{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.9992034411343,
  "global_step": 3920,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "learning_rate": 3.7499999999999997e-06,
      "loss": 12.5263,
      "step": 10
    },
    {
      "epoch": 0.03,
      "learning_rate": 1.05e-05,
      "loss": 9.3683,
      "step": 20
    },
    {
      "epoch": 0.04,
      "learning_rate": 1.7999999999999997e-05,
      "loss": 6.2513,
      "step": 30
    },
    {
      "epoch": 0.05,
      "learning_rate": 2.55e-05,
      "loss": 5.2618,
      "step": 40
    },
    {
      "epoch": 0.06,
      "learning_rate": 3.2999999999999996e-05,
      "loss": 4.6846,
      "step": 50
    },
    {
      "epoch": 0.08,
      "learning_rate": 4.05e-05,
      "loss": 5.3931,
      "step": 60
    },
    {
      "epoch": 0.09,
      "learning_rate": 4.7999999999999994e-05,
      "loss": 4.5171,
      "step": 70
    },
    {
      "epoch": 0.1,
      "learning_rate": 5.5499999999999994e-05,
      "loss": 4.2914,
      "step": 80
    },
    {
      "epoch": 0.11,
      "learning_rate": 6.299999999999999e-05,
      "loss": 4.0871,
      "step": 90
    },
    {
      "epoch": 0.13,
      "learning_rate": 7.049999999999999e-05,
      "loss": 3.7123,
      "step": 100
    },
    {
      "epoch": 0.14,
      "learning_rate": 7.8e-05,
      "loss": 4.4868,
      "step": 110
    },
    {
      "epoch": 0.15,
      "learning_rate": 8.549999999999999e-05,
      "loss": 3.8837,
      "step": 120
    },
    {
      "epoch": 0.17,
      "learning_rate": 9.3e-05,
      "loss": 3.7156,
      "step": 130
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0001005,
      "loss": 3.6339,
      "step": 140
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.00010799999999999998,
      "loss": 3.1646,
      "step": 150
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.00011549999999999999,
      "loss": 4.1434,
      "step": 160
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.00012299999999999998,
      "loss": 3.612,
      "step": 170
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0001305,
      "loss": 3.5163,
      "step": 180
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.000138,
      "loss": 3.4249,
      "step": 190
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.00014549999999999999,
      "loss": 2.9259,
      "step": 200
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.00015299999999999998,
      "loss": 3.9771,
      "step": 210
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0001605,
      "loss": 3.4041,
      "step": 220
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.000168,
      "loss": 3.3504,
      "step": 230
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.00017549999999999998,
      "loss": 3.2431,
      "step": 240
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.00018299999999999998,
      "loss": 2.674,
      "step": 250
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0001905,
      "loss": 3.6828,
      "step": 260
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.000198,
      "loss": 3.3201,
      "step": 270
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002055,
      "loss": 3.1553,
      "step": 280
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.00021299999999999997,
      "loss": 3.1109,
      "step": 290
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.00022049999999999997,
      "loss": 2.5012,
      "step": 300
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.00022799999999999999,
      "loss": 3.6064,
      "step": 310
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.00023549999999999998,
      "loss": 3.2355,
      "step": 320
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.000243,
      "loss": 3.1318,
      "step": 330
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.00025049999999999996,
      "loss": 3.0418,
      "step": 340
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.000258,
      "loss": 2.3601,
      "step": 350
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002655,
      "loss": 3.6744,
      "step": 360
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.00027299999999999997,
      "loss": 3.1284,
      "step": 370
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002805,
      "loss": 3.0404,
      "step": 380
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.00028799999999999995,
      "loss": 2.8827,
      "step": 390
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.00029549999999999997,
      "loss": 2.3036,
      "step": 400
    },
    {
      "epoch": 0.51,
      "eval_loss": 4.219922065734863,
      "eval_runtime": 978.7091,
      "eval_samples_per_second": 6.369,
      "eval_steps_per_second": 0.797,
      "eval_wer": 0.9499730488280888,
      "step": 400
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002996590909090909,
      "loss": 3.7398,
      "step": 410
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002988068181818181,
      "loss": 3.1013,
      "step": 420
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002979545454545454,
      "loss": 2.9651,
      "step": 430
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002971022727272727,
      "loss": 2.8712,
      "step": 440
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.00029624999999999996,
      "loss": 2.1188,
      "step": 450
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.00029539772727272726,
      "loss": 3.7197,
      "step": 460
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002945454545454545,
      "loss": 3.0581,
      "step": 470
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002936931818181818,
      "loss": 2.9541,
      "step": 480
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002928409090909091,
      "loss": 2.7738,
      "step": 490
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.00029198863636363634,
      "loss": 2.1907,
      "step": 500
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002911363636363636,
      "loss": 3.617,
      "step": 510
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002902840909090909,
      "loss": 2.9914,
      "step": 520
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.00028943181818181813,
      "loss": 2.9248,
      "step": 530
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002885795454545454,
      "loss": 2.7295,
      "step": 540
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002877272727272727,
      "loss": 2.1081,
      "step": 550
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.00028687499999999997,
      "loss": 3.4299,
      "step": 560
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002860227272727272,
      "loss": 2.917,
      "step": 570
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002851704545454545,
      "loss": 2.8774,
      "step": 580
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002843181818181818,
      "loss": 2.7049,
      "step": 590
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002834659090909091,
      "loss": 2.0966,
      "step": 600
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.00028261363636363635,
      "loss": 3.381,
      "step": 610
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002817613636363636,
      "loss": 2.9034,
      "step": 620
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002809090909090909,
      "loss": 2.8067,
      "step": 630
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.00028005681818181813,
      "loss": 2.6366,
      "step": 640
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.00027920454545454543,
      "loss": 2.057,
      "step": 650
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002783522727272727,
      "loss": 3.3471,
      "step": 660
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.00027749999999999997,
      "loss": 2.895,
      "step": 670
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002766477272727272,
      "loss": 2.7621,
      "step": 680
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002757954545454545,
      "loss": 2.6106,
      "step": 690
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002749431818181818,
      "loss": 1.965,
      "step": 700
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002740909090909091,
      "loss": 3.4336,
      "step": 710
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.00027323863636363635,
      "loss": 2.9011,
      "step": 720
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002723863636363636,
      "loss": 2.7347,
      "step": 730
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002715340909090909,
      "loss": 2.6067,
      "step": 740
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.00027068181818181813,
      "loss": 2.0804,
      "step": 750
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.00026982954545454543,
      "loss": 3.5563,
      "step": 760
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.00026897727272727273,
      "loss": 2.8215,
      "step": 770
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.00026812499999999997,
      "loss": 2.5426,
      "step": 780
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002672727272727272,
      "loss": 2.848,
      "step": 790
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002664204545454545,
      "loss": 2.836,
      "step": 800
    },
    {
      "epoch": 1.02,
      "eval_loss": 2.7747511863708496,
      "eval_runtime": 1386.8434,
      "eval_samples_per_second": 4.494,
      "eval_steps_per_second": 0.562,
      "eval_wer": 1.2757569561904054,
      "step": 800
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002655681818181818,
      "loss": 2.6687,
      "step": 810
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002647159090909091,
      "loss": 2.5294,
      "step": 820
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.00026386363636363635,
      "loss": 2.2344,
      "step": 830
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002630113636363636,
      "loss": 2.8092,
      "step": 840
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002621590909090909,
      "loss": 2.8033,
      "step": 850
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.00026130681818181814,
      "loss": 2.6581,
      "step": 860
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.00026045454545454543,
      "loss": 2.5385,
      "step": 870
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.00025960227272727273,
      "loss": 2.2285,
      "step": 880
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.00025875,
      "loss": 2.7362,
      "step": 890
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002578977272727272,
      "loss": 2.7753,
      "step": 900
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002570454545454545,
      "loss": 2.6267,
      "step": 910
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002561931818181818,
      "loss": 2.4796,
      "step": 920
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002553409090909091,
      "loss": 2.1629,
      "step": 930
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.00025448863636363635,
      "loss": 2.7308,
      "step": 940
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002536363636363636,
      "loss": 2.8126,
      "step": 950
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002527840909090909,
      "loss": 2.6516,
      "step": 960
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.00025193181818181814,
      "loss": 2.4509,
      "step": 970
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.00025107954545454544,
      "loss": 2.1525,
      "step": 980
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.00025022727272727273,
      "loss": 2.6477,
      "step": 990
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.000249375,
      "loss": 2.7505,
      "step": 1000
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002485227272727272,
      "loss": 2.629,
      "step": 1010
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002476704545454545,
      "loss": 2.4232,
      "step": 1020
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002468181818181818,
      "loss": 2.0637,
      "step": 1030
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002459659090909091,
      "loss": 2.6852,
      "step": 1040
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.00024511363636363636,
      "loss": 2.753,
      "step": 1050
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002442613636363636,
      "loss": 2.5709,
      "step": 1060
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002434090909090909,
      "loss": 2.4777,
      "step": 1070
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.00024255681818181814,
      "loss": 2.1415,
      "step": 1080
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.00024170454545454544,
      "loss": 2.5522,
      "step": 1090
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002408522727272727,
      "loss": 2.6903,
      "step": 1100
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.00023999999999999998,
      "loss": 2.5554,
      "step": 1110
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.00023914772727272725,
      "loss": 2.4458,
      "step": 1120
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.00023829545454545452,
      "loss": 2.1449,
      "step": 1130
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.00023744318181818182,
      "loss": 2.6099,
      "step": 1140
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002365909090909091,
      "loss": 2.6866,
      "step": 1150
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.00023573863636363633,
      "loss": 2.5955,
      "step": 1160
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.00023488636363636363,
      "loss": 2.4689,
      "step": 1170
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002340340909090909,
      "loss": 2.0902,
      "step": 1180
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.00023318181818181814,
      "loss": 2.4791,
      "step": 1190
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.00023232954545454544,
      "loss": 2.7021,
      "step": 1200
    },
    {
      "epoch": 1.53,
      "eval_loss": 2.621112585067749,
      "eval_runtime": 1372.4396,
      "eval_samples_per_second": 4.542,
      "eval_steps_per_second": 0.568,
      "eval_wer": 1.1724967937398934,
      "step": 1200
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002314772727272727,
      "loss": 2.5888,
      "step": 1210
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.00023062499999999998,
      "loss": 2.4004,
      "step": 1220
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.00022977272727272725,
      "loss": 2.0814,
      "step": 1230
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.00022892045454545452,
      "loss": 2.5302,
      "step": 1240
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002280681818181818,
      "loss": 2.6644,
      "step": 1250
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002272159090909091,
      "loss": 2.5413,
      "step": 1260
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.00022636363636363633,
      "loss": 2.3677,
      "step": 1270
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002255113636363636,
      "loss": 2.0107,
      "step": 1280
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002246590909090909,
      "loss": 2.3977,
      "step": 1290
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.00022380681818181815,
      "loss": 2.6459,
      "step": 1300
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.00022295454545454544,
      "loss": 2.5077,
      "step": 1310
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002221022727272727,
      "loss": 2.3272,
      "step": 1320
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.00022124999999999998,
      "loss": 1.9862,
      "step": 1330
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.00022039772727272725,
      "loss": 2.3923,
      "step": 1340
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.00021954545454545452,
      "loss": 2.6402,
      "step": 1350
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002186931818181818,
      "loss": 2.4859,
      "step": 1360
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002178409090909091,
      "loss": 2.3107,
      "step": 1370
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.00021698863636363634,
      "loss": 2.0173,
      "step": 1380
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002161363636363636,
      "loss": 2.3328,
      "step": 1390
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002152840909090909,
      "loss": 2.5905,
      "step": 1400
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.00021443181818181815,
      "loss": 2.4717,
      "step": 1410
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.00021357954545454542,
      "loss": 2.39,
      "step": 1420
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.00021272727272727272,
      "loss": 1.9524,
      "step": 1430
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.000211875,
      "loss": 2.365,
      "step": 1440
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.00021102272727272723,
      "loss": 2.558,
      "step": 1450
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.00021017045454545453,
      "loss": 2.4819,
      "step": 1460
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002093181818181818,
      "loss": 2.3133,
      "step": 1470
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002084659090909091,
      "loss": 1.9815,
      "step": 1480
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.00020761363636363634,
      "loss": 2.4231,
      "step": 1490
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002067613636363636,
      "loss": 2.6569,
      "step": 1500
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002059090909090909,
      "loss": 2.4576,
      "step": 1510
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.00020505681818181815,
      "loss": 2.3225,
      "step": 1520
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.00020420454545454542,
      "loss": 2.0289,
      "step": 1530
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.00020335227272727272,
      "loss": 2.3312,
      "step": 1540
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002025,
      "loss": 2.5694,
      "step": 1550
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.00020164772727272723,
      "loss": 2.4322,
      "step": 1560
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.00020079545454545453,
      "loss": 2.1472,
      "step": 1570
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0001999431818181818,
      "loss": 2.6277,
      "step": 1580
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0001990909090909091,
      "loss": 2.3966,
      "step": 1590
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.00019823863636363634,
      "loss": 2.2322,
      "step": 1600
    },
    {
      "epoch": 2.04,
      "eval_loss": 2.5714354515075684,
      "eval_runtime": 1370.556,
      "eval_samples_per_second": 4.548,
      "eval_steps_per_second": 0.569,
      "eval_wer": 1.1041616326834074,
      "step": 1600
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0001973863636363636,
      "loss": 1.9566,
      "step": 1610
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0001965340909090909,
      "loss": 1.7948,
      "step": 1620
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.00019568181818181815,
      "loss": 2.6895,
      "step": 1630
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.00019482954545454542,
      "loss": 2.3899,
      "step": 1640
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.00019397727272727272,
      "loss": 2.2251,
      "step": 1650
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.000193125,
      "loss": 1.9633,
      "step": 1660
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.00019227272727272723,
      "loss": 1.7536,
      "step": 1670
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.00019142045454545453,
      "loss": 2.684,
      "step": 1680
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0001905681818181818,
      "loss": 2.448,
      "step": 1690
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0001897159090909091,
      "loss": 2.2632,
      "step": 1700
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.00018886363636363634,
      "loss": 2.0362,
      "step": 1710
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.00018801136363636361,
      "loss": 1.8685,
      "step": 1720
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0001871590909090909,
      "loss": 2.6846,
      "step": 1730
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.00018630681818181816,
      "loss": 2.3884,
      "step": 1740
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.00018545454545454543,
      "loss": 2.337,
      "step": 1750
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.00018460227272727272,
      "loss": 1.965,
      "step": 1760
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.00018375,
      "loss": 1.7979,
      "step": 1770
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.00018289772727272724,
      "loss": 2.6436,
      "step": 1780
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.00018204545454545454,
      "loss": 2.4121,
      "step": 1790
    },
    {
      "epoch": 2.3,
      "learning_rate": 0.0001811931818181818,
      "loss": 2.2591,
      "step": 1800
    },
    {
      "epoch": 2.31,
      "learning_rate": 0.0001803409090909091,
      "loss": 2.0358,
      "step": 1810
    },
    {
      "epoch": 2.32,
      "learning_rate": 0.00017948863636363635,
      "loss": 1.8178,
      "step": 1820
    },
    {
      "epoch": 2.33,
      "learning_rate": 0.00017863636363636362,
      "loss": 2.6027,
      "step": 1830
    },
    {
      "epoch": 2.35,
      "learning_rate": 0.00017778409090909091,
      "loss": 2.3807,
      "step": 1840
    },
    {
      "epoch": 2.36,
      "learning_rate": 0.00017693181818181816,
      "loss": 2.2269,
      "step": 1850
    },
    {
      "epoch": 2.37,
      "learning_rate": 0.00017607954545454543,
      "loss": 1.9806,
      "step": 1860
    },
    {
      "epoch": 2.38,
      "learning_rate": 0.00017522727272727273,
      "loss": 1.7612,
      "step": 1870
    },
    {
      "epoch": 2.4,
      "learning_rate": 0.000174375,
      "loss": 2.5986,
      "step": 1880
    },
    {
      "epoch": 2.41,
      "learning_rate": 0.00017352272727272724,
      "loss": 2.3444,
      "step": 1890
    },
    {
      "epoch": 2.42,
      "learning_rate": 0.00017267045454545454,
      "loss": 2.1959,
      "step": 1900
    },
    {
      "epoch": 2.44,
      "learning_rate": 0.0001718181818181818,
      "loss": 1.8388,
      "step": 1910
    },
    {
      "epoch": 2.45,
      "learning_rate": 0.0001709659090909091,
      "loss": 1.7236,
      "step": 1920
    },
    {
      "epoch": 2.46,
      "learning_rate": 0.00017011363636363635,
      "loss": 2.6009,
      "step": 1930
    },
    {
      "epoch": 2.47,
      "learning_rate": 0.00016926136363636362,
      "loss": 2.3328,
      "step": 1940
    },
    {
      "epoch": 2.49,
      "learning_rate": 0.00016840909090909092,
      "loss": 2.134,
      "step": 1950
    },
    {
      "epoch": 2.5,
      "learning_rate": 0.00016755681818181816,
      "loss": 1.8768,
      "step": 1960
    },
    {
      "epoch": 2.51,
      "learning_rate": 0.00016670454545454543,
      "loss": 1.6798,
      "step": 1970
    },
    {
      "epoch": 2.53,
      "learning_rate": 0.00016585227272727273,
      "loss": 2.5955,
      "step": 1980
    },
    {
      "epoch": 2.54,
      "learning_rate": 0.000165,
      "loss": 2.3165,
      "step": 1990
    },
    {
      "epoch": 2.55,
      "learning_rate": 0.00016414772727272724,
      "loss": 2.1435,
      "step": 2000
    },
    {
      "epoch": 2.55,
      "eval_loss": 2.4204635620117188,
      "eval_runtime": 1359.7511,
      "eval_samples_per_second": 4.584,
      "eval_steps_per_second": 0.574,
      "eval_wer": 1.0092656270329547,
      "step": 2000
    },
    {
      "epoch": 2.56,
      "learning_rate": 0.00016329545454545454,
      "loss": 1.8971,
      "step": 2010
    },
    {
      "epoch": 2.58,
      "learning_rate": 0.0001624431818181818,
      "loss": 1.6731,
      "step": 2020
    },
    {
      "epoch": 2.59,
      "learning_rate": 0.00016159090909090908,
      "loss": 2.6731,
      "step": 2030
    },
    {
      "epoch": 2.6,
      "learning_rate": 0.00016073863636363635,
      "loss": 2.3067,
      "step": 2040
    },
    {
      "epoch": 2.61,
      "learning_rate": 0.00015988636363636362,
      "loss": 2.1082,
      "step": 2050
    },
    {
      "epoch": 2.63,
      "learning_rate": 0.0001590340909090909,
      "loss": 1.849,
      "step": 2060
    },
    {
      "epoch": 2.64,
      "learning_rate": 0.00015818181818181816,
      "loss": 1.6493,
      "step": 2070
    },
    {
      "epoch": 2.65,
      "learning_rate": 0.00015732954545454543,
      "loss": 2.6306,
      "step": 2080
    },
    {
      "epoch": 2.67,
      "learning_rate": 0.00015647727272727273,
      "loss": 2.269,
      "step": 2090
    },
    {
      "epoch": 2.68,
      "learning_rate": 0.000155625,
      "loss": 2.1135,
      "step": 2100
    },
    {
      "epoch": 2.69,
      "learning_rate": 0.00015477272727272725,
      "loss": 1.8049,
      "step": 2110
    },
    {
      "epoch": 2.7,
      "learning_rate": 0.00015392045454545454,
      "loss": 1.695,
      "step": 2120
    },
    {
      "epoch": 2.72,
      "learning_rate": 0.0001530681818181818,
      "loss": 2.5111,
      "step": 2130
    },
    {
      "epoch": 2.73,
      "learning_rate": 0.00015221590909090908,
      "loss": 2.3195,
      "step": 2140
    },
    {
      "epoch": 2.74,
      "learning_rate": 0.00015136363636363635,
      "loss": 2.077,
      "step": 2150
    },
    {
      "epoch": 2.75,
      "learning_rate": 0.00015051136363636362,
      "loss": 1.8057,
      "step": 2160
    },
    {
      "epoch": 2.77,
      "learning_rate": 0.0001496590909090909,
      "loss": 1.5783,
      "step": 2170
    },
    {
      "epoch": 2.78,
      "learning_rate": 0.00014880681818181817,
      "loss": 2.5461,
      "step": 2180
    },
    {
      "epoch": 2.79,
      "learning_rate": 0.00014795454545454544,
      "loss": 2.2407,
      "step": 2190
    },
    {
      "epoch": 2.81,
      "learning_rate": 0.0001471022727272727,
      "loss": 2.0751,
      "step": 2200
    },
    {
      "epoch": 2.82,
      "learning_rate": 0.00014624999999999998,
      "loss": 1.6996,
      "step": 2210
    },
    {
      "epoch": 2.83,
      "learning_rate": 0.00014539772727272727,
      "loss": 1.6127,
      "step": 2220
    },
    {
      "epoch": 2.84,
      "learning_rate": 0.00014454545454545452,
      "loss": 2.4968,
      "step": 2230
    },
    {
      "epoch": 2.86,
      "learning_rate": 0.00014369318181818182,
      "loss": 2.213,
      "step": 2240
    },
    {
      "epoch": 2.87,
      "learning_rate": 0.00014284090909090909,
      "loss": 2.0047,
      "step": 2250
    },
    {
      "epoch": 2.88,
      "learning_rate": 0.00014198863636363636,
      "loss": 1.7818,
      "step": 2260
    },
    {
      "epoch": 2.89,
      "learning_rate": 0.00014113636363636363,
      "loss": 1.435,
      "step": 2270
    },
    {
      "epoch": 2.91,
      "learning_rate": 0.0001402840909090909,
      "loss": 2.5173,
      "step": 2280
    },
    {
      "epoch": 2.92,
      "learning_rate": 0.00013943181818181817,
      "loss": 2.2263,
      "step": 2290
    },
    {
      "epoch": 2.93,
      "learning_rate": 0.00013857954545454544,
      "loss": 2.0696,
      "step": 2300
    },
    {
      "epoch": 2.95,
      "learning_rate": 0.0001377272727272727,
      "loss": 1.7433,
      "step": 2310
    },
    {
      "epoch": 2.96,
      "learning_rate": 0.00013687499999999998,
      "loss": 1.6106,
      "step": 2320
    },
    {
      "epoch": 2.97,
      "learning_rate": 0.00013602272727272728,
      "loss": 2.5292,
      "step": 2330
    },
    {
      "epoch": 2.98,
      "learning_rate": 0.00013517045454545452,
      "loss": 2.0994,
      "step": 2340
    },
    {
      "epoch": 3.0,
      "learning_rate": 0.00013431818181818182,
      "loss": 1.6276,
      "step": 2350
    },
    {
      "epoch": 3.01,
      "learning_rate": 0.0001334659090909091,
      "loss": 2.3773,
      "step": 2360
    },
    {
      "epoch": 3.02,
      "learning_rate": 0.00013261363636363633,
      "loss": 2.1608,
      "step": 2370
    },
    {
      "epoch": 3.04,
      "learning_rate": 0.00013176136363636363,
      "loss": 1.8948,
      "step": 2380
    },
    {
      "epoch": 3.05,
      "learning_rate": 0.0001309090909090909,
      "loss": 1.642,
      "step": 2390
    },
    {
      "epoch": 3.06,
      "learning_rate": 0.00013005681818181817,
      "loss": 1.1205,
      "step": 2400
    },
    {
      "epoch": 3.06,
      "eval_loss": 2.7611794471740723,
      "eval_runtime": 1138.461,
      "eval_samples_per_second": 5.475,
      "eval_steps_per_second": 0.685,
      "eval_wer": 0.909806509172692,
      "step": 2400
    },
    {
      "epoch": 3.07,
      "learning_rate": 0.00012920454545454544,
      "loss": 2.3052,
      "step": 2410
    },
    {
      "epoch": 3.09,
      "learning_rate": 0.0001283522727272727,
      "loss": 2.1374,
      "step": 2420
    },
    {
      "epoch": 3.1,
      "learning_rate": 0.00012749999999999998,
      "loss": 1.8912,
      "step": 2430
    },
    {
      "epoch": 3.11,
      "learning_rate": 0.00012664772727272728,
      "loss": 1.5719,
      "step": 2440
    },
    {
      "epoch": 3.12,
      "learning_rate": 0.00012579545454545452,
      "loss": 1.1514,
      "step": 2450
    },
    {
      "epoch": 3.14,
      "learning_rate": 0.00012494318181818182,
      "loss": 2.3621,
      "step": 2460
    },
    {
      "epoch": 3.15,
      "learning_rate": 0.0001240909090909091,
      "loss": 2.1977,
      "step": 2470
    },
    {
      "epoch": 3.16,
      "learning_rate": 0.00012323863636363633,
      "loss": 1.8366,
      "step": 2480
    },
    {
      "epoch": 3.18,
      "learning_rate": 0.00012238636363636363,
      "loss": 1.5233,
      "step": 2490
    },
    {
      "epoch": 3.19,
      "learning_rate": 0.00012153409090909089,
      "loss": 1.0517,
      "step": 2500
    },
    {
      "epoch": 3.2,
      "learning_rate": 0.00012068181818181817,
      "loss": 2.3243,
      "step": 2510
    },
    {
      "epoch": 3.21,
      "learning_rate": 0.00011982954545454544,
      "loss": 2.0906,
      "step": 2520
    },
    {
      "epoch": 3.23,
      "learning_rate": 0.00011897727272727273,
      "loss": 1.8305,
      "step": 2530
    },
    {
      "epoch": 3.24,
      "learning_rate": 0.00011812499999999998,
      "loss": 1.4674,
      "step": 2540
    },
    {
      "epoch": 3.25,
      "learning_rate": 0.00011727272727272727,
      "loss": 1.0586,
      "step": 2550
    },
    {
      "epoch": 3.27,
      "learning_rate": 0.00011642045454545454,
      "loss": 2.2792,
      "step": 2560
    },
    {
      "epoch": 3.28,
      "learning_rate": 0.00011556818181818181,
      "loss": 2.0311,
      "step": 2570
    },
    {
      "epoch": 3.29,
      "learning_rate": 0.00011471590909090908,
      "loss": 1.7596,
      "step": 2580
    },
    {
      "epoch": 3.3,
      "learning_rate": 0.00011386363636363635,
      "loss": 1.4542,
      "step": 2590
    },
    {
      "epoch": 3.32,
      "learning_rate": 0.00011301136363636362,
      "loss": 1.0427,
      "step": 2600
    },
    {
      "epoch": 3.33,
      "learning_rate": 0.00011215909090909089,
      "loss": 2.2565,
      "step": 2610
    },
    {
      "epoch": 3.34,
      "learning_rate": 0.00011130681818181818,
      "loss": 1.9995,
      "step": 2620
    },
    {
      "epoch": 3.35,
      "learning_rate": 0.00011045454545454545,
      "loss": 1.696,
      "step": 2630
    },
    {
      "epoch": 3.37,
      "learning_rate": 0.00010960227272727272,
      "loss": 1.3768,
      "step": 2640
    },
    {
      "epoch": 3.38,
      "learning_rate": 0.00010874999999999999,
      "loss": 0.9125,
      "step": 2650
    },
    {
      "epoch": 3.39,
      "learning_rate": 0.00010789772727272727,
      "loss": 2.1988,
      "step": 2660
    },
    {
      "epoch": 3.41,
      "learning_rate": 0.00010704545454545453,
      "loss": 1.9406,
      "step": 2670
    },
    {
      "epoch": 3.42,
      "learning_rate": 0.00010619318181818181,
      "loss": 1.6374,
      "step": 2680
    },
    {
      "epoch": 3.43,
      "learning_rate": 0.00010534090909090908,
      "loss": 1.3302,
      "step": 2690
    },
    {
      "epoch": 3.44,
      "learning_rate": 0.00010448863636363635,
      "loss": 0.925,
      "step": 2700
    },
    {
      "epoch": 3.46,
      "learning_rate": 0.00010363636363636362,
      "loss": 2.1624,
      "step": 2710
    },
    {
      "epoch": 3.47,
      "learning_rate": 0.0001027840909090909,
      "loss": 1.8209,
      "step": 2720
    },
    {
      "epoch": 3.48,
      "learning_rate": 0.00010193181818181818,
      "loss": 1.4644,
      "step": 2730
    },
    {
      "epoch": 3.49,
      "learning_rate": 0.00010107954545454544,
      "loss": 1.2247,
      "step": 2740
    },
    {
      "epoch": 3.51,
      "learning_rate": 0.00010022727272727272,
      "loss": 0.8943,
      "step": 2750
    },
    {
      "epoch": 3.52,
      "learning_rate": 9.937499999999999e-05,
      "loss": 2.1089,
      "step": 2760
    },
    {
      "epoch": 3.53,
      "learning_rate": 9.852272727272727e-05,
      "loss": 1.7865,
      "step": 2770
    },
    {
      "epoch": 3.55,
      "learning_rate": 9.767045454545453e-05,
      "loss": 1.4291,
      "step": 2780
    },
    {
      "epoch": 3.56,
      "learning_rate": 9.681818181818181e-05,
      "loss": 1.1092,
      "step": 2790
    },
    {
      "epoch": 3.57,
      "learning_rate": 9.596590909090909e-05,
      "loss": 0.7821,
      "step": 2800
    },
    {
      "epoch": 3.57,
      "eval_loss": 2.290031671524048,
      "eval_runtime": 1222.3047,
      "eval_samples_per_second": 5.099,
      "eval_steps_per_second": 0.638,
      "eval_wer": 0.8477723462389175,
      "step": 2800
    },
    {
      "epoch": 3.58,
      "learning_rate": 9.511363636363634e-05,
      "loss": 2.0107,
      "step": 2810
    },
    {
      "epoch": 3.6,
      "learning_rate": 9.426136363636363e-05,
      "loss": 1.6357,
      "step": 2820
    },
    {
      "epoch": 3.61,
      "learning_rate": 9.34090909090909e-05,
      "loss": 1.278,
      "step": 2830
    },
    {
      "epoch": 3.62,
      "learning_rate": 9.255681818181818e-05,
      "loss": 1.0386,
      "step": 2840
    },
    {
      "epoch": 3.63,
      "learning_rate": 9.170454545454544e-05,
      "loss": 0.8257,
      "step": 2850
    },
    {
      "epoch": 3.65,
      "learning_rate": 9.085227272727272e-05,
      "loss": 1.9624,
      "step": 2860
    },
    {
      "epoch": 3.66,
      "learning_rate": 8.999999999999999e-05,
      "loss": 1.531,
      "step": 2870
    },
    {
      "epoch": 3.67,
      "learning_rate": 8.914772727272728e-05,
      "loss": 1.1868,
      "step": 2880
    },
    {
      "epoch": 3.69,
      "learning_rate": 8.829545454545453e-05,
      "loss": 0.9335,
      "step": 2890
    },
    {
      "epoch": 3.7,
      "learning_rate": 8.744318181818182e-05,
      "loss": 0.7307,
      "step": 2900
    },
    {
      "epoch": 3.71,
      "learning_rate": 8.659090909090909e-05,
      "loss": 1.8045,
      "step": 2910
    },
    {
      "epoch": 3.72,
      "learning_rate": 8.573863636363634e-05,
      "loss": 1.3896,
      "step": 2920
    },
    {
      "epoch": 3.74,
      "learning_rate": 8.488636363636363e-05,
      "loss": 1.1009,
      "step": 2930
    },
    {
      "epoch": 3.75,
      "learning_rate": 8.40340909090909e-05,
      "loss": 0.8753,
      "step": 2940
    },
    {
      "epoch": 3.76,
      "learning_rate": 8.318181818181818e-05,
      "loss": 0.7443,
      "step": 2950
    },
    {
      "epoch": 3.77,
      "learning_rate": 8.232954545454544e-05,
      "loss": 1.6411,
      "step": 2960
    },
    {
      "epoch": 3.79,
      "learning_rate": 8.147727272727272e-05,
      "loss": 1.2826,
      "step": 2970
    },
    {
      "epoch": 3.8,
      "learning_rate": 8.0625e-05,
      "loss": 0.9912,
      "step": 2980
    },
    {
      "epoch": 3.81,
      "learning_rate": 7.977272727272727e-05,
      "loss": 0.8302,
      "step": 2990
    },
    {
      "epoch": 3.83,
      "learning_rate": 7.892045454545454e-05,
      "loss": 0.732,
      "step": 3000
    },
    {
      "epoch": 3.84,
      "learning_rate": 7.806818181818182e-05,
      "loss": 1.4946,
      "step": 3010
    },
    {
      "epoch": 3.85,
      "learning_rate": 7.721590909090909e-05,
      "loss": 1.1615,
      "step": 3020
    },
    {
      "epoch": 3.86,
      "learning_rate": 7.636363636363635e-05,
      "loss": 0.935,
      "step": 3030
    },
    {
      "epoch": 3.88,
      "learning_rate": 7.551136363636363e-05,
      "loss": 0.7748,
      "step": 3040
    },
    {
      "epoch": 3.89,
      "learning_rate": 7.46590909090909e-05,
      "loss": 0.6236,
      "step": 3050
    },
    {
      "epoch": 3.9,
      "learning_rate": 7.380681818181817e-05,
      "loss": 1.4838,
      "step": 3060
    },
    {
      "epoch": 3.92,
      "learning_rate": 7.295454545454546e-05,
      "loss": 1.0713,
      "step": 3070
    },
    {
      "epoch": 3.93,
      "learning_rate": 7.210227272727271e-05,
      "loss": 0.8608,
      "step": 3080
    },
    {
      "epoch": 3.94,
      "learning_rate": 7.125e-05,
      "loss": 0.7445,
      "step": 3090
    },
    {
      "epoch": 3.95,
      "learning_rate": 7.039772727272727e-05,
      "loss": 0.6349,
      "step": 3100
    },
    {
      "epoch": 3.97,
      "learning_rate": 6.954545454545454e-05,
      "loss": 1.3051,
      "step": 3110
    },
    {
      "epoch": 3.98,
      "learning_rate": 6.869318181818181e-05,
      "loss": 0.9718,
      "step": 3120
    },
    {
      "epoch": 3.99,
      "learning_rate": 6.784090909090908e-05,
      "loss": 0.7319,
      "step": 3130
    },
    {
      "epoch": 4.01,
      "learning_rate": 6.698863636363636e-05,
      "loss": 1.0037,
      "step": 3140
    },
    {
      "epoch": 4.02,
      "learning_rate": 6.613636363636363e-05,
      "loss": 0.9463,
      "step": 3150
    },
    {
      "epoch": 4.03,
      "learning_rate": 6.52840909090909e-05,
      "loss": 0.7057,
      "step": 3160
    },
    {
      "epoch": 4.04,
      "learning_rate": 6.443181818181817e-05,
      "loss": 0.5883,
      "step": 3170
    },
    {
      "epoch": 4.06,
      "learning_rate": 6.357954545454546e-05,
      "loss": 0.5031,
      "step": 3180
    },
    {
      "epoch": 4.07,
      "learning_rate": 6.272727272727272e-05,
      "loss": 0.9323,
      "step": 3190
    },
    {
      "epoch": 4.08,
      "learning_rate": 6.187499999999999e-05,
      "loss": 1.0018,
      "step": 3200
    },
    {
      "epoch": 4.08,
      "eval_loss": 0.7969938516616821,
      "eval_runtime": 1370.1693,
      "eval_samples_per_second": 4.549,
      "eval_steps_per_second": 0.569,
      "eval_wer": 0.6274604561253508,
      "step": 3200
    },
    {
      "epoch": 4.09,
      "learning_rate": 6.102272727272726e-05,
      "loss": 0.7135,
      "step": 3210
    },
    {
      "epoch": 4.11,
      "learning_rate": 6.017045454545454e-05,
      "loss": 0.5946,
      "step": 3220
    },
    {
      "epoch": 4.12,
      "learning_rate": 5.931818181818181e-05,
      "loss": 0.4899,
      "step": 3230
    },
    {
      "epoch": 4.13,
      "learning_rate": 5.846590909090909e-05,
      "loss": 0.8917,
      "step": 3240
    },
    {
      "epoch": 4.15,
      "learning_rate": 5.761363636363636e-05,
      "loss": 0.9219,
      "step": 3250
    },
    {
      "epoch": 4.16,
      "learning_rate": 5.676136363636363e-05,
      "loss": 0.6885,
      "step": 3260
    },
    {
      "epoch": 4.17,
      "learning_rate": 5.590909090909091e-05,
      "loss": 0.5763,
      "step": 3270
    },
    {
      "epoch": 4.18,
      "learning_rate": 5.505681818181818e-05,
      "loss": 0.5006,
      "step": 3280
    },
    {
      "epoch": 4.2,
      "learning_rate": 5.4204545454545454e-05,
      "loss": 0.7933,
      "step": 3290
    },
    {
      "epoch": 4.21,
      "learning_rate": 5.335227272727272e-05,
      "loss": 0.8369,
      "step": 3300
    },
    {
      "epoch": 4.22,
      "learning_rate": 5.2499999999999995e-05,
      "loss": 0.6639,
      "step": 3310
    },
    {
      "epoch": 4.23,
      "learning_rate": 5.1647727272727266e-05,
      "loss": 0.5441,
      "step": 3320
    },
    {
      "epoch": 4.25,
      "learning_rate": 5.0795454545454536e-05,
      "loss": 0.4834,
      "step": 3330
    },
    {
      "epoch": 4.26,
      "learning_rate": 4.9943181818181814e-05,
      "loss": 0.7445,
      "step": 3340
    },
    {
      "epoch": 4.27,
      "learning_rate": 4.9090909090909084e-05,
      "loss": 0.784,
      "step": 3350
    },
    {
      "epoch": 4.29,
      "learning_rate": 4.823863636363636e-05,
      "loss": 0.6199,
      "step": 3360
    },
    {
      "epoch": 4.3,
      "learning_rate": 4.738636363636363e-05,
      "loss": 0.5361,
      "step": 3370
    },
    {
      "epoch": 4.31,
      "learning_rate": 4.653409090909091e-05,
      "loss": 0.46,
      "step": 3380
    },
    {
      "epoch": 4.32,
      "learning_rate": 4.568181818181818e-05,
      "loss": 0.7264,
      "step": 3390
    },
    {
      "epoch": 4.34,
      "learning_rate": 4.482954545454546e-05,
      "loss": 0.7638,
      "step": 3400
    },
    {
      "epoch": 4.35,
      "learning_rate": 4.397727272727272e-05,
      "loss": 0.601,
      "step": 3410
    },
    {
      "epoch": 4.36,
      "learning_rate": 4.312499999999999e-05,
      "loss": 0.5139,
      "step": 3420
    },
    {
      "epoch": 4.37,
      "learning_rate": 4.227272727272727e-05,
      "loss": 0.4643,
      "step": 3430
    },
    {
      "epoch": 4.39,
      "learning_rate": 4.142045454545454e-05,
      "loss": 0.6763,
      "step": 3440
    },
    {
      "epoch": 4.4,
      "learning_rate": 4.0568181818181816e-05,
      "loss": 0.7152,
      "step": 3450
    },
    {
      "epoch": 4.41,
      "learning_rate": 3.971590909090909e-05,
      "loss": 0.5839,
      "step": 3460
    },
    {
      "epoch": 4.43,
      "learning_rate": 3.8863636363636364e-05,
      "loss": 0.5277,
      "step": 3470
    },
    {
      "epoch": 4.44,
      "learning_rate": 3.8011363636363634e-05,
      "loss": 0.4187,
      "step": 3480
    },
    {
      "epoch": 4.45,
      "learning_rate": 3.7159090909090905e-05,
      "loss": 0.6736,
      "step": 3490
    },
    {
      "epoch": 4.46,
      "learning_rate": 3.6306818181818175e-05,
      "loss": 0.6799,
      "step": 3500
    },
    {
      "epoch": 4.48,
      "learning_rate": 3.545454545454545e-05,
      "loss": 0.5369,
      "step": 3510
    },
    {
      "epoch": 4.49,
      "learning_rate": 3.460227272727272e-05,
      "loss": 0.4819,
      "step": 3520
    },
    {
      "epoch": 4.5,
      "learning_rate": 3.375e-05,
      "loss": 0.4336,
      "step": 3530
    },
    {
      "epoch": 4.51,
      "learning_rate": 3.289772727272727e-05,
      "loss": 0.6435,
      "step": 3540
    },
    {
      "epoch": 4.53,
      "learning_rate": 3.204545454545454e-05,
      "loss": 0.6614,
      "step": 3550
    },
    {
      "epoch": 4.54,
      "learning_rate": 3.119318181818181e-05,
      "loss": 0.5409,
      "step": 3560
    },
    {
      "epoch": 4.55,
      "learning_rate": 3.034090909090909e-05,
      "loss": 0.4494,
      "step": 3570
    },
    {
      "epoch": 4.57,
      "learning_rate": 2.9488636363636363e-05,
      "loss": 0.428,
      "step": 3580
    },
    {
      "epoch": 4.58,
      "learning_rate": 2.8636363636363634e-05,
      "loss": 0.6517,
      "step": 3590
    },
    {
      "epoch": 4.59,
      "learning_rate": 2.7784090909090907e-05,
      "loss": 0.6312,
      "step": 3600
    },
    {
      "epoch": 4.59,
      "eval_loss": 0.5575253963470459,
      "eval_runtime": 1367.3048,
      "eval_samples_per_second": 4.559,
      "eval_steps_per_second": 0.57,
      "eval_wer": 0.5218769167859334,
      "step": 3600
    },
    {
      "epoch": 4.6,
      "learning_rate": 2.6931818181818178e-05,
      "loss": 0.5375,
      "step": 3610
    },
    {
      "epoch": 4.62,
      "learning_rate": 2.6079545454545452e-05,
      "loss": 0.4596,
      "step": 3620
    },
    {
      "epoch": 4.63,
      "learning_rate": 2.5227272727272726e-05,
      "loss": 0.3776,
      "step": 3630
    },
    {
      "epoch": 4.64,
      "learning_rate": 2.4375e-05,
      "loss": 0.611,
      "step": 3640
    },
    {
      "epoch": 4.66,
      "learning_rate": 2.3522727272727273e-05,
      "loss": 0.6419,
      "step": 3650
    },
    {
      "epoch": 4.67,
      "learning_rate": 2.267045454545454e-05,
      "loss": 0.5316,
      "step": 3660
    },
    {
      "epoch": 4.68,
      "learning_rate": 2.1818181818181814e-05,
      "loss": 0.4529,
      "step": 3670
    },
    {
      "epoch": 4.69,
      "learning_rate": 2.0965909090909088e-05,
      "loss": 0.4011,
      "step": 3680
    },
    {
      "epoch": 4.71,
      "learning_rate": 2.0113636363636362e-05,
      "loss": 0.5759,
      "step": 3690
    },
    {
      "epoch": 4.72,
      "learning_rate": 1.9261363636363636e-05,
      "loss": 0.5917,
      "step": 3700
    },
    {
      "epoch": 4.73,
      "learning_rate": 1.8409090909090907e-05,
      "loss": 0.5131,
      "step": 3710
    },
    {
      "epoch": 4.74,
      "learning_rate": 1.755681818181818e-05,
      "loss": 0.4384,
      "step": 3720
    },
    {
      "epoch": 4.76,
      "learning_rate": 1.670454545454545e-05,
      "loss": 0.386,
      "step": 3730
    },
    {
      "epoch": 4.77,
      "learning_rate": 1.5852272727272725e-05,
      "loss": 0.5591,
      "step": 3740
    },
    {
      "epoch": 4.78,
      "learning_rate": 1.4999999999999999e-05,
      "loss": 0.6046,
      "step": 3750
    },
    {
      "epoch": 4.8,
      "learning_rate": 1.4147727272727271e-05,
      "loss": 0.4815,
      "step": 3760
    },
    {
      "epoch": 4.81,
      "learning_rate": 1.3295454545454545e-05,
      "loss": 0.4244,
      "step": 3770
    },
    {
      "epoch": 4.82,
      "learning_rate": 1.2443181818181819e-05,
      "loss": 0.3968,
      "step": 3780
    },
    {
      "epoch": 4.83,
      "learning_rate": 1.1590909090909089e-05,
      "loss": 0.5641,
      "step": 3790
    },
    {
      "epoch": 4.85,
      "learning_rate": 1.0738636363636363e-05,
      "loss": 0.5865,
      "step": 3800
    },
    {
      "epoch": 4.86,
      "learning_rate": 9.886363636363635e-06,
      "loss": 0.4954,
      "step": 3810
    },
    {
      "epoch": 4.87,
      "learning_rate": 9.034090909090907e-06,
      "loss": 0.4475,
      "step": 3820
    },
    {
      "epoch": 4.88,
      "learning_rate": 8.181818181818181e-06,
      "loss": 0.3937,
      "step": 3830
    },
    {
      "epoch": 4.9,
      "learning_rate": 7.329545454545454e-06,
      "loss": 0.5145,
      "step": 3840
    },
    {
      "epoch": 4.91,
      "learning_rate": 6.4772727272727265e-06,
      "loss": 0.5883,
      "step": 3850
    },
    {
      "epoch": 4.92,
      "learning_rate": 5.6249999999999995e-06,
      "loss": 0.4964,
      "step": 3860
    },
    {
      "epoch": 4.94,
      "learning_rate": 4.772727272727272e-06,
      "loss": 0.4294,
      "step": 3870
    },
    {
      "epoch": 4.95,
      "learning_rate": 3.920454545454545e-06,
      "loss": 0.3601,
      "step": 3880
    },
    {
      "epoch": 4.96,
      "learning_rate": 3.0681818181818178e-06,
      "loss": 0.5155,
      "step": 3890
    },
    {
      "epoch": 4.97,
      "learning_rate": 2.215909090909091e-06,
      "loss": 0.5326,
      "step": 3900
    },
    {
      "epoch": 4.99,
      "learning_rate": 1.3636363636363634e-06,
      "loss": 0.4416,
      "step": 3910
    },
    {
      "epoch": 5.0,
      "learning_rate": 5.113636363636363e-07,
      "loss": 0.3953,
      "step": 3920
    },
    {
      "epoch": 5.0,
      "step": 3920,
      "total_flos": 0.0,
      "train_loss": 2.0088163537030317,
      "train_runtime": 35632.9758,
      "train_samples_per_second": 7.045,
      "train_steps_per_second": 0.11
    }
  ],
  "max_steps": 3920,
  "num_train_epochs": 5,
  "total_flos": 0.0,
  "trial_name": null,
  "trial_params": null
}