{
  "best_metric": 0.994535519125683,
  "best_model_checkpoint": "swin-tiny-patch4-window7-224-shortSleeveCleanedData/checkpoint-2352",
  "epoch": 20.0,
  "global_step": 2940,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.07,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 1.098,
      "step": 10
    },
    {
      "epoch": 0.14,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.9229,
      "step": 20
    },
    {
      "epoch": 0.2,
      "learning_rate": 5e-05,
      "loss": 0.6598,
      "step": 30
    },
    {
      "epoch": 0.27,
      "learning_rate": 4.982817869415808e-05,
      "loss": 0.3916,
      "step": 40
    },
    {
      "epoch": 0.34,
      "learning_rate": 4.965635738831615e-05,
      "loss": 0.2737,
      "step": 50
    },
    {
      "epoch": 0.41,
      "learning_rate": 4.948453608247423e-05,
      "loss": 0.2705,
      "step": 60
    },
    {
      "epoch": 0.48,
      "learning_rate": 4.931271477663231e-05,
      "loss": 0.245,
      "step": 70
    },
    {
      "epoch": 0.54,
      "learning_rate": 4.9140893470790375e-05,
      "loss": 0.1672,
      "step": 80
    },
    {
      "epoch": 0.61,
      "learning_rate": 4.8969072164948454e-05,
      "loss": 0.1976,
      "step": 90
    },
    {
      "epoch": 0.68,
      "learning_rate": 4.879725085910653e-05,
      "loss": 0.2397,
      "step": 100
    },
    {
      "epoch": 0.75,
      "learning_rate": 4.8625429553264604e-05,
      "loss": 0.168,
      "step": 110
    },
    {
      "epoch": 0.82,
      "learning_rate": 4.845360824742268e-05,
      "loss": 0.1745,
      "step": 120
    },
    {
      "epoch": 0.88,
      "learning_rate": 4.828178694158076e-05,
      "loss": 0.202,
      "step": 130
    },
    {
      "epoch": 0.95,
      "learning_rate": 4.810996563573883e-05,
      "loss": 0.1819,
      "step": 140
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9879781420765027,
      "eval_loss": 0.04708021134138107,
      "eval_runtime": 29.2491,
      "eval_samples_per_second": 31.283,
      "eval_steps_per_second": 3.932,
      "step": 147
    },
    {
      "epoch": 1.02,
      "learning_rate": 4.793814432989691e-05,
      "loss": 0.1651,
      "step": 150
    },
    {
      "epoch": 1.09,
      "learning_rate": 4.776632302405499e-05,
      "loss": 0.1301,
      "step": 160
    },
    {
      "epoch": 1.16,
      "learning_rate": 4.7594501718213055e-05,
      "loss": 0.1177,
      "step": 170
    },
    {
      "epoch": 1.22,
      "learning_rate": 4.7422680412371134e-05,
      "loss": 0.1653,
      "step": 180
    },
    {
      "epoch": 1.29,
      "learning_rate": 4.725085910652921e-05,
      "loss": 0.1691,
      "step": 190
    },
    {
      "epoch": 1.36,
      "learning_rate": 4.7079037800687284e-05,
      "loss": 0.1266,
      "step": 200
    },
    {
      "epoch": 1.43,
      "learning_rate": 4.690721649484536e-05,
      "loss": 0.1334,
      "step": 210
    },
    {
      "epoch": 1.5,
      "learning_rate": 4.673539518900344e-05,
      "loss": 0.1716,
      "step": 220
    },
    {
      "epoch": 1.56,
      "learning_rate": 4.656357388316151e-05,
      "loss": 0.1662,
      "step": 230
    },
    {
      "epoch": 1.63,
      "learning_rate": 4.639175257731959e-05,
      "loss": 0.1527,
      "step": 240
    },
    {
      "epoch": 1.7,
      "learning_rate": 4.621993127147767e-05,
      "loss": 0.1171,
      "step": 250
    },
    {
      "epoch": 1.77,
      "learning_rate": 4.6048109965635736e-05,
      "loss": 0.1966,
      "step": 260
    },
    {
      "epoch": 1.84,
      "learning_rate": 4.5876288659793814e-05,
      "loss": 0.1274,
      "step": 270
    },
    {
      "epoch": 1.9,
      "learning_rate": 4.570446735395189e-05,
      "loss": 0.1589,
      "step": 280
    },
    {
      "epoch": 1.97,
      "learning_rate": 4.5532646048109965e-05,
      "loss": 0.1431,
      "step": 290
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9890710382513661,
      "eval_loss": 0.045684538781642914,
      "eval_runtime": 29.1141,
      "eval_samples_per_second": 31.428,
      "eval_steps_per_second": 3.95,
      "step": 294
    },
    {
      "epoch": 2.04,
      "learning_rate": 4.536082474226804e-05,
      "loss": 0.1321,
      "step": 300
    },
    {
      "epoch": 2.11,
      "learning_rate": 4.518900343642612e-05,
      "loss": 0.1454,
      "step": 310
    },
    {
      "epoch": 2.18,
      "learning_rate": 4.5017182130584194e-05,
      "loss": 0.1106,
      "step": 320
    },
    {
      "epoch": 2.24,
      "learning_rate": 4.484536082474227e-05,
      "loss": 0.1536,
      "step": 330
    },
    {
      "epoch": 2.31,
      "learning_rate": 4.467353951890035e-05,
      "loss": 0.1166,
      "step": 340
    },
    {
      "epoch": 2.38,
      "learning_rate": 4.4501718213058416e-05,
      "loss": 0.1172,
      "step": 350
    },
    {
      "epoch": 2.45,
      "learning_rate": 4.4329896907216494e-05,
      "loss": 0.1021,
      "step": 360
    },
    {
      "epoch": 2.52,
      "learning_rate": 4.415807560137457e-05,
      "loss": 0.1234,
      "step": 370
    },
    {
      "epoch": 2.59,
      "learning_rate": 4.3986254295532645e-05,
      "loss": 0.101,
      "step": 380
    },
    {
      "epoch": 2.65,
      "learning_rate": 4.3814432989690723e-05,
      "loss": 0.1254,
      "step": 390
    },
    {
      "epoch": 2.72,
      "learning_rate": 4.36426116838488e-05,
      "loss": 0.1676,
      "step": 400
    },
    {
      "epoch": 2.79,
      "learning_rate": 4.3470790378006874e-05,
      "loss": 0.1297,
      "step": 410
    },
    {
      "epoch": 2.86,
      "learning_rate": 4.329896907216495e-05,
      "loss": 0.1346,
      "step": 420
    },
    {
      "epoch": 2.93,
      "learning_rate": 4.312714776632303e-05,
      "loss": 0.1123,
      "step": 430
    },
    {
      "epoch": 2.99,
      "learning_rate": 4.2955326460481096e-05,
      "loss": 0.1001,
      "step": 440
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9890710382513661,
      "eval_loss": 0.039226751774549484,
      "eval_runtime": 29.509,
      "eval_samples_per_second": 31.008,
      "eval_steps_per_second": 3.897,
      "step": 441
    },
    {
      "epoch": 3.06,
      "learning_rate": 4.2783505154639175e-05,
      "loss": 0.1042,
      "step": 450
    },
    {
      "epoch": 3.13,
      "learning_rate": 4.261168384879725e-05,
      "loss": 0.1238,
      "step": 460
    },
    {
      "epoch": 3.2,
      "learning_rate": 4.2439862542955325e-05,
      "loss": 0.0994,
      "step": 470
    },
    {
      "epoch": 3.27,
      "learning_rate": 4.2268041237113404e-05,
      "loss": 0.1055,
      "step": 480
    },
    {
      "epoch": 3.33,
      "learning_rate": 4.209621993127148e-05,
      "loss": 0.1545,
      "step": 490
    },
    {
      "epoch": 3.4,
      "learning_rate": 4.1924398625429554e-05,
      "loss": 0.0874,
      "step": 500
    },
    {
      "epoch": 3.47,
      "learning_rate": 4.175257731958763e-05,
      "loss": 0.1033,
      "step": 510
    },
    {
      "epoch": 3.54,
      "learning_rate": 4.158075601374571e-05,
      "loss": 0.1123,
      "step": 520
    },
    {
      "epoch": 3.61,
      "learning_rate": 4.140893470790378e-05,
      "loss": 0.0974,
      "step": 530
    },
    {
      "epoch": 3.67,
      "learning_rate": 4.1237113402061855e-05,
      "loss": 0.1005,
      "step": 540
    },
    {
      "epoch": 3.74,
      "learning_rate": 4.1065292096219934e-05,
      "loss": 0.095,
      "step": 550
    },
    {
      "epoch": 3.81,
      "learning_rate": 4.0893470790378005e-05,
      "loss": 0.1275,
      "step": 560
    },
    {
      "epoch": 3.88,
      "learning_rate": 4.0721649484536084e-05,
      "loss": 0.1174,
      "step": 570
    },
    {
      "epoch": 3.95,
      "learning_rate": 4.054982817869416e-05,
      "loss": 0.116,
      "step": 580
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9879781420765027,
      "eval_loss": 0.045129090547561646,
      "eval_runtime": 29.719,
      "eval_samples_per_second": 30.788,
      "eval_steps_per_second": 3.87,
      "step": 588
    },
    {
      "epoch": 4.01,
      "learning_rate": 4.0378006872852234e-05,
      "loss": 0.0961,
      "step": 590
    },
    {
      "epoch": 4.08,
      "learning_rate": 4.020618556701031e-05,
      "loss": 0.0953,
      "step": 600
    },
    {
      "epoch": 4.15,
      "learning_rate": 4.003436426116839e-05,
      "loss": 0.0921,
      "step": 610
    },
    {
      "epoch": 4.22,
      "learning_rate": 3.9862542955326463e-05,
      "loss": 0.1213,
      "step": 620
    },
    {
      "epoch": 4.29,
      "learning_rate": 3.9690721649484535e-05,
      "loss": 0.1133,
      "step": 630
    },
    {
      "epoch": 4.35,
      "learning_rate": 3.9518900343642614e-05,
      "loss": 0.1442,
      "step": 640
    },
    {
      "epoch": 4.42,
      "learning_rate": 3.9347079037800686e-05,
      "loss": 0.0827,
      "step": 650
    },
    {
      "epoch": 4.49,
      "learning_rate": 3.9175257731958764e-05,
      "loss": 0.0651,
      "step": 660
    },
    {
      "epoch": 4.56,
      "learning_rate": 3.900343642611684e-05,
      "loss": 0.0906,
      "step": 670
    },
    {
      "epoch": 4.63,
      "learning_rate": 3.8831615120274915e-05,
      "loss": 0.0607,
      "step": 680
    },
    {
      "epoch": 4.69,
      "learning_rate": 3.865979381443299e-05,
      "loss": 0.0998,
      "step": 690
    },
    {
      "epoch": 4.76,
      "learning_rate": 3.848797250859107e-05,
      "loss": 0.0882,
      "step": 700
    },
    {
      "epoch": 4.83,
      "learning_rate": 3.8316151202749144e-05,
      "loss": 0.0704,
      "step": 710
    },
    {
      "epoch": 4.9,
      "learning_rate": 3.8144329896907216e-05,
      "loss": 0.0689,
      "step": 720
    },
    {
      "epoch": 4.97,
      "learning_rate": 3.7972508591065294e-05,
      "loss": 0.1144,
      "step": 730
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9901639344262295,
      "eval_loss": 0.03983403742313385,
      "eval_runtime": 29.5173,
      "eval_samples_per_second": 30.999,
      "eval_steps_per_second": 3.896,
      "step": 735
    },
    {
      "epoch": 5.03,
      "learning_rate": 3.7800687285223366e-05,
      "loss": 0.0698,
      "step": 740
    },
    {
      "epoch": 5.1,
      "learning_rate": 3.7628865979381445e-05,
      "loss": 0.0875,
      "step": 750
    },
    {
      "epoch": 5.17,
      "learning_rate": 3.745704467353952e-05,
      "loss": 0.0932,
      "step": 760
    },
    {
      "epoch": 5.24,
      "learning_rate": 3.7285223367697595e-05,
      "loss": 0.0789,
      "step": 770
    },
    {
      "epoch": 5.31,
      "learning_rate": 3.7113402061855674e-05,
      "loss": 0.0793,
      "step": 780
    },
    {
      "epoch": 5.37,
      "learning_rate": 3.694158075601375e-05,
      "loss": 0.1196,
      "step": 790
    },
    {
      "epoch": 5.44,
      "learning_rate": 3.6769759450171824e-05,
      "loss": 0.0964,
      "step": 800
    },
    {
      "epoch": 5.51,
      "learning_rate": 3.6597938144329896e-05,
      "loss": 0.1047,
      "step": 810
    },
    {
      "epoch": 5.58,
      "learning_rate": 3.6426116838487974e-05,
      "loss": 0.0932,
      "step": 820
    },
    {
      "epoch": 5.65,
      "learning_rate": 3.6254295532646046e-05,
      "loss": 0.0789,
      "step": 830
    },
    {
      "epoch": 5.71,
      "learning_rate": 3.6082474226804125e-05,
      "loss": 0.0619,
      "step": 840
    },
    {
      "epoch": 5.78,
      "learning_rate": 3.5910652920962203e-05,
      "loss": 0.0654,
      "step": 850
    },
    {
      "epoch": 5.85,
      "learning_rate": 3.5738831615120275e-05,
      "loss": 0.1124,
      "step": 860
    },
    {
      "epoch": 5.92,
      "learning_rate": 3.5567010309278354e-05,
      "loss": 0.1295,
      "step": 870
    },
    {
      "epoch": 5.99,
      "learning_rate": 3.539518900343643e-05,
      "loss": 0.0787,
      "step": 880
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9901639344262295,
      "eval_loss": 0.04411642625927925,
      "eval_runtime": 29.3848,
      "eval_samples_per_second": 31.139,
      "eval_steps_per_second": 3.914,
      "step": 882
    },
    {
      "epoch": 6.05,
      "learning_rate": 3.5223367697594504e-05,
      "loss": 0.0664,
      "step": 890
    },
    {
      "epoch": 6.12,
      "learning_rate": 3.5051546391752576e-05,
      "loss": 0.1042,
      "step": 900
    },
    {
      "epoch": 6.19,
      "learning_rate": 3.4879725085910655e-05,
      "loss": 0.0516,
      "step": 910
    },
    {
      "epoch": 6.26,
      "learning_rate": 3.4707903780068726e-05,
      "loss": 0.1106,
      "step": 920
    },
    {
      "epoch": 6.33,
      "learning_rate": 3.4536082474226805e-05,
      "loss": 0.1176,
      "step": 930
    },
    {
      "epoch": 6.39,
      "learning_rate": 3.4364261168384884e-05,
      "loss": 0.1239,
      "step": 940
    },
    {
      "epoch": 6.46,
      "learning_rate": 3.4192439862542955e-05,
      "loss": 0.0996,
      "step": 950
    },
    {
      "epoch": 6.53,
      "learning_rate": 3.4020618556701034e-05,
      "loss": 0.0705,
      "step": 960
    },
    {
      "epoch": 6.6,
      "learning_rate": 3.384879725085911e-05,
      "loss": 0.0674,
      "step": 970
    },
    {
      "epoch": 6.67,
      "learning_rate": 3.3676975945017185e-05,
      "loss": 0.0857,
      "step": 980
    },
    {
      "epoch": 6.73,
      "learning_rate": 3.3505154639175256e-05,
      "loss": 0.1103,
      "step": 990
    },
    {
      "epoch": 6.8,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.0778,
      "step": 1000
    },
    {
      "epoch": 6.87,
      "learning_rate": 3.316151202749141e-05,
      "loss": 0.0963,
      "step": 1010
    },
    {
      "epoch": 6.94,
      "learning_rate": 3.2989690721649485e-05,
      "loss": 0.0998,
      "step": 1020
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9901639344262295,
      "eval_loss": 0.03199444338679314,
      "eval_runtime": 30.1676,
      "eval_samples_per_second": 30.331,
      "eval_steps_per_second": 3.812,
      "step": 1029
    },
    {
      "epoch": 7.01,
      "learning_rate": 3.2817869415807564e-05,
      "loss": 0.0896,
      "step": 1030
    },
    {
      "epoch": 7.07,
      "learning_rate": 3.2646048109965636e-05,
      "loss": 0.0835,
      "step": 1040
    },
    {
      "epoch": 7.14,
      "learning_rate": 3.2474226804123714e-05,
      "loss": 0.0665,
      "step": 1050
    },
    {
      "epoch": 7.21,
      "learning_rate": 3.230240549828179e-05,
      "loss": 0.0733,
      "step": 1060
    },
    {
      "epoch": 7.28,
      "learning_rate": 3.2130584192439865e-05,
      "loss": 0.0852,
      "step": 1070
    },
    {
      "epoch": 7.35,
      "learning_rate": 3.1958762886597937e-05,
      "loss": 0.0618,
      "step": 1080
    },
    {
      "epoch": 7.41,
      "learning_rate": 3.1786941580756015e-05,
      "loss": 0.1017,
      "step": 1090
    },
    {
      "epoch": 7.48,
      "learning_rate": 3.161512027491409e-05,
      "loss": 0.0451,
      "step": 1100
    },
    {
      "epoch": 7.55,
      "learning_rate": 3.1443298969072166e-05,
      "loss": 0.0663,
      "step": 1110
    },
    {
      "epoch": 7.62,
      "learning_rate": 3.1271477663230244e-05,
      "loss": 0.1017,
      "step": 1120
    },
    {
      "epoch": 7.69,
      "learning_rate": 3.1099656357388316e-05,
      "loss": 0.1026,
      "step": 1130
    },
    {
      "epoch": 7.76,
      "learning_rate": 3.0927835051546395e-05,
      "loss": 0.0812,
      "step": 1140
    },
    {
      "epoch": 7.82,
      "learning_rate": 3.075601374570447e-05,
      "loss": 0.0913,
      "step": 1150
    },
    {
      "epoch": 7.89,
      "learning_rate": 3.0584192439862545e-05,
      "loss": 0.0783,
      "step": 1160
    },
    {
      "epoch": 7.96,
      "learning_rate": 3.0412371134020617e-05,
      "loss": 0.124,
      "step": 1170
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9901639344262295,
      "eval_loss": 0.0364375114440918,
      "eval_runtime": 29.8491,
      "eval_samples_per_second": 30.654,
      "eval_steps_per_second": 3.853,
      "step": 1176
    },
    {
      "epoch": 8.03,
      "learning_rate": 3.0240549828178692e-05,
      "loss": 0.049,
      "step": 1180
    },
    {
      "epoch": 8.1,
      "learning_rate": 3.006872852233677e-05,
      "loss": 0.0884,
      "step": 1190
    },
    {
      "epoch": 8.16,
      "learning_rate": 2.9896907216494846e-05,
      "loss": 0.0843,
      "step": 1200
    },
    {
      "epoch": 8.23,
      "learning_rate": 2.972508591065292e-05,
      "loss": 0.0892,
      "step": 1210
    },
    {
      "epoch": 8.3,
      "learning_rate": 2.9553264604811e-05,
      "loss": 0.0468,
      "step": 1220
    },
    {
      "epoch": 8.37,
      "learning_rate": 2.9381443298969075e-05,
      "loss": 0.0669,
      "step": 1230
    },
    {
      "epoch": 8.44,
      "learning_rate": 2.920962199312715e-05,
      "loss": 0.067,
      "step": 1240
    },
    {
      "epoch": 8.5,
      "learning_rate": 2.903780068728523e-05,
      "loss": 0.0585,
      "step": 1250
    },
    {
      "epoch": 8.57,
      "learning_rate": 2.8865979381443297e-05,
      "loss": 0.032,
      "step": 1260
    },
    {
      "epoch": 8.64,
      "learning_rate": 2.8694158075601372e-05,
      "loss": 0.1195,
      "step": 1270
    },
    {
      "epoch": 8.71,
      "learning_rate": 2.852233676975945e-05,
      "loss": 0.0733,
      "step": 1280
    },
    {
      "epoch": 8.78,
      "learning_rate": 2.8350515463917526e-05,
      "loss": 0.0664,
      "step": 1290
    },
    {
      "epoch": 8.84,
      "learning_rate": 2.81786941580756e-05,
      "loss": 0.0792,
      "step": 1300
    },
    {
      "epoch": 8.91,
      "learning_rate": 2.800687285223368e-05,
      "loss": 0.0866,
      "step": 1310
    },
    {
      "epoch": 8.98,
      "learning_rate": 2.7835051546391755e-05,
      "loss": 0.103,
      "step": 1320
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.9879781420765027,
      "eval_loss": 0.039538685232400894,
      "eval_runtime": 30.6734,
      "eval_samples_per_second": 29.83,
      "eval_steps_per_second": 3.749,
      "step": 1323
    },
    {
      "epoch": 9.05,
      "learning_rate": 2.766323024054983e-05,
      "loss": 0.0559,
      "step": 1330
    },
    {
      "epoch": 9.12,
      "learning_rate": 2.749140893470791e-05,
      "loss": 0.0774,
      "step": 1340
    },
    {
      "epoch": 9.18,
      "learning_rate": 2.7319587628865977e-05,
      "loss": 0.0809,
      "step": 1350
    },
    {
      "epoch": 9.25,
      "learning_rate": 2.7147766323024053e-05,
      "loss": 0.0984,
      "step": 1360
    },
    {
      "epoch": 9.32,
      "learning_rate": 2.697594501718213e-05,
      "loss": 0.0838,
      "step": 1370
    },
    {
      "epoch": 9.39,
      "learning_rate": 2.6804123711340206e-05,
      "loss": 0.0894,
      "step": 1380
    },
    {
      "epoch": 9.46,
      "learning_rate": 2.663230240549828e-05,
      "loss": 0.0501,
      "step": 1390
    },
    {
      "epoch": 9.52,
      "learning_rate": 2.646048109965636e-05,
      "loss": 0.0829,
      "step": 1400
    },
    {
      "epoch": 9.59,
      "learning_rate": 2.6288659793814435e-05,
      "loss": 0.0885,
      "step": 1410
    },
    {
      "epoch": 9.66,
      "learning_rate": 2.611683848797251e-05,
      "loss": 0.0925,
      "step": 1420
    },
    {
      "epoch": 9.73,
      "learning_rate": 2.594501718213059e-05,
      "loss": 0.0645,
      "step": 1430
    },
    {
      "epoch": 9.8,
      "learning_rate": 2.5773195876288658e-05,
      "loss": 0.091,
      "step": 1440
    },
    {
      "epoch": 9.86,
      "learning_rate": 2.5601374570446733e-05,
      "loss": 0.0785,
      "step": 1450
    },
    {
      "epoch": 9.93,
      "learning_rate": 2.542955326460481e-05,
      "loss": 0.0794,
      "step": 1460
    },
    {
      "epoch": 10.0,
      "learning_rate": 2.5257731958762887e-05,
      "loss": 0.0591,
      "step": 1470
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.9912568306010929,
      "eval_loss": 0.029904531314969063,
      "eval_runtime": 31.3335,
      "eval_samples_per_second": 29.202,
      "eval_steps_per_second": 3.67,
      "step": 1470
    },
    {
      "epoch": 10.07,
      "learning_rate": 2.5085910652920962e-05,
      "loss": 0.0475,
      "step": 1480
    },
    {
      "epoch": 10.14,
      "learning_rate": 2.491408934707904e-05,
      "loss": 0.0637,
      "step": 1490
    },
    {
      "epoch": 10.2,
      "learning_rate": 2.4742268041237116e-05,
      "loss": 0.0445,
      "step": 1500
    },
    {
      "epoch": 10.27,
      "learning_rate": 2.4570446735395188e-05,
      "loss": 0.0421,
      "step": 1510
    },
    {
      "epoch": 10.34,
      "learning_rate": 2.4398625429553266e-05,
      "loss": 0.0807,
      "step": 1520
    },
    {
      "epoch": 10.41,
      "learning_rate": 2.422680412371134e-05,
      "loss": 0.0552,
      "step": 1530
    },
    {
      "epoch": 10.48,
      "learning_rate": 2.4054982817869417e-05,
      "loss": 0.0447,
      "step": 1540
    },
    {
      "epoch": 10.54,
      "learning_rate": 2.3883161512027495e-05,
      "loss": 0.054,
      "step": 1550
    },
    {
      "epoch": 10.61,
      "learning_rate": 2.3711340206185567e-05,
      "loss": 0.0445,
      "step": 1560
    },
    {
      "epoch": 10.68,
      "learning_rate": 2.3539518900343642e-05,
      "loss": 0.0747,
      "step": 1570
    },
    {
      "epoch": 10.75,
      "learning_rate": 2.336769759450172e-05,
      "loss": 0.0705,
      "step": 1580
    },
    {
      "epoch": 10.82,
      "learning_rate": 2.3195876288659796e-05,
      "loss": 0.0846,
      "step": 1590
    },
    {
      "epoch": 10.88,
      "learning_rate": 2.3024054982817868e-05,
      "loss": 0.0832,
      "step": 1600
    },
    {
      "epoch": 10.95,
      "learning_rate": 2.2852233676975946e-05,
      "loss": 0.0445,
      "step": 1610
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.9912568306010929,
      "eval_loss": 0.030193105340003967,
      "eval_runtime": 29.7828,
      "eval_samples_per_second": 30.722,
      "eval_steps_per_second": 3.861,
      "step": 1617
    },
    {
      "epoch": 11.02,
      "learning_rate": 2.268041237113402e-05,
      "loss": 0.0648,
      "step": 1620
    },
    {
      "epoch": 11.09,
      "learning_rate": 2.2508591065292097e-05,
      "loss": 0.0355,
      "step": 1630
    },
    {
      "epoch": 11.16,
      "learning_rate": 2.2336769759450175e-05,
      "loss": 0.0566,
      "step": 1640
    },
    {
      "epoch": 11.22,
      "learning_rate": 2.2164948453608247e-05,
      "loss": 0.0849,
      "step": 1650
    },
    {
      "epoch": 11.29,
      "learning_rate": 2.1993127147766322e-05,
      "loss": 0.0469,
      "step": 1660
    },
    {
      "epoch": 11.36,
      "learning_rate": 2.18213058419244e-05,
      "loss": 0.0702,
      "step": 1670
    },
    {
      "epoch": 11.43,
      "learning_rate": 2.1649484536082476e-05,
      "loss": 0.0442,
      "step": 1680
    },
    {
      "epoch": 11.5,
      "learning_rate": 2.1477663230240548e-05,
      "loss": 0.0591,
      "step": 1690
    },
    {
      "epoch": 11.56,
      "learning_rate": 2.1305841924398627e-05,
      "loss": 0.0651,
      "step": 1700
    },
    {
      "epoch": 11.63,
      "learning_rate": 2.1134020618556702e-05,
      "loss": 0.042,
      "step": 1710
    },
    {
      "epoch": 11.7,
      "learning_rate": 2.0962199312714777e-05,
      "loss": 0.0726,
      "step": 1720
    },
    {
      "epoch": 11.77,
      "learning_rate": 2.0790378006872856e-05,
      "loss": 0.0792,
      "step": 1730
    },
    {
      "epoch": 11.84,
      "learning_rate": 2.0618556701030927e-05,
      "loss": 0.0599,
      "step": 1740
    },
    {
      "epoch": 11.9,
      "learning_rate": 2.0446735395189003e-05,
      "loss": 0.0702,
      "step": 1750
    },
    {
      "epoch": 11.97,
      "learning_rate": 2.027491408934708e-05,
      "loss": 0.0684,
      "step": 1760
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.9879781420765027,
      "eval_loss": 0.03504302725195885,
      "eval_runtime": 29.3732,
      "eval_samples_per_second": 31.151,
      "eval_steps_per_second": 3.915,
      "step": 1764
    },
    {
      "epoch": 12.04,
      "learning_rate": 2.0103092783505157e-05,
      "loss": 0.0412,
      "step": 1770
    },
    {
      "epoch": 12.11,
      "learning_rate": 1.9931271477663232e-05,
      "loss": 0.0772,
      "step": 1780
    },
    {
      "epoch": 12.18,
      "learning_rate": 1.9759450171821307e-05,
      "loss": 0.0542,
      "step": 1790
    },
    {
      "epoch": 12.24,
      "learning_rate": 1.9587628865979382e-05,
      "loss": 0.0628,
      "step": 1800
    },
    {
      "epoch": 12.31,
      "learning_rate": 1.9415807560137457e-05,
      "loss": 0.0542,
      "step": 1810
    },
    {
      "epoch": 12.38,
      "learning_rate": 1.9243986254295536e-05,
      "loss": 0.061,
      "step": 1820
    },
    {
      "epoch": 12.45,
      "learning_rate": 1.9072164948453608e-05,
      "loss": 0.0461,
      "step": 1830
    },
    {
      "epoch": 12.52,
      "learning_rate": 1.8900343642611683e-05,
      "loss": 0.0551,
      "step": 1840
    },
    {
      "epoch": 12.59,
      "learning_rate": 1.872852233676976e-05,
      "loss": 0.0588,
      "step": 1850
    },
    {
      "epoch": 12.65,
      "learning_rate": 1.8556701030927837e-05,
      "loss": 0.0463,
      "step": 1860
    },
    {
      "epoch": 12.72,
      "learning_rate": 1.8384879725085912e-05,
      "loss": 0.0401,
      "step": 1870
    },
    {
      "epoch": 12.79,
      "learning_rate": 1.8213058419243987e-05,
      "loss": 0.0597,
      "step": 1880
    },
    {
      "epoch": 12.86,
      "learning_rate": 1.8041237113402062e-05,
      "loss": 0.0449,
      "step": 1890
    },
    {
      "epoch": 12.93,
      "learning_rate": 1.7869415807560138e-05,
      "loss": 0.0586,
      "step": 1900
    },
    {
      "epoch": 12.99,
      "learning_rate": 1.7697594501718216e-05,
      "loss": 0.0358,
      "step": 1910
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.9890710382513661,
      "eval_loss": 0.040750984102487564,
      "eval_runtime": 29.934,
      "eval_samples_per_second": 30.567,
      "eval_steps_per_second": 3.842,
      "step": 1911
    },
    {
      "epoch": 13.06,
      "learning_rate": 1.7525773195876288e-05,
      "loss": 0.0635,
      "step": 1920
    },
    {
      "epoch": 13.13,
      "learning_rate": 1.7353951890034363e-05,
      "loss": 0.0545,
      "step": 1930
    },
    {
      "epoch": 13.2,
      "learning_rate": 1.7182130584192442e-05,
      "loss": 0.0378,
      "step": 1940
    },
    {
      "epoch": 13.27,
      "learning_rate": 1.7010309278350517e-05,
      "loss": 0.0744,
      "step": 1950
    },
    {
      "epoch": 13.33,
      "learning_rate": 1.6838487972508592e-05,
      "loss": 0.0406,
      "step": 1960
    },
    {
      "epoch": 13.4,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 0.0506,
      "step": 1970
    },
    {
      "epoch": 13.47,
      "learning_rate": 1.6494845360824743e-05,
      "loss": 0.0676,
      "step": 1980
    },
    {
      "epoch": 13.54,
      "learning_rate": 1.6323024054982818e-05,
      "loss": 0.0553,
      "step": 1990
    },
    {
      "epoch": 13.61,
      "learning_rate": 1.6151202749140896e-05,
      "loss": 0.0594,
      "step": 2000
    },
    {
      "epoch": 13.67,
      "learning_rate": 1.5979381443298968e-05,
      "loss": 0.0378,
      "step": 2010
    },
    {
      "epoch": 13.74,
      "learning_rate": 1.5807560137457044e-05,
      "loss": 0.0576,
      "step": 2020
    },
    {
      "epoch": 13.81,
      "learning_rate": 1.5635738831615122e-05,
      "loss": 0.0621,
      "step": 2030
    },
    {
      "epoch": 13.88,
      "learning_rate": 1.5463917525773197e-05,
      "loss": 0.0611,
      "step": 2040
    },
    {
      "epoch": 13.95,
      "learning_rate": 1.5292096219931273e-05,
      "loss": 0.0548,
      "step": 2050
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.9901639344262295,
      "eval_loss": 0.038173872977495193,
      "eval_runtime": 30.2908,
      "eval_samples_per_second": 30.207,
      "eval_steps_per_second": 3.797,
      "step": 2058
    },
    {
      "epoch": 14.01,
      "learning_rate": 1.5120274914089346e-05,
      "loss": 0.0702,
      "step": 2060
    },
    {
      "epoch": 14.08,
      "learning_rate": 1.4948453608247423e-05,
      "loss": 0.0667,
      "step": 2070
    },
    {
      "epoch": 14.15,
      "learning_rate": 1.47766323024055e-05,
      "loss": 0.0563,
      "step": 2080
    },
    {
      "epoch": 14.22,
      "learning_rate": 1.4604810996563575e-05,
      "loss": 0.0601,
      "step": 2090
    },
    {
      "epoch": 14.29,
      "learning_rate": 1.4432989690721649e-05,
      "loss": 0.0535,
      "step": 2100
    },
    {
      "epoch": 14.35,
      "learning_rate": 1.4261168384879725e-05,
      "loss": 0.0753,
      "step": 2110
    },
    {
      "epoch": 14.42,
      "learning_rate": 1.40893470790378e-05,
      "loss": 0.0677,
      "step": 2120
    },
    {
      "epoch": 14.49,
      "learning_rate": 1.3917525773195878e-05,
      "loss": 0.0451,
      "step": 2130
    },
    {
      "epoch": 14.56,
      "learning_rate": 1.3745704467353954e-05,
      "loss": 0.0631,
      "step": 2140
    },
    {
      "epoch": 14.63,
      "learning_rate": 1.3573883161512026e-05,
      "loss": 0.0177,
      "step": 2150
    },
    {
      "epoch": 14.69,
      "learning_rate": 1.3402061855670103e-05,
      "loss": 0.0559,
      "step": 2160
    },
    {
      "epoch": 14.76,
      "learning_rate": 1.323024054982818e-05,
      "loss": 0.0352,
      "step": 2170
    },
    {
      "epoch": 14.83,
      "learning_rate": 1.3058419243986255e-05,
      "loss": 0.0398,
      "step": 2180
    },
    {
      "epoch": 14.9,
      "learning_rate": 1.2886597938144329e-05,
      "loss": 0.0134,
      "step": 2190
    },
    {
      "epoch": 14.97,
      "learning_rate": 1.2714776632302406e-05,
      "loss": 0.0611,
      "step": 2200
    },
    {
      "epoch": 15.0,
      "eval_accuracy": 0.9923497267759562,
      "eval_loss": 0.03305086866021156,
      "eval_runtime": 29.8991,
      "eval_samples_per_second": 30.603,
      "eval_steps_per_second": 3.846,
      "step": 2205
    },
    {
      "epoch": 15.03,
      "learning_rate": 1.2542955326460481e-05,
      "loss": 0.0722,
      "step": 2210
    },
    {
      "epoch": 15.1,
      "learning_rate": 1.2371134020618558e-05,
      "loss": 0.0704,
      "step": 2220
    },
    {
      "epoch": 15.17,
      "learning_rate": 1.2199312714776633e-05,
      "loss": 0.0528,
      "step": 2230
    },
    {
      "epoch": 15.24,
      "learning_rate": 1.2027491408934708e-05,
      "loss": 0.043,
      "step": 2240
    },
    {
      "epoch": 15.31,
      "learning_rate": 1.1855670103092783e-05,
      "loss": 0.0529,
      "step": 2250
    },
    {
      "epoch": 15.37,
      "learning_rate": 1.168384879725086e-05,
      "loss": 0.0297,
      "step": 2260
    },
    {
      "epoch": 15.44,
      "learning_rate": 1.1512027491408934e-05,
      "loss": 0.0522,
      "step": 2270
    },
    {
      "epoch": 15.51,
      "learning_rate": 1.134020618556701e-05,
      "loss": 0.0401,
      "step": 2280
    },
    {
      "epoch": 15.58,
      "learning_rate": 1.1168384879725088e-05,
      "loss": 0.0455,
      "step": 2290
    },
    {
      "epoch": 15.65,
      "learning_rate": 1.0996563573883161e-05,
      "loss": 0.0303,
      "step": 2300
    },
    {
      "epoch": 15.71,
      "learning_rate": 1.0824742268041238e-05,
      "loss": 0.0263,
      "step": 2310
    },
    {
      "epoch": 15.78,
      "learning_rate": 1.0652920962199313e-05,
      "loss": 0.1067,
      "step": 2320
    },
    {
      "epoch": 15.85,
      "learning_rate": 1.0481099656357389e-05,
      "loss": 0.049,
      "step": 2330
    },
    {
      "epoch": 15.92,
      "learning_rate": 1.0309278350515464e-05,
      "loss": 0.044,
      "step": 2340
    },
    {
      "epoch": 15.99,
      "learning_rate": 1.013745704467354e-05,
      "loss": 0.0231,
      "step": 2350
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.994535519125683,
      "eval_loss": 0.035519640892744064,
      "eval_runtime": 30.0972,
      "eval_samples_per_second": 30.401,
      "eval_steps_per_second": 3.821,
      "step": 2352
    },
    {
      "epoch": 16.05,
      "learning_rate": 9.965635738831616e-06,
      "loss": 0.0456,
      "step": 2360
    },
    {
      "epoch": 16.12,
      "learning_rate": 9.793814432989691e-06,
      "loss": 0.0517,
      "step": 2370
    },
    {
      "epoch": 16.19,
      "learning_rate": 9.621993127147768e-06,
      "loss": 0.0601,
      "step": 2380
    },
    {
      "epoch": 16.26,
      "learning_rate": 9.450171821305841e-06,
      "loss": 0.0695,
      "step": 2390
    },
    {
      "epoch": 16.33,
      "learning_rate": 9.278350515463918e-06,
      "loss": 0.0709,
      "step": 2400
    },
    {
      "epoch": 16.39,
      "learning_rate": 9.106529209621994e-06,
      "loss": 0.0691,
      "step": 2410
    },
    {
      "epoch": 16.46,
      "learning_rate": 8.934707903780069e-06,
      "loss": 0.0514,
      "step": 2420
    },
    {
      "epoch": 16.53,
      "learning_rate": 8.762886597938144e-06,
      "loss": 0.0507,
      "step": 2430
    },
    {
      "epoch": 16.6,
      "learning_rate": 8.591065292096221e-06,
      "loss": 0.0497,
      "step": 2440
    },
    {
      "epoch": 16.67,
      "learning_rate": 8.419243986254296e-06,
      "loss": 0.0539,
      "step": 2450
    },
    {
      "epoch": 16.73,
      "learning_rate": 8.247422680412371e-06,
      "loss": 0.0144,
      "step": 2460
    },
    {
      "epoch": 16.8,
      "learning_rate": 8.075601374570448e-06,
      "loss": 0.039,
      "step": 2470
    },
    {
      "epoch": 16.87,
      "learning_rate": 7.903780068728522e-06,
      "loss": 0.0495,
      "step": 2480
    },
    {
      "epoch": 16.94,
      "learning_rate": 7.731958762886599e-06,
      "loss": 0.046,
      "step": 2490
    },
    {
      "epoch": 17.0,
      "eval_accuracy": 0.9934426229508196,
      "eval_loss": 0.03209972009062767,
      "eval_runtime": 30.1657,
      "eval_samples_per_second": 30.332,
      "eval_steps_per_second": 3.812,
      "step": 2499
    },
    {
      "epoch": 17.01,
      "learning_rate": 7.560137457044673e-06,
      "loss": 0.0709,
      "step": 2500
    },
    {
      "epoch": 17.07,
      "learning_rate": 7.38831615120275e-06,
      "loss": 0.0541,
      "step": 2510
    },
    {
      "epoch": 17.14,
      "learning_rate": 7.216494845360824e-06,
      "loss": 0.0432,
      "step": 2520
    },
    {
      "epoch": 17.21,
      "learning_rate": 7.0446735395189e-06,
      "loss": 0.0525,
      "step": 2530
    },
    {
      "epoch": 17.28,
      "learning_rate": 6.872852233676977e-06,
      "loss": 0.0454,
      "step": 2540
    },
    {
      "epoch": 17.35,
      "learning_rate": 6.701030927835052e-06,
      "loss": 0.0283,
      "step": 2550
    },
    {
      "epoch": 17.41,
      "learning_rate": 6.529209621993128e-06,
      "loss": 0.0578,
      "step": 2560
    },
    {
      "epoch": 17.48,
      "learning_rate": 6.357388316151203e-06,
      "loss": 0.0652,
      "step": 2570
    },
    {
      "epoch": 17.55,
      "learning_rate": 6.185567010309279e-06,
      "loss": 0.0418,
      "step": 2580
    },
    {
      "epoch": 17.62,
      "learning_rate": 6.013745704467354e-06,
      "loss": 0.0795,
      "step": 2590
    },
    {
      "epoch": 17.69,
      "learning_rate": 5.84192439862543e-06,
      "loss": 0.0294,
      "step": 2600
    },
    {
      "epoch": 17.76,
      "learning_rate": 5.670103092783505e-06,
      "loss": 0.0702,
      "step": 2610
    },
    {
      "epoch": 17.82,
      "learning_rate": 5.498281786941581e-06,
      "loss": 0.0267,
      "step": 2620
    },
    {
      "epoch": 17.89,
      "learning_rate": 5.326460481099657e-06,
      "loss": 0.0397,
      "step": 2630
    },
    {
      "epoch": 17.96,
      "learning_rate": 5.154639175257732e-06,
      "loss": 0.0648,
      "step": 2640
    },
    {
      "epoch": 18.0,
      "eval_accuracy": 0.9923497267759562,
      "eval_loss": 0.0327494814991951,
      "eval_runtime": 30.289,
      "eval_samples_per_second": 30.209,
      "eval_steps_per_second": 3.797,
      "step": 2646
    },
    {
      "epoch": 18.03,
      "learning_rate": 4.982817869415808e-06,
      "loss": 0.033,
      "step": 2650
    },
    {
      "epoch": 18.1,
      "learning_rate": 4.810996563573884e-06,
      "loss": 0.0447,
      "step": 2660
    },
    {
      "epoch": 18.16,
      "learning_rate": 4.639175257731959e-06,
      "loss": 0.0324,
      "step": 2670
    },
    {
      "epoch": 18.23,
      "learning_rate": 4.467353951890034e-06,
      "loss": 0.0382,
      "step": 2680
    },
    {
      "epoch": 18.3,
      "learning_rate": 4.2955326460481105e-06,
      "loss": 0.0628,
      "step": 2690
    },
    {
      "epoch": 18.37,
      "learning_rate": 4.123711340206186e-06,
      "loss": 0.0808,
      "step": 2700
    },
    {
      "epoch": 18.44,
      "learning_rate": 3.951890034364261e-06,
      "loss": 0.0463,
      "step": 2710
    },
    {
      "epoch": 18.5,
      "learning_rate": 3.7800687285223365e-06,
      "loss": 0.0369,
      "step": 2720
    },
    {
      "epoch": 18.57,
      "learning_rate": 3.608247422680412e-06,
      "loss": 0.0273,
      "step": 2730
    },
    {
      "epoch": 18.64,
      "learning_rate": 3.4364261168384886e-06,
      "loss": 0.0327,
      "step": 2740
    },
    {
      "epoch": 18.71,
      "learning_rate": 3.264604810996564e-06,
      "loss": 0.0331,
      "step": 2750
    },
    {
      "epoch": 18.78,
      "learning_rate": 3.0927835051546395e-06,
      "loss": 0.0461,
      "step": 2760
    },
    {
      "epoch": 18.84,
      "learning_rate": 2.920962199312715e-06,
      "loss": 0.0525,
      "step": 2770
    },
    {
      "epoch": 18.91,
      "learning_rate": 2.7491408934707903e-06,
      "loss": 0.053,
      "step": 2780
    },
    {
      "epoch": 18.98,
      "learning_rate": 2.577319587628866e-06,
      "loss": 0.0565,
      "step": 2790
    },
    {
      "epoch": 19.0,
      "eval_accuracy": 0.9923497267759562,
      "eval_loss": 0.03197157755494118,
      "eval_runtime": 30.3146,
      "eval_samples_per_second": 30.184,
      "eval_steps_per_second": 3.794,
      "step": 2793
    },
    {
      "epoch": 19.05,
      "learning_rate": 2.405498281786942e-06,
      "loss": 0.0308,
      "step": 2800
    },
    {
      "epoch": 19.12,
      "learning_rate": 2.233676975945017e-06,
      "loss": 0.0538,
      "step": 2810
    },
    {
      "epoch": 19.18,
      "learning_rate": 2.061855670103093e-06,
      "loss": 0.0483,
      "step": 2820
    },
    {
      "epoch": 19.25,
      "learning_rate": 1.8900343642611683e-06,
      "loss": 0.0546,
      "step": 2830
    },
    {
      "epoch": 19.32,
      "learning_rate": 1.7182130584192443e-06,
      "loss": 0.0346,
      "step": 2840
    },
    {
      "epoch": 19.39,
      "learning_rate": 1.5463917525773197e-06,
      "loss": 0.0423,
      "step": 2850
    },
    {
      "epoch": 19.46,
      "learning_rate": 1.3745704467353952e-06,
      "loss": 0.0241,
      "step": 2860
    },
    {
      "epoch": 19.52,
      "learning_rate": 1.202749140893471e-06,
      "loss": 0.0426,
      "step": 2870
    },
    {
      "epoch": 19.59,
      "learning_rate": 1.0309278350515464e-06,
      "loss": 0.0409,
      "step": 2880
    },
    {
      "epoch": 19.66,
      "learning_rate": 8.591065292096222e-07,
      "loss": 0.0323,
      "step": 2890
    },
    {
      "epoch": 19.73,
      "learning_rate": 6.872852233676976e-07,
      "loss": 0.0261,
      "step": 2900
    },
    {
      "epoch": 19.8,
      "learning_rate": 5.154639175257732e-07,
      "loss": 0.04,
      "step": 2910
    },
    {
      "epoch": 19.86,
      "learning_rate": 3.436426116838488e-07,
      "loss": 0.0575,
      "step": 2920
    },
    {
      "epoch": 19.93,
      "learning_rate": 1.718213058419244e-07,
      "loss": 0.0495,
      "step": 2930
    },
    {
      "epoch": 20.0,
      "learning_rate": 0.0,
      "loss": 0.0413,
      "step": 2940
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.9923497267759562,
      "eval_loss": 0.032683759927749634,
      "eval_runtime": 29.7008,
      "eval_samples_per_second": 30.807,
      "eval_steps_per_second": 3.872,
      "step": 2940
    },
    {
      "epoch": 20.0,
      "step": 2940,
      "total_flos": 4.089918706717778e+18,
      "train_loss": 0.08860804434029423,
      "train_runtime": 6732.3279,
      "train_samples_per_second": 24.44,
      "train_steps_per_second": 0.437
    }
  ],
  "max_steps": 2940,
  "num_train_epochs": 20,
  "total_flos": 4.089918706717778e+18,
  "trial_name": null,
  "trial_params": null
}