{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.9964104098115465,
  "eval_steps": 500,
  "global_step": 3340,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.02,
      "learning_rate": 1.1976047904191617e-05,
      "loss": 1.194,
      "step": 20
    },
    {
      "epoch": 0.05,
      "learning_rate": 2.3952095808383234e-05,
      "loss": 1.0484,
      "step": 40
    },
    {
      "epoch": 0.07,
      "learning_rate": 3.592814371257485e-05,
      "loss": 0.8359,
      "step": 60
    },
    {
      "epoch": 0.1,
      "learning_rate": 4.790419161676647e-05,
      "loss": 0.6286,
      "step": 80
    },
    {
      "epoch": 0.12,
      "learning_rate": 5.988023952095808e-05,
      "loss": 0.4686,
      "step": 100
    },
    {
      "epoch": 0.14,
      "learning_rate": 7.18562874251497e-05,
      "loss": 0.3526,
      "step": 120
    },
    {
      "epoch": 0.17,
      "learning_rate": 8.383233532934131e-05,
      "loss": 0.2483,
      "step": 140
    },
    {
      "epoch": 0.19,
      "learning_rate": 9.580838323353294e-05,
      "loss": 0.1862,
      "step": 160
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.00010778443113772456,
      "loss": 0.1524,
      "step": 180
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.00011976047904191617,
      "loss": 0.1389,
      "step": 200
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0001317365269461078,
      "loss": 0.1361,
      "step": 220
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0001437125748502994,
      "loss": 0.1286,
      "step": 240
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.00015568862275449103,
      "loss": 0.1239,
      "step": 260
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.00016766467065868263,
      "loss": 0.1205,
      "step": 280
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.00017964071856287425,
      "loss": 0.1206,
      "step": 300
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.00019161676646706587,
      "loss": 0.1166,
      "step": 320
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0001996007984031936,
      "loss": 0.1174,
      "step": 340
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.000198270126413839,
      "loss": 0.115,
      "step": 360
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.00019693945442448438,
      "loss": 0.1158,
      "step": 380
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.00019560878243512974,
      "loss": 0.1157,
      "step": 400
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.00019427811044577512,
      "loss": 0.114,
      "step": 420
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.00019294743845642048,
      "loss": 0.1115,
      "step": 440
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.00019161676646706587,
      "loss": 0.1152,
      "step": 460
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.00019028609447771126,
      "loss": 0.1107,
      "step": 480
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.00018895542248835662,
      "loss": 0.1117,
      "step": 500
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.000187624750499002,
      "loss": 0.1096,
      "step": 520
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.00018629407850964737,
      "loss": 0.1108,
      "step": 540
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.00018496340652029275,
      "loss": 0.1103,
      "step": 560
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.00018363273453093811,
      "loss": 0.106,
      "step": 580
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0001823020625415835,
      "loss": 0.1066,
      "step": 600
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0001809713905522289,
      "loss": 0.1078,
      "step": 620
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.00017964071856287425,
      "loss": 0.1074,
      "step": 640
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.00017831004657351964,
      "loss": 0.1047,
      "step": 660
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.000176979374584165,
      "loss": 0.1051,
      "step": 680
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.00017564870259481038,
      "loss": 0.1052,
      "step": 700
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.00017431803060545577,
      "loss": 0.1068,
      "step": 720
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.00017298735861610113,
      "loss": 0.1081,
      "step": 740
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.00017165668662674652,
      "loss": 0.1069,
      "step": 760
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.00017032601463739188,
      "loss": 0.1072,
      "step": 780
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.00016899534264803727,
      "loss": 0.1057,
      "step": 800
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.00016766467065868263,
      "loss": 0.1057,
      "step": 820
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.00016633399866932801,
      "loss": 0.1072,
      "step": 840
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0001650033266799734,
      "loss": 0.1054,
      "step": 860
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.00016367265469061876,
      "loss": 0.1064,
      "step": 880
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.00016234198270126415,
      "loss": 0.1069,
      "step": 900
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0001610113107119095,
      "loss": 0.1062,
      "step": 920
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0001596806387225549,
      "loss": 0.1033,
      "step": 940
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.00015834996673320028,
      "loss": 0.1022,
      "step": 960
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.00015701929474384565,
      "loss": 0.1023,
      "step": 980
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.00015568862275449103,
      "loss": 0.1024,
      "step": 1000
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0001543579507651364,
      "loss": 0.1008,
      "step": 1020
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.00015302727877578178,
      "loss": 0.1011,
      "step": 1040
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.00015169660678642714,
      "loss": 0.1035,
      "step": 1060
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.00015036593479707253,
      "loss": 0.1012,
      "step": 1080
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.00014903526280771792,
      "loss": 0.1004,
      "step": 1100
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.00014770459081836328,
      "loss": 0.0996,
      "step": 1120
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.00014637391882900866,
      "loss": 0.102,
      "step": 1140
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.00014504324683965402,
      "loss": 0.0976,
      "step": 1160
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0001437125748502994,
      "loss": 0.1024,
      "step": 1180
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0001423819028609448,
      "loss": 0.1006,
      "step": 1200
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.00014105123087159016,
      "loss": 0.101,
      "step": 1220
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.00013972055888223555,
      "loss": 0.102,
      "step": 1240
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0001383898868928809,
      "loss": 0.1018,
      "step": 1260
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0001370592149035263,
      "loss": 0.0999,
      "step": 1280
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.00013572854291417165,
      "loss": 0.1021,
      "step": 1300
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.00013439787092481704,
      "loss": 0.0996,
      "step": 1320
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.00013306719893546243,
      "loss": 0.1005,
      "step": 1340
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0001317365269461078,
      "loss": 0.0993,
      "step": 1360
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.00013040585495675318,
      "loss": 0.0999,
      "step": 1380
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.00012907518296739854,
      "loss": 0.1011,
      "step": 1400
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.00012774451097804392,
      "loss": 0.0968,
      "step": 1420
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0001264138389886893,
      "loss": 0.0983,
      "step": 1440
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.00012508316699933467,
      "loss": 0.0987,
      "step": 1460
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.00012375249500998006,
      "loss": 0.0984,
      "step": 1480
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.00012242182302062542,
      "loss": 0.0956,
      "step": 1500
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0001210911510312708,
      "loss": 0.0971,
      "step": 1520
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.00011976047904191617,
      "loss": 0.0986,
      "step": 1540
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.00011842980705256155,
      "loss": 0.0995,
      "step": 1560
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.00011709913506320693,
      "loss": 0.0989,
      "step": 1580
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0001157684630738523,
      "loss": 0.1006,
      "step": 1600
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.00011443779108449768,
      "loss": 0.0982,
      "step": 1620
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.00011310711909514305,
      "loss": 0.0982,
      "step": 1640
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.00011177644710578842,
      "loss": 0.0998,
      "step": 1660
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.00011044577511643381,
      "loss": 0.0986,
      "step": 1680
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.00010911510312707917,
      "loss": 0.0996,
      "step": 1700
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.00010778443113772456,
      "loss": 0.0994,
      "step": 1720
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.00010645375914836992,
      "loss": 0.0999,
      "step": 1740
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0001051230871590153,
      "loss": 0.1006,
      "step": 1760
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.00010379241516966068,
      "loss": 0.0954,
      "step": 1780
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.00010246174318030605,
      "loss": 0.0958,
      "step": 1800
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.00010113107119095144,
      "loss": 0.0969,
      "step": 1820
    },
    {
      "epoch": 2.2,
      "learning_rate": 9.98003992015968e-05,
      "loss": 0.0949,
      "step": 1840
    },
    {
      "epoch": 2.23,
      "learning_rate": 9.846972721224219e-05,
      "loss": 0.0957,
      "step": 1860
    },
    {
      "epoch": 2.25,
      "learning_rate": 9.713905522288756e-05,
      "loss": 0.0951,
      "step": 1880
    },
    {
      "epoch": 2.27,
      "learning_rate": 9.580838323353294e-05,
      "loss": 0.0966,
      "step": 1900
    },
    {
      "epoch": 2.3,
      "learning_rate": 9.447771124417831e-05,
      "loss": 0.0946,
      "step": 1920
    },
    {
      "epoch": 2.32,
      "learning_rate": 9.314703925482368e-05,
      "loss": 0.0947,
      "step": 1940
    },
    {
      "epoch": 2.35,
      "learning_rate": 9.181636726546906e-05,
      "loss": 0.0947,
      "step": 1960
    },
    {
      "epoch": 2.37,
      "learning_rate": 9.048569527611444e-05,
      "loss": 0.0961,
      "step": 1980
    },
    {
      "epoch": 2.39,
      "learning_rate": 8.915502328675982e-05,
      "loss": 0.0926,
      "step": 2000
    },
    {
      "epoch": 2.42,
      "learning_rate": 8.782435129740519e-05,
      "loss": 0.0972,
      "step": 2020
    },
    {
      "epoch": 2.44,
      "learning_rate": 8.649367930805057e-05,
      "loss": 0.095,
      "step": 2040
    },
    {
      "epoch": 2.46,
      "learning_rate": 8.516300731869594e-05,
      "loss": 0.0969,
      "step": 2060
    },
    {
      "epoch": 2.49,
      "learning_rate": 8.383233532934131e-05,
      "loss": 0.0964,
      "step": 2080
    },
    {
      "epoch": 2.51,
      "learning_rate": 8.25016633399867e-05,
      "loss": 0.0944,
      "step": 2100
    },
    {
      "epoch": 2.54,
      "learning_rate": 8.117099135063207e-05,
      "loss": 0.097,
      "step": 2120
    },
    {
      "epoch": 2.56,
      "learning_rate": 7.984031936127745e-05,
      "loss": 0.0955,
      "step": 2140
    },
    {
      "epoch": 2.58,
      "learning_rate": 7.850964737192282e-05,
      "loss": 0.0957,
      "step": 2160
    },
    {
      "epoch": 2.61,
      "learning_rate": 7.71789753825682e-05,
      "loss": 0.095,
      "step": 2180
    },
    {
      "epoch": 2.63,
      "learning_rate": 7.584830339321357e-05,
      "loss": 0.0941,
      "step": 2200
    },
    {
      "epoch": 2.66,
      "learning_rate": 7.451763140385896e-05,
      "loss": 0.096,
      "step": 2220
    },
    {
      "epoch": 2.68,
      "learning_rate": 7.318695941450433e-05,
      "loss": 0.0942,
      "step": 2240
    },
    {
      "epoch": 2.7,
      "learning_rate": 7.18562874251497e-05,
      "loss": 0.0936,
      "step": 2260
    },
    {
      "epoch": 2.73,
      "learning_rate": 7.052561543579508e-05,
      "loss": 0.0928,
      "step": 2280
    },
    {
      "epoch": 2.75,
      "learning_rate": 6.919494344644045e-05,
      "loss": 0.0948,
      "step": 2300
    },
    {
      "epoch": 2.78,
      "learning_rate": 6.786427145708583e-05,
      "loss": 0.0937,
      "step": 2320
    },
    {
      "epoch": 2.8,
      "learning_rate": 6.653359946773121e-05,
      "loss": 0.0902,
      "step": 2340
    },
    {
      "epoch": 2.82,
      "learning_rate": 6.520292747837659e-05,
      "loss": 0.093,
      "step": 2360
    },
    {
      "epoch": 2.85,
      "learning_rate": 6.387225548902196e-05,
      "loss": 0.0936,
      "step": 2380
    },
    {
      "epoch": 2.87,
      "learning_rate": 6.254158349966734e-05,
      "loss": 0.0958,
      "step": 2400
    },
    {
      "epoch": 2.9,
      "learning_rate": 6.121091151031271e-05,
      "loss": 0.0948,
      "step": 2420
    },
    {
      "epoch": 2.92,
      "learning_rate": 5.988023952095808e-05,
      "loss": 0.0965,
      "step": 2440
    },
    {
      "epoch": 2.94,
      "learning_rate": 5.8549567531603464e-05,
      "loss": 0.0932,
      "step": 2460
    },
    {
      "epoch": 2.97,
      "learning_rate": 5.721889554224884e-05,
      "loss": 0.0941,
      "step": 2480
    },
    {
      "epoch": 2.99,
      "learning_rate": 5.588822355289421e-05,
      "loss": 0.0957,
      "step": 2500
    },
    {
      "epoch": 3.02,
      "learning_rate": 5.4557551563539585e-05,
      "loss": 0.0943,
      "step": 2520
    },
    {
      "epoch": 3.04,
      "learning_rate": 5.322687957418496e-05,
      "loss": 0.0954,
      "step": 2540
    },
    {
      "epoch": 3.06,
      "learning_rate": 5.189620758483034e-05,
      "loss": 0.095,
      "step": 2560
    },
    {
      "epoch": 3.09,
      "learning_rate": 5.056553559547572e-05,
      "loss": 0.096,
      "step": 2580
    },
    {
      "epoch": 3.11,
      "learning_rate": 4.9234863606121094e-05,
      "loss": 0.0954,
      "step": 2600
    },
    {
      "epoch": 3.13,
      "learning_rate": 4.790419161676647e-05,
      "loss": 0.0913,
      "step": 2620
    },
    {
      "epoch": 3.16,
      "learning_rate": 4.657351962741184e-05,
      "loss": 0.092,
      "step": 2640
    },
    {
      "epoch": 3.18,
      "learning_rate": 4.524284763805722e-05,
      "loss": 0.0933,
      "step": 2660
    },
    {
      "epoch": 3.21,
      "learning_rate": 4.3912175648702596e-05,
      "loss": 0.0915,
      "step": 2680
    },
    {
      "epoch": 3.23,
      "learning_rate": 4.258150365934797e-05,
      "loss": 0.0903,
      "step": 2700
    },
    {
      "epoch": 3.25,
      "learning_rate": 4.125083166999335e-05,
      "loss": 0.0924,
      "step": 2720
    },
    {
      "epoch": 3.28,
      "learning_rate": 3.9920159680638724e-05,
      "loss": 0.0926,
      "step": 2740
    },
    {
      "epoch": 3.3,
      "learning_rate": 3.85894876912841e-05,
      "loss": 0.0923,
      "step": 2760
    },
    {
      "epoch": 3.33,
      "learning_rate": 3.725881570192948e-05,
      "loss": 0.0899,
      "step": 2780
    },
    {
      "epoch": 3.35,
      "learning_rate": 3.592814371257485e-05,
      "loss": 0.0922,
      "step": 2800
    },
    {
      "epoch": 3.37,
      "learning_rate": 3.4597471723220226e-05,
      "loss": 0.0905,
      "step": 2820
    },
    {
      "epoch": 3.4,
      "learning_rate": 3.326679973386561e-05,
      "loss": 0.0897,
      "step": 2840
    },
    {
      "epoch": 3.42,
      "learning_rate": 3.193612774451098e-05,
      "loss": 0.0923,
      "step": 2860
    },
    {
      "epoch": 3.45,
      "learning_rate": 3.0605455755156355e-05,
      "loss": 0.0922,
      "step": 2880
    },
    {
      "epoch": 3.47,
      "learning_rate": 2.9274783765801732e-05,
      "loss": 0.0939,
      "step": 2900
    },
    {
      "epoch": 3.49,
      "learning_rate": 2.7944111776447106e-05,
      "loss": 0.0931,
      "step": 2920
    },
    {
      "epoch": 3.52,
      "learning_rate": 2.661343978709248e-05,
      "loss": 0.0898,
      "step": 2940
    },
    {
      "epoch": 3.54,
      "learning_rate": 2.528276779773786e-05,
      "loss": 0.0943,
      "step": 2960
    },
    {
      "epoch": 3.57,
      "learning_rate": 2.3952095808383234e-05,
      "loss": 0.0932,
      "step": 2980
    },
    {
      "epoch": 3.59,
      "learning_rate": 2.2687957418496342e-05,
      "loss": 0.0909,
      "step": 3000
    },
    {
      "epoch": 3.61,
      "learning_rate": 2.135728542914172e-05,
      "loss": 0.0926,
      "step": 3020
    },
    {
      "epoch": 3.64,
      "learning_rate": 2.0026613439787093e-05,
      "loss": 0.0902,
      "step": 3040
    },
    {
      "epoch": 3.66,
      "learning_rate": 1.869594145043247e-05,
      "loss": 0.0931,
      "step": 3060
    },
    {
      "epoch": 3.69,
      "learning_rate": 1.7365269461077845e-05,
      "loss": 0.0892,
      "step": 3080
    },
    {
      "epoch": 3.71,
      "learning_rate": 1.603459747172322e-05,
      "loss": 0.0912,
      "step": 3100
    },
    {
      "epoch": 3.73,
      "learning_rate": 1.4703925482368597e-05,
      "loss": 0.0897,
      "step": 3120
    },
    {
      "epoch": 3.76,
      "learning_rate": 1.3373253493013973e-05,
      "loss": 0.0912,
      "step": 3140
    },
    {
      "epoch": 3.78,
      "learning_rate": 1.2042581503659348e-05,
      "loss": 0.0918,
      "step": 3160
    },
    {
      "epoch": 3.8,
      "learning_rate": 1.0711909514304724e-05,
      "loss": 0.0861,
      "step": 3180
    },
    {
      "epoch": 3.83,
      "learning_rate": 9.3812375249501e-06,
      "loss": 0.091,
      "step": 3200
    },
    {
      "epoch": 3.85,
      "learning_rate": 8.050565535595477e-06,
      "loss": 0.0904,
      "step": 3220
    },
    {
      "epoch": 3.88,
      "learning_rate": 6.719893546240852e-06,
      "loss": 0.093,
      "step": 3240
    },
    {
      "epoch": 3.9,
      "learning_rate": 5.3892215568862275e-06,
      "loss": 0.0928,
      "step": 3260
    },
    {
      "epoch": 3.92,
      "learning_rate": 4.058549567531603e-06,
      "loss": 0.092,
      "step": 3280
    },
    {
      "epoch": 3.95,
      "learning_rate": 2.7278775781769794e-06,
      "loss": 0.0905,
      "step": 3300
    },
    {
      "epoch": 3.97,
      "learning_rate": 1.3972055888223554e-06,
      "loss": 0.0913,
      "step": 3320
    },
    {
      "epoch": 4.0,
      "learning_rate": 6.65335994677312e-08,
      "loss": 0.0927,
      "step": 3340
    }
  ],
  "logging_steps": 20,
  "max_steps": 3340,
  "num_train_epochs": 4,
  "save_steps": 500,
  "total_flos": 1.6896316557662093e+19,
  "trial_name": null,
  "trial_params": null
}