diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,16637 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 2000,
+  "global_step": 20638,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00048454307587944567,
+      "grad_norm": 0.4255552887916565,
+      "learning_rate": 8.064516129032258e-07,
+      "loss": 1.221,
+      "num_input_tokens_seen": 392540,
+      "step": 10
+    },
+    {
+      "epoch": 0.0009690861517588913,
+      "grad_norm": 0.4066425859928131,
+      "learning_rate": 1.6129032258064516e-06,
+      "loss": 1.2327,
+      "num_input_tokens_seen": 788876,
+      "step": 20
+    },
+    {
+      "epoch": 0.0014536292276383371,
+      "grad_norm": 0.37268057465553284,
+      "learning_rate": 2.4193548387096776e-06,
+      "loss": 1.2174,
+      "num_input_tokens_seen": 1187392,
+      "step": 30
+    },
+    {
+      "epoch": 0.0019381723035177827,
+      "grad_norm": 0.38770782947540283,
+      "learning_rate": 3.225806451612903e-06,
+      "loss": 1.1503,
+      "num_input_tokens_seen": 1556468,
+      "step": 40
+    },
+    {
+      "epoch": 0.0024227153793972282,
+      "grad_norm": 0.40057116746902466,
+      "learning_rate": 4.032258064516129e-06,
+      "loss": 1.2156,
+      "num_input_tokens_seen": 1950296,
+      "step": 50
+    },
+    {
+      "epoch": 0.0029072584552766742,
+      "grad_norm": 0.4051766097545624,
+      "learning_rate": 4.838709677419355e-06,
+      "loss": 1.2149,
+      "num_input_tokens_seen": 2336152,
+      "step": 60
+    },
+    {
+      "epoch": 0.00339180153115612,
+      "grad_norm": 0.41299352049827576,
+      "learning_rate": 5.64516129032258e-06,
+      "loss": 1.19,
+      "num_input_tokens_seen": 2728108,
+      "step": 70
+    },
+    {
+      "epoch": 0.0038763446070355654,
+      "grad_norm": 0.4118805527687073,
+      "learning_rate": 6.451612903225806e-06,
+      "loss": 1.1994,
+      "num_input_tokens_seen": 3111872,
+      "step": 80
+    },
+    {
+      "epoch": 0.004360887682915011,
+      "grad_norm": 0.42130813002586365,
+      "learning_rate": 7.258064516129033e-06,
+      "loss": 1.2706,
+      "num_input_tokens_seen": 3506908,
+      "step": 90
+    },
+    {
+      "epoch": 0.0048454307587944565,
+      "grad_norm": 0.3949407637119293,
+      "learning_rate": 8.064516129032258e-06,
+      "loss": 1.187,
+      "num_input_tokens_seen": 3920340,
+      "step": 100
+    },
+    {
+      "epoch": 0.005329973834673903,
+      "grad_norm": 0.42650577425956726,
+      "learning_rate": 8.870967741935484e-06,
+      "loss": 1.2195,
+      "num_input_tokens_seen": 4303752,
+      "step": 110
+    },
+    {
+      "epoch": 0.0058145169105533485,
+      "grad_norm": 0.43526706099510193,
+      "learning_rate": 9.67741935483871e-06,
+      "loss": 1.2141,
+      "num_input_tokens_seen": 4684992,
+      "step": 120
+    },
+    {
+      "epoch": 0.006299059986432794,
+      "grad_norm": 0.4030572474002838,
+      "learning_rate": 1.0483870967741936e-05,
+      "loss": 1.2873,
+      "num_input_tokens_seen": 5052616,
+      "step": 130
+    },
+    {
+      "epoch": 0.00678360306231224,
+      "grad_norm": 0.41993653774261475,
+      "learning_rate": 1.129032258064516e-05,
+      "loss": 1.2175,
+      "num_input_tokens_seen": 5427828,
+      "step": 140
+    },
+    {
+      "epoch": 0.007268146138191685,
+      "grad_norm": 0.4032376706600189,
+      "learning_rate": 1.2096774193548388e-05,
+      "loss": 1.2747,
+      "num_input_tokens_seen": 5819148,
+      "step": 150
+    },
+    {
+      "epoch": 0.007752689214071131,
+      "grad_norm": 0.4497855603694916,
+      "learning_rate": 1.2903225806451613e-05,
+      "loss": 1.2422,
+      "num_input_tokens_seen": 6198028,
+      "step": 160
+    },
+    {
+      "epoch": 0.008237232289950576,
+      "grad_norm": 0.4343348741531372,
+      "learning_rate": 1.3709677419354839e-05,
+      "loss": 1.2621,
+      "num_input_tokens_seen": 6604716,
+      "step": 170
+    },
+    {
+      "epoch": 0.008721775365830022,
+      "grad_norm": 0.4532233774662018,
+      "learning_rate": 1.4516129032258066e-05,
+      "loss": 1.2263,
+      "num_input_tokens_seen": 6995768,
+      "step": 180
+    },
+    {
+      "epoch": 0.009206318441709467,
+      "grad_norm": 0.4360114634037018,
+      "learning_rate": 1.5322580645161292e-05,
+      "loss": 1.2022,
+      "num_input_tokens_seen": 7395348,
+      "step": 190
+    },
+    {
+      "epoch": 0.009690861517588913,
+      "grad_norm": 0.4088006913661957,
+      "learning_rate": 1.6129032258064517e-05,
+      "loss": 1.2199,
+      "num_input_tokens_seen": 7783256,
+      "step": 200
+    },
+    {
+      "epoch": 0.010175404593468359,
+      "grad_norm": 0.4394921064376831,
+      "learning_rate": 1.693548387096774e-05,
+      "loss": 1.2227,
+      "num_input_tokens_seen": 8183636,
+      "step": 210
+    },
+    {
+      "epoch": 0.010659947669347806,
+      "grad_norm": 0.4413774311542511,
+      "learning_rate": 1.774193548387097e-05,
+      "loss": 1.2134,
+      "num_input_tokens_seen": 8541160,
+      "step": 220
+    },
+    {
+      "epoch": 0.011144490745227251,
+      "grad_norm": 0.4048616588115692,
+      "learning_rate": 1.8548387096774193e-05,
+      "loss": 1.1969,
+      "num_input_tokens_seen": 8913056,
+      "step": 230
+    },
+    {
+      "epoch": 0.011629033821106697,
+      "grad_norm": 0.4245803654193878,
+      "learning_rate": 1.935483870967742e-05,
+      "loss": 1.2161,
+      "num_input_tokens_seen": 9304116,
+      "step": 240
+    },
+    {
+      "epoch": 0.012113576896986143,
+      "grad_norm": 0.4485531151294708,
+      "learning_rate": 2.0161290322580645e-05,
+      "loss": 1.1966,
+      "num_input_tokens_seen": 9690224,
+      "step": 250
+    },
+    {
+      "epoch": 0.012598119972865588,
+      "grad_norm": 0.44514381885528564,
+      "learning_rate": 2.0967741935483873e-05,
+      "loss": 1.2736,
+      "num_input_tokens_seen": 10093420,
+      "step": 260
+    },
+    {
+      "epoch": 0.013082663048745034,
+      "grad_norm": 0.47046637535095215,
+      "learning_rate": 2.1774193548387097e-05,
+      "loss": 1.2217,
+      "num_input_tokens_seen": 10480092,
+      "step": 270
+    },
+    {
+      "epoch": 0.01356720612462448,
+      "grad_norm": 0.4320948123931885,
+      "learning_rate": 2.258064516129032e-05,
+      "loss": 1.2185,
+      "num_input_tokens_seen": 10856260,
+      "step": 280
+    },
+    {
+      "epoch": 0.014051749200503925,
+      "grad_norm": 0.418040931224823,
+      "learning_rate": 2.338709677419355e-05,
+      "loss": 1.2504,
+      "num_input_tokens_seen": 11231836,
+      "step": 290
+    },
+    {
+      "epoch": 0.01453629227638337,
+      "grad_norm": 0.4360697865486145,
+      "learning_rate": 2.4193548387096777e-05,
+      "loss": 1.2199,
+      "num_input_tokens_seen": 11622080,
+      "step": 300
+    },
+    {
+      "epoch": 0.015020835352262816,
+      "grad_norm": 0.4039739668369293,
+      "learning_rate": 2.5e-05,
+      "loss": 1.1947,
+      "num_input_tokens_seen": 12009244,
+      "step": 310
+    },
+    {
+      "epoch": 0.015505378428142261,
+      "grad_norm": 0.4114651381969452,
+      "learning_rate": 2.5806451612903226e-05,
+      "loss": 1.2345,
+      "num_input_tokens_seen": 12404068,
+      "step": 320
+    },
+    {
+      "epoch": 0.015989921504021707,
+      "grad_norm": 0.44177761673927307,
+      "learning_rate": 2.661290322580645e-05,
+      "loss": 1.2303,
+      "num_input_tokens_seen": 12774432,
+      "step": 330
+    },
+    {
+      "epoch": 0.016474464579901153,
+      "grad_norm": 0.4257271885871887,
+      "learning_rate": 2.7419354838709678e-05,
+      "loss": 1.2513,
+      "num_input_tokens_seen": 13155544,
+      "step": 340
+    },
+    {
+      "epoch": 0.016959007655780598,
+      "grad_norm": 0.4340864419937134,
+      "learning_rate": 2.822580645161291e-05,
+      "loss": 1.2486,
+      "num_input_tokens_seen": 13554616,
+      "step": 350
+    },
+    {
+      "epoch": 0.017443550731660044,
+      "grad_norm": 0.4148235023021698,
+      "learning_rate": 2.9032258064516133e-05,
+      "loss": 1.2193,
+      "num_input_tokens_seen": 13937664,
+      "step": 360
+    },
+    {
+      "epoch": 0.01792809380753949,
+      "grad_norm": 0.417041152715683,
+      "learning_rate": 2.9838709677419357e-05,
+      "loss": 1.2142,
+      "num_input_tokens_seen": 14331280,
+      "step": 370
+    },
+    {
+      "epoch": 0.018412636883418935,
+      "grad_norm": 0.39417609572410583,
+      "learning_rate": 3.0645161290322585e-05,
+      "loss": 1.2865,
+      "num_input_tokens_seen": 14707672,
+      "step": 380
+    },
+    {
+      "epoch": 0.01889717995929838,
+      "grad_norm": 0.4522784948348999,
+      "learning_rate": 3.1451612903225806e-05,
+      "loss": 1.2437,
+      "num_input_tokens_seen": 15106840,
+      "step": 390
+    },
+    {
+      "epoch": 0.019381723035177826,
+      "grad_norm": 0.43782681226730347,
+      "learning_rate": 3.2258064516129034e-05,
+      "loss": 1.2215,
+      "num_input_tokens_seen": 15480172,
+      "step": 400
+    },
+    {
+      "epoch": 0.01986626611105727,
+      "grad_norm": 0.4035712480545044,
+      "learning_rate": 3.306451612903226e-05,
+      "loss": 1.2133,
+      "num_input_tokens_seen": 15879184,
+      "step": 410
+    },
+    {
+      "epoch": 0.020350809186936717,
+      "grad_norm": 0.4507812261581421,
+      "learning_rate": 3.387096774193548e-05,
+      "loss": 1.2469,
+      "num_input_tokens_seen": 16282872,
+      "step": 420
+    },
+    {
+      "epoch": 0.020835352262816166,
+      "grad_norm": 0.41557714343070984,
+      "learning_rate": 3.467741935483872e-05,
+      "loss": 1.2694,
+      "num_input_tokens_seen": 16649572,
+      "step": 430
+    },
+    {
+      "epoch": 0.02131989533869561,
+      "grad_norm": 0.43545469641685486,
+      "learning_rate": 3.548387096774194e-05,
+      "loss": 1.2469,
+      "num_input_tokens_seen": 17047428,
+      "step": 440
+    },
+    {
+      "epoch": 0.021804438414575057,
+      "grad_norm": 0.46730995178222656,
+      "learning_rate": 3.6290322580645165e-05,
+      "loss": 1.2082,
+      "num_input_tokens_seen": 17447240,
+      "step": 450
+    },
+    {
+      "epoch": 0.022288981490454503,
+      "grad_norm": 0.45304545760154724,
+      "learning_rate": 3.7096774193548386e-05,
+      "loss": 1.1941,
+      "num_input_tokens_seen": 17831524,
+      "step": 460
+    },
+    {
+      "epoch": 0.02277352456633395,
+      "grad_norm": 0.44148966670036316,
+      "learning_rate": 3.7903225806451614e-05,
+      "loss": 1.2507,
+      "num_input_tokens_seen": 18226524,
+      "step": 470
+    },
+    {
+      "epoch": 0.023258067642213394,
+      "grad_norm": 0.46780791878700256,
+      "learning_rate": 3.870967741935484e-05,
+      "loss": 1.2136,
+      "num_input_tokens_seen": 18621280,
+      "step": 480
+    },
+    {
+      "epoch": 0.02374261071809284,
+      "grad_norm": 0.40209197998046875,
+      "learning_rate": 3.951612903225806e-05,
+      "loss": 1.3024,
+      "num_input_tokens_seen": 19023172,
+      "step": 490
+    },
+    {
+      "epoch": 0.024227153793972285,
+      "grad_norm": 0.45693570375442505,
+      "learning_rate": 4.032258064516129e-05,
+      "loss": 1.2317,
+      "num_input_tokens_seen": 19384076,
+      "step": 500
+    },
+    {
+      "epoch": 0.02471169686985173,
+      "grad_norm": 0.4480474293231964,
+      "learning_rate": 4.112903225806452e-05,
+      "loss": 1.244,
+      "num_input_tokens_seen": 19772672,
+      "step": 510
+    },
+    {
+      "epoch": 0.025196239945731176,
+      "grad_norm": 0.3898276686668396,
+      "learning_rate": 4.1935483870967746e-05,
+      "loss": 1.2248,
+      "num_input_tokens_seen": 20147580,
+      "step": 520
+    },
+    {
+      "epoch": 0.02568078302161062,
+      "grad_norm": 0.45244449377059937,
+      "learning_rate": 4.2741935483870973e-05,
+      "loss": 1.2479,
+      "num_input_tokens_seen": 20538480,
+      "step": 530
+    },
+    {
+      "epoch": 0.026165326097490067,
+      "grad_norm": 0.4444933235645294,
+      "learning_rate": 4.3548387096774194e-05,
+      "loss": 1.2349,
+      "num_input_tokens_seen": 20938472,
+      "step": 540
+    },
+    {
+      "epoch": 0.026649869173369513,
+      "grad_norm": 0.4245850741863251,
+      "learning_rate": 4.435483870967742e-05,
+      "loss": 1.2061,
+      "num_input_tokens_seen": 21323560,
+      "step": 550
+    },
+    {
+      "epoch": 0.02713441224924896,
+      "grad_norm": 0.4357333481311798,
+      "learning_rate": 4.516129032258064e-05,
+      "loss": 1.2532,
+      "num_input_tokens_seen": 21725356,
+      "step": 560
+    },
+    {
+      "epoch": 0.027618955325128404,
+      "grad_norm": 0.4161827564239502,
+      "learning_rate": 4.596774193548387e-05,
+      "loss": 1.241,
+      "num_input_tokens_seen": 22107816,
+      "step": 570
+    },
+    {
+      "epoch": 0.02810349840100785,
+      "grad_norm": 0.44058066606521606,
+      "learning_rate": 4.67741935483871e-05,
+      "loss": 1.284,
+      "num_input_tokens_seen": 22490484,
+      "step": 580
+    },
+    {
+      "epoch": 0.028588041476887295,
+      "grad_norm": 0.4422559142112732,
+      "learning_rate": 4.7580645161290326e-05,
+      "loss": 1.2633,
+      "num_input_tokens_seen": 22890776,
+      "step": 590
+    },
+    {
+      "epoch": 0.02907258455276674,
+      "grad_norm": 0.4280403256416321,
+      "learning_rate": 4.8387096774193554e-05,
+      "loss": 1.2649,
+      "num_input_tokens_seen": 23274324,
+      "step": 600
+    },
+    {
+      "epoch": 0.029557127628646186,
+      "grad_norm": 0.4213055968284607,
+      "learning_rate": 4.9193548387096775e-05,
+      "loss": 1.2544,
+      "num_input_tokens_seen": 23668904,
+      "step": 610
+    },
+    {
+      "epoch": 0.030041670704525632,
+      "grad_norm": 0.4733617901802063,
+      "learning_rate": 5e-05,
+      "loss": 1.2238,
+      "num_input_tokens_seen": 24062116,
+      "step": 620
+    },
+    {
+      "epoch": 0.030526213780405077,
+      "grad_norm": 0.4112738370895386,
+      "learning_rate": 4.999996921293424e-05,
+      "loss": 1.2544,
+      "num_input_tokens_seen": 24430924,
+      "step": 630
+    },
+    {
+      "epoch": 0.031010756856284523,
+      "grad_norm": 0.3989699184894562,
+      "learning_rate": 4.999987685181276e-05,
+      "loss": 1.2679,
+      "num_input_tokens_seen": 24803116,
+      "step": 640
+    },
+    {
+      "epoch": 0.03149529993216397,
+      "grad_norm": 0.46107640862464905,
+      "learning_rate": 4.9999722916863064e-05,
+      "loss": 1.2498,
+      "num_input_tokens_seen": 25174648,
+      "step": 650
+    },
+    {
+      "epoch": 0.031979843008043414,
+      "grad_norm": 0.4159105122089386,
+      "learning_rate": 4.999950740846427e-05,
+      "loss": 1.2148,
+      "num_input_tokens_seen": 25572492,
+      "step": 660
+    },
+    {
+      "epoch": 0.03246438608392286,
+      "grad_norm": 0.41995078325271606,
+      "learning_rate": 4.9999230327147187e-05,
+      "loss": 1.2921,
+      "num_input_tokens_seen": 25970816,
+      "step": 670
+    },
+    {
+      "epoch": 0.032948929159802305,
+      "grad_norm": 0.4316340386867523,
+      "learning_rate": 4.999889167359425e-05,
+      "loss": 1.2401,
+      "num_input_tokens_seen": 26346612,
+      "step": 680
+    },
+    {
+      "epoch": 0.03343347223568175,
+      "grad_norm": 0.4364665448665619,
+      "learning_rate": 4.999849144863954e-05,
+      "loss": 1.228,
+      "num_input_tokens_seen": 26750808,
+      "step": 690
+    },
+    {
+      "epoch": 0.033918015311561196,
+      "grad_norm": 0.418192595243454,
+      "learning_rate": 4.9998029653268805e-05,
+      "loss": 1.2565,
+      "num_input_tokens_seen": 27144672,
+      "step": 700
+    },
+    {
+      "epoch": 0.03440255838744064,
+      "grad_norm": 0.42366668581962585,
+      "learning_rate": 4.9997506288619436e-05,
+      "loss": 1.3009,
+      "num_input_tokens_seen": 27537568,
+      "step": 710
+    },
+    {
+      "epoch": 0.03488710146332009,
+      "grad_norm": 0.40567639470100403,
+      "learning_rate": 4.9996921355980456e-05,
+      "loss": 1.1824,
+      "num_input_tokens_seen": 27934864,
+      "step": 720
+    },
+    {
+      "epoch": 0.03537164453919953,
+      "grad_norm": 0.4429372549057007,
+      "learning_rate": 4.999627485679254e-05,
+      "loss": 1.2941,
+      "num_input_tokens_seen": 28315316,
+      "step": 730
+    },
+    {
+      "epoch": 0.03585618761507898,
+      "grad_norm": 0.4326581358909607,
+      "learning_rate": 4.999556679264798e-05,
+      "loss": 1.2178,
+      "num_input_tokens_seen": 28695628,
+      "step": 740
+    },
+    {
+      "epoch": 0.036340730690958424,
+      "grad_norm": 0.39658111333847046,
+      "learning_rate": 4.9994797165290724e-05,
+      "loss": 1.2194,
+      "num_input_tokens_seen": 29075004,
+      "step": 750
+    },
+    {
+      "epoch": 0.03682527376683787,
+      "grad_norm": 0.42921292781829834,
+      "learning_rate": 4.999396597661634e-05,
+      "loss": 1.262,
+      "num_input_tokens_seen": 29452648,
+      "step": 760
+    },
+    {
+      "epoch": 0.037309816842717315,
+      "grad_norm": 0.42564624547958374,
+      "learning_rate": 4.999307322867201e-05,
+      "loss": 1.2686,
+      "num_input_tokens_seen": 29806040,
+      "step": 770
+    },
+    {
+      "epoch": 0.03779435991859676,
+      "grad_norm": 0.42206162214279175,
+      "learning_rate": 4.9992118923656525e-05,
+      "loss": 1.2644,
+      "num_input_tokens_seen": 30187964,
+      "step": 780
+    },
+    {
+      "epoch": 0.038278902994476206,
+      "grad_norm": 0.40607237815856934,
+      "learning_rate": 4.999110306392034e-05,
+      "loss": 1.2095,
+      "num_input_tokens_seen": 30595388,
+      "step": 790
+    },
+    {
+      "epoch": 0.03876344607035565,
+      "grad_norm": 0.4239601492881775,
+      "learning_rate": 4.999002565196546e-05,
+      "loss": 1.2785,
+      "num_input_tokens_seen": 30969404,
+      "step": 800
+    },
+    {
+      "epoch": 0.0392479891462351,
+      "grad_norm": 0.37221038341522217,
+      "learning_rate": 4.9988886690445524e-05,
+      "loss": 1.2645,
+      "num_input_tokens_seen": 31352172,
+      "step": 810
+    },
+    {
+      "epoch": 0.03973253222211454,
+      "grad_norm": 0.42992502450942993,
+      "learning_rate": 4.998768618216575e-05,
+      "loss": 1.2646,
+      "num_input_tokens_seen": 31707436,
+      "step": 820
+    },
+    {
+      "epoch": 0.04021707529799399,
+      "grad_norm": 0.40228039026260376,
+      "learning_rate": 4.998642413008294e-05,
+      "loss": 1.2796,
+      "num_input_tokens_seen": 32113760,
+      "step": 830
+    },
+    {
+      "epoch": 0.040701618373873434,
+      "grad_norm": 0.4015443027019501,
+      "learning_rate": 4.9985100537305494e-05,
+      "loss": 1.271,
+      "num_input_tokens_seen": 32495944,
+      "step": 840
+    },
+    {
+      "epoch": 0.04118616144975288,
+      "grad_norm": 0.4232306480407715,
+      "learning_rate": 4.998371540709338e-05,
+      "loss": 1.2403,
+      "num_input_tokens_seen": 32878288,
+      "step": 850
+    },
+    {
+      "epoch": 0.04167070452563233,
+      "grad_norm": 0.4329904317855835,
+      "learning_rate": 4.998226874285811e-05,
+      "loss": 1.2433,
+      "num_input_tokens_seen": 33283124,
+      "step": 860
+    },
+    {
+      "epoch": 0.04215524760151178,
+      "grad_norm": 0.3939256966114044,
+      "learning_rate": 4.998076054816278e-05,
+      "loss": 1.2513,
+      "num_input_tokens_seen": 33675788,
+      "step": 870
+    },
+    {
+      "epoch": 0.04263979067739122,
+      "grad_norm": 0.40896937251091003,
+      "learning_rate": 4.997919082672201e-05,
+      "loss": 1.2058,
+      "num_input_tokens_seen": 34078876,
+      "step": 880
+    },
+    {
+      "epoch": 0.04312433375327067,
+      "grad_norm": 0.41988757252693176,
+      "learning_rate": 4.997755958240198e-05,
+      "loss": 1.2704,
+      "num_input_tokens_seen": 34476892,
+      "step": 890
+    },
+    {
+      "epoch": 0.043608876829150114,
+      "grad_norm": 0.3970952332019806,
+      "learning_rate": 4.997586681922039e-05,
+      "loss": 1.2592,
+      "num_input_tokens_seen": 34879628,
+      "step": 900
+    },
+    {
+      "epoch": 0.04409341990502956,
+      "grad_norm": 0.3762570917606354,
+      "learning_rate": 4.997411254134645e-05,
+      "loss": 1.2744,
+      "num_input_tokens_seen": 35263368,
+      "step": 910
+    },
+    {
+      "epoch": 0.044577962980909006,
+      "grad_norm": 0.44425544142723083,
+      "learning_rate": 4.9972296753100875e-05,
+      "loss": 1.27,
+      "num_input_tokens_seen": 35669980,
+      "step": 920
+    },
+    {
+      "epoch": 0.04506250605678845,
+      "grad_norm": 0.3981640636920929,
+      "learning_rate": 4.9970419458955916e-05,
+      "loss": 1.2627,
+      "num_input_tokens_seen": 36075236,
+      "step": 930
+    },
+    {
+      "epoch": 0.0455470491326679,
+      "grad_norm": 0.42503446340560913,
+      "learning_rate": 4.996848066353526e-05,
+      "loss": 1.2742,
+      "num_input_tokens_seen": 36460584,
+      "step": 940
+    },
+    {
+      "epoch": 0.04603159220854734,
+      "grad_norm": 0.44177576899528503,
+      "learning_rate": 4.99664803716141e-05,
+      "loss": 1.2338,
+      "num_input_tokens_seen": 36844140,
+      "step": 950
+    },
+    {
+      "epoch": 0.04651613528442679,
+      "grad_norm": 0.4258316159248352,
+      "learning_rate": 4.996441858811909e-05,
+      "loss": 1.251,
+      "num_input_tokens_seen": 37245224,
+      "step": 960
+    },
+    {
+      "epoch": 0.04700067836030623,
+      "grad_norm": 0.39270251989364624,
+      "learning_rate": 4.996229531812833e-05,
+      "loss": 1.2682,
+      "num_input_tokens_seen": 37623120,
+      "step": 970
+    },
+    {
+      "epoch": 0.04748522143618568,
+      "grad_norm": 0.40389484167099,
+      "learning_rate": 4.996011056687135e-05,
+      "loss": 1.2599,
+      "num_input_tokens_seen": 37992076,
+      "step": 980
+    },
+    {
+      "epoch": 0.047969764512065124,
+      "grad_norm": 0.4183795154094696,
+      "learning_rate": 4.9957864339729126e-05,
+      "loss": 1.2647,
+      "num_input_tokens_seen": 38392008,
+      "step": 990
+    },
+    {
+      "epoch": 0.04845430758794457,
+      "grad_norm": 0.3972471356391907,
+      "learning_rate": 4.9955556642234034e-05,
+      "loss": 1.2387,
+      "num_input_tokens_seen": 38802860,
+      "step": 1000
+    },
+    {
+      "epoch": 0.048938850663824016,
+      "grad_norm": 0.3918885886669159,
+      "learning_rate": 4.9953187480069854e-05,
+      "loss": 1.2611,
+      "num_input_tokens_seen": 39175600,
+      "step": 1010
+    },
+    {
+      "epoch": 0.04942339373970346,
+      "grad_norm": 0.4439409673213959,
+      "learning_rate": 4.9950756859071755e-05,
+      "loss": 1.1817,
+      "num_input_tokens_seen": 39586740,
+      "step": 1020
+    },
+    {
+      "epoch": 0.04990793681558291,
+      "grad_norm": 0.43080589175224304,
+      "learning_rate": 4.994826478522626e-05,
+      "loss": 1.2453,
+      "num_input_tokens_seen": 39993292,
+      "step": 1030
+    },
+    {
+      "epoch": 0.05039247989146235,
+      "grad_norm": 0.41559743881225586,
+      "learning_rate": 4.9945711264671276e-05,
+      "loss": 1.2635,
+      "num_input_tokens_seen": 40412108,
+      "step": 1040
+    },
+    {
+      "epoch": 0.0508770229673418,
+      "grad_norm": 0.4444926381111145,
+      "learning_rate": 4.994309630369602e-05,
+      "loss": 1.2742,
+      "num_input_tokens_seen": 40815268,
+      "step": 1050
+    },
+    {
+      "epoch": 0.05136156604322124,
+      "grad_norm": 0.43432703614234924,
+      "learning_rate": 4.9940419908741065e-05,
+      "loss": 1.2782,
+      "num_input_tokens_seen": 41228008,
+      "step": 1060
+    },
+    {
+      "epoch": 0.05184610911910069,
+      "grad_norm": 0.46701622009277344,
+      "learning_rate": 4.993768208639826e-05,
+      "loss": 1.1921,
+      "num_input_tokens_seen": 41641820,
+      "step": 1070
+    },
+    {
+      "epoch": 0.052330652194980135,
+      "grad_norm": 0.42707207798957825,
+      "learning_rate": 4.993488284341078e-05,
+      "loss": 1.221,
+      "num_input_tokens_seen": 42035360,
+      "step": 1080
+    },
+    {
+      "epoch": 0.05281519527085958,
+      "grad_norm": 0.4334350824356079,
+      "learning_rate": 4.993202218667307e-05,
+      "loss": 1.2273,
+      "num_input_tokens_seen": 42437020,
+      "step": 1090
+    },
+    {
+      "epoch": 0.053299738346739026,
+      "grad_norm": 0.3986518681049347,
+      "learning_rate": 4.99291001232308e-05,
+      "loss": 1.2333,
+      "num_input_tokens_seen": 42804100,
+      "step": 1100
+    },
+    {
+      "epoch": 0.05378428142261847,
+      "grad_norm": 0.4267319142818451,
+      "learning_rate": 4.992611666028094e-05,
+      "loss": 1.2582,
+      "num_input_tokens_seen": 43211524,
+      "step": 1110
+    },
+    {
+      "epoch": 0.05426882449849792,
+      "grad_norm": 0.4689897894859314,
+      "learning_rate": 4.992307180517165e-05,
+      "loss": 1.2354,
+      "num_input_tokens_seen": 43626756,
+      "step": 1120
+    },
+    {
+      "epoch": 0.05475336757437736,
+      "grad_norm": 0.45038825273513794,
+      "learning_rate": 4.991996556540229e-05,
+      "loss": 1.2588,
+      "num_input_tokens_seen": 44045992,
+      "step": 1130
+    },
+    {
+      "epoch": 0.05523791065025681,
+      "grad_norm": 0.4214176535606384,
+      "learning_rate": 4.991679794862343e-05,
+      "loss": 1.1964,
+      "num_input_tokens_seen": 44421180,
+      "step": 1140
+    },
+    {
+      "epoch": 0.055722453726136253,
+      "grad_norm": 0.41900694370269775,
+      "learning_rate": 4.99135689626368e-05,
+      "loss": 1.2565,
+      "num_input_tokens_seen": 44831096,
+      "step": 1150
+    },
+    {
+      "epoch": 0.0562069968020157,
+      "grad_norm": 0.39536434412002563,
+      "learning_rate": 4.9910278615395276e-05,
+      "loss": 1.2181,
+      "num_input_tokens_seen": 45272232,
+      "step": 1160
+    },
+    {
+      "epoch": 0.056691539877895145,
+      "grad_norm": 0.4110620319843292,
+      "learning_rate": 4.9906926915002875e-05,
+      "loss": 1.2794,
+      "num_input_tokens_seen": 45672992,
+      "step": 1170
+    },
+    {
+      "epoch": 0.05717608295377459,
+      "grad_norm": 0.4034215211868286,
+      "learning_rate": 4.9903513869714704e-05,
+      "loss": 1.2561,
+      "num_input_tokens_seen": 46085848,
+      "step": 1180
+    },
+    {
+      "epoch": 0.057660626029654036,
+      "grad_norm": 0.456302672624588,
+      "learning_rate": 4.990003948793699e-05,
+      "loss": 1.258,
+      "num_input_tokens_seen": 46486968,
+      "step": 1190
+    },
+    {
+      "epoch": 0.05814516910553348,
+      "grad_norm": 0.4716334044933319,
+      "learning_rate": 4.989650377822702e-05,
+      "loss": 1.2495,
+      "num_input_tokens_seen": 46892488,
+      "step": 1200
+    },
+    {
+      "epoch": 0.05862971218141293,
+      "grad_norm": 0.4383469820022583,
+      "learning_rate": 4.98929067492931e-05,
+      "loss": 1.2278,
+      "num_input_tokens_seen": 47271524,
+      "step": 1210
+    },
+    {
+      "epoch": 0.05911425525729237,
+      "grad_norm": 0.4059995710849762,
+      "learning_rate": 4.988924840999462e-05,
+      "loss": 1.2907,
+      "num_input_tokens_seen": 47666308,
+      "step": 1220
+    },
+    {
+      "epoch": 0.05959879833317182,
+      "grad_norm": 0.42955759167671204,
+      "learning_rate": 4.9885528769341905e-05,
+      "loss": 1.2455,
+      "num_input_tokens_seen": 48036456,
+      "step": 1230
+    },
+    {
+      "epoch": 0.060083341409051264,
+      "grad_norm": 0.43056222796440125,
+      "learning_rate": 4.988174783649633e-05,
+      "loss": 1.2205,
+      "num_input_tokens_seen": 48426548,
+      "step": 1240
+    },
+    {
+      "epoch": 0.06056788448493071,
+      "grad_norm": 0.408191442489624,
+      "learning_rate": 4.987790562077019e-05,
+      "loss": 1.2517,
+      "num_input_tokens_seen": 48803056,
+      "step": 1250
+    },
+    {
+      "epoch": 0.061052427560810155,
+      "grad_norm": 0.3784200847148895,
+      "learning_rate": 4.987400213162673e-05,
+      "loss": 1.2009,
+      "num_input_tokens_seen": 49204892,
+      "step": 1260
+    },
+    {
+      "epoch": 0.0615369706366896,
+      "grad_norm": 0.4229642450809479,
+      "learning_rate": 4.987003737868011e-05,
+      "loss": 1.2979,
+      "num_input_tokens_seen": 49597520,
+      "step": 1270
+    },
+    {
+      "epoch": 0.062021513712569046,
+      "grad_norm": 0.4075496792793274,
+      "learning_rate": 4.9866011371695374e-05,
+      "loss": 1.2041,
+      "num_input_tokens_seen": 49973816,
+      "step": 1280
+    },
+    {
+      "epoch": 0.06250605678844849,
+      "grad_norm": 0.4234859049320221,
+      "learning_rate": 4.9861924120588445e-05,
+      "loss": 1.2489,
+      "num_input_tokens_seen": 50382180,
+      "step": 1290
+    },
+    {
+      "epoch": 0.06299059986432794,
+      "grad_norm": 0.4083632528781891,
+      "learning_rate": 4.985777563542607e-05,
+      "loss": 1.2459,
+      "num_input_tokens_seen": 50746876,
+      "step": 1300
+    },
+    {
+      "epoch": 0.06347514294020738,
+      "grad_norm": 0.4187311828136444,
+      "learning_rate": 4.985356592642584e-05,
+      "loss": 1.2366,
+      "num_input_tokens_seen": 51150280,
+      "step": 1310
+    },
+    {
+      "epoch": 0.06395968601608683,
+      "grad_norm": 0.4542880952358246,
+      "learning_rate": 4.984929500395611e-05,
+      "loss": 1.2511,
+      "num_input_tokens_seen": 51527856,
+      "step": 1320
+    },
+    {
+      "epoch": 0.06444422909196627,
+      "grad_norm": 0.40190425515174866,
+      "learning_rate": 4.9844962878536004e-05,
+      "loss": 1.2314,
+      "num_input_tokens_seen": 51885820,
+      "step": 1330
+    },
+    {
+      "epoch": 0.06492877216784572,
+      "grad_norm": 0.3928161859512329,
+      "learning_rate": 4.9840569560835416e-05,
+      "loss": 1.2156,
+      "num_input_tokens_seen": 52276412,
+      "step": 1340
+    },
+    {
+      "epoch": 0.06541331524372516,
+      "grad_norm": 0.407649964094162,
+      "learning_rate": 4.9836115061674925e-05,
+      "loss": 1.2379,
+      "num_input_tokens_seen": 52647892,
+      "step": 1350
+    },
+    {
+      "epoch": 0.06589785831960461,
+      "grad_norm": 0.4136994183063507,
+      "learning_rate": 4.983159939202582e-05,
+      "loss": 1.2633,
+      "num_input_tokens_seen": 53053832,
+      "step": 1360
+    },
+    {
+      "epoch": 0.06638240139548406,
+      "grad_norm": 0.4096525311470032,
+      "learning_rate": 4.9827022563010016e-05,
+      "loss": 1.2265,
+      "num_input_tokens_seen": 53431196,
+      "step": 1370
+    },
+    {
+      "epoch": 0.0668669444713635,
+      "grad_norm": 0.4357371926307678,
+      "learning_rate": 4.982238458590009e-05,
+      "loss": 1.2617,
+      "num_input_tokens_seen": 53833560,
+      "step": 1380
+    },
+    {
+      "epoch": 0.06735148754724295,
+      "grad_norm": 0.4323122799396515,
+      "learning_rate": 4.9817685472119246e-05,
+      "loss": 1.3117,
+      "num_input_tokens_seen": 54217320,
+      "step": 1390
+    },
+    {
+      "epoch": 0.06783603062312239,
+      "grad_norm": 0.4947691261768341,
+      "learning_rate": 4.98129252332412e-05,
+      "loss": 1.2197,
+      "num_input_tokens_seen": 54609344,
+      "step": 1400
+    },
+    {
+      "epoch": 0.06832057369900184,
+      "grad_norm": 0.43488138914108276,
+      "learning_rate": 4.980810388099028e-05,
+      "loss": 1.256,
+      "num_input_tokens_seen": 55028116,
+      "step": 1410
+    },
+    {
+      "epoch": 0.06880511677488128,
+      "grad_norm": 0.4741380512714386,
+      "learning_rate": 4.980322142724129e-05,
+      "loss": 1.2525,
+      "num_input_tokens_seen": 55416832,
+      "step": 1420
+    },
+    {
+      "epoch": 0.06928965985076073,
+      "grad_norm": 0.4468291103839874,
+      "learning_rate": 4.979827788401956e-05,
+      "loss": 1.2414,
+      "num_input_tokens_seen": 55798836,
+      "step": 1430
+    },
+    {
+      "epoch": 0.06977420292664017,
+      "grad_norm": 0.42211663722991943,
+      "learning_rate": 4.979327326350086e-05,
+      "loss": 1.2589,
+      "num_input_tokens_seen": 56187416,
+      "step": 1440
+    },
+    {
+      "epoch": 0.07025874600251962,
+      "grad_norm": 0.4140244722366333,
+      "learning_rate": 4.9788207578011405e-05,
+      "loss": 1.239,
+      "num_input_tokens_seen": 56577512,
+      "step": 1450
+    },
+    {
+      "epoch": 0.07074328907839907,
+      "grad_norm": 0.39189738035202026,
+      "learning_rate": 4.978308084002779e-05,
+      "loss": 1.1905,
+      "num_input_tokens_seen": 56978600,
+      "step": 1460
+    },
+    {
+      "epoch": 0.07122783215427851,
+      "grad_norm": 0.3976607024669647,
+      "learning_rate": 4.9777893062176986e-05,
+      "loss": 1.252,
+      "num_input_tokens_seen": 57367068,
+      "step": 1470
+    },
+    {
+      "epoch": 0.07171237523015796,
+      "grad_norm": 0.3938475251197815,
+      "learning_rate": 4.977264425723632e-05,
+      "loss": 1.2644,
+      "num_input_tokens_seen": 57756716,
+      "step": 1480
+    },
+    {
+      "epoch": 0.0721969183060374,
+      "grad_norm": 0.3958350718021393,
+      "learning_rate": 4.976733443813343e-05,
+      "loss": 1.2834,
+      "num_input_tokens_seen": 58127588,
+      "step": 1490
+    },
+    {
+      "epoch": 0.07268146138191685,
+      "grad_norm": 0.41049420833587646,
+      "learning_rate": 4.976196361794619e-05,
+      "loss": 1.2359,
+      "num_input_tokens_seen": 58531680,
+      "step": 1500
+    },
+    {
+      "epoch": 0.0731660044577963,
+      "grad_norm": 0.40144941210746765,
+      "learning_rate": 4.9756531809902765e-05,
+      "loss": 1.2358,
+      "num_input_tokens_seen": 58917960,
+      "step": 1510
+    },
+    {
+      "epoch": 0.07365054753367574,
+      "grad_norm": 0.4596763551235199,
+      "learning_rate": 4.975103902738149e-05,
+      "loss": 1.2676,
+      "num_input_tokens_seen": 59324348,
+      "step": 1520
+    },
+    {
+      "epoch": 0.07413509060955518,
+      "grad_norm": 0.41802385449409485,
+      "learning_rate": 4.974548528391091e-05,
+      "loss": 1.2669,
+      "num_input_tokens_seen": 59734700,
+      "step": 1530
+    },
+    {
+      "epoch": 0.07461963368543463,
+      "grad_norm": 0.409807026386261,
+      "learning_rate": 4.9739870593169705e-05,
+      "loss": 1.2579,
+      "num_input_tokens_seen": 60135636,
+      "step": 1540
+    },
+    {
+      "epoch": 0.07510417676131408,
+      "grad_norm": 0.41342952847480774,
+      "learning_rate": 4.9734194968986656e-05,
+      "loss": 1.2319,
+      "num_input_tokens_seen": 60516912,
+      "step": 1550
+    },
+    {
+      "epoch": 0.07558871983719352,
+      "grad_norm": 0.4141559898853302,
+      "learning_rate": 4.972845842534063e-05,
+      "loss": 1.2452,
+      "num_input_tokens_seen": 60899940,
+      "step": 1560
+    },
+    {
+      "epoch": 0.07607326291307297,
+      "grad_norm": 0.4386405646800995,
+      "learning_rate": 4.9722660976360534e-05,
+      "loss": 1.2262,
+      "num_input_tokens_seen": 61295608,
+      "step": 1570
+    },
+    {
+      "epoch": 0.07655780598895241,
+      "grad_norm": 0.40798524022102356,
+      "learning_rate": 4.9716802636325286e-05,
+      "loss": 1.2671,
+      "num_input_tokens_seen": 61693040,
+      "step": 1580
+    },
+    {
+      "epoch": 0.07704234906483186,
+      "grad_norm": 0.37625306844711304,
+      "learning_rate": 4.9710883419663774e-05,
+      "loss": 1.2962,
+      "num_input_tokens_seen": 62072768,
+      "step": 1590
+    },
+    {
+      "epoch": 0.0775268921407113,
+      "grad_norm": 0.4342903792858124,
+      "learning_rate": 4.970490334095482e-05,
+      "loss": 1.257,
+      "num_input_tokens_seen": 62446884,
+      "step": 1600
+    },
+    {
+      "epoch": 0.07801143521659075,
+      "grad_norm": 0.41230422258377075,
+      "learning_rate": 4.969886241492715e-05,
+      "loss": 1.27,
+      "num_input_tokens_seen": 62836772,
+      "step": 1610
+    },
+    {
+      "epoch": 0.0784959782924702,
+      "grad_norm": 0.4433565139770508,
+      "learning_rate": 4.969276065645936e-05,
+      "loss": 1.2592,
+      "num_input_tokens_seen": 63200816,
+      "step": 1620
+    },
+    {
+      "epoch": 0.07898052136834964,
+      "grad_norm": 0.47754809260368347,
+      "learning_rate": 4.968659808057986e-05,
+      "loss": 1.2752,
+      "num_input_tokens_seen": 63605328,
+      "step": 1630
+    },
+    {
+      "epoch": 0.07946506444422909,
+      "grad_norm": 0.3927769958972931,
+      "learning_rate": 4.968037470246687e-05,
+      "loss": 1.2468,
+      "num_input_tokens_seen": 63972724,
+      "step": 1640
+    },
+    {
+      "epoch": 0.07994960752010853,
+      "grad_norm": 0.39971116185188293,
+      "learning_rate": 4.9674090537448346e-05,
+      "loss": 1.1927,
+      "num_input_tokens_seen": 64354840,
+      "step": 1650
+    },
+    {
+      "epoch": 0.08043415059598798,
+      "grad_norm": 0.41240766644477844,
+      "learning_rate": 4.966774560100198e-05,
+      "loss": 1.2495,
+      "num_input_tokens_seen": 64781948,
+      "step": 1660
+    },
+    {
+      "epoch": 0.08091869367186742,
+      "grad_norm": 0.38843321800231934,
+      "learning_rate": 4.966133990875512e-05,
+      "loss": 1.2167,
+      "num_input_tokens_seen": 65145908,
+      "step": 1670
+    },
+    {
+      "epoch": 0.08140323674774687,
+      "grad_norm": 0.42320936918258667,
+      "learning_rate": 4.965487347648476e-05,
+      "loss": 1.2934,
+      "num_input_tokens_seen": 65551472,
+      "step": 1680
+    },
+    {
+      "epoch": 0.08188777982362631,
+      "grad_norm": 0.39081892371177673,
+      "learning_rate": 4.964834632011751e-05,
+      "loss": 1.2347,
+      "num_input_tokens_seen": 65941084,
+      "step": 1690
+    },
+    {
+      "epoch": 0.08237232289950576,
+      "grad_norm": 0.4150112271308899,
+      "learning_rate": 4.964175845572952e-05,
+      "loss": 1.2232,
+      "num_input_tokens_seen": 66336864,
+      "step": 1700
+    },
+    {
+      "epoch": 0.0828568659753852,
+      "grad_norm": 0.37995874881744385,
+      "learning_rate": 4.9635109899546476e-05,
+      "loss": 1.263,
+      "num_input_tokens_seen": 66718060,
+      "step": 1710
+    },
+    {
+      "epoch": 0.08334140905126466,
+      "grad_norm": 0.39722344279289246,
+      "learning_rate": 4.962840066794354e-05,
+      "loss": 1.276,
+      "num_input_tokens_seen": 67124980,
+      "step": 1720
+    },
+    {
+      "epoch": 0.08382595212714411,
+      "grad_norm": 0.4169159233570099,
+      "learning_rate": 4.9621630777445316e-05,
+      "loss": 1.2465,
+      "num_input_tokens_seen": 67528936,
+      "step": 1730
+    },
+    {
+      "epoch": 0.08431049520302356,
+      "grad_norm": 0.426110178232193,
+      "learning_rate": 4.96148002447258e-05,
+      "loss": 1.2534,
+      "num_input_tokens_seen": 67929316,
+      "step": 1740
+    },
+    {
+      "epoch": 0.084795038278903,
+      "grad_norm": 0.44997116923332214,
+      "learning_rate": 4.960790908660838e-05,
+      "loss": 1.2193,
+      "num_input_tokens_seen": 68318772,
+      "step": 1750
+    },
+    {
+      "epoch": 0.08527958135478245,
+      "grad_norm": 0.39495953917503357,
+      "learning_rate": 4.9600957320065715e-05,
+      "loss": 1.2464,
+      "num_input_tokens_seen": 68728948,
+      "step": 1760
+    },
+    {
+      "epoch": 0.08576412443066189,
+      "grad_norm": 0.39029765129089355,
+      "learning_rate": 4.959394496221977e-05,
+      "loss": 1.251,
+      "num_input_tokens_seen": 69116412,
+      "step": 1770
+    },
+    {
+      "epoch": 0.08624866750654134,
+      "grad_norm": 0.3958011865615845,
+      "learning_rate": 4.958687203034176e-05,
+      "loss": 1.2569,
+      "num_input_tokens_seen": 69515196,
+      "step": 1780
+    },
+    {
+      "epoch": 0.08673321058242078,
+      "grad_norm": 0.41105154156684875,
+      "learning_rate": 4.957973854185204e-05,
+      "loss": 1.2641,
+      "num_input_tokens_seen": 69919164,
+      "step": 1790
+    },
+    {
+      "epoch": 0.08721775365830023,
+      "grad_norm": 0.4101089835166931,
+      "learning_rate": 4.957254451432016e-05,
+      "loss": 1.2562,
+      "num_input_tokens_seen": 70312276,
+      "step": 1800
+    },
+    {
+      "epoch": 0.08770229673417967,
+      "grad_norm": 0.4429624080657959,
+      "learning_rate": 4.956528996546476e-05,
+      "loss": 1.2673,
+      "num_input_tokens_seen": 70700052,
+      "step": 1810
+    },
+    {
+      "epoch": 0.08818683981005912,
+      "grad_norm": 0.38688287138938904,
+      "learning_rate": 4.9557974913153536e-05,
+      "loss": 1.2294,
+      "num_input_tokens_seen": 71078868,
+      "step": 1820
+    },
+    {
+      "epoch": 0.08867138288593857,
+      "grad_norm": 0.4344911277294159,
+      "learning_rate": 4.955059937540322e-05,
+      "loss": 1.2436,
+      "num_input_tokens_seen": 71470396,
+      "step": 1830
+    },
+    {
+      "epoch": 0.08915592596181801,
+      "grad_norm": 0.4057205021381378,
+      "learning_rate": 4.9543163370379484e-05,
+      "loss": 1.2561,
+      "num_input_tokens_seen": 71856368,
+      "step": 1840
+    },
+    {
+      "epoch": 0.08964046903769746,
+      "grad_norm": 0.4123441278934479,
+      "learning_rate": 4.953566691639697e-05,
+      "loss": 1.2738,
+      "num_input_tokens_seen": 72253720,
+      "step": 1850
+    },
+    {
+      "epoch": 0.0901250121135769,
+      "grad_norm": 0.42719489336013794,
+      "learning_rate": 4.9528110031919175e-05,
+      "loss": 1.2306,
+      "num_input_tokens_seen": 72625716,
+      "step": 1860
+    },
+    {
+      "epoch": 0.09060955518945635,
+      "grad_norm": 0.41707441210746765,
+      "learning_rate": 4.9520492735558444e-05,
+      "loss": 1.2504,
+      "num_input_tokens_seen": 73013404,
+      "step": 1870
+    },
+    {
+      "epoch": 0.0910940982653358,
+      "grad_norm": 0.4267444312572479,
+      "learning_rate": 4.951281504607591e-05,
+      "loss": 1.2637,
+      "num_input_tokens_seen": 73406604,
+      "step": 1880
+    },
+    {
+      "epoch": 0.09157864134121524,
+      "grad_norm": 0.42108404636383057,
+      "learning_rate": 4.950507698238147e-05,
+      "loss": 1.2222,
+      "num_input_tokens_seen": 73783332,
+      "step": 1890
+    },
+    {
+      "epoch": 0.09206318441709468,
+      "grad_norm": 0.4063856303691864,
+      "learning_rate": 4.949727856353369e-05,
+      "loss": 1.2441,
+      "num_input_tokens_seen": 74175440,
+      "step": 1900
+    },
+    {
+      "epoch": 0.09254772749297413,
+      "grad_norm": 0.37594878673553467,
+      "learning_rate": 4.94894198087398e-05,
+      "loss": 1.2485,
+      "num_input_tokens_seen": 74581572,
+      "step": 1910
+    },
+    {
+      "epoch": 0.09303227056885358,
+      "grad_norm": 0.3911046087741852,
+      "learning_rate": 4.9481500737355656e-05,
+      "loss": 1.2702,
+      "num_input_tokens_seen": 74943092,
+      "step": 1920
+    },
+    {
+      "epoch": 0.09351681364473302,
+      "grad_norm": 0.4102419912815094,
+      "learning_rate": 4.947352136888566e-05,
+      "loss": 1.2468,
+      "num_input_tokens_seen": 75331720,
+      "step": 1930
+    },
+    {
+      "epoch": 0.09400135672061247,
+      "grad_norm": 0.39770299196243286,
+      "learning_rate": 4.94654817229827e-05,
+      "loss": 1.2201,
+      "num_input_tokens_seen": 75736304,
+      "step": 1940
+    },
+    {
+      "epoch": 0.09448589979649191,
+      "grad_norm": 0.43506377935409546,
+      "learning_rate": 4.945738181944815e-05,
+      "loss": 1.2235,
+      "num_input_tokens_seen": 76130884,
+      "step": 1950
+    },
+    {
+      "epoch": 0.09497044287237136,
+      "grad_norm": 0.3792125880718231,
+      "learning_rate": 4.9449221678231795e-05,
+      "loss": 1.2698,
+      "num_input_tokens_seen": 76512888,
+      "step": 1960
+    },
+    {
+      "epoch": 0.0954549859482508,
+      "grad_norm": 0.39077404141426086,
+      "learning_rate": 4.9441001319431784e-05,
+      "loss": 1.2458,
+      "num_input_tokens_seen": 76922052,
+      "step": 1970
+    },
+    {
+      "epoch": 0.09593952902413025,
+      "grad_norm": 0.38054585456848145,
+      "learning_rate": 4.943272076329457e-05,
+      "loss": 1.2281,
+      "num_input_tokens_seen": 77306336,
+      "step": 1980
+    },
+    {
+      "epoch": 0.0964240721000097,
+      "grad_norm": 0.3908326029777527,
+      "learning_rate": 4.942438003021487e-05,
+      "loss": 1.2028,
+      "num_input_tokens_seen": 77689480,
+      "step": 1990
+    },
+    {
+      "epoch": 0.09690861517588914,
+      "grad_norm": 0.43404319882392883,
+      "learning_rate": 4.941597914073563e-05,
+      "loss": 1.234,
+      "num_input_tokens_seen": 78067836,
+      "step": 2000
+    },
+    {
+      "epoch": 0.09690861517588914,
+      "eval_loss": 1.2438511848449707,
+      "eval_runtime": 5.3239,
+      "eval_samples_per_second": 28.175,
+      "eval_steps_per_second": 3.569,
+      "num_input_tokens_seen": 78067836,
+      "step": 2000
+    },
+    {
+      "epoch": 0.09739315825176859,
+      "grad_norm": 0.4713827967643738,
+      "learning_rate": 4.9407518115547945e-05,
+      "loss": 1.2405,
+      "num_input_tokens_seen": 78461236,
+      "step": 2010
+    },
+    {
+      "epoch": 0.09787770132764803,
+      "grad_norm": 0.4206187427043915,
+      "learning_rate": 4.939899697549102e-05,
+      "loss": 1.2046,
+      "num_input_tokens_seen": 78840632,
+      "step": 2020
+    },
+    {
+      "epoch": 0.09836224440352748,
+      "grad_norm": 0.3922071158885956,
+      "learning_rate": 4.939041574155213e-05,
+      "loss": 1.2593,
+      "num_input_tokens_seen": 79253124,
+      "step": 2030
+    },
+    {
+      "epoch": 0.09884678747940692,
+      "grad_norm": 0.4534291625022888,
+      "learning_rate": 4.938177443486657e-05,
+      "loss": 1.2696,
+      "num_input_tokens_seen": 79629736,
+      "step": 2040
+    },
+    {
+      "epoch": 0.09933133055528637,
+      "grad_norm": 0.424139142036438,
+      "learning_rate": 4.937307307671756e-05,
+      "loss": 1.2686,
+      "num_input_tokens_seen": 80023928,
+      "step": 2050
+    },
+    {
+      "epoch": 0.09981587363116581,
+      "grad_norm": 0.45865899324417114,
+      "learning_rate": 4.9364311688536245e-05,
+      "loss": 1.2834,
+      "num_input_tokens_seen": 80420160,
+      "step": 2060
+    },
+    {
+      "epoch": 0.10030041670704526,
+      "grad_norm": 0.42138755321502686,
+      "learning_rate": 4.935549029190163e-05,
+      "loss": 1.2079,
+      "num_input_tokens_seen": 80810044,
+      "step": 2070
+    },
+    {
+      "epoch": 0.1007849597829247,
+      "grad_norm": 0.3841094672679901,
+      "learning_rate": 4.934660890854049e-05,
+      "loss": 1.2496,
+      "num_input_tokens_seen": 81210716,
+      "step": 2080
+    },
+    {
+      "epoch": 0.10126950285880415,
+      "grad_norm": 0.38037049770355225,
+      "learning_rate": 4.9337667560327396e-05,
+      "loss": 1.2724,
+      "num_input_tokens_seen": 81585980,
+      "step": 2090
+    },
+    {
+      "epoch": 0.1017540459346836,
+      "grad_norm": 0.4014841318130493,
+      "learning_rate": 4.932866626928454e-05,
+      "loss": 1.227,
+      "num_input_tokens_seen": 81980652,
+      "step": 2100
+    },
+    {
+      "epoch": 0.10223858901056304,
+      "grad_norm": 0.41437438130378723,
+      "learning_rate": 4.9319605057581816e-05,
+      "loss": 1.2472,
+      "num_input_tokens_seen": 82387492,
+      "step": 2110
+    },
+    {
+      "epoch": 0.10272313208644249,
+      "grad_norm": 0.4091525971889496,
+      "learning_rate": 4.931048394753666e-05,
+      "loss": 1.188,
+      "num_input_tokens_seen": 82759272,
+      "step": 2120
+    },
+    {
+      "epoch": 0.10320767516232193,
+      "grad_norm": 0.43405330181121826,
+      "learning_rate": 4.930130296161406e-05,
+      "loss": 1.2372,
+      "num_input_tokens_seen": 83158564,
+      "step": 2130
+    },
+    {
+      "epoch": 0.10369221823820138,
+      "grad_norm": 0.43749940395355225,
+      "learning_rate": 4.929206212242646e-05,
+      "loss": 1.2085,
+      "num_input_tokens_seen": 83536672,
+      "step": 2140
+    },
+    {
+      "epoch": 0.10417676131408082,
+      "grad_norm": 0.43483173847198486,
+      "learning_rate": 4.928276145273372e-05,
+      "loss": 1.2199,
+      "num_input_tokens_seen": 83943180,
+      "step": 2150
+    },
+    {
+      "epoch": 0.10466130438996027,
+      "grad_norm": 0.42413613200187683,
+      "learning_rate": 4.9273400975443076e-05,
+      "loss": 1.2012,
+      "num_input_tokens_seen": 84356164,
+      "step": 2160
+    },
+    {
+      "epoch": 0.10514584746583971,
+      "grad_norm": 0.445049911737442,
+      "learning_rate": 4.926398071360905e-05,
+      "loss": 1.2046,
+      "num_input_tokens_seen": 84761956,
+      "step": 2170
+    },
+    {
+      "epoch": 0.10563039054171916,
+      "grad_norm": 0.4001222848892212,
+      "learning_rate": 4.925450069043342e-05,
+      "loss": 1.246,
+      "num_input_tokens_seen": 85138340,
+      "step": 2180
+    },
+    {
+      "epoch": 0.1061149336175986,
+      "grad_norm": 0.401049941778183,
+      "learning_rate": 4.924496092926517e-05,
+      "loss": 1.2164,
+      "num_input_tokens_seen": 85513616,
+      "step": 2190
+    },
+    {
+      "epoch": 0.10659947669347805,
+      "grad_norm": 0.4612962603569031,
+      "learning_rate": 4.923536145360038e-05,
+      "loss": 1.2358,
+      "num_input_tokens_seen": 85931448,
+      "step": 2200
+    },
+    {
+      "epoch": 0.1070840197693575,
+      "grad_norm": 0.41920316219329834,
+      "learning_rate": 4.922570228708223e-05,
+      "loss": 1.3026,
+      "num_input_tokens_seen": 86315108,
+      "step": 2210
+    },
+    {
+      "epoch": 0.10756856284523694,
+      "grad_norm": 0.39290520548820496,
+      "learning_rate": 4.921598345350092e-05,
+      "loss": 1.2535,
+      "num_input_tokens_seen": 86709944,
+      "step": 2220
+    },
+    {
+      "epoch": 0.10805310592111639,
+      "grad_norm": 0.4282558262348175,
+      "learning_rate": 4.92062049767936e-05,
+      "loss": 1.2274,
+      "num_input_tokens_seen": 87113940,
+      "step": 2230
+    },
+    {
+      "epoch": 0.10853764899699583,
+      "grad_norm": 0.4753318428993225,
+      "learning_rate": 4.91963668810443e-05,
+      "loss": 1.2483,
+      "num_input_tokens_seen": 87501616,
+      "step": 2240
+    },
+    {
+      "epoch": 0.10902219207287528,
+      "grad_norm": 0.423225075006485,
+      "learning_rate": 4.918646919048393e-05,
+      "loss": 1.2327,
+      "num_input_tokens_seen": 87894360,
+      "step": 2250
+    },
+    {
+      "epoch": 0.10950673514875472,
+      "grad_norm": 0.39466777443885803,
+      "learning_rate": 4.9176511929490144e-05,
+      "loss": 1.2604,
+      "num_input_tokens_seen": 88317576,
+      "step": 2260
+    },
+    {
+      "epoch": 0.10999127822463417,
+      "grad_norm": 0.41255709528923035,
+      "learning_rate": 4.916649512258733e-05,
+      "loss": 1.2341,
+      "num_input_tokens_seen": 88690104,
+      "step": 2270
+    },
+    {
+      "epoch": 0.11047582130051362,
+      "grad_norm": 0.42974668741226196,
+      "learning_rate": 4.9156418794446545e-05,
+      "loss": 1.2159,
+      "num_input_tokens_seen": 89090732,
+      "step": 2280
+    },
+    {
+      "epoch": 0.11096036437639306,
+      "grad_norm": 0.44297677278518677,
+      "learning_rate": 4.914628296988543e-05,
+      "loss": 1.2344,
+      "num_input_tokens_seen": 89486096,
+      "step": 2290
+    },
+    {
+      "epoch": 0.11144490745227251,
+      "grad_norm": 0.4597613513469696,
+      "learning_rate": 4.913608767386817e-05,
+      "loss": 1.2482,
+      "num_input_tokens_seen": 89880732,
+      "step": 2300
+    },
+    {
+      "epoch": 0.11192945052815195,
+      "grad_norm": 0.4195135235786438,
+      "learning_rate": 4.912583293150542e-05,
+      "loss": 1.2158,
+      "num_input_tokens_seen": 90328148,
+      "step": 2310
+    },
+    {
+      "epoch": 0.1124139936040314,
+      "grad_norm": 0.44454923272132874,
+      "learning_rate": 4.9115518768054264e-05,
+      "loss": 1.2103,
+      "num_input_tokens_seen": 90703168,
+      "step": 2320
+    },
+    {
+      "epoch": 0.11289853667991084,
+      "grad_norm": 0.3934001624584198,
+      "learning_rate": 4.910514520891812e-05,
+      "loss": 1.2893,
+      "num_input_tokens_seen": 91052408,
+      "step": 2330
+    },
+    {
+      "epoch": 0.11338307975579029,
+      "grad_norm": 0.38367339968681335,
+      "learning_rate": 4.90947122796467e-05,
+      "loss": 1.2125,
+      "num_input_tokens_seen": 91430836,
+      "step": 2340
+    },
+    {
+      "epoch": 0.11386762283166973,
+      "grad_norm": 0.40738987922668457,
+      "learning_rate": 4.908422000593596e-05,
+      "loss": 1.2289,
+      "num_input_tokens_seen": 91848392,
+      "step": 2350
+    },
+    {
+      "epoch": 0.11435216590754918,
+      "grad_norm": 0.4179443418979645,
+      "learning_rate": 4.907366841362799e-05,
+      "loss": 1.2494,
+      "num_input_tokens_seen": 92240852,
+      "step": 2360
+    },
+    {
+      "epoch": 0.11483670898342863,
+      "grad_norm": 0.38464492559432983,
+      "learning_rate": 4.906305752871102e-05,
+      "loss": 1.1879,
+      "num_input_tokens_seen": 92619204,
+      "step": 2370
+    },
+    {
+      "epoch": 0.11532125205930807,
+      "grad_norm": 0.42252519726753235,
+      "learning_rate": 4.905238737731926e-05,
+      "loss": 1.2608,
+      "num_input_tokens_seen": 93002632,
+      "step": 2380
+    },
+    {
+      "epoch": 0.11580579513518752,
+      "grad_norm": 0.4146505296230316,
+      "learning_rate": 4.9041657985732936e-05,
+      "loss": 1.2486,
+      "num_input_tokens_seen": 93399964,
+      "step": 2390
+    },
+    {
+      "epoch": 0.11629033821106696,
+      "grad_norm": 0.438689261674881,
+      "learning_rate": 4.903086938037818e-05,
+      "loss": 1.27,
+      "num_input_tokens_seen": 93791652,
+      "step": 2400
+    },
+    {
+      "epoch": 0.11677488128694641,
+      "grad_norm": 0.42062869668006897,
+      "learning_rate": 4.9020021587826926e-05,
+      "loss": 1.2591,
+      "num_input_tokens_seen": 94191368,
+      "step": 2410
+    },
+    {
+      "epoch": 0.11725942436282585,
+      "grad_norm": 0.4142090976238251,
+      "learning_rate": 4.9009114634796926e-05,
+      "loss": 1.2627,
+      "num_input_tokens_seen": 94571872,
+      "step": 2420
+    },
+    {
+      "epoch": 0.1177439674387053,
+      "grad_norm": 0.4038289487361908,
+      "learning_rate": 4.8998148548151625e-05,
+      "loss": 1.2551,
+      "num_input_tokens_seen": 94979976,
+      "step": 2430
+    },
+    {
+      "epoch": 0.11822851051458474,
+      "grad_norm": 0.38102611899375916,
+      "learning_rate": 4.898712335490012e-05,
+      "loss": 1.2559,
+      "num_input_tokens_seen": 95358664,
+      "step": 2440
+    },
+    {
+      "epoch": 0.11871305359046419,
+      "grad_norm": 0.3852735757827759,
+      "learning_rate": 4.897603908219706e-05,
+      "loss": 1.2448,
+      "num_input_tokens_seen": 95763084,
+      "step": 2450
+    },
+    {
+      "epoch": 0.11919759666634364,
+      "grad_norm": 0.3712899684906006,
+      "learning_rate": 4.8964895757342643e-05,
+      "loss": 1.2488,
+      "num_input_tokens_seen": 96188292,
+      "step": 2460
+    },
+    {
+      "epoch": 0.11968213974222308,
+      "grad_norm": 0.3657510578632355,
+      "learning_rate": 4.8953693407782484e-05,
+      "loss": 1.253,
+      "num_input_tokens_seen": 96601596,
+      "step": 2470
+    },
+    {
+      "epoch": 0.12016668281810253,
+      "grad_norm": 0.4605284631252289,
+      "learning_rate": 4.894243206110758e-05,
+      "loss": 1.236,
+      "num_input_tokens_seen": 96973176,
+      "step": 2480
+    },
+    {
+      "epoch": 0.12065122589398197,
+      "grad_norm": 0.41020140051841736,
+      "learning_rate": 4.8931111745054226e-05,
+      "loss": 1.2364,
+      "num_input_tokens_seen": 97378752,
+      "step": 2490
+    },
+    {
+      "epoch": 0.12113576896986142,
+      "grad_norm": 0.36746346950531006,
+      "learning_rate": 4.891973248750399e-05,
+      "loss": 1.2666,
+      "num_input_tokens_seen": 97771668,
+      "step": 2500
+    },
+    {
+      "epoch": 0.12162031204574086,
+      "grad_norm": 0.4322774112224579,
+      "learning_rate": 4.890829431648357e-05,
+      "loss": 1.1992,
+      "num_input_tokens_seen": 98151028,
+      "step": 2510
+    },
+    {
+      "epoch": 0.12210485512162031,
+      "grad_norm": 0.42371341586112976,
+      "learning_rate": 4.8896797260164795e-05,
+      "loss": 1.3052,
+      "num_input_tokens_seen": 98531364,
+      "step": 2520
+    },
+    {
+      "epoch": 0.12258939819749975,
+      "grad_norm": 0.43509674072265625,
+      "learning_rate": 4.888524134686451e-05,
+      "loss": 1.2146,
+      "num_input_tokens_seen": 98929648,
+      "step": 2530
+    },
+    {
+      "epoch": 0.1230739412733792,
+      "grad_norm": 0.38487133383750916,
+      "learning_rate": 4.887362660504452e-05,
+      "loss": 1.2251,
+      "num_input_tokens_seen": 99326780,
+      "step": 2540
+    },
+    {
+      "epoch": 0.12355848434925865,
+      "grad_norm": 0.47879770398139954,
+      "learning_rate": 4.8861953063311544e-05,
+      "loss": 1.2862,
+      "num_input_tokens_seen": 99746768,
+      "step": 2550
+    },
+    {
+      "epoch": 0.12404302742513809,
+      "grad_norm": 0.3904902935028076,
+      "learning_rate": 4.8850220750417107e-05,
+      "loss": 1.2565,
+      "num_input_tokens_seen": 100114696,
+      "step": 2560
+    },
+    {
+      "epoch": 0.12452757050101754,
+      "grad_norm": 0.3858630359172821,
+      "learning_rate": 4.883842969525748e-05,
+      "loss": 1.2145,
+      "num_input_tokens_seen": 100542848,
+      "step": 2570
+    },
+    {
+      "epoch": 0.12501211357689698,
+      "grad_norm": 0.42013123631477356,
+      "learning_rate": 4.882657992687363e-05,
+      "loss": 1.2271,
+      "num_input_tokens_seen": 100976524,
+      "step": 2580
+    },
+    {
+      "epoch": 0.12549665665277643,
+      "grad_norm": 0.3980698585510254,
+      "learning_rate": 4.8814671474451126e-05,
+      "loss": 1.2801,
+      "num_input_tokens_seen": 101345460,
+      "step": 2590
+    },
+    {
+      "epoch": 0.12598119972865587,
+      "grad_norm": 0.3923351764678955,
+      "learning_rate": 4.880270436732006e-05,
+      "loss": 1.2064,
+      "num_input_tokens_seen": 101759284,
+      "step": 2600
+    },
+    {
+      "epoch": 0.12646574280453532,
+      "grad_norm": 0.40927964448928833,
+      "learning_rate": 4.879067863495502e-05,
+      "loss": 1.2739,
+      "num_input_tokens_seen": 102167588,
+      "step": 2610
+    },
+    {
+      "epoch": 0.12695028588041476,
+      "grad_norm": 0.44859156012535095,
+      "learning_rate": 4.8778594306974956e-05,
+      "loss": 1.2309,
+      "num_input_tokens_seen": 102539236,
+      "step": 2620
+    },
+    {
+      "epoch": 0.1274348289562942,
+      "grad_norm": 0.39487892389297485,
+      "learning_rate": 4.876645141314315e-05,
+      "loss": 1.2866,
+      "num_input_tokens_seen": 102925936,
+      "step": 2630
+    },
+    {
+      "epoch": 0.12791937203217366,
+      "grad_norm": 0.3923860490322113,
+      "learning_rate": 4.875424998336713e-05,
+      "loss": 1.2757,
+      "num_input_tokens_seen": 103325248,
+      "step": 2640
+    },
+    {
+      "epoch": 0.1284039151080531,
+      "grad_norm": 0.4167238175868988,
+      "learning_rate": 4.874199004769858e-05,
+      "loss": 1.2338,
+      "num_input_tokens_seen": 103712300,
+      "step": 2650
+    },
+    {
+      "epoch": 0.12888845818393255,
+      "grad_norm": 0.38341742753982544,
+      "learning_rate": 4.872967163633332e-05,
+      "loss": 1.2383,
+      "num_input_tokens_seen": 104124208,
+      "step": 2660
+    },
+    {
+      "epoch": 0.129373001259812,
+      "grad_norm": 0.3938426375389099,
+      "learning_rate": 4.871729477961115e-05,
+      "loss": 1.2011,
+      "num_input_tokens_seen": 104502028,
+      "step": 2670
+    },
+    {
+      "epoch": 0.12985754433569144,
+      "grad_norm": 0.3831293284893036,
+      "learning_rate": 4.8704859508015847e-05,
+      "loss": 1.2634,
+      "num_input_tokens_seen": 104876704,
+      "step": 2680
+    },
+    {
+      "epoch": 0.13034208741157088,
+      "grad_norm": 0.4196309745311737,
+      "learning_rate": 4.869236585217504e-05,
+      "loss": 1.2585,
+      "num_input_tokens_seen": 105302956,
+      "step": 2690
+    },
+    {
+      "epoch": 0.13082663048745033,
+      "grad_norm": 0.4367752969264984,
+      "learning_rate": 4.867981384286019e-05,
+      "loss": 1.1848,
+      "num_input_tokens_seen": 105695812,
+      "step": 2700
+    },
+    {
+      "epoch": 0.13131117356332977,
+      "grad_norm": 0.39066222310066223,
+      "learning_rate": 4.8667203510986444e-05,
+      "loss": 1.2545,
+      "num_input_tokens_seen": 106084708,
+      "step": 2710
+    },
+    {
+      "epoch": 0.13179571663920922,
+      "grad_norm": 0.4448121190071106,
+      "learning_rate": 4.865453488761262e-05,
+      "loss": 1.275,
+      "num_input_tokens_seen": 106499276,
+      "step": 2720
+    },
+    {
+      "epoch": 0.13228025971508867,
+      "grad_norm": 0.4147442877292633,
+      "learning_rate": 4.864180800394109e-05,
+      "loss": 1.277,
+      "num_input_tokens_seen": 106866988,
+      "step": 2730
+    },
+    {
+      "epoch": 0.1327648027909681,
+      "grad_norm": 0.40027832984924316,
+      "learning_rate": 4.862902289131773e-05,
+      "loss": 1.2567,
+      "num_input_tokens_seen": 107259856,
+      "step": 2740
+    },
+    {
+      "epoch": 0.13324934586684756,
+      "grad_norm": 0.36126065254211426,
+      "learning_rate": 4.861617958123183e-05,
+      "loss": 1.2553,
+      "num_input_tokens_seen": 107646308,
+      "step": 2750
+    },
+    {
+      "epoch": 0.133733888942727,
+      "grad_norm": 0.41318050026893616,
+      "learning_rate": 4.860327810531602e-05,
+      "loss": 1.244,
+      "num_input_tokens_seen": 108059036,
+      "step": 2760
+    },
+    {
+      "epoch": 0.13421843201860645,
+      "grad_norm": 0.3924005925655365,
+      "learning_rate": 4.859031849534618e-05,
+      "loss": 1.223,
+      "num_input_tokens_seen": 108463880,
+      "step": 2770
+    },
+    {
+      "epoch": 0.1347029750944859,
+      "grad_norm": 0.41446420550346375,
+      "learning_rate": 4.857730078324138e-05,
+      "loss": 1.2424,
+      "num_input_tokens_seen": 108859416,
+      "step": 2780
+    },
+    {
+      "epoch": 0.13518751817036534,
+      "grad_norm": 0.4131425619125366,
+      "learning_rate": 4.856422500106379e-05,
+      "loss": 1.2368,
+      "num_input_tokens_seen": 109284028,
+      "step": 2790
+    },
+    {
+      "epoch": 0.13567206124624479,
+      "grad_norm": 0.40505287051200867,
+      "learning_rate": 4.855109118101862e-05,
+      "loss": 1.2731,
+      "num_input_tokens_seen": 109666092,
+      "step": 2800
+    },
+    {
+      "epoch": 0.13615660432212423,
+      "grad_norm": 0.4067140817642212,
+      "learning_rate": 4.8537899355454e-05,
+      "loss": 1.2945,
+      "num_input_tokens_seen": 110023200,
+      "step": 2810
+    },
+    {
+      "epoch": 0.13664114739800368,
+      "grad_norm": 0.42721500992774963,
+      "learning_rate": 4.8524649556860934e-05,
+      "loss": 1.2448,
+      "num_input_tokens_seen": 110400364,
+      "step": 2820
+    },
+    {
+      "epoch": 0.13712569047388312,
+      "grad_norm": 0.40935906767845154,
+      "learning_rate": 4.851134181787323e-05,
+      "loss": 1.2359,
+      "num_input_tokens_seen": 110793504,
+      "step": 2830
+    },
+    {
+      "epoch": 0.13761023354976257,
+      "grad_norm": 0.4633960425853729,
+      "learning_rate": 4.849797617126738e-05,
+      "loss": 1.2571,
+      "num_input_tokens_seen": 111170796,
+      "step": 2840
+    },
+    {
+      "epoch": 0.138094776625642,
+      "grad_norm": 0.43465399742126465,
+      "learning_rate": 4.848455264996251e-05,
+      "loss": 1.2171,
+      "num_input_tokens_seen": 111576676,
+      "step": 2850
+    },
+    {
+      "epoch": 0.13857931970152146,
+      "grad_norm": 0.3760763704776764,
+      "learning_rate": 4.847107128702028e-05,
+      "loss": 1.2611,
+      "num_input_tokens_seen": 112001188,
+      "step": 2860
+    },
+    {
+      "epoch": 0.1390638627774009,
+      "grad_norm": 0.4067819118499756,
+      "learning_rate": 4.845753211564482e-05,
+      "loss": 1.1932,
+      "num_input_tokens_seen": 112357276,
+      "step": 2870
+    },
+    {
+      "epoch": 0.13954840585328035,
+      "grad_norm": 0.44168785214424133,
+      "learning_rate": 4.844393516918265e-05,
+      "loss": 1.2458,
+      "num_input_tokens_seen": 112745724,
+      "step": 2880
+    },
+    {
+      "epoch": 0.1400329489291598,
+      "grad_norm": 0.42745277285575867,
+      "learning_rate": 4.8430280481122556e-05,
+      "loss": 1.1844,
+      "num_input_tokens_seen": 113158204,
+      "step": 2890
+    },
+    {
+      "epoch": 0.14051749200503924,
+      "grad_norm": 0.4196942150592804,
+      "learning_rate": 4.8416568085095585e-05,
+      "loss": 1.2464,
+      "num_input_tokens_seen": 113534300,
+      "step": 2900
+    },
+    {
+      "epoch": 0.1410020350809187,
+      "grad_norm": 0.3699077069759369,
+      "learning_rate": 4.840279801487488e-05,
+      "loss": 1.2435,
+      "num_input_tokens_seen": 113941828,
+      "step": 2910
+    },
+    {
+      "epoch": 0.14148657815679813,
+      "grad_norm": 0.4300510287284851,
+      "learning_rate": 4.8388970304375636e-05,
+      "loss": 1.2438,
+      "num_input_tokens_seen": 114360276,
+      "step": 2920
+    },
+    {
+      "epoch": 0.14197112123267758,
+      "grad_norm": 0.39781638979911804,
+      "learning_rate": 4.837508498765504e-05,
+      "loss": 1.2151,
+      "num_input_tokens_seen": 114724572,
+      "step": 2930
+    },
+    {
+      "epoch": 0.14245566430855702,
+      "grad_norm": 0.3915679156780243,
+      "learning_rate": 4.836114209891214e-05,
+      "loss": 1.2486,
+      "num_input_tokens_seen": 115115652,
+      "step": 2940
+    },
+    {
+      "epoch": 0.14294020738443647,
+      "grad_norm": 0.3921675682067871,
+      "learning_rate": 4.834714167248778e-05,
+      "loss": 1.2237,
+      "num_input_tokens_seen": 115528916,
+      "step": 2950
+    },
+    {
+      "epoch": 0.14342475046031591,
+      "grad_norm": 0.4208355247974396,
+      "learning_rate": 4.8333083742864524e-05,
+      "loss": 1.3102,
+      "num_input_tokens_seen": 115883916,
+      "step": 2960
+    },
+    {
+      "epoch": 0.14390929353619536,
+      "grad_norm": 0.4001530408859253,
+      "learning_rate": 4.831896834466658e-05,
+      "loss": 1.2382,
+      "num_input_tokens_seen": 116270492,
+      "step": 2970
+    },
+    {
+      "epoch": 0.1443938366120748,
+      "grad_norm": 0.4217351973056793,
+      "learning_rate": 4.830479551265966e-05,
+      "loss": 1.1906,
+      "num_input_tokens_seen": 116680092,
+      "step": 2980
+    },
+    {
+      "epoch": 0.14487837968795425,
+      "grad_norm": 0.37808090448379517,
+      "learning_rate": 4.8290565281750974e-05,
+      "loss": 1.2392,
+      "num_input_tokens_seen": 117085684,
+      "step": 2990
+    },
+    {
+      "epoch": 0.1453629227638337,
+      "grad_norm": 0.3951645493507385,
+      "learning_rate": 4.827627768698909e-05,
+      "loss": 1.2293,
+      "num_input_tokens_seen": 117454120,
+      "step": 3000
+    },
+    {
+      "epoch": 0.14584746583971314,
+      "grad_norm": 0.35314205288887024,
+      "learning_rate": 4.8261932763563834e-05,
+      "loss": 1.1356,
+      "num_input_tokens_seen": 117866504,
+      "step": 3010
+    },
+    {
+      "epoch": 0.1463320089155926,
+      "grad_norm": 0.4276081323623657,
+      "learning_rate": 4.824753054680628e-05,
+      "loss": 1.2778,
+      "num_input_tokens_seen": 118250676,
+      "step": 3020
+    },
+    {
+      "epoch": 0.14681655199147203,
+      "grad_norm": 0.40378737449645996,
+      "learning_rate": 4.823307107218857e-05,
+      "loss": 1.2868,
+      "num_input_tokens_seen": 118607944,
+      "step": 3030
+    },
+    {
+      "epoch": 0.14730109506735148,
+      "grad_norm": 0.41640186309814453,
+      "learning_rate": 4.82185543753239e-05,
+      "loss": 1.2456,
+      "num_input_tokens_seen": 119009904,
+      "step": 3040
+    },
+    {
+      "epoch": 0.14778563814323092,
+      "grad_norm": 0.4186214804649353,
+      "learning_rate": 4.820398049196638e-05,
+      "loss": 1.2287,
+      "num_input_tokens_seen": 119411820,
+      "step": 3050
+    },
+    {
+      "epoch": 0.14827018121911037,
+      "grad_norm": 0.42039960622787476,
+      "learning_rate": 4.818934945801098e-05,
+      "loss": 1.2247,
+      "num_input_tokens_seen": 119798480,
+      "step": 3060
+    },
+    {
+      "epoch": 0.14875472429498982,
+      "grad_norm": 0.42505961656570435,
+      "learning_rate": 4.8174661309493436e-05,
+      "loss": 1.2242,
+      "num_input_tokens_seen": 120184372,
+      "step": 3070
+    },
+    {
+      "epoch": 0.14923926737086926,
+      "grad_norm": 0.4036807417869568,
+      "learning_rate": 4.815991608259014e-05,
+      "loss": 1.2154,
+      "num_input_tokens_seen": 120579184,
+      "step": 3080
+    },
+    {
+      "epoch": 0.1497238104467487,
+      "grad_norm": 0.40509092807769775,
+      "learning_rate": 4.8145113813618083e-05,
+      "loss": 1.2086,
+      "num_input_tokens_seen": 120941008,
+      "step": 3090
+    },
+    {
+      "epoch": 0.15020835352262815,
+      "grad_norm": 0.4269082546234131,
+      "learning_rate": 4.813025453903472e-05,
+      "loss": 1.2496,
+      "num_input_tokens_seen": 121322824,
+      "step": 3100
+    },
+    {
+      "epoch": 0.1506928965985076,
+      "grad_norm": 0.4054916203022003,
+      "learning_rate": 4.811533829543795e-05,
+      "loss": 1.2079,
+      "num_input_tokens_seen": 121738192,
+      "step": 3110
+    },
+    {
+      "epoch": 0.15117743967438704,
+      "grad_norm": 0.413043349981308,
+      "learning_rate": 4.8100365119565946e-05,
+      "loss": 1.2609,
+      "num_input_tokens_seen": 122108864,
+      "step": 3120
+    },
+    {
+      "epoch": 0.1516619827502665,
+      "grad_norm": 0.41178593039512634,
+      "learning_rate": 4.8085335048297135e-05,
+      "loss": 1.2548,
+      "num_input_tokens_seen": 122510356,
+      "step": 3130
+    },
+    {
+      "epoch": 0.15214652582614593,
+      "grad_norm": 0.39382249116897583,
+      "learning_rate": 4.8070248118650055e-05,
+      "loss": 1.1809,
+      "num_input_tokens_seen": 122865892,
+      "step": 3140
+    },
+    {
+      "epoch": 0.15263106890202538,
+      "grad_norm": 0.46355170011520386,
+      "learning_rate": 4.8055104367783275e-05,
+      "loss": 1.2281,
+      "num_input_tokens_seen": 123237536,
+      "step": 3150
+    },
+    {
+      "epoch": 0.15311561197790483,
+      "grad_norm": 0.40591225028038025,
+      "learning_rate": 4.803990383299535e-05,
+      "loss": 1.2243,
+      "num_input_tokens_seen": 123611572,
+      "step": 3160
+    },
+    {
+      "epoch": 0.15360015505378427,
+      "grad_norm": 0.41489487886428833,
+      "learning_rate": 4.802464655172466e-05,
+      "loss": 1.239,
+      "num_input_tokens_seen": 124010748,
+      "step": 3170
+    },
+    {
+      "epoch": 0.15408469812966372,
+      "grad_norm": 0.3760206997394562,
+      "learning_rate": 4.800933256154935e-05,
+      "loss": 1.2013,
+      "num_input_tokens_seen": 124392148,
+      "step": 3180
+    },
+    {
+      "epoch": 0.15456924120554316,
+      "grad_norm": 0.43629947304725647,
+      "learning_rate": 4.7993961900187255e-05,
+      "loss": 1.2253,
+      "num_input_tokens_seen": 124776000,
+      "step": 3190
+    },
+    {
+      "epoch": 0.1550537842814226,
+      "grad_norm": 0.4242326021194458,
+      "learning_rate": 4.7978534605495784e-05,
+      "loss": 1.2461,
+      "num_input_tokens_seen": 125158264,
+      "step": 3200
+    },
+    {
+      "epoch": 0.15553832735730205,
+      "grad_norm": 0.40662136673927307,
+      "learning_rate": 4.796305071547182e-05,
+      "loss": 1.1913,
+      "num_input_tokens_seen": 125544848,
+      "step": 3210
+    },
+    {
+      "epoch": 0.1560228704331815,
+      "grad_norm": 0.4275949001312256,
+      "learning_rate": 4.794751026825165e-05,
+      "loss": 1.2397,
+      "num_input_tokens_seen": 125940532,
+      "step": 3220
+    },
+    {
+      "epoch": 0.15650741350906094,
+      "grad_norm": 0.41079017519950867,
+      "learning_rate": 4.793191330211085e-05,
+      "loss": 1.2478,
+      "num_input_tokens_seen": 126357472,
+      "step": 3230
+    },
+    {
+      "epoch": 0.1569919565849404,
+      "grad_norm": 0.4043791890144348,
+      "learning_rate": 4.791625985546422e-05,
+      "loss": 1.2263,
+      "num_input_tokens_seen": 126745452,
+      "step": 3240
+    },
+    {
+      "epoch": 0.15747649966081984,
+      "grad_norm": 0.4353187680244446,
+      "learning_rate": 4.790054996686564e-05,
+      "loss": 1.1988,
+      "num_input_tokens_seen": 127138656,
+      "step": 3250
+    },
+    {
+      "epoch": 0.15796104273669928,
+      "grad_norm": 0.4300646483898163,
+      "learning_rate": 4.7884783675008016e-05,
+      "loss": 1.183,
+      "num_input_tokens_seen": 127522676,
+      "step": 3260
+    },
+    {
+      "epoch": 0.15844558581257873,
+      "grad_norm": 0.3753831684589386,
+      "learning_rate": 4.7868961018723194e-05,
+      "loss": 1.2412,
+      "num_input_tokens_seen": 127917336,
+      "step": 3270
+    },
+    {
+      "epoch": 0.15893012888845817,
+      "grad_norm": 0.36102235317230225,
+      "learning_rate": 4.785308203698182e-05,
+      "loss": 1.2764,
+      "num_input_tokens_seen": 128336744,
+      "step": 3280
+    },
+    {
+      "epoch": 0.15941467196433762,
+      "grad_norm": 0.40448567271232605,
+      "learning_rate": 4.783714676889327e-05,
+      "loss": 1.2122,
+      "num_input_tokens_seen": 128733104,
+      "step": 3290
+    },
+    {
+      "epoch": 0.15989921504021706,
+      "grad_norm": 0.3721427023410797,
+      "learning_rate": 4.7821155253705564e-05,
+      "loss": 1.2684,
+      "num_input_tokens_seen": 129154876,
+      "step": 3300
+    },
+    {
+      "epoch": 0.1603837581160965,
+      "grad_norm": 0.4208812415599823,
+      "learning_rate": 4.7805107530805244e-05,
+      "loss": 1.2343,
+      "num_input_tokens_seen": 129547976,
+      "step": 3310
+    },
+    {
+      "epoch": 0.16086830119197595,
+      "grad_norm": 0.39913055300712585,
+      "learning_rate": 4.778900363971729e-05,
+      "loss": 1.2078,
+      "num_input_tokens_seen": 129961780,
+      "step": 3320
+    },
+    {
+      "epoch": 0.1613528442678554,
+      "grad_norm": 0.4164119362831116,
+      "learning_rate": 4.777284362010504e-05,
+      "loss": 1.2501,
+      "num_input_tokens_seen": 130358684,
+      "step": 3330
+    },
+    {
+      "epoch": 0.16183738734373485,
+      "grad_norm": 0.3951362073421478,
+      "learning_rate": 4.775662751177003e-05,
+      "loss": 1.206,
+      "num_input_tokens_seen": 130762100,
+      "step": 3340
+    },
+    {
+      "epoch": 0.1623219304196143,
+      "grad_norm": 0.4006355404853821,
+      "learning_rate": 4.774035535465201e-05,
+      "loss": 1.2516,
+      "num_input_tokens_seen": 131146316,
+      "step": 3350
+    },
+    {
+      "epoch": 0.16280647349549374,
+      "grad_norm": 0.41116824746131897,
+      "learning_rate": 4.7724027188828716e-05,
+      "loss": 1.1868,
+      "num_input_tokens_seen": 131538384,
+      "step": 3360
+    },
+    {
+      "epoch": 0.16329101657137318,
+      "grad_norm": 0.4331505596637726,
+      "learning_rate": 4.7707643054515855e-05,
+      "loss": 1.2212,
+      "num_input_tokens_seen": 131926864,
+      "step": 3370
+    },
+    {
+      "epoch": 0.16377555964725263,
+      "grad_norm": 0.3957429826259613,
+      "learning_rate": 4.7691202992066984e-05,
+      "loss": 1.1938,
+      "num_input_tokens_seen": 132309648,
+      "step": 3380
+    },
+    {
+      "epoch": 0.16426010272313207,
+      "grad_norm": 0.3914317786693573,
+      "learning_rate": 4.767470704197341e-05,
+      "loss": 1.2082,
+      "num_input_tokens_seen": 132725476,
+      "step": 3390
+    },
+    {
+      "epoch": 0.16474464579901152,
+      "grad_norm": 0.4648467004299164,
+      "learning_rate": 4.765815524486407e-05,
+      "loss": 1.2276,
+      "num_input_tokens_seen": 133107108,
+      "step": 3400
+    },
+    {
+      "epoch": 0.16522918887489096,
+      "grad_norm": 0.3745146691799164,
+      "learning_rate": 4.764154764150548e-05,
+      "loss": 1.223,
+      "num_input_tokens_seen": 133483520,
+      "step": 3410
+    },
+    {
+      "epoch": 0.1657137319507704,
+      "grad_norm": 0.38040289282798767,
+      "learning_rate": 4.762488427280159e-05,
+      "loss": 1.1943,
+      "num_input_tokens_seen": 133871252,
+      "step": 3420
+    },
+    {
+      "epoch": 0.16619827502664986,
+      "grad_norm": 0.41370365023612976,
+      "learning_rate": 4.760816517979369e-05,
+      "loss": 1.2548,
+      "num_input_tokens_seen": 134284456,
+      "step": 3430
+    },
+    {
+      "epoch": 0.16668281810252933,
+      "grad_norm": 0.4398741126060486,
+      "learning_rate": 4.7591390403660326e-05,
+      "loss": 1.2549,
+      "num_input_tokens_seen": 134684752,
+      "step": 3440
+    },
+    {
+      "epoch": 0.16716736117840877,
+      "grad_norm": 0.41080906987190247,
+      "learning_rate": 4.7574559985717196e-05,
+      "loss": 1.1814,
+      "num_input_tokens_seen": 135115868,
+      "step": 3450
+    },
+    {
+      "epoch": 0.16765190425428822,
+      "grad_norm": 0.4656144678592682,
+      "learning_rate": 4.7557673967417024e-05,
+      "loss": 1.1833,
+      "num_input_tokens_seen": 135511836,
+      "step": 3460
+    },
+    {
+      "epoch": 0.16813644733016767,
+      "grad_norm": 0.3965156376361847,
+      "learning_rate": 4.7540732390349494e-05,
+      "loss": 1.2127,
+      "num_input_tokens_seen": 135920396,
+      "step": 3470
+    },
+    {
+      "epoch": 0.1686209904060471,
+      "grad_norm": 0.400642991065979,
+      "learning_rate": 4.752373529624113e-05,
+      "loss": 1.2603,
+      "num_input_tokens_seen": 136333964,
+      "step": 3480
+    },
+    {
+      "epoch": 0.16910553348192656,
+      "grad_norm": 0.4004031717777252,
+      "learning_rate": 4.7506682726955166e-05,
+      "loss": 1.2232,
+      "num_input_tokens_seen": 136718116,
+      "step": 3490
+    },
+    {
+      "epoch": 0.169590076557806,
+      "grad_norm": 0.40622180700302124,
+      "learning_rate": 4.74895747244915e-05,
+      "loss": 1.1966,
+      "num_input_tokens_seen": 137109584,
+      "step": 3500
+    },
+    {
+      "epoch": 0.17007461963368545,
+      "grad_norm": 0.41456958651542664,
+      "learning_rate": 4.747241133098655e-05,
+      "loss": 1.2557,
+      "num_input_tokens_seen": 137510040,
+      "step": 3510
+    },
+    {
+      "epoch": 0.1705591627095649,
+      "grad_norm": 0.38758885860443115,
+      "learning_rate": 4.745519258871314e-05,
+      "loss": 1.2414,
+      "num_input_tokens_seen": 137901036,
+      "step": 3520
+    },
+    {
+      "epoch": 0.17104370578544434,
+      "grad_norm": 0.4364662766456604,
+      "learning_rate": 4.743791854008045e-05,
+      "loss": 1.2312,
+      "num_input_tokens_seen": 138292348,
+      "step": 3530
+    },
+    {
+      "epoch": 0.17152824886132378,
+      "grad_norm": 0.40788963437080383,
+      "learning_rate": 4.742058922763386e-05,
+      "loss": 1.2465,
+      "num_input_tokens_seen": 138699080,
+      "step": 3540
+    },
+    {
+      "epoch": 0.17201279193720323,
+      "grad_norm": 0.39772936701774597,
+      "learning_rate": 4.740320469405487e-05,
+      "loss": 1.2562,
+      "num_input_tokens_seen": 139086780,
+      "step": 3550
+    },
+    {
+      "epoch": 0.17249733501308268,
+      "grad_norm": 0.4041747748851776,
+      "learning_rate": 4.738576498216097e-05,
+      "loss": 1.193,
+      "num_input_tokens_seen": 139486640,
+      "step": 3560
+    },
+    {
+      "epoch": 0.17298187808896212,
+      "grad_norm": 0.4034312069416046,
+      "learning_rate": 4.7368270134905565e-05,
+      "loss": 1.1797,
+      "num_input_tokens_seen": 139891212,
+      "step": 3570
+    },
+    {
+      "epoch": 0.17346642116484157,
+      "grad_norm": 0.3874165713787079,
+      "learning_rate": 4.735072019537786e-05,
+      "loss": 1.2149,
+      "num_input_tokens_seen": 140300916,
+      "step": 3580
+    },
+    {
+      "epoch": 0.173950964240721,
+      "grad_norm": 0.3958899974822998,
+      "learning_rate": 4.733311520680276e-05,
+      "loss": 1.2303,
+      "num_input_tokens_seen": 140685408,
+      "step": 3590
+    },
+    {
+      "epoch": 0.17443550731660046,
+      "grad_norm": 0.42453455924987793,
+      "learning_rate": 4.7315455212540714e-05,
+      "loss": 1.2372,
+      "num_input_tokens_seen": 141095616,
+      "step": 3600
+    },
+    {
+      "epoch": 0.1749200503924799,
+      "grad_norm": 0.4058285057544708,
+      "learning_rate": 4.7297740256087695e-05,
+      "loss": 1.1825,
+      "num_input_tokens_seen": 141494816,
+      "step": 3610
+    },
+    {
+      "epoch": 0.17540459346835935,
+      "grad_norm": 0.3956435024738312,
+      "learning_rate": 4.727997038107501e-05,
+      "loss": 1.1843,
+      "num_input_tokens_seen": 141885796,
+      "step": 3620
+    },
+    {
+      "epoch": 0.1758891365442388,
+      "grad_norm": 0.40899357199668884,
+      "learning_rate": 4.726214563126926e-05,
+      "loss": 1.2292,
+      "num_input_tokens_seen": 142285860,
+      "step": 3630
+    },
+    {
+      "epoch": 0.17637367962011824,
+      "grad_norm": 0.41558772325515747,
+      "learning_rate": 4.7244266050572175e-05,
+      "loss": 1.2355,
+      "num_input_tokens_seen": 142706648,
+      "step": 3640
+    },
+    {
+      "epoch": 0.17685822269599769,
+      "grad_norm": 0.40557295083999634,
+      "learning_rate": 4.722633168302054e-05,
+      "loss": 1.2054,
+      "num_input_tokens_seen": 143098084,
+      "step": 3650
+    },
+    {
+      "epoch": 0.17734276577187713,
+      "grad_norm": 0.41156765818595886,
+      "learning_rate": 4.720834257278609e-05,
+      "loss": 1.1816,
+      "num_input_tokens_seen": 143490440,
+      "step": 3660
+    },
+    {
+      "epoch": 0.17782730884775658,
+      "grad_norm": 0.4566863179206848,
+      "learning_rate": 4.719029876417537e-05,
+      "loss": 1.2687,
+      "num_input_tokens_seen": 143900176,
+      "step": 3670
+    },
+    {
+      "epoch": 0.17831185192363602,
+      "grad_norm": 0.38572782278060913,
+      "learning_rate": 4.717220030162964e-05,
+      "loss": 1.2074,
+      "num_input_tokens_seen": 144266036,
+      "step": 3680
+    },
+    {
+      "epoch": 0.17879639499951547,
+      "grad_norm": 0.4045168459415436,
+      "learning_rate": 4.715404722972481e-05,
+      "loss": 1.2177,
+      "num_input_tokens_seen": 144661236,
+      "step": 3690
+    },
+    {
+      "epoch": 0.1792809380753949,
+      "grad_norm": 0.43136268854141235,
+      "learning_rate": 4.713583959317125e-05,
+      "loss": 1.231,
+      "num_input_tokens_seen": 145039188,
+      "step": 3700
+    },
+    {
+      "epoch": 0.17976548115127436,
+      "grad_norm": 0.37638697028160095,
+      "learning_rate": 4.711757743681375e-05,
+      "loss": 1.265,
+      "num_input_tokens_seen": 145471096,
+      "step": 3710
+    },
+    {
+      "epoch": 0.1802500242271538,
+      "grad_norm": 0.418754905462265,
+      "learning_rate": 4.7099260805631354e-05,
+      "loss": 1.2135,
+      "num_input_tokens_seen": 145875596,
+      "step": 3720
+    },
+    {
+      "epoch": 0.18073456730303325,
+      "grad_norm": 0.3730713725090027,
+      "learning_rate": 4.7080889744737275e-05,
+      "loss": 1.219,
+      "num_input_tokens_seen": 146269480,
+      "step": 3730
+    },
+    {
+      "epoch": 0.1812191103789127,
+      "grad_norm": 0.3674296438694,
+      "learning_rate": 4.7062464299378835e-05,
+      "loss": 1.2366,
+      "num_input_tokens_seen": 146676848,
+      "step": 3740
+    },
+    {
+      "epoch": 0.18170365345479214,
+      "grad_norm": 0.4344138205051422,
+      "learning_rate": 4.7043984514937236e-05,
+      "loss": 1.2243,
+      "num_input_tokens_seen": 147058324,
+      "step": 3750
+    },
+    {
+      "epoch": 0.1821881965306716,
+      "grad_norm": 0.3907421827316284,
+      "learning_rate": 4.7025450436927555e-05,
+      "loss": 1.2282,
+      "num_input_tokens_seen": 147448108,
+      "step": 3760
+    },
+    {
+      "epoch": 0.18267273960655103,
+      "grad_norm": 0.40709200501441956,
+      "learning_rate": 4.700686211099859e-05,
+      "loss": 1.2951,
+      "num_input_tokens_seen": 147840480,
+      "step": 3770
+    },
+    {
+      "epoch": 0.18315728268243048,
+      "grad_norm": 0.4162179231643677,
+      "learning_rate": 4.698821958293273e-05,
+      "loss": 1.2506,
+      "num_input_tokens_seen": 148246316,
+      "step": 3780
+    },
+    {
+      "epoch": 0.18364182575830992,
+      "grad_norm": 0.41495996713638306,
+      "learning_rate": 4.696952289864587e-05,
+      "loss": 1.2097,
+      "num_input_tokens_seen": 148608520,
+      "step": 3790
+    },
+    {
+      "epoch": 0.18412636883418937,
+      "grad_norm": 0.3995479643344879,
+      "learning_rate": 4.6950772104187303e-05,
+      "loss": 1.2055,
+      "num_input_tokens_seen": 148986420,
+      "step": 3800
+    },
+    {
+      "epoch": 0.18461091191006881,
+      "grad_norm": 0.3885250985622406,
+      "learning_rate": 4.6931967245739586e-05,
+      "loss": 1.1965,
+      "num_input_tokens_seen": 149370912,
+      "step": 3810
+    },
+    {
+      "epoch": 0.18509545498594826,
+      "grad_norm": 0.4217512905597687,
+      "learning_rate": 4.691310836961843e-05,
+      "loss": 1.2325,
+      "num_input_tokens_seen": 149784152,
+      "step": 3820
+    },
+    {
+      "epoch": 0.1855799980618277,
+      "grad_norm": 0.4255019724369049,
+      "learning_rate": 4.689419552227259e-05,
+      "loss": 1.2151,
+      "num_input_tokens_seen": 150163748,
+      "step": 3830
+    },
+    {
+      "epoch": 0.18606454113770715,
+      "grad_norm": 0.3889770209789276,
+      "learning_rate": 4.687522875028376e-05,
+      "loss": 1.1787,
+      "num_input_tokens_seen": 150567952,
+      "step": 3840
+    },
+    {
+      "epoch": 0.1865490842135866,
+      "grad_norm": 0.39190900325775146,
+      "learning_rate": 4.685620810036642e-05,
+      "loss": 1.2238,
+      "num_input_tokens_seen": 150990236,
+      "step": 3850
+    },
+    {
+      "epoch": 0.18703362728946604,
+      "grad_norm": 0.380062997341156,
+      "learning_rate": 4.683713361936779e-05,
+      "loss": 1.2109,
+      "num_input_tokens_seen": 151403764,
+      "step": 3860
+    },
+    {
+      "epoch": 0.1875181703653455,
+      "grad_norm": 0.38644662499427795,
+      "learning_rate": 4.681800535426765e-05,
+      "loss": 1.1836,
+      "num_input_tokens_seen": 151781412,
+      "step": 3870
+    },
+    {
+      "epoch": 0.18800271344122493,
+      "grad_norm": 0.41668441891670227,
+      "learning_rate": 4.679882335217825e-05,
+      "loss": 1.1741,
+      "num_input_tokens_seen": 152190552,
+      "step": 3880
+    },
+    {
+      "epoch": 0.18848725651710438,
+      "grad_norm": 0.43978846073150635,
+      "learning_rate": 4.6779587660344195e-05,
+      "loss": 1.2325,
+      "num_input_tokens_seen": 152621828,
+      "step": 3890
+    },
+    {
+      "epoch": 0.18897179959298382,
+      "grad_norm": 0.3876698911190033,
+      "learning_rate": 4.676029832614231e-05,
+      "loss": 1.2592,
+      "num_input_tokens_seen": 153007936,
+      "step": 3900
+    },
+    {
+      "epoch": 0.18945634266886327,
+      "grad_norm": 0.40177682042121887,
+      "learning_rate": 4.6740955397081594e-05,
+      "loss": 1.2129,
+      "num_input_tokens_seen": 153391480,
+      "step": 3910
+    },
+    {
+      "epoch": 0.18994088574474272,
+      "grad_norm": 0.44533804059028625,
+      "learning_rate": 4.672155892080298e-05,
+      "loss": 1.2284,
+      "num_input_tokens_seen": 153772340,
+      "step": 3920
+    },
+    {
+      "epoch": 0.19042542882062216,
+      "grad_norm": 0.4126012921333313,
+      "learning_rate": 4.670210894507932e-05,
+      "loss": 1.2157,
+      "num_input_tokens_seen": 154169508,
+      "step": 3930
+    },
+    {
+      "epoch": 0.1909099718965016,
+      "grad_norm": 0.4048556685447693,
+      "learning_rate": 4.668260551781522e-05,
+      "loss": 1.2128,
+      "num_input_tokens_seen": 154556080,
+      "step": 3940
+    },
+    {
+      "epoch": 0.19139451497238105,
+      "grad_norm": 0.41531628370285034,
+      "learning_rate": 4.6663048687046965e-05,
+      "loss": 1.2164,
+      "num_input_tokens_seen": 154956220,
+      "step": 3950
+    },
+    {
+      "epoch": 0.1918790580482605,
+      "grad_norm": 0.4042895436286926,
+      "learning_rate": 4.6643438500942324e-05,
+      "loss": 1.2069,
+      "num_input_tokens_seen": 155328904,
+      "step": 3960
+    },
+    {
+      "epoch": 0.19236360112413994,
+      "grad_norm": 0.38158828020095825,
+      "learning_rate": 4.662377500780053e-05,
+      "loss": 1.2578,
+      "num_input_tokens_seen": 155742432,
+      "step": 3970
+    },
+    {
+      "epoch": 0.1928481442000194,
+      "grad_norm": 0.3938775956630707,
+      "learning_rate": 4.660405825605207e-05,
+      "loss": 1.2278,
+      "num_input_tokens_seen": 156106304,
+      "step": 3980
+    },
+    {
+      "epoch": 0.19333268727589883,
+      "grad_norm": 0.4105328917503357,
+      "learning_rate": 4.6584288294258623e-05,
+      "loss": 1.2094,
+      "num_input_tokens_seen": 156496196,
+      "step": 3990
+    },
+    {
+      "epoch": 0.19381723035177828,
+      "grad_norm": 0.3988109827041626,
+      "learning_rate": 4.6564465171112916e-05,
+      "loss": 1.2248,
+      "num_input_tokens_seen": 156868756,
+      "step": 4000
+    },
+    {
+      "epoch": 0.19381723035177828,
+      "eval_loss": 1.2256019115447998,
+      "eval_runtime": 5.1669,
+      "eval_samples_per_second": 29.031,
+      "eval_steps_per_second": 3.677,
+      "num_input_tokens_seen": 156868756,
+      "step": 4000
+    },
+    {
+      "epoch": 0.19430177342765773,
+      "grad_norm": 0.3892788887023926,
+      "learning_rate": 4.654458893543861e-05,
+      "loss": 1.1957,
+      "num_input_tokens_seen": 157256772,
+      "step": 4010
+    },
+    {
+      "epoch": 0.19478631650353717,
+      "grad_norm": 0.42166975140571594,
+      "learning_rate": 4.6524659636190183e-05,
+      "loss": 1.2207,
+      "num_input_tokens_seen": 157653400,
+      "step": 4020
+    },
+    {
+      "epoch": 0.19527085957941662,
+      "grad_norm": 0.4215097725391388,
+      "learning_rate": 4.650467732245282e-05,
+      "loss": 1.257,
+      "num_input_tokens_seen": 158042052,
+      "step": 4030
+    },
+    {
+      "epoch": 0.19575540265529606,
+      "grad_norm": 0.4513319432735443,
+      "learning_rate": 4.648464204344224e-05,
+      "loss": 1.2021,
+      "num_input_tokens_seen": 158421912,
+      "step": 4040
+    },
+    {
+      "epoch": 0.1962399457311755,
+      "grad_norm": 0.3815930485725403,
+      "learning_rate": 4.646455384850466e-05,
+      "loss": 1.2464,
+      "num_input_tokens_seen": 158794420,
+      "step": 4050
+    },
+    {
+      "epoch": 0.19672448880705495,
+      "grad_norm": 0.4318133294582367,
+      "learning_rate": 4.64444127871166e-05,
+      "loss": 1.1736,
+      "num_input_tokens_seen": 159183072,
+      "step": 4060
+    },
+    {
+      "epoch": 0.1972090318829344,
+      "grad_norm": 0.4135242700576782,
+      "learning_rate": 4.6424218908884795e-05,
+      "loss": 1.2094,
+      "num_input_tokens_seen": 159567032,
+      "step": 4070
+    },
+    {
+      "epoch": 0.19769357495881384,
+      "grad_norm": 0.410666823387146,
+      "learning_rate": 4.640397226354607e-05,
+      "loss": 1.2323,
+      "num_input_tokens_seen": 159937052,
+      "step": 4080
+    },
+    {
+      "epoch": 0.1981781180346933,
+      "grad_norm": 0.42202097177505493,
+      "learning_rate": 4.63836729009672e-05,
+      "loss": 1.209,
+      "num_input_tokens_seen": 160346176,
+      "step": 4090
+    },
+    {
+      "epoch": 0.19866266111057274,
+      "grad_norm": 0.42181599140167236,
+      "learning_rate": 4.636332087114481e-05,
+      "loss": 1.1995,
+      "num_input_tokens_seen": 160734656,
+      "step": 4100
+    },
+    {
+      "epoch": 0.19914720418645218,
+      "grad_norm": 0.42154404520988464,
+      "learning_rate": 4.6342916224205254e-05,
+      "loss": 1.2422,
+      "num_input_tokens_seen": 161143420,
+      "step": 4110
+    },
+    {
+      "epoch": 0.19963174726233163,
+      "grad_norm": 0.38917213678359985,
+      "learning_rate": 4.632245901040446e-05,
+      "loss": 1.2135,
+      "num_input_tokens_seen": 161523856,
+      "step": 4120
+    },
+    {
+      "epoch": 0.20011629033821107,
+      "grad_norm": 0.41019120812416077,
+      "learning_rate": 4.6301949280127835e-05,
+      "loss": 1.2309,
+      "num_input_tokens_seen": 161906012,
+      "step": 4130
+    },
+    {
+      "epoch": 0.20060083341409052,
+      "grad_norm": 0.3908578157424927,
+      "learning_rate": 4.6281387083890134e-05,
+      "loss": 1.2335,
+      "num_input_tokens_seen": 162302056,
+      "step": 4140
+    },
+    {
+      "epoch": 0.20108537648996996,
+      "grad_norm": 0.4184846878051758,
+      "learning_rate": 4.626077247233533e-05,
+      "loss": 1.207,
+      "num_input_tokens_seen": 162674300,
+      "step": 4150
+    },
+    {
+      "epoch": 0.2015699195658494,
+      "grad_norm": 0.41114839911460876,
+      "learning_rate": 4.62401054962365e-05,
+      "loss": 1.2287,
+      "num_input_tokens_seen": 163082608,
+      "step": 4160
+    },
+    {
+      "epoch": 0.20205446264172885,
+      "grad_norm": 0.4034859538078308,
+      "learning_rate": 4.6219386206495684e-05,
+      "loss": 1.2536,
+      "num_input_tokens_seen": 163463012,
+      "step": 4170
+    },
+    {
+      "epoch": 0.2025390057176083,
+      "grad_norm": 0.37627366185188293,
+      "learning_rate": 4.6198614654143765e-05,
+      "loss": 1.2428,
+      "num_input_tokens_seen": 163855604,
+      "step": 4180
+    },
+    {
+      "epoch": 0.20302354879348775,
+      "grad_norm": 0.3891449272632599,
+      "learning_rate": 4.6177790890340376e-05,
+      "loss": 1.2113,
+      "num_input_tokens_seen": 164258760,
+      "step": 4190
+    },
+    {
+      "epoch": 0.2035080918693672,
+      "grad_norm": 0.3885304033756256,
+      "learning_rate": 4.615691496637371e-05,
+      "loss": 1.2276,
+      "num_input_tokens_seen": 164630872,
+      "step": 4200
+    },
+    {
+      "epoch": 0.20399263494524664,
+      "grad_norm": 0.4263809323310852,
+      "learning_rate": 4.613598693366044e-05,
+      "loss": 1.1905,
+      "num_input_tokens_seen": 165040256,
+      "step": 4210
+    },
+    {
+      "epoch": 0.20447717802112608,
+      "grad_norm": 0.3751036524772644,
+      "learning_rate": 4.611500684374559e-05,
+      "loss": 1.2174,
+      "num_input_tokens_seen": 165424264,
+      "step": 4220
+    },
+    {
+      "epoch": 0.20496172109700553,
+      "grad_norm": 0.40289247035980225,
+      "learning_rate": 4.6093974748302385e-05,
+      "loss": 1.1899,
+      "num_input_tokens_seen": 165812068,
+      "step": 4230
+    },
+    {
+      "epoch": 0.20544626417288497,
+      "grad_norm": 0.4211735427379608,
+      "learning_rate": 4.6072890699132155e-05,
+      "loss": 1.1992,
+      "num_input_tokens_seen": 166177560,
+      "step": 4240
+    },
+    {
+      "epoch": 0.20593080724876442,
+      "grad_norm": 0.36460644006729126,
+      "learning_rate": 4.605175474816418e-05,
+      "loss": 1.2103,
+      "num_input_tokens_seen": 166578756,
+      "step": 4250
+    },
+    {
+      "epoch": 0.20641535032464386,
+      "grad_norm": 0.3921453356742859,
+      "learning_rate": 4.603056694745556e-05,
+      "loss": 1.2143,
+      "num_input_tokens_seen": 166975748,
+      "step": 4260
+    },
+    {
+      "epoch": 0.2068998934005233,
+      "grad_norm": 0.3987452983856201,
+      "learning_rate": 4.600932734919113e-05,
+      "loss": 1.2292,
+      "num_input_tokens_seen": 167388080,
+      "step": 4270
+    },
+    {
+      "epoch": 0.20738443647640276,
+      "grad_norm": 0.4142529368400574,
+      "learning_rate": 4.5988036005683265e-05,
+      "loss": 1.2048,
+      "num_input_tokens_seen": 167784884,
+      "step": 4280
+    },
+    {
+      "epoch": 0.2078689795522822,
+      "grad_norm": 0.39146703481674194,
+      "learning_rate": 4.596669296937182e-05,
+      "loss": 1.2422,
+      "num_input_tokens_seen": 168184036,
+      "step": 4290
+    },
+    {
+      "epoch": 0.20835352262816165,
+      "grad_norm": 0.3954419791698456,
+      "learning_rate": 4.594529829282395e-05,
+      "loss": 1.2297,
+      "num_input_tokens_seen": 168575028,
+      "step": 4300
+    },
+    {
+      "epoch": 0.2088380657040411,
+      "grad_norm": 0.4144536554813385,
+      "learning_rate": 4.5923852028733985e-05,
+      "loss": 1.1951,
+      "num_input_tokens_seen": 168948296,
+      "step": 4310
+    },
+    {
+      "epoch": 0.20932260877992054,
+      "grad_norm": 0.3958514630794525,
+      "learning_rate": 4.590235422992335e-05,
+      "loss": 1.1684,
+      "num_input_tokens_seen": 169357988,
+      "step": 4320
+    },
+    {
+      "epoch": 0.20980715185579998,
+      "grad_norm": 0.40073680877685547,
+      "learning_rate": 4.588080494934036e-05,
+      "loss": 1.2098,
+      "num_input_tokens_seen": 169761752,
+      "step": 4330
+    },
+    {
+      "epoch": 0.21029169493167943,
+      "grad_norm": 0.37582847476005554,
+      "learning_rate": 4.585920424006015e-05,
+      "loss": 1.2668,
+      "num_input_tokens_seen": 170160840,
+      "step": 4340
+    },
+    {
+      "epoch": 0.21077623800755887,
+      "grad_norm": 0.36600205302238464,
+      "learning_rate": 4.5837552155284516e-05,
+      "loss": 1.2413,
+      "num_input_tokens_seen": 170562748,
+      "step": 4350
+    },
+    {
+      "epoch": 0.21126078108343832,
+      "grad_norm": 0.41803476214408875,
+      "learning_rate": 4.581584874834179e-05,
+      "loss": 1.2522,
+      "num_input_tokens_seen": 170956952,
+      "step": 4360
+    },
+    {
+      "epoch": 0.21174532415931777,
+      "grad_norm": 0.43290525674819946,
+      "learning_rate": 4.5794094072686716e-05,
+      "loss": 1.254,
+      "num_input_tokens_seen": 171333724,
+      "step": 4370
+    },
+    {
+      "epoch": 0.2122298672351972,
+      "grad_norm": 0.4034164845943451,
+      "learning_rate": 4.5772288181900294e-05,
+      "loss": 1.2486,
+      "num_input_tokens_seen": 171727380,
+      "step": 4380
+    },
+    {
+      "epoch": 0.21271441031107666,
+      "grad_norm": 0.4040049910545349,
+      "learning_rate": 4.575043112968969e-05,
+      "loss": 1.24,
+      "num_input_tokens_seen": 172114780,
+      "step": 4390
+    },
+    {
+      "epoch": 0.2131989533869561,
+      "grad_norm": 0.39204099774360657,
+      "learning_rate": 4.5728522969888044e-05,
+      "loss": 1.2371,
+      "num_input_tokens_seen": 172504760,
+      "step": 4400
+    },
+    {
+      "epoch": 0.21368349646283555,
+      "grad_norm": 0.41550543904304504,
+      "learning_rate": 4.5706563756454414e-05,
+      "loss": 1.1737,
+      "num_input_tokens_seen": 172866336,
+      "step": 4410
+    },
+    {
+      "epoch": 0.214168039538715,
+      "grad_norm": 0.3796961009502411,
+      "learning_rate": 4.5684553543473555e-05,
+      "loss": 1.2245,
+      "num_input_tokens_seen": 173274608,
+      "step": 4420
+    },
+    {
+      "epoch": 0.21465258261459444,
+      "grad_norm": 0.39554864168167114,
+      "learning_rate": 4.5662492385155886e-05,
+      "loss": 1.2163,
+      "num_input_tokens_seen": 173670996,
+      "step": 4430
+    },
+    {
+      "epoch": 0.21513712569047388,
+      "grad_norm": 0.41738855838775635,
+      "learning_rate": 4.564038033583725e-05,
+      "loss": 1.1744,
+      "num_input_tokens_seen": 174089336,
+      "step": 4440
+    },
+    {
+      "epoch": 0.21562166876635333,
+      "grad_norm": 0.38417306542396545,
+      "learning_rate": 4.561821744997887e-05,
+      "loss": 1.2079,
+      "num_input_tokens_seen": 174457260,
+      "step": 4450
+    },
+    {
+      "epoch": 0.21610621184223278,
+      "grad_norm": 0.3893676698207855,
+      "learning_rate": 4.5596003782167154e-05,
+      "loss": 1.2236,
+      "num_input_tokens_seen": 174845308,
+      "step": 4460
+    },
+    {
+      "epoch": 0.21659075491811222,
+      "grad_norm": 0.5625600218772888,
+      "learning_rate": 4.55737393871136e-05,
+      "loss": 1.1457,
+      "num_input_tokens_seen": 175217384,
+      "step": 4470
+    },
+    {
+      "epoch": 0.21707529799399167,
+      "grad_norm": 0.4266904294490814,
+      "learning_rate": 4.555142431965465e-05,
+      "loss": 1.2847,
+      "num_input_tokens_seen": 175597084,
+      "step": 4480
+    },
+    {
+      "epoch": 0.2175598410698711,
+      "grad_norm": 0.40159592032432556,
+      "learning_rate": 4.552905863475152e-05,
+      "loss": 1.2064,
+      "num_input_tokens_seen": 175985196,
+      "step": 4490
+    },
+    {
+      "epoch": 0.21804438414575056,
+      "grad_norm": 0.3944200873374939,
+      "learning_rate": 4.550664238749012e-05,
+      "loss": 1.1969,
+      "num_input_tokens_seen": 176390156,
+      "step": 4500
+    },
+    {
+      "epoch": 0.21852892722163,
+      "grad_norm": 0.37802281975746155,
+      "learning_rate": 4.54841756330809e-05,
+      "loss": 1.1757,
+      "num_input_tokens_seen": 176751292,
+      "step": 4510
+    },
+    {
+      "epoch": 0.21901347029750945,
+      "grad_norm": 0.46125710010528564,
+      "learning_rate": 4.546165842685869e-05,
+      "loss": 1.1816,
+      "num_input_tokens_seen": 177136472,
+      "step": 4520
+    },
+    {
+      "epoch": 0.2194980133733889,
+      "grad_norm": 0.40457674860954285,
+      "learning_rate": 4.543909082428257e-05,
+      "loss": 1.2635,
+      "num_input_tokens_seen": 177564052,
+      "step": 4530
+    },
+    {
+      "epoch": 0.21998255644926834,
+      "grad_norm": 0.3985607326030731,
+      "learning_rate": 4.541647288093579e-05,
+      "loss": 1.2209,
+      "num_input_tokens_seen": 177977356,
+      "step": 4540
+    },
+    {
+      "epoch": 0.22046709952514779,
+      "grad_norm": 0.41993504762649536,
+      "learning_rate": 4.539380465252554e-05,
+      "loss": 1.1924,
+      "num_input_tokens_seen": 178361500,
+      "step": 4550
+    },
+    {
+      "epoch": 0.22095164260102723,
+      "grad_norm": 0.4131138324737549,
+      "learning_rate": 4.5371086194882886e-05,
+      "loss": 1.2114,
+      "num_input_tokens_seen": 178786424,
+      "step": 4560
+    },
+    {
+      "epoch": 0.22143618567690668,
+      "grad_norm": 0.3843264579772949,
+      "learning_rate": 4.5348317563962594e-05,
+      "loss": 1.2211,
+      "num_input_tokens_seen": 179173444,
+      "step": 4570
+    },
+    {
+      "epoch": 0.22192072875278612,
+      "grad_norm": 0.39719441533088684,
+      "learning_rate": 4.532549881584301e-05,
+      "loss": 1.2099,
+      "num_input_tokens_seen": 179550012,
+      "step": 4580
+    },
+    {
+      "epoch": 0.22240527182866557,
+      "grad_norm": 0.41848018765449524,
+      "learning_rate": 4.5302630006725934e-05,
+      "loss": 1.2015,
+      "num_input_tokens_seen": 179934896,
+      "step": 4590
+    },
+    {
+      "epoch": 0.22288981490454501,
+      "grad_norm": 0.37484750151634216,
+      "learning_rate": 4.527971119293643e-05,
+      "loss": 1.2331,
+      "num_input_tokens_seen": 180319432,
+      "step": 4600
+    },
+    {
+      "epoch": 0.22337435798042446,
+      "grad_norm": 0.40925487875938416,
+      "learning_rate": 4.525674243092275e-05,
+      "loss": 1.2214,
+      "num_input_tokens_seen": 180713196,
+      "step": 4610
+    },
+    {
+      "epoch": 0.2238589010563039,
+      "grad_norm": 0.381757915019989,
+      "learning_rate": 4.523372377725615e-05,
+      "loss": 1.1964,
+      "num_input_tokens_seen": 181112116,
+      "step": 4620
+    },
+    {
+      "epoch": 0.22434344413218335,
+      "grad_norm": 0.3882332742214203,
+      "learning_rate": 4.5210655288630774e-05,
+      "loss": 1.189,
+      "num_input_tokens_seen": 181537372,
+      "step": 4630
+    },
+    {
+      "epoch": 0.2248279872080628,
+      "grad_norm": 0.3924133777618408,
+      "learning_rate": 4.518753702186352e-05,
+      "loss": 1.2174,
+      "num_input_tokens_seen": 181917440,
+      "step": 4640
+    },
+    {
+      "epoch": 0.22531253028394224,
+      "grad_norm": 0.3810102045536041,
+      "learning_rate": 4.5164369033893874e-05,
+      "loss": 1.235,
+      "num_input_tokens_seen": 182302576,
+      "step": 4650
+    },
+    {
+      "epoch": 0.2257970733598217,
+      "grad_norm": 0.4031634032726288,
+      "learning_rate": 4.514115138178378e-05,
+      "loss": 1.2213,
+      "num_input_tokens_seen": 182677192,
+      "step": 4660
+    },
+    {
+      "epoch": 0.22628161643570113,
+      "grad_norm": 0.4006711542606354,
+      "learning_rate": 4.511788412271749e-05,
+      "loss": 1.1881,
+      "num_input_tokens_seen": 183057516,
+      "step": 4670
+    },
+    {
+      "epoch": 0.22676615951158058,
+      "grad_norm": 0.42727628350257874,
+      "learning_rate": 4.509456731400149e-05,
+      "loss": 1.1956,
+      "num_input_tokens_seen": 183446504,
+      "step": 4680
+    },
+    {
+      "epoch": 0.22725070258746002,
+      "grad_norm": 0.419729620218277,
+      "learning_rate": 4.507120101306425e-05,
+      "loss": 1.2287,
+      "num_input_tokens_seen": 183853916,
+      "step": 4690
+    },
+    {
+      "epoch": 0.22773524566333947,
+      "grad_norm": 0.3630729019641876,
+      "learning_rate": 4.504778527745616e-05,
+      "loss": 1.179,
+      "num_input_tokens_seen": 184251140,
+      "step": 4700
+    },
+    {
+      "epoch": 0.22821978873921892,
+      "grad_norm": 0.3897242248058319,
+      "learning_rate": 4.5024320164849366e-05,
+      "loss": 1.1704,
+      "num_input_tokens_seen": 184655808,
+      "step": 4710
+    },
+    {
+      "epoch": 0.22870433181509836,
+      "grad_norm": 0.3991836905479431,
+      "learning_rate": 4.500080573303762e-05,
+      "loss": 1.2208,
+      "num_input_tokens_seen": 185054436,
+      "step": 4720
+    },
+    {
+      "epoch": 0.2291888748909778,
+      "grad_norm": 0.39902886748313904,
+      "learning_rate": 4.497724203993615e-05,
+      "loss": 1.2088,
+      "num_input_tokens_seen": 185439916,
+      "step": 4730
+    },
+    {
+      "epoch": 0.22967341796685725,
+      "grad_norm": 0.39412978291511536,
+      "learning_rate": 4.495362914358152e-05,
+      "loss": 1.2317,
+      "num_input_tokens_seen": 185844756,
+      "step": 4740
+    },
+    {
+      "epoch": 0.2301579610427367,
+      "grad_norm": 0.37934547662734985,
+      "learning_rate": 4.4929967102131473e-05,
+      "loss": 1.2081,
+      "num_input_tokens_seen": 186226308,
+      "step": 4750
+    },
+    {
+      "epoch": 0.23064250411861614,
+      "grad_norm": 0.405785471200943,
+      "learning_rate": 4.49062559738648e-05,
+      "loss": 1.1877,
+      "num_input_tokens_seen": 186592700,
+      "step": 4760
+    },
+    {
+      "epoch": 0.2311270471944956,
+      "grad_norm": 0.36941656470298767,
+      "learning_rate": 4.488249581718118e-05,
+      "loss": 1.2322,
+      "num_input_tokens_seen": 186963268,
+      "step": 4770
+    },
+    {
+      "epoch": 0.23161159027037503,
+      "grad_norm": 0.3917679488658905,
+      "learning_rate": 4.485868669060104e-05,
+      "loss": 1.2006,
+      "num_input_tokens_seen": 187364124,
+      "step": 4780
+    },
+    {
+      "epoch": 0.23209613334625448,
+      "grad_norm": 0.38910412788391113,
+      "learning_rate": 4.483482865276545e-05,
+      "loss": 1.2244,
+      "num_input_tokens_seen": 187761036,
+      "step": 4790
+    },
+    {
+      "epoch": 0.23258067642213393,
+      "grad_norm": 0.42575618624687195,
+      "learning_rate": 4.481092176243592e-05,
+      "loss": 1.2119,
+      "num_input_tokens_seen": 188156176,
+      "step": 4800
+    },
+    {
+      "epoch": 0.23306521949801337,
+      "grad_norm": 0.3848058879375458,
+      "learning_rate": 4.4786966078494296e-05,
+      "loss": 1.2539,
+      "num_input_tokens_seen": 188549760,
+      "step": 4810
+    },
+    {
+      "epoch": 0.23354976257389282,
+      "grad_norm": 0.4218595623970032,
+      "learning_rate": 4.47629616599426e-05,
+      "loss": 1.2262,
+      "num_input_tokens_seen": 188935732,
+      "step": 4820
+    },
+    {
+      "epoch": 0.23403430564977226,
+      "grad_norm": 0.3957028388977051,
+      "learning_rate": 4.473890856590287e-05,
+      "loss": 1.2091,
+      "num_input_tokens_seen": 189364744,
+      "step": 4830
+    },
+    {
+      "epoch": 0.2345188487256517,
+      "grad_norm": 0.3964502811431885,
+      "learning_rate": 4.471480685561704e-05,
+      "loss": 1.1825,
+      "num_input_tokens_seen": 189761404,
+      "step": 4840
+    },
+    {
+      "epoch": 0.23500339180153115,
+      "grad_norm": 0.44072818756103516,
+      "learning_rate": 4.469065658844679e-05,
+      "loss": 1.1775,
+      "num_input_tokens_seen": 190143308,
+      "step": 4850
+    },
+    {
+      "epoch": 0.2354879348774106,
+      "grad_norm": 0.4088262617588043,
+      "learning_rate": 4.466645782387339e-05,
+      "loss": 1.1938,
+      "num_input_tokens_seen": 190553260,
+      "step": 4860
+    },
+    {
+      "epoch": 0.23597247795329004,
+      "grad_norm": 0.3793879449367523,
+      "learning_rate": 4.464221062149756e-05,
+      "loss": 1.184,
+      "num_input_tokens_seen": 190944324,
+      "step": 4870
+    },
+    {
+      "epoch": 0.2364570210291695,
+      "grad_norm": 0.3843640089035034,
+      "learning_rate": 4.461791504103931e-05,
+      "loss": 1.1906,
+      "num_input_tokens_seen": 191314092,
+      "step": 4880
+    },
+    {
+      "epoch": 0.23694156410504894,
+      "grad_norm": 0.42063283920288086,
+      "learning_rate": 4.459357114233781e-05,
+      "loss": 1.2077,
+      "num_input_tokens_seen": 191680088,
+      "step": 4890
+    },
+    {
+      "epoch": 0.23742610718092838,
+      "grad_norm": 0.3904326558113098,
+      "learning_rate": 4.4569178985351246e-05,
+      "loss": 1.1678,
+      "num_input_tokens_seen": 192097904,
+      "step": 4900
+    },
+    {
+      "epoch": 0.23791065025680783,
+      "grad_norm": 0.40638238191604614,
+      "learning_rate": 4.4544738630156644e-05,
+      "loss": 1.2103,
+      "num_input_tokens_seen": 192484692,
+      "step": 4910
+    },
+    {
+      "epoch": 0.23839519333268727,
+      "grad_norm": 0.4055958092212677,
+      "learning_rate": 4.4520250136949755e-05,
+      "loss": 1.2244,
+      "num_input_tokens_seen": 192862996,
+      "step": 4920
+    },
+    {
+      "epoch": 0.23887973640856672,
+      "grad_norm": 0.42126190662384033,
+      "learning_rate": 4.449571356604488e-05,
+      "loss": 1.1551,
+      "num_input_tokens_seen": 193256264,
+      "step": 4930
+    },
+    {
+      "epoch": 0.23936427948444616,
+      "grad_norm": 0.3869550824165344,
+      "learning_rate": 4.4471128977874755e-05,
+      "loss": 1.2033,
+      "num_input_tokens_seen": 193631576,
+      "step": 4940
+    },
+    {
+      "epoch": 0.2398488225603256,
+      "grad_norm": 0.4147154986858368,
+      "learning_rate": 4.4446496432990345e-05,
+      "loss": 1.1741,
+      "num_input_tokens_seen": 194010208,
+      "step": 4950
+    },
+    {
+      "epoch": 0.24033336563620505,
+      "grad_norm": 0.39161938428878784,
+      "learning_rate": 4.442181599206078e-05,
+      "loss": 1.2042,
+      "num_input_tokens_seen": 194370712,
+      "step": 4960
+    },
+    {
+      "epoch": 0.2408179087120845,
+      "grad_norm": 0.40375518798828125,
+      "learning_rate": 4.439708771587311e-05,
+      "loss": 1.2074,
+      "num_input_tokens_seen": 194764676,
+      "step": 4970
+    },
+    {
+      "epoch": 0.24130245178796395,
+      "grad_norm": 0.37387263774871826,
+      "learning_rate": 4.437231166533222e-05,
+      "loss": 1.23,
+      "num_input_tokens_seen": 195173608,
+      "step": 4980
+    },
+    {
+      "epoch": 0.2417869948638434,
+      "grad_norm": 0.45811185240745544,
+      "learning_rate": 4.434748790146066e-05,
+      "loss": 1.1996,
+      "num_input_tokens_seen": 195550528,
+      "step": 4990
+    },
+    {
+      "epoch": 0.24227153793972284,
+      "grad_norm": 0.40425029397010803,
+      "learning_rate": 4.432261648539852e-05,
+      "loss": 1.1837,
+      "num_input_tokens_seen": 195956732,
+      "step": 5000
+    },
+    {
+      "epoch": 0.24275608101560228,
+      "grad_norm": 0.40268510580062866,
+      "learning_rate": 4.42976974784032e-05,
+      "loss": 1.25,
+      "num_input_tokens_seen": 196340952,
+      "step": 5010
+    },
+    {
+      "epoch": 0.24324062409148173,
+      "grad_norm": 0.43269723653793335,
+      "learning_rate": 4.427273094184938e-05,
+      "loss": 1.174,
+      "num_input_tokens_seen": 196711452,
+      "step": 5020
+    },
+    {
+      "epoch": 0.24372516716736117,
+      "grad_norm": 0.3802777826786041,
+      "learning_rate": 4.424771693722875e-05,
+      "loss": 1.2558,
+      "num_input_tokens_seen": 197112496,
+      "step": 5030
+    },
+    {
+      "epoch": 0.24420971024324062,
+      "grad_norm": 0.39778804779052734,
+      "learning_rate": 4.422265552614995e-05,
+      "loss": 1.1993,
+      "num_input_tokens_seen": 197504404,
+      "step": 5040
+    },
+    {
+      "epoch": 0.24469425331912006,
+      "grad_norm": 0.4463038146495819,
+      "learning_rate": 4.4197546770338346e-05,
+      "loss": 1.196,
+      "num_input_tokens_seen": 197904548,
+      "step": 5050
+    },
+    {
+      "epoch": 0.2451787963949995,
+      "grad_norm": 0.37699079513549805,
+      "learning_rate": 4.417239073163596e-05,
+      "loss": 1.1525,
+      "num_input_tokens_seen": 198319024,
+      "step": 5060
+    },
+    {
+      "epoch": 0.24566333947087896,
+      "grad_norm": 0.3865511119365692,
+      "learning_rate": 4.414718747200121e-05,
+      "loss": 1.2275,
+      "num_input_tokens_seen": 198714528,
+      "step": 5070
+    },
+    {
+      "epoch": 0.2461478825467584,
+      "grad_norm": 0.3743216395378113,
+      "learning_rate": 4.412193705350886e-05,
+      "loss": 1.2307,
+      "num_input_tokens_seen": 199085164,
+      "step": 5080
+    },
+    {
+      "epoch": 0.24663242562263785,
+      "grad_norm": 0.41756245493888855,
+      "learning_rate": 4.4096639538349835e-05,
+      "loss": 1.2028,
+      "num_input_tokens_seen": 199462308,
+      "step": 5090
+    },
+    {
+      "epoch": 0.2471169686985173,
+      "grad_norm": 0.35077378153800964,
+      "learning_rate": 4.407129498883101e-05,
+      "loss": 1.1789,
+      "num_input_tokens_seen": 199882788,
+      "step": 5100
+    },
+    {
+      "epoch": 0.24760151177439674,
+      "grad_norm": 0.4078513979911804,
+      "learning_rate": 4.404590346737514e-05,
+      "loss": 1.2188,
+      "num_input_tokens_seen": 200271336,
+      "step": 5110
+    },
+    {
+      "epoch": 0.24808605485027618,
+      "grad_norm": 0.42663541436195374,
+      "learning_rate": 4.402046503652065e-05,
+      "loss": 1.2293,
+      "num_input_tokens_seen": 200660788,
+      "step": 5120
+    },
+    {
+      "epoch": 0.24857059792615563,
+      "grad_norm": 0.3861826956272125,
+      "learning_rate": 4.399497975892153e-05,
+      "loss": 1.2182,
+      "num_input_tokens_seen": 201075832,
+      "step": 5130
+    },
+    {
+      "epoch": 0.24905514100203507,
+      "grad_norm": 0.38945886492729187,
+      "learning_rate": 4.3969447697347116e-05,
+      "loss": 1.2464,
+      "num_input_tokens_seen": 201447360,
+      "step": 5140
+    },
+    {
+      "epoch": 0.24953968407791452,
+      "grad_norm": 0.3884553611278534,
+      "learning_rate": 4.3943868914681996e-05,
+      "loss": 1.2353,
+      "num_input_tokens_seen": 201863948,
+      "step": 5150
+    },
+    {
+      "epoch": 0.25002422715379397,
+      "grad_norm": 0.409061461687088,
+      "learning_rate": 4.3918243473925823e-05,
+      "loss": 1.2235,
+      "num_input_tokens_seen": 202265196,
+      "step": 5160
+    },
+    {
+      "epoch": 0.25050877022967344,
+      "grad_norm": 0.43314066529273987,
+      "learning_rate": 4.389257143819318e-05,
+      "loss": 1.2076,
+      "num_input_tokens_seen": 202665352,
+      "step": 5170
+    },
+    {
+      "epoch": 0.25099331330555286,
+      "grad_norm": 0.43011194467544556,
+      "learning_rate": 4.386685287071337e-05,
+      "loss": 1.1834,
+      "num_input_tokens_seen": 203076796,
+      "step": 5180
+    },
+    {
+      "epoch": 0.25147785638143233,
+      "grad_norm": 0.4096457064151764,
+      "learning_rate": 4.3841087834830354e-05,
+      "loss": 1.2336,
+      "num_input_tokens_seen": 203488952,
+      "step": 5190
+    },
+    {
+      "epoch": 0.25196239945731175,
+      "grad_norm": 0.4316791594028473,
+      "learning_rate": 4.381527639400251e-05,
+      "loss": 1.2205,
+      "num_input_tokens_seen": 203855896,
+      "step": 5200
+    },
+    {
+      "epoch": 0.2524469425331912,
+      "grad_norm": 0.4344671666622162,
+      "learning_rate": 4.3789418611802533e-05,
+      "loss": 1.1693,
+      "num_input_tokens_seen": 204240628,
+      "step": 5210
+    },
+    {
+      "epoch": 0.25293148560907064,
+      "grad_norm": 0.378844678401947,
+      "learning_rate": 4.3763514551917236e-05,
+      "loss": 1.1722,
+      "num_input_tokens_seen": 204638860,
+      "step": 5220
+    },
+    {
+      "epoch": 0.2534160286849501,
+      "grad_norm": 0.38779839873313904,
+      "learning_rate": 4.3737564278147406e-05,
+      "loss": 1.2179,
+      "num_input_tokens_seen": 205028820,
+      "step": 5230
+    },
+    {
+      "epoch": 0.25390057176082953,
+      "grad_norm": 0.4080614745616913,
+      "learning_rate": 4.3711567854407683e-05,
+      "loss": 1.1904,
+      "num_input_tokens_seen": 205408752,
+      "step": 5240
+    },
+    {
+      "epoch": 0.254385114836709,
+      "grad_norm": 0.40406113862991333,
+      "learning_rate": 4.3685525344726345e-05,
+      "loss": 1.2017,
+      "num_input_tokens_seen": 205800048,
+      "step": 5250
+    },
+    {
+      "epoch": 0.2548696579125884,
+      "grad_norm": 0.3979373872280121,
+      "learning_rate": 4.365943681324519e-05,
+      "loss": 1.2127,
+      "num_input_tokens_seen": 206178400,
+      "step": 5260
+    },
+    {
+      "epoch": 0.2553542009884679,
+      "grad_norm": 0.406194269657135,
+      "learning_rate": 4.363330232421938e-05,
+      "loss": 1.2299,
+      "num_input_tokens_seen": 206550496,
+      "step": 5270
+    },
+    {
+      "epoch": 0.2558387440643473,
+      "grad_norm": 0.43176546692848206,
+      "learning_rate": 4.3607121942017225e-05,
+      "loss": 1.1843,
+      "num_input_tokens_seen": 206938216,
+      "step": 5280
+    },
+    {
+      "epoch": 0.2563232871402268,
+      "grad_norm": 0.43860721588134766,
+      "learning_rate": 4.358089573112012e-05,
+      "loss": 1.1568,
+      "num_input_tokens_seen": 207342028,
+      "step": 5290
+    },
+    {
+      "epoch": 0.2568078302161062,
+      "grad_norm": 0.40631529688835144,
+      "learning_rate": 4.355462375612231e-05,
+      "loss": 1.2707,
+      "num_input_tokens_seen": 207740984,
+      "step": 5300
+    },
+    {
+      "epoch": 0.2572923732919857,
+      "grad_norm": 0.3890661895275116,
+      "learning_rate": 4.352830608173074e-05,
+      "loss": 1.1957,
+      "num_input_tokens_seen": 208160252,
+      "step": 5310
+    },
+    {
+      "epoch": 0.2577769163678651,
+      "grad_norm": 0.40882420539855957,
+      "learning_rate": 4.350194277276495e-05,
+      "loss": 1.1938,
+      "num_input_tokens_seen": 208537272,
+      "step": 5320
+    },
+    {
+      "epoch": 0.25826145944374457,
+      "grad_norm": 0.4055963456630707,
+      "learning_rate": 4.347553389415684e-05,
+      "loss": 1.1603,
+      "num_input_tokens_seen": 208927108,
+      "step": 5330
+    },
+    {
+      "epoch": 0.258746002519624,
+      "grad_norm": 0.3766281008720398,
+      "learning_rate": 4.3449079510950555e-05,
+      "loss": 1.2169,
+      "num_input_tokens_seen": 209320208,
+      "step": 5340
+    },
+    {
+      "epoch": 0.25923054559550346,
+      "grad_norm": 0.3955957591533661,
+      "learning_rate": 4.3422579688302337e-05,
+      "loss": 1.2096,
+      "num_input_tokens_seen": 209704852,
+      "step": 5350
+    },
+    {
+      "epoch": 0.2597150886713829,
+      "grad_norm": 0.416856050491333,
+      "learning_rate": 4.339603449148031e-05,
+      "loss": 1.2155,
+      "num_input_tokens_seen": 210092272,
+      "step": 5360
+    },
+    {
+      "epoch": 0.26019963174726235,
+      "grad_norm": 0.3843337893486023,
+      "learning_rate": 4.33694439858644e-05,
+      "loss": 1.2163,
+      "num_input_tokens_seen": 210513236,
+      "step": 5370
+    },
+    {
+      "epoch": 0.26068417482314177,
+      "grad_norm": 0.3668496310710907,
+      "learning_rate": 4.334280823694607e-05,
+      "loss": 1.2402,
+      "num_input_tokens_seen": 210927300,
+      "step": 5380
+    },
+    {
+      "epoch": 0.26116871789902124,
+      "grad_norm": 0.38131269812583923,
+      "learning_rate": 4.331612731032826e-05,
+      "loss": 1.1913,
+      "num_input_tokens_seen": 211339824,
+      "step": 5390
+    },
+    {
+      "epoch": 0.26165326097490066,
+      "grad_norm": 0.4179877042770386,
+      "learning_rate": 4.328940127172516e-05,
+      "loss": 1.197,
+      "num_input_tokens_seen": 211752352,
+      "step": 5400
+    },
+    {
+      "epoch": 0.26213780405078013,
+      "grad_norm": 0.42718562483787537,
+      "learning_rate": 4.326263018696208e-05,
+      "loss": 1.199,
+      "num_input_tokens_seen": 212135324,
+      "step": 5410
+    },
+    {
+      "epoch": 0.26262234712665955,
+      "grad_norm": 0.43352535367012024,
+      "learning_rate": 4.3235814121975274e-05,
+      "loss": 1.2094,
+      "num_input_tokens_seen": 212524932,
+      "step": 5420
+    },
+    {
+      "epoch": 0.263106890202539,
+      "grad_norm": 0.396100789308548,
+      "learning_rate": 4.320895314281177e-05,
+      "loss": 1.2254,
+      "num_input_tokens_seen": 212938816,
+      "step": 5430
+    },
+    {
+      "epoch": 0.26359143327841844,
+      "grad_norm": 0.3762286901473999,
+      "learning_rate": 4.318204731562922e-05,
+      "loss": 1.2321,
+      "num_input_tokens_seen": 213336448,
+      "step": 5440
+    },
+    {
+      "epoch": 0.2640759763542979,
+      "grad_norm": 0.35878944396972656,
+      "learning_rate": 4.3155096706695755e-05,
+      "loss": 1.2227,
+      "num_input_tokens_seen": 213756868,
+      "step": 5450
+    },
+    {
+      "epoch": 0.26456051943017733,
+      "grad_norm": 0.4023889899253845,
+      "learning_rate": 4.312810138238979e-05,
+      "loss": 1.2823,
+      "num_input_tokens_seen": 214154532,
+      "step": 5460
+    },
+    {
+      "epoch": 0.2650450625060568,
+      "grad_norm": 0.40626269578933716,
+      "learning_rate": 4.310106140919986e-05,
+      "loss": 1.2193,
+      "num_input_tokens_seen": 214554500,
+      "step": 5470
+    },
+    {
+      "epoch": 0.2655296055819362,
+      "grad_norm": 0.4005719721317291,
+      "learning_rate": 4.307397685372448e-05,
+      "loss": 1.1867,
+      "num_input_tokens_seen": 214941668,
+      "step": 5480
+    },
+    {
+      "epoch": 0.2660141486578157,
+      "grad_norm": 0.4291759431362152,
+      "learning_rate": 4.304684778267199e-05,
+      "loss": 1.2184,
+      "num_input_tokens_seen": 215343812,
+      "step": 5490
+    },
+    {
+      "epoch": 0.2664986917336951,
+      "grad_norm": 0.41158977150917053,
+      "learning_rate": 4.3019674262860324e-05,
+      "loss": 1.1839,
+      "num_input_tokens_seen": 215688820,
+      "step": 5500
+    },
+    {
+      "epoch": 0.2669832348095746,
+      "grad_norm": 0.4709673821926117,
+      "learning_rate": 4.2992456361216934e-05,
+      "loss": 1.2018,
+      "num_input_tokens_seen": 216078056,
+      "step": 5510
+    },
+    {
+      "epoch": 0.267467777885454,
+      "grad_norm": 0.47514909505844116,
+      "learning_rate": 4.2965194144778554e-05,
+      "loss": 1.2325,
+      "num_input_tokens_seen": 216461372,
+      "step": 5520
+    },
+    {
+      "epoch": 0.2679523209613335,
+      "grad_norm": 0.38686904311180115,
+      "learning_rate": 4.293788768069108e-05,
+      "loss": 1.2933,
+      "num_input_tokens_seen": 216842388,
+      "step": 5530
+    },
+    {
+      "epoch": 0.2684368640372129,
+      "grad_norm": 0.3987712860107422,
+      "learning_rate": 4.291053703620939e-05,
+      "loss": 1.2467,
+      "num_input_tokens_seen": 217231240,
+      "step": 5540
+    },
+    {
+      "epoch": 0.26892140711309237,
+      "grad_norm": 0.4047813415527344,
+      "learning_rate": 4.288314227869716e-05,
+      "loss": 1.2023,
+      "num_input_tokens_seen": 217648192,
+      "step": 5550
+    },
+    {
+      "epoch": 0.2694059501889718,
+      "grad_norm": 0.424683541059494,
+      "learning_rate": 4.285570347562674e-05,
+      "loss": 1.2134,
+      "num_input_tokens_seen": 218017588,
+      "step": 5560
+    },
+    {
+      "epoch": 0.26989049326485126,
+      "grad_norm": 0.4166494607925415,
+      "learning_rate": 4.2828220694578935e-05,
+      "loss": 1.1818,
+      "num_input_tokens_seen": 218425896,
+      "step": 5570
+    },
+    {
+      "epoch": 0.2703750363407307,
+      "grad_norm": 0.3912692964076996,
+      "learning_rate": 4.28006940032429e-05,
+      "loss": 1.1785,
+      "num_input_tokens_seen": 218807588,
+      "step": 5580
+    },
+    {
+      "epoch": 0.27085957941661015,
+      "grad_norm": 0.36660903692245483,
+      "learning_rate": 4.2773123469415897e-05,
+      "loss": 1.1937,
+      "num_input_tokens_seen": 219185200,
+      "step": 5590
+    },
+    {
+      "epoch": 0.27134412249248957,
+      "grad_norm": 0.39641043543815613,
+      "learning_rate": 4.274550916100321e-05,
+      "loss": 1.1904,
+      "num_input_tokens_seen": 219546888,
+      "step": 5600
+    },
+    {
+      "epoch": 0.27182866556836904,
+      "grad_norm": 0.371076375246048,
+      "learning_rate": 4.271785114601791e-05,
+      "loss": 1.2116,
+      "num_input_tokens_seen": 219938908,
+      "step": 5610
+    },
+    {
+      "epoch": 0.27231320864424846,
+      "grad_norm": 0.4308777153491974,
+      "learning_rate": 4.269014949258072e-05,
+      "loss": 1.1918,
+      "num_input_tokens_seen": 220337796,
+      "step": 5620
+    },
+    {
+      "epoch": 0.27279775172012793,
+      "grad_norm": 0.4023754894733429,
+      "learning_rate": 4.266240426891987e-05,
+      "loss": 1.1797,
+      "num_input_tokens_seen": 220717632,
+      "step": 5630
+    },
+    {
+      "epoch": 0.27328229479600735,
+      "grad_norm": 0.3996109366416931,
+      "learning_rate": 4.2634615543370885e-05,
+      "loss": 1.1916,
+      "num_input_tokens_seen": 221115656,
+      "step": 5640
+    },
+    {
+      "epoch": 0.2737668378718868,
+      "grad_norm": 0.43148279190063477,
+      "learning_rate": 4.26067833843764e-05,
+      "loss": 1.1981,
+      "num_input_tokens_seen": 221450512,
+      "step": 5650
+    },
+    {
+      "epoch": 0.27425138094776624,
+      "grad_norm": 0.4533494710922241,
+      "learning_rate": 4.257890786048609e-05,
+      "loss": 1.1904,
+      "num_input_tokens_seen": 221855800,
+      "step": 5660
+    },
+    {
+      "epoch": 0.2747359240236457,
+      "grad_norm": 0.38806283473968506,
+      "learning_rate": 4.255098904035638e-05,
+      "loss": 1.2631,
+      "num_input_tokens_seen": 222267776,
+      "step": 5670
+    },
+    {
+      "epoch": 0.27522046709952513,
+      "grad_norm": 0.3938799202442169,
+      "learning_rate": 4.252302699275037e-05,
+      "loss": 1.1635,
+      "num_input_tokens_seen": 222682008,
+      "step": 5680
+    },
+    {
+      "epoch": 0.2757050101754046,
+      "grad_norm": 0.4024967551231384,
+      "learning_rate": 4.249502178653759e-05,
+      "loss": 1.1915,
+      "num_input_tokens_seen": 223053692,
+      "step": 5690
+    },
+    {
+      "epoch": 0.276189553251284,
+      "grad_norm": 0.3679750859737396,
+      "learning_rate": 4.246697349069391e-05,
+      "loss": 1.1711,
+      "num_input_tokens_seen": 223438432,
+      "step": 5700
+    },
+    {
+      "epoch": 0.2766740963271635,
+      "grad_norm": 0.42832136154174805,
+      "learning_rate": 4.243888217430129e-05,
+      "loss": 1.1753,
+      "num_input_tokens_seen": 223831660,
+      "step": 5710
+    },
+    {
+      "epoch": 0.2771586394030429,
+      "grad_norm": 0.4067901074886322,
+      "learning_rate": 4.241074790654769e-05,
+      "loss": 1.2218,
+      "num_input_tokens_seen": 224233376,
+      "step": 5720
+    },
+    {
+      "epoch": 0.2776431824789224,
+      "grad_norm": 0.38691502809524536,
+      "learning_rate": 4.2382570756726815e-05,
+      "loss": 1.2041,
+      "num_input_tokens_seen": 224640328,
+      "step": 5730
+    },
+    {
+      "epoch": 0.2781277255548018,
+      "grad_norm": 0.40320757031440735,
+      "learning_rate": 4.235435079423802e-05,
+      "loss": 1.2297,
+      "num_input_tokens_seen": 225039904,
+      "step": 5740
+    },
+    {
+      "epoch": 0.2786122686306813,
+      "grad_norm": 0.43891477584838867,
+      "learning_rate": 4.232608808858608e-05,
+      "loss": 1.1493,
+      "num_input_tokens_seen": 225441684,
+      "step": 5750
+    },
+    {
+      "epoch": 0.2790968117065607,
+      "grad_norm": 0.4412339925765991,
+      "learning_rate": 4.2297782709381065e-05,
+      "loss": 1.2154,
+      "num_input_tokens_seen": 225837976,
+      "step": 5760
+    },
+    {
+      "epoch": 0.2795813547824402,
+      "grad_norm": 0.39056098461151123,
+      "learning_rate": 4.226943472633813e-05,
+      "loss": 1.1641,
+      "num_input_tokens_seen": 226187316,
+      "step": 5770
+    },
+    {
+      "epoch": 0.2800658978583196,
+      "grad_norm": 0.41356363892555237,
+      "learning_rate": 4.2241044209277384e-05,
+      "loss": 1.2058,
+      "num_input_tokens_seen": 226579696,
+      "step": 5780
+    },
+    {
+      "epoch": 0.28055044093419906,
+      "grad_norm": 0.40139320492744446,
+      "learning_rate": 4.2212611228123686e-05,
+      "loss": 1.2125,
+      "num_input_tokens_seen": 226972904,
+      "step": 5790
+    },
+    {
+      "epoch": 0.2810349840100785,
+      "grad_norm": 0.40096282958984375,
+      "learning_rate": 4.218413585290647e-05,
+      "loss": 1.2027,
+      "num_input_tokens_seen": 227364980,
+      "step": 5800
+    },
+    {
+      "epoch": 0.28151952708595795,
+      "grad_norm": 0.41258716583251953,
+      "learning_rate": 4.21556181537596e-05,
+      "loss": 1.2129,
+      "num_input_tokens_seen": 227750424,
+      "step": 5810
+    },
+    {
+      "epoch": 0.2820040701618374,
+      "grad_norm": 0.4123449921607971,
+      "learning_rate": 4.2127058200921186e-05,
+      "loss": 1.1705,
+      "num_input_tokens_seen": 228150768,
+      "step": 5820
+    },
+    {
+      "epoch": 0.28248861323771685,
+      "grad_norm": 0.3969346582889557,
+      "learning_rate": 4.209845606473339e-05,
+      "loss": 1.1913,
+      "num_input_tokens_seen": 228525776,
+      "step": 5830
+    },
+    {
+      "epoch": 0.28297315631359626,
+      "grad_norm": 0.378460168838501,
+      "learning_rate": 4.206981181564229e-05,
+      "loss": 1.1467,
+      "num_input_tokens_seen": 228912604,
+      "step": 5840
+    },
+    {
+      "epoch": 0.28345769938947574,
+      "grad_norm": 0.38569942116737366,
+      "learning_rate": 4.2041125524197664e-05,
+      "loss": 1.2075,
+      "num_input_tokens_seen": 229298524,
+      "step": 5850
+    },
+    {
+      "epoch": 0.28394224246535515,
+      "grad_norm": 0.3867596983909607,
+      "learning_rate": 4.201239726105286e-05,
+      "loss": 1.1713,
+      "num_input_tokens_seen": 229686964,
+      "step": 5860
+    },
+    {
+      "epoch": 0.28442678554123463,
+      "grad_norm": 0.373037725687027,
+      "learning_rate": 4.198362709696458e-05,
+      "loss": 1.2551,
+      "num_input_tokens_seen": 230060948,
+      "step": 5870
+    },
+    {
+      "epoch": 0.28491132861711405,
+      "grad_norm": 0.3632396161556244,
+      "learning_rate": 4.195481510279276e-05,
+      "loss": 1.2301,
+      "num_input_tokens_seen": 230448628,
+      "step": 5880
+    },
+    {
+      "epoch": 0.2853958716929935,
+      "grad_norm": 0.4397088289260864,
+      "learning_rate": 4.192596134950033e-05,
+      "loss": 1.2441,
+      "num_input_tokens_seen": 230837484,
+      "step": 5890
+    },
+    {
+      "epoch": 0.28588041476887294,
+      "grad_norm": 0.37855035066604614,
+      "learning_rate": 4.189706590815307e-05,
+      "loss": 1.2049,
+      "num_input_tokens_seen": 231247448,
+      "step": 5900
+    },
+    {
+      "epoch": 0.2863649578447524,
+      "grad_norm": 0.3964214324951172,
+      "learning_rate": 4.186812884991946e-05,
+      "loss": 1.1927,
+      "num_input_tokens_seen": 231653776,
+      "step": 5910
+    },
+    {
+      "epoch": 0.28684950092063183,
+      "grad_norm": 0.35895830392837524,
+      "learning_rate": 4.183915024607048e-05,
+      "loss": 1.2302,
+      "num_input_tokens_seen": 232045136,
+      "step": 5920
+    },
+    {
+      "epoch": 0.2873340439965113,
+      "grad_norm": 0.375405877828598,
+      "learning_rate": 4.18101301679794e-05,
+      "loss": 1.1242,
+      "num_input_tokens_seen": 232402564,
+      "step": 5930
+    },
+    {
+      "epoch": 0.2878185870723907,
+      "grad_norm": 0.37960904836654663,
+      "learning_rate": 4.178106868712168e-05,
+      "loss": 1.2209,
+      "num_input_tokens_seen": 232790844,
+      "step": 5940
+    },
+    {
+      "epoch": 0.2883031301482702,
+      "grad_norm": 0.4015379250049591,
+      "learning_rate": 4.1751965875074746e-05,
+      "loss": 1.1929,
+      "num_input_tokens_seen": 233194892,
+      "step": 5950
+    },
+    {
+      "epoch": 0.2887876732241496,
+      "grad_norm": 0.42236605286598206,
+      "learning_rate": 4.172282180351779e-05,
+      "loss": 1.2123,
+      "num_input_tokens_seen": 233566416,
+      "step": 5960
+    },
+    {
+      "epoch": 0.2892722163000291,
+      "grad_norm": 0.4050694406032562,
+      "learning_rate": 4.169363654423166e-05,
+      "loss": 1.1981,
+      "num_input_tokens_seen": 233992752,
+      "step": 5970
+    },
+    {
+      "epoch": 0.2897567593759085,
+      "grad_norm": 0.4495182931423187,
+      "learning_rate": 4.166441016909864e-05,
+      "loss": 1.2207,
+      "num_input_tokens_seen": 234347140,
+      "step": 5980
+    },
+    {
+      "epoch": 0.290241302451788,
+      "grad_norm": 0.3789389431476593,
+      "learning_rate": 4.163514275010228e-05,
+      "loss": 1.1707,
+      "num_input_tokens_seen": 234751484,
+      "step": 5990
+    },
+    {
+      "epoch": 0.2907258455276674,
+      "grad_norm": 0.3984537720680237,
+      "learning_rate": 4.160583435932719e-05,
+      "loss": 1.2024,
+      "num_input_tokens_seen": 235148092,
+      "step": 6000
+    },
+    {
+      "epoch": 0.2907258455276674,
+      "eval_loss": 1.200870156288147,
+      "eval_runtime": 5.1147,
+      "eval_samples_per_second": 29.327,
+      "eval_steps_per_second": 3.715,
+      "num_input_tokens_seen": 235148092,
+      "step": 6000
+    },
+    {
+      "epoch": 0.29121038860354687,
+      "grad_norm": 0.409307062625885,
+      "learning_rate": 4.157648506895895e-05,
+      "loss": 1.2106,
+      "num_input_tokens_seen": 235543492,
+      "step": 6010
+    },
+    {
+      "epoch": 0.2916949316794263,
+      "grad_norm": 0.3580097258090973,
+      "learning_rate": 4.154709495128383e-05,
+      "loss": 1.1765,
+      "num_input_tokens_seen": 235908876,
+      "step": 6020
+    },
+    {
+      "epoch": 0.29217947475530576,
+      "grad_norm": 0.38826310634613037,
+      "learning_rate": 4.151766407868866e-05,
+      "loss": 1.1798,
+      "num_input_tokens_seen": 236298644,
+      "step": 6030
+    },
+    {
+      "epoch": 0.2926640178311852,
+      "grad_norm": 0.3992711901664734,
+      "learning_rate": 4.1488192523660676e-05,
+      "loss": 1.2077,
+      "num_input_tokens_seen": 236681740,
+      "step": 6040
+    },
+    {
+      "epoch": 0.29314856090706465,
+      "grad_norm": 0.4223003089427948,
+      "learning_rate": 4.1458680358787275e-05,
+      "loss": 1.1904,
+      "num_input_tokens_seen": 237098732,
+      "step": 6050
+    },
+    {
+      "epoch": 0.29363310398294407,
+      "grad_norm": 0.3620993494987488,
+      "learning_rate": 4.14291276567559e-05,
+      "loss": 1.2307,
+      "num_input_tokens_seen": 237483400,
+      "step": 6060
+    },
+    {
+      "epoch": 0.29411764705882354,
+      "grad_norm": 0.3823581039905548,
+      "learning_rate": 4.139953449035383e-05,
+      "loss": 1.1695,
+      "num_input_tokens_seen": 237865200,
+      "step": 6070
+    },
+    {
+      "epoch": 0.29460219013470296,
+      "grad_norm": 0.37260711193084717,
+      "learning_rate": 4.136990093246802e-05,
+      "loss": 1.2297,
+      "num_input_tokens_seen": 238233548,
+      "step": 6080
+    },
+    {
+      "epoch": 0.29508673321058243,
+      "grad_norm": 0.4020680785179138,
+      "learning_rate": 4.134022705608487e-05,
+      "loss": 1.2263,
+      "num_input_tokens_seen": 238623180,
+      "step": 6090
+    },
+    {
+      "epoch": 0.29557127628646185,
+      "grad_norm": 0.41028186678886414,
+      "learning_rate": 4.1310512934290124e-05,
+      "loss": 1.2341,
+      "num_input_tokens_seen": 238985848,
+      "step": 6100
+    },
+    {
+      "epoch": 0.2960558193623413,
+      "grad_norm": 0.4138743579387665,
+      "learning_rate": 4.1280758640268625e-05,
+      "loss": 1.2075,
+      "num_input_tokens_seen": 239355272,
+      "step": 6110
+    },
+    {
+      "epoch": 0.29654036243822074,
+      "grad_norm": 0.35869237780570984,
+      "learning_rate": 4.125096424730417e-05,
+      "loss": 1.2002,
+      "num_input_tokens_seen": 239742404,
+      "step": 6120
+    },
+    {
+      "epoch": 0.2970249055141002,
+      "grad_norm": 0.39054998755455017,
+      "learning_rate": 4.12211298287793e-05,
+      "loss": 1.2103,
+      "num_input_tokens_seen": 240150336,
+      "step": 6130
+    },
+    {
+      "epoch": 0.29750944858997963,
+      "grad_norm": 0.42105597257614136,
+      "learning_rate": 4.119125545817517e-05,
+      "loss": 1.2114,
+      "num_input_tokens_seen": 240519488,
+      "step": 6140
+    },
+    {
+      "epoch": 0.2979939916658591,
+      "grad_norm": 0.37524622678756714,
+      "learning_rate": 4.11613412090713e-05,
+      "loss": 1.2072,
+      "num_input_tokens_seen": 240898472,
+      "step": 6150
+    },
+    {
+      "epoch": 0.2984785347417385,
+      "grad_norm": 0.3830278217792511,
+      "learning_rate": 4.113138715514546e-05,
+      "loss": 1.2617,
+      "num_input_tokens_seen": 241258828,
+      "step": 6160
+    },
+    {
+      "epoch": 0.298963077817618,
+      "grad_norm": 0.3757288157939911,
+      "learning_rate": 4.110139337017345e-05,
+      "loss": 1.201,
+      "num_input_tokens_seen": 241648504,
+      "step": 6170
+    },
+    {
+      "epoch": 0.2994476208934974,
+      "grad_norm": 0.500215470790863,
+      "learning_rate": 4.1071359928028896e-05,
+      "loss": 1.218,
+      "num_input_tokens_seen": 242026552,
+      "step": 6180
+    },
+    {
+      "epoch": 0.2999321639693769,
+      "grad_norm": 0.3983130156993866,
+      "learning_rate": 4.104128690268314e-05,
+      "loss": 1.1784,
+      "num_input_tokens_seen": 242415488,
+      "step": 6190
+    },
+    {
+      "epoch": 0.3004167070452563,
+      "grad_norm": 0.4273653030395508,
+      "learning_rate": 4.101117436820499e-05,
+      "loss": 1.2151,
+      "num_input_tokens_seen": 242796596,
+      "step": 6200
+    },
+    {
+      "epoch": 0.3009012501211358,
+      "grad_norm": 0.4290236830711365,
+      "learning_rate": 4.098102239876058e-05,
+      "loss": 1.1989,
+      "num_input_tokens_seen": 243188916,
+      "step": 6210
+    },
+    {
+      "epoch": 0.3013857931970152,
+      "grad_norm": 0.4325619339942932,
+      "learning_rate": 4.095083106861317e-05,
+      "loss": 1.1811,
+      "num_input_tokens_seen": 243612644,
+      "step": 6220
+    },
+    {
+      "epoch": 0.30187033627289467,
+      "grad_norm": 0.38956135511398315,
+      "learning_rate": 4.0920600452122934e-05,
+      "loss": 1.2224,
+      "num_input_tokens_seen": 244023064,
+      "step": 6230
+    },
+    {
+      "epoch": 0.3023548793487741,
+      "grad_norm": 0.3935461938381195,
+      "learning_rate": 4.0890330623746856e-05,
+      "loss": 1.175,
+      "num_input_tokens_seen": 244414972,
+      "step": 6240
+    },
+    {
+      "epoch": 0.30283942242465356,
+      "grad_norm": 0.3803289830684662,
+      "learning_rate": 4.086002165803845e-05,
+      "loss": 1.1878,
+      "num_input_tokens_seen": 244793504,
+      "step": 6250
+    },
+    {
+      "epoch": 0.303323965500533,
+      "grad_norm": 0.3836755156517029,
+      "learning_rate": 4.082967362964766e-05,
+      "loss": 1.2004,
+      "num_input_tokens_seen": 245195808,
+      "step": 6260
+    },
+    {
+      "epoch": 0.30380850857641245,
+      "grad_norm": 0.36543598771095276,
+      "learning_rate": 4.079928661332062e-05,
+      "loss": 1.2333,
+      "num_input_tokens_seen": 245600428,
+      "step": 6270
+    },
+    {
+      "epoch": 0.30429305165229187,
+      "grad_norm": 0.46092289686203003,
+      "learning_rate": 4.07688606838995e-05,
+      "loss": 1.1549,
+      "num_input_tokens_seen": 246013016,
+      "step": 6280
+    },
+    {
+      "epoch": 0.30477759472817134,
+      "grad_norm": 0.364113986492157,
+      "learning_rate": 4.0738395916322305e-05,
+      "loss": 1.213,
+      "num_input_tokens_seen": 246396572,
+      "step": 6290
+    },
+    {
+      "epoch": 0.30526213780405076,
+      "grad_norm": 0.4042966067790985,
+      "learning_rate": 4.0707892385622695e-05,
+      "loss": 1.1444,
+      "num_input_tokens_seen": 246823552,
+      "step": 6300
+    },
+    {
+      "epoch": 0.30574668087993023,
+      "grad_norm": 0.3558143079280853,
+      "learning_rate": 4.0677350166929805e-05,
+      "loss": 1.2313,
+      "num_input_tokens_seen": 247219524,
+      "step": 6310
+    },
+    {
+      "epoch": 0.30623122395580965,
+      "grad_norm": 0.4053792655467987,
+      "learning_rate": 4.064676933546807e-05,
+      "loss": 1.2071,
+      "num_input_tokens_seen": 247632420,
+      "step": 6320
+    },
+    {
+      "epoch": 0.3067157670316891,
+      "grad_norm": 0.37768182158470154,
+      "learning_rate": 4.061614996655701e-05,
+      "loss": 1.2165,
+      "num_input_tokens_seen": 248035440,
+      "step": 6330
+    },
+    {
+      "epoch": 0.30720031010756854,
+      "grad_norm": 0.38902053236961365,
+      "learning_rate": 4.0585492135611064e-05,
+      "loss": 1.2016,
+      "num_input_tokens_seen": 248418968,
+      "step": 6340
+    },
+    {
+      "epoch": 0.307684853183448,
+      "grad_norm": 0.4327777028083801,
+      "learning_rate": 4.055479591813941e-05,
+      "loss": 1.248,
+      "num_input_tokens_seen": 248804400,
+      "step": 6350
+    },
+    {
+      "epoch": 0.30816939625932743,
+      "grad_norm": 0.38908207416534424,
+      "learning_rate": 4.052406138974576e-05,
+      "loss": 1.2374,
+      "num_input_tokens_seen": 249179376,
+      "step": 6360
+    },
+    {
+      "epoch": 0.3086539393352069,
+      "grad_norm": 0.45117059350013733,
+      "learning_rate": 4.049328862612819e-05,
+      "loss": 1.1853,
+      "num_input_tokens_seen": 249591320,
+      "step": 6370
+    },
+    {
+      "epoch": 0.3091384824110863,
+      "grad_norm": 0.3981532156467438,
+      "learning_rate": 4.0462477703078944e-05,
+      "loss": 1.1586,
+      "num_input_tokens_seen": 249973744,
+      "step": 6380
+    },
+    {
+      "epoch": 0.3096230254869658,
+      "grad_norm": 0.39917775988578796,
+      "learning_rate": 4.043162869648427e-05,
+      "loss": 1.2109,
+      "num_input_tokens_seen": 250419824,
+      "step": 6390
+    },
+    {
+      "epoch": 0.3101075685628452,
+      "grad_norm": 0.45799973607063293,
+      "learning_rate": 4.0400741682324185e-05,
+      "loss": 1.2142,
+      "num_input_tokens_seen": 250837544,
+      "step": 6400
+    },
+    {
+      "epoch": 0.3105921116387247,
+      "grad_norm": 0.4319687783718109,
+      "learning_rate": 4.036981673667234e-05,
+      "loss": 1.2561,
+      "num_input_tokens_seen": 251224108,
+      "step": 6410
+    },
+    {
+      "epoch": 0.3110766547146041,
+      "grad_norm": 0.3861522376537323,
+      "learning_rate": 4.0338853935695795e-05,
+      "loss": 1.1854,
+      "num_input_tokens_seen": 251600636,
+      "step": 6420
+    },
+    {
+      "epoch": 0.3115611977904836,
+      "grad_norm": 0.4382316768169403,
+      "learning_rate": 4.0307853355654856e-05,
+      "loss": 1.1935,
+      "num_input_tokens_seen": 251988072,
+      "step": 6430
+    },
+    {
+      "epoch": 0.312045740866363,
+      "grad_norm": 0.41492384672164917,
+      "learning_rate": 4.027681507290288e-05,
+      "loss": 1.2185,
+      "num_input_tokens_seen": 252366656,
+      "step": 6440
+    },
+    {
+      "epoch": 0.31253028394224247,
+      "grad_norm": 0.4099133014678955,
+      "learning_rate": 4.0245739163886076e-05,
+      "loss": 1.2092,
+      "num_input_tokens_seen": 252784056,
+      "step": 6450
+    },
+    {
+      "epoch": 0.3130148270181219,
+      "grad_norm": 0.41487085819244385,
+      "learning_rate": 4.021462570514333e-05,
+      "loss": 1.1743,
+      "num_input_tokens_seen": 253186828,
+      "step": 6460
+    },
+    {
+      "epoch": 0.31349937009400136,
+      "grad_norm": 0.370210736989975,
+      "learning_rate": 4.0183474773305995e-05,
+      "loss": 1.2137,
+      "num_input_tokens_seen": 253574680,
+      "step": 6470
+    },
+    {
+      "epoch": 0.3139839131698808,
+      "grad_norm": 0.3662930727005005,
+      "learning_rate": 4.015228644509776e-05,
+      "loss": 1.1403,
+      "num_input_tokens_seen": 253967160,
+      "step": 6480
+    },
+    {
+      "epoch": 0.31446845624576025,
+      "grad_norm": 0.41104060411453247,
+      "learning_rate": 4.0121060797334375e-05,
+      "loss": 1.2502,
+      "num_input_tokens_seen": 254357996,
+      "step": 6490
+    },
+    {
+      "epoch": 0.31495299932163967,
+      "grad_norm": 0.3852168023586273,
+      "learning_rate": 4.008979790692353e-05,
+      "loss": 1.1896,
+      "num_input_tokens_seen": 254758512,
+      "step": 6500
+    },
+    {
+      "epoch": 0.31543754239751914,
+      "grad_norm": 0.40157365798950195,
+      "learning_rate": 4.005849785086464e-05,
+      "loss": 1.2011,
+      "num_input_tokens_seen": 255148092,
+      "step": 6510
+    },
+    {
+      "epoch": 0.31592208547339856,
+      "grad_norm": 0.41574087738990784,
+      "learning_rate": 4.002716070624866e-05,
+      "loss": 1.2243,
+      "num_input_tokens_seen": 255552832,
+      "step": 6520
+    },
+    {
+      "epoch": 0.31640662854927804,
+      "grad_norm": 0.3938582241535187,
+      "learning_rate": 3.999578655025787e-05,
+      "loss": 1.1956,
+      "num_input_tokens_seen": 255953292,
+      "step": 6530
+    },
+    {
+      "epoch": 0.31689117162515745,
+      "grad_norm": 0.39169254899024963,
+      "learning_rate": 3.996437546016575e-05,
+      "loss": 1.2158,
+      "num_input_tokens_seen": 256352132,
+      "step": 6540
+    },
+    {
+      "epoch": 0.3173757147010369,
+      "grad_norm": 0.38570165634155273,
+      "learning_rate": 3.993292751333671e-05,
+      "loss": 1.2131,
+      "num_input_tokens_seen": 256744884,
+      "step": 6550
+    },
+    {
+      "epoch": 0.31786025777691634,
+      "grad_norm": 0.38711732625961304,
+      "learning_rate": 3.9901442787225955e-05,
+      "loss": 1.1811,
+      "num_input_tokens_seen": 257140072,
+      "step": 6560
+    },
+    {
+      "epoch": 0.3183448008527958,
+      "grad_norm": 0.4001631438732147,
+      "learning_rate": 3.986992135937927e-05,
+      "loss": 1.2469,
+      "num_input_tokens_seen": 257526996,
+      "step": 6570
+    },
+    {
+      "epoch": 0.31882934392867524,
+      "grad_norm": 0.4085213243961334,
+      "learning_rate": 3.9838363307432844e-05,
+      "loss": 1.173,
+      "num_input_tokens_seen": 257934640,
+      "step": 6580
+    },
+    {
+      "epoch": 0.3193138870045547,
+      "grad_norm": 0.3954554498195648,
+      "learning_rate": 3.980676870911305e-05,
+      "loss": 1.1913,
+      "num_input_tokens_seen": 258291948,
+      "step": 6590
+    },
+    {
+      "epoch": 0.3197984300804341,
+      "grad_norm": 0.3921205401420593,
+      "learning_rate": 3.977513764223629e-05,
+      "loss": 1.1919,
+      "num_input_tokens_seen": 258703912,
+      "step": 6600
+    },
+    {
+      "epoch": 0.3202829731563136,
+      "grad_norm": 0.39368802309036255,
+      "learning_rate": 3.974347018470879e-05,
+      "loss": 1.1346,
+      "num_input_tokens_seen": 259074912,
+      "step": 6610
+    },
+    {
+      "epoch": 0.320767516232193,
+      "grad_norm": 0.4187362790107727,
+      "learning_rate": 3.9711766414526386e-05,
+      "loss": 1.2207,
+      "num_input_tokens_seen": 259450412,
+      "step": 6620
+    },
+    {
+      "epoch": 0.3212520593080725,
+      "grad_norm": 0.4282675087451935,
+      "learning_rate": 3.9680026409774385e-05,
+      "loss": 1.1791,
+      "num_input_tokens_seen": 259839588,
+      "step": 6630
+    },
+    {
+      "epoch": 0.3217366023839519,
+      "grad_norm": 0.41149836778640747,
+      "learning_rate": 3.9648250248627285e-05,
+      "loss": 1.1941,
+      "num_input_tokens_seen": 260210136,
+      "step": 6640
+    },
+    {
+      "epoch": 0.3222211454598314,
+      "grad_norm": 0.40860268473625183,
+      "learning_rate": 3.961643800934869e-05,
+      "loss": 1.1643,
+      "num_input_tokens_seen": 260596296,
+      "step": 6650
+    },
+    {
+      "epoch": 0.3227056885357108,
+      "grad_norm": 0.40710315108299255,
+      "learning_rate": 3.958458977029103e-05,
+      "loss": 1.1714,
+      "num_input_tokens_seen": 260991860,
+      "step": 6660
+    },
+    {
+      "epoch": 0.3231902316115903,
+      "grad_norm": 0.36702919006347656,
+      "learning_rate": 3.955270560989542e-05,
+      "loss": 1.2253,
+      "num_input_tokens_seen": 261370536,
+      "step": 6670
+    },
+    {
+      "epoch": 0.3236747746874697,
+      "grad_norm": 0.36147916316986084,
+      "learning_rate": 3.952078560669142e-05,
+      "loss": 1.1437,
+      "num_input_tokens_seen": 261745012,
+      "step": 6680
+    },
+    {
+      "epoch": 0.32415931776334916,
+      "grad_norm": 0.38143739104270935,
+      "learning_rate": 3.9488829839296914e-05,
+      "loss": 1.2353,
+      "num_input_tokens_seen": 262122240,
+      "step": 6690
+    },
+    {
+      "epoch": 0.3246438608392286,
+      "grad_norm": 0.4425183832645416,
+      "learning_rate": 3.9456838386417835e-05,
+      "loss": 1.1905,
+      "num_input_tokens_seen": 262497788,
+      "step": 6700
+    },
+    {
+      "epoch": 0.32512840391510806,
+      "grad_norm": 0.4275406301021576,
+      "learning_rate": 3.9424811326848024e-05,
+      "loss": 1.2369,
+      "num_input_tokens_seen": 262888292,
+      "step": 6710
+    },
+    {
+      "epoch": 0.3256129469909875,
+      "grad_norm": 0.4195478558540344,
+      "learning_rate": 3.939274873946901e-05,
+      "loss": 1.2022,
+      "num_input_tokens_seen": 263277288,
+      "step": 6720
+    },
+    {
+      "epoch": 0.32609749006686695,
+      "grad_norm": 0.38973620533943176,
+      "learning_rate": 3.936065070324984e-05,
+      "loss": 1.1681,
+      "num_input_tokens_seen": 263661148,
+      "step": 6730
+    },
+    {
+      "epoch": 0.32658203314274636,
+      "grad_norm": 0.40185558795928955,
+      "learning_rate": 3.932851729724685e-05,
+      "loss": 1.2155,
+      "num_input_tokens_seen": 264039708,
+      "step": 6740
+    },
+    {
+      "epoch": 0.32706657621862584,
+      "grad_norm": 0.3969719409942627,
+      "learning_rate": 3.929634860060351e-05,
+      "loss": 1.2022,
+      "num_input_tokens_seen": 264431972,
+      "step": 6750
+    },
+    {
+      "epoch": 0.32755111929450526,
+      "grad_norm": 0.4183984696865082,
+      "learning_rate": 3.92641446925502e-05,
+      "loss": 1.2105,
+      "num_input_tokens_seen": 264799512,
+      "step": 6760
+    },
+    {
+      "epoch": 0.32803566237038473,
+      "grad_norm": 0.42662981152534485,
+      "learning_rate": 3.9231905652404034e-05,
+      "loss": 1.218,
+      "num_input_tokens_seen": 265163968,
+      "step": 6770
+    },
+    {
+      "epoch": 0.32852020544626415,
+      "grad_norm": 0.4058518707752228,
+      "learning_rate": 3.919963155956864e-05,
+      "loss": 1.1482,
+      "num_input_tokens_seen": 265559800,
+      "step": 6780
+    },
+    {
+      "epoch": 0.3290047485221436,
+      "grad_norm": 0.3763887584209442,
+      "learning_rate": 3.916732249353399e-05,
+      "loss": 1.1912,
+      "num_input_tokens_seen": 265950740,
+      "step": 6790
+    },
+    {
+      "epoch": 0.32948929159802304,
+      "grad_norm": 0.36359450221061707,
+      "learning_rate": 3.91349785338762e-05,
+      "loss": 1.178,
+      "num_input_tokens_seen": 266339700,
+      "step": 6800
+    },
+    {
+      "epoch": 0.3299738346739025,
+      "grad_norm": 0.3989261984825134,
+      "learning_rate": 3.91025997602573e-05,
+      "loss": 1.23,
+      "num_input_tokens_seen": 266740236,
+      "step": 6810
+    },
+    {
+      "epoch": 0.33045837774978193,
+      "grad_norm": 0.37006011605262756,
+      "learning_rate": 3.90701862524251e-05,
+      "loss": 1.1739,
+      "num_input_tokens_seen": 267116000,
+      "step": 6820
+    },
+    {
+      "epoch": 0.3309429208256614,
+      "grad_norm": 0.3907497525215149,
+      "learning_rate": 3.903773809021294e-05,
+      "loss": 1.1801,
+      "num_input_tokens_seen": 267526384,
+      "step": 6830
+    },
+    {
+      "epoch": 0.3314274639015408,
+      "grad_norm": 0.3850066363811493,
+      "learning_rate": 3.900525535353952e-05,
+      "loss": 1.2116,
+      "num_input_tokens_seen": 267893528,
+      "step": 6840
+    },
+    {
+      "epoch": 0.3319120069774203,
+      "grad_norm": 0.4400016665458679,
+      "learning_rate": 3.897273812240868e-05,
+      "loss": 1.1532,
+      "num_input_tokens_seen": 268282380,
+      "step": 6850
+    },
+    {
+      "epoch": 0.3323965500532997,
+      "grad_norm": 0.3924698233604431,
+      "learning_rate": 3.894018647690924e-05,
+      "loss": 1.2103,
+      "num_input_tokens_seen": 268696688,
+      "step": 6860
+    },
+    {
+      "epoch": 0.3328810931291792,
+      "grad_norm": 0.3710118234157562,
+      "learning_rate": 3.890760049721477e-05,
+      "loss": 1.1848,
+      "num_input_tokens_seen": 269125764,
+      "step": 6870
+    },
+    {
+      "epoch": 0.33336563620505866,
+      "grad_norm": 0.37334132194519043,
+      "learning_rate": 3.887498026358341e-05,
+      "loss": 1.2415,
+      "num_input_tokens_seen": 269507904,
+      "step": 6880
+    },
+    {
+      "epoch": 0.3338501792809381,
+      "grad_norm": 0.46161389350891113,
+      "learning_rate": 3.8842325856357656e-05,
+      "loss": 1.2383,
+      "num_input_tokens_seen": 269890536,
+      "step": 6890
+    },
+    {
+      "epoch": 0.33433472235681755,
+      "grad_norm": 0.43443042039871216,
+      "learning_rate": 3.8809637355964176e-05,
+      "loss": 1.1867,
+      "num_input_tokens_seen": 270291048,
+      "step": 6900
+    },
+    {
+      "epoch": 0.33481926543269697,
+      "grad_norm": 0.39960533380508423,
+      "learning_rate": 3.8776914842913626e-05,
+      "loss": 1.2211,
+      "num_input_tokens_seen": 270695776,
+      "step": 6910
+    },
+    {
+      "epoch": 0.33530380850857644,
+      "grad_norm": 0.4040696322917938,
+      "learning_rate": 3.8744158397800404e-05,
+      "loss": 1.2013,
+      "num_input_tokens_seen": 271104472,
+      "step": 6920
+    },
+    {
+      "epoch": 0.33578835158445586,
+      "grad_norm": 0.4039623737335205,
+      "learning_rate": 3.87113681013025e-05,
+      "loss": 1.1545,
+      "num_input_tokens_seen": 271507036,
+      "step": 6930
+    },
+    {
+      "epoch": 0.33627289466033533,
+      "grad_norm": 0.40450507402420044,
+      "learning_rate": 3.867854403418128e-05,
+      "loss": 1.1504,
+      "num_input_tokens_seen": 271911112,
+      "step": 6940
+    },
+    {
+      "epoch": 0.33675743773621475,
+      "grad_norm": 0.4363791346549988,
+      "learning_rate": 3.864568627728128e-05,
+      "loss": 1.1979,
+      "num_input_tokens_seen": 272282304,
+      "step": 6950
+    },
+    {
+      "epoch": 0.3372419808120942,
+      "grad_norm": 0.38292601704597473,
+      "learning_rate": 3.861279491153e-05,
+      "loss": 1.23,
+      "num_input_tokens_seen": 272673128,
+      "step": 6960
+    },
+    {
+      "epoch": 0.33772652388797364,
+      "grad_norm": 0.39683130383491516,
+      "learning_rate": 3.857987001793775e-05,
+      "loss": 1.1738,
+      "num_input_tokens_seen": 273079120,
+      "step": 6970
+    },
+    {
+      "epoch": 0.3382110669638531,
+      "grad_norm": 0.4041154682636261,
+      "learning_rate": 3.8546911677597395e-05,
+      "loss": 1.1948,
+      "num_input_tokens_seen": 273482888,
+      "step": 6980
+    },
+    {
+      "epoch": 0.33869561003973253,
+      "grad_norm": 0.3793328106403351,
+      "learning_rate": 3.8513919971684175e-05,
+      "loss": 1.1841,
+      "num_input_tokens_seen": 273889924,
+      "step": 6990
+    },
+    {
+      "epoch": 0.339180153115612,
+      "grad_norm": 0.3985520005226135,
+      "learning_rate": 3.848089498145552e-05,
+      "loss": 1.1821,
+      "num_input_tokens_seen": 274289204,
+      "step": 7000
+    },
+    {
+      "epoch": 0.3396646961914914,
+      "grad_norm": 0.42367762327194214,
+      "learning_rate": 3.844783678825083e-05,
+      "loss": 1.2186,
+      "num_input_tokens_seen": 274669384,
+      "step": 7010
+    },
+    {
+      "epoch": 0.3401492392673709,
+      "grad_norm": 0.38627681136131287,
+      "learning_rate": 3.8414745473491295e-05,
+      "loss": 1.1691,
+      "num_input_tokens_seen": 275045420,
+      "step": 7020
+    },
+    {
+      "epoch": 0.3406337823432503,
+      "grad_norm": 0.40813496708869934,
+      "learning_rate": 3.838162111867967e-05,
+      "loss": 1.1912,
+      "num_input_tokens_seen": 275452320,
+      "step": 7030
+    },
+    {
+      "epoch": 0.3411183254191298,
+      "grad_norm": 0.3678280711174011,
+      "learning_rate": 3.834846380540009e-05,
+      "loss": 1.2274,
+      "num_input_tokens_seen": 275844788,
+      "step": 7040
+    },
+    {
+      "epoch": 0.3416028684950092,
+      "grad_norm": 0.3878428637981415,
+      "learning_rate": 3.831527361531786e-05,
+      "loss": 1.2127,
+      "num_input_tokens_seen": 276211512,
+      "step": 7050
+    },
+    {
+      "epoch": 0.3420874115708887,
+      "grad_norm": 0.4239497482776642,
+      "learning_rate": 3.828205063017927e-05,
+      "loss": 1.2089,
+      "num_input_tokens_seen": 276596564,
+      "step": 7060
+    },
+    {
+      "epoch": 0.3425719546467681,
+      "grad_norm": 0.3910852074623108,
+      "learning_rate": 3.824879493181138e-05,
+      "loss": 1.153,
+      "num_input_tokens_seen": 276989732,
+      "step": 7070
+    },
+    {
+      "epoch": 0.34305649772264757,
+      "grad_norm": 0.5242650508880615,
+      "learning_rate": 3.821550660212182e-05,
+      "loss": 1.1671,
+      "num_input_tokens_seen": 277396804,
+      "step": 7080
+    },
+    {
+      "epoch": 0.343541040798527,
+      "grad_norm": 0.3888712525367737,
+      "learning_rate": 3.8182185723098584e-05,
+      "loss": 1.1751,
+      "num_input_tokens_seen": 277776744,
+      "step": 7090
+    },
+    {
+      "epoch": 0.34402558387440646,
+      "grad_norm": 0.40083321928977966,
+      "learning_rate": 3.814883237680984e-05,
+      "loss": 1.262,
+      "num_input_tokens_seen": 278174104,
+      "step": 7100
+    },
+    {
+      "epoch": 0.3445101269502859,
+      "grad_norm": 0.4279809594154358,
+      "learning_rate": 3.811544664540373e-05,
+      "loss": 1.1803,
+      "num_input_tokens_seen": 278545544,
+      "step": 7110
+    },
+    {
+      "epoch": 0.34499467002616535,
+      "grad_norm": 0.39681699872016907,
+      "learning_rate": 3.8082028611108144e-05,
+      "loss": 1.1754,
+      "num_input_tokens_seen": 278922000,
+      "step": 7120
+    },
+    {
+      "epoch": 0.34547921310204477,
+      "grad_norm": 0.4334243834018707,
+      "learning_rate": 3.804857835623054e-05,
+      "loss": 1.1708,
+      "num_input_tokens_seen": 279321312,
+      "step": 7130
+    },
+    {
+      "epoch": 0.34596375617792424,
+      "grad_norm": 0.3777294158935547,
+      "learning_rate": 3.801509596315773e-05,
+      "loss": 1.1567,
+      "num_input_tokens_seen": 279699076,
+      "step": 7140
+    },
+    {
+      "epoch": 0.34644829925380366,
+      "grad_norm": 0.4178134500980377,
+      "learning_rate": 3.798158151435569e-05,
+      "loss": 1.1974,
+      "num_input_tokens_seen": 280103592,
+      "step": 7150
+    },
+    {
+      "epoch": 0.34693284232968313,
+      "grad_norm": 0.3830433189868927,
+      "learning_rate": 3.794803509236935e-05,
+      "loss": 1.1979,
+      "num_input_tokens_seen": 280490292,
+      "step": 7160
+    },
+    {
+      "epoch": 0.34741738540556255,
+      "grad_norm": 0.4182297885417938,
+      "learning_rate": 3.791445677982237e-05,
+      "loss": 1.1493,
+      "num_input_tokens_seen": 280873500,
+      "step": 7170
+    },
+    {
+      "epoch": 0.347901928481442,
+      "grad_norm": 0.3943808674812317,
+      "learning_rate": 3.788084665941697e-05,
+      "loss": 1.2254,
+      "num_input_tokens_seen": 281272964,
+      "step": 7180
+    },
+    {
+      "epoch": 0.34838647155732144,
+      "grad_norm": 0.3648427128791809,
+      "learning_rate": 3.7847204813933715e-05,
+      "loss": 1.1815,
+      "num_input_tokens_seen": 281658580,
+      "step": 7190
+    },
+    {
+      "epoch": 0.3488710146332009,
+      "grad_norm": 0.4216112494468689,
+      "learning_rate": 3.78135313262313e-05,
+      "loss": 1.2728,
+      "num_input_tokens_seen": 282047632,
+      "step": 7200
+    },
+    {
+      "epoch": 0.34935555770908033,
+      "grad_norm": 0.4069361686706543,
+      "learning_rate": 3.7779826279246344e-05,
+      "loss": 1.2193,
+      "num_input_tokens_seen": 282442860,
+      "step": 7210
+    },
+    {
+      "epoch": 0.3498401007849598,
+      "grad_norm": 0.3994755148887634,
+      "learning_rate": 3.774608975599323e-05,
+      "loss": 1.1925,
+      "num_input_tokens_seen": 282829096,
+      "step": 7220
+    },
+    {
+      "epoch": 0.3503246438608392,
+      "grad_norm": 0.40187373757362366,
+      "learning_rate": 3.7712321839563816e-05,
+      "loss": 1.1644,
+      "num_input_tokens_seen": 283214644,
+      "step": 7230
+    },
+    {
+      "epoch": 0.3508091869367187,
+      "grad_norm": 0.40469735860824585,
+      "learning_rate": 3.7678522613127315e-05,
+      "loss": 1.1836,
+      "num_input_tokens_seen": 283597500,
+      "step": 7240
+    },
+    {
+      "epoch": 0.3512937300125981,
+      "grad_norm": 0.3870331943035126,
+      "learning_rate": 3.764469215993006e-05,
+      "loss": 1.1341,
+      "num_input_tokens_seen": 283969576,
+      "step": 7250
+    },
+    {
+      "epoch": 0.3517782730884776,
+      "grad_norm": 0.38946250081062317,
+      "learning_rate": 3.761083056329527e-05,
+      "loss": 1.2089,
+      "num_input_tokens_seen": 284368172,
+      "step": 7260
+    },
+    {
+      "epoch": 0.352262816164357,
+      "grad_norm": 0.41155683994293213,
+      "learning_rate": 3.757693790662289e-05,
+      "loss": 1.2169,
+      "num_input_tokens_seen": 284753172,
+      "step": 7270
+    },
+    {
+      "epoch": 0.3527473592402365,
+      "grad_norm": 0.3682643175125122,
+      "learning_rate": 3.754301427338935e-05,
+      "loss": 1.2065,
+      "num_input_tokens_seen": 285154396,
+      "step": 7280
+    },
+    {
+      "epoch": 0.3532319023161159,
+      "grad_norm": 0.384960800409317,
+      "learning_rate": 3.750905974714739e-05,
+      "loss": 1.194,
+      "num_input_tokens_seen": 285556552,
+      "step": 7290
+    },
+    {
+      "epoch": 0.35371644539199537,
+      "grad_norm": 0.39383670687675476,
+      "learning_rate": 3.747507441152581e-05,
+      "loss": 1.1667,
+      "num_input_tokens_seen": 285937684,
+      "step": 7300
+    },
+    {
+      "epoch": 0.3542009884678748,
+      "grad_norm": 0.4078581631183624,
+      "learning_rate": 3.744105835022932e-05,
+      "loss": 1.168,
+      "num_input_tokens_seen": 286339704,
+      "step": 7310
+    },
+    {
+      "epoch": 0.35468553154375426,
+      "grad_norm": 0.4079924523830414,
+      "learning_rate": 3.740701164703831e-05,
+      "loss": 1.2325,
+      "num_input_tokens_seen": 286724008,
+      "step": 7320
+    },
+    {
+      "epoch": 0.3551700746196337,
+      "grad_norm": 0.40963229537010193,
+      "learning_rate": 3.737293438580861e-05,
+      "loss": 1.217,
+      "num_input_tokens_seen": 287109268,
+      "step": 7330
+    },
+    {
+      "epoch": 0.35565461769551315,
+      "grad_norm": 0.40729522705078125,
+      "learning_rate": 3.7338826650471335e-05,
+      "loss": 1.1671,
+      "num_input_tokens_seen": 287487492,
+      "step": 7340
+    },
+    {
+      "epoch": 0.35613916077139257,
+      "grad_norm": 0.4351021647453308,
+      "learning_rate": 3.730468852503265e-05,
+      "loss": 1.1748,
+      "num_input_tokens_seen": 287878812,
+      "step": 7350
+    },
+    {
+      "epoch": 0.35662370384727204,
+      "grad_norm": 0.3914582431316376,
+      "learning_rate": 3.727052009357358e-05,
+      "loss": 1.1588,
+      "num_input_tokens_seen": 288271328,
+      "step": 7360
+    },
+    {
+      "epoch": 0.35710824692315146,
+      "grad_norm": 0.4153572618961334,
+      "learning_rate": 3.723632144024979e-05,
+      "loss": 1.2123,
+      "num_input_tokens_seen": 288659084,
+      "step": 7370
+    },
+    {
+      "epoch": 0.35759278999903094,
+      "grad_norm": 0.38702714443206787,
+      "learning_rate": 3.7202092649291356e-05,
+      "loss": 1.1788,
+      "num_input_tokens_seen": 289060932,
+      "step": 7380
+    },
+    {
+      "epoch": 0.35807733307491035,
+      "grad_norm": 0.40125608444213867,
+      "learning_rate": 3.7167833805002614e-05,
+      "loss": 1.1521,
+      "num_input_tokens_seen": 289437452,
+      "step": 7390
+    },
+    {
+      "epoch": 0.3585618761507898,
+      "grad_norm": 0.43882331252098083,
+      "learning_rate": 3.7133544991761896e-05,
+      "loss": 1.1843,
+      "num_input_tokens_seen": 289845972,
+      "step": 7400
+    },
+    {
+      "epoch": 0.35904641922666924,
+      "grad_norm": 0.37727952003479004,
+      "learning_rate": 3.7099226294021375e-05,
+      "loss": 1.2051,
+      "num_input_tokens_seen": 290209868,
+      "step": 7410
+    },
+    {
+      "epoch": 0.3595309623025487,
+      "grad_norm": 0.3910406231880188,
+      "learning_rate": 3.706487779630679e-05,
+      "loss": 1.1358,
+      "num_input_tokens_seen": 290641272,
+      "step": 7420
+    },
+    {
+      "epoch": 0.36001550537842814,
+      "grad_norm": 0.3660062551498413,
+      "learning_rate": 3.703049958321733e-05,
+      "loss": 1.1969,
+      "num_input_tokens_seen": 291057612,
+      "step": 7430
+    },
+    {
+      "epoch": 0.3605000484543076,
+      "grad_norm": 0.3746223449707031,
+      "learning_rate": 3.69960917394253e-05,
+      "loss": 1.1809,
+      "num_input_tokens_seen": 291459584,
+      "step": 7440
+    },
+    {
+      "epoch": 0.360984591530187,
+      "grad_norm": 0.3975430428981781,
+      "learning_rate": 3.696165434967605e-05,
+      "loss": 1.2374,
+      "num_input_tokens_seen": 291868768,
+      "step": 7450
+    },
+    {
+      "epoch": 0.3614691346060665,
+      "grad_norm": 0.40742921829223633,
+      "learning_rate": 3.692718749878767e-05,
+      "loss": 1.2081,
+      "num_input_tokens_seen": 292249240,
+      "step": 7460
+    },
+    {
+      "epoch": 0.3619536776819459,
+      "grad_norm": 0.44806450605392456,
+      "learning_rate": 3.6892691271650814e-05,
+      "loss": 1.2251,
+      "num_input_tokens_seen": 292658816,
+      "step": 7470
+    },
+    {
+      "epoch": 0.3624382207578254,
+      "grad_norm": 0.37986406683921814,
+      "learning_rate": 3.68581657532285e-05,
+      "loss": 1.1863,
+      "num_input_tokens_seen": 293031712,
+      "step": 7480
+    },
+    {
+      "epoch": 0.3629227638337048,
+      "grad_norm": 0.4018678069114685,
+      "learning_rate": 3.682361102855586e-05,
+      "loss": 1.1884,
+      "num_input_tokens_seen": 293444948,
+      "step": 7490
+    },
+    {
+      "epoch": 0.3634073069095843,
+      "grad_norm": 0.3642045855522156,
+      "learning_rate": 3.6789027182739996e-05,
+      "loss": 1.1818,
+      "num_input_tokens_seen": 293852108,
+      "step": 7500
+    },
+    {
+      "epoch": 0.3638918499854637,
+      "grad_norm": 0.3775769770145416,
+      "learning_rate": 3.675441430095972e-05,
+      "loss": 1.2457,
+      "num_input_tokens_seen": 294219892,
+      "step": 7510
+    },
+    {
+      "epoch": 0.3643763930613432,
+      "grad_norm": 0.37454187870025635,
+      "learning_rate": 3.6719772468465345e-05,
+      "loss": 1.1447,
+      "num_input_tokens_seen": 294604332,
+      "step": 7520
+    },
+    {
+      "epoch": 0.3648609361372226,
+      "grad_norm": 0.38920724391937256,
+      "learning_rate": 3.6685101770578515e-05,
+      "loss": 1.2058,
+      "num_input_tokens_seen": 295015584,
+      "step": 7530
+    },
+    {
+      "epoch": 0.36534547921310206,
+      "grad_norm": 0.4151008725166321,
+      "learning_rate": 3.6650402292691946e-05,
+      "loss": 1.1891,
+      "num_input_tokens_seen": 295420200,
+      "step": 7540
+    },
+    {
+      "epoch": 0.3658300222889815,
+      "grad_norm": 0.4080232083797455,
+      "learning_rate": 3.661567412026925e-05,
+      "loss": 1.2027,
+      "num_input_tokens_seen": 295806112,
+      "step": 7550
+    },
+    {
+      "epoch": 0.36631456536486096,
+      "grad_norm": 0.43082767724990845,
+      "learning_rate": 3.658091733884471e-05,
+      "loss": 1.1592,
+      "num_input_tokens_seen": 296215744,
+      "step": 7560
+    },
+    {
+      "epoch": 0.3667991084407404,
+      "grad_norm": 0.3868250250816345,
+      "learning_rate": 3.654613203402307e-05,
+      "loss": 1.1657,
+      "num_input_tokens_seen": 296611788,
+      "step": 7570
+    },
+    {
+      "epoch": 0.36728365151661985,
+      "grad_norm": 0.3819040358066559,
+      "learning_rate": 3.6511318291479324e-05,
+      "loss": 1.1796,
+      "num_input_tokens_seen": 297006960,
+      "step": 7580
+    },
+    {
+      "epoch": 0.36776819459249926,
+      "grad_norm": 0.4080538749694824,
+      "learning_rate": 3.6476476196958514e-05,
+      "loss": 1.2078,
+      "num_input_tokens_seen": 297411416,
+      "step": 7590
+    },
+    {
+      "epoch": 0.36825273766837874,
+      "grad_norm": 0.40407416224479675,
+      "learning_rate": 3.644160583627551e-05,
+      "loss": 1.1925,
+      "num_input_tokens_seen": 297793232,
+      "step": 7600
+    },
+    {
+      "epoch": 0.36873728074425816,
+      "grad_norm": 0.37448978424072266,
+      "learning_rate": 3.6406707295314795e-05,
+      "loss": 1.1922,
+      "num_input_tokens_seen": 298204896,
+      "step": 7610
+    },
+    {
+      "epoch": 0.36922182382013763,
+      "grad_norm": 0.37111660838127136,
+      "learning_rate": 3.6371780660030266e-05,
+      "loss": 1.148,
+      "num_input_tokens_seen": 298561544,
+      "step": 7620
+    },
+    {
+      "epoch": 0.36970636689601705,
+      "grad_norm": 0.3740837574005127,
+      "learning_rate": 3.633682601644501e-05,
+      "loss": 1.2045,
+      "num_input_tokens_seen": 298944532,
+      "step": 7630
+    },
+    {
+      "epoch": 0.3701909099718965,
+      "grad_norm": 0.39125657081604004,
+      "learning_rate": 3.6301843450651096e-05,
+      "loss": 1.1708,
+      "num_input_tokens_seen": 299346324,
+      "step": 7640
+    },
+    {
+      "epoch": 0.37067545304777594,
+      "grad_norm": 0.3972494304180145,
+      "learning_rate": 3.6266833048809386e-05,
+      "loss": 1.1542,
+      "num_input_tokens_seen": 299741248,
+      "step": 7650
+    },
+    {
+      "epoch": 0.3711599961236554,
+      "grad_norm": 0.3871600329875946,
+      "learning_rate": 3.623179489714926e-05,
+      "loss": 1.1652,
+      "num_input_tokens_seen": 300143696,
+      "step": 7660
+    },
+    {
+      "epoch": 0.37164453919953483,
+      "grad_norm": 0.3876330852508545,
+      "learning_rate": 3.619672908196849e-05,
+      "loss": 1.1383,
+      "num_input_tokens_seen": 300512992,
+      "step": 7670
+    },
+    {
+      "epoch": 0.3721290822754143,
+      "grad_norm": 0.4013025760650635,
+      "learning_rate": 3.616163568963295e-05,
+      "loss": 1.1484,
+      "num_input_tokens_seen": 300904468,
+      "step": 7680
+    },
+    {
+      "epoch": 0.3726136253512937,
+      "grad_norm": 0.36160480976104736,
+      "learning_rate": 3.6126514806576437e-05,
+      "loss": 1.205,
+      "num_input_tokens_seen": 301325360,
+      "step": 7690
+    },
+    {
+      "epoch": 0.3730981684271732,
+      "grad_norm": 0.35836082696914673,
+      "learning_rate": 3.6091366519300476e-05,
+      "loss": 1.1436,
+      "num_input_tokens_seen": 301732708,
+      "step": 7700
+    },
+    {
+      "epoch": 0.3735827115030526,
+      "grad_norm": 0.43142881989479065,
+      "learning_rate": 3.605619091437408e-05,
+      "loss": 1.193,
+      "num_input_tokens_seen": 302120864,
+      "step": 7710
+    },
+    {
+      "epoch": 0.3740672545789321,
+      "grad_norm": 0.41082125902175903,
+      "learning_rate": 3.6020988078433537e-05,
+      "loss": 1.1767,
+      "num_input_tokens_seen": 302492456,
+      "step": 7720
+    },
+    {
+      "epoch": 0.3745517976548115,
+      "grad_norm": 0.393816739320755,
+      "learning_rate": 3.598575809818221e-05,
+      "loss": 1.2283,
+      "num_input_tokens_seen": 302846348,
+      "step": 7730
+    },
+    {
+      "epoch": 0.375036340730691,
+      "grad_norm": 0.4002731144428253,
+      "learning_rate": 3.5950501060390315e-05,
+      "loss": 1.1765,
+      "num_input_tokens_seen": 303200188,
+      "step": 7740
+    },
+    {
+      "epoch": 0.3755208838065704,
+      "grad_norm": 0.41621705889701843,
+      "learning_rate": 3.5915217051894725e-05,
+      "loss": 1.1259,
+      "num_input_tokens_seen": 303616612,
+      "step": 7750
+    },
+    {
+      "epoch": 0.37600542688244987,
+      "grad_norm": 0.352398544549942,
+      "learning_rate": 3.587990615959871e-05,
+      "loss": 1.1968,
+      "num_input_tokens_seen": 304005224,
+      "step": 7760
+    },
+    {
+      "epoch": 0.3764899699583293,
+      "grad_norm": 0.36988088488578796,
+      "learning_rate": 3.584456847047177e-05,
+      "loss": 1.1962,
+      "num_input_tokens_seen": 304405536,
+      "step": 7770
+    },
+    {
+      "epoch": 0.37697451303420876,
+      "grad_norm": 0.377210795879364,
+      "learning_rate": 3.5809204071549415e-05,
+      "loss": 1.2039,
+      "num_input_tokens_seen": 304791388,
+      "step": 7780
+    },
+    {
+      "epoch": 0.3774590561100882,
+      "grad_norm": 0.35832905769348145,
+      "learning_rate": 3.577381304993294e-05,
+      "loss": 1.1936,
+      "num_input_tokens_seen": 305184000,
+      "step": 7790
+    },
+    {
+      "epoch": 0.37794359918596765,
+      "grad_norm": 0.3834468424320221,
+      "learning_rate": 3.5738395492789176e-05,
+      "loss": 1.2239,
+      "num_input_tokens_seen": 305552272,
+      "step": 7800
+    },
+    {
+      "epoch": 0.37842814226184707,
+      "grad_norm": 0.3998955488204956,
+      "learning_rate": 3.570295148735036e-05,
+      "loss": 1.1768,
+      "num_input_tokens_seen": 305942616,
+      "step": 7810
+    },
+    {
+      "epoch": 0.37891268533772654,
+      "grad_norm": 0.382485955953598,
+      "learning_rate": 3.5667481120913836e-05,
+      "loss": 1.1988,
+      "num_input_tokens_seen": 306364336,
+      "step": 7820
+    },
+    {
+      "epoch": 0.37939722841360596,
+      "grad_norm": 0.43074601888656616,
+      "learning_rate": 3.5631984480841885e-05,
+      "loss": 1.2357,
+      "num_input_tokens_seen": 306775904,
+      "step": 7830
+    },
+    {
+      "epoch": 0.37988177148948543,
+      "grad_norm": 0.3977073132991791,
+      "learning_rate": 3.55964616545615e-05,
+      "loss": 1.1833,
+      "num_input_tokens_seen": 307121492,
+      "step": 7840
+    },
+    {
+      "epoch": 0.38036631456536485,
+      "grad_norm": 0.38828161358833313,
+      "learning_rate": 3.5560912729564155e-05,
+      "loss": 1.1875,
+      "num_input_tokens_seen": 307493488,
+      "step": 7850
+    },
+    {
+      "epoch": 0.3808508576412443,
+      "grad_norm": 0.404365211725235,
+      "learning_rate": 3.5525337793405636e-05,
+      "loss": 1.1878,
+      "num_input_tokens_seen": 307890584,
+      "step": 7860
+    },
+    {
+      "epoch": 0.38133540071712374,
+      "grad_norm": 0.42278143763542175,
+      "learning_rate": 3.548973693370576e-05,
+      "loss": 1.1444,
+      "num_input_tokens_seen": 308295900,
+      "step": 7870
+    },
+    {
+      "epoch": 0.3818199437930032,
+      "grad_norm": 0.38250425457954407,
+      "learning_rate": 3.545411023814823e-05,
+      "loss": 1.1498,
+      "num_input_tokens_seen": 308723400,
+      "step": 7880
+    },
+    {
+      "epoch": 0.38230448686888263,
+      "grad_norm": 0.4000505805015564,
+      "learning_rate": 3.541845779448034e-05,
+      "loss": 1.1917,
+      "num_input_tokens_seen": 309139432,
+      "step": 7890
+    },
+    {
+      "epoch": 0.3827890299447621,
+      "grad_norm": 0.4143424928188324,
+      "learning_rate": 3.5382779690512824e-05,
+      "loss": 1.2196,
+      "num_input_tokens_seen": 309543828,
+      "step": 7900
+    },
+    {
+      "epoch": 0.3832735730206415,
+      "grad_norm": 0.4070720672607422,
+      "learning_rate": 3.5347076014119606e-05,
+      "loss": 1.1941,
+      "num_input_tokens_seen": 309945560,
+      "step": 7910
+    },
+    {
+      "epoch": 0.383758116096521,
+      "grad_norm": 0.4233192801475525,
+      "learning_rate": 3.5311346853237614e-05,
+      "loss": 1.2011,
+      "num_input_tokens_seen": 310335412,
+      "step": 7920
+    },
+    {
+      "epoch": 0.3842426591724004,
+      "grad_norm": 0.4479762017726898,
+      "learning_rate": 3.527559229586653e-05,
+      "loss": 1.2178,
+      "num_input_tokens_seen": 310706360,
+      "step": 7930
+    },
+    {
+      "epoch": 0.3847272022482799,
+      "grad_norm": 0.41147053241729736,
+      "learning_rate": 3.523981243006857e-05,
+      "loss": 1.2376,
+      "num_input_tokens_seen": 311088520,
+      "step": 7940
+    },
+    {
+      "epoch": 0.3852117453241593,
+      "grad_norm": 0.38983359932899475,
+      "learning_rate": 3.5204007343968326e-05,
+      "loss": 1.2019,
+      "num_input_tokens_seen": 311467552,
+      "step": 7950
+    },
+    {
+      "epoch": 0.3856962884000388,
+      "grad_norm": 0.3905565142631531,
+      "learning_rate": 3.516817712575246e-05,
+      "loss": 1.1972,
+      "num_input_tokens_seen": 311857180,
+      "step": 7960
+    },
+    {
+      "epoch": 0.3861808314759182,
+      "grad_norm": 0.390117883682251,
+      "learning_rate": 3.513232186366956e-05,
+      "loss": 1.1901,
+      "num_input_tokens_seen": 312250400,
+      "step": 7970
+    },
+    {
+      "epoch": 0.38666537455179767,
+      "grad_norm": 0.3978836238384247,
+      "learning_rate": 3.50964416460299e-05,
+      "loss": 1.1729,
+      "num_input_tokens_seen": 312652712,
+      "step": 7980
+    },
+    {
+      "epoch": 0.3871499176276771,
+      "grad_norm": 0.4010826349258423,
+      "learning_rate": 3.5060536561205195e-05,
+      "loss": 1.1981,
+      "num_input_tokens_seen": 313069828,
+      "step": 7990
+    },
+    {
+      "epoch": 0.38763446070355656,
+      "grad_norm": 0.42469522356987,
+      "learning_rate": 3.5024606697628424e-05,
+      "loss": 1.2074,
+      "num_input_tokens_seen": 313452856,
+      "step": 8000
+    },
+    {
+      "epoch": 0.38763446070355656,
+      "eval_loss": 1.1777091026306152,
+      "eval_runtime": 5.3785,
+      "eval_samples_per_second": 27.889,
+      "eval_steps_per_second": 3.533,
+      "num_input_tokens_seen": 313452856,
+      "step": 8000
+    },
+    {
+      "epoch": 0.388119003779436,
+      "grad_norm": 0.3903387486934662,
+      "learning_rate": 3.49886521437936e-05,
+      "loss": 1.153,
+      "num_input_tokens_seen": 313830208,
+      "step": 8010
+    },
+    {
+      "epoch": 0.38860354685531545,
+      "grad_norm": 0.398037314414978,
+      "learning_rate": 3.495267298825555e-05,
+      "loss": 1.1707,
+      "num_input_tokens_seen": 314242840,
+      "step": 8020
+    },
+    {
+      "epoch": 0.38908808993119487,
+      "grad_norm": 0.3712460398674011,
+      "learning_rate": 3.4916669319629664e-05,
+      "loss": 1.161,
+      "num_input_tokens_seen": 314602680,
+      "step": 8030
+    },
+    {
+      "epoch": 0.38957263300707434,
+      "grad_norm": 0.4191719591617584,
+      "learning_rate": 3.4880641226591733e-05,
+      "loss": 1.1705,
+      "num_input_tokens_seen": 315002656,
+      "step": 8040
+    },
+    {
+      "epoch": 0.39005717608295376,
+      "grad_norm": 0.4063269793987274,
+      "learning_rate": 3.48445887978777e-05,
+      "loss": 1.1794,
+      "num_input_tokens_seen": 315420940,
+      "step": 8050
+    },
+    {
+      "epoch": 0.39054171915883323,
+      "grad_norm": 0.40375033020973206,
+      "learning_rate": 3.480851212228345e-05,
+      "loss": 1.1876,
+      "num_input_tokens_seen": 315794464,
+      "step": 8060
+    },
+    {
+      "epoch": 0.39102626223471265,
+      "grad_norm": 0.43218469619750977,
+      "learning_rate": 3.4772411288664576e-05,
+      "loss": 1.2107,
+      "num_input_tokens_seen": 316178848,
+      "step": 8070
+    },
+    {
+      "epoch": 0.3915108053105921,
+      "grad_norm": 0.3920303285121918,
+      "learning_rate": 3.473628638593618e-05,
+      "loss": 1.1755,
+      "num_input_tokens_seen": 316541328,
+      "step": 8080
+    },
+    {
+      "epoch": 0.39199534838647154,
+      "grad_norm": 0.405202716588974,
+      "learning_rate": 3.470013750307263e-05,
+      "loss": 1.1929,
+      "num_input_tokens_seen": 316916528,
+      "step": 8090
+    },
+    {
+      "epoch": 0.392479891462351,
+      "grad_norm": 0.39330440759658813,
+      "learning_rate": 3.466396472910739e-05,
+      "loss": 1.1855,
+      "num_input_tokens_seen": 317357952,
+      "step": 8100
+    },
+    {
+      "epoch": 0.39296443453823043,
+      "grad_norm": 0.3834272623062134,
+      "learning_rate": 3.462776815313274e-05,
+      "loss": 1.1863,
+      "num_input_tokens_seen": 317785332,
+      "step": 8110
+    },
+    {
+      "epoch": 0.3934489776141099,
+      "grad_norm": 0.40610063076019287,
+      "learning_rate": 3.4591547864299576e-05,
+      "loss": 1.1825,
+      "num_input_tokens_seen": 318172140,
+      "step": 8120
+    },
+    {
+      "epoch": 0.3939335206899893,
+      "grad_norm": 0.41190090775489807,
+      "learning_rate": 3.455530395181722e-05,
+      "loss": 1.1949,
+      "num_input_tokens_seen": 318556536,
+      "step": 8130
+    },
+    {
+      "epoch": 0.3944180637658688,
+      "grad_norm": 0.3720203936100006,
+      "learning_rate": 3.451903650495317e-05,
+      "loss": 1.1492,
+      "num_input_tokens_seen": 318939604,
+      "step": 8140
+    },
+    {
+      "epoch": 0.3949026068417482,
+      "grad_norm": 0.39604607224464417,
+      "learning_rate": 3.448274561303288e-05,
+      "loss": 1.2164,
+      "num_input_tokens_seen": 319344624,
+      "step": 8150
+    },
+    {
+      "epoch": 0.3953871499176277,
+      "grad_norm": 0.40863263607025146,
+      "learning_rate": 3.444643136543957e-05,
+      "loss": 1.1766,
+      "num_input_tokens_seen": 319738832,
+      "step": 8160
+    },
+    {
+      "epoch": 0.3958716929935071,
+      "grad_norm": 0.4493870437145233,
+      "learning_rate": 3.4410093851613964e-05,
+      "loss": 1.2016,
+      "num_input_tokens_seen": 320145484,
+      "step": 8170
+    },
+    {
+      "epoch": 0.3963562360693866,
+      "grad_norm": 0.3747680187225342,
+      "learning_rate": 3.4373733161054095e-05,
+      "loss": 1.1803,
+      "num_input_tokens_seen": 320561880,
+      "step": 8180
+    },
+    {
+      "epoch": 0.396840779145266,
+      "grad_norm": 0.39447423815727234,
+      "learning_rate": 3.433734938331508e-05,
+      "loss": 1.1853,
+      "num_input_tokens_seen": 320944848,
+      "step": 8190
+    },
+    {
+      "epoch": 0.39732532222114547,
+      "grad_norm": 0.4037897288799286,
+      "learning_rate": 3.43009426080089e-05,
+      "loss": 1.1759,
+      "num_input_tokens_seen": 321319020,
+      "step": 8200
+    },
+    {
+      "epoch": 0.3978098652970249,
+      "grad_norm": 0.3832351565361023,
+      "learning_rate": 3.4264512924804175e-05,
+      "loss": 1.1788,
+      "num_input_tokens_seen": 321694824,
+      "step": 8210
+    },
+    {
+      "epoch": 0.39829440837290436,
+      "grad_norm": 0.42781588435173035,
+      "learning_rate": 3.422806042342596e-05,
+      "loss": 1.1784,
+      "num_input_tokens_seen": 322077948,
+      "step": 8220
+    },
+    {
+      "epoch": 0.3987789514487838,
+      "grad_norm": 0.3955729305744171,
+      "learning_rate": 3.419158519365548e-05,
+      "loss": 1.234,
+      "num_input_tokens_seen": 322505044,
+      "step": 8230
+    },
+    {
+      "epoch": 0.39926349452466325,
+      "grad_norm": 0.419911652803421,
+      "learning_rate": 3.4155087325329985e-05,
+      "loss": 1.1763,
+      "num_input_tokens_seen": 322911104,
+      "step": 8240
+    },
+    {
+      "epoch": 0.39974803760054267,
+      "grad_norm": 0.4269978702068329,
+      "learning_rate": 3.411856690834243e-05,
+      "loss": 1.17,
+      "num_input_tokens_seen": 323320996,
+      "step": 8250
+    },
+    {
+      "epoch": 0.40023258067642214,
+      "grad_norm": 0.36580172181129456,
+      "learning_rate": 3.408202403264135e-05,
+      "loss": 1.1565,
+      "num_input_tokens_seen": 323709072,
+      "step": 8260
+    },
+    {
+      "epoch": 0.40071712375230156,
+      "grad_norm": 0.3848220705986023,
+      "learning_rate": 3.404545878823058e-05,
+      "loss": 1.1959,
+      "num_input_tokens_seen": 324080944,
+      "step": 8270
+    },
+    {
+      "epoch": 0.40120166682818104,
+      "grad_norm": 0.4162658452987671,
+      "learning_rate": 3.400887126516904e-05,
+      "loss": 1.1497,
+      "num_input_tokens_seen": 324485356,
+      "step": 8280
+    },
+    {
+      "epoch": 0.40168620990406045,
+      "grad_norm": 0.40066495537757874,
+      "learning_rate": 3.3972261553570536e-05,
+      "loss": 1.1657,
+      "num_input_tokens_seen": 324871712,
+      "step": 8290
+    },
+    {
+      "epoch": 0.4021707529799399,
+      "grad_norm": 0.3958793878555298,
+      "learning_rate": 3.3935629743603496e-05,
+      "loss": 1.1941,
+      "num_input_tokens_seen": 325257684,
+      "step": 8300
+    },
+    {
+      "epoch": 0.40265529605581935,
+      "grad_norm": 0.408408522605896,
+      "learning_rate": 3.389897592549082e-05,
+      "loss": 1.1954,
+      "num_input_tokens_seen": 325638416,
+      "step": 8310
+    },
+    {
+      "epoch": 0.4031398391316988,
+      "grad_norm": 0.36070606112480164,
+      "learning_rate": 3.3862300189509564e-05,
+      "loss": 1.2009,
+      "num_input_tokens_seen": 326018508,
+      "step": 8320
+    },
+    {
+      "epoch": 0.40362438220757824,
+      "grad_norm": 0.3974900543689728,
+      "learning_rate": 3.382560262599082e-05,
+      "loss": 1.1711,
+      "num_input_tokens_seen": 326405592,
+      "step": 8330
+    },
+    {
+      "epoch": 0.4041089252834577,
+      "grad_norm": 0.3950560390949249,
+      "learning_rate": 3.378888332531939e-05,
+      "loss": 1.1634,
+      "num_input_tokens_seen": 326784284,
+      "step": 8340
+    },
+    {
+      "epoch": 0.4045934683593371,
+      "grad_norm": 0.36520737409591675,
+      "learning_rate": 3.3752142377933646e-05,
+      "loss": 1.1633,
+      "num_input_tokens_seen": 327186212,
+      "step": 8350
+    },
+    {
+      "epoch": 0.4050780114352166,
+      "grad_norm": 0.4279801547527313,
+      "learning_rate": 3.371537987432526e-05,
+      "loss": 1.1751,
+      "num_input_tokens_seen": 327580260,
+      "step": 8360
+    },
+    {
+      "epoch": 0.405562554511096,
+      "grad_norm": 0.4253396987915039,
+      "learning_rate": 3.367859590503901e-05,
+      "loss": 1.1813,
+      "num_input_tokens_seen": 327974240,
+      "step": 8370
+    },
+    {
+      "epoch": 0.4060470975869755,
+      "grad_norm": 0.46800392866134644,
+      "learning_rate": 3.364179056067253e-05,
+      "loss": 1.2001,
+      "num_input_tokens_seen": 328378912,
+      "step": 8380
+    },
+    {
+      "epoch": 0.4065316406628549,
+      "grad_norm": 0.3665182292461395,
+      "learning_rate": 3.360496393187609e-05,
+      "loss": 1.2047,
+      "num_input_tokens_seen": 328787716,
+      "step": 8390
+    },
+    {
+      "epoch": 0.4070161837387344,
+      "grad_norm": 0.38969382643699646,
+      "learning_rate": 3.356811610935241e-05,
+      "loss": 1.1611,
+      "num_input_tokens_seen": 329162816,
+      "step": 8400
+    },
+    {
+      "epoch": 0.4075007268146138,
+      "grad_norm": 0.3802407383918762,
+      "learning_rate": 3.353124718385641e-05,
+      "loss": 1.2203,
+      "num_input_tokens_seen": 329545976,
+      "step": 8410
+    },
+    {
+      "epoch": 0.4079852698904933,
+      "grad_norm": 0.4015510082244873,
+      "learning_rate": 3.3494357246194956e-05,
+      "loss": 1.1915,
+      "num_input_tokens_seen": 329952104,
+      "step": 8420
+    },
+    {
+      "epoch": 0.4084698129663727,
+      "grad_norm": 0.3899809420108795,
+      "learning_rate": 3.345744638722669e-05,
+      "loss": 1.1899,
+      "num_input_tokens_seen": 330329116,
+      "step": 8430
+    },
+    {
+      "epoch": 0.40895435604225217,
+      "grad_norm": 0.3643754720687866,
+      "learning_rate": 3.3420514697861766e-05,
+      "loss": 1.2037,
+      "num_input_tokens_seen": 330723032,
+      "step": 8440
+    },
+    {
+      "epoch": 0.4094388991181316,
+      "grad_norm": 0.3885078728199005,
+      "learning_rate": 3.338356226906166e-05,
+      "loss": 1.2045,
+      "num_input_tokens_seen": 331128052,
+      "step": 8450
+    },
+    {
+      "epoch": 0.40992344219401106,
+      "grad_norm": 0.4163413643836975,
+      "learning_rate": 3.3346589191838915e-05,
+      "loss": 1.1901,
+      "num_input_tokens_seen": 331532708,
+      "step": 8460
+    },
+    {
+      "epoch": 0.4104079852698905,
+      "grad_norm": 0.36400115489959717,
+      "learning_rate": 3.330959555725695e-05,
+      "loss": 1.1846,
+      "num_input_tokens_seen": 331914620,
+      "step": 8470
+    },
+    {
+      "epoch": 0.41089252834576995,
+      "grad_norm": 0.36604899168014526,
+      "learning_rate": 3.327258145642979e-05,
+      "loss": 1.1852,
+      "num_input_tokens_seen": 332313264,
+      "step": 8480
+    },
+    {
+      "epoch": 0.41137707142164937,
+      "grad_norm": 0.4256085455417633,
+      "learning_rate": 3.323554698052187e-05,
+      "loss": 1.1764,
+      "num_input_tokens_seen": 332721544,
+      "step": 8490
+    },
+    {
+      "epoch": 0.41186161449752884,
+      "grad_norm": 0.39886176586151123,
+      "learning_rate": 3.3198492220747834e-05,
+      "loss": 1.1835,
+      "num_input_tokens_seen": 333090772,
+      "step": 8500
+    },
+    {
+      "epoch": 0.41234615757340826,
+      "grad_norm": 0.3774135410785675,
+      "learning_rate": 3.316141726837226e-05,
+      "loss": 1.1219,
+      "num_input_tokens_seen": 333473080,
+      "step": 8510
+    },
+    {
+      "epoch": 0.41283070064928773,
+      "grad_norm": 0.4200323522090912,
+      "learning_rate": 3.312432221470947e-05,
+      "loss": 1.187,
+      "num_input_tokens_seen": 333860928,
+      "step": 8520
+    },
+    {
+      "epoch": 0.41331524372516715,
+      "grad_norm": 0.433433473110199,
+      "learning_rate": 3.30872071511233e-05,
+      "loss": 1.1961,
+      "num_input_tokens_seen": 334240240,
+      "step": 8530
+    },
+    {
+      "epoch": 0.4137997868010466,
+      "grad_norm": 0.38179147243499756,
+      "learning_rate": 3.3050072169026855e-05,
+      "loss": 1.2174,
+      "num_input_tokens_seen": 334633964,
+      "step": 8540
+    },
+    {
+      "epoch": 0.41428432987692604,
+      "grad_norm": 0.3912273049354553,
+      "learning_rate": 3.30129173598823e-05,
+      "loss": 1.1801,
+      "num_input_tokens_seen": 335042032,
+      "step": 8550
+    },
+    {
+      "epoch": 0.4147688729528055,
+      "grad_norm": 0.3975712060928345,
+      "learning_rate": 3.297574281520065e-05,
+      "loss": 1.1647,
+      "num_input_tokens_seen": 335436124,
+      "step": 8560
+    },
+    {
+      "epoch": 0.41525341602868493,
+      "grad_norm": 0.45118552446365356,
+      "learning_rate": 3.293854862654151e-05,
+      "loss": 1.1916,
+      "num_input_tokens_seen": 335836836,
+      "step": 8570
+    },
+    {
+      "epoch": 0.4157379591045644,
+      "grad_norm": 0.3697469234466553,
+      "learning_rate": 3.290133488551288e-05,
+      "loss": 1.1983,
+      "num_input_tokens_seen": 336228900,
+      "step": 8580
+    },
+    {
+      "epoch": 0.4162225021804438,
+      "grad_norm": 0.37667909264564514,
+      "learning_rate": 3.286410168377091e-05,
+      "loss": 1.1968,
+      "num_input_tokens_seen": 336622764,
+      "step": 8590
+    },
+    {
+      "epoch": 0.4167070452563233,
+      "grad_norm": 0.36621060967445374,
+      "learning_rate": 3.282684911301968e-05,
+      "loss": 1.1642,
+      "num_input_tokens_seen": 337027652,
+      "step": 8600
+    },
+    {
+      "epoch": 0.4171915883322027,
+      "grad_norm": 0.39898422360420227,
+      "learning_rate": 3.2789577265010974e-05,
+      "loss": 1.2309,
+      "num_input_tokens_seen": 337410656,
+      "step": 8610
+    },
+    {
+      "epoch": 0.4176761314080822,
+      "grad_norm": 0.40708452463150024,
+      "learning_rate": 3.275228623154406e-05,
+      "loss": 1.1755,
+      "num_input_tokens_seen": 337815548,
+      "step": 8620
+    },
+    {
+      "epoch": 0.4181606744839616,
+      "grad_norm": 0.39701101183891296,
+      "learning_rate": 3.271497610446547e-05,
+      "loss": 1.2235,
+      "num_input_tokens_seen": 338219464,
+      "step": 8630
+    },
+    {
+      "epoch": 0.4186452175598411,
+      "grad_norm": 0.3921484351158142,
+      "learning_rate": 3.267764697566874e-05,
+      "loss": 1.1719,
+      "num_input_tokens_seen": 338635512,
+      "step": 8640
+    },
+    {
+      "epoch": 0.4191297606357205,
+      "grad_norm": 0.37697601318359375,
+      "learning_rate": 3.264029893709422e-05,
+      "loss": 1.1836,
+      "num_input_tokens_seen": 339021496,
+      "step": 8650
+    },
+    {
+      "epoch": 0.41961430371159997,
+      "grad_norm": 0.3878484070301056,
+      "learning_rate": 3.260293208072883e-05,
+      "loss": 1.1847,
+      "num_input_tokens_seen": 339403824,
+      "step": 8660
+    },
+    {
+      "epoch": 0.4200988467874794,
+      "grad_norm": 0.42603814601898193,
+      "learning_rate": 3.2565546498605834e-05,
+      "loss": 1.1747,
+      "num_input_tokens_seen": 339808580,
+      "step": 8670
+    },
+    {
+      "epoch": 0.42058338986335886,
+      "grad_norm": 0.3939875662326813,
+      "learning_rate": 3.252814228280464e-05,
+      "loss": 1.2169,
+      "num_input_tokens_seen": 340199464,
+      "step": 8680
+    },
+    {
+      "epoch": 0.4210679329392383,
+      "grad_norm": 0.3872847855091095,
+      "learning_rate": 3.249071952545052e-05,
+      "loss": 1.1669,
+      "num_input_tokens_seen": 340608396,
+      "step": 8690
+    },
+    {
+      "epoch": 0.42155247601511775,
+      "grad_norm": 0.4044415056705475,
+      "learning_rate": 3.245327831871442e-05,
+      "loss": 1.2167,
+      "num_input_tokens_seen": 340997568,
+      "step": 8700
+    },
+    {
+      "epoch": 0.42203701909099717,
+      "grad_norm": 0.4061294198036194,
+      "learning_rate": 3.2415818754812735e-05,
+      "loss": 1.179,
+      "num_input_tokens_seen": 341399468,
+      "step": 8710
+    },
+    {
+      "epoch": 0.42252156216687664,
+      "grad_norm": 0.3923657238483429,
+      "learning_rate": 3.237834092600709e-05,
+      "loss": 1.1482,
+      "num_input_tokens_seen": 341803448,
+      "step": 8720
+    },
+    {
+      "epoch": 0.42300610524275606,
+      "grad_norm": 0.3841986060142517,
+      "learning_rate": 3.234084492460404e-05,
+      "loss": 1.1648,
+      "num_input_tokens_seen": 342189656,
+      "step": 8730
+    },
+    {
+      "epoch": 0.42349064831863553,
+      "grad_norm": 0.4120123088359833,
+      "learning_rate": 3.230333084295496e-05,
+      "loss": 1.1752,
+      "num_input_tokens_seen": 342569416,
+      "step": 8740
+    },
+    {
+      "epoch": 0.42397519139451495,
+      "grad_norm": 0.44047102332115173,
+      "learning_rate": 3.226579877345572e-05,
+      "loss": 1.2093,
+      "num_input_tokens_seen": 342959072,
+      "step": 8750
+    },
+    {
+      "epoch": 0.4244597344703944,
+      "grad_norm": 0.38502237200737,
+      "learning_rate": 3.22282488085465e-05,
+      "loss": 1.1982,
+      "num_input_tokens_seen": 343375272,
+      "step": 8760
+    },
+    {
+      "epoch": 0.42494427754627384,
+      "grad_norm": 0.37675246596336365,
+      "learning_rate": 3.2190681040711566e-05,
+      "loss": 1.168,
+      "num_input_tokens_seen": 343750180,
+      "step": 8770
+    },
+    {
+      "epoch": 0.4254288206221533,
+      "grad_norm": 0.36886003613471985,
+      "learning_rate": 3.215309556247903e-05,
+      "loss": 1.2242,
+      "num_input_tokens_seen": 344158160,
+      "step": 8780
+    },
+    {
+      "epoch": 0.42591336369803273,
+      "grad_norm": 0.40360134840011597,
+      "learning_rate": 3.21154924664206e-05,
+      "loss": 1.2046,
+      "num_input_tokens_seen": 344541580,
+      "step": 8790
+    },
+    {
+      "epoch": 0.4263979067739122,
+      "grad_norm": 0.39642900228500366,
+      "learning_rate": 3.207787184515142e-05,
+      "loss": 1.2041,
+      "num_input_tokens_seen": 344912764,
+      "step": 8800
+    },
+    {
+      "epoch": 0.4268824498497916,
+      "grad_norm": 0.40765756368637085,
+      "learning_rate": 3.204023379132975e-05,
+      "loss": 1.1939,
+      "num_input_tokens_seen": 345319956,
+      "step": 8810
+    },
+    {
+      "epoch": 0.4273669929256711,
+      "grad_norm": 0.4131956100463867,
+      "learning_rate": 3.2002578397656826e-05,
+      "loss": 1.1394,
+      "num_input_tokens_seen": 345697944,
+      "step": 8820
+    },
+    {
+      "epoch": 0.4278515360015505,
+      "grad_norm": 0.4072543978691101,
+      "learning_rate": 3.196490575687657e-05,
+      "loss": 1.1678,
+      "num_input_tokens_seen": 346082640,
+      "step": 8830
+    },
+    {
+      "epoch": 0.42833607907743,
+      "grad_norm": 0.41665875911712646,
+      "learning_rate": 3.192721596177538e-05,
+      "loss": 1.1764,
+      "num_input_tokens_seen": 346505640,
+      "step": 8840
+    },
+    {
+      "epoch": 0.4288206221533094,
+      "grad_norm": 0.4276581406593323,
+      "learning_rate": 3.188950910518193e-05,
+      "loss": 1.1871,
+      "num_input_tokens_seen": 346926848,
+      "step": 8850
+    },
+    {
+      "epoch": 0.4293051652291889,
+      "grad_norm": 0.40989550948143005,
+      "learning_rate": 3.185178527996687e-05,
+      "loss": 1.2048,
+      "num_input_tokens_seen": 347308076,
+      "step": 8860
+    },
+    {
+      "epoch": 0.4297897083050683,
+      "grad_norm": 0.3860456347465515,
+      "learning_rate": 3.18140445790427e-05,
+      "loss": 1.1526,
+      "num_input_tokens_seen": 347719608,
+      "step": 8870
+    },
+    {
+      "epoch": 0.43027425138094777,
+      "grad_norm": 0.360655277967453,
+      "learning_rate": 3.1776287095363435e-05,
+      "loss": 1.1543,
+      "num_input_tokens_seen": 348104112,
+      "step": 8880
+    },
+    {
+      "epoch": 0.4307587944568272,
+      "grad_norm": 0.42288416624069214,
+      "learning_rate": 3.173851292192446e-05,
+      "loss": 1.1498,
+      "num_input_tokens_seen": 348488872,
+      "step": 8890
+    },
+    {
+      "epoch": 0.43124333753270666,
+      "grad_norm": 0.39610555768013,
+      "learning_rate": 3.170072215176224e-05,
+      "loss": 1.1501,
+      "num_input_tokens_seen": 348882096,
+      "step": 8900
+    },
+    {
+      "epoch": 0.4317278806085861,
+      "grad_norm": 0.40388280153274536,
+      "learning_rate": 3.166291487795413e-05,
+      "loss": 1.214,
+      "num_input_tokens_seen": 349271852,
+      "step": 8910
+    },
+    {
+      "epoch": 0.43221242368446555,
+      "grad_norm": 0.41157808899879456,
+      "learning_rate": 3.1625091193618144e-05,
+      "loss": 1.1712,
+      "num_input_tokens_seen": 349662936,
+      "step": 8920
+    },
+    {
+      "epoch": 0.43269696676034497,
+      "grad_norm": 0.41196179389953613,
+      "learning_rate": 3.158725119191269e-05,
+      "loss": 1.1742,
+      "num_input_tokens_seen": 350080888,
+      "step": 8930
+    },
+    {
+      "epoch": 0.43318150983622444,
+      "grad_norm": 0.3659604489803314,
+      "learning_rate": 3.1549394966036384e-05,
+      "loss": 1.1984,
+      "num_input_tokens_seen": 350479392,
+      "step": 8940
+    },
+    {
+      "epoch": 0.43366605291210386,
+      "grad_norm": 0.3877184987068176,
+      "learning_rate": 3.1511522609227795e-05,
+      "loss": 1.1329,
+      "num_input_tokens_seen": 350885976,
+      "step": 8950
+    },
+    {
+      "epoch": 0.43415059598798333,
+      "grad_norm": 0.4030652642250061,
+      "learning_rate": 3.147363421476522e-05,
+      "loss": 1.2162,
+      "num_input_tokens_seen": 351282332,
+      "step": 8960
+    },
+    {
+      "epoch": 0.43463513906386275,
+      "grad_norm": 0.38181596994400024,
+      "learning_rate": 3.1435729875966455e-05,
+      "loss": 1.1721,
+      "num_input_tokens_seen": 351676412,
+      "step": 8970
+    },
+    {
+      "epoch": 0.4351196821397422,
+      "grad_norm": 0.4003159701824188,
+      "learning_rate": 3.139780968618858e-05,
+      "loss": 1.1416,
+      "num_input_tokens_seen": 352076632,
+      "step": 8980
+    },
+    {
+      "epoch": 0.43560422521562164,
+      "grad_norm": 0.3967534303665161,
+      "learning_rate": 3.13598737388277e-05,
+      "loss": 1.1532,
+      "num_input_tokens_seen": 352466764,
+      "step": 8990
+    },
+    {
+      "epoch": 0.4360887682915011,
+      "grad_norm": 0.3866839110851288,
+      "learning_rate": 3.1321922127318724e-05,
+      "loss": 1.1999,
+      "num_input_tokens_seen": 352848332,
+      "step": 9000
+    },
+    {
+      "epoch": 0.43657331136738053,
+      "grad_norm": 0.3901658356189728,
+      "learning_rate": 3.128395494513517e-05,
+      "loss": 1.2005,
+      "num_input_tokens_seen": 353205300,
+      "step": 9010
+    },
+    {
+      "epoch": 0.43705785444326,
+      "grad_norm": 0.3866805136203766,
+      "learning_rate": 3.1245972285788885e-05,
+      "loss": 1.1484,
+      "num_input_tokens_seen": 353606764,
+      "step": 9020
+    },
+    {
+      "epoch": 0.4375423975191394,
+      "grad_norm": 0.3832225799560547,
+      "learning_rate": 3.120797424282984e-05,
+      "loss": 1.1939,
+      "num_input_tokens_seen": 353998028,
+      "step": 9030
+    },
+    {
+      "epoch": 0.4380269405950189,
+      "grad_norm": 0.39194390177726746,
+      "learning_rate": 3.116996090984588e-05,
+      "loss": 1.1963,
+      "num_input_tokens_seen": 354399420,
+      "step": 9040
+    },
+    {
+      "epoch": 0.4385114836708983,
+      "grad_norm": 0.3819257915019989,
+      "learning_rate": 3.113193238046255e-05,
+      "loss": 1.1531,
+      "num_input_tokens_seen": 354797548,
+      "step": 9050
+    },
+    {
+      "epoch": 0.4389960267467778,
+      "grad_norm": 0.3843849301338196,
+      "learning_rate": 3.1093888748342765e-05,
+      "loss": 1.1642,
+      "num_input_tokens_seen": 355200164,
+      "step": 9060
+    },
+    {
+      "epoch": 0.4394805698226572,
+      "grad_norm": 0.404051274061203,
+      "learning_rate": 3.10558301071867e-05,
+      "loss": 1.1794,
+      "num_input_tokens_seen": 355583100,
+      "step": 9070
+    },
+    {
+      "epoch": 0.4399651128985367,
+      "grad_norm": 0.3765638768672943,
+      "learning_rate": 3.1017756550731437e-05,
+      "loss": 1.2015,
+      "num_input_tokens_seen": 355999884,
+      "step": 9080
+    },
+    {
+      "epoch": 0.4404496559744161,
+      "grad_norm": 0.4142244756221771,
+      "learning_rate": 3.097966817275085e-05,
+      "loss": 1.1928,
+      "num_input_tokens_seen": 356363412,
+      "step": 9090
+    },
+    {
+      "epoch": 0.44093419905029557,
+      "grad_norm": 0.4150756597518921,
+      "learning_rate": 3.0941565067055275e-05,
+      "loss": 1.1665,
+      "num_input_tokens_seen": 356759816,
+      "step": 9100
+    },
+    {
+      "epoch": 0.441418742126175,
+      "grad_norm": 0.4147997498512268,
+      "learning_rate": 3.090344732749134e-05,
+      "loss": 1.1488,
+      "num_input_tokens_seen": 357163868,
+      "step": 9110
+    },
+    {
+      "epoch": 0.44190328520205446,
+      "grad_norm": 0.3696651756763458,
+      "learning_rate": 3.086531504794172e-05,
+      "loss": 1.1948,
+      "num_input_tokens_seen": 357578232,
+      "step": 9120
+    },
+    {
+      "epoch": 0.4423878282779339,
+      "grad_norm": 0.38870376348495483,
+      "learning_rate": 3.0827168322324884e-05,
+      "loss": 1.141,
+      "num_input_tokens_seen": 357996944,
+      "step": 9130
+    },
+    {
+      "epoch": 0.44287237135381335,
+      "grad_norm": 0.4145285487174988,
+      "learning_rate": 3.0789007244594904e-05,
+      "loss": 1.1882,
+      "num_input_tokens_seen": 358391992,
+      "step": 9140
+    },
+    {
+      "epoch": 0.44335691442969277,
+      "grad_norm": 0.4269642233848572,
+      "learning_rate": 3.0750831908741176e-05,
+      "loss": 1.1631,
+      "num_input_tokens_seen": 358795456,
+      "step": 9150
+    },
+    {
+      "epoch": 0.44384145750557225,
+      "grad_norm": 0.4123391807079315,
+      "learning_rate": 3.071264240878824e-05,
+      "loss": 1.223,
+      "num_input_tokens_seen": 359217840,
+      "step": 9160
+    },
+    {
+      "epoch": 0.44432600058145166,
+      "grad_norm": 0.3737592399120331,
+      "learning_rate": 3.06744388387955e-05,
+      "loss": 1.1953,
+      "num_input_tokens_seen": 359596700,
+      "step": 9170
+    },
+    {
+      "epoch": 0.44481054365733114,
+      "grad_norm": 0.35300344228744507,
+      "learning_rate": 3.0636221292857014e-05,
+      "loss": 1.1526,
+      "num_input_tokens_seen": 359971336,
+      "step": 9180
+    },
+    {
+      "epoch": 0.4452950867332106,
+      "grad_norm": 0.42388081550598145,
+      "learning_rate": 3.059798986510129e-05,
+      "loss": 1.2093,
+      "num_input_tokens_seen": 360365460,
+      "step": 9190
+    },
+    {
+      "epoch": 0.44577962980909003,
+      "grad_norm": 0.4228881895542145,
+      "learning_rate": 3.055974464969099e-05,
+      "loss": 1.183,
+      "num_input_tokens_seen": 360783972,
+      "step": 9200
+    },
+    {
+      "epoch": 0.4462641728849695,
+      "grad_norm": 0.47396156191825867,
+      "learning_rate": 3.0521485740822756e-05,
+      "loss": 1.1713,
+      "num_input_tokens_seen": 361168724,
+      "step": 9210
+    },
+    {
+      "epoch": 0.4467487159608489,
+      "grad_norm": 0.3625274896621704,
+      "learning_rate": 3.0483213232726943e-05,
+      "loss": 1.1959,
+      "num_input_tokens_seen": 361586496,
+      "step": 9220
+    },
+    {
+      "epoch": 0.4472332590367284,
+      "grad_norm": 0.3853507936000824,
+      "learning_rate": 3.044492721966743e-05,
+      "loss": 1.1483,
+      "num_input_tokens_seen": 361979848,
+      "step": 9230
+    },
+    {
+      "epoch": 0.4477178021126078,
+      "grad_norm": 0.3805350959300995,
+      "learning_rate": 3.040662779594131e-05,
+      "loss": 1.145,
+      "num_input_tokens_seen": 362373352,
+      "step": 9240
+    },
+    {
+      "epoch": 0.4482023451884873,
+      "grad_norm": 0.3566093444824219,
+      "learning_rate": 3.036831505587876e-05,
+      "loss": 1.1617,
+      "num_input_tokens_seen": 362795508,
+      "step": 9250
+    },
+    {
+      "epoch": 0.4486868882643667,
+      "grad_norm": 0.4072115123271942,
+      "learning_rate": 3.0329989093842708e-05,
+      "loss": 1.1401,
+      "num_input_tokens_seen": 363198416,
+      "step": 9260
+    },
+    {
+      "epoch": 0.4491714313402462,
+      "grad_norm": 0.3717459440231323,
+      "learning_rate": 3.0291650004228676e-05,
+      "loss": 1.1726,
+      "num_input_tokens_seen": 363578608,
+      "step": 9270
+    },
+    {
+      "epoch": 0.4496559744161256,
+      "grad_norm": 0.3549041450023651,
+      "learning_rate": 3.025329788146451e-05,
+      "loss": 1.1873,
+      "num_input_tokens_seen": 363973344,
+      "step": 9280
+    },
+    {
+      "epoch": 0.45014051749200507,
+      "grad_norm": 0.40732115507125854,
+      "learning_rate": 3.0214932820010155e-05,
+      "loss": 1.1968,
+      "num_input_tokens_seen": 364371260,
+      "step": 9290
+    },
+    {
+      "epoch": 0.4506250605678845,
+      "grad_norm": 0.39712613821029663,
+      "learning_rate": 3.017655491435742e-05,
+      "loss": 1.1529,
+      "num_input_tokens_seen": 364781760,
+      "step": 9300
+    },
+    {
+      "epoch": 0.45110960364376396,
+      "grad_norm": 0.4143085181713104,
+      "learning_rate": 3.0138164259029757e-05,
+      "loss": 1.2141,
+      "num_input_tokens_seen": 365169144,
+      "step": 9310
+    },
+    {
+      "epoch": 0.4515941467196434,
+      "grad_norm": 0.41558942198753357,
+      "learning_rate": 3.0099760948582023e-05,
+      "loss": 1.1541,
+      "num_input_tokens_seen": 365542904,
+      "step": 9320
+    },
+    {
+      "epoch": 0.45207868979552285,
+      "grad_norm": 0.38127419352531433,
+      "learning_rate": 3.0061345077600228e-05,
+      "loss": 1.1399,
+      "num_input_tokens_seen": 365944000,
+      "step": 9330
+    },
+    {
+      "epoch": 0.45256323287140227,
+      "grad_norm": 0.3887827396392822,
+      "learning_rate": 3.0022916740701334e-05,
+      "loss": 1.1597,
+      "num_input_tokens_seen": 366332944,
+      "step": 9340
+    },
+    {
+      "epoch": 0.45304777594728174,
+      "grad_norm": 0.38723742961883545,
+      "learning_rate": 2.9984476032532992e-05,
+      "loss": 1.1531,
+      "num_input_tokens_seen": 366697388,
+      "step": 9350
+    },
+    {
+      "epoch": 0.45353231902316116,
+      "grad_norm": 0.39358997344970703,
+      "learning_rate": 2.9946023047773336e-05,
+      "loss": 1.1241,
+      "num_input_tokens_seen": 367099924,
+      "step": 9360
+    },
+    {
+      "epoch": 0.45401686209904063,
+      "grad_norm": 0.3835636079311371,
+      "learning_rate": 2.9907557881130737e-05,
+      "loss": 1.2058,
+      "num_input_tokens_seen": 367476372,
+      "step": 9370
+    },
+    {
+      "epoch": 0.45450140517492005,
+      "grad_norm": 0.40065279603004456,
+      "learning_rate": 2.9869080627343555e-05,
+      "loss": 1.1679,
+      "num_input_tokens_seen": 367872588,
+      "step": 9380
+    },
+    {
+      "epoch": 0.4549859482507995,
+      "grad_norm": 0.38955920934677124,
+      "learning_rate": 2.9830591381179928e-05,
+      "loss": 1.1893,
+      "num_input_tokens_seen": 368260348,
+      "step": 9390
+    },
+    {
+      "epoch": 0.45547049132667894,
+      "grad_norm": 0.4499861001968384,
+      "learning_rate": 2.9792090237437543e-05,
+      "loss": 1.1168,
+      "num_input_tokens_seen": 368644964,
+      "step": 9400
+    },
+    {
+      "epoch": 0.4559550344025584,
+      "grad_norm": 0.386730432510376,
+      "learning_rate": 2.9753577290943376e-05,
+      "loss": 1.2014,
+      "num_input_tokens_seen": 369055912,
+      "step": 9410
+    },
+    {
+      "epoch": 0.45643957747843783,
+      "grad_norm": 0.39342692494392395,
+      "learning_rate": 2.971505263655347e-05,
+      "loss": 1.1853,
+      "num_input_tokens_seen": 369472516,
+      "step": 9420
+    },
+    {
+      "epoch": 0.4569241205543173,
+      "grad_norm": 0.4010606110095978,
+      "learning_rate": 2.9676516369152713e-05,
+      "loss": 1.0836,
+      "num_input_tokens_seen": 369855504,
+      "step": 9430
+    },
+    {
+      "epoch": 0.4574086636301967,
+      "grad_norm": 0.41329410672187805,
+      "learning_rate": 2.9637968583654597e-05,
+      "loss": 1.1578,
+      "num_input_tokens_seen": 370249352,
+      "step": 9440
+    },
+    {
+      "epoch": 0.4578932067060762,
+      "grad_norm": 0.3728090226650238,
+      "learning_rate": 2.9599409375000975e-05,
+      "loss": 1.1782,
+      "num_input_tokens_seen": 370653324,
+      "step": 9450
+    },
+    {
+      "epoch": 0.4583777497819556,
+      "grad_norm": 0.37402981519699097,
+      "learning_rate": 2.9560838838161843e-05,
+      "loss": 1.1704,
+      "num_input_tokens_seen": 371025900,
+      "step": 9460
+    },
+    {
+      "epoch": 0.4588622928578351,
+      "grad_norm": 0.39774417877197266,
+      "learning_rate": 2.9522257068135086e-05,
+      "loss": 1.1619,
+      "num_input_tokens_seen": 371424328,
+      "step": 9470
+    },
+    {
+      "epoch": 0.4593468359337145,
+      "grad_norm": 0.383232444524765,
+      "learning_rate": 2.9483664159946273e-05,
+      "loss": 1.1858,
+      "num_input_tokens_seen": 371838760,
+      "step": 9480
+    },
+    {
+      "epoch": 0.459831379009594,
+      "grad_norm": 0.37605851888656616,
+      "learning_rate": 2.9445060208648383e-05,
+      "loss": 1.1487,
+      "num_input_tokens_seen": 372233964,
+      "step": 9490
+    },
+    {
+      "epoch": 0.4603159220854734,
+      "grad_norm": 0.376461386680603,
+      "learning_rate": 2.9406445309321623e-05,
+      "loss": 1.2029,
+      "num_input_tokens_seen": 372599424,
+      "step": 9500
+    },
+    {
+      "epoch": 0.46080046516135287,
+      "grad_norm": 0.413618266582489,
+      "learning_rate": 2.9367819557073134e-05,
+      "loss": 1.1964,
+      "num_input_tokens_seen": 372979896,
+      "step": 9510
+    },
+    {
+      "epoch": 0.4612850082372323,
+      "grad_norm": 0.40922775864601135,
+      "learning_rate": 2.9329183047036807e-05,
+      "loss": 1.1989,
+      "num_input_tokens_seen": 373358408,
+      "step": 9520
+    },
+    {
+      "epoch": 0.46176955131311176,
+      "grad_norm": 0.37975606322288513,
+      "learning_rate": 2.9290535874373022e-05,
+      "loss": 1.2165,
+      "num_input_tokens_seen": 373734508,
+      "step": 9530
+    },
+    {
+      "epoch": 0.4622540943889912,
+      "grad_norm": 0.35514163970947266,
+      "learning_rate": 2.925187813426843e-05,
+      "loss": 1.1367,
+      "num_input_tokens_seen": 374105356,
+      "step": 9540
+    },
+    {
+      "epoch": 0.46273863746487065,
+      "grad_norm": 0.3892764151096344,
+      "learning_rate": 2.9213209921935703e-05,
+      "loss": 1.194,
+      "num_input_tokens_seen": 374514008,
+      "step": 9550
+    },
+    {
+      "epoch": 0.46322318054075007,
+      "grad_norm": 0.35142743587493896,
+      "learning_rate": 2.9174531332613292e-05,
+      "loss": 1.1837,
+      "num_input_tokens_seen": 374918944,
+      "step": 9560
+    },
+    {
+      "epoch": 0.46370772361662954,
+      "grad_norm": 0.42724543809890747,
+      "learning_rate": 2.9135842461565225e-05,
+      "loss": 1.1408,
+      "num_input_tokens_seen": 375331548,
+      "step": 9570
+    },
+    {
+      "epoch": 0.46419226669250896,
+      "grad_norm": 0.38824784755706787,
+      "learning_rate": 2.9097143404080853e-05,
+      "loss": 1.1919,
+      "num_input_tokens_seen": 375707484,
+      "step": 9580
+    },
+    {
+      "epoch": 0.46467680976838843,
+      "grad_norm": 0.3674311637878418,
+      "learning_rate": 2.905843425547461e-05,
+      "loss": 1.1473,
+      "num_input_tokens_seen": 376112184,
+      "step": 9590
+    },
+    {
+      "epoch": 0.46516135284426785,
+      "grad_norm": 0.3765534460544586,
+      "learning_rate": 2.9019715111085773e-05,
+      "loss": 1.1647,
+      "num_input_tokens_seen": 376494180,
+      "step": 9600
+    },
+    {
+      "epoch": 0.4656458959201473,
+      "grad_norm": 0.38776230812072754,
+      "learning_rate": 2.8980986066278255e-05,
+      "loss": 1.2236,
+      "num_input_tokens_seen": 376885136,
+      "step": 9610
+    },
+    {
+      "epoch": 0.46613043899602674,
+      "grad_norm": 0.37993282079696655,
+      "learning_rate": 2.8942247216440354e-05,
+      "loss": 1.2307,
+      "num_input_tokens_seen": 377277268,
+      "step": 9620
+    },
+    {
+      "epoch": 0.4666149820719062,
+      "grad_norm": 0.4081166088581085,
+      "learning_rate": 2.8903498656984514e-05,
+      "loss": 1.1428,
+      "num_input_tokens_seen": 377661420,
+      "step": 9630
+    },
+    {
+      "epoch": 0.46709952514778563,
+      "grad_norm": 0.3825088441371918,
+      "learning_rate": 2.8864740483347074e-05,
+      "loss": 1.1167,
+      "num_input_tokens_seen": 378079588,
+      "step": 9640
+    },
+    {
+      "epoch": 0.4675840682236651,
+      "grad_norm": 0.39228224754333496,
+      "learning_rate": 2.8825972790988082e-05,
+      "loss": 1.2089,
+      "num_input_tokens_seen": 378473248,
+      "step": 9650
+    },
+    {
+      "epoch": 0.4680686112995445,
+      "grad_norm": 0.40170297026634216,
+      "learning_rate": 2.8787195675391015e-05,
+      "loss": 1.1505,
+      "num_input_tokens_seen": 378849372,
+      "step": 9660
+    },
+    {
+      "epoch": 0.468553154375424,
+      "grad_norm": 0.39289790391921997,
+      "learning_rate": 2.874840923206256e-05,
+      "loss": 1.1567,
+      "num_input_tokens_seen": 379273348,
+      "step": 9670
+    },
+    {
+      "epoch": 0.4690376974513034,
+      "grad_norm": 0.3959096372127533,
+      "learning_rate": 2.870961355653239e-05,
+      "loss": 1.1711,
+      "num_input_tokens_seen": 379660408,
+      "step": 9680
+    },
+    {
+      "epoch": 0.4695222405271829,
+      "grad_norm": 0.43645361065864563,
+      "learning_rate": 2.8670808744352884e-05,
+      "loss": 1.1985,
+      "num_input_tokens_seen": 380054412,
+      "step": 9690
+    },
+    {
+      "epoch": 0.4700067836030623,
+      "grad_norm": 0.39692050218582153,
+      "learning_rate": 2.863199489109897e-05,
+      "loss": 1.182,
+      "num_input_tokens_seen": 380485684,
+      "step": 9700
+    },
+    {
+      "epoch": 0.4704913266789418,
+      "grad_norm": 0.42202410101890564,
+      "learning_rate": 2.8593172092367797e-05,
+      "loss": 1.1693,
+      "num_input_tokens_seen": 380889292,
+      "step": 9710
+    },
+    {
+      "epoch": 0.4709758697548212,
+      "grad_norm": 0.403148889541626,
+      "learning_rate": 2.8554340443778594e-05,
+      "loss": 1.1838,
+      "num_input_tokens_seen": 381290904,
+      "step": 9720
+    },
+    {
+      "epoch": 0.47146041283070067,
+      "grad_norm": 0.3972223699092865,
+      "learning_rate": 2.8515500040972347e-05,
+      "loss": 1.2143,
+      "num_input_tokens_seen": 381694440,
+      "step": 9730
+    },
+    {
+      "epoch": 0.4719449559065801,
+      "grad_norm": 0.3653228282928467,
+      "learning_rate": 2.8476650979611623e-05,
+      "loss": 1.1622,
+      "num_input_tokens_seen": 382095168,
+      "step": 9740
+    },
+    {
+      "epoch": 0.47242949898245956,
+      "grad_norm": 0.3957580626010895,
+      "learning_rate": 2.843779335538031e-05,
+      "loss": 1.1705,
+      "num_input_tokens_seen": 382467792,
+      "step": 9750
+    },
+    {
+      "epoch": 0.472914042058339,
+      "grad_norm": 0.4030523896217346,
+      "learning_rate": 2.8398927263983382e-05,
+      "loss": 1.1782,
+      "num_input_tokens_seen": 382880656,
+      "step": 9760
+    },
+    {
+      "epoch": 0.47339858513421845,
+      "grad_norm": 0.4061778485774994,
+      "learning_rate": 2.8360052801146687e-05,
+      "loss": 1.1652,
+      "num_input_tokens_seen": 383259940,
+      "step": 9770
+    },
+    {
+      "epoch": 0.47388312821009787,
+      "grad_norm": 0.38041362166404724,
+      "learning_rate": 2.8321170062616664e-05,
+      "loss": 1.1677,
+      "num_input_tokens_seen": 383648640,
+      "step": 9780
+    },
+    {
+      "epoch": 0.47436767128597734,
+      "grad_norm": 0.39732810854911804,
+      "learning_rate": 2.828227914416015e-05,
+      "loss": 1.1045,
+      "num_input_tokens_seen": 384047384,
+      "step": 9790
+    },
+    {
+      "epoch": 0.47485221436185676,
+      "grad_norm": 0.4323562681674957,
+      "learning_rate": 2.824338014156412e-05,
+      "loss": 1.1875,
+      "num_input_tokens_seen": 384447200,
+      "step": 9800
+    },
+    {
+      "epoch": 0.47533675743773623,
+      "grad_norm": 0.3938267230987549,
+      "learning_rate": 2.8204473150635476e-05,
+      "loss": 1.1853,
+      "num_input_tokens_seen": 384839616,
+      "step": 9810
+    },
+    {
+      "epoch": 0.47582130051361565,
+      "grad_norm": 0.3984321355819702,
+      "learning_rate": 2.816555826720078e-05,
+      "loss": 1.1566,
+      "num_input_tokens_seen": 385227704,
+      "step": 9820
+    },
+    {
+      "epoch": 0.4763058435894951,
+      "grad_norm": 0.37600892782211304,
+      "learning_rate": 2.8126635587106036e-05,
+      "loss": 1.1883,
+      "num_input_tokens_seen": 385642608,
+      "step": 9830
+    },
+    {
+      "epoch": 0.47679038666537454,
+      "grad_norm": 0.43121588230133057,
+      "learning_rate": 2.808770520621646e-05,
+      "loss": 1.1532,
+      "num_input_tokens_seen": 386042136,
+      "step": 9840
+    },
+    {
+      "epoch": 0.477274929741254,
+      "grad_norm": 0.4222434163093567,
+      "learning_rate": 2.8048767220416223e-05,
+      "loss": 1.175,
+      "num_input_tokens_seen": 386420708,
+      "step": 9850
+    },
+    {
+      "epoch": 0.47775947281713343,
+      "grad_norm": 0.423554927110672,
+      "learning_rate": 2.800982172560823e-05,
+      "loss": 1.1398,
+      "num_input_tokens_seen": 386777908,
+      "step": 9860
+    },
+    {
+      "epoch": 0.4782440158930129,
+      "grad_norm": 0.3951357305049896,
+      "learning_rate": 2.7970868817713887e-05,
+      "loss": 1.1527,
+      "num_input_tokens_seen": 387176340,
+      "step": 9870
+    },
+    {
+      "epoch": 0.4787285589688923,
+      "grad_norm": 0.3626028001308441,
+      "learning_rate": 2.7931908592672844e-05,
+      "loss": 1.1641,
+      "num_input_tokens_seen": 387547504,
+      "step": 9880
+    },
+    {
+      "epoch": 0.4792131020447718,
+      "grad_norm": 0.4093586504459381,
+      "learning_rate": 2.78929411464428e-05,
+      "loss": 1.1164,
+      "num_input_tokens_seen": 387947832,
+      "step": 9890
+    },
+    {
+      "epoch": 0.4796976451206512,
+      "grad_norm": 0.3790941536426544,
+      "learning_rate": 2.7853966574999197e-05,
+      "loss": 1.1421,
+      "num_input_tokens_seen": 388371140,
+      "step": 9900
+    },
+    {
+      "epoch": 0.4801821881965307,
+      "grad_norm": 0.44463813304901123,
+      "learning_rate": 2.7814984974335067e-05,
+      "loss": 1.178,
+      "num_input_tokens_seen": 388758584,
+      "step": 9910
+    },
+    {
+      "epoch": 0.4806667312724101,
+      "grad_norm": 0.40607205033302307,
+      "learning_rate": 2.7775996440460733e-05,
+      "loss": 1.2158,
+      "num_input_tokens_seen": 389136284,
+      "step": 9920
+    },
+    {
+      "epoch": 0.4811512743482896,
+      "grad_norm": 0.41725072264671326,
+      "learning_rate": 2.7737001069403608e-05,
+      "loss": 1.1731,
+      "num_input_tokens_seen": 389528592,
+      "step": 9930
+    },
+    {
+      "epoch": 0.481635817424169,
+      "grad_norm": 0.43715402483940125,
+      "learning_rate": 2.7697998957207915e-05,
+      "loss": 1.1359,
+      "num_input_tokens_seen": 389924908,
+      "step": 9940
+    },
+    {
+      "epoch": 0.4821203605000485,
+      "grad_norm": 0.4026664197444916,
+      "learning_rate": 2.765899019993453e-05,
+      "loss": 1.1334,
+      "num_input_tokens_seen": 390327408,
+      "step": 9950
+    },
+    {
+      "epoch": 0.4826049035759279,
+      "grad_norm": 0.38899385929107666,
+      "learning_rate": 2.7619974893660643e-05,
+      "loss": 1.1628,
+      "num_input_tokens_seen": 390716640,
+      "step": 9960
+    },
+    {
+      "epoch": 0.48308944665180736,
+      "grad_norm": 0.3987119197845459,
+      "learning_rate": 2.758095313447961e-05,
+      "loss": 1.1804,
+      "num_input_tokens_seen": 391147420,
+      "step": 9970
+    },
+    {
+      "epoch": 0.4835739897276868,
+      "grad_norm": 0.4419252872467041,
+      "learning_rate": 2.754192501850066e-05,
+      "loss": 1.1857,
+      "num_input_tokens_seen": 391523552,
+      "step": 9980
+    },
+    {
+      "epoch": 0.48405853280356625,
+      "grad_norm": 0.40142005681991577,
+      "learning_rate": 2.7502890641848696e-05,
+      "loss": 1.2214,
+      "num_input_tokens_seen": 391906876,
+      "step": 9990
+    },
+    {
+      "epoch": 0.4845430758794457,
+      "grad_norm": 0.36947381496429443,
+      "learning_rate": 2.7463850100664028e-05,
+      "loss": 1.1617,
+      "num_input_tokens_seen": 392316428,
+      "step": 10000
+    },
+    {
+      "epoch": 0.4845430758794457,
+      "eval_loss": 1.1596810817718506,
+      "eval_runtime": 5.1226,
+      "eval_samples_per_second": 29.282,
+      "eval_steps_per_second": 3.709,
+      "num_input_tokens_seen": 392316428,
+      "step": 10000
+    },
+    {
+      "epoch": 0.48502761895532515,
+      "grad_norm": 0.38736334443092346,
+      "learning_rate": 2.7424803491102158e-05,
+      "loss": 1.1817,
+      "num_input_tokens_seen": 392732788,
+      "step": 10010
+    },
+    {
+      "epoch": 0.48551216203120456,
+      "grad_norm": 0.40144097805023193,
+      "learning_rate": 2.738575090933352e-05,
+      "loss": 1.1628,
+      "num_input_tokens_seen": 393114108,
+      "step": 10020
+    },
+    {
+      "epoch": 0.48599670510708404,
+      "grad_norm": 0.42609161138534546,
+      "learning_rate": 2.7346692451543264e-05,
+      "loss": 1.1679,
+      "num_input_tokens_seen": 393509264,
+      "step": 10030
+    },
+    {
+      "epoch": 0.48648124818296345,
+      "grad_norm": 0.4124153256416321,
+      "learning_rate": 2.730762821393103e-05,
+      "loss": 1.1491,
+      "num_input_tokens_seen": 393915040,
+      "step": 10040
+    },
+    {
+      "epoch": 0.48696579125884293,
+      "grad_norm": 0.42232540249824524,
+      "learning_rate": 2.726855829271066e-05,
+      "loss": 1.1626,
+      "num_input_tokens_seen": 394289168,
+      "step": 10050
+    },
+    {
+      "epoch": 0.48745033433472235,
+      "grad_norm": 0.4197705388069153,
+      "learning_rate": 2.722948278411003e-05,
+      "loss": 1.1708,
+      "num_input_tokens_seen": 394680292,
+      "step": 10060
+    },
+    {
+      "epoch": 0.4879348774106018,
+      "grad_norm": 0.3949076533317566,
+      "learning_rate": 2.7190401784370743e-05,
+      "loss": 1.1591,
+      "num_input_tokens_seen": 395070520,
+      "step": 10070
+    },
+    {
+      "epoch": 0.48841942048648124,
+      "grad_norm": 0.42010289430618286,
+      "learning_rate": 2.7151315389747967e-05,
+      "loss": 1.1333,
+      "num_input_tokens_seen": 395500444,
+      "step": 10080
+    },
+    {
+      "epoch": 0.4889039635623607,
+      "grad_norm": 0.36050793528556824,
+      "learning_rate": 2.7112223696510108e-05,
+      "loss": 1.1303,
+      "num_input_tokens_seen": 395889484,
+      "step": 10090
+    },
+    {
+      "epoch": 0.48938850663824013,
+      "grad_norm": 0.40366074442863464,
+      "learning_rate": 2.7073126800938666e-05,
+      "loss": 1.1765,
+      "num_input_tokens_seen": 396277572,
+      "step": 10100
+    },
+    {
+      "epoch": 0.4898730497141196,
+      "grad_norm": 0.410087913274765,
+      "learning_rate": 2.7034024799327928e-05,
+      "loss": 1.1247,
+      "num_input_tokens_seen": 396678980,
+      "step": 10110
+    },
+    {
+      "epoch": 0.490357592789999,
+      "grad_norm": 0.38809484243392944,
+      "learning_rate": 2.6994917787984764e-05,
+      "loss": 1.1675,
+      "num_input_tokens_seen": 397061400,
+      "step": 10120
+    },
+    {
+      "epoch": 0.4908421358658785,
+      "grad_norm": 0.4384767711162567,
+      "learning_rate": 2.69558058632284e-05,
+      "loss": 1.1633,
+      "num_input_tokens_seen": 397457472,
+      "step": 10130
+    },
+    {
+      "epoch": 0.4913266789417579,
+      "grad_norm": 0.4038788974285126,
+      "learning_rate": 2.6916689121390127e-05,
+      "loss": 1.1532,
+      "num_input_tokens_seen": 397885272,
+      "step": 10140
+    },
+    {
+      "epoch": 0.4918112220176374,
+      "grad_norm": 0.4135740399360657,
+      "learning_rate": 2.6877567658813134e-05,
+      "loss": 1.1828,
+      "num_input_tokens_seen": 398292228,
+      "step": 10150
+    },
+    {
+      "epoch": 0.4922957650935168,
+      "grad_norm": 0.3874277174472809,
+      "learning_rate": 2.6838441571852223e-05,
+      "loss": 1.1119,
+      "num_input_tokens_seen": 398678720,
+      "step": 10160
+    },
+    {
+      "epoch": 0.4927803081693963,
+      "grad_norm": 0.4087202548980713,
+      "learning_rate": 2.679931095687358e-05,
+      "loss": 1.1294,
+      "num_input_tokens_seen": 399066144,
+      "step": 10170
+    },
+    {
+      "epoch": 0.4932648512452757,
+      "grad_norm": 0.39371687173843384,
+      "learning_rate": 2.6760175910254565e-05,
+      "loss": 1.1805,
+      "num_input_tokens_seen": 399480648,
+      "step": 10180
+    },
+    {
+      "epoch": 0.49374939432115517,
+      "grad_norm": 0.39417147636413574,
+      "learning_rate": 2.6721036528383424e-05,
+      "loss": 1.1356,
+      "num_input_tokens_seen": 399859596,
+      "step": 10190
+    },
+    {
+      "epoch": 0.4942339373970346,
+      "grad_norm": 0.394105464220047,
+      "learning_rate": 2.66818929076591e-05,
+      "loss": 1.1561,
+      "num_input_tokens_seen": 400214040,
+      "step": 10200
+    },
+    {
+      "epoch": 0.49471848047291406,
+      "grad_norm": 0.3794916868209839,
+      "learning_rate": 2.664274514449097e-05,
+      "loss": 1.1644,
+      "num_input_tokens_seen": 400607764,
+      "step": 10210
+    },
+    {
+      "epoch": 0.4952030235487935,
+      "grad_norm": 0.43505197763442993,
+      "learning_rate": 2.660359333529862e-05,
+      "loss": 1.1151,
+      "num_input_tokens_seen": 400977376,
+      "step": 10220
+    },
+    {
+      "epoch": 0.49568756662467295,
+      "grad_norm": 0.373087614774704,
+      "learning_rate": 2.6564437576511587e-05,
+      "loss": 1.1589,
+      "num_input_tokens_seen": 401363624,
+      "step": 10230
+    },
+    {
+      "epoch": 0.49617210970055237,
+      "grad_norm": 0.41389840841293335,
+      "learning_rate": 2.6525277964569155e-05,
+      "loss": 1.1478,
+      "num_input_tokens_seen": 401733716,
+      "step": 10240
+    },
+    {
+      "epoch": 0.49665665277643184,
+      "grad_norm": 0.41126182675361633,
+      "learning_rate": 2.648611459592008e-05,
+      "loss": 1.1653,
+      "num_input_tokens_seen": 402141292,
+      "step": 10250
+    },
+    {
+      "epoch": 0.49714119585231126,
+      "grad_norm": 0.3853723108768463,
+      "learning_rate": 2.644694756702238e-05,
+      "loss": 1.1522,
+      "num_input_tokens_seen": 402544236,
+      "step": 10260
+    },
+    {
+      "epoch": 0.49762573892819073,
+      "grad_norm": 0.39461591839790344,
+      "learning_rate": 2.640777697434309e-05,
+      "loss": 1.175,
+      "num_input_tokens_seen": 402961036,
+      "step": 10270
+    },
+    {
+      "epoch": 0.49811028200407015,
+      "grad_norm": 0.4081992506980896,
+      "learning_rate": 2.6368602914358008e-05,
+      "loss": 1.16,
+      "num_input_tokens_seen": 403377968,
+      "step": 10280
+    },
+    {
+      "epoch": 0.4985948250799496,
+      "grad_norm": 0.4176690876483917,
+      "learning_rate": 2.63294254835515e-05,
+      "loss": 1.1912,
+      "num_input_tokens_seen": 403777800,
+      "step": 10290
+    },
+    {
+      "epoch": 0.49907936815582904,
+      "grad_norm": 0.38343364000320435,
+      "learning_rate": 2.6290244778416202e-05,
+      "loss": 1.1198,
+      "num_input_tokens_seen": 404136320,
+      "step": 10300
+    },
+    {
+      "epoch": 0.4995639112317085,
+      "grad_norm": 0.4266665577888489,
+      "learning_rate": 2.6251060895452844e-05,
+      "loss": 1.2289,
+      "num_input_tokens_seen": 404515816,
+      "step": 10310
+    },
+    {
+      "epoch": 0.5000484543075879,
+      "grad_norm": 0.36221781373023987,
+      "learning_rate": 2.6211873931169955e-05,
+      "loss": 1.1902,
+      "num_input_tokens_seen": 404931996,
+      "step": 10320
+    },
+    {
+      "epoch": 0.5005329973834673,
+      "grad_norm": 0.4041613042354584,
+      "learning_rate": 2.6172683982083675e-05,
+      "loss": 1.1757,
+      "num_input_tokens_seen": 405288152,
+      "step": 10330
+    },
+    {
+      "epoch": 0.5010175404593469,
+      "grad_norm": 0.38727837800979614,
+      "learning_rate": 2.6133491144717475e-05,
+      "loss": 1.1697,
+      "num_input_tokens_seen": 405676608,
+      "step": 10340
+    },
+    {
+      "epoch": 0.5015020835352263,
+      "grad_norm": 0.41572102904319763,
+      "learning_rate": 2.6094295515601967e-05,
+      "loss": 1.1641,
+      "num_input_tokens_seen": 406068632,
+      "step": 10350
+    },
+    {
+      "epoch": 0.5019866266111057,
+      "grad_norm": 0.42120370268821716,
+      "learning_rate": 2.605509719127463e-05,
+      "loss": 1.1326,
+      "num_input_tokens_seen": 406465000,
+      "step": 10360
+    },
+    {
+      "epoch": 0.5024711696869851,
+      "grad_norm": 0.4009106755256653,
+      "learning_rate": 2.6015896268279555e-05,
+      "loss": 1.1493,
+      "num_input_tokens_seen": 406845928,
+      "step": 10370
+    },
+    {
+      "epoch": 0.5029557127628647,
+      "grad_norm": 0.3664901554584503,
+      "learning_rate": 2.5976692843167255e-05,
+      "loss": 1.1917,
+      "num_input_tokens_seen": 407225352,
+      "step": 10380
+    },
+    {
+      "epoch": 0.5034402558387441,
+      "grad_norm": 0.4137326180934906,
+      "learning_rate": 2.5937487012494422e-05,
+      "loss": 1.1673,
+      "num_input_tokens_seen": 407601740,
+      "step": 10390
+    },
+    {
+      "epoch": 0.5039247989146235,
+      "grad_norm": 0.3960126042366028,
+      "learning_rate": 2.5898278872823646e-05,
+      "loss": 1.1413,
+      "num_input_tokens_seen": 407997996,
+      "step": 10400
+    },
+    {
+      "epoch": 0.5044093419905029,
+      "grad_norm": 0.42062610387802124,
+      "learning_rate": 2.5859068520723205e-05,
+      "loss": 1.1753,
+      "num_input_tokens_seen": 408377408,
+      "step": 10410
+    },
+    {
+      "epoch": 0.5048938850663824,
+      "grad_norm": 0.35909387469291687,
+      "learning_rate": 2.581985605276684e-05,
+      "loss": 1.1511,
+      "num_input_tokens_seen": 408721880,
+      "step": 10420
+    },
+    {
+      "epoch": 0.5053784281422619,
+      "grad_norm": 0.42842748761177063,
+      "learning_rate": 2.578064156553349e-05,
+      "loss": 1.1312,
+      "num_input_tokens_seen": 409143260,
+      "step": 10430
+    },
+    {
+      "epoch": 0.5058629712181413,
+      "grad_norm": 0.3839662969112396,
+      "learning_rate": 2.5741425155607097e-05,
+      "loss": 1.1704,
+      "num_input_tokens_seen": 409535852,
+      "step": 10440
+    },
+    {
+      "epoch": 0.5063475142940207,
+      "grad_norm": 0.41793426871299744,
+      "learning_rate": 2.57022069195763e-05,
+      "loss": 1.1989,
+      "num_input_tokens_seen": 409937740,
+      "step": 10450
+    },
+    {
+      "epoch": 0.5068320573699002,
+      "grad_norm": 0.4150349199771881,
+      "learning_rate": 2.5662986954034245e-05,
+      "loss": 1.1293,
+      "num_input_tokens_seen": 410319088,
+      "step": 10460
+    },
+    {
+      "epoch": 0.5073166004457796,
+      "grad_norm": 0.37542667984962463,
+      "learning_rate": 2.5623765355578354e-05,
+      "loss": 1.2247,
+      "num_input_tokens_seen": 410678500,
+      "step": 10470
+    },
+    {
+      "epoch": 0.5078011435216591,
+      "grad_norm": 0.4027327001094818,
+      "learning_rate": 2.5584542220810065e-05,
+      "loss": 1.1796,
+      "num_input_tokens_seen": 411079744,
+      "step": 10480
+    },
+    {
+      "epoch": 0.5082856865975385,
+      "grad_norm": 0.39059481024742126,
+      "learning_rate": 2.55453176463346e-05,
+      "loss": 1.1849,
+      "num_input_tokens_seen": 411488444,
+      "step": 10490
+    },
+    {
+      "epoch": 0.508770229673418,
+      "grad_norm": 0.40780818462371826,
+      "learning_rate": 2.5506091728760702e-05,
+      "loss": 1.1187,
+      "num_input_tokens_seen": 411879348,
+      "step": 10500
+    },
+    {
+      "epoch": 0.5092547727492974,
+      "grad_norm": 0.3999616503715515,
+      "learning_rate": 2.546686456470046e-05,
+      "loss": 1.1775,
+      "num_input_tokens_seen": 412294412,
+      "step": 10510
+    },
+    {
+      "epoch": 0.5097393158251768,
+      "grad_norm": 0.4117770195007324,
+      "learning_rate": 2.5427636250769016e-05,
+      "loss": 1.1678,
+      "num_input_tokens_seen": 412688988,
+      "step": 10520
+    },
+    {
+      "epoch": 0.5102238589010563,
+      "grad_norm": 0.3824278712272644,
+      "learning_rate": 2.538840688358435e-05,
+      "loss": 1.1374,
+      "num_input_tokens_seen": 413069468,
+      "step": 10530
+    },
+    {
+      "epoch": 0.5107084019769358,
+      "grad_norm": 0.3502415418624878,
+      "learning_rate": 2.5349176559767007e-05,
+      "loss": 1.2013,
+      "num_input_tokens_seen": 413440188,
+      "step": 10540
+    },
+    {
+      "epoch": 0.5111929450528152,
+      "grad_norm": 0.3400461971759796,
+      "learning_rate": 2.530994537593994e-05,
+      "loss": 1.1755,
+      "num_input_tokens_seen": 413826580,
+      "step": 10550
+    },
+    {
+      "epoch": 0.5116774881286946,
+      "grad_norm": 0.4381474554538727,
+      "learning_rate": 2.527071342872817e-05,
+      "loss": 1.1644,
+      "num_input_tokens_seen": 414235248,
+      "step": 10560
+    },
+    {
+      "epoch": 0.512162031204574,
+      "grad_norm": 0.3964703381061554,
+      "learning_rate": 2.5231480814758633e-05,
+      "loss": 1.1745,
+      "num_input_tokens_seen": 414635544,
+      "step": 10570
+    },
+    {
+      "epoch": 0.5126465742804536,
+      "grad_norm": 0.3687261939048767,
+      "learning_rate": 2.51922476306599e-05,
+      "loss": 1.166,
+      "num_input_tokens_seen": 415043552,
+      "step": 10580
+    },
+    {
+      "epoch": 0.513131117356333,
+      "grad_norm": 0.3944050669670105,
+      "learning_rate": 2.5153013973061916e-05,
+      "loss": 1.2209,
+      "num_input_tokens_seen": 415413920,
+      "step": 10590
+    },
+    {
+      "epoch": 0.5136156604322124,
+      "grad_norm": 0.4199015498161316,
+      "learning_rate": 2.511377993859584e-05,
+      "loss": 1.1452,
+      "num_input_tokens_seen": 415791632,
+      "step": 10600
+    },
+    {
+      "epoch": 0.5141002035080918,
+      "grad_norm": 0.4487917721271515,
+      "learning_rate": 2.507454562389372e-05,
+      "loss": 1.1334,
+      "num_input_tokens_seen": 416204172,
+      "step": 10610
+    },
+    {
+      "epoch": 0.5145847465839714,
+      "grad_norm": 0.4062841832637787,
+      "learning_rate": 2.5035311125588322e-05,
+      "loss": 1.1398,
+      "num_input_tokens_seen": 416591364,
+      "step": 10620
+    },
+    {
+      "epoch": 0.5150692896598508,
+      "grad_norm": 0.41555893421173096,
+      "learning_rate": 2.4996076540312854e-05,
+      "loss": 1.1504,
+      "num_input_tokens_seen": 416957332,
+      "step": 10630
+    },
+    {
+      "epoch": 0.5155538327357302,
+      "grad_norm": 0.35754913091659546,
+      "learning_rate": 2.4956841964700718e-05,
+      "loss": 1.1095,
+      "num_input_tokens_seen": 417326068,
+      "step": 10640
+    },
+    {
+      "epoch": 0.5160383758116096,
+      "grad_norm": 0.381197065114975,
+      "learning_rate": 2.4917607495385338e-05,
+      "loss": 1.1859,
+      "num_input_tokens_seen": 417756524,
+      "step": 10650
+    },
+    {
+      "epoch": 0.5165229188874891,
+      "grad_norm": 0.41809821128845215,
+      "learning_rate": 2.4878373228999835e-05,
+      "loss": 1.1568,
+      "num_input_tokens_seen": 418159804,
+      "step": 10660
+    },
+    {
+      "epoch": 0.5170074619633686,
+      "grad_norm": 0.39222896099090576,
+      "learning_rate": 2.4839139262176837e-05,
+      "loss": 1.1618,
+      "num_input_tokens_seen": 418565656,
+      "step": 10670
+    },
+    {
+      "epoch": 0.517492005039248,
+      "grad_norm": 0.38486379384994507,
+      "learning_rate": 2.4799905691548245e-05,
+      "loss": 1.1738,
+      "num_input_tokens_seen": 418952312,
+      "step": 10680
+    },
+    {
+      "epoch": 0.5179765481151274,
+      "grad_norm": 0.3719366788864136,
+      "learning_rate": 2.476067261374499e-05,
+      "loss": 1.1973,
+      "num_input_tokens_seen": 419369500,
+      "step": 10690
+    },
+    {
+      "epoch": 0.5184610911910069,
+      "grad_norm": 0.3997963070869446,
+      "learning_rate": 2.4721440125396773e-05,
+      "loss": 1.1287,
+      "num_input_tokens_seen": 419730868,
+      "step": 10700
+    },
+    {
+      "epoch": 0.5189456342668863,
+      "grad_norm": 0.38283371925354004,
+      "learning_rate": 2.4682208323131852e-05,
+      "loss": 1.1742,
+      "num_input_tokens_seen": 420139208,
+      "step": 10710
+    },
+    {
+      "epoch": 0.5194301773427658,
+      "grad_norm": 0.3874126970767975,
+      "learning_rate": 2.464297730357678e-05,
+      "loss": 1.1564,
+      "num_input_tokens_seen": 420521152,
+      "step": 10720
+    },
+    {
+      "epoch": 0.5199147204186452,
+      "grad_norm": 0.3984072804450989,
+      "learning_rate": 2.460374716335622e-05,
+      "loss": 1.1473,
+      "num_input_tokens_seen": 420929760,
+      "step": 10730
+    },
+    {
+      "epoch": 0.5203992634945247,
+      "grad_norm": 0.3916292190551758,
+      "learning_rate": 2.456451799909263e-05,
+      "loss": 1.1685,
+      "num_input_tokens_seen": 421298444,
+      "step": 10740
+    },
+    {
+      "epoch": 0.5208838065704041,
+      "grad_norm": 0.4141142964363098,
+      "learning_rate": 2.4525289907406068e-05,
+      "loss": 1.139,
+      "num_input_tokens_seen": 421670376,
+      "step": 10750
+    },
+    {
+      "epoch": 0.5213683496462835,
+      "grad_norm": 0.4091172218322754,
+      "learning_rate": 2.448606298491399e-05,
+      "loss": 1.1174,
+      "num_input_tokens_seen": 422062488,
+      "step": 10760
+    },
+    {
+      "epoch": 0.521852892722163,
+      "grad_norm": 0.413764089345932,
+      "learning_rate": 2.4446837328230907e-05,
+      "loss": 1.157,
+      "num_input_tokens_seen": 422469080,
+      "step": 10770
+    },
+    {
+      "epoch": 0.5223374357980425,
+      "grad_norm": 0.35379403829574585,
+      "learning_rate": 2.4407613033968287e-05,
+      "loss": 1.1344,
+      "num_input_tokens_seen": 422867012,
+      "step": 10780
+    },
+    {
+      "epoch": 0.5228219788739219,
+      "grad_norm": 0.3597969114780426,
+      "learning_rate": 2.436839019873418e-05,
+      "loss": 1.1457,
+      "num_input_tokens_seen": 423225644,
+      "step": 10790
+    },
+    {
+      "epoch": 0.5233065219498013,
+      "grad_norm": 0.3982402980327606,
+      "learning_rate": 2.4329168919133062e-05,
+      "loss": 1.1819,
+      "num_input_tokens_seen": 423605548,
+      "step": 10800
+    },
+    {
+      "epoch": 0.5237910650256807,
+      "grad_norm": 0.452061265707016,
+      "learning_rate": 2.4289949291765608e-05,
+      "loss": 1.1438,
+      "num_input_tokens_seen": 423993544,
+      "step": 10810
+    },
+    {
+      "epoch": 0.5242756081015603,
+      "grad_norm": 0.4269421100616455,
+      "learning_rate": 2.4250731413228374e-05,
+      "loss": 1.1333,
+      "num_input_tokens_seen": 424365256,
+      "step": 10820
+    },
+    {
+      "epoch": 0.5247601511774397,
+      "grad_norm": 0.3826684057712555,
+      "learning_rate": 2.421151538011364e-05,
+      "loss": 1.1673,
+      "num_input_tokens_seen": 424768824,
+      "step": 10830
+    },
+    {
+      "epoch": 0.5252446942533191,
+      "grad_norm": 0.3775523602962494,
+      "learning_rate": 2.4172301289009137e-05,
+      "loss": 1.1264,
+      "num_input_tokens_seen": 425175704,
+      "step": 10840
+    },
+    {
+      "epoch": 0.5257292373291985,
+      "grad_norm": 0.4121640920639038,
+      "learning_rate": 2.4133089236497804e-05,
+      "loss": 1.1525,
+      "num_input_tokens_seen": 425568528,
+      "step": 10850
+    },
+    {
+      "epoch": 0.526213780405078,
+      "grad_norm": 0.4088398218154907,
+      "learning_rate": 2.4093879319157572e-05,
+      "loss": 1.1616,
+      "num_input_tokens_seen": 425944588,
+      "step": 10860
+    },
+    {
+      "epoch": 0.5266983234809575,
+      "grad_norm": 0.4086717665195465,
+      "learning_rate": 2.4054671633561094e-05,
+      "loss": 1.1615,
+      "num_input_tokens_seen": 426354932,
+      "step": 10870
+    },
+    {
+      "epoch": 0.5271828665568369,
+      "grad_norm": 0.40191206336021423,
+      "learning_rate": 2.401546627627554e-05,
+      "loss": 1.1574,
+      "num_input_tokens_seen": 426750536,
+      "step": 10880
+    },
+    {
+      "epoch": 0.5276674096327163,
+      "grad_norm": 0.4215683043003082,
+      "learning_rate": 2.3976263343862357e-05,
+      "loss": 1.1275,
+      "num_input_tokens_seen": 427145692,
+      "step": 10890
+    },
+    {
+      "epoch": 0.5281519527085958,
+      "grad_norm": 0.428448885679245,
+      "learning_rate": 2.393706293287698e-05,
+      "loss": 1.1513,
+      "num_input_tokens_seen": 427517468,
+      "step": 10900
+    },
+    {
+      "epoch": 0.5286364957844752,
+      "grad_norm": 0.41464531421661377,
+      "learning_rate": 2.3897865139868685e-05,
+      "loss": 1.2119,
+      "num_input_tokens_seen": 427940524,
+      "step": 10910
+    },
+    {
+      "epoch": 0.5291210388603547,
+      "grad_norm": 0.40263885259628296,
+      "learning_rate": 2.3858670061380267e-05,
+      "loss": 1.1092,
+      "num_input_tokens_seen": 428306984,
+      "step": 10920
+    },
+    {
+      "epoch": 0.5296055819362341,
+      "grad_norm": 0.4086105227470398,
+      "learning_rate": 2.3819477793947825e-05,
+      "loss": 1.147,
+      "num_input_tokens_seen": 428680608,
+      "step": 10930
+    },
+    {
+      "epoch": 0.5300901250121136,
+      "grad_norm": 0.40625572204589844,
+      "learning_rate": 2.378028843410058e-05,
+      "loss": 1.1671,
+      "num_input_tokens_seen": 429089700,
+      "step": 10940
+    },
+    {
+      "epoch": 0.530574668087993,
+      "grad_norm": 0.4443352520465851,
+      "learning_rate": 2.374110207836054e-05,
+      "loss": 1.1634,
+      "num_input_tokens_seen": 429511972,
+      "step": 10950
+    },
+    {
+      "epoch": 0.5310592111638724,
+      "grad_norm": 0.3834231495857239,
+      "learning_rate": 2.3701918823242357e-05,
+      "loss": 1.1406,
+      "num_input_tokens_seen": 429933992,
+      "step": 10960
+    },
+    {
+      "epoch": 0.5315437542397519,
+      "grad_norm": 0.4218290150165558,
+      "learning_rate": 2.366273876525302e-05,
+      "loss": 1.1097,
+      "num_input_tokens_seen": 430335652,
+      "step": 10970
+    },
+    {
+      "epoch": 0.5320282973156314,
+      "grad_norm": 0.38524654507637024,
+      "learning_rate": 2.3623562000891646e-05,
+      "loss": 1.176,
+      "num_input_tokens_seen": 430737500,
+      "step": 10980
+    },
+    {
+      "epoch": 0.5325128403915108,
+      "grad_norm": 0.4242812395095825,
+      "learning_rate": 2.3584388626649246e-05,
+      "loss": 1.1395,
+      "num_input_tokens_seen": 431134616,
+      "step": 10990
+    },
+    {
+      "epoch": 0.5329973834673902,
+      "grad_norm": 0.42048659920692444,
+      "learning_rate": 2.3545218739008483e-05,
+      "loss": 1.1276,
+      "num_input_tokens_seen": 431534440,
+      "step": 11000
+    },
+    {
+      "epoch": 0.5334819265432696,
+      "grad_norm": 0.43740230798721313,
+      "learning_rate": 2.3506052434443436e-05,
+      "loss": 1.1547,
+      "num_input_tokens_seen": 431956964,
+      "step": 11010
+    },
+    {
+      "epoch": 0.5339664696191492,
+      "grad_norm": 0.40744540095329285,
+      "learning_rate": 2.3466889809419342e-05,
+      "loss": 1.1674,
+      "num_input_tokens_seen": 432337932,
+      "step": 11020
+    },
+    {
+      "epoch": 0.5344510126950286,
+      "grad_norm": 0.4600279927253723,
+      "learning_rate": 2.342773096039238e-05,
+      "loss": 1.1226,
+      "num_input_tokens_seen": 432737924,
+      "step": 11030
+    },
+    {
+      "epoch": 0.534935555770908,
+      "grad_norm": 0.4031646251678467,
+      "learning_rate": 2.3388575983809456e-05,
+      "loss": 1.1667,
+      "num_input_tokens_seen": 433161476,
+      "step": 11040
+    },
+    {
+      "epoch": 0.5354200988467874,
+      "grad_norm": 0.3792802691459656,
+      "learning_rate": 2.3349424976107903e-05,
+      "loss": 1.2043,
+      "num_input_tokens_seen": 433547952,
+      "step": 11050
+    },
+    {
+      "epoch": 0.535904641922667,
+      "grad_norm": 0.4132523536682129,
+      "learning_rate": 2.3310278033715285e-05,
+      "loss": 1.1405,
+      "num_input_tokens_seen": 433943164,
+      "step": 11060
+    },
+    {
+      "epoch": 0.5363891849985464,
+      "grad_norm": 0.38873347640037537,
+      "learning_rate": 2.3271135253049183e-05,
+      "loss": 1.1494,
+      "num_input_tokens_seen": 434332620,
+      "step": 11070
+    },
+    {
+      "epoch": 0.5368737280744258,
+      "grad_norm": 0.3987562358379364,
+      "learning_rate": 2.3231996730516884e-05,
+      "loss": 1.1497,
+      "num_input_tokens_seen": 434702280,
+      "step": 11080
+    },
+    {
+      "epoch": 0.5373582711503052,
+      "grad_norm": 0.41692429780960083,
+      "learning_rate": 2.3192862562515226e-05,
+      "loss": 1.1397,
+      "num_input_tokens_seen": 435102792,
+      "step": 11090
+    },
+    {
+      "epoch": 0.5378428142261847,
+      "grad_norm": 0.39025530219078064,
+      "learning_rate": 2.3153732845430302e-05,
+      "loss": 1.122,
+      "num_input_tokens_seen": 435481228,
+      "step": 11100
+    },
+    {
+      "epoch": 0.5383273573020642,
+      "grad_norm": 0.3755995035171509,
+      "learning_rate": 2.3114607675637233e-05,
+      "loss": 1.1926,
+      "num_input_tokens_seen": 435894796,
+      "step": 11110
+    },
+    {
+      "epoch": 0.5388119003779436,
+      "grad_norm": 0.4043687582015991,
+      "learning_rate": 2.3075487149499974e-05,
+      "loss": 1.1358,
+      "num_input_tokens_seen": 436289408,
+      "step": 11120
+    },
+    {
+      "epoch": 0.539296443453823,
+      "grad_norm": 0.3815999925136566,
+      "learning_rate": 2.3036371363371008e-05,
+      "loss": 1.1443,
+      "num_input_tokens_seen": 436692436,
+      "step": 11130
+    },
+    {
+      "epoch": 0.5397809865297025,
+      "grad_norm": 0.38345128297805786,
+      "learning_rate": 2.2997260413591156e-05,
+      "loss": 1.1892,
+      "num_input_tokens_seen": 437100692,
+      "step": 11140
+    },
+    {
+      "epoch": 0.5402655296055819,
+      "grad_norm": 0.4016791880130768,
+      "learning_rate": 2.295815439648934e-05,
+      "loss": 1.1396,
+      "num_input_tokens_seen": 437475216,
+      "step": 11150
+    },
+    {
+      "epoch": 0.5407500726814614,
+      "grad_norm": 0.387517511844635,
+      "learning_rate": 2.2919053408382306e-05,
+      "loss": 1.1204,
+      "num_input_tokens_seen": 437882256,
+      "step": 11160
+    },
+    {
+      "epoch": 0.5412346157573408,
+      "grad_norm": 0.3928276598453522,
+      "learning_rate": 2.287995754557445e-05,
+      "loss": 1.1519,
+      "num_input_tokens_seen": 438262044,
+      "step": 11170
+    },
+    {
+      "epoch": 0.5417191588332203,
+      "grad_norm": 0.4152311682701111,
+      "learning_rate": 2.2840866904357495e-05,
+      "loss": 1.1304,
+      "num_input_tokens_seen": 438668324,
+      "step": 11180
+    },
+    {
+      "epoch": 0.5422037019090997,
+      "grad_norm": 0.4021238684654236,
+      "learning_rate": 2.2801781581010362e-05,
+      "loss": 1.1578,
+      "num_input_tokens_seen": 439061528,
+      "step": 11190
+    },
+    {
+      "epoch": 0.5426882449849791,
+      "grad_norm": 0.3985769748687744,
+      "learning_rate": 2.2762701671798833e-05,
+      "loss": 1.1735,
+      "num_input_tokens_seen": 439445292,
+      "step": 11200
+    },
+    {
+      "epoch": 0.5431727880608586,
+      "grad_norm": 0.4368831515312195,
+      "learning_rate": 2.2723627272975352e-05,
+      "loss": 1.155,
+      "num_input_tokens_seen": 439862688,
+      "step": 11210
+    },
+    {
+      "epoch": 0.5436573311367381,
+      "grad_norm": 0.4021163880825043,
+      "learning_rate": 2.2684558480778833e-05,
+      "loss": 1.1758,
+      "num_input_tokens_seen": 440251824,
+      "step": 11220
+    },
+    {
+      "epoch": 0.5441418742126175,
+      "grad_norm": 0.41846776008605957,
+      "learning_rate": 2.264549539143434e-05,
+      "loss": 1.1374,
+      "num_input_tokens_seen": 440647480,
+      "step": 11230
+    },
+    {
+      "epoch": 0.5446264172884969,
+      "grad_norm": 0.38505181670188904,
+      "learning_rate": 2.2606438101152893e-05,
+      "loss": 1.1702,
+      "num_input_tokens_seen": 441032116,
+      "step": 11240
+    },
+    {
+      "epoch": 0.5451109603643763,
+      "grad_norm": 0.3722958564758301,
+      "learning_rate": 2.2567386706131268e-05,
+      "loss": 1.1441,
+      "num_input_tokens_seen": 441435016,
+      "step": 11250
+    },
+    {
+      "epoch": 0.5455955034402559,
+      "grad_norm": 0.38270628452301025,
+      "learning_rate": 2.2528341302551666e-05,
+      "loss": 1.1893,
+      "num_input_tokens_seen": 441836580,
+      "step": 11260
+    },
+    {
+      "epoch": 0.5460800465161353,
+      "grad_norm": 0.40590032935142517,
+      "learning_rate": 2.2489301986581586e-05,
+      "loss": 1.116,
+      "num_input_tokens_seen": 442232080,
+      "step": 11270
+    },
+    {
+      "epoch": 0.5465645895920147,
+      "grad_norm": 0.36881017684936523,
+      "learning_rate": 2.2450268854373497e-05,
+      "loss": 1.1247,
+      "num_input_tokens_seen": 442637844,
+      "step": 11280
+    },
+    {
+      "epoch": 0.5470491326678941,
+      "grad_norm": 0.38571688532829285,
+      "learning_rate": 2.2411242002064637e-05,
+      "loss": 1.1632,
+      "num_input_tokens_seen": 443047888,
+      "step": 11290
+    },
+    {
+      "epoch": 0.5475336757437737,
+      "grad_norm": 0.402986079454422,
+      "learning_rate": 2.23722215257768e-05,
+      "loss": 1.1353,
+      "num_input_tokens_seen": 443450848,
+      "step": 11300
+    },
+    {
+      "epoch": 0.5480182188196531,
+      "grad_norm": 0.43902143836021423,
+      "learning_rate": 2.2333207521616056e-05,
+      "loss": 1.1409,
+      "num_input_tokens_seen": 443825900,
+      "step": 11310
+    },
+    {
+      "epoch": 0.5485027618955325,
+      "grad_norm": 0.4374961853027344,
+      "learning_rate": 2.2294200085672552e-05,
+      "loss": 1.1557,
+      "num_input_tokens_seen": 444255800,
+      "step": 11320
+    },
+    {
+      "epoch": 0.5489873049714119,
+      "grad_norm": 0.38471776247024536,
+      "learning_rate": 2.225519931402024e-05,
+      "loss": 1.1903,
+      "num_input_tokens_seen": 444647372,
+      "step": 11330
+    },
+    {
+      "epoch": 0.5494718480472914,
+      "grad_norm": 0.3870150148868561,
+      "learning_rate": 2.2216205302716656e-05,
+      "loss": 1.165,
+      "num_input_tokens_seen": 445028328,
+      "step": 11340
+    },
+    {
+      "epoch": 0.5499563911231709,
+      "grad_norm": 0.3954487144947052,
+      "learning_rate": 2.217721814780272e-05,
+      "loss": 1.1201,
+      "num_input_tokens_seen": 445435648,
+      "step": 11350
+    },
+    {
+      "epoch": 0.5504409341990503,
+      "grad_norm": 0.4321858286857605,
+      "learning_rate": 2.2138237945302412e-05,
+      "loss": 1.0994,
+      "num_input_tokens_seen": 445822760,
+      "step": 11360
+    },
+    {
+      "epoch": 0.5509254772749297,
+      "grad_norm": 0.3976816236972809,
+      "learning_rate": 2.2099264791222643e-05,
+      "loss": 1.1399,
+      "num_input_tokens_seen": 446231820,
+      "step": 11370
+    },
+    {
+      "epoch": 0.5514100203508092,
+      "grad_norm": 0.3802758455276489,
+      "learning_rate": 2.2060298781552927e-05,
+      "loss": 1.1505,
+      "num_input_tokens_seen": 446644088,
+      "step": 11380
+    },
+    {
+      "epoch": 0.5518945634266886,
+      "grad_norm": 0.39267027378082275,
+      "learning_rate": 2.2021340012265177e-05,
+      "loss": 1.1583,
+      "num_input_tokens_seen": 447026360,
+      "step": 11390
+    },
+    {
+      "epoch": 0.552379106502568,
+      "grad_norm": 0.41610804200172424,
+      "learning_rate": 2.198238857931352e-05,
+      "loss": 1.1108,
+      "num_input_tokens_seen": 447431376,
+      "step": 11400
+    },
+    {
+      "epoch": 0.5528636495784475,
+      "grad_norm": 0.40714430809020996,
+      "learning_rate": 2.1943444578633957e-05,
+      "loss": 1.1341,
+      "num_input_tokens_seen": 447810844,
+      "step": 11410
+    },
+    {
+      "epoch": 0.553348192654327,
+      "grad_norm": 0.3614059090614319,
+      "learning_rate": 2.1904508106144208e-05,
+      "loss": 1.165,
+      "num_input_tokens_seen": 448200712,
+      "step": 11420
+    },
+    {
+      "epoch": 0.5538327357302064,
+      "grad_norm": 0.3718690574169159,
+      "learning_rate": 2.1865579257743475e-05,
+      "loss": 1.0994,
+      "num_input_tokens_seen": 448585024,
+      "step": 11430
+    },
+    {
+      "epoch": 0.5543172788060858,
+      "grad_norm": 0.4221544861793518,
+      "learning_rate": 2.1826658129312133e-05,
+      "loss": 1.1858,
+      "num_input_tokens_seen": 448991036,
+      "step": 11440
+    },
+    {
+      "epoch": 0.5548018218819653,
+      "grad_norm": 0.40332892537117004,
+      "learning_rate": 2.178774481671159e-05,
+      "loss": 1.1761,
+      "num_input_tokens_seen": 449381496,
+      "step": 11450
+    },
+    {
+      "epoch": 0.5552863649578448,
+      "grad_norm": 0.4298264682292938,
+      "learning_rate": 2.174883941578397e-05,
+      "loss": 1.1593,
+      "num_input_tokens_seen": 449808272,
+      "step": 11460
+    },
+    {
+      "epoch": 0.5557709080337242,
+      "grad_norm": 0.4092737138271332,
+      "learning_rate": 2.1709942022351924e-05,
+      "loss": 1.1434,
+      "num_input_tokens_seen": 450187828,
+      "step": 11470
+    },
+    {
+      "epoch": 0.5562554511096036,
+      "grad_norm": 0.4071583151817322,
+      "learning_rate": 2.1671052732218392e-05,
+      "loss": 1.1578,
+      "num_input_tokens_seen": 450578040,
+      "step": 11480
+    },
+    {
+      "epoch": 0.5567399941854831,
+      "grad_norm": 0.4042060673236847,
+      "learning_rate": 2.1632171641166326e-05,
+      "loss": 1.114,
+      "num_input_tokens_seen": 450979104,
+      "step": 11490
+    },
+    {
+      "epoch": 0.5572245372613626,
+      "grad_norm": 0.41152459383010864,
+      "learning_rate": 2.1593298844958526e-05,
+      "loss": 1.1724,
+      "num_input_tokens_seen": 451369344,
+      "step": 11500
+    },
+    {
+      "epoch": 0.557709080337242,
+      "grad_norm": 0.3623616099357605,
+      "learning_rate": 2.1554434439337326e-05,
+      "loss": 1.1765,
+      "num_input_tokens_seen": 451771544,
+      "step": 11510
+    },
+    {
+      "epoch": 0.5581936234131214,
+      "grad_norm": 0.41300123929977417,
+      "learning_rate": 2.15155785200244e-05,
+      "loss": 1.1843,
+      "num_input_tokens_seen": 452152352,
+      "step": 11520
+    },
+    {
+      "epoch": 0.5586781664890009,
+      "grad_norm": 0.41391250491142273,
+      "learning_rate": 2.147673118272054e-05,
+      "loss": 1.141,
+      "num_input_tokens_seen": 452586876,
+      "step": 11530
+    },
+    {
+      "epoch": 0.5591627095648803,
+      "grad_norm": 0.37428995966911316,
+      "learning_rate": 2.1437892523105378e-05,
+      "loss": 1.1809,
+      "num_input_tokens_seen": 452992560,
+      "step": 11540
+    },
+    {
+      "epoch": 0.5596472526407598,
+      "grad_norm": 0.42412886023521423,
+      "learning_rate": 2.1399062636837197e-05,
+      "loss": 1.1461,
+      "num_input_tokens_seen": 453342104,
+      "step": 11550
+    },
+    {
+      "epoch": 0.5601317957166392,
+      "grad_norm": 0.3886395990848541,
+      "learning_rate": 2.1360241619552652e-05,
+      "loss": 1.1342,
+      "num_input_tokens_seen": 453754860,
+      "step": 11560
+    },
+    {
+      "epoch": 0.5606163387925187,
+      "grad_norm": 0.3729265332221985,
+      "learning_rate": 2.1321429566866542e-05,
+      "loss": 1.1745,
+      "num_input_tokens_seen": 454158212,
+      "step": 11570
+    },
+    {
+      "epoch": 0.5611008818683981,
+      "grad_norm": 0.3990989625453949,
+      "learning_rate": 2.1282626574371635e-05,
+      "loss": 1.1643,
+      "num_input_tokens_seen": 454533420,
+      "step": 11580
+    },
+    {
+      "epoch": 0.5615854249442775,
+      "grad_norm": 0.39453041553497314,
+      "learning_rate": 2.124383273763834e-05,
+      "loss": 1.1723,
+      "num_input_tokens_seen": 454918724,
+      "step": 11590
+    },
+    {
+      "epoch": 0.562069968020157,
+      "grad_norm": 0.3955458700656891,
+      "learning_rate": 2.120504815221452e-05,
+      "loss": 1.1748,
+      "num_input_tokens_seen": 455320488,
+      "step": 11600
+    },
+    {
+      "epoch": 0.5625545110960365,
+      "grad_norm": 0.3768325448036194,
+      "learning_rate": 2.1166272913625273e-05,
+      "loss": 1.1634,
+      "num_input_tokens_seen": 455716484,
+      "step": 11610
+    },
+    {
+      "epoch": 0.5630390541719159,
+      "grad_norm": 0.3783731758594513,
+      "learning_rate": 2.1127507117372657e-05,
+      "loss": 1.1953,
+      "num_input_tokens_seen": 456099764,
+      "step": 11620
+    },
+    {
+      "epoch": 0.5635235972477953,
+      "grad_norm": 0.41152095794677734,
+      "learning_rate": 2.1088750858935495e-05,
+      "loss": 1.1831,
+      "num_input_tokens_seen": 456506480,
+      "step": 11630
+    },
+    {
+      "epoch": 0.5640081403236747,
+      "grad_norm": 0.414093017578125,
+      "learning_rate": 2.1050004233769094e-05,
+      "loss": 1.1475,
+      "num_input_tokens_seen": 456916444,
+      "step": 11640
+    },
+    {
+      "epoch": 0.5644926833995543,
+      "grad_norm": 0.41902777552604675,
+      "learning_rate": 2.1011267337305036e-05,
+      "loss": 1.1574,
+      "num_input_tokens_seen": 457314824,
+      "step": 11650
+    },
+    {
+      "epoch": 0.5649772264754337,
+      "grad_norm": 0.40867382287979126,
+      "learning_rate": 2.0972540264950976e-05,
+      "loss": 1.0926,
+      "num_input_tokens_seen": 457711980,
+      "step": 11660
+    },
+    {
+      "epoch": 0.5654617695513131,
+      "grad_norm": 0.3919362425804138,
+      "learning_rate": 2.0933823112090322e-05,
+      "loss": 1.2073,
+      "num_input_tokens_seen": 458128016,
+      "step": 11670
+    },
+    {
+      "epoch": 0.5659463126271925,
+      "grad_norm": 0.3909580707550049,
+      "learning_rate": 2.0895115974082106e-05,
+      "loss": 1.1525,
+      "num_input_tokens_seen": 458509664,
+      "step": 11680
+    },
+    {
+      "epoch": 0.5664308557030721,
+      "grad_norm": 0.38012978434562683,
+      "learning_rate": 2.0856418946260643e-05,
+      "loss": 1.146,
+      "num_input_tokens_seen": 458918900,
+      "step": 11690
+    },
+    {
+      "epoch": 0.5669153987789515,
+      "grad_norm": 0.378833144903183,
+      "learning_rate": 2.0817732123935364e-05,
+      "loss": 1.1811,
+      "num_input_tokens_seen": 459322936,
+      "step": 11700
+    },
+    {
+      "epoch": 0.5673999418548309,
+      "grad_norm": 0.37148529291152954,
+      "learning_rate": 2.0779055602390583e-05,
+      "loss": 1.1755,
+      "num_input_tokens_seen": 459735016,
+      "step": 11710
+    },
+    {
+      "epoch": 0.5678844849307103,
+      "grad_norm": 0.3815120458602905,
+      "learning_rate": 2.0740389476885223e-05,
+      "loss": 1.1825,
+      "num_input_tokens_seen": 460131788,
+      "step": 11720
+    },
+    {
+      "epoch": 0.5683690280065898,
+      "grad_norm": 0.3902805745601654,
+      "learning_rate": 2.070173384265261e-05,
+      "loss": 1.1369,
+      "num_input_tokens_seen": 460515608,
+      "step": 11730
+    },
+    {
+      "epoch": 0.5688535710824693,
+      "grad_norm": 0.40333831310272217,
+      "learning_rate": 2.066308879490023e-05,
+      "loss": 1.1596,
+      "num_input_tokens_seen": 460897436,
+      "step": 11740
+    },
+    {
+      "epoch": 0.5693381141583487,
+      "grad_norm": 0.39738982915878296,
+      "learning_rate": 2.0624454428809484e-05,
+      "loss": 1.182,
+      "num_input_tokens_seen": 461288352,
+      "step": 11750
+    },
+    {
+      "epoch": 0.5698226572342281,
+      "grad_norm": 0.38377317786216736,
+      "learning_rate": 2.0585830839535487e-05,
+      "loss": 1.1371,
+      "num_input_tokens_seen": 461665964,
+      "step": 11760
+    },
+    {
+      "epoch": 0.5703072003101076,
+      "grad_norm": 0.40340209007263184,
+      "learning_rate": 2.054721812220678e-05,
+      "loss": 1.1388,
+      "num_input_tokens_seen": 462099048,
+      "step": 11770
+    },
+    {
+      "epoch": 0.570791743385987,
+      "grad_norm": 0.39503028988838196,
+      "learning_rate": 2.0508616371925156e-05,
+      "loss": 1.1972,
+      "num_input_tokens_seen": 462477648,
+      "step": 11780
+    },
+    {
+      "epoch": 0.5712762864618665,
+      "grad_norm": 0.3878142237663269,
+      "learning_rate": 2.0470025683765392e-05,
+      "loss": 1.1394,
+      "num_input_tokens_seen": 462857744,
+      "step": 11790
+    },
+    {
+      "epoch": 0.5717608295377459,
+      "grad_norm": 0.430828720331192,
+      "learning_rate": 2.043144615277499e-05,
+      "loss": 1.1204,
+      "num_input_tokens_seen": 463231476,
+      "step": 11800
+    },
+    {
+      "epoch": 0.5722453726136254,
+      "grad_norm": 0.38779589533805847,
+      "learning_rate": 2.039287787397402e-05,
+      "loss": 1.1468,
+      "num_input_tokens_seen": 463608072,
+      "step": 11810
+    },
+    {
+      "epoch": 0.5727299156895048,
+      "grad_norm": 0.40600141882896423,
+      "learning_rate": 2.0354320942354804e-05,
+      "loss": 1.1389,
+      "num_input_tokens_seen": 464007424,
+      "step": 11820
+    },
+    {
+      "epoch": 0.5732144587653842,
+      "grad_norm": 0.3781258463859558,
+      "learning_rate": 2.0315775452881708e-05,
+      "loss": 1.172,
+      "num_input_tokens_seen": 464436116,
+      "step": 11830
+    },
+    {
+      "epoch": 0.5736990018412637,
+      "grad_norm": 0.41185423731803894,
+      "learning_rate": 2.027724150049096e-05,
+      "loss": 1.1674,
+      "num_input_tokens_seen": 464834636,
+      "step": 11840
+    },
+    {
+      "epoch": 0.5741835449171432,
+      "grad_norm": 0.39568862318992615,
+      "learning_rate": 2.0238719180090323e-05,
+      "loss": 1.1769,
+      "num_input_tokens_seen": 465219480,
+      "step": 11850
+    },
+    {
+      "epoch": 0.5746680879930226,
+      "grad_norm": 0.3740871846675873,
+      "learning_rate": 2.0200208586558954e-05,
+      "loss": 1.1428,
+      "num_input_tokens_seen": 465608864,
+      "step": 11860
+    },
+    {
+      "epoch": 0.575152631068902,
+      "grad_norm": 0.39056453108787537,
+      "learning_rate": 2.0161709814747102e-05,
+      "loss": 1.1356,
+      "num_input_tokens_seen": 466002328,
+      "step": 11870
+    },
+    {
+      "epoch": 0.5756371741447814,
+      "grad_norm": 0.46836405992507935,
+      "learning_rate": 2.012322295947589e-05,
+      "loss": 1.1183,
+      "num_input_tokens_seen": 466418136,
+      "step": 11880
+    },
+    {
+      "epoch": 0.576121717220661,
+      "grad_norm": 0.41011425852775574,
+      "learning_rate": 2.0084748115537126e-05,
+      "loss": 1.1868,
+      "num_input_tokens_seen": 466816508,
+      "step": 11890
+    },
+    {
+      "epoch": 0.5766062602965404,
+      "grad_norm": 0.4022364914417267,
+      "learning_rate": 2.0046285377692998e-05,
+      "loss": 1.1622,
+      "num_input_tokens_seen": 467208132,
+      "step": 11900
+    },
+    {
+      "epoch": 0.5770908033724198,
+      "grad_norm": 0.4446437656879425,
+      "learning_rate": 2.0007834840675905e-05,
+      "loss": 1.2062,
+      "num_input_tokens_seen": 467607752,
+      "step": 11910
+    },
+    {
+      "epoch": 0.5775753464482992,
+      "grad_norm": 0.4171381890773773,
+      "learning_rate": 1.9969396599188177e-05,
+      "loss": 1.1675,
+      "num_input_tokens_seen": 467989092,
+      "step": 11920
+    },
+    {
+      "epoch": 0.5780598895241787,
+      "grad_norm": 0.44026637077331543,
+      "learning_rate": 1.993097074790186e-05,
+      "loss": 1.1417,
+      "num_input_tokens_seen": 468356440,
+      "step": 11930
+    },
+    {
+      "epoch": 0.5785444326000582,
+      "grad_norm": 0.3875192105770111,
+      "learning_rate": 1.989255738145851e-05,
+      "loss": 1.1676,
+      "num_input_tokens_seen": 468736536,
+      "step": 11940
+    },
+    {
+      "epoch": 0.5790289756759376,
+      "grad_norm": 0.37714606523513794,
+      "learning_rate": 1.9854156594468905e-05,
+      "loss": 1.1685,
+      "num_input_tokens_seen": 469122032,
+      "step": 11950
+    },
+    {
+      "epoch": 0.579513518751817,
+      "grad_norm": 0.3938728868961334,
+      "learning_rate": 1.9815768481512837e-05,
+      "loss": 1.1166,
+      "num_input_tokens_seen": 469516476,
+      "step": 11960
+    },
+    {
+      "epoch": 0.5799980618276965,
+      "grad_norm": 0.4008789360523224,
+      "learning_rate": 1.9777393137138916e-05,
+      "loss": 1.1795,
+      "num_input_tokens_seen": 469938440,
+      "step": 11970
+    },
+    {
+      "epoch": 0.580482604903576,
+      "grad_norm": 0.3709718585014343,
+      "learning_rate": 1.9739030655864263e-05,
+      "loss": 1.1366,
+      "num_input_tokens_seen": 470313748,
+      "step": 11980
+    },
+    {
+      "epoch": 0.5809671479794554,
+      "grad_norm": 0.4143054485321045,
+      "learning_rate": 1.9700681132174356e-05,
+      "loss": 1.1339,
+      "num_input_tokens_seen": 470678456,
+      "step": 11990
+    },
+    {
+      "epoch": 0.5814516910553348,
+      "grad_norm": 0.3943856656551361,
+      "learning_rate": 1.9662344660522726e-05,
+      "loss": 1.1755,
+      "num_input_tokens_seen": 471101508,
+      "step": 12000
+    },
+    {
+      "epoch": 0.5814516910553348,
+      "eval_loss": 1.1437398195266724,
+      "eval_runtime": 7.1807,
+      "eval_samples_per_second": 20.889,
+      "eval_steps_per_second": 2.646,
+      "num_input_tokens_seen": 471101508,
+      "step": 12000
+    },
+    {
+      "epoch": 0.5819362341312143,
+      "grad_norm": 0.40113458037376404,
+      "learning_rate": 1.9624021335330767e-05,
+      "loss": 1.1483,
+      "num_input_tokens_seen": 471482556,
+      "step": 12010
+    },
+    {
+      "epoch": 0.5824207772070937,
+      "grad_norm": 0.3723471760749817,
+      "learning_rate": 1.9585711250987515e-05,
+      "loss": 1.0867,
+      "num_input_tokens_seen": 471888220,
+      "step": 12020
+    },
+    {
+      "epoch": 0.5829053202829731,
+      "grad_norm": 0.3816375732421875,
+      "learning_rate": 1.9547414501849363e-05,
+      "loss": 1.1351,
+      "num_input_tokens_seen": 472248060,
+      "step": 12030
+    },
+    {
+      "epoch": 0.5833898633588526,
+      "grad_norm": 0.41664794087409973,
+      "learning_rate": 1.9509131182239875e-05,
+      "loss": 1.1705,
+      "num_input_tokens_seen": 472639804,
+      "step": 12040
+    },
+    {
+      "epoch": 0.5838744064347321,
+      "grad_norm": 0.3829415440559387,
+      "learning_rate": 1.9470861386449546e-05,
+      "loss": 1.1315,
+      "num_input_tokens_seen": 473018176,
+      "step": 12050
+    },
+    {
+      "epoch": 0.5843589495106115,
+      "grad_norm": 0.408346027135849,
+      "learning_rate": 1.9432605208735543e-05,
+      "loss": 1.1286,
+      "num_input_tokens_seen": 473403200,
+      "step": 12060
+    },
+    {
+      "epoch": 0.5848434925864909,
+      "grad_norm": 0.3734830915927887,
+      "learning_rate": 1.9394362743321516e-05,
+      "loss": 1.1347,
+      "num_input_tokens_seen": 473796676,
+      "step": 12070
+    },
+    {
+      "epoch": 0.5853280356623703,
+      "grad_norm": 0.3995228111743927,
+      "learning_rate": 1.9356134084397305e-05,
+      "loss": 1.159,
+      "num_input_tokens_seen": 474192924,
+      "step": 12080
+    },
+    {
+      "epoch": 0.5858125787382499,
+      "grad_norm": 0.38744309544563293,
+      "learning_rate": 1.9317919326118793e-05,
+      "loss": 1.1302,
+      "num_input_tokens_seen": 474587284,
+      "step": 12090
+    },
+    {
+      "epoch": 0.5862971218141293,
+      "grad_norm": 0.40833401679992676,
+      "learning_rate": 1.9279718562607595e-05,
+      "loss": 1.1293,
+      "num_input_tokens_seen": 474984376,
+      "step": 12100
+    },
+    {
+      "epoch": 0.5867816648900087,
+      "grad_norm": 0.40676310658454895,
+      "learning_rate": 1.9241531887950853e-05,
+      "loss": 1.1125,
+      "num_input_tokens_seen": 475382208,
+      "step": 12110
+    },
+    {
+      "epoch": 0.5872662079658881,
+      "grad_norm": 0.4175487756729126,
+      "learning_rate": 1.9203359396201038e-05,
+      "loss": 1.1302,
+      "num_input_tokens_seen": 475803188,
+      "step": 12120
+    },
+    {
+      "epoch": 0.5877507510417677,
+      "grad_norm": 0.39633074402809143,
+      "learning_rate": 1.9165201181375663e-05,
+      "loss": 1.1573,
+      "num_input_tokens_seen": 476200608,
+      "step": 12130
+    },
+    {
+      "epoch": 0.5882352941176471,
+      "grad_norm": 0.38707002997398376,
+      "learning_rate": 1.9127057337457077e-05,
+      "loss": 1.1641,
+      "num_input_tokens_seen": 476617248,
+      "step": 12140
+    },
+    {
+      "epoch": 0.5887198371935265,
+      "grad_norm": 0.38113346695899963,
+      "learning_rate": 1.908892795839226e-05,
+      "loss": 1.1381,
+      "num_input_tokens_seen": 476978804,
+      "step": 12150
+    },
+    {
+      "epoch": 0.5892043802694059,
+      "grad_norm": 0.40942734479904175,
+      "learning_rate": 1.905081313809253e-05,
+      "loss": 1.1546,
+      "num_input_tokens_seen": 477371880,
+      "step": 12160
+    },
+    {
+      "epoch": 0.5896889233452854,
+      "grad_norm": 0.3737110495567322,
+      "learning_rate": 1.9012712970433384e-05,
+      "loss": 1.1463,
+      "num_input_tokens_seen": 477751464,
+      "step": 12170
+    },
+    {
+      "epoch": 0.5901734664211649,
+      "grad_norm": 0.37110334634780884,
+      "learning_rate": 1.8974627549254205e-05,
+      "loss": 1.1607,
+      "num_input_tokens_seen": 478117284,
+      "step": 12180
+    },
+    {
+      "epoch": 0.5906580094970443,
+      "grad_norm": 0.38362741470336914,
+      "learning_rate": 1.8936556968358047e-05,
+      "loss": 1.1784,
+      "num_input_tokens_seen": 478503476,
+      "step": 12190
+    },
+    {
+      "epoch": 0.5911425525729237,
+      "grad_norm": 0.4000930190086365,
+      "learning_rate": 1.8898501321511445e-05,
+      "loss": 1.0922,
+      "num_input_tokens_seen": 478897364,
+      "step": 12200
+    },
+    {
+      "epoch": 0.5916270956488032,
+      "grad_norm": 0.39831140637397766,
+      "learning_rate": 1.8860460702444126e-05,
+      "loss": 1.1765,
+      "num_input_tokens_seen": 479272824,
+      "step": 12210
+    },
+    {
+      "epoch": 0.5921116387246826,
+      "grad_norm": 0.37732118368148804,
+      "learning_rate": 1.8822435204848827e-05,
+      "loss": 1.1729,
+      "num_input_tokens_seen": 479653192,
+      "step": 12220
+    },
+    {
+      "epoch": 0.5925961818005621,
+      "grad_norm": 0.3948591649532318,
+      "learning_rate": 1.8784424922381015e-05,
+      "loss": 1.136,
+      "num_input_tokens_seen": 480072752,
+      "step": 12230
+    },
+    {
+      "epoch": 0.5930807248764415,
+      "grad_norm": 0.3917677700519562,
+      "learning_rate": 1.8746429948658693e-05,
+      "loss": 1.1598,
+      "num_input_tokens_seen": 480459552,
+      "step": 12240
+    },
+    {
+      "epoch": 0.593565267952321,
+      "grad_norm": 0.4181142747402191,
+      "learning_rate": 1.8708450377262178e-05,
+      "loss": 1.109,
+      "num_input_tokens_seen": 480856544,
+      "step": 12250
+    },
+    {
+      "epoch": 0.5940498110282004,
+      "grad_norm": 0.4221828281879425,
+      "learning_rate": 1.867048630173381e-05,
+      "loss": 1.1264,
+      "num_input_tokens_seen": 481254676,
+      "step": 12260
+    },
+    {
+      "epoch": 0.5945343541040798,
+      "grad_norm": 0.3877246379852295,
+      "learning_rate": 1.8632537815577812e-05,
+      "loss": 1.1524,
+      "num_input_tokens_seen": 481664936,
+      "step": 12270
+    },
+    {
+      "epoch": 0.5950188971799593,
+      "grad_norm": 0.41641271114349365,
+      "learning_rate": 1.859460501225998e-05,
+      "loss": 1.1492,
+      "num_input_tokens_seen": 482076064,
+      "step": 12280
+    },
+    {
+      "epoch": 0.5955034402558388,
+      "grad_norm": 0.38876596093177795,
+      "learning_rate": 1.8556687985207473e-05,
+      "loss": 1.1841,
+      "num_input_tokens_seen": 482481220,
+      "step": 12290
+    },
+    {
+      "epoch": 0.5959879833317182,
+      "grad_norm": 0.372016966342926,
+      "learning_rate": 1.851878682780864e-05,
+      "loss": 1.1443,
+      "num_input_tokens_seen": 482858776,
+      "step": 12300
+    },
+    {
+      "epoch": 0.5964725264075976,
+      "grad_norm": 0.4116594195365906,
+      "learning_rate": 1.84809016334127e-05,
+      "loss": 1.168,
+      "num_input_tokens_seen": 483260116,
+      "step": 12310
+    },
+    {
+      "epoch": 0.596957069483477,
+      "grad_norm": 0.42143404483795166,
+      "learning_rate": 1.8443032495329564e-05,
+      "loss": 1.1707,
+      "num_input_tokens_seen": 483647520,
+      "step": 12320
+    },
+    {
+      "epoch": 0.5974416125593566,
+      "grad_norm": 0.41614919900894165,
+      "learning_rate": 1.8405179506829622e-05,
+      "loss": 1.1575,
+      "num_input_tokens_seen": 484057976,
+      "step": 12330
+    },
+    {
+      "epoch": 0.597926155635236,
+      "grad_norm": 0.39331093430519104,
+      "learning_rate": 1.836734276114346e-05,
+      "loss": 1.1582,
+      "num_input_tokens_seen": 484441132,
+      "step": 12340
+    },
+    {
+      "epoch": 0.5984106987111154,
+      "grad_norm": 0.41087356209754944,
+      "learning_rate": 1.832952235146166e-05,
+      "loss": 1.1851,
+      "num_input_tokens_seen": 484824128,
+      "step": 12350
+    },
+    {
+      "epoch": 0.5988952417869948,
+      "grad_norm": 0.3742915093898773,
+      "learning_rate": 1.829171837093459e-05,
+      "loss": 1.1018,
+      "num_input_tokens_seen": 485241940,
+      "step": 12360
+    },
+    {
+      "epoch": 0.5993797848628744,
+      "grad_norm": 0.3956594169139862,
+      "learning_rate": 1.8253930912672136e-05,
+      "loss": 1.1769,
+      "num_input_tokens_seen": 485653000,
+      "step": 12370
+    },
+    {
+      "epoch": 0.5998643279387538,
+      "grad_norm": 0.4172208905220032,
+      "learning_rate": 1.8216160069743498e-05,
+      "loss": 1.1215,
+      "num_input_tokens_seen": 486031148,
+      "step": 12380
+    },
+    {
+      "epoch": 0.6003488710146332,
+      "grad_norm": 0.3961176872253418,
+      "learning_rate": 1.8178405935176933e-05,
+      "loss": 1.1929,
+      "num_input_tokens_seen": 486411448,
+      "step": 12390
+    },
+    {
+      "epoch": 0.6008334140905126,
+      "grad_norm": 0.3706035315990448,
+      "learning_rate": 1.8140668601959593e-05,
+      "loss": 1.0883,
+      "num_input_tokens_seen": 486792408,
+      "step": 12400
+    },
+    {
+      "epoch": 0.6013179571663921,
+      "grad_norm": 0.3640589416027069,
+      "learning_rate": 1.81029481630372e-05,
+      "loss": 1.1478,
+      "num_input_tokens_seen": 487182316,
+      "step": 12410
+    },
+    {
+      "epoch": 0.6018025002422716,
+      "grad_norm": 0.4026361107826233,
+      "learning_rate": 1.806524471131388e-05,
+      "loss": 1.1486,
+      "num_input_tokens_seen": 487604672,
+      "step": 12420
+    },
+    {
+      "epoch": 0.602287043318151,
+      "grad_norm": 0.41077134013175964,
+      "learning_rate": 1.8027558339651936e-05,
+      "loss": 1.1469,
+      "num_input_tokens_seen": 487986272,
+      "step": 12430
+    },
+    {
+      "epoch": 0.6027715863940304,
+      "grad_norm": 0.394106924533844,
+      "learning_rate": 1.7989889140871583e-05,
+      "loss": 1.0973,
+      "num_input_tokens_seen": 488373412,
+      "step": 12440
+    },
+    {
+      "epoch": 0.6032561294699099,
+      "grad_norm": 0.3956287205219269,
+      "learning_rate": 1.795223720775076e-05,
+      "loss": 1.118,
+      "num_input_tokens_seen": 488785752,
+      "step": 12450
+    },
+    {
+      "epoch": 0.6037406725457893,
+      "grad_norm": 0.3825465142726898,
+      "learning_rate": 1.791460263302487e-05,
+      "loss": 1.1358,
+      "num_input_tokens_seen": 489205628,
+      "step": 12460
+    },
+    {
+      "epoch": 0.6042252156216688,
+      "grad_norm": 0.3835325539112091,
+      "learning_rate": 1.7876985509386547e-05,
+      "loss": 1.1225,
+      "num_input_tokens_seen": 489595928,
+      "step": 12470
+    },
+    {
+      "epoch": 0.6047097586975482,
+      "grad_norm": 0.3921869397163391,
+      "learning_rate": 1.7839385929485482e-05,
+      "loss": 1.1232,
+      "num_input_tokens_seen": 490009524,
+      "step": 12480
+    },
+    {
+      "epoch": 0.6051943017734277,
+      "grad_norm": 0.4052605628967285,
+      "learning_rate": 1.7801803985928117e-05,
+      "loss": 1.1823,
+      "num_input_tokens_seen": 490436288,
+      "step": 12490
+    },
+    {
+      "epoch": 0.6056788448493071,
+      "grad_norm": 0.37396615743637085,
+      "learning_rate": 1.7764239771277477e-05,
+      "loss": 1.1319,
+      "num_input_tokens_seen": 490820096,
+      "step": 12500
+    },
+    {
+      "epoch": 0.6061633879251865,
+      "grad_norm": 0.3906390368938446,
+      "learning_rate": 1.772669337805292e-05,
+      "loss": 1.1324,
+      "num_input_tokens_seen": 491197804,
+      "step": 12510
+    },
+    {
+      "epoch": 0.606647931001066,
+      "grad_norm": 0.3880655765533447,
+      "learning_rate": 1.768916489872991e-05,
+      "loss": 1.1389,
+      "num_input_tokens_seen": 491590844,
+      "step": 12520
+    },
+    {
+      "epoch": 0.6071324740769455,
+      "grad_norm": 0.42076969146728516,
+      "learning_rate": 1.765165442573979e-05,
+      "loss": 1.1312,
+      "num_input_tokens_seen": 491987104,
+      "step": 12530
+    },
+    {
+      "epoch": 0.6076170171528249,
+      "grad_norm": 0.4096836745738983,
+      "learning_rate": 1.7614162051469545e-05,
+      "loss": 1.1339,
+      "num_input_tokens_seen": 492387748,
+      "step": 12540
+    },
+    {
+      "epoch": 0.6081015602287043,
+      "grad_norm": 0.3795408606529236,
+      "learning_rate": 1.7576687868261587e-05,
+      "loss": 1.1246,
+      "num_input_tokens_seen": 492784488,
+      "step": 12550
+    },
+    {
+      "epoch": 0.6085861033045837,
+      "grad_norm": 0.38204097747802734,
+      "learning_rate": 1.7539231968413546e-05,
+      "loss": 1.1139,
+      "num_input_tokens_seen": 493156420,
+      "step": 12560
+    },
+    {
+      "epoch": 0.6090706463804633,
+      "grad_norm": 0.4025273621082306,
+      "learning_rate": 1.7501794444177975e-05,
+      "loss": 1.1592,
+      "num_input_tokens_seen": 493529324,
+      "step": 12570
+    },
+    {
+      "epoch": 0.6095551894563427,
+      "grad_norm": 0.39617040753364563,
+      "learning_rate": 1.746437538776222e-05,
+      "loss": 1.1537,
+      "num_input_tokens_seen": 493929468,
+      "step": 12580
+    },
+    {
+      "epoch": 0.6100397325322221,
+      "grad_norm": 0.39346054196357727,
+      "learning_rate": 1.742697489132811e-05,
+      "loss": 1.1642,
+      "num_input_tokens_seen": 494356540,
+      "step": 12590
+    },
+    {
+      "epoch": 0.6105242756081015,
+      "grad_norm": 0.4220488369464874,
+      "learning_rate": 1.738959304699176e-05,
+      "loss": 1.1657,
+      "num_input_tokens_seen": 494737364,
+      "step": 12600
+    },
+    {
+      "epoch": 0.611008818683981,
+      "grad_norm": 0.38811561465263367,
+      "learning_rate": 1.735222994682336e-05,
+      "loss": 1.131,
+      "num_input_tokens_seen": 495145924,
+      "step": 12610
+    },
+    {
+      "epoch": 0.6114933617598605,
+      "grad_norm": 0.3839203417301178,
+      "learning_rate": 1.7314885682846925e-05,
+      "loss": 1.1502,
+      "num_input_tokens_seen": 495543252,
+      "step": 12620
+    },
+    {
+      "epoch": 0.6119779048357399,
+      "grad_norm": 0.4147859811782837,
+      "learning_rate": 1.7277560347040094e-05,
+      "loss": 1.219,
+      "num_input_tokens_seen": 495924020,
+      "step": 12630
+    },
+    {
+      "epoch": 0.6124624479116193,
+      "grad_norm": 0.3909815549850464,
+      "learning_rate": 1.7240254031333862e-05,
+      "loss": 1.1642,
+      "num_input_tokens_seen": 496317616,
+      "step": 12640
+    },
+    {
+      "epoch": 0.6129469909874988,
+      "grad_norm": 0.3842679560184479,
+      "learning_rate": 1.720296682761238e-05,
+      "loss": 1.142,
+      "num_input_tokens_seen": 496696256,
+      "step": 12650
+    },
+    {
+      "epoch": 0.6134315340633782,
+      "grad_norm": 0.4022013545036316,
+      "learning_rate": 1.716569882771274e-05,
+      "loss": 1.1284,
+      "num_input_tokens_seen": 497103872,
+      "step": 12660
+    },
+    {
+      "epoch": 0.6139160771392577,
+      "grad_norm": 0.39619848132133484,
+      "learning_rate": 1.7128450123424746e-05,
+      "loss": 1.1532,
+      "num_input_tokens_seen": 497505888,
+      "step": 12670
+    },
+    {
+      "epoch": 0.6144006202151371,
+      "grad_norm": 0.3860088586807251,
+      "learning_rate": 1.709122080649064e-05,
+      "loss": 1.1758,
+      "num_input_tokens_seen": 497903692,
+      "step": 12680
+    },
+    {
+      "epoch": 0.6148851632910166,
+      "grad_norm": 0.377780020236969,
+      "learning_rate": 1.705401096860496e-05,
+      "loss": 1.1119,
+      "num_input_tokens_seen": 498287500,
+      "step": 12690
+    },
+    {
+      "epoch": 0.615369706366896,
+      "grad_norm": 0.3980233371257782,
+      "learning_rate": 1.7016820701414215e-05,
+      "loss": 1.1614,
+      "num_input_tokens_seen": 498685408,
+      "step": 12700
+    },
+    {
+      "epoch": 0.6158542494427754,
+      "grad_norm": 0.399446964263916,
+      "learning_rate": 1.697965009651677e-05,
+      "loss": 1.1788,
+      "num_input_tokens_seen": 499066376,
+      "step": 12710
+    },
+    {
+      "epoch": 0.6163387925186549,
+      "grad_norm": 0.3826245069503784,
+      "learning_rate": 1.6942499245462525e-05,
+      "loss": 1.1421,
+      "num_input_tokens_seen": 499446672,
+      "step": 12720
+    },
+    {
+      "epoch": 0.6168233355945344,
+      "grad_norm": 0.39101335406303406,
+      "learning_rate": 1.6905368239752718e-05,
+      "loss": 1.1129,
+      "num_input_tokens_seen": 499832256,
+      "step": 12730
+    },
+    {
+      "epoch": 0.6173078786704138,
+      "grad_norm": 0.41338834166526794,
+      "learning_rate": 1.686825717083975e-05,
+      "loss": 1.1999,
+      "num_input_tokens_seen": 500220888,
+      "step": 12740
+    },
+    {
+      "epoch": 0.6177924217462932,
+      "grad_norm": 0.4286590814590454,
+      "learning_rate": 1.6831166130126872e-05,
+      "loss": 1.1657,
+      "num_input_tokens_seen": 500641332,
+      "step": 12750
+    },
+    {
+      "epoch": 0.6182769648221726,
+      "grad_norm": 0.4031652510166168,
+      "learning_rate": 1.6794095208968058e-05,
+      "loss": 1.1097,
+      "num_input_tokens_seen": 501041940,
+      "step": 12760
+    },
+    {
+      "epoch": 0.6187615078980522,
+      "grad_norm": 0.3872615098953247,
+      "learning_rate": 1.675704449866768e-05,
+      "loss": 1.1681,
+      "num_input_tokens_seen": 501421856,
+      "step": 12770
+    },
+    {
+      "epoch": 0.6192460509739316,
+      "grad_norm": 0.4028913676738739,
+      "learning_rate": 1.672001409048034e-05,
+      "loss": 1.1452,
+      "num_input_tokens_seen": 501803240,
+      "step": 12780
+    },
+    {
+      "epoch": 0.619730594049811,
+      "grad_norm": 0.4178786277770996,
+      "learning_rate": 1.668300407561067e-05,
+      "loss": 1.1493,
+      "num_input_tokens_seen": 502206632,
+      "step": 12790
+    },
+    {
+      "epoch": 0.6202151371256904,
+      "grad_norm": 0.4024272859096527,
+      "learning_rate": 1.6646014545213042e-05,
+      "loss": 1.1246,
+      "num_input_tokens_seen": 502597668,
+      "step": 12800
+    },
+    {
+      "epoch": 0.62069968020157,
+      "grad_norm": 0.38439643383026123,
+      "learning_rate": 1.6609045590391383e-05,
+      "loss": 1.1421,
+      "num_input_tokens_seen": 502992140,
+      "step": 12810
+    },
+    {
+      "epoch": 0.6211842232774494,
+      "grad_norm": 0.44478264451026917,
+      "learning_rate": 1.6572097302198935e-05,
+      "loss": 1.116,
+      "num_input_tokens_seen": 503374384,
+      "step": 12820
+    },
+    {
+      "epoch": 0.6216687663533288,
+      "grad_norm": 0.44358116388320923,
+      "learning_rate": 1.6535169771638066e-05,
+      "loss": 1.127,
+      "num_input_tokens_seen": 503789460,
+      "step": 12830
+    },
+    {
+      "epoch": 0.6221533094292082,
+      "grad_norm": 0.3874736428260803,
+      "learning_rate": 1.6498263089659992e-05,
+      "loss": 1.1076,
+      "num_input_tokens_seen": 504149520,
+      "step": 12840
+    },
+    {
+      "epoch": 0.6226378525050877,
+      "grad_norm": 0.3996816873550415,
+      "learning_rate": 1.6461377347164593e-05,
+      "loss": 1.1729,
+      "num_input_tokens_seen": 504529724,
+      "step": 12850
+    },
+    {
+      "epoch": 0.6231223955809672,
+      "grad_norm": 0.39820224046707153,
+      "learning_rate": 1.6424512635000158e-05,
+      "loss": 1.1639,
+      "num_input_tokens_seen": 504913244,
+      "step": 12860
+    },
+    {
+      "epoch": 0.6236069386568466,
+      "grad_norm": 0.4100916385650635,
+      "learning_rate": 1.638766904396321e-05,
+      "loss": 1.1485,
+      "num_input_tokens_seen": 505324308,
+      "step": 12870
+    },
+    {
+      "epoch": 0.624091481732726,
+      "grad_norm": 0.38851940631866455,
+      "learning_rate": 1.635084666479822e-05,
+      "loss": 1.1161,
+      "num_input_tokens_seen": 505696332,
+      "step": 12880
+    },
+    {
+      "epoch": 0.6245760248086055,
+      "grad_norm": 0.4355764091014862,
+      "learning_rate": 1.6314045588197442e-05,
+      "loss": 1.143,
+      "num_input_tokens_seen": 506083568,
+      "step": 12890
+    },
+    {
+      "epoch": 0.6250605678844849,
+      "grad_norm": 0.38407522439956665,
+      "learning_rate": 1.6277265904800643e-05,
+      "loss": 1.1396,
+      "num_input_tokens_seen": 506497556,
+      "step": 12900
+    },
+    {
+      "epoch": 0.6255451109603644,
+      "grad_norm": 0.38687968254089355,
+      "learning_rate": 1.6240507705194896e-05,
+      "loss": 1.1215,
+      "num_input_tokens_seen": 506881424,
+      "step": 12910
+    },
+    {
+      "epoch": 0.6260296540362438,
+      "grad_norm": 0.36193808913230896,
+      "learning_rate": 1.6203771079914387e-05,
+      "loss": 1.1487,
+      "num_input_tokens_seen": 507256568,
+      "step": 12920
+    },
+    {
+      "epoch": 0.6265141971121233,
+      "grad_norm": 0.3949004113674164,
+      "learning_rate": 1.6167056119440125e-05,
+      "loss": 1.1569,
+      "num_input_tokens_seen": 507669200,
+      "step": 12930
+    },
+    {
+      "epoch": 0.6269987401880027,
+      "grad_norm": 0.39322030544281006,
+      "learning_rate": 1.6130362914199814e-05,
+      "loss": 1.0929,
+      "num_input_tokens_seen": 508034880,
+      "step": 12940
+    },
+    {
+      "epoch": 0.6274832832638821,
+      "grad_norm": 0.40019360184669495,
+      "learning_rate": 1.6093691554567524e-05,
+      "loss": 1.1317,
+      "num_input_tokens_seen": 508462956,
+      "step": 12950
+    },
+    {
+      "epoch": 0.6279678263397616,
+      "grad_norm": 0.40196293592453003,
+      "learning_rate": 1.6057042130863538e-05,
+      "loss": 1.0989,
+      "num_input_tokens_seen": 508848948,
+      "step": 12960
+    },
+    {
+      "epoch": 0.6284523694156411,
+      "grad_norm": 0.4627667963504791,
+      "learning_rate": 1.6020414733354117e-05,
+      "loss": 1.1436,
+      "num_input_tokens_seen": 509221768,
+      "step": 12970
+    },
+    {
+      "epoch": 0.6289369124915205,
+      "grad_norm": 0.42447420954704285,
+      "learning_rate": 1.598380945225127e-05,
+      "loss": 1.1207,
+      "num_input_tokens_seen": 509643908,
+      "step": 12980
+    },
+    {
+      "epoch": 0.6294214555673999,
+      "grad_norm": 0.3753610849380493,
+      "learning_rate": 1.594722637771253e-05,
+      "loss": 1.1599,
+      "num_input_tokens_seen": 510024072,
+      "step": 12990
+    },
+    {
+      "epoch": 0.6299059986432793,
+      "grad_norm": 0.39578545093536377,
+      "learning_rate": 1.5910665599840745e-05,
+      "loss": 1.1634,
+      "num_input_tokens_seen": 510426748,
+      "step": 13000
+    },
+    {
+      "epoch": 0.6303905417191589,
+      "grad_norm": 0.4332728087902069,
+      "learning_rate": 1.5874127208683824e-05,
+      "loss": 1.0877,
+      "num_input_tokens_seen": 510818184,
+      "step": 13010
+    },
+    {
+      "epoch": 0.6308750847950383,
+      "grad_norm": 0.3980608284473419,
+      "learning_rate": 1.5837611294234583e-05,
+      "loss": 1.102,
+      "num_input_tokens_seen": 511196012,
+      "step": 13020
+    },
+    {
+      "epoch": 0.6313596278709177,
+      "grad_norm": 0.3887230455875397,
+      "learning_rate": 1.5801117946430434e-05,
+      "loss": 1.1257,
+      "num_input_tokens_seen": 511590376,
+      "step": 13030
+    },
+    {
+      "epoch": 0.6318441709467971,
+      "grad_norm": 0.4023469388484955,
+      "learning_rate": 1.576464725515322e-05,
+      "loss": 1.13,
+      "num_input_tokens_seen": 511985844,
+      "step": 13040
+    },
+    {
+      "epoch": 0.6323287140226767,
+      "grad_norm": 0.3778144121170044,
+      "learning_rate": 1.5728199310229e-05,
+      "loss": 1.0993,
+      "num_input_tokens_seen": 512401748,
+      "step": 13050
+    },
+    {
+      "epoch": 0.6328132570985561,
+      "grad_norm": 0.3801812529563904,
+      "learning_rate": 1.5691774201427772e-05,
+      "loss": 1.0846,
+      "num_input_tokens_seen": 512818064,
+      "step": 13060
+    },
+    {
+      "epoch": 0.6332978001744355,
+      "grad_norm": 0.3784249722957611,
+      "learning_rate": 1.565537201846335e-05,
+      "loss": 1.1541,
+      "num_input_tokens_seen": 513220236,
+      "step": 13070
+    },
+    {
+      "epoch": 0.6337823432503149,
+      "grad_norm": 0.4744737148284912,
+      "learning_rate": 1.5618992850993025e-05,
+      "loss": 1.1496,
+      "num_input_tokens_seen": 513594808,
+      "step": 13080
+    },
+    {
+      "epoch": 0.6342668863261944,
+      "grad_norm": 0.3975614607334137,
+      "learning_rate": 1.5582636788617412e-05,
+      "loss": 1.176,
+      "num_input_tokens_seen": 513981028,
+      "step": 13090
+    },
+    {
+      "epoch": 0.6347514294020739,
+      "grad_norm": 0.3835643231868744,
+      "learning_rate": 1.5546303920880256e-05,
+      "loss": 1.1276,
+      "num_input_tokens_seen": 514381624,
+      "step": 13100
+    },
+    {
+      "epoch": 0.6352359724779533,
+      "grad_norm": 0.3918743431568146,
+      "learning_rate": 1.5509994337268126e-05,
+      "loss": 1.1648,
+      "num_input_tokens_seen": 514799552,
+      "step": 13110
+    },
+    {
+      "epoch": 0.6357205155538327,
+      "grad_norm": 0.3977094888687134,
+      "learning_rate": 1.547370812721028e-05,
+      "loss": 1.0923,
+      "num_input_tokens_seen": 515216184,
+      "step": 13120
+    },
+    {
+      "epoch": 0.6362050586297122,
+      "grad_norm": 0.41210541129112244,
+      "learning_rate": 1.5437445380078383e-05,
+      "loss": 1.1419,
+      "num_input_tokens_seen": 515617056,
+      "step": 13130
+    },
+    {
+      "epoch": 0.6366896017055916,
+      "grad_norm": 0.3935719132423401,
+      "learning_rate": 1.5401206185186323e-05,
+      "loss": 1.1472,
+      "num_input_tokens_seen": 516021628,
+      "step": 13140
+    },
+    {
+      "epoch": 0.637174144781471,
+      "grad_norm": 0.4016217291355133,
+      "learning_rate": 1.536499063178999e-05,
+      "loss": 1.1494,
+      "num_input_tokens_seen": 516424464,
+      "step": 13150
+    },
+    {
+      "epoch": 0.6376586878573505,
+      "grad_norm": 0.41335245966911316,
+      "learning_rate": 1.5328798809087012e-05,
+      "loss": 1.1859,
+      "num_input_tokens_seen": 516827724,
+      "step": 13160
+    },
+    {
+      "epoch": 0.63814323093323,
+      "grad_norm": 0.38563454151153564,
+      "learning_rate": 1.5292630806216603e-05,
+      "loss": 1.1049,
+      "num_input_tokens_seen": 517224524,
+      "step": 13170
+    },
+    {
+      "epoch": 0.6386277740091094,
+      "grad_norm": 0.4143526256084442,
+      "learning_rate": 1.5256486712259304e-05,
+      "loss": 1.1405,
+      "num_input_tokens_seen": 517641892,
+      "step": 13180
+    },
+    {
+      "epoch": 0.6391123170849888,
+      "grad_norm": 0.3733633756637573,
+      "learning_rate": 1.5220366616236748e-05,
+      "loss": 1.1688,
+      "num_input_tokens_seen": 518022120,
+      "step": 13190
+    },
+    {
+      "epoch": 0.6395968601608683,
+      "grad_norm": 0.412308931350708,
+      "learning_rate": 1.5184270607111494e-05,
+      "loss": 1.1234,
+      "num_input_tokens_seen": 518438460,
+      "step": 13200
+    },
+    {
+      "epoch": 0.6400814032367478,
+      "grad_norm": 0.3862762749195099,
+      "learning_rate": 1.5148198773786754e-05,
+      "loss": 1.14,
+      "num_input_tokens_seen": 518813036,
+      "step": 13210
+    },
+    {
+      "epoch": 0.6405659463126272,
+      "grad_norm": 0.4147486984729767,
+      "learning_rate": 1.5112151205106182e-05,
+      "loss": 1.1246,
+      "num_input_tokens_seen": 519216952,
+      "step": 13220
+    },
+    {
+      "epoch": 0.6410504893885066,
+      "grad_norm": 0.40653976798057556,
+      "learning_rate": 1.507612798985371e-05,
+      "loss": 1.115,
+      "num_input_tokens_seen": 519604276,
+      "step": 13230
+    },
+    {
+      "epoch": 0.641535032464386,
+      "grad_norm": 0.4051119089126587,
+      "learning_rate": 1.5040129216753257e-05,
+      "loss": 1.179,
+      "num_input_tokens_seen": 519988992,
+      "step": 13240
+    },
+    {
+      "epoch": 0.6420195755402656,
+      "grad_norm": 0.39531031250953674,
+      "learning_rate": 1.5004154974468548e-05,
+      "loss": 1.1244,
+      "num_input_tokens_seen": 520363784,
+      "step": 13250
+    },
+    {
+      "epoch": 0.642504118616145,
+      "grad_norm": 0.38629281520843506,
+      "learning_rate": 1.4968205351602899e-05,
+      "loss": 1.115,
+      "num_input_tokens_seen": 520766492,
+      "step": 13260
+    },
+    {
+      "epoch": 0.6429886616920244,
+      "grad_norm": 0.3945416808128357,
+      "learning_rate": 1.4932280436698976e-05,
+      "loss": 1.1265,
+      "num_input_tokens_seen": 521157024,
+      "step": 13270
+    },
+    {
+      "epoch": 0.6434732047679038,
+      "grad_norm": 0.3776805102825165,
+      "learning_rate": 1.48963803182386e-05,
+      "loss": 1.1165,
+      "num_input_tokens_seen": 521579912,
+      "step": 13280
+    },
+    {
+      "epoch": 0.6439577478437833,
+      "grad_norm": 0.4211878776550293,
+      "learning_rate": 1.4860505084642506e-05,
+      "loss": 1.1249,
+      "num_input_tokens_seen": 521991792,
+      "step": 13290
+    },
+    {
+      "epoch": 0.6444422909196628,
+      "grad_norm": 0.4258326292037964,
+      "learning_rate": 1.4824654824270168e-05,
+      "loss": 1.1046,
+      "num_input_tokens_seen": 522392304,
+      "step": 13300
+    },
+    {
+      "epoch": 0.6449268339955422,
+      "grad_norm": 0.40192142128944397,
+      "learning_rate": 1.478882962541952e-05,
+      "loss": 1.1577,
+      "num_input_tokens_seen": 522783384,
+      "step": 13310
+    },
+    {
+      "epoch": 0.6454113770714216,
+      "grad_norm": 0.3894496262073517,
+      "learning_rate": 1.475302957632677e-05,
+      "loss": 1.1205,
+      "num_input_tokens_seen": 523150196,
+      "step": 13320
+    },
+    {
+      "epoch": 0.6458959201473011,
+      "grad_norm": 0.4126818776130676,
+      "learning_rate": 1.471725476516622e-05,
+      "loss": 1.1076,
+      "num_input_tokens_seen": 523551404,
+      "step": 13330
+    },
+    {
+      "epoch": 0.6463804632231805,
+      "grad_norm": 0.37312477827072144,
+      "learning_rate": 1.4681505280049974e-05,
+      "loss": 1.1852,
+      "num_input_tokens_seen": 523955836,
+      "step": 13340
+    },
+    {
+      "epoch": 0.64686500629906,
+      "grad_norm": 0.40087607502937317,
+      "learning_rate": 1.4645781209027764e-05,
+      "loss": 1.1758,
+      "num_input_tokens_seen": 524356324,
+      "step": 13350
+    },
+    {
+      "epoch": 0.6473495493749394,
+      "grad_norm": 0.38394778966903687,
+      "learning_rate": 1.4610082640086754e-05,
+      "loss": 1.108,
+      "num_input_tokens_seen": 524759624,
+      "step": 13360
+    },
+    {
+      "epoch": 0.6478340924508189,
+      "grad_norm": 0.4227401912212372,
+      "learning_rate": 1.4574409661151264e-05,
+      "loss": 1.1216,
+      "num_input_tokens_seen": 525129624,
+      "step": 13370
+    },
+    {
+      "epoch": 0.6483186355266983,
+      "grad_norm": 0.3866701126098633,
+      "learning_rate": 1.4538762360082608e-05,
+      "loss": 1.1328,
+      "num_input_tokens_seen": 525533224,
+      "step": 13380
+    },
+    {
+      "epoch": 0.6488031786025777,
+      "grad_norm": 0.39452120661735535,
+      "learning_rate": 1.4503140824678863e-05,
+      "loss": 1.124,
+      "num_input_tokens_seen": 525921192,
+      "step": 13390
+    },
+    {
+      "epoch": 0.6492877216784572,
+      "grad_norm": 0.3937327563762665,
+      "learning_rate": 1.4467545142674599e-05,
+      "loss": 1.132,
+      "num_input_tokens_seen": 526287908,
+      "step": 13400
+    },
+    {
+      "epoch": 0.6497722647543367,
+      "grad_norm": 0.427117258310318,
+      "learning_rate": 1.4431975401740783e-05,
+      "loss": 1.1116,
+      "num_input_tokens_seen": 526650256,
+      "step": 13410
+    },
+    {
+      "epoch": 0.6502568078302161,
+      "grad_norm": 0.36921098828315735,
+      "learning_rate": 1.4396431689484416e-05,
+      "loss": 1.0976,
+      "num_input_tokens_seen": 527032076,
+      "step": 13420
+    },
+    {
+      "epoch": 0.6507413509060955,
+      "grad_norm": 0.3914394676685333,
+      "learning_rate": 1.4360914093448463e-05,
+      "loss": 1.2131,
+      "num_input_tokens_seen": 527466664,
+      "step": 13430
+    },
+    {
+      "epoch": 0.651225893981975,
+      "grad_norm": 0.4258521497249603,
+      "learning_rate": 1.4325422701111502e-05,
+      "loss": 1.1817,
+      "num_input_tokens_seen": 527850952,
+      "step": 13440
+    },
+    {
+      "epoch": 0.6517104370578545,
+      "grad_norm": 0.401259183883667,
+      "learning_rate": 1.428995759988761e-05,
+      "loss": 1.158,
+      "num_input_tokens_seen": 528263332,
+      "step": 13450
+    },
+    {
+      "epoch": 0.6521949801337339,
+      "grad_norm": 0.3809134066104889,
+      "learning_rate": 1.4254518877126095e-05,
+      "loss": 1.1234,
+      "num_input_tokens_seen": 528686588,
+      "step": 13460
+    },
+    {
+      "epoch": 0.6526795232096133,
+      "grad_norm": 0.406120240688324,
+      "learning_rate": 1.42191066201113e-05,
+      "loss": 1.1635,
+      "num_input_tokens_seen": 529070212,
+      "step": 13470
+    },
+    {
+      "epoch": 0.6531640662854927,
+      "grad_norm": 0.43414533138275146,
+      "learning_rate": 1.418372091606239e-05,
+      "loss": 1.1747,
+      "num_input_tokens_seen": 529477424,
+      "step": 13480
+    },
+    {
+      "epoch": 0.6536486093613723,
+      "grad_norm": 0.4020240306854248,
+      "learning_rate": 1.4148361852133129e-05,
+      "loss": 1.1394,
+      "num_input_tokens_seen": 529879856,
+      "step": 13490
+    },
+    {
+      "epoch": 0.6541331524372517,
+      "grad_norm": 0.39953121542930603,
+      "learning_rate": 1.4113029515411647e-05,
+      "loss": 1.1363,
+      "num_input_tokens_seen": 530288184,
+      "step": 13500
+    },
+    {
+      "epoch": 0.6546176955131311,
+      "grad_norm": 0.39526060223579407,
+      "learning_rate": 1.407772399292027e-05,
+      "loss": 1.1483,
+      "num_input_tokens_seen": 530697004,
+      "step": 13510
+    },
+    {
+      "epoch": 0.6551022385890105,
+      "grad_norm": 0.4059954583644867,
+      "learning_rate": 1.4042445371615271e-05,
+      "loss": 1.125,
+      "num_input_tokens_seen": 531092792,
+      "step": 13520
+    },
+    {
+      "epoch": 0.65558678166489,
+      "grad_norm": 0.40413933992385864,
+      "learning_rate": 1.4007193738386675e-05,
+      "loss": 1.1345,
+      "num_input_tokens_seen": 531481940,
+      "step": 13530
+    },
+    {
+      "epoch": 0.6560713247407695,
+      "grad_norm": 0.3758378326892853,
+      "learning_rate": 1.3971969180058032e-05,
+      "loss": 1.139,
+      "num_input_tokens_seen": 531883224,
+      "step": 13540
+    },
+    {
+      "epoch": 0.6565558678166489,
+      "grad_norm": 0.3990527093410492,
+      "learning_rate": 1.3936771783386183e-05,
+      "loss": 1.1284,
+      "num_input_tokens_seen": 532298904,
+      "step": 13550
+    },
+    {
+      "epoch": 0.6570404108925283,
+      "grad_norm": 0.403590589761734,
+      "learning_rate": 1.390160163506113e-05,
+      "loss": 1.1493,
+      "num_input_tokens_seen": 532676364,
+      "step": 13560
+    },
+    {
+      "epoch": 0.6575249539684078,
+      "grad_norm": 0.41753387451171875,
+      "learning_rate": 1.3866458821705697e-05,
+      "loss": 1.1165,
+      "num_input_tokens_seen": 533090008,
+      "step": 13570
+    },
+    {
+      "epoch": 0.6580094970442872,
+      "grad_norm": 0.40354228019714355,
+      "learning_rate": 1.3831343429875421e-05,
+      "loss": 1.1542,
+      "num_input_tokens_seen": 533475684,
+      "step": 13580
+    },
+    {
+      "epoch": 0.6584940401201667,
+      "grad_norm": 0.40463098883628845,
+      "learning_rate": 1.3796255546058293e-05,
+      "loss": 1.1574,
+      "num_input_tokens_seen": 533869320,
+      "step": 13590
+    },
+    {
+      "epoch": 0.6589785831960461,
+      "grad_norm": 0.40255841612815857,
+      "learning_rate": 1.3761195256674554e-05,
+      "loss": 1.1158,
+      "num_input_tokens_seen": 534226148,
+      "step": 13600
+    },
+    {
+      "epoch": 0.6594631262719256,
+      "grad_norm": 0.4016367197036743,
+      "learning_rate": 1.3726162648076474e-05,
+      "loss": 1.1712,
+      "num_input_tokens_seen": 534636196,
+      "step": 13610
+    },
+    {
+      "epoch": 0.659947669347805,
+      "grad_norm": 0.4226371645927429,
+      "learning_rate": 1.3691157806548167e-05,
+      "loss": 1.1584,
+      "num_input_tokens_seen": 535027492,
+      "step": 13620
+    },
+    {
+      "epoch": 0.6604322124236844,
+      "grad_norm": 0.4005202651023865,
+      "learning_rate": 1.365618081830532e-05,
+      "loss": 1.1075,
+      "num_input_tokens_seen": 535420836,
+      "step": 13630
+    },
+    {
+      "epoch": 0.6609167554995639,
+      "grad_norm": 0.38361644744873047,
+      "learning_rate": 1.3621231769495047e-05,
+      "loss": 1.1318,
+      "num_input_tokens_seen": 535798348,
+      "step": 13640
+    },
+    {
+      "epoch": 0.6614012985754434,
+      "grad_norm": 0.36114785075187683,
+      "learning_rate": 1.3586310746195641e-05,
+      "loss": 1.1685,
+      "num_input_tokens_seen": 536163432,
+      "step": 13650
+    },
+    {
+      "epoch": 0.6618858416513228,
+      "grad_norm": 0.452426940202713,
+      "learning_rate": 1.3551417834416375e-05,
+      "loss": 1.1241,
+      "num_input_tokens_seen": 536559076,
+      "step": 13660
+    },
+    {
+      "epoch": 0.6623703847272022,
+      "grad_norm": 0.43133866786956787,
+      "learning_rate": 1.3516553120097281e-05,
+      "loss": 1.1362,
+      "num_input_tokens_seen": 536949788,
+      "step": 13670
+    },
+    {
+      "epoch": 0.6628549278030816,
+      "grad_norm": 0.3915277421474457,
+      "learning_rate": 1.3481716689108915e-05,
+      "loss": 1.1215,
+      "num_input_tokens_seen": 537337428,
+      "step": 13680
+    },
+    {
+      "epoch": 0.6633394708789612,
+      "grad_norm": 0.38239341974258423,
+      "learning_rate": 1.3446908627252236e-05,
+      "loss": 1.1594,
+      "num_input_tokens_seen": 537726764,
+      "step": 13690
+    },
+    {
+      "epoch": 0.6638240139548406,
+      "grad_norm": 0.4042087495326996,
+      "learning_rate": 1.3412129020258257e-05,
+      "loss": 1.1515,
+      "num_input_tokens_seen": 538088164,
+      "step": 13700
+    },
+    {
+      "epoch": 0.66430855703072,
+      "grad_norm": 0.39968568086624146,
+      "learning_rate": 1.3377377953787956e-05,
+      "loss": 1.116,
+      "num_input_tokens_seen": 538462580,
+      "step": 13710
+    },
+    {
+      "epoch": 0.6647931001065994,
+      "grad_norm": 0.3845781683921814,
+      "learning_rate": 1.3342655513432001e-05,
+      "loss": 1.1221,
+      "num_input_tokens_seen": 538858660,
+      "step": 13720
+    },
+    {
+      "epoch": 0.665277643182479,
+      "grad_norm": 0.37981632351875305,
+      "learning_rate": 1.3307961784710554e-05,
+      "loss": 1.1724,
+      "num_input_tokens_seen": 539260832,
+      "step": 13730
+    },
+    {
+      "epoch": 0.6657621862583584,
+      "grad_norm": 0.4112668037414551,
+      "learning_rate": 1.327329685307307e-05,
+      "loss": 1.1521,
+      "num_input_tokens_seen": 539654112,
+      "step": 13740
+    },
+    {
+      "epoch": 0.6662467293342378,
+      "grad_norm": 0.3671356439590454,
+      "learning_rate": 1.3238660803898074e-05,
+      "loss": 1.1909,
+      "num_input_tokens_seen": 540039560,
+      "step": 13750
+    },
+    {
+      "epoch": 0.6667312724101173,
+      "grad_norm": 0.3972877562046051,
+      "learning_rate": 1.3204053722492927e-05,
+      "loss": 1.1792,
+      "num_input_tokens_seen": 540430488,
+      "step": 13760
+    },
+    {
+      "epoch": 0.6672158154859967,
+      "grad_norm": 0.39763057231903076,
+      "learning_rate": 1.3169475694093703e-05,
+      "loss": 1.1322,
+      "num_input_tokens_seen": 540829364,
+      "step": 13770
+    },
+    {
+      "epoch": 0.6677003585618762,
+      "grad_norm": 0.37179917097091675,
+      "learning_rate": 1.3134926803864861e-05,
+      "loss": 1.1592,
+      "num_input_tokens_seen": 541255364,
+      "step": 13780
+    },
+    {
+      "epoch": 0.6681849016377556,
+      "grad_norm": 0.3995148241519928,
+      "learning_rate": 1.3100407136899123e-05,
+      "loss": 1.103,
+      "num_input_tokens_seen": 541631036,
+      "step": 13790
+    },
+    {
+      "epoch": 0.6686694447136351,
+      "grad_norm": 0.4086041748523712,
+      "learning_rate": 1.3065916778217235e-05,
+      "loss": 1.1492,
+      "num_input_tokens_seen": 542006572,
+      "step": 13800
+    },
+    {
+      "epoch": 0.6691539877895145,
+      "grad_norm": 0.3766721785068512,
+      "learning_rate": 1.3031455812767746e-05,
+      "loss": 1.1115,
+      "num_input_tokens_seen": 542379296,
+      "step": 13810
+    },
+    {
+      "epoch": 0.6696385308653939,
+      "grad_norm": 0.4227388799190521,
+      "learning_rate": 1.299702432542683e-05,
+      "loss": 1.0872,
+      "num_input_tokens_seen": 542769344,
+      "step": 13820
+    },
+    {
+      "epoch": 0.6701230739412734,
+      "grad_norm": 0.4330199658870697,
+      "learning_rate": 1.2962622400998012e-05,
+      "loss": 1.0849,
+      "num_input_tokens_seen": 543147408,
+      "step": 13830
+    },
+    {
+      "epoch": 0.6706076170171529,
+      "grad_norm": 0.37317484617233276,
+      "learning_rate": 1.292825012421208e-05,
+      "loss": 1.1378,
+      "num_input_tokens_seen": 543533556,
+      "step": 13840
+    },
+    {
+      "epoch": 0.6710921600930323,
+      "grad_norm": 0.39739182591438293,
+      "learning_rate": 1.2893907579726728e-05,
+      "loss": 1.1708,
+      "num_input_tokens_seen": 543906020,
+      "step": 13850
+    },
+    {
+      "epoch": 0.6715767031689117,
+      "grad_norm": 0.4071529507637024,
+      "learning_rate": 1.2859594852126456e-05,
+      "loss": 1.1109,
+      "num_input_tokens_seen": 544321144,
+      "step": 13860
+    },
+    {
+      "epoch": 0.6720612462447911,
+      "grad_norm": 0.398400217294693,
+      "learning_rate": 1.282531202592232e-05,
+      "loss": 1.0842,
+      "num_input_tokens_seen": 544708388,
+      "step": 13870
+    },
+    {
+      "epoch": 0.6725457893206707,
+      "grad_norm": 0.37718960642814636,
+      "learning_rate": 1.2791059185551744e-05,
+      "loss": 1.1508,
+      "num_input_tokens_seen": 545104336,
+      "step": 13880
+    },
+    {
+      "epoch": 0.6730303323965501,
+      "grad_norm": 0.41166213154792786,
+      "learning_rate": 1.2756836415378254e-05,
+      "loss": 1.1302,
+      "num_input_tokens_seen": 545501896,
+      "step": 13890
+    },
+    {
+      "epoch": 0.6735148754724295,
+      "grad_norm": 0.4132916033267975,
+      "learning_rate": 1.2722643799691378e-05,
+      "loss": 1.1038,
+      "num_input_tokens_seen": 545881328,
+      "step": 13900
+    },
+    {
+      "epoch": 0.6739994185483089,
+      "grad_norm": 0.40238797664642334,
+      "learning_rate": 1.2688481422706309e-05,
+      "loss": 1.146,
+      "num_input_tokens_seen": 546279016,
+      "step": 13910
+    },
+    {
+      "epoch": 0.6744839616241884,
+      "grad_norm": 0.4064416289329529,
+      "learning_rate": 1.2654349368563828e-05,
+      "loss": 1.1636,
+      "num_input_tokens_seen": 546677216,
+      "step": 13920
+    },
+    {
+      "epoch": 0.6749685047000679,
+      "grad_norm": 0.37984856963157654,
+      "learning_rate": 1.2620247721329973e-05,
+      "loss": 1.1299,
+      "num_input_tokens_seen": 547080472,
+      "step": 13930
+    },
+    {
+      "epoch": 0.6754530477759473,
+      "grad_norm": 0.4335133135318756,
+      "learning_rate": 1.2586176564995922e-05,
+      "loss": 1.1437,
+      "num_input_tokens_seen": 547472432,
+      "step": 13940
+    },
+    {
+      "epoch": 0.6759375908518267,
+      "grad_norm": 0.3851531445980072,
+      "learning_rate": 1.2552135983477756e-05,
+      "loss": 1.1503,
+      "num_input_tokens_seen": 547869104,
+      "step": 13950
+    },
+    {
+      "epoch": 0.6764221339277062,
+      "grad_norm": 0.37984663248062134,
+      "learning_rate": 1.251812606061624e-05,
+      "loss": 1.1761,
+      "num_input_tokens_seen": 548278152,
+      "step": 13960
+    },
+    {
+      "epoch": 0.6769066770035856,
+      "grad_norm": 0.40836653113365173,
+      "learning_rate": 1.2484146880176642e-05,
+      "loss": 1.066,
+      "num_input_tokens_seen": 548692648,
+      "step": 13970
+    },
+    {
+      "epoch": 0.6773912200794651,
+      "grad_norm": 0.38975247740745544,
+      "learning_rate": 1.2450198525848487e-05,
+      "loss": 1.0798,
+      "num_input_tokens_seen": 549066912,
+      "step": 13980
+    },
+    {
+      "epoch": 0.6778757631553445,
+      "grad_norm": 0.3853222727775574,
+      "learning_rate": 1.2416281081245398e-05,
+      "loss": 1.1685,
+      "num_input_tokens_seen": 549440944,
+      "step": 13990
+    },
+    {
+      "epoch": 0.678360306231224,
+      "grad_norm": 0.45054277777671814,
+      "learning_rate": 1.2382394629904864e-05,
+      "loss": 1.1473,
+      "num_input_tokens_seen": 549831184,
+      "step": 14000
+    },
+    {
+      "epoch": 0.678360306231224,
+      "eval_loss": 1.1321347951889038,
+      "eval_runtime": 5.0521,
+      "eval_samples_per_second": 29.69,
+      "eval_steps_per_second": 3.761,
+      "num_input_tokens_seen": 549831184,
+      "step": 14000
+    },
+    {
+      "epoch": 0.6788448493071034,
+      "grad_norm": 0.3913303315639496,
+      "learning_rate": 1.2348539255288038e-05,
+      "loss": 1.1504,
+      "num_input_tokens_seen": 550197984,
+      "step": 14010
+    },
+    {
+      "epoch": 0.6793293923829828,
+      "grad_norm": 0.38616275787353516,
+      "learning_rate": 1.2314715040779534e-05,
+      "loss": 1.1412,
+      "num_input_tokens_seen": 550565288,
+      "step": 14020
+    },
+    {
+      "epoch": 0.6798139354588623,
+      "grad_norm": 0.376402884721756,
+      "learning_rate": 1.2280922069687225e-05,
+      "loss": 1.1712,
+      "num_input_tokens_seen": 550966008,
+      "step": 14030
+    },
+    {
+      "epoch": 0.6802984785347418,
+      "grad_norm": 0.4074733853340149,
+      "learning_rate": 1.2247160425241996e-05,
+      "loss": 1.0974,
+      "num_input_tokens_seen": 551358528,
+      "step": 14040
+    },
+    {
+      "epoch": 0.6807830216106212,
+      "grad_norm": 0.41833576560020447,
+      "learning_rate": 1.221343019059764e-05,
+      "loss": 1.1279,
+      "num_input_tokens_seen": 551759700,
+      "step": 14050
+    },
+    {
+      "epoch": 0.6812675646865006,
+      "grad_norm": 0.36511725187301636,
+      "learning_rate": 1.217973144883053e-05,
+      "loss": 1.1514,
+      "num_input_tokens_seen": 552164368,
+      "step": 14060
+    },
+    {
+      "epoch": 0.68175210776238,
+      "grad_norm": 0.3876798152923584,
+      "learning_rate": 1.2146064282939501e-05,
+      "loss": 1.169,
+      "num_input_tokens_seen": 552571428,
+      "step": 14070
+    },
+    {
+      "epoch": 0.6822366508382596,
+      "grad_norm": 0.3926098942756653,
+      "learning_rate": 1.2112428775845616e-05,
+      "loss": 1.1665,
+      "num_input_tokens_seen": 552952628,
+      "step": 14080
+    },
+    {
+      "epoch": 0.682721193914139,
+      "grad_norm": 0.3909384608268738,
+      "learning_rate": 1.2078825010391958e-05,
+      "loss": 1.133,
+      "num_input_tokens_seen": 553359608,
+      "step": 14090
+    },
+    {
+      "epoch": 0.6832057369900184,
+      "grad_norm": 0.38221150636672974,
+      "learning_rate": 1.2045253069343448e-05,
+      "loss": 1.1773,
+      "num_input_tokens_seen": 553757744,
+      "step": 14100
+    },
+    {
+      "epoch": 0.6836902800658978,
+      "grad_norm": 0.42475947737693787,
+      "learning_rate": 1.2011713035386588e-05,
+      "loss": 1.1495,
+      "num_input_tokens_seen": 554165892,
+      "step": 14110
+    },
+    {
+      "epoch": 0.6841748231417774,
+      "grad_norm": 0.41192540526390076,
+      "learning_rate": 1.1978204991129324e-05,
+      "loss": 1.1358,
+      "num_input_tokens_seen": 554549428,
+      "step": 14120
+    },
+    {
+      "epoch": 0.6846593662176568,
+      "grad_norm": 0.390726238489151,
+      "learning_rate": 1.1944729019100808e-05,
+      "loss": 1.1225,
+      "num_input_tokens_seen": 554948408,
+      "step": 14130
+    },
+    {
+      "epoch": 0.6851439092935362,
+      "grad_norm": 0.40107494592666626,
+      "learning_rate": 1.19112852017512e-05,
+      "loss": 1.1305,
+      "num_input_tokens_seen": 555381500,
+      "step": 14140
+    },
+    {
+      "epoch": 0.6856284523694156,
+      "grad_norm": 0.41240257024765015,
+      "learning_rate": 1.1877873621451453e-05,
+      "loss": 1.1169,
+      "num_input_tokens_seen": 555772300,
+      "step": 14150
+    },
+    {
+      "epoch": 0.6861129954452951,
+      "grad_norm": 0.42891547083854675,
+      "learning_rate": 1.1844494360493141e-05,
+      "loss": 1.095,
+      "num_input_tokens_seen": 556200948,
+      "step": 14160
+    },
+    {
+      "epoch": 0.6865975385211746,
+      "grad_norm": 0.4002200663089752,
+      "learning_rate": 1.1811147501088196e-05,
+      "loss": 1.188,
+      "num_input_tokens_seen": 556610528,
+      "step": 14170
+    },
+    {
+      "epoch": 0.687082081597054,
+      "grad_norm": 0.4111470878124237,
+      "learning_rate": 1.1777833125368812e-05,
+      "loss": 1.1927,
+      "num_input_tokens_seen": 557009708,
+      "step": 14180
+    },
+    {
+      "epoch": 0.6875666246729334,
+      "grad_norm": 0.43058955669403076,
+      "learning_rate": 1.17445513153871e-05,
+      "loss": 1.1684,
+      "num_input_tokens_seen": 557438504,
+      "step": 14190
+    },
+    {
+      "epoch": 0.6880511677488129,
+      "grad_norm": 0.36157649755477905,
+      "learning_rate": 1.1711302153115045e-05,
+      "loss": 1.1805,
+      "num_input_tokens_seen": 557855252,
+      "step": 14200
+    },
+    {
+      "epoch": 0.6885357108246923,
+      "grad_norm": 0.3946475088596344,
+      "learning_rate": 1.1678085720444142e-05,
+      "loss": 1.0994,
+      "num_input_tokens_seen": 558247492,
+      "step": 14210
+    },
+    {
+      "epoch": 0.6890202539005718,
+      "grad_norm": 0.38171178102493286,
+      "learning_rate": 1.1644902099185328e-05,
+      "loss": 1.1378,
+      "num_input_tokens_seen": 558657816,
+      "step": 14220
+    },
+    {
+      "epoch": 0.6895047969764512,
+      "grad_norm": 0.38918769359588623,
+      "learning_rate": 1.1611751371068706e-05,
+      "loss": 1.1926,
+      "num_input_tokens_seen": 559061308,
+      "step": 14230
+    },
+    {
+      "epoch": 0.6899893400523307,
+      "grad_norm": 0.4294280707836151,
+      "learning_rate": 1.1578633617743373e-05,
+      "loss": 1.1388,
+      "num_input_tokens_seen": 559449040,
+      "step": 14240
+    },
+    {
+      "epoch": 0.6904738831282101,
+      "grad_norm": 0.4078969359397888,
+      "learning_rate": 1.1545548920777194e-05,
+      "loss": 1.1432,
+      "num_input_tokens_seen": 559819924,
+      "step": 14250
+    },
+    {
+      "epoch": 0.6909584262040895,
+      "grad_norm": 0.3768039643764496,
+      "learning_rate": 1.1512497361656633e-05,
+      "loss": 1.1597,
+      "num_input_tokens_seen": 560234120,
+      "step": 14260
+    },
+    {
+      "epoch": 0.691442969279969,
+      "grad_norm": 0.3907431364059448,
+      "learning_rate": 1.1479479021786533e-05,
+      "loss": 1.1175,
+      "num_input_tokens_seen": 560648180,
+      "step": 14270
+    },
+    {
+      "epoch": 0.6919275123558485,
+      "grad_norm": 0.415067195892334,
+      "learning_rate": 1.1446493982489916e-05,
+      "loss": 1.1103,
+      "num_input_tokens_seen": 561025820,
+      "step": 14280
+    },
+    {
+      "epoch": 0.6924120554317279,
+      "grad_norm": 0.3955281674861908,
+      "learning_rate": 1.1413542325007804e-05,
+      "loss": 1.1182,
+      "num_input_tokens_seen": 561453128,
+      "step": 14290
+    },
+    {
+      "epoch": 0.6928965985076073,
+      "grad_norm": 0.3895666003227234,
+      "learning_rate": 1.1380624130498946e-05,
+      "loss": 1.165,
+      "num_input_tokens_seen": 561854100,
+      "step": 14300
+    },
+    {
+      "epoch": 0.6933811415834867,
+      "grad_norm": 0.40961936116218567,
+      "learning_rate": 1.134773948003976e-05,
+      "loss": 1.1452,
+      "num_input_tokens_seen": 562259444,
+      "step": 14310
+    },
+    {
+      "epoch": 0.6938656846593663,
+      "grad_norm": 0.4259204864501953,
+      "learning_rate": 1.1314888454623951e-05,
+      "loss": 1.1066,
+      "num_input_tokens_seen": 562672632,
+      "step": 14320
+    },
+    {
+      "epoch": 0.6943502277352457,
+      "grad_norm": 0.44359973073005676,
+      "learning_rate": 1.1282071135162498e-05,
+      "loss": 1.1223,
+      "num_input_tokens_seen": 563095012,
+      "step": 14330
+    },
+    {
+      "epoch": 0.6948347708111251,
+      "grad_norm": 0.4470522701740265,
+      "learning_rate": 1.1249287602483285e-05,
+      "loss": 1.1879,
+      "num_input_tokens_seen": 563479840,
+      "step": 14340
+    },
+    {
+      "epoch": 0.6953193138870045,
+      "grad_norm": 0.39908698201179504,
+      "learning_rate": 1.1216537937331028e-05,
+      "loss": 1.1204,
+      "num_input_tokens_seen": 563889984,
+      "step": 14350
+    },
+    {
+      "epoch": 0.695803856962884,
+      "grad_norm": 0.4189169704914093,
+      "learning_rate": 1.1183822220367014e-05,
+      "loss": 1.1517,
+      "num_input_tokens_seen": 564268180,
+      "step": 14360
+    },
+    {
+      "epoch": 0.6962884000387635,
+      "grad_norm": 0.3651569187641144,
+      "learning_rate": 1.1151140532168917e-05,
+      "loss": 1.1388,
+      "num_input_tokens_seen": 564649240,
+      "step": 14370
+    },
+    {
+      "epoch": 0.6967729431146429,
+      "grad_norm": 0.4237881898880005,
+      "learning_rate": 1.1118492953230594e-05,
+      "loss": 1.1211,
+      "num_input_tokens_seen": 565027528,
+      "step": 14380
+    },
+    {
+      "epoch": 0.6972574861905223,
+      "grad_norm": 0.41956138610839844,
+      "learning_rate": 1.1085879563961915e-05,
+      "loss": 1.1297,
+      "num_input_tokens_seen": 565436040,
+      "step": 14390
+    },
+    {
+      "epoch": 0.6977420292664018,
+      "grad_norm": 0.4476149082183838,
+      "learning_rate": 1.1053300444688502e-05,
+      "loss": 1.108,
+      "num_input_tokens_seen": 565816684,
+      "step": 14400
+    },
+    {
+      "epoch": 0.6982265723422812,
+      "grad_norm": 0.40152573585510254,
+      "learning_rate": 1.1020755675651605e-05,
+      "loss": 1.1131,
+      "num_input_tokens_seen": 566192928,
+      "step": 14410
+    },
+    {
+      "epoch": 0.6987111154181607,
+      "grad_norm": 0.4253696799278259,
+      "learning_rate": 1.0988245337007863e-05,
+      "loss": 1.1218,
+      "num_input_tokens_seen": 566589328,
+      "step": 14420
+    },
+    {
+      "epoch": 0.6991956584940401,
+      "grad_norm": 0.39764001965522766,
+      "learning_rate": 1.0955769508829103e-05,
+      "loss": 1.1141,
+      "num_input_tokens_seen": 566991652,
+      "step": 14430
+    },
+    {
+      "epoch": 0.6996802015699196,
+      "grad_norm": 0.3731628656387329,
+      "learning_rate": 1.0923328271102174e-05,
+      "loss": 1.1397,
+      "num_input_tokens_seen": 567388660,
+      "step": 14440
+    },
+    {
+      "epoch": 0.700164744645799,
+      "grad_norm": 0.4285131096839905,
+      "learning_rate": 1.0890921703728693e-05,
+      "loss": 1.1473,
+      "num_input_tokens_seen": 567783500,
+      "step": 14450
+    },
+    {
+      "epoch": 0.7006492877216784,
+      "grad_norm": 0.3825373649597168,
+      "learning_rate": 1.0858549886524944e-05,
+      "loss": 1.1373,
+      "num_input_tokens_seen": 568160888,
+      "step": 14460
+    },
+    {
+      "epoch": 0.7011338307975579,
+      "grad_norm": 0.38323846459388733,
+      "learning_rate": 1.0826212899221559e-05,
+      "loss": 1.1175,
+      "num_input_tokens_seen": 568542356,
+      "step": 14470
+    },
+    {
+      "epoch": 0.7016183738734374,
+      "grad_norm": 0.4034413695335388,
+      "learning_rate": 1.0793910821463424e-05,
+      "loss": 1.1396,
+      "num_input_tokens_seen": 568922996,
+      "step": 14480
+    },
+    {
+      "epoch": 0.7021029169493168,
+      "grad_norm": 0.4333634078502655,
+      "learning_rate": 1.076164373280944e-05,
+      "loss": 1.117,
+      "num_input_tokens_seen": 569333292,
+      "step": 14490
+    },
+    {
+      "epoch": 0.7025874600251962,
+      "grad_norm": 0.39302992820739746,
+      "learning_rate": 1.0729411712732319e-05,
+      "loss": 1.1303,
+      "num_input_tokens_seen": 569719056,
+      "step": 14500
+    },
+    {
+      "epoch": 0.7030720031010756,
+      "grad_norm": 0.39154550433158875,
+      "learning_rate": 1.0697214840618409e-05,
+      "loss": 1.2008,
+      "num_input_tokens_seen": 570111844,
+      "step": 14510
+    },
+    {
+      "epoch": 0.7035565461769552,
+      "grad_norm": 0.406896710395813,
+      "learning_rate": 1.0665053195767493e-05,
+      "loss": 1.1581,
+      "num_input_tokens_seen": 570534852,
+      "step": 14520
+    },
+    {
+      "epoch": 0.7040410892528346,
+      "grad_norm": 0.4043082892894745,
+      "learning_rate": 1.0632926857392567e-05,
+      "loss": 1.1445,
+      "num_input_tokens_seen": 570909832,
+      "step": 14530
+    },
+    {
+      "epoch": 0.704525632328714,
+      "grad_norm": 0.392946720123291,
+      "learning_rate": 1.0600835904619713e-05,
+      "loss": 1.1079,
+      "num_input_tokens_seen": 571297336,
+      "step": 14540
+    },
+    {
+      "epoch": 0.7050101754045934,
+      "grad_norm": 0.4595877528190613,
+      "learning_rate": 1.0568780416487811e-05,
+      "loss": 1.1688,
+      "num_input_tokens_seen": 571674808,
+      "step": 14550
+    },
+    {
+      "epoch": 0.705494718480473,
+      "grad_norm": 0.39945316314697266,
+      "learning_rate": 1.0536760471948423e-05,
+      "loss": 1.0807,
+      "num_input_tokens_seen": 572049484,
+      "step": 14560
+    },
+    {
+      "epoch": 0.7059792615563524,
+      "grad_norm": 0.4135000705718994,
+      "learning_rate": 1.0504776149865559e-05,
+      "loss": 1.0744,
+      "num_input_tokens_seen": 572426520,
+      "step": 14570
+    },
+    {
+      "epoch": 0.7064638046322318,
+      "grad_norm": 0.40249672532081604,
+      "learning_rate": 1.0472827529015494e-05,
+      "loss": 1.1261,
+      "num_input_tokens_seen": 572824836,
+      "step": 14580
+    },
+    {
+      "epoch": 0.7069483477081112,
+      "grad_norm": 0.4083980321884155,
+      "learning_rate": 1.0440914688086581e-05,
+      "loss": 1.1333,
+      "num_input_tokens_seen": 573191304,
+      "step": 14590
+    },
+    {
+      "epoch": 0.7074328907839907,
+      "grad_norm": 0.43228352069854736,
+      "learning_rate": 1.0409037705679018e-05,
+      "loss": 1.1248,
+      "num_input_tokens_seen": 573607132,
+      "step": 14600
+    },
+    {
+      "epoch": 0.7079174338598702,
+      "grad_norm": 0.3942347764968872,
+      "learning_rate": 1.0377196660304717e-05,
+      "loss": 1.1319,
+      "num_input_tokens_seen": 573989044,
+      "step": 14610
+    },
+    {
+      "epoch": 0.7084019769357496,
+      "grad_norm": 0.3767456114292145,
+      "learning_rate": 1.0345391630387064e-05,
+      "loss": 1.1459,
+      "num_input_tokens_seen": 574401880,
+      "step": 14620
+    },
+    {
+      "epoch": 0.708886520011629,
+      "grad_norm": 0.41538006067276,
+      "learning_rate": 1.0313622694260747e-05,
+      "loss": 1.136,
+      "num_input_tokens_seen": 574809836,
+      "step": 14630
+    },
+    {
+      "epoch": 0.7093710630875085,
+      "grad_norm": 0.3873327970504761,
+      "learning_rate": 1.0281889930171546e-05,
+      "loss": 1.2085,
+      "num_input_tokens_seen": 575187448,
+      "step": 14640
+    },
+    {
+      "epoch": 0.7098556061633879,
+      "grad_norm": 0.3875696063041687,
+      "learning_rate": 1.0250193416276171e-05,
+      "loss": 1.131,
+      "num_input_tokens_seen": 575619396,
+      "step": 14650
+    },
+    {
+      "epoch": 0.7103401492392674,
+      "grad_norm": 0.40218356251716614,
+      "learning_rate": 1.0218533230642005e-05,
+      "loss": 1.1638,
+      "num_input_tokens_seen": 576039756,
+      "step": 14660
+    },
+    {
+      "epoch": 0.7108246923151468,
+      "grad_norm": 0.360257089138031,
+      "learning_rate": 1.0186909451247023e-05,
+      "loss": 1.1354,
+      "num_input_tokens_seen": 576409768,
+      "step": 14670
+    },
+    {
+      "epoch": 0.7113092353910263,
+      "grad_norm": 0.3771324157714844,
+      "learning_rate": 1.0155322155979468e-05,
+      "loss": 1.1132,
+      "num_input_tokens_seen": 576806004,
+      "step": 14680
+    },
+    {
+      "epoch": 0.7117937784669057,
+      "grad_norm": 0.41122815012931824,
+      "learning_rate": 1.0123771422637757e-05,
+      "loss": 1.178,
+      "num_input_tokens_seen": 577197312,
+      "step": 14690
+    },
+    {
+      "epoch": 0.7122783215427851,
+      "grad_norm": 0.3926630914211273,
+      "learning_rate": 1.0092257328930255e-05,
+      "loss": 1.0902,
+      "num_input_tokens_seen": 577597560,
+      "step": 14700
+    },
+    {
+      "epoch": 0.7127628646186646,
+      "grad_norm": 0.4445978105068207,
+      "learning_rate": 1.0060779952475074e-05,
+      "loss": 1.1167,
+      "num_input_tokens_seen": 577999220,
+      "step": 14710
+    },
+    {
+      "epoch": 0.7132474076945441,
+      "grad_norm": 0.42254573106765747,
+      "learning_rate": 1.0029339370799912e-05,
+      "loss": 1.129,
+      "num_input_tokens_seen": 578383384,
+      "step": 14720
+    },
+    {
+      "epoch": 0.7137319507704235,
+      "grad_norm": 0.4229062795639038,
+      "learning_rate": 9.997935661341804e-06,
+      "loss": 1.1274,
+      "num_input_tokens_seen": 578758652,
+      "step": 14730
+    },
+    {
+      "epoch": 0.7142164938463029,
+      "grad_norm": 0.39834490418434143,
+      "learning_rate": 9.966568901447026e-06,
+      "loss": 1.1414,
+      "num_input_tokens_seen": 579152316,
+      "step": 14740
+    },
+    {
+      "epoch": 0.7147010369221823,
+      "grad_norm": 0.3739629089832306,
+      "learning_rate": 9.935239168370795e-06,
+      "loss": 1.1303,
+      "num_input_tokens_seen": 579581096,
+      "step": 14750
+    },
+    {
+      "epoch": 0.7151855799980619,
+      "grad_norm": 0.412434458732605,
+      "learning_rate": 9.903946539277162e-06,
+      "loss": 1.0987,
+      "num_input_tokens_seen": 579998920,
+      "step": 14760
+    },
+    {
+      "epoch": 0.7156701230739413,
+      "grad_norm": 0.401985764503479,
+      "learning_rate": 9.872691091238789e-06,
+      "loss": 1.0892,
+      "num_input_tokens_seen": 580392288,
+      "step": 14770
+    },
+    {
+      "epoch": 0.7161546661498207,
+      "grad_norm": 0.42549392580986023,
+      "learning_rate": 9.841472901236765e-06,
+      "loss": 1.1417,
+      "num_input_tokens_seen": 580778920,
+      "step": 14780
+    },
+    {
+      "epoch": 0.7166392092257001,
+      "grad_norm": 0.42221537232398987,
+      "learning_rate": 9.81029204616038e-06,
+      "loss": 1.1389,
+      "num_input_tokens_seen": 581160372,
+      "step": 14790
+    },
+    {
+      "epoch": 0.7171237523015797,
+      "grad_norm": 0.4084275960922241,
+      "learning_rate": 9.779148602807032e-06,
+      "loss": 1.1119,
+      "num_input_tokens_seen": 581518392,
+      "step": 14800
+    },
+    {
+      "epoch": 0.7176082953774591,
+      "grad_norm": 0.3924039900302887,
+      "learning_rate": 9.748042647881909e-06,
+      "loss": 1.1196,
+      "num_input_tokens_seen": 581916744,
+      "step": 14810
+    },
+    {
+      "epoch": 0.7180928384533385,
+      "grad_norm": 0.4364875555038452,
+      "learning_rate": 9.716974257997927e-06,
+      "loss": 1.1251,
+      "num_input_tokens_seen": 582306144,
+      "step": 14820
+    },
+    {
+      "epoch": 0.7185773815292179,
+      "grad_norm": 0.4177420735359192,
+      "learning_rate": 9.68594350967543e-06,
+      "loss": 1.1217,
+      "num_input_tokens_seen": 582701516,
+      "step": 14830
+    },
+    {
+      "epoch": 0.7190619246050974,
+      "grad_norm": 0.3962678015232086,
+      "learning_rate": 9.654950479342079e-06,
+      "loss": 1.1467,
+      "num_input_tokens_seen": 583110044,
+      "step": 14840
+    },
+    {
+      "epoch": 0.7195464676809769,
+      "grad_norm": 0.3973855972290039,
+      "learning_rate": 9.62399524333263e-06,
+      "loss": 1.0922,
+      "num_input_tokens_seen": 583517076,
+      "step": 14850
+    },
+    {
+      "epoch": 0.7200310107568563,
+      "grad_norm": 0.3978596329689026,
+      "learning_rate": 9.593077877888757e-06,
+      "loss": 1.1018,
+      "num_input_tokens_seen": 583892052,
+      "step": 14860
+    },
+    {
+      "epoch": 0.7205155538327357,
+      "grad_norm": 0.3927263617515564,
+      "learning_rate": 9.56219845915886e-06,
+      "loss": 1.1682,
+      "num_input_tokens_seen": 584282980,
+      "step": 14870
+    },
+    {
+      "epoch": 0.7210000969086152,
+      "grad_norm": 0.4256937801837921,
+      "learning_rate": 9.531357063197867e-06,
+      "loss": 1.1401,
+      "num_input_tokens_seen": 584680944,
+      "step": 14880
+    },
+    {
+      "epoch": 0.7214846399844946,
+      "grad_norm": 0.36345407366752625,
+      "learning_rate": 9.500553765967066e-06,
+      "loss": 1.1402,
+      "num_input_tokens_seen": 585085904,
+      "step": 14890
+    },
+    {
+      "epoch": 0.721969183060374,
+      "grad_norm": 0.38229337334632874,
+      "learning_rate": 9.46978864333391e-06,
+      "loss": 1.1655,
+      "num_input_tokens_seen": 585477160,
+      "step": 14900
+    },
+    {
+      "epoch": 0.7224537261362535,
+      "grad_norm": 0.41841405630111694,
+      "learning_rate": 9.439061771071824e-06,
+      "loss": 1.1179,
+      "num_input_tokens_seen": 585880592,
+      "step": 14910
+    },
+    {
+      "epoch": 0.722938269212133,
+      "grad_norm": 0.4083956480026245,
+      "learning_rate": 9.408373224860035e-06,
+      "loss": 1.1698,
+      "num_input_tokens_seen": 586284256,
+      "step": 14920
+    },
+    {
+      "epoch": 0.7234228122880124,
+      "grad_norm": 0.4051735997200012,
+      "learning_rate": 9.377723080283368e-06,
+      "loss": 1.1011,
+      "num_input_tokens_seen": 586669452,
+      "step": 14930
+    },
+    {
+      "epoch": 0.7239073553638918,
+      "grad_norm": 0.40003690123558044,
+      "learning_rate": 9.347111412832041e-06,
+      "loss": 1.0989,
+      "num_input_tokens_seen": 587054752,
+      "step": 14940
+    },
+    {
+      "epoch": 0.7243918984397713,
+      "grad_norm": 0.41723689436912537,
+      "learning_rate": 9.31653829790156e-06,
+      "loss": 1.1647,
+      "num_input_tokens_seen": 587436984,
+      "step": 14950
+    },
+    {
+      "epoch": 0.7248764415156508,
+      "grad_norm": 0.3850226402282715,
+      "learning_rate": 9.286003810792423e-06,
+      "loss": 1.1224,
+      "num_input_tokens_seen": 587849120,
+      "step": 14960
+    },
+    {
+      "epoch": 0.7253609845915302,
+      "grad_norm": 0.38576218485832214,
+      "learning_rate": 9.255508026710017e-06,
+      "loss": 1.1768,
+      "num_input_tokens_seen": 588230752,
+      "step": 14970
+    },
+    {
+      "epoch": 0.7258455276674096,
+      "grad_norm": 0.44978147745132446,
+      "learning_rate": 9.225051020764396e-06,
+      "loss": 1.1512,
+      "num_input_tokens_seen": 588620412,
+      "step": 14980
+    },
+    {
+      "epoch": 0.726330070743289,
+      "grad_norm": 0.4420928359031677,
+      "learning_rate": 9.194632867970115e-06,
+      "loss": 1.1517,
+      "num_input_tokens_seen": 588987072,
+      "step": 14990
+    },
+    {
+      "epoch": 0.7268146138191686,
+      "grad_norm": 0.40865370631217957,
+      "learning_rate": 9.16425364324602e-06,
+      "loss": 1.1299,
+      "num_input_tokens_seen": 589381344,
+      "step": 15000
+    },
+    {
+      "epoch": 0.727299156895048,
+      "grad_norm": 0.4068050682544708,
+      "learning_rate": 9.133913421415103e-06,
+      "loss": 1.1,
+      "num_input_tokens_seen": 589790080,
+      "step": 15010
+    },
+    {
+      "epoch": 0.7277836999709274,
+      "grad_norm": 0.37048131227493286,
+      "learning_rate": 9.10361227720425e-06,
+      "loss": 1.1035,
+      "num_input_tokens_seen": 590204672,
+      "step": 15020
+    },
+    {
+      "epoch": 0.7282682430468068,
+      "grad_norm": 0.3541117012500763,
+      "learning_rate": 9.073350285244142e-06,
+      "loss": 1.0962,
+      "num_input_tokens_seen": 590633676,
+      "step": 15030
+    },
+    {
+      "epoch": 0.7287527861226863,
+      "grad_norm": 0.4002024531364441,
+      "learning_rate": 9.04312752006901e-06,
+      "loss": 1.1322,
+      "num_input_tokens_seen": 590999504,
+      "step": 15040
+    },
+    {
+      "epoch": 0.7292373291985658,
+      "grad_norm": 0.4075698256492615,
+      "learning_rate": 9.012944056116477e-06,
+      "loss": 1.1658,
+      "num_input_tokens_seen": 591409344,
+      "step": 15050
+    },
+    {
+      "epoch": 0.7297218722744452,
+      "grad_norm": 0.41059961915016174,
+      "learning_rate": 8.982799967727374e-06,
+      "loss": 1.1439,
+      "num_input_tokens_seen": 591798840,
+      "step": 15060
+    },
+    {
+      "epoch": 0.7302064153503246,
+      "grad_norm": 0.3906279504299164,
+      "learning_rate": 8.952695329145517e-06,
+      "loss": 1.1835,
+      "num_input_tokens_seen": 592185512,
+      "step": 15070
+    },
+    {
+      "epoch": 0.7306909584262041,
+      "grad_norm": 0.41761744022369385,
+      "learning_rate": 8.922630214517618e-06,
+      "loss": 1.1068,
+      "num_input_tokens_seen": 592622168,
+      "step": 15080
+    },
+    {
+      "epoch": 0.7311755015020835,
+      "grad_norm": 0.408543199300766,
+      "learning_rate": 8.892604697892981e-06,
+      "loss": 1.1234,
+      "num_input_tokens_seen": 592987956,
+      "step": 15090
+    },
+    {
+      "epoch": 0.731660044577963,
+      "grad_norm": 0.4270235002040863,
+      "learning_rate": 8.862618853223442e-06,
+      "loss": 1.115,
+      "num_input_tokens_seen": 593378144,
+      "step": 15100
+    },
+    {
+      "epoch": 0.7321445876538424,
+      "grad_norm": 0.39898785948753357,
+      "learning_rate": 8.832672754363066e-06,
+      "loss": 1.1088,
+      "num_input_tokens_seen": 593782856,
+      "step": 15110
+    },
+    {
+      "epoch": 0.7326291307297219,
+      "grad_norm": 0.3919895887374878,
+      "learning_rate": 8.802766475068066e-06,
+      "loss": 1.0846,
+      "num_input_tokens_seen": 594160824,
+      "step": 15120
+    },
+    {
+      "epoch": 0.7331136738056013,
+      "grad_norm": 0.4317933917045593,
+      "learning_rate": 8.77290008899657e-06,
+      "loss": 1.0992,
+      "num_input_tokens_seen": 594564940,
+      "step": 15130
+    },
+    {
+      "epoch": 0.7335982168814807,
+      "grad_norm": 0.4144594073295593,
+      "learning_rate": 8.743073669708454e-06,
+      "loss": 1.1512,
+      "num_input_tokens_seen": 594969124,
+      "step": 15140
+    },
+    {
+      "epoch": 0.7340827599573602,
+      "grad_norm": 0.41548287868499756,
+      "learning_rate": 8.713287290665139e-06,
+      "loss": 1.109,
+      "num_input_tokens_seen": 595337996,
+      "step": 15150
+    },
+    {
+      "epoch": 0.7345673030332397,
+      "grad_norm": 0.396049827337265,
+      "learning_rate": 8.68354102522945e-06,
+      "loss": 1.1377,
+      "num_input_tokens_seen": 595759896,
+      "step": 15160
+    },
+    {
+      "epoch": 0.7350518461091191,
+      "grad_norm": 0.3859698176383972,
+      "learning_rate": 8.6538349466654e-06,
+      "loss": 1.1388,
+      "num_input_tokens_seen": 596196968,
+      "step": 15170
+    },
+    {
+      "epoch": 0.7355363891849985,
+      "grad_norm": 0.4152047634124756,
+      "learning_rate": 8.624169128138038e-06,
+      "loss": 1.1374,
+      "num_input_tokens_seen": 596608088,
+      "step": 15180
+    },
+    {
+      "epoch": 0.736020932260878,
+      "grad_norm": 0.3938763439655304,
+      "learning_rate": 8.594543642713245e-06,
+      "loss": 1.1061,
+      "num_input_tokens_seen": 596999532,
+      "step": 15190
+    },
+    {
+      "epoch": 0.7365054753367575,
+      "grad_norm": 0.3672252595424652,
+      "learning_rate": 8.564958563357543e-06,
+      "loss": 1.1148,
+      "num_input_tokens_seen": 597383556,
+      "step": 15200
+    },
+    {
+      "epoch": 0.7369900184126369,
+      "grad_norm": 0.4167373478412628,
+      "learning_rate": 8.535413962937983e-06,
+      "loss": 1.1183,
+      "num_input_tokens_seen": 597766768,
+      "step": 15210
+    },
+    {
+      "epoch": 0.7374745614885163,
+      "grad_norm": 0.395749568939209,
+      "learning_rate": 8.50590991422186e-06,
+      "loss": 1.1175,
+      "num_input_tokens_seen": 598156584,
+      "step": 15220
+    },
+    {
+      "epoch": 0.7379591045643957,
+      "grad_norm": 0.4062468409538269,
+      "learning_rate": 8.476446489876651e-06,
+      "loss": 1.1551,
+      "num_input_tokens_seen": 598562200,
+      "step": 15230
+    },
+    {
+      "epoch": 0.7384436476402753,
+      "grad_norm": 0.3954980969429016,
+      "learning_rate": 8.447023762469725e-06,
+      "loss": 1.1301,
+      "num_input_tokens_seen": 598961604,
+      "step": 15240
+    },
+    {
+      "epoch": 0.7389281907161547,
+      "grad_norm": 0.39155393838882446,
+      "learning_rate": 8.417641804468243e-06,
+      "loss": 1.1221,
+      "num_input_tokens_seen": 599354072,
+      "step": 15250
+    },
+    {
+      "epoch": 0.7394127337920341,
+      "grad_norm": 0.39971819519996643,
+      "learning_rate": 8.388300688238951e-06,
+      "loss": 1.127,
+      "num_input_tokens_seen": 599752608,
+      "step": 15260
+    },
+    {
+      "epoch": 0.7398972768679135,
+      "grad_norm": 0.4002969563007355,
+      "learning_rate": 8.359000486047994e-06,
+      "loss": 1.1052,
+      "num_input_tokens_seen": 600141532,
+      "step": 15270
+    },
+    {
+      "epoch": 0.740381819943793,
+      "grad_norm": 0.4160708487033844,
+      "learning_rate": 8.329741270060754e-06,
+      "loss": 1.0865,
+      "num_input_tokens_seen": 600526872,
+      "step": 15280
+    },
+    {
+      "epoch": 0.7408663630196725,
+      "grad_norm": 0.41618549823760986,
+      "learning_rate": 8.300523112341674e-06,
+      "loss": 1.1429,
+      "num_input_tokens_seen": 600899788,
+      "step": 15290
+    },
+    {
+      "epoch": 0.7413509060955519,
+      "grad_norm": 0.4318315386772156,
+      "learning_rate": 8.271346084854042e-06,
+      "loss": 1.0822,
+      "num_input_tokens_seen": 601307316,
+      "step": 15300
+    },
+    {
+      "epoch": 0.7418354491714313,
+      "grad_norm": 0.38331708312034607,
+      "learning_rate": 8.242210259459872e-06,
+      "loss": 1.1082,
+      "num_input_tokens_seen": 601712016,
+      "step": 15310
+    },
+    {
+      "epoch": 0.7423199922473108,
+      "grad_norm": 0.3820512294769287,
+      "learning_rate": 8.213115707919692e-06,
+      "loss": 1.1066,
+      "num_input_tokens_seen": 602117976,
+      "step": 15320
+    },
+    {
+      "epoch": 0.7428045353231902,
+      "grad_norm": 0.385928750038147,
+      "learning_rate": 8.18406250189237e-06,
+      "loss": 1.1274,
+      "num_input_tokens_seen": 602497772,
+      "step": 15330
+    },
+    {
+      "epoch": 0.7432890783990697,
+      "grad_norm": 0.40160906314849854,
+      "learning_rate": 8.15505071293495e-06,
+      "loss": 1.1348,
+      "num_input_tokens_seen": 602904540,
+      "step": 15340
+    },
+    {
+      "epoch": 0.7437736214749491,
+      "grad_norm": 0.3810666501522064,
+      "learning_rate": 8.126080412502437e-06,
+      "loss": 1.1229,
+      "num_input_tokens_seen": 603301296,
+      "step": 15350
+    },
+    {
+      "epoch": 0.7442581645508286,
+      "grad_norm": 0.37731239199638367,
+      "learning_rate": 8.097151671947709e-06,
+      "loss": 1.1554,
+      "num_input_tokens_seen": 603671376,
+      "step": 15360
+    },
+    {
+      "epoch": 0.744742707626708,
+      "grad_norm": 0.3665080666542053,
+      "learning_rate": 8.068264562521221e-06,
+      "loss": 1.1406,
+      "num_input_tokens_seen": 604069780,
+      "step": 15370
+    },
+    {
+      "epoch": 0.7452272507025874,
+      "grad_norm": 0.37638604640960693,
+      "learning_rate": 8.039419155370933e-06,
+      "loss": 1.2152,
+      "num_input_tokens_seen": 604460932,
+      "step": 15380
+    },
+    {
+      "epoch": 0.7457117937784669,
+      "grad_norm": 0.4055761396884918,
+      "learning_rate": 8.010615521542077e-06,
+      "loss": 1.075,
+      "num_input_tokens_seen": 604849832,
+      "step": 15390
+    },
+    {
+      "epoch": 0.7461963368543464,
+      "grad_norm": 0.41014060378074646,
+      "learning_rate": 7.981853731977005e-06,
+      "loss": 1.0986,
+      "num_input_tokens_seen": 605250684,
+      "step": 15400
+    },
+    {
+      "epoch": 0.7466808799302258,
+      "grad_norm": 0.44504401087760925,
+      "learning_rate": 7.953133857514999e-06,
+      "loss": 1.1245,
+      "num_input_tokens_seen": 605657060,
+      "step": 15410
+    },
+    {
+      "epoch": 0.7471654230061052,
+      "grad_norm": 0.3899877071380615,
+      "learning_rate": 7.92445596889213e-06,
+      "loss": 1.096,
+      "num_input_tokens_seen": 606046780,
+      "step": 15420
+    },
+    {
+      "epoch": 0.7476499660819846,
+      "grad_norm": 0.40705353021621704,
+      "learning_rate": 7.89582013674101e-06,
+      "loss": 1.1188,
+      "num_input_tokens_seen": 606446820,
+      "step": 15430
+    },
+    {
+      "epoch": 0.7481345091578642,
+      "grad_norm": 0.3865094482898712,
+      "learning_rate": 7.867226431590735e-06,
+      "loss": 1.1454,
+      "num_input_tokens_seen": 606828080,
+      "step": 15440
+    },
+    {
+      "epoch": 0.7486190522337436,
+      "grad_norm": 0.42546987533569336,
+      "learning_rate": 7.838674923866585e-06,
+      "loss": 1.0962,
+      "num_input_tokens_seen": 607213936,
+      "step": 15450
+    },
+    {
+      "epoch": 0.749103595309623,
+      "grad_norm": 0.39779505133628845,
+      "learning_rate": 7.81016568388993e-06,
+      "loss": 1.1534,
+      "num_input_tokens_seen": 607602128,
+      "step": 15460
+    },
+    {
+      "epoch": 0.7495881383855024,
+      "grad_norm": 0.43804842233657837,
+      "learning_rate": 7.78169878187805e-06,
+      "loss": 1.1426,
+      "num_input_tokens_seen": 607980840,
+      "step": 15470
+    },
+    {
+      "epoch": 0.750072681461382,
+      "grad_norm": 0.42345863580703735,
+      "learning_rate": 7.753274287943927e-06,
+      "loss": 1.1451,
+      "num_input_tokens_seen": 608389256,
+      "step": 15480
+    },
+    {
+      "epoch": 0.7505572245372614,
+      "grad_norm": 0.4104250371456146,
+      "learning_rate": 7.724892272096115e-06,
+      "loss": 1.1078,
+      "num_input_tokens_seen": 608763772,
+      "step": 15490
+    },
+    {
+      "epoch": 0.7510417676131408,
+      "grad_norm": 0.37295570969581604,
+      "learning_rate": 7.696552804238514e-06,
+      "loss": 1.1495,
+      "num_input_tokens_seen": 609154616,
+      "step": 15500
+    },
+    {
+      "epoch": 0.7515263106890202,
+      "grad_norm": 0.42971867322921753,
+      "learning_rate": 7.668255954170258e-06,
+      "loss": 1.1204,
+      "num_input_tokens_seen": 609563980,
+      "step": 15510
+    },
+    {
+      "epoch": 0.7520108537648997,
+      "grad_norm": 0.3849759101867676,
+      "learning_rate": 7.640001791585507e-06,
+      "loss": 1.1601,
+      "num_input_tokens_seen": 609984756,
+      "step": 15520
+    },
+    {
+      "epoch": 0.7524953968407792,
+      "grad_norm": 0.41971349716186523,
+      "learning_rate": 7.611790386073281e-06,
+      "loss": 1.1537,
+      "num_input_tokens_seen": 610384836,
+      "step": 15530
+    },
+    {
+      "epoch": 0.7529799399166586,
+      "grad_norm": 0.3679080903530121,
+      "learning_rate": 7.583621807117288e-06,
+      "loss": 1.1229,
+      "num_input_tokens_seen": 610787156,
+      "step": 15540
+    },
+    {
+      "epoch": 0.753464482992538,
+      "grad_norm": 0.3937299847602844,
+      "learning_rate": 7.555496124095776e-06,
+      "loss": 1.1227,
+      "num_input_tokens_seen": 611161580,
+      "step": 15550
+    },
+    {
+      "epoch": 0.7539490260684175,
+      "grad_norm": 0.4039948284626007,
+      "learning_rate": 7.527413406281294e-06,
+      "loss": 1.163,
+      "num_input_tokens_seen": 611556308,
+      "step": 15560
+    },
+    {
+      "epoch": 0.7544335691442969,
+      "grad_norm": 0.38725948333740234,
+      "learning_rate": 7.499373722840636e-06,
+      "loss": 1.1422,
+      "num_input_tokens_seen": 611967848,
+      "step": 15570
+    },
+    {
+      "epoch": 0.7549181122201764,
+      "grad_norm": 0.3972112536430359,
+      "learning_rate": 7.471377142834532e-06,
+      "loss": 1.102,
+      "num_input_tokens_seen": 612364408,
+      "step": 15580
+    },
+    {
+      "epoch": 0.7554026552960558,
+      "grad_norm": 0.39917463064193726,
+      "learning_rate": 7.443423735217622e-06,
+      "loss": 1.136,
+      "num_input_tokens_seen": 612775300,
+      "step": 15590
+    },
+    {
+      "epoch": 0.7558871983719353,
+      "grad_norm": 0.39816567301750183,
+      "learning_rate": 7.415513568838153e-06,
+      "loss": 1.1112,
+      "num_input_tokens_seen": 613151088,
+      "step": 15600
+    },
+    {
+      "epoch": 0.7563717414478147,
+      "grad_norm": 0.4336942136287689,
+      "learning_rate": 7.387646712437904e-06,
+      "loss": 1.1323,
+      "num_input_tokens_seen": 613546524,
+      "step": 15610
+    },
+    {
+      "epoch": 0.7568562845236941,
+      "grad_norm": 0.4093795716762543,
+      "learning_rate": 7.35982323465198e-06,
+      "loss": 1.081,
+      "num_input_tokens_seen": 613944468,
+      "step": 15620
+    },
+    {
+      "epoch": 0.7573408275995736,
+      "grad_norm": 0.3967530131340027,
+      "learning_rate": 7.332043204008621e-06,
+      "loss": 1.2024,
+      "num_input_tokens_seen": 614316784,
+      "step": 15630
+    },
+    {
+      "epoch": 0.7578253706754531,
+      "grad_norm": 0.3960801362991333,
+      "learning_rate": 7.304306688929102e-06,
+      "loss": 1.1074,
+      "num_input_tokens_seen": 614744716,
+      "step": 15640
+    },
+    {
+      "epoch": 0.7583099137513325,
+      "grad_norm": 0.40353167057037354,
+      "learning_rate": 7.2766137577274765e-06,
+      "loss": 1.1534,
+      "num_input_tokens_seen": 615162168,
+      "step": 15650
+    },
+    {
+      "epoch": 0.7587944568272119,
+      "grad_norm": 0.4028729200363159,
+      "learning_rate": 7.248964478610482e-06,
+      "loss": 1.1388,
+      "num_input_tokens_seen": 615552592,
+      "step": 15660
+    },
+    {
+      "epoch": 0.7592789999030913,
+      "grad_norm": 0.4119277596473694,
+      "learning_rate": 7.221358919677329e-06,
+      "loss": 1.1401,
+      "num_input_tokens_seen": 615985000,
+      "step": 15670
+    },
+    {
+      "epoch": 0.7597635429789709,
+      "grad_norm": 0.4111025035381317,
+      "learning_rate": 7.193797148919557e-06,
+      "loss": 1.1402,
+      "num_input_tokens_seen": 616372156,
+      "step": 15680
+    },
+    {
+      "epoch": 0.7602480860548503,
+      "grad_norm": 0.3870091438293457,
+      "learning_rate": 7.166279234220829e-06,
+      "loss": 1.1222,
+      "num_input_tokens_seen": 616794564,
+      "step": 15690
+    },
+    {
+      "epoch": 0.7607326291307297,
+      "grad_norm": 0.4028170704841614,
+      "learning_rate": 7.138805243356847e-06,
+      "loss": 1.1556,
+      "num_input_tokens_seen": 617166456,
+      "step": 15700
+    },
+    {
+      "epoch": 0.7612171722066091,
+      "grad_norm": 0.4021094739437103,
+      "learning_rate": 7.111375243995058e-06,
+      "loss": 1.1548,
+      "num_input_tokens_seen": 617553256,
+      "step": 15710
+    },
+    {
+      "epoch": 0.7617017152824886,
+      "grad_norm": 0.4055640399456024,
+      "learning_rate": 7.083989303694635e-06,
+      "loss": 1.1371,
+      "num_input_tokens_seen": 617951232,
+      "step": 15720
+    },
+    {
+      "epoch": 0.7621862583583681,
+      "grad_norm": 0.3847959339618683,
+      "learning_rate": 7.05664748990617e-06,
+      "loss": 1.1114,
+      "num_input_tokens_seen": 618343460,
+      "step": 15730
+    },
+    {
+      "epoch": 0.7626708014342475,
+      "grad_norm": 0.38287290930747986,
+      "learning_rate": 7.0293498699716105e-06,
+      "loss": 1.1387,
+      "num_input_tokens_seen": 618773472,
+      "step": 15740
+    },
+    {
+      "epoch": 0.7631553445101269,
+      "grad_norm": 0.41064536571502686,
+      "learning_rate": 7.0020965111240454e-06,
+      "loss": 1.1396,
+      "num_input_tokens_seen": 619175808,
+      "step": 15750
+    },
+    {
+      "epoch": 0.7636398875860064,
+      "grad_norm": 0.42379313707351685,
+      "learning_rate": 6.9748874804875516e-06,
+      "loss": 1.1246,
+      "num_input_tokens_seen": 619583744,
+      "step": 15760
+    },
+    {
+      "epoch": 0.7641244306618858,
+      "grad_norm": 0.4114890396595001,
+      "learning_rate": 6.947722845077032e-06,
+      "loss": 1.1252,
+      "num_input_tokens_seen": 619983760,
+      "step": 15770
+    },
+    {
+      "epoch": 0.7646089737377653,
+      "grad_norm": 0.4021173119544983,
+      "learning_rate": 6.9206026717980265e-06,
+      "loss": 1.1981,
+      "num_input_tokens_seen": 620365344,
+      "step": 15780
+    },
+    {
+      "epoch": 0.7650935168136447,
+      "grad_norm": 0.42168235778808594,
+      "learning_rate": 6.893527027446589e-06,
+      "loss": 1.0703,
+      "num_input_tokens_seen": 620746756,
+      "step": 15790
+    },
+    {
+      "epoch": 0.7655780598895242,
+      "grad_norm": 0.38521257042884827,
+      "learning_rate": 6.866495978709087e-06,
+      "loss": 1.1277,
+      "num_input_tokens_seen": 621139404,
+      "step": 15800
+    },
+    {
+      "epoch": 0.7660626029654036,
+      "grad_norm": 0.4103443920612335,
+      "learning_rate": 6.839509592162055e-06,
+      "loss": 1.1429,
+      "num_input_tokens_seen": 621564096,
+      "step": 15810
+    },
+    {
+      "epoch": 0.766547146041283,
+      "grad_norm": 0.36137986183166504,
+      "learning_rate": 6.8125679342720294e-06,
+      "loss": 1.1534,
+      "num_input_tokens_seen": 621947492,
+      "step": 15820
+    },
+    {
+      "epoch": 0.7670316891171625,
+      "grad_norm": 0.4212779104709625,
+      "learning_rate": 6.7856710713953805e-06,
+      "loss": 1.1169,
+      "num_input_tokens_seen": 622366700,
+      "step": 15830
+    },
+    {
+      "epoch": 0.767516232193042,
+      "grad_norm": 0.4411621689796448,
+      "learning_rate": 6.75881906977813e-06,
+      "loss": 1.1259,
+      "num_input_tokens_seen": 622768796,
+      "step": 15840
+    },
+    {
+      "epoch": 0.7680007752689214,
+      "grad_norm": 0.39934778213500977,
+      "learning_rate": 6.732011995555851e-06,
+      "loss": 1.1114,
+      "num_input_tokens_seen": 623147136,
+      "step": 15850
+    },
+    {
+      "epoch": 0.7684853183448008,
+      "grad_norm": 0.43150418996810913,
+      "learning_rate": 6.705249914753414e-06,
+      "loss": 1.1407,
+      "num_input_tokens_seen": 623483104,
+      "step": 15860
+    },
+    {
+      "epoch": 0.7689698614206802,
+      "grad_norm": 0.3736652731895447,
+      "learning_rate": 6.6785328932849e-06,
+      "loss": 1.1541,
+      "num_input_tokens_seen": 623845764,
+      "step": 15870
+    },
+    {
+      "epoch": 0.7694544044965598,
+      "grad_norm": 0.436292439699173,
+      "learning_rate": 6.65186099695341e-06,
+      "loss": 1.1381,
+      "num_input_tokens_seen": 624231876,
+      "step": 15880
+    },
+    {
+      "epoch": 0.7699389475724392,
+      "grad_norm": 0.44637537002563477,
+      "learning_rate": 6.6252342914508965e-06,
+      "loss": 1.1727,
+      "num_input_tokens_seen": 624626492,
+      "step": 15890
+    },
+    {
+      "epoch": 0.7704234906483186,
+      "grad_norm": 0.4382629990577698,
+      "learning_rate": 6.598652842358008e-06,
+      "loss": 1.0927,
+      "num_input_tokens_seen": 625027572,
+      "step": 15900
+    },
+    {
+      "epoch": 0.770908033724198,
+      "grad_norm": 0.37075766921043396,
+      "learning_rate": 6.572116715143939e-06,
+      "loss": 1.1196,
+      "num_input_tokens_seen": 625436036,
+      "step": 15910
+    },
+    {
+      "epoch": 0.7713925768000776,
+      "grad_norm": 0.3835522532463074,
+      "learning_rate": 6.545625975166231e-06,
+      "loss": 1.1169,
+      "num_input_tokens_seen": 625817956,
+      "step": 15920
+    },
+    {
+      "epoch": 0.771877119875957,
+      "grad_norm": 0.4104726016521454,
+      "learning_rate": 6.51918068767067e-06,
+      "loss": 1.1782,
+      "num_input_tokens_seen": 626184948,
+      "step": 15930
+    },
+    {
+      "epoch": 0.7723616629518364,
+      "grad_norm": 0.4460158944129944,
+      "learning_rate": 6.492780917791075e-06,
+      "loss": 1.1137,
+      "num_input_tokens_seen": 626596420,
+      "step": 15940
+    },
+    {
+      "epoch": 0.7728462060277158,
+      "grad_norm": 0.4014306664466858,
+      "learning_rate": 6.466426730549166e-06,
+      "loss": 1.154,
+      "num_input_tokens_seen": 627007456,
+      "step": 15950
+    },
+    {
+      "epoch": 0.7733307491035953,
+      "grad_norm": 0.3667517304420471,
+      "learning_rate": 6.440118190854394e-06,
+      "loss": 1.1304,
+      "num_input_tokens_seen": 627390576,
+      "step": 15960
+    },
+    {
+      "epoch": 0.7738152921794748,
+      "grad_norm": 0.39722877740859985,
+      "learning_rate": 6.41385536350376e-06,
+      "loss": 1.1548,
+      "num_input_tokens_seen": 627755000,
+      "step": 15970
+    },
+    {
+      "epoch": 0.7742998352553542,
+      "grad_norm": 0.41055047512054443,
+      "learning_rate": 6.387638313181721e-06,
+      "loss": 1.1473,
+      "num_input_tokens_seen": 628145832,
+      "step": 15980
+    },
+    {
+      "epoch": 0.7747843783312336,
+      "grad_norm": 0.4256828725337982,
+      "learning_rate": 6.3614671044599364e-06,
+      "loss": 1.1001,
+      "num_input_tokens_seen": 628544212,
+      "step": 15990
+    },
+    {
+      "epoch": 0.7752689214071131,
+      "grad_norm": 0.4094095230102539,
+      "learning_rate": 6.335341801797209e-06,
+      "loss": 1.1743,
+      "num_input_tokens_seen": 628937800,
+      "step": 16000
+    },
+    {
+      "epoch": 0.7752689214071131,
+      "eval_loss": 1.1244345903396606,
+      "eval_runtime": 5.2554,
+      "eval_samples_per_second": 28.542,
+      "eval_steps_per_second": 3.615,
+      "num_input_tokens_seen": 628937800,
+      "step": 16000
+    },
+    {
+      "epoch": 0.7757534644829925,
+      "grad_norm": 0.44040119647979736,
+      "learning_rate": 6.309262469539235e-06,
+      "loss": 1.0965,
+      "num_input_tokens_seen": 629321000,
+      "step": 16010
+    },
+    {
+      "epoch": 0.776238007558872,
+      "grad_norm": 0.3980225920677185,
+      "learning_rate": 6.283229171918506e-06,
+      "loss": 1.1363,
+      "num_input_tokens_seen": 629740836,
+      "step": 16020
+    },
+    {
+      "epoch": 0.7767225506347514,
+      "grad_norm": 0.4174353778362274,
+      "learning_rate": 6.257241973054132e-06,
+      "loss": 1.1164,
+      "num_input_tokens_seen": 630136212,
+      "step": 16030
+    },
+    {
+      "epoch": 0.7772070937106309,
+      "grad_norm": 0.44722503423690796,
+      "learning_rate": 6.231300936951686e-06,
+      "loss": 1.1585,
+      "num_input_tokens_seen": 630539300,
+      "step": 16040
+    },
+    {
+      "epoch": 0.7776916367865103,
+      "grad_norm": 0.41615843772888184,
+      "learning_rate": 6.205406127503021e-06,
+      "loss": 1.1508,
+      "num_input_tokens_seen": 630936060,
+      "step": 16050
+    },
+    {
+      "epoch": 0.7781761798623897,
+      "grad_norm": 0.39735278487205505,
+      "learning_rate": 6.1795576084861804e-06,
+      "loss": 1.0761,
+      "num_input_tokens_seen": 631334484,
+      "step": 16060
+    },
+    {
+      "epoch": 0.7786607229382693,
+      "grad_norm": 0.4064568281173706,
+      "learning_rate": 6.153755443565146e-06,
+      "loss": 1.1212,
+      "num_input_tokens_seen": 631740948,
+      "step": 16070
+    },
+    {
+      "epoch": 0.7791452660141487,
+      "grad_norm": 0.4038830101490021,
+      "learning_rate": 6.12799969628976e-06,
+      "loss": 1.1316,
+      "num_input_tokens_seen": 632141196,
+      "step": 16080
+    },
+    {
+      "epoch": 0.7796298090900281,
+      "grad_norm": 0.41059982776641846,
+      "learning_rate": 6.102290430095536e-06,
+      "loss": 1.1087,
+      "num_input_tokens_seen": 632532608,
+      "step": 16090
+    },
+    {
+      "epoch": 0.7801143521659075,
+      "grad_norm": 0.40716275572776794,
+      "learning_rate": 6.0766277083035035e-06,
+      "loss": 1.1119,
+      "num_input_tokens_seen": 632916136,
+      "step": 16100
+    },
+    {
+      "epoch": 0.780598895241787,
+      "grad_norm": 0.39503198862075806,
+      "learning_rate": 6.05101159412006e-06,
+      "loss": 1.1406,
+      "num_input_tokens_seen": 633327552,
+      "step": 16110
+    },
+    {
+      "epoch": 0.7810834383176665,
+      "grad_norm": 0.3686826229095459,
+      "learning_rate": 6.025442150636781e-06,
+      "loss": 1.0964,
+      "num_input_tokens_seen": 633698396,
+      "step": 16120
+    },
+    {
+      "epoch": 0.7815679813935459,
+      "grad_norm": 0.4384057819843292,
+      "learning_rate": 5.999919440830354e-06,
+      "loss": 1.1353,
+      "num_input_tokens_seen": 634084992,
+      "step": 16130
+    },
+    {
+      "epoch": 0.7820525244694253,
+      "grad_norm": 0.4173773527145386,
+      "learning_rate": 5.974443527562296e-06,
+      "loss": 1.0915,
+      "num_input_tokens_seen": 634487672,
+      "step": 16140
+    },
+    {
+      "epoch": 0.7825370675453048,
+      "grad_norm": 0.4255557358264923,
+      "learning_rate": 5.949014473578909e-06,
+      "loss": 1.133,
+      "num_input_tokens_seen": 634851792,
+      "step": 16150
+    },
+    {
+      "epoch": 0.7830216106211842,
+      "grad_norm": 0.40748724341392517,
+      "learning_rate": 5.923632341511071e-06,
+      "loss": 1.156,
+      "num_input_tokens_seen": 635237128,
+      "step": 16160
+    },
+    {
+      "epoch": 0.7835061536970637,
+      "grad_norm": 0.38821229338645935,
+      "learning_rate": 5.898297193874086e-06,
+      "loss": 1.1218,
+      "num_input_tokens_seen": 635617944,
+      "step": 16170
+    },
+    {
+      "epoch": 0.7839906967729431,
+      "grad_norm": 0.39625513553619385,
+      "learning_rate": 5.873009093067547e-06,
+      "loss": 1.1036,
+      "num_input_tokens_seen": 636028208,
+      "step": 16180
+    },
+    {
+      "epoch": 0.7844752398488226,
+      "grad_norm": 0.4132024943828583,
+      "learning_rate": 5.8477681013751724e-06,
+      "loss": 1.1117,
+      "num_input_tokens_seen": 636400024,
+      "step": 16190
+    },
+    {
+      "epoch": 0.784959782924702,
+      "grad_norm": 0.4017508029937744,
+      "learning_rate": 5.822574280964637e-06,
+      "loss": 1.1399,
+      "num_input_tokens_seen": 636803444,
+      "step": 16200
+    },
+    {
+      "epoch": 0.7854443260005814,
+      "grad_norm": 0.3757273852825165,
+      "learning_rate": 5.79742769388745e-06,
+      "loss": 1.1278,
+      "num_input_tokens_seen": 637208612,
+      "step": 16210
+    },
+    {
+      "epoch": 0.7859288690764609,
+      "grad_norm": 0.4408009648323059,
+      "learning_rate": 5.77232840207878e-06,
+      "loss": 1.1813,
+      "num_input_tokens_seen": 637569988,
+      "step": 16220
+    },
+    {
+      "epoch": 0.7864134121523404,
+      "grad_norm": 0.4416961371898651,
+      "learning_rate": 5.747276467357313e-06,
+      "loss": 1.0945,
+      "num_input_tokens_seen": 637969128,
+      "step": 16230
+    },
+    {
+      "epoch": 0.7868979552282198,
+      "grad_norm": 0.3773948848247528,
+      "learning_rate": 5.722271951425101e-06,
+      "loss": 1.097,
+      "num_input_tokens_seen": 638377296,
+      "step": 16240
+    },
+    {
+      "epoch": 0.7873824983040992,
+      "grad_norm": 0.4033990502357483,
+      "learning_rate": 5.6973149158673775e-06,
+      "loss": 1.1328,
+      "num_input_tokens_seen": 638763704,
+      "step": 16250
+    },
+    {
+      "epoch": 0.7878670413799786,
+      "grad_norm": 0.40168893337249756,
+      "learning_rate": 5.6724054221524825e-06,
+      "loss": 1.1358,
+      "num_input_tokens_seen": 639184012,
+      "step": 16260
+    },
+    {
+      "epoch": 0.7883515844558582,
+      "grad_norm": 0.4118176996707916,
+      "learning_rate": 5.647543531631611e-06,
+      "loss": 1.152,
+      "num_input_tokens_seen": 639586888,
+      "step": 16270
+    },
+    {
+      "epoch": 0.7888361275317376,
+      "grad_norm": 0.4122507572174072,
+      "learning_rate": 5.62272930553874e-06,
+      "loss": 1.1098,
+      "num_input_tokens_seen": 640002384,
+      "step": 16280
+    },
+    {
+      "epoch": 0.789320670607617,
+      "grad_norm": 0.42295822501182556,
+      "learning_rate": 5.597962804990453e-06,
+      "loss": 1.1196,
+      "num_input_tokens_seen": 640390420,
+      "step": 16290
+    },
+    {
+      "epoch": 0.7898052136834964,
+      "grad_norm": 0.37593498826026917,
+      "learning_rate": 5.573244090985777e-06,
+      "loss": 1.0817,
+      "num_input_tokens_seen": 640735112,
+      "step": 16300
+    },
+    {
+      "epoch": 0.790289756759376,
+      "grad_norm": 0.40528619289398193,
+      "learning_rate": 5.548573224406045e-06,
+      "loss": 1.163,
+      "num_input_tokens_seen": 641116036,
+      "step": 16310
+    },
+    {
+      "epoch": 0.7907742998352554,
+      "grad_norm": 0.4023245573043823,
+      "learning_rate": 5.523950266014754e-06,
+      "loss": 1.1551,
+      "num_input_tokens_seen": 641501348,
+      "step": 16320
+    },
+    {
+      "epoch": 0.7912588429111348,
+      "grad_norm": 0.3965412676334381,
+      "learning_rate": 5.499375276457372e-06,
+      "loss": 1.0989,
+      "num_input_tokens_seen": 641919160,
+      "step": 16330
+    },
+    {
+      "epoch": 0.7917433859870142,
+      "grad_norm": 0.4392828643321991,
+      "learning_rate": 5.4748483162612716e-06,
+      "loss": 1.1252,
+      "num_input_tokens_seen": 642318860,
+      "step": 16340
+    },
+    {
+      "epoch": 0.7922279290628937,
+      "grad_norm": 0.38923805952072144,
+      "learning_rate": 5.450369445835485e-06,
+      "loss": 1.1219,
+      "num_input_tokens_seen": 642734704,
+      "step": 16350
+    },
+    {
+      "epoch": 0.7927124721387732,
+      "grad_norm": 0.42891883850097656,
+      "learning_rate": 5.425938725470628e-06,
+      "loss": 1.1488,
+      "num_input_tokens_seen": 643108016,
+      "step": 16360
+    },
+    {
+      "epoch": 0.7931970152146526,
+      "grad_norm": 0.42537081241607666,
+      "learning_rate": 5.4015562153387126e-06,
+      "loss": 1.1634,
+      "num_input_tokens_seen": 643481944,
+      "step": 16370
+    },
+    {
+      "epoch": 0.793681558290532,
+      "grad_norm": 0.39477813243865967,
+      "learning_rate": 5.377221975493016e-06,
+      "loss": 1.091,
+      "num_input_tokens_seen": 643855924,
+      "step": 16380
+    },
+    {
+      "epoch": 0.7941661013664115,
+      "grad_norm": 0.4207918345928192,
+      "learning_rate": 5.352936065867931e-06,
+      "loss": 1.0861,
+      "num_input_tokens_seen": 644257816,
+      "step": 16390
+    },
+    {
+      "epoch": 0.7946506444422909,
+      "grad_norm": 0.44012096524238586,
+      "learning_rate": 5.328698546278798e-06,
+      "loss": 1.1237,
+      "num_input_tokens_seen": 644645828,
+      "step": 16400
+    },
+    {
+      "epoch": 0.7951351875181704,
+      "grad_norm": 0.4076230525970459,
+      "learning_rate": 5.30450947642179e-06,
+      "loss": 1.1243,
+      "num_input_tokens_seen": 645029716,
+      "step": 16410
+    },
+    {
+      "epoch": 0.7956197305940498,
+      "grad_norm": 0.3962908983230591,
+      "learning_rate": 5.280368915873749e-06,
+      "loss": 1.149,
+      "num_input_tokens_seen": 645399720,
+      "step": 16420
+    },
+    {
+      "epoch": 0.7961042736699293,
+      "grad_norm": 0.4037371575832367,
+      "learning_rate": 5.256276924092035e-06,
+      "loss": 1.1561,
+      "num_input_tokens_seen": 645804128,
+      "step": 16430
+    },
+    {
+      "epoch": 0.7965888167458087,
+      "grad_norm": 0.4168999493122101,
+      "learning_rate": 5.232233560414387e-06,
+      "loss": 1.1377,
+      "num_input_tokens_seen": 646212012,
+      "step": 16440
+    },
+    {
+      "epoch": 0.7970733598216881,
+      "grad_norm": 0.40669918060302734,
+      "learning_rate": 5.208238884058783e-06,
+      "loss": 1.1135,
+      "num_input_tokens_seen": 646620356,
+      "step": 16450
+    },
+    {
+      "epoch": 0.7975579028975676,
+      "grad_norm": 0.3902994692325592,
+      "learning_rate": 5.1842929541232544e-06,
+      "loss": 1.1916,
+      "num_input_tokens_seen": 646994340,
+      "step": 16460
+    },
+    {
+      "epoch": 0.7980424459734471,
+      "grad_norm": 0.451654314994812,
+      "learning_rate": 5.1603958295858205e-06,
+      "loss": 1.1252,
+      "num_input_tokens_seen": 647370752,
+      "step": 16470
+    },
+    {
+      "epoch": 0.7985269890493265,
+      "grad_norm": 0.39699918031692505,
+      "learning_rate": 5.136547569304246e-06,
+      "loss": 1.1475,
+      "num_input_tokens_seen": 647750888,
+      "step": 16480
+    },
+    {
+      "epoch": 0.7990115321252059,
+      "grad_norm": 0.37806710600852966,
+      "learning_rate": 5.112748232015993e-06,
+      "loss": 1.1349,
+      "num_input_tokens_seen": 648103956,
+      "step": 16490
+    },
+    {
+      "epoch": 0.7994960752010853,
+      "grad_norm": 0.38627612590789795,
+      "learning_rate": 5.088997876337981e-06,
+      "loss": 1.1531,
+      "num_input_tokens_seen": 648508460,
+      "step": 16500
+    },
+    {
+      "epoch": 0.7999806182769649,
+      "grad_norm": 0.414298951625824,
+      "learning_rate": 5.065296560766522e-06,
+      "loss": 1.1369,
+      "num_input_tokens_seen": 648868944,
+      "step": 16510
+    },
+    {
+      "epoch": 0.8004651613528443,
+      "grad_norm": 0.38814184069633484,
+      "learning_rate": 5.041644343677126e-06,
+      "loss": 1.1623,
+      "num_input_tokens_seen": 649301060,
+      "step": 16520
+    },
+    {
+      "epoch": 0.8009497044287237,
+      "grad_norm": 0.380260169506073,
+      "learning_rate": 5.018041283324387e-06,
+      "loss": 1.142,
+      "num_input_tokens_seen": 649700576,
+      "step": 16530
+    },
+    {
+      "epoch": 0.8014342475046031,
+      "grad_norm": 0.406630277633667,
+      "learning_rate": 4.994487437841827e-06,
+      "loss": 1.1287,
+      "num_input_tokens_seen": 650094908,
+      "step": 16540
+    },
+    {
+      "epoch": 0.8019187905804827,
+      "grad_norm": 0.4198836088180542,
+      "learning_rate": 4.9709828652417385e-06,
+      "loss": 1.1155,
+      "num_input_tokens_seen": 650509396,
+      "step": 16550
+    },
+    {
+      "epoch": 0.8024033336563621,
+      "grad_norm": 0.3846551775932312,
+      "learning_rate": 4.947527623415071e-06,
+      "loss": 1.084,
+      "num_input_tokens_seen": 650893528,
+      "step": 16560
+    },
+    {
+      "epoch": 0.8028878767322415,
+      "grad_norm": 0.36685115098953247,
+      "learning_rate": 4.924121770131274e-06,
+      "loss": 1.1465,
+      "num_input_tokens_seen": 651297116,
+      "step": 16570
+    },
+    {
+      "epoch": 0.8033724198081209,
+      "grad_norm": 0.4003463387489319,
+      "learning_rate": 4.900765363038151e-06,
+      "loss": 1.1613,
+      "num_input_tokens_seen": 651677048,
+      "step": 16580
+    },
+    {
+      "epoch": 0.8038569628840004,
+      "grad_norm": 0.4008733928203583,
+      "learning_rate": 4.87745845966171e-06,
+      "loss": 1.1023,
+      "num_input_tokens_seen": 652077724,
+      "step": 16590
+    },
+    {
+      "epoch": 0.8043415059598799,
+      "grad_norm": 0.40616297721862793,
+      "learning_rate": 4.854201117406066e-06,
+      "loss": 1.1272,
+      "num_input_tokens_seen": 652466264,
+      "step": 16600
+    },
+    {
+      "epoch": 0.8048260490357593,
+      "grad_norm": 0.41137826442718506,
+      "learning_rate": 4.830993393553215e-06,
+      "loss": 1.1,
+      "num_input_tokens_seen": 652835720,
+      "step": 16610
+    },
+    {
+      "epoch": 0.8053105921116387,
+      "grad_norm": 0.44767943024635315,
+      "learning_rate": 4.807835345263009e-06,
+      "loss": 1.1096,
+      "num_input_tokens_seen": 653218356,
+      "step": 16620
+    },
+    {
+      "epoch": 0.8057951351875182,
+      "grad_norm": 0.38211873173713684,
+      "learning_rate": 4.784727029572894e-06,
+      "loss": 1.1586,
+      "num_input_tokens_seen": 653630732,
+      "step": 16630
+    },
+    {
+      "epoch": 0.8062796782633976,
+      "grad_norm": 0.4150942862033844,
+      "learning_rate": 4.761668503397851e-06,
+      "loss": 1.1344,
+      "num_input_tokens_seen": 654032908,
+      "step": 16640
+    },
+    {
+      "epoch": 0.806764221339277,
+      "grad_norm": 0.3996165096759796,
+      "learning_rate": 4.738659823530236e-06,
+      "loss": 1.1124,
+      "num_input_tokens_seen": 654438804,
+      "step": 16650
+    },
+    {
+      "epoch": 0.8072487644151565,
+      "grad_norm": 0.3866037428379059,
+      "learning_rate": 4.715701046639623e-06,
+      "loss": 1.1109,
+      "num_input_tokens_seen": 654823000,
+      "step": 16660
+    },
+    {
+      "epoch": 0.807733307491036,
+      "grad_norm": 0.4089403450489044,
+      "learning_rate": 4.6927922292726886e-06,
+      "loss": 1.1196,
+      "num_input_tokens_seen": 655212888,
+      "step": 16670
+    },
+    {
+      "epoch": 0.8082178505669154,
+      "grad_norm": 0.43001115322113037,
+      "learning_rate": 4.669933427853043e-06,
+      "loss": 1.0934,
+      "num_input_tokens_seen": 655622528,
+      "step": 16680
+    },
+    {
+      "epoch": 0.8087023936427948,
+      "grad_norm": 0.40845033526420593,
+      "learning_rate": 4.647124698681127e-06,
+      "loss": 1.1133,
+      "num_input_tokens_seen": 656007064,
+      "step": 16690
+    },
+    {
+      "epoch": 0.8091869367186743,
+      "grad_norm": 0.38767552375793457,
+      "learning_rate": 4.624366097934046e-06,
+      "loss": 1.1088,
+      "num_input_tokens_seen": 656399380,
+      "step": 16700
+    },
+    {
+      "epoch": 0.8096714797945538,
+      "grad_norm": 0.39583420753479004,
+      "learning_rate": 4.60165768166545e-06,
+      "loss": 1.1914,
+      "num_input_tokens_seen": 656798972,
+      "step": 16710
+    },
+    {
+      "epoch": 0.8101560228704332,
+      "grad_norm": 0.4240066707134247,
+      "learning_rate": 4.578999505805362e-06,
+      "loss": 1.1416,
+      "num_input_tokens_seen": 657198652,
+      "step": 16720
+    },
+    {
+      "epoch": 0.8106405659463126,
+      "grad_norm": 0.3843323290348053,
+      "learning_rate": 4.556391626160109e-06,
+      "loss": 1.1275,
+      "num_input_tokens_seen": 657637540,
+      "step": 16730
+    },
+    {
+      "epoch": 0.811125109022192,
+      "grad_norm": 0.41490888595581055,
+      "learning_rate": 4.533834098412082e-06,
+      "loss": 1.154,
+      "num_input_tokens_seen": 658012976,
+      "step": 16740
+    },
+    {
+      "epoch": 0.8116096520980716,
+      "grad_norm": 0.4508754014968872,
+      "learning_rate": 4.511326978119718e-06,
+      "loss": 1.1442,
+      "num_input_tokens_seen": 658403516,
+      "step": 16750
+    },
+    {
+      "epoch": 0.812094195173951,
+      "grad_norm": 0.3963550925254822,
+      "learning_rate": 4.488870320717251e-06,
+      "loss": 1.1307,
+      "num_input_tokens_seen": 658799828,
+      "step": 16760
+    },
+    {
+      "epoch": 0.8125787382498304,
+      "grad_norm": 0.4240289032459259,
+      "learning_rate": 4.466464181514657e-06,
+      "loss": 1.1333,
+      "num_input_tokens_seen": 659188772,
+      "step": 16770
+    },
+    {
+      "epoch": 0.8130632813257098,
+      "grad_norm": 0.4394795596599579,
+      "learning_rate": 4.444108615697476e-06,
+      "loss": 1.1301,
+      "num_input_tokens_seen": 659605584,
+      "step": 16780
+    },
+    {
+      "epoch": 0.8135478244015893,
+      "grad_norm": 0.391284704208374,
+      "learning_rate": 4.421803678326691e-06,
+      "loss": 1.0913,
+      "num_input_tokens_seen": 660023780,
+      "step": 16790
+    },
+    {
+      "epoch": 0.8140323674774688,
+      "grad_norm": 0.37516680359840393,
+      "learning_rate": 4.399549424338589e-06,
+      "loss": 1.0831,
+      "num_input_tokens_seen": 660418884,
+      "step": 16800
+    },
+    {
+      "epoch": 0.8145169105533482,
+      "grad_norm": 0.415872186422348,
+      "learning_rate": 4.3773459085446275e-06,
+      "loss": 1.1123,
+      "num_input_tokens_seen": 660814748,
+      "step": 16810
+    },
+    {
+      "epoch": 0.8150014536292276,
+      "grad_norm": 0.4023378789424896,
+      "learning_rate": 4.355193185631287e-06,
+      "loss": 1.087,
+      "num_input_tokens_seen": 661192768,
+      "step": 16820
+    },
+    {
+      "epoch": 0.8154859967051071,
+      "grad_norm": 0.3767937421798706,
+      "learning_rate": 4.333091310159956e-06,
+      "loss": 1.1545,
+      "num_input_tokens_seen": 661606668,
+      "step": 16830
+    },
+    {
+      "epoch": 0.8159705397809865,
+      "grad_norm": 0.5043548941612244,
+      "learning_rate": 4.311040336566791e-06,
+      "loss": 1.1445,
+      "num_input_tokens_seen": 661982448,
+      "step": 16840
+    },
+    {
+      "epoch": 0.816455082856866,
+      "grad_norm": 0.3812048137187958,
+      "learning_rate": 4.289040319162569e-06,
+      "loss": 1.136,
+      "num_input_tokens_seen": 662373168,
+      "step": 16850
+    },
+    {
+      "epoch": 0.8169396259327454,
+      "grad_norm": 0.40610408782958984,
+      "learning_rate": 4.267091312132576e-06,
+      "loss": 1.1151,
+      "num_input_tokens_seen": 662758704,
+      "step": 16860
+    },
+    {
+      "epoch": 0.8174241690086249,
+      "grad_norm": 0.4081581234931946,
+      "learning_rate": 4.245193369536437e-06,
+      "loss": 1.1532,
+      "num_input_tokens_seen": 663150068,
+      "step": 16870
+    },
+    {
+      "epoch": 0.8179087120845043,
+      "grad_norm": 0.41858088970184326,
+      "learning_rate": 4.2233465453080486e-06,
+      "loss": 1.1054,
+      "num_input_tokens_seen": 663562480,
+      "step": 16880
+    },
+    {
+      "epoch": 0.8183932551603837,
+      "grad_norm": 0.45081600546836853,
+      "learning_rate": 4.20155089325536e-06,
+      "loss": 1.128,
+      "num_input_tokens_seen": 663941916,
+      "step": 16890
+    },
+    {
+      "epoch": 0.8188777982362632,
+      "grad_norm": 0.41167008876800537,
+      "learning_rate": 4.17980646706031e-06,
+      "loss": 1.1354,
+      "num_input_tokens_seen": 664345336,
+      "step": 16900
+    },
+    {
+      "epoch": 0.8193623413121427,
+      "grad_norm": 0.39505717158317566,
+      "learning_rate": 4.158113320278667e-06,
+      "loss": 1.084,
+      "num_input_tokens_seen": 664762920,
+      "step": 16910
+    },
+    {
+      "epoch": 0.8198468843880221,
+      "grad_norm": 0.3788187503814697,
+      "learning_rate": 4.136471506339901e-06,
+      "loss": 1.1064,
+      "num_input_tokens_seen": 665156684,
+      "step": 16920
+    },
+    {
+      "epoch": 0.8203314274639015,
+      "grad_norm": 0.43902385234832764,
+      "learning_rate": 4.114881078547042e-06,
+      "loss": 1.1317,
+      "num_input_tokens_seen": 665557436,
+      "step": 16930
+    },
+    {
+      "epoch": 0.820815970539781,
+      "grad_norm": 0.3830069303512573,
+      "learning_rate": 4.093342090076571e-06,
+      "loss": 1.127,
+      "num_input_tokens_seen": 665945288,
+      "step": 16940
+    },
+    {
+      "epoch": 0.8213005136156605,
+      "grad_norm": 0.4254535734653473,
+      "learning_rate": 4.071854593978253e-06,
+      "loss": 1.1208,
+      "num_input_tokens_seen": 666344140,
+      "step": 16950
+    },
+    {
+      "epoch": 0.8217850566915399,
+      "grad_norm": 0.38880154490470886,
+      "learning_rate": 4.050418643175067e-06,
+      "loss": 1.1566,
+      "num_input_tokens_seen": 666728028,
+      "step": 16960
+    },
+    {
+      "epoch": 0.8222695997674193,
+      "grad_norm": 0.4109187126159668,
+      "learning_rate": 4.029034290462996e-06,
+      "loss": 1.1369,
+      "num_input_tokens_seen": 667095588,
+      "step": 16970
+    },
+    {
+      "epoch": 0.8227541428432987,
+      "grad_norm": 0.36832395195961,
+      "learning_rate": 4.0077015885109676e-06,
+      "loss": 1.1458,
+      "num_input_tokens_seen": 667490212,
+      "step": 16980
+    },
+    {
+      "epoch": 0.8232386859191783,
+      "grad_norm": 0.38334590196609497,
+      "learning_rate": 3.986420589860682e-06,
+      "loss": 1.1087,
+      "num_input_tokens_seen": 667892264,
+      "step": 16990
+    },
+    {
+      "epoch": 0.8237232289950577,
+      "grad_norm": 0.41069987416267395,
+      "learning_rate": 3.9651913469265e-06,
+      "loss": 1.0849,
+      "num_input_tokens_seen": 668289636,
+      "step": 17000
+    },
+    {
+      "epoch": 0.8242077720709371,
+      "grad_norm": 0.3793525695800781,
+      "learning_rate": 3.944013911995317e-06,
+      "loss": 1.1524,
+      "num_input_tokens_seen": 668676344,
+      "step": 17010
+    },
+    {
+      "epoch": 0.8246923151468165,
+      "grad_norm": 0.41535207629203796,
+      "learning_rate": 3.922888337226399e-06,
+      "loss": 1.1035,
+      "num_input_tokens_seen": 669042236,
+      "step": 17020
+    },
+    {
+      "epoch": 0.825176858222696,
+      "grad_norm": 0.3818122148513794,
+      "learning_rate": 3.90181467465133e-06,
+      "loss": 1.103,
+      "num_input_tokens_seen": 669438068,
+      "step": 17030
+    },
+    {
+      "epoch": 0.8256614012985755,
+      "grad_norm": 0.4072776436805725,
+      "learning_rate": 3.880792976173788e-06,
+      "loss": 1.1103,
+      "num_input_tokens_seen": 669833020,
+      "step": 17040
+    },
+    {
+      "epoch": 0.8261459443744549,
+      "grad_norm": 0.3956544101238251,
+      "learning_rate": 3.859823293569495e-06,
+      "loss": 1.1063,
+      "num_input_tokens_seen": 670234476,
+      "step": 17050
+    },
+    {
+      "epoch": 0.8266304874503343,
+      "grad_norm": 0.4025129973888397,
+      "learning_rate": 3.838905678486049e-06,
+      "loss": 1.0968,
+      "num_input_tokens_seen": 670625508,
+      "step": 17060
+    },
+    {
+      "epoch": 0.8271150305262138,
+      "grad_norm": 0.48538658022880554,
+      "learning_rate": 3.818040182442814e-06,
+      "loss": 1.1978,
+      "num_input_tokens_seen": 671013016,
+      "step": 17070
+    },
+    {
+      "epoch": 0.8275995736020932,
+      "grad_norm": 0.39329686760902405,
+      "learning_rate": 3.7972268568307685e-06,
+      "loss": 1.1016,
+      "num_input_tokens_seen": 671418600,
+      "step": 17080
+    },
+    {
+      "epoch": 0.8280841166779727,
+      "grad_norm": 0.39455854892730713,
+      "learning_rate": 3.776465752912428e-06,
+      "loss": 1.16,
+      "num_input_tokens_seen": 671802384,
+      "step": 17090
+    },
+    {
+      "epoch": 0.8285686597538521,
+      "grad_norm": 0.5405839681625366,
+      "learning_rate": 3.7557569218216516e-06,
+      "loss": 1.164,
+      "num_input_tokens_seen": 672200552,
+      "step": 17100
+    },
+    {
+      "epoch": 0.8290532028297316,
+      "grad_norm": 0.369508296251297,
+      "learning_rate": 3.735100414563594e-06,
+      "loss": 1.1282,
+      "num_input_tokens_seen": 672618664,
+      "step": 17110
+    },
+    {
+      "epoch": 0.829537745905611,
+      "grad_norm": 0.41533350944519043,
+      "learning_rate": 3.7144962820144928e-06,
+      "loss": 1.1445,
+      "num_input_tokens_seen": 673004068,
+      "step": 17120
+    },
+    {
+      "epoch": 0.8300222889814904,
+      "grad_norm": 0.38036444783210754,
+      "learning_rate": 3.6939445749216235e-06,
+      "loss": 1.1596,
+      "num_input_tokens_seen": 673389044,
+      "step": 17130
+    },
+    {
+      "epoch": 0.8305068320573699,
+      "grad_norm": 0.37696197628974915,
+      "learning_rate": 3.6734453439031257e-06,
+      "loss": 1.1471,
+      "num_input_tokens_seen": 673724844,
+      "step": 17140
+    },
+    {
+      "epoch": 0.8309913751332494,
+      "grad_norm": 0.40430694818496704,
+      "learning_rate": 3.65299863944788e-06,
+      "loss": 1.095,
+      "num_input_tokens_seen": 674106168,
+      "step": 17150
+    },
+    {
+      "epoch": 0.8314759182091288,
+      "grad_norm": 0.468234658241272,
+      "learning_rate": 3.6326045119154327e-06,
+      "loss": 1.108,
+      "num_input_tokens_seen": 674469340,
+      "step": 17160
+    },
+    {
+      "epoch": 0.8319604612850082,
+      "grad_norm": 0.40100011229515076,
+      "learning_rate": 3.612263011535791e-06,
+      "loss": 1.1609,
+      "num_input_tokens_seen": 674828188,
+      "step": 17170
+    },
+    {
+      "epoch": 0.8324450043608876,
+      "grad_norm": 0.42651060223579407,
+      "learning_rate": 3.5919741884093723e-06,
+      "loss": 1.1289,
+      "num_input_tokens_seen": 675221792,
+      "step": 17180
+    },
+    {
+      "epoch": 0.8329295474367672,
+      "grad_norm": 0.3936256468296051,
+      "learning_rate": 3.5717380925068405e-06,
+      "loss": 1.1308,
+      "num_input_tokens_seen": 675644508,
+      "step": 17190
+    },
+    {
+      "epoch": 0.8334140905126466,
+      "grad_norm": 0.41212528944015503,
+      "learning_rate": 3.5515547736689995e-06,
+      "loss": 1.0939,
+      "num_input_tokens_seen": 676038672,
+      "step": 17200
+    },
+    {
+      "epoch": 0.833898633588526,
+      "grad_norm": 0.4407350420951843,
+      "learning_rate": 3.531424281606663e-06,
+      "loss": 1.1637,
+      "num_input_tokens_seen": 676431380,
+      "step": 17210
+    },
+    {
+      "epoch": 0.8343831766644054,
+      "grad_norm": 0.37873637676239014,
+      "learning_rate": 3.511346665900536e-06,
+      "loss": 1.1442,
+      "num_input_tokens_seen": 676821844,
+      "step": 17220
+    },
+    {
+      "epoch": 0.834867719740285,
+      "grad_norm": 0.39286068081855774,
+      "learning_rate": 3.491321976001072e-06,
+      "loss": 1.1215,
+      "num_input_tokens_seen": 677199996,
+      "step": 17230
+    },
+    {
+      "epoch": 0.8353522628161644,
+      "grad_norm": 0.42986762523651123,
+      "learning_rate": 3.471350261228412e-06,
+      "loss": 1.0996,
+      "num_input_tokens_seen": 677592120,
+      "step": 17240
+    },
+    {
+      "epoch": 0.8358368058920438,
+      "grad_norm": 0.4484102725982666,
+      "learning_rate": 3.451431570772179e-06,
+      "loss": 1.1268,
+      "num_input_tokens_seen": 677999524,
+      "step": 17250
+    },
+    {
+      "epoch": 0.8363213489679232,
+      "grad_norm": 0.407656729221344,
+      "learning_rate": 3.431565953691418e-06,
+      "loss": 1.1505,
+      "num_input_tokens_seen": 678368908,
+      "step": 17260
+    },
+    {
+      "epoch": 0.8368058920438027,
+      "grad_norm": 0.398433119058609,
+      "learning_rate": 3.4117534589144547e-06,
+      "loss": 1.1351,
+      "num_input_tokens_seen": 678739932,
+      "step": 17270
+    },
+    {
+      "epoch": 0.8372904351196822,
+      "grad_norm": 0.47509506344795227,
+      "learning_rate": 3.3919941352387768e-06,
+      "loss": 1.0815,
+      "num_input_tokens_seen": 679131488,
+      "step": 17280
+    },
+    {
+      "epoch": 0.8377749781955616,
+      "grad_norm": 0.3671634793281555,
+      "learning_rate": 3.372288031330917e-06,
+      "loss": 1.1059,
+      "num_input_tokens_seen": 679542760,
+      "step": 17290
+    },
+    {
+      "epoch": 0.838259521271441,
+      "grad_norm": 0.4009820818901062,
+      "learning_rate": 3.3526351957263115e-06,
+      "loss": 1.1132,
+      "num_input_tokens_seen": 679961904,
+      "step": 17300
+    },
+    {
+      "epoch": 0.8387440643473205,
+      "grad_norm": 0.3820185363292694,
+      "learning_rate": 3.3330356768292215e-06,
+      "loss": 1.0903,
+      "num_input_tokens_seen": 680347620,
+      "step": 17310
+    },
+    {
+      "epoch": 0.8392286074231999,
+      "grad_norm": 0.45867863297462463,
+      "learning_rate": 3.3134895229125772e-06,
+      "loss": 1.1581,
+      "num_input_tokens_seen": 680716452,
+      "step": 17320
+    },
+    {
+      "epoch": 0.8397131504990794,
+      "grad_norm": 0.4050085246562958,
+      "learning_rate": 3.293996782117881e-06,
+      "loss": 1.0828,
+      "num_input_tokens_seen": 681101364,
+      "step": 17330
+    },
+    {
+      "epoch": 0.8401976935749588,
+      "grad_norm": 0.39128220081329346,
+      "learning_rate": 3.2745575024550695e-06,
+      "loss": 1.0908,
+      "num_input_tokens_seen": 681520008,
+      "step": 17340
+    },
+    {
+      "epoch": 0.8406822366508383,
+      "grad_norm": 0.47006210684776306,
+      "learning_rate": 3.2551717318024255e-06,
+      "loss": 1.1409,
+      "num_input_tokens_seen": 681899268,
+      "step": 17350
+    },
+    {
+      "epoch": 0.8411667797267177,
+      "grad_norm": 0.41448378562927246,
+      "learning_rate": 3.235839517906411e-06,
+      "loss": 1.1815,
+      "num_input_tokens_seen": 682286376,
+      "step": 17360
+    },
+    {
+      "epoch": 0.8416513228025971,
+      "grad_norm": 0.3959016501903534,
+      "learning_rate": 3.216560908381616e-06,
+      "loss": 1.1594,
+      "num_input_tokens_seen": 682681272,
+      "step": 17370
+    },
+    {
+      "epoch": 0.8421358658784766,
+      "grad_norm": 0.39144328236579895,
+      "learning_rate": 3.1973359507105645e-06,
+      "loss": 1.0814,
+      "num_input_tokens_seen": 683085540,
+      "step": 17380
+    },
+    {
+      "epoch": 0.8426204089543561,
+      "grad_norm": 0.5121538043022156,
+      "learning_rate": 3.1781646922436848e-06,
+      "loss": 1.1284,
+      "num_input_tokens_seen": 683457100,
+      "step": 17390
+    },
+    {
+      "epoch": 0.8431049520302355,
+      "grad_norm": 0.4063737094402313,
+      "learning_rate": 3.1590471801991012e-06,
+      "loss": 1.1375,
+      "num_input_tokens_seen": 683831048,
+      "step": 17400
+    },
+    {
+      "epoch": 0.8435894951061149,
+      "grad_norm": 0.3836385905742645,
+      "learning_rate": 3.1399834616625907e-06,
+      "loss": 1.1183,
+      "num_input_tokens_seen": 684224324,
+      "step": 17410
+    },
+    {
+      "epoch": 0.8440740381819943,
+      "grad_norm": 0.39683645963668823,
+      "learning_rate": 3.120973583587425e-06,
+      "loss": 1.1197,
+      "num_input_tokens_seen": 684624492,
+      "step": 17420
+    },
+    {
+      "epoch": 0.8445585812578739,
+      "grad_norm": 0.3896249234676361,
+      "learning_rate": 3.102017592794279e-06,
+      "loss": 1.0956,
+      "num_input_tokens_seen": 685029460,
+      "step": 17430
+    },
+    {
+      "epoch": 0.8450431243337533,
+      "grad_norm": 0.39281994104385376,
+      "learning_rate": 3.0831155359710927e-06,
+      "loss": 1.158,
+      "num_input_tokens_seen": 685406276,
+      "step": 17440
+    },
+    {
+      "epoch": 0.8455276674096327,
+      "grad_norm": 0.4067038893699646,
+      "learning_rate": 3.0642674596729785e-06,
+      "loss": 1.1395,
+      "num_input_tokens_seen": 685780380,
+      "step": 17450
+    },
+    {
+      "epoch": 0.8460122104855121,
+      "grad_norm": 0.3911641836166382,
+      "learning_rate": 3.0454734103220942e-06,
+      "loss": 1.1158,
+      "num_input_tokens_seen": 686162356,
+      "step": 17460
+    },
+    {
+      "epoch": 0.8464967535613916,
+      "grad_norm": 0.3925408720970154,
+      "learning_rate": 3.026733434207532e-06,
+      "loss": 1.1467,
+      "num_input_tokens_seen": 686553108,
+      "step": 17470
+    },
+    {
+      "epoch": 0.8469812966372711,
+      "grad_norm": 0.42923226952552795,
+      "learning_rate": 3.008047577485204e-06,
+      "loss": 1.1584,
+      "num_input_tokens_seen": 686914652,
+      "step": 17480
+    },
+    {
+      "epoch": 0.8474658397131505,
+      "grad_norm": 0.4102732837200165,
+      "learning_rate": 2.98941588617771e-06,
+      "loss": 1.1051,
+      "num_input_tokens_seen": 687275992,
+      "step": 17490
+    },
+    {
+      "epoch": 0.8479503827890299,
+      "grad_norm": 0.41153576970100403,
+      "learning_rate": 2.970838406174284e-06,
+      "loss": 1.1446,
+      "num_input_tokens_seen": 687689920,
+      "step": 17500
+    },
+    {
+      "epoch": 0.8484349258649094,
+      "grad_norm": 0.4032898545265198,
+      "learning_rate": 2.952315183230589e-06,
+      "loss": 1.1283,
+      "num_input_tokens_seen": 688080568,
+      "step": 17510
+    },
+    {
+      "epoch": 0.8489194689407888,
+      "grad_norm": 0.4179829955101013,
+      "learning_rate": 2.9338462629687034e-06,
+      "loss": 1.1031,
+      "num_input_tokens_seen": 688453284,
+      "step": 17520
+    },
+    {
+      "epoch": 0.8494040120166683,
+      "grad_norm": 0.4297243058681488,
+      "learning_rate": 2.915431690876916e-06,
+      "loss": 1.1442,
+      "num_input_tokens_seen": 688838776,
+      "step": 17530
+    },
+    {
+      "epoch": 0.8498885550925477,
+      "grad_norm": 0.4191533029079437,
+      "learning_rate": 2.8970715123096877e-06,
+      "loss": 1.1041,
+      "num_input_tokens_seen": 689225084,
+      "step": 17540
+    },
+    {
+      "epoch": 0.8503730981684272,
+      "grad_norm": 0.4235047996044159,
+      "learning_rate": 2.8787657724875016e-06,
+      "loss": 1.0696,
+      "num_input_tokens_seen": 689607636,
+      "step": 17550
+    },
+    {
+      "epoch": 0.8508576412443066,
+      "grad_norm": 0.409962922334671,
+      "learning_rate": 2.860514516496754e-06,
+      "loss": 1.1092,
+      "num_input_tokens_seen": 689951392,
+      "step": 17560
+    },
+    {
+      "epoch": 0.851342184320186,
+      "grad_norm": 0.4196973443031311,
+      "learning_rate": 2.8423177892896585e-06,
+      "loss": 1.0901,
+      "num_input_tokens_seen": 690348064,
+      "step": 17570
+    },
+    {
+      "epoch": 0.8518267273960655,
+      "grad_norm": 0.3807368278503418,
+      "learning_rate": 2.8241756356841233e-06,
+      "loss": 1.1281,
+      "num_input_tokens_seen": 690749348,
+      "step": 17580
+    },
+    {
+      "epoch": 0.852311270471945,
+      "grad_norm": 0.40411800146102905,
+      "learning_rate": 2.806088100363635e-06,
+      "loss": 1.0834,
+      "num_input_tokens_seen": 691138376,
+      "step": 17590
+    },
+    {
+      "epoch": 0.8527958135478244,
+      "grad_norm": 0.39498665928840637,
+      "learning_rate": 2.7880552278771703e-06,
+      "loss": 1.0952,
+      "num_input_tokens_seen": 691567648,
+      "step": 17600
+    },
+    {
+      "epoch": 0.8532803566237038,
+      "grad_norm": 0.434725284576416,
+      "learning_rate": 2.7700770626390677e-06,
+      "loss": 1.1152,
+      "num_input_tokens_seen": 691920228,
+      "step": 17610
+    },
+    {
+      "epoch": 0.8537648996995832,
+      "grad_norm": 0.3900477886199951,
+      "learning_rate": 2.7521536489289233e-06,
+      "loss": 1.1517,
+      "num_input_tokens_seen": 692294800,
+      "step": 17620
+    },
+    {
+      "epoch": 0.8542494427754628,
+      "grad_norm": 0.42599180340766907,
+      "learning_rate": 2.7342850308914843e-06,
+      "loss": 1.1066,
+      "num_input_tokens_seen": 692670100,
+      "step": 17630
+    },
+    {
+      "epoch": 0.8547339858513422,
+      "grad_norm": 0.416204571723938,
+      "learning_rate": 2.716471252536526e-06,
+      "loss": 1.0739,
+      "num_input_tokens_seen": 693061084,
+      "step": 17640
+    },
+    {
+      "epoch": 0.8552185289272216,
+      "grad_norm": 0.40145227313041687,
+      "learning_rate": 2.6987123577387833e-06,
+      "loss": 1.1394,
+      "num_input_tokens_seen": 693458880,
+      "step": 17650
+    },
+    {
+      "epoch": 0.855703072003101,
+      "grad_norm": 0.38669687509536743,
+      "learning_rate": 2.6810083902377825e-06,
+      "loss": 1.1175,
+      "num_input_tokens_seen": 693867900,
+      "step": 17660
+    },
+    {
+      "epoch": 0.8561876150789806,
+      "grad_norm": 0.42774924635887146,
+      "learning_rate": 2.663359393637785e-06,
+      "loss": 1.1,
+      "num_input_tokens_seen": 694285092,
+      "step": 17670
+    },
+    {
+      "epoch": 0.85667215815486,
+      "grad_norm": 0.39861494302749634,
+      "learning_rate": 2.645765411407655e-06,
+      "loss": 1.1243,
+      "num_input_tokens_seen": 694666360,
+      "step": 17680
+    },
+    {
+      "epoch": 0.8571567012307394,
+      "grad_norm": 0.3979116678237915,
+      "learning_rate": 2.6282264868807637e-06,
+      "loss": 1.1226,
+      "num_input_tokens_seen": 695052444,
+      "step": 17690
+    },
+    {
+      "epoch": 0.8576412443066188,
+      "grad_norm": 0.4196152985095978,
+      "learning_rate": 2.61074266325487e-06,
+      "loss": 1.1283,
+      "num_input_tokens_seen": 695441168,
+      "step": 17700
+    },
+    {
+      "epoch": 0.8581257873824983,
+      "grad_norm": 0.41752830147743225,
+      "learning_rate": 2.59331398359203e-06,
+      "loss": 1.0917,
+      "num_input_tokens_seen": 695864664,
+      "step": 17710
+    },
+    {
+      "epoch": 0.8586103304583778,
+      "grad_norm": 0.41962143778800964,
+      "learning_rate": 2.5759404908184654e-06,
+      "loss": 1.1223,
+      "num_input_tokens_seen": 696267288,
+      "step": 17720
+    },
+    {
+      "epoch": 0.8590948735342572,
+      "grad_norm": 0.4000556170940399,
+      "learning_rate": 2.5586222277244887e-06,
+      "loss": 1.1531,
+      "num_input_tokens_seen": 696650996,
+      "step": 17730
+    },
+    {
+      "epoch": 0.8595794166101366,
+      "grad_norm": 0.4215715229511261,
+      "learning_rate": 2.541359236964386e-06,
+      "loss": 1.0838,
+      "num_input_tokens_seen": 697023960,
+      "step": 17740
+    },
+    {
+      "epoch": 0.8600639596860161,
+      "grad_norm": 0.3923567533493042,
+      "learning_rate": 2.5241515610562983e-06,
+      "loss": 1.1086,
+      "num_input_tokens_seen": 697411632,
+      "step": 17750
+    },
+    {
+      "epoch": 0.8605485027618955,
+      "grad_norm": 0.39392712712287903,
+      "learning_rate": 2.506999242382141e-06,
+      "loss": 1.1011,
+      "num_input_tokens_seen": 697800768,
+      "step": 17760
+    },
+    {
+      "epoch": 0.861033045837775,
+      "grad_norm": 0.39785829186439514,
+      "learning_rate": 2.489902323187465e-06,
+      "loss": 1.0928,
+      "num_input_tokens_seen": 698174508,
+      "step": 17770
+    },
+    {
+      "epoch": 0.8615175889136544,
+      "grad_norm": 0.4119921028614044,
+      "learning_rate": 2.472860845581404e-06,
+      "loss": 1.1081,
+      "num_input_tokens_seen": 698560436,
+      "step": 17780
+    },
+    {
+      "epoch": 0.8620021319895339,
+      "grad_norm": 0.4029790461063385,
+      "learning_rate": 2.455874851536516e-06,
+      "loss": 1.1645,
+      "num_input_tokens_seen": 698924028,
+      "step": 17790
+    },
+    {
+      "epoch": 0.8624866750654133,
+      "grad_norm": 0.4168151915073395,
+      "learning_rate": 2.4389443828887166e-06,
+      "loss": 1.1385,
+      "num_input_tokens_seen": 699315512,
+      "step": 17800
+    },
+    {
+      "epoch": 0.8629712181412927,
+      "grad_norm": 0.44244882464408875,
+      "learning_rate": 2.422069481337161e-06,
+      "loss": 1.1355,
+      "num_input_tokens_seen": 699695032,
+      "step": 17810
+    },
+    {
+      "epoch": 0.8634557612171722,
+      "grad_norm": 0.3886070251464844,
+      "learning_rate": 2.405250188444147e-06,
+      "loss": 1.1019,
+      "num_input_tokens_seen": 700102396,
+      "step": 17820
+    },
+    {
+      "epoch": 0.8639403042930517,
+      "grad_norm": 0.4062063992023468,
+      "learning_rate": 2.3884865456350103e-06,
+      "loss": 1.1314,
+      "num_input_tokens_seen": 700467740,
+      "step": 17830
+    },
+    {
+      "epoch": 0.8644248473689311,
+      "grad_norm": 0.4632991552352905,
+      "learning_rate": 2.371778594198021e-06,
+      "loss": 1.1259,
+      "num_input_tokens_seen": 700860544,
+      "step": 17840
+    },
+    {
+      "epoch": 0.8649093904448105,
+      "grad_norm": 0.3729296624660492,
+      "learning_rate": 2.355126375284272e-06,
+      "loss": 1.1139,
+      "num_input_tokens_seen": 701260456,
+      "step": 17850
+    },
+    {
+      "epoch": 0.8653939335206899,
+      "grad_norm": 0.3984206020832062,
+      "learning_rate": 2.338529929907618e-06,
+      "loss": 1.1629,
+      "num_input_tokens_seen": 701665356,
+      "step": 17860
+    },
+    {
+      "epoch": 0.8658784765965695,
+      "grad_norm": 0.39754825830459595,
+      "learning_rate": 2.321989298944513e-06,
+      "loss": 1.1423,
+      "num_input_tokens_seen": 702049100,
+      "step": 17870
+    },
+    {
+      "epoch": 0.8663630196724489,
+      "grad_norm": 0.4021519720554352,
+      "learning_rate": 2.305504523133964e-06,
+      "loss": 1.1645,
+      "num_input_tokens_seen": 702418804,
+      "step": 17880
+    },
+    {
+      "epoch": 0.8668475627483283,
+      "grad_norm": 0.41587236523628235,
+      "learning_rate": 2.2890756430773956e-06,
+      "loss": 1.1121,
+      "num_input_tokens_seen": 702806248,
+      "step": 17890
+    },
+    {
+      "epoch": 0.8673321058242077,
+      "grad_norm": 0.38616082072257996,
+      "learning_rate": 2.272702699238574e-06,
+      "loss": 1.1448,
+      "num_input_tokens_seen": 703198880,
+      "step": 17900
+    },
+    {
+      "epoch": 0.8678166489000873,
+      "grad_norm": 0.4032549560070038,
+      "learning_rate": 2.2563857319434945e-06,
+      "loss": 1.1484,
+      "num_input_tokens_seen": 703603228,
+      "step": 17910
+    },
+    {
+      "epoch": 0.8683011919759667,
+      "grad_norm": 0.4154454171657562,
+      "learning_rate": 2.2401247813802652e-06,
+      "loss": 1.0953,
+      "num_input_tokens_seen": 703977824,
+      "step": 17920
+    },
+    {
+      "epoch": 0.8687857350518461,
+      "grad_norm": 0.3659140169620514,
+      "learning_rate": 2.223919887599063e-06,
+      "loss": 1.136,
+      "num_input_tokens_seen": 704375196,
+      "step": 17930
+    },
+    {
+      "epoch": 0.8692702781277255,
+      "grad_norm": 0.4203336834907532,
+      "learning_rate": 2.2077710905119618e-06,
+      "loss": 1.1147,
+      "num_input_tokens_seen": 704771996,
+      "step": 17940
+    },
+    {
+      "epoch": 0.869754821203605,
+      "grad_norm": 0.4060530364513397,
+      "learning_rate": 2.191678429892893e-06,
+      "loss": 1.1429,
+      "num_input_tokens_seen": 705149948,
+      "step": 17950
+    },
+    {
+      "epoch": 0.8702393642794845,
+      "grad_norm": 0.4052405059337616,
+      "learning_rate": 2.175641945377524e-06,
+      "loss": 1.1261,
+      "num_input_tokens_seen": 705545528,
+      "step": 17960
+    },
+    {
+      "epoch": 0.8707239073553639,
+      "grad_norm": 0.39759910106658936,
+      "learning_rate": 2.1596616764631595e-06,
+      "loss": 1.1454,
+      "num_input_tokens_seen": 705922212,
+      "step": 17970
+    },
+    {
+      "epoch": 0.8712084504312433,
+      "grad_norm": 0.3922360837459564,
+      "learning_rate": 2.143737662508635e-06,
+      "loss": 1.0771,
+      "num_input_tokens_seen": 706339472,
+      "step": 17980
+    },
+    {
+      "epoch": 0.8716929935071228,
+      "grad_norm": 0.39637112617492676,
+      "learning_rate": 2.127869942734262e-06,
+      "loss": 1.0886,
+      "num_input_tokens_seen": 706709976,
+      "step": 17990
+    },
+    {
+      "epoch": 0.8721775365830022,
+      "grad_norm": 0.36679452657699585,
+      "learning_rate": 2.112058556221663e-06,
+      "loss": 1.137,
+      "num_input_tokens_seen": 707117360,
+      "step": 18000
+    },
+    {
+      "epoch": 0.8721775365830022,
+      "eval_loss": 1.1178553104400635,
+      "eval_runtime": 5.0781,
+      "eval_samples_per_second": 29.539,
+      "eval_steps_per_second": 3.742,
+      "num_input_tokens_seen": 707117360,
+      "step": 18000
+    },
+    {
+      "epoch": 0.8726620796588817,
+      "grad_norm": 0.3815447986125946,
+      "learning_rate": 2.0963035419137577e-06,
+      "loss": 1.1119,
+      "num_input_tokens_seen": 707514656,
+      "step": 18010
+    },
+    {
+      "epoch": 0.8731466227347611,
+      "grad_norm": 0.39872461557388306,
+      "learning_rate": 2.0806049386145774e-06,
+      "loss": 1.1278,
+      "num_input_tokens_seen": 707917760,
+      "step": 18020
+    },
+    {
+      "epoch": 0.8736311658106406,
+      "grad_norm": 0.3805377781391144,
+      "learning_rate": 2.0649627849892466e-06,
+      "loss": 1.1064,
+      "num_input_tokens_seen": 708298628,
+      "step": 18030
+    },
+    {
+      "epoch": 0.87411570888652,
+      "grad_norm": 0.3966962695121765,
+      "learning_rate": 2.0493771195638443e-06,
+      "loss": 1.1155,
+      "num_input_tokens_seen": 708713480,
+      "step": 18040
+    },
+    {
+      "epoch": 0.8746002519623994,
+      "grad_norm": 0.42251577973365784,
+      "learning_rate": 2.033847980725323e-06,
+      "loss": 1.1489,
+      "num_input_tokens_seen": 709099180,
+      "step": 18050
+    },
+    {
+      "epoch": 0.8750847950382789,
+      "grad_norm": 0.4105037450790405,
+      "learning_rate": 2.018375406721415e-06,
+      "loss": 1.1305,
+      "num_input_tokens_seen": 709499996,
+      "step": 18060
+    },
+    {
+      "epoch": 0.8755693381141584,
+      "grad_norm": 0.4050074517726898,
+      "learning_rate": 2.0029594356605286e-06,
+      "loss": 1.1278,
+      "num_input_tokens_seen": 709902500,
+      "step": 18070
+    },
+    {
+      "epoch": 0.8760538811900378,
+      "grad_norm": 0.3788783550262451,
+      "learning_rate": 1.9876001055116664e-06,
+      "loss": 1.0709,
+      "num_input_tokens_seen": 710294012,
+      "step": 18080
+    },
+    {
+      "epoch": 0.8765384242659172,
+      "grad_norm": 0.4039725959300995,
+      "learning_rate": 1.9722974541043244e-06,
+      "loss": 1.131,
+      "num_input_tokens_seen": 710669864,
+      "step": 18090
+    },
+    {
+      "epoch": 0.8770229673417966,
+      "grad_norm": 0.3935058116912842,
+      "learning_rate": 1.957051519128403e-06,
+      "loss": 1.092,
+      "num_input_tokens_seen": 711071604,
+      "step": 18100
+    },
+    {
+      "epoch": 0.8775075104176762,
+      "grad_norm": 0.42256423830986023,
+      "learning_rate": 1.9418623381341094e-06,
+      "loss": 1.1371,
+      "num_input_tokens_seen": 711476612,
+      "step": 18110
+    },
+    {
+      "epoch": 0.8779920534935556,
+      "grad_norm": 0.3739723265171051,
+      "learning_rate": 1.926729948531872e-06,
+      "loss": 1.14,
+      "num_input_tokens_seen": 711872752,
+      "step": 18120
+    },
+    {
+      "epoch": 0.878476596569435,
+      "grad_norm": 0.3799099624156952,
+      "learning_rate": 1.9116543875922294e-06,
+      "loss": 1.1055,
+      "num_input_tokens_seen": 712244912,
+      "step": 18130
+    },
+    {
+      "epoch": 0.8789611396453144,
+      "grad_norm": 0.3699980080127716,
+      "learning_rate": 1.8966356924457833e-06,
+      "loss": 1.1191,
+      "num_input_tokens_seen": 712647364,
+      "step": 18140
+    },
+    {
+      "epoch": 0.8794456827211939,
+      "grad_norm": 0.3966718316078186,
+      "learning_rate": 1.881673900083042e-06,
+      "loss": 1.1611,
+      "num_input_tokens_seen": 713028800,
+      "step": 18150
+    },
+    {
+      "epoch": 0.8799302257970734,
+      "grad_norm": 0.3989785313606262,
+      "learning_rate": 1.8667690473543858e-06,
+      "loss": 1.1063,
+      "num_input_tokens_seen": 713412044,
+      "step": 18160
+    },
+    {
+      "epoch": 0.8804147688729528,
+      "grad_norm": 0.3920852541923523,
+      "learning_rate": 1.8519211709699475e-06,
+      "loss": 1.1163,
+      "num_input_tokens_seen": 713796584,
+      "step": 18170
+    },
+    {
+      "epoch": 0.8808993119488322,
+      "grad_norm": 0.4311419129371643,
+      "learning_rate": 1.837130307499535e-06,
+      "loss": 1.0843,
+      "num_input_tokens_seen": 714199172,
+      "step": 18180
+    },
+    {
+      "epoch": 0.8813838550247117,
+      "grad_norm": 0.3938891887664795,
+      "learning_rate": 1.822396493372533e-06,
+      "loss": 1.0989,
+      "num_input_tokens_seen": 714590532,
+      "step": 18190
+    },
+    {
+      "epoch": 0.8818683981005911,
+      "grad_norm": 0.42306891083717346,
+      "learning_rate": 1.807719764877805e-06,
+      "loss": 1.0779,
+      "num_input_tokens_seen": 715001368,
+      "step": 18200
+    },
+    {
+      "epoch": 0.8823529411764706,
+      "grad_norm": 0.39101406931877136,
+      "learning_rate": 1.793100158163627e-06,
+      "loss": 1.1353,
+      "num_input_tokens_seen": 715383004,
+      "step": 18210
+    },
+    {
+      "epoch": 0.88283748425235,
+      "grad_norm": 0.3895913362503052,
+      "learning_rate": 1.7785377092375848e-06,
+      "loss": 1.1042,
+      "num_input_tokens_seen": 715769884,
+      "step": 18220
+    },
+    {
+      "epoch": 0.8833220273282295,
+      "grad_norm": 0.40741434693336487,
+      "learning_rate": 1.7640324539664827e-06,
+      "loss": 1.1079,
+      "num_input_tokens_seen": 716157804,
+      "step": 18230
+    },
+    {
+      "epoch": 0.8838065704041089,
+      "grad_norm": 0.40758630633354187,
+      "learning_rate": 1.749584428076262e-06,
+      "loss": 1.1415,
+      "num_input_tokens_seen": 716535668,
+      "step": 18240
+    },
+    {
+      "epoch": 0.8842911134799883,
+      "grad_norm": 0.39152994751930237,
+      "learning_rate": 1.7351936671519104e-06,
+      "loss": 1.1472,
+      "num_input_tokens_seen": 716917736,
+      "step": 18250
+    },
+    {
+      "epoch": 0.8847756565558678,
+      "grad_norm": 0.3884604275226593,
+      "learning_rate": 1.7208602066373592e-06,
+      "loss": 1.1329,
+      "num_input_tokens_seen": 717307888,
+      "step": 18260
+    },
+    {
+      "epoch": 0.8852601996317473,
+      "grad_norm": 0.4423676133155823,
+      "learning_rate": 1.706584081835444e-06,
+      "loss": 1.154,
+      "num_input_tokens_seen": 717697032,
+      "step": 18270
+    },
+    {
+      "epoch": 0.8857447427076267,
+      "grad_norm": 0.39519426226615906,
+      "learning_rate": 1.6923653279077468e-06,
+      "loss": 1.106,
+      "num_input_tokens_seen": 718093096,
+      "step": 18280
+    },
+    {
+      "epoch": 0.8862292857835061,
+      "grad_norm": 0.4418719708919525,
+      "learning_rate": 1.6782039798745791e-06,
+      "loss": 1.1005,
+      "num_input_tokens_seen": 718458596,
+      "step": 18290
+    },
+    {
+      "epoch": 0.8867138288593855,
+      "grad_norm": 0.39423951506614685,
+      "learning_rate": 1.6641000726148353e-06,
+      "loss": 1.109,
+      "num_input_tokens_seen": 718874968,
+      "step": 18300
+    },
+    {
+      "epoch": 0.8871983719352651,
+      "grad_norm": 0.40183213353157043,
+      "learning_rate": 1.650053640865959e-06,
+      "loss": 1.1351,
+      "num_input_tokens_seen": 719282808,
+      "step": 18310
+    },
+    {
+      "epoch": 0.8876829150111445,
+      "grad_norm": 0.3977072238922119,
+      "learning_rate": 1.6360647192238176e-06,
+      "loss": 1.1096,
+      "num_input_tokens_seen": 719670048,
+      "step": 18320
+    },
+    {
+      "epoch": 0.8881674580870239,
+      "grad_norm": 0.4405307173728943,
+      "learning_rate": 1.6221333421426483e-06,
+      "loss": 1.1682,
+      "num_input_tokens_seen": 720046844,
+      "step": 18330
+    },
+    {
+      "epoch": 0.8886520011629033,
+      "grad_norm": 0.38534072041511536,
+      "learning_rate": 1.6082595439349368e-06,
+      "loss": 1.1598,
+      "num_input_tokens_seen": 720440272,
+      "step": 18340
+    },
+    {
+      "epoch": 0.8891365442387829,
+      "grad_norm": 0.3926476836204529,
+      "learning_rate": 1.5944433587713693e-06,
+      "loss": 1.1201,
+      "num_input_tokens_seen": 720832848,
+      "step": 18350
+    },
+    {
+      "epoch": 0.8896210873146623,
+      "grad_norm": 0.38580575585365295,
+      "learning_rate": 1.5806848206807362e-06,
+      "loss": 1.1723,
+      "num_input_tokens_seen": 721201392,
+      "step": 18360
+    },
+    {
+      "epoch": 0.8901056303905417,
+      "grad_norm": 0.41746583580970764,
+      "learning_rate": 1.566983963549834e-06,
+      "loss": 1.0699,
+      "num_input_tokens_seen": 721622692,
+      "step": 18370
+    },
+    {
+      "epoch": 0.8905901734664212,
+      "grad_norm": 0.3924805223941803,
+      "learning_rate": 1.5533408211234002e-06,
+      "loss": 1.1053,
+      "num_input_tokens_seen": 722027036,
+      "step": 18380
+    },
+    {
+      "epoch": 0.8910747165423006,
+      "grad_norm": 0.37664535641670227,
+      "learning_rate": 1.5397554270040137e-06,
+      "loss": 1.1089,
+      "num_input_tokens_seen": 722405744,
+      "step": 18390
+    },
+    {
+      "epoch": 0.8915592596181801,
+      "grad_norm": 0.4099685847759247,
+      "learning_rate": 1.5262278146520426e-06,
+      "loss": 1.1115,
+      "num_input_tokens_seen": 722805360,
+      "step": 18400
+    },
+    {
+      "epoch": 0.8920438026940595,
+      "grad_norm": 0.3878704607486725,
+      "learning_rate": 1.5127580173855071e-06,
+      "loss": 1.1438,
+      "num_input_tokens_seen": 723211584,
+      "step": 18410
+    },
+    {
+      "epoch": 0.892528345769939,
+      "grad_norm": 0.438963919878006,
+      "learning_rate": 1.4993460683800698e-06,
+      "loss": 1.1211,
+      "num_input_tokens_seen": 723640776,
+      "step": 18420
+    },
+    {
+      "epoch": 0.8930128888458184,
+      "grad_norm": 0.3852384090423584,
+      "learning_rate": 1.4859920006688789e-06,
+      "loss": 1.1244,
+      "num_input_tokens_seen": 724035952,
+      "step": 18430
+    },
+    {
+      "epoch": 0.8934974319216978,
+      "grad_norm": 0.36654970049858093,
+      "learning_rate": 1.4726958471425495e-06,
+      "loss": 1.125,
+      "num_input_tokens_seen": 724439364,
+      "step": 18440
+    },
+    {
+      "epoch": 0.8939819749975773,
+      "grad_norm": 0.39647069573402405,
+      "learning_rate": 1.4594576405490417e-06,
+      "loss": 1.083,
+      "num_input_tokens_seen": 724842880,
+      "step": 18450
+    },
+    {
+      "epoch": 0.8944665180734568,
+      "grad_norm": 0.35190171003341675,
+      "learning_rate": 1.4462774134935963e-06,
+      "loss": 1.1442,
+      "num_input_tokens_seen": 725280940,
+      "step": 18460
+    },
+    {
+      "epoch": 0.8949510611493362,
+      "grad_norm": 0.40093109011650085,
+      "learning_rate": 1.43315519843866e-06,
+      "loss": 1.1372,
+      "num_input_tokens_seen": 725666144,
+      "step": 18470
+    },
+    {
+      "epoch": 0.8954356042252156,
+      "grad_norm": 0.4389098286628723,
+      "learning_rate": 1.4200910277037932e-06,
+      "loss": 1.1656,
+      "num_input_tokens_seen": 726077688,
+      "step": 18480
+    },
+    {
+      "epoch": 0.895920147301095,
+      "grad_norm": 0.401603102684021,
+      "learning_rate": 1.4070849334655883e-06,
+      "loss": 1.1504,
+      "num_input_tokens_seen": 726457876,
+      "step": 18490
+    },
+    {
+      "epoch": 0.8964046903769746,
+      "grad_norm": 0.4019956588745117,
+      "learning_rate": 1.3941369477576072e-06,
+      "loss": 1.1493,
+      "num_input_tokens_seen": 726837760,
+      "step": 18500
+    },
+    {
+      "epoch": 0.896889233452854,
+      "grad_norm": 0.3990192413330078,
+      "learning_rate": 1.3812471024702873e-06,
+      "loss": 1.0712,
+      "num_input_tokens_seen": 727203844,
+      "step": 18510
+    },
+    {
+      "epoch": 0.8973737765287334,
+      "grad_norm": 0.396411269903183,
+      "learning_rate": 1.3684154293508722e-06,
+      "loss": 1.0972,
+      "num_input_tokens_seen": 727595240,
+      "step": 18520
+    },
+    {
+      "epoch": 0.8978583196046128,
+      "grad_norm": 0.3998691737651825,
+      "learning_rate": 1.3556419600033288e-06,
+      "loss": 1.0947,
+      "num_input_tokens_seen": 727979392,
+      "step": 18530
+    },
+    {
+      "epoch": 0.8983428626804923,
+      "grad_norm": 0.4026288390159607,
+      "learning_rate": 1.3429267258882578e-06,
+      "loss": 1.1136,
+      "num_input_tokens_seen": 728342232,
+      "step": 18540
+    },
+    {
+      "epoch": 0.8988274057563718,
+      "grad_norm": 0.3861730396747589,
+      "learning_rate": 1.3302697583228523e-06,
+      "loss": 1.0925,
+      "num_input_tokens_seen": 728749172,
+      "step": 18550
+    },
+    {
+      "epoch": 0.8993119488322512,
+      "grad_norm": 0.38084274530410767,
+      "learning_rate": 1.3176710884807791e-06,
+      "loss": 1.115,
+      "num_input_tokens_seen": 729125920,
+      "step": 18560
+    },
+    {
+      "epoch": 0.8997964919081306,
+      "grad_norm": 0.43145671486854553,
+      "learning_rate": 1.3051307473921193e-06,
+      "loss": 1.1619,
+      "num_input_tokens_seen": 729532004,
+      "step": 18570
+    },
+    {
+      "epoch": 0.9002810349840101,
+      "grad_norm": 0.41663506627082825,
+      "learning_rate": 1.2926487659433024e-06,
+      "loss": 1.1116,
+      "num_input_tokens_seen": 729903788,
+      "step": 18580
+    },
+    {
+      "epoch": 0.9007655780598895,
+      "grad_norm": 0.4152432680130005,
+      "learning_rate": 1.2802251748770144e-06,
+      "loss": 1.1174,
+      "num_input_tokens_seen": 730309520,
+      "step": 18590
+    },
+    {
+      "epoch": 0.901250121135769,
+      "grad_norm": 0.3903380334377289,
+      "learning_rate": 1.267860004792129e-06,
+      "loss": 1.0801,
+      "num_input_tokens_seen": 730705760,
+      "step": 18600
+    },
+    {
+      "epoch": 0.9017346642116484,
+      "grad_norm": 0.4450056552886963,
+      "learning_rate": 1.255553286143632e-06,
+      "loss": 1.1158,
+      "num_input_tokens_seen": 731091508,
+      "step": 18610
+    },
+    {
+      "epoch": 0.9022192072875279,
+      "grad_norm": 0.39127764105796814,
+      "learning_rate": 1.2433050492425352e-06,
+      "loss": 1.1237,
+      "num_input_tokens_seen": 731486328,
+      "step": 18620
+    },
+    {
+      "epoch": 0.9027037503634073,
+      "grad_norm": 0.3897622227668762,
+      "learning_rate": 1.2311153242558354e-06,
+      "loss": 1.0822,
+      "num_input_tokens_seen": 731875972,
+      "step": 18630
+    },
+    {
+      "epoch": 0.9031882934392867,
+      "grad_norm": 0.4100290536880493,
+      "learning_rate": 1.218984141206389e-06,
+      "loss": 1.0725,
+      "num_input_tokens_seen": 732293268,
+      "step": 18640
+    },
+    {
+      "epoch": 0.9036728365151662,
+      "grad_norm": 0.3897705674171448,
+      "learning_rate": 1.2069115299728845e-06,
+      "loss": 1.1306,
+      "num_input_tokens_seen": 732690944,
+      "step": 18650
+    },
+    {
+      "epoch": 0.9041573795910457,
+      "grad_norm": 0.4191875755786896,
+      "learning_rate": 1.1948975202897423e-06,
+      "loss": 1.1358,
+      "num_input_tokens_seen": 733075324,
+      "step": 18660
+    },
+    {
+      "epoch": 0.9046419226669251,
+      "grad_norm": 0.3774699568748474,
+      "learning_rate": 1.1829421417470481e-06,
+      "loss": 1.1105,
+      "num_input_tokens_seen": 733466324,
+      "step": 18670
+    },
+    {
+      "epoch": 0.9051264657428045,
+      "grad_norm": 0.39872831106185913,
+      "learning_rate": 1.17104542379049e-06,
+      "loss": 1.1047,
+      "num_input_tokens_seen": 733853980,
+      "step": 18680
+    },
+    {
+      "epoch": 0.905611008818684,
+      "grad_norm": 0.38337400555610657,
+      "learning_rate": 1.159207395721268e-06,
+      "loss": 1.0915,
+      "num_input_tokens_seen": 734284336,
+      "step": 18690
+    },
+    {
+      "epoch": 0.9060955518945635,
+      "grad_norm": 0.4249227046966553,
+      "learning_rate": 1.1474280866960314e-06,
+      "loss": 1.1328,
+      "num_input_tokens_seen": 734691688,
+      "step": 18700
+    },
+    {
+      "epoch": 0.9065800949704429,
+      "grad_norm": 0.40089496970176697,
+      "learning_rate": 1.1357075257268147e-06,
+      "loss": 1.1384,
+      "num_input_tokens_seen": 735099992,
+      "step": 18710
+    },
+    {
+      "epoch": 0.9070646380463223,
+      "grad_norm": 0.42720267176628113,
+      "learning_rate": 1.1240457416809458e-06,
+      "loss": 1.1603,
+      "num_input_tokens_seen": 735517812,
+      "step": 18720
+    },
+    {
+      "epoch": 0.9075491811222017,
+      "grad_norm": 0.41853755712509155,
+      "learning_rate": 1.112442763280999e-06,
+      "loss": 1.0719,
+      "num_input_tokens_seen": 735875148,
+      "step": 18730
+    },
+    {
+      "epoch": 0.9080337241980813,
+      "grad_norm": 0.4119112491607666,
+      "learning_rate": 1.1008986191047095e-06,
+      "loss": 1.12,
+      "num_input_tokens_seen": 736279692,
+      "step": 18740
+    },
+    {
+      "epoch": 0.9085182672739607,
+      "grad_norm": 0.40979915857315063,
+      "learning_rate": 1.0894133375848907e-06,
+      "loss": 1.0894,
+      "num_input_tokens_seen": 736678652,
+      "step": 18750
+    },
+    {
+      "epoch": 0.9090028103498401,
+      "grad_norm": 0.3658255338668823,
+      "learning_rate": 1.0779869470094072e-06,
+      "loss": 1.1409,
+      "num_input_tokens_seen": 737076332,
+      "step": 18760
+    },
+    {
+      "epoch": 0.9094873534257195,
+      "grad_norm": 0.3819989264011383,
+      "learning_rate": 1.0666194755210524e-06,
+      "loss": 1.0877,
+      "num_input_tokens_seen": 737463700,
+      "step": 18770
+    },
+    {
+      "epoch": 0.909971896501599,
+      "grad_norm": 0.3775273859500885,
+      "learning_rate": 1.0553109511175192e-06,
+      "loss": 1.0883,
+      "num_input_tokens_seen": 737833988,
+      "step": 18780
+    },
+    {
+      "epoch": 0.9104564395774785,
+      "grad_norm": 0.4142516553401947,
+      "learning_rate": 1.0440614016513056e-06,
+      "loss": 1.0953,
+      "num_input_tokens_seen": 738251768,
+      "step": 18790
+    },
+    {
+      "epoch": 0.9109409826533579,
+      "grad_norm": 0.4076782166957855,
+      "learning_rate": 1.0328708548296622e-06,
+      "loss": 1.0818,
+      "num_input_tokens_seen": 738646584,
+      "step": 18800
+    },
+    {
+      "epoch": 0.9114255257292373,
+      "grad_norm": 0.3837725520133972,
+      "learning_rate": 1.0217393382145224e-06,
+      "loss": 1.1542,
+      "num_input_tokens_seen": 739027612,
+      "step": 18810
+    },
+    {
+      "epoch": 0.9119100688051168,
+      "grad_norm": 0.4405462145805359,
+      "learning_rate": 1.0106668792224134e-06,
+      "loss": 1.1116,
+      "num_input_tokens_seen": 739421692,
+      "step": 18820
+    },
+    {
+      "epoch": 0.9123946118809962,
+      "grad_norm": 0.4041633605957031,
+      "learning_rate": 9.996535051244316e-07,
+      "loss": 1.062,
+      "num_input_tokens_seen": 739822296,
+      "step": 18830
+    },
+    {
+      "epoch": 0.9128791549568757,
+      "grad_norm": 0.4356255829334259,
+      "learning_rate": 9.88699243046126e-07,
+      "loss": 1.0675,
+      "num_input_tokens_seen": 740216416,
+      "step": 18840
+    },
+    {
+      "epoch": 0.9133636980327551,
+      "grad_norm": 0.365655779838562,
+      "learning_rate": 9.778041199674626e-07,
+      "loss": 1.13,
+      "num_input_tokens_seen": 740584032,
+      "step": 18850
+    },
+    {
+      "epoch": 0.9138482411086346,
+      "grad_norm": 0.38094544410705566,
+      "learning_rate": 9.669681627227562e-07,
+      "loss": 1.1065,
+      "num_input_tokens_seen": 740973480,
+      "step": 18860
+    },
+    {
+      "epoch": 0.914332784184514,
+      "grad_norm": 0.41554176807403564,
+      "learning_rate": 9.561913980005916e-07,
+      "loss": 1.148,
+      "num_input_tokens_seen": 741350984,
+      "step": 18870
+    },
+    {
+      "epoch": 0.9148173272603934,
+      "grad_norm": 0.40062519907951355,
+      "learning_rate": 9.45473852343759e-07,
+      "loss": 1.1589,
+      "num_input_tokens_seen": 741724776,
+      "step": 18880
+    },
+    {
+      "epoch": 0.9153018703362729,
+      "grad_norm": 0.3932797610759735,
+      "learning_rate": 9.348155521492125e-07,
+      "loss": 1.1375,
+      "num_input_tokens_seen": 742129168,
+      "step": 18890
+    },
+    {
+      "epoch": 0.9157864134121524,
+      "grad_norm": 0.4039793610572815,
+      "learning_rate": 9.242165236679646e-07,
+      "loss": 1.1235,
+      "num_input_tokens_seen": 742522860,
+      "step": 18900
+    },
+    {
+      "epoch": 0.9162709564880318,
+      "grad_norm": 0.4156935513019562,
+      "learning_rate": 9.136767930050666e-07,
+      "loss": 1.1931,
+      "num_input_tokens_seen": 742904860,
+      "step": 18910
+    },
+    {
+      "epoch": 0.9167554995639112,
+      "grad_norm": 0.419506698846817,
+      "learning_rate": 9.031963861194953e-07,
+      "loss": 1.0917,
+      "num_input_tokens_seen": 743285040,
+      "step": 18920
+    },
+    {
+      "epoch": 0.9172400426397906,
+      "grad_norm": 0.417361855506897,
+      "learning_rate": 8.927753288241386e-07,
+      "loss": 1.1108,
+      "num_input_tokens_seen": 743675480,
+      "step": 18930
+    },
+    {
+      "epoch": 0.9177245857156702,
+      "grad_norm": 0.38126012682914734,
+      "learning_rate": 8.82413646785693e-07,
+      "loss": 1.1074,
+      "num_input_tokens_seen": 744082016,
+      "step": 18940
+    },
+    {
+      "epoch": 0.9182091287915496,
+      "grad_norm": 0.401082843542099,
+      "learning_rate": 8.721113655246222e-07,
+      "loss": 1.092,
+      "num_input_tokens_seen": 744477944,
+      "step": 18950
+    },
+    {
+      "epoch": 0.918693671867429,
+      "grad_norm": 0.41044408082962036,
+      "learning_rate": 8.618685104150925e-07,
+      "loss": 1.1382,
+      "num_input_tokens_seen": 744870004,
+      "step": 18960
+    },
+    {
+      "epoch": 0.9191782149433084,
+      "grad_norm": 0.40474820137023926,
+      "learning_rate": 8.516851066848936e-07,
+      "loss": 1.0912,
+      "num_input_tokens_seen": 745260580,
+      "step": 18970
+    },
+    {
+      "epoch": 0.919662758019188,
+      "grad_norm": 0.405483603477478,
+      "learning_rate": 8.415611794153982e-07,
+      "loss": 1.1106,
+      "num_input_tokens_seen": 745671092,
+      "step": 18980
+    },
+    {
+      "epoch": 0.9201473010950674,
+      "grad_norm": 0.43090713024139404,
+      "learning_rate": 8.314967535414858e-07,
+      "loss": 1.076,
+      "num_input_tokens_seen": 746080284,
+      "step": 18990
+    },
+    {
+      "epoch": 0.9206318441709468,
+      "grad_norm": 0.4164203107357025,
+      "learning_rate": 8.214918538514915e-07,
+      "loss": 1.1442,
+      "num_input_tokens_seen": 746446148,
+      "step": 19000
+    },
+    {
+      "epoch": 0.9211163872468262,
+      "grad_norm": 0.4038824737071991,
+      "learning_rate": 8.11546504987129e-07,
+      "loss": 1.1099,
+      "num_input_tokens_seen": 746836508,
+      "step": 19010
+    },
+    {
+      "epoch": 0.9216009303227057,
+      "grad_norm": 0.39919909834861755,
+      "learning_rate": 8.016607314434571e-07,
+      "loss": 1.0966,
+      "num_input_tokens_seen": 747221668,
+      "step": 19020
+    },
+    {
+      "epoch": 0.9220854733985852,
+      "grad_norm": 0.4156663417816162,
+      "learning_rate": 7.918345575687796e-07,
+      "loss": 1.1268,
+      "num_input_tokens_seen": 747595608,
+      "step": 19030
+    },
+    {
+      "epoch": 0.9225700164744646,
+      "grad_norm": 0.40613898634910583,
+      "learning_rate": 7.820680075646319e-07,
+      "loss": 1.101,
+      "num_input_tokens_seen": 747986884,
+      "step": 19040
+    },
+    {
+      "epoch": 0.923054559550344,
+      "grad_norm": 0.3905068039894104,
+      "learning_rate": 7.723611054856833e-07,
+      "loss": 1.1111,
+      "num_input_tokens_seen": 748374856,
+      "step": 19050
+    },
+    {
+      "epoch": 0.9235391026262235,
+      "grad_norm": 0.3992663621902466,
+      "learning_rate": 7.627138752396984e-07,
+      "loss": 1.0937,
+      "num_input_tokens_seen": 748752876,
+      "step": 19060
+    },
+    {
+      "epoch": 0.9240236457021029,
+      "grad_norm": 0.4008621871471405,
+      "learning_rate": 7.531263405874678e-07,
+      "loss": 1.0564,
+      "num_input_tokens_seen": 749125480,
+      "step": 19070
+    },
+    {
+      "epoch": 0.9245081887779824,
+      "grad_norm": 0.4372722804546356,
+      "learning_rate": 7.435985251427552e-07,
+      "loss": 1.1177,
+      "num_input_tokens_seen": 749506916,
+      "step": 19080
+    },
+    {
+      "epoch": 0.9249927318538618,
+      "grad_norm": 0.41504257917404175,
+      "learning_rate": 7.34130452372242e-07,
+      "loss": 1.0958,
+      "num_input_tokens_seen": 749895376,
+      "step": 19090
+    },
+    {
+      "epoch": 0.9254772749297413,
+      "grad_norm": 0.41004976630210876,
+      "learning_rate": 7.247221455954662e-07,
+      "loss": 1.109,
+      "num_input_tokens_seen": 750282460,
+      "step": 19100
+    },
+    {
+      "epoch": 0.9259618180056207,
+      "grad_norm": 0.40631359815597534,
+      "learning_rate": 7.1537362798475e-07,
+      "loss": 1.1214,
+      "num_input_tokens_seen": 750694156,
+      "step": 19110
+    },
+    {
+      "epoch": 0.9264463610815001,
+      "grad_norm": 0.3669877350330353,
+      "learning_rate": 7.060849225651756e-07,
+      "loss": 1.0916,
+      "num_input_tokens_seen": 751090800,
+      "step": 19120
+    },
+    {
+      "epoch": 0.9269309041573796,
+      "grad_norm": 0.39109060168266296,
+      "learning_rate": 6.968560522145007e-07,
+      "loss": 1.1302,
+      "num_input_tokens_seen": 751504408,
+      "step": 19130
+    },
+    {
+      "epoch": 0.9274154472332591,
+      "grad_norm": 0.4094466269016266,
+      "learning_rate": 6.876870396631097e-07,
+      "loss": 1.1174,
+      "num_input_tokens_seen": 751899136,
+      "step": 19140
+    },
+    {
+      "epoch": 0.9278999903091385,
+      "grad_norm": 0.39219051599502563,
+      "learning_rate": 6.785779074939657e-07,
+      "loss": 1.1205,
+      "num_input_tokens_seen": 752306356,
+      "step": 19150
+    },
+    {
+      "epoch": 0.9283845333850179,
+      "grad_norm": 0.39333903789520264,
+      "learning_rate": 6.695286781425392e-07,
+      "loss": 1.0771,
+      "num_input_tokens_seen": 752706400,
+      "step": 19160
+    },
+    {
+      "epoch": 0.9288690764608973,
+      "grad_norm": 0.4072161614894867,
+      "learning_rate": 6.605393738967763e-07,
+      "loss": 1.1592,
+      "num_input_tokens_seen": 753110792,
+      "step": 19170
+    },
+    {
+      "epoch": 0.9293536195367769,
+      "grad_norm": 0.39852002263069153,
+      "learning_rate": 6.516100168970113e-07,
+      "loss": 1.1284,
+      "num_input_tokens_seen": 753499472,
+      "step": 19180
+    },
+    {
+      "epoch": 0.9298381626126563,
+      "grad_norm": 0.4045473039150238,
+      "learning_rate": 6.427406291359489e-07,
+      "loss": 1.1153,
+      "num_input_tokens_seen": 753919864,
+      "step": 19190
+    },
+    {
+      "epoch": 0.9303227056885357,
+      "grad_norm": 0.37104934453964233,
+      "learning_rate": 6.339312324585761e-07,
+      "loss": 1.1329,
+      "num_input_tokens_seen": 754331304,
+      "step": 19200
+    },
+    {
+      "epoch": 0.9308072487644151,
+      "grad_norm": 0.40764960646629333,
+      "learning_rate": 6.251818485621341e-07,
+      "loss": 1.0779,
+      "num_input_tokens_seen": 754724460,
+      "step": 19210
+    },
+    {
+      "epoch": 0.9312917918402946,
+      "grad_norm": 0.3814258277416229,
+      "learning_rate": 6.164924989960519e-07,
+      "loss": 1.0927,
+      "num_input_tokens_seen": 755092192,
+      "step": 19220
+    },
+    {
+      "epoch": 0.9317763349161741,
+      "grad_norm": 0.4385196566581726,
+      "learning_rate": 6.078632051618988e-07,
+      "loss": 1.086,
+      "num_input_tokens_seen": 755461096,
+      "step": 19230
+    },
+    {
+      "epoch": 0.9322608779920535,
+      "grad_norm": 0.42675715684890747,
+      "learning_rate": 5.99293988313318e-07,
+      "loss": 1.1314,
+      "num_input_tokens_seen": 755853476,
+      "step": 19240
+    },
+    {
+      "epoch": 0.9327454210679329,
+      "grad_norm": 0.4081270396709442,
+      "learning_rate": 5.907848695559964e-07,
+      "loss": 1.0827,
+      "num_input_tokens_seen": 756266064,
+      "step": 19250
+    },
+    {
+      "epoch": 0.9332299641438124,
+      "grad_norm": 0.40428853034973145,
+      "learning_rate": 5.823358698476e-07,
+      "loss": 1.1371,
+      "num_input_tokens_seen": 756658988,
+      "step": 19260
+    },
+    {
+      "epoch": 0.9337145072196918,
+      "grad_norm": 0.3700270652770996,
+      "learning_rate": 5.73947009997719e-07,
+      "loss": 1.1292,
+      "num_input_tokens_seen": 757041432,
+      "step": 19270
+    },
+    {
+      "epoch": 0.9341990502955713,
+      "grad_norm": 0.4112299978733063,
+      "learning_rate": 5.656183106678287e-07,
+      "loss": 1.0989,
+      "num_input_tokens_seen": 757439484,
+      "step": 19280
+    },
+    {
+      "epoch": 0.9346835933714507,
+      "grad_norm": 0.3882913291454315,
+      "learning_rate": 5.573497923712173e-07,
+      "loss": 1.1112,
+      "num_input_tokens_seen": 757858200,
+      "step": 19290
+    },
+    {
+      "epoch": 0.9351681364473302,
+      "grad_norm": 0.3786594271659851,
+      "learning_rate": 5.491414754729667e-07,
+      "loss": 1.0912,
+      "num_input_tokens_seen": 758250852,
+      "step": 19300
+    },
+    {
+      "epoch": 0.9356526795232096,
+      "grad_norm": 0.4327850043773651,
+      "learning_rate": 5.409933801898692e-07,
+      "loss": 1.1164,
+      "num_input_tokens_seen": 758642612,
+      "step": 19310
+    },
+    {
+      "epoch": 0.936137222599089,
+      "grad_norm": 0.38797247409820557,
+      "learning_rate": 5.329055265904076e-07,
+      "loss": 1.0734,
+      "num_input_tokens_seen": 759069940,
+      "step": 19320
+    },
+    {
+      "epoch": 0.9366217656749685,
+      "grad_norm": 0.40884557366371155,
+      "learning_rate": 5.248779345946808e-07,
+      "loss": 1.0806,
+      "num_input_tokens_seen": 759446920,
+      "step": 19330
+    },
+    {
+      "epoch": 0.937106308750848,
+      "grad_norm": 0.4048992097377777,
+      "learning_rate": 5.169106239743648e-07,
+      "loss": 1.1333,
+      "num_input_tokens_seen": 759815092,
+      "step": 19340
+    },
+    {
+      "epoch": 0.9375908518267274,
+      "grad_norm": 0.39969974756240845,
+      "learning_rate": 5.090036143526767e-07,
+      "loss": 1.1397,
+      "num_input_tokens_seen": 760232476,
+      "step": 19350
+    },
+    {
+      "epoch": 0.9380753949026068,
+      "grad_norm": 0.41090163588523865,
+      "learning_rate": 5.011569252043019e-07,
+      "loss": 1.0991,
+      "num_input_tokens_seen": 760610356,
+      "step": 19360
+    },
+    {
+      "epoch": 0.9385599379784862,
+      "grad_norm": 0.42378801107406616,
+      "learning_rate": 4.933705758553619e-07,
+      "loss": 1.1243,
+      "num_input_tokens_seen": 761002188,
+      "step": 19370
+    },
+    {
+      "epoch": 0.9390444810543658,
+      "grad_norm": 0.40979644656181335,
+      "learning_rate": 4.856445854833719e-07,
+      "loss": 1.1623,
+      "num_input_tokens_seen": 761397232,
+      "step": 19380
+    },
+    {
+      "epoch": 0.9395290241302452,
+      "grad_norm": 0.3961712718009949,
+      "learning_rate": 4.779789731171713e-07,
+      "loss": 1.1282,
+      "num_input_tokens_seen": 761789268,
+      "step": 19390
+    },
+    {
+      "epoch": 0.9400135672061246,
+      "grad_norm": 0.3944045603275299,
+      "learning_rate": 4.7037375763689664e-07,
+      "loss": 1.1045,
+      "num_input_tokens_seen": 762178272,
+      "step": 19400
+    },
+    {
+      "epoch": 0.940498110282004,
+      "grad_norm": 0.40869519114494324,
+      "learning_rate": 4.628289577739309e-07,
+      "loss": 1.1048,
+      "num_input_tokens_seen": 762586012,
+      "step": 19410
+    },
+    {
+      "epoch": 0.9409826533578836,
+      "grad_norm": 0.46123749017715454,
+      "learning_rate": 4.55344592110854e-07,
+      "loss": 1.1706,
+      "num_input_tokens_seen": 763010200,
+      "step": 19420
+    },
+    {
+      "epoch": 0.941467196433763,
+      "grad_norm": 0.40070682764053345,
+      "learning_rate": 4.4792067908140387e-07,
+      "loss": 1.1299,
+      "num_input_tokens_seen": 763393780,
+      "step": 19430
+    },
+    {
+      "epoch": 0.9419517395096424,
+      "grad_norm": 0.381905734539032,
+      "learning_rate": 4.4055723697040976e-07,
+      "loss": 1.1155,
+      "num_input_tokens_seen": 763809116,
+      "step": 19440
+    },
+    {
+      "epoch": 0.9424362825855218,
+      "grad_norm": 0.41140425205230713,
+      "learning_rate": 4.3325428391378377e-07,
+      "loss": 1.1394,
+      "num_input_tokens_seen": 764205148,
+      "step": 19450
+    },
+    {
+      "epoch": 0.9429208256614013,
+      "grad_norm": 0.43021172285079956,
+      "learning_rate": 4.260118378984407e-07,
+      "loss": 1.1463,
+      "num_input_tokens_seen": 764595480,
+      "step": 19460
+    },
+    {
+      "epoch": 0.9434053687372808,
+      "grad_norm": 0.43823298811912537,
+      "learning_rate": 4.188299167622728e-07,
+      "loss": 1.1693,
+      "num_input_tokens_seen": 764964992,
+      "step": 19470
+    },
+    {
+      "epoch": 0.9438899118131602,
+      "grad_norm": 0.3995303213596344,
+      "learning_rate": 4.117085381941055e-07,
+      "loss": 1.1092,
+      "num_input_tokens_seen": 765363980,
+      "step": 19480
+    },
+    {
+      "epoch": 0.9443744548890396,
+      "grad_norm": 0.36656254529953003,
+      "learning_rate": 4.0464771973364456e-07,
+      "loss": 1.128,
+      "num_input_tokens_seen": 765753044,
+      "step": 19490
+    },
+    {
+      "epoch": 0.9448589979649191,
+      "grad_norm": 0.34931907057762146,
+      "learning_rate": 3.9764747877144015e-07,
+      "loss": 1.1266,
+      "num_input_tokens_seen": 766153976,
+      "step": 19500
+    },
+    {
+      "epoch": 0.9453435410407985,
+      "grad_norm": 0.38153091073036194,
+      "learning_rate": 3.90707832548845e-07,
+      "loss": 1.0923,
+      "num_input_tokens_seen": 766544700,
+      "step": 19510
+    },
+    {
+      "epoch": 0.945828084116678,
+      "grad_norm": 0.3994804620742798,
+      "learning_rate": 3.838287981579619e-07,
+      "loss": 1.1353,
+      "num_input_tokens_seen": 766951124,
+      "step": 19520
+    },
+    {
+      "epoch": 0.9463126271925574,
+      "grad_norm": 0.4302452504634857,
+      "learning_rate": 3.7701039254162405e-07,
+      "loss": 1.1205,
+      "num_input_tokens_seen": 767361300,
+      "step": 19530
+    },
+    {
+      "epoch": 0.9467971702684369,
+      "grad_norm": 0.4123125970363617,
+      "learning_rate": 3.702526324933148e-07,
+      "loss": 1.1401,
+      "num_input_tokens_seen": 767727640,
+      "step": 19540
+    },
+    {
+      "epoch": 0.9472817133443163,
+      "grad_norm": 0.3907496929168701,
+      "learning_rate": 3.635555346571701e-07,
+      "loss": 1.1068,
+      "num_input_tokens_seen": 768096588,
+      "step": 19550
+    },
+    {
+      "epoch": 0.9477662564201957,
+      "grad_norm": 0.47248342633247375,
+      "learning_rate": 3.569191155279067e-07,
+      "loss": 1.0979,
+      "num_input_tokens_seen": 768467260,
+      "step": 19560
+    },
+    {
+      "epoch": 0.9482507994960752,
+      "grad_norm": 0.38330796360969543,
+      "learning_rate": 3.503433914507942e-07,
+      "loss": 1.0569,
+      "num_input_tokens_seen": 768874516,
+      "step": 19570
+    },
+    {
+      "epoch": 0.9487353425719547,
+      "grad_norm": 0.40380969643592834,
+      "learning_rate": 3.438283786216134e-07,
+      "loss": 1.1573,
+      "num_input_tokens_seen": 769280172,
+      "step": 19580
+    },
+    {
+      "epoch": 0.9492198856478341,
+      "grad_norm": 0.37804293632507324,
+      "learning_rate": 3.373740930866176e-07,
+      "loss": 1.1426,
+      "num_input_tokens_seen": 769668856,
+      "step": 19590
+    },
+    {
+      "epoch": 0.9497044287237135,
+      "grad_norm": 0.39080822467803955,
+      "learning_rate": 3.309805507424796e-07,
+      "loss": 1.0912,
+      "num_input_tokens_seen": 770088040,
+      "step": 19600
+    },
+    {
+      "epoch": 0.9501889717995929,
+      "grad_norm": 0.4255511462688446,
+      "learning_rate": 3.2464776733628075e-07,
+      "loss": 1.1104,
+      "num_input_tokens_seen": 770463620,
+      "step": 19610
+    },
+    {
+      "epoch": 0.9506735148754725,
+      "grad_norm": 0.3901178240776062,
+      "learning_rate": 3.183757584654418e-07,
+      "loss": 1.1788,
+      "num_input_tokens_seen": 770856772,
+      "step": 19620
+    },
+    {
+      "epoch": 0.9511580579513519,
+      "grad_norm": 0.4149034321308136,
+      "learning_rate": 3.1216453957770565e-07,
+      "loss": 1.118,
+      "num_input_tokens_seen": 771274300,
+      "step": 19630
+    },
+    {
+      "epoch": 0.9516426010272313,
+      "grad_norm": 0.43007585406303406,
+      "learning_rate": 3.0601412597108527e-07,
+      "loss": 1.0905,
+      "num_input_tokens_seen": 771659776,
+      "step": 19640
+    },
+    {
+      "epoch": 0.9521271441031107,
+      "grad_norm": 0.41749218106269836,
+      "learning_rate": 2.9992453279383825e-07,
+      "loss": 1.1274,
+      "num_input_tokens_seen": 772049360,
+      "step": 19650
+    },
+    {
+      "epoch": 0.9526116871789903,
+      "grad_norm": 0.39733201265335083,
+      "learning_rate": 2.938957750444199e-07,
+      "loss": 1.0949,
+      "num_input_tokens_seen": 772424652,
+      "step": 19660
+    },
+    {
+      "epoch": 0.9530962302548697,
+      "grad_norm": 0.3866783380508423,
+      "learning_rate": 2.879278675714497e-07,
+      "loss": 1.1138,
+      "num_input_tokens_seen": 772839880,
+      "step": 19670
+    },
+    {
+      "epoch": 0.9535807733307491,
+      "grad_norm": 0.40627822279930115,
+      "learning_rate": 2.820208250736839e-07,
+      "loss": 1.1247,
+      "num_input_tokens_seen": 773238608,
+      "step": 19680
+    },
+    {
+      "epoch": 0.9540653164066285,
+      "grad_norm": 0.3711201548576355,
+      "learning_rate": 2.7617466209995115e-07,
+      "loss": 1.0661,
+      "num_input_tokens_seen": 773603544,
+      "step": 19690
+    },
+    {
+      "epoch": 0.954549859482508,
+      "grad_norm": 0.4101531505584717,
+      "learning_rate": 2.703893930491558e-07,
+      "loss": 1.1215,
+      "num_input_tokens_seen": 773975992,
+      "step": 19700
+    },
+    {
+      "epoch": 0.9550344025583875,
+      "grad_norm": 0.39880767464637756,
+      "learning_rate": 2.6466503217021654e-07,
+      "loss": 1.1503,
+      "num_input_tokens_seen": 774381780,
+      "step": 19710
+    },
+    {
+      "epoch": 0.9555189456342669,
+      "grad_norm": 0.3977833688259125,
+      "learning_rate": 2.5900159356202493e-07,
+      "loss": 1.1096,
+      "num_input_tokens_seen": 774779660,
+      "step": 19720
+    },
+    {
+      "epoch": 0.9560034887101463,
+      "grad_norm": 0.3905205726623535,
+      "learning_rate": 2.5339909117344515e-07,
+      "loss": 1.1316,
+      "num_input_tokens_seen": 775185260,
+      "step": 19730
+    },
+    {
+      "epoch": 0.9564880317860258,
+      "grad_norm": 0.3854650557041168,
+      "learning_rate": 2.4785753880323945e-07,
+      "loss": 1.0807,
+      "num_input_tokens_seen": 775588504,
+      "step": 19740
+    },
+    {
+      "epoch": 0.9569725748619052,
+      "grad_norm": 0.3917495608329773,
+      "learning_rate": 2.4237695010005945e-07,
+      "loss": 1.1197,
+      "num_input_tokens_seen": 775995628,
+      "step": 19750
+    },
+    {
+      "epoch": 0.9574571179377847,
+      "grad_norm": 0.401319295167923,
+      "learning_rate": 2.3695733856240466e-07,
+      "loss": 1.1239,
+      "num_input_tokens_seen": 776399704,
+      "step": 19760
+    },
+    {
+      "epoch": 0.9579416610136641,
+      "grad_norm": 0.3838230073451996,
+      "learning_rate": 2.3159871753859475e-07,
+      "loss": 1.1737,
+      "num_input_tokens_seen": 776806144,
+      "step": 19770
+    },
+    {
+      "epoch": 0.9584262040895436,
+      "grad_norm": 0.39468008279800415,
+      "learning_rate": 2.263011002267168e-07,
+      "loss": 1.1331,
+      "num_input_tokens_seen": 777150100,
+      "step": 19780
+    },
+    {
+      "epoch": 0.958910747165423,
+      "grad_norm": 0.390552282333374,
+      "learning_rate": 2.2106449967463084e-07,
+      "loss": 1.0929,
+      "num_input_tokens_seen": 777529280,
+      "step": 19790
+    },
+    {
+      "epoch": 0.9593952902413024,
+      "grad_norm": 0.38140344619750977,
+      "learning_rate": 2.158889287798921e-07,
+      "loss": 1.102,
+      "num_input_tokens_seen": 777934560,
+      "step": 19800
+    },
+    {
+      "epoch": 0.9598798333171819,
+      "grad_norm": 0.3968132734298706,
+      "learning_rate": 2.1077440028975936e-07,
+      "loss": 1.096,
+      "num_input_tokens_seen": 778325752,
+      "step": 19810
+    },
+    {
+      "epoch": 0.9603643763930614,
+      "grad_norm": 0.39363691210746765,
+      "learning_rate": 2.057209268011312e-07,
+      "loss": 1.1293,
+      "num_input_tokens_seen": 778719844,
+      "step": 19820
+    },
+    {
+      "epoch": 0.9608489194689408,
+      "grad_norm": 0.4126024544239044,
+      "learning_rate": 2.0072852076054305e-07,
+      "loss": 1.0956,
+      "num_input_tokens_seen": 779089772,
+      "step": 19830
+    },
+    {
+      "epoch": 0.9613334625448202,
+      "grad_norm": 0.4260122776031494,
+      "learning_rate": 1.957971944641146e-07,
+      "loss": 1.1148,
+      "num_input_tokens_seen": 779469980,
+      "step": 19840
+    },
+    {
+      "epoch": 0.9618180056206996,
+      "grad_norm": 0.43452244997024536,
+      "learning_rate": 1.9092696005753309e-07,
+      "loss": 1.1507,
+      "num_input_tokens_seen": 779871528,
+      "step": 19850
+    },
+    {
+      "epoch": 0.9623025486965792,
+      "grad_norm": 0.3910650312900543,
+      "learning_rate": 1.861178295360172e-07,
+      "loss": 1.1368,
+      "num_input_tokens_seen": 780275872,
+      "step": 19860
+    },
+    {
+      "epoch": 0.9627870917724586,
+      "grad_norm": 0.38760048151016235,
+      "learning_rate": 1.813698147442866e-07,
+      "loss": 1.1478,
+      "num_input_tokens_seen": 780644748,
+      "step": 19870
+    },
+    {
+      "epoch": 0.963271634848338,
+      "grad_norm": 0.4468711316585541,
+      "learning_rate": 1.7668292737653692e-07,
+      "loss": 1.0858,
+      "num_input_tokens_seen": 781036480,
+      "step": 19880
+    },
+    {
+      "epoch": 0.9637561779242174,
+      "grad_norm": 0.4267124533653259,
+      "learning_rate": 1.7205717897640638e-07,
+      "loss": 1.1375,
+      "num_input_tokens_seen": 781423288,
+      "step": 19890
+    },
+    {
+      "epoch": 0.964240721000097,
+      "grad_norm": 0.3964008092880249,
+      "learning_rate": 1.674925809369593e-07,
+      "loss": 1.1471,
+      "num_input_tokens_seen": 781830760,
+      "step": 19900
+    },
+    {
+      "epoch": 0.9647252640759764,
+      "grad_norm": 0.39716637134552,
+      "learning_rate": 1.6298914450063596e-07,
+      "loss": 1.1797,
+      "num_input_tokens_seen": 782212080,
+      "step": 19910
+    },
+    {
+      "epoch": 0.9652098071518558,
+      "grad_norm": 0.372944712638855,
+      "learning_rate": 1.5854688075924718e-07,
+      "loss": 1.1394,
+      "num_input_tokens_seen": 782640040,
+      "step": 19920
+    },
+    {
+      "epoch": 0.9656943502277352,
+      "grad_norm": 0.3637344539165497,
+      "learning_rate": 1.5416580065392984e-07,
+      "loss": 1.1311,
+      "num_input_tokens_seen": 783024516,
+      "step": 19930
+    },
+    {
+      "epoch": 0.9661788933036147,
+      "grad_norm": 0.42081308364868164,
+      "learning_rate": 1.4984591497513856e-07,
+      "loss": 1.1429,
+      "num_input_tokens_seen": 783413016,
+      "step": 19940
+    },
+    {
+      "epoch": 0.9666634363794941,
+      "grad_norm": 0.4066031873226166,
+      "learning_rate": 1.4558723436259857e-07,
+      "loss": 1.0993,
+      "num_input_tokens_seen": 783799984,
+      "step": 19950
+    },
+    {
+      "epoch": 0.9671479794553736,
+      "grad_norm": 0.39138638973236084,
+      "learning_rate": 1.413897693052918e-07,
+      "loss": 1.1108,
+      "num_input_tokens_seen": 784171868,
+      "step": 19960
+    },
+    {
+      "epoch": 0.967632522531253,
+      "grad_norm": 0.39998748898506165,
+      "learning_rate": 1.3725353014142627e-07,
+      "loss": 1.1042,
+      "num_input_tokens_seen": 784567228,
+      "step": 19970
+    },
+    {
+      "epoch": 0.9681170656071325,
+      "grad_norm": 0.42037269473075867,
+      "learning_rate": 1.3317852705842239e-07,
+      "loss": 1.1562,
+      "num_input_tokens_seen": 784983108,
+      "step": 19980
+    },
+    {
+      "epoch": 0.9686016086830119,
+      "grad_norm": 0.4196394979953766,
+      "learning_rate": 1.2916477009286553e-07,
+      "loss": 1.1435,
+      "num_input_tokens_seen": 785363296,
+      "step": 19990
+    },
+    {
+      "epoch": 0.9690861517588913,
+      "grad_norm": 0.405160129070282,
+      "learning_rate": 1.2521226913050077e-07,
+      "loss": 1.0713,
+      "num_input_tokens_seen": 785755388,
+      "step": 20000
+    },
+    {
+      "epoch": 0.9690861517588913,
+      "eval_loss": 1.1160393953323364,
+      "eval_runtime": 4.9905,
+      "eval_samples_per_second": 30.057,
+      "eval_steps_per_second": 3.807,
+      "num_input_tokens_seen": 785755388,
+      "step": 20000
+    },
+    {
+      "epoch": 0.9695706948347708,
+      "grad_norm": 0.38724902272224426,
+      "learning_rate": 1.2132103390620208e-07,
+      "loss": 1.1259,
+      "num_input_tokens_seen": 786168264,
+      "step": 20010
+    },
+    {
+      "epoch": 0.9700552379106503,
+      "grad_norm": 0.4315202236175537,
+      "learning_rate": 1.1749107400394477e-07,
+      "loss": 1.1549,
+      "num_input_tokens_seen": 786578560,
+      "step": 20020
+    },
+    {
+      "epoch": 0.9705397809865297,
+      "grad_norm": 0.4061446785926819,
+      "learning_rate": 1.1372239885678871e-07,
+      "loss": 1.1396,
+      "num_input_tokens_seen": 786961092,
+      "step": 20030
+    },
+    {
+      "epoch": 0.9710243240624091,
+      "grad_norm": 0.38450926542282104,
+      "learning_rate": 1.1001501774684785e-07,
+      "loss": 1.119,
+      "num_input_tokens_seen": 787377820,
+      "step": 20040
+    },
+    {
+      "epoch": 0.9715088671382885,
+      "grad_norm": 0.3797142803668976,
+      "learning_rate": 1.063689398052764e-07,
+      "loss": 1.1298,
+      "num_input_tokens_seen": 787789388,
+      "step": 20050
+    },
+    {
+      "epoch": 0.9719934102141681,
+      "grad_norm": 0.38401252031326294,
+      "learning_rate": 1.0278417401223539e-07,
+      "loss": 1.1608,
+      "num_input_tokens_seen": 788175444,
+      "step": 20060
+    },
+    {
+      "epoch": 0.9724779532900475,
+      "grad_norm": 0.39400792121887207,
+      "learning_rate": 9.92607291968789e-08,
+      "loss": 1.1329,
+      "num_input_tokens_seen": 788565228,
+      "step": 20070
+    },
+    {
+      "epoch": 0.9729624963659269,
+      "grad_norm": 0.36722561717033386,
+      "learning_rate": 9.579861403732627e-08,
+      "loss": 1.1117,
+      "num_input_tokens_seen": 788948208,
+      "step": 20080
+    },
+    {
+      "epoch": 0.9734470394418063,
+      "grad_norm": 0.39072954654693604,
+      "learning_rate": 9.239783706065375e-08,
+      "loss": 1.1896,
+      "num_input_tokens_seen": 789332972,
+      "step": 20090
+    },
+    {
+      "epoch": 0.9739315825176859,
+      "grad_norm": 0.4184485077857971,
+      "learning_rate": 8.905840664284736e-08,
+      "loss": 1.114,
+      "num_input_tokens_seen": 789721976,
+      "step": 20100
+    },
+    {
+      "epoch": 0.9744161255935653,
+      "grad_norm": 0.3963470160961151,
+      "learning_rate": 8.578033100881677e-08,
+      "loss": 1.1298,
+      "num_input_tokens_seen": 790100792,
+      "step": 20110
+    },
+    {
+      "epoch": 0.9749006686694447,
+      "grad_norm": 0.39767026901245117,
+      "learning_rate": 8.256361823234527e-08,
+      "loss": 1.1244,
+      "num_input_tokens_seen": 790492440,
+      "step": 20120
+    },
+    {
+      "epoch": 0.9753852117453241,
+      "grad_norm": 0.4353410005569458,
+      "learning_rate": 7.940827623608427e-08,
+      "loss": 1.1209,
+      "num_input_tokens_seen": 790900392,
+      "step": 20130
+    },
+    {
+      "epoch": 0.9758697548212036,
+      "grad_norm": 0.4203384518623352,
+      "learning_rate": 7.631431279153111e-08,
+      "loss": 1.1472,
+      "num_input_tokens_seen": 791285628,
+      "step": 20140
+    },
+    {
+      "epoch": 0.9763542978970831,
+      "grad_norm": 0.3865092098712921,
+      "learning_rate": 7.328173551901241e-08,
+      "loss": 1.1548,
+      "num_input_tokens_seen": 791701900,
+      "step": 20150
+    },
+    {
+      "epoch": 0.9768388409729625,
+      "grad_norm": 0.4074331521987915,
+      "learning_rate": 7.031055188765622e-08,
+      "loss": 1.0969,
+      "num_input_tokens_seen": 792103252,
+      "step": 20160
+    },
+    {
+      "epoch": 0.9773233840488419,
+      "grad_norm": 0.4258996546268463,
+      "learning_rate": 6.74007692153894e-08,
+      "loss": 1.1043,
+      "num_input_tokens_seen": 792516300,
+      "step": 20170
+    },
+    {
+      "epoch": 0.9778079271247214,
+      "grad_norm": 0.42485523223876953,
+      "learning_rate": 6.455239466890418e-08,
+      "loss": 1.1026,
+      "num_input_tokens_seen": 792886424,
+      "step": 20180
+    },
+    {
+      "epoch": 0.9782924702006008,
+      "grad_norm": 0.4095394015312195,
+      "learning_rate": 6.176543526364709e-08,
+      "loss": 1.1432,
+      "num_input_tokens_seen": 793294236,
+      "step": 20190
+    },
+    {
+      "epoch": 0.9787770132764803,
+      "grad_norm": 0.38347864151000977,
+      "learning_rate": 5.903989786380515e-08,
+      "loss": 1.126,
+      "num_input_tokens_seen": 793704988,
+      "step": 20200
+    },
+    {
+      "epoch": 0.9792615563523597,
+      "grad_norm": 0.39956969022750854,
+      "learning_rate": 5.637578918227526e-08,
+      "loss": 1.1295,
+      "num_input_tokens_seen": 794085224,
+      "step": 20210
+    },
+    {
+      "epoch": 0.9797460994282392,
+      "grad_norm": 0.4015878736972809,
+      "learning_rate": 5.377311578067257e-08,
+      "loss": 1.1144,
+      "num_input_tokens_seen": 794453220,
+      "step": 20220
+    },
+    {
+      "epoch": 0.9802306425041186,
+      "grad_norm": 0.4096335172653198,
+      "learning_rate": 5.1231884069288805e-08,
+      "loss": 1.0987,
+      "num_input_tokens_seen": 794843816,
+      "step": 20230
+    },
+    {
+      "epoch": 0.980715185579998,
+      "grad_norm": 0.3710688650608063,
+      "learning_rate": 4.875210030708677e-08,
+      "loss": 1.1324,
+      "num_input_tokens_seen": 795210116,
+      "step": 20240
+    },
+    {
+      "epoch": 0.9811997286558775,
+      "grad_norm": 0.39916378259658813,
+      "learning_rate": 4.6333770601689195e-08,
+      "loss": 1.1467,
+      "num_input_tokens_seen": 795613888,
+      "step": 20250
+    },
+    {
+      "epoch": 0.981684271731757,
+      "grad_norm": 0.4291175603866577,
+      "learning_rate": 4.397690090935935e-08,
+      "loss": 1.1484,
+      "num_input_tokens_seen": 796019220,
+      "step": 20260
+    },
+    {
+      "epoch": 0.9821688148076364,
+      "grad_norm": 0.37853673100471497,
+      "learning_rate": 4.1681497034984364e-08,
+      "loss": 1.0957,
+      "num_input_tokens_seen": 796422340,
+      "step": 20270
+    },
+    {
+      "epoch": 0.9826533578835158,
+      "grad_norm": 0.3903977572917938,
+      "learning_rate": 3.9447564632066894e-08,
+      "loss": 1.1658,
+      "num_input_tokens_seen": 796813048,
+      "step": 20280
+    },
+    {
+      "epoch": 0.9831379009593952,
+      "grad_norm": 0.39838188886642456,
+      "learning_rate": 3.7275109202700165e-08,
+      "loss": 1.0973,
+      "num_input_tokens_seen": 797185304,
+      "step": 20290
+    },
+    {
+      "epoch": 0.9836224440352748,
+      "grad_norm": 0.3897262513637543,
+      "learning_rate": 3.5164136097567965e-08,
+      "loss": 1.1272,
+      "num_input_tokens_seen": 797583336,
+      "step": 20300
+    },
+    {
+      "epoch": 0.9841069871111542,
+      "grad_norm": 0.400979220867157,
+      "learning_rate": 3.311465051592522e-08,
+      "loss": 1.144,
+      "num_input_tokens_seen": 797981308,
+      "step": 20310
+    },
+    {
+      "epoch": 0.9845915301870336,
+      "grad_norm": 0.4422793686389923,
+      "learning_rate": 3.112665750558408e-08,
+      "loss": 1.07,
+      "num_input_tokens_seen": 798370388,
+      "step": 20320
+    },
+    {
+      "epoch": 0.985076073262913,
+      "grad_norm": 0.37973761558532715,
+      "learning_rate": 2.9200161962900098e-08,
+      "loss": 1.0582,
+      "num_input_tokens_seen": 798765140,
+      "step": 20330
+    },
+    {
+      "epoch": 0.9855606163387926,
+      "grad_norm": 0.42133259773254395,
+      "learning_rate": 2.7335168632763863e-08,
+      "loss": 1.1353,
+      "num_input_tokens_seen": 799159564,
+      "step": 20340
+    },
+    {
+      "epoch": 0.986045159414672,
+      "grad_norm": 0.5169322490692139,
+      "learning_rate": 2.5531682108595468e-08,
+      "loss": 1.1157,
+      "num_input_tokens_seen": 799545036,
+      "step": 20350
+    },
+    {
+      "epoch": 0.9865297024905514,
+      "grad_norm": 0.4151138365268707,
+      "learning_rate": 2.3789706832311186e-08,
+      "loss": 1.0991,
+      "num_input_tokens_seen": 799911908,
+      "step": 20360
+    },
+    {
+      "epoch": 0.9870142455664308,
+      "grad_norm": 0.3700861930847168,
+      "learning_rate": 2.2109247094342922e-08,
+      "loss": 1.0973,
+      "num_input_tokens_seen": 800292720,
+      "step": 20370
+    },
+    {
+      "epoch": 0.9874987886423103,
+      "grad_norm": 0.3943372070789337,
+      "learning_rate": 2.049030703359933e-08,
+      "loss": 1.0944,
+      "num_input_tokens_seen": 800674764,
+      "step": 20380
+    },
+    {
+      "epoch": 0.9879833317181898,
+      "grad_norm": 0.40062424540519714,
+      "learning_rate": 1.893289063747694e-08,
+      "loss": 1.1316,
+      "num_input_tokens_seen": 801063864,
+      "step": 20390
+    },
+    {
+      "epoch": 0.9884678747940692,
+      "grad_norm": 0.40976980328559875,
+      "learning_rate": 1.7437001741835157e-08,
+      "loss": 1.1555,
+      "num_input_tokens_seen": 801463696,
+      "step": 20400
+    },
+    {
+      "epoch": 0.9889524178699486,
+      "grad_norm": 0.36616411805152893,
+      "learning_rate": 1.6002644031001823e-08,
+      "loss": 1.1345,
+      "num_input_tokens_seen": 801848428,
+      "step": 20410
+    },
+    {
+      "epoch": 0.9894369609458281,
+      "grad_norm": 0.42311692237854004,
+      "learning_rate": 1.4629821037742686e-08,
+      "loss": 1.1274,
+      "num_input_tokens_seen": 802239180,
+      "step": 20420
+    },
+    {
+      "epoch": 0.9899215040217075,
+      "grad_norm": 0.39517006278038025,
+      "learning_rate": 1.331853614327805e-08,
+      "loss": 1.1543,
+      "num_input_tokens_seen": 802651824,
+      "step": 20430
+    },
+    {
+      "epoch": 0.990406047097587,
+      "grad_norm": 0.4204373359680176,
+      "learning_rate": 1.2068792577255017e-08,
+      "loss": 1.1141,
+      "num_input_tokens_seen": 803056060,
+      "step": 20440
+    },
+    {
+      "epoch": 0.9908905901734664,
+      "grad_norm": 0.42741021513938904,
+      "learning_rate": 1.0880593417753049e-08,
+      "loss": 1.107,
+      "num_input_tokens_seen": 803449480,
+      "step": 20450
+    },
+    {
+      "epoch": 0.9913751332493459,
+      "grad_norm": 0.3925841450691223,
+      "learning_rate": 9.753941591258974e-09,
+      "loss": 1.0623,
+      "num_input_tokens_seen": 803845032,
+      "step": 20460
+    },
+    {
+      "epoch": 0.9918596763252253,
+      "grad_norm": 0.38680553436279297,
+      "learning_rate": 8.6888398726781e-09,
+      "loss": 1.0997,
+      "num_input_tokens_seen": 804253704,
+      "step": 20470
+    },
+    {
+      "epoch": 0.9923442194011047,
+      "grad_norm": 0.3853330612182617,
+      "learning_rate": 7.685290885323104e-09,
+      "loss": 1.1338,
+      "num_input_tokens_seen": 804638340,
+      "step": 20480
+    },
+    {
+      "epoch": 0.9928287624769842,
+      "grad_norm": 0.378024160861969,
+      "learning_rate": 6.743297100897383e-09,
+      "loss": 1.1218,
+      "num_input_tokens_seen": 805045548,
+      "step": 20490
+    },
+    {
+      "epoch": 0.9933133055528637,
+      "grad_norm": 0.4238669276237488,
+      "learning_rate": 5.862860839500606e-09,
+      "loss": 1.1136,
+      "num_input_tokens_seen": 805444148,
+      "step": 20500
+    },
+    {
+      "epoch": 0.9937978486287431,
+      "grad_norm": 0.40654098987579346,
+      "learning_rate": 5.0439842696148324e-09,
+      "loss": 1.1313,
+      "num_input_tokens_seen": 805854612,
+      "step": 20510
+    },
+    {
+      "epoch": 0.9942823917046225,
+      "grad_norm": 0.38661834597587585,
+      "learning_rate": 4.286669408104516e-09,
+      "loss": 1.0887,
+      "num_input_tokens_seen": 806230656,
+      "step": 20520
+    },
+    {
+      "epoch": 0.9947669347805019,
+      "grad_norm": 0.42557960748672485,
+      "learning_rate": 3.5909181202137267e-09,
+      "loss": 1.1222,
+      "num_input_tokens_seen": 806626356,
+      "step": 20530
+    },
+    {
+      "epoch": 0.9952514778563815,
+      "grad_norm": 0.40263572335243225,
+      "learning_rate": 2.9567321195467236e-09,
+      "loss": 1.0986,
+      "num_input_tokens_seen": 807005260,
+      "step": 20540
+    },
+    {
+      "epoch": 0.9957360209322609,
+      "grad_norm": 0.39904552698135376,
+      "learning_rate": 2.384112968087382e-09,
+      "loss": 1.0754,
+      "num_input_tokens_seen": 807374212,
+      "step": 20550
+    },
+    {
+      "epoch": 0.9962205640081403,
+      "grad_norm": 0.41701799631118774,
+      "learning_rate": 1.8730620761742147e-09,
+      "loss": 1.1052,
+      "num_input_tokens_seen": 807812204,
+      "step": 20560
+    },
+    {
+      "epoch": 0.9967051070840197,
+      "grad_norm": 0.3973786532878876,
+      "learning_rate": 1.4235807025114733e-09,
+      "loss": 1.1094,
+      "num_input_tokens_seen": 808181824,
+      "step": 20570
+    },
+    {
+      "epoch": 0.9971896501598992,
+      "grad_norm": 0.3856273889541626,
+      "learning_rate": 1.0356699541497205e-09,
+      "loss": 1.1272,
+      "num_input_tokens_seen": 808554948,
+      "step": 20580
+    },
+    {
+      "epoch": 0.9976741932357787,
+      "grad_norm": 0.4207695424556732,
+      "learning_rate": 7.093307865052578e-10,
+      "loss": 1.1468,
+      "num_input_tokens_seen": 808980516,
+      "step": 20590
+    },
+    {
+      "epoch": 0.9981587363116581,
+      "grad_norm": 0.4193539619445801,
+      "learning_rate": 4.4456400333514616e-10,
+      "loss": 1.1278,
+      "num_input_tokens_seen": 809358268,
+      "step": 20600
+    },
+    {
+      "epoch": 0.9986432793875375,
+      "grad_norm": 0.4102073609828949,
+      "learning_rate": 2.4137025675663447e-10,
+      "loss": 1.0943,
+      "num_input_tokens_seen": 809733876,
+      "step": 20610
+    },
+    {
+      "epoch": 0.999127822463417,
+      "grad_norm": 0.41482511162757874,
+      "learning_rate": 9.975004722495573e-11,
+      "loss": 1.1314,
+      "num_input_tokens_seen": 810140020,
+      "step": 20620
+    },
+    {
+      "epoch": 0.9996123655392964,
+      "grad_norm": 0.3799668252468109,
+      "learning_rate": 1.9703723547204178e-11,
+      "loss": 1.1301,
+      "num_input_tokens_seen": 810535920,
+      "step": 20630
+    },
+    {
+      "epoch": 1.0,
+      "num_input_tokens_seen": 810839096,
+      "step": 20638,
+      "total_flos": 4.0772707408255386e+18,
+      "train_loss": 1.171214375919653,
+      "train_runtime": 248208.7734,
+      "train_samples_per_second": 10.643,
+      "train_steps_per_second": 0.083
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 20638,
+  "num_input_tokens_seen": 810839096,
+  "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": 4.0772707408255386e+18,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}