{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 2424,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00041254125412541255,
      "grad_norm": 25.173291317801098,
      "learning_rate": 4.1152263374485605e-08,
      "loss": 1.4315,
      "step": 1
    },
    {
      "epoch": 0.0020627062706270625,
      "grad_norm": 23.244172144374918,
      "learning_rate": 2.05761316872428e-07,
      "loss": 1.4238,
      "step": 5
    },
    {
      "epoch": 0.004125412541254125,
      "grad_norm": 15.407649804349377,
      "learning_rate": 4.11522633744856e-07,
      "loss": 1.4148,
      "step": 10
    },
    {
      "epoch": 0.006188118811881188,
      "grad_norm": 8.402022752616448,
      "learning_rate": 6.17283950617284e-07,
      "loss": 1.278,
      "step": 15
    },
    {
      "epoch": 0.00825082508250825,
      "grad_norm": 11.034069123055462,
      "learning_rate": 8.23045267489712e-07,
      "loss": 1.1772,
      "step": 20
    },
    {
      "epoch": 0.010313531353135313,
      "grad_norm": 4.332735330250593,
      "learning_rate": 1.02880658436214e-06,
      "loss": 1.0531,
      "step": 25
    },
    {
      "epoch": 0.012376237623762377,
      "grad_norm": 3.55215196003044,
      "learning_rate": 1.234567901234568e-06,
      "loss": 0.9881,
      "step": 30
    },
    {
      "epoch": 0.014438943894389438,
      "grad_norm": 3.3491262120808307,
      "learning_rate": 1.440329218106996e-06,
      "loss": 0.9604,
      "step": 35
    },
    {
      "epoch": 0.0165016501650165,
      "grad_norm": 3.022805529228879,
      "learning_rate": 1.646090534979424e-06,
      "loss": 0.9382,
      "step": 40
    },
    {
      "epoch": 0.018564356435643563,
      "grad_norm": 3.068991080972076,
      "learning_rate": 1.8518518518518519e-06,
      "loss": 0.9276,
      "step": 45
    },
    {
      "epoch": 0.020627062706270627,
      "grad_norm": 3.1689853717520595,
      "learning_rate": 2.05761316872428e-06,
      "loss": 0.9118,
      "step": 50
    },
    {
      "epoch": 0.02268976897689769,
      "grad_norm": 3.095012272596912,
      "learning_rate": 2.263374485596708e-06,
      "loss": 0.9069,
      "step": 55
    },
    {
      "epoch": 0.024752475247524754,
      "grad_norm": 3.1980712687602972,
      "learning_rate": 2.469135802469136e-06,
      "loss": 0.9114,
      "step": 60
    },
    {
      "epoch": 0.026815181518151814,
      "grad_norm": 3.0048714944711734,
      "learning_rate": 2.674897119341564e-06,
      "loss": 0.8915,
      "step": 65
    },
    {
      "epoch": 0.028877887788778877,
      "grad_norm": 3.2438404533032807,
      "learning_rate": 2.880658436213992e-06,
      "loss": 0.9011,
      "step": 70
    },
    {
      "epoch": 0.03094059405940594,
      "grad_norm": 3.0946505652248786,
      "learning_rate": 3.08641975308642e-06,
      "loss": 0.8911,
      "step": 75
    },
    {
      "epoch": 0.033003300330033,
      "grad_norm": 2.976460498240933,
      "learning_rate": 3.292181069958848e-06,
      "loss": 0.8905,
      "step": 80
    },
    {
      "epoch": 0.03506600660066007,
      "grad_norm": 2.964486117776111,
      "learning_rate": 3.4979423868312762e-06,
      "loss": 0.8711,
      "step": 85
    },
    {
      "epoch": 0.03712871287128713,
      "grad_norm": 3.1472293237072018,
      "learning_rate": 3.7037037037037037e-06,
      "loss": 0.8776,
      "step": 90
    },
    {
      "epoch": 0.039191419141914194,
      "grad_norm": 2.9491584240395547,
      "learning_rate": 3.909465020576132e-06,
      "loss": 0.8731,
      "step": 95
    },
    {
      "epoch": 0.041254125412541254,
      "grad_norm": 3.1222334377001575,
      "learning_rate": 4.11522633744856e-06,
      "loss": 0.8673,
      "step": 100
    },
    {
      "epoch": 0.043316831683168314,
      "grad_norm": 3.1224801634532744,
      "learning_rate": 4.3209876543209875e-06,
      "loss": 0.8568,
      "step": 105
    },
    {
      "epoch": 0.04537953795379538,
      "grad_norm": 3.1837554638007255,
      "learning_rate": 4.526748971193416e-06,
      "loss": 0.8641,
      "step": 110
    },
    {
      "epoch": 0.04744224422442244,
      "grad_norm": 3.1279379452438336,
      "learning_rate": 4.732510288065844e-06,
      "loss": 0.8654,
      "step": 115
    },
    {
      "epoch": 0.04950495049504951,
      "grad_norm": 3.165713359110299,
      "learning_rate": 4.938271604938272e-06,
      "loss": 0.8593,
      "step": 120
    },
    {
      "epoch": 0.05156765676567657,
      "grad_norm": 2.9608774490517282,
      "learning_rate": 5.1440329218107e-06,
      "loss": 0.8548,
      "step": 125
    },
    {
      "epoch": 0.05363036303630363,
      "grad_norm": 2.9604320712666485,
      "learning_rate": 5.349794238683128e-06,
      "loss": 0.8506,
      "step": 130
    },
    {
      "epoch": 0.055693069306930694,
      "grad_norm": 3.0433993058342765,
      "learning_rate": 5.555555555555557e-06,
      "loss": 0.8439,
      "step": 135
    },
    {
      "epoch": 0.057755775577557754,
      "grad_norm": 3.3372432183258036,
      "learning_rate": 5.761316872427984e-06,
      "loss": 0.8352,
      "step": 140
    },
    {
      "epoch": 0.05981848184818482,
      "grad_norm": 3.166742755838131,
      "learning_rate": 5.967078189300412e-06,
      "loss": 0.8584,
      "step": 145
    },
    {
      "epoch": 0.06188118811881188,
      "grad_norm": 3.2086952524615326,
      "learning_rate": 6.17283950617284e-06,
      "loss": 0.825,
      "step": 150
    },
    {
      "epoch": 0.06394389438943894,
      "grad_norm": 3.4577671140527557,
      "learning_rate": 6.3786008230452675e-06,
      "loss": 0.824,
      "step": 155
    },
    {
      "epoch": 0.066006600660066,
      "grad_norm": 3.3939747315335613,
      "learning_rate": 6.584362139917696e-06,
      "loss": 0.8412,
      "step": 160
    },
    {
      "epoch": 0.06806930693069307,
      "grad_norm": 3.1642303616218195,
      "learning_rate": 6.790123456790124e-06,
      "loss": 0.8372,
      "step": 165
    },
    {
      "epoch": 0.07013201320132013,
      "grad_norm": 3.9078331437297975,
      "learning_rate": 6.9958847736625525e-06,
      "loss": 0.8393,
      "step": 170
    },
    {
      "epoch": 0.0721947194719472,
      "grad_norm": 3.086012476362717,
      "learning_rate": 7.201646090534981e-06,
      "loss": 0.8549,
      "step": 175
    },
    {
      "epoch": 0.07425742574257425,
      "grad_norm": 3.163457682817171,
      "learning_rate": 7.4074074074074075e-06,
      "loss": 0.8319,
      "step": 180
    },
    {
      "epoch": 0.07632013201320131,
      "grad_norm": 3.1200124817434607,
      "learning_rate": 7.613168724279836e-06,
      "loss": 0.8229,
      "step": 185
    },
    {
      "epoch": 0.07838283828382839,
      "grad_norm": 3.010419511128134,
      "learning_rate": 7.818930041152263e-06,
      "loss": 0.8329,
      "step": 190
    },
    {
      "epoch": 0.08044554455445545,
      "grad_norm": 3.0362752294117317,
      "learning_rate": 8.024691358024692e-06,
      "loss": 0.8288,
      "step": 195
    },
    {
      "epoch": 0.08250825082508251,
      "grad_norm": 3.1939826585064997,
      "learning_rate": 8.23045267489712e-06,
      "loss": 0.8134,
      "step": 200
    },
    {
      "epoch": 0.08457095709570957,
      "grad_norm": 2.9811252237712056,
      "learning_rate": 8.43621399176955e-06,
      "loss": 0.8293,
      "step": 205
    },
    {
      "epoch": 0.08663366336633663,
      "grad_norm": 3.2289018691242264,
      "learning_rate": 8.641975308641975e-06,
      "loss": 0.8108,
      "step": 210
    },
    {
      "epoch": 0.0886963696369637,
      "grad_norm": 2.920715386036233,
      "learning_rate": 8.847736625514404e-06,
      "loss": 0.8199,
      "step": 215
    },
    {
      "epoch": 0.09075907590759076,
      "grad_norm": 2.8888271523586835,
      "learning_rate": 9.053497942386832e-06,
      "loss": 0.831,
      "step": 220
    },
    {
      "epoch": 0.09282178217821782,
      "grad_norm": 2.8414889865764312,
      "learning_rate": 9.25925925925926e-06,
      "loss": 0.8188,
      "step": 225
    },
    {
      "epoch": 0.09488448844884488,
      "grad_norm": 3.3784549124194743,
      "learning_rate": 9.465020576131688e-06,
      "loss": 0.8083,
      "step": 230
    },
    {
      "epoch": 0.09694719471947194,
      "grad_norm": 3.2323221908817996,
      "learning_rate": 9.670781893004116e-06,
      "loss": 0.8186,
      "step": 235
    },
    {
      "epoch": 0.09900990099009901,
      "grad_norm": 2.8554149805575846,
      "learning_rate": 9.876543209876543e-06,
      "loss": 0.8035,
      "step": 240
    },
    {
      "epoch": 0.10107260726072607,
      "grad_norm": 3.027573880753027,
      "learning_rate": 9.9999792514327e-06,
      "loss": 0.8091,
      "step": 245
    },
    {
      "epoch": 0.10313531353135313,
      "grad_norm": 3.0804491286944145,
      "learning_rate": 9.999745832028163e-06,
      "loss": 0.7989,
      "step": 250
    },
    {
      "epoch": 0.1051980198019802,
      "grad_norm": 3.2290033265195848,
      "learning_rate": 9.999253069658074e-06,
      "loss": 0.8178,
      "step": 255
    },
    {
      "epoch": 0.10726072607260725,
      "grad_norm": 3.4736769564268233,
      "learning_rate": 9.998500989882627e-06,
      "loss": 0.8178,
      "step": 260
    },
    {
      "epoch": 0.10932343234323433,
      "grad_norm": 3.0428405529671263,
      "learning_rate": 9.997489631713117e-06,
      "loss": 0.8006,
      "step": 265
    },
    {
      "epoch": 0.11138613861386139,
      "grad_norm": 2.8688354965301426,
      "learning_rate": 9.996219047609943e-06,
      "loss": 0.7866,
      "step": 270
    },
    {
      "epoch": 0.11344884488448845,
      "grad_norm": 3.0556606109594644,
      "learning_rate": 9.99468930347986e-06,
      "loss": 0.8174,
      "step": 275
    },
    {
      "epoch": 0.11551155115511551,
      "grad_norm": 3.0647994656665603,
      "learning_rate": 9.99290047867258e-06,
      "loss": 0.8044,
      "step": 280
    },
    {
      "epoch": 0.11757425742574257,
      "grad_norm": 2.856082139020889,
      "learning_rate": 9.990852665976648e-06,
      "loss": 0.7871,
      "step": 285
    },
    {
      "epoch": 0.11963696369636964,
      "grad_norm": 2.894869532549818,
      "learning_rate": 9.98854597161462e-06,
      "loss": 0.7967,
      "step": 290
    },
    {
      "epoch": 0.1216996699669967,
      "grad_norm": 2.9795368246449483,
      "learning_rate": 9.98598051523758e-06,
      "loss": 0.7987,
      "step": 295
    },
    {
      "epoch": 0.12376237623762376,
      "grad_norm": 3.16205487552237,
      "learning_rate": 9.983156429918895e-06,
      "loss": 0.7675,
      "step": 300
    },
    {
      "epoch": 0.12582508250825084,
      "grad_norm": 3.245576749246031,
      "learning_rate": 9.98007386214735e-06,
      "loss": 0.7915,
      "step": 305
    },
    {
      "epoch": 0.12788778877887788,
      "grad_norm": 3.025661649853723,
      "learning_rate": 9.976732971819526e-06,
      "loss": 0.7895,
      "step": 310
    },
    {
      "epoch": 0.12995049504950495,
      "grad_norm": 2.843144690280715,
      "learning_rate": 9.973133932231514e-06,
      "loss": 0.7782,
      "step": 315
    },
    {
      "epoch": 0.132013201320132,
      "grad_norm": 3.105956078671983,
      "learning_rate": 9.96927693006992e-06,
      "loss": 0.7754,
      "step": 320
    },
    {
      "epoch": 0.13407590759075907,
      "grad_norm": 2.7999853476012877,
      "learning_rate": 9.965162165402194e-06,
      "loss": 0.7652,
      "step": 325
    },
    {
      "epoch": 0.13613861386138615,
      "grad_norm": 2.8609256056169987,
      "learning_rate": 9.960789851666237e-06,
      "loss": 0.7786,
      "step": 330
    },
    {
      "epoch": 0.1382013201320132,
      "grad_norm": 3.1630409852297454,
      "learning_rate": 9.956160215659342e-06,
      "loss": 0.7828,
      "step": 335
    },
    {
      "epoch": 0.14026402640264027,
      "grad_norm": 2.795363964969132,
      "learning_rate": 9.951273497526423e-06,
      "loss": 0.7686,
      "step": 340
    },
    {
      "epoch": 0.14232673267326731,
      "grad_norm": 2.9053928333152723,
      "learning_rate": 9.94612995074756e-06,
      "loss": 0.7635,
      "step": 345
    },
    {
      "epoch": 0.1443894389438944,
      "grad_norm": 2.7319984337656478,
      "learning_rate": 9.94072984212485e-06,
      "loss": 0.7396,
      "step": 350
    },
    {
      "epoch": 0.14645214521452146,
      "grad_norm": 2.90922485756749,
      "learning_rate": 9.935073451768567e-06,
      "loss": 0.7675,
      "step": 355
    },
    {
      "epoch": 0.1485148514851485,
      "grad_norm": 2.7539644778310852,
      "learning_rate": 9.929161073082636e-06,
      "loss": 0.7559,
      "step": 360
    },
    {
      "epoch": 0.15057755775577558,
      "grad_norm": 2.8959145263782644,
      "learning_rate": 9.922993012749413e-06,
      "loss": 0.748,
      "step": 365
    },
    {
      "epoch": 0.15264026402640263,
      "grad_norm": 2.8562652818653516,
      "learning_rate": 9.916569590713775e-06,
      "loss": 0.7412,
      "step": 370
    },
    {
      "epoch": 0.1547029702970297,
      "grad_norm": 2.860687978591301,
      "learning_rate": 9.90989114016652e-06,
      "loss": 0.7648,
      "step": 375
    },
    {
      "epoch": 0.15676567656765678,
      "grad_norm": 2.7743547846726995,
      "learning_rate": 9.902958007527092e-06,
      "loss": 0.7469,
      "step": 380
    },
    {
      "epoch": 0.15882838283828382,
      "grad_norm": 2.933130436544695,
      "learning_rate": 9.89577055242561e-06,
      "loss": 0.7415,
      "step": 385
    },
    {
      "epoch": 0.1608910891089109,
      "grad_norm": 2.777976255262721,
      "learning_rate": 9.88832914768421e-06,
      "loss": 0.719,
      "step": 390
    },
    {
      "epoch": 0.16295379537953794,
      "grad_norm": 2.8659385000525983,
      "learning_rate": 9.880634179297706e-06,
      "loss": 0.742,
      "step": 395
    },
    {
      "epoch": 0.16501650165016502,
      "grad_norm": 2.8897535873983418,
      "learning_rate": 9.872686046413575e-06,
      "loss": 0.7346,
      "step": 400
    },
    {
      "epoch": 0.1670792079207921,
      "grad_norm": 2.8576246139096892,
      "learning_rate": 9.864485161311242e-06,
      "loss": 0.7325,
      "step": 405
    },
    {
      "epoch": 0.16914191419141913,
      "grad_norm": 3.1030879684908346,
      "learning_rate": 9.856031949380707e-06,
      "loss": 0.7606,
      "step": 410
    },
    {
      "epoch": 0.1712046204620462,
      "grad_norm": 2.713632760138165,
      "learning_rate": 9.847326849100467e-06,
      "loss": 0.7379,
      "step": 415
    },
    {
      "epoch": 0.17326732673267325,
      "grad_norm": 2.80346541082337,
      "learning_rate": 9.838370312014783e-06,
      "loss": 0.7381,
      "step": 420
    },
    {
      "epoch": 0.17533003300330033,
      "grad_norm": 3.200752732606346,
      "learning_rate": 9.829162802710246e-06,
      "loss": 0.7328,
      "step": 425
    },
    {
      "epoch": 0.1773927392739274,
      "grad_norm": 3.1792821336770567,
      "learning_rate": 9.819704798791691e-06,
      "loss": 0.7305,
      "step": 430
    },
    {
      "epoch": 0.17945544554455445,
      "grad_norm": 2.8356276160053424,
      "learning_rate": 9.80999679085741e-06,
      "loss": 0.714,
      "step": 435
    },
    {
      "epoch": 0.18151815181518152,
      "grad_norm": 2.6413041173544043,
      "learning_rate": 9.800039282473719e-06,
      "loss": 0.735,
      "step": 440
    },
    {
      "epoch": 0.18358085808580857,
      "grad_norm": 2.9983638727309856,
      "learning_rate": 9.789832790148822e-06,
      "loss": 0.7148,
      "step": 445
    },
    {
      "epoch": 0.18564356435643564,
      "grad_norm": 2.995170348181315,
      "learning_rate": 9.77937784330603e-06,
      "loss": 0.7137,
      "step": 450
    },
    {
      "epoch": 0.18770627062706272,
      "grad_norm": 3.0031426949210456,
      "learning_rate": 9.768674984256292e-06,
      "loss": 0.7334,
      "step": 455
    },
    {
      "epoch": 0.18976897689768976,
      "grad_norm": 2.789041969236007,
      "learning_rate": 9.757724768170074e-06,
      "loss": 0.7161,
      "step": 460
    },
    {
      "epoch": 0.19183168316831684,
      "grad_norm": 2.65991213841404,
      "learning_rate": 9.74652776304855e-06,
      "loss": 0.709,
      "step": 465
    },
    {
      "epoch": 0.19389438943894388,
      "grad_norm": 2.7707492384242007,
      "learning_rate": 9.73508454969415e-06,
      "loss": 0.7162,
      "step": 470
    },
    {
      "epoch": 0.19595709570957096,
      "grad_norm": 2.6804566008309654,
      "learning_rate": 9.723395721680418e-06,
      "loss": 0.7173,
      "step": 475
    },
    {
      "epoch": 0.19801980198019803,
      "grad_norm": 2.8279805454985545,
      "learning_rate": 9.711461885321247e-06,
      "loss": 0.6969,
      "step": 480
    },
    {
      "epoch": 0.20008250825082508,
      "grad_norm": 3.1208245898107045,
      "learning_rate": 9.699283659639402e-06,
      "loss": 0.7091,
      "step": 485
    },
    {
      "epoch": 0.20214521452145215,
      "grad_norm": 2.6589644032154363,
      "learning_rate": 9.68686167633443e-06,
      "loss": 0.6952,
      "step": 490
    },
    {
      "epoch": 0.2042079207920792,
      "grad_norm": 2.9084930227852137,
      "learning_rate": 9.67419657974988e-06,
      "loss": 0.7134,
      "step": 495
    },
    {
      "epoch": 0.20627062706270627,
      "grad_norm": 2.8303925657999853,
      "learning_rate": 9.661289026839889e-06,
      "loss": 0.6975,
      "step": 500
    },
    {
      "epoch": 0.20833333333333334,
      "grad_norm": 2.784037458254395,
      "learning_rate": 9.648139687135106e-06,
      "loss": 0.6979,
      "step": 505
    },
    {
      "epoch": 0.2103960396039604,
      "grad_norm": 2.7696913715189897,
      "learning_rate": 9.634749242707948e-06,
      "loss": 0.6733,
      "step": 510
    },
    {
      "epoch": 0.21245874587458746,
      "grad_norm": 2.9247340254888545,
      "learning_rate": 9.62111838813724e-06,
      "loss": 0.6945,
      "step": 515
    },
    {
      "epoch": 0.2145214521452145,
      "grad_norm": 2.719893437871808,
      "learning_rate": 9.607247830472174e-06,
      "loss": 0.6761,
      "step": 520
    },
    {
      "epoch": 0.21658415841584158,
      "grad_norm": 2.8726720120792386,
      "learning_rate": 9.593138289195634e-06,
      "loss": 0.6913,
      "step": 525
    },
    {
      "epoch": 0.21864686468646866,
      "grad_norm": 2.781250153342701,
      "learning_rate": 9.578790496186879e-06,
      "loss": 0.6903,
      "step": 530
    },
    {
      "epoch": 0.2207095709570957,
      "grad_norm": 2.7701813874151244,
      "learning_rate": 9.56420519568358e-06,
      "loss": 0.6831,
      "step": 535
    },
    {
      "epoch": 0.22277227722772278,
      "grad_norm": 2.7057701780207983,
      "learning_rate": 9.549383144243213e-06,
      "loss": 0.6956,
      "step": 540
    },
    {
      "epoch": 0.22483498349834982,
      "grad_norm": 2.7687957358970112,
      "learning_rate": 9.534325110703814e-06,
      "loss": 0.6709,
      "step": 545
    },
    {
      "epoch": 0.2268976897689769,
      "grad_norm": 2.7158766270699983,
      "learning_rate": 9.519031876144106e-06,
      "loss": 0.6608,
      "step": 550
    },
    {
      "epoch": 0.22896039603960397,
      "grad_norm": 2.7406561041756095,
      "learning_rate": 9.503504233842973e-06,
      "loss": 0.6857,
      "step": 555
    },
    {
      "epoch": 0.23102310231023102,
      "grad_norm": 2.7133954058150525,
      "learning_rate": 9.487742989238318e-06,
      "loss": 0.6723,
      "step": 560
    },
    {
      "epoch": 0.2330858085808581,
      "grad_norm": 2.7907877653122797,
      "learning_rate": 9.471748959885284e-06,
      "loss": 0.6809,
      "step": 565
    },
    {
      "epoch": 0.23514851485148514,
      "grad_norm": 2.8110823806166922,
      "learning_rate": 9.455522975413846e-06,
      "loss": 0.6695,
      "step": 570
    },
    {
      "epoch": 0.2372112211221122,
      "grad_norm": 2.9142519407957543,
      "learning_rate": 9.439065877485774e-06,
      "loss": 0.6506,
      "step": 575
    },
    {
      "epoch": 0.23927392739273928,
      "grad_norm": 2.7148544728300195,
      "learning_rate": 9.422378519750978e-06,
      "loss": 0.6499,
      "step": 580
    },
    {
      "epoch": 0.24133663366336633,
      "grad_norm": 2.713997914652474,
      "learning_rate": 9.40546176780323e-06,
      "loss": 0.6525,
      "step": 585
    },
    {
      "epoch": 0.2433993399339934,
      "grad_norm": 2.692751552453192,
      "learning_rate": 9.388316499135259e-06,
      "loss": 0.6496,
      "step": 590
    },
    {
      "epoch": 0.24546204620462045,
      "grad_norm": 2.6802330864864854,
      "learning_rate": 9.370943603093235e-06,
      "loss": 0.6417,
      "step": 595
    },
    {
      "epoch": 0.24752475247524752,
      "grad_norm": 2.8231163075907437,
      "learning_rate": 9.353343980830644e-06,
      "loss": 0.652,
      "step": 600
    },
    {
      "epoch": 0.2495874587458746,
      "grad_norm": 2.667983921285145,
      "learning_rate": 9.33551854526154e-06,
      "loss": 0.6516,
      "step": 605
    },
    {
      "epoch": 0.25165016501650167,
      "grad_norm": 2.7364796827272038,
      "learning_rate": 9.31746822101319e-06,
      "loss": 0.6402,
      "step": 610
    },
    {
      "epoch": 0.2537128712871287,
      "grad_norm": 2.6289768861588256,
      "learning_rate": 9.299193944378112e-06,
      "loss": 0.651,
      "step": 615
    },
    {
      "epoch": 0.25577557755775576,
      "grad_norm": 3.115748743305885,
      "learning_rate": 9.280696663265512e-06,
      "loss": 0.6409,
      "step": 620
    },
    {
      "epoch": 0.25783828382838286,
      "grad_norm": 2.7801532039806727,
      "learning_rate": 9.261977337152107e-06,
      "loss": 0.6232,
      "step": 625
    },
    {
      "epoch": 0.2599009900990099,
      "grad_norm": 2.8268570410512415,
      "learning_rate": 9.243036937032373e-06,
      "loss": 0.652,
      "step": 630
    },
    {
      "epoch": 0.26196369636963696,
      "grad_norm": 2.734252985059684,
      "learning_rate": 9.223876445368153e-06,
      "loss": 0.6408,
      "step": 635
    },
    {
      "epoch": 0.264026402640264,
      "grad_norm": 2.8594124607696862,
      "learning_rate": 9.204496856037718e-06,
      "loss": 0.6608,
      "step": 640
    },
    {
      "epoch": 0.2660891089108911,
      "grad_norm": 2.7632847146041937,
      "learning_rate": 9.184899174284201e-06,
      "loss": 0.6435,
      "step": 645
    },
    {
      "epoch": 0.26815181518151815,
      "grad_norm": 2.7892742600413056,
      "learning_rate": 9.16508441666346e-06,
      "loss": 0.6382,
      "step": 650
    },
    {
      "epoch": 0.2702145214521452,
      "grad_norm": 2.9224506916430926,
      "learning_rate": 9.14505361099134e-06,
      "loss": 0.6249,
      "step": 655
    },
    {
      "epoch": 0.2722772277227723,
      "grad_norm": 2.916964064023966,
      "learning_rate": 9.124807796290366e-06,
      "loss": 0.6284,
      "step": 660
    },
    {
      "epoch": 0.27433993399339934,
      "grad_norm": 2.7731836971623953,
      "learning_rate": 9.104348022735853e-06,
      "loss": 0.633,
      "step": 665
    },
    {
      "epoch": 0.2764026402640264,
      "grad_norm": 2.7535471272900196,
      "learning_rate": 9.083675351601417e-06,
      "loss": 0.6194,
      "step": 670
    },
    {
      "epoch": 0.2784653465346535,
      "grad_norm": 2.856389390689046,
      "learning_rate": 9.062790855203932e-06,
      "loss": 0.6376,
      "step": 675
    },
    {
      "epoch": 0.28052805280528054,
      "grad_norm": 2.9038867407258224,
      "learning_rate": 9.041695616847915e-06,
      "loss": 0.6263,
      "step": 680
    },
    {
      "epoch": 0.2825907590759076,
      "grad_norm": 2.5804402002125415,
      "learning_rate": 9.020390730769324e-06,
      "loss": 0.6393,
      "step": 685
    },
    {
      "epoch": 0.28465346534653463,
      "grad_norm": 2.632496010078948,
      "learning_rate": 8.998877302078803e-06,
      "loss": 0.6076,
      "step": 690
    },
    {
      "epoch": 0.28671617161716173,
      "grad_norm": 4.356850125546882,
      "learning_rate": 8.97715644670436e-06,
      "loss": 0.6133,
      "step": 695
    },
    {
      "epoch": 0.2887788778877888,
      "grad_norm": 2.559920441209793,
      "learning_rate": 8.955229291333473e-06,
      "loss": 0.6167,
      "step": 700
    },
    {
      "epoch": 0.2908415841584158,
      "grad_norm": 2.7757209022740823,
      "learning_rate": 8.933096973354665e-06,
      "loss": 0.6171,
      "step": 705
    },
    {
      "epoch": 0.2929042904290429,
      "grad_norm": 2.673673023680112,
      "learning_rate": 8.910760640798487e-06,
      "loss": 0.6142,
      "step": 710
    },
    {
      "epoch": 0.29496699669966997,
      "grad_norm": 2.8111377522291856,
      "learning_rate": 8.88822145227798e-06,
      "loss": 0.609,
      "step": 715
    },
    {
      "epoch": 0.297029702970297,
      "grad_norm": 2.6340814150599017,
      "learning_rate": 8.865480576928578e-06,
      "loss": 0.6205,
      "step": 720
    },
    {
      "epoch": 0.2990924092409241,
      "grad_norm": 2.6824532363700566,
      "learning_rate": 8.842539194347448e-06,
      "loss": 0.5866,
      "step": 725
    },
    {
      "epoch": 0.30115511551155116,
      "grad_norm": 2.619983825517638,
      "learning_rate": 8.819398494532328e-06,
      "loss": 0.6133,
      "step": 730
    },
    {
      "epoch": 0.3032178217821782,
      "grad_norm": 2.686298593112456,
      "learning_rate": 8.796059677819773e-06,
      "loss": 0.6011,
      "step": 735
    },
    {
      "epoch": 0.30528052805280526,
      "grad_norm": 2.9363501744882132,
      "learning_rate": 8.77252395482291e-06,
      "loss": 0.6084,
      "step": 740
    },
    {
      "epoch": 0.30734323432343236,
      "grad_norm": 2.752287473506256,
      "learning_rate": 8.748792546368641e-06,
      "loss": 0.5966,
      "step": 745
    },
    {
      "epoch": 0.3094059405940594,
      "grad_norm": 2.71844410518512,
      "learning_rate": 8.72486668343431e-06,
      "loss": 0.614,
      "step": 750
    },
    {
      "epoch": 0.31146864686468645,
      "grad_norm": 3.145817966285669,
      "learning_rate": 8.700747607083851e-06,
      "loss": 0.5957,
      "step": 755
    },
    {
      "epoch": 0.31353135313531355,
      "grad_norm": 3.0310831048054303,
      "learning_rate": 8.676436568403422e-06,
      "loss": 0.5978,
      "step": 760
    },
    {
      "epoch": 0.3155940594059406,
      "grad_norm": 2.701739599075158,
      "learning_rate": 8.651934828436497e-06,
      "loss": 0.5923,
      "step": 765
    },
    {
      "epoch": 0.31765676567656764,
      "grad_norm": 2.8448862694962944,
      "learning_rate": 8.627243658118466e-06,
      "loss": 0.5981,
      "step": 770
    },
    {
      "epoch": 0.31971947194719474,
      "grad_norm": 2.719726931310169,
      "learning_rate": 8.602364338210699e-06,
      "loss": 0.5937,
      "step": 775
    },
    {
      "epoch": 0.3217821782178218,
      "grad_norm": 2.8019550606602652,
      "learning_rate": 8.57729815923412e-06,
      "loss": 0.6029,
      "step": 780
    },
    {
      "epoch": 0.32384488448844884,
      "grad_norm": 2.503107027258152,
      "learning_rate": 8.55204642140226e-06,
      "loss": 0.5952,
      "step": 785
    },
    {
      "epoch": 0.3259075907590759,
      "grad_norm": 2.6677149194472762,
      "learning_rate": 8.52661043455382e-06,
      "loss": 0.5917,
      "step": 790
    },
    {
      "epoch": 0.327970297029703,
      "grad_norm": 2.746672046408305,
      "learning_rate": 8.50099151808472e-06,
      "loss": 0.5973,
      "step": 795
    },
    {
      "epoch": 0.33003300330033003,
      "grad_norm": 2.6263940638611554,
      "learning_rate": 8.47519100087967e-06,
      "loss": 0.5898,
      "step": 800
    },
    {
      "epoch": 0.3320957095709571,
      "grad_norm": 2.503813501518845,
      "learning_rate": 8.449210221243225e-06,
      "loss": 0.5991,
      "step": 805
    },
    {
      "epoch": 0.3341584158415842,
      "grad_norm": 2.4979356769923573,
      "learning_rate": 8.42305052683038e-06,
      "loss": 0.5912,
      "step": 810
    },
    {
      "epoch": 0.3362211221122112,
      "grad_norm": 2.847043832115963,
      "learning_rate": 8.39671327457666e-06,
      "loss": 0.5799,
      "step": 815
    },
    {
      "epoch": 0.33828382838283827,
      "grad_norm": 2.7801453739069633,
      "learning_rate": 8.370199830627732e-06,
      "loss": 0.5823,
      "step": 820
    },
    {
      "epoch": 0.34034653465346537,
      "grad_norm": 2.627346750326928,
      "learning_rate": 8.343511570268541e-06,
      "loss": 0.5865,
      "step": 825
    },
    {
      "epoch": 0.3424092409240924,
      "grad_norm": 2.8090361015173246,
      "learning_rate": 8.316649877851977e-06,
      "loss": 0.568,
      "step": 830
    },
    {
      "epoch": 0.34447194719471946,
      "grad_norm": 2.6857687788227107,
      "learning_rate": 8.289616146727062e-06,
      "loss": 0.5811,
      "step": 835
    },
    {
      "epoch": 0.3465346534653465,
      "grad_norm": 2.7827938400617964,
      "learning_rate": 8.262411779166681e-06,
      "loss": 0.5608,
      "step": 840
    },
    {
      "epoch": 0.3485973597359736,
      "grad_norm": 2.7509505960907084,
      "learning_rate": 8.235038186294836e-06,
      "loss": 0.5888,
      "step": 845
    },
    {
      "epoch": 0.35066006600660066,
      "grad_norm": 2.6638572156656584,
      "learning_rate": 8.207496788013456e-06,
      "loss": 0.562,
      "step": 850
    },
    {
      "epoch": 0.3527227722772277,
      "grad_norm": 2.591282044159344,
      "learning_rate": 8.179789012928747e-06,
      "loss": 0.5712,
      "step": 855
    },
    {
      "epoch": 0.3547854785478548,
      "grad_norm": 2.5109540424205576,
      "learning_rate": 8.151916298277078e-06,
      "loss": 0.553,
      "step": 860
    },
    {
      "epoch": 0.35684818481848185,
      "grad_norm": 2.892814594553888,
      "learning_rate": 8.123880089850438e-06,
      "loss": 0.5686,
      "step": 865
    },
    {
      "epoch": 0.3589108910891089,
      "grad_norm": 2.5359593127713898,
      "learning_rate": 8.095681841921441e-06,
      "loss": 0.5454,
      "step": 870
    },
    {
      "epoch": 0.360973597359736,
      "grad_norm": 2.6515995361569202,
      "learning_rate": 8.06732301716789e-06,
      "loss": 0.562,
      "step": 875
    },
    {
      "epoch": 0.36303630363036304,
      "grad_norm": 2.485706938051772,
      "learning_rate": 8.038805086596903e-06,
      "loss": 0.5532,
      "step": 880
    },
    {
      "epoch": 0.3650990099009901,
      "grad_norm": 2.5183163477684576,
      "learning_rate": 8.010129529468614e-06,
      "loss": 0.5666,
      "step": 885
    },
    {
      "epoch": 0.36716171617161714,
      "grad_norm": 2.675472695073968,
      "learning_rate": 7.981297833219435e-06,
      "loss": 0.5631,
      "step": 890
    },
    {
      "epoch": 0.36922442244224424,
      "grad_norm": 2.6653612064334045,
      "learning_rate": 7.952311493384916e-06,
      "loss": 0.5681,
      "step": 895
    },
    {
      "epoch": 0.3712871287128713,
      "grad_norm": 2.57514354803963,
      "learning_rate": 7.923172013522153e-06,
      "loss": 0.5668,
      "step": 900
    },
    {
      "epoch": 0.37334983498349833,
      "grad_norm": 2.707762540885758,
      "learning_rate": 7.893880905131807e-06,
      "loss": 0.558,
      "step": 905
    },
    {
      "epoch": 0.37541254125412543,
      "grad_norm": 2.6980529523902472,
      "learning_rate": 7.864439687579695e-06,
      "loss": 0.5447,
      "step": 910
    },
    {
      "epoch": 0.3774752475247525,
      "grad_norm": 2.7666172686862267,
      "learning_rate": 7.834849888017979e-06,
      "loss": 0.555,
      "step": 915
    },
    {
      "epoch": 0.3795379537953795,
      "grad_norm": 2.8852303996201876,
      "learning_rate": 7.805113041305958e-06,
      "loss": 0.5381,
      "step": 920
    },
    {
      "epoch": 0.3816006600660066,
      "grad_norm": 2.583967025080346,
      "learning_rate": 7.775230689930445e-06,
      "loss": 0.5465,
      "step": 925
    },
    {
      "epoch": 0.38366336633663367,
      "grad_norm": 2.612939793985183,
      "learning_rate": 7.745204383925753e-06,
      "loss": 0.5542,
      "step": 930
    },
    {
      "epoch": 0.3857260726072607,
      "grad_norm": 2.779912992420415,
      "learning_rate": 7.715035680793311e-06,
      "loss": 0.5527,
      "step": 935
    },
    {
      "epoch": 0.38778877887788776,
      "grad_norm": 2.644547318860223,
      "learning_rate": 7.684726145420853e-06,
      "loss": 0.5405,
      "step": 940
    },
    {
      "epoch": 0.38985148514851486,
      "grad_norm": 2.5007394214042433,
      "learning_rate": 7.654277350001255e-06,
      "loss": 0.5635,
      "step": 945
    },
    {
      "epoch": 0.3919141914191419,
      "grad_norm": 2.806021602965884,
      "learning_rate": 7.623690873950988e-06,
      "loss": 0.5273,
      "step": 950
    },
    {
      "epoch": 0.39397689768976896,
      "grad_norm": 2.5596611924464754,
      "learning_rate": 7.592968303828181e-06,
      "loss": 0.5275,
      "step": 955
    },
    {
      "epoch": 0.39603960396039606,
      "grad_norm": 2.6100310716117994,
      "learning_rate": 7.5621112332503325e-06,
      "loss": 0.5296,
      "step": 960
    },
    {
      "epoch": 0.3981023102310231,
      "grad_norm": 2.4652574522767017,
      "learning_rate": 7.531121262811645e-06,
      "loss": 0.5327,
      "step": 965
    },
    {
      "epoch": 0.40016501650165015,
      "grad_norm": 2.5835355525531054,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.5313,
      "step": 970
    },
    {
      "epoch": 0.40222772277227725,
      "grad_norm": 2.556870458540024,
      "learning_rate": 7.468749059113578e-06,
      "loss": 0.5219,
      "step": 975
    },
    {
      "epoch": 0.4042904290429043,
      "grad_norm": 2.6560183770094166,
      "learning_rate": 7.437370061177116e-06,
      "loss": 0.529,
      "step": 980
    },
    {
      "epoch": 0.40635313531353134,
      "grad_norm": 2.6201898522482683,
      "learning_rate": 7.40586463385783e-06,
      "loss": 0.534,
      "step": 985
    },
    {
      "epoch": 0.4084158415841584,
      "grad_norm": 2.6803677560952712,
      "learning_rate": 7.374234411380987e-06,
      "loss": 0.5249,
      "step": 990
    },
    {
      "epoch": 0.4104785478547855,
      "grad_norm": 2.5303749301843586,
      "learning_rate": 7.342481034445127e-06,
      "loss": 0.5318,
      "step": 995
    },
    {
      "epoch": 0.41254125412541254,
      "grad_norm": 2.5847848779090694,
      "learning_rate": 7.310606150136965e-06,
      "loss": 0.5137,
      "step": 1000
    },
    {
      "epoch": 0.4146039603960396,
      "grad_norm": 2.624837840218052,
      "learning_rate": 7.2786114118459564e-06,
      "loss": 0.5275,
      "step": 1005
    },
    {
      "epoch": 0.4166666666666667,
      "grad_norm": 2.7154629758532574,
      "learning_rate": 7.246498479178523e-06,
      "loss": 0.5311,
      "step": 1010
    },
    {
      "epoch": 0.41872937293729373,
      "grad_norm": 2.7217754848899234,
      "learning_rate": 7.214269017871981e-06,
      "loss": 0.5388,
      "step": 1015
    },
    {
      "epoch": 0.4207920792079208,
      "grad_norm": 2.8374602937448956,
      "learning_rate": 7.181924699708127e-06,
      "loss": 0.5114,
      "step": 1020
    },
    {
      "epoch": 0.4228547854785479,
      "grad_norm": 2.669653807310483,
      "learning_rate": 7.149467202426525e-06,
      "loss": 0.5162,
      "step": 1025
    },
    {
      "epoch": 0.4249174917491749,
      "grad_norm": 2.8259143043104853,
      "learning_rate": 7.116898209637478e-06,
      "loss": 0.5207,
      "step": 1030
    },
    {
      "epoch": 0.42698019801980197,
      "grad_norm": 2.4487631528416034,
      "learning_rate": 7.084219410734701e-06,
      "loss": 0.5239,
      "step": 1035
    },
    {
      "epoch": 0.429042904290429,
      "grad_norm": 2.5467769274691157,
      "learning_rate": 7.051432500807682e-06,
      "loss": 0.5143,
      "step": 1040
    },
    {
      "epoch": 0.4311056105610561,
      "grad_norm": 2.674655727993759,
      "learning_rate": 7.018539180553768e-06,
      "loss": 0.5179,
      "step": 1045
    },
    {
      "epoch": 0.43316831683168316,
      "grad_norm": 2.5887768807891107,
      "learning_rate": 6.985541156189932e-06,
      "loss": 0.5084,
      "step": 1050
    },
    {
      "epoch": 0.4352310231023102,
      "grad_norm": 2.7440060414213394,
      "learning_rate": 6.952440139364286e-06,
      "loss": 0.512,
      "step": 1055
    },
    {
      "epoch": 0.4372937293729373,
      "grad_norm": 2.645722871973474,
      "learning_rate": 6.919237847067282e-06,
      "loss": 0.5199,
      "step": 1060
    },
    {
      "epoch": 0.43935643564356436,
      "grad_norm": 2.611780386705664,
      "learning_rate": 6.885936001542658e-06,
      "loss": 0.5078,
      "step": 1065
    },
    {
      "epoch": 0.4414191419141914,
      "grad_norm": 2.5086528986935703,
      "learning_rate": 6.852536330198099e-06,
      "loss": 0.5201,
      "step": 1070
    },
    {
      "epoch": 0.4434818481848185,
      "grad_norm": 2.4576048338159486,
      "learning_rate": 6.819040565515636e-06,
      "loss": 0.5126,
      "step": 1075
    },
    {
      "epoch": 0.44554455445544555,
      "grad_norm": 2.7304284435440542,
      "learning_rate": 6.785450444961783e-06,
      "loss": 0.5145,
      "step": 1080
    },
    {
      "epoch": 0.4476072607260726,
      "grad_norm": 2.6166324259831435,
      "learning_rate": 6.751767710897404e-06,
      "loss": 0.5141,
      "step": 1085
    },
    {
      "epoch": 0.44966996699669964,
      "grad_norm": 2.528382351784196,
      "learning_rate": 6.71799411048734e-06,
      "loss": 0.4858,
      "step": 1090
    },
    {
      "epoch": 0.45173267326732675,
      "grad_norm": 2.633514406168111,
      "learning_rate": 6.684131395609784e-06,
      "loss": 0.5084,
      "step": 1095
    },
    {
      "epoch": 0.4537953795379538,
      "grad_norm": 2.534482871115365,
      "learning_rate": 6.650181322765407e-06,
      "loss": 0.4981,
      "step": 1100
    },
    {
      "epoch": 0.45585808580858084,
      "grad_norm": 2.4508414675719505,
      "learning_rate": 6.61614565298624e-06,
      "loss": 0.5008,
      "step": 1105
    },
    {
      "epoch": 0.45792079207920794,
      "grad_norm": 2.4274494034766203,
      "learning_rate": 6.5820261517443365e-06,
      "loss": 0.4981,
      "step": 1110
    },
    {
      "epoch": 0.459983498349835,
      "grad_norm": 2.538858548171987,
      "learning_rate": 6.54782458886019e-06,
      "loss": 0.4891,
      "step": 1115
    },
    {
      "epoch": 0.46204620462046203,
      "grad_norm": 2.5483259090729575,
      "learning_rate": 6.5135427384109315e-06,
      "loss": 0.4999,
      "step": 1120
    },
    {
      "epoch": 0.46410891089108913,
      "grad_norm": 2.5232721697679166,
      "learning_rate": 6.479182378638308e-06,
      "loss": 0.5024,
      "step": 1125
    },
    {
      "epoch": 0.4661716171617162,
      "grad_norm": 2.591445497366531,
      "learning_rate": 6.444745291856442e-06,
      "loss": 0.4964,
      "step": 1130
    },
    {
      "epoch": 0.4682343234323432,
      "grad_norm": 2.568897490439017,
      "learning_rate": 6.410233264359379e-06,
      "loss": 0.4978,
      "step": 1135
    },
    {
      "epoch": 0.47029702970297027,
      "grad_norm": 2.494168999146032,
      "learning_rate": 6.375648086328431e-06,
      "loss": 0.5058,
      "step": 1140
    },
    {
      "epoch": 0.47235973597359737,
      "grad_norm": 2.4409245464037967,
      "learning_rate": 6.340991551739319e-06,
      "loss": 0.4885,
      "step": 1145
    },
    {
      "epoch": 0.4744224422442244,
      "grad_norm": 2.603683438326256,
      "learning_rate": 6.3062654582691175e-06,
      "loss": 0.4983,
      "step": 1150
    },
    {
      "epoch": 0.47648514851485146,
      "grad_norm": 2.5017059333187404,
      "learning_rate": 6.271471607203006e-06,
      "loss": 0.4868,
      "step": 1155
    },
    {
      "epoch": 0.47854785478547857,
      "grad_norm": 2.664734878049621,
      "learning_rate": 6.236611803340829e-06,
      "loss": 0.4946,
      "step": 1160
    },
    {
      "epoch": 0.4806105610561056,
      "grad_norm": 2.6730486310091037,
      "learning_rate": 6.201687854903492e-06,
      "loss": 0.4811,
      "step": 1165
    },
    {
      "epoch": 0.48267326732673266,
      "grad_norm": 2.8065530727044927,
      "learning_rate": 6.16670157343915e-06,
      "loss": 0.4974,
      "step": 1170
    },
    {
      "epoch": 0.48473597359735976,
      "grad_norm": 2.660140416974183,
      "learning_rate": 6.131654773729255e-06,
      "loss": 0.485,
      "step": 1175
    },
    {
      "epoch": 0.4867986798679868,
      "grad_norm": 2.494718317338255,
      "learning_rate": 6.096549273694411e-06,
      "loss": 0.4832,
      "step": 1180
    },
    {
      "epoch": 0.48886138613861385,
      "grad_norm": 2.6067794556409427,
      "learning_rate": 6.061386894300082e-06,
      "loss": 0.4822,
      "step": 1185
    },
    {
      "epoch": 0.4909240924092409,
      "grad_norm": 2.411365711902098,
      "learning_rate": 6.026169459462132e-06,
      "loss": 0.488,
      "step": 1190
    },
    {
      "epoch": 0.492986798679868,
      "grad_norm": 2.5304361708398253,
      "learning_rate": 5.990898795952225e-06,
      "loss": 0.4808,
      "step": 1195
    },
    {
      "epoch": 0.49504950495049505,
      "grad_norm": 2.542256557791119,
      "learning_rate": 5.955576733303053e-06,
      "loss": 0.4922,
      "step": 1200
    },
    {
      "epoch": 0.4971122112211221,
      "grad_norm": 2.416811649448376,
      "learning_rate": 5.920205103713449e-06,
      "loss": 0.4857,
      "step": 1205
    },
    {
      "epoch": 0.4991749174917492,
      "grad_norm": 2.4894521032789023,
      "learning_rate": 5.884785741953345e-06,
      "loss": 0.4742,
      "step": 1210
    },
    {
      "epoch": 0.5012376237623762,
      "grad_norm": 2.4053368303294995,
      "learning_rate": 5.849320485268597e-06,
      "loss": 0.4691,
      "step": 1215
    },
    {
      "epoch": 0.5033003300330033,
      "grad_norm": 2.440300312110612,
      "learning_rate": 5.8138111732856906e-06,
      "loss": 0.4794,
      "step": 1220
    },
    {
      "epoch": 0.5053630363036303,
      "grad_norm": 2.507917784275635,
      "learning_rate": 5.778259647916309e-06,
      "loss": 0.4807,
      "step": 1225
    },
    {
      "epoch": 0.5074257425742574,
      "grad_norm": 2.564227750985733,
      "learning_rate": 5.7426677532618e-06,
      "loss": 0.4623,
      "step": 1230
    },
    {
      "epoch": 0.5094884488448845,
      "grad_norm": 2.5559157486419037,
      "learning_rate": 5.707037335517514e-06,
      "loss": 0.4599,
      "step": 1235
    },
    {
      "epoch": 0.5115511551155115,
      "grad_norm": 2.4886710886810666,
      "learning_rate": 5.67137024287704e-06,
      "loss": 0.4539,
      "step": 1240
    },
    {
      "epoch": 0.5136138613861386,
      "grad_norm": 2.533851130168081,
      "learning_rate": 5.635668325436343e-06,
      "loss": 0.4762,
      "step": 1245
    },
    {
      "epoch": 0.5156765676567657,
      "grad_norm": 2.5682484148678593,
      "learning_rate": 5.599933435097791e-06,
      "loss": 0.461,
      "step": 1250
    },
    {
      "epoch": 0.5177392739273927,
      "grad_norm": 2.7619029888240334,
      "learning_rate": 5.564167425474093e-06,
      "loss": 0.4708,
      "step": 1255
    },
    {
      "epoch": 0.5198019801980198,
      "grad_norm": 2.65748720318593,
      "learning_rate": 5.528372151792161e-06,
      "loss": 0.4835,
      "step": 1260
    },
    {
      "epoch": 0.5218646864686468,
      "grad_norm": 2.343200657687914,
      "learning_rate": 5.492549470796865e-06,
      "loss": 0.4562,
      "step": 1265
    },
    {
      "epoch": 0.5239273927392739,
      "grad_norm": 2.398515835994456,
      "learning_rate": 5.456701240654726e-06,
      "loss": 0.4448,
      "step": 1270
    },
    {
      "epoch": 0.525990099009901,
      "grad_norm": 2.480816874843841,
      "learning_rate": 5.420829320857532e-06,
      "loss": 0.4599,
      "step": 1275
    },
    {
      "epoch": 0.528052805280528,
      "grad_norm": 2.5823441471809643,
      "learning_rate": 5.384935572125882e-06,
      "loss": 0.4352,
      "step": 1280
    },
    {
      "epoch": 0.5301155115511551,
      "grad_norm": 2.4908465249480853,
      "learning_rate": 5.349021856312669e-06,
      "loss": 0.4594,
      "step": 1285
    },
    {
      "epoch": 0.5321782178217822,
      "grad_norm": 2.399600740783908,
      "learning_rate": 5.3130900363065055e-06,
      "loss": 0.4466,
      "step": 1290
    },
    {
      "epoch": 0.5342409240924092,
      "grad_norm": 2.454620597807439,
      "learning_rate": 5.277141975935083e-06,
      "loss": 0.4585,
      "step": 1295
    },
    {
      "epoch": 0.5363036303630363,
      "grad_norm": 2.372249644843968,
      "learning_rate": 5.24117953986851e-06,
      "loss": 0.4512,
      "step": 1300
    },
    {
      "epoch": 0.5383663366336634,
      "grad_norm": 2.3367801961389407,
      "learning_rate": 5.2052045935225725e-06,
      "loss": 0.4568,
      "step": 1305
    },
    {
      "epoch": 0.5404290429042904,
      "grad_norm": 2.4464214105982856,
      "learning_rate": 5.169219002961987e-06,
      "loss": 0.4561,
      "step": 1310
    },
    {
      "epoch": 0.5424917491749175,
      "grad_norm": 10.09033249652104,
      "learning_rate": 5.133224634803594e-06,
      "loss": 0.4506,
      "step": 1315
    },
    {
      "epoch": 0.5445544554455446,
      "grad_norm": 2.525626316753546,
      "learning_rate": 5.097223356119538e-06,
      "loss": 0.4651,
      "step": 1320
    },
    {
      "epoch": 0.5466171617161716,
      "grad_norm": 2.521794410914058,
      "learning_rate": 5.061217034340426e-06,
      "loss": 0.4729,
      "step": 1325
    },
    {
      "epoch": 0.5486798679867987,
      "grad_norm": 2.5260338421626445,
      "learning_rate": 5.02520753715845e-06,
      "loss": 0.4525,
      "step": 1330
    },
    {
      "epoch": 0.5507425742574258,
      "grad_norm": 2.446840796366364,
      "learning_rate": 4.989196732430518e-06,
      "loss": 0.4567,
      "step": 1335
    },
    {
      "epoch": 0.5528052805280528,
      "grad_norm": 2.3859413099257893,
      "learning_rate": 4.953186488081362e-06,
      "loss": 0.4471,
      "step": 1340
    },
    {
      "epoch": 0.5548679867986799,
      "grad_norm": 2.5520983341553105,
      "learning_rate": 4.9171786720066465e-06,
      "loss": 0.4523,
      "step": 1345
    },
    {
      "epoch": 0.556930693069307,
      "grad_norm": 2.470761509016717,
      "learning_rate": 4.881175151976075e-06,
      "loss": 0.4589,
      "step": 1350
    },
    {
      "epoch": 0.558993399339934,
      "grad_norm": 2.6345528475282847,
      "learning_rate": 4.845177795536516e-06,
      "loss": 0.4429,
      "step": 1355
    },
    {
      "epoch": 0.5610561056105611,
      "grad_norm": 2.4651932232525366,
      "learning_rate": 4.809188469915121e-06,
      "loss": 0.4341,
      "step": 1360
    },
    {
      "epoch": 0.5631188118811881,
      "grad_norm": 2.423072698536209,
      "learning_rate": 4.773209041922472e-06,
      "loss": 0.4416,
      "step": 1365
    },
    {
      "epoch": 0.5651815181518152,
      "grad_norm": 2.453471069490191,
      "learning_rate": 4.737241377855751e-06,
      "loss": 0.4307,
      "step": 1370
    },
    {
      "epoch": 0.5672442244224423,
      "grad_norm": 2.373055640370215,
      "learning_rate": 4.7012873434019296e-06,
      "loss": 0.4355,
      "step": 1375
    },
    {
      "epoch": 0.5693069306930693,
      "grad_norm": 2.5437533790570646,
      "learning_rate": 4.6653488035409975e-06,
      "loss": 0.4415,
      "step": 1380
    },
    {
      "epoch": 0.5713696369636964,
      "grad_norm": 2.391498166308705,
      "learning_rate": 4.629427622449217e-06,
      "loss": 0.4313,
      "step": 1385
    },
    {
      "epoch": 0.5734323432343235,
      "grad_norm": 2.643038273859464,
      "learning_rate": 4.59352566340243e-06,
      "loss": 0.4531,
      "step": 1390
    },
    {
      "epoch": 0.5754950495049505,
      "grad_norm": 2.515338122652954,
      "learning_rate": 4.557644788679413e-06,
      "loss": 0.4372,
      "step": 1395
    },
    {
      "epoch": 0.5775577557755776,
      "grad_norm": 2.5228479375899355,
      "learning_rate": 4.521786859465263e-06,
      "loss": 0.4343,
      "step": 1400
    },
    {
      "epoch": 0.5796204620462047,
      "grad_norm": 2.526725915460239,
      "learning_rate": 4.485953735754872e-06,
      "loss": 0.4236,
      "step": 1405
    },
    {
      "epoch": 0.5816831683168316,
      "grad_norm": 2.3164427894674224,
      "learning_rate": 4.450147276256439e-06,
      "loss": 0.4164,
      "step": 1410
    },
    {
      "epoch": 0.5837458745874587,
      "grad_norm": 2.3425181209777963,
      "learning_rate": 4.414369338295056e-06,
      "loss": 0.4427,
      "step": 1415
    },
    {
      "epoch": 0.5858085808580858,
      "grad_norm": 2.450541853696667,
      "learning_rate": 4.37862177771637e-06,
      "loss": 0.4406,
      "step": 1420
    },
    {
      "epoch": 0.5878712871287128,
      "grad_norm": 2.420195141307401,
      "learning_rate": 4.342906448790315e-06,
      "loss": 0.4475,
      "step": 1425
    },
    {
      "epoch": 0.5899339933993399,
      "grad_norm": 2.4136105308611624,
      "learning_rate": 4.307225204114927e-06,
      "loss": 0.435,
      "step": 1430
    },
    {
      "epoch": 0.591996699669967,
      "grad_norm": 2.4087084238655962,
      "learning_rate": 4.271579894520254e-06,
      "loss": 0.44,
      "step": 1435
    },
    {
      "epoch": 0.594059405940594,
      "grad_norm": 2.457797382505865,
      "learning_rate": 4.235972368972343e-06,
      "loss": 0.4368,
      "step": 1440
    },
    {
      "epoch": 0.5961221122112211,
      "grad_norm": 2.3957561493714827,
      "learning_rate": 4.200404474477341e-06,
      "loss": 0.4163,
      "step": 1445
    },
    {
      "epoch": 0.5981848184818482,
      "grad_norm": 2.3457771252145587,
      "learning_rate": 4.16487805598568e-06,
      "loss": 0.4094,
      "step": 1450
    },
    {
      "epoch": 0.6002475247524752,
      "grad_norm": 2.573777931737885,
      "learning_rate": 4.12939495629638e-06,
      "loss": 0.4162,
      "step": 1455
    },
    {
      "epoch": 0.6023102310231023,
      "grad_norm": 2.348703417322529,
      "learning_rate": 4.093957015961465e-06,
      "loss": 0.4208,
      "step": 1460
    },
    {
      "epoch": 0.6043729372937293,
      "grad_norm": 2.6136735137117997,
      "learning_rate": 4.0585660731904855e-06,
      "loss": 0.4357,
      "step": 1465
    },
    {
      "epoch": 0.6064356435643564,
      "grad_norm": 2.3717596586411633,
      "learning_rate": 4.023223963755168e-06,
      "loss": 0.4111,
      "step": 1470
    },
    {
      "epoch": 0.6084983498349835,
      "grad_norm": 2.3925003782413117,
      "learning_rate": 3.987932520894201e-06,
      "loss": 0.422,
      "step": 1475
    },
    {
      "epoch": 0.6105610561056105,
      "grad_norm": 2.4066751622931086,
      "learning_rate": 3.9526935752181275e-06,
      "loss": 0.4262,
      "step": 1480
    },
    {
      "epoch": 0.6126237623762376,
      "grad_norm": 2.2489948265452435,
      "learning_rate": 3.917508954614401e-06,
      "loss": 0.4137,
      "step": 1485
    },
    {
      "epoch": 0.6146864686468647,
      "grad_norm": 2.3814132782379636,
      "learning_rate": 3.882380484152567e-06,
      "loss": 0.4114,
      "step": 1490
    },
    {
      "epoch": 0.6167491749174917,
      "grad_norm": 2.503060690023786,
      "learning_rate": 3.847309985989593e-06,
      "loss": 0.4222,
      "step": 1495
    },
    {
      "epoch": 0.6188118811881188,
      "grad_norm": 2.485844712788725,
      "learning_rate": 3.8122992792753534e-06,
      "loss": 0.4212,
      "step": 1500
    },
    {
      "epoch": 0.6208745874587459,
      "grad_norm": 2.428549383590827,
      "learning_rate": 3.777350180058264e-06,
      "loss": 0.4187,
      "step": 1505
    },
    {
      "epoch": 0.6229372937293729,
      "grad_norm": 2.6875741250581533,
      "learning_rate": 3.7424645011910847e-06,
      "loss": 0.4123,
      "step": 1510
    },
    {
      "epoch": 0.625,
      "grad_norm": 2.425981196955932,
      "learning_rate": 3.707644052236887e-06,
      "loss": 0.4096,
      "step": 1515
    },
    {
      "epoch": 0.6270627062706271,
      "grad_norm": 2.7328115668919155,
      "learning_rate": 3.672890639375184e-06,
      "loss": 0.4206,
      "step": 1520
    },
    {
      "epoch": 0.6291254125412541,
      "grad_norm": 2.3245488388421593,
      "learning_rate": 3.6382060653082434e-06,
      "loss": 0.4052,
      "step": 1525
    },
    {
      "epoch": 0.6311881188118812,
      "grad_norm": 2.384682866362113,
      "learning_rate": 3.6035921291675815e-06,
      "loss": 0.4156,
      "step": 1530
    },
    {
      "epoch": 0.6332508250825083,
      "grad_norm": 2.4129965482745486,
      "learning_rate": 3.569050626420636e-06,
      "loss": 0.4093,
      "step": 1535
    },
    {
      "epoch": 0.6353135313531353,
      "grad_norm": 2.4547665669069776,
      "learning_rate": 3.5345833487776404e-06,
      "loss": 0.425,
      "step": 1540
    },
    {
      "epoch": 0.6373762376237624,
      "grad_norm": 2.489381449919041,
      "learning_rate": 3.500192084098677e-06,
      "loss": 0.4215,
      "step": 1545
    },
    {
      "epoch": 0.6394389438943895,
      "grad_norm": 2.382148749916793,
      "learning_rate": 3.4658786163009416e-06,
      "loss": 0.4152,
      "step": 1550
    },
    {
      "epoch": 0.6415016501650165,
      "grad_norm": 2.3828897057876746,
      "learning_rate": 3.4316447252662142e-06,
      "loss": 0.4084,
      "step": 1555
    },
    {
      "epoch": 0.6435643564356436,
      "grad_norm": 2.446574644486715,
      "learning_rate": 3.3974921867485238e-06,
      "loss": 0.407,
      "step": 1560
    },
    {
      "epoch": 0.6456270627062707,
      "grad_norm": 2.3536881611551035,
      "learning_rate": 3.3634227722820496e-06,
      "loss": 0.4007,
      "step": 1565
    },
    {
      "epoch": 0.6476897689768977,
      "grad_norm": 2.48572252066864,
      "learning_rate": 3.3294382490892226e-06,
      "loss": 0.401,
      "step": 1570
    },
    {
      "epoch": 0.6497524752475248,
      "grad_norm": 2.4835672414143604,
      "learning_rate": 3.2955403799890567e-06,
      "loss": 0.4005,
      "step": 1575
    },
    {
      "epoch": 0.6518151815181518,
      "grad_norm": 2.3799907386582224,
      "learning_rate": 3.261730923305717e-06,
      "loss": 0.4026,
      "step": 1580
    },
    {
      "epoch": 0.6538778877887789,
      "grad_norm": 2.486466648614617,
      "learning_rate": 3.2280116327773028e-06,
      "loss": 0.4048,
      "step": 1585
    },
    {
      "epoch": 0.655940594059406,
      "grad_norm": 2.4042700940181523,
      "learning_rate": 3.194384257464884e-06,
      "loss": 0.3945,
      "step": 1590
    },
    {
      "epoch": 0.658003300330033,
      "grad_norm": 2.327134482890906,
      "learning_rate": 3.160850541661779e-06,
      "loss": 0.4047,
      "step": 1595
    },
    {
      "epoch": 0.6600660066006601,
      "grad_norm": 2.4535001026550844,
      "learning_rate": 3.127412224803068e-06,
      "loss": 0.3885,
      "step": 1600
    },
    {
      "epoch": 0.6621287128712872,
      "grad_norm": 2.6395091605070786,
      "learning_rate": 3.094071041375375e-06,
      "loss": 0.4085,
      "step": 1605
    },
    {
      "epoch": 0.6641914191419142,
      "grad_norm": 2.45831697505499,
      "learning_rate": 3.060828720826889e-06,
      "loss": 0.4063,
      "step": 1610
    },
    {
      "epoch": 0.6662541254125413,
      "grad_norm": 2.25651867695093,
      "learning_rate": 3.0276869874776632e-06,
      "loss": 0.3991,
      "step": 1615
    },
    {
      "epoch": 0.6683168316831684,
      "grad_norm": 2.36719278095781,
      "learning_rate": 2.994647560430167e-06,
      "loss": 0.3943,
      "step": 1620
    },
    {
      "epoch": 0.6703795379537953,
      "grad_norm": 2.4049216832251035,
      "learning_rate": 2.961712153480118e-06,
      "loss": 0.4029,
      "step": 1625
    },
    {
      "epoch": 0.6724422442244224,
      "grad_norm": 2.559007669700738,
      "learning_rate": 2.9288824750275803e-06,
      "loss": 0.4032,
      "step": 1630
    },
    {
      "epoch": 0.6745049504950495,
      "grad_norm": 2.310036214506543,
      "learning_rate": 2.896160227988357e-06,
      "loss": 0.4005,
      "step": 1635
    },
    {
      "epoch": 0.6765676567656765,
      "grad_norm": 2.43914641154883,
      "learning_rate": 2.8635471097056423e-06,
      "loss": 0.3943,
      "step": 1640
    },
    {
      "epoch": 0.6786303630363036,
      "grad_norm": 2.443819723387033,
      "learning_rate": 2.8310448118619967e-06,
      "loss": 0.407,
      "step": 1645
    },
    {
      "epoch": 0.6806930693069307,
      "grad_norm": 2.4768276955867794,
      "learning_rate": 2.7986550203915807e-06,
      "loss": 0.3909,
      "step": 1650
    },
    {
      "epoch": 0.6827557755775577,
      "grad_norm": 2.3876332218325365,
      "learning_rate": 2.7663794153927165e-06,
      "loss": 0.3989,
      "step": 1655
    },
    {
      "epoch": 0.6848184818481848,
      "grad_norm": 2.373869717825122,
      "learning_rate": 2.7342196710407337e-06,
      "loss": 0.3848,
      "step": 1660
    },
    {
      "epoch": 0.6868811881188119,
      "grad_norm": 2.2944477847209703,
      "learning_rate": 2.7021774555011214e-06,
      "loss": 0.3952,
      "step": 1665
    },
    {
      "epoch": 0.6889438943894389,
      "grad_norm": 2.4458499611784044,
      "learning_rate": 2.6702544308430122e-06,
      "loss": 0.3841,
      "step": 1670
    },
    {
      "epoch": 0.691006600660066,
      "grad_norm": 2.7433908365448443,
      "learning_rate": 2.6384522529529542e-06,
      "loss": 0.383,
      "step": 1675
    },
    {
      "epoch": 0.693069306930693,
      "grad_norm": 2.4164071285356052,
      "learning_rate": 2.6067725714490307e-06,
      "loss": 0.3841,
      "step": 1680
    },
    {
      "epoch": 0.6951320132013201,
      "grad_norm": 2.5329201914328077,
      "learning_rate": 2.5752170295952856e-06,
      "loss": 0.4024,
      "step": 1685
    },
    {
      "epoch": 0.6971947194719472,
      "grad_norm": 2.362743279232679,
      "learning_rate": 2.5437872642164818e-06,
      "loss": 0.3965,
      "step": 1690
    },
    {
      "epoch": 0.6992574257425742,
      "grad_norm": 2.3597073251243064,
      "learning_rate": 2.5124849056132094e-06,
      "loss": 0.3658,
      "step": 1695
    },
    {
      "epoch": 0.7013201320132013,
      "grad_norm": 2.645876828690864,
      "learning_rate": 2.4813115774773046e-06,
      "loss": 0.3884,
      "step": 1700
    },
    {
      "epoch": 0.7033828382838284,
      "grad_norm": 2.392821189181109,
      "learning_rate": 2.4502688968076416e-06,
      "loss": 0.3827,
      "step": 1705
    },
    {
      "epoch": 0.7054455445544554,
      "grad_norm": 2.4614918453264565,
      "learning_rate": 2.4193584738262426e-06,
      "loss": 0.3825,
      "step": 1710
    },
    {
      "epoch": 0.7075082508250825,
      "grad_norm": 2.3794205872949252,
      "learning_rate": 2.388581911894767e-06,
      "loss": 0.3825,
      "step": 1715
    },
    {
      "epoch": 0.7095709570957096,
      "grad_norm": 2.4464619731962634,
      "learning_rate": 2.357940807431339e-06,
      "loss": 0.3785,
      "step": 1720
    },
    {
      "epoch": 0.7116336633663366,
      "grad_norm": 2.3852404958442013,
      "learning_rate": 2.3274367498277246e-06,
      "loss": 0.3773,
      "step": 1725
    },
    {
      "epoch": 0.7136963696369637,
      "grad_norm": 2.376518354617449,
      "learning_rate": 2.2970713213669127e-06,
      "loss": 0.3831,
      "step": 1730
    },
    {
      "epoch": 0.7157590759075908,
      "grad_norm": 2.6079917176596448,
      "learning_rate": 2.266846097141026e-06,
      "loss": 0.3857,
      "step": 1735
    },
    {
      "epoch": 0.7178217821782178,
      "grad_norm": 2.3458483796617418,
      "learning_rate": 2.2367626449696168e-06,
      "loss": 0.3751,
      "step": 1740
    },
    {
      "epoch": 0.7198844884488449,
      "grad_norm": 2.270229411474977,
      "learning_rate": 2.206822525318352e-06,
      "loss": 0.3784,
      "step": 1745
    },
    {
      "epoch": 0.721947194719472,
      "grad_norm": 2.3820977460708077,
      "learning_rate": 2.1770272912180577e-06,
      "loss": 0.3726,
      "step": 1750
    },
    {
      "epoch": 0.724009900990099,
      "grad_norm": 2.3917376427271746,
      "learning_rate": 2.1473784881841753e-06,
      "loss": 0.3749,
      "step": 1755
    },
    {
      "epoch": 0.7260726072607261,
      "grad_norm": 2.430451456865234,
      "learning_rate": 2.117877654136584e-06,
      "loss": 0.3621,
      "step": 1760
    },
    {
      "epoch": 0.7281353135313532,
      "grad_norm": 2.3121061406028107,
      "learning_rate": 2.088526319319827e-06,
      "loss": 0.3672,
      "step": 1765
    },
    {
      "epoch": 0.7301980198019802,
      "grad_norm": 2.360101764240328,
      "learning_rate": 2.059326006223743e-06,
      "loss": 0.3717,
      "step": 1770
    },
    {
      "epoch": 0.7322607260726073,
      "grad_norm": 2.573193311702909,
      "learning_rate": 2.030278229504484e-06,
      "loss": 0.379,
      "step": 1775
    },
    {
      "epoch": 0.7343234323432343,
      "grad_norm": 2.3389637944355943,
      "learning_rate": 2.001384495905954e-06,
      "loss": 0.3752,
      "step": 1780
    },
    {
      "epoch": 0.7363861386138614,
      "grad_norm": 2.488613631746104,
      "learning_rate": 1.972646304181656e-06,
      "loss": 0.3782,
      "step": 1785
    },
    {
      "epoch": 0.7384488448844885,
      "grad_norm": 2.4022170179819096,
      "learning_rate": 1.944065145016935e-06,
      "loss": 0.3726,
      "step": 1790
    },
    {
      "epoch": 0.7405115511551155,
      "grad_norm": 2.256385150261715,
      "learning_rate": 1.9156425009516736e-06,
      "loss": 0.3688,
      "step": 1795
    },
    {
      "epoch": 0.7425742574257426,
      "grad_norm": 2.4167248024578813,
      "learning_rate": 1.8873798463033742e-06,
      "loss": 0.3684,
      "step": 1800
    },
    {
      "epoch": 0.7446369636963697,
      "grad_norm": 2.451766042702295,
      "learning_rate": 1.8592786470906932e-06,
      "loss": 0.375,
      "step": 1805
    },
    {
      "epoch": 0.7466996699669967,
      "grad_norm": 2.383004059186879,
      "learning_rate": 1.8313403609573976e-06,
      "loss": 0.3759,
      "step": 1810
    },
    {
      "epoch": 0.7487623762376238,
      "grad_norm": 2.12270311435573,
      "learning_rate": 1.8035664370967493e-06,
      "loss": 0.3763,
      "step": 1815
    },
    {
      "epoch": 0.7508250825082509,
      "grad_norm": 2.4415551404560847,
      "learning_rate": 1.775958316176339e-06,
      "loss": 0.3676,
      "step": 1820
    },
    {
      "epoch": 0.7528877887788779,
      "grad_norm": 2.381919666851577,
      "learning_rate": 1.7485174302633557e-06,
      "loss": 0.3707,
      "step": 1825
    },
    {
      "epoch": 0.754950495049505,
      "grad_norm": 2.468112725751951,
      "learning_rate": 1.721245202750299e-06,
      "loss": 0.3743,
      "step": 1830
    },
    {
      "epoch": 0.7570132013201321,
      "grad_norm": 2.4412319646605707,
      "learning_rate": 1.694143048281156e-06,
      "loss": 0.3697,
      "step": 1835
    },
    {
      "epoch": 0.759075907590759,
      "grad_norm": 2.3933092088835592,
      "learning_rate": 1.6672123726780083e-06,
      "loss": 0.362,
      "step": 1840
    },
    {
      "epoch": 0.7611386138613861,
      "grad_norm": 2.3599050961874357,
      "learning_rate": 1.6404545728681232e-06,
      "loss": 0.3621,
      "step": 1845
    },
    {
      "epoch": 0.7632013201320133,
      "grad_norm": 2.3969987242185407,
      "learning_rate": 1.613871036811489e-06,
      "loss": 0.3631,
      "step": 1850
    },
    {
      "epoch": 0.7652640264026402,
      "grad_norm": 2.3728886726584144,
      "learning_rate": 1.5874631434288128e-06,
      "loss": 0.3747,
      "step": 1855
    },
    {
      "epoch": 0.7673267326732673,
      "grad_norm": 2.334377590582883,
      "learning_rate": 1.5612322625300064e-06,
      "loss": 0.3647,
      "step": 1860
    },
    {
      "epoch": 0.7693894389438944,
      "grad_norm": 2.3804587872682865,
      "learning_rate": 1.5351797547431212e-06,
      "loss": 0.3694,
      "step": 1865
    },
    {
      "epoch": 0.7714521452145214,
      "grad_norm": 2.3600418503523444,
      "learning_rate": 1.5093069714437803e-06,
      "loss": 0.3636,
      "step": 1870
    },
    {
      "epoch": 0.7735148514851485,
      "grad_norm": 2.553398892828675,
      "learning_rate": 1.483615254685075e-06,
      "loss": 0.3611,
      "step": 1875
    },
    {
      "epoch": 0.7755775577557755,
      "grad_norm": 2.3499524569210157,
      "learning_rate": 1.4581059371279516e-06,
      "loss": 0.3674,
      "step": 1880
    },
    {
      "epoch": 0.7776402640264026,
      "grad_norm": 2.4025948209352848,
      "learning_rate": 1.4327803419720836e-06,
      "loss": 0.3609,
      "step": 1885
    },
    {
      "epoch": 0.7797029702970297,
      "grad_norm": 2.44960933153857,
      "learning_rate": 1.4076397828872441e-06,
      "loss": 0.3546,
      "step": 1890
    },
    {
      "epoch": 0.7817656765676567,
      "grad_norm": 2.453171947739372,
      "learning_rate": 1.3826855639451492e-06,
      "loss": 0.3675,
      "step": 1895
    },
    {
      "epoch": 0.7838283828382838,
      "grad_norm": 2.243857634486722,
      "learning_rate": 1.357918979551831e-06,
      "loss": 0.3655,
      "step": 1900
    },
    {
      "epoch": 0.7858910891089109,
      "grad_norm": 2.334627109038789,
      "learning_rate": 1.333341314380479e-06,
      "loss": 0.3528,
      "step": 1905
    },
    {
      "epoch": 0.7879537953795379,
      "grad_norm": 2.4450603903304367,
      "learning_rate": 1.308953843304816e-06,
      "loss": 0.3696,
      "step": 1910
    },
    {
      "epoch": 0.790016501650165,
      "grad_norm": 2.469101543725218,
      "learning_rate": 1.2847578313329623e-06,
      "loss": 0.3722,
      "step": 1915
    },
    {
      "epoch": 0.7920792079207921,
      "grad_norm": 2.347298229927237,
      "learning_rate": 1.2607545335418154e-06,
      "loss": 0.364,
      "step": 1920
    },
    {
      "epoch": 0.7941419141914191,
      "grad_norm": 2.478215187683226,
      "learning_rate": 1.2369451950119553e-06,
      "loss": 0.3634,
      "step": 1925
    },
    {
      "epoch": 0.7962046204620462,
      "grad_norm": 2.321071149582765,
      "learning_rate": 1.2133310507630535e-06,
      "loss": 0.3639,
      "step": 1930
    },
    {
      "epoch": 0.7982673267326733,
      "grad_norm": 2.363187519551145,
      "learning_rate": 1.189913325689816e-06,
      "loss": 0.3694,
      "step": 1935
    },
    {
      "epoch": 0.8003300330033003,
      "grad_norm": 2.1565024791128615,
      "learning_rate": 1.166693234498446e-06,
      "loss": 0.3444,
      "step": 1940
    },
    {
      "epoch": 0.8023927392739274,
      "grad_norm": 2.238362330023504,
      "learning_rate": 1.1436719816436293e-06,
      "loss": 0.3535,
      "step": 1945
    },
    {
      "epoch": 0.8044554455445545,
      "grad_norm": 2.682507072878571,
      "learning_rate": 1.120850761266068e-06,
      "loss": 0.3622,
      "step": 1950
    },
    {
      "epoch": 0.8065181518151815,
      "grad_norm": 2.4616448463288974,
      "learning_rate": 1.098230757130529e-06,
      "loss": 0.3521,
      "step": 1955
    },
    {
      "epoch": 0.8085808580858086,
      "grad_norm": 2.518356814157187,
      "learning_rate": 1.075813142564448e-06,
      "loss": 0.3621,
      "step": 1960
    },
    {
      "epoch": 0.8106435643564357,
      "grad_norm": 2.41717183499352,
      "learning_rate": 1.053599080397068e-06,
      "loss": 0.3534,
      "step": 1965
    },
    {
      "epoch": 0.8127062706270627,
      "grad_norm": 2.3274563555835357,
      "learning_rate": 1.031589722899109e-06,
      "loss": 0.3497,
      "step": 1970
    },
    {
      "epoch": 0.8147689768976898,
      "grad_norm": 2.41058923252155,
      "learning_rate": 1.0097862117230162e-06,
      "loss": 0.3453,
      "step": 1975
    },
    {
      "epoch": 0.8168316831683168,
      "grad_norm": 2.441792027931522,
      "learning_rate": 9.881896778437328e-07,
      "loss": 0.3444,
      "step": 1980
    },
    {
      "epoch": 0.8188943894389439,
      "grad_norm": 2.4018876695962383,
      "learning_rate": 9.6680124150003e-07,
      "loss": 0.3654,
      "step": 1985
    },
    {
      "epoch": 0.820957095709571,
      "grad_norm": 2.3575482669225902,
      "learning_rate": 9.456220121364091e-07,
      "loss": 0.348,
      "step": 1990
    },
    {
      "epoch": 0.823019801980198,
      "grad_norm": 2.525821025154249,
      "learning_rate": 9.24653088345544e-07,
      "loss": 0.3523,
      "step": 1995
    },
    {
      "epoch": 0.8250825082508251,
      "grad_norm": 2.7783807540038294,
      "learning_rate": 9.038955578113018e-07,
      "loss": 0.3582,
      "step": 2000
    },
    {
      "epoch": 0.8271452145214522,
      "grad_norm": 2.50572956814031,
      "learning_rate": 8.833504972523238e-07,
      "loss": 0.3593,
      "step": 2005
    },
    {
      "epoch": 0.8292079207920792,
      "grad_norm": 2.252557918607386,
      "learning_rate": 8.630189723661663e-07,
      "loss": 0.3536,
      "step": 2010
    },
    {
      "epoch": 0.8312706270627063,
      "grad_norm": 2.4651130321437758,
      "learning_rate": 8.429020377740338e-07,
      "loss": 0.3535,
      "step": 2015
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 2.400123594291822,
      "learning_rate": 8.230007369660636e-07,
      "loss": 0.3543,
      "step": 2020
    },
    {
      "epoch": 0.8353960396039604,
      "grad_norm": 2.2893414346006806,
      "learning_rate": 8.033161022472063e-07,
      "loss": 0.3633,
      "step": 2025
    },
    {
      "epoch": 0.8374587458745875,
      "grad_norm": 2.387886294048468,
      "learning_rate": 7.838491546836763e-07,
      "loss": 0.3501,
      "step": 2030
    },
    {
      "epoch": 0.8395214521452146,
      "grad_norm": 2.380305733427394,
      "learning_rate": 7.646009040499846e-07,
      "loss": 0.3396,
      "step": 2035
    },
    {
      "epoch": 0.8415841584158416,
      "grad_norm": 2.6014881264846164,
      "learning_rate": 7.455723487765664e-07,
      "loss": 0.3586,
      "step": 2040
    },
    {
      "epoch": 0.8436468646864687,
      "grad_norm": 2.4907278984166106,
      "learning_rate": 7.267644758979869e-07,
      "loss": 0.3468,
      "step": 2045
    },
    {
      "epoch": 0.8457095709570958,
      "grad_norm": 2.34019203542543,
      "learning_rate": 7.08178261001743e-07,
      "loss": 0.3514,
      "step": 2050
    },
    {
      "epoch": 0.8477722772277227,
      "grad_norm": 2.3922323702113637,
      "learning_rate": 6.898146681776629e-07,
      "loss": 0.3486,
      "step": 2055
    },
    {
      "epoch": 0.8498349834983498,
      "grad_norm": 2.3826898973626127,
      "learning_rate": 6.7167464996789e-07,
      "loss": 0.3621,
      "step": 2060
    },
    {
      "epoch": 0.851897689768977,
      "grad_norm": 2.3165319710343324,
      "learning_rate": 6.537591473174814e-07,
      "loss": 0.3413,
      "step": 2065
    },
    {
      "epoch": 0.8539603960396039,
      "grad_norm": 2.5692025408049064,
      "learning_rate": 6.360690895255916e-07,
      "loss": 0.3461,
      "step": 2070
    },
    {
      "epoch": 0.856023102310231,
      "grad_norm": 2.272769119087517,
      "learning_rate": 6.186053941972775e-07,
      "loss": 0.348,
      "step": 2075
    },
    {
      "epoch": 0.858085808580858,
      "grad_norm": 2.4550162262909447,
      "learning_rate": 6.013689671958944e-07,
      "loss": 0.3531,
      "step": 2080
    },
    {
      "epoch": 0.8601485148514851,
      "grad_norm": 2.388763908670921,
      "learning_rate": 5.84360702596109e-07,
      "loss": 0.3608,
      "step": 2085
    },
    {
      "epoch": 0.8622112211221122,
      "grad_norm": 2.5005604744641174,
      "learning_rate": 5.67581482637527e-07,
      "loss": 0.3444,
      "step": 2090
    },
    {
      "epoch": 0.8642739273927392,
      "grad_norm": 2.4560197080279234,
      "learning_rate": 5.510321776789213e-07,
      "loss": 0.3463,
      "step": 2095
    },
    {
      "epoch": 0.8663366336633663,
      "grad_norm": 2.365882695902703,
      "learning_rate": 5.347136461530966e-07,
      "loss": 0.3366,
      "step": 2100
    },
    {
      "epoch": 0.8683993399339934,
      "grad_norm": 2.4380355384519925,
      "learning_rate": 5.186267345223539e-07,
      "loss": 0.3385,
      "step": 2105
    },
    {
      "epoch": 0.8704620462046204,
      "grad_norm": 2.399986551951994,
      "learning_rate": 5.027722772345828e-07,
      "loss": 0.3448,
      "step": 2110
    },
    {
      "epoch": 0.8725247524752475,
      "grad_norm": 2.454557551769108,
      "learning_rate": 4.871510966799847e-07,
      "loss": 0.3507,
      "step": 2115
    },
    {
      "epoch": 0.8745874587458746,
      "grad_norm": 2.456742477580607,
      "learning_rate": 4.717640031484055e-07,
      "loss": 0.3525,
      "step": 2120
    },
    {
      "epoch": 0.8766501650165016,
      "grad_norm": 2.41826728768274,
      "learning_rate": 4.566117947873139e-07,
      "loss": 0.3526,
      "step": 2125
    },
    {
      "epoch": 0.8787128712871287,
      "grad_norm": 2.4671547541589183,
      "learning_rate": 4.4169525756039164e-07,
      "loss": 0.3375,
      "step": 2130
    },
    {
      "epoch": 0.8807755775577558,
      "grad_norm": 2.3142164149499793,
      "learning_rate": 4.2701516520677054e-07,
      "loss": 0.3428,
      "step": 2135
    },
    {
      "epoch": 0.8828382838283828,
      "grad_norm": 2.282591060676799,
      "learning_rate": 4.1257227920089684e-07,
      "loss": 0.315,
      "step": 2140
    },
    {
      "epoch": 0.8849009900990099,
      "grad_norm": 2.2945870317486445,
      "learning_rate": 3.983673487130313e-07,
      "loss": 0.3438,
      "step": 2145
    },
    {
      "epoch": 0.886963696369637,
      "grad_norm": 2.4023456680454003,
      "learning_rate": 3.8440111057038874e-07,
      "loss": 0.3466,
      "step": 2150
    },
    {
      "epoch": 0.889026402640264,
      "grad_norm": 2.4628827167617278,
      "learning_rate": 3.706742892189197e-07,
      "loss": 0.3578,
      "step": 2155
    },
    {
      "epoch": 0.8910891089108911,
      "grad_norm": 2.518821614068097,
      "learning_rate": 3.5718759668572913e-07,
      "loss": 0.3484,
      "step": 2160
    },
    {
      "epoch": 0.8931518151815182,
      "grad_norm": 2.4044153870357734,
      "learning_rate": 3.439417325421468e-07,
      "loss": 0.3331,
      "step": 2165
    },
    {
      "epoch": 0.8952145214521452,
      "grad_norm": 2.263400317365007,
      "learning_rate": 3.3093738386743734e-07,
      "loss": 0.3359,
      "step": 2170
    },
    {
      "epoch": 0.8972772277227723,
      "grad_norm": 2.4013680975289162,
      "learning_rate": 3.1817522521316034e-07,
      "loss": 0.3509,
      "step": 2175
    },
    {
      "epoch": 0.8993399339933993,
      "grad_norm": 2.4793048291891324,
      "learning_rate": 3.0565591856818236e-07,
      "loss": 0.3499,
      "step": 2180
    },
    {
      "epoch": 0.9014026402640264,
      "grad_norm": 2.524550570591509,
      "learning_rate": 2.9338011332433525e-07,
      "loss": 0.3314,
      "step": 2185
    },
    {
      "epoch": 0.9034653465346535,
      "grad_norm": 2.4135168075751294,
      "learning_rate": 2.813484462427357e-07,
      "loss": 0.3359,
      "step": 2190
    },
    {
      "epoch": 0.9055280528052805,
      "grad_norm": 2.4118495326051996,
      "learning_rate": 2.695615414207542e-07,
      "loss": 0.3499,
      "step": 2195
    },
    {
      "epoch": 0.9075907590759076,
      "grad_norm": 2.392533982344082,
      "learning_rate": 2.5802001025963917e-07,
      "loss": 0.3417,
      "step": 2200
    },
    {
      "epoch": 0.9096534653465347,
      "grad_norm": 2.2880229451881418,
      "learning_rate": 2.467244514328082e-07,
      "loss": 0.3493,
      "step": 2205
    },
    {
      "epoch": 0.9117161716171617,
      "grad_norm": 2.3091247857735504,
      "learning_rate": 2.3567545085478983e-07,
      "loss": 0.3383,
      "step": 2210
    },
    {
      "epoch": 0.9137788778877888,
      "grad_norm": 2.4901336136450136,
      "learning_rate": 2.248735816508324e-07,
      "loss": 0.3398,
      "step": 2215
    },
    {
      "epoch": 0.9158415841584159,
      "grad_norm": 2.3142795015199114,
      "learning_rate": 2.1431940412717843e-07,
      "loss": 0.3427,
      "step": 2220
    },
    {
      "epoch": 0.9179042904290429,
      "grad_norm": 2.4277647469057744,
      "learning_rate": 2.040134657419951e-07,
      "loss": 0.3387,
      "step": 2225
    },
    {
      "epoch": 0.91996699669967,
      "grad_norm": 2.3617745475614034,
      "learning_rate": 1.9395630107698293e-07,
      "loss": 0.3392,
      "step": 2230
    },
    {
      "epoch": 0.9220297029702971,
      "grad_norm": 2.5202449484112615,
      "learning_rate": 1.8414843180964316e-07,
      "loss": 0.3451,
      "step": 2235
    },
    {
      "epoch": 0.9240924092409241,
      "grad_norm": 2.6186027806455727,
      "learning_rate": 1.7459036668621586e-07,
      "loss": 0.345,
      "step": 2240
    },
    {
      "epoch": 0.9261551155115512,
      "grad_norm": 2.357402116923623,
      "learning_rate": 1.6528260149529573e-07,
      "loss": 0.3426,
      "step": 2245
    },
    {
      "epoch": 0.9282178217821783,
      "grad_norm": 2.7714941079884365,
      "learning_rate": 1.562256190421102e-07,
      "loss": 0.3444,
      "step": 2250
    },
    {
      "epoch": 0.9302805280528053,
      "grad_norm": 2.384372509379296,
      "learning_rate": 1.4741988912347848e-07,
      "loss": 0.349,
      "step": 2255
    },
    {
      "epoch": 0.9323432343234324,
      "grad_norm": 2.423642340690504,
      "learning_rate": 1.3886586850344276e-07,
      "loss": 0.3411,
      "step": 2260
    },
    {
      "epoch": 0.9344059405940595,
      "grad_norm": 2.4976812347623762,
      "learning_rate": 1.30564000889572e-07,
      "loss": 0.3427,
      "step": 2265
    },
    {
      "epoch": 0.9364686468646864,
      "grad_norm": 2.4409748992106244,
      "learning_rate": 1.225147169099511e-07,
      "loss": 0.3344,
      "step": 2270
    },
    {
      "epoch": 0.9385313531353136,
      "grad_norm": 2.4594177758611226,
      "learning_rate": 1.147184340908386e-07,
      "loss": 0.3466,
      "step": 2275
    },
    {
      "epoch": 0.9405940594059405,
      "grad_norm": 2.3563533473196054,
      "learning_rate": 1.0717555683501413e-07,
      "loss": 0.3437,
      "step": 2280
    },
    {
      "epoch": 0.9426567656765676,
      "grad_norm": 2.438141654343306,
      "learning_rate": 9.988647640079785e-08,
      "loss": 0.3481,
      "step": 2285
    },
    {
      "epoch": 0.9447194719471947,
      "grad_norm": 2.3653534995422465,
      "learning_rate": 9.285157088175678e-08,
      "loss": 0.3305,
      "step": 2290
    },
    {
      "epoch": 0.9467821782178217,
      "grad_norm": 2.5595084581881893,
      "learning_rate": 8.607120518709156e-08,
      "loss": 0.3447,
      "step": 2295
    },
    {
      "epoch": 0.9488448844884488,
      "grad_norm": 2.298821712031628,
      "learning_rate": 7.954573102271157e-08,
      "loss": 0.3427,
      "step": 2300
    },
    {
      "epoch": 0.9509075907590759,
      "grad_norm": 2.3999698309349786,
      "learning_rate": 7.327548687298625e-08,
      "loss": 0.3387,
      "step": 2305
    },
    {
      "epoch": 0.9529702970297029,
      "grad_norm": 2.2525707814009617,
      "learning_rate": 6.726079798319185e-08,
      "loss": 0.3561,
      "step": 2310
    },
    {
      "epoch": 0.95503300330033,
      "grad_norm": 2.3217448485216248,
      "learning_rate": 6.150197634263888e-08,
      "loss": 0.3469,
      "step": 2315
    },
    {
      "epoch": 0.9570957095709571,
      "grad_norm": 2.365101163841893,
      "learning_rate": 5.599932066848834e-08,
      "loss": 0.3367,
      "step": 2320
    },
    {
      "epoch": 0.9591584158415841,
      "grad_norm": 2.425841031506538,
      "learning_rate": 5.0753116390258594e-08,
      "loss": 0.3401,
      "step": 2325
    },
    {
      "epoch": 0.9612211221122112,
      "grad_norm": 2.563836501970768,
      "learning_rate": 4.576363563501718e-08,
      "loss": 0.3409,
      "step": 2330
    },
    {
      "epoch": 0.9632838283828383,
      "grad_norm": 2.34358481427922,
      "learning_rate": 4.103113721326768e-08,
      "loss": 0.3484,
      "step": 2335
    },
    {
      "epoch": 0.9653465346534653,
      "grad_norm": 2.402352860387557,
      "learning_rate": 3.655586660552324e-08,
      "loss": 0.3419,
      "step": 2340
    },
    {
      "epoch": 0.9674092409240924,
      "grad_norm": 2.412967620073454,
      "learning_rate": 3.233805594957506e-08,
      "loss": 0.3272,
      "step": 2345
    },
    {
      "epoch": 0.9694719471947195,
      "grad_norm": 2.3840732180526487,
      "learning_rate": 2.8377924028449855e-08,
      "loss": 0.3383,
      "step": 2350
    },
    {
      "epoch": 0.9715346534653465,
      "grad_norm": 2.508928823648118,
      "learning_rate": 2.4675676259059976e-08,
      "loss": 0.3447,
      "step": 2355
    },
    {
      "epoch": 0.9735973597359736,
      "grad_norm": 2.1993358612568317,
      "learning_rate": 2.123150468155144e-08,
      "loss": 0.3294,
      "step": 2360
    },
    {
      "epoch": 0.9756600660066007,
      "grad_norm": 2.4190680800537567,
      "learning_rate": 1.8045587949339637e-08,
      "loss": 0.3351,
      "step": 2365
    },
    {
      "epoch": 0.9777227722772277,
      "grad_norm": 2.2890929930955037,
      "learning_rate": 1.5118091319843985e-08,
      "loss": 0.34,
      "step": 2370
    },
    {
      "epoch": 0.9797854785478548,
      "grad_norm": 2.5688702019838883,
      "learning_rate": 1.2449166645915333e-08,
      "loss": 0.3304,
      "step": 2375
    },
    {
      "epoch": 0.9818481848184818,
      "grad_norm": 2.483690216637316,
      "learning_rate": 1.0038952367958376e-08,
      "loss": 0.341,
      "step": 2380
    },
    {
      "epoch": 0.9839108910891089,
      "grad_norm": 2.179239333588943,
      "learning_rate": 7.887573506752954e-09,
      "loss": 0.3384,
      "step": 2385
    },
    {
      "epoch": 0.985973597359736,
      "grad_norm": 2.512101769493497,
      "learning_rate": 5.9951416569659085e-09,
      "loss": 0.3348,
      "step": 2390
    },
    {
      "epoch": 0.988036303630363,
      "grad_norm": 2.4449313798457597,
      "learning_rate": 4.361754981365152e-09,
      "loss": 0.3339,
      "step": 2395
    },
    {
      "epoch": 0.9900990099009901,
      "grad_norm": 2.2966537435115444,
      "learning_rate": 2.98749820572708e-09,
      "loss": 0.3263,
      "step": 2400
    },
    {
      "epoch": 0.9921617161716172,
      "grad_norm": 2.5187709917417727,
      "learning_rate": 1.8724426144395293e-09,
      "loss": 0.3357,
      "step": 2405
    },
    {
      "epoch": 0.9942244224422442,
      "grad_norm": 2.340909896310804,
      "learning_rate": 1.0166460468080674e-09,
      "loss": 0.3426,
      "step": 2410
    },
    {
      "epoch": 0.9962871287128713,
      "grad_norm": 2.3871751186126207,
      "learning_rate": 4.2015289405339386e-10,
      "loss": 0.3369,
      "step": 2415
    },
    {
      "epoch": 0.9983498349834984,
      "grad_norm": 2.377304997896102,
      "learning_rate": 8.29940970092924e-11,
      "loss": 0.3308,
      "step": 2420
    },
    {
      "epoch": 1.0,
      "eval_runtime": 3.3729,
      "eval_samples_per_second": 2.965,
      "eval_steps_per_second": 0.889,
      "step": 2424
    },
    {
      "epoch": 1.0,
      "step": 2424,
      "total_flos": 253768142684160.0,
      "train_loss": 0.5300087753695624,
      "train_runtime": 21368.6318,
      "train_samples_per_second": 1.814,
      "train_steps_per_second": 0.113
    }
  ],
  "logging_steps": 5,
  "max_steps": 2424,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 253768142684160.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}