diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -1,9 +1,9 @@
 {
-  "best_metric": 2.046032667160034,
-  "best_model_checkpoint": "/home/sunggeunan/data/ICL/outputs/lora/SKIML-ICL_mrqa_nq_v3/Meta-Llama-3-8B-Instruct-unanswerable-5Q-0U-0C-qa_first/checkpoint-512",
-  "epoch": 0.9982939312698026,
+  "best_metric": 2.0117971897125244,
+  "best_model_checkpoint": "/home/sunggeunan/data/ICL/outputs/lora/SKIML-ICL_mrqa_nq_v3/Meta-Llama-3-8B-Instruct-unanswerable-5Q-0U-0C-qa_first/checkpoint-1536",
+  "epoch": 2.994881793809408,
   "eval_steps": 500,
-  "global_step": 512,
+  "global_step": 1536,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -3599,6 +3599,7190 @@
       "eval_samples_per_second": 1.293,
       "eval_steps_per_second": 0.324,
       "step": 512
+    },
+    {
+      "epoch": 1.0002437241043138,
+      "grad_norm": 0.41433945298194885,
+      "learning_rate": 3.505825908156271e-07,
+      "loss": 2.0366,
+      "step": 513
+    },
+    {
+      "epoch": 1.0021935169388252,
+      "grad_norm": 0.42399510741233826,
+      "learning_rate": 3.5023989033584646e-07,
+      "loss": 1.991,
+      "step": 514
+    },
+    {
+      "epoch": 1.0041433097733365,
+      "grad_norm": 0.45652541518211365,
+      "learning_rate": 3.4989718985606576e-07,
+      "loss": 2.0066,
+      "step": 515
+    },
+    {
+      "epoch": 1.006093102607848,
+      "grad_norm": 0.43585795164108276,
+      "learning_rate": 3.495544893762851e-07,
+      "loss": 2.0025,
+      "step": 516
+    },
+    {
+      "epoch": 1.0080428954423593,
+      "grad_norm": 0.43803489208221436,
+      "learning_rate": 3.4921178889650445e-07,
+      "loss": 2.0654,
+      "step": 517
+    },
+    {
+      "epoch": 1.0099926882768706,
+      "grad_norm": 0.43803176283836365,
+      "learning_rate": 3.4886908841672375e-07,
+      "loss": 2.0896,
+      "step": 518
+    },
+    {
+      "epoch": 1.011942481111382,
+      "grad_norm": 0.41983944177627563,
+      "learning_rate": 3.485263879369431e-07,
+      "loss": 2.0335,
+      "step": 519
+    },
+    {
+      "epoch": 1.0138922739458933,
+      "grad_norm": 0.4354363977909088,
+      "learning_rate": 3.4818368745716244e-07,
+      "loss": 2.0699,
+      "step": 520
+    },
+    {
+      "epoch": 1.0158420667804047,
+      "grad_norm": 0.42140671610832214,
+      "learning_rate": 3.4784098697738174e-07,
+      "loss": 1.9646,
+      "step": 521
+    },
+    {
+      "epoch": 1.0177918596149158,
+      "grad_norm": 0.4265493154525757,
+      "learning_rate": 3.474982864976011e-07,
+      "loss": 2.0735,
+      "step": 522
+    },
+    {
+      "epoch": 1.0197416524494272,
+      "grad_norm": 0.43847259879112244,
+      "learning_rate": 3.4715558601782043e-07,
+      "loss": 2.0986,
+      "step": 523
+    },
+    {
+      "epoch": 1.0216914452839385,
+      "grad_norm": 0.4600801467895508,
+      "learning_rate": 3.468128855380397e-07,
+      "loss": 2.0643,
+      "step": 524
+    },
+    {
+      "epoch": 1.0236412381184499,
+      "grad_norm": 0.42904648184776306,
+      "learning_rate": 3.4647018505825907e-07,
+      "loss": 2.0056,
+      "step": 525
+    },
+    {
+      "epoch": 1.0255910309529612,
+      "grad_norm": 0.46431151032447815,
+      "learning_rate": 3.461274845784784e-07,
+      "loss": 2.1056,
+      "step": 526
+    },
+    {
+      "epoch": 1.0275408237874726,
+      "grad_norm": 0.455836683511734,
+      "learning_rate": 3.457847840986977e-07,
+      "loss": 2.0187,
+      "step": 527
+    },
+    {
+      "epoch": 1.029490616621984,
+      "grad_norm": 0.4192461669445038,
+      "learning_rate": 3.45442083618917e-07,
+      "loss": 2.0832,
+      "step": 528
+    },
+    {
+      "epoch": 1.0314404094564953,
+      "grad_norm": 0.4513595402240753,
+      "learning_rate": 3.450993831391364e-07,
+      "loss": 2.058,
+      "step": 529
+    },
+    {
+      "epoch": 1.0333902022910066,
+      "grad_norm": 0.4370152950286865,
+      "learning_rate": 3.447566826593557e-07,
+      "loss": 2.0537,
+      "step": 530
+    },
+    {
+      "epoch": 1.035339995125518,
+      "grad_norm": 0.4199161231517792,
+      "learning_rate": 3.44413982179575e-07,
+      "loss": 1.9518,
+      "step": 531
+    },
+    {
+      "epoch": 1.0372897879600294,
+      "grad_norm": 0.43688762187957764,
+      "learning_rate": 3.440712816997944e-07,
+      "loss": 2.0444,
+      "step": 532
+    },
+    {
+      "epoch": 1.0392395807945405,
+      "grad_norm": 0.49809253215789795,
+      "learning_rate": 3.437285812200137e-07,
+      "loss": 2.0401,
+      "step": 533
+    },
+    {
+      "epoch": 1.0411893736290518,
+      "grad_norm": 0.4518781900405884,
+      "learning_rate": 3.43385880740233e-07,
+      "loss": 2.0605,
+      "step": 534
+    },
+    {
+      "epoch": 1.0431391664635632,
+      "grad_norm": 0.45353132486343384,
+      "learning_rate": 3.430431802604524e-07,
+      "loss": 2.0402,
+      "step": 535
+    },
+    {
+      "epoch": 1.0450889592980745,
+      "grad_norm": 0.4396359622478485,
+      "learning_rate": 3.427004797806717e-07,
+      "loss": 2.0643,
+      "step": 536
+    },
+    {
+      "epoch": 1.047038752132586,
+      "grad_norm": 0.4434252083301544,
+      "learning_rate": 3.42357779300891e-07,
+      "loss": 2.0188,
+      "step": 537
+    },
+    {
+      "epoch": 1.0489885449670973,
+      "grad_norm": 0.4241044819355011,
+      "learning_rate": 3.420150788211104e-07,
+      "loss": 1.9556,
+      "step": 538
+    },
+    {
+      "epoch": 1.0509383378016086,
+      "grad_norm": 0.4382232129573822,
+      "learning_rate": 3.4167237834132967e-07,
+      "loss": 1.9855,
+      "step": 539
+    },
+    {
+      "epoch": 1.05288813063612,
+      "grad_norm": 0.4357564151287079,
+      "learning_rate": 3.4132967786154897e-07,
+      "loss": 2.0524,
+      "step": 540
+    },
+    {
+      "epoch": 1.0548379234706313,
+      "grad_norm": 0.46050140261650085,
+      "learning_rate": 3.409869773817683e-07,
+      "loss": 2.0461,
+      "step": 541
+    },
+    {
+      "epoch": 1.0567877163051427,
+      "grad_norm": 0.44581982493400574,
+      "learning_rate": 3.4064427690198766e-07,
+      "loss": 1.9955,
+      "step": 542
+    },
+    {
+      "epoch": 1.0587375091396538,
+      "grad_norm": 0.4502599835395813,
+      "learning_rate": 3.4030157642220696e-07,
+      "loss": 2.0864,
+      "step": 543
+    },
+    {
+      "epoch": 1.0606873019741652,
+      "grad_norm": 0.44767019152641296,
+      "learning_rate": 3.399588759424263e-07,
+      "loss": 2.0447,
+      "step": 544
+    },
+    {
+      "epoch": 1.0626370948086765,
+      "grad_norm": 0.44603490829467773,
+      "learning_rate": 3.3961617546264565e-07,
+      "loss": 2.0709,
+      "step": 545
+    },
+    {
+      "epoch": 1.0645868876431879,
+      "grad_norm": 0.4321264922618866,
+      "learning_rate": 3.3927347498286495e-07,
+      "loss": 2.0157,
+      "step": 546
+    },
+    {
+      "epoch": 1.0665366804776992,
+      "grad_norm": 0.4479556083679199,
+      "learning_rate": 3.389307745030843e-07,
+      "loss": 2.1088,
+      "step": 547
+    },
+    {
+      "epoch": 1.0684864733122106,
+      "grad_norm": 0.4292636513710022,
+      "learning_rate": 3.3858807402330364e-07,
+      "loss": 2.0847,
+      "step": 548
+    },
+    {
+      "epoch": 1.070436266146722,
+      "grad_norm": 0.43631821870803833,
+      "learning_rate": 3.3824537354352294e-07,
+      "loss": 2.034,
+      "step": 549
+    },
+    {
+      "epoch": 1.0723860589812333,
+      "grad_norm": 0.43201327323913574,
+      "learning_rate": 3.379026730637423e-07,
+      "loss": 1.9633,
+      "step": 550
+    },
+    {
+      "epoch": 1.0743358518157446,
+      "grad_norm": 0.4389747679233551,
+      "learning_rate": 3.3755997258396163e-07,
+      "loss": 2.0331,
+      "step": 551
+    },
+    {
+      "epoch": 1.076285644650256,
+      "grad_norm": 0.46588924527168274,
+      "learning_rate": 3.372172721041809e-07,
+      "loss": 2.0748,
+      "step": 552
+    },
+    {
+      "epoch": 1.0782354374847674,
+      "grad_norm": 0.45190852880477905,
+      "learning_rate": 3.3687457162440027e-07,
+      "loss": 1.9639,
+      "step": 553
+    },
+    {
+      "epoch": 1.0801852303192785,
+      "grad_norm": 0.4458979070186615,
+      "learning_rate": 3.3653187114461957e-07,
+      "loss": 2.1124,
+      "step": 554
+    },
+    {
+      "epoch": 1.0821350231537898,
+      "grad_norm": 0.40400832891464233,
+      "learning_rate": 3.361891706648389e-07,
+      "loss": 1.9776,
+      "step": 555
+    },
+    {
+      "epoch": 1.0840848159883012,
+      "grad_norm": 0.4538462460041046,
+      "learning_rate": 3.3584647018505826e-07,
+      "loss": 1.9962,
+      "step": 556
+    },
+    {
+      "epoch": 1.0860346088228126,
+      "grad_norm": 0.44181132316589355,
+      "learning_rate": 3.3550376970527756e-07,
+      "loss": 2.0973,
+      "step": 557
+    },
+    {
+      "epoch": 1.087984401657324,
+      "grad_norm": 0.43516308069229126,
+      "learning_rate": 3.351610692254969e-07,
+      "loss": 1.9923,
+      "step": 558
+    },
+    {
+      "epoch": 1.0899341944918353,
+      "grad_norm": 0.4485546052455902,
+      "learning_rate": 3.3481836874571625e-07,
+      "loss": 2.0242,
+      "step": 559
+    },
+    {
+      "epoch": 1.0918839873263466,
+      "grad_norm": 0.45358070731163025,
+      "learning_rate": 3.3447566826593555e-07,
+      "loss": 2.0603,
+      "step": 560
+    },
+    {
+      "epoch": 1.093833780160858,
+      "grad_norm": 0.43879690766334534,
+      "learning_rate": 3.341329677861549e-07,
+      "loss": 1.9869,
+      "step": 561
+    },
+    {
+      "epoch": 1.0957835729953693,
+      "grad_norm": 0.4376320242881775,
+      "learning_rate": 3.3379026730637424e-07,
+      "loss": 2.0447,
+      "step": 562
+    },
+    {
+      "epoch": 1.0977333658298805,
+      "grad_norm": 0.4591986835002899,
+      "learning_rate": 3.3344756682659354e-07,
+      "loss": 2.0188,
+      "step": 563
+    },
+    {
+      "epoch": 1.0996831586643918,
+      "grad_norm": 0.4306589961051941,
+      "learning_rate": 3.331048663468129e-07,
+      "loss": 2.0223,
+      "step": 564
+    },
+    {
+      "epoch": 1.1016329514989032,
+      "grad_norm": 0.43692710995674133,
+      "learning_rate": 3.3276216586703223e-07,
+      "loss": 2.0507,
+      "step": 565
+    },
+    {
+      "epoch": 1.1035827443334145,
+      "grad_norm": 0.4663935601711273,
+      "learning_rate": 3.324194653872515e-07,
+      "loss": 2.0444,
+      "step": 566
+    },
+    {
+      "epoch": 1.1055325371679259,
+      "grad_norm": 0.45090562105178833,
+      "learning_rate": 3.320767649074708e-07,
+      "loss": 1.9944,
+      "step": 567
+    },
+    {
+      "epoch": 1.1074823300024372,
+      "grad_norm": 0.4450632631778717,
+      "learning_rate": 3.317340644276902e-07,
+      "loss": 2.0264,
+      "step": 568
+    },
+    {
+      "epoch": 1.1094321228369486,
+      "grad_norm": 0.45126745104789734,
+      "learning_rate": 3.313913639479095e-07,
+      "loss": 2.081,
+      "step": 569
+    },
+    {
+      "epoch": 1.11138191567146,
+      "grad_norm": 0.44254472851753235,
+      "learning_rate": 3.310486634681288e-07,
+      "loss": 2.0223,
+      "step": 570
+    },
+    {
+      "epoch": 1.1133317085059713,
+      "grad_norm": 0.43211621046066284,
+      "learning_rate": 3.307059629883482e-07,
+      "loss": 2.0363,
+      "step": 571
+    },
+    {
+      "epoch": 1.1152815013404827,
+      "grad_norm": 0.4256265163421631,
+      "learning_rate": 3.303632625085675e-07,
+      "loss": 2.0363,
+      "step": 572
+    },
+    {
+      "epoch": 1.117231294174994,
+      "grad_norm": 0.4462417960166931,
+      "learning_rate": 3.300205620287868e-07,
+      "loss": 2.0394,
+      "step": 573
+    },
+    {
+      "epoch": 1.1191810870095051,
+      "grad_norm": 0.4583437442779541,
+      "learning_rate": 3.296778615490062e-07,
+      "loss": 2.0878,
+      "step": 574
+    },
+    {
+      "epoch": 1.1211308798440165,
+      "grad_norm": 0.4595088064670563,
+      "learning_rate": 3.293351610692255e-07,
+      "loss": 2.111,
+      "step": 575
+    },
+    {
+      "epoch": 1.1230806726785278,
+      "grad_norm": 0.4117080569267273,
+      "learning_rate": 3.289924605894448e-07,
+      "loss": 1.999,
+      "step": 576
+    },
+    {
+      "epoch": 1.1250304655130392,
+      "grad_norm": 0.4381641149520874,
+      "learning_rate": 3.286497601096642e-07,
+      "loss": 2.044,
+      "step": 577
+    },
+    {
+      "epoch": 1.1269802583475506,
+      "grad_norm": 0.43854039907455444,
+      "learning_rate": 3.283070596298835e-07,
+      "loss": 2.0272,
+      "step": 578
+    },
+    {
+      "epoch": 1.128930051182062,
+      "grad_norm": 0.4721965789794922,
+      "learning_rate": 3.279643591501028e-07,
+      "loss": 2.0697,
+      "step": 579
+    },
+    {
+      "epoch": 1.1308798440165733,
+      "grad_norm": 0.4373783767223358,
+      "learning_rate": 3.2762165867032207e-07,
+      "loss": 2.0102,
+      "step": 580
+    },
+    {
+      "epoch": 1.1328296368510846,
+      "grad_norm": 0.4286502003669739,
+      "learning_rate": 3.2727895819054147e-07,
+      "loss": 1.9695,
+      "step": 581
+    },
+    {
+      "epoch": 1.134779429685596,
+      "grad_norm": 0.4373305141925812,
+      "learning_rate": 3.2693625771076077e-07,
+      "loss": 1.9823,
+      "step": 582
+    },
+    {
+      "epoch": 1.1367292225201073,
+      "grad_norm": 0.4659106433391571,
+      "learning_rate": 3.2659355723098006e-07,
+      "loss": 2.081,
+      "step": 583
+    },
+    {
+      "epoch": 1.1386790153546187,
+      "grad_norm": 0.4315546154975891,
+      "learning_rate": 3.2625085675119946e-07,
+      "loss": 2.0336,
+      "step": 584
+    },
+    {
+      "epoch": 1.1406288081891298,
+      "grad_norm": 0.4512901306152344,
+      "learning_rate": 3.2590815627141876e-07,
+      "loss": 2.0642,
+      "step": 585
+    },
+    {
+      "epoch": 1.1425786010236412,
+      "grad_norm": 0.4398232400417328,
+      "learning_rate": 3.2556545579163805e-07,
+      "loss": 2.0401,
+      "step": 586
+    },
+    {
+      "epoch": 1.1445283938581525,
+      "grad_norm": 0.45262405276298523,
+      "learning_rate": 3.2522275531185745e-07,
+      "loss": 2.0999,
+      "step": 587
+    },
+    {
+      "epoch": 1.1464781866926639,
+      "grad_norm": 0.4210640490055084,
+      "learning_rate": 3.2488005483207675e-07,
+      "loss": 1.992,
+      "step": 588
+    },
+    {
+      "epoch": 1.1484279795271752,
+      "grad_norm": 0.4530121386051178,
+      "learning_rate": 3.2453735435229604e-07,
+      "loss": 2.0119,
+      "step": 589
+    },
+    {
+      "epoch": 1.1503777723616866,
+      "grad_norm": 0.43637722730636597,
+      "learning_rate": 3.2419465387251544e-07,
+      "loss": 2.0022,
+      "step": 590
+    },
+    {
+      "epoch": 1.152327565196198,
+      "grad_norm": 0.46872228384017944,
+      "learning_rate": 3.2385195339273474e-07,
+      "loss": 2.0545,
+      "step": 591
+    },
+    {
+      "epoch": 1.1542773580307093,
+      "grad_norm": 0.45964333415031433,
+      "learning_rate": 3.2350925291295403e-07,
+      "loss": 2.0313,
+      "step": 592
+    },
+    {
+      "epoch": 1.1562271508652207,
+      "grad_norm": 0.4444529414176941,
+      "learning_rate": 3.231665524331734e-07,
+      "loss": 2.0463,
+      "step": 593
+    },
+    {
+      "epoch": 1.1581769436997318,
+      "grad_norm": 0.4702310264110565,
+      "learning_rate": 3.228238519533927e-07,
+      "loss": 2.0055,
+      "step": 594
+    },
+    {
+      "epoch": 1.1601267365342431,
+      "grad_norm": 0.4435891807079315,
+      "learning_rate": 3.22481151473612e-07,
+      "loss": 2.1027,
+      "step": 595
+    },
+    {
+      "epoch": 1.1620765293687545,
+      "grad_norm": 0.4557732343673706,
+      "learning_rate": 3.2213845099383137e-07,
+      "loss": 2.0307,
+      "step": 596
+    },
+    {
+      "epoch": 1.1640263222032659,
+      "grad_norm": 0.4286348819732666,
+      "learning_rate": 3.217957505140507e-07,
+      "loss": 2.0196,
+      "step": 597
+    },
+    {
+      "epoch": 1.1659761150377772,
+      "grad_norm": 0.4475346803665161,
+      "learning_rate": 3.2145305003427e-07,
+      "loss": 2.1014,
+      "step": 598
+    },
+    {
+      "epoch": 1.1679259078722886,
+      "grad_norm": 0.418293297290802,
+      "learning_rate": 3.2111034955448936e-07,
+      "loss": 2.078,
+      "step": 599
+    },
+    {
+      "epoch": 1.1698757007068,
+      "grad_norm": 0.42740973830223083,
+      "learning_rate": 3.207676490747087e-07,
+      "loss": 1.9695,
+      "step": 600
+    },
+    {
+      "epoch": 1.1718254935413113,
+      "grad_norm": 0.41325512528419495,
+      "learning_rate": 3.20424948594928e-07,
+      "loss": 2.0297,
+      "step": 601
+    },
+    {
+      "epoch": 1.1737752863758226,
+      "grad_norm": 0.4326270818710327,
+      "learning_rate": 3.2008224811514735e-07,
+      "loss": 2.0059,
+      "step": 602
+    },
+    {
+      "epoch": 1.175725079210334,
+      "grad_norm": 0.44774889945983887,
+      "learning_rate": 3.197395476353667e-07,
+      "loss": 2.0427,
+      "step": 603
+    },
+    {
+      "epoch": 1.1776748720448453,
+      "grad_norm": 0.446158766746521,
+      "learning_rate": 3.19396847155586e-07,
+      "loss": 2.0748,
+      "step": 604
+    },
+    {
+      "epoch": 1.1796246648793565,
+      "grad_norm": 0.4654727280139923,
+      "learning_rate": 3.1905414667580534e-07,
+      "loss": 1.9297,
+      "step": 605
+    },
+    {
+      "epoch": 1.1815744577138678,
+      "grad_norm": 0.45213672518730164,
+      "learning_rate": 3.1871144619602463e-07,
+      "loss": 2.1087,
+      "step": 606
+    },
+    {
+      "epoch": 1.1835242505483792,
+      "grad_norm": 0.45228397846221924,
+      "learning_rate": 3.18368745716244e-07,
+      "loss": 2.0961,
+      "step": 607
+    },
+    {
+      "epoch": 1.1854740433828905,
+      "grad_norm": 0.4470541477203369,
+      "learning_rate": 3.180260452364633e-07,
+      "loss": 2.0073,
+      "step": 608
+    },
+    {
+      "epoch": 1.1874238362174019,
+      "grad_norm": 0.4324132204055786,
+      "learning_rate": 3.176833447566826e-07,
+      "loss": 2.0334,
+      "step": 609
+    },
+    {
+      "epoch": 1.1893736290519132,
+      "grad_norm": 0.47044241428375244,
+      "learning_rate": 3.1734064427690197e-07,
+      "loss": 2.1086,
+      "step": 610
+    },
+    {
+      "epoch": 1.1913234218864246,
+      "grad_norm": 0.43018707633018494,
+      "learning_rate": 3.169979437971213e-07,
+      "loss": 2.0289,
+      "step": 611
+    },
+    {
+      "epoch": 1.193273214720936,
+      "grad_norm": 0.44133853912353516,
+      "learning_rate": 3.166552433173406e-07,
+      "loss": 2.0333,
+      "step": 612
+    },
+    {
+      "epoch": 1.1952230075554473,
+      "grad_norm": 0.45557719469070435,
+      "learning_rate": 3.1631254283755996e-07,
+      "loss": 2.0399,
+      "step": 613
+    },
+    {
+      "epoch": 1.1971728003899587,
+      "grad_norm": 0.4350452125072479,
+      "learning_rate": 3.159698423577793e-07,
+      "loss": 2.0224,
+      "step": 614
+    },
+    {
+      "epoch": 1.19912259322447,
+      "grad_norm": 0.4687999188899994,
+      "learning_rate": 3.156271418779986e-07,
+      "loss": 2.0228,
+      "step": 615
+    },
+    {
+      "epoch": 1.2010723860589811,
+      "grad_norm": 0.43684178590774536,
+      "learning_rate": 3.1528444139821795e-07,
+      "loss": 2.0776,
+      "step": 616
+    },
+    {
+      "epoch": 1.2030221788934925,
+      "grad_norm": 0.45561161637306213,
+      "learning_rate": 3.149417409184373e-07,
+      "loss": 2.0022,
+      "step": 617
+    },
+    {
+      "epoch": 1.2049719717280039,
+      "grad_norm": 0.4689810276031494,
+      "learning_rate": 3.145990404386566e-07,
+      "loss": 2.0173,
+      "step": 618
+    },
+    {
+      "epoch": 1.2069217645625152,
+      "grad_norm": 0.4293496310710907,
+      "learning_rate": 3.142563399588759e-07,
+      "loss": 1.9824,
+      "step": 619
+    },
+    {
+      "epoch": 1.2088715573970266,
+      "grad_norm": 0.4662802219390869,
+      "learning_rate": 3.139136394790953e-07,
+      "loss": 2.0784,
+      "step": 620
+    },
+    {
+      "epoch": 1.210821350231538,
+      "grad_norm": 0.45310187339782715,
+      "learning_rate": 3.135709389993146e-07,
+      "loss": 1.9844,
+      "step": 621
+    },
+    {
+      "epoch": 1.2127711430660493,
+      "grad_norm": 0.4419795870780945,
+      "learning_rate": 3.1322823851953387e-07,
+      "loss": 2.0515,
+      "step": 622
+    },
+    {
+      "epoch": 1.2147209359005606,
+      "grad_norm": 0.4516865611076355,
+      "learning_rate": 3.1288553803975327e-07,
+      "loss": 2.0879,
+      "step": 623
+    },
+    {
+      "epoch": 1.216670728735072,
+      "grad_norm": 0.46178489923477173,
+      "learning_rate": 3.1254283755997257e-07,
+      "loss": 2.0498,
+      "step": 624
+    },
+    {
+      "epoch": 1.2186205215695831,
+      "grad_norm": 0.4678952097892761,
+      "learning_rate": 3.1220013708019186e-07,
+      "loss": 2.0408,
+      "step": 625
+    },
+    {
+      "epoch": 1.2205703144040945,
+      "grad_norm": 0.4456236660480499,
+      "learning_rate": 3.1185743660041126e-07,
+      "loss": 1.9694,
+      "step": 626
+    },
+    {
+      "epoch": 1.2225201072386058,
+      "grad_norm": 0.4397581219673157,
+      "learning_rate": 3.1151473612063056e-07,
+      "loss": 2.0048,
+      "step": 627
+    },
+    {
+      "epoch": 1.2244699000731172,
+      "grad_norm": 0.4338027238845825,
+      "learning_rate": 3.1117203564084985e-07,
+      "loss": 2.0194,
+      "step": 628
+    },
+    {
+      "epoch": 1.2264196929076285,
+      "grad_norm": 0.4413823187351227,
+      "learning_rate": 3.108293351610692e-07,
+      "loss": 2.025,
+      "step": 629
+    },
+    {
+      "epoch": 1.2283694857421399,
+      "grad_norm": 0.43685299158096313,
+      "learning_rate": 3.1048663468128855e-07,
+      "loss": 2.0051,
+      "step": 630
+    },
+    {
+      "epoch": 1.2303192785766512,
+      "grad_norm": 0.4644426107406616,
+      "learning_rate": 3.1014393420150784e-07,
+      "loss": 2.0313,
+      "step": 631
+    },
+    {
+      "epoch": 1.2322690714111626,
+      "grad_norm": 0.4478755593299866,
+      "learning_rate": 3.098012337217272e-07,
+      "loss": 1.9669,
+      "step": 632
+    },
+    {
+      "epoch": 1.234218864245674,
+      "grad_norm": 0.43452218174934387,
+      "learning_rate": 3.0945853324194654e-07,
+      "loss": 1.9927,
+      "step": 633
+    },
+    {
+      "epoch": 1.2361686570801853,
+      "grad_norm": 0.4408141076564789,
+      "learning_rate": 3.0911583276216583e-07,
+      "loss": 2.136,
+      "step": 634
+    },
+    {
+      "epoch": 1.2381184499146967,
+      "grad_norm": 0.42754924297332764,
+      "learning_rate": 3.087731322823852e-07,
+      "loss": 2.0247,
+      "step": 635
+    },
+    {
+      "epoch": 1.2400682427492078,
+      "grad_norm": 0.4387798607349396,
+      "learning_rate": 3.084304318026045e-07,
+      "loss": 1.9643,
+      "step": 636
+    },
+    {
+      "epoch": 1.2420180355837191,
+      "grad_norm": 0.46978920698165894,
+      "learning_rate": 3.080877313228238e-07,
+      "loss": 2.0776,
+      "step": 637
+    },
+    {
+      "epoch": 1.2439678284182305,
+      "grad_norm": 0.41821563243865967,
+      "learning_rate": 3.0774503084304317e-07,
+      "loss": 2.0355,
+      "step": 638
+    },
+    {
+      "epoch": 1.2459176212527419,
+      "grad_norm": 0.4664837419986725,
+      "learning_rate": 3.074023303632625e-07,
+      "loss": 2.0328,
+      "step": 639
+    },
+    {
+      "epoch": 1.2478674140872532,
+      "grad_norm": 0.4467378258705139,
+      "learning_rate": 3.070596298834818e-07,
+      "loss": 2.0058,
+      "step": 640
+    },
+    {
+      "epoch": 1.2498172069217646,
+      "grad_norm": 0.442058265209198,
+      "learning_rate": 3.0671692940370116e-07,
+      "loss": 2.0565,
+      "step": 641
+    },
+    {
+      "epoch": 1.251766999756276,
+      "grad_norm": 0.4655166268348694,
+      "learning_rate": 3.0637422892392045e-07,
+      "loss": 2.0628,
+      "step": 642
+    },
+    {
+      "epoch": 1.2537167925907873,
+      "grad_norm": 0.4388466477394104,
+      "learning_rate": 3.060315284441398e-07,
+      "loss": 2.0716,
+      "step": 643
+    },
+    {
+      "epoch": 1.2556665854252986,
+      "grad_norm": 0.48705416917800903,
+      "learning_rate": 3.0568882796435915e-07,
+      "loss": 1.9872,
+      "step": 644
+    },
+    {
+      "epoch": 1.2576163782598098,
+      "grad_norm": 0.4618842899799347,
+      "learning_rate": 3.0534612748457844e-07,
+      "loss": 2.0306,
+      "step": 645
+    },
+    {
+      "epoch": 1.2595661710943213,
+      "grad_norm": 0.46533843874931335,
+      "learning_rate": 3.050034270047978e-07,
+      "loss": 2.0827,
+      "step": 646
+    },
+    {
+      "epoch": 1.2615159639288325,
+      "grad_norm": 0.4898700714111328,
+      "learning_rate": 3.0466072652501714e-07,
+      "loss": 1.9585,
+      "step": 647
+    },
+    {
+      "epoch": 1.2634657567633438,
+      "grad_norm": 0.4561532735824585,
+      "learning_rate": 3.0431802604523643e-07,
+      "loss": 2.0689,
+      "step": 648
+    },
+    {
+      "epoch": 1.2654155495978552,
+      "grad_norm": 0.4628736078739166,
+      "learning_rate": 3.039753255654558e-07,
+      "loss": 2.0307,
+      "step": 649
+    },
+    {
+      "epoch": 1.2673653424323665,
+      "grad_norm": 0.4475798010826111,
+      "learning_rate": 3.036326250856751e-07,
+      "loss": 2.0372,
+      "step": 650
+    },
+    {
+      "epoch": 1.269315135266878,
+      "grad_norm": 0.44448035955429077,
+      "learning_rate": 3.032899246058944e-07,
+      "loss": 2.0334,
+      "step": 651
+    },
+    {
+      "epoch": 1.2712649281013892,
+      "grad_norm": 0.4554859697818756,
+      "learning_rate": 3.0294722412611377e-07,
+      "loss": 2.0487,
+      "step": 652
+    },
+    {
+      "epoch": 1.2732147209359006,
+      "grad_norm": 0.44150403141975403,
+      "learning_rate": 3.026045236463331e-07,
+      "loss": 2.085,
+      "step": 653
+    },
+    {
+      "epoch": 1.275164513770412,
+      "grad_norm": 0.4476960301399231,
+      "learning_rate": 3.022618231665524e-07,
+      "loss": 1.9762,
+      "step": 654
+    },
+    {
+      "epoch": 1.2771143066049233,
+      "grad_norm": 0.4773290753364563,
+      "learning_rate": 3.019191226867717e-07,
+      "loss": 2.0565,
+      "step": 655
+    },
+    {
+      "epoch": 1.2790640994394344,
+      "grad_norm": 0.43788987398147583,
+      "learning_rate": 3.015764222069911e-07,
+      "loss": 2.0629,
+      "step": 656
+    },
+    {
+      "epoch": 1.281013892273946,
+      "grad_norm": 0.4314157962799072,
+      "learning_rate": 3.012337217272104e-07,
+      "loss": 2.0554,
+      "step": 657
+    },
+    {
+      "epoch": 1.2829636851084572,
+      "grad_norm": 0.45381680130958557,
+      "learning_rate": 3.008910212474297e-07,
+      "loss": 2.0514,
+      "step": 658
+    },
+    {
+      "epoch": 1.2849134779429685,
+      "grad_norm": 0.47213441133499146,
+      "learning_rate": 3.005483207676491e-07,
+      "loss": 2.0267,
+      "step": 659
+    },
+    {
+      "epoch": 1.2868632707774799,
+      "grad_norm": 0.4460486173629761,
+      "learning_rate": 3.002056202878684e-07,
+      "loss": 2.0717,
+      "step": 660
+    },
+    {
+      "epoch": 1.2888130636119912,
+      "grad_norm": 0.452747642993927,
+      "learning_rate": 2.998629198080877e-07,
+      "loss": 2.0634,
+      "step": 661
+    },
+    {
+      "epoch": 1.2907628564465026,
+      "grad_norm": 0.4495120942592621,
+      "learning_rate": 2.995202193283071e-07,
+      "loss": 2.042,
+      "step": 662
+    },
+    {
+      "epoch": 1.292712649281014,
+      "grad_norm": 0.433224081993103,
+      "learning_rate": 2.991775188485264e-07,
+      "loss": 2.0565,
+      "step": 663
+    },
+    {
+      "epoch": 1.2946624421155253,
+      "grad_norm": 0.4596520960330963,
+      "learning_rate": 2.9883481836874567e-07,
+      "loss": 2.0272,
+      "step": 664
+    },
+    {
+      "epoch": 1.2966122349500366,
+      "grad_norm": 0.433887243270874,
+      "learning_rate": 2.9849211788896507e-07,
+      "loss": 1.965,
+      "step": 665
+    },
+    {
+      "epoch": 1.298562027784548,
+      "grad_norm": 0.44755810499191284,
+      "learning_rate": 2.9814941740918437e-07,
+      "loss": 1.9915,
+      "step": 666
+    },
+    {
+      "epoch": 1.3005118206190591,
+      "grad_norm": 0.48203861713409424,
+      "learning_rate": 2.9780671692940366e-07,
+      "loss": 2.0296,
+      "step": 667
+    },
+    {
+      "epoch": 1.3024616134535705,
+      "grad_norm": 0.4314959943294525,
+      "learning_rate": 2.97464016449623e-07,
+      "loss": 2.0282,
+      "step": 668
+    },
+    {
+      "epoch": 1.3044114062880818,
+      "grad_norm": 0.4476211369037628,
+      "learning_rate": 2.9712131596984236e-07,
+      "loss": 2.0348,
+      "step": 669
+    },
+    {
+      "epoch": 1.3063611991225932,
+      "grad_norm": 0.45356854796409607,
+      "learning_rate": 2.9677861549006165e-07,
+      "loss": 2.0369,
+      "step": 670
+    },
+    {
+      "epoch": 1.3083109919571045,
+      "grad_norm": 0.4637032747268677,
+      "learning_rate": 2.96435915010281e-07,
+      "loss": 2.1002,
+      "step": 671
+    },
+    {
+      "epoch": 1.310260784791616,
+      "grad_norm": 0.4258365333080292,
+      "learning_rate": 2.9609321453050035e-07,
+      "loss": 2.0184,
+      "step": 672
+    },
+    {
+      "epoch": 1.3122105776261273,
+      "grad_norm": 0.4571716785430908,
+      "learning_rate": 2.9575051405071964e-07,
+      "loss": 2.0711,
+      "step": 673
+    },
+    {
+      "epoch": 1.3141603704606386,
+      "grad_norm": 0.4479144215583801,
+      "learning_rate": 2.95407813570939e-07,
+      "loss": 2.1037,
+      "step": 674
+    },
+    {
+      "epoch": 1.31611016329515,
+      "grad_norm": 0.463773638010025,
+      "learning_rate": 2.9506511309115834e-07,
+      "loss": 2.087,
+      "step": 675
+    },
+    {
+      "epoch": 1.318059956129661,
+      "grad_norm": 0.4595959782600403,
+      "learning_rate": 2.9472241261137763e-07,
+      "loss": 2.0246,
+      "step": 676
+    },
+    {
+      "epoch": 1.3200097489641727,
+      "grad_norm": 0.41977226734161377,
+      "learning_rate": 2.94379712131597e-07,
+      "loss": 2.0132,
+      "step": 677
+    },
+    {
+      "epoch": 1.3219595417986838,
+      "grad_norm": 0.4429217576980591,
+      "learning_rate": 2.940370116518163e-07,
+      "loss": 2.0414,
+      "step": 678
+    },
+    {
+      "epoch": 1.3239093346331952,
+      "grad_norm": 0.46036285161972046,
+      "learning_rate": 2.936943111720356e-07,
+      "loss": 2.0474,
+      "step": 679
+    },
+    {
+      "epoch": 1.3258591274677065,
+      "grad_norm": 0.4518478512763977,
+      "learning_rate": 2.9335161069225497e-07,
+      "loss": 1.991,
+      "step": 680
+    },
+    {
+      "epoch": 1.3278089203022179,
+      "grad_norm": 0.4507528841495514,
+      "learning_rate": 2.9300891021247426e-07,
+      "loss": 2.0038,
+      "step": 681
+    },
+    {
+      "epoch": 1.3297587131367292,
+      "grad_norm": 0.45446595549583435,
+      "learning_rate": 2.926662097326936e-07,
+      "loss": 1.9257,
+      "step": 682
+    },
+    {
+      "epoch": 1.3317085059712406,
+      "grad_norm": 0.45073091983795166,
+      "learning_rate": 2.9232350925291296e-07,
+      "loss": 2.0667,
+      "step": 683
+    },
+    {
+      "epoch": 1.333658298805752,
+      "grad_norm": 0.43848779797554016,
+      "learning_rate": 2.9198080877313225e-07,
+      "loss": 2.0127,
+      "step": 684
+    },
+    {
+      "epoch": 1.3356080916402633,
+      "grad_norm": 0.44587504863739014,
+      "learning_rate": 2.916381082933516e-07,
+      "loss": 2.0694,
+      "step": 685
+    },
+    {
+      "epoch": 1.3375578844747746,
+      "grad_norm": 0.46157652139663696,
+      "learning_rate": 2.9129540781357095e-07,
+      "loss": 2.112,
+      "step": 686
+    },
+    {
+      "epoch": 1.3395076773092858,
+      "grad_norm": 0.461897075176239,
+      "learning_rate": 2.9095270733379024e-07,
+      "loss": 2.0431,
+      "step": 687
+    },
+    {
+      "epoch": 1.3414574701437973,
+      "grad_norm": 0.42506590485572815,
+      "learning_rate": 2.906100068540096e-07,
+      "loss": 2.0612,
+      "step": 688
+    },
+    {
+      "epoch": 1.3434072629783085,
+      "grad_norm": 0.43368127942085266,
+      "learning_rate": 2.9026730637422894e-07,
+      "loss": 2.0253,
+      "step": 689
+    },
+    {
+      "epoch": 1.3453570558128198,
+      "grad_norm": 0.4484082758426666,
+      "learning_rate": 2.8992460589444823e-07,
+      "loss": 1.9962,
+      "step": 690
+    },
+    {
+      "epoch": 1.3473068486473312,
+      "grad_norm": 0.44570791721343994,
+      "learning_rate": 2.895819054146676e-07,
+      "loss": 2.018,
+      "step": 691
+    },
+    {
+      "epoch": 1.3492566414818425,
+      "grad_norm": 0.4472144842147827,
+      "learning_rate": 2.892392049348869e-07,
+      "loss": 2.0254,
+      "step": 692
+    },
+    {
+      "epoch": 1.351206434316354,
+      "grad_norm": 0.4680030047893524,
+      "learning_rate": 2.888965044551062e-07,
+      "loss": 2.1265,
+      "step": 693
+    },
+    {
+      "epoch": 1.3531562271508653,
+      "grad_norm": 0.44323253631591797,
+      "learning_rate": 2.885538039753255e-07,
+      "loss": 2.0222,
+      "step": 694
+    },
+    {
+      "epoch": 1.3551060199853766,
+      "grad_norm": 0.4732964038848877,
+      "learning_rate": 2.882111034955449e-07,
+      "loss": 2.0219,
+      "step": 695
+    },
+    {
+      "epoch": 1.357055812819888,
+      "grad_norm": 0.4392209053039551,
+      "learning_rate": 2.878684030157642e-07,
+      "loss": 1.9841,
+      "step": 696
+    },
+    {
+      "epoch": 1.3590056056543993,
+      "grad_norm": 0.46177539229393005,
+      "learning_rate": 2.875257025359835e-07,
+      "loss": 2.0461,
+      "step": 697
+    },
+    {
+      "epoch": 1.3609553984889105,
+      "grad_norm": 0.4625999927520752,
+      "learning_rate": 2.871830020562029e-07,
+      "loss": 2.0137,
+      "step": 698
+    },
+    {
+      "epoch": 1.3629051913234218,
+      "grad_norm": 0.43552806973457336,
+      "learning_rate": 2.868403015764222e-07,
+      "loss": 2.0408,
+      "step": 699
+    },
+    {
+      "epoch": 1.3648549841579332,
+      "grad_norm": 0.47674480080604553,
+      "learning_rate": 2.864976010966415e-07,
+      "loss": 2.0021,
+      "step": 700
+    },
+    {
+      "epoch": 1.3668047769924445,
+      "grad_norm": 0.46479421854019165,
+      "learning_rate": 2.861549006168609e-07,
+      "loss": 1.9898,
+      "step": 701
+    },
+    {
+      "epoch": 1.3687545698269559,
+      "grad_norm": 0.4399622976779938,
+      "learning_rate": 2.858122001370802e-07,
+      "loss": 1.9638,
+      "step": 702
+    },
+    {
+      "epoch": 1.3707043626614672,
+      "grad_norm": 0.442557692527771,
+      "learning_rate": 2.854694996572995e-07,
+      "loss": 2.0099,
+      "step": 703
+    },
+    {
+      "epoch": 1.3726541554959786,
+      "grad_norm": 0.4601743817329407,
+      "learning_rate": 2.851267991775189e-07,
+      "loss": 2.057,
+      "step": 704
+    },
+    {
+      "epoch": 1.37460394833049,
+      "grad_norm": 0.4959220290184021,
+      "learning_rate": 2.847840986977382e-07,
+      "loss": 2.092,
+      "step": 705
+    },
+    {
+      "epoch": 1.3765537411650013,
+      "grad_norm": 0.40172404050827026,
+      "learning_rate": 2.8444139821795747e-07,
+      "loss": 2.0074,
+      "step": 706
+    },
+    {
+      "epoch": 1.3785035339995124,
+      "grad_norm": 0.4572814404964447,
+      "learning_rate": 2.840986977381768e-07,
+      "loss": 1.9777,
+      "step": 707
+    },
+    {
+      "epoch": 1.380453326834024,
+      "grad_norm": 0.4464624524116516,
+      "learning_rate": 2.8375599725839617e-07,
+      "loss": 2.0183,
+      "step": 708
+    },
+    {
+      "epoch": 1.3824031196685351,
+      "grad_norm": 0.4498922526836395,
+      "learning_rate": 2.8341329677861546e-07,
+      "loss": 2.0975,
+      "step": 709
+    },
+    {
+      "epoch": 1.3843529125030465,
+      "grad_norm": 0.4430985748767853,
+      "learning_rate": 2.830705962988348e-07,
+      "loss": 2.027,
+      "step": 710
+    },
+    {
+      "epoch": 1.3863027053375578,
+      "grad_norm": 0.4422641694545746,
+      "learning_rate": 2.8272789581905416e-07,
+      "loss": 2.0625,
+      "step": 711
+    },
+    {
+      "epoch": 1.3882524981720692,
+      "grad_norm": 0.46121206879615784,
+      "learning_rate": 2.8238519533927345e-07,
+      "loss": 2.0135,
+      "step": 712
+    },
+    {
+      "epoch": 1.3902022910065805,
+      "grad_norm": 0.4685353934764862,
+      "learning_rate": 2.820424948594928e-07,
+      "loss": 2.071,
+      "step": 713
+    },
+    {
+      "epoch": 1.392152083841092,
+      "grad_norm": 0.43733134865760803,
+      "learning_rate": 2.8169979437971215e-07,
+      "loss": 2.0531,
+      "step": 714
+    },
+    {
+      "epoch": 1.3941018766756033,
+      "grad_norm": 0.4479463994503021,
+      "learning_rate": 2.8135709389993144e-07,
+      "loss": 2.0192,
+      "step": 715
+    },
+    {
+      "epoch": 1.3960516695101146,
+      "grad_norm": 0.4477840065956116,
+      "learning_rate": 2.810143934201508e-07,
+      "loss": 2.0408,
+      "step": 716
+    },
+    {
+      "epoch": 1.398001462344626,
+      "grad_norm": 0.44232964515686035,
+      "learning_rate": 2.8067169294037014e-07,
+      "loss": 2.0992,
+      "step": 717
+    },
+    {
+      "epoch": 1.399951255179137,
+      "grad_norm": 0.4573095142841339,
+      "learning_rate": 2.8032899246058943e-07,
+      "loss": 1.9958,
+      "step": 718
+    },
+    {
+      "epoch": 1.4019010480136487,
+      "grad_norm": 0.4734794497489929,
+      "learning_rate": 2.799862919808088e-07,
+      "loss": 2.0268,
+      "step": 719
+    },
+    {
+      "epoch": 1.4038508408481598,
+      "grad_norm": 0.4753987193107605,
+      "learning_rate": 2.7964359150102807e-07,
+      "loss": 2.0436,
+      "step": 720
+    },
+    {
+      "epoch": 1.4058006336826712,
+      "grad_norm": 0.4515923261642456,
+      "learning_rate": 2.793008910212474e-07,
+      "loss": 2.0018,
+      "step": 721
+    },
+    {
+      "epoch": 1.4077504265171825,
+      "grad_norm": 0.45925289392471313,
+      "learning_rate": 2.7895819054146677e-07,
+      "loss": 2.0454,
+      "step": 722
+    },
+    {
+      "epoch": 1.4097002193516939,
+      "grad_norm": 0.4684261083602905,
+      "learning_rate": 2.7861549006168606e-07,
+      "loss": 2.0355,
+      "step": 723
+    },
+    {
+      "epoch": 1.4116500121862052,
+      "grad_norm": 0.4723130464553833,
+      "learning_rate": 2.782727895819054e-07,
+      "loss": 2.0189,
+      "step": 724
+    },
+    {
+      "epoch": 1.4135998050207166,
+      "grad_norm": 0.43946054577827454,
+      "learning_rate": 2.7793008910212476e-07,
+      "loss": 2.0165,
+      "step": 725
+    },
+    {
+      "epoch": 1.415549597855228,
+      "grad_norm": 0.45172879099845886,
+      "learning_rate": 2.7758738862234405e-07,
+      "loss": 1.9966,
+      "step": 726
+    },
+    {
+      "epoch": 1.4174993906897393,
+      "grad_norm": 0.4361145496368408,
+      "learning_rate": 2.772446881425634e-07,
+      "loss": 1.982,
+      "step": 727
+    },
+    {
+      "epoch": 1.4194491835242506,
+      "grad_norm": 0.4422454237937927,
+      "learning_rate": 2.7690198766278275e-07,
+      "loss": 2.0032,
+      "step": 728
+    },
+    {
+      "epoch": 1.4213989763587618,
+      "grad_norm": 0.4438495934009552,
+      "learning_rate": 2.7655928718300204e-07,
+      "loss": 2.0198,
+      "step": 729
+    },
+    {
+      "epoch": 1.4233487691932731,
+      "grad_norm": 0.4422749876976013,
+      "learning_rate": 2.762165867032214e-07,
+      "loss": 1.992,
+      "step": 730
+    },
+    {
+      "epoch": 1.4252985620277845,
+      "grad_norm": 0.4652174115180969,
+      "learning_rate": 2.7587388622344074e-07,
+      "loss": 2.0345,
+      "step": 731
+    },
+    {
+      "epoch": 1.4272483548622958,
+      "grad_norm": 0.46277597546577454,
+      "learning_rate": 2.7553118574366003e-07,
+      "loss": 2.0406,
+      "step": 732
+    },
+    {
+      "epoch": 1.4291981476968072,
+      "grad_norm": 0.45579442381858826,
+      "learning_rate": 2.751884852638793e-07,
+      "loss": 2.0671,
+      "step": 733
+    },
+    {
+      "epoch": 1.4311479405313186,
+      "grad_norm": 0.43527230620384216,
+      "learning_rate": 2.748457847840987e-07,
+      "loss": 2.0433,
+      "step": 734
+    },
+    {
+      "epoch": 1.43309773336583,
+      "grad_norm": 0.4699551463127136,
+      "learning_rate": 2.74503084304318e-07,
+      "loss": 2.0366,
+      "step": 735
+    },
+    {
+      "epoch": 1.4350475262003413,
+      "grad_norm": 0.4446089565753937,
+      "learning_rate": 2.741603838245373e-07,
+      "loss": 1.9986,
+      "step": 736
+    },
+    {
+      "epoch": 1.4369973190348526,
+      "grad_norm": 0.4645906686782837,
+      "learning_rate": 2.738176833447567e-07,
+      "loss": 2.1331,
+      "step": 737
+    },
+    {
+      "epoch": 1.4389471118693637,
+      "grad_norm": 0.46871501207351685,
+      "learning_rate": 2.73474982864976e-07,
+      "loss": 2.0402,
+      "step": 738
+    },
+    {
+      "epoch": 1.4408969047038753,
+      "grad_norm": 0.4507101774215698,
+      "learning_rate": 2.731322823851953e-07,
+      "loss": 2.0027,
+      "step": 739
+    },
+    {
+      "epoch": 1.4428466975383865,
+      "grad_norm": 0.4642309546470642,
+      "learning_rate": 2.727895819054147e-07,
+      "loss": 2.0613,
+      "step": 740
+    },
+    {
+      "epoch": 1.4447964903728978,
+      "grad_norm": 0.4762292206287384,
+      "learning_rate": 2.72446881425634e-07,
+      "loss": 2.0315,
+      "step": 741
+    },
+    {
+      "epoch": 1.4467462832074092,
+      "grad_norm": 0.4549463391304016,
+      "learning_rate": 2.721041809458533e-07,
+      "loss": 2.0492,
+      "step": 742
+    },
+    {
+      "epoch": 1.4486960760419205,
+      "grad_norm": 0.4566596448421478,
+      "learning_rate": 2.717614804660727e-07,
+      "loss": 1.9571,
+      "step": 743
+    },
+    {
+      "epoch": 1.4506458688764319,
+      "grad_norm": 0.4666212797164917,
+      "learning_rate": 2.71418779986292e-07,
+      "loss": 1.9897,
+      "step": 744
+    },
+    {
+      "epoch": 1.4525956617109432,
+      "grad_norm": 0.45651644468307495,
+      "learning_rate": 2.710760795065113e-07,
+      "loss": 2.0471,
+      "step": 745
+    },
+    {
+      "epoch": 1.4545454545454546,
+      "grad_norm": 0.43935099244117737,
+      "learning_rate": 2.707333790267306e-07,
+      "loss": 1.9525,
+      "step": 746
+    },
+    {
+      "epoch": 1.456495247379966,
+      "grad_norm": 0.4813799560070038,
+      "learning_rate": 2.7039067854695e-07,
+      "loss": 2.0396,
+      "step": 747
+    },
+    {
+      "epoch": 1.4584450402144773,
+      "grad_norm": 0.4743799567222595,
+      "learning_rate": 2.7004797806716927e-07,
+      "loss": 2.0824,
+      "step": 748
+    },
+    {
+      "epoch": 1.4603948330489884,
+      "grad_norm": 0.4927983283996582,
+      "learning_rate": 2.6970527758738857e-07,
+      "loss": 2.0257,
+      "step": 749
+    },
+    {
+      "epoch": 1.4623446258835,
+      "grad_norm": 0.4711035192012787,
+      "learning_rate": 2.6936257710760797e-07,
+      "loss": 2.0487,
+      "step": 750
+    },
+    {
+      "epoch": 1.4642944187180111,
+      "grad_norm": 0.4515864849090576,
+      "learning_rate": 2.6901987662782726e-07,
+      "loss": 2.0244,
+      "step": 751
+    },
+    {
+      "epoch": 1.4662442115525225,
+      "grad_norm": 0.46076542139053345,
+      "learning_rate": 2.6867717614804656e-07,
+      "loss": 2.07,
+      "step": 752
+    },
+    {
+      "epoch": 1.4681940043870338,
+      "grad_norm": 0.44762691855430603,
+      "learning_rate": 2.6833447566826596e-07,
+      "loss": 2.0297,
+      "step": 753
+    },
+    {
+      "epoch": 1.4701437972215452,
+      "grad_norm": 0.4801499843597412,
+      "learning_rate": 2.6799177518848525e-07,
+      "loss": 2.0683,
+      "step": 754
+    },
+    {
+      "epoch": 1.4720935900560566,
+      "grad_norm": 0.45053598284721375,
+      "learning_rate": 2.6764907470870455e-07,
+      "loss": 1.9783,
+      "step": 755
+    },
+    {
+      "epoch": 1.474043382890568,
+      "grad_norm": 0.45730066299438477,
+      "learning_rate": 2.673063742289239e-07,
+      "loss": 2.0548,
+      "step": 756
+    },
+    {
+      "epoch": 1.4759931757250793,
+      "grad_norm": 0.4543995261192322,
+      "learning_rate": 2.6696367374914324e-07,
+      "loss": 2.0306,
+      "step": 757
+    },
+    {
+      "epoch": 1.4779429685595906,
+      "grad_norm": 0.4372531473636627,
+      "learning_rate": 2.6662097326936254e-07,
+      "loss": 2.0164,
+      "step": 758
+    },
+    {
+      "epoch": 1.479892761394102,
+      "grad_norm": 0.44617414474487305,
+      "learning_rate": 2.662782727895819e-07,
+      "loss": 1.9891,
+      "step": 759
+    },
+    {
+      "epoch": 1.481842554228613,
+      "grad_norm": 0.4605617821216583,
+      "learning_rate": 2.6593557230980123e-07,
+      "loss": 2.01,
+      "step": 760
+    },
+    {
+      "epoch": 1.4837923470631245,
+      "grad_norm": 0.4638999402523041,
+      "learning_rate": 2.655928718300205e-07,
+      "loss": 2.0685,
+      "step": 761
+    },
+    {
+      "epoch": 1.4857421398976358,
+      "grad_norm": 0.4548538327217102,
+      "learning_rate": 2.6525017135023987e-07,
+      "loss": 2.0665,
+      "step": 762
+    },
+    {
+      "epoch": 1.4876919327321472,
+      "grad_norm": 0.44948044419288635,
+      "learning_rate": 2.649074708704592e-07,
+      "loss": 1.9921,
+      "step": 763
+    },
+    {
+      "epoch": 1.4896417255666585,
+      "grad_norm": 0.4577581286430359,
+      "learning_rate": 2.645647703906785e-07,
+      "loss": 2.0392,
+      "step": 764
+    },
+    {
+      "epoch": 1.4915915184011699,
+      "grad_norm": 0.4821256101131439,
+      "learning_rate": 2.6422206991089786e-07,
+      "loss": 2.1304,
+      "step": 765
+    },
+    {
+      "epoch": 1.4935413112356812,
+      "grad_norm": 0.48839786648750305,
+      "learning_rate": 2.638793694311172e-07,
+      "loss": 2.0773,
+      "step": 766
+    },
+    {
+      "epoch": 1.4954911040701926,
+      "grad_norm": 0.43702590465545654,
+      "learning_rate": 2.635366689513365e-07,
+      "loss": 2.02,
+      "step": 767
+    },
+    {
+      "epoch": 1.497440896904704,
+      "grad_norm": 0.45477136969566345,
+      "learning_rate": 2.6319396847155585e-07,
+      "loss": 1.9962,
+      "step": 768
+    },
+    {
+      "epoch": 1.499390689739215,
+      "grad_norm": 0.47229456901550293,
+      "learning_rate": 2.6285126799177515e-07,
+      "loss": 2.0281,
+      "step": 769
+    },
+    {
+      "epoch": 1.5013404825737267,
+      "grad_norm": 0.4817400276660919,
+      "learning_rate": 2.625085675119945e-07,
+      "loss": 2.1009,
+      "step": 770
+    },
+    {
+      "epoch": 1.5032902754082378,
+      "grad_norm": 0.4645569324493408,
+      "learning_rate": 2.6216586703221384e-07,
+      "loss": 2.083,
+      "step": 771
+    },
+    {
+      "epoch": 1.5052400682427494,
+      "grad_norm": 0.44810667634010315,
+      "learning_rate": 2.6182316655243314e-07,
+      "loss": 2.09,
+      "step": 772
+    },
+    {
+      "epoch": 1.5071898610772605,
+      "grad_norm": 0.44432902336120605,
+      "learning_rate": 2.614804660726525e-07,
+      "loss": 2.0126,
+      "step": 773
+    },
+    {
+      "epoch": 1.5091396539117719,
+      "grad_norm": 0.4630286991596222,
+      "learning_rate": 2.6113776559287183e-07,
+      "loss": 2.0136,
+      "step": 774
+    },
+    {
+      "epoch": 1.5110894467462832,
+      "grad_norm": 0.44443148374557495,
+      "learning_rate": 2.607950651130911e-07,
+      "loss": 1.9979,
+      "step": 775
+    },
+    {
+      "epoch": 1.5130392395807946,
+      "grad_norm": 0.44903403520584106,
+      "learning_rate": 2.6045236463331047e-07,
+      "loss": 1.9788,
+      "step": 776
+    },
+    {
+      "epoch": 1.514989032415306,
+      "grad_norm": 0.45394134521484375,
+      "learning_rate": 2.601096641535298e-07,
+      "loss": 1.9529,
+      "step": 777
+    },
+    {
+      "epoch": 1.516938825249817,
+      "grad_norm": 0.46713778376579285,
+      "learning_rate": 2.597669636737491e-07,
+      "loss": 2.0212,
+      "step": 778
+    },
+    {
+      "epoch": 1.5188886180843286,
+      "grad_norm": 0.45262840390205383,
+      "learning_rate": 2.5942426319396846e-07,
+      "loss": 2.0723,
+      "step": 779
+    },
+    {
+      "epoch": 1.5208384109188398,
+      "grad_norm": 0.4648626446723938,
+      "learning_rate": 2.590815627141878e-07,
+      "loss": 2.0046,
+      "step": 780
+    },
+    {
+      "epoch": 1.5227882037533513,
+      "grad_norm": 0.4754423201084137,
+      "learning_rate": 2.587388622344071e-07,
+      "loss": 2.0434,
+      "step": 781
+    },
+    {
+      "epoch": 1.5247379965878625,
+      "grad_norm": 0.4271760880947113,
+      "learning_rate": 2.583961617546264e-07,
+      "loss": 2.0843,
+      "step": 782
+    },
+    {
+      "epoch": 1.5266877894223738,
+      "grad_norm": 0.48139727115631104,
+      "learning_rate": 2.580534612748458e-07,
+      "loss": 2.098,
+      "step": 783
+    },
+    {
+      "epoch": 1.5286375822568852,
+      "grad_norm": 0.473366379737854,
+      "learning_rate": 2.577107607950651e-07,
+      "loss": 2.0422,
+      "step": 784
+    },
+    {
+      "epoch": 1.5305873750913965,
+      "grad_norm": 0.4580918848514557,
+      "learning_rate": 2.573680603152844e-07,
+      "loss": 2.006,
+      "step": 785
+    },
+    {
+      "epoch": 1.5325371679259079,
+      "grad_norm": 0.4635441303253174,
+      "learning_rate": 2.570253598355038e-07,
+      "loss": 1.9736,
+      "step": 786
+    },
+    {
+      "epoch": 1.5344869607604192,
+      "grad_norm": 0.4621422290802002,
+      "learning_rate": 2.566826593557231e-07,
+      "loss": 2.1078,
+      "step": 787
+    },
+    {
+      "epoch": 1.5364367535949306,
+      "grad_norm": 0.4151935279369354,
+      "learning_rate": 2.563399588759424e-07,
+      "loss": 2.0092,
+      "step": 788
+    },
+    {
+      "epoch": 1.5383865464294417,
+      "grad_norm": 0.4793336093425751,
+      "learning_rate": 2.559972583961618e-07,
+      "loss": 2.0173,
+      "step": 789
+    },
+    {
+      "epoch": 1.5403363392639533,
+      "grad_norm": 0.4768364429473877,
+      "learning_rate": 2.5565455791638107e-07,
+      "loss": 2.0813,
+      "step": 790
+    },
+    {
+      "epoch": 1.5422861320984644,
+      "grad_norm": 0.452411949634552,
+      "learning_rate": 2.5531185743660037e-07,
+      "loss": 2.0527,
+      "step": 791
+    },
+    {
+      "epoch": 1.544235924932976,
+      "grad_norm": 0.44334676861763,
+      "learning_rate": 2.5496915695681977e-07,
+      "loss": 1.9701,
+      "step": 792
+    },
+    {
+      "epoch": 1.5461857177674871,
+      "grad_norm": 0.4465942978858948,
+      "learning_rate": 2.5462645647703906e-07,
+      "loss": 1.9905,
+      "step": 793
+    },
+    {
+      "epoch": 1.5481355106019985,
+      "grad_norm": 0.4681743085384369,
+      "learning_rate": 2.5428375599725836e-07,
+      "loss": 2.0654,
+      "step": 794
+    },
+    {
+      "epoch": 1.5500853034365099,
+      "grad_norm": 0.46780961751937866,
+      "learning_rate": 2.539410555174777e-07,
+      "loss": 2.0336,
+      "step": 795
+    },
+    {
+      "epoch": 1.5520350962710212,
+      "grad_norm": 0.44133254885673523,
+      "learning_rate": 2.5359835503769705e-07,
+      "loss": 1.9668,
+      "step": 796
+    },
+    {
+      "epoch": 1.5539848891055326,
+      "grad_norm": 0.45011645555496216,
+      "learning_rate": 2.5325565455791635e-07,
+      "loss": 2.0099,
+      "step": 797
+    },
+    {
+      "epoch": 1.555934681940044,
+      "grad_norm": 0.41162246465682983,
+      "learning_rate": 2.529129540781357e-07,
+      "loss": 1.9684,
+      "step": 798
+    },
+    {
+      "epoch": 1.5578844747745553,
+      "grad_norm": 0.438760906457901,
+      "learning_rate": 2.5257025359835504e-07,
+      "loss": 1.9934,
+      "step": 799
+    },
+    {
+      "epoch": 1.5598342676090664,
+      "grad_norm": 0.45921608805656433,
+      "learning_rate": 2.5222755311857434e-07,
+      "loss": 2.0447,
+      "step": 800
+    },
+    {
+      "epoch": 1.561784060443578,
+      "grad_norm": 0.4474433958530426,
+      "learning_rate": 2.518848526387937e-07,
+      "loss": 2.0508,
+      "step": 801
+    },
+    {
+      "epoch": 1.5637338532780891,
+      "grad_norm": 0.42901015281677246,
+      "learning_rate": 2.5154215215901303e-07,
+      "loss": 2.0607,
+      "step": 802
+    },
+    {
+      "epoch": 1.5656836461126007,
+      "grad_norm": 0.4604319632053375,
+      "learning_rate": 2.511994516792323e-07,
+      "loss": 2.0142,
+      "step": 803
+    },
+    {
+      "epoch": 1.5676334389471118,
+      "grad_norm": 0.4305102527141571,
+      "learning_rate": 2.5085675119945167e-07,
+      "loss": 1.9828,
+      "step": 804
+    },
+    {
+      "epoch": 1.5695832317816232,
+      "grad_norm": 0.4656990170478821,
+      "learning_rate": 2.50514050719671e-07,
+      "loss": 2.0302,
+      "step": 805
+    },
+    {
+      "epoch": 1.5715330246161345,
+      "grad_norm": 0.4602496325969696,
+      "learning_rate": 2.501713502398903e-07,
+      "loss": 2.0412,
+      "step": 806
+    },
+    {
+      "epoch": 1.5734828174506459,
+      "grad_norm": 0.4626891314983368,
+      "learning_rate": 2.4982864976010966e-07,
+      "loss": 2.0513,
+      "step": 807
+    },
+    {
+      "epoch": 1.5754326102851572,
+      "grad_norm": 0.4671951234340668,
+      "learning_rate": 2.4948594928032896e-07,
+      "loss": 2.003,
+      "step": 808
+    },
+    {
+      "epoch": 1.5773824031196684,
+      "grad_norm": 0.4399751126766205,
+      "learning_rate": 2.491432488005483e-07,
+      "loss": 2.0532,
+      "step": 809
+    },
+    {
+      "epoch": 1.57933219595418,
+      "grad_norm": 0.4228038191795349,
+      "learning_rate": 2.4880054832076765e-07,
+      "loss": 2.0078,
+      "step": 810
+    },
+    {
+      "epoch": 1.581281988788691,
+      "grad_norm": 0.4445479214191437,
+      "learning_rate": 2.4845784784098695e-07,
+      "loss": 2.0142,
+      "step": 811
+    },
+    {
+      "epoch": 1.5832317816232027,
+      "grad_norm": 0.4397488534450531,
+      "learning_rate": 2.481151473612063e-07,
+      "loss": 2.0468,
+      "step": 812
+    },
+    {
+      "epoch": 1.5851815744577138,
+      "grad_norm": 0.48187440633773804,
+      "learning_rate": 2.4777244688142564e-07,
+      "loss": 2.0444,
+      "step": 813
+    },
+    {
+      "epoch": 1.5871313672922251,
+      "grad_norm": 0.4355807304382324,
+      "learning_rate": 2.4742974640164494e-07,
+      "loss": 1.9955,
+      "step": 814
+    },
+    {
+      "epoch": 1.5890811601267365,
+      "grad_norm": 0.4219972491264343,
+      "learning_rate": 2.470870459218643e-07,
+      "loss": 1.9971,
+      "step": 815
+    },
+    {
+      "epoch": 1.5910309529612479,
+      "grad_norm": 0.44700267910957336,
+      "learning_rate": 2.4674434544208363e-07,
+      "loss": 2.0297,
+      "step": 816
+    },
+    {
+      "epoch": 1.5929807457957592,
+      "grad_norm": 0.45433923602104187,
+      "learning_rate": 2.464016449623029e-07,
+      "loss": 2.0064,
+      "step": 817
+    },
+    {
+      "epoch": 1.5949305386302706,
+      "grad_norm": 0.4188825488090515,
+      "learning_rate": 2.4605894448252227e-07,
+      "loss": 2.0236,
+      "step": 818
+    },
+    {
+      "epoch": 1.596880331464782,
+      "grad_norm": 0.4635048508644104,
+      "learning_rate": 2.457162440027416e-07,
+      "loss": 2.0652,
+      "step": 819
+    },
+    {
+      "epoch": 1.598830124299293,
+      "grad_norm": 0.4555036127567291,
+      "learning_rate": 2.453735435229609e-07,
+      "loss": 2.079,
+      "step": 820
+    },
+    {
+      "epoch": 1.6007799171338046,
+      "grad_norm": 0.45152541995048523,
+      "learning_rate": 2.4503084304318026e-07,
+      "loss": 1.9724,
+      "step": 821
+    },
+    {
+      "epoch": 1.6027297099683158,
+      "grad_norm": 0.4355667233467102,
+      "learning_rate": 2.446881425633996e-07,
+      "loss": 2.0444,
+      "step": 822
+    },
+    {
+      "epoch": 1.6046795028028273,
+      "grad_norm": 0.42853429913520813,
+      "learning_rate": 2.443454420836189e-07,
+      "loss": 1.9451,
+      "step": 823
+    },
+    {
+      "epoch": 1.6066292956373385,
+      "grad_norm": 0.4546351134777069,
+      "learning_rate": 2.4400274160383825e-07,
+      "loss": 2.015,
+      "step": 824
+    },
+    {
+      "epoch": 1.6085790884718498,
+      "grad_norm": 0.45015424489974976,
+      "learning_rate": 2.4366004112405755e-07,
+      "loss": 2.0171,
+      "step": 825
+    },
+    {
+      "epoch": 1.6105288813063612,
+      "grad_norm": 0.446065217256546,
+      "learning_rate": 2.433173406442769e-07,
+      "loss": 2.0085,
+      "step": 826
+    },
+    {
+      "epoch": 1.6124786741408725,
+      "grad_norm": 0.46771183609962463,
+      "learning_rate": 2.4297464016449624e-07,
+      "loss": 1.9844,
+      "step": 827
+    },
+    {
+      "epoch": 1.614428466975384,
+      "grad_norm": 0.4590853452682495,
+      "learning_rate": 2.4263193968471554e-07,
+      "loss": 2.0031,
+      "step": 828
+    },
+    {
+      "epoch": 1.6163782598098952,
+      "grad_norm": 0.4465842545032501,
+      "learning_rate": 2.422892392049349e-07,
+      "loss": 2.0344,
+      "step": 829
+    },
+    {
+      "epoch": 1.6183280526444066,
+      "grad_norm": 0.40251830220222473,
+      "learning_rate": 2.419465387251542e-07,
+      "loss": 2.0129,
+      "step": 830
+    },
+    {
+      "epoch": 1.6202778454789177,
+      "grad_norm": 0.45284631848335266,
+      "learning_rate": 2.416038382453735e-07,
+      "loss": 2.0354,
+      "step": 831
+    },
+    {
+      "epoch": 1.6222276383134293,
+      "grad_norm": 0.4733079969882965,
+      "learning_rate": 2.4126113776559287e-07,
+      "loss": 1.993,
+      "step": 832
+    },
+    {
+      "epoch": 1.6241774311479404,
+      "grad_norm": 0.4264031946659088,
+      "learning_rate": 2.4091843728581217e-07,
+      "loss": 2.007,
+      "step": 833
+    },
+    {
+      "epoch": 1.626127223982452,
+      "grad_norm": 0.46400555968284607,
+      "learning_rate": 2.405757368060315e-07,
+      "loss": 1.9825,
+      "step": 834
+    },
+    {
+      "epoch": 1.6280770168169632,
+      "grad_norm": 0.4408418834209442,
+      "learning_rate": 2.4023303632625086e-07,
+      "loss": 2.0199,
+      "step": 835
+    },
+    {
+      "epoch": 1.6300268096514745,
+      "grad_norm": 0.4353219270706177,
+      "learning_rate": 2.3989033584647016e-07,
+      "loss": 1.9767,
+      "step": 836
+    },
+    {
+      "epoch": 1.6319766024859859,
+      "grad_norm": 0.47256654500961304,
+      "learning_rate": 2.395476353666895e-07,
+      "loss": 2.0708,
+      "step": 837
+    },
+    {
+      "epoch": 1.6339263953204972,
+      "grad_norm": 0.44208547472953796,
+      "learning_rate": 2.392049348869088e-07,
+      "loss": 2.0518,
+      "step": 838
+    },
+    {
+      "epoch": 1.6358761881550086,
+      "grad_norm": 0.4937672019004822,
+      "learning_rate": 2.3886223440712815e-07,
+      "loss": 2.043,
+      "step": 839
+    },
+    {
+      "epoch": 1.6378259809895197,
+      "grad_norm": 0.46095776557922363,
+      "learning_rate": 2.385195339273475e-07,
+      "loss": 2.0421,
+      "step": 840
+    },
+    {
+      "epoch": 1.6397757738240313,
+      "grad_norm": 0.4658643901348114,
+      "learning_rate": 2.3817683344756682e-07,
+      "loss": 2.0225,
+      "step": 841
+    },
+    {
+      "epoch": 1.6417255666585424,
+      "grad_norm": 0.4451207220554352,
+      "learning_rate": 2.3783413296778616e-07,
+      "loss": 2.0244,
+      "step": 842
+    },
+    {
+      "epoch": 1.643675359493054,
+      "grad_norm": 0.43841567635536194,
+      "learning_rate": 2.3749143248800546e-07,
+      "loss": 1.9797,
+      "step": 843
+    },
+    {
+      "epoch": 1.6456251523275651,
+      "grad_norm": 0.45495790243148804,
+      "learning_rate": 2.371487320082248e-07,
+      "loss": 2.039,
+      "step": 844
+    },
+    {
+      "epoch": 1.6475749451620765,
+      "grad_norm": 0.4694961607456207,
+      "learning_rate": 2.3680603152844415e-07,
+      "loss": 2.0232,
+      "step": 845
+    },
+    {
+      "epoch": 1.6495247379965878,
+      "grad_norm": 0.4593546986579895,
+      "learning_rate": 2.3646333104866345e-07,
+      "loss": 2.0495,
+      "step": 846
+    },
+    {
+      "epoch": 1.6514745308310992,
+      "grad_norm": 0.4738862216472626,
+      "learning_rate": 2.361206305688828e-07,
+      "loss": 2.0105,
+      "step": 847
+    },
+    {
+      "epoch": 1.6534243236656105,
+      "grad_norm": 0.45088139176368713,
+      "learning_rate": 2.357779300891021e-07,
+      "loss": 2.0418,
+      "step": 848
+    },
+    {
+      "epoch": 1.655374116500122,
+      "grad_norm": 0.4501790702342987,
+      "learning_rate": 2.3543522960932144e-07,
+      "loss": 2.0531,
+      "step": 849
+    },
+    {
+      "epoch": 1.6573239093346332,
+      "grad_norm": 0.47187909483909607,
+      "learning_rate": 2.3509252912954078e-07,
+      "loss": 1.9907,
+      "step": 850
+    },
+    {
+      "epoch": 1.6592737021691444,
+      "grad_norm": 0.46769675612449646,
+      "learning_rate": 2.3474982864976008e-07,
+      "loss": 2.0145,
+      "step": 851
+    },
+    {
+      "epoch": 1.661223495003656,
+      "grad_norm": 0.44854676723480225,
+      "learning_rate": 2.3440712816997943e-07,
+      "loss": 2.0381,
+      "step": 852
+    },
+    {
+      "epoch": 1.663173287838167,
+      "grad_norm": 0.4576641023159027,
+      "learning_rate": 2.3406442769019877e-07,
+      "loss": 1.9722,
+      "step": 853
+    },
+    {
+      "epoch": 1.6651230806726787,
+      "grad_norm": 0.4568294584751129,
+      "learning_rate": 2.3372172721041807e-07,
+      "loss": 1.9744,
+      "step": 854
+    },
+    {
+      "epoch": 1.6670728735071898,
+      "grad_norm": 0.4591883718967438,
+      "learning_rate": 2.3337902673063742e-07,
+      "loss": 1.9666,
+      "step": 855
+    },
+    {
+      "epoch": 1.6690226663417012,
+      "grad_norm": 0.44672197103500366,
+      "learning_rate": 2.3303632625085674e-07,
+      "loss": 1.9944,
+      "step": 856
+    },
+    {
+      "epoch": 1.6709724591762125,
+      "grad_norm": 0.4896506667137146,
+      "learning_rate": 2.3269362577107606e-07,
+      "loss": 2.0492,
+      "step": 857
+    },
+    {
+      "epoch": 1.6729222520107239,
+      "grad_norm": 0.4453061521053314,
+      "learning_rate": 2.323509252912954e-07,
+      "loss": 1.9757,
+      "step": 858
+    },
+    {
+      "epoch": 1.6748720448452352,
+      "grad_norm": 0.4569021761417389,
+      "learning_rate": 2.3200822481151473e-07,
+      "loss": 2.0523,
+      "step": 859
+    },
+    {
+      "epoch": 1.6768218376797466,
+      "grad_norm": 0.4553905427455902,
+      "learning_rate": 2.3166552433173405e-07,
+      "loss": 2.0189,
+      "step": 860
+    },
+    {
+      "epoch": 1.678771630514258,
+      "grad_norm": 0.4560829699039459,
+      "learning_rate": 2.3132282385195337e-07,
+      "loss": 2.0833,
+      "step": 861
+    },
+    {
+      "epoch": 1.680721423348769,
+      "grad_norm": 0.4487151503562927,
+      "learning_rate": 2.3098012337217272e-07,
+      "loss": 1.9806,
+      "step": 862
+    },
+    {
+      "epoch": 1.6826712161832806,
+      "grad_norm": 0.440891832113266,
+      "learning_rate": 2.3063742289239204e-07,
+      "loss": 1.9989,
+      "step": 863
+    },
+    {
+      "epoch": 1.6846210090177918,
+      "grad_norm": 0.469881534576416,
+      "learning_rate": 2.3029472241261136e-07,
+      "loss": 2.0626,
+      "step": 864
+    },
+    {
+      "epoch": 1.6865708018523033,
+      "grad_norm": 0.43621349334716797,
+      "learning_rate": 2.299520219328307e-07,
+      "loss": 2.063,
+      "step": 865
+    },
+    {
+      "epoch": 1.6885205946868145,
+      "grad_norm": 0.45750436186790466,
+      "learning_rate": 2.2960932145305003e-07,
+      "loss": 2.0164,
+      "step": 866
+    },
+    {
+      "epoch": 1.6904703875213258,
+      "grad_norm": 0.46832090616226196,
+      "learning_rate": 2.2926662097326935e-07,
+      "loss": 2.0459,
+      "step": 867
+    },
+    {
+      "epoch": 1.6924201803558372,
+      "grad_norm": 0.4424852728843689,
+      "learning_rate": 2.289239204934887e-07,
+      "loss": 2.0148,
+      "step": 868
+    },
+    {
+      "epoch": 1.6943699731903485,
+      "grad_norm": 0.4639265239238739,
+      "learning_rate": 2.28581220013708e-07,
+      "loss": 2.0453,
+      "step": 869
+    },
+    {
+      "epoch": 1.69631976602486,
+      "grad_norm": 0.42720574140548706,
+      "learning_rate": 2.2823851953392734e-07,
+      "loss": 2.0164,
+      "step": 870
+    },
+    {
+      "epoch": 1.698269558859371,
+      "grad_norm": 0.46615973114967346,
+      "learning_rate": 2.2789581905414668e-07,
+      "loss": 2.0235,
+      "step": 871
+    },
+    {
+      "epoch": 1.7002193516938826,
+      "grad_norm": 0.46956273913383484,
+      "learning_rate": 2.2755311857436598e-07,
+      "loss": 2.0668,
+      "step": 872
+    },
+    {
+      "epoch": 1.7021691445283937,
+      "grad_norm": 0.45590096712112427,
+      "learning_rate": 2.2721041809458533e-07,
+      "loss": 2.0767,
+      "step": 873
+    },
+    {
+      "epoch": 1.7041189373629053,
+      "grad_norm": 0.4419032037258148,
+      "learning_rate": 2.2686771761480465e-07,
+      "loss": 2.0298,
+      "step": 874
+    },
+    {
+      "epoch": 1.7060687301974164,
+      "grad_norm": 0.48438993096351624,
+      "learning_rate": 2.2652501713502397e-07,
+      "loss": 2.0881,
+      "step": 875
+    },
+    {
+      "epoch": 1.7080185230319278,
+      "grad_norm": 0.4674246609210968,
+      "learning_rate": 2.2618231665524332e-07,
+      "loss": 1.9858,
+      "step": 876
+    },
+    {
+      "epoch": 1.7099683158664392,
+      "grad_norm": 0.4731968641281128,
+      "learning_rate": 2.2583961617546264e-07,
+      "loss": 2.0684,
+      "step": 877
+    },
+    {
+      "epoch": 1.7119181087009505,
+      "grad_norm": 0.44370540976524353,
+      "learning_rate": 2.2549691569568196e-07,
+      "loss": 2.0222,
+      "step": 878
+    },
+    {
+      "epoch": 1.7138679015354619,
+      "grad_norm": 0.43057727813720703,
+      "learning_rate": 2.251542152159013e-07,
+      "loss": 2.0054,
+      "step": 879
+    },
+    {
+      "epoch": 1.7158176943699732,
+      "grad_norm": 0.4575825035572052,
+      "learning_rate": 2.2481151473612063e-07,
+      "loss": 2.0194,
+      "step": 880
+    },
+    {
+      "epoch": 1.7177674872044846,
+      "grad_norm": 0.46100616455078125,
+      "learning_rate": 2.2446881425633995e-07,
+      "loss": 2.0362,
+      "step": 881
+    },
+    {
+      "epoch": 1.7197172800389957,
+      "grad_norm": 0.46780040860176086,
+      "learning_rate": 2.2412611377655927e-07,
+      "loss": 2.0458,
+      "step": 882
+    },
+    {
+      "epoch": 1.7216670728735073,
+      "grad_norm": 0.4316709339618683,
+      "learning_rate": 2.2378341329677862e-07,
+      "loss": 2.0401,
+      "step": 883
+    },
+    {
+      "epoch": 1.7236168657080184,
+      "grad_norm": 0.43883568048477173,
+      "learning_rate": 2.2344071281699794e-07,
+      "loss": 2.0407,
+      "step": 884
+    },
+    {
+      "epoch": 1.72556665854253,
+      "grad_norm": 0.44989317655563354,
+      "learning_rate": 2.2309801233721726e-07,
+      "loss": 2.0253,
+      "step": 885
+    },
+    {
+      "epoch": 1.7275164513770411,
+      "grad_norm": 0.4468737840652466,
+      "learning_rate": 2.227553118574366e-07,
+      "loss": 2.0336,
+      "step": 886
+    },
+    {
+      "epoch": 1.7294662442115525,
+      "grad_norm": 0.45126405358314514,
+      "learning_rate": 2.224126113776559e-07,
+      "loss": 2.0259,
+      "step": 887
+    },
+    {
+      "epoch": 1.7314160370460638,
+      "grad_norm": 0.43270209431648254,
+      "learning_rate": 2.2206991089787525e-07,
+      "loss": 2.0071,
+      "step": 888
+    },
+    {
+      "epoch": 1.7333658298805752,
+      "grad_norm": 0.4503726363182068,
+      "learning_rate": 2.217272104180946e-07,
+      "loss": 2.1025,
+      "step": 889
+    },
+    {
+      "epoch": 1.7353156227150865,
+      "grad_norm": 0.44900792837142944,
+      "learning_rate": 2.213845099383139e-07,
+      "loss": 1.9883,
+      "step": 890
+    },
+    {
+      "epoch": 1.737265415549598,
+      "grad_norm": 0.4531221091747284,
+      "learning_rate": 2.2104180945853324e-07,
+      "loss": 2.0095,
+      "step": 891
+    },
+    {
+      "epoch": 1.7392152083841093,
+      "grad_norm": 0.46359124779701233,
+      "learning_rate": 2.2069910897875258e-07,
+      "loss": 2.003,
+      "step": 892
+    },
+    {
+      "epoch": 1.7411650012186204,
+      "grad_norm": 0.4506163001060486,
+      "learning_rate": 2.2035640849897188e-07,
+      "loss": 1.9438,
+      "step": 893
+    },
+    {
+      "epoch": 1.743114794053132,
+      "grad_norm": 0.4618943929672241,
+      "learning_rate": 2.2001370801919123e-07,
+      "loss": 2.0772,
+      "step": 894
+    },
+    {
+      "epoch": 1.745064586887643,
+      "grad_norm": 0.4341379404067993,
+      "learning_rate": 2.1967100753941055e-07,
+      "loss": 1.9443,
+      "step": 895
+    },
+    {
+      "epoch": 1.7470143797221547,
+      "grad_norm": 0.4800126254558563,
+      "learning_rate": 2.1932830705962987e-07,
+      "loss": 1.9994,
+      "step": 896
+    },
+    {
+      "epoch": 1.7489641725566658,
+      "grad_norm": 0.45474764704704285,
+      "learning_rate": 2.1898560657984922e-07,
+      "loss": 2.0635,
+      "step": 897
+    },
+    {
+      "epoch": 1.7509139653911772,
+      "grad_norm": 0.44301092624664307,
+      "learning_rate": 2.1864290610006854e-07,
+      "loss": 1.9752,
+      "step": 898
+    },
+    {
+      "epoch": 1.7528637582256885,
+      "grad_norm": 0.4428479075431824,
+      "learning_rate": 2.1830020562028786e-07,
+      "loss": 1.9371,
+      "step": 899
+    },
+    {
+      "epoch": 1.7548135510601999,
+      "grad_norm": 0.4576126039028168,
+      "learning_rate": 2.1795750514050718e-07,
+      "loss": 2.063,
+      "step": 900
+    },
+    {
+      "epoch": 1.7567633438947112,
+      "grad_norm": 0.47722387313842773,
+      "learning_rate": 2.1761480466072653e-07,
+      "loss": 2.0743,
+      "step": 901
+    },
+    {
+      "epoch": 1.7587131367292224,
+      "grad_norm": 0.4575481712818146,
+      "learning_rate": 2.1727210418094585e-07,
+      "loss": 1.9873,
+      "step": 902
+    },
+    {
+      "epoch": 1.760662929563734,
+      "grad_norm": 0.4340214729309082,
+      "learning_rate": 2.1692940370116517e-07,
+      "loss": 1.9459,
+      "step": 903
+    },
+    {
+      "epoch": 1.762612722398245,
+      "grad_norm": 0.41616639494895935,
+      "learning_rate": 2.1658670322138452e-07,
+      "loss": 1.9505,
+      "step": 904
+    },
+    {
+      "epoch": 1.7645625152327566,
+      "grad_norm": 0.472650408744812,
+      "learning_rate": 2.162440027416038e-07,
+      "loss": 2.0594,
+      "step": 905
+    },
+    {
+      "epoch": 1.7665123080672678,
+      "grad_norm": 0.4756447374820709,
+      "learning_rate": 2.1590130226182316e-07,
+      "loss": 1.9695,
+      "step": 906
+    },
+    {
+      "epoch": 1.7684621009017791,
+      "grad_norm": 0.44738152623176575,
+      "learning_rate": 2.155586017820425e-07,
+      "loss": 2.0771,
+      "step": 907
+    },
+    {
+      "epoch": 1.7704118937362905,
+      "grad_norm": 0.4602157771587372,
+      "learning_rate": 2.152159013022618e-07,
+      "loss": 2.0813,
+      "step": 908
+    },
+    {
+      "epoch": 1.7723616865708018,
+      "grad_norm": 0.46765050292015076,
+      "learning_rate": 2.1487320082248115e-07,
+      "loss": 2.0801,
+      "step": 909
+    },
+    {
+      "epoch": 1.7743114794053132,
+      "grad_norm": 0.4703747034072876,
+      "learning_rate": 2.145305003427005e-07,
+      "loss": 2.0093,
+      "step": 910
+    },
+    {
+      "epoch": 1.7762612722398246,
+      "grad_norm": 0.48457059264183044,
+      "learning_rate": 2.141877998629198e-07,
+      "loss": 2.0528,
+      "step": 911
+    },
+    {
+      "epoch": 1.778211065074336,
+      "grad_norm": 0.478710412979126,
+      "learning_rate": 2.1384509938313914e-07,
+      "loss": 2.1099,
+      "step": 912
+    },
+    {
+      "epoch": 1.780160857908847,
+      "grad_norm": 0.4458109438419342,
+      "learning_rate": 2.1350239890335843e-07,
+      "loss": 2.0592,
+      "step": 913
+    },
+    {
+      "epoch": 1.7821106507433586,
+      "grad_norm": 0.4474625885486603,
+      "learning_rate": 2.1315969842357778e-07,
+      "loss": 2.0055,
+      "step": 914
+    },
+    {
+      "epoch": 1.7840604435778697,
+      "grad_norm": 0.4586813151836395,
+      "learning_rate": 2.1281699794379713e-07,
+      "loss": 2.0131,
+      "step": 915
+    },
+    {
+      "epoch": 1.7860102364123813,
+      "grad_norm": 0.45083218812942505,
+      "learning_rate": 2.1247429746401642e-07,
+      "loss": 2.0437,
+      "step": 916
+    },
+    {
+      "epoch": 1.7879600292468925,
+      "grad_norm": 0.44078171253204346,
+      "learning_rate": 2.1213159698423577e-07,
+      "loss": 1.9792,
+      "step": 917
+    },
+    {
+      "epoch": 1.7899098220814038,
+      "grad_norm": 0.4346940219402313,
+      "learning_rate": 2.117888965044551e-07,
+      "loss": 1.9933,
+      "step": 918
+    },
+    {
+      "epoch": 1.7918596149159152,
+      "grad_norm": 0.45846906304359436,
+      "learning_rate": 2.114461960246744e-07,
+      "loss": 1.9682,
+      "step": 919
+    },
+    {
+      "epoch": 1.7938094077504265,
+      "grad_norm": 0.4335155785083771,
+      "learning_rate": 2.1110349554489376e-07,
+      "loss": 2.03,
+      "step": 920
+    },
+    {
+      "epoch": 1.7957592005849379,
+      "grad_norm": 0.4618023633956909,
+      "learning_rate": 2.1076079506511308e-07,
+      "loss": 2.0966,
+      "step": 921
+    },
+    {
+      "epoch": 1.7977089934194492,
+      "grad_norm": 0.46044906973838806,
+      "learning_rate": 2.104180945853324e-07,
+      "loss": 2.0873,
+      "step": 922
+    },
+    {
+      "epoch": 1.7996587862539606,
+      "grad_norm": 0.4635170102119446,
+      "learning_rate": 2.1007539410555175e-07,
+      "loss": 1.9897,
+      "step": 923
+    },
+    {
+      "epoch": 1.8016085790884717,
+      "grad_norm": 0.4335494637489319,
+      "learning_rate": 2.0973269362577107e-07,
+      "loss": 2.0228,
+      "step": 924
+    },
+    {
+      "epoch": 1.8035583719229833,
+      "grad_norm": 0.44605642557144165,
+      "learning_rate": 2.093899931459904e-07,
+      "loss": 2.0561,
+      "step": 925
+    },
+    {
+      "epoch": 1.8055081647574944,
+      "grad_norm": 0.4611765146255493,
+      "learning_rate": 2.090472926662097e-07,
+      "loss": 2.0329,
+      "step": 926
+    },
+    {
+      "epoch": 1.807457957592006,
+      "grad_norm": 0.443036288022995,
+      "learning_rate": 2.0870459218642906e-07,
+      "loss": 1.9565,
+      "step": 927
+    },
+    {
+      "epoch": 1.8094077504265171,
+      "grad_norm": 0.4552265405654907,
+      "learning_rate": 2.0836189170664838e-07,
+      "loss": 2.0842,
+      "step": 928
+    },
+    {
+      "epoch": 1.8113575432610285,
+      "grad_norm": 0.41511160135269165,
+      "learning_rate": 2.080191912268677e-07,
+      "loss": 2.0043,
+      "step": 929
+    },
+    {
+      "epoch": 1.8133073360955398,
+      "grad_norm": 0.44421470165252686,
+      "learning_rate": 2.0767649074708705e-07,
+      "loss": 2.0433,
+      "step": 930
+    },
+    {
+      "epoch": 1.8152571289300512,
+      "grad_norm": 0.43709036707878113,
+      "learning_rate": 2.0733379026730634e-07,
+      "loss": 2.0405,
+      "step": 931
+    },
+    {
+      "epoch": 1.8172069217645626,
+      "grad_norm": 0.429074227809906,
+      "learning_rate": 2.069910897875257e-07,
+      "loss": 1.964,
+      "step": 932
+    },
+    {
+      "epoch": 1.8191567145990737,
+      "grad_norm": 0.4392930269241333,
+      "learning_rate": 2.0664838930774504e-07,
+      "loss": 1.9819,
+      "step": 933
+    },
+    {
+      "epoch": 1.8211065074335853,
+      "grad_norm": 0.41590166091918945,
+      "learning_rate": 2.0630568882796433e-07,
+      "loss": 1.9821,
+      "step": 934
+    },
+    {
+      "epoch": 1.8230563002680964,
+      "grad_norm": 0.445362389087677,
+      "learning_rate": 2.0596298834818368e-07,
+      "loss": 2.092,
+      "step": 935
+    },
+    {
+      "epoch": 1.825006093102608,
+      "grad_norm": 0.43674713373184204,
+      "learning_rate": 2.0562028786840303e-07,
+      "loss": 2.0371,
+      "step": 936
+    },
+    {
+      "epoch": 1.826955885937119,
+      "grad_norm": 0.4520663022994995,
+      "learning_rate": 2.0527758738862232e-07,
+      "loss": 2.0329,
+      "step": 937
+    },
+    {
+      "epoch": 1.8289056787716305,
+      "grad_norm": 0.4744395613670349,
+      "learning_rate": 2.0493488690884167e-07,
+      "loss": 2.0828,
+      "step": 938
+    },
+    {
+      "epoch": 1.8308554716061418,
+      "grad_norm": 0.45714208483695984,
+      "learning_rate": 2.04592186429061e-07,
+      "loss": 2.017,
+      "step": 939
+    },
+    {
+      "epoch": 1.8328052644406532,
+      "grad_norm": 0.4604392647743225,
+      "learning_rate": 2.042494859492803e-07,
+      "loss": 1.9813,
+      "step": 940
+    },
+    {
+      "epoch": 1.8347550572751645,
+      "grad_norm": 0.43890222907066345,
+      "learning_rate": 2.0390678546949966e-07,
+      "loss": 1.9902,
+      "step": 941
+    },
+    {
+      "epoch": 1.8367048501096759,
+      "grad_norm": 0.44383513927459717,
+      "learning_rate": 2.0356408498971898e-07,
+      "loss": 2.0434,
+      "step": 942
+    },
+    {
+      "epoch": 1.8386546429441872,
+      "grad_norm": 0.43706512451171875,
+      "learning_rate": 2.032213845099383e-07,
+      "loss": 2.052,
+      "step": 943
+    },
+    {
+      "epoch": 1.8406044357786984,
+      "grad_norm": 0.427843302488327,
+      "learning_rate": 2.0287868403015762e-07,
+      "loss": 1.8841,
+      "step": 944
+    },
+    {
+      "epoch": 1.84255422861321,
+      "grad_norm": 0.4639602601528168,
+      "learning_rate": 2.0253598355037697e-07,
+      "loss": 2.0831,
+      "step": 945
+    },
+    {
+      "epoch": 1.844504021447721,
+      "grad_norm": 0.44139614701271057,
+      "learning_rate": 2.021932830705963e-07,
+      "loss": 1.9867,
+      "step": 946
+    },
+    {
+      "epoch": 1.8464538142822327,
+      "grad_norm": 0.4408351182937622,
+      "learning_rate": 2.018505825908156e-07,
+      "loss": 2.0199,
+      "step": 947
+    },
+    {
+      "epoch": 1.8484036071167438,
+      "grad_norm": 0.49647897481918335,
+      "learning_rate": 2.0150788211103496e-07,
+      "loss": 2.0877,
+      "step": 948
+    },
+    {
+      "epoch": 1.8503533999512551,
+      "grad_norm": 0.46033725142478943,
+      "learning_rate": 2.0116518163125428e-07,
+      "loss": 2.0584,
+      "step": 949
+    },
+    {
+      "epoch": 1.8523031927857665,
+      "grad_norm": 0.4471881687641144,
+      "learning_rate": 2.008224811514736e-07,
+      "loss": 1.9694,
+      "step": 950
+    },
+    {
+      "epoch": 1.8542529856202778,
+      "grad_norm": 0.435660183429718,
+      "learning_rate": 2.0047978067169295e-07,
+      "loss": 2.0025,
+      "step": 951
+    },
+    {
+      "epoch": 1.8562027784547892,
+      "grad_norm": 0.4504587650299072,
+      "learning_rate": 2.0013708019191224e-07,
+      "loss": 2.0403,
+      "step": 952
+    },
+    {
+      "epoch": 1.8581525712893006,
+      "grad_norm": 0.446451336145401,
+      "learning_rate": 1.997943797121316e-07,
+      "loss": 1.9817,
+      "step": 953
+    },
+    {
+      "epoch": 1.860102364123812,
+      "grad_norm": 0.46191105246543884,
+      "learning_rate": 1.9945167923235094e-07,
+      "loss": 2.0329,
+      "step": 954
+    },
+    {
+      "epoch": 1.862052156958323,
+      "grad_norm": 0.4477747976779938,
+      "learning_rate": 1.9910897875257023e-07,
+      "loss": 2.0113,
+      "step": 955
+    },
+    {
+      "epoch": 1.8640019497928346,
+      "grad_norm": 0.46400219202041626,
+      "learning_rate": 1.9876627827278958e-07,
+      "loss": 2.0142,
+      "step": 956
+    },
+    {
+      "epoch": 1.8659517426273458,
+      "grad_norm": 0.45763564109802246,
+      "learning_rate": 1.984235777930089e-07,
+      "loss": 2.0555,
+      "step": 957
+    },
+    {
+      "epoch": 1.8679015354618573,
+      "grad_norm": 0.4603627920150757,
+      "learning_rate": 1.9808087731322822e-07,
+      "loss": 2.0022,
+      "step": 958
+    },
+    {
+      "epoch": 1.8698513282963685,
+      "grad_norm": 0.5134696364402771,
+      "learning_rate": 1.9773817683344757e-07,
+      "loss": 2.0396,
+      "step": 959
+    },
+    {
+      "epoch": 1.8718011211308798,
+      "grad_norm": 0.46097123622894287,
+      "learning_rate": 1.973954763536669e-07,
+      "loss": 2.0887,
+      "step": 960
+    },
+    {
+      "epoch": 1.8737509139653912,
+      "grad_norm": 0.45269545912742615,
+      "learning_rate": 1.970527758738862e-07,
+      "loss": 2.0184,
+      "step": 961
+    },
+    {
+      "epoch": 1.8757007067999025,
+      "grad_norm": 0.463885635137558,
+      "learning_rate": 1.9671007539410553e-07,
+      "loss": 2.0701,
+      "step": 962
+    },
+    {
+      "epoch": 1.8776504996344139,
+      "grad_norm": 0.4765574634075165,
+      "learning_rate": 1.9636737491432488e-07,
+      "loss": 1.9951,
+      "step": 963
+    },
+    {
+      "epoch": 1.879600292468925,
+      "grad_norm": 0.48183631896972656,
+      "learning_rate": 1.960246744345442e-07,
+      "loss": 2.0723,
+      "step": 964
+    },
+    {
+      "epoch": 1.8815500853034366,
+      "grad_norm": 0.44266360998153687,
+      "learning_rate": 1.9568197395476352e-07,
+      "loss": 2.0134,
+      "step": 965
+    },
+    {
+      "epoch": 1.8834998781379477,
+      "grad_norm": 0.4508133828639984,
+      "learning_rate": 1.9533927347498287e-07,
+      "loss": 1.9951,
+      "step": 966
+    },
+    {
+      "epoch": 1.8854496709724593,
+      "grad_norm": 0.4255620539188385,
+      "learning_rate": 1.949965729952022e-07,
+      "loss": 1.9663,
+      "step": 967
+    },
+    {
+      "epoch": 1.8873994638069704,
+      "grad_norm": 0.45423394441604614,
+      "learning_rate": 1.946538725154215e-07,
+      "loss": 2.0072,
+      "step": 968
+    },
+    {
+      "epoch": 1.8893492566414818,
+      "grad_norm": 0.4226663112640381,
+      "learning_rate": 1.9431117203564086e-07,
+      "loss": 1.9598,
+      "step": 969
+    },
+    {
+      "epoch": 1.8912990494759931,
+      "grad_norm": 0.47366762161254883,
+      "learning_rate": 1.9396847155586015e-07,
+      "loss": 1.9927,
+      "step": 970
+    },
+    {
+      "epoch": 1.8932488423105045,
+      "grad_norm": 0.44758790731430054,
+      "learning_rate": 1.936257710760795e-07,
+      "loss": 1.9628,
+      "step": 971
+    },
+    {
+      "epoch": 1.8951986351450159,
+      "grad_norm": 0.48197463154792786,
+      "learning_rate": 1.9328307059629885e-07,
+      "loss": 2.1004,
+      "step": 972
+    },
+    {
+      "epoch": 1.8971484279795272,
+      "grad_norm": 0.4538448750972748,
+      "learning_rate": 1.9294037011651814e-07,
+      "loss": 2.0199,
+      "step": 973
+    },
+    {
+      "epoch": 1.8990982208140386,
+      "grad_norm": 0.47362738847732544,
+      "learning_rate": 1.925976696367375e-07,
+      "loss": 2.0746,
+      "step": 974
+    },
+    {
+      "epoch": 1.9010480136485497,
+      "grad_norm": 0.47095638513565063,
+      "learning_rate": 1.922549691569568e-07,
+      "loss": 1.9897,
+      "step": 975
+    },
+    {
+      "epoch": 1.9029978064830613,
+      "grad_norm": 0.4763641059398651,
+      "learning_rate": 1.9191226867717613e-07,
+      "loss": 2.0156,
+      "step": 976
+    },
+    {
+      "epoch": 1.9049475993175724,
+      "grad_norm": 0.4224942922592163,
+      "learning_rate": 1.9156956819739548e-07,
+      "loss": 2.0114,
+      "step": 977
+    },
+    {
+      "epoch": 1.906897392152084,
+      "grad_norm": 0.44930440187454224,
+      "learning_rate": 1.912268677176148e-07,
+      "loss": 2.0121,
+      "step": 978
+    },
+    {
+      "epoch": 1.9088471849865951,
+      "grad_norm": 0.45916110277175903,
+      "learning_rate": 1.9088416723783412e-07,
+      "loss": 2.0053,
+      "step": 979
+    },
+    {
+      "epoch": 1.9107969778211065,
+      "grad_norm": 0.42759600281715393,
+      "learning_rate": 1.9054146675805347e-07,
+      "loss": 2.0109,
+      "step": 980
+    },
+    {
+      "epoch": 1.9127467706556178,
+      "grad_norm": 0.49347975850105286,
+      "learning_rate": 1.901987662782728e-07,
+      "loss": 2.0657,
+      "step": 981
+    },
+    {
+      "epoch": 1.9146965634901292,
+      "grad_norm": 0.4315294027328491,
+      "learning_rate": 1.898560657984921e-07,
+      "loss": 1.9473,
+      "step": 982
+    },
+    {
+      "epoch": 1.9166463563246405,
+      "grad_norm": 0.42915600538253784,
+      "learning_rate": 1.8951336531871143e-07,
+      "loss": 1.9958,
+      "step": 983
+    },
+    {
+      "epoch": 1.9185961491591519,
+      "grad_norm": 0.48152124881744385,
+      "learning_rate": 1.8917066483893078e-07,
+      "loss": 2.0815,
+      "step": 984
+    },
+    {
+      "epoch": 1.9205459419936632,
+      "grad_norm": 0.44423532485961914,
+      "learning_rate": 1.888279643591501e-07,
+      "loss": 2.0227,
+      "step": 985
+    },
+    {
+      "epoch": 1.9224957348281744,
+      "grad_norm": 0.4499359130859375,
+      "learning_rate": 1.8848526387936942e-07,
+      "loss": 1.961,
+      "step": 986
+    },
+    {
+      "epoch": 1.924445527662686,
+      "grad_norm": 0.4560549855232239,
+      "learning_rate": 1.8814256339958877e-07,
+      "loss": 2.03,
+      "step": 987
+    },
+    {
+      "epoch": 1.926395320497197,
+      "grad_norm": 0.48396381735801697,
+      "learning_rate": 1.8779986291980806e-07,
+      "loss": 1.985,
+      "step": 988
+    },
+    {
+      "epoch": 1.9283451133317087,
+      "grad_norm": 0.456910103559494,
+      "learning_rate": 1.874571624400274e-07,
+      "loss": 1.9802,
+      "step": 989
+    },
+    {
+      "epoch": 1.9302949061662198,
+      "grad_norm": 0.46041303873062134,
+      "learning_rate": 1.8711446196024676e-07,
+      "loss": 1.9507,
+      "step": 990
+    },
+    {
+      "epoch": 1.9322446990007311,
+      "grad_norm": 0.4496663510799408,
+      "learning_rate": 1.8677176148046605e-07,
+      "loss": 2.0329,
+      "step": 991
+    },
+    {
+      "epoch": 1.9341944918352425,
+      "grad_norm": 0.4381345212459564,
+      "learning_rate": 1.864290610006854e-07,
+      "loss": 1.9643,
+      "step": 992
+    },
+    {
+      "epoch": 1.9361442846697539,
+      "grad_norm": 0.43699464201927185,
+      "learning_rate": 1.8608636052090475e-07,
+      "loss": 2.026,
+      "step": 993
+    },
+    {
+      "epoch": 1.9380940775042652,
+      "grad_norm": 0.4496040344238281,
+      "learning_rate": 1.8574366004112404e-07,
+      "loss": 1.9318,
+      "step": 994
+    },
+    {
+      "epoch": 1.9400438703387763,
+      "grad_norm": 0.45028945803642273,
+      "learning_rate": 1.854009595613434e-07,
+      "loss": 2.0254,
+      "step": 995
+    },
+    {
+      "epoch": 1.941993663173288,
+      "grad_norm": 0.46241873502731323,
+      "learning_rate": 1.8505825908156268e-07,
+      "loss": 2.0224,
+      "step": 996
+    },
+    {
+      "epoch": 1.943943456007799,
+      "grad_norm": 0.4494277238845825,
+      "learning_rate": 1.8471555860178203e-07,
+      "loss": 2.0734,
+      "step": 997
+    },
+    {
+      "epoch": 1.9458932488423106,
+      "grad_norm": 0.44225579500198364,
+      "learning_rate": 1.8437285812200138e-07,
+      "loss": 2.0548,
+      "step": 998
+    },
+    {
+      "epoch": 1.9478430416768218,
+      "grad_norm": 0.4850820004940033,
+      "learning_rate": 1.8403015764222067e-07,
+      "loss": 1.9961,
+      "step": 999
+    },
+    {
+      "epoch": 1.9497928345113331,
+      "grad_norm": 0.46442610025405884,
+      "learning_rate": 1.8368745716244002e-07,
+      "loss": 1.9777,
+      "step": 1000
+    },
+    {
+      "epoch": 1.9517426273458445,
+      "grad_norm": 0.457109272480011,
+      "learning_rate": 1.8334475668265934e-07,
+      "loss": 2.0949,
+      "step": 1001
+    },
+    {
+      "epoch": 1.9536924201803558,
+      "grad_norm": 0.4514349699020386,
+      "learning_rate": 1.8300205620287866e-07,
+      "loss": 2.0933,
+      "step": 1002
+    },
+    {
+      "epoch": 1.9556422130148672,
+      "grad_norm": 0.4601777195930481,
+      "learning_rate": 1.82659355723098e-07,
+      "loss": 1.9975,
+      "step": 1003
+    },
+    {
+      "epoch": 1.9575920058493785,
+      "grad_norm": 0.4604569673538208,
+      "learning_rate": 1.8231665524331733e-07,
+      "loss": 2.0364,
+      "step": 1004
+    },
+    {
+      "epoch": 1.95954179868389,
+      "grad_norm": 0.4434170424938202,
+      "learning_rate": 1.8197395476353665e-07,
+      "loss": 1.9835,
+      "step": 1005
+    },
+    {
+      "epoch": 1.961491591518401,
+      "grad_norm": 0.45063334703445435,
+      "learning_rate": 1.81631254283756e-07,
+      "loss": 1.9904,
+      "step": 1006
+    },
+    {
+      "epoch": 1.9634413843529126,
+      "grad_norm": 0.45276153087615967,
+      "learning_rate": 1.8128855380397532e-07,
+      "loss": 2.021,
+      "step": 1007
+    },
+    {
+      "epoch": 1.9653911771874237,
+      "grad_norm": 0.44774502515792847,
+      "learning_rate": 1.8094585332419464e-07,
+      "loss": 2.0024,
+      "step": 1008
+    },
+    {
+      "epoch": 1.9673409700219353,
+      "grad_norm": 0.43734362721443176,
+      "learning_rate": 1.8060315284441396e-07,
+      "loss": 2.0261,
+      "step": 1009
+    },
+    {
+      "epoch": 1.9692907628564464,
+      "grad_norm": 0.45293501019477844,
+      "learning_rate": 1.802604523646333e-07,
+      "loss": 2.0781,
+      "step": 1010
+    },
+    {
+      "epoch": 1.9712405556909578,
+      "grad_norm": 0.4538004994392395,
+      "learning_rate": 1.7991775188485263e-07,
+      "loss": 2.0081,
+      "step": 1011
+    },
+    {
+      "epoch": 1.9731903485254692,
+      "grad_norm": 0.45042964816093445,
+      "learning_rate": 1.7957505140507195e-07,
+      "loss": 2.0121,
+      "step": 1012
+    },
+    {
+      "epoch": 1.9751401413599805,
+      "grad_norm": 0.4721399247646332,
+      "learning_rate": 1.792323509252913e-07,
+      "loss": 2.0071,
+      "step": 1013
+    },
+    {
+      "epoch": 1.9770899341944919,
+      "grad_norm": 0.4297287166118622,
+      "learning_rate": 1.788896504455106e-07,
+      "loss": 2.0213,
+      "step": 1014
+    },
+    {
+      "epoch": 1.9790397270290032,
+      "grad_norm": 0.4454828202724457,
+      "learning_rate": 1.7854694996572994e-07,
+      "loss": 2.0093,
+      "step": 1015
+    },
+    {
+      "epoch": 1.9809895198635146,
+      "grad_norm": 0.4550788700580597,
+      "learning_rate": 1.782042494859493e-07,
+      "loss": 2.0599,
+      "step": 1016
+    },
+    {
+      "epoch": 1.9829393126980257,
+      "grad_norm": 0.44854849576950073,
+      "learning_rate": 1.7786154900616858e-07,
+      "loss": 2.0262,
+      "step": 1017
+    },
+    {
+      "epoch": 1.9848891055325373,
+      "grad_norm": 0.4477459192276001,
+      "learning_rate": 1.7751884852638793e-07,
+      "loss": 1.9533,
+      "step": 1018
+    },
+    {
+      "epoch": 1.9868388983670484,
+      "grad_norm": 0.43663471937179565,
+      "learning_rate": 1.7717614804660728e-07,
+      "loss": 2.0122,
+      "step": 1019
+    },
+    {
+      "epoch": 1.98878869120156,
+      "grad_norm": 0.45281800627708435,
+      "learning_rate": 1.7683344756682657e-07,
+      "loss": 2.0711,
+      "step": 1020
+    },
+    {
+      "epoch": 1.9907384840360711,
+      "grad_norm": 0.44143861532211304,
+      "learning_rate": 1.7649074708704592e-07,
+      "loss": 2.0198,
+      "step": 1021
+    },
+    {
+      "epoch": 1.9926882768705825,
+      "grad_norm": 0.4464763402938843,
+      "learning_rate": 1.7614804660726524e-07,
+      "loss": 2.0117,
+      "step": 1022
+    },
+    {
+      "epoch": 1.9946380697050938,
+      "grad_norm": 0.42707762122154236,
+      "learning_rate": 1.7580534612748456e-07,
+      "loss": 1.9629,
+      "step": 1023
+    },
+    {
+      "epoch": 1.9965878625396052,
+      "grad_norm": 0.4683617949485779,
+      "learning_rate": 1.754626456477039e-07,
+      "loss": 2.0467,
+      "step": 1024
+    },
+    {
+      "epoch": 1.9985376553741165,
+      "grad_norm": 0.4215565025806427,
+      "learning_rate": 1.7511994516792323e-07,
+      "loss": 1.9545,
+      "step": 1025
+    },
+    {
+      "epoch": 1.9985376553741165,
+      "eval_loss": 2.0196783542633057,
+      "eval_runtime": 480.5583,
+      "eval_samples_per_second": 1.294,
+      "eval_steps_per_second": 0.325,
+      "step": 1025
+    },
+    {
+      "epoch": 2.0004874482086277,
+      "grad_norm": 0.42005443572998047,
+      "learning_rate": 1.7477724468814255e-07,
+      "loss": 2.0291,
+      "step": 1026
+    },
+    {
+      "epoch": 2.0024372410431392,
+      "grad_norm": 0.44807538390159607,
+      "learning_rate": 1.7443454420836187e-07,
+      "loss": 1.9539,
+      "step": 1027
+    },
+    {
+      "epoch": 2.0043870338776504,
+      "grad_norm": 0.47760045528411865,
+      "learning_rate": 1.7409184372858122e-07,
+      "loss": 1.9968,
+      "step": 1028
+    },
+    {
+      "epoch": 2.006336826712162,
+      "grad_norm": 0.42226237058639526,
+      "learning_rate": 1.7374914324880054e-07,
+      "loss": 1.9312,
+      "step": 1029
+    },
+    {
+      "epoch": 2.008286619546673,
+      "grad_norm": 0.44275936484336853,
+      "learning_rate": 1.7340644276901986e-07,
+      "loss": 2.0539,
+      "step": 1030
+    },
+    {
+      "epoch": 2.0102364123811847,
+      "grad_norm": 0.4601239264011383,
+      "learning_rate": 1.730637422892392e-07,
+      "loss": 2.028,
+      "step": 1031
+    },
+    {
+      "epoch": 2.012186205215696,
+      "grad_norm": 0.46528592705726624,
+      "learning_rate": 1.727210418094585e-07,
+      "loss": 2.0602,
+      "step": 1032
+    },
+    {
+      "epoch": 2.0141359980502074,
+      "grad_norm": 0.45640939474105835,
+      "learning_rate": 1.7237834132967785e-07,
+      "loss": 1.9717,
+      "step": 1033
+    },
+    {
+      "epoch": 2.0160857908847185,
+      "grad_norm": 0.4541582763195038,
+      "learning_rate": 1.720356408498972e-07,
+      "loss": 1.9875,
+      "step": 1034
+    },
+    {
+      "epoch": 2.0180355837192296,
+      "grad_norm": 0.48910263180732727,
+      "learning_rate": 1.716929403701165e-07,
+      "loss": 2.0361,
+      "step": 1035
+    },
+    {
+      "epoch": 2.019985376553741,
+      "grad_norm": 0.4677620232105255,
+      "learning_rate": 1.7135023989033584e-07,
+      "loss": 2.0022,
+      "step": 1036
+    },
+    {
+      "epoch": 2.0219351693882524,
+      "grad_norm": 0.4541827440261841,
+      "learning_rate": 1.710075394105552e-07,
+      "loss": 1.978,
+      "step": 1037
+    },
+    {
+      "epoch": 2.023884962222764,
+      "grad_norm": 0.49816232919692993,
+      "learning_rate": 1.7066483893077448e-07,
+      "loss": 2.0942,
+      "step": 1038
+    },
+    {
+      "epoch": 2.025834755057275,
+      "grad_norm": 0.46215084195137024,
+      "learning_rate": 1.7032213845099383e-07,
+      "loss": 1.9678,
+      "step": 1039
+    },
+    {
+      "epoch": 2.0277845478917866,
+      "grad_norm": 0.45608967542648315,
+      "learning_rate": 1.6997943797121315e-07,
+      "loss": 2.0144,
+      "step": 1040
+    },
+    {
+      "epoch": 2.0297343407262978,
+      "grad_norm": 0.43837353587150574,
+      "learning_rate": 1.6963673749143247e-07,
+      "loss": 2.0576,
+      "step": 1041
+    },
+    {
+      "epoch": 2.0316841335608093,
+      "grad_norm": 0.43267446756362915,
+      "learning_rate": 1.6929403701165182e-07,
+      "loss": 2.0134,
+      "step": 1042
+    },
+    {
+      "epoch": 2.0336339263953205,
+      "grad_norm": 0.4117604196071625,
+      "learning_rate": 1.6895133653187114e-07,
+      "loss": 1.9955,
+      "step": 1043
+    },
+    {
+      "epoch": 2.0355837192298316,
+      "grad_norm": 0.43197301030158997,
+      "learning_rate": 1.6860863605209046e-07,
+      "loss": 2.0108,
+      "step": 1044
+    },
+    {
+      "epoch": 2.037533512064343,
+      "grad_norm": 0.4347788095474243,
+      "learning_rate": 1.6826593557230978e-07,
+      "loss": 1.9693,
+      "step": 1045
+    },
+    {
+      "epoch": 2.0394833048988543,
+      "grad_norm": 0.4551312029361725,
+      "learning_rate": 1.6792323509252913e-07,
+      "loss": 1.9717,
+      "step": 1046
+    },
+    {
+      "epoch": 2.041433097733366,
+      "grad_norm": 0.43120816349983215,
+      "learning_rate": 1.6758053461274845e-07,
+      "loss": 1.9727,
+      "step": 1047
+    },
+    {
+      "epoch": 2.043382890567877,
+      "grad_norm": 0.4371218979358673,
+      "learning_rate": 1.6723783413296777e-07,
+      "loss": 2.0241,
+      "step": 1048
+    },
+    {
+      "epoch": 2.0453326834023886,
+      "grad_norm": 0.43102675676345825,
+      "learning_rate": 1.6689513365318712e-07,
+      "loss": 1.9871,
+      "step": 1049
+    },
+    {
+      "epoch": 2.0472824762368997,
+      "grad_norm": 0.4514808654785156,
+      "learning_rate": 1.6655243317340644e-07,
+      "loss": 1.9983,
+      "step": 1050
+    },
+    {
+      "epoch": 2.0492322690714113,
+      "grad_norm": 0.41947314143180847,
+      "learning_rate": 1.6620973269362576e-07,
+      "loss": 1.9818,
+      "step": 1051
+    },
+    {
+      "epoch": 2.0511820619059224,
+      "grad_norm": 0.4618643522262573,
+      "learning_rate": 1.658670322138451e-07,
+      "loss": 2.0139,
+      "step": 1052
+    },
+    {
+      "epoch": 2.053131854740434,
+      "grad_norm": 0.4574257433414459,
+      "learning_rate": 1.655243317340644e-07,
+      "loss": 1.9757,
+      "step": 1053
+    },
+    {
+      "epoch": 2.055081647574945,
+      "grad_norm": 0.41532963514328003,
+      "learning_rate": 1.6518163125428375e-07,
+      "loss": 1.9949,
+      "step": 1054
+    },
+    {
+      "epoch": 2.0570314404094563,
+      "grad_norm": 0.4592721462249756,
+      "learning_rate": 1.648389307745031e-07,
+      "loss": 2.0015,
+      "step": 1055
+    },
+    {
+      "epoch": 2.058981233243968,
+      "grad_norm": 0.4672160744667053,
+      "learning_rate": 1.644962302947224e-07,
+      "loss": 2.109,
+      "step": 1056
+    },
+    {
+      "epoch": 2.060931026078479,
+      "grad_norm": 0.439223051071167,
+      "learning_rate": 1.6415352981494174e-07,
+      "loss": 2.0451,
+      "step": 1057
+    },
+    {
+      "epoch": 2.0628808189129906,
+      "grad_norm": 0.442984014749527,
+      "learning_rate": 1.6381082933516104e-07,
+      "loss": 2.0049,
+      "step": 1058
+    },
+    {
+      "epoch": 2.0648306117475017,
+      "grad_norm": 0.4305538535118103,
+      "learning_rate": 1.6346812885538038e-07,
+      "loss": 2.0219,
+      "step": 1059
+    },
+    {
+      "epoch": 2.0667804045820133,
+      "grad_norm": 0.41040942072868347,
+      "learning_rate": 1.6312542837559973e-07,
+      "loss": 1.9563,
+      "step": 1060
+    },
+    {
+      "epoch": 2.0687301974165244,
+      "grad_norm": 0.43184876441955566,
+      "learning_rate": 1.6278272789581903e-07,
+      "loss": 2.0343,
+      "step": 1061
+    },
+    {
+      "epoch": 2.070679990251036,
+      "grad_norm": 0.45109209418296814,
+      "learning_rate": 1.6244002741603837e-07,
+      "loss": 2.0315,
+      "step": 1062
+    },
+    {
+      "epoch": 2.072629783085547,
+      "grad_norm": 0.437923789024353,
+      "learning_rate": 1.6209732693625772e-07,
+      "loss": 2.0265,
+      "step": 1063
+    },
+    {
+      "epoch": 2.0745795759200587,
+      "grad_norm": 0.4145483374595642,
+      "learning_rate": 1.6175462645647702e-07,
+      "loss": 2.0416,
+      "step": 1064
+    },
+    {
+      "epoch": 2.07652936875457,
+      "grad_norm": 0.44786587357521057,
+      "learning_rate": 1.6141192597669636e-07,
+      "loss": 1.9311,
+      "step": 1065
+    },
+    {
+      "epoch": 2.078479161589081,
+      "grad_norm": 0.44810619950294495,
+      "learning_rate": 1.6106922549691568e-07,
+      "loss": 1.9878,
+      "step": 1066
+    },
+    {
+      "epoch": 2.0804289544235925,
+      "grad_norm": 0.4340622127056122,
+      "learning_rate": 1.60726525017135e-07,
+      "loss": 1.9779,
+      "step": 1067
+    },
+    {
+      "epoch": 2.0823787472581037,
+      "grad_norm": 0.4348512887954712,
+      "learning_rate": 1.6038382453735435e-07,
+      "loss": 1.9908,
+      "step": 1068
+    },
+    {
+      "epoch": 2.0843285400926153,
+      "grad_norm": 0.4324460029602051,
+      "learning_rate": 1.6004112405757367e-07,
+      "loss": 1.9812,
+      "step": 1069
+    },
+    {
+      "epoch": 2.0862783329271264,
+      "grad_norm": 0.44518154859542847,
+      "learning_rate": 1.59698423577793e-07,
+      "loss": 2.0472,
+      "step": 1070
+    },
+    {
+      "epoch": 2.088228125761638,
+      "grad_norm": 0.449823796749115,
+      "learning_rate": 1.5935572309801232e-07,
+      "loss": 2.0346,
+      "step": 1071
+    },
+    {
+      "epoch": 2.090177918596149,
+      "grad_norm": 0.44722941517829895,
+      "learning_rate": 1.5901302261823166e-07,
+      "loss": 1.9449,
+      "step": 1072
+    },
+    {
+      "epoch": 2.0921277114306607,
+      "grad_norm": 0.42540016770362854,
+      "learning_rate": 1.5867032213845098e-07,
+      "loss": 2.0294,
+      "step": 1073
+    },
+    {
+      "epoch": 2.094077504265172,
+      "grad_norm": 0.42311087250709534,
+      "learning_rate": 1.583276216586703e-07,
+      "loss": 2.0389,
+      "step": 1074
+    },
+    {
+      "epoch": 2.096027297099683,
+      "grad_norm": 0.44461527466773987,
+      "learning_rate": 1.5798492117888965e-07,
+      "loss": 2.0042,
+      "step": 1075
+    },
+    {
+      "epoch": 2.0979770899341945,
+      "grad_norm": 0.469026118516922,
+      "learning_rate": 1.5764222069910897e-07,
+      "loss": 2.0641,
+      "step": 1076
+    },
+    {
+      "epoch": 2.0999268827687056,
+      "grad_norm": 0.42003822326660156,
+      "learning_rate": 1.572995202193283e-07,
+      "loss": 2.0329,
+      "step": 1077
+    },
+    {
+      "epoch": 2.1018766756032172,
+      "grad_norm": 0.4529522657394409,
+      "learning_rate": 1.5695681973954764e-07,
+      "loss": 2.0006,
+      "step": 1078
+    },
+    {
+      "epoch": 2.1038264684377284,
+      "grad_norm": 0.44563665986061096,
+      "learning_rate": 1.5661411925976694e-07,
+      "loss": 2.0605,
+      "step": 1079
+    },
+    {
+      "epoch": 2.10577626127224,
+      "grad_norm": 0.42427098751068115,
+      "learning_rate": 1.5627141877998628e-07,
+      "loss": 1.9327,
+      "step": 1080
+    },
+    {
+      "epoch": 2.107726054106751,
+      "grad_norm": 0.438480406999588,
+      "learning_rate": 1.5592871830020563e-07,
+      "loss": 2.0477,
+      "step": 1081
+    },
+    {
+      "epoch": 2.1096758469412626,
+      "grad_norm": 0.459905207157135,
+      "learning_rate": 1.5558601782042493e-07,
+      "loss": 2.0103,
+      "step": 1082
+    },
+    {
+      "epoch": 2.1116256397757738,
+      "grad_norm": 0.4529953896999359,
+      "learning_rate": 1.5524331734064427e-07,
+      "loss": 2.045,
+      "step": 1083
+    },
+    {
+      "epoch": 2.1135754326102854,
+      "grad_norm": 0.4429239332675934,
+      "learning_rate": 1.549006168608636e-07,
+      "loss": 1.9767,
+      "step": 1084
+    },
+    {
+      "epoch": 2.1155252254447965,
+      "grad_norm": 0.4611859619617462,
+      "learning_rate": 1.5455791638108292e-07,
+      "loss": 2.004,
+      "step": 1085
+    },
+    {
+      "epoch": 2.1174750182793076,
+      "grad_norm": 0.4560939371585846,
+      "learning_rate": 1.5421521590130226e-07,
+      "loss": 2.0251,
+      "step": 1086
+    },
+    {
+      "epoch": 2.119424811113819,
+      "grad_norm": 0.46503308415412903,
+      "learning_rate": 1.5387251542152158e-07,
+      "loss": 1.9608,
+      "step": 1087
+    },
+    {
+      "epoch": 2.1213746039483303,
+      "grad_norm": 0.45832183957099915,
+      "learning_rate": 1.535298149417409e-07,
+      "loss": 2.0145,
+      "step": 1088
+    },
+    {
+      "epoch": 2.123324396782842,
+      "grad_norm": 0.4266718626022339,
+      "learning_rate": 1.5318711446196023e-07,
+      "loss": 1.9256,
+      "step": 1089
+    },
+    {
+      "epoch": 2.125274189617353,
+      "grad_norm": 0.43129250407218933,
+      "learning_rate": 1.5284441398217957e-07,
+      "loss": 2.0501,
+      "step": 1090
+    },
+    {
+      "epoch": 2.1272239824518646,
+      "grad_norm": 0.4121784567832947,
+      "learning_rate": 1.525017135023989e-07,
+      "loss": 1.9625,
+      "step": 1091
+    },
+    {
+      "epoch": 2.1291737752863757,
+      "grad_norm": 0.4729436933994293,
+      "learning_rate": 1.5215901302261822e-07,
+      "loss": 2.0381,
+      "step": 1092
+    },
+    {
+      "epoch": 2.1311235681208873,
+      "grad_norm": 0.46315380930900574,
+      "learning_rate": 1.5181631254283756e-07,
+      "loss": 2.0422,
+      "step": 1093
+    },
+    {
+      "epoch": 2.1330733609553985,
+      "grad_norm": 0.4314703047275543,
+      "learning_rate": 1.5147361206305688e-07,
+      "loss": 2.0789,
+      "step": 1094
+    },
+    {
+      "epoch": 2.13502315378991,
+      "grad_norm": 0.4514041543006897,
+      "learning_rate": 1.511309115832762e-07,
+      "loss": 2.012,
+      "step": 1095
+    },
+    {
+      "epoch": 2.136972946624421,
+      "grad_norm": 0.4410359561443329,
+      "learning_rate": 1.5078821110349555e-07,
+      "loss": 2.0134,
+      "step": 1096
+    },
+    {
+      "epoch": 2.1389227394589323,
+      "grad_norm": 0.4480474293231964,
+      "learning_rate": 1.5044551062371485e-07,
+      "loss": 2.0422,
+      "step": 1097
+    },
+    {
+      "epoch": 2.140872532293444,
+      "grad_norm": 0.46355828642845154,
+      "learning_rate": 1.501028101439342e-07,
+      "loss": 2.0701,
+      "step": 1098
+    },
+    {
+      "epoch": 2.142822325127955,
+      "grad_norm": 0.4337591230869293,
+      "learning_rate": 1.4976010966415354e-07,
+      "loss": 2.0295,
+      "step": 1099
+    },
+    {
+      "epoch": 2.1447721179624666,
+      "grad_norm": 0.4611411988735199,
+      "learning_rate": 1.4941740918437284e-07,
+      "loss": 2.044,
+      "step": 1100
+    },
+    {
+      "epoch": 2.1467219107969777,
+      "grad_norm": 0.45693400502204895,
+      "learning_rate": 1.4907470870459218e-07,
+      "loss": 1.9968,
+      "step": 1101
+    },
+    {
+      "epoch": 2.1486717036314893,
+      "grad_norm": 0.4180225729942322,
+      "learning_rate": 1.487320082248115e-07,
+      "loss": 1.9708,
+      "step": 1102
+    },
+    {
+      "epoch": 2.1506214964660004,
+      "grad_norm": 0.4424188435077667,
+      "learning_rate": 1.4838930774503083e-07,
+      "loss": 1.987,
+      "step": 1103
+    },
+    {
+      "epoch": 2.152571289300512,
+      "grad_norm": 0.4601939916610718,
+      "learning_rate": 1.4804660726525017e-07,
+      "loss": 2.0478,
+      "step": 1104
+    },
+    {
+      "epoch": 2.154521082135023,
+      "grad_norm": 0.4408921003341675,
+      "learning_rate": 1.477039067854695e-07,
+      "loss": 1.9208,
+      "step": 1105
+    },
+    {
+      "epoch": 2.1564708749695347,
+      "grad_norm": 0.4693887233734131,
+      "learning_rate": 1.4736120630568882e-07,
+      "loss": 2.0303,
+      "step": 1106
+    },
+    {
+      "epoch": 2.158420667804046,
+      "grad_norm": 0.4544321894645691,
+      "learning_rate": 1.4701850582590816e-07,
+      "loss": 2.0507,
+      "step": 1107
+    },
+    {
+      "epoch": 2.160370460638557,
+      "grad_norm": 0.4257730543613434,
+      "learning_rate": 1.4667580534612748e-07,
+      "loss": 1.9841,
+      "step": 1108
+    },
+    {
+      "epoch": 2.1623202534730686,
+      "grad_norm": 0.4435708522796631,
+      "learning_rate": 1.463331048663468e-07,
+      "loss": 1.9875,
+      "step": 1109
+    },
+    {
+      "epoch": 2.1642700463075797,
+      "grad_norm": 0.45505550503730774,
+      "learning_rate": 1.4599040438656613e-07,
+      "loss": 2.0299,
+      "step": 1110
+    },
+    {
+      "epoch": 2.1662198391420913,
+      "grad_norm": 0.42817404866218567,
+      "learning_rate": 1.4564770390678547e-07,
+      "loss": 2.0039,
+      "step": 1111
+    },
+    {
+      "epoch": 2.1681696319766024,
+      "grad_norm": 0.46066412329673767,
+      "learning_rate": 1.453050034270048e-07,
+      "loss": 2.0501,
+      "step": 1112
+    },
+    {
+      "epoch": 2.170119424811114,
+      "grad_norm": 0.43956106901168823,
+      "learning_rate": 1.4496230294722412e-07,
+      "loss": 2.0459,
+      "step": 1113
+    },
+    {
+      "epoch": 2.172069217645625,
+      "grad_norm": 0.4616369903087616,
+      "learning_rate": 1.4461960246744346e-07,
+      "loss": 2.0696,
+      "step": 1114
+    },
+    {
+      "epoch": 2.1740190104801367,
+      "grad_norm": 0.4470338523387909,
+      "learning_rate": 1.4427690198766276e-07,
+      "loss": 1.973,
+      "step": 1115
+    },
+    {
+      "epoch": 2.175968803314648,
+      "grad_norm": 0.46773573756217957,
+      "learning_rate": 1.439342015078821e-07,
+      "loss": 1.9794,
+      "step": 1116
+    },
+    {
+      "epoch": 2.177918596149159,
+      "grad_norm": 0.4283302128314972,
+      "learning_rate": 1.4359150102810145e-07,
+      "loss": 2.0106,
+      "step": 1117
+    },
+    {
+      "epoch": 2.1798683889836705,
+      "grad_norm": 0.4649440348148346,
+      "learning_rate": 1.4324880054832075e-07,
+      "loss": 2.0682,
+      "step": 1118
+    },
+    {
+      "epoch": 2.1818181818181817,
+      "grad_norm": 0.4346439242362976,
+      "learning_rate": 1.429061000685401e-07,
+      "loss": 2.0001,
+      "step": 1119
+    },
+    {
+      "epoch": 2.1837679746526932,
+      "grad_norm": 0.4359522759914398,
+      "learning_rate": 1.4256339958875944e-07,
+      "loss": 2.0299,
+      "step": 1120
+    },
+    {
+      "epoch": 2.1857177674872044,
+      "grad_norm": 0.4563926160335541,
+      "learning_rate": 1.4222069910897874e-07,
+      "loss": 1.9896,
+      "step": 1121
+    },
+    {
+      "epoch": 2.187667560321716,
+      "grad_norm": 0.4313761293888092,
+      "learning_rate": 1.4187799862919808e-07,
+      "loss": 1.9882,
+      "step": 1122
+    },
+    {
+      "epoch": 2.189617353156227,
+      "grad_norm": 0.44476938247680664,
+      "learning_rate": 1.415352981494174e-07,
+      "loss": 1.9293,
+      "step": 1123
+    },
+    {
+      "epoch": 2.1915671459907387,
+      "grad_norm": 0.4599853754043579,
+      "learning_rate": 1.4119259766963673e-07,
+      "loss": 1.9964,
+      "step": 1124
+    },
+    {
+      "epoch": 2.19351693882525,
+      "grad_norm": 0.4537375271320343,
+      "learning_rate": 1.4084989718985607e-07,
+      "loss": 2.0956,
+      "step": 1125
+    },
+    {
+      "epoch": 2.195466731659761,
+      "grad_norm": 0.4554738998413086,
+      "learning_rate": 1.405071967100754e-07,
+      "loss": 2.0423,
+      "step": 1126
+    },
+    {
+      "epoch": 2.1974165244942725,
+      "grad_norm": 0.43387341499328613,
+      "learning_rate": 1.4016449623029472e-07,
+      "loss": 1.998,
+      "step": 1127
+    },
+    {
+      "epoch": 2.1993663173287836,
+      "grad_norm": 0.4407186806201935,
+      "learning_rate": 1.3982179575051404e-07,
+      "loss": 2.0791,
+      "step": 1128
+    },
+    {
+      "epoch": 2.201316110163295,
+      "grad_norm": 0.44602102041244507,
+      "learning_rate": 1.3947909527073338e-07,
+      "loss": 2.0093,
+      "step": 1129
+    },
+    {
+      "epoch": 2.2032659029978063,
+      "grad_norm": 0.42528602480888367,
+      "learning_rate": 1.391363947909527e-07,
+      "loss": 1.992,
+      "step": 1130
+    },
+    {
+      "epoch": 2.205215695832318,
+      "grad_norm": 0.39986541867256165,
+      "learning_rate": 1.3879369431117203e-07,
+      "loss": 1.9602,
+      "step": 1131
+    },
+    {
+      "epoch": 2.207165488666829,
+      "grad_norm": 0.4568888545036316,
+      "learning_rate": 1.3845099383139137e-07,
+      "loss": 2.0055,
+      "step": 1132
+    },
+    {
+      "epoch": 2.2091152815013406,
+      "grad_norm": 0.45194804668426514,
+      "learning_rate": 1.381082933516107e-07,
+      "loss": 2.0293,
+      "step": 1133
+    },
+    {
+      "epoch": 2.2110650743358518,
+      "grad_norm": 0.43082454800605774,
+      "learning_rate": 1.3776559287183002e-07,
+      "loss": 2.0219,
+      "step": 1134
+    },
+    {
+      "epoch": 2.2130148671703633,
+      "grad_norm": 0.43497541546821594,
+      "learning_rate": 1.3742289239204936e-07,
+      "loss": 2.0042,
+      "step": 1135
+    },
+    {
+      "epoch": 2.2149646600048745,
+      "grad_norm": 0.4378391206264496,
+      "learning_rate": 1.3708019191226866e-07,
+      "loss": 1.9613,
+      "step": 1136
+    },
+    {
+      "epoch": 2.2169144528393856,
+      "grad_norm": 0.4528283476829529,
+      "learning_rate": 1.36737491432488e-07,
+      "loss": 2.0555,
+      "step": 1137
+    },
+    {
+      "epoch": 2.218864245673897,
+      "grad_norm": 0.440682590007782,
+      "learning_rate": 1.3639479095270735e-07,
+      "loss": 2.0322,
+      "step": 1138
+    },
+    {
+      "epoch": 2.2208140385084083,
+      "grad_norm": 0.45496833324432373,
+      "learning_rate": 1.3605209047292665e-07,
+      "loss": 2.0366,
+      "step": 1139
+    },
+    {
+      "epoch": 2.22276383134292,
+      "grad_norm": 0.41444405913352966,
+      "learning_rate": 1.35709389993146e-07,
+      "loss": 1.9774,
+      "step": 1140
+    },
+    {
+      "epoch": 2.224713624177431,
+      "grad_norm": 0.4410904049873352,
+      "learning_rate": 1.353666895133653e-07,
+      "loss": 2.0427,
+      "step": 1141
+    },
+    {
+      "epoch": 2.2266634170119426,
+      "grad_norm": 0.44298315048217773,
+      "learning_rate": 1.3502398903358464e-07,
+      "loss": 1.976,
+      "step": 1142
+    },
+    {
+      "epoch": 2.2286132098464537,
+      "grad_norm": 0.45140451192855835,
+      "learning_rate": 1.3468128855380398e-07,
+      "loss": 2.0178,
+      "step": 1143
+    },
+    {
+      "epoch": 2.2305630026809653,
+      "grad_norm": 0.4413256049156189,
+      "learning_rate": 1.3433858807402328e-07,
+      "loss": 2.0457,
+      "step": 1144
+    },
+    {
+      "epoch": 2.2325127955154764,
+      "grad_norm": 0.4329676628112793,
+      "learning_rate": 1.3399588759424263e-07,
+      "loss": 1.993,
+      "step": 1145
+    },
+    {
+      "epoch": 2.234462588349988,
+      "grad_norm": 0.4443471133708954,
+      "learning_rate": 1.3365318711446195e-07,
+      "loss": 1.9932,
+      "step": 1146
+    },
+    {
+      "epoch": 2.236412381184499,
+      "grad_norm": 0.43921956419944763,
+      "learning_rate": 1.3331048663468127e-07,
+      "loss": 1.9404,
+      "step": 1147
+    },
+    {
+      "epoch": 2.2383621740190103,
+      "grad_norm": 0.4288458526134491,
+      "learning_rate": 1.3296778615490062e-07,
+      "loss": 2.0023,
+      "step": 1148
+    },
+    {
+      "epoch": 2.240311966853522,
+      "grad_norm": 0.48166126012802124,
+      "learning_rate": 1.3262508567511994e-07,
+      "loss": 2.0247,
+      "step": 1149
+    },
+    {
+      "epoch": 2.242261759688033,
+      "grad_norm": 0.42935001850128174,
+      "learning_rate": 1.3228238519533926e-07,
+      "loss": 2.0893,
+      "step": 1150
+    },
+    {
+      "epoch": 2.2442115525225446,
+      "grad_norm": 0.4443514347076416,
+      "learning_rate": 1.319396847155586e-07,
+      "loss": 1.9194,
+      "step": 1151
+    },
+    {
+      "epoch": 2.2461613453570557,
+      "grad_norm": 0.4799755811691284,
+      "learning_rate": 1.3159698423577793e-07,
+      "loss": 2.002,
+      "step": 1152
+    },
+    {
+      "epoch": 2.2481111381915673,
+      "grad_norm": 0.42361047863960266,
+      "learning_rate": 1.3125428375599725e-07,
+      "loss": 2.0243,
+      "step": 1153
+    },
+    {
+      "epoch": 2.2500609310260784,
+      "grad_norm": 0.44116365909576416,
+      "learning_rate": 1.3091158327621657e-07,
+      "loss": 1.93,
+      "step": 1154
+    },
+    {
+      "epoch": 2.25201072386059,
+      "grad_norm": 0.4595193862915039,
+      "learning_rate": 1.3056888279643592e-07,
+      "loss": 2.0667,
+      "step": 1155
+    },
+    {
+      "epoch": 2.253960516695101,
+      "grad_norm": 0.43922775983810425,
+      "learning_rate": 1.3022618231665524e-07,
+      "loss": 2.0263,
+      "step": 1156
+    },
+    {
+      "epoch": 2.2559103095296127,
+      "grad_norm": 0.4693455100059509,
+      "learning_rate": 1.2988348183687456e-07,
+      "loss": 2.0534,
+      "step": 1157
+    },
+    {
+      "epoch": 2.257860102364124,
+      "grad_norm": 0.4336584210395813,
+      "learning_rate": 1.295407813570939e-07,
+      "loss": 2.0013,
+      "step": 1158
+    },
+    {
+      "epoch": 2.259809895198635,
+      "grad_norm": 0.45184361934661865,
+      "learning_rate": 1.291980808773132e-07,
+      "loss": 1.9839,
+      "step": 1159
+    },
+    {
+      "epoch": 2.2617596880331465,
+      "grad_norm": 0.45428767800331116,
+      "learning_rate": 1.2885538039753255e-07,
+      "loss": 1.978,
+      "step": 1160
+    },
+    {
+      "epoch": 2.2637094808676577,
+      "grad_norm": 0.43744540214538574,
+      "learning_rate": 1.285126799177519e-07,
+      "loss": 2.0031,
+      "step": 1161
+    },
+    {
+      "epoch": 2.2656592737021692,
+      "grad_norm": 0.4603250026702881,
+      "learning_rate": 1.281699794379712e-07,
+      "loss": 2.0069,
+      "step": 1162
+    },
+    {
+      "epoch": 2.2676090665366804,
+      "grad_norm": 0.41848501563072205,
+      "learning_rate": 1.2782727895819054e-07,
+      "loss": 1.9591,
+      "step": 1163
+    },
+    {
+      "epoch": 2.269558859371192,
+      "grad_norm": 0.4679056704044342,
+      "learning_rate": 1.2748457847840988e-07,
+      "loss": 1.9896,
+      "step": 1164
+    },
+    {
+      "epoch": 2.271508652205703,
+      "grad_norm": 0.4484792947769165,
+      "learning_rate": 1.2714187799862918e-07,
+      "loss": 1.981,
+      "step": 1165
+    },
+    {
+      "epoch": 2.2734584450402147,
+      "grad_norm": 0.4249541759490967,
+      "learning_rate": 1.2679917751884853e-07,
+      "loss": 2.0056,
+      "step": 1166
+    },
+    {
+      "epoch": 2.275408237874726,
+      "grad_norm": 0.44611814618110657,
+      "learning_rate": 1.2645647703906785e-07,
+      "loss": 1.9827,
+      "step": 1167
+    },
+    {
+      "epoch": 2.2773580307092374,
+      "grad_norm": 0.45555368065834045,
+      "learning_rate": 1.2611377655928717e-07,
+      "loss": 2.0409,
+      "step": 1168
+    },
+    {
+      "epoch": 2.2793078235437485,
+      "grad_norm": 0.43659961223602295,
+      "learning_rate": 1.2577107607950652e-07,
+      "loss": 2.0553,
+      "step": 1169
+    },
+    {
+      "epoch": 2.2812576163782596,
+      "grad_norm": 0.4498869478702545,
+      "learning_rate": 1.2542837559972584e-07,
+      "loss": 2.04,
+      "step": 1170
+    },
+    {
+      "epoch": 2.283207409212771,
+      "grad_norm": 0.4552645981311798,
+      "learning_rate": 1.2508567511994516e-07,
+      "loss": 2.0162,
+      "step": 1171
+    },
+    {
+      "epoch": 2.2851572020472823,
+      "grad_norm": 0.4378924071788788,
+      "learning_rate": 1.2474297464016448e-07,
+      "loss": 1.9875,
+      "step": 1172
+    },
+    {
+      "epoch": 2.287106994881794,
+      "grad_norm": 0.4232163429260254,
+      "learning_rate": 1.2440027416038383e-07,
+      "loss": 2.0426,
+      "step": 1173
+    },
+    {
+      "epoch": 2.289056787716305,
+      "grad_norm": 0.44864141941070557,
+      "learning_rate": 1.2405757368060315e-07,
+      "loss": 2.0094,
+      "step": 1174
+    },
+    {
+      "epoch": 2.2910065805508166,
+      "grad_norm": 0.45312970876693726,
+      "learning_rate": 1.2371487320082247e-07,
+      "loss": 2.0105,
+      "step": 1175
+    },
+    {
+      "epoch": 2.2929563733853278,
+      "grad_norm": 0.43205204606056213,
+      "learning_rate": 1.2337217272104182e-07,
+      "loss": 2.0116,
+      "step": 1176
+    },
+    {
+      "epoch": 2.294906166219839,
+      "grad_norm": 0.4289553761482239,
+      "learning_rate": 1.2302947224126114e-07,
+      "loss": 1.9694,
+      "step": 1177
+    },
+    {
+      "epoch": 2.2968559590543505,
+      "grad_norm": 0.48636192083358765,
+      "learning_rate": 1.2268677176148046e-07,
+      "loss": 2.0412,
+      "step": 1178
+    },
+    {
+      "epoch": 2.298805751888862,
+      "grad_norm": 0.453530877828598,
+      "learning_rate": 1.223440712816998e-07,
+      "loss": 1.9954,
+      "step": 1179
+    },
+    {
+      "epoch": 2.300755544723373,
+      "grad_norm": 0.47017619013786316,
+      "learning_rate": 1.2200137080191913e-07,
+      "loss": 1.9938,
+      "step": 1180
+    },
+    {
+      "epoch": 2.3027053375578843,
+      "grad_norm": 0.47395429015159607,
+      "learning_rate": 1.2165867032213845e-07,
+      "loss": 2.163,
+      "step": 1181
+    },
+    {
+      "epoch": 2.304655130392396,
+      "grad_norm": 0.460896760225296,
+      "learning_rate": 1.2131596984235777e-07,
+      "loss": 2.0602,
+      "step": 1182
+    },
+    {
+      "epoch": 2.306604923226907,
+      "grad_norm": 0.44715631008148193,
+      "learning_rate": 1.209732693625771e-07,
+      "loss": 1.9946,
+      "step": 1183
+    },
+    {
+      "epoch": 2.3085547160614186,
+      "grad_norm": 0.4538576900959015,
+      "learning_rate": 1.2063056888279644e-07,
+      "loss": 2.0119,
+      "step": 1184
+    },
+    {
+      "epoch": 2.3105045088959297,
+      "grad_norm": 0.45894429087638855,
+      "learning_rate": 1.2028786840301576e-07,
+      "loss": 2.0613,
+      "step": 1185
+    },
+    {
+      "epoch": 2.3124543017304413,
+      "grad_norm": 0.4307364523410797,
+      "learning_rate": 1.1994516792323508e-07,
+      "loss": 1.9579,
+      "step": 1186
+    },
+    {
+      "epoch": 2.3144040945649524,
+      "grad_norm": 0.455282598733902,
+      "learning_rate": 1.196024674434544e-07,
+      "loss": 2.057,
+      "step": 1187
+    },
+    {
+      "epoch": 2.3163538873994636,
+      "grad_norm": 0.41711223125457764,
+      "learning_rate": 1.1925976696367375e-07,
+      "loss": 2.0075,
+      "step": 1188
+    },
+    {
+      "epoch": 2.318303680233975,
+      "grad_norm": 0.44980695843696594,
+      "learning_rate": 1.1891706648389308e-07,
+      "loss": 2.0534,
+      "step": 1189
+    },
+    {
+      "epoch": 2.3202534730684863,
+      "grad_norm": 0.42400816082954407,
+      "learning_rate": 1.185743660041124e-07,
+      "loss": 1.9931,
+      "step": 1190
+    },
+    {
+      "epoch": 2.322203265902998,
+      "grad_norm": 0.45932522416114807,
+      "learning_rate": 1.1823166552433172e-07,
+      "loss": 2.0005,
+      "step": 1191
+    },
+    {
+      "epoch": 2.324153058737509,
+      "grad_norm": 0.4707227647304535,
+      "learning_rate": 1.1788896504455104e-07,
+      "loss": 2.0125,
+      "step": 1192
+    },
+    {
+      "epoch": 2.3261028515720206,
+      "grad_norm": 0.4686226546764374,
+      "learning_rate": 1.1754626456477039e-07,
+      "loss": 2.0177,
+      "step": 1193
+    },
+    {
+      "epoch": 2.3280526444065317,
+      "grad_norm": 0.45334458351135254,
+      "learning_rate": 1.1720356408498971e-07,
+      "loss": 2.0139,
+      "step": 1194
+    },
+    {
+      "epoch": 2.3300024372410433,
+      "grad_norm": 0.4258855879306793,
+      "learning_rate": 1.1686086360520903e-07,
+      "loss": 2.0112,
+      "step": 1195
+    },
+    {
+      "epoch": 2.3319522300755544,
+      "grad_norm": 0.4407455325126648,
+      "learning_rate": 1.1651816312542837e-07,
+      "loss": 2.0678,
+      "step": 1196
+    },
+    {
+      "epoch": 2.333902022910066,
+      "grad_norm": 0.44041523337364197,
+      "learning_rate": 1.161754626456477e-07,
+      "loss": 2.012,
+      "step": 1197
+    },
+    {
+      "epoch": 2.335851815744577,
+      "grad_norm": 0.44474172592163086,
+      "learning_rate": 1.1583276216586702e-07,
+      "loss": 2.0197,
+      "step": 1198
+    },
+    {
+      "epoch": 2.3378016085790883,
+      "grad_norm": 0.4264999032020569,
+      "learning_rate": 1.1549006168608636e-07,
+      "loss": 2.0351,
+      "step": 1199
+    },
+    {
+      "epoch": 2.3397514014136,
+      "grad_norm": 0.4538833498954773,
+      "learning_rate": 1.1514736120630568e-07,
+      "loss": 1.9862,
+      "step": 1200
+    },
+    {
+      "epoch": 2.341701194248111,
+      "grad_norm": 0.4511955678462982,
+      "learning_rate": 1.1480466072652501e-07,
+      "loss": 1.9563,
+      "step": 1201
+    },
+    {
+      "epoch": 2.3436509870826225,
+      "grad_norm": 0.4677155315876007,
+      "learning_rate": 1.1446196024674435e-07,
+      "loss": 2.0302,
+      "step": 1202
+    },
+    {
+      "epoch": 2.3456007799171337,
+      "grad_norm": 0.44334524869918823,
+      "learning_rate": 1.1411925976696367e-07,
+      "loss": 1.9415,
+      "step": 1203
+    },
+    {
+      "epoch": 2.3475505727516452,
+      "grad_norm": 0.43790337443351746,
+      "learning_rate": 1.1377655928718299e-07,
+      "loss": 1.9417,
+      "step": 1204
+    },
+    {
+      "epoch": 2.3495003655861564,
+      "grad_norm": 0.43173927068710327,
+      "learning_rate": 1.1343385880740232e-07,
+      "loss": 2.0638,
+      "step": 1205
+    },
+    {
+      "epoch": 2.351450158420668,
+      "grad_norm": 0.41969236731529236,
+      "learning_rate": 1.1309115832762166e-07,
+      "loss": 1.9791,
+      "step": 1206
+    },
+    {
+      "epoch": 2.353399951255179,
+      "grad_norm": 0.44412854313850403,
+      "learning_rate": 1.1274845784784098e-07,
+      "loss": 2.0437,
+      "step": 1207
+    },
+    {
+      "epoch": 2.3553497440896907,
+      "grad_norm": 0.4438328444957733,
+      "learning_rate": 1.1240575736806031e-07,
+      "loss": 1.9868,
+      "step": 1208
+    },
+    {
+      "epoch": 2.357299536924202,
+      "grad_norm": 0.47638362646102905,
+      "learning_rate": 1.1206305688827963e-07,
+      "loss": 2.0087,
+      "step": 1209
+    },
+    {
+      "epoch": 2.359249329758713,
+      "grad_norm": 0.4346645176410675,
+      "learning_rate": 1.1172035640849897e-07,
+      "loss": 2.0362,
+      "step": 1210
+    },
+    {
+      "epoch": 2.3611991225932245,
+      "grad_norm": 0.42794761061668396,
+      "learning_rate": 1.113776559287183e-07,
+      "loss": 2.0047,
+      "step": 1211
+    },
+    {
+      "epoch": 2.3631489154277356,
+      "grad_norm": 0.4326521158218384,
+      "learning_rate": 1.1103495544893762e-07,
+      "loss": 1.962,
+      "step": 1212
+    },
+    {
+      "epoch": 2.365098708262247,
+      "grad_norm": 0.4463393986225128,
+      "learning_rate": 1.1069225496915694e-07,
+      "loss": 2.0585,
+      "step": 1213
+    },
+    {
+      "epoch": 2.3670485010967584,
+      "grad_norm": 0.39195939898490906,
+      "learning_rate": 1.1034955448937629e-07,
+      "loss": 2.004,
+      "step": 1214
+    },
+    {
+      "epoch": 2.36899829393127,
+      "grad_norm": 0.4264962077140808,
+      "learning_rate": 1.1000685400959561e-07,
+      "loss": 1.9817,
+      "step": 1215
+    },
+    {
+      "epoch": 2.370948086765781,
+      "grad_norm": 0.43186262249946594,
+      "learning_rate": 1.0966415352981493e-07,
+      "loss": 1.9599,
+      "step": 1216
+    },
+    {
+      "epoch": 2.3728978796002926,
+      "grad_norm": 0.4679785668849945,
+      "learning_rate": 1.0932145305003427e-07,
+      "loss": 2.042,
+      "step": 1217
+    },
+    {
+      "epoch": 2.3748476724348038,
+      "grad_norm": 0.45915287733078003,
+      "learning_rate": 1.0897875257025359e-07,
+      "loss": 1.9846,
+      "step": 1218
+    },
+    {
+      "epoch": 2.3767974652693153,
+      "grad_norm": 0.42811834812164307,
+      "learning_rate": 1.0863605209047292e-07,
+      "loss": 2.0096,
+      "step": 1219
+    },
+    {
+      "epoch": 2.3787472581038265,
+      "grad_norm": 0.44724828004837036,
+      "learning_rate": 1.0829335161069226e-07,
+      "loss": 2.0716,
+      "step": 1220
+    },
+    {
+      "epoch": 2.3806970509383376,
+      "grad_norm": 0.46073630452156067,
+      "learning_rate": 1.0795065113091158e-07,
+      "loss": 2.0203,
+      "step": 1221
+    },
+    {
+      "epoch": 2.382646843772849,
+      "grad_norm": 0.47367486357688904,
+      "learning_rate": 1.076079506511309e-07,
+      "loss": 2.008,
+      "step": 1222
+    },
+    {
+      "epoch": 2.3845966366073603,
+      "grad_norm": 0.42962348461151123,
+      "learning_rate": 1.0726525017135025e-07,
+      "loss": 1.9825,
+      "step": 1223
+    },
+    {
+      "epoch": 2.386546429441872,
+      "grad_norm": 0.4696093797683716,
+      "learning_rate": 1.0692254969156957e-07,
+      "loss": 2.1004,
+      "step": 1224
+    },
+    {
+      "epoch": 2.388496222276383,
+      "grad_norm": 0.4257887303829193,
+      "learning_rate": 1.0657984921178889e-07,
+      "loss": 1.9573,
+      "step": 1225
+    },
+    {
+      "epoch": 2.3904460151108946,
+      "grad_norm": 0.467960387468338,
+      "learning_rate": 1.0623714873200821e-07,
+      "loss": 2.052,
+      "step": 1226
+    },
+    {
+      "epoch": 2.3923958079454057,
+      "grad_norm": 0.4748067557811737,
+      "learning_rate": 1.0589444825222754e-07,
+      "loss": 2.0189,
+      "step": 1227
+    },
+    {
+      "epoch": 2.3943456007799173,
+      "grad_norm": 0.4576222896575928,
+      "learning_rate": 1.0555174777244688e-07,
+      "loss": 1.9823,
+      "step": 1228
+    },
+    {
+      "epoch": 2.3962953936144284,
+      "grad_norm": 0.4228207468986511,
+      "learning_rate": 1.052090472926662e-07,
+      "loss": 2.0092,
+      "step": 1229
+    },
+    {
+      "epoch": 2.39824518644894,
+      "grad_norm": 0.45505014061927795,
+      "learning_rate": 1.0486634681288553e-07,
+      "loss": 2.001,
+      "step": 1230
+    },
+    {
+      "epoch": 2.400194979283451,
+      "grad_norm": 0.4578765630722046,
+      "learning_rate": 1.0452364633310485e-07,
+      "loss": 2.0533,
+      "step": 1231
+    },
+    {
+      "epoch": 2.4021447721179623,
+      "grad_norm": 0.4225928485393524,
+      "learning_rate": 1.0418094585332419e-07,
+      "loss": 2.0138,
+      "step": 1232
+    },
+    {
+      "epoch": 2.404094564952474,
+      "grad_norm": 0.4521108567714691,
+      "learning_rate": 1.0383824537354352e-07,
+      "loss": 2.0642,
+      "step": 1233
+    },
+    {
+      "epoch": 2.406044357786985,
+      "grad_norm": 0.46423792839050293,
+      "learning_rate": 1.0349554489376284e-07,
+      "loss": 1.9876,
+      "step": 1234
+    },
+    {
+      "epoch": 2.4079941506214966,
+      "grad_norm": 0.41854438185691833,
+      "learning_rate": 1.0315284441398217e-07,
+      "loss": 1.9795,
+      "step": 1235
+    },
+    {
+      "epoch": 2.4099439434560077,
+      "grad_norm": 0.4496399462223053,
+      "learning_rate": 1.0281014393420151e-07,
+      "loss": 2.0066,
+      "step": 1236
+    },
+    {
+      "epoch": 2.4118937362905193,
+      "grad_norm": 0.4492359757423401,
+      "learning_rate": 1.0246744345442083e-07,
+      "loss": 1.9813,
+      "step": 1237
+    },
+    {
+      "epoch": 2.4138435291250304,
+      "grad_norm": 0.4545169770717621,
+      "learning_rate": 1.0212474297464015e-07,
+      "loss": 2.0467,
+      "step": 1238
+    },
+    {
+      "epoch": 2.4157933219595416,
+      "grad_norm": 0.42319947481155396,
+      "learning_rate": 1.0178204249485949e-07,
+      "loss": 1.9979,
+      "step": 1239
+    },
+    {
+      "epoch": 2.417743114794053,
+      "grad_norm": 0.42224758863449097,
+      "learning_rate": 1.0143934201507881e-07,
+      "loss": 2.0232,
+      "step": 1240
+    },
+    {
+      "epoch": 2.4196929076285647,
+      "grad_norm": 0.4655146598815918,
+      "learning_rate": 1.0109664153529814e-07,
+      "loss": 2.0967,
+      "step": 1241
+    },
+    {
+      "epoch": 2.421642700463076,
+      "grad_norm": 0.4494486451148987,
+      "learning_rate": 1.0075394105551748e-07,
+      "loss": 1.9977,
+      "step": 1242
+    },
+    {
+      "epoch": 2.423592493297587,
+      "grad_norm": 0.451652467250824,
+      "learning_rate": 1.004112405757368e-07,
+      "loss": 2.0475,
+      "step": 1243
+    },
+    {
+      "epoch": 2.4255422861320985,
+      "grad_norm": 0.40768539905548096,
+      "learning_rate": 1.0006854009595612e-07,
+      "loss": 2.0364,
+      "step": 1244
+    },
+    {
+      "epoch": 2.4274920789666097,
+      "grad_norm": 0.44125187397003174,
+      "learning_rate": 9.972583961617547e-08,
+      "loss": 1.9314,
+      "step": 1245
+    },
+    {
+      "epoch": 2.4294418718011213,
+      "grad_norm": 0.44899818301200867,
+      "learning_rate": 9.938313913639479e-08,
+      "loss": 2.0219,
+      "step": 1246
+    },
+    {
+      "epoch": 2.4313916646356324,
+      "grad_norm": 0.43204620480537415,
+      "learning_rate": 9.904043865661411e-08,
+      "loss": 2.0006,
+      "step": 1247
+    },
+    {
+      "epoch": 2.433341457470144,
+      "grad_norm": 0.4570924639701843,
+      "learning_rate": 9.869773817683344e-08,
+      "loss": 1.9907,
+      "step": 1248
+    },
+    {
+      "epoch": 2.435291250304655,
+      "grad_norm": 0.43851619958877563,
+      "learning_rate": 9.835503769705277e-08,
+      "loss": 2.0273,
+      "step": 1249
+    },
+    {
+      "epoch": 2.4372410431391662,
+      "grad_norm": 0.4350932836532593,
+      "learning_rate": 9.80123372172721e-08,
+      "loss": 2.0083,
+      "step": 1250
+    },
+    {
+      "epoch": 2.439190835973678,
+      "grad_norm": 0.4317277669906616,
+      "learning_rate": 9.766963673749143e-08,
+      "loss": 1.9669,
+      "step": 1251
+    },
+    {
+      "epoch": 2.441140628808189,
+      "grad_norm": 0.44647184014320374,
+      "learning_rate": 9.732693625771075e-08,
+      "loss": 2.0481,
+      "step": 1252
+    },
+    {
+      "epoch": 2.4430904216427005,
+      "grad_norm": 0.42302149534225464,
+      "learning_rate": 9.698423577793008e-08,
+      "loss": 2.0113,
+      "step": 1253
+    },
+    {
+      "epoch": 2.4450402144772116,
+      "grad_norm": 0.4655074179172516,
+      "learning_rate": 9.664153529814942e-08,
+      "loss": 1.9743,
+      "step": 1254
+    },
+    {
+      "epoch": 2.4469900073117232,
+      "grad_norm": 0.4577556550502777,
+      "learning_rate": 9.629883481836874e-08,
+      "loss": 2.0582,
+      "step": 1255
+    },
+    {
+      "epoch": 2.4489398001462344,
+      "grad_norm": 0.42997199296951294,
+      "learning_rate": 9.595613433858807e-08,
+      "loss": 1.9791,
+      "step": 1256
+    },
+    {
+      "epoch": 2.450889592980746,
+      "grad_norm": 0.44211360812187195,
+      "learning_rate": 9.56134338588074e-08,
+      "loss": 2.0297,
+      "step": 1257
+    },
+    {
+      "epoch": 2.452839385815257,
+      "grad_norm": 0.45790159702301025,
+      "learning_rate": 9.527073337902673e-08,
+      "loss": 2.1102,
+      "step": 1258
+    },
+    {
+      "epoch": 2.4547891786497686,
+      "grad_norm": 0.42864009737968445,
+      "learning_rate": 9.492803289924605e-08,
+      "loss": 2.0318,
+      "step": 1259
+    },
+    {
+      "epoch": 2.4567389714842798,
+      "grad_norm": 0.44925981760025024,
+      "learning_rate": 9.458533241946539e-08,
+      "loss": 2.0127,
+      "step": 1260
+    },
+    {
+      "epoch": 2.458688764318791,
+      "grad_norm": 0.45514166355133057,
+      "learning_rate": 9.424263193968471e-08,
+      "loss": 2.0644,
+      "step": 1261
+    },
+    {
+      "epoch": 2.4606385571533025,
+      "grad_norm": 0.4267461597919464,
+      "learning_rate": 9.389993145990403e-08,
+      "loss": 2.0096,
+      "step": 1262
+    },
+    {
+      "epoch": 2.4625883499878136,
+      "grad_norm": 0.43666982650756836,
+      "learning_rate": 9.355723098012338e-08,
+      "loss": 1.9746,
+      "step": 1263
+    },
+    {
+      "epoch": 2.464538142822325,
+      "grad_norm": 0.4164760112762451,
+      "learning_rate": 9.32145305003427e-08,
+      "loss": 1.9952,
+      "step": 1264
+    },
+    {
+      "epoch": 2.4664879356568363,
+      "grad_norm": 0.46930649876594543,
+      "learning_rate": 9.287183002056202e-08,
+      "loss": 2.0536,
+      "step": 1265
+    },
+    {
+      "epoch": 2.468437728491348,
+      "grad_norm": 0.44051653146743774,
+      "learning_rate": 9.252912954078134e-08,
+      "loss": 1.9565,
+      "step": 1266
+    },
+    {
+      "epoch": 2.470387521325859,
+      "grad_norm": 0.4479975700378418,
+      "learning_rate": 9.218642906100069e-08,
+      "loss": 1.9897,
+      "step": 1267
+    },
+    {
+      "epoch": 2.4723373141603706,
+      "grad_norm": 0.42661628127098083,
+      "learning_rate": 9.184372858122001e-08,
+      "loss": 2.0528,
+      "step": 1268
+    },
+    {
+      "epoch": 2.4742871069948817,
+      "grad_norm": 0.44977331161499023,
+      "learning_rate": 9.150102810143933e-08,
+      "loss": 2.0217,
+      "step": 1269
+    },
+    {
+      "epoch": 2.4762368998293933,
+      "grad_norm": 0.43319931626319885,
+      "learning_rate": 9.115832762165867e-08,
+      "loss": 2.0457,
+      "step": 1270
+    },
+    {
+      "epoch": 2.4781866926639045,
+      "grad_norm": 0.42555347084999084,
+      "learning_rate": 9.0815627141878e-08,
+      "loss": 2.0342,
+      "step": 1271
+    },
+    {
+      "epoch": 2.4801364854984156,
+      "grad_norm": 0.4264991581439972,
+      "learning_rate": 9.047292666209732e-08,
+      "loss": 2.0189,
+      "step": 1272
+    },
+    {
+      "epoch": 2.482086278332927,
+      "grad_norm": 0.4302367568016052,
+      "learning_rate": 9.013022618231666e-08,
+      "loss": 2.0182,
+      "step": 1273
+    },
+    {
+      "epoch": 2.4840360711674383,
+      "grad_norm": 0.46240586042404175,
+      "learning_rate": 8.978752570253598e-08,
+      "loss": 1.9985,
+      "step": 1274
+    },
+    {
+      "epoch": 2.48598586400195,
+      "grad_norm": 0.4554547965526581,
+      "learning_rate": 8.94448252227553e-08,
+      "loss": 2.023,
+      "step": 1275
+    },
+    {
+      "epoch": 2.487935656836461,
+      "grad_norm": 0.4174085855484009,
+      "learning_rate": 8.910212474297464e-08,
+      "loss": 1.9843,
+      "step": 1276
+    },
+    {
+      "epoch": 2.4898854496709726,
+      "grad_norm": 0.422527939081192,
+      "learning_rate": 8.875942426319397e-08,
+      "loss": 2.0498,
+      "step": 1277
+    },
+    {
+      "epoch": 2.4918352425054837,
+      "grad_norm": 0.4390697777271271,
+      "learning_rate": 8.841672378341329e-08,
+      "loss": 2.0527,
+      "step": 1278
+    },
+    {
+      "epoch": 2.4937850353399953,
+      "grad_norm": 0.4202776551246643,
+      "learning_rate": 8.807402330363262e-08,
+      "loss": 2.0324,
+      "step": 1279
+    },
+    {
+      "epoch": 2.4957348281745064,
+      "grad_norm": 0.45423048734664917,
+      "learning_rate": 8.773132282385196e-08,
+      "loss": 2.0534,
+      "step": 1280
+    },
+    {
+      "epoch": 2.497684621009018,
+      "grad_norm": 0.4252302944660187,
+      "learning_rate": 8.738862234407128e-08,
+      "loss": 2.038,
+      "step": 1281
+    },
+    {
+      "epoch": 2.499634413843529,
+      "grad_norm": 0.4438997209072113,
+      "learning_rate": 8.704592186429061e-08,
+      "loss": 1.9951,
+      "step": 1282
+    },
+    {
+      "epoch": 2.5015842066780403,
+      "grad_norm": 0.4525878429412842,
+      "learning_rate": 8.670322138450993e-08,
+      "loss": 1.9968,
+      "step": 1283
+    },
+    {
+      "epoch": 2.503533999512552,
+      "grad_norm": 0.42212188243865967,
+      "learning_rate": 8.636052090472925e-08,
+      "loss": 1.9487,
+      "step": 1284
+    },
+    {
+      "epoch": 2.505483792347063,
+      "grad_norm": 0.43147486448287964,
+      "learning_rate": 8.60178204249486e-08,
+      "loss": 2.0526,
+      "step": 1285
+    },
+    {
+      "epoch": 2.5074335851815746,
+      "grad_norm": 0.4312281012535095,
+      "learning_rate": 8.567511994516792e-08,
+      "loss": 2.0348,
+      "step": 1286
+    },
+    {
+      "epoch": 2.5093833780160857,
+      "grad_norm": 0.4578983783721924,
+      "learning_rate": 8.533241946538724e-08,
+      "loss": 2.001,
+      "step": 1287
+    },
+    {
+      "epoch": 2.5113331708505973,
+      "grad_norm": 0.4519219994544983,
+      "learning_rate": 8.498971898560658e-08,
+      "loss": 1.9922,
+      "step": 1288
+    },
+    {
+      "epoch": 2.5132829636851084,
+      "grad_norm": 0.4265437424182892,
+      "learning_rate": 8.464701850582591e-08,
+      "loss": 1.9976,
+      "step": 1289
+    },
+    {
+      "epoch": 2.5152327565196195,
+      "grad_norm": 0.4292197823524475,
+      "learning_rate": 8.430431802604523e-08,
+      "loss": 2.013,
+      "step": 1290
+    },
+    {
+      "epoch": 2.517182549354131,
+      "grad_norm": 0.45781049132347107,
+      "learning_rate": 8.396161754626457e-08,
+      "loss": 2.0487,
+      "step": 1291
+    },
+    {
+      "epoch": 2.5191323421886427,
+      "grad_norm": 0.45349356532096863,
+      "learning_rate": 8.361891706648389e-08,
+      "loss": 2.0226,
+      "step": 1292
+    },
+    {
+      "epoch": 2.521082135023154,
+      "grad_norm": 0.4041767120361328,
+      "learning_rate": 8.327621658670322e-08,
+      "loss": 1.946,
+      "step": 1293
+    },
+    {
+      "epoch": 2.523031927857665,
+      "grad_norm": 0.42952778935432434,
+      "learning_rate": 8.293351610692256e-08,
+      "loss": 2.0116,
+      "step": 1294
+    },
+    {
+      "epoch": 2.5249817206921765,
+      "grad_norm": 0.4263194501399994,
+      "learning_rate": 8.259081562714188e-08,
+      "loss": 1.9454,
+      "step": 1295
+    },
+    {
+      "epoch": 2.5269315135266877,
+      "grad_norm": 0.44140732288360596,
+      "learning_rate": 8.22481151473612e-08,
+      "loss": 1.9816,
+      "step": 1296
+    },
+    {
+      "epoch": 2.5288813063611992,
+      "grad_norm": 0.425795316696167,
+      "learning_rate": 8.190541466758052e-08,
+      "loss": 1.9621,
+      "step": 1297
+    },
+    {
+      "epoch": 2.5308310991957104,
+      "grad_norm": 0.481499582529068,
+      "learning_rate": 8.156271418779987e-08,
+      "loss": 2.0685,
+      "step": 1298
+    },
+    {
+      "epoch": 2.532780892030222,
+      "grad_norm": 0.45349955558776855,
+      "learning_rate": 8.122001370801919e-08,
+      "loss": 2.0093,
+      "step": 1299
+    },
+    {
+      "epoch": 2.534730684864733,
+      "grad_norm": 0.45293182134628296,
+      "learning_rate": 8.087731322823851e-08,
+      "loss": 2.0175,
+      "step": 1300
+    },
+    {
+      "epoch": 2.536680477699244,
+      "grad_norm": 0.43124717473983765,
+      "learning_rate": 8.053461274845784e-08,
+      "loss": 2.013,
+      "step": 1301
+    },
+    {
+      "epoch": 2.538630270533756,
+      "grad_norm": 0.4451233744621277,
+      "learning_rate": 8.019191226867718e-08,
+      "loss": 2.0267,
+      "step": 1302
+    },
+    {
+      "epoch": 2.5405800633682674,
+      "grad_norm": 0.4524950683116913,
+      "learning_rate": 7.98492117888965e-08,
+      "loss": 1.9566,
+      "step": 1303
+    },
+    {
+      "epoch": 2.5425298562027785,
+      "grad_norm": 0.43343812227249146,
+      "learning_rate": 7.950651130911583e-08,
+      "loss": 2.0161,
+      "step": 1304
+    },
+    {
+      "epoch": 2.5444796490372896,
+      "grad_norm": 0.44210267066955566,
+      "learning_rate": 7.916381082933515e-08,
+      "loss": 2.0397,
+      "step": 1305
+    },
+    {
+      "epoch": 2.546429441871801,
+      "grad_norm": 0.432013601064682,
+      "learning_rate": 7.882111034955449e-08,
+      "loss": 1.9897,
+      "step": 1306
+    },
+    {
+      "epoch": 2.5483792347063123,
+      "grad_norm": 0.4709438979625702,
+      "learning_rate": 7.847840986977382e-08,
+      "loss": 2.0318,
+      "step": 1307
+    },
+    {
+      "epoch": 2.550329027540824,
+      "grad_norm": 0.421413391828537,
+      "learning_rate": 7.813570938999314e-08,
+      "loss": 1.9603,
+      "step": 1308
+    },
+    {
+      "epoch": 2.552278820375335,
+      "grad_norm": 0.45838281512260437,
+      "learning_rate": 7.779300891021246e-08,
+      "loss": 2.0385,
+      "step": 1309
+    },
+    {
+      "epoch": 2.5542286132098466,
+      "grad_norm": 0.4583630859851837,
+      "learning_rate": 7.74503084304318e-08,
+      "loss": 2.0503,
+      "step": 1310
+    },
+    {
+      "epoch": 2.5561784060443578,
+      "grad_norm": 0.4589591920375824,
+      "learning_rate": 7.710760795065113e-08,
+      "loss": 2.0642,
+      "step": 1311
+    },
+    {
+      "epoch": 2.558128198878869,
+      "grad_norm": 0.4392520487308502,
+      "learning_rate": 7.676490747087045e-08,
+      "loss": 2.1387,
+      "step": 1312
+    },
+    {
+      "epoch": 2.5600779917133805,
+      "grad_norm": 0.42284518480300903,
+      "learning_rate": 7.642220699108979e-08,
+      "loss": 1.9936,
+      "step": 1313
+    },
+    {
+      "epoch": 2.562027784547892,
+      "grad_norm": 0.41540029644966125,
+      "learning_rate": 7.607950651130911e-08,
+      "loss": 2.0042,
+      "step": 1314
+    },
+    {
+      "epoch": 2.563977577382403,
+      "grad_norm": 0.438179075717926,
+      "learning_rate": 7.573680603152844e-08,
+      "loss": 2.0479,
+      "step": 1315
+    },
+    {
+      "epoch": 2.5659273702169143,
+      "grad_norm": 0.4616820812225342,
+      "learning_rate": 7.539410555174778e-08,
+      "loss": 2.0521,
+      "step": 1316
+    },
+    {
+      "epoch": 2.567877163051426,
+      "grad_norm": 0.4477052092552185,
+      "learning_rate": 7.50514050719671e-08,
+      "loss": 2.0674,
+      "step": 1317
+    },
+    {
+      "epoch": 2.569826955885937,
+      "grad_norm": 0.45091158151626587,
+      "learning_rate": 7.470870459218642e-08,
+      "loss": 2.0692,
+      "step": 1318
+    },
+    {
+      "epoch": 2.5717767487204486,
+      "grad_norm": 0.4183453917503357,
+      "learning_rate": 7.436600411240575e-08,
+      "loss": 1.9472,
+      "step": 1319
+    },
+    {
+      "epoch": 2.5737265415549597,
+      "grad_norm": 0.44900861382484436,
+      "learning_rate": 7.402330363262509e-08,
+      "loss": 2.0387,
+      "step": 1320
+    },
+    {
+      "epoch": 2.5756763343894713,
+      "grad_norm": 0.441074401140213,
+      "learning_rate": 7.368060315284441e-08,
+      "loss": 2.0313,
+      "step": 1321
+    },
+    {
+      "epoch": 2.5776261272239824,
+      "grad_norm": 0.4497835338115692,
+      "learning_rate": 7.333790267306374e-08,
+      "loss": 2.0052,
+      "step": 1322
+    },
+    {
+      "epoch": 2.5795759200584936,
+      "grad_norm": 0.42318490147590637,
+      "learning_rate": 7.299520219328306e-08,
+      "loss": 2.0348,
+      "step": 1323
+    },
+    {
+      "epoch": 2.581525712893005,
+      "grad_norm": 0.4562874436378479,
+      "learning_rate": 7.26525017135024e-08,
+      "loss": 1.998,
+      "step": 1324
+    },
+    {
+      "epoch": 2.5834755057275167,
+      "grad_norm": 0.41354840993881226,
+      "learning_rate": 7.230980123372173e-08,
+      "loss": 1.9226,
+      "step": 1325
+    },
+    {
+      "epoch": 2.585425298562028,
+      "grad_norm": 0.4205332398414612,
+      "learning_rate": 7.196710075394105e-08,
+      "loss": 2.0012,
+      "step": 1326
+    },
+    {
+      "epoch": 2.587375091396539,
+      "grad_norm": 0.4440540075302124,
+      "learning_rate": 7.162440027416037e-08,
+      "loss": 1.9842,
+      "step": 1327
+    },
+    {
+      "epoch": 2.5893248842310506,
+      "grad_norm": 0.4503772556781769,
+      "learning_rate": 7.128169979437972e-08,
+      "loss": 2.0297,
+      "step": 1328
+    },
+    {
+      "epoch": 2.5912746770655617,
+      "grad_norm": 0.42969727516174316,
+      "learning_rate": 7.093899931459904e-08,
+      "loss": 1.9969,
+      "step": 1329
+    },
+    {
+      "epoch": 2.5932244699000733,
+      "grad_norm": 0.4641987085342407,
+      "learning_rate": 7.059629883481836e-08,
+      "loss": 2.0081,
+      "step": 1330
+    },
+    {
+      "epoch": 2.5951742627345844,
+      "grad_norm": 0.4018316864967346,
+      "learning_rate": 7.02535983550377e-08,
+      "loss": 2.0136,
+      "step": 1331
+    },
+    {
+      "epoch": 2.597124055569096,
+      "grad_norm": 0.42933347821235657,
+      "learning_rate": 6.991089787525702e-08,
+      "loss": 2.0135,
+      "step": 1332
+    },
+    {
+      "epoch": 2.599073848403607,
+      "grad_norm": 0.4433072805404663,
+      "learning_rate": 6.956819739547635e-08,
+      "loss": 1.9786,
+      "step": 1333
+    },
+    {
+      "epoch": 2.6010236412381182,
+      "grad_norm": 0.4240866005420685,
+      "learning_rate": 6.922549691569569e-08,
+      "loss": 2.0294,
+      "step": 1334
+    },
+    {
+      "epoch": 2.60297343407263,
+      "grad_norm": 0.41409748792648315,
+      "learning_rate": 6.888279643591501e-08,
+      "loss": 1.9859,
+      "step": 1335
+    },
+    {
+      "epoch": 2.604923226907141,
+      "grad_norm": 0.4448906481266022,
+      "learning_rate": 6.854009595613433e-08,
+      "loss": 2.0377,
+      "step": 1336
+    },
+    {
+      "epoch": 2.6068730197416525,
+      "grad_norm": 0.43834927678108215,
+      "learning_rate": 6.819739547635368e-08,
+      "loss": 1.9926,
+      "step": 1337
+    },
+    {
+      "epoch": 2.6088228125761637,
+      "grad_norm": 0.45507144927978516,
+      "learning_rate": 6.7854694996573e-08,
+      "loss": 1.9994,
+      "step": 1338
+    },
+    {
+      "epoch": 2.6107726054106752,
+      "grad_norm": 0.4331389367580414,
+      "learning_rate": 6.751199451679232e-08,
+      "loss": 1.9778,
+      "step": 1339
+    },
+    {
+      "epoch": 2.6127223982451864,
+      "grad_norm": 0.42573344707489014,
+      "learning_rate": 6.716929403701164e-08,
+      "loss": 2.0348,
+      "step": 1340
+    },
+    {
+      "epoch": 2.6146721910796975,
+      "grad_norm": 0.4336106479167938,
+      "learning_rate": 6.682659355723097e-08,
+      "loss": 1.9955,
+      "step": 1341
+    },
+    {
+      "epoch": 2.616621983914209,
+      "grad_norm": 0.4435541331768036,
+      "learning_rate": 6.648389307745031e-08,
+      "loss": 2.005,
+      "step": 1342
+    },
+    {
+      "epoch": 2.6185717767487207,
+      "grad_norm": 0.4269503951072693,
+      "learning_rate": 6.614119259766963e-08,
+      "loss": 1.9855,
+      "step": 1343
+    },
+    {
+      "epoch": 2.620521569583232,
+      "grad_norm": 0.4256771504878998,
+      "learning_rate": 6.579849211788896e-08,
+      "loss": 1.9991,
+      "step": 1344
+    },
+    {
+      "epoch": 2.622471362417743,
+      "grad_norm": 0.42985865473747253,
+      "learning_rate": 6.545579163810828e-08,
+      "loss": 1.9968,
+      "step": 1345
+    },
+    {
+      "epoch": 2.6244211552522545,
+      "grad_norm": 0.4007837772369385,
+      "learning_rate": 6.511309115832762e-08,
+      "loss": 2.0383,
+      "step": 1346
+    },
+    {
+      "epoch": 2.6263709480867656,
+      "grad_norm": 0.4280641973018646,
+      "learning_rate": 6.477039067854695e-08,
+      "loss": 2.023,
+      "step": 1347
+    },
+    {
+      "epoch": 2.628320740921277,
+      "grad_norm": 0.4413083791732788,
+      "learning_rate": 6.442769019876627e-08,
+      "loss": 2.0358,
+      "step": 1348
+    },
+    {
+      "epoch": 2.6302705337557883,
+      "grad_norm": 0.46252715587615967,
+      "learning_rate": 6.40849897189856e-08,
+      "loss": 2.0432,
+      "step": 1349
+    },
+    {
+      "epoch": 2.6322203265903,
+      "grad_norm": 0.4754340350627899,
+      "learning_rate": 6.374228923920494e-08,
+      "loss": 2.0282,
+      "step": 1350
+    },
+    {
+      "epoch": 2.634170119424811,
+      "grad_norm": 0.43686631321907043,
+      "learning_rate": 6.339958875942426e-08,
+      "loss": 1.9914,
+      "step": 1351
+    },
+    {
+      "epoch": 2.636119912259322,
+      "grad_norm": 0.4519065022468567,
+      "learning_rate": 6.305688827964358e-08,
+      "loss": 1.9913,
+      "step": 1352
+    },
+    {
+      "epoch": 2.6380697050938338,
+      "grad_norm": 0.4338942766189575,
+      "learning_rate": 6.271418779986292e-08,
+      "loss": 1.9278,
+      "step": 1353
+    },
+    {
+      "epoch": 2.6400194979283453,
+      "grad_norm": 0.4273185729980469,
+      "learning_rate": 6.237148732008224e-08,
+      "loss": 1.9683,
+      "step": 1354
+    },
+    {
+      "epoch": 2.6419692907628565,
+      "grad_norm": 0.4474540650844574,
+      "learning_rate": 6.202878684030157e-08,
+      "loss": 1.9811,
+      "step": 1355
+    },
+    {
+      "epoch": 2.6439190835973676,
+      "grad_norm": 0.43969810009002686,
+      "learning_rate": 6.168608636052091e-08,
+      "loss": 2.0206,
+      "step": 1356
+    },
+    {
+      "epoch": 2.645868876431879,
+      "grad_norm": 0.434159517288208,
+      "learning_rate": 6.134338588074023e-08,
+      "loss": 1.941,
+      "step": 1357
+    },
+    {
+      "epoch": 2.6478186692663903,
+      "grad_norm": 0.42709431052207947,
+      "learning_rate": 6.100068540095956e-08,
+      "loss": 2.0679,
+      "step": 1358
+    },
+    {
+      "epoch": 2.649768462100902,
+      "grad_norm": 0.4570341408252716,
+      "learning_rate": 6.065798492117888e-08,
+      "loss": 1.9906,
+      "step": 1359
+    },
+    {
+      "epoch": 2.651718254935413,
+      "grad_norm": 0.45039740204811096,
+      "learning_rate": 6.031528444139822e-08,
+      "loss": 2.0602,
+      "step": 1360
+    },
+    {
+      "epoch": 2.6536680477699246,
+      "grad_norm": 0.4509207606315613,
+      "learning_rate": 5.997258396161754e-08,
+      "loss": 1.9592,
+      "step": 1361
+    },
+    {
+      "epoch": 2.6556178406044357,
+      "grad_norm": 0.4447914958000183,
+      "learning_rate": 5.962988348183687e-08,
+      "loss": 1.977,
+      "step": 1362
+    },
+    {
+      "epoch": 2.657567633438947,
+      "grad_norm": 0.43301117420196533,
+      "learning_rate": 5.92871830020562e-08,
+      "loss": 1.8566,
+      "step": 1363
+    },
+    {
+      "epoch": 2.6595174262734584,
+      "grad_norm": 0.43681600689888,
+      "learning_rate": 5.894448252227552e-08,
+      "loss": 1.9502,
+      "step": 1364
+    },
+    {
+      "epoch": 2.66146721910797,
+      "grad_norm": 0.45358729362487793,
+      "learning_rate": 5.8601782042494856e-08,
+      "loss": 1.9963,
+      "step": 1365
+    },
+    {
+      "epoch": 2.663417011942481,
+      "grad_norm": 0.4327445328235626,
+      "learning_rate": 5.8259081562714184e-08,
+      "loss": 2.0573,
+      "step": 1366
+    },
+    {
+      "epoch": 2.6653668047769923,
+      "grad_norm": 0.4307156503200531,
+      "learning_rate": 5.791638108293351e-08,
+      "loss": 2.026,
+      "step": 1367
+    },
+    {
+      "epoch": 2.667316597611504,
+      "grad_norm": 0.4320853054523468,
+      "learning_rate": 5.757368060315284e-08,
+      "loss": 1.9879,
+      "step": 1368
+    },
+    {
+      "epoch": 2.669266390446015,
+      "grad_norm": 0.4349263608455658,
+      "learning_rate": 5.7230980123372174e-08,
+      "loss": 1.9817,
+      "step": 1369
+    },
+    {
+      "epoch": 2.6712161832805266,
+      "grad_norm": 0.477649986743927,
+      "learning_rate": 5.6888279643591495e-08,
+      "loss": 2.0803,
+      "step": 1370
+    },
+    {
+      "epoch": 2.6731659761150377,
+      "grad_norm": 0.43707796931266785,
+      "learning_rate": 5.654557916381083e-08,
+      "loss": 2.0257,
+      "step": 1371
+    },
+    {
+      "epoch": 2.6751157689495493,
+      "grad_norm": 0.43572092056274414,
+      "learning_rate": 5.6202878684030156e-08,
+      "loss": 1.9668,
+      "step": 1372
+    },
+    {
+      "epoch": 2.6770655617840604,
+      "grad_norm": 0.45613327622413635,
+      "learning_rate": 5.5860178204249484e-08,
+      "loss": 1.9635,
+      "step": 1373
+    },
+    {
+      "epoch": 2.6790153546185715,
+      "grad_norm": 0.4179806113243103,
+      "learning_rate": 5.551747772446881e-08,
+      "loss": 1.9834,
+      "step": 1374
+    },
+    {
+      "epoch": 2.680965147453083,
+      "grad_norm": 0.4401552677154541,
+      "learning_rate": 5.5174777244688146e-08,
+      "loss": 2.0564,
+      "step": 1375
+    },
+    {
+      "epoch": 2.6829149402875947,
+      "grad_norm": 0.44003936648368835,
+      "learning_rate": 5.483207676490747e-08,
+      "loss": 2.012,
+      "step": 1376
+    },
+    {
+      "epoch": 2.684864733122106,
+      "grad_norm": 0.4311721920967102,
+      "learning_rate": 5.4489376285126795e-08,
+      "loss": 1.9477,
+      "step": 1377
+    },
+    {
+      "epoch": 2.686814525956617,
+      "grad_norm": 0.44712117314338684,
+      "learning_rate": 5.414667580534613e-08,
+      "loss": 2.0537,
+      "step": 1378
+    },
+    {
+      "epoch": 2.6887643187911285,
+      "grad_norm": 0.4397601783275604,
+      "learning_rate": 5.380397532556545e-08,
+      "loss": 1.9269,
+      "step": 1379
+    },
+    {
+      "epoch": 2.6907141116256397,
+      "grad_norm": 0.413853257894516,
+      "learning_rate": 5.3461274845784784e-08,
+      "loss": 1.9361,
+      "step": 1380
+    },
+    {
+      "epoch": 2.6926639044601512,
+      "grad_norm": 0.4286174774169922,
+      "learning_rate": 5.3118574366004105e-08,
+      "loss": 1.9624,
+      "step": 1381
+    },
+    {
+      "epoch": 2.6946136972946624,
+      "grad_norm": 0.4571874439716339,
+      "learning_rate": 5.277587388622344e-08,
+      "loss": 2.0422,
+      "step": 1382
+    },
+    {
+      "epoch": 2.696563490129174,
+      "grad_norm": 0.4140987694263458,
+      "learning_rate": 5.243317340644277e-08,
+      "loss": 1.9979,
+      "step": 1383
+    },
+    {
+      "epoch": 2.698513282963685,
+      "grad_norm": 0.4402693808078766,
+      "learning_rate": 5.2090472926662095e-08,
+      "loss": 2.0034,
+      "step": 1384
+    },
+    {
+      "epoch": 2.700463075798196,
+      "grad_norm": 0.43446001410484314,
+      "learning_rate": 5.174777244688142e-08,
+      "loss": 2.0259,
+      "step": 1385
+    },
+    {
+      "epoch": 2.702412868632708,
+      "grad_norm": 0.4453582763671875,
+      "learning_rate": 5.1405071967100756e-08,
+      "loss": 1.9536,
+      "step": 1386
+    },
+    {
+      "epoch": 2.7043626614672194,
+      "grad_norm": 0.4660840630531311,
+      "learning_rate": 5.106237148732008e-08,
+      "loss": 2.0633,
+      "step": 1387
+    },
+    {
+      "epoch": 2.7063124543017305,
+      "grad_norm": 0.45734548568725586,
+      "learning_rate": 5.0719671007539405e-08,
+      "loss": 2.0105,
+      "step": 1388
+    },
+    {
+      "epoch": 2.7082622471362416,
+      "grad_norm": 0.4270325303077698,
+      "learning_rate": 5.037697052775874e-08,
+      "loss": 1.9925,
+      "step": 1389
+    },
+    {
+      "epoch": 2.710212039970753,
+      "grad_norm": 0.4218302369117737,
+      "learning_rate": 5.003427004797806e-08,
+      "loss": 1.9478,
+      "step": 1390
+    },
+    {
+      "epoch": 2.7121618328052643,
+      "grad_norm": 0.42878419160842896,
+      "learning_rate": 4.9691569568197395e-08,
+      "loss": 1.9816,
+      "step": 1391
+    },
+    {
+      "epoch": 2.714111625639776,
+      "grad_norm": 0.457177996635437,
+      "learning_rate": 4.934886908841672e-08,
+      "loss": 2.0461,
+      "step": 1392
+    },
+    {
+      "epoch": 2.716061418474287,
+      "grad_norm": 0.4213179647922516,
+      "learning_rate": 4.900616860863605e-08,
+      "loss": 2.0361,
+      "step": 1393
+    },
+    {
+      "epoch": 2.7180112113087986,
+      "grad_norm": 0.43558916449546814,
+      "learning_rate": 4.866346812885538e-08,
+      "loss": 2.0455,
+      "step": 1394
+    },
+    {
+      "epoch": 2.7199610041433098,
+      "grad_norm": 0.4477299451828003,
+      "learning_rate": 4.832076764907471e-08,
+      "loss": 2.0509,
+      "step": 1395
+    },
+    {
+      "epoch": 2.721910796977821,
+      "grad_norm": 0.44148439168930054,
+      "learning_rate": 4.797806716929403e-08,
+      "loss": 1.9977,
+      "step": 1396
+    },
+    {
+      "epoch": 2.7238605898123325,
+      "grad_norm": 0.4374631345272064,
+      "learning_rate": 4.763536668951337e-08,
+      "loss": 1.9677,
+      "step": 1397
+    },
+    {
+      "epoch": 2.7258103826468436,
+      "grad_norm": 0.4173222780227661,
+      "learning_rate": 4.7292666209732695e-08,
+      "loss": 2.0426,
+      "step": 1398
+    },
+    {
+      "epoch": 2.727760175481355,
+      "grad_norm": 0.4424826204776764,
+      "learning_rate": 4.6949965729952016e-08,
+      "loss": 2.0592,
+      "step": 1399
+    },
+    {
+      "epoch": 2.7297099683158663,
+      "grad_norm": 0.45383262634277344,
+      "learning_rate": 4.660726525017135e-08,
+      "loss": 2.0148,
+      "step": 1400
+    },
+    {
+      "epoch": 2.731659761150378,
+      "grad_norm": 0.40516209602355957,
+      "learning_rate": 4.626456477039067e-08,
+      "loss": 1.9739,
+      "step": 1401
+    },
+    {
+      "epoch": 2.733609553984889,
+      "grad_norm": 0.4282223880290985,
+      "learning_rate": 4.5921864290610005e-08,
+      "loss": 1.977,
+      "step": 1402
+    },
+    {
+      "epoch": 2.7355593468194,
+      "grad_norm": 0.4480572044849396,
+      "learning_rate": 4.557916381082933e-08,
+      "loss": 1.9405,
+      "step": 1403
+    },
+    {
+      "epoch": 2.7375091396539117,
+      "grad_norm": 0.43524089455604553,
+      "learning_rate": 4.523646333104866e-08,
+      "loss": 2.0241,
+      "step": 1404
+    },
+    {
+      "epoch": 2.7394589324884233,
+      "grad_norm": 0.455409973859787,
+      "learning_rate": 4.489376285126799e-08,
+      "loss": 2.0057,
+      "step": 1405
+    },
+    {
+      "epoch": 2.7414087253229344,
+      "grad_norm": 0.43779870867729187,
+      "learning_rate": 4.455106237148732e-08,
+      "loss": 1.9976,
+      "step": 1406
+    },
+    {
+      "epoch": 2.7433585181574456,
+      "grad_norm": 0.42375847697257996,
+      "learning_rate": 4.420836189170664e-08,
+      "loss": 1.9067,
+      "step": 1407
+    },
+    {
+      "epoch": 2.745308310991957,
+      "grad_norm": 0.4355579912662506,
+      "learning_rate": 4.386566141192598e-08,
+      "loss": 1.974,
+      "step": 1408
+    },
+    {
+      "epoch": 2.7472581038264683,
+      "grad_norm": 0.430085688829422,
+      "learning_rate": 4.3522960932145305e-08,
+      "loss": 2.0081,
+      "step": 1409
+    },
+    {
+      "epoch": 2.74920789666098,
+      "grad_norm": 0.4447627365589142,
+      "learning_rate": 4.3180260452364626e-08,
+      "loss": 1.9915,
+      "step": 1410
+    },
+    {
+      "epoch": 2.751157689495491,
+      "grad_norm": 0.41174575686454773,
+      "learning_rate": 4.283755997258396e-08,
+      "loss": 1.9572,
+      "step": 1411
+    },
+    {
+      "epoch": 2.7531074823300026,
+      "grad_norm": 0.4330975413322449,
+      "learning_rate": 4.249485949280329e-08,
+      "loss": 2.0492,
+      "step": 1412
+    },
+    {
+      "epoch": 2.7550572751645137,
+      "grad_norm": 0.4706032872200012,
+      "learning_rate": 4.2152159013022616e-08,
+      "loss": 2.0414,
+      "step": 1413
+    },
+    {
+      "epoch": 2.757007067999025,
+      "grad_norm": 0.45478367805480957,
+      "learning_rate": 4.180945853324194e-08,
+      "loss": 2.0053,
+      "step": 1414
+    },
+    {
+      "epoch": 2.7589568608335364,
+      "grad_norm": 0.4465593099594116,
+      "learning_rate": 4.146675805346128e-08,
+      "loss": 1.932,
+      "step": 1415
+    },
+    {
+      "epoch": 2.760906653668048,
+      "grad_norm": 0.4564541280269623,
+      "learning_rate": 4.11240575736806e-08,
+      "loss": 2.0402,
+      "step": 1416
+    },
+    {
+      "epoch": 2.762856446502559,
+      "grad_norm": 0.4326637387275696,
+      "learning_rate": 4.078135709389993e-08,
+      "loss": 2.0531,
+      "step": 1417
+    },
+    {
+      "epoch": 2.7648062393370703,
+      "grad_norm": 0.46069779992103577,
+      "learning_rate": 4.0438656614119254e-08,
+      "loss": 2.0529,
+      "step": 1418
+    },
+    {
+      "epoch": 2.766756032171582,
+      "grad_norm": 0.42713242769241333,
+      "learning_rate": 4.009595613433859e-08,
+      "loss": 2.0555,
+      "step": 1419
+    },
+    {
+      "epoch": 2.768705825006093,
+      "grad_norm": 0.46604225039482117,
+      "learning_rate": 3.9753255654557916e-08,
+      "loss": 2.0123,
+      "step": 1420
+    },
+    {
+      "epoch": 2.7706556178406045,
+      "grad_norm": 0.44637739658355713,
+      "learning_rate": 3.941055517477724e-08,
+      "loss": 1.9867,
+      "step": 1421
+    },
+    {
+      "epoch": 2.7726054106751157,
+      "grad_norm": 0.44564053416252136,
+      "learning_rate": 3.906785469499657e-08,
+      "loss": 1.9587,
+      "step": 1422
+    },
+    {
+      "epoch": 2.7745552035096273,
+      "grad_norm": 0.413260281085968,
+      "learning_rate": 3.87251542152159e-08,
+      "loss": 2.0343,
+      "step": 1423
+    },
+    {
+      "epoch": 2.7765049963441384,
+      "grad_norm": 0.4345310926437378,
+      "learning_rate": 3.8382453735435226e-08,
+      "loss": 2.0119,
+      "step": 1424
+    },
+    {
+      "epoch": 2.7784547891786495,
+      "grad_norm": 0.4478965997695923,
+      "learning_rate": 3.8039753255654554e-08,
+      "loss": 2.02,
+      "step": 1425
+    },
+    {
+      "epoch": 2.780404582013161,
+      "grad_norm": 0.44151195883750916,
+      "learning_rate": 3.769705277587389e-08,
+      "loss": 2.0169,
+      "step": 1426
+    },
+    {
+      "epoch": 2.7823543748476727,
+      "grad_norm": 0.4505918025970459,
+      "learning_rate": 3.735435229609321e-08,
+      "loss": 2.0449,
+      "step": 1427
+    },
+    {
+      "epoch": 2.784304167682184,
+      "grad_norm": 0.4261781871318817,
+      "learning_rate": 3.7011651816312543e-08,
+      "loss": 2.0652,
+      "step": 1428
+    },
+    {
+      "epoch": 2.786253960516695,
+      "grad_norm": 0.44098833203315735,
+      "learning_rate": 3.666895133653187e-08,
+      "loss": 2.0455,
+      "step": 1429
+    },
+    {
+      "epoch": 2.7882037533512065,
+      "grad_norm": 0.42615678906440735,
+      "learning_rate": 3.63262508567512e-08,
+      "loss": 2.0186,
+      "step": 1430
+    },
+    {
+      "epoch": 2.7901535461857176,
+      "grad_norm": 0.45692741870880127,
+      "learning_rate": 3.5983550376970526e-08,
+      "loss": 1.9664,
+      "step": 1431
+    },
+    {
+      "epoch": 2.7921033390202292,
+      "grad_norm": 0.437515527009964,
+      "learning_rate": 3.564084989718986e-08,
+      "loss": 2.0042,
+      "step": 1432
+    },
+    {
+      "epoch": 2.7940531318547404,
+      "grad_norm": 0.4577179253101349,
+      "learning_rate": 3.529814941740918e-08,
+      "loss": 2.0074,
+      "step": 1433
+    },
+    {
+      "epoch": 2.796002924689252,
+      "grad_norm": 0.4680233299732208,
+      "learning_rate": 3.495544893762851e-08,
+      "loss": 2.0081,
+      "step": 1434
+    },
+    {
+      "epoch": 2.797952717523763,
+      "grad_norm": 0.440106600522995,
+      "learning_rate": 3.4612748457847843e-08,
+      "loss": 2.0441,
+      "step": 1435
+    },
+    {
+      "epoch": 2.799902510358274,
+      "grad_norm": 0.4136613607406616,
+      "learning_rate": 3.4270047978067164e-08,
+      "loss": 1.9053,
+      "step": 1436
+    },
+    {
+      "epoch": 2.8018523031927858,
+      "grad_norm": 0.42198479175567627,
+      "learning_rate": 3.39273474982865e-08,
+      "loss": 2.0024,
+      "step": 1437
+    },
+    {
+      "epoch": 2.8038020960272974,
+      "grad_norm": 0.45030972361564636,
+      "learning_rate": 3.358464701850582e-08,
+      "loss": 2.0419,
+      "step": 1438
+    },
+    {
+      "epoch": 2.8057518888618085,
+      "grad_norm": 0.3976515233516693,
+      "learning_rate": 3.3241946538725154e-08,
+      "loss": 1.9997,
+      "step": 1439
+    },
+    {
+      "epoch": 2.8077016816963196,
+      "grad_norm": 0.4644767940044403,
+      "learning_rate": 3.289924605894448e-08,
+      "loss": 2.0295,
+      "step": 1440
+    },
+    {
+      "epoch": 2.809651474530831,
+      "grad_norm": 0.45879313349723816,
+      "learning_rate": 3.255654557916381e-08,
+      "loss": 2.0531,
+      "step": 1441
+    },
+    {
+      "epoch": 2.8116012673653423,
+      "grad_norm": 0.4370739161968231,
+      "learning_rate": 3.221384509938314e-08,
+      "loss": 1.9524,
+      "step": 1442
+    },
+    {
+      "epoch": 2.813551060199854,
+      "grad_norm": 0.4276754856109619,
+      "learning_rate": 3.187114461960247e-08,
+      "loss": 2.0108,
+      "step": 1443
+    },
+    {
+      "epoch": 2.815500853034365,
+      "grad_norm": 0.4218864142894745,
+      "learning_rate": 3.152844413982179e-08,
+      "loss": 1.9939,
+      "step": 1444
+    },
+    {
+      "epoch": 2.8174506458688766,
+      "grad_norm": 0.44251778721809387,
+      "learning_rate": 3.118574366004112e-08,
+      "loss": 2.0217,
+      "step": 1445
+    },
+    {
+      "epoch": 2.8194004387033877,
+      "grad_norm": 0.4414883553981781,
+      "learning_rate": 3.0843043180260454e-08,
+      "loss": 2.0019,
+      "step": 1446
+    },
+    {
+      "epoch": 2.821350231537899,
+      "grad_norm": 0.4485674798488617,
+      "learning_rate": 3.050034270047978e-08,
+      "loss": 1.9658,
+      "step": 1447
+    },
+    {
+      "epoch": 2.8233000243724105,
+      "grad_norm": 0.45192667841911316,
+      "learning_rate": 3.015764222069911e-08,
+      "loss": 2.0346,
+      "step": 1448
+    },
+    {
+      "epoch": 2.825249817206922,
+      "grad_norm": 0.43689438700675964,
+      "learning_rate": 2.981494174091844e-08,
+      "loss": 1.9979,
+      "step": 1449
+    },
+    {
+      "epoch": 2.827199610041433,
+      "grad_norm": 0.4323086142539978,
+      "learning_rate": 2.947224126113776e-08,
+      "loss": 1.982,
+      "step": 1450
+    },
+    {
+      "epoch": 2.8291494028759443,
+      "grad_norm": 0.42197027802467346,
+      "learning_rate": 2.9129540781357092e-08,
+      "loss": 1.9698,
+      "step": 1451
+    },
+    {
+      "epoch": 2.831099195710456,
+      "grad_norm": 0.42834025621414185,
+      "learning_rate": 2.878684030157642e-08,
+      "loss": 2.0422,
+      "step": 1452
+    },
+    {
+      "epoch": 2.833048988544967,
+      "grad_norm": 0.4507107138633728,
+      "learning_rate": 2.8444139821795747e-08,
+      "loss": 2.0761,
+      "step": 1453
+    },
+    {
+      "epoch": 2.8349987813794786,
+      "grad_norm": 0.4503672122955322,
+      "learning_rate": 2.8101439342015078e-08,
+      "loss": 1.9846,
+      "step": 1454
+    },
+    {
+      "epoch": 2.8369485742139897,
+      "grad_norm": 0.43898439407348633,
+      "learning_rate": 2.7758738862234406e-08,
+      "loss": 2.0528,
+      "step": 1455
+    },
+    {
+      "epoch": 2.8388983670485013,
+      "grad_norm": 0.4379539489746094,
+      "learning_rate": 2.7416038382453733e-08,
+      "loss": 2.0037,
+      "step": 1456
+    },
+    {
+      "epoch": 2.8408481598830124,
+      "grad_norm": 0.432386577129364,
+      "learning_rate": 2.7073337902673064e-08,
+      "loss": 2.0275,
+      "step": 1457
+    },
+    {
+      "epoch": 2.8427979527175236,
+      "grad_norm": 0.43408092856407166,
+      "learning_rate": 2.6730637422892392e-08,
+      "loss": 2.074,
+      "step": 1458
+    },
+    {
+      "epoch": 2.844747745552035,
+      "grad_norm": 0.4475395679473877,
+      "learning_rate": 2.638793694311172e-08,
+      "loss": 2.0375,
+      "step": 1459
+    },
+    {
+      "epoch": 2.8466975383865463,
+      "grad_norm": 0.4225776493549347,
+      "learning_rate": 2.6045236463331047e-08,
+      "loss": 1.9984,
+      "step": 1460
+    },
+    {
+      "epoch": 2.848647331221058,
+      "grad_norm": 0.43169689178466797,
+      "learning_rate": 2.5702535983550378e-08,
+      "loss": 2.0475,
+      "step": 1461
+    },
+    {
+      "epoch": 2.850597124055569,
+      "grad_norm": 0.433133989572525,
+      "learning_rate": 2.5359835503769703e-08,
+      "loss": 2.0557,
+      "step": 1462
+    },
+    {
+      "epoch": 2.8525469168900806,
+      "grad_norm": 0.45559263229370117,
+      "learning_rate": 2.501713502398903e-08,
+      "loss": 2.004,
+      "step": 1463
+    },
+    {
+      "epoch": 2.8544967097245917,
+      "grad_norm": 0.4383031129837036,
+      "learning_rate": 2.467443454420836e-08,
+      "loss": 2.0464,
+      "step": 1464
+    },
+    {
+      "epoch": 2.856446502559103,
+      "grad_norm": 0.43648001551628113,
+      "learning_rate": 2.433173406442769e-08,
+      "loss": 1.9961,
+      "step": 1465
+    },
+    {
+      "epoch": 2.8583962953936144,
+      "grad_norm": 0.43689587712287903,
+      "learning_rate": 2.3989033584647016e-08,
+      "loss": 1.9981,
+      "step": 1466
+    },
+    {
+      "epoch": 2.860346088228126,
+      "grad_norm": 0.43933171033859253,
+      "learning_rate": 2.3646333104866347e-08,
+      "loss": 2.0508,
+      "step": 1467
+    },
+    {
+      "epoch": 2.862295881062637,
+      "grad_norm": 0.44383934140205383,
+      "learning_rate": 2.3303632625085675e-08,
+      "loss": 1.9662,
+      "step": 1468
+    },
+    {
+      "epoch": 2.8642456738971482,
+      "grad_norm": 0.44308724999427795,
+      "learning_rate": 2.2960932145305003e-08,
+      "loss": 2.0537,
+      "step": 1469
+    },
+    {
+      "epoch": 2.86619546673166,
+      "grad_norm": 0.4277627766132355,
+      "learning_rate": 2.261823166552433e-08,
+      "loss": 2.0042,
+      "step": 1470
+    },
+    {
+      "epoch": 2.868145259566171,
+      "grad_norm": 0.45172715187072754,
+      "learning_rate": 2.227553118574366e-08,
+      "loss": 2.0579,
+      "step": 1471
+    },
+    {
+      "epoch": 2.8700950524006825,
+      "grad_norm": 0.4282797873020172,
+      "learning_rate": 2.193283070596299e-08,
+      "loss": 1.9594,
+      "step": 1472
+    },
+    {
+      "epoch": 2.8720448452351937,
+      "grad_norm": 0.4169709384441376,
+      "learning_rate": 2.1590130226182313e-08,
+      "loss": 2.0176,
+      "step": 1473
+    },
+    {
+      "epoch": 2.8739946380697052,
+      "grad_norm": 0.4406467378139496,
+      "learning_rate": 2.1247429746401644e-08,
+      "loss": 2.0083,
+      "step": 1474
+    },
+    {
+      "epoch": 2.8759444309042164,
+      "grad_norm": 0.44870543479919434,
+      "learning_rate": 2.090472926662097e-08,
+      "loss": 2.0036,
+      "step": 1475
+    },
+    {
+      "epoch": 2.8778942237387275,
+      "grad_norm": 0.45419618487358093,
+      "learning_rate": 2.05620287868403e-08,
+      "loss": 2.0227,
+      "step": 1476
+    },
+    {
+      "epoch": 2.879844016573239,
+      "grad_norm": 0.40358567237854004,
+      "learning_rate": 2.0219328307059627e-08,
+      "loss": 1.9999,
+      "step": 1477
+    },
+    {
+      "epoch": 2.8817938094077507,
+      "grad_norm": 0.4595142602920532,
+      "learning_rate": 1.9876627827278958e-08,
+      "loss": 2.0622,
+      "step": 1478
+    },
+    {
+      "epoch": 2.883743602242262,
+      "grad_norm": 0.46610069274902344,
+      "learning_rate": 1.9533927347498285e-08,
+      "loss": 2.0356,
+      "step": 1479
+    },
+    {
+      "epoch": 2.885693395076773,
+      "grad_norm": 0.4541744291782379,
+      "learning_rate": 1.9191226867717613e-08,
+      "loss": 2.0371,
+      "step": 1480
+    },
+    {
+      "epoch": 2.8876431879112845,
+      "grad_norm": 0.416298508644104,
+      "learning_rate": 1.8848526387936944e-08,
+      "loss": 1.9623,
+      "step": 1481
+    },
+    {
+      "epoch": 2.8895929807457956,
+      "grad_norm": 0.4404451251029968,
+      "learning_rate": 1.8505825908156272e-08,
+      "loss": 2.1124,
+      "step": 1482
+    },
+    {
+      "epoch": 2.891542773580307,
+      "grad_norm": 0.4394998848438263,
+      "learning_rate": 1.81631254283756e-08,
+      "loss": 2.0782,
+      "step": 1483
+    },
+    {
+      "epoch": 2.8934925664148183,
+      "grad_norm": 0.46584269404411316,
+      "learning_rate": 1.782042494859493e-08,
+      "loss": 2.0165,
+      "step": 1484
+    },
+    {
+      "epoch": 2.89544235924933,
+      "grad_norm": 0.45088499784469604,
+      "learning_rate": 1.7477724468814255e-08,
+      "loss": 1.9465,
+      "step": 1485
+    },
+    {
+      "epoch": 2.897392152083841,
+      "grad_norm": 0.417900949716568,
+      "learning_rate": 1.7135023989033582e-08,
+      "loss": 2.0229,
+      "step": 1486
+    },
+    {
+      "epoch": 2.899341944918352,
+      "grad_norm": 0.4679426848888397,
+      "learning_rate": 1.679232350925291e-08,
+      "loss": 2.0294,
+      "step": 1487
+    },
+    {
+      "epoch": 2.9012917377528638,
+      "grad_norm": 0.4508228302001953,
+      "learning_rate": 1.644962302947224e-08,
+      "loss": 1.9921,
+      "step": 1488
+    },
+    {
+      "epoch": 2.9032415305873753,
+      "grad_norm": 0.41129690408706665,
+      "learning_rate": 1.610692254969157e-08,
+      "loss": 2.0249,
+      "step": 1489
+    },
+    {
+      "epoch": 2.9051913234218865,
+      "grad_norm": 0.397612988948822,
+      "learning_rate": 1.5764222069910896e-08,
+      "loss": 1.9656,
+      "step": 1490
+    },
+    {
+      "epoch": 2.9071411162563976,
+      "grad_norm": 0.41494330763816833,
+      "learning_rate": 1.5421521590130227e-08,
+      "loss": 2.0063,
+      "step": 1491
+    },
+    {
+      "epoch": 2.909090909090909,
+      "grad_norm": 0.41872090101242065,
+      "learning_rate": 1.5078821110349555e-08,
+      "loss": 2.0596,
+      "step": 1492
+    },
+    {
+      "epoch": 2.9110407019254203,
+      "grad_norm": 0.4372239112854004,
+      "learning_rate": 1.473612063056888e-08,
+      "loss": 2.0347,
+      "step": 1493
+    },
+    {
+      "epoch": 2.912990494759932,
+      "grad_norm": 0.5028005838394165,
+      "learning_rate": 1.439342015078821e-08,
+      "loss": 2.0734,
+      "step": 1494
+    },
+    {
+      "epoch": 2.914940287594443,
+      "grad_norm": 0.4472177028656006,
+      "learning_rate": 1.4050719671007539e-08,
+      "loss": 1.984,
+      "step": 1495
+    },
+    {
+      "epoch": 2.9168900804289546,
+      "grad_norm": 0.4196500778198242,
+      "learning_rate": 1.3708019191226867e-08,
+      "loss": 2.0038,
+      "step": 1496
+    },
+    {
+      "epoch": 2.9188398732634657,
+      "grad_norm": 0.4304303526878357,
+      "learning_rate": 1.3365318711446196e-08,
+      "loss": 2.0214,
+      "step": 1497
+    },
+    {
+      "epoch": 2.920789666097977,
+      "grad_norm": 0.43512046337127686,
+      "learning_rate": 1.3022618231665524e-08,
+      "loss": 2.0264,
+      "step": 1498
+    },
+    {
+      "epoch": 2.9227394589324884,
+      "grad_norm": 0.4143558442592621,
+      "learning_rate": 1.2679917751884851e-08,
+      "loss": 2.0229,
+      "step": 1499
+    },
+    {
+      "epoch": 2.924689251767,
+      "grad_norm": 0.45177680253982544,
+      "learning_rate": 1.233721727210418e-08,
+      "loss": 2.0142,
+      "step": 1500
+    },
+    {
+      "epoch": 2.926639044601511,
+      "grad_norm": 0.4237738251686096,
+      "learning_rate": 1.1994516792323508e-08,
+      "loss": 2.1034,
+      "step": 1501
+    },
+    {
+      "epoch": 2.9285888374360223,
+      "grad_norm": 0.45249003171920776,
+      "learning_rate": 1.1651816312542837e-08,
+      "loss": 2.0412,
+      "step": 1502
+    },
+    {
+      "epoch": 2.930538630270534,
+      "grad_norm": 0.44523778557777405,
+      "learning_rate": 1.1309115832762165e-08,
+      "loss": 2.0167,
+      "step": 1503
+    },
+    {
+      "epoch": 2.932488423105045,
+      "grad_norm": 0.4320509433746338,
+      "learning_rate": 1.0966415352981494e-08,
+      "loss": 2.0711,
+      "step": 1504
+    },
+    {
+      "epoch": 2.9344382159395566,
+      "grad_norm": 0.44118785858154297,
+      "learning_rate": 1.0623714873200822e-08,
+      "loss": 2.0188,
+      "step": 1505
+    },
+    {
+      "epoch": 2.9363880087740677,
+      "grad_norm": 0.464070588350296,
+      "learning_rate": 1.028101439342015e-08,
+      "loss": 1.9916,
+      "step": 1506
+    },
+    {
+      "epoch": 2.9383378016085793,
+      "grad_norm": 0.4519140422344208,
+      "learning_rate": 9.938313913639479e-09,
+      "loss": 1.9818,
+      "step": 1507
+    },
+    {
+      "epoch": 2.9402875944430904,
+      "grad_norm": 0.4367213845252991,
+      "learning_rate": 9.595613433858807e-09,
+      "loss": 1.9969,
+      "step": 1508
+    },
+    {
+      "epoch": 2.9422373872776015,
+      "grad_norm": 0.44010135531425476,
+      "learning_rate": 9.252912954078136e-09,
+      "loss": 2.0042,
+      "step": 1509
+    },
+    {
+      "epoch": 2.944187180112113,
+      "grad_norm": 0.42604339122772217,
+      "learning_rate": 8.910212474297465e-09,
+      "loss": 1.9714,
+      "step": 1510
+    },
+    {
+      "epoch": 2.9461369729466247,
+      "grad_norm": 0.41623058915138245,
+      "learning_rate": 8.567511994516791e-09,
+      "loss": 1.876,
+      "step": 1511
+    },
+    {
+      "epoch": 2.948086765781136,
+      "grad_norm": 0.4275459945201874,
+      "learning_rate": 8.22481151473612e-09,
+      "loss": 1.9769,
+      "step": 1512
+    },
+    {
+      "epoch": 2.950036558615647,
+      "grad_norm": 0.42414721846580505,
+      "learning_rate": 7.882111034955448e-09,
+      "loss": 2.0048,
+      "step": 1513
+    },
+    {
+      "epoch": 2.9519863514501585,
+      "grad_norm": 0.44245725870132446,
+      "learning_rate": 7.539410555174777e-09,
+      "loss": 2.0181,
+      "step": 1514
+    },
+    {
+      "epoch": 2.9539361442846697,
+      "grad_norm": 0.4433532655239105,
+      "learning_rate": 7.196710075394105e-09,
+      "loss": 1.99,
+      "step": 1515
+    },
+    {
+      "epoch": 2.9558859371191812,
+      "grad_norm": 0.4201236665248871,
+      "learning_rate": 6.854009595613433e-09,
+      "loss": 2.0502,
+      "step": 1516
+    },
+    {
+      "epoch": 2.9578357299536924,
+      "grad_norm": 0.4342474043369293,
+      "learning_rate": 6.511309115832762e-09,
+      "loss": 2.0019,
+      "step": 1517
+    },
+    {
+      "epoch": 2.959785522788204,
+      "grad_norm": 0.4743070900440216,
+      "learning_rate": 6.16860863605209e-09,
+      "loss": 2.0605,
+      "step": 1518
+    },
+    {
+      "epoch": 2.961735315622715,
+      "grad_norm": 0.4557639956474304,
+      "learning_rate": 5.825908156271419e-09,
+      "loss": 1.9851,
+      "step": 1519
+    },
+    {
+      "epoch": 2.963685108457226,
+      "grad_norm": 0.4486991763114929,
+      "learning_rate": 5.483207676490747e-09,
+      "loss": 2.0393,
+      "step": 1520
+    },
+    {
+      "epoch": 2.965634901291738,
+      "grad_norm": 0.4367319643497467,
+      "learning_rate": 5.140507196710075e-09,
+      "loss": 2.0389,
+      "step": 1521
+    },
+    {
+      "epoch": 2.967584694126249,
+      "grad_norm": 0.45666617155075073,
+      "learning_rate": 4.797806716929403e-09,
+      "loss": 2.0502,
+      "step": 1522
+    },
+    {
+      "epoch": 2.9695344869607605,
+      "grad_norm": 0.43143293261528015,
+      "learning_rate": 4.4551062371487326e-09,
+      "loss": 2.091,
+      "step": 1523
+    },
+    {
+      "epoch": 2.9714842797952716,
+      "grad_norm": 0.4313979744911194,
+      "learning_rate": 4.11240575736806e-09,
+      "loss": 1.8739,
+      "step": 1524
+    },
+    {
+      "epoch": 2.973434072629783,
+      "grad_norm": 0.49060067534446716,
+      "learning_rate": 3.769705277587389e-09,
+      "loss": 2.0295,
+      "step": 1525
+    },
+    {
+      "epoch": 2.9753838654642943,
+      "grad_norm": 0.44060608744621277,
+      "learning_rate": 3.4270047978067167e-09,
+      "loss": 2.0589,
+      "step": 1526
+    },
+    {
+      "epoch": 2.9773336582988055,
+      "grad_norm": 0.4408791661262512,
+      "learning_rate": 3.084304318026045e-09,
+      "loss": 2.0144,
+      "step": 1527
+    },
+    {
+      "epoch": 2.979283451133317,
+      "grad_norm": 0.4273541271686554,
+      "learning_rate": 2.7416038382453736e-09,
+      "loss": 2.0209,
+      "step": 1528
+    },
+    {
+      "epoch": 2.9812332439678286,
+      "grad_norm": 0.4148176908493042,
+      "learning_rate": 2.3989033584647016e-09,
+      "loss": 2.0129,
+      "step": 1529
+    },
+    {
+      "epoch": 2.9831830368023398,
+      "grad_norm": 0.4230509400367737,
+      "learning_rate": 2.05620287868403e-09,
+      "loss": 1.9813,
+      "step": 1530
+    },
+    {
+      "epoch": 2.985132829636851,
+      "grad_norm": 0.49288874864578247,
+      "learning_rate": 1.7135023989033583e-09,
+      "loss": 2.0771,
+      "step": 1531
+    },
+    {
+      "epoch": 2.9870826224713625,
+      "grad_norm": 0.41819414496421814,
+      "learning_rate": 1.3708019191226868e-09,
+      "loss": 1.9969,
+      "step": 1532
+    },
+    {
+      "epoch": 2.9890324153058736,
+      "grad_norm": 0.4301490783691406,
+      "learning_rate": 1.028101439342015e-09,
+      "loss": 1.9538,
+      "step": 1533
+    },
+    {
+      "epoch": 2.990982208140385,
+      "grad_norm": 0.4619876444339752,
+      "learning_rate": 6.854009595613434e-10,
+      "loss": 2.0211,
+      "step": 1534
+    },
+    {
+      "epoch": 2.9929320009748963,
+      "grad_norm": 0.4294811487197876,
+      "learning_rate": 3.427004797806717e-10,
+      "loss": 2.0085,
+      "step": 1535
+    },
+    {
+      "epoch": 2.994881793809408,
+      "grad_norm": 0.44153717160224915,
+      "learning_rate": 0.0,
+      "loss": 2.0167,
+      "step": 1536
+    },
+    {
+      "epoch": 2.994881793809408,
+      "eval_loss": 2.0117971897125244,
+      "eval_runtime": 480.4707,
+      "eval_samples_per_second": 1.295,
+      "eval_steps_per_second": 0.325,
+      "step": 1536
     }
   ],
   "logging_steps": 1,
@@ -3613,12 +10797,12 @@
         "should_evaluate": false,
         "should_log": false,
         "should_save": true,
-        "should_training_stop": false
+        "should_training_stop": true
       },
       "attributes": {}
     }
   },
-  "total_flos": 1.5142103496146289e+18,
+  "total_flos": 4.53488100805863e+18,
   "train_batch_size": 2,
   "trial_name": null,
   "trial_params": null