{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.4276080920269734,
  "eval_steps": 500,
  "global_step": 1800,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.007933359777865925,
      "grad_norm": 19.14600299430099,
      "learning_rate": 8.9197224975223e-07,
      "loss": 8.2489,
      "step": 10
    },
    {
      "epoch": 0.01586671955573185,
      "grad_norm": 18.01915458135243,
      "learning_rate": 1.8830525272547078e-06,
      "loss": 8.1177,
      "step": 20
    },
    {
      "epoch": 0.02380007933359778,
      "grad_norm": 9.33822311099433,
      "learning_rate": 2.8741328047571853e-06,
      "loss": 7.4735,
      "step": 30
    },
    {
      "epoch": 0.0317334391114637,
      "grad_norm": 4.906820841029322,
      "learning_rate": 3.8652130822596635e-06,
      "loss": 6.7427,
      "step": 40
    },
    {
      "epoch": 0.039666798889329634,
      "grad_norm": 3.0516622673060785,
      "learning_rate": 4.8562933597621405e-06,
      "loss": 6.2707,
      "step": 50
    },
    {
      "epoch": 0.04760015866719556,
      "grad_norm": 3.0497788345357493,
      "learning_rate": 5.847373637264618e-06,
      "loss": 5.9319,
      "step": 60
    },
    {
      "epoch": 0.055533518445061485,
      "grad_norm": 3.678614138586657,
      "learning_rate": 6.838453914767097e-06,
      "loss": 5.5292,
      "step": 70
    },
    {
      "epoch": 0.0634668782229274,
      "grad_norm": 4.261516409919679,
      "learning_rate": 7.829534192269576e-06,
      "loss": 4.9099,
      "step": 80
    },
    {
      "epoch": 0.07140023800079334,
      "grad_norm": 4.6638020698067875,
      "learning_rate": 8.820614469772052e-06,
      "loss": 4.2313,
      "step": 90
    },
    {
      "epoch": 0.07933359777865927,
      "grad_norm": 4.5951745742169,
      "learning_rate": 9.81169474727453e-06,
      "loss": 3.663,
      "step": 100
    },
    {
      "epoch": 0.0872669575565252,
      "grad_norm": 3.9132985800446947,
      "learning_rate": 1.0802775024777007e-05,
      "loss": 3.2806,
      "step": 110
    },
    {
      "epoch": 0.09520031733439112,
      "grad_norm": 3.864223439251188,
      "learning_rate": 1.1793855302279485e-05,
      "loss": 3.0464,
      "step": 120
    },
    {
      "epoch": 0.10313367711225704,
      "grad_norm": 3.5022825994548232,
      "learning_rate": 1.2784935579781963e-05,
      "loss": 2.8886,
      "step": 130
    },
    {
      "epoch": 0.11106703689012297,
      "grad_norm": 5.1510905388003,
      "learning_rate": 1.377601585728444e-05,
      "loss": 2.7668,
      "step": 140
    },
    {
      "epoch": 0.1190003966679889,
      "grad_norm": 7.21771904799698,
      "learning_rate": 1.4767096134786917e-05,
      "loss": 2.6692,
      "step": 150
    },
    {
      "epoch": 0.1269337564458548,
      "grad_norm": 9.497867401163038,
      "learning_rate": 1.5758176412289396e-05,
      "loss": 2.5934,
      "step": 160
    },
    {
      "epoch": 0.13486711622372075,
      "grad_norm": 8.479404348386907,
      "learning_rate": 1.6749256689791874e-05,
      "loss": 2.5471,
      "step": 170
    },
    {
      "epoch": 0.14280047600158668,
      "grad_norm": 13.33424723235721,
      "learning_rate": 1.7740336967294352e-05,
      "loss": 2.4827,
      "step": 180
    },
    {
      "epoch": 0.1507338357794526,
      "grad_norm": 9.736118445504315,
      "learning_rate": 1.873141724479683e-05,
      "loss": 2.4223,
      "step": 190
    },
    {
      "epoch": 0.15866719555731854,
      "grad_norm": 12.548385792238333,
      "learning_rate": 1.9722497522299307e-05,
      "loss": 2.3752,
      "step": 200
    },
    {
      "epoch": 0.16660055533518445,
      "grad_norm": 15.910150113440148,
      "learning_rate": 2.0713577799801785e-05,
      "loss": 2.3339,
      "step": 210
    },
    {
      "epoch": 0.1745339151130504,
      "grad_norm": 12.024121798784853,
      "learning_rate": 2.1704658077304263e-05,
      "loss": 2.2732,
      "step": 220
    },
    {
      "epoch": 0.1824672748909163,
      "grad_norm": 18.143793557621525,
      "learning_rate": 2.269573835480674e-05,
      "loss": 2.2154,
      "step": 230
    },
    {
      "epoch": 0.19040063466878224,
      "grad_norm": 11.930035885996817,
      "learning_rate": 2.368681863230922e-05,
      "loss": 2.1602,
      "step": 240
    },
    {
      "epoch": 0.19833399444664815,
      "grad_norm": 12.424586861990159,
      "learning_rate": 2.4677898909811696e-05,
      "loss": 2.0887,
      "step": 250
    },
    {
      "epoch": 0.2062673542245141,
      "grad_norm": 11.063682700121975,
      "learning_rate": 2.5668979187314174e-05,
      "loss": 1.9843,
      "step": 260
    },
    {
      "epoch": 0.21420071400238,
      "grad_norm": 9.80005982816964,
      "learning_rate": 2.6660059464816652e-05,
      "loss": 1.871,
      "step": 270
    },
    {
      "epoch": 0.22213407378024594,
      "grad_norm": 6.8486046511969985,
      "learning_rate": 2.7651139742319126e-05,
      "loss": 1.7654,
      "step": 280
    },
    {
      "epoch": 0.23006743355811185,
      "grad_norm": 4.650133260417432,
      "learning_rate": 2.8642220019821607e-05,
      "loss": 1.6631,
      "step": 290
    },
    {
      "epoch": 0.2380007933359778,
      "grad_norm": 8.484902376448611,
      "learning_rate": 2.9633300297324085e-05,
      "loss": 1.587,
      "step": 300
    },
    {
      "epoch": 0.2459341531138437,
      "grad_norm": 8.266296207037957,
      "learning_rate": 3.062438057482656e-05,
      "loss": 1.5209,
      "step": 310
    },
    {
      "epoch": 0.2538675128917096,
      "grad_norm": 6.035258103391185,
      "learning_rate": 3.161546085232904e-05,
      "loss": 1.4691,
      "step": 320
    },
    {
      "epoch": 0.2618008726695756,
      "grad_norm": 6.090335574638085,
      "learning_rate": 3.260654112983152e-05,
      "loss": 1.4213,
      "step": 330
    },
    {
      "epoch": 0.2697342324474415,
      "grad_norm": 7.504625709152101,
      "learning_rate": 3.359762140733399e-05,
      "loss": 1.3809,
      "step": 340
    },
    {
      "epoch": 0.2776675922253074,
      "grad_norm": 5.216278847793629,
      "learning_rate": 3.458870168483647e-05,
      "loss": 1.3304,
      "step": 350
    },
    {
      "epoch": 0.28560095200317337,
      "grad_norm": 5.001664767324889,
      "learning_rate": 3.5579781962338955e-05,
      "loss": 1.2886,
      "step": 360
    },
    {
      "epoch": 0.2935343117810393,
      "grad_norm": 4.035122456050597,
      "learning_rate": 3.657086223984143e-05,
      "loss": 1.2441,
      "step": 370
    },
    {
      "epoch": 0.3014676715589052,
      "grad_norm": 4.3131477681191015,
      "learning_rate": 3.7561942517343904e-05,
      "loss": 1.1973,
      "step": 380
    },
    {
      "epoch": 0.3094010313367711,
      "grad_norm": 8.188203831814278,
      "learning_rate": 3.855302279484638e-05,
      "loss": 1.175,
      "step": 390
    },
    {
      "epoch": 0.31733439111463707,
      "grad_norm": 4.755933403654607,
      "learning_rate": 3.9544103072348866e-05,
      "loss": 1.1342,
      "step": 400
    },
    {
      "epoch": 0.325267750892503,
      "grad_norm": 6.57796715979494,
      "learning_rate": 4.0535183349851344e-05,
      "loss": 1.0971,
      "step": 410
    },
    {
      "epoch": 0.3332011106703689,
      "grad_norm": 4.216167153902993,
      "learning_rate": 4.1526263627353815e-05,
      "loss": 1.0513,
      "step": 420
    },
    {
      "epoch": 0.3411344704482348,
      "grad_norm": 8.418073662598946,
      "learning_rate": 4.251734390485629e-05,
      "loss": 1.0278,
      "step": 430
    },
    {
      "epoch": 0.3490678302261008,
      "grad_norm": 4.293832971560352,
      "learning_rate": 4.350842418235878e-05,
      "loss": 0.9989,
      "step": 440
    },
    {
      "epoch": 0.3570011900039667,
      "grad_norm": 4.738542701520305,
      "learning_rate": 4.4499504459861255e-05,
      "loss": 0.9646,
      "step": 450
    },
    {
      "epoch": 0.3649345497818326,
      "grad_norm": 3.836235781620353,
      "learning_rate": 4.5490584737363726e-05,
      "loss": 0.9282,
      "step": 460
    },
    {
      "epoch": 0.3728679095596985,
      "grad_norm": 4.385628595929679,
      "learning_rate": 4.6481665014866204e-05,
      "loss": 0.9069,
      "step": 470
    },
    {
      "epoch": 0.3808012693375645,
      "grad_norm": 4.795190260431852,
      "learning_rate": 4.747274529236869e-05,
      "loss": 0.8767,
      "step": 480
    },
    {
      "epoch": 0.3887346291154304,
      "grad_norm": 5.15771361003216,
      "learning_rate": 4.846382556987116e-05,
      "loss": 0.848,
      "step": 490
    },
    {
      "epoch": 0.3966679888932963,
      "grad_norm": 5.00866147389186,
      "learning_rate": 4.945490584737364e-05,
      "loss": 0.832,
      "step": 500
    },
    {
      "epoch": 0.40460134867116226,
      "grad_norm": 3.3749478011742617,
      "learning_rate": 5.044598612487612e-05,
      "loss": 0.801,
      "step": 510
    },
    {
      "epoch": 0.4125347084490282,
      "grad_norm": 5.648100107599949,
      "learning_rate": 5.14370664023786e-05,
      "loss": 0.7907,
      "step": 520
    },
    {
      "epoch": 0.4204680682268941,
      "grad_norm": 5.180183508710987,
      "learning_rate": 5.242814667988107e-05,
      "loss": 0.7641,
      "step": 530
    },
    {
      "epoch": 0.42840142800476,
      "grad_norm": 4.795683231567134,
      "learning_rate": 5.341922695738355e-05,
      "loss": 0.7428,
      "step": 540
    },
    {
      "epoch": 0.43633478778262597,
      "grad_norm": 4.474423469358904,
      "learning_rate": 5.4410307234886027e-05,
      "loss": 0.7178,
      "step": 550
    },
    {
      "epoch": 0.4442681475604919,
      "grad_norm": 5.965818045702901,
      "learning_rate": 5.5401387512388504e-05,
      "loss": 0.7008,
      "step": 560
    },
    {
      "epoch": 0.4522015073383578,
      "grad_norm": 4.4181008640757256,
      "learning_rate": 5.639246778989098e-05,
      "loss": 0.678,
      "step": 570
    },
    {
      "epoch": 0.4601348671162237,
      "grad_norm": 4.925348313364693,
      "learning_rate": 5.738354806739347e-05,
      "loss": 0.6516,
      "step": 580
    },
    {
      "epoch": 0.46806822689408967,
      "grad_norm": 4.995771437156914,
      "learning_rate": 5.8374628344895944e-05,
      "loss": 0.6261,
      "step": 590
    },
    {
      "epoch": 0.4760015866719556,
      "grad_norm": 4.3400967454545185,
      "learning_rate": 5.936570862239842e-05,
      "loss": 0.6099,
      "step": 600
    },
    {
      "epoch": 0.4839349464498215,
      "grad_norm": 3.8437851403940013,
      "learning_rate": 6.035678889990089e-05,
      "loss": 0.5905,
      "step": 610
    },
    {
      "epoch": 0.4918683062276874,
      "grad_norm": 4.673291642824443,
      "learning_rate": 6.134786917740337e-05,
      "loss": 0.5739,
      "step": 620
    },
    {
      "epoch": 0.49980166600555337,
      "grad_norm": 4.691220121981611,
      "learning_rate": 6.233894945490585e-05,
      "loss": 0.5543,
      "step": 630
    },
    {
      "epoch": 0.5077350257834192,
      "grad_norm": 4.609768045302665,
      "learning_rate": 6.333002973240833e-05,
      "loss": 0.5265,
      "step": 640
    },
    {
      "epoch": 0.5156683855612852,
      "grad_norm": 3.2621270260578967,
      "learning_rate": 6.43211100099108e-05,
      "loss": 0.5175,
      "step": 650
    },
    {
      "epoch": 0.5236017453391512,
      "grad_norm": 3.5703471859456806,
      "learning_rate": 6.531219028741328e-05,
      "loss": 0.4972,
      "step": 660
    },
    {
      "epoch": 0.5315351051170171,
      "grad_norm": 3.530519835264167,
      "learning_rate": 6.630327056491576e-05,
      "loss": 0.4717,
      "step": 670
    },
    {
      "epoch": 0.539468464894883,
      "grad_norm": 2.392046097161058,
      "learning_rate": 6.729435084241824e-05,
      "loss": 0.4583,
      "step": 680
    },
    {
      "epoch": 0.5474018246727489,
      "grad_norm": 3.149236166223247,
      "learning_rate": 6.828543111992072e-05,
      "loss": 0.4412,
      "step": 690
    },
    {
      "epoch": 0.5553351844506148,
      "grad_norm": 2.9354688050931586,
      "learning_rate": 6.92765113974232e-05,
      "loss": 0.423,
      "step": 700
    },
    {
      "epoch": 0.5632685442284807,
      "grad_norm": 2.7127436235307125,
      "learning_rate": 7.026759167492567e-05,
      "loss": 0.4062,
      "step": 710
    },
    {
      "epoch": 0.5712019040063467,
      "grad_norm": 2.6410601628390777,
      "learning_rate": 7.125867195242815e-05,
      "loss": 0.3923,
      "step": 720
    },
    {
      "epoch": 0.5791352637842127,
      "grad_norm": 2.7996527641631386,
      "learning_rate": 7.224975222993063e-05,
      "loss": 0.3739,
      "step": 730
    },
    {
      "epoch": 0.5870686235620786,
      "grad_norm": 3.53218129873484,
      "learning_rate": 7.32408325074331e-05,
      "loss": 0.3558,
      "step": 740
    },
    {
      "epoch": 0.5950019833399445,
      "grad_norm": 3.256471177277615,
      "learning_rate": 7.423191278493558e-05,
      "loss": 0.3378,
      "step": 750
    },
    {
      "epoch": 0.6029353431178104,
      "grad_norm": 2.448855772936633,
      "learning_rate": 7.522299306243806e-05,
      "loss": 0.3318,
      "step": 760
    },
    {
      "epoch": 0.6108687028956763,
      "grad_norm": 2.8206966008765355,
      "learning_rate": 7.621407333994054e-05,
      "loss": 0.3145,
      "step": 770
    },
    {
      "epoch": 0.6188020626735422,
      "grad_norm": 2.262400615934418,
      "learning_rate": 7.720515361744302e-05,
      "loss": 0.2981,
      "step": 780
    },
    {
      "epoch": 0.6267354224514081,
      "grad_norm": 2.6691511066584894,
      "learning_rate": 7.81962338949455e-05,
      "loss": 0.2878,
      "step": 790
    },
    {
      "epoch": 0.6346687822292741,
      "grad_norm": 2.1502320771866494,
      "learning_rate": 7.918731417244797e-05,
      "loss": 0.2724,
      "step": 800
    },
    {
      "epoch": 0.64260214200714,
      "grad_norm": 2.1869290626436806,
      "learning_rate": 8.017839444995045e-05,
      "loss": 0.2589,
      "step": 810
    },
    {
      "epoch": 0.650535501785006,
      "grad_norm": 2.6277829837493476,
      "learning_rate": 8.116947472745293e-05,
      "loss": 0.2481,
      "step": 820
    },
    {
      "epoch": 0.6584688615628719,
      "grad_norm": 1.8401496589236042,
      "learning_rate": 8.21605550049554e-05,
      "loss": 0.2361,
      "step": 830
    },
    {
      "epoch": 0.6664022213407378,
      "grad_norm": 1.7862155383026863,
      "learning_rate": 8.315163528245788e-05,
      "loss": 0.222,
      "step": 840
    },
    {
      "epoch": 0.6743355811186037,
      "grad_norm": 1.9765212410305726,
      "learning_rate": 8.414271555996036e-05,
      "loss": 0.2103,
      "step": 850
    },
    {
      "epoch": 0.6822689408964696,
      "grad_norm": 2.4397348134664556,
      "learning_rate": 8.513379583746284e-05,
      "loss": 0.2021,
      "step": 860
    },
    {
      "epoch": 0.6902023006743356,
      "grad_norm": 2.296758487647809,
      "learning_rate": 8.612487611496532e-05,
      "loss": 0.1924,
      "step": 870
    },
    {
      "epoch": 0.6981356604522015,
      "grad_norm": 2.104859870669743,
      "learning_rate": 8.71159563924678e-05,
      "loss": 0.1817,
      "step": 880
    },
    {
      "epoch": 0.7060690202300675,
      "grad_norm": 1.3003724019670297,
      "learning_rate": 8.810703666997027e-05,
      "loss": 0.1732,
      "step": 890
    },
    {
      "epoch": 0.7140023800079334,
      "grad_norm": 1.3709932626535646,
      "learning_rate": 8.909811694747275e-05,
      "loss": 0.1593,
      "step": 900
    },
    {
      "epoch": 0.7219357397857993,
      "grad_norm": 1.8342489151171628,
      "learning_rate": 9.008919722497523e-05,
      "loss": 0.1551,
      "step": 910
    },
    {
      "epoch": 0.7298690995636652,
      "grad_norm": 1.762934595290678,
      "learning_rate": 9.10802775024777e-05,
      "loss": 0.1489,
      "step": 920
    },
    {
      "epoch": 0.7378024593415311,
      "grad_norm": 1.4075349928757257,
      "learning_rate": 9.207135777998018e-05,
      "loss": 0.1391,
      "step": 930
    },
    {
      "epoch": 0.745735819119397,
      "grad_norm": 1.3962752805987728,
      "learning_rate": 9.306243805748266e-05,
      "loss": 0.1332,
      "step": 940
    },
    {
      "epoch": 0.753669178897263,
      "grad_norm": 3.5595641910775173,
      "learning_rate": 9.405351833498514e-05,
      "loss": 0.1507,
      "step": 950
    },
    {
      "epoch": 0.761602538675129,
      "grad_norm": 1.2799346026465042,
      "learning_rate": 9.504459861248762e-05,
      "loss": 0.1706,
      "step": 960
    },
    {
      "epoch": 0.7695358984529949,
      "grad_norm": 0.9866434355012127,
      "learning_rate": 9.60356788899901e-05,
      "loss": 0.1241,
      "step": 970
    },
    {
      "epoch": 0.7774692582308608,
      "grad_norm": 1.2605525685265588,
      "learning_rate": 9.702675916749257e-05,
      "loss": 0.1133,
      "step": 980
    },
    {
      "epoch": 0.7854026180087267,
      "grad_norm": 1.0654588120046724,
      "learning_rate": 9.801783944499505e-05,
      "loss": 0.1066,
      "step": 990
    },
    {
      "epoch": 0.7933359777865926,
      "grad_norm": 0.761881486606698,
      "learning_rate": 9.900891972249753e-05,
      "loss": 0.1015,
      "step": 1000
    },
    {
      "epoch": 0.8012693375644585,
      "grad_norm": 1.0601392633604383,
      "learning_rate": 0.0001,
      "loss": 0.0953,
      "step": 1010
    },
    {
      "epoch": 0.8092026973423245,
      "grad_norm": 0.8224204361157935,
      "learning_rate": 9.999970066101032e-05,
      "loss": 0.0906,
      "step": 1020
    },
    {
      "epoch": 0.8171360571201904,
      "grad_norm": 0.7080480512831684,
      "learning_rate": 9.99988026476254e-05,
      "loss": 0.0866,
      "step": 1030
    },
    {
      "epoch": 0.8250694168980564,
      "grad_norm": 0.7547638843757503,
      "learning_rate": 9.999730597059764e-05,
      "loss": 0.0809,
      "step": 1040
    },
    {
      "epoch": 0.8330027766759223,
      "grad_norm": 1.1588422105514353,
      "learning_rate": 9.999521064784764e-05,
      "loss": 0.0844,
      "step": 1050
    },
    {
      "epoch": 0.8409361364537882,
      "grad_norm": 0.6385523334282596,
      "learning_rate": 9.999251670446386e-05,
      "loss": 0.0804,
      "step": 1060
    },
    {
      "epoch": 0.8488694962316541,
      "grad_norm": 0.5623095327146973,
      "learning_rate": 9.998922417270235e-05,
      "loss": 0.072,
      "step": 1070
    },
    {
      "epoch": 0.85680285600952,
      "grad_norm": 0.43678572122080983,
      "learning_rate": 9.99853330919865e-05,
      "loss": 0.0679,
      "step": 1080
    },
    {
      "epoch": 0.8647362157873859,
      "grad_norm": 0.4972727247312438,
      "learning_rate": 9.998084350890633e-05,
      "loss": 0.065,
      "step": 1090
    },
    {
      "epoch": 0.8726695755652519,
      "grad_norm": 0.5035917631799433,
      "learning_rate": 9.997575547721818e-05,
      "loss": 0.0627,
      "step": 1100
    },
    {
      "epoch": 0.8806029353431178,
      "grad_norm": 0.5182605794595788,
      "learning_rate": 9.997006905784388e-05,
      "loss": 0.0602,
      "step": 1110
    },
    {
      "epoch": 0.8885362951209838,
      "grad_norm": 0.41920699625736624,
      "learning_rate": 9.996378431887012e-05,
      "loss": 0.0579,
      "step": 1120
    },
    {
      "epoch": 0.8964696548988497,
      "grad_norm": 0.414907297365465,
      "learning_rate": 9.995690133554758e-05,
      "loss": 0.0546,
      "step": 1130
    },
    {
      "epoch": 0.9044030146767156,
      "grad_norm": 0.3702005062489374,
      "learning_rate": 9.99494201902901e-05,
      "loss": 0.0523,
      "step": 1140
    },
    {
      "epoch": 0.9123363744545815,
      "grad_norm": 0.42037834689750175,
      "learning_rate": 9.994134097267358e-05,
      "loss": 0.0516,
      "step": 1150
    },
    {
      "epoch": 0.9202697342324474,
      "grad_norm": 0.38928699328314564,
      "learning_rate": 9.993266377943504e-05,
      "loss": 0.0505,
      "step": 1160
    },
    {
      "epoch": 0.9282030940103134,
      "grad_norm": 0.3941895011795836,
      "learning_rate": 9.992338871447138e-05,
      "loss": 0.0495,
      "step": 1170
    },
    {
      "epoch": 0.9361364537881793,
      "grad_norm": 0.37452365492894446,
      "learning_rate": 9.991351588883811e-05,
      "loss": 0.0474,
      "step": 1180
    },
    {
      "epoch": 0.9440698135660452,
      "grad_norm": 0.4395025405712049,
      "learning_rate": 9.99030454207481e-05,
      "loss": 0.0467,
      "step": 1190
    },
    {
      "epoch": 0.9520031733439112,
      "grad_norm": 0.4639716702934286,
      "learning_rate": 9.989197743557017e-05,
      "loss": 0.045,
      "step": 1200
    },
    {
      "epoch": 0.9599365331217771,
      "grad_norm": 0.34908102760028864,
      "learning_rate": 9.988031206582746e-05,
      "loss": 0.0439,
      "step": 1210
    },
    {
      "epoch": 0.967869892899643,
      "grad_norm": 0.29211915294831964,
      "learning_rate": 9.986804945119597e-05,
      "loss": 0.0429,
      "step": 1220
    },
    {
      "epoch": 0.9758032526775089,
      "grad_norm": 0.2801643208729514,
      "learning_rate": 9.985518973850285e-05,
      "loss": 0.041,
      "step": 1230
    },
    {
      "epoch": 0.9837366124553748,
      "grad_norm": 0.3013251709022931,
      "learning_rate": 9.984173308172465e-05,
      "loss": 0.0403,
      "step": 1240
    },
    {
      "epoch": 0.9916699722332408,
      "grad_norm": 0.28830525385588507,
      "learning_rate": 9.982767964198542e-05,
      "loss": 0.04,
      "step": 1250
    },
    {
      "epoch": 0.9996033320111067,
      "grad_norm": 0.2756668810883745,
      "learning_rate": 9.98130295875549e-05,
      "loss": 0.0386,
      "step": 1260
    },
    {
      "epoch": 1.0071400238000794,
      "grad_norm": 0.2673295493921112,
      "learning_rate": 9.979778309384638e-05,
      "loss": 0.0369,
      "step": 1270
    },
    {
      "epoch": 1.0150733835779453,
      "grad_norm": 0.28881709817045276,
      "learning_rate": 9.978194034341462e-05,
      "loss": 0.0366,
      "step": 1280
    },
    {
      "epoch": 1.0230067433558112,
      "grad_norm": 0.2443064331996704,
      "learning_rate": 9.976550152595379e-05,
      "loss": 0.035,
      "step": 1290
    },
    {
      "epoch": 1.0309401031336771,
      "grad_norm": 0.2586793532004466,
      "learning_rate": 9.974846683829501e-05,
      "loss": 0.0351,
      "step": 1300
    },
    {
      "epoch": 1.038873462911543,
      "grad_norm": 0.6237223735436187,
      "learning_rate": 9.973083648440416e-05,
      "loss": 0.0495,
      "step": 1310
    },
    {
      "epoch": 1.046806822689409,
      "grad_norm": 0.982888186735296,
      "learning_rate": 9.971261067537932e-05,
      "loss": 0.0428,
      "step": 1320
    },
    {
      "epoch": 1.0547401824672749,
      "grad_norm": 0.3721310946214082,
      "learning_rate": 9.969378962944829e-05,
      "loss": 0.0396,
      "step": 1330
    },
    {
      "epoch": 1.0626735422451408,
      "grad_norm": 0.3035794519690265,
      "learning_rate": 9.9674373571966e-05,
      "loss": 0.0352,
      "step": 1340
    },
    {
      "epoch": 1.0706069020230067,
      "grad_norm": 0.2564381074742219,
      "learning_rate": 9.965436273541175e-05,
      "loss": 0.034,
      "step": 1350
    },
    {
      "epoch": 1.0785402618008726,
      "grad_norm": 0.2627161314780947,
      "learning_rate": 9.963375735938651e-05,
      "loss": 0.0328,
      "step": 1360
    },
    {
      "epoch": 1.0864736215787385,
      "grad_norm": 0.23965693542315172,
      "learning_rate": 9.961255769060997e-05,
      "loss": 0.0318,
      "step": 1370
    },
    {
      "epoch": 1.0944069813566046,
      "grad_norm": 0.24468200257324066,
      "learning_rate": 9.95907639829176e-05,
      "loss": 0.0312,
      "step": 1380
    },
    {
      "epoch": 1.1023403411344705,
      "grad_norm": 0.2090058523545183,
      "learning_rate": 9.95683764972577e-05,
      "loss": 0.0301,
      "step": 1390
    },
    {
      "epoch": 1.1102737009123365,
      "grad_norm": 0.2583539701995491,
      "learning_rate": 9.954539550168816e-05,
      "loss": 0.0301,
      "step": 1400
    },
    {
      "epoch": 1.1182070606902024,
      "grad_norm": 0.22716588824206302,
      "learning_rate": 9.952182127137327e-05,
      "loss": 0.0293,
      "step": 1410
    },
    {
      "epoch": 1.1261404204680683,
      "grad_norm": 0.24087202301512833,
      "learning_rate": 9.94976540885805e-05,
      "loss": 0.0287,
      "step": 1420
    },
    {
      "epoch": 1.1340737802459342,
      "grad_norm": 0.23388912586015503,
      "learning_rate": 9.947289424267703e-05,
      "loss": 0.0283,
      "step": 1430
    },
    {
      "epoch": 1.1420071400238,
      "grad_norm": 0.21775302377697844,
      "learning_rate": 9.944754203012639e-05,
      "loss": 0.0273,
      "step": 1440
    },
    {
      "epoch": 1.149940499801666,
      "grad_norm": 0.2120232453240847,
      "learning_rate": 9.942159775448478e-05,
      "loss": 0.0273,
      "step": 1450
    },
    {
      "epoch": 1.157873859579532,
      "grad_norm": 0.2055361894837688,
      "learning_rate": 9.939506172639753e-05,
      "loss": 0.027,
      "step": 1460
    },
    {
      "epoch": 1.1658072193573978,
      "grad_norm": 0.22045835966574195,
      "learning_rate": 9.936793426359536e-05,
      "loss": 0.0264,
      "step": 1470
    },
    {
      "epoch": 1.1737405791352638,
      "grad_norm": 0.19936431245788608,
      "learning_rate": 9.934021569089058e-05,
      "loss": 0.0262,
      "step": 1480
    },
    {
      "epoch": 1.1816739389131297,
      "grad_norm": 0.2182877537167163,
      "learning_rate": 9.931190634017314e-05,
      "loss": 0.0256,
      "step": 1490
    },
    {
      "epoch": 1.1896072986909956,
      "grad_norm": 0.20894413268256554,
      "learning_rate": 9.928300655040676e-05,
      "loss": 0.0255,
      "step": 1500
    },
    {
      "epoch": 1.1975406584688615,
      "grad_norm": 0.217040492922683,
      "learning_rate": 9.925351666762479e-05,
      "loss": 0.0248,
      "step": 1510
    },
    {
      "epoch": 1.2054740182467274,
      "grad_norm": 0.2080798426449999,
      "learning_rate": 9.92234370449261e-05,
      "loss": 0.0247,
      "step": 1520
    },
    {
      "epoch": 1.2134073780245935,
      "grad_norm": 0.21422582180243674,
      "learning_rate": 9.919276804247084e-05,
      "loss": 0.0246,
      "step": 1530
    },
    {
      "epoch": 1.2213407378024592,
      "grad_norm": 0.22109698138341372,
      "learning_rate": 9.916151002747614e-05,
      "loss": 0.0242,
      "step": 1540
    },
    {
      "epoch": 1.2292740975803254,
      "grad_norm": 0.2191386707736148,
      "learning_rate": 9.912966337421169e-05,
      "loss": 0.0241,
      "step": 1550
    },
    {
      "epoch": 1.2372074573581913,
      "grad_norm": 0.21285746880929995,
      "learning_rate": 9.909722846399531e-05,
      "loss": 0.0235,
      "step": 1560
    },
    {
      "epoch": 1.2451408171360572,
      "grad_norm": 0.2060608693515892,
      "learning_rate": 9.906420568518832e-05,
      "loss": 0.0235,
      "step": 1570
    },
    {
      "epoch": 1.253074176913923,
      "grad_norm": 0.2165821829643641,
      "learning_rate": 9.903059543319096e-05,
      "loss": 0.0228,
      "step": 1580
    },
    {
      "epoch": 1.261007536691789,
      "grad_norm": 0.20784176983531677,
      "learning_rate": 9.899639811043755e-05,
      "loss": 0.0224,
      "step": 1590
    },
    {
      "epoch": 1.268940896469655,
      "grad_norm": 0.1941327009060369,
      "learning_rate": 9.896161412639178e-05,
      "loss": 0.0224,
      "step": 1600
    },
    {
      "epoch": 1.2768742562475208,
      "grad_norm": 0.19719426217271882,
      "learning_rate": 9.892624389754175e-05,
      "loss": 0.0218,
      "step": 1610
    },
    {
      "epoch": 1.2848076160253867,
      "grad_norm": 0.1987404294569414,
      "learning_rate": 9.889028784739502e-05,
      "loss": 0.0217,
      "step": 1620
    },
    {
      "epoch": 1.2927409758032526,
      "grad_norm": 0.36446508935174654,
      "learning_rate": 9.885374640647349e-05,
      "loss": 0.0218,
      "step": 1630
    },
    {
      "epoch": 1.3006743355811186,
      "grad_norm": 0.2153981685427444,
      "learning_rate": 9.881662001230827e-05,
      "loss": 0.0219,
      "step": 1640
    },
    {
      "epoch": 1.3086076953589845,
      "grad_norm": 0.23124643936791714,
      "learning_rate": 9.877890910943448e-05,
      "loss": 0.0215,
      "step": 1650
    },
    {
      "epoch": 1.3165410551368504,
      "grad_norm": 0.19794561936484395,
      "learning_rate": 9.874061414938583e-05,
      "loss": 0.0209,
      "step": 1660
    },
    {
      "epoch": 1.3244744149147163,
      "grad_norm": 0.1939423361167064,
      "learning_rate": 9.870173559068933e-05,
      "loss": 0.0204,
      "step": 1670
    },
    {
      "epoch": 1.3324077746925824,
      "grad_norm": 0.2157419658468051,
      "learning_rate": 9.86622738988597e-05,
      "loss": 0.02,
      "step": 1680
    },
    {
      "epoch": 1.3403411344704481,
      "grad_norm": 0.19391991216609314,
      "learning_rate": 9.862222954639389e-05,
      "loss": 0.0196,
      "step": 1690
    },
    {
      "epoch": 1.3482744942483142,
      "grad_norm": 0.19977705728669262,
      "learning_rate": 9.858160301276529e-05,
      "loss": 0.0191,
      "step": 1700
    },
    {
      "epoch": 1.3562078540261802,
      "grad_norm": 0.1658808248673058,
      "learning_rate": 9.854039478441817e-05,
      "loss": 0.0194,
      "step": 1710
    },
    {
      "epoch": 1.364141213804046,
      "grad_norm": 0.18655747400847833,
      "learning_rate": 9.849860535476168e-05,
      "loss": 0.0193,
      "step": 1720
    },
    {
      "epoch": 1.372074573581912,
      "grad_norm": 0.18591273213926962,
      "learning_rate": 9.845623522416407e-05,
      "loss": 0.0191,
      "step": 1730
    },
    {
      "epoch": 1.380007933359778,
      "grad_norm": 0.19943417251594622,
      "learning_rate": 9.841328489994659e-05,
      "loss": 0.0189,
      "step": 1740
    },
    {
      "epoch": 1.3879412931376438,
      "grad_norm": 0.20534200797212154,
      "learning_rate": 9.836975489637754e-05,
      "loss": 0.0184,
      "step": 1750
    },
    {
      "epoch": 1.3958746529155097,
      "grad_norm": 0.1911015742718259,
      "learning_rate": 9.832564573466598e-05,
      "loss": 0.0182,
      "step": 1760
    },
    {
      "epoch": 1.4038080126933756,
      "grad_norm": 0.18521458075594055,
      "learning_rate": 9.82809579429556e-05,
      "loss": 0.0178,
      "step": 1770
    },
    {
      "epoch": 1.4117413724712415,
      "grad_norm": 0.19914721587866432,
      "learning_rate": 9.823569205631836e-05,
      "loss": 0.0179,
      "step": 1780
    },
    {
      "epoch": 1.4196747322491075,
      "grad_norm": 0.19845608937674608,
      "learning_rate": 9.818984861674803e-05,
      "loss": 0.0176,
      "step": 1790
    },
    {
      "epoch": 1.4276080920269734,
      "grad_norm": 0.19674825577684443,
      "learning_rate": 9.814342817315374e-05,
      "loss": 0.0177,
      "step": 1800
    }
  ],
  "logging_steps": 10,
  "max_steps": 10088,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "save_steps": 1800,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 129584125181952.0,
  "train_batch_size": 128,
  "trial_name": null,
  "trial_params": null
}