{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.7378939277478862,
  "eval_steps": 500,
  "global_step": 3000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.004919292851652575,
      "grad_norm": 2.650216579437256,
      "learning_rate": 1.557377049180328e-06,
      "loss": 1.9055,
      "step": 20
    },
    {
      "epoch": 0.00983858570330515,
      "grad_norm": 2.133456230163574,
      "learning_rate": 3.1967213114754097e-06,
      "loss": 1.8043,
      "step": 40
    },
    {
      "epoch": 0.014757878554957724,
      "grad_norm": 2.4590532779693604,
      "learning_rate": 4.836065573770492e-06,
      "loss": 1.7683,
      "step": 60
    },
    {
      "epoch": 0.0196771714066103,
      "grad_norm": 1.522373080253601,
      "learning_rate": 6.475409836065574e-06,
      "loss": 1.6977,
      "step": 80
    },
    {
      "epoch": 0.024596464258262875,
      "grad_norm": 1.6893309354782104,
      "learning_rate": 8.114754098360657e-06,
      "loss": 1.6298,
      "step": 100
    },
    {
      "epoch": 0.02951575710991545,
      "grad_norm": 1.5092189311981201,
      "learning_rate": 9.754098360655738e-06,
      "loss": 1.4915,
      "step": 120
    },
    {
      "epoch": 0.03443504996156802,
      "grad_norm": 1.3016860485076904,
      "learning_rate": 1.139344262295082e-05,
      "loss": 1.3592,
      "step": 140
    },
    {
      "epoch": 0.0393543428132206,
      "grad_norm": 1.8532716035842896,
      "learning_rate": 1.3032786885245902e-05,
      "loss": 1.2786,
      "step": 160
    },
    {
      "epoch": 0.044273635664873176,
      "grad_norm": 1.5091300010681152,
      "learning_rate": 1.4672131147540986e-05,
      "loss": 1.1882,
      "step": 180
    },
    {
      "epoch": 0.04919292851652575,
      "grad_norm": 2.3855371475219727,
      "learning_rate": 1.6311475409836068e-05,
      "loss": 1.2969,
      "step": 200
    },
    {
      "epoch": 0.05411222136817832,
      "grad_norm": 1.8920749425888062,
      "learning_rate": 1.7950819672131146e-05,
      "loss": 1.1571,
      "step": 220
    },
    {
      "epoch": 0.0590315142198309,
      "grad_norm": 1.7724186182022095,
      "learning_rate": 1.9590163934426232e-05,
      "loss": 1.1782,
      "step": 240
    },
    {
      "epoch": 0.06395080707148347,
      "grad_norm": 1.53757643699646,
      "learning_rate": 2.122950819672131e-05,
      "loss": 1.1917,
      "step": 260
    },
    {
      "epoch": 0.06887009992313604,
      "grad_norm": 1.773545503616333,
      "learning_rate": 2.2868852459016393e-05,
      "loss": 1.1343,
      "step": 280
    },
    {
      "epoch": 0.07378939277478862,
      "grad_norm": 2.211615800857544,
      "learning_rate": 2.4508196721311478e-05,
      "loss": 1.1242,
      "step": 300
    },
    {
      "epoch": 0.0787086856264412,
      "grad_norm": 1.4129002094268799,
      "learning_rate": 2.614754098360656e-05,
      "loss": 1.0075,
      "step": 320
    },
    {
      "epoch": 0.08362797847809378,
      "grad_norm": 1.8837412595748901,
      "learning_rate": 2.778688524590164e-05,
      "loss": 1.0791,
      "step": 340
    },
    {
      "epoch": 0.08854727132974635,
      "grad_norm": 2.0091936588287354,
      "learning_rate": 2.9426229508196725e-05,
      "loss": 1.0921,
      "step": 360
    },
    {
      "epoch": 0.09346656418139893,
      "grad_norm": 2.447704792022705,
      "learning_rate": 3.106557377049181e-05,
      "loss": 1.0776,
      "step": 380
    },
    {
      "epoch": 0.0983858570330515,
      "grad_norm": 2.063652753829956,
      "learning_rate": 3.2704918032786885e-05,
      "loss": 1.0928,
      "step": 400
    },
    {
      "epoch": 0.10330514988470407,
      "grad_norm": 1.985748529434204,
      "learning_rate": 3.434426229508197e-05,
      "loss": 1.1167,
      "step": 420
    },
    {
      "epoch": 0.10822444273635665,
      "grad_norm": 2.069704532623291,
      "learning_rate": 3.5983606557377056e-05,
      "loss": 1.1432,
      "step": 440
    },
    {
      "epoch": 0.11314373558800922,
      "grad_norm": 1.5993927717208862,
      "learning_rate": 3.762295081967213e-05,
      "loss": 1.0484,
      "step": 460
    },
    {
      "epoch": 0.1180630284396618,
      "grad_norm": 1.8127434253692627,
      "learning_rate": 3.9262295081967214e-05,
      "loss": 1.0379,
      "step": 480
    },
    {
      "epoch": 0.12298232129131437,
      "grad_norm": 1.3972879648208618,
      "learning_rate": 4.09016393442623e-05,
      "loss": 1.0084,
      "step": 500
    },
    {
      "epoch": 0.12790161414296694,
      "grad_norm": 2.6233468055725098,
      "learning_rate": 4.254098360655738e-05,
      "loss": 0.9489,
      "step": 520
    },
    {
      "epoch": 0.13282090699461951,
      "grad_norm": 1.9032576084136963,
      "learning_rate": 4.4180327868852463e-05,
      "loss": 0.9932,
      "step": 540
    },
    {
      "epoch": 0.1377401998462721,
      "grad_norm": 2.0064117908477783,
      "learning_rate": 4.581967213114754e-05,
      "loss": 0.987,
      "step": 560
    },
    {
      "epoch": 0.14265949269792466,
      "grad_norm": 1.6437580585479736,
      "learning_rate": 4.745901639344262e-05,
      "loss": 1.0212,
      "step": 580
    },
    {
      "epoch": 0.14757878554957723,
      "grad_norm": 1.3102104663848877,
      "learning_rate": 4.9098360655737706e-05,
      "loss": 1.0538,
      "step": 600
    },
    {
      "epoch": 0.15249807840122984,
      "grad_norm": 1.6655702590942383,
      "learning_rate": 5.073770491803279e-05,
      "loss": 0.9762,
      "step": 620
    },
    {
      "epoch": 0.1574173712528824,
      "grad_norm": 1.488173007965088,
      "learning_rate": 5.237704918032788e-05,
      "loss": 1.0271,
      "step": 640
    },
    {
      "epoch": 0.16233666410453498,
      "grad_norm": 2.1534035205841064,
      "learning_rate": 5.401639344262295e-05,
      "loss": 1.0124,
      "step": 660
    },
    {
      "epoch": 0.16725595695618756,
      "grad_norm": 2.0473647117614746,
      "learning_rate": 5.5655737704918035e-05,
      "loss": 1.0228,
      "step": 680
    },
    {
      "epoch": 0.17217524980784013,
      "grad_norm": 1.5801570415496826,
      "learning_rate": 5.7295081967213114e-05,
      "loss": 0.9668,
      "step": 700
    },
    {
      "epoch": 0.1770945426594927,
      "grad_norm": 1.749709963798523,
      "learning_rate": 5.89344262295082e-05,
      "loss": 1.0632,
      "step": 720
    },
    {
      "epoch": 0.18201383551114528,
      "grad_norm": 2.072209596633911,
      "learning_rate": 6.0573770491803284e-05,
      "loss": 0.9709,
      "step": 740
    },
    {
      "epoch": 0.18693312836279785,
      "grad_norm": 1.6242225170135498,
      "learning_rate": 6.221311475409836e-05,
      "loss": 1.0601,
      "step": 760
    },
    {
      "epoch": 0.19185242121445042,
      "grad_norm": 1.41958749294281,
      "learning_rate": 6.385245901639346e-05,
      "loss": 0.9846,
      "step": 780
    },
    {
      "epoch": 0.196771714066103,
      "grad_norm": 1.4413362741470337,
      "learning_rate": 6.549180327868852e-05,
      "loss": 1.0331,
      "step": 800
    },
    {
      "epoch": 0.20169100691775557,
      "grad_norm": 1.8630566596984863,
      "learning_rate": 6.713114754098361e-05,
      "loss": 1.0214,
      "step": 820
    },
    {
      "epoch": 0.20661029976940815,
      "grad_norm": 1.689773678779602,
      "learning_rate": 6.877049180327869e-05,
      "loss": 0.9745,
      "step": 840
    },
    {
      "epoch": 0.21152959262106072,
      "grad_norm": 1.9186772108078003,
      "learning_rate": 7.040983606557377e-05,
      "loss": 1.0628,
      "step": 860
    },
    {
      "epoch": 0.2164488854727133,
      "grad_norm": 1.933374285697937,
      "learning_rate": 7.204918032786886e-05,
      "loss": 0.9864,
      "step": 880
    },
    {
      "epoch": 0.22136817832436587,
      "grad_norm": 1.5824335813522339,
      "learning_rate": 7.368852459016394e-05,
      "loss": 1.0016,
      "step": 900
    },
    {
      "epoch": 0.22628747117601844,
      "grad_norm": 1.591445803642273,
      "learning_rate": 7.532786885245902e-05,
      "loss": 1.005,
      "step": 920
    },
    {
      "epoch": 0.231206764027671,
      "grad_norm": 1.7586395740509033,
      "learning_rate": 7.69672131147541e-05,
      "loss": 0.9775,
      "step": 940
    },
    {
      "epoch": 0.2361260568793236,
      "grad_norm": 2.071493148803711,
      "learning_rate": 7.860655737704918e-05,
      "loss": 1.0228,
      "step": 960
    },
    {
      "epoch": 0.24104534973097616,
      "grad_norm": 1.3450851440429688,
      "learning_rate": 8.024590163934427e-05,
      "loss": 1.0164,
      "step": 980
    },
    {
      "epoch": 0.24596464258262873,
      "grad_norm": 1.8442620038986206,
      "learning_rate": 8.188524590163935e-05,
      "loss": 0.9911,
      "step": 1000
    },
    {
      "epoch": 0.25088393543428134,
      "grad_norm": 1.098396897315979,
      "learning_rate": 8.352459016393444e-05,
      "loss": 0.9718,
      "step": 1020
    },
    {
      "epoch": 0.2558032282859339,
      "grad_norm": 1.005915880203247,
      "learning_rate": 8.51639344262295e-05,
      "loss": 0.9436,
      "step": 1040
    },
    {
      "epoch": 0.2607225211375865,
      "grad_norm": 1.113786220550537,
      "learning_rate": 8.68032786885246e-05,
      "loss": 1.0079,
      "step": 1060
    },
    {
      "epoch": 0.26564181398923903,
      "grad_norm": 1.1104156970977783,
      "learning_rate": 8.844262295081968e-05,
      "loss": 0.9534,
      "step": 1080
    },
    {
      "epoch": 0.27056110684089163,
      "grad_norm": 0.9686466455459595,
      "learning_rate": 9.008196721311476e-05,
      "loss": 0.9199,
      "step": 1100
    },
    {
      "epoch": 0.2754803996925442,
      "grad_norm": 1.78913414478302,
      "learning_rate": 9.172131147540985e-05,
      "loss": 0.9132,
      "step": 1120
    },
    {
      "epoch": 0.2803996925441968,
      "grad_norm": 1.5726219415664673,
      "learning_rate": 9.336065573770493e-05,
      "loss": 1.053,
      "step": 1140
    },
    {
      "epoch": 0.2853189853958493,
      "grad_norm": 1.7471468448638916,
      "learning_rate": 9.5e-05,
      "loss": 0.9108,
      "step": 1160
    },
    {
      "epoch": 0.2902382782475019,
      "grad_norm": 1.045501947402954,
      "learning_rate": 9.663934426229508e-05,
      "loss": 1.0033,
      "step": 1180
    },
    {
      "epoch": 0.29515757109915447,
      "grad_norm": 1.3126252889633179,
      "learning_rate": 9.827868852459016e-05,
      "loss": 0.9574,
      "step": 1200
    },
    {
      "epoch": 0.30007686395080707,
      "grad_norm": 1.1474781036376953,
      "learning_rate": 9.991803278688525e-05,
      "loss": 0.9781,
      "step": 1220
    },
    {
      "epoch": 0.3049961568024597,
      "grad_norm": 1.435939073562622,
      "learning_rate": 9.999926090641448e-05,
      "loss": 0.9573,
      "step": 1240
    },
    {
      "epoch": 0.3099154496541122,
      "grad_norm": 2.488921642303467,
      "learning_rate": 9.999688600430955e-05,
      "loss": 1.02,
      "step": 1260
    },
    {
      "epoch": 0.3148347425057648,
      "grad_norm": 1.1515672206878662,
      "learning_rate": 9.999287332412564e-05,
      "loss": 0.9107,
      "step": 1280
    },
    {
      "epoch": 0.31975403535741737,
      "grad_norm": 1.6522116661071777,
      "learning_rate": 9.998722299730854e-05,
      "loss": 0.8688,
      "step": 1300
    },
    {
      "epoch": 0.32467332820906997,
      "grad_norm": 1.14382803440094,
      "learning_rate": 9.997993520894937e-05,
      "loss": 0.9889,
      "step": 1320
    },
    {
      "epoch": 0.3295926210607225,
      "grad_norm": 1.205665111541748,
      "learning_rate": 9.997101019777864e-05,
      "loss": 1.014,
      "step": 1340
    },
    {
      "epoch": 0.3345119139123751,
      "grad_norm": 1.1363279819488525,
      "learning_rate": 9.996044825615826e-05,
      "loss": 0.9774,
      "step": 1360
    },
    {
      "epoch": 0.33943120676402766,
      "grad_norm": 1.490625023841858,
      "learning_rate": 9.994824973007218e-05,
      "loss": 1.0072,
      "step": 1380
    },
    {
      "epoch": 0.34435049961568026,
      "grad_norm": 1.5652107000350952,
      "learning_rate": 9.993441501911479e-05,
      "loss": 1.0104,
      "step": 1400
    },
    {
      "epoch": 0.3492697924673328,
      "grad_norm": 1.1277583837509155,
      "learning_rate": 9.991894457647808e-05,
      "loss": 0.9207,
      "step": 1420
    },
    {
      "epoch": 0.3541890853189854,
      "grad_norm": 1.0140403509140015,
      "learning_rate": 9.990183890893663e-05,
      "loss": 0.9267,
      "step": 1440
    },
    {
      "epoch": 0.35910837817063795,
      "grad_norm": 1.0925248861312866,
      "learning_rate": 9.98830985768311e-05,
      "loss": 0.989,
      "step": 1460
    },
    {
      "epoch": 0.36402767102229056,
      "grad_norm": 1.3068804740905762,
      "learning_rate": 9.986272419404982e-05,
      "loss": 0.9893,
      "step": 1480
    },
    {
      "epoch": 0.3689469638739431,
      "grad_norm": 1.398160457611084,
      "learning_rate": 9.98407164280087e-05,
      "loss": 0.9825,
      "step": 1500
    },
    {
      "epoch": 0.3738662567255957,
      "grad_norm": 1.2048150300979614,
      "learning_rate": 9.981707599962937e-05,
      "loss": 0.916,
      "step": 1520
    },
    {
      "epoch": 0.37878554957724825,
      "grad_norm": 0.9273918271064758,
      "learning_rate": 9.979180368331558e-05,
      "loss": 0.9421,
      "step": 1540
    },
    {
      "epoch": 0.38370484242890085,
      "grad_norm": 1.247428297996521,
      "learning_rate": 9.97649003069278e-05,
      "loss": 0.9482,
      "step": 1560
    },
    {
      "epoch": 0.3886241352805534,
      "grad_norm": 1.952083706855774,
      "learning_rate": 9.973636675175613e-05,
      "loss": 0.9189,
      "step": 1580
    },
    {
      "epoch": 0.393543428132206,
      "grad_norm": 0.979943037033081,
      "learning_rate": 9.970620395249137e-05,
      "loss": 0.9506,
      "step": 1600
    },
    {
      "epoch": 0.39846272098385854,
      "grad_norm": 0.9653918147087097,
      "learning_rate": 9.967441289719452e-05,
      "loss": 0.927,
      "step": 1620
    },
    {
      "epoch": 0.40338201383551114,
      "grad_norm": 1.7574175596237183,
      "learning_rate": 9.96409946272643e-05,
      "loss": 0.9486,
      "step": 1640
    },
    {
      "epoch": 0.40830130668716375,
      "grad_norm": 0.8972555994987488,
      "learning_rate": 9.960595023740307e-05,
      "loss": 0.9739,
      "step": 1660
    },
    {
      "epoch": 0.4132205995388163,
      "grad_norm": 0.8190364837646484,
      "learning_rate": 9.956928087558102e-05,
      "loss": 0.943,
      "step": 1680
    },
    {
      "epoch": 0.4181398923904689,
      "grad_norm": 0.9662739634513855,
      "learning_rate": 9.953098774299847e-05,
      "loss": 0.982,
      "step": 1700
    },
    {
      "epoch": 0.42305918524212144,
      "grad_norm": 0.9831930994987488,
      "learning_rate": 9.949107209404665e-05,
      "loss": 1.0335,
      "step": 1720
    },
    {
      "epoch": 0.42797847809377404,
      "grad_norm": 0.8506195545196533,
      "learning_rate": 9.944953523626643e-05,
      "loss": 0.9193,
      "step": 1740
    },
    {
      "epoch": 0.4328977709454266,
      "grad_norm": 0.9357776641845703,
      "learning_rate": 9.940637853030572e-05,
      "loss": 0.9683,
      "step": 1760
    },
    {
      "epoch": 0.4378170637970792,
      "grad_norm": 1.3345489501953125,
      "learning_rate": 9.936160338987466e-05,
      "loss": 0.956,
      "step": 1780
    },
    {
      "epoch": 0.44273635664873173,
      "grad_norm": 0.8419784307479858,
      "learning_rate": 9.931521128169947e-05,
      "loss": 0.9407,
      "step": 1800
    },
    {
      "epoch": 0.44765564950038433,
      "grad_norm": 0.9009922742843628,
      "learning_rate": 9.926720372547438e-05,
      "loss": 0.9174,
      "step": 1820
    },
    {
      "epoch": 0.4525749423520369,
      "grad_norm": 1.1509584188461304,
      "learning_rate": 9.921758229381177e-05,
      "loss": 0.9644,
      "step": 1840
    },
    {
      "epoch": 0.4574942352036895,
      "grad_norm": 1.1815032958984375,
      "learning_rate": 9.916634861219076e-05,
      "loss": 0.9572,
      "step": 1860
    },
    {
      "epoch": 0.462413528055342,
      "grad_norm": 1.1411418914794922,
      "learning_rate": 9.911350435890386e-05,
      "loss": 0.9424,
      "step": 1880
    },
    {
      "epoch": 0.46733282090699463,
      "grad_norm": 1.2081893682479858,
      "learning_rate": 9.905905126500208e-05,
      "loss": 0.9643,
      "step": 1900
    },
    {
      "epoch": 0.4722521137586472,
      "grad_norm": 1.0384275913238525,
      "learning_rate": 9.900299111423819e-05,
      "loss": 0.9573,
      "step": 1920
    },
    {
      "epoch": 0.4771714066102998,
      "grad_norm": 1.0825401544570923,
      "learning_rate": 9.894532574300827e-05,
      "loss": 0.9533,
      "step": 1940
    },
    {
      "epoch": 0.4820906994619523,
      "grad_norm": 0.9118188619613647,
      "learning_rate": 9.888605704029161e-05,
      "loss": 0.952,
      "step": 1960
    },
    {
      "epoch": 0.4870099923136049,
      "grad_norm": 1.1044224500656128,
      "learning_rate": 9.882518694758875e-05,
      "loss": 0.8827,
      "step": 1980
    },
    {
      "epoch": 0.49192928516525747,
      "grad_norm": 0.9803332090377808,
      "learning_rate": 9.876271745885792e-05,
      "loss": 0.9254,
      "step": 2000
    },
    {
      "epoch": 0.49684857801691007,
      "grad_norm": 1.0061184167861938,
      "learning_rate": 9.869865062044979e-05,
      "loss": 0.9428,
      "step": 2020
    },
    {
      "epoch": 0.5017678708685627,
      "grad_norm": 0.9520462155342102,
      "learning_rate": 9.863298853104032e-05,
      "loss": 0.9292,
      "step": 2040
    },
    {
      "epoch": 0.5066871637202153,
      "grad_norm": 1.5129801034927368,
      "learning_rate": 9.856573334156209e-05,
      "loss": 0.9351,
      "step": 2060
    },
    {
      "epoch": 0.5116064565718678,
      "grad_norm": 1.4102259874343872,
      "learning_rate": 9.84968872551338e-05,
      "loss": 0.908,
      "step": 2080
    },
    {
      "epoch": 0.5165257494235204,
      "grad_norm": 1.0974596738815308,
      "learning_rate": 9.842645252698813e-05,
      "loss": 0.9548,
      "step": 2100
    },
    {
      "epoch": 0.521445042275173,
      "grad_norm": 0.9075796008110046,
      "learning_rate": 9.835443146439786e-05,
      "loss": 0.8862,
      "step": 2120
    },
    {
      "epoch": 0.5263643351268256,
      "grad_norm": 1.2785767316818237,
      "learning_rate": 9.828082642660026e-05,
      "loss": 0.9333,
      "step": 2140
    },
    {
      "epoch": 0.5312836279784781,
      "grad_norm": 1.128395915031433,
      "learning_rate": 9.820563982471987e-05,
      "loss": 0.9688,
      "step": 2160
    },
    {
      "epoch": 0.5362029208301307,
      "grad_norm": 1.3578219413757324,
      "learning_rate": 9.812887412168941e-05,
      "loss": 0.9747,
      "step": 2180
    },
    {
      "epoch": 0.5411222136817833,
      "grad_norm": 1.431168794631958,
      "learning_rate": 9.805053183216923e-05,
      "loss": 0.9358,
      "step": 2200
    },
    {
      "epoch": 0.5460415065334359,
      "grad_norm": 0.8082225918769836,
      "learning_rate": 9.797061552246486e-05,
      "loss": 0.9133,
      "step": 2220
    },
    {
      "epoch": 0.5509607993850884,
      "grad_norm": 1.1131757497787476,
      "learning_rate": 9.788912781044292e-05,
      "loss": 0.9679,
      "step": 2240
    },
    {
      "epoch": 0.555880092236741,
      "grad_norm": 1.163212776184082,
      "learning_rate": 9.780607136544543e-05,
      "loss": 0.9928,
      "step": 2260
    },
    {
      "epoch": 0.5607993850883936,
      "grad_norm": 1.0671954154968262,
      "learning_rate": 9.772144890820234e-05,
      "loss": 0.8596,
      "step": 2280
    },
    {
      "epoch": 0.5657186779400462,
      "grad_norm": 1.1239769458770752,
      "learning_rate": 9.763526321074241e-05,
      "loss": 0.9382,
      "step": 2300
    },
    {
      "epoch": 0.5706379707916986,
      "grad_norm": 1.2082741260528564,
      "learning_rate": 9.754751709630237e-05,
      "loss": 0.9354,
      "step": 2320
    },
    {
      "epoch": 0.5755572636433512,
      "grad_norm": 1.0991088151931763,
      "learning_rate": 9.745821343923451e-05,
      "loss": 0.9695,
      "step": 2340
    },
    {
      "epoch": 0.5804765564950038,
      "grad_norm": 1.2947261333465576,
      "learning_rate": 9.736735516491247e-05,
      "loss": 0.9162,
      "step": 2360
    },
    {
      "epoch": 0.5853958493466564,
      "grad_norm": 1.3943086862564087,
      "learning_rate": 9.727494524963536e-05,
      "loss": 0.9511,
      "step": 2380
    },
    {
      "epoch": 0.5903151421983089,
      "grad_norm": 1.0838426351547241,
      "learning_rate": 9.718098672053044e-05,
      "loss": 0.983,
      "step": 2400
    },
    {
      "epoch": 0.5952344350499615,
      "grad_norm": 1.285778522491455,
      "learning_rate": 9.708548265545375e-05,
      "loss": 0.8961,
      "step": 2420
    },
    {
      "epoch": 0.6001537279016141,
      "grad_norm": 0.9513759016990662,
      "learning_rate": 9.698843618288942e-05,
      "loss": 0.9459,
      "step": 2440
    },
    {
      "epoch": 0.6050730207532667,
      "grad_norm": 0.8295150399208069,
      "learning_rate": 9.688985048184717e-05,
      "loss": 0.8766,
      "step": 2460
    },
    {
      "epoch": 0.6099923136049193,
      "grad_norm": 0.8916402459144592,
      "learning_rate": 9.67897287817581e-05,
      "loss": 0.9322,
      "step": 2480
    },
    {
      "epoch": 0.6149116064565718,
      "grad_norm": 1.0418959856033325,
      "learning_rate": 9.6688074362369e-05,
      "loss": 1.0029,
      "step": 2500
    },
    {
      "epoch": 0.6198308993082244,
      "grad_norm": 1.5775642395019531,
      "learning_rate": 9.658489055363486e-05,
      "loss": 0.9241,
      "step": 2520
    },
    {
      "epoch": 0.624750192159877,
      "grad_norm": 1.4654349088668823,
      "learning_rate": 9.648018073560976e-05,
      "loss": 0.9755,
      "step": 2540
    },
    {
      "epoch": 0.6296694850115296,
      "grad_norm": 1.1109009981155396,
      "learning_rate": 9.637394833833627e-05,
      "loss": 0.9492,
      "step": 2560
    },
    {
      "epoch": 0.6345887778631821,
      "grad_norm": 1.0579159259796143,
      "learning_rate": 9.626619684173288e-05,
      "loss": 0.9327,
      "step": 2580
    },
    {
      "epoch": 0.6395080707148347,
      "grad_norm": 1.344231128692627,
      "learning_rate": 9.615692977548026e-05,
      "loss": 0.9245,
      "step": 2600
    },
    {
      "epoch": 0.6444273635664873,
      "grad_norm": 1.1978901624679565,
      "learning_rate": 9.604615071890541e-05,
      "loss": 0.8989,
      "step": 2620
    },
    {
      "epoch": 0.6493466564181399,
      "grad_norm": 1.2430976629257202,
      "learning_rate": 9.593386330086458e-05,
      "loss": 0.922,
      "step": 2640
    },
    {
      "epoch": 0.6542659492697924,
      "grad_norm": 0.8552722334861755,
      "learning_rate": 9.582007119962424e-05,
      "loss": 0.9013,
      "step": 2660
    },
    {
      "epoch": 0.659185242121445,
      "grad_norm": 0.8925425410270691,
      "learning_rate": 9.570477814274077e-05,
      "loss": 1.0079,
      "step": 2680
    },
    {
      "epoch": 0.6641045349730976,
      "grad_norm": 1.0146712064743042,
      "learning_rate": 9.558798790693817e-05,
      "loss": 0.9355,
      "step": 2700
    },
    {
      "epoch": 0.6690238278247502,
      "grad_norm": 1.0288127660751343,
      "learning_rate": 9.54697043179845e-05,
      "loss": 0.9069,
      "step": 2720
    },
    {
      "epoch": 0.6739431206764027,
      "grad_norm": 1.180992603302002,
      "learning_rate": 9.534993125056643e-05,
      "loss": 0.9002,
      "step": 2740
    },
    {
      "epoch": 0.6788624135280553,
      "grad_norm": 1.1395238637924194,
      "learning_rate": 9.522867262816243e-05,
      "loss": 0.9654,
      "step": 2760
    },
    {
      "epoch": 0.6837817063797079,
      "grad_norm": 1.1949700117111206,
      "learning_rate": 9.510593242291414e-05,
      "loss": 0.8939,
      "step": 2780
    },
    {
      "epoch": 0.6887009992313605,
      "grad_norm": 1.1609656810760498,
      "learning_rate": 9.49817146554963e-05,
      "loss": 0.9344,
      "step": 2800
    },
    {
      "epoch": 0.693620292083013,
      "grad_norm": 1.0648009777069092,
      "learning_rate": 9.485602339498509e-05,
      "loss": 0.969,
      "step": 2820
    },
    {
      "epoch": 0.6985395849346656,
      "grad_norm": 0.7220922112464905,
      "learning_rate": 9.472886275872475e-05,
      "loss": 0.8687,
      "step": 2840
    },
    {
      "epoch": 0.7034588777863182,
      "grad_norm": 1.2389453649520874,
      "learning_rate": 9.460023691219277e-05,
      "loss": 0.8678,
      "step": 2860
    },
    {
      "epoch": 0.7083781706379708,
      "grad_norm": 0.894183874130249,
      "learning_rate": 9.447015006886338e-05,
      "loss": 0.9072,
      "step": 2880
    },
    {
      "epoch": 0.7132974634896234,
      "grad_norm": 1.036689281463623,
      "learning_rate": 9.433860649006961e-05,
      "loss": 0.9594,
      "step": 2900
    },
    {
      "epoch": 0.7182167563412759,
      "grad_norm": 1.0629465579986572,
      "learning_rate": 9.420561048486359e-05,
      "loss": 0.9467,
      "step": 2920
    },
    {
      "epoch": 0.7231360491929285,
      "grad_norm": 1.2931263446807861,
      "learning_rate": 9.40711664098755e-05,
      "loss": 0.9452,
      "step": 2940
    },
    {
      "epoch": 0.7280553420445811,
      "grad_norm": 1.4071450233459473,
      "learning_rate": 9.393527866917082e-05,
      "loss": 0.8886,
      "step": 2960
    },
    {
      "epoch": 0.7329746348962337,
      "grad_norm": 0.9684018492698669,
      "learning_rate": 9.379795171410601e-05,
      "loss": 0.8815,
      "step": 2980
    },
    {
      "epoch": 0.7378939277478862,
      "grad_norm": 1.0788882970809937,
      "learning_rate": 9.36591900431828e-05,
      "loss": 0.9255,
      "step": 3000
    }
  ],
  "logging_steps": 20,
  "max_steps": 12198,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.19351181312e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}