{
  "best_metric": 0.03601429983973503,
  "best_model_checkpoint": "saves/psy-course/Llama3-OpenBioLLM-8B/train/fold8/checkpoint-1300",
  "epoch": 4.996682778883518,
  "eval_steps": 50,
  "global_step": 3295,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.015164439389631315,
      "grad_norm": 6.69179105758667,
      "learning_rate": 3.0303030303030305e-06,
      "loss": 1.3822,
      "step": 10
    },
    {
      "epoch": 0.03032887877926263,
      "grad_norm": 6.022090435028076,
      "learning_rate": 6.060606060606061e-06,
      "loss": 1.3562,
      "step": 20
    },
    {
      "epoch": 0.045493318168893945,
      "grad_norm": 3.7154126167297363,
      "learning_rate": 9.090909090909091e-06,
      "loss": 1.0285,
      "step": 30
    },
    {
      "epoch": 0.06065775755852526,
      "grad_norm": 2.2793216705322266,
      "learning_rate": 1.2121212121212122e-05,
      "loss": 0.7714,
      "step": 40
    },
    {
      "epoch": 0.07582219694815658,
      "grad_norm": 1.6856296062469482,
      "learning_rate": 1.5151515151515153e-05,
      "loss": 0.5651,
      "step": 50
    },
    {
      "epoch": 0.07582219694815658,
      "eval_loss": 0.34537237882614136,
      "eval_runtime": 206.8033,
      "eval_samples_per_second": 5.672,
      "eval_steps_per_second": 5.672,
      "step": 50
    },
    {
      "epoch": 0.09098663633778789,
      "grad_norm": 1.1443841457366943,
      "learning_rate": 1.8181818181818182e-05,
      "loss": 0.2852,
      "step": 60
    },
    {
      "epoch": 0.1061510757274192,
      "grad_norm": 1.259319543838501,
      "learning_rate": 2.1212121212121215e-05,
      "loss": 0.1781,
      "step": 70
    },
    {
      "epoch": 0.12131551511705052,
      "grad_norm": 1.664919376373291,
      "learning_rate": 2.4242424242424244e-05,
      "loss": 0.1862,
      "step": 80
    },
    {
      "epoch": 0.13647995450668182,
      "grad_norm": 1.8507658243179321,
      "learning_rate": 2.7272727272727273e-05,
      "loss": 0.1156,
      "step": 90
    },
    {
      "epoch": 0.15164439389631315,
      "grad_norm": 1.0127924680709839,
      "learning_rate": 3.0303030303030306e-05,
      "loss": 0.159,
      "step": 100
    },
    {
      "epoch": 0.15164439389631315,
      "eval_loss": 0.09158477187156677,
      "eval_runtime": 206.5291,
      "eval_samples_per_second": 5.68,
      "eval_steps_per_second": 5.68,
      "step": 100
    },
    {
      "epoch": 0.16680883328594445,
      "grad_norm": 0.5882529020309448,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.0949,
      "step": 110
    },
    {
      "epoch": 0.18197327267557578,
      "grad_norm": 1.612291932106018,
      "learning_rate": 3.6363636363636364e-05,
      "loss": 0.0905,
      "step": 120
    },
    {
      "epoch": 0.19713771206520708,
      "grad_norm": 1.462253212928772,
      "learning_rate": 3.939393939393939e-05,
      "loss": 0.0932,
      "step": 130
    },
    {
      "epoch": 0.2123021514548384,
      "grad_norm": 0.7238670587539673,
      "learning_rate": 4.242424242424243e-05,
      "loss": 0.0719,
      "step": 140
    },
    {
      "epoch": 0.2274665908444697,
      "grad_norm": 1.0951787233352661,
      "learning_rate": 4.545454545454546e-05,
      "loss": 0.0775,
      "step": 150
    },
    {
      "epoch": 0.2274665908444697,
      "eval_loss": 0.06597652286291122,
      "eval_runtime": 206.2354,
      "eval_samples_per_second": 5.688,
      "eval_steps_per_second": 5.688,
      "step": 150
    },
    {
      "epoch": 0.24263103023410104,
      "grad_norm": 0.9825966358184814,
      "learning_rate": 4.848484848484849e-05,
      "loss": 0.0688,
      "step": 160
    },
    {
      "epoch": 0.25779546962373234,
      "grad_norm": 1.0078524351119995,
      "learning_rate": 5.151515151515152e-05,
      "loss": 0.0726,
      "step": 170
    },
    {
      "epoch": 0.27295990901336364,
      "grad_norm": 1.0524179935455322,
      "learning_rate": 5.4545454545454546e-05,
      "loss": 0.084,
      "step": 180
    },
    {
      "epoch": 0.288124348402995,
      "grad_norm": 0.7553159594535828,
      "learning_rate": 5.757575757575758e-05,
      "loss": 0.0741,
      "step": 190
    },
    {
      "epoch": 0.3032887877926263,
      "grad_norm": 1.2318599224090576,
      "learning_rate": 6.060606060606061e-05,
      "loss": 0.0565,
      "step": 200
    },
    {
      "epoch": 0.3032887877926263,
      "eval_loss": 0.05869923532009125,
      "eval_runtime": 205.8994,
      "eval_samples_per_second": 5.697,
      "eval_steps_per_second": 5.697,
      "step": 200
    },
    {
      "epoch": 0.3184532271822576,
      "grad_norm": 0.7203672528266907,
      "learning_rate": 6.363636363636364e-05,
      "loss": 0.0557,
      "step": 210
    },
    {
      "epoch": 0.3336176665718889,
      "grad_norm": 0.9632550477981567,
      "learning_rate": 6.666666666666667e-05,
      "loss": 0.0588,
      "step": 220
    },
    {
      "epoch": 0.34878210596152026,
      "grad_norm": 0.7189677357673645,
      "learning_rate": 6.96969696969697e-05,
      "loss": 0.0813,
      "step": 230
    },
    {
      "epoch": 0.36394654535115156,
      "grad_norm": 0.7692343592643738,
      "learning_rate": 7.272727272727273e-05,
      "loss": 0.053,
      "step": 240
    },
    {
      "epoch": 0.37911098474078286,
      "grad_norm": 0.8544824123382568,
      "learning_rate": 7.575757575757576e-05,
      "loss": 0.0563,
      "step": 250
    },
    {
      "epoch": 0.37911098474078286,
      "eval_loss": 0.05937855690717697,
      "eval_runtime": 206.5519,
      "eval_samples_per_second": 5.679,
      "eval_steps_per_second": 5.679,
      "step": 250
    },
    {
      "epoch": 0.39427542413041416,
      "grad_norm": 0.6232314109802246,
      "learning_rate": 7.878787878787879e-05,
      "loss": 0.0534,
      "step": 260
    },
    {
      "epoch": 0.40943986352004547,
      "grad_norm": 0.350496768951416,
      "learning_rate": 8.181818181818183e-05,
      "loss": 0.0794,
      "step": 270
    },
    {
      "epoch": 0.4246043029096768,
      "grad_norm": 0.9284454584121704,
      "learning_rate": 8.484848484848486e-05,
      "loss": 0.0607,
      "step": 280
    },
    {
      "epoch": 0.4397687422993081,
      "grad_norm": 0.5696659088134766,
      "learning_rate": 8.787878787878789e-05,
      "loss": 0.0464,
      "step": 290
    },
    {
      "epoch": 0.4549331816889394,
      "grad_norm": 0.7423091530799866,
      "learning_rate": 9.090909090909092e-05,
      "loss": 0.0631,
      "step": 300
    },
    {
      "epoch": 0.4549331816889394,
      "eval_loss": 0.05754981189966202,
      "eval_runtime": 207.1337,
      "eval_samples_per_second": 5.663,
      "eval_steps_per_second": 5.663,
      "step": 300
    },
    {
      "epoch": 0.4700976210785707,
      "grad_norm": 0.46020084619522095,
      "learning_rate": 9.393939393939395e-05,
      "loss": 0.0479,
      "step": 310
    },
    {
      "epoch": 0.4852620604682021,
      "grad_norm": 0.6511795520782471,
      "learning_rate": 9.696969696969698e-05,
      "loss": 0.047,
      "step": 320
    },
    {
      "epoch": 0.5004264998578334,
      "grad_norm": 0.9548239707946777,
      "learning_rate": 0.0001,
      "loss": 0.0567,
      "step": 330
    },
    {
      "epoch": 0.5155909392474647,
      "grad_norm": 0.6067355275154114,
      "learning_rate": 9.999719336268101e-05,
      "loss": 0.0527,
      "step": 340
    },
    {
      "epoch": 0.530755378637096,
      "grad_norm": 0.4896825850009918,
      "learning_rate": 9.998877376581251e-05,
      "loss": 0.0677,
      "step": 350
    },
    {
      "epoch": 0.530755378637096,
      "eval_loss": 0.05034831538796425,
      "eval_runtime": 205.9619,
      "eval_samples_per_second": 5.695,
      "eval_steps_per_second": 5.695,
      "step": 350
    },
    {
      "epoch": 0.5459198180267273,
      "grad_norm": 0.5949435234069824,
      "learning_rate": 9.997474215462472e-05,
      "loss": 0.0531,
      "step": 360
    },
    {
      "epoch": 0.5610842574163586,
      "grad_norm": 0.7067845463752747,
      "learning_rate": 9.995510010438337e-05,
      "loss": 0.041,
      "step": 370
    },
    {
      "epoch": 0.57624869680599,
      "grad_norm": 0.6521092057228088,
      "learning_rate": 9.992984982021295e-05,
      "loss": 0.0584,
      "step": 380
    },
    {
      "epoch": 0.5914131361956213,
      "grad_norm": 0.5313027501106262,
      "learning_rate": 9.9898994136849e-05,
      "loss": 0.0549,
      "step": 390
    },
    {
      "epoch": 0.6065775755852526,
      "grad_norm": 1.3004082441329956,
      "learning_rate": 9.986253651832005e-05,
      "loss": 0.0366,
      "step": 400
    },
    {
      "epoch": 0.6065775755852526,
      "eval_loss": 0.04712074622511864,
      "eval_runtime": 206.1479,
      "eval_samples_per_second": 5.69,
      "eval_steps_per_second": 5.69,
      "step": 400
    },
    {
      "epoch": 0.6217420149748839,
      "grad_norm": 0.5177161693572998,
      "learning_rate": 9.982048105755859e-05,
      "loss": 0.0414,
      "step": 410
    },
    {
      "epoch": 0.6369064543645152,
      "grad_norm": 0.36376678943634033,
      "learning_rate": 9.977283247594166e-05,
      "loss": 0.0556,
      "step": 420
    },
    {
      "epoch": 0.6520708937541465,
      "grad_norm": 0.46044835448265076,
      "learning_rate": 9.971959612276076e-05,
      "loss": 0.0463,
      "step": 430
    },
    {
      "epoch": 0.6672353331437778,
      "grad_norm": 0.5747679471969604,
      "learning_rate": 9.966077797462129e-05,
      "loss": 0.05,
      "step": 440
    },
    {
      "epoch": 0.6823997725334091,
      "grad_norm": 0.3989450931549072,
      "learning_rate": 9.959638463477165e-05,
      "loss": 0.0397,
      "step": 450
    },
    {
      "epoch": 0.6823997725334091,
      "eval_loss": 0.04302780702710152,
      "eval_runtime": 207.0673,
      "eval_samples_per_second": 5.665,
      "eval_steps_per_second": 5.665,
      "step": 450
    },
    {
      "epoch": 0.6975642119230405,
      "grad_norm": 0.37973183393478394,
      "learning_rate": 9.952642333236186e-05,
      "loss": 0.0496,
      "step": 460
    },
    {
      "epoch": 0.7127286513126718,
      "grad_norm": 0.4910580515861511,
      "learning_rate": 9.945090192163202e-05,
      "loss": 0.0476,
      "step": 470
    },
    {
      "epoch": 0.7278930907023031,
      "grad_norm": 0.44105058908462524,
      "learning_rate": 9.936982888103051e-05,
      "loss": 0.0394,
      "step": 480
    },
    {
      "epoch": 0.7430575300919344,
      "grad_norm": 0.519553542137146,
      "learning_rate": 9.928321331226219e-05,
      "loss": 0.045,
      "step": 490
    },
    {
      "epoch": 0.7582219694815657,
      "grad_norm": 0.28150394558906555,
      "learning_rate": 9.919106493926655e-05,
      "loss": 0.0383,
      "step": 500
    },
    {
      "epoch": 0.7582219694815657,
      "eval_loss": 0.04788108915090561,
      "eval_runtime": 206.939,
      "eval_samples_per_second": 5.668,
      "eval_steps_per_second": 5.668,
      "step": 500
    },
    {
      "epoch": 0.773386408871197,
      "grad_norm": 0.28955650329589844,
      "learning_rate": 9.909339410712612e-05,
      "loss": 0.0482,
      "step": 510
    },
    {
      "epoch": 0.7885508482608283,
      "grad_norm": 0.6213341355323792,
      "learning_rate": 9.8990211780905e-05,
      "loss": 0.0394,
      "step": 520
    },
    {
      "epoch": 0.8037152876504596,
      "grad_norm": 0.44944509863853455,
      "learning_rate": 9.888152954441785e-05,
      "loss": 0.0454,
      "step": 530
    },
    {
      "epoch": 0.8188797270400909,
      "grad_norm": 0.3272422254085541,
      "learning_rate": 9.876735959892953e-05,
      "loss": 0.0392,
      "step": 540
    },
    {
      "epoch": 0.8340441664297223,
      "grad_norm": 0.7294989824295044,
      "learning_rate": 9.864771476178522e-05,
      "loss": 0.0508,
      "step": 550
    },
    {
      "epoch": 0.8340441664297223,
      "eval_loss": 0.04269608110189438,
      "eval_runtime": 206.1894,
      "eval_samples_per_second": 5.689,
      "eval_steps_per_second": 5.689,
      "step": 550
    },
    {
      "epoch": 0.8492086058193536,
      "grad_norm": 0.41436007618904114,
      "learning_rate": 9.852260846497153e-05,
      "loss": 0.0486,
      "step": 560
    },
    {
      "epoch": 0.864373045208985,
      "grad_norm": 0.8043579459190369,
      "learning_rate": 9.839205475360851e-05,
      "loss": 0.0505,
      "step": 570
    },
    {
      "epoch": 0.8795374845986162,
      "grad_norm": 0.4937809705734253,
      "learning_rate": 9.825606828437291e-05,
      "loss": 0.0561,
      "step": 580
    },
    {
      "epoch": 0.8947019239882475,
      "grad_norm": 0.4687022268772125,
      "learning_rate": 9.811466432385267e-05,
      "loss": 0.0428,
      "step": 590
    },
    {
      "epoch": 0.9098663633778789,
      "grad_norm": 0.5363739132881165,
      "learning_rate": 9.796785874683314e-05,
      "loss": 0.0346,
      "step": 600
    },
    {
      "epoch": 0.9098663633778789,
      "eval_loss": 0.04343252256512642,
      "eval_runtime": 206.7862,
      "eval_samples_per_second": 5.673,
      "eval_steps_per_second": 5.673,
      "step": 600
    },
    {
      "epoch": 0.9250308027675102,
      "grad_norm": 0.446513295173645,
      "learning_rate": 9.781566803451475e-05,
      "loss": 0.0566,
      "step": 610
    },
    {
      "epoch": 0.9401952421571415,
      "grad_norm": 0.27837061882019043,
      "learning_rate": 9.765810927266281e-05,
      "loss": 0.0415,
      "step": 620
    },
    {
      "epoch": 0.9553596815467729,
      "grad_norm": 0.27945834398269653,
      "learning_rate": 9.749520014968934e-05,
      "loss": 0.0398,
      "step": 630
    },
    {
      "epoch": 0.9705241209364042,
      "grad_norm": 0.33713576197624207,
      "learning_rate": 9.732695895466735e-05,
      "loss": 0.0437,
      "step": 640
    },
    {
      "epoch": 0.9856885603260355,
      "grad_norm": 0.32925042510032654,
      "learning_rate": 9.715340457527746e-05,
      "loss": 0.0513,
      "step": 650
    },
    {
      "epoch": 0.9856885603260355,
      "eval_loss": 0.044393427670001984,
      "eval_runtime": 207.4663,
      "eval_samples_per_second": 5.654,
      "eval_steps_per_second": 5.654,
      "step": 650
    },
    {
      "epoch": 1.0008529997156668,
      "grad_norm": 0.25668787956237793,
      "learning_rate": 9.697455649568761e-05,
      "loss": 0.0373,
      "step": 660
    },
    {
      "epoch": 1.016017439105298,
      "grad_norm": 0.575491726398468,
      "learning_rate": 9.679043479436556e-05,
      "loss": 0.0316,
      "step": 670
    },
    {
      "epoch": 1.0311818784949294,
      "grad_norm": 0.21307440102100372,
      "learning_rate": 9.660106014182489e-05,
      "loss": 0.0272,
      "step": 680
    },
    {
      "epoch": 1.0463463178845607,
      "grad_norm": 0.27312636375427246,
      "learning_rate": 9.640645379830424e-05,
      "loss": 0.0453,
      "step": 690
    },
    {
      "epoch": 1.061510757274192,
      "grad_norm": 0.286176860332489,
      "learning_rate": 9.620663761138067e-05,
      "loss": 0.0339,
      "step": 700
    },
    {
      "epoch": 1.061510757274192,
      "eval_loss": 0.04173321649432182,
      "eval_runtime": 206.3468,
      "eval_samples_per_second": 5.685,
      "eval_steps_per_second": 5.685,
      "step": 700
    },
    {
      "epoch": 1.0766751966638233,
      "grad_norm": 0.1278642863035202,
      "learning_rate": 9.600163401351688e-05,
      "loss": 0.0254,
      "step": 710
    },
    {
      "epoch": 1.0918396360534546,
      "grad_norm": 0.23175828158855438,
      "learning_rate": 9.579146601954276e-05,
      "loss": 0.0494,
      "step": 720
    },
    {
      "epoch": 1.1070040754430859,
      "grad_norm": 0.10618554055690765,
      "learning_rate": 9.557615722407177e-05,
      "loss": 0.0378,
      "step": 730
    },
    {
      "epoch": 1.1221685148327172,
      "grad_norm": 0.11078658699989319,
      "learning_rate": 9.535573179885191e-05,
      "loss": 0.032,
      "step": 740
    },
    {
      "epoch": 1.1373329542223485,
      "grad_norm": 0.2334222048521042,
      "learning_rate": 9.513021449005214e-05,
      "loss": 0.0296,
      "step": 750
    },
    {
      "epoch": 1.1373329542223485,
      "eval_loss": 0.04424874484539032,
      "eval_runtime": 206.976,
      "eval_samples_per_second": 5.667,
      "eval_steps_per_second": 5.667,
      "step": 750
    },
    {
      "epoch": 1.15249739361198,
      "grad_norm": 0.2527366578578949,
      "learning_rate": 9.489963061548428e-05,
      "loss": 0.0338,
      "step": 760
    },
    {
      "epoch": 1.1676618330016113,
      "grad_norm": 0.23535087704658508,
      "learning_rate": 9.466400606176062e-05,
      "loss": 0.0226,
      "step": 770
    },
    {
      "epoch": 1.1828262723912426,
      "grad_norm": 0.33640772104263306,
      "learning_rate": 9.442336728138779e-05,
      "loss": 0.0234,
      "step": 780
    },
    {
      "epoch": 1.197990711780874,
      "grad_norm": 0.18896321952342987,
      "learning_rate": 9.417774128979706e-05,
      "loss": 0.0368,
      "step": 790
    },
    {
      "epoch": 1.2131551511705052,
      "grad_norm": 0.24871914088726044,
      "learning_rate": 9.39271556623114e-05,
      "loss": 0.0288,
      "step": 800
    },
    {
      "epoch": 1.2131551511705052,
      "eval_loss": 0.039680108428001404,
      "eval_runtime": 207.6653,
      "eval_samples_per_second": 5.649,
      "eval_steps_per_second": 5.649,
      "step": 800
    },
    {
      "epoch": 1.2283195905601365,
      "grad_norm": 0.40964576601982117,
      "learning_rate": 9.367163853104975e-05,
      "loss": 0.036,
      "step": 810
    },
    {
      "epoch": 1.2434840299497678,
      "grad_norm": 0.22146688401699066,
      "learning_rate": 9.341121858176876e-05,
      "loss": 0.0369,
      "step": 820
    },
    {
      "epoch": 1.2586484693393991,
      "grad_norm": 0.2575783133506775,
      "learning_rate": 9.314592505064239e-05,
      "loss": 0.0344,
      "step": 830
    },
    {
      "epoch": 1.2738129087290304,
      "grad_norm": 0.19507430493831635,
      "learning_rate": 9.28757877209796e-05,
      "loss": 0.0265,
      "step": 840
    },
    {
      "epoch": 1.2889773481186617,
      "grad_norm": 0.2289421558380127,
      "learning_rate": 9.260083691988084e-05,
      "loss": 0.0299,
      "step": 850
    },
    {
      "epoch": 1.2889773481186617,
      "eval_loss": 0.04206722602248192,
      "eval_runtime": 207.1881,
      "eval_samples_per_second": 5.662,
      "eval_steps_per_second": 5.662,
      "step": 850
    },
    {
      "epoch": 1.304141787508293,
      "grad_norm": 0.5384100675582886,
      "learning_rate": 9.232110351483327e-05,
      "loss": 0.0373,
      "step": 860
    },
    {
      "epoch": 1.3193062268979243,
      "grad_norm": 0.5763593912124634,
      "learning_rate": 9.203661891024547e-05,
      "loss": 0.0376,
      "step": 870
    },
    {
      "epoch": 1.3344706662875556,
      "grad_norm": 0.34299781918525696,
      "learning_rate": 9.174741504392173e-05,
      "loss": 0.0311,
      "step": 880
    },
    {
      "epoch": 1.349635105677187,
      "grad_norm": 0.254780650138855,
      "learning_rate": 9.145352438347662e-05,
      "loss": 0.0355,
      "step": 890
    },
    {
      "epoch": 1.3647995450668184,
      "grad_norm": 0.46846649050712585,
      "learning_rate": 9.115497992268995e-05,
      "loss": 0.0293,
      "step": 900
    },
    {
      "epoch": 1.3647995450668184,
      "eval_loss": 0.04009329527616501,
      "eval_runtime": 206.6929,
      "eval_samples_per_second": 5.675,
      "eval_steps_per_second": 5.675,
      "step": 900
    },
    {
      "epoch": 1.3799639844564497,
      "grad_norm": 0.07255344092845917,
      "learning_rate": 9.085181517780273e-05,
      "loss": 0.0312,
      "step": 910
    },
    {
      "epoch": 1.395128423846081,
      "grad_norm": 0.18492460250854492,
      "learning_rate": 9.054406418375443e-05,
      "loss": 0.0329,
      "step": 920
    },
    {
      "epoch": 1.4102928632357123,
      "grad_norm": 0.34338879585266113,
      "learning_rate": 9.023176149036203e-05,
      "loss": 0.0288,
      "step": 930
    },
    {
      "epoch": 1.4254573026253436,
      "grad_norm": 0.22292286157608032,
      "learning_rate": 8.991494215844132e-05,
      "loss": 0.0441,
      "step": 940
    },
    {
      "epoch": 1.440621742014975,
      "grad_norm": 0.37994903326034546,
      "learning_rate": 8.959364175587069e-05,
      "loss": 0.0278,
      "step": 950
    },
    {
      "epoch": 1.440621742014975,
      "eval_loss": 0.03929847106337547,
      "eval_runtime": 207.7851,
      "eval_samples_per_second": 5.645,
      "eval_steps_per_second": 5.645,
      "step": 950
    },
    {
      "epoch": 1.4557861814046062,
      "grad_norm": 0.24896001815795898,
      "learning_rate": 8.926789635359817e-05,
      "loss": 0.043,
      "step": 960
    },
    {
      "epoch": 1.4709506207942376,
      "grad_norm": 0.16315113008022308,
      "learning_rate": 8.893774252159187e-05,
      "loss": 0.0298,
      "step": 970
    },
    {
      "epoch": 1.4861150601838689,
      "grad_norm": 0.19432376325130463,
      "learning_rate": 8.860321732473439e-05,
      "loss": 0.0364,
      "step": 980
    },
    {
      "epoch": 1.5012794995735002,
      "grad_norm": 0.3546680212020874,
      "learning_rate": 8.826435831866184e-05,
      "loss": 0.0373,
      "step": 990
    },
    {
      "epoch": 1.5164439389631315,
      "grad_norm": 0.09447364509105682,
      "learning_rate": 8.79212035455475e-05,
      "loss": 0.0283,
      "step": 1000
    },
    {
      "epoch": 1.5164439389631315,
      "eval_loss": 0.040457405149936676,
      "eval_runtime": 207.9569,
      "eval_samples_per_second": 5.641,
      "eval_steps_per_second": 5.641,
      "step": 1000
    },
    {
      "epoch": 1.5316083783527628,
      "grad_norm": 0.2906639873981476,
      "learning_rate": 8.757379152983103e-05,
      "loss": 0.0273,
      "step": 1010
    },
    {
      "epoch": 1.546772817742394,
      "grad_norm": 0.4261516332626343,
      "learning_rate": 8.722216127389363e-05,
      "loss": 0.0337,
      "step": 1020
    },
    {
      "epoch": 1.5619372571320254,
      "grad_norm": 0.3574506640434265,
      "learning_rate": 8.686635225367919e-05,
      "loss": 0.0285,
      "step": 1030
    },
    {
      "epoch": 1.5771016965216567,
      "grad_norm": 0.2925097346305847,
      "learning_rate": 8.650640441426274e-05,
      "loss": 0.0367,
      "step": 1040
    },
    {
      "epoch": 1.592266135911288,
      "grad_norm": 0.4504016935825348,
      "learning_rate": 8.614235816536582e-05,
      "loss": 0.0493,
      "step": 1050
    },
    {
      "epoch": 1.592266135911288,
      "eval_loss": 0.039341144263744354,
      "eval_runtime": 206.8668,
      "eval_samples_per_second": 5.67,
      "eval_steps_per_second": 5.67,
      "step": 1050
    },
    {
      "epoch": 1.6074305753009193,
      "grad_norm": 0.28596726059913635,
      "learning_rate": 8.577425437681994e-05,
      "loss": 0.0352,
      "step": 1060
    },
    {
      "epoch": 1.6225950146905506,
      "grad_norm": 0.12631350755691528,
      "learning_rate": 8.540213437397833e-05,
      "loss": 0.0317,
      "step": 1070
    },
    {
      "epoch": 1.6377594540801819,
      "grad_norm": 0.1901707649230957,
      "learning_rate": 8.502603993307647e-05,
      "loss": 0.0271,
      "step": 1080
    },
    {
      "epoch": 1.6529238934698132,
      "grad_norm": 0.34583041071891785,
      "learning_rate": 8.464601327654207e-05,
      "loss": 0.0261,
      "step": 1090
    },
    {
      "epoch": 1.6680883328594445,
      "grad_norm": 0.35897573828697205,
      "learning_rate": 8.4262097068255e-05,
      "loss": 0.0287,
      "step": 1100
    },
    {
      "epoch": 1.6680883328594445,
      "eval_loss": 0.03920057788491249,
      "eval_runtime": 205.8057,
      "eval_samples_per_second": 5.7,
      "eval_steps_per_second": 5.7,
      "step": 1100
    },
    {
      "epoch": 1.6832527722490758,
      "grad_norm": 0.5413476228713989,
      "learning_rate": 8.387433440875758e-05,
      "loss": 0.0332,
      "step": 1110
    },
    {
      "epoch": 1.698417211638707,
      "grad_norm": 0.5442347526550293,
      "learning_rate": 8.348276883041583e-05,
      "loss": 0.0313,
      "step": 1120
    },
    {
      "epoch": 1.7135816510283386,
      "grad_norm": 0.3635435998439789,
      "learning_rate": 8.308744429253238e-05,
      "loss": 0.0388,
      "step": 1130
    },
    {
      "epoch": 1.72874609041797,
      "grad_norm": 0.43414217233657837,
      "learning_rate": 8.268840517641123e-05,
      "loss": 0.0352,
      "step": 1140
    },
    {
      "epoch": 1.7439105298076012,
      "grad_norm": 0.571806788444519,
      "learning_rate": 8.228569628037543e-05,
      "loss": 0.0383,
      "step": 1150
    },
    {
      "epoch": 1.7439105298076012,
      "eval_loss": 0.03788575902581215,
      "eval_runtime": 207.4735,
      "eval_samples_per_second": 5.654,
      "eval_steps_per_second": 5.654,
      "step": 1150
    },
    {
      "epoch": 1.7590749691972325,
      "grad_norm": 0.30155709385871887,
      "learning_rate": 8.187936281473758e-05,
      "loss": 0.0311,
      "step": 1160
    },
    {
      "epoch": 1.7742394085868638,
      "grad_norm": 0.11790592968463898,
      "learning_rate": 8.146945039672446e-05,
      "loss": 0.0288,
      "step": 1170
    },
    {
      "epoch": 1.789403847976495,
      "grad_norm": 0.3467343747615814,
      "learning_rate": 8.105600504535567e-05,
      "loss": 0.0309,
      "step": 1180
    },
    {
      "epoch": 1.8045682873661264,
      "grad_norm": 0.4226018786430359,
      "learning_rate": 8.06390731762773e-05,
      "loss": 0.0439,
      "step": 1190
    },
    {
      "epoch": 1.8197327267557577,
      "grad_norm": 0.5903601050376892,
      "learning_rate": 8.021870159655109e-05,
      "loss": 0.0312,
      "step": 1200
    },
    {
      "epoch": 1.8197327267557577,
      "eval_loss": 0.037754252552986145,
      "eval_runtime": 206.9483,
      "eval_samples_per_second": 5.668,
      "eval_steps_per_second": 5.668,
      "step": 1200
    },
    {
      "epoch": 1.8348971661453892,
      "grad_norm": 0.2787196636199951,
      "learning_rate": 7.979493749939955e-05,
      "loss": 0.034,
      "step": 1210
    },
    {
      "epoch": 1.8500616055350205,
      "grad_norm": 0.23407863080501556,
      "learning_rate": 7.93678284589079e-05,
      "loss": 0.0418,
      "step": 1220
    },
    {
      "epoch": 1.8652260449246518,
      "grad_norm": 0.17310842871665955,
      "learning_rate": 7.893742242468301e-05,
      "loss": 0.0261,
      "step": 1230
    },
    {
      "epoch": 1.8803904843142831,
      "grad_norm": 0.7467524409294128,
      "learning_rate": 7.850376771647038e-05,
      "loss": 0.0406,
      "step": 1240
    },
    {
      "epoch": 1.8955549237039144,
      "grad_norm": 0.2921977639198303,
      "learning_rate": 7.806691301872958e-05,
      "loss": 0.0353,
      "step": 1250
    },
    {
      "epoch": 1.8955549237039144,
      "eval_loss": 0.03786836564540863,
      "eval_runtime": 208.333,
      "eval_samples_per_second": 5.63,
      "eval_steps_per_second": 5.63,
      "step": 1250
    },
    {
      "epoch": 1.9107193630935457,
      "grad_norm": 0.18109281361103058,
      "learning_rate": 7.762690737516846e-05,
      "loss": 0.0344,
      "step": 1260
    },
    {
      "epoch": 1.925883802483177,
      "grad_norm": 0.21812854707241058,
      "learning_rate": 7.718380018323742e-05,
      "loss": 0.0286,
      "step": 1270
    },
    {
      "epoch": 1.9410482418728083,
      "grad_norm": 0.3241231143474579,
      "learning_rate": 7.673764118858371e-05,
      "loss": 0.0287,
      "step": 1280
    },
    {
      "epoch": 1.9562126812624396,
      "grad_norm": 0.18561425805091858,
      "learning_rate": 7.628848047946675e-05,
      "loss": 0.0392,
      "step": 1290
    },
    {
      "epoch": 1.971377120652071,
      "grad_norm": 0.2008410096168518,
      "learning_rate": 7.583636848113483e-05,
      "loss": 0.0242,
      "step": 1300
    },
    {
      "epoch": 1.971377120652071,
      "eval_loss": 0.03601429983973503,
      "eval_runtime": 208.4866,
      "eval_samples_per_second": 5.626,
      "eval_steps_per_second": 5.626,
      "step": 1300
    },
    {
      "epoch": 1.9865415600417022,
      "grad_norm": 0.42017701268196106,
      "learning_rate": 7.538135595016423e-05,
      "loss": 0.0329,
      "step": 1310
    },
    {
      "epoch": 2.0017059994313335,
      "grad_norm": 0.23573864996433258,
      "learning_rate": 7.492349396876096e-05,
      "loss": 0.0248,
      "step": 1320
    },
    {
      "epoch": 2.016870438820965,
      "grad_norm": 0.12982603907585144,
      "learning_rate": 7.4462833939026e-05,
      "loss": 0.0258,
      "step": 1330
    },
    {
      "epoch": 2.032034878210596,
      "grad_norm": 0.20572739839553833,
      "learning_rate": 7.399942757718455e-05,
      "loss": 0.0179,
      "step": 1340
    },
    {
      "epoch": 2.0471993176002274,
      "grad_norm": 0.3729046881198883,
      "learning_rate": 7.35333269077802e-05,
      "loss": 0.0176,
      "step": 1350
    },
    {
      "epoch": 2.0471993176002274,
      "eval_loss": 0.04126604646444321,
      "eval_runtime": 207.5727,
      "eval_samples_per_second": 5.651,
      "eval_steps_per_second": 5.651,
      "step": 1350
    },
    {
      "epoch": 2.0623637569898587,
      "grad_norm": 0.27359482645988464,
      "learning_rate": 7.306458425783426e-05,
      "loss": 0.0309,
      "step": 1360
    },
    {
      "epoch": 2.07752819637949,
      "grad_norm": 0.2222670614719391,
      "learning_rate": 7.25932522509713e-05,
      "loss": 0.0191,
      "step": 1370
    },
    {
      "epoch": 2.0926926357691213,
      "grad_norm": 0.35448694229125977,
      "learning_rate": 7.211938380151133e-05,
      "loss": 0.0213,
      "step": 1380
    },
    {
      "epoch": 2.1078570751587526,
      "grad_norm": 0.29310059547424316,
      "learning_rate": 7.164303210852934e-05,
      "loss": 0.0245,
      "step": 1390
    },
    {
      "epoch": 2.123021514548384,
      "grad_norm": 0.2385050356388092,
      "learning_rate": 7.116425064988286e-05,
      "loss": 0.0132,
      "step": 1400
    },
    {
      "epoch": 2.123021514548384,
      "eval_loss": 0.038636982440948486,
      "eval_runtime": 208.1606,
      "eval_samples_per_second": 5.635,
      "eval_steps_per_second": 5.635,
      "step": 1400
    },
    {
      "epoch": 2.1381859539380152,
      "grad_norm": 0.2776142358779907,
      "learning_rate": 7.068309317620827e-05,
      "loss": 0.021,
      "step": 1410
    },
    {
      "epoch": 2.1533503933276466,
      "grad_norm": 0.3040802478790283,
      "learning_rate": 7.019961370488645e-05,
      "loss": 0.02,
      "step": 1420
    },
    {
      "epoch": 2.168514832717278,
      "grad_norm": 0.3864327073097229,
      "learning_rate": 6.971386651397849e-05,
      "loss": 0.0191,
      "step": 1430
    },
    {
      "epoch": 2.183679272106909,
      "grad_norm": 0.3276340663433075,
      "learning_rate": 6.922590613613211e-05,
      "loss": 0.017,
      "step": 1440
    },
    {
      "epoch": 2.1988437114965405,
      "grad_norm": 0.18806734681129456,
      "learning_rate": 6.873578735245961e-05,
      "loss": 0.0224,
      "step": 1450
    },
    {
      "epoch": 2.1988437114965405,
      "eval_loss": 0.04134543985128403,
      "eval_runtime": 208.1264,
      "eval_samples_per_second": 5.636,
      "eval_steps_per_second": 5.636,
      "step": 1450
    },
    {
      "epoch": 2.2140081508861718,
      "grad_norm": 0.1309497058391571,
      "learning_rate": 6.824356518638775e-05,
      "loss": 0.0188,
      "step": 1460
    },
    {
      "epoch": 2.229172590275803,
      "grad_norm": 0.2113678902387619,
      "learning_rate": 6.774929489748052e-05,
      "loss": 0.0169,
      "step": 1470
    },
    {
      "epoch": 2.2443370296654344,
      "grad_norm": 0.5515885353088379,
      "learning_rate": 6.725303197523548e-05,
      "loss": 0.0285,
      "step": 1480
    },
    {
      "epoch": 2.259501469055066,
      "grad_norm": 0.17094577848911285,
      "learning_rate": 6.675483213285412e-05,
      "loss": 0.019,
      "step": 1490
    },
    {
      "epoch": 2.274665908444697,
      "grad_norm": 0.4091382920742035,
      "learning_rate": 6.625475130098728e-05,
      "loss": 0.0198,
      "step": 1500
    },
    {
      "epoch": 2.274665908444697,
      "eval_loss": 0.042340174317359924,
      "eval_runtime": 207.7019,
      "eval_samples_per_second": 5.648,
      "eval_steps_per_second": 5.648,
      "step": 1500
    },
    {
      "epoch": 2.2898303478343287,
      "grad_norm": 0.26022475957870483,
      "learning_rate": 6.575284562145593e-05,
      "loss": 0.0198,
      "step": 1510
    },
    {
      "epoch": 2.30499478722396,
      "grad_norm": 0.46716049313545227,
      "learning_rate": 6.524917144094851e-05,
      "loss": 0.0213,
      "step": 1520
    },
    {
      "epoch": 2.3201592266135913,
      "grad_norm": 0.09904512763023376,
      "learning_rate": 6.474378530469509e-05,
      "loss": 0.0249,
      "step": 1530
    },
    {
      "epoch": 2.3353236660032226,
      "grad_norm": 0.35088425874710083,
      "learning_rate": 6.42367439501193e-05,
      "loss": 0.0149,
      "step": 1540
    },
    {
      "epoch": 2.350488105392854,
      "grad_norm": 0.38264918327331543,
      "learning_rate": 6.372810430046862e-05,
      "loss": 0.0191,
      "step": 1550
    },
    {
      "epoch": 2.350488105392854,
      "eval_loss": 0.04294043034315109,
      "eval_runtime": 206.9625,
      "eval_samples_per_second": 5.668,
      "eval_steps_per_second": 5.668,
      "step": 1550
    },
    {
      "epoch": 2.365652544782485,
      "grad_norm": 0.13157694041728973,
      "learning_rate": 6.321792345842402e-05,
      "loss": 0.0228,
      "step": 1560
    },
    {
      "epoch": 2.3808169841721165,
      "grad_norm": 0.2602044939994812,
      "learning_rate": 6.270625869968906e-05,
      "loss": 0.0264,
      "step": 1570
    },
    {
      "epoch": 2.395981423561748,
      "grad_norm": 0.20206056535243988,
      "learning_rate": 6.219316746656007e-05,
      "loss": 0.0243,
      "step": 1580
    },
    {
      "epoch": 2.411145862951379,
      "grad_norm": 0.3280366361141205,
      "learning_rate": 6.167870736147713e-05,
      "loss": 0.0233,
      "step": 1590
    },
    {
      "epoch": 2.4263103023410104,
      "grad_norm": 0.258645623922348,
      "learning_rate": 6.116293614055744e-05,
      "loss": 0.017,
      "step": 1600
    },
    {
      "epoch": 2.4263103023410104,
      "eval_loss": 0.041183970868587494,
      "eval_runtime": 208.1675,
      "eval_samples_per_second": 5.635,
      "eval_steps_per_second": 5.635,
      "step": 1600
    },
    {
      "epoch": 2.4414747417306417,
      "grad_norm": 0.29392683506011963,
      "learning_rate": 6.06459117071113e-05,
      "loss": 0.0177,
      "step": 1610
    },
    {
      "epoch": 2.456639181120273,
      "grad_norm": 0.24882225692272186,
      "learning_rate": 6.012769210514146e-05,
      "loss": 0.0164,
      "step": 1620
    },
    {
      "epoch": 2.4718036205099043,
      "grad_norm": 0.2122090458869934,
      "learning_rate": 5.9608335512826915e-05,
      "loss": 0.0197,
      "step": 1630
    },
    {
      "epoch": 2.4869680598995356,
      "grad_norm": 0.12246949225664139,
      "learning_rate": 5.908790023599144e-05,
      "loss": 0.0116,
      "step": 1640
    },
    {
      "epoch": 2.502132499289167,
      "grad_norm": 0.2111271470785141,
      "learning_rate": 5.856644470155781e-05,
      "loss": 0.0194,
      "step": 1650
    },
    {
      "epoch": 2.502132499289167,
      "eval_loss": 0.046471014618873596,
      "eval_runtime": 207.0102,
      "eval_samples_per_second": 5.666,
      "eval_steps_per_second": 5.666,
      "step": 1650
    },
    {
      "epoch": 2.5172969386787982,
      "grad_norm": 0.33697161078453064,
      "learning_rate": 5.8044027450988546e-05,
      "loss": 0.0242,
      "step": 1660
    },
    {
      "epoch": 2.5324613780684295,
      "grad_norm": 0.2361498922109604,
      "learning_rate": 5.752070713371371e-05,
      "loss": 0.0164,
      "step": 1670
    },
    {
      "epoch": 2.547625817458061,
      "grad_norm": 0.25567626953125,
      "learning_rate": 5.699654250054662e-05,
      "loss": 0.0153,
      "step": 1680
    },
    {
      "epoch": 2.562790256847692,
      "grad_norm": 0.24401170015335083,
      "learning_rate": 5.647159239708809e-05,
      "loss": 0.0132,
      "step": 1690
    },
    {
      "epoch": 2.5779546962373234,
      "grad_norm": 0.2683306038379669,
      "learning_rate": 5.5945915757120146e-05,
      "loss": 0.0178,
      "step": 1700
    },
    {
      "epoch": 2.5779546962373234,
      "eval_loss": 0.04394778236746788,
      "eval_runtime": 199.3527,
      "eval_samples_per_second": 5.884,
      "eval_steps_per_second": 5.884,
      "step": 1700
    },
    {
      "epoch": 2.5931191356269547,
      "grad_norm": 0.54051673412323,
      "learning_rate": 5.5419571595989825e-05,
      "loss": 0.0211,
      "step": 1710
    },
    {
      "epoch": 2.608283575016586,
      "grad_norm": 0.23906239867210388,
      "learning_rate": 5.4892619003983734e-05,
      "loss": 0.0173,
      "step": 1720
    },
    {
      "epoch": 2.6234480144062173,
      "grad_norm": 0.07637752592563629,
      "learning_rate": 5.436511713969428e-05,
      "loss": 0.0127,
      "step": 1730
    },
    {
      "epoch": 2.6386124537958486,
      "grad_norm": 0.07885332405567169,
      "learning_rate": 5.383712522337817e-05,
      "loss": 0.02,
      "step": 1740
    },
    {
      "epoch": 2.65377689318548,
      "grad_norm": 0.47538188099861145,
      "learning_rate": 5.3308702530308076e-05,
      "loss": 0.0238,
      "step": 1750
    },
    {
      "epoch": 2.65377689318548,
      "eval_loss": 0.041050076484680176,
      "eval_runtime": 195.9057,
      "eval_samples_per_second": 5.988,
      "eval_steps_per_second": 5.988,
      "step": 1750
    },
    {
      "epoch": 2.6689413325751112,
      "grad_norm": 0.32186976075172424,
      "learning_rate": 5.2779908384118025e-05,
      "loss": 0.0208,
      "step": 1760
    },
    {
      "epoch": 2.6841057719647425,
      "grad_norm": 0.5302672982215881,
      "learning_rate": 5.22508021501434e-05,
      "loss": 0.0242,
      "step": 1770
    },
    {
      "epoch": 2.699270211354374,
      "grad_norm": 0.28542810678482056,
      "learning_rate": 5.1721443228756284e-05,
      "loss": 0.0162,
      "step": 1780
    },
    {
      "epoch": 2.714434650744005,
      "grad_norm": 0.11375605314970016,
      "learning_rate": 5.119189104869683e-05,
      "loss": 0.0176,
      "step": 1790
    },
    {
      "epoch": 2.729599090133637,
      "grad_norm": 0.24023477733135223,
      "learning_rate": 5.066220506040148e-05,
      "loss": 0.0181,
      "step": 1800
    },
    {
      "epoch": 2.729599090133637,
      "eval_loss": 0.04137849807739258,
      "eval_runtime": 193.9483,
      "eval_samples_per_second": 6.048,
      "eval_steps_per_second": 6.048,
      "step": 1800
    },
    {
      "epoch": 2.7447635295232677,
      "grad_norm": 0.19859518110752106,
      "learning_rate": 5.013244472932872e-05,
      "loss": 0.0223,
      "step": 1810
    },
    {
      "epoch": 2.7599279689128995,
      "grad_norm": 0.32186752557754517,
      "learning_rate": 4.960266952928316e-05,
      "loss": 0.0193,
      "step": 1820
    },
    {
      "epoch": 2.7750924083025303,
      "grad_norm": 0.4007152318954468,
      "learning_rate": 4.907293893573867e-05,
      "loss": 0.0172,
      "step": 1830
    },
    {
      "epoch": 2.790256847692162,
      "grad_norm": 0.26739656925201416,
      "learning_rate": 4.8543312419161396e-05,
      "loss": 0.0164,
      "step": 1840
    },
    {
      "epoch": 2.805421287081793,
      "grad_norm": 0.2896629571914673,
      "learning_rate": 4.8013849438333165e-05,
      "loss": 0.0128,
      "step": 1850
    },
    {
      "epoch": 2.805421287081793,
      "eval_loss": 0.04389198124408722,
      "eval_runtime": 192.7356,
      "eval_samples_per_second": 6.086,
      "eval_steps_per_second": 6.086,
      "step": 1850
    },
    {
      "epoch": 2.8205857264714247,
      "grad_norm": 0.5798066854476929,
      "learning_rate": 4.748460943367643e-05,
      "loss": 0.0179,
      "step": 1860
    },
    {
      "epoch": 2.8357501658610555,
      "grad_norm": 0.3237079083919525,
      "learning_rate": 4.695565182058113e-05,
      "loss": 0.0286,
      "step": 1870
    },
    {
      "epoch": 2.8509146052506873,
      "grad_norm": 0.2319541871547699,
      "learning_rate": 4.642703598273431e-05,
      "loss": 0.0186,
      "step": 1880
    },
    {
      "epoch": 2.8660790446403186,
      "grad_norm": 0.32056906819343567,
      "learning_rate": 4.589882126545352e-05,
      "loss": 0.0187,
      "step": 1890
    },
    {
      "epoch": 2.88124348402995,
      "grad_norm": 0.27325454354286194,
      "learning_rate": 4.537106696902425e-05,
      "loss": 0.0287,
      "step": 1900
    },
    {
      "epoch": 2.88124348402995,
      "eval_loss": 0.04099781811237335,
      "eval_runtime": 191.609,
      "eval_samples_per_second": 6.122,
      "eval_steps_per_second": 6.122,
      "step": 1900
    },
    {
      "epoch": 2.896407923419581,
      "grad_norm": 1.0726760625839233,
      "learning_rate": 4.484383234204266e-05,
      "loss": 0.0261,
      "step": 1910
    },
    {
      "epoch": 2.9115723628092125,
      "grad_norm": 0.3901512026786804,
      "learning_rate": 4.4317176574763935e-05,
      "loss": 0.0208,
      "step": 1920
    },
    {
      "epoch": 2.926736802198844,
      "grad_norm": 0.19363577663898468,
      "learning_rate": 4.3791158792457334e-05,
      "loss": 0.0219,
      "step": 1930
    },
    {
      "epoch": 2.941901241588475,
      "grad_norm": 0.21966907382011414,
      "learning_rate": 4.3265838048768334e-05,
      "loss": 0.0194,
      "step": 1940
    },
    {
      "epoch": 2.9570656809781064,
      "grad_norm": 0.31120771169662476,
      "learning_rate": 4.274127331908915e-05,
      "loss": 0.0202,
      "step": 1950
    },
    {
      "epoch": 2.9570656809781064,
      "eval_loss": 0.041776422411203384,
      "eval_runtime": 190.7741,
      "eval_samples_per_second": 6.149,
      "eval_steps_per_second": 6.149,
      "step": 1950
    },
    {
      "epoch": 2.9722301203677377,
      "grad_norm": 0.149943545460701,
      "learning_rate": 4.221752349393758e-05,
      "loss": 0.0123,
      "step": 1960
    },
    {
      "epoch": 2.987394559757369,
      "grad_norm": 0.24167421460151672,
      "learning_rate": 4.1694647372345886e-05,
      "loss": 0.0241,
      "step": 1970
    },
    {
      "epoch": 3.0025589991470003,
      "grad_norm": 0.09040270000696182,
      "learning_rate": 4.117270365525946e-05,
      "loss": 0.0157,
      "step": 1980
    },
    {
      "epoch": 3.0177234385366316,
      "grad_norm": 0.05640542134642601,
      "learning_rate": 4.065175093894694e-05,
      "loss": 0.0106,
      "step": 1990
    },
    {
      "epoch": 3.032887877926263,
      "grad_norm": 0.1316869705915451,
      "learning_rate": 4.013184770842167e-05,
      "loss": 0.011,
      "step": 2000
    },
    {
      "epoch": 3.032887877926263,
      "eval_loss": 0.042993590235710144,
      "eval_runtime": 190.4338,
      "eval_samples_per_second": 6.16,
      "eval_steps_per_second": 6.16,
      "step": 2000
    },
    {
      "epoch": 3.048052317315894,
      "grad_norm": 0.09240477532148361,
      "learning_rate": 3.9613052330876064e-05,
      "loss": 0.0117,
      "step": 2010
    },
    {
      "epoch": 3.0632167567055255,
      "grad_norm": 0.5564406514167786,
      "learning_rate": 3.909542304912881e-05,
      "loss": 0.0114,
      "step": 2020
    },
    {
      "epoch": 3.078381196095157,
      "grad_norm": 0.06670588254928589,
      "learning_rate": 3.857901797508628e-05,
      "loss": 0.0131,
      "step": 2030
    },
    {
      "epoch": 3.093545635484788,
      "grad_norm": 0.21447496116161346,
      "learning_rate": 3.80638950832186e-05,
      "loss": 0.0111,
      "step": 2040
    },
    {
      "epoch": 3.1087100748744194,
      "grad_norm": 0.2611368000507355,
      "learning_rate": 3.7550112204051014e-05,
      "loss": 0.005,
      "step": 2050
    },
    {
      "epoch": 3.1087100748744194,
      "eval_loss": 0.048680998384952545,
      "eval_runtime": 190.3304,
      "eval_samples_per_second": 6.163,
      "eval_steps_per_second": 6.163,
      "step": 2050
    },
    {
      "epoch": 3.1238745142640507,
      "grad_norm": 0.06609263271093369,
      "learning_rate": 3.703772701767167e-05,
      "loss": 0.0058,
      "step": 2060
    },
    {
      "epoch": 3.139038953653682,
      "grad_norm": 0.1730610877275467,
      "learning_rate": 3.652679704725596e-05,
      "loss": 0.0053,
      "step": 2070
    },
    {
      "epoch": 3.1542033930433133,
      "grad_norm": 0.15803933143615723,
      "learning_rate": 3.601737965260882e-05,
      "loss": 0.0093,
      "step": 2080
    },
    {
      "epoch": 3.1693678324329446,
      "grad_norm": 0.16534733772277832,
      "learning_rate": 3.550953202372503e-05,
      "loss": 0.0046,
      "step": 2090
    },
    {
      "epoch": 3.184532271822576,
      "grad_norm": 0.25180765986442566,
      "learning_rate": 3.500331117436895e-05,
      "loss": 0.0045,
      "step": 2100
    },
    {
      "epoch": 3.184532271822576,
      "eval_loss": 0.0501817911863327,
      "eval_runtime": 190.1913,
      "eval_samples_per_second": 6.167,
      "eval_steps_per_second": 6.167,
      "step": 2100
    },
    {
      "epoch": 3.1996967112122072,
      "grad_norm": 0.143573597073555,
      "learning_rate": 3.4498773935673626e-05,
      "loss": 0.0034,
      "step": 2110
    },
    {
      "epoch": 3.2148611506018385,
      "grad_norm": 0.1678372025489807,
      "learning_rate": 3.399597694976081e-05,
      "loss": 0.0081,
      "step": 2120
    },
    {
      "epoch": 3.23002558999147,
      "grad_norm": 0.46680811047554016,
      "learning_rate": 3.349497666338187e-05,
      "loss": 0.0117,
      "step": 2130
    },
    {
      "epoch": 3.245190029381101,
      "grad_norm": 0.32169198989868164,
      "learning_rate": 3.299582932158085e-05,
      "loss": 0.0125,
      "step": 2140
    },
    {
      "epoch": 3.260354468770733,
      "grad_norm": 0.20941679179668427,
      "learning_rate": 3.2498590961379996e-05,
      "loss": 0.0072,
      "step": 2150
    },
    {
      "epoch": 3.260354468770733,
      "eval_loss": 0.04960820451378822,
      "eval_runtime": 189.4256,
      "eval_samples_per_second": 6.192,
      "eval_steps_per_second": 6.192,
      "step": 2150
    },
    {
      "epoch": 3.2755189081603637,
      "grad_norm": 0.17324012517929077,
      "learning_rate": 3.200331740548887e-05,
      "loss": 0.0083,
      "step": 2160
    },
    {
      "epoch": 3.2906833475499955,
      "grad_norm": 0.5363154411315918,
      "learning_rate": 3.1510064256037274e-05,
      "loss": 0.0136,
      "step": 2170
    },
    {
      "epoch": 3.3058477869396263,
      "grad_norm": 0.03524131327867508,
      "learning_rate": 3.1018886888333065e-05,
      "loss": 0.0091,
      "step": 2180
    },
    {
      "epoch": 3.321012226329258,
      "grad_norm": 0.24780172109603882,
      "learning_rate": 3.052984044464548e-05,
      "loss": 0.0088,
      "step": 2190
    },
    {
      "epoch": 3.3361766657188894,
      "grad_norm": 0.07167384773492813,
      "learning_rate": 3.0042979828014496e-05,
      "loss": 0.0098,
      "step": 2200
    },
    {
      "epoch": 3.3361766657188894,
      "eval_loss": 0.04823986068367958,
      "eval_runtime": 188.5523,
      "eval_samples_per_second": 6.221,
      "eval_steps_per_second": 6.221,
      "step": 2200
    },
    {
      "epoch": 3.3513411051085207,
      "grad_norm": 0.10964875668287277,
      "learning_rate": 2.9558359696087152e-05,
      "loss": 0.0114,
      "step": 2210
    },
    {
      "epoch": 3.366505544498152,
      "grad_norm": 0.4698887765407562,
      "learning_rate": 2.9076034454981332e-05,
      "loss": 0.0099,
      "step": 2220
    },
    {
      "epoch": 3.3816699838877833,
      "grad_norm": 0.4254396855831146,
      "learning_rate": 2.8596058253177932e-05,
      "loss": 0.0073,
      "step": 2230
    },
    {
      "epoch": 3.3968344232774146,
      "grad_norm": 0.22847697138786316,
      "learning_rate": 2.811848497544175e-05,
      "loss": 0.0109,
      "step": 2240
    },
    {
      "epoch": 3.411998862667046,
      "grad_norm": 0.08946174383163452,
      "learning_rate": 2.764336823677216e-05,
      "loss": 0.0089,
      "step": 2250
    },
    {
      "epoch": 3.411998862667046,
      "eval_loss": 0.049223024398088455,
      "eval_runtime": 188.3627,
      "eval_samples_per_second": 6.227,
      "eval_steps_per_second": 6.227,
      "step": 2250
    },
    {
      "epoch": 3.427163302056677,
      "grad_norm": 0.2524348497390747,
      "learning_rate": 2.717076137638388e-05,
      "loss": 0.0077,
      "step": 2260
    },
    {
      "epoch": 3.4423277414463085,
      "grad_norm": 0.3513559401035309,
      "learning_rate": 2.6700717451719066e-05,
      "loss": 0.013,
      "step": 2270
    },
    {
      "epoch": 3.45749218083594,
      "grad_norm": 0.35712215304374695,
      "learning_rate": 2.6233289232490447e-05,
      "loss": 0.0095,
      "step": 2280
    },
    {
      "epoch": 3.472656620225571,
      "grad_norm": 0.06787083297967911,
      "learning_rate": 2.5768529194757474e-05,
      "loss": 0.006,
      "step": 2290
    },
    {
      "epoch": 3.4878210596152024,
      "grad_norm": 0.028745301067829132,
      "learning_rate": 2.5306489515034713e-05,
      "loss": 0.0072,
      "step": 2300
    },
    {
      "epoch": 3.4878210596152024,
      "eval_loss": 0.04863268509507179,
      "eval_runtime": 188.4036,
      "eval_samples_per_second": 6.226,
      "eval_steps_per_second": 6.226,
      "step": 2300
    },
    {
      "epoch": 3.5029854990048337,
      "grad_norm": 0.26812678575515747,
      "learning_rate": 2.484722206443455e-05,
      "loss": 0.0088,
      "step": 2310
    },
    {
      "epoch": 3.518149938394465,
      "grad_norm": 0.04494244605302811,
      "learning_rate": 2.4390778402843605e-05,
      "loss": 0.0105,
      "step": 2320
    },
    {
      "epoch": 3.5333143777840963,
      "grad_norm": 0.2933897078037262,
      "learning_rate": 2.3937209773134472e-05,
      "loss": 0.0072,
      "step": 2330
    },
    {
      "epoch": 3.5484788171737276,
      "grad_norm": 0.32120901346206665,
      "learning_rate": 2.3486567095412864e-05,
      "loss": 0.0084,
      "step": 2340
    },
    {
      "epoch": 3.563643256563359,
      "grad_norm": 0.6362977027893066,
      "learning_rate": 2.3038900961301053e-05,
      "loss": 0.0116,
      "step": 2350
    },
    {
      "epoch": 3.563643256563359,
      "eval_loss": 0.04956178739666939,
      "eval_runtime": 188.2167,
      "eval_samples_per_second": 6.232,
      "eval_steps_per_second": 6.232,
      "step": 2350
    },
    {
      "epoch": 3.57880769595299,
      "grad_norm": 0.03769369050860405,
      "learning_rate": 2.259426162825819e-05,
      "loss": 0.0068,
      "step": 2360
    },
    {
      "epoch": 3.5939721353426215,
      "grad_norm": 0.4940986931324005,
      "learning_rate": 2.215269901393805e-05,
      "loss": 0.0122,
      "step": 2370
    },
    {
      "epoch": 3.609136574732253,
      "grad_norm": 0.11686809360980988,
      "learning_rate": 2.1714262690585192e-05,
      "loss": 0.007,
      "step": 2380
    },
    {
      "epoch": 3.624301014121884,
      "grad_norm": 0.11242389678955078,
      "learning_rate": 2.1279001879469424e-05,
      "loss": 0.0058,
      "step": 2390
    },
    {
      "epoch": 3.6394654535115154,
      "grad_norm": 0.25813475251197815,
      "learning_rate": 2.084696544536025e-05,
      "loss": 0.0094,
      "step": 2400
    },
    {
      "epoch": 3.6394654535115154,
      "eval_loss": 0.0489325188100338,
      "eval_runtime": 188.2757,
      "eval_samples_per_second": 6.23,
      "eval_steps_per_second": 6.23,
      "step": 2400
    },
    {
      "epoch": 3.6546298929011467,
      "grad_norm": 0.05024724453687668,
      "learning_rate": 2.0418201891040778e-05,
      "loss": 0.0066,
      "step": 2410
    },
    {
      "epoch": 3.669794332290778,
      "grad_norm": 0.07094204425811768,
      "learning_rate": 1.9992759351862772e-05,
      "loss": 0.015,
      "step": 2420
    },
    {
      "epoch": 3.6849587716804093,
      "grad_norm": 0.21597981452941895,
      "learning_rate": 1.9570685590342463e-05,
      "loss": 0.0071,
      "step": 2430
    },
    {
      "epoch": 3.7001232110700406,
      "grad_norm": 0.4047929644584656,
      "learning_rate": 1.9152027990798748e-05,
      "loss": 0.0092,
      "step": 2440
    },
    {
      "epoch": 3.715287650459672,
      "grad_norm": 1.1898387670516968,
      "learning_rate": 1.8736833554033263e-05,
      "loss": 0.0055,
      "step": 2450
    },
    {
      "epoch": 3.715287650459672,
      "eval_loss": 0.050134897232055664,
      "eval_runtime": 187.4638,
      "eval_samples_per_second": 6.257,
      "eval_steps_per_second": 6.257,
      "step": 2450
    },
    {
      "epoch": 3.7304520898493037,
      "grad_norm": 0.17783983051776886,
      "learning_rate": 1.8325148892054093e-05,
      "loss": 0.0068,
      "step": 2460
    },
    {
      "epoch": 3.7456165292389345,
      "grad_norm": 0.0921979695558548,
      "learning_rate": 1.7917020222842662e-05,
      "loss": 0.0066,
      "step": 2470
    },
    {
      "epoch": 3.7607809686285663,
      "grad_norm": 0.07260593771934509,
      "learning_rate": 1.751249336516513e-05,
      "loss": 0.0083,
      "step": 2480
    },
    {
      "epoch": 3.775945408018197,
      "grad_norm": 0.07632392644882202,
      "learning_rate": 1.7111613733428522e-05,
      "loss": 0.0078,
      "step": 2490
    },
    {
      "epoch": 3.791109847407829,
      "grad_norm": 0.1986379325389862,
      "learning_rate": 1.671442633258218e-05,
      "loss": 0.0095,
      "step": 2500
    },
    {
      "epoch": 3.791109847407829,
      "eval_loss": 0.052877526730298996,
      "eval_runtime": 187.1604,
      "eval_samples_per_second": 6.267,
      "eval_steps_per_second": 6.267,
      "step": 2500
    },
    {
      "epoch": 3.8062742867974597,
      "grad_norm": 0.33867862820625305,
      "learning_rate": 1.6320975753065453e-05,
      "loss": 0.0084,
      "step": 2510
    },
    {
      "epoch": 3.8214387261870915,
      "grad_norm": 0.5895967483520508,
      "learning_rate": 1.5931306165801452e-05,
      "loss": 0.0071,
      "step": 2520
    },
    {
      "epoch": 3.8366031655767223,
      "grad_norm": 0.15865935385227203,
      "learning_rate": 1.554546131723848e-05,
      "loss": 0.0064,
      "step": 2530
    },
    {
      "epoch": 3.851767604966354,
      "grad_norm": 0.7080567479133606,
      "learning_rate": 1.5163484524438516e-05,
      "loss": 0.0066,
      "step": 2540
    },
    {
      "epoch": 3.8669320443559854,
      "grad_norm": 0.34300854802131653,
      "learning_rate": 1.4785418670214496e-05,
      "loss": 0.0113,
      "step": 2550
    },
    {
      "epoch": 3.8669320443559854,
      "eval_loss": 0.05170406401157379,
      "eval_runtime": 187.6804,
      "eval_samples_per_second": 6.25,
      "eval_steps_per_second": 6.25,
      "step": 2550
    },
    {
      "epoch": 3.8820964837456167,
      "grad_norm": 0.03344808518886566,
      "learning_rate": 1.4411306198315777e-05,
      "loss": 0.0071,
      "step": 2560
    },
    {
      "epoch": 3.897260923135248,
      "grad_norm": 0.22047798335552216,
      "learning_rate": 1.4041189108663421e-05,
      "loss": 0.0072,
      "step": 2570
    },
    {
      "epoch": 3.9124253625248793,
      "grad_norm": 0.36693868041038513,
      "learning_rate": 1.3675108952634824e-05,
      "loss": 0.0041,
      "step": 2580
    },
    {
      "epoch": 3.9275898019145106,
      "grad_norm": 0.6838831901550293,
      "learning_rate": 1.3313106828399147e-05,
      "loss": 0.01,
      "step": 2590
    },
    {
      "epoch": 3.942754241304142,
      "grad_norm": 0.04084847494959831,
      "learning_rate": 1.2955223376303205e-05,
      "loss": 0.0042,
      "step": 2600
    },
    {
      "epoch": 3.942754241304142,
      "eval_loss": 0.05176025629043579,
      "eval_runtime": 187.7358,
      "eval_samples_per_second": 6.248,
      "eval_steps_per_second": 6.248,
      "step": 2600
    },
    {
      "epoch": 3.957918680693773,
      "grad_norm": 0.13362419605255127,
      "learning_rate": 1.2601498774309112e-05,
      "loss": 0.006,
      "step": 2610
    },
    {
      "epoch": 3.9730831200834045,
      "grad_norm": 0.21809060871601105,
      "learning_rate": 1.2251972733483612e-05,
      "loss": 0.0086,
      "step": 2620
    },
    {
      "epoch": 3.9882475594730358,
      "grad_norm": 0.6125285625457764,
      "learning_rate": 1.19066844935399e-05,
      "loss": 0.0048,
      "step": 2630
    },
    {
      "epoch": 4.003411998862667,
      "grad_norm": 0.09248276054859161,
      "learning_rate": 1.156567281843241e-05,
      "loss": 0.0059,
      "step": 2640
    },
    {
      "epoch": 4.018576438252298,
      "grad_norm": 0.17377474904060364,
      "learning_rate": 1.1228975992004842e-05,
      "loss": 0.0021,
      "step": 2650
    },
    {
      "epoch": 4.018576438252298,
      "eval_loss": 0.053896982222795486,
      "eval_runtime": 187.5167,
      "eval_samples_per_second": 6.255,
      "eval_steps_per_second": 6.255,
      "step": 2650
    },
    {
      "epoch": 4.03374087764193,
      "grad_norm": 0.09760443866252899,
      "learning_rate": 1.0896631813692377e-05,
      "loss": 0.0032,
      "step": 2660
    },
    {
      "epoch": 4.048905317031561,
      "grad_norm": 0.09826330095529556,
      "learning_rate": 1.0568677594277899e-05,
      "loss": 0.0035,
      "step": 2670
    },
    {
      "epoch": 4.064069756421192,
      "grad_norm": 0.029800906777381897,
      "learning_rate": 1.0245150151703509e-05,
      "loss": 0.0023,
      "step": 2680
    },
    {
      "epoch": 4.079234195810824,
      "grad_norm": 0.13375523686408997,
      "learning_rate": 9.926085806936918e-06,
      "loss": 0.0029,
      "step": 2690
    },
    {
      "epoch": 4.094398635200455,
      "grad_norm": 0.26318782567977905,
      "learning_rate": 9.611520379894068e-06,
      "loss": 0.0027,
      "step": 2700
    },
    {
      "epoch": 4.094398635200455,
      "eval_loss": 0.0573224201798439,
      "eval_runtime": 187.1068,
      "eval_samples_per_second": 6.269,
      "eval_steps_per_second": 6.269,
      "step": 2700
    },
    {
      "epoch": 4.109563074590087,
      "grad_norm": 0.11672820150852203,
      "learning_rate": 9.301489185417578e-06,
      "loss": 0.0036,
      "step": 2710
    },
    {
      "epoch": 4.1247275139797175,
      "grad_norm": 0.06662098318338394,
      "learning_rate": 8.996027029312314e-06,
      "loss": 0.0018,
      "step": 2720
    },
    {
      "epoch": 4.139891953369349,
      "grad_norm": 0.18935194611549377,
      "learning_rate": 8.695168204437721e-06,
      "loss": 0.0065,
      "step": 2730
    },
    {
      "epoch": 4.15505639275898,
      "grad_norm": 0.212269127368927,
      "learning_rate": 8.398946486858029e-06,
      "loss": 0.0027,
      "step": 2740
    },
    {
      "epoch": 4.170220832148612,
      "grad_norm": 0.0438799224793911,
      "learning_rate": 8.107395132050333e-06,
      "loss": 0.0017,
      "step": 2750
    },
    {
      "epoch": 4.170220832148612,
      "eval_loss": 0.058968957513570786,
      "eval_runtime": 186.8676,
      "eval_samples_per_second": 6.277,
      "eval_steps_per_second": 6.277,
      "step": 2750
    },
    {
      "epoch": 4.185385271538243,
      "grad_norm": 0.09194748103618622,
      "learning_rate": 7.820546871171142e-06,
      "loss": 0.0034,
      "step": 2760
    },
    {
      "epoch": 4.200549710927874,
      "grad_norm": 0.07000330835580826,
      "learning_rate": 7.5384339073818145e-06,
      "loss": 0.0031,
      "step": 2770
    },
    {
      "epoch": 4.215714150317505,
      "grad_norm": 0.0696445107460022,
      "learning_rate": 7.261087912233228e-06,
      "loss": 0.0047,
      "step": 2780
    },
    {
      "epoch": 4.230878589707137,
      "grad_norm": 0.15083633363246918,
      "learning_rate": 6.988540022110235e-06,
      "loss": 0.0033,
      "step": 2790
    },
    {
      "epoch": 4.246043029096768,
      "grad_norm": 0.18565095961093903,
      "learning_rate": 6.720820834735969e-06,
      "loss": 0.0033,
      "step": 2800
    },
    {
      "epoch": 4.246043029096768,
      "eval_loss": 0.060299091041088104,
      "eval_runtime": 186.5266,
      "eval_samples_per_second": 6.289,
      "eval_steps_per_second": 6.289,
      "step": 2800
    },
    {
      "epoch": 4.2612074684864,
      "grad_norm": 0.2135116308927536,
      "learning_rate": 6.457960405736968e-06,
      "loss": 0.0052,
      "step": 2810
    },
    {
      "epoch": 4.2763719078760305,
      "grad_norm": 0.030328113585710526,
      "learning_rate": 6.199988245268778e-06,
      "loss": 0.0019,
      "step": 2820
    },
    {
      "epoch": 4.291536347265662,
      "grad_norm": 0.35592690110206604,
      "learning_rate": 5.946933314703157e-06,
      "loss": 0.0017,
      "step": 2830
    },
    {
      "epoch": 4.306700786655293,
      "grad_norm": 0.027882764115929604,
      "learning_rate": 5.698824023376531e-06,
      "loss": 0.002,
      "step": 2840
    },
    {
      "epoch": 4.321865226044925,
      "grad_norm": 0.12716230750083923,
      "learning_rate": 5.455688225400802e-06,
      "loss": 0.003,
      "step": 2850
    },
    {
      "epoch": 4.321865226044925,
      "eval_loss": 0.06180414929986,
      "eval_runtime": 186.8078,
      "eval_samples_per_second": 6.279,
      "eval_steps_per_second": 6.279,
      "step": 2850
    },
    {
      "epoch": 4.337029665434556,
      "grad_norm": 0.05450417846441269,
      "learning_rate": 5.217553216536098e-06,
      "loss": 0.0024,
      "step": 2860
    },
    {
      "epoch": 4.3521941048241874,
      "grad_norm": 0.2045290321111679,
      "learning_rate": 4.984445731126542e-06,
      "loss": 0.0045,
      "step": 2870
    },
    {
      "epoch": 4.367358544213818,
      "grad_norm": 0.022383777424693108,
      "learning_rate": 4.756391939098853e-06,
      "loss": 0.0013,
      "step": 2880
    },
    {
      "epoch": 4.38252298360345,
      "grad_norm": 0.011135186068713665,
      "learning_rate": 4.533417443024374e-06,
      "loss": 0.0059,
      "step": 2890
    },
    {
      "epoch": 4.397687422993081,
      "grad_norm": 0.04015190526843071,
      "learning_rate": 4.315547275244769e-06,
      "loss": 0.0013,
      "step": 2900
    },
    {
      "epoch": 4.397687422993081,
      "eval_loss": 0.0623171404004097,
      "eval_runtime": 186.7545,
      "eval_samples_per_second": 6.281,
      "eval_steps_per_second": 6.281,
      "step": 2900
    },
    {
      "epoch": 4.412851862382713,
      "grad_norm": 0.21536949276924133,
      "learning_rate": 4.10280589506179e-06,
      "loss": 0.0035,
      "step": 2910
    },
    {
      "epoch": 4.4280163017723435,
      "grad_norm": 0.08297796547412872,
      "learning_rate": 3.895217185991312e-06,
      "loss": 0.0021,
      "step": 2920
    },
    {
      "epoch": 4.443180741161975,
      "grad_norm": 0.25166183710098267,
      "learning_rate": 3.692804453082038e-06,
      "loss": 0.0056,
      "step": 2930
    },
    {
      "epoch": 4.458345180551606,
      "grad_norm": 0.12861423194408417,
      "learning_rate": 3.495590420299194e-06,
      "loss": 0.0016,
      "step": 2940
    },
    {
      "epoch": 4.473509619941238,
      "grad_norm": 0.8641133308410645,
      "learning_rate": 3.3035972279733207e-06,
      "loss": 0.003,
      "step": 2950
    },
    {
      "epoch": 4.473509619941238,
      "eval_loss": 0.06252659857273102,
      "eval_runtime": 186.6625,
      "eval_samples_per_second": 6.284,
      "eval_steps_per_second": 6.284,
      "step": 2950
    },
    {
      "epoch": 4.488674059330869,
      "grad_norm": 0.05727919191122055,
      "learning_rate": 3.116846430314796e-06,
      "loss": 0.0016,
      "step": 2960
    },
    {
      "epoch": 4.5038384987205005,
      "grad_norm": 0.019385578110814095,
      "learning_rate": 2.9353589929939087e-06,
      "loss": 0.0011,
      "step": 2970
    },
    {
      "epoch": 4.519002938110132,
      "grad_norm": 0.07699675858020782,
      "learning_rate": 2.7591552907872574e-06,
      "loss": 0.0022,
      "step": 2980
    },
    {
      "epoch": 4.534167377499763,
      "grad_norm": 0.04615270346403122,
      "learning_rate": 2.5882551052902883e-06,
      "loss": 0.0035,
      "step": 2990
    },
    {
      "epoch": 4.549331816889394,
      "grad_norm": 0.5508703589439392,
      "learning_rate": 2.4226776226965453e-06,
      "loss": 0.0036,
      "step": 3000
    },
    {
      "epoch": 4.549331816889394,
      "eval_loss": 0.06311533600091934,
      "eval_runtime": 186.2603,
      "eval_samples_per_second": 6.298,
      "eval_steps_per_second": 6.298,
      "step": 3000
    },
    {
      "epoch": 4.564496256279026,
      "grad_norm": 0.025140024721622467,
      "learning_rate": 2.262441431643697e-06,
      "loss": 0.0062,
      "step": 3010
    },
    {
      "epoch": 4.579660695668657,
      "grad_norm": 0.018112068995833397,
      "learning_rate": 2.1075645211266927e-06,
      "loss": 0.001,
      "step": 3020
    },
    {
      "epoch": 4.594825135058288,
      "grad_norm": 0.22902579605579376,
      "learning_rate": 1.9580642784782053e-06,
      "loss": 0.0028,
      "step": 3030
    },
    {
      "epoch": 4.60998957444792,
      "grad_norm": 0.021765662357211113,
      "learning_rate": 1.813957487416651e-06,
      "loss": 0.0011,
      "step": 3040
    },
    {
      "epoch": 4.625154013837551,
      "grad_norm": 0.30804678797721863,
      "learning_rate": 1.6752603261619315e-06,
      "loss": 0.0017,
      "step": 3050
    },
    {
      "epoch": 4.625154013837551,
      "eval_loss": 0.06340394914150238,
      "eval_runtime": 186.2512,
      "eval_samples_per_second": 6.298,
      "eval_steps_per_second": 6.298,
      "step": 3050
    },
    {
      "epoch": 4.640318453227183,
      "grad_norm": 0.058663006871938705,
      "learning_rate": 1.541988365619207e-06,
      "loss": 0.0043,
      "step": 3060
    },
    {
      "epoch": 4.6554828926168135,
      "grad_norm": 0.22114038467407227,
      "learning_rate": 1.4141565676307865e-06,
      "loss": 0.0033,
      "step": 3070
    },
    {
      "epoch": 4.670647332006445,
      "grad_norm": 0.013990496285259724,
      "learning_rate": 1.291779283296468e-06,
      "loss": 0.0018,
      "step": 3080
    },
    {
      "epoch": 4.685811771396076,
      "grad_norm": 0.06152665242552757,
      "learning_rate": 1.1748702513623922e-06,
      "loss": 0.0027,
      "step": 3090
    },
    {
      "epoch": 4.700976210785708,
      "grad_norm": 0.13100837171077728,
      "learning_rate": 1.0634425966786155e-06,
      "loss": 0.0023,
      "step": 3100
    },
    {
      "epoch": 4.700976210785708,
      "eval_loss": 0.06349016726016998,
      "eval_runtime": 186.2271,
      "eval_samples_per_second": 6.299,
      "eval_steps_per_second": 6.299,
      "step": 3100
    },
    {
      "epoch": 4.716140650175339,
      "grad_norm": 0.24262085556983948,
      "learning_rate": 9.575088287257118e-07,
      "loss": 0.0025,
      "step": 3110
    },
    {
      "epoch": 4.73130508956497,
      "grad_norm": 0.5570153594017029,
      "learning_rate": 8.570808402103392e-07,
      "loss": 0.0051,
      "step": 3120
    },
    {
      "epoch": 4.746469528954601,
      "grad_norm": 0.22754321992397308,
      "learning_rate": 7.62169905730109e-07,
      "loss": 0.0032,
      "step": 3130
    },
    {
      "epoch": 4.761633968344233,
      "grad_norm": 0.4912855625152588,
      "learning_rate": 6.727866805078531e-07,
      "loss": 0.0021,
      "step": 3140
    },
    {
      "epoch": 4.776798407733864,
      "grad_norm": 0.017126476392149925,
      "learning_rate": 5.889411991953975e-07,
      "loss": 0.0028,
      "step": 3150
    },
    {
      "epoch": 4.776798407733864,
      "eval_loss": 0.06346834450960159,
      "eval_runtime": 186.8935,
      "eval_samples_per_second": 6.276,
      "eval_steps_per_second": 6.276,
      "step": 3150
    },
    {
      "epoch": 4.791962847123496,
      "grad_norm": 0.2492581158876419,
      "learning_rate": 5.106428747470137e-07,
      "loss": 0.0021,
      "step": 3160
    },
    {
      "epoch": 4.8071272865131265,
      "grad_norm": 0.10539235919713974,
      "learning_rate": 4.37900497362681e-07,
      "loss": 0.0027,
      "step": 3170
    },
    {
      "epoch": 4.822291725902758,
      "grad_norm": 0.04554608091711998,
      "learning_rate": 3.7072223350124167e-07,
      "loss": 0.0021,
      "step": 3180
    },
    {
      "epoch": 4.837456165292389,
      "grad_norm": 0.2119264155626297,
      "learning_rate": 3.0911562496358517e-07,
      "loss": 0.0032,
      "step": 3190
    },
    {
      "epoch": 4.852620604682021,
      "grad_norm": 0.1288806051015854,
      "learning_rate": 2.53087588045986e-07,
      "loss": 0.0028,
      "step": 3200
    },
    {
      "epoch": 4.852620604682021,
      "eval_loss": 0.06373777985572815,
      "eval_runtime": 186.9337,
      "eval_samples_per_second": 6.275,
      "eval_steps_per_second": 6.275,
      "step": 3200
    },
    {
      "epoch": 4.867785044071652,
      "grad_norm": 0.12755931913852692,
      "learning_rate": 2.026444127636029e-07,
      "loss": 0.0023,
      "step": 3210
    },
    {
      "epoch": 4.882949483461283,
      "grad_norm": 0.00676093390211463,
      "learning_rate": 1.577917621443825e-07,
      "loss": 0.0038,
      "step": 3220
    },
    {
      "epoch": 4.898113922850914,
      "grad_norm": 0.26253750920295715,
      "learning_rate": 1.185346715932345e-07,
      "loss": 0.005,
      "step": 3230
    },
    {
      "epoch": 4.913278362240546,
      "grad_norm": 0.42117446660995483,
      "learning_rate": 8.48775483267783e-08,
      "loss": 0.0029,
      "step": 3240
    },
    {
      "epoch": 4.928442801630177,
      "grad_norm": 0.05510617420077324,
      "learning_rate": 5.6824170878544414e-08,
      "loss": 0.0021,
      "step": 3250
    },
    {
      "epoch": 4.928442801630177,
      "eval_loss": 0.06360332667827606,
      "eval_runtime": 186.8974,
      "eval_samples_per_second": 6.276,
      "eval_steps_per_second": 6.276,
      "step": 3250
    },
    {
      "epoch": 4.943607241019809,
      "grad_norm": 0.033535584807395935,
      "learning_rate": 3.4377688674774866e-08,
      "loss": 0.002,
      "step": 3260
    },
    {
      "epoch": 4.9587716804094395,
      "grad_norm": 0.037935420870780945,
      "learning_rate": 1.7540621680850466e-08,
      "loss": 0.0012,
      "step": 3270
    },
    {
      "epoch": 4.973936119799071,
      "grad_norm": 0.057161327451467514,
      "learning_rate": 6.314860118400434e-09,
      "loss": 0.0038,
      "step": 3280
    },
    {
      "epoch": 4.989100559188703,
      "grad_norm": 0.027261769399046898,
      "learning_rate": 7.016642530777162e-10,
      "loss": 0.0017,
      "step": 3290
    },
    {
      "epoch": 4.996682778883518,
      "step": 3295,
      "total_flos": 8.269502901505229e+17,
      "train_loss": 0.04054494600613038,
      "train_runtime": 44308.6539,
      "train_samples_per_second": 1.191,
      "train_steps_per_second": 0.074
    }
  ],
  "logging_steps": 10,
  "max_steps": 3295,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 50,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 8.269502901505229e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}