1.5530551453102707e-05,
+      "loss": 8.9408,
+      "step": 96840
+    },
+    {
+      "epoch": 0.4836575195385653,
+      "grad_norm": 0.08915037661790848,
+      "learning_rate": 1.5529049538161154e-05,
+      "loss": 8.9423,
+      "step": 96850
+    },
+    {
+      "epoch": 0.4837074583635047,
+      "grad_norm": 0.09625165164470673,
+      "learning_rate": 1.5527547623219607e-05,
+      "loss": 8.948,
+      "step": 96860
+    },
+    {
+      "epoch": 0.4837573971884442,
+      "grad_norm": 0.0959184393286705,
+      "learning_rate": 1.5526045708278054e-05,
+      "loss": 8.9383,
+      "step": 96870
+    },
+    {
+      "epoch": 0.4838073360133836,
+      "grad_norm": 0.09560030698776245,
+      "learning_rate": 1.5524543793336504e-05,
+      "loss": 8.9419,
+      "step": 96880
+    },
+    {
+      "epoch": 0.4838572748383231,
+      "grad_norm": 0.09560041129589081,
+      "learning_rate": 1.5523041878394954e-05,
+      "loss": 8.9391,
+      "step": 96890
+    },
+    {
+      "epoch": 0.4839072136632625,
+      "grad_norm": 0.09270980209112167,
+      "learning_rate": 1.55215399634534e-05,
+      "loss": 8.9466,
+      "step": 96900
+    },
+    {
+      "epoch": 0.4839571524882019,
+      "grad_norm": 0.08851654082536697,
+      "learning_rate": 1.5520038048511855e-05,
+      "loss": 8.9342,
+      "step": 96910
+    },
+    {
+      "epoch": 0.4840070913131414,
+      "grad_norm": 0.08883985877037048,
+      "learning_rate": 1.55185361335703e-05,
+      "loss": 8.9362,
+      "step": 96920
+    },
+    {
+      "epoch": 0.4840570301380808,
+      "grad_norm": 0.09435227513313293,
+      "learning_rate": 1.5517034218628752e-05,
+      "loss": 8.9312,
+      "step": 96930
+    },
+    {
+      "epoch": 0.4841069689630203,
+      "grad_norm": 0.09650822728872299,
+      "learning_rate": 1.5515532303687202e-05,
+      "loss": 8.9235,
+      "step": 96940
+    },
+    {
+      "epoch": 0.4841569077879597,
+      "grad_norm": 0.09342892467975616,
+      "learning_rate": 1.551403038874565e-05,
+      "loss": 8.9339,
+      "step": 96950
+    },
+    {
+      "epoch": 0.4842068466128992,
+      "grad_norm": 0.0937606543302536,
+      "learning_rate": 1.5512528473804102e-05,
+      "loss": 8.9385,
+      "step": 96960
+    },
+    {
+      "epoch": 0.4842567854378386,
+      "grad_norm": 0.09186763316392899,
+      "learning_rate": 1.551102655886255e-05,
+      "loss": 8.9308,
+      "step": 96970
+    },
+    {
+      "epoch": 0.4843067242627781,
+      "grad_norm": 0.09301009774208069,
+      "learning_rate": 1.5509524643921e-05,
+      "loss": 8.9326,
+      "step": 96980
+    },
+    {
+      "epoch": 0.4843566630877175,
+      "grad_norm": 0.09419457614421844,
+      "learning_rate": 1.550802272897945e-05,
+      "loss": 8.9292,
+      "step": 96990
+    },
+    {
+      "epoch": 0.484406601912657,
+      "grad_norm": 0.09167234599590302,
+      "learning_rate": 1.5506520814037896e-05,
+      "loss": 8.9431,
+      "step": 97000
+    },
+    {
+      "epoch": 0.4844565407375964,
+      "grad_norm": 0.0926911011338234,
+      "learning_rate": 1.550501889909635e-05,
+      "loss": 8.9269,
+      "step": 97010
+    },
+    {
+      "epoch": 0.4845064795625359,
+      "grad_norm": 0.09434627741575241,
+      "learning_rate": 1.5503516984154797e-05,
+      "loss": 8.9336,
+      "step": 97020
+    },
+    {
+      "epoch": 0.4845564183874753,
+      "grad_norm": 0.09322822839021683,
+      "learning_rate": 1.5502015069213247e-05,
+      "loss": 8.9313,
+      "step": 97030
+    },
+    {
+      "epoch": 0.4846063572124148,
+      "grad_norm": 0.09197692573070526,
+      "learning_rate": 1.5500513154271697e-05,
+      "loss": 8.9304,
+      "step": 97040
+    },
+    {
+      "epoch": 0.4846562960373542,
+      "grad_norm": 0.09229908883571625,
+      "learning_rate": 1.5499011239330144e-05,
+      "loss": 8.9265,
+      "step": 97050
+    },
+    {
+      "epoch": 0.4847062348622937,
+      "grad_norm": 0.09997537732124329,
+      "learning_rate": 1.5497509324388597e-05,
+      "loss": 8.9292,
+      "step": 97060
+    },
+    {
+      "epoch": 0.4847561736872331,
+      "grad_norm": 0.09335023909807205,
+      "learning_rate": 1.5496007409447044e-05,
+      "loss": 8.9234,
+      "step": 97070
+    },
+    {
+      "epoch": 0.4848061125121726,
+      "grad_norm": 0.09383931756019592,
+      "learning_rate": 1.5494505494505498e-05,
+      "loss": 8.9355,
+      "step": 97080
+    },
+    {
+      "epoch": 0.484856051337112,
+      "grad_norm": 0.09556584805250168,
+      "learning_rate": 1.5493003579563944e-05,
+      "loss": 8.9404,
+      "step": 97090
+    },
+    {
+      "epoch": 0.4849059901620515,
+      "grad_norm": 0.09331072121858597,
+      "learning_rate": 1.549150166462239e-05,
+      "loss": 8.935,
+      "step": 97100
+    },
+    {
+      "epoch": 0.4849559289869909,
+      "grad_norm": 0.09537593275308609,
+      "learning_rate": 1.5489999749680845e-05,
+      "loss": 8.9442,
+      "step": 97110
+    },
+    {
+      "epoch": 0.4850058678119304,
+      "grad_norm": 0.0932040587067604,
+      "learning_rate": 1.548849783473929e-05,
+      "loss": 8.9356,
+      "step": 97120
+    },
+    {
+      "epoch": 0.4850558066368698,
+      "grad_norm": 0.09548848122358322,
+      "learning_rate": 1.5486995919797745e-05,
+      "loss": 8.9351,
+      "step": 97130
+    },
+    {
+      "epoch": 0.4851057454618093,
+      "grad_norm": 0.088288314640522,
+      "learning_rate": 1.5485494004856192e-05,
+      "loss": 8.9277,
+      "step": 97140
+    },
+    {
+      "epoch": 0.4851556842867487,
+      "grad_norm": 0.0967058390378952,
+      "learning_rate": 1.548399208991464e-05,
+      "loss": 8.9245,
+      "step": 97150
+    },
+    {
+      "epoch": 0.4852056231116882,
+      "grad_norm": 0.0918177142739296,
+      "learning_rate": 1.5482490174973092e-05,
+      "loss": 8.9246,
+      "step": 97160
+    },
+    {
+      "epoch": 0.4852555619366276,
+      "grad_norm": 0.09275839477777481,
+      "learning_rate": 1.548098826003154e-05,
+      "loss": 8.9332,
+      "step": 97170
+    },
+    {
+      "epoch": 0.4853055007615671,
+      "grad_norm": 0.09571007639169693,
+      "learning_rate": 1.5479486345089993e-05,
+      "loss": 8.9214,
+      "step": 97180
+    },
+    {
+      "epoch": 0.4853554395865065,
+      "grad_norm": 0.09551675617694855,
+      "learning_rate": 1.547798443014844e-05,
+      "loss": 8.9341,
+      "step": 97190
+    },
+    {
+      "epoch": 0.485405378411446,
+      "grad_norm": 0.09668447822332382,
+      "learning_rate": 1.5476482515206886e-05,
+      "loss": 8.9389,
+      "step": 97200
+    },
+    {
+      "epoch": 0.4854553172363854,
+      "grad_norm": 0.09511365741491318,
+      "learning_rate": 1.547498060026534e-05,
+      "loss": 8.9348,
+      "step": 97210
+    },
+    {
+      "epoch": 0.4855052560613249,
+      "grad_norm": 0.09372630715370178,
+      "learning_rate": 1.5473478685323787e-05,
+      "loss": 8.9325,
+      "step": 97220
+    },
+    {
+      "epoch": 0.4855551948862643,
+      "grad_norm": 0.09546605497598648,
+      "learning_rate": 1.547197677038224e-05,
+      "loss": 8.9311,
+      "step": 97230
+    },
+    {
+      "epoch": 0.4856051337112038,
+      "grad_norm": 0.09527111798524857,
+      "learning_rate": 1.5470474855440687e-05,
+      "loss": 8.929,
+      "step": 97240
+    },
+    {
+      "epoch": 0.4856550725361432,
+      "grad_norm": 0.09080222994089127,
+      "learning_rate": 1.5468972940499134e-05,
+      "loss": 8.931,
+      "step": 97250
+    },
+    {
+      "epoch": 0.4857050113610827,
+      "grad_norm": 0.09486258774995804,
+      "learning_rate": 1.5467471025557587e-05,
+      "loss": 8.9347,
+      "step": 97260
+    },
+    {
+      "epoch": 0.4857549501860221,
+      "grad_norm": 0.09705396741628647,
+      "learning_rate": 1.5465969110616034e-05,
+      "loss": 8.9462,
+      "step": 97270
+    },
+    {
+      "epoch": 0.4858048890109616,
+      "grad_norm": 0.09238271415233612,
+      "learning_rate": 1.5464467195674488e-05,
+      "loss": 8.922,
+      "step": 97280
+    },
+    {
+      "epoch": 0.485854827835901,
+      "grad_norm": 0.09559265524148941,
+      "learning_rate": 1.5462965280732934e-05,
+      "loss": 8.9331,
+      "step": 97290
+    },
+    {
+      "epoch": 0.4859047666608405,
+      "grad_norm": 0.09076651930809021,
+      "learning_rate": 1.546146336579138e-05,
+      "loss": 8.9372,
+      "step": 97300
+    },
+    {
+      "epoch": 0.4859547054857799,
+      "grad_norm": 0.09629295766353607,
+      "learning_rate": 1.5459961450849835e-05,
+      "loss": 8.9217,
+      "step": 97310
+    },
+    {
+      "epoch": 0.4860046443107194,
+      "grad_norm": 0.09511082619428635,
+      "learning_rate": 1.545845953590828e-05,
+      "loss": 8.9201,
+      "step": 97320
+    },
+    {
+      "epoch": 0.4860545831356588,
+      "grad_norm": 0.09342075139284134,
+      "learning_rate": 1.5456957620966735e-05,
+      "loss": 8.9417,
+      "step": 97330
+    },
+    {
+      "epoch": 0.4861045219605983,
+      "grad_norm": 0.09244676679372787,
+      "learning_rate": 1.5455455706025182e-05,
+      "loss": 8.9413,
+      "step": 97340
+    },
+    {
+      "epoch": 0.4861544607855377,
+      "grad_norm": 0.08962726593017578,
+      "learning_rate": 1.545395379108363e-05,
+      "loss": 8.9414,
+      "step": 97350
+    },
+    {
+      "epoch": 0.4862043996104772,
+      "grad_norm": 0.09315306693315506,
+      "learning_rate": 1.5452451876142082e-05,
+      "loss": 8.9322,
+      "step": 97360
+    },
+    {
+      "epoch": 0.4862543384354166,
+      "grad_norm": 0.0958314910531044,
+      "learning_rate": 1.545094996120053e-05,
+      "loss": 8.93,
+      "step": 97370
+    },
+    {
+      "epoch": 0.4863042772603561,
+      "grad_norm": 0.09707090258598328,
+      "learning_rate": 1.5449448046258983e-05,
+      "loss": 8.9318,
+      "step": 97380
+    },
+    {
+      "epoch": 0.4863542160852955,
+      "grad_norm": 0.09278113394975662,
+      "learning_rate": 1.544794613131743e-05,
+      "loss": 8.9208,
+      "step": 97390
+    },
+    {
+      "epoch": 0.486404154910235,
+      "grad_norm": 0.09409945458173752,
+      "learning_rate": 1.544644421637588e-05,
+      "loss": 8.9421,
+      "step": 97400
+    },
+    {
+      "epoch": 0.4864540937351744,
+      "grad_norm": 0.08893690258264542,
+      "learning_rate": 1.544494230143433e-05,
+      "loss": 8.9299,
+      "step": 97410
+    },
+    {
+      "epoch": 0.4865040325601139,
+      "grad_norm": 0.09356432408094406,
+      "learning_rate": 1.5443440386492777e-05,
+      "loss": 8.9403,
+      "step": 97420
+    },
+    {
+      "epoch": 0.4865539713850533,
+      "grad_norm": 0.09327803552150726,
+      "learning_rate": 1.544193847155123e-05,
+      "loss": 8.9288,
+      "step": 97430
+    },
+    {
+      "epoch": 0.4866039102099928,
+      "grad_norm": 0.0947476327419281,
+      "learning_rate": 1.5440436556609677e-05,
+      "loss": 8.9193,
+      "step": 97440
+    },
+    {
+      "epoch": 0.4866538490349322,
+      "grad_norm": 0.09564851969480515,
+      "learning_rate": 1.5438934641668127e-05,
+      "loss": 8.9157,
+      "step": 97450
+    },
+    {
+      "epoch": 0.4867037878598717,
+      "grad_norm": 0.09227867424488068,
+      "learning_rate": 1.5437432726726577e-05,
+      "loss": 8.9285,
+      "step": 97460
+    },
+    {
+      "epoch": 0.4867537266848111,
+      "grad_norm": 0.09659763425588608,
+      "learning_rate": 1.5435930811785024e-05,
+      "loss": 8.9351,
+      "step": 97470
+    },
+    {
+      "epoch": 0.48680366550975057,
+      "grad_norm": 0.09299914538860321,
+      "learning_rate": 1.5434428896843478e-05,
+      "loss": 8.9386,
+      "step": 97480
+    },
+    {
+      "epoch": 0.48685360433469,
+      "grad_norm": 0.09589099138975143,
+      "learning_rate": 1.5432926981901925e-05,
+      "loss": 8.9294,
+      "step": 97490
+    },
+    {
+      "epoch": 0.48690354315962947,
+      "grad_norm": 0.08684620261192322,
+      "learning_rate": 1.5431425066960375e-05,
+      "loss": 8.9267,
+      "step": 97500
+    },
+    {
+      "epoch": 0.4869534819845689,
+      "grad_norm": 0.09718476235866547,
+      "learning_rate": 1.5429923152018825e-05,
+      "loss": 8.9199,
+      "step": 97510
+    },
+    {
+      "epoch": 0.48700342080950837,
+      "grad_norm": 0.09026718139648438,
+      "learning_rate": 1.542842123707727e-05,
+      "loss": 8.9492,
+      "step": 97520
+    },
+    {
+      "epoch": 0.4870533596344478,
+      "grad_norm": 0.09231576323509216,
+      "learning_rate": 1.5426919322135725e-05,
+      "loss": 8.931,
+      "step": 97530
+    },
+    {
+      "epoch": 0.48710329845938727,
+      "grad_norm": 0.09282558411359787,
+      "learning_rate": 1.5425417407194172e-05,
+      "loss": 8.9311,
+      "step": 97540
+    },
+    {
+      "epoch": 0.4871532372843267,
+      "grad_norm": 0.10046909749507904,
+      "learning_rate": 1.5423915492252622e-05,
+      "loss": 8.9389,
+      "step": 97550
+    },
+    {
+      "epoch": 0.48720317610926617,
+      "grad_norm": 0.09517937153577805,
+      "learning_rate": 1.5422413577311072e-05,
+      "loss": 8.9321,
+      "step": 97560
+    },
+    {
+      "epoch": 0.4872531149342056,
+      "grad_norm": 0.09505096822977066,
+      "learning_rate": 1.542091166236952e-05,
+      "loss": 8.9198,
+      "step": 97570
+    },
+    {
+      "epoch": 0.48730305375914507,
+      "grad_norm": 0.08916744589805603,
+      "learning_rate": 1.5419409747427973e-05,
+      "loss": 8.9174,
+      "step": 97580
+    },
+    {
+      "epoch": 0.4873529925840845,
+      "grad_norm": 0.08957254886627197,
+      "learning_rate": 1.541790783248642e-05,
+      "loss": 8.9441,
+      "step": 97590
+    },
+    {
+      "epoch": 0.48740293140902397,
+      "grad_norm": 0.08980575203895569,
+      "learning_rate": 1.541640591754487e-05,
+      "loss": 8.9318,
+      "step": 97600
+    },
+    {
+      "epoch": 0.4874528702339634,
+      "grad_norm": 0.09375443309545517,
+      "learning_rate": 1.541490400260332e-05,
+      "loss": 8.9393,
+      "step": 97610
+    },
+    {
+      "epoch": 0.48750280905890286,
+      "grad_norm": 0.08852649480104446,
+      "learning_rate": 1.5413402087661767e-05,
+      "loss": 8.9289,
+      "step": 97620
+    },
+    {
+      "epoch": 0.4875527478838423,
+      "grad_norm": 0.09620748460292816,
+      "learning_rate": 1.541190017272022e-05,
+      "loss": 8.9414,
+      "step": 97630
+    },
+    {
+      "epoch": 0.48760268670878176,
+      "grad_norm": 0.09633340686559677,
+      "learning_rate": 1.5410398257778667e-05,
+      "loss": 8.9263,
+      "step": 97640
+    },
+    {
+      "epoch": 0.4876526255337212,
+      "grad_norm": 0.0938490703701973,
+      "learning_rate": 1.5408896342837117e-05,
+      "loss": 8.932,
+      "step": 97650
+    },
+    {
+      "epoch": 0.48770256435866066,
+      "grad_norm": 0.08980219811201096,
+      "learning_rate": 1.5407394427895567e-05,
+      "loss": 8.913,
+      "step": 97660
+    },
+    {
+      "epoch": 0.4877525031836001,
+      "grad_norm": 0.09491949528455734,
+      "learning_rate": 1.5405892512954014e-05,
+      "loss": 8.9317,
+      "step": 97670
+    },
+    {
+      "epoch": 0.48780244200853956,
+      "grad_norm": 0.0898803249001503,
+      "learning_rate": 1.5404390598012468e-05,
+      "loss": 8.938,
+      "step": 97680
+    },
+    {
+      "epoch": 0.487852380833479,
+      "grad_norm": 0.09298001229763031,
+      "learning_rate": 1.5402888683070915e-05,
+      "loss": 8.9168,
+      "step": 97690
+    },
+    {
+      "epoch": 0.48790231965841846,
+      "grad_norm": 0.09501669555902481,
+      "learning_rate": 1.5401386768129365e-05,
+      "loss": 8.9396,
+      "step": 97700
+    },
+    {
+      "epoch": 0.4879522584833579,
+      "grad_norm": 0.09508873522281647,
+      "learning_rate": 1.5399884853187815e-05,
+      "loss": 8.9306,
+      "step": 97710
+    },
+    {
+      "epoch": 0.48800219730829736,
+      "grad_norm": 0.0931331217288971,
+      "learning_rate": 1.5398382938246265e-05,
+      "loss": 8.9261,
+      "step": 97720
+    },
+    {
+      "epoch": 0.4880521361332368,
+      "grad_norm": 0.0985851138830185,
+      "learning_rate": 1.5396881023304715e-05,
+      "loss": 8.9298,
+      "step": 97730
+    },
+    {
+      "epoch": 0.48810207495817626,
+      "grad_norm": 0.08908767253160477,
+      "learning_rate": 1.5395379108363162e-05,
+      "loss": 8.9204,
+      "step": 97740
+    },
+    {
+      "epoch": 0.4881520137831157,
+      "grad_norm": 0.09742926806211472,
+      "learning_rate": 1.5393877193421612e-05,
+      "loss": 8.9244,
+      "step": 97750
+    },
+    {
+      "epoch": 0.48820195260805516,
+      "grad_norm": 0.09041773527860641,
+      "learning_rate": 1.5392375278480062e-05,
+      "loss": 8.9151,
+      "step": 97760
+    },
+    {
+      "epoch": 0.4882518914329946,
+      "grad_norm": 0.094319187104702,
+      "learning_rate": 1.5390873363538513e-05,
+      "loss": 8.9254,
+      "step": 97770
+    },
+    {
+      "epoch": 0.48830183025793406,
+      "grad_norm": 0.09709497541189194,
+      "learning_rate": 1.5389371448596963e-05,
+      "loss": 8.9349,
+      "step": 97780
+    },
+    {
+      "epoch": 0.4883517690828735,
+      "grad_norm": 0.09084603935480118,
+      "learning_rate": 1.538786953365541e-05,
+      "loss": 8.9242,
+      "step": 97790
+    },
+    {
+      "epoch": 0.48840170790781295,
+      "grad_norm": 0.09305722266435623,
+      "learning_rate": 1.538636761871386e-05,
+      "loss": 8.9348,
+      "step": 97800
+    },
+    {
+      "epoch": 0.4884516467327524,
+      "grad_norm": 0.09015587717294693,
+      "learning_rate": 1.538486570377231e-05,
+      "loss": 8.9306,
+      "step": 97810
+    },
+    {
+      "epoch": 0.48850158555769185,
+      "grad_norm": 0.0915970653295517,
+      "learning_rate": 1.538336378883076e-05,
+      "loss": 8.9168,
+      "step": 97820
+    },
+    {
+      "epoch": 0.4885515243826313,
+      "grad_norm": 0.0917646586894989,
+      "learning_rate": 1.538186187388921e-05,
+      "loss": 8.9245,
+      "step": 97830
+    },
+    {
+      "epoch": 0.48860146320757075,
+      "grad_norm": 0.09395934641361237,
+      "learning_rate": 1.5380359958947657e-05,
+      "loss": 8.9253,
+      "step": 97840
+    },
+    {
+      "epoch": 0.4886514020325102,
+      "grad_norm": 0.09345404803752899,
+      "learning_rate": 1.5378858044006107e-05,
+      "loss": 8.9186,
+      "step": 97850
+    },
+    {
+      "epoch": 0.48870134085744965,
+      "grad_norm": 0.0977400466799736,
+      "learning_rate": 1.5377356129064557e-05,
+      "loss": 8.9381,
+      "step": 97860
+    },
+    {
+      "epoch": 0.4887512796823891,
+      "grad_norm": 0.09431370347738266,
+      "learning_rate": 1.5375854214123008e-05,
+      "loss": 8.925,
+      "step": 97870
+    },
+    {
+      "epoch": 0.48880121850732855,
+      "grad_norm": 0.09120336174964905,
+      "learning_rate": 1.5374352299181458e-05,
+      "loss": 8.9313,
+      "step": 97880
+    },
+    {
+      "epoch": 0.48885115733226797,
+      "grad_norm": 0.09345964342355728,
+      "learning_rate": 1.5372850384239905e-05,
+      "loss": 8.9307,
+      "step": 97890
+    },
+    {
+      "epoch": 0.4889010961572074,
+      "grad_norm": 0.09451073408126831,
+      "learning_rate": 1.5371348469298355e-05,
+      "loss": 8.9228,
+      "step": 97900
+    },
+    {
+      "epoch": 0.48895103498214687,
+      "grad_norm": 0.08943965286016464,
+      "learning_rate": 1.5369846554356805e-05,
+      "loss": 8.9363,
+      "step": 97910
+    },
+    {
+      "epoch": 0.4890009738070863,
+      "grad_norm": 0.09674328565597534,
+      "learning_rate": 1.5368344639415255e-05,
+      "loss": 8.9236,
+      "step": 97920
+    },
+    {
+      "epoch": 0.48905091263202577,
+      "grad_norm": 0.09601275622844696,
+      "learning_rate": 1.5366842724473705e-05,
+      "loss": 8.9287,
+      "step": 97930
+    },
+    {
+      "epoch": 0.4891008514569652,
+      "grad_norm": 0.09064171463251114,
+      "learning_rate": 1.5365340809532152e-05,
+      "loss": 8.9479,
+      "step": 97940
+    },
+    {
+      "epoch": 0.48915079028190467,
+      "grad_norm": 0.09523607045412064,
+      "learning_rate": 1.5363838894590602e-05,
+      "loss": 8.9308,
+      "step": 97950
+    },
+    {
+      "epoch": 0.4892007291068441,
+      "grad_norm": 0.09979385882616043,
+      "learning_rate": 1.5362336979649052e-05,
+      "loss": 8.9273,
+      "step": 97960
+    },
+    {
+      "epoch": 0.48925066793178357,
+      "grad_norm": 0.09213593602180481,
+      "learning_rate": 1.5360835064707503e-05,
+      "loss": 8.9341,
+      "step": 97970
+    },
+    {
+      "epoch": 0.489300606756723,
+      "grad_norm": 0.09432204067707062,
+      "learning_rate": 1.5359333149765953e-05,
+      "loss": 8.9229,
+      "step": 97980
+    },
+    {
+      "epoch": 0.48935054558166247,
+      "grad_norm": 0.09289470314979553,
+      "learning_rate": 1.53578312348244e-05,
+      "loss": 8.9194,
+      "step": 97990
+    },
+    {
+      "epoch": 0.4894004844066019,
+      "grad_norm": 0.09178686141967773,
+      "learning_rate": 1.535632931988285e-05,
+      "loss": 8.9441,
+      "step": 98000
+    },
+    {
+      "epoch": 0.48945042323154136,
+      "grad_norm": 0.08877766132354736,
+      "learning_rate": 1.53548274049413e-05,
+      "loss": 8.9273,
+      "step": 98010
+    },
+    {
+      "epoch": 0.4895003620564808,
+      "grad_norm": 0.09612318128347397,
+      "learning_rate": 1.535332548999975e-05,
+      "loss": 8.9139,
+      "step": 98020
+    },
+    {
+      "epoch": 0.48955030088142026,
+      "grad_norm": 0.09008832275867462,
+      "learning_rate": 1.53518235750582e-05,
+      "loss": 8.9228,
+      "step": 98030
+    },
+    {
+      "epoch": 0.4896002397063597,
+      "grad_norm": 0.0915464237332344,
+      "learning_rate": 1.535032166011665e-05,
+      "loss": 8.9361,
+      "step": 98040
+    },
+    {
+      "epoch": 0.48965017853129916,
+      "grad_norm": 0.093266561627388,
+      "learning_rate": 1.5348819745175097e-05,
+      "loss": 8.9302,
+      "step": 98050
+    },
+    {
+      "epoch": 0.4897001173562386,
+      "grad_norm": 0.08844608813524246,
+      "learning_rate": 1.5347317830233547e-05,
+      "loss": 8.9241,
+      "step": 98060
+    },
+    {
+      "epoch": 0.48975005618117806,
+      "grad_norm": 0.0893082469701767,
+      "learning_rate": 1.5345815915291998e-05,
+      "loss": 8.927,
+      "step": 98070
+    },
+    {
+      "epoch": 0.4897999950061175,
+      "grad_norm": 0.09313119947910309,
+      "learning_rate": 1.5344314000350448e-05,
+      "loss": 8.9324,
+      "step": 98080
+    },
+    {
+      "epoch": 0.48984993383105696,
+      "grad_norm": 0.09175945073366165,
+      "learning_rate": 1.5342812085408898e-05,
+      "loss": 8.9129,
+      "step": 98090
+    },
+    {
+      "epoch": 0.4898998726559964,
+      "grad_norm": 0.09570962190628052,
+      "learning_rate": 1.5341310170467345e-05,
+      "loss": 8.9264,
+      "step": 98100
+    },
+    {
+      "epoch": 0.48994981148093586,
+      "grad_norm": 0.09470510482788086,
+      "learning_rate": 1.5339808255525795e-05,
+      "loss": 8.9235,
+      "step": 98110
+    },
+    {
+      "epoch": 0.4899997503058753,
+      "grad_norm": 0.09118285030126572,
+      "learning_rate": 1.5338306340584245e-05,
+      "loss": 8.9291,
+      "step": 98120
+    },
+    {
+      "epoch": 0.49004968913081476,
+      "grad_norm": 0.09788258373737335,
+      "learning_rate": 1.5336804425642695e-05,
+      "loss": 8.919,
+      "step": 98130
+    },
+    {
+      "epoch": 0.4900996279557542,
+      "grad_norm": 0.09753192961215973,
+      "learning_rate": 1.5335302510701145e-05,
+      "loss": 8.9069,
+      "step": 98140
+    },
+    {
+      "epoch": 0.49014956678069366,
+      "grad_norm": 0.09691204130649567,
+      "learning_rate": 1.5333800595759592e-05,
+      "loss": 8.9225,
+      "step": 98150
+    },
+    {
+      "epoch": 0.4901995056056331,
+      "grad_norm": 0.0969519168138504,
+      "learning_rate": 1.5332298680818042e-05,
+      "loss": 8.9203,
+      "step": 98160
+    },
+    {
+      "epoch": 0.49024944443057256,
+      "grad_norm": 0.09179043769836426,
+      "learning_rate": 1.5330796765876493e-05,
+      "loss": 8.9177,
+      "step": 98170
+    },
+    {
+      "epoch": 0.490299383255512,
+      "grad_norm": 0.09258826822042465,
+      "learning_rate": 1.5329294850934943e-05,
+      "loss": 8.9295,
+      "step": 98180
+    },
+    {
+      "epoch": 0.49034932208045146,
+      "grad_norm": 0.09368129819631577,
+      "learning_rate": 1.5327792935993393e-05,
+      "loss": 8.9187,
+      "step": 98190
+    },
+    {
+      "epoch": 0.4903992609053909,
+      "grad_norm": 0.09183964878320694,
+      "learning_rate": 1.5326291021051843e-05,
+      "loss": 8.9198,
+      "step": 98200
+    },
+    {
+      "epoch": 0.49044919973033035,
+      "grad_norm": 0.0940631628036499,
+      "learning_rate": 1.532478910611029e-05,
+      "loss": 8.9311,
+      "step": 98210
+    },
+    {
+      "epoch": 0.4904991385552698,
+      "grad_norm": 0.09668076038360596,
+      "learning_rate": 1.532328719116874e-05,
+      "loss": 8.9115,
+      "step": 98220
+    },
+    {
+      "epoch": 0.49054907738020925,
+      "grad_norm": 0.09049348533153534,
+      "learning_rate": 1.532178527622719e-05,
+      "loss": 8.9309,
+      "step": 98230
+    },
+    {
+      "epoch": 0.4905990162051487,
+      "grad_norm": 0.08599221706390381,
+      "learning_rate": 1.532028336128564e-05,
+      "loss": 8.9212,
+      "step": 98240
+    },
+    {
+      "epoch": 0.49064895503008815,
+      "grad_norm": 0.09467752277851105,
+      "learning_rate": 1.531878144634409e-05,
+      "loss": 8.9257,
+      "step": 98250
+    },
+    {
+      "epoch": 0.4906988938550276,
+      "grad_norm": 0.10057593882083893,
+      "learning_rate": 1.5317279531402537e-05,
+      "loss": 8.9434,
+      "step": 98260
+    },
+    {
+      "epoch": 0.49074883267996705,
+      "grad_norm": 0.09494142979383469,
+      "learning_rate": 1.5315777616460988e-05,
+      "loss": 8.9185,
+      "step": 98270
+    },
+    {
+      "epoch": 0.4907987715049065,
+      "grad_norm": 0.09219949692487717,
+      "learning_rate": 1.5314275701519438e-05,
+      "loss": 8.9139,
+      "step": 98280
+    },
+    {
+      "epoch": 0.49084871032984595,
+      "grad_norm": 0.09445098787546158,
+      "learning_rate": 1.5312773786577888e-05,
+      "loss": 8.9353,
+      "step": 98290
+    },
+    {
+      "epoch": 0.49089864915478537,
+      "grad_norm": 0.09249778091907501,
+      "learning_rate": 1.5311271871636338e-05,
+      "loss": 8.9202,
+      "step": 98300
+    },
+    {
+      "epoch": 0.49094858797972485,
+      "grad_norm": 0.09504956007003784,
+      "learning_rate": 1.5309769956694785e-05,
+      "loss": 8.9148,
+      "step": 98310
+    },
+    {
+      "epoch": 0.49099852680466427,
+      "grad_norm": 0.09462475776672363,
+      "learning_rate": 1.5308268041753235e-05,
+      "loss": 8.9387,
+      "step": 98320
+    },
+    {
+      "epoch": 0.49104846562960375,
+      "grad_norm": 0.09308874607086182,
+      "learning_rate": 1.5306766126811685e-05,
+      "loss": 8.9143,
+      "step": 98330
+    },
+    {
+      "epoch": 0.49109840445454317,
+      "grad_norm": 0.09277541190385818,
+      "learning_rate": 1.5305264211870135e-05,
+      "loss": 8.9223,
+      "step": 98340
+    },
+    {
+      "epoch": 0.49114834327948265,
+      "grad_norm": 0.09326548874378204,
+      "learning_rate": 1.5303762296928586e-05,
+      "loss": 8.926,
+      "step": 98350
+    },
+    {
+      "epoch": 0.49119828210442207,
+      "grad_norm": 0.09251368790864944,
+      "learning_rate": 1.5302260381987036e-05,
+      "loss": 8.94,
+      "step": 98360
+    },
+    {
+      "epoch": 0.49124822092936155,
+      "grad_norm": 0.0935811847448349,
+      "learning_rate": 1.5300758467045483e-05,
+      "loss": 8.9264,
+      "step": 98370
+    },
+    {
+      "epoch": 0.49129815975430097,
+      "grad_norm": 0.09226620942354202,
+      "learning_rate": 1.5299256552103933e-05,
+      "loss": 8.9254,
+      "step": 98380
+    },
+    {
+      "epoch": 0.49134809857924044,
+      "grad_norm": 0.09316042810678482,
+      "learning_rate": 1.5297754637162383e-05,
+      "loss": 8.9175,
+      "step": 98390
+    },
+    {
+      "epoch": 0.49139803740417987,
+      "grad_norm": 0.09391681849956512,
+      "learning_rate": 1.5296252722220833e-05,
+      "loss": 8.9204,
+      "step": 98400
+    },
+    {
+      "epoch": 0.49144797622911934,
+      "grad_norm": 0.09176868200302124,
+      "learning_rate": 1.5294750807279283e-05,
+      "loss": 8.919,
+      "step": 98410
+    },
+    {
+      "epoch": 0.49149791505405876,
+      "grad_norm": 0.09538892656564713,
+      "learning_rate": 1.529324889233773e-05,
+      "loss": 8.9293,
+      "step": 98420
+    },
+    {
+      "epoch": 0.49154785387899824,
+      "grad_norm": 0.09601672738790512,
+      "learning_rate": 1.529174697739618e-05,
+      "loss": 8.9258,
+      "step": 98430
+    },
+    {
+      "epoch": 0.49159779270393766,
+      "grad_norm": 0.09024684131145477,
+      "learning_rate": 1.529024506245463e-05,
+      "loss": 8.9218,
+      "step": 98440
+    },
+    {
+      "epoch": 0.49164773152887714,
+      "grad_norm": 0.09166863560676575,
+      "learning_rate": 1.528874314751308e-05,
+      "loss": 8.9383,
+      "step": 98450
+    },
+    {
+      "epoch": 0.49169767035381656,
+      "grad_norm": 0.09486201405525208,
+      "learning_rate": 1.528724123257153e-05,
+      "loss": 8.9158,
+      "step": 98460
+    },
+    {
+      "epoch": 0.49174760917875604,
+      "grad_norm": 0.09765074402093887,
+      "learning_rate": 1.5285739317629978e-05,
+      "loss": 8.9163,
+      "step": 98470
+    },
+    {
+      "epoch": 0.49179754800369546,
+      "grad_norm": 0.09160161763429642,
+      "learning_rate": 1.5284237402688428e-05,
+      "loss": 8.9048,
+      "step": 98480
+    },
+    {
+      "epoch": 0.49184748682863494,
+      "grad_norm": 0.09425348788499832,
+      "learning_rate": 1.5282735487746878e-05,
+      "loss": 8.923,
+      "step": 98490
+    },
+    {
+      "epoch": 0.49189742565357436,
+      "grad_norm": 0.09459834545850754,
+      "learning_rate": 1.5281233572805328e-05,
+      "loss": 8.9128,
+      "step": 98500
+    },
+    {
+      "epoch": 0.49194736447851384,
+      "grad_norm": 0.08558713644742966,
+      "learning_rate": 1.527973165786378e-05,
+      "loss": 8.9286,
+      "step": 98510
+    },
+    {
+      "epoch": 0.49199730330345326,
+      "grad_norm": 0.09137829393148422,
+      "learning_rate": 1.5278229742922225e-05,
+      "loss": 8.9298,
+      "step": 98520
+    },
+    {
+      "epoch": 0.49204724212839274,
+      "grad_norm": 0.09085921198129654,
+      "learning_rate": 1.5276727827980675e-05,
+      "loss": 8.8977,
+      "step": 98530
+    },
+    {
+      "epoch": 0.49209718095333216,
+      "grad_norm": 0.08780453354120255,
+      "learning_rate": 1.5275225913039125e-05,
+      "loss": 8.9185,
+      "step": 98540
+    },
+    {
+      "epoch": 0.49214711977827164,
+      "grad_norm": 0.09832119941711426,
+      "learning_rate": 1.5273723998097576e-05,
+      "loss": 8.9146,
+      "step": 98550
+    },
+    {
+      "epoch": 0.49219705860321106,
+      "grad_norm": 0.09514303505420685,
+      "learning_rate": 1.5272222083156026e-05,
+      "loss": 8.9141,
+      "step": 98560
+    },
+    {
+      "epoch": 0.49224699742815053,
+      "grad_norm": 0.09371118247509003,
+      "learning_rate": 1.5270720168214473e-05,
+      "loss": 8.9301,
+      "step": 98570
+    },
+    {
+      "epoch": 0.49229693625308996,
+      "grad_norm": 0.09175911545753479,
+      "learning_rate": 1.5269218253272923e-05,
+      "loss": 8.9148,
+      "step": 98580
+    },
+    {
+      "epoch": 0.49234687507802943,
+      "grad_norm": 0.09370813518762589,
+      "learning_rate": 1.5267716338331373e-05,
+      "loss": 8.917,
+      "step": 98590
+    },
+    {
+      "epoch": 0.49239681390296886,
+      "grad_norm": 0.09525982290506363,
+      "learning_rate": 1.5266214423389823e-05,
+      "loss": 8.9276,
+      "step": 98600
+    },
+    {
+      "epoch": 0.49244675272790833,
+      "grad_norm": 0.09428591281175613,
+      "learning_rate": 1.5264712508448273e-05,
+      "loss": 8.9137,
+      "step": 98610
+    },
+    {
+      "epoch": 0.49249669155284775,
+      "grad_norm": 0.09742698073387146,
+      "learning_rate": 1.526321059350672e-05,
+      "loss": 8.9217,
+      "step": 98620
+    },
+    {
+      "epoch": 0.49254663037778723,
+      "grad_norm": 0.09069808572530746,
+      "learning_rate": 1.526170867856517e-05,
+      "loss": 8.9151,
+      "step": 98630
+    },
+    {
+      "epoch": 0.49259656920272665,
+      "grad_norm": 0.08807901293039322,
+      "learning_rate": 1.526020676362362e-05,
+      "loss": 8.9236,
+      "step": 98640
+    },
+    {
+      "epoch": 0.49264650802766613,
+      "grad_norm": 0.10254672169685364,
+      "learning_rate": 1.5258704848682069e-05,
+      "loss": 8.9263,
+      "step": 98650
+    },
+    {
+      "epoch": 0.49269644685260555,
+      "grad_norm": 0.09022139757871628,
+      "learning_rate": 1.525720293374052e-05,
+      "loss": 8.937,
+      "step": 98660
+    },
+    {
+      "epoch": 0.49274638567754503,
+      "grad_norm": 0.0958266630768776,
+      "learning_rate": 1.525570101879897e-05,
+      "loss": 8.9133,
+      "step": 98670
+    },
+    {
+      "epoch": 0.49279632450248445,
+      "grad_norm": 0.09768004715442657,
+      "learning_rate": 1.525419910385742e-05,
+      "loss": 8.9149,
+      "step": 98680
+    },
+    {
+      "epoch": 0.49284626332742393,
+      "grad_norm": 0.09794740378856659,
+      "learning_rate": 1.5252697188915868e-05,
+      "loss": 8.9157,
+      "step": 98690
+    },
+    {
+      "epoch": 0.49289620215236335,
+      "grad_norm": 0.09140866249799728,
+      "learning_rate": 1.5251195273974316e-05,
+      "loss": 8.931,
+      "step": 98700
+    },
+    {
+      "epoch": 0.4929461409773028,
+      "grad_norm": 0.09415680915117264,
+      "learning_rate": 1.5249693359032768e-05,
+      "loss": 8.9115,
+      "step": 98710
+    },
+    {
+      "epoch": 0.49299607980224225,
+      "grad_norm": 0.09118063002824783,
+      "learning_rate": 1.5248191444091217e-05,
+      "loss": 8.9061,
+      "step": 98720
+    },
+    {
+      "epoch": 0.4930460186271817,
+      "grad_norm": 0.09346728771924973,
+      "learning_rate": 1.5246689529149667e-05,
+      "loss": 8.9145,
+      "step": 98730
+    },
+    {
+      "epoch": 0.49309595745212115,
+      "grad_norm": 0.0916360467672348,
+      "learning_rate": 1.5245187614208115e-05,
+      "loss": 8.9112,
+      "step": 98740
+    },
+    {
+      "epoch": 0.4931458962770606,
+      "grad_norm": 0.09017834067344666,
+      "learning_rate": 1.5243685699266564e-05,
+      "loss": 8.9181,
+      "step": 98750
+    },
+    {
+      "epoch": 0.49319583510200005,
+      "grad_norm": 0.09475754201412201,
+      "learning_rate": 1.5242183784325016e-05,
+      "loss": 8.9081,
+      "step": 98760
+    },
+    {
+      "epoch": 0.4932457739269395,
+      "grad_norm": 0.09371718764305115,
+      "learning_rate": 1.5240681869383464e-05,
+      "loss": 8.9163,
+      "step": 98770
+    },
+    {
+      "epoch": 0.49329571275187895,
+      "grad_norm": 0.09911168366670609,
+      "learning_rate": 1.5239179954441915e-05,
+      "loss": 8.9001,
+      "step": 98780
+    },
+    {
+      "epoch": 0.4933456515768184,
+      "grad_norm": 0.09791523963212967,
+      "learning_rate": 1.5237678039500363e-05,
+      "loss": 8.9206,
+      "step": 98790
+    },
+    {
+      "epoch": 0.49339559040175784,
+      "grad_norm": 0.09776368737220764,
+      "learning_rate": 1.5236176124558811e-05,
+      "loss": 8.9072,
+      "step": 98800
+    },
+    {
+      "epoch": 0.4934455292266973,
+      "grad_norm": 0.09171529114246368,
+      "learning_rate": 1.5234674209617263e-05,
+      "loss": 8.9162,
+      "step": 98810
+    },
+    {
+      "epoch": 0.49349546805163674,
+      "grad_norm": 0.08931485563516617,
+      "learning_rate": 1.5233172294675712e-05,
+      "loss": 8.9238,
+      "step": 98820
+    },
+    {
+      "epoch": 0.4935454068765762,
+      "grad_norm": 0.09580953419208527,
+      "learning_rate": 1.5231670379734162e-05,
+      "loss": 8.9204,
+      "step": 98830
+    },
+    {
+      "epoch": 0.49359534570151564,
+      "grad_norm": 0.09444009512662888,
+      "learning_rate": 1.523016846479261e-05,
+      "loss": 8.9138,
+      "step": 98840
+    },
+    {
+      "epoch": 0.4936452845264551,
+      "grad_norm": 0.0973835289478302,
+      "learning_rate": 1.5228666549851059e-05,
+      "loss": 8.8994,
+      "step": 98850
+    },
+    {
+      "epoch": 0.49369522335139454,
+      "grad_norm": 0.0919685810804367,
+      "learning_rate": 1.5227164634909511e-05,
+      "loss": 8.9286,
+      "step": 98860
+    },
+    {
+      "epoch": 0.493745162176334,
+      "grad_norm": 0.09695620089769363,
+      "learning_rate": 1.522566271996796e-05,
+      "loss": 8.9083,
+      "step": 98870
+    },
+    {
+      "epoch": 0.49379510100127344,
+      "grad_norm": 0.08825941383838654,
+      "learning_rate": 1.522416080502641e-05,
+      "loss": 8.9236,
+      "step": 98880
+    },
+    {
+      "epoch": 0.49384503982621286,
+      "grad_norm": 0.08976638317108154,
+      "learning_rate": 1.5222658890084858e-05,
+      "loss": 8.9358,
+      "step": 98890
+    },
+    {
+      "epoch": 0.49389497865115234,
+      "grad_norm": 0.09463369101285934,
+      "learning_rate": 1.5221156975143306e-05,
+      "loss": 8.9231,
+      "step": 98900
+    },
+    {
+      "epoch": 0.49394491747609176,
+      "grad_norm": 0.09321442991495132,
+      "learning_rate": 1.5219655060201758e-05,
+      "loss": 8.9213,
+      "step": 98910
+    },
+    {
+      "epoch": 0.49399485630103124,
+      "grad_norm": 0.0944615826010704,
+      "learning_rate": 1.5218153145260207e-05,
+      "loss": 8.9208,
+      "step": 98920
+    },
+    {
+      "epoch": 0.49404479512597066,
+      "grad_norm": 0.09346310794353485,
+      "learning_rate": 1.5216651230318657e-05,
+      "loss": 8.9182,
+      "step": 98930
+    },
+    {
+      "epoch": 0.49409473395091014,
+      "grad_norm": 0.09217849373817444,
+      "learning_rate": 1.5215149315377106e-05,
+      "loss": 8.9081,
+      "step": 98940
+    },
+    {
+      "epoch": 0.49414467277584956,
+      "grad_norm": 0.09414231032133102,
+      "learning_rate": 1.5213647400435554e-05,
+      "loss": 8.9238,
+      "step": 98950
+    },
+    {
+      "epoch": 0.49419461160078904,
+      "grad_norm": 0.09089777618646622,
+      "learning_rate": 1.5212145485494006e-05,
+      "loss": 8.9218,
+      "step": 98960
+    },
+    {
+      "epoch": 0.49424455042572846,
+      "grad_norm": 0.09277298301458359,
+      "learning_rate": 1.5210643570552454e-05,
+      "loss": 8.9305,
+      "step": 98970
+    },
+    {
+      "epoch": 0.49429448925066793,
+      "grad_norm": 0.09524578601121902,
+      "learning_rate": 1.5209141655610905e-05,
+      "loss": 8.9063,
+      "step": 98980
+    },
+    {
+      "epoch": 0.49434442807560736,
+      "grad_norm": 0.09183613210916519,
+      "learning_rate": 1.5207639740669353e-05,
+      "loss": 8.9204,
+      "step": 98990
+    },
+    {
+      "epoch": 0.49439436690054683,
+      "grad_norm": 0.09120965749025345,
+      "learning_rate": 1.5206137825727805e-05,
+      "loss": 8.92,
+      "step": 99000
+    },
+    {
+      "epoch": 0.49444430572548626,
+      "grad_norm": 0.09432142972946167,
+      "learning_rate": 1.5204635910786253e-05,
+      "loss": 8.9168,
+      "step": 99010
+    },
+    {
+      "epoch": 0.49449424455042573,
+      "grad_norm": 0.0890686959028244,
+      "learning_rate": 1.5203133995844702e-05,
+      "loss": 8.9168,
+      "step": 99020
+    },
+    {
+      "epoch": 0.49454418337536515,
+      "grad_norm": 0.09807530790567398,
+      "learning_rate": 1.5201632080903152e-05,
+      "loss": 8.9078,
+      "step": 99030
+    },
+    {
+      "epoch": 0.49459412220030463,
+      "grad_norm": 0.09491509944200516,
+      "learning_rate": 1.52001301659616e-05,
+      "loss": 8.909,
+      "step": 99040
+    },
+    {
+      "epoch": 0.49464406102524405,
+      "grad_norm": 0.09278037399053574,
+      "learning_rate": 1.5198628251020052e-05,
+      "loss": 8.9086,
+      "step": 99050
+    },
+    {
+      "epoch": 0.49469399985018353,
+      "grad_norm": 0.09126697480678558,
+      "learning_rate": 1.5197126336078501e-05,
+      "loss": 8.9129,
+      "step": 99060
+    },
+    {
+      "epoch": 0.49474393867512295,
+      "grad_norm": 0.0916236862540245,
+      "learning_rate": 1.519562442113695e-05,
+      "loss": 8.9122,
+      "step": 99070
+    },
+    {
+      "epoch": 0.49479387750006243,
+      "grad_norm": 0.08973243832588196,
+      "learning_rate": 1.51941225061954e-05,
+      "loss": 8.9116,
+      "step": 99080
+    },
+    {
+      "epoch": 0.49484381632500185,
+      "grad_norm": 0.09231727570295334,
+      "learning_rate": 1.5192620591253848e-05,
+      "loss": 8.9159,
+      "step": 99090
+    },
+    {
+      "epoch": 0.49489375514994133,
+      "grad_norm": 0.09403295069932938,
+      "learning_rate": 1.51911186763123e-05,
+      "loss": 8.9249,
+      "step": 99100
+    },
+    {
+      "epoch": 0.49494369397488075,
+      "grad_norm": 0.09361027926206589,
+      "learning_rate": 1.5189616761370748e-05,
+      "loss": 8.9205,
+      "step": 99110
+    },
+    {
+      "epoch": 0.4949936327998202,
+      "grad_norm": 0.09127476811408997,
+      "learning_rate": 1.5188114846429197e-05,
+      "loss": 8.9237,
+      "step": 99120
+    },
+    {
+      "epoch": 0.49504357162475965,
+      "grad_norm": 0.090853251516819,
+      "learning_rate": 1.5186612931487647e-05,
+      "loss": 8.9382,
+      "step": 99130
+    },
+    {
+      "epoch": 0.4950935104496991,
+      "grad_norm": 0.09697920083999634,
+      "learning_rate": 1.5185111016546096e-05,
+      "loss": 8.9156,
+      "step": 99140
+    },
+    {
+      "epoch": 0.49514344927463855,
+      "grad_norm": 0.09126929193735123,
+      "learning_rate": 1.5183609101604547e-05,
+      "loss": 8.9194,
+      "step": 99150
+    },
+    {
+      "epoch": 0.495193388099578,
+      "grad_norm": 0.0899655893445015,
+      "learning_rate": 1.5182107186662996e-05,
+      "loss": 8.9068,
+      "step": 99160
+    },
+    {
+      "epoch": 0.49524332692451745,
+      "grad_norm": 0.08864665776491165,
+      "learning_rate": 1.5180605271721444e-05,
+      "loss": 8.8933,
+      "step": 99170
+    },
+    {
+      "epoch": 0.4952932657494569,
+      "grad_norm": 0.09634734690189362,
+      "learning_rate": 1.5179103356779895e-05,
+      "loss": 8.9322,
+      "step": 99180
+    },
+    {
+      "epoch": 0.49534320457439635,
+      "grad_norm": 0.0953470766544342,
+      "learning_rate": 1.5177601441838343e-05,
+      "loss": 8.9094,
+      "step": 99190
+    },
+    {
+      "epoch": 0.4953931433993358,
+      "grad_norm": 0.0901167094707489,
+      "learning_rate": 1.5176099526896795e-05,
+      "loss": 8.9126,
+      "step": 99200
+    },
+    {
+      "epoch": 0.49544308222427524,
+      "grad_norm": 0.09353036433458328,
+      "learning_rate": 1.5174597611955243e-05,
+      "loss": 8.9145,
+      "step": 99210
+    },
+    {
+      "epoch": 0.4954930210492147,
+      "grad_norm": 0.0913802906870842,
+      "learning_rate": 1.5173095697013692e-05,
+      "loss": 8.9135,
+      "step": 99220
+    },
+    {
+      "epoch": 0.49554295987415414,
+      "grad_norm": 0.09367360174655914,
+      "learning_rate": 1.5171593782072142e-05,
+      "loss": 8.9098,
+      "step": 99230
+    },
+    {
+      "epoch": 0.4955928986990936,
+      "grad_norm": 0.08964614570140839,
+      "learning_rate": 1.517009186713059e-05,
+      "loss": 8.9137,
+      "step": 99240
+    },
+    {
+      "epoch": 0.49564283752403304,
+      "grad_norm": 0.0972425788640976,
+      "learning_rate": 1.5168589952189042e-05,
+      "loss": 8.9079,
+      "step": 99250
+    },
+    {
+      "epoch": 0.4956927763489725,
+      "grad_norm": 0.08969489485025406,
+      "learning_rate": 1.5167088037247491e-05,
+      "loss": 8.9144,
+      "step": 99260
+    },
+    {
+      "epoch": 0.49574271517391194,
+      "grad_norm": 0.09431774914264679,
+      "learning_rate": 1.516558612230594e-05,
+      "loss": 8.9184,
+      "step": 99270
+    },
+    {
+      "epoch": 0.4957926539988514,
+      "grad_norm": 0.0908089205622673,
+      "learning_rate": 1.516408420736439e-05,
+      "loss": 8.9106,
+      "step": 99280
+    },
+    {
+      "epoch": 0.49584259282379084,
+      "grad_norm": 0.08945730328559875,
+      "learning_rate": 1.5162582292422838e-05,
+      "loss": 8.9202,
+      "step": 99290
+    },
+    {
+      "epoch": 0.4958925316487303,
+      "grad_norm": 0.09563491493463516,
+      "learning_rate": 1.516108037748129e-05,
+      "loss": 8.8976,
+      "step": 99300
+    },
+    {
+      "epoch": 0.49594247047366974,
+      "grad_norm": 0.0948091596364975,
+      "learning_rate": 1.5159578462539738e-05,
+      "loss": 8.916,
+      "step": 99310
+    },
+    {
+      "epoch": 0.4959924092986092,
+      "grad_norm": 0.09705643355846405,
+      "learning_rate": 1.515807654759819e-05,
+      "loss": 8.9191,
+      "step": 99320
+    },
+    {
+      "epoch": 0.49604234812354864,
+      "grad_norm": 0.08930245041847229,
+      "learning_rate": 1.5156574632656639e-05,
+      "loss": 8.9,
+      "step": 99330
+    },
+    {
+      "epoch": 0.4960922869484881,
+      "grad_norm": 0.09772713482379913,
+      "learning_rate": 1.5155072717715086e-05,
+      "loss": 8.9179,
+      "step": 99340
+    },
+    {
+      "epoch": 0.49614222577342754,
+      "grad_norm": 0.0901922956109047,
+      "learning_rate": 1.5153570802773537e-05,
+      "loss": 8.9004,
+      "step": 99350
+    },
+    {
+      "epoch": 0.496192164598367,
+      "grad_norm": 0.09526271373033524,
+      "learning_rate": 1.5152068887831986e-05,
+      "loss": 8.9225,
+      "step": 99360
+    },
+    {
+      "epoch": 0.49624210342330644,
+      "grad_norm": 0.09357525408267975,
+      "learning_rate": 1.5150566972890438e-05,
+      "loss": 8.9093,
+      "step": 99370
+    },
+    {
+      "epoch": 0.4962920422482459,
+      "grad_norm": 0.09719167649745941,
+      "learning_rate": 1.5149065057948886e-05,
+      "loss": 8.9092,
+      "step": 99380
+    },
+    {
+      "epoch": 0.49634198107318533,
+      "grad_norm": 0.09308057278394699,
+      "learning_rate": 1.5147563143007333e-05,
+      "loss": 8.8939,
+      "step": 99390
+    },
+    {
+      "epoch": 0.4963919198981248,
+      "grad_norm": 0.0987476035952568,
+      "learning_rate": 1.5146061228065785e-05,
+      "loss": 8.9098,
+      "step": 99400
+    },
+    {
+      "epoch": 0.49644185872306423,
+      "grad_norm": 0.09410141408443451,
+      "learning_rate": 1.5144559313124233e-05,
+      "loss": 8.9034,
+      "step": 99410
+    },
+    {
+      "epoch": 0.4964917975480037,
+      "grad_norm": 0.09079299867153168,
+      "learning_rate": 1.5143057398182685e-05,
+      "loss": 8.9117,
+      "step": 99420
+    },
+    {
+      "epoch": 0.49654173637294313,
+      "grad_norm": 0.0934637263417244,
+      "learning_rate": 1.5141555483241134e-05,
+      "loss": 8.9108,
+      "step": 99430
+    },
+    {
+      "epoch": 0.4965916751978826,
+      "grad_norm": 0.09323903918266296,
+      "learning_rate": 1.514005356829958e-05,
+      "loss": 8.9113,
+      "step": 99440
+    },
+    {
+      "epoch": 0.49664161402282203,
+      "grad_norm": 0.09298430383205414,
+      "learning_rate": 1.5138551653358032e-05,
+      "loss": 8.9119,
+      "step": 99450
+    },
+    {
+      "epoch": 0.4966915528477615,
+      "grad_norm": 0.09651913493871689,
+      "learning_rate": 1.5137049738416481e-05,
+      "loss": 8.8993,
+      "step": 99460
+    },
+    {
+      "epoch": 0.49674149167270093,
+      "grad_norm": 0.0914125069975853,
+      "learning_rate": 1.5135547823474933e-05,
+      "loss": 8.9252,
+      "step": 99470
+    },
+    {
+      "epoch": 0.4967914304976404,
+      "grad_norm": 0.09153977036476135,
+      "learning_rate": 1.5134045908533381e-05,
+      "loss": 8.9252,
+      "step": 99480
+    },
+    {
+      "epoch": 0.49684136932257983,
+      "grad_norm": 0.08837665617465973,
+      "learning_rate": 1.5132543993591828e-05,
+      "loss": 8.8961,
+      "step": 99490
+    },
+    {
+      "epoch": 0.4968913081475193,
+      "grad_norm": 0.0936015248298645,
+      "learning_rate": 1.513104207865028e-05,
+      "loss": 8.9022,
+      "step": 99500
+    },
+    {
+      "epoch": 0.49694124697245873,
+      "grad_norm": 0.09733084589242935,
+      "learning_rate": 1.5129540163708728e-05,
+      "loss": 8.8925,
+      "step": 99510
+    },
+    {
+      "epoch": 0.4969911857973982,
+      "grad_norm": 0.09360146522521973,
+      "learning_rate": 1.512803824876718e-05,
+      "loss": 8.903,
+      "step": 99520
+    },
+    {
+      "epoch": 0.4970411246223376,
+      "grad_norm": 0.09262694418430328,
+      "learning_rate": 1.5126536333825629e-05,
+      "loss": 8.9168,
+      "step": 99530
+    },
+    {
+      "epoch": 0.4970910634472771,
+      "grad_norm": 0.0910312682390213,
+      "learning_rate": 1.5125034418884076e-05,
+      "loss": 8.9155,
+      "step": 99540
+    },
+    {
+      "epoch": 0.4971410022722165,
+      "grad_norm": 0.09175118058919907,
+      "learning_rate": 1.5123532503942527e-05,
+      "loss": 8.905,
+      "step": 99550
+    },
+    {
+      "epoch": 0.497190941097156,
+      "grad_norm": 0.08740215748548508,
+      "learning_rate": 1.5122030589000976e-05,
+      "loss": 8.9235,
+      "step": 99560
+    },
+    {
+      "epoch": 0.4972408799220954,
+      "grad_norm": 0.09946295619010925,
+      "learning_rate": 1.5120528674059428e-05,
+      "loss": 8.9026,
+      "step": 99570
+    },
+    {
+      "epoch": 0.4972908187470349,
+      "grad_norm": 0.09400948882102966,
+      "learning_rate": 1.5119026759117876e-05,
+      "loss": 8.9128,
+      "step": 99580
+    },
+    {
+      "epoch": 0.4973407575719743,
+      "grad_norm": 0.09512431174516678,
+      "learning_rate": 1.5117524844176323e-05,
+      "loss": 8.9141,
+      "step": 99590
+    },
+    {
+      "epoch": 0.4973906963969138,
+      "grad_norm": 0.09360211342573166,
+      "learning_rate": 1.5116022929234775e-05,
+      "loss": 8.9206,
+      "step": 99600
+    },
+    {
+      "epoch": 0.4974406352218532,
+      "grad_norm": 0.08853219449520111,
+      "learning_rate": 1.5114521014293223e-05,
+      "loss": 8.9061,
+      "step": 99610
+    },
+    {
+      "epoch": 0.4974905740467927,
+      "grad_norm": 0.09550435096025467,
+      "learning_rate": 1.5113019099351675e-05,
+      "loss": 8.909,
+      "step": 99620
+    },
+    {
+      "epoch": 0.4975405128717321,
+      "grad_norm": 0.09092508256435394,
+      "learning_rate": 1.5111517184410124e-05,
+      "loss": 8.9078,
+      "step": 99630
+    },
+    {
+      "epoch": 0.4975904516966716,
+      "grad_norm": 0.09557131677865982,
+      "learning_rate": 1.5110015269468574e-05,
+      "loss": 8.9048,
+      "step": 99640
+    },
+    {
+      "epoch": 0.497640390521611,
+      "grad_norm": 0.09490132331848145,
+      "learning_rate": 1.5108513354527022e-05,
+      "loss": 8.9167,
+      "step": 99650
+    },
+    {
+      "epoch": 0.4976903293465505,
+      "grad_norm": 0.09305530786514282,
+      "learning_rate": 1.5107011439585471e-05,
+      "loss": 8.9106,
+      "step": 99660
+    },
+    {
+      "epoch": 0.4977402681714899,
+      "grad_norm": 0.09075649082660675,
+      "learning_rate": 1.5105509524643923e-05,
+      "loss": 8.9161,
+      "step": 99670
+    },
+    {
+      "epoch": 0.4977902069964294,
+      "grad_norm": 0.0945330560207367,
+      "learning_rate": 1.5104007609702371e-05,
+      "loss": 8.9061,
+      "step": 99680
+    },
+    {
+      "epoch": 0.4978401458213688,
+      "grad_norm": 0.09516996145248413,
+      "learning_rate": 1.5102505694760821e-05,
+      "loss": 8.9028,
+      "step": 99690
+    },
+    {
+      "epoch": 0.4978900846463083,
+      "grad_norm": 0.09055611491203308,
+      "learning_rate": 1.510100377981927e-05,
+      "loss": 8.9215,
+      "step": 99700
+    },
+    {
+      "epoch": 0.4979400234712477,
+      "grad_norm": 0.0914728119969368,
+      "learning_rate": 1.5099501864877718e-05,
+      "loss": 8.9231,
+      "step": 99710
+    },
+    {
+      "epoch": 0.4979899622961872,
+      "grad_norm": 0.08997715264558792,
+      "learning_rate": 1.509799994993617e-05,
+      "loss": 8.9054,
+      "step": 99720
+    },
+    {
+      "epoch": 0.4980399011211266,
+      "grad_norm": 0.10086820274591446,
+      "learning_rate": 1.5096498034994619e-05,
+      "loss": 8.9084,
+      "step": 99730
+    },
+    {
+      "epoch": 0.4980898399460661,
+      "grad_norm": 0.09388542920351028,
+      "learning_rate": 1.5094996120053069e-05,
+      "loss": 8.9193,
+      "step": 99740
+    },
+    {
+      "epoch": 0.4981397787710055,
+      "grad_norm": 0.09304860979318619,
+      "learning_rate": 1.5093494205111517e-05,
+      "loss": 8.9002,
+      "step": 99750
+    },
+    {
+      "epoch": 0.498189717595945,
+      "grad_norm": 0.0892772451043129,
+      "learning_rate": 1.5091992290169966e-05,
+      "loss": 8.8994,
+      "step": 99760
+    },
+    {
+      "epoch": 0.4982396564208844,
+      "grad_norm": 0.09076282382011414,
+      "learning_rate": 1.5090490375228418e-05,
+      "loss": 8.9014,
+      "step": 99770
+    },
+    {
+      "epoch": 0.4982895952458239,
+      "grad_norm": 0.09353473037481308,
+      "learning_rate": 1.5088988460286866e-05,
+      "loss": 8.9208,
+      "step": 99780
+    },
+    {
+      "epoch": 0.4983395340707633,
+      "grad_norm": 0.09274925291538239,
+      "learning_rate": 1.5087486545345316e-05,
+      "loss": 8.9072,
+      "step": 99790
+    },
+    {
+      "epoch": 0.4983894728957028,
+      "grad_norm": 0.09765703231096268,
+      "learning_rate": 1.5085984630403765e-05,
+      "loss": 8.9014,
+      "step": 99800
+    },
+    {
+      "epoch": 0.4984394117206422,
+      "grad_norm": 0.08879239112138748,
+      "learning_rate": 1.5084482715462213e-05,
+      "loss": 8.9185,
+      "step": 99810
+    },
+    {
+      "epoch": 0.4984893505455817,
+      "grad_norm": 0.09869293868541718,
+      "learning_rate": 1.5082980800520665e-05,
+      "loss": 8.9048,
+      "step": 99820
+    },
+    {
+      "epoch": 0.4985392893705211,
+      "grad_norm": 0.09303834289312363,
+      "learning_rate": 1.5081478885579114e-05,
+      "loss": 8.9107,
+      "step": 99830
+    },
+    {
+      "epoch": 0.4985892281954606,
+      "grad_norm": 0.09813646227121353,
+      "learning_rate": 1.5079976970637564e-05,
+      "loss": 8.9082,
+      "step": 99840
+    },
+    {
+      "epoch": 0.4986391670204,
+      "grad_norm": 0.09101364761590958,
+      "learning_rate": 1.5078475055696012e-05,
+      "loss": 8.8986,
+      "step": 99850
+    },
+    {
+      "epoch": 0.4986891058453395,
+      "grad_norm": 0.09561823308467865,
+      "learning_rate": 1.5076973140754461e-05,
+      "loss": 8.901,
+      "step": 99860
+    },
+    {
+      "epoch": 0.4987390446702789,
+      "grad_norm": 0.09230560064315796,
+      "learning_rate": 1.5075471225812913e-05,
+      "loss": 8.9104,
+      "step": 99870
+    },
+    {
+      "epoch": 0.49878898349521833,
+      "grad_norm": 0.10049329698085785,
+      "learning_rate": 1.5073969310871361e-05,
+      "loss": 8.9125,
+      "step": 99880
+    },
+    {
+      "epoch": 0.4988389223201578,
+      "grad_norm": 0.08787727355957031,
+      "learning_rate": 1.5072467395929811e-05,
+      "loss": 8.9013,
+      "step": 99890
+    },
+    {
+      "epoch": 0.49888886114509723,
+      "grad_norm": 0.09034966677427292,
+      "learning_rate": 1.507096548098826e-05,
+      "loss": 8.9037,
+      "step": 99900
+    },
+    {
+      "epoch": 0.4989387999700367,
+      "grad_norm": 0.09500369429588318,
+      "learning_rate": 1.5069463566046708e-05,
+      "loss": 8.8933,
+      "step": 99910
+    },
+    {
+      "epoch": 0.4989887387949761,
+      "grad_norm": 0.09762393683195114,
+      "learning_rate": 1.506796165110516e-05,
+      "loss": 8.9136,
+      "step": 99920
+    },
+    {
+      "epoch": 0.4990386776199156,
+      "grad_norm": 0.09636267274618149,
+      "learning_rate": 1.5066459736163609e-05,
+      "loss": 8.9171,
+      "step": 99930
+    },
+    {
+      "epoch": 0.499088616444855,
+      "grad_norm": 0.09169818460941315,
+      "learning_rate": 1.5064957821222059e-05,
+      "loss": 8.9056,
+      "step": 99940
+    },
+    {
+      "epoch": 0.4991385552697945,
+      "grad_norm": 0.09505058825016022,
+      "learning_rate": 1.5063455906280507e-05,
+      "loss": 8.9185,
+      "step": 99950
+    },
+    {
+      "epoch": 0.4991884940947339,
+      "grad_norm": 0.09004746377468109,
+      "learning_rate": 1.5061953991338956e-05,
+      "loss": 8.9026,
+      "step": 99960
+    },
+    {
+      "epoch": 0.4992384329196734,
+      "grad_norm": 0.08870364725589752,
+      "learning_rate": 1.5060452076397408e-05,
+      "loss": 8.9053,
+      "step": 99970
+    },
+    {
+      "epoch": 0.4992883717446128,
+      "grad_norm": 0.08827803283929825,
+      "learning_rate": 1.5058950161455856e-05,
+      "loss": 8.9115,
+      "step": 99980
+    },
+    {
+      "epoch": 0.4993383105695523,
+      "grad_norm": 0.09465722739696503,
+      "learning_rate": 1.5057448246514306e-05,
+      "loss": 8.9056,
+      "step": 99990
+    },
+    {
+      "epoch": 0.4993882493944917,
+      "grad_norm": 0.0972822979092598,
+      "learning_rate": 1.5055946331572755e-05,
+      "loss": 8.9103,
+      "step": 100000
+    },
+    {
+      "epoch": 0.4994381882194312,
+      "grad_norm": 0.0914982259273529,
+      "learning_rate": 1.5054444416631207e-05,
+      "loss": 8.9047,
+      "step": 100010
+    },
+    {
+      "epoch": 0.4994881270443706,
+      "grad_norm": 0.09300156682729721,
+      "learning_rate": 1.5052942501689655e-05,
+      "loss": 8.9028,
+      "step": 100020
+    },
+    {
+      "epoch": 0.4995380658693101,
+      "grad_norm": 0.0923018679022789,
+      "learning_rate": 1.5051440586748104e-05,
+      "loss": 8.9091,
+      "step": 100030
+    },
+    {
+      "epoch": 0.4995880046942495,
+      "grad_norm": 0.09287258982658386,
+      "learning_rate": 1.5049938671806554e-05,
+      "loss": 8.9109,
+      "step": 100040
+    },
+    {
+      "epoch": 0.499637943519189,
+      "grad_norm": 0.0904991626739502,
+      "learning_rate": 1.5048436756865002e-05,
+      "loss": 8.9109,
+      "step": 100050
+    },
+    {
+      "epoch": 0.4996878823441284,
+      "grad_norm": 0.09570524841547012,
+      "learning_rate": 1.5046934841923454e-05,
+      "loss": 8.9005,
+      "step": 100060
+    },
+    {
+      "epoch": 0.4997378211690679,
+      "grad_norm": 0.09132358431816101,
+      "learning_rate": 1.5045432926981903e-05,
+      "loss": 8.9182,
+      "step": 100070
+    },
+    {
+      "epoch": 0.4997877599940073,
+      "grad_norm": 0.09124463051557541,
+      "learning_rate": 1.5043931012040351e-05,
+      "loss": 8.9059,
+      "step": 100080
+    },
+    {
+      "epoch": 0.4998376988189468,
+      "grad_norm": 0.08917669951915741,
+      "learning_rate": 1.5042429097098801e-05,
+      "loss": 8.9083,
+      "step": 100090
+    },
+    {
+      "epoch": 0.4998876376438862,
+      "grad_norm": 0.09558670222759247,
+      "learning_rate": 1.504092718215725e-05,
+      "loss": 8.9074,
+      "step": 100100
+    },
+    {
+      "epoch": 0.4999375764688257,
+      "grad_norm": 0.09686680883169174,
+      "learning_rate": 1.5039425267215702e-05,
+      "loss": 8.9087,
+      "step": 100110
+    },
+    {
+      "epoch": 0.4999875152937651,
+      "grad_norm": 0.08919398486614227,
+      "learning_rate": 1.503792335227415e-05,
+      "loss": 8.911,
+      "step": 100120
+    },
+    {
+      "epoch": 0.5000374541187046,
+      "grad_norm": 0.09795615077018738,
+      "learning_rate": 1.5036421437332599e-05,
+      "loss": 8.8899,
+      "step": 100130
+    },
+    {
+      "epoch": 0.500087392943644,
+      "grad_norm": 0.09374431520700455,
+      "learning_rate": 1.5034919522391049e-05,
+      "loss": 8.896,
+      "step": 100140
+    },
+    {
+      "epoch": 0.5001373317685834,
+      "grad_norm": 0.09322655200958252,
+      "learning_rate": 1.5033417607449497e-05,
+      "loss": 8.901,
+      "step": 100150
+    },
+    {
+      "epoch": 0.500187270593523,
+      "grad_norm": 0.0963103175163269,
+      "learning_rate": 1.503191569250795e-05,
+      "loss": 8.9022,
+      "step": 100160
+    },
+    {
+      "epoch": 0.5002372094184624,
+      "grad_norm": 0.08915145695209503,
+      "learning_rate": 1.5030413777566398e-05,
+      "loss": 8.902,
+      "step": 100170
+    },
+    {
+      "epoch": 0.5002871482434018,
+      "grad_norm": 0.08930926024913788,
+      "learning_rate": 1.5028911862624846e-05,
+      "loss": 8.9118,
+      "step": 100180
+    },
+    {
+      "epoch": 0.5003370870683412,
+      "grad_norm": 0.09767061471939087,
+      "learning_rate": 1.5027409947683296e-05,
+      "loss": 8.9057,
+      "step": 100190
+    },
+    {
+      "epoch": 0.5003870258932808,
+      "grad_norm": 0.09693524241447449,
+      "learning_rate": 1.5025908032741745e-05,
+      "loss": 8.8938,
+      "step": 100200
+    },
+    {
+      "epoch": 0.5004369647182202,
+      "grad_norm": 0.09122538566589355,
+      "learning_rate": 1.5024406117800197e-05,
+      "loss": 8.9038,
+      "step": 100210
+    },
+    {
+      "epoch": 0.5004869035431596,
+      "grad_norm": 0.0925898402929306,
+      "learning_rate": 1.5022904202858645e-05,
+      "loss": 8.9147,
+      "step": 100220
+    },
+    {
+      "epoch": 0.500536842368099,
+      "grad_norm": 0.08967383205890656,
+      "learning_rate": 1.5021402287917094e-05,
+      "loss": 8.9091,
+      "step": 100230
+    },
+    {
+      "epoch": 0.5005867811930386,
+      "grad_norm": 0.09207455813884735,
+      "learning_rate": 1.5019900372975544e-05,
+      "loss": 8.9131,
+      "step": 100240
+    },
+    {
+      "epoch": 0.500636720017978,
+      "grad_norm": 0.09674876183271408,
+      "learning_rate": 1.5018398458033992e-05,
+      "loss": 8.9054,
+      "step": 100250
+    },
+    {
+      "epoch": 0.5006866588429174,
+      "grad_norm": 0.09284425526857376,
+      "learning_rate": 1.5016896543092444e-05,
+      "loss": 8.9048,
+      "step": 100260
+    },
+    {
+      "epoch": 0.5007365976678568,
+      "grad_norm": 0.09239511936903,
+      "learning_rate": 1.5015394628150893e-05,
+      "loss": 8.9162,
+      "step": 100270
+    },
+    {
+      "epoch": 0.5007865364927964,
+      "grad_norm": 0.0956454798579216,
+      "learning_rate": 1.5013892713209341e-05,
+      "loss": 8.9134,
+      "step": 100280
+    },
+    {
+      "epoch": 0.5008364753177358,
+      "grad_norm": 0.0959239974617958,
+      "learning_rate": 1.5012390798267791e-05,
+      "loss": 8.8996,
+      "step": 100290
+    },
+    {
+      "epoch": 0.5008864141426752,
+      "grad_norm": 0.0976717472076416,
+      "learning_rate": 1.501088888332624e-05,
+      "loss": 8.9049,
+      "step": 100300
+    },
+    {
+      "epoch": 0.5009363529676146,
+      "grad_norm": 0.09285130351781845,
+      "learning_rate": 1.5009386968384692e-05,
+      "loss": 8.9057,
+      "step": 100310
+    },
+    {
+      "epoch": 0.5009862917925542,
+      "grad_norm": 0.08767900615930557,
+      "learning_rate": 1.500788505344314e-05,
+      "loss": 8.9035,
+      "step": 100320
+    },
+    {
+      "epoch": 0.5010362306174936,
+      "grad_norm": 0.09420838207006454,
+      "learning_rate": 1.5006383138501592e-05,
+      "loss": 8.8962,
+      "step": 100330
+    },
+    {
+      "epoch": 0.501086169442433,
+      "grad_norm": 0.09375480562448502,
+      "learning_rate": 1.5004881223560039e-05,
+      "loss": 8.9055,
+      "step": 100340
+    },
+    {
+      "epoch": 0.5011361082673724,
+      "grad_norm": 0.09297312796115875,
+      "learning_rate": 1.5003379308618487e-05,
+      "loss": 8.9098,
+      "step": 100350
+    },
+    {
+      "epoch": 0.501186047092312,
+      "grad_norm": 0.09020781517028809,
+      "learning_rate": 1.500187739367694e-05,
+      "loss": 8.9313,
+      "step": 100360
+    },
+    {
+      "epoch": 0.5012359859172514,
+      "grad_norm": 0.09823616594076157,
+      "learning_rate": 1.5000375478735388e-05,
+      "loss": 8.9211,
+      "step": 100370
+    },
+    {
+      "epoch": 0.5012859247421908,
+      "grad_norm": 0.09264297038316727,
+      "learning_rate": 1.4998873563793838e-05,
+      "loss": 8.9147,
+      "step": 100380
+    },
+    {
+      "epoch": 0.5013358635671302,
+      "grad_norm": 0.09266399592161179,
+      "learning_rate": 1.4997371648852286e-05,
+      "loss": 8.9005,
+      "step": 100390
+    },
+    {
+      "epoch": 0.5013858023920698,
+      "grad_norm": 0.0926387682557106,
+      "learning_rate": 1.4995869733910737e-05,
+      "loss": 8.9014,
+      "step": 100400
+    },
+    {
+      "epoch": 0.5014357412170092,
+      "grad_norm": 0.09379039704799652,
+      "learning_rate": 1.4994367818969187e-05,
+      "loss": 8.9031,
+      "step": 100410
+    },
+    {
+      "epoch": 0.5014856800419486,
+      "grad_norm": 0.09653208404779434,
+      "learning_rate": 1.4992865904027635e-05,
+      "loss": 8.9117,
+      "step": 100420
+    },
+    {
+      "epoch": 0.501535618866888,
+      "grad_norm": 0.09672179073095322,
+      "learning_rate": 1.4991363989086086e-05,
+      "loss": 8.9151,
+      "step": 100430
+    },
+    {
+      "epoch": 0.5015855576918276,
+      "grad_norm": 0.09220828860998154,
+      "learning_rate": 1.4989862074144534e-05,
+      "loss": 8.918,
+      "step": 100440
+    },
+    {
+      "epoch": 0.501635496516767,
+      "grad_norm": 0.08738303184509277,
+      "learning_rate": 1.4988360159202984e-05,
+      "loss": 8.8946,
+      "step": 100450
+    },
+    {
+      "epoch": 0.5016854353417064,
+      "grad_norm": 0.0968918427824974,
+      "learning_rate": 1.4986858244261434e-05,
+      "loss": 8.9172,
+      "step": 100460
+    },
+    {
+      "epoch": 0.5017353741666458,
+      "grad_norm": 0.08795435726642609,
+      "learning_rate": 1.4985356329319883e-05,
+      "loss": 8.8957,
+      "step": 100470
+    },
+    {
+      "epoch": 0.5017853129915854,
+      "grad_norm": 0.09090543538331985,
+      "learning_rate": 1.4983854414378333e-05,
+      "loss": 8.9172,
+      "step": 100480
+    },
+    {
+      "epoch": 0.5018352518165248,
+      "grad_norm": 0.09485392272472382,
+      "learning_rate": 1.4982352499436782e-05,
+      "loss": 8.908,
+      "step": 100490
+    },
+    {
+      "epoch": 0.5018851906414642,
+      "grad_norm": 0.094363734126091,
+      "learning_rate": 1.4980850584495232e-05,
+      "loss": 8.9137,
+      "step": 100500
+    },
+    {
+      "epoch": 0.5019351294664036,
+      "grad_norm": 0.08762264251708984,
+      "learning_rate": 1.4979348669553682e-05,
+      "loss": 8.916,
+      "step": 100510
+    },
+    {
+      "epoch": 0.5019850682913432,
+      "grad_norm": 0.09093135595321655,
+      "learning_rate": 1.497784675461213e-05,
+      "loss": 8.9092,
+      "step": 100520
+    },
+    {
+      "epoch": 0.5020350071162826,
+      "grad_norm": 0.08741986751556396,
+      "learning_rate": 1.497634483967058e-05,
+      "loss": 8.9032,
+      "step": 100530
+    },
+    {
+      "epoch": 0.502084945941222,
+      "grad_norm": 0.09750941395759583,
+      "learning_rate": 1.4974842924729029e-05,
+      "loss": 8.908,
+      "step": 100540
+    },
+    {
+      "epoch": 0.5021348847661614,
+      "grad_norm": 0.09141114354133606,
+      "learning_rate": 1.497334100978748e-05,
+      "loss": 8.8956,
+      "step": 100550
+    },
+    {
+      "epoch": 0.502184823591101,
+      "grad_norm": 0.08919312804937363,
+      "learning_rate": 1.497183909484593e-05,
+      "loss": 8.9119,
+      "step": 100560
+    },
+    {
+      "epoch": 0.5022347624160404,
+      "grad_norm": 0.09717653691768646,
+      "learning_rate": 1.497033717990438e-05,
+      "loss": 8.8956,
+      "step": 100570
+    },
+    {
+      "epoch": 0.5022847012409798,
+      "grad_norm": 0.09845616668462753,
+      "learning_rate": 1.4968835264962828e-05,
+      "loss": 8.8947,
+      "step": 100580
+    },
+    {
+      "epoch": 0.5023346400659192,
+      "grad_norm": 0.09342017769813538,
+      "learning_rate": 1.4967333350021277e-05,
+      "loss": 8.903,
+      "step": 100590
+    },
+    {
+      "epoch": 0.5023845788908587,
+      "grad_norm": 0.08772523701190948,
+      "learning_rate": 1.4965831435079727e-05,
+      "loss": 8.9031,
+      "step": 100600
+    },
+    {
+      "epoch": 0.5024345177157982,
+      "grad_norm": 0.09380695968866348,
+      "learning_rate": 1.4964329520138177e-05,
+      "loss": 8.9074,
+      "step": 100610
+    },
+    {
+      "epoch": 0.5024844565407376,
+      "grad_norm": 0.09516066312789917,
+      "learning_rate": 1.4962827605196627e-05,
+      "loss": 8.8997,
+      "step": 100620
+    },
+    {
+      "epoch": 0.502534395365677,
+      "grad_norm": 0.0951061025261879,
+      "learning_rate": 1.4961325690255076e-05,
+      "loss": 8.8925,
+      "step": 100630
+    },
+    {
+      "epoch": 0.5025843341906165,
+      "grad_norm": 0.09456291794776917,
+      "learning_rate": 1.4959823775313524e-05,
+      "loss": 8.8921,
+      "step": 100640
+    },
+    {
+      "epoch": 0.502634273015556,
+      "grad_norm": 0.09144452959299088,
+      "learning_rate": 1.4958321860371974e-05,
+      "loss": 8.8916,
+      "step": 100650
+    },
+    {
+      "epoch": 0.5026842118404954,
+      "grad_norm": 0.09398289024829865,
+      "learning_rate": 1.4956819945430424e-05,
+      "loss": 8.9007,
+      "step": 100660
+    },
+    {
+      "epoch": 0.5027341506654348,
+      "grad_norm": 0.09931785613298416,
+      "learning_rate": 1.4955318030488875e-05,
+      "loss": 8.903,
+      "step": 100670
+    },
+    {
+      "epoch": 0.5027840894903743,
+      "grad_norm": 0.09348601847887039,
+      "learning_rate": 1.4953816115547323e-05,
+      "loss": 8.9051,
+      "step": 100680
+    },
+    {
+      "epoch": 0.5028340283153138,
+      "grad_norm": 0.08900678902864456,
+      "learning_rate": 1.4952314200605772e-05,
+      "loss": 8.9082,
+      "step": 100690
+    },
+    {
+      "epoch": 0.5028839671402532,
+      "grad_norm": 0.09216514974832535,
+      "learning_rate": 1.4950812285664222e-05,
+      "loss": 8.9075,
+      "step": 100700
+    },
+    {
+      "epoch": 0.5029339059651926,
+      "grad_norm": 0.09171604365110397,
+      "learning_rate": 1.4949310370722672e-05,
+      "loss": 8.9241,
+      "step": 100710
+    },
+    {
+      "epoch": 0.5029838447901321,
+      "grad_norm": 0.08902275562286377,
+      "learning_rate": 1.4947808455781122e-05,
+      "loss": 8.9145,
+      "step": 100720
+    },
+    {
+      "epoch": 0.5030337836150716,
+      "grad_norm": 0.09136269986629486,
+      "learning_rate": 1.4946306540839572e-05,
+      "loss": 8.9118,
+      "step": 100730
+    },
+    {
+      "epoch": 0.503083722440011,
+      "grad_norm": 0.09541922062635422,
+      "learning_rate": 1.4944804625898019e-05,
+      "loss": 8.9042,
+      "step": 100740
+    },
+    {
+      "epoch": 0.5031336612649504,
+      "grad_norm": 0.08974391967058182,
+      "learning_rate": 1.494330271095647e-05,
+      "loss": 8.9063,
+      "step": 100750
+    },
+    {
+      "epoch": 0.5031836000898899,
+      "grad_norm": 0.0941217839717865,
+      "learning_rate": 1.494180079601492e-05,
+      "loss": 8.9088,
+      "step": 100760
+    },
+    {
+      "epoch": 0.5032335389148294,
+      "grad_norm": 0.0899343267083168,
+      "learning_rate": 1.494029888107337e-05,
+      "loss": 8.8967,
+      "step": 100770
+    },
+    {
+      "epoch": 0.5032834777397688,
+      "grad_norm": 0.09517884254455566,
+      "learning_rate": 1.493879696613182e-05,
+      "loss": 8.8919,
+      "step": 100780
+    },
+    {
+      "epoch": 0.5033334165647082,
+      "grad_norm": 0.08698021620512009,
+      "learning_rate": 1.4937295051190267e-05,
+      "loss": 8.9165,
+      "step": 100790
+    },
+    {
+      "epoch": 0.5033833553896477,
+      "grad_norm": 0.09504187107086182,
+      "learning_rate": 1.4935793136248717e-05,
+      "loss": 8.9146,
+      "step": 100800
+    },
+    {
+      "epoch": 0.5034332942145872,
+      "grad_norm": 0.09907368570566177,
+      "learning_rate": 1.4934291221307167e-05,
+      "loss": 8.8958,
+      "step": 100810
+    },
+    {
+      "epoch": 0.5034832330395266,
+      "grad_norm": 0.09145436435937881,
+      "learning_rate": 1.4932789306365617e-05,
+      "loss": 8.9008,
+      "step": 100820
+    },
+    {
+      "epoch": 0.503533171864466,
+      "grad_norm": 0.09592849761247635,
+      "learning_rate": 1.4931287391424067e-05,
+      "loss": 8.9033,
+      "step": 100830
+    },
+    {
+      "epoch": 0.5035831106894055,
+      "grad_norm": 0.0957927331328392,
+      "learning_rate": 1.4929785476482514e-05,
+      "loss": 8.91,
+      "step": 100840
+    },
+    {
+      "epoch": 0.503633049514345,
+      "grad_norm": 0.09893249720335007,
+      "learning_rate": 1.4928283561540964e-05,
+      "loss": 8.9059,
+      "step": 100850
+    },
+    {
+      "epoch": 0.5036829883392844,
+      "grad_norm": 0.0861205905675888,
+      "learning_rate": 1.4926781646599414e-05,
+      "loss": 8.9054,
+      "step": 100860
+    },
+    {
+      "epoch": 0.5037329271642238,
+      "grad_norm": 0.08853903412818909,
+      "learning_rate": 1.4925279731657865e-05,
+      "loss": 8.9096,
+      "step": 100870
+    },
+    {
+      "epoch": 0.5037828659891632,
+      "grad_norm": 0.09132624417543411,
+      "learning_rate": 1.4923777816716315e-05,
+      "loss": 8.8916,
+      "step": 100880
+    },
+    {
+      "epoch": 0.5038328048141028,
+      "grad_norm": 0.08741272985935211,
+      "learning_rate": 1.4922275901774763e-05,
+      "loss": 8.8881,
+      "step": 100890
+    },
+    {
+      "epoch": 0.5038827436390422,
+      "grad_norm": 0.09232524782419205,
+      "learning_rate": 1.4920773986833212e-05,
+      "loss": 8.9061,
+      "step": 100900
+    },
+    {
+      "epoch": 0.5039326824639816,
+      "grad_norm": 0.09237165004014969,
+      "learning_rate": 1.4919272071891662e-05,
+      "loss": 8.9059,
+      "step": 100910
+    },
+    {
+      "epoch": 0.503982621288921,
+      "grad_norm": 0.09467703849077225,
+      "learning_rate": 1.4917770156950112e-05,
+      "loss": 8.9001,
+      "step": 100920
+    },
+    {
+      "epoch": 0.5040325601138606,
+      "grad_norm": 0.09434116631746292,
+      "learning_rate": 1.4916268242008562e-05,
+      "loss": 8.8948,
+      "step": 100930
+    },
+    {
+      "epoch": 0.5040824989388,
+      "grad_norm": 0.0881483182311058,
+      "learning_rate": 1.491476632706701e-05,
+      "loss": 8.8947,
+      "step": 100940
+    },
+    {
+      "epoch": 0.5041324377637394,
+      "grad_norm": 0.09188562631607056,
+      "learning_rate": 1.491326441212546e-05,
+      "loss": 8.9034,
+      "step": 100950
+    },
+    {
+      "epoch": 0.5041823765886788,
+      "grad_norm": 0.09108608961105347,
+      "learning_rate": 1.491176249718391e-05,
+      "loss": 8.8958,
+      "step": 100960
+    },
+    {
+      "epoch": 0.5042323154136183,
+      "grad_norm": 0.09667716175317764,
+      "learning_rate": 1.491026058224236e-05,
+      "loss": 8.8876,
+      "step": 100970
+    },
+    {
+      "epoch": 0.5042822542385578,
+      "grad_norm": 0.0894923135638237,
+      "learning_rate": 1.490875866730081e-05,
+      "loss": 8.8906,
+      "step": 100980
+    },
+    {
+      "epoch": 0.5043321930634972,
+      "grad_norm": 0.08996430784463882,
+      "learning_rate": 1.4907256752359258e-05,
+      "loss": 8.8851,
+      "step": 100990
+    },
+    {
+      "epoch": 0.5043821318884366,
+      "grad_norm": 0.0906483381986618,
+      "learning_rate": 1.4905754837417707e-05,
+      "loss": 8.8891,
+      "step": 101000
+    },
+    {
+      "epoch": 0.5044320707133761,
+      "grad_norm": 0.09572287648916245,
+      "learning_rate": 1.4904252922476157e-05,
+      "loss": 8.906,
+      "step": 101010
+    },
+    {
+      "epoch": 0.5044820095383156,
+      "grad_norm": 0.09146282821893692,
+      "learning_rate": 1.4902751007534607e-05,
+      "loss": 8.9083,
+      "step": 101020
+    },
+    {
+      "epoch": 0.504531948363255,
+      "grad_norm": 0.10019837319850922,
+      "learning_rate": 1.4901249092593057e-05,
+      "loss": 8.9084,
+      "step": 101030
+    },
+    {
+      "epoch": 0.5045818871881944,
+      "grad_norm": 0.09378591179847717,
+      "learning_rate": 1.4899747177651506e-05,
+      "loss": 8.899,
+      "step": 101040
+    },
+    {
+      "epoch": 0.5046318260131339,
+      "grad_norm": 0.0919695645570755,
+      "learning_rate": 1.4898245262709956e-05,
+      "loss": 8.899,
+      "step": 101050
+    },
+    {
+      "epoch": 0.5046817648380734,
+      "grad_norm": 0.09100176393985748,
+      "learning_rate": 1.4896743347768404e-05,
+      "loss": 8.9073,
+      "step": 101060
+    },
+    {
+      "epoch": 0.5047317036630128,
+      "grad_norm": 0.09281841665506363,
+      "learning_rate": 1.4895241432826855e-05,
+      "loss": 8.9013,
+      "step": 101070
+    },
+    {
+      "epoch": 0.5047816424879522,
+      "grad_norm": 0.09110787510871887,
+      "learning_rate": 1.4893739517885305e-05,
+      "loss": 8.885,
+      "step": 101080
+    },
+    {
+      "epoch": 0.5048315813128917,
+      "grad_norm": 0.09434214234352112,
+      "learning_rate": 1.4892237602943753e-05,
+      "loss": 8.9101,
+      "step": 101090
+    },
+    {
+      "epoch": 0.5048815201378312,
+      "grad_norm": 0.09112465381622314,
+      "learning_rate": 1.4890735688002203e-05,
+      "loss": 8.8961,
+      "step": 101100
+    },
+    {
+      "epoch": 0.5049314589627706,
+      "grad_norm": 0.0931396409869194,
+      "learning_rate": 1.4889233773060652e-05,
+      "loss": 8.9027,
+      "step": 101110
+    },
+    {
+      "epoch": 0.50498139778771,
+      "grad_norm": 0.0969899371266365,
+      "learning_rate": 1.4887731858119102e-05,
+      "loss": 8.8754,
+      "step": 101120
+    },
+    {
+      "epoch": 0.5050313366126495,
+      "grad_norm": 0.09737983345985413,
+      "learning_rate": 1.4886229943177552e-05,
+      "loss": 8.896,
+      "step": 101130
+    },
+    {
+      "epoch": 0.505081275437589,
+      "grad_norm": 0.0903887152671814,
+      "learning_rate": 1.4884728028236e-05,
+      "loss": 8.8901,
+      "step": 101140
+    },
+    {
+      "epoch": 0.5051312142625284,
+      "grad_norm": 0.09329243004322052,
+      "learning_rate": 1.4883226113294451e-05,
+      "loss": 8.9111,
+      "step": 101150
+    },
+    {
+      "epoch": 0.5051811530874678,
+      "grad_norm": 0.09709962457418442,
+      "learning_rate": 1.48817241983529e-05,
+      "loss": 8.9032,
+      "step": 101160
+    },
+    {
+      "epoch": 0.5052310919124073,
+      "grad_norm": 0.09550347924232483,
+      "learning_rate": 1.488022228341135e-05,
+      "loss": 8.8953,
+      "step": 101170
+    },
+    {
+      "epoch": 0.5052810307373468,
+      "grad_norm": 0.09034687280654907,
+      "learning_rate": 1.48787203684698e-05,
+      "loss": 8.8965,
+      "step": 101180
+    },
+    {
+      "epoch": 0.5053309695622862,
+      "grad_norm": 0.08858486264944077,
+      "learning_rate": 1.4877218453528248e-05,
+      "loss": 8.8872,
+      "step": 101190
+    },
+    {
+      "epoch": 0.5053809083872256,
+      "grad_norm": 0.09409047663211823,
+      "learning_rate": 1.4875716538586698e-05,
+      "loss": 8.8993,
+      "step": 101200
+    },
+    {
+      "epoch": 0.5054308472121651,
+      "grad_norm": 0.09161051362752914,
+      "learning_rate": 1.4874214623645149e-05,
+      "loss": 8.8977,
+      "step": 101210
+    },
+    {
+      "epoch": 0.5054807860371046,
+      "grad_norm": 0.09429701417684555,
+      "learning_rate": 1.4872712708703597e-05,
+      "loss": 8.9064,
+      "step": 101220
+    },
+    {
+      "epoch": 0.505530724862044,
+      "grad_norm": 0.09554371237754822,
+      "learning_rate": 1.4871210793762047e-05,
+      "loss": 8.8981,
+      "step": 101230
+    },
+    {
+      "epoch": 0.5055806636869834,
+      "grad_norm": 0.10310646891593933,
+      "learning_rate": 1.4869708878820496e-05,
+      "loss": 8.899,
+      "step": 101240
+    },
+    {
+      "epoch": 0.5056306025119229,
+      "grad_norm": 0.09066363424062729,
+      "learning_rate": 1.4868206963878946e-05,
+      "loss": 8.9038,
+      "step": 101250
+    },
+    {
+      "epoch": 0.5056805413368624,
+      "grad_norm": 0.09133642911911011,
+      "learning_rate": 1.4866705048937396e-05,
+      "loss": 8.893,
+      "step": 101260
+    },
+    {
+      "epoch": 0.5057304801618018,
+      "grad_norm": 0.09266974031925201,
+      "learning_rate": 1.4865203133995845e-05,
+      "loss": 8.8918,
+      "step": 101270
+    },
+    {
+      "epoch": 0.5057804189867412,
+      "grad_norm": 0.09828782081604004,
+      "learning_rate": 1.4863701219054295e-05,
+      "loss": 8.9079,
+      "step": 101280
+    },
+    {
+      "epoch": 0.5058303578116807,
+      "grad_norm": 0.09121903032064438,
+      "learning_rate": 1.4862199304112743e-05,
+      "loss": 8.9018,
+      "step": 101290
+    },
+    {
+      "epoch": 0.5058802966366202,
+      "grad_norm": 0.09173334389925003,
+      "learning_rate": 1.4860697389171193e-05,
+      "loss": 8.8951,
+      "step": 101300
+    },
+    {
+      "epoch": 0.5059302354615596,
+      "grad_norm": 0.09970010817050934,
+      "learning_rate": 1.4859195474229644e-05,
+      "loss": 8.8949,
+      "step": 101310
+    },
+    {
+      "epoch": 0.505980174286499,
+      "grad_norm": 0.0909634530544281,
+      "learning_rate": 1.4857693559288092e-05,
+      "loss": 8.8874,
+      "step": 101320
+    },
+    {
+      "epoch": 0.5060301131114385,
+      "grad_norm": 0.09572920203208923,
+      "learning_rate": 1.4856191644346542e-05,
+      "loss": 8.882,
+      "step": 101330
+    },
+    {
+      "epoch": 0.506080051936378,
+      "grad_norm": 0.08662997931241989,
+      "learning_rate": 1.485468972940499e-05,
+      "loss": 8.9034,
+      "step": 101340
+    },
+    {
+      "epoch": 0.5061299907613174,
+      "grad_norm": 0.09258969128131866,
+      "learning_rate": 1.4853187814463441e-05,
+      "loss": 8.8739,
+      "step": 101350
+    },
+    {
+      "epoch": 0.5061799295862568,
+      "grad_norm": 0.08739626407623291,
+      "learning_rate": 1.4851685899521891e-05,
+      "loss": 8.8976,
+      "step": 101360
+    },
+    {
+      "epoch": 0.5062298684111963,
+      "grad_norm": 0.09354405850172043,
+      "learning_rate": 1.4850183984580341e-05,
+      "loss": 8.8916,
+      "step": 101370
+    },
+    {
+      "epoch": 0.5062798072361357,
+      "grad_norm": 0.0959988608956337,
+      "learning_rate": 1.484868206963879e-05,
+      "loss": 8.9009,
+      "step": 101380
+    },
+    {
+      "epoch": 0.5063297460610752,
+      "grad_norm": 0.09055614471435547,
+      "learning_rate": 1.4847180154697238e-05,
+      "loss": 8.8916,
+      "step": 101390
+    },
+    {
+      "epoch": 0.5063796848860146,
+      "grad_norm": 0.09432981163263321,
+      "learning_rate": 1.4845678239755688e-05,
+      "loss": 8.9087,
+      "step": 101400
+    },
+    {
+      "epoch": 0.5064296237109541,
+      "grad_norm": 0.09564248472452164,
+      "learning_rate": 1.4844176324814139e-05,
+      "loss": 8.891,
+      "step": 101410
+    },
+    {
+      "epoch": 0.5064795625358935,
+      "grad_norm": 0.09236615151166916,
+      "learning_rate": 1.4842674409872589e-05,
+      "loss": 8.8975,
+      "step": 101420
+    },
+    {
+      "epoch": 0.506529501360833,
+      "grad_norm": 0.09159623831510544,
+      "learning_rate": 1.4841172494931037e-05,
+      "loss": 8.9015,
+      "step": 101430
+    },
+    {
+      "epoch": 0.5065794401857724,
+      "grad_norm": 0.09691286832094193,
+      "learning_rate": 1.4839670579989487e-05,
+      "loss": 8.8919,
+      "step": 101440
+    },
+    {
+      "epoch": 0.5066293790107119,
+      "grad_norm": 0.09283226728439331,
+      "learning_rate": 1.4838168665047936e-05,
+      "loss": 8.8946,
+      "step": 101450
+    },
+    {
+      "epoch": 0.5066793178356513,
+      "grad_norm": 0.09346681088209152,
+      "learning_rate": 1.4836666750106386e-05,
+      "loss": 8.8962,
+      "step": 101460
+    },
+    {
+      "epoch": 0.5067292566605908,
+      "grad_norm": 0.09086382389068604,
+      "learning_rate": 1.4835164835164836e-05,
+      "loss": 8.8942,
+      "step": 101470
+    },
+    {
+      "epoch": 0.5067791954855302,
+      "grad_norm": 0.09305468201637268,
+      "learning_rate": 1.4833662920223285e-05,
+      "loss": 8.8986,
+      "step": 101480
+    },
+    {
+      "epoch": 0.5068291343104697,
+      "grad_norm": 0.09416045248508453,
+      "learning_rate": 1.4832161005281735e-05,
+      "loss": 8.8897,
+      "step": 101490
+    },
+    {
+      "epoch": 0.5068790731354091,
+      "grad_norm": 0.09527656435966492,
+      "learning_rate": 1.4830659090340183e-05,
+      "loss": 8.8883,
+      "step": 101500
+    },
+    {
+      "epoch": 0.5069290119603486,
+      "grad_norm": 0.09679878503084183,
+      "learning_rate": 1.4829157175398634e-05,
+      "loss": 8.8931,
+      "step": 101510
+    },
+    {
+      "epoch": 0.506978950785288,
+      "grad_norm": 0.09845301508903503,
+      "learning_rate": 1.4827655260457084e-05,
+      "loss": 8.8859,
+      "step": 101520
+    },
+    {
+      "epoch": 0.5070288896102275,
+      "grad_norm": 0.09258875995874405,
+      "learning_rate": 1.4826153345515534e-05,
+      "loss": 8.8809,
+      "step": 101530
+    },
+    {
+      "epoch": 0.5070788284351669,
+      "grad_norm": 0.08814681321382523,
+      "learning_rate": 1.4824651430573982e-05,
+      "loss": 8.8913,
+      "step": 101540
+    },
+    {
+      "epoch": 0.5071287672601064,
+      "grad_norm": 0.09510982036590576,
+      "learning_rate": 1.4823149515632431e-05,
+      "loss": 8.9071,
+      "step": 101550
+    },
+    {
+      "epoch": 0.5071787060850458,
+      "grad_norm": 0.09275613725185394,
+      "learning_rate": 1.4821647600690881e-05,
+      "loss": 8.9034,
+      "step": 101560
+    },
+    {
+      "epoch": 0.5072286449099853,
+      "grad_norm": 0.09569612890481949,
+      "learning_rate": 1.4820145685749331e-05,
+      "loss": 8.8981,
+      "step": 101570
+    },
+    {
+      "epoch": 0.5072785837349247,
+      "grad_norm": 0.10368134826421738,
+      "learning_rate": 1.4818643770807781e-05,
+      "loss": 8.8932,
+      "step": 101580
+    },
+    {
+      "epoch": 0.5073285225598642,
+      "grad_norm": 0.0873536765575409,
+      "learning_rate": 1.481714185586623e-05,
+      "loss": 8.8866,
+      "step": 101590
+    },
+    {
+      "epoch": 0.5073784613848036,
+      "grad_norm": 0.0927792489528656,
+      "learning_rate": 1.4815639940924678e-05,
+      "loss": 8.8916,
+      "step": 101600
+    },
+    {
+      "epoch": 0.5074284002097431,
+      "grad_norm": 0.09179840981960297,
+      "learning_rate": 1.4814138025983129e-05,
+      "loss": 8.8925,
+      "step": 101610
+    },
+    {
+      "epoch": 0.5074783390346825,
+      "grad_norm": 0.09306514263153076,
+      "learning_rate": 1.4812636111041579e-05,
+      "loss": 8.9127,
+      "step": 101620
+    },
+    {
+      "epoch": 0.507528277859622,
+      "grad_norm": 0.09609080106019974,
+      "learning_rate": 1.4811134196100029e-05,
+      "loss": 8.8901,
+      "step": 101630
+    },
+    {
+      "epoch": 0.5075782166845614,
+      "grad_norm": 0.09421303868293762,
+      "learning_rate": 1.4809632281158477e-05,
+      "loss": 8.8944,
+      "step": 101640
+    },
+    {
+      "epoch": 0.5076281555095009,
+      "grad_norm": 0.09363938122987747,
+      "learning_rate": 1.4808130366216926e-05,
+      "loss": 8.9101,
+      "step": 101650
+    },
+    {
+      "epoch": 0.5076780943344403,
+      "grad_norm": 0.09263626486063004,
+      "learning_rate": 1.4806628451275376e-05,
+      "loss": 8.8904,
+      "step": 101660
+    },
+    {
+      "epoch": 0.5077280331593798,
+      "grad_norm": 0.08773566037416458,
+      "learning_rate": 1.4805126536333826e-05,
+      "loss": 8.8914,
+      "step": 101670
+    },
+    {
+      "epoch": 0.5077779719843192,
+      "grad_norm": 0.09525804966688156,
+      "learning_rate": 1.4803624621392276e-05,
+      "loss": 8.8931,
+      "step": 101680
+    },
+    {
+      "epoch": 0.5078279108092587,
+      "grad_norm": 0.09385735541582108,
+      "learning_rate": 1.4802122706450727e-05,
+      "loss": 8.904,
+      "step": 101690
+    },
+    {
+      "epoch": 0.5078778496341981,
+      "grad_norm": 0.0941053256392479,
+      "learning_rate": 1.4800620791509173e-05,
+      "loss": 8.8882,
+      "step": 101700
+    },
+    {
+      "epoch": 0.5079277884591376,
+      "grad_norm": 0.10088019073009491,
+      "learning_rate": 1.4799118876567624e-05,
+      "loss": 8.8875,
+      "step": 101710
+    },
+    {
+      "epoch": 0.507977727284077,
+      "grad_norm": 0.09125470370054245,
+      "learning_rate": 1.4797616961626074e-05,
+      "loss": 8.8858,
+      "step": 101720
+    },
+    {
+      "epoch": 0.5080276661090165,
+      "grad_norm": 0.08886610716581345,
+      "learning_rate": 1.4796115046684524e-05,
+      "loss": 8.8908,
+      "step": 101730
+    },
+    {
+      "epoch": 0.5080776049339559,
+      "grad_norm": 0.09195410460233688,
+      "learning_rate": 1.4794613131742974e-05,
+      "loss": 8.8764,
+      "step": 101740
+    },
+    {
+      "epoch": 0.5081275437588954,
+      "grad_norm": 0.09641274064779282,
+      "learning_rate": 1.4793111216801421e-05,
+      "loss": 8.88,
+      "step": 101750
+    },
+    {
+      "epoch": 0.5081774825838348,
+      "grad_norm": 0.08937504887580872,
+      "learning_rate": 1.4791609301859871e-05,
+      "loss": 8.8912,
+      "step": 101760
+    },
+    {
+      "epoch": 0.5082274214087743,
+      "grad_norm": 0.10038053244352341,
+      "learning_rate": 1.4790107386918321e-05,
+      "loss": 8.8898,
+      "step": 101770
+    },
+    {
+      "epoch": 0.5082773602337137,
+      "grad_norm": 0.08841981738805771,
+      "learning_rate": 1.4788605471976772e-05,
+      "loss": 8.902,
+      "step": 101780
+    },
+    {
+      "epoch": 0.5083272990586531,
+      "grad_norm": 0.09414731711149216,
+      "learning_rate": 1.4787103557035222e-05,
+      "loss": 8.8862,
+      "step": 101790
+    },
+    {
+      "epoch": 0.5083772378835926,
+      "grad_norm": 0.09005799144506454,
+      "learning_rate": 1.4785601642093668e-05,
+      "loss": 8.884,
+      "step": 101800
+    },
+    {
+      "epoch": 0.5084271767085321,
+      "grad_norm": 0.09526651352643967,
+      "learning_rate": 1.4784099727152119e-05,
+      "loss": 8.8958,
+      "step": 101810
+    },
+    {
+      "epoch": 0.5084771155334715,
+      "grad_norm": 0.094487763941288,
+      "learning_rate": 1.4782597812210569e-05,
+      "loss": 8.9035,
+      "step": 101820
+    },
+    {
+      "epoch": 0.508527054358411,
+      "grad_norm": 0.0912264958024025,
+      "learning_rate": 1.4781095897269019e-05,
+      "loss": 8.8972,
+      "step": 101830
+    },
+    {
+      "epoch": 0.5085769931833504,
+      "grad_norm": 0.09153269231319427,
+      "learning_rate": 1.477959398232747e-05,
+      "loss": 8.8956,
+      "step": 101840
+    },
+    {
+      "epoch": 0.5086269320082899,
+      "grad_norm": 0.09265515953302383,
+      "learning_rate": 1.4778092067385918e-05,
+      "loss": 8.8747,
+      "step": 101850
+    },
+    {
+      "epoch": 0.5086768708332293,
+      "grad_norm": 0.09390241652727127,
+      "learning_rate": 1.4776590152444366e-05,
+      "loss": 8.9175,
+      "step": 101860
+    },
+    {
+      "epoch": 0.5087268096581687,
+      "grad_norm": 0.09035301208496094,
+      "learning_rate": 1.4775088237502816e-05,
+      "loss": 8.8821,
+      "step": 101870
+    },
+    {
+      "epoch": 0.5087767484831082,
+      "grad_norm": 0.08922744542360306,
+      "learning_rate": 1.4773586322561267e-05,
+      "loss": 8.8883,
+      "step": 101880
+    },
+    {
+      "epoch": 0.5088266873080476,
+      "grad_norm": 0.09103063493967056,
+      "learning_rate": 1.4772084407619717e-05,
+      "loss": 8.8571,
+      "step": 101890
+    },
+    {
+      "epoch": 0.5088766261329871,
+      "grad_norm": 0.09650906175374985,
+      "learning_rate": 1.4770582492678165e-05,
+      "loss": 8.884,
+      "step": 101900
+    },
+    {
+      "epoch": 0.5089265649579265,
+      "grad_norm": 0.09376221150159836,
+      "learning_rate": 1.4769080577736614e-05,
+      "loss": 8.8981,
+      "step": 101910
+    },
+    {
+      "epoch": 0.508976503782866,
+      "grad_norm": 0.0935838520526886,
+      "learning_rate": 1.4767578662795064e-05,
+      "loss": 8.8788,
+      "step": 101920
+    },
+    {
+      "epoch": 0.5090264426078054,
+      "grad_norm": 0.09474118798971176,
+      "learning_rate": 1.4766076747853514e-05,
+      "loss": 8.8921,
+      "step": 101930
+    },
+    {
+      "epoch": 0.5090763814327449,
+      "grad_norm": 0.0944557636976242,
+      "learning_rate": 1.4764574832911964e-05,
+      "loss": 8.9048,
+      "step": 101940
+    },
+    {
+      "epoch": 0.5091263202576843,
+      "grad_norm": 0.08811389654874802,
+      "learning_rate": 1.4763072917970413e-05,
+      "loss": 8.9097,
+      "step": 101950
+    },
+    {
+      "epoch": 0.5091762590826238,
+      "grad_norm": 0.0878419429063797,
+      "learning_rate": 1.4761571003028861e-05,
+      "loss": 8.8932,
+      "step": 101960
+    },
+    {
+      "epoch": 0.5092261979075632,
+      "grad_norm": 0.09334632754325867,
+      "learning_rate": 1.4760069088087311e-05,
+      "loss": 8.8877,
+      "step": 101970
+    },
+    {
+      "epoch": 0.5092761367325027,
+      "grad_norm": 0.09611135721206665,
+      "learning_rate": 1.4758567173145762e-05,
+      "loss": 8.8956,
+      "step": 101980
+    },
+    {
+      "epoch": 0.5093260755574421,
+      "grad_norm": 0.09940996766090393,
+      "learning_rate": 1.4757065258204212e-05,
+      "loss": 8.8933,
+      "step": 101990
+    },
+    {
+      "epoch": 0.5093760143823816,
+      "grad_norm": 0.08453790843486786,
+      "learning_rate": 1.475556334326266e-05,
+      "loss": 8.8861,
+      "step": 102000
+    },
+    {
+      "epoch": 0.509425953207321,
+      "grad_norm": 0.09275045245885849,
+      "learning_rate": 1.475406142832111e-05,
+      "loss": 8.8927,
+      "step": 102010
+    },
+    {
+      "epoch": 0.5094758920322605,
+      "grad_norm": 0.09172289073467255,
+      "learning_rate": 1.4752559513379559e-05,
+      "loss": 8.8762,
+      "step": 102020
+    },
+    {
+      "epoch": 0.5095258308571999,
+      "grad_norm": 0.09009222686290741,
+      "learning_rate": 1.4751057598438009e-05,
+      "loss": 8.9011,
+      "step": 102030
+    },
+    {
+      "epoch": 0.5095757696821394,
+      "grad_norm": 0.09388335049152374,
+      "learning_rate": 1.474955568349646e-05,
+      "loss": 8.8875,
+      "step": 102040
+    },
+    {
+      "epoch": 0.5096257085070788,
+      "grad_norm": 0.09411590546369553,
+      "learning_rate": 1.4748053768554908e-05,
+      "loss": 8.8942,
+      "step": 102050
+    },
+    {
+      "epoch": 0.5096756473320183,
+      "grad_norm": 0.08968400955200195,
+      "learning_rate": 1.4746551853613358e-05,
+      "loss": 8.8953,
+      "step": 102060
+    },
+    {
+      "epoch": 0.5097255861569577,
+      "grad_norm": 0.09362632036209106,
+      "learning_rate": 1.4745049938671806e-05,
+      "loss": 8.9011,
+      "step": 102070
+    },
+    {
+      "epoch": 0.5097755249818972,
+      "grad_norm": 0.093928761780262,
+      "learning_rate": 1.4743548023730257e-05,
+      "loss": 8.8858,
+      "step": 102080
+    },
+    {
+      "epoch": 0.5098254638068366,
+      "grad_norm": 0.09098640829324722,
+      "learning_rate": 1.4742046108788707e-05,
+      "loss": 8.8899,
+      "step": 102090
+    },
+    {
+      "epoch": 0.5098754026317761,
+      "grad_norm": 0.09484419226646423,
+      "learning_rate": 1.4740544193847155e-05,
+      "loss": 8.8725,
+      "step": 102100
+    },
+    {
+      "epoch": 0.5099253414567155,
+      "grad_norm": 0.0966237410902977,
+      "learning_rate": 1.4739042278905605e-05,
+      "loss": 8.8903,
+      "step": 102110
+    },
+    {
+      "epoch": 0.509975280281655,
+      "grad_norm": 0.09729520231485367,
+      "learning_rate": 1.4737540363964054e-05,
+      "loss": 8.8814,
+      "step": 102120
+    },
+    {
+      "epoch": 0.5100252191065944,
+      "grad_norm": 0.09406805038452148,
+      "learning_rate": 1.4736038449022504e-05,
+      "loss": 8.8763,
+      "step": 102130
+    },
+    {
+      "epoch": 0.5100751579315339,
+      "grad_norm": 0.0967145785689354,
+      "learning_rate": 1.4734536534080954e-05,
+      "loss": 8.8686,
+      "step": 102140
+    },
+    {
+      "epoch": 0.5101250967564733,
+      "grad_norm": 0.10264673829078674,
+      "learning_rate": 1.4733034619139403e-05,
+      "loss": 8.8916,
+      "step": 102150
+    },
+    {
+      "epoch": 0.5101750355814128,
+      "grad_norm": 0.09559258073568344,
+      "learning_rate": 1.4731532704197853e-05,
+      "loss": 8.889,
+      "step": 102160
+    },
+    {
+      "epoch": 0.5102249744063522,
+      "grad_norm": 0.09675032645463943,
+      "learning_rate": 1.4730030789256303e-05,
+      "loss": 8.8761,
+      "step": 102170
+    },
+    {
+      "epoch": 0.5102749132312917,
+      "grad_norm": 0.09106460958719254,
+      "learning_rate": 1.4728528874314752e-05,
+      "loss": 8.89,
+      "step": 102180
+    },
+    {
+      "epoch": 0.5103248520562311,
+      "grad_norm": 0.09537220746278763,
+      "learning_rate": 1.4727026959373202e-05,
+      "loss": 8.892,
+      "step": 102190
+    },
+    {
+      "epoch": 0.5103747908811705,
+      "grad_norm": 0.08975797146558762,
+      "learning_rate": 1.472552504443165e-05,
+      "loss": 8.8853,
+      "step": 102200
+    },
+    {
+      "epoch": 0.51042472970611,
+      "grad_norm": 0.09914405643939972,
+      "learning_rate": 1.47240231294901e-05,
+      "loss": 8.8851,
+      "step": 102210
+    },
+    {
+      "epoch": 0.5104746685310495,
+      "grad_norm": 0.09528080374002457,
+      "learning_rate": 1.472252121454855e-05,
+      "loss": 8.9045,
+      "step": 102220
+    },
+    {
+      "epoch": 0.5105246073559889,
+      "grad_norm": 0.09561271965503693,
+      "learning_rate": 1.4721019299606999e-05,
+      "loss": 8.8875,
+      "step": 102230
+    },
+    {
+      "epoch": 0.5105745461809283,
+      "grad_norm": 0.09324440360069275,
+      "learning_rate": 1.471951738466545e-05,
+      "loss": 8.882,
+      "step": 102240
+    },
+    {
+      "epoch": 0.5106244850058678,
+      "grad_norm": 0.10084079951047897,
+      "learning_rate": 1.4718015469723898e-05,
+      "loss": 8.8891,
+      "step": 102250
+    },
+    {
+      "epoch": 0.5106744238308073,
+      "grad_norm": 0.09249073266983032,
+      "learning_rate": 1.4716513554782348e-05,
+      "loss": 8.8773,
+      "step": 102260
+    },
+    {
+      "epoch": 0.5107243626557467,
+      "grad_norm": 0.09199045598506927,
+      "learning_rate": 1.4715011639840798e-05,
+      "loss": 8.8801,
+      "step": 102270
+    },
+    {
+      "epoch": 0.5107743014806861,
+      "grad_norm": 0.09335954487323761,
+      "learning_rate": 1.4713509724899247e-05,
+      "loss": 8.8854,
+      "step": 102280
+    },
+    {
+      "epoch": 0.5108242403056256,
+      "grad_norm": 0.09267157316207886,
+      "learning_rate": 1.4712007809957697e-05,
+      "loss": 8.8834,
+      "step": 102290
+    },
+    {
+      "epoch": 0.5108741791305651,
+      "grad_norm": 0.0931268110871315,
+      "learning_rate": 1.4710505895016145e-05,
+      "loss": 8.8788,
+      "step": 102300
+    },
+    {
+      "epoch": 0.5109241179555045,
+      "grad_norm": 0.09288343042135239,
+      "learning_rate": 1.4709003980074595e-05,
+      "loss": 8.8813,
+      "step": 102310
+    },
+    {
+      "epoch": 0.5109740567804439,
+      "grad_norm": 0.0869889035820961,
+      "learning_rate": 1.4707502065133046e-05,
+      "loss": 8.8893,
+      "step": 102320
+    },
+    {
+      "epoch": 0.5110239956053834,
+      "grad_norm": 0.0896022841334343,
+      "learning_rate": 1.4706000150191496e-05,
+      "loss": 8.8793,
+      "step": 102330
+    },
+    {
+      "epoch": 0.5110739344303229,
+      "grad_norm": 0.09461673349142075,
+      "learning_rate": 1.4704498235249944e-05,
+      "loss": 8.8852,
+      "step": 102340
+    },
+    {
+      "epoch": 0.5111238732552623,
+      "grad_norm": 0.09535450488328934,
+      "learning_rate": 1.4702996320308393e-05,
+      "loss": 8.8921,
+      "step": 102350
+    },
+    {
+      "epoch": 0.5111738120802017,
+      "grad_norm": 0.09272834658622742,
+      "learning_rate": 1.4701494405366843e-05,
+      "loss": 8.9021,
+      "step": 102360
+    },
+    {
+      "epoch": 0.5112237509051412,
+      "grad_norm": 0.09534373134374619,
+      "learning_rate": 1.4699992490425293e-05,
+      "loss": 8.8848,
+      "step": 102370
+    },
+    {
+      "epoch": 0.5112736897300807,
+      "grad_norm": 0.09193330258131027,
+      "learning_rate": 1.4698490575483743e-05,
+      "loss": 8.8843,
+      "step": 102380
+    },
+    {
+      "epoch": 0.5113236285550201,
+      "grad_norm": 0.09175775945186615,
+      "learning_rate": 1.4696988660542192e-05,
+      "loss": 8.895,
+      "step": 102390
+    },
+    {
+      "epoch": 0.5113735673799595,
+      "grad_norm": 0.09013816714286804,
+      "learning_rate": 1.469548674560064e-05,
+      "loss": 8.8983,
+      "step": 102400
+    },
+    {
+      "epoch": 0.511423506204899,
+      "grad_norm": 0.09730950742959976,
+      "learning_rate": 1.469398483065909e-05,
+      "loss": 8.8866,
+      "step": 102410
+    },
+    {
+      "epoch": 0.5114734450298385,
+      "grad_norm": 0.09352081269025803,
+      "learning_rate": 1.469248291571754e-05,
+      "loss": 8.8842,
+      "step": 102420
+    },
+    {
+      "epoch": 0.5115233838547779,
+      "grad_norm": 0.09047748893499374,
+      "learning_rate": 1.469098100077599e-05,
+      "loss": 8.8971,
+      "step": 102430
+    },
+    {
+      "epoch": 0.5115733226797173,
+      "grad_norm": 0.09623833000659943,
+      "learning_rate": 1.468947908583444e-05,
+      "loss": 8.8938,
+      "step": 102440
+    },
+    {
+      "epoch": 0.5116232615046568,
+      "grad_norm": 0.09208162873983383,
+      "learning_rate": 1.4687977170892888e-05,
+      "loss": 8.902,
+      "step": 102450
+    },
+    {
+      "epoch": 0.5116732003295963,
+      "grad_norm": 0.0962912067770958,
+      "learning_rate": 1.4686475255951338e-05,
+      "loss": 8.8727,
+      "step": 102460
+    },
+    {
+      "epoch": 0.5117231391545357,
+      "grad_norm": 0.0914090946316719,
+      "learning_rate": 1.4684973341009788e-05,
+      "loss": 8.8823,
+      "step": 102470
+    },
+    {
+      "epoch": 0.5117730779794751,
+      "grad_norm": 0.09146260470151901,
+      "learning_rate": 1.4683471426068238e-05,
+      "loss": 8.9058,
+      "step": 102480
+    },
+    {
+      "epoch": 0.5118230168044146,
+      "grad_norm": 0.09242929518222809,
+      "learning_rate": 1.4681969511126688e-05,
+      "loss": 8.8762,
+      "step": 102490
+    },
+    {
+      "epoch": 0.5118729556293541,
+      "grad_norm": 0.09700184315443039,
+      "learning_rate": 1.4680467596185135e-05,
+      "loss": 8.8787,
+      "step": 102500
+    },
+    {
+      "epoch": 0.5119228944542935,
+      "grad_norm": 0.08677946031093597,
+      "learning_rate": 1.4678965681243585e-05,
+      "loss": 8.8883,
+      "step": 102510
+    },
+    {
+      "epoch": 0.5119728332792329,
+      "grad_norm": 0.09197265654802322,
+      "learning_rate": 1.4677463766302036e-05,
+      "loss": 8.8903,
+      "step": 102520
+    },
+    {
+      "epoch": 0.5120227721041724,
+      "grad_norm": 0.08945990353822708,
+      "learning_rate": 1.4675961851360486e-05,
+      "loss": 8.8724,
+      "step": 102530
+    },
+    {
+      "epoch": 0.5120727109291119,
+      "grad_norm": 0.08991879224777222,
+      "learning_rate": 1.4674459936418936e-05,
+      "loss": 8.8745,
+      "step": 102540
+    },
+    {
+      "epoch": 0.5121226497540513,
+      "grad_norm": 0.09267672151327133,
+      "learning_rate": 1.4672958021477383e-05,
+      "loss": 8.8855,
+      "step": 102550
+    },
+    {
+      "epoch": 0.5121725885789907,
+      "grad_norm": 0.09635213762521744,
+      "learning_rate": 1.4671456106535833e-05,
+      "loss": 8.8905,
+      "step": 102560
+    },
+    {
+      "epoch": 0.5122225274039302,
+      "grad_norm": 0.0958060547709465,
+      "learning_rate": 1.4669954191594283e-05,
+      "loss": 8.9029,
+      "step": 102570
+    },
+    {
+      "epoch": 0.5122724662288697,
+      "grad_norm": 0.09407208114862442,
+      "learning_rate": 1.4668452276652733e-05,
+      "loss": 8.8742,
+      "step": 102580
+    },
+    {
+      "epoch": 0.5123224050538091,
+      "grad_norm": 0.09680324047803879,
+      "learning_rate": 1.4666950361711183e-05,
+      "loss": 8.8967,
+      "step": 102590
+    },
+    {
+      "epoch": 0.5123723438787485,
+      "grad_norm": 0.09674806892871857,
+      "learning_rate": 1.466544844676963e-05,
+      "loss": 8.8878,
+      "step": 102600
+    },
+    {
+      "epoch": 0.512422282703688,
+      "grad_norm": 0.08954492211341858,
+      "learning_rate": 1.466394653182808e-05,
+      "loss": 8.8912,
+      "step": 102610
+    },
+    {
+      "epoch": 0.5124722215286275,
+      "grad_norm": 0.08924522995948792,
+      "learning_rate": 1.466244461688653e-05,
+      "loss": 8.8811,
+      "step": 102620
+    },
+    {
+      "epoch": 0.5125221603535669,
+      "grad_norm": 0.09569413959980011,
+      "learning_rate": 1.466094270194498e-05,
+      "loss": 8.8828,
+      "step": 102630
+    },
+    {
+      "epoch": 0.5125720991785063,
+      "grad_norm": 0.09132692217826843,
+      "learning_rate": 1.4659440787003431e-05,
+      "loss": 8.8799,
+      "step": 102640
+    },
+    {
+      "epoch": 0.5126220380034457,
+      "grad_norm": 0.08919345587491989,
+      "learning_rate": 1.4657938872061878e-05,
+      "loss": 8.8885,
+      "step": 102650
+    },
+    {
+      "epoch": 0.5126719768283853,
+      "grad_norm": 0.09804876893758774,
+      "learning_rate": 1.4656436957120328e-05,
+      "loss": 8.8798,
+      "step": 102660
+    },
+    {
+      "epoch": 0.5127219156533247,
+      "grad_norm": 0.10200927406549454,
+      "learning_rate": 1.4654935042178778e-05,
+      "loss": 8.8672,
+      "step": 102670
+    },
+    {
+      "epoch": 0.5127718544782641,
+      "grad_norm": 0.09064894169569016,
+      "learning_rate": 1.4653433127237228e-05,
+      "loss": 8.8719,
+      "step": 102680
+    },
+    {
+      "epoch": 0.5128217933032035,
+      "grad_norm": 0.09494546800851822,
+      "learning_rate": 1.4651931212295678e-05,
+      "loss": 8.8996,
+      "step": 102690
+    },
+    {
+      "epoch": 0.5128717321281431,
+      "grad_norm": 0.09214048832654953,
+      "learning_rate": 1.4650429297354127e-05,
+      "loss": 8.8922,
+      "step": 102700
+    },
+    {
+      "epoch": 0.5129216709530825,
+      "grad_norm": 0.09070252627134323,
+      "learning_rate": 1.4648927382412575e-05,
+      "loss": 8.8887,
+      "step": 102710
+    },
+    {
+      "epoch": 0.5129716097780219,
+      "grad_norm": 0.09582309424877167,
+      "learning_rate": 1.4647425467471026e-05,
+      "loss": 8.8841,
+      "step": 102720
+    },
+    {
+      "epoch": 0.5130215486029613,
+      "grad_norm": 0.0906309261918068,
+      "learning_rate": 1.4645923552529476e-05,
+      "loss": 8.8794,
+      "step": 102730
+    },
+    {
+      "epoch": 0.5130714874279009,
+      "grad_norm": 0.0934000313282013,
+      "learning_rate": 1.4644421637587926e-05,
+      "loss": 8.8729,
+      "step": 102740
+    },
+    {
+      "epoch": 0.5131214262528403,
+      "grad_norm": 0.09085663408041,
+      "learning_rate": 1.4642919722646374e-05,
+      "loss": 8.8923,
+      "step": 102750
+    },
+    {
+      "epoch": 0.5131713650777797,
+      "grad_norm": 0.0955960750579834,
+      "learning_rate": 1.4641417807704823e-05,
+      "loss": 8.8932,
+      "step": 102760
+    },
+    {
+      "epoch": 0.5132213039027191,
+      "grad_norm": 0.09510260075330734,
+      "learning_rate": 1.4639915892763273e-05,
+      "loss": 8.8844,
+      "step": 102770
+    },
+    {
+      "epoch": 0.5132712427276587,
+      "grad_norm": 0.0931367352604866,
+      "learning_rate": 1.4638413977821723e-05,
+      "loss": 8.8895,
+      "step": 102780
+    },
+    {
+      "epoch": 0.5133211815525981,
+      "grad_norm": 0.09127708524465561,
+      "learning_rate": 1.4636912062880173e-05,
+      "loss": 8.8706,
+      "step": 102790
+    },
+    {
+      "epoch": 0.5133711203775375,
+      "grad_norm": 0.09185762703418732,
+      "learning_rate": 1.4635410147938622e-05,
+      "loss": 8.8691,
+      "step": 102800
+    },
+    {
+      "epoch": 0.5134210592024769,
+      "grad_norm": 0.08858191221952438,
+      "learning_rate": 1.463390823299707e-05,
+      "loss": 8.8763,
+      "step": 102810
+    },
+    {
+      "epoch": 0.5134709980274165,
+      "grad_norm": 0.08831470459699631,
+      "learning_rate": 1.463240631805552e-05,
+      "loss": 8.8851,
+      "step": 102820
+    },
+    {
+      "epoch": 0.5135209368523559,
+      "grad_norm": 0.08874993771314621,
+      "learning_rate": 1.463090440311397e-05,
+      "loss": 8.8826,
+      "step": 102830
+    },
+    {
+      "epoch": 0.5135708756772953,
+      "grad_norm": 0.09777554869651794,
+      "learning_rate": 1.4629402488172421e-05,
+      "loss": 8.888,
+      "step": 102840
+    },
+    {
+      "epoch": 0.5136208145022347,
+      "grad_norm": 0.08795005083084106,
+      "learning_rate": 1.462790057323087e-05,
+      "loss": 8.8808,
+      "step": 102850
+    },
+    {
+      "epoch": 0.5136707533271742,
+      "grad_norm": 0.08897390216588974,
+      "learning_rate": 1.462639865828932e-05,
+      "loss": 8.8769,
+      "step": 102860
+    },
+    {
+      "epoch": 0.5137206921521137,
+      "grad_norm": 0.09718848019838333,
+      "learning_rate": 1.4624896743347768e-05,
+      "loss": 8.8757,
+      "step": 102870
+    },
+    {
+      "epoch": 0.5137706309770531,
+      "grad_norm": 0.09486740082502365,
+      "learning_rate": 1.4623394828406218e-05,
+      "loss": 8.8906,
+      "step": 102880
+    },
+    {
+      "epoch": 0.5138205698019925,
+      "grad_norm": 0.09383400529623032,
+      "learning_rate": 1.4621892913464668e-05,
+      "loss": 8.8932,
+      "step": 102890
+    },
+    {
+      "epoch": 0.513870508626932,
+      "grad_norm": 0.09229724854230881,
+      "learning_rate": 1.4620390998523117e-05,
+      "loss": 8.8677,
+      "step": 102900
+    },
+    {
+      "epoch": 0.5139204474518715,
+      "grad_norm": 0.09365616738796234,
+      "learning_rate": 1.4618889083581567e-05,
+      "loss": 8.8787,
+      "step": 102910
+    },
+    {
+      "epoch": 0.5139703862768109,
+      "grad_norm": 0.09437093883752823,
+      "learning_rate": 1.4617387168640016e-05,
+      "loss": 8.9,
+      "step": 102920
+    },
+    {
+      "epoch": 0.5140203251017503,
+      "grad_norm": 0.09300998598337173,
+      "learning_rate": 1.4615885253698466e-05,
+      "loss": 8.88,
+      "step": 102930
+    },
+    {
+      "epoch": 0.5140702639266898,
+      "grad_norm": 0.09167800098657608,
+      "learning_rate": 1.4614383338756916e-05,
+      "loss": 8.8927,
+      "step": 102940
+    },
+    {
+      "epoch": 0.5141202027516293,
+      "grad_norm": 0.09113366156816483,
+      "learning_rate": 1.4612881423815364e-05,
+      "loss": 8.8802,
+      "step": 102950
+    },
+    {
+      "epoch": 0.5141701415765687,
+      "grad_norm": 0.09514357149600983,
+      "learning_rate": 1.4611379508873815e-05,
+      "loss": 8.8725,
+      "step": 102960
+    },
+    {
+      "epoch": 0.5142200804015081,
+      "grad_norm": 0.09392548352479935,
+      "learning_rate": 1.4609877593932263e-05,
+      "loss": 8.869,
+      "step": 102970
+    },
+    {
+      "epoch": 0.5142700192264476,
+      "grad_norm": 0.08608120679855347,
+      "learning_rate": 1.4608375678990713e-05,
+      "loss": 8.8806,
+      "step": 102980
+    },
+    {
+      "epoch": 0.5143199580513871,
+      "grad_norm": 0.09436332434415817,
+      "learning_rate": 1.4606873764049163e-05,
+      "loss": 8.8717,
+      "step": 102990
+    },
+    {
+      "epoch": 0.5143698968763265,
+      "grad_norm": 0.09002230316400528,
+      "learning_rate": 1.4605371849107612e-05,
+      "loss": 8.8916,
+      "step": 103000
+    },
+    {
+      "epoch": 0.5144198357012659,
+      "grad_norm": 0.09780488163232803,
+      "learning_rate": 1.4603869934166062e-05,
+      "loss": 8.8725,
+      "step": 103010
+    },
+    {
+      "epoch": 0.5144697745262053,
+      "grad_norm": 0.09558107703924179,
+      "learning_rate": 1.4602368019224512e-05,
+      "loss": 8.8836,
+      "step": 103020
+    },
+    {
+      "epoch": 0.5145197133511449,
+      "grad_norm": 0.09473099559545517,
+      "learning_rate": 1.460086610428296e-05,
+      "loss": 8.8739,
+      "step": 103030
+    },
+    {
+      "epoch": 0.5145696521760843,
+      "grad_norm": 0.09364248812198639,
+      "learning_rate": 1.4599364189341411e-05,
+      "loss": 8.8709,
+      "step": 103040
+    },
+    {
+      "epoch": 0.5146195910010237,
+      "grad_norm": 0.09198643267154694,
+      "learning_rate": 1.459786227439986e-05,
+      "loss": 8.8838,
+      "step": 103050
+    },
+    {
+      "epoch": 0.5146695298259631,
+      "grad_norm": 0.0934210941195488,
+      "learning_rate": 1.459636035945831e-05,
+      "loss": 8.8783,
+      "step": 103060
+    },
+    {
+      "epoch": 0.5147194686509027,
+      "grad_norm": 0.09671808034181595,
+      "learning_rate": 1.459485844451676e-05,
+      "loss": 8.8814,
+      "step": 103070
+    },
+    {
+      "epoch": 0.5147694074758421,
+      "grad_norm": 0.0925278291106224,
+      "learning_rate": 1.4593356529575208e-05,
+      "loss": 8.8924,
+      "step": 103080
+    },
+    {
+      "epoch": 0.5148193463007815,
+      "grad_norm": 0.09130702912807465,
+      "learning_rate": 1.4591854614633658e-05,
+      "loss": 8.8789,
+      "step": 103090
+    },
+    {
+      "epoch": 0.5148692851257209,
+      "grad_norm": 0.0948934555053711,
+      "learning_rate": 1.4590352699692107e-05,
+      "loss": 8.8758,
+      "step": 103100
+    },
+    {
+      "epoch": 0.5149192239506605,
+      "grad_norm": 0.09707729518413544,
+      "learning_rate": 1.4588850784750557e-05,
+      "loss": 8.8773,
+      "step": 103110
+    },
+    {
+      "epoch": 0.5149691627755999,
+      "grad_norm": 0.09346991032361984,
+      "learning_rate": 1.4587348869809007e-05,
+      "loss": 8.8743,
+      "step": 103120
+    },
+    {
+      "epoch": 0.5150191016005393,
+      "grad_norm": 0.09797349572181702,
+      "learning_rate": 1.4585846954867456e-05,
+      "loss": 8.8825,
+      "step": 103130
+    },
+    {
+      "epoch": 0.5150690404254787,
+      "grad_norm": 0.08969218283891678,
+      "learning_rate": 1.4584345039925906e-05,
+      "loss": 8.889,
+      "step": 103140
+    },
+    {
+      "epoch": 0.5151189792504183,
+      "grad_norm": 0.0924285352230072,
+      "learning_rate": 1.4582843124984354e-05,
+      "loss": 8.876,
+      "step": 103150
+    },
+    {
+      "epoch": 0.5151689180753577,
+      "grad_norm": 0.0896601751446724,
+      "learning_rate": 1.4581341210042805e-05,
+      "loss": 8.8611,
+      "step": 103160
+    },
+    {
+      "epoch": 0.5152188569002971,
+      "grad_norm": 0.08862532675266266,
+      "learning_rate": 1.4579839295101255e-05,
+      "loss": 8.8764,
+      "step": 103170
+    },
+    {
+      "epoch": 0.5152687957252365,
+      "grad_norm": 0.09978237003087997,
+      "learning_rate": 1.4578337380159705e-05,
+      "loss": 8.883,
+      "step": 103180
+    },
+    {
+      "epoch": 0.5153187345501761,
+      "grad_norm": 0.09299599379301071,
+      "learning_rate": 1.4576835465218153e-05,
+      "loss": 8.8806,
+      "step": 103190
+    },
+    {
+      "epoch": 0.5153686733751155,
+      "grad_norm": 0.09699616581201553,
+      "learning_rate": 1.4575333550276602e-05,
+      "loss": 8.8871,
+      "step": 103200
+    },
+    {
+      "epoch": 0.5154186122000549,
+      "grad_norm": 0.09479953348636627,
+      "learning_rate": 1.4573831635335052e-05,
+      "loss": 8.8893,
+      "step": 103210
+    },
+    {
+      "epoch": 0.5154685510249943,
+      "grad_norm": 0.09587261080741882,
+      "learning_rate": 1.4572329720393502e-05,
+      "loss": 8.8802,
+      "step": 103220
+    },
+    {
+      "epoch": 0.5155184898499339,
+      "grad_norm": 0.09195046871900558,
+      "learning_rate": 1.4570827805451952e-05,
+      "loss": 8.8887,
+      "step": 103230
+    },
+    {
+      "epoch": 0.5155684286748733,
+      "grad_norm": 0.09111469984054565,
+      "learning_rate": 1.4569325890510401e-05,
+      "loss": 8.868,
+      "step": 103240
+    },
+    {
+      "epoch": 0.5156183674998127,
+      "grad_norm": 0.09878696501255035,
+      "learning_rate": 1.456782397556885e-05,
+      "loss": 8.8859,
+      "step": 103250
+    },
+    {
+      "epoch": 0.5156683063247521,
+      "grad_norm": 0.09132800251245499,
+      "learning_rate": 1.45663220606273e-05,
+      "loss": 8.8826,
+      "step": 103260
+    },
+    {
+      "epoch": 0.5157182451496917,
+      "grad_norm": 0.0929095596075058,
+      "learning_rate": 1.456482014568575e-05,
+      "loss": 8.8912,
+      "step": 103270
+    },
+    {
+      "epoch": 0.5157681839746311,
+      "grad_norm": 0.0883350670337677,
+      "learning_rate": 1.45633182307442e-05,
+      "loss": 8.8899,
+      "step": 103280
+    },
+    {
+      "epoch": 0.5158181227995705,
+      "grad_norm": 0.0903296247124672,
+      "learning_rate": 1.4561816315802648e-05,
+      "loss": 8.8713,
+      "step": 103290
+    },
+    {
+      "epoch": 0.5158680616245099,
+      "grad_norm": 0.09251569956541061,
+      "learning_rate": 1.4560314400861097e-05,
+      "loss": 8.8755,
+      "step": 103300
+    },
+    {
+      "epoch": 0.5159180004494495,
+      "grad_norm": 0.10021259635686874,
+      "learning_rate": 1.4558812485919547e-05,
+      "loss": 8.8717,
+      "step": 103310
+    },
+    {
+      "epoch": 0.5159679392743889,
+      "grad_norm": 0.09307809174060822,
+      "learning_rate": 1.4557310570977997e-05,
+      "loss": 8.861,
+      "step": 103320
+    },
+    {
+      "epoch": 0.5160178780993283,
+      "grad_norm": 0.09546101093292236,
+      "learning_rate": 1.4555808656036448e-05,
+      "loss": 8.8763,
+      "step": 103330
+    },
+    {
+      "epoch": 0.5160678169242677,
+      "grad_norm": 0.09137708693742752,
+      "learning_rate": 1.4554306741094898e-05,
+      "loss": 8.8986,
+      "step": 103340
+    },
+    {
+      "epoch": 0.5161177557492073,
+      "grad_norm": 0.09687686711549759,
+      "learning_rate": 1.4552804826153344e-05,
+      "loss": 8.8711,
+      "step": 103350
+    },
+    {
+      "epoch": 0.5161676945741467,
+      "grad_norm": 0.09281541407108307,
+      "learning_rate": 1.4551302911211795e-05,
+      "loss": 8.8645,
+      "step": 103360
+    },
+    {
+      "epoch": 0.5162176333990861,
+      "grad_norm": 0.08926443755626678,
+      "learning_rate": 1.4549800996270245e-05,
+      "loss": 8.8634,
+      "step": 103370
+    },
+    {
+      "epoch": 0.5162675722240255,
+      "grad_norm": 0.09436703473329544,
+      "learning_rate": 1.4548299081328695e-05,
+      "loss": 8.878,
+      "step": 103380
+    },
+    {
+      "epoch": 0.5163175110489651,
+      "grad_norm": 0.09582871198654175,
+      "learning_rate": 1.4546797166387145e-05,
+      "loss": 8.8757,
+      "step": 103390
+    },
+    {
+      "epoch": 0.5163674498739045,
+      "grad_norm": 0.09050548076629639,
+      "learning_rate": 1.4545295251445592e-05,
+      "loss": 8.8923,
+      "step": 103400
+    },
+    {
+      "epoch": 0.5164173886988439,
+      "grad_norm": 0.0901259183883667,
+      "learning_rate": 1.4543793336504042e-05,
+      "loss": 8.8696,
+      "step": 103410
+    },
+    {
+      "epoch": 0.5164673275237833,
+      "grad_norm": 0.09428418427705765,
+      "learning_rate": 1.4542291421562492e-05,
+      "loss": 8.882,
+      "step": 103420
+    },
+    {
+      "epoch": 0.5165172663487229,
+      "grad_norm": 0.09372241050004959,
+      "learning_rate": 1.4540789506620943e-05,
+      "loss": 8.874,
+      "step": 103430
+    },
+    {
+      "epoch": 0.5165672051736623,
+      "grad_norm": 0.09623238444328308,
+      "learning_rate": 1.4539287591679393e-05,
+      "loss": 8.8679,
+      "step": 103440
+    },
+    {
+      "epoch": 0.5166171439986017,
+      "grad_norm": 0.09057226777076721,
+      "learning_rate": 1.453778567673784e-05,
+      "loss": 8.8789,
+      "step": 103450
+    },
+    {
+      "epoch": 0.5166670828235411,
+      "grad_norm": 0.09329625219106674,
+      "learning_rate": 1.453628376179629e-05,
+      "loss": 8.874,
+      "step": 103460
+    },
+    {
+      "epoch": 0.5167170216484807,
+      "grad_norm": 0.0941077321767807,
+      "learning_rate": 1.453478184685474e-05,
+      "loss": 8.8716,
+      "step": 103470
+    },
+    {
+      "epoch": 0.5167669604734201,
+      "grad_norm": 0.09424999356269836,
+      "learning_rate": 1.453327993191319e-05,
+      "loss": 8.8834,
+      "step": 103480
+    },
+    {
+      "epoch": 0.5168168992983595,
+      "grad_norm": 0.0937827080488205,
+      "learning_rate": 1.453177801697164e-05,
+      "loss": 8.8731,
+      "step": 103490
+    },
+    {
+      "epoch": 0.5168668381232989,
+      "grad_norm": 0.0896645337343216,
+      "learning_rate": 1.4530276102030089e-05,
+      "loss": 8.8669,
+      "step": 103500
+    },
+    {
+      "epoch": 0.5169167769482385,
+      "grad_norm": 0.08836094290018082,
+      "learning_rate": 1.4528774187088537e-05,
+      "loss": 8.8832,
+      "step": 103510
+    },
+    {
+      "epoch": 0.5169667157731779,
+      "grad_norm": 0.0928412526845932,
+      "learning_rate": 1.4527272272146987e-05,
+      "loss": 8.8857,
+      "step": 103520
+    },
+    {
+      "epoch": 0.5170166545981173,
+      "grad_norm": 0.09539926797151566,
+      "learning_rate": 1.4525770357205438e-05,
+      "loss": 8.892,
+      "step": 103530
+    },
+    {
+      "epoch": 0.5170665934230567,
+      "grad_norm": 0.09582491964101791,
+      "learning_rate": 1.4524268442263888e-05,
+      "loss": 8.8688,
+      "step": 103540
+    },
+    {
+      "epoch": 0.5171165322479963,
+      "grad_norm": 0.09870574623346329,
+      "learning_rate": 1.4522766527322336e-05,
+      "loss": 8.8783,
+      "step": 103550
+    },
+    {
+      "epoch": 0.5171664710729357,
+      "grad_norm": 0.09343771636486053,
+      "learning_rate": 1.4521264612380785e-05,
+      "loss": 8.8707,
+      "step": 103560
+    },
+    {
+      "epoch": 0.5172164098978751,
+      "grad_norm": 0.09553004056215286,
+      "learning_rate": 1.4519762697439235e-05,
+      "loss": 8.8827,
+      "step": 103570
+    },
+    {
+      "epoch": 0.5172663487228145,
+      "grad_norm": 0.09276316314935684,
+      "learning_rate": 1.4518260782497685e-05,
+      "loss": 8.8747,
+      "step": 103580
+    },
+    {
+      "epoch": 0.517316287547754,
+      "grad_norm": 0.09070324152708054,
+      "learning_rate": 1.4516758867556135e-05,
+      "loss": 8.876,
+      "step": 103590
+    },
+    {
+      "epoch": 0.5173662263726935,
+      "grad_norm": 0.09403187781572342,
+      "learning_rate": 1.4515256952614584e-05,
+      "loss": 8.8583,
+      "step": 103600
+    },
+    {
+      "epoch": 0.5174161651976329,
+      "grad_norm": 0.08794444054365158,
+      "learning_rate": 1.4513755037673032e-05,
+      "loss": 8.8818,
+      "step": 103610
+    },
+    {
+      "epoch": 0.5174661040225723,
+      "grad_norm": 0.09099708497524261,
+      "learning_rate": 1.4512253122731482e-05,
+      "loss": 8.8783,
+      "step": 103620
+    },
+    {
+      "epoch": 0.5175160428475118,
+      "grad_norm": 0.09580422937870026,
+      "learning_rate": 1.4510751207789933e-05,
+      "loss": 8.8714,
+      "step": 103630
+    },
+    {
+      "epoch": 0.5175659816724513,
+      "grad_norm": 0.09501136839389801,
+      "learning_rate": 1.4509249292848383e-05,
+      "loss": 8.8734,
+      "step": 103640
+    },
+    {
+      "epoch": 0.5176159204973907,
+      "grad_norm": 0.09438387304544449,
+      "learning_rate": 1.4507747377906831e-05,
+      "loss": 8.8821,
+      "step": 103650
+    },
+    {
+      "epoch": 0.5176658593223301,
+      "grad_norm": 0.09567675739526749,
+      "learning_rate": 1.4506245462965281e-05,
+      "loss": 8.8686,
+      "step": 103660
+    },
+    {
+      "epoch": 0.5177157981472696,
+      "grad_norm": 0.09035927057266235,
+      "learning_rate": 1.450474354802373e-05,
+      "loss": 8.8584,
+      "step": 103670
+    },
+    {
+      "epoch": 0.5177657369722091,
+      "grad_norm": 0.09373793751001358,
+      "learning_rate": 1.450324163308218e-05,
+      "loss": 8.8612,
+      "step": 103680
+    },
+    {
+      "epoch": 0.5178156757971485,
+      "grad_norm": 0.0980074554681778,
+      "learning_rate": 1.450173971814063e-05,
+      "loss": 8.8687,
+      "step": 103690
+    },
+    {
+      "epoch": 0.5178656146220879,
+      "grad_norm": 0.09776370227336884,
+      "learning_rate": 1.450023780319908e-05,
+      "loss": 8.8841,
+      "step": 103700
+    },
+    {
+      "epoch": 0.5179155534470274,
+      "grad_norm": 0.0890578031539917,
+      "learning_rate": 1.4498735888257529e-05,
+      "loss": 8.8819,
+      "step": 103710
+    },
+    {
+      "epoch": 0.5179654922719669,
+      "grad_norm": 0.09932509809732437,
+      "learning_rate": 1.4497233973315977e-05,
+      "loss": 8.8684,
+      "step": 103720
+    },
+    {
+      "epoch": 0.5180154310969063,
+      "grad_norm": 0.09279154241085052,
+      "learning_rate": 1.4495732058374428e-05,
+      "loss": 8.8734,
+      "step": 103730
+    },
+    {
+      "epoch": 0.5180653699218457,
+      "grad_norm": 0.09163205325603485,
+      "learning_rate": 1.4494230143432878e-05,
+      "loss": 8.8798,
+      "step": 103740
+    },
+    {
+      "epoch": 0.5181153087467852,
+      "grad_norm": 0.08860352635383606,
+      "learning_rate": 1.4492728228491328e-05,
+      "loss": 8.8876,
+      "step": 103750
+    },
+    {
+      "epoch": 0.5181652475717247,
+      "grad_norm": 0.08968942612409592,
+      "learning_rate": 1.4491226313549776e-05,
+      "loss": 8.8661,
+      "step": 103760
+    },
+    {
+      "epoch": 0.5182151863966641,
+      "grad_norm": 0.09817136079072952,
+      "learning_rate": 1.4489724398608225e-05,
+      "loss": 8.88,
+      "step": 103770
+    },
+    {
+      "epoch": 0.5182651252216035,
+      "grad_norm": 0.09469393640756607,
+      "learning_rate": 1.4488222483666675e-05,
+      "loss": 8.8826,
+      "step": 103780
+    },
+    {
+      "epoch": 0.518315064046543,
+      "grad_norm": 0.08931383490562439,
+      "learning_rate": 1.4486720568725125e-05,
+      "loss": 8.874,
+      "step": 103790
+    },
+    {
+      "epoch": 0.5183650028714825,
+      "grad_norm": 0.09799165278673172,
+      "learning_rate": 1.4485218653783575e-05,
+      "loss": 8.8699,
+      "step": 103800
+    },
+    {
+      "epoch": 0.5184149416964219,
+      "grad_norm": 0.09098178893327713,
+      "learning_rate": 1.4483716738842024e-05,
+      "loss": 8.8739,
+      "step": 103810
+    },
+    {
+      "epoch": 0.5184648805213613,
+      "grad_norm": 0.0937158390879631,
+      "learning_rate": 1.4482214823900474e-05,
+      "loss": 8.8743,
+      "step": 103820
+    },
+    {
+      "epoch": 0.5185148193463008,
+      "grad_norm": 0.09612367302179337,
+      "learning_rate": 1.4480712908958923e-05,
+      "loss": 8.8872,
+      "step": 103830
+    },
+    {
+      "epoch": 0.5185647581712403,
+      "grad_norm": 0.08803288638591766,
+      "learning_rate": 1.4479210994017373e-05,
+      "loss": 8.8751,
+      "step": 103840
+    },
+    {
+      "epoch": 0.5186146969961797,
+      "grad_norm": 0.09393215924501419,
+      "learning_rate": 1.4477709079075823e-05,
+      "loss": 8.8681,
+      "step": 103850
+    },
+    {
+      "epoch": 0.5186646358211191,
+      "grad_norm": 0.09139379113912582,
+      "learning_rate": 1.4476207164134271e-05,
+      "loss": 8.8752,
+      "step": 103860
+    },
+    {
+      "epoch": 0.5187145746460585,
+      "grad_norm": 0.08722906559705734,
+      "learning_rate": 1.4474705249192722e-05,
+      "loss": 8.8791,
+      "step": 103870
+    },
+    {
+      "epoch": 0.5187645134709981,
+      "grad_norm": 0.09366084635257721,
+      "learning_rate": 1.447320333425117e-05,
+      "loss": 8.8777,
+      "step": 103880
+    },
+    {
+      "epoch": 0.5188144522959375,
+      "grad_norm": 0.08771378546953201,
+      "learning_rate": 1.447170141930962e-05,
+      "loss": 8.8769,
+      "step": 103890
+    },
+    {
+      "epoch": 0.5188643911208769,
+      "grad_norm": 0.09123623371124268,
+      "learning_rate": 1.447019950436807e-05,
+      "loss": 8.8813,
+      "step": 103900
+    },
+    {
+      "epoch": 0.5189143299458163,
+      "grad_norm": 0.0887846052646637,
+      "learning_rate": 1.4468697589426519e-05,
+      "loss": 8.8682,
+      "step": 103910
+    },
+    {
+      "epoch": 0.5189642687707559,
+      "grad_norm": 0.091937355697155,
+      "learning_rate": 1.4467195674484969e-05,
+      "loss": 8.8748,
+      "step": 103920
+    },
+    {
+      "epoch": 0.5190142075956953,
+      "grad_norm": 0.09356644749641418,
+      "learning_rate": 1.4465693759543418e-05,
+      "loss": 8.8539,
+      "step": 103930
+    },
+    {
+      "epoch": 0.5190641464206347,
+      "grad_norm": 0.08848299831151962,
+      "learning_rate": 1.4464191844601868e-05,
+      "loss": 8.8697,
+      "step": 103940
+    },
+    {
+      "epoch": 0.5191140852455741,
+      "grad_norm": 0.0895775780081749,
+      "learning_rate": 1.4462689929660318e-05,
+      "loss": 8.8636,
+      "step": 103950
+    },
+    {
+      "epoch": 0.5191640240705137,
+      "grad_norm": 0.09382645785808563,
+      "learning_rate": 1.4461188014718766e-05,
+      "loss": 8.8704,
+      "step": 103960
+    },
+    {
+      "epoch": 0.5192139628954531,
+      "grad_norm": 0.0946570485830307,
+      "learning_rate": 1.4459686099777217e-05,
+      "loss": 8.8538,
+      "step": 103970
+    },
+    {
+      "epoch": 0.5192639017203925,
+      "grad_norm": 0.09343206882476807,
+      "learning_rate": 1.4458184184835667e-05,
+      "loss": 8.8671,
+      "step": 103980
+    },
+    {
+      "epoch": 0.5193138405453319,
+      "grad_norm": 0.09253936260938644,
+      "learning_rate": 1.4456682269894115e-05,
+      "loss": 8.8672,
+      "step": 103990
+    },
+    {
+      "epoch": 0.5193637793702715,
+      "grad_norm": 0.0950196385383606,
+      "learning_rate": 1.4455180354952565e-05,
+      "loss": 8.8815,
+      "step": 104000
+    },
+    {
+      "epoch": 0.5194137181952109,
+      "grad_norm": 0.09575220942497253,
+      "learning_rate": 1.4453678440011014e-05,
+      "loss": 8.8704,
+      "step": 104010
+    },
+    {
+      "epoch": 0.5194636570201503,
+      "grad_norm": 0.09172710031270981,
+      "learning_rate": 1.4452176525069464e-05,
+      "loss": 8.8797,
+      "step": 104020
+    },
+    {
+      "epoch": 0.5195135958450897,
+      "grad_norm": 0.09384613484144211,
+      "learning_rate": 1.4450674610127914e-05,
+      "loss": 8.8628,
+      "step": 104030
+    },
+    {
+      "epoch": 0.5195635346700292,
+      "grad_norm": 0.09063451737165451,
+      "learning_rate": 1.4449172695186363e-05,
+      "loss": 8.8736,
+      "step": 104040
+    },
+    {
+      "epoch": 0.5196134734949687,
+      "grad_norm": 0.09290114045143127,
+      "learning_rate": 1.4447670780244813e-05,
+      "loss": 8.8818,
+      "step": 104050
+    },
+    {
+      "epoch": 0.5196634123199081,
+      "grad_norm": 0.09309238195419312,
+      "learning_rate": 1.4446168865303261e-05,
+      "loss": 8.8818,
+      "step": 104060
+    },
+    {
+      "epoch": 0.5197133511448475,
+      "grad_norm": 0.09484042227268219,
+      "learning_rate": 1.4444666950361712e-05,
+      "loss": 8.882,
+      "step": 104070
+    },
+    {
+      "epoch": 0.519763289969787,
+      "grad_norm": 0.09208900481462479,
+      "learning_rate": 1.4443165035420162e-05,
+      "loss": 8.8948,
+      "step": 104080
+    },
+    {
+      "epoch": 0.5198132287947265,
+      "grad_norm": 0.09136916697025299,
+      "learning_rate": 1.444166312047861e-05,
+      "loss": 8.8841,
+      "step": 104090
+    },
+    {
+      "epoch": 0.5198631676196659,
+      "grad_norm": 0.09338723123073578,
+      "learning_rate": 1.444016120553706e-05,
+      "loss": 8.8688,
+      "step": 104100
+    },
+    {
+      "epoch": 0.5199131064446053,
+      "grad_norm": 0.09019609540700912,
+      "learning_rate": 1.4438659290595509e-05,
+      "loss": 8.8557,
+      "step": 104110
+    },
+    {
+      "epoch": 0.5199630452695448,
+      "grad_norm": 0.09477616846561432,
+      "learning_rate": 1.4437157375653959e-05,
+      "loss": 8.8702,
+      "step": 104120
+    },
+    {
+      "epoch": 0.5200129840944843,
+      "grad_norm": 0.09462156146764755,
+      "learning_rate": 1.443565546071241e-05,
+      "loss": 8.8836,
+      "step": 104130
+    },
+    {
+      "epoch": 0.5200629229194237,
+      "grad_norm": 0.09140991419553757,
+      "learning_rate": 1.443415354577086e-05,
+      "loss": 8.8589,
+      "step": 104140
+    },
+    {
+      "epoch": 0.5201128617443631,
+      "grad_norm": 0.09621613472700119,
+      "learning_rate": 1.4432651630829308e-05,
+      "loss": 8.8712,
+      "step": 104150
+    },
+    {
+      "epoch": 0.5201628005693026,
+      "grad_norm": 0.09273172914981842,
+      "learning_rate": 1.4431149715887756e-05,
+      "loss": 8.8778,
+      "step": 104160
+    },
+    {
+      "epoch": 0.5202127393942421,
+      "grad_norm": 0.09596338123083115,
+      "learning_rate": 1.4429647800946207e-05,
+      "loss": 8.8749,
+      "step": 104170
+    },
+    {
+      "epoch": 0.5202626782191815,
+      "grad_norm": 0.09606953710317612,
+      "learning_rate": 1.4428145886004657e-05,
+      "loss": 8.8644,
+      "step": 104180
+    },
+    {
+      "epoch": 0.5203126170441209,
+      "grad_norm": 0.08977267146110535,
+      "learning_rate": 1.4426643971063107e-05,
+      "loss": 8.8545,
+      "step": 104190
+    },
+    {
+      "epoch": 0.5203625558690604,
+      "grad_norm": 0.09220623224973679,
+      "learning_rate": 1.4425142056121555e-05,
+      "loss": 8.8825,
+      "step": 104200
+    },
+    {
+      "epoch": 0.5204124946939999,
+      "grad_norm": 0.09000903367996216,
+      "learning_rate": 1.4423640141180004e-05,
+      "loss": 8.8679,
+      "step": 104210
+    },
+    {
+      "epoch": 0.5204624335189393,
+      "grad_norm": 0.09188482165336609,
+      "learning_rate": 1.4422138226238454e-05,
+      "loss": 8.8606,
+      "step": 104220
+    },
+    {
+      "epoch": 0.5205123723438787,
+      "grad_norm": 0.09331043064594269,
+      "learning_rate": 1.4420636311296904e-05,
+      "loss": 8.8716,
+      "step": 104230
+    },
+    {
+      "epoch": 0.5205623111688182,
+      "grad_norm": 0.0948970839381218,
+      "learning_rate": 1.4419134396355354e-05,
+      "loss": 8.8613,
+      "step": 104240
+    },
+    {
+      "epoch": 0.5206122499937577,
+      "grad_norm": 0.09408802539110184,
+      "learning_rate": 1.4417632481413803e-05,
+      "loss": 8.8697,
+      "step": 104250
+    },
+    {
+      "epoch": 0.5206621888186971,
+      "grad_norm": 0.10012494027614594,
+      "learning_rate": 1.4416130566472251e-05,
+      "loss": 8.8821,
+      "step": 104260
+    },
+    {
+      "epoch": 0.5207121276436365,
+      "grad_norm": 0.08951479196548462,
+      "learning_rate": 1.4414628651530702e-05,
+      "loss": 8.8749,
+      "step": 104270
+    },
+    {
+      "epoch": 0.520762066468576,
+      "grad_norm": 0.09365096688270569,
+      "learning_rate": 1.4413126736589152e-05,
+      "loss": 8.8729,
+      "step": 104280
+    },
+    {
+      "epoch": 0.5208120052935155,
+      "grad_norm": 0.08924220502376556,
+      "learning_rate": 1.4411624821647602e-05,
+      "loss": 8.8587,
+      "step": 104290
+    },
+    {
+      "epoch": 0.5208619441184549,
+      "grad_norm": 0.09431713819503784,
+      "learning_rate": 1.4410122906706052e-05,
+      "loss": 8.8691,
+      "step": 104300
+    },
+    {
+      "epoch": 0.5209118829433943,
+      "grad_norm": 0.0971921905875206,
+      "learning_rate": 1.4408620991764499e-05,
+      "loss": 8.8626,
+      "step": 104310
+    },
+    {
+      "epoch": 0.5209618217683338,
+      "grad_norm": 0.09287826716899872,
+      "learning_rate": 1.4407119076822949e-05,
+      "loss": 8.8534,
+      "step": 104320
+    },
+    {
+      "epoch": 0.5210117605932733,
+      "grad_norm": 0.09646140784025192,
+      "learning_rate": 1.44056171618814e-05,
+      "loss": 8.8684,
+      "step": 104330
+    },
+    {
+      "epoch": 0.5210616994182127,
+      "grad_norm": 0.09730952978134155,
+      "learning_rate": 1.440411524693985e-05,
+      "loss": 8.8817,
+      "step": 104340
+    },
+    {
+      "epoch": 0.5211116382431521,
+      "grad_norm": 0.09068258851766586,
+      "learning_rate": 1.44026133319983e-05,
+      "loss": 8.8648,
+      "step": 104350
+    },
+    {
+      "epoch": 0.5211615770680916,
+      "grad_norm": 0.09530194848775864,
+      "learning_rate": 1.4401111417056746e-05,
+      "loss": 8.8584,
+      "step": 104360
+    },
+    {
+      "epoch": 0.521211515893031,
+      "grad_norm": 0.09383071959018707,
+      "learning_rate": 1.4399609502115197e-05,
+      "loss": 8.8508,
+      "step": 104370
+    },
+    {
+      "epoch": 0.5212614547179705,
+      "grad_norm": 0.096278615295887,
+      "learning_rate": 1.4398107587173647e-05,
+      "loss": 8.8657,
+      "step": 104380
+    },
+    {
+      "epoch": 0.5213113935429099,
+      "grad_norm": 0.09329640120267868,
+      "learning_rate": 1.4396605672232097e-05,
+      "loss": 8.8752,
+      "step": 104390
+    },
+    {
+      "epoch": 0.5213613323678494,
+      "grad_norm": 0.0894843116402626,
+      "learning_rate": 1.4395103757290547e-05,
+      "loss": 8.8745,
+      "step": 104400
+    },
+    {
+      "epoch": 0.5214112711927888,
+      "grad_norm": 0.0973014086484909,
+      "learning_rate": 1.4393601842348994e-05,
+      "loss": 8.8633,
+      "step": 104410
+    },
+    {
+      "epoch": 0.5214612100177283,
+      "grad_norm": 0.09077692031860352,
+      "learning_rate": 1.4392099927407444e-05,
+      "loss": 8.8766,
+      "step": 104420
+    },
+    {
+      "epoch": 0.5215111488426677,
+      "grad_norm": 0.09806881844997406,
+      "learning_rate": 1.4390598012465894e-05,
+      "loss": 8.868,
+      "step": 104430
+    },
+    {
+      "epoch": 0.5215610876676072,
+      "grad_norm": 0.09557566046714783,
+      "learning_rate": 1.4389096097524344e-05,
+      "loss": 8.8595,
+      "step": 104440
+    },
+    {
+      "epoch": 0.5216110264925466,
+      "grad_norm": 0.09814813733100891,
+      "learning_rate": 1.4387594182582795e-05,
+      "loss": 8.8594,
+      "step": 104450
+    },
+    {
+      "epoch": 0.5216609653174861,
+      "grad_norm": 0.09561028331518173,
+      "learning_rate": 1.4386092267641243e-05,
+      "loss": 8.8795,
+      "step": 104460
+    },
+    {
+      "epoch": 0.5217109041424255,
+      "grad_norm": 0.0946744978427887,
+      "learning_rate": 1.4384590352699692e-05,
+      "loss": 8.8686,
+      "step": 104470
+    },
+    {
+      "epoch": 0.521760842967365,
+      "grad_norm": 0.08944974094629288,
+      "learning_rate": 1.4383088437758142e-05,
+      "loss": 8.851,
+      "step": 104480
+    },
+    {
+      "epoch": 0.5218107817923044,
+      "grad_norm": 0.09138962626457214,
+      "learning_rate": 1.4381586522816592e-05,
+      "loss": 8.8796,
+      "step": 104490
+    },
+    {
+      "epoch": 0.5218607206172439,
+      "grad_norm": 0.09289447963237762,
+      "learning_rate": 1.4380084607875042e-05,
+      "loss": 8.8598,
+      "step": 104500
+    },
+    {
+      "epoch": 0.5219106594421833,
+      "grad_norm": 0.0965084508061409,
+      "learning_rate": 1.437858269293349e-05,
+      "loss": 8.8582,
+      "step": 104510
+    },
+    {
+      "epoch": 0.5219605982671228,
+      "grad_norm": 0.08889494836330414,
+      "learning_rate": 1.4377080777991939e-05,
+      "loss": 8.8847,
+      "step": 104520
+    },
+    {
+      "epoch": 0.5220105370920622,
+      "grad_norm": 0.09522352367639542,
+      "learning_rate": 1.437557886305039e-05,
+      "loss": 8.8416,
+      "step": 104530
+    },
+    {
+      "epoch": 0.5220604759170017,
+      "grad_norm": 0.09857291728258133,
+      "learning_rate": 1.437407694810884e-05,
+      "loss": 8.8689,
+      "step": 104540
+    },
+    {
+      "epoch": 0.5221104147419411,
+      "grad_norm": 0.0958976224064827,
+      "learning_rate": 1.437257503316729e-05,
+      "loss": 8.859,
+      "step": 104550
+    },
+    {
+      "epoch": 0.5221603535668806,
+      "grad_norm": 0.08890080451965332,
+      "learning_rate": 1.4371073118225738e-05,
+      "loss": 8.8703,
+      "step": 104560
+    },
+    {
+      "epoch": 0.52221029239182,
+      "grad_norm": 0.09524325281381607,
+      "learning_rate": 1.4369571203284187e-05,
+      "loss": 8.876,
+      "step": 104570
+    },
+    {
+      "epoch": 0.5222602312167595,
+      "grad_norm": 0.0954352393746376,
+      "learning_rate": 1.4368069288342637e-05,
+      "loss": 8.867,
+      "step": 104580
+    },
+    {
+      "epoch": 0.5223101700416989,
+      "grad_norm": 0.09157583862543106,
+      "learning_rate": 1.4366567373401087e-05,
+      "loss": 8.8728,
+      "step": 104590
+    },
+    {
+      "epoch": 0.5223601088666384,
+      "grad_norm": 0.08985961973667145,
+      "learning_rate": 1.4365065458459537e-05,
+      "loss": 8.8768,
+      "step": 104600
+    },
+    {
+      "epoch": 0.5224100476915778,
+      "grad_norm": 0.09679489582777023,
+      "learning_rate": 1.4363563543517986e-05,
+      "loss": 8.8566,
+      "step": 104610
+    },
+    {
+      "epoch": 0.5224599865165173,
+      "grad_norm": 0.09118127077817917,
+      "learning_rate": 1.4362061628576436e-05,
+      "loss": 8.8611,
+      "step": 104620
+    },
+    {
+      "epoch": 0.5225099253414567,
+      "grad_norm": 0.09494442492723465,
+      "learning_rate": 1.4360559713634884e-05,
+      "loss": 8.8769,
+      "step": 104630
+    },
+    {
+      "epoch": 0.5225598641663962,
+      "grad_norm": 0.08483417332172394,
+      "learning_rate": 1.4359057798693334e-05,
+      "loss": 8.8627,
+      "step": 104640
+    },
+    {
+      "epoch": 0.5226098029913356,
+      "grad_norm": 0.09283437579870224,
+      "learning_rate": 1.4357555883751785e-05,
+      "loss": 8.8751,
+      "step": 104650
+    },
+    {
+      "epoch": 0.5226597418162751,
+      "grad_norm": 0.09502757340669632,
+      "learning_rate": 1.4356053968810233e-05,
+      "loss": 8.8703,
+      "step": 104660
+    },
+    {
+      "epoch": 0.5227096806412145,
+      "grad_norm": 0.09698113799095154,
+      "learning_rate": 1.4354552053868683e-05,
+      "loss": 8.8747,
+      "step": 104670
+    },
+    {
+      "epoch": 0.522759619466154,
+      "grad_norm": 0.09982874244451523,
+      "learning_rate": 1.4353050138927132e-05,
+      "loss": 8.8763,
+      "step": 104680
+    },
+    {
+      "epoch": 0.5228095582910934,
+      "grad_norm": 0.09431293606758118,
+      "learning_rate": 1.4351548223985582e-05,
+      "loss": 8.875,
+      "step": 104690
+    },
+    {
+      "epoch": 0.5228594971160329,
+      "grad_norm": 0.09267256408929825,
+      "learning_rate": 1.4350046309044032e-05,
+      "loss": 8.8649,
+      "step": 104700
+    },
+    {
+      "epoch": 0.5229094359409723,
+      "grad_norm": 0.09294337034225464,
+      "learning_rate": 1.434854439410248e-05,
+      "loss": 8.8886,
+      "step": 104710
+    },
+    {
+      "epoch": 0.5229593747659118,
+      "grad_norm": 0.09335044771432877,
+      "learning_rate": 1.434704247916093e-05,
+      "loss": 8.862,
+      "step": 104720
+    },
+    {
+      "epoch": 0.5230093135908512,
+      "grad_norm": 0.094884492456913,
+      "learning_rate": 1.434554056421938e-05,
+      "loss": 8.8607,
+      "step": 104730
+    },
+    {
+      "epoch": 0.5230592524157907,
+      "grad_norm": 0.10318083316087723,
+      "learning_rate": 1.434403864927783e-05,
+      "loss": 8.8615,
+      "step": 104740
+    },
+    {
+      "epoch": 0.5231091912407301,
+      "grad_norm": 0.08785218000411987,
+      "learning_rate": 1.434253673433628e-05,
+      "loss": 8.8801,
+      "step": 104750
+    },
+    {
+      "epoch": 0.5231591300656696,
+      "grad_norm": 0.09105756133794785,
+      "learning_rate": 1.4341034819394728e-05,
+      "loss": 8.88,
+      "step": 104760
+    },
+    {
+      "epoch": 0.523209068890609,
+      "grad_norm": 0.09258095175027847,
+      "learning_rate": 1.4339532904453178e-05,
+      "loss": 8.8629,
+      "step": 104770
+    },
+    {
+      "epoch": 0.5232590077155485,
+      "grad_norm": 0.09019656479358673,
+      "learning_rate": 1.4338030989511628e-05,
+      "loss": 8.876,
+      "step": 104780
+    },
+    {
+      "epoch": 0.5233089465404879,
+      "grad_norm": 0.09507337957620621,
+      "learning_rate": 1.4336529074570077e-05,
+      "loss": 8.8558,
+      "step": 104790
+    },
+    {
+      "epoch": 0.5233588853654274,
+      "grad_norm": 0.09072145074605942,
+      "learning_rate": 1.4335027159628527e-05,
+      "loss": 8.8707,
+      "step": 104800
+    },
+    {
+      "epoch": 0.5234088241903668,
+      "grad_norm": 0.08817720413208008,
+      "learning_rate": 1.4333525244686976e-05,
+      "loss": 8.8654,
+      "step": 104810
+    },
+    {
+      "epoch": 0.5234587630153062,
+      "grad_norm": 0.09037192910909653,
+      "learning_rate": 1.4332023329745426e-05,
+      "loss": 8.8517,
+      "step": 104820
+    },
+    {
+      "epoch": 0.5235087018402457,
+      "grad_norm": 0.087149478495121,
+      "learning_rate": 1.4330521414803876e-05,
+      "loss": 8.8782,
+      "step": 104830
+    },
+    {
+      "epoch": 0.5235586406651851,
+      "grad_norm": 0.10001854598522186,
+      "learning_rate": 1.4329019499862324e-05,
+      "loss": 8.8626,
+      "step": 104840
+    },
+    {
+      "epoch": 0.5236085794901246,
+      "grad_norm": 0.09408778697252274,
+      "learning_rate": 1.4327517584920775e-05,
+      "loss": 8.8606,
+      "step": 104850
+    },
+    {
+      "epoch": 0.523658518315064,
+      "grad_norm": 0.092310830950737,
+      "learning_rate": 1.4326015669979223e-05,
+      "loss": 8.8692,
+      "step": 104860
+    },
+    {
+      "epoch": 0.5237084571400035,
+      "grad_norm": 0.0982789471745491,
+      "learning_rate": 1.4324513755037673e-05,
+      "loss": 8.8585,
+      "step": 104870
+    },
+    {
+      "epoch": 0.5237583959649429,
+      "grad_norm": 0.09160380810499191,
+      "learning_rate": 1.4323011840096124e-05,
+      "loss": 8.8631,
+      "step": 104880
+    },
+    {
+      "epoch": 0.5238083347898824,
+      "grad_norm": 0.09572508931159973,
+      "learning_rate": 1.4321509925154572e-05,
+      "loss": 8.8558,
+      "step": 104890
+    },
+    {
+      "epoch": 0.5238582736148218,
+      "grad_norm": 0.09284794330596924,
+      "learning_rate": 1.4320008010213022e-05,
+      "loss": 8.8739,
+      "step": 104900
+    },
+    {
+      "epoch": 0.5239082124397613,
+      "grad_norm": 0.09144847840070724,
+      "learning_rate": 1.431850609527147e-05,
+      "loss": 8.8543,
+      "step": 104910
+    },
+    {
+      "epoch": 0.5239581512647007,
+      "grad_norm": 0.10097398608922958,
+      "learning_rate": 1.431700418032992e-05,
+      "loss": 8.8667,
+      "step": 104920
+    },
+    {
+      "epoch": 0.5240080900896402,
+      "grad_norm": 0.10776495188474655,
+      "learning_rate": 1.4315502265388371e-05,
+      "loss": 8.8659,
+      "step": 104930
+    },
+    {
+      "epoch": 0.5240580289145796,
+      "grad_norm": 0.08633417636156082,
+      "learning_rate": 1.4314000350446821e-05,
+      "loss": 8.8753,
+      "step": 104940
+    },
+    {
+      "epoch": 0.5241079677395191,
+      "grad_norm": 0.08752107620239258,
+      "learning_rate": 1.431249843550527e-05,
+      "loss": 8.8829,
+      "step": 104950
+    },
+    {
+      "epoch": 0.5241579065644585,
+      "grad_norm": 0.08872366696596146,
+      "learning_rate": 1.4310996520563718e-05,
+      "loss": 8.862,
+      "step": 104960
+    },
+    {
+      "epoch": 0.524207845389398,
+      "grad_norm": 0.09275726228952408,
+      "learning_rate": 1.4309494605622168e-05,
+      "loss": 8.8588,
+      "step": 104970
+    },
+    {
+      "epoch": 0.5242577842143374,
+      "grad_norm": 0.09212324768304825,
+      "learning_rate": 1.4307992690680619e-05,
+      "loss": 8.8612,
+      "step": 104980
+    },
+    {
+      "epoch": 0.5243077230392769,
+      "grad_norm": 0.09363345056772232,
+      "learning_rate": 1.4306490775739069e-05,
+      "loss": 8.8581,
+      "step": 104990
+    },
+    {
+      "epoch": 0.5243576618642163,
+      "grad_norm": 0.09286009520292282,
+      "learning_rate": 1.4304988860797517e-05,
+      "loss": 8.8635,
+      "step": 105000
+    },
+    {
+      "epoch": 0.5244076006891558,
+      "grad_norm": 0.0911487564444542,
+      "learning_rate": 1.4303486945855966e-05,
+      "loss": 8.8541,
+      "step": 105010
+    },
+    {
+      "epoch": 0.5244575395140952,
+      "grad_norm": 0.08649508655071259,
+      "learning_rate": 1.4301985030914416e-05,
+      "loss": 8.8619,
+      "step": 105020
+    },
+    {
+      "epoch": 0.5245074783390347,
+      "grad_norm": 0.09261025488376617,
+      "learning_rate": 1.4300483115972866e-05,
+      "loss": 8.8467,
+      "step": 105030
+    },
+    {
+      "epoch": 0.5245574171639741,
+      "grad_norm": 0.0908532589673996,
+      "learning_rate": 1.4298981201031316e-05,
+      "loss": 8.8492,
+      "step": 105040
+    },
+    {
+      "epoch": 0.5246073559889136,
+      "grad_norm": 0.09164856374263763,
+      "learning_rate": 1.4297479286089765e-05,
+      "loss": 8.863,
+      "step": 105050
+    },
+    {
+      "epoch": 0.524657294813853,
+      "grad_norm": 0.09361206740140915,
+      "learning_rate": 1.4295977371148213e-05,
+      "loss": 8.8672,
+      "step": 105060
+    },
+    {
+      "epoch": 0.5247072336387925,
+      "grad_norm": 0.09860822558403015,
+      "learning_rate": 1.4294475456206663e-05,
+      "loss": 8.8629,
+      "step": 105070
+    },
+    {
+      "epoch": 0.5247571724637319,
+      "grad_norm": 0.09220654517412186,
+      "learning_rate": 1.4292973541265114e-05,
+      "loss": 8.8683,
+      "step": 105080
+    },
+    {
+      "epoch": 0.5248071112886714,
+      "grad_norm": 0.09145854413509369,
+      "learning_rate": 1.4291471626323564e-05,
+      "loss": 8.8745,
+      "step": 105090
+    },
+    {
+      "epoch": 0.5248570501136108,
+      "grad_norm": 0.09256047010421753,
+      "learning_rate": 1.4289969711382014e-05,
+      "loss": 8.8524,
+      "step": 105100
+    },
+    {
+      "epoch": 0.5249069889385503,
+      "grad_norm": 0.09105725586414337,
+      "learning_rate": 1.428846779644046e-05,
+      "loss": 8.8741,
+      "step": 105110
+    },
+    {
+      "epoch": 0.5249569277634897,
+      "grad_norm": 0.09319113194942474,
+      "learning_rate": 1.428696588149891e-05,
+      "loss": 8.8626,
+      "step": 105120
+    },
+    {
+      "epoch": 0.5250068665884292,
+      "grad_norm": 0.09418758749961853,
+      "learning_rate": 1.4285463966557361e-05,
+      "loss": 8.8647,
+      "step": 105130
+    },
+    {
+      "epoch": 0.5250568054133686,
+      "grad_norm": 0.09596774727106094,
+      "learning_rate": 1.4283962051615811e-05,
+      "loss": 8.8574,
+      "step": 105140
+    },
+    {
+      "epoch": 0.525106744238308,
+      "grad_norm": 0.09476704150438309,
+      "learning_rate": 1.4282460136674261e-05,
+      "loss": 8.8575,
+      "step": 105150
+    },
+    {
+      "epoch": 0.5251566830632475,
+      "grad_norm": 0.09615634381771088,
+      "learning_rate": 1.4280958221732708e-05,
+      "loss": 8.8502,
+      "step": 105160
+    },
+    {
+      "epoch": 0.525206621888187,
+      "grad_norm": 0.09071250259876251,
+      "learning_rate": 1.4279456306791158e-05,
+      "loss": 8.8623,
+      "step": 105170
+    },
+    {
+      "epoch": 0.5252565607131264,
+      "grad_norm": 0.08948079496622086,
+      "learning_rate": 1.4277954391849609e-05,
+      "loss": 8.8616,
+      "step": 105180
+    },
+    {
+      "epoch": 0.5253064995380659,
+      "grad_norm": 0.09837137907743454,
+      "learning_rate": 1.4276452476908059e-05,
+      "loss": 8.856,
+      "step": 105190
+    },
+    {
+      "epoch": 0.5253564383630053,
+      "grad_norm": 0.09280554950237274,
+      "learning_rate": 1.4274950561966509e-05,
+      "loss": 8.8634,
+      "step": 105200
+    },
+    {
+      "epoch": 0.5254063771879448,
+      "grad_norm": 0.09170987457036972,
+      "learning_rate": 1.4273448647024956e-05,
+      "loss": 8.8582,
+      "step": 105210
+    },
+    {
+      "epoch": 0.5254563160128842,
+      "grad_norm": 0.09427595138549805,
+      "learning_rate": 1.4271946732083406e-05,
+      "loss": 8.8591,
+      "step": 105220
+    },
+    {
+      "epoch": 0.5255062548378236,
+      "grad_norm": 0.08826853334903717,
+      "learning_rate": 1.4270444817141856e-05,
+      "loss": 8.8695,
+      "step": 105230
+    },
+    {
+      "epoch": 0.5255561936627631,
+      "grad_norm": 0.09086894989013672,
+      "learning_rate": 1.4268942902200306e-05,
+      "loss": 8.8574,
+      "step": 105240
+    },
+    {
+      "epoch": 0.5256061324877026,
+      "grad_norm": 0.09478379786014557,
+      "learning_rate": 1.4267440987258756e-05,
+      "loss": 8.8617,
+      "step": 105250
+    },
+    {
+      "epoch": 0.525656071312642,
+      "grad_norm": 0.09793008863925934,
+      "learning_rate": 1.4265939072317205e-05,
+      "loss": 8.8654,
+      "step": 105260
+    },
+    {
+      "epoch": 0.5257060101375814,
+      "grad_norm": 0.09823063015937805,
+      "learning_rate": 1.4264437157375653e-05,
+      "loss": 8.8549,
+      "step": 105270
+    },
+    {
+      "epoch": 0.5257559489625209,
+      "grad_norm": 0.08841749280691147,
+      "learning_rate": 1.4262935242434104e-05,
+      "loss": 8.865,
+      "step": 105280
+    },
+    {
+      "epoch": 0.5258058877874604,
+      "grad_norm": 0.09652768075466156,
+      "learning_rate": 1.4261433327492554e-05,
+      "loss": 8.8668,
+      "step": 105290
+    },
+    {
+      "epoch": 0.5258558266123998,
+      "grad_norm": 0.09202666580677032,
+      "learning_rate": 1.4259931412551004e-05,
+      "loss": 8.8702,
+      "step": 105300
+    },
+    {
+      "epoch": 0.5259057654373392,
+      "grad_norm": 0.09303510934114456,
+      "learning_rate": 1.4258429497609452e-05,
+      "loss": 8.8646,
+      "step": 105310
+    },
+    {
+      "epoch": 0.5259557042622787,
+      "grad_norm": 0.09391511976718903,
+      "learning_rate": 1.4256927582667901e-05,
+      "loss": 8.8435,
+      "step": 105320
+    },
+    {
+      "epoch": 0.5260056430872182,
+      "grad_norm": 0.10001968592405319,
+      "learning_rate": 1.4255425667726351e-05,
+      "loss": 8.8473,
+      "step": 105330
+    },
+    {
+      "epoch": 0.5260555819121576,
+      "grad_norm": 0.09100840985774994,
+      "learning_rate": 1.4253923752784801e-05,
+      "loss": 8.8602,
+      "step": 105340
+    },
+    {
+      "epoch": 0.526105520737097,
+      "grad_norm": 0.0969325453042984,
+      "learning_rate": 1.4252421837843251e-05,
+      "loss": 8.8614,
+      "step": 105350
+    },
+    {
+      "epoch": 0.5261554595620365,
+      "grad_norm": 0.09638604521751404,
+      "learning_rate": 1.42509199229017e-05,
+      "loss": 8.8666,
+      "step": 105360
+    },
+    {
+      "epoch": 0.526205398386976,
+      "grad_norm": 0.09278382360935211,
+      "learning_rate": 1.4249418007960148e-05,
+      "loss": 8.8661,
+      "step": 105370
+    },
+    {
+      "epoch": 0.5262553372119154,
+      "grad_norm": 0.09172243624925613,
+      "learning_rate": 1.4247916093018599e-05,
+      "loss": 8.8472,
+      "step": 105380
+    },
+    {
+      "epoch": 0.5263052760368548,
+      "grad_norm": 0.09195327013731003,
+      "learning_rate": 1.4246414178077049e-05,
+      "loss": 8.8642,
+      "step": 105390
+    },
+    {
+      "epoch": 0.5263552148617943,
+      "grad_norm": 0.0913047194480896,
+      "learning_rate": 1.4244912263135499e-05,
+      "loss": 8.8626,
+      "step": 105400
+    },
+    {
+      "epoch": 0.5264051536867338,
+      "grad_norm": 0.09170885384082794,
+      "learning_rate": 1.4243410348193947e-05,
+      "loss": 8.8508,
+      "step": 105410
+    },
+    {
+      "epoch": 0.5264550925116732,
+      "grad_norm": 0.0922761857509613,
+      "learning_rate": 1.4241908433252398e-05,
+      "loss": 8.8671,
+      "step": 105420
+    },
+    {
+      "epoch": 0.5265050313366126,
+      "grad_norm": 0.0899137631058693,
+      "learning_rate": 1.4240406518310846e-05,
+      "loss": 8.866,
+      "step": 105430
+    },
+    {
+      "epoch": 0.5265549701615521,
+      "grad_norm": 0.10037946701049805,
+      "learning_rate": 1.4238904603369296e-05,
+      "loss": 8.8532,
+      "step": 105440
+    },
+    {
+      "epoch": 0.5266049089864916,
+      "grad_norm": 0.08979129791259766,
+      "learning_rate": 1.4237402688427746e-05,
+      "loss": 8.8657,
+      "step": 105450
+    },
+    {
+      "epoch": 0.526654847811431,
+      "grad_norm": 0.08900034427642822,
+      "learning_rate": 1.4235900773486195e-05,
+      "loss": 8.8669,
+      "step": 105460
+    },
+    {
+      "epoch": 0.5267047866363704,
+      "grad_norm": 0.08913334459066391,
+      "learning_rate": 1.4234398858544645e-05,
+      "loss": 8.8633,
+      "step": 105470
+    },
+    {
+      "epoch": 0.5267547254613099,
+      "grad_norm": 0.09597323089838028,
+      "learning_rate": 1.4232896943603094e-05,
+      "loss": 8.8568,
+      "step": 105480
+    },
+    {
+      "epoch": 0.5268046642862494,
+      "grad_norm": 0.09855466336011887,
+      "learning_rate": 1.4231395028661544e-05,
+      "loss": 8.8733,
+      "step": 105490
+    },
+    {
+      "epoch": 0.5268546031111888,
+      "grad_norm": 0.096231609582901,
+      "learning_rate": 1.4229893113719994e-05,
+      "loss": 8.8701,
+      "step": 105500
+    },
+    {
+      "epoch": 0.5269045419361282,
+      "grad_norm": 0.09357398748397827,
+      "learning_rate": 1.4228391198778442e-05,
+      "loss": 8.8504,
+      "step": 105510
+    },
+    {
+      "epoch": 0.5269544807610677,
+      "grad_norm": 0.09753704071044922,
+      "learning_rate": 1.4226889283836893e-05,
+      "loss": 8.8658,
+      "step": 105520
+    },
+    {
+      "epoch": 0.5270044195860072,
+      "grad_norm": 0.08956428617238998,
+      "learning_rate": 1.4225387368895341e-05,
+      "loss": 8.8621,
+      "step": 105530
+    },
+    {
+      "epoch": 0.5270543584109466,
+      "grad_norm": 0.10666527599096298,
+      "learning_rate": 1.4223885453953791e-05,
+      "loss": 8.8537,
+      "step": 105540
+    },
+    {
+      "epoch": 0.527104297235886,
+      "grad_norm": 0.0884803831577301,
+      "learning_rate": 1.4222383539012241e-05,
+      "loss": 8.8577,
+      "step": 105550
+    },
+    {
+      "epoch": 0.5271542360608255,
+      "grad_norm": 0.0955420434474945,
+      "learning_rate": 1.422088162407069e-05,
+      "loss": 8.8548,
+      "step": 105560
+    },
+    {
+      "epoch": 0.527204174885765,
+      "grad_norm": 0.09400387853384018,
+      "learning_rate": 1.421937970912914e-05,
+      "loss": 8.8536,
+      "step": 105570
+    },
+    {
+      "epoch": 0.5272541137107044,
+      "grad_norm": 0.09268182516098022,
+      "learning_rate": 1.4217877794187589e-05,
+      "loss": 8.8589,
+      "step": 105580
+    },
+    {
+      "epoch": 0.5273040525356438,
+      "grad_norm": 0.09028761833906174,
+      "learning_rate": 1.4216375879246039e-05,
+      "loss": 8.8849,
+      "step": 105590
+    },
+    {
+      "epoch": 0.5273539913605833,
+      "grad_norm": 0.09481693059206009,
+      "learning_rate": 1.4214873964304489e-05,
+      "loss": 8.8536,
+      "step": 105600
+    },
+    {
+      "epoch": 0.5274039301855228,
+      "grad_norm": 0.09226229786872864,
+      "learning_rate": 1.4213372049362937e-05,
+      "loss": 8.8569,
+      "step": 105610
+    },
+    {
+      "epoch": 0.5274538690104622,
+      "grad_norm": 0.0896633192896843,
+      "learning_rate": 1.4211870134421388e-05,
+      "loss": 8.8696,
+      "step": 105620
+    },
+    {
+      "epoch": 0.5275038078354016,
+      "grad_norm": 0.09037711471319199,
+      "learning_rate": 1.4210368219479838e-05,
+      "loss": 8.8657,
+      "step": 105630
+    },
+    {
+      "epoch": 0.527553746660341,
+      "grad_norm": 0.09311304241418839,
+      "learning_rate": 1.4208866304538286e-05,
+      "loss": 8.8558,
+      "step": 105640
+    },
+    {
+      "epoch": 0.5276036854852806,
+      "grad_norm": 0.09499665349721909,
+      "learning_rate": 1.4207364389596736e-05,
+      "loss": 8.8671,
+      "step": 105650
+    },
+    {
+      "epoch": 0.52765362431022,
+      "grad_norm": 0.09343437850475311,
+      "learning_rate": 1.4205862474655185e-05,
+      "loss": 8.8469,
+      "step": 105660
+    },
+    {
+      "epoch": 0.5277035631351594,
+      "grad_norm": 0.09089982509613037,
+      "learning_rate": 1.4204360559713635e-05,
+      "loss": 8.8575,
+      "step": 105670
+    },
+    {
+      "epoch": 0.5277535019600988,
+      "grad_norm": 0.09557026624679565,
+      "learning_rate": 1.4202858644772085e-05,
+      "loss": 8.8554,
+      "step": 105680
+    },
+    {
+      "epoch": 0.5278034407850384,
+      "grad_norm": 0.09490370750427246,
+      "learning_rate": 1.4201356729830534e-05,
+      "loss": 8.8517,
+      "step": 105690
+    },
+    {
+      "epoch": 0.5278533796099778,
+      "grad_norm": 0.09626539051532745,
+      "learning_rate": 1.4199854814888984e-05,
+      "loss": 8.8529,
+      "step": 105700
+    },
+    {
+      "epoch": 0.5279033184349172,
+      "grad_norm": 0.09501594305038452,
+      "learning_rate": 1.4198352899947432e-05,
+      "loss": 8.8488,
+      "step": 105710
+    },
+    {
+      "epoch": 0.5279532572598566,
+      "grad_norm": 0.09240279346704483,
+      "learning_rate": 1.4196850985005883e-05,
+      "loss": 8.855,
+      "step": 105720
+    },
+    {
+      "epoch": 0.5280031960847962,
+      "grad_norm": 0.09040988236665726,
+      "learning_rate": 1.4195349070064333e-05,
+      "loss": 8.8664,
+      "step": 105730
+    },
+    {
+      "epoch": 0.5280531349097356,
+      "grad_norm": 0.0923992171883583,
+      "learning_rate": 1.4193847155122781e-05,
+      "loss": 8.8492,
+      "step": 105740
+    },
+    {
+      "epoch": 0.528103073734675,
+      "grad_norm": 0.09252621978521347,
+      "learning_rate": 1.4192345240181231e-05,
+      "loss": 8.8525,
+      "step": 105750
+    },
+    {
+      "epoch": 0.5281530125596144,
+      "grad_norm": 0.09642771631479263,
+      "learning_rate": 1.419084332523968e-05,
+      "loss": 8.8541,
+      "step": 105760
+    },
+    {
+      "epoch": 0.528202951384554,
+      "grad_norm": 0.09207062423229218,
+      "learning_rate": 1.418934141029813e-05,
+      "loss": 8.8631,
+      "step": 105770
+    },
+    {
+      "epoch": 0.5282528902094934,
+      "grad_norm": 0.09391027688980103,
+      "learning_rate": 1.418783949535658e-05,
+      "loss": 8.8586,
+      "step": 105780
+    },
+    {
+      "epoch": 0.5283028290344328,
+      "grad_norm": 0.09609290212392807,
+      "learning_rate": 1.418633758041503e-05,
+      "loss": 8.8679,
+      "step": 105790
+    },
+    {
+      "epoch": 0.5283527678593722,
+      "grad_norm": 0.0963875949382782,
+      "learning_rate": 1.4184835665473479e-05,
+      "loss": 8.8622,
+      "step": 105800
+    },
+    {
+      "epoch": 0.5284027066843117,
+      "grad_norm": 0.08848047256469727,
+      "learning_rate": 1.4183333750531929e-05,
+      "loss": 8.8517,
+      "step": 105810
+    },
+    {
+      "epoch": 0.5284526455092512,
+      "grad_norm": 0.08758027851581573,
+      "learning_rate": 1.4181831835590378e-05,
+      "loss": 8.8608,
+      "step": 105820
+    },
+    {
+      "epoch": 0.5285025843341906,
+      "grad_norm": 0.09200868755578995,
+      "learning_rate": 1.4180329920648828e-05,
+      "loss": 8.8544,
+      "step": 105830
+    },
+    {
+      "epoch": 0.52855252315913,
+      "grad_norm": 0.09613315761089325,
+      "learning_rate": 1.4178828005707278e-05,
+      "loss": 8.8563,
+      "step": 105840
+    },
+    {
+      "epoch": 0.5286024619840695,
+      "grad_norm": 0.09787869453430176,
+      "learning_rate": 1.4177326090765726e-05,
+      "loss": 8.8435,
+      "step": 105850
+    },
+    {
+      "epoch": 0.528652400809009,
+      "grad_norm": 0.09556427597999573,
+      "learning_rate": 1.4175824175824177e-05,
+      "loss": 8.861,
+      "step": 105860
+    },
+    {
+      "epoch": 0.5287023396339484,
+      "grad_norm": 0.09230974316596985,
+      "learning_rate": 1.4174322260882625e-05,
+      "loss": 8.8409,
+      "step": 105870
+    },
+    {
+      "epoch": 0.5287522784588878,
+      "grad_norm": 0.0904126837849617,
+      "learning_rate": 1.4172820345941075e-05,
+      "loss": 8.8483,
+      "step": 105880
+    },
+    {
+      "epoch": 0.5288022172838273,
+      "grad_norm": 0.09018577635288239,
+      "learning_rate": 1.4171318430999525e-05,
+      "loss": 8.8603,
+      "step": 105890
+    },
+    {
+      "epoch": 0.5288521561087668,
+      "grad_norm": 0.09044212102890015,
+      "learning_rate": 1.4169816516057974e-05,
+      "loss": 8.845,
+      "step": 105900
+    },
+    {
+      "epoch": 0.5289020949337062,
+      "grad_norm": 0.08735643327236176,
+      "learning_rate": 1.4168314601116424e-05,
+      "loss": 8.8518,
+      "step": 105910
+    },
+    {
+      "epoch": 0.5289520337586456,
+      "grad_norm": 0.08877028524875641,
+      "learning_rate": 1.4166812686174873e-05,
+      "loss": 8.8487,
+      "step": 105920
+    },
+    {
+      "epoch": 0.5290019725835851,
+      "grad_norm": 0.09317787736654282,
+      "learning_rate": 1.4165310771233323e-05,
+      "loss": 8.8551,
+      "step": 105930
+    },
+    {
+      "epoch": 0.5290519114085246,
+      "grad_norm": 0.09516378492116928,
+      "learning_rate": 1.4163808856291773e-05,
+      "loss": 8.8644,
+      "step": 105940
+    },
+    {
+      "epoch": 0.529101850233464,
+      "grad_norm": 0.08973415195941925,
+      "learning_rate": 1.4162306941350223e-05,
+      "loss": 8.876,
+      "step": 105950
+    },
+    {
+      "epoch": 0.5291517890584034,
+      "grad_norm": 0.09267198294401169,
+      "learning_rate": 1.4160805026408672e-05,
+      "loss": 8.8482,
+      "step": 105960
+    },
+    {
+      "epoch": 0.5292017278833429,
+      "grad_norm": 0.0925462618470192,
+      "learning_rate": 1.415930311146712e-05,
+      "loss": 8.8538,
+      "step": 105970
+    },
+    {
+      "epoch": 0.5292516667082824,
+      "grad_norm": 0.09187739342451096,
+      "learning_rate": 1.415780119652557e-05,
+      "loss": 8.8542,
+      "step": 105980
+    },
+    {
+      "epoch": 0.5293016055332218,
+      "grad_norm": 0.08892559260129929,
+      "learning_rate": 1.415629928158402e-05,
+      "loss": 8.8422,
+      "step": 105990
+    },
+    {
+      "epoch": 0.5293515443581612,
+      "grad_norm": 0.09179945290088654,
+      "learning_rate": 1.415479736664247e-05,
+      "loss": 8.8534,
+      "step": 106000
+    },
+    {
+      "epoch": 0.5294014831831007,
+      "grad_norm": 0.09214461594820023,
+      "learning_rate": 1.4153295451700919e-05,
+      "loss": 8.8695,
+      "step": 106010
+    },
+    {
+      "epoch": 0.5294514220080402,
+      "grad_norm": 0.09491454809904099,
+      "learning_rate": 1.4151793536759368e-05,
+      "loss": 8.8725,
+      "step": 106020
+    },
+    {
+      "epoch": 0.5295013608329796,
+      "grad_norm": 0.09679799526929855,
+      "learning_rate": 1.4150291621817818e-05,
+      "loss": 8.845,
+      "step": 106030
+    },
+    {
+      "epoch": 0.529551299657919,
+      "grad_norm": 0.08733250945806503,
+      "learning_rate": 1.4148789706876268e-05,
+      "loss": 8.8541,
+      "step": 106040
+    },
+    {
+      "epoch": 0.5296012384828584,
+      "grad_norm": 0.09057143330574036,
+      "learning_rate": 1.4147287791934718e-05,
+      "loss": 8.8618,
+      "step": 106050
+    },
+    {
+      "epoch": 0.529651177307798,
+      "grad_norm": 0.09323775768280029,
+      "learning_rate": 1.4145785876993167e-05,
+      "loss": 8.8594,
+      "step": 106060
+    },
+    {
+      "epoch": 0.5297011161327374,
+      "grad_norm": 0.09307857602834702,
+      "learning_rate": 1.4144283962051615e-05,
+      "loss": 8.8627,
+      "step": 106070
+    },
+    {
+      "epoch": 0.5297510549576768,
+      "grad_norm": 0.09620136767625809,
+      "learning_rate": 1.4142782047110065e-05,
+      "loss": 8.8487,
+      "step": 106080
+    },
+    {
+      "epoch": 0.5298009937826162,
+      "grad_norm": 0.08902984112501144,
+      "learning_rate": 1.4141280132168515e-05,
+      "loss": 8.866,
+      "step": 106090
+    },
+    {
+      "epoch": 0.5298509326075558,
+      "grad_norm": 0.09401166439056396,
+      "learning_rate": 1.4139778217226966e-05,
+      "loss": 8.8448,
+      "step": 106100
+    },
+    {
+      "epoch": 0.5299008714324952,
+      "grad_norm": 0.09507770091295242,
+      "learning_rate": 1.4138276302285416e-05,
+      "loss": 8.852,
+      "step": 106110
+    },
+    {
+      "epoch": 0.5299508102574346,
+      "grad_norm": 0.09667683392763138,
+      "learning_rate": 1.4136774387343863e-05,
+      "loss": 8.8424,
+      "step": 106120
+    },
+    {
+      "epoch": 0.530000749082374,
+      "grad_norm": 0.09889981150627136,
+      "learning_rate": 1.4135272472402313e-05,
+      "loss": 8.8661,
+      "step": 106130
+    },
+    {
+      "epoch": 0.5300506879073136,
+      "grad_norm": 0.09269087016582489,
+      "learning_rate": 1.4133770557460763e-05,
+      "loss": 8.8558,
+      "step": 106140
+    },
+    {
+      "epoch": 0.530100626732253,
+      "grad_norm": 0.09363644570112228,
+      "learning_rate": 1.4132268642519213e-05,
+      "loss": 8.8576,
+      "step": 106150
+    },
+    {
+      "epoch": 0.5301505655571924,
+      "grad_norm": 0.09286314994096756,
+      "learning_rate": 1.4130766727577663e-05,
+      "loss": 8.8647,
+      "step": 106160
+    },
+    {
+      "epoch": 0.5302005043821318,
+      "grad_norm": 0.09268610924482346,
+      "learning_rate": 1.412926481263611e-05,
+      "loss": 8.8578,
+      "step": 106170
+    },
+    {
+      "epoch": 0.5302504432070714,
+      "grad_norm": 0.09902572631835938,
+      "learning_rate": 1.412776289769456e-05,
+      "loss": 8.8533,
+      "step": 106180
+    },
+    {
+      "epoch": 0.5303003820320108,
+      "grad_norm": 0.093776173889637,
+      "learning_rate": 1.412626098275301e-05,
+      "loss": 8.8647,
+      "step": 106190
+    },
+    {
+      "epoch": 0.5303503208569502,
+      "grad_norm": 0.0955788791179657,
+      "learning_rate": 1.412475906781146e-05,
+      "loss": 8.8529,
+      "step": 106200
+    },
+    {
+      "epoch": 0.5304002596818896,
+      "grad_norm": 0.09192150086164474,
+      "learning_rate": 1.412325715286991e-05,
+      "loss": 8.8548,
+      "step": 106210
+    },
+    {
+      "epoch": 0.5304501985068292,
+      "grad_norm": 0.09892907738685608,
+      "learning_rate": 1.4121755237928358e-05,
+      "loss": 8.8589,
+      "step": 106220
+    },
+    {
+      "epoch": 0.5305001373317686,
+      "grad_norm": 0.09268810600042343,
+      "learning_rate": 1.4120253322986808e-05,
+      "loss": 8.8518,
+      "step": 106230
+    },
+    {
+      "epoch": 0.530550076156708,
+      "grad_norm": 0.09231319278478622,
+      "learning_rate": 1.4118751408045258e-05,
+      "loss": 8.8329,
+      "step": 106240
+    },
+    {
+      "epoch": 0.5306000149816474,
+      "grad_norm": 0.08841013163328171,
+      "learning_rate": 1.4117249493103708e-05,
+      "loss": 8.8476,
+      "step": 106250
+    },
+    {
+      "epoch": 0.530649953806587,
+      "grad_norm": 0.09667849540710449,
+      "learning_rate": 1.4115747578162158e-05,
+      "loss": 8.8692,
+      "step": 106260
+    },
+    {
+      "epoch": 0.5306998926315264,
+      "grad_norm": 0.0873061865568161,
+      "learning_rate": 1.4114245663220607e-05,
+      "loss": 8.8684,
+      "step": 106270
+    },
+    {
+      "epoch": 0.5307498314564658,
+      "grad_norm": 0.0922415629029274,
+      "learning_rate": 1.4112743748279055e-05,
+      "loss": 8.8375,
+      "step": 106280
+    },
+    {
+      "epoch": 0.5307997702814052,
+      "grad_norm": 0.09075237065553665,
+      "learning_rate": 1.4111241833337505e-05,
+      "loss": 8.8494,
+      "step": 106290
+    },
+    {
+      "epoch": 0.5308497091063448,
+      "grad_norm": 0.09417222440242767,
+      "learning_rate": 1.4109739918395956e-05,
+      "loss": 8.8416,
+      "step": 106300
+    },
+    {
+      "epoch": 0.5308996479312842,
+      "grad_norm": 0.09558428823947906,
+      "learning_rate": 1.4108238003454406e-05,
+      "loss": 8.8383,
+      "step": 106310
+    },
+    {
+      "epoch": 0.5309495867562236,
+      "grad_norm": 0.10185667872428894,
+      "learning_rate": 1.4106736088512854e-05,
+      "loss": 8.8451,
+      "step": 106320
+    },
+    {
+      "epoch": 0.530999525581163,
+      "grad_norm": 0.09136248379945755,
+      "learning_rate": 1.4105234173571303e-05,
+      "loss": 8.8492,
+      "step": 106330
+    },
+    {
+      "epoch": 0.5310494644061026,
+      "grad_norm": 0.09055881202220917,
+      "learning_rate": 1.4103732258629753e-05,
+      "loss": 8.8517,
+      "step": 106340
+    },
+    {
+      "epoch": 0.531099403231042,
+      "grad_norm": 0.08744243532419205,
+      "learning_rate": 1.4102230343688203e-05,
+      "loss": 8.8688,
+      "step": 106350
+    },
+    {
+      "epoch": 0.5311493420559814,
+      "grad_norm": 0.0952538251876831,
+      "learning_rate": 1.4100728428746653e-05,
+      "loss": 8.8506,
+      "step": 106360
+    },
+    {
+      "epoch": 0.5311992808809208,
+      "grad_norm": 0.0906098335981369,
+      "learning_rate": 1.4099226513805102e-05,
+      "loss": 8.8518,
+      "step": 106370
+    },
+    {
+      "epoch": 0.5312492197058604,
+      "grad_norm": 0.0909523218870163,
+      "learning_rate": 1.409772459886355e-05,
+      "loss": 8.8542,
+      "step": 106380
+    },
+    {
+      "epoch": 0.5312991585307998,
+      "grad_norm": 0.0970349833369255,
+      "learning_rate": 1.4096222683922e-05,
+      "loss": 8.8624,
+      "step": 106390
+    },
+    {
+      "epoch": 0.5313490973557392,
+      "grad_norm": 0.09205695241689682,
+      "learning_rate": 1.409472076898045e-05,
+      "loss": 8.8506,
+      "step": 106400
+    },
+    {
+      "epoch": 0.5313990361806786,
+      "grad_norm": 0.09964334964752197,
+      "learning_rate": 1.40932188540389e-05,
+      "loss": 8.8655,
+      "step": 106410
+    },
+    {
+      "epoch": 0.5314489750056182,
+      "grad_norm": 0.0929349958896637,
+      "learning_rate": 1.409171693909735e-05,
+      "loss": 8.8504,
+      "step": 106420
+    },
+    {
+      "epoch": 0.5314989138305576,
+      "grad_norm": 0.08960101008415222,
+      "learning_rate": 1.40902150241558e-05,
+      "loss": 8.8597,
+      "step": 106430
+    },
+    {
+      "epoch": 0.531548852655497,
+      "grad_norm": 0.09919091314077377,
+      "learning_rate": 1.4088713109214248e-05,
+      "loss": 8.8413,
+      "step": 106440
+    },
+    {
+      "epoch": 0.5315987914804364,
+      "grad_norm": 0.10037015378475189,
+      "learning_rate": 1.4087211194272698e-05,
+      "loss": 8.8489,
+      "step": 106450
+    },
+    {
+      "epoch": 0.531648730305376,
+      "grad_norm": 0.09269418567419052,
+      "learning_rate": 1.4085709279331148e-05,
+      "loss": 8.8392,
+      "step": 106460
+    },
+    {
+      "epoch": 0.5316986691303154,
+      "grad_norm": 0.09641455113887787,
+      "learning_rate": 1.4084207364389597e-05,
+      "loss": 8.8553,
+      "step": 106470
+    },
+    {
+      "epoch": 0.5317486079552548,
+      "grad_norm": 0.09411516040563583,
+      "learning_rate": 1.4082705449448047e-05,
+      "loss": 8.845,
+      "step": 106480
+    },
+    {
+      "epoch": 0.5317985467801942,
+      "grad_norm": 0.08999986946582794,
+      "learning_rate": 1.4081203534506495e-05,
+      "loss": 8.8416,
+      "step": 106490
+    },
+    {
+      "epoch": 0.5318484856051338,
+      "grad_norm": 0.0922091007232666,
+      "learning_rate": 1.4079701619564946e-05,
+      "loss": 8.8469,
+      "step": 106500
+    },
+    {
+      "epoch": 0.5318984244300732,
+      "grad_norm": 0.09420941770076752,
+      "learning_rate": 1.4078199704623396e-05,
+      "loss": 8.862,
+      "step": 106510
+    },
+    {
+      "epoch": 0.5319483632550126,
+      "grad_norm": 0.09016049653291702,
+      "learning_rate": 1.4076697789681844e-05,
+      "loss": 8.8476,
+      "step": 106520
+    },
+    {
+      "epoch": 0.531998302079952,
+      "grad_norm": 0.104548379778862,
+      "learning_rate": 1.4075195874740295e-05,
+      "loss": 8.8476,
+      "step": 106530
+    },
+    {
+      "epoch": 0.5320482409048916,
+      "grad_norm": 0.08957797288894653,
+      "learning_rate": 1.4073693959798743e-05,
+      "loss": 8.856,
+      "step": 106540
+    },
+    {
+      "epoch": 0.532098179729831,
+      "grad_norm": 0.09814672917127609,
+      "learning_rate": 1.4072192044857193e-05,
+      "loss": 8.8517,
+      "step": 106550
+    },
+    {
+      "epoch": 0.5321481185547704,
+      "grad_norm": 0.09064110368490219,
+      "learning_rate": 1.4070690129915643e-05,
+      "loss": 8.8456,
+      "step": 106560
+    },
+    {
+      "epoch": 0.5321980573797098,
+      "grad_norm": 0.09723041951656342,
+      "learning_rate": 1.4069188214974092e-05,
+      "loss": 8.8518,
+      "step": 106570
+    },
+    {
+      "epoch": 0.5322479962046494,
+      "grad_norm": 0.09290885925292969,
+      "learning_rate": 1.4067686300032542e-05,
+      "loss": 8.8574,
+      "step": 106580
+    },
+    {
+      "epoch": 0.5322979350295888,
+      "grad_norm": 0.09443046897649765,
+      "learning_rate": 1.4066184385090992e-05,
+      "loss": 8.8631,
+      "step": 106590
+    },
+    {
+      "epoch": 0.5323478738545282,
+      "grad_norm": 0.09200415760278702,
+      "learning_rate": 1.406468247014944e-05,
+      "loss": 8.8491,
+      "step": 106600
+    },
+    {
+      "epoch": 0.5323978126794676,
+      "grad_norm": 0.09204615652561188,
+      "learning_rate": 1.4063180555207891e-05,
+      "loss": 8.8424,
+      "step": 106610
+    },
+    {
+      "epoch": 0.5324477515044072,
+      "grad_norm": 0.0978984385728836,
+      "learning_rate": 1.406167864026634e-05,
+      "loss": 8.8551,
+      "step": 106620
+    },
+    {
+      "epoch": 0.5324976903293466,
+      "grad_norm": 0.10067642480134964,
+      "learning_rate": 1.406017672532479e-05,
+      "loss": 8.8473,
+      "step": 106630
+    },
+    {
+      "epoch": 0.532547629154286,
+      "grad_norm": 0.09151472896337509,
+      "learning_rate": 1.405867481038324e-05,
+      "loss": 8.8594,
+      "step": 106640
+    },
+    {
+      "epoch": 0.5325975679792254,
+      "grad_norm": 0.09709201753139496,
+      "learning_rate": 1.4057172895441688e-05,
+      "loss": 8.8443,
+      "step": 106650
+    },
+    {
+      "epoch": 0.532647506804165,
+      "grad_norm": 0.09483177214860916,
+      "learning_rate": 1.4055670980500138e-05,
+      "loss": 8.8499,
+      "step": 106660
+    },
+    {
+      "epoch": 0.5326974456291044,
+      "grad_norm": 0.09062449634075165,
+      "learning_rate": 1.4054169065558587e-05,
+      "loss": 8.8465,
+      "step": 106670
+    },
+    {
+      "epoch": 0.5327473844540438,
+      "grad_norm": 0.08802146464586258,
+      "learning_rate": 1.4052667150617037e-05,
+      "loss": 8.8475,
+      "step": 106680
+    },
+    {
+      "epoch": 0.5327973232789832,
+      "grad_norm": 0.09027039259672165,
+      "learning_rate": 1.4051165235675487e-05,
+      "loss": 8.8396,
+      "step": 106690
+    },
+    {
+      "epoch": 0.5328472621039227,
+      "grad_norm": 0.08948853611946106,
+      "learning_rate": 1.4049663320733936e-05,
+      "loss": 8.855,
+      "step": 106700
+    },
+    {
+      "epoch": 0.5328972009288622,
+      "grad_norm": 0.09329523891210556,
+      "learning_rate": 1.4048161405792386e-05,
+      "loss": 8.8439,
+      "step": 106710
+    },
+    {
+      "epoch": 0.5329471397538016,
+      "grad_norm": 0.09256135672330856,
+      "learning_rate": 1.4046659490850834e-05,
+      "loss": 8.8441,
+      "step": 106720
+    },
+    {
+      "epoch": 0.532997078578741,
+      "grad_norm": 0.0897359699010849,
+      "learning_rate": 1.4045157575909285e-05,
+      "loss": 8.8476,
+      "step": 106730
+    },
+    {
+      "epoch": 0.5330470174036805,
+      "grad_norm": 0.10047424584627151,
+      "learning_rate": 1.4043655660967735e-05,
+      "loss": 8.858,
+      "step": 106740
+    },
+    {
+      "epoch": 0.53309695622862,
+      "grad_norm": 0.09750477969646454,
+      "learning_rate": 1.4042153746026185e-05,
+      "loss": 8.8484,
+      "step": 106750
+    },
+    {
+      "epoch": 0.5331468950535594,
+      "grad_norm": 0.09380090236663818,
+      "learning_rate": 1.4040651831084633e-05,
+      "loss": 8.8272,
+      "step": 106760
+    },
+    {
+      "epoch": 0.5331968338784988,
+      "grad_norm": 0.09145848453044891,
+      "learning_rate": 1.4039149916143082e-05,
+      "loss": 8.8505,
+      "step": 106770
+    },
+    {
+      "epoch": 0.5332467727034383,
+      "grad_norm": 0.09653259068727493,
+      "learning_rate": 1.4037648001201532e-05,
+      "loss": 8.837,
+      "step": 106780
+    },
+    {
+      "epoch": 0.5332967115283778,
+      "grad_norm": 0.09636209160089493,
+      "learning_rate": 1.4036146086259982e-05,
+      "loss": 8.8567,
+      "step": 106790
+    },
+    {
+      "epoch": 0.5333466503533172,
+      "grad_norm": 0.09500285983085632,
+      "learning_rate": 1.4034644171318432e-05,
+      "loss": 8.8523,
+      "step": 106800
+    },
+    {
+      "epoch": 0.5333965891782566,
+      "grad_norm": 0.09158430993556976,
+      "learning_rate": 1.4033142256376881e-05,
+      "loss": 8.8477,
+      "step": 106810
+    },
+    {
+      "epoch": 0.533446528003196,
+      "grad_norm": 0.09533253312110901,
+      "learning_rate": 1.403164034143533e-05,
+      "loss": 8.8366,
+      "step": 106820
+    },
+    {
+      "epoch": 0.5334964668281356,
+      "grad_norm": 0.09380104392766953,
+      "learning_rate": 1.403013842649378e-05,
+      "loss": 8.8238,
+      "step": 106830
+    },
+    {
+      "epoch": 0.533546405653075,
+      "grad_norm": 0.09858004003763199,
+      "learning_rate": 1.402863651155223e-05,
+      "loss": 8.8324,
+      "step": 106840
+    },
+    {
+      "epoch": 0.5335963444780144,
+      "grad_norm": 0.09267483651638031,
+      "learning_rate": 1.402713459661068e-05,
+      "loss": 8.8385,
+      "step": 106850
+    },
+    {
+      "epoch": 0.5336462833029538,
+      "grad_norm": 0.0965743437409401,
+      "learning_rate": 1.4025632681669128e-05,
+      "loss": 8.8602,
+      "step": 106860
+    },
+    {
+      "epoch": 0.5336962221278934,
+      "grad_norm": 0.09583596140146255,
+      "learning_rate": 1.4024130766727577e-05,
+      "loss": 8.8519,
+      "step": 106870
+    },
+    {
+      "epoch": 0.5337461609528328,
+      "grad_norm": 0.095204658806324,
+      "learning_rate": 1.4022628851786027e-05,
+      "loss": 8.8489,
+      "step": 106880
+    },
+    {
+      "epoch": 0.5337960997777722,
+      "grad_norm": 0.09433889389038086,
+      "learning_rate": 1.4021126936844477e-05,
+      "loss": 8.8466,
+      "step": 106890
+    },
+    {
+      "epoch": 0.5338460386027116,
+      "grad_norm": 0.09470517933368683,
+      "learning_rate": 1.4019625021902927e-05,
+      "loss": 8.8563,
+      "step": 106900
+    },
+    {
+      "epoch": 0.5338959774276512,
+      "grad_norm": 0.09337898343801498,
+      "learning_rate": 1.4018123106961378e-05,
+      "loss": 8.842,
+      "step": 106910
+    },
+    {
+      "epoch": 0.5339459162525906,
+      "grad_norm": 0.09626980870962143,
+      "learning_rate": 1.4016621192019824e-05,
+      "loss": 8.8568,
+      "step": 106920
+    },
+    {
+      "epoch": 0.53399585507753,
+      "grad_norm": 0.09709392488002777,
+      "learning_rate": 1.4015119277078275e-05,
+      "loss": 8.8477,
+      "step": 106930
+    },
+    {
+      "epoch": 0.5340457939024694,
+      "grad_norm": 0.09491003304719925,
+      "learning_rate": 1.4013617362136725e-05,
+      "loss": 8.8506,
+      "step": 106940
+    },
+    {
+      "epoch": 0.534095732727409,
+      "grad_norm": 0.09251262992620468,
+      "learning_rate": 1.4012115447195175e-05,
+      "loss": 8.8449,
+      "step": 106950
+    },
+    {
+      "epoch": 0.5341456715523484,
+      "grad_norm": 0.0920642837882042,
+      "learning_rate": 1.4010613532253625e-05,
+      "loss": 8.8396,
+      "step": 106960
+    },
+    {
+      "epoch": 0.5341956103772878,
+      "grad_norm": 0.09323357790708542,
+      "learning_rate": 1.4009111617312072e-05,
+      "loss": 8.8498,
+      "step": 106970
+    },
+    {
+      "epoch": 0.5342455492022272,
+      "grad_norm": 0.09278106689453125,
+      "learning_rate": 1.4007609702370522e-05,
+      "loss": 8.8429,
+      "step": 106980
+    },
+    {
+      "epoch": 0.5342954880271668,
+      "grad_norm": 0.09746766835451126,
+      "learning_rate": 1.4006107787428972e-05,
+      "loss": 8.8352,
+      "step": 106990
+    },
+    {
+      "epoch": 0.5343454268521062,
+      "grad_norm": 0.10587667673826218,
+      "learning_rate": 1.4004605872487422e-05,
+      "loss": 8.8457,
+      "step": 107000
+    },
+    {
+      "epoch": 0.5343953656770456,
+      "grad_norm": 0.09702330827713013,
+      "learning_rate": 1.4003103957545873e-05,
+      "loss": 8.8515,
+      "step": 107010
+    },
+    {
+      "epoch": 0.534445304501985,
+      "grad_norm": 0.09104038029909134,
+      "learning_rate": 1.400160204260432e-05,
+      "loss": 8.847,
+      "step": 107020
+    },
+    {
+      "epoch": 0.5344952433269246,
+      "grad_norm": 0.08740893006324768,
+      "learning_rate": 1.400010012766277e-05,
+      "loss": 8.856,
+      "step": 107030
+    },
+    {
+      "epoch": 0.534545182151864,
+      "grad_norm": 0.09178484231233597,
+      "learning_rate": 1.399859821272122e-05,
+      "loss": 8.8457,
+      "step": 107040
+    },
+    {
+      "epoch": 0.5345951209768034,
+      "grad_norm": 0.09630772471427917,
+      "learning_rate": 1.399709629777967e-05,
+      "loss": 8.8488,
+      "step": 107050
+    },
+    {
+      "epoch": 0.5346450598017428,
+      "grad_norm": 0.09321489930152893,
+      "learning_rate": 1.399559438283812e-05,
+      "loss": 8.845,
+      "step": 107060
+    },
+    {
+      "epoch": 0.5346949986266823,
+      "grad_norm": 0.09223134070634842,
+      "learning_rate": 1.3994092467896569e-05,
+      "loss": 8.8505,
+      "step": 107070
+    },
+    {
+      "epoch": 0.5347449374516218,
+      "grad_norm": 0.09391648322343826,
+      "learning_rate": 1.3992590552955017e-05,
+      "loss": 8.8556,
+      "step": 107080
+    },
+    {
+      "epoch": 0.5347948762765612,
+      "grad_norm": 0.08899156749248505,
+      "learning_rate": 1.3991088638013467e-05,
+      "loss": 8.8542,
+      "step": 107090
+    },
+    {
+      "epoch": 0.5348448151015006,
+      "grad_norm": 0.09277118742465973,
+      "learning_rate": 1.3989586723071917e-05,
+      "loss": 8.8285,
+      "step": 107100
+    },
+    {
+      "epoch": 0.5348947539264401,
+      "grad_norm": 0.09124419838190079,
+      "learning_rate": 1.3988084808130368e-05,
+      "loss": 8.8385,
+      "step": 107110
+    },
+    {
+      "epoch": 0.5349446927513796,
+      "grad_norm": 0.08537127077579498,
+      "learning_rate": 1.3986582893188816e-05,
+      "loss": 8.8404,
+      "step": 107120
+    },
+    {
+      "epoch": 0.534994631576319,
+      "grad_norm": 0.09226098656654358,
+      "learning_rate": 1.3985080978247265e-05,
+      "loss": 8.8535,
+      "step": 107130
+    },
+    {
+      "epoch": 0.5350445704012584,
+      "grad_norm": 0.09150153398513794,
+      "learning_rate": 1.3983579063305715e-05,
+      "loss": 8.8599,
+      "step": 107140
+    },
+    {
+      "epoch": 0.535094509226198,
+      "grad_norm": 0.09004461765289307,
+      "learning_rate": 1.3982077148364165e-05,
+      "loss": 8.836,
+      "step": 107150
+    },
+    {
+      "epoch": 0.5351444480511374,
+      "grad_norm": 0.08848071843385696,
+      "learning_rate": 1.3980575233422615e-05,
+      "loss": 8.84,
+      "step": 107160
+    },
+    {
+      "epoch": 0.5351943868760768,
+      "grad_norm": 0.09761686623096466,
+      "learning_rate": 1.3979073318481064e-05,
+      "loss": 8.8489,
+      "step": 107170
+    },
+    {
+      "epoch": 0.5352443257010162,
+      "grad_norm": 0.09823184460401535,
+      "learning_rate": 1.3977571403539512e-05,
+      "loss": 8.8443,
+      "step": 107180
+    },
+    {
+      "epoch": 0.5352942645259557,
+      "grad_norm": 0.10020698606967926,
+      "learning_rate": 1.3976069488597962e-05,
+      "loss": 8.8401,
+      "step": 107190
+    },
+    {
+      "epoch": 0.5353442033508952,
+      "grad_norm": 0.08891124278306961,
+      "learning_rate": 1.3974567573656412e-05,
+      "loss": 8.8526,
+      "step": 107200
+    },
+    {
+      "epoch": 0.5353941421758346,
+      "grad_norm": 0.10143271088600159,
+      "learning_rate": 1.3973065658714863e-05,
+      "loss": 8.8288,
+      "step": 107210
+    },
+    {
+      "epoch": 0.535444081000774,
+      "grad_norm": 0.09501948952674866,
+      "learning_rate": 1.3971563743773311e-05,
+      "loss": 8.8283,
+      "step": 107220
+    },
+    {
+      "epoch": 0.5354940198257135,
+      "grad_norm": 0.09250184148550034,
+      "learning_rate": 1.3970061828831761e-05,
+      "loss": 8.8371,
+      "step": 107230
+    },
+    {
+      "epoch": 0.535543958650653,
+      "grad_norm": 0.10043402016162872,
+      "learning_rate": 1.396855991389021e-05,
+      "loss": 8.8402,
+      "step": 107240
+    },
+    {
+      "epoch": 0.5355938974755924,
+      "grad_norm": 0.0956840068101883,
+      "learning_rate": 1.396705799894866e-05,
+      "loss": 8.8494,
+      "step": 107250
+    },
+    {
+      "epoch": 0.5356438363005318,
+      "grad_norm": 0.09393565356731415,
+      "learning_rate": 1.396555608400711e-05,
+      "loss": 8.8335,
+      "step": 107260
+    },
+    {
+      "epoch": 0.5356937751254713,
+      "grad_norm": 0.09771479666233063,
+      "learning_rate": 1.3964054169065559e-05,
+      "loss": 8.8376,
+      "step": 107270
+    },
+    {
+      "epoch": 0.5357437139504108,
+      "grad_norm": 0.08908728510141373,
+      "learning_rate": 1.3962552254124009e-05,
+      "loss": 8.8401,
+      "step": 107280
+    },
+    {
+      "epoch": 0.5357936527753502,
+      "grad_norm": 0.08806288987398148,
+      "learning_rate": 1.3961050339182457e-05,
+      "loss": 8.8491,
+      "step": 107290
+    },
+    {
+      "epoch": 0.5358435916002896,
+      "grad_norm": 0.09544865041971207,
+      "learning_rate": 1.3959548424240907e-05,
+      "loss": 8.8399,
+      "step": 107300
+    },
+    {
+      "epoch": 0.5358935304252291,
+      "grad_norm": 0.0903419628739357,
+      "learning_rate": 1.3958046509299358e-05,
+      "loss": 8.8217,
+      "step": 107310
+    },
+    {
+      "epoch": 0.5359434692501686,
+      "grad_norm": 0.09364979714155197,
+      "learning_rate": 1.3956544594357806e-05,
+      "loss": 8.8447,
+      "step": 107320
+    },
+    {
+      "epoch": 0.535993408075108,
+      "grad_norm": 0.09100334346294403,
+      "learning_rate": 1.3955042679416256e-05,
+      "loss": 8.8389,
+      "step": 107330
+    },
+    {
+      "epoch": 0.5360433469000474,
+      "grad_norm": 0.09545484930276871,
+      "learning_rate": 1.3953540764474705e-05,
+      "loss": 8.8371,
+      "step": 107340
+    },
+    {
+      "epoch": 0.5360932857249869,
+      "grad_norm": 0.09141393005847931,
+      "learning_rate": 1.3952038849533155e-05,
+      "loss": 8.8562,
+      "step": 107350
+    },
+    {
+      "epoch": 0.5361432245499264,
+      "grad_norm": 0.09346693754196167,
+      "learning_rate": 1.3950536934591605e-05,
+      "loss": 8.8436,
+      "step": 107360
+    },
+    {
+      "epoch": 0.5361931633748658,
+      "grad_norm": 0.09371886402368546,
+      "learning_rate": 1.3949035019650054e-05,
+      "loss": 8.8336,
+      "step": 107370
+    },
+    {
+      "epoch": 0.5362431021998052,
+      "grad_norm": 0.08820048719644547,
+      "learning_rate": 1.3947533104708504e-05,
+      "loss": 8.8447,
+      "step": 107380
+    },
+    {
+      "epoch": 0.5362930410247447,
+      "grad_norm": 0.09496960043907166,
+      "learning_rate": 1.3946031189766954e-05,
+      "loss": 8.8354,
+      "step": 107390
+    },
+    {
+      "epoch": 0.5363429798496842,
+      "grad_norm": 0.09885689616203308,
+      "learning_rate": 1.3944529274825402e-05,
+      "loss": 8.8273,
+      "step": 107400
+    },
+    {
+      "epoch": 0.5363929186746236,
+      "grad_norm": 0.09907709807157516,
+      "learning_rate": 1.3943027359883853e-05,
+      "loss": 8.8314,
+      "step": 107410
+    },
+    {
+      "epoch": 0.536442857499563,
+      "grad_norm": 0.08787566423416138,
+      "learning_rate": 1.3941525444942301e-05,
+      "loss": 8.8396,
+      "step": 107420
+    },
+    {
+      "epoch": 0.5364927963245025,
+      "grad_norm": 0.0915251299738884,
+      "learning_rate": 1.3940023530000751e-05,
+      "loss": 8.8471,
+      "step": 107430
+    },
+    {
+      "epoch": 0.536542735149442,
+      "grad_norm": 0.09245522320270538,
+      "learning_rate": 1.3938521615059201e-05,
+      "loss": 8.8446,
+      "step": 107440
+    },
+    {
+      "epoch": 0.5365926739743814,
+      "grad_norm": 0.09490872919559479,
+      "learning_rate": 1.393701970011765e-05,
+      "loss": 8.8361,
+      "step": 107450
+    },
+    {
+      "epoch": 0.5366426127993208,
+      "grad_norm": 0.1009124219417572,
+      "learning_rate": 1.39355177851761e-05,
+      "loss": 8.8215,
+      "step": 107460
+    },
+    {
+      "epoch": 0.5366925516242603,
+      "grad_norm": 0.10078973323106766,
+      "learning_rate": 1.3934015870234549e-05,
+      "loss": 8.8342,
+      "step": 107470
+    },
+    {
+      "epoch": 0.5367424904491997,
+      "grad_norm": 0.09087687730789185,
+      "learning_rate": 1.3932513955292999e-05,
+      "loss": 8.8498,
+      "step": 107480
+    },
+    {
+      "epoch": 0.5367924292741392,
+      "grad_norm": 0.09109964966773987,
+      "learning_rate": 1.3931012040351449e-05,
+      "loss": 8.8493,
+      "step": 107490
+    },
+    {
+      "epoch": 0.5368423680990786,
+      "grad_norm": 0.09533953666687012,
+      "learning_rate": 1.3929510125409897e-05,
+      "loss": 8.8549,
+      "step": 107500
+    },
+    {
+      "epoch": 0.5368923069240181,
+      "grad_norm": 0.09448252618312836,
+      "learning_rate": 1.3928008210468348e-05,
+      "loss": 8.8328,
+      "step": 107510
+    },
+    {
+      "epoch": 0.5369422457489575,
+      "grad_norm": 0.0955210030078888,
+      "learning_rate": 1.3926506295526796e-05,
+      "loss": 8.8328,
+      "step": 107520
+    },
+    {
+      "epoch": 0.536992184573897,
+      "grad_norm": 0.10150817781686783,
+      "learning_rate": 1.3925004380585246e-05,
+      "loss": 8.8467,
+      "step": 107530
+    },
+    {
+      "epoch": 0.5370421233988364,
+      "grad_norm": 0.09742490202188492,
+      "learning_rate": 1.3923502465643696e-05,
+      "loss": 8.8382,
+      "step": 107540
+    },
+    {
+      "epoch": 0.5370920622237759,
+      "grad_norm": 0.09283208101987839,
+      "learning_rate": 1.3922000550702147e-05,
+      "loss": 8.8348,
+      "step": 107550
+    },
+    {
+      "epoch": 0.5371420010487153,
+      "grad_norm": 0.09196760505437851,
+      "learning_rate": 1.3920498635760595e-05,
+      "loss": 8.8501,
+      "step": 107560
+    },
+    {
+      "epoch": 0.5371919398736548,
+      "grad_norm": 0.09395366907119751,
+      "learning_rate": 1.3918996720819044e-05,
+      "loss": 8.8297,
+      "step": 107570
+    },
+    {
+      "epoch": 0.5372418786985942,
+      "grad_norm": 0.09952932596206665,
+      "learning_rate": 1.3917494805877494e-05,
+      "loss": 8.8295,
+      "step": 107580
+    },
+    {
+      "epoch": 0.5372918175235337,
+      "grad_norm": 0.10399974137544632,
+      "learning_rate": 1.3915992890935944e-05,
+      "loss": 8.8323,
+      "step": 107590
+    },
+    {
+      "epoch": 0.5373417563484731,
+      "grad_norm": 0.08807992935180664,
+      "learning_rate": 1.3914490975994394e-05,
+      "loss": 8.8514,
+      "step": 107600
+    },
+    {
+      "epoch": 0.5373916951734126,
+      "grad_norm": 0.09528198093175888,
+      "learning_rate": 1.3912989061052843e-05,
+      "loss": 8.8302,
+      "step": 107610
+    },
+    {
+      "epoch": 0.537441633998352,
+      "grad_norm": 0.09758550673723221,
+      "learning_rate": 1.3911487146111291e-05,
+      "loss": 8.8455,
+      "step": 107620
+    },
+    {
+      "epoch": 0.5374915728232915,
+      "grad_norm": 0.0921962559223175,
+      "learning_rate": 1.3909985231169741e-05,
+      "loss": 8.8425,
+      "step": 107630
+    },
+    {
+      "epoch": 0.5375415116482309,
+      "grad_norm": 0.09394194185733795,
+      "learning_rate": 1.3908483316228191e-05,
+      "loss": 8.8426,
+      "step": 107640
+    },
+    {
+      "epoch": 0.5375914504731704,
+      "grad_norm": 0.0899246484041214,
+      "learning_rate": 1.3906981401286642e-05,
+      "loss": 8.8407,
+      "step": 107650
+    },
+    {
+      "epoch": 0.5376413892981098,
+      "grad_norm": 0.09396055340766907,
+      "learning_rate": 1.390547948634509e-05,
+      "loss": 8.8325,
+      "step": 107660
+    },
+    {
+      "epoch": 0.5376913281230493,
+      "grad_norm": 0.09703001379966736,
+      "learning_rate": 1.3903977571403539e-05,
+      "loss": 8.8322,
+      "step": 107670
+    },
+    {
+      "epoch": 0.5377412669479887,
+      "grad_norm": 0.08996451646089554,
+      "learning_rate": 1.3902475656461989e-05,
+      "loss": 8.8422,
+      "step": 107680
+    },
+    {
+      "epoch": 0.5377912057729282,
+      "grad_norm": 0.09080103039741516,
+      "learning_rate": 1.3900973741520439e-05,
+      "loss": 8.8472,
+      "step": 107690
+    },
+    {
+      "epoch": 0.5378411445978676,
+      "grad_norm": 0.09670452773571014,
+      "learning_rate": 1.3899471826578889e-05,
+      "loss": 8.8514,
+      "step": 107700
+    },
+    {
+      "epoch": 0.5378910834228071,
+      "grad_norm": 0.0983625054359436,
+      "learning_rate": 1.389796991163734e-05,
+      "loss": 8.8386,
+      "step": 107710
+    },
+    {
+      "epoch": 0.5379410222477465,
+      "grad_norm": 0.09571649879217148,
+      "learning_rate": 1.3896467996695786e-05,
+      "loss": 8.8353,
+      "step": 107720
+    },
+    {
+      "epoch": 0.537990961072686,
+      "grad_norm": 0.0922119989991188,
+      "learning_rate": 1.3894966081754236e-05,
+      "loss": 8.8401,
+      "step": 107730
+    },
+    {
+      "epoch": 0.5380408998976254,
+      "grad_norm": 0.08857207745313644,
+      "learning_rate": 1.3893464166812686e-05,
+      "loss": 8.842,
+      "step": 107740
+    },
+    {
+      "epoch": 0.5380908387225649,
+      "grad_norm": 0.10182986408472061,
+      "learning_rate": 1.3891962251871137e-05,
+      "loss": 8.8435,
+      "step": 107750
+    },
+    {
+      "epoch": 0.5381407775475043,
+      "grad_norm": 0.09429343789815903,
+      "learning_rate": 1.3890460336929587e-05,
+      "loss": 8.8431,
+      "step": 107760
+    },
+    {
+      "epoch": 0.5381907163724438,
+      "grad_norm": 0.09139162302017212,
+      "learning_rate": 1.3888958421988034e-05,
+      "loss": 8.8568,
+      "step": 107770
+    },
+    {
+      "epoch": 0.5382406551973832,
+      "grad_norm": 0.08870648592710495,
+      "learning_rate": 1.3887456507046484e-05,
+      "loss": 8.8555,
+      "step": 107780
+    },
+    {
+      "epoch": 0.5382905940223226,
+      "grad_norm": 0.09010998159646988,
+      "learning_rate": 1.3885954592104934e-05,
+      "loss": 8.8466,
+      "step": 107790
+    },
+    {
+      "epoch": 0.5383405328472621,
+      "grad_norm": 0.09655839204788208,
+      "learning_rate": 1.3884452677163384e-05,
+      "loss": 8.8326,
+      "step": 107800
+    },
+    {
+      "epoch": 0.5383904716722016,
+      "grad_norm": 0.09276573359966278,
+      "learning_rate": 1.3882950762221834e-05,
+      "loss": 8.8574,
+      "step": 107810
+    },
+    {
+      "epoch": 0.538440410497141,
+      "grad_norm": 0.0921722874045372,
+      "learning_rate": 1.3881448847280281e-05,
+      "loss": 8.8477,
+      "step": 107820
+    },
+    {
+      "epoch": 0.5384903493220804,
+      "grad_norm": 0.0946490541100502,
+      "learning_rate": 1.3879946932338731e-05,
+      "loss": 8.8401,
+      "step": 107830
+    },
+    {
+      "epoch": 0.5385402881470199,
+      "grad_norm": 0.0921718031167984,
+      "learning_rate": 1.3878445017397181e-05,
+      "loss": 8.8461,
+      "step": 107840
+    },
+    {
+      "epoch": 0.5385902269719594,
+      "grad_norm": 0.08847850561141968,
+      "learning_rate": 1.3876943102455632e-05,
+      "loss": 8.8462,
+      "step": 107850
+    },
+    {
+      "epoch": 0.5386401657968988,
+      "grad_norm": 0.09322067350149155,
+      "learning_rate": 1.3875441187514082e-05,
+      "loss": 8.8422,
+      "step": 107860
+    },
+    {
+      "epoch": 0.5386901046218382,
+      "grad_norm": 0.09097786247730255,
+      "learning_rate": 1.387393927257253e-05,
+      "loss": 8.8435,
+      "step": 107870
+    },
+    {
+      "epoch": 0.5387400434467777,
+      "grad_norm": 0.09428925067186356,
+      "learning_rate": 1.3872437357630979e-05,
+      "loss": 8.8407,
+      "step": 107880
+    },
+    {
+      "epoch": 0.5387899822717171,
+      "grad_norm": 0.09316663444042206,
+      "learning_rate": 1.3870935442689429e-05,
+      "loss": 8.8444,
+      "step": 107890
+    },
+    {
+      "epoch": 0.5388399210966566,
+      "grad_norm": 0.08978422731161118,
+      "learning_rate": 1.386943352774788e-05,
+      "loss": 8.839,
+      "step": 107900
+    },
+    {
+      "epoch": 0.538889859921596,
+      "grad_norm": 0.09495079517364502,
+      "learning_rate": 1.386793161280633e-05,
+      "loss": 8.8695,
+      "step": 107910
+    },
+    {
+      "epoch": 0.5389397987465355,
+      "grad_norm": 0.09116814285516739,
+      "learning_rate": 1.3866429697864778e-05,
+      "loss": 8.8255,
+      "step": 107920
+    },
+    {
+      "epoch": 0.538989737571475,
+      "grad_norm": 0.0922946184873581,
+      "learning_rate": 1.3864927782923226e-05,
+      "loss": 8.8413,
+      "step": 107930
+    },
+    {
+      "epoch": 0.5390396763964144,
+      "grad_norm": 0.09681878238916397,
+      "learning_rate": 1.3863425867981676e-05,
+      "loss": 8.8529,
+      "step": 107940
+    },
+    {
+      "epoch": 0.5390896152213538,
+      "grad_norm": 0.0935935452580452,
+      "learning_rate": 1.3861923953040127e-05,
+      "loss": 8.8531,
+      "step": 107950
+    },
+    {
+      "epoch": 0.5391395540462933,
+      "grad_norm": 0.09184682369232178,
+      "learning_rate": 1.3860422038098577e-05,
+      "loss": 8.8414,
+      "step": 107960
+    },
+    {
+      "epoch": 0.5391894928712327,
+      "grad_norm": 0.10028470307588577,
+      "learning_rate": 1.3858920123157025e-05,
+      "loss": 8.8435,
+      "step": 107970
+    },
+    {
+      "epoch": 0.5392394316961722,
+      "grad_norm": 0.09753447771072388,
+      "learning_rate": 1.3857418208215474e-05,
+      "loss": 8.8459,
+      "step": 107980
+    },
+    {
+      "epoch": 0.5392893705211116,
+      "grad_norm": 0.10108836740255356,
+      "learning_rate": 1.3855916293273924e-05,
+      "loss": 8.8328,
+      "step": 107990
+    },
+    {
+      "epoch": 0.5393393093460511,
+      "grad_norm": 0.08985279500484467,
+      "learning_rate": 1.3854414378332374e-05,
+      "loss": 8.8288,
+      "step": 108000
+    },
+    {
+      "epoch": 0.5393892481709905,
+      "grad_norm": 0.08839620649814606,
+      "learning_rate": 1.3852912463390824e-05,
+      "loss": 8.8446,
+      "step": 108010
+    },
+    {
+      "epoch": 0.53943918699593,
+      "grad_norm": 0.09369973093271255,
+      "learning_rate": 1.3851410548449273e-05,
+      "loss": 8.8246,
+      "step": 108020
+    },
+    {
+      "epoch": 0.5394891258208694,
+      "grad_norm": 0.09340804815292358,
+      "learning_rate": 1.3849908633507723e-05,
+      "loss": 8.8316,
+      "step": 108030
+    },
+    {
+      "epoch": 0.5395390646458089,
+      "grad_norm": 0.09237315505743027,
+      "learning_rate": 1.3848406718566171e-05,
+      "loss": 8.8389,
+      "step": 108040
+    },
+    {
+      "epoch": 0.5395890034707483,
+      "grad_norm": 0.09002888947725296,
+      "learning_rate": 1.3846904803624622e-05,
+      "loss": 8.8388,
+      "step": 108050
+    },
+    {
+      "epoch": 0.5396389422956878,
+      "grad_norm": 0.0946909487247467,
+      "learning_rate": 1.3845402888683072e-05,
+      "loss": 8.8354,
+      "step": 108060
+    },
+    {
+      "epoch": 0.5396888811206272,
+      "grad_norm": 0.09334079921245575,
+      "learning_rate": 1.3843900973741522e-05,
+      "loss": 8.8316,
+      "step": 108070
+    },
+    {
+      "epoch": 0.5397388199455667,
+      "grad_norm": 0.09467971324920654,
+      "learning_rate": 1.384239905879997e-05,
+      "loss": 8.8425,
+      "step": 108080
+    },
+    {
+      "epoch": 0.5397887587705061,
+      "grad_norm": 0.09707115590572357,
+      "learning_rate": 1.3840897143858419e-05,
+      "loss": 8.8327,
+      "step": 108090
+    },
+    {
+      "epoch": 0.5398386975954456,
+      "grad_norm": 0.09300867468118668,
+      "learning_rate": 1.383939522891687e-05,
+      "loss": 8.8463,
+      "step": 108100
+    },
+    {
+      "epoch": 0.539888636420385,
+      "grad_norm": 0.09528551995754242,
+      "learning_rate": 1.383789331397532e-05,
+      "loss": 8.8379,
+      "step": 108110
+    },
+    {
+      "epoch": 0.5399385752453245,
+      "grad_norm": 0.09320961683988571,
+      "learning_rate": 1.383639139903377e-05,
+      "loss": 8.8346,
+      "step": 108120
+    },
+    {
+      "epoch": 0.5399885140702639,
+      "grad_norm": 0.09055766463279724,
+      "learning_rate": 1.3834889484092218e-05,
+      "loss": 8.8463,
+      "step": 108130
+    },
+    {
+      "epoch": 0.5400384528952034,
+      "grad_norm": 0.09693289548158646,
+      "learning_rate": 1.3833387569150667e-05,
+      "loss": 8.8457,
+      "step": 108140
+    },
+    {
+      "epoch": 0.5400883917201428,
+      "grad_norm": 0.10044632852077484,
+      "learning_rate": 1.3831885654209117e-05,
+      "loss": 8.8156,
+      "step": 108150
+    },
+    {
+      "epoch": 0.5401383305450823,
+      "grad_norm": 0.09970605373382568,
+      "learning_rate": 1.3830383739267567e-05,
+      "loss": 8.8469,
+      "step": 108160
+    },
+    {
+      "epoch": 0.5401882693700217,
+      "grad_norm": 0.09658732265233994,
+      "learning_rate": 1.3828881824326017e-05,
+      "loss": 8.8184,
+      "step": 108170
+    },
+    {
+      "epoch": 0.5402382081949612,
+      "grad_norm": 0.0948546975851059,
+      "learning_rate": 1.3827379909384466e-05,
+      "loss": 8.8466,
+      "step": 108180
+    },
+    {
+      "epoch": 0.5402881470199006,
+      "grad_norm": 0.09518400579690933,
+      "learning_rate": 1.3825877994442916e-05,
+      "loss": 8.8373,
+      "step": 108190
+    },
+    {
+      "epoch": 0.5403380858448401,
+      "grad_norm": 0.09109831601381302,
+      "learning_rate": 1.3824376079501364e-05,
+      "loss": 8.8272,
+      "step": 108200
+    },
+    {
+      "epoch": 0.5403880246697795,
+      "grad_norm": 0.08981281518936157,
+      "learning_rate": 1.3822874164559814e-05,
+      "loss": 8.8456,
+      "step": 108210
+    },
+    {
+      "epoch": 0.540437963494719,
+      "grad_norm": 0.10105878114700317,
+      "learning_rate": 1.3821372249618265e-05,
+      "loss": 8.8423,
+      "step": 108220
+    },
+    {
+      "epoch": 0.5404879023196584,
+      "grad_norm": 0.08803227543830872,
+      "learning_rate": 1.3819870334676713e-05,
+      "loss": 8.8416,
+      "step": 108230
+    },
+    {
+      "epoch": 0.5405378411445979,
+      "grad_norm": 0.09488827735185623,
+      "learning_rate": 1.3818368419735163e-05,
+      "loss": 8.8317,
+      "step": 108240
+    },
+    {
+      "epoch": 0.5405877799695373,
+      "grad_norm": 0.0969134047627449,
+      "learning_rate": 1.3816866504793612e-05,
+      "loss": 8.837,
+      "step": 108250
+    },
+    {
+      "epoch": 0.5406377187944768,
+      "grad_norm": 0.09399174153804779,
+      "learning_rate": 1.3815364589852062e-05,
+      "loss": 8.8314,
+      "step": 108260
+    },
+    {
+      "epoch": 0.5406876576194162,
+      "grad_norm": 0.08893413096666336,
+      "learning_rate": 1.3813862674910512e-05,
+      "loss": 8.8467,
+      "step": 108270
+    },
+    {
+      "epoch": 0.5407375964443557,
+      "grad_norm": 0.09072422981262207,
+      "learning_rate": 1.381236075996896e-05,
+      "loss": 8.8399,
+      "step": 108280
+    },
+    {
+      "epoch": 0.5407875352692951,
+      "grad_norm": 0.09168332815170288,
+      "learning_rate": 1.381085884502741e-05,
+      "loss": 8.8355,
+      "step": 108290
+    },
+    {
+      "epoch": 0.5408374740942345,
+      "grad_norm": 0.09577570855617523,
+      "learning_rate": 1.380935693008586e-05,
+      "loss": 8.8358,
+      "step": 108300
+    },
+    {
+      "epoch": 0.540887412919174,
+      "grad_norm": 0.09399871528148651,
+      "learning_rate": 1.380785501514431e-05,
+      "loss": 8.8198,
+      "step": 108310
+    },
+    {
+      "epoch": 0.5409373517441135,
+      "grad_norm": 0.09029237180948257,
+      "learning_rate": 1.380635310020276e-05,
+      "loss": 8.8317,
+      "step": 108320
+    },
+    {
+      "epoch": 0.5409872905690529,
+      "grad_norm": 0.09074974060058594,
+      "learning_rate": 1.3804851185261208e-05,
+      "loss": 8.8355,
+      "step": 108330
+    },
+    {
+      "epoch": 0.5410372293939923,
+      "grad_norm": 0.08936341851949692,
+      "learning_rate": 1.3803349270319658e-05,
+      "loss": 8.8351,
+      "step": 108340
+    },
+    {
+      "epoch": 0.5410871682189318,
+      "grad_norm": 0.08986354619264603,
+      "learning_rate": 1.3801847355378108e-05,
+      "loss": 8.8396,
+      "step": 108350
+    },
+    {
+      "epoch": 0.5411371070438713,
+      "grad_norm": 0.0931684747338295,
+      "learning_rate": 1.3800345440436557e-05,
+      "loss": 8.8502,
+      "step": 108360
+    },
+    {
+      "epoch": 0.5411870458688107,
+      "grad_norm": 0.09337104111909866,
+      "learning_rate": 1.3798843525495007e-05,
+      "loss": 8.841,
+      "step": 108370
+    },
+    {
+      "epoch": 0.5412369846937501,
+      "grad_norm": 0.09039416164159775,
+      "learning_rate": 1.3797341610553456e-05,
+      "loss": 8.8413,
+      "step": 108380
+    },
+    {
+      "epoch": 0.5412869235186896,
+      "grad_norm": 0.09265271574258804,
+      "learning_rate": 1.3795839695611906e-05,
+      "loss": 8.8387,
+      "step": 108390
+    },
+    {
+      "epoch": 0.5413368623436291,
+      "grad_norm": 0.09323190897703171,
+      "learning_rate": 1.3794337780670356e-05,
+      "loss": 8.8338,
+      "step": 108400
+    },
+    {
+      "epoch": 0.5413868011685685,
+      "grad_norm": 0.0942569449543953,
+      "learning_rate": 1.3792835865728804e-05,
+      "loss": 8.829,
+      "step": 108410
+    },
+    {
+      "epoch": 0.5414367399935079,
+      "grad_norm": 0.09136468917131424,
+      "learning_rate": 1.3791333950787255e-05,
+      "loss": 8.8375,
+      "step": 108420
+    },
+    {
+      "epoch": 0.5414866788184474,
+      "grad_norm": 0.09518197923898697,
+      "learning_rate": 1.3789832035845703e-05,
+      "loss": 8.8347,
+      "step": 108430
+    },
+    {
+      "epoch": 0.5415366176433869,
+      "grad_norm": 0.09513925015926361,
+      "learning_rate": 1.3788330120904153e-05,
+      "loss": 8.8337,
+      "step": 108440
+    },
+    {
+      "epoch": 0.5415865564683263,
+      "grad_norm": 0.09242982417345047,
+      "learning_rate": 1.3786828205962603e-05,
+      "loss": 8.8264,
+      "step": 108450
+    },
+    {
+      "epoch": 0.5416364952932657,
+      "grad_norm": 0.09404724091291428,
+      "learning_rate": 1.3785326291021052e-05,
+      "loss": 8.8328,
+      "step": 108460
+    },
+    {
+      "epoch": 0.5416864341182052,
+      "grad_norm": 0.09275107085704803,
+      "learning_rate": 1.3783824376079502e-05,
+      "loss": 8.8397,
+      "step": 108470
+    },
+    {
+      "epoch": 0.5417363729431447,
+      "grad_norm": 0.09128131717443466,
+      "learning_rate": 1.378232246113795e-05,
+      "loss": 8.8265,
+      "step": 108480
+    },
+    {
+      "epoch": 0.5417863117680841,
+      "grad_norm": 0.09248005598783493,
+      "learning_rate": 1.37808205461964e-05,
+      "loss": 8.8399,
+      "step": 108490
+    },
+    {
+      "epoch": 0.5418362505930235,
+      "grad_norm": 0.09171829372644424,
+      "learning_rate": 1.3779318631254851e-05,
+      "loss": 8.8332,
+      "step": 108500
+    },
+    {
+      "epoch": 0.541886189417963,
+      "grad_norm": 0.09409584105014801,
+      "learning_rate": 1.37778167163133e-05,
+      "loss": 8.8347,
+      "step": 108510
+    },
+    {
+      "epoch": 0.5419361282429025,
+      "grad_norm": 0.09212516993284225,
+      "learning_rate": 1.377631480137175e-05,
+      "loss": 8.839,
+      "step": 108520
+    },
+    {
+      "epoch": 0.5419860670678419,
+      "grad_norm": 0.0963674783706665,
+      "learning_rate": 1.3774812886430198e-05,
+      "loss": 8.8263,
+      "step": 108530
+    },
+    {
+      "epoch": 0.5420360058927813,
+      "grad_norm": 0.09333939105272293,
+      "learning_rate": 1.3773310971488648e-05,
+      "loss": 8.8339,
+      "step": 108540
+    },
+    {
+      "epoch": 0.5420859447177208,
+      "grad_norm": 0.0930102989077568,
+      "learning_rate": 1.3771809056547098e-05,
+      "loss": 8.8331,
+      "step": 108550
+    },
+    {
+      "epoch": 0.5421358835426603,
+      "grad_norm": 0.09327764064073563,
+      "learning_rate": 1.3770307141605549e-05,
+      "loss": 8.8169,
+      "step": 108560
+    },
+    {
+      "epoch": 0.5421858223675997,
+      "grad_norm": 0.09788885712623596,
+      "learning_rate": 1.3768805226663997e-05,
+      "loss": 8.8428,
+      "step": 108570
+    },
+    {
+      "epoch": 0.5422357611925391,
+      "grad_norm": 0.08719142526388168,
+      "learning_rate": 1.3767303311722446e-05,
+      "loss": 8.835,
+      "step": 108580
+    },
+    {
+      "epoch": 0.5422857000174786,
+      "grad_norm": 0.08899932354688644,
+      "learning_rate": 1.3765801396780896e-05,
+      "loss": 8.8379,
+      "step": 108590
+    },
+    {
+      "epoch": 0.5423356388424181,
+      "grad_norm": 0.09376303106546402,
+      "learning_rate": 1.3764299481839346e-05,
+      "loss": 8.8337,
+      "step": 108600
+    },
+    {
+      "epoch": 0.5423855776673575,
+      "grad_norm": 0.09086206555366516,
+      "learning_rate": 1.3762797566897796e-05,
+      "loss": 8.8522,
+      "step": 108610
+    },
+    {
+      "epoch": 0.5424355164922969,
+      "grad_norm": 0.09616357088088989,
+      "learning_rate": 1.3761295651956245e-05,
+      "loss": 8.8464,
+      "step": 108620
+    },
+    {
+      "epoch": 0.5424854553172364,
+      "grad_norm": 0.09458029270172119,
+      "learning_rate": 1.3759793737014693e-05,
+      "loss": 8.8334,
+      "step": 108630
+    },
+    {
+      "epoch": 0.5425353941421759,
+      "grad_norm": 0.0973440557718277,
+      "learning_rate": 1.3758291822073143e-05,
+      "loss": 8.8272,
+      "step": 108640
+    },
+    {
+      "epoch": 0.5425853329671153,
+      "grad_norm": 0.09475944191217422,
+      "learning_rate": 1.3756789907131593e-05,
+      "loss": 8.828,
+      "step": 108650
+    },
+    {
+      "epoch": 0.5426352717920547,
+      "grad_norm": 0.09611568599939346,
+      "learning_rate": 1.3755287992190044e-05,
+      "loss": 8.8075,
+      "step": 108660
+    },
+    {
+      "epoch": 0.5426852106169942,
+      "grad_norm": 0.09558840841054916,
+      "learning_rate": 1.3753786077248492e-05,
+      "loss": 8.8357,
+      "step": 108670
+    },
+    {
+      "epoch": 0.5427351494419337,
+      "grad_norm": 0.09221713989973068,
+      "learning_rate": 1.375228416230694e-05,
+      "loss": 8.8487,
+      "step": 108680
+    },
+    {
+      "epoch": 0.5427850882668731,
+      "grad_norm": 0.09657151997089386,
+      "learning_rate": 1.375078224736539e-05,
+      "loss": 8.8293,
+      "step": 108690
+    },
+    {
+      "epoch": 0.5428350270918125,
+      "grad_norm": 0.09163092821836472,
+      "learning_rate": 1.3749280332423841e-05,
+      "loss": 8.8348,
+      "step": 108700
+    },
+    {
+      "epoch": 0.542884965916752,
+      "grad_norm": 0.09311132878065109,
+      "learning_rate": 1.3747778417482291e-05,
+      "loss": 8.837,
+      "step": 108710
+    },
+    {
+      "epoch": 0.5429349047416915,
+      "grad_norm": 0.08675689995288849,
+      "learning_rate": 1.3746276502540741e-05,
+      "loss": 8.8421,
+      "step": 108720
+    },
+    {
+      "epoch": 0.5429848435666309,
+      "grad_norm": 0.08896356076002121,
+      "learning_rate": 1.3744774587599188e-05,
+      "loss": 8.8465,
+      "step": 108730
+    },
+    {
+      "epoch": 0.5430347823915703,
+      "grad_norm": 0.09236080199480057,
+      "learning_rate": 1.3743272672657638e-05,
+      "loss": 8.8397,
+      "step": 108740
+    },
+    {
+      "epoch": 0.5430847212165097,
+      "grad_norm": 0.09245762974023819,
+      "learning_rate": 1.3741770757716088e-05,
+      "loss": 8.835,
+      "step": 108750
+    },
+    {
+      "epoch": 0.5431346600414493,
+      "grad_norm": 0.09627767652273178,
+      "learning_rate": 1.3740268842774539e-05,
+      "loss": 8.8153,
+      "step": 108760
+    },
+    {
+      "epoch": 0.5431845988663887,
+      "grad_norm": 0.09583393484354019,
+      "learning_rate": 1.3738766927832989e-05,
+      "loss": 8.8408,
+      "step": 108770
+    },
+    {
+      "epoch": 0.5432345376913281,
+      "grad_norm": 0.10037499666213989,
+      "learning_rate": 1.3737265012891436e-05,
+      "loss": 8.8385,
+      "step": 108780
+    },
+    {
+      "epoch": 0.5432844765162675,
+      "grad_norm": 0.08847088366746902,
+      "learning_rate": 1.3735763097949886e-05,
+      "loss": 8.8244,
+      "step": 108790
+    },
+    {
+      "epoch": 0.543334415341207,
+      "grad_norm": 0.0916975736618042,
+      "learning_rate": 1.3734261183008336e-05,
+      "loss": 8.8241,
+      "step": 108800
+    },
+    {
+      "epoch": 0.5433843541661465,
+      "grad_norm": 0.08768840879201889,
+      "learning_rate": 1.3732759268066786e-05,
+      "loss": 8.84,
+      "step": 108810
+    },
+    {
+      "epoch": 0.5434342929910859,
+      "grad_norm": 0.09283578395843506,
+      "learning_rate": 1.3731257353125236e-05,
+      "loss": 8.8298,
+      "step": 108820
+    },
+    {
+      "epoch": 0.5434842318160253,
+      "grad_norm": 0.09757833927869797,
+      "learning_rate": 1.3729755438183683e-05,
+      "loss": 8.8251,
+      "step": 108830
+    },
+    {
+      "epoch": 0.5435341706409648,
+      "grad_norm": 0.10008548200130463,
+      "learning_rate": 1.3728253523242133e-05,
+      "loss": 8.8276,
+      "step": 108840
+    },
+    {
+      "epoch": 0.5435841094659043,
+      "grad_norm": 0.09164326637983322,
+      "learning_rate": 1.3726751608300583e-05,
+      "loss": 8.8318,
+      "step": 108850
+    },
+    {
+      "epoch": 0.5436340482908437,
+      "grad_norm": 0.09030098468065262,
+      "learning_rate": 1.3725249693359034e-05,
+      "loss": 8.8277,
+      "step": 108860
+    },
+    {
+      "epoch": 0.5436839871157831,
+      "grad_norm": 0.08942774683237076,
+      "learning_rate": 1.3723747778417484e-05,
+      "loss": 8.8314,
+      "step": 108870
+    },
+    {
+      "epoch": 0.5437339259407226,
+      "grad_norm": 0.10168751329183578,
+      "learning_rate": 1.3722245863475932e-05,
+      "loss": 8.8215,
+      "step": 108880
+    },
+    {
+      "epoch": 0.5437838647656621,
+      "grad_norm": 0.09753987193107605,
+      "learning_rate": 1.372074394853438e-05,
+      "loss": 8.8198,
+      "step": 108890
+    },
+    {
+      "epoch": 0.5438338035906015,
+      "grad_norm": 0.10058535635471344,
+      "learning_rate": 1.3719242033592831e-05,
+      "loss": 8.8108,
+      "step": 108900
+    },
+    {
+      "epoch": 0.5438837424155409,
+      "grad_norm": 0.09578339755535126,
+      "learning_rate": 1.3717740118651281e-05,
+      "loss": 8.8362,
+      "step": 108910
+    },
+    {
+      "epoch": 0.5439336812404804,
+      "grad_norm": 0.08314099907875061,
+      "learning_rate": 1.3716238203709731e-05,
+      "loss": 8.8229,
+      "step": 108920
+    },
+    {
+      "epoch": 0.5439836200654199,
+      "grad_norm": 0.08905912190675735,
+      "learning_rate": 1.371473628876818e-05,
+      "loss": 8.8382,
+      "step": 108930
+    },
+    {
+      "epoch": 0.5440335588903593,
+      "grad_norm": 0.09270746260881424,
+      "learning_rate": 1.3713234373826628e-05,
+      "loss": 8.8224,
+      "step": 108940
+    },
+    {
+      "epoch": 0.5440834977152987,
+      "grad_norm": 0.09719360619783401,
+      "learning_rate": 1.3711732458885078e-05,
+      "loss": 8.8304,
+      "step": 108950
+    },
+    {
+      "epoch": 0.5441334365402382,
+      "grad_norm": 0.08986958116292953,
+      "learning_rate": 1.3710230543943529e-05,
+      "loss": 8.8256,
+      "step": 108960
+    },
+    {
+      "epoch": 0.5441833753651777,
+      "grad_norm": 0.09314291179180145,
+      "learning_rate": 1.3708728629001979e-05,
+      "loss": 8.8262,
+      "step": 108970
+    },
+    {
+      "epoch": 0.5442333141901171,
+      "grad_norm": 0.09784272313117981,
+      "learning_rate": 1.3707226714060427e-05,
+      "loss": 8.8172,
+      "step": 108980
+    },
+    {
+      "epoch": 0.5442832530150565,
+      "grad_norm": 0.09231096506118774,
+      "learning_rate": 1.3705724799118876e-05,
+      "loss": 8.8399,
+      "step": 108990
+    },
+    {
+      "epoch": 0.544333191839996,
+      "grad_norm": 0.09249245375394821,
+      "learning_rate": 1.3704222884177326e-05,
+      "loss": 8.8255,
+      "step": 109000
+    },
+    {
+      "epoch": 0.5443831306649355,
+      "grad_norm": 0.09249752014875412,
+      "learning_rate": 1.3702720969235776e-05,
+      "loss": 8.8309,
+      "step": 109010
+    },
+    {
+      "epoch": 0.5444330694898749,
+      "grad_norm": 0.09660474956035614,
+      "learning_rate": 1.3701219054294226e-05,
+      "loss": 8.8297,
+      "step": 109020
+    },
+    {
+      "epoch": 0.5444830083148143,
+      "grad_norm": 0.09552586078643799,
+      "learning_rate": 1.3699717139352675e-05,
+      "loss": 8.8145,
+      "step": 109030
+    },
+    {
+      "epoch": 0.5445329471397538,
+      "grad_norm": 0.09530143439769745,
+      "learning_rate": 1.3698215224411125e-05,
+      "loss": 8.82,
+      "step": 109040
+    },
+    {
+      "epoch": 0.5445828859646933,
+      "grad_norm": 0.10076169669628143,
+      "learning_rate": 1.3696713309469573e-05,
+      "loss": 8.8208,
+      "step": 109050
+    },
+    {
+      "epoch": 0.5446328247896327,
+      "grad_norm": 0.09382011741399765,
+      "learning_rate": 1.3695211394528024e-05,
+      "loss": 8.8229,
+      "step": 109060
+    },
+    {
+      "epoch": 0.5446827636145721,
+      "grad_norm": 0.09154395759105682,
+      "learning_rate": 1.3693709479586474e-05,
+      "loss": 8.833,
+      "step": 109070
+    },
+    {
+      "epoch": 0.5447327024395116,
+      "grad_norm": 0.0922236368060112,
+      "learning_rate": 1.3692207564644922e-05,
+      "loss": 8.8162,
+      "step": 109080
+    },
+    {
+      "epoch": 0.5447826412644511,
+      "grad_norm": 0.09294591844081879,
+      "learning_rate": 1.3690705649703372e-05,
+      "loss": 8.8393,
+      "step": 109090
+    },
+    {
+      "epoch": 0.5448325800893905,
+      "grad_norm": 0.09426495432853699,
+      "learning_rate": 1.3689203734761821e-05,
+      "loss": 8.8455,
+      "step": 109100
+    },
+    {
+      "epoch": 0.5448825189143299,
+      "grad_norm": 0.09277299046516418,
+      "learning_rate": 1.3687701819820271e-05,
+      "loss": 8.8368,
+      "step": 109110
+    },
+    {
+      "epoch": 0.5449324577392693,
+      "grad_norm": 0.09092520922422409,
+      "learning_rate": 1.3686199904878721e-05,
+      "loss": 8.8391,
+      "step": 109120
+    },
+    {
+      "epoch": 0.5449823965642089,
+      "grad_norm": 0.0939490795135498,
+      "learning_rate": 1.368469798993717e-05,
+      "loss": 8.832,
+      "step": 109130
+    },
+    {
+      "epoch": 0.5450323353891483,
+      "grad_norm": 0.09599693864583969,
+      "learning_rate": 1.368319607499562e-05,
+      "loss": 8.8155,
+      "step": 109140
+    },
+    {
+      "epoch": 0.5450822742140877,
+      "grad_norm": 0.10206658393144608,
+      "learning_rate": 1.3681694160054068e-05,
+      "loss": 8.8281,
+      "step": 109150
+    },
+    {
+      "epoch": 0.5451322130390271,
+      "grad_norm": 0.10047373175621033,
+      "learning_rate": 1.3680192245112519e-05,
+      "loss": 8.8277,
+      "step": 109160
+    },
+    {
+      "epoch": 0.5451821518639667,
+      "grad_norm": 0.09348422288894653,
+      "learning_rate": 1.3678690330170969e-05,
+      "loss": 8.8309,
+      "step": 109170
+    },
+    {
+      "epoch": 0.5452320906889061,
+      "grad_norm": 0.09665758162736893,
+      "learning_rate": 1.3677188415229417e-05,
+      "loss": 8.8489,
+      "step": 109180
+    },
+    {
+      "epoch": 0.5452820295138455,
+      "grad_norm": 0.09331992268562317,
+      "learning_rate": 1.3675686500287867e-05,
+      "loss": 8.8213,
+      "step": 109190
+    },
+    {
+      "epoch": 0.545331968338785,
+      "grad_norm": 0.08620268851518631,
+      "learning_rate": 1.3674184585346318e-05,
+      "loss": 8.8371,
+      "step": 109200
+    },
+    {
+      "epoch": 0.5453819071637245,
+      "grad_norm": 0.09555631130933762,
+      "learning_rate": 1.3672682670404766e-05,
+      "loss": 8.819,
+      "step": 109210
+    },
+    {
+      "epoch": 0.5454318459886639,
+      "grad_norm": 0.09607243537902832,
+      "learning_rate": 1.3671180755463216e-05,
+      "loss": 8.8082,
+      "step": 109220
+    },
+    {
+      "epoch": 0.5454817848136033,
+      "grad_norm": 0.09805332869291306,
+      "learning_rate": 1.3669678840521665e-05,
+      "loss": 8.8269,
+      "step": 109230
+    },
+    {
+      "epoch": 0.5455317236385427,
+      "grad_norm": 0.09287256002426147,
+      "learning_rate": 1.3668176925580115e-05,
+      "loss": 8.8131,
+      "step": 109240
+    },
+    {
+      "epoch": 0.5455816624634823,
+      "grad_norm": 0.08924273401498795,
+      "learning_rate": 1.3666675010638565e-05,
+      "loss": 8.8197,
+      "step": 109250
+    },
+    {
+      "epoch": 0.5456316012884217,
+      "grad_norm": 0.09110093861818314,
+      "learning_rate": 1.3665173095697014e-05,
+      "loss": 8.8382,
+      "step": 109260
+    },
+    {
+      "epoch": 0.5456815401133611,
+      "grad_norm": 0.09054429084062576,
+      "learning_rate": 1.3663671180755464e-05,
+      "loss": 8.8259,
+      "step": 109270
+    },
+    {
+      "epoch": 0.5457314789383005,
+      "grad_norm": 0.09228412806987762,
+      "learning_rate": 1.3662169265813912e-05,
+      "loss": 8.8154,
+      "step": 109280
+    },
+    {
+      "epoch": 0.5457814177632401,
+      "grad_norm": 0.09101639688014984,
+      "learning_rate": 1.3660667350872362e-05,
+      "loss": 8.8172,
+      "step": 109290
+    },
+    {
+      "epoch": 0.5458313565881795,
+      "grad_norm": 0.09964627027511597,
+      "learning_rate": 1.3659165435930813e-05,
+      "loss": 8.8302,
+      "step": 109300
+    },
+    {
+      "epoch": 0.5458812954131189,
+      "grad_norm": 0.0952049046754837,
+      "learning_rate": 1.3657663520989261e-05,
+      "loss": 8.8477,
+      "step": 109310
+    },
+    {
+      "epoch": 0.5459312342380583,
+      "grad_norm": 0.094510518014431,
+      "learning_rate": 1.3656161606047711e-05,
+      "loss": 8.8274,
+      "step": 109320
+    },
+    {
+      "epoch": 0.5459811730629979,
+      "grad_norm": 0.09812169522047043,
+      "learning_rate": 1.365465969110616e-05,
+      "loss": 8.8186,
+      "step": 109330
+    },
+    {
+      "epoch": 0.5460311118879373,
+      "grad_norm": 0.09070379287004471,
+      "learning_rate": 1.365315777616461e-05,
+      "loss": 8.8214,
+      "step": 109340
+    },
+    {
+      "epoch": 0.5460810507128767,
+      "grad_norm": 0.09096910059452057,
+      "learning_rate": 1.365165586122306e-05,
+      "loss": 8.8309,
+      "step": 109350
+    },
+    {
+      "epoch": 0.5461309895378161,
+      "grad_norm": 0.0929383784532547,
+      "learning_rate": 1.365015394628151e-05,
+      "loss": 8.8311,
+      "step": 109360
+    },
+    {
+      "epoch": 0.5461809283627557,
+      "grad_norm": 0.09064075350761414,
+      "learning_rate": 1.3648652031339959e-05,
+      "loss": 8.8385,
+      "step": 109370
+    },
+    {
+      "epoch": 0.5462308671876951,
+      "grad_norm": 0.09253614395856857,
+      "learning_rate": 1.3647150116398407e-05,
+      "loss": 8.8222,
+      "step": 109380
+    },
+    {
+      "epoch": 0.5462808060126345,
+      "grad_norm": 0.09299398958683014,
+      "learning_rate": 1.3645648201456857e-05,
+      "loss": 8.8109,
+      "step": 109390
+    },
+    {
+      "epoch": 0.5463307448375739,
+      "grad_norm": 0.09167538583278656,
+      "learning_rate": 1.3644146286515308e-05,
+      "loss": 8.8226,
+      "step": 109400
+    },
+    {
+      "epoch": 0.5463806836625135,
+      "grad_norm": 0.09108640998601913,
+      "learning_rate": 1.3642644371573758e-05,
+      "loss": 8.823,
+      "step": 109410
+    },
+    {
+      "epoch": 0.5464306224874529,
+      "grad_norm": 0.09583522379398346,
+      "learning_rate": 1.3641142456632206e-05,
+      "loss": 8.83,
+      "step": 109420
+    },
+    {
+      "epoch": 0.5464805613123923,
+      "grad_norm": 0.08716997504234314,
+      "learning_rate": 1.3639640541690655e-05,
+      "loss": 8.8318,
+      "step": 109430
+    },
+    {
+      "epoch": 0.5465305001373317,
+      "grad_norm": 0.08698141574859619,
+      "learning_rate": 1.3638138626749105e-05,
+      "loss": 8.8372,
+      "step": 109440
+    },
+    {
+      "epoch": 0.5465804389622713,
+      "grad_norm": 0.08734197914600372,
+      "learning_rate": 1.3636636711807555e-05,
+      "loss": 8.8252,
+      "step": 109450
+    },
+    {
+      "epoch": 0.5466303777872107,
+      "grad_norm": 0.08939076960086823,
+      "learning_rate": 1.3635134796866005e-05,
+      "loss": 8.8243,
+      "step": 109460
+    },
+    {
+      "epoch": 0.5466803166121501,
+      "grad_norm": 0.09037525951862335,
+      "learning_rate": 1.3633632881924454e-05,
+      "loss": 8.8242,
+      "step": 109470
+    },
+    {
+      "epoch": 0.5467302554370895,
+      "grad_norm": 0.09184879809617996,
+      "learning_rate": 1.3632130966982902e-05,
+      "loss": 8.8219,
+      "step": 109480
+    },
+    {
+      "epoch": 0.5467801942620291,
+      "grad_norm": 0.09089411050081253,
+      "learning_rate": 1.3630629052041352e-05,
+      "loss": 8.8242,
+      "step": 109490
+    },
+    {
+      "epoch": 0.5468301330869685,
+      "grad_norm": 0.09041132777929306,
+      "learning_rate": 1.3629127137099803e-05,
+      "loss": 8.8208,
+      "step": 109500
+    },
+    {
+      "epoch": 0.5468800719119079,
+      "grad_norm": 0.08768140524625778,
+      "learning_rate": 1.3627625222158253e-05,
+      "loss": 8.8216,
+      "step": 109510
+    },
+    {
+      "epoch": 0.5469300107368473,
+      "grad_norm": 0.08897078037261963,
+      "learning_rate": 1.3626123307216703e-05,
+      "loss": 8.8137,
+      "step": 109520
+    },
+    {
+      "epoch": 0.5469799495617869,
+      "grad_norm": 0.0939832553267479,
+      "learning_rate": 1.362462139227515e-05,
+      "loss": 8.8204,
+      "step": 109530
+    },
+    {
+      "epoch": 0.5470298883867263,
+      "grad_norm": 0.09440961480140686,
+      "learning_rate": 1.36231194773336e-05,
+      "loss": 8.8241,
+      "step": 109540
+    },
+    {
+      "epoch": 0.5470798272116657,
+      "grad_norm": 0.08818075060844421,
+      "learning_rate": 1.362161756239205e-05,
+      "loss": 8.8393,
+      "step": 109550
+    },
+    {
+      "epoch": 0.5471297660366051,
+      "grad_norm": 0.09029530733823776,
+      "learning_rate": 1.36201156474505e-05,
+      "loss": 8.8109,
+      "step": 109560
+    },
+    {
+      "epoch": 0.5471797048615447,
+      "grad_norm": 0.09235231578350067,
+      "learning_rate": 1.361861373250895e-05,
+      "loss": 8.8183,
+      "step": 109570
+    },
+    {
+      "epoch": 0.5472296436864841,
+      "grad_norm": 0.09568846970796585,
+      "learning_rate": 1.3617111817567397e-05,
+      "loss": 8.8189,
+      "step": 109580
+    },
+    {
+      "epoch": 0.5472795825114235,
+      "grad_norm": 0.09308406710624695,
+      "learning_rate": 1.3615609902625847e-05,
+      "loss": 8.827,
+      "step": 109590
+    },
+    {
+      "epoch": 0.5473295213363629,
+      "grad_norm": 0.09407919645309448,
+      "learning_rate": 1.3614107987684298e-05,
+      "loss": 8.8238,
+      "step": 109600
+    },
+    {
+      "epoch": 0.5473794601613025,
+      "grad_norm": 0.09674926102161407,
+      "learning_rate": 1.3612606072742748e-05,
+      "loss": 8.8228,
+      "step": 109610
+    },
+    {
+      "epoch": 0.5474293989862419,
+      "grad_norm": 0.09146232157945633,
+      "learning_rate": 1.3611104157801198e-05,
+      "loss": 8.8094,
+      "step": 109620
+    },
+    {
+      "epoch": 0.5474793378111813,
+      "grad_norm": 0.09639618545770645,
+      "learning_rate": 1.3609602242859645e-05,
+      "loss": 8.8284,
+      "step": 109630
+    },
+    {
+      "epoch": 0.5475292766361207,
+      "grad_norm": 0.09209806472063065,
+      "learning_rate": 1.3608100327918095e-05,
+      "loss": 8.8338,
+      "step": 109640
+    },
+    {
+      "epoch": 0.5475792154610603,
+      "grad_norm": 0.09675581008195877,
+      "learning_rate": 1.3606598412976545e-05,
+      "loss": 8.8283,
+      "step": 109650
+    },
+    {
+      "epoch": 0.5476291542859997,
+      "grad_norm": 0.09263081103563309,
+      "learning_rate": 1.3605096498034995e-05,
+      "loss": 8.8038,
+      "step": 109660
+    },
+    {
+      "epoch": 0.5476790931109391,
+      "grad_norm": 0.09105950593948364,
+      "learning_rate": 1.3603594583093446e-05,
+      "loss": 8.8312,
+      "step": 109670
+    },
+    {
+      "epoch": 0.5477290319358785,
+      "grad_norm": 0.09509497880935669,
+      "learning_rate": 1.3602092668151894e-05,
+      "loss": 8.831,
+      "step": 109680
+    },
+    {
+      "epoch": 0.547778970760818,
+      "grad_norm": 0.09219329804182053,
+      "learning_rate": 1.3600590753210343e-05,
+      "loss": 8.8016,
+      "step": 109690
+    },
+    {
+      "epoch": 0.5478289095857575,
+      "grad_norm": 0.09572474658489227,
+      "learning_rate": 1.3599088838268793e-05,
+      "loss": 8.8219,
+      "step": 109700
+    },
+    {
+      "epoch": 0.5478788484106969,
+      "grad_norm": 0.10046906024217606,
+      "learning_rate": 1.3597586923327243e-05,
+      "loss": 8.8214,
+      "step": 109710
+    },
+    {
+      "epoch": 0.5479287872356363,
+      "grad_norm": 0.09304700791835785,
+      "learning_rate": 1.3596085008385693e-05,
+      "loss": 8.8302,
+      "step": 109720
+    },
+    {
+      "epoch": 0.5479787260605758,
+      "grad_norm": 0.09092891216278076,
+      "learning_rate": 1.3594583093444142e-05,
+      "loss": 8.8251,
+      "step": 109730
+    },
+    {
+      "epoch": 0.5480286648855153,
+      "grad_norm": 0.09128131717443466,
+      "learning_rate": 1.359308117850259e-05,
+      "loss": 8.829,
+      "step": 109740
+    },
+    {
+      "epoch": 0.5480786037104547,
+      "grad_norm": 0.08997046947479248,
+      "learning_rate": 1.359157926356104e-05,
+      "loss": 8.8308,
+      "step": 109750
+    },
+    {
+      "epoch": 0.5481285425353941,
+      "grad_norm": 0.09499360620975494,
+      "learning_rate": 1.359007734861949e-05,
+      "loss": 8.8291,
+      "step": 109760
+    },
+    {
+      "epoch": 0.5481784813603335,
+      "grad_norm": 0.09103532880544662,
+      "learning_rate": 1.358857543367794e-05,
+      "loss": 8.8168,
+      "step": 109770
+    },
+    {
+      "epoch": 0.5482284201852731,
+      "grad_norm": 0.08995458483695984,
+      "learning_rate": 1.3587073518736389e-05,
+      "loss": 8.8261,
+      "step": 109780
+    },
+    {
+      "epoch": 0.5482783590102125,
+      "grad_norm": 0.09766585379838943,
+      "learning_rate": 1.3585571603794838e-05,
+      "loss": 8.8222,
+      "step": 109790
+    },
+    {
+      "epoch": 0.5483282978351519,
+      "grad_norm": 0.09343060851097107,
+      "learning_rate": 1.3584069688853288e-05,
+      "loss": 8.817,
+      "step": 109800
+    },
+    {
+      "epoch": 0.5483782366600913,
+      "grad_norm": 0.09289675951004028,
+      "learning_rate": 1.3582567773911738e-05,
+      "loss": 8.8271,
+      "step": 109810
+    },
+    {
+      "epoch": 0.5484281754850309,
+      "grad_norm": 0.0947795957326889,
+      "learning_rate": 1.3581065858970188e-05,
+      "loss": 8.8265,
+      "step": 109820
+    },
+    {
+      "epoch": 0.5484781143099703,
+      "grad_norm": 0.09564196318387985,
+      "learning_rate": 1.3579563944028637e-05,
+      "loss": 8.823,
+      "step": 109830
+    },
+    {
+      "epoch": 0.5485280531349097,
+      "grad_norm": 0.08977336436510086,
+      "learning_rate": 1.3578062029087087e-05,
+      "loss": 8.8181,
+      "step": 109840
+    },
+    {
+      "epoch": 0.5485779919598491,
+      "grad_norm": 0.0958486795425415,
+      "learning_rate": 1.3576560114145535e-05,
+      "loss": 8.8321,
+      "step": 109850
+    },
+    {
+      "epoch": 0.5486279307847887,
+      "grad_norm": 0.09179677069187164,
+      "learning_rate": 1.3575058199203985e-05,
+      "loss": 8.8342,
+      "step": 109860
+    },
+    {
+      "epoch": 0.5486778696097281,
+      "grad_norm": 0.09745611250400543,
+      "learning_rate": 1.3573556284262436e-05,
+      "loss": 8.8287,
+      "step": 109870
+    },
+    {
+      "epoch": 0.5487278084346675,
+      "grad_norm": 0.09767097979784012,
+      "learning_rate": 1.3572054369320884e-05,
+      "loss": 8.8168,
+      "step": 109880
+    },
+    {
+      "epoch": 0.5487777472596069,
+      "grad_norm": 0.0993611216545105,
+      "learning_rate": 1.3570552454379334e-05,
+      "loss": 8.8124,
+      "step": 109890
+    },
+    {
+      "epoch": 0.5488276860845465,
+      "grad_norm": 0.09202627837657928,
+      "learning_rate": 1.3569050539437783e-05,
+      "loss": 8.8306,
+      "step": 109900
+    },
+    {
+      "epoch": 0.5488776249094859,
+      "grad_norm": 0.09640886634588242,
+      "learning_rate": 1.3567548624496233e-05,
+      "loss": 8.8248,
+      "step": 109910
+    },
+    {
+      "epoch": 0.5489275637344253,
+      "grad_norm": 0.09196940064430237,
+      "learning_rate": 1.3566046709554683e-05,
+      "loss": 8.8267,
+      "step": 109920
+    },
+    {
+      "epoch": 0.5489775025593647,
+      "grad_norm": 0.09446310251951218,
+      "learning_rate": 1.3564544794613132e-05,
+      "loss": 8.8257,
+      "step": 109930
+    },
+    {
+      "epoch": 0.5490274413843043,
+      "grad_norm": 0.0930735394358635,
+      "learning_rate": 1.3563042879671582e-05,
+      "loss": 8.8324,
+      "step": 109940
+    },
+    {
+      "epoch": 0.5490773802092437,
+      "grad_norm": 0.0945110097527504,
+      "learning_rate": 1.356154096473003e-05,
+      "loss": 8.8224,
+      "step": 109950
+    },
+    {
+      "epoch": 0.5491273190341831,
+      "grad_norm": 0.09478355199098587,
+      "learning_rate": 1.356003904978848e-05,
+      "loss": 8.8194,
+      "step": 109960
+    },
+    {
+      "epoch": 0.5491772578591225,
+      "grad_norm": 0.09091106802225113,
+      "learning_rate": 1.355853713484693e-05,
+      "loss": 8.8413,
+      "step": 109970
+    },
+    {
+      "epoch": 0.5492271966840621,
+      "grad_norm": 0.0927327498793602,
+      "learning_rate": 1.3557035219905379e-05,
+      "loss": 8.8189,
+      "step": 109980
+    },
+    {
+      "epoch": 0.5492771355090015,
+      "grad_norm": 0.08513586223125458,
+      "learning_rate": 1.355553330496383e-05,
+      "loss": 8.8279,
+      "step": 109990
+    },
+    {
+      "epoch": 0.5493270743339409,
+      "grad_norm": 0.08400266617536545,
+      "learning_rate": 1.355403139002228e-05,
+      "loss": 8.8285,
+      "step": 110000
+    },
+    {
+      "epoch": 0.5493770131588803,
+      "grad_norm": 0.09037226438522339,
+      "learning_rate": 1.3552529475080728e-05,
+      "loss": 8.8035,
+      "step": 110010
+    },
+    {
+      "epoch": 0.5494269519838199,
+      "grad_norm": 0.09207148849964142,
+      "learning_rate": 1.3551027560139178e-05,
+      "loss": 8.8098,
+      "step": 110020
+    },
+    {
+      "epoch": 0.5494768908087593,
+      "grad_norm": 0.09380757808685303,
+      "learning_rate": 1.3549525645197627e-05,
+      "loss": 8.8365,
+      "step": 110030
+    },
+    {
+      "epoch": 0.5495268296336987,
+      "grad_norm": 0.10270896553993225,
+      "learning_rate": 1.3548023730256077e-05,
+      "loss": 8.8368,
+      "step": 110040
+    },
+    {
+      "epoch": 0.5495767684586381,
+      "grad_norm": 0.09271609783172607,
+      "learning_rate": 1.3546521815314527e-05,
+      "loss": 8.8293,
+      "step": 110050
+    },
+    {
+      "epoch": 0.5496267072835777,
+      "grad_norm": 0.0941663458943367,
+      "learning_rate": 1.3545019900372975e-05,
+      "loss": 8.8156,
+      "step": 110060
+    },
+    {
+      "epoch": 0.5496766461085171,
+      "grad_norm": 0.09754271805286407,
+      "learning_rate": 1.3543517985431426e-05,
+      "loss": 8.8233,
+      "step": 110070
+    },
+    {
+      "epoch": 0.5497265849334565,
+      "grad_norm": 0.09003855288028717,
+      "learning_rate": 1.3542016070489874e-05,
+      "loss": 8.8332,
+      "step": 110080
+    },
+    {
+      "epoch": 0.5497765237583959,
+      "grad_norm": 0.08914513885974884,
+      "learning_rate": 1.3540514155548324e-05,
+      "loss": 8.8161,
+      "step": 110090
+    },
+    {
+      "epoch": 0.5498264625833355,
+      "grad_norm": 0.09949744492769241,
+      "learning_rate": 1.3539012240606774e-05,
+      "loss": 8.8143,
+      "step": 110100
+    },
+    {
+      "epoch": 0.5498764014082749,
+      "grad_norm": 0.0968717560172081,
+      "learning_rate": 1.3537510325665223e-05,
+      "loss": 8.8253,
+      "step": 110110
+    },
+    {
+      "epoch": 0.5499263402332143,
+      "grad_norm": 0.09448512643575668,
+      "learning_rate": 1.3536008410723673e-05,
+      "loss": 8.8268,
+      "step": 110120
+    },
+    {
+      "epoch": 0.5499762790581537,
+      "grad_norm": 0.09558204561471939,
+      "learning_rate": 1.3534506495782123e-05,
+      "loss": 8.8139,
+      "step": 110130
+    },
+    {
+      "epoch": 0.5500262178830932,
+      "grad_norm": 0.09001177549362183,
+      "learning_rate": 1.3533004580840572e-05,
+      "loss": 8.8172,
+      "step": 110140
+    },
+    {
+      "epoch": 0.5500761567080327,
+      "grad_norm": 0.09885483235120773,
+      "learning_rate": 1.3531502665899022e-05,
+      "loss": 8.8299,
+      "step": 110150
+    },
+    {
+      "epoch": 0.5501260955329721,
+      "grad_norm": 0.09453386068344116,
+      "learning_rate": 1.3530000750957472e-05,
+      "loss": 8.8231,
+      "step": 110160
+    },
+    {
+      "epoch": 0.5501760343579115,
+      "grad_norm": 0.09200888872146606,
+      "learning_rate": 1.352849883601592e-05,
+      "loss": 8.8221,
+      "step": 110170
+    },
+    {
+      "epoch": 0.550225973182851,
+      "grad_norm": 0.09591957181692123,
+      "learning_rate": 1.352699692107437e-05,
+      "loss": 8.8092,
+      "step": 110180
+    },
+    {
+      "epoch": 0.5502759120077905,
+      "grad_norm": 0.09270251542329788,
+      "learning_rate": 1.352549500613282e-05,
+      "loss": 8.824,
+      "step": 110190
+    },
+    {
+      "epoch": 0.5503258508327299,
+      "grad_norm": 0.0944395586848259,
+      "learning_rate": 1.352399309119127e-05,
+      "loss": 8.8054,
+      "step": 110200
+    },
+    {
+      "epoch": 0.5503757896576693,
+      "grad_norm": 0.09260264784097672,
+      "learning_rate": 1.352249117624972e-05,
+      "loss": 8.8058,
+      "step": 110210
+    },
+    {
+      "epoch": 0.5504257284826088,
+      "grad_norm": 0.09659282863140106,
+      "learning_rate": 1.3520989261308168e-05,
+      "loss": 8.8116,
+      "step": 110220
+    },
+    {
+      "epoch": 0.5504756673075483,
+      "grad_norm": 0.10288282483816147,
+      "learning_rate": 1.3519487346366618e-05,
+      "loss": 8.8174,
+      "step": 110230
+    },
+    {
+      "epoch": 0.5505256061324877,
+      "grad_norm": 0.08334256708621979,
+      "learning_rate": 1.3517985431425067e-05,
+      "loss": 8.8262,
+      "step": 110240
+    },
+    {
+      "epoch": 0.5505755449574271,
+      "grad_norm": 0.0915384441614151,
+      "learning_rate": 1.3516483516483517e-05,
+      "loss": 8.8317,
+      "step": 110250
+    },
+    {
+      "epoch": 0.5506254837823666,
+      "grad_norm": 0.09057852625846863,
+      "learning_rate": 1.3514981601541967e-05,
+      "loss": 8.8207,
+      "step": 110260
+    },
+    {
+      "epoch": 0.5506754226073061,
+      "grad_norm": 0.09614181518554688,
+      "learning_rate": 1.3513479686600416e-05,
+      "loss": 8.8095,
+      "step": 110270
+    },
+    {
+      "epoch": 0.5507253614322455,
+      "grad_norm": 0.09297794103622437,
+      "learning_rate": 1.3511977771658866e-05,
+      "loss": 8.8024,
+      "step": 110280
+    },
+    {
+      "epoch": 0.5507753002571849,
+      "grad_norm": 0.09804685413837433,
+      "learning_rate": 1.3510475856717314e-05,
+      "loss": 8.8088,
+      "step": 110290
+    },
+    {
+      "epoch": 0.5508252390821244,
+      "grad_norm": 0.09318984299898148,
+      "learning_rate": 1.3508973941775764e-05,
+      "loss": 8.8085,
+      "step": 110300
+    },
+    {
+      "epoch": 0.5508751779070639,
+      "grad_norm": 0.09332066029310226,
+      "learning_rate": 1.3507472026834215e-05,
+      "loss": 8.809,
+      "step": 110310
+    },
+    {
+      "epoch": 0.5509251167320033,
+      "grad_norm": 0.0949983298778534,
+      "learning_rate": 1.3505970111892665e-05,
+      "loss": 8.8238,
+      "step": 110320
+    },
+    {
+      "epoch": 0.5509750555569427,
+      "grad_norm": 0.09308378398418427,
+      "learning_rate": 1.3504468196951113e-05,
+      "loss": 8.844,
+      "step": 110330
+    },
+    {
+      "epoch": 0.5510249943818822,
+      "grad_norm": 0.09120608866214752,
+      "learning_rate": 1.3502966282009562e-05,
+      "loss": 8.8099,
+      "step": 110340
+    },
+    {
+      "epoch": 0.5510749332068217,
+      "grad_norm": 0.09323354065418243,
+      "learning_rate": 1.3501464367068012e-05,
+      "loss": 8.8136,
+      "step": 110350
+    },
+    {
+      "epoch": 0.5511248720317611,
+      "grad_norm": 0.0914924144744873,
+      "learning_rate": 1.3499962452126462e-05,
+      "loss": 8.8114,
+      "step": 110360
+    },
+    {
+      "epoch": 0.5511748108567005,
+      "grad_norm": 0.09144311398267746,
+      "learning_rate": 1.3498460537184912e-05,
+      "loss": 8.8016,
+      "step": 110370
+    },
+    {
+      "epoch": 0.55122474968164,
+      "grad_norm": 0.0942022055387497,
+      "learning_rate": 1.349695862224336e-05,
+      "loss": 8.8147,
+      "step": 110380
+    },
+    {
+      "epoch": 0.5512746885065795,
+      "grad_norm": 0.09916293621063232,
+      "learning_rate": 1.349545670730181e-05,
+      "loss": 8.8379,
+      "step": 110390
+    },
+    {
+      "epoch": 0.5513246273315189,
+      "grad_norm": 0.09357352554798126,
+      "learning_rate": 1.349395479236026e-05,
+      "loss": 8.8308,
+      "step": 110400
+    },
+    {
+      "epoch": 0.5513745661564583,
+      "grad_norm": 0.09357526153326035,
+      "learning_rate": 1.349245287741871e-05,
+      "loss": 8.8173,
+      "step": 110410
+    },
+    {
+      "epoch": 0.5514245049813978,
+      "grad_norm": 0.08764225989580154,
+      "learning_rate": 1.349095096247716e-05,
+      "loss": 8.8153,
+      "step": 110420
+    },
+    {
+      "epoch": 0.5514744438063373,
+      "grad_norm": 0.09387609362602234,
+      "learning_rate": 1.3489449047535608e-05,
+      "loss": 8.8175,
+      "step": 110430
+    },
+    {
+      "epoch": 0.5515243826312767,
+      "grad_norm": 0.09247896075248718,
+      "learning_rate": 1.3487947132594057e-05,
+      "loss": 8.809,
+      "step": 110440
+    },
+    {
+      "epoch": 0.5515743214562161,
+      "grad_norm": 0.09111214429140091,
+      "learning_rate": 1.3486445217652507e-05,
+      "loss": 8.8131,
+      "step": 110450
+    },
+    {
+      "epoch": 0.5516242602811556,
+      "grad_norm": 0.09693354368209839,
+      "learning_rate": 1.3484943302710957e-05,
+      "loss": 8.8086,
+      "step": 110460
+    },
+    {
+      "epoch": 0.551674199106095,
+      "grad_norm": 0.09754697233438492,
+      "learning_rate": 1.3483441387769407e-05,
+      "loss": 8.8201,
+      "step": 110470
+    },
+    {
+      "epoch": 0.5517241379310345,
+      "grad_norm": 0.09200212359428406,
+      "learning_rate": 1.3481939472827857e-05,
+      "loss": 8.8175,
+      "step": 110480
+    },
+    {
+      "epoch": 0.5517740767559739,
+      "grad_norm": 0.10172619670629501,
+      "learning_rate": 1.3480437557886304e-05,
+      "loss": 8.8139,
+      "step": 110490
+    },
+    {
+      "epoch": 0.5518240155809134,
+      "grad_norm": 0.09638337045907974,
+      "learning_rate": 1.3478935642944754e-05,
+      "loss": 8.8247,
+      "step": 110500
+    },
+    {
+      "epoch": 0.5518739544058529,
+      "grad_norm": 0.09530887752771378,
+      "learning_rate": 1.3477433728003205e-05,
+      "loss": 8.8224,
+      "step": 110510
+    },
+    {
+      "epoch": 0.5519238932307923,
+      "grad_norm": 0.09654687345027924,
+      "learning_rate": 1.3475931813061655e-05,
+      "loss": 8.8239,
+      "step": 110520
+    },
+    {
+      "epoch": 0.5519738320557317,
+      "grad_norm": 0.0979146957397461,
+      "learning_rate": 1.3474429898120105e-05,
+      "loss": 8.827,
+      "step": 110530
+    },
+    {
+      "epoch": 0.5520237708806712,
+      "grad_norm": 0.09774132072925568,
+      "learning_rate": 1.3472927983178552e-05,
+      "loss": 8.7931,
+      "step": 110540
+    },
+    {
+      "epoch": 0.5520737097056106,
+      "grad_norm": 0.09289306402206421,
+      "learning_rate": 1.3471426068237002e-05,
+      "loss": 8.8123,
+      "step": 110550
+    },
+    {
+      "epoch": 0.5521236485305501,
+      "grad_norm": 0.08744509518146515,
+      "learning_rate": 1.3469924153295452e-05,
+      "loss": 8.8039,
+      "step": 110560
+    },
+    {
+      "epoch": 0.5521735873554895,
+      "grad_norm": 0.08979026228189468,
+      "learning_rate": 1.3468422238353902e-05,
+      "loss": 8.8056,
+      "step": 110570
+    },
+    {
+      "epoch": 0.552223526180429,
+      "grad_norm": 0.09543202817440033,
+      "learning_rate": 1.3466920323412352e-05,
+      "loss": 8.8093,
+      "step": 110580
+    },
+    {
+      "epoch": 0.5522734650053684,
+      "grad_norm": 0.09224719554185867,
+      "learning_rate": 1.34654184084708e-05,
+      "loss": 8.8181,
+      "step": 110590
+    },
+    {
+      "epoch": 0.5523234038303079,
+      "grad_norm": 0.0893092080950737,
+      "learning_rate": 1.346391649352925e-05,
+      "loss": 8.8243,
+      "step": 110600
+    },
+    {
+      "epoch": 0.5523733426552473,
+      "grad_norm": 0.09298286586999893,
+      "learning_rate": 1.34624145785877e-05,
+      "loss": 8.8224,
+      "step": 110610
+    },
+    {
+      "epoch": 0.5524232814801868,
+      "grad_norm": 0.09143733233213425,
+      "learning_rate": 1.346091266364615e-05,
+      "loss": 8.8141,
+      "step": 110620
+    },
+    {
+      "epoch": 0.5524732203051262,
+      "grad_norm": 0.09493601322174072,
+      "learning_rate": 1.34594107487046e-05,
+      "loss": 8.8273,
+      "step": 110630
+    },
+    {
+      "epoch": 0.5525231591300657,
+      "grad_norm": 0.09791026264429092,
+      "learning_rate": 1.3457908833763048e-05,
+      "loss": 8.801,
+      "step": 110640
+    },
+    {
+      "epoch": 0.5525730979550051,
+      "grad_norm": 0.08950292319059372,
+      "learning_rate": 1.3456406918821497e-05,
+      "loss": 8.8103,
+      "step": 110650
+    },
+    {
+      "epoch": 0.5526230367799446,
+      "grad_norm": 0.09202896058559418,
+      "learning_rate": 1.3454905003879947e-05,
+      "loss": 8.8157,
+      "step": 110660
+    },
+    {
+      "epoch": 0.552672975604884,
+      "grad_norm": 0.09055034816265106,
+      "learning_rate": 1.3453403088938397e-05,
+      "loss": 8.8016,
+      "step": 110670
+    },
+    {
+      "epoch": 0.5527229144298235,
+      "grad_norm": 0.09382042288780212,
+      "learning_rate": 1.3451901173996847e-05,
+      "loss": 8.8212,
+      "step": 110680
+    },
+    {
+      "epoch": 0.5527728532547629,
+      "grad_norm": 0.09259592741727829,
+      "learning_rate": 1.3450399259055296e-05,
+      "loss": 8.799,
+      "step": 110690
+    },
+    {
+      "epoch": 0.5528227920797024,
+      "grad_norm": 0.09515759348869324,
+      "learning_rate": 1.3448897344113744e-05,
+      "loss": 8.8096,
+      "step": 110700
+    },
+    {
+      "epoch": 0.5528727309046418,
+      "grad_norm": 0.09235917776823044,
+      "learning_rate": 1.3447395429172195e-05,
+      "loss": 8.8192,
+      "step": 110710
+    },
+    {
+      "epoch": 0.5529226697295813,
+      "grad_norm": 0.09815475344657898,
+      "learning_rate": 1.3445893514230645e-05,
+      "loss": 8.8107,
+      "step": 110720
+    },
+    {
+      "epoch": 0.5529726085545207,
+      "grad_norm": 0.09027951210737228,
+      "learning_rate": 1.3444391599289095e-05,
+      "loss": 8.8183,
+      "step": 110730
+    },
+    {
+      "epoch": 0.5530225473794602,
+      "grad_norm": 0.09175124019384384,
+      "learning_rate": 1.3442889684347543e-05,
+      "loss": 8.8093,
+      "step": 110740
+    },
+    {
+      "epoch": 0.5530724862043996,
+      "grad_norm": 0.0899006649851799,
+      "learning_rate": 1.3441387769405992e-05,
+      "loss": 8.8136,
+      "step": 110750
+    },
+    {
+      "epoch": 0.5531224250293391,
+      "grad_norm": 0.0890357568860054,
+      "learning_rate": 1.3439885854464442e-05,
+      "loss": 8.8141,
+      "step": 110760
+    },
+    {
+      "epoch": 0.5531723638542785,
+      "grad_norm": 0.09818465262651443,
+      "learning_rate": 1.3438383939522892e-05,
+      "loss": 8.8089,
+      "step": 110770
+    },
+    {
+      "epoch": 0.5532223026792179,
+      "grad_norm": 0.09356721490621567,
+      "learning_rate": 1.3436882024581342e-05,
+      "loss": 8.8082,
+      "step": 110780
+    },
+    {
+      "epoch": 0.5532722415041574,
+      "grad_norm": 0.09822935611009598,
+      "learning_rate": 1.3435380109639791e-05,
+      "loss": 8.8022,
+      "step": 110790
+    },
+    {
+      "epoch": 0.5533221803290969,
+      "grad_norm": 0.09416577219963074,
+      "learning_rate": 1.3433878194698241e-05,
+      "loss": 8.8152,
+      "step": 110800
+    },
+    {
+      "epoch": 0.5533721191540363,
+      "grad_norm": 0.08785787969827652,
+      "learning_rate": 1.343237627975669e-05,
+      "loss": 8.8243,
+      "step": 110810
+    },
+    {
+      "epoch": 0.5534220579789757,
+      "grad_norm": 0.09121017158031464,
+      "learning_rate": 1.343087436481514e-05,
+      "loss": 8.8282,
+      "step": 110820
+    },
+    {
+      "epoch": 0.5534719968039152,
+      "grad_norm": 0.09230327606201172,
+      "learning_rate": 1.342937244987359e-05,
+      "loss": 8.807,
+      "step": 110830
+    },
+    {
+      "epoch": 0.5535219356288547,
+      "grad_norm": 0.09492280334234238,
+      "learning_rate": 1.3427870534932038e-05,
+      "loss": 8.805,
+      "step": 110840
+    },
+    {
+      "epoch": 0.5535718744537941,
+      "grad_norm": 0.09432464092969894,
+      "learning_rate": 1.3426368619990489e-05,
+      "loss": 8.8028,
+      "step": 110850
+    },
+    {
+      "epoch": 0.5536218132787335,
+      "grad_norm": 0.08931145071983337,
+      "learning_rate": 1.3424866705048937e-05,
+      "loss": 8.8191,
+      "step": 110860
+    },
+    {
+      "epoch": 0.553671752103673,
+      "grad_norm": 0.08781544119119644,
+      "learning_rate": 1.3423364790107387e-05,
+      "loss": 8.8121,
+      "step": 110870
+    },
+    {
+      "epoch": 0.5537216909286125,
+      "grad_norm": 0.09263685345649719,
+      "learning_rate": 1.3421862875165837e-05,
+      "loss": 8.8321,
+      "step": 110880
+    },
+    {
+      "epoch": 0.5537716297535519,
+      "grad_norm": 0.0927627757191658,
+      "learning_rate": 1.3420360960224286e-05,
+      "loss": 8.8092,
+      "step": 110890
+    },
+    {
+      "epoch": 0.5538215685784913,
+      "grad_norm": 0.09033826738595963,
+      "learning_rate": 1.3418859045282736e-05,
+      "loss": 8.8183,
+      "step": 110900
+    },
+    {
+      "epoch": 0.5538715074034308,
+      "grad_norm": 0.09372294694185257,
+      "learning_rate": 1.3417357130341185e-05,
+      "loss": 8.807,
+      "step": 110910
+    },
+    {
+      "epoch": 0.5539214462283703,
+      "grad_norm": 0.09317562729120255,
+      "learning_rate": 1.3415855215399635e-05,
+      "loss": 8.8221,
+      "step": 110920
+    },
+    {
+      "epoch": 0.5539713850533097,
+      "grad_norm": 0.09567903727293015,
+      "learning_rate": 1.3414353300458085e-05,
+      "loss": 8.8016,
+      "step": 110930
+    },
+    {
+      "epoch": 0.5540213238782491,
+      "grad_norm": 0.09196696430444717,
+      "learning_rate": 1.3412851385516533e-05,
+      "loss": 8.7924,
+      "step": 110940
+    },
+    {
+      "epoch": 0.5540712627031886,
+      "grad_norm": 0.09338729828596115,
+      "learning_rate": 1.3411349470574984e-05,
+      "loss": 8.8085,
+      "step": 110950
+    },
+    {
+      "epoch": 0.554121201528128,
+      "grad_norm": 0.08997715264558792,
+      "learning_rate": 1.3409847555633434e-05,
+      "loss": 8.8152,
+      "step": 110960
+    },
+    {
+      "epoch": 0.5541711403530675,
+      "grad_norm": 0.09378908574581146,
+      "learning_rate": 1.3408345640691882e-05,
+      "loss": 8.8156,
+      "step": 110970
+    },
+    {
+      "epoch": 0.5542210791780069,
+      "grad_norm": 0.09587135910987854,
+      "learning_rate": 1.3406843725750333e-05,
+      "loss": 8.8153,
+      "step": 110980
+    },
+    {
+      "epoch": 0.5542710180029464,
+      "grad_norm": 0.09000777453184128,
+      "learning_rate": 1.3405341810808781e-05,
+      "loss": 8.8142,
+      "step": 110990
+    },
+    {
+      "epoch": 0.5543209568278858,
+      "grad_norm": 0.09232640266418457,
+      "learning_rate": 1.3403839895867231e-05,
+      "loss": 8.8225,
+      "step": 111000
+    },
+    {
+      "epoch": 0.5543708956528253,
+      "grad_norm": 0.0918559655547142,
+      "learning_rate": 1.3402337980925681e-05,
+      "loss": 8.8173,
+      "step": 111010
+    },
+    {
+      "epoch": 0.5544208344777647,
+      "grad_norm": 0.09692919254302979,
+      "learning_rate": 1.340083606598413e-05,
+      "loss": 8.8041,
+      "step": 111020
+    },
+    {
+      "epoch": 0.5544707733027042,
+      "grad_norm": 0.09214792400598526,
+      "learning_rate": 1.339933415104258e-05,
+      "loss": 8.8139,
+      "step": 111030
+    },
+    {
+      "epoch": 0.5545207121276436,
+      "grad_norm": 0.10509046912193298,
+      "learning_rate": 1.3397832236101028e-05,
+      "loss": 8.8006,
+      "step": 111040
+    },
+    {
+      "epoch": 0.5545706509525831,
+      "grad_norm": 0.09465457499027252,
+      "learning_rate": 1.3396330321159479e-05,
+      "loss": 8.8227,
+      "step": 111050
+    },
+    {
+      "epoch": 0.5546205897775225,
+      "grad_norm": 0.09613688290119171,
+      "learning_rate": 1.3394828406217929e-05,
+      "loss": 8.8222,
+      "step": 111060
+    },
+    {
+      "epoch": 0.554670528602462,
+      "grad_norm": 0.09685207903385162,
+      "learning_rate": 1.3393326491276377e-05,
+      "loss": 8.805,
+      "step": 111070
+    },
+    {
+      "epoch": 0.5547204674274014,
+      "grad_norm": 0.09501661360263824,
+      "learning_rate": 1.3391824576334828e-05,
+      "loss": 8.8135,
+      "step": 111080
+    },
+    {
+      "epoch": 0.5547704062523409,
+      "grad_norm": 0.08681835234165192,
+      "learning_rate": 1.3390322661393276e-05,
+      "loss": 8.8151,
+      "step": 111090
+    },
+    {
+      "epoch": 0.5548203450772803,
+      "grad_norm": 0.09274887293577194,
+      "learning_rate": 1.3388820746451726e-05,
+      "loss": 8.8216,
+      "step": 111100
+    },
+    {
+      "epoch": 0.5548702839022198,
+      "grad_norm": 0.08929115533828735,
+      "learning_rate": 1.3387318831510176e-05,
+      "loss": 8.8071,
+      "step": 111110
+    },
+    {
+      "epoch": 0.5549202227271592,
+      "grad_norm": 0.0890386626124382,
+      "learning_rate": 1.3385816916568627e-05,
+      "loss": 8.8029,
+      "step": 111120
+    },
+    {
+      "epoch": 0.5549701615520987,
+      "grad_norm": 0.09131958335638046,
+      "learning_rate": 1.3384315001627075e-05,
+      "loss": 8.8071,
+      "step": 111130
+    },
+    {
+      "epoch": 0.5550201003770381,
+      "grad_norm": 0.09491490572690964,
+      "learning_rate": 1.3382813086685524e-05,
+      "loss": 8.8114,
+      "step": 111140
+    },
+    {
+      "epoch": 0.5550700392019776,
+      "grad_norm": 0.09056004881858826,
+      "learning_rate": 1.3381311171743974e-05,
+      "loss": 8.8142,
+      "step": 111150
+    },
+    {
+      "epoch": 0.555119978026917,
+      "grad_norm": 0.09362094849348068,
+      "learning_rate": 1.3379809256802424e-05,
+      "loss": 8.8108,
+      "step": 111160
+    },
+    {
+      "epoch": 0.5551699168518565,
+      "grad_norm": 0.08976966887712479,
+      "learning_rate": 1.3378307341860874e-05,
+      "loss": 8.8087,
+      "step": 111170
+    },
+    {
+      "epoch": 0.5552198556767959,
+      "grad_norm": 0.0920504778623581,
+      "learning_rate": 1.3376805426919323e-05,
+      "loss": 8.8133,
+      "step": 111180
+    },
+    {
+      "epoch": 0.5552697945017354,
+      "grad_norm": 0.09543319791555405,
+      "learning_rate": 1.3375303511977771e-05,
+      "loss": 8.8236,
+      "step": 111190
+    },
+    {
+      "epoch": 0.5553197333266748,
+      "grad_norm": 0.08917569369077682,
+      "learning_rate": 1.3373801597036221e-05,
+      "loss": 8.8106,
+      "step": 111200
+    },
+    {
+      "epoch": 0.5553696721516143,
+      "grad_norm": 0.0944250300526619,
+      "learning_rate": 1.3372299682094671e-05,
+      "loss": 8.8051,
+      "step": 111210
+    },
+    {
+      "epoch": 0.5554196109765537,
+      "grad_norm": 0.09428086876869202,
+      "learning_rate": 1.3370797767153122e-05,
+      "loss": 8.8179,
+      "step": 111220
+    },
+    {
+      "epoch": 0.5554695498014932,
+      "grad_norm": 0.09255554527044296,
+      "learning_rate": 1.336929585221157e-05,
+      "loss": 8.8145,
+      "step": 111230
+    },
+    {
+      "epoch": 0.5555194886264326,
+      "grad_norm": 0.09141411632299423,
+      "learning_rate": 1.3367793937270019e-05,
+      "loss": 8.8118,
+      "step": 111240
+    },
+    {
+      "epoch": 0.555569427451372,
+      "grad_norm": 0.0931018739938736,
+      "learning_rate": 1.3366292022328469e-05,
+      "loss": 8.8022,
+      "step": 111250
+    },
+    {
+      "epoch": 0.5556193662763115,
+      "grad_norm": 0.09240280836820602,
+      "learning_rate": 1.3364790107386919e-05,
+      "loss": 8.8012,
+      "step": 111260
+    },
+    {
+      "epoch": 0.555669305101251,
+      "grad_norm": 0.09618735313415527,
+      "learning_rate": 1.3363288192445369e-05,
+      "loss": 8.8128,
+      "step": 111270
+    },
+    {
+      "epoch": 0.5557192439261904,
+      "grad_norm": 0.097829669713974,
+      "learning_rate": 1.336178627750382e-05,
+      "loss": 8.8086,
+      "step": 111280
+    },
+    {
+      "epoch": 0.5557691827511299,
+      "grad_norm": 0.09861978143453598,
+      "learning_rate": 1.3360284362562266e-05,
+      "loss": 8.8114,
+      "step": 111290
+    },
+    {
+      "epoch": 0.5558191215760693,
+      "grad_norm": 0.08871912211179733,
+      "learning_rate": 1.3358782447620716e-05,
+      "loss": 8.8007,
+      "step": 111300
+    },
+    {
+      "epoch": 0.5558690604010088,
+      "grad_norm": 0.09760592132806778,
+      "learning_rate": 1.3357280532679166e-05,
+      "loss": 8.8187,
+      "step": 111310
+    },
+    {
+      "epoch": 0.5559189992259482,
+      "grad_norm": 0.09599092602729797,
+      "learning_rate": 1.3355778617737617e-05,
+      "loss": 8.8205,
+      "step": 111320
+    },
+    {
+      "epoch": 0.5559689380508877,
+      "grad_norm": 0.0927274227142334,
+      "learning_rate": 1.3354276702796067e-05,
+      "loss": 8.8211,
+      "step": 111330
+    },
+    {
+      "epoch": 0.5560188768758271,
+      "grad_norm": 0.0984937772154808,
+      "learning_rate": 1.3352774787854514e-05,
+      "loss": 8.8142,
+      "step": 111340
+    },
+    {
+      "epoch": 0.5560688157007666,
+      "grad_norm": 0.09656939655542374,
+      "learning_rate": 1.3351272872912964e-05,
+      "loss": 8.8249,
+      "step": 111350
+    },
+    {
+      "epoch": 0.556118754525706,
+      "grad_norm": 0.0973270907998085,
+      "learning_rate": 1.3349770957971414e-05,
+      "loss": 8.8217,
+      "step": 111360
+    },
+    {
+      "epoch": 0.5561686933506454,
+      "grad_norm": 0.09051787108182907,
+      "learning_rate": 1.3348269043029864e-05,
+      "loss": 8.791,
+      "step": 111370
+    },
+    {
+      "epoch": 0.5562186321755849,
+      "grad_norm": 0.09165889024734497,
+      "learning_rate": 1.3346767128088314e-05,
+      "loss": 8.8026,
+      "step": 111380
+    },
+    {
+      "epoch": 0.5562685710005244,
+      "grad_norm": 0.09260290116071701,
+      "learning_rate": 1.3345265213146761e-05,
+      "loss": 8.8246,
+      "step": 111390
+    },
+    {
+      "epoch": 0.5563185098254638,
+      "grad_norm": 0.08657737821340561,
+      "learning_rate": 1.3343763298205211e-05,
+      "loss": 8.8278,
+      "step": 111400
+    },
+    {
+      "epoch": 0.5563684486504032,
+      "grad_norm": 0.09201980382204056,
+      "learning_rate": 1.3342261383263661e-05,
+      "loss": 8.8152,
+      "step": 111410
+    },
+    {
+      "epoch": 0.5564183874753427,
+      "grad_norm": 0.09139979630708694,
+      "learning_rate": 1.3340759468322112e-05,
+      "loss": 8.8084,
+      "step": 111420
+    },
+    {
+      "epoch": 0.5564683263002822,
+      "grad_norm": 0.08889581263065338,
+      "learning_rate": 1.3339257553380562e-05,
+      "loss": 8.8078,
+      "step": 111430
+    },
+    {
+      "epoch": 0.5565182651252216,
+      "grad_norm": 0.09374164789915085,
+      "learning_rate": 1.3337755638439009e-05,
+      "loss": 8.8154,
+      "step": 111440
+    },
+    {
+      "epoch": 0.556568203950161,
+      "grad_norm": 0.08838196843862534,
+      "learning_rate": 1.3336253723497459e-05,
+      "loss": 8.8182,
+      "step": 111450
+    },
+    {
+      "epoch": 0.5566181427751005,
+      "grad_norm": 0.09526786208152771,
+      "learning_rate": 1.3334751808555909e-05,
+      "loss": 8.8142,
+      "step": 111460
+    },
+    {
+      "epoch": 0.55666808160004,
+      "grad_norm": 0.09561596810817719,
+      "learning_rate": 1.3333249893614359e-05,
+      "loss": 8.809,
+      "step": 111470
+    },
+    {
+      "epoch": 0.5567180204249794,
+      "grad_norm": 0.09850651025772095,
+      "learning_rate": 1.333174797867281e-05,
+      "loss": 8.8064,
+      "step": 111480
+    },
+    {
+      "epoch": 0.5567679592499188,
+      "grad_norm": 0.09541216492652893,
+      "learning_rate": 1.3330246063731258e-05,
+      "loss": 8.8111,
+      "step": 111490
+    },
+    {
+      "epoch": 0.5568178980748583,
+      "grad_norm": 0.08966288715600967,
+      "learning_rate": 1.3328744148789706e-05,
+      "loss": 8.8197,
+      "step": 111500
+    },
+    {
+      "epoch": 0.5568678368997978,
+      "grad_norm": 0.09588926285505295,
+      "learning_rate": 1.3327242233848156e-05,
+      "loss": 8.81,
+      "step": 111510
+    },
+    {
+      "epoch": 0.5569177757247372,
+      "grad_norm": 0.09180673956871033,
+      "learning_rate": 1.3325740318906607e-05,
+      "loss": 8.7976,
+      "step": 111520
+    },
+    {
+      "epoch": 0.5569677145496766,
+      "grad_norm": 0.09179525822401047,
+      "learning_rate": 1.3324238403965057e-05,
+      "loss": 8.8198,
+      "step": 111530
+    },
+    {
+      "epoch": 0.5570176533746161,
+      "grad_norm": 0.09658701717853546,
+      "learning_rate": 1.3322736489023505e-05,
+      "loss": 8.8084,
+      "step": 111540
+    },
+    {
+      "epoch": 0.5570675921995556,
+      "grad_norm": 0.09038843214511871,
+      "learning_rate": 1.3321234574081954e-05,
+      "loss": 8.8118,
+      "step": 111550
+    },
+    {
+      "epoch": 0.557117531024495,
+      "grad_norm": 0.09037397056818008,
+      "learning_rate": 1.3319732659140404e-05,
+      "loss": 8.8033,
+      "step": 111560
+    },
+    {
+      "epoch": 0.5571674698494344,
+      "grad_norm": 0.09661097079515457,
+      "learning_rate": 1.3318230744198854e-05,
+      "loss": 8.8129,
+      "step": 111570
+    },
+    {
+      "epoch": 0.5572174086743739,
+      "grad_norm": 0.09827481955289841,
+      "learning_rate": 1.3316728829257304e-05,
+      "loss": 8.8086,
+      "step": 111580
+    },
+    {
+      "epoch": 0.5572673474993134,
+      "grad_norm": 0.09280990809202194,
+      "learning_rate": 1.3315226914315753e-05,
+      "loss": 8.822,
+      "step": 111590
+    },
+    {
+      "epoch": 0.5573172863242528,
+      "grad_norm": 0.09101619571447372,
+      "learning_rate": 1.3313724999374201e-05,
+      "loss": 8.8055,
+      "step": 111600
+    },
+    {
+      "epoch": 0.5573672251491922,
+      "grad_norm": 0.09069830179214478,
+      "learning_rate": 1.3312223084432651e-05,
+      "loss": 8.8176,
+      "step": 111610
+    },
+    {
+      "epoch": 0.5574171639741317,
+      "grad_norm": 0.08932122588157654,
+      "learning_rate": 1.3310721169491102e-05,
+      "loss": 8.8185,
+      "step": 111620
+    },
+    {
+      "epoch": 0.5574671027990712,
+      "grad_norm": 0.08982568234205246,
+      "learning_rate": 1.3309219254549552e-05,
+      "loss": 8.8209,
+      "step": 111630
+    },
+    {
+      "epoch": 0.5575170416240106,
+      "grad_norm": 0.09528286010026932,
+      "learning_rate": 1.3307717339608e-05,
+      "loss": 8.8101,
+      "step": 111640
+    },
+    {
+      "epoch": 0.55756698044895,
+      "grad_norm": 0.09759854525327682,
+      "learning_rate": 1.330621542466645e-05,
+      "loss": 8.7972,
+      "step": 111650
+    },
+    {
+      "epoch": 0.5576169192738895,
+      "grad_norm": 0.09867121279239655,
+      "learning_rate": 1.3304713509724899e-05,
+      "loss": 8.817,
+      "step": 111660
+    },
+    {
+      "epoch": 0.557666858098829,
+      "grad_norm": 0.09287247806787491,
+      "learning_rate": 1.3303211594783349e-05,
+      "loss": 8.8053,
+      "step": 111670
+    },
+    {
+      "epoch": 0.5577167969237684,
+      "grad_norm": 0.09172237664461136,
+      "learning_rate": 1.33017096798418e-05,
+      "loss": 8.8155,
+      "step": 111680
+    },
+    {
+      "epoch": 0.5577667357487078,
+      "grad_norm": 0.08916451781988144,
+      "learning_rate": 1.3300207764900248e-05,
+      "loss": 8.8164,
+      "step": 111690
+    },
+    {
+      "epoch": 0.5578166745736473,
+      "grad_norm": 0.08944099396467209,
+      "learning_rate": 1.3298705849958698e-05,
+      "loss": 8.8193,
+      "step": 111700
+    },
+    {
+      "epoch": 0.5578666133985868,
+      "grad_norm": 0.09073956310749054,
+      "learning_rate": 1.3297203935017146e-05,
+      "loss": 8.8127,
+      "step": 111710
+    },
+    {
+      "epoch": 0.5579165522235262,
+      "grad_norm": 0.09659486263990402,
+      "learning_rate": 1.3295702020075597e-05,
+      "loss": 8.7983,
+      "step": 111720
+    },
+    {
+      "epoch": 0.5579664910484656,
+      "grad_norm": 0.09616279602050781,
+      "learning_rate": 1.3294200105134047e-05,
+      "loss": 8.8022,
+      "step": 111730
+    },
+    {
+      "epoch": 0.558016429873405,
+      "grad_norm": 0.09376160800457001,
+      "learning_rate": 1.3292698190192495e-05,
+      "loss": 8.8009,
+      "step": 111740
+    },
+    {
+      "epoch": 0.5580663686983445,
+      "grad_norm": 0.10133592039346695,
+      "learning_rate": 1.3291196275250945e-05,
+      "loss": 8.8166,
+      "step": 111750
+    },
+    {
+      "epoch": 0.558116307523284,
+      "grad_norm": 0.09426382184028625,
+      "learning_rate": 1.3289694360309394e-05,
+      "loss": 8.8035,
+      "step": 111760
+    },
+    {
+      "epoch": 0.5581662463482234,
+      "grad_norm": 0.09278400242328644,
+      "learning_rate": 1.3288192445367844e-05,
+      "loss": 8.8152,
+      "step": 111770
+    },
+    {
+      "epoch": 0.5582161851731628,
+      "grad_norm": 0.09196425974369049,
+      "learning_rate": 1.3286690530426294e-05,
+      "loss": 8.8105,
+      "step": 111780
+    },
+    {
+      "epoch": 0.5582661239981023,
+      "grad_norm": 0.09005366265773773,
+      "learning_rate": 1.3285188615484743e-05,
+      "loss": 8.8056,
+      "step": 111790
+    },
+    {
+      "epoch": 0.5583160628230418,
+      "grad_norm": 0.09192261844873428,
+      "learning_rate": 1.3283686700543193e-05,
+      "loss": 8.8059,
+      "step": 111800
+    },
+    {
+      "epoch": 0.5583660016479812,
+      "grad_norm": 0.09564708918333054,
+      "learning_rate": 1.3282184785601643e-05,
+      "loss": 8.8168,
+      "step": 111810
+    },
+    {
+      "epoch": 0.5584159404729206,
+      "grad_norm": 0.09438828378915787,
+      "learning_rate": 1.3280682870660092e-05,
+      "loss": 8.8086,
+      "step": 111820
+    },
+    {
+      "epoch": 0.5584658792978601,
+      "grad_norm": 0.0901927575469017,
+      "learning_rate": 1.3279180955718542e-05,
+      "loss": 8.8193,
+      "step": 111830
+    },
+    {
+      "epoch": 0.5585158181227996,
+      "grad_norm": 0.0903163030743599,
+      "learning_rate": 1.327767904077699e-05,
+      "loss": 8.8052,
+      "step": 111840
+    },
+    {
+      "epoch": 0.558565756947739,
+      "grad_norm": 0.08931218832731247,
+      "learning_rate": 1.327617712583544e-05,
+      "loss": 8.8079,
+      "step": 111850
+    },
+    {
+      "epoch": 0.5586156957726784,
+      "grad_norm": 0.0938708558678627,
+      "learning_rate": 1.327467521089389e-05,
+      "loss": 8.7983,
+      "step": 111860
+    },
+    {
+      "epoch": 0.5586656345976179,
+      "grad_norm": 0.09172417223453522,
+      "learning_rate": 1.3273173295952339e-05,
+      "loss": 8.8037,
+      "step": 111870
+    },
+    {
+      "epoch": 0.5587155734225574,
+      "grad_norm": 0.0920783281326294,
+      "learning_rate": 1.327167138101079e-05,
+      "loss": 8.7911,
+      "step": 111880
+    },
+    {
+      "epoch": 0.5587655122474968,
+      "grad_norm": 0.09476281702518463,
+      "learning_rate": 1.3270169466069238e-05,
+      "loss": 8.7965,
+      "step": 111890
+    },
+    {
+      "epoch": 0.5588154510724362,
+      "grad_norm": 0.09168004244565964,
+      "learning_rate": 1.3268667551127688e-05,
+      "loss": 8.8016,
+      "step": 111900
+    },
+    {
+      "epoch": 0.5588653898973757,
+      "grad_norm": 0.09263158589601517,
+      "learning_rate": 1.3267165636186138e-05,
+      "loss": 8.8066,
+      "step": 111910
+    },
+    {
+      "epoch": 0.5589153287223152,
+      "grad_norm": 0.09056047350168228,
+      "learning_rate": 1.3265663721244587e-05,
+      "loss": 8.8106,
+      "step": 111920
+    },
+    {
+      "epoch": 0.5589652675472546,
+      "grad_norm": 0.09256269037723541,
+      "learning_rate": 1.3264161806303037e-05,
+      "loss": 8.8149,
+      "step": 111930
+    },
+    {
+      "epoch": 0.559015206372194,
+      "grad_norm": 0.09468802064657211,
+      "learning_rate": 1.3262659891361485e-05,
+      "loss": 8.7959,
+      "step": 111940
+    },
+    {
+      "epoch": 0.5590651451971335,
+      "grad_norm": 0.09123507142066956,
+      "learning_rate": 1.3261157976419935e-05,
+      "loss": 8.7775,
+      "step": 111950
+    },
+    {
+      "epoch": 0.559115084022073,
+      "grad_norm": 0.08695657551288605,
+      "learning_rate": 1.3259656061478386e-05,
+      "loss": 8.8059,
+      "step": 111960
+    },
+    {
+      "epoch": 0.5591650228470124,
+      "grad_norm": 0.09323490411043167,
+      "learning_rate": 1.3258154146536836e-05,
+      "loss": 8.8008,
+      "step": 111970
+    },
+    {
+      "epoch": 0.5592149616719518,
+      "grad_norm": 0.09762334078550339,
+      "learning_rate": 1.3256652231595284e-05,
+      "loss": 8.8023,
+      "step": 111980
+    },
+    {
+      "epoch": 0.5592649004968913,
+      "grad_norm": 0.0962090715765953,
+      "learning_rate": 1.3255150316653733e-05,
+      "loss": 8.8002,
+      "step": 111990
+    },
+    {
+      "epoch": 0.5593148393218308,
+      "grad_norm": 0.09336007386445999,
+      "learning_rate": 1.3253648401712183e-05,
+      "loss": 8.8044,
+      "step": 112000
+    },
+    {
+      "epoch": 0.5593647781467702,
+      "grad_norm": 0.09324468672275543,
+      "learning_rate": 1.3252146486770633e-05,
+      "loss": 8.7984,
+      "step": 112010
+    },
+    {
+      "epoch": 0.5594147169717096,
+      "grad_norm": 0.09045629948377609,
+      "learning_rate": 1.3250644571829083e-05,
+      "loss": 8.7981,
+      "step": 112020
+    },
+    {
+      "epoch": 0.5594646557966491,
+      "grad_norm": 0.09312763065099716,
+      "learning_rate": 1.3249142656887532e-05,
+      "loss": 8.7998,
+      "step": 112030
+    },
+    {
+      "epoch": 0.5595145946215886,
+      "grad_norm": 0.09190090000629425,
+      "learning_rate": 1.324764074194598e-05,
+      "loss": 8.8192,
+      "step": 112040
+    },
+    {
+      "epoch": 0.559564533446528,
+      "grad_norm": 0.09083940088748932,
+      "learning_rate": 1.324613882700443e-05,
+      "loss": 8.8084,
+      "step": 112050
+    },
+    {
+      "epoch": 0.5596144722714674,
+      "grad_norm": 0.09351202100515366,
+      "learning_rate": 1.324463691206288e-05,
+      "loss": 8.8105,
+      "step": 112060
+    },
+    {
+      "epoch": 0.5596644110964069,
+      "grad_norm": 0.09421859681606293,
+      "learning_rate": 1.324313499712133e-05,
+      "loss": 8.7967,
+      "step": 112070
+    },
+    {
+      "epoch": 0.5597143499213464,
+      "grad_norm": 0.09057319909334183,
+      "learning_rate": 1.324163308217978e-05,
+      "loss": 8.8044,
+      "step": 112080
+    },
+    {
+      "epoch": 0.5597642887462858,
+      "grad_norm": 0.09181645512580872,
+      "learning_rate": 1.3240131167238228e-05,
+      "loss": 8.798,
+      "step": 112090
+    },
+    {
+      "epoch": 0.5598142275712252,
+      "grad_norm": 0.08992860466241837,
+      "learning_rate": 1.3238629252296678e-05,
+      "loss": 8.7979,
+      "step": 112100
+    },
+    {
+      "epoch": 0.5598641663961647,
+      "grad_norm": 0.09091092646121979,
+      "learning_rate": 1.3237127337355128e-05,
+      "loss": 8.8105,
+      "step": 112110
+    },
+    {
+      "epoch": 0.5599141052211042,
+      "grad_norm": 0.0932513177394867,
+      "learning_rate": 1.3235625422413578e-05,
+      "loss": 8.7957,
+      "step": 112120
+    },
+    {
+      "epoch": 0.5599640440460436,
+      "grad_norm": 0.08964350819587708,
+      "learning_rate": 1.3234123507472028e-05,
+      "loss": 8.8212,
+      "step": 112130
+    },
+    {
+      "epoch": 0.560013982870983,
+      "grad_norm": 0.09178796410560608,
+      "learning_rate": 1.3232621592530475e-05,
+      "loss": 8.7955,
+      "step": 112140
+    },
+    {
+      "epoch": 0.5600639216959225,
+      "grad_norm": 0.09413108229637146,
+      "learning_rate": 1.3231119677588925e-05,
+      "loss": 8.799,
+      "step": 112150
+    },
+    {
+      "epoch": 0.560113860520862,
+      "grad_norm": 0.09713994711637497,
+      "learning_rate": 1.3229617762647376e-05,
+      "loss": 8.8062,
+      "step": 112160
+    },
+    {
+      "epoch": 0.5601637993458014,
+      "grad_norm": 0.09629582613706589,
+      "learning_rate": 1.3228115847705826e-05,
+      "loss": 8.8064,
+      "step": 112170
+    },
+    {
+      "epoch": 0.5602137381707408,
+      "grad_norm": 0.09238281100988388,
+      "learning_rate": 1.3226613932764276e-05,
+      "loss": 8.8076,
+      "step": 112180
+    },
+    {
+      "epoch": 0.5602636769956802,
+      "grad_norm": 0.0891227275133133,
+      "learning_rate": 1.3225112017822723e-05,
+      "loss": 8.8078,
+      "step": 112190
+    },
+    {
+      "epoch": 0.5603136158206198,
+      "grad_norm": 0.09361886978149414,
+      "learning_rate": 1.3223610102881173e-05,
+      "loss": 8.8078,
+      "step": 112200
+    },
+    {
+      "epoch": 0.5603635546455592,
+      "grad_norm": 0.09347978234291077,
+      "learning_rate": 1.3222108187939623e-05,
+      "loss": 8.7925,
+      "step": 112210
+    },
+    {
+      "epoch": 0.5604134934704986,
+      "grad_norm": 0.09022147953510284,
+      "learning_rate": 1.3220606272998073e-05,
+      "loss": 8.7998,
+      "step": 112220
+    },
+    {
+      "epoch": 0.560463432295438,
+      "grad_norm": 0.09901180863380432,
+      "learning_rate": 1.3219104358056523e-05,
+      "loss": 8.7882,
+      "step": 112230
+    },
+    {
+      "epoch": 0.5605133711203776,
+      "grad_norm": 0.09106682986021042,
+      "learning_rate": 1.3217602443114972e-05,
+      "loss": 8.7916,
+      "step": 112240
+    },
+    {
+      "epoch": 0.560563309945317,
+      "grad_norm": 0.09734653681516647,
+      "learning_rate": 1.321610052817342e-05,
+      "loss": 8.7859,
+      "step": 112250
+    },
+    {
+      "epoch": 0.5606132487702564,
+      "grad_norm": 0.09442363679409027,
+      "learning_rate": 1.321459861323187e-05,
+      "loss": 8.7861,
+      "step": 112260
+    },
+    {
+      "epoch": 0.5606631875951958,
+      "grad_norm": 0.09389246255159378,
+      "learning_rate": 1.321309669829032e-05,
+      "loss": 8.8063,
+      "step": 112270
+    },
+    {
+      "epoch": 0.5607131264201354,
+      "grad_norm": 0.09472633898258209,
+      "learning_rate": 1.3211594783348771e-05,
+      "loss": 8.7963,
+      "step": 112280
+    },
+    {
+      "epoch": 0.5607630652450748,
+      "grad_norm": 0.0962495282292366,
+      "learning_rate": 1.321009286840722e-05,
+      "loss": 8.8155,
+      "step": 112290
+    },
+    {
+      "epoch": 0.5608130040700142,
+      "grad_norm": 0.09074840694665909,
+      "learning_rate": 1.3208590953465668e-05,
+      "loss": 8.8103,
+      "step": 112300
+    },
+    {
+      "epoch": 0.5608629428949536,
+      "grad_norm": 0.09516070038080215,
+      "learning_rate": 1.3207089038524118e-05,
+      "loss": 8.7933,
+      "step": 112310
+    },
+    {
+      "epoch": 0.5609128817198932,
+      "grad_norm": 0.09915341436862946,
+      "learning_rate": 1.3205587123582568e-05,
+      "loss": 8.8079,
+      "step": 112320
+    },
+    {
+      "epoch": 0.5609628205448326,
+      "grad_norm": 0.09369552135467529,
+      "learning_rate": 1.3204085208641018e-05,
+      "loss": 8.8061,
+      "step": 112330
+    },
+    {
+      "epoch": 0.561012759369772,
+      "grad_norm": 0.09473207592964172,
+      "learning_rate": 1.3202583293699467e-05,
+      "loss": 8.8014,
+      "step": 112340
+    },
+    {
+      "epoch": 0.5610626981947114,
+      "grad_norm": 0.09303644299507141,
+      "learning_rate": 1.3201081378757915e-05,
+      "loss": 8.8019,
+      "step": 112350
+    },
+    {
+      "epoch": 0.561112637019651,
+      "grad_norm": 0.09537196904420853,
+      "learning_rate": 1.3199579463816366e-05,
+      "loss": 8.8123,
+      "step": 112360
+    },
+    {
+      "epoch": 0.5611625758445904,
+      "grad_norm": 0.08959188312292099,
+      "learning_rate": 1.3198077548874816e-05,
+      "loss": 8.7895,
+      "step": 112370
+    },
+    {
+      "epoch": 0.5612125146695298,
+      "grad_norm": 0.0941653922200203,
+      "learning_rate": 1.3196575633933266e-05,
+      "loss": 8.7977,
+      "step": 112380
+    },
+    {
+      "epoch": 0.5612624534944692,
+      "grad_norm": 0.09285327792167664,
+      "learning_rate": 1.3195073718991716e-05,
+      "loss": 8.8058,
+      "step": 112390
+    },
+    {
+      "epoch": 0.5613123923194088,
+      "grad_norm": 0.09470658004283905,
+      "learning_rate": 1.3193571804050163e-05,
+      "loss": 8.7981,
+      "step": 112400
+    },
+    {
+      "epoch": 0.5613623311443482,
+      "grad_norm": 0.0920797735452652,
+      "learning_rate": 1.3192069889108613e-05,
+      "loss": 8.8114,
+      "step": 112410
+    },
+    {
+      "epoch": 0.5614122699692876,
+      "grad_norm": 0.09267135709524155,
+      "learning_rate": 1.3190567974167063e-05,
+      "loss": 8.8038,
+      "step": 112420
+    },
+    {
+      "epoch": 0.561462208794227,
+      "grad_norm": 0.0882105603814125,
+      "learning_rate": 1.3189066059225513e-05,
+      "loss": 8.804,
+      "step": 112430
+    },
+    {
+      "epoch": 0.5615121476191666,
+      "grad_norm": 0.09088388085365295,
+      "learning_rate": 1.3187564144283964e-05,
+      "loss": 8.7979,
+      "step": 112440
+    },
+    {
+      "epoch": 0.561562086444106,
+      "grad_norm": 0.09451723098754883,
+      "learning_rate": 1.3186062229342412e-05,
+      "loss": 8.8055,
+      "step": 112450
+    },
+    {
+      "epoch": 0.5616120252690454,
+      "grad_norm": 0.08706700801849365,
+      "learning_rate": 1.318456031440086e-05,
+      "loss": 8.8008,
+      "step": 112460
+    },
+    {
+      "epoch": 0.5616619640939848,
+      "grad_norm": 0.09282489866018295,
+      "learning_rate": 1.318305839945931e-05,
+      "loss": 8.7917,
+      "step": 112470
+    },
+    {
+      "epoch": 0.5617119029189244,
+      "grad_norm": 0.08970408886671066,
+      "learning_rate": 1.3181556484517761e-05,
+      "loss": 8.8134,
+      "step": 112480
+    },
+    {
+      "epoch": 0.5617618417438638,
+      "grad_norm": 0.09117847681045532,
+      "learning_rate": 1.3180054569576211e-05,
+      "loss": 8.8164,
+      "step": 112490
+    },
+    {
+      "epoch": 0.5618117805688032,
+      "grad_norm": 0.0940316915512085,
+      "learning_rate": 1.317855265463466e-05,
+      "loss": 8.81,
+      "step": 112500
+    },
+    {
+      "epoch": 0.5618617193937426,
+      "grad_norm": 0.08806363493204117,
+      "learning_rate": 1.3177050739693108e-05,
+      "loss": 8.8093,
+      "step": 112510
+    },
+    {
+      "epoch": 0.5619116582186822,
+      "grad_norm": 0.0929834246635437,
+      "learning_rate": 1.3175548824751558e-05,
+      "loss": 8.79,
+      "step": 112520
+    },
+    {
+      "epoch": 0.5619615970436216,
+      "grad_norm": 0.09379177540540695,
+      "learning_rate": 1.3174046909810009e-05,
+      "loss": 8.7881,
+      "step": 112530
+    },
+    {
+      "epoch": 0.562011535868561,
+      "grad_norm": 0.0958777591586113,
+      "learning_rate": 1.3172544994868459e-05,
+      "loss": 8.7917,
+      "step": 112540
+    },
+    {
+      "epoch": 0.5620614746935004,
+      "grad_norm": 0.09718500822782516,
+      "learning_rate": 1.3171043079926907e-05,
+      "loss": 8.807,
+      "step": 112550
+    },
+    {
+      "epoch": 0.56211141351844,
+      "grad_norm": 0.09176292270421982,
+      "learning_rate": 1.3169541164985356e-05,
+      "loss": 8.8189,
+      "step": 112560
+    },
+    {
+      "epoch": 0.5621613523433794,
+      "grad_norm": 0.09247379750013351,
+      "learning_rate": 1.3168039250043806e-05,
+      "loss": 8.798,
+      "step": 112570
+    },
+    {
+      "epoch": 0.5622112911683188,
+      "grad_norm": 0.10087962448596954,
+      "learning_rate": 1.3166537335102256e-05,
+      "loss": 8.7897,
+      "step": 112580
+    },
+    {
+      "epoch": 0.5622612299932582,
+      "grad_norm": 0.08978106826543808,
+      "learning_rate": 1.3165035420160706e-05,
+      "loss": 8.801,
+      "step": 112590
+    },
+    {
+      "epoch": 0.5623111688181978,
+      "grad_norm": 0.09266199916601181,
+      "learning_rate": 1.3163533505219155e-05,
+      "loss": 8.7918,
+      "step": 112600
+    },
+    {
+      "epoch": 0.5623611076431372,
+      "grad_norm": 0.087232805788517,
+      "learning_rate": 1.3162031590277605e-05,
+      "loss": 8.8059,
+      "step": 112610
+    },
+    {
+      "epoch": 0.5624110464680766,
+      "grad_norm": 0.09095864742994308,
+      "learning_rate": 1.3160529675336053e-05,
+      "loss": 8.8065,
+      "step": 112620
+    },
+    {
+      "epoch": 0.562460985293016,
+      "grad_norm": 0.09693886339664459,
+      "learning_rate": 1.3159027760394504e-05,
+      "loss": 8.7969,
+      "step": 112630
+    },
+    {
+      "epoch": 0.5625109241179556,
+      "grad_norm": 0.09412913024425507,
+      "learning_rate": 1.3157525845452954e-05,
+      "loss": 8.8179,
+      "step": 112640
+    },
+    {
+      "epoch": 0.562560862942895,
+      "grad_norm": 0.0964449867606163,
+      "learning_rate": 1.3156023930511402e-05,
+      "loss": 8.7957,
+      "step": 112650
+    },
+    {
+      "epoch": 0.5626108017678344,
+      "grad_norm": 0.09539590030908585,
+      "learning_rate": 1.3154522015569852e-05,
+      "loss": 8.7885,
+      "step": 112660
+    },
+    {
+      "epoch": 0.5626607405927738,
+      "grad_norm": 0.0972166359424591,
+      "learning_rate": 1.31530201006283e-05,
+      "loss": 8.8137,
+      "step": 112670
+    },
+    {
+      "epoch": 0.5627106794177134,
+      "grad_norm": 0.092460498213768,
+      "learning_rate": 1.3151518185686751e-05,
+      "loss": 8.8034,
+      "step": 112680
+    },
+    {
+      "epoch": 0.5627606182426528,
+      "grad_norm": 0.0895271971821785,
+      "learning_rate": 1.3150016270745201e-05,
+      "loss": 8.8099,
+      "step": 112690
+    },
+    {
+      "epoch": 0.5628105570675922,
+      "grad_norm": 0.09291508048772812,
+      "learning_rate": 1.314851435580365e-05,
+      "loss": 8.802,
+      "step": 112700
+    },
+    {
+      "epoch": 0.5628604958925316,
+      "grad_norm": 0.08741452544927597,
+      "learning_rate": 1.31470124408621e-05,
+      "loss": 8.7987,
+      "step": 112710
+    },
+    {
+      "epoch": 0.5629104347174712,
+      "grad_norm": 0.09074393659830093,
+      "learning_rate": 1.3145510525920548e-05,
+      "loss": 8.7936,
+      "step": 112720
+    },
+    {
+      "epoch": 0.5629603735424106,
+      "grad_norm": 0.09287057816982269,
+      "learning_rate": 1.3144008610978999e-05,
+      "loss": 8.7991,
+      "step": 112730
+    },
+    {
+      "epoch": 0.56301031236735,
+      "grad_norm": 0.09430692344903946,
+      "learning_rate": 1.3142506696037449e-05,
+      "loss": 8.8014,
+      "step": 112740
+    },
+    {
+      "epoch": 0.5630602511922894,
+      "grad_norm": 0.09498278051614761,
+      "learning_rate": 1.3141004781095897e-05,
+      "loss": 8.8221,
+      "step": 112750
+    },
+    {
+      "epoch": 0.5631101900172288,
+      "grad_norm": 0.09470874071121216,
+      "learning_rate": 1.3139502866154347e-05,
+      "loss": 8.7972,
+      "step": 112760
+    },
+    {
+      "epoch": 0.5631601288421684,
+      "grad_norm": 0.09391231089830399,
+      "learning_rate": 1.3138000951212798e-05,
+      "loss": 8.781,
+      "step": 112770
+    },
+    {
+      "epoch": 0.5632100676671078,
+      "grad_norm": 0.09245941042900085,
+      "learning_rate": 1.3136499036271246e-05,
+      "loss": 8.8051,
+      "step": 112780
+    },
+    {
+      "epoch": 0.5632600064920472,
+      "grad_norm": 0.09180227667093277,
+      "learning_rate": 1.3134997121329696e-05,
+      "loss": 8.7889,
+      "step": 112790
+    },
+    {
+      "epoch": 0.5633099453169866,
+      "grad_norm": 0.10003501176834106,
+      "learning_rate": 1.3133495206388145e-05,
+      "loss": 8.7907,
+      "step": 112800
+    },
+    {
+      "epoch": 0.5633598841419262,
+      "grad_norm": 0.09212306886911392,
+      "learning_rate": 1.3131993291446595e-05,
+      "loss": 8.7978,
+      "step": 112810
+    },
+    {
+      "epoch": 0.5634098229668656,
+      "grad_norm": 0.0921301320195198,
+      "learning_rate": 1.3130491376505045e-05,
+      "loss": 8.797,
+      "step": 112820
+    },
+    {
+      "epoch": 0.563459761791805,
+      "grad_norm": 0.09204480797052383,
+      "learning_rate": 1.3128989461563494e-05,
+      "loss": 8.7892,
+      "step": 112830
+    },
+    {
+      "epoch": 0.5635097006167444,
+      "grad_norm": 0.09469878673553467,
+      "learning_rate": 1.3127487546621944e-05,
+      "loss": 8.8042,
+      "step": 112840
+    },
+    {
+      "epoch": 0.563559639441684,
+      "grad_norm": 0.09218306094408035,
+      "learning_rate": 1.3125985631680392e-05,
+      "loss": 8.8014,
+      "step": 112850
+    },
+    {
+      "epoch": 0.5636095782666234,
+      "grad_norm": 0.09181991964578629,
+      "learning_rate": 1.3124483716738842e-05,
+      "loss": 8.7919,
+      "step": 112860
+    },
+    {
+      "epoch": 0.5636595170915628,
+      "grad_norm": 0.09692985564470291,
+      "learning_rate": 1.3122981801797293e-05,
+      "loss": 8.8035,
+      "step": 112870
+    },
+    {
+      "epoch": 0.5637094559165022,
+      "grad_norm": 0.09414488822221756,
+      "learning_rate": 1.3121479886855741e-05,
+      "loss": 8.7791,
+      "step": 112880
+    },
+    {
+      "epoch": 0.5637593947414418,
+      "grad_norm": 0.09031105786561966,
+      "learning_rate": 1.3119977971914191e-05,
+      "loss": 8.7972,
+      "step": 112890
+    },
+    {
+      "epoch": 0.5638093335663812,
+      "grad_norm": 0.08820466697216034,
+      "learning_rate": 1.311847605697264e-05,
+      "loss": 8.8065,
+      "step": 112900
+    },
+    {
+      "epoch": 0.5638592723913206,
+      "grad_norm": 0.09765250235795975,
+      "learning_rate": 1.311697414203109e-05,
+      "loss": 8.8143,
+      "step": 112910
+    },
+    {
+      "epoch": 0.56390921121626,
+      "grad_norm": 0.09003455191850662,
+      "learning_rate": 1.311547222708954e-05,
+      "loss": 8.7868,
+      "step": 112920
+    },
+    {
+      "epoch": 0.5639591500411996,
+      "grad_norm": 0.09266629070043564,
+      "learning_rate": 1.311397031214799e-05,
+      "loss": 8.8078,
+      "step": 112930
+    },
+    {
+      "epoch": 0.564009088866139,
+      "grad_norm": 0.08923901617527008,
+      "learning_rate": 1.3112468397206439e-05,
+      "loss": 8.8024,
+      "step": 112940
+    },
+    {
+      "epoch": 0.5640590276910784,
+      "grad_norm": 0.09165368974208832,
+      "learning_rate": 1.3110966482264887e-05,
+      "loss": 8.8027,
+      "step": 112950
+    },
+    {
+      "epoch": 0.5641089665160178,
+      "grad_norm": 0.09243005514144897,
+      "learning_rate": 1.3109464567323337e-05,
+      "loss": 8.8092,
+      "step": 112960
+    },
+    {
+      "epoch": 0.5641589053409574,
+      "grad_norm": 0.08696895092725754,
+      "learning_rate": 1.3107962652381788e-05,
+      "loss": 8.8019,
+      "step": 112970
+    },
+    {
+      "epoch": 0.5642088441658968,
+      "grad_norm": 0.0982133075594902,
+      "learning_rate": 1.3106460737440238e-05,
+      "loss": 8.7784,
+      "step": 112980
+    },
+    {
+      "epoch": 0.5642587829908362,
+      "grad_norm": 0.09230280667543411,
+      "learning_rate": 1.3104958822498686e-05,
+      "loss": 8.8033,
+      "step": 112990
+    },
+    {
+      "epoch": 0.5643087218157756,
+      "grad_norm": 0.09832031279802322,
+      "learning_rate": 1.3103456907557135e-05,
+      "loss": 8.8019,
+      "step": 113000
+    },
+    {
+      "epoch": 0.5643586606407152,
+      "grad_norm": 0.09023145586252213,
+      "learning_rate": 1.3101954992615585e-05,
+      "loss": 8.8122,
+      "step": 113010
+    },
+    {
+      "epoch": 0.5644085994656546,
+      "grad_norm": 0.09258860349655151,
+      "learning_rate": 1.3100453077674035e-05,
+      "loss": 8.8005,
+      "step": 113020
+    },
+    {
+      "epoch": 0.564458538290594,
+      "grad_norm": 0.09020715951919556,
+      "learning_rate": 1.3098951162732485e-05,
+      "loss": 8.8047,
+      "step": 113030
+    },
+    {
+      "epoch": 0.5645084771155334,
+      "grad_norm": 0.09244858473539352,
+      "learning_rate": 1.3097449247790934e-05,
+      "loss": 8.7822,
+      "step": 113040
+    },
+    {
+      "epoch": 0.564558415940473,
+      "grad_norm": 0.09436847269535065,
+      "learning_rate": 1.3095947332849382e-05,
+      "loss": 8.8044,
+      "step": 113050
+    },
+    {
+      "epoch": 0.5646083547654124,
+      "grad_norm": 0.0903688296675682,
+      "learning_rate": 1.3094445417907832e-05,
+      "loss": 8.7983,
+      "step": 113060
+    },
+    {
+      "epoch": 0.5646582935903518,
+      "grad_norm": 0.08826575428247452,
+      "learning_rate": 1.3092943502966283e-05,
+      "loss": 8.7907,
+      "step": 113070
+    },
+    {
+      "epoch": 0.5647082324152912,
+      "grad_norm": 0.09253015369176865,
+      "learning_rate": 1.3091441588024733e-05,
+      "loss": 8.7995,
+      "step": 113080
+    },
+    {
+      "epoch": 0.5647581712402308,
+      "grad_norm": 0.08847915381193161,
+      "learning_rate": 1.3089939673083183e-05,
+      "loss": 8.7927,
+      "step": 113090
+    },
+    {
+      "epoch": 0.5648081100651702,
+      "grad_norm": 0.09462785720825195,
+      "learning_rate": 1.308843775814163e-05,
+      "loss": 8.7956,
+      "step": 113100
+    },
+    {
+      "epoch": 0.5648580488901096,
+      "grad_norm": 0.09631918370723724,
+      "learning_rate": 1.308693584320008e-05,
+      "loss": 8.8033,
+      "step": 113110
+    },
+    {
+      "epoch": 0.564907987715049,
+      "grad_norm": 0.09395527094602585,
+      "learning_rate": 1.308543392825853e-05,
+      "loss": 8.7929,
+      "step": 113120
+    },
+    {
+      "epoch": 0.5649579265399886,
+      "grad_norm": 0.09398861974477768,
+      "learning_rate": 1.308393201331698e-05,
+      "loss": 8.7886,
+      "step": 113130
+    },
+    {
+      "epoch": 0.565007865364928,
+      "grad_norm": 0.09166686236858368,
+      "learning_rate": 1.308243009837543e-05,
+      "loss": 8.7877,
+      "step": 113140
+    },
+    {
+      "epoch": 0.5650578041898674,
+      "grad_norm": 0.09073825925588608,
+      "learning_rate": 1.3080928183433877e-05,
+      "loss": 8.8058,
+      "step": 113150
+    },
+    {
+      "epoch": 0.5651077430148068,
+      "grad_norm": 0.09182006120681763,
+      "learning_rate": 1.3079426268492327e-05,
+      "loss": 8.7976,
+      "step": 113160
+    },
+    {
+      "epoch": 0.5651576818397464,
+      "grad_norm": 0.09310693293809891,
+      "learning_rate": 1.3077924353550778e-05,
+      "loss": 8.7859,
+      "step": 113170
+    },
+    {
+      "epoch": 0.5652076206646858,
+      "grad_norm": 0.09456516802310944,
+      "learning_rate": 1.3076422438609228e-05,
+      "loss": 8.8033,
+      "step": 113180
+    },
+    {
+      "epoch": 0.5652575594896252,
+      "grad_norm": 0.08986683934926987,
+      "learning_rate": 1.3074920523667678e-05,
+      "loss": 8.808,
+      "step": 113190
+    },
+    {
+      "epoch": 0.5653074983145646,
+      "grad_norm": 0.09818089008331299,
+      "learning_rate": 1.3073418608726125e-05,
+      "loss": 8.7929,
+      "step": 113200
+    },
+    {
+      "epoch": 0.5653574371395041,
+      "grad_norm": 0.09592483192682266,
+      "learning_rate": 1.3071916693784575e-05,
+      "loss": 8.8035,
+      "step": 113210
+    },
+    {
+      "epoch": 0.5654073759644436,
+      "grad_norm": 0.09427493065595627,
+      "learning_rate": 1.3070414778843025e-05,
+      "loss": 8.7978,
+      "step": 113220
+    },
+    {
+      "epoch": 0.565457314789383,
+      "grad_norm": 0.09667732566595078,
+      "learning_rate": 1.3068912863901475e-05,
+      "loss": 8.7764,
+      "step": 113230
+    },
+    {
+      "epoch": 0.5655072536143224,
+      "grad_norm": 0.10168339312076569,
+      "learning_rate": 1.3067410948959925e-05,
+      "loss": 8.7989,
+      "step": 113240
+    },
+    {
+      "epoch": 0.565557192439262,
+      "grad_norm": 0.09616504609584808,
+      "learning_rate": 1.3065909034018374e-05,
+      "loss": 8.8068,
+      "step": 113250
+    },
+    {
+      "epoch": 0.5656071312642014,
+      "grad_norm": 0.09338529407978058,
+      "learning_rate": 1.3064407119076822e-05,
+      "loss": 8.7902,
+      "step": 113260
+    },
+    {
+      "epoch": 0.5656570700891408,
+      "grad_norm": 0.09293079376220703,
+      "learning_rate": 1.3062905204135273e-05,
+      "loss": 8.7938,
+      "step": 113270
+    },
+    {
+      "epoch": 0.5657070089140802,
+      "grad_norm": 0.09134235978126526,
+      "learning_rate": 1.3061403289193723e-05,
+      "loss": 8.7961,
+      "step": 113280
+    },
+    {
+      "epoch": 0.5657569477390197,
+      "grad_norm": 0.0925818383693695,
+      "learning_rate": 1.3059901374252173e-05,
+      "loss": 8.7874,
+      "step": 113290
+    },
+    {
+      "epoch": 0.5658068865639592,
+      "grad_norm": 0.09122247248888016,
+      "learning_rate": 1.3058399459310621e-05,
+      "loss": 8.807,
+      "step": 113300
+    },
+    {
+      "epoch": 0.5658568253888986,
+      "grad_norm": 0.08886589109897614,
+      "learning_rate": 1.305689754436907e-05,
+      "loss": 8.792,
+      "step": 113310
+    },
+    {
+      "epoch": 0.565906764213838,
+      "grad_norm": 0.09335944801568985,
+      "learning_rate": 1.305539562942752e-05,
+      "loss": 8.7786,
+      "step": 113320
+    },
+    {
+      "epoch": 0.5659567030387775,
+      "grad_norm": 0.088760145008564,
+      "learning_rate": 1.305389371448597e-05,
+      "loss": 8.791,
+      "step": 113330
+    },
+    {
+      "epoch": 0.566006641863717,
+      "grad_norm": 0.09438331425189972,
+      "learning_rate": 1.305239179954442e-05,
+      "loss": 8.7911,
+      "step": 113340
+    },
+    {
+      "epoch": 0.5660565806886564,
+      "grad_norm": 0.09462893754243851,
+      "learning_rate": 1.3050889884602869e-05,
+      "loss": 8.7829,
+      "step": 113350
+    },
+    {
+      "epoch": 0.5661065195135958,
+      "grad_norm": 0.09259051084518433,
+      "learning_rate": 1.3049387969661317e-05,
+      "loss": 8.8002,
+      "step": 113360
+    },
+    {
+      "epoch": 0.5661564583385353,
+      "grad_norm": 0.09230745583772659,
+      "learning_rate": 1.3047886054719768e-05,
+      "loss": 8.7846,
+      "step": 113370
+    },
+    {
+      "epoch": 0.5662063971634748,
+      "grad_norm": 0.09083818644285202,
+      "learning_rate": 1.3046384139778218e-05,
+      "loss": 8.7832,
+      "step": 113380
+    },
+    {
+      "epoch": 0.5662563359884142,
+      "grad_norm": 0.09257438033819199,
+      "learning_rate": 1.3044882224836668e-05,
+      "loss": 8.7789,
+      "step": 113390
+    },
+    {
+      "epoch": 0.5663062748133536,
+      "grad_norm": 0.09438567608594894,
+      "learning_rate": 1.3043380309895116e-05,
+      "loss": 8.7963,
+      "step": 113400
+    },
+    {
+      "epoch": 0.5663562136382931,
+      "grad_norm": 0.09249507635831833,
+      "learning_rate": 1.3041878394953567e-05,
+      "loss": 8.7973,
+      "step": 113410
+    },
+    {
+      "epoch": 0.5664061524632326,
+      "grad_norm": 0.09157245606184006,
+      "learning_rate": 1.3040376480012015e-05,
+      "loss": 8.7952,
+      "step": 113420
+    },
+    {
+      "epoch": 0.566456091288172,
+      "grad_norm": 0.09889290481805801,
+      "learning_rate": 1.3038874565070465e-05,
+      "loss": 8.8033,
+      "step": 113430
+    },
+    {
+      "epoch": 0.5665060301131114,
+      "grad_norm": 0.09277231991291046,
+      "learning_rate": 1.3037372650128915e-05,
+      "loss": 8.8005,
+      "step": 113440
+    },
+    {
+      "epoch": 0.5665559689380509,
+      "grad_norm": 0.09674065560102463,
+      "learning_rate": 1.3035870735187364e-05,
+      "loss": 8.7933,
+      "step": 113450
+    },
+    {
+      "epoch": 0.5666059077629904,
+      "grad_norm": 0.09464483708143234,
+      "learning_rate": 1.3034368820245814e-05,
+      "loss": 8.7983,
+      "step": 113460
+    },
+    {
+      "epoch": 0.5666558465879298,
+      "grad_norm": 0.09485206753015518,
+      "learning_rate": 1.3032866905304263e-05,
+      "loss": 8.797,
+      "step": 113470
+    },
+    {
+      "epoch": 0.5667057854128692,
+      "grad_norm": 0.08628489077091217,
+      "learning_rate": 1.3031364990362713e-05,
+      "loss": 8.8084,
+      "step": 113480
+    },
+    {
+      "epoch": 0.5667557242378087,
+      "grad_norm": 0.09606333822011948,
+      "learning_rate": 1.3029863075421163e-05,
+      "loss": 8.7985,
+      "step": 113490
+    },
+    {
+      "epoch": 0.5668056630627482,
+      "grad_norm": 0.09314040094614029,
+      "learning_rate": 1.3028361160479611e-05,
+      "loss": 8.8068,
+      "step": 113500
+    },
+    {
+      "epoch": 0.5668556018876876,
+      "grad_norm": 0.09132982790470123,
+      "learning_rate": 1.3026859245538062e-05,
+      "loss": 8.7969,
+      "step": 113510
+    },
+    {
+      "epoch": 0.566905540712627,
+      "grad_norm": 0.09420787543058395,
+      "learning_rate": 1.302535733059651e-05,
+      "loss": 8.7799,
+      "step": 113520
+    },
+    {
+      "epoch": 0.5669554795375665,
+      "grad_norm": 0.09258543699979782,
+      "learning_rate": 1.302385541565496e-05,
+      "loss": 8.8007,
+      "step": 113530
+    },
+    {
+      "epoch": 0.567005418362506,
+      "grad_norm": 0.09835248440504074,
+      "learning_rate": 1.302235350071341e-05,
+      "loss": 8.8051,
+      "step": 113540
+    },
+    {
+      "epoch": 0.5670553571874454,
+      "grad_norm": 0.093469999730587,
+      "learning_rate": 1.3020851585771859e-05,
+      "loss": 8.81,
+      "step": 113550
+    },
+    {
+      "epoch": 0.5671052960123848,
+      "grad_norm": 0.09576337039470673,
+      "learning_rate": 1.3019349670830309e-05,
+      "loss": 8.7956,
+      "step": 113560
+    },
+    {
+      "epoch": 0.5671552348373243,
+      "grad_norm": 0.09052767604589462,
+      "learning_rate": 1.301784775588876e-05,
+      "loss": 8.7867,
+      "step": 113570
+    },
+    {
+      "epoch": 0.5672051736622638,
+      "grad_norm": 0.09173665940761566,
+      "learning_rate": 1.3016345840947208e-05,
+      "loss": 8.7889,
+      "step": 113580
+    },
+    {
+      "epoch": 0.5672551124872032,
+      "grad_norm": 0.09585422277450562,
+      "learning_rate": 1.3014843926005658e-05,
+      "loss": 8.799,
+      "step": 113590
+    },
+    {
+      "epoch": 0.5673050513121426,
+      "grad_norm": 0.09513553977012634,
+      "learning_rate": 1.3013342011064106e-05,
+      "loss": 8.767,
+      "step": 113600
+    },
+    {
+      "epoch": 0.5673549901370821,
+      "grad_norm": 0.08924262225627899,
+      "learning_rate": 1.3011840096122557e-05,
+      "loss": 8.8002,
+      "step": 113610
+    },
+    {
+      "epoch": 0.5674049289620215,
+      "grad_norm": 0.0911087617278099,
+      "learning_rate": 1.3010338181181007e-05,
+      "loss": 8.794,
+      "step": 113620
+    },
+    {
+      "epoch": 0.567454867786961,
+      "grad_norm": 0.09030284732580185,
+      "learning_rate": 1.3008836266239455e-05,
+      "loss": 8.8027,
+      "step": 113630
+    },
+    {
+      "epoch": 0.5675048066119004,
+      "grad_norm": 0.09411875158548355,
+      "learning_rate": 1.3007334351297905e-05,
+      "loss": 8.7871,
+      "step": 113640
+    },
+    {
+      "epoch": 0.5675547454368399,
+      "grad_norm": 0.09722502529621124,
+      "learning_rate": 1.3005832436356354e-05,
+      "loss": 8.7956,
+      "step": 113650
+    },
+    {
+      "epoch": 0.5676046842617793,
+      "grad_norm": 0.09281644970178604,
+      "learning_rate": 1.3004330521414804e-05,
+      "loss": 8.787,
+      "step": 113660
+    },
+    {
+      "epoch": 0.5676546230867188,
+      "grad_norm": 0.0939922034740448,
+      "learning_rate": 1.3002828606473254e-05,
+      "loss": 8.7992,
+      "step": 113670
+    },
+    {
+      "epoch": 0.5677045619116582,
+      "grad_norm": 0.08864696323871613,
+      "learning_rate": 1.3001326691531703e-05,
+      "loss": 8.7967,
+      "step": 113680
+    },
+    {
+      "epoch": 0.5677545007365977,
+      "grad_norm": 0.09028666466474533,
+      "learning_rate": 1.2999824776590153e-05,
+      "loss": 8.792,
+      "step": 113690
+    },
+    {
+      "epoch": 0.5678044395615371,
+      "grad_norm": 0.09309808909893036,
+      "learning_rate": 1.2998322861648601e-05,
+      "loss": 8.7893,
+      "step": 113700
+    },
+    {
+      "epoch": 0.5678543783864766,
+      "grad_norm": 0.08978434652090073,
+      "learning_rate": 1.2996820946707052e-05,
+      "loss": 8.7902,
+      "step": 113710
+    },
+    {
+      "epoch": 0.567904317211416,
+      "grad_norm": 0.09246104210615158,
+      "learning_rate": 1.2995319031765502e-05,
+      "loss": 8.7838,
+      "step": 113720
+    },
+    {
+      "epoch": 0.5679542560363554,
+      "grad_norm": 0.0925283282995224,
+      "learning_rate": 1.2993817116823952e-05,
+      "loss": 8.7888,
+      "step": 113730
+    },
+    {
+      "epoch": 0.5680041948612949,
+      "grad_norm": 0.09267081320285797,
+      "learning_rate": 1.29923152018824e-05,
+      "loss": 8.7942,
+      "step": 113740
+    },
+    {
+      "epoch": 0.5680541336862344,
+      "grad_norm": 0.09131681174039841,
+      "learning_rate": 1.2990813286940849e-05,
+      "loss": 8.7979,
+      "step": 113750
+    },
+    {
+      "epoch": 0.5681040725111738,
+      "grad_norm": 0.09731698036193848,
+      "learning_rate": 1.2989311371999299e-05,
+      "loss": 8.7936,
+      "step": 113760
+    },
+    {
+      "epoch": 0.5681540113361132,
+      "grad_norm": 0.09286783635616302,
+      "learning_rate": 1.298780945705775e-05,
+      "loss": 8.8072,
+      "step": 113770
+    },
+    {
+      "epoch": 0.5682039501610527,
+      "grad_norm": 0.08999673277139664,
+      "learning_rate": 1.29863075421162e-05,
+      "loss": 8.7972,
+      "step": 113780
+    },
+    {
+      "epoch": 0.5682538889859922,
+      "grad_norm": 0.09515384584665298,
+      "learning_rate": 1.2984805627174648e-05,
+      "loss": 8.8005,
+      "step": 113790
+    },
+    {
+      "epoch": 0.5683038278109316,
+      "grad_norm": 0.09427673369646072,
+      "learning_rate": 1.2983303712233096e-05,
+      "loss": 8.7889,
+      "step": 113800
+    },
+    {
+      "epoch": 0.568353766635871,
+      "grad_norm": 0.09467284381389618,
+      "learning_rate": 1.2981801797291547e-05,
+      "loss": 8.8057,
+      "step": 113810
+    },
+    {
+      "epoch": 0.5684037054608105,
+      "grad_norm": 0.09350381791591644,
+      "learning_rate": 1.2980299882349997e-05,
+      "loss": 8.7858,
+      "step": 113820
+    },
+    {
+      "epoch": 0.56845364428575,
+      "grad_norm": 0.08991306275129318,
+      "learning_rate": 1.2978797967408447e-05,
+      "loss": 8.7852,
+      "step": 113830
+    },
+    {
+      "epoch": 0.5685035831106894,
+      "grad_norm": 0.08979171514511108,
+      "learning_rate": 1.2977296052466895e-05,
+      "loss": 8.7917,
+      "step": 113840
+    },
+    {
+      "epoch": 0.5685535219356288,
+      "grad_norm": 0.09261246025562286,
+      "learning_rate": 1.2975794137525344e-05,
+      "loss": 8.8051,
+      "step": 113850
+    },
+    {
+      "epoch": 0.5686034607605683,
+      "grad_norm": 0.09397750347852707,
+      "learning_rate": 1.2974292222583794e-05,
+      "loss": 8.7922,
+      "step": 113860
+    },
+    {
+      "epoch": 0.5686533995855078,
+      "grad_norm": 0.08991680294275284,
+      "learning_rate": 1.2972790307642244e-05,
+      "loss": 8.8009,
+      "step": 113870
+    },
+    {
+      "epoch": 0.5687033384104472,
+      "grad_norm": 0.08592606335878372,
+      "learning_rate": 1.2971288392700694e-05,
+      "loss": 8.7829,
+      "step": 113880
+    },
+    {
+      "epoch": 0.5687532772353866,
+      "grad_norm": 0.08787564933300018,
+      "learning_rate": 1.2969786477759145e-05,
+      "loss": 8.7925,
+      "step": 113890
+    },
+    {
+      "epoch": 0.5688032160603261,
+      "grad_norm": 0.09271685034036636,
+      "learning_rate": 1.2968284562817591e-05,
+      "loss": 8.7928,
+      "step": 113900
+    },
+    {
+      "epoch": 0.5688531548852656,
+      "grad_norm": 0.09563078731298447,
+      "learning_rate": 1.2966782647876042e-05,
+      "loss": 8.7901,
+      "step": 113910
+    },
+    {
+      "epoch": 0.568903093710205,
+      "grad_norm": 0.08803986757993698,
+      "learning_rate": 1.2965280732934492e-05,
+      "loss": 8.7871,
+      "step": 113920
+    },
+    {
+      "epoch": 0.5689530325351444,
+      "grad_norm": 0.09317762404680252,
+      "learning_rate": 1.2963778817992942e-05,
+      "loss": 8.7962,
+      "step": 113930
+    },
+    {
+      "epoch": 0.5690029713600839,
+      "grad_norm": 0.09467476606369019,
+      "learning_rate": 1.2962276903051392e-05,
+      "loss": 8.8057,
+      "step": 113940
+    },
+    {
+      "epoch": 0.5690529101850234,
+      "grad_norm": 0.0917622447013855,
+      "learning_rate": 1.2960774988109839e-05,
+      "loss": 8.7854,
+      "step": 113950
+    },
+    {
+      "epoch": 0.5691028490099628,
+      "grad_norm": 0.09286537766456604,
+      "learning_rate": 1.2959273073168289e-05,
+      "loss": 8.7928,
+      "step": 113960
+    },
+    {
+      "epoch": 0.5691527878349022,
+      "grad_norm": 0.09457229822874069,
+      "learning_rate": 1.295777115822674e-05,
+      "loss": 8.8018,
+      "step": 113970
+    },
+    {
+      "epoch": 0.5692027266598417,
+      "grad_norm": 0.09735032916069031,
+      "learning_rate": 1.295626924328519e-05,
+      "loss": 8.7764,
+      "step": 113980
+    },
+    {
+      "epoch": 0.5692526654847812,
+      "grad_norm": 0.09656371176242828,
+      "learning_rate": 1.295476732834364e-05,
+      "loss": 8.7861,
+      "step": 113990
+    },
+    {
+      "epoch": 0.5693026043097206,
+      "grad_norm": 0.0926496610045433,
+      "learning_rate": 1.2953265413402086e-05,
+      "loss": 8.7779,
+      "step": 114000
+    },
+    {
+      "epoch": 0.56935254313466,
+      "grad_norm": 0.09257472306489944,
+      "learning_rate": 1.2951763498460537e-05,
+      "loss": 8.7978,
+      "step": 114010
+    },
+    {
+      "epoch": 0.5694024819595995,
+      "grad_norm": 0.09308171272277832,
+      "learning_rate": 1.2950261583518987e-05,
+      "loss": 8.7984,
+      "step": 114020
+    },
+    {
+      "epoch": 0.569452420784539,
+      "grad_norm": 0.09056945145130157,
+      "learning_rate": 1.2948759668577437e-05,
+      "loss": 8.8057,
+      "step": 114030
+    },
+    {
+      "epoch": 0.5695023596094784,
+      "grad_norm": 0.09501245617866516,
+      "learning_rate": 1.2947257753635887e-05,
+      "loss": 8.7859,
+      "step": 114040
+    },
+    {
+      "epoch": 0.5695522984344178,
+      "grad_norm": 0.08686729520559311,
+      "learning_rate": 1.2945755838694336e-05,
+      "loss": 8.7834,
+      "step": 114050
+    },
+    {
+      "epoch": 0.5696022372593573,
+      "grad_norm": 0.08709965646266937,
+      "learning_rate": 1.2944253923752784e-05,
+      "loss": 8.7863,
+      "step": 114060
+    },
+    {
+      "epoch": 0.5696521760842967,
+      "grad_norm": 0.09584898501634598,
+      "learning_rate": 1.2942752008811234e-05,
+      "loss": 8.7893,
+      "step": 114070
+    },
+    {
+      "epoch": 0.5697021149092362,
+      "grad_norm": 0.09185926616191864,
+      "learning_rate": 1.2941250093869685e-05,
+      "loss": 8.7821,
+      "step": 114080
+    },
+    {
+      "epoch": 0.5697520537341756,
+      "grad_norm": 0.09461195766925812,
+      "learning_rate": 1.2939748178928135e-05,
+      "loss": 8.7804,
+      "step": 114090
+    },
+    {
+      "epoch": 0.5698019925591151,
+      "grad_norm": 0.09803737699985504,
+      "learning_rate": 1.2938246263986583e-05,
+      "loss": 8.7888,
+      "step": 114100
+    },
+    {
+      "epoch": 0.5698519313840545,
+      "grad_norm": 0.08924354612827301,
+      "learning_rate": 1.2936744349045032e-05,
+      "loss": 8.7978,
+      "step": 114110
+    },
+    {
+      "epoch": 0.569901870208994,
+      "grad_norm": 0.0952063798904419,
+      "learning_rate": 1.2935242434103482e-05,
+      "loss": 8.8019,
+      "step": 114120
+    },
+    {
+      "epoch": 0.5699518090339334,
+      "grad_norm": 0.09857337176799774,
+      "learning_rate": 1.2933740519161932e-05,
+      "loss": 8.7956,
+      "step": 114130
+    },
+    {
+      "epoch": 0.5700017478588729,
+      "grad_norm": 0.09166187793016434,
+      "learning_rate": 1.2932238604220382e-05,
+      "loss": 8.7832,
+      "step": 114140
+    },
+    {
+      "epoch": 0.5700516866838123,
+      "grad_norm": 0.09211006760597229,
+      "learning_rate": 1.293073668927883e-05,
+      "loss": 8.7998,
+      "step": 114150
+    },
+    {
+      "epoch": 0.5701016255087518,
+      "grad_norm": 0.09411497414112091,
+      "learning_rate": 1.292923477433728e-05,
+      "loss": 8.7912,
+      "step": 114160
+    },
+    {
+      "epoch": 0.5701515643336912,
+      "grad_norm": 0.09173117578029633,
+      "learning_rate": 1.292773285939573e-05,
+      "loss": 8.8004,
+      "step": 114170
+    },
+    {
+      "epoch": 0.5702015031586307,
+      "grad_norm": 0.09168483316898346,
+      "learning_rate": 1.292623094445418e-05,
+      "loss": 8.7954,
+      "step": 114180
+    },
+    {
+      "epoch": 0.5702514419835701,
+      "grad_norm": 0.09110462665557861,
+      "learning_rate": 1.292472902951263e-05,
+      "loss": 8.7839,
+      "step": 114190
+    },
+    {
+      "epoch": 0.5703013808085096,
+      "grad_norm": 0.0920116975903511,
+      "learning_rate": 1.2923227114571078e-05,
+      "loss": 8.7785,
+      "step": 114200
+    },
+    {
+      "epoch": 0.570351319633449,
+      "grad_norm": 0.08953049778938293,
+      "learning_rate": 1.2921725199629528e-05,
+      "loss": 8.7941,
+      "step": 114210
+    },
+    {
+      "epoch": 0.5704012584583885,
+      "grad_norm": 0.09635299444198608,
+      "learning_rate": 1.2920223284687977e-05,
+      "loss": 8.7874,
+      "step": 114220
+    },
+    {
+      "epoch": 0.5704511972833279,
+      "grad_norm": 0.09404581040143967,
+      "learning_rate": 1.2918721369746427e-05,
+      "loss": 8.7777,
+      "step": 114230
+    },
+    {
+      "epoch": 0.5705011361082674,
+      "grad_norm": 0.08653539419174194,
+      "learning_rate": 1.2917219454804877e-05,
+      "loss": 8.7868,
+      "step": 114240
+    },
+    {
+      "epoch": 0.5705510749332068,
+      "grad_norm": 0.08841050416231155,
+      "learning_rate": 1.2915717539863326e-05,
+      "loss": 8.7813,
+      "step": 114250
+    },
+    {
+      "epoch": 0.5706010137581463,
+      "grad_norm": 0.0912218913435936,
+      "learning_rate": 1.2914215624921776e-05,
+      "loss": 8.8012,
+      "step": 114260
+    },
+    {
+      "epoch": 0.5706509525830857,
+      "grad_norm": 0.09305412322282791,
+      "learning_rate": 1.2912713709980224e-05,
+      "loss": 8.8028,
+      "step": 114270
+    },
+    {
+      "epoch": 0.5707008914080252,
+      "grad_norm": 0.09391267597675323,
+      "learning_rate": 1.2911211795038675e-05,
+      "loss": 8.7863,
+      "step": 114280
+    },
+    {
+      "epoch": 0.5707508302329646,
+      "grad_norm": 0.09395688027143478,
+      "learning_rate": 1.2909709880097125e-05,
+      "loss": 8.796,
+      "step": 114290
+    },
+    {
+      "epoch": 0.5708007690579041,
+      "grad_norm": 0.09076563268899918,
+      "learning_rate": 1.2908207965155573e-05,
+      "loss": 8.7863,
+      "step": 114300
+    },
+    {
+      "epoch": 0.5708507078828435,
+      "grad_norm": 0.09484805911779404,
+      "learning_rate": 1.2906706050214023e-05,
+      "loss": 8.7788,
+      "step": 114310
+    },
+    {
+      "epoch": 0.570900646707783,
+      "grad_norm": 0.0906333476305008,
+      "learning_rate": 1.2905204135272472e-05,
+      "loss": 8.7787,
+      "step": 114320
+    },
+    {
+      "epoch": 0.5709505855327224,
+      "grad_norm": 0.0936955064535141,
+      "learning_rate": 1.2903702220330922e-05,
+      "loss": 8.7984,
+      "step": 114330
+    },
+    {
+      "epoch": 0.5710005243576619,
+      "grad_norm": 0.09082095324993134,
+      "learning_rate": 1.2902200305389372e-05,
+      "loss": 8.7958,
+      "step": 114340
+    },
+    {
+      "epoch": 0.5710504631826013,
+      "grad_norm": 0.096087247133255,
+      "learning_rate": 1.290069839044782e-05,
+      "loss": 8.7895,
+      "step": 114350
+    },
+    {
+      "epoch": 0.5711004020075408,
+      "grad_norm": 0.09378685802221298,
+      "learning_rate": 1.2899196475506271e-05,
+      "loss": 8.7897,
+      "step": 114360
+    },
+    {
+      "epoch": 0.5711503408324802,
+      "grad_norm": 0.09574133902788162,
+      "learning_rate": 1.289769456056472e-05,
+      "loss": 8.7893,
+      "step": 114370
+    },
+    {
+      "epoch": 0.5712002796574197,
+      "grad_norm": 0.09573321789503098,
+      "learning_rate": 1.289619264562317e-05,
+      "loss": 8.7822,
+      "step": 114380
+    },
+    {
+      "epoch": 0.5712502184823591,
+      "grad_norm": 0.0985926166176796,
+      "learning_rate": 1.289469073068162e-05,
+      "loss": 8.7745,
+      "step": 114390
+    },
+    {
+      "epoch": 0.5713001573072986,
+      "grad_norm": 0.09592541307210922,
+      "learning_rate": 1.2893188815740068e-05,
+      "loss": 8.7653,
+      "step": 114400
+    },
+    {
+      "epoch": 0.571350096132238,
+      "grad_norm": 0.09706228971481323,
+      "learning_rate": 1.2891686900798518e-05,
+      "loss": 8.787,
+      "step": 114410
+    },
+    {
+      "epoch": 0.5714000349571775,
+      "grad_norm": 0.09211396425962448,
+      "learning_rate": 1.2890184985856969e-05,
+      "loss": 8.7994,
+      "step": 114420
+    },
+    {
+      "epoch": 0.5714499737821169,
+      "grad_norm": 0.08897708356380463,
+      "learning_rate": 1.2888683070915417e-05,
+      "loss": 8.7914,
+      "step": 114430
+    },
+    {
+      "epoch": 0.5714999126070563,
+      "grad_norm": 0.09851206839084625,
+      "learning_rate": 1.2887181155973867e-05,
+      "loss": 8.7865,
+      "step": 114440
+    },
+    {
+      "epoch": 0.5715498514319958,
+      "grad_norm": 0.0946231335401535,
+      "learning_rate": 1.2885679241032316e-05,
+      "loss": 8.7873,
+      "step": 114450
+    },
+    {
+      "epoch": 0.5715997902569353,
+      "grad_norm": 0.096876360476017,
+      "learning_rate": 1.2884177326090766e-05,
+      "loss": 8.7855,
+      "step": 114460
+    },
+    {
+      "epoch": 0.5716497290818747,
+      "grad_norm": 0.09491303563117981,
+      "learning_rate": 1.2882675411149216e-05,
+      "loss": 8.806,
+      "step": 114470
+    },
+    {
+      "epoch": 0.5716996679068141,
+      "grad_norm": 0.08952588587999344,
+      "learning_rate": 1.2881173496207665e-05,
+      "loss": 8.7896,
+      "step": 114480
+    },
+    {
+      "epoch": 0.5717496067317536,
+      "grad_norm": 0.08914480358362198,
+      "learning_rate": 1.2879671581266115e-05,
+      "loss": 8.7765,
+      "step": 114490
+    },
+    {
+      "epoch": 0.5717995455566931,
+      "grad_norm": 0.09196481853723526,
+      "learning_rate": 1.2878169666324565e-05,
+      "loss": 8.7767,
+      "step": 114500
+    },
+    {
+      "epoch": 0.5718494843816325,
+      "grad_norm": 0.08663848787546158,
+      "learning_rate": 1.2876667751383013e-05,
+      "loss": 8.7991,
+      "step": 114510
+    },
+    {
+      "epoch": 0.5718994232065719,
+      "grad_norm": 0.08811726421117783,
+      "learning_rate": 1.2875165836441464e-05,
+      "loss": 8.786,
+      "step": 114520
+    },
+    {
+      "epoch": 0.5719493620315114,
+      "grad_norm": 0.09745927155017853,
+      "learning_rate": 1.2873663921499912e-05,
+      "loss": 8.7802,
+      "step": 114530
+    },
+    {
+      "epoch": 0.5719993008564509,
+      "grad_norm": 0.09472652524709702,
+      "learning_rate": 1.2872162006558362e-05,
+      "loss": 8.7909,
+      "step": 114540
+    },
+    {
+      "epoch": 0.5720492396813903,
+      "grad_norm": 0.10268702358007431,
+      "learning_rate": 1.2870660091616812e-05,
+      "loss": 8.7917,
+      "step": 114550
+    },
+    {
+      "epoch": 0.5720991785063297,
+      "grad_norm": 0.09564602375030518,
+      "learning_rate": 1.2869158176675261e-05,
+      "loss": 8.7963,
+      "step": 114560
+    },
+    {
+      "epoch": 0.5721491173312692,
+      "grad_norm": 0.08768244832754135,
+      "learning_rate": 1.2867656261733711e-05,
+      "loss": 8.7989,
+      "step": 114570
+    },
+    {
+      "epoch": 0.5721990561562087,
+      "grad_norm": 0.09054630994796753,
+      "learning_rate": 1.2866154346792161e-05,
+      "loss": 8.7797,
+      "step": 114580
+    },
+    {
+      "epoch": 0.5722489949811481,
+      "grad_norm": 0.09199676662683487,
+      "learning_rate": 1.286465243185061e-05,
+      "loss": 8.7763,
+      "step": 114590
+    },
+    {
+      "epoch": 0.5722989338060875,
+      "grad_norm": 0.09266217797994614,
+      "learning_rate": 1.286315051690906e-05,
+      "loss": 8.793,
+      "step": 114600
+    },
+    {
+      "epoch": 0.572348872631027,
+      "grad_norm": 0.09249533712863922,
+      "learning_rate": 1.2861648601967508e-05,
+      "loss": 8.7705,
+      "step": 114610
+    },
+    {
+      "epoch": 0.5723988114559665,
+      "grad_norm": 0.0948820635676384,
+      "learning_rate": 1.2860146687025959e-05,
+      "loss": 8.7812,
+      "step": 114620
+    },
+    {
+      "epoch": 0.5724487502809059,
+      "grad_norm": 0.08931874483823776,
+      "learning_rate": 1.2858644772084409e-05,
+      "loss": 8.782,
+      "step": 114630
+    },
+    {
+      "epoch": 0.5724986891058453,
+      "grad_norm": 0.09403346478939056,
+      "learning_rate": 1.2857142857142857e-05,
+      "loss": 8.7976,
+      "step": 114640
+    },
+    {
+      "epoch": 0.5725486279307848,
+      "grad_norm": 0.0919768437743187,
+      "learning_rate": 1.2855640942201307e-05,
+      "loss": 8.7873,
+      "step": 114650
+    },
+    {
+      "epoch": 0.5725985667557243,
+      "grad_norm": 0.09091559052467346,
+      "learning_rate": 1.2854139027259756e-05,
+      "loss": 8.8047,
+      "step": 114660
+    },
+    {
+      "epoch": 0.5726485055806637,
+      "grad_norm": 0.09633230417966843,
+      "learning_rate": 1.2852637112318206e-05,
+      "loss": 8.7847,
+      "step": 114670
+    },
+    {
+      "epoch": 0.5726984444056031,
+      "grad_norm": 0.0949721410870552,
+      "learning_rate": 1.2851135197376656e-05,
+      "loss": 8.8071,
+      "step": 114680
+    },
+    {
+      "epoch": 0.5727483832305426,
+      "grad_norm": 0.09185748547315598,
+      "learning_rate": 1.2849633282435105e-05,
+      "loss": 8.7908,
+      "step": 114690
+    },
+    {
+      "epoch": 0.5727983220554821,
+      "grad_norm": 0.0928587019443512,
+      "learning_rate": 1.2848131367493555e-05,
+      "loss": 8.787,
+      "step": 114700
+    },
+    {
+      "epoch": 0.5728482608804215,
+      "grad_norm": 0.09270650893449783,
+      "learning_rate": 1.2846629452552003e-05,
+      "loss": 8.7901,
+      "step": 114710
+    },
+    {
+      "epoch": 0.5728981997053609,
+      "grad_norm": 0.09454198181629181,
+      "learning_rate": 1.2845127537610454e-05,
+      "loss": 8.7726,
+      "step": 114720
+    },
+    {
+      "epoch": 0.5729481385303004,
+      "grad_norm": 0.09954354166984558,
+      "learning_rate": 1.2843625622668904e-05,
+      "loss": 8.7879,
+      "step": 114730
+    },
+    {
+      "epoch": 0.5729980773552398,
+      "grad_norm": 0.09377159178256989,
+      "learning_rate": 1.2842123707727354e-05,
+      "loss": 8.7744,
+      "step": 114740
+    },
+    {
+      "epoch": 0.5730480161801793,
+      "grad_norm": 0.08639419823884964,
+      "learning_rate": 1.2840621792785802e-05,
+      "loss": 8.7816,
+      "step": 114750
+    },
+    {
+      "epoch": 0.5730979550051187,
+      "grad_norm": 0.0929587110877037,
+      "learning_rate": 1.2839119877844251e-05,
+      "loss": 8.7803,
+      "step": 114760
+    },
+    {
+      "epoch": 0.5731478938300582,
+      "grad_norm": 0.09413690865039825,
+      "learning_rate": 1.2837617962902701e-05,
+      "loss": 8.7931,
+      "step": 114770
+    },
+    {
+      "epoch": 0.5731978326549976,
+      "grad_norm": 0.09625568985939026,
+      "learning_rate": 1.2836116047961151e-05,
+      "loss": 8.774,
+      "step": 114780
+    },
+    {
+      "epoch": 0.5732477714799371,
+      "grad_norm": 0.09371865540742874,
+      "learning_rate": 1.2834614133019601e-05,
+      "loss": 8.797,
+      "step": 114790
+    },
+    {
+      "epoch": 0.5732977103048765,
+      "grad_norm": 0.09570909291505814,
+      "learning_rate": 1.283311221807805e-05,
+      "loss": 8.7847,
+      "step": 114800
+    },
+    {
+      "epoch": 0.573347649129816,
+      "grad_norm": 0.09836754947900772,
+      "learning_rate": 1.2831610303136498e-05,
+      "loss": 8.7969,
+      "step": 114810
+    },
+    {
+      "epoch": 0.5733975879547554,
+      "grad_norm": 0.08818056434392929,
+      "learning_rate": 1.2830108388194949e-05,
+      "loss": 8.7888,
+      "step": 114820
+    },
+    {
+      "epoch": 0.5734475267796949,
+      "grad_norm": 0.09481636434793472,
+      "learning_rate": 1.2828606473253399e-05,
+      "loss": 8.7942,
+      "step": 114830
+    },
+    {
+      "epoch": 0.5734974656046343,
+      "grad_norm": 0.09127473831176758,
+      "learning_rate": 1.2827104558311849e-05,
+      "loss": 8.7794,
+      "step": 114840
+    },
+    {
+      "epoch": 0.5735474044295737,
+      "grad_norm": 0.08786597847938538,
+      "learning_rate": 1.2825602643370297e-05,
+      "loss": 8.7874,
+      "step": 114850
+    },
+    {
+      "epoch": 0.5735973432545132,
+      "grad_norm": 0.10287149995565414,
+      "learning_rate": 1.2824100728428746e-05,
+      "loss": 8.7911,
+      "step": 114860
+    },
+    {
+      "epoch": 0.5736472820794527,
+      "grad_norm": 0.09321471303701401,
+      "learning_rate": 1.2822598813487196e-05,
+      "loss": 8.7907,
+      "step": 114870
+    },
+    {
+      "epoch": 0.5736972209043921,
+      "grad_norm": 0.08879370242357254,
+      "learning_rate": 1.2821096898545646e-05,
+      "loss": 8.7877,
+      "step": 114880
+    },
+    {
+      "epoch": 0.5737471597293315,
+      "grad_norm": 0.0937252789735794,
+      "learning_rate": 1.2819594983604096e-05,
+      "loss": 8.7674,
+      "step": 114890
+    },
+    {
+      "epoch": 0.573797098554271,
+      "grad_norm": 0.09370169043540955,
+      "learning_rate": 1.2818093068662547e-05,
+      "loss": 8.7896,
+      "step": 114900
+    },
+    {
+      "epoch": 0.5738470373792105,
+      "grad_norm": 0.09470123052597046,
+      "learning_rate": 1.2816591153720993e-05,
+      "loss": 8.7953,
+      "step": 114910
+    },
+    {
+      "epoch": 0.5738969762041499,
+      "grad_norm": 0.09248079359531403,
+      "learning_rate": 1.2815089238779444e-05,
+      "loss": 8.79,
+      "step": 114920
+    },
+    {
+      "epoch": 0.5739469150290893,
+      "grad_norm": 0.09325503557920456,
+      "learning_rate": 1.2813587323837894e-05,
+      "loss": 8.7838,
+      "step": 114930
+    },
+    {
+      "epoch": 0.5739968538540288,
+      "grad_norm": 0.0931696742773056,
+      "learning_rate": 1.2812085408896344e-05,
+      "loss": 8.7742,
+      "step": 114940
+    },
+    {
+      "epoch": 0.5740467926789683,
+      "grad_norm": 0.09374930709600449,
+      "learning_rate": 1.2810583493954794e-05,
+      "loss": 8.7925,
+      "step": 114950
+    },
+    {
+      "epoch": 0.5740967315039077,
+      "grad_norm": 0.08893626183271408,
+      "learning_rate": 1.2809081579013241e-05,
+      "loss": 8.7882,
+      "step": 114960
+    },
+    {
+      "epoch": 0.5741466703288471,
+      "grad_norm": 0.08865047246217728,
+      "learning_rate": 1.2807579664071691e-05,
+      "loss": 8.7788,
+      "step": 114970
+    },
+    {
+      "epoch": 0.5741966091537866,
+      "grad_norm": 0.09421852231025696,
+      "learning_rate": 1.2806077749130141e-05,
+      "loss": 8.7768,
+      "step": 114980
+    },
+    {
+      "epoch": 0.5742465479787261,
+      "grad_norm": 0.08916622400283813,
+      "learning_rate": 1.2804575834188591e-05,
+      "loss": 8.785,
+      "step": 114990
+    },
+    {
+      "epoch": 0.5742964868036655,
+      "grad_norm": 0.08971542119979858,
+      "learning_rate": 1.2803073919247042e-05,
+      "loss": 8.7801,
+      "step": 115000
+    },
+    {
+      "epoch": 0.5743464256286049,
+      "grad_norm": 0.09803227335214615,
+      "learning_rate": 1.2801572004305488e-05,
+      "loss": 8.7843,
+      "step": 115010
+    },
+    {
+      "epoch": 0.5743963644535444,
+      "grad_norm": 0.0927937850356102,
+      "learning_rate": 1.2800070089363939e-05,
+      "loss": 8.7938,
+      "step": 115020
+    },
+    {
+      "epoch": 0.5744463032784839,
+      "grad_norm": 0.09266925603151321,
+      "learning_rate": 1.2798568174422389e-05,
+      "loss": 8.7783,
+      "step": 115030
+    },
+    {
+      "epoch": 0.5744962421034233,
+      "grad_norm": 0.09375444054603577,
+      "learning_rate": 1.2797066259480839e-05,
+      "loss": 8.7916,
+      "step": 115040
+    },
+    {
+      "epoch": 0.5745461809283627,
+      "grad_norm": 0.09235719591379166,
+      "learning_rate": 1.2795564344539289e-05,
+      "loss": 8.79,
+      "step": 115050
+    },
+    {
+      "epoch": 0.5745961197533022,
+      "grad_norm": 0.09506914019584656,
+      "learning_rate": 1.2794062429597738e-05,
+      "loss": 8.7758,
+      "step": 115060
+    },
+    {
+      "epoch": 0.5746460585782417,
+      "grad_norm": 0.09061712771654129,
+      "learning_rate": 1.2792560514656186e-05,
+      "loss": 8.785,
+      "step": 115070
+    },
+    {
+      "epoch": 0.5746959974031811,
+      "grad_norm": 0.09398726373910904,
+      "learning_rate": 1.2791058599714636e-05,
+      "loss": 8.7835,
+      "step": 115080
+    },
+    {
+      "epoch": 0.5747459362281205,
+      "grad_norm": 0.09454075247049332,
+      "learning_rate": 1.2789556684773086e-05,
+      "loss": 8.7984,
+      "step": 115090
+    },
+    {
+      "epoch": 0.57479587505306,
+      "grad_norm": 0.09011261910200119,
+      "learning_rate": 1.2788054769831537e-05,
+      "loss": 8.7811,
+      "step": 115100
+    },
+    {
+      "epoch": 0.5748458138779995,
+      "grad_norm": 0.09368349611759186,
+      "learning_rate": 1.2786552854889985e-05,
+      "loss": 8.7701,
+      "step": 115110
+    },
+    {
+      "epoch": 0.5748957527029389,
+      "grad_norm": 0.09060084819793701,
+      "learning_rate": 1.2785050939948434e-05,
+      "loss": 8.78,
+      "step": 115120
+    },
+    {
+      "epoch": 0.5749456915278783,
+      "grad_norm": 0.09300053119659424,
+      "learning_rate": 1.2783549025006884e-05,
+      "loss": 8.7671,
+      "step": 115130
+    },
+    {
+      "epoch": 0.5749956303528178,
+      "grad_norm": 0.09098996222019196,
+      "learning_rate": 1.2782047110065334e-05,
+      "loss": 8.7889,
+      "step": 115140
+    },
+    {
+      "epoch": 0.5750455691777573,
+      "grad_norm": 0.09513608366250992,
+      "learning_rate": 1.2780545195123784e-05,
+      "loss": 8.7721,
+      "step": 115150
+    },
+    {
+      "epoch": 0.5750955080026967,
+      "grad_norm": 0.09404340386390686,
+      "learning_rate": 1.2779043280182233e-05,
+      "loss": 8.7911,
+      "step": 115160
+    },
+    {
+      "epoch": 0.5751454468276361,
+      "grad_norm": 0.0925978347659111,
+      "learning_rate": 1.2777541365240681e-05,
+      "loss": 8.784,
+      "step": 115170
+    },
+    {
+      "epoch": 0.5751953856525756,
+      "grad_norm": 0.09595629572868347,
+      "learning_rate": 1.2776039450299131e-05,
+      "loss": 8.7783,
+      "step": 115180
+    },
+    {
+      "epoch": 0.5752453244775151,
+      "grad_norm": 0.09918168187141418,
+      "learning_rate": 1.2774537535357581e-05,
+      "loss": 8.7753,
+      "step": 115190
+    },
+    {
+      "epoch": 0.5752952633024545,
+      "grad_norm": 0.09095029532909393,
+      "learning_rate": 1.2773035620416032e-05,
+      "loss": 8.7743,
+      "step": 115200
+    },
+    {
+      "epoch": 0.5753452021273939,
+      "grad_norm": 0.09520848840475082,
+      "learning_rate": 1.277153370547448e-05,
+      "loss": 8.7882,
+      "step": 115210
+    },
+    {
+      "epoch": 0.5753951409523334,
+      "grad_norm": 0.0901302695274353,
+      "learning_rate": 1.277003179053293e-05,
+      "loss": 8.7796,
+      "step": 115220
+    },
+    {
+      "epoch": 0.5754450797772729,
+      "grad_norm": 0.08923668414354324,
+      "learning_rate": 1.2768529875591379e-05,
+      "loss": 8.7896,
+      "step": 115230
+    },
+    {
+      "epoch": 0.5754950186022123,
+      "grad_norm": 0.09207792580127716,
+      "learning_rate": 1.2767027960649829e-05,
+      "loss": 8.7796,
+      "step": 115240
+    },
+    {
+      "epoch": 0.5755449574271517,
+      "grad_norm": 0.09478235244750977,
+      "learning_rate": 1.2765526045708279e-05,
+      "loss": 8.7792,
+      "step": 115250
+    },
+    {
+      "epoch": 0.5755948962520911,
+      "grad_norm": 0.09395542740821838,
+      "learning_rate": 1.2764024130766728e-05,
+      "loss": 8.791,
+      "step": 115260
+    },
+    {
+      "epoch": 0.5756448350770307,
+      "grad_norm": 0.09286878257989883,
+      "learning_rate": 1.2762522215825178e-05,
+      "loss": 8.7711,
+      "step": 115270
+    },
+    {
+      "epoch": 0.5756947739019701,
+      "grad_norm": 0.09151309728622437,
+      "learning_rate": 1.2761020300883626e-05,
+      "loss": 8.785,
+      "step": 115280
+    },
+    {
+      "epoch": 0.5757447127269095,
+      "grad_norm": 0.09942897409200668,
+      "learning_rate": 1.2759518385942076e-05,
+      "loss": 8.7799,
+      "step": 115290
+    },
+    {
+      "epoch": 0.575794651551849,
+      "grad_norm": 0.09082534909248352,
+      "learning_rate": 1.2758016471000527e-05,
+      "loss": 8.7904,
+      "step": 115300
+    },
+    {
+      "epoch": 0.5758445903767885,
+      "grad_norm": 0.09576486796140671,
+      "learning_rate": 1.2756514556058975e-05,
+      "loss": 8.7816,
+      "step": 115310
+    },
+    {
+      "epoch": 0.5758945292017279,
+      "grad_norm": 0.09835550934076309,
+      "learning_rate": 1.2755012641117425e-05,
+      "loss": 8.7852,
+      "step": 115320
+    },
+    {
+      "epoch": 0.5759444680266673,
+      "grad_norm": 0.09621113538742065,
+      "learning_rate": 1.2753510726175874e-05,
+      "loss": 8.7841,
+      "step": 115330
+    },
+    {
+      "epoch": 0.5759944068516067,
+      "grad_norm": 0.09935037046670914,
+      "learning_rate": 1.2752008811234324e-05,
+      "loss": 8.7874,
+      "step": 115340
+    },
+    {
+      "epoch": 0.5760443456765463,
+      "grad_norm": 0.09516027569770813,
+      "learning_rate": 1.2750506896292774e-05,
+      "loss": 8.7866,
+      "step": 115350
+    },
+    {
+      "epoch": 0.5760942845014857,
+      "grad_norm": 0.09592787176370621,
+      "learning_rate": 1.2749004981351223e-05,
+      "loss": 8.7759,
+      "step": 115360
+    },
+    {
+      "epoch": 0.5761442233264251,
+      "grad_norm": 0.09324520081281662,
+      "learning_rate": 1.2747503066409673e-05,
+      "loss": 8.7756,
+      "step": 115370
+    },
+    {
+      "epoch": 0.5761941621513645,
+      "grad_norm": 0.09475477784872055,
+      "learning_rate": 1.2746001151468123e-05,
+      "loss": 8.7751,
+      "step": 115380
+    },
+    {
+      "epoch": 0.5762441009763041,
+      "grad_norm": 0.09500320255756378,
+      "learning_rate": 1.2744499236526571e-05,
+      "loss": 8.7876,
+      "step": 115390
+    },
+    {
+      "epoch": 0.5762940398012435,
+      "grad_norm": 0.09369548410177231,
+      "learning_rate": 1.2742997321585022e-05,
+      "loss": 8.7883,
+      "step": 115400
+    },
+    {
+      "epoch": 0.5763439786261829,
+      "grad_norm": 0.0879288762807846,
+      "learning_rate": 1.274149540664347e-05,
+      "loss": 8.7739,
+      "step": 115410
+    },
+    {
+      "epoch": 0.5763939174511223,
+      "grad_norm": 0.09155362099409103,
+      "learning_rate": 1.273999349170192e-05,
+      "loss": 8.7865,
+      "step": 115420
+    },
+    {
+      "epoch": 0.5764438562760619,
+      "grad_norm": 0.09589493274688721,
+      "learning_rate": 1.273849157676037e-05,
+      "loss": 8.7779,
+      "step": 115430
+    },
+    {
+      "epoch": 0.5764937951010013,
+      "grad_norm": 0.09291618317365646,
+      "learning_rate": 1.2736989661818819e-05,
+      "loss": 8.7973,
+      "step": 115440
+    },
+    {
+      "epoch": 0.5765437339259407,
+      "grad_norm": 0.08438339829444885,
+      "learning_rate": 1.273548774687727e-05,
+      "loss": 8.779,
+      "step": 115450
+    },
+    {
+      "epoch": 0.5765936727508801,
+      "grad_norm": 0.09597896039485931,
+      "learning_rate": 1.2733985831935718e-05,
+      "loss": 8.7717,
+      "step": 115460
+    },
+    {
+      "epoch": 0.5766436115758197,
+      "grad_norm": 0.09120340645313263,
+      "learning_rate": 1.2732483916994168e-05,
+      "loss": 8.7793,
+      "step": 115470
+    },
+    {
+      "epoch": 0.5766935504007591,
+      "grad_norm": 0.09236761182546616,
+      "learning_rate": 1.2730982002052618e-05,
+      "loss": 8.7749,
+      "step": 115480
+    },
+    {
+      "epoch": 0.5767434892256985,
+      "grad_norm": 0.09631557762622833,
+      "learning_rate": 1.2729480087111066e-05,
+      "loss": 8.7847,
+      "step": 115490
+    },
+    {
+      "epoch": 0.5767934280506379,
+      "grad_norm": 0.09939119219779968,
+      "learning_rate": 1.2727978172169517e-05,
+      "loss": 8.7732,
+      "step": 115500
+    },
+    {
+      "epoch": 0.5768433668755775,
+      "grad_norm": 0.08828873187303543,
+      "learning_rate": 1.2726476257227965e-05,
+      "loss": 8.7684,
+      "step": 115510
+    },
+    {
+      "epoch": 0.5768933057005169,
+      "grad_norm": 0.09669400006532669,
+      "learning_rate": 1.2724974342286415e-05,
+      "loss": 8.782,
+      "step": 115520
+    },
+    {
+      "epoch": 0.5769432445254563,
+      "grad_norm": 0.08688271045684814,
+      "learning_rate": 1.2723472427344866e-05,
+      "loss": 8.7748,
+      "step": 115530
+    },
+    {
+      "epoch": 0.5769931833503957,
+      "grad_norm": 0.09424971789121628,
+      "learning_rate": 1.2721970512403316e-05,
+      "loss": 8.7726,
+      "step": 115540
+    },
+    {
+      "epoch": 0.5770431221753353,
+      "grad_norm": 0.09156131744384766,
+      "learning_rate": 1.2720468597461764e-05,
+      "loss": 8.7737,
+      "step": 115550
+    },
+    {
+      "epoch": 0.5770930610002747,
+      "grad_norm": 0.09321281313896179,
+      "learning_rate": 1.2718966682520213e-05,
+      "loss": 8.7779,
+      "step": 115560
+    },
+    {
+      "epoch": 0.5771429998252141,
+      "grad_norm": 0.09447400271892548,
+      "learning_rate": 1.2717464767578663e-05,
+      "loss": 8.7736,
+      "step": 115570
+    },
+    {
+      "epoch": 0.5771929386501535,
+      "grad_norm": 0.09802529215812683,
+      "learning_rate": 1.2715962852637113e-05,
+      "loss": 8.7908,
+      "step": 115580
+    },
+    {
+      "epoch": 0.5772428774750931,
+      "grad_norm": 0.09059146791696548,
+      "learning_rate": 1.2714460937695563e-05,
+      "loss": 8.7868,
+      "step": 115590
+    },
+    {
+      "epoch": 0.5772928163000325,
+      "grad_norm": 0.09354998171329498,
+      "learning_rate": 1.2712959022754012e-05,
+      "loss": 8.7936,
+      "step": 115600
+    },
+    {
+      "epoch": 0.5773427551249719,
+      "grad_norm": 0.09096765518188477,
+      "learning_rate": 1.271145710781246e-05,
+      "loss": 8.7762,
+      "step": 115610
+    },
+    {
+      "epoch": 0.5773926939499113,
+      "grad_norm": 0.09621896594762802,
+      "learning_rate": 1.270995519287091e-05,
+      "loss": 8.7725,
+      "step": 115620
+    },
+    {
+      "epoch": 0.5774426327748509,
+      "grad_norm": 0.09030432254076004,
+      "learning_rate": 1.270845327792936e-05,
+      "loss": 8.7869,
+      "step": 115630
+    },
+    {
+      "epoch": 0.5774925715997903,
+      "grad_norm": 0.0910486951470375,
+      "learning_rate": 1.270695136298781e-05,
+      "loss": 8.7925,
+      "step": 115640
+    },
+    {
+      "epoch": 0.5775425104247297,
+      "grad_norm": 0.09086786955595016,
+      "learning_rate": 1.270544944804626e-05,
+      "loss": 8.7813,
+      "step": 115650
+    },
+    {
+      "epoch": 0.5775924492496691,
+      "grad_norm": 0.08619526028633118,
+      "learning_rate": 1.2703947533104708e-05,
+      "loss": 8.7701,
+      "step": 115660
+    },
+    {
+      "epoch": 0.5776423880746087,
+      "grad_norm": 0.09702730923891068,
+      "learning_rate": 1.2702445618163158e-05,
+      "loss": 8.78,
+      "step": 115670
+    },
+    {
+      "epoch": 0.5776923268995481,
+      "grad_norm": 0.0912908986210823,
+      "learning_rate": 1.2700943703221608e-05,
+      "loss": 8.7743,
+      "step": 115680
+    },
+    {
+      "epoch": 0.5777422657244875,
+      "grad_norm": 0.09520822018384933,
+      "learning_rate": 1.2699441788280058e-05,
+      "loss": 8.7819,
+      "step": 115690
+    },
+    {
+      "epoch": 0.5777922045494269,
+      "grad_norm": 0.0918937474489212,
+      "learning_rate": 1.2697939873338508e-05,
+      "loss": 8.7656,
+      "step": 115700
+    },
+    {
+      "epoch": 0.5778421433743663,
+      "grad_norm": 0.09397950768470764,
+      "learning_rate": 1.2696437958396955e-05,
+      "loss": 8.7783,
+      "step": 115710
+    },
+    {
+      "epoch": 0.5778920821993059,
+      "grad_norm": 0.09476295858621597,
+      "learning_rate": 1.2694936043455405e-05,
+      "loss": 8.7867,
+      "step": 115720
+    },
+    {
+      "epoch": 0.5779420210242453,
+      "grad_norm": 0.09248529374599457,
+      "learning_rate": 1.2693434128513856e-05,
+      "loss": 8.7845,
+      "step": 115730
+    },
+    {
+      "epoch": 0.5779919598491847,
+      "grad_norm": 0.0914454385638237,
+      "learning_rate": 1.2691932213572306e-05,
+      "loss": 8.7718,
+      "step": 115740
+    },
+    {
+      "epoch": 0.5780418986741241,
+      "grad_norm": 0.09722574055194855,
+      "learning_rate": 1.2690430298630756e-05,
+      "loss": 8.7653,
+      "step": 115750
+    },
+    {
+      "epoch": 0.5780918374990637,
+      "grad_norm": 0.09682019799947739,
+      "learning_rate": 1.2688928383689203e-05,
+      "loss": 8.7623,
+      "step": 115760
+    },
+    {
+      "epoch": 0.5781417763240031,
+      "grad_norm": 0.08837232738733292,
+      "learning_rate": 1.2687426468747653e-05,
+      "loss": 8.78,
+      "step": 115770
+    },
+    {
+      "epoch": 0.5781917151489425,
+      "grad_norm": 0.0944056510925293,
+      "learning_rate": 1.2685924553806103e-05,
+      "loss": 8.7849,
+      "step": 115780
+    },
+    {
+      "epoch": 0.5782416539738819,
+      "grad_norm": 0.0868343710899353,
+      "learning_rate": 1.2684422638864553e-05,
+      "loss": 8.7678,
+      "step": 115790
+    },
+    {
+      "epoch": 0.5782915927988215,
+      "grad_norm": 0.09442039579153061,
+      "learning_rate": 1.2682920723923003e-05,
+      "loss": 8.7778,
+      "step": 115800
+    },
+    {
+      "epoch": 0.5783415316237609,
+      "grad_norm": 0.08963579684495926,
+      "learning_rate": 1.268141880898145e-05,
+      "loss": 8.7754,
+      "step": 115810
+    },
+    {
+      "epoch": 0.5783914704487003,
+      "grad_norm": 0.09150784462690353,
+      "learning_rate": 1.26799168940399e-05,
+      "loss": 8.7636,
+      "step": 115820
+    },
+    {
+      "epoch": 0.5784414092736397,
+      "grad_norm": 0.09078077971935272,
+      "learning_rate": 1.267841497909835e-05,
+      "loss": 8.7658,
+      "step": 115830
+    },
+    {
+      "epoch": 0.5784913480985793,
+      "grad_norm": 0.09242097288370132,
+      "learning_rate": 1.26769130641568e-05,
+      "loss": 8.7743,
+      "step": 115840
+    },
+    {
+      "epoch": 0.5785412869235187,
+      "grad_norm": 0.09405604004859924,
+      "learning_rate": 1.2675411149215251e-05,
+      "loss": 8.7668,
+      "step": 115850
+    },
+    {
+      "epoch": 0.5785912257484581,
+      "grad_norm": 0.09389903396368027,
+      "learning_rate": 1.26739092342737e-05,
+      "loss": 8.7732,
+      "step": 115860
+    },
+    {
+      "epoch": 0.5786411645733975,
+      "grad_norm": 0.08914590626955032,
+      "learning_rate": 1.2672407319332148e-05,
+      "loss": 8.7808,
+      "step": 115870
+    },
+    {
+      "epoch": 0.5786911033983371,
+      "grad_norm": 0.09185650944709778,
+      "learning_rate": 1.2670905404390598e-05,
+      "loss": 8.779,
+      "step": 115880
+    },
+    {
+      "epoch": 0.5787410422232765,
+      "grad_norm": 0.091721311211586,
+      "learning_rate": 1.2669403489449048e-05,
+      "loss": 8.778,
+      "step": 115890
+    },
+    {
+      "epoch": 0.5787909810482159,
+      "grad_norm": 0.08753747493028641,
+      "learning_rate": 1.2667901574507498e-05,
+      "loss": 8.7874,
+      "step": 115900
+    },
+    {
+      "epoch": 0.5788409198731553,
+      "grad_norm": 0.09685467183589935,
+      "learning_rate": 1.2666399659565947e-05,
+      "loss": 8.7786,
+      "step": 115910
+    },
+    {
+      "epoch": 0.5788908586980949,
+      "grad_norm": 0.09479455649852753,
+      "learning_rate": 1.2664897744624395e-05,
+      "loss": 8.7859,
+      "step": 115920
+    },
+    {
+      "epoch": 0.5789407975230343,
+      "grad_norm": 0.09902866929769516,
+      "learning_rate": 1.2663395829682846e-05,
+      "loss": 8.7763,
+      "step": 115930
+    },
+    {
+      "epoch": 0.5789907363479737,
+      "grad_norm": 0.09214528650045395,
+      "learning_rate": 1.2661893914741296e-05,
+      "loss": 8.7802,
+      "step": 115940
+    },
+    {
+      "epoch": 0.5790406751729131,
+      "grad_norm": 0.09406938403844833,
+      "learning_rate": 1.2660391999799746e-05,
+      "loss": 8.7633,
+      "step": 115950
+    },
+    {
+      "epoch": 0.5790906139978527,
+      "grad_norm": 0.0909062996506691,
+      "learning_rate": 1.2658890084858194e-05,
+      "loss": 8.7809,
+      "step": 115960
+    },
+    {
+      "epoch": 0.5791405528227921,
+      "grad_norm": 0.08931753784418106,
+      "learning_rate": 1.2657388169916643e-05,
+      "loss": 8.781,
+      "step": 115970
+    },
+    {
+      "epoch": 0.5791904916477315,
+      "grad_norm": 0.100962333381176,
+      "learning_rate": 1.2655886254975093e-05,
+      "loss": 8.7796,
+      "step": 115980
+    },
+    {
+      "epoch": 0.5792404304726709,
+      "grad_norm": 0.09826654195785522,
+      "learning_rate": 1.2654384340033543e-05,
+      "loss": 8.7885,
+      "step": 115990
+    },
+    {
+      "epoch": 0.5792903692976105,
+      "grad_norm": 0.09005758911371231,
+      "learning_rate": 1.2652882425091993e-05,
+      "loss": 8.7879,
+      "step": 116000
+    },
+    {
+      "epoch": 0.5793403081225499,
+      "grad_norm": 0.09193582087755203,
+      "learning_rate": 1.2651380510150442e-05,
+      "loss": 8.7698,
+      "step": 116010
+    },
+    {
+      "epoch": 0.5793902469474893,
+      "grad_norm": 0.09053417295217514,
+      "learning_rate": 1.2649878595208892e-05,
+      "loss": 8.7836,
+      "step": 116020
+    },
+    {
+      "epoch": 0.5794401857724287,
+      "grad_norm": 0.09887261688709259,
+      "learning_rate": 1.264837668026734e-05,
+      "loss": 8.7667,
+      "step": 116030
+    },
+    {
+      "epoch": 0.5794901245973683,
+      "grad_norm": 0.09462026506662369,
+      "learning_rate": 1.264687476532579e-05,
+      "loss": 8.7674,
+      "step": 116040
+    },
+    {
+      "epoch": 0.5795400634223077,
+      "grad_norm": 0.0957246646285057,
+      "learning_rate": 1.2645372850384241e-05,
+      "loss": 8.7798,
+      "step": 116050
+    },
+    {
+      "epoch": 0.5795900022472471,
+      "grad_norm": 0.10332443565130234,
+      "learning_rate": 1.264387093544269e-05,
+      "loss": 8.7828,
+      "step": 116060
+    },
+    {
+      "epoch": 0.5796399410721865,
+      "grad_norm": 0.08929960429668427,
+      "learning_rate": 1.264236902050114e-05,
+      "loss": 8.7821,
+      "step": 116070
+    },
+    {
+      "epoch": 0.5796898798971261,
+      "grad_norm": 0.09699279814958572,
+      "learning_rate": 1.2640867105559588e-05,
+      "loss": 8.7763,
+      "step": 116080
+    },
+    {
+      "epoch": 0.5797398187220655,
+      "grad_norm": 0.0911206379532814,
+      "learning_rate": 1.2639365190618038e-05,
+      "loss": 8.7705,
+      "step": 116090
+    },
+    {
+      "epoch": 0.5797897575470049,
+      "grad_norm": 0.09146827459335327,
+      "learning_rate": 1.2637863275676488e-05,
+      "loss": 8.7882,
+      "step": 116100
+    },
+    {
+      "epoch": 0.5798396963719443,
+      "grad_norm": 0.09372538328170776,
+      "learning_rate": 1.2636361360734937e-05,
+      "loss": 8.7599,
+      "step": 116110
+    },
+    {
+      "epoch": 0.5798896351968839,
+      "grad_norm": 0.09762945026159286,
+      "learning_rate": 1.2634859445793387e-05,
+      "loss": 8.7786,
+      "step": 116120
+    },
+    {
+      "epoch": 0.5799395740218233,
+      "grad_norm": 0.09360945969820023,
+      "learning_rate": 1.2633357530851836e-05,
+      "loss": 8.7724,
+      "step": 116130
+    },
+    {
+      "epoch": 0.5799895128467627,
+      "grad_norm": 0.09288088977336884,
+      "learning_rate": 1.2631855615910286e-05,
+      "loss": 8.7698,
+      "step": 116140
+    },
+    {
+      "epoch": 0.5800394516717021,
+      "grad_norm": 0.09539160132408142,
+      "learning_rate": 1.2630353700968736e-05,
+      "loss": 8.7872,
+      "step": 116150
+    },
+    {
+      "epoch": 0.5800893904966417,
+      "grad_norm": 0.08673470467329025,
+      "learning_rate": 1.2628851786027184e-05,
+      "loss": 8.7901,
+      "step": 116160
+    },
+    {
+      "epoch": 0.5801393293215811,
+      "grad_norm": 0.09566935151815414,
+      "learning_rate": 1.2627349871085635e-05,
+      "loss": 8.7816,
+      "step": 116170
+    },
+    {
+      "epoch": 0.5801892681465205,
+      "grad_norm": 0.09150587022304535,
+      "learning_rate": 1.2625847956144085e-05,
+      "loss": 8.7651,
+      "step": 116180
+    },
+    {
+      "epoch": 0.5802392069714599,
+      "grad_norm": 0.08590695261955261,
+      "learning_rate": 1.2624346041202533e-05,
+      "loss": 8.7845,
+      "step": 116190
+    },
+    {
+      "epoch": 0.5802891457963995,
+      "grad_norm": 0.09772205352783203,
+      "learning_rate": 1.2622844126260983e-05,
+      "loss": 8.7651,
+      "step": 116200
+    },
+    {
+      "epoch": 0.5803390846213389,
+      "grad_norm": 0.09127950668334961,
+      "learning_rate": 1.2621342211319432e-05,
+      "loss": 8.7905,
+      "step": 116210
+    },
+    {
+      "epoch": 0.5803890234462783,
+      "grad_norm": 0.09437844902276993,
+      "learning_rate": 1.2619840296377882e-05,
+      "loss": 8.7782,
+      "step": 116220
+    },
+    {
+      "epoch": 0.5804389622712177,
+      "grad_norm": 0.08986880630254745,
+      "learning_rate": 1.2618338381436332e-05,
+      "loss": 8.7814,
+      "step": 116230
+    },
+    {
+      "epoch": 0.5804889010961573,
+      "grad_norm": 0.09137626737356186,
+      "learning_rate": 1.261683646649478e-05,
+      "loss": 8.7681,
+      "step": 116240
+    },
+    {
+      "epoch": 0.5805388399210967,
+      "grad_norm": 0.08857541531324387,
+      "learning_rate": 1.2615334551553231e-05,
+      "loss": 8.776,
+      "step": 116250
+    },
+    {
+      "epoch": 0.5805887787460361,
+      "grad_norm": 0.0896126851439476,
+      "learning_rate": 1.261383263661168e-05,
+      "loss": 8.7803,
+      "step": 116260
+    },
+    {
+      "epoch": 0.5806387175709755,
+      "grad_norm": 0.08985196799039841,
+      "learning_rate": 1.261233072167013e-05,
+      "loss": 8.7829,
+      "step": 116270
+    },
+    {
+      "epoch": 0.580688656395915,
+      "grad_norm": 0.09273276478052139,
+      "learning_rate": 1.261082880672858e-05,
+      "loss": 8.7826,
+      "step": 116280
+    },
+    {
+      "epoch": 0.5807385952208545,
+      "grad_norm": 0.08667043596506119,
+      "learning_rate": 1.2609326891787028e-05,
+      "loss": 8.7681,
+      "step": 116290
+    },
+    {
+      "epoch": 0.5807885340457939,
+      "grad_norm": 0.09057953208684921,
+      "learning_rate": 1.2607824976845478e-05,
+      "loss": 8.7694,
+      "step": 116300
+    },
+    {
+      "epoch": 0.5808384728707333,
+      "grad_norm": 0.09501492232084274,
+      "learning_rate": 1.2606323061903927e-05,
+      "loss": 8.7712,
+      "step": 116310
+    },
+    {
+      "epoch": 0.5808884116956728,
+      "grad_norm": 0.08906516432762146,
+      "learning_rate": 1.2604821146962377e-05,
+      "loss": 8.788,
+      "step": 116320
+    },
+    {
+      "epoch": 0.5809383505206123,
+      "grad_norm": 0.09163739532232285,
+      "learning_rate": 1.2603319232020827e-05,
+      "loss": 8.7738,
+      "step": 116330
+    },
+    {
+      "epoch": 0.5809882893455517,
+      "grad_norm": 0.08950494229793549,
+      "learning_rate": 1.2601817317079277e-05,
+      "loss": 8.7851,
+      "step": 116340
+    },
+    {
+      "epoch": 0.5810382281704911,
+      "grad_norm": 0.09091368317604065,
+      "learning_rate": 1.2600315402137726e-05,
+      "loss": 8.7618,
+      "step": 116350
+    },
+    {
+      "epoch": 0.5810881669954306,
+      "grad_norm": 0.09785959124565125,
+      "learning_rate": 1.2598813487196174e-05,
+      "loss": 8.7833,
+      "step": 116360
+    },
+    {
+      "epoch": 0.5811381058203701,
+      "grad_norm": 0.09358986467123032,
+      "learning_rate": 1.2597311572254625e-05,
+      "loss": 8.7742,
+      "step": 116370
+    },
+    {
+      "epoch": 0.5811880446453095,
+      "grad_norm": 0.09452328830957413,
+      "learning_rate": 1.2595809657313075e-05,
+      "loss": 8.7679,
+      "step": 116380
+    },
+    {
+      "epoch": 0.5812379834702489,
+      "grad_norm": 0.09855862706899643,
+      "learning_rate": 1.2594307742371525e-05,
+      "loss": 8.7724,
+      "step": 116390
+    },
+    {
+      "epoch": 0.5812879222951884,
+      "grad_norm": 0.09987233579158783,
+      "learning_rate": 1.2592805827429973e-05,
+      "loss": 8.7737,
+      "step": 116400
+    },
+    {
+      "epoch": 0.5813378611201279,
+      "grad_norm": 0.09587401151657104,
+      "learning_rate": 1.2591303912488422e-05,
+      "loss": 8.7721,
+      "step": 116410
+    },
+    {
+      "epoch": 0.5813877999450673,
+      "grad_norm": 0.08716830611228943,
+      "learning_rate": 1.2589801997546872e-05,
+      "loss": 8.7768,
+      "step": 116420
+    },
+    {
+      "epoch": 0.5814377387700067,
+      "grad_norm": 0.09757670015096664,
+      "learning_rate": 1.2588300082605322e-05,
+      "loss": 8.7556,
+      "step": 116430
+    },
+    {
+      "epoch": 0.5814876775949462,
+      "grad_norm": 0.09444671124219894,
+      "learning_rate": 1.2586798167663772e-05,
+      "loss": 8.763,
+      "step": 116440
+    },
+    {
+      "epoch": 0.5815376164198857,
+      "grad_norm": 0.091017946600914,
+      "learning_rate": 1.2585296252722221e-05,
+      "loss": 8.7753,
+      "step": 116450
+    },
+    {
+      "epoch": 0.5815875552448251,
+      "grad_norm": 0.09647946804761887,
+      "learning_rate": 1.258379433778067e-05,
+      "loss": 8.7778,
+      "step": 116460
+    },
+    {
+      "epoch": 0.5816374940697645,
+      "grad_norm": 0.09835417568683624,
+      "learning_rate": 1.258229242283912e-05,
+      "loss": 8.7821,
+      "step": 116470
+    },
+    {
+      "epoch": 0.581687432894704,
+      "grad_norm": 0.09176066517829895,
+      "learning_rate": 1.258079050789757e-05,
+      "loss": 8.7742,
+      "step": 116480
+    },
+    {
+      "epoch": 0.5817373717196435,
+      "grad_norm": 0.09349832683801651,
+      "learning_rate": 1.257928859295602e-05,
+      "loss": 8.7629,
+      "step": 116490
+    },
+    {
+      "epoch": 0.5817873105445829,
+      "grad_norm": 0.0905727818608284,
+      "learning_rate": 1.257778667801447e-05,
+      "loss": 8.7742,
+      "step": 116500
+    },
+    {
+      "epoch": 0.5818372493695223,
+      "grad_norm": 0.09379737824201584,
+      "learning_rate": 1.2576284763072917e-05,
+      "loss": 8.7643,
+      "step": 116510
+    },
+    {
+      "epoch": 0.5818871881944618,
+      "grad_norm": 0.09791654348373413,
+      "learning_rate": 1.2574782848131367e-05,
+      "loss": 8.7811,
+      "step": 116520
+    },
+    {
+      "epoch": 0.5819371270194013,
+      "grad_norm": 0.09698036313056946,
+      "learning_rate": 1.2573280933189817e-05,
+      "loss": 8.7858,
+      "step": 116530
+    },
+    {
+      "epoch": 0.5819870658443407,
+      "grad_norm": 0.09158887714147568,
+      "learning_rate": 1.2571779018248267e-05,
+      "loss": 8.7815,
+      "step": 116540
+    },
+    {
+      "epoch": 0.5820370046692801,
+      "grad_norm": 0.08619420230388641,
+      "learning_rate": 1.2570277103306718e-05,
+      "loss": 8.7721,
+      "step": 116550
+    },
+    {
+      "epoch": 0.5820869434942196,
+      "grad_norm": 0.09012347459793091,
+      "learning_rate": 1.2568775188365164e-05,
+      "loss": 8.7651,
+      "step": 116560
+    },
+    {
+      "epoch": 0.582136882319159,
+      "grad_norm": 0.0954323559999466,
+      "learning_rate": 1.2567273273423615e-05,
+      "loss": 8.7664,
+      "step": 116570
+    },
+    {
+      "epoch": 0.5821868211440985,
+      "grad_norm": 0.09117588400840759,
+      "learning_rate": 1.2565771358482065e-05,
+      "loss": 8.7764,
+      "step": 116580
+    },
+    {
+      "epoch": 0.5822367599690379,
+      "grad_norm": 0.09088102728128433,
+      "learning_rate": 1.2564269443540515e-05,
+      "loss": 8.7799,
+      "step": 116590
+    },
+    {
+      "epoch": 0.5822866987939774,
+      "grad_norm": 0.09504615515470505,
+      "learning_rate": 1.2562767528598965e-05,
+      "loss": 8.7671,
+      "step": 116600
+    },
+    {
+      "epoch": 0.5823366376189169,
+      "grad_norm": 0.09126445651054382,
+      "learning_rate": 1.2561265613657414e-05,
+      "loss": 8.7732,
+      "step": 116610
+    },
+    {
+      "epoch": 0.5823865764438563,
+      "grad_norm": 0.0928245261311531,
+      "learning_rate": 1.2559763698715862e-05,
+      "loss": 8.7712,
+      "step": 116620
+    },
+    {
+      "epoch": 0.5824365152687957,
+      "grad_norm": 0.09340432286262512,
+      "learning_rate": 1.2558261783774312e-05,
+      "loss": 8.7847,
+      "step": 116630
+    },
+    {
+      "epoch": 0.5824864540937352,
+      "grad_norm": 0.09035646170377731,
+      "learning_rate": 1.2556759868832762e-05,
+      "loss": 8.7678,
+      "step": 116640
+    },
+    {
+      "epoch": 0.5825363929186747,
+      "grad_norm": 0.09360393136739731,
+      "learning_rate": 1.2555257953891213e-05,
+      "loss": 8.7652,
+      "step": 116650
+    },
+    {
+      "epoch": 0.5825863317436141,
+      "grad_norm": 0.09272781759500504,
+      "learning_rate": 1.2553756038949661e-05,
+      "loss": 8.7708,
+      "step": 116660
+    },
+    {
+      "epoch": 0.5826362705685535,
+      "grad_norm": 0.09432695060968399,
+      "learning_rate": 1.255225412400811e-05,
+      "loss": 8.7648,
+      "step": 116670
+    },
+    {
+      "epoch": 0.582686209393493,
+      "grad_norm": 0.10040241479873657,
+      "learning_rate": 1.255075220906656e-05,
+      "loss": 8.7673,
+      "step": 116680
+    },
+    {
+      "epoch": 0.5827361482184324,
+      "grad_norm": 0.09128203243017197,
+      "learning_rate": 1.254925029412501e-05,
+      "loss": 8.7639,
+      "step": 116690
+    },
+    {
+      "epoch": 0.5827860870433719,
+      "grad_norm": 0.08589877933263779,
+      "learning_rate": 1.254774837918346e-05,
+      "loss": 8.7859,
+      "step": 116700
+    },
+    {
+      "epoch": 0.5828360258683113,
+      "grad_norm": 0.09354274719953537,
+      "learning_rate": 1.2546246464241909e-05,
+      "loss": 8.7649,
+      "step": 116710
+    },
+    {
+      "epoch": 0.5828859646932507,
+      "grad_norm": 0.09442020207643509,
+      "learning_rate": 1.2544744549300357e-05,
+      "loss": 8.7609,
+      "step": 116720
+    },
+    {
+      "epoch": 0.5829359035181902,
+      "grad_norm": 0.0930701419711113,
+      "learning_rate": 1.2543242634358807e-05,
+      "loss": 8.7645,
+      "step": 116730
+    },
+    {
+      "epoch": 0.5829858423431297,
+      "grad_norm": 0.09396909922361374,
+      "learning_rate": 1.2541740719417257e-05,
+      "loss": 8.7725,
+      "step": 116740
+    },
+    {
+      "epoch": 0.5830357811680691,
+      "grad_norm": 0.09508174657821655,
+      "learning_rate": 1.2540238804475708e-05,
+      "loss": 8.7659,
+      "step": 116750
+    },
+    {
+      "epoch": 0.5830857199930085,
+      "grad_norm": 0.09655360132455826,
+      "learning_rate": 1.2538736889534158e-05,
+      "loss": 8.7918,
+      "step": 116760
+    },
+    {
+      "epoch": 0.583135658817948,
+      "grad_norm": 0.09458701312541962,
+      "learning_rate": 1.2537234974592605e-05,
+      "loss": 8.7626,
+      "step": 116770
+    },
+    {
+      "epoch": 0.5831855976428875,
+      "grad_norm": 0.09564093500375748,
+      "learning_rate": 1.2535733059651055e-05,
+      "loss": 8.77,
+      "step": 116780
+    },
+    {
+      "epoch": 0.5832355364678269,
+      "grad_norm": 0.09138777107000351,
+      "learning_rate": 1.2534231144709505e-05,
+      "loss": 8.7656,
+      "step": 116790
+    },
+    {
+      "epoch": 0.5832854752927663,
+      "grad_norm": 0.09551417827606201,
+      "learning_rate": 1.2532729229767955e-05,
+      "loss": 8.755,
+      "step": 116800
+    },
+    {
+      "epoch": 0.5833354141177058,
+      "grad_norm": 0.09259354323148727,
+      "learning_rate": 1.2531227314826405e-05,
+      "loss": 8.7714,
+      "step": 116810
+    },
+    {
+      "epoch": 0.5833853529426453,
+      "grad_norm": 0.09137726575136185,
+      "learning_rate": 1.2529725399884854e-05,
+      "loss": 8.7712,
+      "step": 116820
+    },
+    {
+      "epoch": 0.5834352917675847,
+      "grad_norm": 0.08896147459745407,
+      "learning_rate": 1.2528223484943302e-05,
+      "loss": 8.7618,
+      "step": 116830
+    },
+    {
+      "epoch": 0.5834852305925241,
+      "grad_norm": 0.09602434188127518,
+      "learning_rate": 1.2526721570001752e-05,
+      "loss": 8.7643,
+      "step": 116840
+    },
+    {
+      "epoch": 0.5835351694174636,
+      "grad_norm": 0.0915607139468193,
+      "learning_rate": 1.2525219655060203e-05,
+      "loss": 8.7753,
+      "step": 116850
+    },
+    {
+      "epoch": 0.5835851082424031,
+      "grad_norm": 0.09016492962837219,
+      "learning_rate": 1.2523717740118653e-05,
+      "loss": 8.7609,
+      "step": 116860
+    },
+    {
+      "epoch": 0.5836350470673425,
+      "grad_norm": 0.08857151120901108,
+      "learning_rate": 1.2522215825177101e-05,
+      "loss": 8.7689,
+      "step": 116870
+    },
+    {
+      "epoch": 0.5836849858922819,
+      "grad_norm": 0.09395987540483475,
+      "learning_rate": 1.252071391023555e-05,
+      "loss": 8.765,
+      "step": 116880
+    },
+    {
+      "epoch": 0.5837349247172214,
+      "grad_norm": 0.09694436937570572,
+      "learning_rate": 1.2519211995294e-05,
+      "loss": 8.7747,
+      "step": 116890
+    },
+    {
+      "epoch": 0.5837848635421609,
+      "grad_norm": 0.0934179425239563,
+      "learning_rate": 1.251771008035245e-05,
+      "loss": 8.7531,
+      "step": 116900
+    },
+    {
+      "epoch": 0.5838348023671003,
+      "grad_norm": 0.08705223351716995,
+      "learning_rate": 1.25162081654109e-05,
+      "loss": 8.7724,
+      "step": 116910
+    },
+    {
+      "epoch": 0.5838847411920397,
+      "grad_norm": 0.09162668138742447,
+      "learning_rate": 1.2514706250469349e-05,
+      "loss": 8.7788,
+      "step": 116920
+    },
+    {
+      "epoch": 0.5839346800169792,
+      "grad_norm": 0.08836338669061661,
+      "learning_rate": 1.2513204335527797e-05,
+      "loss": 8.7742,
+      "step": 116930
+    },
+    {
+      "epoch": 0.5839846188419187,
+      "grad_norm": 0.08714136481285095,
+      "learning_rate": 1.2511702420586247e-05,
+      "loss": 8.7651,
+      "step": 116940
+    },
+    {
+      "epoch": 0.5840345576668581,
+      "grad_norm": 0.09328541159629822,
+      "learning_rate": 1.2510200505644698e-05,
+      "loss": 8.7717,
+      "step": 116950
+    },
+    {
+      "epoch": 0.5840844964917975,
+      "grad_norm": 0.09145410358905792,
+      "learning_rate": 1.2508698590703148e-05,
+      "loss": 8.77,
+      "step": 116960
+    },
+    {
+      "epoch": 0.584134435316737,
+      "grad_norm": 0.08732394874095917,
+      "learning_rate": 1.2507196675761596e-05,
+      "loss": 8.7813,
+      "step": 116970
+    },
+    {
+      "epoch": 0.5841843741416765,
+      "grad_norm": 0.09267169237136841,
+      "learning_rate": 1.2505694760820046e-05,
+      "loss": 8.7717,
+      "step": 116980
+    },
+    {
+      "epoch": 0.5842343129666159,
+      "grad_norm": 0.08930670469999313,
+      "learning_rate": 1.2504192845878495e-05,
+      "loss": 8.7648,
+      "step": 116990
+    },
+    {
+      "epoch": 0.5842842517915553,
+      "grad_norm": 0.09315798431634903,
+      "learning_rate": 1.2502690930936945e-05,
+      "loss": 8.7702,
+      "step": 117000
+    },
+    {
+      "epoch": 0.5843341906164948,
+      "grad_norm": 0.0956578180193901,
+      "learning_rate": 1.2501189015995395e-05,
+      "loss": 8.7584,
+      "step": 117010
+    },
+    {
+      "epoch": 0.5843841294414343,
+      "grad_norm": 0.0876522958278656,
+      "learning_rate": 1.2499687101053844e-05,
+      "loss": 8.7719,
+      "step": 117020
+    },
+    {
+      "epoch": 0.5844340682663737,
+      "grad_norm": 0.09839426726102829,
+      "learning_rate": 1.2498185186112294e-05,
+      "loss": 8.7733,
+      "step": 117030
+    },
+    {
+      "epoch": 0.5844840070913131,
+      "grad_norm": 0.09518906474113464,
+      "learning_rate": 1.2496683271170742e-05,
+      "loss": 8.7664,
+      "step": 117040
+    },
+    {
+      "epoch": 0.5845339459162526,
+      "grad_norm": 0.09161265939474106,
+      "learning_rate": 1.2495181356229193e-05,
+      "loss": 8.7651,
+      "step": 117050
+    },
+    {
+      "epoch": 0.584583884741192,
+      "grad_norm": 0.08892537653446198,
+      "learning_rate": 1.2493679441287643e-05,
+      "loss": 8.7663,
+      "step": 117060
+    },
+    {
+      "epoch": 0.5846338235661315,
+      "grad_norm": 0.09384747594594955,
+      "learning_rate": 1.2492177526346091e-05,
+      "loss": 8.7786,
+      "step": 117070
+    },
+    {
+      "epoch": 0.5846837623910709,
+      "grad_norm": 0.09078042209148407,
+      "learning_rate": 1.2490675611404542e-05,
+      "loss": 8.7735,
+      "step": 117080
+    },
+    {
+      "epoch": 0.5847337012160104,
+      "grad_norm": 0.09043184667825699,
+      "learning_rate": 1.248917369646299e-05,
+      "loss": 8.7555,
+      "step": 117090
+    },
+    {
+      "epoch": 0.5847836400409498,
+      "grad_norm": 0.09139489382505417,
+      "learning_rate": 1.248767178152144e-05,
+      "loss": 8.7699,
+      "step": 117100
+    },
+    {
+      "epoch": 0.5848335788658893,
+      "grad_norm": 0.09362711757421494,
+      "learning_rate": 1.248616986657989e-05,
+      "loss": 8.7641,
+      "step": 117110
+    },
+    {
+      "epoch": 0.5848835176908287,
+      "grad_norm": 0.09047945588827133,
+      "learning_rate": 1.2484667951638339e-05,
+      "loss": 8.766,
+      "step": 117120
+    },
+    {
+      "epoch": 0.5849334565157682,
+      "grad_norm": 0.09149231016635895,
+      "learning_rate": 1.2483166036696789e-05,
+      "loss": 8.7654,
+      "step": 117130
+    },
+    {
+      "epoch": 0.5849833953407076,
+      "grad_norm": 0.09510745108127594,
+      "learning_rate": 1.2481664121755237e-05,
+      "loss": 8.7704,
+      "step": 117140
+    },
+    {
+      "epoch": 0.5850333341656471,
+      "grad_norm": 0.09530346840620041,
+      "learning_rate": 1.2480162206813688e-05,
+      "loss": 8.777,
+      "step": 117150
+    },
+    {
+      "epoch": 0.5850832729905865,
+      "grad_norm": 0.09704668074846268,
+      "learning_rate": 1.2478660291872138e-05,
+      "loss": 8.7655,
+      "step": 117160
+    },
+    {
+      "epoch": 0.585133211815526,
+      "grad_norm": 0.095854252576828,
+      "learning_rate": 1.2477158376930586e-05,
+      "loss": 8.7692,
+      "step": 117170
+    },
+    {
+      "epoch": 0.5851831506404654,
+      "grad_norm": 0.09275175631046295,
+      "learning_rate": 1.2475656461989037e-05,
+      "loss": 8.7775,
+      "step": 117180
+    },
+    {
+      "epoch": 0.5852330894654049,
+      "grad_norm": 0.09659481048583984,
+      "learning_rate": 1.2474154547047487e-05,
+      "loss": 8.7822,
+      "step": 117190
+    },
+    {
+      "epoch": 0.5852830282903443,
+      "grad_norm": 0.08653581887483597,
+      "learning_rate": 1.2472652632105935e-05,
+      "loss": 8.774,
+      "step": 117200
+    },
+    {
+      "epoch": 0.5853329671152838,
+      "grad_norm": 0.0911843553185463,
+      "learning_rate": 1.2471150717164385e-05,
+      "loss": 8.7828,
+      "step": 117210
+    },
+    {
+      "epoch": 0.5853829059402232,
+      "grad_norm": 0.09394507855176926,
+      "learning_rate": 1.2469648802222834e-05,
+      "loss": 8.7798,
+      "step": 117220
+    },
+    {
+      "epoch": 0.5854328447651627,
+      "grad_norm": 0.0936194658279419,
+      "learning_rate": 1.2468146887281284e-05,
+      "loss": 8.7758,
+      "step": 117230
+    },
+    {
+      "epoch": 0.5854827835901021,
+      "grad_norm": 0.09510088711977005,
+      "learning_rate": 1.2466644972339734e-05,
+      "loss": 8.7771,
+      "step": 117240
+    },
+    {
+      "epoch": 0.5855327224150416,
+      "grad_norm": 0.09112706780433655,
+      "learning_rate": 1.2465143057398183e-05,
+      "loss": 8.7511,
+      "step": 117250
+    },
+    {
+      "epoch": 0.585582661239981,
+      "grad_norm": 0.08992750197649002,
+      "learning_rate": 1.2463641142456633e-05,
+      "loss": 8.766,
+      "step": 117260
+    },
+    {
+      "epoch": 0.5856326000649205,
+      "grad_norm": 0.09287506341934204,
+      "learning_rate": 1.2462139227515081e-05,
+      "loss": 8.7652,
+      "step": 117270
+    },
+    {
+      "epoch": 0.5856825388898599,
+      "grad_norm": 0.09044741094112396,
+      "learning_rate": 1.2460637312573532e-05,
+      "loss": 8.7573,
+      "step": 117280
+    },
+    {
+      "epoch": 0.5857324777147994,
+      "grad_norm": 0.09210161119699478,
+      "learning_rate": 1.2459135397631982e-05,
+      "loss": 8.7738,
+      "step": 117290
+    },
+    {
+      "epoch": 0.5857824165397388,
+      "grad_norm": 0.08944103866815567,
+      "learning_rate": 1.245763348269043e-05,
+      "loss": 8.766,
+      "step": 117300
+    },
+    {
+      "epoch": 0.5858323553646783,
+      "grad_norm": 0.09100234508514404,
+      "learning_rate": 1.245613156774888e-05,
+      "loss": 8.78,
+      "step": 117310
+    },
+    {
+      "epoch": 0.5858822941896177,
+      "grad_norm": 0.09310677647590637,
+      "learning_rate": 1.2454629652807329e-05,
+      "loss": 8.7623,
+      "step": 117320
+    },
+    {
+      "epoch": 0.5859322330145572,
+      "grad_norm": 0.09582335501909256,
+      "learning_rate": 1.2453127737865779e-05,
+      "loss": 8.7645,
+      "step": 117330
+    },
+    {
+      "epoch": 0.5859821718394966,
+      "grad_norm": 0.09254664927721024,
+      "learning_rate": 1.245162582292423e-05,
+      "loss": 8.7731,
+      "step": 117340
+    },
+    {
+      "epoch": 0.5860321106644361,
+      "grad_norm": 0.09486523270606995,
+      "learning_rate": 1.245012390798268e-05,
+      "loss": 8.7579,
+      "step": 117350
+    },
+    {
+      "epoch": 0.5860820494893755,
+      "grad_norm": 0.101737841963768,
+      "learning_rate": 1.2448621993041128e-05,
+      "loss": 8.7666,
+      "step": 117360
+    },
+    {
+      "epoch": 0.586131988314315,
+      "grad_norm": 0.08502329140901566,
+      "learning_rate": 1.2447120078099576e-05,
+      "loss": 8.7802,
+      "step": 117370
+    },
+    {
+      "epoch": 0.5861819271392544,
+      "grad_norm": 0.09242632240056992,
+      "learning_rate": 1.2445618163158027e-05,
+      "loss": 8.7565,
+      "step": 117380
+    },
+    {
+      "epoch": 0.5862318659641939,
+      "grad_norm": 0.08945038169622421,
+      "learning_rate": 1.2444116248216477e-05,
+      "loss": 8.7817,
+      "step": 117390
+    },
+    {
+      "epoch": 0.5862818047891333,
+      "grad_norm": 0.09282662719488144,
+      "learning_rate": 1.2442614333274927e-05,
+      "loss": 8.7616,
+      "step": 117400
+    },
+    {
+      "epoch": 0.5863317436140728,
+      "grad_norm": 0.08877874165773392,
+      "learning_rate": 1.2441112418333375e-05,
+      "loss": 8.7676,
+      "step": 117410
+    },
+    {
+      "epoch": 0.5863816824390122,
+      "grad_norm": 0.09669318050146103,
+      "learning_rate": 1.2439610503391824e-05,
+      "loss": 8.7787,
+      "step": 117420
+    },
+    {
+      "epoch": 0.5864316212639517,
+      "grad_norm": 0.08970874547958374,
+      "learning_rate": 1.2438108588450274e-05,
+      "loss": 8.7688,
+      "step": 117430
+    },
+    {
+      "epoch": 0.5864815600888911,
+      "grad_norm": 0.09568735212087631,
+      "learning_rate": 1.2436606673508724e-05,
+      "loss": 8.7597,
+      "step": 117440
+    },
+    {
+      "epoch": 0.5865314989138306,
+      "grad_norm": 0.09088582545518875,
+      "learning_rate": 1.2435104758567174e-05,
+      "loss": 8.7621,
+      "step": 117450
+    },
+    {
+      "epoch": 0.58658143773877,
+      "grad_norm": 0.08822001516819,
+      "learning_rate": 1.2433602843625623e-05,
+      "loss": 8.7695,
+      "step": 117460
+    },
+    {
+      "epoch": 0.5866313765637095,
+      "grad_norm": 0.09296117722988129,
+      "learning_rate": 1.2432100928684071e-05,
+      "loss": 8.7744,
+      "step": 117470
+    },
+    {
+      "epoch": 0.5866813153886489,
+      "grad_norm": 0.09976507723331451,
+      "learning_rate": 1.2430599013742522e-05,
+      "loss": 8.7743,
+      "step": 117480
+    },
+    {
+      "epoch": 0.5867312542135884,
+      "grad_norm": 0.09362848848104477,
+      "learning_rate": 1.2429097098800972e-05,
+      "loss": 8.7735,
+      "step": 117490
+    },
+    {
+      "epoch": 0.5867811930385278,
+      "grad_norm": 0.08755030483007431,
+      "learning_rate": 1.2427595183859422e-05,
+      "loss": 8.772,
+      "step": 117500
+    },
+    {
+      "epoch": 0.5868311318634672,
+      "grad_norm": 0.09280279278755188,
+      "learning_rate": 1.2426093268917872e-05,
+      "loss": 8.7612,
+      "step": 117510
+    },
+    {
+      "epoch": 0.5868810706884067,
+      "grad_norm": 0.08822941035032272,
+      "learning_rate": 1.2424591353976319e-05,
+      "loss": 8.7682,
+      "step": 117520
+    },
+    {
+      "epoch": 0.5869310095133462,
+      "grad_norm": 0.09698329120874405,
+      "learning_rate": 1.2423089439034769e-05,
+      "loss": 8.753,
+      "step": 117530
+    },
+    {
+      "epoch": 0.5869809483382856,
+      "grad_norm": 0.09743697941303253,
+      "learning_rate": 1.242158752409322e-05,
+      "loss": 8.7608,
+      "step": 117540
+    },
+    {
+      "epoch": 0.587030887163225,
+      "grad_norm": 0.09233416616916656,
+      "learning_rate": 1.242008560915167e-05,
+      "loss": 8.7681,
+      "step": 117550
+    },
+    {
+      "epoch": 0.5870808259881645,
+      "grad_norm": 0.09132979065179825,
+      "learning_rate": 1.241858369421012e-05,
+      "loss": 8.7779,
+      "step": 117560
+    },
+    {
+      "epoch": 0.587130764813104,
+      "grad_norm": 0.08776623755693436,
+      "learning_rate": 1.2417081779268566e-05,
+      "loss": 8.7762,
+      "step": 117570
+    },
+    {
+      "epoch": 0.5871807036380434,
+      "grad_norm": 0.09042856097221375,
+      "learning_rate": 1.2415579864327017e-05,
+      "loss": 8.7604,
+      "step": 117580
+    },
+    {
+      "epoch": 0.5872306424629828,
+      "grad_norm": 0.09197208285331726,
+      "learning_rate": 1.2414077949385467e-05,
+      "loss": 8.7629,
+      "step": 117590
+    },
+    {
+      "epoch": 0.5872805812879223,
+      "grad_norm": 0.09105636179447174,
+      "learning_rate": 1.2412576034443917e-05,
+      "loss": 8.7535,
+      "step": 117600
+    },
+    {
+      "epoch": 0.5873305201128618,
+      "grad_norm": 0.0907004103064537,
+      "learning_rate": 1.2411074119502367e-05,
+      "loss": 8.7604,
+      "step": 117610
+    },
+    {
+      "epoch": 0.5873804589378012,
+      "grad_norm": 0.09452099353075027,
+      "learning_rate": 1.2409572204560814e-05,
+      "loss": 8.757,
+      "step": 117620
+    },
+    {
+      "epoch": 0.5874303977627406,
+      "grad_norm": 0.09557855874300003,
+      "learning_rate": 1.2408070289619264e-05,
+      "loss": 8.7649,
+      "step": 117630
+    },
+    {
+      "epoch": 0.5874803365876801,
+      "grad_norm": 0.09412212669849396,
+      "learning_rate": 1.2406568374677714e-05,
+      "loss": 8.7602,
+      "step": 117640
+    },
+    {
+      "epoch": 0.5875302754126196,
+      "grad_norm": 0.09182430058717728,
+      "learning_rate": 1.2405066459736164e-05,
+      "loss": 8.7597,
+      "step": 117650
+    },
+    {
+      "epoch": 0.587580214237559,
+      "grad_norm": 0.08901126682758331,
+      "learning_rate": 1.2403564544794615e-05,
+      "loss": 8.7648,
+      "step": 117660
+    },
+    {
+      "epoch": 0.5876301530624984,
+      "grad_norm": 0.09120889008045197,
+      "learning_rate": 1.2402062629853063e-05,
+      "loss": 8.752,
+      "step": 117670
+    },
+    {
+      "epoch": 0.5876800918874379,
+      "grad_norm": 0.09408021718263626,
+      "learning_rate": 1.2400560714911512e-05,
+      "loss": 8.7826,
+      "step": 117680
+    },
+    {
+      "epoch": 0.5877300307123773,
+      "grad_norm": 0.0881025642156601,
+      "learning_rate": 1.2399058799969962e-05,
+      "loss": 8.7617,
+      "step": 117690
+    },
+    {
+      "epoch": 0.5877799695373168,
+      "grad_norm": 0.09191694110631943,
+      "learning_rate": 1.2397556885028412e-05,
+      "loss": 8.7643,
+      "step": 117700
+    },
+    {
+      "epoch": 0.5878299083622562,
+      "grad_norm": 0.08953646570444107,
+      "learning_rate": 1.2396054970086862e-05,
+      "loss": 8.7694,
+      "step": 117710
+    },
+    {
+      "epoch": 0.5878798471871957,
+      "grad_norm": 0.08859028667211533,
+      "learning_rate": 1.239455305514531e-05,
+      "loss": 8.773,
+      "step": 117720
+    },
+    {
+      "epoch": 0.5879297860121351,
+      "grad_norm": 0.09517339617013931,
+      "learning_rate": 1.2393051140203759e-05,
+      "loss": 8.7629,
+      "step": 117730
+    },
+    {
+      "epoch": 0.5879797248370746,
+      "grad_norm": 0.09675818681716919,
+      "learning_rate": 1.239154922526221e-05,
+      "loss": 8.7473,
+      "step": 117740
+    },
+    {
+      "epoch": 0.588029663662014,
+      "grad_norm": 0.09151841700077057,
+      "learning_rate": 1.239004731032066e-05,
+      "loss": 8.7576,
+      "step": 117750
+    },
+    {
+      "epoch": 0.5880796024869535,
+      "grad_norm": 0.09673187136650085,
+      "learning_rate": 1.238854539537911e-05,
+      "loss": 8.7646,
+      "step": 117760
+    },
+    {
+      "epoch": 0.5881295413118929,
+      "grad_norm": 0.0907597616314888,
+      "learning_rate": 1.2387043480437558e-05,
+      "loss": 8.7586,
+      "step": 117770
+    },
+    {
+      "epoch": 0.5881794801368324,
+      "grad_norm": 0.09877236932516098,
+      "learning_rate": 1.2385541565496007e-05,
+      "loss": 8.759,
+      "step": 117780
+    },
+    {
+      "epoch": 0.5882294189617718,
+      "grad_norm": 0.09087306261062622,
+      "learning_rate": 1.2384039650554457e-05,
+      "loss": 8.7683,
+      "step": 117790
+    },
+    {
+      "epoch": 0.5882793577867113,
+      "grad_norm": 0.09113863110542297,
+      "learning_rate": 1.2382537735612907e-05,
+      "loss": 8.7498,
+      "step": 117800
+    },
+    {
+      "epoch": 0.5883292966116507,
+      "grad_norm": 0.09708728641271591,
+      "learning_rate": 1.2381035820671357e-05,
+      "loss": 8.7612,
+      "step": 117810
+    },
+    {
+      "epoch": 0.5883792354365902,
+      "grad_norm": 0.09207798540592194,
+      "learning_rate": 1.2379533905729806e-05,
+      "loss": 8.771,
+      "step": 117820
+    },
+    {
+      "epoch": 0.5884291742615296,
+      "grad_norm": 0.09187018126249313,
+      "learning_rate": 1.2378031990788256e-05,
+      "loss": 8.7732,
+      "step": 117830
+    },
+    {
+      "epoch": 0.588479113086469,
+      "grad_norm": 0.09126574546098709,
+      "learning_rate": 1.2376530075846704e-05,
+      "loss": 8.7423,
+      "step": 117840
+    },
+    {
+      "epoch": 0.5885290519114085,
+      "grad_norm": 0.09289278835058212,
+      "learning_rate": 1.2375028160905154e-05,
+      "loss": 8.7448,
+      "step": 117850
+    },
+    {
+      "epoch": 0.588578990736348,
+      "grad_norm": 0.09014803171157837,
+      "learning_rate": 1.2373526245963605e-05,
+      "loss": 8.7721,
+      "step": 117860
+    },
+    {
+      "epoch": 0.5886289295612874,
+      "grad_norm": 0.09293773025274277,
+      "learning_rate": 1.2372024331022053e-05,
+      "loss": 8.7588,
+      "step": 117870
+    },
+    {
+      "epoch": 0.5886788683862269,
+      "grad_norm": 0.08857384324073792,
+      "learning_rate": 1.2370522416080503e-05,
+      "loss": 8.7656,
+      "step": 117880
+    },
+    {
+      "epoch": 0.5887288072111663,
+      "grad_norm": 0.09308899194002151,
+      "learning_rate": 1.2369020501138952e-05,
+      "loss": 8.7587,
+      "step": 117890
+    },
+    {
+      "epoch": 0.5887787460361058,
+      "grad_norm": 0.09509921073913574,
+      "learning_rate": 1.2367518586197402e-05,
+      "loss": 8.7589,
+      "step": 117900
+    },
+    {
+      "epoch": 0.5888286848610452,
+      "grad_norm": 0.090652696788311,
+      "learning_rate": 1.2366016671255852e-05,
+      "loss": 8.7635,
+      "step": 117910
+    },
+    {
+      "epoch": 0.5888786236859846,
+      "grad_norm": 0.08929238468408585,
+      "learning_rate": 1.23645147563143e-05,
+      "loss": 8.7458,
+      "step": 117920
+    },
+    {
+      "epoch": 0.5889285625109241,
+      "grad_norm": 0.09252835810184479,
+      "learning_rate": 1.236301284137275e-05,
+      "loss": 8.7504,
+      "step": 117930
+    },
+    {
+      "epoch": 0.5889785013358636,
+      "grad_norm": 0.09552726149559021,
+      "learning_rate": 1.23615109264312e-05,
+      "loss": 8.7493,
+      "step": 117940
+    },
+    {
+      "epoch": 0.589028440160803,
+      "grad_norm": 0.08871292322874069,
+      "learning_rate": 1.236000901148965e-05,
+      "loss": 8.7727,
+      "step": 117950
+    },
+    {
+      "epoch": 0.5890783789857424,
+      "grad_norm": 0.09849249571561813,
+      "learning_rate": 1.23585070965481e-05,
+      "loss": 8.7701,
+      "step": 117960
+    },
+    {
+      "epoch": 0.5891283178106819,
+      "grad_norm": 0.09064419567584991,
+      "learning_rate": 1.2357005181606548e-05,
+      "loss": 8.7677,
+      "step": 117970
+    },
+    {
+      "epoch": 0.5891782566356214,
+      "grad_norm": 0.096795454621315,
+      "learning_rate": 1.2355503266664998e-05,
+      "loss": 8.7601,
+      "step": 117980
+    },
+    {
+      "epoch": 0.5892281954605608,
+      "grad_norm": 0.09287415444850922,
+      "learning_rate": 1.2354001351723448e-05,
+      "loss": 8.7754,
+      "step": 117990
+    },
+    {
+      "epoch": 0.5892781342855002,
+      "grad_norm": 0.09310322254896164,
+      "learning_rate": 1.2352499436781897e-05,
+      "loss": 8.7775,
+      "step": 118000
+    },
+    {
+      "epoch": 0.5893280731104397,
+      "grad_norm": 0.09387247264385223,
+      "learning_rate": 1.2350997521840347e-05,
+      "loss": 8.7668,
+      "step": 118010
+    },
+    {
+      "epoch": 0.5893780119353792,
+      "grad_norm": 0.09476472437381744,
+      "learning_rate": 1.2349495606898796e-05,
+      "loss": 8.7517,
+      "step": 118020
+    },
+    {
+      "epoch": 0.5894279507603186,
+      "grad_norm": 0.09105034172534943,
+      "learning_rate": 1.2347993691957246e-05,
+      "loss": 8.7663,
+      "step": 118030
+    },
+    {
+      "epoch": 0.589477889585258,
+      "grad_norm": 0.09070300310850143,
+      "learning_rate": 1.2346491777015696e-05,
+      "loss": 8.7565,
+      "step": 118040
+    },
+    {
+      "epoch": 0.5895278284101975,
+      "grad_norm": 0.09422162175178528,
+      "learning_rate": 1.2344989862074144e-05,
+      "loss": 8.7575,
+      "step": 118050
+    },
+    {
+      "epoch": 0.589577767235137,
+      "grad_norm": 0.08815252780914307,
+      "learning_rate": 1.2343487947132595e-05,
+      "loss": 8.7475,
+      "step": 118060
+    },
+    {
+      "epoch": 0.5896277060600764,
+      "grad_norm": 0.08904334157705307,
+      "learning_rate": 1.2341986032191043e-05,
+      "loss": 8.7518,
+      "step": 118070
+    },
+    {
+      "epoch": 0.5896776448850158,
+      "grad_norm": 0.09929101914167404,
+      "learning_rate": 1.2340484117249493e-05,
+      "loss": 8.7426,
+      "step": 118080
+    },
+    {
+      "epoch": 0.5897275837099553,
+      "grad_norm": 0.09539781510829926,
+      "learning_rate": 1.2338982202307943e-05,
+      "loss": 8.7484,
+      "step": 118090
+    },
+    {
+      "epoch": 0.5897775225348948,
+      "grad_norm": 0.09773596376180649,
+      "learning_rate": 1.2337480287366392e-05,
+      "loss": 8.7371,
+      "step": 118100
+    },
+    {
+      "epoch": 0.5898274613598342,
+      "grad_norm": 0.09675204008817673,
+      "learning_rate": 1.2335978372424842e-05,
+      "loss": 8.7507,
+      "step": 118110
+    },
+    {
+      "epoch": 0.5898774001847736,
+      "grad_norm": 0.09443136304616928,
+      "learning_rate": 1.233447645748329e-05,
+      "loss": 8.7688,
+      "step": 118120
+    },
+    {
+      "epoch": 0.5899273390097131,
+      "grad_norm": 0.09123744815587997,
+      "learning_rate": 1.233297454254174e-05,
+      "loss": 8.7677,
+      "step": 118130
+    },
+    {
+      "epoch": 0.5899772778346526,
+      "grad_norm": 0.09106560051441193,
+      "learning_rate": 1.2331472627600191e-05,
+      "loss": 8.7569,
+      "step": 118140
+    },
+    {
+      "epoch": 0.590027216659592,
+      "grad_norm": 0.08924289792776108,
+      "learning_rate": 1.2329970712658641e-05,
+      "loss": 8.7577,
+      "step": 118150
+    },
+    {
+      "epoch": 0.5900771554845314,
+      "grad_norm": 0.08827322721481323,
+      "learning_rate": 1.232846879771709e-05,
+      "loss": 8.7717,
+      "step": 118160
+    },
+    {
+      "epoch": 0.5901270943094709,
+      "grad_norm": 0.08779691904783249,
+      "learning_rate": 1.2326966882775538e-05,
+      "loss": 8.7642,
+      "step": 118170
+    },
+    {
+      "epoch": 0.5901770331344104,
+      "grad_norm": 0.09139509499073029,
+      "learning_rate": 1.2325464967833988e-05,
+      "loss": 8.777,
+      "step": 118180
+    },
+    {
+      "epoch": 0.5902269719593498,
+      "grad_norm": 0.0951077789068222,
+      "learning_rate": 1.2323963052892438e-05,
+      "loss": 8.7487,
+      "step": 118190
+    },
+    {
+      "epoch": 0.5902769107842892,
+      "grad_norm": 0.09444812685251236,
+      "learning_rate": 1.2322461137950889e-05,
+      "loss": 8.752,
+      "step": 118200
+    },
+    {
+      "epoch": 0.5903268496092287,
+      "grad_norm": 0.0971173644065857,
+      "learning_rate": 1.2320959223009337e-05,
+      "loss": 8.7693,
+      "step": 118210
+    },
+    {
+      "epoch": 0.5903767884341682,
+      "grad_norm": 0.08963154256343842,
+      "learning_rate": 1.2319457308067786e-05,
+      "loss": 8.7772,
+      "step": 118220
+    },
+    {
+      "epoch": 0.5904267272591076,
+      "grad_norm": 0.09564877301454544,
+      "learning_rate": 1.2317955393126236e-05,
+      "loss": 8.771,
+      "step": 118230
+    },
+    {
+      "epoch": 0.590476666084047,
+      "grad_norm": 0.09680573642253876,
+      "learning_rate": 1.2316453478184686e-05,
+      "loss": 8.7641,
+      "step": 118240
+    },
+    {
+      "epoch": 0.5905266049089865,
+      "grad_norm": 0.09979761391878128,
+      "learning_rate": 1.2314951563243136e-05,
+      "loss": 8.7556,
+      "step": 118250
+    },
+    {
+      "epoch": 0.590576543733926,
+      "grad_norm": 0.0942818745970726,
+      "learning_rate": 1.2313449648301585e-05,
+      "loss": 8.7516,
+      "step": 118260
+    },
+    {
+      "epoch": 0.5906264825588654,
+      "grad_norm": 0.0858750268816948,
+      "learning_rate": 1.2311947733360033e-05,
+      "loss": 8.7616,
+      "step": 118270
+    },
+    {
+      "epoch": 0.5906764213838048,
+      "grad_norm": 0.09308760613203049,
+      "learning_rate": 1.2310445818418483e-05,
+      "loss": 8.7686,
+      "step": 118280
+    },
+    {
+      "epoch": 0.5907263602087443,
+      "grad_norm": 0.09751647710800171,
+      "learning_rate": 1.2308943903476933e-05,
+      "loss": 8.7575,
+      "step": 118290
+    },
+    {
+      "epoch": 0.5907762990336838,
+      "grad_norm": 0.09501097351312637,
+      "learning_rate": 1.2307441988535384e-05,
+      "loss": 8.7436,
+      "step": 118300
+    },
+    {
+      "epoch": 0.5908262378586232,
+      "grad_norm": 0.09237045794725418,
+      "learning_rate": 1.2305940073593834e-05,
+      "loss": 8.7537,
+      "step": 118310
+    },
+    {
+      "epoch": 0.5908761766835626,
+      "grad_norm": 0.08817274123430252,
+      "learning_rate": 1.230443815865228e-05,
+      "loss": 8.7618,
+      "step": 118320
+    },
+    {
+      "epoch": 0.590926115508502,
+      "grad_norm": 0.08590534329414368,
+      "learning_rate": 1.230293624371073e-05,
+      "loss": 8.7611,
+      "step": 118330
+    },
+    {
+      "epoch": 0.5909760543334416,
+      "grad_norm": 0.09068264812231064,
+      "learning_rate": 1.2301434328769181e-05,
+      "loss": 8.7521,
+      "step": 118340
+    },
+    {
+      "epoch": 0.591025993158381,
+      "grad_norm": 0.08929526805877686,
+      "learning_rate": 1.2299932413827631e-05,
+      "loss": 8.7686,
+      "step": 118350
+    },
+    {
+      "epoch": 0.5910759319833204,
+      "grad_norm": 0.09027723968029022,
+      "learning_rate": 1.2298430498886081e-05,
+      "loss": 8.7556,
+      "step": 118360
+    },
+    {
+      "epoch": 0.5911258708082598,
+      "grad_norm": 0.08829135447740555,
+      "learning_rate": 1.2296928583944528e-05,
+      "loss": 8.7631,
+      "step": 118370
+    },
+    {
+      "epoch": 0.5911758096331994,
+      "grad_norm": 0.08966977894306183,
+      "learning_rate": 1.2295426669002978e-05,
+      "loss": 8.7739,
+      "step": 118380
+    },
+    {
+      "epoch": 0.5912257484581388,
+      "grad_norm": 0.09342916309833527,
+      "learning_rate": 1.2293924754061428e-05,
+      "loss": 8.76,
+      "step": 118390
+    },
+    {
+      "epoch": 0.5912756872830782,
+      "grad_norm": 0.09105119109153748,
+      "learning_rate": 1.2292422839119879e-05,
+      "loss": 8.7592,
+      "step": 118400
+    },
+    {
+      "epoch": 0.5913256261080176,
+      "grad_norm": 0.09159651398658752,
+      "learning_rate": 1.2290920924178329e-05,
+      "loss": 8.7641,
+      "step": 118410
+    },
+    {
+      "epoch": 0.5913755649329572,
+      "grad_norm": 0.08399590849876404,
+      "learning_rate": 1.2289419009236776e-05,
+      "loss": 8.7847,
+      "step": 118420
+    },
+    {
+      "epoch": 0.5914255037578966,
+      "grad_norm": 0.08947955816984177,
+      "learning_rate": 1.2287917094295226e-05,
+      "loss": 8.7599,
+      "step": 118430
+    },
+    {
+      "epoch": 0.591475442582836,
+      "grad_norm": 0.08784180134534836,
+      "learning_rate": 1.2286415179353676e-05,
+      "loss": 8.7619,
+      "step": 118440
+    },
+    {
+      "epoch": 0.5915253814077754,
+      "grad_norm": 0.09851372241973877,
+      "learning_rate": 1.2284913264412126e-05,
+      "loss": 8.7478,
+      "step": 118450
+    },
+    {
+      "epoch": 0.591575320232715,
+      "grad_norm": 0.09141961485147476,
+      "learning_rate": 1.2283411349470576e-05,
+      "loss": 8.7752,
+      "step": 118460
+    },
+    {
+      "epoch": 0.5916252590576544,
+      "grad_norm": 0.09081381559371948,
+      "learning_rate": 1.2281909434529025e-05,
+      "loss": 8.7681,
+      "step": 118470
+    },
+    {
+      "epoch": 0.5916751978825938,
+      "grad_norm": 0.0934201031923294,
+      "learning_rate": 1.2280407519587473e-05,
+      "loss": 8.7523,
+      "step": 118480
+    },
+    {
+      "epoch": 0.5917251367075332,
+      "grad_norm": 0.09097079187631607,
+      "learning_rate": 1.2278905604645923e-05,
+      "loss": 8.7642,
+      "step": 118490
+    },
+    {
+      "epoch": 0.5917750755324728,
+      "grad_norm": 0.09448742121458054,
+      "learning_rate": 1.2277403689704374e-05,
+      "loss": 8.7639,
+      "step": 118500
+    },
+    {
+      "epoch": 0.5918250143574122,
+      "grad_norm": 0.09550192207098007,
+      "learning_rate": 1.2275901774762824e-05,
+      "loss": 8.7621,
+      "step": 118510
+    },
+    {
+      "epoch": 0.5918749531823516,
+      "grad_norm": 0.0902007445693016,
+      "learning_rate": 1.2274399859821272e-05,
+      "loss": 8.7534,
+      "step": 118520
+    },
+    {
+      "epoch": 0.591924892007291,
+      "grad_norm": 0.09537496417760849,
+      "learning_rate": 1.227289794487972e-05,
+      "loss": 8.7706,
+      "step": 118530
+    },
+    {
+      "epoch": 0.5919748308322306,
+      "grad_norm": 0.09085094183683395,
+      "learning_rate": 1.2271396029938171e-05,
+      "loss": 8.755,
+      "step": 118540
+    },
+    {
+      "epoch": 0.59202476965717,
+      "grad_norm": 0.09145759046077728,
+      "learning_rate": 1.2269894114996621e-05,
+      "loss": 8.7584,
+      "step": 118550
+    },
+    {
+      "epoch": 0.5920747084821094,
+      "grad_norm": 0.08843749761581421,
+      "learning_rate": 1.2268392200055071e-05,
+      "loss": 8.7606,
+      "step": 118560
+    },
+    {
+      "epoch": 0.5921246473070488,
+      "grad_norm": 0.08855432271957397,
+      "learning_rate": 1.226689028511352e-05,
+      "loss": 8.7522,
+      "step": 118570
+    },
+    {
+      "epoch": 0.5921745861319884,
+      "grad_norm": 0.09247451275587082,
+      "learning_rate": 1.2265388370171968e-05,
+      "loss": 8.756,
+      "step": 118580
+    },
+    {
+      "epoch": 0.5922245249569278,
+      "grad_norm": 0.09341256320476532,
+      "learning_rate": 1.2263886455230418e-05,
+      "loss": 8.7578,
+      "step": 118590
+    },
+    {
+      "epoch": 0.5922744637818672,
+      "grad_norm": 0.08816874027252197,
+      "learning_rate": 1.2262384540288869e-05,
+      "loss": 8.7486,
+      "step": 118600
+    },
+    {
+      "epoch": 0.5923244026068066,
+      "grad_norm": 0.08888726681470871,
+      "learning_rate": 1.2260882625347319e-05,
+      "loss": 8.7589,
+      "step": 118610
+    },
+    {
+      "epoch": 0.5923743414317462,
+      "grad_norm": 0.08801273256540298,
+      "learning_rate": 1.2259380710405767e-05,
+      "loss": 8.7634,
+      "step": 118620
+    },
+    {
+      "epoch": 0.5924242802566856,
+      "grad_norm": 0.09357626736164093,
+      "learning_rate": 1.2257878795464218e-05,
+      "loss": 8.7542,
+      "step": 118630
+    },
+    {
+      "epoch": 0.592474219081625,
+      "grad_norm": 0.09778086096048355,
+      "learning_rate": 1.2256376880522666e-05,
+      "loss": 8.7537,
+      "step": 118640
+    },
+    {
+      "epoch": 0.5925241579065644,
+      "grad_norm": 0.09161866456270218,
+      "learning_rate": 1.2254874965581116e-05,
+      "loss": 8.7466,
+      "step": 118650
+    },
+    {
+      "epoch": 0.592574096731504,
+      "grad_norm": 0.09025179594755173,
+      "learning_rate": 1.2253373050639566e-05,
+      "loss": 8.7628,
+      "step": 118660
+    },
+    {
+      "epoch": 0.5926240355564434,
+      "grad_norm": 0.09360655397176743,
+      "learning_rate": 1.2251871135698015e-05,
+      "loss": 8.7526,
+      "step": 118670
+    },
+    {
+      "epoch": 0.5926739743813828,
+      "grad_norm": 0.09320328384637833,
+      "learning_rate": 1.2250369220756465e-05,
+      "loss": 8.7654,
+      "step": 118680
+    },
+    {
+      "epoch": 0.5927239132063222,
+      "grad_norm": 0.08667218685150146,
+      "learning_rate": 1.2248867305814913e-05,
+      "loss": 8.763,
+      "step": 118690
+    },
+    {
+      "epoch": 0.5927738520312616,
+      "grad_norm": 0.08786769956350327,
+      "learning_rate": 1.2247365390873364e-05,
+      "loss": 8.7849,
+      "step": 118700
+    },
+    {
+      "epoch": 0.5928237908562012,
+      "grad_norm": 0.09707433730363846,
+      "learning_rate": 1.2245863475931814e-05,
+      "loss": 8.75,
+      "step": 118710
+    },
+    {
+      "epoch": 0.5928737296811406,
+      "grad_norm": 0.09028863161802292,
+      "learning_rate": 1.2244361560990262e-05,
+      "loss": 8.7534,
+      "step": 118720
+    },
+    {
+      "epoch": 0.59292366850608,
+      "grad_norm": 0.08839409798383713,
+      "learning_rate": 1.2242859646048713e-05,
+      "loss": 8.7801,
+      "step": 118730
+    },
+    {
+      "epoch": 0.5929736073310194,
+      "grad_norm": 0.09288914501667023,
+      "learning_rate": 1.2241357731107161e-05,
+      "loss": 8.7674,
+      "step": 118740
+    },
+    {
+      "epoch": 0.593023546155959,
+      "grad_norm": 0.08941594511270523,
+      "learning_rate": 1.2239855816165611e-05,
+      "loss": 8.7506,
+      "step": 118750
+    },
+    {
+      "epoch": 0.5930734849808984,
+      "grad_norm": 0.08502016216516495,
+      "learning_rate": 1.2238353901224061e-05,
+      "loss": 8.7585,
+      "step": 118760
+    },
+    {
+      "epoch": 0.5931234238058378,
+      "grad_norm": 0.09299997240304947,
+      "learning_rate": 1.223685198628251e-05,
+      "loss": 8.7511,
+      "step": 118770
+    },
+    {
+      "epoch": 0.5931733626307772,
+      "grad_norm": 0.09595504403114319,
+      "learning_rate": 1.223535007134096e-05,
+      "loss": 8.7624,
+      "step": 118780
+    },
+    {
+      "epoch": 0.5932233014557168,
+      "grad_norm": 0.09487087279558182,
+      "learning_rate": 1.223384815639941e-05,
+      "loss": 8.7711,
+      "step": 118790
+    },
+    {
+      "epoch": 0.5932732402806562,
+      "grad_norm": 0.09738994389772415,
+      "learning_rate": 1.2232346241457859e-05,
+      "loss": 8.7638,
+      "step": 118800
+    },
+    {
+      "epoch": 0.5933231791055956,
+      "grad_norm": 0.09411431103944778,
+      "learning_rate": 1.2230844326516309e-05,
+      "loss": 8.7554,
+      "step": 118810
+    },
+    {
+      "epoch": 0.593373117930535,
+      "grad_norm": 0.09381046146154404,
+      "learning_rate": 1.2229342411574757e-05,
+      "loss": 8.7498,
+      "step": 118820
+    },
+    {
+      "epoch": 0.5934230567554746,
+      "grad_norm": 0.08876951038837433,
+      "learning_rate": 1.2227840496633208e-05,
+      "loss": 8.7571,
+      "step": 118830
+    },
+    {
+      "epoch": 0.593472995580414,
+      "grad_norm": 0.08969464153051376,
+      "learning_rate": 1.2226338581691658e-05,
+      "loss": 8.7615,
+      "step": 118840
+    },
+    {
+      "epoch": 0.5935229344053534,
+      "grad_norm": 0.09235970675945282,
+      "learning_rate": 1.2224836666750106e-05,
+      "loss": 8.7604,
+      "step": 118850
+    },
+    {
+      "epoch": 0.5935728732302928,
+      "grad_norm": 0.09277419745922089,
+      "learning_rate": 1.2223334751808556e-05,
+      "loss": 8.7511,
+      "step": 118860
+    },
+    {
+      "epoch": 0.5936228120552324,
+      "grad_norm": 0.08957847207784653,
+      "learning_rate": 1.2221832836867007e-05,
+      "loss": 8.7614,
+      "step": 118870
+    },
+    {
+      "epoch": 0.5936727508801718,
+      "grad_norm": 0.09086732566356659,
+      "learning_rate": 1.2220330921925455e-05,
+      "loss": 8.7601,
+      "step": 118880
+    },
+    {
+      "epoch": 0.5937226897051112,
+      "grad_norm": 0.09175397455692291,
+      "learning_rate": 1.2218829006983905e-05,
+      "loss": 8.7584,
+      "step": 118890
+    },
+    {
+      "epoch": 0.5937726285300506,
+      "grad_norm": 0.09422596544027328,
+      "learning_rate": 1.2217327092042354e-05,
+      "loss": 8.7508,
+      "step": 118900
+    },
+    {
+      "epoch": 0.5938225673549902,
+      "grad_norm": 0.09395137429237366,
+      "learning_rate": 1.2215825177100804e-05,
+      "loss": 8.753,
+      "step": 118910
+    },
+    {
+      "epoch": 0.5938725061799296,
+      "grad_norm": 0.08856262266635895,
+      "learning_rate": 1.2214323262159254e-05,
+      "loss": 8.7564,
+      "step": 118920
+    },
+    {
+      "epoch": 0.593922445004869,
+      "grad_norm": 0.08992553502321243,
+      "learning_rate": 1.2212821347217703e-05,
+      "loss": 8.7466,
+      "step": 118930
+    },
+    {
+      "epoch": 0.5939723838298084,
+      "grad_norm": 0.094720758497715,
+      "learning_rate": 1.2211319432276153e-05,
+      "loss": 8.7473,
+      "step": 118940
+    },
+    {
+      "epoch": 0.594022322654748,
+      "grad_norm": 0.09277505427598953,
+      "learning_rate": 1.2209817517334603e-05,
+      "loss": 8.7496,
+      "step": 118950
+    },
+    {
+      "epoch": 0.5940722614796874,
+      "grad_norm": 0.09039611369371414,
+      "learning_rate": 1.2208315602393051e-05,
+      "loss": 8.746,
+      "step": 118960
+    },
+    {
+      "epoch": 0.5941222003046268,
+      "grad_norm": 0.09134664386510849,
+      "learning_rate": 1.2206813687451502e-05,
+      "loss": 8.7584,
+      "step": 118970
+    },
+    {
+      "epoch": 0.5941721391295662,
+      "grad_norm": 0.09162914752960205,
+      "learning_rate": 1.220531177250995e-05,
+      "loss": 8.754,
+      "step": 118980
+    },
+    {
+      "epoch": 0.5942220779545058,
+      "grad_norm": 0.0885622650384903,
+      "learning_rate": 1.22038098575684e-05,
+      "loss": 8.7476,
+      "step": 118990
+    },
+    {
+      "epoch": 0.5942720167794452,
+      "grad_norm": 0.09717550873756409,
+      "learning_rate": 1.220230794262685e-05,
+      "loss": 8.759,
+      "step": 119000
+    },
+    {
+      "epoch": 0.5943219556043846,
+      "grad_norm": 0.0922437235713005,
+      "learning_rate": 1.2200806027685299e-05,
+      "loss": 8.7528,
+      "step": 119010
+    },
+    {
+      "epoch": 0.594371894429324,
+      "grad_norm": 0.08450714498758316,
+      "learning_rate": 1.2199304112743749e-05,
+      "loss": 8.7721,
+      "step": 119020
+    },
+    {
+      "epoch": 0.5944218332542636,
+      "grad_norm": 0.09239663183689117,
+      "learning_rate": 1.2197802197802198e-05,
+      "loss": 8.7442,
+      "step": 119030
+    },
+    {
+      "epoch": 0.594471772079203,
+      "grad_norm": 0.09686051309108734,
+      "learning_rate": 1.2196300282860648e-05,
+      "loss": 8.7684,
+      "step": 119040
+    },
+    {
+      "epoch": 0.5945217109041424,
+      "grad_norm": 0.0940653383731842,
+      "learning_rate": 1.2194798367919098e-05,
+      "loss": 8.7579,
+      "step": 119050
+    },
+    {
+      "epoch": 0.5945716497290818,
+      "grad_norm": 0.08960103243589401,
+      "learning_rate": 1.2193296452977546e-05,
+      "loss": 8.7545,
+      "step": 119060
+    },
+    {
+      "epoch": 0.5946215885540214,
+      "grad_norm": 0.09707477688789368,
+      "learning_rate": 1.2191794538035997e-05,
+      "loss": 8.7575,
+      "step": 119070
+    },
+    {
+      "epoch": 0.5946715273789608,
+      "grad_norm": 0.09536729753017426,
+      "learning_rate": 1.2190292623094445e-05,
+      "loss": 8.7516,
+      "step": 119080
+    },
+    {
+      "epoch": 0.5947214662039002,
+      "grad_norm": 0.09189159423112869,
+      "learning_rate": 1.2188790708152895e-05,
+      "loss": 8.7709,
+      "step": 119090
+    },
+    {
+      "epoch": 0.5947714050288396,
+      "grad_norm": 0.09311476349830627,
+      "learning_rate": 1.2187288793211345e-05,
+      "loss": 8.7534,
+      "step": 119100
+    },
+    {
+      "epoch": 0.5948213438537792,
+      "grad_norm": 0.09297338128089905,
+      "learning_rate": 1.2185786878269796e-05,
+      "loss": 8.7442,
+      "step": 119110
+    },
+    {
+      "epoch": 0.5948712826787186,
+      "grad_norm": 0.09508251398801804,
+      "learning_rate": 1.2184284963328244e-05,
+      "loss": 8.7543,
+      "step": 119120
+    },
+    {
+      "epoch": 0.594921221503658,
+      "grad_norm": 0.09149755537509918,
+      "learning_rate": 1.2182783048386693e-05,
+      "loss": 8.741,
+      "step": 119130
+    },
+    {
+      "epoch": 0.5949711603285974,
+      "grad_norm": 0.09345537424087524,
+      "learning_rate": 1.2181281133445143e-05,
+      "loss": 8.7353,
+      "step": 119140
+    },
+    {
+      "epoch": 0.595021099153537,
+      "grad_norm": 0.08903185278177261,
+      "learning_rate": 1.2179779218503593e-05,
+      "loss": 8.7532,
+      "step": 119150
+    },
+    {
+      "epoch": 0.5950710379784764,
+      "grad_norm": 0.08982034772634506,
+      "learning_rate": 1.2178277303562043e-05,
+      "loss": 8.7417,
+      "step": 119160
+    },
+    {
+      "epoch": 0.5951209768034158,
+      "grad_norm": 0.09356318414211273,
+      "learning_rate": 1.2176775388620492e-05,
+      "loss": 8.7377,
+      "step": 119170
+    },
+    {
+      "epoch": 0.5951709156283552,
+      "grad_norm": 0.09444409608840942,
+      "learning_rate": 1.217527347367894e-05,
+      "loss": 8.7521,
+      "step": 119180
+    },
+    {
+      "epoch": 0.5952208544532948,
+      "grad_norm": 0.09872134029865265,
+      "learning_rate": 1.217377155873739e-05,
+      "loss": 8.7463,
+      "step": 119190
+    },
+    {
+      "epoch": 0.5952707932782342,
+      "grad_norm": 0.09588941931724548,
+      "learning_rate": 1.217226964379584e-05,
+      "loss": 8.7578,
+      "step": 119200
+    },
+    {
+      "epoch": 0.5953207321031736,
+      "grad_norm": 0.09109771251678467,
+      "learning_rate": 1.217076772885429e-05,
+      "loss": 8.7556,
+      "step": 119210
+    },
+    {
+      "epoch": 0.595370670928113,
+      "grad_norm": 0.09148535132408142,
+      "learning_rate": 1.2169265813912739e-05,
+      "loss": 8.7584,
+      "step": 119220
+    },
+    {
+      "epoch": 0.5954206097530526,
+      "grad_norm": 0.09875836223363876,
+      "learning_rate": 1.2167763898971188e-05,
+      "loss": 8.7525,
+      "step": 119230
+    },
+    {
+      "epoch": 0.595470548577992,
+      "grad_norm": 0.09247757494449615,
+      "learning_rate": 1.2166261984029638e-05,
+      "loss": 8.7402,
+      "step": 119240
+    },
+    {
+      "epoch": 0.5955204874029314,
+      "grad_norm": 0.09525955468416214,
+      "learning_rate": 1.2164760069088088e-05,
+      "loss": 8.7735,
+      "step": 119250
+    },
+    {
+      "epoch": 0.5955704262278708,
+      "grad_norm": 0.09496864676475525,
+      "learning_rate": 1.2163258154146538e-05,
+      "loss": 8.7534,
+      "step": 119260
+    },
+    {
+      "epoch": 0.5956203650528104,
+      "grad_norm": 0.09750999510288239,
+      "learning_rate": 1.2161756239204988e-05,
+      "loss": 8.7551,
+      "step": 119270
+    },
+    {
+      "epoch": 0.5956703038777498,
+      "grad_norm": 0.0901816189289093,
+      "learning_rate": 1.2160254324263435e-05,
+      "loss": 8.7569,
+      "step": 119280
+    },
+    {
+      "epoch": 0.5957202427026892,
+      "grad_norm": 0.09008877724409103,
+      "learning_rate": 1.2158752409321885e-05,
+      "loss": 8.7433,
+      "step": 119290
+    },
+    {
+      "epoch": 0.5957701815276286,
+      "grad_norm": 0.08970709890127182,
+      "learning_rate": 1.2157250494380335e-05,
+      "loss": 8.7651,
+      "step": 119300
+    },
+    {
+      "epoch": 0.5958201203525681,
+      "grad_norm": 0.09239272028207779,
+      "learning_rate": 1.2155748579438786e-05,
+      "loss": 8.746,
+      "step": 119310
+    },
+    {
+      "epoch": 0.5958700591775076,
+      "grad_norm": 0.0948515385389328,
+      "learning_rate": 1.2154246664497236e-05,
+      "loss": 8.7497,
+      "step": 119320
+    },
+    {
+      "epoch": 0.595919998002447,
+      "grad_norm": 0.09544006735086441,
+      "learning_rate": 1.2152744749555683e-05,
+      "loss": 8.7772,
+      "step": 119330
+    },
+    {
+      "epoch": 0.5959699368273864,
+      "grad_norm": 0.092537522315979,
+      "learning_rate": 1.2151242834614133e-05,
+      "loss": 8.7515,
+      "step": 119340
+    },
+    {
+      "epoch": 0.596019875652326,
+      "grad_norm": 0.09258683770895004,
+      "learning_rate": 1.2149740919672583e-05,
+      "loss": 8.7646,
+      "step": 119350
+    },
+    {
+      "epoch": 0.5960698144772654,
+      "grad_norm": 0.09671740233898163,
+      "learning_rate": 1.2148239004731033e-05,
+      "loss": 8.7454,
+      "step": 119360
+    },
+    {
+      "epoch": 0.5961197533022048,
+      "grad_norm": 0.0934329703450203,
+      "learning_rate": 1.2146737089789483e-05,
+      "loss": 8.7486,
+      "step": 119370
+    },
+    {
+      "epoch": 0.5961696921271442,
+      "grad_norm": 0.09332381933927536,
+      "learning_rate": 1.214523517484793e-05,
+      "loss": 8.7586,
+      "step": 119380
+    },
+    {
+      "epoch": 0.5962196309520837,
+      "grad_norm": 0.09403794258832932,
+      "learning_rate": 1.214373325990638e-05,
+      "loss": 8.7459,
+      "step": 119390
+    },
+    {
+      "epoch": 0.5962695697770232,
+      "grad_norm": 0.097443126142025,
+      "learning_rate": 1.214223134496483e-05,
+      "loss": 8.7482,
+      "step": 119400
+    },
+    {
+      "epoch": 0.5963195086019626,
+      "grad_norm": 0.09268827736377716,
+      "learning_rate": 1.214072943002328e-05,
+      "loss": 8.75,
+      "step": 119410
+    },
+    {
+      "epoch": 0.596369447426902,
+      "grad_norm": 0.09062253683805466,
+      "learning_rate": 1.213922751508173e-05,
+      "loss": 8.7481,
+      "step": 119420
+    },
+    {
+      "epoch": 0.5964193862518415,
+      "grad_norm": 0.08736632764339447,
+      "learning_rate": 1.213772560014018e-05,
+      "loss": 8.7588,
+      "step": 119430
+    },
+    {
+      "epoch": 0.596469325076781,
+      "grad_norm": 0.09273076057434082,
+      "learning_rate": 1.2136223685198628e-05,
+      "loss": 8.7602,
+      "step": 119440
+    },
+    {
+      "epoch": 0.5965192639017204,
+      "grad_norm": 0.0936063900589943,
+      "learning_rate": 1.2134721770257078e-05,
+      "loss": 8.7539,
+      "step": 119450
+    },
+    {
+      "epoch": 0.5965692027266598,
+      "grad_norm": 0.09075089544057846,
+      "learning_rate": 1.2133219855315528e-05,
+      "loss": 8.7523,
+      "step": 119460
+    },
+    {
+      "epoch": 0.5966191415515993,
+      "grad_norm": 0.09237466752529144,
+      "learning_rate": 1.2131717940373978e-05,
+      "loss": 8.7609,
+      "step": 119470
+    },
+    {
+      "epoch": 0.5966690803765388,
+      "grad_norm": 0.09360762685537338,
+      "learning_rate": 1.2130216025432427e-05,
+      "loss": 8.7495,
+      "step": 119480
+    },
+    {
+      "epoch": 0.5967190192014782,
+      "grad_norm": 0.09769395738840103,
+      "learning_rate": 1.2128714110490875e-05,
+      "loss": 8.7676,
+      "step": 119490
+    },
+    {
+      "epoch": 0.5967689580264176,
+      "grad_norm": 0.0937671959400177,
+      "learning_rate": 1.2127212195549325e-05,
+      "loss": 8.7536,
+      "step": 119500
+    },
+    {
+      "epoch": 0.5968188968513571,
+      "grad_norm": 0.09220635890960693,
+      "learning_rate": 1.2125710280607776e-05,
+      "loss": 8.7522,
+      "step": 119510
+    },
+    {
+      "epoch": 0.5968688356762966,
+      "grad_norm": 0.0912616029381752,
+      "learning_rate": 1.2124208365666226e-05,
+      "loss": 8.7607,
+      "step": 119520
+    },
+    {
+      "epoch": 0.596918774501236,
+      "grad_norm": 0.09352380782365799,
+      "learning_rate": 1.2122706450724674e-05,
+      "loss": 8.7563,
+      "step": 119530
+    },
+    {
+      "epoch": 0.5969687133261754,
+      "grad_norm": 0.09166228026151657,
+      "learning_rate": 1.2121204535783123e-05,
+      "loss": 8.7546,
+      "step": 119540
+    },
+    {
+      "epoch": 0.5970186521511149,
+      "grad_norm": 0.09042517840862274,
+      "learning_rate": 1.2119702620841573e-05,
+      "loss": 8.7494,
+      "step": 119550
+    },
+    {
+      "epoch": 0.5970685909760544,
+      "grad_norm": 0.09637751430273056,
+      "learning_rate": 1.2118200705900023e-05,
+      "loss": 8.7369,
+      "step": 119560
+    },
+    {
+      "epoch": 0.5971185298009938,
+      "grad_norm": 0.09652339667081833,
+      "learning_rate": 1.2116698790958473e-05,
+      "loss": 8.7499,
+      "step": 119570
+    },
+    {
+      "epoch": 0.5971684686259332,
+      "grad_norm": 0.08974555879831314,
+      "learning_rate": 1.2115196876016922e-05,
+      "loss": 8.7598,
+      "step": 119580
+    },
+    {
+      "epoch": 0.5972184074508727,
+      "grad_norm": 0.09256293624639511,
+      "learning_rate": 1.2113694961075372e-05,
+      "loss": 8.7288,
+      "step": 119590
+    },
+    {
+      "epoch": 0.5972683462758122,
+      "grad_norm": 0.09114140272140503,
+      "learning_rate": 1.211219304613382e-05,
+      "loss": 8.77,
+      "step": 119600
+    },
+    {
+      "epoch": 0.5973182851007516,
+      "grad_norm": 0.09448247402906418,
+      "learning_rate": 1.211069113119227e-05,
+      "loss": 8.7269,
+      "step": 119610
+    },
+    {
+      "epoch": 0.597368223925691,
+      "grad_norm": 0.09690418839454651,
+      "learning_rate": 1.210918921625072e-05,
+      "loss": 8.7554,
+      "step": 119620
+    },
+    {
+      "epoch": 0.5974181627506305,
+      "grad_norm": 0.09861066192388535,
+      "learning_rate": 1.210768730130917e-05,
+      "loss": 8.76,
+      "step": 119630
+    },
+    {
+      "epoch": 0.59746810157557,
+      "grad_norm": 0.09121115505695343,
+      "learning_rate": 1.210618538636762e-05,
+      "loss": 8.7617,
+      "step": 119640
+    },
+    {
+      "epoch": 0.5975180404005094,
+      "grad_norm": 0.09317606687545776,
+      "learning_rate": 1.2104683471426068e-05,
+      "loss": 8.7548,
+      "step": 119650
+    },
+    {
+      "epoch": 0.5975679792254488,
+      "grad_norm": 0.08856779336929321,
+      "learning_rate": 1.2103181556484518e-05,
+      "loss": 8.7495,
+      "step": 119660
+    },
+    {
+      "epoch": 0.5976179180503882,
+      "grad_norm": 0.09526504576206207,
+      "learning_rate": 1.2101679641542968e-05,
+      "loss": 8.7478,
+      "step": 119670
+    },
+    {
+      "epoch": 0.5976678568753278,
+      "grad_norm": 0.0859864354133606,
+      "learning_rate": 1.2100177726601417e-05,
+      "loss": 8.7426,
+      "step": 119680
+    },
+    {
+      "epoch": 0.5977177957002672,
+      "grad_norm": 0.08864925801753998,
+      "learning_rate": 1.2098675811659867e-05,
+      "loss": 8.7371,
+      "step": 119690
+    },
+    {
+      "epoch": 0.5977677345252066,
+      "grad_norm": 0.08991964906454086,
+      "learning_rate": 1.2097173896718315e-05,
+      "loss": 8.7372,
+      "step": 119700
+    },
+    {
+      "epoch": 0.597817673350146,
+      "grad_norm": 0.09090539813041687,
+      "learning_rate": 1.2095671981776766e-05,
+      "loss": 8.7734,
+      "step": 119710
+    },
+    {
+      "epoch": 0.5978676121750855,
+      "grad_norm": 0.09099775552749634,
+      "learning_rate": 1.2094170066835216e-05,
+      "loss": 8.7409,
+      "step": 119720
+    },
+    {
+      "epoch": 0.597917551000025,
+      "grad_norm": 0.08712445944547653,
+      "learning_rate": 1.2092668151893664e-05,
+      "loss": 8.7475,
+      "step": 119730
+    },
+    {
+      "epoch": 0.5979674898249644,
+      "grad_norm": 0.08881444483995438,
+      "learning_rate": 1.2091166236952114e-05,
+      "loss": 8.7383,
+      "step": 119740
+    },
+    {
+      "epoch": 0.5980174286499038,
+      "grad_norm": 0.0887836366891861,
+      "learning_rate": 1.2089664322010565e-05,
+      "loss": 8.7615,
+      "step": 119750
+    },
+    {
+      "epoch": 0.5980673674748433,
+      "grad_norm": 0.09183062613010406,
+      "learning_rate": 1.2088162407069013e-05,
+      "loss": 8.7624,
+      "step": 119760
+    },
+    {
+      "epoch": 0.5981173062997828,
+      "grad_norm": 0.09438557177782059,
+      "learning_rate": 1.2086660492127463e-05,
+      "loss": 8.7546,
+      "step": 119770
+    },
+    {
+      "epoch": 0.5981672451247222,
+      "grad_norm": 0.09165972471237183,
+      "learning_rate": 1.2085158577185912e-05,
+      "loss": 8.7515,
+      "step": 119780
+    },
+    {
+      "epoch": 0.5982171839496616,
+      "grad_norm": 0.0887795090675354,
+      "learning_rate": 1.2083656662244362e-05,
+      "loss": 8.7527,
+      "step": 119790
+    },
+    {
+      "epoch": 0.5982671227746011,
+      "grad_norm": 0.09268735349178314,
+      "learning_rate": 1.2082154747302812e-05,
+      "loss": 8.7633,
+      "step": 119800
+    },
+    {
+      "epoch": 0.5983170615995406,
+      "grad_norm": 0.09194294363260269,
+      "learning_rate": 1.208065283236126e-05,
+      "loss": 8.7406,
+      "step": 119810
+    },
+    {
+      "epoch": 0.59836700042448,
+      "grad_norm": 0.09086401760578156,
+      "learning_rate": 1.207915091741971e-05,
+      "loss": 8.7622,
+      "step": 119820
+    },
+    {
+      "epoch": 0.5984169392494194,
+      "grad_norm": 0.0923188179731369,
+      "learning_rate": 1.207764900247816e-05,
+      "loss": 8.7406,
+      "step": 119830
+    },
+    {
+      "epoch": 0.5984668780743589,
+      "grad_norm": 0.09538555890321732,
+      "learning_rate": 1.207614708753661e-05,
+      "loss": 8.7567,
+      "step": 119840
+    },
+    {
+      "epoch": 0.5985168168992984,
+      "grad_norm": 0.09424526989459991,
+      "learning_rate": 1.207464517259506e-05,
+      "loss": 8.7516,
+      "step": 119850
+    },
+    {
+      "epoch": 0.5985667557242378,
+      "grad_norm": 0.09127725660800934,
+      "learning_rate": 1.2073143257653508e-05,
+      "loss": 8.762,
+      "step": 119860
+    },
+    {
+      "epoch": 0.5986166945491772,
+      "grad_norm": 0.09290522336959839,
+      "learning_rate": 1.2071641342711958e-05,
+      "loss": 8.752,
+      "step": 119870
+    },
+    {
+      "epoch": 0.5986666333741167,
+      "grad_norm": 0.09055527299642563,
+      "learning_rate": 1.2070139427770407e-05,
+      "loss": 8.7407,
+      "step": 119880
+    },
+    {
+      "epoch": 0.5987165721990562,
+      "grad_norm": 0.09295734018087387,
+      "learning_rate": 1.2068637512828857e-05,
+      "loss": 8.7368,
+      "step": 119890
+    },
+    {
+      "epoch": 0.5987665110239956,
+      "grad_norm": 0.09094783663749695,
+      "learning_rate": 1.2067135597887307e-05,
+      "loss": 8.7513,
+      "step": 119900
+    },
+    {
+      "epoch": 0.598816449848935,
+      "grad_norm": 0.09828267991542816,
+      "learning_rate": 1.2065633682945757e-05,
+      "loss": 8.7433,
+      "step": 119910
+    },
+    {
+      "epoch": 0.5988663886738745,
+      "grad_norm": 0.10480784624814987,
+      "learning_rate": 1.2064131768004206e-05,
+      "loss": 8.729,
+      "step": 119920
+    },
+    {
+      "epoch": 0.598916327498814,
+      "grad_norm": 0.09408173710107803,
+      "learning_rate": 1.2062629853062654e-05,
+      "loss": 8.748,
+      "step": 119930
+    },
+    {
+      "epoch": 0.5989662663237534,
+      "grad_norm": 0.09466306120157242,
+      "learning_rate": 1.2061127938121104e-05,
+      "loss": 8.739,
+      "step": 119940
+    },
+    {
+      "epoch": 0.5990162051486928,
+      "grad_norm": 0.09082759916782379,
+      "learning_rate": 1.2059626023179555e-05,
+      "loss": 8.751,
+      "step": 119950
+    },
+    {
+      "epoch": 0.5990661439736323,
+      "grad_norm": 0.09120157361030579,
+      "learning_rate": 1.2058124108238005e-05,
+      "loss": 8.7516,
+      "step": 119960
+    },
+    {
+      "epoch": 0.5991160827985718,
+      "grad_norm": 0.08926384896039963,
+      "learning_rate": 1.2056622193296453e-05,
+      "loss": 8.7449,
+      "step": 119970
+    },
+    {
+      "epoch": 0.5991660216235112,
+      "grad_norm": 0.09408469498157501,
+      "learning_rate": 1.2055120278354902e-05,
+      "loss": 8.7444,
+      "step": 119980
+    },
+    {
+      "epoch": 0.5992159604484506,
+      "grad_norm": 0.09376782178878784,
+      "learning_rate": 1.2053618363413352e-05,
+      "loss": 8.7396,
+      "step": 119990
+    },
+    {
+      "epoch": 0.5992658992733901,
+      "grad_norm": 0.0928616002202034,
+      "learning_rate": 1.2052116448471802e-05,
+      "loss": 8.7482,
+      "step": 120000
+    },
+    {
+      "epoch": 0.5993158380983296,
+      "grad_norm": 0.09431707859039307,
+      "learning_rate": 1.2050614533530252e-05,
+      "loss": 8.7567,
+      "step": 120010
+    },
+    {
+      "epoch": 0.599365776923269,
+      "grad_norm": 0.08612378686666489,
+      "learning_rate": 1.20491126185887e-05,
+      "loss": 8.7719,
+      "step": 120020
+    },
+    {
+      "epoch": 0.5994157157482084,
+      "grad_norm": 0.08852338045835495,
+      "learning_rate": 1.204761070364715e-05,
+      "loss": 8.7456,
+      "step": 120030
+    },
+    {
+      "epoch": 0.5994656545731479,
+      "grad_norm": 0.09709616750478745,
+      "learning_rate": 1.20461087887056e-05,
+      "loss": 8.7524,
+      "step": 120040
+    },
+    {
+      "epoch": 0.5995155933980874,
+      "grad_norm": 0.08951878547668457,
+      "learning_rate": 1.204460687376405e-05,
+      "loss": 8.7406,
+      "step": 120050
+    },
+    {
+      "epoch": 0.5995655322230268,
+      "grad_norm": 0.08870140463113785,
+      "learning_rate": 1.20431049588225e-05,
+      "loss": 8.746,
+      "step": 120060
+    },
+    {
+      "epoch": 0.5996154710479662,
+      "grad_norm": 0.0954098254442215,
+      "learning_rate": 1.2041603043880948e-05,
+      "loss": 8.7421,
+      "step": 120070
+    },
+    {
+      "epoch": 0.5996654098729057,
+      "grad_norm": 0.09251780807971954,
+      "learning_rate": 1.2040101128939397e-05,
+      "loss": 8.726,
+      "step": 120080
+    },
+    {
+      "epoch": 0.5997153486978452,
+      "grad_norm": 0.09680180251598358,
+      "learning_rate": 1.2038599213997847e-05,
+      "loss": 8.7528,
+      "step": 120090
+    },
+    {
+      "epoch": 0.5997652875227846,
+      "grad_norm": 0.08500494062900543,
+      "learning_rate": 1.2037097299056297e-05,
+      "loss": 8.7571,
+      "step": 120100
+    },
+    {
+      "epoch": 0.599815226347724,
+      "grad_norm": 0.08889224380254745,
+      "learning_rate": 1.2035595384114747e-05,
+      "loss": 8.758,
+      "step": 120110
+    },
+    {
+      "epoch": 0.5998651651726635,
+      "grad_norm": 0.09139248728752136,
+      "learning_rate": 1.2034093469173198e-05,
+      "loss": 8.7463,
+      "step": 120120
+    },
+    {
+      "epoch": 0.599915103997603,
+      "grad_norm": 0.09338033199310303,
+      "learning_rate": 1.2032591554231644e-05,
+      "loss": 8.7608,
+      "step": 120130
+    },
+    {
+      "epoch": 0.5999650428225424,
+      "grad_norm": 0.09576421976089478,
+      "learning_rate": 1.2031089639290094e-05,
+      "loss": 8.7344,
+      "step": 120140
+    },
+    {
+      "epoch": 0.6000149816474818,
+      "grad_norm": 0.0958356186747551,
+      "learning_rate": 1.2029587724348545e-05,
+      "loss": 8.7332,
+      "step": 120150
+    },
+    {
+      "epoch": 0.6000649204724213,
+      "grad_norm": 0.10234443098306656,
+      "learning_rate": 1.2028085809406995e-05,
+      "loss": 8.7325,
+      "step": 120160
+    },
+    {
+      "epoch": 0.6001148592973607,
+      "grad_norm": 0.08857684582471848,
+      "learning_rate": 1.2026583894465445e-05,
+      "loss": 8.7573,
+      "step": 120170
+    },
+    {
+      "epoch": 0.6001647981223002,
+      "grad_norm": 0.09901963919401169,
+      "learning_rate": 1.2025081979523892e-05,
+      "loss": 8.7471,
+      "step": 120180
+    },
+    {
+      "epoch": 0.6002147369472396,
+      "grad_norm": 0.09748201072216034,
+      "learning_rate": 1.2023580064582342e-05,
+      "loss": 8.7285,
+      "step": 120190
+    },
+    {
+      "epoch": 0.6002646757721791,
+      "grad_norm": 0.09091214835643768,
+      "learning_rate": 1.2022078149640792e-05,
+      "loss": 8.7214,
+      "step": 120200
+    },
+    {
+      "epoch": 0.6003146145971185,
+      "grad_norm": 0.09007634967565536,
+      "learning_rate": 1.2020576234699242e-05,
+      "loss": 8.7256,
+      "step": 120210
+    },
+    {
+      "epoch": 0.600364553422058,
+      "grad_norm": 0.0863836258649826,
+      "learning_rate": 1.2019074319757693e-05,
+      "loss": 8.7418,
+      "step": 120220
+    },
+    {
+      "epoch": 0.6004144922469974,
+      "grad_norm": 0.09045583009719849,
+      "learning_rate": 1.201757240481614e-05,
+      "loss": 8.7637,
+      "step": 120230
+    },
+    {
+      "epoch": 0.6004644310719369,
+      "grad_norm": 0.09015191346406937,
+      "learning_rate": 1.201607048987459e-05,
+      "loss": 8.7327,
+      "step": 120240
+    },
+    {
+      "epoch": 0.6005143698968763,
+      "grad_norm": 0.09545261412858963,
+      "learning_rate": 1.201456857493304e-05,
+      "loss": 8.7351,
+      "step": 120250
+    },
+    {
+      "epoch": 0.6005643087218158,
+      "grad_norm": 0.09164721518754959,
+      "learning_rate": 1.201306665999149e-05,
+      "loss": 8.7474,
+      "step": 120260
+    },
+    {
+      "epoch": 0.6006142475467552,
+      "grad_norm": 0.09745185077190399,
+      "learning_rate": 1.201156474504994e-05,
+      "loss": 8.7403,
+      "step": 120270
+    },
+    {
+      "epoch": 0.6006641863716947,
+      "grad_norm": 0.08840557932853699,
+      "learning_rate": 1.2010062830108389e-05,
+      "loss": 8.7636,
+      "step": 120280
+    },
+    {
+      "epoch": 0.6007141251966341,
+      "grad_norm": 0.10116366297006607,
+      "learning_rate": 1.2008560915166837e-05,
+      "loss": 8.7441,
+      "step": 120290
+    },
+    {
+      "epoch": 0.6007640640215736,
+      "grad_norm": 0.09359109401702881,
+      "learning_rate": 1.2007059000225287e-05,
+      "loss": 8.7407,
+      "step": 120300
+    },
+    {
+      "epoch": 0.600814002846513,
+      "grad_norm": 0.09176816046237946,
+      "learning_rate": 1.2005557085283737e-05,
+      "loss": 8.7537,
+      "step": 120310
+    },
+    {
+      "epoch": 0.6008639416714525,
+      "grad_norm": 0.09760214388370514,
+      "learning_rate": 1.2004055170342188e-05,
+      "loss": 8.7465,
+      "step": 120320
+    },
+    {
+      "epoch": 0.6009138804963919,
+      "grad_norm": 0.09278165549039841,
+      "learning_rate": 1.2002553255400636e-05,
+      "loss": 8.737,
+      "step": 120330
+    },
+    {
+      "epoch": 0.6009638193213314,
+      "grad_norm": 0.09701333940029144,
+      "learning_rate": 1.2001051340459085e-05,
+      "loss": 8.7367,
+      "step": 120340
+    },
+    {
+      "epoch": 0.6010137581462708,
+      "grad_norm": 0.09638296812772751,
+      "learning_rate": 1.1999549425517535e-05,
+      "loss": 8.7463,
+      "step": 120350
+    },
+    {
+      "epoch": 0.6010636969712103,
+      "grad_norm": 0.09861765056848526,
+      "learning_rate": 1.1998047510575985e-05,
+      "loss": 8.7523,
+      "step": 120360
+    },
+    {
+      "epoch": 0.6011136357961497,
+      "grad_norm": 0.08915422856807709,
+      "learning_rate": 1.1996545595634435e-05,
+      "loss": 8.7447,
+      "step": 120370
+    },
+    {
+      "epoch": 0.6011635746210892,
+      "grad_norm": 0.09036835283041,
+      "learning_rate": 1.1995043680692884e-05,
+      "loss": 8.7562,
+      "step": 120380
+    },
+    {
+      "epoch": 0.6012135134460286,
+      "grad_norm": 0.08992304652929306,
+      "learning_rate": 1.1993541765751332e-05,
+      "loss": 8.7521,
+      "step": 120390
+    },
+    {
+      "epoch": 0.6012634522709681,
+      "grad_norm": 0.0943126529455185,
+      "learning_rate": 1.1992039850809782e-05,
+      "loss": 8.7471,
+      "step": 120400
+    },
+    {
+      "epoch": 0.6013133910959075,
+      "grad_norm": 0.09551472961902618,
+      "learning_rate": 1.1990537935868232e-05,
+      "loss": 8.7448,
+      "step": 120410
+    },
+    {
+      "epoch": 0.601363329920847,
+      "grad_norm": 0.08826518803834915,
+      "learning_rate": 1.1989036020926683e-05,
+      "loss": 8.7382,
+      "step": 120420
+    },
+    {
+      "epoch": 0.6014132687457864,
+      "grad_norm": 0.09639231860637665,
+      "learning_rate": 1.1987534105985131e-05,
+      "loss": 8.7435,
+      "step": 120430
+    },
+    {
+      "epoch": 0.6014632075707259,
+      "grad_norm": 0.09125574678182602,
+      "learning_rate": 1.1986032191043581e-05,
+      "loss": 8.7593,
+      "step": 120440
+    },
+    {
+      "epoch": 0.6015131463956653,
+      "grad_norm": 0.08899689465761185,
+      "learning_rate": 1.198453027610203e-05,
+      "loss": 8.761,
+      "step": 120450
+    },
+    {
+      "epoch": 0.6015630852206048,
+      "grad_norm": 0.09570741653442383,
+      "learning_rate": 1.198302836116048e-05,
+      "loss": 8.7517,
+      "step": 120460
+    },
+    {
+      "epoch": 0.6016130240455442,
+      "grad_norm": 0.09493447840213776,
+      "learning_rate": 1.198152644621893e-05,
+      "loss": 8.7354,
+      "step": 120470
+    },
+    {
+      "epoch": 0.6016629628704837,
+      "grad_norm": 0.09143593907356262,
+      "learning_rate": 1.1980024531277379e-05,
+      "loss": 8.7383,
+      "step": 120480
+    },
+    {
+      "epoch": 0.6017129016954231,
+      "grad_norm": 0.09237073361873627,
+      "learning_rate": 1.1978522616335829e-05,
+      "loss": 8.7553,
+      "step": 120490
+    },
+    {
+      "epoch": 0.6017628405203626,
+      "grad_norm": 0.0897151529788971,
+      "learning_rate": 1.1977020701394277e-05,
+      "loss": 8.7465,
+      "step": 120500
+    },
+    {
+      "epoch": 0.601812779345302,
+      "grad_norm": 0.08980879932641983,
+      "learning_rate": 1.1975518786452727e-05,
+      "loss": 8.7424,
+      "step": 120510
+    },
+    {
+      "epoch": 0.6018627181702415,
+      "grad_norm": 0.09323468804359436,
+      "learning_rate": 1.1974016871511178e-05,
+      "loss": 8.7519,
+      "step": 120520
+    },
+    {
+      "epoch": 0.6019126569951809,
+      "grad_norm": 0.092279814183712,
+      "learning_rate": 1.1972514956569626e-05,
+      "loss": 8.7398,
+      "step": 120530
+    },
+    {
+      "epoch": 0.6019625958201203,
+      "grad_norm": 0.09498938918113708,
+      "learning_rate": 1.1971013041628076e-05,
+      "loss": 8.7409,
+      "step": 120540
+    },
+    {
+      "epoch": 0.6020125346450598,
+      "grad_norm": 0.09677129238843918,
+      "learning_rate": 1.1969511126686525e-05,
+      "loss": 8.7443,
+      "step": 120550
+    },
+    {
+      "epoch": 0.6020624734699993,
+      "grad_norm": 0.0944163128733635,
+      "learning_rate": 1.1968009211744975e-05,
+      "loss": 8.745,
+      "step": 120560
+    },
+    {
+      "epoch": 0.6021124122949387,
+      "grad_norm": 0.08608414977788925,
+      "learning_rate": 1.1966507296803425e-05,
+      "loss": 8.7553,
+      "step": 120570
+    },
+    {
+      "epoch": 0.6021623511198781,
+      "grad_norm": 0.094789057970047,
+      "learning_rate": 1.1965005381861874e-05,
+      "loss": 8.7447,
+      "step": 120580
+    },
+    {
+      "epoch": 0.6022122899448176,
+      "grad_norm": 0.09222722053527832,
+      "learning_rate": 1.1963503466920324e-05,
+      "loss": 8.7381,
+      "step": 120590
+    },
+    {
+      "epoch": 0.6022622287697571,
+      "grad_norm": 0.08701804280281067,
+      "learning_rate": 1.1962001551978774e-05,
+      "loss": 8.7559,
+      "step": 120600
+    },
+    {
+      "epoch": 0.6023121675946965,
+      "grad_norm": 0.09150071442127228,
+      "learning_rate": 1.1960499637037222e-05,
+      "loss": 8.735,
+      "step": 120610
+    },
+    {
+      "epoch": 0.602362106419636,
+      "grad_norm": 0.09063006192445755,
+      "learning_rate": 1.1958997722095673e-05,
+      "loss": 8.7467,
+      "step": 120620
+    },
+    {
+      "epoch": 0.6024120452445754,
+      "grad_norm": 0.09188646078109741,
+      "learning_rate": 1.1957495807154121e-05,
+      "loss": 8.752,
+      "step": 120630
+    },
+    {
+      "epoch": 0.6024619840695149,
+      "grad_norm": 0.09145389497280121,
+      "learning_rate": 1.1955993892212571e-05,
+      "loss": 8.767,
+      "step": 120640
+    },
+    {
+      "epoch": 0.6025119228944543,
+      "grad_norm": 0.0874108150601387,
+      "learning_rate": 1.1954491977271021e-05,
+      "loss": 8.7427,
+      "step": 120650
+    },
+    {
+      "epoch": 0.6025618617193937,
+      "grad_norm": 0.09589160233736038,
+      "learning_rate": 1.195299006232947e-05,
+      "loss": 8.7374,
+      "step": 120660
+    },
+    {
+      "epoch": 0.6026118005443332,
+      "grad_norm": 0.0927971825003624,
+      "learning_rate": 1.195148814738792e-05,
+      "loss": 8.7538,
+      "step": 120670
+    },
+    {
+      "epoch": 0.6026617393692726,
+      "grad_norm": 0.09196257591247559,
+      "learning_rate": 1.1949986232446369e-05,
+      "loss": 8.7561,
+      "step": 120680
+    },
+    {
+      "epoch": 0.6027116781942121,
+      "grad_norm": 0.0907442569732666,
+      "learning_rate": 1.1948484317504819e-05,
+      "loss": 8.7335,
+      "step": 120690
+    },
+    {
+      "epoch": 0.6027616170191515,
+      "grad_norm": 0.09081535786390305,
+      "learning_rate": 1.1946982402563269e-05,
+      "loss": 8.7354,
+      "step": 120700
+    },
+    {
+      "epoch": 0.602811555844091,
+      "grad_norm": 0.0993124321103096,
+      "learning_rate": 1.1945480487621717e-05,
+      "loss": 8.7478,
+      "step": 120710
+    },
+    {
+      "epoch": 0.6028614946690304,
+      "grad_norm": 0.09617297351360321,
+      "learning_rate": 1.1943978572680168e-05,
+      "loss": 8.7466,
+      "step": 120720
+    },
+    {
+      "epoch": 0.6029114334939699,
+      "grad_norm": 0.09750398993492126,
+      "learning_rate": 1.1942476657738616e-05,
+      "loss": 8.7461,
+      "step": 120730
+    },
+    {
+      "epoch": 0.6029613723189093,
+      "grad_norm": 0.09524727612733841,
+      "learning_rate": 1.1940974742797066e-05,
+      "loss": 8.7452,
+      "step": 120740
+    },
+    {
+      "epoch": 0.6030113111438488,
+      "grad_norm": 0.08593402057886124,
+      "learning_rate": 1.1939472827855516e-05,
+      "loss": 8.7495,
+      "step": 120750
+    },
+    {
+      "epoch": 0.6030612499687882,
+      "grad_norm": 0.0903341993689537,
+      "learning_rate": 1.1937970912913967e-05,
+      "loss": 8.7438,
+      "step": 120760
+    },
+    {
+      "epoch": 0.6031111887937277,
+      "grad_norm": 0.09396269917488098,
+      "learning_rate": 1.1936468997972415e-05,
+      "loss": 8.7225,
+      "step": 120770
+    },
+    {
+      "epoch": 0.6031611276186671,
+      "grad_norm": 0.09335647523403168,
+      "learning_rate": 1.1934967083030864e-05,
+      "loss": 8.7488,
+      "step": 120780
+    },
+    {
+      "epoch": 0.6032110664436066,
+      "grad_norm": 0.09076468646526337,
+      "learning_rate": 1.1933465168089314e-05,
+      "loss": 8.7418,
+      "step": 120790
+    },
+    {
+      "epoch": 0.603261005268546,
+      "grad_norm": 0.09674550592899323,
+      "learning_rate": 1.1931963253147764e-05,
+      "loss": 8.7379,
+      "step": 120800
+    },
+    {
+      "epoch": 0.6033109440934855,
+      "grad_norm": 0.09369368851184845,
+      "learning_rate": 1.1930461338206214e-05,
+      "loss": 8.7453,
+      "step": 120810
+    },
+    {
+      "epoch": 0.6033608829184249,
+      "grad_norm": 0.09193939715623856,
+      "learning_rate": 1.1928959423264663e-05,
+      "loss": 8.7374,
+      "step": 120820
+    },
+    {
+      "epoch": 0.6034108217433644,
+      "grad_norm": 0.08715419471263885,
+      "learning_rate": 1.1927457508323111e-05,
+      "loss": 8.7427,
+      "step": 120830
+    },
+    {
+      "epoch": 0.6034607605683038,
+      "grad_norm": 0.0879080593585968,
+      "learning_rate": 1.1925955593381561e-05,
+      "loss": 8.7488,
+      "step": 120840
+    },
+    {
+      "epoch": 0.6035106993932433,
+      "grad_norm": 0.09147755801677704,
+      "learning_rate": 1.1924453678440011e-05,
+      "loss": 8.7333,
+      "step": 120850
+    },
+    {
+      "epoch": 0.6035606382181827,
+      "grad_norm": 0.09098514169454575,
+      "learning_rate": 1.1922951763498462e-05,
+      "loss": 8.7172,
+      "step": 120860
+    },
+    {
+      "epoch": 0.6036105770431222,
+      "grad_norm": 0.08796340972185135,
+      "learning_rate": 1.192144984855691e-05,
+      "loss": 8.7379,
+      "step": 120870
+    },
+    {
+      "epoch": 0.6036605158680616,
+      "grad_norm": 0.09227326512336731,
+      "learning_rate": 1.1919947933615359e-05,
+      "loss": 8.739,
+      "step": 120880
+    },
+    {
+      "epoch": 0.6037104546930011,
+      "grad_norm": 0.09190311282873154,
+      "learning_rate": 1.1918446018673809e-05,
+      "loss": 8.7631,
+      "step": 120890
+    },
+    {
+      "epoch": 0.6037603935179405,
+      "grad_norm": 0.09431488811969757,
+      "learning_rate": 1.1916944103732259e-05,
+      "loss": 8.7499,
+      "step": 120900
+    },
+    {
+      "epoch": 0.60381033234288,
+      "grad_norm": 0.09246110171079636,
+      "learning_rate": 1.1915442188790709e-05,
+      "loss": 8.7489,
+      "step": 120910
+    },
+    {
+      "epoch": 0.6038602711678194,
+      "grad_norm": 0.093075692653656,
+      "learning_rate": 1.191394027384916e-05,
+      "loss": 8.7325,
+      "step": 120920
+    },
+    {
+      "epoch": 0.6039102099927589,
+      "grad_norm": 0.09103310108184814,
+      "learning_rate": 1.1912438358907606e-05,
+      "loss": 8.7337,
+      "step": 120930
+    },
+    {
+      "epoch": 0.6039601488176983,
+      "grad_norm": 0.1019456684589386,
+      "learning_rate": 1.1910936443966056e-05,
+      "loss": 8.7356,
+      "step": 120940
+    },
+    {
+      "epoch": 0.6040100876426377,
+      "grad_norm": 0.08999264985322952,
+      "learning_rate": 1.1909434529024506e-05,
+      "loss": 8.7455,
+      "step": 120950
+    },
+    {
+      "epoch": 0.6040600264675772,
+      "grad_norm": 0.09247510880231857,
+      "learning_rate": 1.1907932614082957e-05,
+      "loss": 8.7358,
+      "step": 120960
+    },
+    {
+      "epoch": 0.6041099652925167,
+      "grad_norm": 0.0908162072300911,
+      "learning_rate": 1.1906430699141407e-05,
+      "loss": 8.7354,
+      "step": 120970
+    },
+    {
+      "epoch": 0.6041599041174561,
+      "grad_norm": 0.09764672070741653,
+      "learning_rate": 1.1904928784199855e-05,
+      "loss": 8.7378,
+      "step": 120980
+    },
+    {
+      "epoch": 0.6042098429423955,
+      "grad_norm": 0.09343189001083374,
+      "learning_rate": 1.1903426869258304e-05,
+      "loss": 8.7326,
+      "step": 120990
+    },
+    {
+      "epoch": 0.604259781767335,
+      "grad_norm": 0.0914471447467804,
+      "learning_rate": 1.1901924954316754e-05,
+      "loss": 8.7343,
+      "step": 121000
+    },
+    {
+      "epoch": 0.6043097205922745,
+      "grad_norm": 0.09239111840724945,
+      "learning_rate": 1.1900423039375204e-05,
+      "loss": 8.7356,
+      "step": 121010
+    },
+    {
+      "epoch": 0.6043596594172139,
+      "grad_norm": 0.09467372298240662,
+      "learning_rate": 1.1898921124433654e-05,
+      "loss": 8.7418,
+      "step": 121020
+    },
+    {
+      "epoch": 0.6044095982421533,
+      "grad_norm": 0.09846384823322296,
+      "learning_rate": 1.1897419209492103e-05,
+      "loss": 8.7513,
+      "step": 121030
+    },
+    {
+      "epoch": 0.6044595370670928,
+      "grad_norm": 0.09008269757032394,
+      "learning_rate": 1.1895917294550551e-05,
+      "loss": 8.7377,
+      "step": 121040
+    },
+    {
+      "epoch": 0.6045094758920323,
+      "grad_norm": 0.0888042226433754,
+      "learning_rate": 1.1894415379609001e-05,
+      "loss": 8.7362,
+      "step": 121050
+    },
+    {
+      "epoch": 0.6045594147169717,
+      "grad_norm": 0.09224848449230194,
+      "learning_rate": 1.1892913464667452e-05,
+      "loss": 8.768,
+      "step": 121060
+    },
+    {
+      "epoch": 0.6046093535419111,
+      "grad_norm": 0.0956854522228241,
+      "learning_rate": 1.1891411549725902e-05,
+      "loss": 8.7352,
+      "step": 121070
+    },
+    {
+      "epoch": 0.6046592923668506,
+      "grad_norm": 0.09198862314224243,
+      "learning_rate": 1.188990963478435e-05,
+      "loss": 8.7352,
+      "step": 121080
+    },
+    {
+      "epoch": 0.6047092311917901,
+      "grad_norm": 0.09782185405492783,
+      "learning_rate": 1.1888407719842799e-05,
+      "loss": 8.7306,
+      "step": 121090
+    },
+    {
+      "epoch": 0.6047591700167295,
+      "grad_norm": 0.0954626128077507,
+      "learning_rate": 1.1886905804901249e-05,
+      "loss": 8.7577,
+      "step": 121100
+    },
+    {
+      "epoch": 0.6048091088416689,
+      "grad_norm": 0.0895036831498146,
+      "learning_rate": 1.1885403889959699e-05,
+      "loss": 8.7356,
+      "step": 121110
+    },
+    {
+      "epoch": 0.6048590476666084,
+      "grad_norm": 0.09198148548603058,
+      "learning_rate": 1.188390197501815e-05,
+      "loss": 8.743,
+      "step": 121120
+    },
+    {
+      "epoch": 0.6049089864915479,
+      "grad_norm": 0.0872054398059845,
+      "learning_rate": 1.18824000600766e-05,
+      "loss": 8.7409,
+      "step": 121130
+    },
+    {
+      "epoch": 0.6049589253164873,
+      "grad_norm": 0.0946134701371193,
+      "learning_rate": 1.1880898145135046e-05,
+      "loss": 8.7335,
+      "step": 121140
+    },
+    {
+      "epoch": 0.6050088641414267,
+      "grad_norm": 0.08696244657039642,
+      "learning_rate": 1.1879396230193496e-05,
+      "loss": 8.7231,
+      "step": 121150
+    },
+    {
+      "epoch": 0.6050588029663662,
+      "grad_norm": 0.09812266379594803,
+      "learning_rate": 1.1877894315251947e-05,
+      "loss": 8.7275,
+      "step": 121160
+    },
+    {
+      "epoch": 0.6051087417913057,
+      "grad_norm": 0.09273399412631989,
+      "learning_rate": 1.1876392400310397e-05,
+      "loss": 8.738,
+      "step": 121170
+    },
+    {
+      "epoch": 0.6051586806162451,
+      "grad_norm": 0.09106171131134033,
+      "learning_rate": 1.1874890485368847e-05,
+      "loss": 8.7407,
+      "step": 121180
+    },
+    {
+      "epoch": 0.6052086194411845,
+      "grad_norm": 0.08590860664844513,
+      "learning_rate": 1.1873388570427294e-05,
+      "loss": 8.743,
+      "step": 121190
+    },
+    {
+      "epoch": 0.605258558266124,
+      "grad_norm": 0.09189659357070923,
+      "learning_rate": 1.1871886655485744e-05,
+      "loss": 8.7424,
+      "step": 121200
+    },
+    {
+      "epoch": 0.6053084970910635,
+      "grad_norm": 0.10102283954620361,
+      "learning_rate": 1.1870384740544194e-05,
+      "loss": 8.7519,
+      "step": 121210
+    },
+    {
+      "epoch": 0.6053584359160029,
+      "grad_norm": 0.09137275069952011,
+      "learning_rate": 1.1868882825602644e-05,
+      "loss": 8.7475,
+      "step": 121220
+    },
+    {
+      "epoch": 0.6054083747409423,
+      "grad_norm": 0.09499847143888474,
+      "learning_rate": 1.1867380910661094e-05,
+      "loss": 8.7518,
+      "step": 121230
+    },
+    {
+      "epoch": 0.6054583135658818,
+      "grad_norm": 0.09718917310237885,
+      "learning_rate": 1.1865878995719543e-05,
+      "loss": 8.7258,
+      "step": 121240
+    },
+    {
+      "epoch": 0.6055082523908213,
+      "grad_norm": 0.08984164148569107,
+      "learning_rate": 1.1864377080777991e-05,
+      "loss": 8.751,
+      "step": 121250
+    },
+    {
+      "epoch": 0.6055581912157607,
+      "grad_norm": 0.09550242125988007,
+      "learning_rate": 1.1862875165836442e-05,
+      "loss": 8.7504,
+      "step": 121260
+    },
+    {
+      "epoch": 0.6056081300407001,
+      "grad_norm": 0.09198052436113358,
+      "learning_rate": 1.1861373250894892e-05,
+      "loss": 8.7366,
+      "step": 121270
+    },
+    {
+      "epoch": 0.6056580688656396,
+      "grad_norm": 0.09440373629331589,
+      "learning_rate": 1.1859871335953342e-05,
+      "loss": 8.7489,
+      "step": 121280
+    },
+    {
+      "epoch": 0.6057080076905791,
+      "grad_norm": 0.09069590270519257,
+      "learning_rate": 1.185836942101179e-05,
+      "loss": 8.7381,
+      "step": 121290
+    },
+    {
+      "epoch": 0.6057579465155185,
+      "grad_norm": 0.09249866008758545,
+      "learning_rate": 1.1856867506070239e-05,
+      "loss": 8.7219,
+      "step": 121300
+    },
+    {
+      "epoch": 0.6058078853404579,
+      "grad_norm": 0.09518305212259293,
+      "learning_rate": 1.1855365591128689e-05,
+      "loss": 8.743,
+      "step": 121310
+    },
+    {
+      "epoch": 0.6058578241653974,
+      "grad_norm": 0.09274056553840637,
+      "learning_rate": 1.185386367618714e-05,
+      "loss": 8.7459,
+      "step": 121320
+    },
+    {
+      "epoch": 0.6059077629903369,
+      "grad_norm": 0.09070045500993729,
+      "learning_rate": 1.185236176124559e-05,
+      "loss": 8.7359,
+      "step": 121330
+    },
+    {
+      "epoch": 0.6059577018152763,
+      "grad_norm": 0.09082917869091034,
+      "learning_rate": 1.1850859846304038e-05,
+      "loss": 8.742,
+      "step": 121340
+    },
+    {
+      "epoch": 0.6060076406402157,
+      "grad_norm": 0.09195704758167267,
+      "learning_rate": 1.1849357931362486e-05,
+      "loss": 8.7441,
+      "step": 121350
+    },
+    {
+      "epoch": 0.6060575794651551,
+      "grad_norm": 0.0921049565076828,
+      "learning_rate": 1.1847856016420937e-05,
+      "loss": 8.73,
+      "step": 121360
+    },
+    {
+      "epoch": 0.6061075182900947,
+      "grad_norm": 0.09024003148078918,
+      "learning_rate": 1.1846354101479387e-05,
+      "loss": 8.7474,
+      "step": 121370
+    },
+    {
+      "epoch": 0.6061574571150341,
+      "grad_norm": 0.09810420870780945,
+      "learning_rate": 1.1844852186537837e-05,
+      "loss": 8.7297,
+      "step": 121380
+    },
+    {
+      "epoch": 0.6062073959399735,
+      "grad_norm": 0.08957557380199432,
+      "learning_rate": 1.1843350271596285e-05,
+      "loss": 8.7334,
+      "step": 121390
+    },
+    {
+      "epoch": 0.606257334764913,
+      "grad_norm": 0.095646932721138,
+      "learning_rate": 1.1841848356654736e-05,
+      "loss": 8.7491,
+      "step": 121400
+    },
+    {
+      "epoch": 0.6063072735898525,
+      "grad_norm": 0.09269688278436661,
+      "learning_rate": 1.1840346441713184e-05,
+      "loss": 8.7294,
+      "step": 121410
+    },
+    {
+      "epoch": 0.6063572124147919,
+      "grad_norm": 0.09186460822820663,
+      "learning_rate": 1.1838844526771634e-05,
+      "loss": 8.7326,
+      "step": 121420
+    },
+    {
+      "epoch": 0.6064071512397313,
+      "grad_norm": 0.09095993638038635,
+      "learning_rate": 1.1837342611830084e-05,
+      "loss": 8.7583,
+      "step": 121430
+    },
+    {
+      "epoch": 0.6064570900646707,
+      "grad_norm": 0.0951274111866951,
+      "learning_rate": 1.1835840696888533e-05,
+      "loss": 8.7412,
+      "step": 121440
+    },
+    {
+      "epoch": 0.6065070288896103,
+      "grad_norm": 0.09069649875164032,
+      "learning_rate": 1.1834338781946983e-05,
+      "loss": 8.7235,
+      "step": 121450
+    },
+    {
+      "epoch": 0.6065569677145497,
+      "grad_norm": 0.09192630648612976,
+      "learning_rate": 1.1832836867005432e-05,
+      "loss": 8.7422,
+      "step": 121460
+    },
+    {
+      "epoch": 0.6066069065394891,
+      "grad_norm": 0.09435165673494339,
+      "learning_rate": 1.1831334952063882e-05,
+      "loss": 8.7295,
+      "step": 121470
+    },
+    {
+      "epoch": 0.6066568453644285,
+      "grad_norm": 0.09608633071184158,
+      "learning_rate": 1.1829833037122332e-05,
+      "loss": 8.7352,
+      "step": 121480
+    },
+    {
+      "epoch": 0.6067067841893681,
+      "grad_norm": 0.09027303010225296,
+      "learning_rate": 1.182833112218078e-05,
+      "loss": 8.7404,
+      "step": 121490
+    },
+    {
+      "epoch": 0.6067567230143075,
+      "grad_norm": 0.08934234827756882,
+      "learning_rate": 1.182682920723923e-05,
+      "loss": 8.7349,
+      "step": 121500
+    },
+    {
+      "epoch": 0.6068066618392469,
+      "grad_norm": 0.09478524327278137,
+      "learning_rate": 1.1825327292297679e-05,
+      "loss": 8.754,
+      "step": 121510
+    },
+    {
+      "epoch": 0.6068566006641863,
+      "grad_norm": 0.09414006769657135,
+      "learning_rate": 1.182382537735613e-05,
+      "loss": 8.7425,
+      "step": 121520
+    },
+    {
+      "epoch": 0.6069065394891259,
+      "grad_norm": 0.09682093560695648,
+      "learning_rate": 1.182232346241458e-05,
+      "loss": 8.7536,
+      "step": 121530
+    },
+    {
+      "epoch": 0.6069564783140653,
+      "grad_norm": 0.09124701470136642,
+      "learning_rate": 1.1820821547473028e-05,
+      "loss": 8.7449,
+      "step": 121540
+    },
+    {
+      "epoch": 0.6070064171390047,
+      "grad_norm": 0.08872292190790176,
+      "learning_rate": 1.1819319632531478e-05,
+      "loss": 8.7353,
+      "step": 121550
+    },
+    {
+      "epoch": 0.6070563559639441,
+      "grad_norm": 0.09001273661851883,
+      "learning_rate": 1.1817817717589928e-05,
+      "loss": 8.7428,
+      "step": 121560
+    },
+    {
+      "epoch": 0.6071062947888837,
+      "grad_norm": 0.08708401024341583,
+      "learning_rate": 1.1816315802648377e-05,
+      "loss": 8.7396,
+      "step": 121570
+    },
+    {
+      "epoch": 0.6071562336138231,
+      "grad_norm": 0.09001478552818298,
+      "learning_rate": 1.1814813887706827e-05,
+      "loss": 8.7425,
+      "step": 121580
+    },
+    {
+      "epoch": 0.6072061724387625,
+      "grad_norm": 0.0907810628414154,
+      "learning_rate": 1.1813311972765275e-05,
+      "loss": 8.7552,
+      "step": 121590
+    },
+    {
+      "epoch": 0.6072561112637019,
+      "grad_norm": 0.08939887583255768,
+      "learning_rate": 1.1811810057823726e-05,
+      "loss": 8.7563,
+      "step": 121600
+    },
+    {
+      "epoch": 0.6073060500886415,
+      "grad_norm": 0.09514924883842468,
+      "learning_rate": 1.1810308142882176e-05,
+      "loss": 8.7546,
+      "step": 121610
+    },
+    {
+      "epoch": 0.6073559889135809,
+      "grad_norm": 0.09000418335199356,
+      "learning_rate": 1.1808806227940624e-05,
+      "loss": 8.7427,
+      "step": 121620
+    },
+    {
+      "epoch": 0.6074059277385203,
+      "grad_norm": 0.09303897619247437,
+      "learning_rate": 1.1807304312999075e-05,
+      "loss": 8.7311,
+      "step": 121630
+    },
+    {
+      "epoch": 0.6074558665634597,
+      "grad_norm": 0.09306817501783371,
+      "learning_rate": 1.1805802398057523e-05,
+      "loss": 8.7395,
+      "step": 121640
+    },
+    {
+      "epoch": 0.6075058053883992,
+      "grad_norm": 0.08681345731019974,
+      "learning_rate": 1.1804300483115973e-05,
+      "loss": 8.7435,
+      "step": 121650
+    },
+    {
+      "epoch": 0.6075557442133387,
+      "grad_norm": 0.09274129569530487,
+      "learning_rate": 1.1802798568174423e-05,
+      "loss": 8.7352,
+      "step": 121660
+    },
+    {
+      "epoch": 0.6076056830382781,
+      "grad_norm": 0.090678870677948,
+      "learning_rate": 1.1801296653232872e-05,
+      "loss": 8.7323,
+      "step": 121670
+    },
+    {
+      "epoch": 0.6076556218632175,
+      "grad_norm": 0.09506350755691528,
+      "learning_rate": 1.1799794738291322e-05,
+      "loss": 8.7349,
+      "step": 121680
+    },
+    {
+      "epoch": 0.607705560688157,
+      "grad_norm": 0.09269612282514572,
+      "learning_rate": 1.179829282334977e-05,
+      "loss": 8.7393,
+      "step": 121690
+    },
+    {
+      "epoch": 0.6077554995130965,
+      "grad_norm": 0.09288372099399567,
+      "learning_rate": 1.179679090840822e-05,
+      "loss": 8.7256,
+      "step": 121700
+    },
+    {
+      "epoch": 0.6078054383380359,
+      "grad_norm": 0.09377559274435043,
+      "learning_rate": 1.179528899346667e-05,
+      "loss": 8.7359,
+      "step": 121710
+    },
+    {
+      "epoch": 0.6078553771629753,
+      "grad_norm": 0.088566854596138,
+      "learning_rate": 1.1793787078525121e-05,
+      "loss": 8.7409,
+      "step": 121720
+    },
+    {
+      "epoch": 0.6079053159879148,
+      "grad_norm": 0.09255025535821915,
+      "learning_rate": 1.179228516358357e-05,
+      "loss": 8.7415,
+      "step": 121730
+    },
+    {
+      "epoch": 0.6079552548128543,
+      "grad_norm": 0.09494113177061081,
+      "learning_rate": 1.1790783248642018e-05,
+      "loss": 8.7397,
+      "step": 121740
+    },
+    {
+      "epoch": 0.6080051936377937,
+      "grad_norm": 0.08850625157356262,
+      "learning_rate": 1.1789281333700468e-05,
+      "loss": 8.7394,
+      "step": 121750
+    },
+    {
+      "epoch": 0.6080551324627331,
+      "grad_norm": 0.09062092751264572,
+      "learning_rate": 1.1787779418758918e-05,
+      "loss": 8.7512,
+      "step": 121760
+    },
+    {
+      "epoch": 0.6081050712876725,
+      "grad_norm": 0.0890820100903511,
+      "learning_rate": 1.1786277503817369e-05,
+      "loss": 8.7326,
+      "step": 121770
+    },
+    {
+      "epoch": 0.6081550101126121,
+      "grad_norm": 0.09788678586483002,
+      "learning_rate": 1.1784775588875817e-05,
+      "loss": 8.7336,
+      "step": 121780
+    },
+    {
+      "epoch": 0.6082049489375515,
+      "grad_norm": 0.09738252311944962,
+      "learning_rate": 1.1783273673934265e-05,
+      "loss": 8.7479,
+      "step": 121790
+    },
+    {
+      "epoch": 0.6082548877624909,
+      "grad_norm": 0.09170432388782501,
+      "learning_rate": 1.1781771758992716e-05,
+      "loss": 8.7248,
+      "step": 121800
+    },
+    {
+      "epoch": 0.6083048265874303,
+      "grad_norm": 0.0888880267739296,
+      "learning_rate": 1.1780269844051166e-05,
+      "loss": 8.7402,
+      "step": 121810
+    },
+    {
+      "epoch": 0.6083547654123699,
+      "grad_norm": 0.09058916568756104,
+      "learning_rate": 1.1778767929109616e-05,
+      "loss": 8.7334,
+      "step": 121820
+    },
+    {
+      "epoch": 0.6084047042373093,
+      "grad_norm": 0.09217052161693573,
+      "learning_rate": 1.1777266014168065e-05,
+      "loss": 8.7499,
+      "step": 121830
+    },
+    {
+      "epoch": 0.6084546430622487,
+      "grad_norm": 0.09062588959932327,
+      "learning_rate": 1.1775764099226513e-05,
+      "loss": 8.7287,
+      "step": 121840
+    },
+    {
+      "epoch": 0.6085045818871881,
+      "grad_norm": 0.09828715771436691,
+      "learning_rate": 1.1774262184284963e-05,
+      "loss": 8.7276,
+      "step": 121850
+    },
+    {
+      "epoch": 0.6085545207121277,
+      "grad_norm": 0.08966988325119019,
+      "learning_rate": 1.1772760269343413e-05,
+      "loss": 8.7422,
+      "step": 121860
+    },
+    {
+      "epoch": 0.6086044595370671,
+      "grad_norm": 0.08841178566217422,
+      "learning_rate": 1.1771258354401864e-05,
+      "loss": 8.748,
+      "step": 121870
+    },
+    {
+      "epoch": 0.6086543983620065,
+      "grad_norm": 0.08782627433538437,
+      "learning_rate": 1.1769756439460314e-05,
+      "loss": 8.7351,
+      "step": 121880
+    },
+    {
+      "epoch": 0.6087043371869459,
+      "grad_norm": 0.08963312953710556,
+      "learning_rate": 1.176825452451876e-05,
+      "loss": 8.7342,
+      "step": 121890
+    },
+    {
+      "epoch": 0.6087542760118855,
+      "grad_norm": 0.08687534928321838,
+      "learning_rate": 1.176675260957721e-05,
+      "loss": 8.7416,
+      "step": 121900
+    },
+    {
+      "epoch": 0.6088042148368249,
+      "grad_norm": 0.08947011828422546,
+      "learning_rate": 1.1765250694635661e-05,
+      "loss": 8.7506,
+      "step": 121910
+    },
+    {
+      "epoch": 0.6088541536617643,
+      "grad_norm": 0.09872882813215256,
+      "learning_rate": 1.1763748779694111e-05,
+      "loss": 8.7542,
+      "step": 121920
+    },
+    {
+      "epoch": 0.6089040924867037,
+      "grad_norm": 0.09389559924602509,
+      "learning_rate": 1.1762246864752561e-05,
+      "loss": 8.7213,
+      "step": 121930
+    },
+    {
+      "epoch": 0.6089540313116433,
+      "grad_norm": 0.09409285336732864,
+      "learning_rate": 1.1760744949811008e-05,
+      "loss": 8.7131,
+      "step": 121940
+    },
+    {
+      "epoch": 0.6090039701365827,
+      "grad_norm": 0.08693234622478485,
+      "learning_rate": 1.1759243034869458e-05,
+      "loss": 8.7416,
+      "step": 121950
+    },
+    {
+      "epoch": 0.6090539089615221,
+      "grad_norm": 0.09217488020658493,
+      "learning_rate": 1.1757741119927908e-05,
+      "loss": 8.7226,
+      "step": 121960
+    },
+    {
+      "epoch": 0.6091038477864615,
+      "grad_norm": 0.09177994728088379,
+      "learning_rate": 1.1756239204986359e-05,
+      "loss": 8.7294,
+      "step": 121970
+    },
+    {
+      "epoch": 0.6091537866114011,
+      "grad_norm": 0.09113775938749313,
+      "learning_rate": 1.1754737290044809e-05,
+      "loss": 8.7382,
+      "step": 121980
+    },
+    {
+      "epoch": 0.6092037254363405,
+      "grad_norm": 0.09432590007781982,
+      "learning_rate": 1.1753235375103256e-05,
+      "loss": 8.7495,
+      "step": 121990
+    },
+    {
+      "epoch": 0.6092536642612799,
+      "grad_norm": 0.09742003679275513,
+      "learning_rate": 1.1751733460161706e-05,
+      "loss": 8.7237,
+      "step": 122000
+    },
+    {
+      "epoch": 0.6093036030862193,
+      "grad_norm": 0.09325713664293289,
+      "learning_rate": 1.1750231545220156e-05,
+      "loss": 8.7452,
+      "step": 122010
+    },
+    {
+      "epoch": 0.6093535419111589,
+      "grad_norm": 0.09275127202272415,
+      "learning_rate": 1.1748729630278606e-05,
+      "loss": 8.7461,
+      "step": 122020
+    },
+    {
+      "epoch": 0.6094034807360983,
+      "grad_norm": 0.09381189197301865,
+      "learning_rate": 1.1747227715337056e-05,
+      "loss": 8.7323,
+      "step": 122030
+    },
+    {
+      "epoch": 0.6094534195610377,
+      "grad_norm": 0.09303104132413864,
+      "learning_rate": 1.1745725800395505e-05,
+      "loss": 8.7343,
+      "step": 122040
+    },
+    {
+      "epoch": 0.6095033583859771,
+      "grad_norm": 0.09670752286911011,
+      "learning_rate": 1.1744223885453953e-05,
+      "loss": 8.7394,
+      "step": 122050
+    },
+    {
+      "epoch": 0.6095532972109167,
+      "grad_norm": 0.09118695557117462,
+      "learning_rate": 1.1742721970512403e-05,
+      "loss": 8.7528,
+      "step": 122060
+    },
+    {
+      "epoch": 0.6096032360358561,
+      "grad_norm": 0.09087878465652466,
+      "learning_rate": 1.1741220055570854e-05,
+      "loss": 8.7145,
+      "step": 122070
+    },
+    {
+      "epoch": 0.6096531748607955,
+      "grad_norm": 0.0934610515832901,
+      "learning_rate": 1.1739718140629304e-05,
+      "loss": 8.7274,
+      "step": 122080
+    },
+    {
+      "epoch": 0.6097031136857349,
+      "grad_norm": 0.0913190245628357,
+      "learning_rate": 1.1738216225687752e-05,
+      "loss": 8.7205,
+      "step": 122090
+    },
+    {
+      "epoch": 0.6097530525106745,
+      "grad_norm": 0.09773539006710052,
+      "learning_rate": 1.17367143107462e-05,
+      "loss": 8.738,
+      "step": 122100
+    },
+    {
+      "epoch": 0.6098029913356139,
+      "grad_norm": 0.09631845355033875,
+      "learning_rate": 1.1735212395804651e-05,
+      "loss": 8.7265,
+      "step": 122110
+    },
+    {
+      "epoch": 0.6098529301605533,
+      "grad_norm": 0.08930249512195587,
+      "learning_rate": 1.1733710480863101e-05,
+      "loss": 8.7458,
+      "step": 122120
+    },
+    {
+      "epoch": 0.6099028689854927,
+      "grad_norm": 0.08900859951972961,
+      "learning_rate": 1.1732208565921551e-05,
+      "loss": 8.7685,
+      "step": 122130
+    },
+    {
+      "epoch": 0.6099528078104323,
+      "grad_norm": 0.0874401181936264,
+      "learning_rate": 1.173070665098e-05,
+      "loss": 8.739,
+      "step": 122140
+    },
+    {
+      "epoch": 0.6100027466353717,
+      "grad_norm": 0.10340723395347595,
+      "learning_rate": 1.1729204736038448e-05,
+      "loss": 8.7406,
+      "step": 122150
+    },
+    {
+      "epoch": 0.6100526854603111,
+      "grad_norm": 0.09558750689029694,
+      "learning_rate": 1.1727702821096898e-05,
+      "loss": 8.7341,
+      "step": 122160
+    },
+    {
+      "epoch": 0.6101026242852505,
+      "grad_norm": 0.09469115734100342,
+      "learning_rate": 1.1726200906155349e-05,
+      "loss": 8.7295,
+      "step": 122170
+    },
+    {
+      "epoch": 0.6101525631101901,
+      "grad_norm": 0.09226387739181519,
+      "learning_rate": 1.1724698991213799e-05,
+      "loss": 8.7433,
+      "step": 122180
+    },
+    {
+      "epoch": 0.6102025019351295,
+      "grad_norm": 0.09578849375247955,
+      "learning_rate": 1.1723197076272247e-05,
+      "loss": 8.7148,
+      "step": 122190
+    },
+    {
+      "epoch": 0.6102524407600689,
+      "grad_norm": 0.08702340722084045,
+      "learning_rate": 1.1721695161330697e-05,
+      "loss": 8.7389,
+      "step": 122200
+    },
+    {
+      "epoch": 0.6103023795850083,
+      "grad_norm": 0.0902688279747963,
+      "learning_rate": 1.1720193246389146e-05,
+      "loss": 8.7441,
+      "step": 122210
+    },
+    {
+      "epoch": 0.6103523184099479,
+      "grad_norm": 0.08830910176038742,
+      "learning_rate": 1.1718691331447596e-05,
+      "loss": 8.7218,
+      "step": 122220
+    },
+    {
+      "epoch": 0.6104022572348873,
+      "grad_norm": 0.09016358107328415,
+      "learning_rate": 1.1717189416506046e-05,
+      "loss": 8.7398,
+      "step": 122230
+    },
+    {
+      "epoch": 0.6104521960598267,
+      "grad_norm": 0.09306544810533524,
+      "learning_rate": 1.1715687501564495e-05,
+      "loss": 8.7381,
+      "step": 122240
+    },
+    {
+      "epoch": 0.6105021348847661,
+      "grad_norm": 0.09133855253458023,
+      "learning_rate": 1.1714185586622945e-05,
+      "loss": 8.7466,
+      "step": 122250
+    },
+    {
+      "epoch": 0.6105520737097057,
+      "grad_norm": 0.09112872183322906,
+      "learning_rate": 1.1712683671681393e-05,
+      "loss": 8.7304,
+      "step": 122260
+    },
+    {
+      "epoch": 0.6106020125346451,
+      "grad_norm": 0.08973613381385803,
+      "learning_rate": 1.1711181756739844e-05,
+      "loss": 8.7265,
+      "step": 122270
+    },
+    {
+      "epoch": 0.6106519513595845,
+      "grad_norm": 0.08785367757081985,
+      "learning_rate": 1.1709679841798294e-05,
+      "loss": 8.7227,
+      "step": 122280
+    },
+    {
+      "epoch": 0.6107018901845239,
+      "grad_norm": 0.09560448676347733,
+      "learning_rate": 1.1708177926856742e-05,
+      "loss": 8.7347,
+      "step": 122290
+    },
+    {
+      "epoch": 0.6107518290094635,
+      "grad_norm": 0.09626337140798569,
+      "learning_rate": 1.1706676011915192e-05,
+      "loss": 8.7301,
+      "step": 122300
+    },
+    {
+      "epoch": 0.6108017678344029,
+      "grad_norm": 0.08685650676488876,
+      "learning_rate": 1.1705174096973641e-05,
+      "loss": 8.7399,
+      "step": 122310
+    },
+    {
+      "epoch": 0.6108517066593423,
+      "grad_norm": 0.09066285192966461,
+      "learning_rate": 1.1703672182032091e-05,
+      "loss": 8.7531,
+      "step": 122320
+    },
+    {
+      "epoch": 0.6109016454842817,
+      "grad_norm": 0.09555506706237793,
+      "learning_rate": 1.1702170267090541e-05,
+      "loss": 8.7348,
+      "step": 122330
+    },
+    {
+      "epoch": 0.6109515843092213,
+      "grad_norm": 0.09356008470058441,
+      "learning_rate": 1.170066835214899e-05,
+      "loss": 8.7236,
+      "step": 122340
+    },
+    {
+      "epoch": 0.6110015231341607,
+      "grad_norm": 0.0910077691078186,
+      "learning_rate": 1.169916643720744e-05,
+      "loss": 8.7322,
+      "step": 122350
+    },
+    {
+      "epoch": 0.6110514619591001,
+      "grad_norm": 0.09046470373868942,
+      "learning_rate": 1.169766452226589e-05,
+      "loss": 8.7302,
+      "step": 122360
+    },
+    {
+      "epoch": 0.6111014007840395,
+      "grad_norm": 0.09235422313213348,
+      "learning_rate": 1.1696162607324339e-05,
+      "loss": 8.7346,
+      "step": 122370
+    },
+    {
+      "epoch": 0.611151339608979,
+      "grad_norm": 0.09250746667385101,
+      "learning_rate": 1.1694660692382789e-05,
+      "loss": 8.747,
+      "step": 122380
+    },
+    {
+      "epoch": 0.6112012784339185,
+      "grad_norm": 0.09652043879032135,
+      "learning_rate": 1.1693158777441237e-05,
+      "loss": 8.7317,
+      "step": 122390
+    },
+    {
+      "epoch": 0.6112512172588579,
+      "grad_norm": 0.0955178514122963,
+      "learning_rate": 1.1691656862499687e-05,
+      "loss": 8.7403,
+      "step": 122400
+    },
+    {
+      "epoch": 0.6113011560837973,
+      "grad_norm": 0.08899587392807007,
+      "learning_rate": 1.1690154947558138e-05,
+      "loss": 8.7483,
+      "step": 122410
+    },
+    {
+      "epoch": 0.6113510949087368,
+      "grad_norm": 0.08871007710695267,
+      "learning_rate": 1.1688653032616586e-05,
+      "loss": 8.7446,
+      "step": 122420
+    },
+    {
+      "epoch": 0.6114010337336763,
+      "grad_norm": 0.09327740967273712,
+      "learning_rate": 1.1687151117675036e-05,
+      "loss": 8.7326,
+      "step": 122430
+    },
+    {
+      "epoch": 0.6114509725586157,
+      "grad_norm": 0.09347925335168839,
+      "learning_rate": 1.1685649202733485e-05,
+      "loss": 8.7262,
+      "step": 122440
+    },
+    {
+      "epoch": 0.6115009113835551,
+      "grad_norm": 0.08919844776391983,
+      "learning_rate": 1.1684147287791935e-05,
+      "loss": 8.7354,
+      "step": 122450
+    },
+    {
+      "epoch": 0.6115508502084946,
+      "grad_norm": 0.09506843984127045,
+      "learning_rate": 1.1682645372850385e-05,
+      "loss": 8.7218,
+      "step": 122460
+    },
+    {
+      "epoch": 0.6116007890334341,
+      "grad_norm": 0.09201342612504959,
+      "learning_rate": 1.1681143457908834e-05,
+      "loss": 8.7338,
+      "step": 122470
+    },
+    {
+      "epoch": 0.6116507278583735,
+      "grad_norm": 0.09908469021320343,
+      "learning_rate": 1.1679641542967284e-05,
+      "loss": 8.7404,
+      "step": 122480
+    },
+    {
+      "epoch": 0.6117006666833129,
+      "grad_norm": 0.09065460413694382,
+      "learning_rate": 1.1678139628025732e-05,
+      "loss": 8.7288,
+      "step": 122490
+    },
+    {
+      "epoch": 0.6117506055082524,
+      "grad_norm": 0.09107660502195358,
+      "learning_rate": 1.1676637713084182e-05,
+      "loss": 8.7291,
+      "step": 122500
+    },
+    {
+      "epoch": 0.6118005443331919,
+      "grad_norm": 0.0926177054643631,
+      "learning_rate": 1.1675135798142633e-05,
+      "loss": 8.734,
+      "step": 122510
+    },
+    {
+      "epoch": 0.6118504831581313,
+      "grad_norm": 0.09085045754909515,
+      "learning_rate": 1.1673633883201083e-05,
+      "loss": 8.7309,
+      "step": 122520
+    },
+    {
+      "epoch": 0.6119004219830707,
+      "grad_norm": 0.09174468368291855,
+      "learning_rate": 1.1672131968259531e-05,
+      "loss": 8.7436,
+      "step": 122530
+    },
+    {
+      "epoch": 0.6119503608080102,
+      "grad_norm": 0.09456495195627213,
+      "learning_rate": 1.167063005331798e-05,
+      "loss": 8.7272,
+      "step": 122540
+    },
+    {
+      "epoch": 0.6120002996329497,
+      "grad_norm": 0.09345948696136475,
+      "learning_rate": 1.166912813837643e-05,
+      "loss": 8.7226,
+      "step": 122550
+    },
+    {
+      "epoch": 0.6120502384578891,
+      "grad_norm": 0.0974060595035553,
+      "learning_rate": 1.166762622343488e-05,
+      "loss": 8.7425,
+      "step": 122560
+    },
+    {
+      "epoch": 0.6121001772828285,
+      "grad_norm": 0.09621378034353256,
+      "learning_rate": 1.166612430849333e-05,
+      "loss": 8.7256,
+      "step": 122570
+    },
+    {
+      "epoch": 0.612150116107768,
+      "grad_norm": 0.09098070859909058,
+      "learning_rate": 1.1664622393551779e-05,
+      "loss": 8.7195,
+      "step": 122580
+    },
+    {
+      "epoch": 0.6122000549327075,
+      "grad_norm": 0.08891347795724869,
+      "learning_rate": 1.1663120478610227e-05,
+      "loss": 8.739,
+      "step": 122590
+    },
+    {
+      "epoch": 0.6122499937576469,
+      "grad_norm": 0.09389282763004303,
+      "learning_rate": 1.1661618563668677e-05,
+      "loss": 8.717,
+      "step": 122600
+    },
+    {
+      "epoch": 0.6122999325825863,
+      "grad_norm": 0.0891786739230156,
+      "learning_rate": 1.1660116648727128e-05,
+      "loss": 8.7401,
+      "step": 122610
+    },
+    {
+      "epoch": 0.6123498714075257,
+      "grad_norm": 0.09159950911998749,
+      "learning_rate": 1.1658614733785578e-05,
+      "loss": 8.7326,
+      "step": 122620
+    },
+    {
+      "epoch": 0.6123998102324653,
+      "grad_norm": 0.0896199643611908,
+      "learning_rate": 1.1657112818844026e-05,
+      "loss": 8.7444,
+      "step": 122630
+    },
+    {
+      "epoch": 0.6124497490574047,
+      "grad_norm": 0.0880565494298935,
+      "learning_rate": 1.1655610903902475e-05,
+      "loss": 8.7348,
+      "step": 122640
+    },
+    {
+      "epoch": 0.6124996878823441,
+      "grad_norm": 0.08740682154893875,
+      "learning_rate": 1.1654108988960925e-05,
+      "loss": 8.7177,
+      "step": 122650
+    },
+    {
+      "epoch": 0.6125496267072835,
+      "grad_norm": 0.09321615844964981,
+      "learning_rate": 1.1652607074019375e-05,
+      "loss": 8.7422,
+      "step": 122660
+    },
+    {
+      "epoch": 0.6125995655322231,
+      "grad_norm": 0.09611321985721588,
+      "learning_rate": 1.1651105159077825e-05,
+      "loss": 8.7159,
+      "step": 122670
+    },
+    {
+      "epoch": 0.6126495043571625,
+      "grad_norm": 0.08695535361766815,
+      "learning_rate": 1.1649603244136275e-05,
+      "loss": 8.7261,
+      "step": 122680
+    },
+    {
+      "epoch": 0.6126994431821019,
+      "grad_norm": 0.09775250405073166,
+      "learning_rate": 1.1648101329194722e-05,
+      "loss": 8.7415,
+      "step": 122690
+    },
+    {
+      "epoch": 0.6127493820070413,
+      "grad_norm": 0.094340018928051,
+      "learning_rate": 1.1646599414253172e-05,
+      "loss": 8.7466,
+      "step": 122700
+    },
+    {
+      "epoch": 0.6127993208319809,
+      "grad_norm": 0.09047188609838486,
+      "learning_rate": 1.1645097499311623e-05,
+      "loss": 8.7268,
+      "step": 122710
+    },
+    {
+      "epoch": 0.6128492596569203,
+      "grad_norm": 0.0914597436785698,
+      "learning_rate": 1.1643595584370073e-05,
+      "loss": 8.7365,
+      "step": 122720
+    },
+    {
+      "epoch": 0.6128991984818597,
+      "grad_norm": 0.08962114155292511,
+      "learning_rate": 1.1642093669428523e-05,
+      "loss": 8.731,
+      "step": 122730
+    },
+    {
+      "epoch": 0.6129491373067991,
+      "grad_norm": 0.0948721170425415,
+      "learning_rate": 1.164059175448697e-05,
+      "loss": 8.7166,
+      "step": 122740
+    },
+    {
+      "epoch": 0.6129990761317387,
+      "grad_norm": 0.08996075391769409,
+      "learning_rate": 1.163908983954542e-05,
+      "loss": 8.7231,
+      "step": 122750
+    },
+    {
+      "epoch": 0.6130490149566781,
+      "grad_norm": 0.09304836392402649,
+      "learning_rate": 1.163758792460387e-05,
+      "loss": 8.7423,
+      "step": 122760
+    },
+    {
+      "epoch": 0.6130989537816175,
+      "grad_norm": 0.08591028302907944,
+      "learning_rate": 1.163608600966232e-05,
+      "loss": 8.723,
+      "step": 122770
+    },
+    {
+      "epoch": 0.6131488926065569,
+      "grad_norm": 0.08985080569982529,
+      "learning_rate": 1.163458409472077e-05,
+      "loss": 8.7389,
+      "step": 122780
+    },
+    {
+      "epoch": 0.6131988314314964,
+      "grad_norm": 0.09522908180952072,
+      "learning_rate": 1.1633082179779217e-05,
+      "loss": 8.7362,
+      "step": 122790
+    },
+    {
+      "epoch": 0.6132487702564359,
+      "grad_norm": 0.08915939182043076,
+      "learning_rate": 1.1631580264837667e-05,
+      "loss": 8.7425,
+      "step": 122800
+    },
+    {
+      "epoch": 0.6132987090813753,
+      "grad_norm": 0.0966191291809082,
+      "learning_rate": 1.1630078349896118e-05,
+      "loss": 8.7361,
+      "step": 122810
+    },
+    {
+      "epoch": 0.6133486479063147,
+      "grad_norm": 0.08906877040863037,
+      "learning_rate": 1.1628576434954568e-05,
+      "loss": 8.726,
+      "step": 122820
+    },
+    {
+      "epoch": 0.6133985867312542,
+      "grad_norm": 0.0922885611653328,
+      "learning_rate": 1.1627074520013018e-05,
+      "loss": 8.7094,
+      "step": 122830
+    },
+    {
+      "epoch": 0.6134485255561937,
+      "grad_norm": 0.0915333479642868,
+      "learning_rate": 1.1625572605071466e-05,
+      "loss": 8.7258,
+      "step": 122840
+    },
+    {
+      "epoch": 0.6134984643811331,
+      "grad_norm": 0.09396089613437653,
+      "learning_rate": 1.1624070690129915e-05,
+      "loss": 8.7101,
+      "step": 122850
+    },
+    {
+      "epoch": 0.6135484032060725,
+      "grad_norm": 0.09335985779762268,
+      "learning_rate": 1.1622568775188365e-05,
+      "loss": 8.735,
+      "step": 122860
+    },
+    {
+      "epoch": 0.613598342031012,
+      "grad_norm": 0.09260796010494232,
+      "learning_rate": 1.1621066860246815e-05,
+      "loss": 8.7354,
+      "step": 122870
+    },
+    {
+      "epoch": 0.6136482808559515,
+      "grad_norm": 0.09389949589967728,
+      "learning_rate": 1.1619564945305265e-05,
+      "loss": 8.7136,
+      "step": 122880
+    },
+    {
+      "epoch": 0.6136982196808909,
+      "grad_norm": 0.0865430235862732,
+      "learning_rate": 1.1618063030363714e-05,
+      "loss": 8.743,
+      "step": 122890
+    },
+    {
+      "epoch": 0.6137481585058303,
+      "grad_norm": 0.08791029453277588,
+      "learning_rate": 1.1616561115422162e-05,
+      "loss": 8.7337,
+      "step": 122900
+    },
+    {
+      "epoch": 0.6137980973307698,
+      "grad_norm": 0.0919414758682251,
+      "learning_rate": 1.1615059200480613e-05,
+      "loss": 8.7336,
+      "step": 122910
+    },
+    {
+      "epoch": 0.6138480361557093,
+      "grad_norm": 0.08799885213375092,
+      "learning_rate": 1.1613557285539063e-05,
+      "loss": 8.7203,
+      "step": 122920
+    },
+    {
+      "epoch": 0.6138979749806487,
+      "grad_norm": 0.08847959339618683,
+      "learning_rate": 1.1612055370597513e-05,
+      "loss": 8.7178,
+      "step": 122930
+    },
+    {
+      "epoch": 0.6139479138055881,
+      "grad_norm": 0.10149681568145752,
+      "learning_rate": 1.1610553455655961e-05,
+      "loss": 8.7265,
+      "step": 122940
+    },
+    {
+      "epoch": 0.6139978526305276,
+      "grad_norm": 0.09459761530160904,
+      "learning_rate": 1.160905154071441e-05,
+      "loss": 8.7224,
+      "step": 122950
+    },
+    {
+      "epoch": 0.6140477914554671,
+      "grad_norm": 0.09227167069911957,
+      "learning_rate": 1.160754962577286e-05,
+      "loss": 8.7274,
+      "step": 122960
+    },
+    {
+      "epoch": 0.6140977302804065,
+      "grad_norm": 0.08995702862739563,
+      "learning_rate": 1.160604771083131e-05,
+      "loss": 8.718,
+      "step": 122970
+    },
+    {
+      "epoch": 0.6141476691053459,
+      "grad_norm": 0.08922126144170761,
+      "learning_rate": 1.160454579588976e-05,
+      "loss": 8.7368,
+      "step": 122980
+    },
+    {
+      "epoch": 0.6141976079302854,
+      "grad_norm": 0.08837445080280304,
+      "learning_rate": 1.1603043880948209e-05,
+      "loss": 8.7421,
+      "step": 122990
+    },
+    {
+      "epoch": 0.6142475467552249,
+      "grad_norm": 0.0875106155872345,
+      "learning_rate": 1.1601541966006657e-05,
+      "loss": 8.7198,
+      "step": 123000
+    },
+    {
+      "epoch": 0.6142974855801643,
+      "grad_norm": 0.0882338210940361,
+      "learning_rate": 1.1600040051065108e-05,
+      "loss": 8.7162,
+      "step": 123010
+    },
+    {
+      "epoch": 0.6143474244051037,
+      "grad_norm": 0.08920801430940628,
+      "learning_rate": 1.1598538136123558e-05,
+      "loss": 8.717,
+      "step": 123020
+    },
+    {
+      "epoch": 0.6143973632300432,
+      "grad_norm": 0.08952096849679947,
+      "learning_rate": 1.1597036221182008e-05,
+      "loss": 8.7223,
+      "step": 123030
+    },
+    {
+      "epoch": 0.6144473020549827,
+      "grad_norm": 0.09667718410491943,
+      "learning_rate": 1.1595534306240456e-05,
+      "loss": 8.7406,
+      "step": 123040
+    },
+    {
+      "epoch": 0.6144972408799221,
+      "grad_norm": 0.08800283074378967,
+      "learning_rate": 1.1594032391298907e-05,
+      "loss": 8.741,
+      "step": 123050
+    },
+    {
+      "epoch": 0.6145471797048615,
+      "grad_norm": 0.09324777871370316,
+      "learning_rate": 1.1592530476357355e-05,
+      "loss": 8.7304,
+      "step": 123060
+    },
+    {
+      "epoch": 0.614597118529801,
+      "grad_norm": 0.09224963933229446,
+      "learning_rate": 1.1591028561415805e-05,
+      "loss": 8.7463,
+      "step": 123070
+    },
+    {
+      "epoch": 0.6146470573547405,
+      "grad_norm": 0.09018932282924652,
+      "learning_rate": 1.1589526646474255e-05,
+      "loss": 8.7387,
+      "step": 123080
+    },
+    {
+      "epoch": 0.6146969961796799,
+      "grad_norm": 0.09440066665410995,
+      "learning_rate": 1.1588024731532704e-05,
+      "loss": 8.7013,
+      "step": 123090
+    },
+    {
+      "epoch": 0.6147469350046193,
+      "grad_norm": 0.09074559062719345,
+      "learning_rate": 1.1586522816591154e-05,
+      "loss": 8.7298,
+      "step": 123100
+    },
+    {
+      "epoch": 0.6147968738295588,
+      "grad_norm": 0.09388012439012527,
+      "learning_rate": 1.1585020901649603e-05,
+      "loss": 8.7206,
+      "step": 123110
+    },
+    {
+      "epoch": 0.6148468126544983,
+      "grad_norm": 0.09513651579618454,
+      "learning_rate": 1.1583518986708053e-05,
+      "loss": 8.7438,
+      "step": 123120
+    },
+    {
+      "epoch": 0.6148967514794377,
+      "grad_norm": 0.08741681277751923,
+      "learning_rate": 1.1582017071766503e-05,
+      "loss": 8.7586,
+      "step": 123130
+    },
+    {
+      "epoch": 0.6149466903043771,
+      "grad_norm": 0.0910298153758049,
+      "learning_rate": 1.1580515156824951e-05,
+      "loss": 8.7249,
+      "step": 123140
+    },
+    {
+      "epoch": 0.6149966291293166,
+      "grad_norm": 0.0909765362739563,
+      "learning_rate": 1.1579013241883402e-05,
+      "loss": 8.7324,
+      "step": 123150
+    },
+    {
+      "epoch": 0.615046567954256,
+      "grad_norm": 0.09115078300237656,
+      "learning_rate": 1.157751132694185e-05,
+      "loss": 8.7132,
+      "step": 123160
+    },
+    {
+      "epoch": 0.6150965067791955,
+      "grad_norm": 0.09587119519710541,
+      "learning_rate": 1.15760094120003e-05,
+      "loss": 8.7349,
+      "step": 123170
+    },
+    {
+      "epoch": 0.6151464456041349,
+      "grad_norm": 0.09301666915416718,
+      "learning_rate": 1.157450749705875e-05,
+      "loss": 8.7374,
+      "step": 123180
+    },
+    {
+      "epoch": 0.6151963844290744,
+      "grad_norm": 0.0892057865858078,
+      "learning_rate": 1.15730055821172e-05,
+      "loss": 8.7265,
+      "step": 123190
+    },
+    {
+      "epoch": 0.6152463232540138,
+      "grad_norm": 0.09476073831319809,
+      "learning_rate": 1.157150366717565e-05,
+      "loss": 8.7264,
+      "step": 123200
+    },
+    {
+      "epoch": 0.6152962620789533,
+      "grad_norm": 0.08770003169775009,
+      "learning_rate": 1.15700017522341e-05,
+      "loss": 8.734,
+      "step": 123210
+    },
+    {
+      "epoch": 0.6153462009038927,
+      "grad_norm": 0.09061672538518906,
+      "learning_rate": 1.1568499837292548e-05,
+      "loss": 8.737,
+      "step": 123220
+    },
+    {
+      "epoch": 0.6153961397288322,
+      "grad_norm": 0.09397213160991669,
+      "learning_rate": 1.1566997922350998e-05,
+      "loss": 8.7247,
+      "step": 123230
+    },
+    {
+      "epoch": 0.6154460785537716,
+      "grad_norm": 0.09056683629751205,
+      "learning_rate": 1.1565496007409448e-05,
+      "loss": 8.7302,
+      "step": 123240
+    },
+    {
+      "epoch": 0.6154960173787111,
+      "grad_norm": 0.09550472348928452,
+      "learning_rate": 1.1563994092467897e-05,
+      "loss": 8.7345,
+      "step": 123250
+    },
+    {
+      "epoch": 0.6155459562036505,
+      "grad_norm": 0.09115252643823624,
+      "learning_rate": 1.1562492177526347e-05,
+      "loss": 8.7392,
+      "step": 123260
+    },
+    {
+      "epoch": 0.61559589502859,
+      "grad_norm": 0.09301244467496872,
+      "learning_rate": 1.1560990262584795e-05,
+      "loss": 8.7234,
+      "step": 123270
+    },
+    {
+      "epoch": 0.6156458338535294,
+      "grad_norm": 0.0930706262588501,
+      "learning_rate": 1.1559488347643246e-05,
+      "loss": 8.7305,
+      "step": 123280
+    },
+    {
+      "epoch": 0.6156957726784689,
+      "grad_norm": 0.09021922200918198,
+      "learning_rate": 1.1557986432701696e-05,
+      "loss": 8.7383,
+      "step": 123290
+    },
+    {
+      "epoch": 0.6157457115034083,
+      "grad_norm": 0.09439912438392639,
+      "learning_rate": 1.1556484517760144e-05,
+      "loss": 8.7176,
+      "step": 123300
+    },
+    {
+      "epoch": 0.6157956503283478,
+      "grad_norm": 0.09691648930311203,
+      "learning_rate": 1.1554982602818594e-05,
+      "loss": 8.7301,
+      "step": 123310
+    },
+    {
+      "epoch": 0.6158455891532872,
+      "grad_norm": 0.09814375638961792,
+      "learning_rate": 1.1553480687877043e-05,
+      "loss": 8.7233,
+      "step": 123320
+    },
+    {
+      "epoch": 0.6158955279782267,
+      "grad_norm": 0.09281103312969208,
+      "learning_rate": 1.1551978772935493e-05,
+      "loss": 8.741,
+      "step": 123330
+    },
+    {
+      "epoch": 0.6159454668031661,
+      "grad_norm": 0.09556780755519867,
+      "learning_rate": 1.1550476857993943e-05,
+      "loss": 8.7346,
+      "step": 123340
+    },
+    {
+      "epoch": 0.6159954056281056,
+      "grad_norm": 0.09661609679460526,
+      "learning_rate": 1.1548974943052392e-05,
+      "loss": 8.7364,
+      "step": 123350
+    },
+    {
+      "epoch": 0.616045344453045,
+      "grad_norm": 0.08993745595216751,
+      "learning_rate": 1.1547473028110842e-05,
+      "loss": 8.7326,
+      "step": 123360
+    },
+    {
+      "epoch": 0.6160952832779845,
+      "grad_norm": 0.08858168870210648,
+      "learning_rate": 1.1545971113169292e-05,
+      "loss": 8.7335,
+      "step": 123370
+    },
+    {
+      "epoch": 0.6161452221029239,
+      "grad_norm": 0.0938563272356987,
+      "learning_rate": 1.154446919822774e-05,
+      "loss": 8.731,
+      "step": 123380
+    },
+    {
+      "epoch": 0.6161951609278634,
+      "grad_norm": 0.09345029294490814,
+      "learning_rate": 1.154296728328619e-05,
+      "loss": 8.7307,
+      "step": 123390
+    },
+    {
+      "epoch": 0.6162450997528028,
+      "grad_norm": 0.09097786992788315,
+      "learning_rate": 1.154146536834464e-05,
+      "loss": 8.7149,
+      "step": 123400
+    },
+    {
+      "epoch": 0.6162950385777423,
+      "grad_norm": 0.09178051352500916,
+      "learning_rate": 1.153996345340309e-05,
+      "loss": 8.7436,
+      "step": 123410
+    },
+    {
+      "epoch": 0.6163449774026817,
+      "grad_norm": 0.08830767124891281,
+      "learning_rate": 1.153846153846154e-05,
+      "loss": 8.7355,
+      "step": 123420
+    },
+    {
+      "epoch": 0.6163949162276212,
+      "grad_norm": 0.09343507140874863,
+      "learning_rate": 1.1536959623519988e-05,
+      "loss": 8.7295,
+      "step": 123430
+    },
+    {
+      "epoch": 0.6164448550525606,
+      "grad_norm": 0.09138908237218857,
+      "learning_rate": 1.1535457708578438e-05,
+      "loss": 8.7319,
+      "step": 123440
+    },
+    {
+      "epoch": 0.6164947938775001,
+      "grad_norm": 0.09196456521749496,
+      "learning_rate": 1.1533955793636887e-05,
+      "loss": 8.7181,
+      "step": 123450
+    },
+    {
+      "epoch": 0.6165447327024395,
+      "grad_norm": 0.09027263522148132,
+      "learning_rate": 1.1532453878695337e-05,
+      "loss": 8.7359,
+      "step": 123460
+    },
+    {
+      "epoch": 0.616594671527379,
+      "grad_norm": 0.09021011739969254,
+      "learning_rate": 1.1530951963753787e-05,
+      "loss": 8.7455,
+      "step": 123470
+    },
+    {
+      "epoch": 0.6166446103523184,
+      "grad_norm": 0.08662664890289307,
+      "learning_rate": 1.1529450048812236e-05,
+      "loss": 8.7415,
+      "step": 123480
+    },
+    {
+      "epoch": 0.6166945491772579,
+      "grad_norm": 0.09319479018449783,
+      "learning_rate": 1.1527948133870686e-05,
+      "loss": 8.7173,
+      "step": 123490
+    },
+    {
+      "epoch": 0.6167444880021973,
+      "grad_norm": 0.08931265026330948,
+      "learning_rate": 1.1526446218929134e-05,
+      "loss": 8.7091,
+      "step": 123500
+    },
+    {
+      "epoch": 0.6167944268271368,
+      "grad_norm": 0.09478646516799927,
+      "learning_rate": 1.1524944303987584e-05,
+      "loss": 8.7149,
+      "step": 123510
+    },
+    {
+      "epoch": 0.6168443656520762,
+      "grad_norm": 0.08900546282529831,
+      "learning_rate": 1.1523442389046035e-05,
+      "loss": 8.7382,
+      "step": 123520
+    },
+    {
+      "epoch": 0.6168943044770157,
+      "grad_norm": 0.09752871096134186,
+      "learning_rate": 1.1521940474104485e-05,
+      "loss": 8.7297,
+      "step": 123530
+    },
+    {
+      "epoch": 0.6169442433019551,
+      "grad_norm": 0.09565001726150513,
+      "learning_rate": 1.1520438559162933e-05,
+      "loss": 8.7219,
+      "step": 123540
+    },
+    {
+      "epoch": 0.6169941821268946,
+      "grad_norm": 0.09056803584098816,
+      "learning_rate": 1.1518936644221382e-05,
+      "loss": 8.7327,
+      "step": 123550
+    },
+    {
+      "epoch": 0.617044120951834,
+      "grad_norm": 0.09336548298597336,
+      "learning_rate": 1.1517434729279832e-05,
+      "loss": 8.7223,
+      "step": 123560
+    },
+    {
+      "epoch": 0.6170940597767735,
+      "grad_norm": 0.09907118231058121,
+      "learning_rate": 1.1515932814338282e-05,
+      "loss": 8.7273,
+      "step": 123570
+    },
+    {
+      "epoch": 0.6171439986017129,
+      "grad_norm": 0.08720168471336365,
+      "learning_rate": 1.1514430899396732e-05,
+      "loss": 8.7245,
+      "step": 123580
+    },
+    {
+      "epoch": 0.6171939374266524,
+      "grad_norm": 0.09530545026063919,
+      "learning_rate": 1.151292898445518e-05,
+      "loss": 8.7177,
+      "step": 123590
+    },
+    {
+      "epoch": 0.6172438762515918,
+      "grad_norm": 0.08918207138776779,
+      "learning_rate": 1.151142706951363e-05,
+      "loss": 8.7255,
+      "step": 123600
+    },
+    {
+      "epoch": 0.6172938150765312,
+      "grad_norm": 0.09224686026573181,
+      "learning_rate": 1.150992515457208e-05,
+      "loss": 8.7282,
+      "step": 123610
+    },
+    {
+      "epoch": 0.6173437539014707,
+      "grad_norm": 0.08980294317007065,
+      "learning_rate": 1.150842323963053e-05,
+      "loss": 8.7338,
+      "step": 123620
+    },
+    {
+      "epoch": 0.6173936927264101,
+      "grad_norm": 0.08682075142860413,
+      "learning_rate": 1.150692132468898e-05,
+      "loss": 8.7079,
+      "step": 123630
+    },
+    {
+      "epoch": 0.6174436315513496,
+      "grad_norm": 0.0889846533536911,
+      "learning_rate": 1.1505419409747428e-05,
+      "loss": 8.7336,
+      "step": 123640
+    },
+    {
+      "epoch": 0.617493570376289,
+      "grad_norm": 0.09437645971775055,
+      "learning_rate": 1.1503917494805877e-05,
+      "loss": 8.73,
+      "step": 123650
+    },
+    {
+      "epoch": 0.6175435092012285,
+      "grad_norm": 0.08967822045087814,
+      "learning_rate": 1.1502415579864327e-05,
+      "loss": 8.7262,
+      "step": 123660
+    },
+    {
+      "epoch": 0.6175934480261679,
+      "grad_norm": 0.09480708837509155,
+      "learning_rate": 1.1500913664922777e-05,
+      "loss": 8.7245,
+      "step": 123670
+    },
+    {
+      "epoch": 0.6176433868511074,
+      "grad_norm": 0.095451220870018,
+      "learning_rate": 1.1499411749981227e-05,
+      "loss": 8.7531,
+      "step": 123680
+    },
+    {
+      "epoch": 0.6176933256760468,
+      "grad_norm": 0.08989094942808151,
+      "learning_rate": 1.1497909835039677e-05,
+      "loss": 8.7383,
+      "step": 123690
+    },
+    {
+      "epoch": 0.6177432645009863,
+      "grad_norm": 0.08974773436784744,
+      "learning_rate": 1.1496407920098124e-05,
+      "loss": 8.7298,
+      "step": 123700
+    },
+    {
+      "epoch": 0.6177932033259257,
+      "grad_norm": 0.08904717862606049,
+      "learning_rate": 1.1494906005156574e-05,
+      "loss": 8.7198,
+      "step": 123710
+    },
+    {
+      "epoch": 0.6178431421508652,
+      "grad_norm": 0.0950782522559166,
+      "learning_rate": 1.1493404090215025e-05,
+      "loss": 8.7192,
+      "step": 123720
+    },
+    {
+      "epoch": 0.6178930809758046,
+      "grad_norm": 0.08988340198993683,
+      "learning_rate": 1.1491902175273475e-05,
+      "loss": 8.7233,
+      "step": 123730
+    },
+    {
+      "epoch": 0.6179430198007441,
+      "grad_norm": 0.08889536559581757,
+      "learning_rate": 1.1490400260331925e-05,
+      "loss": 8.7234,
+      "step": 123740
+    },
+    {
+      "epoch": 0.6179929586256835,
+      "grad_norm": 0.08637700229883194,
+      "learning_rate": 1.1488898345390372e-05,
+      "loss": 8.7312,
+      "step": 123750
+    },
+    {
+      "epoch": 0.618042897450623,
+      "grad_norm": 0.09260913729667664,
+      "learning_rate": 1.1487396430448822e-05,
+      "loss": 8.7229,
+      "step": 123760
+    },
+    {
+      "epoch": 0.6180928362755624,
+      "grad_norm": 0.10173122584819794,
+      "learning_rate": 1.1485894515507272e-05,
+      "loss": 8.7177,
+      "step": 123770
+    },
+    {
+      "epoch": 0.6181427751005019,
+      "grad_norm": 0.09701031446456909,
+      "learning_rate": 1.1484392600565722e-05,
+      "loss": 8.7225,
+      "step": 123780
+    },
+    {
+      "epoch": 0.6181927139254413,
+      "grad_norm": 0.09118404984474182,
+      "learning_rate": 1.1482890685624172e-05,
+      "loss": 8.7159,
+      "step": 123790
+    },
+    {
+      "epoch": 0.6182426527503808,
+      "grad_norm": 0.09571003168821335,
+      "learning_rate": 1.148138877068262e-05,
+      "loss": 8.7244,
+      "step": 123800
+    },
+    {
+      "epoch": 0.6182925915753202,
+      "grad_norm": 0.09160356968641281,
+      "learning_rate": 1.147988685574107e-05,
+      "loss": 8.7271,
+      "step": 123810
+    },
+    {
+      "epoch": 0.6183425304002597,
+      "grad_norm": 0.10198367387056351,
+      "learning_rate": 1.147838494079952e-05,
+      "loss": 8.728,
+      "step": 123820
+    },
+    {
+      "epoch": 0.6183924692251991,
+      "grad_norm": 0.090346559882164,
+      "learning_rate": 1.147688302585797e-05,
+      "loss": 8.7252,
+      "step": 123830
+    },
+    {
+      "epoch": 0.6184424080501386,
+      "grad_norm": 0.08791203796863556,
+      "learning_rate": 1.147538111091642e-05,
+      "loss": 8.7254,
+      "step": 123840
+    },
+    {
+      "epoch": 0.618492346875078,
+      "grad_norm": 0.09534545242786407,
+      "learning_rate": 1.1473879195974868e-05,
+      "loss": 8.7175,
+      "step": 123850
+    },
+    {
+      "epoch": 0.6185422857000175,
+      "grad_norm": 0.09206069260835648,
+      "learning_rate": 1.1472377281033317e-05,
+      "loss": 8.732,
+      "step": 123860
+    },
+    {
+      "epoch": 0.6185922245249569,
+      "grad_norm": 0.09079962968826294,
+      "learning_rate": 1.1470875366091767e-05,
+      "loss": 8.7245,
+      "step": 123870
+    },
+    {
+      "epoch": 0.6186421633498964,
+      "grad_norm": 0.09148161858320236,
+      "learning_rate": 1.1469373451150217e-05,
+      "loss": 8.7172,
+      "step": 123880
+    },
+    {
+      "epoch": 0.6186921021748358,
+      "grad_norm": 0.0891844853758812,
+      "learning_rate": 1.1467871536208667e-05,
+      "loss": 8.721,
+      "step": 123890
+    },
+    {
+      "epoch": 0.6187420409997753,
+      "grad_norm": 0.08975985646247864,
+      "learning_rate": 1.1466369621267116e-05,
+      "loss": 8.7102,
+      "step": 123900
+    },
+    {
+      "epoch": 0.6187919798247147,
+      "grad_norm": 0.10155780613422394,
+      "learning_rate": 1.1464867706325564e-05,
+      "loss": 8.7252,
+      "step": 123910
+    },
+    {
+      "epoch": 0.6188419186496542,
+      "grad_norm": 0.09575510025024414,
+      "learning_rate": 1.1463365791384015e-05,
+      "loss": 8.716,
+      "step": 123920
+    },
+    {
+      "epoch": 0.6188918574745936,
+      "grad_norm": 0.0914383977651596,
+      "learning_rate": 1.1461863876442465e-05,
+      "loss": 8.7268,
+      "step": 123930
+    },
+    {
+      "epoch": 0.618941796299533,
+      "grad_norm": 0.09883058816194534,
+      "learning_rate": 1.1460361961500915e-05,
+      "loss": 8.7068,
+      "step": 123940
+    },
+    {
+      "epoch": 0.6189917351244725,
+      "grad_norm": 0.09328742325305939,
+      "learning_rate": 1.1458860046559363e-05,
+      "loss": 8.7305,
+      "step": 123950
+    },
+    {
+      "epoch": 0.619041673949412,
+      "grad_norm": 0.09385626018047333,
+      "learning_rate": 1.1457358131617812e-05,
+      "loss": 8.7238,
+      "step": 123960
+    },
+    {
+      "epoch": 0.6190916127743514,
+      "grad_norm": 0.0958520770072937,
+      "learning_rate": 1.1455856216676262e-05,
+      "loss": 8.7139,
+      "step": 123970
+    },
+    {
+      "epoch": 0.6191415515992909,
+      "grad_norm": 0.09502740204334259,
+      "learning_rate": 1.1454354301734712e-05,
+      "loss": 8.7134,
+      "step": 123980
+    },
+    {
+      "epoch": 0.6191914904242303,
+      "grad_norm": 0.09292402863502502,
+      "learning_rate": 1.1452852386793162e-05,
+      "loss": 8.7092,
+      "step": 123990
+    },
+    {
+      "epoch": 0.6192414292491698,
+      "grad_norm": 0.09679784625768661,
+      "learning_rate": 1.1451350471851611e-05,
+      "loss": 8.7259,
+      "step": 124000
+    },
+    {
+      "epoch": 0.6192913680741092,
+      "grad_norm": 0.09584274888038635,
+      "learning_rate": 1.1449848556910061e-05,
+      "loss": 8.715,
+      "step": 124010
+    },
+    {
+      "epoch": 0.6193413068990486,
+      "grad_norm": 0.08926044404506683,
+      "learning_rate": 1.144834664196851e-05,
+      "loss": 8.7075,
+      "step": 124020
+    },
+    {
+      "epoch": 0.6193912457239881,
+      "grad_norm": 0.09470371901988983,
+      "learning_rate": 1.144684472702696e-05,
+      "loss": 8.7184,
+      "step": 124030
+    },
+    {
+      "epoch": 0.6194411845489276,
+      "grad_norm": 0.09544707089662552,
+      "learning_rate": 1.144534281208541e-05,
+      "loss": 8.7218,
+      "step": 124040
+    },
+    {
+      "epoch": 0.619491123373867,
+      "grad_norm": 0.09077138453722,
+      "learning_rate": 1.1443840897143858e-05,
+      "loss": 8.7214,
+      "step": 124050
+    },
+    {
+      "epoch": 0.6195410621988064,
+      "grad_norm": 0.08742060512304306,
+      "learning_rate": 1.1442338982202309e-05,
+      "loss": 8.7439,
+      "step": 124060
+    },
+    {
+      "epoch": 0.6195910010237459,
+      "grad_norm": 0.09070980548858643,
+      "learning_rate": 1.1440837067260757e-05,
+      "loss": 8.7162,
+      "step": 124070
+    },
+    {
+      "epoch": 0.6196409398486854,
+      "grad_norm": 0.09399957954883575,
+      "learning_rate": 1.1439335152319207e-05,
+      "loss": 8.7223,
+      "step": 124080
+    },
+    {
+      "epoch": 0.6196908786736248,
+      "grad_norm": 0.08889596909284592,
+      "learning_rate": 1.1437833237377657e-05,
+      "loss": 8.7342,
+      "step": 124090
+    },
+    {
+      "epoch": 0.6197408174985642,
+      "grad_norm": 0.09190160781145096,
+      "learning_rate": 1.1436331322436106e-05,
+      "loss": 8.7455,
+      "step": 124100
+    },
+    {
+      "epoch": 0.6197907563235037,
+      "grad_norm": 0.09013140201568604,
+      "learning_rate": 1.1434829407494556e-05,
+      "loss": 8.7439,
+      "step": 124110
+    },
+    {
+      "epoch": 0.6198406951484432,
+      "grad_norm": 0.0921657457947731,
+      "learning_rate": 1.1433327492553005e-05,
+      "loss": 8.7196,
+      "step": 124120
+    },
+    {
+      "epoch": 0.6198906339733826,
+      "grad_norm": 0.09478151053190231,
+      "learning_rate": 1.1431825577611455e-05,
+      "loss": 8.7308,
+      "step": 124130
+    },
+    {
+      "epoch": 0.619940572798322,
+      "grad_norm": 0.09333883225917816,
+      "learning_rate": 1.1430323662669905e-05,
+      "loss": 8.7126,
+      "step": 124140
+    },
+    {
+      "epoch": 0.6199905116232615,
+      "grad_norm": 0.0867578536272049,
+      "learning_rate": 1.1428821747728353e-05,
+      "loss": 8.7471,
+      "step": 124150
+    },
+    {
+      "epoch": 0.620040450448201,
+      "grad_norm": 0.08902301639318466,
+      "learning_rate": 1.1427319832786804e-05,
+      "loss": 8.7455,
+      "step": 124160
+    },
+    {
+      "epoch": 0.6200903892731404,
+      "grad_norm": 0.09362560510635376,
+      "learning_rate": 1.1425817917845254e-05,
+      "loss": 8.7118,
+      "step": 124170
+    },
+    {
+      "epoch": 0.6201403280980798,
+      "grad_norm": 0.09488838911056519,
+      "learning_rate": 1.1424316002903702e-05,
+      "loss": 8.7172,
+      "step": 124180
+    },
+    {
+      "epoch": 0.6201902669230193,
+      "grad_norm": 0.09388437122106552,
+      "learning_rate": 1.1422814087962152e-05,
+      "loss": 8.7249,
+      "step": 124190
+    },
+    {
+      "epoch": 0.6202402057479588,
+      "grad_norm": 0.09056483954191208,
+      "learning_rate": 1.1421312173020601e-05,
+      "loss": 8.7256,
+      "step": 124200
+    },
+    {
+      "epoch": 0.6202901445728982,
+      "grad_norm": 0.09356372058391571,
+      "learning_rate": 1.1419810258079051e-05,
+      "loss": 8.7165,
+      "step": 124210
+    },
+    {
+      "epoch": 0.6203400833978376,
+      "grad_norm": 0.09414329379796982,
+      "learning_rate": 1.1418308343137501e-05,
+      "loss": 8.7177,
+      "step": 124220
+    },
+    {
+      "epoch": 0.6203900222227771,
+      "grad_norm": 0.08978509902954102,
+      "learning_rate": 1.141680642819595e-05,
+      "loss": 8.7225,
+      "step": 124230
+    },
+    {
+      "epoch": 0.6204399610477166,
+      "grad_norm": 0.09141732007265091,
+      "learning_rate": 1.14153045132544e-05,
+      "loss": 8.7325,
+      "step": 124240
+    },
+    {
+      "epoch": 0.620489899872656,
+      "grad_norm": 0.09441263973712921,
+      "learning_rate": 1.1413802598312848e-05,
+      "loss": 8.719,
+      "step": 124250
+    },
+    {
+      "epoch": 0.6205398386975954,
+      "grad_norm": 0.09462641179561615,
+      "learning_rate": 1.1412300683371299e-05,
+      "loss": 8.716,
+      "step": 124260
+    },
+    {
+      "epoch": 0.6205897775225349,
+      "grad_norm": 0.09261373430490494,
+      "learning_rate": 1.1410798768429749e-05,
+      "loss": 8.722,
+      "step": 124270
+    },
+    {
+      "epoch": 0.6206397163474744,
+      "grad_norm": 0.09427915513515472,
+      "learning_rate": 1.1409296853488197e-05,
+      "loss": 8.7251,
+      "step": 124280
+    },
+    {
+      "epoch": 0.6206896551724138,
+      "grad_norm": 0.09489326179027557,
+      "learning_rate": 1.1407794938546647e-05,
+      "loss": 8.7197,
+      "step": 124290
+    },
+    {
+      "epoch": 0.6207395939973532,
+      "grad_norm": 0.09282520413398743,
+      "learning_rate": 1.1406293023605096e-05,
+      "loss": 8.7217,
+      "step": 124300
+    },
+    {
+      "epoch": 0.6207895328222927,
+      "grad_norm": 0.09794396162033081,
+      "learning_rate": 1.1404791108663546e-05,
+      "loss": 8.7228,
+      "step": 124310
+    },
+    {
+      "epoch": 0.6208394716472322,
+      "grad_norm": 0.09146995097398758,
+      "learning_rate": 1.1403289193721996e-05,
+      "loss": 8.7181,
+      "step": 124320
+    },
+    {
+      "epoch": 0.6208894104721716,
+      "grad_norm": 0.0925363302230835,
+      "learning_rate": 1.1401787278780446e-05,
+      "loss": 8.7161,
+      "step": 124330
+    },
+    {
+      "epoch": 0.620939349297111,
+      "grad_norm": 0.09043951332569122,
+      "learning_rate": 1.1400285363838895e-05,
+      "loss": 8.7191,
+      "step": 124340
+    },
+    {
+      "epoch": 0.6209892881220505,
+      "grad_norm": 0.09174945205450058,
+      "learning_rate": 1.1398783448897343e-05,
+      "loss": 8.7329,
+      "step": 124350
+    },
+    {
+      "epoch": 0.62103922694699,
+      "grad_norm": 0.09292636811733246,
+      "learning_rate": 1.1397281533955794e-05,
+      "loss": 8.7225,
+      "step": 124360
+    },
+    {
+      "epoch": 0.6210891657719294,
+      "grad_norm": 0.09614364802837372,
+      "learning_rate": 1.1395779619014244e-05,
+      "loss": 8.7178,
+      "step": 124370
+    },
+    {
+      "epoch": 0.6211391045968688,
+      "grad_norm": 0.09423890709877014,
+      "learning_rate": 1.1394277704072694e-05,
+      "loss": 8.7239,
+      "step": 124380
+    },
+    {
+      "epoch": 0.6211890434218083,
+      "grad_norm": 0.08773567527532578,
+      "learning_rate": 1.1392775789131142e-05,
+      "loss": 8.6973,
+      "step": 124390
+    },
+    {
+      "epoch": 0.6212389822467478,
+      "grad_norm": 0.09796379506587982,
+      "learning_rate": 1.1391273874189591e-05,
+      "loss": 8.7243,
+      "step": 124400
+    },
+    {
+      "epoch": 0.6212889210716872,
+      "grad_norm": 0.08884675800800323,
+      "learning_rate": 1.1389771959248041e-05,
+      "loss": 8.7269,
+      "step": 124410
+    },
+    {
+      "epoch": 0.6213388598966266,
+      "grad_norm": 0.09190382063388824,
+      "learning_rate": 1.1388270044306491e-05,
+      "loss": 8.7166,
+      "step": 124420
+    },
+    {
+      "epoch": 0.621388798721566,
+      "grad_norm": 0.09155641496181488,
+      "learning_rate": 1.1386768129364941e-05,
+      "loss": 8.712,
+      "step": 124430
+    },
+    {
+      "epoch": 0.6214387375465056,
+      "grad_norm": 0.09598276764154434,
+      "learning_rate": 1.138526621442339e-05,
+      "loss": 8.7123,
+      "step": 124440
+    },
+    {
+      "epoch": 0.621488676371445,
+      "grad_norm": 0.0909108817577362,
+      "learning_rate": 1.1383764299481838e-05,
+      "loss": 8.7224,
+      "step": 124450
+    },
+    {
+      "epoch": 0.6215386151963844,
+      "grad_norm": 0.0917460024356842,
+      "learning_rate": 1.1382262384540289e-05,
+      "loss": 8.72,
+      "step": 124460
+    },
+    {
+      "epoch": 0.6215885540213238,
+      "grad_norm": 0.08814991265535355,
+      "learning_rate": 1.1380760469598739e-05,
+      "loss": 8.7155,
+      "step": 124470
+    },
+    {
+      "epoch": 0.6216384928462634,
+      "grad_norm": 0.09747850894927979,
+      "learning_rate": 1.1379258554657189e-05,
+      "loss": 8.7379,
+      "step": 124480
+    },
+    {
+      "epoch": 0.6216884316712028,
+      "grad_norm": 0.09266173839569092,
+      "learning_rate": 1.137775663971564e-05,
+      "loss": 8.7156,
+      "step": 124490
+    },
+    {
+      "epoch": 0.6217383704961422,
+      "grad_norm": 0.09457944333553314,
+      "learning_rate": 1.1376254724774086e-05,
+      "loss": 8.7216,
+      "step": 124500
+    },
+    {
+      "epoch": 0.6217883093210816,
+      "grad_norm": 0.08899705111980438,
+      "learning_rate": 1.1374752809832536e-05,
+      "loss": 8.7244,
+      "step": 124510
+    },
+    {
+      "epoch": 0.6218382481460212,
+      "grad_norm": 0.0950658842921257,
+      "learning_rate": 1.1373250894890986e-05,
+      "loss": 8.7242,
+      "step": 124520
+    },
+    {
+      "epoch": 0.6218881869709606,
+      "grad_norm": 0.0944470465183258,
+      "learning_rate": 1.1371748979949436e-05,
+      "loss": 8.724,
+      "step": 124530
+    },
+    {
+      "epoch": 0.6219381257959,
+      "grad_norm": 0.08662784099578857,
+      "learning_rate": 1.1370247065007887e-05,
+      "loss": 8.7335,
+      "step": 124540
+    },
+    {
+      "epoch": 0.6219880646208394,
+      "grad_norm": 0.09047723561525345,
+      "learning_rate": 1.1368745150066333e-05,
+      "loss": 8.7357,
+      "step": 124550
+    },
+    {
+      "epoch": 0.622038003445779,
+      "grad_norm": 0.09099981188774109,
+      "learning_rate": 1.1367243235124784e-05,
+      "loss": 8.7242,
+      "step": 124560
+    },
+    {
+      "epoch": 0.6220879422707184,
+      "grad_norm": 0.0873718112707138,
+      "learning_rate": 1.1365741320183234e-05,
+      "loss": 8.7197,
+      "step": 124570
+    },
+    {
+      "epoch": 0.6221378810956578,
+      "grad_norm": 0.0861561968922615,
+      "learning_rate": 1.1364239405241684e-05,
+      "loss": 8.7394,
+      "step": 124580
+    },
+    {
+      "epoch": 0.6221878199205972,
+      "grad_norm": 0.10405902564525604,
+      "learning_rate": 1.1362737490300134e-05,
+      "loss": 8.7154,
+      "step": 124590
+    },
+    {
+      "epoch": 0.6222377587455367,
+      "grad_norm": 0.09721992909908295,
+      "learning_rate": 1.1361235575358581e-05,
+      "loss": 8.7003,
+      "step": 124600
+    },
+    {
+      "epoch": 0.6222876975704762,
+      "grad_norm": 0.09676691889762878,
+      "learning_rate": 1.1359733660417031e-05,
+      "loss": 8.7209,
+      "step": 124610
+    },
+    {
+      "epoch": 0.6223376363954156,
+      "grad_norm": 0.09333143383264542,
+      "learning_rate": 1.1358231745475481e-05,
+      "loss": 8.7237,
+      "step": 124620
+    },
+    {
+      "epoch": 0.622387575220355,
+      "grad_norm": 0.08884768187999725,
+      "learning_rate": 1.1356729830533931e-05,
+      "loss": 8.7182,
+      "step": 124630
+    },
+    {
+      "epoch": 0.6224375140452945,
+      "grad_norm": 0.09150359779596329,
+      "learning_rate": 1.1355227915592382e-05,
+      "loss": 8.7136,
+      "step": 124640
+    },
+    {
+      "epoch": 0.622487452870234,
+      "grad_norm": 0.09673064947128296,
+      "learning_rate": 1.135372600065083e-05,
+      "loss": 8.7233,
+      "step": 124650
+    },
+    {
+      "epoch": 0.6225373916951734,
+      "grad_norm": 0.09233259409666061,
+      "learning_rate": 1.1352224085709279e-05,
+      "loss": 8.7057,
+      "step": 124660
+    },
+    {
+      "epoch": 0.6225873305201128,
+      "grad_norm": 0.09211067855358124,
+      "learning_rate": 1.1350722170767729e-05,
+      "loss": 8.7187,
+      "step": 124670
+    },
+    {
+      "epoch": 0.6226372693450523,
+      "grad_norm": 0.09305453300476074,
+      "learning_rate": 1.1349220255826179e-05,
+      "loss": 8.7037,
+      "step": 124680
+    },
+    {
+      "epoch": 0.6226872081699918,
+      "grad_norm": 0.08884977549314499,
+      "learning_rate": 1.134771834088463e-05,
+      "loss": 8.7114,
+      "step": 124690
+    },
+    {
+      "epoch": 0.6227371469949312,
+      "grad_norm": 0.09412440657615662,
+      "learning_rate": 1.1346216425943078e-05,
+      "loss": 8.7272,
+      "step": 124700
+    },
+    {
+      "epoch": 0.6227870858198706,
+      "grad_norm": 0.09514528512954712,
+      "learning_rate": 1.1344714511001526e-05,
+      "loss": 8.7156,
+      "step": 124710
+    },
+    {
+      "epoch": 0.6228370246448101,
+      "grad_norm": 0.0961642786860466,
+      "learning_rate": 1.1343212596059976e-05,
+      "loss": 8.7162,
+      "step": 124720
+    },
+    {
+      "epoch": 0.6228869634697496,
+      "grad_norm": 0.09910131245851517,
+      "learning_rate": 1.1341710681118427e-05,
+      "loss": 8.7031,
+      "step": 124730
+    },
+    {
+      "epoch": 0.622936902294689,
+      "grad_norm": 0.09175023436546326,
+      "learning_rate": 1.1340208766176877e-05,
+      "loss": 8.7131,
+      "step": 124740
+    },
+    {
+      "epoch": 0.6229868411196284,
+      "grad_norm": 0.09244109690189362,
+      "learning_rate": 1.1338706851235325e-05,
+      "loss": 8.7122,
+      "step": 124750
+    },
+    {
+      "epoch": 0.6230367799445679,
+      "grad_norm": 0.09264490753412247,
+      "learning_rate": 1.1337204936293774e-05,
+      "loss": 8.7155,
+      "step": 124760
+    },
+    {
+      "epoch": 0.6230867187695074,
+      "grad_norm": 0.09691507369279861,
+      "learning_rate": 1.1335703021352224e-05,
+      "loss": 8.7146,
+      "step": 124770
+    },
+    {
+      "epoch": 0.6231366575944468,
+      "grad_norm": 0.08869668841362,
+      "learning_rate": 1.1334201106410674e-05,
+      "loss": 8.7173,
+      "step": 124780
+    },
+    {
+      "epoch": 0.6231865964193862,
+      "grad_norm": 0.09124024212360382,
+      "learning_rate": 1.1332699191469124e-05,
+      "loss": 8.7296,
+      "step": 124790
+    },
+    {
+      "epoch": 0.6232365352443257,
+      "grad_norm": 0.09584067761898041,
+      "learning_rate": 1.1331197276527573e-05,
+      "loss": 8.7047,
+      "step": 124800
+    },
+    {
+      "epoch": 0.6232864740692652,
+      "grad_norm": 0.08927422761917114,
+      "learning_rate": 1.1329695361586023e-05,
+      "loss": 8.7281,
+      "step": 124810
+    },
+    {
+      "epoch": 0.6233364128942046,
+      "grad_norm": 0.09715837985277176,
+      "learning_rate": 1.1328193446644471e-05,
+      "loss": 8.7242,
+      "step": 124820
+    },
+    {
+      "epoch": 0.623386351719144,
+      "grad_norm": 0.09197895228862762,
+      "learning_rate": 1.1326691531702922e-05,
+      "loss": 8.7181,
+      "step": 124830
+    },
+    {
+      "epoch": 0.6234362905440834,
+      "grad_norm": 0.0883011445403099,
+      "learning_rate": 1.1325189616761372e-05,
+      "loss": 8.7173,
+      "step": 124840
+    },
+    {
+      "epoch": 0.623486229369023,
+      "grad_norm": 0.08731172233819962,
+      "learning_rate": 1.132368770181982e-05,
+      "loss": 8.7087,
+      "step": 124850
+    },
+    {
+      "epoch": 0.6235361681939624,
+      "grad_norm": 0.09282612055540085,
+      "learning_rate": 1.132218578687827e-05,
+      "loss": 8.7295,
+      "step": 124860
+    },
+    {
+      "epoch": 0.6235861070189018,
+      "grad_norm": 0.0898682028055191,
+      "learning_rate": 1.1320683871936719e-05,
+      "loss": 8.7312,
+      "step": 124870
+    },
+    {
+      "epoch": 0.6236360458438412,
+      "grad_norm": 0.09973151236772537,
+      "learning_rate": 1.1319181956995169e-05,
+      "loss": 8.7089,
+      "step": 124880
+    },
+    {
+      "epoch": 0.6236859846687808,
+      "grad_norm": 0.10166613757610321,
+      "learning_rate": 1.131768004205362e-05,
+      "loss": 8.7065,
+      "step": 124890
+    },
+    {
+      "epoch": 0.6237359234937202,
+      "grad_norm": 0.10042425245046616,
+      "learning_rate": 1.1316178127112068e-05,
+      "loss": 8.716,
+      "step": 124900
+    },
+    {
+      "epoch": 0.6237858623186596,
+      "grad_norm": 0.09376679360866547,
+      "learning_rate": 1.1314676212170518e-05,
+      "loss": 8.7169,
+      "step": 124910
+    },
+    {
+      "epoch": 0.623835801143599,
+      "grad_norm": 0.09402406215667725,
+      "learning_rate": 1.1313174297228966e-05,
+      "loss": 8.7226,
+      "step": 124920
+    },
+    {
+      "epoch": 0.6238857399685386,
+      "grad_norm": 0.0880286768078804,
+      "learning_rate": 1.1311672382287417e-05,
+      "loss": 8.7149,
+      "step": 124930
+    },
+    {
+      "epoch": 0.623935678793478,
+      "grad_norm": 0.09521077573299408,
+      "learning_rate": 1.1310170467345867e-05,
+      "loss": 8.7109,
+      "step": 124940
+    },
+    {
+      "epoch": 0.6239856176184174,
+      "grad_norm": 0.09367305040359497,
+      "learning_rate": 1.1308668552404315e-05,
+      "loss": 8.7349,
+      "step": 124950
+    },
+    {
+      "epoch": 0.6240355564433568,
+      "grad_norm": 0.08822840452194214,
+      "learning_rate": 1.1307166637462765e-05,
+      "loss": 8.7135,
+      "step": 124960
+    },
+    {
+      "epoch": 0.6240854952682964,
+      "grad_norm": 0.09069608896970749,
+      "learning_rate": 1.1305664722521216e-05,
+      "loss": 8.7098,
+      "step": 124970
+    },
+    {
+      "epoch": 0.6241354340932358,
+      "grad_norm": 0.09736108034849167,
+      "learning_rate": 1.1304162807579664e-05,
+      "loss": 8.7174,
+      "step": 124980
+    },
+    {
+      "epoch": 0.6241853729181752,
+      "grad_norm": 0.09728176891803741,
+      "learning_rate": 1.1302660892638114e-05,
+      "loss": 8.7019,
+      "step": 124990
+    },
+    {
+      "epoch": 0.6242353117431146,
+      "grad_norm": 0.0918160006403923,
+      "learning_rate": 1.1301158977696563e-05,
+      "loss": 8.7286,
+      "step": 125000
+    },
+    {
+      "epoch": 0.6242852505680542,
+      "grad_norm": 0.09253043681383133,
+      "learning_rate": 1.1299657062755013e-05,
+      "loss": 8.7112,
+      "step": 125010
+    },
+    {
+      "epoch": 0.6243351893929936,
+      "grad_norm": 0.09095212817192078,
+      "learning_rate": 1.1298155147813463e-05,
+      "loss": 8.7239,
+      "step": 125020
+    },
+    {
+      "epoch": 0.624385128217933,
+      "grad_norm": 0.09760838001966476,
+      "learning_rate": 1.1296653232871912e-05,
+      "loss": 8.7218,
+      "step": 125030
+    },
+    {
+      "epoch": 0.6244350670428724,
+      "grad_norm": 0.09743019938468933,
+      "learning_rate": 1.1295151317930362e-05,
+      "loss": 8.7253,
+      "step": 125040
+    },
+    {
+      "epoch": 0.624485005867812,
+      "grad_norm": 0.09515538066625595,
+      "learning_rate": 1.129364940298881e-05,
+      "loss": 8.7092,
+      "step": 125050
+    },
+    {
+      "epoch": 0.6245349446927514,
+      "grad_norm": 0.08986768871545792,
+      "learning_rate": 1.129214748804726e-05,
+      "loss": 8.7231,
+      "step": 125060
+    },
+    {
+      "epoch": 0.6245848835176908,
+      "grad_norm": 0.09468900412321091,
+      "learning_rate": 1.129064557310571e-05,
+      "loss": 8.7225,
+      "step": 125070
+    },
+    {
+      "epoch": 0.6246348223426302,
+      "grad_norm": 0.0945427417755127,
+      "learning_rate": 1.1289143658164159e-05,
+      "loss": 8.7031,
+      "step": 125080
+    },
+    {
+      "epoch": 0.6246847611675698,
+      "grad_norm": 0.0921517014503479,
+      "learning_rate": 1.128764174322261e-05,
+      "loss": 8.712,
+      "step": 125090
+    },
+    {
+      "epoch": 0.6247346999925092,
+      "grad_norm": 0.09056264162063599,
+      "learning_rate": 1.1286139828281058e-05,
+      "loss": 8.7088,
+      "step": 125100
+    },
+    {
+      "epoch": 0.6247846388174486,
+      "grad_norm": 0.0919748991727829,
+      "learning_rate": 1.1284637913339508e-05,
+      "loss": 8.7197,
+      "step": 125110
+    },
+    {
+      "epoch": 0.624834577642388,
+      "grad_norm": 0.0946049764752388,
+      "learning_rate": 1.1283135998397958e-05,
+      "loss": 8.7279,
+      "step": 125120
+    },
+    {
+      "epoch": 0.6248845164673276,
+      "grad_norm": 0.09442543238401413,
+      "learning_rate": 1.1281634083456408e-05,
+      "loss": 8.7228,
+      "step": 125130
+    },
+    {
+      "epoch": 0.624934455292267,
+      "grad_norm": 0.09224551171064377,
+      "learning_rate": 1.1280132168514857e-05,
+      "loss": 8.7196,
+      "step": 125140
+    },
+    {
+      "epoch": 0.6249843941172064,
+      "grad_norm": 0.0918484553694725,
+      "learning_rate": 1.1278630253573305e-05,
+      "loss": 8.7279,
+      "step": 125150
+    },
+    {
+      "epoch": 0.6250343329421458,
+      "grad_norm": 0.08815157413482666,
+      "learning_rate": 1.1277128338631755e-05,
+      "loss": 8.7135,
+      "step": 125160
+    },
+    {
+      "epoch": 0.6250842717670854,
+      "grad_norm": 0.09147334843873978,
+      "learning_rate": 1.1275626423690206e-05,
+      "loss": 8.7166,
+      "step": 125170
+    },
+    {
+      "epoch": 0.6251342105920248,
+      "grad_norm": 0.09084770083427429,
+      "learning_rate": 1.1274124508748656e-05,
+      "loss": 8.7209,
+      "step": 125180
+    },
+    {
+      "epoch": 0.6251841494169642,
+      "grad_norm": 0.0943610891699791,
+      "learning_rate": 1.1272622593807104e-05,
+      "loss": 8.7096,
+      "step": 125190
+    },
+    {
+      "epoch": 0.6252340882419036,
+      "grad_norm": 0.09305550903081894,
+      "learning_rate": 1.1271120678865553e-05,
+      "loss": 8.7275,
+      "step": 125200
+    },
+    {
+      "epoch": 0.6252840270668432,
+      "grad_norm": 0.09299591183662415,
+      "learning_rate": 1.1269618763924003e-05,
+      "loss": 8.7027,
+      "step": 125210
+    },
+    {
+      "epoch": 0.6253339658917826,
+      "grad_norm": 0.0949455201625824,
+      "learning_rate": 1.1268116848982453e-05,
+      "loss": 8.7092,
+      "step": 125220
+    },
+    {
+      "epoch": 0.625383904716722,
+      "grad_norm": 0.09270406514406204,
+      "learning_rate": 1.1266614934040903e-05,
+      "loss": 8.7045,
+      "step": 125230
+    },
+    {
+      "epoch": 0.6254338435416614,
+      "grad_norm": 0.09729525446891785,
+      "learning_rate": 1.1265113019099352e-05,
+      "loss": 8.7111,
+      "step": 125240
+    },
+    {
+      "epoch": 0.625483782366601,
+      "grad_norm": 0.09073136001825333,
+      "learning_rate": 1.12636111041578e-05,
+      "loss": 8.7025,
+      "step": 125250
+    },
+    {
+      "epoch": 0.6255337211915404,
+      "grad_norm": 0.08994272351264954,
+      "learning_rate": 1.126210918921625e-05,
+      "loss": 8.7221,
+      "step": 125260
+    },
+    {
+      "epoch": 0.6255836600164798,
+      "grad_norm": 0.09095024317502975,
+      "learning_rate": 1.12606072742747e-05,
+      "loss": 8.7107,
+      "step": 125270
+    },
+    {
+      "epoch": 0.6256335988414192,
+      "grad_norm": 0.09382519870996475,
+      "learning_rate": 1.125910535933315e-05,
+      "loss": 8.6977,
+      "step": 125280
+    },
+    {
+      "epoch": 0.6256835376663588,
+      "grad_norm": 0.09506542235612869,
+      "learning_rate": 1.1257603444391601e-05,
+      "loss": 8.694,
+      "step": 125290
+    },
+    {
+      "epoch": 0.6257334764912982,
+      "grad_norm": 0.08988244086503983,
+      "learning_rate": 1.125610152945005e-05,
+      "loss": 8.7117,
+      "step": 125300
+    },
+    {
+      "epoch": 0.6257834153162376,
+      "grad_norm": 0.0954940915107727,
+      "learning_rate": 1.1254599614508498e-05,
+      "loss": 8.7261,
+      "step": 125310
+    },
+    {
+      "epoch": 0.625833354141177,
+      "grad_norm": 0.0926600992679596,
+      "learning_rate": 1.1253097699566948e-05,
+      "loss": 8.7071,
+      "step": 125320
+    },
+    {
+      "epoch": 0.6258832929661166,
+      "grad_norm": 0.09239501506090164,
+      "learning_rate": 1.1251595784625398e-05,
+      "loss": 8.7221,
+      "step": 125330
+    },
+    {
+      "epoch": 0.625933231791056,
+      "grad_norm": 0.09140519052743912,
+      "learning_rate": 1.1250093869683848e-05,
+      "loss": 8.7182,
+      "step": 125340
+    },
+    {
+      "epoch": 0.6259831706159954,
+      "grad_norm": 0.09351127594709396,
+      "learning_rate": 1.1248591954742297e-05,
+      "loss": 8.7191,
+      "step": 125350
+    },
+    {
+      "epoch": 0.6260331094409348,
+      "grad_norm": 0.09365622699260712,
+      "learning_rate": 1.1247090039800745e-05,
+      "loss": 8.7149,
+      "step": 125360
+    },
+    {
+      "epoch": 0.6260830482658744,
+      "grad_norm": 0.09343401342630386,
+      "learning_rate": 1.1245588124859196e-05,
+      "loss": 8.7062,
+      "step": 125370
+    },
+    {
+      "epoch": 0.6261329870908138,
+      "grad_norm": 0.09367363899946213,
+      "learning_rate": 1.1244086209917646e-05,
+      "loss": 8.7105,
+      "step": 125380
+    },
+    {
+      "epoch": 0.6261829259157532,
+      "grad_norm": 0.09021138399839401,
+      "learning_rate": 1.1242584294976096e-05,
+      "loss": 8.714,
+      "step": 125390
+    },
+    {
+      "epoch": 0.6262328647406926,
+      "grad_norm": 0.10367359220981598,
+      "learning_rate": 1.1241082380034544e-05,
+      "loss": 8.7316,
+      "step": 125400
+    },
+    {
+      "epoch": 0.6262828035656322,
+      "grad_norm": 0.09110212326049805,
+      "learning_rate": 1.1239580465092993e-05,
+      "loss": 8.723,
+      "step": 125410
+    },
+    {
+      "epoch": 0.6263327423905716,
+      "grad_norm": 0.09497211873531342,
+      "learning_rate": 1.1238078550151443e-05,
+      "loss": 8.719,
+      "step": 125420
+    },
+    {
+      "epoch": 0.626382681215511,
+      "grad_norm": 0.08793065696954727,
+      "learning_rate": 1.1236576635209893e-05,
+      "loss": 8.7118,
+      "step": 125430
+    },
+    {
+      "epoch": 0.6264326200404504,
+      "grad_norm": 0.09238415211439133,
+      "learning_rate": 1.1235074720268343e-05,
+      "loss": 8.7091,
+      "step": 125440
+    },
+    {
+      "epoch": 0.62648255886539,
+      "grad_norm": 0.08847446739673615,
+      "learning_rate": 1.1233572805326794e-05,
+      "loss": 8.7071,
+      "step": 125450
+    },
+    {
+      "epoch": 0.6265324976903294,
+      "grad_norm": 0.09379098564386368,
+      "learning_rate": 1.123207089038524e-05,
+      "loss": 8.7228,
+      "step": 125460
+    },
+    {
+      "epoch": 0.6265824365152688,
+      "grad_norm": 0.09281666576862335,
+      "learning_rate": 1.123056897544369e-05,
+      "loss": 8.7114,
+      "step": 125470
+    },
+    {
+      "epoch": 0.6266323753402082,
+      "grad_norm": 0.0954156294465065,
+      "learning_rate": 1.122906706050214e-05,
+      "loss": 8.7198,
+      "step": 125480
+    },
+    {
+      "epoch": 0.6266823141651477,
+      "grad_norm": 0.091185063123703,
+      "learning_rate": 1.1227565145560591e-05,
+      "loss": 8.7046,
+      "step": 125490
+    },
+    {
+      "epoch": 0.6267322529900872,
+      "grad_norm": 0.09116750210523605,
+      "learning_rate": 1.1226063230619041e-05,
+      "loss": 8.7127,
+      "step": 125500
+    },
+    {
+      "epoch": 0.6267821918150266,
+      "grad_norm": 0.09332124143838882,
+      "learning_rate": 1.1224561315677488e-05,
+      "loss": 8.7211,
+      "step": 125510
+    },
+    {
+      "epoch": 0.626832130639966,
+      "grad_norm": 0.08798982203006744,
+      "learning_rate": 1.1223059400735938e-05,
+      "loss": 8.6972,
+      "step": 125520
+    },
+    {
+      "epoch": 0.6268820694649055,
+      "grad_norm": 0.08885789662599564,
+      "learning_rate": 1.1221557485794388e-05,
+      "loss": 8.7219,
+      "step": 125530
+    },
+    {
+      "epoch": 0.626932008289845,
+      "grad_norm": 0.10069695860147476,
+      "learning_rate": 1.1220055570852838e-05,
+      "loss": 8.699,
+      "step": 125540
+    },
+    {
+      "epoch": 0.6269819471147844,
+      "grad_norm": 0.0930192843079567,
+      "learning_rate": 1.1218553655911289e-05,
+      "loss": 8.7112,
+      "step": 125550
+    },
+    {
+      "epoch": 0.6270318859397238,
+      "grad_norm": 0.09061025083065033,
+      "learning_rate": 1.1217051740969735e-05,
+      "loss": 8.7135,
+      "step": 125560
+    },
+    {
+      "epoch": 0.6270818247646633,
+      "grad_norm": 0.08885504305362701,
+      "learning_rate": 1.1215549826028186e-05,
+      "loss": 8.7175,
+      "step": 125570
+    },
+    {
+      "epoch": 0.6271317635896028,
+      "grad_norm": 0.09031138569116592,
+      "learning_rate": 1.1214047911086636e-05,
+      "loss": 8.7082,
+      "step": 125580
+    },
+    {
+      "epoch": 0.6271817024145422,
+      "grad_norm": 0.09189961850643158,
+      "learning_rate": 1.1212545996145086e-05,
+      "loss": 8.7175,
+      "step": 125590
+    },
+    {
+      "epoch": 0.6272316412394816,
+      "grad_norm": 0.09408345818519592,
+      "learning_rate": 1.1211044081203536e-05,
+      "loss": 8.7174,
+      "step": 125600
+    },
+    {
+      "epoch": 0.627281580064421,
+      "grad_norm": 0.09308337420225143,
+      "learning_rate": 1.1209542166261985e-05,
+      "loss": 8.7101,
+      "step": 125610
+    },
+    {
+      "epoch": 0.6273315188893606,
+      "grad_norm": 0.09275424480438232,
+      "learning_rate": 1.1208040251320433e-05,
+      "loss": 8.7109,
+      "step": 125620
+    },
+    {
+      "epoch": 0.6273814577143,
+      "grad_norm": 0.10062972456216812,
+      "learning_rate": 1.1206538336378883e-05,
+      "loss": 8.6987,
+      "step": 125630
+    },
+    {
+      "epoch": 0.6274313965392394,
+      "grad_norm": 0.09465427696704865,
+      "learning_rate": 1.1205036421437333e-05,
+      "loss": 8.7145,
+      "step": 125640
+    },
+    {
+      "epoch": 0.6274813353641788,
+      "grad_norm": 0.09311853349208832,
+      "learning_rate": 1.1203534506495784e-05,
+      "loss": 8.7253,
+      "step": 125650
+    },
+    {
+      "epoch": 0.6275312741891184,
+      "grad_norm": 0.088384248316288,
+      "learning_rate": 1.1202032591554232e-05,
+      "loss": 8.7049,
+      "step": 125660
+    },
+    {
+      "epoch": 0.6275812130140578,
+      "grad_norm": 0.09492412954568863,
+      "learning_rate": 1.120053067661268e-05,
+      "loss": 8.7136,
+      "step": 125670
+    },
+    {
+      "epoch": 0.6276311518389972,
+      "grad_norm": 0.10065683722496033,
+      "learning_rate": 1.119902876167113e-05,
+      "loss": 8.7108,
+      "step": 125680
+    },
+    {
+      "epoch": 0.6276810906639366,
+      "grad_norm": 0.0918295755982399,
+      "learning_rate": 1.1197526846729581e-05,
+      "loss": 8.7085,
+      "step": 125690
+    },
+    {
+      "epoch": 0.6277310294888762,
+      "grad_norm": 0.09792115539312363,
+      "learning_rate": 1.1196024931788031e-05,
+      "loss": 8.6919,
+      "step": 125700
+    },
+    {
+      "epoch": 0.6277809683138156,
+      "grad_norm": 0.08935079723596573,
+      "learning_rate": 1.119452301684648e-05,
+      "loss": 8.7047,
+      "step": 125710
+    },
+    {
+      "epoch": 0.627830907138755,
+      "grad_norm": 0.09444247931241989,
+      "learning_rate": 1.1193021101904928e-05,
+      "loss": 8.7183,
+      "step": 125720
+    },
+    {
+      "epoch": 0.6278808459636944,
+      "grad_norm": 0.09201852232217789,
+      "learning_rate": 1.1191519186963378e-05,
+      "loss": 8.7243,
+      "step": 125730
+    },
+    {
+      "epoch": 0.627930784788634,
+      "grad_norm": 0.09014883637428284,
+      "learning_rate": 1.1190017272021828e-05,
+      "loss": 8.7091,
+      "step": 125740
+    },
+    {
+      "epoch": 0.6279807236135734,
+      "grad_norm": 0.091966412961483,
+      "learning_rate": 1.1188515357080279e-05,
+      "loss": 8.7245,
+      "step": 125750
+    },
+    {
+      "epoch": 0.6280306624385128,
+      "grad_norm": 0.08749305456876755,
+      "learning_rate": 1.1187013442138727e-05,
+      "loss": 8.7155,
+      "step": 125760
+    },
+    {
+      "epoch": 0.6280806012634522,
+      "grad_norm": 0.09407436102628708,
+      "learning_rate": 1.1185511527197177e-05,
+      "loss": 8.7034,
+      "step": 125770
+    },
+    {
+      "epoch": 0.6281305400883918,
+      "grad_norm": 0.09107963740825653,
+      "learning_rate": 1.1184009612255626e-05,
+      "loss": 8.7061,
+      "step": 125780
+    },
+    {
+      "epoch": 0.6281804789133312,
+      "grad_norm": 0.09338110685348511,
+      "learning_rate": 1.1182507697314076e-05,
+      "loss": 8.7152,
+      "step": 125790
+    },
+    {
+      "epoch": 0.6282304177382706,
+      "grad_norm": 0.09084051847457886,
+      "learning_rate": 1.1181005782372526e-05,
+      "loss": 8.7097,
+      "step": 125800
+    },
+    {
+      "epoch": 0.62828035656321,
+      "grad_norm": 0.09835775941610336,
+      "learning_rate": 1.1179503867430975e-05,
+      "loss": 8.7238,
+      "step": 125810
+    },
+    {
+      "epoch": 0.6283302953881496,
+      "grad_norm": 0.09301155060529709,
+      "learning_rate": 1.1178001952489425e-05,
+      "loss": 8.7027,
+      "step": 125820
+    },
+    {
+      "epoch": 0.628380234213089,
+      "grad_norm": 0.08785326033830643,
+      "learning_rate": 1.1176500037547873e-05,
+      "loss": 8.7189,
+      "step": 125830
+    },
+    {
+      "epoch": 0.6284301730380284,
+      "grad_norm": 0.08803771436214447,
+      "learning_rate": 1.1174998122606323e-05,
+      "loss": 8.7123,
+      "step": 125840
+    },
+    {
+      "epoch": 0.6284801118629678,
+      "grad_norm": 0.0908784344792366,
+      "learning_rate": 1.1173496207664774e-05,
+      "loss": 8.7182,
+      "step": 125850
+    },
+    {
+      "epoch": 0.6285300506879073,
+      "grad_norm": 0.0862756222486496,
+      "learning_rate": 1.1171994292723222e-05,
+      "loss": 8.7111,
+      "step": 125860
+    },
+    {
+      "epoch": 0.6285799895128468,
+      "grad_norm": 0.09046090394258499,
+      "learning_rate": 1.1170492377781672e-05,
+      "loss": 8.7195,
+      "step": 125870
+    },
+    {
+      "epoch": 0.6286299283377862,
+      "grad_norm": 0.08847752958536148,
+      "learning_rate": 1.116899046284012e-05,
+      "loss": 8.7092,
+      "step": 125880
+    },
+    {
+      "epoch": 0.6286798671627256,
+      "grad_norm": 0.09415357559919357,
+      "learning_rate": 1.1167488547898571e-05,
+      "loss": 8.6979,
+      "step": 125890
+    },
+    {
+      "epoch": 0.6287298059876651,
+      "grad_norm": 0.09321362525224686,
+      "learning_rate": 1.1165986632957021e-05,
+      "loss": 8.7019,
+      "step": 125900
+    },
+    {
+      "epoch": 0.6287797448126046,
+      "grad_norm": 0.09205802530050278,
+      "learning_rate": 1.116448471801547e-05,
+      "loss": 8.7374,
+      "step": 125910
+    },
+    {
+      "epoch": 0.628829683637544,
+      "grad_norm": 0.0888669490814209,
+      "learning_rate": 1.116298280307392e-05,
+      "loss": 8.7149,
+      "step": 125920
+    },
+    {
+      "epoch": 0.6288796224624834,
+      "grad_norm": 0.08783555775880814,
+      "learning_rate": 1.1161480888132368e-05,
+      "loss": 8.7195,
+      "step": 125930
+    },
+    {
+      "epoch": 0.628929561287423,
+      "grad_norm": 0.09451974183320999,
+      "learning_rate": 1.1159978973190818e-05,
+      "loss": 8.7245,
+      "step": 125940
+    },
+    {
+      "epoch": 0.6289795001123624,
+      "grad_norm": 0.09017028659582138,
+      "learning_rate": 1.1158477058249269e-05,
+      "loss": 8.7217,
+      "step": 125950
+    },
+    {
+      "epoch": 0.6290294389373018,
+      "grad_norm": 0.09248178452253342,
+      "learning_rate": 1.1156975143307717e-05,
+      "loss": 8.7001,
+      "step": 125960
+    },
+    {
+      "epoch": 0.6290793777622412,
+      "grad_norm": 0.09288008511066437,
+      "learning_rate": 1.1155473228366167e-05,
+      "loss": 8.6956,
+      "step": 125970
+    },
+    {
+      "epoch": 0.6291293165871807,
+      "grad_norm": 0.08859378844499588,
+      "learning_rate": 1.1153971313424617e-05,
+      "loss": 8.7073,
+      "step": 125980
+    },
+    {
+      "epoch": 0.6291792554121202,
+      "grad_norm": 0.09456798434257507,
+      "learning_rate": 1.1152469398483066e-05,
+      "loss": 8.7203,
+      "step": 125990
+    },
+    {
+      "epoch": 0.6292291942370596,
+      "grad_norm": 0.09040806442499161,
+      "learning_rate": 1.1150967483541516e-05,
+      "loss": 8.711,
+      "step": 126000
+    },
+    {
+      "epoch": 0.629279133061999,
+      "grad_norm": 0.09092914313077927,
+      "learning_rate": 1.1149465568599965e-05,
+      "loss": 8.7157,
+      "step": 126010
+    },
+    {
+      "epoch": 0.6293290718869385,
+      "grad_norm": 0.08959808945655823,
+      "learning_rate": 1.1147963653658415e-05,
+      "loss": 8.7037,
+      "step": 126020
+    },
+    {
+      "epoch": 0.629379010711878,
+      "grad_norm": 0.09273937344551086,
+      "learning_rate": 1.1146461738716865e-05,
+      "loss": 8.7036,
+      "step": 126030
+    },
+    {
+      "epoch": 0.6294289495368174,
+      "grad_norm": 0.09658925980329514,
+      "learning_rate": 1.1144959823775313e-05,
+      "loss": 8.7071,
+      "step": 126040
+    },
+    {
+      "epoch": 0.6294788883617568,
+      "grad_norm": 0.09372569620609283,
+      "learning_rate": 1.1143457908833764e-05,
+      "loss": 8.7113,
+      "step": 126050
+    },
+    {
+      "epoch": 0.6295288271866963,
+      "grad_norm": 0.08820081502199173,
+      "learning_rate": 1.1141955993892212e-05,
+      "loss": 8.7279,
+      "step": 126060
+    },
+    {
+      "epoch": 0.6295787660116358,
+      "grad_norm": 0.09399138391017914,
+      "learning_rate": 1.1140454078950662e-05,
+      "loss": 8.7203,
+      "step": 126070
+    },
+    {
+      "epoch": 0.6296287048365752,
+      "grad_norm": 0.09462126344442368,
+      "learning_rate": 1.1138952164009112e-05,
+      "loss": 8.7135,
+      "step": 126080
+    },
+    {
+      "epoch": 0.6296786436615146,
+      "grad_norm": 0.09695866703987122,
+      "learning_rate": 1.1137450249067561e-05,
+      "loss": 8.6986,
+      "step": 126090
+    },
+    {
+      "epoch": 0.6297285824864541,
+      "grad_norm": 0.09450877457857132,
+      "learning_rate": 1.1135948334126011e-05,
+      "loss": 8.7052,
+      "step": 126100
+    },
+    {
+      "epoch": 0.6297785213113936,
+      "grad_norm": 0.09167155623435974,
+      "learning_rate": 1.113444641918446e-05,
+      "loss": 8.7027,
+      "step": 126110
+    },
+    {
+      "epoch": 0.629828460136333,
+      "grad_norm": 0.09358896315097809,
+      "learning_rate": 1.113294450424291e-05,
+      "loss": 8.7079,
+      "step": 126120
+    },
+    {
+      "epoch": 0.6298783989612724,
+      "grad_norm": 0.10010679066181183,
+      "learning_rate": 1.113144258930136e-05,
+      "loss": 8.701,
+      "step": 126130
+    },
+    {
+      "epoch": 0.6299283377862119,
+      "grad_norm": 0.09216713160276413,
+      "learning_rate": 1.112994067435981e-05,
+      "loss": 8.7026,
+      "step": 126140
+    },
+    {
+      "epoch": 0.6299782766111514,
+      "grad_norm": 0.08950921893119812,
+      "learning_rate": 1.1128438759418259e-05,
+      "loss": 8.6985,
+      "step": 126150
+    },
+    {
+      "epoch": 0.6300282154360908,
+      "grad_norm": 0.09360364824533463,
+      "learning_rate": 1.1126936844476707e-05,
+      "loss": 8.728,
+      "step": 126160
+    },
+    {
+      "epoch": 0.6300781542610302,
+      "grad_norm": 0.09034190326929092,
+      "learning_rate": 1.1125434929535157e-05,
+      "loss": 8.7074,
+      "step": 126170
+    },
+    {
+      "epoch": 0.6301280930859697,
+      "grad_norm": 0.09222965687513351,
+      "learning_rate": 1.1123933014593607e-05,
+      "loss": 8.7075,
+      "step": 126180
+    },
+    {
+      "epoch": 0.6301780319109092,
+      "grad_norm": 0.09020216763019562,
+      "learning_rate": 1.1122431099652058e-05,
+      "loss": 8.6951,
+      "step": 126190
+    },
+    {
+      "epoch": 0.6302279707358486,
+      "grad_norm": 0.09144456684589386,
+      "learning_rate": 1.1120929184710506e-05,
+      "loss": 8.7062,
+      "step": 126200
+    },
+    {
+      "epoch": 0.630277909560788,
+      "grad_norm": 0.09432315826416016,
+      "learning_rate": 1.1119427269768955e-05,
+      "loss": 8.712,
+      "step": 126210
+    },
+    {
+      "epoch": 0.6303278483857275,
+      "grad_norm": 0.09091579914093018,
+      "learning_rate": 1.1117925354827405e-05,
+      "loss": 8.7356,
+      "step": 126220
+    },
+    {
+      "epoch": 0.630377787210667,
+      "grad_norm": 0.08974793553352356,
+      "learning_rate": 1.1116423439885855e-05,
+      "loss": 8.7139,
+      "step": 126230
+    },
+    {
+      "epoch": 0.6304277260356064,
+      "grad_norm": 0.09444157779216766,
+      "learning_rate": 1.1114921524944305e-05,
+      "loss": 8.7199,
+      "step": 126240
+    },
+    {
+      "epoch": 0.6304776648605458,
+      "grad_norm": 0.09196534752845764,
+      "learning_rate": 1.1113419610002754e-05,
+      "loss": 8.7038,
+      "step": 126250
+    },
+    {
+      "epoch": 0.6305276036854853,
+      "grad_norm": 0.09097559005022049,
+      "learning_rate": 1.1111917695061202e-05,
+      "loss": 8.7136,
+      "step": 126260
+    },
+    {
+      "epoch": 0.6305775425104247,
+      "grad_norm": 0.08905323594808578,
+      "learning_rate": 1.1110415780119652e-05,
+      "loss": 8.7215,
+      "step": 126270
+    },
+    {
+      "epoch": 0.6306274813353642,
+      "grad_norm": 0.09112866222858429,
+      "learning_rate": 1.1108913865178103e-05,
+      "loss": 8.7085,
+      "step": 126280
+    },
+    {
+      "epoch": 0.6306774201603036,
+      "grad_norm": 0.08942396938800812,
+      "learning_rate": 1.1107411950236553e-05,
+      "loss": 8.7192,
+      "step": 126290
+    },
+    {
+      "epoch": 0.6307273589852431,
+      "grad_norm": 0.09311738610267639,
+      "learning_rate": 1.1105910035295003e-05,
+      "loss": 8.6981,
+      "step": 126300
+    },
+    {
+      "epoch": 0.6307772978101825,
+      "grad_norm": 0.09299565851688385,
+      "learning_rate": 1.110440812035345e-05,
+      "loss": 8.71,
+      "step": 126310
+    },
+    {
+      "epoch": 0.630827236635122,
+      "grad_norm": 0.09112560003995895,
+      "learning_rate": 1.11029062054119e-05,
+      "loss": 8.6983,
+      "step": 126320
+    },
+    {
+      "epoch": 0.6308771754600614,
+      "grad_norm": 0.09203418344259262,
+      "learning_rate": 1.110140429047035e-05,
+      "loss": 8.7096,
+      "step": 126330
+    },
+    {
+      "epoch": 0.6309271142850009,
+      "grad_norm": 0.09365582466125488,
+      "learning_rate": 1.10999023755288e-05,
+      "loss": 8.6942,
+      "step": 126340
+    },
+    {
+      "epoch": 0.6309770531099403,
+      "grad_norm": 0.09594764560461044,
+      "learning_rate": 1.109840046058725e-05,
+      "loss": 8.7237,
+      "step": 126350
+    },
+    {
+      "epoch": 0.6310269919348798,
+      "grad_norm": 0.08934248983860016,
+      "learning_rate": 1.1096898545645697e-05,
+      "loss": 8.702,
+      "step": 126360
+    },
+    {
+      "epoch": 0.6310769307598192,
+      "grad_norm": 0.09041931480169296,
+      "learning_rate": 1.1095396630704147e-05,
+      "loss": 8.7068,
+      "step": 126370
+    },
+    {
+      "epoch": 0.6311268695847587,
+      "grad_norm": 0.09372342377901077,
+      "learning_rate": 1.1093894715762598e-05,
+      "loss": 8.6993,
+      "step": 126380
+    },
+    {
+      "epoch": 0.6311768084096981,
+      "grad_norm": 0.09366505593061447,
+      "learning_rate": 1.1092392800821048e-05,
+      "loss": 8.7181,
+      "step": 126390
+    },
+    {
+      "epoch": 0.6312267472346376,
+      "grad_norm": 0.09063727408647537,
+      "learning_rate": 1.1090890885879498e-05,
+      "loss": 8.7059,
+      "step": 126400
+    },
+    {
+      "epoch": 0.631276686059577,
+      "grad_norm": 0.0880524069070816,
+      "learning_rate": 1.1089388970937945e-05,
+      "loss": 8.7184,
+      "step": 126410
+    },
+    {
+      "epoch": 0.6313266248845165,
+      "grad_norm": 0.09939765185117722,
+      "learning_rate": 1.1087887055996395e-05,
+      "loss": 8.7115,
+      "step": 126420
+    },
+    {
+      "epoch": 0.6313765637094559,
+      "grad_norm": 0.09165541082620621,
+      "learning_rate": 1.1086385141054845e-05,
+      "loss": 8.7044,
+      "step": 126430
+    },
+    {
+      "epoch": 0.6314265025343954,
+      "grad_norm": 0.08993414044380188,
+      "learning_rate": 1.1084883226113295e-05,
+      "loss": 8.7114,
+      "step": 126440
+    },
+    {
+      "epoch": 0.6314764413593348,
+      "grad_norm": 0.09997765719890594,
+      "learning_rate": 1.1083381311171745e-05,
+      "loss": 8.7137,
+      "step": 126450
+    },
+    {
+      "epoch": 0.6315263801842743,
+      "grad_norm": 0.09372782707214355,
+      "learning_rate": 1.1081879396230194e-05,
+      "loss": 8.6919,
+      "step": 126460
+    },
+    {
+      "epoch": 0.6315763190092137,
+      "grad_norm": 0.09590387344360352,
+      "learning_rate": 1.1080377481288642e-05,
+      "loss": 8.7069,
+      "step": 126470
+    },
+    {
+      "epoch": 0.6316262578341532,
+      "grad_norm": 0.09233605861663818,
+      "learning_rate": 1.1078875566347093e-05,
+      "loss": 8.694,
+      "step": 126480
+    },
+    {
+      "epoch": 0.6316761966590926,
+      "grad_norm": 0.09469950944185257,
+      "learning_rate": 1.1077373651405543e-05,
+      "loss": 8.7218,
+      "step": 126490
+    },
+    {
+      "epoch": 0.6317261354840321,
+      "grad_norm": 0.08625823259353638,
+      "learning_rate": 1.1075871736463993e-05,
+      "loss": 8.7083,
+      "step": 126500
+    },
+    {
+      "epoch": 0.6317760743089715,
+      "grad_norm": 0.09083887934684753,
+      "learning_rate": 1.1074369821522441e-05,
+      "loss": 8.6994,
+      "step": 126510
+    },
+    {
+      "epoch": 0.631826013133911,
+      "grad_norm": 0.097719706594944,
+      "learning_rate": 1.107286790658089e-05,
+      "loss": 8.7016,
+      "step": 126520
+    },
+    {
+      "epoch": 0.6318759519588504,
+      "grad_norm": 0.09654133021831512,
+      "learning_rate": 1.107136599163934e-05,
+      "loss": 8.7169,
+      "step": 126530
+    },
+    {
+      "epoch": 0.6319258907837899,
+      "grad_norm": 0.09275779873132706,
+      "learning_rate": 1.106986407669779e-05,
+      "loss": 8.7157,
+      "step": 126540
+    },
+    {
+      "epoch": 0.6319758296087293,
+      "grad_norm": 0.08812788128852844,
+      "learning_rate": 1.106836216175624e-05,
+      "loss": 8.6997,
+      "step": 126550
+    },
+    {
+      "epoch": 0.6320257684336688,
+      "grad_norm": 0.09277535229921341,
+      "learning_rate": 1.1066860246814689e-05,
+      "loss": 8.7109,
+      "step": 126560
+    },
+    {
+      "epoch": 0.6320757072586082,
+      "grad_norm": 0.09549902379512787,
+      "learning_rate": 1.1065358331873137e-05,
+      "loss": 8.701,
+      "step": 126570
+    },
+    {
+      "epoch": 0.6321256460835476,
+      "grad_norm": 0.08873405307531357,
+      "learning_rate": 1.1063856416931588e-05,
+      "loss": 8.7066,
+      "step": 126580
+    },
+    {
+      "epoch": 0.6321755849084871,
+      "grad_norm": 0.09165298938751221,
+      "learning_rate": 1.1062354501990038e-05,
+      "loss": 8.7016,
+      "step": 126590
+    },
+    {
+      "epoch": 0.6322255237334266,
+      "grad_norm": 0.09139636904001236,
+      "learning_rate": 1.1060852587048488e-05,
+      "loss": 8.7104,
+      "step": 126600
+    },
+    {
+      "epoch": 0.632275462558366,
+      "grad_norm": 0.09368384629487991,
+      "learning_rate": 1.1059350672106936e-05,
+      "loss": 8.7085,
+      "step": 126610
+    },
+    {
+      "epoch": 0.6323254013833054,
+      "grad_norm": 0.0939926728606224,
+      "learning_rate": 1.1057848757165387e-05,
+      "loss": 8.7038,
+      "step": 126620
+    },
+    {
+      "epoch": 0.6323753402082449,
+      "grad_norm": 0.0963914766907692,
+      "learning_rate": 1.1056346842223835e-05,
+      "loss": 8.7162,
+      "step": 126630
+    },
+    {
+      "epoch": 0.6324252790331844,
+      "grad_norm": 0.08765320479869843,
+      "learning_rate": 1.1054844927282285e-05,
+      "loss": 8.7177,
+      "step": 126640
+    },
+    {
+      "epoch": 0.6324752178581238,
+      "grad_norm": 0.09075529873371124,
+      "learning_rate": 1.1053343012340735e-05,
+      "loss": 8.7002,
+      "step": 126650
+    },
+    {
+      "epoch": 0.6325251566830632,
+      "grad_norm": 0.09381550550460815,
+      "learning_rate": 1.1051841097399184e-05,
+      "loss": 8.7145,
+      "step": 126660
+    },
+    {
+      "epoch": 0.6325750955080027,
+      "grad_norm": 0.09270758181810379,
+      "learning_rate": 1.1050339182457634e-05,
+      "loss": 8.7087,
+      "step": 126670
+    },
+    {
+      "epoch": 0.6326250343329421,
+      "grad_norm": 0.08957044035196304,
+      "learning_rate": 1.1048837267516083e-05,
+      "loss": 8.7222,
+      "step": 126680
+    },
+    {
+      "epoch": 0.6326749731578816,
+      "grad_norm": 0.09274608641862869,
+      "learning_rate": 1.1047335352574533e-05,
+      "loss": 8.7215,
+      "step": 126690
+    },
+    {
+      "epoch": 0.632724911982821,
+      "grad_norm": 0.08785900473594666,
+      "learning_rate": 1.1045833437632983e-05,
+      "loss": 8.7077,
+      "step": 126700
+    },
+    {
+      "epoch": 0.6327748508077605,
+      "grad_norm": 0.0963745191693306,
+      "learning_rate": 1.1044331522691431e-05,
+      "loss": 8.7214,
+      "step": 126710
+    },
+    {
+      "epoch": 0.6328247896327,
+      "grad_norm": 0.09362554550170898,
+      "learning_rate": 1.1042829607749882e-05,
+      "loss": 8.6988,
+      "step": 126720
+    },
+    {
+      "epoch": 0.6328747284576394,
+      "grad_norm": 0.09830887615680695,
+      "learning_rate": 1.104132769280833e-05,
+      "loss": 8.7249,
+      "step": 126730
+    },
+    {
+      "epoch": 0.6329246672825788,
+      "grad_norm": 0.09366270899772644,
+      "learning_rate": 1.103982577786678e-05,
+      "loss": 8.714,
+      "step": 126740
+    },
+    {
+      "epoch": 0.6329746061075183,
+      "grad_norm": 0.09500569850206375,
+      "learning_rate": 1.103832386292523e-05,
+      "loss": 8.7131,
+      "step": 126750
+    },
+    {
+      "epoch": 0.6330245449324577,
+      "grad_norm": 0.09202902019023895,
+      "learning_rate": 1.1036821947983679e-05,
+      "loss": 8.71,
+      "step": 126760
+    },
+    {
+      "epoch": 0.6330744837573972,
+      "grad_norm": 0.09248954057693481,
+      "learning_rate": 1.1035320033042129e-05,
+      "loss": 8.6971,
+      "step": 126770
+    },
+    {
+      "epoch": 0.6331244225823366,
+      "grad_norm": 0.09548656642436981,
+      "learning_rate": 1.103381811810058e-05,
+      "loss": 8.7115,
+      "step": 126780
+    },
+    {
+      "epoch": 0.6331743614072761,
+      "grad_norm": 0.09098765254020691,
+      "learning_rate": 1.1032316203159028e-05,
+      "loss": 8.6862,
+      "step": 126790
+    },
+    {
+      "epoch": 0.6332243002322155,
+      "grad_norm": 0.08989952504634857,
+      "learning_rate": 1.1030814288217478e-05,
+      "loss": 8.7133,
+      "step": 126800
+    },
+    {
+      "epoch": 0.633274239057155,
+      "grad_norm": 0.09226398915052414,
+      "learning_rate": 1.1029312373275926e-05,
+      "loss": 8.7104,
+      "step": 126810
+    },
+    {
+      "epoch": 0.6333241778820944,
+      "grad_norm": 0.09326343238353729,
+      "learning_rate": 1.1027810458334377e-05,
+      "loss": 8.7034,
+      "step": 126820
+    },
+    {
+      "epoch": 0.6333741167070339,
+      "grad_norm": 0.09733281284570694,
+      "learning_rate": 1.1026308543392827e-05,
+      "loss": 8.6897,
+      "step": 126830
+    },
+    {
+      "epoch": 0.6334240555319733,
+      "grad_norm": 0.0914359837770462,
+      "learning_rate": 1.1024806628451275e-05,
+      "loss": 8.7046,
+      "step": 126840
+    },
+    {
+      "epoch": 0.6334739943569128,
+      "grad_norm": 0.08934281766414642,
+      "learning_rate": 1.1023304713509725e-05,
+      "loss": 8.7141,
+      "step": 126850
+    },
+    {
+      "epoch": 0.6335239331818522,
+      "grad_norm": 0.09313992410898209,
+      "learning_rate": 1.1021802798568174e-05,
+      "loss": 8.7073,
+      "step": 126860
+    },
+    {
+      "epoch": 0.6335738720067917,
+      "grad_norm": 0.09446176141500473,
+      "learning_rate": 1.1020300883626624e-05,
+      "loss": 8.6979,
+      "step": 126870
+    },
+    {
+      "epoch": 0.6336238108317311,
+      "grad_norm": 0.09279874712228775,
+      "learning_rate": 1.1018798968685074e-05,
+      "loss": 8.6945,
+      "step": 126880
+    },
+    {
+      "epoch": 0.6336737496566706,
+      "grad_norm": 0.09727776795625687,
+      "learning_rate": 1.1017297053743523e-05,
+      "loss": 8.6937,
+      "step": 126890
+    },
+    {
+      "epoch": 0.63372368848161,
+      "grad_norm": 0.09010564535856247,
+      "learning_rate": 1.1015795138801973e-05,
+      "loss": 8.7025,
+      "step": 126900
+    },
+    {
+      "epoch": 0.6337736273065495,
+      "grad_norm": 0.08602187037467957,
+      "learning_rate": 1.1014293223860421e-05,
+      "loss": 8.7096,
+      "step": 126910
+    },
+    {
+      "epoch": 0.6338235661314889,
+      "grad_norm": 0.09525911509990692,
+      "learning_rate": 1.1012791308918872e-05,
+      "loss": 8.7116,
+      "step": 126920
+    },
+    {
+      "epoch": 0.6338735049564284,
+      "grad_norm": 0.08880488574504852,
+      "learning_rate": 1.1011289393977322e-05,
+      "loss": 8.7006,
+      "step": 126930
+    },
+    {
+      "epoch": 0.6339234437813678,
+      "grad_norm": 0.09625755250453949,
+      "learning_rate": 1.1009787479035772e-05,
+      "loss": 8.6934,
+      "step": 126940
+    },
+    {
+      "epoch": 0.6339733826063073,
+      "grad_norm": 0.0941665768623352,
+      "learning_rate": 1.100828556409422e-05,
+      "loss": 8.7106,
+      "step": 126950
+    },
+    {
+      "epoch": 0.6340233214312467,
+      "grad_norm": 0.08782870322465897,
+      "learning_rate": 1.1006783649152669e-05,
+      "loss": 8.7069,
+      "step": 126960
+    },
+    {
+      "epoch": 0.6340732602561862,
+      "grad_norm": 0.08948568999767303,
+      "learning_rate": 1.1005281734211119e-05,
+      "loss": 8.7111,
+      "step": 126970
+    },
+    {
+      "epoch": 0.6341231990811256,
+      "grad_norm": 0.09504511207342148,
+      "learning_rate": 1.100377981926957e-05,
+      "loss": 8.7132,
+      "step": 126980
+    },
+    {
+      "epoch": 0.6341731379060651,
+      "grad_norm": 0.09073743224143982,
+      "learning_rate": 1.100227790432802e-05,
+      "loss": 8.6928,
+      "step": 126990
+    },
+    {
+      "epoch": 0.6342230767310045,
+      "grad_norm": 0.09450805932283401,
+      "learning_rate": 1.1000775989386468e-05,
+      "loss": 8.6902,
+      "step": 127000
+    },
+    {
+      "epoch": 0.634273015555944,
+      "grad_norm": 0.08821328729391098,
+      "learning_rate": 1.0999274074444916e-05,
+      "loss": 8.7069,
+      "step": 127010
+    },
+    {
+      "epoch": 0.6343229543808834,
+      "grad_norm": 0.09215594828128815,
+      "learning_rate": 1.0997772159503367e-05,
+      "loss": 8.6957,
+      "step": 127020
+    },
+    {
+      "epoch": 0.6343728932058229,
+      "grad_norm": 0.09893883019685745,
+      "learning_rate": 1.0996270244561817e-05,
+      "loss": 8.6884,
+      "step": 127030
+    },
+    {
+      "epoch": 0.6344228320307623,
+      "grad_norm": 0.0937436893582344,
+      "learning_rate": 1.0994768329620267e-05,
+      "loss": 8.7161,
+      "step": 127040
+    },
+    {
+      "epoch": 0.6344727708557018,
+      "grad_norm": 0.08972315490245819,
+      "learning_rate": 1.0993266414678715e-05,
+      "loss": 8.6941,
+      "step": 127050
+    },
+    {
+      "epoch": 0.6345227096806412,
+      "grad_norm": 0.09105207026004791,
+      "learning_rate": 1.0991764499737164e-05,
+      "loss": 8.7012,
+      "step": 127060
+    },
+    {
+      "epoch": 0.6345726485055807,
+      "grad_norm": 0.09180324524641037,
+      "learning_rate": 1.0990262584795614e-05,
+      "loss": 8.6961,
+      "step": 127070
+    },
+    {
+      "epoch": 0.6346225873305201,
+      "grad_norm": 0.09050474315881729,
+      "learning_rate": 1.0988760669854064e-05,
+      "loss": 8.7129,
+      "step": 127080
+    },
+    {
+      "epoch": 0.6346725261554595,
+      "grad_norm": 0.08975604176521301,
+      "learning_rate": 1.0987258754912514e-05,
+      "loss": 8.7069,
+      "step": 127090
+    },
+    {
+      "epoch": 0.634722464980399,
+      "grad_norm": 0.09653442353010178,
+      "learning_rate": 1.0985756839970965e-05,
+      "loss": 8.6998,
+      "step": 127100
+    },
+    {
+      "epoch": 0.6347724038053385,
+      "grad_norm": 0.09031674265861511,
+      "learning_rate": 1.0984254925029411e-05,
+      "loss": 8.7128,
+      "step": 127110
+    },
+    {
+      "epoch": 0.6348223426302779,
+      "grad_norm": 0.09196629375219345,
+      "learning_rate": 1.0982753010087862e-05,
+      "loss": 8.7115,
+      "step": 127120
+    },
+    {
+      "epoch": 0.6348722814552173,
+      "grad_norm": 0.09125792980194092,
+      "learning_rate": 1.0981251095146312e-05,
+      "loss": 8.6958,
+      "step": 127130
+    },
+    {
+      "epoch": 0.6349222202801568,
+      "grad_norm": 0.09218353778123856,
+      "learning_rate": 1.0979749180204762e-05,
+      "loss": 8.7032,
+      "step": 127140
+    },
+    {
+      "epoch": 0.6349721591050963,
+      "grad_norm": 0.09349720180034637,
+      "learning_rate": 1.0978247265263212e-05,
+      "loss": 8.698,
+      "step": 127150
+    },
+    {
+      "epoch": 0.6350220979300357,
+      "grad_norm": 0.0936102643609047,
+      "learning_rate": 1.0976745350321659e-05,
+      "loss": 8.6991,
+      "step": 127160
+    },
+    {
+      "epoch": 0.6350720367549751,
+      "grad_norm": 0.0921725407242775,
+      "learning_rate": 1.0975243435380109e-05,
+      "loss": 8.7104,
+      "step": 127170
+    },
+    {
+      "epoch": 0.6351219755799146,
+      "grad_norm": 0.09616363793611526,
+      "learning_rate": 1.097374152043856e-05,
+      "loss": 8.7191,
+      "step": 127180
+    },
+    {
+      "epoch": 0.6351719144048541,
+      "grad_norm": 0.09801583737134933,
+      "learning_rate": 1.097223960549701e-05,
+      "loss": 8.702,
+      "step": 127190
+    },
+    {
+      "epoch": 0.6352218532297935,
+      "grad_norm": 0.09138066321611404,
+      "learning_rate": 1.097073769055546e-05,
+      "loss": 8.7066,
+      "step": 127200
+    },
+    {
+      "epoch": 0.6352717920547329,
+      "grad_norm": 0.09149125218391418,
+      "learning_rate": 1.0969235775613906e-05,
+      "loss": 8.7125,
+      "step": 127210
+    },
+    {
+      "epoch": 0.6353217308796724,
+      "grad_norm": 0.0975177139043808,
+      "learning_rate": 1.0967733860672357e-05,
+      "loss": 8.7044,
+      "step": 127220
+    },
+    {
+      "epoch": 0.6353716697046119,
+      "grad_norm": 0.08766981214284897,
+      "learning_rate": 1.0966231945730807e-05,
+      "loss": 8.7172,
+      "step": 127230
+    },
+    {
+      "epoch": 0.6354216085295513,
+      "grad_norm": 0.08835364878177643,
+      "learning_rate": 1.0964730030789257e-05,
+      "loss": 8.7012,
+      "step": 127240
+    },
+    {
+      "epoch": 0.6354715473544907,
+      "grad_norm": 0.09096872806549072,
+      "learning_rate": 1.0963228115847707e-05,
+      "loss": 8.702,
+      "step": 127250
+    },
+    {
+      "epoch": 0.6355214861794302,
+      "grad_norm": 0.09311988204717636,
+      "learning_rate": 1.0961726200906156e-05,
+      "loss": 8.6887,
+      "step": 127260
+    },
+    {
+      "epoch": 0.6355714250043697,
+      "grad_norm": 0.0934610664844513,
+      "learning_rate": 1.0960224285964604e-05,
+      "loss": 8.7063,
+      "step": 127270
+    },
+    {
+      "epoch": 0.6356213638293091,
+      "grad_norm": 0.098729208111763,
+      "learning_rate": 1.0958722371023054e-05,
+      "loss": 8.7185,
+      "step": 127280
+    },
+    {
+      "epoch": 0.6356713026542485,
+      "grad_norm": 0.08784567564725876,
+      "learning_rate": 1.0957220456081504e-05,
+      "loss": 8.6937,
+      "step": 127290
+    },
+    {
+      "epoch": 0.635721241479188,
+      "grad_norm": 0.08716242760419846,
+      "learning_rate": 1.0955718541139955e-05,
+      "loss": 8.7219,
+      "step": 127300
+    },
+    {
+      "epoch": 0.6357711803041275,
+      "grad_norm": 0.093071848154068,
+      "learning_rate": 1.0954216626198403e-05,
+      "loss": 8.7022,
+      "step": 127310
+    },
+    {
+      "epoch": 0.6358211191290669,
+      "grad_norm": 0.09252569824457169,
+      "learning_rate": 1.0952714711256852e-05,
+      "loss": 8.7088,
+      "step": 127320
+    },
+    {
+      "epoch": 0.6358710579540063,
+      "grad_norm": 0.09083006531000137,
+      "learning_rate": 1.0951212796315302e-05,
+      "loss": 8.7163,
+      "step": 127330
+    },
+    {
+      "epoch": 0.6359209967789458,
+      "grad_norm": 0.08883119374513626,
+      "learning_rate": 1.0949710881373752e-05,
+      "loss": 8.6963,
+      "step": 127340
+    },
+    {
+      "epoch": 0.6359709356038853,
+      "grad_norm": 0.09394201636314392,
+      "learning_rate": 1.0948208966432202e-05,
+      "loss": 8.7119,
+      "step": 127350
+    },
+    {
+      "epoch": 0.6360208744288247,
+      "grad_norm": 0.10139136761426926,
+      "learning_rate": 1.094670705149065e-05,
+      "loss": 8.704,
+      "step": 127360
+    },
+    {
+      "epoch": 0.6360708132537641,
+      "grad_norm": 0.09166630357503891,
+      "learning_rate": 1.0945205136549099e-05,
+      "loss": 8.6997,
+      "step": 127370
+    },
+    {
+      "epoch": 0.6361207520787036,
+      "grad_norm": 0.10158146917819977,
+      "learning_rate": 1.094370322160755e-05,
+      "loss": 8.7121,
+      "step": 127380
+    },
+    {
+      "epoch": 0.6361706909036431,
+      "grad_norm": 0.08964638411998749,
+      "learning_rate": 1.0942201306666e-05,
+      "loss": 8.6963,
+      "step": 127390
+    },
+    {
+      "epoch": 0.6362206297285825,
+      "grad_norm": 0.09380380064249039,
+      "learning_rate": 1.094069939172445e-05,
+      "loss": 8.7023,
+      "step": 127400
+    },
+    {
+      "epoch": 0.6362705685535219,
+      "grad_norm": 0.09249448776245117,
+      "learning_rate": 1.0939197476782898e-05,
+      "loss": 8.7212,
+      "step": 127410
+    },
+    {
+      "epoch": 0.6363205073784614,
+      "grad_norm": 0.09347539395093918,
+      "learning_rate": 1.0937695561841348e-05,
+      "loss": 8.6968,
+      "step": 127420
+    },
+    {
+      "epoch": 0.6363704462034009,
+      "grad_norm": 0.09082402288913727,
+      "learning_rate": 1.0936193646899797e-05,
+      "loss": 8.7038,
+      "step": 127430
+    },
+    {
+      "epoch": 0.6364203850283403,
+      "grad_norm": 0.09499524533748627,
+      "learning_rate": 1.0934691731958247e-05,
+      "loss": 8.7015,
+      "step": 127440
+    },
+    {
+      "epoch": 0.6364703238532797,
+      "grad_norm": 0.09662103652954102,
+      "learning_rate": 1.0933189817016697e-05,
+      "loss": 8.6867,
+      "step": 127450
+    },
+    {
+      "epoch": 0.6365202626782192,
+      "grad_norm": 0.10028364509344101,
+      "learning_rate": 1.0931687902075146e-05,
+      "loss": 8.697,
+      "step": 127460
+    },
+    {
+      "epoch": 0.6365702015031587,
+      "grad_norm": 0.09399396926164627,
+      "learning_rate": 1.0930185987133596e-05,
+      "loss": 8.7017,
+      "step": 127470
+    },
+    {
+      "epoch": 0.6366201403280981,
+      "grad_norm": 0.0959121435880661,
+      "learning_rate": 1.0928684072192044e-05,
+      "loss": 8.6941,
+      "step": 127480
+    },
+    {
+      "epoch": 0.6366700791530375,
+      "grad_norm": 0.09251075237989426,
+      "learning_rate": 1.0927182157250494e-05,
+      "loss": 8.6763,
+      "step": 127490
+    },
+    {
+      "epoch": 0.636720017977977,
+      "grad_norm": 0.09388341009616852,
+      "learning_rate": 1.0925680242308945e-05,
+      "loss": 8.698,
+      "step": 127500
+    },
+    {
+      "epoch": 0.6367699568029165,
+      "grad_norm": 0.09439099580049515,
+      "learning_rate": 1.0924178327367393e-05,
+      "loss": 8.721,
+      "step": 127510
+    },
+    {
+      "epoch": 0.6368198956278559,
+      "grad_norm": 0.0929148718714714,
+      "learning_rate": 1.0922676412425843e-05,
+      "loss": 8.7023,
+      "step": 127520
+    },
+    {
+      "epoch": 0.6368698344527953,
+      "grad_norm": 0.08788886666297913,
+      "learning_rate": 1.0921174497484292e-05,
+      "loss": 8.7087,
+      "step": 127530
+    },
+    {
+      "epoch": 0.6369197732777347,
+      "grad_norm": 0.09236535429954529,
+      "learning_rate": 1.0919672582542742e-05,
+      "loss": 8.7026,
+      "step": 127540
+    },
+    {
+      "epoch": 0.6369697121026743,
+      "grad_norm": 0.08757699280977249,
+      "learning_rate": 1.0918170667601192e-05,
+      "loss": 8.6932,
+      "step": 127550
+    },
+    {
+      "epoch": 0.6370196509276137,
+      "grad_norm": 0.09204018861055374,
+      "learning_rate": 1.0916668752659642e-05,
+      "loss": 8.7039,
+      "step": 127560
+    },
+    {
+      "epoch": 0.6370695897525531,
+      "grad_norm": 0.0933895856142044,
+      "learning_rate": 1.091516683771809e-05,
+      "loss": 8.6862,
+      "step": 127570
+    },
+    {
+      "epoch": 0.6371195285774925,
+      "grad_norm": 0.09441374242305756,
+      "learning_rate": 1.0913664922776541e-05,
+      "loss": 8.712,
+      "step": 127580
+    },
+    {
+      "epoch": 0.637169467402432,
+      "grad_norm": 0.08799993246793747,
+      "learning_rate": 1.091216300783499e-05,
+      "loss": 8.69,
+      "step": 127590
+    },
+    {
+      "epoch": 0.6372194062273715,
+      "grad_norm": 0.08958584070205688,
+      "learning_rate": 1.091066109289344e-05,
+      "loss": 8.6959,
+      "step": 127600
+    },
+    {
+      "epoch": 0.6372693450523109,
+      "grad_norm": 0.09037229418754578,
+      "learning_rate": 1.090915917795189e-05,
+      "loss": 8.7096,
+      "step": 127610
+    },
+    {
+      "epoch": 0.6373192838772503,
+      "grad_norm": 0.08842466026544571,
+      "learning_rate": 1.0907657263010338e-05,
+      "loss": 8.7115,
+      "step": 127620
+    },
+    {
+      "epoch": 0.6373692227021898,
+      "grad_norm": 0.09405070543289185,
+      "learning_rate": 1.0906155348068788e-05,
+      "loss": 8.7059,
+      "step": 127630
+    },
+    {
+      "epoch": 0.6374191615271293,
+      "grad_norm": 0.09778458625078201,
+      "learning_rate": 1.0904653433127237e-05,
+      "loss": 8.7081,
+      "step": 127640
+    },
+    {
+      "epoch": 0.6374691003520687,
+      "grad_norm": 0.09865093231201172,
+      "learning_rate": 1.0903151518185687e-05,
+      "loss": 8.7095,
+      "step": 127650
+    },
+    {
+      "epoch": 0.6375190391770081,
+      "grad_norm": 0.09222712367773056,
+      "learning_rate": 1.0901649603244137e-05,
+      "loss": 8.6945,
+      "step": 127660
+    },
+    {
+      "epoch": 0.6375689780019476,
+      "grad_norm": 0.09677907079458237,
+      "learning_rate": 1.0900147688302586e-05,
+      "loss": 8.6983,
+      "step": 127670
+    },
+    {
+      "epoch": 0.6376189168268871,
+      "grad_norm": 0.08906341344118118,
+      "learning_rate": 1.0898645773361036e-05,
+      "loss": 8.7033,
+      "step": 127680
+    },
+    {
+      "epoch": 0.6376688556518265,
+      "grad_norm": 0.09072484076023102,
+      "learning_rate": 1.0897143858419484e-05,
+      "loss": 8.7092,
+      "step": 127690
+    },
+    {
+      "epoch": 0.6377187944767659,
+      "grad_norm": 0.09344867616891861,
+      "learning_rate": 1.0895641943477935e-05,
+      "loss": 8.6877,
+      "step": 127700
+    },
+    {
+      "epoch": 0.6377687333017054,
+      "grad_norm": 0.093172587454319,
+      "learning_rate": 1.0894140028536385e-05,
+      "loss": 8.679,
+      "step": 127710
+    },
+    {
+      "epoch": 0.6378186721266449,
+      "grad_norm": 0.08979550749063492,
+      "learning_rate": 1.0892638113594833e-05,
+      "loss": 8.7013,
+      "step": 127720
+    },
+    {
+      "epoch": 0.6378686109515843,
+      "grad_norm": 0.09762118011713028,
+      "learning_rate": 1.0891136198653283e-05,
+      "loss": 8.7104,
+      "step": 127730
+    },
+    {
+      "epoch": 0.6379185497765237,
+      "grad_norm": 0.09290577471256256,
+      "learning_rate": 1.0889634283711734e-05,
+      "loss": 8.7004,
+      "step": 127740
+    },
+    {
+      "epoch": 0.6379684886014632,
+      "grad_norm": 0.0906912311911583,
+      "learning_rate": 1.0888132368770182e-05,
+      "loss": 8.7001,
+      "step": 127750
+    },
+    {
+      "epoch": 0.6380184274264027,
+      "grad_norm": 0.08934336155653,
+      "learning_rate": 1.0886630453828632e-05,
+      "loss": 8.6983,
+      "step": 127760
+    },
+    {
+      "epoch": 0.6380683662513421,
+      "grad_norm": 0.09342038631439209,
+      "learning_rate": 1.088512853888708e-05,
+      "loss": 8.6941,
+      "step": 127770
+    },
+    {
+      "epoch": 0.6381183050762815,
+      "grad_norm": 0.0915367603302002,
+      "learning_rate": 1.0883626623945531e-05,
+      "loss": 8.696,
+      "step": 127780
+    },
+    {
+      "epoch": 0.638168243901221,
+      "grad_norm": 0.08702074736356735,
+      "learning_rate": 1.0882124709003981e-05,
+      "loss": 8.6942,
+      "step": 127790
+    },
+    {
+      "epoch": 0.6382181827261605,
+      "grad_norm": 0.09180362522602081,
+      "learning_rate": 1.088062279406243e-05,
+      "loss": 8.7023,
+      "step": 127800
+    },
+    {
+      "epoch": 0.6382681215510999,
+      "grad_norm": 0.0904599204659462,
+      "learning_rate": 1.087912087912088e-05,
+      "loss": 8.7151,
+      "step": 127810
+    },
+    {
+      "epoch": 0.6383180603760393,
+      "grad_norm": 0.09203129261732101,
+      "learning_rate": 1.0877618964179328e-05,
+      "loss": 8.7078,
+      "step": 127820
+    },
+    {
+      "epoch": 0.6383679992009788,
+      "grad_norm": 0.08823869377374649,
+      "learning_rate": 1.0876117049237779e-05,
+      "loss": 8.7192,
+      "step": 127830
+    },
+    {
+      "epoch": 0.6384179380259183,
+      "grad_norm": 0.09404473751783371,
+      "learning_rate": 1.0874615134296229e-05,
+      "loss": 8.7175,
+      "step": 127840
+    },
+    {
+      "epoch": 0.6384678768508577,
+      "grad_norm": 0.09303843975067139,
+      "learning_rate": 1.0873113219354677e-05,
+      "loss": 8.7135,
+      "step": 127850
+    },
+    {
+      "epoch": 0.6385178156757971,
+      "grad_norm": 0.09125247597694397,
+      "learning_rate": 1.0871611304413127e-05,
+      "loss": 8.7052,
+      "step": 127860
+    },
+    {
+      "epoch": 0.6385677545007366,
+      "grad_norm": 0.0958453044295311,
+      "learning_rate": 1.0870109389471576e-05,
+      "loss": 8.6845,
+      "step": 127870
+    },
+    {
+      "epoch": 0.6386176933256761,
+      "grad_norm": 0.09031480550765991,
+      "learning_rate": 1.0868607474530026e-05,
+      "loss": 8.7015,
+      "step": 127880
+    },
+    {
+      "epoch": 0.6386676321506155,
+      "grad_norm": 0.10492537170648575,
+      "learning_rate": 1.0867105559588476e-05,
+      "loss": 8.7136,
+      "step": 127890
+    },
+    {
+      "epoch": 0.6387175709755549,
+      "grad_norm": 0.09470899403095245,
+      "learning_rate": 1.0865603644646926e-05,
+      "loss": 8.6868,
+      "step": 127900
+    },
+    {
+      "epoch": 0.6387675098004943,
+      "grad_norm": 0.09199722856283188,
+      "learning_rate": 1.0864101729705375e-05,
+      "loss": 8.6939,
+      "step": 127910
+    },
+    {
+      "epoch": 0.6388174486254339,
+      "grad_norm": 0.08992622792720795,
+      "learning_rate": 1.0862599814763823e-05,
+      "loss": 8.6928,
+      "step": 127920
+    },
+    {
+      "epoch": 0.6388673874503733,
+      "grad_norm": 0.09477774053812027,
+      "learning_rate": 1.0861097899822274e-05,
+      "loss": 8.6969,
+      "step": 127930
+    },
+    {
+      "epoch": 0.6389173262753127,
+      "grad_norm": 0.09002884477376938,
+      "learning_rate": 1.0859595984880724e-05,
+      "loss": 8.6955,
+      "step": 127940
+    },
+    {
+      "epoch": 0.6389672651002521,
+      "grad_norm": 0.09048038721084595,
+      "learning_rate": 1.0858094069939174e-05,
+      "loss": 8.7043,
+      "step": 127950
+    },
+    {
+      "epoch": 0.6390172039251917,
+      "grad_norm": 0.08667916804552078,
+      "learning_rate": 1.0856592154997622e-05,
+      "loss": 8.6913,
+      "step": 127960
+    },
+    {
+      "epoch": 0.6390671427501311,
+      "grad_norm": 0.09208723902702332,
+      "learning_rate": 1.085509024005607e-05,
+      "loss": 8.7021,
+      "step": 127970
+    },
+    {
+      "epoch": 0.6391170815750705,
+      "grad_norm": 0.09197891503572464,
+      "learning_rate": 1.0853588325114521e-05,
+      "loss": 8.7092,
+      "step": 127980
+    },
+    {
+      "epoch": 0.6391670204000099,
+      "grad_norm": 0.09264016896486282,
+      "learning_rate": 1.0852086410172971e-05,
+      "loss": 8.6802,
+      "step": 127990
+    },
+    {
+      "epoch": 0.6392169592249495,
+      "grad_norm": 0.0906987339258194,
+      "learning_rate": 1.0850584495231421e-05,
+      "loss": 8.6835,
+      "step": 128000
+    },
+    {
+      "epoch": 0.6392668980498889,
+      "grad_norm": 0.09663955122232437,
+      "learning_rate": 1.084908258028987e-05,
+      "loss": 8.6855,
+      "step": 128010
+    },
+    {
+      "epoch": 0.6393168368748283,
+      "grad_norm": 0.08810370415449142,
+      "learning_rate": 1.0847580665348318e-05,
+      "loss": 8.7045,
+      "step": 128020
+    },
+    {
+      "epoch": 0.6393667756997677,
+      "grad_norm": 0.09239162504673004,
+      "learning_rate": 1.0846078750406769e-05,
+      "loss": 8.7063,
+      "step": 128030
+    },
+    {
+      "epoch": 0.6394167145247073,
+      "grad_norm": 0.09910788387060165,
+      "learning_rate": 1.0844576835465219e-05,
+      "loss": 8.6895,
+      "step": 128040
+    },
+    {
+      "epoch": 0.6394666533496467,
+      "grad_norm": 0.09570180624723434,
+      "learning_rate": 1.0843074920523669e-05,
+      "loss": 8.7115,
+      "step": 128050
+    },
+    {
+      "epoch": 0.6395165921745861,
+      "grad_norm": 0.09080067276954651,
+      "learning_rate": 1.0841573005582119e-05,
+      "loss": 8.6992,
+      "step": 128060
+    },
+    {
+      "epoch": 0.6395665309995255,
+      "grad_norm": 0.09045545756816864,
+      "learning_rate": 1.0840071090640566e-05,
+      "loss": 8.7097,
+      "step": 128070
+    },
+    {
+      "epoch": 0.6396164698244651,
+      "grad_norm": 0.09077303856611252,
+      "learning_rate": 1.0838569175699016e-05,
+      "loss": 8.6892,
+      "step": 128080
+    },
+    {
+      "epoch": 0.6396664086494045,
+      "grad_norm": 0.09273795038461685,
+      "learning_rate": 1.0837067260757466e-05,
+      "loss": 8.6948,
+      "step": 128090
+    },
+    {
+      "epoch": 0.6397163474743439,
+      "grad_norm": 0.09408220648765564,
+      "learning_rate": 1.0835565345815916e-05,
+      "loss": 8.6957,
+      "step": 128100
+    },
+    {
+      "epoch": 0.6397662862992833,
+      "grad_norm": 0.09319932758808136,
+      "learning_rate": 1.0834063430874367e-05,
+      "loss": 8.6996,
+      "step": 128110
+    },
+    {
+      "epoch": 0.6398162251242229,
+      "grad_norm": 0.09193559736013412,
+      "learning_rate": 1.0832561515932813e-05,
+      "loss": 8.7085,
+      "step": 128120
+    },
+    {
+      "epoch": 0.6398661639491623,
+      "grad_norm": 0.09312306344509125,
+      "learning_rate": 1.0831059600991264e-05,
+      "loss": 8.6832,
+      "step": 128130
+    },
+    {
+      "epoch": 0.6399161027741017,
+      "grad_norm": 0.09437835961580276,
+      "learning_rate": 1.0829557686049714e-05,
+      "loss": 8.7107,
+      "step": 128140
+    },
+    {
+      "epoch": 0.6399660415990411,
+      "grad_norm": 0.09023882448673248,
+      "learning_rate": 1.0828055771108164e-05,
+      "loss": 8.6888,
+      "step": 128150
+    },
+    {
+      "epoch": 0.6400159804239807,
+      "grad_norm": 0.10062866657972336,
+      "learning_rate": 1.0826553856166614e-05,
+      "loss": 8.6884,
+      "step": 128160
+    },
+    {
+      "epoch": 0.6400659192489201,
+      "grad_norm": 0.09322546422481537,
+      "learning_rate": 1.0825051941225061e-05,
+      "loss": 8.7025,
+      "step": 128170
+    },
+    {
+      "epoch": 0.6401158580738595,
+      "grad_norm": 0.09134891629219055,
+      "learning_rate": 1.0823550026283511e-05,
+      "loss": 8.7113,
+      "step": 128180
+    },
+    {
+      "epoch": 0.6401657968987989,
+      "grad_norm": 0.0888613611459732,
+      "learning_rate": 1.0822048111341961e-05,
+      "loss": 8.6908,
+      "step": 128190
+    },
+    {
+      "epoch": 0.6402157357237385,
+      "grad_norm": 0.0894928053021431,
+      "learning_rate": 1.0820546196400411e-05,
+      "loss": 8.7003,
+      "step": 128200
+    },
+    {
+      "epoch": 0.6402656745486779,
+      "grad_norm": 0.09011103212833405,
+      "learning_rate": 1.0819044281458862e-05,
+      "loss": 8.6731,
+      "step": 128210
+    },
+    {
+      "epoch": 0.6403156133736173,
+      "grad_norm": 0.08623740077018738,
+      "learning_rate": 1.081754236651731e-05,
+      "loss": 8.699,
+      "step": 128220
+    },
+    {
+      "epoch": 0.6403655521985567,
+      "grad_norm": 0.09192583709955215,
+      "learning_rate": 1.0816040451575759e-05,
+      "loss": 8.7016,
+      "step": 128230
+    },
+    {
+      "epoch": 0.6404154910234963,
+      "grad_norm": 0.0926465094089508,
+      "learning_rate": 1.0814538536634209e-05,
+      "loss": 8.69,
+      "step": 128240
+    },
+    {
+      "epoch": 0.6404654298484357,
+      "grad_norm": 0.09219156205654144,
+      "learning_rate": 1.0813036621692659e-05,
+      "loss": 8.6993,
+      "step": 128250
+    },
+    {
+      "epoch": 0.6405153686733751,
+      "grad_norm": 0.09209728986024857,
+      "learning_rate": 1.0811534706751109e-05,
+      "loss": 8.6976,
+      "step": 128260
+    },
+    {
+      "epoch": 0.6405653074983145,
+      "grad_norm": 0.09860700368881226,
+      "learning_rate": 1.0810032791809558e-05,
+      "loss": 8.6936,
+      "step": 128270
+    },
+    {
+      "epoch": 0.6406152463232541,
+      "grad_norm": 0.09730758517980576,
+      "learning_rate": 1.0808530876868006e-05,
+      "loss": 8.7031,
+      "step": 128280
+    },
+    {
+      "epoch": 0.6406651851481935,
+      "grad_norm": 0.09535746276378632,
+      "learning_rate": 1.0807028961926456e-05,
+      "loss": 8.6945,
+      "step": 128290
+    },
+    {
+      "epoch": 0.6407151239731329,
+      "grad_norm": 0.08811869472265244,
+      "learning_rate": 1.0805527046984906e-05,
+      "loss": 8.7013,
+      "step": 128300
+    },
+    {
+      "epoch": 0.6407650627980723,
+      "grad_norm": 0.08866571635007858,
+      "learning_rate": 1.0804025132043357e-05,
+      "loss": 8.7039,
+      "step": 128310
+    },
+    {
+      "epoch": 0.6408150016230119,
+      "grad_norm": 0.0883893147110939,
+      "learning_rate": 1.0802523217101805e-05,
+      "loss": 8.7151,
+      "step": 128320
+    },
+    {
+      "epoch": 0.6408649404479513,
+      "grad_norm": 0.09385412186384201,
+      "learning_rate": 1.0801021302160254e-05,
+      "loss": 8.6897,
+      "step": 128330
+    },
+    {
+      "epoch": 0.6409148792728907,
+      "grad_norm": 0.09245074540376663,
+      "learning_rate": 1.0799519387218704e-05,
+      "loss": 8.7014,
+      "step": 128340
+    },
+    {
+      "epoch": 0.6409648180978301,
+      "grad_norm": 0.09483639895915985,
+      "learning_rate": 1.0798017472277154e-05,
+      "loss": 8.6969,
+      "step": 128350
+    },
+    {
+      "epoch": 0.6410147569227697,
+      "grad_norm": 0.09574125707149506,
+      "learning_rate": 1.0796515557335604e-05,
+      "loss": 8.6871,
+      "step": 128360
+    },
+    {
+      "epoch": 0.6410646957477091,
+      "grad_norm": 0.09570538997650146,
+      "learning_rate": 1.0795013642394053e-05,
+      "loss": 8.6784,
+      "step": 128370
+    },
+    {
+      "epoch": 0.6411146345726485,
+      "grad_norm": 0.08854909241199493,
+      "learning_rate": 1.0793511727452503e-05,
+      "loss": 8.7082,
+      "step": 128380
+    },
+    {
+      "epoch": 0.6411645733975879,
+      "grad_norm": 0.09440670907497406,
+      "learning_rate": 1.0792009812510951e-05,
+      "loss": 8.7027,
+      "step": 128390
+    },
+    {
+      "epoch": 0.6412145122225275,
+      "grad_norm": 0.08996603637933731,
+      "learning_rate": 1.0790507897569401e-05,
+      "loss": 8.7015,
+      "step": 128400
+    },
+    {
+      "epoch": 0.6412644510474669,
+      "grad_norm": 0.0914885401725769,
+      "learning_rate": 1.0789005982627852e-05,
+      "loss": 8.698,
+      "step": 128410
+    },
+    {
+      "epoch": 0.6413143898724063,
+      "grad_norm": 0.08596347272396088,
+      "learning_rate": 1.07875040676863e-05,
+      "loss": 8.7038,
+      "step": 128420
+    },
+    {
+      "epoch": 0.6413643286973457,
+      "grad_norm": 0.09728407114744186,
+      "learning_rate": 1.078600215274475e-05,
+      "loss": 8.6863,
+      "step": 128430
+    },
+    {
+      "epoch": 0.6414142675222853,
+      "grad_norm": 0.08998657763004303,
+      "learning_rate": 1.0784500237803199e-05,
+      "loss": 8.6925,
+      "step": 128440
+    },
+    {
+      "epoch": 0.6414642063472247,
+      "grad_norm": 0.09361448884010315,
+      "learning_rate": 1.0782998322861649e-05,
+      "loss": 8.7054,
+      "step": 128450
+    },
+    {
+      "epoch": 0.6415141451721641,
+      "grad_norm": 0.09714844077825546,
+      "learning_rate": 1.0781496407920099e-05,
+      "loss": 8.6882,
+      "step": 128460
+    },
+    {
+      "epoch": 0.6415640839971035,
+      "grad_norm": 0.09411020576953888,
+      "learning_rate": 1.0779994492978548e-05,
+      "loss": 8.6916,
+      "step": 128470
+    },
+    {
+      "epoch": 0.641614022822043,
+      "grad_norm": 0.09303989261388779,
+      "learning_rate": 1.0778492578036998e-05,
+      "loss": 8.6849,
+      "step": 128480
+    },
+    {
+      "epoch": 0.6416639616469825,
+      "grad_norm": 0.0919375792145729,
+      "learning_rate": 1.0776990663095446e-05,
+      "loss": 8.6868,
+      "step": 128490
+    },
+    {
+      "epoch": 0.6417139004719219,
+      "grad_norm": 0.08758861571550369,
+      "learning_rate": 1.0775488748153896e-05,
+      "loss": 8.715,
+      "step": 128500
+    },
+    {
+      "epoch": 0.6417638392968613,
+      "grad_norm": 0.09430398046970367,
+      "learning_rate": 1.0773986833212347e-05,
+      "loss": 8.6882,
+      "step": 128510
+    },
+    {
+      "epoch": 0.6418137781218008,
+      "grad_norm": 0.09385448694229126,
+      "learning_rate": 1.0772484918270795e-05,
+      "loss": 8.699,
+      "step": 128520
+    },
+    {
+      "epoch": 0.6418637169467403,
+      "grad_norm": 0.08902585506439209,
+      "learning_rate": 1.0770983003329245e-05,
+      "loss": 8.714,
+      "step": 128530
+    },
+    {
+      "epoch": 0.6419136557716797,
+      "grad_norm": 0.08999849855899811,
+      "learning_rate": 1.0769481088387695e-05,
+      "loss": 8.6958,
+      "step": 128540
+    },
+    {
+      "epoch": 0.6419635945966191,
+      "grad_norm": 0.09851477295160294,
+      "learning_rate": 1.0767979173446144e-05,
+      "loss": 8.7072,
+      "step": 128550
+    },
+    {
+      "epoch": 0.6420135334215585,
+      "grad_norm": 0.0901130884885788,
+      "learning_rate": 1.0766477258504594e-05,
+      "loss": 8.7057,
+      "step": 128560
+    },
+    {
+      "epoch": 0.6420634722464981,
+      "grad_norm": 0.0934019386768341,
+      "learning_rate": 1.0764975343563043e-05,
+      "loss": 8.6784,
+      "step": 128570
+    },
+    {
+      "epoch": 0.6421134110714375,
+      "grad_norm": 0.0917850062251091,
+      "learning_rate": 1.0763473428621493e-05,
+      "loss": 8.696,
+      "step": 128580
+    },
+    {
+      "epoch": 0.6421633498963769,
+      "grad_norm": 0.09072545170783997,
+      "learning_rate": 1.0761971513679943e-05,
+      "loss": 8.7021,
+      "step": 128590
+    },
+    {
+      "epoch": 0.6422132887213163,
+      "grad_norm": 0.08813299983739853,
+      "learning_rate": 1.0760469598738391e-05,
+      "loss": 8.6961,
+      "step": 128600
+    },
+    {
+      "epoch": 0.6422632275462559,
+      "grad_norm": 0.08745348453521729,
+      "learning_rate": 1.0758967683796842e-05,
+      "loss": 8.678,
+      "step": 128610
+    },
+    {
+      "epoch": 0.6423131663711953,
+      "grad_norm": 0.09347406774759293,
+      "learning_rate": 1.075746576885529e-05,
+      "loss": 8.7121,
+      "step": 128620
+    },
+    {
+      "epoch": 0.6423631051961347,
+      "grad_norm": 0.09024196863174438,
+      "learning_rate": 1.075596385391374e-05,
+      "loss": 8.6969,
+      "step": 128630
+    },
+    {
+      "epoch": 0.6424130440210741,
+      "grad_norm": 0.0957035943865776,
+      "learning_rate": 1.075446193897219e-05,
+      "loss": 8.6935,
+      "step": 128640
+    },
+    {
+      "epoch": 0.6424629828460137,
+      "grad_norm": 0.09494053572416306,
+      "learning_rate": 1.0752960024030639e-05,
+      "loss": 8.6902,
+      "step": 128650
+    },
+    {
+      "epoch": 0.6425129216709531,
+      "grad_norm": 0.09406720846891403,
+      "learning_rate": 1.0751458109089089e-05,
+      "loss": 8.6785,
+      "step": 128660
+    },
+    {
+      "epoch": 0.6425628604958925,
+      "grad_norm": 0.10135558992624283,
+      "learning_rate": 1.0749956194147538e-05,
+      "loss": 8.6853,
+      "step": 128670
+    },
+    {
+      "epoch": 0.6426127993208319,
+      "grad_norm": 0.09328481554985046,
+      "learning_rate": 1.0748454279205988e-05,
+      "loss": 8.6928,
+      "step": 128680
+    },
+    {
+      "epoch": 0.6426627381457715,
+      "grad_norm": 0.09335169196128845,
+      "learning_rate": 1.0746952364264438e-05,
+      "loss": 8.7002,
+      "step": 128690
+    },
+    {
+      "epoch": 0.6427126769707109,
+      "grad_norm": 0.09129825979471207,
+      "learning_rate": 1.0745450449322888e-05,
+      "loss": 8.7094,
+      "step": 128700
+    },
+    {
+      "epoch": 0.6427626157956503,
+      "grad_norm": 0.09808053821325302,
+      "learning_rate": 1.0743948534381337e-05,
+      "loss": 8.693,
+      "step": 128710
+    },
+    {
+      "epoch": 0.6428125546205897,
+      "grad_norm": 0.0911967009305954,
+      "learning_rate": 1.0742446619439785e-05,
+      "loss": 8.7,
+      "step": 128720
+    },
+    {
+      "epoch": 0.6428624934455293,
+      "grad_norm": 0.09128003567457199,
+      "learning_rate": 1.0740944704498235e-05,
+      "loss": 8.6906,
+      "step": 128730
+    },
+    {
+      "epoch": 0.6429124322704687,
+      "grad_norm": 0.09504960477352142,
+      "learning_rate": 1.0739442789556685e-05,
+      "loss": 8.7072,
+      "step": 128740
+    },
+    {
+      "epoch": 0.6429623710954081,
+      "grad_norm": 0.09051109105348587,
+      "learning_rate": 1.0737940874615136e-05,
+      "loss": 8.692,
+      "step": 128750
+    },
+    {
+      "epoch": 0.6430123099203475,
+      "grad_norm": 0.09383983165025711,
+      "learning_rate": 1.0736438959673584e-05,
+      "loss": 8.6927,
+      "step": 128760
+    },
+    {
+      "epoch": 0.6430622487452871,
+      "grad_norm": 0.09124093502759933,
+      "learning_rate": 1.0734937044732033e-05,
+      "loss": 8.6917,
+      "step": 128770
+    },
+    {
+      "epoch": 0.6431121875702265,
+      "grad_norm": 0.09348580986261368,
+      "learning_rate": 1.0733435129790483e-05,
+      "loss": 8.6897,
+      "step": 128780
+    },
+    {
+      "epoch": 0.6431621263951659,
+      "grad_norm": 0.08800901472568512,
+      "learning_rate": 1.0731933214848933e-05,
+      "loss": 8.692,
+      "step": 128790
+    },
+    {
+      "epoch": 0.6432120652201053,
+      "grad_norm": 0.08735651522874832,
+      "learning_rate": 1.0730431299907383e-05,
+      "loss": 8.697,
+      "step": 128800
+    },
+    {
+      "epoch": 0.6432620040450449,
+      "grad_norm": 0.09434789419174194,
+      "learning_rate": 1.0728929384965832e-05,
+      "loss": 8.6856,
+      "step": 128810
+    },
+    {
+      "epoch": 0.6433119428699843,
+      "grad_norm": 0.09529183804988861,
+      "learning_rate": 1.072742747002428e-05,
+      "loss": 8.7082,
+      "step": 128820
+    },
+    {
+      "epoch": 0.6433618816949237,
+      "grad_norm": 0.08833926171064377,
+      "learning_rate": 1.072592555508273e-05,
+      "loss": 8.6816,
+      "step": 128830
+    },
+    {
+      "epoch": 0.6434118205198631,
+      "grad_norm": 0.09038601070642471,
+      "learning_rate": 1.072442364014118e-05,
+      "loss": 8.6931,
+      "step": 128840
+    },
+    {
+      "epoch": 0.6434617593448027,
+      "grad_norm": 0.09003665298223495,
+      "learning_rate": 1.072292172519963e-05,
+      "loss": 8.685,
+      "step": 128850
+    },
+    {
+      "epoch": 0.6435116981697421,
+      "grad_norm": 0.08987587690353394,
+      "learning_rate": 1.0721419810258079e-05,
+      "loss": 8.707,
+      "step": 128860
+    },
+    {
+      "epoch": 0.6435616369946815,
+      "grad_norm": 0.09192398190498352,
+      "learning_rate": 1.0719917895316528e-05,
+      "loss": 8.6899,
+      "step": 128870
+    },
+    {
+      "epoch": 0.6436115758196209,
+      "grad_norm": 0.09186169505119324,
+      "learning_rate": 1.0718415980374978e-05,
+      "loss": 8.682,
+      "step": 128880
+    },
+    {
+      "epoch": 0.6436615146445605,
+      "grad_norm": 0.0946180447936058,
+      "learning_rate": 1.0716914065433428e-05,
+      "loss": 8.6664,
+      "step": 128890
+    },
+    {
+      "epoch": 0.6437114534694999,
+      "grad_norm": 0.08595947921276093,
+      "learning_rate": 1.0715412150491878e-05,
+      "loss": 8.6969,
+      "step": 128900
+    },
+    {
+      "epoch": 0.6437613922944393,
+      "grad_norm": 0.0891309604048729,
+      "learning_rate": 1.0713910235550328e-05,
+      "loss": 8.7002,
+      "step": 128910
+    },
+    {
+      "epoch": 0.6438113311193787,
+      "grad_norm": 0.08861633390188217,
+      "learning_rate": 1.0712408320608775e-05,
+      "loss": 8.7062,
+      "step": 128920
+    },
+    {
+      "epoch": 0.6438612699443182,
+      "grad_norm": 0.0909549817442894,
+      "learning_rate": 1.0710906405667225e-05,
+      "loss": 8.7044,
+      "step": 128930
+    },
+    {
+      "epoch": 0.6439112087692577,
+      "grad_norm": 0.09526665508747101,
+      "learning_rate": 1.0709404490725675e-05,
+      "loss": 8.6959,
+      "step": 128940
+    },
+    {
+      "epoch": 0.6439611475941971,
+      "grad_norm": 0.0844038724899292,
+      "learning_rate": 1.0707902575784126e-05,
+      "loss": 8.6776,
+      "step": 128950
+    },
+    {
+      "epoch": 0.6440110864191365,
+      "grad_norm": 0.09909462928771973,
+      "learning_rate": 1.0706400660842576e-05,
+      "loss": 8.6958,
+      "step": 128960
+    },
+    {
+      "epoch": 0.644061025244076,
+      "grad_norm": 0.09768630564212799,
+      "learning_rate": 1.0704898745901023e-05,
+      "loss": 8.6804,
+      "step": 128970
+    },
+    {
+      "epoch": 0.6441109640690155,
+      "grad_norm": 0.09060562402009964,
+      "learning_rate": 1.0703396830959473e-05,
+      "loss": 8.6874,
+      "step": 128980
+    },
+    {
+      "epoch": 0.6441609028939549,
+      "grad_norm": 0.09415677189826965,
+      "learning_rate": 1.0701894916017923e-05,
+      "loss": 8.6893,
+      "step": 128990
+    },
+    {
+      "epoch": 0.6442108417188943,
+      "grad_norm": 0.08882970362901688,
+      "learning_rate": 1.0700393001076373e-05,
+      "loss": 8.685,
+      "step": 129000
+    },
+    {
+      "epoch": 0.6442607805438338,
+      "grad_norm": 0.09341847151517868,
+      "learning_rate": 1.0698891086134823e-05,
+      "loss": 8.6986,
+      "step": 129010
+    },
+    {
+      "epoch": 0.6443107193687733,
+      "grad_norm": 0.08802028000354767,
+      "learning_rate": 1.069738917119327e-05,
+      "loss": 8.6974,
+      "step": 129020
+    },
+    {
+      "epoch": 0.6443606581937127,
+      "grad_norm": 0.09880810230970383,
+      "learning_rate": 1.069588725625172e-05,
+      "loss": 8.7002,
+      "step": 129030
+    },
+    {
+      "epoch": 0.6444105970186521,
+      "grad_norm": 0.09482931345701218,
+      "learning_rate": 1.069438534131017e-05,
+      "loss": 8.7117,
+      "step": 129040
+    },
+    {
+      "epoch": 0.6444605358435916,
+      "grad_norm": 0.09154761582612991,
+      "learning_rate": 1.069288342636862e-05,
+      "loss": 8.7054,
+      "step": 129050
+    },
+    {
+      "epoch": 0.6445104746685311,
+      "grad_norm": 0.08899923413991928,
+      "learning_rate": 1.069138151142707e-05,
+      "loss": 8.7215,
+      "step": 129060
+    },
+    {
+      "epoch": 0.6445604134934705,
+      "grad_norm": 0.09512390941381454,
+      "learning_rate": 1.068987959648552e-05,
+      "loss": 8.6868,
+      "step": 129070
+    },
+    {
+      "epoch": 0.6446103523184099,
+      "grad_norm": 0.09039557725191116,
+      "learning_rate": 1.0688377681543968e-05,
+      "loss": 8.6942,
+      "step": 129080
+    },
+    {
+      "epoch": 0.6446602911433494,
+      "grad_norm": 0.09328074753284454,
+      "learning_rate": 1.0686875766602418e-05,
+      "loss": 8.6928,
+      "step": 129090
+    },
+    {
+      "epoch": 0.6447102299682889,
+      "grad_norm": 0.0936678946018219,
+      "learning_rate": 1.0685373851660868e-05,
+      "loss": 8.6925,
+      "step": 129100
+    },
+    {
+      "epoch": 0.6447601687932283,
+      "grad_norm": 0.09343737363815308,
+      "learning_rate": 1.0683871936719318e-05,
+      "loss": 8.698,
+      "step": 129110
+    },
+    {
+      "epoch": 0.6448101076181677,
+      "grad_norm": 0.08791700750589371,
+      "learning_rate": 1.0682370021777767e-05,
+      "loss": 8.6843,
+      "step": 129120
+    },
+    {
+      "epoch": 0.6448600464431072,
+      "grad_norm": 0.09414328634738922,
+      "learning_rate": 1.0680868106836215e-05,
+      "loss": 8.7037,
+      "step": 129130
+    },
+    {
+      "epoch": 0.6449099852680467,
+      "grad_norm": 0.0889165922999382,
+      "learning_rate": 1.0679366191894665e-05,
+      "loss": 8.701,
+      "step": 129140
+    },
+    {
+      "epoch": 0.6449599240929861,
+      "grad_norm": 0.09565069526433945,
+      "learning_rate": 1.0677864276953116e-05,
+      "loss": 8.6856,
+      "step": 129150
+    },
+    {
+      "epoch": 0.6450098629179255,
+      "grad_norm": 0.09142804890871048,
+      "learning_rate": 1.0676362362011566e-05,
+      "loss": 8.694,
+      "step": 129160
+    },
+    {
+      "epoch": 0.645059801742865,
+      "grad_norm": 0.09595470130443573,
+      "learning_rate": 1.0674860447070014e-05,
+      "loss": 8.7063,
+      "step": 129170
+    },
+    {
+      "epoch": 0.6451097405678045,
+      "grad_norm": 0.09262866526842117,
+      "learning_rate": 1.0673358532128463e-05,
+      "loss": 8.687,
+      "step": 129180
+    },
+    {
+      "epoch": 0.6451596793927439,
+      "grad_norm": 0.09345003217458725,
+      "learning_rate": 1.0671856617186913e-05,
+      "loss": 8.6825,
+      "step": 129190
+    },
+    {
+      "epoch": 0.6452096182176833,
+      "grad_norm": 0.09604539722204208,
+      "learning_rate": 1.0670354702245363e-05,
+      "loss": 8.685,
+      "step": 129200
+    },
+    {
+      "epoch": 0.6452595570426228,
+      "grad_norm": 0.09135550260543823,
+      "learning_rate": 1.0668852787303813e-05,
+      "loss": 8.6817,
+      "step": 129210
+    },
+    {
+      "epoch": 0.6453094958675623,
+      "grad_norm": 0.09230076521635056,
+      "learning_rate": 1.0667350872362262e-05,
+      "loss": 8.7034,
+      "step": 129220
+    },
+    {
+      "epoch": 0.6453594346925017,
+      "grad_norm": 0.08917170763015747,
+      "learning_rate": 1.0665848957420712e-05,
+      "loss": 8.6952,
+      "step": 129230
+    },
+    {
+      "epoch": 0.6454093735174411,
+      "grad_norm": 0.09029002487659454,
+      "learning_rate": 1.066434704247916e-05,
+      "loss": 8.7049,
+      "step": 129240
+    },
+    {
+      "epoch": 0.6454593123423806,
+      "grad_norm": 0.09072355180978775,
+      "learning_rate": 1.066284512753761e-05,
+      "loss": 8.6823,
+      "step": 129250
+    },
+    {
+      "epoch": 0.64550925116732,
+      "grad_norm": 0.09148232638835907,
+      "learning_rate": 1.066134321259606e-05,
+      "loss": 8.6949,
+      "step": 129260
+    },
+    {
+      "epoch": 0.6455591899922595,
+      "grad_norm": 0.09294047951698303,
+      "learning_rate": 1.065984129765451e-05,
+      "loss": 8.6964,
+      "step": 129270
+    },
+    {
+      "epoch": 0.6456091288171989,
+      "grad_norm": 0.09537001699209213,
+      "learning_rate": 1.065833938271296e-05,
+      "loss": 8.6854,
+      "step": 129280
+    },
+    {
+      "epoch": 0.6456590676421384,
+      "grad_norm": 0.08766493201255798,
+      "learning_rate": 1.0656837467771408e-05,
+      "loss": 8.6707,
+      "step": 129290
+    },
+    {
+      "epoch": 0.6457090064670779,
+      "grad_norm": 0.09294747561216354,
+      "learning_rate": 1.0655335552829858e-05,
+      "loss": 8.691,
+      "step": 129300
+    },
+    {
+      "epoch": 0.6457589452920173,
+      "grad_norm": 0.09923656284809113,
+      "learning_rate": 1.0653833637888308e-05,
+      "loss": 8.6914,
+      "step": 129310
+    },
+    {
+      "epoch": 0.6458088841169567,
+      "grad_norm": 0.09559011459350586,
+      "learning_rate": 1.0652331722946757e-05,
+      "loss": 8.6932,
+      "step": 129320
+    },
+    {
+      "epoch": 0.6458588229418962,
+      "grad_norm": 0.09223086386919022,
+      "learning_rate": 1.0650829808005207e-05,
+      "loss": 8.6739,
+      "step": 129330
+    },
+    {
+      "epoch": 0.6459087617668356,
+      "grad_norm": 0.09738924354314804,
+      "learning_rate": 1.0649327893063655e-05,
+      "loss": 8.6669,
+      "step": 129340
+    },
+    {
+      "epoch": 0.6459587005917751,
+      "grad_norm": 0.09521890431642532,
+      "learning_rate": 1.0647825978122106e-05,
+      "loss": 8.7012,
+      "step": 129350
+    },
+    {
+      "epoch": 0.6460086394167145,
+      "grad_norm": 0.08749169111251831,
+      "learning_rate": 1.0646324063180556e-05,
+      "loss": 8.7065,
+      "step": 129360
+    },
+    {
+      "epoch": 0.646058578241654,
+      "grad_norm": 0.09419877082109451,
+      "learning_rate": 1.0644822148239004e-05,
+      "loss": 8.6892,
+      "step": 129370
+    },
+    {
+      "epoch": 0.6461085170665934,
+      "grad_norm": 0.1010696068406105,
+      "learning_rate": 1.0643320233297455e-05,
+      "loss": 8.6767,
+      "step": 129380
+    },
+    {
+      "epoch": 0.6461584558915329,
+      "grad_norm": 0.09025359898805618,
+      "learning_rate": 1.0641818318355905e-05,
+      "loss": 8.6869,
+      "step": 129390
+    },
+    {
+      "epoch": 0.6462083947164723,
+      "grad_norm": 0.10446906089782715,
+      "learning_rate": 1.0640316403414353e-05,
+      "loss": 8.6966,
+      "step": 129400
+    },
+    {
+      "epoch": 0.6462583335414118,
+      "grad_norm": 0.09092262387275696,
+      "learning_rate": 1.0638814488472803e-05,
+      "loss": 8.6835,
+      "step": 129410
+    },
+    {
+      "epoch": 0.6463082723663512,
+      "grad_norm": 0.08932240307331085,
+      "learning_rate": 1.0637312573531252e-05,
+      "loss": 8.6936,
+      "step": 129420
+    },
+    {
+      "epoch": 0.6463582111912907,
+      "grad_norm": 0.09258059412240982,
+      "learning_rate": 1.0635810658589702e-05,
+      "loss": 8.6914,
+      "step": 129430
+    },
+    {
+      "epoch": 0.6464081500162301,
+      "grad_norm": 0.09351666271686554,
+      "learning_rate": 1.0634308743648152e-05,
+      "loss": 8.7064,
+      "step": 129440
+    },
+    {
+      "epoch": 0.6464580888411696,
+      "grad_norm": 0.08932211250066757,
+      "learning_rate": 1.06328068287066e-05,
+      "loss": 8.7065,
+      "step": 129450
+    },
+    {
+      "epoch": 0.646508027666109,
+      "grad_norm": 0.0951404944062233,
+      "learning_rate": 1.0631304913765051e-05,
+      "loss": 8.6996,
+      "step": 129460
+    },
+    {
+      "epoch": 0.6465579664910485,
+      "grad_norm": 0.09667675197124481,
+      "learning_rate": 1.06298029988235e-05,
+      "loss": 8.6869,
+      "step": 129470
+    },
+    {
+      "epoch": 0.6466079053159879,
+      "grad_norm": 0.08983483165502548,
+      "learning_rate": 1.062830108388195e-05,
+      "loss": 8.6919,
+      "step": 129480
+    },
+    {
+      "epoch": 0.6466578441409274,
+      "grad_norm": 0.09399787336587906,
+      "learning_rate": 1.06267991689404e-05,
+      "loss": 8.6946,
+      "step": 129490
+    },
+    {
+      "epoch": 0.6467077829658668,
+      "grad_norm": 0.09137116372585297,
+      "learning_rate": 1.0625297253998848e-05,
+      "loss": 8.7024,
+      "step": 129500
+    },
+    {
+      "epoch": 0.6467577217908063,
+      "grad_norm": 0.09180816262960434,
+      "learning_rate": 1.0623795339057298e-05,
+      "loss": 8.6889,
+      "step": 129510
+    },
+    {
+      "epoch": 0.6468076606157457,
+      "grad_norm": 0.08744052797555923,
+      "learning_rate": 1.0622293424115747e-05,
+      "loss": 8.6927,
+      "step": 129520
+    },
+    {
+      "epoch": 0.6468575994406852,
+      "grad_norm": 0.09338006377220154,
+      "learning_rate": 1.0620791509174197e-05,
+      "loss": 8.6732,
+      "step": 129530
+    },
+    {
+      "epoch": 0.6469075382656246,
+      "grad_norm": 0.0890432596206665,
+      "learning_rate": 1.0619289594232647e-05,
+      "loss": 8.682,
+      "step": 129540
+    },
+    {
+      "epoch": 0.6469574770905641,
+      "grad_norm": 0.08677448332309723,
+      "learning_rate": 1.0617787679291097e-05,
+      "loss": 8.6879,
+      "step": 129550
+    },
+    {
+      "epoch": 0.6470074159155035,
+      "grad_norm": 0.08942339569330215,
+      "learning_rate": 1.0616285764349546e-05,
+      "loss": 8.6885,
+      "step": 129560
+    },
+    {
+      "epoch": 0.6470573547404429,
+      "grad_norm": 0.09570731967687607,
+      "learning_rate": 1.0614783849407994e-05,
+      "loss": 8.7066,
+      "step": 129570
+    },
+    {
+      "epoch": 0.6471072935653824,
+      "grad_norm": 0.09744849801063538,
+      "learning_rate": 1.0613281934466445e-05,
+      "loss": 8.6889,
+      "step": 129580
+    },
+    {
+      "epoch": 0.6471572323903219,
+      "grad_norm": 0.09440560638904572,
+      "learning_rate": 1.0611780019524895e-05,
+      "loss": 8.6876,
+      "step": 129590
+    },
+    {
+      "epoch": 0.6472071712152613,
+      "grad_norm": 0.09413042664527893,
+      "learning_rate": 1.0610278104583345e-05,
+      "loss": 8.7024,
+      "step": 129600
+    },
+    {
+      "epoch": 0.6472571100402007,
+      "grad_norm": 0.09831690788269043,
+      "learning_rate": 1.0608776189641793e-05,
+      "loss": 8.702,
+      "step": 129610
+    },
+    {
+      "epoch": 0.6473070488651402,
+      "grad_norm": 0.0940280631184578,
+      "learning_rate": 1.0607274274700242e-05,
+      "loss": 8.6957,
+      "step": 129620
+    },
+    {
+      "epoch": 0.6473569876900797,
+      "grad_norm": 0.09738873690366745,
+      "learning_rate": 1.0605772359758692e-05,
+      "loss": 8.6943,
+      "step": 129630
+    },
+    {
+      "epoch": 0.6474069265150191,
+      "grad_norm": 0.09439673274755478,
+      "learning_rate": 1.0604270444817142e-05,
+      "loss": 8.6919,
+      "step": 129640
+    },
+    {
+      "epoch": 0.6474568653399585,
+      "grad_norm": 0.08990797400474548,
+      "learning_rate": 1.0602768529875592e-05,
+      "loss": 8.6936,
+      "step": 129650
+    },
+    {
+      "epoch": 0.647506804164898,
+      "grad_norm": 0.09267980605363846,
+      "learning_rate": 1.0601266614934041e-05,
+      "loss": 8.6866,
+      "step": 129660
+    },
+    {
+      "epoch": 0.6475567429898375,
+      "grad_norm": 0.09749335050582886,
+      "learning_rate": 1.0599764699992491e-05,
+      "loss": 8.6843,
+      "step": 129670
+    },
+    {
+      "epoch": 0.6476066818147769,
+      "grad_norm": 0.09197567403316498,
+      "learning_rate": 1.059826278505094e-05,
+      "loss": 8.6853,
+      "step": 129680
+    },
+    {
+      "epoch": 0.6476566206397163,
+      "grad_norm": 0.08990548551082611,
+      "learning_rate": 1.059676087010939e-05,
+      "loss": 8.6833,
+      "step": 129690
+    },
+    {
+      "epoch": 0.6477065594646558,
+      "grad_norm": 0.09163473546504974,
+      "learning_rate": 1.059525895516784e-05,
+      "loss": 8.6803,
+      "step": 129700
+    },
+    {
+      "epoch": 0.6477564982895953,
+      "grad_norm": 0.08452335745096207,
+      "learning_rate": 1.059375704022629e-05,
+      "loss": 8.6832,
+      "step": 129710
+    },
+    {
+      "epoch": 0.6478064371145347,
+      "grad_norm": 0.09139879792928696,
+      "learning_rate": 1.0592255125284739e-05,
+      "loss": 8.6855,
+      "step": 129720
+    },
+    {
+      "epoch": 0.6478563759394741,
+      "grad_norm": 0.09565936774015427,
+      "learning_rate": 1.0590753210343187e-05,
+      "loss": 8.6953,
+      "step": 129730
+    },
+    {
+      "epoch": 0.6479063147644136,
+      "grad_norm": 0.09335167706012726,
+      "learning_rate": 1.0589251295401637e-05,
+      "loss": 8.6954,
+      "step": 129740
+    },
+    {
+      "epoch": 0.647956253589353,
+      "grad_norm": 0.09695900231599808,
+      "learning_rate": 1.0587749380460087e-05,
+      "loss": 8.6963,
+      "step": 129750
+    },
+    {
+      "epoch": 0.6480061924142925,
+      "grad_norm": 0.09421364217996597,
+      "learning_rate": 1.0586247465518538e-05,
+      "loss": 8.7094,
+      "step": 129760
+    },
+    {
+      "epoch": 0.6480561312392319,
+      "grad_norm": 0.09024903178215027,
+      "learning_rate": 1.0584745550576986e-05,
+      "loss": 8.6867,
+      "step": 129770
+    },
+    {
+      "epoch": 0.6481060700641714,
+      "grad_norm": 0.09890251606702805,
+      "learning_rate": 1.0583243635635435e-05,
+      "loss": 8.6979,
+      "step": 129780
+    },
+    {
+      "epoch": 0.6481560088891108,
+      "grad_norm": 0.09448616206645966,
+      "learning_rate": 1.0581741720693885e-05,
+      "loss": 8.6953,
+      "step": 129790
+    },
+    {
+      "epoch": 0.6482059477140503,
+      "grad_norm": 0.08602569997310638,
+      "learning_rate": 1.0580239805752335e-05,
+      "loss": 8.6916,
+      "step": 129800
+    },
+    {
+      "epoch": 0.6482558865389897,
+      "grad_norm": 0.08742332458496094,
+      "learning_rate": 1.0578737890810785e-05,
+      "loss": 8.6816,
+      "step": 129810
+    },
+    {
+      "epoch": 0.6483058253639292,
+      "grad_norm": 0.09227047860622406,
+      "learning_rate": 1.0577235975869234e-05,
+      "loss": 8.7025,
+      "step": 129820
+    },
+    {
+      "epoch": 0.6483557641888686,
+      "grad_norm": 0.09513872116804123,
+      "learning_rate": 1.0575734060927682e-05,
+      "loss": 8.6642,
+      "step": 129830
+    },
+    {
+      "epoch": 0.6484057030138081,
+      "grad_norm": 0.08885648846626282,
+      "learning_rate": 1.0574232145986132e-05,
+      "loss": 8.6826,
+      "step": 129840
+    },
+    {
+      "epoch": 0.6484556418387475,
+      "grad_norm": 0.09655725210905075,
+      "learning_rate": 1.0572730231044582e-05,
+      "loss": 8.6796,
+      "step": 129850
+    },
+    {
+      "epoch": 0.648505580663687,
+      "grad_norm": 0.0928000807762146,
+      "learning_rate": 1.0571228316103033e-05,
+      "loss": 8.6778,
+      "step": 129860
+    },
+    {
+      "epoch": 0.6485555194886264,
+      "grad_norm": 0.09117357432842255,
+      "learning_rate": 1.0569726401161483e-05,
+      "loss": 8.6699,
+      "step": 129870
+    },
+    {
+      "epoch": 0.6486054583135659,
+      "grad_norm": 0.09002866595983505,
+      "learning_rate": 1.056822448621993e-05,
+      "loss": 8.686,
+      "step": 129880
+    },
+    {
+      "epoch": 0.6486553971385053,
+      "grad_norm": 0.09259200841188431,
+      "learning_rate": 1.056672257127838e-05,
+      "loss": 8.6844,
+      "step": 129890
+    },
+    {
+      "epoch": 0.6487053359634448,
+      "grad_norm": 0.0892748162150383,
+      "learning_rate": 1.056522065633683e-05,
+      "loss": 8.6899,
+      "step": 129900
+    },
+    {
+      "epoch": 0.6487552747883842,
+      "grad_norm": 0.0917670950293541,
+      "learning_rate": 1.056371874139528e-05,
+      "loss": 8.6836,
+      "step": 129910
+    },
+    {
+      "epoch": 0.6488052136133237,
+      "grad_norm": 0.09188613295555115,
+      "learning_rate": 1.056221682645373e-05,
+      "loss": 8.7191,
+      "step": 129920
+    },
+    {
+      "epoch": 0.6488551524382631,
+      "grad_norm": 0.09064072370529175,
+      "learning_rate": 1.0560714911512177e-05,
+      "loss": 8.6913,
+      "step": 129930
+    },
+    {
+      "epoch": 0.6489050912632026,
+      "grad_norm": 0.09041624516248703,
+      "learning_rate": 1.0559212996570627e-05,
+      "loss": 8.6797,
+      "step": 129940
+    },
+    {
+      "epoch": 0.648955030088142,
+      "grad_norm": 0.09263037890195847,
+      "learning_rate": 1.0557711081629077e-05,
+      "loss": 8.6839,
+      "step": 129950
+    },
+    {
+      "epoch": 0.6490049689130815,
+      "grad_norm": 0.0885956883430481,
+      "learning_rate": 1.0556209166687528e-05,
+      "loss": 8.6841,
+      "step": 129960
+    },
+    {
+      "epoch": 0.6490549077380209,
+      "grad_norm": 0.08921640366315842,
+      "learning_rate": 1.0554707251745978e-05,
+      "loss": 8.6874,
+      "step": 129970
+    },
+    {
+      "epoch": 0.6491048465629604,
+      "grad_norm": 0.09531452506780624,
+      "learning_rate": 1.0553205336804425e-05,
+      "loss": 8.6952,
+      "step": 129980
+    },
+    {
+      "epoch": 0.6491547853878998,
+      "grad_norm": 0.09411237388849258,
+      "learning_rate": 1.0551703421862875e-05,
+      "loss": 8.6968,
+      "step": 129990
+    },
+    {
+      "epoch": 0.6492047242128393,
+      "grad_norm": 0.09411599487066269,
+      "learning_rate": 1.0550201506921325e-05,
+      "loss": 8.6859,
+      "step": 130000
+    },
+    {
+      "epoch": 0.6492546630377787,
+      "grad_norm": 0.09481288492679596,
+      "learning_rate": 1.0548699591979775e-05,
+      "loss": 8.695,
+      "step": 130010
+    },
+    {
+      "epoch": 0.6493046018627182,
+      "grad_norm": 0.09503282606601715,
+      "learning_rate": 1.0547197677038225e-05,
+      "loss": 8.6801,
+      "step": 130020
+    },
+    {
+      "epoch": 0.6493545406876576,
+      "grad_norm": 0.09125657379627228,
+      "learning_rate": 1.0545695762096674e-05,
+      "loss": 8.6882,
+      "step": 130030
+    },
+    {
+      "epoch": 0.649404479512597,
+      "grad_norm": 0.09231728315353394,
+      "learning_rate": 1.0544193847155122e-05,
+      "loss": 8.6794,
+      "step": 130040
+    },
+    {
+      "epoch": 0.6494544183375365,
+      "grad_norm": 0.09627514332532883,
+      "learning_rate": 1.0542691932213572e-05,
+      "loss": 8.6888,
+      "step": 130050
+    },
+    {
+      "epoch": 0.649504357162476,
+      "grad_norm": 0.0894688069820404,
+      "learning_rate": 1.0541190017272023e-05,
+      "loss": 8.6832,
+      "step": 130060
+    },
+    {
+      "epoch": 0.6495542959874154,
+      "grad_norm": 0.09295932203531265,
+      "learning_rate": 1.0539688102330473e-05,
+      "loss": 8.6835,
+      "step": 130070
+    },
+    {
+      "epoch": 0.6496042348123549,
+      "grad_norm": 0.09354085475206375,
+      "learning_rate": 1.0538186187388921e-05,
+      "loss": 8.6896,
+      "step": 130080
+    },
+    {
+      "epoch": 0.6496541736372943,
+      "grad_norm": 0.09849437326192856,
+      "learning_rate": 1.053668427244737e-05,
+      "loss": 8.67,
+      "step": 130090
+    },
+    {
+      "epoch": 0.6497041124622338,
+      "grad_norm": 0.09284999966621399,
+      "learning_rate": 1.053518235750582e-05,
+      "loss": 8.6605,
+      "step": 130100
+    },
+    {
+      "epoch": 0.6497540512871732,
+      "grad_norm": 0.0957464724779129,
+      "learning_rate": 1.053368044256427e-05,
+      "loss": 8.6809,
+      "step": 130110
+    },
+    {
+      "epoch": 0.6498039901121127,
+      "grad_norm": 0.09025140851736069,
+      "learning_rate": 1.053217852762272e-05,
+      "loss": 8.6906,
+      "step": 130120
+    },
+    {
+      "epoch": 0.6498539289370521,
+      "grad_norm": 0.0904148668050766,
+      "learning_rate": 1.0530676612681169e-05,
+      "loss": 8.6876,
+      "step": 130130
+    },
+    {
+      "epoch": 0.6499038677619916,
+      "grad_norm": 0.08794094622135162,
+      "learning_rate": 1.0529174697739617e-05,
+      "loss": 8.6838,
+      "step": 130140
+    },
+    {
+      "epoch": 0.649953806586931,
+      "grad_norm": 0.09247595816850662,
+      "learning_rate": 1.0527672782798067e-05,
+      "loss": 8.68,
+      "step": 130150
+    },
+    {
+      "epoch": 0.6500037454118704,
+      "grad_norm": 0.09172753244638443,
+      "learning_rate": 1.0526170867856518e-05,
+      "loss": 8.6877,
+      "step": 130160
+    },
+    {
+      "epoch": 0.6500536842368099,
+      "grad_norm": 0.09523764252662659,
+      "learning_rate": 1.0524668952914968e-05,
+      "loss": 8.6913,
+      "step": 130170
+    },
+    {
+      "epoch": 0.6501036230617494,
+      "grad_norm": 0.08924904465675354,
+      "learning_rate": 1.0523167037973416e-05,
+      "loss": 8.681,
+      "step": 130180
+    },
+    {
+      "epoch": 0.6501535618866888,
+      "grad_norm": 0.08906625956296921,
+      "learning_rate": 1.0521665123031866e-05,
+      "loss": 8.675,
+      "step": 130190
+    },
+    {
+      "epoch": 0.6502035007116282,
+      "grad_norm": 0.09328590333461761,
+      "learning_rate": 1.0520163208090315e-05,
+      "loss": 8.6944,
+      "step": 130200
+    },
+    {
+      "epoch": 0.6502534395365677,
+      "grad_norm": 0.09101434797048569,
+      "learning_rate": 1.0518661293148765e-05,
+      "loss": 8.6864,
+      "step": 130210
+    },
+    {
+      "epoch": 0.6503033783615072,
+      "grad_norm": 0.09618061780929565,
+      "learning_rate": 1.0517159378207215e-05,
+      "loss": 8.6875,
+      "step": 130220
+    },
+    {
+      "epoch": 0.6503533171864466,
+      "grad_norm": 0.0945608913898468,
+      "learning_rate": 1.0515657463265664e-05,
+      "loss": 8.689,
+      "step": 130230
+    },
+    {
+      "epoch": 0.650403256011386,
+      "grad_norm": 0.0871725007891655,
+      "learning_rate": 1.0514155548324114e-05,
+      "loss": 8.6919,
+      "step": 130240
+    },
+    {
+      "epoch": 0.6504531948363255,
+      "grad_norm": 0.09153078496456146,
+      "learning_rate": 1.0512653633382562e-05,
+      "loss": 8.6813,
+      "step": 130250
+    },
+    {
+      "epoch": 0.650503133661265,
+      "grad_norm": 0.09095324575901031,
+      "learning_rate": 1.0511151718441013e-05,
+      "loss": 8.6822,
+      "step": 130260
+    },
+    {
+      "epoch": 0.6505530724862044,
+      "grad_norm": 0.09055155515670776,
+      "learning_rate": 1.0509649803499463e-05,
+      "loss": 8.676,
+      "step": 130270
+    },
+    {
+      "epoch": 0.6506030113111438,
+      "grad_norm": 0.08829725533723831,
+      "learning_rate": 1.0508147888557911e-05,
+      "loss": 8.6923,
+      "step": 130280
+    },
+    {
+      "epoch": 0.6506529501360833,
+      "grad_norm": 0.09116893261671066,
+      "learning_rate": 1.0506645973616361e-05,
+      "loss": 8.6698,
+      "step": 130290
+    },
+    {
+      "epoch": 0.6507028889610228,
+      "grad_norm": 0.08940432965755463,
+      "learning_rate": 1.050514405867481e-05,
+      "loss": 8.6763,
+      "step": 130300
+    },
+    {
+      "epoch": 0.6507528277859622,
+      "grad_norm": 0.09223427623510361,
+      "learning_rate": 1.050364214373326e-05,
+      "loss": 8.6929,
+      "step": 130310
+    },
+    {
+      "epoch": 0.6508027666109016,
+      "grad_norm": 0.09368874877691269,
+      "learning_rate": 1.050214022879171e-05,
+      "loss": 8.6783,
+      "step": 130320
+    },
+    {
+      "epoch": 0.6508527054358411,
+      "grad_norm": 0.09133809059858322,
+      "learning_rate": 1.0500638313850159e-05,
+      "loss": 8.6861,
+      "step": 130330
+    },
+    {
+      "epoch": 0.6509026442607806,
+      "grad_norm": 0.09096644818782806,
+      "learning_rate": 1.0499136398908609e-05,
+      "loss": 8.6741,
+      "step": 130340
+    },
+    {
+      "epoch": 0.65095258308572,
+      "grad_norm": 0.09941566735506058,
+      "learning_rate": 1.0497634483967059e-05,
+      "loss": 8.6934,
+      "step": 130350
+    },
+    {
+      "epoch": 0.6510025219106594,
+      "grad_norm": 0.09109698235988617,
+      "learning_rate": 1.0496132569025508e-05,
+      "loss": 8.6798,
+      "step": 130360
+    },
+    {
+      "epoch": 0.6510524607355989,
+      "grad_norm": 0.09425774961709976,
+      "learning_rate": 1.0494630654083958e-05,
+      "loss": 8.6717,
+      "step": 130370
+    },
+    {
+      "epoch": 0.6511023995605384,
+      "grad_norm": 0.09574547410011292,
+      "learning_rate": 1.0493128739142406e-05,
+      "loss": 8.6847,
+      "step": 130380
+    },
+    {
+      "epoch": 0.6511523383854778,
+      "grad_norm": 0.08903608471155167,
+      "learning_rate": 1.0491626824200856e-05,
+      "loss": 8.6854,
+      "step": 130390
+    },
+    {
+      "epoch": 0.6512022772104172,
+      "grad_norm": 0.10021170973777771,
+      "learning_rate": 1.0490124909259307e-05,
+      "loss": 8.6747,
+      "step": 130400
+    },
+    {
+      "epoch": 0.6512522160353567,
+      "grad_norm": 0.09193012118339539,
+      "learning_rate": 1.0488622994317755e-05,
+      "loss": 8.6764,
+      "step": 130410
+    },
+    {
+      "epoch": 0.6513021548602962,
+      "grad_norm": 0.09240026026964188,
+      "learning_rate": 1.0487121079376205e-05,
+      "loss": 8.7007,
+      "step": 130420
+    },
+    {
+      "epoch": 0.6513520936852356,
+      "grad_norm": 0.0931842103600502,
+      "learning_rate": 1.0485619164434654e-05,
+      "loss": 8.6803,
+      "step": 130430
+    },
+    {
+      "epoch": 0.651402032510175,
+      "grad_norm": 0.0900469720363617,
+      "learning_rate": 1.0484117249493104e-05,
+      "loss": 8.6859,
+      "step": 130440
+    },
+    {
+      "epoch": 0.6514519713351145,
+      "grad_norm": 0.08829685300588608,
+      "learning_rate": 1.0482615334551554e-05,
+      "loss": 8.679,
+      "step": 130450
+    },
+    {
+      "epoch": 0.651501910160054,
+      "grad_norm": 0.08664678037166595,
+      "learning_rate": 1.0481113419610003e-05,
+      "loss": 8.6902,
+      "step": 130460
+    },
+    {
+      "epoch": 0.6515518489849934,
+      "grad_norm": 0.08698765188455582,
+      "learning_rate": 1.0479611504668453e-05,
+      "loss": 8.686,
+      "step": 130470
+    },
+    {
+      "epoch": 0.6516017878099328,
+      "grad_norm": 0.09792865067720413,
+      "learning_rate": 1.0478109589726901e-05,
+      "loss": 8.6769,
+      "step": 130480
+    },
+    {
+      "epoch": 0.6516517266348723,
+      "grad_norm": 0.08901020884513855,
+      "learning_rate": 1.0476607674785351e-05,
+      "loss": 8.6788,
+      "step": 130490
+    },
+    {
+      "epoch": 0.6517016654598118,
+      "grad_norm": 0.0966918021440506,
+      "learning_rate": 1.0475105759843802e-05,
+      "loss": 8.6681,
+      "step": 130500
+    },
+    {
+      "epoch": 0.6517516042847512,
+      "grad_norm": 0.0899975523352623,
+      "learning_rate": 1.0473603844902252e-05,
+      "loss": 8.7017,
+      "step": 130510
+    },
+    {
+      "epoch": 0.6518015431096906,
+      "grad_norm": 0.09501145780086517,
+      "learning_rate": 1.04721019299607e-05,
+      "loss": 8.698,
+      "step": 130520
+    },
+    {
+      "epoch": 0.65185148193463,
+      "grad_norm": 0.09583033621311188,
+      "learning_rate": 1.0470600015019149e-05,
+      "loss": 8.6757,
+      "step": 130530
+    },
+    {
+      "epoch": 0.6519014207595695,
+      "grad_norm": 0.09208723902702332,
+      "learning_rate": 1.0469098100077599e-05,
+      "loss": 8.6855,
+      "step": 130540
+    },
+    {
+      "epoch": 0.651951359584509,
+      "grad_norm": 0.09319637715816498,
+      "learning_rate": 1.0467596185136049e-05,
+      "loss": 8.6914,
+      "step": 130550
+    },
+    {
+      "epoch": 0.6520012984094484,
+      "grad_norm": 0.09361417591571808,
+      "learning_rate": 1.04660942701945e-05,
+      "loss": 8.683,
+      "step": 130560
+    },
+    {
+      "epoch": 0.6520512372343878,
+      "grad_norm": 0.0895400270819664,
+      "learning_rate": 1.0464592355252948e-05,
+      "loss": 8.6841,
+      "step": 130570
+    },
+    {
+      "epoch": 0.6521011760593273,
+      "grad_norm": 0.09408006817102432,
+      "learning_rate": 1.0463090440311396e-05,
+      "loss": 8.6757,
+      "step": 130580
+    },
+    {
+      "epoch": 0.6521511148842668,
+      "grad_norm": 0.09309608489274979,
+      "learning_rate": 1.0461588525369846e-05,
+      "loss": 8.6885,
+      "step": 130590
+    },
+    {
+      "epoch": 0.6522010537092062,
+      "grad_norm": 0.1002884954214096,
+      "learning_rate": 1.0460086610428297e-05,
+      "loss": 8.6925,
+      "step": 130600
+    },
+    {
+      "epoch": 0.6522509925341456,
+      "grad_norm": 0.09665568172931671,
+      "learning_rate": 1.0458584695486747e-05,
+      "loss": 8.6746,
+      "step": 130610
+    },
+    {
+      "epoch": 0.6523009313590851,
+      "grad_norm": 0.09286937117576599,
+      "learning_rate": 1.0457082780545195e-05,
+      "loss": 8.6669,
+      "step": 130620
+    },
+    {
+      "epoch": 0.6523508701840246,
+      "grad_norm": 0.094349205493927,
+      "learning_rate": 1.0455580865603644e-05,
+      "loss": 8.674,
+      "step": 130630
+    },
+    {
+      "epoch": 0.652400809008964,
+      "grad_norm": 0.0887681245803833,
+      "learning_rate": 1.0454078950662094e-05,
+      "loss": 8.6897,
+      "step": 130640
+    },
+    {
+      "epoch": 0.6524507478339034,
+      "grad_norm": 0.09428498893976212,
+      "learning_rate": 1.0452577035720544e-05,
+      "loss": 8.6805,
+      "step": 130650
+    },
+    {
+      "epoch": 0.6525006866588429,
+      "grad_norm": 0.09223032742738724,
+      "learning_rate": 1.0451075120778994e-05,
+      "loss": 8.685,
+      "step": 130660
+    },
+    {
+      "epoch": 0.6525506254837824,
+      "grad_norm": 0.09503738582134247,
+      "learning_rate": 1.0449573205837445e-05,
+      "loss": 8.6903,
+      "step": 130670
+    },
+    {
+      "epoch": 0.6526005643087218,
+      "grad_norm": 0.08950851857662201,
+      "learning_rate": 1.0448071290895891e-05,
+      "loss": 8.6922,
+      "step": 130680
+    },
+    {
+      "epoch": 0.6526505031336612,
+      "grad_norm": 0.08932535350322723,
+      "learning_rate": 1.0446569375954341e-05,
+      "loss": 8.6843,
+      "step": 130690
+    },
+    {
+      "epoch": 0.6527004419586007,
+      "grad_norm": 0.093544140458107,
+      "learning_rate": 1.0445067461012792e-05,
+      "loss": 8.6965,
+      "step": 130700
+    },
+    {
+      "epoch": 0.6527503807835402,
+      "grad_norm": 0.09570880979299545,
+      "learning_rate": 1.0443565546071242e-05,
+      "loss": 8.6699,
+      "step": 130710
+    },
+    {
+      "epoch": 0.6528003196084796,
+      "grad_norm": 0.095237135887146,
+      "learning_rate": 1.0442063631129692e-05,
+      "loss": 8.6778,
+      "step": 130720
+    },
+    {
+      "epoch": 0.652850258433419,
+      "grad_norm": 0.10413196682929993,
+      "learning_rate": 1.0440561716188139e-05,
+      "loss": 8.6738,
+      "step": 130730
+    },
+    {
+      "epoch": 0.6529001972583585,
+      "grad_norm": 0.08957455307245255,
+      "learning_rate": 1.0439059801246589e-05,
+      "loss": 8.6821,
+      "step": 130740
+    },
+    {
+      "epoch": 0.652950136083298,
+      "grad_norm": 0.1010318323969841,
+      "learning_rate": 1.043755788630504e-05,
+      "loss": 8.6882,
+      "step": 130750
+    },
+    {
+      "epoch": 0.6530000749082374,
+      "grad_norm": 0.09466128796339035,
+      "learning_rate": 1.043605597136349e-05,
+      "loss": 8.6749,
+      "step": 130760
+    },
+    {
+      "epoch": 0.6530500137331768,
+      "grad_norm": 0.0850997343659401,
+      "learning_rate": 1.043455405642194e-05,
+      "loss": 8.6898,
+      "step": 130770
+    },
+    {
+      "epoch": 0.6530999525581163,
+      "grad_norm": 0.09206338971853256,
+      "learning_rate": 1.0433052141480386e-05,
+      "loss": 8.6861,
+      "step": 130780
+    },
+    {
+      "epoch": 0.6531498913830558,
+      "grad_norm": 0.09347966313362122,
+      "learning_rate": 1.0431550226538836e-05,
+      "loss": 8.7026,
+      "step": 130790
+    },
+    {
+      "epoch": 0.6531998302079952,
+      "grad_norm": 0.08403073996305466,
+      "learning_rate": 1.0430048311597287e-05,
+      "loss": 8.6844,
+      "step": 130800
+    },
+    {
+      "epoch": 0.6532497690329346,
+      "grad_norm": 0.08621867746114731,
+      "learning_rate": 1.0428546396655737e-05,
+      "loss": 8.6709,
+      "step": 130810
+    },
+    {
+      "epoch": 0.6532997078578741,
+      "grad_norm": 0.08873015642166138,
+      "learning_rate": 1.0427044481714187e-05,
+      "loss": 8.6875,
+      "step": 130820
+    },
+    {
+      "epoch": 0.6533496466828136,
+      "grad_norm": 0.08667482435703278,
+      "learning_rate": 1.0425542566772636e-05,
+      "loss": 8.679,
+      "step": 130830
+    },
+    {
+      "epoch": 0.653399585507753,
+      "grad_norm": 0.09268020838499069,
+      "learning_rate": 1.0424040651831084e-05,
+      "loss": 8.6803,
+      "step": 130840
+    },
+    {
+      "epoch": 0.6534495243326924,
+      "grad_norm": 0.1045554131269455,
+      "learning_rate": 1.0422538736889534e-05,
+      "loss": 8.6686,
+      "step": 130850
+    },
+    {
+      "epoch": 0.6534994631576319,
+      "grad_norm": 0.08979924023151398,
+      "learning_rate": 1.0421036821947984e-05,
+      "loss": 8.6894,
+      "step": 130860
+    },
+    {
+      "epoch": 0.6535494019825714,
+      "grad_norm": 0.09369828552007675,
+      "learning_rate": 1.0419534907006435e-05,
+      "loss": 8.6807,
+      "step": 130870
+    },
+    {
+      "epoch": 0.6535993408075108,
+      "grad_norm": 0.09341225028038025,
+      "learning_rate": 1.0418032992064883e-05,
+      "loss": 8.6915,
+      "step": 130880
+    },
+    {
+      "epoch": 0.6536492796324502,
+      "grad_norm": 0.09005949646234512,
+      "learning_rate": 1.0416531077123331e-05,
+      "loss": 8.6985,
+      "step": 130890
+    },
+    {
+      "epoch": 0.6536992184573897,
+      "grad_norm": 0.08998000621795654,
+      "learning_rate": 1.0415029162181782e-05,
+      "loss": 8.6858,
+      "step": 130900
+    },
+    {
+      "epoch": 0.6537491572823292,
+      "grad_norm": 0.09265704452991486,
+      "learning_rate": 1.0413527247240232e-05,
+      "loss": 8.667,
+      "step": 130910
+    },
+    {
+      "epoch": 0.6537990961072686,
+      "grad_norm": 0.09195978194475174,
+      "learning_rate": 1.0412025332298682e-05,
+      "loss": 8.6813,
+      "step": 130920
+    },
+    {
+      "epoch": 0.653849034932208,
+      "grad_norm": 0.09529465436935425,
+      "learning_rate": 1.041052341735713e-05,
+      "loss": 8.6916,
+      "step": 130930
+    },
+    {
+      "epoch": 0.6538989737571475,
+      "grad_norm": 0.0947846919298172,
+      "learning_rate": 1.0409021502415579e-05,
+      "loss": 8.6761,
+      "step": 130940
+    },
+    {
+      "epoch": 0.653948912582087,
+      "grad_norm": 0.08519770950078964,
+      "learning_rate": 1.040751958747403e-05,
+      "loss": 8.6598,
+      "step": 130950
+    },
+    {
+      "epoch": 0.6539988514070264,
+      "grad_norm": 0.0898909941315651,
+      "learning_rate": 1.040601767253248e-05,
+      "loss": 8.6744,
+      "step": 130960
+    },
+    {
+      "epoch": 0.6540487902319658,
+      "grad_norm": 0.09745825827121735,
+      "learning_rate": 1.040451575759093e-05,
+      "loss": 8.6766,
+      "step": 130970
+    },
+    {
+      "epoch": 0.6540987290569052,
+      "grad_norm": 0.09335152804851532,
+      "learning_rate": 1.0403013842649378e-05,
+      "loss": 8.6872,
+      "step": 130980
+    },
+    {
+      "epoch": 0.6541486678818448,
+      "grad_norm": 0.09373294562101364,
+      "learning_rate": 1.0401511927707828e-05,
+      "loss": 8.6794,
+      "step": 130990
+    },
+    {
+      "epoch": 0.6541986067067842,
+      "grad_norm": 0.09135058522224426,
+      "learning_rate": 1.0400010012766277e-05,
+      "loss": 8.6976,
+      "step": 131000
+    },
+    {
+      "epoch": 0.6542485455317236,
+      "grad_norm": 0.09245934337377548,
+      "learning_rate": 1.0398508097824727e-05,
+      "loss": 8.6859,
+      "step": 131010
+    },
+    {
+      "epoch": 0.654298484356663,
+      "grad_norm": 0.08991315215826035,
+      "learning_rate": 1.0397006182883177e-05,
+      "loss": 8.6771,
+      "step": 131020
+    },
+    {
+      "epoch": 0.6543484231816026,
+      "grad_norm": 0.09018691629171371,
+      "learning_rate": 1.0395504267941626e-05,
+      "loss": 8.6944,
+      "step": 131030
+    },
+    {
+      "epoch": 0.654398362006542,
+      "grad_norm": 0.09176444262266159,
+      "learning_rate": 1.0394002353000076e-05,
+      "loss": 8.6827,
+      "step": 131040
+    },
+    {
+      "epoch": 0.6544483008314814,
+      "grad_norm": 0.09148535877466202,
+      "learning_rate": 1.0392500438058524e-05,
+      "loss": 8.68,
+      "step": 131050
+    },
+    {
+      "epoch": 0.6544982396564208,
+      "grad_norm": 0.09757846593856812,
+      "learning_rate": 1.0390998523116974e-05,
+      "loss": 8.6791,
+      "step": 131060
+    },
+    {
+      "epoch": 0.6545481784813604,
+      "grad_norm": 0.0926646739244461,
+      "learning_rate": 1.0389496608175425e-05,
+      "loss": 8.6906,
+      "step": 131070
+    },
+    {
+      "epoch": 0.6545981173062998,
+      "grad_norm": 0.09516071528196335,
+      "learning_rate": 1.0387994693233873e-05,
+      "loss": 8.674,
+      "step": 131080
+    },
+    {
+      "epoch": 0.6546480561312392,
+      "grad_norm": 0.09518009424209595,
+      "learning_rate": 1.0386492778292323e-05,
+      "loss": 8.68,
+      "step": 131090
+    },
+    {
+      "epoch": 0.6546979949561786,
+      "grad_norm": 0.09322939813137054,
+      "learning_rate": 1.0384990863350772e-05,
+      "loss": 8.677,
+      "step": 131100
+    },
+    {
+      "epoch": 0.6547479337811182,
+      "grad_norm": 0.0915323942899704,
+      "learning_rate": 1.0383488948409222e-05,
+      "loss": 8.6732,
+      "step": 131110
+    },
+    {
+      "epoch": 0.6547978726060576,
+      "grad_norm": 0.0918693020939827,
+      "learning_rate": 1.0381987033467672e-05,
+      "loss": 8.6658,
+      "step": 131120
+    },
+    {
+      "epoch": 0.654847811430997,
+      "grad_norm": 0.09240451455116272,
+      "learning_rate": 1.038048511852612e-05,
+      "loss": 8.6933,
+      "step": 131130
+    },
+    {
+      "epoch": 0.6548977502559364,
+      "grad_norm": 0.09553481638431549,
+      "learning_rate": 1.037898320358457e-05,
+      "loss": 8.6893,
+      "step": 131140
+    },
+    {
+      "epoch": 0.654947689080876,
+      "grad_norm": 0.09170468896627426,
+      "learning_rate": 1.0377481288643021e-05,
+      "loss": 8.6906,
+      "step": 131150
+    },
+    {
+      "epoch": 0.6549976279058154,
+      "grad_norm": 0.09421038627624512,
+      "learning_rate": 1.037597937370147e-05,
+      "loss": 8.6669,
+      "step": 131160
+    },
+    {
+      "epoch": 0.6550475667307548,
+      "grad_norm": 0.08922944962978363,
+      "learning_rate": 1.037447745875992e-05,
+      "loss": 8.6888,
+      "step": 131170
+    },
+    {
+      "epoch": 0.6550975055556942,
+      "grad_norm": 0.09422342479228973,
+      "learning_rate": 1.0372975543818368e-05,
+      "loss": 8.6767,
+      "step": 131180
+    },
+    {
+      "epoch": 0.6551474443806338,
+      "grad_norm": 0.09131404012441635,
+      "learning_rate": 1.0371473628876818e-05,
+      "loss": 8.6735,
+      "step": 131190
+    },
+    {
+      "epoch": 0.6551973832055732,
+      "grad_norm": 0.08546370267868042,
+      "learning_rate": 1.0369971713935268e-05,
+      "loss": 8.6832,
+      "step": 131200
+    },
+    {
+      "epoch": 0.6552473220305126,
+      "grad_norm": 0.09401803463697433,
+      "learning_rate": 1.0368469798993717e-05,
+      "loss": 8.68,
+      "step": 131210
+    },
+    {
+      "epoch": 0.655297260855452,
+      "grad_norm": 0.09075015783309937,
+      "learning_rate": 1.0366967884052167e-05,
+      "loss": 8.6723,
+      "step": 131220
+    },
+    {
+      "epoch": 0.6553471996803916,
+      "grad_norm": 0.09611096978187561,
+      "learning_rate": 1.0365465969110616e-05,
+      "loss": 8.6929,
+      "step": 131230
+    },
+    {
+      "epoch": 0.655397138505331,
+      "grad_norm": 0.09546373039484024,
+      "learning_rate": 1.0363964054169066e-05,
+      "loss": 8.6781,
+      "step": 131240
+    },
+    {
+      "epoch": 0.6554470773302704,
+      "grad_norm": 0.09710092097520828,
+      "learning_rate": 1.0362462139227516e-05,
+      "loss": 8.6739,
+      "step": 131250
+    },
+    {
+      "epoch": 0.6554970161552098,
+      "grad_norm": 0.09392837435007095,
+      "learning_rate": 1.0360960224285964e-05,
+      "loss": 8.6824,
+      "step": 131260
+    },
+    {
+      "epoch": 0.6555469549801494,
+      "grad_norm": 0.09128575772047043,
+      "learning_rate": 1.0359458309344415e-05,
+      "loss": 8.6771,
+      "step": 131270
+    },
+    {
+      "epoch": 0.6555968938050888,
+      "grad_norm": 0.09115424007177353,
+      "learning_rate": 1.0357956394402863e-05,
+      "loss": 8.6769,
+      "step": 131280
+    },
+    {
+      "epoch": 0.6556468326300282,
+      "grad_norm": 0.09360767900943756,
+      "learning_rate": 1.0356454479461313e-05,
+      "loss": 8.6999,
+      "step": 131290
+    },
+    {
+      "epoch": 0.6556967714549676,
+      "grad_norm": 0.09261327236890793,
+      "learning_rate": 1.0354952564519763e-05,
+      "loss": 8.6671,
+      "step": 131300
+    },
+    {
+      "epoch": 0.6557467102799072,
+      "grad_norm": 0.09055551886558533,
+      "learning_rate": 1.0353450649578214e-05,
+      "loss": 8.6813,
+      "step": 131310
+    },
+    {
+      "epoch": 0.6557966491048466,
+      "grad_norm": 0.0952824130654335,
+      "learning_rate": 1.0351948734636662e-05,
+      "loss": 8.682,
+      "step": 131320
+    },
+    {
+      "epoch": 0.655846587929786,
+      "grad_norm": 0.09080826491117477,
+      "learning_rate": 1.035044681969511e-05,
+      "loss": 8.6825,
+      "step": 131330
+    },
+    {
+      "epoch": 0.6558965267547254,
+      "grad_norm": 0.0927935540676117,
+      "learning_rate": 1.034894490475356e-05,
+      "loss": 8.6942,
+      "step": 131340
+    },
+    {
+      "epoch": 0.655946465579665,
+      "grad_norm": 0.09244295954704285,
+      "learning_rate": 1.0347442989812011e-05,
+      "loss": 8.6901,
+      "step": 131350
+    },
+    {
+      "epoch": 0.6559964044046044,
+      "grad_norm": 0.09503054618835449,
+      "learning_rate": 1.0345941074870461e-05,
+      "loss": 8.6633,
+      "step": 131360
+    },
+    {
+      "epoch": 0.6560463432295438,
+      "grad_norm": 0.09178107976913452,
+      "learning_rate": 1.034443915992891e-05,
+      "loss": 8.6971,
+      "step": 131370
+    },
+    {
+      "epoch": 0.6560962820544832,
+      "grad_norm": 0.09662067145109177,
+      "learning_rate": 1.0342937244987358e-05,
+      "loss": 8.675,
+      "step": 131380
+    },
+    {
+      "epoch": 0.6561462208794228,
+      "grad_norm": 0.09024110436439514,
+      "learning_rate": 1.0341435330045808e-05,
+      "loss": 8.6675,
+      "step": 131390
+    },
+    {
+      "epoch": 0.6561961597043622,
+      "grad_norm": 0.09327442944049835,
+      "learning_rate": 1.0339933415104258e-05,
+      "loss": 8.6674,
+      "step": 131400
+    },
+    {
+      "epoch": 0.6562460985293016,
+      "grad_norm": 0.09160305559635162,
+      "learning_rate": 1.0338431500162709e-05,
+      "loss": 8.6841,
+      "step": 131410
+    },
+    {
+      "epoch": 0.656296037354241,
+      "grad_norm": 0.09517280757427216,
+      "learning_rate": 1.0336929585221157e-05,
+      "loss": 8.6751,
+      "step": 131420
+    },
+    {
+      "epoch": 0.6563459761791806,
+      "grad_norm": 0.09651865065097809,
+      "learning_rate": 1.0335427670279606e-05,
+      "loss": 8.6741,
+      "step": 131430
+    },
+    {
+      "epoch": 0.65639591500412,
+      "grad_norm": 0.09696118533611298,
+      "learning_rate": 1.0333925755338056e-05,
+      "loss": 8.6646,
+      "step": 131440
+    },
+    {
+      "epoch": 0.6564458538290594,
+      "grad_norm": 0.09164685755968094,
+      "learning_rate": 1.0332423840396506e-05,
+      "loss": 8.6919,
+      "step": 131450
+    },
+    {
+      "epoch": 0.6564957926539988,
+      "grad_norm": 0.09211573749780655,
+      "learning_rate": 1.0330921925454956e-05,
+      "loss": 8.6674,
+      "step": 131460
+    },
+    {
+      "epoch": 0.6565457314789384,
+      "grad_norm": 0.09793127328157425,
+      "learning_rate": 1.0329420010513406e-05,
+      "loss": 8.6897,
+      "step": 131470
+    },
+    {
+      "epoch": 0.6565956703038778,
+      "grad_norm": 0.09273996949195862,
+      "learning_rate": 1.0327918095571853e-05,
+      "loss": 8.6733,
+      "step": 131480
+    },
+    {
+      "epoch": 0.6566456091288172,
+      "grad_norm": 0.08516659587621689,
+      "learning_rate": 1.0326416180630303e-05,
+      "loss": 8.6978,
+      "step": 131490
+    },
+    {
+      "epoch": 0.6566955479537566,
+      "grad_norm": 0.09348917752504349,
+      "learning_rate": 1.0324914265688753e-05,
+      "loss": 8.6678,
+      "step": 131500
+    },
+    {
+      "epoch": 0.6567454867786962,
+      "grad_norm": 0.09061364084482193,
+      "learning_rate": 1.0323412350747204e-05,
+      "loss": 8.688,
+      "step": 131510
+    },
+    {
+      "epoch": 0.6567954256036356,
+      "grad_norm": 0.09217441827058792,
+      "learning_rate": 1.0321910435805654e-05,
+      "loss": 8.6757,
+      "step": 131520
+    },
+    {
+      "epoch": 0.656845364428575,
+      "grad_norm": 0.09263032674789429,
+      "learning_rate": 1.03204085208641e-05,
+      "loss": 8.682,
+      "step": 131530
+    },
+    {
+      "epoch": 0.6568953032535144,
+      "grad_norm": 0.09283848851919174,
+      "learning_rate": 1.031890660592255e-05,
+      "loss": 8.6898,
+      "step": 131540
+    },
+    {
+      "epoch": 0.6569452420784538,
+      "grad_norm": 0.09214113652706146,
+      "learning_rate": 1.0317404690981001e-05,
+      "loss": 8.6754,
+      "step": 131550
+    },
+    {
+      "epoch": 0.6569951809033934,
+      "grad_norm": 0.09393681585788727,
+      "learning_rate": 1.0315902776039451e-05,
+      "loss": 8.6623,
+      "step": 131560
+    },
+    {
+      "epoch": 0.6570451197283328,
+      "grad_norm": 0.08560718595981598,
+      "learning_rate": 1.0314400861097901e-05,
+      "loss": 8.6695,
+      "step": 131570
+    },
+    {
+      "epoch": 0.6570950585532722,
+      "grad_norm": 0.09654083102941513,
+      "learning_rate": 1.0312898946156348e-05,
+      "loss": 8.6811,
+      "step": 131580
+    },
+    {
+      "epoch": 0.6571449973782116,
+      "grad_norm": 0.09015607833862305,
+      "learning_rate": 1.0311397031214798e-05,
+      "loss": 8.6836,
+      "step": 131590
+    },
+    {
+      "epoch": 0.6571949362031512,
+      "grad_norm": 0.08412002772092819,
+      "learning_rate": 1.0309895116273248e-05,
+      "loss": 8.6766,
+      "step": 131600
+    },
+    {
+      "epoch": 0.6572448750280906,
+      "grad_norm": 0.09406512975692749,
+      "learning_rate": 1.0308393201331699e-05,
+      "loss": 8.6751,
+      "step": 131610
+    },
+    {
+      "epoch": 0.65729481385303,
+      "grad_norm": 0.09097196906805038,
+      "learning_rate": 1.0306891286390149e-05,
+      "loss": 8.6757,
+      "step": 131620
+    },
+    {
+      "epoch": 0.6573447526779694,
+      "grad_norm": 0.09862224012613297,
+      "learning_rate": 1.0305389371448597e-05,
+      "loss": 8.6832,
+      "step": 131630
+    },
+    {
+      "epoch": 0.657394691502909,
+      "grad_norm": 0.09050843119621277,
+      "learning_rate": 1.0303887456507046e-05,
+      "loss": 8.6709,
+      "step": 131640
+    },
+    {
+      "epoch": 0.6574446303278484,
+      "grad_norm": 0.09337736666202545,
+      "learning_rate": 1.0302385541565496e-05,
+      "loss": 8.6862,
+      "step": 131650
+    },
+    {
+      "epoch": 0.6574945691527878,
+      "grad_norm": 0.08687882125377655,
+      "learning_rate": 1.0300883626623946e-05,
+      "loss": 8.684,
+      "step": 131660
+    },
+    {
+      "epoch": 0.6575445079777272,
+      "grad_norm": 0.09270884841680527,
+      "learning_rate": 1.0299381711682396e-05,
+      "loss": 8.6805,
+      "step": 131670
+    },
+    {
+      "epoch": 0.6575944468026668,
+      "grad_norm": 0.08802757412195206,
+      "learning_rate": 1.0297879796740845e-05,
+      "loss": 8.6734,
+      "step": 131680
+    },
+    {
+      "epoch": 0.6576443856276062,
+      "grad_norm": 0.09502147138118744,
+      "learning_rate": 1.0296377881799293e-05,
+      "loss": 8.6866,
+      "step": 131690
+    },
+    {
+      "epoch": 0.6576943244525456,
+      "grad_norm": 0.08749578148126602,
+      "learning_rate": 1.0294875966857743e-05,
+      "loss": 8.664,
+      "step": 131700
+    },
+    {
+      "epoch": 0.657744263277485,
+      "grad_norm": 0.09266682714223862,
+      "learning_rate": 1.0293374051916194e-05,
+      "loss": 8.6682,
+      "step": 131710
+    },
+    {
+      "epoch": 0.6577942021024246,
+      "grad_norm": 0.08939957618713379,
+      "learning_rate": 1.0291872136974644e-05,
+      "loss": 8.6629,
+      "step": 131720
+    },
+    {
+      "epoch": 0.657844140927364,
+      "grad_norm": 0.09014017879962921,
+      "learning_rate": 1.0290370222033092e-05,
+      "loss": 8.6807,
+      "step": 131730
+    },
+    {
+      "epoch": 0.6578940797523034,
+      "grad_norm": 0.0930858924984932,
+      "learning_rate": 1.028886830709154e-05,
+      "loss": 8.673,
+      "step": 131740
+    },
+    {
+      "epoch": 0.6579440185772428,
+      "grad_norm": 0.09012752026319504,
+      "learning_rate": 1.0287366392149991e-05,
+      "loss": 8.6753,
+      "step": 131750
+    },
+    {
+      "epoch": 0.6579939574021824,
+      "grad_norm": 0.09860425442457199,
+      "learning_rate": 1.0285864477208441e-05,
+      "loss": 8.6818,
+      "step": 131760
+    },
+    {
+      "epoch": 0.6580438962271218,
+      "grad_norm": 0.08776333183050156,
+      "learning_rate": 1.0284362562266891e-05,
+      "loss": 8.6717,
+      "step": 131770
+    },
+    {
+      "epoch": 0.6580938350520612,
+      "grad_norm": 0.09285827726125717,
+      "learning_rate": 1.028286064732534e-05,
+      "loss": 8.6907,
+      "step": 131780
+    },
+    {
+      "epoch": 0.6581437738770006,
+      "grad_norm": 0.09714893996715546,
+      "learning_rate": 1.0281358732383788e-05,
+      "loss": 8.6782,
+      "step": 131790
+    },
+    {
+      "epoch": 0.6581937127019402,
+      "grad_norm": 0.09962324798107147,
+      "learning_rate": 1.0279856817442238e-05,
+      "loss": 8.6517,
+      "step": 131800
+    },
+    {
+      "epoch": 0.6582436515268796,
+      "grad_norm": 0.0935245156288147,
+      "learning_rate": 1.0278354902500689e-05,
+      "loss": 8.6855,
+      "step": 131810
+    },
+    {
+      "epoch": 0.658293590351819,
+      "grad_norm": 0.09169341623783112,
+      "learning_rate": 1.0276852987559139e-05,
+      "loss": 8.6795,
+      "step": 131820
+    },
+    {
+      "epoch": 0.6583435291767584,
+      "grad_norm": 0.09153547137975693,
+      "learning_rate": 1.0275351072617587e-05,
+      "loss": 8.6665,
+      "step": 131830
+    },
+    {
+      "epoch": 0.658393468001698,
+      "grad_norm": 0.0904143676161766,
+      "learning_rate": 1.0273849157676037e-05,
+      "loss": 8.6895,
+      "step": 131840
+    },
+    {
+      "epoch": 0.6584434068266374,
+      "grad_norm": 0.09252644330263138,
+      "learning_rate": 1.0272347242734486e-05,
+      "loss": 8.681,
+      "step": 131850
+    },
+    {
+      "epoch": 0.6584933456515768,
+      "grad_norm": 0.08688996732234955,
+      "learning_rate": 1.0270845327792936e-05,
+      "loss": 8.6822,
+      "step": 131860
+    },
+    {
+      "epoch": 0.6585432844765162,
+      "grad_norm": 0.09266286343336105,
+      "learning_rate": 1.0269343412851386e-05,
+      "loss": 8.6785,
+      "step": 131870
+    },
+    {
+      "epoch": 0.6585932233014558,
+      "grad_norm": 0.09570097178220749,
+      "learning_rate": 1.0267841497909835e-05,
+      "loss": 8.6812,
+      "step": 131880
+    },
+    {
+      "epoch": 0.6586431621263952,
+      "grad_norm": 0.08836355060338974,
+      "learning_rate": 1.0266339582968285e-05,
+      "loss": 8.6714,
+      "step": 131890
+    },
+    {
+      "epoch": 0.6586931009513346,
+      "grad_norm": 0.09223395586013794,
+      "learning_rate": 1.0264837668026733e-05,
+      "loss": 8.6717,
+      "step": 131900
+    },
+    {
+      "epoch": 0.658743039776274,
+      "grad_norm": 0.09718294441699982,
+      "learning_rate": 1.0263335753085184e-05,
+      "loss": 8.6782,
+      "step": 131910
+    },
+    {
+      "epoch": 0.6587929786012136,
+      "grad_norm": 0.09103352576494217,
+      "learning_rate": 1.0261833838143634e-05,
+      "loss": 8.6824,
+      "step": 131920
+    },
+    {
+      "epoch": 0.658842917426153,
+      "grad_norm": 0.09198730438947678,
+      "learning_rate": 1.0260331923202084e-05,
+      "loss": 8.6817,
+      "step": 131930
+    },
+    {
+      "epoch": 0.6588928562510924,
+      "grad_norm": 0.09134362637996674,
+      "learning_rate": 1.0258830008260532e-05,
+      "loss": 8.6747,
+      "step": 131940
+    },
+    {
+      "epoch": 0.6589427950760318,
+      "grad_norm": 0.09106438606977463,
+      "learning_rate": 1.0257328093318981e-05,
+      "loss": 8.6674,
+      "step": 131950
+    },
+    {
+      "epoch": 0.6589927339009714,
+      "grad_norm": 0.08696383237838745,
+      "learning_rate": 1.0255826178377431e-05,
+      "loss": 8.6701,
+      "step": 131960
+    },
+    {
+      "epoch": 0.6590426727259108,
+      "grad_norm": 0.0889415293931961,
+      "learning_rate": 1.0254324263435881e-05,
+      "loss": 8.6741,
+      "step": 131970
+    },
+    {
+      "epoch": 0.6590926115508502,
+      "grad_norm": 0.09402123838663101,
+      "learning_rate": 1.0252822348494331e-05,
+      "loss": 8.6776,
+      "step": 131980
+    },
+    {
+      "epoch": 0.6591425503757896,
+      "grad_norm": 0.0896662175655365,
+      "learning_rate": 1.025132043355278e-05,
+      "loss": 8.6766,
+      "step": 131990
+    },
+    {
+      "epoch": 0.6591924892007291,
+      "grad_norm": 0.09504416584968567,
+      "learning_rate": 1.024981851861123e-05,
+      "loss": 8.6715,
+      "step": 132000
+    },
+    {
+      "epoch": 0.6592424280256686,
+      "grad_norm": 0.0929967388510704,
+      "learning_rate": 1.0248316603669679e-05,
+      "loss": 8.6769,
+      "step": 132010
+    },
+    {
+      "epoch": 0.659292366850608,
+      "grad_norm": 0.09154746681451797,
+      "learning_rate": 1.0246814688728129e-05,
+      "loss": 8.6683,
+      "step": 132020
+    },
+    {
+      "epoch": 0.6593423056755474,
+      "grad_norm": 0.09038006514310837,
+      "learning_rate": 1.0245312773786579e-05,
+      "loss": 8.6774,
+      "step": 132030
+    },
+    {
+      "epoch": 0.659392244500487,
+      "grad_norm": 0.08670198172330856,
+      "learning_rate": 1.0243810858845027e-05,
+      "loss": 8.6843,
+      "step": 132040
+    },
+    {
+      "epoch": 0.6594421833254264,
+      "grad_norm": 0.08812010288238525,
+      "learning_rate": 1.0242308943903478e-05,
+      "loss": 8.6761,
+      "step": 132050
+    },
+    {
+      "epoch": 0.6594921221503658,
+      "grad_norm": 0.08771144598722458,
+      "learning_rate": 1.0240807028961926e-05,
+      "loss": 8.6834,
+      "step": 132060
+    },
+    {
+      "epoch": 0.6595420609753052,
+      "grad_norm": 0.08678391575813293,
+      "learning_rate": 1.0239305114020376e-05,
+      "loss": 8.6703,
+      "step": 132070
+    },
+    {
+      "epoch": 0.6595919998002447,
+      "grad_norm": 0.09273753315210342,
+      "learning_rate": 1.0237803199078826e-05,
+      "loss": 8.6742,
+      "step": 132080
+    },
+    {
+      "epoch": 0.6596419386251842,
+      "grad_norm": 0.09024372696876526,
+      "learning_rate": 1.0236301284137275e-05,
+      "loss": 8.6711,
+      "step": 132090
+    },
+    {
+      "epoch": 0.6596918774501236,
+      "grad_norm": 0.09069012850522995,
+      "learning_rate": 1.0234799369195725e-05,
+      "loss": 8.6687,
+      "step": 132100
+    },
+    {
+      "epoch": 0.659741816275063,
+      "grad_norm": 0.0902455523610115,
+      "learning_rate": 1.0233297454254174e-05,
+      "loss": 8.667,
+      "step": 132110
+    },
+    {
+      "epoch": 0.6597917551000025,
+      "grad_norm": 0.08929167687892914,
+      "learning_rate": 1.0231795539312624e-05,
+      "loss": 8.6656,
+      "step": 132120
+    },
+    {
+      "epoch": 0.659841693924942,
+      "grad_norm": 0.08881199359893799,
+      "learning_rate": 1.0230293624371074e-05,
+      "loss": 8.6793,
+      "step": 132130
+    },
+    {
+      "epoch": 0.6598916327498814,
+      "grad_norm": 0.09602124989032745,
+      "learning_rate": 1.0228791709429522e-05,
+      "loss": 8.6794,
+      "step": 132140
+    },
+    {
+      "epoch": 0.6599415715748208,
+      "grad_norm": 0.09030324965715408,
+      "learning_rate": 1.0227289794487973e-05,
+      "loss": 8.6628,
+      "step": 132150
+    },
+    {
+      "epoch": 0.6599915103997603,
+      "grad_norm": 0.09483480453491211,
+      "learning_rate": 1.0225787879546423e-05,
+      "loss": 8.668,
+      "step": 132160
+    },
+    {
+      "epoch": 0.6600414492246998,
+      "grad_norm": 0.09794318675994873,
+      "learning_rate": 1.0224285964604871e-05,
+      "loss": 8.6631,
+      "step": 132170
+    },
+    {
+      "epoch": 0.6600913880496392,
+      "grad_norm": 0.09133700281381607,
+      "learning_rate": 1.0222784049663321e-05,
+      "loss": 8.6633,
+      "step": 132180
+    },
+    {
+      "epoch": 0.6601413268745786,
+      "grad_norm": 0.09112256020307541,
+      "learning_rate": 1.022128213472177e-05,
+      "loss": 8.6794,
+      "step": 132190
+    },
+    {
+      "epoch": 0.6601912656995181,
+      "grad_norm": 0.09767073392868042,
+      "learning_rate": 1.021978021978022e-05,
+      "loss": 8.6609,
+      "step": 132200
+    },
+    {
+      "epoch": 0.6602412045244576,
+      "grad_norm": 0.09043072909116745,
+      "learning_rate": 1.021827830483867e-05,
+      "loss": 8.6865,
+      "step": 132210
+    },
+    {
+      "epoch": 0.660291143349397,
+      "grad_norm": 0.08856266736984253,
+      "learning_rate": 1.0216776389897119e-05,
+      "loss": 8.6811,
+      "step": 132220
+    },
+    {
+      "epoch": 0.6603410821743364,
+      "grad_norm": 0.09094087034463882,
+      "learning_rate": 1.0215274474955569e-05,
+      "loss": 8.6836,
+      "step": 132230
+    },
+    {
+      "epoch": 0.6603910209992759,
+      "grad_norm": 0.09180308133363724,
+      "learning_rate": 1.0213772560014017e-05,
+      "loss": 8.677,
+      "step": 132240
+    },
+    {
+      "epoch": 0.6604409598242154,
+      "grad_norm": 0.09490694850683212,
+      "learning_rate": 1.0212270645072468e-05,
+      "loss": 8.6798,
+      "step": 132250
+    },
+    {
+      "epoch": 0.6604908986491548,
+      "grad_norm": 0.09207696467638016,
+      "learning_rate": 1.0210768730130918e-05,
+      "loss": 8.6676,
+      "step": 132260
+    },
+    {
+      "epoch": 0.6605408374740942,
+      "grad_norm": 0.08791854977607727,
+      "learning_rate": 1.0209266815189366e-05,
+      "loss": 8.6882,
+      "step": 132270
+    },
+    {
+      "epoch": 0.6605907762990337,
+      "grad_norm": 0.09454667568206787,
+      "learning_rate": 1.0207764900247816e-05,
+      "loss": 8.6748,
+      "step": 132280
+    },
+    {
+      "epoch": 0.6606407151239732,
+      "grad_norm": 0.08923207223415375,
+      "learning_rate": 1.0206262985306265e-05,
+      "loss": 8.6915,
+      "step": 132290
+    },
+    {
+      "epoch": 0.6606906539489126,
+      "grad_norm": 0.08957415819168091,
+      "learning_rate": 1.0204761070364715e-05,
+      "loss": 8.6757,
+      "step": 132300
+    },
+    {
+      "epoch": 0.660740592773852,
+      "grad_norm": 0.09061232954263687,
+      "learning_rate": 1.0203259155423165e-05,
+      "loss": 8.675,
+      "step": 132310
+    },
+    {
+      "epoch": 0.6607905315987915,
+      "grad_norm": 0.0913691371679306,
+      "learning_rate": 1.0201757240481616e-05,
+      "loss": 8.6773,
+      "step": 132320
+    },
+    {
+      "epoch": 0.660840470423731,
+      "grad_norm": 0.08861444145441055,
+      "learning_rate": 1.0200255325540064e-05,
+      "loss": 8.6856,
+      "step": 132330
+    },
+    {
+      "epoch": 0.6608904092486704,
+      "grad_norm": 0.08731677383184433,
+      "learning_rate": 1.0198753410598512e-05,
+      "loss": 8.6783,
+      "step": 132340
+    },
+    {
+      "epoch": 0.6609403480736098,
+      "grad_norm": 0.09219668805599213,
+      "learning_rate": 1.0197251495656963e-05,
+      "loss": 8.6718,
+      "step": 132350
+    },
+    {
+      "epoch": 0.6609902868985493,
+      "grad_norm": 0.08887942135334015,
+      "learning_rate": 1.0195749580715413e-05,
+      "loss": 8.6755,
+      "step": 132360
+    },
+    {
+      "epoch": 0.6610402257234888,
+      "grad_norm": 0.09542329609394073,
+      "learning_rate": 1.0194247665773863e-05,
+      "loss": 8.6684,
+      "step": 132370
+    },
+    {
+      "epoch": 0.6610901645484282,
+      "grad_norm": 0.0939764529466629,
+      "learning_rate": 1.0192745750832312e-05,
+      "loss": 8.6675,
+      "step": 132380
+    },
+    {
+      "epoch": 0.6611401033733676,
+      "grad_norm": 0.09207078814506531,
+      "learning_rate": 1.019124383589076e-05,
+      "loss": 8.6756,
+      "step": 132390
+    },
+    {
+      "epoch": 0.6611900421983071,
+      "grad_norm": 0.09722395241260529,
+      "learning_rate": 1.018974192094921e-05,
+      "loss": 8.6686,
+      "step": 132400
+    },
+    {
+      "epoch": 0.6612399810232465,
+      "grad_norm": 0.10081882029771805,
+      "learning_rate": 1.018824000600766e-05,
+      "loss": 8.6651,
+      "step": 132410
+    },
+    {
+      "epoch": 0.661289919848186,
+      "grad_norm": 0.09268742799758911,
+      "learning_rate": 1.018673809106611e-05,
+      "loss": 8.6535,
+      "step": 132420
+    },
+    {
+      "epoch": 0.6613398586731254,
+      "grad_norm": 0.08973268419504166,
+      "learning_rate": 1.0185236176124559e-05,
+      "loss": 8.6679,
+      "step": 132430
+    },
+    {
+      "epoch": 0.6613897974980649,
+      "grad_norm": 0.08811484277248383,
+      "learning_rate": 1.0183734261183007e-05,
+      "loss": 8.6566,
+      "step": 132440
+    },
+    {
+      "epoch": 0.6614397363230043,
+      "grad_norm": 0.09202300012111664,
+      "learning_rate": 1.0182232346241458e-05,
+      "loss": 8.6779,
+      "step": 132450
+    },
+    {
+      "epoch": 0.6614896751479438,
+      "grad_norm": 0.08758775144815445,
+      "learning_rate": 1.0180730431299908e-05,
+      "loss": 8.6919,
+      "step": 132460
+    },
+    {
+      "epoch": 0.6615396139728832,
+      "grad_norm": 0.08985673636198044,
+      "learning_rate": 1.0179228516358358e-05,
+      "loss": 8.678,
+      "step": 132470
+    },
+    {
+      "epoch": 0.6615895527978227,
+      "grad_norm": 0.09474221616983414,
+      "learning_rate": 1.0177726601416808e-05,
+      "loss": 8.6547,
+      "step": 132480
+    },
+    {
+      "epoch": 0.6616394916227621,
+      "grad_norm": 0.09064290672540665,
+      "learning_rate": 1.0176224686475255e-05,
+      "loss": 8.6735,
+      "step": 132490
+    },
+    {
+      "epoch": 0.6616894304477016,
+      "grad_norm": 0.09543025493621826,
+      "learning_rate": 1.0174722771533705e-05,
+      "loss": 8.6677,
+      "step": 132500
+    },
+    {
+      "epoch": 0.661739369272641,
+      "grad_norm": 0.08833234757184982,
+      "learning_rate": 1.0173220856592155e-05,
+      "loss": 8.6784,
+      "step": 132510
+    },
+    {
+      "epoch": 0.6617893080975804,
+      "grad_norm": 0.09526517242193222,
+      "learning_rate": 1.0171718941650606e-05,
+      "loss": 8.6595,
+      "step": 132520
+    },
+    {
+      "epoch": 0.6618392469225199,
+      "grad_norm": 0.09490874409675598,
+      "learning_rate": 1.0170217026709056e-05,
+      "loss": 8.6708,
+      "step": 132530
+    },
+    {
+      "epoch": 0.6618891857474594,
+      "grad_norm": 0.0921601727604866,
+      "learning_rate": 1.0168715111767502e-05,
+      "loss": 8.6814,
+      "step": 132540
+    },
+    {
+      "epoch": 0.6619391245723988,
+      "grad_norm": 0.09030061960220337,
+      "learning_rate": 1.0167213196825953e-05,
+      "loss": 8.6737,
+      "step": 132550
+    },
+    {
+      "epoch": 0.6619890633973382,
+      "grad_norm": 0.09003777801990509,
+      "learning_rate": 1.0165711281884403e-05,
+      "loss": 8.6747,
+      "step": 132560
+    },
+    {
+      "epoch": 0.6620390022222777,
+      "grad_norm": 0.09126746654510498,
+      "learning_rate": 1.0164209366942853e-05,
+      "loss": 8.6858,
+      "step": 132570
+    },
+    {
+      "epoch": 0.6620889410472172,
+      "grad_norm": 0.08809841424226761,
+      "learning_rate": 1.0162707452001303e-05,
+      "loss": 8.6722,
+      "step": 132580
+    },
+    {
+      "epoch": 0.6621388798721566,
+      "grad_norm": 0.09087785333395004,
+      "learning_rate": 1.016120553705975e-05,
+      "loss": 8.6795,
+      "step": 132590
+    },
+    {
+      "epoch": 0.662188818697096,
+      "grad_norm": 0.09890290349721909,
+      "learning_rate": 1.01597036221182e-05,
+      "loss": 8.6719,
+      "step": 132600
+    },
+    {
+      "epoch": 0.6622387575220355,
+      "grad_norm": 0.08723946660757065,
+      "learning_rate": 1.015820170717665e-05,
+      "loss": 8.6765,
+      "step": 132610
+    },
+    {
+      "epoch": 0.662288696346975,
+      "grad_norm": 0.09441614151000977,
+      "learning_rate": 1.01566997922351e-05,
+      "loss": 8.6709,
+      "step": 132620
+    },
+    {
+      "epoch": 0.6623386351719144,
+      "grad_norm": 0.09669451415538788,
+      "learning_rate": 1.015519787729355e-05,
+      "loss": 8.6856,
+      "step": 132630
+    },
+    {
+      "epoch": 0.6623885739968538,
+      "grad_norm": 0.0917108878493309,
+      "learning_rate": 1.0153695962352e-05,
+      "loss": 8.6664,
+      "step": 132640
+    },
+    {
+      "epoch": 0.6624385128217933,
+      "grad_norm": 0.0893852636218071,
+      "learning_rate": 1.0152194047410448e-05,
+      "loss": 8.6694,
+      "step": 132650
+    },
+    {
+      "epoch": 0.6624884516467328,
+      "grad_norm": 0.09049538522958755,
+      "learning_rate": 1.0150692132468898e-05,
+      "loss": 8.6595,
+      "step": 132660
+    },
+    {
+      "epoch": 0.6625383904716722,
+      "grad_norm": 0.0896064043045044,
+      "learning_rate": 1.0149190217527348e-05,
+      "loss": 8.6785,
+      "step": 132670
+    },
+    {
+      "epoch": 0.6625883292966116,
+      "grad_norm": 0.09562838077545166,
+      "learning_rate": 1.0147688302585798e-05,
+      "loss": 8.6778,
+      "step": 132680
+    },
+    {
+      "epoch": 0.6626382681215511,
+      "grad_norm": 0.08565637469291687,
+      "learning_rate": 1.0146186387644247e-05,
+      "loss": 8.6761,
+      "step": 132690
+    },
+    {
+      "epoch": 0.6626882069464906,
+      "grad_norm": 0.08810480684041977,
+      "learning_rate": 1.0144684472702695e-05,
+      "loss": 8.6872,
+      "step": 132700
+    },
+    {
+      "epoch": 0.66273814577143,
+      "grad_norm": 0.09386193752288818,
+      "learning_rate": 1.0143182557761145e-05,
+      "loss": 8.6697,
+      "step": 132710
+    },
+    {
+      "epoch": 0.6627880845963694,
+      "grad_norm": 0.08664627373218536,
+      "learning_rate": 1.0141680642819596e-05,
+      "loss": 8.6757,
+      "step": 132720
+    },
+    {
+      "epoch": 0.6628380234213089,
+      "grad_norm": 0.09146647155284882,
+      "learning_rate": 1.0140178727878046e-05,
+      "loss": 8.6781,
+      "step": 132730
+    },
+    {
+      "epoch": 0.6628879622462484,
+      "grad_norm": 0.09313146770000458,
+      "learning_rate": 1.0138676812936494e-05,
+      "loss": 8.6754,
+      "step": 132740
+    },
+    {
+      "epoch": 0.6629379010711878,
+      "grad_norm": 0.08888036757707596,
+      "learning_rate": 1.0137174897994943e-05,
+      "loss": 8.6916,
+      "step": 132750
+    },
+    {
+      "epoch": 0.6629878398961272,
+      "grad_norm": 0.09621097892522812,
+      "learning_rate": 1.0135672983053393e-05,
+      "loss": 8.6865,
+      "step": 132760
+    },
+    {
+      "epoch": 0.6630377787210667,
+      "grad_norm": 0.09131909161806107,
+      "learning_rate": 1.0134171068111843e-05,
+      "loss": 8.6703,
+      "step": 132770
+    },
+    {
+      "epoch": 0.6630877175460062,
+      "grad_norm": 0.09367981553077698,
+      "learning_rate": 1.0132669153170293e-05,
+      "loss": 8.6669,
+      "step": 132780
+    },
+    {
+      "epoch": 0.6631376563709456,
+      "grad_norm": 0.09454252570867538,
+      "learning_rate": 1.0131167238228742e-05,
+      "loss": 8.6777,
+      "step": 132790
+    },
+    {
+      "epoch": 0.663187595195885,
+      "grad_norm": 0.09207130968570709,
+      "learning_rate": 1.0129665323287192e-05,
+      "loss": 8.6612,
+      "step": 132800
+    },
+    {
+      "epoch": 0.6632375340208245,
+      "grad_norm": 0.08936728537082672,
+      "learning_rate": 1.012816340834564e-05,
+      "loss": 8.6755,
+      "step": 132810
+    },
+    {
+      "epoch": 0.663287472845764,
+      "grad_norm": 0.09480737149715424,
+      "learning_rate": 1.012666149340409e-05,
+      "loss": 8.6621,
+      "step": 132820
+    },
+    {
+      "epoch": 0.6633374116707034,
+      "grad_norm": 0.08680365234613419,
+      "learning_rate": 1.012515957846254e-05,
+      "loss": 8.6732,
+      "step": 132830
+    },
+    {
+      "epoch": 0.6633873504956428,
+      "grad_norm": 0.09133225679397583,
+      "learning_rate": 1.012365766352099e-05,
+      "loss": 8.6831,
+      "step": 132840
+    },
+    {
+      "epoch": 0.6634372893205823,
+      "grad_norm": 0.08942453563213348,
+      "learning_rate": 1.012215574857944e-05,
+      "loss": 8.6746,
+      "step": 132850
+    },
+    {
+      "epoch": 0.6634872281455217,
+      "grad_norm": 0.09342201054096222,
+      "learning_rate": 1.0120653833637888e-05,
+      "loss": 8.6761,
+      "step": 132860
+    },
+    {
+      "epoch": 0.6635371669704612,
+      "grad_norm": 0.09316600114107132,
+      "learning_rate": 1.0119151918696338e-05,
+      "loss": 8.6868,
+      "step": 132870
+    },
+    {
+      "epoch": 0.6635871057954006,
+      "grad_norm": 0.08866464346647263,
+      "learning_rate": 1.0117650003754788e-05,
+      "loss": 8.6818,
+      "step": 132880
+    },
+    {
+      "epoch": 0.6636370446203401,
+      "grad_norm": 0.09229955822229385,
+      "learning_rate": 1.0116148088813237e-05,
+      "loss": 8.6467,
+      "step": 132890
+    },
+    {
+      "epoch": 0.6636869834452795,
+      "grad_norm": 0.09364652633666992,
+      "learning_rate": 1.0114646173871687e-05,
+      "loss": 8.6595,
+      "step": 132900
+    },
+    {
+      "epoch": 0.663736922270219,
+      "grad_norm": 0.08931564539670944,
+      "learning_rate": 1.0113144258930135e-05,
+      "loss": 8.6613,
+      "step": 132910
+    },
+    {
+      "epoch": 0.6637868610951584,
+      "grad_norm": 0.09411066770553589,
+      "learning_rate": 1.0111642343988586e-05,
+      "loss": 8.6725,
+      "step": 132920
+    },
+    {
+      "epoch": 0.6638367999200979,
+      "grad_norm": 0.08930312842130661,
+      "learning_rate": 1.0110140429047036e-05,
+      "loss": 8.6819,
+      "step": 132930
+    },
+    {
+      "epoch": 0.6638867387450373,
+      "grad_norm": 0.10185947269201279,
+      "learning_rate": 1.0108638514105484e-05,
+      "loss": 8.6582,
+      "step": 132940
+    },
+    {
+      "epoch": 0.6639366775699768,
+      "grad_norm": 0.09503977745771408,
+      "learning_rate": 1.0107136599163934e-05,
+      "loss": 8.6604,
+      "step": 132950
+    },
+    {
+      "epoch": 0.6639866163949162,
+      "grad_norm": 0.08833954483270645,
+      "learning_rate": 1.0105634684222385e-05,
+      "loss": 8.6738,
+      "step": 132960
+    },
+    {
+      "epoch": 0.6640365552198557,
+      "grad_norm": 0.09546960145235062,
+      "learning_rate": 1.0104132769280833e-05,
+      "loss": 8.6512,
+      "step": 132970
+    },
+    {
+      "epoch": 0.6640864940447951,
+      "grad_norm": 0.089764803647995,
+      "learning_rate": 1.0102630854339283e-05,
+      "loss": 8.6673,
+      "step": 132980
+    },
+    {
+      "epoch": 0.6641364328697346,
+      "grad_norm": 0.09377101808786392,
+      "learning_rate": 1.0101128939397732e-05,
+      "loss": 8.6686,
+      "step": 132990
+    },
+    {
+      "epoch": 0.664186371694674,
+      "grad_norm": 0.0922766849398613,
+      "learning_rate": 1.0099627024456182e-05,
+      "loss": 8.6927,
+      "step": 133000
+    },
+    {
+      "epoch": 0.6642363105196135,
+      "grad_norm": 0.09250890463590622,
+      "learning_rate": 1.0098125109514632e-05,
+      "loss": 8.6572,
+      "step": 133010
+    },
+    {
+      "epoch": 0.6642862493445529,
+      "grad_norm": 0.09117847681045532,
+      "learning_rate": 1.009662319457308e-05,
+      "loss": 8.67,
+      "step": 133020
+    },
+    {
+      "epoch": 0.6643361881694924,
+      "grad_norm": 0.09532193839550018,
+      "learning_rate": 1.009512127963153e-05,
+      "loss": 8.6688,
+      "step": 133030
+    },
+    {
+      "epoch": 0.6643861269944318,
+      "grad_norm": 0.093431755900383,
+      "learning_rate": 1.009361936468998e-05,
+      "loss": 8.6796,
+      "step": 133040
+    },
+    {
+      "epoch": 0.6644360658193713,
+      "grad_norm": 0.09252150356769562,
+      "learning_rate": 1.009211744974843e-05,
+      "loss": 8.6743,
+      "step": 133050
+    },
+    {
+      "epoch": 0.6644860046443107,
+      "grad_norm": 0.10001608729362488,
+      "learning_rate": 1.009061553480688e-05,
+      "loss": 8.6727,
+      "step": 133060
+    },
+    {
+      "epoch": 0.6645359434692502,
+      "grad_norm": 0.09084002673625946,
+      "learning_rate": 1.0089113619865328e-05,
+      "loss": 8.6693,
+      "step": 133070
+    },
+    {
+      "epoch": 0.6645858822941896,
+      "grad_norm": 0.09234880656003952,
+      "learning_rate": 1.0087611704923778e-05,
+      "loss": 8.6681,
+      "step": 133080
+    },
+    {
+      "epoch": 0.6646358211191291,
+      "grad_norm": 0.09040534496307373,
+      "learning_rate": 1.0086109789982227e-05,
+      "loss": 8.659,
+      "step": 133090
+    },
+    {
+      "epoch": 0.6646857599440685,
+      "grad_norm": 0.0903211459517479,
+      "learning_rate": 1.0084607875040677e-05,
+      "loss": 8.6726,
+      "step": 133100
+    },
+    {
+      "epoch": 0.664735698769008,
+      "grad_norm": 0.08797595649957657,
+      "learning_rate": 1.0083105960099127e-05,
+      "loss": 8.676,
+      "step": 133110
+    },
+    {
+      "epoch": 0.6647856375939474,
+      "grad_norm": 0.0916309654712677,
+      "learning_rate": 1.0081604045157577e-05,
+      "loss": 8.676,
+      "step": 133120
+    },
+    {
+      "epoch": 0.6648355764188869,
+      "grad_norm": 0.08980787545442581,
+      "learning_rate": 1.0080102130216026e-05,
+      "loss": 8.6746,
+      "step": 133130
+    },
+    {
+      "epoch": 0.6648855152438263,
+      "grad_norm": 0.0977618619799614,
+      "learning_rate": 1.0078600215274474e-05,
+      "loss": 8.6623,
+      "step": 133140
+    },
+    {
+      "epoch": 0.6649354540687658,
+      "grad_norm": 0.08751973509788513,
+      "learning_rate": 1.0077098300332924e-05,
+      "loss": 8.6612,
+      "step": 133150
+    },
+    {
+      "epoch": 0.6649853928937052,
+      "grad_norm": 0.09779629111289978,
+      "learning_rate": 1.0075596385391375e-05,
+      "loss": 8.6694,
+      "step": 133160
+    },
+    {
+      "epoch": 0.6650353317186447,
+      "grad_norm": 0.0898473933339119,
+      "learning_rate": 1.0074094470449825e-05,
+      "loss": 8.6637,
+      "step": 133170
+    },
+    {
+      "epoch": 0.6650852705435841,
+      "grad_norm": 0.08798861503601074,
+      "learning_rate": 1.0072592555508273e-05,
+      "loss": 8.6467,
+      "step": 133180
+    },
+    {
+      "epoch": 0.6651352093685236,
+      "grad_norm": 0.09523648023605347,
+      "learning_rate": 1.0071090640566722e-05,
+      "loss": 8.6672,
+      "step": 133190
+    },
+    {
+      "epoch": 0.665185148193463,
+      "grad_norm": 0.08886650949716568,
+      "learning_rate": 1.0069588725625172e-05,
+      "loss": 8.6718,
+      "step": 133200
+    },
+    {
+      "epoch": 0.6652350870184025,
+      "grad_norm": 0.09928663820028305,
+      "learning_rate": 1.0068086810683622e-05,
+      "loss": 8.6745,
+      "step": 133210
+    },
+    {
+      "epoch": 0.6652850258433419,
+      "grad_norm": 0.09062928706407547,
+      "learning_rate": 1.0066584895742072e-05,
+      "loss": 8.6682,
+      "step": 133220
+    },
+    {
+      "epoch": 0.6653349646682813,
+      "grad_norm": 0.09397498518228531,
+      "learning_rate": 1.006508298080052e-05,
+      "loss": 8.6596,
+      "step": 133230
+    },
+    {
+      "epoch": 0.6653849034932208,
+      "grad_norm": 0.08882281184196472,
+      "learning_rate": 1.006358106585897e-05,
+      "loss": 8.6688,
+      "step": 133240
+    },
+    {
+      "epoch": 0.6654348423181603,
+      "grad_norm": 0.09210222959518433,
+      "learning_rate": 1.006207915091742e-05,
+      "loss": 8.663,
+      "step": 133250
+    },
+    {
+      "epoch": 0.6654847811430997,
+      "grad_norm": 0.08811397850513458,
+      "learning_rate": 1.006057723597587e-05,
+      "loss": 8.6646,
+      "step": 133260
+    },
+    {
+      "epoch": 0.6655347199680391,
+      "grad_norm": 0.08988697081804276,
+      "learning_rate": 1.005907532103432e-05,
+      "loss": 8.673,
+      "step": 133270
+    },
+    {
+      "epoch": 0.6655846587929786,
+      "grad_norm": 0.09060483425855637,
+      "learning_rate": 1.005757340609277e-05,
+      "loss": 8.691,
+      "step": 133280
+    },
+    {
+      "epoch": 0.6656345976179181,
+      "grad_norm": 0.09499598294496536,
+      "learning_rate": 1.0056071491151217e-05,
+      "loss": 8.6645,
+      "step": 133290
+    },
+    {
+      "epoch": 0.6656845364428575,
+      "grad_norm": 0.09347851574420929,
+      "learning_rate": 1.0054569576209667e-05,
+      "loss": 8.6652,
+      "step": 133300
+    },
+    {
+      "epoch": 0.6657344752677969,
+      "grad_norm": 0.09178951382637024,
+      "learning_rate": 1.0053067661268117e-05,
+      "loss": 8.6673,
+      "step": 133310
+    },
+    {
+      "epoch": 0.6657844140927364,
+      "grad_norm": 0.08929334580898285,
+      "learning_rate": 1.0051565746326567e-05,
+      "loss": 8.6643,
+      "step": 133320
+    },
+    {
+      "epoch": 0.6658343529176759,
+      "grad_norm": 0.08914398401975632,
+      "learning_rate": 1.0050063831385017e-05,
+      "loss": 8.6549,
+      "step": 133330
+    },
+    {
+      "epoch": 0.6658842917426153,
+      "grad_norm": 0.09610702097415924,
+      "learning_rate": 1.0048561916443464e-05,
+      "loss": 8.6626,
+      "step": 133340
+    },
+    {
+      "epoch": 0.6659342305675547,
+      "grad_norm": 0.09578754752874374,
+      "learning_rate": 1.0047060001501914e-05,
+      "loss": 8.6582,
+      "step": 133350
+    },
+    {
+      "epoch": 0.6659841693924942,
+      "grad_norm": 0.09582225978374481,
+      "learning_rate": 1.0045558086560365e-05,
+      "loss": 8.6685,
+      "step": 133360
+    },
+    {
+      "epoch": 0.6660341082174337,
+      "grad_norm": 0.08942214399576187,
+      "learning_rate": 1.0044056171618815e-05,
+      "loss": 8.6746,
+      "step": 133370
+    },
+    {
+      "epoch": 0.6660840470423731,
+      "grad_norm": 0.08953560143709183,
+      "learning_rate": 1.0042554256677265e-05,
+      "loss": 8.6745,
+      "step": 133380
+    },
+    {
+      "epoch": 0.6661339858673125,
+      "grad_norm": 0.0939660593867302,
+      "learning_rate": 1.0041052341735712e-05,
+      "loss": 8.6744,
+      "step": 133390
+    },
+    {
+      "epoch": 0.666183924692252,
+      "grad_norm": 0.09279930591583252,
+      "learning_rate": 1.0039550426794162e-05,
+      "loss": 8.6558,
+      "step": 133400
+    },
+    {
+      "epoch": 0.6662338635171915,
+      "grad_norm": 0.08562085032463074,
+      "learning_rate": 1.0038048511852612e-05,
+      "loss": 8.6662,
+      "step": 133410
+    },
+    {
+      "epoch": 0.6662838023421309,
+      "grad_norm": 0.09029558300971985,
+      "learning_rate": 1.0036546596911062e-05,
+      "loss": 8.6574,
+      "step": 133420
+    },
+    {
+      "epoch": 0.6663337411670703,
+      "grad_norm": 0.08761605620384216,
+      "learning_rate": 1.0035044681969512e-05,
+      "loss": 8.6631,
+      "step": 133430
+    },
+    {
+      "epoch": 0.6663836799920098,
+      "grad_norm": 0.09495789557695389,
+      "learning_rate": 1.0033542767027961e-05,
+      "loss": 8.66,
+      "step": 133440
+    },
+    {
+      "epoch": 0.6664336188169493,
+      "grad_norm": 0.09556636214256287,
+      "learning_rate": 1.003204085208641e-05,
+      "loss": 8.682,
+      "step": 133450
+    },
+    {
+      "epoch": 0.6664835576418887,
+      "grad_norm": 0.09460949152708054,
+      "learning_rate": 1.003053893714486e-05,
+      "loss": 8.6886,
+      "step": 133460
+    },
+    {
+      "epoch": 0.6665334964668281,
+      "grad_norm": 0.0882893055677414,
+      "learning_rate": 1.002903702220331e-05,
+      "loss": 8.668,
+      "step": 133470
+    },
+    {
+      "epoch": 0.6665834352917676,
+      "grad_norm": 0.09346653521060944,
+      "learning_rate": 1.002753510726176e-05,
+      "loss": 8.6562,
+      "step": 133480
+    },
+    {
+      "epoch": 0.6666333741167071,
+      "grad_norm": 0.09014210850000381,
+      "learning_rate": 1.0026033192320208e-05,
+      "loss": 8.6711,
+      "step": 133490
+    },
+    {
+      "epoch": 0.6666833129416465,
+      "grad_norm": 0.09201137721538544,
+      "learning_rate": 1.0024531277378657e-05,
+      "loss": 8.6789,
+      "step": 133500
+    },
+    {
+      "epoch": 0.6667332517665859,
+      "grad_norm": 0.09237250685691833,
+      "learning_rate": 1.0023029362437107e-05,
+      "loss": 8.6558,
+      "step": 133510
+    },
+    {
+      "epoch": 0.6667831905915254,
+      "grad_norm": 0.09040234982967377,
+      "learning_rate": 1.0021527447495557e-05,
+      "loss": 8.6683,
+      "step": 133520
+    },
+    {
+      "epoch": 0.6668331294164648,
+      "grad_norm": 0.10480639338493347,
+      "learning_rate": 1.0020025532554007e-05,
+      "loss": 8.6698,
+      "step": 133530
+    },
+    {
+      "epoch": 0.6668830682414043,
+      "grad_norm": 0.09072550386190414,
+      "learning_rate": 1.0018523617612456e-05,
+      "loss": 8.668,
+      "step": 133540
+    },
+    {
+      "epoch": 0.6669330070663437,
+      "grad_norm": 0.09083649516105652,
+      "learning_rate": 1.0017021702670904e-05,
+      "loss": 8.6629,
+      "step": 133550
+    },
+    {
+      "epoch": 0.6669829458912832,
+      "grad_norm": 0.08946739882230759,
+      "learning_rate": 1.0015519787729355e-05,
+      "loss": 8.6661,
+      "step": 133560
+    },
+    {
+      "epoch": 0.6670328847162226,
+      "grad_norm": 0.09026233851909637,
+      "learning_rate": 1.0014017872787805e-05,
+      "loss": 8.6714,
+      "step": 133570
+    },
+    {
+      "epoch": 0.6670828235411621,
+      "grad_norm": 0.09073863178491592,
+      "learning_rate": 1.0012515957846255e-05,
+      "loss": 8.6753,
+      "step": 133580
+    },
+    {
+      "epoch": 0.6671327623661015,
+      "grad_norm": 0.08761871606111526,
+      "learning_rate": 1.0011014042904703e-05,
+      "loss": 8.6689,
+      "step": 133590
+    },
+    {
+      "epoch": 0.667182701191041,
+      "grad_norm": 0.09583648294210434,
+      "learning_rate": 1.0009512127963154e-05,
+      "loss": 8.6691,
+      "step": 133600
+    },
+    {
+      "epoch": 0.6672326400159804,
+      "grad_norm": 0.09313435852527618,
+      "learning_rate": 1.0008010213021602e-05,
+      "loss": 8.6623,
+      "step": 133610
+    },
+    {
+      "epoch": 0.6672825788409199,
+      "grad_norm": 0.08645382523536682,
+      "learning_rate": 1.0006508298080052e-05,
+      "loss": 8.6729,
+      "step": 133620
+    },
+    {
+      "epoch": 0.6673325176658593,
+      "grad_norm": 0.09160168468952179,
+      "learning_rate": 1.0005006383138502e-05,
+      "loss": 8.6709,
+      "step": 133630
+    },
+    {
+      "epoch": 0.6673824564907987,
+      "grad_norm": 0.0923621729016304,
+      "learning_rate": 1.0003504468196951e-05,
+      "loss": 8.6641,
+      "step": 133640
+    },
+    {
+      "epoch": 0.6674323953157382,
+      "grad_norm": 0.09282460808753967,
+      "learning_rate": 1.0002002553255401e-05,
+      "loss": 8.6736,
+      "step": 133650
+    },
+    {
+      "epoch": 0.6674823341406777,
+      "grad_norm": 0.09043533354997635,
+      "learning_rate": 1.000050063831385e-05,
+      "loss": 8.6669,
+      "step": 133660
+    },
+    {
+      "epoch": 0.6675322729656171,
+      "grad_norm": 0.09773357957601547,
+      "learning_rate": 9.9989987233723e-06,
+      "loss": 8.6597,
+      "step": 133670
+    },
+    {
+      "epoch": 0.6675822117905565,
+      "grad_norm": 0.09863118827342987,
+      "learning_rate": 9.99749680843075e-06,
+      "loss": 8.683,
+      "step": 133680
+    },
+    {
+      "epoch": 0.667632150615496,
+      "grad_norm": 0.10197389125823975,
+      "learning_rate": 9.995994893489198e-06,
+      "loss": 8.6554,
+      "step": 133690
+    },
+    {
+      "epoch": 0.6676820894404355,
+      "grad_norm": 0.09325721859931946,
+      "learning_rate": 9.994492978547649e-06,
+      "loss": 8.6614,
+      "step": 133700
+    },
+    {
+      "epoch": 0.6677320282653749,
+      "grad_norm": 0.09466391056776047,
+      "learning_rate": 9.992991063606097e-06,
+      "loss": 8.6618,
+      "step": 133710
+    },
+    {
+      "epoch": 0.6677819670903143,
+      "grad_norm": 0.09134448319673538,
+      "learning_rate": 9.991489148664547e-06,
+      "loss": 8.6812,
+      "step": 133720
+    },
+    {
+      "epoch": 0.6678319059152538,
+      "grad_norm": 0.09619180858135223,
+      "learning_rate": 9.989987233722997e-06,
+      "loss": 8.6641,
+      "step": 133730
+    },
+    {
+      "epoch": 0.6678818447401933,
+      "grad_norm": 0.096577949821949,
+      "learning_rate": 9.988485318781446e-06,
+      "loss": 8.6647,
+      "step": 133740
+    },
+    {
+      "epoch": 0.6679317835651327,
+      "grad_norm": 0.08971861004829407,
+      "learning_rate": 9.986983403839896e-06,
+      "loss": 8.6594,
+      "step": 133750
+    },
+    {
+      "epoch": 0.6679817223900721,
+      "grad_norm": 0.08666064590215683,
+      "learning_rate": 9.985481488898346e-06,
+      "loss": 8.6657,
+      "step": 133760
+    },
+    {
+      "epoch": 0.6680316612150116,
+      "grad_norm": 0.09941823035478592,
+      "learning_rate": 9.983979573956795e-06,
+      "loss": 8.6705,
+      "step": 133770
+    },
+    {
+      "epoch": 0.6680816000399511,
+      "grad_norm": 0.09281333535909653,
+      "learning_rate": 9.982477659015245e-06,
+      "loss": 8.6561,
+      "step": 133780
+    },
+    {
+      "epoch": 0.6681315388648905,
+      "grad_norm": 0.09609360247850418,
+      "learning_rate": 9.980975744073693e-06,
+      "loss": 8.6606,
+      "step": 133790
+    },
+    {
+      "epoch": 0.6681814776898299,
+      "grad_norm": 0.09355857223272324,
+      "learning_rate": 9.979473829132144e-06,
+      "loss": 8.6826,
+      "step": 133800
+    },
+    {
+      "epoch": 0.6682314165147694,
+      "grad_norm": 0.09512100368738174,
+      "learning_rate": 9.977971914190594e-06,
+      "loss": 8.662,
+      "step": 133810
+    },
+    {
+      "epoch": 0.6682813553397089,
+      "grad_norm": 0.09662336111068726,
+      "learning_rate": 9.976469999249042e-06,
+      "loss": 8.6586,
+      "step": 133820
+    },
+    {
+      "epoch": 0.6683312941646483,
+      "grad_norm": 0.09392707049846649,
+      "learning_rate": 9.974968084307492e-06,
+      "loss": 8.6483,
+      "step": 133830
+    },
+    {
+      "epoch": 0.6683812329895877,
+      "grad_norm": 0.09166334569454193,
+      "learning_rate": 9.973466169365941e-06,
+      "loss": 8.6611,
+      "step": 133840
+    },
+    {
+      "epoch": 0.6684311718145272,
+      "grad_norm": 0.09149006009101868,
+      "learning_rate": 9.971964254424391e-06,
+      "loss": 8.653,
+      "step": 133850
+    },
+    {
+      "epoch": 0.6684811106394667,
+      "grad_norm": 0.09438381344079971,
+      "learning_rate": 9.970462339482841e-06,
+      "loss": 8.6613,
+      "step": 133860
+    },
+    {
+      "epoch": 0.6685310494644061,
+      "grad_norm": 0.08881976455450058,
+      "learning_rate": 9.96896042454129e-06,
+      "loss": 8.656,
+      "step": 133870
+    },
+    {
+      "epoch": 0.6685809882893455,
+      "grad_norm": 0.09529709070920944,
+      "learning_rate": 9.96745850959974e-06,
+      "loss": 8.6681,
+      "step": 133880
+    },
+    {
+      "epoch": 0.668630927114285,
+      "grad_norm": 0.09204521775245667,
+      "learning_rate": 9.965956594658188e-06,
+      "loss": 8.6722,
+      "step": 133890
+    },
+    {
+      "epoch": 0.6686808659392245,
+      "grad_norm": 0.09275458008050919,
+      "learning_rate": 9.964454679716639e-06,
+      "loss": 8.6581,
+      "step": 133900
+    },
+    {
+      "epoch": 0.6687308047641639,
+      "grad_norm": 0.09179842472076416,
+      "learning_rate": 9.962952764775089e-06,
+      "loss": 8.6784,
+      "step": 133910
+    },
+    {
+      "epoch": 0.6687807435891033,
+      "grad_norm": 0.09390482306480408,
+      "learning_rate": 9.961450849833539e-06,
+      "loss": 8.6541,
+      "step": 133920
+    },
+    {
+      "epoch": 0.6688306824140428,
+      "grad_norm": 0.09328165650367737,
+      "learning_rate": 9.959948934891988e-06,
+      "loss": 8.6696,
+      "step": 133930
+    },
+    {
+      "epoch": 0.6688806212389823,
+      "grad_norm": 0.09216855466365814,
+      "learning_rate": 9.958447019950436e-06,
+      "loss": 8.6686,
+      "step": 133940
+    },
+    {
+      "epoch": 0.6689305600639217,
+      "grad_norm": 0.09541112184524536,
+      "learning_rate": 9.956945105008886e-06,
+      "loss": 8.6747,
+      "step": 133950
+    },
+    {
+      "epoch": 0.6689804988888611,
+      "grad_norm": 0.0967298373579979,
+      "learning_rate": 9.955443190067336e-06,
+      "loss": 8.6533,
+      "step": 133960
+    },
+    {
+      "epoch": 0.6690304377138006,
+      "grad_norm": 0.09267246723175049,
+      "learning_rate": 9.953941275125787e-06,
+      "loss": 8.6666,
+      "step": 133970
+    },
+    {
+      "epoch": 0.6690803765387401,
+      "grad_norm": 0.09133971482515335,
+      "learning_rate": 9.952439360184235e-06,
+      "loss": 8.6595,
+      "step": 133980
+    },
+    {
+      "epoch": 0.6691303153636795,
+      "grad_norm": 0.0926765725016594,
+      "learning_rate": 9.950937445242685e-06,
+      "loss": 8.6628,
+      "step": 133990
+    },
+    {
+      "epoch": 0.6691802541886189,
+      "grad_norm": 0.09417456388473511,
+      "learning_rate": 9.949435530301134e-06,
+      "loss": 8.6531,
+      "step": 134000
+    },
+    {
+      "epoch": 0.6692301930135583,
+      "grad_norm": 0.09054078161716461,
+      "learning_rate": 9.947933615359584e-06,
+      "loss": 8.6442,
+      "step": 134010
+    },
+    {
+      "epoch": 0.6692801318384979,
+      "grad_norm": 0.09623897820711136,
+      "learning_rate": 9.946431700418034e-06,
+      "loss": 8.653,
+      "step": 134020
+    },
+    {
+      "epoch": 0.6693300706634373,
+      "grad_norm": 0.08746648579835892,
+      "learning_rate": 9.944929785476483e-06,
+      "loss": 8.6438,
+      "step": 134030
+    },
+    {
+      "epoch": 0.6693800094883767,
+      "grad_norm": 0.09206461906433105,
+      "learning_rate": 9.943427870534933e-06,
+      "loss": 8.6584,
+      "step": 134040
+    },
+    {
+      "epoch": 0.6694299483133161,
+      "grad_norm": 0.08908689767122269,
+      "learning_rate": 9.941925955593381e-06,
+      "loss": 8.6869,
+      "step": 134050
+    },
+    {
+      "epoch": 0.6694798871382557,
+      "grad_norm": 0.09292318671941757,
+      "learning_rate": 9.940424040651831e-06,
+      "loss": 8.6826,
+      "step": 134060
+    },
+    {
+      "epoch": 0.6695298259631951,
+      "grad_norm": 0.09243188053369522,
+      "learning_rate": 9.938922125710282e-06,
+      "loss": 8.6561,
+      "step": 134070
+    },
+    {
+      "epoch": 0.6695797647881345,
+      "grad_norm": 0.09017886966466904,
+      "learning_rate": 9.937420210768732e-06,
+      "loss": 8.6697,
+      "step": 134080
+    },
+    {
+      "epoch": 0.669629703613074,
+      "grad_norm": 0.09152711927890778,
+      "learning_rate": 9.93591829582718e-06,
+      "loss": 8.6725,
+      "step": 134090
+    },
+    {
+      "epoch": 0.6696796424380135,
+      "grad_norm": 0.08486898243427277,
+      "learning_rate": 9.934416380885629e-06,
+      "loss": 8.6743,
+      "step": 134100
+    },
+    {
+      "epoch": 0.6697295812629529,
+      "grad_norm": 0.09803474694490433,
+      "learning_rate": 9.932914465944079e-06,
+      "loss": 8.6401,
+      "step": 134110
+    },
+    {
+      "epoch": 0.6697795200878923,
+      "grad_norm": 0.08933472633361816,
+      "learning_rate": 9.931412551002529e-06,
+      "loss": 8.6808,
+      "step": 134120
+    },
+    {
+      "epoch": 0.6698294589128317,
+      "grad_norm": 0.09698875993490219,
+      "learning_rate": 9.92991063606098e-06,
+      "loss": 8.6689,
+      "step": 134130
+    },
+    {
+      "epoch": 0.6698793977377713,
+      "grad_norm": 0.08952021598815918,
+      "learning_rate": 9.928408721119428e-06,
+      "loss": 8.6691,
+      "step": 134140
+    },
+    {
+      "epoch": 0.6699293365627107,
+      "grad_norm": 0.09210740774869919,
+      "learning_rate": 9.926906806177876e-06,
+      "loss": 8.6496,
+      "step": 134150
+    },
+    {
+      "epoch": 0.6699792753876501,
+      "grad_norm": 0.09553104639053345,
+      "learning_rate": 9.925404891236326e-06,
+      "loss": 8.6525,
+      "step": 134160
+    },
+    {
+      "epoch": 0.6700292142125895,
+      "grad_norm": 0.09282413125038147,
+      "learning_rate": 9.923902976294777e-06,
+      "loss": 8.6403,
+      "step": 134170
+    },
+    {
+      "epoch": 0.6700791530375291,
+      "grad_norm": 0.08609344810247421,
+      "learning_rate": 9.922401061353227e-06,
+      "loss": 8.6725,
+      "step": 134180
+    },
+    {
+      "epoch": 0.6701290918624685,
+      "grad_norm": 0.09846524894237518,
+      "learning_rate": 9.920899146411675e-06,
+      "loss": 8.6636,
+      "step": 134190
+    },
+    {
+      "epoch": 0.6701790306874079,
+      "grad_norm": 0.09294971078634262,
+      "learning_rate": 9.919397231470124e-06,
+      "loss": 8.6753,
+      "step": 134200
+    },
+    {
+      "epoch": 0.6702289695123473,
+      "grad_norm": 0.08816470950841904,
+      "learning_rate": 9.917895316528574e-06,
+      "loss": 8.6784,
+      "step": 134210
+    },
+    {
+      "epoch": 0.6702789083372869,
+      "grad_norm": 0.08824525773525238,
+      "learning_rate": 9.916393401587024e-06,
+      "loss": 8.6515,
+      "step": 134220
+    },
+    {
+      "epoch": 0.6703288471622263,
+      "grad_norm": 0.09112218022346497,
+      "learning_rate": 9.914891486645474e-06,
+      "loss": 8.6579,
+      "step": 134230
+    },
+    {
+      "epoch": 0.6703787859871657,
+      "grad_norm": 0.09652066230773926,
+      "learning_rate": 9.913389571703924e-06,
+      "loss": 8.6431,
+      "step": 134240
+    },
+    {
+      "epoch": 0.6704287248121051,
+      "grad_norm": 0.09298358857631683,
+      "learning_rate": 9.911887656762371e-06,
+      "loss": 8.6533,
+      "step": 134250
+    },
+    {
+      "epoch": 0.6704786636370447,
+      "grad_norm": 0.09158873558044434,
+      "learning_rate": 9.910385741820821e-06,
+      "loss": 8.6822,
+      "step": 134260
+    },
+    {
+      "epoch": 0.6705286024619841,
+      "grad_norm": 0.09415313601493835,
+      "learning_rate": 9.908883826879272e-06,
+      "loss": 8.6575,
+      "step": 134270
+    },
+    {
+      "epoch": 0.6705785412869235,
+      "grad_norm": 0.09670121222734451,
+      "learning_rate": 9.907381911937722e-06,
+      "loss": 8.6453,
+      "step": 134280
+    },
+    {
+      "epoch": 0.6706284801118629,
+      "grad_norm": 0.08960152417421341,
+      "learning_rate": 9.905879996996172e-06,
+      "loss": 8.6497,
+      "step": 134290
+    },
+    {
+      "epoch": 0.6706784189368025,
+      "grad_norm": 0.09629824757575989,
+      "learning_rate": 9.904378082054619e-06,
+      "loss": 8.6611,
+      "step": 134300
+    },
+    {
+      "epoch": 0.6707283577617419,
+      "grad_norm": 0.09402245283126831,
+      "learning_rate": 9.902876167113069e-06,
+      "loss": 8.6787,
+      "step": 134310
+    },
+    {
+      "epoch": 0.6707782965866813,
+      "grad_norm": 0.09469646960496902,
+      "learning_rate": 9.901374252171519e-06,
+      "loss": 8.6602,
+      "step": 134320
+    },
+    {
+      "epoch": 0.6708282354116207,
+      "grad_norm": 0.09018561244010925,
+      "learning_rate": 9.89987233722997e-06,
+      "loss": 8.6598,
+      "step": 134330
+    },
+    {
+      "epoch": 0.6708781742365603,
+      "grad_norm": 0.0924052819609642,
+      "learning_rate": 9.89837042228842e-06,
+      "loss": 8.6522,
+      "step": 134340
+    },
+    {
+      "epoch": 0.6709281130614997,
+      "grad_norm": 0.09291255474090576,
+      "learning_rate": 9.896868507346866e-06,
+      "loss": 8.6561,
+      "step": 134350
+    },
+    {
+      "epoch": 0.6709780518864391,
+      "grad_norm": 0.09238367527723312,
+      "learning_rate": 9.895366592405316e-06,
+      "loss": 8.6519,
+      "step": 134360
+    },
+    {
+      "epoch": 0.6710279907113785,
+      "grad_norm": 0.10505057871341705,
+      "learning_rate": 9.893864677463767e-06,
+      "loss": 8.6528,
+      "step": 134370
+    },
+    {
+      "epoch": 0.6710779295363181,
+      "grad_norm": 0.09222082048654556,
+      "learning_rate": 9.892362762522217e-06,
+      "loss": 8.6566,
+      "step": 134380
+    },
+    {
+      "epoch": 0.6711278683612575,
+      "grad_norm": 0.09739550203084946,
+      "learning_rate": 9.890860847580667e-06,
+      "loss": 8.6633,
+      "step": 134390
+    },
+    {
+      "epoch": 0.6711778071861969,
+      "grad_norm": 0.09582085907459259,
+      "learning_rate": 9.889358932639115e-06,
+      "loss": 8.6723,
+      "step": 134400
+    },
+    {
+      "epoch": 0.6712277460111363,
+      "grad_norm": 0.09395405650138855,
+      "learning_rate": 9.887857017697564e-06,
+      "loss": 8.6626,
+      "step": 134410
+    },
+    {
+      "epoch": 0.6712776848360759,
+      "grad_norm": 0.09649001806974411,
+      "learning_rate": 9.886355102756014e-06,
+      "loss": 8.6701,
+      "step": 134420
+    },
+    {
+      "epoch": 0.6713276236610153,
+      "grad_norm": 0.09423769265413284,
+      "learning_rate": 9.884853187814464e-06,
+      "loss": 8.6763,
+      "step": 134430
+    },
+    {
+      "epoch": 0.6713775624859547,
+      "grad_norm": 0.09237778931856155,
+      "learning_rate": 9.883351272872914e-06,
+      "loss": 8.6705,
+      "step": 134440
+    },
+    {
+      "epoch": 0.6714275013108941,
+      "grad_norm": 0.08992139250040054,
+      "learning_rate": 9.881849357931363e-06,
+      "loss": 8.6781,
+      "step": 134450
+    },
+    {
+      "epoch": 0.6714774401358337,
+      "grad_norm": 0.09282557666301727,
+      "learning_rate": 9.880347442989811e-06,
+      "loss": 8.6552,
+      "step": 134460
+    },
+    {
+      "epoch": 0.6715273789607731,
+      "grad_norm": 0.0941321849822998,
+      "learning_rate": 9.878845528048262e-06,
+      "loss": 8.6635,
+      "step": 134470
+    },
+    {
+      "epoch": 0.6715773177857125,
+      "grad_norm": 0.09401953965425491,
+      "learning_rate": 9.877343613106712e-06,
+      "loss": 8.6542,
+      "step": 134480
+    },
+    {
+      "epoch": 0.6716272566106519,
+      "grad_norm": 0.0891314148902893,
+      "learning_rate": 9.875841698165162e-06,
+      "loss": 8.6772,
+      "step": 134490
+    },
+    {
+      "epoch": 0.6716771954355913,
+      "grad_norm": 0.09066881239414215,
+      "learning_rate": 9.87433978322361e-06,
+      "loss": 8.6553,
+      "step": 134500
+    },
+    {
+      "epoch": 0.6717271342605309,
+      "grad_norm": 0.09230073541402817,
+      "learning_rate": 9.872837868282059e-06,
+      "loss": 8.6493,
+      "step": 134510
+    },
+    {
+      "epoch": 0.6717770730854703,
+      "grad_norm": 0.09482811391353607,
+      "learning_rate": 9.871335953340509e-06,
+      "loss": 8.656,
+      "step": 134520
+    },
+    {
+      "epoch": 0.6718270119104097,
+      "grad_norm": 0.0987572893500328,
+      "learning_rate": 9.86983403839896e-06,
+      "loss": 8.6604,
+      "step": 134530
+    },
+    {
+      "epoch": 0.6718769507353491,
+      "grad_norm": 0.10041922330856323,
+      "learning_rate": 9.86833212345741e-06,
+      "loss": 8.6664,
+      "step": 134540
+    },
+    {
+      "epoch": 0.6719268895602887,
+      "grad_norm": 0.09294810891151428,
+      "learning_rate": 9.866830208515858e-06,
+      "loss": 8.6552,
+      "step": 134550
+    },
+    {
+      "epoch": 0.6719768283852281,
+      "grad_norm": 0.0915827602148056,
+      "learning_rate": 9.865328293574308e-06,
+      "loss": 8.668,
+      "step": 134560
+    },
+    {
+      "epoch": 0.6720267672101675,
+      "grad_norm": 0.09003446251153946,
+      "learning_rate": 9.863826378632757e-06,
+      "loss": 8.6777,
+      "step": 134570
+    },
+    {
+      "epoch": 0.6720767060351069,
+      "grad_norm": 0.09368298947811127,
+      "learning_rate": 9.862324463691207e-06,
+      "loss": 8.6627,
+      "step": 134580
+    },
+    {
+      "epoch": 0.6721266448600465,
+      "grad_norm": 0.09174887835979462,
+      "learning_rate": 9.860822548749657e-06,
+      "loss": 8.6605,
+      "step": 134590
+    },
+    {
+      "epoch": 0.6721765836849859,
+      "grad_norm": 0.09435159713029861,
+      "learning_rate": 9.859320633808105e-06,
+      "loss": 8.6576,
+      "step": 134600
+    },
+    {
+      "epoch": 0.6722265225099253,
+      "grad_norm": 0.09179011732339859,
+      "learning_rate": 9.857818718866556e-06,
+      "loss": 8.6462,
+      "step": 134610
+    },
+    {
+      "epoch": 0.6722764613348647,
+      "grad_norm": 0.09385015815496445,
+      "learning_rate": 9.856316803925004e-06,
+      "loss": 8.6683,
+      "step": 134620
+    },
+    {
+      "epoch": 0.6723264001598043,
+      "grad_norm": 0.09193267673254013,
+      "learning_rate": 9.854814888983454e-06,
+      "loss": 8.6526,
+      "step": 134630
+    },
+    {
+      "epoch": 0.6723763389847437,
+      "grad_norm": 0.09351233392953873,
+      "learning_rate": 9.853312974041904e-06,
+      "loss": 8.6577,
+      "step": 134640
+    },
+    {
+      "epoch": 0.6724262778096831,
+      "grad_norm": 0.09000106900930405,
+      "learning_rate": 9.851811059100353e-06,
+      "loss": 8.6783,
+      "step": 134650
+    },
+    {
+      "epoch": 0.6724762166346225,
+      "grad_norm": 0.09067153185606003,
+      "learning_rate": 9.850309144158803e-06,
+      "loss": 8.6576,
+      "step": 134660
+    },
+    {
+      "epoch": 0.6725261554595621,
+      "grad_norm": 0.0920468345284462,
+      "learning_rate": 9.848807229217252e-06,
+      "loss": 8.6682,
+      "step": 134670
+    },
+    {
+      "epoch": 0.6725760942845015,
+      "grad_norm": 0.09809605777263641,
+      "learning_rate": 9.847305314275702e-06,
+      "loss": 8.6612,
+      "step": 134680
+    },
+    {
+      "epoch": 0.6726260331094409,
+      "grad_norm": 0.1000465601682663,
+      "learning_rate": 9.845803399334152e-06,
+      "loss": 8.6516,
+      "step": 134690
+    },
+    {
+      "epoch": 0.6726759719343803,
+      "grad_norm": 0.08977311849594116,
+      "learning_rate": 9.8443014843926e-06,
+      "loss": 8.6702,
+      "step": 134700
+    },
+    {
+      "epoch": 0.6727259107593199,
+      "grad_norm": 0.09617611020803452,
+      "learning_rate": 9.84279956945105e-06,
+      "loss": 8.655,
+      "step": 134710
+    },
+    {
+      "epoch": 0.6727758495842593,
+      "grad_norm": 0.08458354324102402,
+      "learning_rate": 9.841297654509499e-06,
+      "loss": 8.6569,
+      "step": 134720
+    },
+    {
+      "epoch": 0.6728257884091987,
+      "grad_norm": 0.09260948747396469,
+      "learning_rate": 9.83979573956795e-06,
+      "loss": 8.6509,
+      "step": 134730
+    },
+    {
+      "epoch": 0.6728757272341381,
+      "grad_norm": 0.09007035940885544,
+      "learning_rate": 9.8382938246264e-06,
+      "loss": 8.6548,
+      "step": 134740
+    },
+    {
+      "epoch": 0.6729256660590777,
+      "grad_norm": 0.09251650422811508,
+      "learning_rate": 9.836791909684848e-06,
+      "loss": 8.6643,
+      "step": 134750
+    },
+    {
+      "epoch": 0.6729756048840171,
+      "grad_norm": 0.09078550338745117,
+      "learning_rate": 9.835289994743298e-06,
+      "loss": 8.6623,
+      "step": 134760
+    },
+    {
+      "epoch": 0.6730255437089565,
+      "grad_norm": 0.09649539738893509,
+      "learning_rate": 9.833788079801748e-06,
+      "loss": 8.6652,
+      "step": 134770
+    },
+    {
+      "epoch": 0.6730754825338959,
+      "grad_norm": 0.09006690233945847,
+      "learning_rate": 9.832286164860197e-06,
+      "loss": 8.6559,
+      "step": 134780
+    },
+    {
+      "epoch": 0.6731254213588355,
+      "grad_norm": 0.09151316434144974,
+      "learning_rate": 9.830784249918647e-06,
+      "loss": 8.656,
+      "step": 134790
+    },
+    {
+      "epoch": 0.6731753601837749,
+      "grad_norm": 0.08936766535043716,
+      "learning_rate": 9.829282334977095e-06,
+      "loss": 8.6594,
+      "step": 134800
+    },
+    {
+      "epoch": 0.6732252990087143,
+      "grad_norm": 0.09173012524843216,
+      "learning_rate": 9.827780420035546e-06,
+      "loss": 8.6488,
+      "step": 134810
+    },
+    {
+      "epoch": 0.6732752378336537,
+      "grad_norm": 0.09206842631101608,
+      "learning_rate": 9.826278505093996e-06,
+      "loss": 8.6666,
+      "step": 134820
+    },
+    {
+      "epoch": 0.6733251766585933,
+      "grad_norm": 0.08679156005382538,
+      "learning_rate": 9.824776590152444e-06,
+      "loss": 8.6556,
+      "step": 134830
+    },
+    {
+      "epoch": 0.6733751154835327,
+      "grad_norm": 0.08848954737186432,
+      "learning_rate": 9.823274675210894e-06,
+      "loss": 8.6569,
+      "step": 134840
+    },
+    {
+      "epoch": 0.6734250543084721,
+      "grad_norm": 0.09259720891714096,
+      "learning_rate": 9.821772760269343e-06,
+      "loss": 8.665,
+      "step": 134850
+    },
+    {
+      "epoch": 0.6734749931334115,
+      "grad_norm": 0.0927811861038208,
+      "learning_rate": 9.820270845327793e-06,
+      "loss": 8.6641,
+      "step": 134860
+    },
+    {
+      "epoch": 0.6735249319583511,
+      "grad_norm": 0.09524352103471756,
+      "learning_rate": 9.818768930386243e-06,
+      "loss": 8.6755,
+      "step": 134870
+    },
+    {
+      "epoch": 0.6735748707832905,
+      "grad_norm": 0.09407173842191696,
+      "learning_rate": 9.817267015444692e-06,
+      "loss": 8.6591,
+      "step": 134880
+    },
+    {
+      "epoch": 0.6736248096082299,
+      "grad_norm": 0.0950080081820488,
+      "learning_rate": 9.815765100503142e-06,
+      "loss": 8.6511,
+      "step": 134890
+    },
+    {
+      "epoch": 0.6736747484331693,
+      "grad_norm": 0.09280173480510712,
+      "learning_rate": 9.81426318556159e-06,
+      "loss": 8.6471,
+      "step": 134900
+    },
+    {
+      "epoch": 0.6737246872581089,
+      "grad_norm": 0.0905299037694931,
+      "learning_rate": 9.81276127062004e-06,
+      "loss": 8.6516,
+      "step": 134910
+    },
+    {
+      "epoch": 0.6737746260830483,
+      "grad_norm": 0.09357500821352005,
+      "learning_rate": 9.81125935567849e-06,
+      "loss": 8.668,
+      "step": 134920
+    },
+    {
+      "epoch": 0.6738245649079877,
+      "grad_norm": 0.09710787981748581,
+      "learning_rate": 9.809757440736941e-06,
+      "loss": 8.6576,
+      "step": 134930
+    },
+    {
+      "epoch": 0.6738745037329271,
+      "grad_norm": 0.08756748586893082,
+      "learning_rate": 9.80825552579539e-06,
+      "loss": 8.6721,
+      "step": 134940
+    },
+    {
+      "epoch": 0.6739244425578667,
+      "grad_norm": 0.08993040025234222,
+      "learning_rate": 9.806753610853838e-06,
+      "loss": 8.6743,
+      "step": 134950
+    },
+    {
+      "epoch": 0.6739743813828061,
+      "grad_norm": 0.0880391076207161,
+      "learning_rate": 9.805251695912288e-06,
+      "loss": 8.6597,
+      "step": 134960
+    },
+    {
+      "epoch": 0.6740243202077455,
+      "grad_norm": 0.08911752700805664,
+      "learning_rate": 9.803749780970738e-06,
+      "loss": 8.6669,
+      "step": 134970
+    },
+    {
+      "epoch": 0.6740742590326849,
+      "grad_norm": 0.09046872705221176,
+      "learning_rate": 9.802247866029188e-06,
+      "loss": 8.6558,
+      "step": 134980
+    },
+    {
+      "epoch": 0.6741241978576245,
+      "grad_norm": 0.09054312109947205,
+      "learning_rate": 9.800745951087637e-06,
+      "loss": 8.6595,
+      "step": 134990
+    },
+    {
+      "epoch": 0.6741741366825639,
+      "grad_norm": 0.09408404678106308,
+      "learning_rate": 9.799244036146085e-06,
+      "loss": 8.6439,
+      "step": 135000
+    },
+    {
+      "epoch": 0.6742240755075033,
+      "grad_norm": 0.08936314284801483,
+      "learning_rate": 9.797742121204536e-06,
+      "loss": 8.6506,
+      "step": 135010
+    },
+    {
+      "epoch": 0.6742740143324427,
+      "grad_norm": 0.09435812383890152,
+      "learning_rate": 9.796240206262986e-06,
+      "loss": 8.6472,
+      "step": 135020
+    },
+    {
+      "epoch": 0.6743239531573822,
+      "grad_norm": 0.09421943128108978,
+      "learning_rate": 9.794738291321436e-06,
+      "loss": 8.6749,
+      "step": 135030
+    },
+    {
+      "epoch": 0.6743738919823217,
+      "grad_norm": 0.08769576996564865,
+      "learning_rate": 9.793236376379884e-06,
+      "loss": 8.6708,
+      "step": 135040
+    },
+    {
+      "epoch": 0.6744238308072611,
+      "grad_norm": 0.09175736457109451,
+      "learning_rate": 9.791734461438333e-06,
+      "loss": 8.6579,
+      "step": 135050
+    },
+    {
+      "epoch": 0.6744737696322005,
+      "grad_norm": 0.08920390158891678,
+      "learning_rate": 9.790232546496783e-06,
+      "loss": 8.6663,
+      "step": 135060
+    },
+    {
+      "epoch": 0.67452370845714,
+      "grad_norm": 0.09278859943151474,
+      "learning_rate": 9.788730631555233e-06,
+      "loss": 8.6719,
+      "step": 135070
+    },
+    {
+      "epoch": 0.6745736472820795,
+      "grad_norm": 0.090042345225811,
+      "learning_rate": 9.787228716613683e-06,
+      "loss": 8.6571,
+      "step": 135080
+    },
+    {
+      "epoch": 0.6746235861070189,
+      "grad_norm": 0.09226316958665848,
+      "learning_rate": 9.785726801672134e-06,
+      "loss": 8.6539,
+      "step": 135090
+    },
+    {
+      "epoch": 0.6746735249319583,
+      "grad_norm": 0.09357789903879166,
+      "learning_rate": 9.78422488673058e-06,
+      "loss": 8.6673,
+      "step": 135100
+    },
+    {
+      "epoch": 0.6747234637568978,
+      "grad_norm": 0.09317333251237869,
+      "learning_rate": 9.78272297178903e-06,
+      "loss": 8.6531,
+      "step": 135110
+    },
+    {
+      "epoch": 0.6747734025818373,
+      "grad_norm": 0.09152083098888397,
+      "learning_rate": 9.78122105684748e-06,
+      "loss": 8.6667,
+      "step": 135120
+    },
+    {
+      "epoch": 0.6748233414067767,
+      "grad_norm": 0.0933714434504509,
+      "learning_rate": 9.779719141905931e-06,
+      "loss": 8.6673,
+      "step": 135130
+    },
+    {
+      "epoch": 0.6748732802317161,
+      "grad_norm": 0.09301735460758209,
+      "learning_rate": 9.778217226964381e-06,
+      "loss": 8.6561,
+      "step": 135140
+    },
+    {
+      "epoch": 0.6749232190566556,
+      "grad_norm": 0.0876002088189125,
+      "learning_rate": 9.776715312022828e-06,
+      "loss": 8.6553,
+      "step": 135150
+    },
+    {
+      "epoch": 0.6749731578815951,
+      "grad_norm": 0.09405035525560379,
+      "learning_rate": 9.775213397081278e-06,
+      "loss": 8.6746,
+      "step": 135160
+    },
+    {
+      "epoch": 0.6750230967065345,
+      "grad_norm": 0.09140407294034958,
+      "learning_rate": 9.773711482139728e-06,
+      "loss": 8.6653,
+      "step": 135170
+    },
+    {
+      "epoch": 0.6750730355314739,
+      "grad_norm": 0.08978252112865448,
+      "learning_rate": 9.772209567198178e-06,
+      "loss": 8.6645,
+      "step": 135180
+    },
+    {
+      "epoch": 0.6751229743564134,
+      "grad_norm": 0.0867505818605423,
+      "learning_rate": 9.770707652256629e-06,
+      "loss": 8.6649,
+      "step": 135190
+    },
+    {
+      "epoch": 0.6751729131813529,
+      "grad_norm": 0.08967334032058716,
+      "learning_rate": 9.769205737315075e-06,
+      "loss": 8.6698,
+      "step": 135200
+    },
+    {
+      "epoch": 0.6752228520062923,
+      "grad_norm": 0.09012636542320251,
+      "learning_rate": 9.767703822373526e-06,
+      "loss": 8.6562,
+      "step": 135210
+    },
+    {
+      "epoch": 0.6752727908312317,
+      "grad_norm": 0.0897894874215126,
+      "learning_rate": 9.766201907431976e-06,
+      "loss": 8.6699,
+      "step": 135220
+    },
+    {
+      "epoch": 0.6753227296561712,
+      "grad_norm": 0.09098558127880096,
+      "learning_rate": 9.764699992490426e-06,
+      "loss": 8.6482,
+      "step": 135230
+    },
+    {
+      "epoch": 0.6753726684811107,
+      "grad_norm": 0.09297525882720947,
+      "learning_rate": 9.763198077548876e-06,
+      "loss": 8.6506,
+      "step": 135240
+    },
+    {
+      "epoch": 0.6754226073060501,
+      "grad_norm": 0.09668441861867905,
+      "learning_rate": 9.761696162607325e-06,
+      "loss": 8.6461,
+      "step": 135250
+    },
+    {
+      "epoch": 0.6754725461309895,
+      "grad_norm": 0.09092225879430771,
+      "learning_rate": 9.760194247665773e-06,
+      "loss": 8.6631,
+      "step": 135260
+    },
+    {
+      "epoch": 0.675522484955929,
+      "grad_norm": 0.09417429566383362,
+      "learning_rate": 9.758692332724223e-06,
+      "loss": 8.6589,
+      "step": 135270
+    },
+    {
+      "epoch": 0.6755724237808685,
+      "grad_norm": 0.08669702708721161,
+      "learning_rate": 9.757190417782673e-06,
+      "loss": 8.6646,
+      "step": 135280
+    },
+    {
+      "epoch": 0.6756223626058079,
+      "grad_norm": 0.08657342940568924,
+      "learning_rate": 9.755688502841124e-06,
+      "loss": 8.6499,
+      "step": 135290
+    },
+    {
+      "epoch": 0.6756723014307473,
+      "grad_norm": 0.09297830611467361,
+      "learning_rate": 9.754186587899572e-06,
+      "loss": 8.6483,
+      "step": 135300
+    },
+    {
+      "epoch": 0.6757222402556868,
+      "grad_norm": 0.09091796725988388,
+      "learning_rate": 9.75268467295802e-06,
+      "loss": 8.6687,
+      "step": 135310
+    },
+    {
+      "epoch": 0.6757721790806263,
+      "grad_norm": 0.09649965912103653,
+      "learning_rate": 9.75118275801647e-06,
+      "loss": 8.6546,
+      "step": 135320
+    },
+    {
+      "epoch": 0.6758221179055657,
+      "grad_norm": 0.09378094226121902,
+      "learning_rate": 9.749680843074921e-06,
+      "loss": 8.6654,
+      "step": 135330
+    },
+    {
+      "epoch": 0.6758720567305051,
+      "grad_norm": 0.09173280000686646,
+      "learning_rate": 9.748178928133371e-06,
+      "loss": 8.6516,
+      "step": 135340
+    },
+    {
+      "epoch": 0.6759219955554446,
+      "grad_norm": 0.096292644739151,
+      "learning_rate": 9.74667701319182e-06,
+      "loss": 8.6403,
+      "step": 135350
+    },
+    {
+      "epoch": 0.675971934380384,
+      "grad_norm": 0.09011993557214737,
+      "learning_rate": 9.745175098250268e-06,
+      "loss": 8.6538,
+      "step": 135360
+    },
+    {
+      "epoch": 0.6760218732053235,
+      "grad_norm": 0.09252528101205826,
+      "learning_rate": 9.743673183308718e-06,
+      "loss": 8.6566,
+      "step": 135370
+    },
+    {
+      "epoch": 0.6760718120302629,
+      "grad_norm": 0.09614871442317963,
+      "learning_rate": 9.742171268367168e-06,
+      "loss": 8.6462,
+      "step": 135380
+    },
+    {
+      "epoch": 0.6761217508552024,
+      "grad_norm": 0.09508166462182999,
+      "learning_rate": 9.740669353425619e-06,
+      "loss": 8.6597,
+      "step": 135390
+    },
+    {
+      "epoch": 0.6761716896801419,
+      "grad_norm": 0.09378086030483246,
+      "learning_rate": 9.739167438484067e-06,
+      "loss": 8.6674,
+      "step": 135400
+    },
+    {
+      "epoch": 0.6762216285050813,
+      "grad_norm": 0.08668700605630875,
+      "learning_rate": 9.737665523542517e-06,
+      "loss": 8.6488,
+      "step": 135410
+    },
+    {
+      "epoch": 0.6762715673300207,
+      "grad_norm": 0.09363897889852524,
+      "learning_rate": 9.736163608600966e-06,
+      "loss": 8.6639,
+      "step": 135420
+    },
+    {
+      "epoch": 0.6763215061549602,
+      "grad_norm": 0.0928199365735054,
+      "learning_rate": 9.734661693659416e-06,
+      "loss": 8.6574,
+      "step": 135430
+    },
+    {
+      "epoch": 0.6763714449798996,
+      "grad_norm": 0.09100401401519775,
+      "learning_rate": 9.733159778717866e-06,
+      "loss": 8.6541,
+      "step": 135440
+    },
+    {
+      "epoch": 0.6764213838048391,
+      "grad_norm": 0.08921308815479279,
+      "learning_rate": 9.731657863776315e-06,
+      "loss": 8.6332,
+      "step": 135450
+    },
+    {
+      "epoch": 0.6764713226297785,
+      "grad_norm": 0.0926818996667862,
+      "learning_rate": 9.730155948834765e-06,
+      "loss": 8.6395,
+      "step": 135460
+    },
+    {
+      "epoch": 0.676521261454718,
+      "grad_norm": 0.08852003514766693,
+      "learning_rate": 9.728654033893213e-06,
+      "loss": 8.6492,
+      "step": 135470
+    },
+    {
+      "epoch": 0.6765712002796574,
+      "grad_norm": 0.08820042759180069,
+      "learning_rate": 9.727152118951664e-06,
+      "loss": 8.6549,
+      "step": 135480
+    },
+    {
+      "epoch": 0.6766211391045969,
+      "grad_norm": 0.09674821048974991,
+      "learning_rate": 9.725650204010114e-06,
+      "loss": 8.6498,
+      "step": 135490
+    },
+    {
+      "epoch": 0.6766710779295363,
+      "grad_norm": 0.0904177725315094,
+      "learning_rate": 9.724148289068562e-06,
+      "loss": 8.6776,
+      "step": 135500
+    },
+    {
+      "epoch": 0.6767210167544757,
+      "grad_norm": 0.09356199204921722,
+      "learning_rate": 9.722646374127012e-06,
+      "loss": 8.6619,
+      "step": 135510
+    },
+    {
+      "epoch": 0.6767709555794152,
+      "grad_norm": 0.09460005164146423,
+      "learning_rate": 9.72114445918546e-06,
+      "loss": 8.6478,
+      "step": 135520
+    },
+    {
+      "epoch": 0.6768208944043547,
+      "grad_norm": 0.09244546294212341,
+      "learning_rate": 9.719642544243911e-06,
+      "loss": 8.6778,
+      "step": 135530
+    },
+    {
+      "epoch": 0.6768708332292941,
+      "grad_norm": 0.09258837252855301,
+      "learning_rate": 9.718140629302361e-06,
+      "loss": 8.6654,
+      "step": 135540
+    },
+    {
+      "epoch": 0.6769207720542335,
+      "grad_norm": 0.08935635536909103,
+      "learning_rate": 9.71663871436081e-06,
+      "loss": 8.6699,
+      "step": 135550
+    },
+    {
+      "epoch": 0.676970710879173,
+      "grad_norm": 0.10672944039106369,
+      "learning_rate": 9.71513679941926e-06,
+      "loss": 8.6493,
+      "step": 135560
+    },
+    {
+      "epoch": 0.6770206497041125,
+      "grad_norm": 0.09256184101104736,
+      "learning_rate": 9.71363488447771e-06,
+      "loss": 8.6716,
+      "step": 135570
+    },
+    {
+      "epoch": 0.6770705885290519,
+      "grad_norm": 0.09515663981437683,
+      "learning_rate": 9.712132969536159e-06,
+      "loss": 8.6385,
+      "step": 135580
+    },
+    {
+      "epoch": 0.6771205273539913,
+      "grad_norm": 0.09241361916065216,
+      "learning_rate": 9.710631054594609e-06,
+      "loss": 8.6506,
+      "step": 135590
+    },
+    {
+      "epoch": 0.6771704661789308,
+      "grad_norm": 0.08679981529712677,
+      "learning_rate": 9.709129139653057e-06,
+      "loss": 8.6791,
+      "step": 135600
+    },
+    {
+      "epoch": 0.6772204050038703,
+      "grad_norm": 0.08532244712114334,
+      "learning_rate": 9.707627224711507e-06,
+      "loss": 8.671,
+      "step": 135610
+    },
+    {
+      "epoch": 0.6772703438288097,
+      "grad_norm": 0.08982454240322113,
+      "learning_rate": 9.706125309769958e-06,
+      "loss": 8.6516,
+      "step": 135620
+    },
+    {
+      "epoch": 0.6773202826537491,
+      "grad_norm": 0.10029303282499313,
+      "learning_rate": 9.704623394828406e-06,
+      "loss": 8.6435,
+      "step": 135630
+    },
+    {
+      "epoch": 0.6773702214786886,
+      "grad_norm": 0.09623725712299347,
+      "learning_rate": 9.703121479886856e-06,
+      "loss": 8.6662,
+      "step": 135640
+    },
+    {
+      "epoch": 0.6774201603036281,
+      "grad_norm": 0.08999323099851608,
+      "learning_rate": 9.701619564945305e-06,
+      "loss": 8.6676,
+      "step": 135650
+    },
+    {
+      "epoch": 0.6774700991285675,
+      "grad_norm": 0.09232403337955475,
+      "learning_rate": 9.700117650003755e-06,
+      "loss": 8.6596,
+      "step": 135660
+    },
+    {
+      "epoch": 0.6775200379535069,
+      "grad_norm": 0.08921338617801666,
+      "learning_rate": 9.698615735062205e-06,
+      "loss": 8.6421,
+      "step": 135670
+    },
+    {
+      "epoch": 0.6775699767784464,
+      "grad_norm": 0.09304777532815933,
+      "learning_rate": 9.697113820120654e-06,
+      "loss": 8.6406,
+      "step": 135680
+    },
+    {
+      "epoch": 0.6776199156033859,
+      "grad_norm": 0.08728094398975372,
+      "learning_rate": 9.695611905179104e-06,
+      "loss": 8.6578,
+      "step": 135690
+    },
+    {
+      "epoch": 0.6776698544283253,
+      "grad_norm": 0.09113137423992157,
+      "learning_rate": 9.694109990237552e-06,
+      "loss": 8.6582,
+      "step": 135700
+    },
+    {
+      "epoch": 0.6777197932532647,
+      "grad_norm": 0.09369809180498123,
+      "learning_rate": 9.692608075296002e-06,
+      "loss": 8.6454,
+      "step": 135710
+    },
+    {
+      "epoch": 0.6777697320782042,
+      "grad_norm": 0.0949154943227768,
+      "learning_rate": 9.691106160354453e-06,
+      "loss": 8.646,
+      "step": 135720
+    },
+    {
+      "epoch": 0.6778196709031437,
+      "grad_norm": 0.09159260988235474,
+      "learning_rate": 9.689604245412903e-06,
+      "loss": 8.6647,
+      "step": 135730
+    },
+    {
+      "epoch": 0.6778696097280831,
+      "grad_norm": 0.08925732225179672,
+      "learning_rate": 9.688102330471351e-06,
+      "loss": 8.6396,
+      "step": 135740
+    },
+    {
+      "epoch": 0.6779195485530225,
+      "grad_norm": 0.10024773329496384,
+      "learning_rate": 9.6866004155298e-06,
+      "loss": 8.6457,
+      "step": 135750
+    },
+    {
+      "epoch": 0.677969487377962,
+      "grad_norm": 0.08866682648658752,
+      "learning_rate": 9.68509850058825e-06,
+      "loss": 8.662,
+      "step": 135760
+    },
+    {
+      "epoch": 0.6780194262029015,
+      "grad_norm": 0.09140399843454361,
+      "learning_rate": 9.6835965856467e-06,
+      "loss": 8.6452,
+      "step": 135770
+    },
+    {
+      "epoch": 0.6780693650278409,
+      "grad_norm": 0.09286113828420639,
+      "learning_rate": 9.68209467070515e-06,
+      "loss": 8.6551,
+      "step": 135780
+    },
+    {
+      "epoch": 0.6781193038527803,
+      "grad_norm": 0.09333377331495285,
+      "learning_rate": 9.680592755763599e-06,
+      "loss": 8.6632,
+      "step": 135790
+    },
+    {
+      "epoch": 0.6781692426777198,
+      "grad_norm": 0.09056348353624344,
+      "learning_rate": 9.679090840822047e-06,
+      "loss": 8.6587,
+      "step": 135800
+    },
+    {
+      "epoch": 0.6782191815026593,
+      "grad_norm": 0.09084860980510712,
+      "learning_rate": 9.677588925880497e-06,
+      "loss": 8.664,
+      "step": 135810
+    },
+    {
+      "epoch": 0.6782691203275987,
+      "grad_norm": 0.09165407717227936,
+      "learning_rate": 9.676087010938948e-06,
+      "loss": 8.6647,
+      "step": 135820
+    },
+    {
+      "epoch": 0.6783190591525381,
+      "grad_norm": 0.08495034277439117,
+      "learning_rate": 9.674585095997398e-06,
+      "loss": 8.6464,
+      "step": 135830
+    },
+    {
+      "epoch": 0.6783689979774776,
+      "grad_norm": 0.09273646771907806,
+      "learning_rate": 9.673083181055846e-06,
+      "loss": 8.6369,
+      "step": 135840
+    },
+    {
+      "epoch": 0.678418936802417,
+      "grad_norm": 0.09448003023862839,
+      "learning_rate": 9.671581266114295e-06,
+      "loss": 8.6468,
+      "step": 135850
+    },
+    {
+      "epoch": 0.6784688756273565,
+      "grad_norm": 0.08844837546348572,
+      "learning_rate": 9.670079351172745e-06,
+      "loss": 8.6591,
+      "step": 135860
+    },
+    {
+      "epoch": 0.6785188144522959,
+      "grad_norm": 0.09456723183393478,
+      "learning_rate": 9.668577436231195e-06,
+      "loss": 8.6537,
+      "step": 135870
+    },
+    {
+      "epoch": 0.6785687532772354,
+      "grad_norm": 0.0897020697593689,
+      "learning_rate": 9.667075521289645e-06,
+      "loss": 8.6545,
+      "step": 135880
+    },
+    {
+      "epoch": 0.6786186921021748,
+      "grad_norm": 0.09378873556852341,
+      "learning_rate": 9.665573606348095e-06,
+      "loss": 8.6497,
+      "step": 135890
+    },
+    {
+      "epoch": 0.6786686309271143,
+      "grad_norm": 0.09122031182050705,
+      "learning_rate": 9.664071691406542e-06,
+      "loss": 8.6599,
+      "step": 135900
+    },
+    {
+      "epoch": 0.6787185697520537,
+      "grad_norm": 0.0962730273604393,
+      "learning_rate": 9.662569776464992e-06,
+      "loss": 8.6579,
+      "step": 135910
+    },
+    {
+      "epoch": 0.6787685085769932,
+      "grad_norm": 0.0921352207660675,
+      "learning_rate": 9.661067861523443e-06,
+      "loss": 8.6433,
+      "step": 135920
+    },
+    {
+      "epoch": 0.6788184474019326,
+      "grad_norm": 0.09395617246627808,
+      "learning_rate": 9.659565946581893e-06,
+      "loss": 8.6507,
+      "step": 135930
+    },
+    {
+      "epoch": 0.6788683862268721,
+      "grad_norm": 0.08777940273284912,
+      "learning_rate": 9.658064031640343e-06,
+      "loss": 8.6626,
+      "step": 135940
+    },
+    {
+      "epoch": 0.6789183250518115,
+      "grad_norm": 0.08930377662181854,
+      "learning_rate": 9.65656211669879e-06,
+      "loss": 8.6573,
+      "step": 135950
+    },
+    {
+      "epoch": 0.678968263876751,
+      "grad_norm": 0.09742255508899689,
+      "learning_rate": 9.65506020175724e-06,
+      "loss": 8.6381,
+      "step": 135960
+    },
+    {
+      "epoch": 0.6790182027016904,
+      "grad_norm": 0.09233038872480392,
+      "learning_rate": 9.65355828681569e-06,
+      "loss": 8.6535,
+      "step": 135970
+    },
+    {
+      "epoch": 0.6790681415266299,
+      "grad_norm": 0.08856892585754395,
+      "learning_rate": 9.65205637187414e-06,
+      "loss": 8.6476,
+      "step": 135980
+    },
+    {
+      "epoch": 0.6791180803515693,
+      "grad_norm": 0.09412689507007599,
+      "learning_rate": 9.65055445693259e-06,
+      "loss": 8.6462,
+      "step": 135990
+    },
+    {
+      "epoch": 0.6791680191765088,
+      "grad_norm": 0.09102386981248856,
+      "learning_rate": 9.649052541991037e-06,
+      "loss": 8.6614,
+      "step": 136000
+    },
+    {
+      "epoch": 0.6792179580014482,
+      "grad_norm": 0.0882454663515091,
+      "learning_rate": 9.647550627049487e-06,
+      "loss": 8.6684,
+      "step": 136010
+    },
+    {
+      "epoch": 0.6792678968263877,
+      "grad_norm": 0.09218038618564606,
+      "learning_rate": 9.646048712107938e-06,
+      "loss": 8.6619,
+      "step": 136020
+    },
+    {
+      "epoch": 0.6793178356513271,
+      "grad_norm": 0.09152485430240631,
+      "learning_rate": 9.644546797166388e-06,
+      "loss": 8.6552,
+      "step": 136030
+    },
+    {
+      "epoch": 0.6793677744762666,
+      "grad_norm": 0.08703062683343887,
+      "learning_rate": 9.643044882224838e-06,
+      "loss": 8.6545,
+      "step": 136040
+    },
+    {
+      "epoch": 0.679417713301206,
+      "grad_norm": 0.09743312001228333,
+      "learning_rate": 9.641542967283286e-06,
+      "loss": 8.6342,
+      "step": 136050
+    },
+    {
+      "epoch": 0.6794676521261455,
+      "grad_norm": 0.09214362502098083,
+      "learning_rate": 9.640041052341735e-06,
+      "loss": 8.6579,
+      "step": 136060
+    },
+    {
+      "epoch": 0.6795175909510849,
+      "grad_norm": 0.0931442603468895,
+      "learning_rate": 9.638539137400185e-06,
+      "loss": 8.6461,
+      "step": 136070
+    },
+    {
+      "epoch": 0.6795675297760244,
+      "grad_norm": 0.09083132445812225,
+      "learning_rate": 9.637037222458635e-06,
+      "loss": 8.6503,
+      "step": 136080
+    },
+    {
+      "epoch": 0.6796174686009638,
+      "grad_norm": 0.0905727818608284,
+      "learning_rate": 9.635535307517085e-06,
+      "loss": 8.6579,
+      "step": 136090
+    },
+    {
+      "epoch": 0.6796674074259033,
+      "grad_norm": 0.09426701813936234,
+      "learning_rate": 9.634033392575534e-06,
+      "loss": 8.6577,
+      "step": 136100
+    },
+    {
+      "epoch": 0.6797173462508427,
+      "grad_norm": 0.0878579169511795,
+      "learning_rate": 9.632531477633982e-06,
+      "loss": 8.6501,
+      "step": 136110
+    },
+    {
+      "epoch": 0.6797672850757822,
+      "grad_norm": 0.0943881943821907,
+      "learning_rate": 9.631029562692433e-06,
+      "loss": 8.6596,
+      "step": 136120
+    },
+    {
+      "epoch": 0.6798172239007216,
+      "grad_norm": 0.0909527838230133,
+      "learning_rate": 9.629527647750883e-06,
+      "loss": 8.6589,
+      "step": 136130
+    },
+    {
+      "epoch": 0.6798671627256611,
+      "grad_norm": 0.09404680877923965,
+      "learning_rate": 9.628025732809333e-06,
+      "loss": 8.6575,
+      "step": 136140
+    },
+    {
+      "epoch": 0.6799171015506005,
+      "grad_norm": 0.09321132302284241,
+      "learning_rate": 9.626523817867781e-06,
+      "loss": 8.6519,
+      "step": 136150
+    },
+    {
+      "epoch": 0.67996704037554,
+      "grad_norm": 0.09817219525575638,
+      "learning_rate": 9.62502190292623e-06,
+      "loss": 8.6566,
+      "step": 136160
+    },
+    {
+      "epoch": 0.6800169792004794,
+      "grad_norm": 0.09167230874300003,
+      "learning_rate": 9.62351998798468e-06,
+      "loss": 8.6576,
+      "step": 136170
+    },
+    {
+      "epoch": 0.6800669180254189,
+      "grad_norm": 0.09567620605230331,
+      "learning_rate": 9.62201807304313e-06,
+      "loss": 8.6495,
+      "step": 136180
+    },
+    {
+      "epoch": 0.6801168568503583,
+      "grad_norm": 0.0942583829164505,
+      "learning_rate": 9.62051615810158e-06,
+      "loss": 8.664,
+      "step": 136190
+    },
+    {
+      "epoch": 0.6801667956752978,
+      "grad_norm": 0.08843056857585907,
+      "learning_rate": 9.619014243160029e-06,
+      "loss": 8.6652,
+      "step": 136200
+    },
+    {
+      "epoch": 0.6802167345002372,
+      "grad_norm": 0.08930647373199463,
+      "learning_rate": 9.617512328218479e-06,
+      "loss": 8.6719,
+      "step": 136210
+    },
+    {
+      "epoch": 0.6802666733251767,
+      "grad_norm": 0.09439358115196228,
+      "learning_rate": 9.616010413276928e-06,
+      "loss": 8.6578,
+      "step": 136220
+    },
+    {
+      "epoch": 0.6803166121501161,
+      "grad_norm": 0.08645504713058472,
+      "learning_rate": 9.614508498335378e-06,
+      "loss": 8.661,
+      "step": 136230
+    },
+    {
+      "epoch": 0.6803665509750556,
+      "grad_norm": 0.0910206213593483,
+      "learning_rate": 9.613006583393828e-06,
+      "loss": 8.6595,
+      "step": 136240
+    },
+    {
+      "epoch": 0.680416489799995,
+      "grad_norm": 0.09109845012426376,
+      "learning_rate": 9.611504668452278e-06,
+      "loss": 8.6505,
+      "step": 136250
+    },
+    {
+      "epoch": 0.6804664286249344,
+      "grad_norm": 0.09436528384685516,
+      "learning_rate": 9.610002753510727e-06,
+      "loss": 8.6472,
+      "step": 136260
+    },
+    {
+      "epoch": 0.6805163674498739,
+      "grad_norm": 0.09152334183454514,
+      "learning_rate": 9.608500838569175e-06,
+      "loss": 8.6531,
+      "step": 136270
+    },
+    {
+      "epoch": 0.6805663062748134,
+      "grad_norm": 0.09130854159593582,
+      "learning_rate": 9.606998923627625e-06,
+      "loss": 8.6697,
+      "step": 136280
+    },
+    {
+      "epoch": 0.6806162450997528,
+      "grad_norm": 0.08787249028682709,
+      "learning_rate": 9.605497008686075e-06,
+      "loss": 8.6474,
+      "step": 136290
+    },
+    {
+      "epoch": 0.6806661839246922,
+      "grad_norm": 0.09564139693975449,
+      "learning_rate": 9.603995093744526e-06,
+      "loss": 8.6475,
+      "step": 136300
+    },
+    {
+      "epoch": 0.6807161227496317,
+      "grad_norm": 0.08936180919408798,
+      "learning_rate": 9.602493178802974e-06,
+      "loss": 8.6639,
+      "step": 136310
+    },
+    {
+      "epoch": 0.6807660615745712,
+      "grad_norm": 0.09855446964502335,
+      "learning_rate": 9.600991263861423e-06,
+      "loss": 8.673,
+      "step": 136320
+    },
+    {
+      "epoch": 0.6808160003995106,
+      "grad_norm": 0.08871414512395859,
+      "learning_rate": 9.599489348919873e-06,
+      "loss": 8.6471,
+      "step": 136330
+    },
+    {
+      "epoch": 0.68086593922445,
+      "grad_norm": 0.09087369590997696,
+      "learning_rate": 9.597987433978323e-06,
+      "loss": 8.6577,
+      "step": 136340
+    },
+    {
+      "epoch": 0.6809158780493895,
+      "grad_norm": 0.09292785823345184,
+      "learning_rate": 9.596485519036773e-06,
+      "loss": 8.6358,
+      "step": 136350
+    },
+    {
+      "epoch": 0.680965816874329,
+      "grad_norm": 0.09038669615983963,
+      "learning_rate": 9.594983604095222e-06,
+      "loss": 8.6422,
+      "step": 136360
+    },
+    {
+      "epoch": 0.6810157556992684,
+      "grad_norm": 0.09672816097736359,
+      "learning_rate": 9.593481689153672e-06,
+      "loss": 8.6498,
+      "step": 136370
+    },
+    {
+      "epoch": 0.6810656945242078,
+      "grad_norm": 0.09116402268409729,
+      "learning_rate": 9.59197977421212e-06,
+      "loss": 8.6377,
+      "step": 136380
+    },
+    {
+      "epoch": 0.6811156333491473,
+      "grad_norm": 0.09753582626581192,
+      "learning_rate": 9.59047785927057e-06,
+      "loss": 8.6643,
+      "step": 136390
+    },
+    {
+      "epoch": 0.6811655721740868,
+      "grad_norm": 0.0920112207531929,
+      "learning_rate": 9.58897594432902e-06,
+      "loss": 8.6488,
+      "step": 136400
+    },
+    {
+      "epoch": 0.6812155109990262,
+      "grad_norm": 0.09360447525978088,
+      "learning_rate": 9.587474029387469e-06,
+      "loss": 8.6664,
+      "step": 136410
+    },
+    {
+      "epoch": 0.6812654498239656,
+      "grad_norm": 0.08685506880283356,
+      "learning_rate": 9.58597211444592e-06,
+      "loss": 8.6444,
+      "step": 136420
+    },
+    {
+      "epoch": 0.6813153886489051,
+      "grad_norm": 0.08979588001966476,
+      "learning_rate": 9.584470199504368e-06,
+      "loss": 8.6541,
+      "step": 136430
+    },
+    {
+      "epoch": 0.6813653274738446,
+      "grad_norm": 0.09099925309419632,
+      "learning_rate": 9.582968284562818e-06,
+      "loss": 8.6319,
+      "step": 136440
+    },
+    {
+      "epoch": 0.681415266298784,
+      "grad_norm": 0.09031230956315994,
+      "learning_rate": 9.581466369621268e-06,
+      "loss": 8.6513,
+      "step": 136450
+    },
+    {
+      "epoch": 0.6814652051237234,
+      "grad_norm": 0.08769738674163818,
+      "learning_rate": 9.579964454679717e-06,
+      "loss": 8.6562,
+      "step": 136460
+    },
+    {
+      "epoch": 0.6815151439486629,
+      "grad_norm": 0.08992310613393784,
+      "learning_rate": 9.578462539738167e-06,
+      "loss": 8.6546,
+      "step": 136470
+    },
+    {
+      "epoch": 0.6815650827736023,
+      "grad_norm": 0.09138327836990356,
+      "learning_rate": 9.576960624796615e-06,
+      "loss": 8.6545,
+      "step": 136480
+    },
+    {
+      "epoch": 0.6816150215985418,
+      "grad_norm": 0.08947155624628067,
+      "learning_rate": 9.575458709855065e-06,
+      "loss": 8.6435,
+      "step": 136490
+    },
+    {
+      "epoch": 0.6816649604234812,
+      "grad_norm": 0.09748796373605728,
+      "learning_rate": 9.573956794913516e-06,
+      "loss": 8.6491,
+      "step": 136500
+    },
+    {
+      "epoch": 0.6817148992484207,
+      "grad_norm": 0.0930284857749939,
+      "learning_rate": 9.572454879971964e-06,
+      "loss": 8.6502,
+      "step": 136510
+    },
+    {
+      "epoch": 0.6817648380733601,
+      "grad_norm": 0.08821598440408707,
+      "learning_rate": 9.570952965030414e-06,
+      "loss": 8.6452,
+      "step": 136520
+    },
+    {
+      "epoch": 0.6818147768982996,
+      "grad_norm": 0.09385675936937332,
+      "learning_rate": 9.569451050088864e-06,
+      "loss": 8.6533,
+      "step": 136530
+    },
+    {
+      "epoch": 0.681864715723239,
+      "grad_norm": 0.09572584182024002,
+      "learning_rate": 9.567949135147313e-06,
+      "loss": 8.6635,
+      "step": 136540
+    },
+    {
+      "epoch": 0.6819146545481785,
+      "grad_norm": 0.09196435660123825,
+      "learning_rate": 9.566447220205763e-06,
+      "loss": 8.6576,
+      "step": 136550
+    },
+    {
+      "epoch": 0.6819645933731179,
+      "grad_norm": 0.09118195623159409,
+      "learning_rate": 9.564945305264212e-06,
+      "loss": 8.6366,
+      "step": 136560
+    },
+    {
+      "epoch": 0.6820145321980574,
+      "grad_norm": 0.09549646824598312,
+      "learning_rate": 9.563443390322662e-06,
+      "loss": 8.6665,
+      "step": 136570
+    },
+    {
+      "epoch": 0.6820644710229968,
+      "grad_norm": 0.09187420457601547,
+      "learning_rate": 9.561941475381112e-06,
+      "loss": 8.6431,
+      "step": 136580
+    },
+    {
+      "epoch": 0.6821144098479363,
+      "grad_norm": 0.09515494108200073,
+      "learning_rate": 9.56043956043956e-06,
+      "loss": 8.6565,
+      "step": 136590
+    },
+    {
+      "epoch": 0.6821643486728757,
+      "grad_norm": 0.09291938692331314,
+      "learning_rate": 9.55893764549801e-06,
+      "loss": 8.6536,
+      "step": 136600
+    },
+    {
+      "epoch": 0.6822142874978152,
+      "grad_norm": 0.09135958552360535,
+      "learning_rate": 9.557435730556459e-06,
+      "loss": 8.6686,
+      "step": 136610
+    },
+    {
+      "epoch": 0.6822642263227546,
+      "grad_norm": 0.09079675376415253,
+      "learning_rate": 9.55593381561491e-06,
+      "loss": 8.6562,
+      "step": 136620
+    },
+    {
+      "epoch": 0.682314165147694,
+      "grad_norm": 0.08791366219520569,
+      "learning_rate": 9.55443190067336e-06,
+      "loss": 8.6422,
+      "step": 136630
+    },
+    {
+      "epoch": 0.6823641039726335,
+      "grad_norm": 0.09008821099996567,
+      "learning_rate": 9.552929985731808e-06,
+      "loss": 8.6484,
+      "step": 136640
+    },
+    {
+      "epoch": 0.682414042797573,
+      "grad_norm": 0.09008509665727615,
+      "learning_rate": 9.551428070790258e-06,
+      "loss": 8.6609,
+      "step": 136650
+    },
+    {
+      "epoch": 0.6824639816225124,
+      "grad_norm": 0.0959354117512703,
+      "learning_rate": 9.549926155848707e-06,
+      "loss": 8.6284,
+      "step": 136660
+    },
+    {
+      "epoch": 0.6825139204474518,
+      "grad_norm": 0.08918027579784393,
+      "learning_rate": 9.548424240907157e-06,
+      "loss": 8.6587,
+      "step": 136670
+    },
+    {
+      "epoch": 0.6825638592723913,
+      "grad_norm": 0.08706361055374146,
+      "learning_rate": 9.546922325965607e-06,
+      "loss": 8.6813,
+      "step": 136680
+    },
+    {
+      "epoch": 0.6826137980973308,
+      "grad_norm": 0.09403030574321747,
+      "learning_rate": 9.545420411024057e-06,
+      "loss": 8.6521,
+      "step": 136690
+    },
+    {
+      "epoch": 0.6826637369222702,
+      "grad_norm": 0.08769270032644272,
+      "learning_rate": 9.543918496082506e-06,
+      "loss": 8.6526,
+      "step": 136700
+    },
+    {
+      "epoch": 0.6827136757472096,
+      "grad_norm": 0.08837845176458359,
+      "learning_rate": 9.542416581140954e-06,
+      "loss": 8.647,
+      "step": 136710
+    },
+    {
+      "epoch": 0.6827636145721491,
+      "grad_norm": 0.08815280348062515,
+      "learning_rate": 9.540914666199404e-06,
+      "loss": 8.6532,
+      "step": 136720
+    },
+    {
+      "epoch": 0.6828135533970886,
+      "grad_norm": 0.09271814674139023,
+      "learning_rate": 9.539412751257854e-06,
+      "loss": 8.6624,
+      "step": 136730
+    },
+    {
+      "epoch": 0.682863492222028,
+      "grad_norm": 0.09003327786922455,
+      "learning_rate": 9.537910836316305e-06,
+      "loss": 8.6456,
+      "step": 136740
+    },
+    {
+      "epoch": 0.6829134310469674,
+      "grad_norm": 0.09336933493614197,
+      "learning_rate": 9.536408921374753e-06,
+      "loss": 8.6364,
+      "step": 136750
+    },
+    {
+      "epoch": 0.6829633698719069,
+      "grad_norm": 0.08961906284093857,
+      "learning_rate": 9.534907006433202e-06,
+      "loss": 8.653,
+      "step": 136760
+    },
+    {
+      "epoch": 0.6830133086968464,
+      "grad_norm": 0.08953448385000229,
+      "learning_rate": 9.533405091491652e-06,
+      "loss": 8.6516,
+      "step": 136770
+    },
+    {
+      "epoch": 0.6830632475217858,
+      "grad_norm": 0.09705983847379684,
+      "learning_rate": 9.531903176550102e-06,
+      "loss": 8.6433,
+      "step": 136780
+    },
+    {
+      "epoch": 0.6831131863467252,
+      "grad_norm": 0.0961066335439682,
+      "learning_rate": 9.530401261608552e-06,
+      "loss": 8.6596,
+      "step": 136790
+    },
+    {
+      "epoch": 0.6831631251716647,
+      "grad_norm": 0.09489046037197113,
+      "learning_rate": 9.528899346667e-06,
+      "loss": 8.6587,
+      "step": 136800
+    },
+    {
+      "epoch": 0.6832130639966042,
+      "grad_norm": 0.0909392461180687,
+      "learning_rate": 9.527397431725449e-06,
+      "loss": 8.6567,
+      "step": 136810
+    },
+    {
+      "epoch": 0.6832630028215436,
+      "grad_norm": 0.09276154637336731,
+      "learning_rate": 9.5258955167839e-06,
+      "loss": 8.6676,
+      "step": 136820
+    },
+    {
+      "epoch": 0.683312941646483,
+      "grad_norm": 0.08885812014341354,
+      "learning_rate": 9.52439360184235e-06,
+      "loss": 8.6517,
+      "step": 136830
+    },
+    {
+      "epoch": 0.6833628804714225,
+      "grad_norm": 0.0934206172823906,
+      "learning_rate": 9.5228916869008e-06,
+      "loss": 8.6566,
+      "step": 136840
+    },
+    {
+      "epoch": 0.683412819296362,
+      "grad_norm": 0.08924253284931183,
+      "learning_rate": 9.52138977195925e-06,
+      "loss": 8.6622,
+      "step": 136850
+    },
+    {
+      "epoch": 0.6834627581213014,
+      "grad_norm": 0.0887620598077774,
+      "learning_rate": 9.519887857017697e-06,
+      "loss": 8.6523,
+      "step": 136860
+    },
+    {
+      "epoch": 0.6835126969462408,
+      "grad_norm": 0.09116550534963608,
+      "learning_rate": 9.518385942076147e-06,
+      "loss": 8.6463,
+      "step": 136870
+    },
+    {
+      "epoch": 0.6835626357711803,
+      "grad_norm": 0.09384380280971527,
+      "learning_rate": 9.516884027134597e-06,
+      "loss": 8.6518,
+      "step": 136880
+    },
+    {
+      "epoch": 0.6836125745961198,
+      "grad_norm": 0.08999653905630112,
+      "learning_rate": 9.515382112193047e-06,
+      "loss": 8.6534,
+      "step": 136890
+    },
+    {
+      "epoch": 0.6836625134210592,
+      "grad_norm": 0.09472544491291046,
+      "learning_rate": 9.513880197251497e-06,
+      "loss": 8.6418,
+      "step": 136900
+    },
+    {
+      "epoch": 0.6837124522459986,
+      "grad_norm": 0.08867818862199783,
+      "learning_rate": 9.512378282309944e-06,
+      "loss": 8.6511,
+      "step": 136910
+    },
+    {
+      "epoch": 0.6837623910709381,
+      "grad_norm": 0.09118511527776718,
+      "learning_rate": 9.510876367368394e-06,
+      "loss": 8.6535,
+      "step": 136920
+    },
+    {
+      "epoch": 0.6838123298958776,
+      "grad_norm": 0.09027484059333801,
+      "learning_rate": 9.509374452426844e-06,
+      "loss": 8.6652,
+      "step": 136930
+    },
+    {
+      "epoch": 0.683862268720817,
+      "grad_norm": 0.09857519716024399,
+      "learning_rate": 9.507872537485295e-06,
+      "loss": 8.6434,
+      "step": 136940
+    },
+    {
+      "epoch": 0.6839122075457564,
+      "grad_norm": 0.09504196047782898,
+      "learning_rate": 9.506370622543745e-06,
+      "loss": 8.6584,
+      "step": 136950
+    },
+    {
+      "epoch": 0.6839621463706959,
+      "grad_norm": 0.09877480566501617,
+      "learning_rate": 9.504868707602192e-06,
+      "loss": 8.6523,
+      "step": 136960
+    },
+    {
+      "epoch": 0.6840120851956354,
+      "grad_norm": 0.09247778356075287,
+      "learning_rate": 9.503366792660642e-06,
+      "loss": 8.6504,
+      "step": 136970
+    },
+    {
+      "epoch": 0.6840620240205748,
+      "grad_norm": 0.09239480644464493,
+      "learning_rate": 9.501864877719092e-06,
+      "loss": 8.6505,
+      "step": 136980
+    },
+    {
+      "epoch": 0.6841119628455142,
+      "grad_norm": 0.09122372418642044,
+      "learning_rate": 9.500362962777542e-06,
+      "loss": 8.6395,
+      "step": 136990
+    },
+    {
+      "epoch": 0.6841619016704537,
+      "grad_norm": 0.09300727397203445,
+      "learning_rate": 9.498861047835992e-06,
+      "loss": 8.6488,
+      "step": 137000
+    },
+    {
+      "epoch": 0.6842118404953932,
+      "grad_norm": 0.0911245048046112,
+      "learning_rate": 9.49735913289444e-06,
+      "loss": 8.6665,
+      "step": 137010
+    },
+    {
+      "epoch": 0.6842617793203326,
+      "grad_norm": 0.0955081507563591,
+      "learning_rate": 9.49585721795289e-06,
+      "loss": 8.6382,
+      "step": 137020
+    },
+    {
+      "epoch": 0.684311718145272,
+      "grad_norm": 0.09284482151269913,
+      "learning_rate": 9.49435530301134e-06,
+      "loss": 8.6343,
+      "step": 137030
+    },
+    {
+      "epoch": 0.6843616569702115,
+      "grad_norm": 0.09098115563392639,
+      "learning_rate": 9.49285338806979e-06,
+      "loss": 8.6491,
+      "step": 137040
+    },
+    {
+      "epoch": 0.684411595795151,
+      "grad_norm": 0.09565536677837372,
+      "learning_rate": 9.49135147312824e-06,
+      "loss": 8.6573,
+      "step": 137050
+    },
+    {
+      "epoch": 0.6844615346200904,
+      "grad_norm": 0.0895000547170639,
+      "learning_rate": 9.489849558186688e-06,
+      "loss": 8.6537,
+      "step": 137060
+    },
+    {
+      "epoch": 0.6845114734450298,
+      "grad_norm": 0.09270433336496353,
+      "learning_rate": 9.488347643245137e-06,
+      "loss": 8.6309,
+      "step": 137070
+    },
+    {
+      "epoch": 0.6845614122699692,
+      "grad_norm": 0.09027011692523956,
+      "learning_rate": 9.486845728303587e-06,
+      "loss": 8.6541,
+      "step": 137080
+    },
+    {
+      "epoch": 0.6846113510949088,
+      "grad_norm": 0.08946502953767776,
+      "learning_rate": 9.485343813362037e-06,
+      "loss": 8.6389,
+      "step": 137090
+    },
+    {
+      "epoch": 0.6846612899198482,
+      "grad_norm": 0.09274662286043167,
+      "learning_rate": 9.483841898420487e-06,
+      "loss": 8.6507,
+      "step": 137100
+    },
+    {
+      "epoch": 0.6847112287447876,
+      "grad_norm": 0.09150737524032593,
+      "learning_rate": 9.482339983478936e-06,
+      "loss": 8.6462,
+      "step": 137110
+    },
+    {
+      "epoch": 0.684761167569727,
+      "grad_norm": 0.09016991406679153,
+      "learning_rate": 9.480838068537384e-06,
+      "loss": 8.662,
+      "step": 137120
+    },
+    {
+      "epoch": 0.6848111063946666,
+      "grad_norm": 0.09275011718273163,
+      "learning_rate": 9.479336153595835e-06,
+      "loss": 8.6538,
+      "step": 137130
+    },
+    {
+      "epoch": 0.684861045219606,
+      "grad_norm": 0.09211241453886032,
+      "learning_rate": 9.477834238654285e-06,
+      "loss": 8.6423,
+      "step": 137140
+    },
+    {
+      "epoch": 0.6849109840445454,
+      "grad_norm": 0.09323053061962128,
+      "learning_rate": 9.476332323712735e-06,
+      "loss": 8.6311,
+      "step": 137150
+    },
+    {
+      "epoch": 0.6849609228694848,
+      "grad_norm": 0.0909585952758789,
+      "learning_rate": 9.474830408771183e-06,
+      "loss": 8.6413,
+      "step": 137160
+    },
+    {
+      "epoch": 0.6850108616944244,
+      "grad_norm": 0.09803111851215363,
+      "learning_rate": 9.473328493829634e-06,
+      "loss": 8.6599,
+      "step": 137170
+    },
+    {
+      "epoch": 0.6850608005193638,
+      "grad_norm": 0.09382762759923935,
+      "learning_rate": 9.471826578888082e-06,
+      "loss": 8.6395,
+      "step": 137180
+    },
+    {
+      "epoch": 0.6851107393443032,
+      "grad_norm": 0.09193086624145508,
+      "learning_rate": 9.470324663946532e-06,
+      "loss": 8.6454,
+      "step": 137190
+    },
+    {
+      "epoch": 0.6851606781692426,
+      "grad_norm": 0.0866667628288269,
+      "learning_rate": 9.468822749004982e-06,
+      "loss": 8.6586,
+      "step": 137200
+    },
+    {
+      "epoch": 0.6852106169941822,
+      "grad_norm": 0.09309227019548416,
+      "learning_rate": 9.467320834063431e-06,
+      "loss": 8.6427,
+      "step": 137210
+    },
+    {
+      "epoch": 0.6852605558191216,
+      "grad_norm": 0.091177798807621,
+      "learning_rate": 9.465818919121881e-06,
+      "loss": 8.6524,
+      "step": 137220
+    },
+    {
+      "epoch": 0.685310494644061,
+      "grad_norm": 0.09202263504266739,
+      "learning_rate": 9.46431700418033e-06,
+      "loss": 8.6487,
+      "step": 137230
+    },
+    {
+      "epoch": 0.6853604334690004,
+      "grad_norm": 0.0925859585404396,
+      "learning_rate": 9.46281508923878e-06,
+      "loss": 8.6511,
+      "step": 137240
+    },
+    {
+      "epoch": 0.68541037229394,
+      "grad_norm": 0.09529491513967514,
+      "learning_rate": 9.46131317429723e-06,
+      "loss": 8.6417,
+      "step": 137250
+    },
+    {
+      "epoch": 0.6854603111188794,
+      "grad_norm": 0.09451477974653244,
+      "learning_rate": 9.459811259355678e-06,
+      "loss": 8.6519,
+      "step": 137260
+    },
+    {
+      "epoch": 0.6855102499438188,
+      "grad_norm": 0.08990958333015442,
+      "learning_rate": 9.458309344414129e-06,
+      "loss": 8.647,
+      "step": 137270
+    },
+    {
+      "epoch": 0.6855601887687582,
+      "grad_norm": 0.09311163425445557,
+      "learning_rate": 9.456807429472577e-06,
+      "loss": 8.642,
+      "step": 137280
+    },
+    {
+      "epoch": 0.6856101275936978,
+      "grad_norm": 0.09261354058980942,
+      "learning_rate": 9.455305514531027e-06,
+      "loss": 8.6528,
+      "step": 137290
+    },
+    {
+      "epoch": 0.6856600664186372,
+      "grad_norm": 0.09306151419878006,
+      "learning_rate": 9.453803599589477e-06,
+      "loss": 8.6436,
+      "step": 137300
+    },
+    {
+      "epoch": 0.6857100052435766,
+      "grad_norm": 0.09042923152446747,
+      "learning_rate": 9.452301684647926e-06,
+      "loss": 8.6518,
+      "step": 137310
+    },
+    {
+      "epoch": 0.685759944068516,
+      "grad_norm": 0.09723861515522003,
+      "learning_rate": 9.450799769706376e-06,
+      "loss": 8.6463,
+      "step": 137320
+    },
+    {
+      "epoch": 0.6858098828934556,
+      "grad_norm": 0.08681852370500565,
+      "learning_rate": 9.449297854764826e-06,
+      "loss": 8.6446,
+      "step": 137330
+    },
+    {
+      "epoch": 0.685859821718395,
+      "grad_norm": 0.09019467979669571,
+      "learning_rate": 9.447795939823275e-06,
+      "loss": 8.6482,
+      "step": 137340
+    },
+    {
+      "epoch": 0.6859097605433344,
+      "grad_norm": 0.08701999485492706,
+      "learning_rate": 9.446294024881725e-06,
+      "loss": 8.6666,
+      "step": 137350
+    },
+    {
+      "epoch": 0.6859596993682738,
+      "grad_norm": 0.09151817858219147,
+      "learning_rate": 9.444792109940173e-06,
+      "loss": 8.6528,
+      "step": 137360
+    },
+    {
+      "epoch": 0.6860096381932134,
+      "grad_norm": 0.0913306251168251,
+      "learning_rate": 9.443290194998624e-06,
+      "loss": 8.6475,
+      "step": 137370
+    },
+    {
+      "epoch": 0.6860595770181528,
+      "grad_norm": 0.09225058555603027,
+      "learning_rate": 9.441788280057074e-06,
+      "loss": 8.6478,
+      "step": 137380
+    },
+    {
+      "epoch": 0.6861095158430922,
+      "grad_norm": 0.09292393922805786,
+      "learning_rate": 9.440286365115522e-06,
+      "loss": 8.6571,
+      "step": 137390
+    },
+    {
+      "epoch": 0.6861594546680316,
+      "grad_norm": 0.09331595152616501,
+      "learning_rate": 9.438784450173972e-06,
+      "loss": 8.6414,
+      "step": 137400
+    },
+    {
+      "epoch": 0.6862093934929712,
+      "grad_norm": 0.09516488760709763,
+      "learning_rate": 9.437282535232421e-06,
+      "loss": 8.6387,
+      "step": 137410
+    },
+    {
+      "epoch": 0.6862593323179106,
+      "grad_norm": 0.09467773884534836,
+      "learning_rate": 9.435780620290871e-06,
+      "loss": 8.6409,
+      "step": 137420
+    },
+    {
+      "epoch": 0.68630927114285,
+      "grad_norm": 0.09190910309553146,
+      "learning_rate": 9.434278705349321e-06,
+      "loss": 8.6683,
+      "step": 137430
+    },
+    {
+      "epoch": 0.6863592099677894,
+      "grad_norm": 0.09423499554395676,
+      "learning_rate": 9.43277679040777e-06,
+      "loss": 8.6736,
+      "step": 137440
+    },
+    {
+      "epoch": 0.686409148792729,
+      "grad_norm": 0.0923381969332695,
+      "learning_rate": 9.43127487546622e-06,
+      "loss": 8.6486,
+      "step": 137450
+    },
+    {
+      "epoch": 0.6864590876176684,
+      "grad_norm": 0.08919500559568405,
+      "learning_rate": 9.429772960524668e-06,
+      "loss": 8.6351,
+      "step": 137460
+    },
+    {
+      "epoch": 0.6865090264426078,
+      "grad_norm": 0.0964340791106224,
+      "learning_rate": 9.428271045583119e-06,
+      "loss": 8.6507,
+      "step": 137470
+    },
+    {
+      "epoch": 0.6865589652675472,
+      "grad_norm": 0.09165029972791672,
+      "learning_rate": 9.426769130641569e-06,
+      "loss": 8.6495,
+      "step": 137480
+    },
+    {
+      "epoch": 0.6866089040924866,
+      "grad_norm": 0.09624398499727249,
+      "learning_rate": 9.425267215700019e-06,
+      "loss": 8.6515,
+      "step": 137490
+    },
+    {
+      "epoch": 0.6866588429174262,
+      "grad_norm": 0.09847095608711243,
+      "learning_rate": 9.423765300758467e-06,
+      "loss": 8.6554,
+      "step": 137500
+    },
+    {
+      "epoch": 0.6867087817423656,
+      "grad_norm": 0.0960099995136261,
+      "learning_rate": 9.422263385816916e-06,
+      "loss": 8.6467,
+      "step": 137510
+    },
+    {
+      "epoch": 0.686758720567305,
+      "grad_norm": 0.09143520146608353,
+      "learning_rate": 9.420761470875366e-06,
+      "loss": 8.6458,
+      "step": 137520
+    },
+    {
+      "epoch": 0.6868086593922444,
+      "grad_norm": 0.08999507874250412,
+      "learning_rate": 9.419259555933816e-06,
+      "loss": 8.6486,
+      "step": 137530
+    },
+    {
+      "epoch": 0.686858598217184,
+      "grad_norm": 0.0936044380068779,
+      "learning_rate": 9.417757640992266e-06,
+      "loss": 8.6722,
+      "step": 137540
+    },
+    {
+      "epoch": 0.6869085370421234,
+      "grad_norm": 0.08734313398599625,
+      "learning_rate": 9.416255726050715e-06,
+      "loss": 8.6346,
+      "step": 137550
+    },
+    {
+      "epoch": 0.6869584758670628,
+      "grad_norm": 0.09482523053884506,
+      "learning_rate": 9.414753811109163e-06,
+      "loss": 8.6509,
+      "step": 137560
+    },
+    {
+      "epoch": 0.6870084146920022,
+      "grad_norm": 0.08811195939779282,
+      "learning_rate": 9.413251896167614e-06,
+      "loss": 8.6456,
+      "step": 137570
+    },
+    {
+      "epoch": 0.6870583535169418,
+      "grad_norm": 0.09007152169942856,
+      "learning_rate": 9.411749981226064e-06,
+      "loss": 8.6496,
+      "step": 137580
+    },
+    {
+      "epoch": 0.6871082923418812,
+      "grad_norm": 0.09563975781202316,
+      "learning_rate": 9.410248066284514e-06,
+      "loss": 8.6543,
+      "step": 137590
+    },
+    {
+      "epoch": 0.6871582311668206,
+      "grad_norm": 0.09100675582885742,
+      "learning_rate": 9.408746151342962e-06,
+      "loss": 8.6472,
+      "step": 137600
+    },
+    {
+      "epoch": 0.68720816999176,
+      "grad_norm": 0.09444539994001389,
+      "learning_rate": 9.407244236401411e-06,
+      "loss": 8.652,
+      "step": 137610
+    },
+    {
+      "epoch": 0.6872581088166996,
+      "grad_norm": 0.09386273473501205,
+      "learning_rate": 9.405742321459861e-06,
+      "loss": 8.6341,
+      "step": 137620
+    },
+    {
+      "epoch": 0.687308047641639,
+      "grad_norm": 0.09169203042984009,
+      "learning_rate": 9.404240406518311e-06,
+      "loss": 8.6365,
+      "step": 137630
+    },
+    {
+      "epoch": 0.6873579864665784,
+      "grad_norm": 0.0909794345498085,
+      "learning_rate": 9.402738491576761e-06,
+      "loss": 8.6598,
+      "step": 137640
+    },
+    {
+      "epoch": 0.6874079252915178,
+      "grad_norm": 0.08933387696743011,
+      "learning_rate": 9.40123657663521e-06,
+      "loss": 8.6547,
+      "step": 137650
+    },
+    {
+      "epoch": 0.6874578641164574,
+      "grad_norm": 0.097355417907238,
+      "learning_rate": 9.399734661693658e-06,
+      "loss": 8.6451,
+      "step": 137660
+    },
+    {
+      "epoch": 0.6875078029413968,
+      "grad_norm": 0.08841703087091446,
+      "learning_rate": 9.398232746752109e-06,
+      "loss": 8.6566,
+      "step": 137670
+    },
+    {
+      "epoch": 0.6875577417663362,
+      "grad_norm": 0.09026035666465759,
+      "learning_rate": 9.396730831810559e-06,
+      "loss": 8.6522,
+      "step": 137680
+    },
+    {
+      "epoch": 0.6876076805912756,
+      "grad_norm": 0.09169178456068039,
+      "learning_rate": 9.395228916869009e-06,
+      "loss": 8.6437,
+      "step": 137690
+    },
+    {
+      "epoch": 0.6876576194162152,
+      "grad_norm": 0.08948150277137756,
+      "learning_rate": 9.393727001927459e-06,
+      "loss": 8.6354,
+      "step": 137700
+    },
+    {
+      "epoch": 0.6877075582411546,
+      "grad_norm": 0.09596080332994461,
+      "learning_rate": 9.392225086985906e-06,
+      "loss": 8.637,
+      "step": 137710
+    },
+    {
+      "epoch": 0.687757497066094,
+      "grad_norm": 0.09012558311223984,
+      "learning_rate": 9.390723172044356e-06,
+      "loss": 8.6362,
+      "step": 137720
+    },
+    {
+      "epoch": 0.6878074358910334,
+      "grad_norm": 0.09151880443096161,
+      "learning_rate": 9.389221257102806e-06,
+      "loss": 8.643,
+      "step": 137730
+    },
+    {
+      "epoch": 0.687857374715973,
+      "grad_norm": 0.09232999384403229,
+      "learning_rate": 9.387719342161256e-06,
+      "loss": 8.648,
+      "step": 137740
+    },
+    {
+      "epoch": 0.6879073135409124,
+      "grad_norm": 0.08922406286001205,
+      "learning_rate": 9.386217427219707e-06,
+      "loss": 8.6526,
+      "step": 137750
+    },
+    {
+      "epoch": 0.6879572523658518,
+      "grad_norm": 0.09624180942773819,
+      "learning_rate": 9.384715512278153e-06,
+      "loss": 8.6546,
+      "step": 137760
+    },
+    {
+      "epoch": 0.6880071911907912,
+      "grad_norm": 0.09329060465097427,
+      "learning_rate": 9.383213597336604e-06,
+      "loss": 8.652,
+      "step": 137770
+    },
+    {
+      "epoch": 0.6880571300157308,
+      "grad_norm": 0.08916980773210526,
+      "learning_rate": 9.381711682395054e-06,
+      "loss": 8.6219,
+      "step": 137780
+    },
+    {
+      "epoch": 0.6881070688406702,
+      "grad_norm": 0.08881982415914536,
+      "learning_rate": 9.380209767453504e-06,
+      "loss": 8.6483,
+      "step": 137790
+    },
+    {
+      "epoch": 0.6881570076656096,
+      "grad_norm": 0.09591236710548401,
+      "learning_rate": 9.378707852511954e-06,
+      "loss": 8.6547,
+      "step": 137800
+    },
+    {
+      "epoch": 0.688206946490549,
+      "grad_norm": 0.0902673676609993,
+      "learning_rate": 9.377205937570401e-06,
+      "loss": 8.6516,
+      "step": 137810
+    },
+    {
+      "epoch": 0.6882568853154886,
+      "grad_norm": 0.09228136390447617,
+      "learning_rate": 9.375704022628851e-06,
+      "loss": 8.6433,
+      "step": 137820
+    },
+    {
+      "epoch": 0.688306824140428,
+      "grad_norm": 0.09464501589536667,
+      "learning_rate": 9.374202107687301e-06,
+      "loss": 8.6375,
+      "step": 137830
+    },
+    {
+      "epoch": 0.6883567629653674,
+      "grad_norm": 0.09438979625701904,
+      "learning_rate": 9.372700192745751e-06,
+      "loss": 8.6417,
+      "step": 137840
+    },
+    {
+      "epoch": 0.6884067017903068,
+      "grad_norm": 0.08938003331422806,
+      "learning_rate": 9.371198277804202e-06,
+      "loss": 8.6508,
+      "step": 137850
+    },
+    {
+      "epoch": 0.6884566406152464,
+      "grad_norm": 0.09114663302898407,
+      "learning_rate": 9.36969636286265e-06,
+      "loss": 8.6371,
+      "step": 137860
+    },
+    {
+      "epoch": 0.6885065794401858,
+      "grad_norm": 0.09571482241153717,
+      "learning_rate": 9.368194447921099e-06,
+      "loss": 8.6337,
+      "step": 137870
+    },
+    {
+      "epoch": 0.6885565182651252,
+      "grad_norm": 0.08686354756355286,
+      "learning_rate": 9.366692532979549e-06,
+      "loss": 8.6501,
+      "step": 137880
+    },
+    {
+      "epoch": 0.6886064570900646,
+      "grad_norm": 0.09277797490358353,
+      "learning_rate": 9.365190618037999e-06,
+      "loss": 8.6464,
+      "step": 137890
+    },
+    {
+      "epoch": 0.6886563959150042,
+      "grad_norm": 0.09338825196027756,
+      "learning_rate": 9.363688703096449e-06,
+      "loss": 8.6573,
+      "step": 137900
+    },
+    {
+      "epoch": 0.6887063347399436,
+      "grad_norm": 0.09079349040985107,
+      "learning_rate": 9.362186788154898e-06,
+      "loss": 8.6472,
+      "step": 137910
+    },
+    {
+      "epoch": 0.688756273564883,
+      "grad_norm": 0.09190376847982407,
+      "learning_rate": 9.360684873213346e-06,
+      "loss": 8.6445,
+      "step": 137920
+    },
+    {
+      "epoch": 0.6888062123898224,
+      "grad_norm": 0.09575271606445312,
+      "learning_rate": 9.359182958271796e-06,
+      "loss": 8.6442,
+      "step": 137930
+    },
+    {
+      "epoch": 0.688856151214762,
+      "grad_norm": 0.09022709727287292,
+      "learning_rate": 9.357681043330246e-06,
+      "loss": 8.6403,
+      "step": 137940
+    },
+    {
+      "epoch": 0.6889060900397014,
+      "grad_norm": 0.09948229044675827,
+      "learning_rate": 9.356179128388697e-06,
+      "loss": 8.636,
+      "step": 137950
+    },
+    {
+      "epoch": 0.6889560288646408,
+      "grad_norm": 0.09341901540756226,
+      "learning_rate": 9.354677213447145e-06,
+      "loss": 8.6431,
+      "step": 137960
+    },
+    {
+      "epoch": 0.6890059676895802,
+      "grad_norm": 0.09408657997846603,
+      "learning_rate": 9.353175298505594e-06,
+      "loss": 8.6304,
+      "step": 137970
+    },
+    {
+      "epoch": 0.6890559065145198,
+      "grad_norm": 0.09293026477098465,
+      "learning_rate": 9.351673383564044e-06,
+      "loss": 8.6281,
+      "step": 137980
+    },
+    {
+      "epoch": 0.6891058453394592,
+      "grad_norm": 0.09260407835245132,
+      "learning_rate": 9.350171468622494e-06,
+      "loss": 8.6426,
+      "step": 137990
+    },
+    {
+      "epoch": 0.6891557841643986,
+      "grad_norm": 0.09389134496450424,
+      "learning_rate": 9.348669553680944e-06,
+      "loss": 8.6538,
+      "step": 138000
+    },
+    {
+      "epoch": 0.689205722989338,
+      "grad_norm": 0.09285546839237213,
+      "learning_rate": 9.347167638739393e-06,
+      "loss": 8.6568,
+      "step": 138010
+    },
+    {
+      "epoch": 0.6892556618142776,
+      "grad_norm": 0.0875903069972992,
+      "learning_rate": 9.345665723797843e-06,
+      "loss": 8.6533,
+      "step": 138020
+    },
+    {
+      "epoch": 0.689305600639217,
+      "grad_norm": 0.08949259668588638,
+      "learning_rate": 9.344163808856291e-06,
+      "loss": 8.6302,
+      "step": 138030
+    },
+    {
+      "epoch": 0.6893555394641564,
+      "grad_norm": 0.09010973572731018,
+      "learning_rate": 9.342661893914741e-06,
+      "loss": 8.6442,
+      "step": 138040
+    },
+    {
+      "epoch": 0.6894054782890958,
+      "grad_norm": 0.09066742658615112,
+      "learning_rate": 9.341159978973192e-06,
+      "loss": 8.6377,
+      "step": 138050
+    },
+    {
+      "epoch": 0.6894554171140354,
+      "grad_norm": 0.09108269959688187,
+      "learning_rate": 9.33965806403164e-06,
+      "loss": 8.6662,
+      "step": 138060
+    },
+    {
+      "epoch": 0.6895053559389748,
+      "grad_norm": 0.08662552386522293,
+      "learning_rate": 9.33815614909009e-06,
+      "loss": 8.6386,
+      "step": 138070
+    },
+    {
+      "epoch": 0.6895552947639142,
+      "grad_norm": 0.08738676458597183,
+      "learning_rate": 9.336654234148539e-06,
+      "loss": 8.6379,
+      "step": 138080
+    },
+    {
+      "epoch": 0.6896052335888536,
+      "grad_norm": 0.0980435162782669,
+      "learning_rate": 9.335152319206989e-06,
+      "loss": 8.6349,
+      "step": 138090
+    },
+    {
+      "epoch": 0.6896551724137931,
+      "grad_norm": 0.09144052863121033,
+      "learning_rate": 9.333650404265439e-06,
+      "loss": 8.638,
+      "step": 138100
+    },
+    {
+      "epoch": 0.6897051112387326,
+      "grad_norm": 0.09114320576190948,
+      "learning_rate": 9.332148489323888e-06,
+      "loss": 8.6406,
+      "step": 138110
+    },
+    {
+      "epoch": 0.689755050063672,
+      "grad_norm": 0.09149006754159927,
+      "learning_rate": 9.330646574382338e-06,
+      "loss": 8.6525,
+      "step": 138120
+    },
+    {
+      "epoch": 0.6898049888886114,
+      "grad_norm": 0.0991503894329071,
+      "learning_rate": 9.329144659440786e-06,
+      "loss": 8.6523,
+      "step": 138130
+    },
+    {
+      "epoch": 0.689854927713551,
+      "grad_norm": 0.09777787327766418,
+      "learning_rate": 9.327642744499236e-06,
+      "loss": 8.6357,
+      "step": 138140
+    },
+    {
+      "epoch": 0.6899048665384904,
+      "grad_norm": 0.09248573333024979,
+      "learning_rate": 9.326140829557687e-06,
+      "loss": 8.6435,
+      "step": 138150
+    },
+    {
+      "epoch": 0.6899548053634298,
+      "grad_norm": 0.09282256662845612,
+      "learning_rate": 9.324638914616135e-06,
+      "loss": 8.6544,
+      "step": 138160
+    },
+    {
+      "epoch": 0.6900047441883692,
+      "grad_norm": 0.08951915800571442,
+      "learning_rate": 9.323136999674585e-06,
+      "loss": 8.6397,
+      "step": 138170
+    },
+    {
+      "epoch": 0.6900546830133087,
+      "grad_norm": 0.09212198853492737,
+      "learning_rate": 9.321635084733035e-06,
+      "loss": 8.6415,
+      "step": 138180
+    },
+    {
+      "epoch": 0.6901046218382482,
+      "grad_norm": 0.09542210400104523,
+      "learning_rate": 9.320133169791484e-06,
+      "loss": 8.6362,
+      "step": 138190
+    },
+    {
+      "epoch": 0.6901545606631876,
+      "grad_norm": 0.08942250162363052,
+      "learning_rate": 9.318631254849934e-06,
+      "loss": 8.6416,
+      "step": 138200
+    },
+    {
+      "epoch": 0.690204499488127,
+      "grad_norm": 0.08706214278936386,
+      "learning_rate": 9.317129339908383e-06,
+      "loss": 8.6381,
+      "step": 138210
+    },
+    {
+      "epoch": 0.6902544383130665,
+      "grad_norm": 0.09015882015228271,
+      "learning_rate": 9.315627424966833e-06,
+      "loss": 8.6384,
+      "step": 138220
+    },
+    {
+      "epoch": 0.690304377138006,
+      "grad_norm": 0.09370545297861099,
+      "learning_rate": 9.314125510025283e-06,
+      "loss": 8.6408,
+      "step": 138230
+    },
+    {
+      "epoch": 0.6903543159629454,
+      "grad_norm": 0.08900182694196701,
+      "learning_rate": 9.312623595083731e-06,
+      "loss": 8.6495,
+      "step": 138240
+    },
+    {
+      "epoch": 0.6904042547878848,
+      "grad_norm": 0.08960168808698654,
+      "learning_rate": 9.311121680142182e-06,
+      "loss": 8.6513,
+      "step": 138250
+    },
+    {
+      "epoch": 0.6904541936128243,
+      "grad_norm": 0.09003665298223495,
+      "learning_rate": 9.30961976520063e-06,
+      "loss": 8.6419,
+      "step": 138260
+    },
+    {
+      "epoch": 0.6905041324377638,
+      "grad_norm": 0.09128106385469437,
+      "learning_rate": 9.30811785025908e-06,
+      "loss": 8.6406,
+      "step": 138270
+    },
+    {
+      "epoch": 0.6905540712627032,
+      "grad_norm": 0.09446307271718979,
+      "learning_rate": 9.30661593531753e-06,
+      "loss": 8.6382,
+      "step": 138280
+    },
+    {
+      "epoch": 0.6906040100876426,
+      "grad_norm": 0.08831710368394852,
+      "learning_rate": 9.305114020375979e-06,
+      "loss": 8.6417,
+      "step": 138290
+    },
+    {
+      "epoch": 0.6906539489125821,
+      "grad_norm": 0.08965827524662018,
+      "learning_rate": 9.303612105434429e-06,
+      "loss": 8.633,
+      "step": 138300
+    },
+    {
+      "epoch": 0.6907038877375216,
+      "grad_norm": 0.09387178719043732,
+      "learning_rate": 9.302110190492878e-06,
+      "loss": 8.6404,
+      "step": 138310
+    },
+    {
+      "epoch": 0.690753826562461,
+      "grad_norm": 0.0937371626496315,
+      "learning_rate": 9.300608275551328e-06,
+      "loss": 8.6382,
+      "step": 138320
+    },
+    {
+      "epoch": 0.6908037653874004,
+      "grad_norm": 0.08978761732578278,
+      "learning_rate": 9.299106360609778e-06,
+      "loss": 8.6647,
+      "step": 138330
+    },
+    {
+      "epoch": 0.6908537042123399,
+      "grad_norm": 0.09310397505760193,
+      "learning_rate": 9.297604445668228e-06,
+      "loss": 8.6311,
+      "step": 138340
+    },
+    {
+      "epoch": 0.6909036430372794,
+      "grad_norm": 0.0897870808839798,
+      "learning_rate": 9.296102530726677e-06,
+      "loss": 8.6466,
+      "step": 138350
+    },
+    {
+      "epoch": 0.6909535818622188,
+      "grad_norm": 0.09369547665119171,
+      "learning_rate": 9.294600615785127e-06,
+      "loss": 8.6451,
+      "step": 138360
+    },
+    {
+      "epoch": 0.6910035206871582,
+      "grad_norm": 0.09182377904653549,
+      "learning_rate": 9.293098700843575e-06,
+      "loss": 8.636,
+      "step": 138370
+    },
+    {
+      "epoch": 0.6910534595120977,
+      "grad_norm": 0.09317004680633545,
+      "learning_rate": 9.291596785902025e-06,
+      "loss": 8.6357,
+      "step": 138380
+    },
+    {
+      "epoch": 0.6911033983370372,
+      "grad_norm": 0.09063698351383209,
+      "learning_rate": 9.290094870960476e-06,
+      "loss": 8.6374,
+      "step": 138390
+    },
+    {
+      "epoch": 0.6911533371619766,
+      "grad_norm": 0.09114688634872437,
+      "learning_rate": 9.288592956018924e-06,
+      "loss": 8.6364,
+      "step": 138400
+    },
+    {
+      "epoch": 0.691203275986916,
+      "grad_norm": 0.08719582855701447,
+      "learning_rate": 9.287091041077374e-06,
+      "loss": 8.6513,
+      "step": 138410
+    },
+    {
+      "epoch": 0.6912532148118555,
+      "grad_norm": 0.09610260277986526,
+      "learning_rate": 9.285589126135823e-06,
+      "loss": 8.6279,
+      "step": 138420
+    },
+    {
+      "epoch": 0.691303153636795,
+      "grad_norm": 0.08931412547826767,
+      "learning_rate": 9.284087211194273e-06,
+      "loss": 8.6179,
+      "step": 138430
+    },
+    {
+      "epoch": 0.6913530924617344,
+      "grad_norm": 0.09096454083919525,
+      "learning_rate": 9.282585296252723e-06,
+      "loss": 8.6538,
+      "step": 138440
+    },
+    {
+      "epoch": 0.6914030312866738,
+      "grad_norm": 0.09301277250051498,
+      "learning_rate": 9.281083381311172e-06,
+      "loss": 8.6439,
+      "step": 138450
+    },
+    {
+      "epoch": 0.6914529701116132,
+      "grad_norm": 0.09470412135124207,
+      "learning_rate": 9.279581466369622e-06,
+      "loss": 8.6572,
+      "step": 138460
+    },
+    {
+      "epoch": 0.6915029089365528,
+      "grad_norm": 0.08895006030797958,
+      "learning_rate": 9.27807955142807e-06,
+      "loss": 8.6397,
+      "step": 138470
+    },
+    {
+      "epoch": 0.6915528477614922,
+      "grad_norm": 0.09335384517908096,
+      "learning_rate": 9.27657763648652e-06,
+      "loss": 8.6424,
+      "step": 138480
+    },
+    {
+      "epoch": 0.6916027865864316,
+      "grad_norm": 0.08956567943096161,
+      "learning_rate": 9.27507572154497e-06,
+      "loss": 8.6657,
+      "step": 138490
+    },
+    {
+      "epoch": 0.691652725411371,
+      "grad_norm": 0.08856837451457977,
+      "learning_rate": 9.273573806603421e-06,
+      "loss": 8.6263,
+      "step": 138500
+    },
+    {
+      "epoch": 0.6917026642363105,
+      "grad_norm": 0.09013232588768005,
+      "learning_rate": 9.27207189166187e-06,
+      "loss": 8.6193,
+      "step": 138510
+    },
+    {
+      "epoch": 0.69175260306125,
+      "grad_norm": 0.09470102936029434,
+      "learning_rate": 9.270569976720318e-06,
+      "loss": 8.6281,
+      "step": 138520
+    },
+    {
+      "epoch": 0.6918025418861894,
+      "grad_norm": 0.08966551721096039,
+      "learning_rate": 9.269068061778768e-06,
+      "loss": 8.6274,
+      "step": 138530
+    },
+    {
+      "epoch": 0.6918524807111288,
+      "grad_norm": 0.09647005796432495,
+      "learning_rate": 9.267566146837218e-06,
+      "loss": 8.652,
+      "step": 138540
+    },
+    {
+      "epoch": 0.6919024195360683,
+      "grad_norm": 0.08961433917284012,
+      "learning_rate": 9.266064231895668e-06,
+      "loss": 8.634,
+      "step": 138550
+    },
+    {
+      "epoch": 0.6919523583610078,
+      "grad_norm": 0.09530511498451233,
+      "learning_rate": 9.264562316954117e-06,
+      "loss": 8.6453,
+      "step": 138560
+    },
+    {
+      "epoch": 0.6920022971859472,
+      "grad_norm": 0.09226834028959274,
+      "learning_rate": 9.263060402012565e-06,
+      "loss": 8.6172,
+      "step": 138570
+    },
+    {
+      "epoch": 0.6920522360108866,
+      "grad_norm": 0.09644404798746109,
+      "learning_rate": 9.261558487071016e-06,
+      "loss": 8.6318,
+      "step": 138580
+    },
+    {
+      "epoch": 0.6921021748358261,
+      "grad_norm": 0.08702649176120758,
+      "learning_rate": 9.260056572129466e-06,
+      "loss": 8.6606,
+      "step": 138590
+    },
+    {
+      "epoch": 0.6921521136607656,
+      "grad_norm": 0.08770040422677994,
+      "learning_rate": 9.258554657187916e-06,
+      "loss": 8.6342,
+      "step": 138600
+    },
+    {
+      "epoch": 0.692202052485705,
+      "grad_norm": 0.09810072183609009,
+      "learning_rate": 9.257052742246364e-06,
+      "loss": 8.6444,
+      "step": 138610
+    },
+    {
+      "epoch": 0.6922519913106444,
+      "grad_norm": 0.09160198271274567,
+      "learning_rate": 9.255550827304813e-06,
+      "loss": 8.6532,
+      "step": 138620
+    },
+    {
+      "epoch": 0.6923019301355839,
+      "grad_norm": 0.09566731005907059,
+      "learning_rate": 9.254048912363263e-06,
+      "loss": 8.6305,
+      "step": 138630
+    },
+    {
+      "epoch": 0.6923518689605234,
+      "grad_norm": 0.0926600992679596,
+      "learning_rate": 9.252546997421713e-06,
+      "loss": 8.6613,
+      "step": 138640
+    },
+    {
+      "epoch": 0.6924018077854628,
+      "grad_norm": 0.0963670089840889,
+      "learning_rate": 9.251045082480163e-06,
+      "loss": 8.6392,
+      "step": 138650
+    },
+    {
+      "epoch": 0.6924517466104022,
+      "grad_norm": 0.09352785348892212,
+      "learning_rate": 9.249543167538614e-06,
+      "loss": 8.6547,
+      "step": 138660
+    },
+    {
+      "epoch": 0.6925016854353417,
+      "grad_norm": 0.08976289629936218,
+      "learning_rate": 9.24804125259706e-06,
+      "loss": 8.6306,
+      "step": 138670
+    },
+    {
+      "epoch": 0.6925516242602812,
+      "grad_norm": 0.0948214903473854,
+      "learning_rate": 9.24653933765551e-06,
+      "loss": 8.6406,
+      "step": 138680
+    },
+    {
+      "epoch": 0.6926015630852206,
+      "grad_norm": 0.08709558844566345,
+      "learning_rate": 9.24503742271396e-06,
+      "loss": 8.6603,
+      "step": 138690
+    },
+    {
+      "epoch": 0.69265150191016,
+      "grad_norm": 0.09074770659208298,
+      "learning_rate": 9.243535507772411e-06,
+      "loss": 8.6567,
+      "step": 138700
+    },
+    {
+      "epoch": 0.6927014407350995,
+      "grad_norm": 0.08999824523925781,
+      "learning_rate": 9.242033592830861e-06,
+      "loss": 8.659,
+      "step": 138710
+    },
+    {
+      "epoch": 0.692751379560039,
+      "grad_norm": 0.09150887280702591,
+      "learning_rate": 9.240531677889308e-06,
+      "loss": 8.6724,
+      "step": 138720
+    },
+    {
+      "epoch": 0.6928013183849784,
+      "grad_norm": 0.09337803721427917,
+      "learning_rate": 9.239029762947758e-06,
+      "loss": 8.64,
+      "step": 138730
+    },
+    {
+      "epoch": 0.6928512572099178,
+      "grad_norm": 0.09013208001852036,
+      "learning_rate": 9.237527848006208e-06,
+      "loss": 8.6463,
+      "step": 138740
+    },
+    {
+      "epoch": 0.6929011960348573,
+      "grad_norm": 0.08989313244819641,
+      "learning_rate": 9.236025933064658e-06,
+      "loss": 8.6294,
+      "step": 138750
+    },
+    {
+      "epoch": 0.6929511348597968,
+      "grad_norm": 0.09004274755716324,
+      "learning_rate": 9.234524018123109e-06,
+      "loss": 8.6373,
+      "step": 138760
+    },
+    {
+      "epoch": 0.6930010736847362,
+      "grad_norm": 0.09396898746490479,
+      "learning_rate": 9.233022103181555e-06,
+      "loss": 8.6434,
+      "step": 138770
+    },
+    {
+      "epoch": 0.6930510125096756,
+      "grad_norm": 0.08722883462905884,
+      "learning_rate": 9.231520188240006e-06,
+      "loss": 8.6448,
+      "step": 138780
+    },
+    {
+      "epoch": 0.6931009513346151,
+      "grad_norm": 0.0960550382733345,
+      "learning_rate": 9.230018273298456e-06,
+      "loss": 8.6277,
+      "step": 138790
+    },
+    {
+      "epoch": 0.6931508901595546,
+      "grad_norm": 0.09497455507516861,
+      "learning_rate": 9.228516358356906e-06,
+      "loss": 8.6294,
+      "step": 138800
+    },
+    {
+      "epoch": 0.693200828984494,
+      "grad_norm": 0.086859792470932,
+      "learning_rate": 9.227014443415356e-06,
+      "loss": 8.632,
+      "step": 138810
+    },
+    {
+      "epoch": 0.6932507678094334,
+      "grad_norm": 0.08842300623655319,
+      "learning_rate": 9.225512528473805e-06,
+      "loss": 8.6624,
+      "step": 138820
+    },
+    {
+      "epoch": 0.6933007066343729,
+      "grad_norm": 0.09223365783691406,
+      "learning_rate": 9.224010613532253e-06,
+      "loss": 8.6479,
+      "step": 138830
+    },
+    {
+      "epoch": 0.6933506454593124,
+      "grad_norm": 0.09159765392541885,
+      "learning_rate": 9.222508698590703e-06,
+      "loss": 8.6521,
+      "step": 138840
+    },
+    {
+      "epoch": 0.6934005842842518,
+      "grad_norm": 0.0935441255569458,
+      "learning_rate": 9.221006783649153e-06,
+      "loss": 8.6378,
+      "step": 138850
+    },
+    {
+      "epoch": 0.6934505231091912,
+      "grad_norm": 0.09290704131126404,
+      "learning_rate": 9.219504868707604e-06,
+      "loss": 8.6148,
+      "step": 138860
+    },
+    {
+      "epoch": 0.6935004619341307,
+      "grad_norm": 0.08997565507888794,
+      "learning_rate": 9.218002953766052e-06,
+      "loss": 8.6324,
+      "step": 138870
+    },
+    {
+      "epoch": 0.6935504007590702,
+      "grad_norm": 0.09210515767335892,
+      "learning_rate": 9.2165010388245e-06,
+      "loss": 8.6539,
+      "step": 138880
+    },
+    {
+      "epoch": 0.6936003395840096,
+      "grad_norm": 0.09159453958272934,
+      "learning_rate": 9.21499912388295e-06,
+      "loss": 8.6455,
+      "step": 138890
+    },
+    {
+      "epoch": 0.693650278408949,
+      "grad_norm": 0.09023798257112503,
+      "learning_rate": 9.213497208941401e-06,
+      "loss": 8.6536,
+      "step": 138900
+    },
+    {
+      "epoch": 0.6937002172338885,
+      "grad_norm": 0.08913420885801315,
+      "learning_rate": 9.211995293999851e-06,
+      "loss": 8.6457,
+      "step": 138910
+    },
+    {
+      "epoch": 0.693750156058828,
+      "grad_norm": 0.09656532853841782,
+      "learning_rate": 9.2104933790583e-06,
+      "loss": 8.6271,
+      "step": 138920
+    },
+    {
+      "epoch": 0.6938000948837674,
+      "grad_norm": 0.08836175501346588,
+      "learning_rate": 9.208991464116748e-06,
+      "loss": 8.6521,
+      "step": 138930
+    },
+    {
+      "epoch": 0.6938500337087068,
+      "grad_norm": 0.09175171703100204,
+      "learning_rate": 9.207489549175198e-06,
+      "loss": 8.6369,
+      "step": 138940
+    },
+    {
+      "epoch": 0.6938999725336463,
+      "grad_norm": 0.09058476984500885,
+      "learning_rate": 9.205987634233648e-06,
+      "loss": 8.637,
+      "step": 138950
+    },
+    {
+      "epoch": 0.6939499113585857,
+      "grad_norm": 0.08583276718854904,
+      "learning_rate": 9.204485719292099e-06,
+      "loss": 8.6521,
+      "step": 138960
+    },
+    {
+      "epoch": 0.6939998501835252,
+      "grad_norm": 0.09340415894985199,
+      "learning_rate": 9.202983804350547e-06,
+      "loss": 8.645,
+      "step": 138970
+    },
+    {
+      "epoch": 0.6940497890084646,
+      "grad_norm": 0.09227906912565231,
+      "learning_rate": 9.201481889408997e-06,
+      "loss": 8.6404,
+      "step": 138980
+    },
+    {
+      "epoch": 0.6940997278334041,
+      "grad_norm": 0.09200102090835571,
+      "learning_rate": 9.199979974467446e-06,
+      "loss": 8.6315,
+      "step": 138990
+    },
+    {
+      "epoch": 0.6941496666583435,
+      "grad_norm": 0.09255097806453705,
+      "learning_rate": 9.198478059525896e-06,
+      "loss": 8.6211,
+      "step": 139000
+    },
+    {
+      "epoch": 0.694199605483283,
+      "grad_norm": 0.09157124161720276,
+      "learning_rate": 9.196976144584346e-06,
+      "loss": 8.648,
+      "step": 139010
+    },
+    {
+      "epoch": 0.6942495443082224,
+      "grad_norm": 0.09558318555355072,
+      "learning_rate": 9.195474229642795e-06,
+      "loss": 8.6472,
+      "step": 139020
+    },
+    {
+      "epoch": 0.6942994831331619,
+      "grad_norm": 0.09198082983493805,
+      "learning_rate": 9.193972314701245e-06,
+      "loss": 8.6296,
+      "step": 139030
+    },
+    {
+      "epoch": 0.6943494219581013,
+      "grad_norm": 0.09510613977909088,
+      "learning_rate": 9.192470399759693e-06,
+      "loss": 8.6449,
+      "step": 139040
+    },
+    {
+      "epoch": 0.6943993607830408,
+      "grad_norm": 0.0944787859916687,
+      "learning_rate": 9.190968484818143e-06,
+      "loss": 8.6387,
+      "step": 139050
+    },
+    {
+      "epoch": 0.6944492996079802,
+      "grad_norm": 0.09569986164569855,
+      "learning_rate": 9.189466569876594e-06,
+      "loss": 8.6267,
+      "step": 139060
+    },
+    {
+      "epoch": 0.6944992384329197,
+      "grad_norm": 0.09105094522237778,
+      "learning_rate": 9.187964654935042e-06,
+      "loss": 8.6455,
+      "step": 139070
+    },
+    {
+      "epoch": 0.6945491772578591,
+      "grad_norm": 0.09231606870889664,
+      "learning_rate": 9.186462739993492e-06,
+      "loss": 8.6481,
+      "step": 139080
+    },
+    {
+      "epoch": 0.6945991160827986,
+      "grad_norm": 0.09214016050100327,
+      "learning_rate": 9.18496082505194e-06,
+      "loss": 8.637,
+      "step": 139090
+    },
+    {
+      "epoch": 0.694649054907738,
+      "grad_norm": 0.09304431825876236,
+      "learning_rate": 9.183458910110391e-06,
+      "loss": 8.6331,
+      "step": 139100
+    },
+    {
+      "epoch": 0.6946989937326775,
+      "grad_norm": 0.09022108465433121,
+      "learning_rate": 9.181956995168841e-06,
+      "loss": 8.6322,
+      "step": 139110
+    },
+    {
+      "epoch": 0.6947489325576169,
+      "grad_norm": 0.09341921657323837,
+      "learning_rate": 9.18045508022729e-06,
+      "loss": 8.6466,
+      "step": 139120
+    },
+    {
+      "epoch": 0.6947988713825564,
+      "grad_norm": 0.09652432799339294,
+      "learning_rate": 9.17895316528574e-06,
+      "loss": 8.6152,
+      "step": 139130
+    },
+    {
+      "epoch": 0.6948488102074958,
+      "grad_norm": 0.08961055427789688,
+      "learning_rate": 9.17745125034419e-06,
+      "loss": 8.6514,
+      "step": 139140
+    },
+    {
+      "epoch": 0.6948987490324353,
+      "grad_norm": 0.09760846197605133,
+      "learning_rate": 9.175949335402638e-06,
+      "loss": 8.635,
+      "step": 139150
+    },
+    {
+      "epoch": 0.6949486878573747,
+      "grad_norm": 0.0923495665192604,
+      "learning_rate": 9.174447420461089e-06,
+      "loss": 8.6445,
+      "step": 139160
+    },
+    {
+      "epoch": 0.6949986266823142,
+      "grad_norm": 0.09096313267946243,
+      "learning_rate": 9.172945505519537e-06,
+      "loss": 8.6452,
+      "step": 139170
+    },
+    {
+      "epoch": 0.6950485655072536,
+      "grad_norm": 0.09231965243816376,
+      "learning_rate": 9.171443590577987e-06,
+      "loss": 8.6316,
+      "step": 139180
+    },
+    {
+      "epoch": 0.6950985043321931,
+      "grad_norm": 0.09298811107873917,
+      "learning_rate": 9.169941675636437e-06,
+      "loss": 8.6394,
+      "step": 139190
+    },
+    {
+      "epoch": 0.6951484431571325,
+      "grad_norm": 0.08991467207670212,
+      "learning_rate": 9.168439760694886e-06,
+      "loss": 8.6401,
+      "step": 139200
+    },
+    {
+      "epoch": 0.695198381982072,
+      "grad_norm": 0.09446527063846588,
+      "learning_rate": 9.166937845753336e-06,
+      "loss": 8.6471,
+      "step": 139210
+    },
+    {
+      "epoch": 0.6952483208070114,
+      "grad_norm": 0.09214930981397629,
+      "learning_rate": 9.165435930811785e-06,
+      "loss": 8.6469,
+      "step": 139220
+    },
+    {
+      "epoch": 0.6952982596319509,
+      "grad_norm": 0.09560856968164444,
+      "learning_rate": 9.163934015870235e-06,
+      "loss": 8.6143,
+      "step": 139230
+    },
+    {
+      "epoch": 0.6953481984568903,
+      "grad_norm": 0.09127038717269897,
+      "learning_rate": 9.162432100928685e-06,
+      "loss": 8.6509,
+      "step": 139240
+    },
+    {
+      "epoch": 0.6953981372818298,
+      "grad_norm": 0.09658581763505936,
+      "learning_rate": 9.160930185987133e-06,
+      "loss": 8.6312,
+      "step": 139250
+    },
+    {
+      "epoch": 0.6954480761067692,
+      "grad_norm": 0.08580411970615387,
+      "learning_rate": 9.159428271045584e-06,
+      "loss": 8.6339,
+      "step": 139260
+    },
+    {
+      "epoch": 0.6954980149317087,
+      "grad_norm": 0.0958205908536911,
+      "learning_rate": 9.157926356104032e-06,
+      "loss": 8.6357,
+      "step": 139270
+    },
+    {
+      "epoch": 0.6955479537566481,
+      "grad_norm": 0.09207163751125336,
+      "learning_rate": 9.156424441162482e-06,
+      "loss": 8.6253,
+      "step": 139280
+    },
+    {
+      "epoch": 0.6955978925815876,
+      "grad_norm": 0.09957921504974365,
+      "learning_rate": 9.154922526220932e-06,
+      "loss": 8.6448,
+      "step": 139290
+    },
+    {
+      "epoch": 0.695647831406527,
+      "grad_norm": 0.08785407990217209,
+      "learning_rate": 9.153420611279383e-06,
+      "loss": 8.6244,
+      "step": 139300
+    },
+    {
+      "epoch": 0.6956977702314665,
+      "grad_norm": 0.08667879551649094,
+      "learning_rate": 9.151918696337831e-06,
+      "loss": 8.6377,
+      "step": 139310
+    },
+    {
+      "epoch": 0.6957477090564059,
+      "grad_norm": 0.09369443356990814,
+      "learning_rate": 9.15041678139628e-06,
+      "loss": 8.6415,
+      "step": 139320
+    },
+    {
+      "epoch": 0.6957976478813453,
+      "grad_norm": 0.08814938366413116,
+      "learning_rate": 9.14891486645473e-06,
+      "loss": 8.629,
+      "step": 139330
+    },
+    {
+      "epoch": 0.6958475867062848,
+      "grad_norm": 0.0961841344833374,
+      "learning_rate": 9.14741295151318e-06,
+      "loss": 8.6497,
+      "step": 139340
+    },
+    {
+      "epoch": 0.6958975255312243,
+      "grad_norm": 0.09453871101140976,
+      "learning_rate": 9.14591103657163e-06,
+      "loss": 8.6451,
+      "step": 139350
+    },
+    {
+      "epoch": 0.6959474643561637,
+      "grad_norm": 0.09213295578956604,
+      "learning_rate": 9.144409121630079e-06,
+      "loss": 8.6234,
+      "step": 139360
+    },
+    {
+      "epoch": 0.6959974031811031,
+      "grad_norm": 0.08923090249300003,
+      "learning_rate": 9.142907206688527e-06,
+      "loss": 8.617,
+      "step": 139370
+    },
+    {
+      "epoch": 0.6960473420060426,
+      "grad_norm": 0.09008591622114182,
+      "learning_rate": 9.141405291746977e-06,
+      "loss": 8.6443,
+      "step": 139380
+    },
+    {
+      "epoch": 0.6960972808309821,
+      "grad_norm": 0.0908149927854538,
+      "learning_rate": 9.139903376805427e-06,
+      "loss": 8.6455,
+      "step": 139390
+    },
+    {
+      "epoch": 0.6961472196559215,
+      "grad_norm": 0.09176421910524368,
+      "learning_rate": 9.138401461863878e-06,
+      "loss": 8.6389,
+      "step": 139400
+    },
+    {
+      "epoch": 0.696197158480861,
+      "grad_norm": 0.09254884719848633,
+      "learning_rate": 9.136899546922326e-06,
+      "loss": 8.6449,
+      "step": 139410
+    },
+    {
+      "epoch": 0.6962470973058004,
+      "grad_norm": 0.08738537132740021,
+      "learning_rate": 9.135397631980775e-06,
+      "loss": 8.6249,
+      "step": 139420
+    },
+    {
+      "epoch": 0.6962970361307398,
+      "grad_norm": 0.09267020225524902,
+      "learning_rate": 9.133895717039225e-06,
+      "loss": 8.6208,
+      "step": 139430
+    },
+    {
+      "epoch": 0.6963469749556793,
+      "grad_norm": 0.09037657082080841,
+      "learning_rate": 9.132393802097675e-06,
+      "loss": 8.6347,
+      "step": 139440
+    },
+    {
+      "epoch": 0.6963969137806187,
+      "grad_norm": 0.08870069682598114,
+      "learning_rate": 9.130891887156125e-06,
+      "loss": 8.6323,
+      "step": 139450
+    },
+    {
+      "epoch": 0.6964468526055582,
+      "grad_norm": 0.08991122990846634,
+      "learning_rate": 9.129389972214575e-06,
+      "loss": 8.6758,
+      "step": 139460
+    },
+    {
+      "epoch": 0.6964967914304976,
+      "grad_norm": 0.08251538872718811,
+      "learning_rate": 9.127888057273022e-06,
+      "loss": 8.6402,
+      "step": 139470
+    },
+    {
+      "epoch": 0.6965467302554371,
+      "grad_norm": 0.08768171072006226,
+      "learning_rate": 9.126386142331472e-06,
+      "loss": 8.6293,
+      "step": 139480
+    },
+    {
+      "epoch": 0.6965966690803765,
+      "grad_norm": 0.09158322215080261,
+      "learning_rate": 9.124884227389922e-06,
+      "loss": 8.6397,
+      "step": 139490
+    },
+    {
+      "epoch": 0.696646607905316,
+      "grad_norm": 0.08981853723526001,
+      "learning_rate": 9.123382312448373e-06,
+      "loss": 8.638,
+      "step": 139500
+    },
+    {
+      "epoch": 0.6966965467302554,
+      "grad_norm": 0.08998404443264008,
+      "learning_rate": 9.121880397506823e-06,
+      "loss": 8.6505,
+      "step": 139510
+    },
+    {
+      "epoch": 0.6967464855551949,
+      "grad_norm": 0.09334992617368698,
+      "learning_rate": 9.12037848256527e-06,
+      "loss": 8.6324,
+      "step": 139520
+    },
+    {
+      "epoch": 0.6967964243801343,
+      "grad_norm": 0.09155179560184479,
+      "learning_rate": 9.11887656762372e-06,
+      "loss": 8.6379,
+      "step": 139530
+    },
+    {
+      "epoch": 0.6968463632050738,
+      "grad_norm": 0.09096579253673553,
+      "learning_rate": 9.11737465268217e-06,
+      "loss": 8.6277,
+      "step": 139540
+    },
+    {
+      "epoch": 0.6968963020300132,
+      "grad_norm": 0.0905652865767479,
+      "learning_rate": 9.11587273774062e-06,
+      "loss": 8.6074,
+      "step": 139550
+    },
+    {
+      "epoch": 0.6969462408549527,
+      "grad_norm": 0.09262055158615112,
+      "learning_rate": 9.11437082279907e-06,
+      "loss": 8.6542,
+      "step": 139560
+    },
+    {
+      "epoch": 0.6969961796798921,
+      "grad_norm": 0.09247998148202896,
+      "learning_rate": 9.112868907857517e-06,
+      "loss": 8.6331,
+      "step": 139570
+    },
+    {
+      "epoch": 0.6970461185048316,
+      "grad_norm": 0.09279992431402206,
+      "learning_rate": 9.111366992915967e-06,
+      "loss": 8.6329,
+      "step": 139580
+    },
+    {
+      "epoch": 0.697096057329771,
+      "grad_norm": 0.09361322969198227,
+      "learning_rate": 9.109865077974417e-06,
+      "loss": 8.6413,
+      "step": 139590
+    },
+    {
+      "epoch": 0.6971459961547105,
+      "grad_norm": 0.09208791702985764,
+      "learning_rate": 9.108363163032868e-06,
+      "loss": 8.6303,
+      "step": 139600
+    },
+    {
+      "epoch": 0.6971959349796499,
+      "grad_norm": 0.09434138238430023,
+      "learning_rate": 9.106861248091318e-06,
+      "loss": 8.626,
+      "step": 139610
+    },
+    {
+      "epoch": 0.6972458738045894,
+      "grad_norm": 0.09279754012823105,
+      "learning_rate": 9.105359333149766e-06,
+      "loss": 8.627,
+      "step": 139620
+    },
+    {
+      "epoch": 0.6972958126295288,
+      "grad_norm": 0.08741941303014755,
+      "learning_rate": 9.103857418208215e-06,
+      "loss": 8.6508,
+      "step": 139630
+    },
+    {
+      "epoch": 0.6973457514544683,
+      "grad_norm": 0.09483351558446884,
+      "learning_rate": 9.102355503266665e-06,
+      "loss": 8.6326,
+      "step": 139640
+    },
+    {
+      "epoch": 0.6973956902794077,
+      "grad_norm": 0.09228503704071045,
+      "learning_rate": 9.100853588325115e-06,
+      "loss": 8.6424,
+      "step": 139650
+    },
+    {
+      "epoch": 0.6974456291043472,
+      "grad_norm": 0.08870236575603485,
+      "learning_rate": 9.099351673383565e-06,
+      "loss": 8.6325,
+      "step": 139660
+    },
+    {
+      "epoch": 0.6974955679292866,
+      "grad_norm": 0.09134595096111298,
+      "learning_rate": 9.097849758442014e-06,
+      "loss": 8.6322,
+      "step": 139670
+    },
+    {
+      "epoch": 0.6975455067542261,
+      "grad_norm": 0.09109502285718918,
+      "learning_rate": 9.096347843500462e-06,
+      "loss": 8.636,
+      "step": 139680
+    },
+    {
+      "epoch": 0.6975954455791655,
+      "grad_norm": 0.09011939913034439,
+      "learning_rate": 9.094845928558912e-06,
+      "loss": 8.6288,
+      "step": 139690
+    },
+    {
+      "epoch": 0.697645384404105,
+      "grad_norm": 0.09101170301437378,
+      "learning_rate": 9.093344013617363e-06,
+      "loss": 8.6427,
+      "step": 139700
+    },
+    {
+      "epoch": 0.6976953232290444,
+      "grad_norm": 0.09532671421766281,
+      "learning_rate": 9.091842098675813e-06,
+      "loss": 8.6448,
+      "step": 139710
+    },
+    {
+      "epoch": 0.6977452620539839,
+      "grad_norm": 0.08789071440696716,
+      "learning_rate": 9.090340183734261e-06,
+      "loss": 8.6285,
+      "step": 139720
+    },
+    {
+      "epoch": 0.6977952008789233,
+      "grad_norm": 0.08861180394887924,
+      "learning_rate": 9.08883826879271e-06,
+      "loss": 8.6353,
+      "step": 139730
+    },
+    {
+      "epoch": 0.6978451397038627,
+      "grad_norm": 0.0869184210896492,
+      "learning_rate": 9.08733635385116e-06,
+      "loss": 8.6337,
+      "step": 139740
+    },
+    {
+      "epoch": 0.6978950785288022,
+      "grad_norm": 0.09149453043937683,
+      "learning_rate": 9.08583443890961e-06,
+      "loss": 8.631,
+      "step": 139750
+    },
+    {
+      "epoch": 0.6979450173537417,
+      "grad_norm": 0.09057032316923141,
+      "learning_rate": 9.08433252396806e-06,
+      "loss": 8.6373,
+      "step": 139760
+    },
+    {
+      "epoch": 0.6979949561786811,
+      "grad_norm": 0.08867466449737549,
+      "learning_rate": 9.082830609026509e-06,
+      "loss": 8.6174,
+      "step": 139770
+    },
+    {
+      "epoch": 0.6980448950036205,
+      "grad_norm": 0.09577593952417374,
+      "learning_rate": 9.081328694084959e-06,
+      "loss": 8.6414,
+      "step": 139780
+    },
+    {
+      "epoch": 0.69809483382856,
+      "grad_norm": 0.08875248581171036,
+      "learning_rate": 9.079826779143407e-06,
+      "loss": 8.6312,
+      "step": 139790
+    },
+    {
+      "epoch": 0.6981447726534995,
+      "grad_norm": 0.09105167537927628,
+      "learning_rate": 9.078324864201858e-06,
+      "loss": 8.6298,
+      "step": 139800
+    },
+    {
+      "epoch": 0.6981947114784389,
+      "grad_norm": 0.09408155083656311,
+      "learning_rate": 9.076822949260308e-06,
+      "loss": 8.6485,
+      "step": 139810
+    },
+    {
+      "epoch": 0.6982446503033783,
+      "grad_norm": 0.09340722113847733,
+      "learning_rate": 9.075321034318756e-06,
+      "loss": 8.6215,
+      "step": 139820
+    },
+    {
+      "epoch": 0.6982945891283178,
+      "grad_norm": 0.10109424591064453,
+      "learning_rate": 9.073819119377206e-06,
+      "loss": 8.631,
+      "step": 139830
+    },
+    {
+      "epoch": 0.6983445279532573,
+      "grad_norm": 0.0897102952003479,
+      "learning_rate": 9.072317204435655e-06,
+      "loss": 8.6321,
+      "step": 139840
+    },
+    {
+      "epoch": 0.6983944667781967,
+      "grad_norm": 0.09369130432605743,
+      "learning_rate": 9.070815289494105e-06,
+      "loss": 8.633,
+      "step": 139850
+    },
+    {
+      "epoch": 0.6984444056031361,
+      "grad_norm": 0.09355512261390686,
+      "learning_rate": 9.069313374552555e-06,
+      "loss": 8.6267,
+      "step": 139860
+    },
+    {
+      "epoch": 0.6984943444280756,
+      "grad_norm": 0.09666591137647629,
+      "learning_rate": 9.067811459611004e-06,
+      "loss": 8.6556,
+      "step": 139870
+    },
+    {
+      "epoch": 0.6985442832530151,
+      "grad_norm": 0.09669771045446396,
+      "learning_rate": 9.066309544669454e-06,
+      "loss": 8.6225,
+      "step": 139880
+    },
+    {
+      "epoch": 0.6985942220779545,
+      "grad_norm": 0.0945110023021698,
+      "learning_rate": 9.064807629727902e-06,
+      "loss": 8.6422,
+      "step": 139890
+    },
+    {
+      "epoch": 0.6986441609028939,
+      "grad_norm": 0.09108574688434601,
+      "learning_rate": 9.063305714786353e-06,
+      "loss": 8.6459,
+      "step": 139900
+    },
+    {
+      "epoch": 0.6986940997278334,
+      "grad_norm": 0.09179191291332245,
+      "learning_rate": 9.061803799844803e-06,
+      "loss": 8.6282,
+      "step": 139910
+    },
+    {
+      "epoch": 0.6987440385527729,
+      "grad_norm": 0.08860944956541061,
+      "learning_rate": 9.060301884903251e-06,
+      "loss": 8.6106,
+      "step": 139920
+    },
+    {
+      "epoch": 0.6987939773777123,
+      "grad_norm": 0.09390242397785187,
+      "learning_rate": 9.058799969961701e-06,
+      "loss": 8.622,
+      "step": 139930
+    },
+    {
+      "epoch": 0.6988439162026517,
+      "grad_norm": 0.08761897683143616,
+      "learning_rate": 9.057298055020152e-06,
+      "loss": 8.6207,
+      "step": 139940
+    },
+    {
+      "epoch": 0.6988938550275912,
+      "grad_norm": 0.0982297733426094,
+      "learning_rate": 9.0557961400786e-06,
+      "loss": 8.6384,
+      "step": 139950
+    },
+    {
+      "epoch": 0.6989437938525307,
+      "grad_norm": 0.09506106376647949,
+      "learning_rate": 9.05429422513705e-06,
+      "loss": 8.6514,
+      "step": 139960
+    },
+    {
+      "epoch": 0.6989937326774701,
+      "grad_norm": 0.09104983508586884,
+      "learning_rate": 9.052792310195499e-06,
+      "loss": 8.629,
+      "step": 139970
+    },
+    {
+      "epoch": 0.6990436715024095,
+      "grad_norm": 0.08770845830440521,
+      "learning_rate": 9.051290395253949e-06,
+      "loss": 8.6221,
+      "step": 139980
+    },
+    {
+      "epoch": 0.699093610327349,
+      "grad_norm": 0.09654659777879715,
+      "learning_rate": 9.0497884803124e-06,
+      "loss": 8.6329,
+      "step": 139990
+    },
+    {
+      "epoch": 0.6991435491522885,
+      "grad_norm": 0.0931318923830986,
+      "learning_rate": 9.048286565370848e-06,
+      "loss": 8.6416,
+      "step": 140000
+    },
+    {
+      "epoch": 0.6991934879772279,
+      "grad_norm": 0.09113478660583496,
+      "learning_rate": 9.046784650429298e-06,
+      "loss": 8.6452,
+      "step": 140010
+    },
+    {
+      "epoch": 0.6992434268021673,
+      "grad_norm": 0.09566029906272888,
+      "learning_rate": 9.045282735487746e-06,
+      "loss": 8.6487,
+      "step": 140020
+    },
+    {
+      "epoch": 0.6992933656271068,
+      "grad_norm": 0.08825311809778214,
+      "learning_rate": 9.043780820546197e-06,
+      "loss": 8.6336,
+      "step": 140030
+    },
+    {
+      "epoch": 0.6993433044520463,
+      "grad_norm": 0.09055044502019882,
+      "learning_rate": 9.042278905604647e-06,
+      "loss": 8.6425,
+      "step": 140040
+    },
+    {
+      "epoch": 0.6993932432769857,
+      "grad_norm": 0.08617035299539566,
+      "learning_rate": 9.040776990663095e-06,
+      "loss": 8.6324,
+      "step": 140050
+    },
+    {
+      "epoch": 0.6994431821019251,
+      "grad_norm": 0.09692193567752838,
+      "learning_rate": 9.039275075721545e-06,
+      "loss": 8.6194,
+      "step": 140060
+    },
+    {
+      "epoch": 0.6994931209268646,
+      "grad_norm": 0.09942319244146347,
+      "learning_rate": 9.037773160779994e-06,
+      "loss": 8.6313,
+      "step": 140070
+    },
+    {
+      "epoch": 0.6995430597518041,
+      "grad_norm": 0.0971224308013916,
+      "learning_rate": 9.036271245838444e-06,
+      "loss": 8.6337,
+      "step": 140080
+    },
+    {
+      "epoch": 0.6995929985767435,
+      "grad_norm": 0.09974560886621475,
+      "learning_rate": 9.034769330896894e-06,
+      "loss": 8.6208,
+      "step": 140090
+    },
+    {
+      "epoch": 0.6996429374016829,
+      "grad_norm": 0.09148532152175903,
+      "learning_rate": 9.033267415955344e-06,
+      "loss": 8.6361,
+      "step": 140100
+    },
+    {
+      "epoch": 0.6996928762266224,
+      "grad_norm": 0.09783138334751129,
+      "learning_rate": 9.031765501013793e-06,
+      "loss": 8.6281,
+      "step": 140110
+    },
+    {
+      "epoch": 0.6997428150515619,
+      "grad_norm": 0.09642284363508224,
+      "learning_rate": 9.030263586072241e-06,
+      "loss": 8.6295,
+      "step": 140120
+    },
+    {
+      "epoch": 0.6997927538765013,
+      "grad_norm": 0.088691346347332,
+      "learning_rate": 9.028761671130692e-06,
+      "loss": 8.6492,
+      "step": 140130
+    },
+    {
+      "epoch": 0.6998426927014407,
+      "grad_norm": 0.09170258790254593,
+      "learning_rate": 9.027259756189142e-06,
+      "loss": 8.6167,
+      "step": 140140
+    },
+    {
+      "epoch": 0.6998926315263801,
+      "grad_norm": 0.09011033177375793,
+      "learning_rate": 9.025757841247592e-06,
+      "loss": 8.6359,
+      "step": 140150
+    },
+    {
+      "epoch": 0.6999425703513197,
+      "grad_norm": 0.09351085126399994,
+      "learning_rate": 9.02425592630604e-06,
+      "loss": 8.6323,
+      "step": 140160
+    },
+    {
+      "epoch": 0.6999925091762591,
+      "grad_norm": 0.09307374805212021,
+      "learning_rate": 9.022754011364489e-06,
+      "loss": 8.6389,
+      "step": 140170
+    },
+    {
+      "epoch": 0.7000424480011985,
+      "grad_norm": 0.09782139211893082,
+      "learning_rate": 9.021252096422939e-06,
+      "loss": 8.6427,
+      "step": 140180
+    },
+    {
+      "epoch": 0.700092386826138,
+      "grad_norm": 0.09016304463148117,
+      "learning_rate": 9.01975018148139e-06,
+      "loss": 8.6376,
+      "step": 140190
+    },
+    {
+      "epoch": 0.7001423256510775,
+      "grad_norm": 0.0927334874868393,
+      "learning_rate": 9.01824826653984e-06,
+      "loss": 8.6239,
+      "step": 140200
+    },
+    {
+      "epoch": 0.7001922644760169,
+      "grad_norm": 0.09416099637746811,
+      "learning_rate": 9.016746351598288e-06,
+      "loss": 8.6107,
+      "step": 140210
+    },
+    {
+      "epoch": 0.7002422033009563,
+      "grad_norm": 0.09217461943626404,
+      "learning_rate": 9.015244436656736e-06,
+      "loss": 8.6225,
+      "step": 140220
+    },
+    {
+      "epoch": 0.7002921421258957,
+      "grad_norm": 0.08289490640163422,
+      "learning_rate": 9.013742521715187e-06,
+      "loss": 8.6443,
+      "step": 140230
+    },
+    {
+      "epoch": 0.7003420809508353,
+      "grad_norm": 0.08873941749334335,
+      "learning_rate": 9.012240606773637e-06,
+      "loss": 8.6334,
+      "step": 140240
+    },
+    {
+      "epoch": 0.7003920197757747,
+      "grad_norm": 0.09045631438493729,
+      "learning_rate": 9.010738691832087e-06,
+      "loss": 8.631,
+      "step": 140250
+    },
+    {
+      "epoch": 0.7004419586007141,
+      "grad_norm": 0.08844128996133804,
+      "learning_rate": 9.009236776890537e-06,
+      "loss": 8.6378,
+      "step": 140260
+    },
+    {
+      "epoch": 0.7004918974256535,
+      "grad_norm": 0.09297400712966919,
+      "learning_rate": 9.007734861948984e-06,
+      "loss": 8.6214,
+      "step": 140270
+    },
+    {
+      "epoch": 0.7005418362505931,
+      "grad_norm": 0.09032850712537766,
+      "learning_rate": 9.006232947007434e-06,
+      "loss": 8.6302,
+      "step": 140280
+    },
+    {
+      "epoch": 0.7005917750755325,
+      "grad_norm": 0.09603878855705261,
+      "learning_rate": 9.004731032065884e-06,
+      "loss": 8.6308,
+      "step": 140290
+    },
+    {
+      "epoch": 0.7006417139004719,
+      "grad_norm": 0.09244571626186371,
+      "learning_rate": 9.003229117124334e-06,
+      "loss": 8.6455,
+      "step": 140300
+    },
+    {
+      "epoch": 0.7006916527254113,
+      "grad_norm": 0.09039461612701416,
+      "learning_rate": 9.001727202182785e-06,
+      "loss": 8.6357,
+      "step": 140310
+    },
+    {
+      "epoch": 0.7007415915503509,
+      "grad_norm": 0.09459111839532852,
+      "learning_rate": 9.000225287241231e-06,
+      "loss": 8.6303,
+      "step": 140320
+    },
+    {
+      "epoch": 0.7007915303752903,
+      "grad_norm": 0.09677302837371826,
+      "learning_rate": 8.998723372299682e-06,
+      "loss": 8.6394,
+      "step": 140330
+    },
+    {
+      "epoch": 0.7008414692002297,
+      "grad_norm": 0.08825703710317612,
+      "learning_rate": 8.997221457358132e-06,
+      "loss": 8.6318,
+      "step": 140340
+    },
+    {
+      "epoch": 0.7008914080251691,
+      "grad_norm": 0.09611725807189941,
+      "learning_rate": 8.995719542416582e-06,
+      "loss": 8.6359,
+      "step": 140350
+    },
+    {
+      "epoch": 0.7009413468501087,
+      "grad_norm": 0.09157848358154297,
+      "learning_rate": 8.994217627475032e-06,
+      "loss": 8.6286,
+      "step": 140360
+    },
+    {
+      "epoch": 0.7009912856750481,
+      "grad_norm": 0.08757877349853516,
+      "learning_rate": 8.992715712533479e-06,
+      "loss": 8.639,
+      "step": 140370
+    },
+    {
+      "epoch": 0.7010412244999875,
+      "grad_norm": 0.09328636527061462,
+      "learning_rate": 8.991213797591929e-06,
+      "loss": 8.6316,
+      "step": 140380
+    },
+    {
+      "epoch": 0.7010911633249269,
+      "grad_norm": 0.08965921401977539,
+      "learning_rate": 8.98971188265038e-06,
+      "loss": 8.624,
+      "step": 140390
+    },
+    {
+      "epoch": 0.7011411021498665,
+      "grad_norm": 0.0938333123922348,
+      "learning_rate": 8.98820996770883e-06,
+      "loss": 8.6466,
+      "step": 140400
+    },
+    {
+      "epoch": 0.7011910409748059,
+      "grad_norm": 0.08751453459262848,
+      "learning_rate": 8.98670805276728e-06,
+      "loss": 8.6313,
+      "step": 140410
+    },
+    {
+      "epoch": 0.7012409797997453,
+      "grad_norm": 0.09409631788730621,
+      "learning_rate": 8.985206137825728e-06,
+      "loss": 8.6354,
+      "step": 140420
+    },
+    {
+      "epoch": 0.7012909186246847,
+      "grad_norm": 0.09224843233823776,
+      "learning_rate": 8.983704222884177e-06,
+      "loss": 8.646,
+      "step": 140430
+    },
+    {
+      "epoch": 0.7013408574496242,
+      "grad_norm": 0.08712205290794373,
+      "learning_rate": 8.982202307942627e-06,
+      "loss": 8.6293,
+      "step": 140440
+    },
+    {
+      "epoch": 0.7013907962745637,
+      "grad_norm": 0.08894158154726028,
+      "learning_rate": 8.980700393001077e-06,
+      "loss": 8.6525,
+      "step": 140450
+    },
+    {
+      "epoch": 0.7014407350995031,
+      "grad_norm": 0.09181137382984161,
+      "learning_rate": 8.979198478059527e-06,
+      "loss": 8.6246,
+      "step": 140460
+    },
+    {
+      "epoch": 0.7014906739244425,
+      "grad_norm": 0.08507677912712097,
+      "learning_rate": 8.977696563117976e-06,
+      "loss": 8.6357,
+      "step": 140470
+    },
+    {
+      "epoch": 0.701540612749382,
+      "grad_norm": 0.09141664952039719,
+      "learning_rate": 8.976194648176424e-06,
+      "loss": 8.6319,
+      "step": 140480
+    },
+    {
+      "epoch": 0.7015905515743215,
+      "grad_norm": 0.09606952220201492,
+      "learning_rate": 8.974692733234874e-06,
+      "loss": 8.6253,
+      "step": 140490
+    },
+    {
+      "epoch": 0.7016404903992609,
+      "grad_norm": 0.0922846719622612,
+      "learning_rate": 8.973190818293324e-06,
+      "loss": 8.6208,
+      "step": 140500
+    },
+    {
+      "epoch": 0.7016904292242003,
+      "grad_norm": 0.0884494036436081,
+      "learning_rate": 8.971688903351775e-06,
+      "loss": 8.6359,
+      "step": 140510
+    },
+    {
+      "epoch": 0.7017403680491398,
+      "grad_norm": 0.09224381297826767,
+      "learning_rate": 8.970186988410223e-06,
+      "loss": 8.6228,
+      "step": 140520
+    },
+    {
+      "epoch": 0.7017903068740793,
+      "grad_norm": 0.09082509577274323,
+      "learning_rate": 8.968685073468672e-06,
+      "loss": 8.6334,
+      "step": 140530
+    },
+    {
+      "epoch": 0.7018402456990187,
+      "grad_norm": 0.0936235710978508,
+      "learning_rate": 8.967183158527122e-06,
+      "loss": 8.6158,
+      "step": 140540
+    },
+    {
+      "epoch": 0.7018901845239581,
+      "grad_norm": 0.08928533643484116,
+      "learning_rate": 8.965681243585572e-06,
+      "loss": 8.6382,
+      "step": 140550
+    },
+    {
+      "epoch": 0.7019401233488975,
+      "grad_norm": 0.08770529180765152,
+      "learning_rate": 8.964179328644022e-06,
+      "loss": 8.6297,
+      "step": 140560
+    },
+    {
+      "epoch": 0.7019900621738371,
+      "grad_norm": 0.08795421570539474,
+      "learning_rate": 8.96267741370247e-06,
+      "loss": 8.6302,
+      "step": 140570
+    },
+    {
+      "epoch": 0.7020400009987765,
+      "grad_norm": 0.09204691648483276,
+      "learning_rate": 8.961175498760919e-06,
+      "loss": 8.6417,
+      "step": 140580
+    },
+    {
+      "epoch": 0.7020899398237159,
+      "grad_norm": 0.09336613863706589,
+      "learning_rate": 8.95967358381937e-06,
+      "loss": 8.6493,
+      "step": 140590
+    },
+    {
+      "epoch": 0.7021398786486553,
+      "grad_norm": 0.09377633780241013,
+      "learning_rate": 8.95817166887782e-06,
+      "loss": 8.6375,
+      "step": 140600
+    },
+    {
+      "epoch": 0.7021898174735949,
+      "grad_norm": 0.0860079750418663,
+      "learning_rate": 8.95666975393627e-06,
+      "loss": 8.6426,
+      "step": 140610
+    },
+    {
+      "epoch": 0.7022397562985343,
+      "grad_norm": 0.0926608294248581,
+      "learning_rate": 8.95516783899472e-06,
+      "loss": 8.6271,
+      "step": 140620
+    },
+    {
+      "epoch": 0.7022896951234737,
+      "grad_norm": 0.09390340000391006,
+      "learning_rate": 8.953665924053168e-06,
+      "loss": 8.6388,
+      "step": 140630
+    },
+    {
+      "epoch": 0.7023396339484131,
+      "grad_norm": 0.09352290630340576,
+      "learning_rate": 8.952164009111617e-06,
+      "loss": 8.6309,
+      "step": 140640
+    },
+    {
+      "epoch": 0.7023895727733527,
+      "grad_norm": 0.09155908972024918,
+      "learning_rate": 8.950662094170067e-06,
+      "loss": 8.6282,
+      "step": 140650
+    },
+    {
+      "epoch": 0.7024395115982921,
+      "grad_norm": 0.08889105916023254,
+      "learning_rate": 8.949160179228517e-06,
+      "loss": 8.6313,
+      "step": 140660
+    },
+    {
+      "epoch": 0.7024894504232315,
+      "grad_norm": 0.0867091491818428,
+      "learning_rate": 8.947658264286967e-06,
+      "loss": 8.6344,
+      "step": 140670
+    },
+    {
+      "epoch": 0.7025393892481709,
+      "grad_norm": 0.08940424025058746,
+      "learning_rate": 8.946156349345416e-06,
+      "loss": 8.641,
+      "step": 140680
+    },
+    {
+      "epoch": 0.7025893280731105,
+      "grad_norm": 0.09525980055332184,
+      "learning_rate": 8.944654434403864e-06,
+      "loss": 8.6463,
+      "step": 140690
+    },
+    {
+      "epoch": 0.7026392668980499,
+      "grad_norm": 0.08965422958135605,
+      "learning_rate": 8.943152519462314e-06,
+      "loss": 8.6191,
+      "step": 140700
+    },
+    {
+      "epoch": 0.7026892057229893,
+      "grad_norm": 0.09086750447750092,
+      "learning_rate": 8.941650604520765e-06,
+      "loss": 8.6479,
+      "step": 140710
+    },
+    {
+      "epoch": 0.7027391445479287,
+      "grad_norm": 0.08906187117099762,
+      "learning_rate": 8.940148689579215e-06,
+      "loss": 8.6292,
+      "step": 140720
+    },
+    {
+      "epoch": 0.7027890833728683,
+      "grad_norm": 0.09022345393896103,
+      "learning_rate": 8.938646774637663e-06,
+      "loss": 8.6404,
+      "step": 140730
+    },
+    {
+      "epoch": 0.7028390221978077,
+      "grad_norm": 0.08862177282571793,
+      "learning_rate": 8.937144859696112e-06,
+      "loss": 8.63,
+      "step": 140740
+    },
+    {
+      "epoch": 0.7028889610227471,
+      "grad_norm": 0.09106519818305969,
+      "learning_rate": 8.935642944754562e-06,
+      "loss": 8.6177,
+      "step": 140750
+    },
+    {
+      "epoch": 0.7029388998476865,
+      "grad_norm": 0.09161339700222015,
+      "learning_rate": 8.934141029813012e-06,
+      "loss": 8.6267,
+      "step": 140760
+    },
+    {
+      "epoch": 0.7029888386726261,
+      "grad_norm": 0.0945722833275795,
+      "learning_rate": 8.932639114871462e-06,
+      "loss": 8.6188,
+      "step": 140770
+    },
+    {
+      "epoch": 0.7030387774975655,
+      "grad_norm": 0.0907975360751152,
+      "learning_rate": 8.93113719992991e-06,
+      "loss": 8.6323,
+      "step": 140780
+    },
+    {
+      "epoch": 0.7030887163225049,
+      "grad_norm": 0.08957850933074951,
+      "learning_rate": 8.929635284988361e-06,
+      "loss": 8.6396,
+      "step": 140790
+    },
+    {
+      "epoch": 0.7031386551474443,
+      "grad_norm": 0.0977301150560379,
+      "learning_rate": 8.92813337004681e-06,
+      "loss": 8.6227,
+      "step": 140800
+    },
+    {
+      "epoch": 0.7031885939723839,
+      "grad_norm": 0.09034769982099533,
+      "learning_rate": 8.92663145510526e-06,
+      "loss": 8.6368,
+      "step": 140810
+    },
+    {
+      "epoch": 0.7032385327973233,
+      "grad_norm": 0.0880945548415184,
+      "learning_rate": 8.92512954016371e-06,
+      "loss": 8.6343,
+      "step": 140820
+    },
+    {
+      "epoch": 0.7032884716222627,
+      "grad_norm": 0.09109944850206375,
+      "learning_rate": 8.923627625222158e-06,
+      "loss": 8.6357,
+      "step": 140830
+    },
+    {
+      "epoch": 0.7033384104472021,
+      "grad_norm": 0.0931500494480133,
+      "learning_rate": 8.922125710280608e-06,
+      "loss": 8.6303,
+      "step": 140840
+    },
+    {
+      "epoch": 0.7033883492721417,
+      "grad_norm": 0.09611831605434418,
+      "learning_rate": 8.920623795339057e-06,
+      "loss": 8.62,
+      "step": 140850
+    },
+    {
+      "epoch": 0.7034382880970811,
+      "grad_norm": 0.08877561241388321,
+      "learning_rate": 8.919121880397507e-06,
+      "loss": 8.6357,
+      "step": 140860
+    },
+    {
+      "epoch": 0.7034882269220205,
+      "grad_norm": 0.0883457213640213,
+      "learning_rate": 8.917619965455957e-06,
+      "loss": 8.6247,
+      "step": 140870
+    },
+    {
+      "epoch": 0.7035381657469599,
+      "grad_norm": 0.08729522675275803,
+      "learning_rate": 8.916118050514406e-06,
+      "loss": 8.6459,
+      "step": 140880
+    },
+    {
+      "epoch": 0.7035881045718995,
+      "grad_norm": 0.09188755601644516,
+      "learning_rate": 8.914616135572856e-06,
+      "loss": 8.6239,
+      "step": 140890
+    },
+    {
+      "epoch": 0.7036380433968389,
+      "grad_norm": 0.09601496160030365,
+      "learning_rate": 8.913114220631304e-06,
+      "loss": 8.6126,
+      "step": 140900
+    },
+    {
+      "epoch": 0.7036879822217783,
+      "grad_norm": 0.0922279953956604,
+      "learning_rate": 8.911612305689755e-06,
+      "loss": 8.6379,
+      "step": 140910
+    },
+    {
+      "epoch": 0.7037379210467177,
+      "grad_norm": 0.08827944844961166,
+      "learning_rate": 8.910110390748205e-06,
+      "loss": 8.6256,
+      "step": 140920
+    },
+    {
+      "epoch": 0.7037878598716573,
+      "grad_norm": 0.09115996211767197,
+      "learning_rate": 8.908608475806653e-06,
+      "loss": 8.6132,
+      "step": 140930
+    },
+    {
+      "epoch": 0.7038377986965967,
+      "grad_norm": 0.09394663572311401,
+      "learning_rate": 8.907106560865103e-06,
+      "loss": 8.6283,
+      "step": 140940
+    },
+    {
+      "epoch": 0.7038877375215361,
+      "grad_norm": 0.08945653587579727,
+      "learning_rate": 8.905604645923554e-06,
+      "loss": 8.6324,
+      "step": 140950
+    },
+    {
+      "epoch": 0.7039376763464755,
+      "grad_norm": 0.09907026588916779,
+      "learning_rate": 8.904102730982002e-06,
+      "loss": 8.6333,
+      "step": 140960
+    },
+    {
+      "epoch": 0.7039876151714151,
+      "grad_norm": 0.09264247864484787,
+      "learning_rate": 8.902600816040452e-06,
+      "loss": 8.6323,
+      "step": 140970
+    },
+    {
+      "epoch": 0.7040375539963545,
+      "grad_norm": 0.08847159892320633,
+      "learning_rate": 8.9010989010989e-06,
+      "loss": 8.6337,
+      "step": 140980
+    },
+    {
+      "epoch": 0.7040874928212939,
+      "grad_norm": 0.08942971378564835,
+      "learning_rate": 8.899596986157351e-06,
+      "loss": 8.6253,
+      "step": 140990
+    },
+    {
+      "epoch": 0.7041374316462333,
+      "grad_norm": 0.0965491458773613,
+      "learning_rate": 8.898095071215801e-06,
+      "loss": 8.6361,
+      "step": 141000
+    },
+    {
+      "epoch": 0.7041873704711729,
+      "grad_norm": 0.09533529728651047,
+      "learning_rate": 8.89659315627425e-06,
+      "loss": 8.6196,
+      "step": 141010
+    },
+    {
+      "epoch": 0.7042373092961123,
+      "grad_norm": 0.09187507629394531,
+      "learning_rate": 8.8950912413327e-06,
+      "loss": 8.6384,
+      "step": 141020
+    },
+    {
+      "epoch": 0.7042872481210517,
+      "grad_norm": 0.09570582956075668,
+      "learning_rate": 8.893589326391148e-06,
+      "loss": 8.6171,
+      "step": 141030
+    },
+    {
+      "epoch": 0.7043371869459911,
+      "grad_norm": 0.09152864664793015,
+      "learning_rate": 8.892087411449598e-06,
+      "loss": 8.6202,
+      "step": 141040
+    },
+    {
+      "epoch": 0.7043871257709307,
+      "grad_norm": 0.08897831290960312,
+      "learning_rate": 8.890585496508049e-06,
+      "loss": 8.6357,
+      "step": 141050
+    },
+    {
+      "epoch": 0.7044370645958701,
+      "grad_norm": 0.09181348979473114,
+      "learning_rate": 8.889083581566497e-06,
+      "loss": 8.6106,
+      "step": 141060
+    },
+    {
+      "epoch": 0.7044870034208095,
+      "grad_norm": 0.09355124831199646,
+      "learning_rate": 8.887581666624947e-06,
+      "loss": 8.631,
+      "step": 141070
+    },
+    {
+      "epoch": 0.7045369422457489,
+      "grad_norm": 0.08978011459112167,
+      "learning_rate": 8.886079751683396e-06,
+      "loss": 8.6163,
+      "step": 141080
+    },
+    {
+      "epoch": 0.7045868810706885,
+      "grad_norm": 0.09040986001491547,
+      "learning_rate": 8.884577836741846e-06,
+      "loss": 8.6161,
+      "step": 141090
+    },
+    {
+      "epoch": 0.7046368198956279,
+      "grad_norm": 0.09216012805700302,
+      "learning_rate": 8.883075921800296e-06,
+      "loss": 8.6069,
+      "step": 141100
+    },
+    {
+      "epoch": 0.7046867587205673,
+      "grad_norm": 0.09498348832130432,
+      "learning_rate": 8.881574006858746e-06,
+      "loss": 8.6323,
+      "step": 141110
+    },
+    {
+      "epoch": 0.7047366975455067,
+      "grad_norm": 0.09142297506332397,
+      "learning_rate": 8.880072091917195e-06,
+      "loss": 8.6146,
+      "step": 141120
+    },
+    {
+      "epoch": 0.7047866363704463,
+      "grad_norm": 0.08392765372991562,
+      "learning_rate": 8.878570176975643e-06,
+      "loss": 8.6236,
+      "step": 141130
+    },
+    {
+      "epoch": 0.7048365751953857,
+      "grad_norm": 0.08954204618930817,
+      "learning_rate": 8.877068262034093e-06,
+      "loss": 8.6459,
+      "step": 141140
+    },
+    {
+      "epoch": 0.7048865140203251,
+      "grad_norm": 0.09618078172206879,
+      "learning_rate": 8.875566347092544e-06,
+      "loss": 8.6317,
+      "step": 141150
+    },
+    {
+      "epoch": 0.7049364528452645,
+      "grad_norm": 0.09143991023302078,
+      "learning_rate": 8.874064432150994e-06,
+      "loss": 8.6466,
+      "step": 141160
+    },
+    {
+      "epoch": 0.704986391670204,
+      "grad_norm": 0.09367083013057709,
+      "learning_rate": 8.872562517209442e-06,
+      "loss": 8.6332,
+      "step": 141170
+    },
+    {
+      "epoch": 0.7050363304951435,
+      "grad_norm": 0.09456432610750198,
+      "learning_rate": 8.87106060226789e-06,
+      "loss": 8.6154,
+      "step": 141180
+    },
+    {
+      "epoch": 0.7050862693200829,
+      "grad_norm": 0.09094260632991791,
+      "learning_rate": 8.869558687326341e-06,
+      "loss": 8.6337,
+      "step": 141190
+    },
+    {
+      "epoch": 0.7051362081450223,
+      "grad_norm": 0.09079168736934662,
+      "learning_rate": 8.868056772384791e-06,
+      "loss": 8.6399,
+      "step": 141200
+    },
+    {
+      "epoch": 0.7051861469699618,
+      "grad_norm": 0.08768593519926071,
+      "learning_rate": 8.866554857443241e-06,
+      "loss": 8.6241,
+      "step": 141210
+    },
+    {
+      "epoch": 0.7052360857949013,
+      "grad_norm": 0.09279026836156845,
+      "learning_rate": 8.86505294250169e-06,
+      "loss": 8.6303,
+      "step": 141220
+    },
+    {
+      "epoch": 0.7052860246198407,
+      "grad_norm": 0.09045767784118652,
+      "learning_rate": 8.863551027560138e-06,
+      "loss": 8.6399,
+      "step": 141230
+    },
+    {
+      "epoch": 0.7053359634447801,
+      "grad_norm": 0.09411778301000595,
+      "learning_rate": 8.862049112618588e-06,
+      "loss": 8.6315,
+      "step": 141240
+    },
+    {
+      "epoch": 0.7053859022697196,
+      "grad_norm": 0.08939050883054733,
+      "learning_rate": 8.860547197677039e-06,
+      "loss": 8.6213,
+      "step": 141250
+    },
+    {
+      "epoch": 0.7054358410946591,
+      "grad_norm": 0.08885614573955536,
+      "learning_rate": 8.859045282735489e-06,
+      "loss": 8.6282,
+      "step": 141260
+    },
+    {
+      "epoch": 0.7054857799195985,
+      "grad_norm": 0.09590033441781998,
+      "learning_rate": 8.857543367793939e-06,
+      "loss": 8.6296,
+      "step": 141270
+    },
+    {
+      "epoch": 0.7055357187445379,
+      "grad_norm": 0.09497443586587906,
+      "learning_rate": 8.856041452852386e-06,
+      "loss": 8.6256,
+      "step": 141280
+    },
+    {
+      "epoch": 0.7055856575694774,
+      "grad_norm": 0.09108743816614151,
+      "learning_rate": 8.854539537910836e-06,
+      "loss": 8.6314,
+      "step": 141290
+    },
+    {
+      "epoch": 0.7056355963944169,
+      "grad_norm": 0.08908271044492722,
+      "learning_rate": 8.853037622969286e-06,
+      "loss": 8.6233,
+      "step": 141300
+    },
+    {
+      "epoch": 0.7056855352193563,
+      "grad_norm": 0.10099402815103531,
+      "learning_rate": 8.851535708027736e-06,
+      "loss": 8.6468,
+      "step": 141310
+    },
+    {
+      "epoch": 0.7057354740442957,
+      "grad_norm": 0.09605363756418228,
+      "learning_rate": 8.850033793086187e-06,
+      "loss": 8.6166,
+      "step": 141320
+    },
+    {
+      "epoch": 0.7057854128692352,
+      "grad_norm": 0.09077009558677673,
+      "learning_rate": 8.848531878144633e-06,
+      "loss": 8.6215,
+      "step": 141330
+    },
+    {
+      "epoch": 0.7058353516941747,
+      "grad_norm": 0.0913512110710144,
+      "learning_rate": 8.847029963203083e-06,
+      "loss": 8.6587,
+      "step": 141340
+    },
+    {
+      "epoch": 0.7058852905191141,
+      "grad_norm": 0.08867861330509186,
+      "learning_rate": 8.845528048261534e-06,
+      "loss": 8.6569,
+      "step": 141350
+    },
+    {
+      "epoch": 0.7059352293440535,
+      "grad_norm": 0.09407979995012283,
+      "learning_rate": 8.844026133319984e-06,
+      "loss": 8.606,
+      "step": 141360
+    },
+    {
+      "epoch": 0.705985168168993,
+      "grad_norm": 0.0922011062502861,
+      "learning_rate": 8.842524218378434e-06,
+      "loss": 8.6163,
+      "step": 141370
+    },
+    {
+      "epoch": 0.7060351069939325,
+      "grad_norm": 0.09292416274547577,
+      "learning_rate": 8.84102230343688e-06,
+      "loss": 8.6203,
+      "step": 141380
+    },
+    {
+      "epoch": 0.7060850458188719,
+      "grad_norm": 0.09504501521587372,
+      "learning_rate": 8.839520388495331e-06,
+      "loss": 8.6259,
+      "step": 141390
+    },
+    {
+      "epoch": 0.7061349846438113,
+      "grad_norm": 0.09113211929798126,
+      "learning_rate": 8.838018473553781e-06,
+      "loss": 8.6264,
+      "step": 141400
+    },
+    {
+      "epoch": 0.7061849234687507,
+      "grad_norm": 0.09200391918420792,
+      "learning_rate": 8.836516558612231e-06,
+      "loss": 8.6251,
+      "step": 141410
+    },
+    {
+      "epoch": 0.7062348622936903,
+      "grad_norm": 0.09039604663848877,
+      "learning_rate": 8.835014643670682e-06,
+      "loss": 8.6115,
+      "step": 141420
+    },
+    {
+      "epoch": 0.7062848011186297,
+      "grad_norm": 0.0915713682770729,
+      "learning_rate": 8.83351272872913e-06,
+      "loss": 8.6198,
+      "step": 141430
+    },
+    {
+      "epoch": 0.7063347399435691,
+      "grad_norm": 0.09520621597766876,
+      "learning_rate": 8.832010813787578e-06,
+      "loss": 8.6043,
+      "step": 141440
+    },
+    {
+      "epoch": 0.7063846787685085,
+      "grad_norm": 0.0928700864315033,
+      "learning_rate": 8.830508898846029e-06,
+      "loss": 8.61,
+      "step": 141450
+    },
+    {
+      "epoch": 0.706434617593448,
+      "grad_norm": 0.09517066925764084,
+      "learning_rate": 8.829006983904479e-06,
+      "loss": 8.6289,
+      "step": 141460
+    },
+    {
+      "epoch": 0.7064845564183875,
+      "grad_norm": 0.0947510227560997,
+      "learning_rate": 8.827505068962929e-06,
+      "loss": 8.6077,
+      "step": 141470
+    },
+    {
+      "epoch": 0.7065344952433269,
+      "grad_norm": 0.09221941232681274,
+      "learning_rate": 8.826003154021377e-06,
+      "loss": 8.6342,
+      "step": 141480
+    },
+    {
+      "epoch": 0.7065844340682663,
+      "grad_norm": 0.08994131535291672,
+      "learning_rate": 8.824501239079826e-06,
+      "loss": 8.6346,
+      "step": 141490
+    },
+    {
+      "epoch": 0.7066343728932059,
+      "grad_norm": 0.09264985471963882,
+      "learning_rate": 8.822999324138276e-06,
+      "loss": 8.6291,
+      "step": 141500
+    },
+    {
+      "epoch": 0.7066843117181453,
+      "grad_norm": 0.09057199954986572,
+      "learning_rate": 8.821497409196726e-06,
+      "loss": 8.6444,
+      "step": 141510
+    },
+    {
+      "epoch": 0.7067342505430847,
+      "grad_norm": 0.08880162984132767,
+      "learning_rate": 8.819995494255177e-06,
+      "loss": 8.6288,
+      "step": 141520
+    },
+    {
+      "epoch": 0.7067841893680241,
+      "grad_norm": 0.08788347989320755,
+      "learning_rate": 8.818493579313625e-06,
+      "loss": 8.6211,
+      "step": 141530
+    },
+    {
+      "epoch": 0.7068341281929637,
+      "grad_norm": 0.08884553611278534,
+      "learning_rate": 8.816991664372073e-06,
+      "loss": 8.6423,
+      "step": 141540
+    },
+    {
+      "epoch": 0.7068840670179031,
+      "grad_norm": 0.090285524725914,
+      "learning_rate": 8.815489749430524e-06,
+      "loss": 8.6358,
+      "step": 141550
+    },
+    {
+      "epoch": 0.7069340058428425,
+      "grad_norm": 0.09516801685094833,
+      "learning_rate": 8.813987834488974e-06,
+      "loss": 8.618,
+      "step": 141560
+    },
+    {
+      "epoch": 0.7069839446677819,
+      "grad_norm": 0.09445168077945709,
+      "learning_rate": 8.812485919547424e-06,
+      "loss": 8.6284,
+      "step": 141570
+    },
+    {
+      "epoch": 0.7070338834927214,
+      "grad_norm": 0.08974064886569977,
+      "learning_rate": 8.810984004605873e-06,
+      "loss": 8.612,
+      "step": 141580
+    },
+    {
+      "epoch": 0.7070838223176609,
+      "grad_norm": 0.09538102895021439,
+      "learning_rate": 8.809482089664323e-06,
+      "loss": 8.6337,
+      "step": 141590
+    },
+    {
+      "epoch": 0.7071337611426003,
+      "grad_norm": 0.08882886171340942,
+      "learning_rate": 8.807980174722771e-06,
+      "loss": 8.6367,
+      "step": 141600
+    },
+    {
+      "epoch": 0.7071836999675397,
+      "grad_norm": 0.09137466549873352,
+      "learning_rate": 8.806478259781221e-06,
+      "loss": 8.6261,
+      "step": 141610
+    },
+    {
+      "epoch": 0.7072336387924792,
+      "grad_norm": 0.08818534016609192,
+      "learning_rate": 8.804976344839672e-06,
+      "loss": 8.6378,
+      "step": 141620
+    },
+    {
+      "epoch": 0.7072835776174187,
+      "grad_norm": 0.09256480634212494,
+      "learning_rate": 8.80347442989812e-06,
+      "loss": 8.6342,
+      "step": 141630
+    },
+    {
+      "epoch": 0.7073335164423581,
+      "grad_norm": 0.08914809674024582,
+      "learning_rate": 8.80197251495657e-06,
+      "loss": 8.6384,
+      "step": 141640
+    },
+    {
+      "epoch": 0.7073834552672975,
+      "grad_norm": 0.09401541948318481,
+      "learning_rate": 8.800470600015019e-06,
+      "loss": 8.6253,
+      "step": 141650
+    },
+    {
+      "epoch": 0.707433394092237,
+      "grad_norm": 0.09287466108798981,
+      "learning_rate": 8.798968685073469e-06,
+      "loss": 8.6103,
+      "step": 141660
+    },
+    {
+      "epoch": 0.7074833329171765,
+      "grad_norm": 0.09313994646072388,
+      "learning_rate": 8.797466770131919e-06,
+      "loss": 8.6352,
+      "step": 141670
+    },
+    {
+      "epoch": 0.7075332717421159,
+      "grad_norm": 0.09282007813453674,
+      "learning_rate": 8.795964855190368e-06,
+      "loss": 8.6173,
+      "step": 141680
+    },
+    {
+      "epoch": 0.7075832105670553,
+      "grad_norm": 0.08810416609048843,
+      "learning_rate": 8.794462940248818e-06,
+      "loss": 8.6341,
+      "step": 141690
+    },
+    {
+      "epoch": 0.7076331493919948,
+      "grad_norm": 0.09486988186836243,
+      "learning_rate": 8.792961025307266e-06,
+      "loss": 8.6187,
+      "step": 141700
+    },
+    {
+      "epoch": 0.7076830882169343,
+      "grad_norm": 0.08923081308603287,
+      "learning_rate": 8.791459110365716e-06,
+      "loss": 8.6303,
+      "step": 141710
+    },
+    {
+      "epoch": 0.7077330270418737,
+      "grad_norm": 0.09007243067026138,
+      "learning_rate": 8.789957195424167e-06,
+      "loss": 8.6309,
+      "step": 141720
+    },
+    {
+      "epoch": 0.7077829658668131,
+      "grad_norm": 0.09199108928442001,
+      "learning_rate": 8.788455280482615e-06,
+      "loss": 8.6047,
+      "step": 141730
+    },
+    {
+      "epoch": 0.7078329046917526,
+      "grad_norm": 0.09315073490142822,
+      "learning_rate": 8.786953365541065e-06,
+      "loss": 8.639,
+      "step": 141740
+    },
+    {
+      "epoch": 0.7078828435166921,
+      "grad_norm": 0.09393049031496048,
+      "learning_rate": 8.785451450599515e-06,
+      "loss": 8.637,
+      "step": 141750
+    },
+    {
+      "epoch": 0.7079327823416315,
+      "grad_norm": 0.08970038592815399,
+      "learning_rate": 8.783949535657964e-06,
+      "loss": 8.6206,
+      "step": 141760
+    },
+    {
+      "epoch": 0.7079827211665709,
+      "grad_norm": 0.09079226106405258,
+      "learning_rate": 8.782447620716414e-06,
+      "loss": 8.6221,
+      "step": 141770
+    },
+    {
+      "epoch": 0.7080326599915104,
+      "grad_norm": 0.09092725813388824,
+      "learning_rate": 8.780945705774863e-06,
+      "loss": 8.6283,
+      "step": 141780
+    },
+    {
+      "epoch": 0.7080825988164499,
+      "grad_norm": 0.08902301639318466,
+      "learning_rate": 8.779443790833313e-06,
+      "loss": 8.6266,
+      "step": 141790
+    },
+    {
+      "epoch": 0.7081325376413893,
+      "grad_norm": 0.09129643440246582,
+      "learning_rate": 8.777941875891763e-06,
+      "loss": 8.6308,
+      "step": 141800
+    },
+    {
+      "epoch": 0.7081824764663287,
+      "grad_norm": 0.09005031734704971,
+      "learning_rate": 8.776439960950211e-06,
+      "loss": 8.6405,
+      "step": 141810
+    },
+    {
+      "epoch": 0.7082324152912682,
+      "grad_norm": 0.0907624140381813,
+      "learning_rate": 8.774938046008662e-06,
+      "loss": 8.6345,
+      "step": 141820
+    },
+    {
+      "epoch": 0.7082823541162077,
+      "grad_norm": 0.09212564677000046,
+      "learning_rate": 8.77343613106711e-06,
+      "loss": 8.6532,
+      "step": 141830
+    },
+    {
+      "epoch": 0.7083322929411471,
+      "grad_norm": 0.09632585197687149,
+      "learning_rate": 8.77193421612556e-06,
+      "loss": 8.611,
+      "step": 141840
+    },
+    {
+      "epoch": 0.7083822317660865,
+      "grad_norm": 0.09049876779317856,
+      "learning_rate": 8.77043230118401e-06,
+      "loss": 8.6288,
+      "step": 141850
+    },
+    {
+      "epoch": 0.708432170591026,
+      "grad_norm": 0.08971288055181503,
+      "learning_rate": 8.768930386242459e-06,
+      "loss": 8.6298,
+      "step": 141860
+    },
+    {
+      "epoch": 0.7084821094159655,
+      "grad_norm": 0.09906616806983948,
+      "learning_rate": 8.767428471300909e-06,
+      "loss": 8.6211,
+      "step": 141870
+    },
+    {
+      "epoch": 0.7085320482409049,
+      "grad_norm": 0.09421475976705551,
+      "learning_rate": 8.765926556359358e-06,
+      "loss": 8.6384,
+      "step": 141880
+    },
+    {
+      "epoch": 0.7085819870658443,
+      "grad_norm": 0.08919620513916016,
+      "learning_rate": 8.764424641417808e-06,
+      "loss": 8.6428,
+      "step": 141890
+    },
+    {
+      "epoch": 0.7086319258907838,
+      "grad_norm": 0.0901108905673027,
+      "learning_rate": 8.762922726476258e-06,
+      "loss": 8.6232,
+      "step": 141900
+    },
+    {
+      "epoch": 0.7086818647157233,
+      "grad_norm": 0.0895710289478302,
+      "learning_rate": 8.761420811534708e-06,
+      "loss": 8.6461,
+      "step": 141910
+    },
+    {
+      "epoch": 0.7087318035406627,
+      "grad_norm": 0.09124090522527695,
+      "learning_rate": 8.759918896593157e-06,
+      "loss": 8.6357,
+      "step": 141920
+    },
+    {
+      "epoch": 0.7087817423656021,
+      "grad_norm": 0.09022539854049683,
+      "learning_rate": 8.758416981651605e-06,
+      "loss": 8.622,
+      "step": 141930
+    },
+    {
+      "epoch": 0.7088316811905416,
+      "grad_norm": 0.08793357014656067,
+      "learning_rate": 8.756915066710055e-06,
+      "loss": 8.6265,
+      "step": 141940
+    },
+    {
+      "epoch": 0.708881620015481,
+      "grad_norm": 0.09152141213417053,
+      "learning_rate": 8.755413151768505e-06,
+      "loss": 8.6229,
+      "step": 141950
+    },
+    {
+      "epoch": 0.7089315588404205,
+      "grad_norm": 0.09446197748184204,
+      "learning_rate": 8.753911236826956e-06,
+      "loss": 8.6395,
+      "step": 141960
+    },
+    {
+      "epoch": 0.7089814976653599,
+      "grad_norm": 0.08696843683719635,
+      "learning_rate": 8.752409321885404e-06,
+      "loss": 8.6207,
+      "step": 141970
+    },
+    {
+      "epoch": 0.7090314364902994,
+      "grad_norm": 0.09359335899353027,
+      "learning_rate": 8.750907406943853e-06,
+      "loss": 8.6385,
+      "step": 141980
+    },
+    {
+      "epoch": 0.7090813753152388,
+      "grad_norm": 0.09001364558935165,
+      "learning_rate": 8.749405492002303e-06,
+      "loss": 8.6352,
+      "step": 141990
+    },
+    {
+      "epoch": 0.7091313141401783,
+      "grad_norm": 0.09298966825008392,
+      "learning_rate": 8.747903577060753e-06,
+      "loss": 8.6246,
+      "step": 142000
+    },
+    {
+      "epoch": 0.7091812529651177,
+      "grad_norm": 0.09238903969526291,
+      "learning_rate": 8.746401662119203e-06,
+      "loss": 8.6325,
+      "step": 142010
+    },
+    {
+      "epoch": 0.7092311917900572,
+      "grad_norm": 0.09266587346792221,
+      "learning_rate": 8.744899747177652e-06,
+      "loss": 8.6339,
+      "step": 142020
+    },
+    {
+      "epoch": 0.7092811306149966,
+      "grad_norm": 0.090972401201725,
+      "learning_rate": 8.7433978322361e-06,
+      "loss": 8.634,
+      "step": 142030
+    },
+    {
+      "epoch": 0.7093310694399361,
+      "grad_norm": 0.09120898693799973,
+      "learning_rate": 8.74189591729455e-06,
+      "loss": 8.6247,
+      "step": 142040
+    },
+    {
+      "epoch": 0.7093810082648755,
+      "grad_norm": 0.09463303536176682,
+      "learning_rate": 8.740394002353e-06,
+      "loss": 8.6374,
+      "step": 142050
+    },
+    {
+      "epoch": 0.709430947089815,
+      "grad_norm": 0.09146030992269516,
+      "learning_rate": 8.73889208741145e-06,
+      "loss": 8.6386,
+      "step": 142060
+    },
+    {
+      "epoch": 0.7094808859147544,
+      "grad_norm": 0.09224395453929901,
+      "learning_rate": 8.7373901724699e-06,
+      "loss": 8.6257,
+      "step": 142070
+    },
+    {
+      "epoch": 0.7095308247396939,
+      "grad_norm": 0.08982054889202118,
+      "learning_rate": 8.735888257528348e-06,
+      "loss": 8.6414,
+      "step": 142080
+    },
+    {
+      "epoch": 0.7095807635646333,
+      "grad_norm": 0.09617902338504791,
+      "learning_rate": 8.734386342586798e-06,
+      "loss": 8.6255,
+      "step": 142090
+    },
+    {
+      "epoch": 0.7096307023895728,
+      "grad_norm": 0.09528889507055283,
+      "learning_rate": 8.732884427645248e-06,
+      "loss": 8.622,
+      "step": 142100
+    },
+    {
+      "epoch": 0.7096806412145122,
+      "grad_norm": 0.0894269272685051,
+      "learning_rate": 8.731382512703698e-06,
+      "loss": 8.6288,
+      "step": 142110
+    },
+    {
+      "epoch": 0.7097305800394517,
+      "grad_norm": 0.09145845472812653,
+      "learning_rate": 8.729880597762148e-06,
+      "loss": 8.6278,
+      "step": 142120
+    },
+    {
+      "epoch": 0.7097805188643911,
+      "grad_norm": 0.09675304591655731,
+      "learning_rate": 8.728378682820595e-06,
+      "loss": 8.6259,
+      "step": 142130
+    },
+    {
+      "epoch": 0.7098304576893306,
+      "grad_norm": 0.09058880060911179,
+      "learning_rate": 8.726876767879045e-06,
+      "loss": 8.6513,
+      "step": 142140
+    },
+    {
+      "epoch": 0.70988039651427,
+      "grad_norm": 0.0966552197933197,
+      "learning_rate": 8.725374852937495e-06,
+      "loss": 8.61,
+      "step": 142150
+    },
+    {
+      "epoch": 0.7099303353392095,
+      "grad_norm": 0.08744119107723236,
+      "learning_rate": 8.723872937995946e-06,
+      "loss": 8.6147,
+      "step": 142160
+    },
+    {
+      "epoch": 0.7099802741641489,
+      "grad_norm": 0.09429807960987091,
+      "learning_rate": 8.722371023054396e-06,
+      "loss": 8.6291,
+      "step": 142170
+    },
+    {
+      "epoch": 0.7100302129890884,
+      "grad_norm": 0.08756644278764725,
+      "learning_rate": 8.720869108112843e-06,
+      "loss": 8.6386,
+      "step": 142180
+    },
+    {
+      "epoch": 0.7100801518140278,
+      "grad_norm": 0.08758307993412018,
+      "learning_rate": 8.719367193171293e-06,
+      "loss": 8.6153,
+      "step": 142190
+    },
+    {
+      "epoch": 0.7101300906389673,
+      "grad_norm": 0.09212186932563782,
+      "learning_rate": 8.717865278229743e-06,
+      "loss": 8.6224,
+      "step": 142200
+    },
+    {
+      "epoch": 0.7101800294639067,
+      "grad_norm": 0.09597406536340714,
+      "learning_rate": 8.716363363288193e-06,
+      "loss": 8.6243,
+      "step": 142210
+    },
+    {
+      "epoch": 0.7102299682888462,
+      "grad_norm": 0.09284794330596924,
+      "learning_rate": 8.714861448346643e-06,
+      "loss": 8.6202,
+      "step": 142220
+    },
+    {
+      "epoch": 0.7102799071137856,
+      "grad_norm": 0.09078753739595413,
+      "learning_rate": 8.713359533405092e-06,
+      "loss": 8.625,
+      "step": 142230
+    },
+    {
+      "epoch": 0.7103298459387251,
+      "grad_norm": 0.09475717693567276,
+      "learning_rate": 8.71185761846354e-06,
+      "loss": 8.6252,
+      "step": 142240
+    },
+    {
+      "epoch": 0.7103797847636645,
+      "grad_norm": 0.09465180337429047,
+      "learning_rate": 8.71035570352199e-06,
+      "loss": 8.6361,
+      "step": 142250
+    },
+    {
+      "epoch": 0.710429723588604,
+      "grad_norm": 0.09588777273893356,
+      "learning_rate": 8.70885378858044e-06,
+      "loss": 8.6176,
+      "step": 142260
+    },
+    {
+      "epoch": 0.7104796624135434,
+      "grad_norm": 0.09004874527454376,
+      "learning_rate": 8.70735187363889e-06,
+      "loss": 8.6325,
+      "step": 142270
+    },
+    {
+      "epoch": 0.7105296012384829,
+      "grad_norm": 0.09361999481916428,
+      "learning_rate": 8.70584995869734e-06,
+      "loss": 8.6366,
+      "step": 142280
+    },
+    {
+      "epoch": 0.7105795400634223,
+      "grad_norm": 0.09634486585855484,
+      "learning_rate": 8.704348043755788e-06,
+      "loss": 8.6259,
+      "step": 142290
+    },
+    {
+      "epoch": 0.7106294788883618,
+      "grad_norm": 0.0938863530755043,
+      "learning_rate": 8.702846128814238e-06,
+      "loss": 8.6259,
+      "step": 142300
+    },
+    {
+      "epoch": 0.7106794177133012,
+      "grad_norm": 0.09446729719638824,
+      "learning_rate": 8.701344213872688e-06,
+      "loss": 8.6237,
+      "step": 142310
+    },
+    {
+      "epoch": 0.7107293565382407,
+      "grad_norm": 0.09354673326015472,
+      "learning_rate": 8.699842298931138e-06,
+      "loss": 8.6212,
+      "step": 142320
+    },
+    {
+      "epoch": 0.7107792953631801,
+      "grad_norm": 0.08772012591362,
+      "learning_rate": 8.698340383989587e-06,
+      "loss": 8.6214,
+      "step": 142330
+    },
+    {
+      "epoch": 0.7108292341881196,
+      "grad_norm": 0.0878673866391182,
+      "learning_rate": 8.696838469048035e-06,
+      "loss": 8.6261,
+      "step": 142340
+    },
+    {
+      "epoch": 0.710879173013059,
+      "grad_norm": 0.08873125165700912,
+      "learning_rate": 8.695336554106485e-06,
+      "loss": 8.6183,
+      "step": 142350
+    },
+    {
+      "epoch": 0.7109291118379985,
+      "grad_norm": 0.0938340574502945,
+      "learning_rate": 8.693834639164936e-06,
+      "loss": 8.6124,
+      "step": 142360
+    },
+    {
+      "epoch": 0.7109790506629379,
+      "grad_norm": 0.09345322847366333,
+      "learning_rate": 8.692332724223386e-06,
+      "loss": 8.6146,
+      "step": 142370
+    },
+    {
+      "epoch": 0.7110289894878774,
+      "grad_norm": 0.09127715975046158,
+      "learning_rate": 8.690830809281834e-06,
+      "loss": 8.6477,
+      "step": 142380
+    },
+    {
+      "epoch": 0.7110789283128168,
+      "grad_norm": 0.09575607627630234,
+      "learning_rate": 8.689328894340284e-06,
+      "loss": 8.6153,
+      "step": 142390
+    },
+    {
+      "epoch": 0.7111288671377562,
+      "grad_norm": 0.08808735758066177,
+      "learning_rate": 8.687826979398733e-06,
+      "loss": 8.6387,
+      "step": 142400
+    },
+    {
+      "epoch": 0.7111788059626957,
+      "grad_norm": 0.09160853177309036,
+      "learning_rate": 8.686325064457183e-06,
+      "loss": 8.6374,
+      "step": 142410
+    },
+    {
+      "epoch": 0.7112287447876351,
+      "grad_norm": 0.09475651383399963,
+      "learning_rate": 8.684823149515633e-06,
+      "loss": 8.6342,
+      "step": 142420
+    },
+    {
+      "epoch": 0.7112786836125746,
+      "grad_norm": 0.08886786550283432,
+      "learning_rate": 8.683321234574082e-06,
+      "loss": 8.6129,
+      "step": 142430
+    },
+    {
+      "epoch": 0.711328622437514,
+      "grad_norm": 0.0918954536318779,
+      "learning_rate": 8.681819319632532e-06,
+      "loss": 8.6083,
+      "step": 142440
+    },
+    {
+      "epoch": 0.7113785612624535,
+      "grad_norm": 0.09870171546936035,
+      "learning_rate": 8.68031740469098e-06,
+      "loss": 8.6117,
+      "step": 142450
+    },
+    {
+      "epoch": 0.7114285000873929,
+      "grad_norm": 0.09634058177471161,
+      "learning_rate": 8.67881548974943e-06,
+      "loss": 8.6249,
+      "step": 142460
+    },
+    {
+      "epoch": 0.7114784389123324,
+      "grad_norm": 0.09128417074680328,
+      "learning_rate": 8.67731357480788e-06,
+      "loss": 8.6259,
+      "step": 142470
+    },
+    {
+      "epoch": 0.7115283777372718,
+      "grad_norm": 0.08941973000764847,
+      "learning_rate": 8.67581165986633e-06,
+      "loss": 8.6221,
+      "step": 142480
+    },
+    {
+      "epoch": 0.7115783165622113,
+      "grad_norm": 0.10016234964132309,
+      "learning_rate": 8.67430974492478e-06,
+      "loss": 8.633,
+      "step": 142490
+    },
+    {
+      "epoch": 0.7116282553871507,
+      "grad_norm": 0.08851133286952972,
+      "learning_rate": 8.672807829983228e-06,
+      "loss": 8.6333,
+      "step": 142500
+    },
+    {
+      "epoch": 0.7116781942120902,
+      "grad_norm": 0.08761074393987656,
+      "learning_rate": 8.671305915041678e-06,
+      "loss": 8.6358,
+      "step": 142510
+    },
+    {
+      "epoch": 0.7117281330370296,
+      "grad_norm": 0.0912003293633461,
+      "learning_rate": 8.669804000100128e-06,
+      "loss": 8.6103,
+      "step": 142520
+    },
+    {
+      "epoch": 0.7117780718619691,
+      "grad_norm": 0.09523065388202667,
+      "learning_rate": 8.668302085158577e-06,
+      "loss": 8.6148,
+      "step": 142530
+    },
+    {
+      "epoch": 0.7118280106869085,
+      "grad_norm": 0.09399678558111191,
+      "learning_rate": 8.666800170217027e-06,
+      "loss": 8.622,
+      "step": 142540
+    },
+    {
+      "epoch": 0.711877949511848,
+      "grad_norm": 0.09112656861543655,
+      "learning_rate": 8.665298255275477e-06,
+      "loss": 8.6207,
+      "step": 142550
+    },
+    {
+      "epoch": 0.7119278883367874,
+      "grad_norm": 0.08925676345825195,
+      "learning_rate": 8.663796340333926e-06,
+      "loss": 8.6353,
+      "step": 142560
+    },
+    {
+      "epoch": 0.7119778271617269,
+      "grad_norm": 0.09203901141881943,
+      "learning_rate": 8.662294425392376e-06,
+      "loss": 8.6239,
+      "step": 142570
+    },
+    {
+      "epoch": 0.7120277659866663,
+      "grad_norm": 0.08438628166913986,
+      "learning_rate": 8.660792510450824e-06,
+      "loss": 8.6296,
+      "step": 142580
+    },
+    {
+      "epoch": 0.7120777048116058,
+      "grad_norm": 0.08993806689977646,
+      "learning_rate": 8.659290595509274e-06,
+      "loss": 8.6178,
+      "step": 142590
+    },
+    {
+      "epoch": 0.7121276436365452,
+      "grad_norm": 0.08885114639997482,
+      "learning_rate": 8.657788680567725e-06,
+      "loss": 8.6262,
+      "step": 142600
+    },
+    {
+      "epoch": 0.7121775824614847,
+      "grad_norm": 0.09332863986492157,
+      "learning_rate": 8.656286765626173e-06,
+      "loss": 8.6335,
+      "step": 142610
+    },
+    {
+      "epoch": 0.7122275212864241,
+      "grad_norm": 0.09431906044483185,
+      "learning_rate": 8.654784850684623e-06,
+      "loss": 8.6263,
+      "step": 142620
+    },
+    {
+      "epoch": 0.7122774601113636,
+      "grad_norm": 0.09060939401388168,
+      "learning_rate": 8.653282935743072e-06,
+      "loss": 8.621,
+      "step": 142630
+    },
+    {
+      "epoch": 0.712327398936303,
+      "grad_norm": 0.09892533719539642,
+      "learning_rate": 8.651781020801522e-06,
+      "loss": 8.6007,
+      "step": 142640
+    },
+    {
+      "epoch": 0.7123773377612425,
+      "grad_norm": 0.09287749230861664,
+      "learning_rate": 8.650279105859972e-06,
+      "loss": 8.6396,
+      "step": 142650
+    },
+    {
+      "epoch": 0.7124272765861819,
+      "grad_norm": 0.09754089266061783,
+      "learning_rate": 8.64877719091842e-06,
+      "loss": 8.6199,
+      "step": 142660
+    },
+    {
+      "epoch": 0.7124772154111214,
+      "grad_norm": 0.09147115051746368,
+      "learning_rate": 8.64727527597687e-06,
+      "loss": 8.6157,
+      "step": 142670
+    },
+    {
+      "epoch": 0.7125271542360608,
+      "grad_norm": 0.09473909437656403,
+      "learning_rate": 8.64577336103532e-06,
+      "loss": 8.6091,
+      "step": 142680
+    },
+    {
+      "epoch": 0.7125770930610003,
+      "grad_norm": 0.08881004899740219,
+      "learning_rate": 8.64427144609377e-06,
+      "loss": 8.6372,
+      "step": 142690
+    },
+    {
+      "epoch": 0.7126270318859397,
+      "grad_norm": 0.08714466542005539,
+      "learning_rate": 8.64276953115222e-06,
+      "loss": 8.6302,
+      "step": 142700
+    },
+    {
+      "epoch": 0.7126769707108792,
+      "grad_norm": 0.08891968429088593,
+      "learning_rate": 8.64126761621067e-06,
+      "loss": 8.6191,
+      "step": 142710
+    },
+    {
+      "epoch": 0.7127269095358186,
+      "grad_norm": 0.09196534752845764,
+      "learning_rate": 8.639765701269118e-06,
+      "loss": 8.62,
+      "step": 142720
+    },
+    {
+      "epoch": 0.712776848360758,
+      "grad_norm": 0.08720948547124863,
+      "learning_rate": 8.638263786327568e-06,
+      "loss": 8.6318,
+      "step": 142730
+    },
+    {
+      "epoch": 0.7128267871856975,
+      "grad_norm": 0.0971934124827385,
+      "learning_rate": 8.636761871386017e-06,
+      "loss": 8.6121,
+      "step": 142740
+    },
+    {
+      "epoch": 0.712876726010637,
+      "grad_norm": 0.09166713804006577,
+      "learning_rate": 8.635259956444467e-06,
+      "loss": 8.6297,
+      "step": 142750
+    },
+    {
+      "epoch": 0.7129266648355764,
+      "grad_norm": 0.09517423063516617,
+      "learning_rate": 8.633758041502917e-06,
+      "loss": 8.6293,
+      "step": 142760
+    },
+    {
+      "epoch": 0.7129766036605159,
+      "grad_norm": 0.09297715872526169,
+      "learning_rate": 8.632256126561366e-06,
+      "loss": 8.6296,
+      "step": 142770
+    },
+    {
+      "epoch": 0.7130265424854553,
+      "grad_norm": 0.09390052407979965,
+      "learning_rate": 8.630754211619816e-06,
+      "loss": 8.6186,
+      "step": 142780
+    },
+    {
+      "epoch": 0.7130764813103948,
+      "grad_norm": 0.09712127596139908,
+      "learning_rate": 8.629252296678264e-06,
+      "loss": 8.6155,
+      "step": 142790
+    },
+    {
+      "epoch": 0.7131264201353342,
+      "grad_norm": 0.0925319716334343,
+      "learning_rate": 8.627750381736715e-06,
+      "loss": 8.6215,
+      "step": 142800
+    },
+    {
+      "epoch": 0.7131763589602736,
+      "grad_norm": 0.09186070412397385,
+      "learning_rate": 8.626248466795165e-06,
+      "loss": 8.6158,
+      "step": 142810
+    },
+    {
+      "epoch": 0.7132262977852131,
+      "grad_norm": 0.09534446150064468,
+      "learning_rate": 8.624746551853613e-06,
+      "loss": 8.624,
+      "step": 142820
+    },
+    {
+      "epoch": 0.7132762366101526,
+      "grad_norm": 0.09688442945480347,
+      "learning_rate": 8.623244636912063e-06,
+      "loss": 8.5915,
+      "step": 142830
+    },
+    {
+      "epoch": 0.713326175435092,
+      "grad_norm": 0.0908428207039833,
+      "learning_rate": 8.621742721970512e-06,
+      "loss": 8.6246,
+      "step": 142840
+    },
+    {
+      "epoch": 0.7133761142600314,
+      "grad_norm": 0.08914895355701447,
+      "learning_rate": 8.620240807028962e-06,
+      "loss": 8.6274,
+      "step": 142850
+    },
+    {
+      "epoch": 0.7134260530849709,
+      "grad_norm": 0.0967710018157959,
+      "learning_rate": 8.618738892087412e-06,
+      "loss": 8.6185,
+      "step": 142860
+    },
+    {
+      "epoch": 0.7134759919099104,
+      "grad_norm": 0.09269560873508453,
+      "learning_rate": 8.617236977145863e-06,
+      "loss": 8.6148,
+      "step": 142870
+    },
+    {
+      "epoch": 0.7135259307348498,
+      "grad_norm": 0.09392331540584564,
+      "learning_rate": 8.615735062204311e-06,
+      "loss": 8.6251,
+      "step": 142880
+    },
+    {
+      "epoch": 0.7135758695597892,
+      "grad_norm": 0.08788944780826569,
+      "learning_rate": 8.61423314726276e-06,
+      "loss": 8.6238,
+      "step": 142890
+    },
+    {
+      "epoch": 0.7136258083847287,
+      "grad_norm": 0.09034030884504318,
+      "learning_rate": 8.61273123232121e-06,
+      "loss": 8.622,
+      "step": 142900
+    },
+    {
+      "epoch": 0.7136757472096682,
+      "grad_norm": 0.08973328024148941,
+      "learning_rate": 8.61122931737966e-06,
+      "loss": 8.6273,
+      "step": 142910
+    },
+    {
+      "epoch": 0.7137256860346076,
+      "grad_norm": 0.08710794895887375,
+      "learning_rate": 8.60972740243811e-06,
+      "loss": 8.6138,
+      "step": 142920
+    },
+    {
+      "epoch": 0.713775624859547,
+      "grad_norm": 0.09544447809457779,
+      "learning_rate": 8.608225487496558e-06,
+      "loss": 8.6187,
+      "step": 142930
+    },
+    {
+      "epoch": 0.7138255636844865,
+      "grad_norm": 0.09114502370357513,
+      "learning_rate": 8.606723572555007e-06,
+      "loss": 8.6234,
+      "step": 142940
+    },
+    {
+      "epoch": 0.713875502509426,
+      "grad_norm": 0.09414034336805344,
+      "learning_rate": 8.605221657613457e-06,
+      "loss": 8.6191,
+      "step": 142950
+    },
+    {
+      "epoch": 0.7139254413343654,
+      "grad_norm": 0.08966190367937088,
+      "learning_rate": 8.603719742671907e-06,
+      "loss": 8.6134,
+      "step": 142960
+    },
+    {
+      "epoch": 0.7139753801593048,
+      "grad_norm": 0.08923894166946411,
+      "learning_rate": 8.602217827730358e-06,
+      "loss": 8.6135,
+      "step": 142970
+    },
+    {
+      "epoch": 0.7140253189842443,
+      "grad_norm": 0.09734483063220978,
+      "learning_rate": 8.600715912788806e-06,
+      "loss": 8.6281,
+      "step": 142980
+    },
+    {
+      "epoch": 0.7140752578091838,
+      "grad_norm": 0.09202791005373001,
+      "learning_rate": 8.599213997847254e-06,
+      "loss": 8.6208,
+      "step": 142990
+    },
+    {
+      "epoch": 0.7141251966341232,
+      "grad_norm": 0.09096809476613998,
+      "learning_rate": 8.597712082905705e-06,
+      "loss": 8.6286,
+      "step": 143000
+    },
+    {
+      "epoch": 0.7141751354590626,
+      "grad_norm": 0.09096328914165497,
+      "learning_rate": 8.596210167964155e-06,
+      "loss": 8.6242,
+      "step": 143010
+    },
+    {
+      "epoch": 0.7142250742840021,
+      "grad_norm": 0.09118296205997467,
+      "learning_rate": 8.594708253022605e-06,
+      "loss": 8.6227,
+      "step": 143020
+    },
+    {
+      "epoch": 0.7142750131089416,
+      "grad_norm": 0.09323076903820038,
+      "learning_rate": 8.593206338081055e-06,
+      "loss": 8.6223,
+      "step": 143030
+    },
+    {
+      "epoch": 0.714324951933881,
+      "grad_norm": 0.08754134923219681,
+      "learning_rate": 8.591704423139502e-06,
+      "loss": 8.6249,
+      "step": 143040
+    },
+    {
+      "epoch": 0.7143748907588204,
+      "grad_norm": 0.09009493887424469,
+      "learning_rate": 8.590202508197952e-06,
+      "loss": 8.6254,
+      "step": 143050
+    },
+    {
+      "epoch": 0.7144248295837599,
+      "grad_norm": 0.09118343889713287,
+      "learning_rate": 8.588700593256402e-06,
+      "loss": 8.6078,
+      "step": 143060
+    },
+    {
+      "epoch": 0.7144747684086994,
+      "grad_norm": 0.0941152274608612,
+      "learning_rate": 8.587198678314853e-06,
+      "loss": 8.6298,
+      "step": 143070
+    },
+    {
+      "epoch": 0.7145247072336388,
+      "grad_norm": 0.08993294835090637,
+      "learning_rate": 8.585696763373303e-06,
+      "loss": 8.6334,
+      "step": 143080
+    },
+    {
+      "epoch": 0.7145746460585782,
+      "grad_norm": 0.0892404243350029,
+      "learning_rate": 8.58419484843175e-06,
+      "loss": 8.5978,
+      "step": 143090
+    },
+    {
+      "epoch": 0.7146245848835177,
+      "grad_norm": 0.08918585628271103,
+      "learning_rate": 8.5826929334902e-06,
+      "loss": 8.6161,
+      "step": 143100
+    },
+    {
+      "epoch": 0.7146745237084572,
+      "grad_norm": 0.08921711891889572,
+      "learning_rate": 8.58119101854865e-06,
+      "loss": 8.6231,
+      "step": 143110
+    },
+    {
+      "epoch": 0.7147244625333966,
+      "grad_norm": 0.10099107027053833,
+      "learning_rate": 8.5796891036071e-06,
+      "loss": 8.6362,
+      "step": 143120
+    },
+    {
+      "epoch": 0.714774401358336,
+      "grad_norm": 0.09065527468919754,
+      "learning_rate": 8.57818718866555e-06,
+      "loss": 8.6122,
+      "step": 143130
+    },
+    {
+      "epoch": 0.7148243401832755,
+      "grad_norm": 0.09168516844511032,
+      "learning_rate": 8.576685273723997e-06,
+      "loss": 8.6107,
+      "step": 143140
+    },
+    {
+      "epoch": 0.714874279008215,
+      "grad_norm": 0.0854279100894928,
+      "learning_rate": 8.575183358782447e-06,
+      "loss": 8.6273,
+      "step": 143150
+    },
+    {
+      "epoch": 0.7149242178331544,
+      "grad_norm": 0.08854588121175766,
+      "learning_rate": 8.573681443840897e-06,
+      "loss": 8.6061,
+      "step": 143160
+    },
+    {
+      "epoch": 0.7149741566580938,
+      "grad_norm": 0.0906798392534256,
+      "learning_rate": 8.572179528899348e-06,
+      "loss": 8.6264,
+      "step": 143170
+    },
+    {
+      "epoch": 0.7150240954830333,
+      "grad_norm": 0.09377674758434296,
+      "learning_rate": 8.570677613957798e-06,
+      "loss": 8.6236,
+      "step": 143180
+    },
+    {
+      "epoch": 0.7150740343079728,
+      "grad_norm": 0.09252772480249405,
+      "learning_rate": 8.569175699016246e-06,
+      "loss": 8.6202,
+      "step": 143190
+    },
+    {
+      "epoch": 0.7151239731329122,
+      "grad_norm": 0.09276933968067169,
+      "learning_rate": 8.567673784074695e-06,
+      "loss": 8.6067,
+      "step": 143200
+    },
+    {
+      "epoch": 0.7151739119578516,
+      "grad_norm": 0.09133142232894897,
+      "learning_rate": 8.566171869133145e-06,
+      "loss": 8.6154,
+      "step": 143210
+    },
+    {
+      "epoch": 0.715223850782791,
+      "grad_norm": 0.09052088111639023,
+      "learning_rate": 8.564669954191595e-06,
+      "loss": 8.6426,
+      "step": 143220
+    },
+    {
+      "epoch": 0.7152737896077306,
+      "grad_norm": 0.0909975916147232,
+      "learning_rate": 8.563168039250045e-06,
+      "loss": 8.6208,
+      "step": 143230
+    },
+    {
+      "epoch": 0.71532372843267,
+      "grad_norm": 0.08892220258712769,
+      "learning_rate": 8.561666124308494e-06,
+      "loss": 8.61,
+      "step": 143240
+    },
+    {
+      "epoch": 0.7153736672576094,
+      "grad_norm": 0.09050557017326355,
+      "learning_rate": 8.560164209366942e-06,
+      "loss": 8.6121,
+      "step": 143250
+    },
+    {
+      "epoch": 0.7154236060825488,
+      "grad_norm": 0.09054877609014511,
+      "learning_rate": 8.558662294425392e-06,
+      "loss": 8.6173,
+      "step": 143260
+    },
+    {
+      "epoch": 0.7154735449074884,
+      "grad_norm": 0.09878473728895187,
+      "learning_rate": 8.557160379483843e-06,
+      "loss": 8.611,
+      "step": 143270
+    },
+    {
+      "epoch": 0.7155234837324278,
+      "grad_norm": 0.08595815300941467,
+      "learning_rate": 8.555658464542293e-06,
+      "loss": 8.6321,
+      "step": 143280
+    },
+    {
+      "epoch": 0.7155734225573672,
+      "grad_norm": 0.0915842056274414,
+      "learning_rate": 8.554156549600741e-06,
+      "loss": 8.6161,
+      "step": 143290
+    },
+    {
+      "epoch": 0.7156233613823066,
+      "grad_norm": 0.08792208880186081,
+      "learning_rate": 8.55265463465919e-06,
+      "loss": 8.6099,
+      "step": 143300
+    },
+    {
+      "epoch": 0.7156733002072462,
+      "grad_norm": 0.0902579203248024,
+      "learning_rate": 8.55115271971764e-06,
+      "loss": 8.6081,
+      "step": 143310
+    },
+    {
+      "epoch": 0.7157232390321856,
+      "grad_norm": 0.09018343687057495,
+      "learning_rate": 8.54965080477609e-06,
+      "loss": 8.6206,
+      "step": 143320
+    },
+    {
+      "epoch": 0.715773177857125,
+      "grad_norm": 0.09098510444164276,
+      "learning_rate": 8.54814888983454e-06,
+      "loss": 8.6374,
+      "step": 143330
+    },
+    {
+      "epoch": 0.7158231166820644,
+      "grad_norm": 0.09270698577165604,
+      "learning_rate": 8.546646974892989e-06,
+      "loss": 8.6342,
+      "step": 143340
+    },
+    {
+      "epoch": 0.715873055507004,
+      "grad_norm": 0.09691673517227173,
+      "learning_rate": 8.545145059951439e-06,
+      "loss": 8.6208,
+      "step": 143350
+    },
+    {
+      "epoch": 0.7159229943319434,
+      "grad_norm": 0.09676571935415268,
+      "learning_rate": 8.543643145009887e-06,
+      "loss": 8.6159,
+      "step": 143360
+    },
+    {
+      "epoch": 0.7159729331568828,
+      "grad_norm": 0.08486426621675491,
+      "learning_rate": 8.542141230068338e-06,
+      "loss": 8.6335,
+      "step": 143370
+    },
+    {
+      "epoch": 0.7160228719818222,
+      "grad_norm": 0.09273499995470047,
+      "learning_rate": 8.540639315126788e-06,
+      "loss": 8.62,
+      "step": 143380
+    },
+    {
+      "epoch": 0.7160728108067617,
+      "grad_norm": 0.09392505139112473,
+      "learning_rate": 8.539137400185236e-06,
+      "loss": 8.6318,
+      "step": 143390
+    },
+    {
+      "epoch": 0.7161227496317012,
+      "grad_norm": 0.08955458551645279,
+      "learning_rate": 8.537635485243686e-06,
+      "loss": 8.6154,
+      "step": 143400
+    },
+    {
+      "epoch": 0.7161726884566406,
+      "grad_norm": 0.08472466468811035,
+      "learning_rate": 8.536133570302135e-06,
+      "loss": 8.6181,
+      "step": 143410
+    },
+    {
+      "epoch": 0.71622262728158,
+      "grad_norm": 0.08813430368900299,
+      "learning_rate": 8.534631655360585e-06,
+      "loss": 8.6169,
+      "step": 143420
+    },
+    {
+      "epoch": 0.7162725661065195,
+      "grad_norm": 0.09203272312879562,
+      "learning_rate": 8.533129740419035e-06,
+      "loss": 8.6113,
+      "step": 143430
+    },
+    {
+      "epoch": 0.716322504931459,
+      "grad_norm": 0.08561133593320847,
+      "learning_rate": 8.531627825477484e-06,
+      "loss": 8.6069,
+      "step": 143440
+    },
+    {
+      "epoch": 0.7163724437563984,
+      "grad_norm": 0.08707552403211594,
+      "learning_rate": 8.530125910535934e-06,
+      "loss": 8.6089,
+      "step": 143450
+    },
+    {
+      "epoch": 0.7164223825813378,
+      "grad_norm": 0.09338618814945221,
+      "learning_rate": 8.528623995594382e-06,
+      "loss": 8.6026,
+      "step": 143460
+    },
+    {
+      "epoch": 0.7164723214062773,
+      "grad_norm": 0.08720285445451736,
+      "learning_rate": 8.527122080652833e-06,
+      "loss": 8.6415,
+      "step": 143470
+    },
+    {
+      "epoch": 0.7165222602312168,
+      "grad_norm": 0.09213987737894058,
+      "learning_rate": 8.525620165711283e-06,
+      "loss": 8.6261,
+      "step": 143480
+    },
+    {
+      "epoch": 0.7165721990561562,
+      "grad_norm": 0.09045135974884033,
+      "learning_rate": 8.524118250769731e-06,
+      "loss": 8.6133,
+      "step": 143490
+    },
+    {
+      "epoch": 0.7166221378810956,
+      "grad_norm": 0.08416689932346344,
+      "learning_rate": 8.522616335828181e-06,
+      "loss": 8.6213,
+      "step": 143500
+    },
+    {
+      "epoch": 0.716672076706035,
+      "grad_norm": 0.09071792662143707,
+      "learning_rate": 8.52111442088663e-06,
+      "loss": 8.606,
+      "step": 143510
+    },
+    {
+      "epoch": 0.7167220155309746,
+      "grad_norm": 0.08766292035579681,
+      "learning_rate": 8.51961250594508e-06,
+      "loss": 8.6364,
+      "step": 143520
+    },
+    {
+      "epoch": 0.716771954355914,
+      "grad_norm": 0.09889094531536102,
+      "learning_rate": 8.51811059100353e-06,
+      "loss": 8.6084,
+      "step": 143530
+    },
+    {
+      "epoch": 0.7168218931808534,
+      "grad_norm": 0.09228768199682236,
+      "learning_rate": 8.516608676061979e-06,
+      "loss": 8.617,
+      "step": 143540
+    },
+    {
+      "epoch": 0.7168718320057929,
+      "grad_norm": 0.08915939927101135,
+      "learning_rate": 8.515106761120429e-06,
+      "loss": 8.6044,
+      "step": 143550
+    },
+    {
+      "epoch": 0.7169217708307324,
+      "grad_norm": 0.08671806752681732,
+      "learning_rate": 8.513604846178879e-06,
+      "loss": 8.6338,
+      "step": 143560
+    },
+    {
+      "epoch": 0.7169717096556718,
+      "grad_norm": 0.090422123670578,
+      "learning_rate": 8.512102931237328e-06,
+      "loss": 8.611,
+      "step": 143570
+    },
+    {
+      "epoch": 0.7170216484806112,
+      "grad_norm": 0.08737920969724655,
+      "learning_rate": 8.510601016295778e-06,
+      "loss": 8.6161,
+      "step": 143580
+    },
+    {
+      "epoch": 0.7170715873055507,
+      "grad_norm": 0.08856778591871262,
+      "learning_rate": 8.509099101354226e-06,
+      "loss": 8.63,
+      "step": 143590
+    },
+    {
+      "epoch": 0.7171215261304902,
+      "grad_norm": 0.0872621163725853,
+      "learning_rate": 8.507597186412676e-06,
+      "loss": 8.6203,
+      "step": 143600
+    },
+    {
+      "epoch": 0.7171714649554296,
+      "grad_norm": 0.08896245807409286,
+      "learning_rate": 8.506095271471127e-06,
+      "loss": 8.6228,
+      "step": 143610
+    },
+    {
+      "epoch": 0.717221403780369,
+      "grad_norm": 0.09879324585199356,
+      "learning_rate": 8.504593356529575e-06,
+      "loss": 8.6155,
+      "step": 143620
+    },
+    {
+      "epoch": 0.7172713426053084,
+      "grad_norm": 0.08767727017402649,
+      "learning_rate": 8.503091441588025e-06,
+      "loss": 8.6018,
+      "step": 143630
+    },
+    {
+      "epoch": 0.717321281430248,
+      "grad_norm": 0.0934598445892334,
+      "learning_rate": 8.501589526646474e-06,
+      "loss": 8.6157,
+      "step": 143640
+    },
+    {
+      "epoch": 0.7173712202551874,
+      "grad_norm": 0.09154853224754333,
+      "learning_rate": 8.500087611704924e-06,
+      "loss": 8.618,
+      "step": 143650
+    },
+    {
+      "epoch": 0.7174211590801268,
+      "grad_norm": 0.09346166253089905,
+      "learning_rate": 8.498585696763374e-06,
+      "loss": 8.6144,
+      "step": 143660
+    },
+    {
+      "epoch": 0.7174710979050662,
+      "grad_norm": 0.09311819821596146,
+      "learning_rate": 8.497083781821823e-06,
+      "loss": 8.6105,
+      "step": 143670
+    },
+    {
+      "epoch": 0.7175210367300058,
+      "grad_norm": 0.08571168035268784,
+      "learning_rate": 8.495581866880273e-06,
+      "loss": 8.6284,
+      "step": 143680
+    },
+    {
+      "epoch": 0.7175709755549452,
+      "grad_norm": 0.09284307062625885,
+      "learning_rate": 8.494079951938721e-06,
+      "loss": 8.6317,
+      "step": 143690
+    },
+    {
+      "epoch": 0.7176209143798846,
+      "grad_norm": 0.08840735256671906,
+      "learning_rate": 8.492578036997171e-06,
+      "loss": 8.6275,
+      "step": 143700
+    },
+    {
+      "epoch": 0.717670853204824,
+      "grad_norm": 0.09564638137817383,
+      "learning_rate": 8.491076122055622e-06,
+      "loss": 8.5992,
+      "step": 143710
+    },
+    {
+      "epoch": 0.7177207920297636,
+      "grad_norm": 0.09461285173892975,
+      "learning_rate": 8.489574207114072e-06,
+      "loss": 8.6209,
+      "step": 143720
+    },
+    {
+      "epoch": 0.717770730854703,
+      "grad_norm": 0.0862046629190445,
+      "learning_rate": 8.48807229217252e-06,
+      "loss": 8.6381,
+      "step": 143730
+    },
+    {
+      "epoch": 0.7178206696796424,
+      "grad_norm": 0.0956760123372078,
+      "learning_rate": 8.486570377230969e-06,
+      "loss": 8.6162,
+      "step": 143740
+    },
+    {
+      "epoch": 0.7178706085045818,
+      "grad_norm": 0.09319616854190826,
+      "learning_rate": 8.485068462289419e-06,
+      "loss": 8.617,
+      "step": 143750
+    },
+    {
+      "epoch": 0.7179205473295214,
+      "grad_norm": 0.08868087083101273,
+      "learning_rate": 8.483566547347869e-06,
+      "loss": 8.6159,
+      "step": 143760
+    },
+    {
+      "epoch": 0.7179704861544608,
+      "grad_norm": 0.0888964906334877,
+      "learning_rate": 8.48206463240632e-06,
+      "loss": 8.6165,
+      "step": 143770
+    },
+    {
+      "epoch": 0.7180204249794002,
+      "grad_norm": 0.0897190198302269,
+      "learning_rate": 8.480562717464768e-06,
+      "loss": 8.6291,
+      "step": 143780
+    },
+    {
+      "epoch": 0.7180703638043396,
+      "grad_norm": 0.0912693589925766,
+      "learning_rate": 8.479060802523216e-06,
+      "loss": 8.6018,
+      "step": 143790
+    },
+    {
+      "epoch": 0.7181203026292792,
+      "grad_norm": 0.09534019976854324,
+      "learning_rate": 8.477558887581666e-06,
+      "loss": 8.6268,
+      "step": 143800
+    },
+    {
+      "epoch": 0.7181702414542186,
+      "grad_norm": 0.0912948027253151,
+      "learning_rate": 8.476056972640117e-06,
+      "loss": 8.6165,
+      "step": 143810
+    },
+    {
+      "epoch": 0.718220180279158,
+      "grad_norm": 0.09484043717384338,
+      "learning_rate": 8.474555057698567e-06,
+      "loss": 8.6171,
+      "step": 143820
+    },
+    {
+      "epoch": 0.7182701191040974,
+      "grad_norm": 0.09343184530735016,
+      "learning_rate": 8.473053142757015e-06,
+      "loss": 8.6158,
+      "step": 143830
+    },
+    {
+      "epoch": 0.718320057929037,
+      "grad_norm": 0.09978407621383667,
+      "learning_rate": 8.471551227815464e-06,
+      "loss": 8.6112,
+      "step": 143840
+    },
+    {
+      "epoch": 0.7183699967539764,
+      "grad_norm": 0.09003512561321259,
+      "learning_rate": 8.470049312873914e-06,
+      "loss": 8.6293,
+      "step": 143850
+    },
+    {
+      "epoch": 0.7184199355789158,
+      "grad_norm": 0.0940660908818245,
+      "learning_rate": 8.468547397932364e-06,
+      "loss": 8.6131,
+      "step": 143860
+    },
+    {
+      "epoch": 0.7184698744038552,
+      "grad_norm": 0.09280209988355637,
+      "learning_rate": 8.467045482990814e-06,
+      "loss": 8.6309,
+      "step": 143870
+    },
+    {
+      "epoch": 0.7185198132287948,
+      "grad_norm": 0.0923769474029541,
+      "learning_rate": 8.465543568049264e-06,
+      "loss": 8.6371,
+      "step": 143880
+    },
+    {
+      "epoch": 0.7185697520537342,
+      "grad_norm": 0.09040191769599915,
+      "learning_rate": 8.464041653107711e-06,
+      "loss": 8.6111,
+      "step": 143890
+    },
+    {
+      "epoch": 0.7186196908786736,
+      "grad_norm": 0.09328994154930115,
+      "learning_rate": 8.462539738166161e-06,
+      "loss": 8.6138,
+      "step": 143900
+    },
+    {
+      "epoch": 0.718669629703613,
+      "grad_norm": 0.0912991613149643,
+      "learning_rate": 8.461037823224612e-06,
+      "loss": 8.601,
+      "step": 143910
+    },
+    {
+      "epoch": 0.7187195685285526,
+      "grad_norm": 0.09061001241207123,
+      "learning_rate": 8.459535908283062e-06,
+      "loss": 8.6132,
+      "step": 143920
+    },
+    {
+      "epoch": 0.718769507353492,
+      "grad_norm": 0.09003961086273193,
+      "learning_rate": 8.458033993341512e-06,
+      "loss": 8.6153,
+      "step": 143930
+    },
+    {
+      "epoch": 0.7188194461784314,
+      "grad_norm": 0.09332337230443954,
+      "learning_rate": 8.456532078399959e-06,
+      "loss": 8.6154,
+      "step": 143940
+    },
+    {
+      "epoch": 0.7188693850033708,
+      "grad_norm": 0.09105270355939865,
+      "learning_rate": 8.455030163458409e-06,
+      "loss": 8.6062,
+      "step": 143950
+    },
+    {
+      "epoch": 0.7189193238283104,
+      "grad_norm": 0.0938153937458992,
+      "learning_rate": 8.453528248516859e-06,
+      "loss": 8.6181,
+      "step": 143960
+    },
+    {
+      "epoch": 0.7189692626532498,
+      "grad_norm": 0.08828999102115631,
+      "learning_rate": 8.45202633357531e-06,
+      "loss": 8.6192,
+      "step": 143970
+    },
+    {
+      "epoch": 0.7190192014781892,
+      "grad_norm": 0.09276334941387177,
+      "learning_rate": 8.45052441863376e-06,
+      "loss": 8.609,
+      "step": 143980
+    },
+    {
+      "epoch": 0.7190691403031286,
+      "grad_norm": 0.09385494142770767,
+      "learning_rate": 8.449022503692206e-06,
+      "loss": 8.6149,
+      "step": 143990
+    },
+    {
+      "epoch": 0.7191190791280682,
+      "grad_norm": 0.09120655804872513,
+      "learning_rate": 8.447520588750656e-06,
+      "loss": 8.6296,
+      "step": 144000
+    },
+    {
+      "epoch": 0.7191690179530076,
+      "grad_norm": 0.08751309663057327,
+      "learning_rate": 8.446018673809107e-06,
+      "loss": 8.6187,
+      "step": 144010
+    },
+    {
+      "epoch": 0.719218956777947,
+      "grad_norm": 0.08984875679016113,
+      "learning_rate": 8.444516758867557e-06,
+      "loss": 8.627,
+      "step": 144020
+    },
+    {
+      "epoch": 0.7192688956028864,
+      "grad_norm": 0.08760925382375717,
+      "learning_rate": 8.443014843926007e-06,
+      "loss": 8.6132,
+      "step": 144030
+    },
+    {
+      "epoch": 0.719318834427826,
+      "grad_norm": 0.09052575379610062,
+      "learning_rate": 8.441512928984455e-06,
+      "loss": 8.6184,
+      "step": 144040
+    },
+    {
+      "epoch": 0.7193687732527654,
+      "grad_norm": 0.08988479524850845,
+      "learning_rate": 8.440011014042904e-06,
+      "loss": 8.6226,
+      "step": 144050
+    },
+    {
+      "epoch": 0.7194187120777048,
+      "grad_norm": 0.0874130055308342,
+      "learning_rate": 8.438509099101354e-06,
+      "loss": 8.6209,
+      "step": 144060
+    },
+    {
+      "epoch": 0.7194686509026442,
+      "grad_norm": 0.09357550740242004,
+      "learning_rate": 8.437007184159804e-06,
+      "loss": 8.5982,
+      "step": 144070
+    },
+    {
+      "epoch": 0.7195185897275838,
+      "grad_norm": 0.0861838087439537,
+      "learning_rate": 8.435505269218254e-06,
+      "loss": 8.6101,
+      "step": 144080
+    },
+    {
+      "epoch": 0.7195685285525232,
+      "grad_norm": 0.0887862890958786,
+      "learning_rate": 8.434003354276703e-06,
+      "loss": 8.6,
+      "step": 144090
+    },
+    {
+      "epoch": 0.7196184673774626,
+      "grad_norm": 0.09258834272623062,
+      "learning_rate": 8.432501439335151e-06,
+      "loss": 8.6318,
+      "step": 144100
+    },
+    {
+      "epoch": 0.719668406202402,
+      "grad_norm": 0.0883084088563919,
+      "learning_rate": 8.430999524393602e-06,
+      "loss": 8.6222,
+      "step": 144110
+    },
+    {
+      "epoch": 0.7197183450273416,
+      "grad_norm": 0.09339895844459534,
+      "learning_rate": 8.429497609452052e-06,
+      "loss": 8.6054,
+      "step": 144120
+    },
+    {
+      "epoch": 0.719768283852281,
+      "grad_norm": 0.0934942364692688,
+      "learning_rate": 8.427995694510502e-06,
+      "loss": 8.6172,
+      "step": 144130
+    },
+    {
+      "epoch": 0.7198182226772204,
+      "grad_norm": 0.09863001853227615,
+      "learning_rate": 8.42649377956895e-06,
+      "loss": 8.6149,
+      "step": 144140
+    },
+    {
+      "epoch": 0.7198681615021598,
+      "grad_norm": 0.08822454512119293,
+      "learning_rate": 8.424991864627399e-06,
+      "loss": 8.614,
+      "step": 144150
+    },
+    {
+      "epoch": 0.7199181003270994,
+      "grad_norm": 0.08899369835853577,
+      "learning_rate": 8.423489949685849e-06,
+      "loss": 8.6169,
+      "step": 144160
+    },
+    {
+      "epoch": 0.7199680391520388,
+      "grad_norm": 0.08983699232339859,
+      "learning_rate": 8.4219880347443e-06,
+      "loss": 8.606,
+      "step": 144170
+    },
+    {
+      "epoch": 0.7200179779769782,
+      "grad_norm": 0.092051200568676,
+      "learning_rate": 8.42048611980275e-06,
+      "loss": 8.6196,
+      "step": 144180
+    },
+    {
+      "epoch": 0.7200679168019176,
+      "grad_norm": 0.09895046800374985,
+      "learning_rate": 8.418984204861198e-06,
+      "loss": 8.6161,
+      "step": 144190
+    },
+    {
+      "epoch": 0.7201178556268572,
+      "grad_norm": 0.09208567440509796,
+      "learning_rate": 8.417482289919648e-06,
+      "loss": 8.6099,
+      "step": 144200
+    },
+    {
+      "epoch": 0.7201677944517966,
+      "grad_norm": 0.08995640277862549,
+      "learning_rate": 8.415980374978097e-06,
+      "loss": 8.626,
+      "step": 144210
+    },
+    {
+      "epoch": 0.720217733276736,
+      "grad_norm": 0.08942251652479172,
+      "learning_rate": 8.414478460036547e-06,
+      "loss": 8.614,
+      "step": 144220
+    },
+    {
+      "epoch": 0.7202676721016754,
+      "grad_norm": 0.08955737203359604,
+      "learning_rate": 8.412976545094997e-06,
+      "loss": 8.6169,
+      "step": 144230
+    },
+    {
+      "epoch": 0.720317610926615,
+      "grad_norm": 0.08591482788324356,
+      "learning_rate": 8.411474630153445e-06,
+      "loss": 8.5999,
+      "step": 144240
+    },
+    {
+      "epoch": 0.7203675497515544,
+      "grad_norm": 0.09663660079240799,
+      "learning_rate": 8.409972715211896e-06,
+      "loss": 8.6279,
+      "step": 144250
+    },
+    {
+      "epoch": 0.7204174885764938,
+      "grad_norm": 0.10267946124076843,
+      "learning_rate": 8.408470800270344e-06,
+      "loss": 8.5984,
+      "step": 144260
+    },
+    {
+      "epoch": 0.7204674274014332,
+      "grad_norm": 0.08785830438137054,
+      "learning_rate": 8.406968885328794e-06,
+      "loss": 8.6214,
+      "step": 144270
+    },
+    {
+      "epoch": 0.7205173662263727,
+      "grad_norm": 0.0935339480638504,
+      "learning_rate": 8.405466970387244e-06,
+      "loss": 8.6095,
+      "step": 144280
+    },
+    {
+      "epoch": 0.7205673050513122,
+      "grad_norm": 0.08838778734207153,
+      "learning_rate": 8.403965055445693e-06,
+      "loss": 8.6216,
+      "step": 144290
+    },
+    {
+      "epoch": 0.7206172438762516,
+      "grad_norm": 0.09756764769554138,
+      "learning_rate": 8.402463140504143e-06,
+      "loss": 8.61,
+      "step": 144300
+    },
+    {
+      "epoch": 0.720667182701191,
+      "grad_norm": 0.0923360213637352,
+      "learning_rate": 8.400961225562592e-06,
+      "loss": 8.5908,
+      "step": 144310
+    },
+    {
+      "epoch": 0.7207171215261305,
+      "grad_norm": 0.08914266526699066,
+      "learning_rate": 8.399459310621042e-06,
+      "loss": 8.5996,
+      "step": 144320
+    },
+    {
+      "epoch": 0.72076706035107,
+      "grad_norm": 0.09269039332866669,
+      "learning_rate": 8.397957395679492e-06,
+      "loss": 8.63,
+      "step": 144330
+    },
+    {
+      "epoch": 0.7208169991760094,
+      "grad_norm": 0.09222870320081711,
+      "learning_rate": 8.39645548073794e-06,
+      "loss": 8.6083,
+      "step": 144340
+    },
+    {
+      "epoch": 0.7208669380009488,
+      "grad_norm": 0.09075245261192322,
+      "learning_rate": 8.39495356579639e-06,
+      "loss": 8.6032,
+      "step": 144350
+    },
+    {
+      "epoch": 0.7209168768258883,
+      "grad_norm": 0.0898914784193039,
+      "learning_rate": 8.39345165085484e-06,
+      "loss": 8.5895,
+      "step": 144360
+    },
+    {
+      "epoch": 0.7209668156508278,
+      "grad_norm": 0.08764638006687164,
+      "learning_rate": 8.39194973591329e-06,
+      "loss": 8.5997,
+      "step": 144370
+    },
+    {
+      "epoch": 0.7210167544757672,
+      "grad_norm": 0.08820385485887527,
+      "learning_rate": 8.39044782097174e-06,
+      "loss": 8.6189,
+      "step": 144380
+    },
+    {
+      "epoch": 0.7210666933007066,
+      "grad_norm": 0.08908724784851074,
+      "learning_rate": 8.388945906030188e-06,
+      "loss": 8.6078,
+      "step": 144390
+    },
+    {
+      "epoch": 0.721116632125646,
+      "grad_norm": 0.0920150950551033,
+      "learning_rate": 8.387443991088638e-06,
+      "loss": 8.6197,
+      "step": 144400
+    },
+    {
+      "epoch": 0.7211665709505856,
+      "grad_norm": 0.09331239759922028,
+      "learning_rate": 8.385942076147088e-06,
+      "loss": 8.6267,
+      "step": 144410
+    },
+    {
+      "epoch": 0.721216509775525,
+      "grad_norm": 0.08835045248270035,
+      "learning_rate": 8.384440161205537e-06,
+      "loss": 8.599,
+      "step": 144420
+    },
+    {
+      "epoch": 0.7212664486004644,
+      "grad_norm": 0.0986829549074173,
+      "learning_rate": 8.382938246263987e-06,
+      "loss": 8.6299,
+      "step": 144430
+    },
+    {
+      "epoch": 0.7213163874254038,
+      "grad_norm": 0.09488464146852493,
+      "learning_rate": 8.381436331322435e-06,
+      "loss": 8.6146,
+      "step": 144440
+    },
+    {
+      "epoch": 0.7213663262503434,
+      "grad_norm": 0.09445064514875412,
+      "learning_rate": 8.379934416380886e-06,
+      "loss": 8.6191,
+      "step": 144450
+    },
+    {
+      "epoch": 0.7214162650752828,
+      "grad_norm": 0.10629602521657944,
+      "learning_rate": 8.378432501439336e-06,
+      "loss": 8.6149,
+      "step": 144460
+    },
+    {
+      "epoch": 0.7214662039002222,
+      "grad_norm": 0.09508983045816422,
+      "learning_rate": 8.376930586497784e-06,
+      "loss": 8.6109,
+      "step": 144470
+    },
+    {
+      "epoch": 0.7215161427251616,
+      "grad_norm": 0.0893370732665062,
+      "learning_rate": 8.375428671556234e-06,
+      "loss": 8.6295,
+      "step": 144480
+    },
+    {
+      "epoch": 0.7215660815501012,
+      "grad_norm": 0.09526433795690536,
+      "learning_rate": 8.373926756614683e-06,
+      "loss": 8.6181,
+      "step": 144490
+    },
+    {
+      "epoch": 0.7216160203750406,
+      "grad_norm": 0.08652014285326004,
+      "learning_rate": 8.372424841673133e-06,
+      "loss": 8.6274,
+      "step": 144500
+    },
+    {
+      "epoch": 0.72166595919998,
+      "grad_norm": 0.09043004363775253,
+      "learning_rate": 8.370922926731583e-06,
+      "loss": 8.6233,
+      "step": 144510
+    },
+    {
+      "epoch": 0.7217158980249194,
+      "grad_norm": 0.09349585324525833,
+      "learning_rate": 8.369421011790034e-06,
+      "loss": 8.6119,
+      "step": 144520
+    },
+    {
+      "epoch": 0.721765836849859,
+      "grad_norm": 0.09839218109846115,
+      "learning_rate": 8.367919096848482e-06,
+      "loss": 8.5964,
+      "step": 144530
+    },
+    {
+      "epoch": 0.7218157756747984,
+      "grad_norm": 0.09267604351043701,
+      "learning_rate": 8.36641718190693e-06,
+      "loss": 8.6024,
+      "step": 144540
+    },
+    {
+      "epoch": 0.7218657144997378,
+      "grad_norm": 0.08900393545627594,
+      "learning_rate": 8.36491526696538e-06,
+      "loss": 8.6205,
+      "step": 144550
+    },
+    {
+      "epoch": 0.7219156533246772,
+      "grad_norm": 0.08693215250968933,
+      "learning_rate": 8.36341335202383e-06,
+      "loss": 8.6103,
+      "step": 144560
+    },
+    {
+      "epoch": 0.7219655921496168,
+      "grad_norm": 0.09515117853879929,
+      "learning_rate": 8.361911437082281e-06,
+      "loss": 8.6159,
+      "step": 144570
+    },
+    {
+      "epoch": 0.7220155309745562,
+      "grad_norm": 0.08883559703826904,
+      "learning_rate": 8.36040952214073e-06,
+      "loss": 8.6283,
+      "step": 144580
+    },
+    {
+      "epoch": 0.7220654697994956,
+      "grad_norm": 0.0927378311753273,
+      "learning_rate": 8.358907607199178e-06,
+      "loss": 8.6171,
+      "step": 144590
+    },
+    {
+      "epoch": 0.722115408624435,
+      "grad_norm": 0.09397240728139877,
+      "learning_rate": 8.357405692257628e-06,
+      "loss": 8.6131,
+      "step": 144600
+    },
+    {
+      "epoch": 0.7221653474493746,
+      "grad_norm": 0.0927707701921463,
+      "learning_rate": 8.355903777316078e-06,
+      "loss": 8.6159,
+      "step": 144610
+    },
+    {
+      "epoch": 0.722215286274314,
+      "grad_norm": 0.09005044400691986,
+      "learning_rate": 8.354401862374529e-06,
+      "loss": 8.6134,
+      "step": 144620
+    },
+    {
+      "epoch": 0.7222652250992534,
+      "grad_norm": 0.08939428627490997,
+      "learning_rate": 8.352899947432977e-06,
+      "loss": 8.6081,
+      "step": 144630
+    },
+    {
+      "epoch": 0.7223151639241928,
+      "grad_norm": 0.09121254086494446,
+      "learning_rate": 8.351398032491425e-06,
+      "loss": 8.6085,
+      "step": 144640
+    },
+    {
+      "epoch": 0.7223651027491323,
+      "grad_norm": 0.08655776083469391,
+      "learning_rate": 8.349896117549876e-06,
+      "loss": 8.6199,
+      "step": 144650
+    },
+    {
+      "epoch": 0.7224150415740718,
+      "grad_norm": 0.09415041655302048,
+      "learning_rate": 8.348394202608326e-06,
+      "loss": 8.605,
+      "step": 144660
+    },
+    {
+      "epoch": 0.7224649803990112,
+      "grad_norm": 0.09268126636743546,
+      "learning_rate": 8.346892287666776e-06,
+      "loss": 8.6162,
+      "step": 144670
+    },
+    {
+      "epoch": 0.7225149192239506,
+      "grad_norm": 0.1028081476688385,
+      "learning_rate": 8.345390372725226e-06,
+      "loss": 8.6293,
+      "step": 144680
+    },
+    {
+      "epoch": 0.7225648580488901,
+      "grad_norm": 0.09145328402519226,
+      "learning_rate": 8.343888457783673e-06,
+      "loss": 8.6075,
+      "step": 144690
+    },
+    {
+      "epoch": 0.7226147968738296,
+      "grad_norm": 0.08995439857244492,
+      "learning_rate": 8.342386542842123e-06,
+      "loss": 8.5961,
+      "step": 144700
+    },
+    {
+      "epoch": 0.722664735698769,
+      "grad_norm": 0.08796056360006332,
+      "learning_rate": 8.340884627900573e-06,
+      "loss": 8.6194,
+      "step": 144710
+    },
+    {
+      "epoch": 0.7227146745237084,
+      "grad_norm": 0.0950382873415947,
+      "learning_rate": 8.339382712959024e-06,
+      "loss": 8.6196,
+      "step": 144720
+    },
+    {
+      "epoch": 0.7227646133486479,
+      "grad_norm": 0.09233977645635605,
+      "learning_rate": 8.337880798017474e-06,
+      "loss": 8.6228,
+      "step": 144730
+    },
+    {
+      "epoch": 0.7228145521735874,
+      "grad_norm": 0.09013422578573227,
+      "learning_rate": 8.33637888307592e-06,
+      "loss": 8.6222,
+      "step": 144740
+    },
+    {
+      "epoch": 0.7228644909985268,
+      "grad_norm": 0.0920189619064331,
+      "learning_rate": 8.33487696813437e-06,
+      "loss": 8.6278,
+      "step": 144750
+    },
+    {
+      "epoch": 0.7229144298234662,
+      "grad_norm": 0.09212841838598251,
+      "learning_rate": 8.333375053192821e-06,
+      "loss": 8.6109,
+      "step": 144760
+    },
+    {
+      "epoch": 0.7229643686484057,
+      "grad_norm": 0.09851231426000595,
+      "learning_rate": 8.331873138251271e-06,
+      "loss": 8.6134,
+      "step": 144770
+    },
+    {
+      "epoch": 0.7230143074733452,
+      "grad_norm": 0.09373413026332855,
+      "learning_rate": 8.330371223309721e-06,
+      "loss": 8.5994,
+      "step": 144780
+    },
+    {
+      "epoch": 0.7230642462982846,
+      "grad_norm": 0.09392852336168289,
+      "learning_rate": 8.328869308368168e-06,
+      "loss": 8.628,
+      "step": 144790
+    },
+    {
+      "epoch": 0.723114185123224,
+      "grad_norm": 0.09037362039089203,
+      "learning_rate": 8.327367393426618e-06,
+      "loss": 8.6152,
+      "step": 144800
+    },
+    {
+      "epoch": 0.7231641239481635,
+      "grad_norm": 0.08910314738750458,
+      "learning_rate": 8.325865478485068e-06,
+      "loss": 8.5945,
+      "step": 144810
+    },
+    {
+      "epoch": 0.723214062773103,
+      "grad_norm": 0.08930026739835739,
+      "learning_rate": 8.324363563543519e-06,
+      "loss": 8.61,
+      "step": 144820
+    },
+    {
+      "epoch": 0.7232640015980424,
+      "grad_norm": 0.09371279925107956,
+      "learning_rate": 8.322861648601969e-06,
+      "loss": 8.6156,
+      "step": 144830
+    },
+    {
+      "epoch": 0.7233139404229818,
+      "grad_norm": 0.0969967246055603,
+      "learning_rate": 8.321359733660417e-06,
+      "loss": 8.6191,
+      "step": 144840
+    },
+    {
+      "epoch": 0.7233638792479213,
+      "grad_norm": 0.0957605391740799,
+      "learning_rate": 8.319857818718866e-06,
+      "loss": 8.6104,
+      "step": 144850
+    },
+    {
+      "epoch": 0.7234138180728608,
+      "grad_norm": 0.09280975908041,
+      "learning_rate": 8.318355903777316e-06,
+      "loss": 8.6152,
+      "step": 144860
+    },
+    {
+      "epoch": 0.7234637568978002,
+      "grad_norm": 0.09555871039628983,
+      "learning_rate": 8.316853988835766e-06,
+      "loss": 8.6135,
+      "step": 144870
+    },
+    {
+      "epoch": 0.7235136957227396,
+      "grad_norm": 0.08848311007022858,
+      "learning_rate": 8.315352073894216e-06,
+      "loss": 8.6073,
+      "step": 144880
+    },
+    {
+      "epoch": 0.7235636345476791,
+      "grad_norm": 0.09260237962007523,
+      "learning_rate": 8.313850158952665e-06,
+      "loss": 8.613,
+      "step": 144890
+    },
+    {
+      "epoch": 0.7236135733726186,
+      "grad_norm": 0.08797964453697205,
+      "learning_rate": 8.312348244011113e-06,
+      "loss": 8.6129,
+      "step": 144900
+    },
+    {
+      "epoch": 0.723663512197558,
+      "grad_norm": 0.09759566187858582,
+      "learning_rate": 8.310846329069563e-06,
+      "loss": 8.6059,
+      "step": 144910
+    },
+    {
+      "epoch": 0.7237134510224974,
+      "grad_norm": 0.09639012068510056,
+      "learning_rate": 8.309344414128014e-06,
+      "loss": 8.607,
+      "step": 144920
+    },
+    {
+      "epoch": 0.7237633898474369,
+      "grad_norm": 0.09098049253225327,
+      "learning_rate": 8.307842499186464e-06,
+      "loss": 8.6158,
+      "step": 144930
+    },
+    {
+      "epoch": 0.7238133286723764,
+      "grad_norm": 0.08897264301776886,
+      "learning_rate": 8.306340584244912e-06,
+      "loss": 8.6348,
+      "step": 144940
+    },
+    {
+      "epoch": 0.7238632674973158,
+      "grad_norm": 0.09100917726755142,
+      "learning_rate": 8.30483866930336e-06,
+      "loss": 8.6146,
+      "step": 144950
+    },
+    {
+      "epoch": 0.7239132063222552,
+      "grad_norm": 0.0964989885687828,
+      "learning_rate": 8.303336754361811e-06,
+      "loss": 8.6144,
+      "step": 144960
+    },
+    {
+      "epoch": 0.7239631451471947,
+      "grad_norm": 0.10053501278162003,
+      "learning_rate": 8.301834839420261e-06,
+      "loss": 8.6179,
+      "step": 144970
+    },
+    {
+      "epoch": 0.7240130839721342,
+      "grad_norm": 0.09664203226566315,
+      "learning_rate": 8.300332924478711e-06,
+      "loss": 8.6196,
+      "step": 144980
+    },
+    {
+      "epoch": 0.7240630227970736,
+      "grad_norm": 0.08678041398525238,
+      "learning_rate": 8.298831009537161e-06,
+      "loss": 8.6115,
+      "step": 144990
+    },
+    {
+      "epoch": 0.724112961622013,
+      "grad_norm": 0.09090323746204376,
+      "learning_rate": 8.29732909459561e-06,
+      "loss": 8.6013,
+      "step": 145000
+    },
+    {
+      "epoch": 0.7241629004469525,
+      "grad_norm": 0.09203115105628967,
+      "learning_rate": 8.295827179654058e-06,
+      "loss": 8.6139,
+      "step": 145010
+    },
+    {
+      "epoch": 0.724212839271892,
+      "grad_norm": 0.0934581458568573,
+      "learning_rate": 8.294325264712509e-06,
+      "loss": 8.6223,
+      "step": 145020
+    },
+    {
+      "epoch": 0.7242627780968314,
+      "grad_norm": 0.09179078042507172,
+      "learning_rate": 8.292823349770959e-06,
+      "loss": 8.6099,
+      "step": 145030
+    },
+    {
+      "epoch": 0.7243127169217708,
+      "grad_norm": 0.09519582986831665,
+      "learning_rate": 8.291321434829409e-06,
+      "loss": 8.6186,
+      "step": 145040
+    },
+    {
+      "epoch": 0.7243626557467103,
+      "grad_norm": 0.08919553458690643,
+      "learning_rate": 8.289819519887857e-06,
+      "loss": 8.607,
+      "step": 145050
+    },
+    {
+      "epoch": 0.7244125945716497,
+      "grad_norm": 0.09014479070901871,
+      "learning_rate": 8.288317604946306e-06,
+      "loss": 8.5982,
+      "step": 145060
+    },
+    {
+      "epoch": 0.7244625333965892,
+      "grad_norm": 0.0924357920885086,
+      "learning_rate": 8.286815690004756e-06,
+      "loss": 8.6198,
+      "step": 145070
+    },
+    {
+      "epoch": 0.7245124722215286,
+      "grad_norm": 0.09133771061897278,
+      "learning_rate": 8.285313775063206e-06,
+      "loss": 8.6248,
+      "step": 145080
+    },
+    {
+      "epoch": 0.7245624110464681,
+      "grad_norm": 0.08713619410991669,
+      "learning_rate": 8.283811860121656e-06,
+      "loss": 8.6429,
+      "step": 145090
+    },
+    {
+      "epoch": 0.7246123498714075,
+      "grad_norm": 0.09331973642110825,
+      "learning_rate": 8.282309945180105e-06,
+      "loss": 8.6161,
+      "step": 145100
+    },
+    {
+      "epoch": 0.724662288696347,
+      "grad_norm": 0.08959534019231796,
+      "learning_rate": 8.280808030238553e-06,
+      "loss": 8.6193,
+      "step": 145110
+    },
+    {
+      "epoch": 0.7247122275212864,
+      "grad_norm": 0.09544025361537933,
+      "learning_rate": 8.279306115297004e-06,
+      "loss": 8.6113,
+      "step": 145120
+    },
+    {
+      "epoch": 0.7247621663462259,
+      "grad_norm": 0.08845238387584686,
+      "learning_rate": 8.277804200355454e-06,
+      "loss": 8.6096,
+      "step": 145130
+    },
+    {
+      "epoch": 0.7248121051711653,
+      "grad_norm": 0.09223693609237671,
+      "learning_rate": 8.276302285413904e-06,
+      "loss": 8.606,
+      "step": 145140
+    },
+    {
+      "epoch": 0.7248620439961048,
+      "grad_norm": 0.09197333455085754,
+      "learning_rate": 8.274800370472352e-06,
+      "loss": 8.6065,
+      "step": 145150
+    },
+    {
+      "epoch": 0.7249119828210442,
+      "grad_norm": 0.0903843566775322,
+      "learning_rate": 8.273298455530803e-06,
+      "loss": 8.6257,
+      "step": 145160
+    },
+    {
+      "epoch": 0.7249619216459837,
+      "grad_norm": 0.08799325674772263,
+      "learning_rate": 8.271796540589251e-06,
+      "loss": 8.6078,
+      "step": 145170
+    },
+    {
+      "epoch": 0.7250118604709231,
+      "grad_norm": 0.09171675890684128,
+      "learning_rate": 8.270294625647701e-06,
+      "loss": 8.6055,
+      "step": 145180
+    },
+    {
+      "epoch": 0.7250617992958626,
+      "grad_norm": 0.09235136955976486,
+      "learning_rate": 8.268792710706151e-06,
+      "loss": 8.5921,
+      "step": 145190
+    },
+    {
+      "epoch": 0.725111738120802,
+      "grad_norm": 0.09448470175266266,
+      "learning_rate": 8.2672907957646e-06,
+      "loss": 8.6096,
+      "step": 145200
+    },
+    {
+      "epoch": 0.7251616769457415,
+      "grad_norm": 0.09224043041467667,
+      "learning_rate": 8.26578888082305e-06,
+      "loss": 8.6311,
+      "step": 145210
+    },
+    {
+      "epoch": 0.7252116157706809,
+      "grad_norm": 0.09602584689855576,
+      "learning_rate": 8.264286965881499e-06,
+      "loss": 8.6032,
+      "step": 145220
+    },
+    {
+      "epoch": 0.7252615545956204,
+      "grad_norm": 0.09572752565145493,
+      "learning_rate": 8.262785050939949e-06,
+      "loss": 8.6264,
+      "step": 145230
+    },
+    {
+      "epoch": 0.7253114934205598,
+      "grad_norm": 0.08928846567869186,
+      "learning_rate": 8.261283135998399e-06,
+      "loss": 8.6099,
+      "step": 145240
+    },
+    {
+      "epoch": 0.7253614322454993,
+      "grad_norm": 0.09179164469242096,
+      "learning_rate": 8.259781221056847e-06,
+      "loss": 8.6001,
+      "step": 145250
+    },
+    {
+      "epoch": 0.7254113710704387,
+      "grad_norm": 0.09306824952363968,
+      "learning_rate": 8.258279306115298e-06,
+      "loss": 8.6105,
+      "step": 145260
+    },
+    {
+      "epoch": 0.7254613098953782,
+      "grad_norm": 0.08858099579811096,
+      "learning_rate": 8.256777391173746e-06,
+      "loss": 8.6093,
+      "step": 145270
+    },
+    {
+      "epoch": 0.7255112487203176,
+      "grad_norm": 0.09503430128097534,
+      "learning_rate": 8.255275476232196e-06,
+      "loss": 8.5933,
+      "step": 145280
+    },
+    {
+      "epoch": 0.7255611875452571,
+      "grad_norm": 0.0941505953669548,
+      "learning_rate": 8.253773561290646e-06,
+      "loss": 8.6164,
+      "step": 145290
+    },
+    {
+      "epoch": 0.7256111263701965,
+      "grad_norm": 0.0895562395453453,
+      "learning_rate": 8.252271646349095e-06,
+      "loss": 8.6136,
+      "step": 145300
+    },
+    {
+      "epoch": 0.725661065195136,
+      "grad_norm": 0.0945998951792717,
+      "learning_rate": 8.250769731407545e-06,
+      "loss": 8.6132,
+      "step": 145310
+    },
+    {
+      "epoch": 0.7257110040200754,
+      "grad_norm": 0.09146866202354431,
+      "learning_rate": 8.249267816465995e-06,
+      "loss": 8.6185,
+      "step": 145320
+    },
+    {
+      "epoch": 0.7257609428450149,
+      "grad_norm": 0.09208477288484573,
+      "learning_rate": 8.247765901524444e-06,
+      "loss": 8.6098,
+      "step": 145330
+    },
+    {
+      "epoch": 0.7258108816699543,
+      "grad_norm": 0.10175745189189911,
+      "learning_rate": 8.246263986582894e-06,
+      "loss": 8.607,
+      "step": 145340
+    },
+    {
+      "epoch": 0.7258608204948938,
+      "grad_norm": 0.08928140252828598,
+      "learning_rate": 8.244762071641342e-06,
+      "loss": 8.611,
+      "step": 145350
+    },
+    {
+      "epoch": 0.7259107593198332,
+      "grad_norm": 0.08949202299118042,
+      "learning_rate": 8.243260156699793e-06,
+      "loss": 8.6125,
+      "step": 145360
+    },
+    {
+      "epoch": 0.7259606981447726,
+      "grad_norm": 0.10165835916996002,
+      "learning_rate": 8.241758241758243e-06,
+      "loss": 8.6136,
+      "step": 145370
+    },
+    {
+      "epoch": 0.7260106369697121,
+      "grad_norm": 0.09630440920591354,
+      "learning_rate": 8.240256326816691e-06,
+      "loss": 8.6119,
+      "step": 145380
+    },
+    {
+      "epoch": 0.7260605757946516,
+      "grad_norm": 0.09105534851551056,
+      "learning_rate": 8.238754411875141e-06,
+      "loss": 8.6037,
+      "step": 145390
+    },
+    {
+      "epoch": 0.726110514619591,
+      "grad_norm": 0.09389536082744598,
+      "learning_rate": 8.23725249693359e-06,
+      "loss": 8.5986,
+      "step": 145400
+    },
+    {
+      "epoch": 0.7261604534445304,
+      "grad_norm": 0.08979053050279617,
+      "learning_rate": 8.23575058199204e-06,
+      "loss": 8.6063,
+      "step": 145410
+    },
+    {
+      "epoch": 0.7262103922694699,
+      "grad_norm": 0.09311927855014801,
+      "learning_rate": 8.23424866705049e-06,
+      "loss": 8.6059,
+      "step": 145420
+    },
+    {
+      "epoch": 0.7262603310944094,
+      "grad_norm": 0.08796468377113342,
+      "learning_rate": 8.232746752108939e-06,
+      "loss": 8.6218,
+      "step": 145430
+    },
+    {
+      "epoch": 0.7263102699193488,
+      "grad_norm": 0.0910877138376236,
+      "learning_rate": 8.231244837167389e-06,
+      "loss": 8.611,
+      "step": 145440
+    },
+    {
+      "epoch": 0.7263602087442882,
+      "grad_norm": 0.09477691352367401,
+      "learning_rate": 8.229742922225837e-06,
+      "loss": 8.6215,
+      "step": 145450
+    },
+    {
+      "epoch": 0.7264101475692277,
+      "grad_norm": 0.0948781967163086,
+      "learning_rate": 8.228241007284288e-06,
+      "loss": 8.5978,
+      "step": 145460
+    },
+    {
+      "epoch": 0.7264600863941671,
+      "grad_norm": 0.08527876436710358,
+      "learning_rate": 8.226739092342738e-06,
+      "loss": 8.6191,
+      "step": 145470
+    },
+    {
+      "epoch": 0.7265100252191066,
+      "grad_norm": 0.09014835208654404,
+      "learning_rate": 8.225237177401188e-06,
+      "loss": 8.6033,
+      "step": 145480
+    },
+    {
+      "epoch": 0.726559964044046,
+      "grad_norm": 0.09688574820756912,
+      "learning_rate": 8.223735262459636e-06,
+      "loss": 8.6061,
+      "step": 145490
+    },
+    {
+      "epoch": 0.7266099028689855,
+      "grad_norm": 0.08917857706546783,
+      "learning_rate": 8.222233347518085e-06,
+      "loss": 8.6284,
+      "step": 145500
+    },
+    {
+      "epoch": 0.726659841693925,
+      "grad_norm": 0.09284509718418121,
+      "learning_rate": 8.220731432576535e-06,
+      "loss": 8.6115,
+      "step": 145510
+    },
+    {
+      "epoch": 0.7267097805188644,
+      "grad_norm": 0.08809034526348114,
+      "learning_rate": 8.219229517634985e-06,
+      "loss": 8.6158,
+      "step": 145520
+    },
+    {
+      "epoch": 0.7267597193438038,
+      "grad_norm": 0.09083345532417297,
+      "learning_rate": 8.217727602693435e-06,
+      "loss": 8.6272,
+      "step": 145530
+    },
+    {
+      "epoch": 0.7268096581687433,
+      "grad_norm": 0.08907760679721832,
+      "learning_rate": 8.216225687751884e-06,
+      "loss": 8.6008,
+      "step": 145540
+    },
+    {
+      "epoch": 0.7268595969936827,
+      "grad_norm": 0.08739221096038818,
+      "learning_rate": 8.214723772810332e-06,
+      "loss": 8.6271,
+      "step": 145550
+    },
+    {
+      "epoch": 0.7269095358186222,
+      "grad_norm": 0.0876481756567955,
+      "learning_rate": 8.213221857868783e-06,
+      "loss": 8.6061,
+      "step": 145560
+    },
+    {
+      "epoch": 0.7269594746435616,
+      "grad_norm": 0.08937806636095047,
+      "learning_rate": 8.211719942927233e-06,
+      "loss": 8.629,
+      "step": 145570
+    },
+    {
+      "epoch": 0.7270094134685011,
+      "grad_norm": 0.08746001869440079,
+      "learning_rate": 8.210218027985683e-06,
+      "loss": 8.6011,
+      "step": 145580
+    },
+    {
+      "epoch": 0.7270593522934405,
+      "grad_norm": 0.09080873429775238,
+      "learning_rate": 8.208716113044131e-06,
+      "loss": 8.6195,
+      "step": 145590
+    },
+    {
+      "epoch": 0.72710929111838,
+      "grad_norm": 0.09297414124011993,
+      "learning_rate": 8.20721419810258e-06,
+      "loss": 8.6009,
+      "step": 145600
+    },
+    {
+      "epoch": 0.7271592299433194,
+      "grad_norm": 0.09286045283079147,
+      "learning_rate": 8.20571228316103e-06,
+      "loss": 8.6177,
+      "step": 145610
+    },
+    {
+      "epoch": 0.7272091687682589,
+      "grad_norm": 0.08479689061641693,
+      "learning_rate": 8.20421036821948e-06,
+      "loss": 8.6263,
+      "step": 145620
+    },
+    {
+      "epoch": 0.7272591075931983,
+      "grad_norm": 0.09006743133068085,
+      "learning_rate": 8.20270845327793e-06,
+      "loss": 8.598,
+      "step": 145630
+    },
+    {
+      "epoch": 0.7273090464181378,
+      "grad_norm": 0.09547232836484909,
+      "learning_rate": 8.20120653833638e-06,
+      "loss": 8.5996,
+      "step": 145640
+    },
+    {
+      "epoch": 0.7273589852430772,
+      "grad_norm": 0.0883583128452301,
+      "learning_rate": 8.199704623394827e-06,
+      "loss": 8.6021,
+      "step": 145650
+    },
+    {
+      "epoch": 0.7274089240680167,
+      "grad_norm": 0.09698734432458878,
+      "learning_rate": 8.198202708453278e-06,
+      "loss": 8.6079,
+      "step": 145660
+    },
+    {
+      "epoch": 0.7274588628929561,
+      "grad_norm": 0.09151437133550644,
+      "learning_rate": 8.196700793511728e-06,
+      "loss": 8.6057,
+      "step": 145670
+    },
+    {
+      "epoch": 0.7275088017178956,
+      "grad_norm": 0.09400162845849991,
+      "learning_rate": 8.195198878570178e-06,
+      "loss": 8.6039,
+      "step": 145680
+    },
+    {
+      "epoch": 0.727558740542835,
+      "grad_norm": 0.08918856829404831,
+      "learning_rate": 8.193696963628628e-06,
+      "loss": 8.6045,
+      "step": 145690
+    },
+    {
+      "epoch": 0.7276086793677745,
+      "grad_norm": 0.09551417827606201,
+      "learning_rate": 8.192195048687075e-06,
+      "loss": 8.6184,
+      "step": 145700
+    },
+    {
+      "epoch": 0.7276586181927139,
+      "grad_norm": 0.09140796959400177,
+      "learning_rate": 8.190693133745525e-06,
+      "loss": 8.5989,
+      "step": 145710
+    },
+    {
+      "epoch": 0.7277085570176534,
+      "grad_norm": 0.09093940258026123,
+      "learning_rate": 8.189191218803975e-06,
+      "loss": 8.6054,
+      "step": 145720
+    },
+    {
+      "epoch": 0.7277584958425928,
+      "grad_norm": 0.08896015584468842,
+      "learning_rate": 8.187689303862425e-06,
+      "loss": 8.6201,
+      "step": 145730
+    },
+    {
+      "epoch": 0.7278084346675323,
+      "grad_norm": 0.08995377272367477,
+      "learning_rate": 8.186187388920876e-06,
+      "loss": 8.6244,
+      "step": 145740
+    },
+    {
+      "epoch": 0.7278583734924717,
+      "grad_norm": 0.08884330838918686,
+      "learning_rate": 8.184685473979322e-06,
+      "loss": 8.6063,
+      "step": 145750
+    },
+    {
+      "epoch": 0.7279083123174112,
+      "grad_norm": 0.09230002015829086,
+      "learning_rate": 8.183183559037773e-06,
+      "loss": 8.6138,
+      "step": 145760
+    },
+    {
+      "epoch": 0.7279582511423506,
+      "grad_norm": 0.0911339670419693,
+      "learning_rate": 8.181681644096223e-06,
+      "loss": 8.5995,
+      "step": 145770
+    },
+    {
+      "epoch": 0.7280081899672901,
+      "grad_norm": 0.08715797960758209,
+      "learning_rate": 8.180179729154673e-06,
+      "loss": 8.613,
+      "step": 145780
+    },
+    {
+      "epoch": 0.7280581287922295,
+      "grad_norm": 0.09273428469896317,
+      "learning_rate": 8.178677814213123e-06,
+      "loss": 8.6247,
+      "step": 145790
+    },
+    {
+      "epoch": 0.728108067617169,
+      "grad_norm": 0.09207670390605927,
+      "learning_rate": 8.177175899271572e-06,
+      "loss": 8.6245,
+      "step": 145800
+    },
+    {
+      "epoch": 0.7281580064421084,
+      "grad_norm": 0.09679616987705231,
+      "learning_rate": 8.17567398433002e-06,
+      "loss": 8.6092,
+      "step": 145810
+    },
+    {
+      "epoch": 0.7282079452670479,
+      "grad_norm": 0.09426999092102051,
+      "learning_rate": 8.17417206938847e-06,
+      "loss": 8.6225,
+      "step": 145820
+    },
+    {
+      "epoch": 0.7282578840919873,
+      "grad_norm": 0.09463541954755783,
+      "learning_rate": 8.17267015444692e-06,
+      "loss": 8.6008,
+      "step": 145830
+    },
+    {
+      "epoch": 0.7283078229169268,
+      "grad_norm": 0.09443298727273941,
+      "learning_rate": 8.17116823950537e-06,
+      "loss": 8.5991,
+      "step": 145840
+    },
+    {
+      "epoch": 0.7283577617418662,
+      "grad_norm": 0.09195612370967865,
+      "learning_rate": 8.169666324563819e-06,
+      "loss": 8.5961,
+      "step": 145850
+    },
+    {
+      "epoch": 0.7284077005668057,
+      "grad_norm": 0.08928360790014267,
+      "learning_rate": 8.168164409622268e-06,
+      "loss": 8.634,
+      "step": 145860
+    },
+    {
+      "epoch": 0.7284576393917451,
+      "grad_norm": 0.09235423803329468,
+      "learning_rate": 8.166662494680718e-06,
+      "loss": 8.6134,
+      "step": 145870
+    },
+    {
+      "epoch": 0.7285075782166845,
+      "grad_norm": 0.09218775480985641,
+      "learning_rate": 8.165160579739168e-06,
+      "loss": 8.6084,
+      "step": 145880
+    },
+    {
+      "epoch": 0.728557517041624,
+      "grad_norm": 0.09094615280628204,
+      "learning_rate": 8.163658664797618e-06,
+      "loss": 8.6155,
+      "step": 145890
+    },
+    {
+      "epoch": 0.7286074558665635,
+      "grad_norm": 0.08593304455280304,
+      "learning_rate": 8.162156749856067e-06,
+      "loss": 8.6124,
+      "step": 145900
+    },
+    {
+      "epoch": 0.7286573946915029,
+      "grad_norm": 0.09099908173084259,
+      "learning_rate": 8.160654834914515e-06,
+      "loss": 8.6163,
+      "step": 145910
+    },
+    {
+      "epoch": 0.7287073335164423,
+      "grad_norm": 0.0902167484164238,
+      "learning_rate": 8.159152919972965e-06,
+      "loss": 8.6227,
+      "step": 145920
+    },
+    {
+      "epoch": 0.7287572723413818,
+      "grad_norm": 0.09443819522857666,
+      "learning_rate": 8.157651005031415e-06,
+      "loss": 8.6165,
+      "step": 145930
+    },
+    {
+      "epoch": 0.7288072111663213,
+      "grad_norm": 0.09116199612617493,
+      "learning_rate": 8.156149090089866e-06,
+      "loss": 8.6201,
+      "step": 145940
+    },
+    {
+      "epoch": 0.7288571499912607,
+      "grad_norm": 0.09759961068630219,
+      "learning_rate": 8.154647175148314e-06,
+      "loss": 8.6189,
+      "step": 145950
+    },
+    {
+      "epoch": 0.7289070888162001,
+      "grad_norm": 0.09084867686033249,
+      "learning_rate": 8.153145260206764e-06,
+      "loss": 8.5922,
+      "step": 145960
+    },
+    {
+      "epoch": 0.7289570276411396,
+      "grad_norm": 0.09356144815683365,
+      "learning_rate": 8.151643345265213e-06,
+      "loss": 8.6069,
+      "step": 145970
+    },
+    {
+      "epoch": 0.7290069664660791,
+      "grad_norm": 0.08888962864875793,
+      "learning_rate": 8.150141430323663e-06,
+      "loss": 8.6001,
+      "step": 145980
+    },
+    {
+      "epoch": 0.7290569052910185,
+      "grad_norm": 0.08829820901155472,
+      "learning_rate": 8.148639515382113e-06,
+      "loss": 8.6117,
+      "step": 145990
+    },
+    {
+      "epoch": 0.7291068441159579,
+      "grad_norm": 0.092521533370018,
+      "learning_rate": 8.147137600440562e-06,
+      "loss": 8.6116,
+      "step": 146000
+    },
+    {
+      "epoch": 0.7291567829408974,
+      "grad_norm": 0.09116620570421219,
+      "learning_rate": 8.145635685499012e-06,
+      "loss": 8.6103,
+      "step": 146010
+    },
+    {
+      "epoch": 0.7292067217658369,
+      "grad_norm": 0.09370893239974976,
+      "learning_rate": 8.14413377055746e-06,
+      "loss": 8.5975,
+      "step": 146020
+    },
+    {
+      "epoch": 0.7292566605907763,
+      "grad_norm": 0.09606023132801056,
+      "learning_rate": 8.14263185561591e-06,
+      "loss": 8.6029,
+      "step": 146030
+    },
+    {
+      "epoch": 0.7293065994157157,
+      "grad_norm": 0.08850355446338654,
+      "learning_rate": 8.14112994067436e-06,
+      "loss": 8.6077,
+      "step": 146040
+    },
+    {
+      "epoch": 0.7293565382406552,
+      "grad_norm": 0.09388591349124908,
+      "learning_rate": 8.13962802573281e-06,
+      "loss": 8.6063,
+      "step": 146050
+    },
+    {
+      "epoch": 0.7294064770655947,
+      "grad_norm": 0.09386676549911499,
+      "learning_rate": 8.13812611079126e-06,
+      "loss": 8.5971,
+      "step": 146060
+    },
+    {
+      "epoch": 0.7294564158905341,
+      "grad_norm": 0.09026592969894409,
+      "learning_rate": 8.136624195849708e-06,
+      "loss": 8.5921,
+      "step": 146070
+    },
+    {
+      "epoch": 0.7295063547154735,
+      "grad_norm": 0.09007930010557175,
+      "learning_rate": 8.135122280908158e-06,
+      "loss": 8.6017,
+      "step": 146080
+    },
+    {
+      "epoch": 0.729556293540413,
+      "grad_norm": 0.0955241322517395,
+      "learning_rate": 8.133620365966608e-06,
+      "loss": 8.6081,
+      "step": 146090
+    },
+    {
+      "epoch": 0.7296062323653525,
+      "grad_norm": 0.09348488599061966,
+      "learning_rate": 8.132118451025057e-06,
+      "loss": 8.5981,
+      "step": 146100
+    },
+    {
+      "epoch": 0.7296561711902919,
+      "grad_norm": 0.08962753415107727,
+      "learning_rate": 8.130616536083507e-06,
+      "loss": 8.6063,
+      "step": 146110
+    },
+    {
+      "epoch": 0.7297061100152313,
+      "grad_norm": 0.09236380457878113,
+      "learning_rate": 8.129114621141957e-06,
+      "loss": 8.5859,
+      "step": 146120
+    },
+    {
+      "epoch": 0.7297560488401708,
+      "grad_norm": 0.08918324112892151,
+      "learning_rate": 8.127612706200406e-06,
+      "loss": 8.6115,
+      "step": 146130
+    },
+    {
+      "epoch": 0.7298059876651103,
+      "grad_norm": 0.090632364153862,
+      "learning_rate": 8.126110791258856e-06,
+      "loss": 8.6016,
+      "step": 146140
+    },
+    {
+      "epoch": 0.7298559264900497,
+      "grad_norm": 0.09161046892404556,
+      "learning_rate": 8.124608876317304e-06,
+      "loss": 8.5882,
+      "step": 146150
+    },
+    {
+      "epoch": 0.7299058653149891,
+      "grad_norm": 0.09371636062860489,
+      "learning_rate": 8.123106961375754e-06,
+      "loss": 8.5995,
+      "step": 146160
+    },
+    {
+      "epoch": 0.7299558041399286,
+      "grad_norm": 0.08753936737775803,
+      "learning_rate": 8.121605046434205e-06,
+      "loss": 8.5968,
+      "step": 146170
+    },
+    {
+      "epoch": 0.7300057429648681,
+      "grad_norm": 0.0922481119632721,
+      "learning_rate": 8.120103131492653e-06,
+      "loss": 8.5951,
+      "step": 146180
+    },
+    {
+      "epoch": 0.7300556817898075,
+      "grad_norm": 0.0926506519317627,
+      "learning_rate": 8.118601216551103e-06,
+      "loss": 8.603,
+      "step": 146190
+    },
+    {
+      "epoch": 0.7301056206147469,
+      "grad_norm": 0.08759573101997375,
+      "learning_rate": 8.117099301609552e-06,
+      "loss": 8.6076,
+      "step": 146200
+    },
+    {
+      "epoch": 0.7301555594396864,
+      "grad_norm": 0.09604066610336304,
+      "learning_rate": 8.115597386668002e-06,
+      "loss": 8.6029,
+      "step": 146210
+    },
+    {
+      "epoch": 0.7302054982646259,
+      "grad_norm": 0.08609198778867722,
+      "learning_rate": 8.114095471726452e-06,
+      "loss": 8.6056,
+      "step": 146220
+    },
+    {
+      "epoch": 0.7302554370895653,
+      "grad_norm": 0.08941318094730377,
+      "learning_rate": 8.1125935567849e-06,
+      "loss": 8.6103,
+      "step": 146230
+    },
+    {
+      "epoch": 0.7303053759145047,
+      "grad_norm": 0.09113612025976181,
+      "learning_rate": 8.11109164184335e-06,
+      "loss": 8.6069,
+      "step": 146240
+    },
+    {
+      "epoch": 0.7303553147394442,
+      "grad_norm": 0.09444758296012878,
+      "learning_rate": 8.1095897269018e-06,
+      "loss": 8.6201,
+      "step": 146250
+    },
+    {
+      "epoch": 0.7304052535643837,
+      "grad_norm": 0.0867748036980629,
+      "learning_rate": 8.10808781196025e-06,
+      "loss": 8.6056,
+      "step": 146260
+    },
+    {
+      "epoch": 0.7304551923893231,
+      "grad_norm": 0.09332028776407242,
+      "learning_rate": 8.1065858970187e-06,
+      "loss": 8.6122,
+      "step": 146270
+    },
+    {
+      "epoch": 0.7305051312142625,
+      "grad_norm": 0.09203875064849854,
+      "learning_rate": 8.10508398207715e-06,
+      "loss": 8.6157,
+      "step": 146280
+    },
+    {
+      "epoch": 0.730555070039202,
+      "grad_norm": 0.09554773569107056,
+      "learning_rate": 8.103582067135598e-06,
+      "loss": 8.5996,
+      "step": 146290
+    },
+    {
+      "epoch": 0.7306050088641415,
+      "grad_norm": 0.09731141477823257,
+      "learning_rate": 8.102080152194047e-06,
+      "loss": 8.5881,
+      "step": 146300
+    },
+    {
+      "epoch": 0.7306549476890809,
+      "grad_norm": 0.0938003659248352,
+      "learning_rate": 8.100578237252497e-06,
+      "loss": 8.6107,
+      "step": 146310
+    },
+    {
+      "epoch": 0.7307048865140203,
+      "grad_norm": 0.09249870479106903,
+      "learning_rate": 8.099076322310947e-06,
+      "loss": 8.6074,
+      "step": 146320
+    },
+    {
+      "epoch": 0.7307548253389597,
+      "grad_norm": 0.08894717693328857,
+      "learning_rate": 8.097574407369397e-06,
+      "loss": 8.6157,
+      "step": 146330
+    },
+    {
+      "epoch": 0.7308047641638993,
+      "grad_norm": 0.08928055316209793,
+      "learning_rate": 8.096072492427846e-06,
+      "loss": 8.611,
+      "step": 146340
+    },
+    {
+      "epoch": 0.7308547029888387,
+      "grad_norm": 0.08695624768733978,
+      "learning_rate": 8.094570577486294e-06,
+      "loss": 8.5939,
+      "step": 146350
+    },
+    {
+      "epoch": 0.7309046418137781,
+      "grad_norm": 0.09484840929508209,
+      "learning_rate": 8.093068662544744e-06,
+      "loss": 8.6148,
+      "step": 146360
+    },
+    {
+      "epoch": 0.7309545806387175,
+      "grad_norm": 0.09645149856805801,
+      "learning_rate": 8.091566747603195e-06,
+      "loss": 8.6244,
+      "step": 146370
+    },
+    {
+      "epoch": 0.731004519463657,
+      "grad_norm": 0.09232309460639954,
+      "learning_rate": 8.090064832661645e-06,
+      "loss": 8.6222,
+      "step": 146380
+    },
+    {
+      "epoch": 0.7310544582885965,
+      "grad_norm": 0.0919228121638298,
+      "learning_rate": 8.088562917720093e-06,
+      "loss": 8.5962,
+      "step": 146390
+    },
+    {
+      "epoch": 0.7311043971135359,
+      "grad_norm": 0.09435618668794632,
+      "learning_rate": 8.087061002778542e-06,
+      "loss": 8.62,
+      "step": 146400
+    },
+    {
+      "epoch": 0.7311543359384753,
+      "grad_norm": 0.09660220891237259,
+      "learning_rate": 8.085559087836992e-06,
+      "loss": 8.5962,
+      "step": 146410
+    },
+    {
+      "epoch": 0.7312042747634148,
+      "grad_norm": 0.0912780910730362,
+      "learning_rate": 8.084057172895442e-06,
+      "loss": 8.6183,
+      "step": 146420
+    },
+    {
+      "epoch": 0.7312542135883543,
+      "grad_norm": 0.09284259378910065,
+      "learning_rate": 8.082555257953892e-06,
+      "loss": 8.61,
+      "step": 146430
+    },
+    {
+      "epoch": 0.7313041524132937,
+      "grad_norm": 0.09274186193943024,
+      "learning_rate": 8.08105334301234e-06,
+      "loss": 8.6186,
+      "step": 146440
+    },
+    {
+      "epoch": 0.7313540912382331,
+      "grad_norm": 0.09855927526950836,
+      "learning_rate": 8.07955142807079e-06,
+      "loss": 8.5975,
+      "step": 146450
+    },
+    {
+      "epoch": 0.7314040300631726,
+      "grad_norm": 0.0931464433670044,
+      "learning_rate": 8.07804951312924e-06,
+      "loss": 8.6029,
+      "step": 146460
+    },
+    {
+      "epoch": 0.7314539688881121,
+      "grad_norm": 0.09151794761419296,
+      "learning_rate": 8.07654759818769e-06,
+      "loss": 8.596,
+      "step": 146470
+    },
+    {
+      "epoch": 0.7315039077130515,
+      "grad_norm": 0.08983919769525528,
+      "learning_rate": 8.07504568324614e-06,
+      "loss": 8.6227,
+      "step": 146480
+    },
+    {
+      "epoch": 0.7315538465379909,
+      "grad_norm": 0.09787040948867798,
+      "learning_rate": 8.07354376830459e-06,
+      "loss": 8.6004,
+      "step": 146490
+    },
+    {
+      "epoch": 0.7316037853629304,
+      "grad_norm": 0.09325719624757767,
+      "learning_rate": 8.072041853363037e-06,
+      "loss": 8.612,
+      "step": 146500
+    },
+    {
+      "epoch": 0.7316537241878699,
+      "grad_norm": 0.09518974274396896,
+      "learning_rate": 8.070539938421487e-06,
+      "loss": 8.6086,
+      "step": 146510
+    },
+    {
+      "epoch": 0.7317036630128093,
+      "grad_norm": 0.09281483292579651,
+      "learning_rate": 8.069038023479937e-06,
+      "loss": 8.6059,
+      "step": 146520
+    },
+    {
+      "epoch": 0.7317536018377487,
+      "grad_norm": 0.09161432087421417,
+      "learning_rate": 8.067536108538387e-06,
+      "loss": 8.6162,
+      "step": 146530
+    },
+    {
+      "epoch": 0.7318035406626882,
+      "grad_norm": 0.08932630717754364,
+      "learning_rate": 8.066034193596837e-06,
+      "loss": 8.6122,
+      "step": 146540
+    },
+    {
+      "epoch": 0.7318534794876277,
+      "grad_norm": 0.10480964183807373,
+      "learning_rate": 8.064532278655284e-06,
+      "loss": 8.5951,
+      "step": 146550
+    },
+    {
+      "epoch": 0.7319034183125671,
+      "grad_norm": 0.09288931638002396,
+      "learning_rate": 8.063030363713734e-06,
+      "loss": 8.6034,
+      "step": 146560
+    },
+    {
+      "epoch": 0.7319533571375065,
+      "grad_norm": 0.09532026201486588,
+      "learning_rate": 8.061528448772185e-06,
+      "loss": 8.6096,
+      "step": 146570
+    },
+    {
+      "epoch": 0.732003295962446,
+      "grad_norm": 0.09342137724161148,
+      "learning_rate": 8.060026533830635e-06,
+      "loss": 8.5875,
+      "step": 146580
+    },
+    {
+      "epoch": 0.7320532347873855,
+      "grad_norm": 0.091585174202919,
+      "learning_rate": 8.058524618889085e-06,
+      "loss": 8.61,
+      "step": 146590
+    },
+    {
+      "epoch": 0.7321031736123249,
+      "grad_norm": 0.08510378003120422,
+      "learning_rate": 8.057022703947532e-06,
+      "loss": 8.6082,
+      "step": 146600
+    },
+    {
+      "epoch": 0.7321531124372643,
+      "grad_norm": 0.0953449234366417,
+      "learning_rate": 8.055520789005982e-06,
+      "loss": 8.6204,
+      "step": 146610
+    },
+    {
+      "epoch": 0.7322030512622038,
+      "grad_norm": 0.0904490277171135,
+      "learning_rate": 8.054018874064432e-06,
+      "loss": 8.6037,
+      "step": 146620
+    },
+    {
+      "epoch": 0.7322529900871433,
+      "grad_norm": 0.09040288627147675,
+      "learning_rate": 8.052516959122882e-06,
+      "loss": 8.6055,
+      "step": 146630
+    },
+    {
+      "epoch": 0.7323029289120827,
+      "grad_norm": 0.09579188376665115,
+      "learning_rate": 8.051015044181332e-06,
+      "loss": 8.6217,
+      "step": 146640
+    },
+    {
+      "epoch": 0.7323528677370221,
+      "grad_norm": 0.09186054021120071,
+      "learning_rate": 8.049513129239781e-06,
+      "loss": 8.5966,
+      "step": 146650
+    },
+    {
+      "epoch": 0.7324028065619616,
+      "grad_norm": 0.08724722266197205,
+      "learning_rate": 8.04801121429823e-06,
+      "loss": 8.6201,
+      "step": 146660
+    },
+    {
+      "epoch": 0.7324527453869011,
+      "grad_norm": 0.09225723147392273,
+      "learning_rate": 8.04650929935668e-06,
+      "loss": 8.604,
+      "step": 146670
+    },
+    {
+      "epoch": 0.7325026842118405,
+      "grad_norm": 0.0851278230547905,
+      "learning_rate": 8.04500738441513e-06,
+      "loss": 8.6148,
+      "step": 146680
+    },
+    {
+      "epoch": 0.7325526230367799,
+      "grad_norm": 0.09377813339233398,
+      "learning_rate": 8.04350546947358e-06,
+      "loss": 8.6004,
+      "step": 146690
+    },
+    {
+      "epoch": 0.7326025618617193,
+      "grad_norm": 0.0915536880493164,
+      "learning_rate": 8.042003554532028e-06,
+      "loss": 8.6146,
+      "step": 146700
+    },
+    {
+      "epoch": 0.7326525006866589,
+      "grad_norm": 0.09648875147104263,
+      "learning_rate": 8.040501639590477e-06,
+      "loss": 8.6067,
+      "step": 146710
+    },
+    {
+      "epoch": 0.7327024395115983,
+      "grad_norm": 0.09112456440925598,
+      "learning_rate": 8.038999724648927e-06,
+      "loss": 8.6042,
+      "step": 146720
+    },
+    {
+      "epoch": 0.7327523783365377,
+      "grad_norm": 0.08740793168544769,
+      "learning_rate": 8.037497809707377e-06,
+      "loss": 8.6129,
+      "step": 146730
+    },
+    {
+      "epoch": 0.7328023171614771,
+      "grad_norm": 0.09064393490552902,
+      "learning_rate": 8.035995894765827e-06,
+      "loss": 8.6168,
+      "step": 146740
+    },
+    {
+      "epoch": 0.7328522559864167,
+      "grad_norm": 0.09343009442090988,
+      "learning_rate": 8.034493979824276e-06,
+      "loss": 8.5964,
+      "step": 146750
+    },
+    {
+      "epoch": 0.7329021948113561,
+      "grad_norm": 0.10053983330726624,
+      "learning_rate": 8.032992064882724e-06,
+      "loss": 8.5884,
+      "step": 146760
+    },
+    {
+      "epoch": 0.7329521336362955,
+      "grad_norm": 0.09632902592420578,
+      "learning_rate": 8.031490149941175e-06,
+      "loss": 8.6121,
+      "step": 146770
+    },
+    {
+      "epoch": 0.7330020724612349,
+      "grad_norm": 0.09040207415819168,
+      "learning_rate": 8.029988234999625e-06,
+      "loss": 8.6016,
+      "step": 146780
+    },
+    {
+      "epoch": 0.7330520112861745,
+      "grad_norm": 0.09205528348684311,
+      "learning_rate": 8.028486320058075e-06,
+      "loss": 8.5859,
+      "step": 146790
+    },
+    {
+      "epoch": 0.7331019501111139,
+      "grad_norm": 0.09228353202342987,
+      "learning_rate": 8.026984405116523e-06,
+      "loss": 8.6206,
+      "step": 146800
+    },
+    {
+      "epoch": 0.7331518889360533,
+      "grad_norm": 0.08614012598991394,
+      "learning_rate": 8.025482490174974e-06,
+      "loss": 8.6226,
+      "step": 146810
+    },
+    {
+      "epoch": 0.7332018277609927,
+      "grad_norm": 0.0924997478723526,
+      "learning_rate": 8.023980575233422e-06,
+      "loss": 8.5959,
+      "step": 146820
+    },
+    {
+      "epoch": 0.7332517665859323,
+      "grad_norm": 0.09448036551475525,
+      "learning_rate": 8.022478660291872e-06,
+      "loss": 8.6089,
+      "step": 146830
+    },
+    {
+      "epoch": 0.7333017054108717,
+      "grad_norm": 0.09214811772108078,
+      "learning_rate": 8.020976745350322e-06,
+      "loss": 8.5967,
+      "step": 146840
+    },
+    {
+      "epoch": 0.7333516442358111,
+      "grad_norm": 0.10002180933952332,
+      "learning_rate": 8.019474830408771e-06,
+      "loss": 8.5945,
+      "step": 146850
+    },
+    {
+      "epoch": 0.7334015830607505,
+      "grad_norm": 0.08861683309078217,
+      "learning_rate": 8.017972915467221e-06,
+      "loss": 8.5851,
+      "step": 146860
+    },
+    {
+      "epoch": 0.7334515218856901,
+      "grad_norm": 0.09217624366283417,
+      "learning_rate": 8.01647100052567e-06,
+      "loss": 8.6059,
+      "step": 146870
+    },
+    {
+      "epoch": 0.7335014607106295,
+      "grad_norm": 0.0931958556175232,
+      "learning_rate": 8.01496908558412e-06,
+      "loss": 8.6036,
+      "step": 146880
+    },
+    {
+      "epoch": 0.7335513995355689,
+      "grad_norm": 0.09389421343803406,
+      "learning_rate": 8.01346717064257e-06,
+      "loss": 8.6066,
+      "step": 146890
+    },
+    {
+      "epoch": 0.7336013383605083,
+      "grad_norm": 0.08797122538089752,
+      "learning_rate": 8.011965255701018e-06,
+      "loss": 8.6272,
+      "step": 146900
+    },
+    {
+      "epoch": 0.7336512771854479,
+      "grad_norm": 0.09919506311416626,
+      "learning_rate": 8.010463340759469e-06,
+      "loss": 8.5877,
+      "step": 146910
+    },
+    {
+      "epoch": 0.7337012160103873,
+      "grad_norm": 0.09523117542266846,
+      "learning_rate": 8.008961425817917e-06,
+      "loss": 8.5809,
+      "step": 146920
+    },
+    {
+      "epoch": 0.7337511548353267,
+      "grad_norm": 0.0903659462928772,
+      "learning_rate": 8.007459510876367e-06,
+      "loss": 8.6084,
+      "step": 146930
+    },
+    {
+      "epoch": 0.7338010936602661,
+      "grad_norm": 0.09395240247249603,
+      "learning_rate": 8.005957595934817e-06,
+      "loss": 8.59,
+      "step": 146940
+    },
+    {
+      "epoch": 0.7338510324852057,
+      "grad_norm": 0.0915970653295517,
+      "learning_rate": 8.004455680993266e-06,
+      "loss": 8.6173,
+      "step": 146950
+    },
+    {
+      "epoch": 0.7339009713101451,
+      "grad_norm": 0.10002275556325912,
+      "learning_rate": 8.002953766051716e-06,
+      "loss": 8.5978,
+      "step": 146960
+    },
+    {
+      "epoch": 0.7339509101350845,
+      "grad_norm": 0.10154590010643005,
+      "learning_rate": 8.001451851110166e-06,
+      "loss": 8.5999,
+      "step": 146970
+    },
+    {
+      "epoch": 0.7340008489600239,
+      "grad_norm": 0.09146125614643097,
+      "learning_rate": 7.999949936168615e-06,
+      "loss": 8.6126,
+      "step": 146980
+    },
+    {
+      "epoch": 0.7340507877849635,
+      "grad_norm": 0.09198717772960663,
+      "learning_rate": 7.998448021227065e-06,
+      "loss": 8.6202,
+      "step": 146990
+    },
+    {
+      "epoch": 0.7341007266099029,
+      "grad_norm": 0.09776240587234497,
+      "learning_rate": 7.996946106285513e-06,
+      "loss": 8.6133,
+      "step": 147000
+    },
+    {
+      "epoch": 0.7341506654348423,
+      "grad_norm": 0.09213253110647202,
+      "learning_rate": 7.995444191343964e-06,
+      "loss": 8.6106,
+      "step": 147010
+    },
+    {
+      "epoch": 0.7342006042597817,
+      "grad_norm": 0.09108315408229828,
+      "learning_rate": 7.993942276402414e-06,
+      "loss": 8.6227,
+      "step": 147020
+    },
+    {
+      "epoch": 0.7342505430847213,
+      "grad_norm": 0.08943723142147064,
+      "learning_rate": 7.992440361460862e-06,
+      "loss": 8.6152,
+      "step": 147030
+    },
+    {
+      "epoch": 0.7343004819096607,
+      "grad_norm": 0.08865174651145935,
+      "learning_rate": 7.990938446519312e-06,
+      "loss": 8.591,
+      "step": 147040
+    },
+    {
+      "epoch": 0.7343504207346001,
+      "grad_norm": 0.0994340255856514,
+      "learning_rate": 7.989436531577763e-06,
+      "loss": 8.6018,
+      "step": 147050
+    },
+    {
+      "epoch": 0.7344003595595395,
+      "grad_norm": 0.09007614850997925,
+      "learning_rate": 7.987934616636211e-06,
+      "loss": 8.5896,
+      "step": 147060
+    },
+    {
+      "epoch": 0.7344502983844791,
+      "grad_norm": 0.08295390009880066,
+      "learning_rate": 7.986432701694661e-06,
+      "loss": 8.6055,
+      "step": 147070
+    },
+    {
+      "epoch": 0.7345002372094185,
+      "grad_norm": 0.09356316924095154,
+      "learning_rate": 7.98493078675311e-06,
+      "loss": 8.6058,
+      "step": 147080
+    },
+    {
+      "epoch": 0.7345501760343579,
+      "grad_norm": 0.09198088943958282,
+      "learning_rate": 7.98342887181156e-06,
+      "loss": 8.6089,
+      "step": 147090
+    },
+    {
+      "epoch": 0.7346001148592973,
+      "grad_norm": 0.09037932008504868,
+      "learning_rate": 7.98192695687001e-06,
+      "loss": 8.6066,
+      "step": 147100
+    },
+    {
+      "epoch": 0.7346500536842369,
+      "grad_norm": 0.0916791781783104,
+      "learning_rate": 7.980425041928459e-06,
+      "loss": 8.6022,
+      "step": 147110
+    },
+    {
+      "epoch": 0.7346999925091763,
+      "grad_norm": 0.09494098275899887,
+      "learning_rate": 7.978923126986909e-06,
+      "loss": 8.5989,
+      "step": 147120
+    },
+    {
+      "epoch": 0.7347499313341157,
+      "grad_norm": 0.0917518213391304,
+      "learning_rate": 7.977421212045359e-06,
+      "loss": 8.5957,
+      "step": 147130
+    },
+    {
+      "epoch": 0.7347998701590551,
+      "grad_norm": 0.08974310755729675,
+      "learning_rate": 7.975919297103807e-06,
+      "loss": 8.6081,
+      "step": 147140
+    },
+    {
+      "epoch": 0.7348498089839947,
+      "grad_norm": 0.09109298884868622,
+      "learning_rate": 7.974417382162258e-06,
+      "loss": 8.5945,
+      "step": 147150
+    },
+    {
+      "epoch": 0.7348997478089341,
+      "grad_norm": 0.08790141344070435,
+      "learning_rate": 7.972915467220706e-06,
+      "loss": 8.6105,
+      "step": 147160
+    },
+    {
+      "epoch": 0.7349496866338735,
+      "grad_norm": 0.09072345495223999,
+      "learning_rate": 7.971413552279156e-06,
+      "loss": 8.6066,
+      "step": 147170
+    },
+    {
+      "epoch": 0.7349996254588129,
+      "grad_norm": 0.08956841379404068,
+      "learning_rate": 7.969911637337606e-06,
+      "loss": 8.6191,
+      "step": 147180
+    },
+    {
+      "epoch": 0.7350495642837525,
+      "grad_norm": 0.09307093918323517,
+      "learning_rate": 7.968409722396055e-06,
+      "loss": 8.6011,
+      "step": 147190
+    },
+    {
+      "epoch": 0.7350995031086919,
+      "grad_norm": 0.09060484170913696,
+      "learning_rate": 7.966907807454505e-06,
+      "loss": 8.6013,
+      "step": 147200
+    },
+    {
+      "epoch": 0.7351494419336313,
+      "grad_norm": 0.08914856612682343,
+      "learning_rate": 7.965405892512954e-06,
+      "loss": 8.6046,
+      "step": 147210
+    },
+    {
+      "epoch": 0.7351993807585707,
+      "grad_norm": 0.0949515700340271,
+      "learning_rate": 7.963903977571404e-06,
+      "loss": 8.6229,
+      "step": 147220
+    },
+    {
+      "epoch": 0.7352493195835103,
+      "grad_norm": 0.09112024307250977,
+      "learning_rate": 7.962402062629854e-06,
+      "loss": 8.6261,
+      "step": 147230
+    },
+    {
+      "epoch": 0.7352992584084497,
+      "grad_norm": 0.0883699506521225,
+      "learning_rate": 7.960900147688302e-06,
+      "loss": 8.612,
+      "step": 147240
+    },
+    {
+      "epoch": 0.7353491972333891,
+      "grad_norm": 0.094344362616539,
+      "learning_rate": 7.959398232746753e-06,
+      "loss": 8.61,
+      "step": 147250
+    },
+    {
+      "epoch": 0.7353991360583285,
+      "grad_norm": 0.08884866535663605,
+      "learning_rate": 7.957896317805201e-06,
+      "loss": 8.6119,
+      "step": 147260
+    },
+    {
+      "epoch": 0.735449074883268,
+      "grad_norm": 0.09071295708417892,
+      "learning_rate": 7.956394402863651e-06,
+      "loss": 8.6011,
+      "step": 147270
+    },
+    {
+      "epoch": 0.7354990137082075,
+      "grad_norm": 0.09404217451810837,
+      "learning_rate": 7.954892487922101e-06,
+      "loss": 8.5964,
+      "step": 147280
+    },
+    {
+      "epoch": 0.7355489525331469,
+      "grad_norm": 0.0958835780620575,
+      "learning_rate": 7.953390572980552e-06,
+      "loss": 8.5959,
+      "step": 147290
+    },
+    {
+      "epoch": 0.7355988913580863,
+      "grad_norm": 0.09104613214731216,
+      "learning_rate": 7.951888658039e-06,
+      "loss": 8.5849,
+      "step": 147300
+    },
+    {
+      "epoch": 0.7356488301830258,
+      "grad_norm": 0.0909975990653038,
+      "learning_rate": 7.950386743097449e-06,
+      "loss": 8.6005,
+      "step": 147310
+    },
+    {
+      "epoch": 0.7356987690079653,
+      "grad_norm": 0.09111376851797104,
+      "learning_rate": 7.948884828155899e-06,
+      "loss": 8.6123,
+      "step": 147320
+    },
+    {
+      "epoch": 0.7357487078329047,
+      "grad_norm": 0.08861295878887177,
+      "learning_rate": 7.947382913214349e-06,
+      "loss": 8.5943,
+      "step": 147330
+    },
+    {
+      "epoch": 0.7357986466578441,
+      "grad_norm": 0.09001336991786957,
+      "learning_rate": 7.9458809982728e-06,
+      "loss": 8.5862,
+      "step": 147340
+    },
+    {
+      "epoch": 0.7358485854827835,
+      "grad_norm": 0.0889565646648407,
+      "learning_rate": 7.944379083331248e-06,
+      "loss": 8.5973,
+      "step": 147350
+    },
+    {
+      "epoch": 0.7358985243077231,
+      "grad_norm": 0.09270651638507843,
+      "learning_rate": 7.942877168389696e-06,
+      "loss": 8.5932,
+      "step": 147360
+    },
+    {
+      "epoch": 0.7359484631326625,
+      "grad_norm": 0.09283141791820526,
+      "learning_rate": 7.941375253448146e-06,
+      "loss": 8.595,
+      "step": 147370
+    },
+    {
+      "epoch": 0.7359984019576019,
+      "grad_norm": 0.09074781090021133,
+      "learning_rate": 7.939873338506596e-06,
+      "loss": 8.6006,
+      "step": 147380
+    },
+    {
+      "epoch": 0.7360483407825413,
+      "grad_norm": 0.09841238707304001,
+      "learning_rate": 7.938371423565047e-06,
+      "loss": 8.6104,
+      "step": 147390
+    },
+    {
+      "epoch": 0.7360982796074809,
+      "grad_norm": 0.09481830149888992,
+      "learning_rate": 7.936869508623495e-06,
+      "loss": 8.6014,
+      "step": 147400
+    },
+    {
+      "epoch": 0.7361482184324203,
+      "grad_norm": 0.08906048536300659,
+      "learning_rate": 7.935367593681944e-06,
+      "loss": 8.6178,
+      "step": 147410
+    },
+    {
+      "epoch": 0.7361981572573597,
+      "grad_norm": 0.08914127200841904,
+      "learning_rate": 7.933865678740394e-06,
+      "loss": 8.6059,
+      "step": 147420
+    },
+    {
+      "epoch": 0.7362480960822991,
+      "grad_norm": 0.09480444341897964,
+      "learning_rate": 7.932363763798844e-06,
+      "loss": 8.5944,
+      "step": 147430
+    },
+    {
+      "epoch": 0.7362980349072387,
+      "grad_norm": 0.09161514043807983,
+      "learning_rate": 7.930861848857294e-06,
+      "loss": 8.6164,
+      "step": 147440
+    },
+    {
+      "epoch": 0.7363479737321781,
+      "grad_norm": 0.09558946639299393,
+      "learning_rate": 7.929359933915744e-06,
+      "loss": 8.5894,
+      "step": 147450
+    },
+    {
+      "epoch": 0.7363979125571175,
+      "grad_norm": 0.09459544718265533,
+      "learning_rate": 7.927858018974191e-06,
+      "loss": 8.598,
+      "step": 147460
+    },
+    {
+      "epoch": 0.7364478513820569,
+      "grad_norm": 0.09031836688518524,
+      "learning_rate": 7.926356104032641e-06,
+      "loss": 8.6046,
+      "step": 147470
+    },
+    {
+      "epoch": 0.7364977902069965,
+      "grad_norm": 0.0889408141374588,
+      "learning_rate": 7.924854189091091e-06,
+      "loss": 8.6006,
+      "step": 147480
+    },
+    {
+      "epoch": 0.7365477290319359,
+      "grad_norm": 0.09583480656147003,
+      "learning_rate": 7.923352274149542e-06,
+      "loss": 8.6201,
+      "step": 147490
+    },
+    {
+      "epoch": 0.7365976678568753,
+      "grad_norm": 0.09124153107404709,
+      "learning_rate": 7.921850359207992e-06,
+      "loss": 8.606,
+      "step": 147500
+    },
+    {
+      "epoch": 0.7366476066818147,
+      "grad_norm": 0.09104292094707489,
+      "learning_rate": 7.920348444266439e-06,
+      "loss": 8.5772,
+      "step": 147510
+    },
+    {
+      "epoch": 0.7366975455067543,
+      "grad_norm": 0.09210095554590225,
+      "learning_rate": 7.918846529324889e-06,
+      "loss": 8.6044,
+      "step": 147520
+    },
+    {
+      "epoch": 0.7367474843316937,
+      "grad_norm": 0.08999733626842499,
+      "learning_rate": 7.917344614383339e-06,
+      "loss": 8.6077,
+      "step": 147530
+    },
+    {
+      "epoch": 0.7367974231566331,
+      "grad_norm": 0.09261172264814377,
+      "learning_rate": 7.91584269944179e-06,
+      "loss": 8.5944,
+      "step": 147540
+    },
+    {
+      "epoch": 0.7368473619815725,
+      "grad_norm": 0.0916711613535881,
+      "learning_rate": 7.91434078450024e-06,
+      "loss": 8.5811,
+      "step": 147550
+    },
+    {
+      "epoch": 0.7368973008065121,
+      "grad_norm": 0.09489434212446213,
+      "learning_rate": 7.912838869558686e-06,
+      "loss": 8.6007,
+      "step": 147560
+    },
+    {
+      "epoch": 0.7369472396314515,
+      "grad_norm": 0.09215337038040161,
+      "learning_rate": 7.911336954617136e-06,
+      "loss": 8.6051,
+      "step": 147570
+    },
+    {
+      "epoch": 0.7369971784563909,
+      "grad_norm": 0.09621989727020264,
+      "learning_rate": 7.909835039675586e-06,
+      "loss": 8.6042,
+      "step": 147580
+    },
+    {
+      "epoch": 0.7370471172813303,
+      "grad_norm": 0.09417767822742462,
+      "learning_rate": 7.908333124734037e-06,
+      "loss": 8.5984,
+      "step": 147590
+    },
+    {
+      "epoch": 0.7370970561062699,
+      "grad_norm": 0.09243947267532349,
+      "learning_rate": 7.906831209792487e-06,
+      "loss": 8.6154,
+      "step": 147600
+    },
+    {
+      "epoch": 0.7371469949312093,
+      "grad_norm": 0.09667365252971649,
+      "learning_rate": 7.905329294850935e-06,
+      "loss": 8.6127,
+      "step": 147610
+    },
+    {
+      "epoch": 0.7371969337561487,
+      "grad_norm": 0.09225791692733765,
+      "learning_rate": 7.903827379909384e-06,
+      "loss": 8.5939,
+      "step": 147620
+    },
+    {
+      "epoch": 0.7372468725810881,
+      "grad_norm": 0.09003566205501556,
+      "learning_rate": 7.902325464967834e-06,
+      "loss": 8.5911,
+      "step": 147630
+    },
+    {
+      "epoch": 0.7372968114060277,
+      "grad_norm": 0.09092370420694351,
+      "learning_rate": 7.900823550026284e-06,
+      "loss": 8.5904,
+      "step": 147640
+    },
+    {
+      "epoch": 0.7373467502309671,
+      "grad_norm": 0.09510780870914459,
+      "learning_rate": 7.899321635084734e-06,
+      "loss": 8.6026,
+      "step": 147650
+    },
+    {
+      "epoch": 0.7373966890559065,
+      "grad_norm": 0.09396243095397949,
+      "learning_rate": 7.897819720143183e-06,
+      "loss": 8.6125,
+      "step": 147660
+    },
+    {
+      "epoch": 0.7374466278808459,
+      "grad_norm": 0.09083011746406555,
+      "learning_rate": 7.896317805201631e-06,
+      "loss": 8.5926,
+      "step": 147670
+    },
+    {
+      "epoch": 0.7374965667057855,
+      "grad_norm": 0.10038017481565475,
+      "learning_rate": 7.894815890260082e-06,
+      "loss": 8.6012,
+      "step": 147680
+    },
+    {
+      "epoch": 0.7375465055307249,
+      "grad_norm": 0.0896848812699318,
+      "learning_rate": 7.893313975318532e-06,
+      "loss": 8.5961,
+      "step": 147690
+    },
+    {
+      "epoch": 0.7375964443556643,
+      "grad_norm": 0.09542018920183182,
+      "learning_rate": 7.891812060376982e-06,
+      "loss": 8.6294,
+      "step": 147700
+    },
+    {
+      "epoch": 0.7376463831806037,
+      "grad_norm": 0.0885300263762474,
+      "learning_rate": 7.89031014543543e-06,
+      "loss": 8.607,
+      "step": 147710
+    },
+    {
+      "epoch": 0.7376963220055432,
+      "grad_norm": 0.0911983922123909,
+      "learning_rate": 7.888808230493879e-06,
+      "loss": 8.5836,
+      "step": 147720
+    },
+    {
+      "epoch": 0.7377462608304827,
+      "grad_norm": 0.08745992183685303,
+      "learning_rate": 7.887306315552329e-06,
+      "loss": 8.5998,
+      "step": 147730
+    },
+    {
+      "epoch": 0.7377961996554221,
+      "grad_norm": 0.09371349215507507,
+      "learning_rate": 7.88580440061078e-06,
+      "loss": 8.5988,
+      "step": 147740
+    },
+    {
+      "epoch": 0.7378461384803615,
+      "grad_norm": 0.09324972331523895,
+      "learning_rate": 7.88430248566923e-06,
+      "loss": 8.5987,
+      "step": 147750
+    },
+    {
+      "epoch": 0.737896077305301,
+      "grad_norm": 0.09328649193048477,
+      "learning_rate": 7.882800570727678e-06,
+      "loss": 8.6026,
+      "step": 147760
+    },
+    {
+      "epoch": 0.7379460161302405,
+      "grad_norm": 0.0888761654496193,
+      "learning_rate": 7.881298655786128e-06,
+      "loss": 8.6012,
+      "step": 147770
+    },
+    {
+      "epoch": 0.7379959549551799,
+      "grad_norm": 0.09246739745140076,
+      "learning_rate": 7.879796740844577e-06,
+      "loss": 8.6072,
+      "step": 147780
+    },
+    {
+      "epoch": 0.7380458937801193,
+      "grad_norm": 0.09637174010276794,
+      "learning_rate": 7.878294825903027e-06,
+      "loss": 8.6062,
+      "step": 147790
+    },
+    {
+      "epoch": 0.7380958326050588,
+      "grad_norm": 0.08997974544763565,
+      "learning_rate": 7.876792910961477e-06,
+      "loss": 8.5919,
+      "step": 147800
+    },
+    {
+      "epoch": 0.7381457714299983,
+      "grad_norm": 0.08694270998239517,
+      "learning_rate": 7.875290996019925e-06,
+      "loss": 8.5971,
+      "step": 147810
+    },
+    {
+      "epoch": 0.7381957102549377,
+      "grad_norm": 0.09957410395145416,
+      "learning_rate": 7.873789081078376e-06,
+      "loss": 8.5931,
+      "step": 147820
+    },
+    {
+      "epoch": 0.7382456490798771,
+      "grad_norm": 0.09661848098039627,
+      "learning_rate": 7.872287166136824e-06,
+      "loss": 8.6017,
+      "step": 147830
+    },
+    {
+      "epoch": 0.7382955879048166,
+      "grad_norm": 0.09177085012197495,
+      "learning_rate": 7.870785251195274e-06,
+      "loss": 8.6021,
+      "step": 147840
+    },
+    {
+      "epoch": 0.7383455267297561,
+      "grad_norm": 0.0871456116437912,
+      "learning_rate": 7.869283336253724e-06,
+      "loss": 8.5894,
+      "step": 147850
+    },
+    {
+      "epoch": 0.7383954655546955,
+      "grad_norm": 0.09130498766899109,
+      "learning_rate": 7.867781421312173e-06,
+      "loss": 8.5875,
+      "step": 147860
+    },
+    {
+      "epoch": 0.7384454043796349,
+      "grad_norm": 0.09617603570222855,
+      "learning_rate": 7.866279506370623e-06,
+      "loss": 8.6145,
+      "step": 147870
+    },
+    {
+      "epoch": 0.7384953432045744,
+      "grad_norm": 0.08777206391096115,
+      "learning_rate": 7.864777591429072e-06,
+      "loss": 8.6098,
+      "step": 147880
+    },
+    {
+      "epoch": 0.7385452820295139,
+      "grad_norm": 0.08969053626060486,
+      "learning_rate": 7.863275676487522e-06,
+      "loss": 8.5935,
+      "step": 147890
+    },
+    {
+      "epoch": 0.7385952208544533,
+      "grad_norm": 0.091539166867733,
+      "learning_rate": 7.861773761545972e-06,
+      "loss": 8.5958,
+      "step": 147900
+    },
+    {
+      "epoch": 0.7386451596793927,
+      "grad_norm": 0.0880018100142479,
+      "learning_rate": 7.86027184660442e-06,
+      "loss": 8.6105,
+      "step": 147910
+    },
+    {
+      "epoch": 0.7386950985043322,
+      "grad_norm": 0.09044984728097916,
+      "learning_rate": 7.85876993166287e-06,
+      "loss": 8.6065,
+      "step": 147920
+    },
+    {
+      "epoch": 0.7387450373292717,
+      "grad_norm": 0.0911531150341034,
+      "learning_rate": 7.85726801672132e-06,
+      "loss": 8.6034,
+      "step": 147930
+    },
+    {
+      "epoch": 0.7387949761542111,
+      "grad_norm": 0.09581080824136734,
+      "learning_rate": 7.85576610177977e-06,
+      "loss": 8.5987,
+      "step": 147940
+    },
+    {
+      "epoch": 0.7388449149791505,
+      "grad_norm": 0.08814167231321335,
+      "learning_rate": 7.85426418683822e-06,
+      "loss": 8.604,
+      "step": 147950
+    },
+    {
+      "epoch": 0.73889485380409,
+      "grad_norm": 0.09530513733625412,
+      "learning_rate": 7.852762271896668e-06,
+      "loss": 8.5862,
+      "step": 147960
+    },
+    {
+      "epoch": 0.7389447926290295,
+      "grad_norm": 0.08761637657880783,
+      "learning_rate": 7.851260356955118e-06,
+      "loss": 8.6251,
+      "step": 147970
+    },
+    {
+      "epoch": 0.7389947314539689,
+      "grad_norm": 0.09182920306921005,
+      "learning_rate": 7.849758442013568e-06,
+      "loss": 8.6002,
+      "step": 147980
+    },
+    {
+      "epoch": 0.7390446702789083,
+      "grad_norm": 0.0961126983165741,
+      "learning_rate": 7.848256527072017e-06,
+      "loss": 8.6096,
+      "step": 147990
+    },
+    {
+      "epoch": 0.7390946091038478,
+      "grad_norm": 0.09371432662010193,
+      "learning_rate": 7.846754612130467e-06,
+      "loss": 8.616,
+      "step": 148000
+    },
+    {
+      "epoch": 0.7391445479287873,
+      "grad_norm": 0.08760569244623184,
+      "learning_rate": 7.845252697188915e-06,
+      "loss": 8.6174,
+      "step": 148010
+    },
+    {
+      "epoch": 0.7391944867537267,
+      "grad_norm": 0.09141170978546143,
+      "learning_rate": 7.843750782247366e-06,
+      "loss": 8.6018,
+      "step": 148020
+    },
+    {
+      "epoch": 0.7392444255786661,
+      "grad_norm": 0.09060084819793701,
+      "learning_rate": 7.842248867305816e-06,
+      "loss": 8.5826,
+      "step": 148030
+    },
+    {
+      "epoch": 0.7392943644036056,
+      "grad_norm": 0.09274831414222717,
+      "learning_rate": 7.840746952364264e-06,
+      "loss": 8.6069,
+      "step": 148040
+    },
+    {
+      "epoch": 0.739344303228545,
+      "grad_norm": 0.09324593096971512,
+      "learning_rate": 7.839245037422714e-06,
+      "loss": 8.6036,
+      "step": 148050
+    },
+    {
+      "epoch": 0.7393942420534845,
+      "grad_norm": 0.09713485836982727,
+      "learning_rate": 7.837743122481163e-06,
+      "loss": 8.6138,
+      "step": 148060
+    },
+    {
+      "epoch": 0.7394441808784239,
+      "grad_norm": 0.08883433789014816,
+      "learning_rate": 7.836241207539613e-06,
+      "loss": 8.6116,
+      "step": 148070
+    },
+    {
+      "epoch": 0.7394941197033634,
+      "grad_norm": 0.09417309612035751,
+      "learning_rate": 7.834739292598063e-06,
+      "loss": 8.5914,
+      "step": 148080
+    },
+    {
+      "epoch": 0.7395440585283029,
+      "grad_norm": 0.09347934275865555,
+      "learning_rate": 7.833237377656513e-06,
+      "loss": 8.6136,
+      "step": 148090
+    },
+    {
+      "epoch": 0.7395939973532423,
+      "grad_norm": 0.09639857709407806,
+      "learning_rate": 7.831735462714962e-06,
+      "loss": 8.5874,
+      "step": 148100
+    },
+    {
+      "epoch": 0.7396439361781817,
+      "grad_norm": 0.09195045381784439,
+      "learning_rate": 7.83023354777341e-06,
+      "loss": 8.5943,
+      "step": 148110
+    },
+    {
+      "epoch": 0.7396938750031212,
+      "grad_norm": 0.09634042531251907,
+      "learning_rate": 7.82873163283186e-06,
+      "loss": 8.6032,
+      "step": 148120
+    },
+    {
+      "epoch": 0.7397438138280606,
+      "grad_norm": 0.08561528474092484,
+      "learning_rate": 7.82722971789031e-06,
+      "loss": 8.5971,
+      "step": 148130
+    },
+    {
+      "epoch": 0.7397937526530001,
+      "grad_norm": 0.08745627850294113,
+      "learning_rate": 7.825727802948761e-06,
+      "loss": 8.5908,
+      "step": 148140
+    },
+    {
+      "epoch": 0.7398436914779395,
+      "grad_norm": 0.09139811247587204,
+      "learning_rate": 7.82422588800721e-06,
+      "loss": 8.5864,
+      "step": 148150
+    },
+    {
+      "epoch": 0.739893630302879,
+      "grad_norm": 0.09640243649482727,
+      "learning_rate": 7.822723973065658e-06,
+      "loss": 8.599,
+      "step": 148160
+    },
+    {
+      "epoch": 0.7399435691278184,
+      "grad_norm": 0.09190690517425537,
+      "learning_rate": 7.821222058124108e-06,
+      "loss": 8.6074,
+      "step": 148170
+    },
+    {
+      "epoch": 0.7399935079527579,
+      "grad_norm": 0.08793210238218307,
+      "learning_rate": 7.819720143182558e-06,
+      "loss": 8.602,
+      "step": 148180
+    },
+    {
+      "epoch": 0.7400434467776973,
+      "grad_norm": 0.09305689483880997,
+      "learning_rate": 7.818218228241008e-06,
+      "loss": 8.5939,
+      "step": 148190
+    },
+    {
+      "epoch": 0.7400933856026368,
+      "grad_norm": 0.09034860879182816,
+      "learning_rate": 7.816716313299457e-06,
+      "loss": 8.6036,
+      "step": 148200
+    },
+    {
+      "epoch": 0.7401433244275762,
+      "grad_norm": 0.09198985993862152,
+      "learning_rate": 7.815214398357905e-06,
+      "loss": 8.5912,
+      "step": 148210
+    },
+    {
+      "epoch": 0.7401932632525157,
+      "grad_norm": 0.0869131088256836,
+      "learning_rate": 7.813712483416356e-06,
+      "loss": 8.6085,
+      "step": 148220
+    },
+    {
+      "epoch": 0.7402432020774551,
+      "grad_norm": 0.09751661121845245,
+      "learning_rate": 7.812210568474806e-06,
+      "loss": 8.6094,
+      "step": 148230
+    },
+    {
+      "epoch": 0.7402931409023946,
+      "grad_norm": 0.09016689658164978,
+      "learning_rate": 7.810708653533256e-06,
+      "loss": 8.6164,
+      "step": 148240
+    },
+    {
+      "epoch": 0.740343079727334,
+      "grad_norm": 0.09467813372612,
+      "learning_rate": 7.809206738591706e-06,
+      "loss": 8.616,
+      "step": 148250
+    },
+    {
+      "epoch": 0.7403930185522735,
+      "grad_norm": 0.09038371592760086,
+      "learning_rate": 7.807704823650153e-06,
+      "loss": 8.6001,
+      "step": 148260
+    },
+    {
+      "epoch": 0.7404429573772129,
+      "grad_norm": 0.09057788550853729,
+      "learning_rate": 7.806202908708603e-06,
+      "loss": 8.6125,
+      "step": 148270
+    },
+    {
+      "epoch": 0.7404928962021524,
+      "grad_norm": 0.09068594127893448,
+      "learning_rate": 7.804700993767053e-06,
+      "loss": 8.5882,
+      "step": 148280
+    },
+    {
+      "epoch": 0.7405428350270918,
+      "grad_norm": 0.09720471501350403,
+      "learning_rate": 7.803199078825503e-06,
+      "loss": 8.6165,
+      "step": 148290
+    },
+    {
+      "epoch": 0.7405927738520313,
+      "grad_norm": 0.09376747906208038,
+      "learning_rate": 7.801697163883954e-06,
+      "loss": 8.5862,
+      "step": 148300
+    },
+    {
+      "epoch": 0.7406427126769707,
+      "grad_norm": 0.09301353991031647,
+      "learning_rate": 7.8001952489424e-06,
+      "loss": 8.5996,
+      "step": 148310
+    },
+    {
+      "epoch": 0.7406926515019102,
+      "grad_norm": 0.09005246311426163,
+      "learning_rate": 7.79869333400085e-06,
+      "loss": 8.6072,
+      "step": 148320
+    },
+    {
+      "epoch": 0.7407425903268496,
+      "grad_norm": 0.09260305762290955,
+      "learning_rate": 7.7971914190593e-06,
+      "loss": 8.59,
+      "step": 148330
+    },
+    {
+      "epoch": 0.7407925291517891,
+      "grad_norm": 0.09277404844760895,
+      "learning_rate": 7.795689504117751e-06,
+      "loss": 8.618,
+      "step": 148340
+    },
+    {
+      "epoch": 0.7408424679767285,
+      "grad_norm": 0.09078875184059143,
+      "learning_rate": 7.794187589176201e-06,
+      "loss": 8.612,
+      "step": 148350
+    },
+    {
+      "epoch": 0.7408924068016679,
+      "grad_norm": 0.0957048162817955,
+      "learning_rate": 7.792685674234648e-06,
+      "loss": 8.5944,
+      "step": 148360
+    },
+    {
+      "epoch": 0.7409423456266074,
+      "grad_norm": 0.08615356683731079,
+      "learning_rate": 7.791183759293098e-06,
+      "loss": 8.6149,
+      "step": 148370
+    },
+    {
+      "epoch": 0.7409922844515469,
+      "grad_norm": 0.0927518680691719,
+      "learning_rate": 7.789681844351548e-06,
+      "loss": 8.5918,
+      "step": 148380
+    },
+    {
+      "epoch": 0.7410422232764863,
+      "grad_norm": 0.09268731623888016,
+      "learning_rate": 7.788179929409998e-06,
+      "loss": 8.5982,
+      "step": 148390
+    },
+    {
+      "epoch": 0.7410921621014257,
+      "grad_norm": 0.09085573256015778,
+      "learning_rate": 7.786678014468449e-06,
+      "loss": 8.6049,
+      "step": 148400
+    },
+    {
+      "epoch": 0.7411421009263652,
+      "grad_norm": 0.09364666044712067,
+      "learning_rate": 7.785176099526897e-06,
+      "loss": 8.5953,
+      "step": 148410
+    },
+    {
+      "epoch": 0.7411920397513047,
+      "grad_norm": 0.09044606983661652,
+      "learning_rate": 7.783674184585346e-06,
+      "loss": 8.5928,
+      "step": 148420
+    },
+    {
+      "epoch": 0.7412419785762441,
+      "grad_norm": 0.08722547441720963,
+      "learning_rate": 7.782172269643796e-06,
+      "loss": 8.5983,
+      "step": 148430
+    },
+    {
+      "epoch": 0.7412919174011835,
+      "grad_norm": 0.0928499698638916,
+      "learning_rate": 7.780670354702246e-06,
+      "loss": 8.615,
+      "step": 148440
+    },
+    {
+      "epoch": 0.741341856226123,
+      "grad_norm": 0.09413442015647888,
+      "learning_rate": 7.779168439760696e-06,
+      "loss": 8.6119,
+      "step": 148450
+    },
+    {
+      "epoch": 0.7413917950510625,
+      "grad_norm": 0.09118451923131943,
+      "learning_rate": 7.777666524819145e-06,
+      "loss": 8.6034,
+      "step": 148460
+    },
+    {
+      "epoch": 0.7414417338760019,
+      "grad_norm": 0.08917475491762161,
+      "learning_rate": 7.776164609877593e-06,
+      "loss": 8.5968,
+      "step": 148470
+    },
+    {
+      "epoch": 0.7414916727009413,
+      "grad_norm": 0.09661485254764557,
+      "learning_rate": 7.774662694936043e-06,
+      "loss": 8.589,
+      "step": 148480
+    },
+    {
+      "epoch": 0.7415416115258808,
+      "grad_norm": 0.0894039124250412,
+      "learning_rate": 7.773160779994493e-06,
+      "loss": 8.5998,
+      "step": 148490
+    },
+    {
+      "epoch": 0.7415915503508202,
+      "grad_norm": 0.09486133605241776,
+      "learning_rate": 7.771658865052944e-06,
+      "loss": 8.6045,
+      "step": 148500
+    },
+    {
+      "epoch": 0.7416414891757597,
+      "grad_norm": 0.09382303059101105,
+      "learning_rate": 7.770156950111392e-06,
+      "loss": 8.5874,
+      "step": 148510
+    },
+    {
+      "epoch": 0.7416914280006991,
+      "grad_norm": 0.0903899222612381,
+      "learning_rate": 7.76865503516984e-06,
+      "loss": 8.5879,
+      "step": 148520
+    },
+    {
+      "epoch": 0.7417413668256386,
+      "grad_norm": 0.08889366686344147,
+      "learning_rate": 7.76715312022829e-06,
+      "loss": 8.5956,
+      "step": 148530
+    },
+    {
+      "epoch": 0.741791305650578,
+      "grad_norm": 0.09097573906183243,
+      "learning_rate": 7.765651205286741e-06,
+      "loss": 8.5917,
+      "step": 148540
+    },
+    {
+      "epoch": 0.7418412444755175,
+      "grad_norm": 0.09453503787517548,
+      "learning_rate": 7.764149290345191e-06,
+      "loss": 8.5922,
+      "step": 148550
+    },
+    {
+      "epoch": 0.7418911833004569,
+      "grad_norm": 0.09525911509990692,
+      "learning_rate": 7.76264737540364e-06,
+      "loss": 8.6012,
+      "step": 148560
+    },
+    {
+      "epoch": 0.7419411221253964,
+      "grad_norm": 0.08901818841695786,
+      "learning_rate": 7.76114546046209e-06,
+      "loss": 8.6012,
+      "step": 148570
+    },
+    {
+      "epoch": 0.7419910609503358,
+      "grad_norm": 0.09080463647842407,
+      "learning_rate": 7.759643545520538e-06,
+      "loss": 8.6011,
+      "step": 148580
+    },
+    {
+      "epoch": 0.7420409997752753,
+      "grad_norm": 0.08285506069660187,
+      "learning_rate": 7.758141630578988e-06,
+      "loss": 8.5975,
+      "step": 148590
+    },
+    {
+      "epoch": 0.7420909386002147,
+      "grad_norm": 0.09270590543746948,
+      "learning_rate": 7.756639715637439e-06,
+      "loss": 8.6014,
+      "step": 148600
+    },
+    {
+      "epoch": 0.7421408774251542,
+      "grad_norm": 0.09853670746088028,
+      "learning_rate": 7.755137800695887e-06,
+      "loss": 8.5918,
+      "step": 148610
+    },
+    {
+      "epoch": 0.7421908162500936,
+      "grad_norm": 0.09975389391183853,
+      "learning_rate": 7.753635885754337e-06,
+      "loss": 8.601,
+      "step": 148620
+    },
+    {
+      "epoch": 0.7422407550750331,
+      "grad_norm": 0.09311430156230927,
+      "learning_rate": 7.752133970812786e-06,
+      "loss": 8.5938,
+      "step": 148630
+    },
+    {
+      "epoch": 0.7422906938999725,
+      "grad_norm": 0.09312795102596283,
+      "learning_rate": 7.750632055871236e-06,
+      "loss": 8.5892,
+      "step": 148640
+    },
+    {
+      "epoch": 0.742340632724912,
+      "grad_norm": 0.09401096403598785,
+      "learning_rate": 7.749130140929686e-06,
+      "loss": 8.6029,
+      "step": 148650
+    },
+    {
+      "epoch": 0.7423905715498514,
+      "grad_norm": 0.09024547040462494,
+      "learning_rate": 7.747628225988135e-06,
+      "loss": 8.5963,
+      "step": 148660
+    },
+    {
+      "epoch": 0.7424405103747909,
+      "grad_norm": 0.09483632445335388,
+      "learning_rate": 7.746126311046585e-06,
+      "loss": 8.611,
+      "step": 148670
+    },
+    {
+      "epoch": 0.7424904491997303,
+      "grad_norm": 0.08980057388544083,
+      "learning_rate": 7.744624396105033e-06,
+      "loss": 8.5968,
+      "step": 148680
+    },
+    {
+      "epoch": 0.7425403880246698,
+      "grad_norm": 0.09561526775360107,
+      "learning_rate": 7.743122481163483e-06,
+      "loss": 8.6023,
+      "step": 148690
+    },
+    {
+      "epoch": 0.7425903268496092,
+      "grad_norm": 0.09579294919967651,
+      "learning_rate": 7.741620566221934e-06,
+      "loss": 8.5962,
+      "step": 148700
+    },
+    {
+      "epoch": 0.7426402656745487,
+      "grad_norm": 0.09053392708301544,
+      "learning_rate": 7.740118651280382e-06,
+      "loss": 8.5915,
+      "step": 148710
+    },
+    {
+      "epoch": 0.7426902044994881,
+      "grad_norm": 0.08874721825122833,
+      "learning_rate": 7.738616736338832e-06,
+      "loss": 8.6017,
+      "step": 148720
+    },
+    {
+      "epoch": 0.7427401433244276,
+      "grad_norm": 0.09630593657493591,
+      "learning_rate": 7.737114821397282e-06,
+      "loss": 8.6011,
+      "step": 148730
+    },
+    {
+      "epoch": 0.742790082149367,
+      "grad_norm": 0.0932571217417717,
+      "learning_rate": 7.735612906455731e-06,
+      "loss": 8.6088,
+      "step": 148740
+    },
+    {
+      "epoch": 0.7428400209743065,
+      "grad_norm": 0.09587745368480682,
+      "learning_rate": 7.734110991514181e-06,
+      "loss": 8.6054,
+      "step": 148750
+    },
+    {
+      "epoch": 0.7428899597992459,
+      "grad_norm": 0.08768852055072784,
+      "learning_rate": 7.73260907657263e-06,
+      "loss": 8.5966,
+      "step": 148760
+    },
+    {
+      "epoch": 0.7429398986241854,
+      "grad_norm": 0.08625376224517822,
+      "learning_rate": 7.73110716163108e-06,
+      "loss": 8.5867,
+      "step": 148770
+    },
+    {
+      "epoch": 0.7429898374491248,
+      "grad_norm": 0.09649079293012619,
+      "learning_rate": 7.72960524668953e-06,
+      "loss": 8.6001,
+      "step": 148780
+    },
+    {
+      "epoch": 0.7430397762740643,
+      "grad_norm": 0.0902811661362648,
+      "learning_rate": 7.728103331747978e-06,
+      "loss": 8.5862,
+      "step": 148790
+    },
+    {
+      "epoch": 0.7430897150990037,
+      "grad_norm": 0.08683760464191437,
+      "learning_rate": 7.726601416806429e-06,
+      "loss": 8.5762,
+      "step": 148800
+    },
+    {
+      "epoch": 0.7431396539239432,
+      "grad_norm": 0.09725631028413773,
+      "learning_rate": 7.725099501864877e-06,
+      "loss": 8.6059,
+      "step": 148810
+    },
+    {
+      "epoch": 0.7431895927488826,
+      "grad_norm": 0.09072460979223251,
+      "learning_rate": 7.723597586923327e-06,
+      "loss": 8.5995,
+      "step": 148820
+    },
+    {
+      "epoch": 0.743239531573822,
+      "grad_norm": 0.0900563895702362,
+      "learning_rate": 7.722095671981777e-06,
+      "loss": 8.6039,
+      "step": 148830
+    },
+    {
+      "epoch": 0.7432894703987615,
+      "grad_norm": 0.08766742050647736,
+      "learning_rate": 7.720593757040226e-06,
+      "loss": 8.5947,
+      "step": 148840
+    },
+    {
+      "epoch": 0.743339409223701,
+      "grad_norm": 0.09723313897848129,
+      "learning_rate": 7.719091842098676e-06,
+      "loss": 8.6027,
+      "step": 148850
+    },
+    {
+      "epoch": 0.7433893480486404,
+      "grad_norm": 0.09286182373762131,
+      "learning_rate": 7.717589927157125e-06,
+      "loss": 8.5978,
+      "step": 148860
+    },
+    {
+      "epoch": 0.7434392868735799,
+      "grad_norm": 0.09512756764888763,
+      "learning_rate": 7.716088012215575e-06,
+      "loss": 8.5881,
+      "step": 148870
+    },
+    {
+      "epoch": 0.7434892256985193,
+      "grad_norm": 0.09215110540390015,
+      "learning_rate": 7.714586097274025e-06,
+      "loss": 8.5987,
+      "step": 148880
+    },
+    {
+      "epoch": 0.7435391645234588,
+      "grad_norm": 0.09654014557600021,
+      "learning_rate": 7.713084182332475e-06,
+      "loss": 8.5833,
+      "step": 148890
+    },
+    {
+      "epoch": 0.7435891033483982,
+      "grad_norm": 0.09393315017223358,
+      "learning_rate": 7.711582267390924e-06,
+      "loss": 8.6084,
+      "step": 148900
+    },
+    {
+      "epoch": 0.7436390421733376,
+      "grad_norm": 0.09271156042814255,
+      "learning_rate": 7.710080352449372e-06,
+      "loss": 8.5837,
+      "step": 148910
+    },
+    {
+      "epoch": 0.7436889809982771,
+      "grad_norm": 0.08688397705554962,
+      "learning_rate": 7.708578437507822e-06,
+      "loss": 8.5953,
+      "step": 148920
+    },
+    {
+      "epoch": 0.7437389198232166,
+      "grad_norm": 0.08975444734096527,
+      "learning_rate": 7.707076522566272e-06,
+      "loss": 8.585,
+      "step": 148930
+    },
+    {
+      "epoch": 0.743788858648156,
+      "grad_norm": 0.09515532851219177,
+      "learning_rate": 7.705574607624723e-06,
+      "loss": 8.5907,
+      "step": 148940
+    },
+    {
+      "epoch": 0.7438387974730954,
+      "grad_norm": 0.08886223286390305,
+      "learning_rate": 7.704072692683171e-06,
+      "loss": 8.5786,
+      "step": 148950
+    },
+    {
+      "epoch": 0.7438887362980349,
+      "grad_norm": 0.09450791031122208,
+      "learning_rate": 7.70257077774162e-06,
+      "loss": 8.5943,
+      "step": 148960
+    },
+    {
+      "epoch": 0.7439386751229744,
+      "grad_norm": 0.08783065527677536,
+      "learning_rate": 7.70106886280007e-06,
+      "loss": 8.6139,
+      "step": 148970
+    },
+    {
+      "epoch": 0.7439886139479138,
+      "grad_norm": 0.0896720141172409,
+      "learning_rate": 7.69956694785852e-06,
+      "loss": 8.6009,
+      "step": 148980
+    },
+    {
+      "epoch": 0.7440385527728532,
+      "grad_norm": 0.08917935937643051,
+      "learning_rate": 7.69806503291697e-06,
+      "loss": 8.5988,
+      "step": 148990
+    },
+    {
+      "epoch": 0.7440884915977927,
+      "grad_norm": 0.09412242472171783,
+      "learning_rate": 7.696563117975419e-06,
+      "loss": 8.5932,
+      "step": 149000
+    },
+    {
+      "epoch": 0.7441384304227322,
+      "grad_norm": 0.09121372550725937,
+      "learning_rate": 7.695061203033867e-06,
+      "loss": 8.5708,
+      "step": 149010
+    },
+    {
+      "epoch": 0.7441883692476716,
+      "grad_norm": 0.09384799003601074,
+      "learning_rate": 7.693559288092317e-06,
+      "loss": 8.6003,
+      "step": 149020
+    },
+    {
+      "epoch": 0.744238308072611,
+      "grad_norm": 0.088929682970047,
+      "learning_rate": 7.692057373150767e-06,
+      "loss": 8.5807,
+      "step": 149030
+    },
+    {
+      "epoch": 0.7442882468975505,
+      "grad_norm": 0.09719398617744446,
+      "learning_rate": 7.690555458209218e-06,
+      "loss": 8.5885,
+      "step": 149040
+    },
+    {
+      "epoch": 0.74433818572249,
+      "grad_norm": 0.09361552447080612,
+      "learning_rate": 7.689053543267668e-06,
+      "loss": 8.5899,
+      "step": 149050
+    },
+    {
+      "epoch": 0.7443881245474294,
+      "grad_norm": 0.08958575874567032,
+      "learning_rate": 7.687551628326115e-06,
+      "loss": 8.6113,
+      "step": 149060
+    },
+    {
+      "epoch": 0.7444380633723688,
+      "grad_norm": 0.09068027883768082,
+      "learning_rate": 7.686049713384565e-06,
+      "loss": 8.6056,
+      "step": 149070
+    },
+    {
+      "epoch": 0.7444880021973083,
+      "grad_norm": 0.09241528064012527,
+      "learning_rate": 7.684547798443015e-06,
+      "loss": 8.5827,
+      "step": 149080
+    },
+    {
+      "epoch": 0.7445379410222478,
+      "grad_norm": 0.09474475681781769,
+      "learning_rate": 7.683045883501465e-06,
+      "loss": 8.6088,
+      "step": 149090
+    },
+    {
+      "epoch": 0.7445878798471872,
+      "grad_norm": 0.09101393073797226,
+      "learning_rate": 7.681543968559915e-06,
+      "loss": 8.5973,
+      "step": 149100
+    },
+    {
+      "epoch": 0.7446378186721266,
+      "grad_norm": 0.09199006855487823,
+      "learning_rate": 7.680042053618362e-06,
+      "loss": 8.6017,
+      "step": 149110
+    },
+    {
+      "epoch": 0.7446877574970661,
+      "grad_norm": 0.09310968220233917,
+      "learning_rate": 7.678540138676812e-06,
+      "loss": 8.5962,
+      "step": 149120
+    },
+    {
+      "epoch": 0.7447376963220056,
+      "grad_norm": 0.08850269764661789,
+      "learning_rate": 7.677038223735262e-06,
+      "loss": 8.5832,
+      "step": 149130
+    },
+    {
+      "epoch": 0.744787635146945,
+      "grad_norm": 0.09304803609848022,
+      "learning_rate": 7.675536308793713e-06,
+      "loss": 8.5942,
+      "step": 149140
+    },
+    {
+      "epoch": 0.7448375739718844,
+      "grad_norm": 0.0940072312951088,
+      "learning_rate": 7.674034393852163e-06,
+      "loss": 8.6149,
+      "step": 149150
+    },
+    {
+      "epoch": 0.7448875127968239,
+      "grad_norm": 0.09144038707017899,
+      "learning_rate": 7.672532478910611e-06,
+      "loss": 8.5879,
+      "step": 149160
+    },
+    {
+      "epoch": 0.7449374516217634,
+      "grad_norm": 0.09090874344110489,
+      "learning_rate": 7.67103056396906e-06,
+      "loss": 8.595,
+      "step": 149170
+    },
+    {
+      "epoch": 0.7449873904467028,
+      "grad_norm": 0.10227791219949722,
+      "learning_rate": 7.66952864902751e-06,
+      "loss": 8.611,
+      "step": 149180
+    },
+    {
+      "epoch": 0.7450373292716422,
+      "grad_norm": 0.09328952431678772,
+      "learning_rate": 7.66802673408596e-06,
+      "loss": 8.574,
+      "step": 149190
+    },
+    {
+      "epoch": 0.7450872680965817,
+      "grad_norm": 0.09365438669919968,
+      "learning_rate": 7.66652481914441e-06,
+      "loss": 8.6266,
+      "step": 149200
+    },
+    {
+      "epoch": 0.7451372069215212,
+      "grad_norm": 0.0949469655752182,
+      "learning_rate": 7.665022904202859e-06,
+      "loss": 8.6018,
+      "step": 149210
+    },
+    {
+      "epoch": 0.7451871457464606,
+      "grad_norm": 0.0964326560497284,
+      "learning_rate": 7.663520989261307e-06,
+      "loss": 8.5941,
+      "step": 149220
+    },
+    {
+      "epoch": 0.7452370845714,
+      "grad_norm": 0.091887928545475,
+      "learning_rate": 7.662019074319758e-06,
+      "loss": 8.5932,
+      "step": 149230
+    },
+    {
+      "epoch": 0.7452870233963395,
+      "grad_norm": 0.09236712753772736,
+      "learning_rate": 7.660517159378208e-06,
+      "loss": 8.5758,
+      "step": 149240
+    },
+    {
+      "epoch": 0.745336962221279,
+      "grad_norm": 0.08846684545278549,
+      "learning_rate": 7.659015244436658e-06,
+      "loss": 8.5856,
+      "step": 149250
+    },
+    {
+      "epoch": 0.7453869010462184,
+      "grad_norm": 0.10501750558614731,
+      "learning_rate": 7.657513329495106e-06,
+      "loss": 8.5935,
+      "step": 149260
+    },
+    {
+      "epoch": 0.7454368398711578,
+      "grad_norm": 0.09326668828725815,
+      "learning_rate": 7.656011414553555e-06,
+      "loss": 8.598,
+      "step": 149270
+    },
+    {
+      "epoch": 0.7454867786960973,
+      "grad_norm": 0.09294133633375168,
+      "learning_rate": 7.654509499612005e-06,
+      "loss": 8.5952,
+      "step": 149280
+    },
+    {
+      "epoch": 0.7455367175210368,
+      "grad_norm": 0.09185370057821274,
+      "learning_rate": 7.653007584670455e-06,
+      "loss": 8.6152,
+      "step": 149290
+    },
+    {
+      "epoch": 0.7455866563459762,
+      "grad_norm": 0.09958349168300629,
+      "learning_rate": 7.651505669728905e-06,
+      "loss": 8.5855,
+      "step": 149300
+    },
+    {
+      "epoch": 0.7456365951709156,
+      "grad_norm": 0.1002863273024559,
+      "learning_rate": 7.650003754787356e-06,
+      "loss": 8.5972,
+      "step": 149310
+    },
+    {
+      "epoch": 0.745686533995855,
+      "grad_norm": 0.09235794842243195,
+      "learning_rate": 7.648501839845802e-06,
+      "loss": 8.5912,
+      "step": 149320
+    },
+    {
+      "epoch": 0.7457364728207945,
+      "grad_norm": 0.09120310097932816,
+      "learning_rate": 7.646999924904253e-06,
+      "loss": 8.5867,
+      "step": 149330
+    },
+    {
+      "epoch": 0.745786411645734,
+      "grad_norm": 0.09124782681465149,
+      "learning_rate": 7.645498009962703e-06,
+      "loss": 8.5841,
+      "step": 149340
+    },
+    {
+      "epoch": 0.7458363504706734,
+      "grad_norm": 0.09484837204217911,
+      "learning_rate": 7.643996095021153e-06,
+      "loss": 8.5788,
+      "step": 149350
+    },
+    {
+      "epoch": 0.7458862892956128,
+      "grad_norm": 0.08904207497835159,
+      "learning_rate": 7.642494180079603e-06,
+      "loss": 8.5973,
+      "step": 149360
+    },
+    {
+      "epoch": 0.7459362281205523,
+      "grad_norm": 0.08774899691343307,
+      "learning_rate": 7.64099226513805e-06,
+      "loss": 8.5849,
+      "step": 149370
+    },
+    {
+      "epoch": 0.7459861669454918,
+      "grad_norm": 0.09321718662977219,
+      "learning_rate": 7.6394903501965e-06,
+      "loss": 8.598,
+      "step": 149380
+    },
+    {
+      "epoch": 0.7460361057704312,
+      "grad_norm": 0.09559307992458344,
+      "learning_rate": 7.63798843525495e-06,
+      "loss": 8.6053,
+      "step": 149390
+    },
+    {
+      "epoch": 0.7460860445953706,
+      "grad_norm": 0.09065421670675278,
+      "learning_rate": 7.6364865203134e-06,
+      "loss": 8.5822,
+      "step": 149400
+    },
+    {
+      "epoch": 0.7461359834203101,
+      "grad_norm": 0.08813871443271637,
+      "learning_rate": 7.63498460537185e-06,
+      "loss": 8.6158,
+      "step": 149410
+    },
+    {
+      "epoch": 0.7461859222452496,
+      "grad_norm": 0.09453373402357101,
+      "learning_rate": 7.633482690430299e-06,
+      "loss": 8.595,
+      "step": 149420
+    },
+    {
+      "epoch": 0.746235861070189,
+      "grad_norm": 0.09520356357097626,
+      "learning_rate": 7.631980775488748e-06,
+      "loss": 8.6,
+      "step": 149430
+    },
+    {
+      "epoch": 0.7462857998951284,
+      "grad_norm": 0.09131111204624176,
+      "learning_rate": 7.630478860547198e-06,
+      "loss": 8.6118,
+      "step": 149440
+    },
+    {
+      "epoch": 0.7463357387200679,
+      "grad_norm": 0.09047097712755203,
+      "learning_rate": 7.628976945605647e-06,
+      "loss": 8.598,
+      "step": 149450
+    },
+    {
+      "epoch": 0.7463856775450074,
+      "grad_norm": 0.09119818359613419,
+      "learning_rate": 7.627475030664097e-06,
+      "loss": 8.6094,
+      "step": 149460
+    },
+    {
+      "epoch": 0.7464356163699468,
+      "grad_norm": 0.0884728953242302,
+      "learning_rate": 7.625973115722547e-06,
+      "loss": 8.6118,
+      "step": 149470
+    },
+    {
+      "epoch": 0.7464855551948862,
+      "grad_norm": 0.0959334447979927,
+      "learning_rate": 7.624471200780995e-06,
+      "loss": 8.5847,
+      "step": 149480
+    },
+    {
+      "epoch": 0.7465354940198257,
+      "grad_norm": 0.08748113363981247,
+      "learning_rate": 7.622969285839445e-06,
+      "loss": 8.5978,
+      "step": 149490
+    },
+    {
+      "epoch": 0.7465854328447652,
+      "grad_norm": 0.09139006584882736,
+      "learning_rate": 7.6214673708978945e-06,
+      "loss": 8.5974,
+      "step": 149500
+    },
+    {
+      "epoch": 0.7466353716697046,
+      "grad_norm": 0.089291051030159,
+      "learning_rate": 7.619965455956345e-06,
+      "loss": 8.5852,
+      "step": 149510
+    },
+    {
+      "epoch": 0.746685310494644,
+      "grad_norm": 0.0897984728217125,
+      "learning_rate": 7.618463541014795e-06,
+      "loss": 8.5904,
+      "step": 149520
+    },
+    {
+      "epoch": 0.7467352493195835,
+      "grad_norm": 0.09387893229722977,
+      "learning_rate": 7.6169616260732425e-06,
+      "loss": 8.5993,
+      "step": 149530
+    },
+    {
+      "epoch": 0.746785188144523,
+      "grad_norm": 0.09599009901285172,
+      "learning_rate": 7.615459711131693e-06,
+      "loss": 8.5884,
+      "step": 149540
+    },
+    {
+      "epoch": 0.7468351269694624,
+      "grad_norm": 0.10485220700502396,
+      "learning_rate": 7.613957796190142e-06,
+      "loss": 8.5845,
+      "step": 149550
+    },
+    {
+      "epoch": 0.7468850657944018,
+      "grad_norm": 0.09003337472677231,
+      "learning_rate": 7.612455881248592e-06,
+      "loss": 8.5792,
+      "step": 149560
+    },
+    {
+      "epoch": 0.7469350046193413,
+      "grad_norm": 0.08689581602811813,
+      "learning_rate": 7.610953966307042e-06,
+      "loss": 8.5919,
+      "step": 149570
+    },
+    {
+      "epoch": 0.7469849434442808,
+      "grad_norm": 0.09046222269535065,
+      "learning_rate": 7.609452051365492e-06,
+      "loss": 8.6032,
+      "step": 149580
+    },
+    {
+      "epoch": 0.7470348822692202,
+      "grad_norm": 0.0931020975112915,
+      "learning_rate": 7.60795013642394e-06,
+      "loss": 8.5898,
+      "step": 149590
+    },
+    {
+      "epoch": 0.7470848210941596,
+      "grad_norm": 0.09115353971719742,
+      "learning_rate": 7.6064482214823895e-06,
+      "loss": 8.5903,
+      "step": 149600
+    },
+    {
+      "epoch": 0.7471347599190991,
+      "grad_norm": 0.095846027135849,
+      "learning_rate": 7.60494630654084e-06,
+      "loss": 8.5989,
+      "step": 149610
+    },
+    {
+      "epoch": 0.7471846987440386,
+      "grad_norm": 0.0902191624045372,
+      "learning_rate": 7.60344439159929e-06,
+      "loss": 8.5844,
+      "step": 149620
+    },
+    {
+      "epoch": 0.747234637568978,
+      "grad_norm": 0.08753103017807007,
+      "learning_rate": 7.601942476657739e-06,
+      "loss": 8.5876,
+      "step": 149630
+    },
+    {
+      "epoch": 0.7472845763939174,
+      "grad_norm": 0.09205235540866852,
+      "learning_rate": 7.600440561716188e-06,
+      "loss": 8.6017,
+      "step": 149640
+    },
+    {
+      "epoch": 0.7473345152188569,
+      "grad_norm": 0.08869083225727081,
+      "learning_rate": 7.598938646774637e-06,
+      "loss": 8.5879,
+      "step": 149650
+    },
+    {
+      "epoch": 0.7473844540437964,
+      "grad_norm": 0.09355755895376205,
+      "learning_rate": 7.597436731833087e-06,
+      "loss": 8.6161,
+      "step": 149660
+    },
+    {
+      "epoch": 0.7474343928687358,
+      "grad_norm": 0.09638141095638275,
+      "learning_rate": 7.595934816891537e-06,
+      "loss": 8.5874,
+      "step": 149670
+    },
+    {
+      "epoch": 0.7474843316936752,
+      "grad_norm": 0.09386901557445526,
+      "learning_rate": 7.594432901949987e-06,
+      "loss": 8.596,
+      "step": 149680
+    },
+    {
+      "epoch": 0.7475342705186147,
+      "grad_norm": 0.08905579149723053,
+      "learning_rate": 7.592930987008435e-06,
+      "loss": 8.5965,
+      "step": 149690
+    },
+    {
+      "epoch": 0.7475842093435542,
+      "grad_norm": 0.09240230172872543,
+      "learning_rate": 7.5914290720668845e-06,
+      "loss": 8.5916,
+      "step": 149700
+    },
+    {
+      "epoch": 0.7476341481684936,
+      "grad_norm": 0.08725178986787796,
+      "learning_rate": 7.589927157125335e-06,
+      "loss": 8.61,
+      "step": 149710
+    },
+    {
+      "epoch": 0.747684086993433,
+      "grad_norm": 0.09127280861139297,
+      "learning_rate": 7.588425242183785e-06,
+      "loss": 8.5987,
+      "step": 149720
+    },
+    {
+      "epoch": 0.7477340258183724,
+      "grad_norm": 0.09402144700288773,
+      "learning_rate": 7.586923327242234e-06,
+      "loss": 8.6062,
+      "step": 149730
+    },
+    {
+      "epoch": 0.747783964643312,
+      "grad_norm": 0.08989568054676056,
+      "learning_rate": 7.585421412300684e-06,
+      "loss": 8.5849,
+      "step": 149740
+    },
+    {
+      "epoch": 0.7478339034682514,
+      "grad_norm": 0.09065549820661545,
+      "learning_rate": 7.583919497359133e-06,
+      "loss": 8.5967,
+      "step": 149750
+    },
+    {
+      "epoch": 0.7478838422931908,
+      "grad_norm": 0.09307155758142471,
+      "learning_rate": 7.582417582417582e-06,
+      "loss": 8.6004,
+      "step": 149760
+    },
+    {
+      "epoch": 0.7479337811181302,
+      "grad_norm": 0.09423384815454483,
+      "learning_rate": 7.580915667476032e-06,
+      "loss": 8.595,
+      "step": 149770
+    },
+    {
+      "epoch": 0.7479837199430698,
+      "grad_norm": 0.09108050912618637,
+      "learning_rate": 7.579413752534482e-06,
+      "loss": 8.5889,
+      "step": 149780
+    },
+    {
+      "epoch": 0.7480336587680092,
+      "grad_norm": 0.0906979888677597,
+      "learning_rate": 7.577911837592932e-06,
+      "loss": 8.5869,
+      "step": 149790
+    },
+    {
+      "epoch": 0.7480835975929486,
+      "grad_norm": 0.09189769625663757,
+      "learning_rate": 7.57640992265138e-06,
+      "loss": 8.5869,
+      "step": 149800
+    },
+    {
+      "epoch": 0.748133536417888,
+      "grad_norm": 0.09509114921092987,
+      "learning_rate": 7.57490800770983e-06,
+      "loss": 8.5979,
+      "step": 149810
+    },
+    {
+      "epoch": 0.7481834752428276,
+      "grad_norm": 0.09381816536188126,
+      "learning_rate": 7.57340609276828e-06,
+      "loss": 8.5848,
+      "step": 149820
+    },
+    {
+      "epoch": 0.748233414067767,
+      "grad_norm": 0.10044172406196594,
+      "learning_rate": 7.571904177826729e-06,
+      "loss": 8.568,
+      "step": 149830
+    },
+    {
+      "epoch": 0.7482833528927064,
+      "grad_norm": 0.09521984308958054,
+      "learning_rate": 7.570402262885179e-06,
+      "loss": 8.5959,
+      "step": 149840
+    },
+    {
+      "epoch": 0.7483332917176458,
+      "grad_norm": 0.09529656916856766,
+      "learning_rate": 7.568900347943628e-06,
+      "loss": 8.5873,
+      "step": 149850
+    },
+    {
+      "epoch": 0.7483832305425854,
+      "grad_norm": 0.09173472225666046,
+      "learning_rate": 7.567398433002077e-06,
+      "loss": 8.6029,
+      "step": 149860
+    },
+    {
+      "epoch": 0.7484331693675248,
+      "grad_norm": 0.0962473452091217,
+      "learning_rate": 7.565896518060527e-06,
+      "loss": 8.6064,
+      "step": 149870
+    },
+    {
+      "epoch": 0.7484831081924642,
+      "grad_norm": 0.0965263694524765,
+      "learning_rate": 7.564394603118977e-06,
+      "loss": 8.5685,
+      "step": 149880
+    },
+    {
+      "epoch": 0.7485330470174036,
+      "grad_norm": 0.09147603064775467,
+      "learning_rate": 7.562892688177427e-06,
+      "loss": 8.5998,
+      "step": 149890
+    },
+    {
+      "epoch": 0.7485829858423432,
+      "grad_norm": 0.09058742225170135,
+      "learning_rate": 7.561390773235877e-06,
+      "loss": 8.5803,
+      "step": 149900
+    },
+    {
+      "epoch": 0.7486329246672826,
+      "grad_norm": 0.09538568556308746,
+      "learning_rate": 7.559888858294325e-06,
+      "loss": 8.5839,
+      "step": 149910
+    },
+    {
+      "epoch": 0.748682863492222,
+      "grad_norm": 0.08751174807548523,
+      "learning_rate": 7.558386943352775e-06,
+      "loss": 8.5857,
+      "step": 149920
+    },
+    {
+      "epoch": 0.7487328023171614,
+      "grad_norm": 0.09170546382665634,
+      "learning_rate": 7.556885028411224e-06,
+      "loss": 8.5967,
+      "step": 149930
+    },
+    {
+      "epoch": 0.748782741142101,
+      "grad_norm": 0.09635576605796814,
+      "learning_rate": 7.5553831134696744e-06,
+      "loss": 8.59,
+      "step": 149940
+    },
+    {
+      "epoch": 0.7488326799670404,
+      "grad_norm": 0.09509419649839401,
+      "learning_rate": 7.553881198528125e-06,
+      "loss": 8.5917,
+      "step": 149950
+    },
+    {
+      "epoch": 0.7488826187919798,
+      "grad_norm": 0.08711056411266327,
+      "learning_rate": 7.552379283586572e-06,
+      "loss": 8.5994,
+      "step": 149960
+    },
+    {
+      "epoch": 0.7489325576169192,
+      "grad_norm": 0.0929313451051712,
+      "learning_rate": 7.550877368645022e-06,
+      "loss": 8.5855,
+      "step": 149970
+    },
+    {
+      "epoch": 0.7489824964418588,
+      "grad_norm": 0.09222406148910522,
+      "learning_rate": 7.549375453703472e-06,
+      "loss": 8.6066,
+      "step": 149980
+    },
+    {
+      "epoch": 0.7490324352667982,
+      "grad_norm": 0.0890507623553276,
+      "learning_rate": 7.547873538761922e-06,
+      "loss": 8.6042,
+      "step": 149990
+    },
+    {
+      "epoch": 0.7490823740917376,
+      "grad_norm": 0.08804387599229813,
+      "learning_rate": 7.546371623820372e-06,
+      "loss": 8.6063,
+      "step": 150000
+    },
+    {
+      "epoch": 0.749132312916677,
+      "grad_norm": 0.0897492840886116,
+      "learning_rate": 7.54486970887882e-06,
+      "loss": 8.5969,
+      "step": 150010
+    },
+    {
+      "epoch": 0.7491822517416166,
+      "grad_norm": 0.08836290240287781,
+      "learning_rate": 7.54336779393727e-06,
+      "loss": 8.5929,
+      "step": 150020
+    },
+    {
+      "epoch": 0.749232190566556,
+      "grad_norm": 0.09971511363983154,
+      "learning_rate": 7.541865878995719e-06,
+      "loss": 8.5952,
+      "step": 150030
+    },
+    {
+      "epoch": 0.7492821293914954,
+      "grad_norm": 0.09459482133388519,
+      "learning_rate": 7.5403639640541694e-06,
+      "loss": 8.5777,
+      "step": 150040
+    },
+    {
+      "epoch": 0.7493320682164348,
+      "grad_norm": 0.09455052018165588,
+      "learning_rate": 7.53886204911262e-06,
+      "loss": 8.5771,
+      "step": 150050
+    },
+    {
+      "epoch": 0.7493820070413744,
+      "grad_norm": 0.09091456979513168,
+      "learning_rate": 7.537360134171069e-06,
+      "loss": 8.6112,
+      "step": 150060
+    },
+    {
+      "epoch": 0.7494319458663138,
+      "grad_norm": 0.08837003260850906,
+      "learning_rate": 7.5358582192295174e-06,
+      "loss": 8.5919,
+      "step": 150070
+    },
+    {
+      "epoch": 0.7494818846912532,
+      "grad_norm": 0.09216312319040298,
+      "learning_rate": 7.534356304287967e-06,
+      "loss": 8.5865,
+      "step": 150080
+    },
+    {
+      "epoch": 0.7495318235161926,
+      "grad_norm": 0.09196959435939789,
+      "learning_rate": 7.532854389346417e-06,
+      "loss": 8.5715,
+      "step": 150090
+    },
+    {
+      "epoch": 0.7495817623411322,
+      "grad_norm": 0.0905485451221466,
+      "learning_rate": 7.531352474404867e-06,
+      "loss": 8.5851,
+      "step": 150100
+    },
+    {
+      "epoch": 0.7496317011660716,
+      "grad_norm": 0.0880269706249237,
+      "learning_rate": 7.5298505594633164e-06,
+      "loss": 8.5813,
+      "step": 150110
+    },
+    {
+      "epoch": 0.749681639991011,
+      "grad_norm": 0.08941976726055145,
+      "learning_rate": 7.528348644521765e-06,
+      "loss": 8.5816,
+      "step": 150120
+    },
+    {
+      "epoch": 0.7497315788159504,
+      "grad_norm": 0.08973130583763123,
+      "learning_rate": 7.526846729580214e-06,
+      "loss": 8.6029,
+      "step": 150130
+    },
+    {
+      "epoch": 0.74978151764089,
+      "grad_norm": 0.0887095034122467,
+      "learning_rate": 7.5253448146386644e-06,
+      "loss": 8.6152,
+      "step": 150140
+    },
+    {
+      "epoch": 0.7498314564658294,
+      "grad_norm": 0.09397480636835098,
+      "learning_rate": 7.523842899697115e-06,
+      "loss": 8.5888,
+      "step": 150150
+    },
+    {
+      "epoch": 0.7498813952907688,
+      "grad_norm": 0.09192110598087311,
+      "learning_rate": 7.522340984755564e-06,
+      "loss": 8.5769,
+      "step": 150160
+    },
+    {
+      "epoch": 0.7499313341157082,
+      "grad_norm": 0.09322792291641235,
+      "learning_rate": 7.5208390698140124e-06,
+      "loss": 8.5972,
+      "step": 150170
+    },
+    {
+      "epoch": 0.7499812729406478,
+      "grad_norm": 0.08950521796941757,
+      "learning_rate": 7.519337154872462e-06,
+      "loss": 8.5905,
+      "step": 150180
+    },
+    {
+      "epoch": 0.7500312117655872,
+      "grad_norm": 0.0880652666091919,
+      "learning_rate": 7.517835239930912e-06,
+      "loss": 8.6012,
+      "step": 150190
+    },
+    {
+      "epoch": 0.7500811505905266,
+      "grad_norm": 0.09776634722948074,
+      "learning_rate": 7.516333324989362e-06,
+      "loss": 8.5777,
+      "step": 150200
+    },
+    {
+      "epoch": 0.750131089415466,
+      "grad_norm": 0.09283629804849625,
+      "learning_rate": 7.5148314100478115e-06,
+      "loss": 8.5807,
+      "step": 150210
+    },
+    {
+      "epoch": 0.7501810282404056,
+      "grad_norm": 0.09522504359483719,
+      "learning_rate": 7.513329495106262e-06,
+      "loss": 8.576,
+      "step": 150220
+    },
+    {
+      "epoch": 0.750230967065345,
+      "grad_norm": 0.09157579392194748,
+      "learning_rate": 7.511827580164709e-06,
+      "loss": 8.5927,
+      "step": 150230
+    },
+    {
+      "epoch": 0.7502809058902844,
+      "grad_norm": 0.09480082988739014,
+      "learning_rate": 7.5103256652231594e-06,
+      "loss": 8.6008,
+      "step": 150240
+    },
+    {
+      "epoch": 0.7503308447152238,
+      "grad_norm": 0.0909399539232254,
+      "learning_rate": 7.50882375028161e-06,
+      "loss": 8.5823,
+      "step": 150250
+    },
+    {
+      "epoch": 0.7503807835401634,
+      "grad_norm": 0.09307267516851425,
+      "learning_rate": 7.507321835340059e-06,
+      "loss": 8.5907,
+      "step": 150260
+    },
+    {
+      "epoch": 0.7504307223651028,
+      "grad_norm": 0.08955024927854538,
+      "learning_rate": 7.505819920398509e-06,
+      "loss": 8.5954,
+      "step": 150270
+    },
+    {
+      "epoch": 0.7504806611900422,
+      "grad_norm": 0.09251957386732101,
+      "learning_rate": 7.504318005456957e-06,
+      "loss": 8.5802,
+      "step": 150280
+    },
+    {
+      "epoch": 0.7505306000149816,
+      "grad_norm": 0.09843941777944565,
+      "learning_rate": 7.502816090515407e-06,
+      "loss": 8.5834,
+      "step": 150290
+    },
+    {
+      "epoch": 0.7505805388399212,
+      "grad_norm": 0.09598127752542496,
+      "learning_rate": 7.501314175573857e-06,
+      "loss": 8.599,
+      "step": 150300
+    },
+    {
+      "epoch": 0.7506304776648606,
+      "grad_norm": 0.09445951879024506,
+      "learning_rate": 7.4998122606323065e-06,
+      "loss": 8.5844,
+      "step": 150310
+    },
+    {
+      "epoch": 0.7506804164898,
+      "grad_norm": 0.08694125711917877,
+      "learning_rate": 7.498310345690756e-06,
+      "loss": 8.606,
+      "step": 150320
+    },
+    {
+      "epoch": 0.7507303553147394,
+      "grad_norm": 0.08730349689722061,
+      "learning_rate": 7.496808430749205e-06,
+      "loss": 8.5941,
+      "step": 150330
+    },
+    {
+      "epoch": 0.7507802941396788,
+      "grad_norm": 0.0891655758023262,
+      "learning_rate": 7.495306515807655e-06,
+      "loss": 8.5755,
+      "step": 150340
+    },
+    {
+      "epoch": 0.7508302329646184,
+      "grad_norm": 0.08927493542432785,
+      "learning_rate": 7.493804600866105e-06,
+      "loss": 8.5948,
+      "step": 150350
+    },
+    {
+      "epoch": 0.7508801717895578,
+      "grad_norm": 0.09330999106168747,
+      "learning_rate": 7.492302685924554e-06,
+      "loss": 8.5993,
+      "step": 150360
+    },
+    {
+      "epoch": 0.7509301106144972,
+      "grad_norm": 0.09538490325212479,
+      "learning_rate": 7.490800770983003e-06,
+      "loss": 8.5914,
+      "step": 150370
+    },
+    {
+      "epoch": 0.7509800494394366,
+      "grad_norm": 0.09310005605220795,
+      "learning_rate": 7.489298856041453e-06,
+      "loss": 8.5874,
+      "step": 150380
+    },
+    {
+      "epoch": 0.7510299882643762,
+      "grad_norm": 0.0930032730102539,
+      "learning_rate": 7.487796941099903e-06,
+      "loss": 8.5895,
+      "step": 150390
+    },
+    {
+      "epoch": 0.7510799270893156,
+      "grad_norm": 0.08846841007471085,
+      "learning_rate": 7.486295026158352e-06,
+      "loss": 8.5724,
+      "step": 150400
+    },
+    {
+      "epoch": 0.751129865914255,
+      "grad_norm": 0.09779760241508484,
+      "learning_rate": 7.4847931112168015e-06,
+      "loss": 8.5736,
+      "step": 150410
+    },
+    {
+      "epoch": 0.7511798047391944,
+      "grad_norm": 0.10576785355806351,
+      "learning_rate": 7.483291196275252e-06,
+      "loss": 8.5726,
+      "step": 150420
+    },
+    {
+      "epoch": 0.751229743564134,
+      "grad_norm": 0.09451400488615036,
+      "learning_rate": 7.4817892813337e-06,
+      "loss": 8.5857,
+      "step": 150430
+    },
+    {
+      "epoch": 0.7512796823890734,
+      "grad_norm": 0.09466114640235901,
+      "learning_rate": 7.48028736639215e-06,
+      "loss": 8.597,
+      "step": 150440
+    },
+    {
+      "epoch": 0.7513296212140128,
+      "grad_norm": 0.08923625946044922,
+      "learning_rate": 7.4787854514506e-06,
+      "loss": 8.6096,
+      "step": 150450
+    },
+    {
+      "epoch": 0.7513795600389522,
+      "grad_norm": 0.08818081766366959,
+      "learning_rate": 7.477283536509049e-06,
+      "loss": 8.604,
+      "step": 150460
+    },
+    {
+      "epoch": 0.7514294988638918,
+      "grad_norm": 0.08820995688438416,
+      "learning_rate": 7.475781621567499e-06,
+      "loss": 8.5916,
+      "step": 150470
+    },
+    {
+      "epoch": 0.7514794376888312,
+      "grad_norm": 0.09002698957920074,
+      "learning_rate": 7.474279706625948e-06,
+      "loss": 8.5879,
+      "step": 150480
+    },
+    {
+      "epoch": 0.7515293765137706,
+      "grad_norm": 0.09170237928628922,
+      "learning_rate": 7.472777791684398e-06,
+      "loss": 8.5907,
+      "step": 150490
+    },
+    {
+      "epoch": 0.75157931533871,
+      "grad_norm": 0.09593456238508224,
+      "learning_rate": 7.471275876742848e-06,
+      "loss": 8.5872,
+      "step": 150500
+    },
+    {
+      "epoch": 0.7516292541636496,
+      "grad_norm": 0.09875084459781647,
+      "learning_rate": 7.4697739618012965e-06,
+      "loss": 8.5823,
+      "step": 150510
+    },
+    {
+      "epoch": 0.751679192988589,
+      "grad_norm": 0.09118742495775223,
+      "learning_rate": 7.468272046859747e-06,
+      "loss": 8.5823,
+      "step": 150520
+    },
+    {
+      "epoch": 0.7517291318135284,
+      "grad_norm": 0.09258361160755157,
+      "learning_rate": 7.466770131918195e-06,
+      "loss": 8.5972,
+      "step": 150530
+    },
+    {
+      "epoch": 0.7517790706384678,
+      "grad_norm": 0.08977973461151123,
+      "learning_rate": 7.465268216976645e-06,
+      "loss": 8.6059,
+      "step": 150540
+    },
+    {
+      "epoch": 0.7518290094634074,
+      "grad_norm": 0.092478908598423,
+      "learning_rate": 7.4637663020350955e-06,
+      "loss": 8.5935,
+      "step": 150550
+    },
+    {
+      "epoch": 0.7518789482883468,
+      "grad_norm": 0.09068572521209717,
+      "learning_rate": 7.462264387093544e-06,
+      "loss": 8.5762,
+      "step": 150560
+    },
+    {
+      "epoch": 0.7519288871132862,
+      "grad_norm": 0.09068078547716141,
+      "learning_rate": 7.460762472151994e-06,
+      "loss": 8.5839,
+      "step": 150570
+    },
+    {
+      "epoch": 0.7519788259382256,
+      "grad_norm": 0.09269540011882782,
+      "learning_rate": 7.4592605572104435e-06,
+      "loss": 8.5909,
+      "step": 150580
+    },
+    {
+      "epoch": 0.7520287647631652,
+      "grad_norm": 0.0924522653222084,
+      "learning_rate": 7.457758642268893e-06,
+      "loss": 8.5919,
+      "step": 150590
+    },
+    {
+      "epoch": 0.7520787035881046,
+      "grad_norm": 0.09281891584396362,
+      "learning_rate": 7.456256727327343e-06,
+      "loss": 8.5784,
+      "step": 150600
+    },
+    {
+      "epoch": 0.752128642413044,
+      "grad_norm": 0.08694092184305191,
+      "learning_rate": 7.4547548123857915e-06,
+      "loss": 8.5848,
+      "step": 150610
+    },
+    {
+      "epoch": 0.7521785812379834,
+      "grad_norm": 0.10442210733890533,
+      "learning_rate": 7.453252897444242e-06,
+      "loss": 8.5919,
+      "step": 150620
+    },
+    {
+      "epoch": 0.752228520062923,
+      "grad_norm": 0.09364094585180283,
+      "learning_rate": 7.451750982502691e-06,
+      "loss": 8.6023,
+      "step": 150630
+    },
+    {
+      "epoch": 0.7522784588878624,
+      "grad_norm": 0.08750166743993759,
+      "learning_rate": 7.45024906756114e-06,
+      "loss": 8.5849,
+      "step": 150640
+    },
+    {
+      "epoch": 0.7523283977128018,
+      "grad_norm": 0.08991952240467072,
+      "learning_rate": 7.4487471526195905e-06,
+      "loss": 8.5911,
+      "step": 150650
+    },
+    {
+      "epoch": 0.7523783365377412,
+      "grad_norm": 0.10013854503631592,
+      "learning_rate": 7.44724523767804e-06,
+      "loss": 8.5982,
+      "step": 150660
+    },
+    {
+      "epoch": 0.7524282753626808,
+      "grad_norm": 0.09347521513700485,
+      "learning_rate": 7.445743322736489e-06,
+      "loss": 8.5847,
+      "step": 150670
+    },
+    {
+      "epoch": 0.7524782141876202,
+      "grad_norm": 0.09399186819791794,
+      "learning_rate": 7.4442414077949385e-06,
+      "loss": 8.6024,
+      "step": 150680
+    },
+    {
+      "epoch": 0.7525281530125596,
+      "grad_norm": 0.09026849269866943,
+      "learning_rate": 7.442739492853388e-06,
+      "loss": 8.5992,
+      "step": 150690
+    },
+    {
+      "epoch": 0.752578091837499,
+      "grad_norm": 0.08900460600852966,
+      "learning_rate": 7.441237577911838e-06,
+      "loss": 8.5872,
+      "step": 150700
+    },
+    {
+      "epoch": 0.7526280306624386,
+      "grad_norm": 0.09363497048616409,
+      "learning_rate": 7.439735662970287e-06,
+      "loss": 8.5889,
+      "step": 150710
+    },
+    {
+      "epoch": 0.752677969487378,
+      "grad_norm": 0.09363177418708801,
+      "learning_rate": 7.438233748028737e-06,
+      "loss": 8.5861,
+      "step": 150720
+    },
+    {
+      "epoch": 0.7527279083123174,
+      "grad_norm": 0.08974208682775497,
+      "learning_rate": 7.436731833087186e-06,
+      "loss": 8.5841,
+      "step": 150730
+    },
+    {
+      "epoch": 0.7527778471372568,
+      "grad_norm": 0.09006848931312561,
+      "learning_rate": 7.435229918145636e-06,
+      "loss": 8.5754,
+      "step": 150740
+    },
+    {
+      "epoch": 0.7528277859621963,
+      "grad_norm": 0.08987919241189957,
+      "learning_rate": 7.4337280032040855e-06,
+      "loss": 8.5911,
+      "step": 150750
+    },
+    {
+      "epoch": 0.7528777247871358,
+      "grad_norm": 0.0890379548072815,
+      "learning_rate": 7.432226088262535e-06,
+      "loss": 8.5964,
+      "step": 150760
+    },
+    {
+      "epoch": 0.7529276636120752,
+      "grad_norm": 0.08632665872573853,
+      "learning_rate": 7.430724173320984e-06,
+      "loss": 8.5989,
+      "step": 150770
+    },
+    {
+      "epoch": 0.7529776024370146,
+      "grad_norm": 0.09325092285871506,
+      "learning_rate": 7.4292222583794335e-06,
+      "loss": 8.5849,
+      "step": 150780
+    },
+    {
+      "epoch": 0.7530275412619541,
+      "grad_norm": 0.09268093854188919,
+      "learning_rate": 7.427720343437884e-06,
+      "loss": 8.5921,
+      "step": 150790
+    },
+    {
+      "epoch": 0.7530774800868936,
+      "grad_norm": 0.09364005923271179,
+      "learning_rate": 7.426218428496333e-06,
+      "loss": 8.5925,
+      "step": 150800
+    },
+    {
+      "epoch": 0.753127418911833,
+      "grad_norm": 0.08780599385499954,
+      "learning_rate": 7.424716513554782e-06,
+      "loss": 8.601,
+      "step": 150810
+    },
+    {
+      "epoch": 0.7531773577367724,
+      "grad_norm": 0.0943470448255539,
+      "learning_rate": 7.4232145986132325e-06,
+      "loss": 8.6098,
+      "step": 150820
+    },
+    {
+      "epoch": 0.753227296561712,
+      "grad_norm": 0.08829768002033234,
+      "learning_rate": 7.421712683671681e-06,
+      "loss": 8.5951,
+      "step": 150830
+    },
+    {
+      "epoch": 0.7532772353866514,
+      "grad_norm": 0.0940447673201561,
+      "learning_rate": 7.420210768730131e-06,
+      "loss": 8.5996,
+      "step": 150840
+    },
+    {
+      "epoch": 0.7533271742115908,
+      "grad_norm": 0.09546596556901932,
+      "learning_rate": 7.4187088537885805e-06,
+      "loss": 8.5794,
+      "step": 150850
+    },
+    {
+      "epoch": 0.7533771130365302,
+      "grad_norm": 0.09206235408782959,
+      "learning_rate": 7.41720693884703e-06,
+      "loss": 8.5881,
+      "step": 150860
+    },
+    {
+      "epoch": 0.7534270518614697,
+      "grad_norm": 0.08884277194738388,
+      "learning_rate": 7.41570502390548e-06,
+      "loss": 8.6059,
+      "step": 150870
+    },
+    {
+      "epoch": 0.7534769906864092,
+      "grad_norm": 0.0894254744052887,
+      "learning_rate": 7.414203108963929e-06,
+      "loss": 8.6018,
+      "step": 150880
+    },
+    {
+      "epoch": 0.7535269295113486,
+      "grad_norm": 0.08869513869285583,
+      "learning_rate": 7.412701194022379e-06,
+      "loss": 8.5989,
+      "step": 150890
+    },
+    {
+      "epoch": 0.753576868336288,
+      "grad_norm": 0.08933467417955399,
+      "learning_rate": 7.411199279080829e-06,
+      "loss": 8.5957,
+      "step": 150900
+    },
+    {
+      "epoch": 0.7536268071612275,
+      "grad_norm": 0.09024026989936829,
+      "learning_rate": 7.409697364139277e-06,
+      "loss": 8.5688,
+      "step": 150910
+    },
+    {
+      "epoch": 0.753676745986167,
+      "grad_norm": 0.08978093415498734,
+      "learning_rate": 7.4081954491977275e-06,
+      "loss": 8.5984,
+      "step": 150920
+    },
+    {
+      "epoch": 0.7537266848111064,
+      "grad_norm": 0.09174445271492004,
+      "learning_rate": 7.406693534256177e-06,
+      "loss": 8.5912,
+      "step": 150930
+    },
+    {
+      "epoch": 0.7537766236360458,
+      "grad_norm": 0.09789343923330307,
+      "learning_rate": 7.405191619314626e-06,
+      "loss": 8.5811,
+      "step": 150940
+    },
+    {
+      "epoch": 0.7538265624609853,
+      "grad_norm": 0.09015634655952454,
+      "learning_rate": 7.403689704373076e-06,
+      "loss": 8.5899,
+      "step": 150950
+    },
+    {
+      "epoch": 0.7538765012859248,
+      "grad_norm": 0.08942896872758865,
+      "learning_rate": 7.402187789431525e-06,
+      "loss": 8.5947,
+      "step": 150960
+    },
+    {
+      "epoch": 0.7539264401108642,
+      "grad_norm": 0.09148923307657242,
+      "learning_rate": 7.400685874489975e-06,
+      "loss": 8.5759,
+      "step": 150970
+    },
+    {
+      "epoch": 0.7539763789358036,
+      "grad_norm": 0.09370266646146774,
+      "learning_rate": 7.399183959548425e-06,
+      "loss": 8.5746,
+      "step": 150980
+    },
+    {
+      "epoch": 0.7540263177607431,
+      "grad_norm": 0.09445227682590485,
+      "learning_rate": 7.397682044606874e-06,
+      "loss": 8.6121,
+      "step": 150990
+    },
+    {
+      "epoch": 0.7540762565856826,
+      "grad_norm": 0.08653920888900757,
+      "learning_rate": 7.396180129665324e-06,
+      "loss": 8.5719,
+      "step": 151000
+    },
+    {
+      "epoch": 0.754126195410622,
+      "grad_norm": 0.09342148154973984,
+      "learning_rate": 7.394678214723772e-06,
+      "loss": 8.5733,
+      "step": 151010
+    },
+    {
+      "epoch": 0.7541761342355614,
+      "grad_norm": 0.09228193759918213,
+      "learning_rate": 7.3931762997822225e-06,
+      "loss": 8.597,
+      "step": 151020
+    },
+    {
+      "epoch": 0.7542260730605009,
+      "grad_norm": 0.08755730837583542,
+      "learning_rate": 7.391674384840673e-06,
+      "loss": 8.5821,
+      "step": 151030
+    },
+    {
+      "epoch": 0.7542760118854404,
+      "grad_norm": 0.08993065357208252,
+      "learning_rate": 7.390172469899121e-06,
+      "loss": 8.6094,
+      "step": 151040
+    },
+    {
+      "epoch": 0.7543259507103798,
+      "grad_norm": 0.08330941945314407,
+      "learning_rate": 7.388670554957571e-06,
+      "loss": 8.6262,
+      "step": 151050
+    },
+    {
+      "epoch": 0.7543758895353192,
+      "grad_norm": 0.09145517647266388,
+      "learning_rate": 7.387168640016021e-06,
+      "loss": 8.5803,
+      "step": 151060
+    },
+    {
+      "epoch": 0.7544258283602587,
+      "grad_norm": 0.09170622378587723,
+      "learning_rate": 7.38566672507447e-06,
+      "loss": 8.568,
+      "step": 151070
+    },
+    {
+      "epoch": 0.7544757671851982,
+      "grad_norm": 0.08807757496833801,
+      "learning_rate": 7.38416481013292e-06,
+      "loss": 8.5906,
+      "step": 151080
+    },
+    {
+      "epoch": 0.7545257060101376,
+      "grad_norm": 0.09455586969852448,
+      "learning_rate": 7.382662895191369e-06,
+      "loss": 8.6016,
+      "step": 151090
+    },
+    {
+      "epoch": 0.754575644835077,
+      "grad_norm": 0.08837214857339859,
+      "learning_rate": 7.381160980249819e-06,
+      "loss": 8.5958,
+      "step": 151100
+    },
+    {
+      "epoch": 0.7546255836600165,
+      "grad_norm": 0.08906712383031845,
+      "learning_rate": 7.379659065308268e-06,
+      "loss": 8.5873,
+      "step": 151110
+    },
+    {
+      "epoch": 0.754675522484956,
+      "grad_norm": 0.10366376489400864,
+      "learning_rate": 7.3781571503667175e-06,
+      "loss": 8.5838,
+      "step": 151120
+    },
+    {
+      "epoch": 0.7547254613098954,
+      "grad_norm": 0.08967901021242142,
+      "learning_rate": 7.376655235425168e-06,
+      "loss": 8.5893,
+      "step": 151130
+    },
+    {
+      "epoch": 0.7547754001348348,
+      "grad_norm": 0.0880189910531044,
+      "learning_rate": 7.375153320483617e-06,
+      "loss": 8.5897,
+      "step": 151140
+    },
+    {
+      "epoch": 0.7548253389597743,
+      "grad_norm": 0.09614763408899307,
+      "learning_rate": 7.373651405542066e-06,
+      "loss": 8.5823,
+      "step": 151150
+    },
+    {
+      "epoch": 0.7548752777847137,
+      "grad_norm": 0.0998152643442154,
+      "learning_rate": 7.372149490600516e-06,
+      "loss": 8.5901,
+      "step": 151160
+    },
+    {
+      "epoch": 0.7549252166096532,
+      "grad_norm": 0.09459684044122696,
+      "learning_rate": 7.370647575658965e-06,
+      "loss": 8.5769,
+      "step": 151170
+    },
+    {
+      "epoch": 0.7549751554345926,
+      "grad_norm": 0.09051591157913208,
+      "learning_rate": 7.369145660717415e-06,
+      "loss": 8.593,
+      "step": 151180
+    },
+    {
+      "epoch": 0.7550250942595321,
+      "grad_norm": 0.0956014096736908,
+      "learning_rate": 7.3676437457758646e-06,
+      "loss": 8.5867,
+      "step": 151190
+    },
+    {
+      "epoch": 0.7550750330844715,
+      "grad_norm": 0.10061989724636078,
+      "learning_rate": 7.366141830834314e-06,
+      "loss": 8.5855,
+      "step": 151200
+    },
+    {
+      "epoch": 0.755124971909411,
+      "grad_norm": 0.09272131323814392,
+      "learning_rate": 7.364639915892763e-06,
+      "loss": 8.5966,
+      "step": 151210
+    },
+    {
+      "epoch": 0.7551749107343504,
+      "grad_norm": 0.09143613278865814,
+      "learning_rate": 7.363138000951213e-06,
+      "loss": 8.6061,
+      "step": 151220
+    },
+    {
+      "epoch": 0.7552248495592899,
+      "grad_norm": 0.09460750222206116,
+      "learning_rate": 7.361636086009663e-06,
+      "loss": 8.5901,
+      "step": 151230
+    },
+    {
+      "epoch": 0.7552747883842293,
+      "grad_norm": 0.09221066534519196,
+      "learning_rate": 7.360134171068112e-06,
+      "loss": 8.5844,
+      "step": 151240
+    },
+    {
+      "epoch": 0.7553247272091688,
+      "grad_norm": 0.09496848285198212,
+      "learning_rate": 7.358632256126561e-06,
+      "loss": 8.5875,
+      "step": 151250
+    },
+    {
+      "epoch": 0.7553746660341082,
+      "grad_norm": 0.0936916247010231,
+      "learning_rate": 7.357130341185011e-06,
+      "loss": 8.5976,
+      "step": 151260
+    },
+    {
+      "epoch": 0.7554246048590477,
+      "grad_norm": 0.09380991011857986,
+      "learning_rate": 7.355628426243461e-06,
+      "loss": 8.6017,
+      "step": 151270
+    },
+    {
+      "epoch": 0.7554745436839871,
+      "grad_norm": 0.09433276206254959,
+      "learning_rate": 7.35412651130191e-06,
+      "loss": 8.5803,
+      "step": 151280
+    },
+    {
+      "epoch": 0.7555244825089266,
+      "grad_norm": 0.08973414450883865,
+      "learning_rate": 7.3526245963603596e-06,
+      "loss": 8.5969,
+      "step": 151290
+    },
+    {
+      "epoch": 0.755574421333866,
+      "grad_norm": 0.09096003323793411,
+      "learning_rate": 7.35112268141881e-06,
+      "loss": 8.5919,
+      "step": 151300
+    },
+    {
+      "epoch": 0.7556243601588054,
+      "grad_norm": 0.0915827602148056,
+      "learning_rate": 7.349620766477258e-06,
+      "loss": 8.5893,
+      "step": 151310
+    },
+    {
+      "epoch": 0.7556742989837449,
+      "grad_norm": 0.09033535420894623,
+      "learning_rate": 7.348118851535708e-06,
+      "loss": 8.5979,
+      "step": 151320
+    },
+    {
+      "epoch": 0.7557242378086844,
+      "grad_norm": 0.0904705822467804,
+      "learning_rate": 7.346616936594158e-06,
+      "loss": 8.5985,
+      "step": 151330
+    },
+    {
+      "epoch": 0.7557741766336238,
+      "grad_norm": 0.0926397368311882,
+      "learning_rate": 7.345115021652607e-06,
+      "loss": 8.5686,
+      "step": 151340
+    },
+    {
+      "epoch": 0.7558241154585632,
+      "grad_norm": 0.09501095861196518,
+      "learning_rate": 7.343613106711057e-06,
+      "loss": 8.5688,
+      "step": 151350
+    },
+    {
+      "epoch": 0.7558740542835027,
+      "grad_norm": 0.09006702154874802,
+      "learning_rate": 7.342111191769506e-06,
+      "loss": 8.588,
+      "step": 151360
+    },
+    {
+      "epoch": 0.7559239931084422,
+      "grad_norm": 0.09147243946790695,
+      "learning_rate": 7.340609276827956e-06,
+      "loss": 8.5847,
+      "step": 151370
+    },
+    {
+      "epoch": 0.7559739319333816,
+      "grad_norm": 0.09361589699983597,
+      "learning_rate": 7.339107361886406e-06,
+      "loss": 8.5841,
+      "step": 151380
+    },
+    {
+      "epoch": 0.756023870758321,
+      "grad_norm": 0.0927942767739296,
+      "learning_rate": 7.3376054469448546e-06,
+      "loss": 8.6122,
+      "step": 151390
+    },
+    {
+      "epoch": 0.7560738095832605,
+      "grad_norm": 0.0871824249625206,
+      "learning_rate": 7.336103532003305e-06,
+      "loss": 8.5985,
+      "step": 151400
+    },
+    {
+      "epoch": 0.7561237484082,
+      "grad_norm": 0.09094393253326416,
+      "learning_rate": 7.334601617061753e-06,
+      "loss": 8.5906,
+      "step": 151410
+    },
+    {
+      "epoch": 0.7561736872331394,
+      "grad_norm": 0.08855657279491425,
+      "learning_rate": 7.333099702120203e-06,
+      "loss": 8.5768,
+      "step": 151420
+    },
+    {
+      "epoch": 0.7562236260580788,
+      "grad_norm": 0.09399604797363281,
+      "learning_rate": 7.331597787178654e-06,
+      "loss": 8.5941,
+      "step": 151430
+    },
+    {
+      "epoch": 0.7562735648830183,
+      "grad_norm": 0.09146516025066376,
+      "learning_rate": 7.330095872237102e-06,
+      "loss": 8.5875,
+      "step": 151440
+    },
+    {
+      "epoch": 0.7563235037079578,
+      "grad_norm": 0.09131018072366714,
+      "learning_rate": 7.328593957295552e-06,
+      "loss": 8.5829,
+      "step": 151450
+    },
+    {
+      "epoch": 0.7563734425328972,
+      "grad_norm": 0.09400021284818649,
+      "learning_rate": 7.327092042354002e-06,
+      "loss": 8.5939,
+      "step": 151460
+    },
+    {
+      "epoch": 0.7564233813578366,
+      "grad_norm": 0.08741091191768646,
+      "learning_rate": 7.325590127412451e-06,
+      "loss": 8.5914,
+      "step": 151470
+    },
+    {
+      "epoch": 0.7564733201827761,
+      "grad_norm": 0.09049748629331589,
+      "learning_rate": 7.324088212470901e-06,
+      "loss": 8.5944,
+      "step": 151480
+    },
+    {
+      "epoch": 0.7565232590077156,
+      "grad_norm": 0.09107566624879837,
+      "learning_rate": 7.32258629752935e-06,
+      "loss": 8.5838,
+      "step": 151490
+    },
+    {
+      "epoch": 0.756573197832655,
+      "grad_norm": 0.08954470604658127,
+      "learning_rate": 7.3210843825878e-06,
+      "loss": 8.5864,
+      "step": 151500
+    },
+    {
+      "epoch": 0.7566231366575944,
+      "grad_norm": 0.09204570204019547,
+      "learning_rate": 7.319582467646249e-06,
+      "loss": 8.5808,
+      "step": 151510
+    },
+    {
+      "epoch": 0.7566730754825339,
+      "grad_norm": 0.09428434073925018,
+      "learning_rate": 7.318080552704698e-06,
+      "loss": 8.5759,
+      "step": 151520
+    },
+    {
+      "epoch": 0.7567230143074734,
+      "grad_norm": 0.08692929893732071,
+      "learning_rate": 7.316578637763149e-06,
+      "loss": 8.5886,
+      "step": 151530
+    },
+    {
+      "epoch": 0.7567729531324128,
+      "grad_norm": 0.08495133370161057,
+      "learning_rate": 7.315076722821598e-06,
+      "loss": 8.5822,
+      "step": 151540
+    },
+    {
+      "epoch": 0.7568228919573522,
+      "grad_norm": 0.09004715830087662,
+      "learning_rate": 7.313574807880047e-06,
+      "loss": 8.5829,
+      "step": 151550
+    },
+    {
+      "epoch": 0.7568728307822917,
+      "grad_norm": 0.09108411520719528,
+      "learning_rate": 7.312072892938497e-06,
+      "loss": 8.5784,
+      "step": 151560
+    },
+    {
+      "epoch": 0.7569227696072311,
+      "grad_norm": 0.0907498225569725,
+      "learning_rate": 7.310570977996946e-06,
+      "loss": 8.5694,
+      "step": 151570
+    },
+    {
+      "epoch": 0.7569727084321706,
+      "grad_norm": 0.08905670791864395,
+      "learning_rate": 7.309069063055396e-06,
+      "loss": 8.5844,
+      "step": 151580
+    },
+    {
+      "epoch": 0.75702264725711,
+      "grad_norm": 0.08316656202077866,
+      "learning_rate": 7.3075671481138454e-06,
+      "loss": 8.5998,
+      "step": 151590
+    },
+    {
+      "epoch": 0.7570725860820495,
+      "grad_norm": 0.09294084459543228,
+      "learning_rate": 7.306065233172295e-06,
+      "loss": 8.5846,
+      "step": 151600
+    },
+    {
+      "epoch": 0.757122524906989,
+      "grad_norm": 0.08754400908946991,
+      "learning_rate": 7.304563318230744e-06,
+      "loss": 8.5693,
+      "step": 151610
+    },
+    {
+      "epoch": 0.7571724637319284,
+      "grad_norm": 0.09371070563793182,
+      "learning_rate": 7.303061403289194e-06,
+      "loss": 8.5871,
+      "step": 151620
+    },
+    {
+      "epoch": 0.7572224025568678,
+      "grad_norm": 0.08874963223934174,
+      "learning_rate": 7.301559488347644e-06,
+      "loss": 8.5855,
+      "step": 151630
+    },
+    {
+      "epoch": 0.7572723413818073,
+      "grad_norm": 0.09325557202100754,
+      "learning_rate": 7.300057573406093e-06,
+      "loss": 8.5884,
+      "step": 151640
+    },
+    {
+      "epoch": 0.7573222802067467,
+      "grad_norm": 0.0924559012055397,
+      "learning_rate": 7.298555658464542e-06,
+      "loss": 8.582,
+      "step": 151650
+    },
+    {
+      "epoch": 0.7573722190316862,
+      "grad_norm": 0.09226152300834656,
+      "learning_rate": 7.297053743522992e-06,
+      "loss": 8.5726,
+      "step": 151660
+    },
+    {
+      "epoch": 0.7574221578566256,
+      "grad_norm": 0.09244561940431595,
+      "learning_rate": 7.295551828581442e-06,
+      "loss": 8.5807,
+      "step": 151670
+    },
+    {
+      "epoch": 0.7574720966815651,
+      "grad_norm": 0.08866026997566223,
+      "learning_rate": 7.294049913639891e-06,
+      "loss": 8.6003,
+      "step": 151680
+    },
+    {
+      "epoch": 0.7575220355065045,
+      "grad_norm": 0.09373471140861511,
+      "learning_rate": 7.2925479986983404e-06,
+      "loss": 8.5914,
+      "step": 151690
+    },
+    {
+      "epoch": 0.757571974331444,
+      "grad_norm": 0.09084302186965942,
+      "learning_rate": 7.291046083756791e-06,
+      "loss": 8.5745,
+      "step": 151700
+    },
+    {
+      "epoch": 0.7576219131563834,
+      "grad_norm": 0.0983499139547348,
+      "learning_rate": 7.289544168815239e-06,
+      "loss": 8.5901,
+      "step": 151710
+    },
+    {
+      "epoch": 0.7576718519813229,
+      "grad_norm": 0.09153123944997787,
+      "learning_rate": 7.288042253873689e-06,
+      "loss": 8.5917,
+      "step": 151720
+    },
+    {
+      "epoch": 0.7577217908062623,
+      "grad_norm": 0.08924426138401031,
+      "learning_rate": 7.286540338932139e-06,
+      "loss": 8.5917,
+      "step": 151730
+    },
+    {
+      "epoch": 0.7577717296312018,
+      "grad_norm": 0.09073387831449509,
+      "learning_rate": 7.285038423990588e-06,
+      "loss": 8.5854,
+      "step": 151740
+    },
+    {
+      "epoch": 0.7578216684561412,
+      "grad_norm": 0.09264533221721649,
+      "learning_rate": 7.283536509049038e-06,
+      "loss": 8.5888,
+      "step": 151750
+    },
+    {
+      "epoch": 0.7578716072810807,
+      "grad_norm": 0.09185033291578293,
+      "learning_rate": 7.282034594107487e-06,
+      "loss": 8.585,
+      "step": 151760
+    },
+    {
+      "epoch": 0.7579215461060201,
+      "grad_norm": 0.09249900281429291,
+      "learning_rate": 7.280532679165937e-06,
+      "loss": 8.5612,
+      "step": 151770
+    },
+    {
+      "epoch": 0.7579714849309596,
+      "grad_norm": 0.08698815107345581,
+      "learning_rate": 7.279030764224387e-06,
+      "loss": 8.6089,
+      "step": 151780
+    },
+    {
+      "epoch": 0.758021423755899,
+      "grad_norm": 0.09104190021753311,
+      "learning_rate": 7.2775288492828355e-06,
+      "loss": 8.5981,
+      "step": 151790
+    },
+    {
+      "epoch": 0.7580713625808385,
+      "grad_norm": 0.08875631541013718,
+      "learning_rate": 7.276026934341286e-06,
+      "loss": 8.5945,
+      "step": 151800
+    },
+    {
+      "epoch": 0.7581213014057779,
+      "grad_norm": 0.08652567863464355,
+      "learning_rate": 7.274525019399734e-06,
+      "loss": 8.5882,
+      "step": 151810
+    },
+    {
+      "epoch": 0.7581712402307174,
+      "grad_norm": 0.0900021344423294,
+      "learning_rate": 7.273023104458184e-06,
+      "loss": 8.5922,
+      "step": 151820
+    },
+    {
+      "epoch": 0.7582211790556568,
+      "grad_norm": 0.09021075814962387,
+      "learning_rate": 7.2715211895166345e-06,
+      "loss": 8.5893,
+      "step": 151830
+    },
+    {
+      "epoch": 0.7582711178805963,
+      "grad_norm": 0.09226670861244202,
+      "learning_rate": 7.270019274575083e-06,
+      "loss": 8.5892,
+      "step": 151840
+    },
+    {
+      "epoch": 0.7583210567055357,
+      "grad_norm": 0.09223869442939758,
+      "learning_rate": 7.268517359633533e-06,
+      "loss": 8.5831,
+      "step": 151850
+    },
+    {
+      "epoch": 0.7583709955304752,
+      "grad_norm": 0.08831837028265,
+      "learning_rate": 7.2670154446919825e-06,
+      "loss": 8.5805,
+      "step": 151860
+    },
+    {
+      "epoch": 0.7584209343554146,
+      "grad_norm": 0.0934990644454956,
+      "learning_rate": 7.265513529750432e-06,
+      "loss": 8.5758,
+      "step": 151870
+    },
+    {
+      "epoch": 0.7584708731803541,
+      "grad_norm": 0.09644410014152527,
+      "learning_rate": 7.264011614808882e-06,
+      "loss": 8.5791,
+      "step": 151880
+    },
+    {
+      "epoch": 0.7585208120052935,
+      "grad_norm": 0.08798360824584961,
+      "learning_rate": 7.2625096998673305e-06,
+      "loss": 8.584,
+      "step": 151890
+    },
+    {
+      "epoch": 0.758570750830233,
+      "grad_norm": 0.08095114678144455,
+      "learning_rate": 7.261007784925781e-06,
+      "loss": 8.6034,
+      "step": 151900
+    },
+    {
+      "epoch": 0.7586206896551724,
+      "grad_norm": 0.09009264409542084,
+      "learning_rate": 7.25950586998423e-06,
+      "loss": 8.5723,
+      "step": 151910
+    },
+    {
+      "epoch": 0.7586706284801119,
+      "grad_norm": 0.08817630261182785,
+      "learning_rate": 7.258003955042679e-06,
+      "loss": 8.5885,
+      "step": 151920
+    },
+    {
+      "epoch": 0.7587205673050513,
+      "grad_norm": 0.09514699131250381,
+      "learning_rate": 7.2565020401011295e-06,
+      "loss": 8.5849,
+      "step": 151930
+    },
+    {
+      "epoch": 0.7587705061299908,
+      "grad_norm": 0.09281176328659058,
+      "learning_rate": 7.255000125159579e-06,
+      "loss": 8.579,
+      "step": 151940
+    },
+    {
+      "epoch": 0.7588204449549302,
+      "grad_norm": 0.08887535333633423,
+      "learning_rate": 7.253498210218028e-06,
+      "loss": 8.6016,
+      "step": 151950
+    },
+    {
+      "epoch": 0.7588703837798697,
+      "grad_norm": 0.08661656081676483,
+      "learning_rate": 7.251996295276478e-06,
+      "loss": 8.5982,
+      "step": 151960
+    },
+    {
+      "epoch": 0.7589203226048091,
+      "grad_norm": 0.09328266978263855,
+      "learning_rate": 7.250494380334927e-06,
+      "loss": 8.5797,
+      "step": 151970
+    },
+    {
+      "epoch": 0.7589702614297485,
+      "grad_norm": 0.09176629781723022,
+      "learning_rate": 7.248992465393377e-06,
+      "loss": 8.5894,
+      "step": 151980
+    },
+    {
+      "epoch": 0.759020200254688,
+      "grad_norm": 0.08731918781995773,
+      "learning_rate": 7.247490550451826e-06,
+      "loss": 8.5794,
+      "step": 151990
+    },
+    {
+      "epoch": 0.7590701390796275,
+      "grad_norm": 0.088376984000206,
+      "learning_rate": 7.245988635510276e-06,
+      "loss": 8.5798,
+      "step": 152000
+    },
+    {
+      "epoch": 0.7591200779045669,
+      "grad_norm": 0.09818810224533081,
+      "learning_rate": 7.244486720568726e-06,
+      "loss": 8.5913,
+      "step": 152010
+    },
+    {
+      "epoch": 0.7591700167295063,
+      "grad_norm": 0.08995293080806732,
+      "learning_rate": 7.242984805627175e-06,
+      "loss": 8.5928,
+      "step": 152020
+    },
+    {
+      "epoch": 0.7592199555544458,
+      "grad_norm": 0.09495987743139267,
+      "learning_rate": 7.2414828906856245e-06,
+      "loss": 8.5887,
+      "step": 152030
+    },
+    {
+      "epoch": 0.7592698943793853,
+      "grad_norm": 0.09125667065382004,
+      "learning_rate": 7.239980975744074e-06,
+      "loss": 8.5905,
+      "step": 152040
+    },
+    {
+      "epoch": 0.7593198332043247,
+      "grad_norm": 0.09113696217536926,
+      "learning_rate": 7.238479060802523e-06,
+      "loss": 8.5895,
+      "step": 152050
+    },
+    {
+      "epoch": 0.7593697720292641,
+      "grad_norm": 0.09109106659889221,
+      "learning_rate": 7.236977145860973e-06,
+      "loss": 8.5895,
+      "step": 152060
+    },
+    {
+      "epoch": 0.7594197108542036,
+      "grad_norm": 0.09017052501440048,
+      "learning_rate": 7.235475230919423e-06,
+      "loss": 8.5818,
+      "step": 152070
+    },
+    {
+      "epoch": 0.7594696496791431,
+      "grad_norm": 0.08849617093801498,
+      "learning_rate": 7.233973315977872e-06,
+      "loss": 8.5957,
+      "step": 152080
+    },
+    {
+      "epoch": 0.7595195885040825,
+      "grad_norm": 0.09171763062477112,
+      "learning_rate": 7.232471401036321e-06,
+      "loss": 8.5808,
+      "step": 152090
+    },
+    {
+      "epoch": 0.7595695273290219,
+      "grad_norm": 0.10110411792993546,
+      "learning_rate": 7.2309694860947715e-06,
+      "loss": 8.5877,
+      "step": 152100
+    },
+    {
+      "epoch": 0.7596194661539614,
+      "grad_norm": 0.09818938374519348,
+      "learning_rate": 7.229467571153221e-06,
+      "loss": 8.6031,
+      "step": 152110
+    },
+    {
+      "epoch": 0.7596694049789009,
+      "grad_norm": 0.09149213880300522,
+      "learning_rate": 7.22796565621167e-06,
+      "loss": 8.5883,
+      "step": 152120
+    },
+    {
+      "epoch": 0.7597193438038403,
+      "grad_norm": 0.09270060062408447,
+      "learning_rate": 7.2264637412701195e-06,
+      "loss": 8.603,
+      "step": 152130
+    },
+    {
+      "epoch": 0.7597692826287797,
+      "grad_norm": 0.08630634099245071,
+      "learning_rate": 7.224961826328569e-06,
+      "loss": 8.5836,
+      "step": 152140
+    },
+    {
+      "epoch": 0.7598192214537192,
+      "grad_norm": 0.0882948487997055,
+      "learning_rate": 7.223459911387019e-06,
+      "loss": 8.5835,
+      "step": 152150
+    },
+    {
+      "epoch": 0.7598691602786587,
+      "grad_norm": 0.09250038117170334,
+      "learning_rate": 7.221957996445468e-06,
+      "loss": 8.5625,
+      "step": 152160
+    },
+    {
+      "epoch": 0.7599190991035981,
+      "grad_norm": 0.09016095101833344,
+      "learning_rate": 7.220456081503918e-06,
+      "loss": 8.5824,
+      "step": 152170
+    },
+    {
+      "epoch": 0.7599690379285375,
+      "grad_norm": 0.08857563883066177,
+      "learning_rate": 7.218954166562367e-06,
+      "loss": 8.5876,
+      "step": 152180
+    },
+    {
+      "epoch": 0.760018976753477,
+      "grad_norm": 0.09116075932979584,
+      "learning_rate": 7.217452251620816e-06,
+      "loss": 8.577,
+      "step": 152190
+    },
+    {
+      "epoch": 0.7600689155784165,
+      "grad_norm": 0.09463620185852051,
+      "learning_rate": 7.2159503366792665e-06,
+      "loss": 8.5771,
+      "step": 152200
+    },
+    {
+      "epoch": 0.7601188544033559,
+      "grad_norm": 0.09192495793104172,
+      "learning_rate": 7.214448421737716e-06,
+      "loss": 8.5971,
+      "step": 152210
+    },
+    {
+      "epoch": 0.7601687932282953,
+      "grad_norm": 0.09425206482410431,
+      "learning_rate": 7.212946506796165e-06,
+      "loss": 8.5846,
+      "step": 152220
+    },
+    {
+      "epoch": 0.7602187320532348,
+      "grad_norm": 0.0923011526465416,
+      "learning_rate": 7.211444591854615e-06,
+      "loss": 8.576,
+      "step": 152230
+    },
+    {
+      "epoch": 0.7602686708781743,
+      "grad_norm": 0.08811124414205551,
+      "learning_rate": 7.209942676913064e-06,
+      "loss": 8.587,
+      "step": 152240
+    },
+    {
+      "epoch": 0.7603186097031137,
+      "grad_norm": 0.08695264905691147,
+      "learning_rate": 7.208440761971514e-06,
+      "loss": 8.59,
+      "step": 152250
+    },
+    {
+      "epoch": 0.7603685485280531,
+      "grad_norm": 0.0906856581568718,
+      "learning_rate": 7.206938847029963e-06,
+      "loss": 8.5963,
+      "step": 152260
+    },
+    {
+      "epoch": 0.7604184873529926,
+      "grad_norm": 0.09710587561130524,
+      "learning_rate": 7.205436932088413e-06,
+      "loss": 8.5624,
+      "step": 152270
+    },
+    {
+      "epoch": 0.7604684261779321,
+      "grad_norm": 0.09217939525842667,
+      "learning_rate": 7.203935017146863e-06,
+      "loss": 8.5973,
+      "step": 152280
+    },
+    {
+      "epoch": 0.7605183650028715,
+      "grad_norm": 0.09139391779899597,
+      "learning_rate": 7.202433102205311e-06,
+      "loss": 8.5674,
+      "step": 152290
+    },
+    {
+      "epoch": 0.7605683038278109,
+      "grad_norm": 0.08444222062826157,
+      "learning_rate": 7.2009311872637615e-06,
+      "loss": 8.5952,
+      "step": 152300
+    },
+    {
+      "epoch": 0.7606182426527504,
+      "grad_norm": 0.09180841594934464,
+      "learning_rate": 7.199429272322212e-06,
+      "loss": 8.596,
+      "step": 152310
+    },
+    {
+      "epoch": 0.7606681814776898,
+      "grad_norm": 0.09961362183094025,
+      "learning_rate": 7.19792735738066e-06,
+      "loss": 8.5865,
+      "step": 152320
+    },
+    {
+      "epoch": 0.7607181203026293,
+      "grad_norm": 0.09278172999620438,
+      "learning_rate": 7.19642544243911e-06,
+      "loss": 8.5995,
+      "step": 152330
+    },
+    {
+      "epoch": 0.7607680591275687,
+      "grad_norm": 0.08981714397668839,
+      "learning_rate": 7.194923527497559e-06,
+      "loss": 8.578,
+      "step": 152340
+    },
+    {
+      "epoch": 0.7608179979525082,
+      "grad_norm": 0.09385214000940323,
+      "learning_rate": 7.193421612556009e-06,
+      "loss": 8.5784,
+      "step": 152350
+    },
+    {
+      "epoch": 0.7608679367774476,
+      "grad_norm": 0.09259159117937088,
+      "learning_rate": 7.191919697614459e-06,
+      "loss": 8.5798,
+      "step": 152360
+    },
+    {
+      "epoch": 0.7609178756023871,
+      "grad_norm": 0.0899551659822464,
+      "learning_rate": 7.190417782672908e-06,
+      "loss": 8.5876,
+      "step": 152370
+    },
+    {
+      "epoch": 0.7609678144273265,
+      "grad_norm": 0.09498420357704163,
+      "learning_rate": 7.188915867731358e-06,
+      "loss": 8.5757,
+      "step": 152380
+    },
+    {
+      "epoch": 0.761017753252266,
+      "grad_norm": 0.09275230020284653,
+      "learning_rate": 7.187413952789807e-06,
+      "loss": 8.5825,
+      "step": 152390
+    },
+    {
+      "epoch": 0.7610676920772054,
+      "grad_norm": 0.09266764670610428,
+      "learning_rate": 7.1859120378482565e-06,
+      "loss": 8.5912,
+      "step": 152400
+    },
+    {
+      "epoch": 0.7611176309021449,
+      "grad_norm": 0.08858975768089294,
+      "learning_rate": 7.184410122906707e-06,
+      "loss": 8.5878,
+      "step": 152410
+    },
+    {
+      "epoch": 0.7611675697270843,
+      "grad_norm": 0.09748895466327667,
+      "learning_rate": 7.182908207965155e-06,
+      "loss": 8.5744,
+      "step": 152420
+    },
+    {
+      "epoch": 0.7612175085520237,
+      "grad_norm": 0.09607774019241333,
+      "learning_rate": 7.181406293023605e-06,
+      "loss": 8.5716,
+      "step": 152430
+    },
+    {
+      "epoch": 0.7612674473769632,
+      "grad_norm": 0.09278061985969543,
+      "learning_rate": 7.179904378082055e-06,
+      "loss": 8.5937,
+      "step": 152440
+    },
+    {
+      "epoch": 0.7613173862019027,
+      "grad_norm": 0.0908476933836937,
+      "learning_rate": 7.178402463140504e-06,
+      "loss": 8.5842,
+      "step": 152450
+    },
+    {
+      "epoch": 0.7613673250268421,
+      "grad_norm": 0.0958239734172821,
+      "learning_rate": 7.176900548198954e-06,
+      "loss": 8.593,
+      "step": 152460
+    },
+    {
+      "epoch": 0.7614172638517815,
+      "grad_norm": 0.08807183057069778,
+      "learning_rate": 7.1753986332574035e-06,
+      "loss": 8.5842,
+      "step": 152470
+    },
+    {
+      "epoch": 0.761467202676721,
+      "grad_norm": 0.09366779029369354,
+      "learning_rate": 7.173896718315853e-06,
+      "loss": 8.5819,
+      "step": 152480
+    },
+    {
+      "epoch": 0.7615171415016605,
+      "grad_norm": 0.09066038578748703,
+      "learning_rate": 7.172394803374302e-06,
+      "loss": 8.5982,
+      "step": 152490
+    },
+    {
+      "epoch": 0.7615670803265999,
+      "grad_norm": 0.08774998039007187,
+      "learning_rate": 7.1708928884327515e-06,
+      "loss": 8.5932,
+      "step": 152500
+    },
+    {
+      "epoch": 0.7616170191515393,
+      "grad_norm": 0.09205503761768341,
+      "learning_rate": 7.169390973491202e-06,
+      "loss": 8.5773,
+      "step": 152510
+    },
+    {
+      "epoch": 0.7616669579764788,
+      "grad_norm": 0.08959227055311203,
+      "learning_rate": 7.167889058549651e-06,
+      "loss": 8.586,
+      "step": 152520
+    },
+    {
+      "epoch": 0.7617168968014183,
+      "grad_norm": 0.09033005684614182,
+      "learning_rate": 7.1663871436081e-06,
+      "loss": 8.5832,
+      "step": 152530
+    },
+    {
+      "epoch": 0.7617668356263577,
+      "grad_norm": 0.08551044017076492,
+      "learning_rate": 7.16488522866655e-06,
+      "loss": 8.5845,
+      "step": 152540
+    },
+    {
+      "epoch": 0.7618167744512971,
+      "grad_norm": 0.09518376737833023,
+      "learning_rate": 7.163383313725e-06,
+      "loss": 8.5857,
+      "step": 152550
+    },
+    {
+      "epoch": 0.7618667132762366,
+      "grad_norm": 0.0905361995100975,
+      "learning_rate": 7.161881398783449e-06,
+      "loss": 8.5745,
+      "step": 152560
+    },
+    {
+      "epoch": 0.7619166521011761,
+      "grad_norm": 0.09344843029975891,
+      "learning_rate": 7.1603794838418985e-06,
+      "loss": 8.5731,
+      "step": 152570
+    },
+    {
+      "epoch": 0.7619665909261155,
+      "grad_norm": 0.08621740341186523,
+      "learning_rate": 7.158877568900348e-06,
+      "loss": 8.582,
+      "step": 152580
+    },
+    {
+      "epoch": 0.7620165297510549,
+      "grad_norm": 0.09143950790166855,
+      "learning_rate": 7.157375653958797e-06,
+      "loss": 8.5648,
+      "step": 152590
+    },
+    {
+      "epoch": 0.7620664685759944,
+      "grad_norm": 0.09078672528266907,
+      "learning_rate": 7.155873739017247e-06,
+      "loss": 8.5895,
+      "step": 152600
+    },
+    {
+      "epoch": 0.7621164074009339,
+      "grad_norm": 0.09113630652427673,
+      "learning_rate": 7.154371824075697e-06,
+      "loss": 8.5785,
+      "step": 152610
+    },
+    {
+      "epoch": 0.7621663462258733,
+      "grad_norm": 0.09150314331054688,
+      "learning_rate": 7.152869909134146e-06,
+      "loss": 8.5799,
+      "step": 152620
+    },
+    {
+      "epoch": 0.7622162850508127,
+      "grad_norm": 0.08845186978578568,
+      "learning_rate": 7.151367994192596e-06,
+      "loss": 8.5894,
+      "step": 152630
+    },
+    {
+      "epoch": 0.7622662238757522,
+      "grad_norm": 0.09440822899341583,
+      "learning_rate": 7.149866079251045e-06,
+      "loss": 8.5837,
+      "step": 152640
+    },
+    {
+      "epoch": 0.7623161627006917,
+      "grad_norm": 0.08900156617164612,
+      "learning_rate": 7.148364164309495e-06,
+      "loss": 8.5828,
+      "step": 152650
+    },
+    {
+      "epoch": 0.7623661015256311,
+      "grad_norm": 0.09400945901870728,
+      "learning_rate": 7.146862249367944e-06,
+      "loss": 8.576,
+      "step": 152660
+    },
+    {
+      "epoch": 0.7624160403505705,
+      "grad_norm": 0.09294036775827408,
+      "learning_rate": 7.1453603344263935e-06,
+      "loss": 8.5688,
+      "step": 152670
+    },
+    {
+      "epoch": 0.76246597917551,
+      "grad_norm": 0.09247752279043198,
+      "learning_rate": 7.143858419484844e-06,
+      "loss": 8.5811,
+      "step": 152680
+    },
+    {
+      "epoch": 0.7625159180004495,
+      "grad_norm": 0.09127040207386017,
+      "learning_rate": 7.142356504543292e-06,
+      "loss": 8.5837,
+      "step": 152690
+    },
+    {
+      "epoch": 0.7625658568253889,
+      "grad_norm": 0.0945616066455841,
+      "learning_rate": 7.140854589601742e-06,
+      "loss": 8.5798,
+      "step": 152700
+    },
+    {
+      "epoch": 0.7626157956503283,
+      "grad_norm": 0.08728938549757004,
+      "learning_rate": 7.1393526746601926e-06,
+      "loss": 8.5811,
+      "step": 152710
+    },
+    {
+      "epoch": 0.7626657344752678,
+      "grad_norm": 0.09225558489561081,
+      "learning_rate": 7.137850759718641e-06,
+      "loss": 8.6107,
+      "step": 152720
+    },
+    {
+      "epoch": 0.7627156733002073,
+      "grad_norm": 0.09513969719409943,
+      "learning_rate": 7.136348844777091e-06,
+      "loss": 8.5764,
+      "step": 152730
+    },
+    {
+      "epoch": 0.7627656121251467,
+      "grad_norm": 0.09618328511714935,
+      "learning_rate": 7.13484692983554e-06,
+      "loss": 8.5874,
+      "step": 152740
+    },
+    {
+      "epoch": 0.7628155509500861,
+      "grad_norm": 0.09063924103975296,
+      "learning_rate": 7.13334501489399e-06,
+      "loss": 8.5823,
+      "step": 152750
+    },
+    {
+      "epoch": 0.7628654897750256,
+      "grad_norm": 0.09331314265727997,
+      "learning_rate": 7.13184309995244e-06,
+      "loss": 8.5897,
+      "step": 152760
+    },
+    {
+      "epoch": 0.7629154285999651,
+      "grad_norm": 0.09065219014883041,
+      "learning_rate": 7.1303411850108886e-06,
+      "loss": 8.5748,
+      "step": 152770
+    },
+    {
+      "epoch": 0.7629653674249045,
+      "grad_norm": 0.09267207235097885,
+      "learning_rate": 7.128839270069339e-06,
+      "loss": 8.5953,
+      "step": 152780
+    },
+    {
+      "epoch": 0.7630153062498439,
+      "grad_norm": 0.094063401222229,
+      "learning_rate": 7.127337355127788e-06,
+      "loss": 8.5751,
+      "step": 152790
+    },
+    {
+      "epoch": 0.7630652450747833,
+      "grad_norm": 0.09036116302013397,
+      "learning_rate": 7.125835440186237e-06,
+      "loss": 8.6051,
+      "step": 152800
+    },
+    {
+      "epoch": 0.7631151838997229,
+      "grad_norm": 0.08873546123504639,
+      "learning_rate": 7.1243335252446876e-06,
+      "loss": 8.5893,
+      "step": 152810
+    },
+    {
+      "epoch": 0.7631651227246623,
+      "grad_norm": 0.0934579074382782,
+      "learning_rate": 7.122831610303136e-06,
+      "loss": 8.5921,
+      "step": 152820
+    },
+    {
+      "epoch": 0.7632150615496017,
+      "grad_norm": 0.08765194565057755,
+      "learning_rate": 7.121329695361586e-06,
+      "loss": 8.5853,
+      "step": 152830
+    },
+    {
+      "epoch": 0.7632650003745411,
+      "grad_norm": 0.09874917566776276,
+      "learning_rate": 7.1198277804200356e-06,
+      "loss": 8.6031,
+      "step": 152840
+    },
+    {
+      "epoch": 0.7633149391994807,
+      "grad_norm": 0.10156397521495819,
+      "learning_rate": 7.118325865478485e-06,
+      "loss": 8.5753,
+      "step": 152850
+    },
+    {
+      "epoch": 0.7633648780244201,
+      "grad_norm": 0.0943882167339325,
+      "learning_rate": 7.116823950536935e-06,
+      "loss": 8.5921,
+      "step": 152860
+    },
+    {
+      "epoch": 0.7634148168493595,
+      "grad_norm": 0.0905870795249939,
+      "learning_rate": 7.115322035595384e-06,
+      "loss": 8.5748,
+      "step": 152870
+    },
+    {
+      "epoch": 0.763464755674299,
+      "grad_norm": 0.09174995124340057,
+      "learning_rate": 7.113820120653834e-06,
+      "loss": 8.5743,
+      "step": 152880
+    },
+    {
+      "epoch": 0.7635146944992385,
+      "grad_norm": 0.08916423469781876,
+      "learning_rate": 7.112318205712283e-06,
+      "loss": 8.5986,
+      "step": 152890
+    },
+    {
+      "epoch": 0.7635646333241779,
+      "grad_norm": 0.0889330729842186,
+      "learning_rate": 7.110816290770732e-06,
+      "loss": 8.5888,
+      "step": 152900
+    },
+    {
+      "epoch": 0.7636145721491173,
+      "grad_norm": 0.08929458260536194,
+      "learning_rate": 7.109314375829183e-06,
+      "loss": 8.5861,
+      "step": 152910
+    },
+    {
+      "epoch": 0.7636645109740567,
+      "grad_norm": 0.08744913339614868,
+      "learning_rate": 7.107812460887632e-06,
+      "loss": 8.5891,
+      "step": 152920
+    },
+    {
+      "epoch": 0.7637144497989963,
+      "grad_norm": 0.08869093656539917,
+      "learning_rate": 7.106310545946081e-06,
+      "loss": 8.5786,
+      "step": 152930
+    },
+    {
+      "epoch": 0.7637643886239357,
+      "grad_norm": 0.09083859622478485,
+      "learning_rate": 7.1048086310045306e-06,
+      "loss": 8.5703,
+      "step": 152940
+    },
+    {
+      "epoch": 0.7638143274488751,
+      "grad_norm": 0.09250235557556152,
+      "learning_rate": 7.103306716062981e-06,
+      "loss": 8.5856,
+      "step": 152950
+    },
+    {
+      "epoch": 0.7638642662738145,
+      "grad_norm": 0.09155931323766708,
+      "learning_rate": 7.10180480112143e-06,
+      "loss": 8.5781,
+      "step": 152960
+    },
+    {
+      "epoch": 0.7639142050987541,
+      "grad_norm": 0.08976297080516815,
+      "learning_rate": 7.100302886179879e-06,
+      "loss": 8.5806,
+      "step": 152970
+    },
+    {
+      "epoch": 0.7639641439236935,
+      "grad_norm": 0.0865217074751854,
+      "learning_rate": 7.098800971238329e-06,
+      "loss": 8.5913,
+      "step": 152980
+    },
+    {
+      "epoch": 0.7640140827486329,
+      "grad_norm": 0.09239277988672256,
+      "learning_rate": 7.097299056296778e-06,
+      "loss": 8.5603,
+      "step": 152990
+    },
+    {
+      "epoch": 0.7640640215735723,
+      "grad_norm": 0.09613004326820374,
+      "learning_rate": 7.095797141355228e-06,
+      "loss": 8.5824,
+      "step": 153000
+    },
+    {
+      "epoch": 0.7641139603985119,
+      "grad_norm": 0.09154446423053741,
+      "learning_rate": 7.094295226413678e-06,
+      "loss": 8.5709,
+      "step": 153010
+    },
+    {
+      "epoch": 0.7641638992234513,
+      "grad_norm": 0.08664754033088684,
+      "learning_rate": 7.092793311472127e-06,
+      "loss": 8.5846,
+      "step": 153020
+    },
+    {
+      "epoch": 0.7642138380483907,
+      "grad_norm": 0.08845943212509155,
+      "learning_rate": 7.091291396530577e-06,
+      "loss": 8.5786,
+      "step": 153030
+    },
+    {
+      "epoch": 0.7642637768733301,
+      "grad_norm": 0.09169842302799225,
+      "learning_rate": 7.0897894815890264e-06,
+      "loss": 8.5761,
+      "step": 153040
+    },
+    {
+      "epoch": 0.7643137156982697,
+      "grad_norm": 0.09660103172063828,
+      "learning_rate": 7.088287566647476e-06,
+      "loss": 8.5811,
+      "step": 153050
+    },
+    {
+      "epoch": 0.7643636545232091,
+      "grad_norm": 0.09721507132053375,
+      "learning_rate": 7.086785651705925e-06,
+      "loss": 8.564,
+      "step": 153060
+    },
+    {
+      "epoch": 0.7644135933481485,
+      "grad_norm": 0.08967658877372742,
+      "learning_rate": 7.0852837367643744e-06,
+      "loss": 8.5768,
+      "step": 153070
+    },
+    {
+      "epoch": 0.7644635321730879,
+      "grad_norm": 0.08986034244298935,
+      "learning_rate": 7.083781821822825e-06,
+      "loss": 8.5778,
+      "step": 153080
+    },
+    {
+      "epoch": 0.7645134709980275,
+      "grad_norm": 0.09163659065961838,
+      "learning_rate": 7.082279906881274e-06,
+      "loss": 8.5838,
+      "step": 153090
+    },
+    {
+      "epoch": 0.7645634098229669,
+      "grad_norm": 0.0884322002530098,
+      "learning_rate": 7.080777991939723e-06,
+      "loss": 8.5956,
+      "step": 153100
+    },
+    {
+      "epoch": 0.7646133486479063,
+      "grad_norm": 0.08686332404613495,
+      "learning_rate": 7.0792760769981734e-06,
+      "loss": 8.5683,
+      "step": 153110
+    },
+    {
+      "epoch": 0.7646632874728457,
+      "grad_norm": 0.09294196218252182,
+      "learning_rate": 7.077774162056622e-06,
+      "loss": 8.5796,
+      "step": 153120
+    },
+    {
+      "epoch": 0.7647132262977853,
+      "grad_norm": 0.08890484273433685,
+      "learning_rate": 7.076272247115072e-06,
+      "loss": 8.5687,
+      "step": 153130
+    },
+    {
+      "epoch": 0.7647631651227247,
+      "grad_norm": 0.08904290199279785,
+      "learning_rate": 7.0747703321735214e-06,
+      "loss": 8.5894,
+      "step": 153140
+    },
+    {
+      "epoch": 0.7648131039476641,
+      "grad_norm": 0.09268071502447128,
+      "learning_rate": 7.073268417231971e-06,
+      "loss": 8.5896,
+      "step": 153150
+    },
+    {
+      "epoch": 0.7648630427726035,
+      "grad_norm": 0.08970651030540466,
+      "learning_rate": 7.071766502290421e-06,
+      "loss": 8.6045,
+      "step": 153160
+    },
+    {
+      "epoch": 0.7649129815975431,
+      "grad_norm": 0.09089753031730652,
+      "learning_rate": 7.0702645873488694e-06,
+      "loss": 8.5732,
+      "step": 153170
+    },
+    {
+      "epoch": 0.7649629204224825,
+      "grad_norm": 0.0952310562133789,
+      "learning_rate": 7.06876267240732e-06,
+      "loss": 8.5883,
+      "step": 153180
+    },
+    {
+      "epoch": 0.7650128592474219,
+      "grad_norm": 0.09604635834693909,
+      "learning_rate": 7.06726075746577e-06,
+      "loss": 8.5674,
+      "step": 153190
+    },
+    {
+      "epoch": 0.7650627980723613,
+      "grad_norm": 0.0942385122179985,
+      "learning_rate": 7.065758842524218e-06,
+      "loss": 8.5867,
+      "step": 153200
+    },
+    {
+      "epoch": 0.7651127368973009,
+      "grad_norm": 0.08863166719675064,
+      "learning_rate": 7.0642569275826685e-06,
+      "loss": 8.5591,
+      "step": 153210
+    },
+    {
+      "epoch": 0.7651626757222403,
+      "grad_norm": 0.09197341650724411,
+      "learning_rate": 7.062755012641117e-06,
+      "loss": 8.5931,
+      "step": 153220
+    },
+    {
+      "epoch": 0.7652126145471797,
+      "grad_norm": 0.09086757153272629,
+      "learning_rate": 7.061253097699567e-06,
+      "loss": 8.578,
+      "step": 153230
+    },
+    {
+      "epoch": 0.7652625533721191,
+      "grad_norm": 0.08695393055677414,
+      "learning_rate": 7.059751182758017e-06,
+      "loss": 8.5728,
+      "step": 153240
+    },
+    {
+      "epoch": 0.7653124921970587,
+      "grad_norm": 0.09496577829122543,
+      "learning_rate": 7.058249267816466e-06,
+      "loss": 8.5701,
+      "step": 153250
+    },
+    {
+      "epoch": 0.7653624310219981,
+      "grad_norm": 0.09787900000810623,
+      "learning_rate": 7.056747352874916e-06,
+      "loss": 8.5682,
+      "step": 153260
+    },
+    {
+      "epoch": 0.7654123698469375,
+      "grad_norm": 0.0917130559682846,
+      "learning_rate": 7.055245437933365e-06,
+      "loss": 8.57,
+      "step": 153270
+    },
+    {
+      "epoch": 0.7654623086718769,
+      "grad_norm": 0.08664479851722717,
+      "learning_rate": 7.053743522991815e-06,
+      "loss": 8.5829,
+      "step": 153280
+    },
+    {
+      "epoch": 0.7655122474968163,
+      "grad_norm": 0.08745972812175751,
+      "learning_rate": 7.052241608050265e-06,
+      "loss": 8.564,
+      "step": 153290
+    },
+    {
+      "epoch": 0.7655621863217559,
+      "grad_norm": 0.09336210042238235,
+      "learning_rate": 7.050739693108713e-06,
+      "loss": 8.5767,
+      "step": 153300
+    },
+    {
+      "epoch": 0.7656121251466953,
+      "grad_norm": 0.0911019816994667,
+      "learning_rate": 7.0492377781671635e-06,
+      "loss": 8.5855,
+      "step": 153310
+    },
+    {
+      "epoch": 0.7656620639716347,
+      "grad_norm": 0.09164589643478394,
+      "learning_rate": 7.047735863225613e-06,
+      "loss": 8.5782,
+      "step": 153320
+    },
+    {
+      "epoch": 0.7657120027965741,
+      "grad_norm": 0.09078686684370041,
+      "learning_rate": 7.046233948284062e-06,
+      "loss": 8.5899,
+      "step": 153330
+    },
+    {
+      "epoch": 0.7657619416215137,
+      "grad_norm": 0.09011105448007584,
+      "learning_rate": 7.044732033342512e-06,
+      "loss": 8.5706,
+      "step": 153340
+    },
+    {
+      "epoch": 0.7658118804464531,
+      "grad_norm": 0.08931349217891693,
+      "learning_rate": 7.043230118400962e-06,
+      "loss": 8.5779,
+      "step": 153350
+    },
+    {
+      "epoch": 0.7658618192713925,
+      "grad_norm": 0.09002659469842911,
+      "learning_rate": 7.041728203459411e-06,
+      "loss": 8.5732,
+      "step": 153360
+    },
+    {
+      "epoch": 0.7659117580963319,
+      "grad_norm": 0.0962076261639595,
+      "learning_rate": 7.04022628851786e-06,
+      "loss": 8.5705,
+      "step": 153370
+    },
+    {
+      "epoch": 0.7659616969212715,
+      "grad_norm": 0.0866306945681572,
+      "learning_rate": 7.03872437357631e-06,
+      "loss": 8.5906,
+      "step": 153380
+    },
+    {
+      "epoch": 0.7660116357462109,
+      "grad_norm": 0.09287164360284805,
+      "learning_rate": 7.03722245863476e-06,
+      "loss": 8.5881,
+      "step": 153390
+    },
+    {
+      "epoch": 0.7660615745711503,
+      "grad_norm": 0.09079669415950775,
+      "learning_rate": 7.035720543693209e-06,
+      "loss": 8.5676,
+      "step": 153400
+    },
+    {
+      "epoch": 0.7661115133960897,
+      "grad_norm": 0.08933369070291519,
+      "learning_rate": 7.0342186287516585e-06,
+      "loss": 8.5854,
+      "step": 153410
+    },
+    {
+      "epoch": 0.7661614522210293,
+      "grad_norm": 0.09217403084039688,
+      "learning_rate": 7.032716713810108e-06,
+      "loss": 8.5743,
+      "step": 153420
+    },
+    {
+      "epoch": 0.7662113910459687,
+      "grad_norm": 0.08750222623348236,
+      "learning_rate": 7.031214798868558e-06,
+      "loss": 8.5838,
+      "step": 153430
+    },
+    {
+      "epoch": 0.7662613298709081,
+      "grad_norm": 0.09187041968107224,
+      "learning_rate": 7.029712883927007e-06,
+      "loss": 8.5779,
+      "step": 153440
+    },
+    {
+      "epoch": 0.7663112686958475,
+      "grad_norm": 0.09300150722265244,
+      "learning_rate": 7.028210968985457e-06,
+      "loss": 8.5867,
+      "step": 153450
+    },
+    {
+      "epoch": 0.7663612075207871,
+      "grad_norm": 0.08883293718099594,
+      "learning_rate": 7.026709054043906e-06,
+      "loss": 8.5779,
+      "step": 153460
+    },
+    {
+      "epoch": 0.7664111463457265,
+      "grad_norm": 0.09035582095384598,
+      "learning_rate": 7.025207139102355e-06,
+      "loss": 8.5666,
+      "step": 153470
+    },
+    {
+      "epoch": 0.7664610851706659,
+      "grad_norm": 0.09275439381599426,
+      "learning_rate": 7.0237052241608055e-06,
+      "loss": 8.5823,
+      "step": 153480
+    },
+    {
+      "epoch": 0.7665110239956053,
+      "grad_norm": 0.08916142582893372,
+      "learning_rate": 7.022203309219255e-06,
+      "loss": 8.5674,
+      "step": 153490
+    },
+    {
+      "epoch": 0.7665609628205449,
+      "grad_norm": 0.09958826005458832,
+      "learning_rate": 7.020701394277704e-06,
+      "loss": 8.5855,
+      "step": 153500
+    },
+    {
+      "epoch": 0.7666109016454843,
+      "grad_norm": 0.09693287312984467,
+      "learning_rate": 7.019199479336154e-06,
+      "loss": 8.6037,
+      "step": 153510
+    },
+    {
+      "epoch": 0.7666608404704237,
+      "grad_norm": 0.09419762343168259,
+      "learning_rate": 7.017697564394603e-06,
+      "loss": 8.6014,
+      "step": 153520
+    },
+    {
+      "epoch": 0.7667107792953631,
+      "grad_norm": 0.08924181014299393,
+      "learning_rate": 7.016195649453053e-06,
+      "loss": 8.5692,
+      "step": 153530
+    },
+    {
+      "epoch": 0.7667607181203027,
+      "grad_norm": 0.09446694701910019,
+      "learning_rate": 7.014693734511502e-06,
+      "loss": 8.5536,
+      "step": 153540
+    },
+    {
+      "epoch": 0.7668106569452421,
+      "grad_norm": 0.09108909219503403,
+      "learning_rate": 7.013191819569952e-06,
+      "loss": 8.572,
+      "step": 153550
+    },
+    {
+      "epoch": 0.7668605957701815,
+      "grad_norm": 0.09171731770038605,
+      "learning_rate": 7.011689904628402e-06,
+      "loss": 8.5945,
+      "step": 153560
+    },
+    {
+      "epoch": 0.7669105345951209,
+      "grad_norm": 0.08465012907981873,
+      "learning_rate": 7.01018798968685e-06,
+      "loss": 8.5643,
+      "step": 153570
+    },
+    {
+      "epoch": 0.7669604734200605,
+      "grad_norm": 0.09661739319562912,
+      "learning_rate": 7.0086860747453005e-06,
+      "loss": 8.5638,
+      "step": 153580
+    },
+    {
+      "epoch": 0.7670104122449999,
+      "grad_norm": 0.0985170304775238,
+      "learning_rate": 7.007184159803751e-06,
+      "loss": 8.5805,
+      "step": 153590
+    },
+    {
+      "epoch": 0.7670603510699393,
+      "grad_norm": 0.08951466530561447,
+      "learning_rate": 7.005682244862199e-06,
+      "loss": 8.5677,
+      "step": 153600
+    },
+    {
+      "epoch": 0.7671102898948787,
+      "grad_norm": 0.08966083079576492,
+      "learning_rate": 7.004180329920649e-06,
+      "loss": 8.5769,
+      "step": 153610
+    },
+    {
+      "epoch": 0.7671602287198183,
+      "grad_norm": 0.09198177605867386,
+      "learning_rate": 7.002678414979098e-06,
+      "loss": 8.5674,
+      "step": 153620
+    },
+    {
+      "epoch": 0.7672101675447577,
+      "grad_norm": 0.0966639444231987,
+      "learning_rate": 7.001176500037548e-06,
+      "loss": 8.5847,
+      "step": 153630
+    },
+    {
+      "epoch": 0.7672601063696971,
+      "grad_norm": 0.10365894436836243,
+      "learning_rate": 6.999674585095998e-06,
+      "loss": 8.5755,
+      "step": 153640
+    },
+    {
+      "epoch": 0.7673100451946365,
+      "grad_norm": 0.08775070309638977,
+      "learning_rate": 6.998172670154447e-06,
+      "loss": 8.5752,
+      "step": 153650
+    },
+    {
+      "epoch": 0.7673599840195761,
+      "grad_norm": 0.09237588942050934,
+      "learning_rate": 6.996670755212897e-06,
+      "loss": 8.5767,
+      "step": 153660
+    },
+    {
+      "epoch": 0.7674099228445155,
+      "grad_norm": 0.0916270911693573,
+      "learning_rate": 6.995168840271346e-06,
+      "loss": 8.5828,
+      "step": 153670
+    },
+    {
+      "epoch": 0.7674598616694549,
+      "grad_norm": 0.08937940746545792,
+      "learning_rate": 6.9936669253297955e-06,
+      "loss": 8.5851,
+      "step": 153680
+    },
+    {
+      "epoch": 0.7675098004943943,
+      "grad_norm": 0.08998386561870575,
+      "learning_rate": 6.992165010388246e-06,
+      "loss": 8.5821,
+      "step": 153690
+    },
+    {
+      "epoch": 0.7675597393193339,
+      "grad_norm": 0.09344659745693207,
+      "learning_rate": 6.990663095446694e-06,
+      "loss": 8.5799,
+      "step": 153700
+    },
+    {
+      "epoch": 0.7676096781442733,
+      "grad_norm": 0.08797649294137955,
+      "learning_rate": 6.989161180505144e-06,
+      "loss": 8.5644,
+      "step": 153710
+    },
+    {
+      "epoch": 0.7676596169692127,
+      "grad_norm": 0.08696374297142029,
+      "learning_rate": 6.987659265563594e-06,
+      "loss": 8.575,
+      "step": 153720
+    },
+    {
+      "epoch": 0.7677095557941521,
+      "grad_norm": 0.08707814663648605,
+      "learning_rate": 6.986157350622043e-06,
+      "loss": 8.5779,
+      "step": 153730
+    },
+    {
+      "epoch": 0.7677594946190917,
+      "grad_norm": 0.0909426361322403,
+      "learning_rate": 6.984655435680493e-06,
+      "loss": 8.581,
+      "step": 153740
+    },
+    {
+      "epoch": 0.7678094334440311,
+      "grad_norm": 0.09398233890533447,
+      "learning_rate": 6.9831535207389425e-06,
+      "loss": 8.5727,
+      "step": 153750
+    },
+    {
+      "epoch": 0.7678593722689705,
+      "grad_norm": 0.09371063113212585,
+      "learning_rate": 6.981651605797392e-06,
+      "loss": 8.5808,
+      "step": 153760
+    },
+    {
+      "epoch": 0.7679093110939099,
+      "grad_norm": 0.09734338521957397,
+      "learning_rate": 6.980149690855841e-06,
+      "loss": 8.5737,
+      "step": 153770
+    },
+    {
+      "epoch": 0.7679592499188495,
+      "grad_norm": 0.09083466976881027,
+      "learning_rate": 6.9786477759142905e-06,
+      "loss": 8.5716,
+      "step": 153780
+    },
+    {
+      "epoch": 0.7680091887437889,
+      "grad_norm": 0.08842228353023529,
+      "learning_rate": 6.977145860972741e-06,
+      "loss": 8.5752,
+      "step": 153790
+    },
+    {
+      "epoch": 0.7680591275687283,
+      "grad_norm": 0.08697518706321716,
+      "learning_rate": 6.97564394603119e-06,
+      "loss": 8.5822,
+      "step": 153800
+    },
+    {
+      "epoch": 0.7681090663936677,
+      "grad_norm": 0.08936888724565506,
+      "learning_rate": 6.974142031089639e-06,
+      "loss": 8.582,
+      "step": 153810
+    },
+    {
+      "epoch": 0.7681590052186072,
+      "grad_norm": 0.08924106508493423,
+      "learning_rate": 6.972640116148089e-06,
+      "loss": 8.5737,
+      "step": 153820
+    },
+    {
+      "epoch": 0.7682089440435467,
+      "grad_norm": 0.08949612826108932,
+      "learning_rate": 6.971138201206539e-06,
+      "loss": 8.5744,
+      "step": 153830
+    },
+    {
+      "epoch": 0.7682588828684861,
+      "grad_norm": 0.08947469294071198,
+      "learning_rate": 6.969636286264988e-06,
+      "loss": 8.5754,
+      "step": 153840
+    },
+    {
+      "epoch": 0.7683088216934255,
+      "grad_norm": 0.09205352514982224,
+      "learning_rate": 6.9681343713234375e-06,
+      "loss": 8.5822,
+      "step": 153850
+    },
+    {
+      "epoch": 0.768358760518365,
+      "grad_norm": 0.09942862391471863,
+      "learning_rate": 6.966632456381887e-06,
+      "loss": 8.582,
+      "step": 153860
+    },
+    {
+      "epoch": 0.7684086993433045,
+      "grad_norm": 0.090413898229599,
+      "learning_rate": 6.965130541440336e-06,
+      "loss": 8.5727,
+      "step": 153870
+    },
+    {
+      "epoch": 0.7684586381682439,
+      "grad_norm": 0.09462569653987885,
+      "learning_rate": 6.963628626498786e-06,
+      "loss": 8.5756,
+      "step": 153880
+    },
+    {
+      "epoch": 0.7685085769931833,
+      "grad_norm": 0.09226179867982864,
+      "learning_rate": 6.962126711557236e-06,
+      "loss": 8.5864,
+      "step": 153890
+    },
+    {
+      "epoch": 0.7685585158181228,
+      "grad_norm": 0.08182681351900101,
+      "learning_rate": 6.960624796615685e-06,
+      "loss": 8.5841,
+      "step": 153900
+    },
+    {
+      "epoch": 0.7686084546430623,
+      "grad_norm": 0.09245304763317108,
+      "learning_rate": 6.959122881674135e-06,
+      "loss": 8.5817,
+      "step": 153910
+    },
+    {
+      "epoch": 0.7686583934680017,
+      "grad_norm": 0.09626985341310501,
+      "learning_rate": 6.957620966732584e-06,
+      "loss": 8.595,
+      "step": 153920
+    },
+    {
+      "epoch": 0.7687083322929411,
+      "grad_norm": 0.08915554732084274,
+      "learning_rate": 6.956119051791034e-06,
+      "loss": 8.5715,
+      "step": 153930
+    },
+    {
+      "epoch": 0.7687582711178806,
+      "grad_norm": 0.09438993036746979,
+      "learning_rate": 6.954617136849483e-06,
+      "loss": 8.5903,
+      "step": 153940
+    },
+    {
+      "epoch": 0.7688082099428201,
+      "grad_norm": 0.08598021417856216,
+      "learning_rate": 6.9531152219079325e-06,
+      "loss": 8.5767,
+      "step": 153950
+    },
+    {
+      "epoch": 0.7688581487677595,
+      "grad_norm": 0.09764537960290909,
+      "learning_rate": 6.951613306966383e-06,
+      "loss": 8.5781,
+      "step": 153960
+    },
+    {
+      "epoch": 0.7689080875926989,
+      "grad_norm": 0.09817574918270111,
+      "learning_rate": 6.950111392024831e-06,
+      "loss": 8.5746,
+      "step": 153970
+    },
+    {
+      "epoch": 0.7689580264176384,
+      "grad_norm": 0.08399667590856552,
+      "learning_rate": 6.948609477083281e-06,
+      "loss": 8.5705,
+      "step": 153980
+    },
+    {
+      "epoch": 0.7690079652425779,
+      "grad_norm": 0.09401653707027435,
+      "learning_rate": 6.9471075621417315e-06,
+      "loss": 8.5782,
+      "step": 153990
+    },
+    {
+      "epoch": 0.7690579040675173,
+      "grad_norm": 0.0929185077548027,
+      "learning_rate": 6.94560564720018e-06,
+      "loss": 8.5698,
+      "step": 154000
+    },
+    {
+      "epoch": 0.7691078428924567,
+      "grad_norm": 0.08887774497270584,
+      "learning_rate": 6.94410373225863e-06,
+      "loss": 8.5957,
+      "step": 154010
+    },
+    {
+      "epoch": 0.7691577817173962,
+      "grad_norm": 0.08719068765640259,
+      "learning_rate": 6.942601817317079e-06,
+      "loss": 8.5963,
+      "step": 154020
+    },
+    {
+      "epoch": 0.7692077205423357,
+      "grad_norm": 0.09064088016748428,
+      "learning_rate": 6.941099902375529e-06,
+      "loss": 8.5823,
+      "step": 154030
+    },
+    {
+      "epoch": 0.7692576593672751,
+      "grad_norm": 0.09441380947828293,
+      "learning_rate": 6.939597987433979e-06,
+      "loss": 8.5988,
+      "step": 154040
+    },
+    {
+      "epoch": 0.7693075981922145,
+      "grad_norm": 0.08752386271953583,
+      "learning_rate": 6.9380960724924275e-06,
+      "loss": 8.5738,
+      "step": 154050
+    },
+    {
+      "epoch": 0.769357537017154,
+      "grad_norm": 0.09293375164270401,
+      "learning_rate": 6.936594157550878e-06,
+      "loss": 8.581,
+      "step": 154060
+    },
+    {
+      "epoch": 0.7694074758420935,
+      "grad_norm": 0.09073223173618317,
+      "learning_rate": 6.935092242609327e-06,
+      "loss": 8.586,
+      "step": 154070
+    },
+    {
+      "epoch": 0.7694574146670329,
+      "grad_norm": 0.09582027792930603,
+      "learning_rate": 6.933590327667776e-06,
+      "loss": 8.5888,
+      "step": 154080
+    },
+    {
+      "epoch": 0.7695073534919723,
+      "grad_norm": 0.09688831865787506,
+      "learning_rate": 6.9320884127262265e-06,
+      "loss": 8.5683,
+      "step": 154090
+    },
+    {
+      "epoch": 0.7695572923169118,
+      "grad_norm": 0.09381963312625885,
+      "learning_rate": 6.930586497784675e-06,
+      "loss": 8.5728,
+      "step": 154100
+    },
+    {
+      "epoch": 0.7696072311418513,
+      "grad_norm": 0.0936012864112854,
+      "learning_rate": 6.929084582843125e-06,
+      "loss": 8.5665,
+      "step": 154110
+    },
+    {
+      "epoch": 0.7696571699667907,
+      "grad_norm": 0.08793763071298599,
+      "learning_rate": 6.9275826679015745e-06,
+      "loss": 8.6069,
+      "step": 154120
+    },
+    {
+      "epoch": 0.7697071087917301,
+      "grad_norm": 0.08785630762577057,
+      "learning_rate": 6.926080752960024e-06,
+      "loss": 8.581,
+      "step": 154130
+    },
+    {
+      "epoch": 0.7697570476166696,
+      "grad_norm": 0.09242880344390869,
+      "learning_rate": 6.924578838018474e-06,
+      "loss": 8.5786,
+      "step": 154140
+    },
+    {
+      "epoch": 0.769806986441609,
+      "grad_norm": 0.08848867565393448,
+      "learning_rate": 6.923076923076923e-06,
+      "loss": 8.578,
+      "step": 154150
+    },
+    {
+      "epoch": 0.7698569252665485,
+      "grad_norm": 0.09217596054077148,
+      "learning_rate": 6.921575008135373e-06,
+      "loss": 8.578,
+      "step": 154160
+    },
+    {
+      "epoch": 0.7699068640914879,
+      "grad_norm": 0.08778587728738785,
+      "learning_rate": 6.920073093193823e-06,
+      "loss": 8.5806,
+      "step": 154170
+    },
+    {
+      "epoch": 0.7699568029164274,
+      "grad_norm": 0.09108664095401764,
+      "learning_rate": 6.918571178252271e-06,
+      "loss": 8.5815,
+      "step": 154180
+    },
+    {
+      "epoch": 0.7700067417413669,
+      "grad_norm": 0.09245019406080246,
+      "learning_rate": 6.9170692633107216e-06,
+      "loss": 8.58,
+      "step": 154190
+    },
+    {
+      "epoch": 0.7700566805663063,
+      "grad_norm": 0.09027982503175735,
+      "learning_rate": 6.915567348369171e-06,
+      "loss": 8.5821,
+      "step": 154200
+    },
+    {
+      "epoch": 0.7701066193912457,
+      "grad_norm": 0.093569815158844,
+      "learning_rate": 6.91406543342762e-06,
+      "loss": 8.562,
+      "step": 154210
+    },
+    {
+      "epoch": 0.7701565582161852,
+      "grad_norm": 0.09079516679048538,
+      "learning_rate": 6.91256351848607e-06,
+      "loss": 8.5822,
+      "step": 154220
+    },
+    {
+      "epoch": 0.7702064970411246,
+      "grad_norm": 0.09518477320671082,
+      "learning_rate": 6.91106160354452e-06,
+      "loss": 8.5833,
+      "step": 154230
+    },
+    {
+      "epoch": 0.7702564358660641,
+      "grad_norm": 0.08532573282718658,
+      "learning_rate": 6.909559688602969e-06,
+      "loss": 8.5973,
+      "step": 154240
+    },
+    {
+      "epoch": 0.7703063746910035,
+      "grad_norm": 0.09336396306753159,
+      "learning_rate": 6.908057773661418e-06,
+      "loss": 8.5737,
+      "step": 154250
+    },
+    {
+      "epoch": 0.770356313515943,
+      "grad_norm": 0.09063505381345749,
+      "learning_rate": 6.906555858719868e-06,
+      "loss": 8.5763,
+      "step": 154260
+    },
+    {
+      "epoch": 0.7704062523408824,
+      "grad_norm": 0.08470286428928375,
+      "learning_rate": 6.905053943778318e-06,
+      "loss": 8.5924,
+      "step": 154270
+    },
+    {
+      "epoch": 0.7704561911658219,
+      "grad_norm": 0.09841927140951157,
+      "learning_rate": 6.903552028836767e-06,
+      "loss": 8.5754,
+      "step": 154280
+    },
+    {
+      "epoch": 0.7705061299907613,
+      "grad_norm": 0.09492678195238113,
+      "learning_rate": 6.9020501138952166e-06,
+      "loss": 8.5874,
+      "step": 154290
+    },
+    {
+      "epoch": 0.7705560688157007,
+      "grad_norm": 0.09044993668794632,
+      "learning_rate": 6.900548198953666e-06,
+      "loss": 8.5628,
+      "step": 154300
+    },
+    {
+      "epoch": 0.7706060076406402,
+      "grad_norm": 0.09086602926254272,
+      "learning_rate": 6.899046284012116e-06,
+      "loss": 8.5546,
+      "step": 154310
+    },
+    {
+      "epoch": 0.7706559464655797,
+      "grad_norm": 0.08962773531675339,
+      "learning_rate": 6.897544369070565e-06,
+      "loss": 8.5762,
+      "step": 154320
+    },
+    {
+      "epoch": 0.7707058852905191,
+      "grad_norm": 0.09127478301525116,
+      "learning_rate": 6.896042454129015e-06,
+      "loss": 8.5902,
+      "step": 154330
+    },
+    {
+      "epoch": 0.7707558241154585,
+      "grad_norm": 0.09498311579227448,
+      "learning_rate": 6.894540539187464e-06,
+      "loss": 8.574,
+      "step": 154340
+    },
+    {
+      "epoch": 0.770805762940398,
+      "grad_norm": 0.09378572553396225,
+      "learning_rate": 6.893038624245913e-06,
+      "loss": 8.5666,
+      "step": 154350
+    },
+    {
+      "epoch": 0.7708557017653375,
+      "grad_norm": 0.09132121503353119,
+      "learning_rate": 6.8915367093043636e-06,
+      "loss": 8.5664,
+      "step": 154360
+    },
+    {
+      "epoch": 0.7709056405902769,
+      "grad_norm": 0.0891404077410698,
+      "learning_rate": 6.890034794362813e-06,
+      "loss": 8.5671,
+      "step": 154370
+    },
+    {
+      "epoch": 0.7709555794152163,
+      "grad_norm": 0.0928260087966919,
+      "learning_rate": 6.888532879421262e-06,
+      "loss": 8.5717,
+      "step": 154380
+    },
+    {
+      "epoch": 0.7710055182401558,
+      "grad_norm": 0.092970110476017,
+      "learning_rate": 6.887030964479712e-06,
+      "loss": 8.5745,
+      "step": 154390
+    },
+    {
+      "epoch": 0.7710554570650953,
+      "grad_norm": 0.09026780724525452,
+      "learning_rate": 6.885529049538161e-06,
+      "loss": 8.5715,
+      "step": 154400
+    },
+    {
+      "epoch": 0.7711053958900347,
+      "grad_norm": 0.09131123125553131,
+      "learning_rate": 6.884027134596611e-06,
+      "loss": 8.5806,
+      "step": 154410
+    },
+    {
+      "epoch": 0.7711553347149741,
+      "grad_norm": 0.09628612548112869,
+      "learning_rate": 6.88252521965506e-06,
+      "loss": 8.5872,
+      "step": 154420
+    },
+    {
+      "epoch": 0.7712052735399136,
+      "grad_norm": 0.10040068626403809,
+      "learning_rate": 6.88102330471351e-06,
+      "loss": 8.5681,
+      "step": 154430
+    },
+    {
+      "epoch": 0.7712552123648531,
+      "grad_norm": 0.09062188118696213,
+      "learning_rate": 6.87952138977196e-06,
+      "loss": 8.5869,
+      "step": 154440
+    },
+    {
+      "epoch": 0.7713051511897925,
+      "grad_norm": 0.09356309473514557,
+      "learning_rate": 6.878019474830408e-06,
+      "loss": 8.58,
+      "step": 154450
+    },
+    {
+      "epoch": 0.7713550900147319,
+      "grad_norm": 0.09048853069543839,
+      "learning_rate": 6.876517559888859e-06,
+      "loss": 8.5787,
+      "step": 154460
+    },
+    {
+      "epoch": 0.7714050288396714,
+      "grad_norm": 0.08828842639923096,
+      "learning_rate": 6.875015644947309e-06,
+      "loss": 8.5799,
+      "step": 154470
+    },
+    {
+      "epoch": 0.7714549676646109,
+      "grad_norm": 0.0911807268857956,
+      "learning_rate": 6.873513730005757e-06,
+      "loss": 8.5726,
+      "step": 154480
+    },
+    {
+      "epoch": 0.7715049064895503,
+      "grad_norm": 0.08841972798109055,
+      "learning_rate": 6.8720118150642074e-06,
+      "loss": 8.5756,
+      "step": 154490
+    },
+    {
+      "epoch": 0.7715548453144897,
+      "grad_norm": 0.09796998649835587,
+      "learning_rate": 6.870509900122656e-06,
+      "loss": 8.559,
+      "step": 154500
+    },
+    {
+      "epoch": 0.7716047841394292,
+      "grad_norm": 0.09058402478694916,
+      "learning_rate": 6.869007985181106e-06,
+      "loss": 8.5675,
+      "step": 154510
+    },
+    {
+      "epoch": 0.7716547229643687,
+      "grad_norm": 0.08772189915180206,
+      "learning_rate": 6.867506070239556e-06,
+      "loss": 8.5848,
+      "step": 154520
+    },
+    {
+      "epoch": 0.7717046617893081,
+      "grad_norm": 0.08951723575592041,
+      "learning_rate": 6.866004155298005e-06,
+      "loss": 8.5791,
+      "step": 154530
+    },
+    {
+      "epoch": 0.7717546006142475,
+      "grad_norm": 0.08602496981620789,
+      "learning_rate": 6.864502240356455e-06,
+      "loss": 8.5794,
+      "step": 154540
+    },
+    {
+      "epoch": 0.771804539439187,
+      "grad_norm": 0.08945851027965546,
+      "learning_rate": 6.863000325414904e-06,
+      "loss": 8.5841,
+      "step": 154550
+    },
+    {
+      "epoch": 0.7718544782641265,
+      "grad_norm": 0.09233725816011429,
+      "learning_rate": 6.861498410473354e-06,
+      "loss": 8.5541,
+      "step": 154560
+    },
+    {
+      "epoch": 0.7719044170890659,
+      "grad_norm": 0.09467992186546326,
+      "learning_rate": 6.859996495531804e-06,
+      "loss": 8.5744,
+      "step": 154570
+    },
+    {
+      "epoch": 0.7719543559140053,
+      "grad_norm": 0.08942775428295135,
+      "learning_rate": 6.858494580590252e-06,
+      "loss": 8.5864,
+      "step": 154580
+    },
+    {
+      "epoch": 0.7720042947389448,
+      "grad_norm": 0.08920110762119293,
+      "learning_rate": 6.8569926656487024e-06,
+      "loss": 8.5959,
+      "step": 154590
+    },
+    {
+      "epoch": 0.7720542335638843,
+      "grad_norm": 0.08930312097072601,
+      "learning_rate": 6.855490750707152e-06,
+      "loss": 8.5837,
+      "step": 154600
+    },
+    {
+      "epoch": 0.7721041723888237,
+      "grad_norm": 0.09447375684976578,
+      "learning_rate": 6.853988835765601e-06,
+      "loss": 8.58,
+      "step": 154610
+    },
+    {
+      "epoch": 0.7721541112137631,
+      "grad_norm": 0.08980347961187363,
+      "learning_rate": 6.852486920824051e-06,
+      "loss": 8.5727,
+      "step": 154620
+    },
+    {
+      "epoch": 0.7722040500387026,
+      "grad_norm": 0.09173062443733215,
+      "learning_rate": 6.850985005882501e-06,
+      "loss": 8.5604,
+      "step": 154630
+    },
+    {
+      "epoch": 0.772253988863642,
+      "grad_norm": 0.08936478197574615,
+      "learning_rate": 6.84948309094095e-06,
+      "loss": 8.5684,
+      "step": 154640
+    },
+    {
+      "epoch": 0.7723039276885815,
+      "grad_norm": 0.09296960383653641,
+      "learning_rate": 6.847981175999399e-06,
+      "loss": 8.5477,
+      "step": 154650
+    },
+    {
+      "epoch": 0.7723538665135209,
+      "grad_norm": 0.09330233186483383,
+      "learning_rate": 6.846479261057849e-06,
+      "loss": 8.5855,
+      "step": 154660
+    },
+    {
+      "epoch": 0.7724038053384604,
+      "grad_norm": 0.08701977878808975,
+      "learning_rate": 6.844977346116299e-06,
+      "loss": 8.5892,
+      "step": 154670
+    },
+    {
+      "epoch": 0.7724537441633998,
+      "grad_norm": 0.10141323506832123,
+      "learning_rate": 6.843475431174748e-06,
+      "loss": 8.5798,
+      "step": 154680
+    },
+    {
+      "epoch": 0.7725036829883393,
+      "grad_norm": 0.08847815543413162,
+      "learning_rate": 6.8419735162331974e-06,
+      "loss": 8.579,
+      "step": 154690
+    },
+    {
+      "epoch": 0.7725536218132787,
+      "grad_norm": 0.10151761770248413,
+      "learning_rate": 6.840471601291647e-06,
+      "loss": 8.5819,
+      "step": 154700
+    },
+    {
+      "epoch": 0.7726035606382182,
+      "grad_norm": 0.09666801989078522,
+      "learning_rate": 6.838969686350097e-06,
+      "loss": 8.5728,
+      "step": 154710
+    },
+    {
+      "epoch": 0.7726534994631576,
+      "grad_norm": 0.09449943155050278,
+      "learning_rate": 6.837467771408546e-06,
+      "loss": 8.5817,
+      "step": 154720
+    },
+    {
+      "epoch": 0.7727034382880971,
+      "grad_norm": 0.09444770961999893,
+      "learning_rate": 6.835965856466996e-06,
+      "loss": 8.5654,
+      "step": 154730
+    },
+    {
+      "epoch": 0.7727533771130365,
+      "grad_norm": 0.09364951401948929,
+      "learning_rate": 6.834463941525445e-06,
+      "loss": 8.5647,
+      "step": 154740
+    },
+    {
+      "epoch": 0.772803315937976,
+      "grad_norm": 0.09371913969516754,
+      "learning_rate": 6.832962026583894e-06,
+      "loss": 8.5827,
+      "step": 154750
+    },
+    {
+      "epoch": 0.7728532547629154,
+      "grad_norm": 0.08978746831417084,
+      "learning_rate": 6.8314601116423445e-06,
+      "loss": 8.5917,
+      "step": 154760
+    },
+    {
+      "epoch": 0.7729031935878549,
+      "grad_norm": 0.0957767516374588,
+      "learning_rate": 6.829958196700794e-06,
+      "loss": 8.566,
+      "step": 154770
+    },
+    {
+      "epoch": 0.7729531324127943,
+      "grad_norm": 0.09105443209409714,
+      "learning_rate": 6.828456281759243e-06,
+      "loss": 8.5709,
+      "step": 154780
+    },
+    {
+      "epoch": 0.7730030712377338,
+      "grad_norm": 0.09461674094200134,
+      "learning_rate": 6.826954366817693e-06,
+      "loss": 8.5735,
+      "step": 154790
+    },
+    {
+      "epoch": 0.7730530100626732,
+      "grad_norm": 0.08995918929576874,
+      "learning_rate": 6.825452451876142e-06,
+      "loss": 8.5817,
+      "step": 154800
+    },
+    {
+      "epoch": 0.7731029488876127,
+      "grad_norm": 0.08911199867725372,
+      "learning_rate": 6.823950536934592e-06,
+      "loss": 8.578,
+      "step": 154810
+    },
+    {
+      "epoch": 0.7731528877125521,
+      "grad_norm": 0.09105192124843597,
+      "learning_rate": 6.822448621993041e-06,
+      "loss": 8.5733,
+      "step": 154820
+    },
+    {
+      "epoch": 0.7732028265374916,
+      "grad_norm": 0.08885302394628525,
+      "learning_rate": 6.820946707051491e-06,
+      "loss": 8.5597,
+      "step": 154830
+    },
+    {
+      "epoch": 0.773252765362431,
+      "grad_norm": 0.09799888730049133,
+      "learning_rate": 6.819444792109941e-06,
+      "loss": 8.5674,
+      "step": 154840
+    },
+    {
+      "epoch": 0.7733027041873705,
+      "grad_norm": 0.0921357050538063,
+      "learning_rate": 6.817942877168389e-06,
+      "loss": 8.5727,
+      "step": 154850
+    },
+    {
+      "epoch": 0.7733526430123099,
+      "grad_norm": 0.0890340581536293,
+      "learning_rate": 6.8164409622268395e-06,
+      "loss": 8.5742,
+      "step": 154860
+    },
+    {
+      "epoch": 0.7734025818372494,
+      "grad_norm": 0.09299721568822861,
+      "learning_rate": 6.81493904728529e-06,
+      "loss": 8.6013,
+      "step": 154870
+    },
+    {
+      "epoch": 0.7734525206621888,
+      "grad_norm": 0.09847550094127655,
+      "learning_rate": 6.813437132343738e-06,
+      "loss": 8.5676,
+      "step": 154880
+    },
+    {
+      "epoch": 0.7735024594871283,
+      "grad_norm": 0.09080928564071655,
+      "learning_rate": 6.811935217402188e-06,
+      "loss": 8.5686,
+      "step": 154890
+    },
+    {
+      "epoch": 0.7735523983120677,
+      "grad_norm": 0.09545505046844482,
+      "learning_rate": 6.810433302460637e-06,
+      "loss": 8.5719,
+      "step": 154900
+    },
+    {
+      "epoch": 0.7736023371370072,
+      "grad_norm": 0.08688024431467056,
+      "learning_rate": 6.808931387519087e-06,
+      "loss": 8.5832,
+      "step": 154910
+    },
+    {
+      "epoch": 0.7736522759619466,
+      "grad_norm": 0.0897180363535881,
+      "learning_rate": 6.807429472577537e-06,
+      "loss": 8.5695,
+      "step": 154920
+    },
+    {
+      "epoch": 0.7737022147868861,
+      "grad_norm": 0.0968603864312172,
+      "learning_rate": 6.805927557635986e-06,
+      "loss": 8.5708,
+      "step": 154930
+    },
+    {
+      "epoch": 0.7737521536118255,
+      "grad_norm": 0.09534350782632828,
+      "learning_rate": 6.804425642694436e-06,
+      "loss": 8.5643,
+      "step": 154940
+    },
+    {
+      "epoch": 0.773802092436765,
+      "grad_norm": 0.09181123226881027,
+      "learning_rate": 6.802923727752885e-06,
+      "loss": 8.5733,
+      "step": 154950
+    },
+    {
+      "epoch": 0.7738520312617044,
+      "grad_norm": 0.08840048313140869,
+      "learning_rate": 6.8014218128113345e-06,
+      "loss": 8.56,
+      "step": 154960
+    },
+    {
+      "epoch": 0.7739019700866439,
+      "grad_norm": 0.09602469950914383,
+      "learning_rate": 6.799919897869785e-06,
+      "loss": 8.582,
+      "step": 154970
+    },
+    {
+      "epoch": 0.7739519089115833,
+      "grad_norm": 0.08890822529792786,
+      "learning_rate": 6.798417982928233e-06,
+      "loss": 8.5752,
+      "step": 154980
+    },
+    {
+      "epoch": 0.7740018477365228,
+      "grad_norm": 0.0951463058590889,
+      "learning_rate": 6.796916067986683e-06,
+      "loss": 8.5782,
+      "step": 154990
+    },
+    {
+      "epoch": 0.7740517865614622,
+      "grad_norm": 0.08842669427394867,
+      "learning_rate": 6.795414153045133e-06,
+      "loss": 8.6005,
+      "step": 155000
+    },
+    {
+      "epoch": 0.7741017253864017,
+      "grad_norm": 0.0843082070350647,
+      "learning_rate": 6.793912238103582e-06,
+      "loss": 8.5825,
+      "step": 155010
+    },
+    {
+      "epoch": 0.7741516642113411,
+      "grad_norm": 0.09507893025875092,
+      "learning_rate": 6.792410323162032e-06,
+      "loss": 8.5753,
+      "step": 155020
+    },
+    {
+      "epoch": 0.7742016030362806,
+      "grad_norm": 0.09141889959573746,
+      "learning_rate": 6.7909084082204815e-06,
+      "loss": 8.5755,
+      "step": 155030
+    },
+    {
+      "epoch": 0.77425154186122,
+      "grad_norm": 0.09243453294038773,
+      "learning_rate": 6.789406493278931e-06,
+      "loss": 8.5904,
+      "step": 155040
+    },
+    {
+      "epoch": 0.7743014806861594,
+      "grad_norm": 0.09155929833650589,
+      "learning_rate": 6.78790457833738e-06,
+      "loss": 8.5654,
+      "step": 155050
+    },
+    {
+      "epoch": 0.7743514195110989,
+      "grad_norm": 0.09300145506858826,
+      "learning_rate": 6.7864026633958295e-06,
+      "loss": 8.5482,
+      "step": 155060
+    },
+    {
+      "epoch": 0.7744013583360384,
+      "grad_norm": 0.09293139725923538,
+      "learning_rate": 6.78490074845428e-06,
+      "loss": 8.5743,
+      "step": 155070
+    },
+    {
+      "epoch": 0.7744512971609778,
+      "grad_norm": 0.08827711641788483,
+      "learning_rate": 6.783398833512729e-06,
+      "loss": 8.5664,
+      "step": 155080
+    },
+    {
+      "epoch": 0.7745012359859172,
+      "grad_norm": 0.08957238495349884,
+      "learning_rate": 6.781896918571178e-06,
+      "loss": 8.586,
+      "step": 155090
+    },
+    {
+      "epoch": 0.7745511748108567,
+      "grad_norm": 0.08923155814409256,
+      "learning_rate": 6.780395003629628e-06,
+      "loss": 8.5718,
+      "step": 155100
+    },
+    {
+      "epoch": 0.7746011136357962,
+      "grad_norm": 0.0950160101056099,
+      "learning_rate": 6.778893088688077e-06,
+      "loss": 8.585,
+      "step": 155110
+    },
+    {
+      "epoch": 0.7746510524607356,
+      "grad_norm": 0.09716098755598068,
+      "learning_rate": 6.777391173746527e-06,
+      "loss": 8.5609,
+      "step": 155120
+    },
+    {
+      "epoch": 0.774700991285675,
+      "grad_norm": 0.09193146228790283,
+      "learning_rate": 6.7758892588049765e-06,
+      "loss": 8.5826,
+      "step": 155130
+    },
+    {
+      "epoch": 0.7747509301106145,
+      "grad_norm": 0.08806554228067398,
+      "learning_rate": 6.774387343863426e-06,
+      "loss": 8.5824,
+      "step": 155140
+    },
+    {
+      "epoch": 0.774800868935554,
+      "grad_norm": 0.09546437114477158,
+      "learning_rate": 6.772885428921875e-06,
+      "loss": 8.5459,
+      "step": 155150
+    },
+    {
+      "epoch": 0.7748508077604934,
+      "grad_norm": 0.08918742090463638,
+      "learning_rate": 6.771383513980325e-06,
+      "loss": 8.5828,
+      "step": 155160
+    },
+    {
+      "epoch": 0.7749007465854328,
+      "grad_norm": 0.0883205235004425,
+      "learning_rate": 6.769881599038775e-06,
+      "loss": 8.5882,
+      "step": 155170
+    },
+    {
+      "epoch": 0.7749506854103723,
+      "grad_norm": 0.09442654252052307,
+      "learning_rate": 6.768379684097224e-06,
+      "loss": 8.5788,
+      "step": 155180
+    },
+    {
+      "epoch": 0.7750006242353118,
+      "grad_norm": 0.08732815831899643,
+      "learning_rate": 6.766877769155673e-06,
+      "loss": 8.5597,
+      "step": 155190
+    },
+    {
+      "epoch": 0.7750505630602512,
+      "grad_norm": 0.09169010072946548,
+      "learning_rate": 6.7653758542141235e-06,
+      "loss": 8.5722,
+      "step": 155200
+    },
+    {
+      "epoch": 0.7751005018851906,
+      "grad_norm": 0.09443099051713943,
+      "learning_rate": 6.763873939272573e-06,
+      "loss": 8.59,
+      "step": 155210
+    },
+    {
+      "epoch": 0.7751504407101301,
+      "grad_norm": 0.09280900657176971,
+      "learning_rate": 6.762372024331022e-06,
+      "loss": 8.5831,
+      "step": 155220
+    },
+    {
+      "epoch": 0.7752003795350696,
+      "grad_norm": 0.09292498975992203,
+      "learning_rate": 6.7608701093894715e-06,
+      "loss": 8.5737,
+      "step": 155230
+    },
+    {
+      "epoch": 0.775250318360009,
+      "grad_norm": 0.08915520459413528,
+      "learning_rate": 6.759368194447922e-06,
+      "loss": 8.5831,
+      "step": 155240
+    },
+    {
+      "epoch": 0.7753002571849484,
+      "grad_norm": 0.08977535367012024,
+      "learning_rate": 6.757866279506371e-06,
+      "loss": 8.5632,
+      "step": 155250
+    },
+    {
+      "epoch": 0.7753501960098879,
+      "grad_norm": 0.09135992079973221,
+      "learning_rate": 6.75636436456482e-06,
+      "loss": 8.5602,
+      "step": 155260
+    },
+    {
+      "epoch": 0.7754001348348273,
+      "grad_norm": 0.0918012335896492,
+      "learning_rate": 6.75486244962327e-06,
+      "loss": 8.5713,
+      "step": 155270
+    },
+    {
+      "epoch": 0.7754500736597668,
+      "grad_norm": 0.0947796106338501,
+      "learning_rate": 6.753360534681719e-06,
+      "loss": 8.5765,
+      "step": 155280
+    },
+    {
+      "epoch": 0.7755000124847062,
+      "grad_norm": 0.09258890151977539,
+      "learning_rate": 6.751858619740169e-06,
+      "loss": 8.5802,
+      "step": 155290
+    },
+    {
+      "epoch": 0.7755499513096457,
+      "grad_norm": 0.09028926491737366,
+      "learning_rate": 6.7503567047986185e-06,
+      "loss": 8.5838,
+      "step": 155300
+    },
+    {
+      "epoch": 0.7755998901345851,
+      "grad_norm": 0.09300950914621353,
+      "learning_rate": 6.748854789857068e-06,
+      "loss": 8.5698,
+      "step": 155310
+    },
+    {
+      "epoch": 0.7756498289595246,
+      "grad_norm": 0.09121584892272949,
+      "learning_rate": 6.747352874915518e-06,
+      "loss": 8.5736,
+      "step": 155320
+    },
+    {
+      "epoch": 0.775699767784464,
+      "grad_norm": 0.0916307121515274,
+      "learning_rate": 6.7458509599739665e-06,
+      "loss": 8.5664,
+      "step": 155330
+    },
+    {
+      "epoch": 0.7757497066094035,
+      "grad_norm": 0.09273859858512878,
+      "learning_rate": 6.744349045032417e-06,
+      "loss": 8.576,
+      "step": 155340
+    },
+    {
+      "epoch": 0.7757996454343429,
+      "grad_norm": 0.08911219984292984,
+      "learning_rate": 6.742847130090866e-06,
+      "loss": 8.5848,
+      "step": 155350
+    },
+    {
+      "epoch": 0.7758495842592824,
+      "grad_norm": 0.0911405086517334,
+      "learning_rate": 6.741345215149315e-06,
+      "loss": 8.5869,
+      "step": 155360
+    },
+    {
+      "epoch": 0.7758995230842218,
+      "grad_norm": 0.09100606292486191,
+      "learning_rate": 6.7398433002077655e-06,
+      "loss": 8.5476,
+      "step": 155370
+    },
+    {
+      "epoch": 0.7759494619091613,
+      "grad_norm": 0.09772035479545593,
+      "learning_rate": 6.738341385266214e-06,
+      "loss": 8.5863,
+      "step": 155380
+    },
+    {
+      "epoch": 0.7759994007341007,
+      "grad_norm": 0.08929920196533203,
+      "learning_rate": 6.736839470324664e-06,
+      "loss": 8.5799,
+      "step": 155390
+    },
+    {
+      "epoch": 0.7760493395590402,
+      "grad_norm": 0.0926336944103241,
+      "learning_rate": 6.735337555383114e-06,
+      "loss": 8.5804,
+      "step": 155400
+    },
+    {
+      "epoch": 0.7760992783839796,
+      "grad_norm": 0.09426475316286087,
+      "learning_rate": 6.733835640441563e-06,
+      "loss": 8.5662,
+      "step": 155410
+    },
+    {
+      "epoch": 0.776149217208919,
+      "grad_norm": 0.08994784206151962,
+      "learning_rate": 6.732333725500013e-06,
+      "loss": 8.568,
+      "step": 155420
+    },
+    {
+      "epoch": 0.7761991560338585,
+      "grad_norm": 0.093609519302845,
+      "learning_rate": 6.7308318105584615e-06,
+      "loss": 8.5902,
+      "step": 155430
+    },
+    {
+      "epoch": 0.776249094858798,
+      "grad_norm": 0.09885302931070328,
+      "learning_rate": 6.729329895616912e-06,
+      "loss": 8.5656,
+      "step": 155440
+    },
+    {
+      "epoch": 0.7762990336837374,
+      "grad_norm": 0.08882102370262146,
+      "learning_rate": 6.727827980675362e-06,
+      "loss": 8.5813,
+      "step": 155450
+    },
+    {
+      "epoch": 0.7763489725086768,
+      "grad_norm": 0.09155222773551941,
+      "learning_rate": 6.72632606573381e-06,
+      "loss": 8.5743,
+      "step": 155460
+    },
+    {
+      "epoch": 0.7763989113336163,
+      "grad_norm": 0.09071797877550125,
+      "learning_rate": 6.7248241507922605e-06,
+      "loss": 8.5807,
+      "step": 155470
+    },
+    {
+      "epoch": 0.7764488501585558,
+      "grad_norm": 0.09299448877573013,
+      "learning_rate": 6.72332223585071e-06,
+      "loss": 8.555,
+      "step": 155480
+    },
+    {
+      "epoch": 0.7764987889834952,
+      "grad_norm": 0.09560713171958923,
+      "learning_rate": 6.721820320909159e-06,
+      "loss": 8.5618,
+      "step": 155490
+    },
+    {
+      "epoch": 0.7765487278084346,
+      "grad_norm": 0.08749803900718689,
+      "learning_rate": 6.720318405967609e-06,
+      "loss": 8.5805,
+      "step": 155500
+    },
+    {
+      "epoch": 0.7765986666333741,
+      "grad_norm": 0.09022912383079529,
+      "learning_rate": 6.718816491026058e-06,
+      "loss": 8.5955,
+      "step": 155510
+    },
+    {
+      "epoch": 0.7766486054583136,
+      "grad_norm": 0.09331797063350677,
+      "learning_rate": 6.717314576084508e-06,
+      "loss": 8.5657,
+      "step": 155520
+    },
+    {
+      "epoch": 0.776698544283253,
+      "grad_norm": 0.09521540254354477,
+      "learning_rate": 6.715812661142957e-06,
+      "loss": 8.5662,
+      "step": 155530
+    },
+    {
+      "epoch": 0.7767484831081924,
+      "grad_norm": 0.09175549447536469,
+      "learning_rate": 6.714310746201407e-06,
+      "loss": 8.5795,
+      "step": 155540
+    },
+    {
+      "epoch": 0.7767984219331319,
+      "grad_norm": 0.09057381004095078,
+      "learning_rate": 6.712808831259857e-06,
+      "loss": 8.5605,
+      "step": 155550
+    },
+    {
+      "epoch": 0.7768483607580714,
+      "grad_norm": 0.09255360811948776,
+      "learning_rate": 6.711306916318306e-06,
+      "loss": 8.5666,
+      "step": 155560
+    },
+    {
+      "epoch": 0.7768982995830108,
+      "grad_norm": 0.08889876306056976,
+      "learning_rate": 6.7098050013767555e-06,
+      "loss": 8.5632,
+      "step": 155570
+    },
+    {
+      "epoch": 0.7769482384079502,
+      "grad_norm": 0.09821149706840515,
+      "learning_rate": 6.708303086435205e-06,
+      "loss": 8.5697,
+      "step": 155580
+    },
+    {
+      "epoch": 0.7769981772328897,
+      "grad_norm": 0.09199816733598709,
+      "learning_rate": 6.706801171493654e-06,
+      "loss": 8.5783,
+      "step": 155590
+    },
+    {
+      "epoch": 0.7770481160578292,
+      "grad_norm": 0.08835037052631378,
+      "learning_rate": 6.705299256552104e-06,
+      "loss": 8.5675,
+      "step": 155600
+    },
+    {
+      "epoch": 0.7770980548827686,
+      "grad_norm": 0.09240995347499847,
+      "learning_rate": 6.703797341610554e-06,
+      "loss": 8.586,
+      "step": 155610
+    },
+    {
+      "epoch": 0.777147993707708,
+      "grad_norm": 0.08770235627889633,
+      "learning_rate": 6.702295426669003e-06,
+      "loss": 8.5867,
+      "step": 155620
+    },
+    {
+      "epoch": 0.7771979325326475,
+      "grad_norm": 0.0896248146891594,
+      "learning_rate": 6.700793511727452e-06,
+      "loss": 8.5803,
+      "step": 155630
+    },
+    {
+      "epoch": 0.777247871357587,
+      "grad_norm": 0.09114081412553787,
+      "learning_rate": 6.6992915967859025e-06,
+      "loss": 8.5682,
+      "step": 155640
+    },
+    {
+      "epoch": 0.7772978101825264,
+      "grad_norm": 0.08897022902965546,
+      "learning_rate": 6.697789681844352e-06,
+      "loss": 8.589,
+      "step": 155650
+    },
+    {
+      "epoch": 0.7773477490074658,
+      "grad_norm": 0.09216853976249695,
+      "learning_rate": 6.696287766902801e-06,
+      "loss": 8.5517,
+      "step": 155660
+    },
+    {
+      "epoch": 0.7773976878324053,
+      "grad_norm": 0.0946219190955162,
+      "learning_rate": 6.6947858519612505e-06,
+      "loss": 8.5782,
+      "step": 155670
+    },
+    {
+      "epoch": 0.7774476266573448,
+      "grad_norm": 0.09455423057079315,
+      "learning_rate": 6.6932839370197e-06,
+      "loss": 8.5608,
+      "step": 155680
+    },
+    {
+      "epoch": 0.7774975654822842,
+      "grad_norm": 0.0957740917801857,
+      "learning_rate": 6.69178202207815e-06,
+      "loss": 8.5761,
+      "step": 155690
+    },
+    {
+      "epoch": 0.7775475043072236,
+      "grad_norm": 0.09181936085224152,
+      "learning_rate": 6.690280107136599e-06,
+      "loss": 8.582,
+      "step": 155700
+    },
+    {
+      "epoch": 0.7775974431321631,
+      "grad_norm": 0.09328048676252365,
+      "learning_rate": 6.688778192195049e-06,
+      "loss": 8.5598,
+      "step": 155710
+    },
+    {
+      "epoch": 0.7776473819571026,
+      "grad_norm": 0.08828567713499069,
+      "learning_rate": 6.687276277253499e-06,
+      "loss": 8.5677,
+      "step": 155720
+    },
+    {
+      "epoch": 0.777697320782042,
+      "grad_norm": 0.0937914326786995,
+      "learning_rate": 6.685774362311947e-06,
+      "loss": 8.5704,
+      "step": 155730
+    },
+    {
+      "epoch": 0.7777472596069814,
+      "grad_norm": 0.09564419835805893,
+      "learning_rate": 6.6842724473703976e-06,
+      "loss": 8.5762,
+      "step": 155740
+    },
+    {
+      "epoch": 0.7777971984319209,
+      "grad_norm": 0.08568717539310455,
+      "learning_rate": 6.682770532428847e-06,
+      "loss": 8.558,
+      "step": 155750
+    },
+    {
+      "epoch": 0.7778471372568604,
+      "grad_norm": 0.09092507511377335,
+      "learning_rate": 6.681268617487296e-06,
+      "loss": 8.5532,
+      "step": 155760
+    },
+    {
+      "epoch": 0.7778970760817998,
+      "grad_norm": 0.09010069072246552,
+      "learning_rate": 6.679766702545746e-06,
+      "loss": 8.5674,
+      "step": 155770
+    },
+    {
+      "epoch": 0.7779470149067392,
+      "grad_norm": 0.0927618071436882,
+      "learning_rate": 6.678264787604195e-06,
+      "loss": 8.5879,
+      "step": 155780
+    },
+    {
+      "epoch": 0.7779969537316787,
+      "grad_norm": 0.09050939232110977,
+      "learning_rate": 6.676762872662645e-06,
+      "loss": 8.5588,
+      "step": 155790
+    },
+    {
+      "epoch": 0.7780468925566182,
+      "grad_norm": 0.10684484988451004,
+      "learning_rate": 6.675260957721095e-06,
+      "loss": 8.5721,
+      "step": 155800
+    },
+    {
+      "epoch": 0.7780968313815576,
+      "grad_norm": 0.08730565011501312,
+      "learning_rate": 6.673759042779544e-06,
+      "loss": 8.5582,
+      "step": 155810
+    },
+    {
+      "epoch": 0.778146770206497,
+      "grad_norm": 0.089006207883358,
+      "learning_rate": 6.672257127837994e-06,
+      "loss": 8.5687,
+      "step": 155820
+    },
+    {
+      "epoch": 0.7781967090314365,
+      "grad_norm": 0.09139891713857651,
+      "learning_rate": 6.670755212896442e-06,
+      "loss": 8.5781,
+      "step": 155830
+    },
+    {
+      "epoch": 0.778246647856376,
+      "grad_norm": 0.09247538447380066,
+      "learning_rate": 6.6692532979548926e-06,
+      "loss": 8.578,
+      "step": 155840
+    },
+    {
+      "epoch": 0.7782965866813154,
+      "grad_norm": 0.09350184351205826,
+      "learning_rate": 6.667751383013343e-06,
+      "loss": 8.5866,
+      "step": 155850
+    },
+    {
+      "epoch": 0.7783465255062548,
+      "grad_norm": 0.08985615521669388,
+      "learning_rate": 6.666249468071791e-06,
+      "loss": 8.5793,
+      "step": 155860
+    },
+    {
+      "epoch": 0.7783964643311942,
+      "grad_norm": 0.08648446202278137,
+      "learning_rate": 6.664747553130241e-06,
+      "loss": 8.5698,
+      "step": 155870
+    },
+    {
+      "epoch": 0.7784464031561338,
+      "grad_norm": 0.0879853144288063,
+      "learning_rate": 6.663245638188691e-06,
+      "loss": 8.5803,
+      "step": 155880
+    },
+    {
+      "epoch": 0.7784963419810732,
+      "grad_norm": 0.09268853813409805,
+      "learning_rate": 6.66174372324714e-06,
+      "loss": 8.5734,
+      "step": 155890
+    },
+    {
+      "epoch": 0.7785462808060126,
+      "grad_norm": 0.09160283207893372,
+      "learning_rate": 6.66024180830559e-06,
+      "loss": 8.5861,
+      "step": 155900
+    },
+    {
+      "epoch": 0.778596219630952,
+      "grad_norm": 0.09253548830747604,
+      "learning_rate": 6.658739893364039e-06,
+      "loss": 8.5787,
+      "step": 155910
+    },
+    {
+      "epoch": 0.7786461584558916,
+      "grad_norm": 0.09596210718154907,
+      "learning_rate": 6.657237978422489e-06,
+      "loss": 8.5661,
+      "step": 155920
+    },
+    {
+      "epoch": 0.778696097280831,
+      "grad_norm": 0.08908601105213165,
+      "learning_rate": 6.655736063480938e-06,
+      "loss": 8.5649,
+      "step": 155930
+    },
+    {
+      "epoch": 0.7787460361057704,
+      "grad_norm": 0.08889824151992798,
+      "learning_rate": 6.6542341485393876e-06,
+      "loss": 8.55,
+      "step": 155940
+    },
+    {
+      "epoch": 0.7787959749307098,
+      "grad_norm": 0.09235508739948273,
+      "learning_rate": 6.652732233597838e-06,
+      "loss": 8.581,
+      "step": 155950
+    },
+    {
+      "epoch": 0.7788459137556494,
+      "grad_norm": 0.09276876598596573,
+      "learning_rate": 6.651230318656287e-06,
+      "loss": 8.5593,
+      "step": 155960
+    },
+    {
+      "epoch": 0.7788958525805888,
+      "grad_norm": 0.08646731078624725,
+      "learning_rate": 6.649728403714736e-06,
+      "loss": 8.5816,
+      "step": 155970
+    },
+    {
+      "epoch": 0.7789457914055282,
+      "grad_norm": 0.08740647882223129,
+      "learning_rate": 6.648226488773186e-06,
+      "loss": 8.5732,
+      "step": 155980
+    },
+    {
+      "epoch": 0.7789957302304676,
+      "grad_norm": 0.09212644398212433,
+      "learning_rate": 6.646724573831635e-06,
+      "loss": 8.5731,
+      "step": 155990
+    },
+    {
+      "epoch": 0.7790456690554072,
+      "grad_norm": 0.0925680547952652,
+      "learning_rate": 6.645222658890085e-06,
+      "loss": 8.5776,
+      "step": 156000
+    },
+    {
+      "epoch": 0.7790956078803466,
+      "grad_norm": 0.08844732493162155,
+      "learning_rate": 6.643720743948535e-06,
+      "loss": 8.5941,
+      "step": 156010
+    },
+    {
+      "epoch": 0.779145546705286,
+      "grad_norm": 0.08517491817474365,
+      "learning_rate": 6.642218829006984e-06,
+      "loss": 8.5737,
+      "step": 156020
+    },
+    {
+      "epoch": 0.7791954855302254,
+      "grad_norm": 0.09461086988449097,
+      "learning_rate": 6.640716914065433e-06,
+      "loss": 8.5723,
+      "step": 156030
+    },
+    {
+      "epoch": 0.779245424355165,
+      "grad_norm": 0.09351557493209839,
+      "learning_rate": 6.6392149991238834e-06,
+      "loss": 8.5885,
+      "step": 156040
+    },
+    {
+      "epoch": 0.7792953631801044,
+      "grad_norm": 0.09356269985437393,
+      "learning_rate": 6.637713084182333e-06,
+      "loss": 8.5521,
+      "step": 156050
+    },
+    {
+      "epoch": 0.7793453020050438,
+      "grad_norm": 0.09160886704921722,
+      "learning_rate": 6.636211169240782e-06,
+      "loss": 8.5711,
+      "step": 156060
+    },
+    {
+      "epoch": 0.7793952408299832,
+      "grad_norm": 0.09123864024877548,
+      "learning_rate": 6.634709254299231e-06,
+      "loss": 8.5658,
+      "step": 156070
+    },
+    {
+      "epoch": 0.7794451796549228,
+      "grad_norm": 0.08630036562681198,
+      "learning_rate": 6.633207339357681e-06,
+      "loss": 8.5817,
+      "step": 156080
+    },
+    {
+      "epoch": 0.7794951184798622,
+      "grad_norm": 0.08908306062221527,
+      "learning_rate": 6.631705424416131e-06,
+      "loss": 8.5747,
+      "step": 156090
+    },
+    {
+      "epoch": 0.7795450573048016,
+      "grad_norm": 0.0923098474740982,
+      "learning_rate": 6.63020350947458e-06,
+      "loss": 8.5524,
+      "step": 156100
+    },
+    {
+      "epoch": 0.779594996129741,
+      "grad_norm": 0.09396307915449142,
+      "learning_rate": 6.62870159453303e-06,
+      "loss": 8.5752,
+      "step": 156110
+    },
+    {
+      "epoch": 0.7796449349546806,
+      "grad_norm": 0.09391772747039795,
+      "learning_rate": 6.62719967959148e-06,
+      "loss": 8.5701,
+      "step": 156120
+    },
+    {
+      "epoch": 0.77969487377962,
+      "grad_norm": 0.0857236236333847,
+      "learning_rate": 6.625697764649928e-06,
+      "loss": 8.5592,
+      "step": 156130
+    },
+    {
+      "epoch": 0.7797448126045594,
+      "grad_norm": 0.09074844419956207,
+      "learning_rate": 6.6241958497083784e-06,
+      "loss": 8.5821,
+      "step": 156140
+    },
+    {
+      "epoch": 0.7797947514294988,
+      "grad_norm": 0.08897120505571365,
+      "learning_rate": 6.622693934766828e-06,
+      "loss": 8.5788,
+      "step": 156150
+    },
+    {
+      "epoch": 0.7798446902544384,
+      "grad_norm": 0.08881950378417969,
+      "learning_rate": 6.621192019825277e-06,
+      "loss": 8.572,
+      "step": 156160
+    },
+    {
+      "epoch": 0.7798946290793778,
+      "grad_norm": 0.09620756655931473,
+      "learning_rate": 6.619690104883727e-06,
+      "loss": 8.5728,
+      "step": 156170
+    },
+    {
+      "epoch": 0.7799445679043172,
+      "grad_norm": 0.08694388717412949,
+      "learning_rate": 6.618188189942176e-06,
+      "loss": 8.5895,
+      "step": 156180
+    },
+    {
+      "epoch": 0.7799945067292566,
+      "grad_norm": 0.08693565428256989,
+      "learning_rate": 6.616686275000626e-06,
+      "loss": 8.5849,
+      "step": 156190
+    },
+    {
+      "epoch": 0.7800444455541962,
+      "grad_norm": 0.08743081241846085,
+      "learning_rate": 6.615184360059076e-06,
+      "loss": 8.579,
+      "step": 156200
+    },
+    {
+      "epoch": 0.7800943843791356,
+      "grad_norm": 0.09228560328483582,
+      "learning_rate": 6.613682445117525e-06,
+      "loss": 8.5556,
+      "step": 156210
+    },
+    {
+      "epoch": 0.780144323204075,
+      "grad_norm": 0.09084375947713852,
+      "learning_rate": 6.612180530175975e-06,
+      "loss": 8.5772,
+      "step": 156220
+    },
+    {
+      "epoch": 0.7801942620290144,
+      "grad_norm": 0.08933619409799576,
+      "learning_rate": 6.610678615234423e-06,
+      "loss": 8.5722,
+      "step": 156230
+    },
+    {
+      "epoch": 0.780244200853954,
+      "grad_norm": 0.09434707462787628,
+      "learning_rate": 6.6091767002928734e-06,
+      "loss": 8.5676,
+      "step": 156240
+    },
+    {
+      "epoch": 0.7802941396788934,
+      "grad_norm": 0.08935230225324631,
+      "learning_rate": 6.607674785351324e-06,
+      "loss": 8.5802,
+      "step": 156250
+    },
+    {
+      "epoch": 0.7803440785038328,
+      "grad_norm": 0.09848465025424957,
+      "learning_rate": 6.606172870409772e-06,
+      "loss": 8.555,
+      "step": 156260
+    },
+    {
+      "epoch": 0.7803940173287722,
+      "grad_norm": 0.09197591245174408,
+      "learning_rate": 6.604670955468222e-06,
+      "loss": 8.5815,
+      "step": 156270
+    },
+    {
+      "epoch": 0.7804439561537116,
+      "grad_norm": 0.0875692144036293,
+      "learning_rate": 6.603169040526672e-06,
+      "loss": 8.5706,
+      "step": 156280
+    },
+    {
+      "epoch": 0.7804938949786512,
+      "grad_norm": 0.09515390545129776,
+      "learning_rate": 6.601667125585121e-06,
+      "loss": 8.5762,
+      "step": 156290
+    },
+    {
+      "epoch": 0.7805438338035906,
+      "grad_norm": 0.0905672088265419,
+      "learning_rate": 6.600165210643571e-06,
+      "loss": 8.5627,
+      "step": 156300
+    },
+    {
+      "epoch": 0.78059377262853,
+      "grad_norm": 0.08882583677768707,
+      "learning_rate": 6.59866329570202e-06,
+      "loss": 8.5586,
+      "step": 156310
+    },
+    {
+      "epoch": 0.7806437114534694,
+      "grad_norm": 0.08932114392518997,
+      "learning_rate": 6.59716138076047e-06,
+      "loss": 8.5591,
+      "step": 156320
+    },
+    {
+      "epoch": 0.780693650278409,
+      "grad_norm": 0.09061101078987122,
+      "learning_rate": 6.59565946581892e-06,
+      "loss": 8.5806,
+      "step": 156330
+    },
+    {
+      "epoch": 0.7807435891033484,
+      "grad_norm": 0.08474583178758621,
+      "learning_rate": 6.5941575508773684e-06,
+      "loss": 8.5841,
+      "step": 156340
+    },
+    {
+      "epoch": 0.7807935279282878,
+      "grad_norm": 0.09312687814235687,
+      "learning_rate": 6.592655635935819e-06,
+      "loss": 8.5691,
+      "step": 156350
+    },
+    {
+      "epoch": 0.7808434667532272,
+      "grad_norm": 0.09034864604473114,
+      "learning_rate": 6.591153720994268e-06,
+      "loss": 8.5726,
+      "step": 156360
+    },
+    {
+      "epoch": 0.7808934055781668,
+      "grad_norm": 0.09045044332742691,
+      "learning_rate": 6.589651806052717e-06,
+      "loss": 8.5599,
+      "step": 156370
+    },
+    {
+      "epoch": 0.7809433444031062,
+      "grad_norm": 0.09198865294456482,
+      "learning_rate": 6.5881498911111675e-06,
+      "loss": 8.5653,
+      "step": 156380
+    },
+    {
+      "epoch": 0.7809932832280456,
+      "grad_norm": 0.08959676325321198,
+      "learning_rate": 6.586647976169616e-06,
+      "loss": 8.5642,
+      "step": 156390
+    },
+    {
+      "epoch": 0.781043222052985,
+      "grad_norm": 0.09269116818904877,
+      "learning_rate": 6.585146061228066e-06,
+      "loss": 8.559,
+      "step": 156400
+    },
+    {
+      "epoch": 0.7810931608779246,
+      "grad_norm": 0.08859947323799133,
+      "learning_rate": 6.5836441462865155e-06,
+      "loss": 8.5707,
+      "step": 156410
+    },
+    {
+      "epoch": 0.781143099702864,
+      "grad_norm": 0.08951582759618759,
+      "learning_rate": 6.582142231344965e-06,
+      "loss": 8.5724,
+      "step": 156420
+    },
+    {
+      "epoch": 0.7811930385278034,
+      "grad_norm": 0.08770976960659027,
+      "learning_rate": 6.580640316403415e-06,
+      "loss": 8.5788,
+      "step": 156430
+    },
+    {
+      "epoch": 0.7812429773527428,
+      "grad_norm": 0.09240131080150604,
+      "learning_rate": 6.579138401461864e-06,
+      "loss": 8.5691,
+      "step": 156440
+    },
+    {
+      "epoch": 0.7812929161776824,
+      "grad_norm": 0.08868956565856934,
+      "learning_rate": 6.577636486520314e-06,
+      "loss": 8.5805,
+      "step": 156450
+    },
+    {
+      "epoch": 0.7813428550026218,
+      "grad_norm": 0.09283566474914551,
+      "learning_rate": 6.576134571578763e-06,
+      "loss": 8.5591,
+      "step": 156460
+    },
+    {
+      "epoch": 0.7813927938275612,
+      "grad_norm": 0.09575477242469788,
+      "learning_rate": 6.574632656637212e-06,
+      "loss": 8.562,
+      "step": 156470
+    },
+    {
+      "epoch": 0.7814427326525006,
+      "grad_norm": 0.09500584006309509,
+      "learning_rate": 6.5731307416956625e-06,
+      "loss": 8.5661,
+      "step": 156480
+    },
+    {
+      "epoch": 0.7814926714774402,
+      "grad_norm": 0.0926256999373436,
+      "learning_rate": 6.571628826754112e-06,
+      "loss": 8.5676,
+      "step": 156490
+    },
+    {
+      "epoch": 0.7815426103023796,
+      "grad_norm": 0.09180507063865662,
+      "learning_rate": 6.570126911812561e-06,
+      "loss": 8.5845,
+      "step": 156500
+    },
+    {
+      "epoch": 0.781592549127319,
+      "grad_norm": 0.08600914478302002,
+      "learning_rate": 6.5686249968710105e-06,
+      "loss": 8.5769,
+      "step": 156510
+    },
+    {
+      "epoch": 0.7816424879522584,
+      "grad_norm": 0.0883520096540451,
+      "learning_rate": 6.567123081929461e-06,
+      "loss": 8.5766,
+      "step": 156520
+    },
+    {
+      "epoch": 0.781692426777198,
+      "grad_norm": 0.0899481549859047,
+      "learning_rate": 6.56562116698791e-06,
+      "loss": 8.5516,
+      "step": 156530
+    },
+    {
+      "epoch": 0.7817423656021374,
+      "grad_norm": 0.09373979270458221,
+      "learning_rate": 6.564119252046359e-06,
+      "loss": 8.5745,
+      "step": 156540
+    },
+    {
+      "epoch": 0.7817923044270768,
+      "grad_norm": 0.08799232542514801,
+      "learning_rate": 6.562617337104809e-06,
+      "loss": 8.5713,
+      "step": 156550
+    },
+    {
+      "epoch": 0.7818422432520162,
+      "grad_norm": 0.09024485945701599,
+      "learning_rate": 6.561115422163258e-06,
+      "loss": 8.5852,
+      "step": 156560
+    },
+    {
+      "epoch": 0.7818921820769558,
+      "grad_norm": 0.0885784775018692,
+      "learning_rate": 6.559613507221708e-06,
+      "loss": 8.5742,
+      "step": 156570
+    },
+    {
+      "epoch": 0.7819421209018952,
+      "grad_norm": 0.09604980051517487,
+      "learning_rate": 6.5581115922801575e-06,
+      "loss": 8.5856,
+      "step": 156580
+    },
+    {
+      "epoch": 0.7819920597268346,
+      "grad_norm": 0.08925112336874008,
+      "learning_rate": 6.556609677338607e-06,
+      "loss": 8.577,
+      "step": 156590
+    },
+    {
+      "epoch": 0.782041998551774,
+      "grad_norm": 0.08953628689050674,
+      "learning_rate": 6.555107762397057e-06,
+      "loss": 8.5945,
+      "step": 156600
+    },
+    {
+      "epoch": 0.7820919373767136,
+      "grad_norm": 0.08811146020889282,
+      "learning_rate": 6.5536058474555055e-06,
+      "loss": 8.5769,
+      "step": 156610
+    },
+    {
+      "epoch": 0.782141876201653,
+      "grad_norm": 0.08883983641862869,
+      "learning_rate": 6.552103932513956e-06,
+      "loss": 8.5647,
+      "step": 156620
+    },
+    {
+      "epoch": 0.7821918150265924,
+      "grad_norm": 0.09047286212444305,
+      "learning_rate": 6.550602017572405e-06,
+      "loss": 8.5689,
+      "step": 156630
+    },
+    {
+      "epoch": 0.7822417538515318,
+      "grad_norm": 0.091248519718647,
+      "learning_rate": 6.549100102630854e-06,
+      "loss": 8.5746,
+      "step": 156640
+    },
+    {
+      "epoch": 0.7822916926764714,
+      "grad_norm": 0.0949409082531929,
+      "learning_rate": 6.5475981876893045e-06,
+      "loss": 8.5657,
+      "step": 156650
+    },
+    {
+      "epoch": 0.7823416315014108,
+      "grad_norm": 0.09160256385803223,
+      "learning_rate": 6.546096272747753e-06,
+      "loss": 8.5602,
+      "step": 156660
+    },
+    {
+      "epoch": 0.7823915703263502,
+      "grad_norm": 0.09412173181772232,
+      "learning_rate": 6.544594357806203e-06,
+      "loss": 8.5698,
+      "step": 156670
+    },
+    {
+      "epoch": 0.7824415091512896,
+      "grad_norm": 0.08734364062547684,
+      "learning_rate": 6.543092442864653e-06,
+      "loss": 8.5699,
+      "step": 156680
+    },
+    {
+      "epoch": 0.7824914479762292,
+      "grad_norm": 0.08681227266788483,
+      "learning_rate": 6.541590527923102e-06,
+      "loss": 8.5734,
+      "step": 156690
+    },
+    {
+      "epoch": 0.7825413868011686,
+      "grad_norm": 0.09137750416994095,
+      "learning_rate": 6.540088612981552e-06,
+      "loss": 8.5507,
+      "step": 156700
+    },
+    {
+      "epoch": 0.782591325626108,
+      "grad_norm": 0.08570120483636856,
+      "learning_rate": 6.5385866980400005e-06,
+      "loss": 8.5657,
+      "step": 156710
+    },
+    {
+      "epoch": 0.7826412644510474,
+      "grad_norm": 0.0939328595995903,
+      "learning_rate": 6.537084783098451e-06,
+      "loss": 8.5711,
+      "step": 156720
+    },
+    {
+      "epoch": 0.782691203275987,
+      "grad_norm": 0.08831031620502472,
+      "learning_rate": 6.535582868156901e-06,
+      "loss": 8.5667,
+      "step": 156730
+    },
+    {
+      "epoch": 0.7827411421009264,
+      "grad_norm": 0.09072265774011612,
+      "learning_rate": 6.534080953215349e-06,
+      "loss": 8.5528,
+      "step": 156740
+    },
+    {
+      "epoch": 0.7827910809258658,
+      "grad_norm": 0.09154432266950607,
+      "learning_rate": 6.5325790382737995e-06,
+      "loss": 8.5787,
+      "step": 156750
+    },
+    {
+      "epoch": 0.7828410197508052,
+      "grad_norm": 0.0947096049785614,
+      "learning_rate": 6.531077123332249e-06,
+      "loss": 8.5638,
+      "step": 156760
+    },
+    {
+      "epoch": 0.7828909585757448,
+      "grad_norm": 0.09594990313053131,
+      "learning_rate": 6.529575208390698e-06,
+      "loss": 8.5761,
+      "step": 156770
+    },
+    {
+      "epoch": 0.7829408974006842,
+      "grad_norm": 0.0905158668756485,
+      "learning_rate": 6.528073293449148e-06,
+      "loss": 8.558,
+      "step": 156780
+    },
+    {
+      "epoch": 0.7829908362256236,
+      "grad_norm": 0.0911213606595993,
+      "learning_rate": 6.526571378507597e-06,
+      "loss": 8.5628,
+      "step": 156790
+    },
+    {
+      "epoch": 0.783040775050563,
+      "grad_norm": 0.08714745938777924,
+      "learning_rate": 6.525069463566047e-06,
+      "loss": 8.5658,
+      "step": 156800
+    },
+    {
+      "epoch": 0.7830907138755026,
+      "grad_norm": 0.09068936109542847,
+      "learning_rate": 6.523567548624496e-06,
+      "loss": 8.5696,
+      "step": 156810
+    },
+    {
+      "epoch": 0.783140652700442,
+      "grad_norm": 0.09490951150655746,
+      "learning_rate": 6.522065633682946e-06,
+      "loss": 8.5601,
+      "step": 156820
+    },
+    {
+      "epoch": 0.7831905915253814,
+      "grad_norm": 0.08792884647846222,
+      "learning_rate": 6.520563718741396e-06,
+      "loss": 8.5703,
+      "step": 156830
+    },
+    {
+      "epoch": 0.7832405303503208,
+      "grad_norm": 0.08713288605213165,
+      "learning_rate": 6.519061803799845e-06,
+      "loss": 8.5737,
+      "step": 156840
+    },
+    {
+      "epoch": 0.7832904691752604,
+      "grad_norm": 0.0946747288107872,
+      "learning_rate": 6.5175598888582945e-06,
+      "loss": 8.5595,
+      "step": 156850
+    },
+    {
+      "epoch": 0.7833404080001998,
+      "grad_norm": 0.08821689337491989,
+      "learning_rate": 6.516057973916744e-06,
+      "loss": 8.5619,
+      "step": 156860
+    },
+    {
+      "epoch": 0.7833903468251392,
+      "grad_norm": 0.09473678469657898,
+      "learning_rate": 6.514556058975193e-06,
+      "loss": 8.5639,
+      "step": 156870
+    },
+    {
+      "epoch": 0.7834402856500786,
+      "grad_norm": 0.09480735659599304,
+      "learning_rate": 6.513054144033643e-06,
+      "loss": 8.5744,
+      "step": 156880
+    },
+    {
+      "epoch": 0.7834902244750181,
+      "grad_norm": 0.09333161264657974,
+      "learning_rate": 6.511552229092093e-06,
+      "loss": 8.5615,
+      "step": 156890
+    },
+    {
+      "epoch": 0.7835401632999576,
+      "grad_norm": 0.09537094086408615,
+      "learning_rate": 6.510050314150542e-06,
+      "loss": 8.5697,
+      "step": 156900
+    },
+    {
+      "epoch": 0.783590102124897,
+      "grad_norm": 0.09100158512592316,
+      "learning_rate": 6.508548399208991e-06,
+      "loss": 8.5796,
+      "step": 156910
+    },
+    {
+      "epoch": 0.7836400409498364,
+      "grad_norm": 0.0884363129734993,
+      "learning_rate": 6.5070464842674415e-06,
+      "loss": 8.5578,
+      "step": 156920
+    },
+    {
+      "epoch": 0.783689979774776,
+      "grad_norm": 0.09135287255048752,
+      "learning_rate": 6.505544569325891e-06,
+      "loss": 8.5509,
+      "step": 156930
+    },
+    {
+      "epoch": 0.7837399185997154,
+      "grad_norm": 0.09011730551719666,
+      "learning_rate": 6.50404265438434e-06,
+      "loss": 8.562,
+      "step": 156940
+    },
+    {
+      "epoch": 0.7837898574246548,
+      "grad_norm": 0.08869774639606476,
+      "learning_rate": 6.5025407394427895e-06,
+      "loss": 8.5689,
+      "step": 156950
+    },
+    {
+      "epoch": 0.7838397962495942,
+      "grad_norm": 0.09371919184923172,
+      "learning_rate": 6.501038824501239e-06,
+      "loss": 8.5672,
+      "step": 156960
+    },
+    {
+      "epoch": 0.7838897350745337,
+      "grad_norm": 0.0950884222984314,
+      "learning_rate": 6.499536909559689e-06,
+      "loss": 8.5585,
+      "step": 156970
+    },
+    {
+      "epoch": 0.7839396738994732,
+      "grad_norm": 0.09123662114143372,
+      "learning_rate": 6.498034994618138e-06,
+      "loss": 8.5601,
+      "step": 156980
+    },
+    {
+      "epoch": 0.7839896127244126,
+      "grad_norm": 0.0944250077009201,
+      "learning_rate": 6.496533079676588e-06,
+      "loss": 8.5766,
+      "step": 156990
+    },
+    {
+      "epoch": 0.784039551549352,
+      "grad_norm": 0.08941183984279633,
+      "learning_rate": 6.495031164735038e-06,
+      "loss": 8.5673,
+      "step": 157000
+    },
+    {
+      "epoch": 0.7840894903742915,
+      "grad_norm": 0.0950687900185585,
+      "learning_rate": 6.493529249793486e-06,
+      "loss": 8.5736,
+      "step": 157010
+    },
+    {
+      "epoch": 0.784139429199231,
+      "grad_norm": 0.09286126494407654,
+      "learning_rate": 6.4920273348519365e-06,
+      "loss": 8.5549,
+      "step": 157020
+    },
+    {
+      "epoch": 0.7841893680241704,
+      "grad_norm": 0.09739340096712112,
+      "learning_rate": 6.490525419910386e-06,
+      "loss": 8.5606,
+      "step": 157030
+    },
+    {
+      "epoch": 0.7842393068491098,
+      "grad_norm": 0.09119179099798203,
+      "learning_rate": 6.489023504968835e-06,
+      "loss": 8.5585,
+      "step": 157040
+    },
+    {
+      "epoch": 0.7842892456740493,
+      "grad_norm": 0.09187173843383789,
+      "learning_rate": 6.487521590027285e-06,
+      "loss": 8.5867,
+      "step": 157050
+    },
+    {
+      "epoch": 0.7843391844989888,
+      "grad_norm": 0.08924529701471329,
+      "learning_rate": 6.486019675085734e-06,
+      "loss": 8.5617,
+      "step": 157060
+    },
+    {
+      "epoch": 0.7843891233239282,
+      "grad_norm": 0.08665250241756439,
+      "learning_rate": 6.484517760144184e-06,
+      "loss": 8.5643,
+      "step": 157070
+    },
+    {
+      "epoch": 0.7844390621488676,
+      "grad_norm": 0.08721933513879776,
+      "learning_rate": 6.483015845202634e-06,
+      "loss": 8.5747,
+      "step": 157080
+    },
+    {
+      "epoch": 0.7844890009738071,
+      "grad_norm": 0.09293626993894577,
+      "learning_rate": 6.481513930261083e-06,
+      "loss": 8.5616,
+      "step": 157090
+    },
+    {
+      "epoch": 0.7845389397987466,
+      "grad_norm": 0.09339628368616104,
+      "learning_rate": 6.480012015319533e-06,
+      "loss": 8.5674,
+      "step": 157100
+    },
+    {
+      "epoch": 0.784588878623686,
+      "grad_norm": 0.09324491769075394,
+      "learning_rate": 6.478510100377981e-06,
+      "loss": 8.561,
+      "step": 157110
+    },
+    {
+      "epoch": 0.7846388174486254,
+      "grad_norm": 0.08938504010438919,
+      "learning_rate": 6.4770081854364315e-06,
+      "loss": 8.5575,
+      "step": 157120
+    },
+    {
+      "epoch": 0.7846887562735649,
+      "grad_norm": 0.09114962816238403,
+      "learning_rate": 6.475506270494882e-06,
+      "loss": 8.5692,
+      "step": 157130
+    },
+    {
+      "epoch": 0.7847386950985044,
+      "grad_norm": 0.0890054702758789,
+      "learning_rate": 6.47400435555333e-06,
+      "loss": 8.5646,
+      "step": 157140
+    },
+    {
+      "epoch": 0.7847886339234438,
+      "grad_norm": 0.09435494989156723,
+      "learning_rate": 6.47250244061178e-06,
+      "loss": 8.5797,
+      "step": 157150
+    },
+    {
+      "epoch": 0.7848385727483832,
+      "grad_norm": 0.09404738992452621,
+      "learning_rate": 6.47100052567023e-06,
+      "loss": 8.5541,
+      "step": 157160
+    },
+    {
+      "epoch": 0.7848885115733227,
+      "grad_norm": 0.09190955758094788,
+      "learning_rate": 6.469498610728679e-06,
+      "loss": 8.5693,
+      "step": 157170
+    },
+    {
+      "epoch": 0.7849384503982622,
+      "grad_norm": 0.08809108287096024,
+      "learning_rate": 6.467996695787129e-06,
+      "loss": 8.5717,
+      "step": 157180
+    },
+    {
+      "epoch": 0.7849883892232016,
+      "grad_norm": 0.0869806781411171,
+      "learning_rate": 6.466494780845578e-06,
+      "loss": 8.5623,
+      "step": 157190
+    },
+    {
+      "epoch": 0.785038328048141,
+      "grad_norm": 0.093222476541996,
+      "learning_rate": 6.464992865904028e-06,
+      "loss": 8.5687,
+      "step": 157200
+    },
+    {
+      "epoch": 0.7850882668730805,
+      "grad_norm": 0.08760460466146469,
+      "learning_rate": 6.463490950962477e-06,
+      "loss": 8.5537,
+      "step": 157210
+    },
+    {
+      "epoch": 0.78513820569802,
+      "grad_norm": 0.09468575567007065,
+      "learning_rate": 6.4619890360209265e-06,
+      "loss": 8.5678,
+      "step": 157220
+    },
+    {
+      "epoch": 0.7851881445229594,
+      "grad_norm": 0.08602230995893478,
+      "learning_rate": 6.460487121079377e-06,
+      "loss": 8.5842,
+      "step": 157230
+    },
+    {
+      "epoch": 0.7852380833478988,
+      "grad_norm": 0.093845434486866,
+      "learning_rate": 6.458985206137826e-06,
+      "loss": 8.5626,
+      "step": 157240
+    },
+    {
+      "epoch": 0.7852880221728382,
+      "grad_norm": 0.08826538175344467,
+      "learning_rate": 6.457483291196275e-06,
+      "loss": 8.5479,
+      "step": 157250
+    },
+    {
+      "epoch": 0.7853379609977778,
+      "grad_norm": 0.09430468082427979,
+      "learning_rate": 6.455981376254725e-06,
+      "loss": 8.5857,
+      "step": 157260
+    },
+    {
+      "epoch": 0.7853878998227172,
+      "grad_norm": 0.08807472139596939,
+      "learning_rate": 6.454479461313174e-06,
+      "loss": 8.5541,
+      "step": 157270
+    },
+    {
+      "epoch": 0.7854378386476566,
+      "grad_norm": 0.09787096083164215,
+      "learning_rate": 6.452977546371624e-06,
+      "loss": 8.5621,
+      "step": 157280
+    },
+    {
+      "epoch": 0.785487777472596,
+      "grad_norm": 0.09540273249149323,
+      "learning_rate": 6.4514756314300736e-06,
+      "loss": 8.5687,
+      "step": 157290
+    },
+    {
+      "epoch": 0.7855377162975355,
+      "grad_norm": 0.09057477116584778,
+      "learning_rate": 6.449973716488523e-06,
+      "loss": 8.5679,
+      "step": 157300
+    },
+    {
+      "epoch": 0.785587655122475,
+      "grad_norm": 0.08641599118709564,
+      "learning_rate": 6.448471801546972e-06,
+      "loss": 8.5723,
+      "step": 157310
+    },
+    {
+      "epoch": 0.7856375939474144,
+      "grad_norm": 0.0898127630352974,
+      "learning_rate": 6.446969886605422e-06,
+      "loss": 8.5729,
+      "step": 157320
+    },
+    {
+      "epoch": 0.7856875327723538,
+      "grad_norm": 0.09772838652133942,
+      "learning_rate": 6.445467971663872e-06,
+      "loss": 8.5604,
+      "step": 157330
+    },
+    {
+      "epoch": 0.7857374715972933,
+      "grad_norm": 0.09075041115283966,
+      "learning_rate": 6.443966056722321e-06,
+      "loss": 8.5621,
+      "step": 157340
+    },
+    {
+      "epoch": 0.7857874104222328,
+      "grad_norm": 0.09176818281412125,
+      "learning_rate": 6.44246414178077e-06,
+      "loss": 8.5765,
+      "step": 157350
+    },
+    {
+      "epoch": 0.7858373492471722,
+      "grad_norm": 0.09130357950925827,
+      "learning_rate": 6.44096222683922e-06,
+      "loss": 8.5659,
+      "step": 157360
+    },
+    {
+      "epoch": 0.7858872880721116,
+      "grad_norm": 0.08570962399244308,
+      "learning_rate": 6.43946031189767e-06,
+      "loss": 8.5623,
+      "step": 157370
+    },
+    {
+      "epoch": 0.7859372268970511,
+      "grad_norm": 0.08622989058494568,
+      "learning_rate": 6.437958396956119e-06,
+      "loss": 8.5748,
+      "step": 157380
+    },
+    {
+      "epoch": 0.7859871657219906,
+      "grad_norm": 0.09150289744138718,
+      "learning_rate": 6.4364564820145686e-06,
+      "loss": 8.5404,
+      "step": 157390
+    },
+    {
+      "epoch": 0.78603710454693,
+      "grad_norm": 0.09090328216552734,
+      "learning_rate": 6.434954567073019e-06,
+      "loss": 8.5627,
+      "step": 157400
+    },
+    {
+      "epoch": 0.7860870433718694,
+      "grad_norm": 0.09460791200399399,
+      "learning_rate": 6.433452652131468e-06,
+      "loss": 8.5751,
+      "step": 157410
+    },
+    {
+      "epoch": 0.7861369821968089,
+      "grad_norm": 0.09171260893344879,
+      "learning_rate": 6.431950737189917e-06,
+      "loss": 8.5494,
+      "step": 157420
+    },
+    {
+      "epoch": 0.7861869210217484,
+      "grad_norm": 0.0896756649017334,
+      "learning_rate": 6.430448822248367e-06,
+      "loss": 8.5658,
+      "step": 157430
+    },
+    {
+      "epoch": 0.7862368598466878,
+      "grad_norm": 0.08946090936660767,
+      "learning_rate": 6.428946907306816e-06,
+      "loss": 8.5715,
+      "step": 157440
+    },
+    {
+      "epoch": 0.7862867986716272,
+      "grad_norm": 0.08953197300434113,
+      "learning_rate": 6.427444992365266e-06,
+      "loss": 8.5723,
+      "step": 157450
+    },
+    {
+      "epoch": 0.7863367374965667,
+      "grad_norm": 0.08706686645746231,
+      "learning_rate": 6.425943077423716e-06,
+      "loss": 8.5687,
+      "step": 157460
+    },
+    {
+      "epoch": 0.7863866763215062,
+      "grad_norm": 0.09402570128440857,
+      "learning_rate": 6.424441162482165e-06,
+      "loss": 8.5646,
+      "step": 157470
+    },
+    {
+      "epoch": 0.7864366151464456,
+      "grad_norm": 0.09386756271123886,
+      "learning_rate": 6.422939247540615e-06,
+      "loss": 8.5686,
+      "step": 157480
+    },
+    {
+      "epoch": 0.786486553971385,
+      "grad_norm": 0.09243669360876083,
+      "learning_rate": 6.4214373325990636e-06,
+      "loss": 8.5747,
+      "step": 157490
+    },
+    {
+      "epoch": 0.7865364927963245,
+      "grad_norm": 0.08882700651884079,
+      "learning_rate": 6.419935417657514e-06,
+      "loss": 8.553,
+      "step": 157500
+    },
+    {
+      "epoch": 0.786586431621264,
+      "grad_norm": 0.091353639960289,
+      "learning_rate": 6.418433502715963e-06,
+      "loss": 8.5603,
+      "step": 157510
+    },
+    {
+      "epoch": 0.7866363704462034,
+      "grad_norm": 0.09125305712223053,
+      "learning_rate": 6.416931587774412e-06,
+      "loss": 8.5857,
+      "step": 157520
+    },
+    {
+      "epoch": 0.7866863092711428,
+      "grad_norm": 0.09622442722320557,
+      "learning_rate": 6.415429672832863e-06,
+      "loss": 8.554,
+      "step": 157530
+    },
+    {
+      "epoch": 0.7867362480960823,
+      "grad_norm": 0.09017577767372131,
+      "learning_rate": 6.413927757891311e-06,
+      "loss": 8.5527,
+      "step": 157540
+    },
+    {
+      "epoch": 0.7867861869210218,
+      "grad_norm": 0.09028837084770203,
+      "learning_rate": 6.412425842949761e-06,
+      "loss": 8.5776,
+      "step": 157550
+    },
+    {
+      "epoch": 0.7868361257459612,
+      "grad_norm": 0.0956842452287674,
+      "learning_rate": 6.4109239280082114e-06,
+      "loss": 8.5637,
+      "step": 157560
+    },
+    {
+      "epoch": 0.7868860645709006,
+      "grad_norm": 0.09367352724075317,
+      "learning_rate": 6.40942201306666e-06,
+      "loss": 8.5664,
+      "step": 157570
+    },
+    {
+      "epoch": 0.7869360033958401,
+      "grad_norm": 0.0918663740158081,
+      "learning_rate": 6.40792009812511e-06,
+      "loss": 8.5806,
+      "step": 157580
+    },
+    {
+      "epoch": 0.7869859422207796,
+      "grad_norm": 0.08775880932807922,
+      "learning_rate": 6.406418183183559e-06,
+      "loss": 8.5442,
+      "step": 157590
+    },
+    {
+      "epoch": 0.787035881045719,
+      "grad_norm": 0.09422892332077026,
+      "learning_rate": 6.404916268242009e-06,
+      "loss": 8.5819,
+      "step": 157600
+    },
+    {
+      "epoch": 0.7870858198706584,
+      "grad_norm": 0.0917881578207016,
+      "learning_rate": 6.403414353300459e-06,
+      "loss": 8.56,
+      "step": 157610
+    },
+    {
+      "epoch": 0.7871357586955979,
+      "grad_norm": 0.08879712224006653,
+      "learning_rate": 6.401912438358907e-06,
+      "loss": 8.5751,
+      "step": 157620
+    },
+    {
+      "epoch": 0.7871856975205374,
+      "grad_norm": 0.0947839617729187,
+      "learning_rate": 6.400410523417358e-06,
+      "loss": 8.5617,
+      "step": 157630
+    },
+    {
+      "epoch": 0.7872356363454768,
+      "grad_norm": 0.09528858214616776,
+      "learning_rate": 6.398908608475807e-06,
+      "loss": 8.5535,
+      "step": 157640
+    },
+    {
+      "epoch": 0.7872855751704162,
+      "grad_norm": 0.08856210112571716,
+      "learning_rate": 6.397406693534256e-06,
+      "loss": 8.5532,
+      "step": 157650
+    },
+    {
+      "epoch": 0.7873355139953557,
+      "grad_norm": 0.08869683742523193,
+      "learning_rate": 6.3959047785927064e-06,
+      "loss": 8.5521,
+      "step": 157660
+    },
+    {
+      "epoch": 0.7873854528202952,
+      "grad_norm": 0.09273868054151535,
+      "learning_rate": 6.394402863651155e-06,
+      "loss": 8.5656,
+      "step": 157670
+    },
+    {
+      "epoch": 0.7874353916452346,
+      "grad_norm": 0.09226199239492416,
+      "learning_rate": 6.392900948709605e-06,
+      "loss": 8.5631,
+      "step": 157680
+    },
+    {
+      "epoch": 0.787485330470174,
+      "grad_norm": 0.09002158790826797,
+      "learning_rate": 6.3913990337680544e-06,
+      "loss": 8.5679,
+      "step": 157690
+    },
+    {
+      "epoch": 0.7875352692951135,
+      "grad_norm": 0.09723229706287384,
+      "learning_rate": 6.389897118826504e-06,
+      "loss": 8.5763,
+      "step": 157700
+    },
+    {
+      "epoch": 0.787585208120053,
+      "grad_norm": 0.0909029170870781,
+      "learning_rate": 6.388395203884954e-06,
+      "loss": 8.5659,
+      "step": 157710
+    },
+    {
+      "epoch": 0.7876351469449924,
+      "grad_norm": 0.08461030572652817,
+      "learning_rate": 6.386893288943403e-06,
+      "loss": 8.5652,
+      "step": 157720
+    },
+    {
+      "epoch": 0.7876850857699318,
+      "grad_norm": 0.0874033197760582,
+      "learning_rate": 6.385391374001853e-06,
+      "loss": 8.553,
+      "step": 157730
+    },
+    {
+      "epoch": 0.7877350245948713,
+      "grad_norm": 0.09039891511201859,
+      "learning_rate": 6.383889459060302e-06,
+      "loss": 8.5708,
+      "step": 157740
+    },
+    {
+      "epoch": 0.7877849634198107,
+      "grad_norm": 0.09345263242721558,
+      "learning_rate": 6.382387544118751e-06,
+      "loss": 8.5643,
+      "step": 157750
+    },
+    {
+      "epoch": 0.7878349022447502,
+      "grad_norm": 0.08719782531261444,
+      "learning_rate": 6.3808856291772014e-06,
+      "loss": 8.5788,
+      "step": 157760
+    },
+    {
+      "epoch": 0.7878848410696896,
+      "grad_norm": 0.09255023300647736,
+      "learning_rate": 6.379383714235651e-06,
+      "loss": 8.5487,
+      "step": 157770
+    },
+    {
+      "epoch": 0.7879347798946291,
+      "grad_norm": 0.08840210735797882,
+      "learning_rate": 6.3778817992941e-06,
+      "loss": 8.5636,
+      "step": 157780
+    },
+    {
+      "epoch": 0.7879847187195685,
+      "grad_norm": 0.08876688033342361,
+      "learning_rate": 6.3763798843525494e-06,
+      "loss": 8.5884,
+      "step": 157790
+    },
+    {
+      "epoch": 0.788034657544508,
+      "grad_norm": 0.0919869989156723,
+      "learning_rate": 6.374877969411e-06,
+      "loss": 8.5666,
+      "step": 157800
+    },
+    {
+      "epoch": 0.7880845963694474,
+      "grad_norm": 0.08744458109140396,
+      "learning_rate": 6.373376054469449e-06,
+      "loss": 8.5712,
+      "step": 157810
+    },
+    {
+      "epoch": 0.7881345351943869,
+      "grad_norm": 0.0894647017121315,
+      "learning_rate": 6.371874139527898e-06,
+      "loss": 8.5758,
+      "step": 157820
+    },
+    {
+      "epoch": 0.7881844740193263,
+      "grad_norm": 0.08984792977571487,
+      "learning_rate": 6.370372224586348e-06,
+      "loss": 8.5658,
+      "step": 157830
+    },
+    {
+      "epoch": 0.7882344128442658,
+      "grad_norm": 0.08753778785467148,
+      "learning_rate": 6.368870309644797e-06,
+      "loss": 8.5606,
+      "step": 157840
+    },
+    {
+      "epoch": 0.7882843516692052,
+      "grad_norm": 0.09227602183818817,
+      "learning_rate": 6.367368394703247e-06,
+      "loss": 8.5656,
+      "step": 157850
+    },
+    {
+      "epoch": 0.7883342904941447,
+      "grad_norm": 0.09825395792722702,
+      "learning_rate": 6.3658664797616965e-06,
+      "loss": 8.5529,
+      "step": 157860
+    },
+    {
+      "epoch": 0.7883842293190841,
+      "grad_norm": 0.08933484554290771,
+      "learning_rate": 6.364364564820146e-06,
+      "loss": 8.5791,
+      "step": 157870
+    },
+    {
+      "epoch": 0.7884341681440236,
+      "grad_norm": 0.09111601114273071,
+      "learning_rate": 6.362862649878596e-06,
+      "loss": 8.5454,
+      "step": 157880
+    },
+    {
+      "epoch": 0.788484106968963,
+      "grad_norm": 0.09824613481760025,
+      "learning_rate": 6.3613607349370444e-06,
+      "loss": 8.5477,
+      "step": 157890
+    },
+    {
+      "epoch": 0.7885340457939025,
+      "grad_norm": 0.09107904136180878,
+      "learning_rate": 6.359858819995495e-06,
+      "loss": 8.5625,
+      "step": 157900
+    },
+    {
+      "epoch": 0.7885839846188419,
+      "grad_norm": 0.08801267296075821,
+      "learning_rate": 6.358356905053944e-06,
+      "loss": 8.5432,
+      "step": 157910
+    },
+    {
+      "epoch": 0.7886339234437814,
+      "grad_norm": 0.09051348268985748,
+      "learning_rate": 6.356854990112393e-06,
+      "loss": 8.5402,
+      "step": 157920
+    },
+    {
+      "epoch": 0.7886838622687208,
+      "grad_norm": 0.0885823667049408,
+      "learning_rate": 6.3553530751708435e-06,
+      "loss": 8.5557,
+      "step": 157930
+    },
+    {
+      "epoch": 0.7887338010936603,
+      "grad_norm": 0.09226511418819427,
+      "learning_rate": 6.353851160229292e-06,
+      "loss": 8.5563,
+      "step": 157940
+    },
+    {
+      "epoch": 0.7887837399185997,
+      "grad_norm": 0.09105234593153,
+      "learning_rate": 6.352349245287742e-06,
+      "loss": 8.572,
+      "step": 157950
+    },
+    {
+      "epoch": 0.7888336787435392,
+      "grad_norm": 0.08796839416027069,
+      "learning_rate": 6.350847330346192e-06,
+      "loss": 8.5714,
+      "step": 157960
+    },
+    {
+      "epoch": 0.7888836175684786,
+      "grad_norm": 0.08775161951780319,
+      "learning_rate": 6.349345415404641e-06,
+      "loss": 8.5559,
+      "step": 157970
+    },
+    {
+      "epoch": 0.7889335563934181,
+      "grad_norm": 0.09324196726083755,
+      "learning_rate": 6.347843500463091e-06,
+      "loss": 8.5625,
+      "step": 157980
+    },
+    {
+      "epoch": 0.7889834952183575,
+      "grad_norm": 0.08922797441482544,
+      "learning_rate": 6.3463415855215395e-06,
+      "loss": 8.5759,
+      "step": 157990
+    },
+    {
+      "epoch": 0.789033434043297,
+      "grad_norm": 0.09467015415430069,
+      "learning_rate": 6.34483967057999e-06,
+      "loss": 8.5574,
+      "step": 158000
+    },
+    {
+      "epoch": 0.7890833728682364,
+      "grad_norm": 0.09067735075950623,
+      "learning_rate": 6.34333775563844e-06,
+      "loss": 8.5545,
+      "step": 158010
+    },
+    {
+      "epoch": 0.7891333116931759,
+      "grad_norm": 0.09173320233821869,
+      "learning_rate": 6.341835840696888e-06,
+      "loss": 8.5527,
+      "step": 158020
+    },
+    {
+      "epoch": 0.7891832505181153,
+      "grad_norm": 0.08913468569517136,
+      "learning_rate": 6.3403339257553385e-06,
+      "loss": 8.567,
+      "step": 158030
+    },
+    {
+      "epoch": 0.7892331893430548,
+      "grad_norm": 0.088861383497715,
+      "learning_rate": 6.338832010813787e-06,
+      "loss": 8.5653,
+      "step": 158040
+    },
+    {
+      "epoch": 0.7892831281679942,
+      "grad_norm": 0.08768237382173538,
+      "learning_rate": 6.337330095872237e-06,
+      "loss": 8.5644,
+      "step": 158050
+    },
+    {
+      "epoch": 0.7893330669929337,
+      "grad_norm": 0.09068477898836136,
+      "learning_rate": 6.335828180930687e-06,
+      "loss": 8.5581,
+      "step": 158060
+    },
+    {
+      "epoch": 0.7893830058178731,
+      "grad_norm": 0.09171970933675766,
+      "learning_rate": 6.334326265989136e-06,
+      "loss": 8.5701,
+      "step": 158070
+    },
+    {
+      "epoch": 0.7894329446428126,
+      "grad_norm": 0.09282483905553818,
+      "learning_rate": 6.332824351047586e-06,
+      "loss": 8.5821,
+      "step": 158080
+    },
+    {
+      "epoch": 0.789482883467752,
+      "grad_norm": 0.09467281401157379,
+      "learning_rate": 6.331322436106035e-06,
+      "loss": 8.5509,
+      "step": 158090
+    },
+    {
+      "epoch": 0.7895328222926915,
+      "grad_norm": 0.09098406136035919,
+      "learning_rate": 6.329820521164485e-06,
+      "loss": 8.5573,
+      "step": 158100
+    },
+    {
+      "epoch": 0.7895827611176309,
+      "grad_norm": 0.08324146270751953,
+      "learning_rate": 6.328318606222935e-06,
+      "loss": 8.557,
+      "step": 158110
+    },
+    {
+      "epoch": 0.7896326999425703,
+      "grad_norm": 0.09208779036998749,
+      "learning_rate": 6.326816691281383e-06,
+      "loss": 8.576,
+      "step": 158120
+    },
+    {
+      "epoch": 0.7896826387675098,
+      "grad_norm": 0.0910223051905632,
+      "learning_rate": 6.3253147763398335e-06,
+      "loss": 8.5419,
+      "step": 158130
+    },
+    {
+      "epoch": 0.7897325775924493,
+      "grad_norm": 0.0862598568201065,
+      "learning_rate": 6.323812861398283e-06,
+      "loss": 8.5799,
+      "step": 158140
+    },
+    {
+      "epoch": 0.7897825164173887,
+      "grad_norm": 0.0855773463845253,
+      "learning_rate": 6.322310946456732e-06,
+      "loss": 8.5602,
+      "step": 158150
+    },
+    {
+      "epoch": 0.7898324552423281,
+      "grad_norm": 0.09082646667957306,
+      "learning_rate": 6.320809031515182e-06,
+      "loss": 8.5751,
+      "step": 158160
+    },
+    {
+      "epoch": 0.7898823940672676,
+      "grad_norm": 0.09423871338367462,
+      "learning_rate": 6.319307116573632e-06,
+      "loss": 8.5738,
+      "step": 158170
+    },
+    {
+      "epoch": 0.7899323328922071,
+      "grad_norm": 0.08389170467853546,
+      "learning_rate": 6.317805201632081e-06,
+      "loss": 8.579,
+      "step": 158180
+    },
+    {
+      "epoch": 0.7899822717171465,
+      "grad_norm": 0.08935504406690598,
+      "learning_rate": 6.31630328669053e-06,
+      "loss": 8.5623,
+      "step": 158190
+    },
+    {
+      "epoch": 0.7900322105420859,
+      "grad_norm": 0.09032989293336868,
+      "learning_rate": 6.31480137174898e-06,
+      "loss": 8.5733,
+      "step": 158200
+    },
+    {
+      "epoch": 0.7900821493670254,
+      "grad_norm": 0.08665549755096436,
+      "learning_rate": 6.31329945680743e-06,
+      "loss": 8.5566,
+      "step": 158210
+    },
+    {
+      "epoch": 0.7901320881919648,
+      "grad_norm": 0.0938415452837944,
+      "learning_rate": 6.311797541865879e-06,
+      "loss": 8.5668,
+      "step": 158220
+    },
+    {
+      "epoch": 0.7901820270169043,
+      "grad_norm": 0.08663594722747803,
+      "learning_rate": 6.3102956269243285e-06,
+      "loss": 8.5642,
+      "step": 158230
+    },
+    {
+      "epoch": 0.7902319658418437,
+      "grad_norm": 0.08731528371572495,
+      "learning_rate": 6.308793711982778e-06,
+      "loss": 8.5803,
+      "step": 158240
+    },
+    {
+      "epoch": 0.7902819046667832,
+      "grad_norm": 0.08931776881217957,
+      "learning_rate": 6.307291797041228e-06,
+      "loss": 8.5432,
+      "step": 158250
+    },
+    {
+      "epoch": 0.7903318434917226,
+      "grad_norm": 0.09380349516868591,
+      "learning_rate": 6.305789882099677e-06,
+      "loss": 8.5696,
+      "step": 158260
+    },
+    {
+      "epoch": 0.7903817823166621,
+      "grad_norm": 0.09893126040697098,
+      "learning_rate": 6.304287967158127e-06,
+      "loss": 8.5641,
+      "step": 158270
+    },
+    {
+      "epoch": 0.7904317211416015,
+      "grad_norm": 0.0894632488489151,
+      "learning_rate": 6.302786052216576e-06,
+      "loss": 8.5541,
+      "step": 158280
+    },
+    {
+      "epoch": 0.790481659966541,
+      "grad_norm": 0.08917111158370972,
+      "learning_rate": 6.301284137275025e-06,
+      "loss": 8.5575,
+      "step": 158290
+    },
+    {
+      "epoch": 0.7905315987914804,
+      "grad_norm": 0.09342500567436218,
+      "learning_rate": 6.2997822223334755e-06,
+      "loss": 8.5596,
+      "step": 158300
+    },
+    {
+      "epoch": 0.7905815376164199,
+      "grad_norm": 0.0935325175523758,
+      "learning_rate": 6.298280307391925e-06,
+      "loss": 8.5477,
+      "step": 158310
+    },
+    {
+      "epoch": 0.7906314764413593,
+      "grad_norm": 0.09898251295089722,
+      "learning_rate": 6.296778392450374e-06,
+      "loss": 8.5758,
+      "step": 158320
+    },
+    {
+      "epoch": 0.7906814152662988,
+      "grad_norm": 0.09187868982553482,
+      "learning_rate": 6.295276477508824e-06,
+      "loss": 8.5561,
+      "step": 158330
+    },
+    {
+      "epoch": 0.7907313540912382,
+      "grad_norm": 0.08766231685876846,
+      "learning_rate": 6.293774562567273e-06,
+      "loss": 8.5648,
+      "step": 158340
+    },
+    {
+      "epoch": 0.7907812929161777,
+      "grad_norm": 0.09430361539125443,
+      "learning_rate": 6.292272647625723e-06,
+      "loss": 8.5563,
+      "step": 158350
+    },
+    {
+      "epoch": 0.7908312317411171,
+      "grad_norm": 0.08926588296890259,
+      "learning_rate": 6.290770732684172e-06,
+      "loss": 8.5654,
+      "step": 158360
+    },
+    {
+      "epoch": 0.7908811705660566,
+      "grad_norm": 0.08842246234416962,
+      "learning_rate": 6.289268817742622e-06,
+      "loss": 8.568,
+      "step": 158370
+    },
+    {
+      "epoch": 0.790931109390996,
+      "grad_norm": 0.08935972303152084,
+      "learning_rate": 6.287766902801072e-06,
+      "loss": 8.556,
+      "step": 158380
+    },
+    {
+      "epoch": 0.7909810482159355,
+      "grad_norm": 0.09105094522237778,
+      "learning_rate": 6.28626498785952e-06,
+      "loss": 8.5633,
+      "step": 158390
+    },
+    {
+      "epoch": 0.7910309870408749,
+      "grad_norm": 0.09042556583881378,
+      "learning_rate": 6.2847630729179705e-06,
+      "loss": 8.5534,
+      "step": 158400
+    },
+    {
+      "epoch": 0.7910809258658144,
+      "grad_norm": 0.09127331525087357,
+      "learning_rate": 6.283261157976421e-06,
+      "loss": 8.5687,
+      "step": 158410
+    },
+    {
+      "epoch": 0.7911308646907538,
+      "grad_norm": 0.09261275827884674,
+      "learning_rate": 6.281759243034869e-06,
+      "loss": 8.5473,
+      "step": 158420
+    },
+    {
+      "epoch": 0.7911808035156933,
+      "grad_norm": 0.08710244297981262,
+      "learning_rate": 6.280257328093319e-06,
+      "loss": 8.578,
+      "step": 158430
+    },
+    {
+      "epoch": 0.7912307423406327,
+      "grad_norm": 0.10005831718444824,
+      "learning_rate": 6.278755413151769e-06,
+      "loss": 8.5261,
+      "step": 158440
+    },
+    {
+      "epoch": 0.7912806811655722,
+      "grad_norm": 0.09465691447257996,
+      "learning_rate": 6.277253498210218e-06,
+      "loss": 8.5627,
+      "step": 158450
+    },
+    {
+      "epoch": 0.7913306199905116,
+      "grad_norm": 0.09323757886886597,
+      "learning_rate": 6.275751583268668e-06,
+      "loss": 8.5629,
+      "step": 158460
+    },
+    {
+      "epoch": 0.7913805588154511,
+      "grad_norm": 0.09241240471601486,
+      "learning_rate": 6.274249668327117e-06,
+      "loss": 8.564,
+      "step": 158470
+    },
+    {
+      "epoch": 0.7914304976403905,
+      "grad_norm": 0.09037698805332184,
+      "learning_rate": 6.272747753385567e-06,
+      "loss": 8.5478,
+      "step": 158480
+    },
+    {
+      "epoch": 0.79148043646533,
+      "grad_norm": 0.09861491620540619,
+      "learning_rate": 6.271245838444017e-06,
+      "loss": 8.5623,
+      "step": 158490
+    },
+    {
+      "epoch": 0.7915303752902694,
+      "grad_norm": 0.09813782572746277,
+      "learning_rate": 6.2697439235024655e-06,
+      "loss": 8.5839,
+      "step": 158500
+    },
+    {
+      "epoch": 0.7915803141152089,
+      "grad_norm": 0.08585695922374725,
+      "learning_rate": 6.268242008560916e-06,
+      "loss": 8.5453,
+      "step": 158510
+    },
+    {
+      "epoch": 0.7916302529401483,
+      "grad_norm": 0.09625940024852753,
+      "learning_rate": 6.266740093619364e-06,
+      "loss": 8.5728,
+      "step": 158520
+    },
+    {
+      "epoch": 0.7916801917650877,
+      "grad_norm": 0.09219558537006378,
+      "learning_rate": 6.265238178677814e-06,
+      "loss": 8.5559,
+      "step": 158530
+    },
+    {
+      "epoch": 0.7917301305900272,
+      "grad_norm": 0.0883294865489006,
+      "learning_rate": 6.2637362637362645e-06,
+      "loss": 8.5629,
+      "step": 158540
+    },
+    {
+      "epoch": 0.7917800694149667,
+      "grad_norm": 0.09103575348854065,
+      "learning_rate": 6.262234348794713e-06,
+      "loss": 8.5675,
+      "step": 158550
+    },
+    {
+      "epoch": 0.7918300082399061,
+      "grad_norm": 0.09251187741756439,
+      "learning_rate": 6.260732433853163e-06,
+      "loss": 8.5794,
+      "step": 158560
+    },
+    {
+      "epoch": 0.7918799470648455,
+      "grad_norm": 0.09562096744775772,
+      "learning_rate": 6.2592305189116125e-06,
+      "loss": 8.5676,
+      "step": 158570
+    },
+    {
+      "epoch": 0.791929885889785,
+      "grad_norm": 0.08727453649044037,
+      "learning_rate": 6.257728603970062e-06,
+      "loss": 8.5519,
+      "step": 158580
+    },
+    {
+      "epoch": 0.7919798247147245,
+      "grad_norm": 0.08689242601394653,
+      "learning_rate": 6.256226689028512e-06,
+      "loss": 8.561,
+      "step": 158590
+    },
+    {
+      "epoch": 0.7920297635396639,
+      "grad_norm": 0.09087180346250534,
+      "learning_rate": 6.2547247740869605e-06,
+      "loss": 8.5644,
+      "step": 158600
+    },
+    {
+      "epoch": 0.7920797023646033,
+      "grad_norm": 0.09163371473550797,
+      "learning_rate": 6.253222859145411e-06,
+      "loss": 8.5527,
+      "step": 158610
+    },
+    {
+      "epoch": 0.7921296411895428,
+      "grad_norm": 0.0966818630695343,
+      "learning_rate": 6.25172094420386e-06,
+      "loss": 8.5677,
+      "step": 158620
+    },
+    {
+      "epoch": 0.7921795800144823,
+      "grad_norm": 0.0895446240901947,
+      "learning_rate": 6.250219029262309e-06,
+      "loss": 8.5491,
+      "step": 158630
+    },
+    {
+      "epoch": 0.7922295188394217,
+      "grad_norm": 0.08966691046953201,
+      "learning_rate": 6.2487171143207595e-06,
+      "loss": 8.5715,
+      "step": 158640
+    },
+    {
+      "epoch": 0.7922794576643611,
+      "grad_norm": 0.09029648452997208,
+      "learning_rate": 6.247215199379209e-06,
+      "loss": 8.5655,
+      "step": 158650
+    },
+    {
+      "epoch": 0.7923293964893006,
+      "grad_norm": 0.09223111718893051,
+      "learning_rate": 6.245713284437658e-06,
+      "loss": 8.5537,
+      "step": 158660
+    },
+    {
+      "epoch": 0.7923793353142401,
+      "grad_norm": 0.09418982267379761,
+      "learning_rate": 6.2442113694961075e-06,
+      "loss": 8.5509,
+      "step": 158670
+    },
+    {
+      "epoch": 0.7924292741391795,
+      "grad_norm": 0.09183704107999802,
+      "learning_rate": 6.242709454554557e-06,
+      "loss": 8.5811,
+      "step": 158680
+    },
+    {
+      "epoch": 0.7924792129641189,
+      "grad_norm": 0.09103389829397202,
+      "learning_rate": 6.241207539613007e-06,
+      "loss": 8.5531,
+      "step": 158690
+    },
+    {
+      "epoch": 0.7925291517890584,
+      "grad_norm": 0.09130579978227615,
+      "learning_rate": 6.239705624671456e-06,
+      "loss": 8.5477,
+      "step": 158700
+    },
+    {
+      "epoch": 0.7925790906139979,
+      "grad_norm": 0.0948038399219513,
+      "learning_rate": 6.238203709729906e-06,
+      "loss": 8.5581,
+      "step": 158710
+    },
+    {
+      "epoch": 0.7926290294389373,
+      "grad_norm": 0.08801276236772537,
+      "learning_rate": 6.236701794788355e-06,
+      "loss": 8.5652,
+      "step": 158720
+    },
+    {
+      "epoch": 0.7926789682638767,
+      "grad_norm": 0.09371495991945267,
+      "learning_rate": 6.235199879846805e-06,
+      "loss": 8.54,
+      "step": 158730
+    },
+    {
+      "epoch": 0.7927289070888162,
+      "grad_norm": 0.09850714355707169,
+      "learning_rate": 6.2336979649052546e-06,
+      "loss": 8.5497,
+      "step": 158740
+    },
+    {
+      "epoch": 0.7927788459137557,
+      "grad_norm": 0.08900289982557297,
+      "learning_rate": 6.232196049963704e-06,
+      "loss": 8.5674,
+      "step": 158750
+    },
+    {
+      "epoch": 0.7928287847386951,
+      "grad_norm": 0.09315358847379684,
+      "learning_rate": 6.230694135022153e-06,
+      "loss": 8.566,
+      "step": 158760
+    },
+    {
+      "epoch": 0.7928787235636345,
+      "grad_norm": 0.09045889228582382,
+      "learning_rate": 6.2291922200806025e-06,
+      "loss": 8.5793,
+      "step": 158770
+    },
+    {
+      "epoch": 0.792928662388574,
+      "grad_norm": 0.0957542434334755,
+      "learning_rate": 6.227690305139053e-06,
+      "loss": 8.5479,
+      "step": 158780
+    },
+    {
+      "epoch": 0.7929786012135135,
+      "grad_norm": 0.08995579183101654,
+      "learning_rate": 6.226188390197502e-06,
+      "loss": 8.5627,
+      "step": 158790
+    },
+    {
+      "epoch": 0.7930285400384529,
+      "grad_norm": 0.09000235050916672,
+      "learning_rate": 6.224686475255951e-06,
+      "loss": 8.5479,
+      "step": 158800
+    },
+    {
+      "epoch": 0.7930784788633923,
+      "grad_norm": 0.08959494531154633,
+      "learning_rate": 6.2231845603144016e-06,
+      "loss": 8.5524,
+      "step": 158810
+    },
+    {
+      "epoch": 0.7931284176883318,
+      "grad_norm": 0.0925586149096489,
+      "learning_rate": 6.22168264537285e-06,
+      "loss": 8.566,
+      "step": 158820
+    },
+    {
+      "epoch": 0.7931783565132713,
+      "grad_norm": 0.08586516976356506,
+      "learning_rate": 6.2201807304313e-06,
+      "loss": 8.5513,
+      "step": 158830
+    },
+    {
+      "epoch": 0.7932282953382107,
+      "grad_norm": 0.0888560339808464,
+      "learning_rate": 6.2186788154897496e-06,
+      "loss": 8.5655,
+      "step": 158840
+    },
+    {
+      "epoch": 0.7932782341631501,
+      "grad_norm": 0.08735716342926025,
+      "learning_rate": 6.217176900548199e-06,
+      "loss": 8.5603,
+      "step": 158850
+    },
+    {
+      "epoch": 0.7933281729880896,
+      "grad_norm": 0.08948913961648941,
+      "learning_rate": 6.215674985606649e-06,
+      "loss": 8.5602,
+      "step": 158860
+    },
+    {
+      "epoch": 0.7933781118130291,
+      "grad_norm": 0.09545059502124786,
+      "learning_rate": 6.2141730706650976e-06,
+      "loss": 8.5611,
+      "step": 158870
+    },
+    {
+      "epoch": 0.7934280506379685,
+      "grad_norm": 0.08870154619216919,
+      "learning_rate": 6.212671155723548e-06,
+      "loss": 8.5613,
+      "step": 158880
+    },
+    {
+      "epoch": 0.7934779894629079,
+      "grad_norm": 0.08690696954727173,
+      "learning_rate": 6.211169240781998e-06,
+      "loss": 8.5648,
+      "step": 158890
+    },
+    {
+      "epoch": 0.7935279282878474,
+      "grad_norm": 0.09118922054767609,
+      "learning_rate": 6.209667325840446e-06,
+      "loss": 8.5562,
+      "step": 158900
+    },
+    {
+      "epoch": 0.7935778671127869,
+      "grad_norm": 0.09019847214221954,
+      "learning_rate": 6.2081654108988966e-06,
+      "loss": 8.5698,
+      "step": 158910
+    },
+    {
+      "epoch": 0.7936278059377263,
+      "grad_norm": 0.09406841546297073,
+      "learning_rate": 6.206663495957345e-06,
+      "loss": 8.5572,
+      "step": 158920
+    },
+    {
+      "epoch": 0.7936777447626657,
+      "grad_norm": 0.09367325156927109,
+      "learning_rate": 6.205161581015795e-06,
+      "loss": 8.5657,
+      "step": 158930
+    },
+    {
+      "epoch": 0.7937276835876051,
+      "grad_norm": 0.09416260570287704,
+      "learning_rate": 6.203659666074245e-06,
+      "loss": 8.5811,
+      "step": 158940
+    },
+    {
+      "epoch": 0.7937776224125447,
+      "grad_norm": 0.09471859037876129,
+      "learning_rate": 6.202157751132694e-06,
+      "loss": 8.5597,
+      "step": 158950
+    },
+    {
+      "epoch": 0.7938275612374841,
+      "grad_norm": 0.09237445890903473,
+      "learning_rate": 6.200655836191144e-06,
+      "loss": 8.5647,
+      "step": 158960
+    },
+    {
+      "epoch": 0.7938775000624235,
+      "grad_norm": 0.09338158369064331,
+      "learning_rate": 6.199153921249593e-06,
+      "loss": 8.5622,
+      "step": 158970
+    },
+    {
+      "epoch": 0.793927438887363,
+      "grad_norm": 0.08678948134183884,
+      "learning_rate": 6.197652006308043e-06,
+      "loss": 8.5693,
+      "step": 158980
+    },
+    {
+      "epoch": 0.7939773777123025,
+      "grad_norm": 0.09783153980970383,
+      "learning_rate": 6.196150091366493e-06,
+      "loss": 8.5626,
+      "step": 158990
+    },
+    {
+      "epoch": 0.7940273165372419,
+      "grad_norm": 0.09704387933015823,
+      "learning_rate": 6.194648176424941e-06,
+      "loss": 8.5705,
+      "step": 159000
+    },
+    {
+      "epoch": 0.7940772553621813,
+      "grad_norm": 0.09087003022432327,
+      "learning_rate": 6.193146261483392e-06,
+      "loss": 8.5679,
+      "step": 159010
+    },
+    {
+      "epoch": 0.7941271941871207,
+      "grad_norm": 0.09224209189414978,
+      "learning_rate": 6.191644346541841e-06,
+      "loss": 8.5447,
+      "step": 159020
+    },
+    {
+      "epoch": 0.7941771330120603,
+      "grad_norm": 0.08932413905858994,
+      "learning_rate": 6.19014243160029e-06,
+      "loss": 8.5509,
+      "step": 159030
+    },
+    {
+      "epoch": 0.7942270718369997,
+      "grad_norm": 0.09286679327487946,
+      "learning_rate": 6.18864051665874e-06,
+      "loss": 8.5673,
+      "step": 159040
+    },
+    {
+      "epoch": 0.7942770106619391,
+      "grad_norm": 0.09589332342147827,
+      "learning_rate": 6.18713860171719e-06,
+      "loss": 8.5603,
+      "step": 159050
+    },
+    {
+      "epoch": 0.7943269494868785,
+      "grad_norm": 0.10077871382236481,
+      "learning_rate": 6.185636686775639e-06,
+      "loss": 8.5626,
+      "step": 159060
+    },
+    {
+      "epoch": 0.7943768883118181,
+      "grad_norm": 0.0931212455034256,
+      "learning_rate": 6.184134771834088e-06,
+      "loss": 8.5467,
+      "step": 159070
+    },
+    {
+      "epoch": 0.7944268271367575,
+      "grad_norm": 0.09118203818798065,
+      "learning_rate": 6.182632856892538e-06,
+      "loss": 8.5558,
+      "step": 159080
+    },
+    {
+      "epoch": 0.7944767659616969,
+      "grad_norm": 0.08937732130289078,
+      "learning_rate": 6.181130941950988e-06,
+      "loss": 8.56,
+      "step": 159090
+    },
+    {
+      "epoch": 0.7945267047866363,
+      "grad_norm": 0.09185148030519485,
+      "learning_rate": 6.179629027009437e-06,
+      "loss": 8.5669,
+      "step": 159100
+    },
+    {
+      "epoch": 0.7945766436115759,
+      "grad_norm": 0.08831166476011276,
+      "learning_rate": 6.178127112067887e-06,
+      "loss": 8.5614,
+      "step": 159110
+    },
+    {
+      "epoch": 0.7946265824365153,
+      "grad_norm": 0.08398475497961044,
+      "learning_rate": 6.176625197126336e-06,
+      "loss": 8.5808,
+      "step": 159120
+    },
+    {
+      "epoch": 0.7946765212614547,
+      "grad_norm": 0.08763767033815384,
+      "learning_rate": 6.175123282184786e-06,
+      "loss": 8.567,
+      "step": 159130
+    },
+    {
+      "epoch": 0.7947264600863941,
+      "grad_norm": 0.10390178859233856,
+      "learning_rate": 6.1736213672432354e-06,
+      "loss": 8.542,
+      "step": 159140
+    },
+    {
+      "epoch": 0.7947763989113337,
+      "grad_norm": 0.08661069720983505,
+      "learning_rate": 6.172119452301685e-06,
+      "loss": 8.5623,
+      "step": 159150
+    },
+    {
+      "epoch": 0.7948263377362731,
+      "grad_norm": 0.0880892425775528,
+      "learning_rate": 6.170617537360134e-06,
+      "loss": 8.5611,
+      "step": 159160
+    },
+    {
+      "epoch": 0.7948762765612125,
+      "grad_norm": 0.0939859077334404,
+      "learning_rate": 6.169115622418583e-06,
+      "loss": 8.5483,
+      "step": 159170
+    },
+    {
+      "epoch": 0.7949262153861519,
+      "grad_norm": 0.09007950127124786,
+      "learning_rate": 6.167613707477034e-06,
+      "loss": 8.5587,
+      "step": 159180
+    },
+    {
+      "epoch": 0.7949761542110915,
+      "grad_norm": 0.09516777098178864,
+      "learning_rate": 6.166111792535483e-06,
+      "loss": 8.574,
+      "step": 159190
+    },
+    {
+      "epoch": 0.7950260930360309,
+      "grad_norm": 0.08972634375095367,
+      "learning_rate": 6.164609877593932e-06,
+      "loss": 8.5578,
+      "step": 159200
+    },
+    {
+      "epoch": 0.7950760318609703,
+      "grad_norm": 0.09445880353450775,
+      "learning_rate": 6.1631079626523824e-06,
+      "loss": 8.5576,
+      "step": 159210
+    },
+    {
+      "epoch": 0.7951259706859097,
+      "grad_norm": 0.08887441456317902,
+      "learning_rate": 6.161606047710831e-06,
+      "loss": 8.5647,
+      "step": 159220
+    },
+    {
+      "epoch": 0.7951759095108492,
+      "grad_norm": 0.09141810983419418,
+      "learning_rate": 6.160104132769281e-06,
+      "loss": 8.5579,
+      "step": 159230
+    },
+    {
+      "epoch": 0.7952258483357887,
+      "grad_norm": 0.0876648873090744,
+      "learning_rate": 6.1586022178277304e-06,
+      "loss": 8.5538,
+      "step": 159240
+    },
+    {
+      "epoch": 0.7952757871607281,
+      "grad_norm": 0.08809340745210648,
+      "learning_rate": 6.15710030288618e-06,
+      "loss": 8.5455,
+      "step": 159250
+    },
+    {
+      "epoch": 0.7953257259856675,
+      "grad_norm": 0.09779562056064606,
+      "learning_rate": 6.15559838794463e-06,
+      "loss": 8.5439,
+      "step": 159260
+    },
+    {
+      "epoch": 0.795375664810607,
+      "grad_norm": 0.08831503242254257,
+      "learning_rate": 6.1540964730030784e-06,
+      "loss": 8.5677,
+      "step": 159270
+    },
+    {
+      "epoch": 0.7954256036355465,
+      "grad_norm": 0.09317151457071304,
+      "learning_rate": 6.152594558061529e-06,
+      "loss": 8.551,
+      "step": 159280
+    },
+    {
+      "epoch": 0.7954755424604859,
+      "grad_norm": 0.08679518103599548,
+      "learning_rate": 6.151092643119979e-06,
+      "loss": 8.577,
+      "step": 159290
+    },
+    {
+      "epoch": 0.7955254812854253,
+      "grad_norm": 0.09124913066625595,
+      "learning_rate": 6.149590728178427e-06,
+      "loss": 8.5719,
+      "step": 159300
+    },
+    {
+      "epoch": 0.7955754201103648,
+      "grad_norm": 0.08849890530109406,
+      "learning_rate": 6.1480888132368774e-06,
+      "loss": 8.5562,
+      "step": 159310
+    },
+    {
+      "epoch": 0.7956253589353043,
+      "grad_norm": 0.09584712982177734,
+      "learning_rate": 6.146586898295326e-06,
+      "loss": 8.5612,
+      "step": 159320
+    },
+    {
+      "epoch": 0.7956752977602437,
+      "grad_norm": 0.09756402671337128,
+      "learning_rate": 6.145084983353776e-06,
+      "loss": 8.5604,
+      "step": 159330
+    },
+    {
+      "epoch": 0.7957252365851831,
+      "grad_norm": 0.09176575392484665,
+      "learning_rate": 6.143583068412226e-06,
+      "loss": 8.5578,
+      "step": 159340
+    },
+    {
+      "epoch": 0.7957751754101225,
+      "grad_norm": 0.09086544811725616,
+      "learning_rate": 6.142081153470675e-06,
+      "loss": 8.5625,
+      "step": 159350
+    },
+    {
+      "epoch": 0.7958251142350621,
+      "grad_norm": 0.09206107258796692,
+      "learning_rate": 6.140579238529125e-06,
+      "loss": 8.5577,
+      "step": 159360
+    },
+    {
+      "epoch": 0.7958750530600015,
+      "grad_norm": 0.09471509605646133,
+      "learning_rate": 6.139077323587574e-06,
+      "loss": 8.5724,
+      "step": 159370
+    },
+    {
+      "epoch": 0.7959249918849409,
+      "grad_norm": 0.09322207421064377,
+      "learning_rate": 6.137575408646024e-06,
+      "loss": 8.5488,
+      "step": 159380
+    },
+    {
+      "epoch": 0.7959749307098803,
+      "grad_norm": 0.09167961776256561,
+      "learning_rate": 6.136073493704474e-06,
+      "loss": 8.5676,
+      "step": 159390
+    },
+    {
+      "epoch": 0.7960248695348199,
+      "grad_norm": 0.09485690295696259,
+      "learning_rate": 6.134571578762922e-06,
+      "loss": 8.5436,
+      "step": 159400
+    },
+    {
+      "epoch": 0.7960748083597593,
+      "grad_norm": 0.08992945402860641,
+      "learning_rate": 6.1330696638213725e-06,
+      "loss": 8.5585,
+      "step": 159410
+    },
+    {
+      "epoch": 0.7961247471846987,
+      "grad_norm": 0.08964666724205017,
+      "learning_rate": 6.131567748879822e-06,
+      "loss": 8.5553,
+      "step": 159420
+    },
+    {
+      "epoch": 0.7961746860096381,
+      "grad_norm": 0.09216075390577316,
+      "learning_rate": 6.130065833938271e-06,
+      "loss": 8.5633,
+      "step": 159430
+    },
+    {
+      "epoch": 0.7962246248345777,
+      "grad_norm": 0.08898287266492844,
+      "learning_rate": 6.128563918996721e-06,
+      "loss": 8.5607,
+      "step": 159440
+    },
+    {
+      "epoch": 0.7962745636595171,
+      "grad_norm": 0.08977809548377991,
+      "learning_rate": 6.127062004055171e-06,
+      "loss": 8.5637,
+      "step": 159450
+    },
+    {
+      "epoch": 0.7963245024844565,
+      "grad_norm": 0.08928190171718597,
+      "learning_rate": 6.12556008911362e-06,
+      "loss": 8.5715,
+      "step": 159460
+    },
+    {
+      "epoch": 0.7963744413093959,
+      "grad_norm": 0.09225127846002579,
+      "learning_rate": 6.124058174172069e-06,
+      "loss": 8.5857,
+      "step": 159470
+    },
+    {
+      "epoch": 0.7964243801343355,
+      "grad_norm": 0.09755774587392807,
+      "learning_rate": 6.122556259230519e-06,
+      "loss": 8.5431,
+      "step": 159480
+    },
+    {
+      "epoch": 0.7964743189592749,
+      "grad_norm": 0.09840092062950134,
+      "learning_rate": 6.121054344288969e-06,
+      "loss": 8.5594,
+      "step": 159490
+    },
+    {
+      "epoch": 0.7965242577842143,
+      "grad_norm": 0.09154287725687027,
+      "learning_rate": 6.119552429347418e-06,
+      "loss": 8.5609,
+      "step": 159500
+    },
+    {
+      "epoch": 0.7965741966091537,
+      "grad_norm": 0.08567217737436295,
+      "learning_rate": 6.1180505144058675e-06,
+      "loss": 8.5555,
+      "step": 159510
+    },
+    {
+      "epoch": 0.7966241354340933,
+      "grad_norm": 0.08857017010450363,
+      "learning_rate": 6.116548599464317e-06,
+      "loss": 8.5614,
+      "step": 159520
+    },
+    {
+      "epoch": 0.7966740742590327,
+      "grad_norm": 0.08761531114578247,
+      "learning_rate": 6.115046684522767e-06,
+      "loss": 8.5644,
+      "step": 159530
+    },
+    {
+      "epoch": 0.7967240130839721,
+      "grad_norm": 0.0912371501326561,
+      "learning_rate": 6.113544769581216e-06,
+      "loss": 8.5736,
+      "step": 159540
+    },
+    {
+      "epoch": 0.7967739519089115,
+      "grad_norm": 0.09295840561389923,
+      "learning_rate": 6.112042854639666e-06,
+      "loss": 8.5424,
+      "step": 159550
+    },
+    {
+      "epoch": 0.7968238907338511,
+      "grad_norm": 0.09031003713607788,
+      "learning_rate": 6.110540939698115e-06,
+      "loss": 8.554,
+      "step": 159560
+    },
+    {
+      "epoch": 0.7968738295587905,
+      "grad_norm": 0.0904209241271019,
+      "learning_rate": 6.109039024756565e-06,
+      "loss": 8.5746,
+      "step": 159570
+    },
+    {
+      "epoch": 0.7969237683837299,
+      "grad_norm": 0.08807667344808578,
+      "learning_rate": 6.1075371098150145e-06,
+      "loss": 8.5526,
+      "step": 159580
+    },
+    {
+      "epoch": 0.7969737072086693,
+      "grad_norm": 0.09461508691310883,
+      "learning_rate": 6.106035194873464e-06,
+      "loss": 8.557,
+      "step": 159590
+    },
+    {
+      "epoch": 0.7970236460336089,
+      "grad_norm": 0.08692062646150589,
+      "learning_rate": 6.104533279931913e-06,
+      "loss": 8.5529,
+      "step": 159600
+    },
+    {
+      "epoch": 0.7970735848585483,
+      "grad_norm": 0.09569226950407028,
+      "learning_rate": 6.103031364990363e-06,
+      "loss": 8.5602,
+      "step": 159610
+    },
+    {
+      "epoch": 0.7971235236834877,
+      "grad_norm": 0.09210385382175446,
+      "learning_rate": 6.101529450048813e-06,
+      "loss": 8.5649,
+      "step": 159620
+    },
+    {
+      "epoch": 0.7971734625084271,
+      "grad_norm": 0.09162899106740952,
+      "learning_rate": 6.100027535107262e-06,
+      "loss": 8.5509,
+      "step": 159630
+    },
+    {
+      "epoch": 0.7972234013333667,
+      "grad_norm": 0.09765185415744781,
+      "learning_rate": 6.098525620165711e-06,
+      "loss": 8.5527,
+      "step": 159640
+    },
+    {
+      "epoch": 0.7972733401583061,
+      "grad_norm": 0.09412235766649246,
+      "learning_rate": 6.097023705224161e-06,
+      "loss": 8.5699,
+      "step": 159650
+    },
+    {
+      "epoch": 0.7973232789832455,
+      "grad_norm": 0.09401925653219223,
+      "learning_rate": 6.095521790282611e-06,
+      "loss": 8.5554,
+      "step": 159660
+    },
+    {
+      "epoch": 0.7973732178081849,
+      "grad_norm": 0.09170075505971909,
+      "learning_rate": 6.09401987534106e-06,
+      "loss": 8.5666,
+      "step": 159670
+    },
+    {
+      "epoch": 0.7974231566331245,
+      "grad_norm": 0.0880158469080925,
+      "learning_rate": 6.0925179603995095e-06,
+      "loss": 8.554,
+      "step": 159680
+    },
+    {
+      "epoch": 0.7974730954580639,
+      "grad_norm": 0.09159617871046066,
+      "learning_rate": 6.09101604545796e-06,
+      "loss": 8.5581,
+      "step": 159690
+    },
+    {
+      "epoch": 0.7975230342830033,
+      "grad_norm": 0.09276486188173294,
+      "learning_rate": 6.089514130516408e-06,
+      "loss": 8.5436,
+      "step": 159700
+    },
+    {
+      "epoch": 0.7975729731079427,
+      "grad_norm": 0.08525187522172928,
+      "learning_rate": 6.088012215574858e-06,
+      "loss": 8.5508,
+      "step": 159710
+    },
+    {
+      "epoch": 0.7976229119328823,
+      "grad_norm": 0.08781725913286209,
+      "learning_rate": 6.086510300633308e-06,
+      "loss": 8.5792,
+      "step": 159720
+    },
+    {
+      "epoch": 0.7976728507578217,
+      "grad_norm": 0.09643100947141647,
+      "learning_rate": 6.085008385691757e-06,
+      "loss": 8.5551,
+      "step": 159730
+    },
+    {
+      "epoch": 0.7977227895827611,
+      "grad_norm": 0.08715363591909409,
+      "learning_rate": 6.083506470750207e-06,
+      "loss": 8.5656,
+      "step": 159740
+    },
+    {
+      "epoch": 0.7977727284077005,
+      "grad_norm": 0.09139413386583328,
+      "learning_rate": 6.082004555808656e-06,
+      "loss": 8.5401,
+      "step": 159750
+    },
+    {
+      "epoch": 0.7978226672326401,
+      "grad_norm": 0.09375610202550888,
+      "learning_rate": 6.080502640867106e-06,
+      "loss": 8.5596,
+      "step": 159760
+    },
+    {
+      "epoch": 0.7978726060575795,
+      "grad_norm": 0.09116285294294357,
+      "learning_rate": 6.079000725925556e-06,
+      "loss": 8.5359,
+      "step": 159770
+    },
+    {
+      "epoch": 0.7979225448825189,
+      "grad_norm": 0.08968761563301086,
+      "learning_rate": 6.0774988109840045e-06,
+      "loss": 8.5511,
+      "step": 159780
+    },
+    {
+      "epoch": 0.7979724837074583,
+      "grad_norm": 0.09019223600625992,
+      "learning_rate": 6.075996896042455e-06,
+      "loss": 8.5646,
+      "step": 159790
+    },
+    {
+      "epoch": 0.7980224225323979,
+      "grad_norm": 0.09245903789997101,
+      "learning_rate": 6.074494981100903e-06,
+      "loss": 8.5769,
+      "step": 159800
+    },
+    {
+      "epoch": 0.7980723613573373,
+      "grad_norm": 0.08857943117618561,
+      "learning_rate": 6.072993066159353e-06,
+      "loss": 8.5494,
+      "step": 159810
+    },
+    {
+      "epoch": 0.7981223001822767,
+      "grad_norm": 0.09926991164684296,
+      "learning_rate": 6.0714911512178035e-06,
+      "loss": 8.569,
+      "step": 159820
+    },
+    {
+      "epoch": 0.7981722390072161,
+      "grad_norm": 0.09205451607704163,
+      "learning_rate": 6.069989236276252e-06,
+      "loss": 8.5501,
+      "step": 159830
+    },
+    {
+      "epoch": 0.7982221778321557,
+      "grad_norm": 0.09034121781587601,
+      "learning_rate": 6.068487321334702e-06,
+      "loss": 8.5465,
+      "step": 159840
+    },
+    {
+      "epoch": 0.7982721166570951,
+      "grad_norm": 0.08652927726507187,
+      "learning_rate": 6.0669854063931515e-06,
+      "loss": 8.5484,
+      "step": 159850
+    },
+    {
+      "epoch": 0.7983220554820345,
+      "grad_norm": 0.09696269780397415,
+      "learning_rate": 6.065483491451601e-06,
+      "loss": 8.5509,
+      "step": 159860
+    },
+    {
+      "epoch": 0.7983719943069739,
+      "grad_norm": 0.09118789434432983,
+      "learning_rate": 6.063981576510051e-06,
+      "loss": 8.5379,
+      "step": 159870
+    },
+    {
+      "epoch": 0.7984219331319135,
+      "grad_norm": 0.08937663584947586,
+      "learning_rate": 6.0624796615684995e-06,
+      "loss": 8.5392,
+      "step": 159880
+    },
+    {
+      "epoch": 0.7984718719568529,
+      "grad_norm": 0.08873353898525238,
+      "learning_rate": 6.06097774662695e-06,
+      "loss": 8.5635,
+      "step": 159890
+    },
+    {
+      "epoch": 0.7985218107817923,
+      "grad_norm": 0.0940752923488617,
+      "learning_rate": 6.059475831685399e-06,
+      "loss": 8.554,
+      "step": 159900
+    },
+    {
+      "epoch": 0.7985717496067317,
+      "grad_norm": 0.09297428280115128,
+      "learning_rate": 6.057973916743848e-06,
+      "loss": 8.5504,
+      "step": 159910
+    },
+    {
+      "epoch": 0.7986216884316713,
+      "grad_norm": 0.09415644407272339,
+      "learning_rate": 6.0564720018022985e-06,
+      "loss": 8.5671,
+      "step": 159920
+    },
+    {
+      "epoch": 0.7986716272566107,
+      "grad_norm": 0.08840908855199814,
+      "learning_rate": 6.054970086860748e-06,
+      "loss": 8.55,
+      "step": 159930
+    },
+    {
+      "epoch": 0.7987215660815501,
+      "grad_norm": 0.0910872295498848,
+      "learning_rate": 6.053468171919197e-06,
+      "loss": 8.5604,
+      "step": 159940
+    },
+    {
+      "epoch": 0.7987715049064895,
+      "grad_norm": 0.09535114467144012,
+      "learning_rate": 6.0519662569776465e-06,
+      "loss": 8.5668,
+      "step": 159950
+    },
+    {
+      "epoch": 0.798821443731429,
+      "grad_norm": 0.0967821255326271,
+      "learning_rate": 6.050464342036096e-06,
+      "loss": 8.5682,
+      "step": 159960
+    },
+    {
+      "epoch": 0.7988713825563685,
+      "grad_norm": 0.0884976014494896,
+      "learning_rate": 6.048962427094546e-06,
+      "loss": 8.5515,
+      "step": 159970
+    },
+    {
+      "epoch": 0.7989213213813079,
+      "grad_norm": 0.09149619191884995,
+      "learning_rate": 6.047460512152995e-06,
+      "loss": 8.5623,
+      "step": 159980
+    },
+    {
+      "epoch": 0.7989712602062473,
+      "grad_norm": 0.08907512575387955,
+      "learning_rate": 6.045958597211445e-06,
+      "loss": 8.558,
+      "step": 159990
+    },
+    {
+      "epoch": 0.7990211990311868,
+      "grad_norm": 0.0927252545952797,
+      "learning_rate": 6.044456682269894e-06,
+      "loss": 8.5367,
+      "step": 160000
+    },
+    {
+      "epoch": 0.7990711378561263,
+      "grad_norm": 0.09032353013753891,
+      "learning_rate": 6.042954767328344e-06,
+      "loss": 8.5554,
+      "step": 160010
+    },
+    {
+      "epoch": 0.7991210766810657,
+      "grad_norm": 0.09220920503139496,
+      "learning_rate": 6.0414528523867935e-06,
+      "loss": 8.5655,
+      "step": 160020
+    },
+    {
+      "epoch": 0.7991710155060051,
+      "grad_norm": 0.09536270052194595,
+      "learning_rate": 6.039950937445243e-06,
+      "loss": 8.5547,
+      "step": 160030
+    },
+    {
+      "epoch": 0.7992209543309446,
+      "grad_norm": 0.08931788057088852,
+      "learning_rate": 6.038449022503692e-06,
+      "loss": 8.5624,
+      "step": 160040
+    },
+    {
+      "epoch": 0.7992708931558841,
+      "grad_norm": 0.09477946907281876,
+      "learning_rate": 6.0369471075621415e-06,
+      "loss": 8.5688,
+      "step": 160050
+    },
+    {
+      "epoch": 0.7993208319808235,
+      "grad_norm": 0.09368328005075455,
+      "learning_rate": 6.035445192620592e-06,
+      "loss": 8.5624,
+      "step": 160060
+    },
+    {
+      "epoch": 0.7993707708057629,
+      "grad_norm": 0.09561215341091156,
+      "learning_rate": 6.033943277679041e-06,
+      "loss": 8.557,
+      "step": 160070
+    },
+    {
+      "epoch": 0.7994207096307024,
+      "grad_norm": 0.08733969926834106,
+      "learning_rate": 6.03244136273749e-06,
+      "loss": 8.5325,
+      "step": 160080
+    },
+    {
+      "epoch": 0.7994706484556419,
+      "grad_norm": 0.09188752621412277,
+      "learning_rate": 6.0309394477959405e-06,
+      "loss": 8.5459,
+      "step": 160090
+    },
+    {
+      "epoch": 0.7995205872805813,
+      "grad_norm": 0.09306297451257706,
+      "learning_rate": 6.029437532854389e-06,
+      "loss": 8.5527,
+      "step": 160100
+    },
+    {
+      "epoch": 0.7995705261055207,
+      "grad_norm": 0.09068245440721512,
+      "learning_rate": 6.027935617912839e-06,
+      "loss": 8.5579,
+      "step": 160110
+    },
+    {
+      "epoch": 0.7996204649304602,
+      "grad_norm": 0.09240369498729706,
+      "learning_rate": 6.0264337029712885e-06,
+      "loss": 8.5459,
+      "step": 160120
+    },
+    {
+      "epoch": 0.7996704037553997,
+      "grad_norm": 0.08835700154304504,
+      "learning_rate": 6.024931788029738e-06,
+      "loss": 8.552,
+      "step": 160130
+    },
+    {
+      "epoch": 0.7997203425803391,
+      "grad_norm": 0.08920862525701523,
+      "learning_rate": 6.023429873088188e-06,
+      "loss": 8.5368,
+      "step": 160140
+    },
+    {
+      "epoch": 0.7997702814052785,
+      "grad_norm": 0.09194939583539963,
+      "learning_rate": 6.0219279581466365e-06,
+      "loss": 8.5641,
+      "step": 160150
+    },
+    {
+      "epoch": 0.799820220230218,
+      "grad_norm": 0.09264218807220459,
+      "learning_rate": 6.020426043205087e-06,
+      "loss": 8.5531,
+      "step": 160160
+    },
+    {
+      "epoch": 0.7998701590551575,
+      "grad_norm": 0.08757354319095612,
+      "learning_rate": 6.018924128263537e-06,
+      "loss": 8.5538,
+      "step": 160170
+    },
+    {
+      "epoch": 0.7999200978800969,
+      "grad_norm": 0.09101276844739914,
+      "learning_rate": 6.017422213321985e-06,
+      "loss": 8.5525,
+      "step": 160180
+    },
+    {
+      "epoch": 0.7999700367050363,
+      "grad_norm": 0.09623536467552185,
+      "learning_rate": 6.0159202983804355e-06,
+      "loss": 8.5407,
+      "step": 160190
+    },
+    {
+      "epoch": 0.8000199755299757,
+      "grad_norm": 0.09045930206775665,
+      "learning_rate": 6.014418383438884e-06,
+      "loss": 8.5333,
+      "step": 160200
+    },
+    {
+      "epoch": 0.8000699143549153,
+      "grad_norm": 0.08761759847402573,
+      "learning_rate": 6.012916468497334e-06,
+      "loss": 8.552,
+      "step": 160210
+    },
+    {
+      "epoch": 0.8001198531798547,
+      "grad_norm": 0.09209952503442764,
+      "learning_rate": 6.011414553555784e-06,
+      "loss": 8.5461,
+      "step": 160220
+    },
+    {
+      "epoch": 0.8001697920047941,
+      "grad_norm": 0.08897832036018372,
+      "learning_rate": 6.009912638614233e-06,
+      "loss": 8.5743,
+      "step": 160230
+    },
+    {
+      "epoch": 0.8002197308297335,
+      "grad_norm": 0.09238340705633163,
+      "learning_rate": 6.008410723672683e-06,
+      "loss": 8.5463,
+      "step": 160240
+    },
+    {
+      "epoch": 0.800269669654673,
+      "grad_norm": 0.09207829087972641,
+      "learning_rate": 6.006908808731132e-06,
+      "loss": 8.5396,
+      "step": 160250
+    },
+    {
+      "epoch": 0.8003196084796125,
+      "grad_norm": 0.0933920294046402,
+      "learning_rate": 6.005406893789582e-06,
+      "loss": 8.5405,
+      "step": 160260
+    },
+    {
+      "epoch": 0.8003695473045519,
+      "grad_norm": 0.09098160266876221,
+      "learning_rate": 6.003904978848032e-06,
+      "loss": 8.548,
+      "step": 160270
+    },
+    {
+      "epoch": 0.8004194861294913,
+      "grad_norm": 0.0946149155497551,
+      "learning_rate": 6.00240306390648e-06,
+      "loss": 8.5488,
+      "step": 160280
+    },
+    {
+      "epoch": 0.8004694249544309,
+      "grad_norm": 0.09164818376302719,
+      "learning_rate": 6.0009011489649306e-06,
+      "loss": 8.5535,
+      "step": 160290
+    },
+    {
+      "epoch": 0.8005193637793703,
+      "grad_norm": 0.09267935901880264,
+      "learning_rate": 5.99939923402338e-06,
+      "loss": 8.571,
+      "step": 160300
+    },
+    {
+      "epoch": 0.8005693026043097,
+      "grad_norm": 0.0906684473156929,
+      "learning_rate": 5.997897319081829e-06,
+      "loss": 8.5541,
+      "step": 160310
+    },
+    {
+      "epoch": 0.8006192414292491,
+      "grad_norm": 0.09602951258420944,
+      "learning_rate": 5.996395404140279e-06,
+      "loss": 8.5577,
+      "step": 160320
+    },
+    {
+      "epoch": 0.8006691802541887,
+      "grad_norm": 0.09331289678812027,
+      "learning_rate": 5.994893489198729e-06,
+      "loss": 8.5534,
+      "step": 160330
+    },
+    {
+      "epoch": 0.8007191190791281,
+      "grad_norm": 0.0883229672908783,
+      "learning_rate": 5.993391574257178e-06,
+      "loss": 8.576,
+      "step": 160340
+    },
+    {
+      "epoch": 0.8007690579040675,
+      "grad_norm": 0.09256965667009354,
+      "learning_rate": 5.991889659315627e-06,
+      "loss": 8.5416,
+      "step": 160350
+    },
+    {
+      "epoch": 0.8008189967290069,
+      "grad_norm": 0.09165097028017044,
+      "learning_rate": 5.990387744374077e-06,
+      "loss": 8.5723,
+      "step": 160360
+    },
+    {
+      "epoch": 0.8008689355539464,
+      "grad_norm": 0.09180103242397308,
+      "learning_rate": 5.988885829432527e-06,
+      "loss": 8.5441,
+      "step": 160370
+    },
+    {
+      "epoch": 0.8009188743788859,
+      "grad_norm": 0.09208500385284424,
+      "learning_rate": 5.987383914490976e-06,
+      "loss": 8.5447,
+      "step": 160380
+    },
+    {
+      "epoch": 0.8009688132038253,
+      "grad_norm": 0.09202751517295837,
+      "learning_rate": 5.9858819995494256e-06,
+      "loss": 8.5482,
+      "step": 160390
+    },
+    {
+      "epoch": 0.8010187520287647,
+      "grad_norm": 0.09229252487421036,
+      "learning_rate": 5.984380084607875e-06,
+      "loss": 8.5692,
+      "step": 160400
+    },
+    {
+      "epoch": 0.8010686908537042,
+      "grad_norm": 0.09639665484428406,
+      "learning_rate": 5.982878169666325e-06,
+      "loss": 8.5544,
+      "step": 160410
+    },
+    {
+      "epoch": 0.8011186296786437,
+      "grad_norm": 0.08971371501684189,
+      "learning_rate": 5.981376254724774e-06,
+      "loss": 8.5706,
+      "step": 160420
+    },
+    {
+      "epoch": 0.8011685685035831,
+      "grad_norm": 0.08654212206602097,
+      "learning_rate": 5.979874339783224e-06,
+      "loss": 8.5672,
+      "step": 160430
+    },
+    {
+      "epoch": 0.8012185073285225,
+      "grad_norm": 0.09568259119987488,
+      "learning_rate": 5.978372424841673e-06,
+      "loss": 8.5543,
+      "step": 160440
+    },
+    {
+      "epoch": 0.801268446153462,
+      "grad_norm": 0.08771758526563644,
+      "learning_rate": 5.976870509900122e-06,
+      "loss": 8.5576,
+      "step": 160450
+    },
+    {
+      "epoch": 0.8013183849784015,
+      "grad_norm": 0.0924898236989975,
+      "learning_rate": 5.9753685949585726e-06,
+      "loss": 8.5352,
+      "step": 160460
+    },
+    {
+      "epoch": 0.8013683238033409,
+      "grad_norm": 0.08928176760673523,
+      "learning_rate": 5.973866680017022e-06,
+      "loss": 8.5789,
+      "step": 160470
+    },
+    {
+      "epoch": 0.8014182626282803,
+      "grad_norm": 0.08957140892744064,
+      "learning_rate": 5.972364765075471e-06,
+      "loss": 8.5642,
+      "step": 160480
+    },
+    {
+      "epoch": 0.8014682014532198,
+      "grad_norm": 0.0887003242969513,
+      "learning_rate": 5.970862850133921e-06,
+      "loss": 8.5642,
+      "step": 160490
+    },
+    {
+      "epoch": 0.8015181402781593,
+      "grad_norm": 0.09041450172662735,
+      "learning_rate": 5.96936093519237e-06,
+      "loss": 8.5659,
+      "step": 160500
+    },
+    {
+      "epoch": 0.8015680791030987,
+      "grad_norm": 0.08909320086240768,
+      "learning_rate": 5.96785902025082e-06,
+      "loss": 8.5378,
+      "step": 160510
+    },
+    {
+      "epoch": 0.8016180179280381,
+      "grad_norm": 0.09449964016675949,
+      "learning_rate": 5.966357105309269e-06,
+      "loss": 8.5644,
+      "step": 160520
+    },
+    {
+      "epoch": 0.8016679567529776,
+      "grad_norm": 0.0875280573964119,
+      "learning_rate": 5.964855190367719e-06,
+      "loss": 8.5552,
+      "step": 160530
+    },
+    {
+      "epoch": 0.8017178955779171,
+      "grad_norm": 0.09800990670919418,
+      "learning_rate": 5.963353275426169e-06,
+      "loss": 8.5692,
+      "step": 160540
+    },
+    {
+      "epoch": 0.8017678344028565,
+      "grad_norm": 0.10497017949819565,
+      "learning_rate": 5.961851360484617e-06,
+      "loss": 8.5514,
+      "step": 160550
+    },
+    {
+      "epoch": 0.8018177732277959,
+      "grad_norm": 0.09578979760408401,
+      "learning_rate": 5.960349445543068e-06,
+      "loss": 8.5585,
+      "step": 160560
+    },
+    {
+      "epoch": 0.8018677120527354,
+      "grad_norm": 0.0902516171336174,
+      "learning_rate": 5.958847530601518e-06,
+      "loss": 8.5426,
+      "step": 160570
+    },
+    {
+      "epoch": 0.8019176508776749,
+      "grad_norm": 0.087619349360466,
+      "learning_rate": 5.957345615659966e-06,
+      "loss": 8.5467,
+      "step": 160580
+    },
+    {
+      "epoch": 0.8019675897026143,
+      "grad_norm": 0.08661577105522156,
+      "learning_rate": 5.955843700718416e-06,
+      "loss": 8.5649,
+      "step": 160590
+    },
+    {
+      "epoch": 0.8020175285275537,
+      "grad_norm": 0.0935804471373558,
+      "learning_rate": 5.954341785776866e-06,
+      "loss": 8.558,
+      "step": 160600
+    },
+    {
+      "epoch": 0.8020674673524932,
+      "grad_norm": 0.08383012562990189,
+      "learning_rate": 5.952839870835315e-06,
+      "loss": 8.5579,
+      "step": 160610
+    },
+    {
+      "epoch": 0.8021174061774327,
+      "grad_norm": 0.09680390357971191,
+      "learning_rate": 5.951337955893765e-06,
+      "loss": 8.5733,
+      "step": 160620
+    },
+    {
+      "epoch": 0.8021673450023721,
+      "grad_norm": 0.0891953781247139,
+      "learning_rate": 5.949836040952214e-06,
+      "loss": 8.5736,
+      "step": 160630
+    },
+    {
+      "epoch": 0.8022172838273115,
+      "grad_norm": 0.0887405052781105,
+      "learning_rate": 5.948334126010664e-06,
+      "loss": 8.5685,
+      "step": 160640
+    },
+    {
+      "epoch": 0.802267222652251,
+      "grad_norm": 0.092610664665699,
+      "learning_rate": 5.946832211069113e-06,
+      "loss": 8.5667,
+      "step": 160650
+    },
+    {
+      "epoch": 0.8023171614771905,
+      "grad_norm": 0.08894860744476318,
+      "learning_rate": 5.945330296127563e-06,
+      "loss": 8.5653,
+      "step": 160660
+    },
+    {
+      "epoch": 0.8023671003021299,
+      "grad_norm": 0.09235876053571701,
+      "learning_rate": 5.943828381186013e-06,
+      "loss": 8.5525,
+      "step": 160670
+    },
+    {
+      "epoch": 0.8024170391270693,
+      "grad_norm": 0.0946565493941307,
+      "learning_rate": 5.942326466244461e-06,
+      "loss": 8.5589,
+      "step": 160680
+    },
+    {
+      "epoch": 0.8024669779520088,
+      "grad_norm": 0.08634993433952332,
+      "learning_rate": 5.9408245513029114e-06,
+      "loss": 8.5665,
+      "step": 160690
+    },
+    {
+      "epoch": 0.8025169167769483,
+      "grad_norm": 0.08891087025403976,
+      "learning_rate": 5.939322636361362e-06,
+      "loss": 8.5576,
+      "step": 160700
+    },
+    {
+      "epoch": 0.8025668556018877,
+      "grad_norm": 0.09388010948896408,
+      "learning_rate": 5.93782072141981e-06,
+      "loss": 8.5506,
+      "step": 160710
+    },
+    {
+      "epoch": 0.8026167944268271,
+      "grad_norm": 0.09189348667860031,
+      "learning_rate": 5.93631880647826e-06,
+      "loss": 8.554,
+      "step": 160720
+    },
+    {
+      "epoch": 0.8026667332517666,
+      "grad_norm": 0.08890991657972336,
+      "learning_rate": 5.93481689153671e-06,
+      "loss": 8.5491,
+      "step": 160730
+    },
+    {
+      "epoch": 0.802716672076706,
+      "grad_norm": 0.0901801586151123,
+      "learning_rate": 5.933314976595159e-06,
+      "loss": 8.5757,
+      "step": 160740
+    },
+    {
+      "epoch": 0.8027666109016455,
+      "grad_norm": 0.09369862824678421,
+      "learning_rate": 5.931813061653609e-06,
+      "loss": 8.5404,
+      "step": 160750
+    },
+    {
+      "epoch": 0.8028165497265849,
+      "grad_norm": 0.09540555626153946,
+      "learning_rate": 5.930311146712058e-06,
+      "loss": 8.552,
+      "step": 160760
+    },
+    {
+      "epoch": 0.8028664885515244,
+      "grad_norm": 0.09585050493478775,
+      "learning_rate": 5.928809231770508e-06,
+      "loss": 8.5557,
+      "step": 160770
+    },
+    {
+      "epoch": 0.8029164273764638,
+      "grad_norm": 0.088337741792202,
+      "learning_rate": 5.927307316828957e-06,
+      "loss": 8.5585,
+      "step": 160780
+    },
+    {
+      "epoch": 0.8029663662014033,
+      "grad_norm": 0.08791206032037735,
+      "learning_rate": 5.9258054018874064e-06,
+      "loss": 8.5687,
+      "step": 160790
+    },
+    {
+      "epoch": 0.8030163050263427,
+      "grad_norm": 0.09240417182445526,
+      "learning_rate": 5.924303486945857e-06,
+      "loss": 8.5421,
+      "step": 160800
+    },
+    {
+      "epoch": 0.8030662438512822,
+      "grad_norm": 0.08927211910486221,
+      "learning_rate": 5.922801572004306e-06,
+      "loss": 8.5463,
+      "step": 160810
+    },
+    {
+      "epoch": 0.8031161826762216,
+      "grad_norm": 0.09254935383796692,
+      "learning_rate": 5.921299657062755e-06,
+      "loss": 8.5417,
+      "step": 160820
+    },
+    {
+      "epoch": 0.8031661215011611,
+      "grad_norm": 0.09360701590776443,
+      "learning_rate": 5.919797742121205e-06,
+      "loss": 8.5466,
+      "step": 160830
+    },
+    {
+      "epoch": 0.8032160603261005,
+      "grad_norm": 0.09263437986373901,
+      "learning_rate": 5.918295827179654e-06,
+      "loss": 8.5576,
+      "step": 160840
+    },
+    {
+      "epoch": 0.80326599915104,
+      "grad_norm": 0.08811653405427933,
+      "learning_rate": 5.916793912238104e-06,
+      "loss": 8.5447,
+      "step": 160850
+    },
+    {
+      "epoch": 0.8033159379759794,
+      "grad_norm": 0.09161476045846939,
+      "learning_rate": 5.9152919972965535e-06,
+      "loss": 8.5664,
+      "step": 160860
+    },
+    {
+      "epoch": 0.8033658768009189,
+      "grad_norm": 0.08815915882587433,
+      "learning_rate": 5.913790082355003e-06,
+      "loss": 8.5393,
+      "step": 160870
+    },
+    {
+      "epoch": 0.8034158156258583,
+      "grad_norm": 0.08650946617126465,
+      "learning_rate": 5.912288167413452e-06,
+      "loss": 8.561,
+      "step": 160880
+    },
+    {
+      "epoch": 0.8034657544507978,
+      "grad_norm": 0.09026594460010529,
+      "learning_rate": 5.9107862524719014e-06,
+      "loss": 8.5529,
+      "step": 160890
+    },
+    {
+      "epoch": 0.8035156932757372,
+      "grad_norm": 0.08889591693878174,
+      "learning_rate": 5.909284337530352e-06,
+      "loss": 8.5591,
+      "step": 160900
+    },
+    {
+      "epoch": 0.8035656321006767,
+      "grad_norm": 0.09031382948160172,
+      "learning_rate": 5.907782422588801e-06,
+      "loss": 8.5549,
+      "step": 160910
+    },
+    {
+      "epoch": 0.8036155709256161,
+      "grad_norm": 0.08855296671390533,
+      "learning_rate": 5.90628050764725e-06,
+      "loss": 8.5692,
+      "step": 160920
+    },
+    {
+      "epoch": 0.8036655097505556,
+      "grad_norm": 0.09416501969099045,
+      "learning_rate": 5.9047785927057e-06,
+      "loss": 8.562,
+      "step": 160930
+    },
+    {
+      "epoch": 0.803715448575495,
+      "grad_norm": 0.09245479851961136,
+      "learning_rate": 5.90327667776415e-06,
+      "loss": 8.5468,
+      "step": 160940
+    },
+    {
+      "epoch": 0.8037653874004345,
+      "grad_norm": 0.09061010181903839,
+      "learning_rate": 5.901774762822599e-06,
+      "loss": 8.5644,
+      "step": 160950
+    },
+    {
+      "epoch": 0.8038153262253739,
+      "grad_norm": 0.08976545184850693,
+      "learning_rate": 5.9002728478810485e-06,
+      "loss": 8.5529,
+      "step": 160960
+    },
+    {
+      "epoch": 0.8038652650503134,
+      "grad_norm": 0.09560992568731308,
+      "learning_rate": 5.898770932939498e-06,
+      "loss": 8.5582,
+      "step": 160970
+    },
+    {
+      "epoch": 0.8039152038752528,
+      "grad_norm": 0.09577993303537369,
+      "learning_rate": 5.897269017997947e-06,
+      "loss": 8.558,
+      "step": 160980
+    },
+    {
+      "epoch": 0.8039651427001923,
+      "grad_norm": 0.08982730656862259,
+      "learning_rate": 5.895767103056397e-06,
+      "loss": 8.5651,
+      "step": 160990
+    },
+    {
+      "epoch": 0.8040150815251317,
+      "grad_norm": 0.08825762569904327,
+      "learning_rate": 5.894265188114847e-06,
+      "loss": 8.5539,
+      "step": 161000
+    },
+    {
+      "epoch": 0.8040650203500712,
+      "grad_norm": 0.0992155522108078,
+      "learning_rate": 5.892763273173296e-06,
+      "loss": 8.572,
+      "step": 161010
+    },
+    {
+      "epoch": 0.8041149591750106,
+      "grad_norm": 0.09141044318675995,
+      "learning_rate": 5.891261358231746e-06,
+      "loss": 8.5596,
+      "step": 161020
+    },
+    {
+      "epoch": 0.8041648979999501,
+      "grad_norm": 0.09531890600919724,
+      "learning_rate": 5.889759443290195e-06,
+      "loss": 8.5501,
+      "step": 161030
+    },
+    {
+      "epoch": 0.8042148368248895,
+      "grad_norm": 0.08679885417222977,
+      "learning_rate": 5.888257528348645e-06,
+      "loss": 8.5506,
+      "step": 161040
+    },
+    {
+      "epoch": 0.804264775649829,
+      "grad_norm": 0.08994447439908981,
+      "learning_rate": 5.886755613407094e-06,
+      "loss": 8.5481,
+      "step": 161050
+    },
+    {
+      "epoch": 0.8043147144747684,
+      "grad_norm": 0.09019139409065247,
+      "learning_rate": 5.8852536984655435e-06,
+      "loss": 8.5627,
+      "step": 161060
+    },
+    {
+      "epoch": 0.8043646532997079,
+      "grad_norm": 0.08854031562805176,
+      "learning_rate": 5.883751783523994e-06,
+      "loss": 8.5427,
+      "step": 161070
+    },
+    {
+      "epoch": 0.8044145921246473,
+      "grad_norm": 0.0935094803571701,
+      "learning_rate": 5.882249868582442e-06,
+      "loss": 8.5412,
+      "step": 161080
+    },
+    {
+      "epoch": 0.8044645309495868,
+      "grad_norm": 0.08779904246330261,
+      "learning_rate": 5.880747953640892e-06,
+      "loss": 8.5639,
+      "step": 161090
+    },
+    {
+      "epoch": 0.8045144697745262,
+      "grad_norm": 0.0949099212884903,
+      "learning_rate": 5.8792460386993425e-06,
+      "loss": 8.5561,
+      "step": 161100
+    },
+    {
+      "epoch": 0.8045644085994657,
+      "grad_norm": 0.0882745087146759,
+      "learning_rate": 5.877744123757791e-06,
+      "loss": 8.5621,
+      "step": 161110
+    },
+    {
+      "epoch": 0.8046143474244051,
+      "grad_norm": 0.09196063131093979,
+      "learning_rate": 5.876242208816241e-06,
+      "loss": 8.5656,
+      "step": 161120
+    },
+    {
+      "epoch": 0.8046642862493446,
+      "grad_norm": 0.08640565723180771,
+      "learning_rate": 5.87474029387469e-06,
+      "loss": 8.5621,
+      "step": 161130
+    },
+    {
+      "epoch": 0.804714225074284,
+      "grad_norm": 0.09424121677875519,
+      "learning_rate": 5.87323837893314e-06,
+      "loss": 8.5532,
+      "step": 161140
+    },
+    {
+      "epoch": 0.8047641638992235,
+      "grad_norm": 0.08842575550079346,
+      "learning_rate": 5.87173646399159e-06,
+      "loss": 8.561,
+      "step": 161150
+    },
+    {
+      "epoch": 0.8048141027241629,
+      "grad_norm": 0.09014535695314407,
+      "learning_rate": 5.8702345490500385e-06,
+      "loss": 8.5514,
+      "step": 161160
+    },
+    {
+      "epoch": 0.8048640415491024,
+      "grad_norm": 0.09411894530057907,
+      "learning_rate": 5.868732634108489e-06,
+      "loss": 8.544,
+      "step": 161170
+    },
+    {
+      "epoch": 0.8049139803740418,
+      "grad_norm": 0.09182292968034744,
+      "learning_rate": 5.867230719166938e-06,
+      "loss": 8.5573,
+      "step": 161180
+    },
+    {
+      "epoch": 0.8049639191989812,
+      "grad_norm": 0.08946768194437027,
+      "learning_rate": 5.865728804225387e-06,
+      "loss": 8.5418,
+      "step": 161190
+    },
+    {
+      "epoch": 0.8050138580239207,
+      "grad_norm": 0.08927719295024872,
+      "learning_rate": 5.8642268892838375e-06,
+      "loss": 8.5637,
+      "step": 161200
+    },
+    {
+      "epoch": 0.8050637968488601,
+      "grad_norm": 0.09145810455083847,
+      "learning_rate": 5.862724974342286e-06,
+      "loss": 8.5511,
+      "step": 161210
+    },
+    {
+      "epoch": 0.8051137356737996,
+      "grad_norm": 0.09392247349023819,
+      "learning_rate": 5.861223059400736e-06,
+      "loss": 8.5375,
+      "step": 161220
+    },
+    {
+      "epoch": 0.805163674498739,
+      "grad_norm": 0.09256572276353836,
+      "learning_rate": 5.8597211444591855e-06,
+      "loss": 8.5547,
+      "step": 161230
+    },
+    {
+      "epoch": 0.8052136133236785,
+      "grad_norm": 0.08906706422567368,
+      "learning_rate": 5.858219229517635e-06,
+      "loss": 8.5664,
+      "step": 161240
+    },
+    {
+      "epoch": 0.8052635521486179,
+      "grad_norm": 0.08576681464910507,
+      "learning_rate": 5.856717314576085e-06,
+      "loss": 8.5525,
+      "step": 161250
+    },
+    {
+      "epoch": 0.8053134909735574,
+      "grad_norm": 0.09053860604763031,
+      "learning_rate": 5.855215399634534e-06,
+      "loss": 8.5489,
+      "step": 161260
+    },
+    {
+      "epoch": 0.8053634297984968,
+      "grad_norm": 0.08544787019491196,
+      "learning_rate": 5.853713484692984e-06,
+      "loss": 8.5576,
+      "step": 161270
+    },
+    {
+      "epoch": 0.8054133686234363,
+      "grad_norm": 0.0925343781709671,
+      "learning_rate": 5.852211569751433e-06,
+      "loss": 8.5416,
+      "step": 161280
+    },
+    {
+      "epoch": 0.8054633074483757,
+      "grad_norm": 0.09434544295072556,
+      "learning_rate": 5.850709654809882e-06,
+      "loss": 8.5593,
+      "step": 161290
+    },
+    {
+      "epoch": 0.8055132462733152,
+      "grad_norm": 0.09324178099632263,
+      "learning_rate": 5.8492077398683325e-06,
+      "loss": 8.5498,
+      "step": 161300
+    },
+    {
+      "epoch": 0.8055631850982546,
+      "grad_norm": 0.09053826332092285,
+      "learning_rate": 5.847705824926782e-06,
+      "loss": 8.5522,
+      "step": 161310
+    },
+    {
+      "epoch": 0.8056131239231941,
+      "grad_norm": 0.09149667620658875,
+      "learning_rate": 5.846203909985231e-06,
+      "loss": 8.536,
+      "step": 161320
+    },
+    {
+      "epoch": 0.8056630627481335,
+      "grad_norm": 0.08661210536956787,
+      "learning_rate": 5.8447019950436805e-06,
+      "loss": 8.5475,
+      "step": 161330
+    },
+    {
+      "epoch": 0.805713001573073,
+      "grad_norm": 0.09454959630966187,
+      "learning_rate": 5.843200080102131e-06,
+      "loss": 8.5584,
+      "step": 161340
+    },
+    {
+      "epoch": 0.8057629403980124,
+      "grad_norm": 0.09166700392961502,
+      "learning_rate": 5.84169816516058e-06,
+      "loss": 8.5538,
+      "step": 161350
+    },
+    {
+      "epoch": 0.8058128792229519,
+      "grad_norm": 0.08817677199840546,
+      "learning_rate": 5.840196250219029e-06,
+      "loss": 8.5504,
+      "step": 161360
+    },
+    {
+      "epoch": 0.8058628180478913,
+      "grad_norm": 0.0908307358622551,
+      "learning_rate": 5.838694335277479e-06,
+      "loss": 8.5528,
+      "step": 161370
+    },
+    {
+      "epoch": 0.8059127568728308,
+      "grad_norm": 0.10027515888214111,
+      "learning_rate": 5.837192420335928e-06,
+      "loss": 8.5598,
+      "step": 161380
+    },
+    {
+      "epoch": 0.8059626956977702,
+      "grad_norm": 0.09583675861358643,
+      "learning_rate": 5.835690505394378e-06,
+      "loss": 8.5463,
+      "step": 161390
+    },
+    {
+      "epoch": 0.8060126345227097,
+      "grad_norm": 0.0983138233423233,
+      "learning_rate": 5.8341885904528275e-06,
+      "loss": 8.5588,
+      "step": 161400
+    },
+    {
+      "epoch": 0.8060625733476491,
+      "grad_norm": 0.08900081366300583,
+      "learning_rate": 5.832686675511277e-06,
+      "loss": 8.5361,
+      "step": 161410
+    },
+    {
+      "epoch": 0.8061125121725886,
+      "grad_norm": 0.093843013048172,
+      "learning_rate": 5.831184760569727e-06,
+      "loss": 8.5498,
+      "step": 161420
+    },
+    {
+      "epoch": 0.806162450997528,
+      "grad_norm": 0.09397601336240768,
+      "learning_rate": 5.8296828456281755e-06,
+      "loss": 8.5497,
+      "step": 161430
+    },
+    {
+      "epoch": 0.8062123898224675,
+      "grad_norm": 0.09373549371957779,
+      "learning_rate": 5.828180930686626e-06,
+      "loss": 8.563,
+      "step": 161440
+    },
+    {
+      "epoch": 0.8062623286474069,
+      "grad_norm": 0.09314602613449097,
+      "learning_rate": 5.826679015745075e-06,
+      "loss": 8.5504,
+      "step": 161450
+    },
+    {
+      "epoch": 0.8063122674723464,
+      "grad_norm": 0.09471584111452103,
+      "learning_rate": 5.825177100803524e-06,
+      "loss": 8.5783,
+      "step": 161460
+    },
+    {
+      "epoch": 0.8063622062972858,
+      "grad_norm": 0.08868781477212906,
+      "learning_rate": 5.8236751858619745e-06,
+      "loss": 8.5653,
+      "step": 161470
+    },
+    {
+      "epoch": 0.8064121451222253,
+      "grad_norm": 0.09533412009477615,
+      "learning_rate": 5.822173270920423e-06,
+      "loss": 8.5538,
+      "step": 161480
+    },
+    {
+      "epoch": 0.8064620839471647,
+      "grad_norm": 0.09250637888908386,
+      "learning_rate": 5.820671355978873e-06,
+      "loss": 8.5697,
+      "step": 161490
+    },
+    {
+      "epoch": 0.8065120227721042,
+      "grad_norm": 0.08923783898353577,
+      "learning_rate": 5.819169441037323e-06,
+      "loss": 8.5524,
+      "step": 161500
+    },
+    {
+      "epoch": 0.8065619615970436,
+      "grad_norm": 0.09027642011642456,
+      "learning_rate": 5.817667526095772e-06,
+      "loss": 8.5462,
+      "step": 161510
+    },
+    {
+      "epoch": 0.806611900421983,
+      "grad_norm": 0.08848720788955688,
+      "learning_rate": 5.816165611154222e-06,
+      "loss": 8.547,
+      "step": 161520
+    },
+    {
+      "epoch": 0.8066618392469225,
+      "grad_norm": 0.09159193187952042,
+      "learning_rate": 5.8146636962126705e-06,
+      "loss": 8.5562,
+      "step": 161530
+    },
+    {
+      "epoch": 0.806711778071862,
+      "grad_norm": 0.09201356023550034,
+      "learning_rate": 5.813161781271121e-06,
+      "loss": 8.5494,
+      "step": 161540
+    },
+    {
+      "epoch": 0.8067617168968014,
+      "grad_norm": 0.090115986764431,
+      "learning_rate": 5.811659866329571e-06,
+      "loss": 8.543,
+      "step": 161550
+    },
+    {
+      "epoch": 0.8068116557217409,
+      "grad_norm": 0.08883675187826157,
+      "learning_rate": 5.810157951388019e-06,
+      "loss": 8.566,
+      "step": 161560
+    },
+    {
+      "epoch": 0.8068615945466803,
+      "grad_norm": 0.09100556373596191,
+      "learning_rate": 5.8086560364464695e-06,
+      "loss": 8.5675,
+      "step": 161570
+    },
+    {
+      "epoch": 0.8069115333716198,
+      "grad_norm": 0.09314378350973129,
+      "learning_rate": 5.807154121504919e-06,
+      "loss": 8.5437,
+      "step": 161580
+    },
+    {
+      "epoch": 0.8069614721965592,
+      "grad_norm": 0.09172666072845459,
+      "learning_rate": 5.805652206563368e-06,
+      "loss": 8.5589,
+      "step": 161590
+    },
+    {
+      "epoch": 0.8070114110214986,
+      "grad_norm": 0.09398376941680908,
+      "learning_rate": 5.804150291621818e-06,
+      "loss": 8.5466,
+      "step": 161600
+    },
+    {
+      "epoch": 0.8070613498464381,
+      "grad_norm": 0.08542805910110474,
+      "learning_rate": 5.802648376680267e-06,
+      "loss": 8.5673,
+      "step": 161610
+    },
+    {
+      "epoch": 0.8071112886713776,
+      "grad_norm": 0.08974550664424896,
+      "learning_rate": 5.801146461738717e-06,
+      "loss": 8.5524,
+      "step": 161620
+    },
+    {
+      "epoch": 0.807161227496317,
+      "grad_norm": 0.09220361709594727,
+      "learning_rate": 5.799644546797166e-06,
+      "loss": 8.5404,
+      "step": 161630
+    },
+    {
+      "epoch": 0.8072111663212564,
+      "grad_norm": 0.08950573951005936,
+      "learning_rate": 5.798142631855616e-06,
+      "loss": 8.5477,
+      "step": 161640
+    },
+    {
+      "epoch": 0.8072611051461959,
+      "grad_norm": 0.0872855931520462,
+      "learning_rate": 5.796640716914066e-06,
+      "loss": 8.5645,
+      "step": 161650
+    },
+    {
+      "epoch": 0.8073110439711354,
+      "grad_norm": 0.09016703069210052,
+      "learning_rate": 5.795138801972515e-06,
+      "loss": 8.5501,
+      "step": 161660
+    },
+    {
+      "epoch": 0.8073609827960748,
+      "grad_norm": 0.09285937994718552,
+      "learning_rate": 5.7936368870309645e-06,
+      "loss": 8.5527,
+      "step": 161670
+    },
+    {
+      "epoch": 0.8074109216210142,
+      "grad_norm": 0.09346285462379456,
+      "learning_rate": 5.792134972089414e-06,
+      "loss": 8.5559,
+      "step": 161680
+    },
+    {
+      "epoch": 0.8074608604459537,
+      "grad_norm": 0.08668027818202972,
+      "learning_rate": 5.790633057147863e-06,
+      "loss": 8.5425,
+      "step": 161690
+    },
+    {
+      "epoch": 0.8075107992708932,
+      "grad_norm": 0.0898892879486084,
+      "learning_rate": 5.789131142206313e-06,
+      "loss": 8.5491,
+      "step": 161700
+    },
+    {
+      "epoch": 0.8075607380958326,
+      "grad_norm": 0.09450709819793701,
+      "learning_rate": 5.787629227264763e-06,
+      "loss": 8.5568,
+      "step": 161710
+    },
+    {
+      "epoch": 0.807610676920772,
+      "grad_norm": 0.08670226484537125,
+      "learning_rate": 5.786127312323212e-06,
+      "loss": 8.5596,
+      "step": 161720
+    },
+    {
+      "epoch": 0.8076606157457115,
+      "grad_norm": 0.08837024122476578,
+      "learning_rate": 5.784625397381662e-06,
+      "loss": 8.5445,
+      "step": 161730
+    },
+    {
+      "epoch": 0.807710554570651,
+      "grad_norm": 0.093661367893219,
+      "learning_rate": 5.7831234824401115e-06,
+      "loss": 8.5325,
+      "step": 161740
+    },
+    {
+      "epoch": 0.8077604933955904,
+      "grad_norm": 0.0894351676106453,
+      "learning_rate": 5.781621567498561e-06,
+      "loss": 8.551,
+      "step": 161750
+    },
+    {
+      "epoch": 0.8078104322205298,
+      "grad_norm": 0.08936372399330139,
+      "learning_rate": 5.78011965255701e-06,
+      "loss": 8.5591,
+      "step": 161760
+    },
+    {
+      "epoch": 0.8078603710454693,
+      "grad_norm": 0.09095845371484756,
+      "learning_rate": 5.7786177376154595e-06,
+      "loss": 8.5491,
+      "step": 161770
+    },
+    {
+      "epoch": 0.8079103098704088,
+      "grad_norm": 0.09978078305721283,
+      "learning_rate": 5.77711582267391e-06,
+      "loss": 8.5385,
+      "step": 161780
+    },
+    {
+      "epoch": 0.8079602486953482,
+      "grad_norm": 0.0931134968996048,
+      "learning_rate": 5.775613907732359e-06,
+      "loss": 8.554,
+      "step": 161790
+    },
+    {
+      "epoch": 0.8080101875202876,
+      "grad_norm": 0.09244008362293243,
+      "learning_rate": 5.774111992790808e-06,
+      "loss": 8.5561,
+      "step": 161800
+    },
+    {
+      "epoch": 0.8080601263452271,
+      "grad_norm": 0.09390776604413986,
+      "learning_rate": 5.772610077849258e-06,
+      "loss": 8.5361,
+      "step": 161810
+    },
+    {
+      "epoch": 0.8081100651701666,
+      "grad_norm": 0.09701921045780182,
+      "learning_rate": 5.771108162907708e-06,
+      "loss": 8.5554,
+      "step": 161820
+    },
+    {
+      "epoch": 0.808160003995106,
+      "grad_norm": 0.09131251275539398,
+      "learning_rate": 5.769606247966157e-06,
+      "loss": 8.5438,
+      "step": 161830
+    },
+    {
+      "epoch": 0.8082099428200454,
+      "grad_norm": 0.09114990383386612,
+      "learning_rate": 5.7681043330246066e-06,
+      "loss": 8.5591,
+      "step": 161840
+    },
+    {
+      "epoch": 0.8082598816449849,
+      "grad_norm": 0.08846677094697952,
+      "learning_rate": 5.766602418083056e-06,
+      "loss": 8.5434,
+      "step": 161850
+    },
+    {
+      "epoch": 0.8083098204699244,
+      "grad_norm": 0.08849672228097916,
+      "learning_rate": 5.765100503141505e-06,
+      "loss": 8.5666,
+      "step": 161860
+    },
+    {
+      "epoch": 0.8083597592948638,
+      "grad_norm": 0.0881495401263237,
+      "learning_rate": 5.763598588199955e-06,
+      "loss": 8.5384,
+      "step": 161870
+    },
+    {
+      "epoch": 0.8084096981198032,
+      "grad_norm": 0.09564726799726486,
+      "learning_rate": 5.762096673258405e-06,
+      "loss": 8.5536,
+      "step": 161880
+    },
+    {
+      "epoch": 0.8084596369447427,
+      "grad_norm": 0.08935084193944931,
+      "learning_rate": 5.760594758316854e-06,
+      "loss": 8.5496,
+      "step": 161890
+    },
+    {
+      "epoch": 0.8085095757696822,
+      "grad_norm": 0.09384889900684357,
+      "learning_rate": 5.759092843375304e-06,
+      "loss": 8.5448,
+      "step": 161900
+    },
+    {
+      "epoch": 0.8085595145946216,
+      "grad_norm": 0.08726388961076736,
+      "learning_rate": 5.757590928433753e-06,
+      "loss": 8.5489,
+      "step": 161910
+    },
+    {
+      "epoch": 0.808609453419561,
+      "grad_norm": 0.09072273224592209,
+      "learning_rate": 5.756089013492203e-06,
+      "loss": 8.5589,
+      "step": 161920
+    },
+    {
+      "epoch": 0.8086593922445005,
+      "grad_norm": 0.09182450920343399,
+      "learning_rate": 5.754587098550652e-06,
+      "loss": 8.5484,
+      "step": 161930
+    },
+    {
+      "epoch": 0.80870933106944,
+      "grad_norm": 0.08764579892158508,
+      "learning_rate": 5.7530851836091016e-06,
+      "loss": 8.5622,
+      "step": 161940
+    },
+    {
+      "epoch": 0.8087592698943794,
+      "grad_norm": 0.0898037999868393,
+      "learning_rate": 5.751583268667552e-06,
+      "loss": 8.5555,
+      "step": 161950
+    },
+    {
+      "epoch": 0.8088092087193188,
+      "grad_norm": 0.09432199597358704,
+      "learning_rate": 5.750081353726e-06,
+      "loss": 8.5663,
+      "step": 161960
+    },
+    {
+      "epoch": 0.8088591475442583,
+      "grad_norm": 0.09024576097726822,
+      "learning_rate": 5.74857943878445e-06,
+      "loss": 8.5594,
+      "step": 161970
+    },
+    {
+      "epoch": 0.8089090863691978,
+      "grad_norm": 0.08990621566772461,
+      "learning_rate": 5.747077523842901e-06,
+      "loss": 8.5471,
+      "step": 161980
+    },
+    {
+      "epoch": 0.8089590251941372,
+      "grad_norm": 0.09168119728565216,
+      "learning_rate": 5.745575608901349e-06,
+      "loss": 8.5391,
+      "step": 161990
+    },
+    {
+      "epoch": 0.8090089640190766,
+      "grad_norm": 0.09170830249786377,
+      "learning_rate": 5.744073693959799e-06,
+      "loss": 8.565,
+      "step": 162000
+    },
+    {
+      "epoch": 0.809058902844016,
+      "grad_norm": 0.08952143043279648,
+      "learning_rate": 5.742571779018248e-06,
+      "loss": 8.5513,
+      "step": 162010
+    },
+    {
+      "epoch": 0.8091088416689556,
+      "grad_norm": 0.08867137879133224,
+      "learning_rate": 5.741069864076698e-06,
+      "loss": 8.5306,
+      "step": 162020
+    },
+    {
+      "epoch": 0.809158780493895,
+      "grad_norm": 0.09371023625135422,
+      "learning_rate": 5.739567949135148e-06,
+      "loss": 8.542,
+      "step": 162030
+    },
+    {
+      "epoch": 0.8092087193188344,
+      "grad_norm": 0.09115874022245407,
+      "learning_rate": 5.7380660341935966e-06,
+      "loss": 8.5642,
+      "step": 162040
+    },
+    {
+      "epoch": 0.8092586581437738,
+      "grad_norm": 0.09029915928840637,
+      "learning_rate": 5.736564119252047e-06,
+      "loss": 8.5507,
+      "step": 162050
+    },
+    {
+      "epoch": 0.8093085969687134,
+      "grad_norm": 0.09160950779914856,
+      "learning_rate": 5.735062204310496e-06,
+      "loss": 8.5746,
+      "step": 162060
+    },
+    {
+      "epoch": 0.8093585357936528,
+      "grad_norm": 0.09389065206050873,
+      "learning_rate": 5.733560289368945e-06,
+      "loss": 8.5682,
+      "step": 162070
+    },
+    {
+      "epoch": 0.8094084746185922,
+      "grad_norm": 0.08822131901979446,
+      "learning_rate": 5.732058374427396e-06,
+      "loss": 8.5532,
+      "step": 162080
+    },
+    {
+      "epoch": 0.8094584134435316,
+      "grad_norm": 0.09017851203680038,
+      "learning_rate": 5.730556459485844e-06,
+      "loss": 8.559,
+      "step": 162090
+    },
+    {
+      "epoch": 0.8095083522684712,
+      "grad_norm": 0.09044285118579865,
+      "learning_rate": 5.729054544544294e-06,
+      "loss": 8.5549,
+      "step": 162100
+    },
+    {
+      "epoch": 0.8095582910934106,
+      "grad_norm": 0.09080391377210617,
+      "learning_rate": 5.727552629602744e-06,
+      "loss": 8.5327,
+      "step": 162110
+    },
+    {
+      "epoch": 0.80960822991835,
+      "grad_norm": 0.08978559076786041,
+      "learning_rate": 5.726050714661193e-06,
+      "loss": 8.5409,
+      "step": 162120
+    },
+    {
+      "epoch": 0.8096581687432894,
+      "grad_norm": 0.09273146092891693,
+      "learning_rate": 5.724548799719643e-06,
+      "loss": 8.5516,
+      "step": 162130
+    },
+    {
+      "epoch": 0.809708107568229,
+      "grad_norm": 0.09049143642187119,
+      "learning_rate": 5.7230468847780924e-06,
+      "loss": 8.5536,
+      "step": 162140
+    },
+    {
+      "epoch": 0.8097580463931684,
+      "grad_norm": 0.09314949065446854,
+      "learning_rate": 5.721544969836542e-06,
+      "loss": 8.5574,
+      "step": 162150
+    },
+    {
+      "epoch": 0.8098079852181078,
+      "grad_norm": 0.09063813835382462,
+      "learning_rate": 5.720043054894991e-06,
+      "loss": 8.5568,
+      "step": 162160
+    },
+    {
+      "epoch": 0.8098579240430472,
+      "grad_norm": 0.09309311211109161,
+      "learning_rate": 5.71854113995344e-06,
+      "loss": 8.5452,
+      "step": 162170
+    },
+    {
+      "epoch": 0.8099078628679867,
+      "grad_norm": 0.09010675549507141,
+      "learning_rate": 5.717039225011891e-06,
+      "loss": 8.5491,
+      "step": 162180
+    },
+    {
+      "epoch": 0.8099578016929262,
+      "grad_norm": 0.09501739591360092,
+      "learning_rate": 5.71553731007034e-06,
+      "loss": 8.5697,
+      "step": 162190
+    },
+    {
+      "epoch": 0.8100077405178656,
+      "grad_norm": 0.09098471701145172,
+      "learning_rate": 5.714035395128789e-06,
+      "loss": 8.5599,
+      "step": 162200
+    },
+    {
+      "epoch": 0.810057679342805,
+      "grad_norm": 0.0888034850358963,
+      "learning_rate": 5.712533480187239e-06,
+      "loss": 8.5362,
+      "step": 162210
+    },
+    {
+      "epoch": 0.8101076181677445,
+      "grad_norm": 0.09654735028743744,
+      "learning_rate": 5.711031565245689e-06,
+      "loss": 8.5474,
+      "step": 162220
+    },
+    {
+      "epoch": 0.810157556992684,
+      "grad_norm": 0.09235084056854248,
+      "learning_rate": 5.709529650304138e-06,
+      "loss": 8.5447,
+      "step": 162230
+    },
+    {
+      "epoch": 0.8102074958176234,
+      "grad_norm": 0.09002131223678589,
+      "learning_rate": 5.7080277353625874e-06,
+      "loss": 8.5462,
+      "step": 162240
+    },
+    {
+      "epoch": 0.8102574346425628,
+      "grad_norm": 0.08661506325006485,
+      "learning_rate": 5.706525820421037e-06,
+      "loss": 8.5672,
+      "step": 162250
+    },
+    {
+      "epoch": 0.8103073734675023,
+      "grad_norm": 0.08980439603328705,
+      "learning_rate": 5.705023905479486e-06,
+      "loss": 8.5427,
+      "step": 162260
+    },
+    {
+      "epoch": 0.8103573122924418,
+      "grad_norm": 0.08393685519695282,
+      "learning_rate": 5.703521990537936e-06,
+      "loss": 8.5484,
+      "step": 162270
+    },
+    {
+      "epoch": 0.8104072511173812,
+      "grad_norm": 0.09369827061891556,
+      "learning_rate": 5.702020075596386e-06,
+      "loss": 8.5434,
+      "step": 162280
+    },
+    {
+      "epoch": 0.8104571899423206,
+      "grad_norm": 0.09700123220682144,
+      "learning_rate": 5.700518160654835e-06,
+      "loss": 8.5523,
+      "step": 162290
+    },
+    {
+      "epoch": 0.81050712876726,
+      "grad_norm": 0.08638820052146912,
+      "learning_rate": 5.699016245713285e-06,
+      "loss": 8.5611,
+      "step": 162300
+    },
+    {
+      "epoch": 0.8105570675921996,
+      "grad_norm": 0.09503623843193054,
+      "learning_rate": 5.697514330771734e-06,
+      "loss": 8.5216,
+      "step": 162310
+    },
+    {
+      "epoch": 0.810607006417139,
+      "grad_norm": 0.09698504954576492,
+      "learning_rate": 5.696012415830184e-06,
+      "loss": 8.5362,
+      "step": 162320
+    },
+    {
+      "epoch": 0.8106569452420784,
+      "grad_norm": 0.09360737353563309,
+      "learning_rate": 5.694510500888633e-06,
+      "loss": 8.5411,
+      "step": 162330
+    },
+    {
+      "epoch": 0.8107068840670179,
+      "grad_norm": 0.08911336213350296,
+      "learning_rate": 5.6930085859470824e-06,
+      "loss": 8.5541,
+      "step": 162340
+    },
+    {
+      "epoch": 0.8107568228919574,
+      "grad_norm": 0.09235373139381409,
+      "learning_rate": 5.691506671005533e-06,
+      "loss": 8.5529,
+      "step": 162350
+    },
+    {
+      "epoch": 0.8108067617168968,
+      "grad_norm": 0.09205646812915802,
+      "learning_rate": 5.690004756063981e-06,
+      "loss": 8.5628,
+      "step": 162360
+    },
+    {
+      "epoch": 0.8108567005418362,
+      "grad_norm": 0.08802807331085205,
+      "learning_rate": 5.688502841122431e-06,
+      "loss": 8.5639,
+      "step": 162370
+    },
+    {
+      "epoch": 0.8109066393667757,
+      "grad_norm": 0.09021410346031189,
+      "learning_rate": 5.6870009261808815e-06,
+      "loss": 8.5609,
+      "step": 162380
+    },
+    {
+      "epoch": 0.8109565781917152,
+      "grad_norm": 0.09240546077489853,
+      "learning_rate": 5.68549901123933e-06,
+      "loss": 8.5595,
+      "step": 162390
+    },
+    {
+      "epoch": 0.8110065170166546,
+      "grad_norm": 0.08953437209129333,
+      "learning_rate": 5.68399709629778e-06,
+      "loss": 8.5495,
+      "step": 162400
+    },
+    {
+      "epoch": 0.811056455841594,
+      "grad_norm": 0.09370911866426468,
+      "learning_rate": 5.682495181356229e-06,
+      "loss": 8.541,
+      "step": 162410
+    },
+    {
+      "epoch": 0.8111063946665334,
+      "grad_norm": 0.09173066914081573,
+      "learning_rate": 5.680993266414679e-06,
+      "loss": 8.5352,
+      "step": 162420
+    },
+    {
+      "epoch": 0.811156333491473,
+      "grad_norm": 0.08917384594678879,
+      "learning_rate": 5.679491351473129e-06,
+      "loss": 8.56,
+      "step": 162430
+    },
+    {
+      "epoch": 0.8112062723164124,
+      "grad_norm": 0.09332337230443954,
+      "learning_rate": 5.6779894365315774e-06,
+      "loss": 8.5452,
+      "step": 162440
+    },
+    {
+      "epoch": 0.8112562111413518,
+      "grad_norm": 0.08929426968097687,
+      "learning_rate": 5.676487521590028e-06,
+      "loss": 8.5625,
+      "step": 162450
+    },
+    {
+      "epoch": 0.8113061499662912,
+      "grad_norm": 0.08893352746963501,
+      "learning_rate": 5.674985606648477e-06,
+      "loss": 8.5519,
+      "step": 162460
+    },
+    {
+      "epoch": 0.8113560887912308,
+      "grad_norm": 0.0881025567650795,
+      "learning_rate": 5.673483691706926e-06,
+      "loss": 8.5514,
+      "step": 162470
+    },
+    {
+      "epoch": 0.8114060276161702,
+      "grad_norm": 0.09232950210571289,
+      "learning_rate": 5.6719817767653765e-06,
+      "loss": 8.563,
+      "step": 162480
+    },
+    {
+      "epoch": 0.8114559664411096,
+      "grad_norm": 0.09270848333835602,
+      "learning_rate": 5.670479861823825e-06,
+      "loss": 8.5434,
+      "step": 162490
+    },
+    {
+      "epoch": 0.811505905266049,
+      "grad_norm": 0.0903157889842987,
+      "learning_rate": 5.668977946882275e-06,
+      "loss": 8.5461,
+      "step": 162500
+    },
+    {
+      "epoch": 0.8115558440909886,
+      "grad_norm": 0.09540119767189026,
+      "learning_rate": 5.6674760319407245e-06,
+      "loss": 8.5403,
+      "step": 162510
+    },
+    {
+      "epoch": 0.811605782915928,
+      "grad_norm": 0.09533382952213287,
+      "learning_rate": 5.665974116999174e-06,
+      "loss": 8.5465,
+      "step": 162520
+    },
+    {
+      "epoch": 0.8116557217408674,
+      "grad_norm": 0.09012457728385925,
+      "learning_rate": 5.664472202057624e-06,
+      "loss": 8.5571,
+      "step": 162530
+    },
+    {
+      "epoch": 0.8117056605658068,
+      "grad_norm": 0.09900517761707306,
+      "learning_rate": 5.662970287116073e-06,
+      "loss": 8.5448,
+      "step": 162540
+    },
+    {
+      "epoch": 0.8117555993907464,
+      "grad_norm": 0.08922384679317474,
+      "learning_rate": 5.661468372174523e-06,
+      "loss": 8.5561,
+      "step": 162550
+    },
+    {
+      "epoch": 0.8118055382156858,
+      "grad_norm": 0.09595987200737,
+      "learning_rate": 5.659966457232972e-06,
+      "loss": 8.5536,
+      "step": 162560
+    },
+    {
+      "epoch": 0.8118554770406252,
+      "grad_norm": 0.08500176668167114,
+      "learning_rate": 5.658464542291421e-06,
+      "loss": 8.552,
+      "step": 162570
+    },
+    {
+      "epoch": 0.8119054158655646,
+      "grad_norm": 0.0854329764842987,
+      "learning_rate": 5.6569626273498715e-06,
+      "loss": 8.5606,
+      "step": 162580
+    },
+    {
+      "epoch": 0.8119553546905042,
+      "grad_norm": 0.08596555888652802,
+      "learning_rate": 5.655460712408321e-06,
+      "loss": 8.5476,
+      "step": 162590
+    },
+    {
+      "epoch": 0.8120052935154436,
+      "grad_norm": 0.09005005657672882,
+      "learning_rate": 5.65395879746677e-06,
+      "loss": 8.5487,
+      "step": 162600
+    },
+    {
+      "epoch": 0.812055232340383,
+      "grad_norm": 0.08684473484754562,
+      "learning_rate": 5.6524568825252195e-06,
+      "loss": 8.5494,
+      "step": 162610
+    },
+    {
+      "epoch": 0.8121051711653224,
+      "grad_norm": 0.08870302140712738,
+      "learning_rate": 5.65095496758367e-06,
+      "loss": 8.5443,
+      "step": 162620
+    },
+    {
+      "epoch": 0.812155109990262,
+      "grad_norm": 0.0927698165178299,
+      "learning_rate": 5.649453052642119e-06,
+      "loss": 8.5475,
+      "step": 162630
+    },
+    {
+      "epoch": 0.8122050488152014,
+      "grad_norm": 0.09097401797771454,
+      "learning_rate": 5.647951137700568e-06,
+      "loss": 8.5476,
+      "step": 162640
+    },
+    {
+      "epoch": 0.8122549876401408,
+      "grad_norm": 0.0920330137014389,
+      "learning_rate": 5.646449222759018e-06,
+      "loss": 8.5655,
+      "step": 162650
+    },
+    {
+      "epoch": 0.8123049264650802,
+      "grad_norm": 0.09149347245693207,
+      "learning_rate": 5.644947307817467e-06,
+      "loss": 8.543,
+      "step": 162660
+    },
+    {
+      "epoch": 0.8123548652900198,
+      "grad_norm": 0.09597709774971008,
+      "learning_rate": 5.643445392875917e-06,
+      "loss": 8.5521,
+      "step": 162670
+    },
+    {
+      "epoch": 0.8124048041149592,
+      "grad_norm": 0.08816757798194885,
+      "learning_rate": 5.6419434779343665e-06,
+      "loss": 8.5451,
+      "step": 162680
+    },
+    {
+      "epoch": 0.8124547429398986,
+      "grad_norm": 0.0923578068614006,
+      "learning_rate": 5.640441562992816e-06,
+      "loss": 8.5497,
+      "step": 162690
+    },
+    {
+      "epoch": 0.812504681764838,
+      "grad_norm": 0.08633648604154587,
+      "learning_rate": 5.638939648051266e-06,
+      "loss": 8.5605,
+      "step": 162700
+    },
+    {
+      "epoch": 0.8125546205897776,
+      "grad_norm": 0.09606485068798065,
+      "learning_rate": 5.6374377331097145e-06,
+      "loss": 8.5607,
+      "step": 162710
+    },
+    {
+      "epoch": 0.812604559414717,
+      "grad_norm": 0.09084507077932358,
+      "learning_rate": 5.635935818168165e-06,
+      "loss": 8.559,
+      "step": 162720
+    },
+    {
+      "epoch": 0.8126544982396564,
+      "grad_norm": 0.0889735296368599,
+      "learning_rate": 5.634433903226614e-06,
+      "loss": 8.5486,
+      "step": 162730
+    },
+    {
+      "epoch": 0.8127044370645958,
+      "grad_norm": 0.09297562390565872,
+      "learning_rate": 5.632931988285063e-06,
+      "loss": 8.5603,
+      "step": 162740
+    },
+    {
+      "epoch": 0.8127543758895354,
+      "grad_norm": 0.08924896270036697,
+      "learning_rate": 5.6314300733435135e-06,
+      "loss": 8.5499,
+      "step": 162750
+    },
+    {
+      "epoch": 0.8128043147144748,
+      "grad_norm": 0.08938944339752197,
+      "learning_rate": 5.629928158401962e-06,
+      "loss": 8.5714,
+      "step": 162760
+    },
+    {
+      "epoch": 0.8128542535394142,
+      "grad_norm": 0.09662458300590515,
+      "learning_rate": 5.628426243460412e-06,
+      "loss": 8.5555,
+      "step": 162770
+    },
+    {
+      "epoch": 0.8129041923643536,
+      "grad_norm": 0.08405207097530365,
+      "learning_rate": 5.626924328518862e-06,
+      "loss": 8.5385,
+      "step": 162780
+    },
+    {
+      "epoch": 0.8129541311892932,
+      "grad_norm": 0.09311365336179733,
+      "learning_rate": 5.625422413577311e-06,
+      "loss": 8.5482,
+      "step": 162790
+    },
+    {
+      "epoch": 0.8130040700142326,
+      "grad_norm": 0.09118988364934921,
+      "learning_rate": 5.623920498635761e-06,
+      "loss": 8.5717,
+      "step": 162800
+    },
+    {
+      "epoch": 0.813054008839172,
+      "grad_norm": 0.09193669259548187,
+      "learning_rate": 5.62241858369421e-06,
+      "loss": 8.5485,
+      "step": 162810
+    },
+    {
+      "epoch": 0.8131039476641114,
+      "grad_norm": 0.09104682505130768,
+      "learning_rate": 5.62091666875266e-06,
+      "loss": 8.5567,
+      "step": 162820
+    },
+    {
+      "epoch": 0.813153886489051,
+      "grad_norm": 0.0893746018409729,
+      "learning_rate": 5.61941475381111e-06,
+      "loss": 8.5666,
+      "step": 162830
+    },
+    {
+      "epoch": 0.8132038253139904,
+      "grad_norm": 0.09030063450336456,
+      "learning_rate": 5.617912838869558e-06,
+      "loss": 8.5506,
+      "step": 162840
+    },
+    {
+      "epoch": 0.8132537641389298,
+      "grad_norm": 0.0945194736123085,
+      "learning_rate": 5.6164109239280085e-06,
+      "loss": 8.5281,
+      "step": 162850
+    },
+    {
+      "epoch": 0.8133037029638692,
+      "grad_norm": 0.09485594183206558,
+      "learning_rate": 5.614909008986459e-06,
+      "loss": 8.5553,
+      "step": 162860
+    },
+    {
+      "epoch": 0.8133536417888088,
+      "grad_norm": 0.08975322544574738,
+      "learning_rate": 5.613407094044907e-06,
+      "loss": 8.5503,
+      "step": 162870
+    },
+    {
+      "epoch": 0.8134035806137482,
+      "grad_norm": 0.09506627917289734,
+      "learning_rate": 5.611905179103357e-06,
+      "loss": 8.5368,
+      "step": 162880
+    },
+    {
+      "epoch": 0.8134535194386876,
+      "grad_norm": 0.09811557084321976,
+      "learning_rate": 5.610403264161806e-06,
+      "loss": 8.5708,
+      "step": 162890
+    },
+    {
+      "epoch": 0.813503458263627,
+      "grad_norm": 0.08716001361608505,
+      "learning_rate": 5.608901349220256e-06,
+      "loss": 8.5295,
+      "step": 162900
+    },
+    {
+      "epoch": 0.8135533970885666,
+      "grad_norm": 0.08750799298286438,
+      "learning_rate": 5.607399434278706e-06,
+      "loss": 8.5557,
+      "step": 162910
+    },
+    {
+      "epoch": 0.813603335913506,
+      "grad_norm": 0.09434546530246735,
+      "learning_rate": 5.605897519337155e-06,
+      "loss": 8.5626,
+      "step": 162920
+    },
+    {
+      "epoch": 0.8136532747384454,
+      "grad_norm": 0.08905541896820068,
+      "learning_rate": 5.604395604395605e-06,
+      "loss": 8.543,
+      "step": 162930
+    },
+    {
+      "epoch": 0.8137032135633848,
+      "grad_norm": 0.09109733253717422,
+      "learning_rate": 5.602893689454054e-06,
+      "loss": 8.5606,
+      "step": 162940
+    },
+    {
+      "epoch": 0.8137531523883244,
+      "grad_norm": 0.09967444837093353,
+      "learning_rate": 5.6013917745125035e-06,
+      "loss": 8.534,
+      "step": 162950
+    },
+    {
+      "epoch": 0.8138030912132638,
+      "grad_norm": 0.08702152967453003,
+      "learning_rate": 5.599889859570954e-06,
+      "loss": 8.5679,
+      "step": 162960
+    },
+    {
+      "epoch": 0.8138530300382032,
+      "grad_norm": 0.09200234711170197,
+      "learning_rate": 5.598387944629402e-06,
+      "loss": 8.5477,
+      "step": 162970
+    },
+    {
+      "epoch": 0.8139029688631426,
+      "grad_norm": 0.08344555646181107,
+      "learning_rate": 5.596886029687852e-06,
+      "loss": 8.5648,
+      "step": 162980
+    },
+    {
+      "epoch": 0.8139529076880822,
+      "grad_norm": 0.09271971136331558,
+      "learning_rate": 5.595384114746302e-06,
+      "loss": 8.5407,
+      "step": 162990
+    },
+    {
+      "epoch": 0.8140028465130216,
+      "grad_norm": 0.09208449721336365,
+      "learning_rate": 5.593882199804751e-06,
+      "loss": 8.5329,
+      "step": 163000
+    },
+    {
+      "epoch": 0.814052785337961,
+      "grad_norm": 0.09522075206041336,
+      "learning_rate": 5.592380284863201e-06,
+      "loss": 8.5586,
+      "step": 163010
+    },
+    {
+      "epoch": 0.8141027241629004,
+      "grad_norm": 0.09683191031217575,
+      "learning_rate": 5.5908783699216505e-06,
+      "loss": 8.5441,
+      "step": 163020
+    },
+    {
+      "epoch": 0.81415266298784,
+      "grad_norm": 0.09313095360994339,
+      "learning_rate": 5.5893764549801e-06,
+      "loss": 8.5572,
+      "step": 163030
+    },
+    {
+      "epoch": 0.8142026018127794,
+      "grad_norm": 0.09198789298534393,
+      "learning_rate": 5.587874540038549e-06,
+      "loss": 8.5576,
+      "step": 163040
+    },
+    {
+      "epoch": 0.8142525406377188,
+      "grad_norm": 0.0924067571759224,
+      "learning_rate": 5.5863726250969985e-06,
+      "loss": 8.5483,
+      "step": 163050
+    },
+    {
+      "epoch": 0.8143024794626582,
+      "grad_norm": 0.09334490448236465,
+      "learning_rate": 5.584870710155449e-06,
+      "loss": 8.5449,
+      "step": 163060
+    },
+    {
+      "epoch": 0.8143524182875977,
+      "grad_norm": 0.08748748898506165,
+      "learning_rate": 5.583368795213898e-06,
+      "loss": 8.5381,
+      "step": 163070
+    },
+    {
+      "epoch": 0.8144023571125372,
+      "grad_norm": 0.08966860920190811,
+      "learning_rate": 5.581866880272347e-06,
+      "loss": 8.5411,
+      "step": 163080
+    },
+    {
+      "epoch": 0.8144522959374766,
+      "grad_norm": 0.09077271819114685,
+      "learning_rate": 5.580364965330797e-06,
+      "loss": 8.545,
+      "step": 163090
+    },
+    {
+      "epoch": 0.814502234762416,
+      "grad_norm": 0.0929906815290451,
+      "learning_rate": 5.578863050389247e-06,
+      "loss": 8.5392,
+      "step": 163100
+    },
+    {
+      "epoch": 0.8145521735873555,
+      "grad_norm": 0.09005864709615707,
+      "learning_rate": 5.577361135447696e-06,
+      "loss": 8.5423,
+      "step": 163110
+    },
+    {
+      "epoch": 0.814602112412295,
+      "grad_norm": 0.09385271370410919,
+      "learning_rate": 5.5758592205061455e-06,
+      "loss": 8.5367,
+      "step": 163120
+    },
+    {
+      "epoch": 0.8146520512372344,
+      "grad_norm": 0.09610309451818466,
+      "learning_rate": 5.574357305564595e-06,
+      "loss": 8.5475,
+      "step": 163130
+    },
+    {
+      "epoch": 0.8147019900621738,
+      "grad_norm": 0.09091010689735413,
+      "learning_rate": 5.572855390623044e-06,
+      "loss": 8.5471,
+      "step": 163140
+    },
+    {
+      "epoch": 0.8147519288871133,
+      "grad_norm": 0.09262845665216446,
+      "learning_rate": 5.571353475681494e-06,
+      "loss": 8.5337,
+      "step": 163150
+    },
+    {
+      "epoch": 0.8148018677120528,
+      "grad_norm": 0.08759453892707825,
+      "learning_rate": 5.569851560739944e-06,
+      "loss": 8.5621,
+      "step": 163160
+    },
+    {
+      "epoch": 0.8148518065369922,
+      "grad_norm": 0.08776848018169403,
+      "learning_rate": 5.568349645798393e-06,
+      "loss": 8.5529,
+      "step": 163170
+    },
+    {
+      "epoch": 0.8149017453619316,
+      "grad_norm": 0.08764344453811646,
+      "learning_rate": 5.566847730856843e-06,
+      "loss": 8.5606,
+      "step": 163180
+    },
+    {
+      "epoch": 0.814951684186871,
+      "grad_norm": 0.0934050902724266,
+      "learning_rate": 5.565345815915292e-06,
+      "loss": 8.5542,
+      "step": 163190
+    },
+    {
+      "epoch": 0.8150016230118106,
+      "grad_norm": 0.08384846895933151,
+      "learning_rate": 5.563843900973742e-06,
+      "loss": 8.5444,
+      "step": 163200
+    },
+    {
+      "epoch": 0.81505156183675,
+      "grad_norm": 0.0902908518910408,
+      "learning_rate": 5.562341986032191e-06,
+      "loss": 8.5412,
+      "step": 163210
+    },
+    {
+      "epoch": 0.8151015006616894,
+      "grad_norm": 0.08977564424276352,
+      "learning_rate": 5.5608400710906405e-06,
+      "loss": 8.5599,
+      "step": 163220
+    },
+    {
+      "epoch": 0.8151514394866288,
+      "grad_norm": 0.09766307473182678,
+      "learning_rate": 5.559338156149091e-06,
+      "loss": 8.5354,
+      "step": 163230
+    },
+    {
+      "epoch": 0.8152013783115684,
+      "grad_norm": 0.08712057769298553,
+      "learning_rate": 5.557836241207539e-06,
+      "loss": 8.5372,
+      "step": 163240
+    },
+    {
+      "epoch": 0.8152513171365078,
+      "grad_norm": 0.0951530709862709,
+      "learning_rate": 5.556334326265989e-06,
+      "loss": 8.5354,
+      "step": 163250
+    },
+    {
+      "epoch": 0.8153012559614472,
+      "grad_norm": 0.09080353379249573,
+      "learning_rate": 5.5548324113244396e-06,
+      "loss": 8.5443,
+      "step": 163260
+    },
+    {
+      "epoch": 0.8153511947863866,
+      "grad_norm": 0.09223667532205582,
+      "learning_rate": 5.553330496382888e-06,
+      "loss": 8.5453,
+      "step": 163270
+    },
+    {
+      "epoch": 0.8154011336113262,
+      "grad_norm": 0.08771520853042603,
+      "learning_rate": 5.551828581441338e-06,
+      "loss": 8.5374,
+      "step": 163280
+    },
+    {
+      "epoch": 0.8154510724362656,
+      "grad_norm": 0.08749589323997498,
+      "learning_rate": 5.550326666499787e-06,
+      "loss": 8.5422,
+      "step": 163290
+    },
+    {
+      "epoch": 0.815501011261205,
+      "grad_norm": 0.08917036652565002,
+      "learning_rate": 5.548824751558237e-06,
+      "loss": 8.5481,
+      "step": 163300
+    },
+    {
+      "epoch": 0.8155509500861444,
+      "grad_norm": 0.0868101641535759,
+      "learning_rate": 5.547322836616687e-06,
+      "loss": 8.5619,
+      "step": 163310
+    },
+    {
+      "epoch": 0.815600888911084,
+      "grad_norm": 0.091803178191185,
+      "learning_rate": 5.5458209216751355e-06,
+      "loss": 8.5356,
+      "step": 163320
+    },
+    {
+      "epoch": 0.8156508277360234,
+      "grad_norm": 0.09425832331180573,
+      "learning_rate": 5.544319006733586e-06,
+      "loss": 8.5293,
+      "step": 163330
+    },
+    {
+      "epoch": 0.8157007665609628,
+      "grad_norm": 0.1101720854640007,
+      "learning_rate": 5.542817091792035e-06,
+      "loss": 8.544,
+      "step": 163340
+    },
+    {
+      "epoch": 0.8157507053859022,
+      "grad_norm": 0.0971333235502243,
+      "learning_rate": 5.541315176850484e-06,
+      "loss": 8.5534,
+      "step": 163350
+    },
+    {
+      "epoch": 0.8158006442108418,
+      "grad_norm": 0.09201129525899887,
+      "learning_rate": 5.5398132619089346e-06,
+      "loss": 8.5444,
+      "step": 163360
+    },
+    {
+      "epoch": 0.8158505830357812,
+      "grad_norm": 0.08867859095335007,
+      "learning_rate": 5.538311346967383e-06,
+      "loss": 8.5282,
+      "step": 163370
+    },
+    {
+      "epoch": 0.8159005218607206,
+      "grad_norm": 0.09262479841709137,
+      "learning_rate": 5.536809432025833e-06,
+      "loss": 8.5319,
+      "step": 163380
+    },
+    {
+      "epoch": 0.81595046068566,
+      "grad_norm": 0.08705681562423706,
+      "learning_rate": 5.5353075170842826e-06,
+      "loss": 8.5469,
+      "step": 163390
+    },
+    {
+      "epoch": 0.8160003995105996,
+      "grad_norm": 0.09213688969612122,
+      "learning_rate": 5.533805602142732e-06,
+      "loss": 8.5515,
+      "step": 163400
+    },
+    {
+      "epoch": 0.816050338335539,
+      "grad_norm": 0.0938788652420044,
+      "learning_rate": 5.532303687201182e-06,
+      "loss": 8.538,
+      "step": 163410
+    },
+    {
+      "epoch": 0.8161002771604784,
+      "grad_norm": 0.09791158139705658,
+      "learning_rate": 5.530801772259631e-06,
+      "loss": 8.5372,
+      "step": 163420
+    },
+    {
+      "epoch": 0.8161502159854178,
+      "grad_norm": 0.09159670770168304,
+      "learning_rate": 5.529299857318081e-06,
+      "loss": 8.551,
+      "step": 163430
+    },
+    {
+      "epoch": 0.8162001548103573,
+      "grad_norm": 0.08808984607458115,
+      "learning_rate": 5.52779794237653e-06,
+      "loss": 8.5498,
+      "step": 163440
+    },
+    {
+      "epoch": 0.8162500936352968,
+      "grad_norm": 0.09138135612010956,
+      "learning_rate": 5.526296027434979e-06,
+      "loss": 8.5526,
+      "step": 163450
+    },
+    {
+      "epoch": 0.8163000324602362,
+      "grad_norm": 0.08727425336837769,
+      "learning_rate": 5.5247941124934296e-06,
+      "loss": 8.5605,
+      "step": 163460
+    },
+    {
+      "epoch": 0.8163499712851756,
+      "grad_norm": 0.0928439050912857,
+      "learning_rate": 5.523292197551879e-06,
+      "loss": 8.5574,
+      "step": 163470
+    },
+    {
+      "epoch": 0.8163999101101151,
+      "grad_norm": 0.09845706075429916,
+      "learning_rate": 5.521790282610328e-06,
+      "loss": 8.5555,
+      "step": 163480
+    },
+    {
+      "epoch": 0.8164498489350546,
+      "grad_norm": 0.09574224054813385,
+      "learning_rate": 5.5202883676687776e-06,
+      "loss": 8.5523,
+      "step": 163490
+    },
+    {
+      "epoch": 0.816499787759994,
+      "grad_norm": 0.09411446005105972,
+      "learning_rate": 5.518786452727228e-06,
+      "loss": 8.5482,
+      "step": 163500
+    },
+    {
+      "epoch": 0.8165497265849334,
+      "grad_norm": 0.09284447878599167,
+      "learning_rate": 5.517284537785677e-06,
+      "loss": 8.5463,
+      "step": 163510
+    },
+    {
+      "epoch": 0.8165996654098729,
+      "grad_norm": 0.08727595210075378,
+      "learning_rate": 5.515782622844126e-06,
+      "loss": 8.5425,
+      "step": 163520
+    },
+    {
+      "epoch": 0.8166496042348124,
+      "grad_norm": 0.09593432396650314,
+      "learning_rate": 5.514280707902576e-06,
+      "loss": 8.5428,
+      "step": 163530
+    },
+    {
+      "epoch": 0.8166995430597518,
+      "grad_norm": 0.09068622440099716,
+      "learning_rate": 5.512778792961025e-06,
+      "loss": 8.5558,
+      "step": 163540
+    },
+    {
+      "epoch": 0.8167494818846912,
+      "grad_norm": 0.09026969224214554,
+      "learning_rate": 5.511276878019475e-06,
+      "loss": 8.5397,
+      "step": 163550
+    },
+    {
+      "epoch": 0.8167994207096307,
+      "grad_norm": 0.08629763126373291,
+      "learning_rate": 5.509774963077925e-06,
+      "loss": 8.5547,
+      "step": 163560
+    },
+    {
+      "epoch": 0.8168493595345702,
+      "grad_norm": 0.09150740504264832,
+      "learning_rate": 5.508273048136374e-06,
+      "loss": 8.5438,
+      "step": 163570
+    },
+    {
+      "epoch": 0.8168992983595096,
+      "grad_norm": 0.0950947180390358,
+      "learning_rate": 5.506771133194824e-06,
+      "loss": 8.556,
+      "step": 163580
+    },
+    {
+      "epoch": 0.816949237184449,
+      "grad_norm": 0.0921635702252388,
+      "learning_rate": 5.5052692182532726e-06,
+      "loss": 8.5484,
+      "step": 163590
+    },
+    {
+      "epoch": 0.8169991760093885,
+      "grad_norm": 0.08401766419410706,
+      "learning_rate": 5.503767303311723e-06,
+      "loss": 8.5567,
+      "step": 163600
+    },
+    {
+      "epoch": 0.817049114834328,
+      "grad_norm": 0.09006239473819733,
+      "learning_rate": 5.502265388370172e-06,
+      "loss": 8.55,
+      "step": 163610
+    },
+    {
+      "epoch": 0.8170990536592674,
+      "grad_norm": 0.08677154034376144,
+      "learning_rate": 5.500763473428621e-06,
+      "loss": 8.5348,
+      "step": 163620
+    },
+    {
+      "epoch": 0.8171489924842068,
+      "grad_norm": 0.08495960384607315,
+      "learning_rate": 5.499261558487072e-06,
+      "loss": 8.5472,
+      "step": 163630
+    },
+    {
+      "epoch": 0.8171989313091463,
+      "grad_norm": 0.08838516473770142,
+      "learning_rate": 5.49775964354552e-06,
+      "loss": 8.5546,
+      "step": 163640
+    },
+    {
+      "epoch": 0.8172488701340858,
+      "grad_norm": 0.09089840948581696,
+      "learning_rate": 5.49625772860397e-06,
+      "loss": 8.5567,
+      "step": 163650
+    },
+    {
+      "epoch": 0.8172988089590252,
+      "grad_norm": 0.08638277649879456,
+      "learning_rate": 5.4947558136624204e-06,
+      "loss": 8.5409,
+      "step": 163660
+    },
+    {
+      "epoch": 0.8173487477839646,
+      "grad_norm": 0.09370175004005432,
+      "learning_rate": 5.493253898720869e-06,
+      "loss": 8.5309,
+      "step": 163670
+    },
+    {
+      "epoch": 0.8173986866089041,
+      "grad_norm": 0.08680180460214615,
+      "learning_rate": 5.491751983779319e-06,
+      "loss": 8.5588,
+      "step": 163680
+    },
+    {
+      "epoch": 0.8174486254338436,
+      "grad_norm": 0.09824364632368088,
+      "learning_rate": 5.490250068837768e-06,
+      "loss": 8.5466,
+      "step": 163690
+    },
+    {
+      "epoch": 0.817498564258783,
+      "grad_norm": 0.09609740972518921,
+      "learning_rate": 5.488748153896218e-06,
+      "loss": 8.5486,
+      "step": 163700
+    },
+    {
+      "epoch": 0.8175485030837224,
+      "grad_norm": 0.09737160056829453,
+      "learning_rate": 5.487246238954668e-06,
+      "loss": 8.5543,
+      "step": 163710
+    },
+    {
+      "epoch": 0.8175984419086619,
+      "grad_norm": 0.09023190289735794,
+      "learning_rate": 5.485744324013116e-06,
+      "loss": 8.5539,
+      "step": 163720
+    },
+    {
+      "epoch": 0.8176483807336014,
+      "grad_norm": 0.08571267127990723,
+      "learning_rate": 5.484242409071567e-06,
+      "loss": 8.5504,
+      "step": 163730
+    },
+    {
+      "epoch": 0.8176983195585408,
+      "grad_norm": 0.08893782645463943,
+      "learning_rate": 5.482740494130016e-06,
+      "loss": 8.5526,
+      "step": 163740
+    },
+    {
+      "epoch": 0.8177482583834802,
+      "grad_norm": 0.08946535736322403,
+      "learning_rate": 5.481238579188465e-06,
+      "loss": 8.5538,
+      "step": 163750
+    },
+    {
+      "epoch": 0.8177981972084197,
+      "grad_norm": 0.09130077809095383,
+      "learning_rate": 5.4797366642469154e-06,
+      "loss": 8.5434,
+      "step": 163760
+    },
+    {
+      "epoch": 0.8178481360333592,
+      "grad_norm": 0.08708508312702179,
+      "learning_rate": 5.478234749305364e-06,
+      "loss": 8.5406,
+      "step": 163770
+    },
+    {
+      "epoch": 0.8178980748582986,
+      "grad_norm": 0.09231363236904144,
+      "learning_rate": 5.476732834363814e-06,
+      "loss": 8.5286,
+      "step": 163780
+    },
+    {
+      "epoch": 0.817948013683238,
+      "grad_norm": 0.09548377245664597,
+      "learning_rate": 5.4752309194222634e-06,
+      "loss": 8.5447,
+      "step": 163790
+    },
+    {
+      "epoch": 0.8179979525081775,
+      "grad_norm": 0.08776544034481049,
+      "learning_rate": 5.473729004480713e-06,
+      "loss": 8.5347,
+      "step": 163800
+    },
+    {
+      "epoch": 0.818047891333117,
+      "grad_norm": 0.09309916198253632,
+      "learning_rate": 5.472227089539163e-06,
+      "loss": 8.5546,
+      "step": 163810
+    },
+    {
+      "epoch": 0.8180978301580564,
+      "grad_norm": 0.09029585868120193,
+      "learning_rate": 5.4707251745976114e-06,
+      "loss": 8.531,
+      "step": 163820
+    },
+    {
+      "epoch": 0.8181477689829958,
+      "grad_norm": 0.08917348831892014,
+      "learning_rate": 5.469223259656062e-06,
+      "loss": 8.5553,
+      "step": 163830
+    },
+    {
+      "epoch": 0.8181977078079353,
+      "grad_norm": 0.09531484544277191,
+      "learning_rate": 5.467721344714511e-06,
+      "loss": 8.5415,
+      "step": 163840
+    },
+    {
+      "epoch": 0.8182476466328747,
+      "grad_norm": 0.08932767063379288,
+      "learning_rate": 5.46621942977296e-06,
+      "loss": 8.5473,
+      "step": 163850
+    },
+    {
+      "epoch": 0.8182975854578142,
+      "grad_norm": 0.08596362173557281,
+      "learning_rate": 5.4647175148314104e-06,
+      "loss": 8.552,
+      "step": 163860
+    },
+    {
+      "epoch": 0.8183475242827536,
+      "grad_norm": 0.09185688197612762,
+      "learning_rate": 5.46321559988986e-06,
+      "loss": 8.5515,
+      "step": 163870
+    },
+    {
+      "epoch": 0.8183974631076931,
+      "grad_norm": 0.09344456344842911,
+      "learning_rate": 5.461713684948309e-06,
+      "loss": 8.5385,
+      "step": 163880
+    },
+    {
+      "epoch": 0.8184474019326325,
+      "grad_norm": 0.08724787086248398,
+      "learning_rate": 5.4602117700067584e-06,
+      "loss": 8.5502,
+      "step": 163890
+    },
+    {
+      "epoch": 0.818497340757572,
+      "grad_norm": 0.09026943147182465,
+      "learning_rate": 5.458709855065208e-06,
+      "loss": 8.5269,
+      "step": 163900
+    },
+    {
+      "epoch": 0.8185472795825114,
+      "grad_norm": 0.0960424542427063,
+      "learning_rate": 5.457207940123658e-06,
+      "loss": 8.5535,
+      "step": 163910
+    },
+    {
+      "epoch": 0.8185972184074509,
+      "grad_norm": 0.09598157554864883,
+      "learning_rate": 5.455706025182107e-06,
+      "loss": 8.5696,
+      "step": 163920
+    },
+    {
+      "epoch": 0.8186471572323903,
+      "grad_norm": 0.10426607728004456,
+      "learning_rate": 5.454204110240557e-06,
+      "loss": 8.5406,
+      "step": 163930
+    },
+    {
+      "epoch": 0.8186970960573298,
+      "grad_norm": 0.09276427328586578,
+      "learning_rate": 5.452702195299007e-06,
+      "loss": 8.5369,
+      "step": 163940
+    },
+    {
+      "epoch": 0.8187470348822692,
+      "grad_norm": 0.09880458563566208,
+      "learning_rate": 5.451200280357456e-06,
+      "loss": 8.5536,
+      "step": 163950
+    },
+    {
+      "epoch": 0.8187969737072087,
+      "grad_norm": 0.09303481876850128,
+      "learning_rate": 5.4496983654159055e-06,
+      "loss": 8.5303,
+      "step": 163960
+    },
+    {
+      "epoch": 0.8188469125321481,
+      "grad_norm": 0.09363751113414764,
+      "learning_rate": 5.448196450474355e-06,
+      "loss": 8.5507,
+      "step": 163970
+    },
+    {
+      "epoch": 0.8188968513570876,
+      "grad_norm": 0.09529575705528259,
+      "learning_rate": 5.446694535532804e-06,
+      "loss": 8.5431,
+      "step": 163980
+    },
+    {
+      "epoch": 0.818946790182027,
+      "grad_norm": 0.09039454907178879,
+      "learning_rate": 5.445192620591254e-06,
+      "loss": 8.5674,
+      "step": 163990
+    },
+    {
+      "epoch": 0.8189967290069665,
+      "grad_norm": 0.08885947614908218,
+      "learning_rate": 5.443690705649704e-06,
+      "loss": 8.5497,
+      "step": 164000
+    },
+    {
+      "epoch": 0.8190466678319059,
+      "grad_norm": 0.09177469462156296,
+      "learning_rate": 5.442188790708153e-06,
+      "loss": 8.5541,
+      "step": 164010
+    },
+    {
+      "epoch": 0.8190966066568454,
+      "grad_norm": 0.08675147593021393,
+      "learning_rate": 5.440686875766602e-06,
+      "loss": 8.5576,
+      "step": 164020
+    },
+    {
+      "epoch": 0.8191465454817848,
+      "grad_norm": 0.09036063402891159,
+      "learning_rate": 5.4391849608250525e-06,
+      "loss": 8.5723,
+      "step": 164030
+    },
+    {
+      "epoch": 0.8191964843067243,
+      "grad_norm": 0.08989081531763077,
+      "learning_rate": 5.437683045883502e-06,
+      "loss": 8.5417,
+      "step": 164040
+    },
+    {
+      "epoch": 0.8192464231316637,
+      "grad_norm": 0.0942939966917038,
+      "learning_rate": 5.436181130941951e-06,
+      "loss": 8.5311,
+      "step": 164050
+    },
+    {
+      "epoch": 0.8192963619566032,
+      "grad_norm": 0.0847172811627388,
+      "learning_rate": 5.4346792160004005e-06,
+      "loss": 8.5449,
+      "step": 164060
+    },
+    {
+      "epoch": 0.8193463007815426,
+      "grad_norm": 0.09628713876008987,
+      "learning_rate": 5.43317730105885e-06,
+      "loss": 8.5535,
+      "step": 164070
+    },
+    {
+      "epoch": 0.8193962396064821,
+      "grad_norm": 0.08984513580799103,
+      "learning_rate": 5.4316753861173e-06,
+      "loss": 8.5574,
+      "step": 164080
+    },
+    {
+      "epoch": 0.8194461784314215,
+      "grad_norm": 0.0882403776049614,
+      "learning_rate": 5.430173471175749e-06,
+      "loss": 8.5421,
+      "step": 164090
+    },
+    {
+      "epoch": 0.819496117256361,
+      "grad_norm": 0.09667932987213135,
+      "learning_rate": 5.428671556234199e-06,
+      "loss": 8.5527,
+      "step": 164100
+    },
+    {
+      "epoch": 0.8195460560813004,
+      "grad_norm": 0.08921948820352554,
+      "learning_rate": 5.427169641292649e-06,
+      "loss": 8.5558,
+      "step": 164110
+    },
+    {
+      "epoch": 0.8195959949062399,
+      "grad_norm": 0.08947332203388214,
+      "learning_rate": 5.425667726351097e-06,
+      "loss": 8.5428,
+      "step": 164120
+    },
+    {
+      "epoch": 0.8196459337311793,
+      "grad_norm": 0.09034401178359985,
+      "learning_rate": 5.4241658114095475e-06,
+      "loss": 8.5296,
+      "step": 164130
+    },
+    {
+      "epoch": 0.8196958725561188,
+      "grad_norm": 0.0902322456240654,
+      "learning_rate": 5.422663896467997e-06,
+      "loss": 8.5575,
+      "step": 164140
+    },
+    {
+      "epoch": 0.8197458113810582,
+      "grad_norm": 0.09341645240783691,
+      "learning_rate": 5.421161981526446e-06,
+      "loss": 8.5492,
+      "step": 164150
+    },
+    {
+      "epoch": 0.8197957502059976,
+      "grad_norm": 0.09417828172445297,
+      "learning_rate": 5.419660066584896e-06,
+      "loss": 8.5322,
+      "step": 164160
+    },
+    {
+      "epoch": 0.8198456890309371,
+      "grad_norm": 0.09243685007095337,
+      "learning_rate": 5.418158151643345e-06,
+      "loss": 8.5467,
+      "step": 164170
+    },
+    {
+      "epoch": 0.8198956278558766,
+      "grad_norm": 0.09273440390825272,
+      "learning_rate": 5.416656236701795e-06,
+      "loss": 8.5316,
+      "step": 164180
+    },
+    {
+      "epoch": 0.819945566680816,
+      "grad_norm": 0.09528848528862,
+      "learning_rate": 5.415154321760245e-06,
+      "loss": 8.5459,
+      "step": 164190
+    },
+    {
+      "epoch": 0.8199955055057554,
+      "grad_norm": 0.08874344080686569,
+      "learning_rate": 5.413652406818694e-06,
+      "loss": 8.5414,
+      "step": 164200
+    },
+    {
+      "epoch": 0.8200454443306949,
+      "grad_norm": 0.09056615829467773,
+      "learning_rate": 5.412150491877144e-06,
+      "loss": 8.5484,
+      "step": 164210
+    },
+    {
+      "epoch": 0.8200953831556343,
+      "grad_norm": 0.09031280875205994,
+      "learning_rate": 5.410648576935592e-06,
+      "loss": 8.5455,
+      "step": 164220
+    },
+    {
+      "epoch": 0.8201453219805738,
+      "grad_norm": 0.09306702762842178,
+      "learning_rate": 5.4091466619940425e-06,
+      "loss": 8.547,
+      "step": 164230
+    },
+    {
+      "epoch": 0.8201952608055132,
+      "grad_norm": 0.09164511412382126,
+      "learning_rate": 5.407644747052493e-06,
+      "loss": 8.556,
+      "step": 164240
+    },
+    {
+      "epoch": 0.8202451996304527,
+      "grad_norm": 0.10014204680919647,
+      "learning_rate": 5.406142832110941e-06,
+      "loss": 8.546,
+      "step": 164250
+    },
+    {
+      "epoch": 0.8202951384553921,
+      "grad_norm": 0.09262999892234802,
+      "learning_rate": 5.404640917169391e-06,
+      "loss": 8.5525,
+      "step": 164260
+    },
+    {
+      "epoch": 0.8203450772803316,
+      "grad_norm": 0.09438300132751465,
+      "learning_rate": 5.403139002227841e-06,
+      "loss": 8.5283,
+      "step": 164270
+    },
+    {
+      "epoch": 0.820395016105271,
+      "grad_norm": 0.09390977770090103,
+      "learning_rate": 5.40163708728629e-06,
+      "loss": 8.5388,
+      "step": 164280
+    },
+    {
+      "epoch": 0.8204449549302105,
+      "grad_norm": 0.09154833853244781,
+      "learning_rate": 5.40013517234474e-06,
+      "loss": 8.5558,
+      "step": 164290
+    },
+    {
+      "epoch": 0.82049489375515,
+      "grad_norm": 0.09532206505537033,
+      "learning_rate": 5.398633257403189e-06,
+      "loss": 8.5325,
+      "step": 164300
+    },
+    {
+      "epoch": 0.8205448325800894,
+      "grad_norm": 0.08532291650772095,
+      "learning_rate": 5.397131342461639e-06,
+      "loss": 8.5563,
+      "step": 164310
+    },
+    {
+      "epoch": 0.8205947714050288,
+      "grad_norm": 0.10053151100873947,
+      "learning_rate": 5.395629427520088e-06,
+      "loss": 8.5493,
+      "step": 164320
+    },
+    {
+      "epoch": 0.8206447102299683,
+      "grad_norm": 0.08957508206367493,
+      "learning_rate": 5.3941275125785375e-06,
+      "loss": 8.5511,
+      "step": 164330
+    },
+    {
+      "epoch": 0.8206946490549077,
+      "grad_norm": 0.08716392517089844,
+      "learning_rate": 5.392625597636988e-06,
+      "loss": 8.5514,
+      "step": 164340
+    },
+    {
+      "epoch": 0.8207445878798472,
+      "grad_norm": 0.09494642913341522,
+      "learning_rate": 5.391123682695437e-06,
+      "loss": 8.5466,
+      "step": 164350
+    },
+    {
+      "epoch": 0.8207945267047866,
+      "grad_norm": 0.09567490220069885,
+      "learning_rate": 5.389621767753886e-06,
+      "loss": 8.5463,
+      "step": 164360
+    },
+    {
+      "epoch": 0.8208444655297261,
+      "grad_norm": 0.09385741502046585,
+      "learning_rate": 5.388119852812336e-06,
+      "loss": 8.5292,
+      "step": 164370
+    },
+    {
+      "epoch": 0.8208944043546655,
+      "grad_norm": 0.09710198640823364,
+      "learning_rate": 5.386617937870785e-06,
+      "loss": 8.5374,
+      "step": 164380
+    },
+    {
+      "epoch": 0.820944343179605,
+      "grad_norm": 0.09070218354463577,
+      "learning_rate": 5.385116022929235e-06,
+      "loss": 8.5362,
+      "step": 164390
+    },
+    {
+      "epoch": 0.8209942820045444,
+      "grad_norm": 0.09835637360811234,
+      "learning_rate": 5.3836141079876845e-06,
+      "loss": 8.5439,
+      "step": 164400
+    },
+    {
+      "epoch": 0.8210442208294839,
+      "grad_norm": 0.09094212204217911,
+      "learning_rate": 5.382112193046134e-06,
+      "loss": 8.5619,
+      "step": 164410
+    },
+    {
+      "epoch": 0.8210941596544233,
+      "grad_norm": 0.09400885552167892,
+      "learning_rate": 5.380610278104583e-06,
+      "loss": 8.5419,
+      "step": 164420
+    },
+    {
+      "epoch": 0.8211440984793628,
+      "grad_norm": 0.09460953623056412,
+      "learning_rate": 5.379108363163033e-06,
+      "loss": 8.5401,
+      "step": 164430
+    },
+    {
+      "epoch": 0.8211940373043022,
+      "grad_norm": 0.08946454524993896,
+      "learning_rate": 5.377606448221483e-06,
+      "loss": 8.5356,
+      "step": 164440
+    },
+    {
+      "epoch": 0.8212439761292417,
+      "grad_norm": 0.08983763307332993,
+      "learning_rate": 5.376104533279932e-06,
+      "loss": 8.5455,
+      "step": 164450
+    },
+    {
+      "epoch": 0.8212939149541811,
+      "grad_norm": 0.08810096234083176,
+      "learning_rate": 5.374602618338381e-06,
+      "loss": 8.5647,
+      "step": 164460
+    },
+    {
+      "epoch": 0.8213438537791206,
+      "grad_norm": 0.08846630901098251,
+      "learning_rate": 5.373100703396831e-06,
+      "loss": 8.5521,
+      "step": 164470
+    },
+    {
+      "epoch": 0.82139379260406,
+      "grad_norm": 0.08851570636034012,
+      "learning_rate": 5.371598788455281e-06,
+      "loss": 8.5412,
+      "step": 164480
+    },
+    {
+      "epoch": 0.8214437314289995,
+      "grad_norm": 0.09497125446796417,
+      "learning_rate": 5.37009687351373e-06,
+      "loss": 8.5321,
+      "step": 164490
+    },
+    {
+      "epoch": 0.8214936702539389,
+      "grad_norm": 0.09288336336612701,
+      "learning_rate": 5.3685949585721795e-06,
+      "loss": 8.5472,
+      "step": 164500
+    },
+    {
+      "epoch": 0.8215436090788784,
+      "grad_norm": 0.0919954776763916,
+      "learning_rate": 5.36709304363063e-06,
+      "loss": 8.5613,
+      "step": 164510
+    },
+    {
+      "epoch": 0.8215935479038178,
+      "grad_norm": 0.09269890189170837,
+      "learning_rate": 5.365591128689078e-06,
+      "loss": 8.546,
+      "step": 164520
+    },
+    {
+      "epoch": 0.8216434867287573,
+      "grad_norm": 0.08948935568332672,
+      "learning_rate": 5.364089213747528e-06,
+      "loss": 8.5627,
+      "step": 164530
+    },
+    {
+      "epoch": 0.8216934255536967,
+      "grad_norm": 0.09390202164649963,
+      "learning_rate": 5.362587298805978e-06,
+      "loss": 8.5363,
+      "step": 164540
+    },
+    {
+      "epoch": 0.8217433643786362,
+      "grad_norm": 0.09372469782829285,
+      "learning_rate": 5.361085383864427e-06,
+      "loss": 8.5518,
+      "step": 164550
+    },
+    {
+      "epoch": 0.8217933032035756,
+      "grad_norm": 0.0866970643401146,
+      "learning_rate": 5.359583468922877e-06,
+      "loss": 8.5598,
+      "step": 164560
+    },
+    {
+      "epoch": 0.8218432420285151,
+      "grad_norm": 0.08886361867189407,
+      "learning_rate": 5.358081553981326e-06,
+      "loss": 8.5504,
+      "step": 164570
+    },
+    {
+      "epoch": 0.8218931808534545,
+      "grad_norm": 0.09430024772882462,
+      "learning_rate": 5.356579639039776e-06,
+      "loss": 8.5434,
+      "step": 164580
+    },
+    {
+      "epoch": 0.821943119678394,
+      "grad_norm": 0.09142236411571503,
+      "learning_rate": 5.355077724098226e-06,
+      "loss": 8.5417,
+      "step": 164590
+    },
+    {
+      "epoch": 0.8219930585033334,
+      "grad_norm": 0.09642058610916138,
+      "learning_rate": 5.3535758091566745e-06,
+      "loss": 8.5808,
+      "step": 164600
+    },
+    {
+      "epoch": 0.8220429973282729,
+      "grad_norm": 0.08818865567445755,
+      "learning_rate": 5.352073894215125e-06,
+      "loss": 8.5353,
+      "step": 164610
+    },
+    {
+      "epoch": 0.8220929361532123,
+      "grad_norm": 0.09273871034383774,
+      "learning_rate": 5.350571979273573e-06,
+      "loss": 8.5444,
+      "step": 164620
+    },
+    {
+      "epoch": 0.8221428749781517,
+      "grad_norm": 0.09132785350084305,
+      "learning_rate": 5.349070064332023e-06,
+      "loss": 8.5611,
+      "step": 164630
+    },
+    {
+      "epoch": 0.8221928138030912,
+      "grad_norm": 0.09053760766983032,
+      "learning_rate": 5.3475681493904735e-06,
+      "loss": 8.5465,
+      "step": 164640
+    },
+    {
+      "epoch": 0.8222427526280307,
+      "grad_norm": 0.1000586673617363,
+      "learning_rate": 5.346066234448922e-06,
+      "loss": 8.5449,
+      "step": 164650
+    },
+    {
+      "epoch": 0.8222926914529701,
+      "grad_norm": 0.09241247922182083,
+      "learning_rate": 5.344564319507372e-06,
+      "loss": 8.5303,
+      "step": 164660
+    },
+    {
+      "epoch": 0.8223426302779095,
+      "grad_norm": 0.09306442737579346,
+      "learning_rate": 5.3430624045658215e-06,
+      "loss": 8.5248,
+      "step": 164670
+    },
+    {
+      "epoch": 0.822392569102849,
+      "grad_norm": 0.09631567448377609,
+      "learning_rate": 5.341560489624271e-06,
+      "loss": 8.5448,
+      "step": 164680
+    },
+    {
+      "epoch": 0.8224425079277885,
+      "grad_norm": 0.09375005215406418,
+      "learning_rate": 5.340058574682721e-06,
+      "loss": 8.5468,
+      "step": 164690
+    },
+    {
+      "epoch": 0.8224924467527279,
+      "grad_norm": 0.08916940540075302,
+      "learning_rate": 5.3385566597411695e-06,
+      "loss": 8.5525,
+      "step": 164700
+    },
+    {
+      "epoch": 0.8225423855776673,
+      "grad_norm": 0.09325413405895233,
+      "learning_rate": 5.33705474479962e-06,
+      "loss": 8.5474,
+      "step": 164710
+    },
+    {
+      "epoch": 0.8225923244026068,
+      "grad_norm": 0.08860702812671661,
+      "learning_rate": 5.335552829858069e-06,
+      "loss": 8.5582,
+      "step": 164720
+    },
+    {
+      "epoch": 0.8226422632275463,
+      "grad_norm": 0.09240319579839706,
+      "learning_rate": 5.334050914916518e-06,
+      "loss": 8.5351,
+      "step": 164730
+    },
+    {
+      "epoch": 0.8226922020524857,
+      "grad_norm": 0.09109298884868622,
+      "learning_rate": 5.3325489999749685e-06,
+      "loss": 8.5478,
+      "step": 164740
+    },
+    {
+      "epoch": 0.8227421408774251,
+      "grad_norm": 0.09379500150680542,
+      "learning_rate": 5.331047085033418e-06,
+      "loss": 8.5532,
+      "step": 164750
+    },
+    {
+      "epoch": 0.8227920797023646,
+      "grad_norm": 0.08987203985452652,
+      "learning_rate": 5.329545170091867e-06,
+      "loss": 8.554,
+      "step": 164760
+    },
+    {
+      "epoch": 0.8228420185273041,
+      "grad_norm": 0.08933199197053909,
+      "learning_rate": 5.3280432551503165e-06,
+      "loss": 8.5402,
+      "step": 164770
+    },
+    {
+      "epoch": 0.8228919573522435,
+      "grad_norm": 0.08755306154489517,
+      "learning_rate": 5.326541340208766e-06,
+      "loss": 8.5478,
+      "step": 164780
+    },
+    {
+      "epoch": 0.8229418961771829,
+      "grad_norm": 0.0940420925617218,
+      "learning_rate": 5.325039425267216e-06,
+      "loss": 8.5383,
+      "step": 164790
+    },
+    {
+      "epoch": 0.8229918350021224,
+      "grad_norm": 0.08608761429786682,
+      "learning_rate": 5.323537510325665e-06,
+      "loss": 8.5539,
+      "step": 164800
+    },
+    {
+      "epoch": 0.8230417738270619,
+      "grad_norm": 0.0908665880560875,
+      "learning_rate": 5.322035595384115e-06,
+      "loss": 8.5249,
+      "step": 164810
+    },
+    {
+      "epoch": 0.8230917126520013,
+      "grad_norm": 0.08952327817678452,
+      "learning_rate": 5.320533680442564e-06,
+      "loss": 8.5568,
+      "step": 164820
+    },
+    {
+      "epoch": 0.8231416514769407,
+      "grad_norm": 0.09053060412406921,
+      "learning_rate": 5.319031765501014e-06,
+      "loss": 8.5554,
+      "step": 164830
+    },
+    {
+      "epoch": 0.8231915903018802,
+      "grad_norm": 0.08716212213039398,
+      "learning_rate": 5.3175298505594636e-06,
+      "loss": 8.5512,
+      "step": 164840
+    },
+    {
+      "epoch": 0.8232415291268197,
+      "grad_norm": 0.08893406391143799,
+      "learning_rate": 5.316027935617913e-06,
+      "loss": 8.557,
+      "step": 164850
+    },
+    {
+      "epoch": 0.8232914679517591,
+      "grad_norm": 0.09567388892173767,
+      "learning_rate": 5.314526020676362e-06,
+      "loss": 8.5658,
+      "step": 164860
+    },
+    {
+      "epoch": 0.8233414067766985,
+      "grad_norm": 0.09032758325338364,
+      "learning_rate": 5.3130241057348115e-06,
+      "loss": 8.5635,
+      "step": 164870
+    },
+    {
+      "epoch": 0.823391345601638,
+      "grad_norm": 0.0944572165608406,
+      "learning_rate": 5.311522190793262e-06,
+      "loss": 8.5377,
+      "step": 164880
+    },
+    {
+      "epoch": 0.8234412844265775,
+      "grad_norm": 0.08695536106824875,
+      "learning_rate": 5.310020275851711e-06,
+      "loss": 8.5491,
+      "step": 164890
+    },
+    {
+      "epoch": 0.8234912232515169,
+      "grad_norm": 0.0900052860379219,
+      "learning_rate": 5.30851836091016e-06,
+      "loss": 8.5341,
+      "step": 164900
+    },
+    {
+      "epoch": 0.8235411620764563,
+      "grad_norm": 0.09355796128511429,
+      "learning_rate": 5.3070164459686106e-06,
+      "loss": 8.5427,
+      "step": 164910
+    },
+    {
+      "epoch": 0.8235911009013958,
+      "grad_norm": 0.09041264653205872,
+      "learning_rate": 5.305514531027059e-06,
+      "loss": 8.5535,
+      "step": 164920
+    },
+    {
+      "epoch": 0.8236410397263353,
+      "grad_norm": 0.09315498918294907,
+      "learning_rate": 5.304012616085509e-06,
+      "loss": 8.5472,
+      "step": 164930
+    },
+    {
+      "epoch": 0.8236909785512747,
+      "grad_norm": 0.08934438228607178,
+      "learning_rate": 5.3025107011439586e-06,
+      "loss": 8.5446,
+      "step": 164940
+    },
+    {
+      "epoch": 0.8237409173762141,
+      "grad_norm": 0.09067961573600769,
+      "learning_rate": 5.301008786202408e-06,
+      "loss": 8.5461,
+      "step": 164950
+    },
+    {
+      "epoch": 0.8237908562011536,
+      "grad_norm": 0.08916930854320526,
+      "learning_rate": 5.299506871260858e-06,
+      "loss": 8.5513,
+      "step": 164960
+    },
+    {
+      "epoch": 0.8238407950260931,
+      "grad_norm": 0.09387677907943726,
+      "learning_rate": 5.298004956319307e-06,
+      "loss": 8.5447,
+      "step": 164970
+    },
+    {
+      "epoch": 0.8238907338510325,
+      "grad_norm": 0.08811572939157486,
+      "learning_rate": 5.296503041377757e-06,
+      "loss": 8.5657,
+      "step": 164980
+    },
+    {
+      "epoch": 0.8239406726759719,
+      "grad_norm": 0.09204726666212082,
+      "learning_rate": 5.295001126436207e-06,
+      "loss": 8.5476,
+      "step": 164990
+    },
+    {
+      "epoch": 0.8239906115009114,
+      "grad_norm": 0.0907144770026207,
+      "learning_rate": 5.293499211494655e-06,
+      "loss": 8.5335,
+      "step": 165000
+    },
+    {
+      "epoch": 0.8240405503258509,
+      "grad_norm": 0.09380396455526352,
+      "learning_rate": 5.2919972965531056e-06,
+      "loss": 8.5313,
+      "step": 165010
+    },
+    {
+      "epoch": 0.8240904891507903,
+      "grad_norm": 0.09004945307970047,
+      "learning_rate": 5.290495381611555e-06,
+      "loss": 8.5442,
+      "step": 165020
+    },
+    {
+      "epoch": 0.8241404279757297,
+      "grad_norm": 0.08934416621923447,
+      "learning_rate": 5.288993466670004e-06,
+      "loss": 8.5422,
+      "step": 165030
+    },
+    {
+      "epoch": 0.8241903668006691,
+      "grad_norm": 0.08731923997402191,
+      "learning_rate": 5.287491551728454e-06,
+      "loss": 8.5354,
+      "step": 165040
+    },
+    {
+      "epoch": 0.8242403056256087,
+      "grad_norm": 0.09137827903032303,
+      "learning_rate": 5.285989636786903e-06,
+      "loss": 8.5595,
+      "step": 165050
+    },
+    {
+      "epoch": 0.8242902444505481,
+      "grad_norm": 0.09268289804458618,
+      "learning_rate": 5.284487721845353e-06,
+      "loss": 8.559,
+      "step": 165060
+    },
+    {
+      "epoch": 0.8243401832754875,
+      "grad_norm": 0.0933336690068245,
+      "learning_rate": 5.282985806903803e-06,
+      "loss": 8.535,
+      "step": 165070
+    },
+    {
+      "epoch": 0.824390122100427,
+      "grad_norm": 0.09035741537809372,
+      "learning_rate": 5.281483891962252e-06,
+      "loss": 8.546,
+      "step": 165080
+    },
+    {
+      "epoch": 0.8244400609253665,
+      "grad_norm": 0.08535844087600708,
+      "learning_rate": 5.279981977020702e-06,
+      "loss": 8.5388,
+      "step": 165090
+    },
+    {
+      "epoch": 0.8244899997503059,
+      "grad_norm": 0.09241851419210434,
+      "learning_rate": 5.27848006207915e-06,
+      "loss": 8.5666,
+      "step": 165100
+    },
+    {
+      "epoch": 0.8245399385752453,
+      "grad_norm": 0.09012720733880997,
+      "learning_rate": 5.276978147137601e-06,
+      "loss": 8.5522,
+      "step": 165110
+    },
+    {
+      "epoch": 0.8245898774001847,
+      "grad_norm": 0.09050553292036057,
+      "learning_rate": 5.275476232196051e-06,
+      "loss": 8.5381,
+      "step": 165120
+    },
+    {
+      "epoch": 0.8246398162251243,
+      "grad_norm": 0.09144661575555801,
+      "learning_rate": 5.273974317254499e-06,
+      "loss": 8.5375,
+      "step": 165130
+    },
+    {
+      "epoch": 0.8246897550500637,
+      "grad_norm": 0.08838358521461487,
+      "learning_rate": 5.272472402312949e-06,
+      "loss": 8.5498,
+      "step": 165140
+    },
+    {
+      "epoch": 0.8247396938750031,
+      "grad_norm": 0.09238158166408539,
+      "learning_rate": 5.270970487371399e-06,
+      "loss": 8.538,
+      "step": 165150
+    },
+    {
+      "epoch": 0.8247896326999425,
+      "grad_norm": 0.09907042235136032,
+      "learning_rate": 5.269468572429848e-06,
+      "loss": 8.5508,
+      "step": 165160
+    },
+    {
+      "epoch": 0.824839571524882,
+      "grad_norm": 0.08940906822681427,
+      "learning_rate": 5.267966657488298e-06,
+      "loss": 8.5438,
+      "step": 165170
+    },
+    {
+      "epoch": 0.8248895103498215,
+      "grad_norm": 0.09530402719974518,
+      "learning_rate": 5.266464742546747e-06,
+      "loss": 8.5375,
+      "step": 165180
+    },
+    {
+      "epoch": 0.8249394491747609,
+      "grad_norm": 0.0869201272726059,
+      "learning_rate": 5.264962827605197e-06,
+      "loss": 8.5371,
+      "step": 165190
+    },
+    {
+      "epoch": 0.8249893879997003,
+      "grad_norm": 0.08808686584234238,
+      "learning_rate": 5.263460912663646e-06,
+      "loss": 8.5476,
+      "step": 165200
+    },
+    {
+      "epoch": 0.8250393268246398,
+      "grad_norm": 0.09098785370588303,
+      "learning_rate": 5.261958997722096e-06,
+      "loss": 8.5324,
+      "step": 165210
+    },
+    {
+      "epoch": 0.8250892656495793,
+      "grad_norm": 0.09204449504613876,
+      "learning_rate": 5.260457082780546e-06,
+      "loss": 8.5425,
+      "step": 165220
+    },
+    {
+      "epoch": 0.8251392044745187,
+      "grad_norm": 0.09334173798561096,
+      "learning_rate": 5.258955167838995e-06,
+      "loss": 8.5315,
+      "step": 165230
+    },
+    {
+      "epoch": 0.8251891432994581,
+      "grad_norm": 0.08749334514141083,
+      "learning_rate": 5.2574532528974444e-06,
+      "loss": 8.5364,
+      "step": 165240
+    },
+    {
+      "epoch": 0.8252390821243976,
+      "grad_norm": 0.09419584274291992,
+      "learning_rate": 5.255951337955894e-06,
+      "loss": 8.5426,
+      "step": 165250
+    },
+    {
+      "epoch": 0.8252890209493371,
+      "grad_norm": 0.09130842983722687,
+      "learning_rate": 5.254449423014343e-06,
+      "loss": 8.5287,
+      "step": 165260
+    },
+    {
+      "epoch": 0.8253389597742765,
+      "grad_norm": 0.0886424109339714,
+      "learning_rate": 5.252947508072793e-06,
+      "loss": 8.5396,
+      "step": 165270
+    },
+    {
+      "epoch": 0.8253888985992159,
+      "grad_norm": 0.09385290741920471,
+      "learning_rate": 5.251445593131243e-06,
+      "loss": 8.5361,
+      "step": 165280
+    },
+    {
+      "epoch": 0.8254388374241554,
+      "grad_norm": 0.0890863835811615,
+      "learning_rate": 5.249943678189692e-06,
+      "loss": 8.5467,
+      "step": 165290
+    },
+    {
+      "epoch": 0.8254887762490949,
+      "grad_norm": 0.08851329982280731,
+      "learning_rate": 5.248441763248141e-06,
+      "loss": 8.5532,
+      "step": 165300
+    },
+    {
+      "epoch": 0.8255387150740343,
+      "grad_norm": 0.09374735504388809,
+      "learning_rate": 5.2469398483065914e-06,
+      "loss": 8.5461,
+      "step": 165310
+    },
+    {
+      "epoch": 0.8255886538989737,
+      "grad_norm": 0.08806613087654114,
+      "learning_rate": 5.245437933365041e-06,
+      "loss": 8.5575,
+      "step": 165320
+    },
+    {
+      "epoch": 0.8256385927239132,
+      "grad_norm": 0.09200981259346008,
+      "learning_rate": 5.24393601842349e-06,
+      "loss": 8.5345,
+      "step": 165330
+    },
+    {
+      "epoch": 0.8256885315488527,
+      "grad_norm": 0.08689451217651367,
+      "learning_rate": 5.2424341034819394e-06,
+      "loss": 8.5552,
+      "step": 165340
+    },
+    {
+      "epoch": 0.8257384703737921,
+      "grad_norm": 0.08550180494785309,
+      "learning_rate": 5.240932188540389e-06,
+      "loss": 8.5494,
+      "step": 165350
+    },
+    {
+      "epoch": 0.8257884091987315,
+      "grad_norm": 0.09254343062639236,
+      "learning_rate": 5.239430273598839e-06,
+      "loss": 8.5475,
+      "step": 165360
+    },
+    {
+      "epoch": 0.825838348023671,
+      "grad_norm": 0.08904233574867249,
+      "learning_rate": 5.237928358657288e-06,
+      "loss": 8.5499,
+      "step": 165370
+    },
+    {
+      "epoch": 0.8258882868486105,
+      "grad_norm": 0.09471232444047928,
+      "learning_rate": 5.236426443715738e-06,
+      "loss": 8.5489,
+      "step": 165380
+    },
+    {
+      "epoch": 0.8259382256735499,
+      "grad_norm": 0.09248729050159454,
+      "learning_rate": 5.234924528774188e-06,
+      "loss": 8.5425,
+      "step": 165390
+    },
+    {
+      "epoch": 0.8259881644984893,
+      "grad_norm": 0.09157528728246689,
+      "learning_rate": 5.233422613832636e-06,
+      "loss": 8.5438,
+      "step": 165400
+    },
+    {
+      "epoch": 0.8260381033234288,
+      "grad_norm": 0.09400694817304611,
+      "learning_rate": 5.2319206988910864e-06,
+      "loss": 8.5369,
+      "step": 165410
+    },
+    {
+      "epoch": 0.8260880421483683,
+      "grad_norm": 0.10043618083000183,
+      "learning_rate": 5.230418783949536e-06,
+      "loss": 8.5332,
+      "step": 165420
+    },
+    {
+      "epoch": 0.8261379809733077,
+      "grad_norm": 0.09534844756126404,
+      "learning_rate": 5.228916869007985e-06,
+      "loss": 8.5501,
+      "step": 165430
+    },
+    {
+      "epoch": 0.8261879197982471,
+      "grad_norm": 0.0883263498544693,
+      "learning_rate": 5.227414954066435e-06,
+      "loss": 8.5382,
+      "step": 165440
+    },
+    {
+      "epoch": 0.8262378586231865,
+      "grad_norm": 0.08956603705883026,
+      "learning_rate": 5.225913039124884e-06,
+      "loss": 8.5308,
+      "step": 165450
+    },
+    {
+      "epoch": 0.8262877974481261,
+      "grad_norm": 0.09134762734174728,
+      "learning_rate": 5.224411124183334e-06,
+      "loss": 8.5469,
+      "step": 165460
+    },
+    {
+      "epoch": 0.8263377362730655,
+      "grad_norm": 0.09247227758169174,
+      "learning_rate": 5.222909209241784e-06,
+      "loss": 8.5423,
+      "step": 165470
+    },
+    {
+      "epoch": 0.8263876750980049,
+      "grad_norm": 0.09392131119966507,
+      "learning_rate": 5.221407294300233e-06,
+      "loss": 8.5436,
+      "step": 165480
+    },
+    {
+      "epoch": 0.8264376139229443,
+      "grad_norm": 0.0913129672408104,
+      "learning_rate": 5.219905379358683e-06,
+      "loss": 8.5342,
+      "step": 165490
+    },
+    {
+      "epoch": 0.8264875527478839,
+      "grad_norm": 0.09089593589305878,
+      "learning_rate": 5.218403464417131e-06,
+      "loss": 8.5391,
+      "step": 165500
+    },
+    {
+      "epoch": 0.8265374915728233,
+      "grad_norm": 0.08943469077348709,
+      "learning_rate": 5.2169015494755815e-06,
+      "loss": 8.537,
+      "step": 165510
+    },
+    {
+      "epoch": 0.8265874303977627,
+      "grad_norm": 0.09096402674913406,
+      "learning_rate": 5.215399634534032e-06,
+      "loss": 8.5331,
+      "step": 165520
+    },
+    {
+      "epoch": 0.8266373692227021,
+      "grad_norm": 0.09284608066082001,
+      "learning_rate": 5.21389771959248e-06,
+      "loss": 8.5359,
+      "step": 165530
+    },
+    {
+      "epoch": 0.8266873080476417,
+      "grad_norm": 0.08914735913276672,
+      "learning_rate": 5.21239580465093e-06,
+      "loss": 8.5522,
+      "step": 165540
+    },
+    {
+      "epoch": 0.8267372468725811,
+      "grad_norm": 0.10136851668357849,
+      "learning_rate": 5.21089388970938e-06,
+      "loss": 8.5232,
+      "step": 165550
+    },
+    {
+      "epoch": 0.8267871856975205,
+      "grad_norm": 0.09322341531515121,
+      "learning_rate": 5.209391974767829e-06,
+      "loss": 8.5354,
+      "step": 165560
+    },
+    {
+      "epoch": 0.8268371245224599,
+      "grad_norm": 0.09235122054815292,
+      "learning_rate": 5.207890059826279e-06,
+      "loss": 8.5405,
+      "step": 165570
+    },
+    {
+      "epoch": 0.8268870633473995,
+      "grad_norm": 0.09125807136297226,
+      "learning_rate": 5.206388144884728e-06,
+      "loss": 8.5406,
+      "step": 165580
+    },
+    {
+      "epoch": 0.8269370021723389,
+      "grad_norm": 0.09730856120586395,
+      "learning_rate": 5.204886229943178e-06,
+      "loss": 8.5739,
+      "step": 165590
+    },
+    {
+      "epoch": 0.8269869409972783,
+      "grad_norm": 0.09331224858760834,
+      "learning_rate": 5.203384315001627e-06,
+      "loss": 8.5544,
+      "step": 165600
+    },
+    {
+      "epoch": 0.8270368798222177,
+      "grad_norm": 0.09145566821098328,
+      "learning_rate": 5.2018824000600765e-06,
+      "loss": 8.5305,
+      "step": 165610
+    },
+    {
+      "epoch": 0.8270868186471573,
+      "grad_norm": 0.09408487379550934,
+      "learning_rate": 5.200380485118527e-06,
+      "loss": 8.536,
+      "step": 165620
+    },
+    {
+      "epoch": 0.8271367574720967,
+      "grad_norm": 0.09319987148046494,
+      "learning_rate": 5.198878570176976e-06,
+      "loss": 8.5615,
+      "step": 165630
+    },
+    {
+      "epoch": 0.8271866962970361,
+      "grad_norm": 0.08938295394182205,
+      "learning_rate": 5.197376655235425e-06,
+      "loss": 8.5486,
+      "step": 165640
+    },
+    {
+      "epoch": 0.8272366351219755,
+      "grad_norm": 0.09183337539434433,
+      "learning_rate": 5.195874740293875e-06,
+      "loss": 8.5406,
+      "step": 165650
+    },
+    {
+      "epoch": 0.8272865739469151,
+      "grad_norm": 0.09237658977508545,
+      "learning_rate": 5.194372825352324e-06,
+      "loss": 8.5519,
+      "step": 165660
+    },
+    {
+      "epoch": 0.8273365127718545,
+      "grad_norm": 0.08938465267419815,
+      "learning_rate": 5.192870910410774e-06,
+      "loss": 8.5421,
+      "step": 165670
+    },
+    {
+      "epoch": 0.8273864515967939,
+      "grad_norm": 0.093377985060215,
+      "learning_rate": 5.1913689954692235e-06,
+      "loss": 8.5533,
+      "step": 165680
+    },
+    {
+      "epoch": 0.8274363904217333,
+      "grad_norm": 0.09096071869134903,
+      "learning_rate": 5.189867080527673e-06,
+      "loss": 8.5406,
+      "step": 165690
+    },
+    {
+      "epoch": 0.8274863292466729,
+      "grad_norm": 0.08967030048370361,
+      "learning_rate": 5.188365165586122e-06,
+      "loss": 8.5387,
+      "step": 165700
+    },
+    {
+      "epoch": 0.8275362680716123,
+      "grad_norm": 0.09120657294988632,
+      "learning_rate": 5.186863250644572e-06,
+      "loss": 8.5298,
+      "step": 165710
+    },
+    {
+      "epoch": 0.8275862068965517,
+      "grad_norm": 0.08661961555480957,
+      "learning_rate": 5.185361335703022e-06,
+      "loss": 8.538,
+      "step": 165720
+    },
+    {
+      "epoch": 0.8276361457214911,
+      "grad_norm": 0.0883423313498497,
+      "learning_rate": 5.183859420761471e-06,
+      "loss": 8.5444,
+      "step": 165730
+    },
+    {
+      "epoch": 0.8276860845464307,
+      "grad_norm": 0.09612096101045609,
+      "learning_rate": 5.18235750581992e-06,
+      "loss": 8.5432,
+      "step": 165740
+    },
+    {
+      "epoch": 0.8277360233713701,
+      "grad_norm": 0.08954691141843796,
+      "learning_rate": 5.18085559087837e-06,
+      "loss": 8.5485,
+      "step": 165750
+    },
+    {
+      "epoch": 0.8277859621963095,
+      "grad_norm": 0.09344494342803955,
+      "learning_rate": 5.17935367593682e-06,
+      "loss": 8.5501,
+      "step": 165760
+    },
+    {
+      "epoch": 0.8278359010212489,
+      "grad_norm": 0.09104925394058228,
+      "learning_rate": 5.177851760995269e-06,
+      "loss": 8.5515,
+      "step": 165770
+    },
+    {
+      "epoch": 0.8278858398461885,
+      "grad_norm": 0.09306161850690842,
+      "learning_rate": 5.1763498460537185e-06,
+      "loss": 8.5442,
+      "step": 165780
+    },
+    {
+      "epoch": 0.8279357786711279,
+      "grad_norm": 0.08997157216072083,
+      "learning_rate": 5.174847931112169e-06,
+      "loss": 8.5439,
+      "step": 165790
+    },
+    {
+      "epoch": 0.8279857174960673,
+      "grad_norm": 0.08354370296001434,
+      "learning_rate": 5.173346016170617e-06,
+      "loss": 8.5438,
+      "step": 165800
+    },
+    {
+      "epoch": 0.8280356563210067,
+      "grad_norm": 0.09156732261180878,
+      "learning_rate": 5.171844101229067e-06,
+      "loss": 8.5204,
+      "step": 165810
+    },
+    {
+      "epoch": 0.8280855951459463,
+      "grad_norm": 0.09536905586719513,
+      "learning_rate": 5.170342186287517e-06,
+      "loss": 8.5429,
+      "step": 165820
+    },
+    {
+      "epoch": 0.8281355339708857,
+      "grad_norm": 0.09520316123962402,
+      "learning_rate": 5.168840271345966e-06,
+      "loss": 8.5403,
+      "step": 165830
+    },
+    {
+      "epoch": 0.8281854727958251,
+      "grad_norm": 0.08938983827829361,
+      "learning_rate": 5.167338356404416e-06,
+      "loss": 8.5504,
+      "step": 165840
+    },
+    {
+      "epoch": 0.8282354116207645,
+      "grad_norm": 0.09314066916704178,
+      "learning_rate": 5.165836441462865e-06,
+      "loss": 8.5305,
+      "step": 165850
+    },
+    {
+      "epoch": 0.8282853504457041,
+      "grad_norm": 0.09131036698818207,
+      "learning_rate": 5.164334526521315e-06,
+      "loss": 8.5455,
+      "step": 165860
+    },
+    {
+      "epoch": 0.8283352892706435,
+      "grad_norm": 0.09065386652946472,
+      "learning_rate": 5.162832611579765e-06,
+      "loss": 8.5364,
+      "step": 165870
+    },
+    {
+      "epoch": 0.8283852280955829,
+      "grad_norm": 0.08912426233291626,
+      "learning_rate": 5.1613306966382135e-06,
+      "loss": 8.5208,
+      "step": 165880
+    },
+    {
+      "epoch": 0.8284351669205223,
+      "grad_norm": 0.0999988466501236,
+      "learning_rate": 5.159828781696664e-06,
+      "loss": 8.5334,
+      "step": 165890
+    },
+    {
+      "epoch": 0.8284851057454619,
+      "grad_norm": 0.09201761335134506,
+      "learning_rate": 5.158326866755112e-06,
+      "loss": 8.5401,
+      "step": 165900
+    },
+    {
+      "epoch": 0.8285350445704013,
+      "grad_norm": 0.08930778503417969,
+      "learning_rate": 5.156824951813562e-06,
+      "loss": 8.5424,
+      "step": 165910
+    },
+    {
+      "epoch": 0.8285849833953407,
+      "grad_norm": 0.08692941069602966,
+      "learning_rate": 5.1553230368720125e-06,
+      "loss": 8.554,
+      "step": 165920
+    },
+    {
+      "epoch": 0.8286349222202801,
+      "grad_norm": 0.09199637919664383,
+      "learning_rate": 5.153821121930461e-06,
+      "loss": 8.5407,
+      "step": 165930
+    },
+    {
+      "epoch": 0.8286848610452197,
+      "grad_norm": 0.09677516669034958,
+      "learning_rate": 5.152319206988911e-06,
+      "loss": 8.5306,
+      "step": 165940
+    },
+    {
+      "epoch": 0.8287347998701591,
+      "grad_norm": 0.0840715691447258,
+      "learning_rate": 5.1508172920473605e-06,
+      "loss": 8.5462,
+      "step": 165950
+    },
+    {
+      "epoch": 0.8287847386950985,
+      "grad_norm": 0.0945180281996727,
+      "learning_rate": 5.14931537710581e-06,
+      "loss": 8.5233,
+      "step": 165960
+    },
+    {
+      "epoch": 0.8288346775200379,
+      "grad_norm": 0.09411770850419998,
+      "learning_rate": 5.14781346216426e-06,
+      "loss": 8.5413,
+      "step": 165970
+    },
+    {
+      "epoch": 0.8288846163449775,
+      "grad_norm": 0.09039691835641861,
+      "learning_rate": 5.1463115472227085e-06,
+      "loss": 8.5273,
+      "step": 165980
+    },
+    {
+      "epoch": 0.8289345551699169,
+      "grad_norm": 0.09304576367139816,
+      "learning_rate": 5.144809632281159e-06,
+      "loss": 8.5407,
+      "step": 165990
+    },
+    {
+      "epoch": 0.8289844939948563,
+      "grad_norm": 0.09121018648147583,
+      "learning_rate": 5.143307717339608e-06,
+      "loss": 8.5493,
+      "step": 166000
+    },
+    {
+      "epoch": 0.8290344328197957,
+      "grad_norm": 0.0907362774014473,
+      "learning_rate": 5.141805802398057e-06,
+      "loss": 8.5375,
+      "step": 166010
+    },
+    {
+      "epoch": 0.8290843716447353,
+      "grad_norm": 0.0894848182797432,
+      "learning_rate": 5.1403038874565075e-06,
+      "loss": 8.5168,
+      "step": 166020
+    },
+    {
+      "epoch": 0.8291343104696747,
+      "grad_norm": 0.09260275959968567,
+      "learning_rate": 5.138801972514957e-06,
+      "loss": 8.5411,
+      "step": 166030
+    },
+    {
+      "epoch": 0.8291842492946141,
+      "grad_norm": 0.08942524343729019,
+      "learning_rate": 5.137300057573406e-06,
+      "loss": 8.5425,
+      "step": 166040
+    },
+    {
+      "epoch": 0.8292341881195535,
+      "grad_norm": 0.08970021456480026,
+      "learning_rate": 5.1357981426318555e-06,
+      "loss": 8.5597,
+      "step": 166050
+    },
+    {
+      "epoch": 0.829284126944493,
+      "grad_norm": 0.09671708196401596,
+      "learning_rate": 5.134296227690305e-06,
+      "loss": 8.5416,
+      "step": 166060
+    },
+    {
+      "epoch": 0.8293340657694325,
+      "grad_norm": 0.09718456864356995,
+      "learning_rate": 5.132794312748755e-06,
+      "loss": 8.5404,
+      "step": 166070
+    },
+    {
+      "epoch": 0.8293840045943719,
+      "grad_norm": 0.09095334261655807,
+      "learning_rate": 5.131292397807204e-06,
+      "loss": 8.5516,
+      "step": 166080
+    },
+    {
+      "epoch": 0.8294339434193113,
+      "grad_norm": 0.09243844449520111,
+      "learning_rate": 5.129790482865654e-06,
+      "loss": 8.5487,
+      "step": 166090
+    },
+    {
+      "epoch": 0.8294838822442508,
+      "grad_norm": 0.09933464974164963,
+      "learning_rate": 5.128288567924104e-06,
+      "loss": 8.5089,
+      "step": 166100
+    },
+    {
+      "epoch": 0.8295338210691903,
+      "grad_norm": 0.09080641716718674,
+      "learning_rate": 5.126786652982553e-06,
+      "loss": 8.5346,
+      "step": 166110
+    },
+    {
+      "epoch": 0.8295837598941297,
+      "grad_norm": 0.09354326874017715,
+      "learning_rate": 5.1252847380410025e-06,
+      "loss": 8.5647,
+      "step": 166120
+    },
+    {
+      "epoch": 0.8296336987190691,
+      "grad_norm": 0.08734377473592758,
+      "learning_rate": 5.123782823099452e-06,
+      "loss": 8.544,
+      "step": 166130
+    },
+    {
+      "epoch": 0.8296836375440085,
+      "grad_norm": 0.09000393003225327,
+      "learning_rate": 5.122280908157901e-06,
+      "loss": 8.532,
+      "step": 166140
+    },
+    {
+      "epoch": 0.8297335763689481,
+      "grad_norm": 0.08724624663591385,
+      "learning_rate": 5.120778993216351e-06,
+      "loss": 8.5537,
+      "step": 166150
+    },
+    {
+      "epoch": 0.8297835151938875,
+      "grad_norm": 0.09009451419115067,
+      "learning_rate": 5.119277078274801e-06,
+      "loss": 8.5666,
+      "step": 166160
+    },
+    {
+      "epoch": 0.8298334540188269,
+      "grad_norm": 0.08647624403238297,
+      "learning_rate": 5.11777516333325e-06,
+      "loss": 8.5261,
+      "step": 166170
+    },
+    {
+      "epoch": 0.8298833928437663,
+      "grad_norm": 0.09104521572589874,
+      "learning_rate": 5.116273248391699e-06,
+      "loss": 8.5332,
+      "step": 166180
+    },
+    {
+      "epoch": 0.8299333316687059,
+      "grad_norm": 0.0890510231256485,
+      "learning_rate": 5.1147713334501495e-06,
+      "loss": 8.5533,
+      "step": 166190
+    },
+    {
+      "epoch": 0.8299832704936453,
+      "grad_norm": 0.10203186422586441,
+      "learning_rate": 5.113269418508599e-06,
+      "loss": 8.5449,
+      "step": 166200
+    },
+    {
+      "epoch": 0.8300332093185847,
+      "grad_norm": 0.08849181234836578,
+      "learning_rate": 5.111767503567048e-06,
+      "loss": 8.5472,
+      "step": 166210
+    },
+    {
+      "epoch": 0.8300831481435241,
+      "grad_norm": 0.08866935223340988,
+      "learning_rate": 5.1102655886254975e-06,
+      "loss": 8.5432,
+      "step": 166220
+    },
+    {
+      "epoch": 0.8301330869684637,
+      "grad_norm": 0.08821461349725723,
+      "learning_rate": 5.108763673683947e-06,
+      "loss": 8.5514,
+      "step": 166230
+    },
+    {
+      "epoch": 0.8301830257934031,
+      "grad_norm": 0.09653317928314209,
+      "learning_rate": 5.107261758742397e-06,
+      "loss": 8.5564,
+      "step": 166240
+    },
+    {
+      "epoch": 0.8302329646183425,
+      "grad_norm": 0.09071271866559982,
+      "learning_rate": 5.105759843800846e-06,
+      "loss": 8.5254,
+      "step": 166250
+    },
+    {
+      "epoch": 0.8302829034432819,
+      "grad_norm": 0.09026788175106049,
+      "learning_rate": 5.104257928859296e-06,
+      "loss": 8.5428,
+      "step": 166260
+    },
+    {
+      "epoch": 0.8303328422682215,
+      "grad_norm": 0.08606173098087311,
+      "learning_rate": 5.102756013917746e-06,
+      "loss": 8.5425,
+      "step": 166270
+    },
+    {
+      "epoch": 0.8303827810931609,
+      "grad_norm": 0.09267769753932953,
+      "learning_rate": 5.101254098976194e-06,
+      "loss": 8.5487,
+      "step": 166280
+    },
+    {
+      "epoch": 0.8304327199181003,
+      "grad_norm": 0.08565813302993774,
+      "learning_rate": 5.0997521840346445e-06,
+      "loss": 8.5572,
+      "step": 166290
+    },
+    {
+      "epoch": 0.8304826587430397,
+      "grad_norm": 0.09123985469341278,
+      "learning_rate": 5.098250269093094e-06,
+      "loss": 8.5305,
+      "step": 166300
+    },
+    {
+      "epoch": 0.8305325975679793,
+      "grad_norm": 0.08855421096086502,
+      "learning_rate": 5.096748354151543e-06,
+      "loss": 8.5374,
+      "step": 166310
+    },
+    {
+      "epoch": 0.8305825363929187,
+      "grad_norm": 0.09398633986711502,
+      "learning_rate": 5.095246439209993e-06,
+      "loss": 8.5517,
+      "step": 166320
+    },
+    {
+      "epoch": 0.8306324752178581,
+      "grad_norm": 0.08958692103624344,
+      "learning_rate": 5.093744524268442e-06,
+      "loss": 8.5346,
+      "step": 166330
+    },
+    {
+      "epoch": 0.8306824140427975,
+      "grad_norm": 0.09231898933649063,
+      "learning_rate": 5.092242609326892e-06,
+      "loss": 8.5458,
+      "step": 166340
+    },
+    {
+      "epoch": 0.8307323528677371,
+      "grad_norm": 0.09124819934368134,
+      "learning_rate": 5.090740694385342e-06,
+      "loss": 8.5355,
+      "step": 166350
+    },
+    {
+      "epoch": 0.8307822916926765,
+      "grad_norm": 0.09533385187387466,
+      "learning_rate": 5.089238779443791e-06,
+      "loss": 8.5478,
+      "step": 166360
+    },
+    {
+      "epoch": 0.8308322305176159,
+      "grad_norm": 0.0879926010966301,
+      "learning_rate": 5.087736864502241e-06,
+      "loss": 8.5447,
+      "step": 166370
+    },
+    {
+      "epoch": 0.8308821693425553,
+      "grad_norm": 0.09048743546009064,
+      "learning_rate": 5.086234949560689e-06,
+      "loss": 8.5385,
+      "step": 166380
+    },
+    {
+      "epoch": 0.8309321081674949,
+      "grad_norm": 0.09626738727092743,
+      "learning_rate": 5.0847330346191396e-06,
+      "loss": 8.5422,
+      "step": 166390
+    },
+    {
+      "epoch": 0.8309820469924343,
+      "grad_norm": 0.09046705067157745,
+      "learning_rate": 5.08323111967759e-06,
+      "loss": 8.5361,
+      "step": 166400
+    },
+    {
+      "epoch": 0.8310319858173737,
+      "grad_norm": 0.0912972241640091,
+      "learning_rate": 5.081729204736038e-06,
+      "loss": 8.5421,
+      "step": 166410
+    },
+    {
+      "epoch": 0.8310819246423131,
+      "grad_norm": 0.08912695944309235,
+      "learning_rate": 5.080227289794488e-06,
+      "loss": 8.5457,
+      "step": 166420
+    },
+    {
+      "epoch": 0.8311318634672527,
+      "grad_norm": 0.09452259540557861,
+      "learning_rate": 5.078725374852938e-06,
+      "loss": 8.5539,
+      "step": 166430
+    },
+    {
+      "epoch": 0.8311818022921921,
+      "grad_norm": 0.0892590880393982,
+      "learning_rate": 5.077223459911387e-06,
+      "loss": 8.5343,
+      "step": 166440
+    },
+    {
+      "epoch": 0.8312317411171315,
+      "grad_norm": 0.09252618998289108,
+      "learning_rate": 5.075721544969837e-06,
+      "loss": 8.5364,
+      "step": 166450
+    },
+    {
+      "epoch": 0.8312816799420709,
+      "grad_norm": 0.0892346054315567,
+      "learning_rate": 5.074219630028286e-06,
+      "loss": 8.5372,
+      "step": 166460
+    },
+    {
+      "epoch": 0.8313316187670104,
+      "grad_norm": 0.08983513712882996,
+      "learning_rate": 5.072717715086736e-06,
+      "loss": 8.5547,
+      "step": 166470
+    },
+    {
+      "epoch": 0.8313815575919499,
+      "grad_norm": 0.09154525399208069,
+      "learning_rate": 5.071215800145185e-06,
+      "loss": 8.5316,
+      "step": 166480
+    },
+    {
+      "epoch": 0.8314314964168893,
+      "grad_norm": 0.09328502416610718,
+      "learning_rate": 5.0697138852036346e-06,
+      "loss": 8.5355,
+      "step": 166490
+    },
+    {
+      "epoch": 0.8314814352418287,
+      "grad_norm": 0.09083747118711472,
+      "learning_rate": 5.068211970262085e-06,
+      "loss": 8.5346,
+      "step": 166500
+    },
+    {
+      "epoch": 0.8315313740667682,
+      "grad_norm": 0.09179224073886871,
+      "learning_rate": 5.066710055320534e-06,
+      "loss": 8.5448,
+      "step": 166510
+    },
+    {
+      "epoch": 0.8315813128917077,
+      "grad_norm": 0.08915836364030838,
+      "learning_rate": 5.065208140378983e-06,
+      "loss": 8.5404,
+      "step": 166520
+    },
+    {
+      "epoch": 0.8316312517166471,
+      "grad_norm": 0.08934465050697327,
+      "learning_rate": 5.063706225437433e-06,
+      "loss": 8.5365,
+      "step": 166530
+    },
+    {
+      "epoch": 0.8316811905415865,
+      "grad_norm": 0.09658721089363098,
+      "learning_rate": 5.062204310495882e-06,
+      "loss": 8.5354,
+      "step": 166540
+    },
+    {
+      "epoch": 0.831731129366526,
+      "grad_norm": 0.09819358587265015,
+      "learning_rate": 5.060702395554332e-06,
+      "loss": 8.5305,
+      "step": 166550
+    },
+    {
+      "epoch": 0.8317810681914655,
+      "grad_norm": 0.0890253409743309,
+      "learning_rate": 5.0592004806127816e-06,
+      "loss": 8.5532,
+      "step": 166560
+    },
+    {
+      "epoch": 0.8318310070164049,
+      "grad_norm": 0.09140022099018097,
+      "learning_rate": 5.057698565671231e-06,
+      "loss": 8.5178,
+      "step": 166570
+    },
+    {
+      "epoch": 0.8318809458413443,
+      "grad_norm": 0.0952766090631485,
+      "learning_rate": 5.05619665072968e-06,
+      "loss": 8.5309,
+      "step": 166580
+    },
+    {
+      "epoch": 0.8319308846662838,
+      "grad_norm": 0.08961894363164902,
+      "learning_rate": 5.05469473578813e-06,
+      "loss": 8.5458,
+      "step": 166590
+    },
+    {
+      "epoch": 0.8319808234912233,
+      "grad_norm": 0.08949257433414459,
+      "learning_rate": 5.05319282084658e-06,
+      "loss": 8.5408,
+      "step": 166600
+    },
+    {
+      "epoch": 0.8320307623161627,
+      "grad_norm": 0.09538450837135315,
+      "learning_rate": 5.051690905905029e-06,
+      "loss": 8.5492,
+      "step": 166610
+    },
+    {
+      "epoch": 0.8320807011411021,
+      "grad_norm": 0.0917927548289299,
+      "learning_rate": 5.050188990963478e-06,
+      "loss": 8.5572,
+      "step": 166620
+    },
+    {
+      "epoch": 0.8321306399660416,
+      "grad_norm": 0.09348586946725845,
+      "learning_rate": 5.048687076021928e-06,
+      "loss": 8.5529,
+      "step": 166630
+    },
+    {
+      "epoch": 0.8321805787909811,
+      "grad_norm": 0.09373587369918823,
+      "learning_rate": 5.047185161080378e-06,
+      "loss": 8.527,
+      "step": 166640
+    },
+    {
+      "epoch": 0.8322305176159205,
+      "grad_norm": 0.08720920234918594,
+      "learning_rate": 5.045683246138827e-06,
+      "loss": 8.5178,
+      "step": 166650
+    },
+    {
+      "epoch": 0.8322804564408599,
+      "grad_norm": 0.08963396400213242,
+      "learning_rate": 5.044181331197277e-06,
+      "loss": 8.533,
+      "step": 166660
+    },
+    {
+      "epoch": 0.8323303952657994,
+      "grad_norm": 0.0868087187409401,
+      "learning_rate": 5.042679416255727e-06,
+      "loss": 8.5402,
+      "step": 166670
+    },
+    {
+      "epoch": 0.8323803340907389,
+      "grad_norm": 0.09393750131130219,
+      "learning_rate": 5.041177501314175e-06,
+      "loss": 8.5407,
+      "step": 166680
+    },
+    {
+      "epoch": 0.8324302729156783,
+      "grad_norm": 0.09192768484354019,
+      "learning_rate": 5.039675586372625e-06,
+      "loss": 8.5346,
+      "step": 166690
+    },
+    {
+      "epoch": 0.8324802117406177,
+      "grad_norm": 0.10022985190153122,
+      "learning_rate": 5.038173671431075e-06,
+      "loss": 8.5368,
+      "step": 166700
+    },
+    {
+      "epoch": 0.8325301505655572,
+      "grad_norm": 0.0937315970659256,
+      "learning_rate": 5.036671756489524e-06,
+      "loss": 8.5186,
+      "step": 166710
+    },
+    {
+      "epoch": 0.8325800893904967,
+      "grad_norm": 0.09312406182289124,
+      "learning_rate": 5.035169841547974e-06,
+      "loss": 8.5229,
+      "step": 166720
+    },
+    {
+      "epoch": 0.8326300282154361,
+      "grad_norm": 0.08995606750249863,
+      "learning_rate": 5.033667926606423e-06,
+      "loss": 8.5457,
+      "step": 166730
+    },
+    {
+      "epoch": 0.8326799670403755,
+      "grad_norm": 0.09093338251113892,
+      "learning_rate": 5.032166011664873e-06,
+      "loss": 8.5067,
+      "step": 166740
+    },
+    {
+      "epoch": 0.832729905865315,
+      "grad_norm": 0.0915585607290268,
+      "learning_rate": 5.030664096723322e-06,
+      "loss": 8.5639,
+      "step": 166750
+    },
+    {
+      "epoch": 0.8327798446902545,
+      "grad_norm": 0.09234713017940521,
+      "learning_rate": 5.029162181781772e-06,
+      "loss": 8.5346,
+      "step": 166760
+    },
+    {
+      "epoch": 0.8328297835151939,
+      "grad_norm": 0.08565306663513184,
+      "learning_rate": 5.027660266840222e-06,
+      "loss": 8.541,
+      "step": 166770
+    },
+    {
+      "epoch": 0.8328797223401333,
+      "grad_norm": 0.08641324937343597,
+      "learning_rate": 5.02615835189867e-06,
+      "loss": 8.5491,
+      "step": 166780
+    },
+    {
+      "epoch": 0.8329296611650728,
+      "grad_norm": 0.09579795598983765,
+      "learning_rate": 5.0246564369571204e-06,
+      "loss": 8.549,
+      "step": 166790
+    },
+    {
+      "epoch": 0.8329795999900123,
+      "grad_norm": 0.09320466220378876,
+      "learning_rate": 5.023154522015571e-06,
+      "loss": 8.5598,
+      "step": 166800
+    },
+    {
+      "epoch": 0.8330295388149517,
+      "grad_norm": 0.09680447727441788,
+      "learning_rate": 5.021652607074019e-06,
+      "loss": 8.5379,
+      "step": 166810
+    },
+    {
+      "epoch": 0.8330794776398911,
+      "grad_norm": 0.08433427661657333,
+      "learning_rate": 5.020150692132469e-06,
+      "loss": 8.5581,
+      "step": 166820
+    },
+    {
+      "epoch": 0.8331294164648306,
+      "grad_norm": 0.09262140095233917,
+      "learning_rate": 5.018648777190918e-06,
+      "loss": 8.524,
+      "step": 166830
+    },
+    {
+      "epoch": 0.83317935528977,
+      "grad_norm": 0.09593355655670166,
+      "learning_rate": 5.017146862249368e-06,
+      "loss": 8.5378,
+      "step": 166840
+    },
+    {
+      "epoch": 0.8332292941147095,
+      "grad_norm": 0.09165123105049133,
+      "learning_rate": 5.015644947307818e-06,
+      "loss": 8.54,
+      "step": 166850
+    },
+    {
+      "epoch": 0.8332792329396489,
+      "grad_norm": 0.09254296123981476,
+      "learning_rate": 5.014143032366267e-06,
+      "loss": 8.5392,
+      "step": 166860
+    },
+    {
+      "epoch": 0.8333291717645884,
+      "grad_norm": 0.0935739278793335,
+      "learning_rate": 5.012641117424717e-06,
+      "loss": 8.5476,
+      "step": 166870
+    },
+    {
+      "epoch": 0.8333791105895278,
+      "grad_norm": 0.09110825508832932,
+      "learning_rate": 5.011139202483166e-06,
+      "loss": 8.5356,
+      "step": 166880
+    },
+    {
+      "epoch": 0.8334290494144673,
+      "grad_norm": 0.09004233777523041,
+      "learning_rate": 5.0096372875416154e-06,
+      "loss": 8.5363,
+      "step": 166890
+    },
+    {
+      "epoch": 0.8334789882394067,
+      "grad_norm": 0.08766531944274902,
+      "learning_rate": 5.008135372600066e-06,
+      "loss": 8.5418,
+      "step": 166900
+    },
+    {
+      "epoch": 0.8335289270643462,
+      "grad_norm": 0.09655581414699554,
+      "learning_rate": 5.006633457658514e-06,
+      "loss": 8.5334,
+      "step": 166910
+    },
+    {
+      "epoch": 0.8335788658892856,
+      "grad_norm": 0.08657383173704147,
+      "learning_rate": 5.005131542716964e-06,
+      "loss": 8.5564,
+      "step": 166920
+    },
+    {
+      "epoch": 0.8336288047142251,
+      "grad_norm": 0.09356348216533661,
+      "learning_rate": 5.003629627775414e-06,
+      "loss": 8.5438,
+      "step": 166930
+    },
+    {
+      "epoch": 0.8336787435391645,
+      "grad_norm": 0.0967157855629921,
+      "learning_rate": 5.002127712833863e-06,
+      "loss": 8.5556,
+      "step": 166940
+    },
+    {
+      "epoch": 0.833728682364104,
+      "grad_norm": 0.09090452641248703,
+      "learning_rate": 5.000625797892313e-06,
+      "loss": 8.5707,
+      "step": 166950
+    },
+    {
+      "epoch": 0.8337786211890434,
+      "grad_norm": 0.09209217876195908,
+      "learning_rate": 4.9991238829507624e-06,
+      "loss": 8.5592,
+      "step": 166960
+    },
+    {
+      "epoch": 0.8338285600139829,
+      "grad_norm": 0.09082140028476715,
+      "learning_rate": 4.997621968009212e-06,
+      "loss": 8.5343,
+      "step": 166970
+    },
+    {
+      "epoch": 0.8338784988389223,
+      "grad_norm": 0.09138956665992737,
+      "learning_rate": 4.996120053067661e-06,
+      "loss": 8.5338,
+      "step": 166980
+    },
+    {
+      "epoch": 0.8339284376638618,
+      "grad_norm": 0.09424415230751038,
+      "learning_rate": 4.9946181381261104e-06,
+      "loss": 8.5322,
+      "step": 166990
+    },
+    {
+      "epoch": 0.8339783764888012,
+      "grad_norm": 0.09288877248764038,
+      "learning_rate": 4.993116223184561e-06,
+      "loss": 8.5364,
+      "step": 167000
+    },
+    {
+      "epoch": 0.8340283153137407,
+      "grad_norm": 0.08676401525735855,
+      "learning_rate": 4.99161430824301e-06,
+      "loss": 8.5354,
+      "step": 167010
+    },
+    {
+      "epoch": 0.8340782541386801,
+      "grad_norm": 0.08432342857122421,
+      "learning_rate": 4.990112393301459e-06,
+      "loss": 8.5373,
+      "step": 167020
+    },
+    {
+      "epoch": 0.8341281929636196,
+      "grad_norm": 0.09226579964160919,
+      "learning_rate": 4.988610478359909e-06,
+      "loss": 8.5151,
+      "step": 167030
+    },
+    {
+      "epoch": 0.834178131788559,
+      "grad_norm": 0.0887642651796341,
+      "learning_rate": 4.987108563418359e-06,
+      "loss": 8.5471,
+      "step": 167040
+    },
+    {
+      "epoch": 0.8342280706134985,
+      "grad_norm": 0.0907331258058548,
+      "learning_rate": 4.985606648476808e-06,
+      "loss": 8.5403,
+      "step": 167050
+    },
+    {
+      "epoch": 0.8342780094384379,
+      "grad_norm": 0.08834332972764969,
+      "learning_rate": 4.9841047335352575e-06,
+      "loss": 8.5228,
+      "step": 167060
+    },
+    {
+      "epoch": 0.8343279482633774,
+      "grad_norm": 0.08750323951244354,
+      "learning_rate": 4.982602818593707e-06,
+      "loss": 8.5424,
+      "step": 167070
+    },
+    {
+      "epoch": 0.8343778870883168,
+      "grad_norm": 0.08858982473611832,
+      "learning_rate": 4.981100903652156e-06,
+      "loss": 8.5551,
+      "step": 167080
+    },
+    {
+      "epoch": 0.8344278259132563,
+      "grad_norm": 0.09022153168916702,
+      "learning_rate": 4.979598988710606e-06,
+      "loss": 8.5344,
+      "step": 167090
+    },
+    {
+      "epoch": 0.8344777647381957,
+      "grad_norm": 0.0896066427230835,
+      "learning_rate": 4.978097073769056e-06,
+      "loss": 8.5426,
+      "step": 167100
+    },
+    {
+      "epoch": 0.8345277035631352,
+      "grad_norm": 0.09046115726232529,
+      "learning_rate": 4.976595158827505e-06,
+      "loss": 8.545,
+      "step": 167110
+    },
+    {
+      "epoch": 0.8345776423880746,
+      "grad_norm": 0.08858959376811981,
+      "learning_rate": 4.975093243885955e-06,
+      "loss": 8.5482,
+      "step": 167120
+    },
+    {
+      "epoch": 0.8346275812130141,
+      "grad_norm": 0.08675581961870193,
+      "learning_rate": 4.9735913289444045e-06,
+      "loss": 8.5382,
+      "step": 167130
+    },
+    {
+      "epoch": 0.8346775200379535,
+      "grad_norm": 0.08715889602899551,
+      "learning_rate": 4.972089414002854e-06,
+      "loss": 8.5394,
+      "step": 167140
+    },
+    {
+      "epoch": 0.8347274588628929,
+      "grad_norm": 0.0885719358921051,
+      "learning_rate": 4.970587499061303e-06,
+      "loss": 8.5276,
+      "step": 167150
+    },
+    {
+      "epoch": 0.8347773976878324,
+      "grad_norm": 0.09617254137992859,
+      "learning_rate": 4.9690855841197525e-06,
+      "loss": 8.5369,
+      "step": 167160
+    },
+    {
+      "epoch": 0.8348273365127719,
+      "grad_norm": 0.09246990084648132,
+      "learning_rate": 4.967583669178203e-06,
+      "loss": 8.5449,
+      "step": 167170
+    },
+    {
+      "epoch": 0.8348772753377113,
+      "grad_norm": 0.08657874912023544,
+      "learning_rate": 4.966081754236652e-06,
+      "loss": 8.5495,
+      "step": 167180
+    },
+    {
+      "epoch": 0.8349272141626507,
+      "grad_norm": 0.08776073902845383,
+      "learning_rate": 4.964579839295101e-06,
+      "loss": 8.5435,
+      "step": 167190
+    },
+    {
+      "epoch": 0.8349771529875902,
+      "grad_norm": 0.08979029953479767,
+      "learning_rate": 4.9630779243535515e-06,
+      "loss": 8.5452,
+      "step": 167200
+    },
+    {
+      "epoch": 0.8350270918125297,
+      "grad_norm": 0.09311855584383011,
+      "learning_rate": 4.961576009412e-06,
+      "loss": 8.5497,
+      "step": 167210
+    },
+    {
+      "epoch": 0.8350770306374691,
+      "grad_norm": 0.09252220392227173,
+      "learning_rate": 4.96007409447045e-06,
+      "loss": 8.5304,
+      "step": 167220
+    },
+    {
+      "epoch": 0.8351269694624085,
+      "grad_norm": 0.08825825899839401,
+      "learning_rate": 4.9585721795288995e-06,
+      "loss": 8.5425,
+      "step": 167230
+    },
+    {
+      "epoch": 0.835176908287348,
+      "grad_norm": 0.09545952081680298,
+      "learning_rate": 4.957070264587349e-06,
+      "loss": 8.5281,
+      "step": 167240
+    },
+    {
+      "epoch": 0.8352268471122875,
+      "grad_norm": 0.09352894127368927,
+      "learning_rate": 4.955568349645799e-06,
+      "loss": 8.5386,
+      "step": 167250
+    },
+    {
+      "epoch": 0.8352767859372269,
+      "grad_norm": 0.09909328073263168,
+      "learning_rate": 4.9540664347042475e-06,
+      "loss": 8.537,
+      "step": 167260
+    },
+    {
+      "epoch": 0.8353267247621663,
+      "grad_norm": 0.09386103600263596,
+      "learning_rate": 4.952564519762698e-06,
+      "loss": 8.545,
+      "step": 167270
+    },
+    {
+      "epoch": 0.8353766635871058,
+      "grad_norm": 0.08503425866365433,
+      "learning_rate": 4.951062604821148e-06,
+      "loss": 8.5802,
+      "step": 167280
+    },
+    {
+      "epoch": 0.8354266024120452,
+      "grad_norm": 0.09073678404092789,
+      "learning_rate": 4.949560689879596e-06,
+      "loss": 8.5313,
+      "step": 167290
+    },
+    {
+      "epoch": 0.8354765412369847,
+      "grad_norm": 0.09098771214485168,
+      "learning_rate": 4.9480587749380465e-06,
+      "loss": 8.5403,
+      "step": 167300
+    },
+    {
+      "epoch": 0.8355264800619241,
+      "grad_norm": 0.09336119145154953,
+      "learning_rate": 4.946556859996495e-06,
+      "loss": 8.5428,
+      "step": 167310
+    },
+    {
+      "epoch": 0.8355764188868636,
+      "grad_norm": 0.08903469145298004,
+      "learning_rate": 4.945054945054945e-06,
+      "loss": 8.5508,
+      "step": 167320
+    },
+    {
+      "epoch": 0.835626357711803,
+      "grad_norm": 0.08720388263463974,
+      "learning_rate": 4.943553030113395e-06,
+      "loss": 8.54,
+      "step": 167330
+    },
+    {
+      "epoch": 0.8356762965367425,
+      "grad_norm": 0.09315963834524155,
+      "learning_rate": 4.942051115171844e-06,
+      "loss": 8.549,
+      "step": 167340
+    },
+    {
+      "epoch": 0.8357262353616819,
+      "grad_norm": 0.08953353762626648,
+      "learning_rate": 4.940549200230294e-06,
+      "loss": 8.5557,
+      "step": 167350
+    },
+    {
+      "epoch": 0.8357761741866214,
+      "grad_norm": 0.0915287509560585,
+      "learning_rate": 4.939047285288743e-06,
+      "loss": 8.5594,
+      "step": 167360
+    },
+    {
+      "epoch": 0.8358261130115608,
+      "grad_norm": 0.09571286290884018,
+      "learning_rate": 4.937545370347193e-06,
+      "loss": 8.541,
+      "step": 167370
+    },
+    {
+      "epoch": 0.8358760518365003,
+      "grad_norm": 0.08816110342741013,
+      "learning_rate": 4.936043455405643e-06,
+      "loss": 8.5397,
+      "step": 167380
+    },
+    {
+      "epoch": 0.8359259906614397,
+      "grad_norm": 0.08931413292884827,
+      "learning_rate": 4.934541540464091e-06,
+      "loss": 8.5379,
+      "step": 167390
+    },
+    {
+      "epoch": 0.8359759294863792,
+      "grad_norm": 0.09298036247491837,
+      "learning_rate": 4.9330396255225415e-06,
+      "loss": 8.532,
+      "step": 167400
+    },
+    {
+      "epoch": 0.8360258683113186,
+      "grad_norm": 0.08843599259853363,
+      "learning_rate": 4.931537710580991e-06,
+      "loss": 8.539,
+      "step": 167410
+    },
+    {
+      "epoch": 0.8360758071362581,
+      "grad_norm": 0.0943143293261528,
+      "learning_rate": 4.93003579563944e-06,
+      "loss": 8.5463,
+      "step": 167420
+    },
+    {
+      "epoch": 0.8361257459611975,
+      "grad_norm": 0.08989337831735611,
+      "learning_rate": 4.92853388069789e-06,
+      "loss": 8.5562,
+      "step": 167430
+    },
+    {
+      "epoch": 0.836175684786137,
+      "grad_norm": 0.09261681884527206,
+      "learning_rate": 4.92703196575634e-06,
+      "loss": 8.5383,
+      "step": 167440
+    },
+    {
+      "epoch": 0.8362256236110764,
+      "grad_norm": 0.088409423828125,
+      "learning_rate": 4.925530050814789e-06,
+      "loss": 8.5393,
+      "step": 167450
+    },
+    {
+      "epoch": 0.8362755624360159,
+      "grad_norm": 0.0936957448720932,
+      "learning_rate": 4.924028135873238e-06,
+      "loss": 8.5375,
+      "step": 167460
+    },
+    {
+      "epoch": 0.8363255012609553,
+      "grad_norm": 0.08874178677797318,
+      "learning_rate": 4.922526220931688e-06,
+      "loss": 8.5396,
+      "step": 167470
+    },
+    {
+      "epoch": 0.8363754400858948,
+      "grad_norm": 0.09026958048343658,
+      "learning_rate": 4.921024305990138e-06,
+      "loss": 8.5348,
+      "step": 167480
+    },
+    {
+      "epoch": 0.8364253789108342,
+      "grad_norm": 0.09545276314020157,
+      "learning_rate": 4.919522391048587e-06,
+      "loss": 8.5391,
+      "step": 167490
+    },
+    {
+      "epoch": 0.8364753177357737,
+      "grad_norm": 0.08839011937379837,
+      "learning_rate": 4.9180204761070365e-06,
+      "loss": 8.5282,
+      "step": 167500
+    },
+    {
+      "epoch": 0.8365252565607131,
+      "grad_norm": 0.08772112429141998,
+      "learning_rate": 4.916518561165486e-06,
+      "loss": 8.5369,
+      "step": 167510
+    },
+    {
+      "epoch": 0.8365751953856526,
+      "grad_norm": 0.09278785437345505,
+      "learning_rate": 4.915016646223936e-06,
+      "loss": 8.5288,
+      "step": 167520
+    },
+    {
+      "epoch": 0.836625134210592,
+      "grad_norm": 0.09600167721509933,
+      "learning_rate": 4.913514731282385e-06,
+      "loss": 8.5258,
+      "step": 167530
+    },
+    {
+      "epoch": 0.8366750730355315,
+      "grad_norm": 0.08788032084703445,
+      "learning_rate": 4.912012816340835e-06,
+      "loss": 8.5342,
+      "step": 167540
+    },
+    {
+      "epoch": 0.8367250118604709,
+      "grad_norm": 0.08811583369970322,
+      "learning_rate": 4.910510901399284e-06,
+      "loss": 8.5487,
+      "step": 167550
+    },
+    {
+      "epoch": 0.8367749506854104,
+      "grad_norm": 0.08983556926250458,
+      "learning_rate": 4.909008986457733e-06,
+      "loss": 8.5451,
+      "step": 167560
+    },
+    {
+      "epoch": 0.8368248895103498,
+      "grad_norm": 0.08975329995155334,
+      "learning_rate": 4.9075070715161835e-06,
+      "loss": 8.5408,
+      "step": 167570
+    },
+    {
+      "epoch": 0.8368748283352893,
+      "grad_norm": 0.09413887560367584,
+      "learning_rate": 4.906005156574633e-06,
+      "loss": 8.5396,
+      "step": 167580
+    },
+    {
+      "epoch": 0.8369247671602287,
+      "grad_norm": 0.09020008891820908,
+      "learning_rate": 4.904503241633082e-06,
+      "loss": 8.5279,
+      "step": 167590
+    },
+    {
+      "epoch": 0.8369747059851682,
+      "grad_norm": 0.09177325665950775,
+      "learning_rate": 4.903001326691532e-06,
+      "loss": 8.5333,
+      "step": 167600
+    },
+    {
+      "epoch": 0.8370246448101076,
+      "grad_norm": 0.09214575588703156,
+      "learning_rate": 4.901499411749981e-06,
+      "loss": 8.5326,
+      "step": 167610
+    },
+    {
+      "epoch": 0.837074583635047,
+      "grad_norm": 0.09108281135559082,
+      "learning_rate": 4.899997496808431e-06,
+      "loss": 8.5556,
+      "step": 167620
+    },
+    {
+      "epoch": 0.8371245224599865,
+      "grad_norm": 0.09346097707748413,
+      "learning_rate": 4.89849558186688e-06,
+      "loss": 8.5301,
+      "step": 167630
+    },
+    {
+      "epoch": 0.837174461284926,
+      "grad_norm": 0.09483732283115387,
+      "learning_rate": 4.89699366692533e-06,
+      "loss": 8.5402,
+      "step": 167640
+    },
+    {
+      "epoch": 0.8372244001098654,
+      "grad_norm": 0.09763989597558975,
+      "learning_rate": 4.89549175198378e-06,
+      "loss": 8.5482,
+      "step": 167650
+    },
+    {
+      "epoch": 0.8372743389348049,
+      "grad_norm": 0.09391061961650848,
+      "learning_rate": 4.893989837042228e-06,
+      "loss": 8.5302,
+      "step": 167660
+    },
+    {
+      "epoch": 0.8373242777597443,
+      "grad_norm": 0.09133297950029373,
+      "learning_rate": 4.8924879221006785e-06,
+      "loss": 8.5292,
+      "step": 167670
+    },
+    {
+      "epoch": 0.8373742165846838,
+      "grad_norm": 0.09636905044317245,
+      "learning_rate": 4.890986007159129e-06,
+      "loss": 8.5271,
+      "step": 167680
+    },
+    {
+      "epoch": 0.8374241554096232,
+      "grad_norm": 0.09621647000312805,
+      "learning_rate": 4.889484092217577e-06,
+      "loss": 8.5309,
+      "step": 167690
+    },
+    {
+      "epoch": 0.8374740942345626,
+      "grad_norm": 0.09955976903438568,
+      "learning_rate": 4.887982177276027e-06,
+      "loss": 8.5377,
+      "step": 167700
+    },
+    {
+      "epoch": 0.8375240330595021,
+      "grad_norm": 0.08797106146812439,
+      "learning_rate": 4.886480262334476e-06,
+      "loss": 8.5314,
+      "step": 167710
+    },
+    {
+      "epoch": 0.8375739718844416,
+      "grad_norm": 0.08975420147180557,
+      "learning_rate": 4.884978347392926e-06,
+      "loss": 8.5436,
+      "step": 167720
+    },
+    {
+      "epoch": 0.837623910709381,
+      "grad_norm": 0.09256165474653244,
+      "learning_rate": 4.883476432451376e-06,
+      "loss": 8.5286,
+      "step": 167730
+    },
+    {
+      "epoch": 0.8376738495343204,
+      "grad_norm": 0.09067859500646591,
+      "learning_rate": 4.881974517509825e-06,
+      "loss": 8.5291,
+      "step": 167740
+    },
+    {
+      "epoch": 0.8377237883592599,
+      "grad_norm": 0.08854568004608154,
+      "learning_rate": 4.880472602568275e-06,
+      "loss": 8.5603,
+      "step": 167750
+    },
+    {
+      "epoch": 0.8377737271841994,
+      "grad_norm": 0.09424934536218643,
+      "learning_rate": 4.878970687626724e-06,
+      "loss": 8.544,
+      "step": 167760
+    },
+    {
+      "epoch": 0.8378236660091388,
+      "grad_norm": 0.08619468659162521,
+      "learning_rate": 4.8774687726851735e-06,
+      "loss": 8.5323,
+      "step": 167770
+    },
+    {
+      "epoch": 0.8378736048340782,
+      "grad_norm": 0.09160079061985016,
+      "learning_rate": 4.875966857743624e-06,
+      "loss": 8.5439,
+      "step": 167780
+    },
+    {
+      "epoch": 0.8379235436590177,
+      "grad_norm": 0.09085732698440552,
+      "learning_rate": 4.874464942802072e-06,
+      "loss": 8.54,
+      "step": 167790
+    },
+    {
+      "epoch": 0.8379734824839572,
+      "grad_norm": 0.08762364834547043,
+      "learning_rate": 4.872963027860522e-06,
+      "loss": 8.531,
+      "step": 167800
+    },
+    {
+      "epoch": 0.8380234213088966,
+      "grad_norm": 0.09410136938095093,
+      "learning_rate": 4.871461112918972e-06,
+      "loss": 8.5436,
+      "step": 167810
+    },
+    {
+      "epoch": 0.838073360133836,
+      "grad_norm": 0.0921444445848465,
+      "learning_rate": 4.869959197977421e-06,
+      "loss": 8.5332,
+      "step": 167820
+    },
+    {
+      "epoch": 0.8381232989587755,
+      "grad_norm": 0.09351623058319092,
+      "learning_rate": 4.868457283035871e-06,
+      "loss": 8.5444,
+      "step": 167830
+    },
+    {
+      "epoch": 0.838173237783715,
+      "grad_norm": 0.08999743312597275,
+      "learning_rate": 4.8669553680943205e-06,
+      "loss": 8.5551,
+      "step": 167840
+    },
+    {
+      "epoch": 0.8382231766086544,
+      "grad_norm": 0.09446761757135391,
+      "learning_rate": 4.86545345315277e-06,
+      "loss": 8.5314,
+      "step": 167850
+    },
+    {
+      "epoch": 0.8382731154335938,
+      "grad_norm": 0.09473822265863419,
+      "learning_rate": 4.863951538211219e-06,
+      "loss": 8.5381,
+      "step": 167860
+    },
+    {
+      "epoch": 0.8383230542585333,
+      "grad_norm": 0.09794218093156815,
+      "learning_rate": 4.8624496232696685e-06,
+      "loss": 8.5321,
+      "step": 167870
+    },
+    {
+      "epoch": 0.8383729930834728,
+      "grad_norm": 0.09353634715080261,
+      "learning_rate": 4.860947708328119e-06,
+      "loss": 8.5426,
+      "step": 167880
+    },
+    {
+      "epoch": 0.8384229319084122,
+      "grad_norm": 0.0915239006280899,
+      "learning_rate": 4.859445793386568e-06,
+      "loss": 8.5361,
+      "step": 167890
+    },
+    {
+      "epoch": 0.8384728707333516,
+      "grad_norm": 0.09556429833173752,
+      "learning_rate": 4.857943878445017e-06,
+      "loss": 8.5238,
+      "step": 167900
+    },
+    {
+      "epoch": 0.8385228095582911,
+      "grad_norm": 0.09241461008787155,
+      "learning_rate": 4.856441963503467e-06,
+      "loss": 8.5465,
+      "step": 167910
+    },
+    {
+      "epoch": 0.8385727483832306,
+      "grad_norm": 0.09638688713312149,
+      "learning_rate": 4.854940048561917e-06,
+      "loss": 8.5311,
+      "step": 167920
+    },
+    {
+      "epoch": 0.83862268720817,
+      "grad_norm": 0.09017990529537201,
+      "learning_rate": 4.853438133620366e-06,
+      "loss": 8.5524,
+      "step": 167930
+    },
+    {
+      "epoch": 0.8386726260331094,
+      "grad_norm": 0.0920003205537796,
+      "learning_rate": 4.8519362186788156e-06,
+      "loss": 8.5257,
+      "step": 167940
+    },
+    {
+      "epoch": 0.8387225648580489,
+      "grad_norm": 0.09186447411775589,
+      "learning_rate": 4.850434303737265e-06,
+      "loss": 8.5343,
+      "step": 167950
+    },
+    {
+      "epoch": 0.8387725036829884,
+      "grad_norm": 0.09250812232494354,
+      "learning_rate": 4.848932388795714e-06,
+      "loss": 8.5416,
+      "step": 167960
+    },
+    {
+      "epoch": 0.8388224425079278,
+      "grad_norm": 0.096126988530159,
+      "learning_rate": 4.847430473854164e-06,
+      "loss": 8.5307,
+      "step": 167970
+    },
+    {
+      "epoch": 0.8388723813328672,
+      "grad_norm": 0.09085419774055481,
+      "learning_rate": 4.845928558912614e-06,
+      "loss": 8.5487,
+      "step": 167980
+    },
+    {
+      "epoch": 0.8389223201578067,
+      "grad_norm": 0.09151645004749298,
+      "learning_rate": 4.844426643971063e-06,
+      "loss": 8.5358,
+      "step": 167990
+    },
+    {
+      "epoch": 0.8389722589827462,
+      "grad_norm": 0.09600605070590973,
+      "learning_rate": 4.842924729029513e-06,
+      "loss": 8.5415,
+      "step": 168000
+    },
+    {
+      "epoch": 0.8390221978076856,
+      "grad_norm": 0.08343446999788284,
+      "learning_rate": 4.841422814087962e-06,
+      "loss": 8.5486,
+      "step": 168010
+    },
+    {
+      "epoch": 0.839072136632625,
+      "grad_norm": 0.0970931351184845,
+      "learning_rate": 4.839920899146412e-06,
+      "loss": 8.5497,
+      "step": 168020
+    },
+    {
+      "epoch": 0.8391220754575645,
+      "grad_norm": 0.09128961712121964,
+      "learning_rate": 4.838418984204861e-06,
+      "loss": 8.5345,
+      "step": 168030
+    },
+    {
+      "epoch": 0.839172014282504,
+      "grad_norm": 0.08671637624502182,
+      "learning_rate": 4.8369170692633106e-06,
+      "loss": 8.5515,
+      "step": 168040
+    },
+    {
+      "epoch": 0.8392219531074434,
+      "grad_norm": 0.09237731248140335,
+      "learning_rate": 4.835415154321761e-06,
+      "loss": 8.4984,
+      "step": 168050
+    },
+    {
+      "epoch": 0.8392718919323828,
+      "grad_norm": 0.08762998133897781,
+      "learning_rate": 4.833913239380209e-06,
+      "loss": 8.5373,
+      "step": 168060
+    },
+    {
+      "epoch": 0.8393218307573223,
+      "grad_norm": 0.08894840627908707,
+      "learning_rate": 4.832411324438659e-06,
+      "loss": 8.5283,
+      "step": 168070
+    },
+    {
+      "epoch": 0.8393717695822618,
+      "grad_norm": 0.0915420651435852,
+      "learning_rate": 4.83090940949711e-06,
+      "loss": 8.5282,
+      "step": 168080
+    },
+    {
+      "epoch": 0.8394217084072012,
+      "grad_norm": 0.09219279140233994,
+      "learning_rate": 4.829407494555558e-06,
+      "loss": 8.5298,
+      "step": 168090
+    },
+    {
+      "epoch": 0.8394716472321406,
+      "grad_norm": 0.09041428565979004,
+      "learning_rate": 4.827905579614008e-06,
+      "loss": 8.5258,
+      "step": 168100
+    },
+    {
+      "epoch": 0.83952158605708,
+      "grad_norm": 0.08865369856357574,
+      "learning_rate": 4.826403664672457e-06,
+      "loss": 8.5421,
+      "step": 168110
+    },
+    {
+      "epoch": 0.8395715248820195,
+      "grad_norm": 0.09181597828865051,
+      "learning_rate": 4.824901749730907e-06,
+      "loss": 8.5309,
+      "step": 168120
+    },
+    {
+      "epoch": 0.839621463706959,
+      "grad_norm": 0.0943143367767334,
+      "learning_rate": 4.823399834789357e-06,
+      "loss": 8.5376,
+      "step": 168130
+    },
+    {
+      "epoch": 0.8396714025318984,
+      "grad_norm": 0.0917147845029831,
+      "learning_rate": 4.8218979198478056e-06,
+      "loss": 8.5423,
+      "step": 168140
+    },
+    {
+      "epoch": 0.8397213413568378,
+      "grad_norm": 0.09135729819536209,
+      "learning_rate": 4.820396004906256e-06,
+      "loss": 8.5337,
+      "step": 168150
+    },
+    {
+      "epoch": 0.8397712801817773,
+      "grad_norm": 0.09105221182107925,
+      "learning_rate": 4.818894089964705e-06,
+      "loss": 8.5277,
+      "step": 168160
+    },
+    {
+      "epoch": 0.8398212190067168,
+      "grad_norm": 0.09912960231304169,
+      "learning_rate": 4.817392175023154e-06,
+      "loss": 8.5492,
+      "step": 168170
+    },
+    {
+      "epoch": 0.8398711578316562,
+      "grad_norm": 0.08730529993772507,
+      "learning_rate": 4.815890260081605e-06,
+      "loss": 8.5422,
+      "step": 168180
+    },
+    {
+      "epoch": 0.8399210966565956,
+      "grad_norm": 0.09402299672365189,
+      "learning_rate": 4.814388345140053e-06,
+      "loss": 8.5343,
+      "step": 168190
+    },
+    {
+      "epoch": 0.8399710354815351,
+      "grad_norm": 0.09148437529802322,
+      "learning_rate": 4.812886430198503e-06,
+      "loss": 8.5231,
+      "step": 168200
+    },
+    {
+      "epoch": 0.8400209743064746,
+      "grad_norm": 0.08780474960803986,
+      "learning_rate": 4.811384515256953e-06,
+      "loss": 8.5322,
+      "step": 168210
+    },
+    {
+      "epoch": 0.840070913131414,
+      "grad_norm": 0.08754914999008179,
+      "learning_rate": 4.809882600315402e-06,
+      "loss": 8.5507,
+      "step": 168220
+    },
+    {
+      "epoch": 0.8401208519563534,
+      "grad_norm": 0.08717711269855499,
+      "learning_rate": 4.808380685373852e-06,
+      "loss": 8.5388,
+      "step": 168230
+    },
+    {
+      "epoch": 0.8401707907812929,
+      "grad_norm": 0.09865214675664902,
+      "learning_rate": 4.806878770432301e-06,
+      "loss": 8.5198,
+      "step": 168240
+    },
+    {
+      "epoch": 0.8402207296062324,
+      "grad_norm": 0.09098966419696808,
+      "learning_rate": 4.805376855490751e-06,
+      "loss": 8.5319,
+      "step": 168250
+    },
+    {
+      "epoch": 0.8402706684311718,
+      "grad_norm": 0.08804722875356674,
+      "learning_rate": 4.803874940549201e-06,
+      "loss": 8.5533,
+      "step": 168260
+    },
+    {
+      "epoch": 0.8403206072561112,
+      "grad_norm": 0.09023669362068176,
+      "learning_rate": 4.802373025607649e-06,
+      "loss": 8.5487,
+      "step": 168270
+    },
+    {
+      "epoch": 0.8403705460810507,
+      "grad_norm": 0.09287792444229126,
+      "learning_rate": 4.8008711106661e-06,
+      "loss": 8.5345,
+      "step": 168280
+    },
+    {
+      "epoch": 0.8404204849059902,
+      "grad_norm": 0.09334767609834671,
+      "learning_rate": 4.799369195724549e-06,
+      "loss": 8.5383,
+      "step": 168290
+    },
+    {
+      "epoch": 0.8404704237309296,
+      "grad_norm": 0.09085261076688766,
+      "learning_rate": 4.797867280782998e-06,
+      "loss": 8.5452,
+      "step": 168300
+    },
+    {
+      "epoch": 0.840520362555869,
+      "grad_norm": 0.08985559642314911,
+      "learning_rate": 4.7963653658414484e-06,
+      "loss": 8.5287,
+      "step": 168310
+    },
+    {
+      "epoch": 0.8405703013808085,
+      "grad_norm": 0.09192880243062973,
+      "learning_rate": 4.794863450899898e-06,
+      "loss": 8.5477,
+      "step": 168320
+    },
+    {
+      "epoch": 0.840620240205748,
+      "grad_norm": 0.08942985534667969,
+      "learning_rate": 4.793361535958347e-06,
+      "loss": 8.5286,
+      "step": 168330
+    },
+    {
+      "epoch": 0.8406701790306874,
+      "grad_norm": 0.09409672021865845,
+      "learning_rate": 4.7918596210167964e-06,
+      "loss": 8.5213,
+      "step": 168340
+    },
+    {
+      "epoch": 0.8407201178556268,
+      "grad_norm": 0.0878412127494812,
+      "learning_rate": 4.790357706075246e-06,
+      "loss": 8.5445,
+      "step": 168350
+    },
+    {
+      "epoch": 0.8407700566805663,
+      "grad_norm": 0.09042507410049438,
+      "learning_rate": 4.788855791133696e-06,
+      "loss": 8.5327,
+      "step": 168360
+    },
+    {
+      "epoch": 0.8408199955055058,
+      "grad_norm": 0.09061606973409653,
+      "learning_rate": 4.787353876192145e-06,
+      "loss": 8.5255,
+      "step": 168370
+    },
+    {
+      "epoch": 0.8408699343304452,
+      "grad_norm": 0.08747304230928421,
+      "learning_rate": 4.785851961250595e-06,
+      "loss": 8.5308,
+      "step": 168380
+    },
+    {
+      "epoch": 0.8409198731553846,
+      "grad_norm": 0.08752010762691498,
+      "learning_rate": 4.784350046309044e-06,
+      "loss": 8.5585,
+      "step": 168390
+    },
+    {
+      "epoch": 0.8409698119803241,
+      "grad_norm": 0.09220509976148605,
+      "learning_rate": 4.782848131367494e-06,
+      "loss": 8.5299,
+      "step": 168400
+    },
+    {
+      "epoch": 0.8410197508052636,
+      "grad_norm": 0.09601190686225891,
+      "learning_rate": 4.7813462164259434e-06,
+      "loss": 8.5213,
+      "step": 168410
+    },
+    {
+      "epoch": 0.841069689630203,
+      "grad_norm": 0.09144755452871323,
+      "learning_rate": 4.779844301484393e-06,
+      "loss": 8.5377,
+      "step": 168420
+    },
+    {
+      "epoch": 0.8411196284551424,
+      "grad_norm": 0.09117061644792557,
+      "learning_rate": 4.778342386542842e-06,
+      "loss": 8.5383,
+      "step": 168430
+    },
+    {
+      "epoch": 0.8411695672800819,
+      "grad_norm": 0.0880817323923111,
+      "learning_rate": 4.7768404716012914e-06,
+      "loss": 8.548,
+      "step": 168440
+    },
+    {
+      "epoch": 0.8412195061050214,
+      "grad_norm": 0.09438186883926392,
+      "learning_rate": 4.775338556659742e-06,
+      "loss": 8.5538,
+      "step": 168450
+    },
+    {
+      "epoch": 0.8412694449299608,
+      "grad_norm": 0.09241968393325806,
+      "learning_rate": 4.773836641718191e-06,
+      "loss": 8.5313,
+      "step": 168460
+    },
+    {
+      "epoch": 0.8413193837549002,
+      "grad_norm": 0.09603623300790787,
+      "learning_rate": 4.77233472677664e-06,
+      "loss": 8.5278,
+      "step": 168470
+    },
+    {
+      "epoch": 0.8413693225798397,
+      "grad_norm": 0.094639852643013,
+      "learning_rate": 4.7708328118350905e-06,
+      "loss": 8.5353,
+      "step": 168480
+    },
+    {
+      "epoch": 0.8414192614047792,
+      "grad_norm": 0.09024223685264587,
+      "learning_rate": 4.769330896893539e-06,
+      "loss": 8.5444,
+      "step": 168490
+    },
+    {
+      "epoch": 0.8414692002297186,
+      "grad_norm": 0.08873730897903442,
+      "learning_rate": 4.767828981951989e-06,
+      "loss": 8.523,
+      "step": 168500
+    },
+    {
+      "epoch": 0.841519139054658,
+      "grad_norm": 0.0888361856341362,
+      "learning_rate": 4.7663270670104384e-06,
+      "loss": 8.5363,
+      "step": 168510
+    },
+    {
+      "epoch": 0.8415690778795974,
+      "grad_norm": 0.0935666486620903,
+      "learning_rate": 4.764825152068888e-06,
+      "loss": 8.5435,
+      "step": 168520
+    },
+    {
+      "epoch": 0.841619016704537,
+      "grad_norm": 0.0903899073600769,
+      "learning_rate": 4.763323237127338e-06,
+      "loss": 8.5419,
+      "step": 168530
+    },
+    {
+      "epoch": 0.8416689555294764,
+      "grad_norm": 0.09109959751367569,
+      "learning_rate": 4.7618213221857864e-06,
+      "loss": 8.5273,
+      "step": 168540
+    },
+    {
+      "epoch": 0.8417188943544158,
+      "grad_norm": 0.08860781043767929,
+      "learning_rate": 4.760319407244237e-06,
+      "loss": 8.5301,
+      "step": 168550
+    },
+    {
+      "epoch": 0.8417688331793552,
+      "grad_norm": 0.09561283141374588,
+      "learning_rate": 4.758817492302687e-06,
+      "loss": 8.5381,
+      "step": 168560
+    },
+    {
+      "epoch": 0.8418187720042948,
+      "grad_norm": 0.09598873555660248,
+      "learning_rate": 4.757315577361135e-06,
+      "loss": 8.5349,
+      "step": 168570
+    },
+    {
+      "epoch": 0.8418687108292342,
+      "grad_norm": 0.09074491262435913,
+      "learning_rate": 4.7558136624195855e-06,
+      "loss": 8.5473,
+      "step": 168580
+    },
+    {
+      "epoch": 0.8419186496541736,
+      "grad_norm": 0.08866456151008606,
+      "learning_rate": 4.754311747478034e-06,
+      "loss": 8.5383,
+      "step": 168590
+    },
+    {
+      "epoch": 0.841968588479113,
+      "grad_norm": 0.09333822876214981,
+      "learning_rate": 4.752809832536484e-06,
+      "loss": 8.5318,
+      "step": 168600
+    },
+    {
+      "epoch": 0.8420185273040526,
+      "grad_norm": 0.09061551094055176,
+      "learning_rate": 4.751307917594934e-06,
+      "loss": 8.5349,
+      "step": 168610
+    },
+    {
+      "epoch": 0.842068466128992,
+      "grad_norm": 0.09228689223527908,
+      "learning_rate": 4.749806002653383e-06,
+      "loss": 8.5363,
+      "step": 168620
+    },
+    {
+      "epoch": 0.8421184049539314,
+      "grad_norm": 0.09009858220815659,
+      "learning_rate": 4.748304087711833e-06,
+      "loss": 8.5303,
+      "step": 168630
+    },
+    {
+      "epoch": 0.8421683437788708,
+      "grad_norm": 0.09529682248830795,
+      "learning_rate": 4.746802172770282e-06,
+      "loss": 8.5392,
+      "step": 168640
+    },
+    {
+      "epoch": 0.8422182826038104,
+      "grad_norm": 0.09646670520305634,
+      "learning_rate": 4.745300257828732e-06,
+      "loss": 8.5259,
+      "step": 168650
+    },
+    {
+      "epoch": 0.8422682214287498,
+      "grad_norm": 0.09157206863164902,
+      "learning_rate": 4.743798342887182e-06,
+      "loss": 8.5305,
+      "step": 168660
+    },
+    {
+      "epoch": 0.8423181602536892,
+      "grad_norm": 0.09323584288358688,
+      "learning_rate": 4.74229642794563e-06,
+      "loss": 8.5238,
+      "step": 168670
+    },
+    {
+      "epoch": 0.8423680990786286,
+      "grad_norm": 0.08792025595903397,
+      "learning_rate": 4.7407945130040805e-06,
+      "loss": 8.5433,
+      "step": 168680
+    },
+    {
+      "epoch": 0.8424180379035682,
+      "grad_norm": 0.08940023183822632,
+      "learning_rate": 4.73929259806253e-06,
+      "loss": 8.5252,
+      "step": 168690
+    },
+    {
+      "epoch": 0.8424679767285076,
+      "grad_norm": 0.09656330198049545,
+      "learning_rate": 4.737790683120979e-06,
+      "loss": 8.5101,
+      "step": 168700
+    },
+    {
+      "epoch": 0.842517915553447,
+      "grad_norm": 0.09180876612663269,
+      "learning_rate": 4.736288768179429e-06,
+      "loss": 8.5169,
+      "step": 168710
+    },
+    {
+      "epoch": 0.8425678543783864,
+      "grad_norm": 0.08640285581350327,
+      "learning_rate": 4.734786853237879e-06,
+      "loss": 8.5429,
+      "step": 168720
+    },
+    {
+      "epoch": 0.842617793203326,
+      "grad_norm": 0.0890103355050087,
+      "learning_rate": 4.733284938296328e-06,
+      "loss": 8.5434,
+      "step": 168730
+    },
+    {
+      "epoch": 0.8426677320282654,
+      "grad_norm": 0.09099707752466202,
+      "learning_rate": 4.731783023354777e-06,
+      "loss": 8.5422,
+      "step": 168740
+    },
+    {
+      "epoch": 0.8427176708532048,
+      "grad_norm": 0.09098941087722778,
+      "learning_rate": 4.730281108413227e-06,
+      "loss": 8.5546,
+      "step": 168750
+    },
+    {
+      "epoch": 0.8427676096781442,
+      "grad_norm": 0.09195592254400253,
+      "learning_rate": 4.728779193471677e-06,
+      "loss": 8.5293,
+      "step": 168760
+    },
+    {
+      "epoch": 0.8428175485030838,
+      "grad_norm": 0.09096460044384003,
+      "learning_rate": 4.727277278530126e-06,
+      "loss": 8.5446,
+      "step": 168770
+    },
+    {
+      "epoch": 0.8428674873280232,
+      "grad_norm": 0.08720100671052933,
+      "learning_rate": 4.7257753635885755e-06,
+      "loss": 8.5341,
+      "step": 168780
+    },
+    {
+      "epoch": 0.8429174261529626,
+      "grad_norm": 0.09274513274431229,
+      "learning_rate": 4.724273448647025e-06,
+      "loss": 8.5603,
+      "step": 168790
+    },
+    {
+      "epoch": 0.842967364977902,
+      "grad_norm": 0.09329745918512344,
+      "learning_rate": 4.722771533705475e-06,
+      "loss": 8.561,
+      "step": 168800
+    },
+    {
+      "epoch": 0.8430173038028416,
+      "grad_norm": 0.08319691568613052,
+      "learning_rate": 4.721269618763924e-06,
+      "loss": 8.531,
+      "step": 168810
+    },
+    {
+      "epoch": 0.843067242627781,
+      "grad_norm": 0.0921143889427185,
+      "learning_rate": 4.719767703822374e-06,
+      "loss": 8.539,
+      "step": 168820
+    },
+    {
+      "epoch": 0.8431171814527204,
+      "grad_norm": 0.09409576654434204,
+      "learning_rate": 4.718265788880823e-06,
+      "loss": 8.5339,
+      "step": 168830
+    },
+    {
+      "epoch": 0.8431671202776598,
+      "grad_norm": 0.08891736716032028,
+      "learning_rate": 4.716763873939272e-06,
+      "loss": 8.541,
+      "step": 168840
+    },
+    {
+      "epoch": 0.8432170591025994,
+      "grad_norm": 0.08662791550159454,
+      "learning_rate": 4.7152619589977225e-06,
+      "loss": 8.5403,
+      "step": 168850
+    },
+    {
+      "epoch": 0.8432669979275388,
+      "grad_norm": 0.09259507060050964,
+      "learning_rate": 4.713760044056172e-06,
+      "loss": 8.5186,
+      "step": 168860
+    },
+    {
+      "epoch": 0.8433169367524782,
+      "grad_norm": 0.09273169934749603,
+      "learning_rate": 4.712258129114621e-06,
+      "loss": 8.533,
+      "step": 168870
+    },
+    {
+      "epoch": 0.8433668755774176,
+      "grad_norm": 0.0853755995631218,
+      "learning_rate": 4.710756214173071e-06,
+      "loss": 8.5447,
+      "step": 168880
+    },
+    {
+      "epoch": 0.8434168144023572,
+      "grad_norm": 0.09415735304355621,
+      "learning_rate": 4.70925429923152e-06,
+      "loss": 8.5433,
+      "step": 168890
+    },
+    {
+      "epoch": 0.8434667532272966,
+      "grad_norm": 0.0893118679523468,
+      "learning_rate": 4.70775238428997e-06,
+      "loss": 8.5281,
+      "step": 168900
+    },
+    {
+      "epoch": 0.843516692052236,
+      "grad_norm": 0.09335058182477951,
+      "learning_rate": 4.706250469348419e-06,
+      "loss": 8.5326,
+      "step": 168910
+    },
+    {
+      "epoch": 0.8435666308771754,
+      "grad_norm": 0.09207750111818314,
+      "learning_rate": 4.704748554406869e-06,
+      "loss": 8.5416,
+      "step": 168920
+    },
+    {
+      "epoch": 0.843616569702115,
+      "grad_norm": 0.08903926610946655,
+      "learning_rate": 4.703246639465319e-06,
+      "loss": 8.5295,
+      "step": 168930
+    },
+    {
+      "epoch": 0.8436665085270544,
+      "grad_norm": 0.09720440953969955,
+      "learning_rate": 4.701744724523767e-06,
+      "loss": 8.5293,
+      "step": 168940
+    },
+    {
+      "epoch": 0.8437164473519938,
+      "grad_norm": 0.09019520878791809,
+      "learning_rate": 4.7002428095822175e-06,
+      "loss": 8.5264,
+      "step": 168950
+    },
+    {
+      "epoch": 0.8437663861769332,
+      "grad_norm": 0.08853203058242798,
+      "learning_rate": 4.698740894640668e-06,
+      "loss": 8.5345,
+      "step": 168960
+    },
+    {
+      "epoch": 0.8438163250018728,
+      "grad_norm": 0.09496580064296722,
+      "learning_rate": 4.697238979699116e-06,
+      "loss": 8.5143,
+      "step": 168970
+    },
+    {
+      "epoch": 0.8438662638268122,
+      "grad_norm": 0.09561099112033844,
+      "learning_rate": 4.695737064757566e-06,
+      "loss": 8.5355,
+      "step": 168980
+    },
+    {
+      "epoch": 0.8439162026517516,
+      "grad_norm": 0.08805878460407257,
+      "learning_rate": 4.694235149816015e-06,
+      "loss": 8.5346,
+      "step": 168990
+    },
+    {
+      "epoch": 0.843966141476691,
+      "grad_norm": 0.08811572194099426,
+      "learning_rate": 4.692733234874465e-06,
+      "loss": 8.5266,
+      "step": 169000
+    },
+    {
+      "epoch": 0.8440160803016306,
+      "grad_norm": 0.08896370977163315,
+      "learning_rate": 4.691231319932915e-06,
+      "loss": 8.5476,
+      "step": 169010
+    },
+    {
+      "epoch": 0.84406601912657,
+      "grad_norm": 0.09269924461841583,
+      "learning_rate": 4.689729404991364e-06,
+      "loss": 8.5217,
+      "step": 169020
+    },
+    {
+      "epoch": 0.8441159579515094,
+      "grad_norm": 0.0937482938170433,
+      "learning_rate": 4.688227490049814e-06,
+      "loss": 8.5131,
+      "step": 169030
+    },
+    {
+      "epoch": 0.8441658967764488,
+      "grad_norm": 0.09169386327266693,
+      "learning_rate": 4.686725575108263e-06,
+      "loss": 8.5409,
+      "step": 169040
+    },
+    {
+      "epoch": 0.8442158356013884,
+      "grad_norm": 0.09201053529977798,
+      "learning_rate": 4.6852236601667125e-06,
+      "loss": 8.5279,
+      "step": 169050
+    },
+    {
+      "epoch": 0.8442657744263278,
+      "grad_norm": 0.08865007758140564,
+      "learning_rate": 4.683721745225163e-06,
+      "loss": 8.5374,
+      "step": 169060
+    },
+    {
+      "epoch": 0.8443157132512672,
+      "grad_norm": 0.09043735265731812,
+      "learning_rate": 4.682219830283611e-06,
+      "loss": 8.5423,
+      "step": 169070
+    },
+    {
+      "epoch": 0.8443656520762066,
+      "grad_norm": 0.08876598626375198,
+      "learning_rate": 4.680717915342061e-06,
+      "loss": 8.5303,
+      "step": 169080
+    },
+    {
+      "epoch": 0.8444155909011462,
+      "grad_norm": 0.09009053558111191,
+      "learning_rate": 4.679216000400511e-06,
+      "loss": 8.5288,
+      "step": 169090
+    },
+    {
+      "epoch": 0.8444655297260856,
+      "grad_norm": 0.09224365651607513,
+      "learning_rate": 4.67771408545896e-06,
+      "loss": 8.5421,
+      "step": 169100
+    },
+    {
+      "epoch": 0.844515468551025,
+      "grad_norm": 0.09646381437778473,
+      "learning_rate": 4.67621217051741e-06,
+      "loss": 8.5252,
+      "step": 169110
+    },
+    {
+      "epoch": 0.8445654073759644,
+      "grad_norm": 0.09529738873243332,
+      "learning_rate": 4.6747102555758595e-06,
+      "loss": 8.5247,
+      "step": 169120
+    },
+    {
+      "epoch": 0.8446153462009038,
+      "grad_norm": 0.09440014511346817,
+      "learning_rate": 4.673208340634309e-06,
+      "loss": 8.5509,
+      "step": 169130
+    },
+    {
+      "epoch": 0.8446652850258434,
+      "grad_norm": 0.09085705131292343,
+      "learning_rate": 4.671706425692758e-06,
+      "loss": 8.5411,
+      "step": 169140
+    },
+    {
+      "epoch": 0.8447152238507828,
+      "grad_norm": 0.08810048550367355,
+      "learning_rate": 4.6702045107512075e-06,
+      "loss": 8.5213,
+      "step": 169150
+    },
+    {
+      "epoch": 0.8447651626757222,
+      "grad_norm": 0.0883641242980957,
+      "learning_rate": 4.668702595809658e-06,
+      "loss": 8.5451,
+      "step": 169160
+    },
+    {
+      "epoch": 0.8448151015006616,
+      "grad_norm": 0.09298266470432281,
+      "learning_rate": 4.667200680868107e-06,
+      "loss": 8.5362,
+      "step": 169170
+    },
+    {
+      "epoch": 0.8448650403256012,
+      "grad_norm": 0.08597780764102936,
+      "learning_rate": 4.665698765926556e-06,
+      "loss": 8.5315,
+      "step": 169180
+    },
+    {
+      "epoch": 0.8449149791505406,
+      "grad_norm": 0.09305087476968765,
+      "learning_rate": 4.664196850985006e-06,
+      "loss": 8.5321,
+      "step": 169190
+    },
+    {
+      "epoch": 0.84496491797548,
+      "grad_norm": 0.09523943066596985,
+      "learning_rate": 4.662694936043456e-06,
+      "loss": 8.5278,
+      "step": 169200
+    },
+    {
+      "epoch": 0.8450148568004194,
+      "grad_norm": 0.08965837210416794,
+      "learning_rate": 4.661193021101905e-06,
+      "loss": 8.5233,
+      "step": 169210
+    },
+    {
+      "epoch": 0.845064795625359,
+      "grad_norm": 0.08698379248380661,
+      "learning_rate": 4.6596911061603545e-06,
+      "loss": 8.5409,
+      "step": 169220
+    },
+    {
+      "epoch": 0.8451147344502984,
+      "grad_norm": 0.0960690900683403,
+      "learning_rate": 4.658189191218804e-06,
+      "loss": 8.5348,
+      "step": 169230
+    },
+    {
+      "epoch": 0.8451646732752378,
+      "grad_norm": 0.09241048991680145,
+      "learning_rate": 4.656687276277253e-06,
+      "loss": 8.5386,
+      "step": 169240
+    },
+    {
+      "epoch": 0.8452146121001772,
+      "grad_norm": 0.09009848535060883,
+      "learning_rate": 4.655185361335703e-06,
+      "loss": 8.5341,
+      "step": 169250
+    },
+    {
+      "epoch": 0.8452645509251168,
+      "grad_norm": 0.0927545353770256,
+      "learning_rate": 4.653683446394153e-06,
+      "loss": 8.5328,
+      "step": 169260
+    },
+    {
+      "epoch": 0.8453144897500562,
+      "grad_norm": 0.09297183901071548,
+      "learning_rate": 4.652181531452602e-06,
+      "loss": 8.5347,
+      "step": 169270
+    },
+    {
+      "epoch": 0.8453644285749956,
+      "grad_norm": 0.0915551409125328,
+      "learning_rate": 4.650679616511052e-06,
+      "loss": 8.5317,
+      "step": 169280
+    },
+    {
+      "epoch": 0.845414367399935,
+      "grad_norm": 0.0902283638715744,
+      "learning_rate": 4.649177701569501e-06,
+      "loss": 8.5346,
+      "step": 169290
+    },
+    {
+      "epoch": 0.8454643062248746,
+      "grad_norm": 0.0934210941195488,
+      "learning_rate": 4.647675786627951e-06,
+      "loss": 8.5406,
+      "step": 169300
+    },
+    {
+      "epoch": 0.845514245049814,
+      "grad_norm": 0.09512994438409805,
+      "learning_rate": 4.6461738716864e-06,
+      "loss": 8.5249,
+      "step": 169310
+    },
+    {
+      "epoch": 0.8455641838747534,
+      "grad_norm": 0.08952774852514267,
+      "learning_rate": 4.6446719567448495e-06,
+      "loss": 8.5413,
+      "step": 169320
+    },
+    {
+      "epoch": 0.8456141226996928,
+      "grad_norm": 0.09026441723108292,
+      "learning_rate": 4.6431700418033e-06,
+      "loss": 8.5283,
+      "step": 169330
+    },
+    {
+      "epoch": 0.8456640615246324,
+      "grad_norm": 0.09348812699317932,
+      "learning_rate": 4.641668126861749e-06,
+      "loss": 8.5296,
+      "step": 169340
+    },
+    {
+      "epoch": 0.8457140003495718,
+      "grad_norm": 0.09530653804540634,
+      "learning_rate": 4.640166211920198e-06,
+      "loss": 8.5413,
+      "step": 169350
+    },
+    {
+      "epoch": 0.8457639391745112,
+      "grad_norm": 0.0886545479297638,
+      "learning_rate": 4.6386642969786486e-06,
+      "loss": 8.5314,
+      "step": 169360
+    },
+    {
+      "epoch": 0.8458138779994506,
+      "grad_norm": 0.08922586590051651,
+      "learning_rate": 4.637162382037097e-06,
+      "loss": 8.5231,
+      "step": 169370
+    },
+    {
+      "epoch": 0.8458638168243902,
+      "grad_norm": 0.09260469675064087,
+      "learning_rate": 4.635660467095547e-06,
+      "loss": 8.5274,
+      "step": 169380
+    },
+    {
+      "epoch": 0.8459137556493296,
+      "grad_norm": 0.0915549173951149,
+      "learning_rate": 4.6341585521539965e-06,
+      "loss": 8.5405,
+      "step": 169390
+    },
+    {
+      "epoch": 0.845963694474269,
+      "grad_norm": 0.09589400887489319,
+      "learning_rate": 4.632656637212446e-06,
+      "loss": 8.5489,
+      "step": 169400
+    },
+    {
+      "epoch": 0.8460136332992084,
+      "grad_norm": 0.09545934200286865,
+      "learning_rate": 4.631154722270896e-06,
+      "loss": 8.5183,
+      "step": 169410
+    },
+    {
+      "epoch": 0.846063572124148,
+      "grad_norm": 0.09627781063318253,
+      "learning_rate": 4.6296528073293445e-06,
+      "loss": 8.5336,
+      "step": 169420
+    },
+    {
+      "epoch": 0.8461135109490874,
+      "grad_norm": 0.08923333883285522,
+      "learning_rate": 4.628150892387795e-06,
+      "loss": 8.5482,
+      "step": 169430
+    },
+    {
+      "epoch": 0.8461634497740268,
+      "grad_norm": 0.0943409726023674,
+      "learning_rate": 4.626648977446245e-06,
+      "loss": 8.5379,
+      "step": 169440
+    },
+    {
+      "epoch": 0.8462133885989662,
+      "grad_norm": 0.09771402925252914,
+      "learning_rate": 4.625147062504693e-06,
+      "loss": 8.5228,
+      "step": 169450
+    },
+    {
+      "epoch": 0.8462633274239058,
+      "grad_norm": 0.09028983116149902,
+      "learning_rate": 4.6236451475631436e-06,
+      "loss": 8.5318,
+      "step": 169460
+    },
+    {
+      "epoch": 0.8463132662488452,
+      "grad_norm": 0.09272675216197968,
+      "learning_rate": 4.622143232621592e-06,
+      "loss": 8.5224,
+      "step": 169470
+    },
+    {
+      "epoch": 0.8463632050737846,
+      "grad_norm": 0.09558822959661484,
+      "learning_rate": 4.620641317680042e-06,
+      "loss": 8.5264,
+      "step": 169480
+    },
+    {
+      "epoch": 0.846413143898724,
+      "grad_norm": 0.09761396795511246,
+      "learning_rate": 4.619139402738492e-06,
+      "loss": 8.5319,
+      "step": 169490
+    },
+    {
+      "epoch": 0.8464630827236636,
+      "grad_norm": 0.08748037368059158,
+      "learning_rate": 4.617637487796941e-06,
+      "loss": 8.543,
+      "step": 169500
+    },
+    {
+      "epoch": 0.846513021548603,
+      "grad_norm": 0.0940655842423439,
+      "learning_rate": 4.616135572855391e-06,
+      "loss": 8.5201,
+      "step": 169510
+    },
+    {
+      "epoch": 0.8465629603735424,
+      "grad_norm": 0.09007947146892548,
+      "learning_rate": 4.61463365791384e-06,
+      "loss": 8.5319,
+      "step": 169520
+    },
+    {
+      "epoch": 0.8466128991984818,
+      "grad_norm": 0.09856663644313812,
+      "learning_rate": 4.61313174297229e-06,
+      "loss": 8.5355,
+      "step": 169530
+    },
+    {
+      "epoch": 0.8466628380234213,
+      "grad_norm": 0.09389778226613998,
+      "learning_rate": 4.61162982803074e-06,
+      "loss": 8.5349,
+      "step": 169540
+    },
+    {
+      "epoch": 0.8467127768483608,
+      "grad_norm": 0.08669053763151169,
+      "learning_rate": 4.610127913089188e-06,
+      "loss": 8.5311,
+      "step": 169550
+    },
+    {
+      "epoch": 0.8467627156733002,
+      "grad_norm": 0.08856794238090515,
+      "learning_rate": 4.6086259981476386e-06,
+      "loss": 8.5275,
+      "step": 169560
+    },
+    {
+      "epoch": 0.8468126544982396,
+      "grad_norm": 0.09139453619718552,
+      "learning_rate": 4.607124083206088e-06,
+      "loss": 8.526,
+      "step": 169570
+    },
+    {
+      "epoch": 0.8468625933231791,
+      "grad_norm": 0.09540039300918579,
+      "learning_rate": 4.605622168264537e-06,
+      "loss": 8.5266,
+      "step": 169580
+    },
+    {
+      "epoch": 0.8469125321481186,
+      "grad_norm": 0.09145918488502502,
+      "learning_rate": 4.604120253322987e-06,
+      "loss": 8.5374,
+      "step": 169590
+    },
+    {
+      "epoch": 0.846962470973058,
+      "grad_norm": 0.08852418512105942,
+      "learning_rate": 4.602618338381437e-06,
+      "loss": 8.5433,
+      "step": 169600
+    },
+    {
+      "epoch": 0.8470124097979974,
+      "grad_norm": 0.09328535944223404,
+      "learning_rate": 4.601116423439886e-06,
+      "loss": 8.5504,
+      "step": 169610
+    },
+    {
+      "epoch": 0.847062348622937,
+      "grad_norm": 0.09695547074079514,
+      "learning_rate": 4.599614508498335e-06,
+      "loss": 8.5343,
+      "step": 169620
+    },
+    {
+      "epoch": 0.8471122874478764,
+      "grad_norm": 0.09098362177610397,
+      "learning_rate": 4.598112593556785e-06,
+      "loss": 8.5452,
+      "step": 169630
+    },
+    {
+      "epoch": 0.8471622262728158,
+      "grad_norm": 0.0890357717871666,
+      "learning_rate": 4.596610678615235e-06,
+      "loss": 8.5353,
+      "step": 169640
+    },
+    {
+      "epoch": 0.8472121650977552,
+      "grad_norm": 0.09094347059726715,
+      "learning_rate": 4.595108763673684e-06,
+      "loss": 8.5324,
+      "step": 169650
+    },
+    {
+      "epoch": 0.8472621039226947,
+      "grad_norm": 0.09218890964984894,
+      "learning_rate": 4.5936068487321336e-06,
+      "loss": 8.5394,
+      "step": 169660
+    },
+    {
+      "epoch": 0.8473120427476342,
+      "grad_norm": 0.09381541609764099,
+      "learning_rate": 4.592104933790583e-06,
+      "loss": 8.5285,
+      "step": 169670
+    },
+    {
+      "epoch": 0.8473619815725736,
+      "grad_norm": 0.09734323620796204,
+      "learning_rate": 4.590603018849032e-06,
+      "loss": 8.5248,
+      "step": 169680
+    },
+    {
+      "epoch": 0.847411920397513,
+      "grad_norm": 0.09610073268413544,
+      "learning_rate": 4.589101103907482e-06,
+      "loss": 8.526,
+      "step": 169690
+    },
+    {
+      "epoch": 0.8474618592224525,
+      "grad_norm": 0.08428271114826202,
+      "learning_rate": 4.587599188965932e-06,
+      "loss": 8.5368,
+      "step": 169700
+    },
+    {
+      "epoch": 0.847511798047392,
+      "grad_norm": 0.09294697642326355,
+      "learning_rate": 4.586097274024381e-06,
+      "loss": 8.537,
+      "step": 169710
+    },
+    {
+      "epoch": 0.8475617368723314,
+      "grad_norm": 0.08724416792392731,
+      "learning_rate": 4.58459535908283e-06,
+      "loss": 8.5352,
+      "step": 169720
+    },
+    {
+      "epoch": 0.8476116756972708,
+      "grad_norm": 0.08911699056625366,
+      "learning_rate": 4.583093444141281e-06,
+      "loss": 8.5249,
+      "step": 169730
+    },
+    {
+      "epoch": 0.8476616145222103,
+      "grad_norm": 0.08514885604381561,
+      "learning_rate": 4.58159152919973e-06,
+      "loss": 8.5576,
+      "step": 169740
+    },
+    {
+      "epoch": 0.8477115533471498,
+      "grad_norm": 0.08839089423418045,
+      "learning_rate": 4.580089614258179e-06,
+      "loss": 8.5511,
+      "step": 169750
+    },
+    {
+      "epoch": 0.8477614921720892,
+      "grad_norm": 0.0928574949502945,
+      "learning_rate": 4.578587699316629e-06,
+      "loss": 8.5221,
+      "step": 169760
+    },
+    {
+      "epoch": 0.8478114309970286,
+      "grad_norm": 0.09458362311124802,
+      "learning_rate": 4.577085784375078e-06,
+      "loss": 8.5231,
+      "step": 169770
+    },
+    {
+      "epoch": 0.8478613698219681,
+      "grad_norm": 0.09411700069904327,
+      "learning_rate": 4.575583869433528e-06,
+      "loss": 8.5104,
+      "step": 169780
+    },
+    {
+      "epoch": 0.8479113086469076,
+      "grad_norm": 0.09059763699769974,
+      "learning_rate": 4.574081954491977e-06,
+      "loss": 8.5246,
+      "step": 169790
+    },
+    {
+      "epoch": 0.847961247471847,
+      "grad_norm": 0.09194177389144897,
+      "learning_rate": 4.572580039550427e-06,
+      "loss": 8.5261,
+      "step": 169800
+    },
+    {
+      "epoch": 0.8480111862967864,
+      "grad_norm": 0.08624966442584991,
+      "learning_rate": 4.571078124608877e-06,
+      "loss": 8.517,
+      "step": 169810
+    },
+    {
+      "epoch": 0.8480611251217259,
+      "grad_norm": 0.09074903279542923,
+      "learning_rate": 4.569576209667325e-06,
+      "loss": 8.5229,
+      "step": 169820
+    },
+    {
+      "epoch": 0.8481110639466654,
+      "grad_norm": 0.09367188811302185,
+      "learning_rate": 4.568074294725776e-06,
+      "loss": 8.5213,
+      "step": 169830
+    },
+    {
+      "epoch": 0.8481610027716048,
+      "grad_norm": 0.09176365286111832,
+      "learning_rate": 4.566572379784225e-06,
+      "loss": 8.5176,
+      "step": 169840
+    },
+    {
+      "epoch": 0.8482109415965442,
+      "grad_norm": 0.09426657855510712,
+      "learning_rate": 4.565070464842674e-06,
+      "loss": 8.5222,
+      "step": 169850
+    },
+    {
+      "epoch": 0.8482608804214837,
+      "grad_norm": 0.09585726261138916,
+      "learning_rate": 4.5635685499011244e-06,
+      "loss": 8.5333,
+      "step": 169860
+    },
+    {
+      "epoch": 0.8483108192464232,
+      "grad_norm": 0.08798620849847794,
+      "learning_rate": 4.562066634959573e-06,
+      "loss": 8.5385,
+      "step": 169870
+    },
+    {
+      "epoch": 0.8483607580713626,
+      "grad_norm": 0.09025800973176956,
+      "learning_rate": 4.560564720018023e-06,
+      "loss": 8.5288,
+      "step": 169880
+    },
+    {
+      "epoch": 0.848410696896302,
+      "grad_norm": 0.09170225262641907,
+      "learning_rate": 4.559062805076473e-06,
+      "loss": 8.5304,
+      "step": 169890
+    },
+    {
+      "epoch": 0.8484606357212415,
+      "grad_norm": 0.08960150182247162,
+      "learning_rate": 4.557560890134922e-06,
+      "loss": 8.5373,
+      "step": 169900
+    },
+    {
+      "epoch": 0.848510574546181,
+      "grad_norm": 0.08712244778871536,
+      "learning_rate": 4.556058975193372e-06,
+      "loss": 8.5318,
+      "step": 169910
+    },
+    {
+      "epoch": 0.8485605133711204,
+      "grad_norm": 0.09316559135913849,
+      "learning_rate": 4.5545570602518204e-06,
+      "loss": 8.5409,
+      "step": 169920
+    },
+    {
+      "epoch": 0.8486104521960598,
+      "grad_norm": 0.0908828005194664,
+      "learning_rate": 4.553055145310271e-06,
+      "loss": 8.5518,
+      "step": 169930
+    },
+    {
+      "epoch": 0.8486603910209993,
+      "grad_norm": 0.09130831062793732,
+      "learning_rate": 4.551553230368721e-06,
+      "loss": 8.5231,
+      "step": 169940
+    },
+    {
+      "epoch": 0.8487103298459387,
+      "grad_norm": 0.0959784984588623,
+      "learning_rate": 4.550051315427169e-06,
+      "loss": 8.542,
+      "step": 169950
+    },
+    {
+      "epoch": 0.8487602686708782,
+      "grad_norm": 0.09128502756357193,
+      "learning_rate": 4.5485494004856194e-06,
+      "loss": 8.5307,
+      "step": 169960
+    },
+    {
+      "epoch": 0.8488102074958176,
+      "grad_norm": 0.10153272747993469,
+      "learning_rate": 4.547047485544069e-06,
+      "loss": 8.5301,
+      "step": 169970
+    },
+    {
+      "epoch": 0.8488601463207571,
+      "grad_norm": 0.09342774748802185,
+      "learning_rate": 4.545545570602518e-06,
+      "loss": 8.5356,
+      "step": 169980
+    },
+    {
+      "epoch": 0.8489100851456965,
+      "grad_norm": 0.0933808907866478,
+      "learning_rate": 4.544043655660968e-06,
+      "loss": 8.5405,
+      "step": 169990
+    },
+    {
+      "epoch": 0.848960023970636,
+      "grad_norm": 0.09019854664802551,
+      "learning_rate": 4.542541740719417e-06,
+      "loss": 8.5424,
+      "step": 170000
+    },
+    {
+      "epoch": 0.8490099627955754,
+      "grad_norm": 0.08802162855863571,
+      "learning_rate": 4.541039825777867e-06,
+      "loss": 8.532,
+      "step": 170010
+    },
+    {
+      "epoch": 0.8490599016205149,
+      "grad_norm": 0.09536787867546082,
+      "learning_rate": 4.539537910836316e-06,
+      "loss": 8.5354,
+      "step": 170020
+    },
+    {
+      "epoch": 0.8491098404454543,
+      "grad_norm": 0.0905720591545105,
+      "learning_rate": 4.538035995894766e-06,
+      "loss": 8.5336,
+      "step": 170030
+    },
+    {
+      "epoch": 0.8491597792703938,
+      "grad_norm": 0.09646139293909073,
+      "learning_rate": 4.536534080953216e-06,
+      "loss": 8.5237,
+      "step": 170040
+    },
+    {
+      "epoch": 0.8492097180953332,
+      "grad_norm": 0.09050284326076508,
+      "learning_rate": 4.535032166011665e-06,
+      "loss": 8.5426,
+      "step": 170050
+    },
+    {
+      "epoch": 0.8492596569202727,
+      "grad_norm": 0.09438873827457428,
+      "learning_rate": 4.5335302510701145e-06,
+      "loss": 8.5094,
+      "step": 170060
+    },
+    {
+      "epoch": 0.8493095957452121,
+      "grad_norm": 0.09040695428848267,
+      "learning_rate": 4.532028336128564e-06,
+      "loss": 8.5364,
+      "step": 170070
+    },
+    {
+      "epoch": 0.8493595345701516,
+      "grad_norm": 0.09196127206087112,
+      "learning_rate": 4.530526421187013e-06,
+      "loss": 8.5276,
+      "step": 170080
+    },
+    {
+      "epoch": 0.849409473395091,
+      "grad_norm": 0.09099120646715164,
+      "learning_rate": 4.529024506245463e-06,
+      "loss": 8.5473,
+      "step": 170090
+    },
+    {
+      "epoch": 0.8494594122200304,
+      "grad_norm": 0.08868347853422165,
+      "learning_rate": 4.527522591303913e-06,
+      "loss": 8.5644,
+      "step": 170100
+    },
+    {
+      "epoch": 0.8495093510449699,
+      "grad_norm": 0.09560364484786987,
+      "learning_rate": 4.526020676362362e-06,
+      "loss": 8.5208,
+      "step": 170110
+    },
+    {
+      "epoch": 0.8495592898699094,
+      "grad_norm": 0.09278056770563126,
+      "learning_rate": 4.524518761420811e-06,
+      "loss": 8.5264,
+      "step": 170120
+    },
+    {
+      "epoch": 0.8496092286948488,
+      "grad_norm": 0.09304804354906082,
+      "learning_rate": 4.5230168464792615e-06,
+      "loss": 8.5363,
+      "step": 170130
+    },
+    {
+      "epoch": 0.8496591675197882,
+      "grad_norm": 0.09038883447647095,
+      "learning_rate": 4.521514931537711e-06,
+      "loss": 8.521,
+      "step": 170140
+    },
+    {
+      "epoch": 0.8497091063447277,
+      "grad_norm": 0.09431783854961395,
+      "learning_rate": 4.52001301659616e-06,
+      "loss": 8.5134,
+      "step": 170150
+    },
+    {
+      "epoch": 0.8497590451696672,
+      "grad_norm": 0.09250378608703613,
+      "learning_rate": 4.5185111016546095e-06,
+      "loss": 8.5085,
+      "step": 170160
+    },
+    {
+      "epoch": 0.8498089839946066,
+      "grad_norm": 0.09243625402450562,
+      "learning_rate": 4.517009186713059e-06,
+      "loss": 8.5148,
+      "step": 170170
+    },
+    {
+      "epoch": 0.849858922819546,
+      "grad_norm": 0.09091030061244965,
+      "learning_rate": 4.515507271771509e-06,
+      "loss": 8.5367,
+      "step": 170180
+    },
+    {
+      "epoch": 0.8499088616444855,
+      "grad_norm": 0.09004100412130356,
+      "learning_rate": 4.514005356829958e-06,
+      "loss": 8.5418,
+      "step": 170190
+    },
+    {
+      "epoch": 0.849958800469425,
+      "grad_norm": 0.08924288302659988,
+      "learning_rate": 4.512503441888408e-06,
+      "loss": 8.5241,
+      "step": 170200
+    },
+    {
+      "epoch": 0.8500087392943644,
+      "grad_norm": 0.0939495712518692,
+      "learning_rate": 4.511001526946858e-06,
+      "loss": 8.5362,
+      "step": 170210
+    },
+    {
+      "epoch": 0.8500586781193038,
+      "grad_norm": 0.09469542652368546,
+      "learning_rate": 4.509499612005306e-06,
+      "loss": 8.5327,
+      "step": 170220
+    },
+    {
+      "epoch": 0.8501086169442433,
+      "grad_norm": 0.08831597119569778,
+      "learning_rate": 4.5079976970637565e-06,
+      "loss": 8.5328,
+      "step": 170230
+    },
+    {
+      "epoch": 0.8501585557691828,
+      "grad_norm": 0.0914112776517868,
+      "learning_rate": 4.506495782122206e-06,
+      "loss": 8.5414,
+      "step": 170240
+    },
+    {
+      "epoch": 0.8502084945941222,
+      "grad_norm": 0.0885150134563446,
+      "learning_rate": 4.504993867180655e-06,
+      "loss": 8.5305,
+      "step": 170250
+    },
+    {
+      "epoch": 0.8502584334190616,
+      "grad_norm": 0.08355226367712021,
+      "learning_rate": 4.503491952239105e-06,
+      "loss": 8.5322,
+      "step": 170260
+    },
+    {
+      "epoch": 0.8503083722440011,
+      "grad_norm": 0.08838168531656265,
+      "learning_rate": 4.501990037297554e-06,
+      "loss": 8.5281,
+      "step": 170270
+    },
+    {
+      "epoch": 0.8503583110689406,
+      "grad_norm": 0.0902080163359642,
+      "learning_rate": 4.500488122356004e-06,
+      "loss": 8.5394,
+      "step": 170280
+    },
+    {
+      "epoch": 0.85040824989388,
+      "grad_norm": 0.09007387608289719,
+      "learning_rate": 4.498986207414454e-06,
+      "loss": 8.5112,
+      "step": 170290
+    },
+    {
+      "epoch": 0.8504581887188194,
+      "grad_norm": 0.08401839435100555,
+      "learning_rate": 4.497484292472903e-06,
+      "loss": 8.533,
+      "step": 170300
+    },
+    {
+      "epoch": 0.8505081275437589,
+      "grad_norm": 0.0899522602558136,
+      "learning_rate": 4.495982377531353e-06,
+      "loss": 8.5388,
+      "step": 170310
+    },
+    {
+      "epoch": 0.8505580663686984,
+      "grad_norm": 0.09382461756467819,
+      "learning_rate": 4.494480462589801e-06,
+      "loss": 8.5133,
+      "step": 170320
+    },
+    {
+      "epoch": 0.8506080051936378,
+      "grad_norm": 0.09093036502599716,
+      "learning_rate": 4.4929785476482515e-06,
+      "loss": 8.5497,
+      "step": 170330
+    },
+    {
+      "epoch": 0.8506579440185772,
+      "grad_norm": 0.08482268452644348,
+      "learning_rate": 4.491476632706702e-06,
+      "loss": 8.5193,
+      "step": 170340
+    },
+    {
+      "epoch": 0.8507078828435167,
+      "grad_norm": 0.09074404090642929,
+      "learning_rate": 4.48997471776515e-06,
+      "loss": 8.5359,
+      "step": 170350
+    },
+    {
+      "epoch": 0.8507578216684561,
+      "grad_norm": 0.09321633726358414,
+      "learning_rate": 4.4884728028236e-06,
+      "loss": 8.5268,
+      "step": 170360
+    },
+    {
+      "epoch": 0.8508077604933956,
+      "grad_norm": 0.09202466160058975,
+      "learning_rate": 4.48697088788205e-06,
+      "loss": 8.5414,
+      "step": 170370
+    },
+    {
+      "epoch": 0.850857699318335,
+      "grad_norm": 0.08913063257932663,
+      "learning_rate": 4.485468972940499e-06,
+      "loss": 8.5411,
+      "step": 170380
+    },
+    {
+      "epoch": 0.8509076381432745,
+      "grad_norm": 0.09542038291692734,
+      "learning_rate": 4.483967057998949e-06,
+      "loss": 8.5335,
+      "step": 170390
+    },
+    {
+      "epoch": 0.850957576968214,
+      "grad_norm": 0.09215617924928665,
+      "learning_rate": 4.482465143057398e-06,
+      "loss": 8.5483,
+      "step": 170400
+    },
+    {
+      "epoch": 0.8510075157931534,
+      "grad_norm": 0.09391964226961136,
+      "learning_rate": 4.480963228115848e-06,
+      "loss": 8.5355,
+      "step": 170410
+    },
+    {
+      "epoch": 0.8510574546180928,
+      "grad_norm": 0.09085561335086823,
+      "learning_rate": 4.479461313174298e-06,
+      "loss": 8.5412,
+      "step": 170420
+    },
+    {
+      "epoch": 0.8511073934430323,
+      "grad_norm": 0.08789962530136108,
+      "learning_rate": 4.4779593982327465e-06,
+      "loss": 8.5273,
+      "step": 170430
+    },
+    {
+      "epoch": 0.8511573322679717,
+      "grad_norm": 0.08881513774394989,
+      "learning_rate": 4.476457483291197e-06,
+      "loss": 8.5312,
+      "step": 170440
+    },
+    {
+      "epoch": 0.8512072710929112,
+      "grad_norm": 0.0915045440196991,
+      "learning_rate": 4.474955568349646e-06,
+      "loss": 8.5231,
+      "step": 170450
+    },
+    {
+      "epoch": 0.8512572099178506,
+      "grad_norm": 0.09604188799858093,
+      "learning_rate": 4.473453653408095e-06,
+      "loss": 8.5366,
+      "step": 170460
+    },
+    {
+      "epoch": 0.8513071487427901,
+      "grad_norm": 0.09160809963941574,
+      "learning_rate": 4.4719517384665455e-06,
+      "loss": 8.5187,
+      "step": 170470
+    },
+    {
+      "epoch": 0.8513570875677295,
+      "grad_norm": 0.08846515417098999,
+      "learning_rate": 4.470449823524994e-06,
+      "loss": 8.5163,
+      "step": 170480
+    },
+    {
+      "epoch": 0.851407026392669,
+      "grad_norm": 0.09089051187038422,
+      "learning_rate": 4.468947908583444e-06,
+      "loss": 8.5191,
+      "step": 170490
+    },
+    {
+      "epoch": 0.8514569652176084,
+      "grad_norm": 0.09371612966060638,
+      "learning_rate": 4.4674459936418935e-06,
+      "loss": 8.5244,
+      "step": 170500
+    },
+    {
+      "epoch": 0.8515069040425479,
+      "grad_norm": 0.09188021719455719,
+      "learning_rate": 4.465944078700343e-06,
+      "loss": 8.5292,
+      "step": 170510
+    },
+    {
+      "epoch": 0.8515568428674873,
+      "grad_norm": 0.08881582319736481,
+      "learning_rate": 4.464442163758793e-06,
+      "loss": 8.5183,
+      "step": 170520
+    },
+    {
+      "epoch": 0.8516067816924268,
+      "grad_norm": 0.09312296658754349,
+      "learning_rate": 4.462940248817242e-06,
+      "loss": 8.5437,
+      "step": 170530
+    },
+    {
+      "epoch": 0.8516567205173662,
+      "grad_norm": 0.08998645842075348,
+      "learning_rate": 4.461438333875692e-06,
+      "loss": 8.5336,
+      "step": 170540
+    },
+    {
+      "epoch": 0.8517066593423057,
+      "grad_norm": 0.09388935565948486,
+      "learning_rate": 4.459936418934141e-06,
+      "loss": 8.5427,
+      "step": 170550
+    },
+    {
+      "epoch": 0.8517565981672451,
+      "grad_norm": 0.08806626498699188,
+      "learning_rate": 4.45843450399259e-06,
+      "loss": 8.5317,
+      "step": 170560
+    },
+    {
+      "epoch": 0.8518065369921846,
+      "grad_norm": 0.09021688997745514,
+      "learning_rate": 4.4569325890510405e-06,
+      "loss": 8.5494,
+      "step": 170570
+    },
+    {
+      "epoch": 0.851856475817124,
+      "grad_norm": 0.08856786042451859,
+      "learning_rate": 4.45543067410949e-06,
+      "loss": 8.5258,
+      "step": 170580
+    },
+    {
+      "epoch": 0.8519064146420635,
+      "grad_norm": 0.09338122606277466,
+      "learning_rate": 4.453928759167939e-06,
+      "loss": 8.5315,
+      "step": 170590
+    },
+    {
+      "epoch": 0.8519563534670029,
+      "grad_norm": 0.09389621019363403,
+      "learning_rate": 4.4524268442263885e-06,
+      "loss": 8.5314,
+      "step": 170600
+    },
+    {
+      "epoch": 0.8520062922919424,
+      "grad_norm": 0.08809863030910492,
+      "learning_rate": 4.450924929284839e-06,
+      "loss": 8.5296,
+      "step": 170610
+    },
+    {
+      "epoch": 0.8520562311168818,
+      "grad_norm": 0.08649566769599915,
+      "learning_rate": 4.449423014343288e-06,
+      "loss": 8.5521,
+      "step": 170620
+    },
+    {
+      "epoch": 0.8521061699418213,
+      "grad_norm": 0.09101077169179916,
+      "learning_rate": 4.447921099401737e-06,
+      "loss": 8.5458,
+      "step": 170630
+    },
+    {
+      "epoch": 0.8521561087667607,
+      "grad_norm": 0.0848645567893982,
+      "learning_rate": 4.446419184460187e-06,
+      "loss": 8.5266,
+      "step": 170640
+    },
+    {
+      "epoch": 0.8522060475917002,
+      "grad_norm": 0.0918324664235115,
+      "learning_rate": 4.444917269518636e-06,
+      "loss": 8.517,
+      "step": 170650
+    },
+    {
+      "epoch": 0.8522559864166396,
+      "grad_norm": 0.08957000821828842,
+      "learning_rate": 4.443415354577086e-06,
+      "loss": 8.5481,
+      "step": 170660
+    },
+    {
+      "epoch": 0.8523059252415791,
+      "grad_norm": 0.09019914269447327,
+      "learning_rate": 4.4419134396355355e-06,
+      "loss": 8.5189,
+      "step": 170670
+    },
+    {
+      "epoch": 0.8523558640665185,
+      "grad_norm": 0.09360361099243164,
+      "learning_rate": 4.440411524693985e-06,
+      "loss": 8.5331,
+      "step": 170680
+    },
+    {
+      "epoch": 0.852405802891458,
+      "grad_norm": 0.09016086161136627,
+      "learning_rate": 4.438909609752435e-06,
+      "loss": 8.5151,
+      "step": 170690
+    },
+    {
+      "epoch": 0.8524557417163974,
+      "grad_norm": 0.09195458889007568,
+      "learning_rate": 4.4374076948108835e-06,
+      "loss": 8.5457,
+      "step": 170700
+    },
+    {
+      "epoch": 0.8525056805413369,
+      "grad_norm": 0.0889841690659523,
+      "learning_rate": 4.435905779869334e-06,
+      "loss": 8.5435,
+      "step": 170710
+    },
+    {
+      "epoch": 0.8525556193662763,
+      "grad_norm": 0.09094245731830597,
+      "learning_rate": 4.434403864927783e-06,
+      "loss": 8.5134,
+      "step": 170720
+    },
+    {
+      "epoch": 0.8526055581912158,
+      "grad_norm": 0.09037609398365021,
+      "learning_rate": 4.432901949986232e-06,
+      "loss": 8.5313,
+      "step": 170730
+    },
+    {
+      "epoch": 0.8526554970161552,
+      "grad_norm": 0.0911756157875061,
+      "learning_rate": 4.4314000350446825e-06,
+      "loss": 8.5236,
+      "step": 170740
+    },
+    {
+      "epoch": 0.8527054358410947,
+      "grad_norm": 0.09256346523761749,
+      "learning_rate": 4.429898120103131e-06,
+      "loss": 8.5366,
+      "step": 170750
+    },
+    {
+      "epoch": 0.8527553746660341,
+      "grad_norm": 0.09345842152833939,
+      "learning_rate": 4.428396205161581e-06,
+      "loss": 8.5308,
+      "step": 170760
+    },
+    {
+      "epoch": 0.8528053134909735,
+      "grad_norm": 0.10060130059719086,
+      "learning_rate": 4.426894290220031e-06,
+      "loss": 8.5234,
+      "step": 170770
+    },
+    {
+      "epoch": 0.852855252315913,
+      "grad_norm": 0.09197411686182022,
+      "learning_rate": 4.42539237527848e-06,
+      "loss": 8.558,
+      "step": 170780
+    },
+    {
+      "epoch": 0.8529051911408525,
+      "grad_norm": 0.09203599393367767,
+      "learning_rate": 4.42389046033693e-06,
+      "loss": 8.5315,
+      "step": 170790
+    },
+    {
+      "epoch": 0.8529551299657919,
+      "grad_norm": 0.09319550544023514,
+      "learning_rate": 4.4223885453953785e-06,
+      "loss": 8.5225,
+      "step": 170800
+    },
+    {
+      "epoch": 0.8530050687907313,
+      "grad_norm": 0.08547905087471008,
+      "learning_rate": 4.420886630453829e-06,
+      "loss": 8.5404,
+      "step": 170810
+    },
+    {
+      "epoch": 0.8530550076156708,
+      "grad_norm": 0.08869150280952454,
+      "learning_rate": 4.419384715512279e-06,
+      "loss": 8.5281,
+      "step": 170820
+    },
+    {
+      "epoch": 0.8531049464406103,
+      "grad_norm": 0.08847400546073914,
+      "learning_rate": 4.417882800570727e-06,
+      "loss": 8.544,
+      "step": 170830
+    },
+    {
+      "epoch": 0.8531548852655497,
+      "grad_norm": 0.09268223494291306,
+      "learning_rate": 4.4163808856291775e-06,
+      "loss": 8.5184,
+      "step": 170840
+    },
+    {
+      "epoch": 0.8532048240904891,
+      "grad_norm": 0.09480337798595428,
+      "learning_rate": 4.414878970687627e-06,
+      "loss": 8.5208,
+      "step": 170850
+    },
+    {
+      "epoch": 0.8532547629154286,
+      "grad_norm": 0.09097526222467422,
+      "learning_rate": 4.413377055746076e-06,
+      "loss": 8.5258,
+      "step": 170860
+    },
+    {
+      "epoch": 0.8533047017403681,
+      "grad_norm": 0.0875224694609642,
+      "learning_rate": 4.411875140804526e-06,
+      "loss": 8.5318,
+      "step": 170870
+    },
+    {
+      "epoch": 0.8533546405653075,
+      "grad_norm": 0.08786869794130325,
+      "learning_rate": 4.410373225862975e-06,
+      "loss": 8.5334,
+      "step": 170880
+    },
+    {
+      "epoch": 0.8534045793902469,
+      "grad_norm": 0.08923707902431488,
+      "learning_rate": 4.408871310921425e-06,
+      "loss": 8.535,
+      "step": 170890
+    },
+    {
+      "epoch": 0.8534545182151864,
+      "grad_norm": 0.0923088863492012,
+      "learning_rate": 4.407369395979874e-06,
+      "loss": 8.5256,
+      "step": 170900
+    },
+    {
+      "epoch": 0.8535044570401259,
+      "grad_norm": 0.0889919325709343,
+      "learning_rate": 4.405867481038324e-06,
+      "loss": 8.5382,
+      "step": 170910
+    },
+    {
+      "epoch": 0.8535543958650653,
+      "grad_norm": 0.09233865886926651,
+      "learning_rate": 4.404365566096774e-06,
+      "loss": 8.5243,
+      "step": 170920
+    },
+    {
+      "epoch": 0.8536043346900047,
+      "grad_norm": 0.09140230715274811,
+      "learning_rate": 4.402863651155223e-06,
+      "loss": 8.5295,
+      "step": 170930
+    },
+    {
+      "epoch": 0.8536542735149442,
+      "grad_norm": 0.08698879182338715,
+      "learning_rate": 4.4013617362136725e-06,
+      "loss": 8.5412,
+      "step": 170940
+    },
+    {
+      "epoch": 0.8537042123398837,
+      "grad_norm": 0.09249941259622574,
+      "learning_rate": 4.399859821272122e-06,
+      "loss": 8.533,
+      "step": 170950
+    },
+    {
+      "epoch": 0.8537541511648231,
+      "grad_norm": 0.0933549553155899,
+      "learning_rate": 4.398357906330571e-06,
+      "loss": 8.5375,
+      "step": 170960
+    },
+    {
+      "epoch": 0.8538040899897625,
+      "grad_norm": 0.09042268991470337,
+      "learning_rate": 4.396855991389021e-06,
+      "loss": 8.5378,
+      "step": 170970
+    },
+    {
+      "epoch": 0.853854028814702,
+      "grad_norm": 0.09002401679754257,
+      "learning_rate": 4.395354076447471e-06,
+      "loss": 8.5339,
+      "step": 170980
+    },
+    {
+      "epoch": 0.8539039676396415,
+      "grad_norm": 0.09069207310676575,
+      "learning_rate": 4.39385216150592e-06,
+      "loss": 8.5289,
+      "step": 170990
+    },
+    {
+      "epoch": 0.8539539064645809,
+      "grad_norm": 0.09139145910739899,
+      "learning_rate": 4.392350246564369e-06,
+      "loss": 8.5352,
+      "step": 171000
+    },
+    {
+      "epoch": 0.8540038452895203,
+      "grad_norm": 0.0929713174700737,
+      "learning_rate": 4.3908483316228196e-06,
+      "loss": 8.5344,
+      "step": 171010
+    },
+    {
+      "epoch": 0.8540537841144598,
+      "grad_norm": 0.09418080747127533,
+      "learning_rate": 4.389346416681269e-06,
+      "loss": 8.5188,
+      "step": 171020
+    },
+    {
+      "epoch": 0.8541037229393993,
+      "grad_norm": 0.08637245744466782,
+      "learning_rate": 4.387844501739718e-06,
+      "loss": 8.5291,
+      "step": 171030
+    },
+    {
+      "epoch": 0.8541536617643387,
+      "grad_norm": 0.08483462035655975,
+      "learning_rate": 4.3863425867981676e-06,
+      "loss": 8.5162,
+      "step": 171040
+    },
+    {
+      "epoch": 0.8542036005892781,
+      "grad_norm": 0.08939448744058609,
+      "learning_rate": 4.384840671856617e-06,
+      "loss": 8.5235,
+      "step": 171050
+    },
+    {
+      "epoch": 0.8542535394142176,
+      "grad_norm": 0.09517613798379898,
+      "learning_rate": 4.383338756915067e-06,
+      "loss": 8.5371,
+      "step": 171060
+    },
+    {
+      "epoch": 0.8543034782391571,
+      "grad_norm": 0.09513095766305923,
+      "learning_rate": 4.381836841973516e-06,
+      "loss": 8.5406,
+      "step": 171070
+    },
+    {
+      "epoch": 0.8543534170640965,
+      "grad_norm": 0.08950556814670563,
+      "learning_rate": 4.380334927031966e-06,
+      "loss": 8.5353,
+      "step": 171080
+    },
+    {
+      "epoch": 0.8544033558890359,
+      "grad_norm": 0.09560289233922958,
+      "learning_rate": 4.378833012090416e-06,
+      "loss": 8.5242,
+      "step": 171090
+    },
+    {
+      "epoch": 0.8544532947139754,
+      "grad_norm": 0.0961078554391861,
+      "learning_rate": 4.377331097148864e-06,
+      "loss": 8.5337,
+      "step": 171100
+    },
+    {
+      "epoch": 0.8545032335389148,
+      "grad_norm": 0.0894903764128685,
+      "learning_rate": 4.3758291822073146e-06,
+      "loss": 8.5246,
+      "step": 171110
+    },
+    {
+      "epoch": 0.8545531723638543,
+      "grad_norm": 0.08703772723674774,
+      "learning_rate": 4.374327267265764e-06,
+      "loss": 8.5198,
+      "step": 171120
+    },
+    {
+      "epoch": 0.8546031111887937,
+      "grad_norm": 0.08781271427869797,
+      "learning_rate": 4.372825352324213e-06,
+      "loss": 8.5344,
+      "step": 171130
+    },
+    {
+      "epoch": 0.8546530500137332,
+      "grad_norm": 0.09302118420600891,
+      "learning_rate": 4.371323437382663e-06,
+      "loss": 8.5205,
+      "step": 171140
+    },
+    {
+      "epoch": 0.8547029888386726,
+      "grad_norm": 0.09112757444381714,
+      "learning_rate": 4.369821522441112e-06,
+      "loss": 8.5174,
+      "step": 171150
+    },
+    {
+      "epoch": 0.8547529276636121,
+      "grad_norm": 0.0904446691274643,
+      "learning_rate": 4.368319607499562e-06,
+      "loss": 8.5156,
+      "step": 171160
+    },
+    {
+      "epoch": 0.8548028664885515,
+      "grad_norm": 0.09799622744321823,
+      "learning_rate": 4.366817692558012e-06,
+      "loss": 8.538,
+      "step": 171170
+    },
+    {
+      "epoch": 0.854852805313491,
+      "grad_norm": 0.08375813066959381,
+      "learning_rate": 4.365315777616461e-06,
+      "loss": 8.5342,
+      "step": 171180
+    },
+    {
+      "epoch": 0.8549027441384304,
+      "grad_norm": 0.08957131952047348,
+      "learning_rate": 4.363813862674911e-06,
+      "loss": 8.5287,
+      "step": 171190
+    },
+    {
+      "epoch": 0.8549526829633699,
+      "grad_norm": 0.0968262180685997,
+      "learning_rate": 4.362311947733359e-06,
+      "loss": 8.5356,
+      "step": 171200
+    },
+    {
+      "epoch": 0.8550026217883093,
+      "grad_norm": 0.09194337576627731,
+      "learning_rate": 4.3608100327918096e-06,
+      "loss": 8.5205,
+      "step": 171210
+    },
+    {
+      "epoch": 0.8550525606132487,
+      "grad_norm": 0.09763994812965393,
+      "learning_rate": 4.35930811785026e-06,
+      "loss": 8.5341,
+      "step": 171220
+    },
+    {
+      "epoch": 0.8551024994381882,
+      "grad_norm": 0.09552284330129623,
+      "learning_rate": 4.357806202908708e-06,
+      "loss": 8.5164,
+      "step": 171230
+    },
+    {
+      "epoch": 0.8551524382631277,
+      "grad_norm": 0.09413187205791473,
+      "learning_rate": 4.356304287967158e-06,
+      "loss": 8.5397,
+      "step": 171240
+    },
+    {
+      "epoch": 0.8552023770880671,
+      "grad_norm": 0.09369870275259018,
+      "learning_rate": 4.354802373025608e-06,
+      "loss": 8.5478,
+      "step": 171250
+    },
+    {
+      "epoch": 0.8552523159130065,
+      "grad_norm": 0.08899905532598495,
+      "learning_rate": 4.353300458084057e-06,
+      "loss": 8.5269,
+      "step": 171260
+    },
+    {
+      "epoch": 0.855302254737946,
+      "grad_norm": 0.09390175342559814,
+      "learning_rate": 4.351798543142507e-06,
+      "loss": 8.5124,
+      "step": 171270
+    },
+    {
+      "epoch": 0.8553521935628855,
+      "grad_norm": 0.09079564362764359,
+      "learning_rate": 4.350296628200956e-06,
+      "loss": 8.5169,
+      "step": 171280
+    },
+    {
+      "epoch": 0.8554021323878249,
+      "grad_norm": 0.09164270758628845,
+      "learning_rate": 4.348794713259406e-06,
+      "loss": 8.5276,
+      "step": 171290
+    },
+    {
+      "epoch": 0.8554520712127643,
+      "grad_norm": 0.0873217061161995,
+      "learning_rate": 4.347292798317855e-06,
+      "loss": 8.5337,
+      "step": 171300
+    },
+    {
+      "epoch": 0.8555020100377038,
+      "grad_norm": 0.09027426689863205,
+      "learning_rate": 4.345790883376305e-06,
+      "loss": 8.5357,
+      "step": 171310
+    },
+    {
+      "epoch": 0.8555519488626433,
+      "grad_norm": 0.09546752274036407,
+      "learning_rate": 4.344288968434755e-06,
+      "loss": 8.5264,
+      "step": 171320
+    },
+    {
+      "epoch": 0.8556018876875827,
+      "grad_norm": 0.0898887887597084,
+      "learning_rate": 4.342787053493204e-06,
+      "loss": 8.5286,
+      "step": 171330
+    },
+    {
+      "epoch": 0.8556518265125221,
+      "grad_norm": 0.0918826311826706,
+      "learning_rate": 4.3412851385516534e-06,
+      "loss": 8.5244,
+      "step": 171340
+    },
+    {
+      "epoch": 0.8557017653374616,
+      "grad_norm": 0.09295768290758133,
+      "learning_rate": 4.339783223610103e-06,
+      "loss": 8.5359,
+      "step": 171350
+    },
+    {
+      "epoch": 0.8557517041624011,
+      "grad_norm": 0.08988602459430695,
+      "learning_rate": 4.338281308668552e-06,
+      "loss": 8.5364,
+      "step": 171360
+    },
+    {
+      "epoch": 0.8558016429873405,
+      "grad_norm": 0.0889463871717453,
+      "learning_rate": 4.336779393727002e-06,
+      "loss": 8.5293,
+      "step": 171370
+    },
+    {
+      "epoch": 0.8558515818122799,
+      "grad_norm": 0.09765569865703583,
+      "learning_rate": 4.335277478785452e-06,
+      "loss": 8.5084,
+      "step": 171380
+    },
+    {
+      "epoch": 0.8559015206372194,
+      "grad_norm": 0.08945247530937195,
+      "learning_rate": 4.333775563843901e-06,
+      "loss": 8.5243,
+      "step": 171390
+    },
+    {
+      "epoch": 0.8559514594621589,
+      "grad_norm": 0.08958765864372253,
+      "learning_rate": 4.33227364890235e-06,
+      "loss": 8.5287,
+      "step": 171400
+    },
+    {
+      "epoch": 0.8560013982870983,
+      "grad_norm": 0.09407828748226166,
+      "learning_rate": 4.3307717339608004e-06,
+      "loss": 8.521,
+      "step": 171410
+    },
+    {
+      "epoch": 0.8560513371120377,
+      "grad_norm": 0.09075424075126648,
+      "learning_rate": 4.32926981901925e-06,
+      "loss": 8.5308,
+      "step": 171420
+    },
+    {
+      "epoch": 0.8561012759369772,
+      "grad_norm": 0.09202433377504349,
+      "learning_rate": 4.327767904077699e-06,
+      "loss": 8.5345,
+      "step": 171430
+    },
+    {
+      "epoch": 0.8561512147619167,
+      "grad_norm": 0.09231635928153992,
+      "learning_rate": 4.3262659891361484e-06,
+      "loss": 8.5238,
+      "step": 171440
+    },
+    {
+      "epoch": 0.8562011535868561,
+      "grad_norm": 0.09123031049966812,
+      "learning_rate": 4.324764074194598e-06,
+      "loss": 8.5197,
+      "step": 171450
+    },
+    {
+      "epoch": 0.8562510924117955,
+      "grad_norm": 0.08414209634065628,
+      "learning_rate": 4.323262159253048e-06,
+      "loss": 8.5187,
+      "step": 171460
+    },
+    {
+      "epoch": 0.856301031236735,
+      "grad_norm": 0.09748626500368118,
+      "learning_rate": 4.321760244311497e-06,
+      "loss": 8.5257,
+      "step": 171470
+    },
+    {
+      "epoch": 0.8563509700616745,
+      "grad_norm": 0.08869002014398575,
+      "learning_rate": 4.320258329369947e-06,
+      "loss": 8.5425,
+      "step": 171480
+    },
+    {
+      "epoch": 0.8564009088866139,
+      "grad_norm": 0.09146328270435333,
+      "learning_rate": 4.318756414428397e-06,
+      "loss": 8.5481,
+      "step": 171490
+    },
+    {
+      "epoch": 0.8564508477115533,
+      "grad_norm": 0.0903269499540329,
+      "learning_rate": 4.317254499486846e-06,
+      "loss": 8.5161,
+      "step": 171500
+    },
+    {
+      "epoch": 0.8565007865364928,
+      "grad_norm": 0.09317917376756668,
+      "learning_rate": 4.3157525845452954e-06,
+      "loss": 8.5282,
+      "step": 171510
+    },
+    {
+      "epoch": 0.8565507253614323,
+      "grad_norm": 0.09211008250713348,
+      "learning_rate": 4.314250669603745e-06,
+      "loss": 8.5379,
+      "step": 171520
+    },
+    {
+      "epoch": 0.8566006641863717,
+      "grad_norm": 0.08755582571029663,
+      "learning_rate": 4.312748754662194e-06,
+      "loss": 8.5338,
+      "step": 171530
+    },
+    {
+      "epoch": 0.8566506030113111,
+      "grad_norm": 0.09266114234924316,
+      "learning_rate": 4.311246839720644e-06,
+      "loss": 8.5368,
+      "step": 171540
+    },
+    {
+      "epoch": 0.8567005418362506,
+      "grad_norm": 0.09207313507795334,
+      "learning_rate": 4.309744924779094e-06,
+      "loss": 8.5345,
+      "step": 171550
+    },
+    {
+      "epoch": 0.8567504806611901,
+      "grad_norm": 0.09798050671815872,
+      "learning_rate": 4.308243009837543e-06,
+      "loss": 8.5375,
+      "step": 171560
+    },
+    {
+      "epoch": 0.8568004194861295,
+      "grad_norm": 0.08942960947751999,
+      "learning_rate": 4.306741094895993e-06,
+      "loss": 8.542,
+      "step": 171570
+    },
+    {
+      "epoch": 0.8568503583110689,
+      "grad_norm": 0.0846295952796936,
+      "learning_rate": 4.305239179954442e-06,
+      "loss": 8.5393,
+      "step": 171580
+    },
+    {
+      "epoch": 0.8569002971360083,
+      "grad_norm": 0.09053222835063934,
+      "learning_rate": 4.303737265012892e-06,
+      "loss": 8.5335,
+      "step": 171590
+    },
+    {
+      "epoch": 0.8569502359609479,
+      "grad_norm": 0.09028905630111694,
+      "learning_rate": 4.302235350071341e-06,
+      "loss": 8.5206,
+      "step": 171600
+    },
+    {
+      "epoch": 0.8570001747858873,
+      "grad_norm": 0.09010299295186996,
+      "learning_rate": 4.3007334351297905e-06,
+      "loss": 8.5239,
+      "step": 171610
+    },
+    {
+      "epoch": 0.8570501136108267,
+      "grad_norm": 0.08910666406154633,
+      "learning_rate": 4.299231520188241e-06,
+      "loss": 8.5292,
+      "step": 171620
+    },
+    {
+      "epoch": 0.8571000524357661,
+      "grad_norm": 0.0874292179942131,
+      "learning_rate": 4.297729605246689e-06,
+      "loss": 8.5219,
+      "step": 171630
+    },
+    {
+      "epoch": 0.8571499912607057,
+      "grad_norm": 0.08846402913331985,
+      "learning_rate": 4.296227690305139e-06,
+      "loss": 8.5341,
+      "step": 171640
+    },
+    {
+      "epoch": 0.8571999300856451,
+      "grad_norm": 0.09597437083721161,
+      "learning_rate": 4.2947257753635895e-06,
+      "loss": 8.52,
+      "step": 171650
+    },
+    {
+      "epoch": 0.8572498689105845,
+      "grad_norm": 0.08995895087718964,
+      "learning_rate": 4.293223860422038e-06,
+      "loss": 8.5257,
+      "step": 171660
+    },
+    {
+      "epoch": 0.8572998077355239,
+      "grad_norm": 0.08784112334251404,
+      "learning_rate": 4.291721945480488e-06,
+      "loss": 8.5186,
+      "step": 171670
+    },
+    {
+      "epoch": 0.8573497465604635,
+      "grad_norm": 0.09828951209783554,
+      "learning_rate": 4.290220030538937e-06,
+      "loss": 8.5305,
+      "step": 171680
+    },
+    {
+      "epoch": 0.8573996853854029,
+      "grad_norm": 0.09064572304487228,
+      "learning_rate": 4.288718115597387e-06,
+      "loss": 8.5354,
+      "step": 171690
+    },
+    {
+      "epoch": 0.8574496242103423,
+      "grad_norm": 0.09304898232221603,
+      "learning_rate": 4.287216200655837e-06,
+      "loss": 8.5223,
+      "step": 171700
+    },
+    {
+      "epoch": 0.8574995630352817,
+      "grad_norm": 0.09134497493505478,
+      "learning_rate": 4.2857142857142855e-06,
+      "loss": 8.5243,
+      "step": 171710
+    },
+    {
+      "epoch": 0.8575495018602213,
+      "grad_norm": 0.0942414253950119,
+      "learning_rate": 4.284212370772736e-06,
+      "loss": 8.5371,
+      "step": 171720
+    },
+    {
+      "epoch": 0.8575994406851607,
+      "grad_norm": 0.10213427990674973,
+      "learning_rate": 4.282710455831185e-06,
+      "loss": 8.5147,
+      "step": 171730
+    },
+    {
+      "epoch": 0.8576493795101001,
+      "grad_norm": 0.09323253482580185,
+      "learning_rate": 4.281208540889634e-06,
+      "loss": 8.5178,
+      "step": 171740
+    },
+    {
+      "epoch": 0.8576993183350395,
+      "grad_norm": 0.09151145815849304,
+      "learning_rate": 4.2797066259480845e-06,
+      "loss": 8.5257,
+      "step": 171750
+    },
+    {
+      "epoch": 0.8577492571599791,
+      "grad_norm": 0.0880388393998146,
+      "learning_rate": 4.278204711006533e-06,
+      "loss": 8.5441,
+      "step": 171760
+    },
+    {
+      "epoch": 0.8577991959849185,
+      "grad_norm": 0.0961805209517479,
+      "learning_rate": 4.276702796064983e-06,
+      "loss": 8.5351,
+      "step": 171770
+    },
+    {
+      "epoch": 0.8578491348098579,
+      "grad_norm": 0.09380482137203217,
+      "learning_rate": 4.2752008811234325e-06,
+      "loss": 8.5219,
+      "step": 171780
+    },
+    {
+      "epoch": 0.8578990736347973,
+      "grad_norm": 0.08797604590654373,
+      "learning_rate": 4.273698966181882e-06,
+      "loss": 8.524,
+      "step": 171790
+    },
+    {
+      "epoch": 0.8579490124597369,
+      "grad_norm": 0.08902735263109207,
+      "learning_rate": 4.272197051240332e-06,
+      "loss": 8.5404,
+      "step": 171800
+    },
+    {
+      "epoch": 0.8579989512846763,
+      "grad_norm": 0.09067311882972717,
+      "learning_rate": 4.270695136298781e-06,
+      "loss": 8.5317,
+      "step": 171810
+    },
+    {
+      "epoch": 0.8580488901096157,
+      "grad_norm": 0.09175264835357666,
+      "learning_rate": 4.269193221357231e-06,
+      "loss": 8.5181,
+      "step": 171820
+    },
+    {
+      "epoch": 0.8580988289345551,
+      "grad_norm": 0.0938398614525795,
+      "learning_rate": 4.26769130641568e-06,
+      "loss": 8.5388,
+      "step": 171830
+    },
+    {
+      "epoch": 0.8581487677594947,
+      "grad_norm": 0.09269008040428162,
+      "learning_rate": 4.266189391474129e-06,
+      "loss": 8.5437,
+      "step": 171840
+    },
+    {
+      "epoch": 0.8581987065844341,
+      "grad_norm": 0.09139380604028702,
+      "learning_rate": 4.2646874765325795e-06,
+      "loss": 8.5397,
+      "step": 171850
+    },
+    {
+      "epoch": 0.8582486454093735,
+      "grad_norm": 0.09328168630599976,
+      "learning_rate": 4.263185561591029e-06,
+      "loss": 8.5047,
+      "step": 171860
+    },
+    {
+      "epoch": 0.8582985842343129,
+      "grad_norm": 0.0878739282488823,
+      "learning_rate": 4.261683646649478e-06,
+      "loss": 8.5314,
+      "step": 171870
+    },
+    {
+      "epoch": 0.8583485230592525,
+      "grad_norm": 0.0920386090874672,
+      "learning_rate": 4.2601817317079275e-06,
+      "loss": 8.5369,
+      "step": 171880
+    },
+    {
+      "epoch": 0.8583984618841919,
+      "grad_norm": 0.09493181109428406,
+      "learning_rate": 4.258679816766378e-06,
+      "loss": 8.5218,
+      "step": 171890
+    },
+    {
+      "epoch": 0.8584484007091313,
+      "grad_norm": 0.09276656806468964,
+      "learning_rate": 4.257177901824827e-06,
+      "loss": 8.5195,
+      "step": 171900
+    },
+    {
+      "epoch": 0.8584983395340707,
+      "grad_norm": 0.09002747386693954,
+      "learning_rate": 4.255675986883276e-06,
+      "loss": 8.5195,
+      "step": 171910
+    },
+    {
+      "epoch": 0.8585482783590103,
+      "grad_norm": 0.09539522230625153,
+      "learning_rate": 4.254174071941726e-06,
+      "loss": 8.5185,
+      "step": 171920
+    },
+    {
+      "epoch": 0.8585982171839497,
+      "grad_norm": 0.09918307512998581,
+      "learning_rate": 4.252672157000175e-06,
+      "loss": 8.5317,
+      "step": 171930
+    },
+    {
+      "epoch": 0.8586481560088891,
+      "grad_norm": 0.09323044866323471,
+      "learning_rate": 4.251170242058625e-06,
+      "loss": 8.5361,
+      "step": 171940
+    },
+    {
+      "epoch": 0.8586980948338285,
+      "grad_norm": 0.09227460622787476,
+      "learning_rate": 4.2496683271170745e-06,
+      "loss": 8.5207,
+      "step": 171950
+    },
+    {
+      "epoch": 0.8587480336587681,
+      "grad_norm": 0.09323973953723907,
+      "learning_rate": 4.248166412175524e-06,
+      "loss": 8.5152,
+      "step": 171960
+    },
+    {
+      "epoch": 0.8587979724837075,
+      "grad_norm": 0.0909019485116005,
+      "learning_rate": 4.246664497233974e-06,
+      "loss": 8.5242,
+      "step": 171970
+    },
+    {
+      "epoch": 0.8588479113086469,
+      "grad_norm": 0.08941137790679932,
+      "learning_rate": 4.2451625822924225e-06,
+      "loss": 8.5398,
+      "step": 171980
+    },
+    {
+      "epoch": 0.8588978501335863,
+      "grad_norm": 0.08684715628623962,
+      "learning_rate": 4.243660667350873e-06,
+      "loss": 8.5233,
+      "step": 171990
+    },
+    {
+      "epoch": 0.8589477889585259,
+      "grad_norm": 0.0915994718670845,
+      "learning_rate": 4.242158752409322e-06,
+      "loss": 8.5258,
+      "step": 172000
+    },
+    {
+      "epoch": 0.8589977277834653,
+      "grad_norm": 0.09104558825492859,
+      "learning_rate": 4.240656837467771e-06,
+      "loss": 8.534,
+      "step": 172010
+    },
+    {
+      "epoch": 0.8590476666084047,
+      "grad_norm": 0.09221166372299194,
+      "learning_rate": 4.2391549225262215e-06,
+      "loss": 8.5333,
+      "step": 172020
+    },
+    {
+      "epoch": 0.8590976054333441,
+      "grad_norm": 0.09822104871273041,
+      "learning_rate": 4.23765300758467e-06,
+      "loss": 8.52,
+      "step": 172030
+    },
+    {
+      "epoch": 0.8591475442582837,
+      "grad_norm": 0.08423840254545212,
+      "learning_rate": 4.23615109264312e-06,
+      "loss": 8.5458,
+      "step": 172040
+    },
+    {
+      "epoch": 0.8591974830832231,
+      "grad_norm": 0.0882524624466896,
+      "learning_rate": 4.23464917770157e-06,
+      "loss": 8.5479,
+      "step": 172050
+    },
+    {
+      "epoch": 0.8592474219081625,
+      "grad_norm": 0.09635046869516373,
+      "learning_rate": 4.233147262760019e-06,
+      "loss": 8.5266,
+      "step": 172060
+    },
+    {
+      "epoch": 0.8592973607331019,
+      "grad_norm": 0.09102864563465118,
+      "learning_rate": 4.231645347818469e-06,
+      "loss": 8.5349,
+      "step": 172070
+    },
+    {
+      "epoch": 0.8593472995580413,
+      "grad_norm": 0.09232207387685776,
+      "learning_rate": 4.2301434328769175e-06,
+      "loss": 8.5276,
+      "step": 172080
+    },
+    {
+      "epoch": 0.8593972383829809,
+      "grad_norm": 0.09616389870643616,
+      "learning_rate": 4.228641517935368e-06,
+      "loss": 8.5269,
+      "step": 172090
+    },
+    {
+      "epoch": 0.8594471772079203,
+      "grad_norm": 0.09412415325641632,
+      "learning_rate": 4.227139602993818e-06,
+      "loss": 8.5328,
+      "step": 172100
+    },
+    {
+      "epoch": 0.8594971160328597,
+      "grad_norm": 0.09291058033704758,
+      "learning_rate": 4.225637688052266e-06,
+      "loss": 8.5311,
+      "step": 172110
+    },
+    {
+      "epoch": 0.8595470548577991,
+      "grad_norm": 0.09364159405231476,
+      "learning_rate": 4.2241357731107165e-06,
+      "loss": 8.5437,
+      "step": 172120
+    },
+    {
+      "epoch": 0.8595969936827387,
+      "grad_norm": 0.09280949085950851,
+      "learning_rate": 4.222633858169166e-06,
+      "loss": 8.5254,
+      "step": 172130
+    },
+    {
+      "epoch": 0.8596469325076781,
+      "grad_norm": 0.09044373035430908,
+      "learning_rate": 4.221131943227615e-06,
+      "loss": 8.5283,
+      "step": 172140
+    },
+    {
+      "epoch": 0.8596968713326175,
+      "grad_norm": 0.0903530940413475,
+      "learning_rate": 4.219630028286065e-06,
+      "loss": 8.5234,
+      "step": 172150
+    },
+    {
+      "epoch": 0.8597468101575569,
+      "grad_norm": 0.09625843167304993,
+      "learning_rate": 4.218128113344514e-06,
+      "loss": 8.5441,
+      "step": 172160
+    },
+    {
+      "epoch": 0.8597967489824965,
+      "grad_norm": 0.08800741285085678,
+      "learning_rate": 4.216626198402964e-06,
+      "loss": 8.5485,
+      "step": 172170
+    },
+    {
+      "epoch": 0.8598466878074359,
+      "grad_norm": 0.0877734124660492,
+      "learning_rate": 4.215124283461413e-06,
+      "loss": 8.5253,
+      "step": 172180
+    },
+    {
+      "epoch": 0.8598966266323753,
+      "grad_norm": 0.09386802464723587,
+      "learning_rate": 4.213622368519863e-06,
+      "loss": 8.522,
+      "step": 172190
+    },
+    {
+      "epoch": 0.8599465654573147,
+      "grad_norm": 0.09531255811452866,
+      "learning_rate": 4.212120453578313e-06,
+      "loss": 8.5214,
+      "step": 172200
+    },
+    {
+      "epoch": 0.8599965042822543,
+      "grad_norm": 0.09443308413028717,
+      "learning_rate": 4.210618538636762e-06,
+      "loss": 8.5338,
+      "step": 172210
+    },
+    {
+      "epoch": 0.8600464431071937,
+      "grad_norm": 0.08627857267856598,
+      "learning_rate": 4.2091166236952115e-06,
+      "loss": 8.5226,
+      "step": 172220
+    },
+    {
+      "epoch": 0.8600963819321331,
+      "grad_norm": 0.09067834168672562,
+      "learning_rate": 4.207614708753661e-06,
+      "loss": 8.5358,
+      "step": 172230
+    },
+    {
+      "epoch": 0.8601463207570725,
+      "grad_norm": 0.08249117434024811,
+      "learning_rate": 4.20611279381211e-06,
+      "loss": 8.5418,
+      "step": 172240
+    },
+    {
+      "epoch": 0.8601962595820121,
+      "grad_norm": 0.08913594484329224,
+      "learning_rate": 4.20461087887056e-06,
+      "loss": 8.5061,
+      "step": 172250
+    },
+    {
+      "epoch": 0.8602461984069515,
+      "grad_norm": 0.09263671189546585,
+      "learning_rate": 4.20310896392901e-06,
+      "loss": 8.5203,
+      "step": 172260
+    },
+    {
+      "epoch": 0.8602961372318909,
+      "grad_norm": 0.09188566356897354,
+      "learning_rate": 4.201607048987459e-06,
+      "loss": 8.5296,
+      "step": 172270
+    },
+    {
+      "epoch": 0.8603460760568303,
+      "grad_norm": 0.0945729911327362,
+      "learning_rate": 4.200105134045908e-06,
+      "loss": 8.5218,
+      "step": 172280
+    },
+    {
+      "epoch": 0.8603960148817699,
+      "grad_norm": 0.09193189442157745,
+      "learning_rate": 4.1986032191043585e-06,
+      "loss": 8.5329,
+      "step": 172290
+    },
+    {
+      "epoch": 0.8604459537067093,
+      "grad_norm": 0.08755156397819519,
+      "learning_rate": 4.197101304162808e-06,
+      "loss": 8.5227,
+      "step": 172300
+    },
+    {
+      "epoch": 0.8604958925316487,
+      "grad_norm": 0.08843012899160385,
+      "learning_rate": 4.195599389221257e-06,
+      "loss": 8.5337,
+      "step": 172310
+    },
+    {
+      "epoch": 0.8605458313565881,
+      "grad_norm": 0.09322084486484528,
+      "learning_rate": 4.1940974742797065e-06,
+      "loss": 8.5342,
+      "step": 172320
+    },
+    {
+      "epoch": 0.8605957701815277,
+      "grad_norm": 0.09291605651378632,
+      "learning_rate": 4.192595559338156e-06,
+      "loss": 8.516,
+      "step": 172330
+    },
+    {
+      "epoch": 0.8606457090064671,
+      "grad_norm": 0.09476222097873688,
+      "learning_rate": 4.191093644396606e-06,
+      "loss": 8.5342,
+      "step": 172340
+    },
+    {
+      "epoch": 0.8606956478314065,
+      "grad_norm": 0.09349306672811508,
+      "learning_rate": 4.189591729455055e-06,
+      "loss": 8.5237,
+      "step": 172350
+    },
+    {
+      "epoch": 0.8607455866563459,
+      "grad_norm": 0.09003052115440369,
+      "learning_rate": 4.188089814513505e-06,
+      "loss": 8.5341,
+      "step": 172360
+    },
+    {
+      "epoch": 0.8607955254812855,
+      "grad_norm": 0.08857399970293045,
+      "learning_rate": 4.186587899571955e-06,
+      "loss": 8.5271,
+      "step": 172370
+    },
+    {
+      "epoch": 0.8608454643062249,
+      "grad_norm": 0.09072058647871017,
+      "learning_rate": 4.185085984630403e-06,
+      "loss": 8.5211,
+      "step": 172380
+    },
+    {
+      "epoch": 0.8608954031311643,
+      "grad_norm": 0.09271258860826492,
+      "learning_rate": 4.1835840696888535e-06,
+      "loss": 8.5312,
+      "step": 172390
+    },
+    {
+      "epoch": 0.8609453419561037,
+      "grad_norm": 0.09086389094591141,
+      "learning_rate": 4.182082154747303e-06,
+      "loss": 8.5093,
+      "step": 172400
+    },
+    {
+      "epoch": 0.8609952807810433,
+      "grad_norm": 0.09174318611621857,
+      "learning_rate": 4.180580239805752e-06,
+      "loss": 8.5343,
+      "step": 172410
+    },
+    {
+      "epoch": 0.8610452196059827,
+      "grad_norm": 0.1004238873720169,
+      "learning_rate": 4.179078324864202e-06,
+      "loss": 8.5097,
+      "step": 172420
+    },
+    {
+      "epoch": 0.8610951584309221,
+      "grad_norm": 0.09396187216043472,
+      "learning_rate": 4.177576409922651e-06,
+      "loss": 8.5219,
+      "step": 172430
+    },
+    {
+      "epoch": 0.8611450972558615,
+      "grad_norm": 0.09725307673215866,
+      "learning_rate": 4.176074494981101e-06,
+      "loss": 8.5057,
+      "step": 172440
+    },
+    {
+      "epoch": 0.8611950360808011,
+      "grad_norm": 0.0942634865641594,
+      "learning_rate": 4.174572580039551e-06,
+      "loss": 8.5479,
+      "step": 172450
+    },
+    {
+      "epoch": 0.8612449749057405,
+      "grad_norm": 0.09349733591079712,
+      "learning_rate": 4.173070665098e-06,
+      "loss": 8.5116,
+      "step": 172460
+    },
+    {
+      "epoch": 0.8612949137306799,
+      "grad_norm": 0.09070730209350586,
+      "learning_rate": 4.17156875015645e-06,
+      "loss": 8.5297,
+      "step": 172470
+    },
+    {
+      "epoch": 0.8613448525556193,
+      "grad_norm": 0.09090225398540497,
+      "learning_rate": 4.170066835214898e-06,
+      "loss": 8.5385,
+      "step": 172480
+    },
+    {
+      "epoch": 0.8613947913805589,
+      "grad_norm": 0.08755583316087723,
+      "learning_rate": 4.1685649202733485e-06,
+      "loss": 8.511,
+      "step": 172490
+    },
+    {
+      "epoch": 0.8614447302054983,
+      "grad_norm": 0.08842223882675171,
+      "learning_rate": 4.167063005331799e-06,
+      "loss": 8.53,
+      "step": 172500
+    },
+    {
+      "epoch": 0.8614946690304377,
+      "grad_norm": 0.08680158108472824,
+      "learning_rate": 4.165561090390247e-06,
+      "loss": 8.541,
+      "step": 172510
+    },
+    {
+      "epoch": 0.8615446078553771,
+      "grad_norm": 0.08641988784074783,
+      "learning_rate": 4.164059175448697e-06,
+      "loss": 8.5344,
+      "step": 172520
+    },
+    {
+      "epoch": 0.8615945466803167,
+      "grad_norm": 0.08488810807466507,
+      "learning_rate": 4.162557260507147e-06,
+      "loss": 8.5253,
+      "step": 172530
+    },
+    {
+      "epoch": 0.8616444855052561,
+      "grad_norm": 0.08614518493413925,
+      "learning_rate": 4.161055345565596e-06,
+      "loss": 8.5317,
+      "step": 172540
+    },
+    {
+      "epoch": 0.8616944243301955,
+      "grad_norm": 0.09032870084047318,
+      "learning_rate": 4.159553430624046e-06,
+      "loss": 8.5269,
+      "step": 172550
+    },
+    {
+      "epoch": 0.8617443631551349,
+      "grad_norm": 0.09359986335039139,
+      "learning_rate": 4.158051515682495e-06,
+      "loss": 8.5229,
+      "step": 172560
+    },
+    {
+      "epoch": 0.8617943019800745,
+      "grad_norm": 0.09470424801111221,
+      "learning_rate": 4.156549600740945e-06,
+      "loss": 8.5238,
+      "step": 172570
+    },
+    {
+      "epoch": 0.8618442408050139,
+      "grad_norm": 0.09623333811759949,
+      "learning_rate": 4.155047685799394e-06,
+      "loss": 8.5354,
+      "step": 172580
+    },
+    {
+      "epoch": 0.8618941796299533,
+      "grad_norm": 0.08704448491334915,
+      "learning_rate": 4.1535457708578436e-06,
+      "loss": 8.5351,
+      "step": 172590
+    },
+    {
+      "epoch": 0.8619441184548927,
+      "grad_norm": 0.09469842165708542,
+      "learning_rate": 4.152043855916294e-06,
+      "loss": 8.5372,
+      "step": 172600
+    },
+    {
+      "epoch": 0.8619940572798322,
+      "grad_norm": 0.09458217769861221,
+      "learning_rate": 4.150541940974742e-06,
+      "loss": 8.543,
+      "step": 172610
+    },
+    {
+      "epoch": 0.8620439961047717,
+      "grad_norm": 0.09154288470745087,
+      "learning_rate": 4.149040026033192e-06,
+      "loss": 8.5297,
+      "step": 172620
+    },
+    {
+      "epoch": 0.8620939349297111,
+      "grad_norm": 0.09419039636850357,
+      "learning_rate": 4.1475381110916426e-06,
+      "loss": 8.5533,
+      "step": 172630
+    },
+    {
+      "epoch": 0.8621438737546505,
+      "grad_norm": 0.09300480782985687,
+      "learning_rate": 4.146036196150091e-06,
+      "loss": 8.5419,
+      "step": 172640
+    },
+    {
+      "epoch": 0.86219381257959,
+      "grad_norm": 0.09122762829065323,
+      "learning_rate": 4.144534281208541e-06,
+      "loss": 8.5167,
+      "step": 172650
+    },
+    {
+      "epoch": 0.8622437514045295,
+      "grad_norm": 0.09213479608297348,
+      "learning_rate": 4.1430323662669906e-06,
+      "loss": 8.5241,
+      "step": 172660
+    },
+    {
+      "epoch": 0.8622936902294689,
+      "grad_norm": 0.09271895885467529,
+      "learning_rate": 4.14153045132544e-06,
+      "loss": 8.5352,
+      "step": 172670
+    },
+    {
+      "epoch": 0.8623436290544083,
+      "grad_norm": 0.08686958998441696,
+      "learning_rate": 4.14002853638389e-06,
+      "loss": 8.5167,
+      "step": 172680
+    },
+    {
+      "epoch": 0.8623935678793478,
+      "grad_norm": 0.08590806275606155,
+      "learning_rate": 4.1385266214423386e-06,
+      "loss": 8.5366,
+      "step": 172690
+    },
+    {
+      "epoch": 0.8624435067042873,
+      "grad_norm": 0.0904327780008316,
+      "learning_rate": 4.137024706500789e-06,
+      "loss": 8.5325,
+      "step": 172700
+    },
+    {
+      "epoch": 0.8624934455292267,
+      "grad_norm": 0.09133438020944595,
+      "learning_rate": 4.135522791559238e-06,
+      "loss": 8.5142,
+      "step": 172710
+    },
+    {
+      "epoch": 0.8625433843541661,
+      "grad_norm": 0.08940550684928894,
+      "learning_rate": 4.134020876617687e-06,
+      "loss": 8.51,
+      "step": 172720
+    },
+    {
+      "epoch": 0.8625933231791056,
+      "grad_norm": 0.09443370252847672,
+      "learning_rate": 4.132518961676138e-06,
+      "loss": 8.5289,
+      "step": 172730
+    },
+    {
+      "epoch": 0.8626432620040451,
+      "grad_norm": 0.09132427722215652,
+      "learning_rate": 4.131017046734587e-06,
+      "loss": 8.5351,
+      "step": 172740
+    },
+    {
+      "epoch": 0.8626932008289845,
+      "grad_norm": 0.08972794562578201,
+      "learning_rate": 4.129515131793036e-06,
+      "loss": 8.5338,
+      "step": 172750
+    },
+    {
+      "epoch": 0.8627431396539239,
+      "grad_norm": 0.09007082879543304,
+      "learning_rate": 4.128013216851486e-06,
+      "loss": 8.514,
+      "step": 172760
+    },
+    {
+      "epoch": 0.8627930784788634,
+      "grad_norm": 0.09930828213691711,
+      "learning_rate": 4.126511301909935e-06,
+      "loss": 8.5355,
+      "step": 172770
+    },
+    {
+      "epoch": 0.8628430173038029,
+      "grad_norm": 0.09435442090034485,
+      "learning_rate": 4.125009386968385e-06,
+      "loss": 8.539,
+      "step": 172780
+    },
+    {
+      "epoch": 0.8628929561287423,
+      "grad_norm": 0.09018246084451675,
+      "learning_rate": 4.123507472026834e-06,
+      "loss": 8.5403,
+      "step": 172790
+    },
+    {
+      "epoch": 0.8629428949536817,
+      "grad_norm": 0.09555966407060623,
+      "learning_rate": 4.122005557085284e-06,
+      "loss": 8.5138,
+      "step": 172800
+    },
+    {
+      "epoch": 0.8629928337786212,
+      "grad_norm": 0.09227462112903595,
+      "learning_rate": 4.120503642143733e-06,
+      "loss": 8.5498,
+      "step": 172810
+    },
+    {
+      "epoch": 0.8630427726035607,
+      "grad_norm": 0.08898046612739563,
+      "learning_rate": 4.119001727202183e-06,
+      "loss": 8.5182,
+      "step": 172820
+    },
+    {
+      "epoch": 0.8630927114285001,
+      "grad_norm": 0.08735623210668564,
+      "learning_rate": 4.117499812260633e-06,
+      "loss": 8.5232,
+      "step": 172830
+    },
+    {
+      "epoch": 0.8631426502534395,
+      "grad_norm": 0.09244896471500397,
+      "learning_rate": 4.115997897319082e-06,
+      "loss": 8.5405,
+      "step": 172840
+    },
+    {
+      "epoch": 0.863192589078379,
+      "grad_norm": 0.08761844038963318,
+      "learning_rate": 4.114495982377531e-06,
+      "loss": 8.5422,
+      "step": 172850
+    },
+    {
+      "epoch": 0.8632425279033185,
+      "grad_norm": 0.0887509286403656,
+      "learning_rate": 4.112994067435981e-06,
+      "loss": 8.5045,
+      "step": 172860
+    },
+    {
+      "epoch": 0.8632924667282579,
+      "grad_norm": 0.09360551834106445,
+      "learning_rate": 4.111492152494431e-06,
+      "loss": 8.5411,
+      "step": 172870
+    },
+    {
+      "epoch": 0.8633424055531973,
+      "grad_norm": 0.0890243873000145,
+      "learning_rate": 4.10999023755288e-06,
+      "loss": 8.5195,
+      "step": 172880
+    },
+    {
+      "epoch": 0.8633923443781368,
+      "grad_norm": 0.09118281304836273,
+      "learning_rate": 4.1084883226113294e-06,
+      "loss": 8.5162,
+      "step": 172890
+    },
+    {
+      "epoch": 0.8634422832030763,
+      "grad_norm": 0.08943572640419006,
+      "learning_rate": 4.10698640766978e-06,
+      "loss": 8.5457,
+      "step": 172900
+    },
+    {
+      "epoch": 0.8634922220280157,
+      "grad_norm": 0.09161094576120377,
+      "learning_rate": 4.105484492728228e-06,
+      "loss": 8.5199,
+      "step": 172910
+    },
+    {
+      "epoch": 0.8635421608529551,
+      "grad_norm": 0.09737364202737808,
+      "learning_rate": 4.103982577786678e-06,
+      "loss": 8.528,
+      "step": 172920
+    },
+    {
+      "epoch": 0.8635920996778946,
+      "grad_norm": 0.08938228338956833,
+      "learning_rate": 4.102480662845128e-06,
+      "loss": 8.5449,
+      "step": 172930
+    },
+    {
+      "epoch": 0.863642038502834,
+      "grad_norm": 0.09068731218576431,
+      "learning_rate": 4.100978747903577e-06,
+      "loss": 8.5252,
+      "step": 172940
+    },
+    {
+      "epoch": 0.8636919773277735,
+      "grad_norm": 0.09687364101409912,
+      "learning_rate": 4.099476832962027e-06,
+      "loss": 8.5316,
+      "step": 172950
+    },
+    {
+      "epoch": 0.8637419161527129,
+      "grad_norm": 0.09029592573642731,
+      "learning_rate": 4.097974918020476e-06,
+      "loss": 8.5268,
+      "step": 172960
+    },
+    {
+      "epoch": 0.8637918549776524,
+      "grad_norm": 0.0884825587272644,
+      "learning_rate": 4.096473003078926e-06,
+      "loss": 8.5217,
+      "step": 172970
+    },
+    {
+      "epoch": 0.8638417938025919,
+      "grad_norm": 0.09405983984470367,
+      "learning_rate": 4.094971088137376e-06,
+      "loss": 8.5221,
+      "step": 172980
+    },
+    {
+      "epoch": 0.8638917326275313,
+      "grad_norm": 0.0932493582367897,
+      "learning_rate": 4.0934691731958244e-06,
+      "loss": 8.5124,
+      "step": 172990
+    },
+    {
+      "epoch": 0.8639416714524707,
+      "grad_norm": 0.09520955383777618,
+      "learning_rate": 4.091967258254275e-06,
+      "loss": 8.5114,
+      "step": 173000
+    },
+    {
+      "epoch": 0.8639916102774102,
+      "grad_norm": 0.08801741153001785,
+      "learning_rate": 4.090465343312723e-06,
+      "loss": 8.5435,
+      "step": 173010
+    },
+    {
+      "epoch": 0.8640415491023496,
+      "grad_norm": 0.09374932199716568,
+      "learning_rate": 4.088963428371173e-06,
+      "loss": 8.513,
+      "step": 173020
+    },
+    {
+      "epoch": 0.8640914879272891,
+      "grad_norm": 0.09420095384120941,
+      "learning_rate": 4.0874615134296235e-06,
+      "loss": 8.5286,
+      "step": 173030
+    },
+    {
+      "epoch": 0.8641414267522285,
+      "grad_norm": 0.09835788607597351,
+      "learning_rate": 4.085959598488072e-06,
+      "loss": 8.514,
+      "step": 173040
+    },
+    {
+      "epoch": 0.864191365577168,
+      "grad_norm": 0.0848371684551239,
+      "learning_rate": 4.084457683546522e-06,
+      "loss": 8.5214,
+      "step": 173050
+    },
+    {
+      "epoch": 0.8642413044021074,
+      "grad_norm": 0.09027153998613358,
+      "learning_rate": 4.0829557686049714e-06,
+      "loss": 8.5318,
+      "step": 173060
+    },
+    {
+      "epoch": 0.8642912432270469,
+      "grad_norm": 0.09047489613294601,
+      "learning_rate": 4.081453853663421e-06,
+      "loss": 8.538,
+      "step": 173070
+    },
+    {
+      "epoch": 0.8643411820519863,
+      "grad_norm": 0.09193349629640579,
+      "learning_rate": 4.079951938721871e-06,
+      "loss": 8.517,
+      "step": 173080
+    },
+    {
+      "epoch": 0.8643911208769257,
+      "grad_norm": 0.09359715133905411,
+      "learning_rate": 4.0784500237803194e-06,
+      "loss": 8.5216,
+      "step": 173090
+    },
+    {
+      "epoch": 0.8644410597018652,
+      "grad_norm": 0.09064842015504837,
+      "learning_rate": 4.07694810883877e-06,
+      "loss": 8.5191,
+      "step": 173100
+    },
+    {
+      "epoch": 0.8644909985268047,
+      "grad_norm": 0.09371822327375412,
+      "learning_rate": 4.075446193897219e-06,
+      "loss": 8.5121,
+      "step": 173110
+    },
+    {
+      "epoch": 0.8645409373517441,
+      "grad_norm": 0.09190354496240616,
+      "learning_rate": 4.073944278955668e-06,
+      "loss": 8.4994,
+      "step": 173120
+    },
+    {
+      "epoch": 0.8645908761766835,
+      "grad_norm": 0.09204579889774323,
+      "learning_rate": 4.0724423640141185e-06,
+      "loss": 8.5339,
+      "step": 173130
+    },
+    {
+      "epoch": 0.864640815001623,
+      "grad_norm": 0.09318622201681137,
+      "learning_rate": 4.070940449072568e-06,
+      "loss": 8.5116,
+      "step": 173140
+    },
+    {
+      "epoch": 0.8646907538265625,
+      "grad_norm": 0.09237584471702576,
+      "learning_rate": 4.069438534131017e-06,
+      "loss": 8.5358,
+      "step": 173150
+    },
+    {
+      "epoch": 0.8647406926515019,
+      "grad_norm": 0.09040313959121704,
+      "learning_rate": 4.0679366191894665e-06,
+      "loss": 8.5232,
+      "step": 173160
+    },
+    {
+      "epoch": 0.8647906314764413,
+      "grad_norm": 0.09376216679811478,
+      "learning_rate": 4.066434704247916e-06,
+      "loss": 8.5028,
+      "step": 173170
+    },
+    {
+      "epoch": 0.8648405703013808,
+      "grad_norm": 0.08734447509050369,
+      "learning_rate": 4.064932789306366e-06,
+      "loss": 8.5056,
+      "step": 173180
+    },
+    {
+      "epoch": 0.8648905091263203,
+      "grad_norm": 0.0926559641957283,
+      "learning_rate": 4.063430874364815e-06,
+      "loss": 8.5099,
+      "step": 173190
+    },
+    {
+      "epoch": 0.8649404479512597,
+      "grad_norm": 0.08731448650360107,
+      "learning_rate": 4.061928959423265e-06,
+      "loss": 8.5165,
+      "step": 173200
+    },
+    {
+      "epoch": 0.8649903867761991,
+      "grad_norm": 0.08865402638912201,
+      "learning_rate": 4.060427044481714e-06,
+      "loss": 8.5291,
+      "step": 173210
+    },
+    {
+      "epoch": 0.8650403256011386,
+      "grad_norm": 0.08650528639554977,
+      "learning_rate": 4.058925129540164e-06,
+      "loss": 8.5137,
+      "step": 173220
+    },
+    {
+      "epoch": 0.8650902644260781,
+      "grad_norm": 0.08930091559886932,
+      "learning_rate": 4.0574232145986135e-06,
+      "loss": 8.5208,
+      "step": 173230
+    },
+    {
+      "epoch": 0.8651402032510175,
+      "grad_norm": 0.09235706925392151,
+      "learning_rate": 4.055921299657063e-06,
+      "loss": 8.5156,
+      "step": 173240
+    },
+    {
+      "epoch": 0.8651901420759569,
+      "grad_norm": 0.09423262625932693,
+      "learning_rate": 4.054419384715512e-06,
+      "loss": 8.5162,
+      "step": 173250
+    },
+    {
+      "epoch": 0.8652400809008964,
+      "grad_norm": 0.0886596292257309,
+      "learning_rate": 4.0529174697739615e-06,
+      "loss": 8.5152,
+      "step": 173260
+    },
+    {
+      "epoch": 0.8652900197258359,
+      "grad_norm": 0.08678395301103592,
+      "learning_rate": 4.051415554832412e-06,
+      "loss": 8.53,
+      "step": 173270
+    },
+    {
+      "epoch": 0.8653399585507753,
+      "grad_norm": 0.08788527548313141,
+      "learning_rate": 4.049913639890861e-06,
+      "loss": 8.5281,
+      "step": 173280
+    },
+    {
+      "epoch": 0.8653898973757147,
+      "grad_norm": 0.08688144385814667,
+      "learning_rate": 4.04841172494931e-06,
+      "loss": 8.5394,
+      "step": 173290
+    },
+    {
+      "epoch": 0.8654398362006542,
+      "grad_norm": 0.09183796495199203,
+      "learning_rate": 4.0469098100077605e-06,
+      "loss": 8.5355,
+      "step": 173300
+    },
+    {
+      "epoch": 0.8654897750255937,
+      "grad_norm": 0.09272154420614243,
+      "learning_rate": 4.045407895066209e-06,
+      "loss": 8.5449,
+      "step": 173310
+    },
+    {
+      "epoch": 0.8655397138505331,
+      "grad_norm": 0.0878337025642395,
+      "learning_rate": 4.043905980124659e-06,
+      "loss": 8.543,
+      "step": 173320
+    },
+    {
+      "epoch": 0.8655896526754725,
+      "grad_norm": 0.0915573239326477,
+      "learning_rate": 4.0424040651831085e-06,
+      "loss": 8.5153,
+      "step": 173330
+    },
+    {
+      "epoch": 0.865639591500412,
+      "grad_norm": 0.09731071442365646,
+      "learning_rate": 4.040902150241558e-06,
+      "loss": 8.5377,
+      "step": 173340
+    },
+    {
+      "epoch": 0.8656895303253515,
+      "grad_norm": 0.09491939097642899,
+      "learning_rate": 4.039400235300008e-06,
+      "loss": 8.5258,
+      "step": 173350
+    },
+    {
+      "epoch": 0.8657394691502909,
+      "grad_norm": 0.09417067468166351,
+      "learning_rate": 4.0378983203584565e-06,
+      "loss": 8.5305,
+      "step": 173360
+    },
+    {
+      "epoch": 0.8657894079752303,
+      "grad_norm": 0.09818973392248154,
+      "learning_rate": 4.036396405416907e-06,
+      "loss": 8.5226,
+      "step": 173370
+    },
+    {
+      "epoch": 0.8658393468001698,
+      "grad_norm": 0.10228383541107178,
+      "learning_rate": 4.034894490475357e-06,
+      "loss": 8.5162,
+      "step": 173380
+    },
+    {
+      "epoch": 0.8658892856251093,
+      "grad_norm": 0.09025364369153976,
+      "learning_rate": 4.033392575533805e-06,
+      "loss": 8.5282,
+      "step": 173390
+    },
+    {
+      "epoch": 0.8659392244500487,
+      "grad_norm": 0.09199435263872147,
+      "learning_rate": 4.0318906605922555e-06,
+      "loss": 8.5222,
+      "step": 173400
+    },
+    {
+      "epoch": 0.8659891632749881,
+      "grad_norm": 0.08623290807008743,
+      "learning_rate": 4.030388745650704e-06,
+      "loss": 8.5133,
+      "step": 173410
+    },
+    {
+      "epoch": 0.8660391020999276,
+      "grad_norm": 0.09399810433387756,
+      "learning_rate": 4.028886830709154e-06,
+      "loss": 8.5254,
+      "step": 173420
+    },
+    {
+      "epoch": 0.866089040924867,
+      "grad_norm": 0.08971837162971497,
+      "learning_rate": 4.027384915767604e-06,
+      "loss": 8.5333,
+      "step": 173430
+    },
+    {
+      "epoch": 0.8661389797498065,
+      "grad_norm": 0.08642110228538513,
+      "learning_rate": 4.025883000826053e-06,
+      "loss": 8.5237,
+      "step": 173440
+    },
+    {
+      "epoch": 0.8661889185747459,
+      "grad_norm": 0.09314007312059402,
+      "learning_rate": 4.024381085884503e-06,
+      "loss": 8.5056,
+      "step": 173450
+    },
+    {
+      "epoch": 0.8662388573996854,
+      "grad_norm": 0.09386899322271347,
+      "learning_rate": 4.022879170942952e-06,
+      "loss": 8.524,
+      "step": 173460
+    },
+    {
+      "epoch": 0.8662887962246248,
+      "grad_norm": 0.09295320510864258,
+      "learning_rate": 4.021377256001402e-06,
+      "loss": 8.5266,
+      "step": 173470
+    },
+    {
+      "epoch": 0.8663387350495643,
+      "grad_norm": 0.09699882566928864,
+      "learning_rate": 4.019875341059852e-06,
+      "loss": 8.5314,
+      "step": 173480
+    },
+    {
+      "epoch": 0.8663886738745037,
+      "grad_norm": 0.08937513083219528,
+      "learning_rate": 4.0183734261183e-06,
+      "loss": 8.5286,
+      "step": 173490
+    },
+    {
+      "epoch": 0.8664386126994432,
+      "grad_norm": 0.0886300727725029,
+      "learning_rate": 4.0168715111767505e-06,
+      "loss": 8.519,
+      "step": 173500
+    },
+    {
+      "epoch": 0.8664885515243826,
+      "grad_norm": 0.08783800154924393,
+      "learning_rate": 4.0153695962352e-06,
+      "loss": 8.5191,
+      "step": 173510
+    },
+    {
+      "epoch": 0.8665384903493221,
+      "grad_norm": 0.09207980334758759,
+      "learning_rate": 4.013867681293649e-06,
+      "loss": 8.5026,
+      "step": 173520
+    },
+    {
+      "epoch": 0.8665884291742615,
+      "grad_norm": 0.09273963421583176,
+      "learning_rate": 4.012365766352099e-06,
+      "loss": 8.5182,
+      "step": 173530
+    },
+    {
+      "epoch": 0.866638367999201,
+      "grad_norm": 0.08379366248846054,
+      "learning_rate": 4.010863851410549e-06,
+      "loss": 8.5278,
+      "step": 173540
+    },
+    {
+      "epoch": 0.8666883068241404,
+      "grad_norm": 0.08762573450803757,
+      "learning_rate": 4.009361936468998e-06,
+      "loss": 8.5255,
+      "step": 173550
+    },
+    {
+      "epoch": 0.8667382456490799,
+      "grad_norm": 0.088675856590271,
+      "learning_rate": 4.007860021527447e-06,
+      "loss": 8.5256,
+      "step": 173560
+    },
+    {
+      "epoch": 0.8667881844740193,
+      "grad_norm": 0.08745112270116806,
+      "learning_rate": 4.006358106585897e-06,
+      "loss": 8.5365,
+      "step": 173570
+    },
+    {
+      "epoch": 0.8668381232989588,
+      "grad_norm": 0.08667652308940887,
+      "learning_rate": 4.004856191644347e-06,
+      "loss": 8.5282,
+      "step": 173580
+    },
+    {
+      "epoch": 0.8668880621238982,
+      "grad_norm": 0.09410617500543594,
+      "learning_rate": 4.003354276702796e-06,
+      "loss": 8.5143,
+      "step": 173590
+    },
+    {
+      "epoch": 0.8669380009488377,
+      "grad_norm": 0.0875682532787323,
+      "learning_rate": 4.0018523617612455e-06,
+      "loss": 8.5359,
+      "step": 173600
+    },
+    {
+      "epoch": 0.8669879397737771,
+      "grad_norm": 0.08985573053359985,
+      "learning_rate": 4.000350446819695e-06,
+      "loss": 8.5462,
+      "step": 173610
+    },
+    {
+      "epoch": 0.8670378785987166,
+      "grad_norm": 0.09056945890188217,
+      "learning_rate": 3.998848531878145e-06,
+      "loss": 8.555,
+      "step": 173620
+    },
+    {
+      "epoch": 0.867087817423656,
+      "grad_norm": 0.09406935423612595,
+      "learning_rate": 3.997346616936594e-06,
+      "loss": 8.5018,
+      "step": 173630
+    },
+    {
+      "epoch": 0.8671377562485955,
+      "grad_norm": 0.08972731977701187,
+      "learning_rate": 3.995844701995044e-06,
+      "loss": 8.5286,
+      "step": 173640
+    },
+    {
+      "epoch": 0.8671876950735349,
+      "grad_norm": 0.09432131797075272,
+      "learning_rate": 3.994342787053493e-06,
+      "loss": 8.527,
+      "step": 173650
+    },
+    {
+      "epoch": 0.8672376338984744,
+      "grad_norm": 0.09251857548952103,
+      "learning_rate": 3.992840872111943e-06,
+      "loss": 8.5157,
+      "step": 173660
+    },
+    {
+      "epoch": 0.8672875727234138,
+      "grad_norm": 0.0903116911649704,
+      "learning_rate": 3.9913389571703925e-06,
+      "loss": 8.5218,
+      "step": 173670
+    },
+    {
+      "epoch": 0.8673375115483533,
+      "grad_norm": 0.09307548403739929,
+      "learning_rate": 3.989837042228842e-06,
+      "loss": 8.5296,
+      "step": 173680
+    },
+    {
+      "epoch": 0.8673874503732927,
+      "grad_norm": 0.08810535818338394,
+      "learning_rate": 3.988335127287291e-06,
+      "loss": 8.5257,
+      "step": 173690
+    },
+    {
+      "epoch": 0.8674373891982322,
+      "grad_norm": 0.09042003750801086,
+      "learning_rate": 3.986833212345741e-06,
+      "loss": 8.5462,
+      "step": 173700
+    },
+    {
+      "epoch": 0.8674873280231716,
+      "grad_norm": 0.09091649949550629,
+      "learning_rate": 3.985331297404191e-06,
+      "loss": 8.5249,
+      "step": 173710
+    },
+    {
+      "epoch": 0.867537266848111,
+      "grad_norm": 0.09188742935657501,
+      "learning_rate": 3.98382938246264e-06,
+      "loss": 8.5144,
+      "step": 173720
+    },
+    {
+      "epoch": 0.8675872056730505,
+      "grad_norm": 0.09003132581710815,
+      "learning_rate": 3.982327467521089e-06,
+      "loss": 8.5341,
+      "step": 173730
+    },
+    {
+      "epoch": 0.86763714449799,
+      "grad_norm": 0.09096615016460419,
+      "learning_rate": 3.980825552579539e-06,
+      "loss": 8.532,
+      "step": 173740
+    },
+    {
+      "epoch": 0.8676870833229294,
+      "grad_norm": 0.09003739058971405,
+      "learning_rate": 3.979323637637989e-06,
+      "loss": 8.5205,
+      "step": 173750
+    },
+    {
+      "epoch": 0.8677370221478689,
+      "grad_norm": 0.09441889822483063,
+      "learning_rate": 3.977821722696438e-06,
+      "loss": 8.5302,
+      "step": 173760
+    },
+    {
+      "epoch": 0.8677869609728083,
+      "grad_norm": 0.09117145091295242,
+      "learning_rate": 3.9763198077548875e-06,
+      "loss": 8.5129,
+      "step": 173770
+    },
+    {
+      "epoch": 0.8678368997977478,
+      "grad_norm": 0.0931628867983818,
+      "learning_rate": 3.974817892813338e-06,
+      "loss": 8.5337,
+      "step": 173780
+    },
+    {
+      "epoch": 0.8678868386226872,
+      "grad_norm": 0.09708821773529053,
+      "learning_rate": 3.973315977871786e-06,
+      "loss": 8.5166,
+      "step": 173790
+    },
+    {
+      "epoch": 0.8679367774476267,
+      "grad_norm": 0.08967326581478119,
+      "learning_rate": 3.971814062930236e-06,
+      "loss": 8.5294,
+      "step": 173800
+    },
+    {
+      "epoch": 0.8679867162725661,
+      "grad_norm": 0.09614343196153641,
+      "learning_rate": 3.970312147988686e-06,
+      "loss": 8.5138,
+      "step": 173810
+    },
+    {
+      "epoch": 0.8680366550975056,
+      "grad_norm": 0.09332430362701416,
+      "learning_rate": 3.968810233047135e-06,
+      "loss": 8.5287,
+      "step": 173820
+    },
+    {
+      "epoch": 0.868086593922445,
+      "grad_norm": 0.09939328581094742,
+      "learning_rate": 3.967308318105585e-06,
+      "loss": 8.5366,
+      "step": 173830
+    },
+    {
+      "epoch": 0.8681365327473844,
+      "grad_norm": 0.09348172694444656,
+      "learning_rate": 3.965806403164034e-06,
+      "loss": 8.524,
+      "step": 173840
+    },
+    {
+      "epoch": 0.8681864715723239,
+      "grad_norm": 0.09327511489391327,
+      "learning_rate": 3.964304488222484e-06,
+      "loss": 8.5344,
+      "step": 173850
+    },
+    {
+      "epoch": 0.8682364103972634,
+      "grad_norm": 0.089813232421875,
+      "learning_rate": 3.962802573280934e-06,
+      "loss": 8.5293,
+      "step": 173860
+    },
+    {
+      "epoch": 0.8682863492222028,
+      "grad_norm": 0.09029410034418106,
+      "learning_rate": 3.9613006583393825e-06,
+      "loss": 8.5281,
+      "step": 173870
+    },
+    {
+      "epoch": 0.8683362880471422,
+      "grad_norm": 0.09171853214502335,
+      "learning_rate": 3.959798743397833e-06,
+      "loss": 8.5053,
+      "step": 173880
+    },
+    {
+      "epoch": 0.8683862268720817,
+      "grad_norm": 0.08777931332588196,
+      "learning_rate": 3.958296828456281e-06,
+      "loss": 8.51,
+      "step": 173890
+    },
+    {
+      "epoch": 0.8684361656970212,
+      "grad_norm": 0.09193766117095947,
+      "learning_rate": 3.956794913514731e-06,
+      "loss": 8.5269,
+      "step": 173900
+    },
+    {
+      "epoch": 0.8684861045219606,
+      "grad_norm": 0.08400430530309677,
+      "learning_rate": 3.9552929985731815e-06,
+      "loss": 8.5356,
+      "step": 173910
+    },
+    {
+      "epoch": 0.8685360433469,
+      "grad_norm": 0.09466172009706497,
+      "learning_rate": 3.95379108363163e-06,
+      "loss": 8.5184,
+      "step": 173920
+    },
+    {
+      "epoch": 0.8685859821718395,
+      "grad_norm": 0.08804745227098465,
+      "learning_rate": 3.95228916869008e-06,
+      "loss": 8.5084,
+      "step": 173930
+    },
+    {
+      "epoch": 0.868635920996779,
+      "grad_norm": 0.08844754099845886,
+      "learning_rate": 3.9507872537485295e-06,
+      "loss": 8.5206,
+      "step": 173940
+    },
+    {
+      "epoch": 0.8686858598217184,
+      "grad_norm": 0.08931338787078857,
+      "learning_rate": 3.949285338806979e-06,
+      "loss": 8.5227,
+      "step": 173950
+    },
+    {
+      "epoch": 0.8687357986466578,
+      "grad_norm": 0.092722587287426,
+      "learning_rate": 3.947783423865429e-06,
+      "loss": 8.5357,
+      "step": 173960
+    },
+    {
+      "epoch": 0.8687857374715973,
+      "grad_norm": 0.08423691242933273,
+      "learning_rate": 3.9462815089238775e-06,
+      "loss": 8.5368,
+      "step": 173970
+    },
+    {
+      "epoch": 0.8688356762965368,
+      "grad_norm": 0.08981618285179138,
+      "learning_rate": 3.944779593982328e-06,
+      "loss": 8.5206,
+      "step": 173980
+    },
+    {
+      "epoch": 0.8688856151214762,
+      "grad_norm": 0.0881986990571022,
+      "learning_rate": 3.943277679040777e-06,
+      "loss": 8.5184,
+      "step": 173990
+    },
+    {
+      "epoch": 0.8689355539464156,
+      "grad_norm": 0.09177501499652863,
+      "learning_rate": 3.941775764099226e-06,
+      "loss": 8.5301,
+      "step": 174000
+    },
+    {
+      "epoch": 0.8689854927713551,
+      "grad_norm": 0.0890003889799118,
+      "learning_rate": 3.9402738491576766e-06,
+      "loss": 8.5185,
+      "step": 174010
+    },
+    {
+      "epoch": 0.8690354315962946,
+      "grad_norm": 0.08756007254123688,
+      "learning_rate": 3.938771934216126e-06,
+      "loss": 8.5242,
+      "step": 174020
+    },
+    {
+      "epoch": 0.869085370421234,
+      "grad_norm": 0.09130045026540756,
+      "learning_rate": 3.937270019274575e-06,
+      "loss": 8.5241,
+      "step": 174030
+    },
+    {
+      "epoch": 0.8691353092461734,
+      "grad_norm": 0.09635766595602036,
+      "learning_rate": 3.9357681043330246e-06,
+      "loss": 8.5303,
+      "step": 174040
+    },
+    {
+      "epoch": 0.8691852480711129,
+      "grad_norm": 0.09500184655189514,
+      "learning_rate": 3.934266189391474e-06,
+      "loss": 8.5289,
+      "step": 174050
+    },
+    {
+      "epoch": 0.8692351868960523,
+      "grad_norm": 0.08810250461101532,
+      "learning_rate": 3.932764274449924e-06,
+      "loss": 8.5216,
+      "step": 174060
+    },
+    {
+      "epoch": 0.8692851257209918,
+      "grad_norm": 0.08967293798923492,
+      "learning_rate": 3.931262359508373e-06,
+      "loss": 8.5238,
+      "step": 174070
+    },
+    {
+      "epoch": 0.8693350645459312,
+      "grad_norm": 0.0923919528722763,
+      "learning_rate": 3.929760444566823e-06,
+      "loss": 8.5099,
+      "step": 174080
+    },
+    {
+      "epoch": 0.8693850033708707,
+      "grad_norm": 0.09465716779232025,
+      "learning_rate": 3.928258529625272e-06,
+      "loss": 8.5313,
+      "step": 174090
+    },
+    {
+      "epoch": 0.8694349421958101,
+      "grad_norm": 0.08958126604557037,
+      "learning_rate": 3.926756614683722e-06,
+      "loss": 8.5358,
+      "step": 174100
+    },
+    {
+      "epoch": 0.8694848810207496,
+      "grad_norm": 0.0895143523812294,
+      "learning_rate": 3.9252546997421716e-06,
+      "loss": 8.5239,
+      "step": 174110
+    },
+    {
+      "epoch": 0.869534819845689,
+      "grad_norm": 0.0893610343337059,
+      "learning_rate": 3.923752784800621e-06,
+      "loss": 8.5262,
+      "step": 174120
+    },
+    {
+      "epoch": 0.8695847586706285,
+      "grad_norm": 0.0977412685751915,
+      "learning_rate": 3.92225086985907e-06,
+      "loss": 8.5102,
+      "step": 174130
+    },
+    {
+      "epoch": 0.8696346974955679,
+      "grad_norm": 0.09177926927804947,
+      "learning_rate": 3.9207489549175196e-06,
+      "loss": 8.5045,
+      "step": 174140
+    },
+    {
+      "epoch": 0.8696846363205074,
+      "grad_norm": 0.08462709933519363,
+      "learning_rate": 3.91924703997597e-06,
+      "loss": 8.5163,
+      "step": 174150
+    },
+    {
+      "epoch": 0.8697345751454468,
+      "grad_norm": 0.09177839756011963,
+      "learning_rate": 3.917745125034419e-06,
+      "loss": 8.5165,
+      "step": 174160
+    },
+    {
+      "epoch": 0.8697845139703863,
+      "grad_norm": 0.08942088484764099,
+      "learning_rate": 3.916243210092868e-06,
+      "loss": 8.542,
+      "step": 174170
+    },
+    {
+      "epoch": 0.8698344527953257,
+      "grad_norm": 0.0959954783320427,
+      "learning_rate": 3.914741295151319e-06,
+      "loss": 8.5206,
+      "step": 174180
+    },
+    {
+      "epoch": 0.8698843916202652,
+      "grad_norm": 0.09030862152576447,
+      "learning_rate": 3.913239380209767e-06,
+      "loss": 8.5127,
+      "step": 174190
+    },
+    {
+      "epoch": 0.8699343304452046,
+      "grad_norm": 0.08896990120410919,
+      "learning_rate": 3.911737465268217e-06,
+      "loss": 8.5275,
+      "step": 174200
+    },
+    {
+      "epoch": 0.869984269270144,
+      "grad_norm": 0.08994099497795105,
+      "learning_rate": 3.9102355503266666e-06,
+      "loss": 8.5147,
+      "step": 174210
+    },
+    {
+      "epoch": 0.8700342080950835,
+      "grad_norm": 0.0879681408405304,
+      "learning_rate": 3.908733635385116e-06,
+      "loss": 8.5341,
+      "step": 174220
+    },
+    {
+      "epoch": 0.870084146920023,
+      "grad_norm": 0.0852343812584877,
+      "learning_rate": 3.907231720443566e-06,
+      "loss": 8.5343,
+      "step": 174230
+    },
+    {
+      "epoch": 0.8701340857449624,
+      "grad_norm": 0.08718617260456085,
+      "learning_rate": 3.9057298055020146e-06,
+      "loss": 8.5334,
+      "step": 174240
+    },
+    {
+      "epoch": 0.8701840245699018,
+      "grad_norm": 0.09171243757009506,
+      "learning_rate": 3.904227890560465e-06,
+      "loss": 8.5341,
+      "step": 174250
+    },
+    {
+      "epoch": 0.8702339633948413,
+      "grad_norm": 0.08820389956235886,
+      "learning_rate": 3.902725975618915e-06,
+      "loss": 8.5417,
+      "step": 174260
+    },
+    {
+      "epoch": 0.8702839022197808,
+      "grad_norm": 0.08782992511987686,
+      "learning_rate": 3.901224060677363e-06,
+      "loss": 8.5333,
+      "step": 174270
+    },
+    {
+      "epoch": 0.8703338410447202,
+      "grad_norm": 0.08638270199298859,
+      "learning_rate": 3.899722145735814e-06,
+      "loss": 8.5396,
+      "step": 174280
+    },
+    {
+      "epoch": 0.8703837798696596,
+      "grad_norm": 0.09227801114320755,
+      "learning_rate": 3.898220230794262e-06,
+      "loss": 8.5034,
+      "step": 174290
+    },
+    {
+      "epoch": 0.8704337186945991,
+      "grad_norm": 0.09477287530899048,
+      "learning_rate": 3.896718315852712e-06,
+      "loss": 8.5067,
+      "step": 174300
+    },
+    {
+      "epoch": 0.8704836575195386,
+      "grad_norm": 0.09321620315313339,
+      "learning_rate": 3.8952164009111624e-06,
+      "loss": 8.5343,
+      "step": 174310
+    },
+    {
+      "epoch": 0.870533596344478,
+      "grad_norm": 0.08606991916894913,
+      "learning_rate": 3.893714485969611e-06,
+      "loss": 8.531,
+      "step": 174320
+    },
+    {
+      "epoch": 0.8705835351694174,
+      "grad_norm": 0.08892033994197845,
+      "learning_rate": 3.892212571028061e-06,
+      "loss": 8.531,
+      "step": 174330
+    },
+    {
+      "epoch": 0.8706334739943569,
+      "grad_norm": 0.08995813876390457,
+      "learning_rate": 3.89071065608651e-06,
+      "loss": 8.512,
+      "step": 174340
+    },
+    {
+      "epoch": 0.8706834128192964,
+      "grad_norm": 0.08822611719369888,
+      "learning_rate": 3.88920874114496e-06,
+      "loss": 8.527,
+      "step": 174350
+    },
+    {
+      "epoch": 0.8707333516442358,
+      "grad_norm": 0.09670012444257736,
+      "learning_rate": 3.88770682620341e-06,
+      "loss": 8.5085,
+      "step": 174360
+    },
+    {
+      "epoch": 0.8707832904691752,
+      "grad_norm": 0.0893353670835495,
+      "learning_rate": 3.886204911261858e-06,
+      "loss": 8.5151,
+      "step": 174370
+    },
+    {
+      "epoch": 0.8708332292941147,
+      "grad_norm": 0.0889049619436264,
+      "learning_rate": 3.884702996320309e-06,
+      "loss": 8.4997,
+      "step": 174380
+    },
+    {
+      "epoch": 0.8708831681190542,
+      "grad_norm": 0.08892576396465302,
+      "learning_rate": 3.883201081378758e-06,
+      "loss": 8.5215,
+      "step": 174390
+    },
+    {
+      "epoch": 0.8709331069439936,
+      "grad_norm": 0.09408024698495865,
+      "learning_rate": 3.881699166437207e-06,
+      "loss": 8.5249,
+      "step": 174400
+    },
+    {
+      "epoch": 0.870983045768933,
+      "grad_norm": 0.09239810705184937,
+      "learning_rate": 3.8801972514956574e-06,
+      "loss": 8.5204,
+      "step": 174410
+    },
+    {
+      "epoch": 0.8710329845938725,
+      "grad_norm": 0.08711476624011993,
+      "learning_rate": 3.878695336554107e-06,
+      "loss": 8.5171,
+      "step": 174420
+    },
+    {
+      "epoch": 0.871082923418812,
+      "grad_norm": 0.08488985896110535,
+      "learning_rate": 3.877193421612556e-06,
+      "loss": 8.5197,
+      "step": 174430
+    },
+    {
+      "epoch": 0.8711328622437514,
+      "grad_norm": 0.0881032943725586,
+      "learning_rate": 3.8756915066710054e-06,
+      "loss": 8.5149,
+      "step": 174440
+    },
+    {
+      "epoch": 0.8711828010686908,
+      "grad_norm": 0.0867091566324234,
+      "learning_rate": 3.874189591729455e-06,
+      "loss": 8.5106,
+      "step": 174450
+    },
+    {
+      "epoch": 0.8712327398936303,
+      "grad_norm": 0.09190431237220764,
+      "learning_rate": 3.872687676787905e-06,
+      "loss": 8.524,
+      "step": 174460
+    },
+    {
+      "epoch": 0.8712826787185698,
+      "grad_norm": 0.08907434344291687,
+      "learning_rate": 3.871185761846354e-06,
+      "loss": 8.5324,
+      "step": 174470
+    },
+    {
+      "epoch": 0.8713326175435092,
+      "grad_norm": 0.09729236364364624,
+      "learning_rate": 3.869683846904804e-06,
+      "loss": 8.5182,
+      "step": 174480
+    },
+    {
+      "epoch": 0.8713825563684486,
+      "grad_norm": 0.09291893243789673,
+      "learning_rate": 3.868181931963253e-06,
+      "loss": 8.5277,
+      "step": 174490
+    },
+    {
+      "epoch": 0.8714324951933881,
+      "grad_norm": 0.08960501104593277,
+      "learning_rate": 3.866680017021703e-06,
+      "loss": 8.5335,
+      "step": 174500
+    },
+    {
+      "epoch": 0.8714824340183276,
+      "grad_norm": 0.09359270334243774,
+      "learning_rate": 3.8651781020801524e-06,
+      "loss": 8.5333,
+      "step": 174510
+    },
+    {
+      "epoch": 0.871532372843267,
+      "grad_norm": 0.09425162523984909,
+      "learning_rate": 3.863676187138602e-06,
+      "loss": 8.5144,
+      "step": 174520
+    },
+    {
+      "epoch": 0.8715823116682064,
+      "grad_norm": 0.09045587480068207,
+      "learning_rate": 3.862174272197051e-06,
+      "loss": 8.5378,
+      "step": 174530
+    },
+    {
+      "epoch": 0.8716322504931459,
+      "grad_norm": 0.0939302071928978,
+      "learning_rate": 3.8606723572555004e-06,
+      "loss": 8.5233,
+      "step": 174540
+    },
+    {
+      "epoch": 0.8716821893180854,
+      "grad_norm": 0.08632202446460724,
+      "learning_rate": 3.859170442313951e-06,
+      "loss": 8.534,
+      "step": 174550
+    },
+    {
+      "epoch": 0.8717321281430248,
+      "grad_norm": 0.08903735131025314,
+      "learning_rate": 3.8576685273724e-06,
+      "loss": 8.537,
+      "step": 174560
+    },
+    {
+      "epoch": 0.8717820669679642,
+      "grad_norm": 0.08908429741859436,
+      "learning_rate": 3.856166612430849e-06,
+      "loss": 8.5324,
+      "step": 174570
+    },
+    {
+      "epoch": 0.8718320057929037,
+      "grad_norm": 0.08882272243499756,
+      "learning_rate": 3.8546646974892995e-06,
+      "loss": 8.5367,
+      "step": 174580
+    },
+    {
+      "epoch": 0.8718819446178432,
+      "grad_norm": 0.08850658684968948,
+      "learning_rate": 3.853162782547748e-06,
+      "loss": 8.5214,
+      "step": 174590
+    },
+    {
+      "epoch": 0.8719318834427826,
+      "grad_norm": 0.08655815571546555,
+      "learning_rate": 3.851660867606198e-06,
+      "loss": 8.5412,
+      "step": 174600
+    },
+    {
+      "epoch": 0.871981822267722,
+      "grad_norm": 0.08702993392944336,
+      "learning_rate": 3.8501589526646474e-06,
+      "loss": 8.5364,
+      "step": 174610
+    },
+    {
+      "epoch": 0.8720317610926615,
+      "grad_norm": 0.08738213032484055,
+      "learning_rate": 3.848657037723097e-06,
+      "loss": 8.517,
+      "step": 174620
+    },
+    {
+      "epoch": 0.872081699917601,
+      "grad_norm": 0.09153701364994049,
+      "learning_rate": 3.847155122781547e-06,
+      "loss": 8.504,
+      "step": 174630
+    },
+    {
+      "epoch": 0.8721316387425404,
+      "grad_norm": 0.09358830749988556,
+      "learning_rate": 3.8456532078399954e-06,
+      "loss": 8.5159,
+      "step": 174640
+    },
+    {
+      "epoch": 0.8721815775674798,
+      "grad_norm": 0.08956989645957947,
+      "learning_rate": 3.844151292898446e-06,
+      "loss": 8.5174,
+      "step": 174650
+    },
+    {
+      "epoch": 0.8722315163924192,
+      "grad_norm": 0.09366779029369354,
+      "learning_rate": 3.842649377956896e-06,
+      "loss": 8.5287,
+      "step": 174660
+    },
+    {
+      "epoch": 0.8722814552173588,
+      "grad_norm": 0.09194490313529968,
+      "learning_rate": 3.841147463015344e-06,
+      "loss": 8.5269,
+      "step": 174670
+    },
+    {
+      "epoch": 0.8723313940422982,
+      "grad_norm": 0.08473306894302368,
+      "learning_rate": 3.8396455480737945e-06,
+      "loss": 8.5332,
+      "step": 174680
+    },
+    {
+      "epoch": 0.8723813328672376,
+      "grad_norm": 0.09393809735774994,
+      "learning_rate": 3.838143633132243e-06,
+      "loss": 8.5116,
+      "step": 174690
+    },
+    {
+      "epoch": 0.872431271692177,
+      "grad_norm": 0.08729804307222366,
+      "learning_rate": 3.836641718190693e-06,
+      "loss": 8.5261,
+      "step": 174700
+    },
+    {
+      "epoch": 0.8724812105171166,
+      "grad_norm": 0.08915986865758896,
+      "learning_rate": 3.835139803249143e-06,
+      "loss": 8.5319,
+      "step": 174710
+    },
+    {
+      "epoch": 0.872531149342056,
+      "grad_norm": 0.08928072452545166,
+      "learning_rate": 3.833637888307592e-06,
+      "loss": 8.5346,
+      "step": 174720
+    },
+    {
+      "epoch": 0.8725810881669954,
+      "grad_norm": 0.09394170343875885,
+      "learning_rate": 3.832135973366042e-06,
+      "loss": 8.5138,
+      "step": 174730
+    },
+    {
+      "epoch": 0.8726310269919348,
+      "grad_norm": 0.09060081094503403,
+      "learning_rate": 3.830634058424491e-06,
+      "loss": 8.5273,
+      "step": 174740
+    },
+    {
+      "epoch": 0.8726809658168744,
+      "grad_norm": 0.08914122730493546,
+      "learning_rate": 3.829132143482941e-06,
+      "loss": 8.5229,
+      "step": 174750
+    },
+    {
+      "epoch": 0.8727309046418138,
+      "grad_norm": 0.09032674878835678,
+      "learning_rate": 3.827630228541391e-06,
+      "loss": 8.5249,
+      "step": 174760
+    },
+    {
+      "epoch": 0.8727808434667532,
+      "grad_norm": 0.09034202992916107,
+      "learning_rate": 3.826128313599839e-06,
+      "loss": 8.5362,
+      "step": 174770
+    },
+    {
+      "epoch": 0.8728307822916926,
+      "grad_norm": 0.08692754805088043,
+      "learning_rate": 3.8246263986582895e-06,
+      "loss": 8.5089,
+      "step": 174780
+    },
+    {
+      "epoch": 0.8728807211166322,
+      "grad_norm": 0.09234897792339325,
+      "learning_rate": 3.82312448371674e-06,
+      "loss": 8.5188,
+      "step": 174790
+    },
+    {
+      "epoch": 0.8729306599415716,
+      "grad_norm": 0.09195923805236816,
+      "learning_rate": 3.821622568775188e-06,
+      "loss": 8.5322,
+      "step": 174800
+    },
+    {
+      "epoch": 0.872980598766511,
+      "grad_norm": 0.0930572971701622,
+      "learning_rate": 3.820120653833638e-06,
+      "loss": 8.538,
+      "step": 174810
+    },
+    {
+      "epoch": 0.8730305375914504,
+      "grad_norm": 0.086077980697155,
+      "learning_rate": 3.818618738892088e-06,
+      "loss": 8.5311,
+      "step": 174820
+    },
+    {
+      "epoch": 0.87308047641639,
+      "grad_norm": 0.09503497928380966,
+      "learning_rate": 3.817116823950537e-06,
+      "loss": 8.5202,
+      "step": 174830
+    },
+    {
+      "epoch": 0.8731304152413294,
+      "grad_norm": 0.09077981114387512,
+      "learning_rate": 3.815614909008987e-06,
+      "loss": 8.5122,
+      "step": 174840
+    },
+    {
+      "epoch": 0.8731803540662688,
+      "grad_norm": 0.08733651787042618,
+      "learning_rate": 3.8141129940674356e-06,
+      "loss": 8.5313,
+      "step": 174850
+    },
+    {
+      "epoch": 0.8732302928912082,
+      "grad_norm": 0.08977235853672028,
+      "learning_rate": 3.8126110791258854e-06,
+      "loss": 8.5195,
+      "step": 174860
+    },
+    {
+      "epoch": 0.8732802317161478,
+      "grad_norm": 0.09327967464923859,
+      "learning_rate": 3.8111091641843356e-06,
+      "loss": 8.5095,
+      "step": 174870
+    },
+    {
+      "epoch": 0.8733301705410872,
+      "grad_norm": 0.08643508702516556,
+      "learning_rate": 3.8096072492427845e-06,
+      "loss": 8.5356,
+      "step": 174880
+    },
+    {
+      "epoch": 0.8733801093660266,
+      "grad_norm": 0.09138485044240952,
+      "learning_rate": 3.8081053343012342e-06,
+      "loss": 8.519,
+      "step": 174890
+    },
+    {
+      "epoch": 0.873430048190966,
+      "grad_norm": 0.09083333611488342,
+      "learning_rate": 3.806603419359684e-06,
+      "loss": 8.5285,
+      "step": 174900
+    },
+    {
+      "epoch": 0.8734799870159056,
+      "grad_norm": 0.09481485933065414,
+      "learning_rate": 3.805101504418133e-06,
+      "loss": 8.5086,
+      "step": 174910
+    },
+    {
+      "epoch": 0.873529925840845,
+      "grad_norm": 0.0904790461063385,
+      "learning_rate": 3.803599589476583e-06,
+      "loss": 8.5069,
+      "step": 174920
+    },
+    {
+      "epoch": 0.8735798646657844,
+      "grad_norm": 0.08950541913509369,
+      "learning_rate": 3.802097674535032e-06,
+      "loss": 8.5393,
+      "step": 174930
+    },
+    {
+      "epoch": 0.8736298034907238,
+      "grad_norm": 0.09654703736305237,
+      "learning_rate": 3.8005957595934817e-06,
+      "loss": 8.5304,
+      "step": 174940
+    },
+    {
+      "epoch": 0.8736797423156634,
+      "grad_norm": 0.09147506207227707,
+      "learning_rate": 3.7990938446519315e-06,
+      "loss": 8.5209,
+      "step": 174950
+    },
+    {
+      "epoch": 0.8737296811406028,
+      "grad_norm": 0.09071029722690582,
+      "learning_rate": 3.7975919297103804e-06,
+      "loss": 8.5118,
+      "step": 174960
+    },
+    {
+      "epoch": 0.8737796199655422,
+      "grad_norm": 0.0901302620768547,
+      "learning_rate": 3.7960900147688306e-06,
+      "loss": 8.5154,
+      "step": 174970
+    },
+    {
+      "epoch": 0.8738295587904816,
+      "grad_norm": 0.0890451967716217,
+      "learning_rate": 3.7945880998272803e-06,
+      "loss": 8.5242,
+      "step": 174980
+    },
+    {
+      "epoch": 0.8738794976154212,
+      "grad_norm": 0.09417781978845596,
+      "learning_rate": 3.7930861848857292e-06,
+      "loss": 8.5159,
+      "step": 174990
+    },
+    {
+      "epoch": 0.8739294364403606,
+      "grad_norm": 0.09130019694566727,
+      "learning_rate": 3.791584269944179e-06,
+      "loss": 8.4936,
+      "step": 175000
+    },
+    {
+      "epoch": 0.8739793752653,
+      "grad_norm": 0.09282857924699783,
+      "learning_rate": 3.7900823550026283e-06,
+      "loss": 8.5298,
+      "step": 175010
+    },
+    {
+      "epoch": 0.8740293140902394,
+      "grad_norm": 0.09820152819156647,
+      "learning_rate": 3.788580440061078e-06,
+      "loss": 8.5198,
+      "step": 175020
+    },
+    {
+      "epoch": 0.8740792529151789,
+      "grad_norm": 0.0940333679318428,
+      "learning_rate": 3.787078525119528e-06,
+      "loss": 8.5207,
+      "step": 175030
+    },
+    {
+      "epoch": 0.8741291917401184,
+      "grad_norm": 0.09212444722652435,
+      "learning_rate": 3.7855766101779767e-06,
+      "loss": 8.5327,
+      "step": 175040
+    },
+    {
+      "epoch": 0.8741791305650578,
+      "grad_norm": 0.09330432116985321,
+      "learning_rate": 3.7840746952364265e-06,
+      "loss": 8.5186,
+      "step": 175050
+    },
+    {
+      "epoch": 0.8742290693899972,
+      "grad_norm": 0.09266900271177292,
+      "learning_rate": 3.7825727802948767e-06,
+      "loss": 8.5161,
+      "step": 175060
+    },
+    {
+      "epoch": 0.8742790082149366,
+      "grad_norm": 0.09173017740249634,
+      "learning_rate": 3.7810708653533256e-06,
+      "loss": 8.5348,
+      "step": 175070
+    },
+    {
+      "epoch": 0.8743289470398762,
+      "grad_norm": 0.09216035157442093,
+      "learning_rate": 3.7795689504117753e-06,
+      "loss": 8.5296,
+      "step": 175080
+    },
+    {
+      "epoch": 0.8743788858648156,
+      "grad_norm": 0.08925978094339371,
+      "learning_rate": 3.7780670354702242e-06,
+      "loss": 8.5101,
+      "step": 175090
+    },
+    {
+      "epoch": 0.874428824689755,
+      "grad_norm": 0.09558253735303879,
+      "learning_rate": 3.776565120528674e-06,
+      "loss": 8.5213,
+      "step": 175100
+    },
+    {
+      "epoch": 0.8744787635146944,
+      "grad_norm": 0.09404308348894119,
+      "learning_rate": 3.775063205587124e-06,
+      "loss": 8.5122,
+      "step": 175110
+    },
+    {
+      "epoch": 0.874528702339634,
+      "grad_norm": 0.09096416085958481,
+      "learning_rate": 3.773561290645573e-06,
+      "loss": 8.5048,
+      "step": 175120
+    },
+    {
+      "epoch": 0.8745786411645734,
+      "grad_norm": 0.0968070700764656,
+      "learning_rate": 3.772059375704023e-06,
+      "loss": 8.5088,
+      "step": 175130
+    },
+    {
+      "epoch": 0.8746285799895128,
+      "grad_norm": 0.09349347651004791,
+      "learning_rate": 3.7705574607624726e-06,
+      "loss": 8.5247,
+      "step": 175140
+    },
+    {
+      "epoch": 0.8746785188144522,
+      "grad_norm": 0.09110265225172043,
+      "learning_rate": 3.7690555458209215e-06,
+      "loss": 8.5225,
+      "step": 175150
+    },
+    {
+      "epoch": 0.8747284576393918,
+      "grad_norm": 0.09158271551132202,
+      "learning_rate": 3.7675536308793717e-06,
+      "loss": 8.5216,
+      "step": 175160
+    },
+    {
+      "epoch": 0.8747783964643312,
+      "grad_norm": 0.09609760344028473,
+      "learning_rate": 3.7660517159378206e-06,
+      "loss": 8.5151,
+      "step": 175170
+    },
+    {
+      "epoch": 0.8748283352892706,
+      "grad_norm": 0.09342726320028305,
+      "learning_rate": 3.7645498009962703e-06,
+      "loss": 8.5056,
+      "step": 175180
+    },
+    {
+      "epoch": 0.87487827411421,
+      "grad_norm": 0.09422183781862259,
+      "learning_rate": 3.76304788605472e-06,
+      "loss": 8.5344,
+      "step": 175190
+    },
+    {
+      "epoch": 0.8749282129391496,
+      "grad_norm": 0.0885339081287384,
+      "learning_rate": 3.761545971113169e-06,
+      "loss": 8.5224,
+      "step": 175200
+    },
+    {
+      "epoch": 0.874978151764089,
+      "grad_norm": 0.09695085138082504,
+      "learning_rate": 3.760044056171619e-06,
+      "loss": 8.5084,
+      "step": 175210
+    },
+    {
+      "epoch": 0.8750280905890284,
+      "grad_norm": 0.08846613019704819,
+      "learning_rate": 3.758542141230069e-06,
+      "loss": 8.5175,
+      "step": 175220
+    },
+    {
+      "epoch": 0.8750780294139678,
+      "grad_norm": 0.09065619111061096,
+      "learning_rate": 3.757040226288518e-06,
+      "loss": 8.5186,
+      "step": 175230
+    },
+    {
+      "epoch": 0.8751279682389074,
+      "grad_norm": 0.09855975955724716,
+      "learning_rate": 3.7555383113469676e-06,
+      "loss": 8.5197,
+      "step": 175240
+    },
+    {
+      "epoch": 0.8751779070638468,
+      "grad_norm": 0.09044459462165833,
+      "learning_rate": 3.7540363964054165e-06,
+      "loss": 8.5144,
+      "step": 175250
+    },
+    {
+      "epoch": 0.8752278458887862,
+      "grad_norm": 0.08905865252017975,
+      "learning_rate": 3.7525344814638667e-06,
+      "loss": 8.5222,
+      "step": 175260
+    },
+    {
+      "epoch": 0.8752777847137256,
+      "grad_norm": 0.11130794882774353,
+      "learning_rate": 3.7510325665223164e-06,
+      "loss": 8.5262,
+      "step": 175270
+    },
+    {
+      "epoch": 0.8753277235386652,
+      "grad_norm": 0.08949120342731476,
+      "learning_rate": 3.7495306515807658e-06,
+      "loss": 8.5224,
+      "step": 175280
+    },
+    {
+      "epoch": 0.8753776623636046,
+      "grad_norm": 0.09382175654172897,
+      "learning_rate": 3.748028736639215e-06,
+      "loss": 8.5169,
+      "step": 175290
+    },
+    {
+      "epoch": 0.875427601188544,
+      "grad_norm": 0.09268946945667267,
+      "learning_rate": 3.7465268216976644e-06,
+      "loss": 8.5137,
+      "step": 175300
+    },
+    {
+      "epoch": 0.8754775400134834,
+      "grad_norm": 0.09075185656547546,
+      "learning_rate": 3.745024906756114e-06,
+      "loss": 8.5204,
+      "step": 175310
+    },
+    {
+      "epoch": 0.875527478838423,
+      "grad_norm": 0.0939631536602974,
+      "learning_rate": 3.743522991814564e-06,
+      "loss": 8.519,
+      "step": 175320
+    },
+    {
+      "epoch": 0.8755774176633624,
+      "grad_norm": 0.08865179121494293,
+      "learning_rate": 3.7420210768730133e-06,
+      "loss": 8.5212,
+      "step": 175330
+    },
+    {
+      "epoch": 0.8756273564883018,
+      "grad_norm": 0.0853448212146759,
+      "learning_rate": 3.7405191619314626e-06,
+      "loss": 8.5186,
+      "step": 175340
+    },
+    {
+      "epoch": 0.8756772953132412,
+      "grad_norm": 0.09132332354784012,
+      "learning_rate": 3.739017246989912e-06,
+      "loss": 8.5169,
+      "step": 175350
+    },
+    {
+      "epoch": 0.8757272341381808,
+      "grad_norm": 0.08930841088294983,
+      "learning_rate": 3.737515332048362e-06,
+      "loss": 8.5192,
+      "step": 175360
+    },
+    {
+      "epoch": 0.8757771729631202,
+      "grad_norm": 0.09462074190378189,
+      "learning_rate": 3.7360134171068115e-06,
+      "loss": 8.521,
+      "step": 175370
+    },
+    {
+      "epoch": 0.8758271117880596,
+      "grad_norm": 0.09071394056081772,
+      "learning_rate": 3.7345115021652608e-06,
+      "loss": 8.5251,
+      "step": 175380
+    },
+    {
+      "epoch": 0.875877050612999,
+      "grad_norm": 0.09404830634593964,
+      "learning_rate": 3.73300958722371e-06,
+      "loss": 8.5084,
+      "step": 175390
+    },
+    {
+      "epoch": 0.8759269894379386,
+      "grad_norm": 0.09391729533672333,
+      "learning_rate": 3.73150767228216e-06,
+      "loss": 8.5464,
+      "step": 175400
+    },
+    {
+      "epoch": 0.875976928262878,
+      "grad_norm": 0.08704748004674911,
+      "learning_rate": 3.7300057573406096e-06,
+      "loss": 8.516,
+      "step": 175410
+    },
+    {
+      "epoch": 0.8760268670878174,
+      "grad_norm": 0.096518374979496,
+      "learning_rate": 3.728503842399059e-06,
+      "loss": 8.5001,
+      "step": 175420
+    },
+    {
+      "epoch": 0.8760768059127568,
+      "grad_norm": 0.09533105045557022,
+      "learning_rate": 3.7270019274575083e-06,
+      "loss": 8.517,
+      "step": 175430
+    },
+    {
+      "epoch": 0.8761267447376964,
+      "grad_norm": 0.09067413210868835,
+      "learning_rate": 3.725500012515958e-06,
+      "loss": 8.5323,
+      "step": 175440
+    },
+    {
+      "epoch": 0.8761766835626358,
+      "grad_norm": 0.09444481879472733,
+      "learning_rate": 3.7239980975744074e-06,
+      "loss": 8.5037,
+      "step": 175450
+    },
+    {
+      "epoch": 0.8762266223875752,
+      "grad_norm": 0.08779852837324142,
+      "learning_rate": 3.722496182632857e-06,
+      "loss": 8.5255,
+      "step": 175460
+    },
+    {
+      "epoch": 0.8762765612125146,
+      "grad_norm": 0.08670119941234589,
+      "learning_rate": 3.7209942676913065e-06,
+      "loss": 8.5018,
+      "step": 175470
+    },
+    {
+      "epoch": 0.8763265000374542,
+      "grad_norm": 0.09427984058856964,
+      "learning_rate": 3.719492352749756e-06,
+      "loss": 8.5081,
+      "step": 175480
+    },
+    {
+      "epoch": 0.8763764388623936,
+      "grad_norm": 0.09233436733484268,
+      "learning_rate": 3.7179904378082055e-06,
+      "loss": 8.5266,
+      "step": 175490
+    },
+    {
+      "epoch": 0.876426377687333,
+      "grad_norm": 0.08909424394369125,
+      "learning_rate": 3.716488522866655e-06,
+      "loss": 8.5278,
+      "step": 175500
+    },
+    {
+      "epoch": 0.8764763165122724,
+      "grad_norm": 0.08954276144504547,
+      "learning_rate": 3.7149866079251046e-06,
+      "loss": 8.5341,
+      "step": 175510
+    },
+    {
+      "epoch": 0.876526255337212,
+      "grad_norm": 0.09071674197912216,
+      "learning_rate": 3.713484692983554e-06,
+      "loss": 8.5145,
+      "step": 175520
+    },
+    {
+      "epoch": 0.8765761941621514,
+      "grad_norm": 0.08852982521057129,
+      "learning_rate": 3.7119827780420037e-06,
+      "loss": 8.519,
+      "step": 175530
+    },
+    {
+      "epoch": 0.8766261329870908,
+      "grad_norm": 0.08784948289394379,
+      "learning_rate": 3.710480863100453e-06,
+      "loss": 8.5251,
+      "step": 175540
+    },
+    {
+      "epoch": 0.8766760718120302,
+      "grad_norm": 0.09313932061195374,
+      "learning_rate": 3.708978948158903e-06,
+      "loss": 8.5231,
+      "step": 175550
+    },
+    {
+      "epoch": 0.8767260106369698,
+      "grad_norm": 0.09138453006744385,
+      "learning_rate": 3.707477033217352e-06,
+      "loss": 8.5293,
+      "step": 175560
+    },
+    {
+      "epoch": 0.8767759494619092,
+      "grad_norm": 0.08735091984272003,
+      "learning_rate": 3.705975118275802e-06,
+      "loss": 8.5286,
+      "step": 175570
+    },
+    {
+      "epoch": 0.8768258882868486,
+      "grad_norm": 0.09122058749198914,
+      "learning_rate": 3.7044732033342512e-06,
+      "loss": 8.4957,
+      "step": 175580
+    },
+    {
+      "epoch": 0.876875827111788,
+      "grad_norm": 0.0900067389011383,
+      "learning_rate": 3.7029712883927006e-06,
+      "loss": 8.5311,
+      "step": 175590
+    },
+    {
+      "epoch": 0.8769257659367276,
+      "grad_norm": 0.08420617878437042,
+      "learning_rate": 3.7014693734511503e-06,
+      "loss": 8.5117,
+      "step": 175600
+    },
+    {
+      "epoch": 0.876975704761667,
+      "grad_norm": 0.08814489841461182,
+      "learning_rate": 3.6999674585096e-06,
+      "loss": 8.5307,
+      "step": 175610
+    },
+    {
+      "epoch": 0.8770256435866064,
+      "grad_norm": 0.09584464132785797,
+      "learning_rate": 3.6984655435680494e-06,
+      "loss": 8.525,
+      "step": 175620
+    },
+    {
+      "epoch": 0.8770755824115458,
+      "grad_norm": 0.08932922780513763,
+      "learning_rate": 3.6969636286264987e-06,
+      "loss": 8.5171,
+      "step": 175630
+    },
+    {
+      "epoch": 0.8771255212364854,
+      "grad_norm": 0.08885538578033447,
+      "learning_rate": 3.695461713684948e-06,
+      "loss": 8.5202,
+      "step": 175640
+    },
+    {
+      "epoch": 0.8771754600614248,
+      "grad_norm": 0.09274578839540482,
+      "learning_rate": 3.6939597987433982e-06,
+      "loss": 8.5244,
+      "step": 175650
+    },
+    {
+      "epoch": 0.8772253988863642,
+      "grad_norm": 0.09348073601722717,
+      "learning_rate": 3.6924578838018476e-06,
+      "loss": 8.5243,
+      "step": 175660
+    },
+    {
+      "epoch": 0.8772753377113036,
+      "grad_norm": 0.0866040512919426,
+      "learning_rate": 3.690955968860297e-06,
+      "loss": 8.5139,
+      "step": 175670
+    },
+    {
+      "epoch": 0.8773252765362431,
+      "grad_norm": 0.08929452300071716,
+      "learning_rate": 3.6894540539187462e-06,
+      "loss": 8.5253,
+      "step": 175680
+    },
+    {
+      "epoch": 0.8773752153611826,
+      "grad_norm": 0.0889781191945076,
+      "learning_rate": 3.687952138977196e-06,
+      "loss": 8.5166,
+      "step": 175690
+    },
+    {
+      "epoch": 0.877425154186122,
+      "grad_norm": 0.09169498085975647,
+      "learning_rate": 3.6864502240356457e-06,
+      "loss": 8.5326,
+      "step": 175700
+    },
+    {
+      "epoch": 0.8774750930110614,
+      "grad_norm": 0.08967631310224533,
+      "learning_rate": 3.684948309094095e-06,
+      "loss": 8.5351,
+      "step": 175710
+    },
+    {
+      "epoch": 0.877525031836001,
+      "grad_norm": 0.09231317043304443,
+      "learning_rate": 3.6834463941525444e-06,
+      "loss": 8.4988,
+      "step": 175720
+    },
+    {
+      "epoch": 0.8775749706609404,
+      "grad_norm": 0.09632280468940735,
+      "learning_rate": 3.681944479210994e-06,
+      "loss": 8.5177,
+      "step": 175730
+    },
+    {
+      "epoch": 0.8776249094858798,
+      "grad_norm": 0.0874035581946373,
+      "learning_rate": 3.6804425642694435e-06,
+      "loss": 8.5378,
+      "step": 175740
+    },
+    {
+      "epoch": 0.8776748483108192,
+      "grad_norm": 0.08852864801883698,
+      "learning_rate": 3.6789406493278932e-06,
+      "loss": 8.522,
+      "step": 175750
+    },
+    {
+      "epoch": 0.8777247871357587,
+      "grad_norm": 0.09550099074840546,
+      "learning_rate": 3.6774387343863426e-06,
+      "loss": 8.5171,
+      "step": 175760
+    },
+    {
+      "epoch": 0.8777747259606982,
+      "grad_norm": 0.09109804779291153,
+      "learning_rate": 3.6759368194447923e-06,
+      "loss": 8.5383,
+      "step": 175770
+    },
+    {
+      "epoch": 0.8778246647856376,
+      "grad_norm": 0.08961420506238937,
+      "learning_rate": 3.6744349045032417e-06,
+      "loss": 8.5162,
+      "step": 175780
+    },
+    {
+      "epoch": 0.877874603610577,
+      "grad_norm": 0.0872698426246643,
+      "learning_rate": 3.672932989561691e-06,
+      "loss": 8.5116,
+      "step": 175790
+    },
+    {
+      "epoch": 0.8779245424355165,
+      "grad_norm": 0.0968591645359993,
+      "learning_rate": 3.6714310746201407e-06,
+      "loss": 8.5413,
+      "step": 175800
+    },
+    {
+      "epoch": 0.877974481260456,
+      "grad_norm": 0.09531614184379578,
+      "learning_rate": 3.6699291596785905e-06,
+      "loss": 8.5404,
+      "step": 175810
+    },
+    {
+      "epoch": 0.8780244200853954,
+      "grad_norm": 0.09465590864419937,
+      "learning_rate": 3.66842724473704e-06,
+      "loss": 8.5129,
+      "step": 175820
+    },
+    {
+      "epoch": 0.8780743589103348,
+      "grad_norm": 0.09236373007297516,
+      "learning_rate": 3.666925329795489e-06,
+      "loss": 8.5317,
+      "step": 175830
+    },
+    {
+      "epoch": 0.8781242977352743,
+      "grad_norm": 0.09544330835342407,
+      "learning_rate": 3.6654234148539385e-06,
+      "loss": 8.5133,
+      "step": 175840
+    },
+    {
+      "epoch": 0.8781742365602138,
+      "grad_norm": 0.08793754130601883,
+      "learning_rate": 3.6639214999123887e-06,
+      "loss": 8.519,
+      "step": 175850
+    },
+    {
+      "epoch": 0.8782241753851532,
+      "grad_norm": 0.09597655385732651,
+      "learning_rate": 3.662419584970838e-06,
+      "loss": 8.5266,
+      "step": 175860
+    },
+    {
+      "epoch": 0.8782741142100926,
+      "grad_norm": 0.09199637174606323,
+      "learning_rate": 3.6609176700292873e-06,
+      "loss": 8.5188,
+      "step": 175870
+    },
+    {
+      "epoch": 0.8783240530350321,
+      "grad_norm": 0.09712479263544083,
+      "learning_rate": 3.6594157550877367e-06,
+      "loss": 8.5235,
+      "step": 175880
+    },
+    {
+      "epoch": 0.8783739918599716,
+      "grad_norm": 0.08960530161857605,
+      "learning_rate": 3.6579138401461864e-06,
+      "loss": 8.5275,
+      "step": 175890
+    },
+    {
+      "epoch": 0.878423930684911,
+      "grad_norm": 0.08847316354513168,
+      "learning_rate": 3.656411925204636e-06,
+      "loss": 8.5109,
+      "step": 175900
+    },
+    {
+      "epoch": 0.8784738695098504,
+      "grad_norm": 0.09034876525402069,
+      "learning_rate": 3.6549100102630855e-06,
+      "loss": 8.54,
+      "step": 175910
+    },
+    {
+      "epoch": 0.8785238083347899,
+      "grad_norm": 0.09454880654811859,
+      "learning_rate": 3.653408095321535e-06,
+      "loss": 8.5137,
+      "step": 175920
+    },
+    {
+      "epoch": 0.8785737471597294,
+      "grad_norm": 0.08720461279153824,
+      "learning_rate": 3.6519061803799846e-06,
+      "loss": 8.5258,
+      "step": 175930
+    },
+    {
+      "epoch": 0.8786236859846688,
+      "grad_norm": 0.08554758876562119,
+      "learning_rate": 3.650404265438434e-06,
+      "loss": 8.514,
+      "step": 175940
+    },
+    {
+      "epoch": 0.8786736248096082,
+      "grad_norm": 0.09344278275966644,
+      "learning_rate": 3.6489023504968837e-06,
+      "loss": 8.5226,
+      "step": 175950
+    },
+    {
+      "epoch": 0.8787235636345477,
+      "grad_norm": 0.09505718946456909,
+      "learning_rate": 3.647400435555333e-06,
+      "loss": 8.4881,
+      "step": 175960
+    },
+    {
+      "epoch": 0.8787735024594872,
+      "grad_norm": 0.08501995354890823,
+      "learning_rate": 3.6458985206137828e-06,
+      "loss": 8.5071,
+      "step": 175970
+    },
+    {
+      "epoch": 0.8788234412844266,
+      "grad_norm": 0.09370040893554688,
+      "learning_rate": 3.644396605672232e-06,
+      "loss": 8.5162,
+      "step": 175980
+    },
+    {
+      "epoch": 0.878873380109366,
+      "grad_norm": 0.08761072903871536,
+      "learning_rate": 3.6428946907306814e-06,
+      "loss": 8.5417,
+      "step": 175990
+    },
+    {
+      "epoch": 0.8789233189343055,
+      "grad_norm": 0.08839986473321915,
+      "learning_rate": 3.641392775789131e-06,
+      "loss": 8.5306,
+      "step": 176000
+    },
+    {
+      "epoch": 0.878973257759245,
+      "grad_norm": 0.09286805242300034,
+      "learning_rate": 3.639890860847581e-06,
+      "loss": 8.526,
+      "step": 176010
+    },
+    {
+      "epoch": 0.8790231965841844,
+      "grad_norm": 0.08983197808265686,
+      "learning_rate": 3.6383889459060303e-06,
+      "loss": 8.5081,
+      "step": 176020
+    },
+    {
+      "epoch": 0.8790731354091238,
+      "grad_norm": 0.09328821301460266,
+      "learning_rate": 3.6368870309644796e-06,
+      "loss": 8.5201,
+      "step": 176030
+    },
+    {
+      "epoch": 0.8791230742340632,
+      "grad_norm": 0.08646514266729355,
+      "learning_rate": 3.635385116022929e-06,
+      "loss": 8.522,
+      "step": 176040
+    },
+    {
+      "epoch": 0.8791730130590028,
+      "grad_norm": 0.09557259827852249,
+      "learning_rate": 3.633883201081379e-06,
+      "loss": 8.5124,
+      "step": 176050
+    },
+    {
+      "epoch": 0.8792229518839422,
+      "grad_norm": 0.08780386298894882,
+      "learning_rate": 3.6323812861398284e-06,
+      "loss": 8.5364,
+      "step": 176060
+    },
+    {
+      "epoch": 0.8792728907088816,
+      "grad_norm": 0.08918431401252747,
+      "learning_rate": 3.6308793711982778e-06,
+      "loss": 8.5319,
+      "step": 176070
+    },
+    {
+      "epoch": 0.879322829533821,
+      "grad_norm": 0.09585704654455185,
+      "learning_rate": 3.629377456256727e-06,
+      "loss": 8.5099,
+      "step": 176080
+    },
+    {
+      "epoch": 0.8793727683587605,
+      "grad_norm": 0.08608123660087585,
+      "learning_rate": 3.627875541315177e-06,
+      "loss": 8.5201,
+      "step": 176090
+    },
+    {
+      "epoch": 0.8794227071837,
+      "grad_norm": 0.09179461002349854,
+      "learning_rate": 3.6263736263736266e-06,
+      "loss": 8.5454,
+      "step": 176100
+    },
+    {
+      "epoch": 0.8794726460086394,
+      "grad_norm": 0.09087244421243668,
+      "learning_rate": 3.624871711432076e-06,
+      "loss": 8.5105,
+      "step": 176110
+    },
+    {
+      "epoch": 0.8795225848335788,
+      "grad_norm": 0.09345267713069916,
+      "learning_rate": 3.6233697964905253e-06,
+      "loss": 8.5111,
+      "step": 176120
+    },
+    {
+      "epoch": 0.8795725236585183,
+      "grad_norm": 0.09425700455904007,
+      "learning_rate": 3.621867881548975e-06,
+      "loss": 8.5227,
+      "step": 176130
+    },
+    {
+      "epoch": 0.8796224624834578,
+      "grad_norm": 0.08979269117116928,
+      "learning_rate": 3.6203659666074248e-06,
+      "loss": 8.5186,
+      "step": 176140
+    },
+    {
+      "epoch": 0.8796724013083972,
+      "grad_norm": 0.09118253737688065,
+      "learning_rate": 3.618864051665874e-06,
+      "loss": 8.5299,
+      "step": 176150
+    },
+    {
+      "epoch": 0.8797223401333366,
+      "grad_norm": 0.09220730513334274,
+      "learning_rate": 3.6173621367243234e-06,
+      "loss": 8.5191,
+      "step": 176160
+    },
+    {
+      "epoch": 0.8797722789582761,
+      "grad_norm": 0.08798353374004364,
+      "learning_rate": 3.615860221782773e-06,
+      "loss": 8.5048,
+      "step": 176170
+    },
+    {
+      "epoch": 0.8798222177832156,
+      "grad_norm": 0.0894918367266655,
+      "learning_rate": 3.6143583068412225e-06,
+      "loss": 8.5088,
+      "step": 176180
+    },
+    {
+      "epoch": 0.879872156608155,
+      "grad_norm": 0.09121794253587723,
+      "learning_rate": 3.6128563918996723e-06,
+      "loss": 8.5221,
+      "step": 176190
+    },
+    {
+      "epoch": 0.8799220954330944,
+      "grad_norm": 0.09195500612258911,
+      "learning_rate": 3.6113544769581216e-06,
+      "loss": 8.5301,
+      "step": 176200
+    },
+    {
+      "epoch": 0.8799720342580339,
+      "grad_norm": 0.08938493579626083,
+      "learning_rate": 3.6098525620165714e-06,
+      "loss": 8.5136,
+      "step": 176210
+    },
+    {
+      "epoch": 0.8800219730829734,
+      "grad_norm": 0.08818230032920837,
+      "learning_rate": 3.6083506470750207e-06,
+      "loss": 8.5135,
+      "step": 176220
+    },
+    {
+      "epoch": 0.8800719119079128,
+      "grad_norm": 0.09114837646484375,
+      "learning_rate": 3.60684873213347e-06,
+      "loss": 8.5182,
+      "step": 176230
+    },
+    {
+      "epoch": 0.8801218507328522,
+      "grad_norm": 0.09198463708162308,
+      "learning_rate": 3.60534681719192e-06,
+      "loss": 8.5281,
+      "step": 176240
+    },
+    {
+      "epoch": 0.8801717895577917,
+      "grad_norm": 0.08858832716941833,
+      "learning_rate": 3.6038449022503695e-06,
+      "loss": 8.5173,
+      "step": 176250
+    },
+    {
+      "epoch": 0.8802217283827312,
+      "grad_norm": 0.09060419350862503,
+      "learning_rate": 3.602342987308819e-06,
+      "loss": 8.5146,
+      "step": 176260
+    },
+    {
+      "epoch": 0.8802716672076706,
+      "grad_norm": 0.09478036314249039,
+      "learning_rate": 3.600841072367268e-06,
+      "loss": 8.5338,
+      "step": 176270
+    },
+    {
+      "epoch": 0.88032160603261,
+      "grad_norm": 0.09485729038715363,
+      "learning_rate": 3.5993391574257175e-06,
+      "loss": 8.5161,
+      "step": 176280
+    },
+    {
+      "epoch": 0.8803715448575495,
+      "grad_norm": 0.08923057466745377,
+      "learning_rate": 3.5978372424841677e-06,
+      "loss": 8.512,
+      "step": 176290
+    },
+    {
+      "epoch": 0.880421483682489,
+      "grad_norm": 0.09072960168123245,
+      "learning_rate": 3.596335327542617e-06,
+      "loss": 8.5052,
+      "step": 176300
+    },
+    {
+      "epoch": 0.8804714225074284,
+      "grad_norm": 0.08910852670669556,
+      "learning_rate": 3.5948334126010664e-06,
+      "loss": 8.511,
+      "step": 176310
+    },
+    {
+      "epoch": 0.8805213613323678,
+      "grad_norm": 0.09066713601350784,
+      "learning_rate": 3.5933314976595157e-06,
+      "loss": 8.5136,
+      "step": 176320
+    },
+    {
+      "epoch": 0.8805713001573073,
+      "grad_norm": 0.08695919066667557,
+      "learning_rate": 3.5918295827179655e-06,
+      "loss": 8.5367,
+      "step": 176330
+    },
+    {
+      "epoch": 0.8806212389822468,
+      "grad_norm": 0.08876381814479828,
+      "learning_rate": 3.5903276677764152e-06,
+      "loss": 8.5279,
+      "step": 176340
+    },
+    {
+      "epoch": 0.8806711778071862,
+      "grad_norm": 0.09524772316217422,
+      "learning_rate": 3.5888257528348646e-06,
+      "loss": 8.5178,
+      "step": 176350
+    },
+    {
+      "epoch": 0.8807211166321256,
+      "grad_norm": 0.0915902629494667,
+      "learning_rate": 3.587323837893314e-06,
+      "loss": 8.5005,
+      "step": 176360
+    },
+    {
+      "epoch": 0.8807710554570651,
+      "grad_norm": 0.091648168861866,
+      "learning_rate": 3.5858219229517636e-06,
+      "loss": 8.522,
+      "step": 176370
+    },
+    {
+      "epoch": 0.8808209942820046,
+      "grad_norm": 0.08661641925573349,
+      "learning_rate": 3.584320008010213e-06,
+      "loss": 8.5298,
+      "step": 176380
+    },
+    {
+      "epoch": 0.880870933106944,
+      "grad_norm": 0.09084923565387726,
+      "learning_rate": 3.5828180930686627e-06,
+      "loss": 8.5104,
+      "step": 176390
+    },
+    {
+      "epoch": 0.8809208719318834,
+      "grad_norm": 0.09532788395881653,
+      "learning_rate": 3.581316178127112e-06,
+      "loss": 8.5164,
+      "step": 176400
+    },
+    {
+      "epoch": 0.8809708107568229,
+      "grad_norm": 0.09041056036949158,
+      "learning_rate": 3.579814263185562e-06,
+      "loss": 8.5083,
+      "step": 176410
+    },
+    {
+      "epoch": 0.8810207495817624,
+      "grad_norm": 0.0913330689072609,
+      "learning_rate": 3.578312348244011e-06,
+      "loss": 8.526,
+      "step": 176420
+    },
+    {
+      "epoch": 0.8810706884067018,
+      "grad_norm": 0.09379500150680542,
+      "learning_rate": 3.5768104333024605e-06,
+      "loss": 8.5027,
+      "step": 176430
+    },
+    {
+      "epoch": 0.8811206272316412,
+      "grad_norm": 0.09111961722373962,
+      "learning_rate": 3.5753085183609102e-06,
+      "loss": 8.5041,
+      "step": 176440
+    },
+    {
+      "epoch": 0.8811705660565807,
+      "grad_norm": 0.09450194984674454,
+      "learning_rate": 3.57380660341936e-06,
+      "loss": 8.5164,
+      "step": 176450
+    },
+    {
+      "epoch": 0.8812205048815202,
+      "grad_norm": 0.08781801164150238,
+      "learning_rate": 3.5723046884778093e-06,
+      "loss": 8.5286,
+      "step": 176460
+    },
+    {
+      "epoch": 0.8812704437064596,
+      "grad_norm": 0.08742903918027878,
+      "learning_rate": 3.5708027735362586e-06,
+      "loss": 8.5282,
+      "step": 176470
+    },
+    {
+      "epoch": 0.881320382531399,
+      "grad_norm": 0.09143572300672531,
+      "learning_rate": 3.569300858594708e-06,
+      "loss": 8.5142,
+      "step": 176480
+    },
+    {
+      "epoch": 0.8813703213563385,
+      "grad_norm": 0.0917043462395668,
+      "learning_rate": 3.567798943653158e-06,
+      "loss": 8.5381,
+      "step": 176490
+    },
+    {
+      "epoch": 0.881420260181278,
+      "grad_norm": 0.08833152055740356,
+      "learning_rate": 3.5662970287116075e-06,
+      "loss": 8.5399,
+      "step": 176500
+    },
+    {
+      "epoch": 0.8814701990062174,
+      "grad_norm": 0.09015662968158722,
+      "learning_rate": 3.564795113770057e-06,
+      "loss": 8.5365,
+      "step": 176510
+    },
+    {
+      "epoch": 0.8815201378311568,
+      "grad_norm": 0.0858858972787857,
+      "learning_rate": 3.563293198828506e-06,
+      "loss": 8.5222,
+      "step": 176520
+    },
+    {
+      "epoch": 0.8815700766560963,
+      "grad_norm": 0.08976481109857559,
+      "learning_rate": 3.561791283886956e-06,
+      "loss": 8.5183,
+      "step": 176530
+    },
+    {
+      "epoch": 0.8816200154810357,
+      "grad_norm": 0.08899136632680893,
+      "learning_rate": 3.5602893689454057e-06,
+      "loss": 8.5094,
+      "step": 176540
+    },
+    {
+      "epoch": 0.8816699543059752,
+      "grad_norm": 0.09167985618114471,
+      "learning_rate": 3.558787454003855e-06,
+      "loss": 8.5039,
+      "step": 176550
+    },
+    {
+      "epoch": 0.8817198931309146,
+      "grad_norm": 0.09655990451574326,
+      "learning_rate": 3.5572855390623043e-06,
+      "loss": 8.5208,
+      "step": 176560
+    },
+    {
+      "epoch": 0.8817698319558541,
+      "grad_norm": 0.0897342711687088,
+      "learning_rate": 3.555783624120754e-06,
+      "loss": 8.511,
+      "step": 176570
+    },
+    {
+      "epoch": 0.8818197707807935,
+      "grad_norm": 0.08929044008255005,
+      "learning_rate": 3.5542817091792034e-06,
+      "loss": 8.5278,
+      "step": 176580
+    },
+    {
+      "epoch": 0.881869709605733,
+      "grad_norm": 0.09658628702163696,
+      "learning_rate": 3.552779794237653e-06,
+      "loss": 8.501,
+      "step": 176590
+    },
+    {
+      "epoch": 0.8819196484306724,
+      "grad_norm": 0.09207036346197128,
+      "learning_rate": 3.5512778792961025e-06,
+      "loss": 8.5236,
+      "step": 176600
+    },
+    {
+      "epoch": 0.8819695872556119,
+      "grad_norm": 0.09149431437253952,
+      "learning_rate": 3.5497759643545523e-06,
+      "loss": 8.5151,
+      "step": 176610
+    },
+    {
+      "epoch": 0.8820195260805513,
+      "grad_norm": 0.08894215524196625,
+      "learning_rate": 3.5482740494130016e-06,
+      "loss": 8.5278,
+      "step": 176620
+    },
+    {
+      "epoch": 0.8820694649054908,
+      "grad_norm": 0.09006328135728836,
+      "learning_rate": 3.5467721344714513e-06,
+      "loss": 8.5218,
+      "step": 176630
+    },
+    {
+      "epoch": 0.8821194037304302,
+      "grad_norm": 0.09437132626771927,
+      "learning_rate": 3.5452702195299007e-06,
+      "loss": 8.5154,
+      "step": 176640
+    },
+    {
+      "epoch": 0.8821693425553697,
+      "grad_norm": 0.09077361971139908,
+      "learning_rate": 3.5437683045883504e-06,
+      "loss": 8.5006,
+      "step": 176650
+    },
+    {
+      "epoch": 0.8822192813803091,
+      "grad_norm": 0.08984161168336868,
+      "learning_rate": 3.5422663896467998e-06,
+      "loss": 8.5379,
+      "step": 176660
+    },
+    {
+      "epoch": 0.8822692202052486,
+      "grad_norm": 0.09607677906751633,
+      "learning_rate": 3.540764474705249e-06,
+      "loss": 8.525,
+      "step": 176670
+    },
+    {
+      "epoch": 0.882319159030188,
+      "grad_norm": 0.09277524054050446,
+      "learning_rate": 3.539262559763699e-06,
+      "loss": 8.5213,
+      "step": 176680
+    },
+    {
+      "epoch": 0.8823690978551275,
+      "grad_norm": 0.09065360575914383,
+      "learning_rate": 3.5377606448221486e-06,
+      "loss": 8.5276,
+      "step": 176690
+    },
+    {
+      "epoch": 0.8824190366800669,
+      "grad_norm": 0.10031012445688248,
+      "learning_rate": 3.536258729880598e-06,
+      "loss": 8.4984,
+      "step": 176700
+    },
+    {
+      "epoch": 0.8824689755050064,
+      "grad_norm": 0.08632688969373703,
+      "learning_rate": 3.5347568149390473e-06,
+      "loss": 8.549,
+      "step": 176710
+    },
+    {
+      "epoch": 0.8825189143299458,
+      "grad_norm": 0.09275210648775101,
+      "learning_rate": 3.5332548999974966e-06,
+      "loss": 8.5149,
+      "step": 176720
+    },
+    {
+      "epoch": 0.8825688531548853,
+      "grad_norm": 0.08953607082366943,
+      "learning_rate": 3.5317529850559468e-06,
+      "loss": 8.5107,
+      "step": 176730
+    },
+    {
+      "epoch": 0.8826187919798247,
+      "grad_norm": 0.09068858623504639,
+      "learning_rate": 3.530251070114396e-06,
+      "loss": 8.5048,
+      "step": 176740
+    },
+    {
+      "epoch": 0.8826687308047642,
+      "grad_norm": 0.08825194090604782,
+      "learning_rate": 3.5287491551728454e-06,
+      "loss": 8.5169,
+      "step": 176750
+    },
+    {
+      "epoch": 0.8827186696297036,
+      "grad_norm": 0.08667933940887451,
+      "learning_rate": 3.5272472402312948e-06,
+      "loss": 8.5104,
+      "step": 176760
+    },
+    {
+      "epoch": 0.8827686084546431,
+      "grad_norm": 0.0986347422003746,
+      "learning_rate": 3.5257453252897445e-06,
+      "loss": 8.5012,
+      "step": 176770
+    },
+    {
+      "epoch": 0.8828185472795825,
+      "grad_norm": 0.09399598091840744,
+      "learning_rate": 3.5242434103481943e-06,
+      "loss": 8.5245,
+      "step": 176780
+    },
+    {
+      "epoch": 0.882868486104522,
+      "grad_norm": 0.0902981236577034,
+      "learning_rate": 3.5227414954066436e-06,
+      "loss": 8.5248,
+      "step": 176790
+    },
+    {
+      "epoch": 0.8829184249294614,
+      "grad_norm": 0.09248172491788864,
+      "learning_rate": 3.521239580465093e-06,
+      "loss": 8.526,
+      "step": 176800
+    },
+    {
+      "epoch": 0.8829683637544009,
+      "grad_norm": 0.08781624585390091,
+      "learning_rate": 3.5197376655235427e-06,
+      "loss": 8.5162,
+      "step": 176810
+    },
+    {
+      "epoch": 0.8830183025793403,
+      "grad_norm": 0.09201288223266602,
+      "learning_rate": 3.518235750581992e-06,
+      "loss": 8.5082,
+      "step": 176820
+    },
+    {
+      "epoch": 0.8830682414042798,
+      "grad_norm": 0.0937962606549263,
+      "learning_rate": 3.5167338356404418e-06,
+      "loss": 8.5262,
+      "step": 176830
+    },
+    {
+      "epoch": 0.8831181802292192,
+      "grad_norm": 0.08981435745954514,
+      "learning_rate": 3.515231920698891e-06,
+      "loss": 8.5141,
+      "step": 176840
+    },
+    {
+      "epoch": 0.8831681190541587,
+      "grad_norm": 0.09788450598716736,
+      "learning_rate": 3.513730005757341e-06,
+      "loss": 8.5209,
+      "step": 176850
+    },
+    {
+      "epoch": 0.8832180578790981,
+      "grad_norm": 0.09376143664121628,
+      "learning_rate": 3.51222809081579e-06,
+      "loss": 8.5123,
+      "step": 176860
+    },
+    {
+      "epoch": 0.8832679967040376,
+      "grad_norm": 0.09434860199689865,
+      "learning_rate": 3.5107261758742395e-06,
+      "loss": 8.5131,
+      "step": 176870
+    },
+    {
+      "epoch": 0.883317935528977,
+      "grad_norm": 0.09667465090751648,
+      "learning_rate": 3.5092242609326893e-06,
+      "loss": 8.5213,
+      "step": 176880
+    },
+    {
+      "epoch": 0.8833678743539165,
+      "grad_norm": 0.09229189902544022,
+      "learning_rate": 3.507722345991139e-06,
+      "loss": 8.5217,
+      "step": 176890
+    },
+    {
+      "epoch": 0.8834178131788559,
+      "grad_norm": 0.08936712145805359,
+      "learning_rate": 3.5062204310495884e-06,
+      "loss": 8.5231,
+      "step": 176900
+    },
+    {
+      "epoch": 0.8834677520037953,
+      "grad_norm": 0.08999984711408615,
+      "learning_rate": 3.5047185161080377e-06,
+      "loss": 8.5265,
+      "step": 176910
+    },
+    {
+      "epoch": 0.8835176908287348,
+      "grad_norm": 0.08811944723129272,
+      "learning_rate": 3.503216601166487e-06,
+      "loss": 8.5266,
+      "step": 176920
+    },
+    {
+      "epoch": 0.8835676296536743,
+      "grad_norm": 0.0963667631149292,
+      "learning_rate": 3.501714686224937e-06,
+      "loss": 8.5098,
+      "step": 176930
+    },
+    {
+      "epoch": 0.8836175684786137,
+      "grad_norm": 0.09169621020555496,
+      "learning_rate": 3.5002127712833865e-06,
+      "loss": 8.5253,
+      "step": 176940
+    },
+    {
+      "epoch": 0.8836675073035531,
+      "grad_norm": 0.095432348549366,
+      "learning_rate": 3.498710856341836e-06,
+      "loss": 8.5165,
+      "step": 176950
+    },
+    {
+      "epoch": 0.8837174461284926,
+      "grad_norm": 0.09653273224830627,
+      "learning_rate": 3.497208941400285e-06,
+      "loss": 8.5344,
+      "step": 176960
+    },
+    {
+      "epoch": 0.8837673849534321,
+      "grad_norm": 0.09007629007101059,
+      "learning_rate": 3.495707026458735e-06,
+      "loss": 8.5275,
+      "step": 176970
+    },
+    {
+      "epoch": 0.8838173237783715,
+      "grad_norm": 0.09573765844106674,
+      "learning_rate": 3.4942051115171847e-06,
+      "loss": 8.5092,
+      "step": 176980
+    },
+    {
+      "epoch": 0.8838672626033109,
+      "grad_norm": 0.09132372587919235,
+      "learning_rate": 3.492703196575634e-06,
+      "loss": 8.518,
+      "step": 176990
+    },
+    {
+      "epoch": 0.8839172014282504,
+      "grad_norm": 0.09207233786582947,
+      "learning_rate": 3.4912012816340834e-06,
+      "loss": 8.5078,
+      "step": 177000
+    },
+    {
+      "epoch": 0.8839671402531898,
+      "grad_norm": 0.09270022809505463,
+      "learning_rate": 3.489699366692533e-06,
+      "loss": 8.549,
+      "step": 177010
+    },
+    {
+      "epoch": 0.8840170790781293,
+      "grad_norm": 0.09225805103778839,
+      "learning_rate": 3.4881974517509825e-06,
+      "loss": 8.518,
+      "step": 177020
+    },
+    {
+      "epoch": 0.8840670179030687,
+      "grad_norm": 0.092227041721344,
+      "learning_rate": 3.4866955368094322e-06,
+      "loss": 8.5222,
+      "step": 177030
+    },
+    {
+      "epoch": 0.8841169567280082,
+      "grad_norm": 0.08810020238161087,
+      "learning_rate": 3.4851936218678815e-06,
+      "loss": 8.5089,
+      "step": 177040
+    },
+    {
+      "epoch": 0.8841668955529476,
+      "grad_norm": 0.08814115077257156,
+      "learning_rate": 3.4836917069263313e-06,
+      "loss": 8.5423,
+      "step": 177050
+    },
+    {
+      "epoch": 0.8842168343778871,
+      "grad_norm": 0.09531360119581223,
+      "learning_rate": 3.4821897919847806e-06,
+      "loss": 8.5245,
+      "step": 177060
+    },
+    {
+      "epoch": 0.8842667732028265,
+      "grad_norm": 0.08866927027702332,
+      "learning_rate": 3.48068787704323e-06,
+      "loss": 8.5009,
+      "step": 177070
+    },
+    {
+      "epoch": 0.884316712027766,
+      "grad_norm": 0.08944374322891235,
+      "learning_rate": 3.4791859621016797e-06,
+      "loss": 8.5293,
+      "step": 177080
+    },
+    {
+      "epoch": 0.8843666508527054,
+      "grad_norm": 0.08847851306200027,
+      "learning_rate": 3.4776840471601295e-06,
+      "loss": 8.5336,
+      "step": 177090
+    },
+    {
+      "epoch": 0.8844165896776449,
+      "grad_norm": 0.0904935672879219,
+      "learning_rate": 3.476182132218579e-06,
+      "loss": 8.5157,
+      "step": 177100
+    },
+    {
+      "epoch": 0.8844665285025843,
+      "grad_norm": 0.0903678685426712,
+      "learning_rate": 3.474680217277028e-06,
+      "loss": 8.517,
+      "step": 177110
+    },
+    {
+      "epoch": 0.8845164673275238,
+      "grad_norm": 0.09201235324144363,
+      "learning_rate": 3.4731783023354775e-06,
+      "loss": 8.5221,
+      "step": 177120
+    },
+    {
+      "epoch": 0.8845664061524632,
+      "grad_norm": 0.09328807145357132,
+      "learning_rate": 3.4716763873939276e-06,
+      "loss": 8.5328,
+      "step": 177130
+    },
+    {
+      "epoch": 0.8846163449774027,
+      "grad_norm": 0.08826225250959396,
+      "learning_rate": 3.470174472452377e-06,
+      "loss": 8.5153,
+      "step": 177140
+    },
+    {
+      "epoch": 0.8846662838023421,
+      "grad_norm": 0.09613511711359024,
+      "learning_rate": 3.4686725575108263e-06,
+      "loss": 8.5224,
+      "step": 177150
+    },
+    {
+      "epoch": 0.8847162226272816,
+      "grad_norm": 0.0888667032122612,
+      "learning_rate": 3.4671706425692756e-06,
+      "loss": 8.5245,
+      "step": 177160
+    },
+    {
+      "epoch": 0.884766161452221,
+      "grad_norm": 0.08986398577690125,
+      "learning_rate": 3.4656687276277254e-06,
+      "loss": 8.5181,
+      "step": 177170
+    },
+    {
+      "epoch": 0.8848161002771605,
+      "grad_norm": 0.09408170729875565,
+      "learning_rate": 3.464166812686175e-06,
+      "loss": 8.5075,
+      "step": 177180
+    },
+    {
+      "epoch": 0.8848660391020999,
+      "grad_norm": 0.08650843799114227,
+      "learning_rate": 3.4626648977446245e-06,
+      "loss": 8.5356,
+      "step": 177190
+    },
+    {
+      "epoch": 0.8849159779270394,
+      "grad_norm": 0.09117493778467178,
+      "learning_rate": 3.461162982803074e-06,
+      "loss": 8.5275,
+      "step": 177200
+    },
+    {
+      "epoch": 0.8849659167519788,
+      "grad_norm": 0.09649976342916489,
+      "learning_rate": 3.4596610678615236e-06,
+      "loss": 8.5083,
+      "step": 177210
+    },
+    {
+      "epoch": 0.8850158555769183,
+      "grad_norm": 0.09083914011716843,
+      "learning_rate": 3.4581591529199733e-06,
+      "loss": 8.5271,
+      "step": 177220
+    },
+    {
+      "epoch": 0.8850657944018577,
+      "grad_norm": 0.0903511494398117,
+      "learning_rate": 3.4566572379784227e-06,
+      "loss": 8.5062,
+      "step": 177230
+    },
+    {
+      "epoch": 0.8851157332267972,
+      "grad_norm": 0.09016105532646179,
+      "learning_rate": 3.455155323036872e-06,
+      "loss": 8.5287,
+      "step": 177240
+    },
+    {
+      "epoch": 0.8851656720517366,
+      "grad_norm": 0.09000012278556824,
+      "learning_rate": 3.4536534080953217e-06,
+      "loss": 8.4927,
+      "step": 177250
+    },
+    {
+      "epoch": 0.8852156108766761,
+      "grad_norm": 0.088304802775383,
+      "learning_rate": 3.452151493153771e-06,
+      "loss": 8.5076,
+      "step": 177260
+    },
+    {
+      "epoch": 0.8852655497016155,
+      "grad_norm": 0.09251948446035385,
+      "learning_rate": 3.450649578212221e-06,
+      "loss": 8.5103,
+      "step": 177270
+    },
+    {
+      "epoch": 0.885315488526555,
+      "grad_norm": 0.08819442987442017,
+      "learning_rate": 3.44914766327067e-06,
+      "loss": 8.5039,
+      "step": 177280
+    },
+    {
+      "epoch": 0.8853654273514944,
+      "grad_norm": 0.0890979915857315,
+      "learning_rate": 3.44764574832912e-06,
+      "loss": 8.5171,
+      "step": 177290
+    },
+    {
+      "epoch": 0.8854153661764339,
+      "grad_norm": 0.0990390032529831,
+      "learning_rate": 3.4461438333875692e-06,
+      "loss": 8.52,
+      "step": 177300
+    },
+    {
+      "epoch": 0.8854653050013733,
+      "grad_norm": 0.08952021598815918,
+      "learning_rate": 3.4446419184460186e-06,
+      "loss": 8.5213,
+      "step": 177310
+    },
+    {
+      "epoch": 0.8855152438263127,
+      "grad_norm": 0.08724192529916763,
+      "learning_rate": 3.4431400035044683e-06,
+      "loss": 8.5327,
+      "step": 177320
+    },
+    {
+      "epoch": 0.8855651826512522,
+      "grad_norm": 0.09806641191244125,
+      "learning_rate": 3.441638088562918e-06,
+      "loss": 8.5187,
+      "step": 177330
+    },
+    {
+      "epoch": 0.8856151214761917,
+      "grad_norm": 0.08996843546628952,
+      "learning_rate": 3.4401361736213674e-06,
+      "loss": 8.5578,
+      "step": 177340
+    },
+    {
+      "epoch": 0.8856650603011311,
+      "grad_norm": 0.09567862004041672,
+      "learning_rate": 3.4386342586798167e-06,
+      "loss": 8.5224,
+      "step": 177350
+    },
+    {
+      "epoch": 0.8857149991260705,
+      "grad_norm": 0.08767600357532501,
+      "learning_rate": 3.437132343738266e-06,
+      "loss": 8.5191,
+      "step": 177360
+    },
+    {
+      "epoch": 0.88576493795101,
+      "grad_norm": 0.09355778992176056,
+      "learning_rate": 3.4356304287967163e-06,
+      "loss": 8.5066,
+      "step": 177370
+    },
+    {
+      "epoch": 0.8858148767759495,
+      "grad_norm": 0.09189339727163315,
+      "learning_rate": 3.4341285138551656e-06,
+      "loss": 8.5076,
+      "step": 177380
+    },
+    {
+      "epoch": 0.8858648156008889,
+      "grad_norm": 0.08906829357147217,
+      "learning_rate": 3.432626598913615e-06,
+      "loss": 8.5052,
+      "step": 177390
+    },
+    {
+      "epoch": 0.8859147544258283,
+      "grad_norm": 0.08834553509950638,
+      "learning_rate": 3.4311246839720643e-06,
+      "loss": 8.5115,
+      "step": 177400
+    },
+    {
+      "epoch": 0.8859646932507678,
+      "grad_norm": 0.09251053631305695,
+      "learning_rate": 3.429622769030514e-06,
+      "loss": 8.5233,
+      "step": 177410
+    },
+    {
+      "epoch": 0.8860146320757073,
+      "grad_norm": 0.08647915720939636,
+      "learning_rate": 3.4281208540889638e-06,
+      "loss": 8.5078,
+      "step": 177420
+    },
+    {
+      "epoch": 0.8860645709006467,
+      "grad_norm": 0.09425395727157593,
+      "learning_rate": 3.426618939147413e-06,
+      "loss": 8.5261,
+      "step": 177430
+    },
+    {
+      "epoch": 0.8861145097255861,
+      "grad_norm": 0.08929485827684402,
+      "learning_rate": 3.4251170242058624e-06,
+      "loss": 8.5276,
+      "step": 177440
+    },
+    {
+      "epoch": 0.8861644485505256,
+      "grad_norm": 0.09134992957115173,
+      "learning_rate": 3.423615109264312e-06,
+      "loss": 8.531,
+      "step": 177450
+    },
+    {
+      "epoch": 0.8862143873754651,
+      "grad_norm": 0.08963406831026077,
+      "learning_rate": 3.4221131943227615e-06,
+      "loss": 8.517,
+      "step": 177460
+    },
+    {
+      "epoch": 0.8862643262004045,
+      "grad_norm": 0.09403959661722183,
+      "learning_rate": 3.4206112793812113e-06,
+      "loss": 8.5015,
+      "step": 177470
+    },
+    {
+      "epoch": 0.8863142650253439,
+      "grad_norm": 0.08884259313344955,
+      "learning_rate": 3.4191093644396606e-06,
+      "loss": 8.511,
+      "step": 177480
+    },
+    {
+      "epoch": 0.8863642038502834,
+      "grad_norm": 0.09389010816812515,
+      "learning_rate": 3.4176074494981104e-06,
+      "loss": 8.4983,
+      "step": 177490
+    },
+    {
+      "epoch": 0.8864141426752229,
+      "grad_norm": 0.08853202313184738,
+      "learning_rate": 3.4161055345565597e-06,
+      "loss": 8.5133,
+      "step": 177500
+    },
+    {
+      "epoch": 0.8864640815001623,
+      "grad_norm": 0.09294895827770233,
+      "learning_rate": 3.414603619615009e-06,
+      "loss": 8.5259,
+      "step": 177510
+    },
+    {
+      "epoch": 0.8865140203251017,
+      "grad_norm": 0.09223725646734238,
+      "learning_rate": 3.4131017046734588e-06,
+      "loss": 8.5059,
+      "step": 177520
+    },
+    {
+      "epoch": 0.8865639591500412,
+      "grad_norm": 0.09366583824157715,
+      "learning_rate": 3.4115997897319085e-06,
+      "loss": 8.5378,
+      "step": 177530
+    },
+    {
+      "epoch": 0.8866138979749807,
+      "grad_norm": 0.08734884113073349,
+      "learning_rate": 3.410097874790358e-06,
+      "loss": 8.5117,
+      "step": 177540
+    },
+    {
+      "epoch": 0.8866638367999201,
+      "grad_norm": 0.09199695289134979,
+      "learning_rate": 3.408595959848807e-06,
+      "loss": 8.5281,
+      "step": 177550
+    },
+    {
+      "epoch": 0.8867137756248595,
+      "grad_norm": 0.09172987192869186,
+      "learning_rate": 3.4070940449072565e-06,
+      "loss": 8.523,
+      "step": 177560
+    },
+    {
+      "epoch": 0.886763714449799,
+      "grad_norm": 0.09687057882547379,
+      "learning_rate": 3.4055921299657067e-06,
+      "loss": 8.532,
+      "step": 177570
+    },
+    {
+      "epoch": 0.8868136532747385,
+      "grad_norm": 0.09877201914787292,
+      "learning_rate": 3.404090215024156e-06,
+      "loss": 8.534,
+      "step": 177580
+    },
+    {
+      "epoch": 0.8868635920996779,
+      "grad_norm": 0.09122820198535919,
+      "learning_rate": 3.4025883000826054e-06,
+      "loss": 8.4999,
+      "step": 177590
+    },
+    {
+      "epoch": 0.8869135309246173,
+      "grad_norm": 0.09195327013731003,
+      "learning_rate": 3.4010863851410547e-06,
+      "loss": 8.5035,
+      "step": 177600
+    },
+    {
+      "epoch": 0.8869634697495568,
+      "grad_norm": 0.09369145333766937,
+      "learning_rate": 3.3995844701995044e-06,
+      "loss": 8.4957,
+      "step": 177610
+    },
+    {
+      "epoch": 0.8870134085744963,
+      "grad_norm": 0.09592372179031372,
+      "learning_rate": 3.398082555257954e-06,
+      "loss": 8.5032,
+      "step": 177620
+    },
+    {
+      "epoch": 0.8870633473994357,
+      "grad_norm": 0.09076791256666183,
+      "learning_rate": 3.3965806403164035e-06,
+      "loss": 8.5275,
+      "step": 177630
+    },
+    {
+      "epoch": 0.8871132862243751,
+      "grad_norm": 0.0895276591181755,
+      "learning_rate": 3.395078725374853e-06,
+      "loss": 8.5322,
+      "step": 177640
+    },
+    {
+      "epoch": 0.8871632250493146,
+      "grad_norm": 0.09077776968479156,
+      "learning_rate": 3.3935768104333026e-06,
+      "loss": 8.5201,
+      "step": 177650
+    },
+    {
+      "epoch": 0.8872131638742541,
+      "grad_norm": 0.09358912706375122,
+      "learning_rate": 3.392074895491752e-06,
+      "loss": 8.5083,
+      "step": 177660
+    },
+    {
+      "epoch": 0.8872631026991935,
+      "grad_norm": 0.09124220907688141,
+      "learning_rate": 3.3905729805502017e-06,
+      "loss": 8.514,
+      "step": 177670
+    },
+    {
+      "epoch": 0.8873130415241329,
+      "grad_norm": 0.09009285271167755,
+      "learning_rate": 3.389071065608651e-06,
+      "loss": 8.5199,
+      "step": 177680
+    },
+    {
+      "epoch": 0.8873629803490723,
+      "grad_norm": 0.08989504724740982,
+      "learning_rate": 3.3875691506671008e-06,
+      "loss": 8.525,
+      "step": 177690
+    },
+    {
+      "epoch": 0.8874129191740119,
+      "grad_norm": 0.09324032813310623,
+      "learning_rate": 3.38606723572555e-06,
+      "loss": 8.5225,
+      "step": 177700
+    },
+    {
+      "epoch": 0.8874628579989513,
+      "grad_norm": 0.09039938449859619,
+      "learning_rate": 3.3845653207839995e-06,
+      "loss": 8.5127,
+      "step": 177710
+    },
+    {
+      "epoch": 0.8875127968238907,
+      "grad_norm": 0.09255453944206238,
+      "learning_rate": 3.383063405842449e-06,
+      "loss": 8.5293,
+      "step": 177720
+    },
+    {
+      "epoch": 0.8875627356488301,
+      "grad_norm": 0.0964113101363182,
+      "learning_rate": 3.381561490900899e-06,
+      "loss": 8.5023,
+      "step": 177730
+    },
+    {
+      "epoch": 0.8876126744737697,
+      "grad_norm": 0.09030985832214355,
+      "learning_rate": 3.3800595759593483e-06,
+      "loss": 8.5003,
+      "step": 177740
+    },
+    {
+      "epoch": 0.8876626132987091,
+      "grad_norm": 0.08826965093612671,
+      "learning_rate": 3.3785576610177976e-06,
+      "loss": 8.5222,
+      "step": 177750
+    },
+    {
+      "epoch": 0.8877125521236485,
+      "grad_norm": 0.09713491797447205,
+      "learning_rate": 3.3770557460762474e-06,
+      "loss": 8.5149,
+      "step": 177760
+    },
+    {
+      "epoch": 0.887762490948588,
+      "grad_norm": 0.09569582343101501,
+      "learning_rate": 3.375553831134697e-06,
+      "loss": 8.5064,
+      "step": 177770
+    },
+    {
+      "epoch": 0.8878124297735275,
+      "grad_norm": 0.09315156936645508,
+      "learning_rate": 3.3740519161931465e-06,
+      "loss": 8.5055,
+      "step": 177780
+    },
+    {
+      "epoch": 0.8878623685984669,
+      "grad_norm": 0.09295808523893356,
+      "learning_rate": 3.372550001251596e-06,
+      "loss": 8.5233,
+      "step": 177790
+    },
+    {
+      "epoch": 0.8879123074234063,
+      "grad_norm": 0.08866126090288162,
+      "learning_rate": 3.371048086310045e-06,
+      "loss": 8.5343,
+      "step": 177800
+    },
+    {
+      "epoch": 0.8879622462483457,
+      "grad_norm": 0.09473447501659393,
+      "learning_rate": 3.3695461713684953e-06,
+      "loss": 8.5097,
+      "step": 177810
+    },
+    {
+      "epoch": 0.8880121850732853,
+      "grad_norm": 0.09259150177240372,
+      "learning_rate": 3.3680442564269446e-06,
+      "loss": 8.5188,
+      "step": 177820
+    },
+    {
+      "epoch": 0.8880621238982247,
+      "grad_norm": 0.09014898538589478,
+      "learning_rate": 3.366542341485394e-06,
+      "loss": 8.5231,
+      "step": 177830
+    },
+    {
+      "epoch": 0.8881120627231641,
+      "grad_norm": 0.09090849757194519,
+      "learning_rate": 3.3650404265438433e-06,
+      "loss": 8.5345,
+      "step": 177840
+    },
+    {
+      "epoch": 0.8881620015481035,
+      "grad_norm": 0.09438339620828629,
+      "learning_rate": 3.363538511602293e-06,
+      "loss": 8.5103,
+      "step": 177850
+    },
+    {
+      "epoch": 0.8882119403730431,
+      "grad_norm": 0.09024190902709961,
+      "learning_rate": 3.362036596660743e-06,
+      "loss": 8.5295,
+      "step": 177860
+    },
+    {
+      "epoch": 0.8882618791979825,
+      "grad_norm": 0.0944947823882103,
+      "learning_rate": 3.360534681719192e-06,
+      "loss": 8.5038,
+      "step": 177870
+    },
+    {
+      "epoch": 0.8883118180229219,
+      "grad_norm": 0.09072814136743546,
+      "learning_rate": 3.3590327667776415e-06,
+      "loss": 8.5217,
+      "step": 177880
+    },
+    {
+      "epoch": 0.8883617568478613,
+      "grad_norm": 0.09160330891609192,
+      "learning_rate": 3.3575308518360912e-06,
+      "loss": 8.4917,
+      "step": 177890
+    },
+    {
+      "epoch": 0.8884116956728009,
+      "grad_norm": 0.08163774758577347,
+      "learning_rate": 3.3560289368945406e-06,
+      "loss": 8.5104,
+      "step": 177900
+    },
+    {
+      "epoch": 0.8884616344977403,
+      "grad_norm": 0.09547600150108337,
+      "learning_rate": 3.3545270219529903e-06,
+      "loss": 8.5128,
+      "step": 177910
+    },
+    {
+      "epoch": 0.8885115733226797,
+      "grad_norm": 0.08949603140354156,
+      "learning_rate": 3.3530251070114396e-06,
+      "loss": 8.5251,
+      "step": 177920
+    },
+    {
+      "epoch": 0.8885615121476191,
+      "grad_norm": 0.09252443164587021,
+      "learning_rate": 3.3515231920698894e-06,
+      "loss": 8.5098,
+      "step": 177930
+    },
+    {
+      "epoch": 0.8886114509725587,
+      "grad_norm": 0.09023504704236984,
+      "learning_rate": 3.3500212771283387e-06,
+      "loss": 8.5021,
+      "step": 177940
+    },
+    {
+      "epoch": 0.8886613897974981,
+      "grad_norm": 0.0936330258846283,
+      "learning_rate": 3.348519362186788e-06,
+      "loss": 8.5215,
+      "step": 177950
+    },
+    {
+      "epoch": 0.8887113286224375,
+      "grad_norm": 0.09109587222337723,
+      "learning_rate": 3.347017447245238e-06,
+      "loss": 8.5083,
+      "step": 177960
+    },
+    {
+      "epoch": 0.8887612674473769,
+      "grad_norm": 0.08897558599710464,
+      "learning_rate": 3.3455155323036876e-06,
+      "loss": 8.5156,
+      "step": 177970
+    },
+    {
+      "epoch": 0.8888112062723165,
+      "grad_norm": 0.09321697801351547,
+      "learning_rate": 3.344013617362137e-06,
+      "loss": 8.5194,
+      "step": 177980
+    },
+    {
+      "epoch": 0.8888611450972559,
+      "grad_norm": 0.08879639208316803,
+      "learning_rate": 3.3425117024205862e-06,
+      "loss": 8.5025,
+      "step": 177990
+    },
+    {
+      "epoch": 0.8889110839221953,
+      "grad_norm": 0.08868400007486343,
+      "learning_rate": 3.3410097874790356e-06,
+      "loss": 8.524,
+      "step": 178000
+    },
+    {
+      "epoch": 0.8889610227471347,
+      "grad_norm": 0.0897117480635643,
+      "learning_rate": 3.3395078725374857e-06,
+      "loss": 8.5224,
+      "step": 178010
+    },
+    {
+      "epoch": 0.8890109615720742,
+      "grad_norm": 0.08525160700082779,
+      "learning_rate": 3.338005957595935e-06,
+      "loss": 8.5154,
+      "step": 178020
+    },
+    {
+      "epoch": 0.8890609003970137,
+      "grad_norm": 0.09332958608865738,
+      "learning_rate": 3.3365040426543844e-06,
+      "loss": 8.5251,
+      "step": 178030
+    },
+    {
+      "epoch": 0.8891108392219531,
+      "grad_norm": 0.08514176309108734,
+      "learning_rate": 3.3350021277128337e-06,
+      "loss": 8.5127,
+      "step": 178040
+    },
+    {
+      "epoch": 0.8891607780468925,
+      "grad_norm": 0.08755916357040405,
+      "learning_rate": 3.3335002127712835e-06,
+      "loss": 8.4974,
+      "step": 178050
+    },
+    {
+      "epoch": 0.889210716871832,
+      "grad_norm": 0.08407943695783615,
+      "learning_rate": 3.3319982978297332e-06,
+      "loss": 8.5274,
+      "step": 178060
+    },
+    {
+      "epoch": 0.8892606556967715,
+      "grad_norm": 0.08439040929079056,
+      "learning_rate": 3.3304963828881826e-06,
+      "loss": 8.5451,
+      "step": 178070
+    },
+    {
+      "epoch": 0.8893105945217109,
+      "grad_norm": 0.09115412086248398,
+      "learning_rate": 3.328994467946632e-06,
+      "loss": 8.515,
+      "step": 178080
+    },
+    {
+      "epoch": 0.8893605333466503,
+      "grad_norm": 0.0921141728758812,
+      "learning_rate": 3.3274925530050817e-06,
+      "loss": 8.5164,
+      "step": 178090
+    },
+    {
+      "epoch": 0.8894104721715897,
+      "grad_norm": 0.09111848473548889,
+      "learning_rate": 3.325990638063531e-06,
+      "loss": 8.5319,
+      "step": 178100
+    },
+    {
+      "epoch": 0.8894604109965293,
+      "grad_norm": 0.0888875350356102,
+      "learning_rate": 3.3244887231219808e-06,
+      "loss": 8.4937,
+      "step": 178110
+    },
+    {
+      "epoch": 0.8895103498214687,
+      "grad_norm": 0.08856623619794846,
+      "learning_rate": 3.32298680818043e-06,
+      "loss": 8.5252,
+      "step": 178120
+    },
+    {
+      "epoch": 0.8895602886464081,
+      "grad_norm": 0.0940069705247879,
+      "learning_rate": 3.32148489323888e-06,
+      "loss": 8.514,
+      "step": 178130
+    },
+    {
+      "epoch": 0.8896102274713475,
+      "grad_norm": 0.09143373370170593,
+      "learning_rate": 3.319982978297329e-06,
+      "loss": 8.5126,
+      "step": 178140
+    },
+    {
+      "epoch": 0.8896601662962871,
+      "grad_norm": 0.08722451329231262,
+      "learning_rate": 3.3184810633557785e-06,
+      "loss": 8.5131,
+      "step": 178150
+    },
+    {
+      "epoch": 0.8897101051212265,
+      "grad_norm": 0.09241468459367752,
+      "learning_rate": 3.3169791484142283e-06,
+      "loss": 8.5162,
+      "step": 178160
+    },
+    {
+      "epoch": 0.8897600439461659,
+      "grad_norm": 0.09892044216394424,
+      "learning_rate": 3.315477233472678e-06,
+      "loss": 8.5069,
+      "step": 178170
+    },
+    {
+      "epoch": 0.8898099827711053,
+      "grad_norm": 0.09388883411884308,
+      "learning_rate": 3.3139753185311273e-06,
+      "loss": 8.5245,
+      "step": 178180
+    },
+    {
+      "epoch": 0.8898599215960449,
+      "grad_norm": 0.08809326589107513,
+      "learning_rate": 3.3124734035895767e-06,
+      "loss": 8.5341,
+      "step": 178190
+    },
+    {
+      "epoch": 0.8899098604209843,
+      "grad_norm": 0.09530996531248093,
+      "learning_rate": 3.310971488648026e-06,
+      "loss": 8.4991,
+      "step": 178200
+    },
+    {
+      "epoch": 0.8899597992459237,
+      "grad_norm": 0.09561774879693985,
+      "learning_rate": 3.309469573706476e-06,
+      "loss": 8.5052,
+      "step": 178210
+    },
+    {
+      "epoch": 0.8900097380708631,
+      "grad_norm": 0.09873489290475845,
+      "learning_rate": 3.3079676587649255e-06,
+      "loss": 8.5081,
+      "step": 178220
+    },
+    {
+      "epoch": 0.8900596768958027,
+      "grad_norm": 0.08937592804431915,
+      "learning_rate": 3.306465743823375e-06,
+      "loss": 8.5305,
+      "step": 178230
+    },
+    {
+      "epoch": 0.8901096157207421,
+      "grad_norm": 0.09414834529161453,
+      "learning_rate": 3.304963828881824e-06,
+      "loss": 8.5262,
+      "step": 178240
+    },
+    {
+      "epoch": 0.8901595545456815,
+      "grad_norm": 0.09100335836410522,
+      "learning_rate": 3.303461913940274e-06,
+      "loss": 8.5105,
+      "step": 178250
+    },
+    {
+      "epoch": 0.8902094933706209,
+      "grad_norm": 0.09098581969738007,
+      "learning_rate": 3.3019599989987237e-06,
+      "loss": 8.521,
+      "step": 178260
+    },
+    {
+      "epoch": 0.8902594321955605,
+      "grad_norm": 0.09130746126174927,
+      "learning_rate": 3.300458084057173e-06,
+      "loss": 8.5235,
+      "step": 178270
+    },
+    {
+      "epoch": 0.8903093710204999,
+      "grad_norm": 0.08628986030817032,
+      "learning_rate": 3.2989561691156223e-06,
+      "loss": 8.5126,
+      "step": 178280
+    },
+    {
+      "epoch": 0.8903593098454393,
+      "grad_norm": 0.09175537526607513,
+      "learning_rate": 3.297454254174072e-06,
+      "loss": 8.5197,
+      "step": 178290
+    },
+    {
+      "epoch": 0.8904092486703787,
+      "grad_norm": 0.09056618064641953,
+      "learning_rate": 3.295952339232522e-06,
+      "loss": 8.5097,
+      "step": 178300
+    },
+    {
+      "epoch": 0.8904591874953183,
+      "grad_norm": 0.09293998777866364,
+      "learning_rate": 3.294450424290971e-06,
+      "loss": 8.5344,
+      "step": 178310
+    },
+    {
+      "epoch": 0.8905091263202577,
+      "grad_norm": 0.09091665595769882,
+      "learning_rate": 3.2929485093494205e-06,
+      "loss": 8.5299,
+      "step": 178320
+    },
+    {
+      "epoch": 0.8905590651451971,
+      "grad_norm": 0.0961182564496994,
+      "learning_rate": 3.2914465944078703e-06,
+      "loss": 8.5307,
+      "step": 178330
+    },
+    {
+      "epoch": 0.8906090039701365,
+      "grad_norm": 0.09689652919769287,
+      "learning_rate": 3.2899446794663196e-06,
+      "loss": 8.5363,
+      "step": 178340
+    },
+    {
+      "epoch": 0.8906589427950761,
+      "grad_norm": 0.09379088878631592,
+      "learning_rate": 3.2884427645247694e-06,
+      "loss": 8.5213,
+      "step": 178350
+    },
+    {
+      "epoch": 0.8907088816200155,
+      "grad_norm": 0.09888619929552078,
+      "learning_rate": 3.2869408495832187e-06,
+      "loss": 8.5056,
+      "step": 178360
+    },
+    {
+      "epoch": 0.8907588204449549,
+      "grad_norm": 0.09750135242938995,
+      "learning_rate": 3.2854389346416684e-06,
+      "loss": 8.5178,
+      "step": 178370
+    },
+    {
+      "epoch": 0.8908087592698943,
+      "grad_norm": 0.09058021754026413,
+      "learning_rate": 3.2839370197001178e-06,
+      "loss": 8.5104,
+      "step": 178380
+    },
+    {
+      "epoch": 0.8908586980948339,
+      "grad_norm": 0.08557943254709244,
+      "learning_rate": 3.282435104758567e-06,
+      "loss": 8.5229,
+      "step": 178390
+    },
+    {
+      "epoch": 0.8909086369197733,
+      "grad_norm": 0.09042729437351227,
+      "learning_rate": 3.280933189817017e-06,
+      "loss": 8.5308,
+      "step": 178400
+    },
+    {
+      "epoch": 0.8909585757447127,
+      "grad_norm": 0.08974535018205643,
+      "learning_rate": 3.279431274875466e-06,
+      "loss": 8.5126,
+      "step": 178410
+    },
+    {
+      "epoch": 0.8910085145696521,
+      "grad_norm": 0.0879034698009491,
+      "learning_rate": 3.277929359933916e-06,
+      "loss": 8.5104,
+      "step": 178420
+    },
+    {
+      "epoch": 0.8910584533945917,
+      "grad_norm": 0.08987148851156235,
+      "learning_rate": 3.2764274449923653e-06,
+      "loss": 8.4953,
+      "step": 178430
+    },
+    {
+      "epoch": 0.8911083922195311,
+      "grad_norm": 0.09033453464508057,
+      "learning_rate": 3.2749255300508146e-06,
+      "loss": 8.5089,
+      "step": 178440
+    },
+    {
+      "epoch": 0.8911583310444705,
+      "grad_norm": 0.09760011732578278,
+      "learning_rate": 3.2734236151092644e-06,
+      "loss": 8.4875,
+      "step": 178450
+    },
+    {
+      "epoch": 0.8912082698694099,
+      "grad_norm": 0.09189391136169434,
+      "learning_rate": 3.271921700167714e-06,
+      "loss": 8.5124,
+      "step": 178460
+    },
+    {
+      "epoch": 0.8912582086943495,
+      "grad_norm": 0.09732663631439209,
+      "learning_rate": 3.2704197852261635e-06,
+      "loss": 8.5091,
+      "step": 178470
+    },
+    {
+      "epoch": 0.8913081475192889,
+      "grad_norm": 0.0918707624077797,
+      "learning_rate": 3.2689178702846128e-06,
+      "loss": 8.5142,
+      "step": 178480
+    },
+    {
+      "epoch": 0.8913580863442283,
+      "grad_norm": 0.09437326341867447,
+      "learning_rate": 3.267415955343062e-06,
+      "loss": 8.5027,
+      "step": 178490
+    },
+    {
+      "epoch": 0.8914080251691677,
+      "grad_norm": 0.09226923435926437,
+      "learning_rate": 3.2659140404015123e-06,
+      "loss": 8.5185,
+      "step": 178500
+    },
+    {
+      "epoch": 0.8914579639941073,
+      "grad_norm": 0.09034575521945953,
+      "learning_rate": 3.2644121254599616e-06,
+      "loss": 8.5148,
+      "step": 178510
+    },
+    {
+      "epoch": 0.8915079028190467,
+      "grad_norm": 0.09127431362867355,
+      "learning_rate": 3.262910210518411e-06,
+      "loss": 8.5053,
+      "step": 178520
+    },
+    {
+      "epoch": 0.8915578416439861,
+      "grad_norm": 0.09689009934663773,
+      "learning_rate": 3.2614082955768603e-06,
+      "loss": 8.4981,
+      "step": 178530
+    },
+    {
+      "epoch": 0.8916077804689255,
+      "grad_norm": 0.09404323995113373,
+      "learning_rate": 3.25990638063531e-06,
+      "loss": 8.5289,
+      "step": 178540
+    },
+    {
+      "epoch": 0.8916577192938651,
+      "grad_norm": 0.08964677900075912,
+      "learning_rate": 3.25840446569376e-06,
+      "loss": 8.5246,
+      "step": 178550
+    },
+    {
+      "epoch": 0.8917076581188045,
+      "grad_norm": 0.08872044086456299,
+      "learning_rate": 3.256902550752209e-06,
+      "loss": 8.5265,
+      "step": 178560
+    },
+    {
+      "epoch": 0.8917575969437439,
+      "grad_norm": 0.08949794620275497,
+      "learning_rate": 3.2554006358106585e-06,
+      "loss": 8.5037,
+      "step": 178570
+    },
+    {
+      "epoch": 0.8918075357686833,
+      "grad_norm": 0.08609029650688171,
+      "learning_rate": 3.2538987208691082e-06,
+      "loss": 8.5024,
+      "step": 178580
+    },
+    {
+      "epoch": 0.8918574745936229,
+      "grad_norm": 0.09221512079238892,
+      "learning_rate": 3.2523968059275575e-06,
+      "loss": 8.5054,
+      "step": 178590
+    },
+    {
+      "epoch": 0.8919074134185623,
+      "grad_norm": 0.08988411724567413,
+      "learning_rate": 3.2508948909860073e-06,
+      "loss": 8.5123,
+      "step": 178600
+    },
+    {
+      "epoch": 0.8919573522435017,
+      "grad_norm": 0.08725704252719879,
+      "learning_rate": 3.2493929760444566e-06,
+      "loss": 8.525,
+      "step": 178610
+    },
+    {
+      "epoch": 0.8920072910684411,
+      "grad_norm": 0.09931132942438126,
+      "learning_rate": 3.2478910611029064e-06,
+      "loss": 8.514,
+      "step": 178620
+    },
+    {
+      "epoch": 0.8920572298933807,
+      "grad_norm": 0.09435875713825226,
+      "learning_rate": 3.2463891461613557e-06,
+      "loss": 8.5047,
+      "step": 178630
+    },
+    {
+      "epoch": 0.8921071687183201,
+      "grad_norm": 0.092837855219841,
+      "learning_rate": 3.244887231219805e-06,
+      "loss": 8.5033,
+      "step": 178640
+    },
+    {
+      "epoch": 0.8921571075432595,
+      "grad_norm": 0.09155397117137909,
+      "learning_rate": 3.243385316278255e-06,
+      "loss": 8.5026,
+      "step": 178650
+    },
+    {
+      "epoch": 0.8922070463681989,
+      "grad_norm": 0.08951444178819656,
+      "learning_rate": 3.2418834013367046e-06,
+      "loss": 8.5213,
+      "step": 178660
+    },
+    {
+      "epoch": 0.8922569851931385,
+      "grad_norm": 0.08599022775888443,
+      "learning_rate": 3.240381486395154e-06,
+      "loss": 8.5207,
+      "step": 178670
+    },
+    {
+      "epoch": 0.8923069240180779,
+      "grad_norm": 0.08356813341379166,
+      "learning_rate": 3.2388795714536032e-06,
+      "loss": 8.5287,
+      "step": 178680
+    },
+    {
+      "epoch": 0.8923568628430173,
+      "grad_norm": 0.09023530781269073,
+      "learning_rate": 3.2373776565120526e-06,
+      "loss": 8.525,
+      "step": 178690
+    },
+    {
+      "epoch": 0.8924068016679567,
+      "grad_norm": 0.09185218065977097,
+      "learning_rate": 3.2358757415705027e-06,
+      "loss": 8.5036,
+      "step": 178700
+    },
+    {
+      "epoch": 0.8924567404928962,
+      "grad_norm": 0.09013650566339493,
+      "learning_rate": 3.234373826628952e-06,
+      "loss": 8.5288,
+      "step": 178710
+    },
+    {
+      "epoch": 0.8925066793178357,
+      "grad_norm": 0.0958433449268341,
+      "learning_rate": 3.2328719116874014e-06,
+      "loss": 8.5167,
+      "step": 178720
+    },
+    {
+      "epoch": 0.8925566181427751,
+      "grad_norm": 0.08722110837697983,
+      "learning_rate": 3.2313699967458507e-06,
+      "loss": 8.5105,
+      "step": 178730
+    },
+    {
+      "epoch": 0.8926065569677145,
+      "grad_norm": 0.09434662014245987,
+      "learning_rate": 3.2298680818043005e-06,
+      "loss": 8.5244,
+      "step": 178740
+    },
+    {
+      "epoch": 0.892656495792654,
+      "grad_norm": 0.09160378575325012,
+      "learning_rate": 3.2283661668627502e-06,
+      "loss": 8.5077,
+      "step": 178750
+    },
+    {
+      "epoch": 0.8927064346175935,
+      "grad_norm": 0.09734846651554108,
+      "learning_rate": 3.2268642519211996e-06,
+      "loss": 8.517,
+      "step": 178760
+    },
+    {
+      "epoch": 0.8927563734425329,
+      "grad_norm": 0.0917469784617424,
+      "learning_rate": 3.225362336979649e-06,
+      "loss": 8.5156,
+      "step": 178770
+    },
+    {
+      "epoch": 0.8928063122674723,
+      "grad_norm": 0.09718631953001022,
+      "learning_rate": 3.2238604220380987e-06,
+      "loss": 8.4938,
+      "step": 178780
+    },
+    {
+      "epoch": 0.8928562510924118,
+      "grad_norm": 0.09104308485984802,
+      "learning_rate": 3.222358507096548e-06,
+      "loss": 8.5254,
+      "step": 178790
+    },
+    {
+      "epoch": 0.8929061899173513,
+      "grad_norm": 0.09059741348028183,
+      "learning_rate": 3.2208565921549977e-06,
+      "loss": 8.529,
+      "step": 178800
+    },
+    {
+      "epoch": 0.8929561287422907,
+      "grad_norm": 0.0890093445777893,
+      "learning_rate": 3.219354677213447e-06,
+      "loss": 8.5212,
+      "step": 178810
+    },
+    {
+      "epoch": 0.8930060675672301,
+      "grad_norm": 0.09546619653701782,
+      "learning_rate": 3.217852762271897e-06,
+      "loss": 8.5247,
+      "step": 178820
+    },
+    {
+      "epoch": 0.8930560063921696,
+      "grad_norm": 0.09299060702323914,
+      "learning_rate": 3.216350847330346e-06,
+      "loss": 8.517,
+      "step": 178830
+    },
+    {
+      "epoch": 0.8931059452171091,
+      "grad_norm": 0.09568481147289276,
+      "learning_rate": 3.214848932388796e-06,
+      "loss": 8.5188,
+      "step": 178840
+    },
+    {
+      "epoch": 0.8931558840420485,
+      "grad_norm": 0.09317091107368469,
+      "learning_rate": 3.2133470174472452e-06,
+      "loss": 8.5156,
+      "step": 178850
+    },
+    {
+      "epoch": 0.8932058228669879,
+      "grad_norm": 0.08648492395877838,
+      "learning_rate": 3.211845102505695e-06,
+      "loss": 8.5408,
+      "step": 178860
+    },
+    {
+      "epoch": 0.8932557616919274,
+      "grad_norm": 0.09608660638332367,
+      "learning_rate": 3.2103431875641443e-06,
+      "loss": 8.5176,
+      "step": 178870
+    },
+    {
+      "epoch": 0.8933057005168669,
+      "grad_norm": 0.09281721711158752,
+      "learning_rate": 3.2088412726225937e-06,
+      "loss": 8.5124,
+      "step": 178880
+    },
+    {
+      "epoch": 0.8933556393418063,
+      "grad_norm": 0.08907417207956314,
+      "learning_rate": 3.2073393576810434e-06,
+      "loss": 8.5096,
+      "step": 178890
+    },
+    {
+      "epoch": 0.8934055781667457,
+      "grad_norm": 0.09067781269550323,
+      "learning_rate": 3.205837442739493e-06,
+      "loss": 8.5257,
+      "step": 178900
+    },
+    {
+      "epoch": 0.8934555169916852,
+      "grad_norm": 0.08892224729061127,
+      "learning_rate": 3.2043355277979425e-06,
+      "loss": 8.5168,
+      "step": 178910
+    },
+    {
+      "epoch": 0.8935054558166247,
+      "grad_norm": 0.08958274871110916,
+      "learning_rate": 3.202833612856392e-06,
+      "loss": 8.5153,
+      "step": 178920
+    },
+    {
+      "epoch": 0.8935553946415641,
+      "grad_norm": 0.08994346857070923,
+      "learning_rate": 3.201331697914841e-06,
+      "loss": 8.5179,
+      "step": 178930
+    },
+    {
+      "epoch": 0.8936053334665035,
+      "grad_norm": 0.09714482724666595,
+      "learning_rate": 3.1998297829732913e-06,
+      "loss": 8.5238,
+      "step": 178940
+    },
+    {
+      "epoch": 0.893655272291443,
+      "grad_norm": 0.09184648096561432,
+      "learning_rate": 3.1983278680317407e-06,
+      "loss": 8.5161,
+      "step": 178950
+    },
+    {
+      "epoch": 0.8937052111163825,
+      "grad_norm": 0.0893314927816391,
+      "learning_rate": 3.19682595309019e-06,
+      "loss": 8.5033,
+      "step": 178960
+    },
+    {
+      "epoch": 0.8937551499413219,
+      "grad_norm": 0.09436094015836716,
+      "learning_rate": 3.1953240381486393e-06,
+      "loss": 8.4939,
+      "step": 178970
+    },
+    {
+      "epoch": 0.8938050887662613,
+      "grad_norm": 0.08903580904006958,
+      "learning_rate": 3.193822123207089e-06,
+      "loss": 8.5137,
+      "step": 178980
+    },
+    {
+      "epoch": 0.8938550275912007,
+      "grad_norm": 0.089007668197155,
+      "learning_rate": 3.192320208265539e-06,
+      "loss": 8.5193,
+      "step": 178990
+    },
+    {
+      "epoch": 0.8939049664161403,
+      "grad_norm": 0.09373140335083008,
+      "learning_rate": 3.190818293323988e-06,
+      "loss": 8.4969,
+      "step": 179000
+    },
+    {
+      "epoch": 0.8939549052410797,
+      "grad_norm": 0.08744634687900543,
+      "learning_rate": 3.1893163783824375e-06,
+      "loss": 8.5342,
+      "step": 179010
+    },
+    {
+      "epoch": 0.8940048440660191,
+      "grad_norm": 0.0939723551273346,
+      "learning_rate": 3.1878144634408873e-06,
+      "loss": 8.5059,
+      "step": 179020
+    },
+    {
+      "epoch": 0.8940547828909585,
+      "grad_norm": 0.09691902250051498,
+      "learning_rate": 3.1863125484993366e-06,
+      "loss": 8.5022,
+      "step": 179030
+    },
+    {
+      "epoch": 0.894104721715898,
+      "grad_norm": 0.09357941895723343,
+      "learning_rate": 3.1848106335577864e-06,
+      "loss": 8.5155,
+      "step": 179040
+    },
+    {
+      "epoch": 0.8941546605408375,
+      "grad_norm": 0.09528280794620514,
+      "learning_rate": 3.1833087186162357e-06,
+      "loss": 8.5088,
+      "step": 179050
+    },
+    {
+      "epoch": 0.8942045993657769,
+      "grad_norm": 0.09074392914772034,
+      "learning_rate": 3.1818068036746854e-06,
+      "loss": 8.4962,
+      "step": 179060
+    },
+    {
+      "epoch": 0.8942545381907163,
+      "grad_norm": 0.09400563687086105,
+      "learning_rate": 3.1803048887331348e-06,
+      "loss": 8.5115,
+      "step": 179070
+    },
+    {
+      "epoch": 0.8943044770156559,
+      "grad_norm": 0.09290362149477005,
+      "learning_rate": 3.178802973791584e-06,
+      "loss": 8.5178,
+      "step": 179080
+    },
+    {
+      "epoch": 0.8943544158405953,
+      "grad_norm": 0.09567950665950775,
+      "learning_rate": 3.177301058850034e-06,
+      "loss": 8.5186,
+      "step": 179090
+    },
+    {
+      "epoch": 0.8944043546655347,
+      "grad_norm": 0.09070657938718796,
+      "learning_rate": 3.1757991439084836e-06,
+      "loss": 8.5161,
+      "step": 179100
+    },
+    {
+      "epoch": 0.8944542934904741,
+      "grad_norm": 0.08824972063302994,
+      "learning_rate": 3.174297228966933e-06,
+      "loss": 8.5189,
+      "step": 179110
+    },
+    {
+      "epoch": 0.8945042323154136,
+      "grad_norm": 0.09949322789907455,
+      "learning_rate": 3.1727953140253823e-06,
+      "loss": 8.5051,
+      "step": 179120
+    },
+    {
+      "epoch": 0.8945541711403531,
+      "grad_norm": 0.09183038771152496,
+      "learning_rate": 3.1712933990838316e-06,
+      "loss": 8.5139,
+      "step": 179130
+    },
+    {
+      "epoch": 0.8946041099652925,
+      "grad_norm": 0.09553587436676025,
+      "learning_rate": 3.1697914841422818e-06,
+      "loss": 8.4899,
+      "step": 179140
+    },
+    {
+      "epoch": 0.8946540487902319,
+      "grad_norm": 0.089163638651371,
+      "learning_rate": 3.168289569200731e-06,
+      "loss": 8.5241,
+      "step": 179150
+    },
+    {
+      "epoch": 0.8947039876151714,
+      "grad_norm": 0.09962226450443268,
+      "learning_rate": 3.1667876542591804e-06,
+      "loss": 8.5203,
+      "step": 179160
+    },
+    {
+      "epoch": 0.8947539264401109,
+      "grad_norm": 0.09414087980985641,
+      "learning_rate": 3.1652857393176298e-06,
+      "loss": 8.5183,
+      "step": 179170
+    },
+    {
+      "epoch": 0.8948038652650503,
+      "grad_norm": 0.09427635371685028,
+      "learning_rate": 3.1637838243760795e-06,
+      "loss": 8.5292,
+      "step": 179180
+    },
+    {
+      "epoch": 0.8948538040899897,
+      "grad_norm": 0.08661588281393051,
+      "learning_rate": 3.1622819094345293e-06,
+      "loss": 8.5173,
+      "step": 179190
+    },
+    {
+      "epoch": 0.8949037429149292,
+      "grad_norm": 0.09423363953828812,
+      "learning_rate": 3.1607799944929786e-06,
+      "loss": 8.5142,
+      "step": 179200
+    },
+    {
+      "epoch": 0.8949536817398687,
+      "grad_norm": 0.08995388448238373,
+      "learning_rate": 3.159278079551428e-06,
+      "loss": 8.529,
+      "step": 179210
+    },
+    {
+      "epoch": 0.8950036205648081,
+      "grad_norm": 0.09153537452220917,
+      "learning_rate": 3.1577761646098777e-06,
+      "loss": 8.5241,
+      "step": 179220
+    },
+    {
+      "epoch": 0.8950535593897475,
+      "grad_norm": 0.09005282819271088,
+      "learning_rate": 3.156274249668327e-06,
+      "loss": 8.5244,
+      "step": 179230
+    },
+    {
+      "epoch": 0.895103498214687,
+      "grad_norm": 0.08991970866918564,
+      "learning_rate": 3.154772334726777e-06,
+      "loss": 8.5312,
+      "step": 179240
+    },
+    {
+      "epoch": 0.8951534370396265,
+      "grad_norm": 0.08929390460252762,
+      "learning_rate": 3.153270419785226e-06,
+      "loss": 8.5027,
+      "step": 179250
+    },
+    {
+      "epoch": 0.8952033758645659,
+      "grad_norm": 0.08898353576660156,
+      "learning_rate": 3.151768504843676e-06,
+      "loss": 8.5349,
+      "step": 179260
+    },
+    {
+      "epoch": 0.8952533146895053,
+      "grad_norm": 0.08719191700220108,
+      "learning_rate": 3.150266589902125e-06,
+      "loss": 8.5089,
+      "step": 179270
+    },
+    {
+      "epoch": 0.8953032535144448,
+      "grad_norm": 0.09334076195955276,
+      "learning_rate": 3.1487646749605745e-06,
+      "loss": 8.5229,
+      "step": 179280
+    },
+    {
+      "epoch": 0.8953531923393843,
+      "grad_norm": 0.09189531207084656,
+      "learning_rate": 3.1472627600190243e-06,
+      "loss": 8.518,
+      "step": 179290
+    },
+    {
+      "epoch": 0.8954031311643237,
+      "grad_norm": 0.08680342137813568,
+      "learning_rate": 3.145760845077474e-06,
+      "loss": 8.5283,
+      "step": 179300
+    },
+    {
+      "epoch": 0.8954530699892631,
+      "grad_norm": 0.0863579586148262,
+      "learning_rate": 3.1442589301359234e-06,
+      "loss": 8.5031,
+      "step": 179310
+    },
+    {
+      "epoch": 0.8955030088142026,
+      "grad_norm": 0.09513013064861298,
+      "learning_rate": 3.1427570151943727e-06,
+      "loss": 8.5172,
+      "step": 179320
+    },
+    {
+      "epoch": 0.8955529476391421,
+      "grad_norm": 0.08853243291378021,
+      "learning_rate": 3.1412551002528225e-06,
+      "loss": 8.5315,
+      "step": 179330
+    },
+    {
+      "epoch": 0.8956028864640815,
+      "grad_norm": 0.09000103175640106,
+      "learning_rate": 3.1397531853112722e-06,
+      "loss": 8.5139,
+      "step": 179340
+    },
+    {
+      "epoch": 0.8956528252890209,
+      "grad_norm": 0.0949261337518692,
+      "learning_rate": 3.1382512703697216e-06,
+      "loss": 8.5104,
+      "step": 179350
+    },
+    {
+      "epoch": 0.8957027641139604,
+      "grad_norm": 0.09614192694425583,
+      "learning_rate": 3.136749355428171e-06,
+      "loss": 8.5061,
+      "step": 179360
+    },
+    {
+      "epoch": 0.8957527029388999,
+      "grad_norm": 0.09426548331975937,
+      "learning_rate": 3.1352474404866202e-06,
+      "loss": 8.5072,
+      "step": 179370
+    },
+    {
+      "epoch": 0.8958026417638393,
+      "grad_norm": 0.09127119183540344,
+      "learning_rate": 3.1337455255450704e-06,
+      "loss": 8.5092,
+      "step": 179380
+    },
+    {
+      "epoch": 0.8958525805887787,
+      "grad_norm": 0.09280092269182205,
+      "learning_rate": 3.1322436106035197e-06,
+      "loss": 8.5003,
+      "step": 179390
+    },
+    {
+      "epoch": 0.8959025194137182,
+      "grad_norm": 0.09022334963083267,
+      "learning_rate": 3.130741695661969e-06,
+      "loss": 8.5157,
+      "step": 179400
+    },
+    {
+      "epoch": 0.8959524582386577,
+      "grad_norm": 0.08918463438749313,
+      "learning_rate": 3.1292397807204184e-06,
+      "loss": 8.5252,
+      "step": 179410
+    },
+    {
+      "epoch": 0.8960023970635971,
+      "grad_norm": 0.09359756857156754,
+      "learning_rate": 3.127737865778868e-06,
+      "loss": 8.5137,
+      "step": 179420
+    },
+    {
+      "epoch": 0.8960523358885365,
+      "grad_norm": 0.08726957440376282,
+      "learning_rate": 3.126235950837318e-06,
+      "loss": 8.5027,
+      "step": 179430
+    },
+    {
+      "epoch": 0.896102274713476,
+      "grad_norm": 0.08867673575878143,
+      "learning_rate": 3.1247340358957672e-06,
+      "loss": 8.5117,
+      "step": 179440
+    },
+    {
+      "epoch": 0.8961522135384155,
+      "grad_norm": 0.0853889063000679,
+      "learning_rate": 3.1232321209542166e-06,
+      "loss": 8.5262,
+      "step": 179450
+    },
+    {
+      "epoch": 0.8962021523633549,
+      "grad_norm": 0.08606785535812378,
+      "learning_rate": 3.1217302060126663e-06,
+      "loss": 8.5221,
+      "step": 179460
+    },
+    {
+      "epoch": 0.8962520911882943,
+      "grad_norm": 0.08745083957910538,
+      "learning_rate": 3.1202282910711156e-06,
+      "loss": 8.5107,
+      "step": 179470
+    },
+    {
+      "epoch": 0.8963020300132338,
+      "grad_norm": 0.09155963361263275,
+      "learning_rate": 3.1187263761295654e-06,
+      "loss": 8.4951,
+      "step": 179480
+    },
+    {
+      "epoch": 0.8963519688381733,
+      "grad_norm": 0.09494645148515701,
+      "learning_rate": 3.1172244611880147e-06,
+      "loss": 8.509,
+      "step": 179490
+    },
+    {
+      "epoch": 0.8964019076631127,
+      "grad_norm": 0.0906292051076889,
+      "learning_rate": 3.1157225462464645e-06,
+      "loss": 8.5103,
+      "step": 179500
+    },
+    {
+      "epoch": 0.8964518464880521,
+      "grad_norm": 0.08988522738218307,
+      "learning_rate": 3.114220631304914e-06,
+      "loss": 8.5154,
+      "step": 179510
+    },
+    {
+      "epoch": 0.8965017853129916,
+      "grad_norm": 0.09074413031339645,
+      "learning_rate": 3.112718716363363e-06,
+      "loss": 8.5289,
+      "step": 179520
+    },
+    {
+      "epoch": 0.896551724137931,
+      "grad_norm": 0.0895250216126442,
+      "learning_rate": 3.111216801421813e-06,
+      "loss": 8.5203,
+      "step": 179530
+    },
+    {
+      "epoch": 0.8966016629628705,
+      "grad_norm": 0.09052935242652893,
+      "learning_rate": 3.1097148864802627e-06,
+      "loss": 8.5309,
+      "step": 179540
+    },
+    {
+      "epoch": 0.8966516017878099,
+      "grad_norm": 0.09444441646337509,
+      "learning_rate": 3.108212971538712e-06,
+      "loss": 8.515,
+      "step": 179550
+    },
+    {
+      "epoch": 0.8967015406127494,
+      "grad_norm": 0.09313017874956131,
+      "learning_rate": 3.1067110565971613e-06,
+      "loss": 8.5094,
+      "step": 179560
+    },
+    {
+      "epoch": 0.8967514794376888,
+      "grad_norm": 0.09455044567584991,
+      "learning_rate": 3.1052091416556107e-06,
+      "loss": 8.5063,
+      "step": 179570
+    },
+    {
+      "epoch": 0.8968014182626283,
+      "grad_norm": 0.09298877418041229,
+      "learning_rate": 3.103707226714061e-06,
+      "loss": 8.5096,
+      "step": 179580
+    },
+    {
+      "epoch": 0.8968513570875677,
+      "grad_norm": 0.09229998290538788,
+      "learning_rate": 3.10220531177251e-06,
+      "loss": 8.5201,
+      "step": 179590
+    },
+    {
+      "epoch": 0.8969012959125072,
+      "grad_norm": 0.08736059069633484,
+      "learning_rate": 3.1007033968309595e-06,
+      "loss": 8.4993,
+      "step": 179600
+    },
+    {
+      "epoch": 0.8969512347374466,
+      "grad_norm": 0.09250369668006897,
+      "learning_rate": 3.099201481889409e-06,
+      "loss": 8.494,
+      "step": 179610
+    },
+    {
+      "epoch": 0.8970011735623861,
+      "grad_norm": 0.08871186524629593,
+      "learning_rate": 3.0976995669478586e-06,
+      "loss": 8.5137,
+      "step": 179620
+    },
+    {
+      "epoch": 0.8970511123873255,
+      "grad_norm": 0.09203743189573288,
+      "learning_rate": 3.0961976520063083e-06,
+      "loss": 8.5138,
+      "step": 179630
+    },
+    {
+      "epoch": 0.897101051212265,
+      "grad_norm": 0.09076462686061859,
+      "learning_rate": 3.0946957370647577e-06,
+      "loss": 8.5117,
+      "step": 179640
+    },
+    {
+      "epoch": 0.8971509900372044,
+      "grad_norm": 0.09100484848022461,
+      "learning_rate": 3.093193822123207e-06,
+      "loss": 8.5129,
+      "step": 179650
+    },
+    {
+      "epoch": 0.8972009288621439,
+      "grad_norm": 0.0954102948307991,
+      "learning_rate": 3.0916919071816568e-06,
+      "loss": 8.4978,
+      "step": 179660
+    },
+    {
+      "epoch": 0.8972508676870833,
+      "grad_norm": 0.09269285947084427,
+      "learning_rate": 3.090189992240106e-06,
+      "loss": 8.5116,
+      "step": 179670
+    },
+    {
+      "epoch": 0.8973008065120228,
+      "grad_norm": 0.08553628623485565,
+      "learning_rate": 3.088688077298556e-06,
+      "loss": 8.5297,
+      "step": 179680
+    },
+    {
+      "epoch": 0.8973507453369622,
+      "grad_norm": 0.08914821594953537,
+      "learning_rate": 3.087186162357005e-06,
+      "loss": 8.5084,
+      "step": 179690
+    },
+    {
+      "epoch": 0.8974006841619017,
+      "grad_norm": 0.08601313084363937,
+      "learning_rate": 3.085684247415455e-06,
+      "loss": 8.507,
+      "step": 179700
+    },
+    {
+      "epoch": 0.8974506229868411,
+      "grad_norm": 0.09413706511259079,
+      "learning_rate": 3.0841823324739043e-06,
+      "loss": 8.5197,
+      "step": 179710
+    },
+    {
+      "epoch": 0.8975005618117806,
+      "grad_norm": 0.09138408303260803,
+      "learning_rate": 3.0826804175323536e-06,
+      "loss": 8.5177,
+      "step": 179720
+    },
+    {
+      "epoch": 0.89755050063672,
+      "grad_norm": 0.09180725365877151,
+      "learning_rate": 3.0811785025908033e-06,
+      "loss": 8.5083,
+      "step": 179730
+    },
+    {
+      "epoch": 0.8976004394616595,
+      "grad_norm": 0.09291674196720123,
+      "learning_rate": 3.079676587649253e-06,
+      "loss": 8.5193,
+      "step": 179740
+    },
+    {
+      "epoch": 0.8976503782865989,
+      "grad_norm": 0.08994516730308533,
+      "learning_rate": 3.0781746727077024e-06,
+      "loss": 8.519,
+      "step": 179750
+    },
+    {
+      "epoch": 0.8977003171115384,
+      "grad_norm": 0.09701023995876312,
+      "learning_rate": 3.0766727577661518e-06,
+      "loss": 8.5088,
+      "step": 179760
+    },
+    {
+      "epoch": 0.8977502559364778,
+      "grad_norm": 0.0889933779835701,
+      "learning_rate": 3.075170842824601e-06,
+      "loss": 8.5115,
+      "step": 179770
+    },
+    {
+      "epoch": 0.8978001947614173,
+      "grad_norm": 0.09396040439605713,
+      "learning_rate": 3.0736689278830513e-06,
+      "loss": 8.5349,
+      "step": 179780
+    },
+    {
+      "epoch": 0.8978501335863567,
+      "grad_norm": 0.09631121158599854,
+      "learning_rate": 3.0721670129415006e-06,
+      "loss": 8.4926,
+      "step": 179790
+    },
+    {
+      "epoch": 0.8979000724112962,
+      "grad_norm": 0.08764296770095825,
+      "learning_rate": 3.07066509799995e-06,
+      "loss": 8.524,
+      "step": 179800
+    },
+    {
+      "epoch": 0.8979500112362356,
+      "grad_norm": 0.09530488401651382,
+      "learning_rate": 3.0691631830583993e-06,
+      "loss": 8.513,
+      "step": 179810
+    },
+    {
+      "epoch": 0.8979999500611751,
+      "grad_norm": 0.08981040865182877,
+      "learning_rate": 3.067661268116849e-06,
+      "loss": 8.5024,
+      "step": 179820
+    },
+    {
+      "epoch": 0.8980498888861145,
+      "grad_norm": 0.09306181967258453,
+      "learning_rate": 3.0661593531752988e-06,
+      "loss": 8.5005,
+      "step": 179830
+    },
+    {
+      "epoch": 0.898099827711054,
+      "grad_norm": 0.09456732124090195,
+      "learning_rate": 3.064657438233748e-06,
+      "loss": 8.5144,
+      "step": 179840
+    },
+    {
+      "epoch": 0.8981497665359934,
+      "grad_norm": 0.08984339982271194,
+      "learning_rate": 3.0631555232921974e-06,
+      "loss": 8.5388,
+      "step": 179850
+    },
+    {
+      "epoch": 0.8981997053609329,
+      "grad_norm": 0.08446503430604935,
+      "learning_rate": 3.061653608350647e-06,
+      "loss": 8.5152,
+      "step": 179860
+    },
+    {
+      "epoch": 0.8982496441858723,
+      "grad_norm": 0.09314480423927307,
+      "learning_rate": 3.0601516934090965e-06,
+      "loss": 8.5108,
+      "step": 179870
+    },
+    {
+      "epoch": 0.8982995830108118,
+      "grad_norm": 0.09413187950849533,
+      "learning_rate": 3.0586497784675463e-06,
+      "loss": 8.5027,
+      "step": 179880
+    },
+    {
+      "epoch": 0.8983495218357512,
+      "grad_norm": 0.09056896716356277,
+      "learning_rate": 3.0571478635259956e-06,
+      "loss": 8.5067,
+      "step": 179890
+    },
+    {
+      "epoch": 0.8983994606606907,
+      "grad_norm": 0.08975179493427277,
+      "learning_rate": 3.0556459485844454e-06,
+      "loss": 8.5157,
+      "step": 179900
+    },
+    {
+      "epoch": 0.8984493994856301,
+      "grad_norm": 0.08911077678203583,
+      "learning_rate": 3.0541440336428947e-06,
+      "loss": 8.516,
+      "step": 179910
+    },
+    {
+      "epoch": 0.8984993383105696,
+      "grad_norm": 0.09924951940774918,
+      "learning_rate": 3.0526421187013444e-06,
+      "loss": 8.4984,
+      "step": 179920
+    },
+    {
+      "epoch": 0.898549277135509,
+      "grad_norm": 0.09128329157829285,
+      "learning_rate": 3.0511402037597938e-06,
+      "loss": 8.5007,
+      "step": 179930
+    },
+    {
+      "epoch": 0.8985992159604484,
+      "grad_norm": 0.08956824243068695,
+      "learning_rate": 3.0496382888182435e-06,
+      "loss": 8.5084,
+      "step": 179940
+    },
+    {
+      "epoch": 0.8986491547853879,
+      "grad_norm": 0.09009304642677307,
+      "learning_rate": 3.048136373876693e-06,
+      "loss": 8.5131,
+      "step": 179950
+    },
+    {
+      "epoch": 0.8986990936103274,
+      "grad_norm": 0.0977887287735939,
+      "learning_rate": 3.046634458935142e-06,
+      "loss": 8.5111,
+      "step": 179960
+    },
+    {
+      "epoch": 0.8987490324352668,
+      "grad_norm": 0.08121143281459808,
+      "learning_rate": 3.045132543993592e-06,
+      "loss": 8.512,
+      "step": 179970
+    },
+    {
+      "epoch": 0.8987989712602062,
+      "grad_norm": 0.09262192249298096,
+      "learning_rate": 3.0436306290520417e-06,
+      "loss": 8.523,
+      "step": 179980
+    },
+    {
+      "epoch": 0.8988489100851457,
+      "grad_norm": 0.09229442477226257,
+      "learning_rate": 3.042128714110491e-06,
+      "loss": 8.5225,
+      "step": 179990
+    },
+    {
+      "epoch": 0.8988988489100851,
+      "grad_norm": 0.092495396733284,
+      "learning_rate": 3.0406267991689404e-06,
+      "loss": 8.5128,
+      "step": 180000
+    },
+    {
+      "epoch": 0.8989487877350246,
+      "grad_norm": 0.09082264453172684,
+      "learning_rate": 3.0391248842273897e-06,
+      "loss": 8.5133,
+      "step": 180010
+    },
+    {
+      "epoch": 0.898998726559964,
+      "grad_norm": 0.09642348438501358,
+      "learning_rate": 3.03762296928584e-06,
+      "loss": 8.4977,
+      "step": 180020
+    },
+    {
+      "epoch": 0.8990486653849035,
+      "grad_norm": 0.08731796592473984,
+      "learning_rate": 3.036121054344289e-06,
+      "loss": 8.5145,
+      "step": 180030
+    },
+    {
+      "epoch": 0.8990986042098429,
+      "grad_norm": 0.0900919958949089,
+      "learning_rate": 3.0346191394027385e-06,
+      "loss": 8.5136,
+      "step": 180040
+    },
+    {
+      "epoch": 0.8991485430347824,
+      "grad_norm": 0.09062129259109497,
+      "learning_rate": 3.033117224461188e-06,
+      "loss": 8.5201,
+      "step": 180050
+    },
+    {
+      "epoch": 0.8991984818597218,
+      "grad_norm": 0.09137134999036789,
+      "learning_rate": 3.0316153095196376e-06,
+      "loss": 8.5163,
+      "step": 180060
+    },
+    {
+      "epoch": 0.8992484206846613,
+      "grad_norm": 0.0882347971200943,
+      "learning_rate": 3.0301133945780874e-06,
+      "loss": 8.516,
+      "step": 180070
+    },
+    {
+      "epoch": 0.8992983595096007,
+      "grad_norm": 0.08691363036632538,
+      "learning_rate": 3.0286114796365367e-06,
+      "loss": 8.5077,
+      "step": 180080
+    },
+    {
+      "epoch": 0.8993482983345402,
+      "grad_norm": 0.08639916777610779,
+      "learning_rate": 3.027109564694986e-06,
+      "loss": 8.5125,
+      "step": 180090
+    },
+    {
+      "epoch": 0.8993982371594796,
+      "grad_norm": 0.08904005587100983,
+      "learning_rate": 3.025607649753436e-06,
+      "loss": 8.5207,
+      "step": 180100
+    },
+    {
+      "epoch": 0.8994481759844191,
+      "grad_norm": 0.09132896363735199,
+      "learning_rate": 3.024105734811885e-06,
+      "loss": 8.5062,
+      "step": 180110
+    },
+    {
+      "epoch": 0.8994981148093585,
+      "grad_norm": 0.08482132852077484,
+      "learning_rate": 3.022603819870335e-06,
+      "loss": 8.5185,
+      "step": 180120
+    },
+    {
+      "epoch": 0.899548053634298,
+      "grad_norm": 0.08881242573261261,
+      "learning_rate": 3.0211019049287842e-06,
+      "loss": 8.5312,
+      "step": 180130
+    },
+    {
+      "epoch": 0.8995979924592374,
+      "grad_norm": 0.08985382318496704,
+      "learning_rate": 3.019599989987234e-06,
+      "loss": 8.5021,
+      "step": 180140
+    },
+    {
+      "epoch": 0.8996479312841769,
+      "grad_norm": 0.08999388664960861,
+      "learning_rate": 3.0180980750456833e-06,
+      "loss": 8.4929,
+      "step": 180150
+    },
+    {
+      "epoch": 0.8996978701091163,
+      "grad_norm": 0.092741958796978,
+      "learning_rate": 3.0165961601041326e-06,
+      "loss": 8.5206,
+      "step": 180160
+    },
+    {
+      "epoch": 0.8997478089340558,
+      "grad_norm": 0.08737698942422867,
+      "learning_rate": 3.0150942451625824e-06,
+      "loss": 8.516,
+      "step": 180170
+    },
+    {
+      "epoch": 0.8997977477589952,
+      "grad_norm": 0.09328185021877289,
+      "learning_rate": 3.013592330221032e-06,
+      "loss": 8.5143,
+      "step": 180180
+    },
+    {
+      "epoch": 0.8998476865839347,
+      "grad_norm": 0.09485041350126266,
+      "learning_rate": 3.0120904152794815e-06,
+      "loss": 8.522,
+      "step": 180190
+    },
+    {
+      "epoch": 0.8998976254088741,
+      "grad_norm": 0.0960816740989685,
+      "learning_rate": 3.010588500337931e-06,
+      "loss": 8.5241,
+      "step": 180200
+    },
+    {
+      "epoch": 0.8999475642338136,
+      "grad_norm": 0.08661402761936188,
+      "learning_rate": 3.00908658539638e-06,
+      "loss": 8.5088,
+      "step": 180210
+    },
+    {
+      "epoch": 0.899997503058753,
+      "grad_norm": 0.08585493266582489,
+      "learning_rate": 3.0075846704548303e-06,
+      "loss": 8.5145,
+      "step": 180220
+    },
+    {
+      "epoch": 0.9000474418836925,
+      "grad_norm": 0.09099698811769485,
+      "learning_rate": 3.0060827555132796e-06,
+      "loss": 8.5141,
+      "step": 180230
+    },
+    {
+      "epoch": 0.9000973807086319,
+      "grad_norm": 0.0931340754032135,
+      "learning_rate": 3.004580840571729e-06,
+      "loss": 8.4905,
+      "step": 180240
+    },
+    {
+      "epoch": 0.9001473195335714,
+      "grad_norm": 0.08374540507793427,
+      "learning_rate": 3.0030789256301783e-06,
+      "loss": 8.5174,
+      "step": 180250
+    },
+    {
+      "epoch": 0.9001972583585108,
+      "grad_norm": 0.09065917134284973,
+      "learning_rate": 3.001577010688628e-06,
+      "loss": 8.5069,
+      "step": 180260
+    },
+    {
+      "epoch": 0.9002471971834503,
+      "grad_norm": 0.09299372136592865,
+      "learning_rate": 3.000075095747078e-06,
+      "loss": 8.5176,
+      "step": 180270
+    },
+    {
+      "epoch": 0.9002971360083897,
+      "grad_norm": 0.08948089927434921,
+      "learning_rate": 2.998573180805527e-06,
+      "loss": 8.4984,
+      "step": 180280
+    },
+    {
+      "epoch": 0.9003470748333292,
+      "grad_norm": 0.0947575643658638,
+      "learning_rate": 2.9970712658639765e-06,
+      "loss": 8.5057,
+      "step": 180290
+    },
+    {
+      "epoch": 0.9003970136582686,
+      "grad_norm": 0.09377171099185944,
+      "learning_rate": 2.9955693509224262e-06,
+      "loss": 8.5169,
+      "step": 180300
+    },
+    {
+      "epoch": 0.900446952483208,
+      "grad_norm": 0.08988793194293976,
+      "learning_rate": 2.9940674359808756e-06,
+      "loss": 8.4968,
+      "step": 180310
+    },
+    {
+      "epoch": 0.9004968913081475,
+      "grad_norm": 0.09155865758657455,
+      "learning_rate": 2.9925655210393253e-06,
+      "loss": 8.5183,
+      "step": 180320
+    },
+    {
+      "epoch": 0.900546830133087,
+      "grad_norm": 0.09261929988861084,
+      "learning_rate": 2.9910636060977747e-06,
+      "loss": 8.4968,
+      "step": 180330
+    },
+    {
+      "epoch": 0.9005967689580264,
+      "grad_norm": 0.09239667654037476,
+      "learning_rate": 2.9895616911562244e-06,
+      "loss": 8.5143,
+      "step": 180340
+    },
+    {
+      "epoch": 0.9006467077829658,
+      "grad_norm": 0.09450769424438477,
+      "learning_rate": 2.9880597762146737e-06,
+      "loss": 8.504,
+      "step": 180350
+    },
+    {
+      "epoch": 0.9006966466079053,
+      "grad_norm": 0.0904548168182373,
+      "learning_rate": 2.986557861273123e-06,
+      "loss": 8.5121,
+      "step": 180360
+    },
+    {
+      "epoch": 0.9007465854328448,
+      "grad_norm": 0.09444373100996017,
+      "learning_rate": 2.985055946331573e-06,
+      "loss": 8.5214,
+      "step": 180370
+    },
+    {
+      "epoch": 0.9007965242577842,
+      "grad_norm": 0.09165211021900177,
+      "learning_rate": 2.9835540313900226e-06,
+      "loss": 8.5255,
+      "step": 180380
+    },
+    {
+      "epoch": 0.9008464630827236,
+      "grad_norm": 0.09335034340620041,
+      "learning_rate": 2.982052116448472e-06,
+      "loss": 8.4975,
+      "step": 180390
+    },
+    {
+      "epoch": 0.9008964019076631,
+      "grad_norm": 0.09302797168493271,
+      "learning_rate": 2.9805502015069212e-06,
+      "loss": 8.5096,
+      "step": 180400
+    },
+    {
+      "epoch": 0.9009463407326026,
+      "grad_norm": 0.09198260307312012,
+      "learning_rate": 2.9790482865653706e-06,
+      "loss": 8.4975,
+      "step": 180410
+    },
+    {
+      "epoch": 0.900996279557542,
+      "grad_norm": 0.08965457230806351,
+      "learning_rate": 2.9775463716238208e-06,
+      "loss": 8.5025,
+      "step": 180420
+    },
+    {
+      "epoch": 0.9010462183824814,
+      "grad_norm": 0.08751804381608963,
+      "learning_rate": 2.97604445668227e-06,
+      "loss": 8.4958,
+      "step": 180430
+    },
+    {
+      "epoch": 0.9010961572074209,
+      "grad_norm": 0.09645684063434601,
+      "learning_rate": 2.9745425417407194e-06,
+      "loss": 8.5389,
+      "step": 180440
+    },
+    {
+      "epoch": 0.9011460960323604,
+      "grad_norm": 0.09055258333683014,
+      "learning_rate": 2.9730406267991687e-06,
+      "loss": 8.501,
+      "step": 180450
+    },
+    {
+      "epoch": 0.9011960348572998,
+      "grad_norm": 0.0941043570637703,
+      "learning_rate": 2.971538711857619e-06,
+      "loss": 8.5124,
+      "step": 180460
+    },
+    {
+      "epoch": 0.9012459736822392,
+      "grad_norm": 0.09053096920251846,
+      "learning_rate": 2.9700367969160683e-06,
+      "loss": 8.4932,
+      "step": 180470
+    },
+    {
+      "epoch": 0.9012959125071787,
+      "grad_norm": 0.08705537766218185,
+      "learning_rate": 2.9685348819745176e-06,
+      "loss": 8.5147,
+      "step": 180480
+    },
+    {
+      "epoch": 0.9013458513321182,
+      "grad_norm": 0.09120982885360718,
+      "learning_rate": 2.967032967032967e-06,
+      "loss": 8.536,
+      "step": 180490
+    },
+    {
+      "epoch": 0.9013957901570576,
+      "grad_norm": 0.09626072645187378,
+      "learning_rate": 2.9655310520914167e-06,
+      "loss": 8.5223,
+      "step": 180500
+    },
+    {
+      "epoch": 0.901445728981997,
+      "grad_norm": 0.09363730996847153,
+      "learning_rate": 2.9640291371498664e-06,
+      "loss": 8.5234,
+      "step": 180510
+    },
+    {
+      "epoch": 0.9014956678069365,
+      "grad_norm": 0.09183649718761444,
+      "learning_rate": 2.9625272222083158e-06,
+      "loss": 8.5249,
+      "step": 180520
+    },
+    {
+      "epoch": 0.901545606631876,
+      "grad_norm": 0.08634661138057709,
+      "learning_rate": 2.961025307266765e-06,
+      "loss": 8.514,
+      "step": 180530
+    },
+    {
+      "epoch": 0.9015955454568154,
+      "grad_norm": 0.08858415484428406,
+      "learning_rate": 2.959523392325215e-06,
+      "loss": 8.5206,
+      "step": 180540
+    },
+    {
+      "epoch": 0.9016454842817548,
+      "grad_norm": 0.08977922797203064,
+      "learning_rate": 2.958021477383664e-06,
+      "loss": 8.5043,
+      "step": 180550
+    },
+    {
+      "epoch": 0.9016954231066943,
+      "grad_norm": 0.0880313590168953,
+      "learning_rate": 2.956519562442114e-06,
+      "loss": 8.5144,
+      "step": 180560
+    },
+    {
+      "epoch": 0.9017453619316338,
+      "grad_norm": 0.08856451511383057,
+      "learning_rate": 2.9550176475005633e-06,
+      "loss": 8.5266,
+      "step": 180570
+    },
+    {
+      "epoch": 0.9017953007565732,
+      "grad_norm": 0.09039363265037537,
+      "learning_rate": 2.953515732559013e-06,
+      "loss": 8.5042,
+      "step": 180580
+    },
+    {
+      "epoch": 0.9018452395815126,
+      "grad_norm": 0.08735457062721252,
+      "learning_rate": 2.9520138176174624e-06,
+      "loss": 8.5157,
+      "step": 180590
+    },
+    {
+      "epoch": 0.9018951784064521,
+      "grad_norm": 0.0854557529091835,
+      "learning_rate": 2.9505119026759117e-06,
+      "loss": 8.514,
+      "step": 180600
+    },
+    {
+      "epoch": 0.9019451172313916,
+      "grad_norm": 0.086369588971138,
+      "learning_rate": 2.9490099877343614e-06,
+      "loss": 8.5198,
+      "step": 180610
+    },
+    {
+      "epoch": 0.901995056056331,
+      "grad_norm": 0.096776083111763,
+      "learning_rate": 2.947508072792811e-06,
+      "loss": 8.4991,
+      "step": 180620
+    },
+    {
+      "epoch": 0.9020449948812704,
+      "grad_norm": 0.08912768959999084,
+      "learning_rate": 2.9460061578512605e-06,
+      "loss": 8.5191,
+      "step": 180630
+    },
+    {
+      "epoch": 0.9020949337062099,
+      "grad_norm": 0.09278588742017746,
+      "learning_rate": 2.94450424290971e-06,
+      "loss": 8.5229,
+      "step": 180640
+    },
+    {
+      "epoch": 0.9021448725311494,
+      "grad_norm": 0.09282492101192474,
+      "learning_rate": 2.943002327968159e-06,
+      "loss": 8.5159,
+      "step": 180650
+    },
+    {
+      "epoch": 0.9021948113560888,
+      "grad_norm": 0.08819557726383209,
+      "learning_rate": 2.9415004130266094e-06,
+      "loss": 8.5337,
+      "step": 180660
+    },
+    {
+      "epoch": 0.9022447501810282,
+      "grad_norm": 0.08956322073936462,
+      "learning_rate": 2.9399984980850587e-06,
+      "loss": 8.5141,
+      "step": 180670
+    },
+    {
+      "epoch": 0.9022946890059677,
+      "grad_norm": 0.0941721498966217,
+      "learning_rate": 2.938496583143508e-06,
+      "loss": 8.5227,
+      "step": 180680
+    },
+    {
+      "epoch": 0.9023446278309072,
+      "grad_norm": 0.09089160710573196,
+      "learning_rate": 2.9369946682019574e-06,
+      "loss": 8.5036,
+      "step": 180690
+    },
+    {
+      "epoch": 0.9023945666558466,
+      "grad_norm": 0.09959226101636887,
+      "learning_rate": 2.935492753260407e-06,
+      "loss": 8.4941,
+      "step": 180700
+    },
+    {
+      "epoch": 0.902444505480786,
+      "grad_norm": 0.0919126570224762,
+      "learning_rate": 2.933990838318857e-06,
+      "loss": 8.4893,
+      "step": 180710
+    },
+    {
+      "epoch": 0.9024944443057255,
+      "grad_norm": 0.0832803025841713,
+      "learning_rate": 2.932488923377306e-06,
+      "loss": 8.5169,
+      "step": 180720
+    },
+    {
+      "epoch": 0.902544383130665,
+      "grad_norm": 0.09311427175998688,
+      "learning_rate": 2.9309870084357555e-06,
+      "loss": 8.4991,
+      "step": 180730
+    },
+    {
+      "epoch": 0.9025943219556044,
+      "grad_norm": 0.09170939028263092,
+      "learning_rate": 2.9294850934942053e-06,
+      "loss": 8.4904,
+      "step": 180740
+    },
+    {
+      "epoch": 0.9026442607805438,
+      "grad_norm": 0.09490250796079636,
+      "learning_rate": 2.9279831785526546e-06,
+      "loss": 8.5021,
+      "step": 180750
+    },
+    {
+      "epoch": 0.9026941996054832,
+      "grad_norm": 0.09353826195001602,
+      "learning_rate": 2.9264812636111044e-06,
+      "loss": 8.517,
+      "step": 180760
+    },
+    {
+      "epoch": 0.9027441384304228,
+      "grad_norm": 0.09115730971097946,
+      "learning_rate": 2.9249793486695537e-06,
+      "loss": 8.4986,
+      "step": 180770
+    },
+    {
+      "epoch": 0.9027940772553622,
+      "grad_norm": 0.09605268388986588,
+      "learning_rate": 2.9234774337280035e-06,
+      "loss": 8.5088,
+      "step": 180780
+    },
+    {
+      "epoch": 0.9028440160803016,
+      "grad_norm": 0.09390175342559814,
+      "learning_rate": 2.921975518786453e-06,
+      "loss": 8.5063,
+      "step": 180790
+    },
+    {
+      "epoch": 0.902893954905241,
+      "grad_norm": 0.09228114783763885,
+      "learning_rate": 2.920473603844902e-06,
+      "loss": 8.5095,
+      "step": 180800
+    },
+    {
+      "epoch": 0.9029438937301806,
+      "grad_norm": 0.09767407178878784,
+      "learning_rate": 2.918971688903352e-06,
+      "loss": 8.5066,
+      "step": 180810
+    },
+    {
+      "epoch": 0.90299383255512,
+      "grad_norm": 0.08856954425573349,
+      "learning_rate": 2.9174697739618016e-06,
+      "loss": 8.5368,
+      "step": 180820
+    },
+    {
+      "epoch": 0.9030437713800594,
+      "grad_norm": 0.09355857968330383,
+      "learning_rate": 2.915967859020251e-06,
+      "loss": 8.5208,
+      "step": 180830
+    },
+    {
+      "epoch": 0.9030937102049988,
+      "grad_norm": 0.08886077255010605,
+      "learning_rate": 2.9144659440787003e-06,
+      "loss": 8.5253,
+      "step": 180840
+    },
+    {
+      "epoch": 0.9031436490299384,
+      "grad_norm": 0.09166674315929413,
+      "learning_rate": 2.9129640291371496e-06,
+      "loss": 8.5249,
+      "step": 180850
+    },
+    {
+      "epoch": 0.9031935878548778,
+      "grad_norm": 0.08610928058624268,
+      "learning_rate": 2.9114621141956e-06,
+      "loss": 8.5264,
+      "step": 180860
+    },
+    {
+      "epoch": 0.9032435266798172,
+      "grad_norm": 0.08989940583705902,
+      "learning_rate": 2.909960199254049e-06,
+      "loss": 8.5241,
+      "step": 180870
+    },
+    {
+      "epoch": 0.9032934655047566,
+      "grad_norm": 0.09518139809370041,
+      "learning_rate": 2.9084582843124985e-06,
+      "loss": 8.5228,
+      "step": 180880
+    },
+    {
+      "epoch": 0.9033434043296962,
+      "grad_norm": 0.09098067134618759,
+      "learning_rate": 2.906956369370948e-06,
+      "loss": 8.5047,
+      "step": 180890
+    },
+    {
+      "epoch": 0.9033933431546356,
+      "grad_norm": 0.0919727236032486,
+      "learning_rate": 2.9054544544293976e-06,
+      "loss": 8.5145,
+      "step": 180900
+    },
+    {
+      "epoch": 0.903443281979575,
+      "grad_norm": 0.09157610684633255,
+      "learning_rate": 2.9039525394878473e-06,
+      "loss": 8.4891,
+      "step": 180910
+    },
+    {
+      "epoch": 0.9034932208045144,
+      "grad_norm": 0.09556107968091965,
+      "learning_rate": 2.9024506245462966e-06,
+      "loss": 8.5144,
+      "step": 180920
+    },
+    {
+      "epoch": 0.903543159629454,
+      "grad_norm": 0.08390878885984421,
+      "learning_rate": 2.900948709604746e-06,
+      "loss": 8.5181,
+      "step": 180930
+    },
+    {
+      "epoch": 0.9035930984543934,
+      "grad_norm": 0.09191958606243134,
+      "learning_rate": 2.8994467946631957e-06,
+      "loss": 8.5095,
+      "step": 180940
+    },
+    {
+      "epoch": 0.9036430372793328,
+      "grad_norm": 0.09150879085063934,
+      "learning_rate": 2.897944879721645e-06,
+      "loss": 8.5215,
+      "step": 180950
+    },
+    {
+      "epoch": 0.9036929761042722,
+      "grad_norm": 0.0931938886642456,
+      "learning_rate": 2.896442964780095e-06,
+      "loss": 8.4941,
+      "step": 180960
+    },
+    {
+      "epoch": 0.9037429149292117,
+      "grad_norm": 0.09147433936595917,
+      "learning_rate": 2.894941049838544e-06,
+      "loss": 8.5119,
+      "step": 180970
+    },
+    {
+      "epoch": 0.9037928537541512,
+      "grad_norm": 0.09065255522727966,
+      "learning_rate": 2.893439134896994e-06,
+      "loss": 8.5166,
+      "step": 180980
+    },
+    {
+      "epoch": 0.9038427925790906,
+      "grad_norm": 0.0912211537361145,
+      "learning_rate": 2.8919372199554432e-06,
+      "loss": 8.4907,
+      "step": 180990
+    },
+    {
+      "epoch": 0.90389273140403,
+      "grad_norm": 0.09441815316677094,
+      "learning_rate": 2.890435305013893e-06,
+      "loss": 8.5097,
+      "step": 181000
+    },
+    {
+      "epoch": 0.9039426702289695,
+      "grad_norm": 0.08536705374717712,
+      "learning_rate": 2.8889333900723423e-06,
+      "loss": 8.5144,
+      "step": 181010
+    },
+    {
+      "epoch": 0.903992609053909,
+      "grad_norm": 0.08860597014427185,
+      "learning_rate": 2.887431475130792e-06,
+      "loss": 8.5068,
+      "step": 181020
+    },
+    {
+      "epoch": 0.9040425478788484,
+      "grad_norm": 0.09018053114414215,
+      "learning_rate": 2.8859295601892414e-06,
+      "loss": 8.5083,
+      "step": 181030
+    },
+    {
+      "epoch": 0.9040924867037878,
+      "grad_norm": 0.08966843783855438,
+      "learning_rate": 2.8844276452476907e-06,
+      "loss": 8.4964,
+      "step": 181040
+    },
+    {
+      "epoch": 0.9041424255287273,
+      "grad_norm": 0.08810800313949585,
+      "learning_rate": 2.8829257303061405e-06,
+      "loss": 8.5215,
+      "step": 181050
+    },
+    {
+      "epoch": 0.9041923643536668,
+      "grad_norm": 0.0902119129896164,
+      "learning_rate": 2.8814238153645902e-06,
+      "loss": 8.5246,
+      "step": 181060
+    },
+    {
+      "epoch": 0.9042423031786062,
+      "grad_norm": 0.09436272829771042,
+      "learning_rate": 2.8799219004230396e-06,
+      "loss": 8.4939,
+      "step": 181070
+    },
+    {
+      "epoch": 0.9042922420035456,
+      "grad_norm": 0.09024659544229507,
+      "learning_rate": 2.878419985481489e-06,
+      "loss": 8.5095,
+      "step": 181080
+    },
+    {
+      "epoch": 0.904342180828485,
+      "grad_norm": 0.089921735227108,
+      "learning_rate": 2.8769180705399382e-06,
+      "loss": 8.5008,
+      "step": 181090
+    },
+    {
+      "epoch": 0.9043921196534246,
+      "grad_norm": 0.08670993894338608,
+      "learning_rate": 2.8754161555983884e-06,
+      "loss": 8.5017,
+      "step": 181100
+    },
+    {
+      "epoch": 0.904442058478364,
+      "grad_norm": 0.09145168215036392,
+      "learning_rate": 2.8739142406568377e-06,
+      "loss": 8.5277,
+      "step": 181110
+    },
+    {
+      "epoch": 0.9044919973033034,
+      "grad_norm": 0.09542462974786758,
+      "learning_rate": 2.872412325715287e-06,
+      "loss": 8.5008,
+      "step": 181120
+    },
+    {
+      "epoch": 0.9045419361282429,
+      "grad_norm": 0.09122487902641296,
+      "learning_rate": 2.8709104107737364e-06,
+      "loss": 8.5092,
+      "step": 181130
+    },
+    {
+      "epoch": 0.9045918749531824,
+      "grad_norm": 0.08779449760913849,
+      "learning_rate": 2.869408495832186e-06,
+      "loss": 8.4968,
+      "step": 181140
+    },
+    {
+      "epoch": 0.9046418137781218,
+      "grad_norm": 0.08945740014314651,
+      "learning_rate": 2.867906580890636e-06,
+      "loss": 8.5206,
+      "step": 181150
+    },
+    {
+      "epoch": 0.9046917526030612,
+      "grad_norm": 0.08815411478281021,
+      "learning_rate": 2.8664046659490852e-06,
+      "loss": 8.512,
+      "step": 181160
+    },
+    {
+      "epoch": 0.9047416914280006,
+      "grad_norm": 0.09432817250490189,
+      "learning_rate": 2.8649027510075346e-06,
+      "loss": 8.5159,
+      "step": 181170
+    },
+    {
+      "epoch": 0.9047916302529402,
+      "grad_norm": 0.09125339239835739,
+      "learning_rate": 2.8634008360659843e-06,
+      "loss": 8.5162,
+      "step": 181180
+    },
+    {
+      "epoch": 0.9048415690778796,
+      "grad_norm": 0.08769851177930832,
+      "learning_rate": 2.8618989211244337e-06,
+      "loss": 8.5112,
+      "step": 181190
+    },
+    {
+      "epoch": 0.904891507902819,
+      "grad_norm": 0.09156162291765213,
+      "learning_rate": 2.8603970061828834e-06,
+      "loss": 8.5177,
+      "step": 181200
+    },
+    {
+      "epoch": 0.9049414467277584,
+      "grad_norm": 0.095248281955719,
+      "learning_rate": 2.8588950912413328e-06,
+      "loss": 8.4942,
+      "step": 181210
+    },
+    {
+      "epoch": 0.904991385552698,
+      "grad_norm": 0.09011649340391159,
+      "learning_rate": 2.8573931762997825e-06,
+      "loss": 8.5229,
+      "step": 181220
+    },
+    {
+      "epoch": 0.9050413243776374,
+      "grad_norm": 0.08675964921712875,
+      "learning_rate": 2.855891261358232e-06,
+      "loss": 8.5231,
+      "step": 181230
+    },
+    {
+      "epoch": 0.9050912632025768,
+      "grad_norm": 0.0876312181353569,
+      "learning_rate": 2.854389346416681e-06,
+      "loss": 8.5145,
+      "step": 181240
+    },
+    {
+      "epoch": 0.9051412020275162,
+      "grad_norm": 0.0970059260725975,
+      "learning_rate": 2.852887431475131e-06,
+      "loss": 8.5047,
+      "step": 181250
+    },
+    {
+      "epoch": 0.9051911408524558,
+      "grad_norm": 0.0930013656616211,
+      "learning_rate": 2.8513855165335807e-06,
+      "loss": 8.4996,
+      "step": 181260
+    },
+    {
+      "epoch": 0.9052410796773952,
+      "grad_norm": 0.09447172284126282,
+      "learning_rate": 2.84988360159203e-06,
+      "loss": 8.5009,
+      "step": 181270
+    },
+    {
+      "epoch": 0.9052910185023346,
+      "grad_norm": 0.0933113843202591,
+      "learning_rate": 2.8483816866504793e-06,
+      "loss": 8.4953,
+      "step": 181280
+    },
+    {
+      "epoch": 0.905340957327274,
+      "grad_norm": 0.09435321390628815,
+      "learning_rate": 2.8468797717089287e-06,
+      "loss": 8.4999,
+      "step": 181290
+    },
+    {
+      "epoch": 0.9053908961522136,
+      "grad_norm": 0.08828049153089523,
+      "learning_rate": 2.8453778567673784e-06,
+      "loss": 8.5145,
+      "step": 181300
+    },
+    {
+      "epoch": 0.905440834977153,
+      "grad_norm": 0.08913722634315491,
+      "learning_rate": 2.843875941825828e-06,
+      "loss": 8.4944,
+      "step": 181310
+    },
+    {
+      "epoch": 0.9054907738020924,
+      "grad_norm": 0.09632635116577148,
+      "learning_rate": 2.8423740268842775e-06,
+      "loss": 8.493,
+      "step": 181320
+    },
+    {
+      "epoch": 0.9055407126270318,
+      "grad_norm": 0.0897509828209877,
+      "learning_rate": 2.840872111942727e-06,
+      "loss": 8.4977,
+      "step": 181330
+    },
+    {
+      "epoch": 0.9055906514519714,
+      "grad_norm": 0.09208657592535019,
+      "learning_rate": 2.839370197001176e-06,
+      "loss": 8.5314,
+      "step": 181340
+    },
+    {
+      "epoch": 0.9056405902769108,
+      "grad_norm": 0.09006085991859436,
+      "learning_rate": 2.8378682820596264e-06,
+      "loss": 8.5163,
+      "step": 181350
+    },
+    {
+      "epoch": 0.9056905291018502,
+      "grad_norm": 0.09176279604434967,
+      "learning_rate": 2.8363663671180757e-06,
+      "loss": 8.5148,
+      "step": 181360
+    },
+    {
+      "epoch": 0.9057404679267896,
+      "grad_norm": 0.09576241672039032,
+      "learning_rate": 2.834864452176525e-06,
+      "loss": 8.5346,
+      "step": 181370
+    },
+    {
+      "epoch": 0.9057904067517292,
+      "grad_norm": 0.09277977049350739,
+      "learning_rate": 2.8333625372349744e-06,
+      "loss": 8.5067,
+      "step": 181380
+    },
+    {
+      "epoch": 0.9058403455766686,
+      "grad_norm": 0.0882072001695633,
+      "learning_rate": 2.831860622293424e-06,
+      "loss": 8.5046,
+      "step": 181390
+    },
+    {
+      "epoch": 0.905890284401608,
+      "grad_norm": 0.08874956518411636,
+      "learning_rate": 2.830358707351874e-06,
+      "loss": 8.5382,
+      "step": 181400
+    },
+    {
+      "epoch": 0.9059402232265474,
+      "grad_norm": 0.08898142725229263,
+      "learning_rate": 2.828856792410323e-06,
+      "loss": 8.5067,
+      "step": 181410
+    },
+    {
+      "epoch": 0.905990162051487,
+      "grad_norm": 0.08590088039636612,
+      "learning_rate": 2.8273548774687725e-06,
+      "loss": 8.5023,
+      "step": 181420
+    },
+    {
+      "epoch": 0.9060401008764264,
+      "grad_norm": 0.08650851249694824,
+      "learning_rate": 2.8258529625272223e-06,
+      "loss": 8.5202,
+      "step": 181430
+    },
+    {
+      "epoch": 0.9060900397013658,
+      "grad_norm": 0.09431696683168411,
+      "learning_rate": 2.8243510475856716e-06,
+      "loss": 8.5401,
+      "step": 181440
+    },
+    {
+      "epoch": 0.9061399785263052,
+      "grad_norm": 0.09654943645000458,
+      "learning_rate": 2.8228491326441214e-06,
+      "loss": 8.5078,
+      "step": 181450
+    },
+    {
+      "epoch": 0.9061899173512448,
+      "grad_norm": 0.08821980655193329,
+      "learning_rate": 2.8213472177025707e-06,
+      "loss": 8.4975,
+      "step": 181460
+    },
+    {
+      "epoch": 0.9062398561761842,
+      "grad_norm": 0.0911247506737709,
+      "learning_rate": 2.8198453027610204e-06,
+      "loss": 8.5129,
+      "step": 181470
+    },
+    {
+      "epoch": 0.9062897950011236,
+      "grad_norm": 0.08649583905935287,
+      "learning_rate": 2.8183433878194698e-06,
+      "loss": 8.5025,
+      "step": 181480
+    },
+    {
+      "epoch": 0.906339733826063,
+      "grad_norm": 0.09082172811031342,
+      "learning_rate": 2.816841472877919e-06,
+      "loss": 8.5101,
+      "step": 181490
+    },
+    {
+      "epoch": 0.9063896726510026,
+      "grad_norm": 0.09340551495552063,
+      "learning_rate": 2.815339557936369e-06,
+      "loss": 8.5192,
+      "step": 181500
+    },
+    {
+      "epoch": 0.906439611475942,
+      "grad_norm": 0.09474127739667892,
+      "learning_rate": 2.8138376429948186e-06,
+      "loss": 8.5036,
+      "step": 181510
+    },
+    {
+      "epoch": 0.9064895503008814,
+      "grad_norm": 0.08885298669338226,
+      "learning_rate": 2.812335728053268e-06,
+      "loss": 8.4964,
+      "step": 181520
+    },
+    {
+      "epoch": 0.9065394891258208,
+      "grad_norm": 0.09157950431108475,
+      "learning_rate": 2.8108338131117173e-06,
+      "loss": 8.4965,
+      "step": 181530
+    },
+    {
+      "epoch": 0.9065894279507604,
+      "grad_norm": 0.08861027657985687,
+      "learning_rate": 2.809331898170167e-06,
+      "loss": 8.518,
+      "step": 181540
+    },
+    {
+      "epoch": 0.9066393667756998,
+      "grad_norm": 0.0883537232875824,
+      "learning_rate": 2.807829983228617e-06,
+      "loss": 8.5024,
+      "step": 181550
+    },
+    {
+      "epoch": 0.9066893056006392,
+      "grad_norm": 0.0927857756614685,
+      "learning_rate": 2.806328068287066e-06,
+      "loss": 8.4976,
+      "step": 181560
+    },
+    {
+      "epoch": 0.9067392444255786,
+      "grad_norm": 0.09182663261890411,
+      "learning_rate": 2.8048261533455155e-06,
+      "loss": 8.4891,
+      "step": 181570
+    },
+    {
+      "epoch": 0.9067891832505182,
+      "grad_norm": 0.09595274925231934,
+      "learning_rate": 2.803324238403965e-06,
+      "loss": 8.5132,
+      "step": 181580
+    },
+    {
+      "epoch": 0.9068391220754576,
+      "grad_norm": 0.09273464232683182,
+      "learning_rate": 2.801822323462415e-06,
+      "loss": 8.5065,
+      "step": 181590
+    },
+    {
+      "epoch": 0.906889060900397,
+      "grad_norm": 0.09522940963506699,
+      "learning_rate": 2.8003204085208643e-06,
+      "loss": 8.5137,
+      "step": 181600
+    },
+    {
+      "epoch": 0.9069389997253364,
+      "grad_norm": 0.08875519037246704,
+      "learning_rate": 2.7988184935793136e-06,
+      "loss": 8.5101,
+      "step": 181610
+    },
+    {
+      "epoch": 0.906988938550276,
+      "grad_norm": 0.09621170908212662,
+      "learning_rate": 2.797316578637763e-06,
+      "loss": 8.5001,
+      "step": 181620
+    },
+    {
+      "epoch": 0.9070388773752154,
+      "grad_norm": 0.09134541451931,
+      "learning_rate": 2.7958146636962127e-06,
+      "loss": 8.5145,
+      "step": 181630
+    },
+    {
+      "epoch": 0.9070888162001548,
+      "grad_norm": 0.09547372162342072,
+      "learning_rate": 2.7943127487546625e-06,
+      "loss": 8.4756,
+      "step": 181640
+    },
+    {
+      "epoch": 0.9071387550250942,
+      "grad_norm": 0.09366821497678757,
+      "learning_rate": 2.792810833813112e-06,
+      "loss": 8.5362,
+      "step": 181650
+    },
+    {
+      "epoch": 0.9071886938500338,
+      "grad_norm": 0.09829097241163254,
+      "learning_rate": 2.791308918871561e-06,
+      "loss": 8.5254,
+      "step": 181660
+    },
+    {
+      "epoch": 0.9072386326749732,
+      "grad_norm": 0.09187762439250946,
+      "learning_rate": 2.789807003930011e-06,
+      "loss": 8.5192,
+      "step": 181670
+    },
+    {
+      "epoch": 0.9072885714999126,
+      "grad_norm": 0.09273143857717514,
+      "learning_rate": 2.7883050889884602e-06,
+      "loss": 8.4973,
+      "step": 181680
+    },
+    {
+      "epoch": 0.907338510324852,
+      "grad_norm": 0.09177732467651367,
+      "learning_rate": 2.78680317404691e-06,
+      "loss": 8.5055,
+      "step": 181690
+    },
+    {
+      "epoch": 0.9073884491497916,
+      "grad_norm": 0.089385986328125,
+      "learning_rate": 2.7853012591053593e-06,
+      "loss": 8.5095,
+      "step": 181700
+    },
+    {
+      "epoch": 0.907438387974731,
+      "grad_norm": 0.09638892114162445,
+      "learning_rate": 2.783799344163809e-06,
+      "loss": 8.4961,
+      "step": 181710
+    },
+    {
+      "epoch": 0.9074883267996704,
+      "grad_norm": 0.08958081156015396,
+      "learning_rate": 2.7822974292222584e-06,
+      "loss": 8.5128,
+      "step": 181720
+    },
+    {
+      "epoch": 0.9075382656246098,
+      "grad_norm": 0.09109731018543243,
+      "learning_rate": 2.7807955142807077e-06,
+      "loss": 8.5253,
+      "step": 181730
+    },
+    {
+      "epoch": 0.9075882044495494,
+      "grad_norm": 0.08715473860502243,
+      "learning_rate": 2.7792935993391575e-06,
+      "loss": 8.5185,
+      "step": 181740
+    },
+    {
+      "epoch": 0.9076381432744888,
+      "grad_norm": 0.09187250584363937,
+      "learning_rate": 2.7777916843976072e-06,
+      "loss": 8.5065,
+      "step": 181750
+    },
+    {
+      "epoch": 0.9076880820994282,
+      "grad_norm": 0.09143325686454773,
+      "learning_rate": 2.7762897694560566e-06,
+      "loss": 8.4969,
+      "step": 181760
+    },
+    {
+      "epoch": 0.9077380209243676,
+      "grad_norm": 0.08711639046669006,
+      "learning_rate": 2.774787854514506e-06,
+      "loss": 8.5134,
+      "step": 181770
+    },
+    {
+      "epoch": 0.9077879597493071,
+      "grad_norm": 0.09443769603967667,
+      "learning_rate": 2.7732859395729552e-06,
+      "loss": 8.5037,
+      "step": 181780
+    },
+    {
+      "epoch": 0.9078378985742466,
+      "grad_norm": 0.09021743386983871,
+      "learning_rate": 2.7717840246314054e-06,
+      "loss": 8.4901,
+      "step": 181790
+    },
+    {
+      "epoch": 0.907887837399186,
+      "grad_norm": 0.08807482570409775,
+      "learning_rate": 2.7702821096898547e-06,
+      "loss": 8.529,
+      "step": 181800
+    },
+    {
+      "epoch": 0.9079377762241254,
+      "grad_norm": 0.09367617219686508,
+      "learning_rate": 2.768780194748304e-06,
+      "loss": 8.534,
+      "step": 181810
+    },
+    {
+      "epoch": 0.907987715049065,
+      "grad_norm": 0.09450183063745499,
+      "learning_rate": 2.7672782798067534e-06,
+      "loss": 8.5267,
+      "step": 181820
+    },
+    {
+      "epoch": 0.9080376538740044,
+      "grad_norm": 0.09131249040365219,
+      "learning_rate": 2.765776364865203e-06,
+      "loss": 8.5342,
+      "step": 181830
+    },
+    {
+      "epoch": 0.9080875926989438,
+      "grad_norm": 0.09016092121601105,
+      "learning_rate": 2.764274449923653e-06,
+      "loss": 8.5188,
+      "step": 181840
+    },
+    {
+      "epoch": 0.9081375315238832,
+      "grad_norm": 0.09213612228631973,
+      "learning_rate": 2.7627725349821022e-06,
+      "loss": 8.5291,
+      "step": 181850
+    },
+    {
+      "epoch": 0.9081874703488227,
+      "grad_norm": 0.09062368422746658,
+      "learning_rate": 2.7612706200405516e-06,
+      "loss": 8.5298,
+      "step": 181860
+    },
+    {
+      "epoch": 0.9082374091737622,
+      "grad_norm": 0.08791705220937729,
+      "learning_rate": 2.7597687050990013e-06,
+      "loss": 8.5115,
+      "step": 181870
+    },
+    {
+      "epoch": 0.9082873479987016,
+      "grad_norm": 0.09510977566242218,
+      "learning_rate": 2.7582667901574507e-06,
+      "loss": 8.5174,
+      "step": 181880
+    },
+    {
+      "epoch": 0.908337286823641,
+      "grad_norm": 0.0934426486492157,
+      "learning_rate": 2.7567648752159004e-06,
+      "loss": 8.5261,
+      "step": 181890
+    },
+    {
+      "epoch": 0.9083872256485805,
+      "grad_norm": 0.09243565052747726,
+      "learning_rate": 2.7552629602743497e-06,
+      "loss": 8.5155,
+      "step": 181900
+    },
+    {
+      "epoch": 0.90843716447352,
+      "grad_norm": 0.09363988786935806,
+      "learning_rate": 2.7537610453327995e-06,
+      "loss": 8.5313,
+      "step": 181910
+    },
+    {
+      "epoch": 0.9084871032984594,
+      "grad_norm": 0.09643617272377014,
+      "learning_rate": 2.752259130391249e-06,
+      "loss": 8.5019,
+      "step": 181920
+    },
+    {
+      "epoch": 0.9085370421233988,
+      "grad_norm": 0.09652823954820633,
+      "learning_rate": 2.750757215449698e-06,
+      "loss": 8.5114,
+      "step": 181930
+    },
+    {
+      "epoch": 0.9085869809483383,
+      "grad_norm": 0.0902838185429573,
+      "learning_rate": 2.749255300508148e-06,
+      "loss": 8.5336,
+      "step": 181940
+    },
+    {
+      "epoch": 0.9086369197732778,
+      "grad_norm": 0.0898505225777626,
+      "learning_rate": 2.7477533855665977e-06,
+      "loss": 8.5068,
+      "step": 181950
+    },
+    {
+      "epoch": 0.9086868585982172,
+      "grad_norm": 0.0900585874915123,
+      "learning_rate": 2.746251470625047e-06,
+      "loss": 8.4995,
+      "step": 181960
+    },
+    {
+      "epoch": 0.9087367974231566,
+      "grad_norm": 0.09357380867004395,
+      "learning_rate": 2.7447495556834963e-06,
+      "loss": 8.5153,
+      "step": 181970
+    },
+    {
+      "epoch": 0.908786736248096,
+      "grad_norm": 0.08812058717012405,
+      "learning_rate": 2.7432476407419457e-06,
+      "loss": 8.5055,
+      "step": 181980
+    },
+    {
+      "epoch": 0.9088366750730356,
+      "grad_norm": 0.08812303096055984,
+      "learning_rate": 2.741745725800396e-06,
+      "loss": 8.5064,
+      "step": 181990
+    },
+    {
+      "epoch": 0.908886613897975,
+      "grad_norm": 0.09171146154403687,
+      "learning_rate": 2.740243810858845e-06,
+      "loss": 8.5135,
+      "step": 182000
+    },
+    {
+      "epoch": 0.9089365527229144,
+      "grad_norm": 0.09255418926477432,
+      "learning_rate": 2.7387418959172945e-06,
+      "loss": 8.5029,
+      "step": 182010
+    },
+    {
+      "epoch": 0.9089864915478538,
+      "grad_norm": 0.08891849964857101,
+      "learning_rate": 2.737239980975744e-06,
+      "loss": 8.5038,
+      "step": 182020
+    },
+    {
+      "epoch": 0.9090364303727934,
+      "grad_norm": 0.09265652298927307,
+      "learning_rate": 2.7357380660341936e-06,
+      "loss": 8.5129,
+      "step": 182030
+    },
+    {
+      "epoch": 0.9090863691977328,
+      "grad_norm": 0.08749819546937943,
+      "learning_rate": 2.7342361510926433e-06,
+      "loss": 8.5183,
+      "step": 182040
+    },
+    {
+      "epoch": 0.9091363080226722,
+      "grad_norm": 0.08859850466251373,
+      "learning_rate": 2.7327342361510927e-06,
+      "loss": 8.5137,
+      "step": 182050
+    },
+    {
+      "epoch": 0.9091862468476116,
+      "grad_norm": 0.09206672012805939,
+      "learning_rate": 2.731232321209542e-06,
+      "loss": 8.5156,
+      "step": 182060
+    },
+    {
+      "epoch": 0.9092361856725512,
+      "grad_norm": 0.08951518684625626,
+      "learning_rate": 2.7297304062679918e-06,
+      "loss": 8.4979,
+      "step": 182070
+    },
+    {
+      "epoch": 0.9092861244974906,
+      "grad_norm": 0.09024099260568619,
+      "learning_rate": 2.7282284913264415e-06,
+      "loss": 8.5223,
+      "step": 182080
+    },
+    {
+      "epoch": 0.90933606332243,
+      "grad_norm": 0.08952950686216354,
+      "learning_rate": 2.726726576384891e-06,
+      "loss": 8.4998,
+      "step": 182090
+    },
+    {
+      "epoch": 0.9093860021473694,
+      "grad_norm": 0.09278613328933716,
+      "learning_rate": 2.72522466144334e-06,
+      "loss": 8.5078,
+      "step": 182100
+    },
+    {
+      "epoch": 0.909435940972309,
+      "grad_norm": 0.08384092897176743,
+      "learning_rate": 2.72372274650179e-06,
+      "loss": 8.5027,
+      "step": 182110
+    },
+    {
+      "epoch": 0.9094858797972484,
+      "grad_norm": 0.08967365324497223,
+      "learning_rate": 2.7222208315602393e-06,
+      "loss": 8.5171,
+      "step": 182120
+    },
+    {
+      "epoch": 0.9095358186221878,
+      "grad_norm": 0.09536395221948624,
+      "learning_rate": 2.720718916618689e-06,
+      "loss": 8.5055,
+      "step": 182130
+    },
+    {
+      "epoch": 0.9095857574471272,
+      "grad_norm": 0.09401460736989975,
+      "learning_rate": 2.7192170016771384e-06,
+      "loss": 8.5148,
+      "step": 182140
+    },
+    {
+      "epoch": 0.9096356962720668,
+      "grad_norm": 0.0905858501791954,
+      "learning_rate": 2.717715086735588e-06,
+      "loss": 8.5112,
+      "step": 182150
+    },
+    {
+      "epoch": 0.9096856350970062,
+      "grad_norm": 0.09035810828208923,
+      "learning_rate": 2.7162131717940374e-06,
+      "loss": 8.4981,
+      "step": 182160
+    },
+    {
+      "epoch": 0.9097355739219456,
+      "grad_norm": 0.09156506508588791,
+      "learning_rate": 2.7147112568524868e-06,
+      "loss": 8.5072,
+      "step": 182170
+    },
+    {
+      "epoch": 0.909785512746885,
+      "grad_norm": 0.09411493688821793,
+      "learning_rate": 2.7132093419109365e-06,
+      "loss": 8.5061,
+      "step": 182180
+    },
+    {
+      "epoch": 0.9098354515718245,
+      "grad_norm": 0.08917105942964554,
+      "learning_rate": 2.7117074269693863e-06,
+      "loss": 8.5034,
+      "step": 182190
+    },
+    {
+      "epoch": 0.909885390396764,
+      "grad_norm": 0.08715265989303589,
+      "learning_rate": 2.7102055120278356e-06,
+      "loss": 8.507,
+      "step": 182200
+    },
+    {
+      "epoch": 0.9099353292217034,
+      "grad_norm": 0.09055981785058975,
+      "learning_rate": 2.708703597086285e-06,
+      "loss": 8.5144,
+      "step": 182210
+    },
+    {
+      "epoch": 0.9099852680466428,
+      "grad_norm": 0.09121181070804596,
+      "learning_rate": 2.7072016821447343e-06,
+      "loss": 8.5191,
+      "step": 182220
+    },
+    {
+      "epoch": 0.9100352068715823,
+      "grad_norm": 0.09437250345945358,
+      "learning_rate": 2.7056997672031845e-06,
+      "loss": 8.5132,
+      "step": 182230
+    },
+    {
+      "epoch": 0.9100851456965218,
+      "grad_norm": 0.09657744318246841,
+      "learning_rate": 2.7041978522616338e-06,
+      "loss": 8.5328,
+      "step": 182240
+    },
+    {
+      "epoch": 0.9101350845214612,
+      "grad_norm": 0.09253308176994324,
+      "learning_rate": 2.702695937320083e-06,
+      "loss": 8.5081,
+      "step": 182250
+    },
+    {
+      "epoch": 0.9101850233464006,
+      "grad_norm": 0.09130608290433884,
+      "learning_rate": 2.7011940223785324e-06,
+      "loss": 8.4981,
+      "step": 182260
+    },
+    {
+      "epoch": 0.9102349621713401,
+      "grad_norm": 0.08788429945707321,
+      "learning_rate": 2.699692107436982e-06,
+      "loss": 8.4895,
+      "step": 182270
+    },
+    {
+      "epoch": 0.9102849009962796,
+      "grad_norm": 0.09099657833576202,
+      "learning_rate": 2.698190192495432e-06,
+      "loss": 8.508,
+      "step": 182280
+    },
+    {
+      "epoch": 0.910334839821219,
+      "grad_norm": 0.08800596743822098,
+      "learning_rate": 2.6966882775538813e-06,
+      "loss": 8.5297,
+      "step": 182290
+    },
+    {
+      "epoch": 0.9103847786461584,
+      "grad_norm": 0.09730488806962967,
+      "learning_rate": 2.6951863626123306e-06,
+      "loss": 8.5148,
+      "step": 182300
+    },
+    {
+      "epoch": 0.9104347174710979,
+      "grad_norm": 0.09206937998533249,
+      "learning_rate": 2.6936844476707804e-06,
+      "loss": 8.5049,
+      "step": 182310
+    },
+    {
+      "epoch": 0.9104846562960374,
+      "grad_norm": 0.08866451680660248,
+      "learning_rate": 2.6921825327292297e-06,
+      "loss": 8.512,
+      "step": 182320
+    },
+    {
+      "epoch": 0.9105345951209768,
+      "grad_norm": 0.09090999513864517,
+      "learning_rate": 2.6906806177876795e-06,
+      "loss": 8.4796,
+      "step": 182330
+    },
+    {
+      "epoch": 0.9105845339459162,
+      "grad_norm": 0.08853746205568314,
+      "learning_rate": 2.689178702846129e-06,
+      "loss": 8.5107,
+      "step": 182340
+    },
+    {
+      "epoch": 0.9106344727708557,
+      "grad_norm": 0.09544672071933746,
+      "learning_rate": 2.6876767879045785e-06,
+      "loss": 8.5033,
+      "step": 182350
+    },
+    {
+      "epoch": 0.9106844115957952,
+      "grad_norm": 0.09194538742303848,
+      "learning_rate": 2.686174872963028e-06,
+      "loss": 8.5087,
+      "step": 182360
+    },
+    {
+      "epoch": 0.9107343504207346,
+      "grad_norm": 0.08848384767770767,
+      "learning_rate": 2.684672958021477e-06,
+      "loss": 8.5096,
+      "step": 182370
+    },
+    {
+      "epoch": 0.910784289245674,
+      "grad_norm": 0.09186051785945892,
+      "learning_rate": 2.683171043079927e-06,
+      "loss": 8.5115,
+      "step": 182380
+    },
+    {
+      "epoch": 0.9108342280706135,
+      "grad_norm": 0.09041396528482437,
+      "learning_rate": 2.6816691281383767e-06,
+      "loss": 8.5148,
+      "step": 182390
+    },
+    {
+      "epoch": 0.910884166895553,
+      "grad_norm": 0.0917699784040451,
+      "learning_rate": 2.680167213196826e-06,
+      "loss": 8.5289,
+      "step": 182400
+    },
+    {
+      "epoch": 0.9109341057204924,
+      "grad_norm": 0.09223365783691406,
+      "learning_rate": 2.6786652982552754e-06,
+      "loss": 8.509,
+      "step": 182410
+    },
+    {
+      "epoch": 0.9109840445454318,
+      "grad_norm": 0.08908481150865555,
+      "learning_rate": 2.6771633833137247e-06,
+      "loss": 8.519,
+      "step": 182420
+    },
+    {
+      "epoch": 0.9110339833703713,
+      "grad_norm": 0.08989152312278748,
+      "learning_rate": 2.675661468372175e-06,
+      "loss": 8.5255,
+      "step": 182430
+    },
+    {
+      "epoch": 0.9110839221953108,
+      "grad_norm": 0.0932261049747467,
+      "learning_rate": 2.6741595534306242e-06,
+      "loss": 8.5118,
+      "step": 182440
+    },
+    {
+      "epoch": 0.9111338610202502,
+      "grad_norm": 0.08758267015218735,
+      "learning_rate": 2.6726576384890736e-06,
+      "loss": 8.528,
+      "step": 182450
+    },
+    {
+      "epoch": 0.9111837998451896,
+      "grad_norm": 0.09773239493370056,
+      "learning_rate": 2.671155723547523e-06,
+      "loss": 8.5038,
+      "step": 182460
+    },
+    {
+      "epoch": 0.9112337386701291,
+      "grad_norm": 0.09452714025974274,
+      "learning_rate": 2.6696538086059726e-06,
+      "loss": 8.512,
+      "step": 182470
+    },
+    {
+      "epoch": 0.9112836774950686,
+      "grad_norm": 0.09245564788579941,
+      "learning_rate": 2.6681518936644224e-06,
+      "loss": 8.5157,
+      "step": 182480
+    },
+    {
+      "epoch": 0.911333616320008,
+      "grad_norm": 0.08789669722318649,
+      "learning_rate": 2.6666499787228717e-06,
+      "loss": 8.5091,
+      "step": 182490
+    },
+    {
+      "epoch": 0.9113835551449474,
+      "grad_norm": 0.08953165262937546,
+      "learning_rate": 2.665148063781321e-06,
+      "loss": 8.4934,
+      "step": 182500
+    },
+    {
+      "epoch": 0.9114334939698869,
+      "grad_norm": 0.09100998938083649,
+      "learning_rate": 2.663646148839771e-06,
+      "loss": 8.5173,
+      "step": 182510
+    },
+    {
+      "epoch": 0.9114834327948264,
+      "grad_norm": 0.10075462609529495,
+      "learning_rate": 2.66214423389822e-06,
+      "loss": 8.5088,
+      "step": 182520
+    },
+    {
+      "epoch": 0.9115333716197658,
+      "grad_norm": 0.09430301189422607,
+      "learning_rate": 2.66064231895667e-06,
+      "loss": 8.4971,
+      "step": 182530
+    },
+    {
+      "epoch": 0.9115833104447052,
+      "grad_norm": 0.08746136724948883,
+      "learning_rate": 2.6591404040151192e-06,
+      "loss": 8.4993,
+      "step": 182540
+    },
+    {
+      "epoch": 0.9116332492696447,
+      "grad_norm": 0.08813217282295227,
+      "learning_rate": 2.657638489073569e-06,
+      "loss": 8.5164,
+      "step": 182550
+    },
+    {
+      "epoch": 0.9116831880945842,
+      "grad_norm": 0.08637049794197083,
+      "learning_rate": 2.6561365741320183e-06,
+      "loss": 8.508,
+      "step": 182560
+    },
+    {
+      "epoch": 0.9117331269195236,
+      "grad_norm": 0.09196769446134567,
+      "learning_rate": 2.6546346591904676e-06,
+      "loss": 8.498,
+      "step": 182570
+    },
+    {
+      "epoch": 0.911783065744463,
+      "grad_norm": 0.09665251523256302,
+      "learning_rate": 2.6531327442489174e-06,
+      "loss": 8.5179,
+      "step": 182580
+    },
+    {
+      "epoch": 0.9118330045694025,
+      "grad_norm": 0.08948566764593124,
+      "learning_rate": 2.651630829307367e-06,
+      "loss": 8.5084,
+      "step": 182590
+    },
+    {
+      "epoch": 0.911882943394342,
+      "grad_norm": 0.08373957872390747,
+      "learning_rate": 2.6501289143658165e-06,
+      "loss": 8.5356,
+      "step": 182600
+    },
+    {
+      "epoch": 0.9119328822192814,
+      "grad_norm": 0.08785004913806915,
+      "learning_rate": 2.648626999424266e-06,
+      "loss": 8.5137,
+      "step": 182610
+    },
+    {
+      "epoch": 0.9119828210442208,
+      "grad_norm": 0.09454427659511566,
+      "learning_rate": 2.6471250844827156e-06,
+      "loss": 8.503,
+      "step": 182620
+    },
+    {
+      "epoch": 0.9120327598691603,
+      "grad_norm": 0.0859336405992508,
+      "learning_rate": 2.6456231695411653e-06,
+      "loss": 8.5194,
+      "step": 182630
+    },
+    {
+      "epoch": 0.9120826986940997,
+      "grad_norm": 0.0904291421175003,
+      "learning_rate": 2.6441212545996147e-06,
+      "loss": 8.5232,
+      "step": 182640
+    },
+    {
+      "epoch": 0.9121326375190392,
+      "grad_norm": 0.0902220755815506,
+      "learning_rate": 2.642619339658064e-06,
+      "loss": 8.4902,
+      "step": 182650
+    },
+    {
+      "epoch": 0.9121825763439786,
+      "grad_norm": 0.09232915937900543,
+      "learning_rate": 2.6411174247165133e-06,
+      "loss": 8.5124,
+      "step": 182660
+    },
+    {
+      "epoch": 0.9122325151689181,
+      "grad_norm": 0.0919727310538292,
+      "learning_rate": 2.6396155097749635e-06,
+      "loss": 8.5165,
+      "step": 182670
+    },
+    {
+      "epoch": 0.9122824539938575,
+      "grad_norm": 0.09011544287204742,
+      "learning_rate": 2.638113594833413e-06,
+      "loss": 8.534,
+      "step": 182680
+    },
+    {
+      "epoch": 0.912332392818797,
+      "grad_norm": 0.08686014264822006,
+      "learning_rate": 2.636611679891862e-06,
+      "loss": 8.5142,
+      "step": 182690
+    },
+    {
+      "epoch": 0.9123823316437364,
+      "grad_norm": 0.09502139687538147,
+      "learning_rate": 2.6351097649503115e-06,
+      "loss": 8.5061,
+      "step": 182700
+    },
+    {
+      "epoch": 0.9124322704686759,
+      "grad_norm": 0.08637500554323196,
+      "learning_rate": 2.6336078500087613e-06,
+      "loss": 8.5126,
+      "step": 182710
+    },
+    {
+      "epoch": 0.9124822092936153,
+      "grad_norm": 0.08577697724103928,
+      "learning_rate": 2.632105935067211e-06,
+      "loss": 8.5109,
+      "step": 182720
+    },
+    {
+      "epoch": 0.9125321481185548,
+      "grad_norm": 0.08816831558942795,
+      "learning_rate": 2.6306040201256603e-06,
+      "loss": 8.5045,
+      "step": 182730
+    },
+    {
+      "epoch": 0.9125820869434942,
+      "grad_norm": 0.09027297049760818,
+      "learning_rate": 2.6291021051841097e-06,
+      "loss": 8.5062,
+      "step": 182740
+    },
+    {
+      "epoch": 0.9126320257684337,
+      "grad_norm": 0.08729089051485062,
+      "learning_rate": 2.6276001902425594e-06,
+      "loss": 8.4982,
+      "step": 182750
+    },
+    {
+      "epoch": 0.9126819645933731,
+      "grad_norm": 0.09011752158403397,
+      "learning_rate": 2.6260982753010088e-06,
+      "loss": 8.5129,
+      "step": 182760
+    },
+    {
+      "epoch": 0.9127319034183126,
+      "grad_norm": 0.09230612218379974,
+      "learning_rate": 2.6245963603594585e-06,
+      "loss": 8.5134,
+      "step": 182770
+    },
+    {
+      "epoch": 0.912781842243252,
+      "grad_norm": 0.08646968007087708,
+      "learning_rate": 2.623094445417908e-06,
+      "loss": 8.5196,
+      "step": 182780
+    },
+    {
+      "epoch": 0.9128317810681915,
+      "grad_norm": 0.09204564988613129,
+      "learning_rate": 2.6215925304763576e-06,
+      "loss": 8.5168,
+      "step": 182790
+    },
+    {
+      "epoch": 0.9128817198931309,
+      "grad_norm": 0.08996839821338654,
+      "learning_rate": 2.620090615534807e-06,
+      "loss": 8.5249,
+      "step": 182800
+    },
+    {
+      "epoch": 0.9129316587180704,
+      "grad_norm": 0.09080091118812561,
+      "learning_rate": 2.6185887005932563e-06,
+      "loss": 8.5052,
+      "step": 182810
+    },
+    {
+      "epoch": 0.9129815975430098,
+      "grad_norm": 0.09134030342102051,
+      "learning_rate": 2.617086785651706e-06,
+      "loss": 8.5065,
+      "step": 182820
+    },
+    {
+      "epoch": 0.9130315363679493,
+      "grad_norm": 0.08714727312326431,
+      "learning_rate": 2.6155848707101558e-06,
+      "loss": 8.4966,
+      "step": 182830
+    },
+    {
+      "epoch": 0.9130814751928887,
+      "grad_norm": 0.08802255988121033,
+      "learning_rate": 2.614082955768605e-06,
+      "loss": 8.5049,
+      "step": 182840
+    },
+    {
+      "epoch": 0.9131314140178282,
+      "grad_norm": 0.091342493891716,
+      "learning_rate": 2.6125810408270544e-06,
+      "loss": 8.5057,
+      "step": 182850
+    },
+    {
+      "epoch": 0.9131813528427676,
+      "grad_norm": 0.0987749993801117,
+      "learning_rate": 2.6110791258855038e-06,
+      "loss": 8.5085,
+      "step": 182860
+    },
+    {
+      "epoch": 0.9132312916677071,
+      "grad_norm": 0.08858326077461243,
+      "learning_rate": 2.609577210943954e-06,
+      "loss": 8.5,
+      "step": 182870
+    },
+    {
+      "epoch": 0.9132812304926465,
+      "grad_norm": 0.09258518368005753,
+      "learning_rate": 2.6080752960024033e-06,
+      "loss": 8.4916,
+      "step": 182880
+    },
+    {
+      "epoch": 0.913331169317586,
+      "grad_norm": 0.09570000320672989,
+      "learning_rate": 2.6065733810608526e-06,
+      "loss": 8.5024,
+      "step": 182890
+    },
+    {
+      "epoch": 0.9133811081425254,
+      "grad_norm": 0.09027928858995438,
+      "learning_rate": 2.605071466119302e-06,
+      "loss": 8.5172,
+      "step": 182900
+    },
+    {
+      "epoch": 0.9134310469674649,
+      "grad_norm": 0.09241816401481628,
+      "learning_rate": 2.6035695511777517e-06,
+      "loss": 8.5002,
+      "step": 182910
+    },
+    {
+      "epoch": 0.9134809857924043,
+      "grad_norm": 0.08846849203109741,
+      "learning_rate": 2.6020676362362014e-06,
+      "loss": 8.5361,
+      "step": 182920
+    },
+    {
+      "epoch": 0.9135309246173438,
+      "grad_norm": 0.09381444752216339,
+      "learning_rate": 2.6005657212946508e-06,
+      "loss": 8.5218,
+      "step": 182930
+    },
+    {
+      "epoch": 0.9135808634422832,
+      "grad_norm": 0.09875883162021637,
+      "learning_rate": 2.5990638063531e-06,
+      "loss": 8.5085,
+      "step": 182940
+    },
+    {
+      "epoch": 0.9136308022672226,
+      "grad_norm": 0.0989549458026886,
+      "learning_rate": 2.59756189141155e-06,
+      "loss": 8.4929,
+      "step": 182950
+    },
+    {
+      "epoch": 0.9136807410921621,
+      "grad_norm": 0.08980342745780945,
+      "learning_rate": 2.596059976469999e-06,
+      "loss": 8.501,
+      "step": 182960
+    },
+    {
+      "epoch": 0.9137306799171016,
+      "grad_norm": 0.0909147560596466,
+      "learning_rate": 2.594558061528449e-06,
+      "loss": 8.5148,
+      "step": 182970
+    },
+    {
+      "epoch": 0.913780618742041,
+      "grad_norm": 0.091719850897789,
+      "learning_rate": 2.5930561465868983e-06,
+      "loss": 8.5061,
+      "step": 182980
+    },
+    {
+      "epoch": 0.9138305575669804,
+      "grad_norm": 0.09617502242326736,
+      "learning_rate": 2.591554231645348e-06,
+      "loss": 8.5005,
+      "step": 182990
+    },
+    {
+      "epoch": 0.9138804963919199,
+      "grad_norm": 0.08993982523679733,
+      "learning_rate": 2.5900523167037974e-06,
+      "loss": 8.5062,
+      "step": 183000
+    },
+    {
+      "epoch": 0.9139304352168593,
+      "grad_norm": 0.08975507318973541,
+      "learning_rate": 2.5885504017622467e-06,
+      "loss": 8.5052,
+      "step": 183010
+    },
+    {
+      "epoch": 0.9139803740417988,
+      "grad_norm": 0.09407518804073334,
+      "learning_rate": 2.5870484868206965e-06,
+      "loss": 8.5038,
+      "step": 183020
+    },
+    {
+      "epoch": 0.9140303128667382,
+      "grad_norm": 0.09323114156723022,
+      "learning_rate": 2.585546571879146e-06,
+      "loss": 8.5105,
+      "step": 183030
+    },
+    {
+      "epoch": 0.9140802516916777,
+      "grad_norm": 0.08849375694990158,
+      "learning_rate": 2.5840446569375955e-06,
+      "loss": 8.5113,
+      "step": 183040
+    },
+    {
+      "epoch": 0.9141301905166171,
+      "grad_norm": 0.08718482404947281,
+      "learning_rate": 2.582542741996045e-06,
+      "loss": 8.51,
+      "step": 183050
+    },
+    {
+      "epoch": 0.9141801293415566,
+      "grad_norm": 0.09222862124443054,
+      "learning_rate": 2.581040827054494e-06,
+      "loss": 8.503,
+      "step": 183060
+    },
+    {
+      "epoch": 0.914230068166496,
+      "grad_norm": 0.09302924573421478,
+      "learning_rate": 2.5795389121129444e-06,
+      "loss": 8.5162,
+      "step": 183070
+    },
+    {
+      "epoch": 0.9142800069914355,
+      "grad_norm": 0.09386178851127625,
+      "learning_rate": 2.5780369971713937e-06,
+      "loss": 8.5088,
+      "step": 183080
+    },
+    {
+      "epoch": 0.914329945816375,
+      "grad_norm": 0.09588906168937683,
+      "learning_rate": 2.576535082229843e-06,
+      "loss": 8.4878,
+      "step": 183090
+    },
+    {
+      "epoch": 0.9143798846413144,
+      "grad_norm": 0.09002941846847534,
+      "learning_rate": 2.5750331672882924e-06,
+      "loss": 8.4933,
+      "step": 183100
+    },
+    {
+      "epoch": 0.9144298234662538,
+      "grad_norm": 0.08975888043642044,
+      "learning_rate": 2.573531252346742e-06,
+      "loss": 8.4971,
+      "step": 183110
+    },
+    {
+      "epoch": 0.9144797622911933,
+      "grad_norm": 0.09490811824798584,
+      "learning_rate": 2.572029337405192e-06,
+      "loss": 8.522,
+      "step": 183120
+    },
+    {
+      "epoch": 0.9145297011161327,
+      "grad_norm": 0.09320612251758575,
+      "learning_rate": 2.5705274224636412e-06,
+      "loss": 8.5048,
+      "step": 183130
+    },
+    {
+      "epoch": 0.9145796399410722,
+      "grad_norm": 0.08958461880683899,
+      "learning_rate": 2.5690255075220905e-06,
+      "loss": 8.511,
+      "step": 183140
+    },
+    {
+      "epoch": 0.9146295787660116,
+      "grad_norm": 0.09119990468025208,
+      "learning_rate": 2.5675235925805403e-06,
+      "loss": 8.4993,
+      "step": 183150
+    },
+    {
+      "epoch": 0.9146795175909511,
+      "grad_norm": 0.093448705971241,
+      "learning_rate": 2.56602167763899e-06,
+      "loss": 8.4941,
+      "step": 183160
+    },
+    {
+      "epoch": 0.9147294564158905,
+      "grad_norm": 0.0903758853673935,
+      "learning_rate": 2.5645197626974394e-06,
+      "loss": 8.5067,
+      "step": 183170
+    },
+    {
+      "epoch": 0.91477939524083,
+      "grad_norm": 0.09165302664041519,
+      "learning_rate": 2.5630178477558887e-06,
+      "loss": 8.5058,
+      "step": 183180
+    },
+    {
+      "epoch": 0.9148293340657694,
+      "grad_norm": 0.08942514657974243,
+      "learning_rate": 2.5615159328143385e-06,
+      "loss": 8.5319,
+      "step": 183190
+    },
+    {
+      "epoch": 0.9148792728907089,
+      "grad_norm": 0.08656205981969833,
+      "learning_rate": 2.560014017872788e-06,
+      "loss": 8.5003,
+      "step": 183200
+    },
+    {
+      "epoch": 0.9149292117156483,
+      "grad_norm": 0.08952057361602783,
+      "learning_rate": 2.5585121029312376e-06,
+      "loss": 8.5137,
+      "step": 183210
+    },
+    {
+      "epoch": 0.9149791505405878,
+      "grad_norm": 0.09631547331809998,
+      "learning_rate": 2.557010187989687e-06,
+      "loss": 8.5028,
+      "step": 183220
+    },
+    {
+      "epoch": 0.9150290893655272,
+      "grad_norm": 0.0935741513967514,
+      "learning_rate": 2.5555082730481366e-06,
+      "loss": 8.5188,
+      "step": 183230
+    },
+    {
+      "epoch": 0.9150790281904667,
+      "grad_norm": 0.09073397517204285,
+      "learning_rate": 2.554006358106586e-06,
+      "loss": 8.5276,
+      "step": 183240
+    },
+    {
+      "epoch": 0.9151289670154061,
+      "grad_norm": 0.09225539863109589,
+      "learning_rate": 2.5525044431650353e-06,
+      "loss": 8.5081,
+      "step": 183250
+    },
+    {
+      "epoch": 0.9151789058403456,
+      "grad_norm": 0.09920675307512283,
+      "learning_rate": 2.551002528223485e-06,
+      "loss": 8.5123,
+      "step": 183260
+    },
+    {
+      "epoch": 0.915228844665285,
+      "grad_norm": 0.09345440566539764,
+      "learning_rate": 2.549500613281935e-06,
+      "loss": 8.505,
+      "step": 183270
+    },
+    {
+      "epoch": 0.9152787834902245,
+      "grad_norm": 0.09154894202947617,
+      "learning_rate": 2.547998698340384e-06,
+      "loss": 8.4907,
+      "step": 183280
+    },
+    {
+      "epoch": 0.9153287223151639,
+      "grad_norm": 0.09122487902641296,
+      "learning_rate": 2.5464967833988335e-06,
+      "loss": 8.5205,
+      "step": 183290
+    },
+    {
+      "epoch": 0.9153786611401034,
+      "grad_norm": 0.09112253040075302,
+      "learning_rate": 2.544994868457283e-06,
+      "loss": 8.4846,
+      "step": 183300
+    },
+    {
+      "epoch": 0.9154285999650428,
+      "grad_norm": 0.08982587605714798,
+      "learning_rate": 2.543492953515733e-06,
+      "loss": 8.5097,
+      "step": 183310
+    },
+    {
+      "epoch": 0.9154785387899823,
+      "grad_norm": 0.09354843199253082,
+      "learning_rate": 2.5419910385741823e-06,
+      "loss": 8.5127,
+      "step": 183320
+    },
+    {
+      "epoch": 0.9155284776149217,
+      "grad_norm": 0.08977293968200684,
+      "learning_rate": 2.5404891236326317e-06,
+      "loss": 8.5254,
+      "step": 183330
+    },
+    {
+      "epoch": 0.9155784164398612,
+      "grad_norm": 0.0911603569984436,
+      "learning_rate": 2.538987208691081e-06,
+      "loss": 8.5047,
+      "step": 183340
+    },
+    {
+      "epoch": 0.9156283552648006,
+      "grad_norm": 0.09315786510705948,
+      "learning_rate": 2.5374852937495307e-06,
+      "loss": 8.4967,
+      "step": 183350
+    },
+    {
+      "epoch": 0.9156782940897401,
+      "grad_norm": 0.09075506776571274,
+      "learning_rate": 2.5359833788079805e-06,
+      "loss": 8.4941,
+      "step": 183360
+    },
+    {
+      "epoch": 0.9157282329146795,
+      "grad_norm": 0.0923544391989708,
+      "learning_rate": 2.53448146386643e-06,
+      "loss": 8.5156,
+      "step": 183370
+    },
+    {
+      "epoch": 0.915778171739619,
+      "grad_norm": 0.09413059055805206,
+      "learning_rate": 2.532979548924879e-06,
+      "loss": 8.4979,
+      "step": 183380
+    },
+    {
+      "epoch": 0.9158281105645584,
+      "grad_norm": 0.09276200085878372,
+      "learning_rate": 2.531477633983329e-06,
+      "loss": 8.5112,
+      "step": 183390
+    },
+    {
+      "epoch": 0.9158780493894979,
+      "grad_norm": 0.09218557178974152,
+      "learning_rate": 2.5299757190417782e-06,
+      "loss": 8.5049,
+      "step": 183400
+    },
+    {
+      "epoch": 0.9159279882144373,
+      "grad_norm": 0.09127632528543472,
+      "learning_rate": 2.528473804100228e-06,
+      "loss": 8.4994,
+      "step": 183410
+    },
+    {
+      "epoch": 0.9159779270393767,
+      "grad_norm": 0.09124505519866943,
+      "learning_rate": 2.5269718891586773e-06,
+      "loss": 8.5037,
+      "step": 183420
+    },
+    {
+      "epoch": 0.9160278658643162,
+      "grad_norm": 0.09244653582572937,
+      "learning_rate": 2.525469974217127e-06,
+      "loss": 8.5091,
+      "step": 183430
+    },
+    {
+      "epoch": 0.9160778046892557,
+      "grad_norm": 0.09359874576330185,
+      "learning_rate": 2.5239680592755764e-06,
+      "loss": 8.5033,
+      "step": 183440
+    },
+    {
+      "epoch": 0.9161277435141951,
+      "grad_norm": 0.08860906213521957,
+      "learning_rate": 2.5224661443340257e-06,
+      "loss": 8.5175,
+      "step": 183450
+    },
+    {
+      "epoch": 0.9161776823391345,
+      "grad_norm": 0.08711325377225876,
+      "learning_rate": 2.5209642293924755e-06,
+      "loss": 8.4991,
+      "step": 183460
+    },
+    {
+      "epoch": 0.916227621164074,
+      "grad_norm": 0.0916765034198761,
+      "learning_rate": 2.5194623144509253e-06,
+      "loss": 8.5133,
+      "step": 183470
+    },
+    {
+      "epoch": 0.9162775599890135,
+      "grad_norm": 0.0957622155547142,
+      "learning_rate": 2.5179603995093746e-06,
+      "loss": 8.5013,
+      "step": 183480
+    },
+    {
+      "epoch": 0.9163274988139529,
+      "grad_norm": 0.08793076127767563,
+      "learning_rate": 2.516458484567824e-06,
+      "loss": 8.5154,
+      "step": 183490
+    },
+    {
+      "epoch": 0.9163774376388923,
+      "grad_norm": 0.08785158395767212,
+      "learning_rate": 2.5149565696262732e-06,
+      "loss": 8.5007,
+      "step": 183500
+    },
+    {
+      "epoch": 0.9164273764638318,
+      "grad_norm": 0.09278276562690735,
+      "learning_rate": 2.5134546546847234e-06,
+      "loss": 8.5113,
+      "step": 183510
+    },
+    {
+      "epoch": 0.9164773152887713,
+      "grad_norm": 0.08464596420526505,
+      "learning_rate": 2.5119527397431728e-06,
+      "loss": 8.5296,
+      "step": 183520
+    },
+    {
+      "epoch": 0.9165272541137107,
+      "grad_norm": 0.08865651488304138,
+      "learning_rate": 2.510450824801622e-06,
+      "loss": 8.5137,
+      "step": 183530
+    },
+    {
+      "epoch": 0.9165771929386501,
+      "grad_norm": 0.09250415861606598,
+      "learning_rate": 2.5089489098600714e-06,
+      "loss": 8.5173,
+      "step": 183540
+    },
+    {
+      "epoch": 0.9166271317635896,
+      "grad_norm": 0.09271103143692017,
+      "learning_rate": 2.507446994918521e-06,
+      "loss": 8.4864,
+      "step": 183550
+    },
+    {
+      "epoch": 0.9166770705885291,
+      "grad_norm": 0.09764368087053299,
+      "learning_rate": 2.505945079976971e-06,
+      "loss": 8.5003,
+      "step": 183560
+    },
+    {
+      "epoch": 0.9167270094134685,
+      "grad_norm": 0.09379107505083084,
+      "learning_rate": 2.5044431650354203e-06,
+      "loss": 8.5047,
+      "step": 183570
+    },
+    {
+      "epoch": 0.9167769482384079,
+      "grad_norm": 0.08978167921304703,
+      "learning_rate": 2.5029412500938696e-06,
+      "loss": 8.5087,
+      "step": 183580
+    },
+    {
+      "epoch": 0.9168268870633474,
+      "grad_norm": 0.0897134467959404,
+      "learning_rate": 2.5014393351523193e-06,
+      "loss": 8.5065,
+      "step": 183590
+    },
+    {
+      "epoch": 0.9168768258882869,
+      "grad_norm": 0.09094201773405075,
+      "learning_rate": 2.4999374202107687e-06,
+      "loss": 8.5085,
+      "step": 183600
+    },
+    {
+      "epoch": 0.9169267647132263,
+      "grad_norm": 0.09478742629289627,
+      "learning_rate": 2.4984355052692184e-06,
+      "loss": 8.5052,
+      "step": 183610
+    },
+    {
+      "epoch": 0.9169767035381657,
+      "grad_norm": 0.0915999785065651,
+      "learning_rate": 2.4969335903276678e-06,
+      "loss": 8.5097,
+      "step": 183620
+    },
+    {
+      "epoch": 0.9170266423631052,
+      "grad_norm": 0.08944950252771378,
+      "learning_rate": 2.4954316753861175e-06,
+      "loss": 8.4869,
+      "step": 183630
+    },
+    {
+      "epoch": 0.9170765811880447,
+      "grad_norm": 0.09154719859361649,
+      "learning_rate": 2.493929760444567e-06,
+      "loss": 8.5146,
+      "step": 183640
+    },
+    {
+      "epoch": 0.9171265200129841,
+      "grad_norm": 0.0902072861790657,
+      "learning_rate": 2.492427845503016e-06,
+      "loss": 8.5102,
+      "step": 183650
+    },
+    {
+      "epoch": 0.9171764588379235,
+      "grad_norm": 0.09381358325481415,
+      "learning_rate": 2.490925930561466e-06,
+      "loss": 8.5281,
+      "step": 183660
+    },
+    {
+      "epoch": 0.917226397662863,
+      "grad_norm": 0.0945509746670723,
+      "learning_rate": 2.4894240156199157e-06,
+      "loss": 8.5014,
+      "step": 183670
+    },
+    {
+      "epoch": 0.9172763364878025,
+      "grad_norm": 0.09148582816123962,
+      "learning_rate": 2.487922100678365e-06,
+      "loss": 8.51,
+      "step": 183680
+    },
+    {
+      "epoch": 0.9173262753127419,
+      "grad_norm": 0.09244761615991592,
+      "learning_rate": 2.4864201857368144e-06,
+      "loss": 8.5024,
+      "step": 183690
+    },
+    {
+      "epoch": 0.9173762141376813,
+      "grad_norm": 0.08857100456953049,
+      "learning_rate": 2.484918270795264e-06,
+      "loss": 8.5167,
+      "step": 183700
+    },
+    {
+      "epoch": 0.9174261529626208,
+      "grad_norm": 0.09274089336395264,
+      "learning_rate": 2.483416355853714e-06,
+      "loss": 8.5078,
+      "step": 183710
+    },
+    {
+      "epoch": 0.9174760917875603,
+      "grad_norm": 0.08995862305164337,
+      "learning_rate": 2.481914440912163e-06,
+      "loss": 8.5171,
+      "step": 183720
+    },
+    {
+      "epoch": 0.9175260306124997,
+      "grad_norm": 0.08714859187602997,
+      "learning_rate": 2.4804125259706125e-06,
+      "loss": 8.4972,
+      "step": 183730
+    },
+    {
+      "epoch": 0.9175759694374391,
+      "grad_norm": 0.09085182845592499,
+      "learning_rate": 2.478910611029062e-06,
+      "loss": 8.5136,
+      "step": 183740
+    },
+    {
+      "epoch": 0.9176259082623786,
+      "grad_norm": 0.09306159615516663,
+      "learning_rate": 2.477408696087512e-06,
+      "loss": 8.5094,
+      "step": 183750
+    },
+    {
+      "epoch": 0.9176758470873181,
+      "grad_norm": 0.08627147227525711,
+      "learning_rate": 2.4759067811459614e-06,
+      "loss": 8.5231,
+      "step": 183760
+    },
+    {
+      "epoch": 0.9177257859122575,
+      "grad_norm": 0.09129956364631653,
+      "learning_rate": 2.4744048662044107e-06,
+      "loss": 8.4941,
+      "step": 183770
+    },
+    {
+      "epoch": 0.9177757247371969,
+      "grad_norm": 0.09526669234037399,
+      "learning_rate": 2.47290295126286e-06,
+      "loss": 8.4966,
+      "step": 183780
+    },
+    {
+      "epoch": 0.9178256635621364,
+      "grad_norm": 0.09122246503829956,
+      "learning_rate": 2.4714010363213098e-06,
+      "loss": 8.4972,
+      "step": 183790
+    },
+    {
+      "epoch": 0.9178756023870759,
+      "grad_norm": 0.09195857495069504,
+      "learning_rate": 2.4698991213797595e-06,
+      "loss": 8.5057,
+      "step": 183800
+    },
+    {
+      "epoch": 0.9179255412120153,
+      "grad_norm": 0.090829037129879,
+      "learning_rate": 2.468397206438209e-06,
+      "loss": 8.5169,
+      "step": 183810
+    },
+    {
+      "epoch": 0.9179754800369547,
+      "grad_norm": 0.09058675169944763,
+      "learning_rate": 2.466895291496658e-06,
+      "loss": 8.511,
+      "step": 183820
+    },
+    {
+      "epoch": 0.9180254188618941,
+      "grad_norm": 0.08892668038606644,
+      "learning_rate": 2.465393376555108e-06,
+      "loss": 8.5127,
+      "step": 183830
+    },
+    {
+      "epoch": 0.9180753576868337,
+      "grad_norm": 0.09829755127429962,
+      "learning_rate": 2.4638914616135573e-06,
+      "loss": 8.4872,
+      "step": 183840
+    },
+    {
+      "epoch": 0.9181252965117731,
+      "grad_norm": 0.08881474286317825,
+      "learning_rate": 2.462389546672007e-06,
+      "loss": 8.4938,
+      "step": 183850
+    },
+    {
+      "epoch": 0.9181752353367125,
+      "grad_norm": 0.08773206174373627,
+      "learning_rate": 2.4608876317304564e-06,
+      "loss": 8.5175,
+      "step": 183860
+    },
+    {
+      "epoch": 0.918225174161652,
+      "grad_norm": 0.08635301142930984,
+      "learning_rate": 2.459385716788906e-06,
+      "loss": 8.5228,
+      "step": 183870
+    },
+    {
+      "epoch": 0.9182751129865915,
+      "grad_norm": 0.08536515384912491,
+      "learning_rate": 2.4578838018473555e-06,
+      "loss": 8.5215,
+      "step": 183880
+    },
+    {
+      "epoch": 0.9183250518115309,
+      "grad_norm": 0.08720611780881882,
+      "learning_rate": 2.456381886905805e-06,
+      "loss": 8.5034,
+      "step": 183890
+    },
+    {
+      "epoch": 0.9183749906364703,
+      "grad_norm": 0.09371429681777954,
+      "learning_rate": 2.4548799719642545e-06,
+      "loss": 8.5029,
+      "step": 183900
+    },
+    {
+      "epoch": 0.9184249294614097,
+      "grad_norm": 0.08567806333303452,
+      "learning_rate": 2.4533780570227043e-06,
+      "loss": 8.5001,
+      "step": 183910
+    },
+    {
+      "epoch": 0.9184748682863493,
+      "grad_norm": 0.09422826766967773,
+      "learning_rate": 2.4518761420811536e-06,
+      "loss": 8.5079,
+      "step": 183920
+    },
+    {
+      "epoch": 0.9185248071112887,
+      "grad_norm": 0.0892164334654808,
+      "learning_rate": 2.450374227139603e-06,
+      "loss": 8.4915,
+      "step": 183930
+    },
+    {
+      "epoch": 0.9185747459362281,
+      "grad_norm": 0.09322153776884079,
+      "learning_rate": 2.4488723121980523e-06,
+      "loss": 8.5038,
+      "step": 183940
+    },
+    {
+      "epoch": 0.9186246847611675,
+      "grad_norm": 0.0924592837691307,
+      "learning_rate": 2.4473703972565025e-06,
+      "loss": 8.5004,
+      "step": 183950
+    },
+    {
+      "epoch": 0.918674623586107,
+      "grad_norm": 0.0903005376458168,
+      "learning_rate": 2.445868482314952e-06,
+      "loss": 8.5133,
+      "step": 183960
+    },
+    {
+      "epoch": 0.9187245624110465,
+      "grad_norm": 0.09269312024116516,
+      "learning_rate": 2.444366567373401e-06,
+      "loss": 8.5266,
+      "step": 183970
+    },
+    {
+      "epoch": 0.9187745012359859,
+      "grad_norm": 0.0908818319439888,
+      "learning_rate": 2.4428646524318505e-06,
+      "loss": 8.5057,
+      "step": 183980
+    },
+    {
+      "epoch": 0.9188244400609253,
+      "grad_norm": 0.09337764978408813,
+      "learning_rate": 2.4413627374903002e-06,
+      "loss": 8.5112,
+      "step": 183990
+    },
+    {
+      "epoch": 0.9188743788858648,
+      "grad_norm": 0.09528010338544846,
+      "learning_rate": 2.43986082254875e-06,
+      "loss": 8.503,
+      "step": 184000
+    },
+    {
+      "epoch": 0.9189243177108043,
+      "grad_norm": 0.09190403670072556,
+      "learning_rate": 2.4383589076071993e-06,
+      "loss": 8.5116,
+      "step": 184010
+    },
+    {
+      "epoch": 0.9189742565357437,
+      "grad_norm": 0.09150946140289307,
+      "learning_rate": 2.4368569926656486e-06,
+      "loss": 8.5047,
+      "step": 184020
+    },
+    {
+      "epoch": 0.9190241953606831,
+      "grad_norm": 0.09026480466127396,
+      "learning_rate": 2.4353550777240984e-06,
+      "loss": 8.5045,
+      "step": 184030
+    },
+    {
+      "epoch": 0.9190741341856226,
+      "grad_norm": 0.08819565176963806,
+      "learning_rate": 2.4338531627825477e-06,
+      "loss": 8.5203,
+      "step": 184040
+    },
+    {
+      "epoch": 0.9191240730105621,
+      "grad_norm": 0.09232760965824127,
+      "learning_rate": 2.4323512478409975e-06,
+      "loss": 8.5237,
+      "step": 184050
+    },
+    {
+      "epoch": 0.9191740118355015,
+      "grad_norm": 0.08861426264047623,
+      "learning_rate": 2.430849332899447e-06,
+      "loss": 8.4923,
+      "step": 184060
+    },
+    {
+      "epoch": 0.9192239506604409,
+      "grad_norm": 0.09452666342258453,
+      "learning_rate": 2.4293474179578966e-06,
+      "loss": 8.513,
+      "step": 184070
+    },
+    {
+      "epoch": 0.9192738894853804,
+      "grad_norm": 0.09179414063692093,
+      "learning_rate": 2.427845503016346e-06,
+      "loss": 8.4931,
+      "step": 184080
+    },
+    {
+      "epoch": 0.9193238283103199,
+      "grad_norm": 0.09043210744857788,
+      "learning_rate": 2.4263435880747952e-06,
+      "loss": 8.5092,
+      "step": 184090
+    },
+    {
+      "epoch": 0.9193737671352593,
+      "grad_norm": 0.09003657847642899,
+      "learning_rate": 2.424841673133245e-06,
+      "loss": 8.517,
+      "step": 184100
+    },
+    {
+      "epoch": 0.9194237059601987,
+      "grad_norm": 0.09282027930021286,
+      "learning_rate": 2.4233397581916947e-06,
+      "loss": 8.5116,
+      "step": 184110
+    },
+    {
+      "epoch": 0.9194736447851382,
+      "grad_norm": 0.08989115059375763,
+      "learning_rate": 2.421837843250144e-06,
+      "loss": 8.4954,
+      "step": 184120
+    },
+    {
+      "epoch": 0.9195235836100777,
+      "grad_norm": 0.09200627356767654,
+      "learning_rate": 2.4203359283085934e-06,
+      "loss": 8.5153,
+      "step": 184130
+    },
+    {
+      "epoch": 0.9195735224350171,
+      "grad_norm": 0.09356432408094406,
+      "learning_rate": 2.4188340133670427e-06,
+      "loss": 8.5144,
+      "step": 184140
+    },
+    {
+      "epoch": 0.9196234612599565,
+      "grad_norm": 0.08702446520328522,
+      "learning_rate": 2.417332098425493e-06,
+      "loss": 8.4921,
+      "step": 184150
+    },
+    {
+      "epoch": 0.919673400084896,
+      "grad_norm": 0.09164425730705261,
+      "learning_rate": 2.4158301834839422e-06,
+      "loss": 8.5054,
+      "step": 184160
+    },
+    {
+      "epoch": 0.9197233389098355,
+      "grad_norm": 0.09433116018772125,
+      "learning_rate": 2.4143282685423916e-06,
+      "loss": 8.5058,
+      "step": 184170
+    },
+    {
+      "epoch": 0.9197732777347749,
+      "grad_norm": 0.09576484560966492,
+      "learning_rate": 2.412826353600841e-06,
+      "loss": 8.5018,
+      "step": 184180
+    },
+    {
+      "epoch": 0.9198232165597143,
+      "grad_norm": 0.09483747184276581,
+      "learning_rate": 2.4113244386592907e-06,
+      "loss": 8.5188,
+      "step": 184190
+    },
+    {
+      "epoch": 0.9198731553846538,
+      "grad_norm": 0.09752367436885834,
+      "learning_rate": 2.4098225237177404e-06,
+      "loss": 8.5201,
+      "step": 184200
+    },
+    {
+      "epoch": 0.9199230942095933,
+      "grad_norm": 0.0914745032787323,
+      "learning_rate": 2.4083206087761897e-06,
+      "loss": 8.5139,
+      "step": 184210
+    },
+    {
+      "epoch": 0.9199730330345327,
+      "grad_norm": 0.08465293794870377,
+      "learning_rate": 2.406818693834639e-06,
+      "loss": 8.5177,
+      "step": 184220
+    },
+    {
+      "epoch": 0.9200229718594721,
+      "grad_norm": 0.09186521172523499,
+      "learning_rate": 2.4053167788930884e-06,
+      "loss": 8.512,
+      "step": 184230
+    },
+    {
+      "epoch": 0.9200729106844115,
+      "grad_norm": 0.09527092427015305,
+      "learning_rate": 2.403814863951538e-06,
+      "loss": 8.5174,
+      "step": 184240
+    },
+    {
+      "epoch": 0.9201228495093511,
+      "grad_norm": 0.08546338975429535,
+      "learning_rate": 2.402312949009988e-06,
+      "loss": 8.5272,
+      "step": 184250
+    },
+    {
+      "epoch": 0.9201727883342905,
+      "grad_norm": 0.09446820616722107,
+      "learning_rate": 2.4008110340684373e-06,
+      "loss": 8.5021,
+      "step": 184260
+    },
+    {
+      "epoch": 0.9202227271592299,
+      "grad_norm": 0.09014095366001129,
+      "learning_rate": 2.3993091191268866e-06,
+      "loss": 8.5166,
+      "step": 184270
+    },
+    {
+      "epoch": 0.9202726659841693,
+      "grad_norm": 0.09346702694892883,
+      "learning_rate": 2.3978072041853363e-06,
+      "loss": 8.5163,
+      "step": 184280
+    },
+    {
+      "epoch": 0.9203226048091089,
+      "grad_norm": 0.09391672164201736,
+      "learning_rate": 2.396305289243786e-06,
+      "loss": 8.5082,
+      "step": 184290
+    },
+    {
+      "epoch": 0.9203725436340483,
+      "grad_norm": 0.09350437670946121,
+      "learning_rate": 2.3948033743022354e-06,
+      "loss": 8.5018,
+      "step": 184300
+    },
+    {
+      "epoch": 0.9204224824589877,
+      "grad_norm": 0.09567178785800934,
+      "learning_rate": 2.3933014593606848e-06,
+      "loss": 8.5052,
+      "step": 184310
+    },
+    {
+      "epoch": 0.9204724212839271,
+      "grad_norm": 0.09135613590478897,
+      "learning_rate": 2.3917995444191345e-06,
+      "loss": 8.5153,
+      "step": 184320
+    },
+    {
+      "epoch": 0.9205223601088667,
+      "grad_norm": 0.08975464105606079,
+      "learning_rate": 2.390297629477584e-06,
+      "loss": 8.5067,
+      "step": 184330
+    },
+    {
+      "epoch": 0.9205722989338061,
+      "grad_norm": 0.0906573086977005,
+      "learning_rate": 2.3887957145360336e-06,
+      "loss": 8.5185,
+      "step": 184340
+    },
+    {
+      "epoch": 0.9206222377587455,
+      "grad_norm": 0.09258557856082916,
+      "learning_rate": 2.387293799594483e-06,
+      "loss": 8.5008,
+      "step": 184350
+    },
+    {
+      "epoch": 0.9206721765836849,
+      "grad_norm": 0.09056249260902405,
+      "learning_rate": 2.3857918846529327e-06,
+      "loss": 8.5061,
+      "step": 184360
+    },
+    {
+      "epoch": 0.9207221154086245,
+      "grad_norm": 0.09652728587388992,
+      "learning_rate": 2.384289969711382e-06,
+      "loss": 8.4943,
+      "step": 184370
+    },
+    {
+      "epoch": 0.9207720542335639,
+      "grad_norm": 0.08950930088758469,
+      "learning_rate": 2.3827880547698313e-06,
+      "loss": 8.5062,
+      "step": 184380
+    },
+    {
+      "epoch": 0.9208219930585033,
+      "grad_norm": 0.09195245802402496,
+      "learning_rate": 2.381286139828281e-06,
+      "loss": 8.4987,
+      "step": 184390
+    },
+    {
+      "epoch": 0.9208719318834427,
+      "grad_norm": 0.08976045995950699,
+      "learning_rate": 2.379784224886731e-06,
+      "loss": 8.5078,
+      "step": 184400
+    },
+    {
+      "epoch": 0.9209218707083823,
+      "grad_norm": 0.08620437234640121,
+      "learning_rate": 2.37828230994518e-06,
+      "loss": 8.5356,
+      "step": 184410
+    },
+    {
+      "epoch": 0.9209718095333217,
+      "grad_norm": 0.08748499304056168,
+      "learning_rate": 2.3767803950036295e-06,
+      "loss": 8.5128,
+      "step": 184420
+    },
+    {
+      "epoch": 0.9210217483582611,
+      "grad_norm": 0.08978407084941864,
+      "learning_rate": 2.375278480062079e-06,
+      "loss": 8.508,
+      "step": 184430
+    },
+    {
+      "epoch": 0.9210716871832005,
+      "grad_norm": 0.09201885014772415,
+      "learning_rate": 2.373776565120529e-06,
+      "loss": 8.4949,
+      "step": 184440
+    },
+    {
+      "epoch": 0.9211216260081401,
+      "grad_norm": 0.09136571735143661,
+      "learning_rate": 2.3722746501789784e-06,
+      "loss": 8.5061,
+      "step": 184450
+    },
+    {
+      "epoch": 0.9211715648330795,
+      "grad_norm": 0.09248650074005127,
+      "learning_rate": 2.3707727352374277e-06,
+      "loss": 8.4964,
+      "step": 184460
+    },
+    {
+      "epoch": 0.9212215036580189,
+      "grad_norm": 0.08990224450826645,
+      "learning_rate": 2.369270820295877e-06,
+      "loss": 8.51,
+      "step": 184470
+    },
+    {
+      "epoch": 0.9212714424829583,
+      "grad_norm": 0.09131959080696106,
+      "learning_rate": 2.3677689053543268e-06,
+      "loss": 8.5172,
+      "step": 184480
+    },
+    {
+      "epoch": 0.9213213813078979,
+      "grad_norm": 0.09705401957035065,
+      "learning_rate": 2.3662669904127765e-06,
+      "loss": 8.5212,
+      "step": 184490
+    },
+    {
+      "epoch": 0.9213713201328373,
+      "grad_norm": 0.09412173926830292,
+      "learning_rate": 2.364765075471226e-06,
+      "loss": 8.4953,
+      "step": 184500
+    },
+    {
+      "epoch": 0.9214212589577767,
+      "grad_norm": 0.08390575647354126,
+      "learning_rate": 2.363263160529675e-06,
+      "loss": 8.5015,
+      "step": 184510
+    },
+    {
+      "epoch": 0.9214711977827161,
+      "grad_norm": 0.09075505286455154,
+      "learning_rate": 2.361761245588125e-06,
+      "loss": 8.488,
+      "step": 184520
+    },
+    {
+      "epoch": 0.9215211366076557,
+      "grad_norm": 0.08866789937019348,
+      "learning_rate": 2.3602593306465743e-06,
+      "loss": 8.5028,
+      "step": 184530
+    },
+    {
+      "epoch": 0.9215710754325951,
+      "grad_norm": 0.09287255257368088,
+      "learning_rate": 2.358757415705024e-06,
+      "loss": 8.5005,
+      "step": 184540
+    },
+    {
+      "epoch": 0.9216210142575345,
+      "grad_norm": 0.08919494599103928,
+      "learning_rate": 2.3572555007634734e-06,
+      "loss": 8.5096,
+      "step": 184550
+    },
+    {
+      "epoch": 0.9216709530824739,
+      "grad_norm": 0.09100577980279922,
+      "learning_rate": 2.355753585821923e-06,
+      "loss": 8.4997,
+      "step": 184560
+    },
+    {
+      "epoch": 0.9217208919074135,
+      "grad_norm": 0.09456155449151993,
+      "learning_rate": 2.3542516708803725e-06,
+      "loss": 8.4924,
+      "step": 184570
+    },
+    {
+      "epoch": 0.9217708307323529,
+      "grad_norm": 0.09156771004199982,
+      "learning_rate": 2.3527497559388218e-06,
+      "loss": 8.5267,
+      "step": 184580
+    },
+    {
+      "epoch": 0.9218207695572923,
+      "grad_norm": 0.08734113723039627,
+      "learning_rate": 2.3512478409972715e-06,
+      "loss": 8.4983,
+      "step": 184590
+    },
+    {
+      "epoch": 0.9218707083822317,
+      "grad_norm": 0.09091001003980637,
+      "learning_rate": 2.3497459260557213e-06,
+      "loss": 8.518,
+      "step": 184600
+    },
+    {
+      "epoch": 0.9219206472071713,
+      "grad_norm": 0.08664854615926743,
+      "learning_rate": 2.3482440111141706e-06,
+      "loss": 8.5167,
+      "step": 184610
+    },
+    {
+      "epoch": 0.9219705860321107,
+      "grad_norm": 0.08995333313941956,
+      "learning_rate": 2.34674209617262e-06,
+      "loss": 8.4968,
+      "step": 184620
+    },
+    {
+      "epoch": 0.9220205248570501,
+      "grad_norm": 0.09020540118217468,
+      "learning_rate": 2.3452401812310693e-06,
+      "loss": 8.4935,
+      "step": 184630
+    },
+    {
+      "epoch": 0.9220704636819895,
+      "grad_norm": 0.08525863289833069,
+      "learning_rate": 2.3437382662895195e-06,
+      "loss": 8.512,
+      "step": 184640
+    },
+    {
+      "epoch": 0.9221204025069291,
+      "grad_norm": 0.09016373753547668,
+      "learning_rate": 2.342236351347969e-06,
+      "loss": 8.4992,
+      "step": 184650
+    },
+    {
+      "epoch": 0.9221703413318685,
+      "grad_norm": 0.09463771432638168,
+      "learning_rate": 2.340734436406418e-06,
+      "loss": 8.4842,
+      "step": 184660
+    },
+    {
+      "epoch": 0.9222202801568079,
+      "grad_norm": 0.08823353797197342,
+      "learning_rate": 2.3392325214648675e-06,
+      "loss": 8.4996,
+      "step": 184670
+    },
+    {
+      "epoch": 0.9222702189817473,
+      "grad_norm": 0.09164861589670181,
+      "learning_rate": 2.3377306065233172e-06,
+      "loss": 8.5169,
+      "step": 184680
+    },
+    {
+      "epoch": 0.9223201578066869,
+      "grad_norm": 0.092039093375206,
+      "learning_rate": 2.336228691581767e-06,
+      "loss": 8.5056,
+      "step": 184690
+    },
+    {
+      "epoch": 0.9223700966316263,
+      "grad_norm": 0.09081801027059555,
+      "learning_rate": 2.3347267766402163e-06,
+      "loss": 8.4922,
+      "step": 184700
+    },
+    {
+      "epoch": 0.9224200354565657,
+      "grad_norm": 0.09290160983800888,
+      "learning_rate": 2.3332248616986656e-06,
+      "loss": 8.5182,
+      "step": 184710
+    },
+    {
+      "epoch": 0.9224699742815051,
+      "grad_norm": 0.08207748085260391,
+      "learning_rate": 2.3317229467571154e-06,
+      "loss": 8.5093,
+      "step": 184720
+    },
+    {
+      "epoch": 0.9225199131064447,
+      "grad_norm": 0.09553519636392593,
+      "learning_rate": 2.3302210318155647e-06,
+      "loss": 8.5143,
+      "step": 184730
+    },
+    {
+      "epoch": 0.9225698519313841,
+      "grad_norm": 0.09908508509397507,
+      "learning_rate": 2.3287191168740145e-06,
+      "loss": 8.5059,
+      "step": 184740
+    },
+    {
+      "epoch": 0.9226197907563235,
+      "grad_norm": 0.0906306803226471,
+      "learning_rate": 2.327217201932464e-06,
+      "loss": 8.5063,
+      "step": 184750
+    },
+    {
+      "epoch": 0.9226697295812629,
+      "grad_norm": 0.09021761268377304,
+      "learning_rate": 2.3257152869909136e-06,
+      "loss": 8.51,
+      "step": 184760
+    },
+    {
+      "epoch": 0.9227196684062025,
+      "grad_norm": 0.09584904462099075,
+      "learning_rate": 2.324213372049363e-06,
+      "loss": 8.4976,
+      "step": 184770
+    },
+    {
+      "epoch": 0.9227696072311419,
+      "grad_norm": 0.08946247398853302,
+      "learning_rate": 2.3227114571078126e-06,
+      "loss": 8.5217,
+      "step": 184780
+    },
+    {
+      "epoch": 0.9228195460560813,
+      "grad_norm": 0.09100887924432755,
+      "learning_rate": 2.321209542166262e-06,
+      "loss": 8.511,
+      "step": 184790
+    },
+    {
+      "epoch": 0.9228694848810207,
+      "grad_norm": 0.08670665323734283,
+      "learning_rate": 2.3197076272247117e-06,
+      "loss": 8.5134,
+      "step": 184800
+    },
+    {
+      "epoch": 0.9229194237059603,
+      "grad_norm": 0.09500810503959656,
+      "learning_rate": 2.318205712283161e-06,
+      "loss": 8.4904,
+      "step": 184810
+    },
+    {
+      "epoch": 0.9229693625308997,
+      "grad_norm": 0.09684941917657852,
+      "learning_rate": 2.3167037973416104e-06,
+      "loss": 8.5025,
+      "step": 184820
+    },
+    {
+      "epoch": 0.9230193013558391,
+      "grad_norm": 0.08873332291841507,
+      "learning_rate": 2.31520188240006e-06,
+      "loss": 8.4923,
+      "step": 184830
+    },
+    {
+      "epoch": 0.9230692401807785,
+      "grad_norm": 0.0964055061340332,
+      "learning_rate": 2.31369996745851e-06,
+      "loss": 8.5126,
+      "step": 184840
+    },
+    {
+      "epoch": 0.923119179005718,
+      "grad_norm": 0.09410259127616882,
+      "learning_rate": 2.3121980525169592e-06,
+      "loss": 8.5124,
+      "step": 184850
+    },
+    {
+      "epoch": 0.9231691178306575,
+      "grad_norm": 0.09654883295297623,
+      "learning_rate": 2.3106961375754086e-06,
+      "loss": 8.5073,
+      "step": 184860
+    },
+    {
+      "epoch": 0.9232190566555969,
+      "grad_norm": 0.08777954429388046,
+      "learning_rate": 2.309194222633858e-06,
+      "loss": 8.5067,
+      "step": 184870
+    },
+    {
+      "epoch": 0.9232689954805363,
+      "grad_norm": 0.09537537395954132,
+      "learning_rate": 2.307692307692308e-06,
+      "loss": 8.4888,
+      "step": 184880
+    },
+    {
+      "epoch": 0.9233189343054758,
+      "grad_norm": 0.08902040123939514,
+      "learning_rate": 2.3061903927507574e-06,
+      "loss": 8.5042,
+      "step": 184890
+    },
+    {
+      "epoch": 0.9233688731304153,
+      "grad_norm": 0.09366223216056824,
+      "learning_rate": 2.3046884778092067e-06,
+      "loss": 8.5255,
+      "step": 184900
+    },
+    {
+      "epoch": 0.9234188119553547,
+      "grad_norm": 0.08827183395624161,
+      "learning_rate": 2.303186562867656e-06,
+      "loss": 8.5054,
+      "step": 184910
+    },
+    {
+      "epoch": 0.9234687507802941,
+      "grad_norm": 0.09339195489883423,
+      "learning_rate": 2.301684647926106e-06,
+      "loss": 8.5159,
+      "step": 184920
+    },
+    {
+      "epoch": 0.9235186896052335,
+      "grad_norm": 0.09930474311113358,
+      "learning_rate": 2.3001827329845556e-06,
+      "loss": 8.4914,
+      "step": 184930
+    },
+    {
+      "epoch": 0.9235686284301731,
+      "grad_norm": 0.09311912208795547,
+      "learning_rate": 2.298680818043005e-06,
+      "loss": 8.513,
+      "step": 184940
+    },
+    {
+      "epoch": 0.9236185672551125,
+      "grad_norm": 0.09054923057556152,
+      "learning_rate": 2.2971789031014542e-06,
+      "loss": 8.5079,
+      "step": 184950
+    },
+    {
+      "epoch": 0.9236685060800519,
+      "grad_norm": 0.0914381593465805,
+      "learning_rate": 2.295676988159904e-06,
+      "loss": 8.5086,
+      "step": 184960
+    },
+    {
+      "epoch": 0.9237184449049913,
+      "grad_norm": 0.09014654159545898,
+      "learning_rate": 2.2941750732183533e-06,
+      "loss": 8.5171,
+      "step": 184970
+    },
+    {
+      "epoch": 0.9237683837299309,
+      "grad_norm": 0.0892617329955101,
+      "learning_rate": 2.292673158276803e-06,
+      "loss": 8.5193,
+      "step": 184980
+    },
+    {
+      "epoch": 0.9238183225548703,
+      "grad_norm": 0.09248284995555878,
+      "learning_rate": 2.2911712433352524e-06,
+      "loss": 8.5127,
+      "step": 184990
+    },
+    {
+      "epoch": 0.9238682613798097,
+      "grad_norm": 0.08743773400783539,
+      "learning_rate": 2.289669328393702e-06,
+      "loss": 8.523,
+      "step": 185000
+    },
+    {
+      "epoch": 0.9239182002047491,
+      "grad_norm": 0.08780092000961304,
+      "learning_rate": 2.2881674134521515e-06,
+      "loss": 8.5022,
+      "step": 185010
+    },
+    {
+      "epoch": 0.9239681390296887,
+      "grad_norm": 0.09048131853342056,
+      "learning_rate": 2.286665498510601e-06,
+      "loss": 8.5004,
+      "step": 185020
+    },
+    {
+      "epoch": 0.9240180778546281,
+      "grad_norm": 0.089464470744133,
+      "learning_rate": 2.2851635835690506e-06,
+      "loss": 8.51,
+      "step": 185030
+    },
+    {
+      "epoch": 0.9240680166795675,
+      "grad_norm": 0.09030899405479431,
+      "learning_rate": 2.2836616686275003e-06,
+      "loss": 8.5044,
+      "step": 185040
+    },
+    {
+      "epoch": 0.9241179555045069,
+      "grad_norm": 0.08874811232089996,
+      "learning_rate": 2.2821597536859497e-06,
+      "loss": 8.5056,
+      "step": 185050
+    },
+    {
+      "epoch": 0.9241678943294465,
+      "grad_norm": 0.08877728879451752,
+      "learning_rate": 2.280657838744399e-06,
+      "loss": 8.5067,
+      "step": 185060
+    },
+    {
+      "epoch": 0.9242178331543859,
+      "grad_norm": 0.09442177414894104,
+      "learning_rate": 2.2791559238028483e-06,
+      "loss": 8.5198,
+      "step": 185070
+    },
+    {
+      "epoch": 0.9242677719793253,
+      "grad_norm": 0.09836707264184952,
+      "learning_rate": 2.2776540088612985e-06,
+      "loss": 8.4915,
+      "step": 185080
+    },
+    {
+      "epoch": 0.9243177108042647,
+      "grad_norm": 0.08862404525279999,
+      "learning_rate": 2.276152093919748e-06,
+      "loss": 8.4892,
+      "step": 185090
+    },
+    {
+      "epoch": 0.9243676496292043,
+      "grad_norm": 0.09015829116106033,
+      "learning_rate": 2.274650178978197e-06,
+      "loss": 8.5203,
+      "step": 185100
+    },
+    {
+      "epoch": 0.9244175884541437,
+      "grad_norm": 0.0880509465932846,
+      "learning_rate": 2.2731482640366465e-06,
+      "loss": 8.5141,
+      "step": 185110
+    },
+    {
+      "epoch": 0.9244675272790831,
+      "grad_norm": 0.09141357243061066,
+      "learning_rate": 2.2716463490950963e-06,
+      "loss": 8.5056,
+      "step": 185120
+    },
+    {
+      "epoch": 0.9245174661040225,
+      "grad_norm": 0.09560871124267578,
+      "learning_rate": 2.270144434153546e-06,
+      "loss": 8.4859,
+      "step": 185130
+    },
+    {
+      "epoch": 0.9245674049289621,
+      "grad_norm": 0.09067047387361526,
+      "learning_rate": 2.2686425192119953e-06,
+      "loss": 8.5161,
+      "step": 185140
+    },
+    {
+      "epoch": 0.9246173437539015,
+      "grad_norm": 0.08690574020147324,
+      "learning_rate": 2.2671406042704447e-06,
+      "loss": 8.5297,
+      "step": 185150
+    },
+    {
+      "epoch": 0.9246672825788409,
+      "grad_norm": 0.09214247763156891,
+      "learning_rate": 2.2656386893288944e-06,
+      "loss": 8.5163,
+      "step": 185160
+    },
+    {
+      "epoch": 0.9247172214037803,
+      "grad_norm": 0.08881056308746338,
+      "learning_rate": 2.2641367743873438e-06,
+      "loss": 8.5007,
+      "step": 185170
+    },
+    {
+      "epoch": 0.9247671602287199,
+      "grad_norm": 0.09248036891222,
+      "learning_rate": 2.2626348594457935e-06,
+      "loss": 8.4997,
+      "step": 185180
+    },
+    {
+      "epoch": 0.9248170990536593,
+      "grad_norm": 0.08973632007837296,
+      "learning_rate": 2.261132944504243e-06,
+      "loss": 8.5202,
+      "step": 185190
+    },
+    {
+      "epoch": 0.9248670378785987,
+      "grad_norm": 0.09075845032930374,
+      "learning_rate": 2.2596310295626926e-06,
+      "loss": 8.4875,
+      "step": 185200
+    },
+    {
+      "epoch": 0.9249169767035381,
+      "grad_norm": 0.09289133548736572,
+      "learning_rate": 2.258129114621142e-06,
+      "loss": 8.5093,
+      "step": 185210
+    },
+    {
+      "epoch": 0.9249669155284777,
+      "grad_norm": 0.0958157405257225,
+      "learning_rate": 2.2566271996795913e-06,
+      "loss": 8.5047,
+      "step": 185220
+    },
+    {
+      "epoch": 0.9250168543534171,
+      "grad_norm": 0.09263189136981964,
+      "learning_rate": 2.255125284738041e-06,
+      "loss": 8.5116,
+      "step": 185230
+    },
+    {
+      "epoch": 0.9250667931783565,
+      "grad_norm": 0.08901172876358032,
+      "learning_rate": 2.2536233697964908e-06,
+      "loss": 8.4862,
+      "step": 185240
+    },
+    {
+      "epoch": 0.9251167320032959,
+      "grad_norm": 0.09158726781606674,
+      "learning_rate": 2.25212145485494e-06,
+      "loss": 8.5025,
+      "step": 185250
+    },
+    {
+      "epoch": 0.9251666708282354,
+      "grad_norm": 0.08551083505153656,
+      "learning_rate": 2.2506195399133894e-06,
+      "loss": 8.5062,
+      "step": 185260
+    },
+    {
+      "epoch": 0.9252166096531749,
+      "grad_norm": 0.09838327020406723,
+      "learning_rate": 2.2491176249718388e-06,
+      "loss": 8.5046,
+      "step": 185270
+    },
+    {
+      "epoch": 0.9252665484781143,
+      "grad_norm": 0.0905366763472557,
+      "learning_rate": 2.247615710030289e-06,
+      "loss": 8.5041,
+      "step": 185280
+    },
+    {
+      "epoch": 0.9253164873030537,
+      "grad_norm": 0.09381816536188126,
+      "learning_rate": 2.2461137950887383e-06,
+      "loss": 8.5001,
+      "step": 185290
+    },
+    {
+      "epoch": 0.9253664261279932,
+      "grad_norm": 0.09462891519069672,
+      "learning_rate": 2.2446118801471876e-06,
+      "loss": 8.505,
+      "step": 185300
+    },
+    {
+      "epoch": 0.9254163649529327,
+      "grad_norm": 0.09308251738548279,
+      "learning_rate": 2.243109965205637e-06,
+      "loss": 8.5048,
+      "step": 185310
+    },
+    {
+      "epoch": 0.9254663037778721,
+      "grad_norm": 0.09172924607992172,
+      "learning_rate": 2.2416080502640867e-06,
+      "loss": 8.5165,
+      "step": 185320
+    },
+    {
+      "epoch": 0.9255162426028115,
+      "grad_norm": 0.09043785184621811,
+      "learning_rate": 2.2401061353225365e-06,
+      "loss": 8.5263,
+      "step": 185330
+    },
+    {
+      "epoch": 0.925566181427751,
+      "grad_norm": 0.08776098489761353,
+      "learning_rate": 2.2386042203809858e-06,
+      "loss": 8.5158,
+      "step": 185340
+    },
+    {
+      "epoch": 0.9256161202526905,
+      "grad_norm": 0.09275338798761368,
+      "learning_rate": 2.237102305439435e-06,
+      "loss": 8.5181,
+      "step": 185350
+    },
+    {
+      "epoch": 0.9256660590776299,
+      "grad_norm": 0.09385108202695847,
+      "learning_rate": 2.235600390497885e-06,
+      "loss": 8.4931,
+      "step": 185360
+    },
+    {
+      "epoch": 0.9257159979025693,
+      "grad_norm": 0.0922423005104065,
+      "learning_rate": 2.2340984755563346e-06,
+      "loss": 8.5035,
+      "step": 185370
+    },
+    {
+      "epoch": 0.9257659367275088,
+      "grad_norm": 0.08742625266313553,
+      "learning_rate": 2.232596560614784e-06,
+      "loss": 8.5218,
+      "step": 185380
+    },
+    {
+      "epoch": 0.9258158755524483,
+      "grad_norm": 0.08818572014570236,
+      "learning_rate": 2.2310946456732333e-06,
+      "loss": 8.4944,
+      "step": 185390
+    },
+    {
+      "epoch": 0.9258658143773877,
+      "grad_norm": 0.0831325426697731,
+      "learning_rate": 2.229592730731683e-06,
+      "loss": 8.5027,
+      "step": 185400
+    },
+    {
+      "epoch": 0.9259157532023271,
+      "grad_norm": 0.09564302861690521,
+      "learning_rate": 2.2280908157901324e-06,
+      "loss": 8.485,
+      "step": 185410
+    },
+    {
+      "epoch": 0.9259656920272666,
+      "grad_norm": 0.08953244984149933,
+      "learning_rate": 2.226588900848582e-06,
+      "loss": 8.5,
+      "step": 185420
+    },
+    {
+      "epoch": 0.9260156308522061,
+      "grad_norm": 0.09346967190504074,
+      "learning_rate": 2.2250869859070315e-06,
+      "loss": 8.5088,
+      "step": 185430
+    },
+    {
+      "epoch": 0.9260655696771455,
+      "grad_norm": 0.09185745567083359,
+      "learning_rate": 2.2235850709654812e-06,
+      "loss": 8.5048,
+      "step": 185440
+    },
+    {
+      "epoch": 0.9261155085020849,
+      "grad_norm": 0.08855016529560089,
+      "learning_rate": 2.2220831560239305e-06,
+      "loss": 8.5062,
+      "step": 185450
+    },
+    {
+      "epoch": 0.9261654473270244,
+      "grad_norm": 0.09117954969406128,
+      "learning_rate": 2.22058124108238e-06,
+      "loss": 8.5,
+      "step": 185460
+    },
+    {
+      "epoch": 0.9262153861519639,
+      "grad_norm": 0.09156261384487152,
+      "learning_rate": 2.2190793261408296e-06,
+      "loss": 8.5025,
+      "step": 185470
+    },
+    {
+      "epoch": 0.9262653249769033,
+      "grad_norm": 0.09079194068908691,
+      "learning_rate": 2.2175774111992794e-06,
+      "loss": 8.5112,
+      "step": 185480
+    },
+    {
+      "epoch": 0.9263152638018427,
+      "grad_norm": 0.09392797201871872,
+      "learning_rate": 2.2160754962577287e-06,
+      "loss": 8.5097,
+      "step": 185490
+    },
+    {
+      "epoch": 0.9263652026267822,
+      "grad_norm": 0.09081744402647018,
+      "learning_rate": 2.214573581316178e-06,
+      "loss": 8.5035,
+      "step": 185500
+    },
+    {
+      "epoch": 0.9264151414517217,
+      "grad_norm": 0.09045954048633575,
+      "learning_rate": 2.2130716663746274e-06,
+      "loss": 8.4953,
+      "step": 185510
+    },
+    {
+      "epoch": 0.9264650802766611,
+      "grad_norm": 0.08934304118156433,
+      "learning_rate": 2.2115697514330776e-06,
+      "loss": 8.4913,
+      "step": 185520
+    },
+    {
+      "epoch": 0.9265150191016005,
+      "grad_norm": 0.0947798490524292,
+      "learning_rate": 2.210067836491527e-06,
+      "loss": 8.4957,
+      "step": 185530
+    },
+    {
+      "epoch": 0.92656495792654,
+      "grad_norm": 0.09003657102584839,
+      "learning_rate": 2.2085659215499762e-06,
+      "loss": 8.4949,
+      "step": 185540
+    },
+    {
+      "epoch": 0.9266148967514795,
+      "grad_norm": 0.08870958536863327,
+      "learning_rate": 2.2070640066084256e-06,
+      "loss": 8.5088,
+      "step": 185550
+    },
+    {
+      "epoch": 0.9266648355764189,
+      "grad_norm": 0.09150302410125732,
+      "learning_rate": 2.2055620916668753e-06,
+      "loss": 8.5054,
+      "step": 185560
+    },
+    {
+      "epoch": 0.9267147744013583,
+      "grad_norm": 0.08964120596647263,
+      "learning_rate": 2.204060176725325e-06,
+      "loss": 8.5195,
+      "step": 185570
+    },
+    {
+      "epoch": 0.9267647132262978,
+      "grad_norm": 0.0831717923283577,
+      "learning_rate": 2.2025582617837744e-06,
+      "loss": 8.5152,
+      "step": 185580
+    },
+    {
+      "epoch": 0.9268146520512373,
+      "grad_norm": 0.09838960319757462,
+      "learning_rate": 2.2010563468422237e-06,
+      "loss": 8.4911,
+      "step": 185590
+    },
+    {
+      "epoch": 0.9268645908761767,
+      "grad_norm": 0.08773133903741837,
+      "learning_rate": 2.1995544319006735e-06,
+      "loss": 8.5124,
+      "step": 185600
+    },
+    {
+      "epoch": 0.9269145297011161,
+      "grad_norm": 0.09229285269975662,
+      "learning_rate": 2.198052516959123e-06,
+      "loss": 8.4904,
+      "step": 185610
+    },
+    {
+      "epoch": 0.9269644685260556,
+      "grad_norm": 0.087303526699543,
+      "learning_rate": 2.1965506020175726e-06,
+      "loss": 8.5135,
+      "step": 185620
+    },
+    {
+      "epoch": 0.927014407350995,
+      "grad_norm": 0.09083783626556396,
+      "learning_rate": 2.195048687076022e-06,
+      "loss": 8.4835,
+      "step": 185630
+    },
+    {
+      "epoch": 0.9270643461759345,
+      "grad_norm": 0.09213201701641083,
+      "learning_rate": 2.1935467721344717e-06,
+      "loss": 8.5163,
+      "step": 185640
+    },
+    {
+      "epoch": 0.9271142850008739,
+      "grad_norm": 0.09071563184261322,
+      "learning_rate": 2.192044857192921e-06,
+      "loss": 8.5098,
+      "step": 185650
+    },
+    {
+      "epoch": 0.9271642238258134,
+      "grad_norm": 0.08385904878377914,
+      "learning_rate": 2.1905429422513703e-06,
+      "loss": 8.503,
+      "step": 185660
+    },
+    {
+      "epoch": 0.9272141626507528,
+      "grad_norm": 0.09102963656187057,
+      "learning_rate": 2.18904102730982e-06,
+      "loss": 8.4991,
+      "step": 185670
+    },
+    {
+      "epoch": 0.9272641014756923,
+      "grad_norm": 0.08781067281961441,
+      "learning_rate": 2.18753911236827e-06,
+      "loss": 8.5112,
+      "step": 185680
+    },
+    {
+      "epoch": 0.9273140403006317,
+      "grad_norm": 0.09064462780952454,
+      "learning_rate": 2.186037197426719e-06,
+      "loss": 8.5103,
+      "step": 185690
+    },
+    {
+      "epoch": 0.9273639791255712,
+      "grad_norm": 0.09395837038755417,
+      "learning_rate": 2.1845352824851685e-06,
+      "loss": 8.5011,
+      "step": 185700
+    },
+    {
+      "epoch": 0.9274139179505106,
+      "grad_norm": 0.08788859099149704,
+      "learning_rate": 2.183033367543618e-06,
+      "loss": 8.4993,
+      "step": 185710
+    },
+    {
+      "epoch": 0.9274638567754501,
+      "grad_norm": 0.09015844762325287,
+      "learning_rate": 2.181531452602068e-06,
+      "loss": 8.5091,
+      "step": 185720
+    },
+    {
+      "epoch": 0.9275137956003895,
+      "grad_norm": 0.08770440518856049,
+      "learning_rate": 2.1800295376605173e-06,
+      "loss": 8.4965,
+      "step": 185730
+    },
+    {
+      "epoch": 0.927563734425329,
+      "grad_norm": 0.09498754143714905,
+      "learning_rate": 2.1785276227189667e-06,
+      "loss": 8.4999,
+      "step": 185740
+    },
+    {
+      "epoch": 0.9276136732502684,
+      "grad_norm": 0.09174376726150513,
+      "learning_rate": 2.177025707777416e-06,
+      "loss": 8.509,
+      "step": 185750
+    },
+    {
+      "epoch": 0.9276636120752079,
+      "grad_norm": 0.08917669951915741,
+      "learning_rate": 2.1755237928358658e-06,
+      "loss": 8.5181,
+      "step": 185760
+    },
+    {
+      "epoch": 0.9277135509001473,
+      "grad_norm": 0.0941305160522461,
+      "learning_rate": 2.1740218778943155e-06,
+      "loss": 8.5051,
+      "step": 185770
+    },
+    {
+      "epoch": 0.9277634897250868,
+      "grad_norm": 0.09079430997371674,
+      "learning_rate": 2.172519962952765e-06,
+      "loss": 8.515,
+      "step": 185780
+    },
+    {
+      "epoch": 0.9278134285500262,
+      "grad_norm": 0.09007134288549423,
+      "learning_rate": 2.171018048011214e-06,
+      "loss": 8.4849,
+      "step": 185790
+    },
+    {
+      "epoch": 0.9278633673749657,
+      "grad_norm": 0.09367641806602478,
+      "learning_rate": 2.169516133069664e-06,
+      "loss": 8.5065,
+      "step": 185800
+    },
+    {
+      "epoch": 0.9279133061999051,
+      "grad_norm": 0.08902136236429214,
+      "learning_rate": 2.1680142181281133e-06,
+      "loss": 8.5059,
+      "step": 185810
+    },
+    {
+      "epoch": 0.9279632450248446,
+      "grad_norm": 0.08988285809755325,
+      "learning_rate": 2.166512303186563e-06,
+      "loss": 8.5104,
+      "step": 185820
+    },
+    {
+      "epoch": 0.928013183849784,
+      "grad_norm": 0.09199918806552887,
+      "learning_rate": 2.1650103882450123e-06,
+      "loss": 8.5187,
+      "step": 185830
+    },
+    {
+      "epoch": 0.9280631226747235,
+      "grad_norm": 0.08583081513643265,
+      "learning_rate": 2.163508473303462e-06,
+      "loss": 8.4948,
+      "step": 185840
+    },
+    {
+      "epoch": 0.9281130614996629,
+      "grad_norm": 0.09111568331718445,
+      "learning_rate": 2.1620065583619114e-06,
+      "loss": 8.5025,
+      "step": 185850
+    },
+    {
+      "epoch": 0.9281630003246024,
+      "grad_norm": 0.08850061148405075,
+      "learning_rate": 2.160504643420361e-06,
+      "loss": 8.5003,
+      "step": 185860
+    },
+    {
+      "epoch": 0.9282129391495418,
+      "grad_norm": 0.09337007999420166,
+      "learning_rate": 2.1590027284788105e-06,
+      "loss": 8.5119,
+      "step": 185870
+    },
+    {
+      "epoch": 0.9282628779744813,
+      "grad_norm": 0.08924791216850281,
+      "learning_rate": 2.1575008135372603e-06,
+      "loss": 8.5174,
+      "step": 185880
+    },
+    {
+      "epoch": 0.9283128167994207,
+      "grad_norm": 0.09130074083805084,
+      "learning_rate": 2.1559988985957096e-06,
+      "loss": 8.5219,
+      "step": 185890
+    },
+    {
+      "epoch": 0.9283627556243602,
+      "grad_norm": 0.0915064811706543,
+      "learning_rate": 2.154496983654159e-06,
+      "loss": 8.5056,
+      "step": 185900
+    },
+    {
+      "epoch": 0.9284126944492996,
+      "grad_norm": 0.09288337826728821,
+      "learning_rate": 2.1529950687126087e-06,
+      "loss": 8.4999,
+      "step": 185910
+    },
+    {
+      "epoch": 0.9284626332742391,
+      "grad_norm": 0.08676610141992569,
+      "learning_rate": 2.1514931537710584e-06,
+      "loss": 8.514,
+      "step": 185920
+    },
+    {
+      "epoch": 0.9285125720991785,
+      "grad_norm": 0.08682654052972794,
+      "learning_rate": 2.1499912388295078e-06,
+      "loss": 8.5044,
+      "step": 185930
+    },
+    {
+      "epoch": 0.9285625109241179,
+      "grad_norm": 0.0899660736322403,
+      "learning_rate": 2.148489323887957e-06,
+      "loss": 8.5023,
+      "step": 185940
+    },
+    {
+      "epoch": 0.9286124497490574,
+      "grad_norm": 0.09490537643432617,
+      "learning_rate": 2.1469874089464064e-06,
+      "loss": 8.4954,
+      "step": 185950
+    },
+    {
+      "epoch": 0.9286623885739969,
+      "grad_norm": 0.08999119699001312,
+      "learning_rate": 2.1454854940048566e-06,
+      "loss": 8.4927,
+      "step": 185960
+    },
+    {
+      "epoch": 0.9287123273989363,
+      "grad_norm": 0.09169169515371323,
+      "learning_rate": 2.143983579063306e-06,
+      "loss": 8.5092,
+      "step": 185970
+    },
+    {
+      "epoch": 0.9287622662238757,
+      "grad_norm": 0.09515541046857834,
+      "learning_rate": 2.1424816641217553e-06,
+      "loss": 8.4973,
+      "step": 185980
+    },
+    {
+      "epoch": 0.9288122050488152,
+      "grad_norm": 0.09675678610801697,
+      "learning_rate": 2.1409797491802046e-06,
+      "loss": 8.49,
+      "step": 185990
+    },
+    {
+      "epoch": 0.9288621438737547,
+      "grad_norm": 0.08823744207620621,
+      "learning_rate": 2.1394778342386544e-06,
+      "loss": 8.512,
+      "step": 186000
+    },
+    {
+      "epoch": 0.9289120826986941,
+      "grad_norm": 0.09296853095293045,
+      "learning_rate": 2.137975919297104e-06,
+      "loss": 8.5053,
+      "step": 186010
+    },
+    {
+      "epoch": 0.9289620215236335,
+      "grad_norm": 0.08902104943990707,
+      "learning_rate": 2.1364740043555534e-06,
+      "loss": 8.5037,
+      "step": 186020
+    },
+    {
+      "epoch": 0.929011960348573,
+      "grad_norm": 0.094472736120224,
+      "learning_rate": 2.1349720894140028e-06,
+      "loss": 8.4983,
+      "step": 186030
+    },
+    {
+      "epoch": 0.9290618991735125,
+      "grad_norm": 0.09332358092069626,
+      "learning_rate": 2.1334701744724525e-06,
+      "loss": 8.5144,
+      "step": 186040
+    },
+    {
+      "epoch": 0.9291118379984519,
+      "grad_norm": 0.08705665916204453,
+      "learning_rate": 2.131968259530902e-06,
+      "loss": 8.5012,
+      "step": 186050
+    },
+    {
+      "epoch": 0.9291617768233913,
+      "grad_norm": 0.08535271137952805,
+      "learning_rate": 2.1304663445893516e-06,
+      "loss": 8.5141,
+      "step": 186060
+    },
+    {
+      "epoch": 0.9292117156483308,
+      "grad_norm": 0.09060922265052795,
+      "learning_rate": 2.128964429647801e-06,
+      "loss": 8.507,
+      "step": 186070
+    },
+    {
+      "epoch": 0.9292616544732702,
+      "grad_norm": 0.09175438433885574,
+      "learning_rate": 2.1274625147062507e-06,
+      "loss": 8.5066,
+      "step": 186080
+    },
+    {
+      "epoch": 0.9293115932982097,
+      "grad_norm": 0.08872448652982712,
+      "learning_rate": 2.1259605997647e-06,
+      "loss": 8.5203,
+      "step": 186090
+    },
+    {
+      "epoch": 0.9293615321231491,
+      "grad_norm": 0.0916357934474945,
+      "learning_rate": 2.1244586848231494e-06,
+      "loss": 8.5022,
+      "step": 186100
+    },
+    {
+      "epoch": 0.9294114709480886,
+      "grad_norm": 0.08553195744752884,
+      "learning_rate": 2.122956769881599e-06,
+      "loss": 8.5101,
+      "step": 186110
+    },
+    {
+      "epoch": 0.929461409773028,
+      "grad_norm": 0.09251870959997177,
+      "learning_rate": 2.121454854940049e-06,
+      "loss": 8.5139,
+      "step": 186120
+    },
+    {
+      "epoch": 0.9295113485979675,
+      "grad_norm": 0.0898064374923706,
+      "learning_rate": 2.119952939998498e-06,
+      "loss": 8.5184,
+      "step": 186130
+    },
+    {
+      "epoch": 0.9295612874229069,
+      "grad_norm": 0.09209618717432022,
+      "learning_rate": 2.1184510250569475e-06,
+      "loss": 8.5068,
+      "step": 186140
+    },
+    {
+      "epoch": 0.9296112262478464,
+      "grad_norm": 0.08947913348674774,
+      "learning_rate": 2.116949110115397e-06,
+      "loss": 8.512,
+      "step": 186150
+    },
+    {
+      "epoch": 0.9296611650727858,
+      "grad_norm": 0.08615361899137497,
+      "learning_rate": 2.115447195173847e-06,
+      "loss": 8.5051,
+      "step": 186160
+    },
+    {
+      "epoch": 0.9297111038977253,
+      "grad_norm": 0.08832980692386627,
+      "learning_rate": 2.1139452802322964e-06,
+      "loss": 8.5106,
+      "step": 186170
+    },
+    {
+      "epoch": 0.9297610427226647,
+      "grad_norm": 0.09645520150661469,
+      "learning_rate": 2.1124433652907457e-06,
+      "loss": 8.4926,
+      "step": 186180
+    },
+    {
+      "epoch": 0.9298109815476042,
+      "grad_norm": 0.09340712428092957,
+      "learning_rate": 2.110941450349195e-06,
+      "loss": 8.4987,
+      "step": 186190
+    },
+    {
+      "epoch": 0.9298609203725436,
+      "grad_norm": 0.08868103474378586,
+      "learning_rate": 2.109439535407645e-06,
+      "loss": 8.4996,
+      "step": 186200
+    },
+    {
+      "epoch": 0.9299108591974831,
+      "grad_norm": 0.09115692973136902,
+      "learning_rate": 2.1079376204660946e-06,
+      "loss": 8.5084,
+      "step": 186210
+    },
+    {
+      "epoch": 0.9299607980224225,
+      "grad_norm": 0.0927361398935318,
+      "learning_rate": 2.106435705524544e-06,
+      "loss": 8.5019,
+      "step": 186220
+    },
+    {
+      "epoch": 0.930010736847362,
+      "grad_norm": 0.08556483685970306,
+      "learning_rate": 2.1049337905829932e-06,
+      "loss": 8.5105,
+      "step": 186230
+    },
+    {
+      "epoch": 0.9300606756723014,
+      "grad_norm": 0.09068150818347931,
+      "learning_rate": 2.103431875641443e-06,
+      "loss": 8.5143,
+      "step": 186240
+    },
+    {
+      "epoch": 0.9301106144972409,
+      "grad_norm": 0.09075603634119034,
+      "learning_rate": 2.1019299606998923e-06,
+      "loss": 8.5138,
+      "step": 186250
+    },
+    {
+      "epoch": 0.9301605533221803,
+      "grad_norm": 0.09026709944009781,
+      "learning_rate": 2.100428045758342e-06,
+      "loss": 8.5206,
+      "step": 186260
+    },
+    {
+      "epoch": 0.9302104921471198,
+      "grad_norm": 0.09280247241258621,
+      "learning_rate": 2.0989261308167914e-06,
+      "loss": 8.5057,
+      "step": 186270
+    },
+    {
+      "epoch": 0.9302604309720592,
+      "grad_norm": 0.08981787413358688,
+      "learning_rate": 2.097424215875241e-06,
+      "loss": 8.5119,
+      "step": 186280
+    },
+    {
+      "epoch": 0.9303103697969987,
+      "grad_norm": 0.09122589975595474,
+      "learning_rate": 2.0959223009336905e-06,
+      "loss": 8.4926,
+      "step": 186290
+    },
+    {
+      "epoch": 0.9303603086219381,
+      "grad_norm": 0.09164507687091827,
+      "learning_rate": 2.09442038599214e-06,
+      "loss": 8.5077,
+      "step": 186300
+    },
+    {
+      "epoch": 0.9304102474468776,
+      "grad_norm": 0.089156374335289,
+      "learning_rate": 2.0929184710505896e-06,
+      "loss": 8.5004,
+      "step": 186310
+    },
+    {
+      "epoch": 0.930460186271817,
+      "grad_norm": 0.08851976692676544,
+      "learning_rate": 2.0914165561090393e-06,
+      "loss": 8.5076,
+      "step": 186320
+    },
+    {
+      "epoch": 0.9305101250967565,
+      "grad_norm": 0.08928805589675903,
+      "learning_rate": 2.0899146411674886e-06,
+      "loss": 8.5166,
+      "step": 186330
+    },
+    {
+      "epoch": 0.9305600639216959,
+      "grad_norm": 0.09302154183387756,
+      "learning_rate": 2.088412726225938e-06,
+      "loss": 8.504,
+      "step": 186340
+    },
+    {
+      "epoch": 0.9306100027466354,
+      "grad_norm": 0.09193503111600876,
+      "learning_rate": 2.0869108112843873e-06,
+      "loss": 8.5065,
+      "step": 186350
+    },
+    {
+      "epoch": 0.9306599415715748,
+      "grad_norm": 0.09432563930749893,
+      "learning_rate": 2.0854088963428375e-06,
+      "loss": 8.5003,
+      "step": 186360
+    },
+    {
+      "epoch": 0.9307098803965143,
+      "grad_norm": 0.0901406854391098,
+      "learning_rate": 2.083906981401287e-06,
+      "loss": 8.5008,
+      "step": 186370
+    },
+    {
+      "epoch": 0.9307598192214537,
+      "grad_norm": 0.09133270382881165,
+      "learning_rate": 2.082405066459736e-06,
+      "loss": 8.4996,
+      "step": 186380
+    },
+    {
+      "epoch": 0.9308097580463932,
+      "grad_norm": 0.08889655023813248,
+      "learning_rate": 2.0809031515181855e-06,
+      "loss": 8.5021,
+      "step": 186390
+    },
+    {
+      "epoch": 0.9308596968713326,
+      "grad_norm": 0.08690587431192398,
+      "learning_rate": 2.0794012365766352e-06,
+      "loss": 8.4902,
+      "step": 186400
+    },
+    {
+      "epoch": 0.930909635696272,
+      "grad_norm": 0.08836494386196136,
+      "learning_rate": 2.077899321635085e-06,
+      "loss": 8.5141,
+      "step": 186410
+    },
+    {
+      "epoch": 0.9309595745212115,
+      "grad_norm": 0.09551975876092911,
+      "learning_rate": 2.0763974066935343e-06,
+      "loss": 8.4937,
+      "step": 186420
+    },
+    {
+      "epoch": 0.931009513346151,
+      "grad_norm": 0.08524799346923828,
+      "learning_rate": 2.0748954917519837e-06,
+      "loss": 8.5133,
+      "step": 186430
+    },
+    {
+      "epoch": 0.9310594521710904,
+      "grad_norm": 0.09045319259166718,
+      "learning_rate": 2.0733935768104334e-06,
+      "loss": 8.5013,
+      "step": 186440
+    },
+    {
+      "epoch": 0.9311093909960299,
+      "grad_norm": 0.08845016360282898,
+      "learning_rate": 2.071891661868883e-06,
+      "loss": 8.5065,
+      "step": 186450
+    },
+    {
+      "epoch": 0.9311593298209693,
+      "grad_norm": 0.08880577981472015,
+      "learning_rate": 2.0703897469273325e-06,
+      "loss": 8.4971,
+      "step": 186460
+    },
+    {
+      "epoch": 0.9312092686459088,
+      "grad_norm": 0.08805317431688309,
+      "learning_rate": 2.068887831985782e-06,
+      "loss": 8.4976,
+      "step": 186470
+    },
+    {
+      "epoch": 0.9312592074708482,
+      "grad_norm": 0.08898577839136124,
+      "learning_rate": 2.0673859170442316e-06,
+      "loss": 8.4908,
+      "step": 186480
+    },
+    {
+      "epoch": 0.9313091462957876,
+      "grad_norm": 0.08940175920724869,
+      "learning_rate": 2.065884002102681e-06,
+      "loss": 8.5027,
+      "step": 186490
+    },
+    {
+      "epoch": 0.9313590851207271,
+      "grad_norm": 0.08669328689575195,
+      "learning_rate": 2.0643820871611307e-06,
+      "loss": 8.5035,
+      "step": 186500
+    },
+    {
+      "epoch": 0.9314090239456666,
+      "grad_norm": 0.08564577996730804,
+      "learning_rate": 2.06288017221958e-06,
+      "loss": 8.4989,
+      "step": 186510
+    },
+    {
+      "epoch": 0.931458962770606,
+      "grad_norm": 0.09756655246019363,
+      "learning_rate": 2.0613782572780298e-06,
+      "loss": 8.5036,
+      "step": 186520
+    },
+    {
+      "epoch": 0.9315089015955454,
+      "grad_norm": 0.08926945179700851,
+      "learning_rate": 2.059876342336479e-06,
+      "loss": 8.5037,
+      "step": 186530
+    },
+    {
+      "epoch": 0.9315588404204849,
+      "grad_norm": 0.08959157764911652,
+      "learning_rate": 2.0583744273949284e-06,
+      "loss": 8.492,
+      "step": 186540
+    },
+    {
+      "epoch": 0.9316087792454244,
+      "grad_norm": 0.09301578253507614,
+      "learning_rate": 2.056872512453378e-06,
+      "loss": 8.5004,
+      "step": 186550
+    },
+    {
+      "epoch": 0.9316587180703638,
+      "grad_norm": 0.09195592999458313,
+      "learning_rate": 2.055370597511828e-06,
+      "loss": 8.5058,
+      "step": 186560
+    },
+    {
+      "epoch": 0.9317086568953032,
+      "grad_norm": 0.08996932953596115,
+      "learning_rate": 2.0538686825702773e-06,
+      "loss": 8.504,
+      "step": 186570
+    },
+    {
+      "epoch": 0.9317585957202427,
+      "grad_norm": 0.09332447499036789,
+      "learning_rate": 2.0523667676287266e-06,
+      "loss": 8.4953,
+      "step": 186580
+    },
+    {
+      "epoch": 0.9318085345451822,
+      "grad_norm": 0.08750541508197784,
+      "learning_rate": 2.050864852687176e-06,
+      "loss": 8.5028,
+      "step": 186590
+    },
+    {
+      "epoch": 0.9318584733701216,
+      "grad_norm": 0.08709759265184402,
+      "learning_rate": 2.049362937745626e-06,
+      "loss": 8.5079,
+      "step": 186600
+    },
+    {
+      "epoch": 0.931908412195061,
+      "grad_norm": 0.09282954037189484,
+      "learning_rate": 2.0478610228040754e-06,
+      "loss": 8.5121,
+      "step": 186610
+    },
+    {
+      "epoch": 0.9319583510200005,
+      "grad_norm": 0.08732730150222778,
+      "learning_rate": 2.0463591078625248e-06,
+      "loss": 8.5083,
+      "step": 186620
+    },
+    {
+      "epoch": 0.93200828984494,
+      "grad_norm": 0.09398452192544937,
+      "learning_rate": 2.044857192920974e-06,
+      "loss": 8.5088,
+      "step": 186630
+    },
+    {
+      "epoch": 0.9320582286698794,
+      "grad_norm": 0.09803569316864014,
+      "learning_rate": 2.043355277979424e-06,
+      "loss": 8.5001,
+      "step": 186640
+    },
+    {
+      "epoch": 0.9321081674948188,
+      "grad_norm": 0.08622153103351593,
+      "learning_rate": 2.0418533630378736e-06,
+      "loss": 8.5101,
+      "step": 186650
+    },
+    {
+      "epoch": 0.9321581063197583,
+      "grad_norm": 0.0959283858537674,
+      "learning_rate": 2.040351448096323e-06,
+      "loss": 8.5072,
+      "step": 186660
+    },
+    {
+      "epoch": 0.9322080451446978,
+      "grad_norm": 0.08934350311756134,
+      "learning_rate": 2.0388495331547723e-06,
+      "loss": 8.4896,
+      "step": 186670
+    },
+    {
+      "epoch": 0.9322579839696372,
+      "grad_norm": 0.09462158381938934,
+      "learning_rate": 2.037347618213222e-06,
+      "loss": 8.5083,
+      "step": 186680
+    },
+    {
+      "epoch": 0.9323079227945766,
+      "grad_norm": 0.09019871056079865,
+      "learning_rate": 2.0358457032716714e-06,
+      "loss": 8.5093,
+      "step": 186690
+    },
+    {
+      "epoch": 0.9323578616195161,
+      "grad_norm": 0.0971764400601387,
+      "learning_rate": 2.034343788330121e-06,
+      "loss": 8.5032,
+      "step": 186700
+    },
+    {
+      "epoch": 0.9324078004444556,
+      "grad_norm": 0.09243464469909668,
+      "learning_rate": 2.0328418733885704e-06,
+      "loss": 8.5077,
+      "step": 186710
+    },
+    {
+      "epoch": 0.932457739269395,
+      "grad_norm": 0.09595301002264023,
+      "learning_rate": 2.03133995844702e-06,
+      "loss": 8.4983,
+      "step": 186720
+    },
+    {
+      "epoch": 0.9325076780943344,
+      "grad_norm": 0.09078612178564072,
+      "learning_rate": 2.0298380435054695e-06,
+      "loss": 8.5082,
+      "step": 186730
+    },
+    {
+      "epoch": 0.9325576169192739,
+      "grad_norm": 0.09375286102294922,
+      "learning_rate": 2.028336128563919e-06,
+      "loss": 8.518,
+      "step": 186740
+    },
+    {
+      "epoch": 0.9326075557442134,
+      "grad_norm": 0.08845958113670349,
+      "learning_rate": 2.0268342136223686e-06,
+      "loss": 8.5098,
+      "step": 186750
+    },
+    {
+      "epoch": 0.9326574945691528,
+      "grad_norm": 0.08810067176818848,
+      "learning_rate": 2.0253322986808184e-06,
+      "loss": 8.4912,
+      "step": 186760
+    },
+    {
+      "epoch": 0.9327074333940922,
+      "grad_norm": 0.08972468227148056,
+      "learning_rate": 2.0238303837392677e-06,
+      "loss": 8.52,
+      "step": 186770
+    },
+    {
+      "epoch": 0.9327573722190317,
+      "grad_norm": 0.09303756058216095,
+      "learning_rate": 2.022328468797717e-06,
+      "loss": 8.5024,
+      "step": 186780
+    },
+    {
+      "epoch": 0.9328073110439712,
+      "grad_norm": 0.09837578982114792,
+      "learning_rate": 2.0208265538561664e-06,
+      "loss": 8.5024,
+      "step": 186790
+    },
+    {
+      "epoch": 0.9328572498689106,
+      "grad_norm": 0.08584380894899368,
+      "learning_rate": 2.0193246389146165e-06,
+      "loss": 8.5302,
+      "step": 186800
+    },
+    {
+      "epoch": 0.93290718869385,
+      "grad_norm": 0.09214811772108078,
+      "learning_rate": 2.017822723973066e-06,
+      "loss": 8.5084,
+      "step": 186810
+    },
+    {
+      "epoch": 0.9329571275187895,
+      "grad_norm": 0.08620260655879974,
+      "learning_rate": 2.016320809031515e-06,
+      "loss": 8.5275,
+      "step": 186820
+    },
+    {
+      "epoch": 0.933007066343729,
+      "grad_norm": 0.09670999646186829,
+      "learning_rate": 2.0148188940899645e-06,
+      "loss": 8.493,
+      "step": 186830
+    },
+    {
+      "epoch": 0.9330570051686684,
+      "grad_norm": 0.08751846104860306,
+      "learning_rate": 2.0133169791484143e-06,
+      "loss": 8.5126,
+      "step": 186840
+    },
+    {
+      "epoch": 0.9331069439936078,
+      "grad_norm": 0.0873984843492508,
+      "learning_rate": 2.011815064206864e-06,
+      "loss": 8.5206,
+      "step": 186850
+    },
+    {
+      "epoch": 0.9331568828185473,
+      "grad_norm": 0.0933077484369278,
+      "learning_rate": 2.0103131492653134e-06,
+      "loss": 8.5069,
+      "step": 186860
+    },
+    {
+      "epoch": 0.9332068216434868,
+      "grad_norm": 0.09100789576768875,
+      "learning_rate": 2.0088112343237627e-06,
+      "loss": 8.5076,
+      "step": 186870
+    },
+    {
+      "epoch": 0.9332567604684262,
+      "grad_norm": 0.09495274722576141,
+      "learning_rate": 2.0073093193822125e-06,
+      "loss": 8.5025,
+      "step": 186880
+    },
+    {
+      "epoch": 0.9333066992933656,
+      "grad_norm": 0.08778039366006851,
+      "learning_rate": 2.005807404440662e-06,
+      "loss": 8.5001,
+      "step": 186890
+    },
+    {
+      "epoch": 0.933356638118305,
+      "grad_norm": 0.09381724148988724,
+      "learning_rate": 2.0043054894991115e-06,
+      "loss": 8.5296,
+      "step": 186900
+    },
+    {
+      "epoch": 0.9334065769432445,
+      "grad_norm": 0.08771809935569763,
+      "learning_rate": 2.002803574557561e-06,
+      "loss": 8.5012,
+      "step": 186910
+    },
+    {
+      "epoch": 0.933456515768184,
+      "grad_norm": 0.08947025239467621,
+      "learning_rate": 2.0013016596160106e-06,
+      "loss": 8.5203,
+      "step": 186920
+    },
+    {
+      "epoch": 0.9335064545931234,
+      "grad_norm": 0.09270720183849335,
+      "learning_rate": 1.99979974467446e-06,
+      "loss": 8.5097,
+      "step": 186930
+    },
+    {
+      "epoch": 0.9335563934180628,
+      "grad_norm": 0.09039568901062012,
+      "learning_rate": 1.9982978297329093e-06,
+      "loss": 8.4951,
+      "step": 186940
+    },
+    {
+      "epoch": 0.9336063322430023,
+      "grad_norm": 0.09502692520618439,
+      "learning_rate": 1.996795914791359e-06,
+      "loss": 8.4986,
+      "step": 186950
+    },
+    {
+      "epoch": 0.9336562710679418,
+      "grad_norm": 0.0894678607583046,
+      "learning_rate": 1.995293999849809e-06,
+      "loss": 8.5048,
+      "step": 186960
+    },
+    {
+      "epoch": 0.9337062098928812,
+      "grad_norm": 0.08719871193170547,
+      "learning_rate": 1.993792084908258e-06,
+      "loss": 8.5154,
+      "step": 186970
+    },
+    {
+      "epoch": 0.9337561487178206,
+      "grad_norm": 0.09486103057861328,
+      "learning_rate": 1.9922901699667075e-06,
+      "loss": 8.523,
+      "step": 186980
+    },
+    {
+      "epoch": 0.9338060875427601,
+      "grad_norm": 0.09258001297712326,
+      "learning_rate": 1.9907882550251572e-06,
+      "loss": 8.5054,
+      "step": 186990
+    },
+    {
+      "epoch": 0.9338560263676996,
+      "grad_norm": 0.08995097130537033,
+      "learning_rate": 1.989286340083607e-06,
+      "loss": 8.4992,
+      "step": 187000
+    },
+    {
+      "epoch": 0.933905965192639,
+      "grad_norm": 0.09509983658790588,
+      "learning_rate": 1.9877844251420563e-06,
+      "loss": 8.5136,
+      "step": 187010
+    },
+    {
+      "epoch": 0.9339559040175784,
+      "grad_norm": 0.09256593883037567,
+      "learning_rate": 1.9862825102005056e-06,
+      "loss": 8.524,
+      "step": 187020
+    },
+    {
+      "epoch": 0.9340058428425179,
+      "grad_norm": 0.09413289278745651,
+      "learning_rate": 1.984780595258955e-06,
+      "loss": 8.4867,
+      "step": 187030
+    },
+    {
+      "epoch": 0.9340557816674574,
+      "grad_norm": 0.09102970361709595,
+      "learning_rate": 1.983278680317405e-06,
+      "loss": 8.4949,
+      "step": 187040
+    },
+    {
+      "epoch": 0.9341057204923968,
+      "grad_norm": 0.09890950471162796,
+      "learning_rate": 1.9817767653758545e-06,
+      "loss": 8.506,
+      "step": 187050
+    },
+    {
+      "epoch": 0.9341556593173362,
+      "grad_norm": 0.09198985993862152,
+      "learning_rate": 1.980274850434304e-06,
+      "loss": 8.5197,
+      "step": 187060
+    },
+    {
+      "epoch": 0.9342055981422757,
+      "grad_norm": 0.09256508946418762,
+      "learning_rate": 1.978772935492753e-06,
+      "loss": 8.501,
+      "step": 187070
+    },
+    {
+      "epoch": 0.9342555369672152,
+      "grad_norm": 0.08654509484767914,
+      "learning_rate": 1.977271020551203e-06,
+      "loss": 8.5061,
+      "step": 187080
+    },
+    {
+      "epoch": 0.9343054757921546,
+      "grad_norm": 0.08977734297513962,
+      "learning_rate": 1.9757691056096527e-06,
+      "loss": 8.5108,
+      "step": 187090
+    },
+    {
+      "epoch": 0.934355414617094,
+      "grad_norm": 0.08983021229505539,
+      "learning_rate": 1.974267190668102e-06,
+      "loss": 8.4969,
+      "step": 187100
+    },
+    {
+      "epoch": 0.9344053534420335,
+      "grad_norm": 0.09085972607135773,
+      "learning_rate": 1.9727652757265513e-06,
+      "loss": 8.5194,
+      "step": 187110
+    },
+    {
+      "epoch": 0.934455292266973,
+      "grad_norm": 0.08963587135076523,
+      "learning_rate": 1.971263360785001e-06,
+      "loss": 8.4978,
+      "step": 187120
+    },
+    {
+      "epoch": 0.9345052310919124,
+      "grad_norm": 0.08823879063129425,
+      "learning_rate": 1.9697614458434504e-06,
+      "loss": 8.5022,
+      "step": 187130
+    },
+    {
+      "epoch": 0.9345551699168518,
+      "grad_norm": 0.08996618539094925,
+      "learning_rate": 1.9682595309019e-06,
+      "loss": 8.5172,
+      "step": 187140
+    },
+    {
+      "epoch": 0.9346051087417913,
+      "grad_norm": 0.08975294977426529,
+      "learning_rate": 1.9667576159603495e-06,
+      "loss": 8.4987,
+      "step": 187150
+    },
+    {
+      "epoch": 0.9346550475667308,
+      "grad_norm": 0.08694805204868317,
+      "learning_rate": 1.965255701018799e-06,
+      "loss": 8.5241,
+      "step": 187160
+    },
+    {
+      "epoch": 0.9347049863916702,
+      "grad_norm": 0.09013424068689346,
+      "learning_rate": 1.9637537860772486e-06,
+      "loss": 8.5266,
+      "step": 187170
+    },
+    {
+      "epoch": 0.9347549252166096,
+      "grad_norm": 0.09890510141849518,
+      "learning_rate": 1.962251871135698e-06,
+      "loss": 8.514,
+      "step": 187180
+    },
+    {
+      "epoch": 0.934804864041549,
+      "grad_norm": 0.08957573026418686,
+      "learning_rate": 1.9607499561941477e-06,
+      "loss": 8.5077,
+      "step": 187190
+    },
+    {
+      "epoch": 0.9348548028664886,
+      "grad_norm": 0.09044113010168076,
+      "learning_rate": 1.959248041252597e-06,
+      "loss": 8.4931,
+      "step": 187200
+    },
+    {
+      "epoch": 0.934904741691428,
+      "grad_norm": 0.08695122599601746,
+      "learning_rate": 1.9577461263110467e-06,
+      "loss": 8.4954,
+      "step": 187210
+    },
+    {
+      "epoch": 0.9349546805163674,
+      "grad_norm": 0.09316051751375198,
+      "learning_rate": 1.956244211369496e-06,
+      "loss": 8.515,
+      "step": 187220
+    },
+    {
+      "epoch": 0.9350046193413069,
+      "grad_norm": 0.0904255136847496,
+      "learning_rate": 1.9547422964279454e-06,
+      "loss": 8.4974,
+      "step": 187230
+    },
+    {
+      "epoch": 0.9350545581662464,
+      "grad_norm": 0.09191304445266724,
+      "learning_rate": 1.953240381486395e-06,
+      "loss": 8.5016,
+      "step": 187240
+    },
+    {
+      "epoch": 0.9351044969911858,
+      "grad_norm": 0.09369964152574539,
+      "learning_rate": 1.951738466544845e-06,
+      "loss": 8.5113,
+      "step": 187250
+    },
+    {
+      "epoch": 0.9351544358161252,
+      "grad_norm": 0.08856088668107986,
+      "learning_rate": 1.9502365516032942e-06,
+      "loss": 8.4985,
+      "step": 187260
+    },
+    {
+      "epoch": 0.9352043746410647,
+      "grad_norm": 0.08918076008558273,
+      "learning_rate": 1.9487346366617436e-06,
+      "loss": 8.4963,
+      "step": 187270
+    },
+    {
+      "epoch": 0.9352543134660042,
+      "grad_norm": 0.08736125379800797,
+      "learning_rate": 1.947232721720193e-06,
+      "loss": 8.5068,
+      "step": 187280
+    },
+    {
+      "epoch": 0.9353042522909436,
+      "grad_norm": 0.08733315765857697,
+      "learning_rate": 1.945730806778643e-06,
+      "loss": 8.5016,
+      "step": 187290
+    },
+    {
+      "epoch": 0.935354191115883,
+      "grad_norm": 0.0980258360505104,
+      "learning_rate": 1.9442288918370924e-06,
+      "loss": 8.4994,
+      "step": 187300
+    },
+    {
+      "epoch": 0.9354041299408224,
+      "grad_norm": 0.09128105640411377,
+      "learning_rate": 1.9427269768955418e-06,
+      "loss": 8.508,
+      "step": 187310
+    },
+    {
+      "epoch": 0.935454068765762,
+      "grad_norm": 0.08948434889316559,
+      "learning_rate": 1.941225061953991e-06,
+      "loss": 8.4966,
+      "step": 187320
+    },
+    {
+      "epoch": 0.9355040075907014,
+      "grad_norm": 0.09193699061870575,
+      "learning_rate": 1.939723147012441e-06,
+      "loss": 8.4927,
+      "step": 187330
+    },
+    {
+      "epoch": 0.9355539464156408,
+      "grad_norm": 0.09148573130369186,
+      "learning_rate": 1.9382212320708906e-06,
+      "loss": 8.4928,
+      "step": 187340
+    },
+    {
+      "epoch": 0.9356038852405802,
+      "grad_norm": 0.09217602759599686,
+      "learning_rate": 1.93671931712934e-06,
+      "loss": 8.5084,
+      "step": 187350
+    },
+    {
+      "epoch": 0.9356538240655198,
+      "grad_norm": 0.09518623352050781,
+      "learning_rate": 1.9352174021877893e-06,
+      "loss": 8.5043,
+      "step": 187360
+    },
+    {
+      "epoch": 0.9357037628904592,
+      "grad_norm": 0.08786571770906448,
+      "learning_rate": 1.933715487246239e-06,
+      "loss": 8.5051,
+      "step": 187370
+    },
+    {
+      "epoch": 0.9357537017153986,
+      "grad_norm": 0.0917324647307396,
+      "learning_rate": 1.9322135723046883e-06,
+      "loss": 8.5073,
+      "step": 187380
+    },
+    {
+      "epoch": 0.935803640540338,
+      "grad_norm": 0.08800705522298813,
+      "learning_rate": 1.930711657363138e-06,
+      "loss": 8.5151,
+      "step": 187390
+    },
+    {
+      "epoch": 0.9358535793652776,
+      "grad_norm": 0.09400518983602524,
+      "learning_rate": 1.9292097424215874e-06,
+      "loss": 8.4974,
+      "step": 187400
+    },
+    {
+      "epoch": 0.935903518190217,
+      "grad_norm": 0.08761896938085556,
+      "learning_rate": 1.927707827480037e-06,
+      "loss": 8.507,
+      "step": 187410
+    },
+    {
+      "epoch": 0.9359534570151564,
+      "grad_norm": 0.09526966512203217,
+      "learning_rate": 1.9262059125384865e-06,
+      "loss": 8.494,
+      "step": 187420
+    },
+    {
+      "epoch": 0.9360033958400958,
+      "grad_norm": 0.08607209473848343,
+      "learning_rate": 1.924703997596936e-06,
+      "loss": 8.5211,
+      "step": 187430
+    },
+    {
+      "epoch": 0.9360533346650354,
+      "grad_norm": 0.0938340425491333,
+      "learning_rate": 1.9232020826553856e-06,
+      "loss": 8.5096,
+      "step": 187440
+    },
+    {
+      "epoch": 0.9361032734899748,
+      "grad_norm": 0.0961814746260643,
+      "learning_rate": 1.9217001677138354e-06,
+      "loss": 8.5094,
+      "step": 187450
+    },
+    {
+      "epoch": 0.9361532123149142,
+      "grad_norm": 0.09486842155456543,
+      "learning_rate": 1.9201982527722847e-06,
+      "loss": 8.5038,
+      "step": 187460
+    },
+    {
+      "epoch": 0.9362031511398536,
+      "grad_norm": 0.09420774132013321,
+      "learning_rate": 1.918696337830734e-06,
+      "loss": 8.4886,
+      "step": 187470
+    },
+    {
+      "epoch": 0.9362530899647932,
+      "grad_norm": 0.09382013976573944,
+      "learning_rate": 1.9171944228891838e-06,
+      "loss": 8.4905,
+      "step": 187480
+    },
+    {
+      "epoch": 0.9363030287897326,
+      "grad_norm": 0.08685174584388733,
+      "learning_rate": 1.9156925079476335e-06,
+      "loss": 8.5125,
+      "step": 187490
+    },
+    {
+      "epoch": 0.936352967614672,
+      "grad_norm": 0.09120272099971771,
+      "learning_rate": 1.914190593006083e-06,
+      "loss": 8.5024,
+      "step": 187500
+    },
+    {
+      "epoch": 0.9364029064396114,
+      "grad_norm": 0.09364549070596695,
+      "learning_rate": 1.912688678064532e-06,
+      "loss": 8.5126,
+      "step": 187510
+    },
+    {
+      "epoch": 0.936452845264551,
+      "grad_norm": 0.0939672514796257,
+      "learning_rate": 1.9111867631229815e-06,
+      "loss": 8.4961,
+      "step": 187520
+    },
+    {
+      "epoch": 0.9365027840894904,
+      "grad_norm": 0.08717799186706543,
+      "learning_rate": 1.9096848481814317e-06,
+      "loss": 8.5275,
+      "step": 187530
+    },
+    {
+      "epoch": 0.9365527229144298,
+      "grad_norm": 0.09963074326515198,
+      "learning_rate": 1.908182933239881e-06,
+      "loss": 8.5056,
+      "step": 187540
+    },
+    {
+      "epoch": 0.9366026617393692,
+      "grad_norm": 0.08945607393980026,
+      "learning_rate": 1.9066810182983304e-06,
+      "loss": 8.4934,
+      "step": 187550
+    },
+    {
+      "epoch": 0.9366526005643088,
+      "grad_norm": 0.093963123857975,
+      "learning_rate": 1.9051791033567797e-06,
+      "loss": 8.5153,
+      "step": 187560
+    },
+    {
+      "epoch": 0.9367025393892482,
+      "grad_norm": 0.09400065243244171,
+      "learning_rate": 1.9036771884152297e-06,
+      "loss": 8.4959,
+      "step": 187570
+    },
+    {
+      "epoch": 0.9367524782141876,
+      "grad_norm": 0.09188316017389297,
+      "learning_rate": 1.902175273473679e-06,
+      "loss": 8.5011,
+      "step": 187580
+    },
+    {
+      "epoch": 0.936802417039127,
+      "grad_norm": 0.09157782793045044,
+      "learning_rate": 1.9006733585321283e-06,
+      "loss": 8.4999,
+      "step": 187590
+    },
+    {
+      "epoch": 0.9368523558640666,
+      "grad_norm": 0.08901472389698029,
+      "learning_rate": 1.8991714435905779e-06,
+      "loss": 8.5101,
+      "step": 187600
+    },
+    {
+      "epoch": 0.936902294689006,
+      "grad_norm": 0.08923201262950897,
+      "learning_rate": 1.8976695286490276e-06,
+      "loss": 8.4988,
+      "step": 187610
+    },
+    {
+      "epoch": 0.9369522335139454,
+      "grad_norm": 0.09177745878696442,
+      "learning_rate": 1.8961676137074772e-06,
+      "loss": 8.5018,
+      "step": 187620
+    },
+    {
+      "epoch": 0.9370021723388848,
+      "grad_norm": 0.09261074662208557,
+      "learning_rate": 1.8946656987659265e-06,
+      "loss": 8.5063,
+      "step": 187630
+    },
+    {
+      "epoch": 0.9370521111638244,
+      "grad_norm": 0.09517964720726013,
+      "learning_rate": 1.893163783824376e-06,
+      "loss": 8.4984,
+      "step": 187640
+    },
+    {
+      "epoch": 0.9371020499887638,
+      "grad_norm": 0.09262313693761826,
+      "learning_rate": 1.8916618688828258e-06,
+      "loss": 8.4988,
+      "step": 187650
+    },
+    {
+      "epoch": 0.9371519888137032,
+      "grad_norm": 0.08536347001791,
+      "learning_rate": 1.8901599539412751e-06,
+      "loss": 8.5007,
+      "step": 187660
+    },
+    {
+      "epoch": 0.9372019276386426,
+      "grad_norm": 0.08875380456447601,
+      "learning_rate": 1.8886580389997247e-06,
+      "loss": 8.5065,
+      "step": 187670
+    },
+    {
+      "epoch": 0.9372518664635822,
+      "grad_norm": 0.08886587619781494,
+      "learning_rate": 1.887156124058174e-06,
+      "loss": 8.5045,
+      "step": 187680
+    },
+    {
+      "epoch": 0.9373018052885216,
+      "grad_norm": 0.09413565695285797,
+      "learning_rate": 1.885654209116624e-06,
+      "loss": 8.5019,
+      "step": 187690
+    },
+    {
+      "epoch": 0.937351744113461,
+      "grad_norm": 0.0883907675743103,
+      "learning_rate": 1.8841522941750733e-06,
+      "loss": 8.4836,
+      "step": 187700
+    },
+    {
+      "epoch": 0.9374016829384004,
+      "grad_norm": 0.09106942266225815,
+      "learning_rate": 1.8826503792335226e-06,
+      "loss": 8.5187,
+      "step": 187710
+    },
+    {
+      "epoch": 0.93745162176334,
+      "grad_norm": 0.08953803032636642,
+      "learning_rate": 1.8811484642919722e-06,
+      "loss": 8.505,
+      "step": 187720
+    },
+    {
+      "epoch": 0.9375015605882794,
+      "grad_norm": 0.09178190678358078,
+      "learning_rate": 1.879646549350422e-06,
+      "loss": 8.5085,
+      "step": 187730
+    },
+    {
+      "epoch": 0.9375514994132188,
+      "grad_norm": 0.09529706090688705,
+      "learning_rate": 1.8781446344088715e-06,
+      "loss": 8.502,
+      "step": 187740
+    },
+    {
+      "epoch": 0.9376014382381582,
+      "grad_norm": 0.08537425100803375,
+      "learning_rate": 1.8766427194673208e-06,
+      "loss": 8.4914,
+      "step": 187750
+    },
+    {
+      "epoch": 0.9376513770630978,
+      "grad_norm": 0.0965605229139328,
+      "learning_rate": 1.8751408045257701e-06,
+      "loss": 8.5017,
+      "step": 187760
+    },
+    {
+      "epoch": 0.9377013158880372,
+      "grad_norm": 0.08984053879976273,
+      "learning_rate": 1.8736388895842199e-06,
+      "loss": 8.5113,
+      "step": 187770
+    },
+    {
+      "epoch": 0.9377512547129766,
+      "grad_norm": 0.0879005640745163,
+      "learning_rate": 1.8721369746426694e-06,
+      "loss": 8.5056,
+      "step": 187780
+    },
+    {
+      "epoch": 0.937801193537916,
+      "grad_norm": 0.09019943326711655,
+      "learning_rate": 1.870635059701119e-06,
+      "loss": 8.4996,
+      "step": 187790
+    },
+    {
+      "epoch": 0.9378511323628556,
+      "grad_norm": 0.08988826721906662,
+      "learning_rate": 1.8691331447595685e-06,
+      "loss": 8.5003,
+      "step": 187800
+    },
+    {
+      "epoch": 0.937901071187795,
+      "grad_norm": 0.0893273875117302,
+      "learning_rate": 1.8676312298180178e-06,
+      "loss": 8.5022,
+      "step": 187810
+    },
+    {
+      "epoch": 0.9379510100127344,
+      "grad_norm": 0.09221018105745316,
+      "learning_rate": 1.8661293148764676e-06,
+      "loss": 8.5128,
+      "step": 187820
+    },
+    {
+      "epoch": 0.9380009488376738,
+      "grad_norm": 0.09288901835680008,
+      "learning_rate": 1.864627399934917e-06,
+      "loss": 8.4876,
+      "step": 187830
+    },
+    {
+      "epoch": 0.9380508876626134,
+      "grad_norm": 0.09186063706874847,
+      "learning_rate": 1.8631254849933667e-06,
+      "loss": 8.4982,
+      "step": 187840
+    },
+    {
+      "epoch": 0.9381008264875528,
+      "grad_norm": 0.09561605751514435,
+      "learning_rate": 1.861623570051816e-06,
+      "loss": 8.507,
+      "step": 187850
+    },
+    {
+      "epoch": 0.9381507653124922,
+      "grad_norm": 0.09141556173563004,
+      "learning_rate": 1.8601216551102656e-06,
+      "loss": 8.5184,
+      "step": 187860
+    },
+    {
+      "epoch": 0.9382007041374316,
+      "grad_norm": 0.08741176128387451,
+      "learning_rate": 1.858619740168715e-06,
+      "loss": 8.5107,
+      "step": 187870
+    },
+    {
+      "epoch": 0.9382506429623712,
+      "grad_norm": 0.0920015349984169,
+      "learning_rate": 1.8571178252271646e-06,
+      "loss": 8.5134,
+      "step": 187880
+    },
+    {
+      "epoch": 0.9383005817873106,
+      "grad_norm": 0.09082834422588348,
+      "learning_rate": 1.8556159102856142e-06,
+      "loss": 8.5058,
+      "step": 187890
+    },
+    {
+      "epoch": 0.93835052061225,
+      "grad_norm": 0.093767449259758,
+      "learning_rate": 1.8541139953440637e-06,
+      "loss": 8.5058,
+      "step": 187900
+    },
+    {
+      "epoch": 0.9384004594371894,
+      "grad_norm": 0.08792780339717865,
+      "learning_rate": 1.8526120804025133e-06,
+      "loss": 8.5191,
+      "step": 187910
+    },
+    {
+      "epoch": 0.9384503982621288,
+      "grad_norm": 0.08794878423213959,
+      "learning_rate": 1.8511101654609628e-06,
+      "loss": 8.5125,
+      "step": 187920
+    },
+    {
+      "epoch": 0.9385003370870684,
+      "grad_norm": 0.08979849517345428,
+      "learning_rate": 1.8496082505194122e-06,
+      "loss": 8.5039,
+      "step": 187930
+    },
+    {
+      "epoch": 0.9385502759120078,
+      "grad_norm": 0.08824494481086731,
+      "learning_rate": 1.848106335577862e-06,
+      "loss": 8.4959,
+      "step": 187940
+    },
+    {
+      "epoch": 0.9386002147369472,
+      "grad_norm": 0.08635149151086807,
+      "learning_rate": 1.8466044206363112e-06,
+      "loss": 8.4952,
+      "step": 187950
+    },
+    {
+      "epoch": 0.9386501535618866,
+      "grad_norm": 0.09149076044559479,
+      "learning_rate": 1.845102505694761e-06,
+      "loss": 8.4943,
+      "step": 187960
+    },
+    {
+      "epoch": 0.9387000923868262,
+      "grad_norm": 0.0918533205986023,
+      "learning_rate": 1.8436005907532103e-06,
+      "loss": 8.5059,
+      "step": 187970
+    },
+    {
+      "epoch": 0.9387500312117656,
+      "grad_norm": 0.09328784048557281,
+      "learning_rate": 1.8420986758116599e-06,
+      "loss": 8.494,
+      "step": 187980
+    },
+    {
+      "epoch": 0.938799970036705,
+      "grad_norm": 0.087800532579422,
+      "learning_rate": 1.8405967608701094e-06,
+      "loss": 8.5055,
+      "step": 187990
+    },
+    {
+      "epoch": 0.9388499088616444,
+      "grad_norm": 0.0951266810297966,
+      "learning_rate": 1.839094845928559e-06,
+      "loss": 8.4924,
+      "step": 188000
+    },
+    {
+      "epoch": 0.938899847686584,
+      "grad_norm": 0.0940432995557785,
+      "learning_rate": 1.8375929309870085e-06,
+      "loss": 8.4983,
+      "step": 188010
+    },
+    {
+      "epoch": 0.9389497865115234,
+      "grad_norm": 0.08336132764816284,
+      "learning_rate": 1.836091016045458e-06,
+      "loss": 8.5174,
+      "step": 188020
+    },
+    {
+      "epoch": 0.9389997253364628,
+      "grad_norm": 0.09214475005865097,
+      "learning_rate": 1.8345891011039074e-06,
+      "loss": 8.5017,
+      "step": 188030
+    },
+    {
+      "epoch": 0.9390496641614022,
+      "grad_norm": 0.0941125676035881,
+      "learning_rate": 1.8330871861623571e-06,
+      "loss": 8.5126,
+      "step": 188040
+    },
+    {
+      "epoch": 0.9390996029863418,
+      "grad_norm": 0.08968133479356766,
+      "learning_rate": 1.8315852712208065e-06,
+      "loss": 8.5143,
+      "step": 188050
+    },
+    {
+      "epoch": 0.9391495418112812,
+      "grad_norm": 0.08894607424736023,
+      "learning_rate": 1.8300833562792562e-06,
+      "loss": 8.5021,
+      "step": 188060
+    },
+    {
+      "epoch": 0.9391994806362206,
+      "grad_norm": 0.09236035495996475,
+      "learning_rate": 1.8285814413377055e-06,
+      "loss": 8.4871,
+      "step": 188070
+    },
+    {
+      "epoch": 0.93924941946116,
+      "grad_norm": 0.09041870385408401,
+      "learning_rate": 1.827079526396155e-06,
+      "loss": 8.4946,
+      "step": 188080
+    },
+    {
+      "epoch": 0.9392993582860996,
+      "grad_norm": 0.08861184120178223,
+      "learning_rate": 1.8255776114546046e-06,
+      "loss": 8.485,
+      "step": 188090
+    },
+    {
+      "epoch": 0.939349297111039,
+      "grad_norm": 0.09006591141223907,
+      "learning_rate": 1.8240756965130542e-06,
+      "loss": 8.5131,
+      "step": 188100
+    },
+    {
+      "epoch": 0.9393992359359784,
+      "grad_norm": 0.09157989919185638,
+      "learning_rate": 1.8225737815715037e-06,
+      "loss": 8.5131,
+      "step": 188110
+    },
+    {
+      "epoch": 0.9394491747609178,
+      "grad_norm": 0.09308748692274094,
+      "learning_rate": 1.8210718666299533e-06,
+      "loss": 8.4979,
+      "step": 188120
+    },
+    {
+      "epoch": 0.9394991135858574,
+      "grad_norm": 0.10365467518568039,
+      "learning_rate": 1.8195699516884026e-06,
+      "loss": 8.4922,
+      "step": 188130
+    },
+    {
+      "epoch": 0.9395490524107968,
+      "grad_norm": 0.08948034048080444,
+      "learning_rate": 1.8180680367468523e-06,
+      "loss": 8.4991,
+      "step": 188140
+    },
+    {
+      "epoch": 0.9395989912357362,
+      "grad_norm": 0.08552392572164536,
+      "learning_rate": 1.8165661218053017e-06,
+      "loss": 8.525,
+      "step": 188150
+    },
+    {
+      "epoch": 0.9396489300606756,
+      "grad_norm": 0.09680628776550293,
+      "learning_rate": 1.8150642068637514e-06,
+      "loss": 8.5061,
+      "step": 188160
+    },
+    {
+      "epoch": 0.9396988688856152,
+      "grad_norm": 0.08866409212350845,
+      "learning_rate": 1.8135622919222008e-06,
+      "loss": 8.5054,
+      "step": 188170
+    },
+    {
+      "epoch": 0.9397488077105546,
+      "grad_norm": 0.09396608918905258,
+      "learning_rate": 1.8120603769806505e-06,
+      "loss": 8.515,
+      "step": 188180
+    },
+    {
+      "epoch": 0.939798746535494,
+      "grad_norm": 0.08692758530378342,
+      "learning_rate": 1.8105584620390998e-06,
+      "loss": 8.5197,
+      "step": 188190
+    },
+    {
+      "epoch": 0.9398486853604334,
+      "grad_norm": 0.0874529555439949,
+      "learning_rate": 1.8090565470975494e-06,
+      "loss": 8.509,
+      "step": 188200
+    },
+    {
+      "epoch": 0.939898624185373,
+      "grad_norm": 0.08824202418327332,
+      "learning_rate": 1.807554632155999e-06,
+      "loss": 8.5272,
+      "step": 188210
+    },
+    {
+      "epoch": 0.9399485630103124,
+      "grad_norm": 0.0905284509062767,
+      "learning_rate": 1.8060527172144485e-06,
+      "loss": 8.5125,
+      "step": 188220
+    },
+    {
+      "epoch": 0.9399985018352518,
+      "grad_norm": 0.0894288644194603,
+      "learning_rate": 1.804550802272898e-06,
+      "loss": 8.4988,
+      "step": 188230
+    },
+    {
+      "epoch": 0.9400484406601912,
+      "grad_norm": 0.09066865593194962,
+      "learning_rate": 1.8030488873313476e-06,
+      "loss": 8.5031,
+      "step": 188240
+    },
+    {
+      "epoch": 0.9400983794851308,
+      "grad_norm": 0.08707483112812042,
+      "learning_rate": 1.801546972389797e-06,
+      "loss": 8.49,
+      "step": 188250
+    },
+    {
+      "epoch": 0.9401483183100702,
+      "grad_norm": 0.08790065348148346,
+      "learning_rate": 1.8000450574482467e-06,
+      "loss": 8.4957,
+      "step": 188260
+    },
+    {
+      "epoch": 0.9401982571350096,
+      "grad_norm": 0.09469541907310486,
+      "learning_rate": 1.798543142506696e-06,
+      "loss": 8.5037,
+      "step": 188270
+    },
+    {
+      "epoch": 0.940248195959949,
+      "grad_norm": 0.0937683954834938,
+      "learning_rate": 1.7970412275651457e-06,
+      "loss": 8.5359,
+      "step": 188280
+    },
+    {
+      "epoch": 0.9402981347848886,
+      "grad_norm": 0.08924166113138199,
+      "learning_rate": 1.795539312623595e-06,
+      "loss": 8.5029,
+      "step": 188290
+    },
+    {
+      "epoch": 0.940348073609828,
+      "grad_norm": 0.08660190552473068,
+      "learning_rate": 1.7940373976820446e-06,
+      "loss": 8.5128,
+      "step": 188300
+    },
+    {
+      "epoch": 0.9403980124347674,
+      "grad_norm": 0.0949644148349762,
+      "learning_rate": 1.7925354827404942e-06,
+      "loss": 8.5099,
+      "step": 188310
+    },
+    {
+      "epoch": 0.9404479512597068,
+      "grad_norm": 0.08653347194194794,
+      "learning_rate": 1.7910335677989437e-06,
+      "loss": 8.4979,
+      "step": 188320
+    },
+    {
+      "epoch": 0.9404978900846463,
+      "grad_norm": 0.09062409400939941,
+      "learning_rate": 1.7895316528573932e-06,
+      "loss": 8.5025,
+      "step": 188330
+    },
+    {
+      "epoch": 0.9405478289095858,
+      "grad_norm": 0.09127449989318848,
+      "learning_rate": 1.7880297379158428e-06,
+      "loss": 8.4861,
+      "step": 188340
+    },
+    {
+      "epoch": 0.9405977677345252,
+      "grad_norm": 0.09160736948251724,
+      "learning_rate": 1.7865278229742921e-06,
+      "loss": 8.498,
+      "step": 188350
+    },
+    {
+      "epoch": 0.9406477065594646,
+      "grad_norm": 0.08944227546453476,
+      "learning_rate": 1.7850259080327419e-06,
+      "loss": 8.4942,
+      "step": 188360
+    },
+    {
+      "epoch": 0.9406976453844041,
+      "grad_norm": 0.08706548810005188,
+      "learning_rate": 1.7835239930911912e-06,
+      "loss": 8.5159,
+      "step": 188370
+    },
+    {
+      "epoch": 0.9407475842093436,
+      "grad_norm": 0.08913147449493408,
+      "learning_rate": 1.782022078149641e-06,
+      "loss": 8.4955,
+      "step": 188380
+    },
+    {
+      "epoch": 0.940797523034283,
+      "grad_norm": 0.09037254750728607,
+      "learning_rate": 1.7805201632080903e-06,
+      "loss": 8.5251,
+      "step": 188390
+    },
+    {
+      "epoch": 0.9408474618592224,
+      "grad_norm": 0.09032527357339859,
+      "learning_rate": 1.7790182482665398e-06,
+      "loss": 8.522,
+      "step": 188400
+    },
+    {
+      "epoch": 0.9408974006841619,
+      "grad_norm": 0.08700371533632278,
+      "learning_rate": 1.7775163333249894e-06,
+      "loss": 8.4951,
+      "step": 188410
+    },
+    {
+      "epoch": 0.9409473395091014,
+      "grad_norm": 0.09280155599117279,
+      "learning_rate": 1.776014418383439e-06,
+      "loss": 8.5057,
+      "step": 188420
+    },
+    {
+      "epoch": 0.9409972783340408,
+      "grad_norm": 0.09698103368282318,
+      "learning_rate": 1.7745125034418885e-06,
+      "loss": 8.4896,
+      "step": 188430
+    },
+    {
+      "epoch": 0.9410472171589802,
+      "grad_norm": 0.09256529062986374,
+      "learning_rate": 1.773010588500338e-06,
+      "loss": 8.5083,
+      "step": 188440
+    },
+    {
+      "epoch": 0.9410971559839197,
+      "grad_norm": 0.0863165482878685,
+      "learning_rate": 1.7715086735587875e-06,
+      "loss": 8.5024,
+      "step": 188450
+    },
+    {
+      "epoch": 0.9411470948088592,
+      "grad_norm": 0.08784864097833633,
+      "learning_rate": 1.770006758617237e-06,
+      "loss": 8.5086,
+      "step": 188460
+    },
+    {
+      "epoch": 0.9411970336337986,
+      "grad_norm": 0.09332052618265152,
+      "learning_rate": 1.7685048436756864e-06,
+      "loss": 8.5082,
+      "step": 188470
+    },
+    {
+      "epoch": 0.941246972458738,
+      "grad_norm": 0.08981883525848389,
+      "learning_rate": 1.7670029287341362e-06,
+      "loss": 8.5098,
+      "step": 188480
+    },
+    {
+      "epoch": 0.9412969112836775,
+      "grad_norm": 0.08930005878210068,
+      "learning_rate": 1.7655010137925855e-06,
+      "loss": 8.5293,
+      "step": 188490
+    },
+    {
+      "epoch": 0.941346850108617,
+      "grad_norm": 0.08967043459415436,
+      "learning_rate": 1.7639990988510353e-06,
+      "loss": 8.5243,
+      "step": 188500
+    },
+    {
+      "epoch": 0.9413967889335564,
+      "grad_norm": 0.09381302446126938,
+      "learning_rate": 1.7624971839094846e-06,
+      "loss": 8.4928,
+      "step": 188510
+    },
+    {
+      "epoch": 0.9414467277584958,
+      "grad_norm": 0.0840909332036972,
+      "learning_rate": 1.7609952689679341e-06,
+      "loss": 8.5254,
+      "step": 188520
+    },
+    {
+      "epoch": 0.9414966665834353,
+      "grad_norm": 0.09190535545349121,
+      "learning_rate": 1.7594933540263837e-06,
+      "loss": 8.5259,
+      "step": 188530
+    },
+    {
+      "epoch": 0.9415466054083748,
+      "grad_norm": 0.0981719046831131,
+      "learning_rate": 1.7579914390848332e-06,
+      "loss": 8.5103,
+      "step": 188540
+    },
+    {
+      "epoch": 0.9415965442333142,
+      "grad_norm": 0.08787670731544495,
+      "learning_rate": 1.7564895241432828e-06,
+      "loss": 8.5087,
+      "step": 188550
+    },
+    {
+      "epoch": 0.9416464830582536,
+      "grad_norm": 0.09270340204238892,
+      "learning_rate": 1.7549876092017323e-06,
+      "loss": 8.5074,
+      "step": 188560
+    },
+    {
+      "epoch": 0.9416964218831931,
+      "grad_norm": 0.09285027533769608,
+      "learning_rate": 1.7534856942601816e-06,
+      "loss": 8.5098,
+      "step": 188570
+    },
+    {
+      "epoch": 0.9417463607081326,
+      "grad_norm": 0.0908166691660881,
+      "learning_rate": 1.7519837793186314e-06,
+      "loss": 8.4977,
+      "step": 188580
+    },
+    {
+      "epoch": 0.941796299533072,
+      "grad_norm": 0.09321475028991699,
+      "learning_rate": 1.7504818643770807e-06,
+      "loss": 8.5135,
+      "step": 188590
+    },
+    {
+      "epoch": 0.9418462383580114,
+      "grad_norm": 0.08778084069490433,
+      "learning_rate": 1.7489799494355305e-06,
+      "loss": 8.5213,
+      "step": 188600
+    },
+    {
+      "epoch": 0.9418961771829509,
+      "grad_norm": 0.09464406967163086,
+      "learning_rate": 1.7474780344939798e-06,
+      "loss": 8.5029,
+      "step": 188610
+    },
+    {
+      "epoch": 0.9419461160078904,
+      "grad_norm": 0.09429196268320084,
+      "learning_rate": 1.7459761195524294e-06,
+      "loss": 8.4987,
+      "step": 188620
+    },
+    {
+      "epoch": 0.9419960548328298,
+      "grad_norm": 0.09396494925022125,
+      "learning_rate": 1.744474204610879e-06,
+      "loss": 8.5222,
+      "step": 188630
+    },
+    {
+      "epoch": 0.9420459936577692,
+      "grad_norm": 0.08751998841762543,
+      "learning_rate": 1.7429722896693284e-06,
+      "loss": 8.4981,
+      "step": 188640
+    },
+    {
+      "epoch": 0.9420959324827087,
+      "grad_norm": 0.08901175856590271,
+      "learning_rate": 1.741470374727778e-06,
+      "loss": 8.5163,
+      "step": 188650
+    },
+    {
+      "epoch": 0.9421458713076482,
+      "grad_norm": 0.08800294995307922,
+      "learning_rate": 1.7399684597862275e-06,
+      "loss": 8.5103,
+      "step": 188660
+    },
+    {
+      "epoch": 0.9421958101325876,
+      "grad_norm": 0.09194714576005936,
+      "learning_rate": 1.7384665448446769e-06,
+      "loss": 8.4861,
+      "step": 188670
+    },
+    {
+      "epoch": 0.942245748957527,
+      "grad_norm": 0.10198995471000671,
+      "learning_rate": 1.7369646299031266e-06,
+      "loss": 8.511,
+      "step": 188680
+    },
+    {
+      "epoch": 0.9422956877824665,
+      "grad_norm": 0.09428546577692032,
+      "learning_rate": 1.735462714961576e-06,
+      "loss": 8.5202,
+      "step": 188690
+    },
+    {
+      "epoch": 0.942345626607406,
+      "grad_norm": 0.09078855067491531,
+      "learning_rate": 1.7339608000200257e-06,
+      "loss": 8.5044,
+      "step": 188700
+    },
+    {
+      "epoch": 0.9423955654323454,
+      "grad_norm": 0.09226953983306885,
+      "learning_rate": 1.732458885078475e-06,
+      "loss": 8.5043,
+      "step": 188710
+    },
+    {
+      "epoch": 0.9424455042572848,
+      "grad_norm": 0.09067227691411972,
+      "learning_rate": 1.7309569701369248e-06,
+      "loss": 8.5106,
+      "step": 188720
+    },
+    {
+      "epoch": 0.9424954430822243,
+      "grad_norm": 0.09468891471624374,
+      "learning_rate": 1.7294550551953741e-06,
+      "loss": 8.4964,
+      "step": 188730
+    },
+    {
+      "epoch": 0.9425453819071637,
+      "grad_norm": 0.0890989601612091,
+      "learning_rate": 1.7279531402538237e-06,
+      "loss": 8.5198,
+      "step": 188740
+    },
+    {
+      "epoch": 0.9425953207321032,
+      "grad_norm": 0.09626948088407516,
+      "learning_rate": 1.7264512253122732e-06,
+      "loss": 8.5006,
+      "step": 188750
+    },
+    {
+      "epoch": 0.9426452595570426,
+      "grad_norm": 0.08856556564569473,
+      "learning_rate": 1.7249493103707227e-06,
+      "loss": 8.487,
+      "step": 188760
+    },
+    {
+      "epoch": 0.9426951983819821,
+      "grad_norm": 0.09216383099555969,
+      "learning_rate": 1.7234473954291723e-06,
+      "loss": 8.5164,
+      "step": 188770
+    },
+    {
+      "epoch": 0.9427451372069215,
+      "grad_norm": 0.08988320082426071,
+      "learning_rate": 1.7219454804876218e-06,
+      "loss": 8.5159,
+      "step": 188780
+    },
+    {
+      "epoch": 0.942795076031861,
+      "grad_norm": 0.09181145578622818,
+      "learning_rate": 1.7204435655460712e-06,
+      "loss": 8.4966,
+      "step": 188790
+    },
+    {
+      "epoch": 0.9428450148568004,
+      "grad_norm": 0.08662037551403046,
+      "learning_rate": 1.718941650604521e-06,
+      "loss": 8.5168,
+      "step": 188800
+    },
+    {
+      "epoch": 0.9428949536817399,
+      "grad_norm": 0.09044136106967926,
+      "learning_rate": 1.7174397356629702e-06,
+      "loss": 8.5264,
+      "step": 188810
+    },
+    {
+      "epoch": 0.9429448925066793,
+      "grad_norm": 0.09236202389001846,
+      "learning_rate": 1.71593782072142e-06,
+      "loss": 8.4977,
+      "step": 188820
+    },
+    {
+      "epoch": 0.9429948313316188,
+      "grad_norm": 0.08900753408670425,
+      "learning_rate": 1.7144359057798693e-06,
+      "loss": 8.5269,
+      "step": 188830
+    },
+    {
+      "epoch": 0.9430447701565582,
+      "grad_norm": 0.09463819116353989,
+      "learning_rate": 1.7129339908383189e-06,
+      "loss": 8.5008,
+      "step": 188840
+    },
+    {
+      "epoch": 0.9430947089814977,
+      "grad_norm": 0.09038529545068741,
+      "learning_rate": 1.7114320758967684e-06,
+      "loss": 8.5102,
+      "step": 188850
+    },
+    {
+      "epoch": 0.9431446478064371,
+      "grad_norm": 0.0910005047917366,
+      "learning_rate": 1.709930160955218e-06,
+      "loss": 8.4995,
+      "step": 188860
+    },
+    {
+      "epoch": 0.9431945866313766,
+      "grad_norm": 0.09129851311445236,
+      "learning_rate": 1.7084282460136675e-06,
+      "loss": 8.5172,
+      "step": 188870
+    },
+    {
+      "epoch": 0.943244525456316,
+      "grad_norm": 0.09213831275701523,
+      "learning_rate": 1.706926331072117e-06,
+      "loss": 8.5083,
+      "step": 188880
+    },
+    {
+      "epoch": 0.9432944642812554,
+      "grad_norm": 0.09839709848165512,
+      "learning_rate": 1.7054244161305664e-06,
+      "loss": 8.4996,
+      "step": 188890
+    },
+    {
+      "epoch": 0.9433444031061949,
+      "grad_norm": 0.09194820374250412,
+      "learning_rate": 1.7039225011890161e-06,
+      "loss": 8.493,
+      "step": 188900
+    },
+    {
+      "epoch": 0.9433943419311344,
+      "grad_norm": 0.0895412266254425,
+      "learning_rate": 1.7024205862474655e-06,
+      "loss": 8.5171,
+      "step": 188910
+    },
+    {
+      "epoch": 0.9434442807560738,
+      "grad_norm": 0.08914782851934433,
+      "learning_rate": 1.7009186713059152e-06,
+      "loss": 8.4905,
+      "step": 188920
+    },
+    {
+      "epoch": 0.9434942195810132,
+      "grad_norm": 0.09766605496406555,
+      "learning_rate": 1.6994167563643646e-06,
+      "loss": 8.5128,
+      "step": 188930
+    },
+    {
+      "epoch": 0.9435441584059527,
+      "grad_norm": 0.0920957624912262,
+      "learning_rate": 1.697914841422814e-06,
+      "loss": 8.4913,
+      "step": 188940
+    },
+    {
+      "epoch": 0.9435940972308922,
+      "grad_norm": 0.08751115202903748,
+      "learning_rate": 1.6964129264812636e-06,
+      "loss": 8.5104,
+      "step": 188950
+    },
+    {
+      "epoch": 0.9436440360558316,
+      "grad_norm": 0.09014736860990524,
+      "learning_rate": 1.6949110115397132e-06,
+      "loss": 8.5043,
+      "step": 188960
+    },
+    {
+      "epoch": 0.943693974880771,
+      "grad_norm": 0.08869920670986176,
+      "learning_rate": 1.6934090965981627e-06,
+      "loss": 8.4973,
+      "step": 188970
+    },
+    {
+      "epoch": 0.9437439137057105,
+      "grad_norm": 0.088557668030262,
+      "learning_rate": 1.6919071816566123e-06,
+      "loss": 8.5082,
+      "step": 188980
+    },
+    {
+      "epoch": 0.94379385253065,
+      "grad_norm": 0.08449137955904007,
+      "learning_rate": 1.6904052667150618e-06,
+      "loss": 8.5086,
+      "step": 188990
+    },
+    {
+      "epoch": 0.9438437913555894,
+      "grad_norm": 0.08829719573259354,
+      "learning_rate": 1.6889033517735114e-06,
+      "loss": 8.5273,
+      "step": 189000
+    },
+    {
+      "epoch": 0.9438937301805288,
+      "grad_norm": 0.09361531585454941,
+      "learning_rate": 1.6874014368319607e-06,
+      "loss": 8.4817,
+      "step": 189010
+    },
+    {
+      "epoch": 0.9439436690054683,
+      "grad_norm": 0.09768420457839966,
+      "learning_rate": 1.6858995218904104e-06,
+      "loss": 8.5082,
+      "step": 189020
+    },
+    {
+      "epoch": 0.9439936078304078,
+      "grad_norm": 0.0924157053232193,
+      "learning_rate": 1.6843976069488598e-06,
+      "loss": 8.5025,
+      "step": 189030
+    },
+    {
+      "epoch": 0.9440435466553472,
+      "grad_norm": 0.09079400449991226,
+      "learning_rate": 1.6828956920073095e-06,
+      "loss": 8.5017,
+      "step": 189040
+    },
+    {
+      "epoch": 0.9440934854802866,
+      "grad_norm": 0.09591791033744812,
+      "learning_rate": 1.6813937770657589e-06,
+      "loss": 8.5061,
+      "step": 189050
+    },
+    {
+      "epoch": 0.9441434243052261,
+      "grad_norm": 0.0918201357126236,
+      "learning_rate": 1.6798918621242084e-06,
+      "loss": 8.515,
+      "step": 189060
+    },
+    {
+      "epoch": 0.9441933631301656,
+      "grad_norm": 0.09003403782844543,
+      "learning_rate": 1.678389947182658e-06,
+      "loss": 8.5041,
+      "step": 189070
+    },
+    {
+      "epoch": 0.944243301955105,
+      "grad_norm": 0.09714401513338089,
+      "learning_rate": 1.6768880322411075e-06,
+      "loss": 8.4791,
+      "step": 189080
+    },
+    {
+      "epoch": 0.9442932407800444,
+      "grad_norm": 0.08682556450366974,
+      "learning_rate": 1.675386117299557e-06,
+      "loss": 8.5136,
+      "step": 189090
+    },
+    {
+      "epoch": 0.9443431796049839,
+      "grad_norm": 0.08617331087589264,
+      "learning_rate": 1.6738842023580066e-06,
+      "loss": 8.5153,
+      "step": 189100
+    },
+    {
+      "epoch": 0.9443931184299234,
+      "grad_norm": 0.09631218761205673,
+      "learning_rate": 1.672382287416456e-06,
+      "loss": 8.4938,
+      "step": 189110
+    },
+    {
+      "epoch": 0.9444430572548628,
+      "grad_norm": 0.09221839159727097,
+      "learning_rate": 1.6708803724749057e-06,
+      "loss": 8.5093,
+      "step": 189120
+    },
+    {
+      "epoch": 0.9444929960798022,
+      "grad_norm": 0.08441857248544693,
+      "learning_rate": 1.669378457533355e-06,
+      "loss": 8.5082,
+      "step": 189130
+    },
+    {
+      "epoch": 0.9445429349047417,
+      "grad_norm": 0.0894455835223198,
+      "learning_rate": 1.6678765425918047e-06,
+      "loss": 8.4961,
+      "step": 189140
+    },
+    {
+      "epoch": 0.9445928737296811,
+      "grad_norm": 0.08852816373109818,
+      "learning_rate": 1.666374627650254e-06,
+      "loss": 8.4978,
+      "step": 189150
+    },
+    {
+      "epoch": 0.9446428125546206,
+      "grad_norm": 0.09122937172651291,
+      "learning_rate": 1.6648727127087036e-06,
+      "loss": 8.5011,
+      "step": 189160
+    },
+    {
+      "epoch": 0.94469275137956,
+      "grad_norm": 0.09438908100128174,
+      "learning_rate": 1.6633707977671532e-06,
+      "loss": 8.5027,
+      "step": 189170
+    },
+    {
+      "epoch": 0.9447426902044995,
+      "grad_norm": 0.08923160284757614,
+      "learning_rate": 1.6618688828256027e-06,
+      "loss": 8.4992,
+      "step": 189180
+    },
+    {
+      "epoch": 0.944792629029439,
+      "grad_norm": 0.08867133408784866,
+      "learning_rate": 1.6603669678840523e-06,
+      "loss": 8.4922,
+      "step": 189190
+    },
+    {
+      "epoch": 0.9448425678543784,
+      "grad_norm": 0.0948406532406807,
+      "learning_rate": 1.6588650529425018e-06,
+      "loss": 8.5108,
+      "step": 189200
+    },
+    {
+      "epoch": 0.9448925066793178,
+      "grad_norm": 0.08932336419820786,
+      "learning_rate": 1.6573631380009511e-06,
+      "loss": 8.5091,
+      "step": 189210
+    },
+    {
+      "epoch": 0.9449424455042573,
+      "grad_norm": 0.09022360295057297,
+      "learning_rate": 1.6558612230594009e-06,
+      "loss": 8.5191,
+      "step": 189220
+    },
+    {
+      "epoch": 0.9449923843291967,
+      "grad_norm": 0.09157562255859375,
+      "learning_rate": 1.6543593081178502e-06,
+      "loss": 8.5045,
+      "step": 189230
+    },
+    {
+      "epoch": 0.9450423231541362,
+      "grad_norm": 0.09014785289764404,
+      "learning_rate": 1.6528573931763e-06,
+      "loss": 8.5158,
+      "step": 189240
+    },
+    {
+      "epoch": 0.9450922619790756,
+      "grad_norm": 0.09535112977027893,
+      "learning_rate": 1.6513554782347493e-06,
+      "loss": 8.5037,
+      "step": 189250
+    },
+    {
+      "epoch": 0.9451422008040151,
+      "grad_norm": 0.09707490354776382,
+      "learning_rate": 1.6498535632931988e-06,
+      "loss": 8.5056,
+      "step": 189260
+    },
+    {
+      "epoch": 0.9451921396289545,
+      "grad_norm": 0.09138431400060654,
+      "learning_rate": 1.6483516483516484e-06,
+      "loss": 8.4955,
+      "step": 189270
+    },
+    {
+      "epoch": 0.945242078453894,
+      "grad_norm": 0.0848042219877243,
+      "learning_rate": 1.646849733410098e-06,
+      "loss": 8.5018,
+      "step": 189280
+    },
+    {
+      "epoch": 0.9452920172788334,
+      "grad_norm": 0.08952652662992477,
+      "learning_rate": 1.6453478184685475e-06,
+      "loss": 8.5142,
+      "step": 189290
+    },
+    {
+      "epoch": 0.9453419561037729,
+      "grad_norm": 0.09085434675216675,
+      "learning_rate": 1.643845903526997e-06,
+      "loss": 8.5238,
+      "step": 189300
+    },
+    {
+      "epoch": 0.9453918949287123,
+      "grad_norm": 0.08714916557073593,
+      "learning_rate": 1.6423439885854466e-06,
+      "loss": 8.531,
+      "step": 189310
+    },
+    {
+      "epoch": 0.9454418337536518,
+      "grad_norm": 0.09484901279211044,
+      "learning_rate": 1.640842073643896e-06,
+      "loss": 8.4926,
+      "step": 189320
+    },
+    {
+      "epoch": 0.9454917725785912,
+      "grad_norm": 0.08953174948692322,
+      "learning_rate": 1.6393401587023454e-06,
+      "loss": 8.5142,
+      "step": 189330
+    },
+    {
+      "epoch": 0.9455417114035307,
+      "grad_norm": 0.09061595797538757,
+      "learning_rate": 1.6378382437607952e-06,
+      "loss": 8.4869,
+      "step": 189340
+    },
+    {
+      "epoch": 0.9455916502284701,
+      "grad_norm": 0.08741827309131622,
+      "learning_rate": 1.6363363288192445e-06,
+      "loss": 8.506,
+      "step": 189350
+    },
+    {
+      "epoch": 0.9456415890534096,
+      "grad_norm": 0.0936361625790596,
+      "learning_rate": 1.6348344138776943e-06,
+      "loss": 8.4956,
+      "step": 189360
+    },
+    {
+      "epoch": 0.945691527878349,
+      "grad_norm": 0.0899646058678627,
+      "learning_rate": 1.6333324989361436e-06,
+      "loss": 8.5156,
+      "step": 189370
+    },
+    {
+      "epoch": 0.9457414667032885,
+      "grad_norm": 0.09246868640184402,
+      "learning_rate": 1.6318305839945931e-06,
+      "loss": 8.5201,
+      "step": 189380
+    },
+    {
+      "epoch": 0.9457914055282279,
+      "grad_norm": 0.08746492117643356,
+      "learning_rate": 1.6303286690530427e-06,
+      "loss": 8.509,
+      "step": 189390
+    },
+    {
+      "epoch": 0.9458413443531674,
+      "grad_norm": 0.09183592349290848,
+      "learning_rate": 1.6288267541114922e-06,
+      "loss": 8.5058,
+      "step": 189400
+    },
+    {
+      "epoch": 0.9458912831781068,
+      "grad_norm": 0.088591568171978,
+      "learning_rate": 1.6273248391699418e-06,
+      "loss": 8.5103,
+      "step": 189410
+    },
+    {
+      "epoch": 0.9459412220030463,
+      "grad_norm": 0.09180916845798492,
+      "learning_rate": 1.6258229242283913e-06,
+      "loss": 8.505,
+      "step": 189420
+    },
+    {
+      "epoch": 0.9459911608279857,
+      "grad_norm": 0.0878295823931694,
+      "learning_rate": 1.6243210092868406e-06,
+      "loss": 8.4924,
+      "step": 189430
+    },
+    {
+      "epoch": 0.9460410996529252,
+      "grad_norm": 0.09343598783016205,
+      "learning_rate": 1.6228190943452904e-06,
+      "loss": 8.4996,
+      "step": 189440
+    },
+    {
+      "epoch": 0.9460910384778646,
+      "grad_norm": 0.08655836433172226,
+      "learning_rate": 1.6213171794037397e-06,
+      "loss": 8.4907,
+      "step": 189450
+    },
+    {
+      "epoch": 0.9461409773028041,
+      "grad_norm": 0.09079290181398392,
+      "learning_rate": 1.6198152644621895e-06,
+      "loss": 8.5051,
+      "step": 189460
+    },
+    {
+      "epoch": 0.9461909161277435,
+      "grad_norm": 0.09278516471385956,
+      "learning_rate": 1.6183133495206388e-06,
+      "loss": 8.5262,
+      "step": 189470
+    },
+    {
+      "epoch": 0.946240854952683,
+      "grad_norm": 0.0923718810081482,
+      "learning_rate": 1.6168114345790884e-06,
+      "loss": 8.4926,
+      "step": 189480
+    },
+    {
+      "epoch": 0.9462907937776224,
+      "grad_norm": 0.0890062153339386,
+      "learning_rate": 1.615309519637538e-06,
+      "loss": 8.5063,
+      "step": 189490
+    },
+    {
+      "epoch": 0.9463407326025619,
+      "grad_norm": 0.0915982574224472,
+      "learning_rate": 1.6138076046959875e-06,
+      "loss": 8.5034,
+      "step": 189500
+    },
+    {
+      "epoch": 0.9463906714275013,
+      "grad_norm": 0.0939352884888649,
+      "learning_rate": 1.612305689754437e-06,
+      "loss": 8.4948,
+      "step": 189510
+    },
+    {
+      "epoch": 0.9464406102524408,
+      "grad_norm": 0.09317021816968918,
+      "learning_rate": 1.6108037748128865e-06,
+      "loss": 8.5007,
+      "step": 189520
+    },
+    {
+      "epoch": 0.9464905490773802,
+      "grad_norm": 0.08831048011779785,
+      "learning_rate": 1.609301859871336e-06,
+      "loss": 8.4932,
+      "step": 189530
+    },
+    {
+      "epoch": 0.9465404879023197,
+      "grad_norm": 0.1017037183046341,
+      "learning_rate": 1.6077999449297856e-06,
+      "loss": 8.5018,
+      "step": 189540
+    },
+    {
+      "epoch": 0.9465904267272591,
+      "grad_norm": 0.09273435175418854,
+      "learning_rate": 1.606298029988235e-06,
+      "loss": 8.4929,
+      "step": 189550
+    },
+    {
+      "epoch": 0.9466403655521985,
+      "grad_norm": 0.09123826771974564,
+      "learning_rate": 1.6047961150466847e-06,
+      "loss": 8.4936,
+      "step": 189560
+    },
+    {
+      "epoch": 0.946690304377138,
+      "grad_norm": 0.09415683150291443,
+      "learning_rate": 1.603294200105134e-06,
+      "loss": 8.4812,
+      "step": 189570
+    },
+    {
+      "epoch": 0.9467402432020775,
+      "grad_norm": 0.08812573552131653,
+      "learning_rate": 1.6017922851635838e-06,
+      "loss": 8.5189,
+      "step": 189580
+    },
+    {
+      "epoch": 0.9467901820270169,
+      "grad_norm": 0.0949978157877922,
+      "learning_rate": 1.6002903702220331e-06,
+      "loss": 8.4971,
+      "step": 189590
+    },
+    {
+      "epoch": 0.9468401208519563,
+      "grad_norm": 0.09173561632633209,
+      "learning_rate": 1.5987884552804827e-06,
+      "loss": 8.4943,
+      "step": 189600
+    },
+    {
+      "epoch": 0.9468900596768958,
+      "grad_norm": 0.09235606342554092,
+      "learning_rate": 1.5972865403389322e-06,
+      "loss": 8.4975,
+      "step": 189610
+    },
+    {
+      "epoch": 0.9469399985018353,
+      "grad_norm": 0.09276280552148819,
+      "learning_rate": 1.5957846253973818e-06,
+      "loss": 8.5019,
+      "step": 189620
+    },
+    {
+      "epoch": 0.9469899373267747,
+      "grad_norm": 0.09055425226688385,
+      "learning_rate": 1.5942827104558313e-06,
+      "loss": 8.5128,
+      "step": 189630
+    },
+    {
+      "epoch": 0.9470398761517141,
+      "grad_norm": 0.09334809333086014,
+      "learning_rate": 1.5927807955142808e-06,
+      "loss": 8.4991,
+      "step": 189640
+    },
+    {
+      "epoch": 0.9470898149766536,
+      "grad_norm": 0.09053409099578857,
+      "learning_rate": 1.5912788805727302e-06,
+      "loss": 8.4977,
+      "step": 189650
+    },
+    {
+      "epoch": 0.9471397538015931,
+      "grad_norm": 0.09045706689357758,
+      "learning_rate": 1.58977696563118e-06,
+      "loss": 8.4876,
+      "step": 189660
+    },
+    {
+      "epoch": 0.9471896926265325,
+      "grad_norm": 0.0899735763669014,
+      "learning_rate": 1.5882750506896293e-06,
+      "loss": 8.5024,
+      "step": 189670
+    },
+    {
+      "epoch": 0.9472396314514719,
+      "grad_norm": 0.09022071212530136,
+      "learning_rate": 1.586773135748079e-06,
+      "loss": 8.5059,
+      "step": 189680
+    },
+    {
+      "epoch": 0.9472895702764114,
+      "grad_norm": 0.08928265422582626,
+      "learning_rate": 1.5852712208065283e-06,
+      "loss": 8.5079,
+      "step": 189690
+    },
+    {
+      "epoch": 0.9473395091013509,
+      "grad_norm": 0.08643801510334015,
+      "learning_rate": 1.5837693058649779e-06,
+      "loss": 8.4948,
+      "step": 189700
+    },
+    {
+      "epoch": 0.9473894479262903,
+      "grad_norm": 0.09584169834852219,
+      "learning_rate": 1.5822673909234274e-06,
+      "loss": 8.4735,
+      "step": 189710
+    },
+    {
+      "epoch": 0.9474393867512297,
+      "grad_norm": 0.08683201670646667,
+      "learning_rate": 1.580765475981877e-06,
+      "loss": 8.4989,
+      "step": 189720
+    },
+    {
+      "epoch": 0.9474893255761692,
+      "grad_norm": 0.08900432288646698,
+      "learning_rate": 1.5792635610403265e-06,
+      "loss": 8.5159,
+      "step": 189730
+    },
+    {
+      "epoch": 0.9475392644011087,
+      "grad_norm": 0.0945209488272667,
+      "learning_rate": 1.577761646098776e-06,
+      "loss": 8.4977,
+      "step": 189740
+    },
+    {
+      "epoch": 0.9475892032260481,
+      "grad_norm": 0.09035822004079819,
+      "learning_rate": 1.5762597311572254e-06,
+      "loss": 8.4902,
+      "step": 189750
+    },
+    {
+      "epoch": 0.9476391420509875,
+      "grad_norm": 0.09136451780796051,
+      "learning_rate": 1.5747578162156751e-06,
+      "loss": 8.5167,
+      "step": 189760
+    },
+    {
+      "epoch": 0.947689080875927,
+      "grad_norm": 0.09294602274894714,
+      "learning_rate": 1.5732559012741245e-06,
+      "loss": 8.4919,
+      "step": 189770
+    },
+    {
+      "epoch": 0.9477390197008665,
+      "grad_norm": 0.08745280653238297,
+      "learning_rate": 1.5717539863325742e-06,
+      "loss": 8.5084,
+      "step": 189780
+    },
+    {
+      "epoch": 0.9477889585258059,
+      "grad_norm": 0.08939157426357269,
+      "learning_rate": 1.5702520713910236e-06,
+      "loss": 8.5243,
+      "step": 189790
+    },
+    {
+      "epoch": 0.9478388973507453,
+      "grad_norm": 0.09142594784498215,
+      "learning_rate": 1.5687501564494731e-06,
+      "loss": 8.5088,
+      "step": 189800
+    },
+    {
+      "epoch": 0.9478888361756848,
+      "grad_norm": 0.08955658972263336,
+      "learning_rate": 1.5672482415079227e-06,
+      "loss": 8.5227,
+      "step": 189810
+    },
+    {
+      "epoch": 0.9479387750006243,
+      "grad_norm": 0.08954078704118729,
+      "learning_rate": 1.5657463265663722e-06,
+      "loss": 8.5083,
+      "step": 189820
+    },
+    {
+      "epoch": 0.9479887138255637,
+      "grad_norm": 0.09221227467060089,
+      "learning_rate": 1.5642444116248217e-06,
+      "loss": 8.5135,
+      "step": 189830
+    },
+    {
+      "epoch": 0.9480386526505031,
+      "grad_norm": 0.09133052825927734,
+      "learning_rate": 1.5627424966832713e-06,
+      "loss": 8.5026,
+      "step": 189840
+    },
+    {
+      "epoch": 0.9480885914754426,
+      "grad_norm": 0.0908413901925087,
+      "learning_rate": 1.5612405817417208e-06,
+      "loss": 8.5082,
+      "step": 189850
+    },
+    {
+      "epoch": 0.9481385303003821,
+      "grad_norm": 0.09032785892486572,
+      "learning_rate": 1.5597386668001704e-06,
+      "loss": 8.5031,
+      "step": 189860
+    },
+    {
+      "epoch": 0.9481884691253215,
+      "grad_norm": 0.09097219258546829,
+      "learning_rate": 1.5582367518586197e-06,
+      "loss": 8.5025,
+      "step": 189870
+    },
+    {
+      "epoch": 0.9482384079502609,
+      "grad_norm": 0.08840753883123398,
+      "learning_rate": 1.5567348369170695e-06,
+      "loss": 8.4975,
+      "step": 189880
+    },
+    {
+      "epoch": 0.9482883467752004,
+      "grad_norm": 0.08880984038114548,
+      "learning_rate": 1.5552329219755188e-06,
+      "loss": 8.5018,
+      "step": 189890
+    },
+    {
+      "epoch": 0.9483382856001398,
+      "grad_norm": 0.08809378743171692,
+      "learning_rate": 1.5537310070339685e-06,
+      "loss": 8.5057,
+      "step": 189900
+    },
+    {
+      "epoch": 0.9483882244250793,
+      "grad_norm": 0.09054802358150482,
+      "learning_rate": 1.5522290920924179e-06,
+      "loss": 8.4954,
+      "step": 189910
+    },
+    {
+      "epoch": 0.9484381632500187,
+      "grad_norm": 0.08676068484783173,
+      "learning_rate": 1.5507271771508674e-06,
+      "loss": 8.4857,
+      "step": 189920
+    },
+    {
+      "epoch": 0.9484881020749582,
+      "grad_norm": 0.0902305543422699,
+      "learning_rate": 1.549225262209317e-06,
+      "loss": 8.4918,
+      "step": 189930
+    },
+    {
+      "epoch": 0.9485380408998976,
+      "grad_norm": 0.08811389654874802,
+      "learning_rate": 1.5477233472677665e-06,
+      "loss": 8.5045,
+      "step": 189940
+    },
+    {
+      "epoch": 0.9485879797248371,
+      "grad_norm": 0.08572709560394287,
+      "learning_rate": 1.546221432326216e-06,
+      "loss": 8.5071,
+      "step": 189950
+    },
+    {
+      "epoch": 0.9486379185497765,
+      "grad_norm": 0.09473098814487457,
+      "learning_rate": 1.5447195173846656e-06,
+      "loss": 8.5074,
+      "step": 189960
+    },
+    {
+      "epoch": 0.948687857374716,
+      "grad_norm": 0.08549437671899796,
+      "learning_rate": 1.543217602443115e-06,
+      "loss": 8.5064,
+      "step": 189970
+    },
+    {
+      "epoch": 0.9487377961996554,
+      "grad_norm": 0.08814796805381775,
+      "learning_rate": 1.5417156875015647e-06,
+      "loss": 8.5284,
+      "step": 189980
+    },
+    {
+      "epoch": 0.9487877350245949,
+      "grad_norm": 0.0959198847413063,
+      "learning_rate": 1.540213772560014e-06,
+      "loss": 8.501,
+      "step": 189990
+    },
+    {
+      "epoch": 0.9488376738495343,
+      "grad_norm": 0.08826687932014465,
+      "learning_rate": 1.5387118576184638e-06,
+      "loss": 8.5092,
+      "step": 190000
+    },
+    {
+      "epoch": 0.9488876126744737,
+      "grad_norm": 0.09331593662500381,
+      "learning_rate": 1.537209942676913e-06,
+      "loss": 8.4996,
+      "step": 190010
+    },
+    {
+      "epoch": 0.9489375514994132,
+      "grad_norm": 0.09332361817359924,
+      "learning_rate": 1.5357080277353626e-06,
+      "loss": 8.5019,
+      "step": 190020
+    },
+    {
+      "epoch": 0.9489874903243527,
+      "grad_norm": 0.09323418140411377,
+      "learning_rate": 1.5342061127938122e-06,
+      "loss": 8.5008,
+      "step": 190030
+    },
+    {
+      "epoch": 0.9490374291492921,
+      "grad_norm": 0.09189082682132721,
+      "learning_rate": 1.5327041978522615e-06,
+      "loss": 8.5162,
+      "step": 190040
+    },
+    {
+      "epoch": 0.9490873679742315,
+      "grad_norm": 0.0889974981546402,
+      "learning_rate": 1.5312022829107113e-06,
+      "loss": 8.4987,
+      "step": 190050
+    },
+    {
+      "epoch": 0.949137306799171,
+      "grad_norm": 0.08503816276788712,
+      "learning_rate": 1.5297003679691606e-06,
+      "loss": 8.5133,
+      "step": 190060
+    },
+    {
+      "epoch": 0.9491872456241105,
+      "grad_norm": 0.09503103792667389,
+      "learning_rate": 1.5281984530276103e-06,
+      "loss": 8.4963,
+      "step": 190070
+    },
+    {
+      "epoch": 0.9492371844490499,
+      "grad_norm": 0.08845657110214233,
+      "learning_rate": 1.5266965380860597e-06,
+      "loss": 8.4878,
+      "step": 190080
+    },
+    {
+      "epoch": 0.9492871232739893,
+      "grad_norm": 0.08706570416688919,
+      "learning_rate": 1.5251946231445092e-06,
+      "loss": 8.5065,
+      "step": 190090
+    },
+    {
+      "epoch": 0.9493370620989288,
+      "grad_norm": 0.08456262946128845,
+      "learning_rate": 1.5236927082029588e-06,
+      "loss": 8.502,
+      "step": 190100
+    },
+    {
+      "epoch": 0.9493870009238683,
+      "grad_norm": 0.09131083637475967,
+      "learning_rate": 1.5221907932614083e-06,
+      "loss": 8.4949,
+      "step": 190110
+    },
+    {
+      "epoch": 0.9494369397488077,
+      "grad_norm": 0.09142878651618958,
+      "learning_rate": 1.5206888783198579e-06,
+      "loss": 8.5019,
+      "step": 190120
+    },
+    {
+      "epoch": 0.9494868785737471,
+      "grad_norm": 0.08769478648900986,
+      "learning_rate": 1.5191869633783074e-06,
+      "loss": 8.5095,
+      "step": 190130
+    },
+    {
+      "epoch": 0.9495368173986866,
+      "grad_norm": 0.09371474385261536,
+      "learning_rate": 1.5176850484367567e-06,
+      "loss": 8.5074,
+      "step": 190140
+    },
+    {
+      "epoch": 0.9495867562236261,
+      "grad_norm": 0.08654850721359253,
+      "learning_rate": 1.5161831334952065e-06,
+      "loss": 8.5099,
+      "step": 190150
+    },
+    {
+      "epoch": 0.9496366950485655,
+      "grad_norm": 0.09408658742904663,
+      "learning_rate": 1.5146812185536558e-06,
+      "loss": 8.4846,
+      "step": 190160
+    },
+    {
+      "epoch": 0.9496866338735049,
+      "grad_norm": 0.0925004854798317,
+      "learning_rate": 1.5131793036121056e-06,
+      "loss": 8.4927,
+      "step": 190170
+    },
+    {
+      "epoch": 0.9497365726984444,
+      "grad_norm": 0.09245412796735764,
+      "learning_rate": 1.511677388670555e-06,
+      "loss": 8.5093,
+      "step": 190180
+    },
+    {
+      "epoch": 0.9497865115233839,
+      "grad_norm": 0.0886409729719162,
+      "learning_rate": 1.5101754737290044e-06,
+      "loss": 8.5086,
+      "step": 190190
+    },
+    {
+      "epoch": 0.9498364503483233,
+      "grad_norm": 0.08819697797298431,
+      "learning_rate": 1.508673558787454e-06,
+      "loss": 8.4979,
+      "step": 190200
+    },
+    {
+      "epoch": 0.9498863891732627,
+      "grad_norm": 0.09439955651760101,
+      "learning_rate": 1.5071716438459035e-06,
+      "loss": 8.5077,
+      "step": 190210
+    },
+    {
+      "epoch": 0.9499363279982022,
+      "grad_norm": 0.09136414527893066,
+      "learning_rate": 1.505669728904353e-06,
+      "loss": 8.5202,
+      "step": 190220
+    },
+    {
+      "epoch": 0.9499862668231417,
+      "grad_norm": 0.09268520772457123,
+      "learning_rate": 1.5041678139628026e-06,
+      "loss": 8.498,
+      "step": 190230
+    },
+    {
+      "epoch": 0.9500362056480811,
+      "grad_norm": 0.08827424794435501,
+      "learning_rate": 1.502665899021252e-06,
+      "loss": 8.5254,
+      "step": 190240
+    },
+    {
+      "epoch": 0.9500861444730205,
+      "grad_norm": 0.09765148162841797,
+      "learning_rate": 1.5011639840797017e-06,
+      "loss": 8.4877,
+      "step": 190250
+    },
+    {
+      "epoch": 0.95013608329796,
+      "grad_norm": 0.09497714787721634,
+      "learning_rate": 1.499662069138151e-06,
+      "loss": 8.495,
+      "step": 190260
+    },
+    {
+      "epoch": 0.9501860221228995,
+      "grad_norm": 0.091663658618927,
+      "learning_rate": 1.4981601541966008e-06,
+      "loss": 8.5161,
+      "step": 190270
+    },
+    {
+      "epoch": 0.9502359609478389,
+      "grad_norm": 0.09366423636674881,
+      "learning_rate": 1.4966582392550501e-06,
+      "loss": 8.4835,
+      "step": 190280
+    },
+    {
+      "epoch": 0.9502858997727783,
+      "grad_norm": 0.09287559986114502,
+      "learning_rate": 1.4951563243134997e-06,
+      "loss": 8.4897,
+      "step": 190290
+    },
+    {
+      "epoch": 0.9503358385977178,
+      "grad_norm": 0.09357742220163345,
+      "learning_rate": 1.4936544093719492e-06,
+      "loss": 8.4976,
+      "step": 190300
+    },
+    {
+      "epoch": 0.9503857774226573,
+      "grad_norm": 0.09103827178478241,
+      "learning_rate": 1.4921524944303987e-06,
+      "loss": 8.4842,
+      "step": 190310
+    },
+    {
+      "epoch": 0.9504357162475967,
+      "grad_norm": 0.08826170116662979,
+      "learning_rate": 1.4906505794888483e-06,
+      "loss": 8.5147,
+      "step": 190320
+    },
+    {
+      "epoch": 0.9504856550725361,
+      "grad_norm": 0.08988435566425323,
+      "learning_rate": 1.4891486645472978e-06,
+      "loss": 8.4985,
+      "step": 190330
+    },
+    {
+      "epoch": 0.9505355938974756,
+      "grad_norm": 0.09322842955589294,
+      "learning_rate": 1.4876467496057474e-06,
+      "loss": 8.5058,
+      "step": 190340
+    },
+    {
+      "epoch": 0.9505855327224151,
+      "grad_norm": 0.09746944904327393,
+      "learning_rate": 1.486144834664197e-06,
+      "loss": 8.4953,
+      "step": 190350
+    },
+    {
+      "epoch": 0.9506354715473545,
+      "grad_norm": 0.08793298900127411,
+      "learning_rate": 1.4846429197226463e-06,
+      "loss": 8.495,
+      "step": 190360
+    },
+    {
+      "epoch": 0.9506854103722939,
+      "grad_norm": 0.08952205628156662,
+      "learning_rate": 1.483141004781096e-06,
+      "loss": 8.5206,
+      "step": 190370
+    },
+    {
+      "epoch": 0.9507353491972333,
+      "grad_norm": 0.09018826484680176,
+      "learning_rate": 1.4816390898395453e-06,
+      "loss": 8.4989,
+      "step": 190380
+    },
+    {
+      "epoch": 0.9507852880221729,
+      "grad_norm": 0.08909120410680771,
+      "learning_rate": 1.480137174897995e-06,
+      "loss": 8.5011,
+      "step": 190390
+    },
+    {
+      "epoch": 0.9508352268471123,
+      "grad_norm": 0.0869319885969162,
+      "learning_rate": 1.4786352599564444e-06,
+      "loss": 8.4915,
+      "step": 190400
+    },
+    {
+      "epoch": 0.9508851656720517,
+      "grad_norm": 0.0899181067943573,
+      "learning_rate": 1.477133345014894e-06,
+      "loss": 8.4897,
+      "step": 190410
+    },
+    {
+      "epoch": 0.9509351044969911,
+      "grad_norm": 0.0868888720870018,
+      "learning_rate": 1.4756314300733435e-06,
+      "loss": 8.5152,
+      "step": 190420
+    },
+    {
+      "epoch": 0.9509850433219307,
+      "grad_norm": 0.09073241055011749,
+      "learning_rate": 1.474129515131793e-06,
+      "loss": 8.5101,
+      "step": 190430
+    },
+    {
+      "epoch": 0.9510349821468701,
+      "grad_norm": 0.08949577063322067,
+      "learning_rate": 1.4726276001902426e-06,
+      "loss": 8.4997,
+      "step": 190440
+    },
+    {
+      "epoch": 0.9510849209718095,
+      "grad_norm": 0.09318594634532928,
+      "learning_rate": 1.4711256852486921e-06,
+      "loss": 8.4928,
+      "step": 190450
+    },
+    {
+      "epoch": 0.9511348597967489,
+      "grad_norm": 0.08922556787729263,
+      "learning_rate": 1.4696237703071415e-06,
+      "loss": 8.5018,
+      "step": 190460
+    },
+    {
+      "epoch": 0.9511847986216885,
+      "grad_norm": 0.08539392054080963,
+      "learning_rate": 1.4681218553655912e-06,
+      "loss": 8.4915,
+      "step": 190470
+    },
+    {
+      "epoch": 0.9512347374466279,
+      "grad_norm": 0.09017814695835114,
+      "learning_rate": 1.4666199404240406e-06,
+      "loss": 8.4938,
+      "step": 190480
+    },
+    {
+      "epoch": 0.9512846762715673,
+      "grad_norm": 0.08659885823726654,
+      "learning_rate": 1.4651180254824903e-06,
+      "loss": 8.487,
+      "step": 190490
+    },
+    {
+      "epoch": 0.9513346150965067,
+      "grad_norm": 0.09293199330568314,
+      "learning_rate": 1.4636161105409396e-06,
+      "loss": 8.5041,
+      "step": 190500
+    },
+    {
+      "epoch": 0.9513845539214463,
+      "grad_norm": 0.08663637191057205,
+      "learning_rate": 1.4621141955993892e-06,
+      "loss": 8.49,
+      "step": 190510
+    },
+    {
+      "epoch": 0.9514344927463857,
+      "grad_norm": 0.09226275235414505,
+      "learning_rate": 1.4606122806578387e-06,
+      "loss": 8.4864,
+      "step": 190520
+    },
+    {
+      "epoch": 0.9514844315713251,
+      "grad_norm": 0.09186282753944397,
+      "learning_rate": 1.4591103657162883e-06,
+      "loss": 8.5054,
+      "step": 190530
+    },
+    {
+      "epoch": 0.9515343703962645,
+      "grad_norm": 0.09182189404964447,
+      "learning_rate": 1.4576084507747378e-06,
+      "loss": 8.5114,
+      "step": 190540
+    },
+    {
+      "epoch": 0.9515843092212041,
+      "grad_norm": 0.09429624676704407,
+      "learning_rate": 1.4561065358331874e-06,
+      "loss": 8.5163,
+      "step": 190550
+    },
+    {
+      "epoch": 0.9516342480461435,
+      "grad_norm": 0.09003430604934692,
+      "learning_rate": 1.4546046208916367e-06,
+      "loss": 8.4997,
+      "step": 190560
+    },
+    {
+      "epoch": 0.9516841868710829,
+      "grad_norm": 0.0899423211812973,
+      "learning_rate": 1.4531027059500864e-06,
+      "loss": 8.5114,
+      "step": 190570
+    },
+    {
+      "epoch": 0.9517341256960223,
+      "grad_norm": 0.0908636823296547,
+      "learning_rate": 1.4516007910085358e-06,
+      "loss": 8.5043,
+      "step": 190580
+    },
+    {
+      "epoch": 0.9517840645209619,
+      "grad_norm": 0.095972940325737,
+      "learning_rate": 1.4500988760669855e-06,
+      "loss": 8.5164,
+      "step": 190590
+    },
+    {
+      "epoch": 0.9518340033459013,
+      "grad_norm": 0.09233302623033524,
+      "learning_rate": 1.4485969611254349e-06,
+      "loss": 8.5112,
+      "step": 190600
+    },
+    {
+      "epoch": 0.9518839421708407,
+      "grad_norm": 0.0877123698592186,
+      "learning_rate": 1.4470950461838844e-06,
+      "loss": 8.5003,
+      "step": 190610
+    },
+    {
+      "epoch": 0.9519338809957801,
+      "grad_norm": 0.08805521577596664,
+      "learning_rate": 1.445593131242334e-06,
+      "loss": 8.5093,
+      "step": 190620
+    },
+    {
+      "epoch": 0.9519838198207197,
+      "grad_norm": 0.09961655735969543,
+      "learning_rate": 1.4440912163007835e-06,
+      "loss": 8.4953,
+      "step": 190630
+    },
+    {
+      "epoch": 0.9520337586456591,
+      "grad_norm": 0.09043627232313156,
+      "learning_rate": 1.442589301359233e-06,
+      "loss": 8.4995,
+      "step": 190640
+    },
+    {
+      "epoch": 0.9520836974705985,
+      "grad_norm": 0.08915705978870392,
+      "learning_rate": 1.4410873864176826e-06,
+      "loss": 8.502,
+      "step": 190650
+    },
+    {
+      "epoch": 0.9521336362955379,
+      "grad_norm": 0.08573364466428757,
+      "learning_rate": 1.4395854714761321e-06,
+      "loss": 8.5038,
+      "step": 190660
+    },
+    {
+      "epoch": 0.9521835751204775,
+      "grad_norm": 0.08786933869123459,
+      "learning_rate": 1.4380835565345817e-06,
+      "loss": 8.5091,
+      "step": 190670
+    },
+    {
+      "epoch": 0.9522335139454169,
+      "grad_norm": 0.08725972473621368,
+      "learning_rate": 1.436581641593031e-06,
+      "loss": 8.4832,
+      "step": 190680
+    },
+    {
+      "epoch": 0.9522834527703563,
+      "grad_norm": 0.08919131010770798,
+      "learning_rate": 1.4350797266514807e-06,
+      "loss": 8.4989,
+      "step": 190690
+    },
+    {
+      "epoch": 0.9523333915952957,
+      "grad_norm": 0.08816230297088623,
+      "learning_rate": 1.43357781170993e-06,
+      "loss": 8.4939,
+      "step": 190700
+    },
+    {
+      "epoch": 0.9523833304202353,
+      "grad_norm": 0.08805028349161148,
+      "learning_rate": 1.4320758967683798e-06,
+      "loss": 8.5108,
+      "step": 190710
+    },
+    {
+      "epoch": 0.9524332692451747,
+      "grad_norm": 0.08876121044158936,
+      "learning_rate": 1.4305739818268292e-06,
+      "loss": 8.5145,
+      "step": 190720
+    },
+    {
+      "epoch": 0.9524832080701141,
+      "grad_norm": 0.09014929085969925,
+      "learning_rate": 1.4290720668852787e-06,
+      "loss": 8.5085,
+      "step": 190730
+    },
+    {
+      "epoch": 0.9525331468950535,
+      "grad_norm": 0.09153331071138382,
+      "learning_rate": 1.4275701519437283e-06,
+      "loss": 8.4938,
+      "step": 190740
+    },
+    {
+      "epoch": 0.9525830857199931,
+      "grad_norm": 0.09297018498182297,
+      "learning_rate": 1.4260682370021778e-06,
+      "loss": 8.5057,
+      "step": 190750
+    },
+    {
+      "epoch": 0.9526330245449325,
+      "grad_norm": 0.09277532249689102,
+      "learning_rate": 1.4245663220606273e-06,
+      "loss": 8.5036,
+      "step": 190760
+    },
+    {
+      "epoch": 0.9526829633698719,
+      "grad_norm": 0.08956871181726456,
+      "learning_rate": 1.4230644071190769e-06,
+      "loss": 8.4772,
+      "step": 190770
+    },
+    {
+      "epoch": 0.9527329021948113,
+      "grad_norm": 0.09383154660463333,
+      "learning_rate": 1.4215624921775262e-06,
+      "loss": 8.5021,
+      "step": 190780
+    },
+    {
+      "epoch": 0.9527828410197509,
+      "grad_norm": 0.08647419512271881,
+      "learning_rate": 1.420060577235976e-06,
+      "loss": 8.4969,
+      "step": 190790
+    },
+    {
+      "epoch": 0.9528327798446903,
+      "grad_norm": 0.08593643456697464,
+      "learning_rate": 1.4185586622944253e-06,
+      "loss": 8.496,
+      "step": 190800
+    },
+    {
+      "epoch": 0.9528827186696297,
+      "grad_norm": 0.09408606588840485,
+      "learning_rate": 1.417056747352875e-06,
+      "loss": 8.4957,
+      "step": 190810
+    },
+    {
+      "epoch": 0.9529326574945691,
+      "grad_norm": 0.08725640922784805,
+      "learning_rate": 1.4155548324113244e-06,
+      "loss": 8.5259,
+      "step": 190820
+    },
+    {
+      "epoch": 0.9529825963195087,
+      "grad_norm": 0.09070757776498795,
+      "learning_rate": 1.414052917469774e-06,
+      "loss": 8.4818,
+      "step": 190830
+    },
+    {
+      "epoch": 0.9530325351444481,
+      "grad_norm": 0.08983830362558365,
+      "learning_rate": 1.4125510025282235e-06,
+      "loss": 8.4856,
+      "step": 190840
+    },
+    {
+      "epoch": 0.9530824739693875,
+      "grad_norm": 0.08678752928972244,
+      "learning_rate": 1.411049087586673e-06,
+      "loss": 8.513,
+      "step": 190850
+    },
+    {
+      "epoch": 0.9531324127943269,
+      "grad_norm": 0.09017140418291092,
+      "learning_rate": 1.4095471726451226e-06,
+      "loss": 8.518,
+      "step": 190860
+    },
+    {
+      "epoch": 0.9531823516192663,
+      "grad_norm": 0.09126543253660202,
+      "learning_rate": 1.408045257703572e-06,
+      "loss": 8.508,
+      "step": 190870
+    },
+    {
+      "epoch": 0.9532322904442059,
+      "grad_norm": 0.09493157267570496,
+      "learning_rate": 1.4065433427620216e-06,
+      "loss": 8.5058,
+      "step": 190880
+    },
+    {
+      "epoch": 0.9532822292691453,
+      "grad_norm": 0.09498272091150284,
+      "learning_rate": 1.4050414278204712e-06,
+      "loss": 8.5069,
+      "step": 190890
+    },
+    {
+      "epoch": 0.9533321680940847,
+      "grad_norm": 0.08452911674976349,
+      "learning_rate": 1.4035395128789205e-06,
+      "loss": 8.5176,
+      "step": 190900
+    },
+    {
+      "epoch": 0.9533821069190241,
+      "grad_norm": 0.08910375088453293,
+      "learning_rate": 1.4020375979373703e-06,
+      "loss": 8.5208,
+      "step": 190910
+    },
+    {
+      "epoch": 0.9534320457439637,
+      "grad_norm": 0.10094549506902695,
+      "learning_rate": 1.4005356829958196e-06,
+      "loss": 8.5057,
+      "step": 190920
+    },
+    {
+      "epoch": 0.9534819845689031,
+      "grad_norm": 0.08806487917900085,
+      "learning_rate": 1.3990337680542694e-06,
+      "loss": 8.4974,
+      "step": 190930
+    },
+    {
+      "epoch": 0.9535319233938425,
+      "grad_norm": 0.09561792016029358,
+      "learning_rate": 1.3975318531127187e-06,
+      "loss": 8.4931,
+      "step": 190940
+    },
+    {
+      "epoch": 0.9535818622187819,
+      "grad_norm": 0.08623037487268448,
+      "learning_rate": 1.3960299381711682e-06,
+      "loss": 8.496,
+      "step": 190950
+    },
+    {
+      "epoch": 0.9536318010437215,
+      "grad_norm": 0.08998066931962967,
+      "learning_rate": 1.3945280232296178e-06,
+      "loss": 8.5056,
+      "step": 190960
+    },
+    {
+      "epoch": 0.9536817398686609,
+      "grad_norm": 0.10207744687795639,
+      "learning_rate": 1.3930261082880673e-06,
+      "loss": 8.4875,
+      "step": 190970
+    },
+    {
+      "epoch": 0.9537316786936003,
+      "grad_norm": 0.09375947713851929,
+      "learning_rate": 1.3915241933465169e-06,
+      "loss": 8.4791,
+      "step": 190980
+    },
+    {
+      "epoch": 0.9537816175185397,
+      "grad_norm": 0.09087695181369781,
+      "learning_rate": 1.3900222784049664e-06,
+      "loss": 8.4961,
+      "step": 190990
+    },
+    {
+      "epoch": 0.9538315563434793,
+      "grad_norm": 0.09509552270174026,
+      "learning_rate": 1.3885203634634157e-06,
+      "loss": 8.4978,
+      "step": 191000
+    },
+    {
+      "epoch": 0.9538814951684187,
+      "grad_norm": 0.08731069415807724,
+      "learning_rate": 1.3870184485218655e-06,
+      "loss": 8.5142,
+      "step": 191010
+    },
+    {
+      "epoch": 0.9539314339933581,
+      "grad_norm": 0.09380168467760086,
+      "learning_rate": 1.3855165335803148e-06,
+      "loss": 8.4932,
+      "step": 191020
+    },
+    {
+      "epoch": 0.9539813728182975,
+      "grad_norm": 0.09554290771484375,
+      "learning_rate": 1.3840146186387646e-06,
+      "loss": 8.4965,
+      "step": 191030
+    },
+    {
+      "epoch": 0.9540313116432371,
+      "grad_norm": 0.09323696792125702,
+      "learning_rate": 1.382512703697214e-06,
+      "loss": 8.509,
+      "step": 191040
+    },
+    {
+      "epoch": 0.9540812504681765,
+      "grad_norm": 0.09462062269449234,
+      "learning_rate": 1.3810107887556635e-06,
+      "loss": 8.5129,
+      "step": 191050
+    },
+    {
+      "epoch": 0.9541311892931159,
+      "grad_norm": 0.08865117281675339,
+      "learning_rate": 1.379508873814113e-06,
+      "loss": 8.5054,
+      "step": 191060
+    },
+    {
+      "epoch": 0.9541811281180553,
+      "grad_norm": 0.09803714603185654,
+      "learning_rate": 1.3780069588725625e-06,
+      "loss": 8.5022,
+      "step": 191070
+    },
+    {
+      "epoch": 0.9542310669429949,
+      "grad_norm": 0.09523754566907883,
+      "learning_rate": 1.376505043931012e-06,
+      "loss": 8.4978,
+      "step": 191080
+    },
+    {
+      "epoch": 0.9542810057679343,
+      "grad_norm": 0.09519781172275543,
+      "learning_rate": 1.3750031289894616e-06,
+      "loss": 8.4971,
+      "step": 191090
+    },
+    {
+      "epoch": 0.9543309445928737,
+      "grad_norm": 0.09077487140893936,
+      "learning_rate": 1.373501214047911e-06,
+      "loss": 8.5253,
+      "step": 191100
+    },
+    {
+      "epoch": 0.9543808834178131,
+      "grad_norm": 0.08766099810600281,
+      "learning_rate": 1.3719992991063607e-06,
+      "loss": 8.4914,
+      "step": 191110
+    },
+    {
+      "epoch": 0.9544308222427527,
+      "grad_norm": 0.08919545263051987,
+      "learning_rate": 1.37049738416481e-06,
+      "loss": 8.4901,
+      "step": 191120
+    },
+    {
+      "epoch": 0.9544807610676921,
+      "grad_norm": 0.091376394033432,
+      "learning_rate": 1.3689954692232598e-06,
+      "loss": 8.4782,
+      "step": 191130
+    },
+    {
+      "epoch": 0.9545306998926315,
+      "grad_norm": 0.09148900955915451,
+      "learning_rate": 1.3674935542817091e-06,
+      "loss": 8.5037,
+      "step": 191140
+    },
+    {
+      "epoch": 0.9545806387175709,
+      "grad_norm": 0.08681777864694595,
+      "learning_rate": 1.3659916393401587e-06,
+      "loss": 8.4946,
+      "step": 191150
+    },
+    {
+      "epoch": 0.9546305775425105,
+      "grad_norm": 0.09286382794380188,
+      "learning_rate": 1.3644897243986082e-06,
+      "loss": 8.4937,
+      "step": 191160
+    },
+    {
+      "epoch": 0.9546805163674499,
+      "grad_norm": 0.08869923651218414,
+      "learning_rate": 1.3629878094570578e-06,
+      "loss": 8.5241,
+      "step": 191170
+    },
+    {
+      "epoch": 0.9547304551923893,
+      "grad_norm": 0.09262396395206451,
+      "learning_rate": 1.3614858945155073e-06,
+      "loss": 8.4772,
+      "step": 191180
+    },
+    {
+      "epoch": 0.9547803940173287,
+      "grad_norm": 0.08976420015096664,
+      "learning_rate": 1.3599839795739568e-06,
+      "loss": 8.5162,
+      "step": 191190
+    },
+    {
+      "epoch": 0.9548303328422683,
+      "grad_norm": 0.08680476248264313,
+      "learning_rate": 1.3584820646324064e-06,
+      "loss": 8.5119,
+      "step": 191200
+    },
+    {
+      "epoch": 0.9548802716672077,
+      "grad_norm": 0.09427861869335175,
+      "learning_rate": 1.356980149690856e-06,
+      "loss": 8.5139,
+      "step": 191210
+    },
+    {
+      "epoch": 0.9549302104921471,
+      "grad_norm": 0.08783675730228424,
+      "learning_rate": 1.3554782347493053e-06,
+      "loss": 8.5041,
+      "step": 191220
+    },
+    {
+      "epoch": 0.9549801493170865,
+      "grad_norm": 0.09209182858467102,
+      "learning_rate": 1.353976319807755e-06,
+      "loss": 8.5103,
+      "step": 191230
+    },
+    {
+      "epoch": 0.9550300881420261,
+      "grad_norm": 0.08795662224292755,
+      "learning_rate": 1.3524744048662043e-06,
+      "loss": 8.5108,
+      "step": 191240
+    },
+    {
+      "epoch": 0.9550800269669655,
+      "grad_norm": 0.09797421097755432,
+      "learning_rate": 1.350972489924654e-06,
+      "loss": 8.4915,
+      "step": 191250
+    },
+    {
+      "epoch": 0.9551299657919049,
+      "grad_norm": 0.0975293442606926,
+      "learning_rate": 1.3494705749831034e-06,
+      "loss": 8.5011,
+      "step": 191260
+    },
+    {
+      "epoch": 0.9551799046168443,
+      "grad_norm": 0.09412509948015213,
+      "learning_rate": 1.347968660041553e-06,
+      "loss": 8.5011,
+      "step": 191270
+    },
+    {
+      "epoch": 0.9552298434417839,
+      "grad_norm": 0.09029638022184372,
+      "learning_rate": 1.3464667451000025e-06,
+      "loss": 8.5052,
+      "step": 191280
+    },
+    {
+      "epoch": 0.9552797822667233,
+      "grad_norm": 0.09370943903923035,
+      "learning_rate": 1.344964830158452e-06,
+      "loss": 8.4932,
+      "step": 191290
+    },
+    {
+      "epoch": 0.9553297210916627,
+      "grad_norm": 0.09150079637765884,
+      "learning_rate": 1.3434629152169016e-06,
+      "loss": 8.5044,
+      "step": 191300
+    },
+    {
+      "epoch": 0.9553796599166021,
+      "grad_norm": 0.09412145614624023,
+      "learning_rate": 1.3419610002753511e-06,
+      "loss": 8.5145,
+      "step": 191310
+    },
+    {
+      "epoch": 0.9554295987415417,
+      "grad_norm": 0.09421419352293015,
+      "learning_rate": 1.3404590853338005e-06,
+      "loss": 8.5175,
+      "step": 191320
+    },
+    {
+      "epoch": 0.9554795375664811,
+      "grad_norm": 0.08627758920192719,
+      "learning_rate": 1.3389571703922502e-06,
+      "loss": 8.497,
+      "step": 191330
+    },
+    {
+      "epoch": 0.9555294763914205,
+      "grad_norm": 0.0896778404712677,
+      "learning_rate": 1.3374552554506996e-06,
+      "loss": 8.4944,
+      "step": 191340
+    },
+    {
+      "epoch": 0.9555794152163599,
+      "grad_norm": 0.09137307852506638,
+      "learning_rate": 1.3359533405091493e-06,
+      "loss": 8.4975,
+      "step": 191350
+    },
+    {
+      "epoch": 0.9556293540412995,
+      "grad_norm": 0.0910387709736824,
+      "learning_rate": 1.3344514255675987e-06,
+      "loss": 8.5027,
+      "step": 191360
+    },
+    {
+      "epoch": 0.9556792928662389,
+      "grad_norm": 0.09388083219528198,
+      "learning_rate": 1.3329495106260482e-06,
+      "loss": 8.5204,
+      "step": 191370
+    },
+    {
+      "epoch": 0.9557292316911783,
+      "grad_norm": 0.0911535695195198,
+      "learning_rate": 1.3314475956844977e-06,
+      "loss": 8.5079,
+      "step": 191380
+    },
+    {
+      "epoch": 0.9557791705161177,
+      "grad_norm": 0.0940667912364006,
+      "learning_rate": 1.3299456807429473e-06,
+      "loss": 8.5054,
+      "step": 191390
+    },
+    {
+      "epoch": 0.9558291093410572,
+      "grad_norm": 0.08681704849004745,
+      "learning_rate": 1.3284437658013968e-06,
+      "loss": 8.4992,
+      "step": 191400
+    },
+    {
+      "epoch": 0.9558790481659967,
+      "grad_norm": 0.09067001193761826,
+      "learning_rate": 1.3269418508598464e-06,
+      "loss": 8.5076,
+      "step": 191410
+    },
+    {
+      "epoch": 0.9559289869909361,
+      "grad_norm": 0.08941010385751724,
+      "learning_rate": 1.325439935918296e-06,
+      "loss": 8.4981,
+      "step": 191420
+    },
+    {
+      "epoch": 0.9559789258158755,
+      "grad_norm": 0.08782663941383362,
+      "learning_rate": 1.3239380209767455e-06,
+      "loss": 8.4801,
+      "step": 191430
+    },
+    {
+      "epoch": 0.956028864640815,
+      "grad_norm": 0.0957464948296547,
+      "learning_rate": 1.3224361060351948e-06,
+      "loss": 8.4795,
+      "step": 191440
+    },
+    {
+      "epoch": 0.9560788034657545,
+      "grad_norm": 0.0979733094573021,
+      "learning_rate": 1.3209341910936445e-06,
+      "loss": 8.4909,
+      "step": 191450
+    },
+    {
+      "epoch": 0.9561287422906939,
+      "grad_norm": 0.09294435381889343,
+      "learning_rate": 1.3194322761520939e-06,
+      "loss": 8.4827,
+      "step": 191460
+    },
+    {
+      "epoch": 0.9561786811156333,
+      "grad_norm": 0.09183111786842346,
+      "learning_rate": 1.3179303612105436e-06,
+      "loss": 8.4947,
+      "step": 191470
+    },
+    {
+      "epoch": 0.9562286199405728,
+      "grad_norm": 0.0917709469795227,
+      "learning_rate": 1.316428446268993e-06,
+      "loss": 8.5117,
+      "step": 191480
+    },
+    {
+      "epoch": 0.9562785587655123,
+      "grad_norm": 0.0903320387005806,
+      "learning_rate": 1.3149265313274425e-06,
+      "loss": 8.4863,
+      "step": 191490
+    },
+    {
+      "epoch": 0.9563284975904517,
+      "grad_norm": 0.09910201281309128,
+      "learning_rate": 1.313424616385892e-06,
+      "loss": 8.519,
+      "step": 191500
+    },
+    {
+      "epoch": 0.9563784364153911,
+      "grad_norm": 0.08993766456842422,
+      "learning_rate": 1.3119227014443416e-06,
+      "loss": 8.5283,
+      "step": 191510
+    },
+    {
+      "epoch": 0.9564283752403306,
+      "grad_norm": 0.09398555755615234,
+      "learning_rate": 1.3104207865027911e-06,
+      "loss": 8.4969,
+      "step": 191520
+    },
+    {
+      "epoch": 0.9564783140652701,
+      "grad_norm": 0.08856987208127975,
+      "learning_rate": 1.3089188715612407e-06,
+      "loss": 8.5304,
+      "step": 191530
+    },
+    {
+      "epoch": 0.9565282528902095,
+      "grad_norm": 0.09370505064725876,
+      "learning_rate": 1.30741695661969e-06,
+      "loss": 8.5073,
+      "step": 191540
+    },
+    {
+      "epoch": 0.9565781917151489,
+      "grad_norm": 0.08654750138521194,
+      "learning_rate": 1.3059150416781398e-06,
+      "loss": 8.502,
+      "step": 191550
+    },
+    {
+      "epoch": 0.9566281305400884,
+      "grad_norm": 0.08909036964178085,
+      "learning_rate": 1.304413126736589e-06,
+      "loss": 8.5031,
+      "step": 191560
+    },
+    {
+      "epoch": 0.9566780693650279,
+      "grad_norm": 0.09467563778162003,
+      "learning_rate": 1.3029112117950388e-06,
+      "loss": 8.4999,
+      "step": 191570
+    },
+    {
+      "epoch": 0.9567280081899673,
+      "grad_norm": 0.08794749528169632,
+      "learning_rate": 1.3014092968534882e-06,
+      "loss": 8.5046,
+      "step": 191580
+    },
+    {
+      "epoch": 0.9567779470149067,
+      "grad_norm": 0.08709059655666351,
+      "learning_rate": 1.2999073819119377e-06,
+      "loss": 8.4894,
+      "step": 191590
+    },
+    {
+      "epoch": 0.9568278858398462,
+      "grad_norm": 0.09558872878551483,
+      "learning_rate": 1.2984054669703873e-06,
+      "loss": 8.5192,
+      "step": 191600
+    },
+    {
+      "epoch": 0.9568778246647857,
+      "grad_norm": 0.09214767813682556,
+      "learning_rate": 1.2969035520288368e-06,
+      "loss": 8.5062,
+      "step": 191610
+    },
+    {
+      "epoch": 0.9569277634897251,
+      "grad_norm": 0.08921287953853607,
+      "learning_rate": 1.2954016370872863e-06,
+      "loss": 8.5002,
+      "step": 191620
+    },
+    {
+      "epoch": 0.9569777023146645,
+      "grad_norm": 0.09433959424495697,
+      "learning_rate": 1.2938997221457359e-06,
+      "loss": 8.497,
+      "step": 191630
+    },
+    {
+      "epoch": 0.957027641139604,
+      "grad_norm": 0.08520886301994324,
+      "learning_rate": 1.2923978072041852e-06,
+      "loss": 8.4983,
+      "step": 191640
+    },
+    {
+      "epoch": 0.9570775799645435,
+      "grad_norm": 0.0877714678645134,
+      "learning_rate": 1.290895892262635e-06,
+      "loss": 8.513,
+      "step": 191650
+    },
+    {
+      "epoch": 0.9571275187894829,
+      "grad_norm": 0.09116245806217194,
+      "learning_rate": 1.2893939773210843e-06,
+      "loss": 8.4974,
+      "step": 191660
+    },
+    {
+      "epoch": 0.9571774576144223,
+      "grad_norm": 0.08422242105007172,
+      "learning_rate": 1.287892062379534e-06,
+      "loss": 8.5094,
+      "step": 191670
+    },
+    {
+      "epoch": 0.9572273964393618,
+      "grad_norm": 0.08759858459234238,
+      "learning_rate": 1.2863901474379834e-06,
+      "loss": 8.4916,
+      "step": 191680
+    },
+    {
+      "epoch": 0.9572773352643013,
+      "grad_norm": 0.08717787265777588,
+      "learning_rate": 1.284888232496433e-06,
+      "loss": 8.5063,
+      "step": 191690
+    },
+    {
+      "epoch": 0.9573272740892407,
+      "grad_norm": 0.09190387278795242,
+      "learning_rate": 1.2833863175548825e-06,
+      "loss": 8.4875,
+      "step": 191700
+    },
+    {
+      "epoch": 0.9573772129141801,
+      "grad_norm": 0.09158718585968018,
+      "learning_rate": 1.281884402613332e-06,
+      "loss": 8.5172,
+      "step": 191710
+    },
+    {
+      "epoch": 0.9574271517391196,
+      "grad_norm": 0.09621883928775787,
+      "learning_rate": 1.2803824876717816e-06,
+      "loss": 8.5133,
+      "step": 191720
+    },
+    {
+      "epoch": 0.957477090564059,
+      "grad_norm": 0.08670492470264435,
+      "learning_rate": 1.2788805727302311e-06,
+      "loss": 8.5062,
+      "step": 191730
+    },
+    {
+      "epoch": 0.9575270293889985,
+      "grad_norm": 0.08832064270973206,
+      "learning_rate": 1.2773786577886807e-06,
+      "loss": 8.4969,
+      "step": 191740
+    },
+    {
+      "epoch": 0.9575769682139379,
+      "grad_norm": 0.08412336558103561,
+      "learning_rate": 1.2758767428471302e-06,
+      "loss": 8.5298,
+      "step": 191750
+    },
+    {
+      "epoch": 0.9576269070388774,
+      "grad_norm": 0.09177365899085999,
+      "learning_rate": 1.2743748279055795e-06,
+      "loss": 8.494,
+      "step": 191760
+    },
+    {
+      "epoch": 0.9576768458638169,
+      "grad_norm": 0.08740728348493576,
+      "learning_rate": 1.2728729129640293e-06,
+      "loss": 8.4938,
+      "step": 191770
+    },
+    {
+      "epoch": 0.9577267846887563,
+      "grad_norm": 0.08786563575267792,
+      "learning_rate": 1.2713709980224786e-06,
+      "loss": 8.5155,
+      "step": 191780
+    },
+    {
+      "epoch": 0.9577767235136957,
+      "grad_norm": 0.09397321939468384,
+      "learning_rate": 1.2698690830809284e-06,
+      "loss": 8.5062,
+      "step": 191790
+    },
+    {
+      "epoch": 0.9578266623386352,
+      "grad_norm": 0.08770589530467987,
+      "learning_rate": 1.2683671681393777e-06,
+      "loss": 8.485,
+      "step": 191800
+    },
+    {
+      "epoch": 0.9578766011635746,
+      "grad_norm": 0.08947274088859558,
+      "learning_rate": 1.2668652531978272e-06,
+      "loss": 8.4993,
+      "step": 191810
+    },
+    {
+      "epoch": 0.9579265399885141,
+      "grad_norm": 0.0907110795378685,
+      "learning_rate": 1.2653633382562768e-06,
+      "loss": 8.5106,
+      "step": 191820
+    },
+    {
+      "epoch": 0.9579764788134535,
+      "grad_norm": 0.08664959669113159,
+      "learning_rate": 1.2638614233147263e-06,
+      "loss": 8.4997,
+      "step": 191830
+    },
+    {
+      "epoch": 0.9580264176383929,
+      "grad_norm": 0.09159454703330994,
+      "learning_rate": 1.2623595083731759e-06,
+      "loss": 8.5125,
+      "step": 191840
+    },
+    {
+      "epoch": 0.9580763564633324,
+      "grad_norm": 0.09214504808187485,
+      "learning_rate": 1.2608575934316254e-06,
+      "loss": 8.4985,
+      "step": 191850
+    },
+    {
+      "epoch": 0.9581262952882719,
+      "grad_norm": 0.09041109681129456,
+      "learning_rate": 1.2593556784900747e-06,
+      "loss": 8.4857,
+      "step": 191860
+    },
+    {
+      "epoch": 0.9581762341132113,
+      "grad_norm": 0.09253381192684174,
+      "learning_rate": 1.2578537635485245e-06,
+      "loss": 8.4968,
+      "step": 191870
+    },
+    {
+      "epoch": 0.9582261729381507,
+      "grad_norm": 0.08627009391784668,
+      "learning_rate": 1.2563518486069738e-06,
+      "loss": 8.4892,
+      "step": 191880
+    },
+    {
+      "epoch": 0.9582761117630902,
+      "grad_norm": 0.08717307448387146,
+      "learning_rate": 1.2548499336654236e-06,
+      "loss": 8.4982,
+      "step": 191890
+    },
+    {
+      "epoch": 0.9583260505880297,
+      "grad_norm": 0.08845217525959015,
+      "learning_rate": 1.253348018723873e-06,
+      "loss": 8.4926,
+      "step": 191900
+    },
+    {
+      "epoch": 0.9583759894129691,
+      "grad_norm": 0.09016247093677521,
+      "learning_rate": 1.2518461037823225e-06,
+      "loss": 8.4968,
+      "step": 191910
+    },
+    {
+      "epoch": 0.9584259282379085,
+      "grad_norm": 0.08623816072940826,
+      "learning_rate": 1.250344188840772e-06,
+      "loss": 8.4969,
+      "step": 191920
+    },
+    {
+      "epoch": 0.958475867062848,
+      "grad_norm": 0.0954393595457077,
+      "learning_rate": 1.2488422738992216e-06,
+      "loss": 8.517,
+      "step": 191930
+    },
+    {
+      "epoch": 0.9585258058877875,
+      "grad_norm": 0.09252654761075974,
+      "learning_rate": 1.247340358957671e-06,
+      "loss": 8.4971,
+      "step": 191940
+    },
+    {
+      "epoch": 0.9585757447127269,
+      "grad_norm": 0.08835329860448837,
+      "learning_rate": 1.2458384440161206e-06,
+      "loss": 8.5138,
+      "step": 191950
+    },
+    {
+      "epoch": 0.9586256835376663,
+      "grad_norm": 0.08202789723873138,
+      "learning_rate": 1.24433652907457e-06,
+      "loss": 8.4846,
+      "step": 191960
+    },
+    {
+      "epoch": 0.9586756223626058,
+      "grad_norm": 0.09334851056337357,
+      "learning_rate": 1.2428346141330197e-06,
+      "loss": 8.5079,
+      "step": 191970
+    },
+    {
+      "epoch": 0.9587255611875453,
+      "grad_norm": 0.08949562907218933,
+      "learning_rate": 1.241332699191469e-06,
+      "loss": 8.4951,
+      "step": 191980
+    },
+    {
+      "epoch": 0.9587755000124847,
+      "grad_norm": 0.09301643818616867,
+      "learning_rate": 1.2398307842499188e-06,
+      "loss": 8.5035,
+      "step": 191990
+    },
+    {
+      "epoch": 0.9588254388374241,
+      "grad_norm": 0.08926796913146973,
+      "learning_rate": 1.2383288693083681e-06,
+      "loss": 8.5123,
+      "step": 192000
+    },
+    {
+      "epoch": 0.9588753776623636,
+      "grad_norm": 0.09037186205387115,
+      "learning_rate": 1.236826954366818e-06,
+      "loss": 8.5027,
+      "step": 192010
+    },
+    {
+      "epoch": 0.9589253164873031,
+      "grad_norm": 0.09102381020784378,
+      "learning_rate": 1.2353250394252672e-06,
+      "loss": 8.5107,
+      "step": 192020
+    },
+    {
+      "epoch": 0.9589752553122425,
+      "grad_norm": 0.08897027373313904,
+      "learning_rate": 1.2338231244837168e-06,
+      "loss": 8.5083,
+      "step": 192030
+    },
+    {
+      "epoch": 0.9590251941371819,
+      "grad_norm": 0.08504099398851395,
+      "learning_rate": 1.2323212095421663e-06,
+      "loss": 8.5118,
+      "step": 192040
+    },
+    {
+      "epoch": 0.9590751329621214,
+      "grad_norm": 0.09035465121269226,
+      "learning_rate": 1.2308192946006159e-06,
+      "loss": 8.502,
+      "step": 192050
+    },
+    {
+      "epoch": 0.9591250717870609,
+      "grad_norm": 0.0969761461019516,
+      "learning_rate": 1.2293173796590654e-06,
+      "loss": 8.5028,
+      "step": 192060
+    },
+    {
+      "epoch": 0.9591750106120003,
+      "grad_norm": 0.0866863876581192,
+      "learning_rate": 1.227815464717515e-06,
+      "loss": 8.5155,
+      "step": 192070
+    },
+    {
+      "epoch": 0.9592249494369397,
+      "grad_norm": 0.08988199383020401,
+      "learning_rate": 1.2263135497759643e-06,
+      "loss": 8.4989,
+      "step": 192080
+    },
+    {
+      "epoch": 0.9592748882618792,
+      "grad_norm": 0.09326018393039703,
+      "learning_rate": 1.224811634834414e-06,
+      "loss": 8.5031,
+      "step": 192090
+    },
+    {
+      "epoch": 0.9593248270868187,
+      "grad_norm": 0.0929257795214653,
+      "learning_rate": 1.2233097198928634e-06,
+      "loss": 8.4906,
+      "step": 192100
+    },
+    {
+      "epoch": 0.9593747659117581,
+      "grad_norm": 0.09300316870212555,
+      "learning_rate": 1.2218078049513131e-06,
+      "loss": 8.5025,
+      "step": 192110
+    },
+    {
+      "epoch": 0.9594247047366975,
+      "grad_norm": 0.09423361718654633,
+      "learning_rate": 1.2203058900097624e-06,
+      "loss": 8.5001,
+      "step": 192120
+    },
+    {
+      "epoch": 0.959474643561637,
+      "grad_norm": 0.09319435060024261,
+      "learning_rate": 1.218803975068212e-06,
+      "loss": 8.4935,
+      "step": 192130
+    },
+    {
+      "epoch": 0.9595245823865765,
+      "grad_norm": 0.09030423313379288,
+      "learning_rate": 1.2173020601266615e-06,
+      "loss": 8.4956,
+      "step": 192140
+    },
+    {
+      "epoch": 0.9595745212115159,
+      "grad_norm": 0.08913615345954895,
+      "learning_rate": 1.215800145185111e-06,
+      "loss": 8.5366,
+      "step": 192150
+    },
+    {
+      "epoch": 0.9596244600364553,
+      "grad_norm": 0.09200779348611832,
+      "learning_rate": 1.2142982302435606e-06,
+      "loss": 8.5005,
+      "step": 192160
+    },
+    {
+      "epoch": 0.9596743988613948,
+      "grad_norm": 0.08741243183612823,
+      "learning_rate": 1.2127963153020102e-06,
+      "loss": 8.4892,
+      "step": 192170
+    },
+    {
+      "epoch": 0.9597243376863343,
+      "grad_norm": 0.09448469430208206,
+      "learning_rate": 1.2112944003604595e-06,
+      "loss": 8.4926,
+      "step": 192180
+    },
+    {
+      "epoch": 0.9597742765112737,
+      "grad_norm": 0.0888211578130722,
+      "learning_rate": 1.2097924854189092e-06,
+      "loss": 8.4874,
+      "step": 192190
+    },
+    {
+      "epoch": 0.9598242153362131,
+      "grad_norm": 0.0948527380824089,
+      "learning_rate": 1.2082905704773586e-06,
+      "loss": 8.5015,
+      "step": 192200
+    },
+    {
+      "epoch": 0.9598741541611526,
+      "grad_norm": 0.09242790937423706,
+      "learning_rate": 1.2067886555358083e-06,
+      "loss": 8.4939,
+      "step": 192210
+    },
+    {
+      "epoch": 0.959924092986092,
+      "grad_norm": 0.09047332406044006,
+      "learning_rate": 1.2052867405942577e-06,
+      "loss": 8.4966,
+      "step": 192220
+    },
+    {
+      "epoch": 0.9599740318110315,
+      "grad_norm": 0.09905733168125153,
+      "learning_rate": 1.2037848256527072e-06,
+      "loss": 8.5018,
+      "step": 192230
+    },
+    {
+      "epoch": 0.9600239706359709,
+      "grad_norm": 0.09126092493534088,
+      "learning_rate": 1.2022829107111568e-06,
+      "loss": 8.5119,
+      "step": 192240
+    },
+    {
+      "epoch": 0.9600739094609104,
+      "grad_norm": 0.09281742572784424,
+      "learning_rate": 1.2007809957696063e-06,
+      "loss": 8.5087,
+      "step": 192250
+    },
+    {
+      "epoch": 0.9601238482858498,
+      "grad_norm": 0.09464964270591736,
+      "learning_rate": 1.1992790808280558e-06,
+      "loss": 8.505,
+      "step": 192260
+    },
+    {
+      "epoch": 0.9601737871107893,
+      "grad_norm": 0.09487392008304596,
+      "learning_rate": 1.1977771658865054e-06,
+      "loss": 8.4981,
+      "step": 192270
+    },
+    {
+      "epoch": 0.9602237259357287,
+      "grad_norm": 0.09476413577795029,
+      "learning_rate": 1.196275250944955e-06,
+      "loss": 8.4889,
+      "step": 192280
+    },
+    {
+      "epoch": 0.9602736647606682,
+      "grad_norm": 0.09175974875688553,
+      "learning_rate": 1.1947733360034045e-06,
+      "loss": 8.4984,
+      "step": 192290
+    },
+    {
+      "epoch": 0.9603236035856076,
+      "grad_norm": 0.09201910346746445,
+      "learning_rate": 1.1932714210618538e-06,
+      "loss": 8.4981,
+      "step": 192300
+    },
+    {
+      "epoch": 0.9603735424105471,
+      "grad_norm": 0.0915134996175766,
+      "learning_rate": 1.1917695061203036e-06,
+      "loss": 8.5039,
+      "step": 192310
+    },
+    {
+      "epoch": 0.9604234812354865,
+      "grad_norm": 0.08767704665660858,
+      "learning_rate": 1.1902675911787529e-06,
+      "loss": 8.5014,
+      "step": 192320
+    },
+    {
+      "epoch": 0.960473420060426,
+      "grad_norm": 0.0916207805275917,
+      "learning_rate": 1.1887656762372026e-06,
+      "loss": 8.5048,
+      "step": 192330
+    },
+    {
+      "epoch": 0.9605233588853654,
+      "grad_norm": 0.09325848519802094,
+      "learning_rate": 1.187263761295652e-06,
+      "loss": 8.5005,
+      "step": 192340
+    },
+    {
+      "epoch": 0.9605732977103049,
+      "grad_norm": 0.08897262811660767,
+      "learning_rate": 1.1857618463541015e-06,
+      "loss": 8.5037,
+      "step": 192350
+    },
+    {
+      "epoch": 0.9606232365352443,
+      "grad_norm": 0.08885764330625534,
+      "learning_rate": 1.184259931412551e-06,
+      "loss": 8.5089,
+      "step": 192360
+    },
+    {
+      "epoch": 0.9606731753601838,
+      "grad_norm": 0.085177481174469,
+      "learning_rate": 1.1827580164710006e-06,
+      "loss": 8.5113,
+      "step": 192370
+    },
+    {
+      "epoch": 0.9607231141851232,
+      "grad_norm": 0.08419841527938843,
+      "learning_rate": 1.1812561015294501e-06,
+      "loss": 8.48,
+      "step": 192380
+    },
+    {
+      "epoch": 0.9607730530100627,
+      "grad_norm": 0.08771277964115143,
+      "learning_rate": 1.1797541865878997e-06,
+      "loss": 8.5046,
+      "step": 192390
+    },
+    {
+      "epoch": 0.9608229918350021,
+      "grad_norm": 0.09421413391828537,
+      "learning_rate": 1.178252271646349e-06,
+      "loss": 8.479,
+      "step": 192400
+    },
+    {
+      "epoch": 0.9608729306599416,
+      "grad_norm": 0.09432334452867508,
+      "learning_rate": 1.1767503567047988e-06,
+      "loss": 8.4848,
+      "step": 192410
+    },
+    {
+      "epoch": 0.960922869484881,
+      "grad_norm": 0.09068858623504639,
+      "learning_rate": 1.175248441763248e-06,
+      "loss": 8.5012,
+      "step": 192420
+    },
+    {
+      "epoch": 0.9609728083098205,
+      "grad_norm": 0.09556064754724503,
+      "learning_rate": 1.1737465268216979e-06,
+      "loss": 8.5111,
+      "step": 192430
+    },
+    {
+      "epoch": 0.9610227471347599,
+      "grad_norm": 0.08801019191741943,
+      "learning_rate": 1.1722446118801472e-06,
+      "loss": 8.4905,
+      "step": 192440
+    },
+    {
+      "epoch": 0.9610726859596994,
+      "grad_norm": 0.09224274009466171,
+      "learning_rate": 1.1707426969385967e-06,
+      "loss": 8.4729,
+      "step": 192450
+    },
+    {
+      "epoch": 0.9611226247846388,
+      "grad_norm": 0.08931384235620499,
+      "learning_rate": 1.1692407819970463e-06,
+      "loss": 8.5073,
+      "step": 192460
+    },
+    {
+      "epoch": 0.9611725636095783,
+      "grad_norm": 0.09076393395662308,
+      "learning_rate": 1.1677388670554958e-06,
+      "loss": 8.4992,
+      "step": 192470
+    },
+    {
+      "epoch": 0.9612225024345177,
+      "grad_norm": 0.08941461890935898,
+      "learning_rate": 1.1662369521139454e-06,
+      "loss": 8.5019,
+      "step": 192480
+    },
+    {
+      "epoch": 0.9612724412594572,
+      "grad_norm": 0.08615435659885406,
+      "learning_rate": 1.164735037172395e-06,
+      "loss": 8.4891,
+      "step": 192490
+    },
+    {
+      "epoch": 0.9613223800843966,
+      "grad_norm": 0.08833363652229309,
+      "learning_rate": 1.1632331222308442e-06,
+      "loss": 8.5133,
+      "step": 192500
+    },
+    {
+      "epoch": 0.961372318909336,
+      "grad_norm": 0.09384980797767639,
+      "learning_rate": 1.161731207289294e-06,
+      "loss": 8.5135,
+      "step": 192510
+    },
+    {
+      "epoch": 0.9614222577342755,
+      "grad_norm": 0.09676310420036316,
+      "learning_rate": 1.1602292923477433e-06,
+      "loss": 8.4996,
+      "step": 192520
+    },
+    {
+      "epoch": 0.961472196559215,
+      "grad_norm": 0.09096337109804153,
+      "learning_rate": 1.158727377406193e-06,
+      "loss": 8.4903,
+      "step": 192530
+    },
+    {
+      "epoch": 0.9615221353841544,
+      "grad_norm": 0.09308335185050964,
+      "learning_rate": 1.1572254624646424e-06,
+      "loss": 8.5021,
+      "step": 192540
+    },
+    {
+      "epoch": 0.9615720742090939,
+      "grad_norm": 0.0906360074877739,
+      "learning_rate": 1.1557235475230922e-06,
+      "loss": 8.5127,
+      "step": 192550
+    },
+    {
+      "epoch": 0.9616220130340333,
+      "grad_norm": 0.09376540780067444,
+      "learning_rate": 1.1542216325815415e-06,
+      "loss": 8.5081,
+      "step": 192560
+    },
+    {
+      "epoch": 0.9616719518589728,
+      "grad_norm": 0.09033221006393433,
+      "learning_rate": 1.152719717639991e-06,
+      "loss": 8.5068,
+      "step": 192570
+    },
+    {
+      "epoch": 0.9617218906839122,
+      "grad_norm": 0.0966610535979271,
+      "learning_rate": 1.1512178026984406e-06,
+      "loss": 8.4919,
+      "step": 192580
+    },
+    {
+      "epoch": 0.9617718295088517,
+      "grad_norm": 0.09105295687913895,
+      "learning_rate": 1.1497158877568901e-06,
+      "loss": 8.5013,
+      "step": 192590
+    },
+    {
+      "epoch": 0.9618217683337911,
+      "grad_norm": 0.08788570016622543,
+      "learning_rate": 1.1482139728153397e-06,
+      "loss": 8.4736,
+      "step": 192600
+    },
+    {
+      "epoch": 0.9618717071587306,
+      "grad_norm": 0.09203591197729111,
+      "learning_rate": 1.1467120578737892e-06,
+      "loss": 8.493,
+      "step": 192610
+    },
+    {
+      "epoch": 0.96192164598367,
+      "grad_norm": 0.09675536304712296,
+      "learning_rate": 1.1452101429322385e-06,
+      "loss": 8.5179,
+      "step": 192620
+    },
+    {
+      "epoch": 0.9619715848086094,
+      "grad_norm": 0.09318403899669647,
+      "learning_rate": 1.1437082279906883e-06,
+      "loss": 8.5007,
+      "step": 192630
+    },
+    {
+      "epoch": 0.9620215236335489,
+      "grad_norm": 0.09018708020448685,
+      "learning_rate": 1.1422063130491376e-06,
+      "loss": 8.4895,
+      "step": 192640
+    },
+    {
+      "epoch": 0.9620714624584884,
+      "grad_norm": 0.09459145367145538,
+      "learning_rate": 1.1407043981075874e-06,
+      "loss": 8.5098,
+      "step": 192650
+    },
+    {
+      "epoch": 0.9621214012834278,
+      "grad_norm": 0.09730951488018036,
+      "learning_rate": 1.1392024831660367e-06,
+      "loss": 8.5015,
+      "step": 192660
+    },
+    {
+      "epoch": 0.9621713401083672,
+      "grad_norm": 0.09757350385189056,
+      "learning_rate": 1.1377005682244863e-06,
+      "loss": 8.5212,
+      "step": 192670
+    },
+    {
+      "epoch": 0.9622212789333067,
+      "grad_norm": 0.08909895271062851,
+      "learning_rate": 1.1361986532829358e-06,
+      "loss": 8.508,
+      "step": 192680
+    },
+    {
+      "epoch": 0.9622712177582462,
+      "grad_norm": 0.09438234567642212,
+      "learning_rate": 1.1346967383413853e-06,
+      "loss": 8.4956,
+      "step": 192690
+    },
+    {
+      "epoch": 0.9623211565831856,
+      "grad_norm": 0.08857671916484833,
+      "learning_rate": 1.1331948233998349e-06,
+      "loss": 8.5006,
+      "step": 192700
+    },
+    {
+      "epoch": 0.962371095408125,
+      "grad_norm": 0.0914953425526619,
+      "learning_rate": 1.1316929084582844e-06,
+      "loss": 8.5024,
+      "step": 192710
+    },
+    {
+      "epoch": 0.9624210342330645,
+      "grad_norm": 0.08889735490083694,
+      "learning_rate": 1.1301909935167338e-06,
+      "loss": 8.5009,
+      "step": 192720
+    },
+    {
+      "epoch": 0.962470973058004,
+      "grad_norm": 0.09225352108478546,
+      "learning_rate": 1.1286890785751835e-06,
+      "loss": 8.5067,
+      "step": 192730
+    },
+    {
+      "epoch": 0.9625209118829434,
+      "grad_norm": 0.09539781510829926,
+      "learning_rate": 1.1271871636336328e-06,
+      "loss": 8.5049,
+      "step": 192740
+    },
+    {
+      "epoch": 0.9625708507078828,
+      "grad_norm": 0.0937805026769638,
+      "learning_rate": 1.1256852486920826e-06,
+      "loss": 8.5277,
+      "step": 192750
+    },
+    {
+      "epoch": 0.9626207895328223,
+      "grad_norm": 0.08874644339084625,
+      "learning_rate": 1.124183333750532e-06,
+      "loss": 8.5063,
+      "step": 192760
+    },
+    {
+      "epoch": 0.9626707283577618,
+      "grad_norm": 0.0926111564040184,
+      "learning_rate": 1.1226814188089815e-06,
+      "loss": 8.495,
+      "step": 192770
+    },
+    {
+      "epoch": 0.9627206671827012,
+      "grad_norm": 0.09661145508289337,
+      "learning_rate": 1.121179503867431e-06,
+      "loss": 8.5039,
+      "step": 192780
+    },
+    {
+      "epoch": 0.9627706060076406,
+      "grad_norm": 0.09471116960048676,
+      "learning_rate": 1.1196775889258806e-06,
+      "loss": 8.4877,
+      "step": 192790
+    },
+    {
+      "epoch": 0.9628205448325801,
+      "grad_norm": 0.09448757022619247,
+      "learning_rate": 1.11817567398433e-06,
+      "loss": 8.4881,
+      "step": 192800
+    },
+    {
+      "epoch": 0.9628704836575196,
+      "grad_norm": 0.09081356972455978,
+      "learning_rate": 1.1166737590427796e-06,
+      "loss": 8.5098,
+      "step": 192810
+    },
+    {
+      "epoch": 0.962920422482459,
+      "grad_norm": 0.08594221621751785,
+      "learning_rate": 1.1151718441012292e-06,
+      "loss": 8.51,
+      "step": 192820
+    },
+    {
+      "epoch": 0.9629703613073984,
+      "grad_norm": 0.08855900168418884,
+      "learning_rate": 1.1136699291596787e-06,
+      "loss": 8.5147,
+      "step": 192830
+    },
+    {
+      "epoch": 0.9630203001323379,
+      "grad_norm": 0.09176114946603775,
+      "learning_rate": 1.112168014218128e-06,
+      "loss": 8.5023,
+      "step": 192840
+    },
+    {
+      "epoch": 0.9630702389572773,
+      "grad_norm": 0.09412090480327606,
+      "learning_rate": 1.1106660992765778e-06,
+      "loss": 8.5108,
+      "step": 192850
+    },
+    {
+      "epoch": 0.9631201777822168,
+      "grad_norm": 0.0874660536646843,
+      "learning_rate": 1.1091641843350272e-06,
+      "loss": 8.5087,
+      "step": 192860
+    },
+    {
+      "epoch": 0.9631701166071562,
+      "grad_norm": 0.08917557448148727,
+      "learning_rate": 1.107662269393477e-06,
+      "loss": 8.502,
+      "step": 192870
+    },
+    {
+      "epoch": 0.9632200554320957,
+      "grad_norm": 0.09307212382555008,
+      "learning_rate": 1.1061603544519262e-06,
+      "loss": 8.5,
+      "step": 192880
+    },
+    {
+      "epoch": 0.9632699942570351,
+      "grad_norm": 0.0892016664147377,
+      "learning_rate": 1.1046584395103758e-06,
+      "loss": 8.4979,
+      "step": 192890
+    },
+    {
+      "epoch": 0.9633199330819746,
+      "grad_norm": 0.08102846145629883,
+      "learning_rate": 1.1031565245688253e-06,
+      "loss": 8.511,
+      "step": 192900
+    },
+    {
+      "epoch": 0.963369871906914,
+      "grad_norm": 0.09051341563463211,
+      "learning_rate": 1.1016546096272749e-06,
+      "loss": 8.5003,
+      "step": 192910
+    },
+    {
+      "epoch": 0.9634198107318535,
+      "grad_norm": 0.09337877482175827,
+      "learning_rate": 1.1001526946857244e-06,
+      "loss": 8.5054,
+      "step": 192920
+    },
+    {
+      "epoch": 0.9634697495567929,
+      "grad_norm": 0.0894298255443573,
+      "learning_rate": 1.098650779744174e-06,
+      "loss": 8.5,
+      "step": 192930
+    },
+    {
+      "epoch": 0.9635196883817324,
+      "grad_norm": 0.09010350704193115,
+      "learning_rate": 1.0971488648026233e-06,
+      "loss": 8.5062,
+      "step": 192940
+    },
+    {
+      "epoch": 0.9635696272066718,
+      "grad_norm": 0.08480501919984818,
+      "learning_rate": 1.095646949861073e-06,
+      "loss": 8.5046,
+      "step": 192950
+    },
+    {
+      "epoch": 0.9636195660316113,
+      "grad_norm": 0.09088391810655594,
+      "learning_rate": 1.0941450349195224e-06,
+      "loss": 8.4948,
+      "step": 192960
+    },
+    {
+      "epoch": 0.9636695048565507,
+      "grad_norm": 0.09037703275680542,
+      "learning_rate": 1.092643119977972e-06,
+      "loss": 8.4929,
+      "step": 192970
+    },
+    {
+      "epoch": 0.9637194436814902,
+      "grad_norm": 0.08855674415826797,
+      "learning_rate": 1.0911412050364215e-06,
+      "loss": 8.508,
+      "step": 192980
+    },
+    {
+      "epoch": 0.9637693825064296,
+      "grad_norm": 0.08638061583042145,
+      "learning_rate": 1.0896392900948708e-06,
+      "loss": 8.5144,
+      "step": 192990
+    },
+    {
+      "epoch": 0.963819321331369,
+      "grad_norm": 0.09214214235544205,
+      "learning_rate": 1.0881373751533205e-06,
+      "loss": 8.4967,
+      "step": 193000
+    },
+    {
+      "epoch": 0.9638692601563085,
+      "grad_norm": 0.09081536531448364,
+      "learning_rate": 1.0866354602117699e-06,
+      "loss": 8.5044,
+      "step": 193010
+    },
+    {
+      "epoch": 0.963919198981248,
+      "grad_norm": 0.09031979739665985,
+      "learning_rate": 1.0851335452702196e-06,
+      "loss": 8.5013,
+      "step": 193020
+    },
+    {
+      "epoch": 0.9639691378061874,
+      "grad_norm": 0.08983229100704193,
+      "learning_rate": 1.083631630328669e-06,
+      "loss": 8.4903,
+      "step": 193030
+    },
+    {
+      "epoch": 0.9640190766311268,
+      "grad_norm": 0.08677353709936142,
+      "learning_rate": 1.0821297153871185e-06,
+      "loss": 8.5,
+      "step": 193040
+    },
+    {
+      "epoch": 0.9640690154560663,
+      "grad_norm": 0.09196817129850388,
+      "learning_rate": 1.080627800445568e-06,
+      "loss": 8.508,
+      "step": 193050
+    },
+    {
+      "epoch": 0.9641189542810058,
+      "grad_norm": 0.09112430363893509,
+      "learning_rate": 1.0791258855040176e-06,
+      "loss": 8.5005,
+      "step": 193060
+    },
+    {
+      "epoch": 0.9641688931059452,
+      "grad_norm": 0.09055691957473755,
+      "learning_rate": 1.0776239705624671e-06,
+      "loss": 8.4864,
+      "step": 193070
+    },
+    {
+      "epoch": 0.9642188319308846,
+      "grad_norm": 0.09631075710058212,
+      "learning_rate": 1.0761220556209167e-06,
+      "loss": 8.5011,
+      "step": 193080
+    },
+    {
+      "epoch": 0.9642687707558241,
+      "grad_norm": 0.091032475233078,
+      "learning_rate": 1.0746201406793662e-06,
+      "loss": 8.4968,
+      "step": 193090
+    },
+    {
+      "epoch": 0.9643187095807636,
+      "grad_norm": 0.09456492960453033,
+      "learning_rate": 1.0731182257378158e-06,
+      "loss": 8.5076,
+      "step": 193100
+    },
+    {
+      "epoch": 0.964368648405703,
+      "grad_norm": 0.09016124159097672,
+      "learning_rate": 1.071616310796265e-06,
+      "loss": 8.5205,
+      "step": 193110
+    },
+    {
+      "epoch": 0.9644185872306424,
+      "grad_norm": 0.09598606824874878,
+      "learning_rate": 1.0701143958547148e-06,
+      "loss": 8.483,
+      "step": 193120
+    },
+    {
+      "epoch": 0.9644685260555819,
+      "grad_norm": 0.08811622112989426,
+      "learning_rate": 1.0686124809131642e-06,
+      "loss": 8.4907,
+      "step": 193130
+    },
+    {
+      "epoch": 0.9645184648805214,
+      "grad_norm": 0.0883399173617363,
+      "learning_rate": 1.067110565971614e-06,
+      "loss": 8.5047,
+      "step": 193140
+    },
+    {
+      "epoch": 0.9645684037054608,
+      "grad_norm": 0.0905977189540863,
+      "learning_rate": 1.0656086510300633e-06,
+      "loss": 8.4943,
+      "step": 193150
+    },
+    {
+      "epoch": 0.9646183425304002,
+      "grad_norm": 0.09106041491031647,
+      "learning_rate": 1.0641067360885128e-06,
+      "loss": 8.5189,
+      "step": 193160
+    },
+    {
+      "epoch": 0.9646682813553397,
+      "grad_norm": 0.08970261365175247,
+      "learning_rate": 1.0626048211469624e-06,
+      "loss": 8.5162,
+      "step": 193170
+    },
+    {
+      "epoch": 0.9647182201802792,
+      "grad_norm": 0.08954989910125732,
+      "learning_rate": 1.061102906205412e-06,
+      "loss": 8.4932,
+      "step": 193180
+    },
+    {
+      "epoch": 0.9647681590052186,
+      "grad_norm": 0.08976080268621445,
+      "learning_rate": 1.0596009912638614e-06,
+      "loss": 8.4927,
+      "step": 193190
+    },
+    {
+      "epoch": 0.964818097830158,
+      "grad_norm": 0.09053555130958557,
+      "learning_rate": 1.058099076322311e-06,
+      "loss": 8.4998,
+      "step": 193200
+    },
+    {
+      "epoch": 0.9648680366550975,
+      "grad_norm": 0.09599043428897858,
+      "learning_rate": 1.0565971613807603e-06,
+      "loss": 8.5194,
+      "step": 193210
+    },
+    {
+      "epoch": 0.964917975480037,
+      "grad_norm": 0.09338483214378357,
+      "learning_rate": 1.05509524643921e-06,
+      "loss": 8.4981,
+      "step": 193220
+    },
+    {
+      "epoch": 0.9649679143049764,
+      "grad_norm": 0.09092381596565247,
+      "learning_rate": 1.0535933314976594e-06,
+      "loss": 8.5404,
+      "step": 193230
+    },
+    {
+      "epoch": 0.9650178531299158,
+      "grad_norm": 0.0899573415517807,
+      "learning_rate": 1.0520914165561092e-06,
+      "loss": 8.5163,
+      "step": 193240
+    },
+    {
+      "epoch": 0.9650677919548553,
+      "grad_norm": 0.09256323426961899,
+      "learning_rate": 1.0505895016145585e-06,
+      "loss": 8.4776,
+      "step": 193250
+    },
+    {
+      "epoch": 0.9651177307797948,
+      "grad_norm": 0.08706999570131302,
+      "learning_rate": 1.049087586673008e-06,
+      "loss": 8.487,
+      "step": 193260
+    },
+    {
+      "epoch": 0.9651676696047342,
+      "grad_norm": 0.09370410442352295,
+      "learning_rate": 1.0475856717314576e-06,
+      "loss": 8.5015,
+      "step": 193270
+    },
+    {
+      "epoch": 0.9652176084296736,
+      "grad_norm": 0.09068583697080612,
+      "learning_rate": 1.0460837567899071e-06,
+      "loss": 8.4835,
+      "step": 193280
+    },
+    {
+      "epoch": 0.9652675472546131,
+      "grad_norm": 0.09026405960321426,
+      "learning_rate": 1.0445818418483567e-06,
+      "loss": 8.4975,
+      "step": 193290
+    },
+    {
+      "epoch": 0.9653174860795526,
+      "grad_norm": 0.0878196433186531,
+      "learning_rate": 1.0430799269068062e-06,
+      "loss": 8.4881,
+      "step": 193300
+    },
+    {
+      "epoch": 0.965367424904492,
+      "grad_norm": 0.08937402069568634,
+      "learning_rate": 1.0415780119652555e-06,
+      "loss": 8.497,
+      "step": 193310
+    },
+    {
+      "epoch": 0.9654173637294314,
+      "grad_norm": 0.09177373349666595,
+      "learning_rate": 1.0400760970237053e-06,
+      "loss": 8.4949,
+      "step": 193320
+    },
+    {
+      "epoch": 0.9654673025543709,
+      "grad_norm": 0.09359072893857956,
+      "learning_rate": 1.0385741820821546e-06,
+      "loss": 8.5039,
+      "step": 193330
+    },
+    {
+      "epoch": 0.9655172413793104,
+      "grad_norm": 0.09130553156137466,
+      "learning_rate": 1.0370722671406044e-06,
+      "loss": 8.489,
+      "step": 193340
+    },
+    {
+      "epoch": 0.9655671802042498,
+      "grad_norm": 0.09181410074234009,
+      "learning_rate": 1.0355703521990537e-06,
+      "loss": 8.5007,
+      "step": 193350
+    },
+    {
+      "epoch": 0.9656171190291892,
+      "grad_norm": 0.09474316984415054,
+      "learning_rate": 1.0340684372575035e-06,
+      "loss": 8.5026,
+      "step": 193360
+    },
+    {
+      "epoch": 0.9656670578541287,
+      "grad_norm": 0.09248656779527664,
+      "learning_rate": 1.0325665223159528e-06,
+      "loss": 8.5032,
+      "step": 193370
+    },
+    {
+      "epoch": 0.9657169966790682,
+      "grad_norm": 0.09201966971158981,
+      "learning_rate": 1.0310646073744023e-06,
+      "loss": 8.5016,
+      "step": 193380
+    },
+    {
+      "epoch": 0.9657669355040076,
+      "grad_norm": 0.09128998965024948,
+      "learning_rate": 1.0295626924328519e-06,
+      "loss": 8.4941,
+      "step": 193390
+    },
+    {
+      "epoch": 0.965816874328947,
+      "grad_norm": 0.08449841290712357,
+      "learning_rate": 1.0280607774913014e-06,
+      "loss": 8.5381,
+      "step": 193400
+    },
+    {
+      "epoch": 0.9658668131538864,
+      "grad_norm": 0.09496010094881058,
+      "learning_rate": 1.026558862549751e-06,
+      "loss": 8.5214,
+      "step": 193410
+    },
+    {
+      "epoch": 0.965916751978826,
+      "grad_norm": 0.08874188363552094,
+      "learning_rate": 1.0250569476082005e-06,
+      "loss": 8.512,
+      "step": 193420
+    },
+    {
+      "epoch": 0.9659666908037654,
+      "grad_norm": 0.08817078918218613,
+      "learning_rate": 1.0235550326666498e-06,
+      "loss": 8.4784,
+      "step": 193430
+    },
+    {
+      "epoch": 0.9660166296287048,
+      "grad_norm": 0.09461791068315506,
+      "learning_rate": 1.0220531177250996e-06,
+      "loss": 8.4886,
+      "step": 193440
+    },
+    {
+      "epoch": 0.9660665684536442,
+      "grad_norm": 0.09889914095401764,
+      "learning_rate": 1.020551202783549e-06,
+      "loss": 8.506,
+      "step": 193450
+    },
+    {
+      "epoch": 0.9661165072785838,
+      "grad_norm": 0.09483031183481216,
+      "learning_rate": 1.0190492878419987e-06,
+      "loss": 8.5026,
+      "step": 193460
+    },
+    {
+      "epoch": 0.9661664461035232,
+      "grad_norm": 0.09063508361577988,
+      "learning_rate": 1.017547372900448e-06,
+      "loss": 8.4843,
+      "step": 193470
+    },
+    {
+      "epoch": 0.9662163849284626,
+      "grad_norm": 0.09315601736307144,
+      "learning_rate": 1.0160454579588976e-06,
+      "loss": 8.4879,
+      "step": 193480
+    },
+    {
+      "epoch": 0.966266323753402,
+      "grad_norm": 0.09240519255399704,
+      "learning_rate": 1.014543543017347e-06,
+      "loss": 8.4912,
+      "step": 193490
+    },
+    {
+      "epoch": 0.9663162625783416,
+      "grad_norm": 0.09266150742769241,
+      "learning_rate": 1.0130416280757966e-06,
+      "loss": 8.5122,
+      "step": 193500
+    },
+    {
+      "epoch": 0.966366201403281,
+      "grad_norm": 0.08804892003536224,
+      "learning_rate": 1.0115397131342462e-06,
+      "loss": 8.498,
+      "step": 193510
+    },
+    {
+      "epoch": 0.9664161402282204,
+      "grad_norm": 0.0914304181933403,
+      "learning_rate": 1.0100377981926957e-06,
+      "loss": 8.5013,
+      "step": 193520
+    },
+    {
+      "epoch": 0.9664660790531598,
+      "grad_norm": 0.08695808053016663,
+      "learning_rate": 1.008535883251145e-06,
+      "loss": 8.5127,
+      "step": 193530
+    },
+    {
+      "epoch": 0.9665160178780994,
+      "grad_norm": 0.09132501482963562,
+      "learning_rate": 1.0070339683095948e-06,
+      "loss": 8.4965,
+      "step": 193540
+    },
+    {
+      "epoch": 0.9665659567030388,
+      "grad_norm": 0.08505229651927948,
+      "learning_rate": 1.0055320533680441e-06,
+      "loss": 8.5024,
+      "step": 193550
+    },
+    {
+      "epoch": 0.9666158955279782,
+      "grad_norm": 0.09099217504262924,
+      "learning_rate": 1.004030138426494e-06,
+      "loss": 8.5031,
+      "step": 193560
+    },
+    {
+      "epoch": 0.9666658343529176,
+      "grad_norm": 0.09543818980455399,
+      "learning_rate": 1.0025282234849432e-06,
+      "loss": 8.497,
+      "step": 193570
+    },
+    {
+      "epoch": 0.9667157731778572,
+      "grad_norm": 0.09109151363372803,
+      "learning_rate": 1.0010263085433928e-06,
+      "loss": 8.4983,
+      "step": 193580
+    },
+    {
+      "epoch": 0.9667657120027966,
+      "grad_norm": 0.09007328748703003,
+      "learning_rate": 9.995243936018423e-07,
+      "loss": 8.5035,
+      "step": 193590
+    },
+    {
+      "epoch": 0.966815650827736,
+      "grad_norm": 0.09476472437381744,
+      "learning_rate": 9.980224786602919e-07,
+      "loss": 8.4901,
+      "step": 193600
+    },
+    {
+      "epoch": 0.9668655896526754,
+      "grad_norm": 0.09251927584409714,
+      "learning_rate": 9.965205637187414e-07,
+      "loss": 8.5123,
+      "step": 193610
+    },
+    {
+      "epoch": 0.966915528477615,
+      "grad_norm": 0.08989694714546204,
+      "learning_rate": 9.95018648777191e-07,
+      "loss": 8.5028,
+      "step": 193620
+    },
+    {
+      "epoch": 0.9669654673025544,
+      "grad_norm": 0.09202085435390472,
+      "learning_rate": 9.935167338356405e-07,
+      "loss": 8.5208,
+      "step": 193630
+    },
+    {
+      "epoch": 0.9670154061274938,
+      "grad_norm": 0.09178955852985382,
+      "learning_rate": 9.9201481889409e-07,
+      "loss": 8.5038,
+      "step": 193640
+    },
+    {
+      "epoch": 0.9670653449524332,
+      "grad_norm": 0.08983652293682098,
+      "learning_rate": 9.905129039525394e-07,
+      "loss": 8.4953,
+      "step": 193650
+    },
+    {
+      "epoch": 0.9671152837773728,
+      "grad_norm": 0.08969952166080475,
+      "learning_rate": 9.890109890109891e-07,
+      "loss": 8.5043,
+      "step": 193660
+    },
+    {
+      "epoch": 0.9671652226023122,
+      "grad_norm": 0.09150168299674988,
+      "learning_rate": 9.875090740694384e-07,
+      "loss": 8.5058,
+      "step": 193670
+    },
+    {
+      "epoch": 0.9672151614272516,
+      "grad_norm": 0.09063497930765152,
+      "learning_rate": 9.860071591278882e-07,
+      "loss": 8.5009,
+      "step": 193680
+    },
+    {
+      "epoch": 0.967265100252191,
+      "grad_norm": 0.08984601497650146,
+      "learning_rate": 9.845052441863375e-07,
+      "loss": 8.5082,
+      "step": 193690
+    },
+    {
+      "epoch": 0.9673150390771306,
+      "grad_norm": 0.09148889034986496,
+      "learning_rate": 9.83003329244787e-07,
+      "loss": 8.5017,
+      "step": 193700
+    },
+    {
+      "epoch": 0.96736497790207,
+      "grad_norm": 0.09478826075792313,
+      "learning_rate": 9.815014143032366e-07,
+      "loss": 8.4963,
+      "step": 193710
+    },
+    {
+      "epoch": 0.9674149167270094,
+      "grad_norm": 0.0906151756644249,
+      "learning_rate": 9.799994993616862e-07,
+      "loss": 8.4947,
+      "step": 193720
+    },
+    {
+      "epoch": 0.9674648555519488,
+      "grad_norm": 0.0921965166926384,
+      "learning_rate": 9.784975844201357e-07,
+      "loss": 8.5126,
+      "step": 193730
+    },
+    {
+      "epoch": 0.9675147943768884,
+      "grad_norm": 0.09627021849155426,
+      "learning_rate": 9.769956694785852e-07,
+      "loss": 8.4987,
+      "step": 193740
+    },
+    {
+      "epoch": 0.9675647332018278,
+      "grad_norm": 0.0912567749619484,
+      "learning_rate": 9.754937545370346e-07,
+      "loss": 8.5012,
+      "step": 193750
+    },
+    {
+      "epoch": 0.9676146720267672,
+      "grad_norm": 0.09098992496728897,
+      "learning_rate": 9.739918395954843e-07,
+      "loss": 8.4986,
+      "step": 193760
+    },
+    {
+      "epoch": 0.9676646108517066,
+      "grad_norm": 0.09001246094703674,
+      "learning_rate": 9.724899246539337e-07,
+      "loss": 8.4776,
+      "step": 193770
+    },
+    {
+      "epoch": 0.9677145496766462,
+      "grad_norm": 0.09463299065828323,
+      "learning_rate": 9.709880097123834e-07,
+      "loss": 8.4844,
+      "step": 193780
+    },
+    {
+      "epoch": 0.9677644885015856,
+      "grad_norm": 0.09130155295133591,
+      "learning_rate": 9.694860947708328e-07,
+      "loss": 8.4935,
+      "step": 193790
+    },
+    {
+      "epoch": 0.967814427326525,
+      "grad_norm": 0.09506887197494507,
+      "learning_rate": 9.679841798292823e-07,
+      "loss": 8.5098,
+      "step": 193800
+    },
+    {
+      "epoch": 0.9678643661514644,
+      "grad_norm": 0.0893334150314331,
+      "learning_rate": 9.664822648877318e-07,
+      "loss": 8.4957,
+      "step": 193810
+    },
+    {
+      "epoch": 0.9679143049764038,
+      "grad_norm": 0.08439191430807114,
+      "learning_rate": 9.649803499461814e-07,
+      "loss": 8.5174,
+      "step": 193820
+    },
+    {
+      "epoch": 0.9679642438013434,
+      "grad_norm": 0.0962853729724884,
+      "learning_rate": 9.63478435004631e-07,
+      "loss": 8.5103,
+      "step": 193830
+    },
+    {
+      "epoch": 0.9680141826262828,
+      "grad_norm": 0.09273120760917664,
+      "learning_rate": 9.619765200630805e-07,
+      "loss": 8.4903,
+      "step": 193840
+    },
+    {
+      "epoch": 0.9680641214512222,
+      "grad_norm": 0.08646199852228165,
+      "learning_rate": 9.604746051215298e-07,
+      "loss": 8.4982,
+      "step": 193850
+    },
+    {
+      "epoch": 0.9681140602761616,
+      "grad_norm": 0.08911348879337311,
+      "learning_rate": 9.589726901799796e-07,
+      "loss": 8.4981,
+      "step": 193860
+    },
+    {
+      "epoch": 0.9681639991011012,
+      "grad_norm": 0.09012620151042938,
+      "learning_rate": 9.574707752384289e-07,
+      "loss": 8.4926,
+      "step": 193870
+    },
+    {
+      "epoch": 0.9682139379260406,
+      "grad_norm": 0.08890789747238159,
+      "learning_rate": 9.559688602968786e-07,
+      "loss": 8.4956,
+      "step": 193880
+    },
+    {
+      "epoch": 0.96826387675098,
+      "grad_norm": 0.0956655889749527,
+      "learning_rate": 9.54466945355328e-07,
+      "loss": 8.5063,
+      "step": 193890
+    },
+    {
+      "epoch": 0.9683138155759194,
+      "grad_norm": 0.0944046676158905,
+      "learning_rate": 9.529650304137776e-07,
+      "loss": 8.4913,
+      "step": 193900
+    },
+    {
+      "epoch": 0.968363754400859,
+      "grad_norm": 0.09098576009273529,
+      "learning_rate": 9.514631154722271e-07,
+      "loss": 8.4847,
+      "step": 193910
+    },
+    {
+      "epoch": 0.9684136932257984,
+      "grad_norm": 0.09032325446605682,
+      "learning_rate": 9.499612005306767e-07,
+      "loss": 8.5017,
+      "step": 193920
+    },
+    {
+      "epoch": 0.9684636320507378,
+      "grad_norm": 0.08898947387933731,
+      "learning_rate": 9.484592855891261e-07,
+      "loss": 8.4929,
+      "step": 193930
+    },
+    {
+      "epoch": 0.9685135708756772,
+      "grad_norm": 0.087337426841259,
+      "learning_rate": 9.469573706475757e-07,
+      "loss": 8.5065,
+      "step": 193940
+    },
+    {
+      "epoch": 0.9685635097006168,
+      "grad_norm": 0.089985191822052,
+      "learning_rate": 9.454554557060251e-07,
+      "loss": 8.4867,
+      "step": 193950
+    },
+    {
+      "epoch": 0.9686134485255562,
+      "grad_norm": 0.09129071980714798,
+      "learning_rate": 9.439535407644748e-07,
+      "loss": 8.5096,
+      "step": 193960
+    },
+    {
+      "epoch": 0.9686633873504956,
+      "grad_norm": 0.08964653313159943,
+      "learning_rate": 9.424516258229242e-07,
+      "loss": 8.5179,
+      "step": 193970
+    },
+    {
+      "epoch": 0.968713326175435,
+      "grad_norm": 0.09389611333608627,
+      "learning_rate": 9.409497108813739e-07,
+      "loss": 8.4891,
+      "step": 193980
+    },
+    {
+      "epoch": 0.9687632650003746,
+      "grad_norm": 0.0910143330693245,
+      "learning_rate": 9.394477959398232e-07,
+      "loss": 8.5042,
+      "step": 193990
+    },
+    {
+      "epoch": 0.968813203825314,
+      "grad_norm": 0.09279486536979675,
+      "learning_rate": 9.379458809982728e-07,
+      "loss": 8.5063,
+      "step": 194000
+    },
+    {
+      "epoch": 0.9688631426502534,
+      "grad_norm": 0.09352979063987732,
+      "learning_rate": 9.364439660567224e-07,
+      "loss": 8.5034,
+      "step": 194010
+    },
+    {
+      "epoch": 0.9689130814751928,
+      "grad_norm": 0.09566289931535721,
+      "learning_rate": 9.349420511151719e-07,
+      "loss": 8.4996,
+      "step": 194020
+    },
+    {
+      "epoch": 0.9689630203001324,
+      "grad_norm": 0.09447862952947617,
+      "learning_rate": 9.334401361736215e-07,
+      "loss": 8.4978,
+      "step": 194030
+    },
+    {
+      "epoch": 0.9690129591250718,
+      "grad_norm": 0.09513899683952332,
+      "learning_rate": 9.319382212320709e-07,
+      "loss": 8.5037,
+      "step": 194040
+    },
+    {
+      "epoch": 0.9690628979500112,
+      "grad_norm": 0.09640028327703476,
+      "learning_rate": 9.304363062905204e-07,
+      "loss": 8.5095,
+      "step": 194050
+    },
+    {
+      "epoch": 0.9691128367749506,
+      "grad_norm": 0.09304415434598923,
+      "learning_rate": 9.2893439134897e-07,
+      "loss": 8.4915,
+      "step": 194060
+    },
+    {
+      "epoch": 0.9691627755998902,
+      "grad_norm": 0.08626173436641693,
+      "learning_rate": 9.274324764074195e-07,
+      "loss": 8.5022,
+      "step": 194070
+    },
+    {
+      "epoch": 0.9692127144248296,
+      "grad_norm": 0.08531943708658218,
+      "learning_rate": 9.259305614658691e-07,
+      "loss": 8.5094,
+      "step": 194080
+    },
+    {
+      "epoch": 0.969262653249769,
+      "grad_norm": 0.08645452558994293,
+      "learning_rate": 9.244286465243186e-07,
+      "loss": 8.4834,
+      "step": 194090
+    },
+    {
+      "epoch": 0.9693125920747084,
+      "grad_norm": 0.08829811960458755,
+      "learning_rate": 9.229267315827681e-07,
+      "loss": 8.4885,
+      "step": 194100
+    },
+    {
+      "epoch": 0.969362530899648,
+      "grad_norm": 0.08838041871786118,
+      "learning_rate": 9.214248166412176e-07,
+      "loss": 8.4871,
+      "step": 194110
+    },
+    {
+      "epoch": 0.9694124697245874,
+      "grad_norm": 0.10451626777648926,
+      "learning_rate": 9.199229016996671e-07,
+      "loss": 8.5013,
+      "step": 194120
+    },
+    {
+      "epoch": 0.9694624085495268,
+      "grad_norm": 0.0896938219666481,
+      "learning_rate": 9.184209867581167e-07,
+      "loss": 8.4912,
+      "step": 194130
+    },
+    {
+      "epoch": 0.9695123473744662,
+      "grad_norm": 0.09267926216125488,
+      "learning_rate": 9.169190718165662e-07,
+      "loss": 8.5015,
+      "step": 194140
+    },
+    {
+      "epoch": 0.9695622861994058,
+      "grad_norm": 0.09529542177915573,
+      "learning_rate": 9.154171568750157e-07,
+      "loss": 8.5052,
+      "step": 194150
+    },
+    {
+      "epoch": 0.9696122250243452,
+      "grad_norm": 0.09126924723386765,
+      "learning_rate": 9.139152419334652e-07,
+      "loss": 8.4985,
+      "step": 194160
+    },
+    {
+      "epoch": 0.9696621638492846,
+      "grad_norm": 0.09080453962087631,
+      "learning_rate": 9.124133269919148e-07,
+      "loss": 8.4891,
+      "step": 194170
+    },
+    {
+      "epoch": 0.969712102674224,
+      "grad_norm": 0.09079496562480927,
+      "learning_rate": 9.109114120503643e-07,
+      "loss": 8.4894,
+      "step": 194180
+    },
+    {
+      "epoch": 0.9697620414991636,
+      "grad_norm": 0.09417524933815002,
+      "learning_rate": 9.094094971088138e-07,
+      "loss": 8.5045,
+      "step": 194190
+    },
+    {
+      "epoch": 0.969811980324103,
+      "grad_norm": 0.09117228537797928,
+      "learning_rate": 9.079075821672633e-07,
+      "loss": 8.5076,
+      "step": 194200
+    },
+    {
+      "epoch": 0.9698619191490424,
+      "grad_norm": 0.09321728348731995,
+      "learning_rate": 9.064056672257128e-07,
+      "loss": 8.4965,
+      "step": 194210
+    },
+    {
+      "epoch": 0.9699118579739818,
+      "grad_norm": 0.08753789961338043,
+      "learning_rate": 9.049037522841624e-07,
+      "loss": 8.4978,
+      "step": 194220
+    },
+    {
+      "epoch": 0.9699617967989214,
+      "grad_norm": 0.09341299533843994,
+      "learning_rate": 9.034018373426119e-07,
+      "loss": 8.4969,
+      "step": 194230
+    },
+    {
+      "epoch": 0.9700117356238608,
+      "grad_norm": 0.09108027070760727,
+      "learning_rate": 9.018999224010614e-07,
+      "loss": 8.4958,
+      "step": 194240
+    },
+    {
+      "epoch": 0.9700616744488002,
+      "grad_norm": 0.09051679074764252,
+      "learning_rate": 9.00398007459511e-07,
+      "loss": 8.5203,
+      "step": 194250
+    },
+    {
+      "epoch": 0.9701116132737396,
+      "grad_norm": 0.0874452218413353,
+      "learning_rate": 8.988960925179604e-07,
+      "loss": 8.4992,
+      "step": 194260
+    },
+    {
+      "epoch": 0.9701615520986792,
+      "grad_norm": 0.09082052111625671,
+      "learning_rate": 8.9739417757641e-07,
+      "loss": 8.5051,
+      "step": 194270
+    },
+    {
+      "epoch": 0.9702114909236186,
+      "grad_norm": 0.08856025338172913,
+      "learning_rate": 8.958922626348595e-07,
+      "loss": 8.5033,
+      "step": 194280
+    },
+    {
+      "epoch": 0.970261429748558,
+      "grad_norm": 0.09108460694551468,
+      "learning_rate": 8.943903476933091e-07,
+      "loss": 8.5058,
+      "step": 194290
+    },
+    {
+      "epoch": 0.9703113685734974,
+      "grad_norm": 0.09241397678852081,
+      "learning_rate": 8.928884327517586e-07,
+      "loss": 8.5089,
+      "step": 194300
+    },
+    {
+      "epoch": 0.970361307398437,
+      "grad_norm": 0.099297434091568,
+      "learning_rate": 8.91386517810208e-07,
+      "loss": 8.4976,
+      "step": 194310
+    },
+    {
+      "epoch": 0.9704112462233764,
+      "grad_norm": 0.08700823783874512,
+      "learning_rate": 8.898846028686576e-07,
+      "loss": 8.4843,
+      "step": 194320
+    },
+    {
+      "epoch": 0.9704611850483158,
+      "grad_norm": 0.09079594165086746,
+      "learning_rate": 8.883826879271071e-07,
+      "loss": 8.5041,
+      "step": 194330
+    },
+    {
+      "epoch": 0.9705111238732552,
+      "grad_norm": 0.08681687712669373,
+      "learning_rate": 8.868807729855567e-07,
+      "loss": 8.4892,
+      "step": 194340
+    },
+    {
+      "epoch": 0.9705610626981948,
+      "grad_norm": 0.09553930163383484,
+      "learning_rate": 8.853788580440062e-07,
+      "loss": 8.5019,
+      "step": 194350
+    },
+    {
+      "epoch": 0.9706110015231342,
+      "grad_norm": 0.08355355262756348,
+      "learning_rate": 8.838769431024558e-07,
+      "loss": 8.5108,
+      "step": 194360
+    },
+    {
+      "epoch": 0.9706609403480736,
+      "grad_norm": 0.0913529172539711,
+      "learning_rate": 8.823750281609052e-07,
+      "loss": 8.5125,
+      "step": 194370
+    },
+    {
+      "epoch": 0.970710879173013,
+      "grad_norm": 0.09661257266998291,
+      "learning_rate": 8.808731132193547e-07,
+      "loss": 8.5045,
+      "step": 194380
+    },
+    {
+      "epoch": 0.9707608179979526,
+      "grad_norm": 0.09083771705627441,
+      "learning_rate": 8.793711982778043e-07,
+      "loss": 8.4961,
+      "step": 194390
+    },
+    {
+      "epoch": 0.970810756822892,
+      "grad_norm": 0.08930204063653946,
+      "learning_rate": 8.778692833362538e-07,
+      "loss": 8.509,
+      "step": 194400
+    },
+    {
+      "epoch": 0.9708606956478314,
+      "grad_norm": 0.08711707592010498,
+      "learning_rate": 8.763673683947034e-07,
+      "loss": 8.5015,
+      "step": 194410
+    },
+    {
+      "epoch": 0.9709106344727708,
+      "grad_norm": 0.09438509494066238,
+      "learning_rate": 8.748654534531527e-07,
+      "loss": 8.4954,
+      "step": 194420
+    },
+    {
+      "epoch": 0.9709605732977103,
+      "grad_norm": 0.09597020596265793,
+      "learning_rate": 8.733635385116022e-07,
+      "loss": 8.4873,
+      "step": 194430
+    },
+    {
+      "epoch": 0.9710105121226498,
+      "grad_norm": 0.09720340371131897,
+      "learning_rate": 8.718616235700518e-07,
+      "loss": 8.5018,
+      "step": 194440
+    },
+    {
+      "epoch": 0.9710604509475892,
+      "grad_norm": 0.09392903745174408,
+      "learning_rate": 8.703597086285013e-07,
+      "loss": 8.498,
+      "step": 194450
+    },
+    {
+      "epoch": 0.9711103897725286,
+      "grad_norm": 0.08631488680839539,
+      "learning_rate": 8.688577936869509e-07,
+      "loss": 8.5162,
+      "step": 194460
+    },
+    {
+      "epoch": 0.9711603285974681,
+      "grad_norm": 0.0900026261806488,
+      "learning_rate": 8.673558787454003e-07,
+      "loss": 8.4875,
+      "step": 194470
+    },
+    {
+      "epoch": 0.9712102674224076,
+      "grad_norm": 0.09366903454065323,
+      "learning_rate": 8.658539638038498e-07,
+      "loss": 8.4926,
+      "step": 194480
+    },
+    {
+      "epoch": 0.971260206247347,
+      "grad_norm": 0.08986707031726837,
+      "learning_rate": 8.643520488622994e-07,
+      "loss": 8.5027,
+      "step": 194490
+    },
+    {
+      "epoch": 0.9713101450722864,
+      "grad_norm": 0.08996418118476868,
+      "learning_rate": 8.628501339207489e-07,
+      "loss": 8.4983,
+      "step": 194500
+    },
+    {
+      "epoch": 0.971360083897226,
+      "grad_norm": 0.09488117694854736,
+      "learning_rate": 8.613482189791985e-07,
+      "loss": 8.4869,
+      "step": 194510
+    },
+    {
+      "epoch": 0.9714100227221654,
+      "grad_norm": 0.0917017012834549,
+      "learning_rate": 8.59846304037648e-07,
+      "loss": 8.5092,
+      "step": 194520
+    },
+    {
+      "epoch": 0.9714599615471048,
+      "grad_norm": 0.0919492170214653,
+      "learning_rate": 8.583443890960975e-07,
+      "loss": 8.4992,
+      "step": 194530
+    },
+    {
+      "epoch": 0.9715099003720442,
+      "grad_norm": 0.08751703053712845,
+      "learning_rate": 8.56842474154547e-07,
+      "loss": 8.5088,
+      "step": 194540
+    },
+    {
+      "epoch": 0.9715598391969837,
+      "grad_norm": 0.09287269413471222,
+      "learning_rate": 8.553405592129965e-07,
+      "loss": 8.496,
+      "step": 194550
+    },
+    {
+      "epoch": 0.9716097780219232,
+      "grad_norm": 0.08905121684074402,
+      "learning_rate": 8.538386442714461e-07,
+      "loss": 8.5047,
+      "step": 194560
+    },
+    {
+      "epoch": 0.9716597168468626,
+      "grad_norm": 0.0883074626326561,
+      "learning_rate": 8.523367293298956e-07,
+      "loss": 8.4918,
+      "step": 194570
+    },
+    {
+      "epoch": 0.971709655671802,
+      "grad_norm": 0.08838310837745667,
+      "learning_rate": 8.508348143883451e-07,
+      "loss": 8.4834,
+      "step": 194580
+    },
+    {
+      "epoch": 0.9717595944967415,
+      "grad_norm": 0.08658456802368164,
+      "learning_rate": 8.493328994467946e-07,
+      "loss": 8.4993,
+      "step": 194590
+    },
+    {
+      "epoch": 0.971809533321681,
+      "grad_norm": 0.08917684108018875,
+      "learning_rate": 8.478309845052442e-07,
+      "loss": 8.5031,
+      "step": 194600
+    },
+    {
+      "epoch": 0.9718594721466204,
+      "grad_norm": 0.09120479226112366,
+      "learning_rate": 8.463290695636937e-07,
+      "loss": 8.495,
+      "step": 194610
+    },
+    {
+      "epoch": 0.9719094109715598,
+      "grad_norm": 0.09076827019453049,
+      "learning_rate": 8.448271546221432e-07,
+      "loss": 8.5126,
+      "step": 194620
+    },
+    {
+      "epoch": 0.9719593497964993,
+      "grad_norm": 0.09772755205631256,
+      "learning_rate": 8.433252396805928e-07,
+      "loss": 8.4906,
+      "step": 194630
+    },
+    {
+      "epoch": 0.9720092886214388,
+      "grad_norm": 0.08785592764616013,
+      "learning_rate": 8.418233247390422e-07,
+      "loss": 8.5035,
+      "step": 194640
+    },
+    {
+      "epoch": 0.9720592274463782,
+      "grad_norm": 0.09105689078569412,
+      "learning_rate": 8.403214097974918e-07,
+      "loss": 8.4977,
+      "step": 194650
+    },
+    {
+      "epoch": 0.9721091662713176,
+      "grad_norm": 0.09146515280008316,
+      "learning_rate": 8.388194948559413e-07,
+      "loss": 8.5028,
+      "step": 194660
+    },
+    {
+      "epoch": 0.9721591050962571,
+      "grad_norm": 0.09533209353685379,
+      "learning_rate": 8.373175799143908e-07,
+      "loss": 8.4955,
+      "step": 194670
+    },
+    {
+      "epoch": 0.9722090439211966,
+      "grad_norm": 0.08939139544963837,
+      "learning_rate": 8.358156649728404e-07,
+      "loss": 8.4918,
+      "step": 194680
+    },
+    {
+      "epoch": 0.972258982746136,
+      "grad_norm": 0.0908391997218132,
+      "learning_rate": 8.343137500312898e-07,
+      "loss": 8.5214,
+      "step": 194690
+    },
+    {
+      "epoch": 0.9723089215710754,
+      "grad_norm": 0.08793427795171738,
+      "learning_rate": 8.328118350897394e-07,
+      "loss": 8.5005,
+      "step": 194700
+    },
+    {
+      "epoch": 0.9723588603960149,
+      "grad_norm": 0.09640876948833466,
+      "learning_rate": 8.313099201481889e-07,
+      "loss": 8.5052,
+      "step": 194710
+    },
+    {
+      "epoch": 0.9724087992209544,
+      "grad_norm": 0.0828898623585701,
+      "learning_rate": 8.298080052066385e-07,
+      "loss": 8.4938,
+      "step": 194720
+    },
+    {
+      "epoch": 0.9724587380458938,
+      "grad_norm": 0.09573552757501602,
+      "learning_rate": 8.28306090265088e-07,
+      "loss": 8.5073,
+      "step": 194730
+    },
+    {
+      "epoch": 0.9725086768708332,
+      "grad_norm": 0.09013304859399796,
+      "learning_rate": 8.268041753235374e-07,
+      "loss": 8.4836,
+      "step": 194740
+    },
+    {
+      "epoch": 0.9725586156957727,
+      "grad_norm": 0.08847355097532272,
+      "learning_rate": 8.25302260381987e-07,
+      "loss": 8.4966,
+      "step": 194750
+    },
+    {
+      "epoch": 0.9726085545207122,
+      "grad_norm": 0.0923420786857605,
+      "learning_rate": 8.238003454404365e-07,
+      "loss": 8.4948,
+      "step": 194760
+    },
+    {
+      "epoch": 0.9726584933456516,
+      "grad_norm": 0.09254893660545349,
+      "learning_rate": 8.222984304988861e-07,
+      "loss": 8.5117,
+      "step": 194770
+    },
+    {
+      "epoch": 0.972708432170591,
+      "grad_norm": 0.09156614542007446,
+      "learning_rate": 8.207965155573356e-07,
+      "loss": 8.4967,
+      "step": 194780
+    },
+    {
+      "epoch": 0.9727583709955305,
+      "grad_norm": 0.09188225865364075,
+      "learning_rate": 8.192946006157852e-07,
+      "loss": 8.4973,
+      "step": 194790
+    },
+    {
+      "epoch": 0.97280830982047,
+      "grad_norm": 0.09390117973089218,
+      "learning_rate": 8.177926856742346e-07,
+      "loss": 8.4866,
+      "step": 194800
+    },
+    {
+      "epoch": 0.9728582486454094,
+      "grad_norm": 0.09112615883350372,
+      "learning_rate": 8.162907707326841e-07,
+      "loss": 8.5038,
+      "step": 194810
+    },
+    {
+      "epoch": 0.9729081874703488,
+      "grad_norm": 0.08958408236503601,
+      "learning_rate": 8.147888557911337e-07,
+      "loss": 8.5039,
+      "step": 194820
+    },
+    {
+      "epoch": 0.9729581262952882,
+      "grad_norm": 0.09056022018194199,
+      "learning_rate": 8.132869408495832e-07,
+      "loss": 8.5089,
+      "step": 194830
+    },
+    {
+      "epoch": 0.9730080651202277,
+      "grad_norm": 0.0902940034866333,
+      "learning_rate": 8.117850259080328e-07,
+      "loss": 8.4841,
+      "step": 194840
+    },
+    {
+      "epoch": 0.9730580039451672,
+      "grad_norm": 0.08953052759170532,
+      "learning_rate": 8.102831109664822e-07,
+      "loss": 8.5099,
+      "step": 194850
+    },
+    {
+      "epoch": 0.9731079427701066,
+      "grad_norm": 0.09145776927471161,
+      "learning_rate": 8.087811960249317e-07,
+      "loss": 8.4842,
+      "step": 194860
+    },
+    {
+      "epoch": 0.973157881595046,
+      "grad_norm": 0.09231266379356384,
+      "learning_rate": 8.072792810833813e-07,
+      "loss": 8.4961,
+      "step": 194870
+    },
+    {
+      "epoch": 0.9732078204199855,
+      "grad_norm": 0.09086192399263382,
+      "learning_rate": 8.057773661418308e-07,
+      "loss": 8.5026,
+      "step": 194880
+    },
+    {
+      "epoch": 0.973257759244925,
+      "grad_norm": 0.0860838070511818,
+      "learning_rate": 8.042754512002804e-07,
+      "loss": 8.4974,
+      "step": 194890
+    },
+    {
+      "epoch": 0.9733076980698644,
+      "grad_norm": 0.09107068926095963,
+      "learning_rate": 8.027735362587299e-07,
+      "loss": 8.509,
+      "step": 194900
+    },
+    {
+      "epoch": 0.9733576368948038,
+      "grad_norm": 0.09526333957910538,
+      "learning_rate": 8.012716213171794e-07,
+      "loss": 8.4865,
+      "step": 194910
+    },
+    {
+      "epoch": 0.9734075757197433,
+      "grad_norm": 0.09245917946100235,
+      "learning_rate": 7.997697063756289e-07,
+      "loss": 8.5036,
+      "step": 194920
+    },
+    {
+      "epoch": 0.9734575145446828,
+      "grad_norm": 0.09295765310525894,
+      "learning_rate": 7.982677914340784e-07,
+      "loss": 8.4988,
+      "step": 194930
+    },
+    {
+      "epoch": 0.9735074533696222,
+      "grad_norm": 0.08823594450950623,
+      "learning_rate": 7.96765876492528e-07,
+      "loss": 8.4804,
+      "step": 194940
+    },
+    {
+      "epoch": 0.9735573921945616,
+      "grad_norm": 0.08996216952800751,
+      "learning_rate": 7.952639615509775e-07,
+      "loss": 8.5019,
+      "step": 194950
+    },
+    {
+      "epoch": 0.9736073310195011,
+      "grad_norm": 0.08816206455230713,
+      "learning_rate": 7.93762046609427e-07,
+      "loss": 8.5015,
+      "step": 194960
+    },
+    {
+      "epoch": 0.9736572698444406,
+      "grad_norm": 0.09345098584890366,
+      "learning_rate": 7.922601316678765e-07,
+      "loss": 8.4941,
+      "step": 194970
+    },
+    {
+      "epoch": 0.97370720866938,
+      "grad_norm": 0.08701256662607193,
+      "learning_rate": 7.90758216726326e-07,
+      "loss": 8.4983,
+      "step": 194980
+    },
+    {
+      "epoch": 0.9737571474943194,
+      "grad_norm": 0.08897973597049713,
+      "learning_rate": 7.892563017847756e-07,
+      "loss": 8.5118,
+      "step": 194990
+    },
+    {
+      "epoch": 0.9738070863192589,
+      "grad_norm": 0.08834607154130936,
+      "learning_rate": 7.877543868432251e-07,
+      "loss": 8.5255,
+      "step": 195000
+    },
+    {
+      "epoch": 0.9738570251441984,
+      "grad_norm": 0.09231365472078323,
+      "learning_rate": 7.862524719016746e-07,
+      "loss": 8.5062,
+      "step": 195010
+    },
+    {
+      "epoch": 0.9739069639691378,
+      "grad_norm": 0.089638851583004,
+      "learning_rate": 7.847505569601241e-07,
+      "loss": 8.4869,
+      "step": 195020
+    },
+    {
+      "epoch": 0.9739569027940772,
+      "grad_norm": 0.09642373770475388,
+      "learning_rate": 7.832486420185737e-07,
+      "loss": 8.5094,
+      "step": 195030
+    },
+    {
+      "epoch": 0.9740068416190167,
+      "grad_norm": 0.09365835040807724,
+      "learning_rate": 7.817467270770232e-07,
+      "loss": 8.5179,
+      "step": 195040
+    },
+    {
+      "epoch": 0.9740567804439562,
+      "grad_norm": 0.09444354474544525,
+      "learning_rate": 7.802448121354727e-07,
+      "loss": 8.5114,
+      "step": 195050
+    },
+    {
+      "epoch": 0.9741067192688956,
+      "grad_norm": 0.09574456512928009,
+      "learning_rate": 7.787428971939223e-07,
+      "loss": 8.4882,
+      "step": 195060
+    },
+    {
+      "epoch": 0.974156658093835,
+      "grad_norm": 0.09401273727416992,
+      "learning_rate": 7.772409822523717e-07,
+      "loss": 8.4983,
+      "step": 195070
+    },
+    {
+      "epoch": 0.9742065969187745,
+      "grad_norm": 0.08458282053470612,
+      "learning_rate": 7.757390673108213e-07,
+      "loss": 8.4861,
+      "step": 195080
+    },
+    {
+      "epoch": 0.974256535743714,
+      "grad_norm": 0.09678689390420914,
+      "learning_rate": 7.742371523692708e-07,
+      "loss": 8.4971,
+      "step": 195090
+    },
+    {
+      "epoch": 0.9743064745686534,
+      "grad_norm": 0.08890896290540695,
+      "learning_rate": 7.727352374277204e-07,
+      "loss": 8.4881,
+      "step": 195100
+    },
+    {
+      "epoch": 0.9743564133935928,
+      "grad_norm": 0.0906500592827797,
+      "learning_rate": 7.712333224861699e-07,
+      "loss": 8.4991,
+      "step": 195110
+    },
+    {
+      "epoch": 0.9744063522185323,
+      "grad_norm": 0.09543705731630325,
+      "learning_rate": 7.697314075446193e-07,
+      "loss": 8.5054,
+      "step": 195120
+    },
+    {
+      "epoch": 0.9744562910434718,
+      "grad_norm": 0.0905163511633873,
+      "learning_rate": 7.682294926030689e-07,
+      "loss": 8.4979,
+      "step": 195130
+    },
+    {
+      "epoch": 0.9745062298684112,
+      "grad_norm": 0.09297658503055573,
+      "learning_rate": 7.667275776615184e-07,
+      "loss": 8.4858,
+      "step": 195140
+    },
+    {
+      "epoch": 0.9745561686933506,
+      "grad_norm": 0.08763201534748077,
+      "learning_rate": 7.65225662719968e-07,
+      "loss": 8.4979,
+      "step": 195150
+    },
+    {
+      "epoch": 0.9746061075182901,
+      "grad_norm": 0.0941351130604744,
+      "learning_rate": 7.637237477784175e-07,
+      "loss": 8.5004,
+      "step": 195160
+    },
+    {
+      "epoch": 0.9746560463432296,
+      "grad_norm": 0.09358199685811996,
+      "learning_rate": 7.62221832836867e-07,
+      "loss": 8.4789,
+      "step": 195170
+    },
+    {
+      "epoch": 0.974705985168169,
+      "grad_norm": 0.09558181464672089,
+      "learning_rate": 7.607199178953165e-07,
+      "loss": 8.5046,
+      "step": 195180
+    },
+    {
+      "epoch": 0.9747559239931084,
+      "grad_norm": 0.08955127000808716,
+      "learning_rate": 7.59218002953766e-07,
+      "loss": 8.5107,
+      "step": 195190
+    },
+    {
+      "epoch": 0.9748058628180479,
+      "grad_norm": 0.08842689543962479,
+      "learning_rate": 7.577160880122156e-07,
+      "loss": 8.4977,
+      "step": 195200
+    },
+    {
+      "epoch": 0.9748558016429874,
+      "grad_norm": 0.08837328106164932,
+      "learning_rate": 7.562141730706651e-07,
+      "loss": 8.4808,
+      "step": 195210
+    },
+    {
+      "epoch": 0.9749057404679268,
+      "grad_norm": 0.09056874364614487,
+      "learning_rate": 7.547122581291147e-07,
+      "loss": 8.5065,
+      "step": 195220
+    },
+    {
+      "epoch": 0.9749556792928662,
+      "grad_norm": 0.0923481434583664,
+      "learning_rate": 7.532103431875641e-07,
+      "loss": 8.4953,
+      "step": 195230
+    },
+    {
+      "epoch": 0.9750056181178057,
+      "grad_norm": 0.09200780093669891,
+      "learning_rate": 7.517084282460136e-07,
+      "loss": 8.5131,
+      "step": 195240
+    },
+    {
+      "epoch": 0.9750555569427451,
+      "grad_norm": 0.09253337979316711,
+      "learning_rate": 7.502065133044632e-07,
+      "loss": 8.503,
+      "step": 195250
+    },
+    {
+      "epoch": 0.9751054957676846,
+      "grad_norm": 0.09560063481330872,
+      "learning_rate": 7.487045983629127e-07,
+      "loss": 8.4932,
+      "step": 195260
+    },
+    {
+      "epoch": 0.975155434592624,
+      "grad_norm": 0.09416811168193817,
+      "learning_rate": 7.472026834213623e-07,
+      "loss": 8.5141,
+      "step": 195270
+    },
+    {
+      "epoch": 0.9752053734175635,
+      "grad_norm": 0.08733347803354263,
+      "learning_rate": 7.457007684798117e-07,
+      "loss": 8.5092,
+      "step": 195280
+    },
+    {
+      "epoch": 0.975255312242503,
+      "grad_norm": 0.08784154802560806,
+      "learning_rate": 7.441988535382612e-07,
+      "loss": 8.5029,
+      "step": 195290
+    },
+    {
+      "epoch": 0.9753052510674424,
+      "grad_norm": 0.09478521347045898,
+      "learning_rate": 7.426969385967108e-07,
+      "loss": 8.5035,
+      "step": 195300
+    },
+    {
+      "epoch": 0.9753551898923818,
+      "grad_norm": 0.09163037687540054,
+      "learning_rate": 7.411950236551603e-07,
+      "loss": 8.4888,
+      "step": 195310
+    },
+    {
+      "epoch": 0.9754051287173213,
+      "grad_norm": 0.10056457668542862,
+      "learning_rate": 7.396931087136099e-07,
+      "loss": 8.5046,
+      "step": 195320
+    },
+    {
+      "epoch": 0.9754550675422607,
+      "grad_norm": 0.09030482172966003,
+      "learning_rate": 7.381911937720594e-07,
+      "loss": 8.505,
+      "step": 195330
+    },
+    {
+      "epoch": 0.9755050063672002,
+      "grad_norm": 0.089601531624794,
+      "learning_rate": 7.366892788305089e-07,
+      "loss": 8.4949,
+      "step": 195340
+    },
+    {
+      "epoch": 0.9755549451921396,
+      "grad_norm": 0.09470442682504654,
+      "learning_rate": 7.351873638889584e-07,
+      "loss": 8.4826,
+      "step": 195350
+    },
+    {
+      "epoch": 0.9756048840170791,
+      "grad_norm": 0.0969916582107544,
+      "learning_rate": 7.336854489474079e-07,
+      "loss": 8.4949,
+      "step": 195360
+    },
+    {
+      "epoch": 0.9756548228420185,
+      "grad_norm": 0.09550881385803223,
+      "learning_rate": 7.321835340058575e-07,
+      "loss": 8.4897,
+      "step": 195370
+    },
+    {
+      "epoch": 0.975704761666958,
+      "grad_norm": 0.08882039040327072,
+      "learning_rate": 7.30681619064307e-07,
+      "loss": 8.4957,
+      "step": 195380
+    },
+    {
+      "epoch": 0.9757547004918974,
+      "grad_norm": 0.09256576001644135,
+      "learning_rate": 7.291797041227565e-07,
+      "loss": 8.5088,
+      "step": 195390
+    },
+    {
+      "epoch": 0.9758046393168369,
+      "grad_norm": 0.09013333171606064,
+      "learning_rate": 7.27677789181206e-07,
+      "loss": 8.5056,
+      "step": 195400
+    },
+    {
+      "epoch": 0.9758545781417763,
+      "grad_norm": 0.09338297694921494,
+      "learning_rate": 7.261758742396556e-07,
+      "loss": 8.4914,
+      "step": 195410
+    },
+    {
+      "epoch": 0.9759045169667158,
+      "grad_norm": 0.08930335938930511,
+      "learning_rate": 7.246739592981051e-07,
+      "loss": 8.4948,
+      "step": 195420
+    },
+    {
+      "epoch": 0.9759544557916552,
+      "grad_norm": 0.09636734426021576,
+      "learning_rate": 7.231720443565546e-07,
+      "loss": 8.4691,
+      "step": 195430
+    },
+    {
+      "epoch": 0.9760043946165947,
+      "grad_norm": 0.09596901386976242,
+      "learning_rate": 7.216701294150042e-07,
+      "loss": 8.5013,
+      "step": 195440
+    },
+    {
+      "epoch": 0.9760543334415341,
+      "grad_norm": 0.08806633204221725,
+      "learning_rate": 7.201682144734536e-07,
+      "loss": 8.4973,
+      "step": 195450
+    },
+    {
+      "epoch": 0.9761042722664736,
+      "grad_norm": 0.08893059194087982,
+      "learning_rate": 7.186662995319032e-07,
+      "loss": 8.5316,
+      "step": 195460
+    },
+    {
+      "epoch": 0.976154211091413,
+      "grad_norm": 0.09048335999250412,
+      "learning_rate": 7.171643845903527e-07,
+      "loss": 8.504,
+      "step": 195470
+    },
+    {
+      "epoch": 0.9762041499163525,
+      "grad_norm": 0.09120376408100128,
+      "learning_rate": 7.156624696488023e-07,
+      "loss": 8.4939,
+      "step": 195480
+    },
+    {
+      "epoch": 0.9762540887412919,
+      "grad_norm": 0.09385009855031967,
+      "learning_rate": 7.141605547072518e-07,
+      "loss": 8.4997,
+      "step": 195490
+    },
+    {
+      "epoch": 0.9763040275662314,
+      "grad_norm": 0.09150715172290802,
+      "learning_rate": 7.126586397657012e-07,
+      "loss": 8.4917,
+      "step": 195500
+    },
+    {
+      "epoch": 0.9763539663911708,
+      "grad_norm": 0.08368659019470215,
+      "learning_rate": 7.111567248241508e-07,
+      "loss": 8.5044,
+      "step": 195510
+    },
+    {
+      "epoch": 0.9764039052161103,
+      "grad_norm": 0.09331393986940384,
+      "learning_rate": 7.096548098826003e-07,
+      "loss": 8.5056,
+      "step": 195520
+    },
+    {
+      "epoch": 0.9764538440410497,
+      "grad_norm": 0.08423703908920288,
+      "learning_rate": 7.081528949410499e-07,
+      "loss": 8.4948,
+      "step": 195530
+    },
+    {
+      "epoch": 0.9765037828659892,
+      "grad_norm": 0.090935617685318,
+      "learning_rate": 7.066509799994994e-07,
+      "loss": 8.5093,
+      "step": 195540
+    },
+    {
+      "epoch": 0.9765537216909286,
+      "grad_norm": 0.089198037981987,
+      "learning_rate": 7.051490650579488e-07,
+      "loss": 8.5092,
+      "step": 195550
+    },
+    {
+      "epoch": 0.9766036605158681,
+      "grad_norm": 0.09370332956314087,
+      "learning_rate": 7.036471501163984e-07,
+      "loss": 8.4945,
+      "step": 195560
+    },
+    {
+      "epoch": 0.9766535993408075,
+      "grad_norm": 0.09572462737560272,
+      "learning_rate": 7.021452351748479e-07,
+      "loss": 8.4806,
+      "step": 195570
+    },
+    {
+      "epoch": 0.976703538165747,
+      "grad_norm": 0.09619718790054321,
+      "learning_rate": 7.006433202332975e-07,
+      "loss": 8.4985,
+      "step": 195580
+    },
+    {
+      "epoch": 0.9767534769906864,
+      "grad_norm": 0.08778340369462967,
+      "learning_rate": 6.99141405291747e-07,
+      "loss": 8.5015,
+      "step": 195590
+    },
+    {
+      "epoch": 0.9768034158156259,
+      "grad_norm": 0.09372592717409134,
+      "learning_rate": 6.976394903501966e-07,
+      "loss": 8.4923,
+      "step": 195600
+    },
+    {
+      "epoch": 0.9768533546405653,
+      "grad_norm": 0.08879277855157852,
+      "learning_rate": 6.96137575408646e-07,
+      "loss": 8.5049,
+      "step": 195610
+    },
+    {
+      "epoch": 0.9769032934655048,
+      "grad_norm": 0.09419248253107071,
+      "learning_rate": 6.946356604670955e-07,
+      "loss": 8.5186,
+      "step": 195620
+    },
+    {
+      "epoch": 0.9769532322904442,
+      "grad_norm": 0.09513996541500092,
+      "learning_rate": 6.931337455255451e-07,
+      "loss": 8.5032,
+      "step": 195630
+    },
+    {
+      "epoch": 0.9770031711153837,
+      "grad_norm": 0.09008798003196716,
+      "learning_rate": 6.916318305839946e-07,
+      "loss": 8.493,
+      "step": 195640
+    },
+    {
+      "epoch": 0.9770531099403231,
+      "grad_norm": 0.08629944920539856,
+      "learning_rate": 6.901299156424442e-07,
+      "loss": 8.4955,
+      "step": 195650
+    },
+    {
+      "epoch": 0.9771030487652625,
+      "grad_norm": 0.09061171859502792,
+      "learning_rate": 6.886280007008936e-07,
+      "loss": 8.5169,
+      "step": 195660
+    },
+    {
+      "epoch": 0.977152987590202,
+      "grad_norm": 0.0888964906334877,
+      "learning_rate": 6.871260857593431e-07,
+      "loss": 8.4928,
+      "step": 195670
+    },
+    {
+      "epoch": 0.9772029264151415,
+      "grad_norm": 0.08664719760417938,
+      "learning_rate": 6.856241708177927e-07,
+      "loss": 8.4964,
+      "step": 195680
+    },
+    {
+      "epoch": 0.9772528652400809,
+      "grad_norm": 0.08702714741230011,
+      "learning_rate": 6.841222558762422e-07,
+      "loss": 8.4972,
+      "step": 195690
+    },
+    {
+      "epoch": 0.9773028040650203,
+      "grad_norm": 0.0985323116183281,
+      "learning_rate": 6.826203409346918e-07,
+      "loss": 8.4806,
+      "step": 195700
+    },
+    {
+      "epoch": 0.9773527428899598,
+      "grad_norm": 0.08669869601726532,
+      "learning_rate": 6.811184259931413e-07,
+      "loss": 8.5084,
+      "step": 195710
+    },
+    {
+      "epoch": 0.9774026817148993,
+      "grad_norm": 0.09846388548612595,
+      "learning_rate": 6.796165110515908e-07,
+      "loss": 8.5034,
+      "step": 195720
+    },
+    {
+      "epoch": 0.9774526205398387,
+      "grad_norm": 0.08850651979446411,
+      "learning_rate": 6.781145961100403e-07,
+      "loss": 8.5063,
+      "step": 195730
+    },
+    {
+      "epoch": 0.9775025593647781,
+      "grad_norm": 0.09550187736749649,
+      "learning_rate": 6.766126811684898e-07,
+      "loss": 8.4943,
+      "step": 195740
+    },
+    {
+      "epoch": 0.9775524981897176,
+      "grad_norm": 0.08764312416315079,
+      "learning_rate": 6.751107662269394e-07,
+      "loss": 8.5015,
+      "step": 195750
+    },
+    {
+      "epoch": 0.9776024370146571,
+      "grad_norm": 0.09270459413528442,
+      "learning_rate": 6.736088512853889e-07,
+      "loss": 8.4993,
+      "step": 195760
+    },
+    {
+      "epoch": 0.9776523758395965,
+      "grad_norm": 0.09069382399320602,
+      "learning_rate": 6.721069363438384e-07,
+      "loss": 8.4991,
+      "step": 195770
+    },
+    {
+      "epoch": 0.9777023146645359,
+      "grad_norm": 0.09017602354288101,
+      "learning_rate": 6.706050214022879e-07,
+      "loss": 8.5079,
+      "step": 195780
+    },
+    {
+      "epoch": 0.9777522534894754,
+      "grad_norm": 0.0948113352060318,
+      "learning_rate": 6.691031064607375e-07,
+      "loss": 8.4913,
+      "step": 195790
+    },
+    {
+      "epoch": 0.9778021923144148,
+      "grad_norm": 0.09429457783699036,
+      "learning_rate": 6.67601191519187e-07,
+      "loss": 8.501,
+      "step": 195800
+    },
+    {
+      "epoch": 0.9778521311393543,
+      "grad_norm": 0.09153127670288086,
+      "learning_rate": 6.660992765776365e-07,
+      "loss": 8.4899,
+      "step": 195810
+    },
+    {
+      "epoch": 0.9779020699642937,
+      "grad_norm": 0.0950402021408081,
+      "learning_rate": 6.64597361636086e-07,
+      "loss": 8.5159,
+      "step": 195820
+    },
+    {
+      "epoch": 0.9779520087892332,
+      "grad_norm": 0.09019756317138672,
+      "learning_rate": 6.630954466945355e-07,
+      "loss": 8.494,
+      "step": 195830
+    },
+    {
+      "epoch": 0.9780019476141726,
+      "grad_norm": 0.08579851686954498,
+      "learning_rate": 6.615935317529851e-07,
+      "loss": 8.5005,
+      "step": 195840
+    },
+    {
+      "epoch": 0.9780518864391121,
+      "grad_norm": 0.09236463159322739,
+      "learning_rate": 6.600916168114346e-07,
+      "loss": 8.4911,
+      "step": 195850
+    },
+    {
+      "epoch": 0.9781018252640515,
+      "grad_norm": 0.08974046260118484,
+      "learning_rate": 6.585897018698841e-07,
+      "loss": 8.5048,
+      "step": 195860
+    },
+    {
+      "epoch": 0.978151764088991,
+      "grad_norm": 0.0937148705124855,
+      "learning_rate": 6.570877869283337e-07,
+      "loss": 8.5085,
+      "step": 195870
+    },
+    {
+      "epoch": 0.9782017029139304,
+      "grad_norm": 0.09331946820020676,
+      "learning_rate": 6.555858719867831e-07,
+      "loss": 8.4902,
+      "step": 195880
+    },
+    {
+      "epoch": 0.9782516417388699,
+      "grad_norm": 0.08849994838237762,
+      "learning_rate": 6.540839570452327e-07,
+      "loss": 8.513,
+      "step": 195890
+    },
+    {
+      "epoch": 0.9783015805638093,
+      "grad_norm": 0.09173839539289474,
+      "learning_rate": 6.525820421036822e-07,
+      "loss": 8.4979,
+      "step": 195900
+    },
+    {
+      "epoch": 0.9783515193887488,
+      "grad_norm": 0.09019637852907181,
+      "learning_rate": 6.510801271621318e-07,
+      "loss": 8.503,
+      "step": 195910
+    },
+    {
+      "epoch": 0.9784014582136882,
+      "grad_norm": 0.09567339718341827,
+      "learning_rate": 6.495782122205813e-07,
+      "loss": 8.5031,
+      "step": 195920
+    },
+    {
+      "epoch": 0.9784513970386277,
+      "grad_norm": 0.09023743122816086,
+      "learning_rate": 6.480762972790307e-07,
+      "loss": 8.4926,
+      "step": 195930
+    },
+    {
+      "epoch": 0.9785013358635671,
+      "grad_norm": 0.08731987327337265,
+      "learning_rate": 6.465743823374803e-07,
+      "loss": 8.4897,
+      "step": 195940
+    },
+    {
+      "epoch": 0.9785512746885066,
+      "grad_norm": 0.09049738943576813,
+      "learning_rate": 6.450724673959298e-07,
+      "loss": 8.514,
+      "step": 195950
+    },
+    {
+      "epoch": 0.978601213513446,
+      "grad_norm": 0.09187744557857513,
+      "learning_rate": 6.435705524543794e-07,
+      "loss": 8.4761,
+      "step": 195960
+    },
+    {
+      "epoch": 0.9786511523383855,
+      "grad_norm": 0.09112652391195297,
+      "learning_rate": 6.420686375128289e-07,
+      "loss": 8.5003,
+      "step": 195970
+    },
+    {
+      "epoch": 0.9787010911633249,
+      "grad_norm": 0.09285633265972137,
+      "learning_rate": 6.405667225712785e-07,
+      "loss": 8.4864,
+      "step": 195980
+    },
+    {
+      "epoch": 0.9787510299882644,
+      "grad_norm": 0.09360034763813019,
+      "learning_rate": 6.390648076297279e-07,
+      "loss": 8.497,
+      "step": 195990
+    },
+    {
+      "epoch": 0.9788009688132038,
+      "grad_norm": 0.09473513811826706,
+      "learning_rate": 6.375628926881774e-07,
+      "loss": 8.5115,
+      "step": 196000
+    },
+    {
+      "epoch": 0.9788509076381433,
+      "grad_norm": 0.09806603938341141,
+      "learning_rate": 6.36060977746627e-07,
+      "loss": 8.5038,
+      "step": 196010
+    },
+    {
+      "epoch": 0.9789008464630827,
+      "grad_norm": 0.0843457505106926,
+      "learning_rate": 6.345590628050765e-07,
+      "loss": 8.5104,
+      "step": 196020
+    },
+    {
+      "epoch": 0.9789507852880222,
+      "grad_norm": 0.0888645127415657,
+      "learning_rate": 6.330571478635261e-07,
+      "loss": 8.5043,
+      "step": 196030
+    },
+    {
+      "epoch": 0.9790007241129616,
+      "grad_norm": 0.09105794876813889,
+      "learning_rate": 6.315552329219755e-07,
+      "loss": 8.5055,
+      "step": 196040
+    },
+    {
+      "epoch": 0.9790506629379011,
+      "grad_norm": 0.08398020267486572,
+      "learning_rate": 6.30053317980425e-07,
+      "loss": 8.5022,
+      "step": 196050
+    },
+    {
+      "epoch": 0.9791006017628405,
+      "grad_norm": 0.09144193679094315,
+      "learning_rate": 6.285514030388746e-07,
+      "loss": 8.4931,
+      "step": 196060
+    },
+    {
+      "epoch": 0.97915054058778,
+      "grad_norm": 0.08951180428266525,
+      "learning_rate": 6.270494880973241e-07,
+      "loss": 8.4928,
+      "step": 196070
+    },
+    {
+      "epoch": 0.9792004794127194,
+      "grad_norm": 0.08442022651433945,
+      "learning_rate": 6.255475731557737e-07,
+      "loss": 8.495,
+      "step": 196080
+    },
+    {
+      "epoch": 0.9792504182376589,
+      "grad_norm": 0.0968695729970932,
+      "learning_rate": 6.240456582142231e-07,
+      "loss": 8.4993,
+      "step": 196090
+    },
+    {
+      "epoch": 0.9793003570625983,
+      "grad_norm": 0.08780843019485474,
+      "learning_rate": 6.225437432726727e-07,
+      "loss": 8.5002,
+      "step": 196100
+    },
+    {
+      "epoch": 0.9793502958875377,
+      "grad_norm": 0.08662664890289307,
+      "learning_rate": 6.210418283311222e-07,
+      "loss": 8.5069,
+      "step": 196110
+    },
+    {
+      "epoch": 0.9794002347124772,
+      "grad_norm": 0.09597373008728027,
+      "learning_rate": 6.195399133895717e-07,
+      "loss": 8.5132,
+      "step": 196120
+    },
+    {
+      "epoch": 0.9794501735374167,
+      "grad_norm": 0.09307907521724701,
+      "learning_rate": 6.180379984480213e-07,
+      "loss": 8.4828,
+      "step": 196130
+    },
+    {
+      "epoch": 0.9795001123623561,
+      "grad_norm": 0.0918552353978157,
+      "learning_rate": 6.165360835064708e-07,
+      "loss": 8.508,
+      "step": 196140
+    },
+    {
+      "epoch": 0.9795500511872955,
+      "grad_norm": 0.08918620645999908,
+      "learning_rate": 6.150341685649203e-07,
+      "loss": 8.5133,
+      "step": 196150
+    },
+    {
+      "epoch": 0.979599990012235,
+      "grad_norm": 0.09026898443698883,
+      "learning_rate": 6.135322536233698e-07,
+      "loss": 8.5028,
+      "step": 196160
+    },
+    {
+      "epoch": 0.9796499288371745,
+      "grad_norm": 0.08795063197612762,
+      "learning_rate": 6.120303386818193e-07,
+      "loss": 8.5184,
+      "step": 196170
+    },
+    {
+      "epoch": 0.9796998676621139,
+      "grad_norm": 0.09220097213983536,
+      "learning_rate": 6.105284237402689e-07,
+      "loss": 8.5082,
+      "step": 196180
+    },
+    {
+      "epoch": 0.9797498064870533,
+      "grad_norm": 0.08570639044046402,
+      "learning_rate": 6.090265087987184e-07,
+      "loss": 8.5005,
+      "step": 196190
+    },
+    {
+      "epoch": 0.9797997453119928,
+      "grad_norm": 0.08638404309749603,
+      "learning_rate": 6.075245938571679e-07,
+      "loss": 8.506,
+      "step": 196200
+    },
+    {
+      "epoch": 0.9798496841369323,
+      "grad_norm": 0.09257509559392929,
+      "learning_rate": 6.060226789156174e-07,
+      "loss": 8.4789,
+      "step": 196210
+    },
+    {
+      "epoch": 0.9798996229618717,
+      "grad_norm": 0.09392807632684708,
+      "learning_rate": 6.04520763974067e-07,
+      "loss": 8.4907,
+      "step": 196220
+    },
+    {
+      "epoch": 0.9799495617868111,
+      "grad_norm": 0.087054044008255,
+      "learning_rate": 6.030188490325165e-07,
+      "loss": 8.4988,
+      "step": 196230
+    },
+    {
+      "epoch": 0.9799995006117506,
+      "grad_norm": 0.09597992151975632,
+      "learning_rate": 6.01516934090966e-07,
+      "loss": 8.4917,
+      "step": 196240
+    },
+    {
+      "epoch": 0.9800494394366901,
+      "grad_norm": 0.08776523172855377,
+      "learning_rate": 6.000150191494156e-07,
+      "loss": 8.5092,
+      "step": 196250
+    },
+    {
+      "epoch": 0.9800993782616295,
+      "grad_norm": 0.08641576766967773,
+      "learning_rate": 5.98513104207865e-07,
+      "loss": 8.5092,
+      "step": 196260
+    },
+    {
+      "epoch": 0.9801493170865689,
+      "grad_norm": 0.08972152322530746,
+      "learning_rate": 5.970111892663146e-07,
+      "loss": 8.4981,
+      "step": 196270
+    },
+    {
+      "epoch": 0.9801992559115084,
+      "grad_norm": 0.09072574973106384,
+      "learning_rate": 5.955092743247641e-07,
+      "loss": 8.5155,
+      "step": 196280
+    },
+    {
+      "epoch": 0.9802491947364479,
+      "grad_norm": 0.09410280734300613,
+      "learning_rate": 5.940073593832137e-07,
+      "loss": 8.5123,
+      "step": 196290
+    },
+    {
+      "epoch": 0.9802991335613873,
+      "grad_norm": 0.09331624209880829,
+      "learning_rate": 5.925054444416632e-07,
+      "loss": 8.4814,
+      "step": 196300
+    },
+    {
+      "epoch": 0.9803490723863267,
+      "grad_norm": 0.09342946112155914,
+      "learning_rate": 5.910035295001126e-07,
+      "loss": 8.4965,
+      "step": 196310
+    },
+    {
+      "epoch": 0.9803990112112662,
+      "grad_norm": 0.08969230949878693,
+      "learning_rate": 5.895016145585622e-07,
+      "loss": 8.4857,
+      "step": 196320
+    },
+    {
+      "epoch": 0.9804489500362057,
+      "grad_norm": 0.08578766137361526,
+      "learning_rate": 5.879996996170117e-07,
+      "loss": 8.4913,
+      "step": 196330
+    },
+    {
+      "epoch": 0.9804988888611451,
+      "grad_norm": 0.0996427908539772,
+      "learning_rate": 5.864977846754613e-07,
+      "loss": 8.4963,
+      "step": 196340
+    },
+    {
+      "epoch": 0.9805488276860845,
+      "grad_norm": 0.09154878556728363,
+      "learning_rate": 5.849958697339108e-07,
+      "loss": 8.4779,
+      "step": 196350
+    },
+    {
+      "epoch": 0.980598766511024,
+      "grad_norm": 0.09681200236082077,
+      "learning_rate": 5.834939547923602e-07,
+      "loss": 8.5062,
+      "step": 196360
+    },
+    {
+      "epoch": 0.9806487053359635,
+      "grad_norm": 0.09222651273012161,
+      "learning_rate": 5.819920398508098e-07,
+      "loss": 8.4888,
+      "step": 196370
+    },
+    {
+      "epoch": 0.9806986441609029,
+      "grad_norm": 0.08910214900970459,
+      "learning_rate": 5.804901249092593e-07,
+      "loss": 8.5117,
+      "step": 196380
+    },
+    {
+      "epoch": 0.9807485829858423,
+      "grad_norm": 0.09133374691009521,
+      "learning_rate": 5.789882099677089e-07,
+      "loss": 8.4863,
+      "step": 196390
+    },
+    {
+      "epoch": 0.9807985218107818,
+      "grad_norm": 0.09324437379837036,
+      "learning_rate": 5.774862950261584e-07,
+      "loss": 8.4807,
+      "step": 196400
+    },
+    {
+      "epoch": 0.9808484606357213,
+      "grad_norm": 0.08947594463825226,
+      "learning_rate": 5.75984380084608e-07,
+      "loss": 8.4995,
+      "step": 196410
+    },
+    {
+      "epoch": 0.9808983994606607,
+      "grad_norm": 0.09630808234214783,
+      "learning_rate": 5.744824651430574e-07,
+      "loss": 8.4807,
+      "step": 196420
+    },
+    {
+      "epoch": 0.9809483382856001,
+      "grad_norm": 0.08872811496257782,
+      "learning_rate": 5.729805502015069e-07,
+      "loss": 8.5092,
+      "step": 196430
+    },
+    {
+      "epoch": 0.9809982771105396,
+      "grad_norm": 0.08594346791505814,
+      "learning_rate": 5.714786352599565e-07,
+      "loss": 8.5042,
+      "step": 196440
+    },
+    {
+      "epoch": 0.9810482159354791,
+      "grad_norm": 0.08785369992256165,
+      "learning_rate": 5.69976720318406e-07,
+      "loss": 8.4873,
+      "step": 196450
+    },
+    {
+      "epoch": 0.9810981547604185,
+      "grad_norm": 0.08623884618282318,
+      "learning_rate": 5.684748053768556e-07,
+      "loss": 8.5087,
+      "step": 196460
+    },
+    {
+      "epoch": 0.9811480935853579,
+      "grad_norm": 0.09336108714342117,
+      "learning_rate": 5.66972890435305e-07,
+      "loss": 8.4784,
+      "step": 196470
+    },
+    {
+      "epoch": 0.9811980324102973,
+      "grad_norm": 0.09492529183626175,
+      "learning_rate": 5.654709754937545e-07,
+      "loss": 8.5124,
+      "step": 196480
+    },
+    {
+      "epoch": 0.9812479712352369,
+      "grad_norm": 0.08852214366197586,
+      "learning_rate": 5.639690605522041e-07,
+      "loss": 8.5098,
+      "step": 196490
+    },
+    {
+      "epoch": 0.9812979100601763,
+      "grad_norm": 0.09364449232816696,
+      "learning_rate": 5.624671456106536e-07,
+      "loss": 8.4938,
+      "step": 196500
+    },
+    {
+      "epoch": 0.9813478488851157,
+      "grad_norm": 0.0880199745297432,
+      "learning_rate": 5.609652306691032e-07,
+      "loss": 8.5006,
+      "step": 196510
+    },
+    {
+      "epoch": 0.9813977877100551,
+      "grad_norm": 0.08908896148204803,
+      "learning_rate": 5.594633157275527e-07,
+      "loss": 8.5116,
+      "step": 196520
+    },
+    {
+      "epoch": 0.9814477265349947,
+      "grad_norm": 0.0934915542602539,
+      "learning_rate": 5.579614007860022e-07,
+      "loss": 8.4807,
+      "step": 196530
+    },
+    {
+      "epoch": 0.9814976653599341,
+      "grad_norm": 0.09108605235815048,
+      "learning_rate": 5.564594858444517e-07,
+      "loss": 8.4912,
+      "step": 196540
+    },
+    {
+      "epoch": 0.9815476041848735,
+      "grad_norm": 0.08718253672122955,
+      "learning_rate": 5.549575709029012e-07,
+      "loss": 8.5017,
+      "step": 196550
+    },
+    {
+      "epoch": 0.981597543009813,
+      "grad_norm": 0.09099813550710678,
+      "learning_rate": 5.534556559613508e-07,
+      "loss": 8.5023,
+      "step": 196560
+    },
+    {
+      "epoch": 0.9816474818347525,
+      "grad_norm": 0.08504286408424377,
+      "learning_rate": 5.519537410198003e-07,
+      "loss": 8.522,
+      "step": 196570
+    },
+    {
+      "epoch": 0.9816974206596919,
+      "grad_norm": 0.08254232257604599,
+      "learning_rate": 5.504518260782498e-07,
+      "loss": 8.4956,
+      "step": 196580
+    },
+    {
+      "epoch": 0.9817473594846313,
+      "grad_norm": 0.08998601138591766,
+      "learning_rate": 5.489499111366993e-07,
+      "loss": 8.5012,
+      "step": 196590
+    },
+    {
+      "epoch": 0.9817972983095707,
+      "grad_norm": 0.09334680438041687,
+      "learning_rate": 5.474479961951489e-07,
+      "loss": 8.5018,
+      "step": 196600
+    },
+    {
+      "epoch": 0.9818472371345103,
+      "grad_norm": 0.09622131288051605,
+      "learning_rate": 5.459460812535984e-07,
+      "loss": 8.5014,
+      "step": 196610
+    },
+    {
+      "epoch": 0.9818971759594497,
+      "grad_norm": 0.0919957160949707,
+      "learning_rate": 5.444441663120479e-07,
+      "loss": 8.5115,
+      "step": 196620
+    },
+    {
+      "epoch": 0.9819471147843891,
+      "grad_norm": 0.08943960070610046,
+      "learning_rate": 5.429422513704974e-07,
+      "loss": 8.5137,
+      "step": 196630
+    },
+    {
+      "epoch": 0.9819970536093285,
+      "grad_norm": 0.09023985266685486,
+      "learning_rate": 5.414403364289469e-07,
+      "loss": 8.5006,
+      "step": 196640
+    },
+    {
+      "epoch": 0.9820469924342681,
+      "grad_norm": 0.09034501016139984,
+      "learning_rate": 5.399384214873965e-07,
+      "loss": 8.4943,
+      "step": 196650
+    },
+    {
+      "epoch": 0.9820969312592075,
+      "grad_norm": 0.08900442719459534,
+      "learning_rate": 5.38436506545846e-07,
+      "loss": 8.5223,
+      "step": 196660
+    },
+    {
+      "epoch": 0.9821468700841469,
+      "grad_norm": 0.08898912370204926,
+      "learning_rate": 5.369345916042955e-07,
+      "loss": 8.51,
+      "step": 196670
+    },
+    {
+      "epoch": 0.9821968089090863,
+      "grad_norm": 0.09157400578260422,
+      "learning_rate": 5.354326766627451e-07,
+      "loss": 8.5056,
+      "step": 196680
+    },
+    {
+      "epoch": 0.9822467477340259,
+      "grad_norm": 0.09418471902608871,
+      "learning_rate": 5.339307617211945e-07,
+      "loss": 8.4981,
+      "step": 196690
+    },
+    {
+      "epoch": 0.9822966865589653,
+      "grad_norm": 0.0900435671210289,
+      "learning_rate": 5.324288467796441e-07,
+      "loss": 8.4888,
+      "step": 196700
+    },
+    {
+      "epoch": 0.9823466253839047,
+      "grad_norm": 0.0917317196726799,
+      "learning_rate": 5.309269318380936e-07,
+      "loss": 8.5087,
+      "step": 196710
+    },
+    {
+      "epoch": 0.9823965642088441,
+      "grad_norm": 0.09196818619966507,
+      "learning_rate": 5.294250168965432e-07,
+      "loss": 8.5133,
+      "step": 196720
+    },
+    {
+      "epoch": 0.9824465030337837,
+      "grad_norm": 0.09314832836389542,
+      "learning_rate": 5.279231019549927e-07,
+      "loss": 8.4901,
+      "step": 196730
+    },
+    {
+      "epoch": 0.9824964418587231,
+      "grad_norm": 0.09629698842763901,
+      "learning_rate": 5.264211870134421e-07,
+      "loss": 8.4865,
+      "step": 196740
+    },
+    {
+      "epoch": 0.9825463806836625,
+      "grad_norm": 0.09784742444753647,
+      "learning_rate": 5.249192720718917e-07,
+      "loss": 8.4937,
+      "step": 196750
+    },
+    {
+      "epoch": 0.9825963195086019,
+      "grad_norm": 0.08975919336080551,
+      "learning_rate": 5.234173571303412e-07,
+      "loss": 8.4921,
+      "step": 196760
+    },
+    {
+      "epoch": 0.9826462583335415,
+      "grad_norm": 0.09287573397159576,
+      "learning_rate": 5.219154421887908e-07,
+      "loss": 8.4988,
+      "step": 196770
+    },
+    {
+      "epoch": 0.9826961971584809,
+      "grad_norm": 0.09043823927640915,
+      "learning_rate": 5.204135272472403e-07,
+      "loss": 8.4797,
+      "step": 196780
+    },
+    {
+      "epoch": 0.9827461359834203,
+      "grad_norm": 0.0993688553571701,
+      "learning_rate": 5.189116123056899e-07,
+      "loss": 8.4846,
+      "step": 196790
+    },
+    {
+      "epoch": 0.9827960748083597,
+      "grad_norm": 0.0893450453877449,
+      "learning_rate": 5.174096973641393e-07,
+      "loss": 8.4981,
+      "step": 196800
+    },
+    {
+      "epoch": 0.9828460136332992,
+      "grad_norm": 0.08755376189947128,
+      "learning_rate": 5.159077824225888e-07,
+      "loss": 8.5012,
+      "step": 196810
+    },
+    {
+      "epoch": 0.9828959524582387,
+      "grad_norm": 0.09437739849090576,
+      "learning_rate": 5.144058674810384e-07,
+      "loss": 8.5008,
+      "step": 196820
+    },
+    {
+      "epoch": 0.9829458912831781,
+      "grad_norm": 0.08554568886756897,
+      "learning_rate": 5.129039525394879e-07,
+      "loss": 8.4992,
+      "step": 196830
+    },
+    {
+      "epoch": 0.9829958301081175,
+      "grad_norm": 0.09264194965362549,
+      "learning_rate": 5.114020375979375e-07,
+      "loss": 8.4875,
+      "step": 196840
+    },
+    {
+      "epoch": 0.983045768933057,
+      "grad_norm": 0.09195772558450699,
+      "learning_rate": 5.099001226563869e-07,
+      "loss": 8.5076,
+      "step": 196850
+    },
+    {
+      "epoch": 0.9830957077579965,
+      "grad_norm": 0.0927330031991005,
+      "learning_rate": 5.083982077148364e-07,
+      "loss": 8.5003,
+      "step": 196860
+    },
+    {
+      "epoch": 0.9831456465829359,
+      "grad_norm": 0.09278733283281326,
+      "learning_rate": 5.06896292773286e-07,
+      "loss": 8.4892,
+      "step": 196870
+    },
+    {
+      "epoch": 0.9831955854078753,
+      "grad_norm": 0.09159595519304276,
+      "learning_rate": 5.053943778317355e-07,
+      "loss": 8.501,
+      "step": 196880
+    },
+    {
+      "epoch": 0.9832455242328147,
+      "grad_norm": 0.08706831187009811,
+      "learning_rate": 5.038924628901851e-07,
+      "loss": 8.5019,
+      "step": 196890
+    },
+    {
+      "epoch": 0.9832954630577543,
+      "grad_norm": 0.08902629464864731,
+      "learning_rate": 5.023905479486345e-07,
+      "loss": 8.4914,
+      "step": 196900
+    },
+    {
+      "epoch": 0.9833454018826937,
+      "grad_norm": 0.09136991202831268,
+      "learning_rate": 5.00888633007084e-07,
+      "loss": 8.4933,
+      "step": 196910
+    },
+    {
+      "epoch": 0.9833953407076331,
+      "grad_norm": 0.09446898847818375,
+      "learning_rate": 4.993867180655336e-07,
+      "loss": 8.4835,
+      "step": 196920
+    },
+    {
+      "epoch": 0.9834452795325725,
+      "grad_norm": 0.09721554815769196,
+      "learning_rate": 4.978848031239831e-07,
+      "loss": 8.4986,
+      "step": 196930
+    },
+    {
+      "epoch": 0.9834952183575121,
+      "grad_norm": 0.09217740595340729,
+      "learning_rate": 4.963828881824327e-07,
+      "loss": 8.4987,
+      "step": 196940
+    },
+    {
+      "epoch": 0.9835451571824515,
+      "grad_norm": 0.09315560013055801,
+      "learning_rate": 4.948809732408822e-07,
+      "loss": 8.4985,
+      "step": 196950
+    },
+    {
+      "epoch": 0.9835950960073909,
+      "grad_norm": 0.08701568096876144,
+      "learning_rate": 4.933790582993317e-07,
+      "loss": 8.5176,
+      "step": 196960
+    },
+    {
+      "epoch": 0.9836450348323303,
+      "grad_norm": 0.09363686293363571,
+      "learning_rate": 4.918771433577812e-07,
+      "loss": 8.5186,
+      "step": 196970
+    },
+    {
+      "epoch": 0.9836949736572699,
+      "grad_norm": 0.0900358110666275,
+      "learning_rate": 4.903752284162307e-07,
+      "loss": 8.5071,
+      "step": 196980
+    },
+    {
+      "epoch": 0.9837449124822093,
+      "grad_norm": 0.08556856215000153,
+      "learning_rate": 4.888733134746803e-07,
+      "loss": 8.508,
+      "step": 196990
+    },
+    {
+      "epoch": 0.9837948513071487,
+      "grad_norm": 0.08887305855751038,
+      "learning_rate": 4.873713985331298e-07,
+      "loss": 8.4958,
+      "step": 197000
+    },
+    {
+      "epoch": 0.9838447901320881,
+      "grad_norm": 0.08783209323883057,
+      "learning_rate": 4.858694835915793e-07,
+      "loss": 8.5032,
+      "step": 197010
+    },
+    {
+      "epoch": 0.9838947289570277,
+      "grad_norm": 0.08912937343120575,
+      "learning_rate": 4.843675686500288e-07,
+      "loss": 8.5039,
+      "step": 197020
+    },
+    {
+      "epoch": 0.9839446677819671,
+      "grad_norm": 0.09269113093614578,
+      "learning_rate": 4.828656537084784e-07,
+      "loss": 8.4991,
+      "step": 197030
+    },
+    {
+      "epoch": 0.9839946066069065,
+      "grad_norm": 0.0884513109922409,
+      "learning_rate": 4.813637387669279e-07,
+      "loss": 8.4936,
+      "step": 197040
+    },
+    {
+      "epoch": 0.9840445454318459,
+      "grad_norm": 0.09053158015012741,
+      "learning_rate": 4.798618238253774e-07,
+      "loss": 8.5058,
+      "step": 197050
+    },
+    {
+      "epoch": 0.9840944842567855,
+      "grad_norm": 0.09161665290594101,
+      "learning_rate": 4.783599088838269e-07,
+      "loss": 8.5196,
+      "step": 197060
+    },
+    {
+      "epoch": 0.9841444230817249,
+      "grad_norm": 0.09395797550678253,
+      "learning_rate": 4.768579939422764e-07,
+      "loss": 8.5149,
+      "step": 197070
+    },
+    {
+      "epoch": 0.9841943619066643,
+      "grad_norm": 0.09031932055950165,
+      "learning_rate": 4.7535607900072597e-07,
+      "loss": 8.5077,
+      "step": 197080
+    },
+    {
+      "epoch": 0.9842443007316037,
+      "grad_norm": 0.08975038677453995,
+      "learning_rate": 4.738541640591755e-07,
+      "loss": 8.5292,
+      "step": 197090
+    },
+    {
+      "epoch": 0.9842942395565433,
+      "grad_norm": 0.09021060913801193,
+      "learning_rate": 4.7235224911762505e-07,
+      "loss": 8.4957,
+      "step": 197100
+    },
+    {
+      "epoch": 0.9843441783814827,
+      "grad_norm": 0.08840472251176834,
+      "learning_rate": 4.7085033417607454e-07,
+      "loss": 8.5036,
+      "step": 197110
+    },
+    {
+      "epoch": 0.9843941172064221,
+      "grad_norm": 0.08798038959503174,
+      "learning_rate": 4.693484192345241e-07,
+      "loss": 8.5017,
+      "step": 197120
+    },
+    {
+      "epoch": 0.9844440560313615,
+      "grad_norm": 0.08750328421592712,
+      "learning_rate": 4.678465042929735e-07,
+      "loss": 8.5018,
+      "step": 197130
+    },
+    {
+      "epoch": 0.9844939948563011,
+      "grad_norm": 0.0955323576927185,
+      "learning_rate": 4.6634458935142307e-07,
+      "loss": 8.4924,
+      "step": 197140
+    },
+    {
+      "epoch": 0.9845439336812405,
+      "grad_norm": 0.08959522098302841,
+      "learning_rate": 4.648426744098726e-07,
+      "loss": 8.5035,
+      "step": 197150
+    },
+    {
+      "epoch": 0.9845938725061799,
+      "grad_norm": 0.08650697767734528,
+      "learning_rate": 4.633407594683221e-07,
+      "loss": 8.4908,
+      "step": 197160
+    },
+    {
+      "epoch": 0.9846438113311193,
+      "grad_norm": 0.091778464615345,
+      "learning_rate": 4.6183884452677164e-07,
+      "loss": 8.4966,
+      "step": 197170
+    },
+    {
+      "epoch": 0.9846937501560589,
+      "grad_norm": 0.08719423413276672,
+      "learning_rate": 4.603369295852212e-07,
+      "loss": 8.5151,
+      "step": 197180
+    },
+    {
+      "epoch": 0.9847436889809983,
+      "grad_norm": 0.09176063537597656,
+      "learning_rate": 4.588350146436707e-07,
+      "loss": 8.4962,
+      "step": 197190
+    },
+    {
+      "epoch": 0.9847936278059377,
+      "grad_norm": 0.08916878700256348,
+      "learning_rate": 4.573330997021202e-07,
+      "loss": 8.4894,
+      "step": 197200
+    },
+    {
+      "epoch": 0.9848435666308771,
+      "grad_norm": 0.08697172999382019,
+      "learning_rate": 4.558311847605697e-07,
+      "loss": 8.5025,
+      "step": 197210
+    },
+    {
+      "epoch": 0.9848935054558167,
+      "grad_norm": 0.09053148329257965,
+      "learning_rate": 4.5432926981901925e-07,
+      "loss": 8.4972,
+      "step": 197220
+    },
+    {
+      "epoch": 0.9849434442807561,
+      "grad_norm": 0.0926218330860138,
+      "learning_rate": 4.528273548774688e-07,
+      "loss": 8.4953,
+      "step": 197230
+    },
+    {
+      "epoch": 0.9849933831056955,
+      "grad_norm": 0.0867324098944664,
+      "learning_rate": 4.513254399359183e-07,
+      "loss": 8.4862,
+      "step": 197240
+    },
+    {
+      "epoch": 0.9850433219306349,
+      "grad_norm": 0.0981505885720253,
+      "learning_rate": 4.4982352499436783e-07,
+      "loss": 8.4912,
+      "step": 197250
+    },
+    {
+      "epoch": 0.9850932607555745,
+      "grad_norm": 0.09056996554136276,
+      "learning_rate": 4.4832161005281737e-07,
+      "loss": 8.508,
+      "step": 197260
+    },
+    {
+      "epoch": 0.9851431995805139,
+      "grad_norm": 0.08942002803087234,
+      "learning_rate": 4.4681969511126686e-07,
+      "loss": 8.5016,
+      "step": 197270
+    },
+    {
+      "epoch": 0.9851931384054533,
+      "grad_norm": 0.09431149810552597,
+      "learning_rate": 4.453177801697164e-07,
+      "loss": 8.5033,
+      "step": 197280
+    },
+    {
+      "epoch": 0.9852430772303927,
+      "grad_norm": 0.08968444168567657,
+      "learning_rate": 4.438158652281659e-07,
+      "loss": 8.5009,
+      "step": 197290
+    },
+    {
+      "epoch": 0.9852930160553323,
+      "grad_norm": 0.09338608384132385,
+      "learning_rate": 4.4231395028661544e-07,
+      "loss": 8.4933,
+      "step": 197300
+    },
+    {
+      "epoch": 0.9853429548802717,
+      "grad_norm": 0.09319829195737839,
+      "learning_rate": 4.40812035345065e-07,
+      "loss": 8.5047,
+      "step": 197310
+    },
+    {
+      "epoch": 0.9853928937052111,
+      "grad_norm": 0.08871421962976456,
+      "learning_rate": 4.3931012040351447e-07,
+      "loss": 8.4965,
+      "step": 197320
+    },
+    {
+      "epoch": 0.9854428325301505,
+      "grad_norm": 0.08973289281129837,
+      "learning_rate": 4.37808205461964e-07,
+      "loss": 8.5169,
+      "step": 197330
+    },
+    {
+      "epoch": 0.9854927713550901,
+      "grad_norm": 0.0932137593626976,
+      "learning_rate": 4.3630629052041356e-07,
+      "loss": 8.5038,
+      "step": 197340
+    },
+    {
+      "epoch": 0.9855427101800295,
+      "grad_norm": 0.09041199833154678,
+      "learning_rate": 4.3480437557886305e-07,
+      "loss": 8.4874,
+      "step": 197350
+    },
+    {
+      "epoch": 0.9855926490049689,
+      "grad_norm": 0.08918718248605728,
+      "learning_rate": 4.333024606373126e-07,
+      "loss": 8.5049,
+      "step": 197360
+    },
+    {
+      "epoch": 0.9856425878299083,
+      "grad_norm": 0.0872543603181839,
+      "learning_rate": 4.3180054569576213e-07,
+      "loss": 8.4965,
+      "step": 197370
+    },
+    {
+      "epoch": 0.9856925266548479,
+      "grad_norm": 0.09441404789686203,
+      "learning_rate": 4.302986307542116e-07,
+      "loss": 8.5013,
+      "step": 197380
+    },
+    {
+      "epoch": 0.9857424654797873,
+      "grad_norm": 0.08817983418703079,
+      "learning_rate": 4.2879671581266117e-07,
+      "loss": 8.4965,
+      "step": 197390
+    },
+    {
+      "epoch": 0.9857924043047267,
+      "grad_norm": 0.09181421250104904,
+      "learning_rate": 4.2729480087111066e-07,
+      "loss": 8.4777,
+      "step": 197400
+    },
+    {
+      "epoch": 0.9858423431296661,
+      "grad_norm": 0.0921471118927002,
+      "learning_rate": 4.257928859295602e-07,
+      "loss": 8.4848,
+      "step": 197410
+    },
+    {
+      "epoch": 0.9858922819546057,
+      "grad_norm": 0.08574623614549637,
+      "learning_rate": 4.2429097098800974e-07,
+      "loss": 8.4989,
+      "step": 197420
+    },
+    {
+      "epoch": 0.9859422207795451,
+      "grad_norm": 0.08747486025094986,
+      "learning_rate": 4.2278905604645923e-07,
+      "loss": 8.4846,
+      "step": 197430
+    },
+    {
+      "epoch": 0.9859921596044845,
+      "grad_norm": 0.09855210781097412,
+      "learning_rate": 4.212871411049088e-07,
+      "loss": 8.505,
+      "step": 197440
+    },
+    {
+      "epoch": 0.9860420984294239,
+      "grad_norm": 0.09506363421678543,
+      "learning_rate": 4.197852261633583e-07,
+      "loss": 8.5137,
+      "step": 197450
+    },
+    {
+      "epoch": 0.9860920372543635,
+      "grad_norm": 0.09487252682447433,
+      "learning_rate": 4.182833112218078e-07,
+      "loss": 8.499,
+      "step": 197460
+    },
+    {
+      "epoch": 0.9861419760793029,
+      "grad_norm": 0.08848896622657776,
+      "learning_rate": 4.1678139628025735e-07,
+      "loss": 8.4939,
+      "step": 197470
+    },
+    {
+      "epoch": 0.9861919149042423,
+      "grad_norm": 0.09323852509260178,
+      "learning_rate": 4.1527948133870684e-07,
+      "loss": 8.5181,
+      "step": 197480
+    },
+    {
+      "epoch": 0.9862418537291817,
+      "grad_norm": 0.09195881336927414,
+      "learning_rate": 4.137775663971564e-07,
+      "loss": 8.5032,
+      "step": 197490
+    },
+    {
+      "epoch": 0.9862917925541212,
+      "grad_norm": 0.08648441731929779,
+      "learning_rate": 4.1227565145560593e-07,
+      "loss": 8.5266,
+      "step": 197500
+    },
+    {
+      "epoch": 0.9863417313790607,
+      "grad_norm": 0.08899568021297455,
+      "learning_rate": 4.107737365140554e-07,
+      "loss": 8.4906,
+      "step": 197510
+    },
+    {
+      "epoch": 0.9863916702040001,
+      "grad_norm": 0.09646333009004593,
+      "learning_rate": 4.0927182157250496e-07,
+      "loss": 8.481,
+      "step": 197520
+    },
+    {
+      "epoch": 0.9864416090289395,
+      "grad_norm": 0.09079481661319733,
+      "learning_rate": 4.077699066309545e-07,
+      "loss": 8.4866,
+      "step": 197530
+    },
+    {
+      "epoch": 0.986491547853879,
+      "grad_norm": 0.09302527457475662,
+      "learning_rate": 4.06267991689404e-07,
+      "loss": 8.4956,
+      "step": 197540
+    },
+    {
+      "epoch": 0.9865414866788185,
+      "grad_norm": 0.08830315619707108,
+      "learning_rate": 4.0476607674785354e-07,
+      "loss": 8.4984,
+      "step": 197550
+    },
+    {
+      "epoch": 0.9865914255037579,
+      "grad_norm": 0.085964635014534,
+      "learning_rate": 4.0326416180630303e-07,
+      "loss": 8.4988,
+      "step": 197560
+    },
+    {
+      "epoch": 0.9866413643286973,
+      "grad_norm": 0.0878470242023468,
+      "learning_rate": 4.0176224686475257e-07,
+      "loss": 8.4891,
+      "step": 197570
+    },
+    {
+      "epoch": 0.9866913031536368,
+      "grad_norm": 0.09524169564247131,
+      "learning_rate": 4.002603319232021e-07,
+      "loss": 8.4738,
+      "step": 197580
+    },
+    {
+      "epoch": 0.9867412419785763,
+      "grad_norm": 0.09403923898935318,
+      "learning_rate": 3.987584169816516e-07,
+      "loss": 8.4943,
+      "step": 197590
+    },
+    {
+      "epoch": 0.9867911808035157,
+      "grad_norm": 0.09067419916391373,
+      "learning_rate": 3.9725650204010115e-07,
+      "loss": 8.5038,
+      "step": 197600
+    },
+    {
+      "epoch": 0.9868411196284551,
+      "grad_norm": 0.09345284849405289,
+      "learning_rate": 3.957545870985507e-07,
+      "loss": 8.4813,
+      "step": 197610
+    },
+    {
+      "epoch": 0.9868910584533946,
+      "grad_norm": 0.0898512527346611,
+      "learning_rate": 3.942526721570002e-07,
+      "loss": 8.4934,
+      "step": 197620
+    },
+    {
+      "epoch": 0.9869409972783341,
+      "grad_norm": 0.08990932255983353,
+      "learning_rate": 3.927507572154497e-07,
+      "loss": 8.5102,
+      "step": 197630
+    },
+    {
+      "epoch": 0.9869909361032735,
+      "grad_norm": 0.08922252804040909,
+      "learning_rate": 3.9124884227389927e-07,
+      "loss": 8.5005,
+      "step": 197640
+    },
+    {
+      "epoch": 0.9870408749282129,
+      "grad_norm": 0.09017761051654816,
+      "learning_rate": 3.8974692733234876e-07,
+      "loss": 8.4891,
+      "step": 197650
+    },
+    {
+      "epoch": 0.9870908137531524,
+      "grad_norm": 0.09690330922603607,
+      "learning_rate": 3.882450123907983e-07,
+      "loss": 8.519,
+      "step": 197660
+    },
+    {
+      "epoch": 0.9871407525780919,
+      "grad_norm": 0.08771155774593353,
+      "learning_rate": 3.867430974492478e-07,
+      "loss": 8.5076,
+      "step": 197670
+    },
+    {
+      "epoch": 0.9871906914030313,
+      "grad_norm": 0.09097769856452942,
+      "learning_rate": 3.8524118250769733e-07,
+      "loss": 8.4906,
+      "step": 197680
+    },
+    {
+      "epoch": 0.9872406302279707,
+      "grad_norm": 0.0901932418346405,
+      "learning_rate": 3.837392675661469e-07,
+      "loss": 8.4956,
+      "step": 197690
+    },
+    {
+      "epoch": 0.9872905690529102,
+      "grad_norm": 0.08922629803419113,
+      "learning_rate": 3.8223735262459637e-07,
+      "loss": 8.5213,
+      "step": 197700
+    },
+    {
+      "epoch": 0.9873405078778497,
+      "grad_norm": 0.09865950047969818,
+      "learning_rate": 3.807354376830459e-07,
+      "loss": 8.5128,
+      "step": 197710
+    },
+    {
+      "epoch": 0.9873904467027891,
+      "grad_norm": 0.0916372612118721,
+      "learning_rate": 3.7923352274149545e-07,
+      "loss": 8.5038,
+      "step": 197720
+    },
+    {
+      "epoch": 0.9874403855277285,
+      "grad_norm": 0.09816200286149979,
+      "learning_rate": 3.7773160779994494e-07,
+      "loss": 8.483,
+      "step": 197730
+    },
+    {
+      "epoch": 0.987490324352668,
+      "grad_norm": 0.0914330706000328,
+      "learning_rate": 3.762296928583945e-07,
+      "loss": 8.5011,
+      "step": 197740
+    },
+    {
+      "epoch": 0.9875402631776075,
+      "grad_norm": 0.09769126027822495,
+      "learning_rate": 3.74727777916844e-07,
+      "loss": 8.5084,
+      "step": 197750
+    },
+    {
+      "epoch": 0.9875902020025469,
+      "grad_norm": 0.09237746149301529,
+      "learning_rate": 3.732258629752935e-07,
+      "loss": 8.4931,
+      "step": 197760
+    },
+    {
+      "epoch": 0.9876401408274863,
+      "grad_norm": 0.09044523537158966,
+      "learning_rate": 3.7172394803374306e-07,
+      "loss": 8.5058,
+      "step": 197770
+    },
+    {
+      "epoch": 0.9876900796524257,
+      "grad_norm": 0.09292086958885193,
+      "learning_rate": 3.7022203309219255e-07,
+      "loss": 8.486,
+      "step": 197780
+    },
+    {
+      "epoch": 0.9877400184773653,
+      "grad_norm": 0.08766113966703415,
+      "learning_rate": 3.687201181506421e-07,
+      "loss": 8.5007,
+      "step": 197790
+    },
+    {
+      "epoch": 0.9877899573023047,
+      "grad_norm": 0.09624643623828888,
+      "learning_rate": 3.6721820320909164e-07,
+      "loss": 8.4815,
+      "step": 197800
+    },
+    {
+      "epoch": 0.9878398961272441,
+      "grad_norm": 0.0919470340013504,
+      "learning_rate": 3.6571628826754113e-07,
+      "loss": 8.5096,
+      "step": 197810
+    },
+    {
+      "epoch": 0.9878898349521835,
+      "grad_norm": 0.09844647347927094,
+      "learning_rate": 3.6421437332599067e-07,
+      "loss": 8.4995,
+      "step": 197820
+    },
+    {
+      "epoch": 0.987939773777123,
+      "grad_norm": 0.0885077640414238,
+      "learning_rate": 3.6271245838444016e-07,
+      "loss": 8.4987,
+      "step": 197830
+    },
+    {
+      "epoch": 0.9879897126020625,
+      "grad_norm": 0.09276360273361206,
+      "learning_rate": 3.612105434428897e-07,
+      "loss": 8.496,
+      "step": 197840
+    },
+    {
+      "epoch": 0.9880396514270019,
+      "grad_norm": 0.09369642287492752,
+      "learning_rate": 3.5970862850133925e-07,
+      "loss": 8.4979,
+      "step": 197850
+    },
+    {
+      "epoch": 0.9880895902519413,
+      "grad_norm": 0.08917860686779022,
+      "learning_rate": 3.5820671355978874e-07,
+      "loss": 8.4985,
+      "step": 197860
+    },
+    {
+      "epoch": 0.9881395290768809,
+      "grad_norm": 0.09895788133144379,
+      "learning_rate": 3.567047986182383e-07,
+      "loss": 8.5084,
+      "step": 197870
+    },
+    {
+      "epoch": 0.9881894679018203,
+      "grad_norm": 0.09372813254594803,
+      "learning_rate": 3.552028836766878e-07,
+      "loss": 8.4956,
+      "step": 197880
+    },
+    {
+      "epoch": 0.9882394067267597,
+      "grad_norm": 0.08584040403366089,
+      "learning_rate": 3.537009687351373e-07,
+      "loss": 8.4953,
+      "step": 197890
+    },
+    {
+      "epoch": 0.9882893455516991,
+      "grad_norm": 0.09182565659284592,
+      "learning_rate": 3.5219905379358686e-07,
+      "loss": 8.4967,
+      "step": 197900
+    },
+    {
+      "epoch": 0.9883392843766386,
+      "grad_norm": 0.09269486367702484,
+      "learning_rate": 3.506971388520364e-07,
+      "loss": 8.486,
+      "step": 197910
+    },
+    {
+      "epoch": 0.9883892232015781,
+      "grad_norm": 0.09027913957834244,
+      "learning_rate": 3.491952239104859e-07,
+      "loss": 8.4925,
+      "step": 197920
+    },
+    {
+      "epoch": 0.9884391620265175,
+      "grad_norm": 0.09345810860395432,
+      "learning_rate": 3.4769330896893544e-07,
+      "loss": 8.5197,
+      "step": 197930
+    },
+    {
+      "epoch": 0.9884891008514569,
+      "grad_norm": 0.09321021288633347,
+      "learning_rate": 3.461913940273849e-07,
+      "loss": 8.5137,
+      "step": 197940
+    },
+    {
+      "epoch": 0.9885390396763964,
+      "grad_norm": 0.09090397506952286,
+      "learning_rate": 3.4468947908583447e-07,
+      "loss": 8.4701,
+      "step": 197950
+    },
+    {
+      "epoch": 0.9885889785013359,
+      "grad_norm": 0.08423801511526108,
+      "learning_rate": 3.43187564144284e-07,
+      "loss": 8.4907,
+      "step": 197960
+    },
+    {
+      "epoch": 0.9886389173262753,
+      "grad_norm": 0.08930182456970215,
+      "learning_rate": 3.416856492027335e-07,
+      "loss": 8.506,
+      "step": 197970
+    },
+    {
+      "epoch": 0.9886888561512147,
+      "grad_norm": 0.09574810415506363,
+      "learning_rate": 3.4018373426118304e-07,
+      "loss": 8.4963,
+      "step": 197980
+    },
+    {
+      "epoch": 0.9887387949761542,
+      "grad_norm": 0.09295687824487686,
+      "learning_rate": 3.386818193196326e-07,
+      "loss": 8.4867,
+      "step": 197990
+    },
+    {
+      "epoch": 0.9887887338010937,
+      "grad_norm": 0.09014062583446503,
+      "learning_rate": 3.371799043780821e-07,
+      "loss": 8.4981,
+      "step": 198000
+    },
+    {
+      "epoch": 0.9888386726260331,
+      "grad_norm": 0.08868231624364853,
+      "learning_rate": 3.356779894365316e-07,
+      "loss": 8.4993,
+      "step": 198010
+    },
+    {
+      "epoch": 0.9888886114509725,
+      "grad_norm": 0.09391395002603531,
+      "learning_rate": 3.341760744949811e-07,
+      "loss": 8.4724,
+      "step": 198020
+    },
+    {
+      "epoch": 0.988938550275912,
+      "grad_norm": 0.08955026417970657,
+      "learning_rate": 3.3267415955343065e-07,
+      "loss": 8.5013,
+      "step": 198030
+    },
+    {
+      "epoch": 0.9889884891008515,
+      "grad_norm": 0.09359218180179596,
+      "learning_rate": 3.311722446118802e-07,
+      "loss": 8.4972,
+      "step": 198040
+    },
+    {
+      "epoch": 0.9890384279257909,
+      "grad_norm": 0.09413956850767136,
+      "learning_rate": 3.296703296703297e-07,
+      "loss": 8.5147,
+      "step": 198050
+    },
+    {
+      "epoch": 0.9890883667507303,
+      "grad_norm": 0.09088245779275894,
+      "learning_rate": 3.281684147287792e-07,
+      "loss": 8.4853,
+      "step": 198060
+    },
+    {
+      "epoch": 0.9891383055756698,
+      "grad_norm": 0.09208165854215622,
+      "learning_rate": 3.266664997872287e-07,
+      "loss": 8.4913,
+      "step": 198070
+    },
+    {
+      "epoch": 0.9891882444006093,
+      "grad_norm": 0.09409578889608383,
+      "learning_rate": 3.251645848456782e-07,
+      "loss": 8.4958,
+      "step": 198080
+    },
+    {
+      "epoch": 0.9892381832255487,
+      "grad_norm": 0.08956706523895264,
+      "learning_rate": 3.2366266990412775e-07,
+      "loss": 8.4896,
+      "step": 198090
+    },
+    {
+      "epoch": 0.9892881220504881,
+      "grad_norm": 0.09481275826692581,
+      "learning_rate": 3.2216075496257724e-07,
+      "loss": 8.4854,
+      "step": 198100
+    },
+    {
+      "epoch": 0.9893380608754276,
+      "grad_norm": 0.09185443073511124,
+      "learning_rate": 3.206588400210268e-07,
+      "loss": 8.5019,
+      "step": 198110
+    },
+    {
+      "epoch": 0.9893879997003671,
+      "grad_norm": 0.09097184240818024,
+      "learning_rate": 3.1915692507947633e-07,
+      "loss": 8.5186,
+      "step": 198120
+    },
+    {
+      "epoch": 0.9894379385253065,
+      "grad_norm": 0.0905596986413002,
+      "learning_rate": 3.176550101379258e-07,
+      "loss": 8.5126,
+      "step": 198130
+    },
+    {
+      "epoch": 0.9894878773502459,
+      "grad_norm": 0.09233243763446808,
+      "learning_rate": 3.1615309519637536e-07,
+      "loss": 8.4988,
+      "step": 198140
+    },
+    {
+      "epoch": 0.9895378161751854,
+      "grad_norm": 0.0961671769618988,
+      "learning_rate": 3.146511802548249e-07,
+      "loss": 8.4913,
+      "step": 198150
+    },
+    {
+      "epoch": 0.9895877550001249,
+      "grad_norm": 0.09645285457372665,
+      "learning_rate": 3.131492653132744e-07,
+      "loss": 8.4986,
+      "step": 198160
+    },
+    {
+      "epoch": 0.9896376938250643,
+      "grad_norm": 0.09721177816390991,
+      "learning_rate": 3.1164735037172394e-07,
+      "loss": 8.5021,
+      "step": 198170
+    },
+    {
+      "epoch": 0.9896876326500037,
+      "grad_norm": 0.09604041278362274,
+      "learning_rate": 3.101454354301735e-07,
+      "loss": 8.5088,
+      "step": 198180
+    },
+    {
+      "epoch": 0.9897375714749432,
+      "grad_norm": 0.09055788815021515,
+      "learning_rate": 3.0864352048862297e-07,
+      "loss": 8.4992,
+      "step": 198190
+    },
+    {
+      "epoch": 0.9897875102998827,
+      "grad_norm": 0.08688069880008698,
+      "learning_rate": 3.071416055470725e-07,
+      "loss": 8.5091,
+      "step": 198200
+    },
+    {
+      "epoch": 0.9898374491248221,
+      "grad_norm": 0.0961107388138771,
+      "learning_rate": 3.05639690605522e-07,
+      "loss": 8.5026,
+      "step": 198210
+    },
+    {
+      "epoch": 0.9898873879497615,
+      "grad_norm": 0.0864131972193718,
+      "learning_rate": 3.0413777566397155e-07,
+      "loss": 8.5183,
+      "step": 198220
+    },
+    {
+      "epoch": 0.989937326774701,
+      "grad_norm": 0.09623812884092331,
+      "learning_rate": 3.026358607224211e-07,
+      "loss": 8.5025,
+      "step": 198230
+    },
+    {
+      "epoch": 0.9899872655996405,
+      "grad_norm": 0.0971948653459549,
+      "learning_rate": 3.011339457808706e-07,
+      "loss": 8.4985,
+      "step": 198240
+    },
+    {
+      "epoch": 0.9900372044245799,
+      "grad_norm": 0.09427471458911896,
+      "learning_rate": 2.996320308393201e-07,
+      "loss": 8.4906,
+      "step": 198250
+    },
+    {
+      "epoch": 0.9900871432495193,
+      "grad_norm": 0.09101586788892746,
+      "learning_rate": 2.9813011589776967e-07,
+      "loss": 8.5086,
+      "step": 198260
+    },
+    {
+      "epoch": 0.9901370820744588,
+      "grad_norm": 0.09006375819444656,
+      "learning_rate": 2.9662820095621916e-07,
+      "loss": 8.493,
+      "step": 198270
+    },
+    {
+      "epoch": 0.9901870208993983,
+      "grad_norm": 0.08943824470043182,
+      "learning_rate": 2.951262860146687e-07,
+      "loss": 8.5086,
+      "step": 198280
+    },
+    {
+      "epoch": 0.9902369597243377,
+      "grad_norm": 0.08912531286478043,
+      "learning_rate": 2.936243710731182e-07,
+      "loss": 8.4813,
+      "step": 198290
+    },
+    {
+      "epoch": 0.9902868985492771,
+      "grad_norm": 0.10092119127511978,
+      "learning_rate": 2.9212245613156774e-07,
+      "loss": 8.4847,
+      "step": 198300
+    },
+    {
+      "epoch": 0.9903368373742166,
+      "grad_norm": 0.093541719019413,
+      "learning_rate": 2.906205411900173e-07,
+      "loss": 8.5023,
+      "step": 198310
+    },
+    {
+      "epoch": 0.990386776199156,
+      "grad_norm": 0.09948543459177017,
+      "learning_rate": 2.8911862624846677e-07,
+      "loss": 8.4979,
+      "step": 198320
+    },
+    {
+      "epoch": 0.9904367150240955,
+      "grad_norm": 0.0884837880730629,
+      "learning_rate": 2.876167113069163e-07,
+      "loss": 8.5008,
+      "step": 198330
+    },
+    {
+      "epoch": 0.9904866538490349,
+      "grad_norm": 0.08703433722257614,
+      "learning_rate": 2.8611479636536585e-07,
+      "loss": 8.4779,
+      "step": 198340
+    },
+    {
+      "epoch": 0.9905365926739744,
+      "grad_norm": 0.09166325628757477,
+      "learning_rate": 2.8461288142381534e-07,
+      "loss": 8.4882,
+      "step": 198350
+    },
+    {
+      "epoch": 0.9905865314989138,
+      "grad_norm": 0.09063871204853058,
+      "learning_rate": 2.831109664822649e-07,
+      "loss": 8.5057,
+      "step": 198360
+    },
+    {
+      "epoch": 0.9906364703238533,
+      "grad_norm": 0.08705553412437439,
+      "learning_rate": 2.816090515407144e-07,
+      "loss": 8.5068,
+      "step": 198370
+    },
+    {
+      "epoch": 0.9906864091487927,
+      "grad_norm": 0.09029226005077362,
+      "learning_rate": 2.801071365991639e-07,
+      "loss": 8.4944,
+      "step": 198380
+    },
+    {
+      "epoch": 0.9907363479737322,
+      "grad_norm": 0.09434723854064941,
+      "learning_rate": 2.7860522165761346e-07,
+      "loss": 8.4833,
+      "step": 198390
+    },
+    {
+      "epoch": 0.9907862867986716,
+      "grad_norm": 0.0964699313044548,
+      "learning_rate": 2.7710330671606295e-07,
+      "loss": 8.4942,
+      "step": 198400
+    },
+    {
+      "epoch": 0.9908362256236111,
+      "grad_norm": 0.08891110122203827,
+      "learning_rate": 2.756013917745125e-07,
+      "loss": 8.4942,
+      "step": 198410
+    },
+    {
+      "epoch": 0.9908861644485505,
+      "grad_norm": 0.09020981937646866,
+      "learning_rate": 2.7409947683296204e-07,
+      "loss": 8.5053,
+      "step": 198420
+    },
+    {
+      "epoch": 0.99093610327349,
+      "grad_norm": 0.09375789016485214,
+      "learning_rate": 2.7259756189141153e-07,
+      "loss": 8.503,
+      "step": 198430
+    },
+    {
+      "epoch": 0.9909860420984294,
+      "grad_norm": 0.08944463729858398,
+      "learning_rate": 2.7109564694986107e-07,
+      "loss": 8.5068,
+      "step": 198440
+    },
+    {
+      "epoch": 0.9910359809233689,
+      "grad_norm": 0.0924299880862236,
+      "learning_rate": 2.695937320083106e-07,
+      "loss": 8.4966,
+      "step": 198450
+    },
+    {
+      "epoch": 0.9910859197483083,
+      "grad_norm": 0.09989719092845917,
+      "learning_rate": 2.680918170667601e-07,
+      "loss": 8.4982,
+      "step": 198460
+    },
+    {
+      "epoch": 0.9911358585732478,
+      "grad_norm": 0.09110445529222488,
+      "learning_rate": 2.6658990212520965e-07,
+      "loss": 8.4971,
+      "step": 198470
+    },
+    {
+      "epoch": 0.9911857973981872,
+      "grad_norm": 0.09000138938426971,
+      "learning_rate": 2.6508798718365914e-07,
+      "loss": 8.496,
+      "step": 198480
+    },
+    {
+      "epoch": 0.9912357362231267,
+      "grad_norm": 0.09816454350948334,
+      "learning_rate": 2.635860722421087e-07,
+      "loss": 8.4862,
+      "step": 198490
+    },
+    {
+      "epoch": 0.9912856750480661,
+      "grad_norm": 0.09222198277711868,
+      "learning_rate": 2.620841573005582e-07,
+      "loss": 8.495,
+      "step": 198500
+    },
+    {
+      "epoch": 0.9913356138730056,
+      "grad_norm": 0.09800387173891068,
+      "learning_rate": 2.605822423590077e-07,
+      "loss": 8.5051,
+      "step": 198510
+    },
+    {
+      "epoch": 0.991385552697945,
+      "grad_norm": 0.08874485641717911,
+      "learning_rate": 2.5908032741745726e-07,
+      "loss": 8.4951,
+      "step": 198520
+    },
+    {
+      "epoch": 0.9914354915228845,
+      "grad_norm": 0.08928778022527695,
+      "learning_rate": 2.575784124759068e-07,
+      "loss": 8.4947,
+      "step": 198530
+    },
+    {
+      "epoch": 0.9914854303478239,
+      "grad_norm": 0.08784019201993942,
+      "learning_rate": 2.560764975343563e-07,
+      "loss": 8.4991,
+      "step": 198540
+    },
+    {
+      "epoch": 0.9915353691727634,
+      "grad_norm": 0.09007895737886429,
+      "learning_rate": 2.5457458259280584e-07,
+      "loss": 8.4845,
+      "step": 198550
+    },
+    {
+      "epoch": 0.9915853079977028,
+      "grad_norm": 0.08894022554159164,
+      "learning_rate": 2.530726676512553e-07,
+      "loss": 8.5006,
+      "step": 198560
+    },
+    {
+      "epoch": 0.9916352468226423,
+      "grad_norm": 0.08828302472829819,
+      "learning_rate": 2.5157075270970487e-07,
+      "loss": 8.52,
+      "step": 198570
+    },
+    {
+      "epoch": 0.9916851856475817,
+      "grad_norm": 0.0931435376405716,
+      "learning_rate": 2.500688377681544e-07,
+      "loss": 8.5057,
+      "step": 198580
+    },
+    {
+      "epoch": 0.9917351244725212,
+      "grad_norm": 0.09200121462345123,
+      "learning_rate": 2.485669228266039e-07,
+      "loss": 8.4947,
+      "step": 198590
+    },
+    {
+      "epoch": 0.9917850632974606,
+      "grad_norm": 0.08936648070812225,
+      "learning_rate": 2.4706500788505345e-07,
+      "loss": 8.4976,
+      "step": 198600
+    },
+    {
+      "epoch": 0.9918350021224001,
+      "grad_norm": 0.08992834389209747,
+      "learning_rate": 2.45563092943503e-07,
+      "loss": 8.5161,
+      "step": 198610
+    },
+    {
+      "epoch": 0.9918849409473395,
+      "grad_norm": 0.0911891758441925,
+      "learning_rate": 2.440611780019525e-07,
+      "loss": 8.4884,
+      "step": 198620
+    },
+    {
+      "epoch": 0.991934879772279,
+      "grad_norm": 0.08730587363243103,
+      "learning_rate": 2.42559263060402e-07,
+      "loss": 8.513,
+      "step": 198630
+    },
+    {
+      "epoch": 0.9919848185972184,
+      "grad_norm": 0.09357264637947083,
+      "learning_rate": 2.410573481188515e-07,
+      "loss": 8.5056,
+      "step": 198640
+    },
+    {
+      "epoch": 0.9920347574221579,
+      "grad_norm": 0.08683697134256363,
+      "learning_rate": 2.3955543317730105e-07,
+      "loss": 8.4803,
+      "step": 198650
+    },
+    {
+      "epoch": 0.9920846962470973,
+      "grad_norm": 0.08779203146696091,
+      "learning_rate": 2.3805351823575057e-07,
+      "loss": 8.4917,
+      "step": 198660
+    },
+    {
+      "epoch": 0.9921346350720368,
+      "grad_norm": 0.096683070063591,
+      "learning_rate": 2.3655160329420011e-07,
+      "loss": 8.5142,
+      "step": 198670
+    },
+    {
+      "epoch": 0.9921845738969762,
+      "grad_norm": 0.08591101318597794,
+      "learning_rate": 2.3504968835264963e-07,
+      "loss": 8.5099,
+      "step": 198680
+    },
+    {
+      "epoch": 0.9922345127219157,
+      "grad_norm": 0.08974206447601318,
+      "learning_rate": 2.3354777341109915e-07,
+      "loss": 8.5142,
+      "step": 198690
+    },
+    {
+      "epoch": 0.9922844515468551,
+      "grad_norm": 0.09247114509344101,
+      "learning_rate": 2.3204585846954866e-07,
+      "loss": 8.4848,
+      "step": 198700
+    },
+    {
+      "epoch": 0.9923343903717946,
+      "grad_norm": 0.08844613283872604,
+      "learning_rate": 2.305439435279982e-07,
+      "loss": 8.4997,
+      "step": 198710
+    },
+    {
+      "epoch": 0.992384329196734,
+      "grad_norm": 0.08982079476118088,
+      "learning_rate": 2.2904202858644772e-07,
+      "loss": 8.4906,
+      "step": 198720
+    },
+    {
+      "epoch": 0.9924342680216734,
+      "grad_norm": 0.08924318850040436,
+      "learning_rate": 2.2754011364489724e-07,
+      "loss": 8.4932,
+      "step": 198730
+    },
+    {
+      "epoch": 0.9924842068466129,
+      "grad_norm": 0.08730906248092651,
+      "learning_rate": 2.2603819870334678e-07,
+      "loss": 8.5035,
+      "step": 198740
+    },
+    {
+      "epoch": 0.9925341456715524,
+      "grad_norm": 0.08822734653949738,
+      "learning_rate": 2.245362837617963e-07,
+      "loss": 8.5039,
+      "step": 198750
+    },
+    {
+      "epoch": 0.9925840844964918,
+      "grad_norm": 0.09212057292461395,
+      "learning_rate": 2.2303436882024582e-07,
+      "loss": 8.4992,
+      "step": 198760
+    },
+    {
+      "epoch": 0.9926340233214312,
+      "grad_norm": 0.08733692765235901,
+      "learning_rate": 2.2153245387869533e-07,
+      "loss": 8.479,
+      "step": 198770
+    },
+    {
+      "epoch": 0.9926839621463707,
+      "grad_norm": 0.08462866395711899,
+      "learning_rate": 2.2003053893714488e-07,
+      "loss": 8.4978,
+      "step": 198780
+    },
+    {
+      "epoch": 0.9927339009713101,
+      "grad_norm": 0.09213852882385254,
+      "learning_rate": 2.185286239955944e-07,
+      "loss": 8.5001,
+      "step": 198790
+    },
+    {
+      "epoch": 0.9927838397962496,
+      "grad_norm": 0.08344164490699768,
+      "learning_rate": 2.170267090540439e-07,
+      "loss": 8.5045,
+      "step": 198800
+    },
+    {
+      "epoch": 0.992833778621189,
+      "grad_norm": 0.08867906033992767,
+      "learning_rate": 2.1552479411249343e-07,
+      "loss": 8.5079,
+      "step": 198810
+    },
+    {
+      "epoch": 0.9928837174461285,
+      "grad_norm": 0.08968056738376617,
+      "learning_rate": 2.1402287917094297e-07,
+      "loss": 8.5275,
+      "step": 198820
+    },
+    {
+      "epoch": 0.9929336562710679,
+      "grad_norm": 0.0994110107421875,
+      "learning_rate": 2.1252096422939249e-07,
+      "loss": 8.506,
+      "step": 198830
+    },
+    {
+      "epoch": 0.9929835950960074,
+      "grad_norm": 0.09683678299188614,
+      "learning_rate": 2.11019049287842e-07,
+      "loss": 8.5073,
+      "step": 198840
+    },
+    {
+      "epoch": 0.9930335339209468,
+      "grad_norm": 0.0983363687992096,
+      "learning_rate": 2.0951713434629152e-07,
+      "loss": 8.4861,
+      "step": 198850
+    },
+    {
+      "epoch": 0.9930834727458863,
+      "grad_norm": 0.09034904837608337,
+      "learning_rate": 2.0801521940474106e-07,
+      "loss": 8.4945,
+      "step": 198860
+    },
+    {
+      "epoch": 0.9931334115708257,
+      "grad_norm": 0.09096697717905045,
+      "learning_rate": 2.0651330446319058e-07,
+      "loss": 8.5022,
+      "step": 198870
+    },
+    {
+      "epoch": 0.9931833503957652,
+      "grad_norm": 0.085932657122612,
+      "learning_rate": 2.050113895216401e-07,
+      "loss": 8.5177,
+      "step": 198880
+    },
+    {
+      "epoch": 0.9932332892207046,
+      "grad_norm": 0.08621320128440857,
+      "learning_rate": 2.035094745800896e-07,
+      "loss": 8.4877,
+      "step": 198890
+    },
+    {
+      "epoch": 0.9932832280456441,
+      "grad_norm": 0.0914321318268776,
+      "learning_rate": 2.0200755963853916e-07,
+      "loss": 8.4897,
+      "step": 198900
+    },
+    {
+      "epoch": 0.9933331668705835,
+      "grad_norm": 0.08765504509210587,
+      "learning_rate": 2.0050564469698867e-07,
+      "loss": 8.502,
+      "step": 198910
+    },
+    {
+      "epoch": 0.993383105695523,
+      "grad_norm": 0.09285347908735275,
+      "learning_rate": 1.990037297554382e-07,
+      "loss": 8.4892,
+      "step": 198920
+    },
+    {
+      "epoch": 0.9934330445204624,
+      "grad_norm": 0.08891350775957108,
+      "learning_rate": 1.975018148138877e-07,
+      "loss": 8.5031,
+      "step": 198930
+    },
+    {
+      "epoch": 0.9934829833454019,
+      "grad_norm": 0.09927456080913544,
+      "learning_rate": 1.9599989987233725e-07,
+      "loss": 8.5036,
+      "step": 198940
+    },
+    {
+      "epoch": 0.9935329221703413,
+      "grad_norm": 0.08990452438592911,
+      "learning_rate": 1.9449798493078676e-07,
+      "loss": 8.4938,
+      "step": 198950
+    },
+    {
+      "epoch": 0.9935828609952808,
+      "grad_norm": 0.08888816833496094,
+      "learning_rate": 1.9299606998923628e-07,
+      "loss": 8.4857,
+      "step": 198960
+    },
+    {
+      "epoch": 0.9936327998202202,
+      "grad_norm": 0.09310969710350037,
+      "learning_rate": 1.914941550476858e-07,
+      "loss": 8.4864,
+      "step": 198970
+    },
+    {
+      "epoch": 0.9936827386451597,
+      "grad_norm": 0.08977384865283966,
+      "learning_rate": 1.8999224010613534e-07,
+      "loss": 8.4948,
+      "step": 198980
+    },
+    {
+      "epoch": 0.9937326774700991,
+      "grad_norm": 0.08601712435483932,
+      "learning_rate": 1.8849032516458486e-07,
+      "loss": 8.5055,
+      "step": 198990
+    },
+    {
+      "epoch": 0.9937826162950386,
+      "grad_norm": 0.08776349574327469,
+      "learning_rate": 1.8698841022303437e-07,
+      "loss": 8.4916,
+      "step": 199000
+    },
+    {
+      "epoch": 0.993832555119978,
+      "grad_norm": 0.09913208335638046,
+      "learning_rate": 1.8548649528148392e-07,
+      "loss": 8.4838,
+      "step": 199010
+    },
+    {
+      "epoch": 0.9938824939449175,
+      "grad_norm": 0.09497472643852234,
+      "learning_rate": 1.8398458033993343e-07,
+      "loss": 8.4997,
+      "step": 199020
+    },
+    {
+      "epoch": 0.9939324327698569,
+      "grad_norm": 0.08809848874807358,
+      "learning_rate": 1.8248266539838295e-07,
+      "loss": 8.4952,
+      "step": 199030
+    },
+    {
+      "epoch": 0.9939823715947964,
+      "grad_norm": 0.09473247826099396,
+      "learning_rate": 1.8098075045683247e-07,
+      "loss": 8.4957,
+      "step": 199040
+    },
+    {
+      "epoch": 0.9940323104197358,
+      "grad_norm": 0.08685030788183212,
+      "learning_rate": 1.79478835515282e-07,
+      "loss": 8.5072,
+      "step": 199050
+    },
+    {
+      "epoch": 0.9940822492446753,
+      "grad_norm": 0.08791281282901764,
+      "learning_rate": 1.7797692057373153e-07,
+      "loss": 8.5036,
+      "step": 199060
+    },
+    {
+      "epoch": 0.9941321880696147,
+      "grad_norm": 0.09338774532079697,
+      "learning_rate": 1.7647500563218104e-07,
+      "loss": 8.4922,
+      "step": 199070
+    },
+    {
+      "epoch": 0.9941821268945542,
+      "grad_norm": 0.09999926388263702,
+      "learning_rate": 1.7497309069063056e-07,
+      "loss": 8.5116,
+      "step": 199080
+    },
+    {
+      "epoch": 0.9942320657194936,
+      "grad_norm": 0.09071855247020721,
+      "learning_rate": 1.734711757490801e-07,
+      "loss": 8.5158,
+      "step": 199090
+    },
+    {
+      "epoch": 0.994282004544433,
+      "grad_norm": 0.0905730351805687,
+      "learning_rate": 1.7196926080752962e-07,
+      "loss": 8.4812,
+      "step": 199100
+    },
+    {
+      "epoch": 0.9943319433693725,
+      "grad_norm": 0.08825069665908813,
+      "learning_rate": 1.7046734586597914e-07,
+      "loss": 8.5107,
+      "step": 199110
+    },
+    {
+      "epoch": 0.994381882194312,
+      "grad_norm": 0.08565869927406311,
+      "learning_rate": 1.6896543092442865e-07,
+      "loss": 8.5244,
+      "step": 199120
+    },
+    {
+      "epoch": 0.9944318210192514,
+      "grad_norm": 0.09292873740196228,
+      "learning_rate": 1.674635159828782e-07,
+      "loss": 8.501,
+      "step": 199130
+    },
+    {
+      "epoch": 0.9944817598441908,
+      "grad_norm": 0.09264521300792694,
+      "learning_rate": 1.659616010413277e-07,
+      "loss": 8.5132,
+      "step": 199140
+    },
+    {
+      "epoch": 0.9945316986691303,
+      "grad_norm": 0.09055811911821365,
+      "learning_rate": 1.6445968609977723e-07,
+      "loss": 8.502,
+      "step": 199150
+    },
+    {
+      "epoch": 0.9945816374940698,
+      "grad_norm": 0.09365162253379822,
+      "learning_rate": 1.6295777115822672e-07,
+      "loss": 8.5061,
+      "step": 199160
+    },
+    {
+      "epoch": 0.9946315763190092,
+      "grad_norm": 0.08964534848928452,
+      "learning_rate": 1.6145585621667626e-07,
+      "loss": 8.4963,
+      "step": 199170
+    },
+    {
+      "epoch": 0.9946815151439486,
+      "grad_norm": 0.08972986787557602,
+      "learning_rate": 1.5995394127512578e-07,
+      "loss": 8.501,
+      "step": 199180
+    },
+    {
+      "epoch": 0.9947314539688881,
+      "grad_norm": 0.09291598945856094,
+      "learning_rate": 1.584520263335753e-07,
+      "loss": 8.5065,
+      "step": 199190
+    },
+    {
+      "epoch": 0.9947813927938276,
+      "grad_norm": 0.09010519832372665,
+      "learning_rate": 1.569501113920248e-07,
+      "loss": 8.5005,
+      "step": 199200
+    },
+    {
+      "epoch": 0.994831331618767,
+      "grad_norm": 0.08901604264974594,
+      "learning_rate": 1.5544819645047436e-07,
+      "loss": 8.4986,
+      "step": 199210
+    },
+    {
+      "epoch": 0.9948812704437064,
+      "grad_norm": 0.09446748346090317,
+      "learning_rate": 1.5394628150892387e-07,
+      "loss": 8.491,
+      "step": 199220
+    },
+    {
+      "epoch": 0.9949312092686459,
+      "grad_norm": 0.08962590992450714,
+      "learning_rate": 1.524443665673734e-07,
+      "loss": 8.499,
+      "step": 199230
+    },
+    {
+      "epoch": 0.9949811480935854,
+      "grad_norm": 0.0894179418683052,
+      "learning_rate": 1.509424516258229e-07,
+      "loss": 8.5048,
+      "step": 199240
+    },
+    {
+      "epoch": 0.9950310869185248,
+      "grad_norm": 0.0883219912648201,
+      "learning_rate": 1.4944053668427245e-07,
+      "loss": 8.5145,
+      "step": 199250
+    },
+    {
+      "epoch": 0.9950810257434642,
+      "grad_norm": 0.09172813594341278,
+      "learning_rate": 1.4793862174272197e-07,
+      "loss": 8.5009,
+      "step": 199260
+    },
+    {
+      "epoch": 0.9951309645684037,
+      "grad_norm": 0.09289974719285965,
+      "learning_rate": 1.4643670680117148e-07,
+      "loss": 8.5079,
+      "step": 199270
+    },
+    {
+      "epoch": 0.9951809033933432,
+      "grad_norm": 0.09185164421796799,
+      "learning_rate": 1.4493479185962102e-07,
+      "loss": 8.4842,
+      "step": 199280
+    },
+    {
+      "epoch": 0.9952308422182826,
+      "grad_norm": 0.089945949614048,
+      "learning_rate": 1.4343287691807054e-07,
+      "loss": 8.4803,
+      "step": 199290
+    },
+    {
+      "epoch": 0.995280781043222,
+      "grad_norm": 0.09140554815530777,
+      "learning_rate": 1.4193096197652006e-07,
+      "loss": 8.5122,
+      "step": 199300
+    },
+    {
+      "epoch": 0.9953307198681615,
+      "grad_norm": 0.0931864008307457,
+      "learning_rate": 1.4042904703496957e-07,
+      "loss": 8.4922,
+      "step": 199310
+    },
+    {
+      "epoch": 0.995380658693101,
+      "grad_norm": 0.08751117438077927,
+      "learning_rate": 1.3892713209341912e-07,
+      "loss": 8.4979,
+      "step": 199320
+    },
+    {
+      "epoch": 0.9954305975180404,
+      "grad_norm": 0.08914308249950409,
+      "learning_rate": 1.3742521715186863e-07,
+      "loss": 8.4794,
+      "step": 199330
+    },
+    {
+      "epoch": 0.9954805363429798,
+      "grad_norm": 0.09185323864221573,
+      "learning_rate": 1.3592330221031815e-07,
+      "loss": 8.4852,
+      "step": 199340
+    },
+    {
+      "epoch": 0.9955304751679193,
+      "grad_norm": 0.0920838937163353,
+      "learning_rate": 1.3442138726876767e-07,
+      "loss": 8.4852,
+      "step": 199350
+    },
+    {
+      "epoch": 0.9955804139928588,
+      "grad_norm": 0.08701695501804352,
+      "learning_rate": 1.329194723272172e-07,
+      "loss": 8.4987,
+      "step": 199360
+    },
+    {
+      "epoch": 0.9956303528177982,
+      "grad_norm": 0.09538272023200989,
+      "learning_rate": 1.3141755738566673e-07,
+      "loss": 8.4892,
+      "step": 199370
+    },
+    {
+      "epoch": 0.9956802916427376,
+      "grad_norm": 0.08852861821651459,
+      "learning_rate": 1.2991564244411624e-07,
+      "loss": 8.5061,
+      "step": 199380
+    },
+    {
+      "epoch": 0.9957302304676771,
+      "grad_norm": 0.09177656471729279,
+      "learning_rate": 1.2841372750256576e-07,
+      "loss": 8.4837,
+      "step": 199390
+    },
+    {
+      "epoch": 0.9957801692926166,
+      "grad_norm": 0.09195193648338318,
+      "learning_rate": 1.269118125610153e-07,
+      "loss": 8.4985,
+      "step": 199400
+    },
+    {
+      "epoch": 0.995830108117556,
+      "grad_norm": 0.09105399250984192,
+      "learning_rate": 1.2540989761946482e-07,
+      "loss": 8.5016,
+      "step": 199410
+    },
+    {
+      "epoch": 0.9958800469424954,
+      "grad_norm": 0.09002146124839783,
+      "learning_rate": 1.2390798267791434e-07,
+      "loss": 8.4919,
+      "step": 199420
+    },
+    {
+      "epoch": 0.9959299857674349,
+      "grad_norm": 0.09366229176521301,
+      "learning_rate": 1.2240606773636385e-07,
+      "loss": 8.5062,
+      "step": 199430
+    },
+    {
+      "epoch": 0.9959799245923744,
+      "grad_norm": 0.09170850366353989,
+      "learning_rate": 1.209041527948134e-07,
+      "loss": 8.4988,
+      "step": 199440
+    },
+    {
+      "epoch": 0.9960298634173138,
+      "grad_norm": 0.0911903902888298,
+      "learning_rate": 1.194022378532629e-07,
+      "loss": 8.4892,
+      "step": 199450
+    },
+    {
+      "epoch": 0.9960798022422532,
+      "grad_norm": 0.09550126641988754,
+      "learning_rate": 1.1790032291171243e-07,
+      "loss": 8.4933,
+      "step": 199460
+    },
+    {
+      "epoch": 0.9961297410671927,
+      "grad_norm": 0.08738602697849274,
+      "learning_rate": 1.1639840797016196e-07,
+      "loss": 8.4957,
+      "step": 199470
+    },
+    {
+      "epoch": 0.9961796798921322,
+      "grad_norm": 0.09028680622577667,
+      "learning_rate": 1.1489649302861148e-07,
+      "loss": 8.5054,
+      "step": 199480
+    },
+    {
+      "epoch": 0.9962296187170716,
+      "grad_norm": 0.08849211782217026,
+      "learning_rate": 1.13394578087061e-07,
+      "loss": 8.4965,
+      "step": 199490
+    },
+    {
+      "epoch": 0.996279557542011,
+      "grad_norm": 0.09421442449092865,
+      "learning_rate": 1.1189266314551052e-07,
+      "loss": 8.5073,
+      "step": 199500
+    },
+    {
+      "epoch": 0.9963294963669505,
+      "grad_norm": 0.09024062752723694,
+      "learning_rate": 1.1039074820396005e-07,
+      "loss": 8.5054,
+      "step": 199510
+    },
+    {
+      "epoch": 0.99637943519189,
+      "grad_norm": 0.09820150583982468,
+      "learning_rate": 1.0888883326240957e-07,
+      "loss": 8.4772,
+      "step": 199520
+    },
+    {
+      "epoch": 0.9964293740168294,
+      "grad_norm": 0.10558462142944336,
+      "learning_rate": 1.073869183208591e-07,
+      "loss": 8.495,
+      "step": 199530
+    },
+    {
+      "epoch": 0.9964793128417688,
+      "grad_norm": 0.09310225397348404,
+      "learning_rate": 1.0588500337930862e-07,
+      "loss": 8.4867,
+      "step": 199540
+    },
+    {
+      "epoch": 0.9965292516667082,
+      "grad_norm": 0.09150728583335876,
+      "learning_rate": 1.0438308843775815e-07,
+      "loss": 8.4969,
+      "step": 199550
+    },
+    {
+      "epoch": 0.9965791904916478,
+      "grad_norm": 0.09338292479515076,
+      "learning_rate": 1.0288117349620768e-07,
+      "loss": 8.4907,
+      "step": 199560
+    },
+    {
+      "epoch": 0.9966291293165872,
+      "grad_norm": 0.09587588161230087,
+      "learning_rate": 1.0137925855465719e-07,
+      "loss": 8.4943,
+      "step": 199570
+    },
+    {
+      "epoch": 0.9966790681415266,
+      "grad_norm": 0.09336939454078674,
+      "learning_rate": 9.987734361310672e-08,
+      "loss": 8.5067,
+      "step": 199580
+    },
+    {
+      "epoch": 0.996729006966466,
+      "grad_norm": 0.08930207043886185,
+      "learning_rate": 9.837542867155624e-08,
+      "loss": 8.5041,
+      "step": 199590
+    },
+    {
+      "epoch": 0.9967789457914056,
+      "grad_norm": 0.08806464821100235,
+      "learning_rate": 9.687351373000577e-08,
+      "loss": 8.5115,
+      "step": 199600
+    },
+    {
+      "epoch": 0.996828884616345,
+      "grad_norm": 0.09173772484064102,
+      "learning_rate": 9.537159878845528e-08,
+      "loss": 8.5085,
+      "step": 199610
+    },
+    {
+      "epoch": 0.9968788234412844,
+      "grad_norm": 0.08926473557949066,
+      "learning_rate": 9.386968384690481e-08,
+      "loss": 8.5104,
+      "step": 199620
+    },
+    {
+      "epoch": 0.9969287622662238,
+      "grad_norm": 0.09517572820186615,
+      "learning_rate": 9.236776890535433e-08,
+      "loss": 8.5092,
+      "step": 199630
+    },
+    {
+      "epoch": 0.9969787010911634,
+      "grad_norm": 0.08836136758327484,
+      "learning_rate": 9.086585396380386e-08,
+      "loss": 8.5023,
+      "step": 199640
+    },
+    {
+      "epoch": 0.9970286399161028,
+      "grad_norm": 0.09061846137046814,
+      "learning_rate": 8.936393902225338e-08,
+      "loss": 8.5222,
+      "step": 199650
+    },
+    {
+      "epoch": 0.9970785787410422,
+      "grad_norm": 0.09391947835683823,
+      "learning_rate": 8.786202408070291e-08,
+      "loss": 8.5009,
+      "step": 199660
+    },
+    {
+      "epoch": 0.9971285175659816,
+      "grad_norm": 0.09239717572927475,
+      "learning_rate": 8.636010913915242e-08,
+      "loss": 8.4959,
+      "step": 199670
+    },
+    {
+      "epoch": 0.9971784563909212,
+      "grad_norm": 0.09715589135885239,
+      "learning_rate": 8.485819419760195e-08,
+      "loss": 8.5019,
+      "step": 199680
+    },
+    {
+      "epoch": 0.9972283952158606,
+      "grad_norm": 0.09166194498538971,
+      "learning_rate": 8.335627925605147e-08,
+      "loss": 8.4841,
+      "step": 199690
+    },
+    {
+      "epoch": 0.9972783340408,
+      "grad_norm": 0.0937894880771637,
+      "learning_rate": 8.185436431450099e-08,
+      "loss": 8.5135,
+      "step": 199700
+    },
+    {
+      "epoch": 0.9973282728657394,
+      "grad_norm": 0.08937085419893265,
+      "learning_rate": 8.03524493729505e-08,
+      "loss": 8.5079,
+      "step": 199710
+    },
+    {
+      "epoch": 0.997378211690679,
+      "grad_norm": 0.09397994726896286,
+      "learning_rate": 7.885053443140003e-08,
+      "loss": 8.4847,
+      "step": 199720
+    },
+    {
+      "epoch": 0.9974281505156184,
+      "grad_norm": 0.09294033795595169,
+      "learning_rate": 7.734861948984955e-08,
+      "loss": 8.498,
+      "step": 199730
+    },
+    {
+      "epoch": 0.9974780893405578,
+      "grad_norm": 0.08595110476016998,
+      "learning_rate": 7.584670454829908e-08,
+      "loss": 8.4964,
+      "step": 199740
+    },
+    {
+      "epoch": 0.9975280281654972,
+      "grad_norm": 0.09340130537748337,
+      "learning_rate": 7.43447896067486e-08,
+      "loss": 8.4942,
+      "step": 199750
+    },
+    {
+      "epoch": 0.9975779669904367,
+      "grad_norm": 0.08756370097398758,
+      "learning_rate": 7.284287466519813e-08,
+      "loss": 8.5186,
+      "step": 199760
+    },
+    {
+      "epoch": 0.9976279058153762,
+      "grad_norm": 0.09118250012397766,
+      "learning_rate": 7.134095972364764e-08,
+      "loss": 8.4982,
+      "step": 199770
+    },
+    {
+      "epoch": 0.9976778446403156,
+      "grad_norm": 0.09316319227218628,
+      "learning_rate": 6.983904478209717e-08,
+      "loss": 8.491,
+      "step": 199780
+    },
+    {
+      "epoch": 0.997727783465255,
+      "grad_norm": 0.08617562055587769,
+      "learning_rate": 6.833712984054669e-08,
+      "loss": 8.4977,
+      "step": 199790
+    },
+    {
+      "epoch": 0.9977777222901945,
+      "grad_norm": 0.09523026645183563,
+      "learning_rate": 6.683521489899622e-08,
+      "loss": 8.5039,
+      "step": 199800
+    },
+    {
+      "epoch": 0.997827661115134,
+      "grad_norm": 0.09021805971860886,
+      "learning_rate": 6.533329995744574e-08,
+      "loss": 8.4985,
+      "step": 199810
+    },
+    {
+      "epoch": 0.9978775999400734,
+      "grad_norm": 0.08965896815061569,
+      "learning_rate": 6.383138501589527e-08,
+      "loss": 8.4904,
+      "step": 199820
+    },
+    {
+      "epoch": 0.9979275387650128,
+      "grad_norm": 0.09753076732158661,
+      "learning_rate": 6.23294700743448e-08,
+      "loss": 8.5033,
+      "step": 199830
+    },
+    {
+      "epoch": 0.9979774775899523,
+      "grad_norm": 0.0896683931350708,
+      "learning_rate": 6.082755513279431e-08,
+      "loss": 8.503,
+      "step": 199840
+    },
+    {
+      "epoch": 0.9980274164148918,
+      "grad_norm": 0.09268392622470856,
+      "learning_rate": 5.9325640191243836e-08,
+      "loss": 8.4949,
+      "step": 199850
+    },
+    {
+      "epoch": 0.9980773552398312,
+      "grad_norm": 0.08605291694402695,
+      "learning_rate": 5.782372524969336e-08,
+      "loss": 8.4886,
+      "step": 199860
+    },
+    {
+      "epoch": 0.9981272940647706,
+      "grad_norm": 0.09285443276166916,
+      "learning_rate": 5.632181030814288e-08,
+      "loss": 8.4869,
+      "step": 199870
+    },
+    {
+      "epoch": 0.99817723288971,
+      "grad_norm": 0.08672482520341873,
+      "learning_rate": 5.4819895366592405e-08,
+      "loss": 8.501,
+      "step": 199880
+    },
+    {
+      "epoch": 0.9982271717146496,
+      "grad_norm": 0.0957999974489212,
+      "learning_rate": 5.331798042504193e-08,
+      "loss": 8.5033,
+      "step": 199890
+    },
+    {
+      "epoch": 0.998277110539589,
+      "grad_norm": 0.08614014834165573,
+      "learning_rate": 5.181606548349145e-08,
+      "loss": 8.5082,
+      "step": 199900
+    },
+    {
+      "epoch": 0.9983270493645284,
+      "grad_norm": 0.09075914323329926,
+      "learning_rate": 5.0314150541940975e-08,
+      "loss": 8.4991,
+      "step": 199910
+    },
+    {
+      "epoch": 0.9983769881894679,
+      "grad_norm": 0.09307090193033218,
+      "learning_rate": 4.88122356003905e-08,
+      "loss": 8.474,
+      "step": 199920
+    },
+    {
+      "epoch": 0.9984269270144074,
+      "grad_norm": 0.09387535601854324,
+      "learning_rate": 4.731032065884002e-08,
+      "loss": 8.4924,
+      "step": 199930
+    },
+    {
+      "epoch": 0.9984768658393468,
+      "grad_norm": 0.08672790229320526,
+      "learning_rate": 4.5808405717289545e-08,
+      "loss": 8.5078,
+      "step": 199940
+    },
+    {
+      "epoch": 0.9985268046642862,
+      "grad_norm": 0.09689149260520935,
+      "learning_rate": 4.430649077573907e-08,
+      "loss": 8.5156,
+      "step": 199950
+    },
+    {
+      "epoch": 0.9985767434892256,
+      "grad_norm": 0.08997116982936859,
+      "learning_rate": 4.280457583418859e-08,
+      "loss": 8.5033,
+      "step": 199960
+    },
+    {
+      "epoch": 0.9986266823141652,
+      "grad_norm": 0.09200499206781387,
+      "learning_rate": 4.130266089263812e-08,
+      "loss": 8.499,
+      "step": 199970
+    },
+    {
+      "epoch": 0.9986766211391046,
+      "grad_norm": 0.09014206379652023,
+      "learning_rate": 3.980074595108764e-08,
+      "loss": 8.4947,
+      "step": 199980
+    },
+    {
+      "epoch": 0.998726559964044,
+      "grad_norm": 0.09259384125471115,
+      "learning_rate": 3.829883100953716e-08,
+      "loss": 8.4803,
+      "step": 199990
+    },
+    {
+      "epoch": 0.9987764987889834,
+      "grad_norm": 0.09015609323978424,
+      "learning_rate": 3.6796916067986684e-08,
+      "loss": 8.4907,
+      "step": 200000
+    },
+    {
+      "epoch": 0.998826437613923,
+      "grad_norm": 0.08941512554883957,
+      "learning_rate": 3.529500112643621e-08,
+      "loss": 8.5041,
+      "step": 200010
+    },
+    {
+      "epoch": 0.9988763764388624,
+      "grad_norm": 0.09428403526544571,
+      "learning_rate": 3.379308618488573e-08,
+      "loss": 8.5041,
+      "step": 200020
+    },
+    {
+      "epoch": 0.9989263152638018,
+      "grad_norm": 0.0898728296160698,
+      "learning_rate": 3.2291171243335254e-08,
+      "loss": 8.5173,
+      "step": 200030
+    },
+    {
+      "epoch": 0.9989762540887412,
+      "grad_norm": 0.08989530801773071,
+      "learning_rate": 3.078925630178478e-08,
+      "loss": 8.5081,
+      "step": 200040
+    },
+    {
+      "epoch": 0.9990261929136808,
+      "grad_norm": 0.09039604663848877,
+      "learning_rate": 2.92873413602343e-08,
+      "loss": 8.5074,
+      "step": 200050
+    },
+    {
+      "epoch": 0.9990761317386202,
+      "grad_norm": 0.08606907725334167,
+      "learning_rate": 2.7785426418683824e-08,
+      "loss": 8.5027,
+      "step": 200060
+    },
+    {
+      "epoch": 0.9991260705635596,
+      "grad_norm": 0.09312465786933899,
+      "learning_rate": 2.6283511477133347e-08,
+      "loss": 8.499,
+      "step": 200070
+    },
+    {
+      "epoch": 0.999176009388499,
+      "grad_norm": 0.09412333369255066,
+      "learning_rate": 2.478159653558287e-08,
+      "loss": 8.4968,
+      "step": 200080
+    },
+    {
+      "epoch": 0.9992259482134386,
+      "grad_norm": 0.09396173804998398,
+      "learning_rate": 2.3279681594032393e-08,
+      "loss": 8.4779,
+      "step": 200090
+    },
+    {
+      "epoch": 0.999275887038378,
+      "grad_norm": 0.08685121685266495,
+      "learning_rate": 2.1777766652481916e-08,
+      "loss": 8.501,
+      "step": 200100
+    },
+    {
+      "epoch": 0.9993258258633174,
+      "grad_norm": 0.091061532497406,
+      "learning_rate": 2.0275851710931436e-08,
+      "loss": 8.5031,
+      "step": 200110
+    },
+    {
+      "epoch": 0.9993757646882568,
+      "grad_norm": 0.08946952223777771,
+      "learning_rate": 1.877393676938096e-08,
+      "loss": 8.5099,
+      "step": 200120
+    },
+    {
+      "epoch": 0.9994257035131964,
+      "grad_norm": 0.09155212342739105,
+      "learning_rate": 1.7272021827830483e-08,
+      "loss": 8.4807,
+      "step": 200130
+    },
+    {
+      "epoch": 0.9994756423381358,
+      "grad_norm": 0.09216777235269547,
+      "learning_rate": 1.5770106886280006e-08,
+      "loss": 8.5083,
+      "step": 200140
+    },
+    {
+      "epoch": 0.9995255811630752,
+      "grad_norm": 0.08744364976882935,
+      "learning_rate": 1.4268191944729531e-08,
+      "loss": 8.4995,
+      "step": 200150
+    },
+    {
+      "epoch": 0.9995755199880146,
+      "grad_norm": 0.09221003204584122,
+      "learning_rate": 1.2766277003179054e-08,
+      "loss": 8.493,
+      "step": 200160
+    },
+    {
+      "epoch": 0.9996254588129542,
+      "grad_norm": 0.09411770850419998,
+      "learning_rate": 1.1264362061628577e-08,
+      "loss": 8.4741,
+      "step": 200170
+    },
+    {
+      "epoch": 0.9996753976378936,
+      "grad_norm": 0.09217695146799088,
+      "learning_rate": 9.762447120078099e-09,
+      "loss": 8.4893,
+      "step": 200180
+    },
+    {
+      "epoch": 0.999725336462833,
+      "grad_norm": 0.0921444296836853,
+      "learning_rate": 8.260532178527622e-09,
+      "loss": 8.4998,
+      "step": 200190
+    },
+    {
+      "epoch": 0.9997752752877724,
+      "grad_norm": 0.0900796502828598,
+      "learning_rate": 6.758617236977146e-09,
+      "loss": 8.4963,
+      "step": 200200
+    },
+    {
+      "epoch": 0.999825214112712,
+      "grad_norm": 0.08964385092258453,
+      "learning_rate": 5.2567022954266695e-09,
+      "loss": 8.4971,
+      "step": 200210
+    },
+    {
+      "epoch": 0.9998751529376514,
+      "grad_norm": 0.08609479665756226,
+      "learning_rate": 3.754787353876192e-09,
+      "loss": 8.4869,
+      "step": 200220
+    },
+    {
+      "epoch": 0.9999250917625908,
+      "grad_norm": 0.08690956234931946,
+      "learning_rate": 2.252872412325715e-09,
+      "loss": 8.4845,
+      "step": 200230
+    },
+    {
+      "epoch": 0.9999750305875302,
+      "grad_norm": 0.09233241528272629,
+      "learning_rate": 7.509574707752385e-10,
+      "loss": 8.4693,
+      "step": 200240
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 200245,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 9.875112543817114e+16,
+  "train_batch_size": 32,
+  "trial_name": null,
+  "trial_params": null
+}