{
  "best_global_step": 4000,
  "best_metric": 0.8547587990760803,
  "best_model_checkpoint": "./results/checkpoint-4000",
  "epoch": 0.08894862722618647,
  "eval_steps": 500,
  "global_step": 4500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00019766361605819217,
      "grad_norm": 12.468854904174805,
      "learning_rate": 9e-07,
      "loss": 3.6886,
      "step": 10
    },
    {
      "epoch": 0.00039532723211638433,
      "grad_norm": 52.02771759033203,
      "learning_rate": 1.9e-06,
      "loss": 3.597,
      "step": 20
    },
    {
      "epoch": 0.0005929908481745765,
      "grad_norm": 27.688196182250977,
      "learning_rate": 2.9e-06,
      "loss": 3.8032,
      "step": 30
    },
    {
      "epoch": 0.0007906544642327687,
      "grad_norm": 12.250176429748535,
      "learning_rate": 3.9e-06,
      "loss": 3.3572,
      "step": 40
    },
    {
      "epoch": 0.000988318080290961,
      "grad_norm": 20.915935516357422,
      "learning_rate": 4.9000000000000005e-06,
      "loss": 3.3315,
      "step": 50
    },
    {
      "epoch": 0.001185981696349153,
      "grad_norm": 11.775025367736816,
      "learning_rate": 5.9e-06,
      "loss": 2.7832,
      "step": 60
    },
    {
      "epoch": 0.0013836453124073452,
      "grad_norm": 12.706207275390625,
      "learning_rate": 6.900000000000001e-06,
      "loss": 2.3718,
      "step": 70
    },
    {
      "epoch": 0.0015813089284655373,
      "grad_norm": 21.651674270629883,
      "learning_rate": 7.9e-06,
      "loss": 2.0937,
      "step": 80
    },
    {
      "epoch": 0.0017789725445237295,
      "grad_norm": 15.798470497131348,
      "learning_rate": 8.9e-06,
      "loss": 1.7039,
      "step": 90
    },
    {
      "epoch": 0.001976636160581922,
      "grad_norm": 9.725250244140625,
      "learning_rate": 9.900000000000002e-06,
      "loss": 1.5623,
      "step": 100
    },
    {
      "epoch": 0.002174299776640114,
      "grad_norm": 3.9574215412139893,
      "learning_rate": 1.09e-05,
      "loss": 1.6024,
      "step": 110
    },
    {
      "epoch": 0.002371963392698306,
      "grad_norm": 8.425467491149902,
      "learning_rate": 1.19e-05,
      "loss": 1.0913,
      "step": 120
    },
    {
      "epoch": 0.0025696270087564982,
      "grad_norm": 9.648313522338867,
      "learning_rate": 1.29e-05,
      "loss": 1.3395,
      "step": 130
    },
    {
      "epoch": 0.0027672906248146904,
      "grad_norm": 0.9009546041488647,
      "learning_rate": 1.3900000000000002e-05,
      "loss": 0.999,
      "step": 140
    },
    {
      "epoch": 0.0029649542408728825,
      "grad_norm": 10.469375610351562,
      "learning_rate": 1.49e-05,
      "loss": 0.8274,
      "step": 150
    },
    {
      "epoch": 0.0031626178569310747,
      "grad_norm": 0.22269444167613983,
      "learning_rate": 1.59e-05,
      "loss": 0.8564,
      "step": 160
    },
    {
      "epoch": 0.003360281472989267,
      "grad_norm": 7.279628276824951,
      "learning_rate": 1.69e-05,
      "loss": 1.0921,
      "step": 170
    },
    {
      "epoch": 0.003557945089047459,
      "grad_norm": 0.6602649092674255,
      "learning_rate": 1.79e-05,
      "loss": 0.9095,
      "step": 180
    },
    {
      "epoch": 0.003755608705105651,
      "grad_norm": 0.1962263137102127,
      "learning_rate": 1.8900000000000002e-05,
      "loss": 0.6642,
      "step": 190
    },
    {
      "epoch": 0.003953272321163844,
      "grad_norm": 0.7642993927001953,
      "learning_rate": 1.9900000000000003e-05,
      "loss": 0.8014,
      "step": 200
    },
    {
      "epoch": 0.004150935937222036,
      "grad_norm": 0.7219191193580627,
      "learning_rate": 2.09e-05,
      "loss": 1.0484,
      "step": 210
    },
    {
      "epoch": 0.004348599553280228,
      "grad_norm": 7.698592185974121,
      "learning_rate": 2.19e-05,
      "loss": 1.2609,
      "step": 220
    },
    {
      "epoch": 0.00454626316933842,
      "grad_norm": 0.9080265760421753,
      "learning_rate": 2.29e-05,
      "loss": 0.9157,
      "step": 230
    },
    {
      "epoch": 0.004743926785396612,
      "grad_norm": 0.198422372341156,
      "learning_rate": 2.39e-05,
      "loss": 1.2826,
      "step": 240
    },
    {
      "epoch": 0.004941590401454804,
      "grad_norm": 0.267986923456192,
      "learning_rate": 2.4900000000000002e-05,
      "loss": 1.0538,
      "step": 250
    },
    {
      "epoch": 0.0051392540175129965,
      "grad_norm": 9.535926818847656,
      "learning_rate": 2.5900000000000003e-05,
      "loss": 0.8342,
      "step": 260
    },
    {
      "epoch": 0.005336917633571189,
      "grad_norm": 0.20855465531349182,
      "learning_rate": 2.6900000000000003e-05,
      "loss": 0.4623,
      "step": 270
    },
    {
      "epoch": 0.005534581249629381,
      "grad_norm": 0.4157664477825165,
      "learning_rate": 2.7900000000000004e-05,
      "loss": 0.5038,
      "step": 280
    },
    {
      "epoch": 0.005732244865687573,
      "grad_norm": 8.263617515563965,
      "learning_rate": 2.8899999999999998e-05,
      "loss": 0.8435,
      "step": 290
    },
    {
      "epoch": 0.005929908481745765,
      "grad_norm": 9.216032981872559,
      "learning_rate": 2.9900000000000002e-05,
      "loss": 0.7289,
      "step": 300
    },
    {
      "epoch": 0.006127572097803957,
      "grad_norm": 8.40715503692627,
      "learning_rate": 3.09e-05,
      "loss": 1.1021,
      "step": 310
    },
    {
      "epoch": 0.006325235713862149,
      "grad_norm": 0.4089910089969635,
      "learning_rate": 3.19e-05,
      "loss": 0.7686,
      "step": 320
    },
    {
      "epoch": 0.0065228993299203415,
      "grad_norm": 8.227189064025879,
      "learning_rate": 3.29e-05,
      "loss": 0.9375,
      "step": 330
    },
    {
      "epoch": 0.006720562945978534,
      "grad_norm": 8.776185989379883,
      "learning_rate": 3.3900000000000004e-05,
      "loss": 0.9814,
      "step": 340
    },
    {
      "epoch": 0.006918226562036726,
      "grad_norm": 7.721053123474121,
      "learning_rate": 3.49e-05,
      "loss": 0.4118,
      "step": 350
    },
    {
      "epoch": 0.007115890178094918,
      "grad_norm": 5.676611423492432,
      "learning_rate": 3.59e-05,
      "loss": 1.1953,
      "step": 360
    },
    {
      "epoch": 0.00731355379415311,
      "grad_norm": 7.924537658691406,
      "learning_rate": 3.69e-05,
      "loss": 1.1895,
      "step": 370
    },
    {
      "epoch": 0.007511217410211302,
      "grad_norm": 8.677846908569336,
      "learning_rate": 3.79e-05,
      "loss": 0.8925,
      "step": 380
    },
    {
      "epoch": 0.007708881026269494,
      "grad_norm": 6.243840217590332,
      "learning_rate": 3.8900000000000004e-05,
      "loss": 0.6888,
      "step": 390
    },
    {
      "epoch": 0.007906544642327687,
      "grad_norm": 7.967802047729492,
      "learning_rate": 3.99e-05,
      "loss": 1.4355,
      "step": 400
    },
    {
      "epoch": 0.008104208258385879,
      "grad_norm": 8.38342571258545,
      "learning_rate": 4.09e-05,
      "loss": 1.0131,
      "step": 410
    },
    {
      "epoch": 0.008301871874444072,
      "grad_norm": 0.07280826568603516,
      "learning_rate": 4.19e-05,
      "loss": 1.3299,
      "step": 420
    },
    {
      "epoch": 0.008499535490502263,
      "grad_norm": 0.2540580630302429,
      "learning_rate": 4.29e-05,
      "loss": 0.6579,
      "step": 430
    },
    {
      "epoch": 0.008697199106560456,
      "grad_norm": 7.481705188751221,
      "learning_rate": 4.39e-05,
      "loss": 0.9635,
      "step": 440
    },
    {
      "epoch": 0.008894862722618647,
      "grad_norm": 0.20026662945747375,
      "learning_rate": 4.49e-05,
      "loss": 0.883,
      "step": 450
    },
    {
      "epoch": 0.00909252633867684,
      "grad_norm": 6.689612865447998,
      "learning_rate": 4.5900000000000004e-05,
      "loss": 0.9595,
      "step": 460
    },
    {
      "epoch": 0.009290189954735031,
      "grad_norm": 0.06714965403079987,
      "learning_rate": 4.69e-05,
      "loss": 0.7816,
      "step": 470
    },
    {
      "epoch": 0.009487853570793224,
      "grad_norm": 7.251598834991455,
      "learning_rate": 4.79e-05,
      "loss": 0.5981,
      "step": 480
    },
    {
      "epoch": 0.009685517186851416,
      "grad_norm": 6.720303058624268,
      "learning_rate": 4.89e-05,
      "loss": 0.9463,
      "step": 490
    },
    {
      "epoch": 0.009883180802909609,
      "grad_norm": 6.996522903442383,
      "learning_rate": 4.99e-05,
      "loss": 0.7919,
      "step": 500
    },
    {
      "epoch": 0.009883180802909609,
      "eval_loss": 0.9214622378349304,
      "eval_runtime": 245.7679,
      "eval_samples_per_second": 86.683,
      "eval_steps_per_second": 10.835,
      "step": 500
    },
    {
      "epoch": 0.0100808444189678,
      "grad_norm": 0.3730940818786621,
      "learning_rate": 4.9997025245747756e-05,
      "loss": 0.8747,
      "step": 510
    },
    {
      "epoch": 0.010278508035025993,
      "grad_norm": 6.384608268737793,
      "learning_rate": 4.999371996324526e-05,
      "loss": 1.1554,
      "step": 520
    },
    {
      "epoch": 0.010476171651084184,
      "grad_norm": 6.133856773376465,
      "learning_rate": 4.999041468074276e-05,
      "loss": 1.1372,
      "step": 530
    },
    {
      "epoch": 0.010673835267142377,
      "grad_norm": 0.2688162624835968,
      "learning_rate": 4.998710939824027e-05,
      "loss": 0.9546,
      "step": 540
    },
    {
      "epoch": 0.010871498883200569,
      "grad_norm": 7.641685962677002,
      "learning_rate": 4.9983804115737774e-05,
      "loss": 0.9482,
      "step": 550
    },
    {
      "epoch": 0.011069162499258762,
      "grad_norm": 6.708981037139893,
      "learning_rate": 4.998049883323528e-05,
      "loss": 0.6834,
      "step": 560
    },
    {
      "epoch": 0.011266826115316953,
      "grad_norm": 7.363180637359619,
      "learning_rate": 4.997719355073278e-05,
      "loss": 0.7817,
      "step": 570
    },
    {
      "epoch": 0.011464489731375146,
      "grad_norm": 6.942836284637451,
      "learning_rate": 4.997388826823029e-05,
      "loss": 1.0409,
      "step": 580
    },
    {
      "epoch": 0.011662153347433339,
      "grad_norm": 5.417139530181885,
      "learning_rate": 4.997058298572779e-05,
      "loss": 1.1497,
      "step": 590
    },
    {
      "epoch": 0.01185981696349153,
      "grad_norm": 0.11549051851034164,
      "learning_rate": 4.9967277703225295e-05,
      "loss": 0.671,
      "step": 600
    },
    {
      "epoch": 0.012057480579549723,
      "grad_norm": 6.788494110107422,
      "learning_rate": 4.99639724207228e-05,
      "loss": 1.1224,
      "step": 610
    },
    {
      "epoch": 0.012255144195607914,
      "grad_norm": 7.265660285949707,
      "learning_rate": 4.996066713822031e-05,
      "loss": 0.7574,
      "step": 620
    },
    {
      "epoch": 0.012452807811666107,
      "grad_norm": 6.8422441482543945,
      "learning_rate": 4.995736185571781e-05,
      "loss": 0.668,
      "step": 630
    },
    {
      "epoch": 0.012650471427724299,
      "grad_norm": 6.232938289642334,
      "learning_rate": 4.995405657321531e-05,
      "loss": 1.0247,
      "step": 640
    },
    {
      "epoch": 0.012848135043782492,
      "grad_norm": 0.2170286625623703,
      "learning_rate": 4.995075129071282e-05,
      "loss": 0.8753,
      "step": 650
    },
    {
      "epoch": 0.013045798659840683,
      "grad_norm": 5.531427383422852,
      "learning_rate": 4.9947446008210326e-05,
      "loss": 1.1531,
      "step": 660
    },
    {
      "epoch": 0.013243462275898876,
      "grad_norm": 6.2044548988342285,
      "learning_rate": 4.994414072570783e-05,
      "loss": 0.8807,
      "step": 670
    },
    {
      "epoch": 0.013441125891957067,
      "grad_norm": 6.0426836013793945,
      "learning_rate": 4.994083544320533e-05,
      "loss": 1.0564,
      "step": 680
    },
    {
      "epoch": 0.01363878950801526,
      "grad_norm": 0.18100695312023163,
      "learning_rate": 4.993753016070284e-05,
      "loss": 0.845,
      "step": 690
    },
    {
      "epoch": 0.013836453124073451,
      "grad_norm": 5.483642101287842,
      "learning_rate": 4.9934224878200344e-05,
      "loss": 1.1766,
      "step": 700
    },
    {
      "epoch": 0.014034116740131644,
      "grad_norm": 4.922057628631592,
      "learning_rate": 4.993091959569785e-05,
      "loss": 1.0413,
      "step": 710
    },
    {
      "epoch": 0.014231780356189836,
      "grad_norm": 5.475461959838867,
      "learning_rate": 4.9927614313195356e-05,
      "loss": 0.9466,
      "step": 720
    },
    {
      "epoch": 0.014429443972248029,
      "grad_norm": 6.56538200378418,
      "learning_rate": 4.992430903069286e-05,
      "loss": 0.9341,
      "step": 730
    },
    {
      "epoch": 0.01462710758830622,
      "grad_norm": 4.976065635681152,
      "learning_rate": 4.992100374819036e-05,
      "loss": 0.8329,
      "step": 740
    },
    {
      "epoch": 0.014824771204364413,
      "grad_norm": 0.05167540907859802,
      "learning_rate": 4.9917698465687865e-05,
      "loss": 0.698,
      "step": 750
    },
    {
      "epoch": 0.015022434820422604,
      "grad_norm": 0.05338999629020691,
      "learning_rate": 4.991439318318537e-05,
      "loss": 0.4403,
      "step": 760
    },
    {
      "epoch": 0.015220098436480797,
      "grad_norm": 6.14265251159668,
      "learning_rate": 4.991108790068287e-05,
      "loss": 0.9864,
      "step": 770
    },
    {
      "epoch": 0.015417762052538989,
      "grad_norm": 0.1925213485956192,
      "learning_rate": 4.9907782618180374e-05,
      "loss": 1.1037,
      "step": 780
    },
    {
      "epoch": 0.015615425668597182,
      "grad_norm": 5.5837554931640625,
      "learning_rate": 4.990447733567788e-05,
      "loss": 0.8571,
      "step": 790
    },
    {
      "epoch": 0.015813089284655375,
      "grad_norm": 3.6623053550720215,
      "learning_rate": 4.9901172053175386e-05,
      "loss": 0.9082,
      "step": 800
    },
    {
      "epoch": 0.016010752900713566,
      "grad_norm": 0.05198395252227783,
      "learning_rate": 4.989786677067289e-05,
      "loss": 0.8335,
      "step": 810
    },
    {
      "epoch": 0.016208416516771757,
      "grad_norm": 5.30173921585083,
      "learning_rate": 4.989456148817039e-05,
      "loss": 0.6003,
      "step": 820
    },
    {
      "epoch": 0.01640608013282995,
      "grad_norm": 4.080887317657471,
      "learning_rate": 4.98912562056679e-05,
      "loss": 0.7937,
      "step": 830
    },
    {
      "epoch": 0.016603743748888143,
      "grad_norm": 5.606449604034424,
      "learning_rate": 4.9887950923165404e-05,
      "loss": 1.2462,
      "step": 840
    },
    {
      "epoch": 0.016801407364946334,
      "grad_norm": 6.573556900024414,
      "learning_rate": 4.988464564066291e-05,
      "loss": 0.819,
      "step": 850
    },
    {
      "epoch": 0.016999070981004526,
      "grad_norm": 0.06031334027647972,
      "learning_rate": 4.988134035816042e-05,
      "loss": 0.6425,
      "step": 860
    },
    {
      "epoch": 0.017196734597062717,
      "grad_norm": 4.252384662628174,
      "learning_rate": 4.987803507565792e-05,
      "loss": 0.8228,
      "step": 870
    },
    {
      "epoch": 0.01739439821312091,
      "grad_norm": 0.12518936395645142,
      "learning_rate": 4.987472979315542e-05,
      "loss": 0.7576,
      "step": 880
    },
    {
      "epoch": 0.017592061829179103,
      "grad_norm": 5.032276630401611,
      "learning_rate": 4.9871424510652925e-05,
      "loss": 0.8783,
      "step": 890
    },
    {
      "epoch": 0.017789725445237294,
      "grad_norm": 0.059717316180467606,
      "learning_rate": 4.9868119228150435e-05,
      "loss": 0.8153,
      "step": 900
    },
    {
      "epoch": 0.01798738906129549,
      "grad_norm": 3.093669891357422,
      "learning_rate": 4.986481394564794e-05,
      "loss": 0.6105,
      "step": 910
    },
    {
      "epoch": 0.01818505267735368,
      "grad_norm": 3.593827962875366,
      "learning_rate": 4.986150866314544e-05,
      "loss": 0.6549,
      "step": 920
    },
    {
      "epoch": 0.01838271629341187,
      "grad_norm": 0.11692095547914505,
      "learning_rate": 4.985820338064295e-05,
      "loss": 1.0544,
      "step": 930
    },
    {
      "epoch": 0.018580379909470063,
      "grad_norm": 0.07805492728948593,
      "learning_rate": 4.985489809814045e-05,
      "loss": 0.8575,
      "step": 940
    },
    {
      "epoch": 0.018778043525528258,
      "grad_norm": 3.621807098388672,
      "learning_rate": 4.9851592815637956e-05,
      "loss": 1.1574,
      "step": 950
    },
    {
      "epoch": 0.01897570714158645,
      "grad_norm": 0.1467035412788391,
      "learning_rate": 4.984828753313546e-05,
      "loss": 0.9261,
      "step": 960
    },
    {
      "epoch": 0.01917337075764464,
      "grad_norm": 5.1112895011901855,
      "learning_rate": 4.984498225063297e-05,
      "loss": 1.1392,
      "step": 970
    },
    {
      "epoch": 0.01937103437370283,
      "grad_norm": 4.554998397827148,
      "learning_rate": 4.984167696813047e-05,
      "loss": 1.0693,
      "step": 980
    },
    {
      "epoch": 0.019568697989761026,
      "grad_norm": 0.08779212087392807,
      "learning_rate": 4.9838371685627974e-05,
      "loss": 0.6191,
      "step": 990
    },
    {
      "epoch": 0.019766361605819217,
      "grad_norm": 0.050688233226537704,
      "learning_rate": 4.983506640312548e-05,
      "loss": 0.9877,
      "step": 1000
    },
    {
      "epoch": 0.019766361605819217,
      "eval_loss": 0.9038341045379639,
      "eval_runtime": 236.3116,
      "eval_samples_per_second": 90.152,
      "eval_steps_per_second": 11.269,
      "step": 1000
    },
    {
      "epoch": 0.01996402522187741,
      "grad_norm": 0.0693553164601326,
      "learning_rate": 4.983176112062298e-05,
      "loss": 1.1317,
      "step": 1010
    },
    {
      "epoch": 0.0201616888379356,
      "grad_norm": 4.626270771026611,
      "learning_rate": 4.982845583812048e-05,
      "loss": 0.9338,
      "step": 1020
    },
    {
      "epoch": 0.020359352453993795,
      "grad_norm": 4.608981132507324,
      "learning_rate": 4.9825150555617986e-05,
      "loss": 1.0849,
      "step": 1030
    },
    {
      "epoch": 0.020557016070051986,
      "grad_norm": 4.660639762878418,
      "learning_rate": 4.9821845273115495e-05,
      "loss": 1.4633,
      "step": 1040
    },
    {
      "epoch": 0.020754679686110177,
      "grad_norm": 4.729643821716309,
      "learning_rate": 4.9818539990613e-05,
      "loss": 1.0748,
      "step": 1050
    },
    {
      "epoch": 0.02095234330216837,
      "grad_norm": 0.027232196182012558,
      "learning_rate": 4.98152347081105e-05,
      "loss": 0.7562,
      "step": 1060
    },
    {
      "epoch": 0.021150006918226563,
      "grad_norm": 0.018449297174811363,
      "learning_rate": 4.981192942560801e-05,
      "loss": 0.4423,
      "step": 1070
    },
    {
      "epoch": 0.021347670534284754,
      "grad_norm": 3.372770309448242,
      "learning_rate": 4.9808624143105514e-05,
      "loss": 0.7371,
      "step": 1080
    },
    {
      "epoch": 0.021545334150342946,
      "grad_norm": 0.24770286679267883,
      "learning_rate": 4.9805318860603016e-05,
      "loss": 1.4439,
      "step": 1090
    },
    {
      "epoch": 0.021742997766401137,
      "grad_norm": 0.11222311854362488,
      "learning_rate": 4.980201357810052e-05,
      "loss": 0.6073,
      "step": 1100
    },
    {
      "epoch": 0.021940661382459332,
      "grad_norm": 4.456133842468262,
      "learning_rate": 4.979870829559803e-05,
      "loss": 0.6697,
      "step": 1110
    },
    {
      "epoch": 0.022138324998517523,
      "grad_norm": 0.05747058987617493,
      "learning_rate": 4.979540301309553e-05,
      "loss": 0.949,
      "step": 1120
    },
    {
      "epoch": 0.022335988614575714,
      "grad_norm": 3.084164619445801,
      "learning_rate": 4.9792097730593035e-05,
      "loss": 0.8189,
      "step": 1130
    },
    {
      "epoch": 0.022533652230633906,
      "grad_norm": 3.456510305404663,
      "learning_rate": 4.9788792448090544e-05,
      "loss": 1.3608,
      "step": 1140
    },
    {
      "epoch": 0.0227313158466921,
      "grad_norm": 3.4595232009887695,
      "learning_rate": 4.978548716558805e-05,
      "loss": 0.8679,
      "step": 1150
    },
    {
      "epoch": 0.02292897946275029,
      "grad_norm": 4.478796005249023,
      "learning_rate": 4.978218188308555e-05,
      "loss": 0.8229,
      "step": 1160
    },
    {
      "epoch": 0.023126643078808483,
      "grad_norm": 0.07277049869298935,
      "learning_rate": 4.977887660058305e-05,
      "loss": 0.5847,
      "step": 1170
    },
    {
      "epoch": 0.023324306694866678,
      "grad_norm": 0.13703665137290955,
      "learning_rate": 4.977557131808056e-05,
      "loss": 1.0406,
      "step": 1180
    },
    {
      "epoch": 0.02352197031092487,
      "grad_norm": 3.7098865509033203,
      "learning_rate": 4.9772266035578065e-05,
      "loss": 0.4147,
      "step": 1190
    },
    {
      "epoch": 0.02371963392698306,
      "grad_norm": 0.03374309837818146,
      "learning_rate": 4.976896075307557e-05,
      "loss": 0.8773,
      "step": 1200
    },
    {
      "epoch": 0.02391729754304125,
      "grad_norm": 4.257831573486328,
      "learning_rate": 4.976565547057308e-05,
      "loss": 0.6989,
      "step": 1210
    },
    {
      "epoch": 0.024114961159099446,
      "grad_norm": 4.520907402038574,
      "learning_rate": 4.976235018807058e-05,
      "loss": 0.6721,
      "step": 1220
    },
    {
      "epoch": 0.024312624775157637,
      "grad_norm": 7.800019264221191,
      "learning_rate": 4.9759044905568084e-05,
      "loss": 0.9356,
      "step": 1230
    },
    {
      "epoch": 0.02451028839121583,
      "grad_norm": 3.6012537479400635,
      "learning_rate": 4.9755739623065586e-05,
      "loss": 0.9297,
      "step": 1240
    },
    {
      "epoch": 0.02470795200727402,
      "grad_norm": 4.273988246917725,
      "learning_rate": 4.975243434056309e-05,
      "loss": 1.1335,
      "step": 1250
    },
    {
      "epoch": 0.024905615623332215,
      "grad_norm": 0.15219643712043762,
      "learning_rate": 4.974912905806059e-05,
      "loss": 0.608,
      "step": 1260
    },
    {
      "epoch": 0.025103279239390406,
      "grad_norm": 4.302224636077881,
      "learning_rate": 4.9745823775558095e-05,
      "loss": 0.974,
      "step": 1270
    },
    {
      "epoch": 0.025300942855448597,
      "grad_norm": 0.09040108323097229,
      "learning_rate": 4.9742518493055605e-05,
      "loss": 0.871,
      "step": 1280
    },
    {
      "epoch": 0.02549860647150679,
      "grad_norm": 0.08080489933490753,
      "learning_rate": 4.973921321055311e-05,
      "loss": 1.0142,
      "step": 1290
    },
    {
      "epoch": 0.025696270087564983,
      "grad_norm": 4.567043781280518,
      "learning_rate": 4.973590792805061e-05,
      "loss": 0.9593,
      "step": 1300
    },
    {
      "epoch": 0.025893933703623175,
      "grad_norm": 3.6397976875305176,
      "learning_rate": 4.973260264554811e-05,
      "loss": 1.0869,
      "step": 1310
    },
    {
      "epoch": 0.026091597319681366,
      "grad_norm": 3.9633944034576416,
      "learning_rate": 4.972929736304562e-05,
      "loss": 0.8078,
      "step": 1320
    },
    {
      "epoch": 0.026289260935739557,
      "grad_norm": 2.7522361278533936,
      "learning_rate": 4.9725992080543126e-05,
      "loss": 1.0228,
      "step": 1330
    },
    {
      "epoch": 0.026486924551797752,
      "grad_norm": 3.5321545600891113,
      "learning_rate": 4.972268679804063e-05,
      "loss": 1.1676,
      "step": 1340
    },
    {
      "epoch": 0.026684588167855943,
      "grad_norm": 3.301154851913452,
      "learning_rate": 4.971938151553814e-05,
      "loss": 0.9431,
      "step": 1350
    },
    {
      "epoch": 0.026882251783914134,
      "grad_norm": 0.0989571139216423,
      "learning_rate": 4.971607623303564e-05,
      "loss": 1.0831,
      "step": 1360
    },
    {
      "epoch": 0.027079915399972326,
      "grad_norm": 3.8333899974823,
      "learning_rate": 4.9712770950533144e-05,
      "loss": 1.3761,
      "step": 1370
    },
    {
      "epoch": 0.02727757901603052,
      "grad_norm": 3.214996099472046,
      "learning_rate": 4.970946566803065e-05,
      "loss": 0.7668,
      "step": 1380
    },
    {
      "epoch": 0.02747524263208871,
      "grad_norm": 3.960110902786255,
      "learning_rate": 4.9706160385528156e-05,
      "loss": 1.002,
      "step": 1390
    },
    {
      "epoch": 0.027672906248146903,
      "grad_norm": 4.702463150024414,
      "learning_rate": 4.970285510302566e-05,
      "loss": 1.013,
      "step": 1400
    },
    {
      "epoch": 0.027870569864205094,
      "grad_norm": 4.348130226135254,
      "learning_rate": 4.969954982052316e-05,
      "loss": 0.828,
      "step": 1410
    },
    {
      "epoch": 0.02806823348026329,
      "grad_norm": 0.1650361716747284,
      "learning_rate": 4.9696244538020665e-05,
      "loss": 0.4713,
      "step": 1420
    },
    {
      "epoch": 0.02826589709632148,
      "grad_norm": 0.0727657675743103,
      "learning_rate": 4.9692939255518175e-05,
      "loss": 0.9447,
      "step": 1430
    },
    {
      "epoch": 0.02846356071237967,
      "grad_norm": 3.5271363258361816,
      "learning_rate": 4.968963397301568e-05,
      "loss": 1.0093,
      "step": 1440
    },
    {
      "epoch": 0.028661224328437863,
      "grad_norm": 0.10453560948371887,
      "learning_rate": 4.968632869051318e-05,
      "loss": 0.7617,
      "step": 1450
    },
    {
      "epoch": 0.028858887944496057,
      "grad_norm": 2.5674312114715576,
      "learning_rate": 4.968302340801069e-05,
      "loss": 1.0254,
      "step": 1460
    },
    {
      "epoch": 0.02905655156055425,
      "grad_norm": 4.187676429748535,
      "learning_rate": 4.967971812550819e-05,
      "loss": 1.119,
      "step": 1470
    },
    {
      "epoch": 0.02925421517661244,
      "grad_norm": 0.09666266292333603,
      "learning_rate": 4.9676412843005696e-05,
      "loss": 0.6603,
      "step": 1480
    },
    {
      "epoch": 0.029451878792670635,
      "grad_norm": 3.545473098754883,
      "learning_rate": 4.96731075605032e-05,
      "loss": 1.0676,
      "step": 1490
    },
    {
      "epoch": 0.029649542408728826,
      "grad_norm": 3.4705913066864014,
      "learning_rate": 4.96698022780007e-05,
      "loss": 0.6608,
      "step": 1500
    },
    {
      "epoch": 0.029649542408728826,
      "eval_loss": 0.8907366394996643,
      "eval_runtime": 236.1612,
      "eval_samples_per_second": 90.21,
      "eval_steps_per_second": 11.276,
      "step": 1500
    },
    {
      "epoch": 0.029847206024787017,
      "grad_norm": 0.06528951227664948,
      "learning_rate": 4.9666496995498204e-05,
      "loss": 0.8131,
      "step": 1510
    },
    {
      "epoch": 0.03004486964084521,
      "grad_norm": 0.11044718325138092,
      "learning_rate": 4.966319171299571e-05,
      "loss": 0.9144,
      "step": 1520
    },
    {
      "epoch": 0.030242533256903403,
      "grad_norm": 3.092930316925049,
      "learning_rate": 4.965988643049322e-05,
      "loss": 1.0863,
      "step": 1530
    },
    {
      "epoch": 0.030440196872961595,
      "grad_norm": 3.9734549522399902,
      "learning_rate": 4.965658114799072e-05,
      "loss": 0.5064,
      "step": 1540
    },
    {
      "epoch": 0.030637860489019786,
      "grad_norm": 0.035740889608860016,
      "learning_rate": 4.965327586548822e-05,
      "loss": 1.222,
      "step": 1550
    },
    {
      "epoch": 0.030835524105077977,
      "grad_norm": 0.06910470873117447,
      "learning_rate": 4.9649970582985725e-05,
      "loss": 0.8428,
      "step": 1560
    },
    {
      "epoch": 0.031033187721136172,
      "grad_norm": 3.261096715927124,
      "learning_rate": 4.9646665300483235e-05,
      "loss": 0.9363,
      "step": 1570
    },
    {
      "epoch": 0.031230851337194363,
      "grad_norm": 0.09766729176044464,
      "learning_rate": 4.964336001798074e-05,
      "loss": 0.803,
      "step": 1580
    },
    {
      "epoch": 0.03142851495325256,
      "grad_norm": 3.2366461753845215,
      "learning_rate": 4.964005473547824e-05,
      "loss": 1.2179,
      "step": 1590
    },
    {
      "epoch": 0.03162617856931075,
      "grad_norm": 0.09953774511814117,
      "learning_rate": 4.963674945297575e-05,
      "loss": 1.0656,
      "step": 1600
    },
    {
      "epoch": 0.03182384218536894,
      "grad_norm": 3.7796168327331543,
      "learning_rate": 4.963344417047325e-05,
      "loss": 1.0625,
      "step": 1610
    },
    {
      "epoch": 0.03202150580142713,
      "grad_norm": 0.16226442158222198,
      "learning_rate": 4.9630138887970756e-05,
      "loss": 0.8619,
      "step": 1620
    },
    {
      "epoch": 0.03221916941748532,
      "grad_norm": 3.7525415420532227,
      "learning_rate": 4.962683360546826e-05,
      "loss": 0.9852,
      "step": 1630
    },
    {
      "epoch": 0.032416833033543514,
      "grad_norm": 0.07482007145881653,
      "learning_rate": 4.962352832296577e-05,
      "loss": 0.8297,
      "step": 1640
    },
    {
      "epoch": 0.032614496649601706,
      "grad_norm": 0.1131066307425499,
      "learning_rate": 4.962022304046327e-05,
      "loss": 1.1424,
      "step": 1650
    },
    {
      "epoch": 0.0328121602656599,
      "grad_norm": 0.05459194257855415,
      "learning_rate": 4.9616917757960774e-05,
      "loss": 0.7202,
      "step": 1660
    },
    {
      "epoch": 0.033009823881718095,
      "grad_norm": 3.9901282787323,
      "learning_rate": 4.9613612475458284e-05,
      "loss": 0.8532,
      "step": 1670
    },
    {
      "epoch": 0.033207487497776286,
      "grad_norm": 4.42426061630249,
      "learning_rate": 4.961030719295579e-05,
      "loss": 0.977,
      "step": 1680
    },
    {
      "epoch": 0.03340515111383448,
      "grad_norm": 0.06116814911365509,
      "learning_rate": 4.960700191045329e-05,
      "loss": 0.5627,
      "step": 1690
    },
    {
      "epoch": 0.03360281472989267,
      "grad_norm": 3.5464470386505127,
      "learning_rate": 4.960369662795079e-05,
      "loss": 0.6769,
      "step": 1700
    },
    {
      "epoch": 0.03380047834595086,
      "grad_norm": 3.3844430446624756,
      "learning_rate": 4.96003913454483e-05,
      "loss": 1.448,
      "step": 1710
    },
    {
      "epoch": 0.03399814196200905,
      "grad_norm": 3.592660427093506,
      "learning_rate": 4.9597086062945805e-05,
      "loss": 0.9505,
      "step": 1720
    },
    {
      "epoch": 0.03419580557806724,
      "grad_norm": 3.355842113494873,
      "learning_rate": 4.959378078044331e-05,
      "loss": 1.0866,
      "step": 1730
    },
    {
      "epoch": 0.034393469194125434,
      "grad_norm": 3.0316247940063477,
      "learning_rate": 4.959047549794081e-05,
      "loss": 0.6762,
      "step": 1740
    },
    {
      "epoch": 0.03459113281018363,
      "grad_norm": 0.06703350692987442,
      "learning_rate": 4.9587170215438314e-05,
      "loss": 0.828,
      "step": 1750
    },
    {
      "epoch": 0.03478879642624182,
      "grad_norm": 3.7815728187561035,
      "learning_rate": 4.9583864932935816e-05,
      "loss": 1.3015,
      "step": 1760
    },
    {
      "epoch": 0.034986460042300015,
      "grad_norm": 3.3532164096832275,
      "learning_rate": 4.958055965043332e-05,
      "loss": 1.2062,
      "step": 1770
    },
    {
      "epoch": 0.035184123658358206,
      "grad_norm": 3.1324400901794434,
      "learning_rate": 4.957725436793083e-05,
      "loss": 0.9484,
      "step": 1780
    },
    {
      "epoch": 0.0353817872744164,
      "grad_norm": 0.03721948713064194,
      "learning_rate": 4.957394908542833e-05,
      "loss": 0.782,
      "step": 1790
    },
    {
      "epoch": 0.03557945089047459,
      "grad_norm": 2.5899410247802734,
      "learning_rate": 4.9570643802925835e-05,
      "loss": 0.9005,
      "step": 1800
    },
    {
      "epoch": 0.03577711450653278,
      "grad_norm": 0.14065319299697876,
      "learning_rate": 4.9567338520423344e-05,
      "loss": 1.2058,
      "step": 1810
    },
    {
      "epoch": 0.03597477812259098,
      "grad_norm": 3.255124568939209,
      "learning_rate": 4.956403323792085e-05,
      "loss": 0.6861,
      "step": 1820
    },
    {
      "epoch": 0.03617244173864917,
      "grad_norm": 2.9060869216918945,
      "learning_rate": 4.956072795541835e-05,
      "loss": 0.6472,
      "step": 1830
    },
    {
      "epoch": 0.03637010535470736,
      "grad_norm": 0.10534641146659851,
      "learning_rate": 4.955742267291585e-05,
      "loss": 1.1794,
      "step": 1840
    },
    {
      "epoch": 0.03656776897076555,
      "grad_norm": 3.3723835945129395,
      "learning_rate": 4.955411739041336e-05,
      "loss": 1.0936,
      "step": 1850
    },
    {
      "epoch": 0.03676543258682374,
      "grad_norm": 3.3948323726654053,
      "learning_rate": 4.9550812107910865e-05,
      "loss": 1.0723,
      "step": 1860
    },
    {
      "epoch": 0.036963096202881934,
      "grad_norm": 2.8475537300109863,
      "learning_rate": 4.954750682540837e-05,
      "loss": 1.1292,
      "step": 1870
    },
    {
      "epoch": 0.037160759818940126,
      "grad_norm": 3.0720722675323486,
      "learning_rate": 4.954420154290588e-05,
      "loss": 0.4767,
      "step": 1880
    },
    {
      "epoch": 0.03735842343499832,
      "grad_norm": 2.8738229274749756,
      "learning_rate": 4.954089626040338e-05,
      "loss": 0.8313,
      "step": 1890
    },
    {
      "epoch": 0.037556087051056515,
      "grad_norm": 0.04877200722694397,
      "learning_rate": 4.9537590977900884e-05,
      "loss": 0.8162,
      "step": 1900
    },
    {
      "epoch": 0.037753750667114706,
      "grad_norm": 0.15653006732463837,
      "learning_rate": 4.9534285695398386e-05,
      "loss": 1.117,
      "step": 1910
    },
    {
      "epoch": 0.0379514142831729,
      "grad_norm": 0.09898870438337326,
      "learning_rate": 4.9530980412895896e-05,
      "loss": 0.8124,
      "step": 1920
    },
    {
      "epoch": 0.03814907789923109,
      "grad_norm": 3.6042723655700684,
      "learning_rate": 4.95276751303934e-05,
      "loss": 0.978,
      "step": 1930
    },
    {
      "epoch": 0.03834674151528928,
      "grad_norm": 2.490565776824951,
      "learning_rate": 4.95243698478909e-05,
      "loss": 0.4846,
      "step": 1940
    },
    {
      "epoch": 0.03854440513134747,
      "grad_norm": 2.3419349193573,
      "learning_rate": 4.952106456538841e-05,
      "loss": 0.7827,
      "step": 1950
    },
    {
      "epoch": 0.03874206874740566,
      "grad_norm": 2.327219009399414,
      "learning_rate": 4.9517759282885914e-05,
      "loss": 0.9868,
      "step": 1960
    },
    {
      "epoch": 0.038939732363463854,
      "grad_norm": 0.18203556537628174,
      "learning_rate": 4.951445400038342e-05,
      "loss": 0.9514,
      "step": 1970
    },
    {
      "epoch": 0.03913739597952205,
      "grad_norm": 2.613198757171631,
      "learning_rate": 4.951114871788092e-05,
      "loss": 0.7549,
      "step": 1980
    },
    {
      "epoch": 0.03933505959558024,
      "grad_norm": 4.419915676116943,
      "learning_rate": 4.950784343537842e-05,
      "loss": 0.6902,
      "step": 1990
    },
    {
      "epoch": 0.039532723211638435,
      "grad_norm": 0.04059309512376785,
      "learning_rate": 4.9504538152875926e-05,
      "loss": 0.8878,
      "step": 2000
    },
    {
      "epoch": 0.039532723211638435,
      "eval_loss": 0.8785109519958496,
      "eval_runtime": 237.1008,
      "eval_samples_per_second": 89.852,
      "eval_steps_per_second": 11.232,
      "step": 2000
    },
    {
      "epoch": 0.039730386827696626,
      "grad_norm": 2.7908315658569336,
      "learning_rate": 4.950123287037343e-05,
      "loss": 1.0453,
      "step": 2010
    },
    {
      "epoch": 0.03992805044375482,
      "grad_norm": 0.19223953783512115,
      "learning_rate": 4.949792758787094e-05,
      "loss": 0.9602,
      "step": 2020
    },
    {
      "epoch": 0.04012571405981301,
      "grad_norm": 2.7216477394104004,
      "learning_rate": 4.949462230536844e-05,
      "loss": 1.1187,
      "step": 2030
    },
    {
      "epoch": 0.0403233776758712,
      "grad_norm": 3.306227922439575,
      "learning_rate": 4.9491317022865944e-05,
      "loss": 0.9302,
      "step": 2040
    },
    {
      "epoch": 0.0405210412919294,
      "grad_norm": 3.259765386581421,
      "learning_rate": 4.948801174036345e-05,
      "loss": 0.9832,
      "step": 2050
    },
    {
      "epoch": 0.04071870490798759,
      "grad_norm": 3.6598448753356934,
      "learning_rate": 4.9484706457860956e-05,
      "loss": 1.0918,
      "step": 2060
    },
    {
      "epoch": 0.04091636852404578,
      "grad_norm": 3.6276862621307373,
      "learning_rate": 4.948140117535846e-05,
      "loss": 0.8682,
      "step": 2070
    },
    {
      "epoch": 0.04111403214010397,
      "grad_norm": 3.547147035598755,
      "learning_rate": 4.947809589285596e-05,
      "loss": 0.5364,
      "step": 2080
    },
    {
      "epoch": 0.04131169575616216,
      "grad_norm": 3.0729477405548096,
      "learning_rate": 4.947479061035347e-05,
      "loss": 1.2961,
      "step": 2090
    },
    {
      "epoch": 0.041509359372220354,
      "grad_norm": 2.370206594467163,
      "learning_rate": 4.9471485327850975e-05,
      "loss": 0.6171,
      "step": 2100
    },
    {
      "epoch": 0.041707022988278546,
      "grad_norm": 3.900785446166992,
      "learning_rate": 4.946818004534848e-05,
      "loss": 1.3057,
      "step": 2110
    },
    {
      "epoch": 0.04190468660433674,
      "grad_norm": 3.226991653442383,
      "learning_rate": 4.946487476284598e-05,
      "loss": 1.0164,
      "step": 2120
    },
    {
      "epoch": 0.042102350220394935,
      "grad_norm": 0.5991417169570923,
      "learning_rate": 4.946156948034349e-05,
      "loss": 1.4067,
      "step": 2130
    },
    {
      "epoch": 0.042300013836453126,
      "grad_norm": 3.0779223442077637,
      "learning_rate": 4.945826419784099e-05,
      "loss": 0.7529,
      "step": 2140
    },
    {
      "epoch": 0.04249767745251132,
      "grad_norm": 0.05838579684495926,
      "learning_rate": 4.9454958915338496e-05,
      "loss": 0.7606,
      "step": 2150
    },
    {
      "epoch": 0.04269534106856951,
      "grad_norm": 3.432065486907959,
      "learning_rate": 4.9451653632836005e-05,
      "loss": 1.1331,
      "step": 2160
    },
    {
      "epoch": 0.0428930046846277,
      "grad_norm": 2.9518392086029053,
      "learning_rate": 4.944834835033351e-05,
      "loss": 0.6276,
      "step": 2170
    },
    {
      "epoch": 0.04309066830068589,
      "grad_norm": 0.06247523054480553,
      "learning_rate": 4.944504306783101e-05,
      "loss": 0.7537,
      "step": 2180
    },
    {
      "epoch": 0.04328833191674408,
      "grad_norm": 2.4885005950927734,
      "learning_rate": 4.9441737785328514e-05,
      "loss": 1.1748,
      "step": 2190
    },
    {
      "epoch": 0.043485995532802274,
      "grad_norm": 0.15998195111751556,
      "learning_rate": 4.9438432502826024e-05,
      "loss": 0.9554,
      "step": 2200
    },
    {
      "epoch": 0.04368365914886047,
      "grad_norm": 3.214895248413086,
      "learning_rate": 4.9435127220323526e-05,
      "loss": 0.8503,
      "step": 2210
    },
    {
      "epoch": 0.043881322764918664,
      "grad_norm": 0.03642280027270317,
      "learning_rate": 4.943182193782103e-05,
      "loss": 0.4855,
      "step": 2220
    },
    {
      "epoch": 0.044078986380976855,
      "grad_norm": 2.707956075668335,
      "learning_rate": 4.942851665531853e-05,
      "loss": 1.0405,
      "step": 2230
    },
    {
      "epoch": 0.044276649997035046,
      "grad_norm": 0.08170021325349808,
      "learning_rate": 4.9425211372816035e-05,
      "loss": 0.9093,
      "step": 2240
    },
    {
      "epoch": 0.04447431361309324,
      "grad_norm": 2.322641611099243,
      "learning_rate": 4.942190609031354e-05,
      "loss": 1.357,
      "step": 2250
    },
    {
      "epoch": 0.04467197722915143,
      "grad_norm": 2.5577518939971924,
      "learning_rate": 4.941860080781104e-05,
      "loss": 1.2809,
      "step": 2260
    },
    {
      "epoch": 0.04486964084520962,
      "grad_norm": 2.471879243850708,
      "learning_rate": 4.941529552530855e-05,
      "loss": 1.1575,
      "step": 2270
    },
    {
      "epoch": 0.04506730446126781,
      "grad_norm": 3.389317274093628,
      "learning_rate": 4.941199024280605e-05,
      "loss": 1.1552,
      "step": 2280
    },
    {
      "epoch": 0.04526496807732601,
      "grad_norm": 0.11907307803630829,
      "learning_rate": 4.9408684960303556e-05,
      "loss": 0.685,
      "step": 2290
    },
    {
      "epoch": 0.0454626316933842,
      "grad_norm": 3.5071091651916504,
      "learning_rate": 4.940537967780106e-05,
      "loss": 0.5446,
      "step": 2300
    },
    {
      "epoch": 0.04566029530944239,
      "grad_norm": 2.466038227081299,
      "learning_rate": 4.940207439529857e-05,
      "loss": 0.9161,
      "step": 2310
    },
    {
      "epoch": 0.04585795892550058,
      "grad_norm": 2.913646936416626,
      "learning_rate": 4.939876911279607e-05,
      "loss": 0.8943,
      "step": 2320
    },
    {
      "epoch": 0.046055622541558774,
      "grad_norm": 2.976774215698242,
      "learning_rate": 4.9395463830293574e-05,
      "loss": 1.4517,
      "step": 2330
    },
    {
      "epoch": 0.046253286157616966,
      "grad_norm": 0.20807339251041412,
      "learning_rate": 4.9392158547791084e-05,
      "loss": 0.535,
      "step": 2340
    },
    {
      "epoch": 0.04645094977367516,
      "grad_norm": 3.2078752517700195,
      "learning_rate": 4.938885326528859e-05,
      "loss": 0.561,
      "step": 2350
    },
    {
      "epoch": 0.046648613389733355,
      "grad_norm": 2.7821130752563477,
      "learning_rate": 4.938554798278609e-05,
      "loss": 1.0918,
      "step": 2360
    },
    {
      "epoch": 0.046846277005791546,
      "grad_norm": 0.18904615938663483,
      "learning_rate": 4.938224270028359e-05,
      "loss": 1.1504,
      "step": 2370
    },
    {
      "epoch": 0.04704394062184974,
      "grad_norm": 2.9167234897613525,
      "learning_rate": 4.93789374177811e-05,
      "loss": 0.8307,
      "step": 2380
    },
    {
      "epoch": 0.04724160423790793,
      "grad_norm": 2.7260210514068604,
      "learning_rate": 4.9375632135278605e-05,
      "loss": 0.7618,
      "step": 2390
    },
    {
      "epoch": 0.04743926785396612,
      "grad_norm": 2.3594138622283936,
      "learning_rate": 4.937232685277611e-05,
      "loss": 0.4669,
      "step": 2400
    },
    {
      "epoch": 0.04763693147002431,
      "grad_norm": 2.587303400039673,
      "learning_rate": 4.936902157027362e-05,
      "loss": 0.6653,
      "step": 2410
    },
    {
      "epoch": 0.0478345950860825,
      "grad_norm": 0.07142865657806396,
      "learning_rate": 4.936571628777112e-05,
      "loss": 0.8185,
      "step": 2420
    },
    {
      "epoch": 0.048032258702140694,
      "grad_norm": 3.059152603149414,
      "learning_rate": 4.936241100526862e-05,
      "loss": 0.9088,
      "step": 2430
    },
    {
      "epoch": 0.04822992231819889,
      "grad_norm": 0.1070665493607521,
      "learning_rate": 4.9359105722766126e-05,
      "loss": 1.1149,
      "step": 2440
    },
    {
      "epoch": 0.048427585934257084,
      "grad_norm": 2.317112684249878,
      "learning_rate": 4.9355800440263636e-05,
      "loss": 0.5108,
      "step": 2450
    },
    {
      "epoch": 0.048625249550315275,
      "grad_norm": 2.8447506427764893,
      "learning_rate": 4.935249515776114e-05,
      "loss": 1.0359,
      "step": 2460
    },
    {
      "epoch": 0.048822913166373466,
      "grad_norm": 3.0613043308258057,
      "learning_rate": 4.934918987525864e-05,
      "loss": 0.7497,
      "step": 2470
    },
    {
      "epoch": 0.04902057678243166,
      "grad_norm": 2.9260942935943604,
      "learning_rate": 4.9345884592756144e-05,
      "loss": 0.9751,
      "step": 2480
    },
    {
      "epoch": 0.04921824039848985,
      "grad_norm": 0.17592434585094452,
      "learning_rate": 4.934257931025365e-05,
      "loss": 0.9219,
      "step": 2490
    },
    {
      "epoch": 0.04941590401454804,
      "grad_norm": 0.07734190672636032,
      "learning_rate": 4.933927402775116e-05,
      "loss": 0.2931,
      "step": 2500
    },
    {
      "epoch": 0.04941590401454804,
      "eval_loss": 0.8793515563011169,
      "eval_runtime": 237.1956,
      "eval_samples_per_second": 89.816,
      "eval_steps_per_second": 11.227,
      "step": 2500
    },
    {
      "epoch": 0.04961356763060623,
      "grad_norm": 3.786978244781494,
      "learning_rate": 4.933596874524866e-05,
      "loss": 0.8909,
      "step": 2510
    },
    {
      "epoch": 0.04981123124666443,
      "grad_norm": 0.1106736958026886,
      "learning_rate": 4.933266346274616e-05,
      "loss": 0.5019,
      "step": 2520
    },
    {
      "epoch": 0.05000889486272262,
      "grad_norm": 0.16707845032215118,
      "learning_rate": 4.9329358180243665e-05,
      "loss": 0.5127,
      "step": 2530
    },
    {
      "epoch": 0.05020655847878081,
      "grad_norm": 0.0881323516368866,
      "learning_rate": 4.932605289774117e-05,
      "loss": 0.4456,
      "step": 2540
    },
    {
      "epoch": 0.050404222094839,
      "grad_norm": 3.054690361022949,
      "learning_rate": 4.932274761523868e-05,
      "loss": 0.8947,
      "step": 2550
    },
    {
      "epoch": 0.050601885710897195,
      "grad_norm": 2.370521068572998,
      "learning_rate": 4.931944233273618e-05,
      "loss": 1.1004,
      "step": 2560
    },
    {
      "epoch": 0.050799549326955386,
      "grad_norm": 3.135436773300171,
      "learning_rate": 4.9316137050233684e-05,
      "loss": 0.7243,
      "step": 2570
    },
    {
      "epoch": 0.05099721294301358,
      "grad_norm": 3.7042624950408936,
      "learning_rate": 4.9312831767731186e-05,
      "loss": 1.1088,
      "step": 2580
    },
    {
      "epoch": 0.05119487655907177,
      "grad_norm": 2.8376870155334473,
      "learning_rate": 4.9309526485228696e-05,
      "loss": 0.7167,
      "step": 2590
    },
    {
      "epoch": 0.051392540175129967,
      "grad_norm": 0.0706353485584259,
      "learning_rate": 4.93062212027262e-05,
      "loss": 0.7822,
      "step": 2600
    },
    {
      "epoch": 0.05159020379118816,
      "grad_norm": 2.7571935653686523,
      "learning_rate": 4.93029159202237e-05,
      "loss": 1.2513,
      "step": 2610
    },
    {
      "epoch": 0.05178786740724635,
      "grad_norm": 0.07373715192079544,
      "learning_rate": 4.929961063772121e-05,
      "loss": 0.341,
      "step": 2620
    },
    {
      "epoch": 0.05198553102330454,
      "grad_norm": 2.6552443504333496,
      "learning_rate": 4.9296305355218714e-05,
      "loss": 0.4626,
      "step": 2630
    },
    {
      "epoch": 0.05218319463936273,
      "grad_norm": 0.07399863749742508,
      "learning_rate": 4.929300007271622e-05,
      "loss": 0.8852,
      "step": 2640
    },
    {
      "epoch": 0.05238085825542092,
      "grad_norm": 0.12832652032375336,
      "learning_rate": 4.928969479021372e-05,
      "loss": 0.9477,
      "step": 2650
    },
    {
      "epoch": 0.052578521871479114,
      "grad_norm": 3.3986313343048096,
      "learning_rate": 4.928638950771123e-05,
      "loss": 0.8851,
      "step": 2660
    },
    {
      "epoch": 0.05277618548753731,
      "grad_norm": 2.217534303665161,
      "learning_rate": 4.928308422520873e-05,
      "loss": 0.8122,
      "step": 2670
    },
    {
      "epoch": 0.052973849103595504,
      "grad_norm": 2.807478427886963,
      "learning_rate": 4.9279778942706235e-05,
      "loss": 0.8963,
      "step": 2680
    },
    {
      "epoch": 0.053171512719653695,
      "grad_norm": 2.5803465843200684,
      "learning_rate": 4.9276473660203745e-05,
      "loss": 0.6245,
      "step": 2690
    },
    {
      "epoch": 0.053369176335711886,
      "grad_norm": 2.3707590103149414,
      "learning_rate": 4.927316837770125e-05,
      "loss": 0.5902,
      "step": 2700
    },
    {
      "epoch": 0.05356683995177008,
      "grad_norm": 4.263098239898682,
      "learning_rate": 4.926986309519875e-05,
      "loss": 1.3733,
      "step": 2710
    },
    {
      "epoch": 0.05376450356782827,
      "grad_norm": 2.7389602661132812,
      "learning_rate": 4.9266557812696254e-05,
      "loss": 1.0306,
      "step": 2720
    },
    {
      "epoch": 0.05396216718388646,
      "grad_norm": 2.0722873210906982,
      "learning_rate": 4.9263252530193756e-05,
      "loss": 0.906,
      "step": 2730
    },
    {
      "epoch": 0.05415983079994465,
      "grad_norm": 0.11271070688962936,
      "learning_rate": 4.9259947247691266e-05,
      "loss": 0.6766,
      "step": 2740
    },
    {
      "epoch": 0.05435749441600285,
      "grad_norm": 3.4508349895477295,
      "learning_rate": 4.925664196518877e-05,
      "loss": 1.1286,
      "step": 2750
    },
    {
      "epoch": 0.05455515803206104,
      "grad_norm": 2.2231955528259277,
      "learning_rate": 4.925333668268627e-05,
      "loss": 0.9619,
      "step": 2760
    },
    {
      "epoch": 0.05475282164811923,
      "grad_norm": 2.770782232284546,
      "learning_rate": 4.9250031400183775e-05,
      "loss": 1.1384,
      "step": 2770
    },
    {
      "epoch": 0.05495048526417742,
      "grad_norm": 2.6763529777526855,
      "learning_rate": 4.924672611768128e-05,
      "loss": 0.6292,
      "step": 2780
    },
    {
      "epoch": 0.055148148880235615,
      "grad_norm": 0.09729930013418198,
      "learning_rate": 4.924342083517878e-05,
      "loss": 1.0811,
      "step": 2790
    },
    {
      "epoch": 0.055345812496293806,
      "grad_norm": 2.2069194316864014,
      "learning_rate": 4.924011555267629e-05,
      "loss": 0.9206,
      "step": 2800
    },
    {
      "epoch": 0.055543476112352,
      "grad_norm": 0.0986907035112381,
      "learning_rate": 4.923681027017379e-05,
      "loss": 0.7723,
      "step": 2810
    },
    {
      "epoch": 0.05574113972841019,
      "grad_norm": 0.09563972800970078,
      "learning_rate": 4.9233504987671296e-05,
      "loss": 0.5127,
      "step": 2820
    },
    {
      "epoch": 0.05593880334446839,
      "grad_norm": 2.910877227783203,
      "learning_rate": 4.9230199705168805e-05,
      "loss": 1.0938,
      "step": 2830
    },
    {
      "epoch": 0.05613646696052658,
      "grad_norm": 0.17678514122962952,
      "learning_rate": 4.922689442266631e-05,
      "loss": 1.0306,
      "step": 2840
    },
    {
      "epoch": 0.05633413057658477,
      "grad_norm": 2.2762157917022705,
      "learning_rate": 4.922358914016381e-05,
      "loss": 0.8836,
      "step": 2850
    },
    {
      "epoch": 0.05653179419264296,
      "grad_norm": 2.3193490505218506,
      "learning_rate": 4.9220283857661314e-05,
      "loss": 1.039,
      "step": 2860
    },
    {
      "epoch": 0.05672945780870115,
      "grad_norm": 0.2969821095466614,
      "learning_rate": 4.9216978575158824e-05,
      "loss": 0.9778,
      "step": 2870
    },
    {
      "epoch": 0.05692712142475934,
      "grad_norm": 3.387187957763672,
      "learning_rate": 4.9213673292656326e-05,
      "loss": 1.194,
      "step": 2880
    },
    {
      "epoch": 0.057124785040817534,
      "grad_norm": 0.2599051296710968,
      "learning_rate": 4.921036801015383e-05,
      "loss": 0.7274,
      "step": 2890
    },
    {
      "epoch": 0.057322448656875725,
      "grad_norm": 2.5939202308654785,
      "learning_rate": 4.920706272765134e-05,
      "loss": 1.091,
      "step": 2900
    },
    {
      "epoch": 0.057520112272933924,
      "grad_norm": 2.483778953552246,
      "learning_rate": 4.920375744514884e-05,
      "loss": 1.2362,
      "step": 2910
    },
    {
      "epoch": 0.057717775888992115,
      "grad_norm": 0.09381911158561707,
      "learning_rate": 4.9200452162646345e-05,
      "loss": 0.5963,
      "step": 2920
    },
    {
      "epoch": 0.057915439505050306,
      "grad_norm": 1.821659803390503,
      "learning_rate": 4.919714688014385e-05,
      "loss": 0.8208,
      "step": 2930
    },
    {
      "epoch": 0.0581131031211085,
      "grad_norm": 2.4229607582092285,
      "learning_rate": 4.919384159764136e-05,
      "loss": 0.6724,
      "step": 2940
    },
    {
      "epoch": 0.05831076673716669,
      "grad_norm": 2.8029041290283203,
      "learning_rate": 4.919053631513886e-05,
      "loss": 0.8238,
      "step": 2950
    },
    {
      "epoch": 0.05850843035322488,
      "grad_norm": 0.0861673504114151,
      "learning_rate": 4.918723103263636e-05,
      "loss": 0.895,
      "step": 2960
    },
    {
      "epoch": 0.05870609396928307,
      "grad_norm": 2.7137832641601562,
      "learning_rate": 4.9183925750133866e-05,
      "loss": 0.6717,
      "step": 2970
    },
    {
      "epoch": 0.05890375758534127,
      "grad_norm": 0.07592540979385376,
      "learning_rate": 4.9180620467631375e-05,
      "loss": 0.4923,
      "step": 2980
    },
    {
      "epoch": 0.05910142120139946,
      "grad_norm": 2.685272693634033,
      "learning_rate": 4.917731518512888e-05,
      "loss": 0.9596,
      "step": 2990
    },
    {
      "epoch": 0.05929908481745765,
      "grad_norm": 2.9105234146118164,
      "learning_rate": 4.917400990262638e-05,
      "loss": 0.912,
      "step": 3000
    },
    {
      "epoch": 0.05929908481745765,
      "eval_loss": 0.8658507466316223,
      "eval_runtime": 237.4787,
      "eval_samples_per_second": 89.709,
      "eval_steps_per_second": 11.214,
      "step": 3000
    },
    {
      "epoch": 0.05949674843351584,
      "grad_norm": 0.28898268938064575,
      "learning_rate": 4.9170704620123884e-05,
      "loss": 1.0526,
      "step": 3010
    },
    {
      "epoch": 0.059694412049574035,
      "grad_norm": 2.529243230819702,
      "learning_rate": 4.916739933762139e-05,
      "loss": 0.7999,
      "step": 3020
    },
    {
      "epoch": 0.059892075665632226,
      "grad_norm": 2.279623031616211,
      "learning_rate": 4.916409405511889e-05,
      "loss": 0.7309,
      "step": 3030
    },
    {
      "epoch": 0.06008973928169042,
      "grad_norm": 0.049957409501075745,
      "learning_rate": 4.916078877261639e-05,
      "loss": 0.8589,
      "step": 3040
    },
    {
      "epoch": 0.06028740289774861,
      "grad_norm": 2.42553448677063,
      "learning_rate": 4.91574834901139e-05,
      "loss": 0.7888,
      "step": 3050
    },
    {
      "epoch": 0.06048506651380681,
      "grad_norm": 2.584897518157959,
      "learning_rate": 4.9154178207611405e-05,
      "loss": 0.8402,
      "step": 3060
    },
    {
      "epoch": 0.060682730129865,
      "grad_norm": 0.06790557503700256,
      "learning_rate": 4.915087292510891e-05,
      "loss": 0.357,
      "step": 3070
    },
    {
      "epoch": 0.06088039374592319,
      "grad_norm": 2.327202558517456,
      "learning_rate": 4.914756764260642e-05,
      "loss": 0.9274,
      "step": 3080
    },
    {
      "epoch": 0.06107805736198138,
      "grad_norm": 0.0687011107802391,
      "learning_rate": 4.914426236010392e-05,
      "loss": 0.5299,
      "step": 3090
    },
    {
      "epoch": 0.06127572097803957,
      "grad_norm": 2.2109317779541016,
      "learning_rate": 4.914095707760142e-05,
      "loss": 0.7901,
      "step": 3100
    },
    {
      "epoch": 0.06147338459409776,
      "grad_norm": 2.547398090362549,
      "learning_rate": 4.9137651795098926e-05,
      "loss": 0.567,
      "step": 3110
    },
    {
      "epoch": 0.061671048210155954,
      "grad_norm": 2.282902717590332,
      "learning_rate": 4.9134346512596436e-05,
      "loss": 0.9336,
      "step": 3120
    },
    {
      "epoch": 0.061868711826214146,
      "grad_norm": 4.029672622680664,
      "learning_rate": 4.913104123009394e-05,
      "loss": 0.5854,
      "step": 3130
    },
    {
      "epoch": 0.062066375442272344,
      "grad_norm": 0.15566882491111755,
      "learning_rate": 4.912773594759144e-05,
      "loss": 0.854,
      "step": 3140
    },
    {
      "epoch": 0.062264039058330535,
      "grad_norm": 0.08527495712041855,
      "learning_rate": 4.912443066508895e-05,
      "loss": 0.2452,
      "step": 3150
    },
    {
      "epoch": 0.062461702674388726,
      "grad_norm": 0.09388203173875809,
      "learning_rate": 4.9121125382586454e-05,
      "loss": 0.8951,
      "step": 3160
    },
    {
      "epoch": 0.06265936629044691,
      "grad_norm": 3.1649973392486572,
      "learning_rate": 4.911782010008396e-05,
      "loss": 0.892,
      "step": 3170
    },
    {
      "epoch": 0.06285702990650512,
      "grad_norm": 3.3925817012786865,
      "learning_rate": 4.911451481758146e-05,
      "loss": 0.7135,
      "step": 3180
    },
    {
      "epoch": 0.06305469352256331,
      "grad_norm": 0.13573919236660004,
      "learning_rate": 4.911120953507897e-05,
      "loss": 0.5978,
      "step": 3190
    },
    {
      "epoch": 0.0632523571386215,
      "grad_norm": 2.88134765625,
      "learning_rate": 4.910790425257647e-05,
      "loss": 1.121,
      "step": 3200
    },
    {
      "epoch": 0.06345002075467969,
      "grad_norm": 0.15743722021579742,
      "learning_rate": 4.9104598970073975e-05,
      "loss": 1.0691,
      "step": 3210
    },
    {
      "epoch": 0.06364768437073788,
      "grad_norm": 2.900428056716919,
      "learning_rate": 4.9101293687571485e-05,
      "loss": 0.6394,
      "step": 3220
    },
    {
      "epoch": 0.06384534798679607,
      "grad_norm": 0.1241888701915741,
      "learning_rate": 4.909798840506899e-05,
      "loss": 0.849,
      "step": 3230
    },
    {
      "epoch": 0.06404301160285426,
      "grad_norm": 2.8092403411865234,
      "learning_rate": 4.909468312256649e-05,
      "loss": 1.059,
      "step": 3240
    },
    {
      "epoch": 0.06424067521891245,
      "grad_norm": 2.493726968765259,
      "learning_rate": 4.909137784006399e-05,
      "loss": 0.5026,
      "step": 3250
    },
    {
      "epoch": 0.06443833883497065,
      "grad_norm": 0.05199737474322319,
      "learning_rate": 4.9088072557561496e-05,
      "loss": 0.659,
      "step": 3260
    },
    {
      "epoch": 0.06463600245102884,
      "grad_norm": 2.3246049880981445,
      "learning_rate": 4.9084767275059e-05,
      "loss": 0.8188,
      "step": 3270
    },
    {
      "epoch": 0.06483366606708703,
      "grad_norm": 0.12682946026325226,
      "learning_rate": 4.90814619925565e-05,
      "loss": 0.632,
      "step": 3280
    },
    {
      "epoch": 0.06503132968314522,
      "grad_norm": 3.34828782081604,
      "learning_rate": 4.907815671005401e-05,
      "loss": 0.7688,
      "step": 3290
    },
    {
      "epoch": 0.06522899329920341,
      "grad_norm": 2.733819007873535,
      "learning_rate": 4.9074851427551514e-05,
      "loss": 1.1474,
      "step": 3300
    },
    {
      "epoch": 0.0654266569152616,
      "grad_norm": 2.8242006301879883,
      "learning_rate": 4.907154614504902e-05,
      "loss": 0.9586,
      "step": 3310
    },
    {
      "epoch": 0.0656243205313198,
      "grad_norm": 2.7257206439971924,
      "learning_rate": 4.906824086254652e-05,
      "loss": 0.5093,
      "step": 3320
    },
    {
      "epoch": 0.065821984147378,
      "grad_norm": 2.696195363998413,
      "learning_rate": 4.906493558004403e-05,
      "loss": 0.7696,
      "step": 3330
    },
    {
      "epoch": 0.06601964776343619,
      "grad_norm": 0.15057361125946045,
      "learning_rate": 4.906163029754153e-05,
      "loss": 0.9096,
      "step": 3340
    },
    {
      "epoch": 0.06621731137949438,
      "grad_norm": 2.854492425918579,
      "learning_rate": 4.9058325015039035e-05,
      "loss": 0.941,
      "step": 3350
    },
    {
      "epoch": 0.06641497499555257,
      "grad_norm": 0.09030754864215851,
      "learning_rate": 4.9055019732536545e-05,
      "loss": 0.7451,
      "step": 3360
    },
    {
      "epoch": 0.06661263861161076,
      "grad_norm": 0.10631457716226578,
      "learning_rate": 4.905171445003405e-05,
      "loss": 0.8131,
      "step": 3370
    },
    {
      "epoch": 0.06681030222766896,
      "grad_norm": 3.1166298389434814,
      "learning_rate": 4.904840916753155e-05,
      "loss": 0.569,
      "step": 3380
    },
    {
      "epoch": 0.06700796584372715,
      "grad_norm": 3.1512789726257324,
      "learning_rate": 4.9045103885029054e-05,
      "loss": 1.0931,
      "step": 3390
    },
    {
      "epoch": 0.06720562945978534,
      "grad_norm": 0.15035057067871094,
      "learning_rate": 4.904179860252656e-05,
      "loss": 0.7084,
      "step": 3400
    },
    {
      "epoch": 0.06740329307584353,
      "grad_norm": 0.11252701282501221,
      "learning_rate": 4.9038493320024066e-05,
      "loss": 0.6624,
      "step": 3410
    },
    {
      "epoch": 0.06760095669190172,
      "grad_norm": 0.09046754240989685,
      "learning_rate": 4.903518803752157e-05,
      "loss": 0.6766,
      "step": 3420
    },
    {
      "epoch": 0.06779862030795991,
      "grad_norm": 2.3272488117218018,
      "learning_rate": 4.903188275501908e-05,
      "loss": 0.9106,
      "step": 3430
    },
    {
      "epoch": 0.0679962839240181,
      "grad_norm": 2.8726844787597656,
      "learning_rate": 4.902857747251658e-05,
      "loss": 0.8752,
      "step": 3440
    },
    {
      "epoch": 0.0681939475400763,
      "grad_norm": 0.061476483941078186,
      "learning_rate": 4.9025272190014084e-05,
      "loss": 0.5522,
      "step": 3450
    },
    {
      "epoch": 0.06839161115613449,
      "grad_norm": 3.8139467239379883,
      "learning_rate": 4.902196690751159e-05,
      "loss": 0.6546,
      "step": 3460
    },
    {
      "epoch": 0.06858927477219268,
      "grad_norm": 0.09583472460508347,
      "learning_rate": 4.90186616250091e-05,
      "loss": 0.7695,
      "step": 3470
    },
    {
      "epoch": 0.06878693838825087,
      "grad_norm": 2.236619234085083,
      "learning_rate": 4.90153563425066e-05,
      "loss": 1.0962,
      "step": 3480
    },
    {
      "epoch": 0.06898460200430907,
      "grad_norm": 2.348360776901245,
      "learning_rate": 4.90120510600041e-05,
      "loss": 0.6132,
      "step": 3490
    },
    {
      "epoch": 0.06918226562036726,
      "grad_norm": 2.741769552230835,
      "learning_rate": 4.9008745777501605e-05,
      "loss": 1.1316,
      "step": 3500
    },
    {
      "epoch": 0.06918226562036726,
      "eval_loss": 0.8600906729698181,
      "eval_runtime": 237.4494,
      "eval_samples_per_second": 89.72,
      "eval_steps_per_second": 11.215,
      "step": 3500
    },
    {
      "epoch": 0.06937992923642546,
      "grad_norm": 3.465202808380127,
      "learning_rate": 4.900544049499911e-05,
      "loss": 0.7284,
      "step": 3510
    },
    {
      "epoch": 0.06957759285248365,
      "grad_norm": 2.913463592529297,
      "learning_rate": 4.900213521249661e-05,
      "loss": 0.9588,
      "step": 3520
    },
    {
      "epoch": 0.06977525646854184,
      "grad_norm": 2.086529493331909,
      "learning_rate": 4.8998829929994114e-05,
      "loss": 0.5935,
      "step": 3530
    },
    {
      "epoch": 0.06997292008460003,
      "grad_norm": 2.521984577178955,
      "learning_rate": 4.8995524647491624e-05,
      "loss": 0.8319,
      "step": 3540
    },
    {
      "epoch": 0.07017058370065822,
      "grad_norm": 3.0919301509857178,
      "learning_rate": 4.8992219364989126e-05,
      "loss": 0.7219,
      "step": 3550
    },
    {
      "epoch": 0.07036824731671641,
      "grad_norm": 3.319823741912842,
      "learning_rate": 4.898891408248663e-05,
      "loss": 1.0935,
      "step": 3560
    },
    {
      "epoch": 0.0705659109327746,
      "grad_norm": 2.7586803436279297,
      "learning_rate": 4.898560879998414e-05,
      "loss": 0.7112,
      "step": 3570
    },
    {
      "epoch": 0.0707635745488328,
      "grad_norm": 0.21173939108848572,
      "learning_rate": 4.898230351748164e-05,
      "loss": 1.1498,
      "step": 3580
    },
    {
      "epoch": 0.07096123816489099,
      "grad_norm": 2.5180115699768066,
      "learning_rate": 4.8978998234979145e-05,
      "loss": 0.79,
      "step": 3590
    },
    {
      "epoch": 0.07115890178094918,
      "grad_norm": 2.2828385829925537,
      "learning_rate": 4.897569295247665e-05,
      "loss": 1.0336,
      "step": 3600
    },
    {
      "epoch": 0.07135656539700737,
      "grad_norm": 2.8861117362976074,
      "learning_rate": 4.897238766997416e-05,
      "loss": 1.2233,
      "step": 3610
    },
    {
      "epoch": 0.07155422901306556,
      "grad_norm": 2.3170483112335205,
      "learning_rate": 4.896908238747166e-05,
      "loss": 1.0557,
      "step": 3620
    },
    {
      "epoch": 0.07175189262912375,
      "grad_norm": 2.515742778778076,
      "learning_rate": 4.896577710496916e-05,
      "loss": 0.8022,
      "step": 3630
    },
    {
      "epoch": 0.07194955624518196,
      "grad_norm": 2.676393508911133,
      "learning_rate": 4.896247182246667e-05,
      "loss": 0.7748,
      "step": 3640
    },
    {
      "epoch": 0.07214721986124015,
      "grad_norm": 0.17055971920490265,
      "learning_rate": 4.8959166539964175e-05,
      "loss": 0.8714,
      "step": 3650
    },
    {
      "epoch": 0.07234488347729834,
      "grad_norm": 2.648733139038086,
      "learning_rate": 4.895586125746168e-05,
      "loss": 1.0416,
      "step": 3660
    },
    {
      "epoch": 0.07254254709335653,
      "grad_norm": 2.829806089401245,
      "learning_rate": 4.895255597495918e-05,
      "loss": 0.6913,
      "step": 3670
    },
    {
      "epoch": 0.07274021070941472,
      "grad_norm": 2.972714900970459,
      "learning_rate": 4.894925069245669e-05,
      "loss": 1.4667,
      "step": 3680
    },
    {
      "epoch": 0.07293787432547291,
      "grad_norm": 2.85085129737854,
      "learning_rate": 4.8945945409954194e-05,
      "loss": 0.7345,
      "step": 3690
    },
    {
      "epoch": 0.0731355379415311,
      "grad_norm": 0.08262930810451508,
      "learning_rate": 4.8942640127451696e-05,
      "loss": 0.7351,
      "step": 3700
    },
    {
      "epoch": 0.0733332015575893,
      "grad_norm": 3.0437707901000977,
      "learning_rate": 4.8939334844949206e-05,
      "loss": 1.0478,
      "step": 3710
    },
    {
      "epoch": 0.07353086517364749,
      "grad_norm": 3.599978446960449,
      "learning_rate": 4.893602956244671e-05,
      "loss": 0.7786,
      "step": 3720
    },
    {
      "epoch": 0.07372852878970568,
      "grad_norm": 0.16308873891830444,
      "learning_rate": 4.893272427994421e-05,
      "loss": 0.7727,
      "step": 3730
    },
    {
      "epoch": 0.07392619240576387,
      "grad_norm": 2.418511390686035,
      "learning_rate": 4.8929418997441715e-05,
      "loss": 0.6492,
      "step": 3740
    },
    {
      "epoch": 0.07412385602182206,
      "grad_norm": 0.1118987649679184,
      "learning_rate": 4.892611371493922e-05,
      "loss": 1.0268,
      "step": 3750
    },
    {
      "epoch": 0.07432151963788025,
      "grad_norm": 2.941243886947632,
      "learning_rate": 4.892280843243672e-05,
      "loss": 1.2206,
      "step": 3760
    },
    {
      "epoch": 0.07451918325393844,
      "grad_norm": 0.2895379662513733,
      "learning_rate": 4.891950314993422e-05,
      "loss": 1.3224,
      "step": 3770
    },
    {
      "epoch": 0.07471684686999663,
      "grad_norm": 2.18247127532959,
      "learning_rate": 4.891619786743173e-05,
      "loss": 0.8141,
      "step": 3780
    },
    {
      "epoch": 0.07491451048605482,
      "grad_norm": 2.3649168014526367,
      "learning_rate": 4.8912892584929236e-05,
      "loss": 0.8939,
      "step": 3790
    },
    {
      "epoch": 0.07511217410211303,
      "grad_norm": 0.10176785290241241,
      "learning_rate": 4.890958730242674e-05,
      "loss": 0.6265,
      "step": 3800
    },
    {
      "epoch": 0.07530983771817122,
      "grad_norm": 2.643597364425659,
      "learning_rate": 4.890628201992424e-05,
      "loss": 0.7282,
      "step": 3810
    },
    {
      "epoch": 0.07550750133422941,
      "grad_norm": 0.14312265813350677,
      "learning_rate": 4.890297673742175e-05,
      "loss": 0.6308,
      "step": 3820
    },
    {
      "epoch": 0.0757051649502876,
      "grad_norm": 3.102372169494629,
      "learning_rate": 4.8899671454919254e-05,
      "loss": 0.9565,
      "step": 3830
    },
    {
      "epoch": 0.0759028285663458,
      "grad_norm": 7.350551128387451,
      "learning_rate": 4.889636617241676e-05,
      "loss": 1.1017,
      "step": 3840
    },
    {
      "epoch": 0.07610049218240399,
      "grad_norm": 2.9206531047821045,
      "learning_rate": 4.889306088991426e-05,
      "loss": 0.9196,
      "step": 3850
    },
    {
      "epoch": 0.07629815579846218,
      "grad_norm": 0.13130345940589905,
      "learning_rate": 4.888975560741177e-05,
      "loss": 0.6516,
      "step": 3860
    },
    {
      "epoch": 0.07649581941452037,
      "grad_norm": 2.7821011543273926,
      "learning_rate": 4.888645032490927e-05,
      "loss": 1.2361,
      "step": 3870
    },
    {
      "epoch": 0.07669348303057856,
      "grad_norm": 0.16820509731769562,
      "learning_rate": 4.8883145042406775e-05,
      "loss": 1.071,
      "step": 3880
    },
    {
      "epoch": 0.07689114664663675,
      "grad_norm": 2.774754285812378,
      "learning_rate": 4.8879839759904285e-05,
      "loss": 0.875,
      "step": 3890
    },
    {
      "epoch": 0.07708881026269494,
      "grad_norm": 0.07848513871431351,
      "learning_rate": 4.887653447740179e-05,
      "loss": 0.4514,
      "step": 3900
    },
    {
      "epoch": 0.07728647387875313,
      "grad_norm": 2.399386167526245,
      "learning_rate": 4.887322919489929e-05,
      "loss": 1.2902,
      "step": 3910
    },
    {
      "epoch": 0.07748413749481133,
      "grad_norm": 0.1369929015636444,
      "learning_rate": 4.886992391239679e-05,
      "loss": 0.3661,
      "step": 3920
    },
    {
      "epoch": 0.07768180111086952,
      "grad_norm": 2.827442169189453,
      "learning_rate": 4.88666186298943e-05,
      "loss": 1.1939,
      "step": 3930
    },
    {
      "epoch": 0.07787946472692771,
      "grad_norm": 0.1344004124403,
      "learning_rate": 4.8863313347391806e-05,
      "loss": 0.5525,
      "step": 3940
    },
    {
      "epoch": 0.07807712834298591,
      "grad_norm": 0.09482813626527786,
      "learning_rate": 4.886000806488931e-05,
      "loss": 1.0307,
      "step": 3950
    },
    {
      "epoch": 0.0782747919590441,
      "grad_norm": 2.756425142288208,
      "learning_rate": 4.885670278238682e-05,
      "loss": 0.3257,
      "step": 3960
    },
    {
      "epoch": 0.0784724555751023,
      "grad_norm": 2.794774293899536,
      "learning_rate": 4.885339749988432e-05,
      "loss": 0.9775,
      "step": 3970
    },
    {
      "epoch": 0.07867011919116049,
      "grad_norm": 2.7555766105651855,
      "learning_rate": 4.8850092217381824e-05,
      "loss": 1.0349,
      "step": 3980
    },
    {
      "epoch": 0.07886778280721868,
      "grad_norm": 2.6329236030578613,
      "learning_rate": 4.884678693487933e-05,
      "loss": 1.0094,
      "step": 3990
    },
    {
      "epoch": 0.07906544642327687,
      "grad_norm": 2.1600284576416016,
      "learning_rate": 4.884348165237683e-05,
      "loss": 1.5775,
      "step": 4000
    },
    {
      "epoch": 0.07906544642327687,
      "eval_loss": 0.8547587990760803,
      "eval_runtime": 236.8992,
      "eval_samples_per_second": 89.929,
      "eval_steps_per_second": 11.241,
      "step": 4000
    },
    {
      "epoch": 0.07926311003933506,
      "grad_norm": 3.087810516357422,
      "learning_rate": 4.884017636987433e-05,
      "loss": 1.2526,
      "step": 4010
    },
    {
      "epoch": 0.07946077365539325,
      "grad_norm": 0.15373891592025757,
      "learning_rate": 4.8836871087371835e-05,
      "loss": 0.5687,
      "step": 4020
    },
    {
      "epoch": 0.07965843727145144,
      "grad_norm": 2.833374500274658,
      "learning_rate": 4.8833565804869345e-05,
      "loss": 0.5532,
      "step": 4030
    },
    {
      "epoch": 0.07985610088750963,
      "grad_norm": 0.040279921144247055,
      "learning_rate": 4.883026052236685e-05,
      "loss": 0.9503,
      "step": 4040
    },
    {
      "epoch": 0.08005376450356783,
      "grad_norm": 3.7951087951660156,
      "learning_rate": 4.882695523986435e-05,
      "loss": 1.3893,
      "step": 4050
    },
    {
      "epoch": 0.08025142811962602,
      "grad_norm": 0.1750183254480362,
      "learning_rate": 4.8823649957361854e-05,
      "loss": 0.8484,
      "step": 4060
    },
    {
      "epoch": 0.08044909173568421,
      "grad_norm": 3.1721439361572266,
      "learning_rate": 4.882034467485936e-05,
      "loss": 0.7802,
      "step": 4070
    },
    {
      "epoch": 0.0806467553517424,
      "grad_norm": 1.9000381231307983,
      "learning_rate": 4.8817039392356866e-05,
      "loss": 1.2243,
      "step": 4080
    },
    {
      "epoch": 0.08084441896780059,
      "grad_norm": 2.135247230529785,
      "learning_rate": 4.881373410985437e-05,
      "loss": 1.1585,
      "step": 4090
    },
    {
      "epoch": 0.0810420825838588,
      "grad_norm": 0.19161462783813477,
      "learning_rate": 4.881042882735188e-05,
      "loss": 0.6117,
      "step": 4100
    },
    {
      "epoch": 0.08123974619991699,
      "grad_norm": 0.07644087076187134,
      "learning_rate": 4.880712354484938e-05,
      "loss": 0.392,
      "step": 4110
    },
    {
      "epoch": 0.08143740981597518,
      "grad_norm": 1.9246127605438232,
      "learning_rate": 4.8803818262346884e-05,
      "loss": 0.6202,
      "step": 4120
    },
    {
      "epoch": 0.08163507343203337,
      "grad_norm": 0.06487145274877548,
      "learning_rate": 4.880051297984439e-05,
      "loss": 0.5594,
      "step": 4130
    },
    {
      "epoch": 0.08183273704809156,
      "grad_norm": 0.1235739216208458,
      "learning_rate": 4.87972076973419e-05,
      "loss": 0.5401,
      "step": 4140
    },
    {
      "epoch": 0.08203040066414975,
      "grad_norm": 2.928098440170288,
      "learning_rate": 4.87939024148394e-05,
      "loss": 0.9297,
      "step": 4150
    },
    {
      "epoch": 0.08222806428020794,
      "grad_norm": 0.07378148287534714,
      "learning_rate": 4.87905971323369e-05,
      "loss": 0.9265,
      "step": 4160
    },
    {
      "epoch": 0.08242572789626613,
      "grad_norm": 2.527186870574951,
      "learning_rate": 4.878729184983441e-05,
      "loss": 0.9325,
      "step": 4170
    },
    {
      "epoch": 0.08262339151232433,
      "grad_norm": 2.3187239170074463,
      "learning_rate": 4.8783986567331915e-05,
      "loss": 0.8406,
      "step": 4180
    },
    {
      "epoch": 0.08282105512838252,
      "grad_norm": 3.236887216567993,
      "learning_rate": 4.878068128482942e-05,
      "loss": 1.2522,
      "step": 4190
    },
    {
      "epoch": 0.08301871874444071,
      "grad_norm": 2.5885400772094727,
      "learning_rate": 4.877737600232692e-05,
      "loss": 1.0799,
      "step": 4200
    },
    {
      "epoch": 0.0832163823604989,
      "grad_norm": 2.590271234512329,
      "learning_rate": 4.877407071982443e-05,
      "loss": 0.7392,
      "step": 4210
    },
    {
      "epoch": 0.08341404597655709,
      "grad_norm": 2.920560598373413,
      "learning_rate": 4.877076543732193e-05,
      "loss": 1.1529,
      "step": 4220
    },
    {
      "epoch": 0.08361170959261528,
      "grad_norm": 3.7466421127319336,
      "learning_rate": 4.8767460154819436e-05,
      "loss": 0.7338,
      "step": 4230
    },
    {
      "epoch": 0.08380937320867347,
      "grad_norm": 2.9110376834869385,
      "learning_rate": 4.876415487231694e-05,
      "loss": 1.0331,
      "step": 4240
    },
    {
      "epoch": 0.08400703682473167,
      "grad_norm": 0.15148070454597473,
      "learning_rate": 4.876084958981444e-05,
      "loss": 0.7851,
      "step": 4250
    },
    {
      "epoch": 0.08420470044078987,
      "grad_norm": 0.21726632118225098,
      "learning_rate": 4.8757544307311945e-05,
      "loss": 0.8461,
      "step": 4260
    },
    {
      "epoch": 0.08440236405684806,
      "grad_norm": 3.1847314834594727,
      "learning_rate": 4.875423902480945e-05,
      "loss": 0.7172,
      "step": 4270
    },
    {
      "epoch": 0.08460002767290625,
      "grad_norm": 1.725106954574585,
      "learning_rate": 4.875093374230696e-05,
      "loss": 1.2461,
      "step": 4280
    },
    {
      "epoch": 0.08479769128896444,
      "grad_norm": 2.6119651794433594,
      "learning_rate": 4.874762845980446e-05,
      "loss": 1.0353,
      "step": 4290
    },
    {
      "epoch": 0.08499535490502264,
      "grad_norm": 3.063328981399536,
      "learning_rate": 4.874432317730196e-05,
      "loss": 0.7984,
      "step": 4300
    },
    {
      "epoch": 0.08519301852108083,
      "grad_norm": 2.4408907890319824,
      "learning_rate": 4.874101789479947e-05,
      "loss": 0.7814,
      "step": 4310
    },
    {
      "epoch": 0.08539068213713902,
      "grad_norm": 2.6137940883636475,
      "learning_rate": 4.8737712612296975e-05,
      "loss": 0.9909,
      "step": 4320
    },
    {
      "epoch": 0.08558834575319721,
      "grad_norm": 2.7464072704315186,
      "learning_rate": 4.873440732979448e-05,
      "loss": 0.9918,
      "step": 4330
    },
    {
      "epoch": 0.0857860093692554,
      "grad_norm": 0.08046642690896988,
      "learning_rate": 4.873110204729198e-05,
      "loss": 0.5577,
      "step": 4340
    },
    {
      "epoch": 0.08598367298531359,
      "grad_norm": 2.888587474822998,
      "learning_rate": 4.872779676478949e-05,
      "loss": 0.6056,
      "step": 4350
    },
    {
      "epoch": 0.08618133660137178,
      "grad_norm": 0.04963793233036995,
      "learning_rate": 4.8724491482286994e-05,
      "loss": 0.8296,
      "step": 4360
    },
    {
      "epoch": 0.08637900021742997,
      "grad_norm": 0.06268492341041565,
      "learning_rate": 4.8721186199784496e-05,
      "loss": 0.7075,
      "step": 4370
    },
    {
      "epoch": 0.08657666383348817,
      "grad_norm": 0.09710423648357391,
      "learning_rate": 4.8717880917282006e-05,
      "loss": 1.0466,
      "step": 4380
    },
    {
      "epoch": 0.08677432744954636,
      "grad_norm": 2.6248815059661865,
      "learning_rate": 4.871457563477951e-05,
      "loss": 0.7315,
      "step": 4390
    },
    {
      "epoch": 0.08697199106560455,
      "grad_norm": 3.0253047943115234,
      "learning_rate": 4.871127035227701e-05,
      "loss": 1.1718,
      "step": 4400
    },
    {
      "epoch": 0.08716965468166275,
      "grad_norm": 0.09341023862361908,
      "learning_rate": 4.8707965069774515e-05,
      "loss": 0.4118,
      "step": 4410
    },
    {
      "epoch": 0.08736731829772094,
      "grad_norm": 2.8832263946533203,
      "learning_rate": 4.8704659787272024e-05,
      "loss": 0.5604,
      "step": 4420
    },
    {
      "epoch": 0.08756498191377914,
      "grad_norm": 2.774013042449951,
      "learning_rate": 4.870135450476953e-05,
      "loss": 0.7993,
      "step": 4430
    },
    {
      "epoch": 0.08776264552983733,
      "grad_norm": 3.224109649658203,
      "learning_rate": 4.869804922226703e-05,
      "loss": 1.058,
      "step": 4440
    },
    {
      "epoch": 0.08796030914589552,
      "grad_norm": 2.3504011631011963,
      "learning_rate": 4.869474393976454e-05,
      "loss": 0.9765,
      "step": 4450
    },
    {
      "epoch": 0.08815797276195371,
      "grad_norm": 2.297046422958374,
      "learning_rate": 4.869143865726204e-05,
      "loss": 0.4738,
      "step": 4460
    },
    {
      "epoch": 0.0883556363780119,
      "grad_norm": 0.08942723274230957,
      "learning_rate": 4.8688133374759545e-05,
      "loss": 0.7576,
      "step": 4470
    },
    {
      "epoch": 0.08855329999407009,
      "grad_norm": 0.05458178371191025,
      "learning_rate": 4.868482809225705e-05,
      "loss": 0.3693,
      "step": 4480
    },
    {
      "epoch": 0.08875096361012828,
      "grad_norm": 2.929097890853882,
      "learning_rate": 4.868152280975455e-05,
      "loss": 1.0083,
      "step": 4490
    },
    {
      "epoch": 0.08894862722618647,
      "grad_norm": 2.923035144805908,
      "learning_rate": 4.8678217527252054e-05,
      "loss": 1.0263,
      "step": 4500
    },
    {
      "epoch": 0.08894862722618647,
      "eval_loss": 0.8550295233726501,
      "eval_runtime": 237.752,
      "eval_samples_per_second": 89.606,
      "eval_steps_per_second": 11.201,
      "step": 4500
    }
  ],
  "logging_steps": 10,
  "max_steps": 151773,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.8817206976512e+16,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}