{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 29.734513274336283,
  "global_step": 1680,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.18,
      "learning_rate": 0.000999929189777604,
      "loss": 4.7362,
      "step": 10
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0009996844394494005,
      "loss": 4.2612,
      "step": 20
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0009992649603785806,
      "loss": 4.2216,
      "step": 30
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0009986708992478202,
      "loss": 4.1602,
      "step": 40
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0009979024637873308,
      "loss": 4.0982,
      "step": 50
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0009969599227022219,
      "loss": 4.0312,
      "step": 60
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.000995843605578539,
      "loss": 3.7638,
      "step": 70
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0009945539027680166,
      "loss": 3.7872,
      "step": 80
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0009930912652515787,
      "loss": 3.7943,
      "step": 90
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0009914562044816423,
      "loss": 3.753,
      "step": 100
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0009900243933150229,
      "loss": 3.7757,
      "step": 110
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.000988080451875917,
      "loss": 3.4788,
      "step": 120
    },
    {
      "epoch": 2.3,
      "learning_rate": 0.0009859658393553863,
      "loss": 3.2514,
      "step": 130
    },
    {
      "epoch": 2.48,
      "learning_rate": 0.0009836812951872773,
      "loss": 3.3447,
      "step": 140
    },
    {
      "epoch": 2.65,
      "learning_rate": 0.0009812276182268236,
      "loss": 3.4213,
      "step": 150
    },
    {
      "epoch": 2.83,
      "learning_rate": 0.000978605666471306,
      "loss": 3.3397,
      "step": 160
    },
    {
      "epoch": 3.01,
      "learning_rate": 0.0009758163567600282,
      "loss": 3.3895,
      "step": 170
    },
    {
      "epoch": 3.19,
      "learning_rate": 0.0009728606644537177,
      "loss": 2.8609,
      "step": 180
    },
    {
      "epoch": 3.36,
      "learning_rate": 0.0009697396230934648,
      "loss": 2.775,
      "step": 190
    },
    {
      "epoch": 3.54,
      "learning_rate": 0.0009664543240393149,
      "loss": 2.9453,
      "step": 200
    },
    {
      "epoch": 3.72,
      "learning_rate": 0.0009630059160886439,
      "loss": 2.9014,
      "step": 210
    },
    {
      "epoch": 3.89,
      "learning_rate": 0.0009593956050744492,
      "loss": 3.0505,
      "step": 220
    },
    {
      "epoch": 4.07,
      "learning_rate": 0.0009556246534436953,
      "loss": 2.6717,
      "step": 230
    },
    {
      "epoch": 4.25,
      "learning_rate": 0.0009516943798158648,
      "loss": 2.5435,
      "step": 240
    },
    {
      "epoch": 4.42,
      "learning_rate": 0.0009476061585218652,
      "loss": 2.4999,
      "step": 250
    },
    {
      "epoch": 4.6,
      "learning_rate": 0.0009433614191234554,
      "loss": 2.4868,
      "step": 260
    },
    {
      "epoch": 4.78,
      "learning_rate": 0.0009389616459133597,
      "loss": 2.4924,
      "step": 270
    },
    {
      "epoch": 4.96,
      "learning_rate": 0.0009344083773962423,
      "loss": 2.6032,
      "step": 280
    },
    {
      "epoch": 5.13,
      "learning_rate": 0.0009297032057507264,
      "loss": 2.1659,
      "step": 290
    },
    {
      "epoch": 5.31,
      "learning_rate": 0.0009248477762726437,
      "loss": 2.222,
      "step": 300
    },
    {
      "epoch": 5.49,
      "learning_rate": 0.0009198437867997105,
      "loss": 2.0641,
      "step": 310
    },
    {
      "epoch": 5.66,
      "learning_rate": 0.0009146929871178311,
      "loss": 2.2005,
      "step": 320
    },
    {
      "epoch": 5.84,
      "learning_rate": 0.0009093971783492354,
      "loss": 2.2747,
      "step": 330
    },
    {
      "epoch": 6.02,
      "learning_rate": 0.000903958212322666,
      "loss": 2.2078,
      "step": 340
    },
    {
      "epoch": 6.19,
      "learning_rate": 0.0008983779909258347,
      "loss": 1.7158,
      "step": 350
    },
    {
      "epoch": 6.37,
      "learning_rate": 0.0008926584654403724,
      "loss": 1.7397,
      "step": 360
    },
    {
      "epoch": 6.55,
      "learning_rate": 0.0008868016358595103,
      "loss": 2.0346,
      "step": 370
    },
    {
      "epoch": 6.73,
      "learning_rate": 0.0008808095501887243,
      "loss": 1.8981,
      "step": 380
    },
    {
      "epoch": 6.9,
      "learning_rate": 0.0008746843037295936,
      "loss": 1.9037,
      "step": 390
    },
    {
      "epoch": 7.08,
      "learning_rate": 0.0008684280383471181,
      "loss": 1.7414,
      "step": 400
    },
    {
      "epoch": 7.26,
      "learning_rate": 0.0008620429417207562,
      "loss": 1.5215,
      "step": 410
    },
    {
      "epoch": 7.43,
      "learning_rate": 0.0008555312465794402,
      "loss": 1.5522,
      "step": 420
    },
    {
      "epoch": 7.61,
      "learning_rate": 0.0008488952299208401,
      "loss": 1.5898,
      "step": 430
    },
    {
      "epoch": 7.79,
      "learning_rate": 0.0008421372122151469,
      "loss": 1.692,
      "step": 440
    },
    {
      "epoch": 7.96,
      "learning_rate": 0.0008352595565936554,
      "loss": 1.6656,
      "step": 450
    },
    {
      "epoch": 8.14,
      "learning_rate": 0.0008282646680224282,
      "loss": 1.3681,
      "step": 460
    },
    {
      "epoch": 8.32,
      "learning_rate": 0.0008211549924613316,
      "loss": 1.323,
      "step": 470
    },
    {
      "epoch": 8.5,
      "learning_rate": 0.0008139330160087374,
      "loss": 1.301,
      "step": 480
    },
    {
      "epoch": 8.67,
      "learning_rate": 0.0008066012640321876,
      "loss": 1.4134,
      "step": 490
    },
    {
      "epoch": 8.85,
      "learning_rate": 0.0007991623002853296,
      "loss": 1.4661,
      "step": 500
    },
    {
      "epoch": 9.03,
      "learning_rate": 0.0007916187260114262,
      "loss": 1.4292,
      "step": 510
    },
    {
      "epoch": 9.2,
      "learning_rate": 0.0007839731790337593,
      "loss": 0.8971,
      "step": 520
    },
    {
      "epoch": 9.38,
      "learning_rate": 0.0007762283328332388,
      "loss": 1.2906,
      "step": 530
    },
    {
      "epoch": 9.56,
      "learning_rate": 0.000768386895613546,
      "loss": 1.2389,
      "step": 540
    },
    {
      "epoch": 9.73,
      "learning_rate": 0.0007604516093541333,
      "loss": 1.2767,
      "step": 550
    },
    {
      "epoch": 9.91,
      "learning_rate": 0.0007524252488514144,
      "loss": 1.1868,
      "step": 560
    },
    {
      "epoch": 10.09,
      "learning_rate": 0.0007443106207484776,
      "loss": 1.0139,
      "step": 570
    },
    {
      "epoch": 10.27,
      "learning_rate": 0.0007361105625536646,
      "loss": 1.021,
      "step": 580
    },
    {
      "epoch": 10.44,
      "learning_rate": 0.0007278279416483561,
      "loss": 0.9439,
      "step": 590
    },
    {
      "epoch": 10.62,
      "learning_rate": 0.0007194656542843102,
      "loss": 0.9772,
      "step": 600
    },
    {
      "epoch": 10.8,
      "learning_rate": 0.0007110266245709072,
      "loss": 1.0697,
      "step": 610
    },
    {
      "epoch": 10.97,
      "learning_rate": 0.0007025138034526517,
      "loss": 1.1471,
      "step": 620
    },
    {
      "epoch": 11.15,
      "learning_rate": 0.0006939301676772927,
      "loss": 0.8037,
      "step": 630
    },
    {
      "epoch": 11.33,
      "learning_rate": 0.0006852787187549182,
      "loss": 0.8488,
      "step": 640
    },
    {
      "epoch": 11.5,
      "learning_rate": 0.0006765624819083944,
      "loss": 0.8035,
      "step": 650
    },
    {
      "epoch": 11.68,
      "learning_rate": 0.0006677845050155106,
      "loss": 0.8825,
      "step": 660
    },
    {
      "epoch": 11.86,
      "learning_rate": 0.0006589478575432023,
      "loss": 0.8639,
      "step": 670
    },
    {
      "epoch": 12.04,
      "learning_rate": 0.0006500556294742265,
      "loss": 0.9365,
      "step": 680
    },
    {
      "epoch": 12.21,
      "learning_rate": 0.0006411109302266615,
      "loss": 0.704,
      "step": 690
    },
    {
      "epoch": 12.39,
      "learning_rate": 0.0006321168875666118,
      "loss": 0.7477,
      "step": 700
    },
    {
      "epoch": 12.57,
      "learning_rate": 0.0006230766465144965,
      "loss": 0.6058,
      "step": 710
    },
    {
      "epoch": 12.74,
      "learning_rate": 0.0006139933682453035,
      "loss": 0.809,
      "step": 720
    },
    {
      "epoch": 12.92,
      "learning_rate": 0.0006048702289831964,
      "loss": 0.8475,
      "step": 730
    },
    {
      "epoch": 13.1,
      "learning_rate": 0.0005957104188908586,
      "loss": 0.6654,
      "step": 740
    },
    {
      "epoch": 13.27,
      "learning_rate": 0.0005865171409539613,
      "loss": 0.6185,
      "step": 750
    },
    {
      "epoch": 13.45,
      "learning_rate": 0.0005772936098611519,
      "loss": 0.4949,
      "step": 760
    },
    {
      "epoch": 13.63,
      "learning_rate": 0.0005680430508799453,
      "loss": 0.5332,
      "step": 770
    },
    {
      "epoch": 13.81,
      "learning_rate": 0.0005587686987289189,
      "loss": 0.6924,
      "step": 780
    },
    {
      "epoch": 13.98,
      "learning_rate": 0.0005494737964466028,
      "loss": 0.6796,
      "step": 790
    },
    {
      "epoch": 14.16,
      "learning_rate": 0.0005401615942574595,
      "loss": 0.5147,
      "step": 800
    },
    {
      "epoch": 14.34,
      "learning_rate": 0.0005308353484353508,
      "loss": 0.4158,
      "step": 810
    },
    {
      "epoch": 14.51,
      "learning_rate": 0.0005214983201648908,
      "loss": 0.5493,
      "step": 820
    },
    {
      "epoch": 14.69,
      "learning_rate": 0.0005121537744010808,
      "loss": 0.5508,
      "step": 830
    },
    {
      "epoch": 14.87,
      "learning_rate": 0.0005028049787276249,
      "loss": 0.513,
      "step": 840
    },
    {
      "epoch": 15.04,
      "learning_rate": 0.0004934552022143279,
      "loss": 0.5027,
      "step": 850
    },
    {
      "epoch": 15.22,
      "learning_rate": 0.00048410771427397304,
      "loss": 0.3954,
      "step": 860
    },
    {
      "epoch": 15.4,
      "learning_rate": 0.0004747657835190795,
      "loss": 0.4376,
      "step": 870
    },
    {
      "epoch": 15.58,
      "learning_rate": 0.0004654326766189404,
      "loss": 0.4182,
      "step": 880
    },
    {
      "epoch": 15.75,
      "learning_rate": 0.0004561116571573388,
      "loss": 0.4889,
      "step": 890
    },
    {
      "epoch": 15.93,
      "learning_rate": 0.00044680598449134434,
      "loss": 0.3596,
      "step": 900
    },
    {
      "epoch": 16.11,
      "learning_rate": 0.00043751891261158783,
      "loss": 0.3129,
      "step": 910
    },
    {
      "epoch": 16.28,
      "learning_rate": 0.0004282536890044104,
      "loss": 0.3088,
      "step": 920
    },
    {
      "epoch": 16.46,
      "learning_rate": 0.0004190135535162894,
      "loss": 0.3212,
      "step": 930
    },
    {
      "epoch": 16.64,
      "learning_rate": 0.00040980173722093496,
      "loss": 0.4326,
      "step": 940
    },
    {
      "epoch": 16.81,
      "learning_rate": 0.00040062146128945275,
      "loss": 0.3264,
      "step": 950
    },
    {
      "epoch": 16.99,
      "learning_rate": 0.0003914759358639719,
      "loss": 0.3636,
      "step": 960
    },
    {
      "epoch": 17.17,
      "learning_rate": 0.0003823683589351284,
      "loss": 0.2175,
      "step": 970
    },
    {
      "epoch": 17.35,
      "learning_rate": 0.0003733019152237965,
      "loss": 0.208,
      "step": 980
    },
    {
      "epoch": 17.52,
      "learning_rate": 0.0003642797750674629,
      "loss": 0.2717,
      "step": 990
    },
    {
      "epoch": 17.7,
      "learning_rate": 0.0003553050933116273,
      "loss": 0.329,
      "step": 1000
    },
    {
      "epoch": 17.88,
      "learning_rate": 0.0003463810082066231,
      "loss": 0.33,
      "step": 1010
    },
    {
      "epoch": 18.05,
      "learning_rate": 0.00033751064031023887,
      "loss": 0.2679,
      "step": 1020
    },
    {
      "epoch": 18.23,
      "learning_rate": 0.0003286970913965275,
      "loss": 0.224,
      "step": 1030
    },
    {
      "epoch": 18.41,
      "learning_rate": 0.00031994344337118454,
      "loss": 0.2048,
      "step": 1040
    },
    {
      "epoch": 18.58,
      "learning_rate": 0.0003112527571938717,
      "loss": 0.2078,
      "step": 1050
    },
    {
      "epoch": 18.76,
      "learning_rate": 0.00030262807180786645,
      "loss": 0.2303,
      "step": 1060
    },
    {
      "epoch": 18.94,
      "learning_rate": 0.0002940724030774104,
      "loss": 0.2142,
      "step": 1070
    },
    {
      "epoch": 19.12,
      "learning_rate": 0.0002855887427331267,
      "loss": 0.2067,
      "step": 1080
    },
    {
      "epoch": 19.29,
      "learning_rate": 0.0002771800573258781,
      "loss": 0.1458,
      "step": 1090
    },
    {
      "epoch": 19.47,
      "learning_rate": 0.0002688492871894276,
      "loss": 0.1774,
      "step": 1100
    },
    {
      "epoch": 19.65,
      "learning_rate": 0.0002605993454122687,
      "loss": 0.1741,
      "step": 1110
    },
    {
      "epoch": 19.82,
      "learning_rate": 0.0002524331168189817,
      "loss": 0.1883,
      "step": 1120
    },
    {
      "epoch": 20.0,
      "learning_rate": 0.00024435345696147403,
      "loss": 0.1689,
      "step": 1130
    },
    {
      "epoch": 20.18,
      "learning_rate": 0.00023636319112045495,
      "loss": 0.1238,
      "step": 1140
    },
    {
      "epoch": 20.35,
      "learning_rate": 0.00022846511331749792,
      "loss": 0.1372,
      "step": 1150
    },
    {
      "epoch": 20.53,
      "learning_rate": 0.00022066198533803106,
      "loss": 0.1298,
      "step": 1160
    },
    {
      "epoch": 20.71,
      "learning_rate": 0.00021295653576560165,
      "loss": 0.1445,
      "step": 1170
    },
    {
      "epoch": 20.88,
      "learning_rate": 0.00020535145902774978,
      "loss": 0.1454,
      "step": 1180
    },
    {
      "epoch": 21.06,
      "learning_rate": 0.00019784941445382642,
      "loss": 0.1417,
      "step": 1190
    },
    {
      "epoch": 21.24,
      "learning_rate": 0.00019045302534508295,
      "loss": 0.1099,
      "step": 1200
    },
    {
      "epoch": 21.42,
      "learning_rate": 0.0001831648780573612,
      "loss": 0.0922,
      "step": 1210
    },
    {
      "epoch": 21.59,
      "learning_rate": 0.00017598752109670014,
      "loss": 0.1258,
      "step": 1220
    },
    {
      "epoch": 21.77,
      "learning_rate": 0.00016892346422817944,
      "loss": 0.123,
      "step": 1230
    },
    {
      "epoch": 21.95,
      "learning_rate": 0.00016197517759830948,
      "loss": 0.1047,
      "step": 1240
    },
    {
      "epoch": 22.12,
      "learning_rate": 0.00015514509087127522,
      "loss": 0.0991,
      "step": 1250
    },
    {
      "epoch": 22.3,
      "learning_rate": 0.00014843559237933475,
      "loss": 0.1057,
      "step": 1260
    },
    {
      "epoch": 22.48,
      "learning_rate": 0.00014184902828767287,
      "loss": 0.0796,
      "step": 1270
    },
    {
      "epoch": 22.65,
      "learning_rate": 0.0001353877017739974,
      "loss": 0.0747,
      "step": 1280
    },
    {
      "epoch": 22.83,
      "learning_rate": 0.00012905387222316822,
      "loss": 0.0807,
      "step": 1290
    },
    {
      "epoch": 23.01,
      "learning_rate": 0.00012284975443714076,
      "loss": 0.1032,
      "step": 1300
    },
    {
      "epoch": 23.19,
      "learning_rate": 0.00011677751786049661,
      "loss": 0.0743,
      "step": 1310
    },
    {
      "epoch": 23.36,
      "learning_rate": 0.0001108392858218371,
      "loss": 0.0823,
      "step": 1320
    },
    {
      "epoch": 23.54,
      "learning_rate": 0.00010503713479130123,
      "loss": 0.0758,
      "step": 1330
    },
    {
      "epoch": 23.72,
      "learning_rate": 9.937309365446973e-05,
      "loss": 0.0642,
      "step": 1340
    },
    {
      "epoch": 23.89,
      "learning_rate": 9.384914300290748e-05,
      "loss": 0.0791,
      "step": 1350
    },
    {
      "epoch": 24.07,
      "learning_rate": 8.846721444159461e-05,
      "loss": 0.0595,
      "step": 1360
    },
    {
      "epoch": 24.25,
      "learning_rate": 8.32291899134856e-05,
      "loss": 0.0676,
      "step": 1370
    },
    {
      "epoch": 24.42,
      "learning_rate": 7.813690104143555e-05,
      "loss": 0.0611,
      "step": 1380
    },
    {
      "epoch": 24.6,
      "learning_rate": 7.319212848772117e-05,
      "loss": 0.0642,
      "step": 1390
    },
    {
      "epoch": 24.78,
      "learning_rate": 6.83966013313826e-05,
      "loss": 0.0602,
      "step": 1400
    },
    {
      "epoch": 24.96,
      "learning_rate": 6.375199646360142e-05,
      "loss": 0.066,
      "step": 1410
    },
    {
      "epoch": 25.13,
      "learning_rate": 5.9259938001329336e-05,
      "loss": 0.0581,
      "step": 1420
    },
    {
      "epoch": 25.31,
      "learning_rate": 5.492199671936915e-05,
      "loss": 0.0527,
      "step": 1430
    },
    {
      "epoch": 25.49,
      "learning_rate": 5.073968950110941e-05,
      "loss": 0.0634,
      "step": 1440
    },
    {
      "epoch": 25.66,
      "learning_rate": 4.6714478808103314e-05,
      "loss": 0.0545,
      "step": 1450
    },
    {
      "epoch": 25.84,
      "learning_rate": 4.2847772168678246e-05,
      "loss": 0.0549,
      "step": 1460
    },
    {
      "epoch": 26.02,
      "learning_rate": 3.9140921685753064e-05,
      "loss": 0.0529,
      "step": 1470
    },
    {
      "epoch": 26.19,
      "learning_rate": 3.559522356403788e-05,
      "loss": 0.0558,
      "step": 1480
    },
    {
      "epoch": 26.37,
      "learning_rate": 3.221191765677928e-05,
      "loss": 0.0529,
      "step": 1490
    },
    {
      "epoch": 26.55,
      "learning_rate": 2.8992187032210516e-05,
      "loss": 0.0552,
      "step": 1500
    },
    {
      "epoch": 26.73,
      "learning_rate": 2.5937157559859093e-05,
      "loss": 0.0499,
      "step": 1510
    },
    {
      "epoch": 26.9,
      "learning_rate": 2.3047897516853767e-05,
      "loss": 0.0451,
      "step": 1520
    },
    {
      "epoch": 27.08,
      "learning_rate": 2.032541721437209e-05,
      "loss": 0.0511,
      "step": 1530
    },
    {
      "epoch": 27.26,
      "learning_rate": 1.777066864435628e-05,
      "loss": 0.0491,
      "step": 1540
    },
    {
      "epoch": 27.43,
      "learning_rate": 1.538454514662285e-05,
      "loss": 0.045,
      "step": 1550
    },
    {
      "epoch": 27.61,
      "learning_rate": 1.3167881096480372e-05,
      "loss": 0.0594,
      "step": 1560
    },
    {
      "epoch": 27.79,
      "learning_rate": 1.1121451612967626e-05,
      "loss": 0.0456,
      "step": 1570
    },
    {
      "epoch": 27.96,
      "learning_rate": 9.245972287810544e-06,
      "loss": 0.0444,
      "step": 1580
    },
    {
      "epoch": 28.14,
      "learning_rate": 7.542098935195918e-06,
      "loss": 0.0615,
      "step": 1590
    },
    {
      "epoch": 28.32,
      "learning_rate": 6.010427362447224e-06,
      "loss": 0.0456,
      "step": 1600
    },
    {
      "epoch": 28.5,
      "learning_rate": 4.6514931616840066e-06,
      "loss": 0.0445,
      "step": 1610
    },
    {
      "epoch": 28.67,
      "learning_rate": 3.4657715225368535e-06,
      "loss": 0.0494,
      "step": 1620
    },
    {
      "epoch": 28.85,
      "learning_rate": 2.453677065984039e-06,
      "loss": 0.0411,
      "step": 1630
    },
    {
      "epoch": 29.03,
      "learning_rate": 1.615563699367495e-06,
      "loss": 0.0558,
      "step": 1640
    },
    {
      "epoch": 29.2,
      "learning_rate": 9.517244926393609e-07,
      "loss": 0.0457,
      "step": 1650
    },
    {
      "epoch": 29.38,
      "learning_rate": 4.623915758816266e-07,
      "loss": 0.0524,
      "step": 1660
    },
    {
      "epoch": 29.56,
      "learning_rate": 1.4773605813522827e-07,
      "loss": 0.0454,
      "step": 1670
    },
    {
      "epoch": 29.73,
      "learning_rate": 7.867967567354306e-09,
      "loss": 0.055,
      "step": 1680
    },
    {
      "epoch": 29.73,
      "step": 1680,
      "total_flos": 2.133691717066752e+17,
      "train_loss": 1.0390962482385693,
      "train_runtime": 7981.5401,
      "train_samples_per_second": 3.39,
      "train_steps_per_second": 0.21
    }
  ],
  "max_steps": 1680,
  "num_train_epochs": 30,
  "total_flos": 2.133691717066752e+17,
  "trial_name": null,
  "trial_params": null
}