{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.0,
  "global_step": 5600,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "learning_rate": 0.00019964285714285716,
      "loss": 0.0279,
      "step": 10
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0001992857142857143,
      "loss": 0.0234,
      "step": 20
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.00019892857142857144,
      "loss": 0.1147,
      "step": 30
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0001985714285714286,
      "loss": 0.0921,
      "step": 40
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.00019821428571428572,
      "loss": 0.0297,
      "step": 50
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.00019785714285714288,
      "loss": 0.0295,
      "step": 60
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.00019750000000000003,
      "loss": 0.079,
      "step": 70
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.00019717857142857143,
      "loss": 0.1888,
      "step": 80
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.00019682142857142858,
      "loss": 0.1174,
      "step": 90
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.00019646428571428573,
      "loss": 0.0452,
      "step": 100
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.00019610714285714286,
      "loss": 0.1854,
      "step": 110
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.00019575000000000001,
      "loss": 0.1035,
      "step": 120
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.00019539285714285717,
      "loss": 0.0937,
      "step": 130
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0001950357142857143,
      "loss": 0.0312,
      "step": 140
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.00019467857142857145,
      "loss": 0.0776,
      "step": 150
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.00019432142857142857,
      "loss": 0.0497,
      "step": 160
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.00019396428571428573,
      "loss": 0.1676,
      "step": 170
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.00019360714285714288,
      "loss": 0.0727,
      "step": 180
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.00019325,
      "loss": 0.0946,
      "step": 190
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.00019289285714285713,
      "loss": 0.078,
      "step": 200
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0001925357142857143,
      "loss": 0.0193,
      "step": 210
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.00019217857142857144,
      "loss": 0.1381,
      "step": 220
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.00019182142857142857,
      "loss": 0.04,
      "step": 230
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.00019146428571428572,
      "loss": 0.0533,
      "step": 240
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.00019110714285714288,
      "loss": 0.0538,
      "step": 250
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.00019075,
      "loss": 0.0614,
      "step": 260
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.00019039285714285716,
      "loss": 0.0522,
      "step": 270
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.00019003571428571428,
      "loss": 0.1157,
      "step": 280
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.00018967857142857144,
      "loss": 0.088,
      "step": 290
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0001893214285714286,
      "loss": 0.0661,
      "step": 300
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.00018896428571428572,
      "loss": 0.0302,
      "step": 310
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.00018860714285714287,
      "loss": 0.2186,
      "step": 320
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.00018825000000000002,
      "loss": 0.0831,
      "step": 330
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.00018789285714285715,
      "loss": 0.0904,
      "step": 340
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0001875357142857143,
      "loss": 0.0607,
      "step": 350
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.00018717857142857146,
      "loss": 0.0794,
      "step": 360
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.00018682142857142858,
      "loss": 0.0367,
      "step": 370
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0001864642857142857,
      "loss": 0.0733,
      "step": 380
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.00018610714285714286,
      "loss": 0.0577,
      "step": 390
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.00018575,
      "loss": 0.0532,
      "step": 400
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.00018539285714285714,
      "loss": 0.0518,
      "step": 410
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0001850357142857143,
      "loss": 0.0604,
      "step": 420
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.00018467857142857142,
      "loss": 0.1067,
      "step": 430
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.00018432142857142858,
      "loss": 0.069,
      "step": 440
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.00018396428571428573,
      "loss": 0.0886,
      "step": 450
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.00018360714285714286,
      "loss": 0.032,
      "step": 460
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.00018325,
      "loss": 0.1077,
      "step": 470
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.00018289285714285716,
      "loss": 0.0826,
      "step": 480
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0001825357142857143,
      "loss": 0.0361,
      "step": 490
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.00018217857142857144,
      "loss": 0.0141,
      "step": 500
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0001818214285714286,
      "loss": 0.173,
      "step": 510
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.00018146428571428572,
      "loss": 0.0598,
      "step": 520
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.00018110714285714288,
      "loss": 0.0885,
      "step": 530
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.00018075000000000003,
      "loss": 0.1444,
      "step": 540
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.00018039285714285713,
      "loss": 0.0568,
      "step": 550
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.00018003571428571428,
      "loss": 0.0707,
      "step": 560
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.00017967857142857144,
      "loss": 0.083,
      "step": 570
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.00017932142857142856,
      "loss": 0.0752,
      "step": 580
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.00017896428571428572,
      "loss": 0.0787,
      "step": 590
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.00017860714285714287,
      "loss": 0.0916,
      "step": 600
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.00017825,
      "loss": 0.0724,
      "step": 610
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.00017789285714285715,
      "loss": 0.0751,
      "step": 620
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0001775357142857143,
      "loss": 0.0858,
      "step": 630
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.00017717857142857143,
      "loss": 0.1406,
      "step": 640
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.00017682142857142858,
      "loss": 0.1601,
      "step": 650
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.00017646428571428574,
      "loss": 0.0523,
      "step": 660
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.00017610714285714286,
      "loss": 0.0872,
      "step": 670
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.00017575000000000002,
      "loss": 0.0446,
      "step": 680
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.00017539285714285717,
      "loss": 0.1026,
      "step": 690
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0001750357142857143,
      "loss": 0.1335,
      "step": 700
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.00017467857142857145,
      "loss": 0.1539,
      "step": 710
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.00017432142857142858,
      "loss": 0.0735,
      "step": 720
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0001739642857142857,
      "loss": 0.0765,
      "step": 730
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.00017360714285714286,
      "loss": 0.1106,
      "step": 740
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.00017325,
      "loss": 0.0611,
      "step": 750
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.00017289285714285714,
      "loss": 0.057,
      "step": 760
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0001725357142857143,
      "loss": 0.0745,
      "step": 770
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.00017217857142857144,
      "loss": 0.1908,
      "step": 780
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.00017182142857142857,
      "loss": 0.0956,
      "step": 790
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.00017146428571428572,
      "loss": 0.155,
      "step": 800
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.00017110714285714288,
      "loss": 0.0507,
      "step": 810
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.00017075,
      "loss": 0.0468,
      "step": 820
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.00017039285714285716,
      "loss": 0.1161,
      "step": 830
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0001700357142857143,
      "loss": 0.2227,
      "step": 840
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.00016967857142857144,
      "loss": 0.1122,
      "step": 850
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0001693214285714286,
      "loss": 0.0972,
      "step": 860
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.00016896428571428574,
      "loss": 0.1269,
      "step": 870
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.00016860714285714287,
      "loss": 0.0924,
      "step": 880
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.00016825000000000002,
      "loss": 0.108,
      "step": 890
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.00016789285714285715,
      "loss": 0.1069,
      "step": 900
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.00016753571428571428,
      "loss": 0.1064,
      "step": 910
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.00016717857142857143,
      "loss": 0.0687,
      "step": 920
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.00016682142857142858,
      "loss": 0.0471,
      "step": 930
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0001664642857142857,
      "loss": 0.0959,
      "step": 940
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.00016610714285714286,
      "loss": 0.0077,
      "step": 950
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.00016575000000000002,
      "loss": 0.0621,
      "step": 960
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.00016539285714285714,
      "loss": 0.0241,
      "step": 970
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0001650357142857143,
      "loss": 0.0855,
      "step": 980
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.00016467857142857145,
      "loss": 0.1522,
      "step": 990
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.00016432142857142858,
      "loss": 0.0306,
      "step": 1000
    },
    {
      "epoch": 0.71,
      "eval_accuracy": 0.9344408717399071,
      "eval_loss": 0.1823781132698059,
      "eval_runtime": 109.1042,
      "eval_samples_per_second": 51.309,
      "eval_steps_per_second": 6.416,
      "step": 1000
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.00016396428571428573,
      "loss": 0.1212,
      "step": 1010
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.00016360714285714286,
      "loss": 0.1488,
      "step": 1020
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.00016325,
      "loss": 0.0428,
      "step": 1030
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.00016289285714285716,
      "loss": 0.1219,
      "step": 1040
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0001625357142857143,
      "loss": 0.0716,
      "step": 1050
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.00016217857142857144,
      "loss": 0.1323,
      "step": 1060
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.00016182142857142857,
      "loss": 0.0733,
      "step": 1070
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.00016146428571428572,
      "loss": 0.0932,
      "step": 1080
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.00016110714285714285,
      "loss": 0.0646,
      "step": 1090
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.00016075,
      "loss": 0.0693,
      "step": 1100
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.00016039285714285716,
      "loss": 0.0873,
      "step": 1110
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.00016003571428571428,
      "loss": 0.0675,
      "step": 1120
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.00015967857142857144,
      "loss": 0.0612,
      "step": 1130
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.00015932142857142856,
      "loss": 0.0493,
      "step": 1140
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.00015896428571428572,
      "loss": 0.0764,
      "step": 1150
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.00015860714285714287,
      "loss": 0.0994,
      "step": 1160
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.00015825,
      "loss": 0.0382,
      "step": 1170
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.00015789285714285715,
      "loss": 0.0332,
      "step": 1180
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0001575357142857143,
      "loss": 0.1113,
      "step": 1190
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.00015717857142857143,
      "loss": 0.0621,
      "step": 1200
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.00015682142857142858,
      "loss": 0.1229,
      "step": 1210
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.00015646428571428574,
      "loss": 0.0985,
      "step": 1220
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.00015610714285714286,
      "loss": 0.107,
      "step": 1230
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.00015575000000000002,
      "loss": 0.0587,
      "step": 1240
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.00015539285714285714,
      "loss": 0.1237,
      "step": 1250
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.00015503571428571427,
      "loss": 0.0653,
      "step": 1260
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.00015467857142857142,
      "loss": 0.1165,
      "step": 1270
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.00015432142857142858,
      "loss": 0.1154,
      "step": 1280
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0001539642857142857,
      "loss": 0.1125,
      "step": 1290
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.00015360714285714286,
      "loss": 0.1698,
      "step": 1300
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.00015325,
      "loss": 0.0522,
      "step": 1310
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.00015289285714285714,
      "loss": 0.1525,
      "step": 1320
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0001525357142857143,
      "loss": 0.063,
      "step": 1330
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.00015217857142857145,
      "loss": 0.1032,
      "step": 1340
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.00015182142857142857,
      "loss": 0.1243,
      "step": 1350
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.00015146428571428573,
      "loss": 0.0202,
      "step": 1360
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.00015110714285714288,
      "loss": 0.0919,
      "step": 1370
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.00015075,
      "loss": 0.057,
      "step": 1380
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.00015039285714285716,
      "loss": 0.0219,
      "step": 1390
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0001500357142857143,
      "loss": 0.0747,
      "step": 1400
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.00014967857142857144,
      "loss": 0.0413,
      "step": 1410
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.00014932142857142857,
      "loss": 0.0172,
      "step": 1420
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.00014896428571428572,
      "loss": 0.0187,
      "step": 1430
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.00014860714285714284,
      "loss": 0.0064,
      "step": 1440
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.00014825,
      "loss": 0.1001,
      "step": 1450
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.00014789285714285715,
      "loss": 0.079,
      "step": 1460
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.00014753571428571428,
      "loss": 0.0227,
      "step": 1470
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.00014717857142857143,
      "loss": 0.0965,
      "step": 1480
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.00014682142857142859,
      "loss": 0.0045,
      "step": 1490
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0001464642857142857,
      "loss": 0.0454,
      "step": 1500
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.00014610714285714287,
      "loss": 0.0507,
      "step": 1510
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.00014575000000000002,
      "loss": 0.0723,
      "step": 1520
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.00014539285714285715,
      "loss": 0.0543,
      "step": 1530
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0001450357142857143,
      "loss": 0.0628,
      "step": 1540
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.00014467857142857145,
      "loss": 0.0362,
      "step": 1550
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.00014432142857142858,
      "loss": 0.0076,
      "step": 1560
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.00014396428571428573,
      "loss": 0.0962,
      "step": 1570
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.00014360714285714289,
      "loss": 0.0425,
      "step": 1580
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.00014325,
      "loss": 0.0151,
      "step": 1590
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.00014289285714285714,
      "loss": 0.0337,
      "step": 1600
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0001425357142857143,
      "loss": 0.0364,
      "step": 1610
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.00014217857142857142,
      "loss": 0.0152,
      "step": 1620
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.00014182142857142857,
      "loss": 0.0563,
      "step": 1630
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.00014146428571428573,
      "loss": 0.0103,
      "step": 1640
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.00014110714285714285,
      "loss": 0.0043,
      "step": 1650
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.00014075,
      "loss": 0.0802,
      "step": 1660
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.00014039285714285716,
      "loss": 0.0977,
      "step": 1670
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.00014003571428571429,
      "loss": 0.0915,
      "step": 1680
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.00013967857142857144,
      "loss": 0.0802,
      "step": 1690
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0001393214285714286,
      "loss": 0.0471,
      "step": 1700
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.00013896428571428572,
      "loss": 0.0459,
      "step": 1710
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.00013860714285714287,
      "loss": 0.0835,
      "step": 1720
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.00013825,
      "loss": 0.0565,
      "step": 1730
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.00013789285714285715,
      "loss": 0.031,
      "step": 1740
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0001375357142857143,
      "loss": 0.0242,
      "step": 1750
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.00013717857142857143,
      "loss": 0.0669,
      "step": 1760
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.00013682142857142856,
      "loss": 0.1233,
      "step": 1770
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0001364642857142857,
      "loss": 0.0462,
      "step": 1780
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.00013610714285714287,
      "loss": 0.0319,
      "step": 1790
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.00013575,
      "loss": 0.0132,
      "step": 1800
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.00013539285714285715,
      "loss": 0.0519,
      "step": 1810
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0001350357142857143,
      "loss": 0.007,
      "step": 1820
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.00013467857142857143,
      "loss": 0.0243,
      "step": 1830
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.00013432142857142858,
      "loss": 0.0875,
      "step": 1840
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.00013396428571428573,
      "loss": 0.0051,
      "step": 1850
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.00013360714285714286,
      "loss": 0.0624,
      "step": 1860
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.00013325,
      "loss": 0.0599,
      "step": 1870
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.00013289285714285714,
      "loss": 0.1281,
      "step": 1880
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0001325357142857143,
      "loss": 0.0308,
      "step": 1890
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.00013217857142857145,
      "loss": 0.0448,
      "step": 1900
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.00013182142857142857,
      "loss": 0.0394,
      "step": 1910
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.00013146428571428573,
      "loss": 0.0295,
      "step": 1920
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.00013110714285714288,
      "loss": 0.0616,
      "step": 1930
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.00013075,
      "loss": 0.0451,
      "step": 1940
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.00013039285714285713,
      "loss": 0.0847,
      "step": 1950
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0001300357142857143,
      "loss": 0.0766,
      "step": 1960
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.00012967857142857144,
      "loss": 0.0566,
      "step": 1970
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.00012932142857142857,
      "loss": 0.0563,
      "step": 1980
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.00012896428571428572,
      "loss": 0.0144,
      "step": 1990
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.00012860714285714285,
      "loss": 0.0483,
      "step": 2000
    },
    {
      "epoch": 1.43,
      "eval_accuracy": 0.9528403001071811,
      "eval_loss": 0.1830383986234665,
      "eval_runtime": 109.51,
      "eval_samples_per_second": 51.119,
      "eval_steps_per_second": 6.392,
      "step": 2000
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.00012825,
      "loss": 0.0737,
      "step": 2010
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.00012789285714285715,
      "loss": 0.0139,
      "step": 2020
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.00012753571428571428,
      "loss": 0.0791,
      "step": 2030
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.00012717857142857143,
      "loss": 0.121,
      "step": 2040
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0001268214285714286,
      "loss": 0.0444,
      "step": 2050
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.00012646428571428571,
      "loss": 0.0467,
      "step": 2060
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.00012610714285714287,
      "loss": 0.0175,
      "step": 2070
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.00012575000000000002,
      "loss": 0.0107,
      "step": 2080
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.00012539285714285715,
      "loss": 0.0115,
      "step": 2090
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0001250357142857143,
      "loss": 0.0339,
      "step": 2100
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.00012467857142857145,
      "loss": 0.1753,
      "step": 2110
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.00012432142857142855,
      "loss": 0.0901,
      "step": 2120
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0001239642857142857,
      "loss": 0.0111,
      "step": 2130
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.00012360714285714286,
      "loss": 0.1577,
      "step": 2140
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.00012325,
      "loss": 0.0335,
      "step": 2150
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.00012289285714285714,
      "loss": 0.0207,
      "step": 2160
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0001225357142857143,
      "loss": 0.0556,
      "step": 2170
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.00012217857142857142,
      "loss": 0.0313,
      "step": 2180
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.00012182142857142857,
      "loss": 0.0379,
      "step": 2190
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.00012146428571428571,
      "loss": 0.0196,
      "step": 2200
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.00012110714285714287,
      "loss": 0.0736,
      "step": 2210
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.00012075000000000001,
      "loss": 0.0173,
      "step": 2220
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.00012039285714285715,
      "loss": 0.0735,
      "step": 2230
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0001200357142857143,
      "loss": 0.1114,
      "step": 2240
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.00011967857142857144,
      "loss": 0.0873,
      "step": 2250
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.00011932142857142858,
      "loss": 0.0435,
      "step": 2260
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.00011896428571428573,
      "loss": 0.0076,
      "step": 2270
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.00011860714285714287,
      "loss": 0.031,
      "step": 2280
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.00011825000000000001,
      "loss": 0.0536,
      "step": 2290
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.00011789285714285714,
      "loss": 0.0808,
      "step": 2300
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.00011753571428571428,
      "loss": 0.037,
      "step": 2310
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.00011717857142857142,
      "loss": 0.0847,
      "step": 2320
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.00011682142857142857,
      "loss": 0.0717,
      "step": 2330
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.00011646428571428571,
      "loss": 0.0449,
      "step": 2340
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.00011610714285714285,
      "loss": 0.0481,
      "step": 2350
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.00011575000000000001,
      "loss": 0.0384,
      "step": 2360
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.00011539285714285715,
      "loss": 0.0451,
      "step": 2370
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.00011503571428571429,
      "loss": 0.0349,
      "step": 2380
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.00011467857142857144,
      "loss": 0.0514,
      "step": 2390
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.00011432142857142858,
      "loss": 0.0415,
      "step": 2400
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.00011396428571428572,
      "loss": 0.1398,
      "step": 2410
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.00011360714285714286,
      "loss": 0.0087,
      "step": 2420
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.00011325000000000002,
      "loss": 0.0705,
      "step": 2430
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.00011289285714285716,
      "loss": 0.1571,
      "step": 2440
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0001125357142857143,
      "loss": 0.0117,
      "step": 2450
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.00011217857142857145,
      "loss": 0.079,
      "step": 2460
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.00011182142857142856,
      "loss": 0.0462,
      "step": 2470
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.00011146428571428572,
      "loss": 0.0944,
      "step": 2480
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.00011110714285714286,
      "loss": 0.0472,
      "step": 2490
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.00011075,
      "loss": 0.042,
      "step": 2500
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.00011039285714285715,
      "loss": 0.0385,
      "step": 2510
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.00011003571428571429,
      "loss": 0.0499,
      "step": 2520
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.00010967857142857143,
      "loss": 0.1122,
      "step": 2530
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.00010932142857142857,
      "loss": 0.0278,
      "step": 2540
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.00010896428571428572,
      "loss": 0.0138,
      "step": 2550
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.00010860714285714286,
      "loss": 0.0572,
      "step": 2560
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.00010825,
      "loss": 0.0312,
      "step": 2570
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.00010789285714285716,
      "loss": 0.1273,
      "step": 2580
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0001075357142857143,
      "loss": 0.059,
      "step": 2590
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.00010717857142857144,
      "loss": 0.0071,
      "step": 2600
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.00010682142857142859,
      "loss": 0.0517,
      "step": 2610
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.00010646428571428573,
      "loss": 0.0495,
      "step": 2620
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.00010610714285714287,
      "loss": 0.0732,
      "step": 2630
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.00010575000000000001,
      "loss": 0.1345,
      "step": 2640
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.00010539285714285714,
      "loss": 0.0652,
      "step": 2650
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.00010503571428571428,
      "loss": 0.0695,
      "step": 2660
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.00010467857142857143,
      "loss": 0.0412,
      "step": 2670
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.00010432142857142857,
      "loss": 0.0253,
      "step": 2680
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.00010396428571428571,
      "loss": 0.0915,
      "step": 2690
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.00010360714285714286,
      "loss": 0.0154,
      "step": 2700
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.00010325,
      "loss": 0.0451,
      "step": 2710
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.00010289285714285714,
      "loss": 0.0579,
      "step": 2720
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0001025357142857143,
      "loss": 0.0864,
      "step": 2730
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.00010217857142857144,
      "loss": 0.0308,
      "step": 2740
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.00010182142857142858,
      "loss": 0.0187,
      "step": 2750
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.00010146428571428573,
      "loss": 0.0759,
      "step": 2760
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.00010110714285714287,
      "loss": 0.0413,
      "step": 2770
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.00010075000000000001,
      "loss": 0.0143,
      "step": 2780
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.00010039285714285715,
      "loss": 0.0427,
      "step": 2790
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0001000357142857143,
      "loss": 0.0677,
      "step": 2800
    },
    {
      "epoch": 2.01,
      "learning_rate": 9.967857142857143e-05,
      "loss": 0.0446,
      "step": 2810
    },
    {
      "epoch": 2.01,
      "learning_rate": 9.932142857142858e-05,
      "loss": 0.0476,
      "step": 2820
    },
    {
      "epoch": 2.02,
      "learning_rate": 9.896428571428572e-05,
      "loss": 0.0026,
      "step": 2830
    },
    {
      "epoch": 2.03,
      "learning_rate": 9.860714285714286e-05,
      "loss": 0.0032,
      "step": 2840
    },
    {
      "epoch": 2.04,
      "learning_rate": 9.825e-05,
      "loss": 0.032,
      "step": 2850
    },
    {
      "epoch": 2.04,
      "learning_rate": 9.789285714285716e-05,
      "loss": 0.0363,
      "step": 2860
    },
    {
      "epoch": 2.05,
      "learning_rate": 9.753571428571428e-05,
      "loss": 0.0345,
      "step": 2870
    },
    {
      "epoch": 2.06,
      "learning_rate": 9.717857142857144e-05,
      "loss": 0.0418,
      "step": 2880
    },
    {
      "epoch": 2.06,
      "learning_rate": 9.682142857142858e-05,
      "loss": 0.0293,
      "step": 2890
    },
    {
      "epoch": 2.07,
      "learning_rate": 9.646428571428572e-05,
      "loss": 0.0182,
      "step": 2900
    },
    {
      "epoch": 2.08,
      "learning_rate": 9.610714285714286e-05,
      "loss": 0.0414,
      "step": 2910
    },
    {
      "epoch": 2.09,
      "learning_rate": 9.575000000000001e-05,
      "loss": 0.045,
      "step": 2920
    },
    {
      "epoch": 2.09,
      "learning_rate": 9.539285714285715e-05,
      "loss": 0.0567,
      "step": 2930
    },
    {
      "epoch": 2.1,
      "learning_rate": 9.503571428571429e-05,
      "loss": 0.0183,
      "step": 2940
    },
    {
      "epoch": 2.11,
      "learning_rate": 9.467857142857143e-05,
      "loss": 0.0039,
      "step": 2950
    },
    {
      "epoch": 2.11,
      "learning_rate": 9.432142857142857e-05,
      "loss": 0.0149,
      "step": 2960
    },
    {
      "epoch": 2.12,
      "learning_rate": 9.396428571428571e-05,
      "loss": 0.0288,
      "step": 2970
    },
    {
      "epoch": 2.13,
      "learning_rate": 9.360714285714286e-05,
      "loss": 0.0272,
      "step": 2980
    },
    {
      "epoch": 2.14,
      "learning_rate": 9.325e-05,
      "loss": 0.0054,
      "step": 2990
    },
    {
      "epoch": 2.14,
      "learning_rate": 9.289285714285714e-05,
      "loss": 0.0428,
      "step": 3000
    },
    {
      "epoch": 2.14,
      "eval_accuracy": 0.9476598785280458,
      "eval_loss": 0.206739142537117,
      "eval_runtime": 107.9202,
      "eval_samples_per_second": 51.872,
      "eval_steps_per_second": 6.486,
      "step": 3000
    },
    {
      "epoch": 2.15,
      "learning_rate": 9.25357142857143e-05,
      "loss": 0.0492,
      "step": 3010
    },
    {
      "epoch": 2.16,
      "learning_rate": 9.217857142857144e-05,
      "loss": 0.0313,
      "step": 3020
    },
    {
      "epoch": 2.16,
      "learning_rate": 9.182142857142858e-05,
      "loss": 0.0101,
      "step": 3030
    },
    {
      "epoch": 2.17,
      "learning_rate": 9.146428571428572e-05,
      "loss": 0.0313,
      "step": 3040
    },
    {
      "epoch": 2.18,
      "learning_rate": 9.110714285714286e-05,
      "loss": 0.004,
      "step": 3050
    },
    {
      "epoch": 2.19,
      "learning_rate": 9.075e-05,
      "loss": 0.014,
      "step": 3060
    },
    {
      "epoch": 2.19,
      "learning_rate": 9.039285714285715e-05,
      "loss": 0.0118,
      "step": 3070
    },
    {
      "epoch": 2.2,
      "learning_rate": 9.003571428571429e-05,
      "loss": 0.0433,
      "step": 3080
    },
    {
      "epoch": 2.21,
      "learning_rate": 8.967857142857143e-05,
      "loss": 0.0189,
      "step": 3090
    },
    {
      "epoch": 2.21,
      "learning_rate": 8.932142857142858e-05,
      "loss": 0.0344,
      "step": 3100
    },
    {
      "epoch": 2.22,
      "learning_rate": 8.896428571428572e-05,
      "loss": 0.0662,
      "step": 3110
    },
    {
      "epoch": 2.23,
      "learning_rate": 8.860714285714286e-05,
      "loss": 0.0481,
      "step": 3120
    },
    {
      "epoch": 2.24,
      "learning_rate": 8.825e-05,
      "loss": 0.0507,
      "step": 3130
    },
    {
      "epoch": 2.24,
      "learning_rate": 8.789285714285714e-05,
      "loss": 0.0288,
      "step": 3140
    },
    {
      "epoch": 2.25,
      "learning_rate": 8.753571428571428e-05,
      "loss": 0.0438,
      "step": 3150
    },
    {
      "epoch": 2.26,
      "learning_rate": 8.717857142857144e-05,
      "loss": 0.0883,
      "step": 3160
    },
    {
      "epoch": 2.26,
      "learning_rate": 8.682142857142858e-05,
      "loss": 0.0578,
      "step": 3170
    },
    {
      "epoch": 2.27,
      "learning_rate": 8.646428571428572e-05,
      "loss": 0.0381,
      "step": 3180
    },
    {
      "epoch": 2.28,
      "learning_rate": 8.610714285714286e-05,
      "loss": 0.0126,
      "step": 3190
    },
    {
      "epoch": 2.29,
      "learning_rate": 8.575000000000001e-05,
      "loss": 0.0299,
      "step": 3200
    },
    {
      "epoch": 2.29,
      "learning_rate": 8.539285714285715e-05,
      "loss": 0.0244,
      "step": 3210
    },
    {
      "epoch": 2.3,
      "learning_rate": 8.503571428571429e-05,
      "loss": 0.0671,
      "step": 3220
    },
    {
      "epoch": 2.31,
      "learning_rate": 8.467857142857143e-05,
      "loss": 0.019,
      "step": 3230
    },
    {
      "epoch": 2.31,
      "learning_rate": 8.432142857142857e-05,
      "loss": 0.0076,
      "step": 3240
    },
    {
      "epoch": 2.32,
      "learning_rate": 8.396428571428572e-05,
      "loss": 0.0497,
      "step": 3250
    },
    {
      "epoch": 2.33,
      "learning_rate": 8.360714285714286e-05,
      "loss": 0.0452,
      "step": 3260
    },
    {
      "epoch": 2.34,
      "learning_rate": 8.325e-05,
      "loss": 0.0199,
      "step": 3270
    },
    {
      "epoch": 2.34,
      "learning_rate": 8.289285714285714e-05,
      "loss": 0.003,
      "step": 3280
    },
    {
      "epoch": 2.35,
      "learning_rate": 8.25357142857143e-05,
      "loss": 0.0315,
      "step": 3290
    },
    {
      "epoch": 2.36,
      "learning_rate": 8.217857142857142e-05,
      "loss": 0.0729,
      "step": 3300
    },
    {
      "epoch": 2.36,
      "learning_rate": 8.182142857142858e-05,
      "loss": 0.0036,
      "step": 3310
    },
    {
      "epoch": 2.37,
      "learning_rate": 8.146428571428572e-05,
      "loss": 0.0152,
      "step": 3320
    },
    {
      "epoch": 2.38,
      "learning_rate": 8.110714285714286e-05,
      "loss": 0.0695,
      "step": 3330
    },
    {
      "epoch": 2.39,
      "learning_rate": 8.075e-05,
      "loss": 0.0125,
      "step": 3340
    },
    {
      "epoch": 2.39,
      "learning_rate": 8.039285714285715e-05,
      "loss": 0.0243,
      "step": 3350
    },
    {
      "epoch": 2.4,
      "learning_rate": 8.003571428571429e-05,
      "loss": 0.0127,
      "step": 3360
    },
    {
      "epoch": 2.41,
      "learning_rate": 7.967857142857143e-05,
      "loss": 0.1011,
      "step": 3370
    },
    {
      "epoch": 2.41,
      "learning_rate": 7.932142857142858e-05,
      "loss": 0.0288,
      "step": 3380
    },
    {
      "epoch": 2.42,
      "learning_rate": 7.896428571428571e-05,
      "loss": 0.0739,
      "step": 3390
    },
    {
      "epoch": 2.43,
      "learning_rate": 7.860714285714285e-05,
      "loss": 0.0055,
      "step": 3400
    },
    {
      "epoch": 2.44,
      "learning_rate": 7.825e-05,
      "loss": 0.0054,
      "step": 3410
    },
    {
      "epoch": 2.44,
      "learning_rate": 7.789285714285714e-05,
      "loss": 0.0665,
      "step": 3420
    },
    {
      "epoch": 2.45,
      "learning_rate": 7.753571428571428e-05,
      "loss": 0.005,
      "step": 3430
    },
    {
      "epoch": 2.46,
      "learning_rate": 7.717857142857144e-05,
      "loss": 0.0944,
      "step": 3440
    },
    {
      "epoch": 2.46,
      "learning_rate": 7.682142857142858e-05,
      "loss": 0.0364,
      "step": 3450
    },
    {
      "epoch": 2.47,
      "learning_rate": 7.646428571428572e-05,
      "loss": 0.0151,
      "step": 3460
    },
    {
      "epoch": 2.48,
      "learning_rate": 7.610714285714287e-05,
      "loss": 0.0269,
      "step": 3470
    },
    {
      "epoch": 2.49,
      "learning_rate": 7.575e-05,
      "loss": 0.0054,
      "step": 3480
    },
    {
      "epoch": 2.49,
      "learning_rate": 7.539285714285714e-05,
      "loss": 0.0173,
      "step": 3490
    },
    {
      "epoch": 2.5,
      "learning_rate": 7.503571428571429e-05,
      "loss": 0.0035,
      "step": 3500
    },
    {
      "epoch": 2.51,
      "learning_rate": 7.467857142857143e-05,
      "loss": 0.0929,
      "step": 3510
    },
    {
      "epoch": 2.51,
      "learning_rate": 7.432142857142857e-05,
      "loss": 0.0199,
      "step": 3520
    },
    {
      "epoch": 2.52,
      "learning_rate": 7.396428571428573e-05,
      "loss": 0.0084,
      "step": 3530
    },
    {
      "epoch": 2.53,
      "learning_rate": 7.360714285714287e-05,
      "loss": 0.0032,
      "step": 3540
    },
    {
      "epoch": 2.54,
      "learning_rate": 7.325e-05,
      "loss": 0.0048,
      "step": 3550
    },
    {
      "epoch": 2.54,
      "learning_rate": 7.289285714285715e-05,
      "loss": 0.0044,
      "step": 3560
    },
    {
      "epoch": 2.55,
      "learning_rate": 7.253571428571429e-05,
      "loss": 0.0355,
      "step": 3570
    },
    {
      "epoch": 2.56,
      "learning_rate": 7.217857142857143e-05,
      "loss": 0.0592,
      "step": 3580
    },
    {
      "epoch": 2.56,
      "learning_rate": 7.182142857142858e-05,
      "loss": 0.0282,
      "step": 3590
    },
    {
      "epoch": 2.57,
      "learning_rate": 7.146428571428572e-05,
      "loss": 0.022,
      "step": 3600
    },
    {
      "epoch": 2.58,
      "learning_rate": 7.110714285714286e-05,
      "loss": 0.009,
      "step": 3610
    },
    {
      "epoch": 2.59,
      "learning_rate": 7.075e-05,
      "loss": 0.0041,
      "step": 3620
    },
    {
      "epoch": 2.59,
      "learning_rate": 7.039285714285715e-05,
      "loss": 0.0235,
      "step": 3630
    },
    {
      "epoch": 2.6,
      "learning_rate": 7.003571428571429e-05,
      "loss": 0.0456,
      "step": 3640
    },
    {
      "epoch": 2.61,
      "learning_rate": 6.967857142857143e-05,
      "loss": 0.047,
      "step": 3650
    },
    {
      "epoch": 2.61,
      "learning_rate": 6.932142857142857e-05,
      "loss": 0.0386,
      "step": 3660
    },
    {
      "epoch": 2.62,
      "learning_rate": 6.9e-05,
      "loss": 0.0186,
      "step": 3670
    },
    {
      "epoch": 2.63,
      "learning_rate": 6.864285714285714e-05,
      "loss": 0.0043,
      "step": 3680
    },
    {
      "epoch": 2.64,
      "learning_rate": 6.828571428571429e-05,
      "loss": 0.0029,
      "step": 3690
    },
    {
      "epoch": 2.64,
      "learning_rate": 6.792857142857143e-05,
      "loss": 0.003,
      "step": 3700
    },
    {
      "epoch": 2.65,
      "learning_rate": 6.757142857142857e-05,
      "loss": 0.0838,
      "step": 3710
    },
    {
      "epoch": 2.66,
      "learning_rate": 6.721428571428572e-05,
      "loss": 0.0033,
      "step": 3720
    },
    {
      "epoch": 2.66,
      "learning_rate": 6.685714285714286e-05,
      "loss": 0.0014,
      "step": 3730
    },
    {
      "epoch": 2.67,
      "learning_rate": 6.65e-05,
      "loss": 0.0993,
      "step": 3740
    },
    {
      "epoch": 2.68,
      "learning_rate": 6.614285714285716e-05,
      "loss": 0.0163,
      "step": 3750
    },
    {
      "epoch": 2.69,
      "learning_rate": 6.578571428571428e-05,
      "loss": 0.0053,
      "step": 3760
    },
    {
      "epoch": 2.69,
      "learning_rate": 6.542857142857142e-05,
      "loss": 0.0024,
      "step": 3770
    },
    {
      "epoch": 2.7,
      "learning_rate": 6.507142857142858e-05,
      "loss": 0.0475,
      "step": 3780
    },
    {
      "epoch": 2.71,
      "learning_rate": 6.471428571428572e-05,
      "loss": 0.0031,
      "step": 3790
    },
    {
      "epoch": 2.71,
      "learning_rate": 6.435714285714286e-05,
      "loss": 0.0444,
      "step": 3800
    },
    {
      "epoch": 2.72,
      "learning_rate": 6.400000000000001e-05,
      "loss": 0.0198,
      "step": 3810
    },
    {
      "epoch": 2.73,
      "learning_rate": 6.364285714285715e-05,
      "loss": 0.0435,
      "step": 3820
    },
    {
      "epoch": 2.74,
      "learning_rate": 6.328571428571429e-05,
      "loss": 0.0366,
      "step": 3830
    },
    {
      "epoch": 2.74,
      "learning_rate": 6.292857142857143e-05,
      "loss": 0.0217,
      "step": 3840
    },
    {
      "epoch": 2.75,
      "learning_rate": 6.257142857142857e-05,
      "loss": 0.02,
      "step": 3850
    },
    {
      "epoch": 2.76,
      "learning_rate": 6.221428571428571e-05,
      "loss": 0.005,
      "step": 3860
    },
    {
      "epoch": 2.76,
      "learning_rate": 6.185714285714286e-05,
      "loss": 0.0238,
      "step": 3870
    },
    {
      "epoch": 2.77,
      "learning_rate": 6.15e-05,
      "loss": 0.0431,
      "step": 3880
    },
    {
      "epoch": 2.78,
      "learning_rate": 6.114285714285714e-05,
      "loss": 0.0057,
      "step": 3890
    },
    {
      "epoch": 2.79,
      "learning_rate": 6.078571428571429e-05,
      "loss": 0.0022,
      "step": 3900
    },
    {
      "epoch": 2.79,
      "learning_rate": 6.042857142857144e-05,
      "loss": 0.0706,
      "step": 3910
    },
    {
      "epoch": 2.8,
      "learning_rate": 6.007142857142858e-05,
      "loss": 0.0135,
      "step": 3920
    },
    {
      "epoch": 2.81,
      "learning_rate": 5.9714285714285724e-05,
      "loss": 0.0029,
      "step": 3930
    },
    {
      "epoch": 2.81,
      "learning_rate": 5.935714285714286e-05,
      "loss": 0.0627,
      "step": 3940
    },
    {
      "epoch": 2.82,
      "learning_rate": 5.9e-05,
      "loss": 0.0545,
      "step": 3950
    },
    {
      "epoch": 2.83,
      "learning_rate": 5.8642857142857144e-05,
      "loss": 0.0158,
      "step": 3960
    },
    {
      "epoch": 2.84,
      "learning_rate": 5.828571428571429e-05,
      "loss": 0.0534,
      "step": 3970
    },
    {
      "epoch": 2.84,
      "learning_rate": 5.792857142857143e-05,
      "loss": 0.0278,
      "step": 3980
    },
    {
      "epoch": 2.85,
      "learning_rate": 5.757142857142858e-05,
      "loss": 0.0446,
      "step": 3990
    },
    {
      "epoch": 2.86,
      "learning_rate": 5.721428571428572e-05,
      "loss": 0.0149,
      "step": 4000
    },
    {
      "epoch": 2.86,
      "eval_accuracy": 0.9530189353340479,
      "eval_loss": 0.1966186761856079,
      "eval_runtime": 109.5835,
      "eval_samples_per_second": 51.084,
      "eval_steps_per_second": 6.388,
      "step": 4000
    },
    {
      "epoch": 2.86,
      "learning_rate": 5.6857142857142865e-05,
      "loss": 0.0098,
      "step": 4010
    },
    {
      "epoch": 2.87,
      "learning_rate": 5.65e-05,
      "loss": 0.0194,
      "step": 4020
    },
    {
      "epoch": 2.88,
      "learning_rate": 5.6142857142857145e-05,
      "loss": 0.1191,
      "step": 4030
    },
    {
      "epoch": 2.89,
      "learning_rate": 5.5785714285714285e-05,
      "loss": 0.0042,
      "step": 4040
    },
    {
      "epoch": 2.89,
      "learning_rate": 5.542857142857143e-05,
      "loss": 0.0115,
      "step": 4050
    },
    {
      "epoch": 2.9,
      "learning_rate": 5.507142857142857e-05,
      "loss": 0.0039,
      "step": 4060
    },
    {
      "epoch": 2.91,
      "learning_rate": 5.471428571428572e-05,
      "loss": 0.0083,
      "step": 4070
    },
    {
      "epoch": 2.91,
      "learning_rate": 5.4357142857142865e-05,
      "loss": 0.0263,
      "step": 4080
    },
    {
      "epoch": 2.92,
      "learning_rate": 5.4000000000000005e-05,
      "loss": 0.0045,
      "step": 4090
    },
    {
      "epoch": 2.93,
      "learning_rate": 5.364285714285715e-05,
      "loss": 0.0364,
      "step": 4100
    },
    {
      "epoch": 2.94,
      "learning_rate": 5.3285714285714285e-05,
      "loss": 0.003,
      "step": 4110
    },
    {
      "epoch": 2.94,
      "learning_rate": 5.2928571428571425e-05,
      "loss": 0.0032,
      "step": 4120
    },
    {
      "epoch": 2.95,
      "learning_rate": 5.257142857142857e-05,
      "loss": 0.0232,
      "step": 4130
    },
    {
      "epoch": 2.96,
      "learning_rate": 5.221428571428572e-05,
      "loss": 0.0037,
      "step": 4140
    },
    {
      "epoch": 2.96,
      "learning_rate": 5.185714285714286e-05,
      "loss": 0.0164,
      "step": 4150
    },
    {
      "epoch": 2.97,
      "learning_rate": 5.1500000000000005e-05,
      "loss": 0.0448,
      "step": 4160
    },
    {
      "epoch": 2.98,
      "learning_rate": 5.1142857142857145e-05,
      "loss": 0.0197,
      "step": 4170
    },
    {
      "epoch": 2.99,
      "learning_rate": 5.078571428571429e-05,
      "loss": 0.0036,
      "step": 4180
    },
    {
      "epoch": 2.99,
      "learning_rate": 5.042857142857144e-05,
      "loss": 0.0033,
      "step": 4190
    },
    {
      "epoch": 3.0,
      "learning_rate": 5.007142857142857e-05,
      "loss": 0.002,
      "step": 4200
    },
    {
      "epoch": 3.01,
      "learning_rate": 4.971428571428572e-05,
      "loss": 0.0032,
      "step": 4210
    },
    {
      "epoch": 3.01,
      "learning_rate": 4.935714285714286e-05,
      "loss": 0.0162,
      "step": 4220
    },
    {
      "epoch": 3.02,
      "learning_rate": 4.9e-05,
      "loss": 0.0074,
      "step": 4230
    },
    {
      "epoch": 3.03,
      "learning_rate": 4.8642857142857145e-05,
      "loss": 0.1163,
      "step": 4240
    },
    {
      "epoch": 3.04,
      "learning_rate": 4.828571428571429e-05,
      "loss": 0.0142,
      "step": 4250
    },
    {
      "epoch": 3.04,
      "learning_rate": 4.7928571428571425e-05,
      "loss": 0.0074,
      "step": 4260
    },
    {
      "epoch": 3.05,
      "learning_rate": 4.757142857142857e-05,
      "loss": 0.0207,
      "step": 4270
    },
    {
      "epoch": 3.06,
      "learning_rate": 4.721428571428572e-05,
      "loss": 0.005,
      "step": 4280
    },
    {
      "epoch": 3.06,
      "learning_rate": 4.685714285714286e-05,
      "loss": 0.0044,
      "step": 4290
    },
    {
      "epoch": 3.07,
      "learning_rate": 4.6500000000000005e-05,
      "loss": 0.0017,
      "step": 4300
    },
    {
      "epoch": 3.08,
      "learning_rate": 4.6142857142857145e-05,
      "loss": 0.0022,
      "step": 4310
    },
    {
      "epoch": 3.09,
      "learning_rate": 4.5785714285714285e-05,
      "loss": 0.0294,
      "step": 4320
    },
    {
      "epoch": 3.09,
      "learning_rate": 4.542857142857143e-05,
      "loss": 0.0034,
      "step": 4330
    },
    {
      "epoch": 3.1,
      "learning_rate": 4.507142857142858e-05,
      "loss": 0.0194,
      "step": 4340
    },
    {
      "epoch": 3.11,
      "learning_rate": 4.471428571428571e-05,
      "loss": 0.0233,
      "step": 4350
    },
    {
      "epoch": 3.11,
      "learning_rate": 4.435714285714286e-05,
      "loss": 0.0364,
      "step": 4360
    },
    {
      "epoch": 3.12,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.0703,
      "step": 4370
    },
    {
      "epoch": 3.13,
      "learning_rate": 4.3642857142857146e-05,
      "loss": 0.0039,
      "step": 4380
    },
    {
      "epoch": 3.14,
      "learning_rate": 4.328571428571429e-05,
      "loss": 0.0456,
      "step": 4390
    },
    {
      "epoch": 3.14,
      "learning_rate": 4.292857142857143e-05,
      "loss": 0.0028,
      "step": 4400
    },
    {
      "epoch": 3.15,
      "learning_rate": 4.257142857142857e-05,
      "loss": 0.0296,
      "step": 4410
    },
    {
      "epoch": 3.16,
      "learning_rate": 4.221428571428572e-05,
      "loss": 0.0367,
      "step": 4420
    },
    {
      "epoch": 3.16,
      "learning_rate": 4.185714285714286e-05,
      "loss": 0.0048,
      "step": 4430
    },
    {
      "epoch": 3.17,
      "learning_rate": 4.15e-05,
      "loss": 0.0483,
      "step": 4440
    },
    {
      "epoch": 3.18,
      "learning_rate": 4.1142857142857146e-05,
      "loss": 0.0228,
      "step": 4450
    },
    {
      "epoch": 3.19,
      "learning_rate": 4.0785714285714286e-05,
      "loss": 0.0488,
      "step": 4460
    },
    {
      "epoch": 3.19,
      "learning_rate": 4.042857142857143e-05,
      "loss": 0.0649,
      "step": 4470
    },
    {
      "epoch": 3.2,
      "learning_rate": 4.007142857142857e-05,
      "loss": 0.0153,
      "step": 4480
    },
    {
      "epoch": 3.21,
      "learning_rate": 3.971428571428571e-05,
      "loss": 0.0034,
      "step": 4490
    },
    {
      "epoch": 3.21,
      "learning_rate": 3.935714285714286e-05,
      "loss": 0.0272,
      "step": 4500
    },
    {
      "epoch": 3.22,
      "learning_rate": 3.9000000000000006e-05,
      "loss": 0.0038,
      "step": 4510
    },
    {
      "epoch": 3.23,
      "learning_rate": 3.8642857142857146e-05,
      "loss": 0.0095,
      "step": 4520
    },
    {
      "epoch": 3.24,
      "learning_rate": 3.8285714285714286e-05,
      "loss": 0.0148,
      "step": 4530
    },
    {
      "epoch": 3.24,
      "learning_rate": 3.792857142857143e-05,
      "loss": 0.0057,
      "step": 4540
    },
    {
      "epoch": 3.25,
      "learning_rate": 3.757142857142857e-05,
      "loss": 0.0024,
      "step": 4550
    },
    {
      "epoch": 3.26,
      "learning_rate": 3.721428571428572e-05,
      "loss": 0.0184,
      "step": 4560
    },
    {
      "epoch": 3.26,
      "learning_rate": 3.685714285714286e-05,
      "loss": 0.0024,
      "step": 4570
    },
    {
      "epoch": 3.27,
      "learning_rate": 3.65e-05,
      "loss": 0.005,
      "step": 4580
    },
    {
      "epoch": 3.28,
      "learning_rate": 3.6142857142857146e-05,
      "loss": 0.0043,
      "step": 4590
    },
    {
      "epoch": 3.29,
      "learning_rate": 3.5785714285714286e-05,
      "loss": 0.0178,
      "step": 4600
    },
    {
      "epoch": 3.29,
      "learning_rate": 3.5428571428571426e-05,
      "loss": 0.0017,
      "step": 4610
    },
    {
      "epoch": 3.3,
      "learning_rate": 3.507142857142857e-05,
      "loss": 0.0226,
      "step": 4620
    },
    {
      "epoch": 3.31,
      "learning_rate": 3.471428571428571e-05,
      "loss": 0.025,
      "step": 4630
    },
    {
      "epoch": 3.31,
      "learning_rate": 3.435714285714286e-05,
      "loss": 0.0016,
      "step": 4640
    },
    {
      "epoch": 3.32,
      "learning_rate": 3.4000000000000007e-05,
      "loss": 0.0013,
      "step": 4650
    },
    {
      "epoch": 3.33,
      "learning_rate": 3.364285714285714e-05,
      "loss": 0.0018,
      "step": 4660
    },
    {
      "epoch": 3.34,
      "learning_rate": 3.3285714285714286e-05,
      "loss": 0.0083,
      "step": 4670
    },
    {
      "epoch": 3.34,
      "learning_rate": 3.292857142857143e-05,
      "loss": 0.0147,
      "step": 4680
    },
    {
      "epoch": 3.35,
      "learning_rate": 3.257142857142857e-05,
      "loss": 0.002,
      "step": 4690
    },
    {
      "epoch": 3.36,
      "learning_rate": 3.221428571428571e-05,
      "loss": 0.0094,
      "step": 4700
    },
    {
      "epoch": 3.36,
      "learning_rate": 3.185714285714286e-05,
      "loss": 0.0821,
      "step": 4710
    },
    {
      "epoch": 3.37,
      "learning_rate": 3.15e-05,
      "loss": 0.0058,
      "step": 4720
    },
    {
      "epoch": 3.38,
      "learning_rate": 3.114285714285715e-05,
      "loss": 0.0026,
      "step": 4730
    },
    {
      "epoch": 3.39,
      "learning_rate": 3.078571428571429e-05,
      "loss": 0.0032,
      "step": 4740
    },
    {
      "epoch": 3.39,
      "learning_rate": 3.042857142857143e-05,
      "loss": 0.0026,
      "step": 4750
    },
    {
      "epoch": 3.4,
      "learning_rate": 3.0071428571428573e-05,
      "loss": 0.0458,
      "step": 4760
    },
    {
      "epoch": 3.41,
      "learning_rate": 2.9714285714285717e-05,
      "loss": 0.0148,
      "step": 4770
    },
    {
      "epoch": 3.41,
      "learning_rate": 2.935714285714286e-05,
      "loss": 0.0022,
      "step": 4780
    },
    {
      "epoch": 3.42,
      "learning_rate": 2.9e-05,
      "loss": 0.008,
      "step": 4790
    },
    {
      "epoch": 3.43,
      "learning_rate": 2.8642857142857144e-05,
      "loss": 0.0311,
      "step": 4800
    },
    {
      "epoch": 3.44,
      "learning_rate": 2.8285714285714287e-05,
      "loss": 0.0411,
      "step": 4810
    },
    {
      "epoch": 3.44,
      "learning_rate": 2.792857142857143e-05,
      "loss": 0.0033,
      "step": 4820
    },
    {
      "epoch": 3.45,
      "learning_rate": 2.757142857142857e-05,
      "loss": 0.0666,
      "step": 4830
    },
    {
      "epoch": 3.46,
      "learning_rate": 2.7214285714285714e-05,
      "loss": 0.0161,
      "step": 4840
    },
    {
      "epoch": 3.46,
      "learning_rate": 2.6857142857142857e-05,
      "loss": 0.0064,
      "step": 4850
    },
    {
      "epoch": 3.47,
      "learning_rate": 2.6500000000000004e-05,
      "loss": 0.0114,
      "step": 4860
    },
    {
      "epoch": 3.48,
      "learning_rate": 2.6142857142857147e-05,
      "loss": 0.0098,
      "step": 4870
    },
    {
      "epoch": 3.49,
      "learning_rate": 2.5785714285714284e-05,
      "loss": 0.0026,
      "step": 4880
    },
    {
      "epoch": 3.49,
      "learning_rate": 2.542857142857143e-05,
      "loss": 0.0339,
      "step": 4890
    },
    {
      "epoch": 3.5,
      "learning_rate": 2.5071428571428574e-05,
      "loss": 0.0037,
      "step": 4900
    },
    {
      "epoch": 3.51,
      "learning_rate": 2.4714285714285714e-05,
      "loss": 0.0252,
      "step": 4910
    },
    {
      "epoch": 3.51,
      "learning_rate": 2.4357142857142857e-05,
      "loss": 0.0018,
      "step": 4920
    },
    {
      "epoch": 3.52,
      "learning_rate": 2.4e-05,
      "loss": 0.04,
      "step": 4930
    },
    {
      "epoch": 3.53,
      "learning_rate": 2.3642857142857144e-05,
      "loss": 0.023,
      "step": 4940
    },
    {
      "epoch": 3.54,
      "learning_rate": 2.3285714285714287e-05,
      "loss": 0.0378,
      "step": 4950
    },
    {
      "epoch": 3.54,
      "learning_rate": 2.292857142857143e-05,
      "loss": 0.049,
      "step": 4960
    },
    {
      "epoch": 3.55,
      "learning_rate": 2.257142857142857e-05,
      "loss": 0.0479,
      "step": 4970
    },
    {
      "epoch": 3.56,
      "learning_rate": 2.2214285714285714e-05,
      "loss": 0.0037,
      "step": 4980
    },
    {
      "epoch": 3.56,
      "learning_rate": 2.185714285714286e-05,
      "loss": 0.034,
      "step": 4990
    },
    {
      "epoch": 3.57,
      "learning_rate": 2.15e-05,
      "loss": 0.0581,
      "step": 5000
    },
    {
      "epoch": 3.57,
      "eval_accuracy": 0.9546266523758485,
      "eval_loss": 0.17252865433692932,
      "eval_runtime": 108.7929,
      "eval_samples_per_second": 51.456,
      "eval_steps_per_second": 6.434,
      "step": 5000
    },
    {
      "epoch": 3.58,
      "learning_rate": 2.1142857142857144e-05,
      "loss": 0.0252,
      "step": 5010
    },
    {
      "epoch": 3.59,
      "learning_rate": 2.0785714285714288e-05,
      "loss": 0.0039,
      "step": 5020
    },
    {
      "epoch": 3.59,
      "learning_rate": 2.042857142857143e-05,
      "loss": 0.0036,
      "step": 5030
    },
    {
      "epoch": 3.6,
      "learning_rate": 2.007142857142857e-05,
      "loss": 0.0046,
      "step": 5040
    },
    {
      "epoch": 3.61,
      "learning_rate": 1.9714285714285714e-05,
      "loss": 0.0028,
      "step": 5050
    },
    {
      "epoch": 3.61,
      "learning_rate": 1.9357142857142858e-05,
      "loss": 0.002,
      "step": 5060
    },
    {
      "epoch": 3.62,
      "learning_rate": 1.9e-05,
      "loss": 0.0504,
      "step": 5070
    },
    {
      "epoch": 3.63,
      "learning_rate": 1.864285714285714e-05,
      "loss": 0.0027,
      "step": 5080
    },
    {
      "epoch": 3.64,
      "learning_rate": 1.8285714285714288e-05,
      "loss": 0.061,
      "step": 5090
    },
    {
      "epoch": 3.64,
      "learning_rate": 1.792857142857143e-05,
      "loss": 0.0055,
      "step": 5100
    },
    {
      "epoch": 3.65,
      "learning_rate": 1.757142857142857e-05,
      "loss": 0.021,
      "step": 5110
    },
    {
      "epoch": 3.66,
      "learning_rate": 1.7214285714285715e-05,
      "loss": 0.0133,
      "step": 5120
    },
    {
      "epoch": 3.66,
      "learning_rate": 1.6857142857142858e-05,
      "loss": 0.0045,
      "step": 5130
    },
    {
      "epoch": 3.67,
      "learning_rate": 1.65e-05,
      "loss": 0.0076,
      "step": 5140
    },
    {
      "epoch": 3.68,
      "learning_rate": 1.614285714285714e-05,
      "loss": 0.004,
      "step": 5150
    },
    {
      "epoch": 3.69,
      "learning_rate": 1.5785714285714288e-05,
      "loss": 0.0013,
      "step": 5160
    },
    {
      "epoch": 3.69,
      "learning_rate": 1.5428571428571428e-05,
      "loss": 0.029,
      "step": 5170
    },
    {
      "epoch": 3.7,
      "learning_rate": 1.5071428571428573e-05,
      "loss": 0.0032,
      "step": 5180
    },
    {
      "epoch": 3.71,
      "learning_rate": 1.4714285714285713e-05,
      "loss": 0.0014,
      "step": 5190
    },
    {
      "epoch": 3.71,
      "learning_rate": 1.4357142857142858e-05,
      "loss": 0.0068,
      "step": 5200
    },
    {
      "epoch": 3.72,
      "learning_rate": 1.4000000000000001e-05,
      "loss": 0.0013,
      "step": 5210
    },
    {
      "epoch": 3.73,
      "learning_rate": 1.3642857142857143e-05,
      "loss": 0.0012,
      "step": 5220
    },
    {
      "epoch": 3.74,
      "learning_rate": 1.3285714285714288e-05,
      "loss": 0.0016,
      "step": 5230
    },
    {
      "epoch": 3.74,
      "learning_rate": 1.2928571428571428e-05,
      "loss": 0.0033,
      "step": 5240
    },
    {
      "epoch": 3.75,
      "learning_rate": 1.2571428571428573e-05,
      "loss": 0.0202,
      "step": 5250
    },
    {
      "epoch": 3.76,
      "learning_rate": 1.2214285714285715e-05,
      "loss": 0.0089,
      "step": 5260
    },
    {
      "epoch": 3.76,
      "learning_rate": 1.1857142857142858e-05,
      "loss": 0.0647,
      "step": 5270
    },
    {
      "epoch": 3.77,
      "learning_rate": 1.1500000000000002e-05,
      "loss": 0.0031,
      "step": 5280
    },
    {
      "epoch": 3.78,
      "learning_rate": 1.1142857142857143e-05,
      "loss": 0.0015,
      "step": 5290
    },
    {
      "epoch": 3.79,
      "learning_rate": 1.0785714285714287e-05,
      "loss": 0.0046,
      "step": 5300
    },
    {
      "epoch": 3.79,
      "learning_rate": 1.0428571428571428e-05,
      "loss": 0.0019,
      "step": 5310
    },
    {
      "epoch": 3.8,
      "learning_rate": 1.0071428571428572e-05,
      "loss": 0.0152,
      "step": 5320
    },
    {
      "epoch": 3.81,
      "learning_rate": 9.714285714285715e-06,
      "loss": 0.0035,
      "step": 5330
    },
    {
      "epoch": 3.81,
      "learning_rate": 9.357142857142857e-06,
      "loss": 0.0266,
      "step": 5340
    },
    {
      "epoch": 3.82,
      "learning_rate": 9e-06,
      "loss": 0.0015,
      "step": 5350
    },
    {
      "epoch": 3.83,
      "learning_rate": 8.642857142857144e-06,
      "loss": 0.0033,
      "step": 5360
    },
    {
      "epoch": 3.84,
      "learning_rate": 8.285714285714285e-06,
      "loss": 0.003,
      "step": 5370
    },
    {
      "epoch": 3.84,
      "learning_rate": 7.928571428571429e-06,
      "loss": 0.0012,
      "step": 5380
    },
    {
      "epoch": 3.85,
      "learning_rate": 7.571428571428572e-06,
      "loss": 0.0018,
      "step": 5390
    },
    {
      "epoch": 3.86,
      "learning_rate": 7.214285714285715e-06,
      "loss": 0.0011,
      "step": 5400
    },
    {
      "epoch": 3.86,
      "learning_rate": 6.857142857142858e-06,
      "loss": 0.0217,
      "step": 5410
    },
    {
      "epoch": 3.87,
      "learning_rate": 6.5000000000000004e-06,
      "loss": 0.0173,
      "step": 5420
    },
    {
      "epoch": 3.88,
      "learning_rate": 6.142857142857143e-06,
      "loss": 0.0019,
      "step": 5430
    },
    {
      "epoch": 3.89,
      "learning_rate": 5.785714285714286e-06,
      "loss": 0.0082,
      "step": 5440
    },
    {
      "epoch": 3.89,
      "learning_rate": 5.428571428571429e-06,
      "loss": 0.0018,
      "step": 5450
    },
    {
      "epoch": 3.9,
      "learning_rate": 5.071428571428571e-06,
      "loss": 0.0014,
      "step": 5460
    },
    {
      "epoch": 3.91,
      "learning_rate": 4.714285714285715e-06,
      "loss": 0.0219,
      "step": 5470
    },
    {
      "epoch": 3.91,
      "learning_rate": 4.357142857142857e-06,
      "loss": 0.0201,
      "step": 5480
    },
    {
      "epoch": 3.92,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.0262,
      "step": 5490
    },
    {
      "epoch": 3.93,
      "learning_rate": 3.642857142857143e-06,
      "loss": 0.0019,
      "step": 5500
    },
    {
      "epoch": 3.94,
      "learning_rate": 3.2857142857142857e-06,
      "loss": 0.0025,
      "step": 5510
    },
    {
      "epoch": 3.94,
      "learning_rate": 2.9285714285714287e-06,
      "loss": 0.0011,
      "step": 5520
    },
    {
      "epoch": 3.95,
      "learning_rate": 2.5714285714285716e-06,
      "loss": 0.0016,
      "step": 5530
    },
    {
      "epoch": 3.96,
      "learning_rate": 2.214285714285714e-06,
      "loss": 0.0174,
      "step": 5540
    },
    {
      "epoch": 3.96,
      "learning_rate": 1.8571428571428573e-06,
      "loss": 0.0012,
      "step": 5550
    },
    {
      "epoch": 3.97,
      "learning_rate": 1.5e-06,
      "loss": 0.0019,
      "step": 5560
    },
    {
      "epoch": 3.98,
      "learning_rate": 1.142857142857143e-06,
      "loss": 0.0015,
      "step": 5570
    },
    {
      "epoch": 3.99,
      "learning_rate": 7.857142857142858e-07,
      "loss": 0.0301,
      "step": 5580
    },
    {
      "epoch": 3.99,
      "learning_rate": 4.285714285714286e-07,
      "loss": 0.0027,
      "step": 5590
    },
    {
      "epoch": 4.0,
      "learning_rate": 7.142857142857144e-08,
      "loss": 0.0033,
      "step": 5600
    },
    {
      "epoch": 4.0,
      "step": 5600,
      "total_flos": 6.939872621924401e+18,
      "train_loss": 0.045772934323841975,
      "train_runtime": 3642.6125,
      "train_samples_per_second": 24.586,
      "train_steps_per_second": 1.537
    }
  ],
  "max_steps": 5600,
  "num_train_epochs": 4,
  "total_flos": 6.939872621924401e+18,
  "trial_name": null,
  "trial_params": null
}