{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 255,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.019801980198019802,
      "grad_norm": 112.43038177490234,
      "learning_rate": 6.25e-07,
      "loss": 1.6969003677368164,
      "step": 1,
      "token_acc": 0.8064516129032258
    },
    {
      "epoch": 0.039603960396039604,
      "grad_norm": 110.48677062988281,
      "learning_rate": 1.25e-06,
      "loss": 1.748899221420288,
      "step": 2,
      "token_acc": 0.7708029197080292
    },
    {
      "epoch": 0.0594059405940594,
      "grad_norm": 115.64219665527344,
      "learning_rate": 1.8750000000000003e-06,
      "loss": 1.8286726474761963,
      "step": 3,
      "token_acc": 0.7410852713178294
    },
    {
      "epoch": 0.07920792079207921,
      "grad_norm": 124.96868896484375,
      "learning_rate": 2.5e-06,
      "loss": 1.7194671630859375,
      "step": 4,
      "token_acc": 0.7702205882352942
    },
    {
      "epoch": 0.09900990099009901,
      "grad_norm": 163.94943237304688,
      "learning_rate": 3.125e-06,
      "loss": 1.1396582126617432,
      "step": 5,
      "token_acc": 0.7951002227171492
    },
    {
      "epoch": 0.1188118811881188,
      "grad_norm": 32.72979736328125,
      "learning_rate": 3.7500000000000005e-06,
      "loss": 0.30865272879600525,
      "step": 6,
      "token_acc": 0.8803571428571428
    },
    {
      "epoch": 0.13861386138613863,
      "grad_norm": 13.428762435913086,
      "learning_rate": 4.3750000000000005e-06,
      "loss": 0.338789701461792,
      "step": 7,
      "token_acc": 0.9037267080745341
    },
    {
      "epoch": 0.15841584158415842,
      "grad_norm": 5.0131683349609375,
      "learning_rate": 5e-06,
      "loss": 0.1692318320274353,
      "step": 8,
      "token_acc": 0.9482758620689655
    },
    {
      "epoch": 0.1782178217821782,
      "grad_norm": 6.603835105895996,
      "learning_rate": 4.999797786569422e-06,
      "loss": 0.034888219088315964,
      "step": 9,
      "token_acc": 0.9873873873873874
    },
    {
      "epoch": 0.19801980198019803,
      "grad_norm": 3.781229019165039,
      "learning_rate": 4.999191178989905e-06,
      "loss": 0.1800294816493988,
      "step": 10,
      "token_acc": 0.9378787878787879
    },
    {
      "epoch": 0.21782178217821782,
      "grad_norm": 5.01358699798584,
      "learning_rate": 4.998180275392809e-06,
      "loss": 0.13575810194015503,
      "step": 11,
      "token_acc": 0.9471947194719472
    },
    {
      "epoch": 0.2376237623762376,
      "grad_norm": 2.0856845378875732,
      "learning_rate": 4.996765239312762e-06,
      "loss": 0.0804593563079834,
      "step": 12,
      "token_acc": 0.9751131221719457
    },
    {
      "epoch": 0.25742574257425743,
      "grad_norm": 1.4063044786453247,
      "learning_rate": 4.994946299661202e-06,
      "loss": 0.05913688614964485,
      "step": 13,
      "token_acc": 0.9794392523364486
    },
    {
      "epoch": 0.27722772277227725,
      "grad_norm": 2.229121685028076,
      "learning_rate": 4.992723750689353e-06,
      "loss": 0.19943833351135254,
      "step": 14,
      "token_acc": 0.9161184210526315
    },
    {
      "epoch": 0.297029702970297,
      "grad_norm": 0.49488377571105957,
      "learning_rate": 4.9900979519406154e-06,
      "loss": 0.000499831628985703,
      "step": 15,
      "token_acc": 1.0
    },
    {
      "epoch": 0.31683168316831684,
      "grad_norm": 2.4041683673858643,
      "learning_rate": 4.9870693281924075e-06,
      "loss": 0.1364714354276657,
      "step": 16,
      "token_acc": 0.9506726457399103
    },
    {
      "epoch": 0.33663366336633666,
      "grad_norm": 2.073594808578491,
      "learning_rate": 4.983638369387449e-06,
      "loss": 0.1324838101863861,
      "step": 17,
      "token_acc": 0.9489247311827957
    },
    {
      "epoch": 0.3564356435643564,
      "grad_norm": 1.244447112083435,
      "learning_rate": 4.9798056305544996e-06,
      "loss": 0.12353045493364334,
      "step": 18,
      "token_acc": 0.9479166666666666
    },
    {
      "epoch": 0.37623762376237624,
      "grad_norm": 1.008650541305542,
      "learning_rate": 4.975571731718572e-06,
      "loss": 0.05529968813061714,
      "step": 19,
      "token_acc": 0.9760956175298805
    },
    {
      "epoch": 0.39603960396039606,
      "grad_norm": 2.1068756580352783,
      "learning_rate": 4.970937357800635e-06,
      "loss": 0.15211760997772217,
      "step": 20,
      "token_acc": 0.9358974358974359
    },
    {
      "epoch": 0.4158415841584158,
      "grad_norm": 6.23456335067749,
      "learning_rate": 4.965903258506806e-06,
      "loss": 0.10873998701572418,
      "step": 21,
      "token_acc": 0.9576988155668359
    },
    {
      "epoch": 0.43564356435643564,
      "grad_norm": 1.3932998180389404,
      "learning_rate": 4.9604702482070765e-06,
      "loss": 0.11840049177408218,
      "step": 22,
      "token_acc": 0.9556025369978859
    },
    {
      "epoch": 0.45544554455445546,
      "grad_norm": 3.151000499725342,
      "learning_rate": 4.954639205803567e-06,
      "loss": 0.001169877708889544,
      "step": 23,
      "token_acc": 1.0
    },
    {
      "epoch": 0.4752475247524752,
      "grad_norm": 1.4540824890136719,
      "learning_rate": 4.9484110745883475e-06,
      "loss": 0.13642601668834686,
      "step": 24,
      "token_acc": 0.9473684210526315
    },
    {
      "epoch": 0.49504950495049505,
      "grad_norm": 2.8562874794006348,
      "learning_rate": 4.941786862090842e-06,
      "loss": 0.13860881328582764,
      "step": 25,
      "token_acc": 0.9502196193265008
    },
    {
      "epoch": 0.5148514851485149,
      "grad_norm": 2.342770576477051,
      "learning_rate": 4.934767639914837e-06,
      "loss": 0.10411743819713593,
      "step": 26,
      "token_acc": 0.9603960396039604
    },
    {
      "epoch": 0.5346534653465347,
      "grad_norm": 1.7055999040603638,
      "learning_rate": 4.927354543565131e-06,
      "loss": 0.17893913388252258,
      "step": 27,
      "token_acc": 0.9277286135693216
    },
    {
      "epoch": 0.5544554455445545,
      "grad_norm": 1.2067103385925293,
      "learning_rate": 4.919548772263837e-06,
      "loss": 0.10788057744503021,
      "step": 28,
      "token_acc": 0.9545454545454546
    },
    {
      "epoch": 0.5742574257425742,
      "grad_norm": 28.483919143676758,
      "learning_rate": 4.911351588756391e-06,
      "loss": 0.08022284507751465,
      "step": 29,
      "token_acc": 0.966131907308378
    },
    {
      "epoch": 0.594059405940594,
      "grad_norm": 1.1291663646697998,
      "learning_rate": 4.902764319107271e-06,
      "loss": 0.14956128597259521,
      "step": 30,
      "token_acc": 0.9387417218543046
    },
    {
      "epoch": 0.6138613861386139,
      "grad_norm": 1.909855842590332,
      "learning_rate": 4.893788352485483e-06,
      "loss": 0.15027566254138947,
      "step": 31,
      "token_acc": 0.9524590163934427
    },
    {
      "epoch": 0.6336633663366337,
      "grad_norm": 0.7861319184303284,
      "learning_rate": 4.884425140939826e-06,
      "loss": 0.060224276036024094,
      "step": 32,
      "token_acc": 0.9787685774946921
    },
    {
      "epoch": 0.6534653465346535,
      "grad_norm": 0.7657933831214905,
      "learning_rate": 4.874676199164005e-06,
      "loss": 0.06414218991994858,
      "step": 33,
      "token_acc": 0.972972972972973
    },
    {
      "epoch": 0.6732673267326733,
      "grad_norm": 1.1537519693374634,
      "learning_rate": 4.864543104251587e-06,
      "loss": 0.20120443403720856,
      "step": 34,
      "token_acc": 0.9267015706806283
    },
    {
      "epoch": 0.693069306930693,
      "grad_norm": 2.486093759536743,
      "learning_rate": 4.854027495440881e-06,
      "loss": 0.14303359389305115,
      "step": 35,
      "token_acc": 0.9379474940334129
    },
    {
      "epoch": 0.7128712871287128,
      "grad_norm": 1.4527450799942017,
      "learning_rate": 4.843131073849753e-06,
      "loss": 0.2470996379852295,
      "step": 36,
      "token_acc": 0.9028925619834711
    },
    {
      "epoch": 0.7326732673267327,
      "grad_norm": 8.935379028320312,
      "learning_rate": 4.831855602200436e-06,
      "loss": 0.07834438234567642,
      "step": 37,
      "token_acc": 0.9705882352941176
    },
    {
      "epoch": 0.7524752475247525,
      "grad_norm": 1.2249703407287598,
      "learning_rate": 4.8202029045343715e-06,
      "loss": 0.1835034340620041,
      "step": 38,
      "token_acc": 0.9294605809128631
    },
    {
      "epoch": 0.7722772277227723,
      "grad_norm": 0.7227316498756409,
      "learning_rate": 4.808174865917136e-06,
      "loss": 0.05235376954078674,
      "step": 39,
      "token_acc": 0.9773156899810964
    },
    {
      "epoch": 0.7920792079207921,
      "grad_norm": 1.185982346534729,
      "learning_rate": 4.795773432133492e-06,
      "loss": 0.1768205463886261,
      "step": 40,
      "token_acc": 0.926865671641791
    },
    {
      "epoch": 0.8118811881188119,
      "grad_norm": 0.0211347583681345,
      "learning_rate": 4.783000609372615e-06,
      "loss": 4.1630286432337016e-05,
      "step": 41,
      "token_acc": 1.0
    },
    {
      "epoch": 0.8316831683168316,
      "grad_norm": 1.0244837999343872,
      "learning_rate": 4.769858463903551e-06,
      "loss": 0.14255551993846893,
      "step": 42,
      "token_acc": 0.9443561208267091
    },
    {
      "epoch": 0.8514851485148515,
      "grad_norm": 1.8504385948181152,
      "learning_rate": 4.756349121740957e-06,
      "loss": 0.13204005360603333,
      "step": 43,
      "token_acc": 0.9509954058192955
    },
    {
      "epoch": 0.8712871287128713,
      "grad_norm": 1.2780109643936157,
      "learning_rate": 4.742474768301172e-06,
      "loss": 0.20053692162036896,
      "step": 44,
      "token_acc": 0.922971114167813
    },
    {
      "epoch": 0.8910891089108911,
      "grad_norm": 0.8156643509864807,
      "learning_rate": 4.72823764804868e-06,
      "loss": 0.09570003300905228,
      "step": 45,
      "token_acc": 0.9638752052545156
    },
    {
      "epoch": 0.9108910891089109,
      "grad_norm": 2.6454577445983887,
      "learning_rate": 4.7136400641330245e-06,
      "loss": 0.056901715695858,
      "step": 46,
      "token_acc": 0.9744525547445255
    },
    {
      "epoch": 0.9306930693069307,
      "grad_norm": 0.534940779209137,
      "learning_rate": 4.698684378016223e-06,
      "loss": 0.04188961535692215,
      "step": 47,
      "token_acc": 0.9814814814814815
    },
    {
      "epoch": 0.9504950495049505,
      "grad_norm": 1.127388834953308,
      "learning_rate": 4.683373009090752e-06,
      "loss": 0.18821795284748077,
      "step": 48,
      "token_acc": 0.9254966887417219
    },
    {
      "epoch": 0.9702970297029703,
      "grad_norm": 0.9354441165924072,
      "learning_rate": 4.66770843428816e-06,
      "loss": 0.12239677459001541,
      "step": 49,
      "token_acc": 0.9536842105263158
    },
    {
      "epoch": 0.9900990099009901,
      "grad_norm": 0.826557457447052,
      "learning_rate": 4.651693187678376e-06,
      "loss": 0.06959648430347443,
      "step": 50,
      "token_acc": 0.9734939759036144
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.266963005065918,
      "learning_rate": 4.6353298600597664e-06,
      "loss": 0.23357868194580078,
      "step": 51,
      "token_acc": 0.9162011173184358
    },
    {
      "epoch": 1.0198019801980198,
      "grad_norm": 1.3645458221435547,
      "learning_rate": 4.618621098540023e-06,
      "loss": 0.08285430073738098,
      "step": 52,
      "token_acc": 0.963855421686747
    },
    {
      "epoch": 1.0396039603960396,
      "grad_norm": 1.2242019176483154,
      "learning_rate": 4.601569606107934e-06,
      "loss": 0.17037668824195862,
      "step": 53,
      "token_acc": 0.9297520661157025
    },
    {
      "epoch": 1.0594059405940595,
      "grad_norm": 1.0588418245315552,
      "learning_rate": 4.5841781411961266e-06,
      "loss": 0.12704727053642273,
      "step": 54,
      "token_acc": 0.9484304932735426
    },
    {
      "epoch": 1.0792079207920793,
      "grad_norm": 0.9267637729644775,
      "learning_rate": 4.566449517234824e-06,
      "loss": 0.14441068470478058,
      "step": 55,
      "token_acc": 0.9452991452991453
    },
    {
      "epoch": 1.099009900990099,
      "grad_norm": 0.9277703166007996,
      "learning_rate": 4.548386602196725e-06,
      "loss": 0.11188974976539612,
      "step": 56,
      "token_acc": 0.9526627218934911
    },
    {
      "epoch": 1.118811881188119,
      "grad_norm": 0.31870558857917786,
      "learning_rate": 4.529992318133041e-06,
      "loss": 0.0002970542118418962,
      "step": 57,
      "token_acc": 1.0
    },
    {
      "epoch": 1.1386138613861387,
      "grad_norm": 0.7992555499076843,
      "learning_rate": 4.511269640700801e-06,
      "loss": 0.14128899574279785,
      "step": 58,
      "token_acc": 0.9463986599664992
    },
    {
      "epoch": 1.1584158415841583,
      "grad_norm": 0.9504163265228271,
      "learning_rate": 4.492221598681469e-06,
      "loss": 0.09462043642997742,
      "step": 59,
      "token_acc": 0.9606299212598425
    },
    {
      "epoch": 1.1782178217821782,
      "grad_norm": 0.0006429307395592332,
      "learning_rate": 4.472851273490985e-06,
      "loss": 3.4975448670593323e-06,
      "step": 60,
      "token_acc": 1.0
    },
    {
      "epoch": 1.198019801980198,
      "grad_norm": 0.9329037666320801,
      "learning_rate": 4.453161798681273e-06,
      "loss": 0.15014111995697021,
      "step": 61,
      "token_acc": 0.9425087108013938
    },
    {
      "epoch": 1.2178217821782178,
      "grad_norm": 1.0982563495635986,
      "learning_rate": 4.433156359433334e-06,
      "loss": 0.17269109189510345,
      "step": 62,
      "token_acc": 0.9347496206373292
    },
    {
      "epoch": 1.2376237623762376,
      "grad_norm": 0.6611605286598206,
      "learning_rate": 4.412838192041968e-06,
      "loss": 0.05397292971611023,
      "step": 63,
      "token_acc": 0.977859778597786
    },
    {
      "epoch": 1.2574257425742574,
      "grad_norm": 0.902929961681366,
      "learning_rate": 4.392210583392238e-06,
      "loss": 0.1418980062007904,
      "step": 64,
      "token_acc": 0.9414715719063546
    },
    {
      "epoch": 1.2772277227722773,
      "grad_norm": 1.0193787813186646,
      "learning_rate": 4.3712768704277535e-06,
      "loss": 0.13208191096782684,
      "step": 65,
      "token_acc": 0.9506531204644412
    },
    {
      "epoch": 1.297029702970297,
      "grad_norm": 2.592942476272583,
      "learning_rate": 4.350040439610844e-06,
      "loss": 0.2708500921726227,
      "step": 66,
      "token_acc": 0.9067103109656302
    },
    {
      "epoch": 1.316831683168317,
      "grad_norm": 0.6828377842903137,
      "learning_rate": 4.328504726374733e-06,
      "loss": 0.06117222458124161,
      "step": 67,
      "token_acc": 0.974155069582505
    },
    {
      "epoch": 1.3366336633663367,
      "grad_norm": 1.250339150428772,
      "learning_rate": 4.306673214567783e-06,
      "loss": 0.15187141299247742,
      "step": 68,
      "token_acc": 0.9449225473321858
    },
    {
      "epoch": 1.3564356435643563,
      "grad_norm": 0.7941989302635193,
      "learning_rate": 4.284549435889914e-06,
      "loss": 0.09086102992296219,
      "step": 69,
      "token_acc": 0.9666666666666667
    },
    {
      "epoch": 1.3762376237623761,
      "grad_norm": 0.7678201794624329,
      "learning_rate": 4.262136969321269e-06,
      "loss": 0.045609161257743835,
      "step": 70,
      "token_acc": 0.9830220713073005
    },
    {
      "epoch": 1.396039603960396,
      "grad_norm": 1.5242784023284912,
      "learning_rate": 4.2394394405432526e-06,
      "loss": 0.15937024354934692,
      "step": 71,
      "token_acc": 0.9418181818181818
    },
    {
      "epoch": 1.4158415841584158,
      "grad_norm": 0.5641005039215088,
      "learning_rate": 4.216460521351993e-06,
      "loss": 0.04410265386104584,
      "step": 72,
      "token_acc": 0.9817905918057663
    },
    {
      "epoch": 1.4356435643564356,
      "grad_norm": 2.2610104084014893,
      "learning_rate": 4.1932039290643534e-06,
      "loss": 0.16153395175933838,
      "step": 73,
      "token_acc": 0.9415807560137457
    },
    {
      "epoch": 1.4554455445544554,
      "grad_norm": 1.0418875217437744,
      "learning_rate": 4.169673425916582e-06,
      "loss": 0.17528606951236725,
      "step": 74,
      "token_acc": 0.9372990353697749
    },
    {
      "epoch": 1.4752475247524752,
      "grad_norm": 1.041953682899475,
      "learning_rate": 4.145872818455691e-06,
      "loss": 0.09952806681394577,
      "step": 75,
      "token_acc": 0.9672413793103448
    },
    {
      "epoch": 1.495049504950495,
      "grad_norm": 3.4738848209381104,
      "learning_rate": 4.121805956923668e-06,
      "loss": 0.18613466620445251,
      "step": 76,
      "token_acc": 0.9263803680981595
    },
    {
      "epoch": 1.5148514851485149,
      "grad_norm": 0.7892954349517822,
      "learning_rate": 4.09747673463462e-06,
      "loss": 0.09873586893081665,
      "step": 77,
      "token_acc": 0.9636048526863085
    },
    {
      "epoch": 1.5346534653465347,
      "grad_norm": 0.5464914441108704,
      "learning_rate": 4.0728890873449486e-06,
      "loss": 0.05909133702516556,
      "step": 78,
      "token_acc": 0.9772727272727273
    },
    {
      "epoch": 1.5544554455445545,
      "grad_norm": 1.515067219734192,
      "learning_rate": 4.048046992616661e-06,
      "loss": 0.19638079404830933,
      "step": 79,
      "token_acc": 0.927536231884058
    },
    {
      "epoch": 1.5742574257425743,
      "grad_norm": 0.9421702027320862,
      "learning_rate": 4.022954469173914e-06,
      "loss": 0.14969073235988617,
      "step": 80,
      "token_acc": 0.9399656946826758
    },
    {
      "epoch": 1.5940594059405941,
      "grad_norm": 1.6792571544647217,
      "learning_rate": 3.997615576252907e-06,
      "loss": 0.16854381561279297,
      "step": 81,
      "token_acc": 0.9386973180076629
    },
    {
      "epoch": 1.613861386138614,
      "grad_norm": 0.0332539901137352,
      "learning_rate": 3.972034412945212e-06,
      "loss": 0.0001334004628006369,
      "step": 82,
      "token_acc": 1.0
    },
    {
      "epoch": 1.6336633663366338,
      "grad_norm": 0.9271730780601501,
      "learning_rate": 3.946215117534658e-06,
      "loss": 0.16093292832374573,
      "step": 83,
      "token_acc": 0.9369202226345084
    },
    {
      "epoch": 1.6534653465346536,
      "grad_norm": 0.9867342710494995,
      "learning_rate": 3.92016186682789e-06,
      "loss": 0.07043827325105667,
      "step": 84,
      "token_acc": 0.9734939759036144
    },
    {
      "epoch": 1.6732673267326734,
      "grad_norm": 0.8659154772758484,
      "learning_rate": 3.8938788754786675e-06,
      "loss": 0.1386394500732422,
      "step": 85,
      "token_acc": 0.9528455284552846
    },
    {
      "epoch": 1.693069306930693,
      "grad_norm": 3.0848546028137207,
      "learning_rate": 3.8673703953060685e-06,
      "loss": 0.1357145607471466,
      "step": 86,
      "token_acc": 0.9578947368421052
    },
    {
      "epoch": 1.7128712871287128,
      "grad_norm": 0.6138753890991211,
      "learning_rate": 3.840640714606664e-06,
      "loss": 0.06022759899497032,
      "step": 87,
      "token_acc": 0.9752252252252253
    },
    {
      "epoch": 1.7326732673267327,
      "grad_norm": 1.006622552871704,
      "learning_rate": 3.813694157460801e-06,
      "loss": 0.15556997060775757,
      "step": 88,
      "token_acc": 0.9411764705882353
    },
    {
      "epoch": 1.7524752475247525,
      "grad_norm": 0.7253024578094482,
      "learning_rate": 3.7865350830330897e-06,
      "loss": 0.138234943151474,
      "step": 89,
      "token_acc": 0.9427662957074722
    },
    {
      "epoch": 1.7722772277227723,
      "grad_norm": 0.0006839027046225965,
      "learning_rate": 3.7591678848672198e-06,
      "loss": 3.070681486860849e-06,
      "step": 90,
      "token_acc": 1.0
    },
    {
      "epoch": 1.7920792079207921,
      "grad_norm": 0.9093657732009888,
      "learning_rate": 3.7315969901752107e-06,
      "loss": 0.12562689185142517,
      "step": 91,
      "token_acc": 0.9551569506726457
    },
    {
      "epoch": 1.811881188118812,
      "grad_norm": 1.1722992658615112,
      "learning_rate": 3.7038268591212245e-06,
      "loss": 0.2242039144039154,
      "step": 92,
      "token_acc": 0.9171483622350675
    },
    {
      "epoch": 1.8316831683168315,
      "grad_norm": 0.5971883535385132,
      "learning_rate": 3.6758619841000337e-06,
      "loss": 0.06140849366784096,
      "step": 93,
      "token_acc": 0.9755555555555555
    },
    {
      "epoch": 1.8514851485148514,
      "grad_norm": 0.3268670439720154,
      "learning_rate": 3.6477068890102895e-06,
      "loss": 0.045806486159563065,
      "step": 94,
      "token_acc": 0.9841017488076311
    },
    {
      "epoch": 1.8712871287128712,
      "grad_norm": 1.1285595893859863,
      "learning_rate": 3.6193661285226862e-06,
      "loss": 0.033183883875608444,
      "step": 95,
      "token_acc": 0.9858657243816255
    },
    {
      "epoch": 1.891089108910891,
      "grad_norm": 0.5610284209251404,
      "learning_rate": 3.590844287343146e-06,
      "loss": 0.05272660031914711,
      "step": 96,
      "token_acc": 0.9779559118236473
    },
    {
      "epoch": 1.9108910891089108,
      "grad_norm": 1.3507022857666016,
      "learning_rate": 3.5621459794711487e-06,
      "loss": 0.19275179505348206,
      "step": 97,
      "token_acc": 0.9241935483870968
    },
    {
      "epoch": 1.9306930693069306,
      "grad_norm": 0.5547268986701965,
      "learning_rate": 3.5332758474533255e-06,
      "loss": 0.052035968750715256,
      "step": 98,
      "token_acc": 0.9798165137614679
    },
    {
      "epoch": 1.9504950495049505,
      "grad_norm": 1.2770878076553345,
      "learning_rate": 3.5042385616324243e-06,
      "loss": 0.22472034394741058,
      "step": 99,
      "token_acc": 0.9131274131274131
    },
    {
      "epoch": 1.9702970297029703,
      "grad_norm": 1.1276659965515137,
      "learning_rate": 3.4750388193917892e-06,
      "loss": 0.1283009797334671,
      "step": 100,
      "token_acc": 0.9481792717086834
    },
    {
      "epoch": 1.99009900990099,
      "grad_norm": 1.4426108598709106,
      "learning_rate": 3.4456813443954623e-06,
      "loss": 0.20597082376480103,
      "step": 101,
      "token_acc": 0.9205673758865248
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.0010104256216436625,
      "learning_rate": 3.4161708858240285e-06,
      "loss": 3.949928213842213e-06,
      "step": 102,
      "token_acc": 1.0
    },
    {
      "epoch": 2.01980198019802,
      "grad_norm": 1.1984245777130127,
      "learning_rate": 3.386512217606339e-06,
      "loss": 0.16492575407028198,
      "step": 103,
      "token_acc": 0.9415204678362573
    },
    {
      "epoch": 2.0396039603960396,
      "grad_norm": 0.8825148344039917,
      "learning_rate": 3.3567101376472333e-06,
      "loss": 0.13931022584438324,
      "step": 104,
      "token_acc": 0.9545454545454546
    },
    {
      "epoch": 2.0594059405940595,
      "grad_norm": 0.8195608854293823,
      "learning_rate": 3.3267694670513717e-06,
      "loss": 0.09142282605171204,
      "step": 105,
      "token_acc": 0.9675767918088737
    },
    {
      "epoch": 2.0792079207920793,
      "grad_norm": 0.4743385910987854,
      "learning_rate": 3.2966950493433264e-06,
      "loss": 0.03967257961630821,
      "step": 106,
      "token_acc": 0.9853420195439739
    },
    {
      "epoch": 2.099009900990099,
      "grad_norm": 0.39722344279289246,
      "learning_rate": 3.266491749684039e-06,
      "loss": 0.04726823419332504,
      "step": 107,
      "token_acc": 0.9801324503311258
    },
    {
      "epoch": 2.118811881188119,
      "grad_norm": 0.8539097309112549,
      "learning_rate": 3.236164454083781e-06,
      "loss": 0.1401786506175995,
      "step": 108,
      "token_acc": 0.9473684210526315
    },
    {
      "epoch": 2.1386138613861387,
      "grad_norm": 0.9900526404380798,
      "learning_rate": 3.2057180686117394e-06,
      "loss": 0.16982492804527283,
      "step": 109,
      "token_acc": 0.9400510204081632
    },
    {
      "epoch": 2.1584158415841586,
      "grad_norm": 0.7716503143310547,
      "learning_rate": 3.175157518602357e-06,
      "loss": 0.05942318215966225,
      "step": 110,
      "token_acc": 0.9772296015180265
    },
    {
      "epoch": 2.1782178217821784,
      "grad_norm": 0.5397495627403259,
      "learning_rate": 3.144487747858562e-06,
      "loss": 0.08918597549200058,
      "step": 111,
      "token_acc": 0.966996699669967
    },
    {
      "epoch": 2.198019801980198,
      "grad_norm": 1.0641586780548096,
      "learning_rate": 3.1137137178519983e-06,
      "loss": 0.1809808313846588,
      "step": 112,
      "token_acc": 0.933184855233853
    },
    {
      "epoch": 2.217821782178218,
      "grad_norm": 0.9593180418014526,
      "learning_rate": 3.082840406920411e-06,
      "loss": 0.1645204722881317,
      "step": 113,
      "token_acc": 0.9459459459459459
    },
    {
      "epoch": 2.237623762376238,
      "grad_norm": 0.7205140590667725,
      "learning_rate": 3.0518728094622935e-06,
      "loss": 0.05737221613526344,
      "step": 114,
      "token_acc": 0.9779411764705882
    },
    {
      "epoch": 2.2574257425742577,
      "grad_norm": 1.0832691192626953,
      "learning_rate": 3.020815935128941e-06,
      "loss": 0.23320218920707703,
      "step": 115,
      "token_acc": 0.9105960264900662
    },
    {
      "epoch": 2.2772277227722775,
      "grad_norm": 0.514970600605011,
      "learning_rate": 2.9896748080140337e-06,
      "loss": 0.05509328097105026,
      "step": 116,
      "token_acc": 0.9801192842942346
    },
    {
      "epoch": 2.297029702970297,
      "grad_norm": 0.7392024397850037,
      "learning_rate": 2.9584544658408907e-06,
      "loss": 0.05270788446068764,
      "step": 117,
      "token_acc": 0.9819277108433735
    },
    {
      "epoch": 2.3168316831683167,
      "grad_norm": 0.6466699838638306,
      "learning_rate": 2.927159959147506e-06,
      "loss": 0.05460197851061821,
      "step": 118,
      "token_acc": 0.9787644787644788
    },
    {
      "epoch": 2.3366336633663365,
      "grad_norm": 0.8660450577735901,
      "learning_rate": 2.895796350469525e-06,
      "loss": 0.10432581603527069,
      "step": 119,
      "token_acc": 0.9673076923076923
    },
    {
      "epoch": 2.3564356435643563,
      "grad_norm": 0.8661943078041077,
      "learning_rate": 2.8643687135212738e-06,
      "loss": 0.11461281031370163,
      "step": 120,
      "token_acc": 0.960167714884696
    },
    {
      "epoch": 2.376237623762376,
      "grad_norm": 0.6902587413787842,
      "learning_rate": 2.8328821323749772e-06,
      "loss": 0.08308986574411392,
      "step": 121,
      "token_acc": 0.9657320872274143
    },
    {
      "epoch": 2.396039603960396,
      "grad_norm": 0.0015966725768521428,
      "learning_rate": 2.8013417006383078e-06,
      "loss": 5.366915047488874e-06,
      "step": 122,
      "token_acc": 1.0
    },
    {
      "epoch": 2.4158415841584158,
      "grad_norm": 1.0549192428588867,
      "learning_rate": 2.7697525206303893e-06,
      "loss": 0.1817295253276825,
      "step": 123,
      "token_acc": 0.9313572542901716
    },
    {
      "epoch": 2.4356435643564356,
      "grad_norm": 0.8230319023132324,
      "learning_rate": 2.7381197025563873e-06,
      "loss": 0.12129511684179306,
      "step": 124,
      "token_acc": 0.9602941176470589
    },
    {
      "epoch": 2.4554455445544554,
      "grad_norm": 1.1076909303665161,
      "learning_rate": 2.7064483636808314e-06,
      "loss": 0.1017247661948204,
      "step": 125,
      "token_acc": 0.9617486338797814
    },
    {
      "epoch": 2.4752475247524752,
      "grad_norm": 0.8721786141395569,
      "learning_rate": 2.6747436274997896e-06,
      "loss": 0.06177304685115814,
      "step": 126,
      "token_acc": 0.9773755656108597
    },
    {
      "epoch": 2.495049504950495,
      "grad_norm": 1.9826033115386963,
      "learning_rate": 2.6430106229120374e-06,
      "loss": 0.22127825021743774,
      "step": 127,
      "token_acc": 0.9185905224787363
    },
    {
      "epoch": 2.514851485148515,
      "grad_norm": 0.0003355354128871113,
      "learning_rate": 2.611254483389351e-06,
      "loss": 1.940394440680393e-06,
      "step": 128,
      "token_acc": 1.0
    },
    {
      "epoch": 2.5346534653465347,
      "grad_norm": 1.5624958276748657,
      "learning_rate": 2.579480346146061e-06,
      "loss": 0.22154107689857483,
      "step": 129,
      "token_acc": 0.9141104294478528
    },
    {
      "epoch": 2.5544554455445545,
      "grad_norm": 0.9364230036735535,
      "learning_rate": 2.547693351308006e-06,
      "loss": 0.05561017245054245,
      "step": 130,
      "token_acc": 0.978515625
    },
    {
      "epoch": 2.5742574257425743,
      "grad_norm": 1.906137466430664,
      "learning_rate": 2.5158986410810034e-06,
      "loss": 0.21886566281318665,
      "step": 131,
      "token_acc": 0.9167927382753404
    },
    {
      "epoch": 2.594059405940594,
      "grad_norm": 1.0787904262542725,
      "learning_rate": 2.484101358918998e-06,
      "loss": 0.1073170006275177,
      "step": 132,
      "token_acc": 0.9607072691552063
    },
    {
      "epoch": 2.613861386138614,
      "grad_norm": 0.7391598224639893,
      "learning_rate": 2.4523066486919948e-06,
      "loss": 0.04957495257258415,
      "step": 133,
      "token_acc": 0.982174688057041
    },
    {
      "epoch": 2.633663366336634,
      "grad_norm": 0.0008166991174221039,
      "learning_rate": 2.420519653853939e-06,
      "loss": 4.6596414904342964e-06,
      "step": 134,
      "token_acc": 1.0
    },
    {
      "epoch": 2.6534653465346536,
      "grad_norm": 0.7014015316963196,
      "learning_rate": 2.3887455166106502e-06,
      "loss": 0.05042169243097305,
      "step": 135,
      "token_acc": 0.9835526315789473
    },
    {
      "epoch": 2.6732673267326734,
      "grad_norm": 1.1513527631759644,
      "learning_rate": 2.356989377087963e-06,
      "loss": 0.13228660821914673,
      "step": 136,
      "token_acc": 0.9497568881685575
    },
    {
      "epoch": 2.693069306930693,
      "grad_norm": 0.9080381989479065,
      "learning_rate": 2.3252563725002108e-06,
      "loss": 0.10027599334716797,
      "step": 137,
      "token_acc": 0.9586919104991394
    },
    {
      "epoch": 2.7128712871287126,
      "grad_norm": 0.9012190103530884,
      "learning_rate": 2.2935516363191695e-06,
      "loss": 0.12134231626987457,
      "step": 138,
      "token_acc": 0.9572271386430679
    },
    {
      "epoch": 2.7326732673267324,
      "grad_norm": 0.0009042822057381272,
      "learning_rate": 2.261880297443614e-06,
      "loss": 5.1436677495075855e-06,
      "step": 139,
      "token_acc": 1.0
    },
    {
      "epoch": 2.7524752475247523,
      "grad_norm": 1.1476218700408936,
      "learning_rate": 2.2302474793696115e-06,
      "loss": 0.09037760645151138,
      "step": 140,
      "token_acc": 0.9656946826758147
    },
    {
      "epoch": 2.772277227722772,
      "grad_norm": 1.0903990268707275,
      "learning_rate": 2.1986582993616926e-06,
      "loss": 0.0636323094367981,
      "step": 141,
      "token_acc": 0.9781659388646288
    },
    {
      "epoch": 2.792079207920792,
      "grad_norm": 1.4802842140197754,
      "learning_rate": 2.1671178676250236e-06,
      "loss": 0.20774291455745697,
      "step": 142,
      "token_acc": 0.9215328467153284
    },
    {
      "epoch": 2.8118811881188117,
      "grad_norm": 0.679743230342865,
      "learning_rate": 2.1356312864787266e-06,
      "loss": 0.053458839654922485,
      "step": 143,
      "token_acc": 0.9785575048732943
    },
    {
      "epoch": 2.8316831683168315,
      "grad_norm": 0.7652466893196106,
      "learning_rate": 2.1042036495304753e-06,
      "loss": 0.08847317099571228,
      "step": 144,
      "token_acc": 0.9720394736842105
    },
    {
      "epoch": 2.8514851485148514,
      "grad_norm": 1.6989758014678955,
      "learning_rate": 2.0728400408524945e-06,
      "loss": 0.20837374031543732,
      "step": 145,
      "token_acc": 0.9178832116788321
    },
    {
      "epoch": 2.871287128712871,
      "grad_norm": 1.2557241916656494,
      "learning_rate": 2.04154553415911e-06,
      "loss": 0.10232782363891602,
      "step": 146,
      "token_acc": 0.955595026642984
    },
    {
      "epoch": 2.891089108910891,
      "grad_norm": 0.8444433808326721,
      "learning_rate": 2.0103251919859667e-06,
      "loss": 0.09674122929573059,
      "step": 147,
      "token_acc": 0.9681978798586572
    },
    {
      "epoch": 2.910891089108911,
      "grad_norm": 1.1800702810287476,
      "learning_rate": 1.9791840648710597e-06,
      "loss": 0.11082691699266434,
      "step": 148,
      "token_acc": 0.9579349904397706
    },
    {
      "epoch": 2.9306930693069306,
      "grad_norm": 1.2196922302246094,
      "learning_rate": 1.948127190537708e-06,
      "loss": 0.13585644960403442,
      "step": 149,
      "token_acc": 0.9494290375203915
    },
    {
      "epoch": 2.9504950495049505,
      "grad_norm": 0.8979001045227051,
      "learning_rate": 1.9171595930795897e-06,
      "loss": 0.057459376752376556,
      "step": 150,
      "token_acc": 0.974609375
    },
    {
      "epoch": 2.9702970297029703,
      "grad_norm": 0.7514238953590393,
      "learning_rate": 1.8862862821480023e-06,
      "loss": 0.05544315651059151,
      "step": 151,
      "token_acc": 0.9801192842942346
    },
    {
      "epoch": 2.99009900990099,
      "grad_norm": 1.8518508672714233,
      "learning_rate": 1.8555122521414393e-06,
      "loss": 0.2566053867340088,
      "step": 152,
      "token_acc": 0.900459418070444
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.371117115020752,
      "learning_rate": 1.8248424813976438e-06,
      "loss": 0.31392282247543335,
      "step": 153,
      "token_acc": 0.8895184135977338
    },
    {
      "epoch": 3.01980198019802,
      "grad_norm": 0.000473017746116966,
      "learning_rate": 1.7942819313882614e-06,
      "loss": 2.869689296858269e-06,
      "step": 154,
      "token_acc": 1.0
    },
    {
      "epoch": 3.0396039603960396,
      "grad_norm": 1.3725121021270752,
      "learning_rate": 1.76383554591622e-06,
      "loss": 0.07852256298065186,
      "step": 155,
      "token_acc": 0.972972972972973
    },
    {
      "epoch": 3.0594059405940595,
      "grad_norm": 1.0758554935455322,
      "learning_rate": 1.7335082503159618e-06,
      "loss": 0.1400071084499359,
      "step": 156,
      "token_acc": 0.9437908496732026
    },
    {
      "epoch": 3.0792079207920793,
      "grad_norm": 1.1335726976394653,
      "learning_rate": 1.7033049506566746e-06,
      "loss": 0.11961829662322998,
      "step": 157,
      "token_acc": 0.9532710280373832
    },
    {
      "epoch": 3.099009900990099,
      "grad_norm": 1.7540920972824097,
      "learning_rate": 1.6732305329486292e-06,
      "loss": 0.20202551782131195,
      "step": 158,
      "token_acc": 0.9227799227799228
    },
    {
      "epoch": 3.118811881188119,
      "grad_norm": 1.7364046573638916,
      "learning_rate": 1.6432898623527677e-06,
      "loss": 0.17151674628257751,
      "step": 159,
      "token_acc": 0.938
    },
    {
      "epoch": 3.1386138613861387,
      "grad_norm": 1.020960807800293,
      "learning_rate": 1.613487782393661e-06,
      "loss": 0.08284834027290344,
      "step": 160,
      "token_acc": 0.9679878048780488
    },
    {
      "epoch": 3.1584158415841586,
      "grad_norm": 1.1220277547836304,
      "learning_rate": 1.5838291141759727e-06,
      "loss": 0.10776171088218689,
      "step": 161,
      "token_acc": 0.957089552238806
    },
    {
      "epoch": 3.1782178217821784,
      "grad_norm": 1.8783152103424072,
      "learning_rate": 1.5543186556045381e-06,
      "loss": 0.0788167268037796,
      "step": 162,
      "token_acc": 0.9818511796733213
    },
    {
      "epoch": 3.198019801980198,
      "grad_norm": 2.0448639392852783,
      "learning_rate": 1.524961180608211e-06,
      "loss": 0.15025168657302856,
      "step": 163,
      "token_acc": 0.9495934959349593
    },
    {
      "epoch": 3.217821782178218,
      "grad_norm": 1.1998469829559326,
      "learning_rate": 1.495761438367577e-06,
      "loss": 0.047870174050331116,
      "step": 164,
      "token_acc": 0.9811676082862524
    },
    {
      "epoch": 3.237623762376238,
      "grad_norm": 2.1205570697784424,
      "learning_rate": 1.4667241525466753e-06,
      "loss": 0.058612160384655,
      "step": 165,
      "token_acc": 0.9783132530120482
    },
    {
      "epoch": 3.2574257425742577,
      "grad_norm": 3.9921395778656006,
      "learning_rate": 1.4378540205288521e-06,
      "loss": 0.08336001634597778,
      "step": 166,
      "token_acc": 0.9720767888307156
    },
    {
      "epoch": 3.2772277227722775,
      "grad_norm": 3.234689712524414,
      "learning_rate": 1.409155712656855e-06,
      "loss": 0.12804031372070312,
      "step": 167,
      "token_acc": 0.9582089552238806
    },
    {
      "epoch": 3.297029702970297,
      "grad_norm": 3.429499626159668,
      "learning_rate": 1.3806338714773144e-06,
      "loss": 0.11839525401592255,
      "step": 168,
      "token_acc": 0.9514866979655712
    },
    {
      "epoch": 3.3168316831683167,
      "grad_norm": 4.234612464904785,
      "learning_rate": 1.352293110989711e-06,
      "loss": 0.1435491144657135,
      "step": 169,
      "token_acc": 0.9514563106796117
    },
    {
      "epoch": 3.3366336633663365,
      "grad_norm": 4.562895774841309,
      "learning_rate": 1.3241380158999669e-06,
      "loss": 0.14821024239063263,
      "step": 170,
      "token_acc": 0.9426048565121413
    },
    {
      "epoch": 3.3564356435643563,
      "grad_norm": 0.9261006116867065,
      "learning_rate": 1.296173140878776e-06,
      "loss": 0.04509488865733147,
      "step": 171,
      "token_acc": 0.9819168173598554
    },
    {
      "epoch": 3.376237623762376,
      "grad_norm": 2.607170581817627,
      "learning_rate": 1.2684030098247892e-06,
      "loss": 0.16471607983112335,
      "step": 172,
      "token_acc": 0.947107438016529
    },
    {
      "epoch": 3.396039603960396,
      "grad_norm": 3.300290822982788,
      "learning_rate": 1.2408321151327817e-06,
      "loss": 0.13392293453216553,
      "step": 173,
      "token_acc": 0.9528455284552846
    },
    {
      "epoch": 3.4158415841584158,
      "grad_norm": 1.8824461698532104,
      "learning_rate": 1.2134649169669102e-06,
      "loss": 0.16140276193618774,
      "step": 174,
      "token_acc": 0.946524064171123
    },
    {
      "epoch": 3.4356435643564356,
      "grad_norm": 1.5867682695388794,
      "learning_rate": 1.1863058425391994e-06,
      "loss": 0.08794964849948883,
      "step": 175,
      "token_acc": 0.970856102003643
    },
    {
      "epoch": 3.4554455445544554,
      "grad_norm": 2.3255279064178467,
      "learning_rate": 1.1593592853933374e-06,
      "loss": 0.06007307022809982,
      "step": 176,
      "token_acc": 0.9805825242718447
    },
    {
      "epoch": 3.4752475247524752,
      "grad_norm": 4.138920783996582,
      "learning_rate": 1.1326296046939334e-06,
      "loss": 0.16897620260715485,
      "step": 177,
      "token_acc": 0.9449275362318841
    },
    {
      "epoch": 3.495049504950495,
      "grad_norm": 1.6672917604446411,
      "learning_rate": 1.1061211245213338e-06,
      "loss": 0.09250764548778534,
      "step": 178,
      "token_acc": 0.9670932358318098
    },
    {
      "epoch": 3.514851485148515,
      "grad_norm": 1.0188837051391602,
      "learning_rate": 1.079838133172111e-06,
      "loss": 0.04335964843630791,
      "step": 179,
      "token_acc": 0.9885807504078303
    },
    {
      "epoch": 3.5346534653465347,
      "grad_norm": 1.052925705909729,
      "learning_rate": 1.0537848824653419e-06,
      "loss": 0.050165336579084396,
      "step": 180,
      "token_acc": 0.98
    },
    {
      "epoch": 3.5544554455445545,
      "grad_norm": 2.140634059906006,
      "learning_rate": 1.0279655870547894e-06,
      "loss": 0.17244857549667358,
      "step": 181,
      "token_acc": 0.9317548746518106
    },
    {
      "epoch": 3.5742574257425743,
      "grad_norm": 1.7252540588378906,
      "learning_rate": 1.002384423747093e-06,
      "loss": 0.09934945404529572,
      "step": 182,
      "token_acc": 0.9650735294117647
    },
    {
      "epoch": 3.594059405940594,
      "grad_norm": 0.000313323806039989,
      "learning_rate": 9.77045530826086e-07,
      "loss": 1.8877468619393767e-06,
      "step": 183,
      "token_acc": 1.0
    },
    {
      "epoch": 3.613861386138614,
      "grad_norm": 2.33498477935791,
      "learning_rate": 9.519530073833405e-07,
      "loss": 0.16016721725463867,
      "step": 184,
      "token_acc": 0.9464285714285714
    },
    {
      "epoch": 3.633663366336634,
      "grad_norm": 2.1279542446136475,
      "learning_rate": 9.271109126550515e-07,
      "loss": 0.09996386617422104,
      "step": 185,
      "token_acc": 0.9643605870020965
    },
    {
      "epoch": 3.6534653465346536,
      "grad_norm": 1.1761771440505981,
      "learning_rate": 9.025232653653801e-07,
      "loss": 0.03842503949999809,
      "step": 186,
      "token_acc": 0.9861111111111112
    },
    {
      "epoch": 3.6732673267326734,
      "grad_norm": 1.884169101715088,
      "learning_rate": 8.781940430763328e-07,
      "loss": 0.13109558820724487,
      "step": 187,
      "token_acc": 0.9562043795620438
    },
    {
      "epoch": 3.693069306930693,
      "grad_norm": 0.9455557465553284,
      "learning_rate": 8.5412718154431e-07,
      "loss": 0.04397479072213173,
      "step": 188,
      "token_acc": 0.9796557120500783
    },
    {
      "epoch": 3.7128712871287126,
      "grad_norm": 2.342214822769165,
      "learning_rate": 8.303265740834182e-07,
      "loss": 0.03308221325278282,
      "step": 189,
      "token_acc": 0.9910554561717353
    },
    {
      "epoch": 3.7326732673267324,
      "grad_norm": 2.943640947341919,
      "learning_rate": 8.067960709356479e-07,
      "loss": 0.09943701326847076,
      "step": 190,
      "token_acc": 0.9624217118997912
    },
    {
      "epoch": 3.7524752475247523,
      "grad_norm": 2.837277412414551,
      "learning_rate": 7.83539478648008e-07,
      "loss": 0.16247156262397766,
      "step": 191,
      "token_acc": 0.9338842975206612
    },
    {
      "epoch": 3.772277227722772,
      "grad_norm": 1.8565586805343628,
      "learning_rate": 7.605605594567477e-07,
      "loss": 0.04198635742068291,
      "step": 192,
      "token_acc": 0.9845261121856866
    },
    {
      "epoch": 3.792079207920792,
      "grad_norm": 2.336308479309082,
      "learning_rate": 7.378630306787315e-07,
      "loss": 0.10302429646253586,
      "step": 193,
      "token_acc": 0.9588785046728971
    },
    {
      "epoch": 3.8118811881188117,
      "grad_norm": 2.848647117614746,
      "learning_rate": 7.15450564110087e-07,
      "loss": 0.044242676347494125,
      "step": 194,
      "token_acc": 0.984822934232715
    },
    {
      "epoch": 3.8316831683168315,
      "grad_norm": 2.8329131603240967,
      "learning_rate": 6.933267854322167e-07,
      "loss": 0.10576768219470978,
      "step": 195,
      "token_acc": 0.9608482871125612
    },
    {
      "epoch": 3.8514851485148514,
      "grad_norm": 3.3709895610809326,
      "learning_rate": 6.714952736252672e-07,
      "loss": 0.08542198687791824,
      "step": 196,
      "token_acc": 0.9712351945854484
    },
    {
      "epoch": 3.871287128712871,
      "grad_norm": 2.1180598735809326,
      "learning_rate": 6.499595603891564e-07,
      "loss": 0.05604492500424385,
      "step": 197,
      "token_acc": 0.9809725158562368
    },
    {
      "epoch": 3.891089108910891,
      "grad_norm": 0.0005274138529784977,
      "learning_rate": 6.28723129572247e-07,
      "loss": 2.883556135202525e-06,
      "step": 198,
      "token_acc": 1.0
    },
    {
      "epoch": 3.910891089108911,
      "grad_norm": 2.1085925102233887,
      "learning_rate": 6.07789416607763e-07,
      "loss": 0.04899070784449577,
      "step": 199,
      "token_acc": 0.9854014598540146
    },
    {
      "epoch": 3.9306930693069306,
      "grad_norm": 1.8721121549606323,
      "learning_rate": 5.871618079580327e-07,
      "loss": 0.04069714993238449,
      "step": 200,
      "token_acc": 0.9861386138613861
    },
    {
      "epoch": 3.9504950495049505,
      "grad_norm": 2.5635242462158203,
      "learning_rate": 5.668436405666655e-07,
      "loss": 0.10191994905471802,
      "step": 201,
      "token_acc": 0.9627228525121556
    },
    {
      "epoch": 3.9702970297029703,
      "grad_norm": 2.964982509613037,
      "learning_rate": 5.468382013187274e-07,
      "loss": 0.15324977040290833,
      "step": 202,
      "token_acc": 0.9467680608365019
    },
    {
      "epoch": 3.99009900990099,
      "grad_norm": 3.3698129653930664,
      "learning_rate": 5.271487265090163e-07,
      "loss": 0.21770858764648438,
      "step": 203,
      "token_acc": 0.9187817258883249
    },
    {
      "epoch": 4.0,
      "grad_norm": 4.455423831939697,
      "learning_rate": 5.077784013185319e-07,
      "loss": 0.11420600861310959,
      "step": 204,
      "token_acc": 0.9642857142857143
    },
    {
      "epoch": 4.01980198019802,
      "grad_norm": 3.918121337890625,
      "learning_rate": 4.887303592992002e-07,
      "loss": 0.06277328729629517,
      "step": 205,
      "token_acc": 0.9772727272727273
    },
    {
      "epoch": 4.03960396039604,
      "grad_norm": 2.2823991775512695,
      "learning_rate": 4.700076818669591e-07,
      "loss": 0.10903944075107574,
      "step": 206,
      "token_acc": 0.9625850340136054
    },
    {
      "epoch": 4.0594059405940595,
      "grad_norm": 2.8564696311950684,
      "learning_rate": 4.516133978032758e-07,
      "loss": 0.06087807938456535,
      "step": 207,
      "token_acc": 0.9810526315789474
    },
    {
      "epoch": 4.079207920792079,
      "grad_norm": 3.087757110595703,
      "learning_rate": 4.3355048276517655e-07,
      "loss": 0.09883318096399307,
      "step": 208,
      "token_acc": 0.969439728353141
    },
    {
      "epoch": 4.099009900990099,
      "grad_norm": 1.6992267370224,
      "learning_rate": 4.1582185880387454e-07,
      "loss": 0.07258477061986923,
      "step": 209,
      "token_acc": 0.976271186440678
    },
    {
      "epoch": 4.118811881188119,
      "grad_norm": 3.171792507171631,
      "learning_rate": 3.9843039389206667e-07,
      "loss": 0.07628844678401947,
      "step": 210,
      "token_acc": 0.9690444145356663
    },
    {
      "epoch": 4.138613861386139,
      "grad_norm": 2.5836098194122314,
      "learning_rate": 3.813789014599781e-07,
      "loss": 0.13898494839668274,
      "step": 211,
      "token_acc": 0.9528130671506352
    },
    {
      "epoch": 4.158415841584159,
      "grad_norm": 2.3464548587799072,
      "learning_rate": 3.6467013994023333e-07,
      "loss": 0.02614203840494156,
      "step": 212,
      "token_acc": 0.9940476190476191
    },
    {
      "epoch": 4.178217821782178,
      "grad_norm": 1.5897901058197021,
      "learning_rate": 3.483068123216246e-07,
      "loss": 0.027943788096308708,
      "step": 213,
      "token_acc": 0.9905660377358491
    },
    {
      "epoch": 4.198019801980198,
      "grad_norm": 3.7007808685302734,
      "learning_rate": 3.3229156571184086e-07,
      "loss": 0.1025439202785492,
      "step": 214,
      "token_acc": 0.9723756906077348
    },
    {
      "epoch": 4.217821782178218,
      "grad_norm": 2.4944381713867188,
      "learning_rate": 3.1662699090924925e-07,
      "loss": 0.11341433227062225,
      "step": 215,
      "token_acc": 0.9559543230016313
    },
    {
      "epoch": 4.237623762376238,
      "grad_norm": 1.5083459615707397,
      "learning_rate": 3.0131562198377763e-07,
      "loss": 0.04218436777591705,
      "step": 216,
      "token_acc": 0.9864077669902913
    },
    {
      "epoch": 4.257425742574258,
      "grad_norm": 2.8699820041656494,
      "learning_rate": 2.8635993586697555e-07,
      "loss": 0.022960849106311798,
      "step": 217,
      "token_acc": 0.9960079840319361
    },
    {
      "epoch": 4.2772277227722775,
      "grad_norm": 2.660264730453491,
      "learning_rate": 2.717623519513199e-07,
      "loss": 0.07656387984752655,
      "step": 218,
      "token_acc": 0.9735099337748344
    },
    {
      "epoch": 4.297029702970297,
      "grad_norm": 3.680818557739258,
      "learning_rate": 2.575252316988286e-07,
      "loss": 0.03125208988785744,
      "step": 219,
      "token_acc": 0.9864559819413092
    },
    {
      "epoch": 4.316831683168317,
      "grad_norm": 1.5799322128295898,
      "learning_rate": 2.4365087825904353e-07,
      "loss": 0.04260074347257614,
      "step": 220,
      "token_acc": 0.9854545454545455
    },
    {
      "epoch": 4.336633663366337,
      "grad_norm": 0.00047183220158331096,
      "learning_rate": 2.3014153609644985e-07,
      "loss": 2.44021703110775e-06,
      "step": 221,
      "token_acc": 1.0
    },
    {
      "epoch": 4.356435643564357,
      "grad_norm": 1.4812023639678955,
      "learning_rate": 2.169993906273865e-07,
      "loss": 0.039052195847034454,
      "step": 222,
      "token_acc": 0.9849340866290018
    },
    {
      "epoch": 4.376237623762377,
      "grad_norm": 2.4544718265533447,
      "learning_rate": 2.0422656786650868e-07,
      "loss": 0.062005531042814255,
      "step": 223,
      "token_acc": 0.9832402234636871
    },
    {
      "epoch": 4.396039603960396,
      "grad_norm": 2.964094877243042,
      "learning_rate": 1.9182513408286446e-07,
      "loss": 0.11347703635692596,
      "step": 224,
      "token_acc": 0.9622302158273381
    },
    {
      "epoch": 4.415841584158416,
      "grad_norm": 2.016446590423584,
      "learning_rate": 1.7979709546562974e-07,
      "loss": 0.06778152287006378,
      "step": 225,
      "token_acc": 0.9770114942528736
    },
    {
      "epoch": 4.435643564356436,
      "grad_norm": 2.8770592212677,
      "learning_rate": 1.6814439779956505e-07,
      "loss": 0.11741024255752563,
      "step": 226,
      "token_acc": 0.9593908629441624
    },
    {
      "epoch": 4.455445544554456,
      "grad_norm": 4.033825397491455,
      "learning_rate": 1.5686892615024678e-07,
      "loss": 0.13604798913002014,
      "step": 227,
      "token_acc": 0.9594320486815415
    },
    {
      "epoch": 4.475247524752476,
      "grad_norm": 2.3651978969573975,
      "learning_rate": 1.45972504559119e-07,
      "loss": 0.029621092602610588,
      "step": 228,
      "token_acc": 0.99375
    },
    {
      "epoch": 4.4950495049504955,
      "grad_norm": 3.787503480911255,
      "learning_rate": 1.3545689574841341e-07,
      "loss": 0.14190229773521423,
      "step": 229,
      "token_acc": 0.9496183206106871
    },
    {
      "epoch": 4.514851485148515,
      "grad_norm": 1.3548166751861572,
      "learning_rate": 1.253238008359961e-07,
      "loss": 0.02568677067756653,
      "step": 230,
      "token_acc": 0.9896907216494846
    },
    {
      "epoch": 4.534653465346535,
      "grad_norm": 2.068086624145508,
      "learning_rate": 1.1557485906017491e-07,
      "loss": 0.016756512224674225,
      "step": 231,
      "token_acc": 0.9951298701298701
    },
    {
      "epoch": 4.554455445544555,
      "grad_norm": 5.429747581481934,
      "learning_rate": 1.0621164751451835e-07,
      "loss": 0.03155110776424408,
      "step": 232,
      "token_acc": 0.9887133182844243
    },
    {
      "epoch": 4.574257425742574,
      "grad_norm": 3.2127909660339355,
      "learning_rate": 9.723568089272883e-08,
      "loss": 0.09574354439973831,
      "step": 233,
      "token_acc": 0.9657534246575342
    },
    {
      "epoch": 4.594059405940594,
      "grad_norm": 4.545624732971191,
      "learning_rate": 8.864841124360924e-08,
      "loss": 0.10262885689735413,
      "step": 234,
      "token_acc": 0.9641638225255973
    },
    {
      "epoch": 4.6138613861386135,
      "grad_norm": 2.7521450519561768,
      "learning_rate": 8.045122773616353e-08,
      "loss": 0.0403362475335598,
      "step": 235,
      "token_acc": 0.9856
    },
    {
      "epoch": 4.633663366336633,
      "grad_norm": 0.0002498556859791279,
      "learning_rate": 7.264545643486997e-08,
      "loss": 1.3765927633357933e-06,
      "step": 236,
      "token_acc": 1.0
    },
    {
      "epoch": 4.653465346534653,
      "grad_norm": 3.516314744949341,
      "learning_rate": 6.523236008516343e-08,
      "loss": 0.09355474263429642,
      "step": 237,
      "token_acc": 0.9681274900398407
    },
    {
      "epoch": 4.673267326732673,
      "grad_norm": 4.174475193023682,
      "learning_rate": 5.821313790915884e-08,
      "loss": 0.05687769874930382,
      "step": 238,
      "token_acc": 0.9831730769230769
    },
    {
      "epoch": 4.693069306930693,
      "grad_norm": 0.00025663376436568797,
      "learning_rate": 5.158892541165278e-08,
      "loss": 1.409870264978963e-06,
      "step": 239,
      "token_acc": 1.0
    },
    {
      "epoch": 4.712871287128713,
      "grad_norm": 2.226823329925537,
      "learning_rate": 4.536079419643374e-08,
      "loss": 0.038166601210832596,
      "step": 240,
      "token_acc": 0.9841269841269841
    },
    {
      "epoch": 4.732673267326732,
      "grad_norm": 3.001652956008911,
      "learning_rate": 3.952975179292384e-08,
      "loss": 0.08970314264297485,
      "step": 241,
      "token_acc": 0.9705400981996727
    },
    {
      "epoch": 4.752475247524752,
      "grad_norm": 2.810194492340088,
      "learning_rate": 3.4096741493194196e-08,
      "loss": 0.07326734066009521,
      "step": 242,
      "token_acc": 0.9768339768339769
    },
    {
      "epoch": 4.772277227722772,
      "grad_norm": 3.358609437942505,
      "learning_rate": 2.9062642199365333e-08,
      "loss": 0.09861033409833908,
      "step": 243,
      "token_acc": 0.9621993127147767
    },
    {
      "epoch": 4.792079207920792,
      "grad_norm": 4.996192932128906,
      "learning_rate": 2.4428268281428134e-08,
      "loss": 0.09353439509868622,
      "step": 244,
      "token_acc": 0.9685230024213075
    },
    {
      "epoch": 4.811881188118812,
      "grad_norm": 1.6667205095291138,
      "learning_rate": 2.0194369445501185e-08,
      "loss": 0.03237545117735863,
      "step": 245,
      "token_acc": 0.988391376451078
    },
    {
      "epoch": 4.8316831683168315,
      "grad_norm": 3.3907883167266846,
      "learning_rate": 1.6361630612551082e-08,
      "loss": 0.09833278506994247,
      "step": 246,
      "token_acc": 0.9689349112426036
    },
    {
      "epoch": 4.851485148514851,
      "grad_norm": 5.285821437835693,
      "learning_rate": 1.2930671807592499e-08,
      "loss": 0.085192009806633,
      "step": 247,
      "token_acc": 0.9688888888888889
    },
    {
      "epoch": 4.871287128712871,
      "grad_norm": 5.49505615234375,
      "learning_rate": 9.90204805938505e-09,
      "loss": 0.11812274158000946,
      "step": 248,
      "token_acc": 0.9538461538461539
    },
    {
      "epoch": 4.891089108910891,
      "grad_norm": 3.13936710357666,
      "learning_rate": 7.276249310647632e-09,
      "loss": 0.08899760246276855,
      "step": 249,
      "token_acc": 0.969551282051282
    },
    {
      "epoch": 4.910891089108911,
      "grad_norm": 3.52523136138916,
      "learning_rate": 5.0537003387982066e-09,
      "loss": 0.12114743143320084,
      "step": 250,
      "token_acc": 0.9664879356568364
    },
    {
      "epoch": 4.930693069306931,
      "grad_norm": 2.4868149757385254,
      "learning_rate": 3.2347606872387695e-09,
      "loss": 0.08555735647678375,
      "step": 251,
      "token_acc": 0.9681208053691275
    },
    {
      "epoch": 4.9504950495049505,
      "grad_norm": 2.2696316242218018,
      "learning_rate": 1.8197246071910424e-09,
      "loss": 0.047724779695272446,
      "step": 252,
      "token_acc": 0.9845201238390093
    },
    {
      "epoch": 4.97029702970297,
      "grad_norm": 4.347671031951904,
      "learning_rate": 8.088210100951066e-10,
      "loss": 0.13052120804786682,
      "step": 253,
      "token_acc": 0.9566003616636528
    },
    {
      "epoch": 4.99009900990099,
      "grad_norm": 3.7752373218536377,
      "learning_rate": 2.0221343057791332e-10,
      "loss": 0.09143036603927612,
      "step": 254,
      "token_acc": 0.9687092568448501
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.00023684637562837452,
      "learning_rate": 0.0,
      "loss": 1.288411453970184e-06,
      "step": 255,
      "token_acc": 1.0
    }
  ],
  "logging_steps": 1,
  "max_steps": 255,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 88857543692288.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}