{
  "best_metric": 0.03357379883527756,
  "best_model_checkpoint": "saves/psy-course/Llama-3.1-8B-Instruct/train/fold5/checkpoint-1700",
  "epoch": 4.9971563981042655,
  "eval_steps": 50,
  "global_step": 3295,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.015165876777251185,
      "grad_norm": 4.091175556182861,
      "learning_rate": 3.0303030303030305e-06,
      "loss": 1.5168,
      "step": 10
    },
    {
      "epoch": 0.03033175355450237,
      "grad_norm": 4.17509651184082,
      "learning_rate": 6.060606060606061e-06,
      "loss": 1.4858,
      "step": 20
    },
    {
      "epoch": 0.04549763033175355,
      "grad_norm": 6.30861234664917,
      "learning_rate": 9.090909090909091e-06,
      "loss": 1.384,
      "step": 30
    },
    {
      "epoch": 0.06066350710900474,
      "grad_norm": 1.8290327787399292,
      "learning_rate": 1.2121212121212122e-05,
      "loss": 0.8608,
      "step": 40
    },
    {
      "epoch": 0.07582938388625593,
      "grad_norm": 1.4423801898956299,
      "learning_rate": 1.5151515151515153e-05,
      "loss": 0.6474,
      "step": 50
    },
    {
      "epoch": 0.07582938388625593,
      "eval_loss": 0.41726887226104736,
      "eval_runtime": 156.9909,
      "eval_samples_per_second": 7.472,
      "eval_steps_per_second": 7.472,
      "step": 50
    },
    {
      "epoch": 0.0909952606635071,
      "grad_norm": 1.9544239044189453,
      "learning_rate": 1.8181818181818182e-05,
      "loss": 0.4223,
      "step": 60
    },
    {
      "epoch": 0.1061611374407583,
      "grad_norm": 1.2844865322113037,
      "learning_rate": 2.1212121212121215e-05,
      "loss": 0.2419,
      "step": 70
    },
    {
      "epoch": 0.12132701421800948,
      "grad_norm": 1.459299921989441,
      "learning_rate": 2.4242424242424244e-05,
      "loss": 0.1873,
      "step": 80
    },
    {
      "epoch": 0.13649289099526066,
      "grad_norm": 0.775138795375824,
      "learning_rate": 2.7272727272727273e-05,
      "loss": 0.1012,
      "step": 90
    },
    {
      "epoch": 0.15165876777251186,
      "grad_norm": 1.243545413017273,
      "learning_rate": 3.0303030303030306e-05,
      "loss": 0.0891,
      "step": 100
    },
    {
      "epoch": 0.15165876777251186,
      "eval_loss": 0.09317315369844437,
      "eval_runtime": 157.6318,
      "eval_samples_per_second": 7.441,
      "eval_steps_per_second": 7.441,
      "step": 100
    },
    {
      "epoch": 0.16682464454976303,
      "grad_norm": 1.063110589981079,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.104,
      "step": 110
    },
    {
      "epoch": 0.1819905213270142,
      "grad_norm": 1.3549401760101318,
      "learning_rate": 3.6363636363636364e-05,
      "loss": 0.0783,
      "step": 120
    },
    {
      "epoch": 0.1971563981042654,
      "grad_norm": 0.8954592943191528,
      "learning_rate": 3.939393939393939e-05,
      "loss": 0.0678,
      "step": 130
    },
    {
      "epoch": 0.2123222748815166,
      "grad_norm": 0.6968070268630981,
      "learning_rate": 4.242424242424243e-05,
      "loss": 0.0687,
      "step": 140
    },
    {
      "epoch": 0.22748815165876776,
      "grad_norm": 1.6180505752563477,
      "learning_rate": 4.545454545454546e-05,
      "loss": 0.0787,
      "step": 150
    },
    {
      "epoch": 0.22748815165876776,
      "eval_loss": 0.08598122000694275,
      "eval_runtime": 157.9467,
      "eval_samples_per_second": 7.427,
      "eval_steps_per_second": 7.427,
      "step": 150
    },
    {
      "epoch": 0.24265402843601896,
      "grad_norm": 1.3334227800369263,
      "learning_rate": 4.848484848484849e-05,
      "loss": 0.0884,
      "step": 160
    },
    {
      "epoch": 0.25781990521327014,
      "grad_norm": 1.23464035987854,
      "learning_rate": 5.151515151515152e-05,
      "loss": 0.0874,
      "step": 170
    },
    {
      "epoch": 0.2729857819905213,
      "grad_norm": 0.6125182509422302,
      "learning_rate": 5.4545454545454546e-05,
      "loss": 0.0667,
      "step": 180
    },
    {
      "epoch": 0.2881516587677725,
      "grad_norm": 0.7370634078979492,
      "learning_rate": 5.757575757575758e-05,
      "loss": 0.0557,
      "step": 190
    },
    {
      "epoch": 0.3033175355450237,
      "grad_norm": 0.6302613019943237,
      "learning_rate": 6.060606060606061e-05,
      "loss": 0.059,
      "step": 200
    },
    {
      "epoch": 0.3033175355450237,
      "eval_loss": 0.05934415012598038,
      "eval_runtime": 158.0337,
      "eval_samples_per_second": 7.422,
      "eval_steps_per_second": 7.422,
      "step": 200
    },
    {
      "epoch": 0.3184834123222749,
      "grad_norm": 0.8791934847831726,
      "learning_rate": 6.363636363636364e-05,
      "loss": 0.0652,
      "step": 210
    },
    {
      "epoch": 0.33364928909952607,
      "grad_norm": 0.3848883807659149,
      "learning_rate": 6.666666666666667e-05,
      "loss": 0.0756,
      "step": 220
    },
    {
      "epoch": 0.34881516587677724,
      "grad_norm": 0.8282119631767273,
      "learning_rate": 6.96969696969697e-05,
      "loss": 0.0479,
      "step": 230
    },
    {
      "epoch": 0.3639810426540284,
      "grad_norm": 0.6479150056838989,
      "learning_rate": 7.272727272727273e-05,
      "loss": 0.0461,
      "step": 240
    },
    {
      "epoch": 0.3791469194312796,
      "grad_norm": 0.7077862620353699,
      "learning_rate": 7.575757575757576e-05,
      "loss": 0.053,
      "step": 250
    },
    {
      "epoch": 0.3791469194312796,
      "eval_loss": 0.05168557167053223,
      "eval_runtime": 158.1878,
      "eval_samples_per_second": 7.415,
      "eval_steps_per_second": 7.415,
      "step": 250
    },
    {
      "epoch": 0.3943127962085308,
      "grad_norm": 1.059857726097107,
      "learning_rate": 7.878787878787879e-05,
      "loss": 0.0578,
      "step": 260
    },
    {
      "epoch": 0.409478672985782,
      "grad_norm": 1.0045078992843628,
      "learning_rate": 8.181818181818183e-05,
      "loss": 0.0458,
      "step": 270
    },
    {
      "epoch": 0.4246445497630332,
      "grad_norm": 0.5126291513442993,
      "learning_rate": 8.484848484848486e-05,
      "loss": 0.0522,
      "step": 280
    },
    {
      "epoch": 0.43981042654028435,
      "grad_norm": 0.7032033205032349,
      "learning_rate": 8.787878787878789e-05,
      "loss": 0.0678,
      "step": 290
    },
    {
      "epoch": 0.4549763033175355,
      "grad_norm": 0.3473586142063141,
      "learning_rate": 9.090909090909092e-05,
      "loss": 0.0489,
      "step": 300
    },
    {
      "epoch": 0.4549763033175355,
      "eval_loss": 0.0462711863219738,
      "eval_runtime": 158.3039,
      "eval_samples_per_second": 7.41,
      "eval_steps_per_second": 7.41,
      "step": 300
    },
    {
      "epoch": 0.47014218009478675,
      "grad_norm": 0.475627601146698,
      "learning_rate": 9.393939393939395e-05,
      "loss": 0.0535,
      "step": 310
    },
    {
      "epoch": 0.48530805687203793,
      "grad_norm": 0.42610666155815125,
      "learning_rate": 9.696969696969698e-05,
      "loss": 0.0439,
      "step": 320
    },
    {
      "epoch": 0.500473933649289,
      "grad_norm": 0.5180782675743103,
      "learning_rate": 0.0001,
      "loss": 0.0577,
      "step": 330
    },
    {
      "epoch": 0.5156398104265403,
      "grad_norm": 0.3036288022994995,
      "learning_rate": 9.999719336268101e-05,
      "loss": 0.0554,
      "step": 340
    },
    {
      "epoch": 0.5308056872037915,
      "grad_norm": 1.1450015306472778,
      "learning_rate": 9.998877376581251e-05,
      "loss": 0.0525,
      "step": 350
    },
    {
      "epoch": 0.5308056872037915,
      "eval_loss": 0.047111984342336655,
      "eval_runtime": 158.1613,
      "eval_samples_per_second": 7.416,
      "eval_steps_per_second": 7.416,
      "step": 350
    },
    {
      "epoch": 0.5459715639810426,
      "grad_norm": 0.5295766592025757,
      "learning_rate": 9.997474215462472e-05,
      "loss": 0.0485,
      "step": 360
    },
    {
      "epoch": 0.5611374407582939,
      "grad_norm": 0.3777503967285156,
      "learning_rate": 9.995510010438337e-05,
      "loss": 0.0308,
      "step": 370
    },
    {
      "epoch": 0.576303317535545,
      "grad_norm": 0.3923189342021942,
      "learning_rate": 9.992984982021295e-05,
      "loss": 0.0618,
      "step": 380
    },
    {
      "epoch": 0.5914691943127962,
      "grad_norm": 0.6084585189819336,
      "learning_rate": 9.9898994136849e-05,
      "loss": 0.0675,
      "step": 390
    },
    {
      "epoch": 0.6066350710900474,
      "grad_norm": 0.5436043739318848,
      "learning_rate": 9.986253651832005e-05,
      "loss": 0.0651,
      "step": 400
    },
    {
      "epoch": 0.6066350710900474,
      "eval_loss": 0.04721478745341301,
      "eval_runtime": 158.1808,
      "eval_samples_per_second": 7.416,
      "eval_steps_per_second": 7.416,
      "step": 400
    },
    {
      "epoch": 0.6218009478672986,
      "grad_norm": 0.2049849033355713,
      "learning_rate": 9.982048105755859e-05,
      "loss": 0.0427,
      "step": 410
    },
    {
      "epoch": 0.6369668246445498,
      "grad_norm": 0.39465782046318054,
      "learning_rate": 9.977283247594166e-05,
      "loss": 0.0367,
      "step": 420
    },
    {
      "epoch": 0.6521327014218009,
      "grad_norm": 0.46949708461761475,
      "learning_rate": 9.971959612276076e-05,
      "loss": 0.0602,
      "step": 430
    },
    {
      "epoch": 0.6672985781990521,
      "grad_norm": 0.47874099016189575,
      "learning_rate": 9.966077797462129e-05,
      "loss": 0.0444,
      "step": 440
    },
    {
      "epoch": 0.6824644549763034,
      "grad_norm": 1.0055354833602905,
      "learning_rate": 9.959638463477165e-05,
      "loss": 0.0552,
      "step": 450
    },
    {
      "epoch": 0.6824644549763034,
      "eval_loss": 0.046703532338142395,
      "eval_runtime": 158.1664,
      "eval_samples_per_second": 7.416,
      "eval_steps_per_second": 7.416,
      "step": 450
    },
    {
      "epoch": 0.6976303317535545,
      "grad_norm": 0.25800299644470215,
      "learning_rate": 9.952642333236186e-05,
      "loss": 0.0474,
      "step": 460
    },
    {
      "epoch": 0.7127962085308057,
      "grad_norm": 0.518119752407074,
      "learning_rate": 9.945090192163202e-05,
      "loss": 0.0496,
      "step": 470
    },
    {
      "epoch": 0.7279620853080568,
      "grad_norm": 0.4063582420349121,
      "learning_rate": 9.936982888103051e-05,
      "loss": 0.0387,
      "step": 480
    },
    {
      "epoch": 0.7431279620853081,
      "grad_norm": 0.2823903560638428,
      "learning_rate": 9.928321331226219e-05,
      "loss": 0.0609,
      "step": 490
    },
    {
      "epoch": 0.7582938388625592,
      "grad_norm": 0.3049658238887787,
      "learning_rate": 9.919106493926655e-05,
      "loss": 0.0317,
      "step": 500
    },
    {
      "epoch": 0.7582938388625592,
      "eval_loss": 0.04159977287054062,
      "eval_runtime": 158.2617,
      "eval_samples_per_second": 7.412,
      "eval_steps_per_second": 7.412,
      "step": 500
    },
    {
      "epoch": 0.7734597156398104,
      "grad_norm": 0.48901668190956116,
      "learning_rate": 9.909339410712612e-05,
      "loss": 0.0484,
      "step": 510
    },
    {
      "epoch": 0.7886255924170616,
      "grad_norm": 0.21568810939788818,
      "learning_rate": 9.8990211780905e-05,
      "loss": 0.0369,
      "step": 520
    },
    {
      "epoch": 0.8037914691943128,
      "grad_norm": 0.6088463664054871,
      "learning_rate": 9.888152954441785e-05,
      "loss": 0.0523,
      "step": 530
    },
    {
      "epoch": 0.818957345971564,
      "grad_norm": 0.3415836989879608,
      "learning_rate": 9.876735959892953e-05,
      "loss": 0.043,
      "step": 540
    },
    {
      "epoch": 0.8341232227488151,
      "grad_norm": 0.22671574354171753,
      "learning_rate": 9.864771476178522e-05,
      "loss": 0.0313,
      "step": 550
    },
    {
      "epoch": 0.8341232227488151,
      "eval_loss": 0.042550310492515564,
      "eval_runtime": 158.3505,
      "eval_samples_per_second": 7.408,
      "eval_steps_per_second": 7.408,
      "step": 550
    },
    {
      "epoch": 0.8492890995260663,
      "grad_norm": 0.27584120631217957,
      "learning_rate": 9.852260846497153e-05,
      "loss": 0.0409,
      "step": 560
    },
    {
      "epoch": 0.8644549763033176,
      "grad_norm": 0.3367135524749756,
      "learning_rate": 9.839205475360851e-05,
      "loss": 0.0488,
      "step": 570
    },
    {
      "epoch": 0.8796208530805687,
      "grad_norm": 0.33391615748405457,
      "learning_rate": 9.825606828437291e-05,
      "loss": 0.0491,
      "step": 580
    },
    {
      "epoch": 0.8947867298578199,
      "grad_norm": 0.3055145740509033,
      "learning_rate": 9.811466432385267e-05,
      "loss": 0.0566,
      "step": 590
    },
    {
      "epoch": 0.909952606635071,
      "grad_norm": 0.18712455034255981,
      "learning_rate": 9.796785874683314e-05,
      "loss": 0.0441,
      "step": 600
    },
    {
      "epoch": 0.909952606635071,
      "eval_loss": 0.04294722527265549,
      "eval_runtime": 158.3343,
      "eval_samples_per_second": 7.408,
      "eval_steps_per_second": 7.408,
      "step": 600
    },
    {
      "epoch": 0.9251184834123223,
      "grad_norm": 0.17119909822940826,
      "learning_rate": 9.781566803451475e-05,
      "loss": 0.0363,
      "step": 610
    },
    {
      "epoch": 0.9402843601895735,
      "grad_norm": 0.3182626962661743,
      "learning_rate": 9.765810927266281e-05,
      "loss": 0.0325,
      "step": 620
    },
    {
      "epoch": 0.9554502369668246,
      "grad_norm": 0.3700137734413147,
      "learning_rate": 9.749520014968934e-05,
      "loss": 0.0497,
      "step": 630
    },
    {
      "epoch": 0.9706161137440759,
      "grad_norm": 0.16355620324611664,
      "learning_rate": 9.732695895466735e-05,
      "loss": 0.0399,
      "step": 640
    },
    {
      "epoch": 0.985781990521327,
      "grad_norm": 0.25550681352615356,
      "learning_rate": 9.715340457527746e-05,
      "loss": 0.0436,
      "step": 650
    },
    {
      "epoch": 0.985781990521327,
      "eval_loss": 0.03899650275707245,
      "eval_runtime": 158.4962,
      "eval_samples_per_second": 7.401,
      "eval_steps_per_second": 7.401,
      "step": 650
    },
    {
      "epoch": 1.000947867298578,
      "grad_norm": 0.4347744286060333,
      "learning_rate": 9.697455649568761e-05,
      "loss": 0.0437,
      "step": 660
    },
    {
      "epoch": 1.0161137440758294,
      "grad_norm": 0.32801559567451477,
      "learning_rate": 9.679043479436556e-05,
      "loss": 0.0358,
      "step": 670
    },
    {
      "epoch": 1.0312796208530806,
      "grad_norm": 0.34967103600502014,
      "learning_rate": 9.660106014182489e-05,
      "loss": 0.0391,
      "step": 680
    },
    {
      "epoch": 1.0464454976303317,
      "grad_norm": 0.2869499623775482,
      "learning_rate": 9.640645379830424e-05,
      "loss": 0.0415,
      "step": 690
    },
    {
      "epoch": 1.061611374407583,
      "grad_norm": 0.4630194902420044,
      "learning_rate": 9.620663761138067e-05,
      "loss": 0.0324,
      "step": 700
    },
    {
      "epoch": 1.061611374407583,
      "eval_loss": 0.04166208207607269,
      "eval_runtime": 158.5595,
      "eval_samples_per_second": 7.398,
      "eval_steps_per_second": 7.398,
      "step": 700
    },
    {
      "epoch": 1.0767772511848341,
      "grad_norm": 0.11792129278182983,
      "learning_rate": 9.600163401351688e-05,
      "loss": 0.0311,
      "step": 710
    },
    {
      "epoch": 1.0919431279620853,
      "grad_norm": 0.34663549065589905,
      "learning_rate": 9.579146601954276e-05,
      "loss": 0.0433,
      "step": 720
    },
    {
      "epoch": 1.1071090047393364,
      "grad_norm": 0.26798030734062195,
      "learning_rate": 9.557615722407177e-05,
      "loss": 0.0375,
      "step": 730
    },
    {
      "epoch": 1.1222748815165877,
      "grad_norm": 0.3786665201187134,
      "learning_rate": 9.535573179885191e-05,
      "loss": 0.0398,
      "step": 740
    },
    {
      "epoch": 1.1374407582938388,
      "grad_norm": 0.11541562527418137,
      "learning_rate": 9.513021449005214e-05,
      "loss": 0.0356,
      "step": 750
    },
    {
      "epoch": 1.1374407582938388,
      "eval_loss": 0.04211043193936348,
      "eval_runtime": 158.5353,
      "eval_samples_per_second": 7.399,
      "eval_steps_per_second": 7.399,
      "step": 750
    },
    {
      "epoch": 1.15260663507109,
      "grad_norm": 0.17363899946212769,
      "learning_rate": 9.489963061548428e-05,
      "loss": 0.0316,
      "step": 760
    },
    {
      "epoch": 1.1677725118483413,
      "grad_norm": 0.2537560760974884,
      "learning_rate": 9.466400606176062e-05,
      "loss": 0.0367,
      "step": 770
    },
    {
      "epoch": 1.1829383886255924,
      "grad_norm": 0.637069582939148,
      "learning_rate": 9.442336728138779e-05,
      "loss": 0.0448,
      "step": 780
    },
    {
      "epoch": 1.1981042654028435,
      "grad_norm": 0.3410698175430298,
      "learning_rate": 9.417774128979706e-05,
      "loss": 0.0375,
      "step": 790
    },
    {
      "epoch": 1.2132701421800949,
      "grad_norm": 0.3418003022670746,
      "learning_rate": 9.39271556623114e-05,
      "loss": 0.0368,
      "step": 800
    },
    {
      "epoch": 1.2132701421800949,
      "eval_loss": 0.037808436900377274,
      "eval_runtime": 158.5396,
      "eval_samples_per_second": 7.399,
      "eval_steps_per_second": 7.399,
      "step": 800
    },
    {
      "epoch": 1.228436018957346,
      "grad_norm": 0.34589603543281555,
      "learning_rate": 9.367163853104975e-05,
      "loss": 0.0349,
      "step": 810
    },
    {
      "epoch": 1.2436018957345971,
      "grad_norm": 0.246941938996315,
      "learning_rate": 9.341121858176876e-05,
      "loss": 0.0387,
      "step": 820
    },
    {
      "epoch": 1.2587677725118485,
      "grad_norm": 0.3432750105857849,
      "learning_rate": 9.314592505064239e-05,
      "loss": 0.0316,
      "step": 830
    },
    {
      "epoch": 1.2739336492890996,
      "grad_norm": 0.2746068239212036,
      "learning_rate": 9.28757877209796e-05,
      "loss": 0.0298,
      "step": 840
    },
    {
      "epoch": 1.2890995260663507,
      "grad_norm": 0.41217970848083496,
      "learning_rate": 9.260083691988084e-05,
      "loss": 0.0322,
      "step": 850
    },
    {
      "epoch": 1.2890995260663507,
      "eval_loss": 0.04183207452297211,
      "eval_runtime": 158.6062,
      "eval_samples_per_second": 7.396,
      "eval_steps_per_second": 7.396,
      "step": 850
    },
    {
      "epoch": 1.3042654028436018,
      "grad_norm": 0.5181209444999695,
      "learning_rate": 9.232110351483327e-05,
      "loss": 0.0496,
      "step": 860
    },
    {
      "epoch": 1.3194312796208532,
      "grad_norm": 0.3147033154964447,
      "learning_rate": 9.203661891024547e-05,
      "loss": 0.04,
      "step": 870
    },
    {
      "epoch": 1.3345971563981043,
      "grad_norm": 0.17610414326190948,
      "learning_rate": 9.174741504392173e-05,
      "loss": 0.0317,
      "step": 880
    },
    {
      "epoch": 1.3497630331753554,
      "grad_norm": 0.49788331985473633,
      "learning_rate": 9.145352438347662e-05,
      "loss": 0.0276,
      "step": 890
    },
    {
      "epoch": 1.3649289099526065,
      "grad_norm": 0.4363148808479309,
      "learning_rate": 9.115497992268995e-05,
      "loss": 0.0242,
      "step": 900
    },
    {
      "epoch": 1.3649289099526065,
      "eval_loss": 0.04041317105293274,
      "eval_runtime": 158.4747,
      "eval_samples_per_second": 7.402,
      "eval_steps_per_second": 7.402,
      "step": 900
    },
    {
      "epoch": 1.3800947867298579,
      "grad_norm": 0.2870277166366577,
      "learning_rate": 9.085181517780273e-05,
      "loss": 0.0372,
      "step": 910
    },
    {
      "epoch": 1.395260663507109,
      "grad_norm": 0.3796568214893341,
      "learning_rate": 9.054406418375443e-05,
      "loss": 0.0419,
      "step": 920
    },
    {
      "epoch": 1.41042654028436,
      "grad_norm": 0.23961886763572693,
      "learning_rate": 9.023176149036203e-05,
      "loss": 0.0352,
      "step": 930
    },
    {
      "epoch": 1.4255924170616114,
      "grad_norm": 0.3385351300239563,
      "learning_rate": 8.991494215844132e-05,
      "loss": 0.0368,
      "step": 940
    },
    {
      "epoch": 1.4407582938388626,
      "grad_norm": 0.18048731982707977,
      "learning_rate": 8.959364175587069e-05,
      "loss": 0.0278,
      "step": 950
    },
    {
      "epoch": 1.4407582938388626,
      "eval_loss": 0.039469122886657715,
      "eval_runtime": 158.4868,
      "eval_samples_per_second": 7.401,
      "eval_steps_per_second": 7.401,
      "step": 950
    },
    {
      "epoch": 1.4559241706161137,
      "grad_norm": 0.25072115659713745,
      "learning_rate": 8.926789635359817e-05,
      "loss": 0.0368,
      "step": 960
    },
    {
      "epoch": 1.471090047393365,
      "grad_norm": 0.19099581241607666,
      "learning_rate": 8.893774252159187e-05,
      "loss": 0.0305,
      "step": 970
    },
    {
      "epoch": 1.4862559241706161,
      "grad_norm": 0.280564546585083,
      "learning_rate": 8.860321732473439e-05,
      "loss": 0.0315,
      "step": 980
    },
    {
      "epoch": 1.5014218009478673,
      "grad_norm": 0.2831890881061554,
      "learning_rate": 8.826435831866184e-05,
      "loss": 0.0364,
      "step": 990
    },
    {
      "epoch": 1.5165876777251186,
      "grad_norm": 0.1274333894252777,
      "learning_rate": 8.79212035455475e-05,
      "loss": 0.0391,
      "step": 1000
    },
    {
      "epoch": 1.5165876777251186,
      "eval_loss": 0.03560240566730499,
      "eval_runtime": 158.4471,
      "eval_samples_per_second": 7.403,
      "eval_steps_per_second": 7.403,
      "step": 1000
    },
    {
      "epoch": 1.5317535545023697,
      "grad_norm": 0.26717591285705566,
      "learning_rate": 8.757379152983103e-05,
      "loss": 0.0295,
      "step": 1010
    },
    {
      "epoch": 1.5469194312796208,
      "grad_norm": 0.36288437247276306,
      "learning_rate": 8.722216127389363e-05,
      "loss": 0.0238,
      "step": 1020
    },
    {
      "epoch": 1.5620853080568722,
      "grad_norm": 0.20808231830596924,
      "learning_rate": 8.686635225367919e-05,
      "loss": 0.0244,
      "step": 1030
    },
    {
      "epoch": 1.577251184834123,
      "grad_norm": 0.3459094166755676,
      "learning_rate": 8.650640441426274e-05,
      "loss": 0.0425,
      "step": 1040
    },
    {
      "epoch": 1.5924170616113744,
      "grad_norm": 0.11016377061605453,
      "learning_rate": 8.614235816536582e-05,
      "loss": 0.0294,
      "step": 1050
    },
    {
      "epoch": 1.5924170616113744,
      "eval_loss": 0.035956691950559616,
      "eval_runtime": 158.5287,
      "eval_samples_per_second": 7.399,
      "eval_steps_per_second": 7.399,
      "step": 1050
    },
    {
      "epoch": 1.6075829383886255,
      "grad_norm": 0.11278466880321503,
      "learning_rate": 8.577425437681994e-05,
      "loss": 0.0385,
      "step": 1060
    },
    {
      "epoch": 1.6227488151658767,
      "grad_norm": 0.28282463550567627,
      "learning_rate": 8.540213437397833e-05,
      "loss": 0.0379,
      "step": 1070
    },
    {
      "epoch": 1.637914691943128,
      "grad_norm": 0.1440175175666809,
      "learning_rate": 8.502603993307647e-05,
      "loss": 0.0336,
      "step": 1080
    },
    {
      "epoch": 1.6530805687203791,
      "grad_norm": 0.27948638796806335,
      "learning_rate": 8.464601327654207e-05,
      "loss": 0.0346,
      "step": 1090
    },
    {
      "epoch": 1.6682464454976302,
      "grad_norm": 0.1889415830373764,
      "learning_rate": 8.4262097068255e-05,
      "loss": 0.023,
      "step": 1100
    },
    {
      "epoch": 1.6682464454976302,
      "eval_loss": 0.03588687628507614,
      "eval_runtime": 158.6065,
      "eval_samples_per_second": 7.396,
      "eval_steps_per_second": 7.396,
      "step": 1100
    },
    {
      "epoch": 1.6834123222748816,
      "grad_norm": 0.24837568402290344,
      "learning_rate": 8.387433440875758e-05,
      "loss": 0.0389,
      "step": 1110
    },
    {
      "epoch": 1.6985781990521327,
      "grad_norm": 0.1272035539150238,
      "learning_rate": 8.348276883041583e-05,
      "loss": 0.0276,
      "step": 1120
    },
    {
      "epoch": 1.7137440758293838,
      "grad_norm": 0.16301031410694122,
      "learning_rate": 8.308744429253238e-05,
      "loss": 0.0287,
      "step": 1130
    },
    {
      "epoch": 1.7289099526066352,
      "grad_norm": 0.3571413457393646,
      "learning_rate": 8.268840517641123e-05,
      "loss": 0.0308,
      "step": 1140
    },
    {
      "epoch": 1.7440758293838863,
      "grad_norm": 0.18815423548221588,
      "learning_rate": 8.228569628037543e-05,
      "loss": 0.0282,
      "step": 1150
    },
    {
      "epoch": 1.7440758293838863,
      "eval_loss": 0.03722427040338516,
      "eval_runtime": 158.5456,
      "eval_samples_per_second": 7.399,
      "eval_steps_per_second": 7.399,
      "step": 1150
    },
    {
      "epoch": 1.7592417061611374,
      "grad_norm": 0.40572747588157654,
      "learning_rate": 8.187936281473758e-05,
      "loss": 0.0368,
      "step": 1160
    },
    {
      "epoch": 1.7744075829383887,
      "grad_norm": 0.21999242901802063,
      "learning_rate": 8.146945039672446e-05,
      "loss": 0.0282,
      "step": 1170
    },
    {
      "epoch": 1.7895734597156399,
      "grad_norm": 0.2959887385368347,
      "learning_rate": 8.105600504535567e-05,
      "loss": 0.024,
      "step": 1180
    },
    {
      "epoch": 1.804739336492891,
      "grad_norm": 0.4860469400882721,
      "learning_rate": 8.06390731762773e-05,
      "loss": 0.0422,
      "step": 1190
    },
    {
      "epoch": 1.8199052132701423,
      "grad_norm": 0.23662057518959045,
      "learning_rate": 8.021870159655109e-05,
      "loss": 0.0343,
      "step": 1200
    },
    {
      "epoch": 1.8199052132701423,
      "eval_loss": 0.03589467704296112,
      "eval_runtime": 158.5983,
      "eval_samples_per_second": 7.396,
      "eval_steps_per_second": 7.396,
      "step": 1200
    },
    {
      "epoch": 1.8350710900473932,
      "grad_norm": 0.2618006765842438,
      "learning_rate": 7.979493749939955e-05,
      "loss": 0.0279,
      "step": 1210
    },
    {
      "epoch": 1.8502369668246446,
      "grad_norm": 0.21385808289051056,
      "learning_rate": 7.93678284589079e-05,
      "loss": 0.0292,
      "step": 1220
    },
    {
      "epoch": 1.865402843601896,
      "grad_norm": 0.2664329707622528,
      "learning_rate": 7.893742242468301e-05,
      "loss": 0.0297,
      "step": 1230
    },
    {
      "epoch": 1.8805687203791468,
      "grad_norm": 0.30637428164482117,
      "learning_rate": 7.850376771647038e-05,
      "loss": 0.0414,
      "step": 1240
    },
    {
      "epoch": 1.8957345971563981,
      "grad_norm": 0.18872077763080597,
      "learning_rate": 7.806691301872958e-05,
      "loss": 0.0335,
      "step": 1250
    },
    {
      "epoch": 1.8957345971563981,
      "eval_loss": 0.03388975188136101,
      "eval_runtime": 158.6564,
      "eval_samples_per_second": 7.393,
      "eval_steps_per_second": 7.393,
      "step": 1250
    },
    {
      "epoch": 1.9109004739336493,
      "grad_norm": 0.13050656020641327,
      "learning_rate": 7.762690737516846e-05,
      "loss": 0.0383,
      "step": 1260
    },
    {
      "epoch": 1.9260663507109004,
      "grad_norm": 1.4665435552597046,
      "learning_rate": 7.718380018323742e-05,
      "loss": 0.0394,
      "step": 1270
    },
    {
      "epoch": 1.9412322274881517,
      "grad_norm": 0.2703875005245209,
      "learning_rate": 7.673764118858371e-05,
      "loss": 0.0242,
      "step": 1280
    },
    {
      "epoch": 1.9563981042654028,
      "grad_norm": 0.3413107991218567,
      "learning_rate": 7.628848047946675e-05,
      "loss": 0.0468,
      "step": 1290
    },
    {
      "epoch": 1.971563981042654,
      "grad_norm": 0.3564504384994507,
      "learning_rate": 7.583636848113483e-05,
      "loss": 0.0389,
      "step": 1300
    },
    {
      "epoch": 1.971563981042654,
      "eval_loss": 0.035709355026483536,
      "eval_runtime": 158.6447,
      "eval_samples_per_second": 7.394,
      "eval_steps_per_second": 7.394,
      "step": 1300
    },
    {
      "epoch": 1.9867298578199053,
      "grad_norm": 0.40223392844200134,
      "learning_rate": 7.538135595016423e-05,
      "loss": 0.0339,
      "step": 1310
    },
    {
      "epoch": 2.001895734597156,
      "grad_norm": 0.19443105161190033,
      "learning_rate": 7.492349396876096e-05,
      "loss": 0.0376,
      "step": 1320
    },
    {
      "epoch": 2.0170616113744075,
      "grad_norm": 0.10920795053243637,
      "learning_rate": 7.4462833939026e-05,
      "loss": 0.0236,
      "step": 1330
    },
    {
      "epoch": 2.032227488151659,
      "grad_norm": 0.06198872998356819,
      "learning_rate": 7.399942757718455e-05,
      "loss": 0.0209,
      "step": 1340
    },
    {
      "epoch": 2.0473933649289098,
      "grad_norm": 1.0219924449920654,
      "learning_rate": 7.35333269077802e-05,
      "loss": 0.0277,
      "step": 1350
    },
    {
      "epoch": 2.0473933649289098,
      "eval_loss": 0.03505294770002365,
      "eval_runtime": 158.7744,
      "eval_samples_per_second": 7.388,
      "eval_steps_per_second": 7.388,
      "step": 1350
    },
    {
      "epoch": 2.062559241706161,
      "grad_norm": 0.20012281835079193,
      "learning_rate": 7.306458425783426e-05,
      "loss": 0.0196,
      "step": 1360
    },
    {
      "epoch": 2.0777251184834125,
      "grad_norm": 0.1867983341217041,
      "learning_rate": 7.25932522509713e-05,
      "loss": 0.0222,
      "step": 1370
    },
    {
      "epoch": 2.0928909952606634,
      "grad_norm": 0.5094829797744751,
      "learning_rate": 7.211938380151133e-05,
      "loss": 0.0183,
      "step": 1380
    },
    {
      "epoch": 2.1080568720379147,
      "grad_norm": 0.17359404265880585,
      "learning_rate": 7.164303210852934e-05,
      "loss": 0.0254,
      "step": 1390
    },
    {
      "epoch": 2.123222748815166,
      "grad_norm": 0.4848185181617737,
      "learning_rate": 7.116425064988286e-05,
      "loss": 0.0193,
      "step": 1400
    },
    {
      "epoch": 2.123222748815166,
      "eval_loss": 0.03434028849005699,
      "eval_runtime": 158.7125,
      "eval_samples_per_second": 7.391,
      "eval_steps_per_second": 7.391,
      "step": 1400
    },
    {
      "epoch": 2.138388625592417,
      "grad_norm": 0.2395314872264862,
      "learning_rate": 7.068309317620827e-05,
      "loss": 0.0234,
      "step": 1410
    },
    {
      "epoch": 2.1535545023696683,
      "grad_norm": 0.20019327104091644,
      "learning_rate": 7.019961370488645e-05,
      "loss": 0.0237,
      "step": 1420
    },
    {
      "epoch": 2.1687203791469196,
      "grad_norm": 0.2033195197582245,
      "learning_rate": 6.971386651397849e-05,
      "loss": 0.0252,
      "step": 1430
    },
    {
      "epoch": 2.1838862559241705,
      "grad_norm": 0.49193018674850464,
      "learning_rate": 6.922590613613211e-05,
      "loss": 0.0222,
      "step": 1440
    },
    {
      "epoch": 2.199052132701422,
      "grad_norm": 0.10803612321615219,
      "learning_rate": 6.873578735245961e-05,
      "loss": 0.0211,
      "step": 1450
    },
    {
      "epoch": 2.199052132701422,
      "eval_loss": 0.03537236899137497,
      "eval_runtime": 158.6885,
      "eval_samples_per_second": 7.392,
      "eval_steps_per_second": 7.392,
      "step": 1450
    },
    {
      "epoch": 2.2142180094786728,
      "grad_norm": 0.09245508164167404,
      "learning_rate": 6.824356518638775e-05,
      "loss": 0.0195,
      "step": 1460
    },
    {
      "epoch": 2.229383886255924,
      "grad_norm": 0.18777677416801453,
      "learning_rate": 6.774929489748052e-05,
      "loss": 0.0236,
      "step": 1470
    },
    {
      "epoch": 2.2445497630331754,
      "grad_norm": 0.09797560423612595,
      "learning_rate": 6.725303197523548e-05,
      "loss": 0.02,
      "step": 1480
    },
    {
      "epoch": 2.259715639810427,
      "grad_norm": 0.28358811140060425,
      "learning_rate": 6.675483213285412e-05,
      "loss": 0.0205,
      "step": 1490
    },
    {
      "epoch": 2.2748815165876777,
      "grad_norm": 0.17343221604824066,
      "learning_rate": 6.625475130098728e-05,
      "loss": 0.0149,
      "step": 1500
    },
    {
      "epoch": 2.2748815165876777,
      "eval_loss": 0.03519297018647194,
      "eval_runtime": 158.7684,
      "eval_samples_per_second": 7.388,
      "eval_steps_per_second": 7.388,
      "step": 1500
    },
    {
      "epoch": 2.290047393364929,
      "grad_norm": 0.19066782295703888,
      "learning_rate": 6.575284562145593e-05,
      "loss": 0.0349,
      "step": 1510
    },
    {
      "epoch": 2.30521327014218,
      "grad_norm": 0.23497186601161957,
      "learning_rate": 6.524917144094851e-05,
      "loss": 0.0231,
      "step": 1520
    },
    {
      "epoch": 2.3203791469194313,
      "grad_norm": 0.22643285989761353,
      "learning_rate": 6.474378530469509e-05,
      "loss": 0.0301,
      "step": 1530
    },
    {
      "epoch": 2.3355450236966826,
      "grad_norm": 0.12768208980560303,
      "learning_rate": 6.42367439501193e-05,
      "loss": 0.0226,
      "step": 1540
    },
    {
      "epoch": 2.3507109004739335,
      "grad_norm": 0.26942113041877747,
      "learning_rate": 6.372810430046862e-05,
      "loss": 0.0258,
      "step": 1550
    },
    {
      "epoch": 2.3507109004739335,
      "eval_loss": 0.03370455652475357,
      "eval_runtime": 158.7945,
      "eval_samples_per_second": 7.387,
      "eval_steps_per_second": 7.387,
      "step": 1550
    },
    {
      "epoch": 2.365876777251185,
      "grad_norm": 0.6548528671264648,
      "learning_rate": 6.321792345842402e-05,
      "loss": 0.0196,
      "step": 1560
    },
    {
      "epoch": 2.381042654028436,
      "grad_norm": 0.37841567397117615,
      "learning_rate": 6.270625869968906e-05,
      "loss": 0.0196,
      "step": 1570
    },
    {
      "epoch": 2.396208530805687,
      "grad_norm": 0.15341871976852417,
      "learning_rate": 6.219316746656007e-05,
      "loss": 0.0198,
      "step": 1580
    },
    {
      "epoch": 2.4113744075829384,
      "grad_norm": 0.11849100887775421,
      "learning_rate": 6.167870736147713e-05,
      "loss": 0.0295,
      "step": 1590
    },
    {
      "epoch": 2.4265402843601898,
      "grad_norm": 0.4521804451942444,
      "learning_rate": 6.116293614055744e-05,
      "loss": 0.0255,
      "step": 1600
    },
    {
      "epoch": 2.4265402843601898,
      "eval_loss": 0.03591766580939293,
      "eval_runtime": 158.7516,
      "eval_samples_per_second": 7.389,
      "eval_steps_per_second": 7.389,
      "step": 1600
    },
    {
      "epoch": 2.4417061611374407,
      "grad_norm": 0.3399631977081299,
      "learning_rate": 6.06459117071113e-05,
      "loss": 0.0214,
      "step": 1610
    },
    {
      "epoch": 2.456872037914692,
      "grad_norm": 0.15789680182933807,
      "learning_rate": 6.012769210514146e-05,
      "loss": 0.0265,
      "step": 1620
    },
    {
      "epoch": 2.472037914691943,
      "grad_norm": 0.2058618664741516,
      "learning_rate": 5.9608335512826915e-05,
      "loss": 0.022,
      "step": 1630
    },
    {
      "epoch": 2.4872037914691942,
      "grad_norm": 0.445004940032959,
      "learning_rate": 5.908790023599144e-05,
      "loss": 0.0225,
      "step": 1640
    },
    {
      "epoch": 2.5023696682464456,
      "grad_norm": 0.3923346996307373,
      "learning_rate": 5.856644470155781e-05,
      "loss": 0.014,
      "step": 1650
    },
    {
      "epoch": 2.5023696682464456,
      "eval_loss": 0.0376654677093029,
      "eval_runtime": 158.5463,
      "eval_samples_per_second": 7.398,
      "eval_steps_per_second": 7.398,
      "step": 1650
    },
    {
      "epoch": 2.517535545023697,
      "grad_norm": 0.10384202748537064,
      "learning_rate": 5.8044027450988546e-05,
      "loss": 0.0239,
      "step": 1660
    },
    {
      "epoch": 2.532701421800948,
      "grad_norm": 0.09303828328847885,
      "learning_rate": 5.752070713371371e-05,
      "loss": 0.0182,
      "step": 1670
    },
    {
      "epoch": 2.547867298578199,
      "grad_norm": 0.16929079592227936,
      "learning_rate": 5.699654250054662e-05,
      "loss": 0.0335,
      "step": 1680
    },
    {
      "epoch": 2.56303317535545,
      "grad_norm": 0.36071956157684326,
      "learning_rate": 5.647159239708809e-05,
      "loss": 0.0309,
      "step": 1690
    },
    {
      "epoch": 2.5781990521327014,
      "grad_norm": 0.1808507889509201,
      "learning_rate": 5.5945915757120146e-05,
      "loss": 0.0265,
      "step": 1700
    },
    {
      "epoch": 2.5781990521327014,
      "eval_loss": 0.03357379883527756,
      "eval_runtime": 158.639,
      "eval_samples_per_second": 7.394,
      "eval_steps_per_second": 7.394,
      "step": 1700
    },
    {
      "epoch": 2.5933649289099527,
      "grad_norm": 0.2682015597820282,
      "learning_rate": 5.5419571595989825e-05,
      "loss": 0.0287,
      "step": 1710
    },
    {
      "epoch": 2.6085308056872036,
      "grad_norm": 0.4746161699295044,
      "learning_rate": 5.4892619003983734e-05,
      "loss": 0.0215,
      "step": 1720
    },
    {
      "epoch": 2.623696682464455,
      "grad_norm": 0.2993459105491638,
      "learning_rate": 5.436511713969428e-05,
      "loss": 0.0256,
      "step": 1730
    },
    {
      "epoch": 2.6388625592417063,
      "grad_norm": 0.2758367657661438,
      "learning_rate": 5.383712522337817e-05,
      "loss": 0.0227,
      "step": 1740
    },
    {
      "epoch": 2.654028436018957,
      "grad_norm": 0.09800033271312714,
      "learning_rate": 5.3308702530308076e-05,
      "loss": 0.0211,
      "step": 1750
    },
    {
      "epoch": 2.654028436018957,
      "eval_loss": 0.03440028801560402,
      "eval_runtime": 158.6318,
      "eval_samples_per_second": 7.394,
      "eval_steps_per_second": 7.394,
      "step": 1750
    },
    {
      "epoch": 2.6691943127962086,
      "grad_norm": 0.17268642783164978,
      "learning_rate": 5.2779908384118025e-05,
      "loss": 0.0201,
      "step": 1760
    },
    {
      "epoch": 2.68436018957346,
      "grad_norm": 0.4179351031780243,
      "learning_rate": 5.22508021501434e-05,
      "loss": 0.0214,
      "step": 1770
    },
    {
      "epoch": 2.699526066350711,
      "grad_norm": 0.12654578685760498,
      "learning_rate": 5.1721443228756284e-05,
      "loss": 0.0165,
      "step": 1780
    },
    {
      "epoch": 2.714691943127962,
      "grad_norm": 0.12455421686172485,
      "learning_rate": 5.119189104869683e-05,
      "loss": 0.028,
      "step": 1790
    },
    {
      "epoch": 2.729857819905213,
      "grad_norm": 0.2781529426574707,
      "learning_rate": 5.066220506040148e-05,
      "loss": 0.0278,
      "step": 1800
    },
    {
      "epoch": 2.729857819905213,
      "eval_loss": 0.03548266366124153,
      "eval_runtime": 158.5694,
      "eval_samples_per_second": 7.397,
      "eval_steps_per_second": 7.397,
      "step": 1800
    },
    {
      "epoch": 2.7450236966824644,
      "grad_norm": 0.1623728722333908,
      "learning_rate": 5.013244472932872e-05,
      "loss": 0.0215,
      "step": 1810
    },
    {
      "epoch": 2.7601895734597157,
      "grad_norm": 0.33020347356796265,
      "learning_rate": 4.960266952928316e-05,
      "loss": 0.0236,
      "step": 1820
    },
    {
      "epoch": 2.775355450236967,
      "grad_norm": 0.3335643708705902,
      "learning_rate": 4.907293893573867e-05,
      "loss": 0.0199,
      "step": 1830
    },
    {
      "epoch": 2.790521327014218,
      "grad_norm": 0.33666765689849854,
      "learning_rate": 4.8543312419161396e-05,
      "loss": 0.0166,
      "step": 1840
    },
    {
      "epoch": 2.8056872037914693,
      "grad_norm": 0.5400834083557129,
      "learning_rate": 4.8013849438333165e-05,
      "loss": 0.0253,
      "step": 1850
    },
    {
      "epoch": 2.8056872037914693,
      "eval_loss": 0.0362522192299366,
      "eval_runtime": 158.496,
      "eval_samples_per_second": 7.401,
      "eval_steps_per_second": 7.401,
      "step": 1850
    },
    {
      "epoch": 2.82085308056872,
      "grad_norm": 0.38801684975624084,
      "learning_rate": 4.748460943367643e-05,
      "loss": 0.0245,
      "step": 1860
    },
    {
      "epoch": 2.8360189573459715,
      "grad_norm": 0.09480082243680954,
      "learning_rate": 4.695565182058113e-05,
      "loss": 0.0164,
      "step": 1870
    },
    {
      "epoch": 2.851184834123223,
      "grad_norm": 0.12024141848087311,
      "learning_rate": 4.642703598273431e-05,
      "loss": 0.0137,
      "step": 1880
    },
    {
      "epoch": 2.8663507109004738,
      "grad_norm": 0.40719497203826904,
      "learning_rate": 4.589882126545352e-05,
      "loss": 0.0175,
      "step": 1890
    },
    {
      "epoch": 2.881516587677725,
      "grad_norm": 0.09840822964906693,
      "learning_rate": 4.537106696902425e-05,
      "loss": 0.0178,
      "step": 1900
    },
    {
      "epoch": 2.881516587677725,
      "eval_loss": 0.03454902395606041,
      "eval_runtime": 158.4699,
      "eval_samples_per_second": 7.402,
      "eval_steps_per_second": 7.402,
      "step": 1900
    },
    {
      "epoch": 2.8966824644549765,
      "grad_norm": 0.1652071326971054,
      "learning_rate": 4.484383234204266e-05,
      "loss": 0.0219,
      "step": 1910
    },
    {
      "epoch": 2.9118483412322274,
      "grad_norm": 0.20622217655181885,
      "learning_rate": 4.4317176574763935e-05,
      "loss": 0.0235,
      "step": 1920
    },
    {
      "epoch": 2.9270142180094787,
      "grad_norm": 0.11598307639360428,
      "learning_rate": 4.3791158792457334e-05,
      "loss": 0.0202,
      "step": 1930
    },
    {
      "epoch": 2.94218009478673,
      "grad_norm": 0.6535903811454773,
      "learning_rate": 4.3265838048768334e-05,
      "loss": 0.0279,
      "step": 1940
    },
    {
      "epoch": 2.957345971563981,
      "grad_norm": 0.30591270327568054,
      "learning_rate": 4.274127331908915e-05,
      "loss": 0.0302,
      "step": 1950
    },
    {
      "epoch": 2.957345971563981,
      "eval_loss": 0.033974312245845795,
      "eval_runtime": 158.5364,
      "eval_samples_per_second": 7.399,
      "eval_steps_per_second": 7.399,
      "step": 1950
    },
    {
      "epoch": 2.9725118483412323,
      "grad_norm": 0.1717509925365448,
      "learning_rate": 4.221752349393758e-05,
      "loss": 0.0267,
      "step": 1960
    },
    {
      "epoch": 2.987677725118483,
      "grad_norm": 0.4148465692996979,
      "learning_rate": 4.1694647372345886e-05,
      "loss": 0.0291,
      "step": 1970
    },
    {
      "epoch": 3.0028436018957345,
      "grad_norm": 0.12019037455320358,
      "learning_rate": 4.117270365525946e-05,
      "loss": 0.0223,
      "step": 1980
    },
    {
      "epoch": 3.018009478672986,
      "grad_norm": 0.387494295835495,
      "learning_rate": 4.065175093894694e-05,
      "loss": 0.0154,
      "step": 1990
    },
    {
      "epoch": 3.0331753554502368,
      "grad_norm": 0.21212008595466614,
      "learning_rate": 4.013184770842167e-05,
      "loss": 0.0091,
      "step": 2000
    },
    {
      "epoch": 3.0331753554502368,
      "eval_loss": 0.035842038691043854,
      "eval_runtime": 158.5508,
      "eval_samples_per_second": 7.398,
      "eval_steps_per_second": 7.398,
      "step": 2000
    },
    {
      "epoch": 3.048341232227488,
      "grad_norm": 0.43005234003067017,
      "learning_rate": 3.9613052330876064e-05,
      "loss": 0.0184,
      "step": 2010
    },
    {
      "epoch": 3.0635071090047394,
      "grad_norm": 0.31887632608413696,
      "learning_rate": 3.909542304912881e-05,
      "loss": 0.0141,
      "step": 2020
    },
    {
      "epoch": 3.0786729857819903,
      "grad_norm": 0.27836355566978455,
      "learning_rate": 3.857901797508628e-05,
      "loss": 0.0113,
      "step": 2030
    },
    {
      "epoch": 3.0938388625592417,
      "grad_norm": 0.22910955548286438,
      "learning_rate": 3.80638950832186e-05,
      "loss": 0.0102,
      "step": 2040
    },
    {
      "epoch": 3.109004739336493,
      "grad_norm": 0.15187719464302063,
      "learning_rate": 3.7550112204051014e-05,
      "loss": 0.0072,
      "step": 2050
    },
    {
      "epoch": 3.109004739336493,
      "eval_loss": 0.0410844124853611,
      "eval_runtime": 158.5371,
      "eval_samples_per_second": 7.399,
      "eval_steps_per_second": 7.399,
      "step": 2050
    },
    {
      "epoch": 3.124170616113744,
      "grad_norm": 0.2116091102361679,
      "learning_rate": 3.703772701767167e-05,
      "loss": 0.0152,
      "step": 2060
    },
    {
      "epoch": 3.1393364928909953,
      "grad_norm": 0.2841944694519043,
      "learning_rate": 3.652679704725596e-05,
      "loss": 0.0109,
      "step": 2070
    },
    {
      "epoch": 3.1545023696682466,
      "grad_norm": 0.12401948124170303,
      "learning_rate": 3.601737965260882e-05,
      "loss": 0.0163,
      "step": 2080
    },
    {
      "epoch": 3.1696682464454975,
      "grad_norm": 0.3338809311389923,
      "learning_rate": 3.550953202372503e-05,
      "loss": 0.0146,
      "step": 2090
    },
    {
      "epoch": 3.184834123222749,
      "grad_norm": 0.2843485474586487,
      "learning_rate": 3.500331117436895e-05,
      "loss": 0.0126,
      "step": 2100
    },
    {
      "epoch": 3.184834123222749,
      "eval_loss": 0.03944702818989754,
      "eval_runtime": 158.4379,
      "eval_samples_per_second": 7.404,
      "eval_steps_per_second": 7.404,
      "step": 2100
    },
    {
      "epoch": 3.2,
      "grad_norm": 0.47763463854789734,
      "learning_rate": 3.4498773935673626e-05,
      "loss": 0.016,
      "step": 2110
    },
    {
      "epoch": 3.215165876777251,
      "grad_norm": 0.49825188517570496,
      "learning_rate": 3.399597694976081e-05,
      "loss": 0.0123,
      "step": 2120
    },
    {
      "epoch": 3.2303317535545024,
      "grad_norm": 0.11938051879405975,
      "learning_rate": 3.349497666338187e-05,
      "loss": 0.0122,
      "step": 2130
    },
    {
      "epoch": 3.2454976303317538,
      "grad_norm": 0.483530730009079,
      "learning_rate": 3.299582932158085e-05,
      "loss": 0.0102,
      "step": 2140
    },
    {
      "epoch": 3.2606635071090047,
      "grad_norm": 0.049740269780159,
      "learning_rate": 3.2498590961379996e-05,
      "loss": 0.0108,
      "step": 2150
    },
    {
      "epoch": 3.2606635071090047,
      "eval_loss": 0.040373362600803375,
      "eval_runtime": 158.4135,
      "eval_samples_per_second": 7.405,
      "eval_steps_per_second": 7.405,
      "step": 2150
    },
    {
      "epoch": 3.275829383886256,
      "grad_norm": 0.44491904973983765,
      "learning_rate": 3.200331740548887e-05,
      "loss": 0.0113,
      "step": 2160
    },
    {
      "epoch": 3.2909952606635073,
      "grad_norm": 0.372631698846817,
      "learning_rate": 3.1510064256037274e-05,
      "loss": 0.0121,
      "step": 2170
    },
    {
      "epoch": 3.3061611374407582,
      "grad_norm": 0.4665113091468811,
      "learning_rate": 3.1018886888333065e-05,
      "loss": 0.011,
      "step": 2180
    },
    {
      "epoch": 3.3213270142180096,
      "grad_norm": 0.17037945985794067,
      "learning_rate": 3.052984044464548e-05,
      "loss": 0.0152,
      "step": 2190
    },
    {
      "epoch": 3.3364928909952605,
      "grad_norm": 0.25758716464042664,
      "learning_rate": 3.0042979828014496e-05,
      "loss": 0.0101,
      "step": 2200
    },
    {
      "epoch": 3.3364928909952605,
      "eval_loss": 0.03813379630446434,
      "eval_runtime": 158.4665,
      "eval_samples_per_second": 7.402,
      "eval_steps_per_second": 7.402,
      "step": 2200
    },
    {
      "epoch": 3.351658767772512,
      "grad_norm": 0.30639883875846863,
      "learning_rate": 2.9558359696087152e-05,
      "loss": 0.0122,
      "step": 2210
    },
    {
      "epoch": 3.366824644549763,
      "grad_norm": 0.34101325273513794,
      "learning_rate": 2.9076034454981332e-05,
      "loss": 0.0107,
      "step": 2220
    },
    {
      "epoch": 3.381990521327014,
      "grad_norm": 0.04556896165013313,
      "learning_rate": 2.8596058253177932e-05,
      "loss": 0.0147,
      "step": 2230
    },
    {
      "epoch": 3.3971563981042654,
      "grad_norm": 0.5291752815246582,
      "learning_rate": 2.811848497544175e-05,
      "loss": 0.014,
      "step": 2240
    },
    {
      "epoch": 3.4123222748815167,
      "grad_norm": 0.20119443535804749,
      "learning_rate": 2.764336823677216e-05,
      "loss": 0.0077,
      "step": 2250
    },
    {
      "epoch": 3.4123222748815167,
      "eval_loss": 0.03824818506836891,
      "eval_runtime": 158.4263,
      "eval_samples_per_second": 7.404,
      "eval_steps_per_second": 7.404,
      "step": 2250
    },
    {
      "epoch": 3.4274881516587676,
      "grad_norm": 0.40690121054649353,
      "learning_rate": 2.717076137638388e-05,
      "loss": 0.0147,
      "step": 2260
    },
    {
      "epoch": 3.442654028436019,
      "grad_norm": 0.29800182580947876,
      "learning_rate": 2.6700717451719066e-05,
      "loss": 0.0131,
      "step": 2270
    },
    {
      "epoch": 3.4578199052132703,
      "grad_norm": 0.2343226969242096,
      "learning_rate": 2.6233289232490447e-05,
      "loss": 0.0079,
      "step": 2280
    },
    {
      "epoch": 3.472985781990521,
      "grad_norm": 0.24278907477855682,
      "learning_rate": 2.5768529194757474e-05,
      "loss": 0.0153,
      "step": 2290
    },
    {
      "epoch": 3.4881516587677726,
      "grad_norm": 0.19784784317016602,
      "learning_rate": 2.5306489515034713e-05,
      "loss": 0.0096,
      "step": 2300
    },
    {
      "epoch": 3.4881516587677726,
      "eval_loss": 0.037863798439502716,
      "eval_runtime": 158.4077,
      "eval_samples_per_second": 7.405,
      "eval_steps_per_second": 7.405,
      "step": 2300
    },
    {
      "epoch": 3.5033175355450235,
      "grad_norm": 0.5835947394371033,
      "learning_rate": 2.484722206443455e-05,
      "loss": 0.0186,
      "step": 2310
    },
    {
      "epoch": 3.518483412322275,
      "grad_norm": 0.055888980627059937,
      "learning_rate": 2.4390778402843605e-05,
      "loss": 0.0139,
      "step": 2320
    },
    {
      "epoch": 3.533649289099526,
      "grad_norm": 0.2058580070734024,
      "learning_rate": 2.3937209773134472e-05,
      "loss": 0.0069,
      "step": 2330
    },
    {
      "epoch": 3.5488151658767775,
      "grad_norm": 0.2727804481983185,
      "learning_rate": 2.3486567095412864e-05,
      "loss": 0.0145,
      "step": 2340
    },
    {
      "epoch": 3.5639810426540284,
      "grad_norm": 0.1926075965166092,
      "learning_rate": 2.3038900961301053e-05,
      "loss": 0.0077,
      "step": 2350
    },
    {
      "epoch": 3.5639810426540284,
      "eval_loss": 0.03924045339226723,
      "eval_runtime": 158.4455,
      "eval_samples_per_second": 7.403,
      "eval_steps_per_second": 7.403,
      "step": 2350
    },
    {
      "epoch": 3.5791469194312797,
      "grad_norm": 0.49359145760536194,
      "learning_rate": 2.259426162825819e-05,
      "loss": 0.0092,
      "step": 2360
    },
    {
      "epoch": 3.5943127962085306,
      "grad_norm": 0.09291447699069977,
      "learning_rate": 2.215269901393805e-05,
      "loss": 0.0087,
      "step": 2370
    },
    {
      "epoch": 3.609478672985782,
      "grad_norm": 0.4390147626399994,
      "learning_rate": 2.1714262690585192e-05,
      "loss": 0.0119,
      "step": 2380
    },
    {
      "epoch": 3.6246445497630333,
      "grad_norm": 0.28357169032096863,
      "learning_rate": 2.1279001879469424e-05,
      "loss": 0.015,
      "step": 2390
    },
    {
      "epoch": 3.639810426540284,
      "grad_norm": 0.40260136127471924,
      "learning_rate": 2.084696544536025e-05,
      "loss": 0.015,
      "step": 2400
    },
    {
      "epoch": 3.639810426540284,
      "eval_loss": 0.03814982622861862,
      "eval_runtime": 158.3999,
      "eval_samples_per_second": 7.405,
      "eval_steps_per_second": 7.405,
      "step": 2400
    },
    {
      "epoch": 3.6549763033175355,
      "grad_norm": 0.18413063883781433,
      "learning_rate": 2.0418201891040778e-05,
      "loss": 0.0131,
      "step": 2410
    },
    {
      "epoch": 3.670142180094787,
      "grad_norm": 0.1047610491514206,
      "learning_rate": 1.9992759351862772e-05,
      "loss": 0.0088,
      "step": 2420
    },
    {
      "epoch": 3.6853080568720378,
      "grad_norm": 0.43591246008872986,
      "learning_rate": 1.9570685590342463e-05,
      "loss": 0.0071,
      "step": 2430
    },
    {
      "epoch": 3.700473933649289,
      "grad_norm": 0.12613852322101593,
      "learning_rate": 1.9152027990798748e-05,
      "loss": 0.0076,
      "step": 2440
    },
    {
      "epoch": 3.7156398104265405,
      "grad_norm": 0.2396720051765442,
      "learning_rate": 1.8736833554033263e-05,
      "loss": 0.0095,
      "step": 2450
    },
    {
      "epoch": 3.7156398104265405,
      "eval_loss": 0.040059156715869904,
      "eval_runtime": 158.3833,
      "eval_samples_per_second": 7.406,
      "eval_steps_per_second": 7.406,
      "step": 2450
    },
    {
      "epoch": 3.7308056872037914,
      "grad_norm": 0.4986625015735626,
      "learning_rate": 1.8325148892054093e-05,
      "loss": 0.0147,
      "step": 2460
    },
    {
      "epoch": 3.7459715639810427,
      "grad_norm": 0.21489527821540833,
      "learning_rate": 1.7917020222842662e-05,
      "loss": 0.0078,
      "step": 2470
    },
    {
      "epoch": 3.7611374407582936,
      "grad_norm": 0.4590434730052948,
      "learning_rate": 1.751249336516513e-05,
      "loss": 0.0101,
      "step": 2480
    },
    {
      "epoch": 3.776303317535545,
      "grad_norm": 0.43470606207847595,
      "learning_rate": 1.7111613733428522e-05,
      "loss": 0.0144,
      "step": 2490
    },
    {
      "epoch": 3.7914691943127963,
      "grad_norm": 0.35228005051612854,
      "learning_rate": 1.671442633258218e-05,
      "loss": 0.0174,
      "step": 2500
    },
    {
      "epoch": 3.7914691943127963,
      "eval_loss": 0.039473824203014374,
      "eval_runtime": 158.3937,
      "eval_samples_per_second": 7.406,
      "eval_steps_per_second": 7.406,
      "step": 2500
    },
    {
      "epoch": 3.8066350710900476,
      "grad_norm": 0.45406851172447205,
      "learning_rate": 1.6320975753065453e-05,
      "loss": 0.0188,
      "step": 2510
    },
    {
      "epoch": 3.8218009478672985,
      "grad_norm": 0.3384013772010803,
      "learning_rate": 1.5931306165801452e-05,
      "loss": 0.0074,
      "step": 2520
    },
    {
      "epoch": 3.83696682464455,
      "grad_norm": 0.07098303735256195,
      "learning_rate": 1.554546131723848e-05,
      "loss": 0.0116,
      "step": 2530
    },
    {
      "epoch": 3.8521327014218008,
      "grad_norm": 0.3607996106147766,
      "learning_rate": 1.5163484524438516e-05,
      "loss": 0.0093,
      "step": 2540
    },
    {
      "epoch": 3.867298578199052,
      "grad_norm": 0.8953911066055298,
      "learning_rate": 1.4785418670214496e-05,
      "loss": 0.0105,
      "step": 2550
    },
    {
      "epoch": 3.867298578199052,
      "eval_loss": 0.03934425115585327,
      "eval_runtime": 158.3395,
      "eval_samples_per_second": 7.408,
      "eval_steps_per_second": 7.408,
      "step": 2550
    },
    {
      "epoch": 3.8824644549763034,
      "grad_norm": 0.14035186171531677,
      "learning_rate": 1.4411306198315777e-05,
      "loss": 0.0088,
      "step": 2560
    },
    {
      "epoch": 3.8976303317535548,
      "grad_norm": 0.3157799541950226,
      "learning_rate": 1.4041189108663421e-05,
      "loss": 0.0098,
      "step": 2570
    },
    {
      "epoch": 3.9127962085308057,
      "grad_norm": 0.900272011756897,
      "learning_rate": 1.3675108952634824e-05,
      "loss": 0.0174,
      "step": 2580
    },
    {
      "epoch": 3.927962085308057,
      "grad_norm": 0.31802159547805786,
      "learning_rate": 1.3313106828399147e-05,
      "loss": 0.0157,
      "step": 2590
    },
    {
      "epoch": 3.943127962085308,
      "grad_norm": 0.24903734028339386,
      "learning_rate": 1.2955223376303205e-05,
      "loss": 0.014,
      "step": 2600
    },
    {
      "epoch": 3.943127962085308,
      "eval_loss": 0.03850428760051727,
      "eval_runtime": 158.3637,
      "eval_samples_per_second": 7.407,
      "eval_steps_per_second": 7.407,
      "step": 2600
    },
    {
      "epoch": 3.9582938388625593,
      "grad_norm": 0.24580879509449005,
      "learning_rate": 1.2601498774309112e-05,
      "loss": 0.0086,
      "step": 2610
    },
    {
      "epoch": 3.9734597156398106,
      "grad_norm": 0.16215670108795166,
      "learning_rate": 1.2251972733483612e-05,
      "loss": 0.0106,
      "step": 2620
    },
    {
      "epoch": 3.9886255924170615,
      "grad_norm": 0.2537078559398651,
      "learning_rate": 1.19066844935399e-05,
      "loss": 0.0094,
      "step": 2630
    },
    {
      "epoch": 4.003791469194312,
      "grad_norm": 0.16900096833705902,
      "learning_rate": 1.156567281843241e-05,
      "loss": 0.0143,
      "step": 2640
    },
    {
      "epoch": 4.018957345971564,
      "grad_norm": 0.25430336594581604,
      "learning_rate": 1.1228975992004842e-05,
      "loss": 0.0086,
      "step": 2650
    },
    {
      "epoch": 4.018957345971564,
      "eval_loss": 0.04015021771192551,
      "eval_runtime": 158.3169,
      "eval_samples_per_second": 7.409,
      "eval_steps_per_second": 7.409,
      "step": 2650
    },
    {
      "epoch": 4.034123222748815,
      "grad_norm": 0.03381699323654175,
      "learning_rate": 1.0896631813692377e-05,
      "loss": 0.0056,
      "step": 2660
    },
    {
      "epoch": 4.049289099526066,
      "grad_norm": 0.0653318464756012,
      "learning_rate": 1.0568677594277899e-05,
      "loss": 0.0049,
      "step": 2670
    },
    {
      "epoch": 4.064454976303318,
      "grad_norm": 0.059762757271528244,
      "learning_rate": 1.0245150151703509e-05,
      "loss": 0.0036,
      "step": 2680
    },
    {
      "epoch": 4.079620853080569,
      "grad_norm": 0.1282765418291092,
      "learning_rate": 9.926085806936918e-06,
      "loss": 0.0055,
      "step": 2690
    },
    {
      "epoch": 4.0947867298578196,
      "grad_norm": 0.09458967298269272,
      "learning_rate": 9.611520379894068e-06,
      "loss": 0.005,
      "step": 2700
    },
    {
      "epoch": 4.0947867298578196,
      "eval_loss": 0.043914105743169785,
      "eval_runtime": 158.2953,
      "eval_samples_per_second": 7.41,
      "eval_steps_per_second": 7.41,
      "step": 2700
    },
    {
      "epoch": 4.109952606635071,
      "grad_norm": 0.34830743074417114,
      "learning_rate": 9.301489185417578e-06,
      "loss": 0.0063,
      "step": 2710
    },
    {
      "epoch": 4.125118483412322,
      "grad_norm": 0.06616072356700897,
      "learning_rate": 8.996027029312314e-06,
      "loss": 0.0043,
      "step": 2720
    },
    {
      "epoch": 4.140284360189574,
      "grad_norm": 0.4457527995109558,
      "learning_rate": 8.695168204437721e-06,
      "loss": 0.0052,
      "step": 2730
    },
    {
      "epoch": 4.155450236966825,
      "grad_norm": 0.08168656378984451,
      "learning_rate": 8.398946486858029e-06,
      "loss": 0.0039,
      "step": 2740
    },
    {
      "epoch": 4.170616113744076,
      "grad_norm": 0.46044808626174927,
      "learning_rate": 8.107395132050333e-06,
      "loss": 0.0044,
      "step": 2750
    },
    {
      "epoch": 4.170616113744076,
      "eval_loss": 0.04870103299617767,
      "eval_runtime": 158.3411,
      "eval_samples_per_second": 7.408,
      "eval_steps_per_second": 7.408,
      "step": 2750
    },
    {
      "epoch": 4.185781990521327,
      "grad_norm": 0.05418262258172035,
      "learning_rate": 7.820546871171142e-06,
      "loss": 0.0047,
      "step": 2760
    },
    {
      "epoch": 4.200947867298578,
      "grad_norm": 0.13520409166812897,
      "learning_rate": 7.5384339073818145e-06,
      "loss": 0.0041,
      "step": 2770
    },
    {
      "epoch": 4.216113744075829,
      "grad_norm": 0.07453792542219162,
      "learning_rate": 7.261087912233228e-06,
      "loss": 0.0043,
      "step": 2780
    },
    {
      "epoch": 4.231279620853081,
      "grad_norm": 0.12752209603786469,
      "learning_rate": 6.988540022110235e-06,
      "loss": 0.0127,
      "step": 2790
    },
    {
      "epoch": 4.246445497630332,
      "grad_norm": 0.05751938000321388,
      "learning_rate": 6.720820834735969e-06,
      "loss": 0.0032,
      "step": 2800
    },
    {
      "epoch": 4.246445497630332,
      "eval_loss": 0.04898425191640854,
      "eval_runtime": 158.3176,
      "eval_samples_per_second": 7.409,
      "eval_steps_per_second": 7.409,
      "step": 2800
    },
    {
      "epoch": 4.2616113744075825,
      "grad_norm": 0.2031586617231369,
      "learning_rate": 6.457960405736968e-06,
      "loss": 0.003,
      "step": 2810
    },
    {
      "epoch": 4.276777251184834,
      "grad_norm": 0.24639886617660522,
      "learning_rate": 6.199988245268778e-06,
      "loss": 0.0075,
      "step": 2820
    },
    {
      "epoch": 4.291943127962085,
      "grad_norm": 0.014945941977202892,
      "learning_rate": 5.946933314703157e-06,
      "loss": 0.0046,
      "step": 2830
    },
    {
      "epoch": 4.3071090047393366,
      "grad_norm": 0.10533477365970612,
      "learning_rate": 5.698824023376531e-06,
      "loss": 0.0037,
      "step": 2840
    },
    {
      "epoch": 4.322274881516588,
      "grad_norm": 0.6028527021408081,
      "learning_rate": 5.455688225400802e-06,
      "loss": 0.0052,
      "step": 2850
    },
    {
      "epoch": 4.322274881516588,
      "eval_loss": 0.04893364757299423,
      "eval_runtime": 158.2922,
      "eval_samples_per_second": 7.41,
      "eval_steps_per_second": 7.41,
      "step": 2850
    },
    {
      "epoch": 4.337440758293839,
      "grad_norm": 0.3932960629463196,
      "learning_rate": 5.217553216536098e-06,
      "loss": 0.0106,
      "step": 2860
    },
    {
      "epoch": 4.35260663507109,
      "grad_norm": 0.074831523001194,
      "learning_rate": 4.984445731126542e-06,
      "loss": 0.0061,
      "step": 2870
    },
    {
      "epoch": 4.367772511848341,
      "grad_norm": 0.1919291764497757,
      "learning_rate": 4.756391939098853e-06,
      "loss": 0.0036,
      "step": 2880
    },
    {
      "epoch": 4.382938388625592,
      "grad_norm": 0.040002696216106415,
      "learning_rate": 4.533417443024374e-06,
      "loss": 0.0063,
      "step": 2890
    },
    {
      "epoch": 4.398104265402844,
      "grad_norm": 0.13234563171863556,
      "learning_rate": 4.315547275244769e-06,
      "loss": 0.0089,
      "step": 2900
    },
    {
      "epoch": 4.398104265402844,
      "eval_loss": 0.04927956685423851,
      "eval_runtime": 158.012,
      "eval_samples_per_second": 7.423,
      "eval_steps_per_second": 7.423,
      "step": 2900
    },
    {
      "epoch": 4.413270142180095,
      "grad_norm": 0.34456437826156616,
      "learning_rate": 4.10280589506179e-06,
      "loss": 0.0031,
      "step": 2910
    },
    {
      "epoch": 4.4284360189573455,
      "grad_norm": 0.5860693454742432,
      "learning_rate": 3.895217185991312e-06,
      "loss": 0.0077,
      "step": 2920
    },
    {
      "epoch": 4.443601895734597,
      "grad_norm": 0.4375387132167816,
      "learning_rate": 3.692804453082038e-06,
      "loss": 0.0055,
      "step": 2930
    },
    {
      "epoch": 4.458767772511848,
      "grad_norm": 0.447825163602829,
      "learning_rate": 3.495590420299194e-06,
      "loss": 0.0049,
      "step": 2940
    },
    {
      "epoch": 4.4739336492890995,
      "grad_norm": 0.1041964516043663,
      "learning_rate": 3.3035972279733207e-06,
      "loss": 0.0095,
      "step": 2950
    },
    {
      "epoch": 4.4739336492890995,
      "eval_loss": 0.04903671145439148,
      "eval_runtime": 157.8361,
      "eval_samples_per_second": 7.432,
      "eval_steps_per_second": 7.432,
      "step": 2950
    },
    {
      "epoch": 4.489099526066351,
      "grad_norm": 0.3477923572063446,
      "learning_rate": 3.116846430314796e-06,
      "loss": 0.0033,
      "step": 2960
    },
    {
      "epoch": 4.504265402843602,
      "grad_norm": 0.03364011272788048,
      "learning_rate": 2.9353589929939087e-06,
      "loss": 0.0027,
      "step": 2970
    },
    {
      "epoch": 4.519431279620854,
      "grad_norm": 0.26229575276374817,
      "learning_rate": 2.7591552907872574e-06,
      "loss": 0.0078,
      "step": 2980
    },
    {
      "epoch": 4.534597156398104,
      "grad_norm": 0.5064939260482788,
      "learning_rate": 2.5882551052902883e-06,
      "loss": 0.0037,
      "step": 2990
    },
    {
      "epoch": 4.549763033175355,
      "grad_norm": 0.33356335759162903,
      "learning_rate": 2.4226776226965453e-06,
      "loss": 0.0049,
      "step": 3000
    },
    {
      "epoch": 4.549763033175355,
      "eval_loss": 0.04871579632163048,
      "eval_runtime": 157.6981,
      "eval_samples_per_second": 7.438,
      "eval_steps_per_second": 7.438,
      "step": 3000
    },
    {
      "epoch": 4.564928909952607,
      "grad_norm": 0.08507398515939713,
      "learning_rate": 2.262441431643697e-06,
      "loss": 0.0062,
      "step": 3010
    },
    {
      "epoch": 4.580094786729858,
      "grad_norm": 0.11064343899488449,
      "learning_rate": 2.1075645211266927e-06,
      "loss": 0.0036,
      "step": 3020
    },
    {
      "epoch": 4.595260663507109,
      "grad_norm": 0.2382802814245224,
      "learning_rate": 1.9580642784782053e-06,
      "loss": 0.0083,
      "step": 3030
    },
    {
      "epoch": 4.61042654028436,
      "grad_norm": 0.054368820041418076,
      "learning_rate": 1.813957487416651e-06,
      "loss": 0.0063,
      "step": 3040
    },
    {
      "epoch": 4.625592417061611,
      "grad_norm": 0.0921395868062973,
      "learning_rate": 1.6752603261619315e-06,
      "loss": 0.0025,
      "step": 3050
    },
    {
      "epoch": 4.625592417061611,
      "eval_loss": 0.049159836024045944,
      "eval_runtime": 157.3869,
      "eval_samples_per_second": 7.453,
      "eval_steps_per_second": 7.453,
      "step": 3050
    },
    {
      "epoch": 4.6407582938388625,
      "grad_norm": 0.040545761585235596,
      "learning_rate": 1.541988365619207e-06,
      "loss": 0.0026,
      "step": 3060
    },
    {
      "epoch": 4.655924170616114,
      "grad_norm": 0.11444900929927826,
      "learning_rate": 1.4141565676307865e-06,
      "loss": 0.0046,
      "step": 3070
    },
    {
      "epoch": 4.671090047393365,
      "grad_norm": 0.09660771489143372,
      "learning_rate": 1.291779283296468e-06,
      "loss": 0.0044,
      "step": 3080
    },
    {
      "epoch": 4.6862559241706165,
      "grad_norm": 1.581464171409607,
      "learning_rate": 1.1748702513623922e-06,
      "loss": 0.0067,
      "step": 3090
    },
    {
      "epoch": 4.701421800947867,
      "grad_norm": 0.047929033637046814,
      "learning_rate": 1.0634425966786155e-06,
      "loss": 0.0066,
      "step": 3100
    },
    {
      "epoch": 4.701421800947867,
      "eval_loss": 0.04977952688932419,
      "eval_runtime": 157.0034,
      "eval_samples_per_second": 7.471,
      "eval_steps_per_second": 7.471,
      "step": 3100
    },
    {
      "epoch": 4.716587677725118,
      "grad_norm": 0.16482333838939667,
      "learning_rate": 9.575088287257118e-07,
      "loss": 0.0052,
      "step": 3110
    },
    {
      "epoch": 4.73175355450237,
      "grad_norm": 0.16279959678649902,
      "learning_rate": 8.570808402103392e-07,
      "loss": 0.0083,
      "step": 3120
    },
    {
      "epoch": 4.746919431279621,
      "grad_norm": 0.118639275431633,
      "learning_rate": 7.62169905730109e-07,
      "loss": 0.0044,
      "step": 3130
    },
    {
      "epoch": 4.762085308056872,
      "grad_norm": 0.3321910798549652,
      "learning_rate": 6.727866805078531e-07,
      "loss": 0.0046,
      "step": 3140
    },
    {
      "epoch": 4.777251184834123,
      "grad_norm": 0.6542768478393555,
      "learning_rate": 5.889411991953975e-07,
      "loss": 0.0089,
      "step": 3150
    },
    {
      "epoch": 4.777251184834123,
      "eval_loss": 0.04997863993048668,
      "eval_runtime": 156.3539,
      "eval_samples_per_second": 7.502,
      "eval_steps_per_second": 7.502,
      "step": 3150
    },
    {
      "epoch": 4.792417061611374,
      "grad_norm": 0.010229017585515976,
      "learning_rate": 5.106428747470137e-07,
      "loss": 0.0027,
      "step": 3160
    },
    {
      "epoch": 4.8075829383886255,
      "grad_norm": 0.1426972895860672,
      "learning_rate": 4.37900497362681e-07,
      "loss": 0.0039,
      "step": 3170
    },
    {
      "epoch": 4.822748815165877,
      "grad_norm": 0.20429013669490814,
      "learning_rate": 3.7072223350124167e-07,
      "loss": 0.0056,
      "step": 3180
    },
    {
      "epoch": 4.837914691943128,
      "grad_norm": 0.25066667795181274,
      "learning_rate": 3.0911562496358517e-07,
      "loss": 0.0063,
      "step": 3190
    },
    {
      "epoch": 4.8530805687203795,
      "grad_norm": 0.07332632690668106,
      "learning_rate": 2.53087588045986e-07,
      "loss": 0.0021,
      "step": 3200
    },
    {
      "epoch": 4.8530805687203795,
      "eval_loss": 0.05003194510936737,
      "eval_runtime": 155.9263,
      "eval_samples_per_second": 7.523,
      "eval_steps_per_second": 7.523,
      "step": 3200
    },
    {
      "epoch": 4.86824644549763,
      "grad_norm": 0.34695953130722046,
      "learning_rate": 2.026444127636029e-07,
      "loss": 0.0026,
      "step": 3210
    },
    {
      "epoch": 4.883412322274881,
      "grad_norm": 0.10707182437181473,
      "learning_rate": 1.577917621443825e-07,
      "loss": 0.0042,
      "step": 3220
    },
    {
      "epoch": 4.898578199052133,
      "grad_norm": 0.8119176030158997,
      "learning_rate": 1.185346715932345e-07,
      "loss": 0.0087,
      "step": 3230
    },
    {
      "epoch": 4.913744075829384,
      "grad_norm": 0.3765670359134674,
      "learning_rate": 8.48775483267783e-08,
      "loss": 0.0056,
      "step": 3240
    },
    {
      "epoch": 4.928909952606635,
      "grad_norm": 0.06188008189201355,
      "learning_rate": 5.6824170878544414e-08,
      "loss": 0.0056,
      "step": 3250
    },
    {
      "epoch": 4.928909952606635,
      "eval_loss": 0.04992802441120148,
      "eval_runtime": 155.2754,
      "eval_samples_per_second": 7.554,
      "eval_steps_per_second": 7.554,
      "step": 3250
    },
    {
      "epoch": 4.944075829383886,
      "grad_norm": 0.1870199590921402,
      "learning_rate": 3.4377688674774866e-08,
      "loss": 0.0045,
      "step": 3260
    },
    {
      "epoch": 4.959241706161137,
      "grad_norm": 0.014189193956553936,
      "learning_rate": 1.7540621680850466e-08,
      "loss": 0.0073,
      "step": 3270
    },
    {
      "epoch": 4.9744075829383885,
      "grad_norm": 0.8881843090057373,
      "learning_rate": 6.314860118400434e-09,
      "loss": 0.0074,
      "step": 3280
    },
    {
      "epoch": 4.98957345971564,
      "grad_norm": 0.15361367166042328,
      "learning_rate": 7.016642530777162e-10,
      "loss": 0.0062,
      "step": 3290
    },
    {
      "epoch": 4.9971563981042655,
      "step": 3295,
      "total_flos": 8.270218162679316e+17,
      "train_loss": 0.045199701213737176,
      "train_runtime": 35768.8652,
      "train_samples_per_second": 1.475,
      "train_steps_per_second": 0.092
    }
  ],
  "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": 8.270218162679316e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}