{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 100,
  "global_step": 3897,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 45.099327087402344,
      "learning_rate": 0.00019948678470618425,
      "loss": 2.0895,
      "step": 10
    },
    {
      "epoch": 0.01,
      "grad_norm": 34.96247100830078,
      "learning_rate": 0.0001989735694123685,
      "loss": 2.3707,
      "step": 20
    },
    {
      "epoch": 0.01,
      "grad_norm": 16.490638732910156,
      "learning_rate": 0.00019846035411855275,
      "loss": 1.3701,
      "step": 30
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.11159859597682953,
      "learning_rate": 0.000197947138824737,
      "loss": 1.0582,
      "step": 40
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.12606161832809448,
      "learning_rate": 0.00019743392353092123,
      "loss": 1.1387,
      "step": 50
    },
    {
      "epoch": 0.02,
      "grad_norm": 4.384986400604248,
      "learning_rate": 0.00019692070823710547,
      "loss": 2.2113,
      "step": 60
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.7131040692329407,
      "learning_rate": 0.00019640749294328974,
      "loss": 0.6345,
      "step": 70
    },
    {
      "epoch": 0.02,
      "grad_norm": 5.308964252471924,
      "learning_rate": 0.00019589427764947397,
      "loss": 1.0728,
      "step": 80
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.04876876622438431,
      "learning_rate": 0.0001953810623556582,
      "loss": 1.6609,
      "step": 90
    },
    {
      "epoch": 0.03,
      "grad_norm": 8.89411735534668,
      "learning_rate": 0.00019486784706184245,
      "loss": 1.8662,
      "step": 100
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.05204994976520538,
      "learning_rate": 0.00019435463176802672,
      "loss": 0.5977,
      "step": 110
    },
    {
      "epoch": 0.03,
      "grad_norm": 34.89448165893555,
      "learning_rate": 0.00019384141647421096,
      "loss": 2.2815,
      "step": 120
    },
    {
      "epoch": 0.03,
      "grad_norm": 15.786956787109375,
      "learning_rate": 0.0001933282011803952,
      "loss": 1.6381,
      "step": 130
    },
    {
      "epoch": 0.04,
      "grad_norm": 29.863065719604492,
      "learning_rate": 0.00019281498588657943,
      "loss": 0.7649,
      "step": 140
    },
    {
      "epoch": 0.04,
      "grad_norm": 16.698570251464844,
      "learning_rate": 0.00019230177059276367,
      "loss": 1.1147,
      "step": 150
    },
    {
      "epoch": 0.04,
      "grad_norm": 34.869384765625,
      "learning_rate": 0.00019178855529894794,
      "loss": 1.4108,
      "step": 160
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.3193562924861908,
      "learning_rate": 0.00019127534000513218,
      "loss": 1.0166,
      "step": 170
    },
    {
      "epoch": 0.05,
      "grad_norm": 23.633769989013672,
      "learning_rate": 0.00019076212471131642,
      "loss": 1.0537,
      "step": 180
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.03772430121898651,
      "learning_rate": 0.00019024890941750065,
      "loss": 0.6254,
      "step": 190
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.246229648590088,
      "learning_rate": 0.0001897356941236849,
      "loss": 2.0714,
      "step": 200
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.4526054859161377,
      "learning_rate": 0.00018922247882986913,
      "loss": 1.0681,
      "step": 210
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.3229526281356812,
      "learning_rate": 0.00018870926353605337,
      "loss": 0.6116,
      "step": 220
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.042341507971286774,
      "learning_rate": 0.0001881960482422376,
      "loss": 1.4669,
      "step": 230
    },
    {
      "epoch": 0.06,
      "grad_norm": 35.0599479675293,
      "learning_rate": 0.00018768283294842188,
      "loss": 0.6897,
      "step": 240
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.21465550363063812,
      "learning_rate": 0.00018716961765460611,
      "loss": 1.8696,
      "step": 250
    },
    {
      "epoch": 0.07,
      "grad_norm": 34.034175872802734,
      "learning_rate": 0.00018665640236079035,
      "loss": 1.5576,
      "step": 260
    },
    {
      "epoch": 0.07,
      "grad_norm": 34.895172119140625,
      "learning_rate": 0.0001861431870669746,
      "loss": 1.6542,
      "step": 270
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.30317962169647217,
      "learning_rate": 0.00018562997177315883,
      "loss": 0.9893,
      "step": 280
    },
    {
      "epoch": 0.07,
      "grad_norm": 31.76571273803711,
      "learning_rate": 0.0001851167564793431,
      "loss": 2.43,
      "step": 290
    },
    {
      "epoch": 0.08,
      "grad_norm": 3.9265520572662354,
      "learning_rate": 0.00018460354118552733,
      "loss": 0.6246,
      "step": 300
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.8710033893585205,
      "learning_rate": 0.00018409032589171157,
      "loss": 1.2509,
      "step": 310
    },
    {
      "epoch": 0.08,
      "grad_norm": 28.637104034423828,
      "learning_rate": 0.0001835771105978958,
      "loss": 1.4902,
      "step": 320
    },
    {
      "epoch": 0.08,
      "grad_norm": 4.821115970611572,
      "learning_rate": 0.00018306389530408008,
      "loss": 0.7693,
      "step": 330
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.466738224029541,
      "learning_rate": 0.00018255068001026432,
      "loss": 1.9563,
      "step": 340
    },
    {
      "epoch": 0.09,
      "grad_norm": 34.78805160522461,
      "learning_rate": 0.00018203746471644856,
      "loss": 1.4862,
      "step": 350
    },
    {
      "epoch": 0.09,
      "grad_norm": 5.344199180603027,
      "learning_rate": 0.0001815242494226328,
      "loss": 1.1602,
      "step": 360
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.4355538785457611,
      "learning_rate": 0.00018101103412881703,
      "loss": 1.2546,
      "step": 370
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.7162923216819763,
      "learning_rate": 0.0001804978188350013,
      "loss": 3.0478,
      "step": 380
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.09367480874061584,
      "learning_rate": 0.00017998460354118554,
      "loss": 1.0921,
      "step": 390
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.20555052161216736,
      "learning_rate": 0.00017947138824736978,
      "loss": 0.0523,
      "step": 400
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.08092090487480164,
      "learning_rate": 0.00017895817295355402,
      "loss": 1.5604,
      "step": 410
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.32685595750808716,
      "learning_rate": 0.00017844495765973828,
      "loss": 0.5272,
      "step": 420
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.5422254800796509,
      "learning_rate": 0.00017793174236592252,
      "loss": 3.4396,
      "step": 430
    },
    {
      "epoch": 0.11,
      "grad_norm": 35.279212951660156,
      "learning_rate": 0.00017741852707210676,
      "loss": 0.8554,
      "step": 440
    },
    {
      "epoch": 0.12,
      "grad_norm": 26.898597717285156,
      "learning_rate": 0.000176905311778291,
      "loss": 0.6776,
      "step": 450
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.39926862716674805,
      "learning_rate": 0.00017639209648447526,
      "loss": 2.0551,
      "step": 460
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.39992421865463257,
      "learning_rate": 0.0001758788811906595,
      "loss": 0.6316,
      "step": 470
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.15786024928092957,
      "learning_rate": 0.00017536566589684374,
      "loss": 2.5885,
      "step": 480
    },
    {
      "epoch": 0.13,
      "grad_norm": 35.12531280517578,
      "learning_rate": 0.00017485245060302798,
      "loss": 1.5028,
      "step": 490
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.712053656578064,
      "learning_rate": 0.00017433923530921222,
      "loss": 0.7324,
      "step": 500
    },
    {
      "epoch": 0.13,
      "grad_norm": 28.602834701538086,
      "learning_rate": 0.00017382602001539648,
      "loss": 2.7516,
      "step": 510
    },
    {
      "epoch": 0.13,
      "grad_norm": 8.03342342376709,
      "learning_rate": 0.00017331280472158072,
      "loss": 0.3827,
      "step": 520
    },
    {
      "epoch": 0.14,
      "grad_norm": 33.861934661865234,
      "learning_rate": 0.00017279958942776496,
      "loss": 1.1201,
      "step": 530
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.0627857968211174,
      "learning_rate": 0.0001722863741339492,
      "loss": 0.6096,
      "step": 540
    },
    {
      "epoch": 0.14,
      "grad_norm": 34.3585319519043,
      "learning_rate": 0.00017177315884013344,
      "loss": 1.4903,
      "step": 550
    },
    {
      "epoch": 0.14,
      "grad_norm": 33.19193649291992,
      "learning_rate": 0.00017125994354631768,
      "loss": 1.4301,
      "step": 560
    },
    {
      "epoch": 0.15,
      "grad_norm": 34.12609100341797,
      "learning_rate": 0.00017074672825250194,
      "loss": 1.1228,
      "step": 570
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.07128182798624039,
      "learning_rate": 0.00017023351295868618,
      "loss": 0.8195,
      "step": 580
    },
    {
      "epoch": 0.15,
      "grad_norm": 34.57550048828125,
      "learning_rate": 0.00016972029766487042,
      "loss": 1.9291,
      "step": 590
    },
    {
      "epoch": 0.15,
      "grad_norm": 60.265777587890625,
      "learning_rate": 0.00016920708237105466,
      "loss": 2.2235,
      "step": 600
    },
    {
      "epoch": 0.16,
      "grad_norm": 33.86371612548828,
      "learning_rate": 0.0001686938670772389,
      "loss": 1.101,
      "step": 610
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.018540620803833,
      "learning_rate": 0.00016818065178342314,
      "loss": 1.0946,
      "step": 620
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.05546221882104874,
      "learning_rate": 0.00016766743648960738,
      "loss": 1.3699,
      "step": 630
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.4712983965873718,
      "learning_rate": 0.00016715422119579164,
      "loss": 2.366,
      "step": 640
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.1462753266096115,
      "learning_rate": 0.00016664100590197588,
      "loss": 1.1087,
      "step": 650
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.0955154225230217,
      "learning_rate": 0.00016612779060816012,
      "loss": 0.0171,
      "step": 660
    },
    {
      "epoch": 0.17,
      "grad_norm": 35.19109344482422,
      "learning_rate": 0.00016561457531434436,
      "loss": 1.6907,
      "step": 670
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.16912657022476196,
      "learning_rate": 0.00016510136002052862,
      "loss": 0.0055,
      "step": 680
    },
    {
      "epoch": 0.18,
      "grad_norm": 34.529685974121094,
      "learning_rate": 0.00016458814472671286,
      "loss": 2.8261,
      "step": 690
    },
    {
      "epoch": 0.18,
      "grad_norm": 13.726387023925781,
      "learning_rate": 0.0001640749294328971,
      "loss": 1.7874,
      "step": 700
    },
    {
      "epoch": 0.18,
      "grad_norm": 29.94759750366211,
      "learning_rate": 0.00016356171413908134,
      "loss": 1.3416,
      "step": 710
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.5144930481910706,
      "learning_rate": 0.0001630484988452656,
      "loss": 0.7381,
      "step": 720
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.4536816477775574,
      "learning_rate": 0.00016253528355144985,
      "loss": 1.7744,
      "step": 730
    },
    {
      "epoch": 0.19,
      "grad_norm": 34.2490119934082,
      "learning_rate": 0.00016202206825763408,
      "loss": 1.6362,
      "step": 740
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.7601956725120544,
      "learning_rate": 0.00016150885296381832,
      "loss": 1.0899,
      "step": 750
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.16101330518722534,
      "learning_rate": 0.00016099563767000256,
      "loss": 1.4228,
      "step": 760
    },
    {
      "epoch": 0.2,
      "grad_norm": 2.7051339149475098,
      "learning_rate": 0.00016048242237618683,
      "loss": 1.2515,
      "step": 770
    },
    {
      "epoch": 0.2,
      "grad_norm": 34.69645690917969,
      "learning_rate": 0.00015996920708237107,
      "loss": 1.1568,
      "step": 780
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.5005641579627991,
      "learning_rate": 0.0001594559917885553,
      "loss": 1.8697,
      "step": 790
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.9808595776557922,
      "learning_rate": 0.00015894277649473954,
      "loss": 1.867,
      "step": 800
    },
    {
      "epoch": 0.21,
      "grad_norm": 34.12071228027344,
      "learning_rate": 0.0001584295612009238,
      "loss": 1.2028,
      "step": 810
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.4215242862701416,
      "learning_rate": 0.00015791634590710805,
      "loss": 1.5028,
      "step": 820
    },
    {
      "epoch": 0.21,
      "grad_norm": 32.7039794921875,
      "learning_rate": 0.0001574031306132923,
      "loss": 1.851,
      "step": 830
    },
    {
      "epoch": 0.22,
      "grad_norm": 11.424134254455566,
      "learning_rate": 0.00015688991531947653,
      "loss": 1.1995,
      "step": 840
    },
    {
      "epoch": 0.22,
      "grad_norm": 18.550819396972656,
      "learning_rate": 0.0001563767000256608,
      "loss": 0.5795,
      "step": 850
    },
    {
      "epoch": 0.22,
      "grad_norm": 25.513269424438477,
      "learning_rate": 0.00015586348473184503,
      "loss": 0.6537,
      "step": 860
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.043114595115184784,
      "learning_rate": 0.00015535026943802927,
      "loss": 1.4115,
      "step": 870
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.06105157732963562,
      "learning_rate": 0.0001548370541442135,
      "loss": 0.5029,
      "step": 880
    },
    {
      "epoch": 0.23,
      "grad_norm": 6.027977466583252,
      "learning_rate": 0.00015432383885039775,
      "loss": 1.1377,
      "step": 890
    },
    {
      "epoch": 0.23,
      "grad_norm": 9.911833763122559,
      "learning_rate": 0.000153810623556582,
      "loss": 0.6785,
      "step": 900
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.08647739142179489,
      "learning_rate": 0.00015329740826276625,
      "loss": 0.3428,
      "step": 910
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.06115918606519699,
      "learning_rate": 0.0001527841929689505,
      "loss": 0.5422,
      "step": 920
    },
    {
      "epoch": 0.24,
      "grad_norm": 33.05713653564453,
      "learning_rate": 0.00015227097767513473,
      "loss": 0.9532,
      "step": 930
    },
    {
      "epoch": 0.24,
      "grad_norm": 2.755101442337036,
      "learning_rate": 0.00015175776238131897,
      "loss": 0.8933,
      "step": 940
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.09266197681427,
      "learning_rate": 0.0001512445470875032,
      "loss": 1.3081,
      "step": 950
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.8350268602371216,
      "learning_rate": 0.00015073133179368745,
      "loss": 1.6349,
      "step": 960
    },
    {
      "epoch": 0.25,
      "grad_norm": 28.465572357177734,
      "learning_rate": 0.00015021811649987168,
      "loss": 1.4432,
      "step": 970
    },
    {
      "epoch": 0.25,
      "grad_norm": 24.18708038330078,
      "learning_rate": 0.00014970490120605595,
      "loss": 0.8608,
      "step": 980
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.06248585879802704,
      "learning_rate": 0.0001491916859122402,
      "loss": 0.7754,
      "step": 990
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.02755439281463623,
      "learning_rate": 0.00014867847061842443,
      "loss": 0.3644,
      "step": 1000
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.03344785049557686,
      "learning_rate": 0.00014816525532460867,
      "loss": 2.7486,
      "step": 1010
    },
    {
      "epoch": 0.26,
      "grad_norm": 4.339983940124512,
      "learning_rate": 0.0001476520400307929,
      "loss": 0.8171,
      "step": 1020
    },
    {
      "epoch": 0.26,
      "grad_norm": 4.660573959350586,
      "learning_rate": 0.00014713882473697717,
      "loss": 2.1821,
      "step": 1030
    },
    {
      "epoch": 0.27,
      "grad_norm": 21.810422897338867,
      "learning_rate": 0.0001466256094431614,
      "loss": 1.0611,
      "step": 1040
    },
    {
      "epoch": 0.27,
      "grad_norm": 33.206607818603516,
      "learning_rate": 0.00014611239414934565,
      "loss": 1.5671,
      "step": 1050
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.678447425365448,
      "learning_rate": 0.0001455991788555299,
      "loss": 1.0065,
      "step": 1060
    },
    {
      "epoch": 0.27,
      "grad_norm": 32.59206771850586,
      "learning_rate": 0.00014508596356171415,
      "loss": 1.8422,
      "step": 1070
    },
    {
      "epoch": 0.28,
      "grad_norm": 28.002819061279297,
      "learning_rate": 0.0001445727482678984,
      "loss": 1.1739,
      "step": 1080
    },
    {
      "epoch": 0.28,
      "grad_norm": 4.193221092224121,
      "learning_rate": 0.00014405953297408263,
      "loss": 0.431,
      "step": 1090
    },
    {
      "epoch": 0.28,
      "grad_norm": 34.987388610839844,
      "learning_rate": 0.00014354631768026687,
      "loss": 1.1974,
      "step": 1100
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.1439794898033142,
      "learning_rate": 0.0001430331023864511,
      "loss": 2.1824,
      "step": 1110
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.07825344055891037,
      "learning_rate": 0.00014251988709263537,
      "loss": 1.103,
      "step": 1120
    },
    {
      "epoch": 0.29,
      "grad_norm": 34.14964294433594,
      "learning_rate": 0.0001420066717988196,
      "loss": 2.1521,
      "step": 1130
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.839199185371399,
      "learning_rate": 0.00014149345650500385,
      "loss": 0.0536,
      "step": 1140
    },
    {
      "epoch": 0.3,
      "grad_norm": 34.573585510253906,
      "learning_rate": 0.0001409802412111881,
      "loss": 1.5661,
      "step": 1150
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.41036948561668396,
      "learning_rate": 0.00014046702591737236,
      "loss": 1.1817,
      "step": 1160
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.08495642244815826,
      "learning_rate": 0.0001399538106235566,
      "loss": 1.5704,
      "step": 1170
    },
    {
      "epoch": 0.3,
      "grad_norm": 3.0547990798950195,
      "learning_rate": 0.00013944059532974083,
      "loss": 1.5402,
      "step": 1180
    },
    {
      "epoch": 0.31,
      "grad_norm": 7.8949713706970215,
      "learning_rate": 0.00013892738003592507,
      "loss": 0.4595,
      "step": 1190
    },
    {
      "epoch": 0.31,
      "grad_norm": 24.375001907348633,
      "learning_rate": 0.00013841416474210934,
      "loss": 0.7511,
      "step": 1200
    },
    {
      "epoch": 0.31,
      "grad_norm": 12.163422584533691,
      "learning_rate": 0.00013790094944829358,
      "loss": 0.5726,
      "step": 1210
    },
    {
      "epoch": 0.31,
      "grad_norm": 34.362632751464844,
      "learning_rate": 0.00013738773415447782,
      "loss": 1.3018,
      "step": 1220
    },
    {
      "epoch": 0.32,
      "grad_norm": 21.605195999145508,
      "learning_rate": 0.00013687451886066205,
      "loss": 1.2047,
      "step": 1230
    },
    {
      "epoch": 0.32,
      "grad_norm": 4.356873035430908,
      "learning_rate": 0.0001363613035668463,
      "loss": 1.3787,
      "step": 1240
    },
    {
      "epoch": 0.32,
      "grad_norm": 12.701129913330078,
      "learning_rate": 0.00013584808827303056,
      "loss": 1.0334,
      "step": 1250
    },
    {
      "epoch": 0.32,
      "grad_norm": 7.2957892417907715,
      "learning_rate": 0.0001353348729792148,
      "loss": 0.2939,
      "step": 1260
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.25156155228614807,
      "learning_rate": 0.00013482165768539904,
      "loss": 0.7811,
      "step": 1270
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.8677279353141785,
      "learning_rate": 0.00013430844239158328,
      "loss": 1.5475,
      "step": 1280
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.13093720376491547,
      "learning_rate": 0.00013379522709776754,
      "loss": 0.7902,
      "step": 1290
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.2733534276485443,
      "learning_rate": 0.00013328201180395178,
      "loss": 0.9059,
      "step": 1300
    },
    {
      "epoch": 0.34,
      "grad_norm": 34.48268508911133,
      "learning_rate": 0.00013276879651013602,
      "loss": 1.5932,
      "step": 1310
    },
    {
      "epoch": 0.34,
      "grad_norm": 33.39665603637695,
      "learning_rate": 0.00013225558121632026,
      "loss": 2.4436,
      "step": 1320
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.6858634948730469,
      "learning_rate": 0.0001317423659225045,
      "loss": 0.8113,
      "step": 1330
    },
    {
      "epoch": 0.34,
      "grad_norm": 32.40642166137695,
      "learning_rate": 0.00013122915062868873,
      "loss": 2.1392,
      "step": 1340
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.05664900690317154,
      "learning_rate": 0.00013071593533487297,
      "loss": 0.8724,
      "step": 1350
    },
    {
      "epoch": 0.35,
      "grad_norm": 30.9477481842041,
      "learning_rate": 0.0001302027200410572,
      "loss": 1.2746,
      "step": 1360
    },
    {
      "epoch": 0.35,
      "grad_norm": 2.9625468254089355,
      "learning_rate": 0.00012968950474724145,
      "loss": 2.0471,
      "step": 1370
    },
    {
      "epoch": 0.35,
      "grad_norm": 6.3701629638671875,
      "learning_rate": 0.00012917628945342572,
      "loss": 0.9569,
      "step": 1380
    },
    {
      "epoch": 0.36,
      "grad_norm": 5.0901007652282715,
      "learning_rate": 0.00012866307415960996,
      "loss": 0.8391,
      "step": 1390
    },
    {
      "epoch": 0.36,
      "grad_norm": 2.827472448348999,
      "learning_rate": 0.0001281498588657942,
      "loss": 0.5071,
      "step": 1400
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.24257418513298035,
      "learning_rate": 0.00012763664357197843,
      "loss": 0.0258,
      "step": 1410
    },
    {
      "epoch": 0.36,
      "grad_norm": 34.39240264892578,
      "learning_rate": 0.0001271234282781627,
      "loss": 1.6826,
      "step": 1420
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.06999631226062775,
      "learning_rate": 0.00012661021298434694,
      "loss": 2.854,
      "step": 1430
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.7731397151947021,
      "learning_rate": 0.00012609699769053118,
      "loss": 1.5991,
      "step": 1440
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.7090578675270081,
      "learning_rate": 0.00012558378239671542,
      "loss": 1.1829,
      "step": 1450
    },
    {
      "epoch": 0.37,
      "grad_norm": 33.13046646118164,
      "learning_rate": 0.00012507056710289968,
      "loss": 1.4736,
      "step": 1460
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.6540558338165283,
      "learning_rate": 0.00012455735180908392,
      "loss": 1.6983,
      "step": 1470
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.6605318784713745,
      "learning_rate": 0.00012404413651526816,
      "loss": 1.5005,
      "step": 1480
    },
    {
      "epoch": 0.38,
      "grad_norm": 31.07722282409668,
      "learning_rate": 0.0001235309212214524,
      "loss": 1.4853,
      "step": 1490
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.135640025138855,
      "learning_rate": 0.00012301770592763664,
      "loss": 0.9949,
      "step": 1500
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.14291101694107056,
      "learning_rate": 0.0001225044906338209,
      "loss": 1.7112,
      "step": 1510
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.07842687517404556,
      "learning_rate": 0.00012199127534000514,
      "loss": 0.677,
      "step": 1520
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.6385894417762756,
      "learning_rate": 0.00012147806004618938,
      "loss": 1.0463,
      "step": 1530
    },
    {
      "epoch": 0.4,
      "grad_norm": 34.2146110534668,
      "learning_rate": 0.00012096484475237362,
      "loss": 2.0908,
      "step": 1540
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.15750457346439362,
      "learning_rate": 0.00012045162945855788,
      "loss": 1.6537,
      "step": 1550
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.15273036062717438,
      "learning_rate": 0.00011993841416474212,
      "loss": 1.8717,
      "step": 1560
    },
    {
      "epoch": 0.4,
      "grad_norm": 4.405086517333984,
      "learning_rate": 0.00011942519887092636,
      "loss": 0.3942,
      "step": 1570
    },
    {
      "epoch": 0.41,
      "grad_norm": 2.592953681945801,
      "learning_rate": 0.0001189119835771106,
      "loss": 0.3217,
      "step": 1580
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.5546024441719055,
      "learning_rate": 0.00011839876828329485,
      "loss": 1.0285,
      "step": 1590
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.09479659050703049,
      "learning_rate": 0.00011788555298947909,
      "loss": 1.7519,
      "step": 1600
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.09771596640348434,
      "learning_rate": 0.00011737233769566333,
      "loss": 1.2341,
      "step": 1610
    },
    {
      "epoch": 0.42,
      "grad_norm": 33.37510681152344,
      "learning_rate": 0.00011685912240184758,
      "loss": 3.3626,
      "step": 1620
    },
    {
      "epoch": 0.42,
      "grad_norm": 32.650489807128906,
      "learning_rate": 0.00011634590710803182,
      "loss": 0.9869,
      "step": 1630
    },
    {
      "epoch": 0.42,
      "grad_norm": 32.52039337158203,
      "learning_rate": 0.00011583269181421607,
      "loss": 1.8117,
      "step": 1640
    },
    {
      "epoch": 0.42,
      "grad_norm": 15.11254596710205,
      "learning_rate": 0.00011531947652040031,
      "loss": 1.288,
      "step": 1650
    },
    {
      "epoch": 0.43,
      "grad_norm": 3.6993765830993652,
      "learning_rate": 0.00011480626122658455,
      "loss": 0.863,
      "step": 1660
    },
    {
      "epoch": 0.43,
      "grad_norm": 34.809791564941406,
      "learning_rate": 0.00011429304593276879,
      "loss": 0.6192,
      "step": 1670
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.10077141970396042,
      "learning_rate": 0.00011377983063895306,
      "loss": 0.571,
      "step": 1680
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.07107020914554596,
      "learning_rate": 0.0001132666153451373,
      "loss": 0.996,
      "step": 1690
    },
    {
      "epoch": 0.44,
      "grad_norm": 33.04214096069336,
      "learning_rate": 0.00011275340005132153,
      "loss": 2.6481,
      "step": 1700
    },
    {
      "epoch": 0.44,
      "grad_norm": 32.640541076660156,
      "learning_rate": 0.00011224018475750577,
      "loss": 2.1778,
      "step": 1710
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.7503925561904907,
      "learning_rate": 0.00011172696946369004,
      "loss": 1.5489,
      "step": 1720
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.348849892616272,
      "learning_rate": 0.00011121375416987428,
      "loss": 1.1267,
      "step": 1730
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.8353062272071838,
      "learning_rate": 0.00011070053887605852,
      "loss": 0.0233,
      "step": 1740
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.4115784466266632,
      "learning_rate": 0.00011018732358224275,
      "loss": 1.3895,
      "step": 1750
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.09716140478849411,
      "learning_rate": 0.00010967410828842699,
      "loss": 0.9971,
      "step": 1760
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.07956521958112717,
      "learning_rate": 0.00010916089299461126,
      "loss": 0.4633,
      "step": 1770
    },
    {
      "epoch": 0.46,
      "grad_norm": 33.34837341308594,
      "learning_rate": 0.0001086476777007955,
      "loss": 0.5945,
      "step": 1780
    },
    {
      "epoch": 0.46,
      "grad_norm": 33.278682708740234,
      "learning_rate": 0.00010813446240697974,
      "loss": 1.0137,
      "step": 1790
    },
    {
      "epoch": 0.46,
      "grad_norm": 32.34352111816406,
      "learning_rate": 0.00010762124711316398,
      "loss": 2.238,
      "step": 1800
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.8423387408256531,
      "learning_rate": 0.00010710803181934823,
      "loss": 0.0119,
      "step": 1810
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.43691009283065796,
      "learning_rate": 0.00010659481652553247,
      "loss": 0.7785,
      "step": 1820
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.27386584877967834,
      "learning_rate": 0.0001060816012317167,
      "loss": 1.4121,
      "step": 1830
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.035622984170913696,
      "learning_rate": 0.00010556838593790094,
      "loss": 1.8601,
      "step": 1840
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.6981528401374817,
      "learning_rate": 0.00010505517064408518,
      "loss": 1.2077,
      "step": 1850
    },
    {
      "epoch": 0.48,
      "grad_norm": 32.1064567565918,
      "learning_rate": 0.00010454195535026945,
      "loss": 1.5198,
      "step": 1860
    },
    {
      "epoch": 0.48,
      "grad_norm": 30.88555908203125,
      "learning_rate": 0.00010402874005645369,
      "loss": 2.0069,
      "step": 1870
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.03494291752576828,
      "learning_rate": 0.00010351552476263793,
      "loss": 0.3124,
      "step": 1880
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.9848475456237793,
      "learning_rate": 0.00010300230946882216,
      "loss": 0.8303,
      "step": 1890
    },
    {
      "epoch": 0.49,
      "grad_norm": 31.325864791870117,
      "learning_rate": 0.00010248909417500643,
      "loss": 0.9595,
      "step": 1900
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.025403378531336784,
      "learning_rate": 0.00010197587888119067,
      "loss": 0.362,
      "step": 1910
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.1182492971420288,
      "learning_rate": 0.00010146266358737491,
      "loss": 1.4183,
      "step": 1920
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.0197270717471838,
      "learning_rate": 0.00010094944829355915,
      "loss": 0.9907,
      "step": 1930
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.5864602327346802,
      "learning_rate": 0.00010043623299974341,
      "loss": 0.9328,
      "step": 1940
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.019860761240124702,
      "learning_rate": 9.992301770592765e-05,
      "loss": 1.2027,
      "step": 1950
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.27393853664398193,
      "learning_rate": 9.940980241211189e-05,
      "loss": 1.2231,
      "step": 1960
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.06520062685012817,
      "learning_rate": 9.889658711829613e-05,
      "loss": 1.1252,
      "step": 1970
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.10603094100952148,
      "learning_rate": 9.838337182448038e-05,
      "loss": 0.5362,
      "step": 1980
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.21443594992160797,
      "learning_rate": 9.787015653066462e-05,
      "loss": 1.6563,
      "step": 1990
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.15189635753631592,
      "learning_rate": 9.735694123684886e-05,
      "loss": 1.5657,
      "step": 2000
    },
    {
      "epoch": 0.52,
      "grad_norm": 32.36518478393555,
      "learning_rate": 9.68437259430331e-05,
      "loss": 2.1702,
      "step": 2010
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.2264794558286667,
      "learning_rate": 9.633051064921735e-05,
      "loss": 1.2663,
      "step": 2020
    },
    {
      "epoch": 0.52,
      "grad_norm": 30.200082778930664,
      "learning_rate": 9.581729535540159e-05,
      "loss": 2.2151,
      "step": 2030
    },
    {
      "epoch": 0.52,
      "grad_norm": 3.5600388050079346,
      "learning_rate": 9.530408006158584e-05,
      "loss": 0.7643,
      "step": 2040
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.1124127209186554,
      "learning_rate": 9.479086476777008e-05,
      "loss": 0.679,
      "step": 2050
    },
    {
      "epoch": 0.53,
      "grad_norm": 28.294071197509766,
      "learning_rate": 9.427764947395433e-05,
      "loss": 1.5172,
      "step": 2060
    },
    {
      "epoch": 0.53,
      "grad_norm": 26.79751205444336,
      "learning_rate": 9.376443418013857e-05,
      "loss": 0.4192,
      "step": 2070
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.1510336697101593,
      "learning_rate": 9.325121888632282e-05,
      "loss": 2.1576,
      "step": 2080
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.32501623034477234,
      "learning_rate": 9.273800359250706e-05,
      "loss": 1.1192,
      "step": 2090
    },
    {
      "epoch": 0.54,
      "grad_norm": 26.4614315032959,
      "learning_rate": 9.22247882986913e-05,
      "loss": 0.7766,
      "step": 2100
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.20047006011009216,
      "learning_rate": 9.171157300487555e-05,
      "loss": 0.6836,
      "step": 2110
    },
    {
      "epoch": 0.54,
      "grad_norm": 10.277036666870117,
      "learning_rate": 9.119835771105979e-05,
      "loss": 0.4951,
      "step": 2120
    },
    {
      "epoch": 0.55,
      "grad_norm": 9.455358505249023,
      "learning_rate": 9.068514241724404e-05,
      "loss": 0.2494,
      "step": 2130
    },
    {
      "epoch": 0.55,
      "grad_norm": 5.371951103210449,
      "learning_rate": 9.017192712342828e-05,
      "loss": 0.7672,
      "step": 2140
    },
    {
      "epoch": 0.55,
      "grad_norm": 4.935911655426025,
      "learning_rate": 8.965871182961254e-05,
      "loss": 0.7815,
      "step": 2150
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.05824290215969086,
      "learning_rate": 8.914549653579677e-05,
      "loss": 0.4563,
      "step": 2160
    },
    {
      "epoch": 0.56,
      "grad_norm": 29.326370239257812,
      "learning_rate": 8.863228124198101e-05,
      "loss": 0.9615,
      "step": 2170
    },
    {
      "epoch": 0.56,
      "grad_norm": 30.599393844604492,
      "learning_rate": 8.811906594816526e-05,
      "loss": 0.617,
      "step": 2180
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.6331340074539185,
      "learning_rate": 8.76058506543495e-05,
      "loss": 1.0741,
      "step": 2190
    },
    {
      "epoch": 0.56,
      "grad_norm": 34.11085891723633,
      "learning_rate": 8.709263536053374e-05,
      "loss": 1.9807,
      "step": 2200
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.4941723942756653,
      "learning_rate": 8.6579420066718e-05,
      "loss": 1.6712,
      "step": 2210
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.8858903646469116,
      "learning_rate": 8.606620477290223e-05,
      "loss": 1.1382,
      "step": 2220
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.4389359951019287,
      "learning_rate": 8.555298947908647e-05,
      "loss": 0.0166,
      "step": 2230
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.27877870202064514,
      "learning_rate": 8.503977418527072e-05,
      "loss": 1.8697,
      "step": 2240
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.0968392938375473,
      "learning_rate": 8.452655889145496e-05,
      "loss": 0.9307,
      "step": 2250
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.07415905594825745,
      "learning_rate": 8.401334359763922e-05,
      "loss": 1.3474,
      "step": 2260
    },
    {
      "epoch": 0.58,
      "grad_norm": 32.0526008605957,
      "learning_rate": 8.350012830382345e-05,
      "loss": 1.9036,
      "step": 2270
    },
    {
      "epoch": 0.59,
      "grad_norm": 31.857463836669922,
      "learning_rate": 8.29869130100077e-05,
      "loss": 1.7843,
      "step": 2280
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.16152064502239227,
      "learning_rate": 8.247369771619195e-05,
      "loss": 1.138,
      "step": 2290
    },
    {
      "epoch": 0.59,
      "grad_norm": 31.4908504486084,
      "learning_rate": 8.19604824223762e-05,
      "loss": 1.348,
      "step": 2300
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.8940775990486145,
      "learning_rate": 8.144726712856044e-05,
      "loss": 1.4317,
      "step": 2310
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.3364328145980835,
      "learning_rate": 8.093405183474469e-05,
      "loss": 1.3774,
      "step": 2320
    },
    {
      "epoch": 0.6,
      "grad_norm": 32.3911018371582,
      "learning_rate": 8.042083654092893e-05,
      "loss": 1.8806,
      "step": 2330
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.0908791646361351,
      "learning_rate": 7.990762124711317e-05,
      "loss": 0.3828,
      "step": 2340
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.7057339549064636,
      "learning_rate": 7.939440595329742e-05,
      "loss": 1.8282,
      "step": 2350
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.9124455451965332,
      "learning_rate": 7.888119065948166e-05,
      "loss": 0.8194,
      "step": 2360
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.6995695233345032,
      "learning_rate": 7.83679753656659e-05,
      "loss": 2.4169,
      "step": 2370
    },
    {
      "epoch": 0.61,
      "grad_norm": 30.961750030517578,
      "learning_rate": 7.785476007185013e-05,
      "loss": 1.0782,
      "step": 2380
    },
    {
      "epoch": 0.61,
      "grad_norm": 30.930931091308594,
      "learning_rate": 7.734154477803439e-05,
      "loss": 1.773,
      "step": 2390
    },
    {
      "epoch": 0.62,
      "grad_norm": 2.427703857421875,
      "learning_rate": 7.682832948421863e-05,
      "loss": 1.1838,
      "step": 2400
    },
    {
      "epoch": 0.62,
      "grad_norm": 28.99689292907715,
      "learning_rate": 7.631511419040288e-05,
      "loss": 0.9657,
      "step": 2410
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.0632254034280777,
      "learning_rate": 7.580189889658712e-05,
      "loss": 0.9273,
      "step": 2420
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.06876543164253235,
      "learning_rate": 7.528868360277137e-05,
      "loss": 0.6686,
      "step": 2430
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.059902504086494446,
      "learning_rate": 7.477546830895561e-05,
      "loss": 0.6242,
      "step": 2440
    },
    {
      "epoch": 0.63,
      "grad_norm": 30.04041290283203,
      "learning_rate": 7.426225301513986e-05,
      "loss": 2.3915,
      "step": 2450
    },
    {
      "epoch": 0.63,
      "grad_norm": 2.442535400390625,
      "learning_rate": 7.37490377213241e-05,
      "loss": 0.5622,
      "step": 2460
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.0483098030090332,
      "learning_rate": 7.323582242750834e-05,
      "loss": 1.5706,
      "step": 2470
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.048585113137960434,
      "learning_rate": 7.272260713369259e-05,
      "loss": 0.8918,
      "step": 2480
    },
    {
      "epoch": 0.64,
      "grad_norm": 29.45044708251953,
      "learning_rate": 7.220939183987683e-05,
      "loss": 1.4446,
      "step": 2490
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.03927851840853691,
      "learning_rate": 7.169617654606108e-05,
      "loss": 0.6336,
      "step": 2500
    },
    {
      "epoch": 0.64,
      "grad_norm": 5.8573174476623535,
      "learning_rate": 7.118296125224532e-05,
      "loss": 0.5349,
      "step": 2510
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.04964911565184593,
      "learning_rate": 7.066974595842957e-05,
      "loss": 1.471,
      "step": 2520
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.5833439826965332,
      "learning_rate": 7.015653066461381e-05,
      "loss": 0.3127,
      "step": 2530
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.6209262609481812,
      "learning_rate": 6.964331537079806e-05,
      "loss": 0.3781,
      "step": 2540
    },
    {
      "epoch": 0.65,
      "grad_norm": 34.23354721069336,
      "learning_rate": 6.91301000769823e-05,
      "loss": 1.8894,
      "step": 2550
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.40863272547721863,
      "learning_rate": 6.861688478316654e-05,
      "loss": 0.8985,
      "step": 2560
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.26213815808296204,
      "learning_rate": 6.810366948935078e-05,
      "loss": 1.038,
      "step": 2570
    },
    {
      "epoch": 0.66,
      "grad_norm": 32.263370513916016,
      "learning_rate": 6.759045419553503e-05,
      "loss": 2.2679,
      "step": 2580
    },
    {
      "epoch": 0.66,
      "grad_norm": 34.00870895385742,
      "learning_rate": 6.707723890171927e-05,
      "loss": 2.075,
      "step": 2590
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.4284890294075012,
      "learning_rate": 6.656402360790351e-05,
      "loss": 0.7597,
      "step": 2600
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.3407612144947052,
      "learning_rate": 6.605080831408776e-05,
      "loss": 0.4383,
      "step": 2610
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.467129111289978,
      "learning_rate": 6.5537593020272e-05,
      "loss": 1.2687,
      "step": 2620
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.13389898836612701,
      "learning_rate": 6.502437772645625e-05,
      "loss": 2.1633,
      "step": 2630
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.13675318658351898,
      "learning_rate": 6.451116243264049e-05,
      "loss": 0.7836,
      "step": 2640
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.5350820422172546,
      "learning_rate": 6.399794713882474e-05,
      "loss": 0.7649,
      "step": 2650
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.48448631167411804,
      "learning_rate": 6.348473184500898e-05,
      "loss": 1.38,
      "step": 2660
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.5778862237930298,
      "learning_rate": 6.297151655119323e-05,
      "loss": 2.2181,
      "step": 2670
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.1682463884353638,
      "learning_rate": 6.245830125737747e-05,
      "loss": 1.0876,
      "step": 2680
    },
    {
      "epoch": 0.69,
      "grad_norm": 31.393110275268555,
      "learning_rate": 6.194508596356173e-05,
      "loss": 1.1128,
      "step": 2690
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.6889828443527222,
      "learning_rate": 6.143187066974596e-05,
      "loss": 0.7252,
      "step": 2700
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.5311009883880615,
      "learning_rate": 6.0918655375930204e-05,
      "loss": 0.3959,
      "step": 2710
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.43320441246032715,
      "learning_rate": 6.040544008211445e-05,
      "loss": 1.2466,
      "step": 2720
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.06079118326306343,
      "learning_rate": 5.989222478829869e-05,
      "loss": 2.0125,
      "step": 2730
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.042944520711898804,
      "learning_rate": 5.937900949448294e-05,
      "loss": 0.779,
      "step": 2740
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.9245113730430603,
      "learning_rate": 5.886579420066718e-05,
      "loss": 0.7976,
      "step": 2750
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.039791032671928406,
      "learning_rate": 5.835257890685143e-05,
      "loss": 1.4881,
      "step": 2760
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.6234956979751587,
      "learning_rate": 5.783936361303567e-05,
      "loss": 0.6912,
      "step": 2770
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.9020193219184875,
      "learning_rate": 5.7326148319219916e-05,
      "loss": 0.7522,
      "step": 2780
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.034982066601514816,
      "learning_rate": 5.681293302540416e-05,
      "loss": 1.2431,
      "step": 2790
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.04073556512594223,
      "learning_rate": 5.6299717731588407e-05,
      "loss": 1.1053,
      "step": 2800
    },
    {
      "epoch": 0.72,
      "grad_norm": 35.25714874267578,
      "learning_rate": 5.5786502437772645e-05,
      "loss": 1.4825,
      "step": 2810
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.7252044081687927,
      "learning_rate": 5.52732871439569e-05,
      "loss": 2.1285,
      "step": 2820
    },
    {
      "epoch": 0.73,
      "grad_norm": 31.510704040527344,
      "learning_rate": 5.4760071850141136e-05,
      "loss": 0.7914,
      "step": 2830
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.44585466384887695,
      "learning_rate": 5.4246856556325375e-05,
      "loss": 1.2094,
      "step": 2840
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.2099919468164444,
      "learning_rate": 5.373364126250963e-05,
      "loss": 2.496,
      "step": 2850
    },
    {
      "epoch": 0.73,
      "grad_norm": 30.501766204833984,
      "learning_rate": 5.3220425968693866e-05,
      "loss": 1.0997,
      "step": 2860
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.36132165789604187,
      "learning_rate": 5.270721067487812e-05,
      "loss": 1.4378,
      "step": 2870
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.6164754629135132,
      "learning_rate": 5.219399538106236e-05,
      "loss": 1.1191,
      "step": 2880
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.9825451374053955,
      "learning_rate": 5.16807800872466e-05,
      "loss": 0.861,
      "step": 2890
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.28137707710266113,
      "learning_rate": 5.116756479343084e-05,
      "loss": 0.4387,
      "step": 2900
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.2143826186656952,
      "learning_rate": 5.0654349499615094e-05,
      "loss": 0.9561,
      "step": 2910
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.25483042001724243,
      "learning_rate": 5.014113420579933e-05,
      "loss": 1.4271,
      "step": 2920
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.36998990178108215,
      "learning_rate": 4.962791891198358e-05,
      "loss": 0.8621,
      "step": 2930
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.08702105283737183,
      "learning_rate": 4.9114703618167824e-05,
      "loss": 0.4751,
      "step": 2940
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.455074667930603,
      "learning_rate": 4.860148832435207e-05,
      "loss": 2.7804,
      "step": 2950
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.41690900921821594,
      "learning_rate": 4.8088273030536315e-05,
      "loss": 0.444,
      "step": 2960
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.0824534222483635,
      "learning_rate": 4.757505773672056e-05,
      "loss": 2.2328,
      "step": 2970
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.7097424268722534,
      "learning_rate": 4.70618424429048e-05,
      "loss": 1.232,
      "step": 2980
    },
    {
      "epoch": 0.77,
      "grad_norm": 32.93928527832031,
      "learning_rate": 4.6548627149089045e-05,
      "loss": 2.1723,
      "step": 2990
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.5564465522766113,
      "learning_rate": 4.603541185527329e-05,
      "loss": 1.3908,
      "step": 3000
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.1192642450332642,
      "learning_rate": 4.5522196561457536e-05,
      "loss": 0.2986,
      "step": 3010
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.09087738394737244,
      "learning_rate": 4.5008981267641774e-05,
      "loss": 1.3239,
      "step": 3020
    },
    {
      "epoch": 0.78,
      "grad_norm": 31.66700553894043,
      "learning_rate": 4.449576597382602e-05,
      "loss": 1.9121,
      "step": 3030
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.08067402243614197,
      "learning_rate": 4.3982550680010266e-05,
      "loss": 1.4794,
      "step": 3040
    },
    {
      "epoch": 0.78,
      "grad_norm": 31.138818740844727,
      "learning_rate": 4.346933538619451e-05,
      "loss": 1.3591,
      "step": 3050
    },
    {
      "epoch": 0.79,
      "grad_norm": 30.739635467529297,
      "learning_rate": 4.2956120092378757e-05,
      "loss": 2.1444,
      "step": 3060
    },
    {
      "epoch": 0.79,
      "grad_norm": 29.64994239807129,
      "learning_rate": 4.2442904798563e-05,
      "loss": 1.1349,
      "step": 3070
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.06835217773914337,
      "learning_rate": 4.192968950474724e-05,
      "loss": 1.1995,
      "step": 3080
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.07994609326124191,
      "learning_rate": 4.1416474210931486e-05,
      "loss": 1.106,
      "step": 3090
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.077377088367939,
      "learning_rate": 4.090325891711573e-05,
      "loss": 0.5515,
      "step": 3100
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.07298921793699265,
      "learning_rate": 4.039004362329998e-05,
      "loss": 1.3587,
      "step": 3110
    },
    {
      "epoch": 0.8,
      "grad_norm": 2.3832900524139404,
      "learning_rate": 3.987682832948422e-05,
      "loss": 0.5702,
      "step": 3120
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.3254412412643433,
      "learning_rate": 3.936361303566847e-05,
      "loss": 0.6837,
      "step": 3130
    },
    {
      "epoch": 0.81,
      "grad_norm": 31.57373046875,
      "learning_rate": 3.885039774185271e-05,
      "loss": 1.4529,
      "step": 3140
    },
    {
      "epoch": 0.81,
      "grad_norm": 31.78739356994629,
      "learning_rate": 3.833718244803695e-05,
      "loss": 1.1592,
      "step": 3150
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.885814368724823,
      "learning_rate": 3.78239671542212e-05,
      "loss": 2.0692,
      "step": 3160
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.11006342619657516,
      "learning_rate": 3.7310751860405444e-05,
      "loss": 1.3088,
      "step": 3170
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.7606900930404663,
      "learning_rate": 3.679753656658968e-05,
      "loss": 0.7503,
      "step": 3180
    },
    {
      "epoch": 0.82,
      "grad_norm": 34.01801681518555,
      "learning_rate": 3.628432127277393e-05,
      "loss": 2.3927,
      "step": 3190
    },
    {
      "epoch": 0.82,
      "grad_norm": 31.359045028686523,
      "learning_rate": 3.5771105978958174e-05,
      "loss": 1.7152,
      "step": 3200
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.4896736741065979,
      "learning_rate": 3.525789068514242e-05,
      "loss": 1.1745,
      "step": 3210
    },
    {
      "epoch": 0.83,
      "grad_norm": 33.61710739135742,
      "learning_rate": 3.4744675391326665e-05,
      "loss": 2.6372,
      "step": 3220
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.2002750039100647,
      "learning_rate": 3.423146009751091e-05,
      "loss": 0.3958,
      "step": 3230
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.5248939394950867,
      "learning_rate": 3.3718244803695156e-05,
      "loss": 1.1428,
      "step": 3240
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.20893482863903046,
      "learning_rate": 3.32050295098794e-05,
      "loss": 1.1695,
      "step": 3250
    },
    {
      "epoch": 0.84,
      "grad_norm": 31.339996337890625,
      "learning_rate": 3.269181421606364e-05,
      "loss": 1.9207,
      "step": 3260
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.20348432660102844,
      "learning_rate": 3.2178598922247886e-05,
      "loss": 1.587,
      "step": 3270
    },
    {
      "epoch": 0.84,
      "grad_norm": 31.108985900878906,
      "learning_rate": 3.1665383628432124e-05,
      "loss": 1.3677,
      "step": 3280
    },
    {
      "epoch": 0.84,
      "grad_norm": 29.706878662109375,
      "learning_rate": 3.115216833461637e-05,
      "loss": 2.0426,
      "step": 3290
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.0215935707092285,
      "learning_rate": 3.0638953040800616e-05,
      "loss": 1.7041,
      "step": 3300
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.23811456561088562,
      "learning_rate": 3.012573774698486e-05,
      "loss": 1.2077,
      "step": 3310
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.21963796019554138,
      "learning_rate": 2.9612522453169107e-05,
      "loss": 1.1672,
      "step": 3320
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.149311900138855,
      "learning_rate": 2.9099307159353352e-05,
      "loss": 1.6651,
      "step": 3330
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.1534388065338135,
      "learning_rate": 2.8586091865537594e-05,
      "loss": 1.2245,
      "step": 3340
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.3270301818847656,
      "learning_rate": 2.807287657172184e-05,
      "loss": 0.6268,
      "step": 3350
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.2713325619697571,
      "learning_rate": 2.7559661277906085e-05,
      "loss": 1.6962,
      "step": 3360
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.25903043150901794,
      "learning_rate": 2.704644598409033e-05,
      "loss": 0.4256,
      "step": 3370
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.456488698720932,
      "learning_rate": 2.6533230690274573e-05,
      "loss": 0.7452,
      "step": 3380
    },
    {
      "epoch": 0.87,
      "grad_norm": 30.572643280029297,
      "learning_rate": 2.6020015396458812e-05,
      "loss": 1.9218,
      "step": 3390
    },
    {
      "epoch": 0.87,
      "grad_norm": 31.96410369873047,
      "learning_rate": 2.5506800102643057e-05,
      "loss": 2.2211,
      "step": 3400
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.5312764644622803,
      "learning_rate": 2.4993584808827306e-05,
      "loss": 0.6983,
      "step": 3410
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.23434561491012573,
      "learning_rate": 2.448036951501155e-05,
      "loss": 0.7466,
      "step": 3420
    },
    {
      "epoch": 0.88,
      "grad_norm": 36.59482192993164,
      "learning_rate": 2.3967154221195794e-05,
      "loss": 1.6694,
      "step": 3430
    },
    {
      "epoch": 0.88,
      "grad_norm": 32.04636001586914,
      "learning_rate": 2.3453938927380036e-05,
      "loss": 1.6496,
      "step": 3440
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.2819555699825287,
      "learning_rate": 2.294072363356428e-05,
      "loss": 1.3424,
      "step": 3450
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.46713873744010925,
      "learning_rate": 2.2427508339748527e-05,
      "loss": 0.0149,
      "step": 3460
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.5181002020835876,
      "learning_rate": 2.1914293045932773e-05,
      "loss": 0.7906,
      "step": 3470
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.20827196538448334,
      "learning_rate": 2.140107775211701e-05,
      "loss": 0.4505,
      "step": 3480
    },
    {
      "epoch": 0.9,
      "grad_norm": 32.157432556152344,
      "learning_rate": 2.0887862458301257e-05,
      "loss": 2.5997,
      "step": 3490
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.1746426224708557,
      "learning_rate": 2.0374647164485502e-05,
      "loss": 1.6267,
      "step": 3500
    },
    {
      "epoch": 0.9,
      "grad_norm": 31.728593826293945,
      "learning_rate": 1.9861431870669748e-05,
      "loss": 0.8132,
      "step": 3510
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.8428208231925964,
      "learning_rate": 1.9348216576853994e-05,
      "loss": 1.4408,
      "step": 3520
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.6710448861122131,
      "learning_rate": 1.8835001283038236e-05,
      "loss": 0.7472,
      "step": 3530
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.15114788711071014,
      "learning_rate": 1.8321785989222478e-05,
      "loss": 0.3549,
      "step": 3540
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.48726460337638855,
      "learning_rate": 1.7808570695406723e-05,
      "loss": 1.0424,
      "step": 3550
    },
    {
      "epoch": 0.91,
      "grad_norm": 31.90473747253418,
      "learning_rate": 1.729535540159097e-05,
      "loss": 1.6014,
      "step": 3560
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.11974696815013885,
      "learning_rate": 1.678214010777521e-05,
      "loss": 1.367,
      "step": 3570
    },
    {
      "epoch": 0.92,
      "grad_norm": 31.7264461517334,
      "learning_rate": 1.6268924813959457e-05,
      "loss": 0.8339,
      "step": 3580
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.9428460001945496,
      "learning_rate": 1.5755709520143702e-05,
      "loss": 1.2831,
      "step": 3590
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.5526345372200012,
      "learning_rate": 1.5242494226327944e-05,
      "loss": 0.7811,
      "step": 3600
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.4995884299278259,
      "learning_rate": 1.4729278932512188e-05,
      "loss": 0.454,
      "step": 3610
    },
    {
      "epoch": 0.93,
      "grad_norm": 32.1862907409668,
      "learning_rate": 1.4216063638696434e-05,
      "loss": 1.7057,
      "step": 3620
    },
    {
      "epoch": 0.93,
      "grad_norm": 30.95826530456543,
      "learning_rate": 1.3702848344880677e-05,
      "loss": 1.6207,
      "step": 3630
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.10005848109722137,
      "learning_rate": 1.3189633051064923e-05,
      "loss": 0.8695,
      "step": 3640
    },
    {
      "epoch": 0.94,
      "grad_norm": 31.5998477935791,
      "learning_rate": 1.2676417757249169e-05,
      "loss": 1.9645,
      "step": 3650
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.797754168510437,
      "learning_rate": 1.216320246343341e-05,
      "loss": 1.1224,
      "step": 3660
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.10280079394578934,
      "learning_rate": 1.1649987169617656e-05,
      "loss": 0.8117,
      "step": 3670
    },
    {
      "epoch": 0.94,
      "grad_norm": 31.812061309814453,
      "learning_rate": 1.1136771875801898e-05,
      "loss": 0.5172,
      "step": 3680
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.08260977268218994,
      "learning_rate": 1.0623556581986144e-05,
      "loss": 1.6816,
      "step": 3690
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.6431272029876709,
      "learning_rate": 1.0110341288170388e-05,
      "loss": 2.2201,
      "step": 3700
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.5408867597579956,
      "learning_rate": 9.597125994354632e-06,
      "loss": 0.9963,
      "step": 3710
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.6773453950881958,
      "learning_rate": 9.083910700538877e-06,
      "loss": 2.3436,
      "step": 3720
    },
    {
      "epoch": 0.96,
      "grad_norm": 33.85165786743164,
      "learning_rate": 8.570695406723121e-06,
      "loss": 1.229,
      "step": 3730
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.09623867273330688,
      "learning_rate": 8.057480112907365e-06,
      "loss": 0.9852,
      "step": 3740
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.6665917038917542,
      "learning_rate": 7.5442648190916095e-06,
      "loss": 0.9811,
      "step": 3750
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.1049143448472023,
      "learning_rate": 7.031049525275854e-06,
      "loss": 1.6639,
      "step": 3760
    },
    {
      "epoch": 0.97,
      "grad_norm": 1.028926968574524,
      "learning_rate": 6.517834231460097e-06,
      "loss": 0.6745,
      "step": 3770
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.5345853567123413,
      "learning_rate": 6.004618937644342e-06,
      "loss": 1.9351,
      "step": 3780
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.09490037709474564,
      "learning_rate": 5.4914036438285865e-06,
      "loss": 1.0257,
      "step": 3790
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.7808061838150024,
      "learning_rate": 4.978188350012831e-06,
      "loss": 1.6728,
      "step": 3800
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.4944613575935364,
      "learning_rate": 4.464973056197075e-06,
      "loss": 1.6118,
      "step": 3810
    },
    {
      "epoch": 0.98,
      "grad_norm": 1.0724005699157715,
      "learning_rate": 3.951757762381319e-06,
      "loss": 1.352,
      "step": 3820
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.8290554285049438,
      "learning_rate": 3.4385424685655636e-06,
      "loss": 2.4513,
      "step": 3830
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.532361626625061,
      "learning_rate": 2.9253271747498074e-06,
      "loss": 0.412,
      "step": 3840
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.1319778561592102,
      "learning_rate": 2.412111880934052e-06,
      "loss": 0.7912,
      "step": 3850
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.6048825979232788,
      "learning_rate": 1.8988965871182964e-06,
      "loss": 0.9813,
      "step": 3860
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.4283379018306732,
      "learning_rate": 1.3856812933025406e-06,
      "loss": 0.7374,
      "step": 3870
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.7439431548118591,
      "learning_rate": 8.724659994867848e-07,
      "loss": 0.3349,
      "step": 3880
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.49157199263572693,
      "learning_rate": 3.5925070567102904e-07,
      "loss": 1.4183,
      "step": 3890
    },
    {
      "epoch": 1.0,
      "step": 3897,
      "total_flos": 9.121574551619174e+17,
      "train_loss": 1.2556480791153037,
      "train_runtime": 633.8251,
      "train_samples_per_second": 18.444,
      "train_steps_per_second": 6.148
    }
  ],
  "logging_steps": 10,
  "max_steps": 3897,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 1000,
  "total_flos": 9.121574551619174e+17,
  "train_batch_size": 3,
  "trial_name": null,
  "trial_params": null
}