{
  "best_metric": 16.436102343697538,
  "best_model_checkpoint": "bengali_models/whisper-medium-bn_alldata_multigpu/checkpoint-3100",
  "epoch": 3.003875968992248,
  "global_step": 3100,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "learning_rate": 2.3372340764642327e-06,
      "loss": 2.1543,
      "step": 10
    },
    {
      "epoch": 0.02,
      "learning_rate": 5.456781345429857e-06,
      "loss": 1.4366,
      "step": 20
    },
    {
      "epoch": 0.03,
      "learning_rate": 6.67058251907103e-06,
      "loss": 1.1089,
      "step": 30
    },
    {
      "epoch": 0.04,
      "learning_rate": 7.438617522822122e-06,
      "loss": 0.8852,
      "step": 40
    },
    {
      "epoch": 0.05,
      "learning_rate": 8.001735616354375e-06,
      "loss": 0.7685,
      "step": 50
    },
    {
      "epoch": 0.06,
      "learning_rate": 8.446566317872072e-06,
      "loss": 0.6912,
      "step": 60
    },
    {
      "epoch": 0.07,
      "learning_rate": 8.814279401192807e-06,
      "loss": 0.6055,
      "step": 70
    },
    {
      "epoch": 0.08,
      "learning_rate": 9.127704207365829e-06,
      "loss": 0.5745,
      "step": 80
    },
    {
      "epoch": 0.09,
      "learning_rate": 9.40082771104536e-06,
      "loss": 0.549,
      "step": 90
    },
    {
      "epoch": 0.1,
      "learning_rate": 9.64284315734854e-06,
      "loss": 0.5302,
      "step": 100
    },
    {
      "epoch": 0.1,
      "eval_loss": 0.1824951171875,
      "eval_runtime": 426.8609,
      "eval_samples_per_second": 6.494,
      "eval_steps_per_second": 0.052,
      "eval_wer": 49.244510106775905,
      "step": 100
    },
    {
      "epoch": 0.11,
      "learning_rate": 9.860117748727801e-06,
      "loss": 0.4967,
      "step": 110
    },
    {
      "epoch": 0.12,
      "learning_rate": 9.998163452708908e-06,
      "loss": 0.5011,
      "step": 120
    },
    {
      "epoch": 0.13,
      "learning_rate": 9.988980716253444e-06,
      "loss": 0.501,
      "step": 130
    },
    {
      "epoch": 0.14,
      "learning_rate": 9.97979797979798e-06,
      "loss": 0.4828,
      "step": 140
    },
    {
      "epoch": 0.15,
      "learning_rate": 9.970615243342517e-06,
      "loss": 0.4781,
      "step": 150
    },
    {
      "epoch": 0.16,
      "learning_rate": 9.961432506887054e-06,
      "loss": 0.4595,
      "step": 160
    },
    {
      "epoch": 0.16,
      "learning_rate": 9.952249770431589e-06,
      "loss": 0.4594,
      "step": 170
    },
    {
      "epoch": 0.17,
      "learning_rate": 9.943067033976126e-06,
      "loss": 0.4444,
      "step": 180
    },
    {
      "epoch": 0.18,
      "learning_rate": 9.933884297520661e-06,
      "loss": 0.4481,
      "step": 190
    },
    {
      "epoch": 0.19,
      "learning_rate": 9.924701561065198e-06,
      "loss": 0.4292,
      "step": 200
    },
    {
      "epoch": 0.19,
      "eval_loss": 0.13525390625,
      "eval_runtime": 377.3366,
      "eval_samples_per_second": 7.346,
      "eval_steps_per_second": 0.058,
      "eval_wer": 36.814183063595465,
      "step": 200
    },
    {
      "epoch": 0.2,
      "learning_rate": 9.915518824609735e-06,
      "loss": 0.4188,
      "step": 210
    },
    {
      "epoch": 0.21,
      "learning_rate": 9.90633608815427e-06,
      "loss": 0.4122,
      "step": 220
    },
    {
      "epoch": 0.22,
      "learning_rate": 9.897153351698807e-06,
      "loss": 0.4094,
      "step": 230
    },
    {
      "epoch": 0.23,
      "learning_rate": 9.887970615243342e-06,
      "loss": 0.4012,
      "step": 240
    },
    {
      "epoch": 0.24,
      "learning_rate": 9.87878787878788e-06,
      "loss": 0.3944,
      "step": 250
    },
    {
      "epoch": 0.25,
      "learning_rate": 9.869605142332416e-06,
      "loss": 0.3903,
      "step": 260
    },
    {
      "epoch": 0.26,
      "learning_rate": 9.860422405876953e-06,
      "loss": 0.3864,
      "step": 270
    },
    {
      "epoch": 0.27,
      "learning_rate": 9.851239669421488e-06,
      "loss": 0.3989,
      "step": 280
    },
    {
      "epoch": 0.28,
      "learning_rate": 9.842056932966024e-06,
      "loss": 0.3816,
      "step": 290
    },
    {
      "epoch": 0.29,
      "learning_rate": 9.83287419651056e-06,
      "loss": 0.3769,
      "step": 300
    },
    {
      "epoch": 0.29,
      "eval_loss": 0.108642578125,
      "eval_runtime": 335.294,
      "eval_samples_per_second": 8.267,
      "eval_steps_per_second": 0.066,
      "eval_wer": 31.092606272244982,
      "step": 300
    },
    {
      "epoch": 0.3,
      "learning_rate": 9.823691460055097e-06,
      "loss": 0.3786,
      "step": 310
    },
    {
      "epoch": 0.31,
      "learning_rate": 9.814508723599634e-06,
      "loss": 0.3706,
      "step": 320
    },
    {
      "epoch": 0.32,
      "learning_rate": 9.80532598714417e-06,
      "loss": 0.3849,
      "step": 330
    },
    {
      "epoch": 0.33,
      "learning_rate": 9.796143250688705e-06,
      "loss": 0.3679,
      "step": 340
    },
    {
      "epoch": 0.34,
      "learning_rate": 9.786960514233242e-06,
      "loss": 0.3618,
      "step": 350
    },
    {
      "epoch": 0.35,
      "learning_rate": 9.777777777777779e-06,
      "loss": 0.3577,
      "step": 360
    },
    {
      "epoch": 0.36,
      "learning_rate": 9.768595041322316e-06,
      "loss": 0.3594,
      "step": 370
    },
    {
      "epoch": 0.37,
      "learning_rate": 9.75941230486685e-06,
      "loss": 0.3606,
      "step": 380
    },
    {
      "epoch": 0.38,
      "learning_rate": 9.750229568411388e-06,
      "loss": 0.3474,
      "step": 390
    },
    {
      "epoch": 0.39,
      "learning_rate": 9.741046831955923e-06,
      "loss": 0.3437,
      "step": 400
    },
    {
      "epoch": 0.39,
      "eval_loss": 0.09869384765625,
      "eval_runtime": 350.3333,
      "eval_samples_per_second": 7.912,
      "eval_steps_per_second": 0.063,
      "eval_wer": 27.875898193539722,
      "step": 400
    },
    {
      "epoch": 0.4,
      "learning_rate": 9.73186409550046e-06,
      "loss": 0.3454,
      "step": 410
    },
    {
      "epoch": 0.41,
      "learning_rate": 9.722681359044997e-06,
      "loss": 0.351,
      "step": 420
    },
    {
      "epoch": 0.42,
      "learning_rate": 9.713498622589532e-06,
      "loss": 0.3433,
      "step": 430
    },
    {
      "epoch": 0.43,
      "learning_rate": 9.704315886134069e-06,
      "loss": 0.3501,
      "step": 440
    },
    {
      "epoch": 0.44,
      "learning_rate": 9.695133149678604e-06,
      "loss": 0.3265,
      "step": 450
    },
    {
      "epoch": 0.45,
      "learning_rate": 9.685950413223141e-06,
      "loss": 0.3403,
      "step": 460
    },
    {
      "epoch": 0.46,
      "learning_rate": 9.676767676767678e-06,
      "loss": 0.3422,
      "step": 470
    },
    {
      "epoch": 0.47,
      "learning_rate": 9.667584940312213e-06,
      "loss": 0.3292,
      "step": 480
    },
    {
      "epoch": 0.47,
      "learning_rate": 9.65840220385675e-06,
      "loss": 0.3318,
      "step": 490
    },
    {
      "epoch": 0.48,
      "learning_rate": 9.649219467401286e-06,
      "loss": 0.3324,
      "step": 500
    },
    {
      "epoch": 0.48,
      "eval_loss": 0.087646484375,
      "eval_runtime": 356.163,
      "eval_samples_per_second": 7.783,
      "eval_steps_per_second": 0.062,
      "eval_wer": 26.03250285407293,
      "step": 500
    },
    {
      "epoch": 0.49,
      "learning_rate": 9.640036730945822e-06,
      "loss": 0.3377,
      "step": 510
    },
    {
      "epoch": 0.5,
      "learning_rate": 9.63085399449036e-06,
      "loss": 0.3278,
      "step": 520
    },
    {
      "epoch": 0.51,
      "learning_rate": 9.621671258034895e-06,
      "loss": 0.3348,
      "step": 530
    },
    {
      "epoch": 0.52,
      "learning_rate": 9.612488521579432e-06,
      "loss": 0.3259,
      "step": 540
    },
    {
      "epoch": 0.53,
      "learning_rate": 9.603305785123967e-06,
      "loss": 0.3267,
      "step": 550
    },
    {
      "epoch": 0.54,
      "learning_rate": 9.594123048668504e-06,
      "loss": 0.3174,
      "step": 560
    },
    {
      "epoch": 0.55,
      "learning_rate": 9.58494031221304e-06,
      "loss": 0.318,
      "step": 570
    },
    {
      "epoch": 0.56,
      "learning_rate": 9.575757575757576e-06,
      "loss": 0.31,
      "step": 580
    },
    {
      "epoch": 0.57,
      "learning_rate": 9.566574839302113e-06,
      "loss": 0.3045,
      "step": 590
    },
    {
      "epoch": 0.58,
      "learning_rate": 9.557392102846648e-06,
      "loss": 0.3111,
      "step": 600
    },
    {
      "epoch": 0.58,
      "eval_loss": 0.0823974609375,
      "eval_runtime": 387.8107,
      "eval_samples_per_second": 7.148,
      "eval_steps_per_second": 0.057,
      "eval_wer": 24.185749781747365,
      "step": 600
    },
    {
      "epoch": 0.59,
      "learning_rate": 9.548209366391185e-06,
      "loss": 0.3105,
      "step": 610
    },
    {
      "epoch": 0.6,
      "learning_rate": 9.539026629935722e-06,
      "loss": 0.3146,
      "step": 620
    },
    {
      "epoch": 0.61,
      "learning_rate": 9.529843893480259e-06,
      "loss": 0.3127,
      "step": 630
    },
    {
      "epoch": 0.62,
      "learning_rate": 9.520661157024794e-06,
      "loss": 0.3144,
      "step": 640
    },
    {
      "epoch": 0.63,
      "learning_rate": 9.51147842056933e-06,
      "loss": 0.3068,
      "step": 650
    },
    {
      "epoch": 0.64,
      "learning_rate": 9.502295684113866e-06,
      "loss": 0.3086,
      "step": 660
    },
    {
      "epoch": 0.65,
      "learning_rate": 9.493112947658403e-06,
      "loss": 0.3165,
      "step": 670
    },
    {
      "epoch": 0.66,
      "learning_rate": 9.48393021120294e-06,
      "loss": 0.3042,
      "step": 680
    },
    {
      "epoch": 0.67,
      "learning_rate": 9.474747474747475e-06,
      "loss": 0.3082,
      "step": 690
    },
    {
      "epoch": 0.68,
      "learning_rate": 9.46556473829201e-06,
      "loss": 0.3,
      "step": 700
    },
    {
      "epoch": 0.68,
      "eval_loss": 0.0762939453125,
      "eval_runtime": 391.6289,
      "eval_samples_per_second": 7.078,
      "eval_steps_per_second": 0.056,
      "eval_wer": 22.80236384393258,
      "step": 700
    },
    {
      "epoch": 0.69,
      "learning_rate": 9.456382001836548e-06,
      "loss": 0.3072,
      "step": 710
    },
    {
      "epoch": 0.7,
      "learning_rate": 9.447199265381084e-06,
      "loss": 0.3019,
      "step": 720
    },
    {
      "epoch": 0.71,
      "learning_rate": 9.438016528925621e-06,
      "loss": 0.3157,
      "step": 730
    },
    {
      "epoch": 0.72,
      "learning_rate": 9.428833792470157e-06,
      "loss": 0.3001,
      "step": 740
    },
    {
      "epoch": 0.73,
      "learning_rate": 9.419651056014694e-06,
      "loss": 0.3072,
      "step": 750
    },
    {
      "epoch": 0.74,
      "learning_rate": 9.410468319559229e-06,
      "loss": 0.3069,
      "step": 760
    },
    {
      "epoch": 0.75,
      "learning_rate": 9.401285583103766e-06,
      "loss": 0.3006,
      "step": 770
    },
    {
      "epoch": 0.76,
      "learning_rate": 9.392102846648303e-06,
      "loss": 0.2963,
      "step": 780
    },
    {
      "epoch": 0.77,
      "learning_rate": 9.382920110192838e-06,
      "loss": 0.3006,
      "step": 790
    },
    {
      "epoch": 0.78,
      "learning_rate": 9.373737373737375e-06,
      "loss": 0.2875,
      "step": 800
    },
    {
      "epoch": 0.78,
      "eval_loss": 0.07421875,
      "eval_runtime": 396.2127,
      "eval_samples_per_second": 6.996,
      "eval_steps_per_second": 0.056,
      "eval_wer": 22.5941844066886,
      "step": 800
    },
    {
      "epoch": 0.78,
      "learning_rate": 9.36455463728191e-06,
      "loss": 0.3015,
      "step": 810
    },
    {
      "epoch": 0.79,
      "learning_rate": 9.355371900826447e-06,
      "loss": 0.3002,
      "step": 820
    },
    {
      "epoch": 0.8,
      "learning_rate": 9.346189164370984e-06,
      "loss": 0.2929,
      "step": 830
    },
    {
      "epoch": 0.81,
      "learning_rate": 9.33700642791552e-06,
      "loss": 0.2926,
      "step": 840
    },
    {
      "epoch": 0.82,
      "learning_rate": 9.327823691460056e-06,
      "loss": 0.2885,
      "step": 850
    },
    {
      "epoch": 0.83,
      "learning_rate": 9.318640955004591e-06,
      "loss": 0.284,
      "step": 860
    },
    {
      "epoch": 0.84,
      "learning_rate": 9.309458218549128e-06,
      "loss": 0.2994,
      "step": 870
    },
    {
      "epoch": 0.85,
      "learning_rate": 9.300275482093665e-06,
      "loss": 0.2926,
      "step": 880
    },
    {
      "epoch": 0.86,
      "learning_rate": 9.2910927456382e-06,
      "loss": 0.286,
      "step": 890
    },
    {
      "epoch": 0.87,
      "learning_rate": 9.281910009182737e-06,
      "loss": 0.2901,
      "step": 900
    },
    {
      "epoch": 0.87,
      "eval_loss": 0.0714111328125,
      "eval_runtime": 396.8824,
      "eval_samples_per_second": 6.984,
      "eval_steps_per_second": 0.055,
      "eval_wer": 21.462628433281846,
      "step": 900
    },
    {
      "epoch": 0.88,
      "learning_rate": 9.272727272727273e-06,
      "loss": 0.2778,
      "step": 910
    },
    {
      "epoch": 0.89,
      "learning_rate": 9.26354453627181e-06,
      "loss": 0.2993,
      "step": 920
    },
    {
      "epoch": 0.9,
      "learning_rate": 9.254361799816347e-06,
      "loss": 0.2811,
      "step": 930
    },
    {
      "epoch": 0.91,
      "learning_rate": 9.245179063360882e-06,
      "loss": 0.2927,
      "step": 940
    },
    {
      "epoch": 0.92,
      "learning_rate": 9.235996326905419e-06,
      "loss": 0.2933,
      "step": 950
    },
    {
      "epoch": 0.93,
      "learning_rate": 9.226813590449954e-06,
      "loss": 0.2768,
      "step": 960
    },
    {
      "epoch": 0.94,
      "learning_rate": 9.217630853994491e-06,
      "loss": 0.2783,
      "step": 970
    },
    {
      "epoch": 0.95,
      "learning_rate": 9.208448117539028e-06,
      "loss": 0.2696,
      "step": 980
    },
    {
      "epoch": 0.96,
      "learning_rate": 9.199265381083565e-06,
      "loss": 0.2813,
      "step": 990
    },
    {
      "epoch": 0.97,
      "learning_rate": 9.1900826446281e-06,
      "loss": 0.2792,
      "step": 1000
    },
    {
      "epoch": 0.97,
      "eval_loss": 0.06939697265625,
      "eval_runtime": 446.7973,
      "eval_samples_per_second": 6.204,
      "eval_steps_per_second": 0.049,
      "eval_wer": 20.9186757101605,
      "step": 1000
    },
    {
      "epoch": 0.98,
      "learning_rate": 9.180899908172635e-06,
      "loss": 0.2774,
      "step": 1010
    },
    {
      "epoch": 0.99,
      "learning_rate": 9.171717171717172e-06,
      "loss": 0.2777,
      "step": 1020
    },
    {
      "epoch": 1.0,
      "learning_rate": 9.162534435261709e-06,
      "loss": 0.274,
      "step": 1030
    },
    {
      "epoch": 1.01,
      "learning_rate": 9.153351698806246e-06,
      "loss": 0.2612,
      "step": 1040
    },
    {
      "epoch": 1.02,
      "learning_rate": 9.144168962350781e-06,
      "loss": 0.2672,
      "step": 1050
    },
    {
      "epoch": 1.03,
      "learning_rate": 9.134986225895316e-06,
      "loss": 0.2675,
      "step": 1060
    },
    {
      "epoch": 1.04,
      "learning_rate": 9.125803489439853e-06,
      "loss": 0.2559,
      "step": 1070
    },
    {
      "epoch": 1.05,
      "learning_rate": 9.11662075298439e-06,
      "loss": 0.259,
      "step": 1080
    },
    {
      "epoch": 1.06,
      "learning_rate": 9.107438016528927e-06,
      "loss": 0.266,
      "step": 1090
    },
    {
      "epoch": 1.07,
      "learning_rate": 9.098255280073463e-06,
      "loss": 0.2673,
      "step": 1100
    },
    {
      "epoch": 1.07,
      "eval_loss": 0.06787109375,
      "eval_runtime": 468.4797,
      "eval_samples_per_second": 5.917,
      "eval_steps_per_second": 0.047,
      "eval_wer": 20.05909609831442,
      "step": 1100
    },
    {
      "epoch": 1.08,
      "learning_rate": 9.089072543618e-06,
      "loss": 0.2654,
      "step": 1110
    },
    {
      "epoch": 1.09,
      "learning_rate": 9.079889807162535e-06,
      "loss": 0.2588,
      "step": 1120
    },
    {
      "epoch": 1.09,
      "learning_rate": 9.070707070707072e-06,
      "loss": 0.2713,
      "step": 1130
    },
    {
      "epoch": 1.1,
      "learning_rate": 9.061524334251609e-06,
      "loss": 0.2658,
      "step": 1140
    },
    {
      "epoch": 1.11,
      "learning_rate": 9.052341597796144e-06,
      "loss": 0.26,
      "step": 1150
    },
    {
      "epoch": 1.12,
      "learning_rate": 9.04315886134068e-06,
      "loss": 0.2759,
      "step": 1160
    },
    {
      "epoch": 1.13,
      "learning_rate": 9.033976124885216e-06,
      "loss": 0.2691,
      "step": 1170
    },
    {
      "epoch": 1.14,
      "learning_rate": 9.024793388429753e-06,
      "loss": 0.2673,
      "step": 1180
    },
    {
      "epoch": 1.15,
      "learning_rate": 9.01561065197429e-06,
      "loss": 0.2581,
      "step": 1190
    },
    {
      "epoch": 1.16,
      "learning_rate": 9.006427915518825e-06,
      "loss": 0.253,
      "step": 1200
    },
    {
      "epoch": 1.16,
      "eval_loss": 0.0672607421875,
      "eval_runtime": 450.6375,
      "eval_samples_per_second": 6.151,
      "eval_steps_per_second": 0.049,
      "eval_wer": 19.80390840104761,
      "step": 1200
    },
    {
      "epoch": 1.17,
      "learning_rate": 8.997245179063362e-06,
      "loss": 0.2554,
      "step": 1210
    },
    {
      "epoch": 1.18,
      "learning_rate": 8.988062442607897e-06,
      "loss": 0.2527,
      "step": 1220
    },
    {
      "epoch": 1.19,
      "learning_rate": 8.978879706152434e-06,
      "loss": 0.2607,
      "step": 1230
    },
    {
      "epoch": 1.2,
      "learning_rate": 8.969696969696971e-06,
      "loss": 0.2537,
      "step": 1240
    },
    {
      "epoch": 1.21,
      "learning_rate": 8.960514233241506e-06,
      "loss": 0.2479,
      "step": 1250
    },
    {
      "epoch": 1.22,
      "learning_rate": 8.951331496786043e-06,
      "loss": 0.26,
      "step": 1260
    },
    {
      "epoch": 1.23,
      "learning_rate": 8.942148760330578e-06,
      "loss": 0.2483,
      "step": 1270
    },
    {
      "epoch": 1.24,
      "learning_rate": 8.932966023875115e-06,
      "loss": 0.2475,
      "step": 1280
    },
    {
      "epoch": 1.25,
      "learning_rate": 8.923783287419652e-06,
      "loss": 0.259,
      "step": 1290
    },
    {
      "epoch": 1.26,
      "learning_rate": 8.914600550964188e-06,
      "loss": 0.254,
      "step": 1300
    },
    {
      "epoch": 1.26,
      "eval_loss": 0.0650634765625,
      "eval_runtime": 457.7466,
      "eval_samples_per_second": 6.056,
      "eval_steps_per_second": 0.048,
      "eval_wer": 19.14915049358673,
      "step": 1300
    },
    {
      "epoch": 1.27,
      "learning_rate": 8.905417814508725e-06,
      "loss": 0.2614,
      "step": 1310
    },
    {
      "epoch": 1.28,
      "learning_rate": 8.89623507805326e-06,
      "loss": 0.2659,
      "step": 1320
    },
    {
      "epoch": 1.29,
      "learning_rate": 8.887052341597797e-06,
      "loss": 0.25,
      "step": 1330
    },
    {
      "epoch": 1.3,
      "learning_rate": 8.877869605142334e-06,
      "loss": 0.2725,
      "step": 1340
    },
    {
      "epoch": 1.31,
      "learning_rate": 8.86868686868687e-06,
      "loss": 0.2634,
      "step": 1350
    },
    {
      "epoch": 1.32,
      "learning_rate": 8.859504132231406e-06,
      "loss": 0.2599,
      "step": 1360
    },
    {
      "epoch": 1.33,
      "learning_rate": 8.850321395775941e-06,
      "loss": 0.2406,
      "step": 1370
    },
    {
      "epoch": 1.34,
      "learning_rate": 8.841138659320478e-06,
      "loss": 0.2475,
      "step": 1380
    },
    {
      "epoch": 1.35,
      "learning_rate": 8.831955922865015e-06,
      "loss": 0.2605,
      "step": 1390
    },
    {
      "epoch": 1.36,
      "learning_rate": 8.822773186409552e-06,
      "loss": 0.2608,
      "step": 1400
    },
    {
      "epoch": 1.36,
      "eval_loss": 0.061737060546875,
      "eval_runtime": 363.3208,
      "eval_samples_per_second": 7.63,
      "eval_steps_per_second": 0.061,
      "eval_wer": 18.877174132026056,
      "step": 1400
    },
    {
      "epoch": 1.37,
      "learning_rate": 8.813590449954087e-06,
      "loss": 0.2526,
      "step": 1410
    },
    {
      "epoch": 1.38,
      "learning_rate": 8.804407713498622e-06,
      "loss": 0.259,
      "step": 1420
    },
    {
      "epoch": 1.39,
      "learning_rate": 8.79522497704316e-06,
      "loss": 0.2524,
      "step": 1430
    },
    {
      "epoch": 1.4,
      "learning_rate": 8.786042240587696e-06,
      "loss": 0.2493,
      "step": 1440
    },
    {
      "epoch": 1.41,
      "learning_rate": 8.776859504132233e-06,
      "loss": 0.2534,
      "step": 1450
    },
    {
      "epoch": 1.41,
      "learning_rate": 8.767676767676768e-06,
      "loss": 0.2482,
      "step": 1460
    },
    {
      "epoch": 1.42,
      "learning_rate": 8.758494031221305e-06,
      "loss": 0.2538,
      "step": 1470
    },
    {
      "epoch": 1.43,
      "learning_rate": 8.74931129476584e-06,
      "loss": 0.2491,
      "step": 1480
    },
    {
      "epoch": 1.44,
      "learning_rate": 8.740128558310377e-06,
      "loss": 0.2577,
      "step": 1490
    },
    {
      "epoch": 1.45,
      "learning_rate": 8.730945821854914e-06,
      "loss": 0.2478,
      "step": 1500
    },
    {
      "epoch": 1.45,
      "eval_loss": 0.0625,
      "eval_runtime": 306.0285,
      "eval_samples_per_second": 9.058,
      "eval_steps_per_second": 0.072,
      "eval_wer": 18.98126385064804,
      "step": 1500
    },
    {
      "epoch": 1.46,
      "learning_rate": 8.72176308539945e-06,
      "loss": 0.2534,
      "step": 1510
    },
    {
      "epoch": 1.47,
      "learning_rate": 8.712580348943987e-06,
      "loss": 0.2453,
      "step": 1520
    },
    {
      "epoch": 1.48,
      "learning_rate": 8.703397612488522e-06,
      "loss": 0.252,
      "step": 1530
    },
    {
      "epoch": 1.49,
      "learning_rate": 8.694214876033059e-06,
      "loss": 0.2597,
      "step": 1540
    },
    {
      "epoch": 1.5,
      "learning_rate": 8.685032139577596e-06,
      "loss": 0.2481,
      "step": 1550
    },
    {
      "epoch": 1.51,
      "learning_rate": 8.675849403122131e-06,
      "loss": 0.2505,
      "step": 1560
    },
    {
      "epoch": 1.52,
      "learning_rate": 8.666666666666668e-06,
      "loss": 0.247,
      "step": 1570
    },
    {
      "epoch": 1.53,
      "learning_rate": 8.657483930211203e-06,
      "loss": 0.2558,
      "step": 1580
    },
    {
      "epoch": 1.54,
      "learning_rate": 8.64830119375574e-06,
      "loss": 0.2539,
      "step": 1590
    },
    {
      "epoch": 1.55,
      "learning_rate": 8.639118457300277e-06,
      "loss": 0.254,
      "step": 1600
    },
    {
      "epoch": 1.55,
      "eval_loss": 0.059967041015625,
      "eval_runtime": 884.4448,
      "eval_samples_per_second": 3.134,
      "eval_steps_per_second": 0.025,
      "eval_wer": 18.668994694782086,
      "step": 1600
    },
    {
      "epoch": 1.56,
      "learning_rate": 8.629935720844812e-06,
      "loss": 0.246,
      "step": 1610
    },
    {
      "epoch": 1.57,
      "learning_rate": 8.620752984389349e-06,
      "loss": 0.2391,
      "step": 1620
    },
    {
      "epoch": 1.58,
      "learning_rate": 8.611570247933884e-06,
      "loss": 0.2408,
      "step": 1630
    },
    {
      "epoch": 1.59,
      "learning_rate": 8.602387511478421e-06,
      "loss": 0.2423,
      "step": 1640
    },
    {
      "epoch": 1.6,
      "learning_rate": 8.593204775022958e-06,
      "loss": 0.2394,
      "step": 1650
    },
    {
      "epoch": 1.61,
      "learning_rate": 8.584022038567493e-06,
      "loss": 0.2499,
      "step": 1660
    },
    {
      "epoch": 1.62,
      "learning_rate": 8.57483930211203e-06,
      "loss": 0.2477,
      "step": 1670
    },
    {
      "epoch": 1.63,
      "learning_rate": 8.565656565656566e-06,
      "loss": 0.2564,
      "step": 1680
    },
    {
      "epoch": 1.64,
      "learning_rate": 8.556473829201103e-06,
      "loss": 0.2484,
      "step": 1690
    },
    {
      "epoch": 1.65,
      "learning_rate": 8.54729109274564e-06,
      "loss": 0.2454,
      "step": 1700
    },
    {
      "epoch": 1.65,
      "eval_loss": 0.060546875,
      "eval_runtime": 344.7465,
      "eval_samples_per_second": 8.041,
      "eval_steps_per_second": 0.064,
      "eval_wer": 18.1552615673897,
      "step": 1700
    },
    {
      "epoch": 1.66,
      "learning_rate": 8.538108356290176e-06,
      "loss": 0.2397,
      "step": 1710
    },
    {
      "epoch": 1.67,
      "learning_rate": 8.528925619834712e-06,
      "loss": 0.2384,
      "step": 1720
    },
    {
      "epoch": 1.68,
      "learning_rate": 8.519742883379247e-06,
      "loss": 0.2484,
      "step": 1730
    },
    {
      "epoch": 1.69,
      "learning_rate": 8.510560146923784e-06,
      "loss": 0.244,
      "step": 1740
    },
    {
      "epoch": 1.7,
      "learning_rate": 8.50137741046832e-06,
      "loss": 0.2371,
      "step": 1750
    },
    {
      "epoch": 1.71,
      "learning_rate": 8.492194674012858e-06,
      "loss": 0.2422,
      "step": 1760
    },
    {
      "epoch": 1.72,
      "learning_rate": 8.483011937557393e-06,
      "loss": 0.2362,
      "step": 1770
    },
    {
      "epoch": 1.72,
      "learning_rate": 8.473829201101928e-06,
      "loss": 0.2386,
      "step": 1780
    },
    {
      "epoch": 1.73,
      "learning_rate": 8.464646464646465e-06,
      "loss": 0.236,
      "step": 1790
    },
    {
      "epoch": 1.74,
      "learning_rate": 8.455463728191002e-06,
      "loss": 0.2471,
      "step": 1800
    },
    {
      "epoch": 1.74,
      "eval_loss": 0.060272216796875,
      "eval_runtime": 521.4543,
      "eval_samples_per_second": 5.316,
      "eval_steps_per_second": 0.042,
      "eval_wer": 17.960513061580823,
      "step": 1800
    },
    {
      "epoch": 1.75,
      "learning_rate": 8.446280991735539e-06,
      "loss": 0.2406,
      "step": 1810
    },
    {
      "epoch": 1.76,
      "learning_rate": 8.437098255280074e-06,
      "loss": 0.2476,
      "step": 1820
    },
    {
      "epoch": 1.77,
      "learning_rate": 8.427915518824611e-06,
      "loss": 0.2411,
      "step": 1830
    },
    {
      "epoch": 1.78,
      "learning_rate": 8.418732782369146e-06,
      "loss": 0.243,
      "step": 1840
    },
    {
      "epoch": 1.79,
      "learning_rate": 8.409550045913683e-06,
      "loss": 0.2425,
      "step": 1850
    },
    {
      "epoch": 1.8,
      "learning_rate": 8.40036730945822e-06,
      "loss": 0.2267,
      "step": 1860
    },
    {
      "epoch": 1.81,
      "learning_rate": 8.391184573002755e-06,
      "loss": 0.2492,
      "step": 1870
    },
    {
      "epoch": 1.82,
      "learning_rate": 8.382001836547292e-06,
      "loss": 0.2305,
      "step": 1880
    },
    {
      "epoch": 1.83,
      "learning_rate": 8.372819100091828e-06,
      "loss": 0.2315,
      "step": 1890
    },
    {
      "epoch": 1.84,
      "learning_rate": 8.363636363636365e-06,
      "loss": 0.2438,
      "step": 1900
    },
    {
      "epoch": 1.84,
      "eval_loss": 0.05859375,
      "eval_runtime": 508.8796,
      "eval_samples_per_second": 5.447,
      "eval_steps_per_second": 0.043,
      "eval_wer": 18.249278087435364,
      "step": 1900
    },
    {
      "epoch": 1.85,
      "learning_rate": 8.354453627180901e-06,
      "loss": 0.2381,
      "step": 1910
    },
    {
      "epoch": 1.86,
      "learning_rate": 8.345270890725437e-06,
      "loss": 0.2429,
      "step": 1920
    },
    {
      "epoch": 1.87,
      "learning_rate": 8.336088154269974e-06,
      "loss": 0.233,
      "step": 1930
    },
    {
      "epoch": 1.88,
      "learning_rate": 8.326905417814509e-06,
      "loss": 0.2411,
      "step": 1940
    },
    {
      "epoch": 1.89,
      "learning_rate": 8.317722681359046e-06,
      "loss": 0.2387,
      "step": 1950
    },
    {
      "epoch": 1.9,
      "learning_rate": 8.308539944903583e-06,
      "loss": 0.2402,
      "step": 1960
    },
    {
      "epoch": 1.91,
      "learning_rate": 8.299357208448118e-06,
      "loss": 0.2327,
      "step": 1970
    },
    {
      "epoch": 1.92,
      "learning_rate": 8.290174471992655e-06,
      "loss": 0.2428,
      "step": 1980
    },
    {
      "epoch": 1.93,
      "learning_rate": 8.28099173553719e-06,
      "loss": 0.241,
      "step": 1990
    },
    {
      "epoch": 1.94,
      "learning_rate": 8.271808999081727e-06,
      "loss": 0.2363,
      "step": 2000
    },
    {
      "epoch": 1.94,
      "eval_loss": 0.058197021484375,
      "eval_runtime": 501.722,
      "eval_samples_per_second": 5.525,
      "eval_steps_per_second": 0.044,
      "eval_wer": 17.829561480088643,
      "step": 2000
    },
    {
      "epoch": 1.95,
      "learning_rate": 8.262626262626264e-06,
      "loss": 0.2357,
      "step": 2010
    },
    {
      "epoch": 1.96,
      "learning_rate": 8.2534435261708e-06,
      "loss": 0.2303,
      "step": 2020
    },
    {
      "epoch": 1.97,
      "learning_rate": 8.244260789715336e-06,
      "loss": 0.2396,
      "step": 2030
    },
    {
      "epoch": 1.98,
      "learning_rate": 8.235078053259871e-06,
      "loss": 0.2369,
      "step": 2040
    },
    {
      "epoch": 1.99,
      "learning_rate": 8.225895316804408e-06,
      "loss": 0.2299,
      "step": 2050
    },
    {
      "epoch": 2.0,
      "learning_rate": 8.216712580348945e-06,
      "loss": 0.2387,
      "step": 2060
    },
    {
      "epoch": 2.01,
      "learning_rate": 8.20752984389348e-06,
      "loss": 0.2192,
      "step": 2070
    },
    {
      "epoch": 2.02,
      "learning_rate": 8.198347107438017e-06,
      "loss": 0.226,
      "step": 2080
    },
    {
      "epoch": 2.03,
      "learning_rate": 8.189164370982553e-06,
      "loss": 0.2267,
      "step": 2090
    },
    {
      "epoch": 2.03,
      "learning_rate": 8.17998163452709e-06,
      "loss": 0.2254,
      "step": 2100
    },
    {
      "epoch": 2.03,
      "eval_loss": 0.057861328125,
      "eval_runtime": 387.4399,
      "eval_samples_per_second": 7.155,
      "eval_steps_per_second": 0.057,
      "eval_wer": 17.60123564569203,
      "step": 2100
    },
    {
      "epoch": 2.04,
      "learning_rate": 8.170798898071625e-06,
      "loss": 0.2199,
      "step": 2110
    },
    {
      "epoch": 2.05,
      "learning_rate": 8.161616161616162e-06,
      "loss": 0.2298,
      "step": 2120
    },
    {
      "epoch": 2.06,
      "learning_rate": 8.152433425160699e-06,
      "loss": 0.2277,
      "step": 2130
    },
    {
      "epoch": 2.07,
      "learning_rate": 8.143250688705234e-06,
      "loss": 0.2202,
      "step": 2140
    },
    {
      "epoch": 2.08,
      "learning_rate": 8.134067952249771e-06,
      "loss": 0.2209,
      "step": 2150
    },
    {
      "epoch": 2.09,
      "learning_rate": 8.124885215794306e-06,
      "loss": 0.2266,
      "step": 2160
    },
    {
      "epoch": 2.1,
      "learning_rate": 8.115702479338843e-06,
      "loss": 0.2216,
      "step": 2170
    },
    {
      "epoch": 2.11,
      "learning_rate": 8.10651974288338e-06,
      "loss": 0.2199,
      "step": 2180
    },
    {
      "epoch": 2.12,
      "learning_rate": 8.097337006427917e-06,
      "loss": 0.2153,
      "step": 2190
    },
    {
      "epoch": 2.13,
      "learning_rate": 8.088154269972452e-06,
      "loss": 0.2229,
      "step": 2200
    },
    {
      "epoch": 2.13,
      "eval_loss": 0.05926513671875,
      "eval_runtime": 369.2175,
      "eval_samples_per_second": 7.508,
      "eval_steps_per_second": 0.06,
      "eval_wer": 17.312470619837487,
      "step": 2200
    },
    {
      "epoch": 2.14,
      "learning_rate": 8.078971533516987e-06,
      "loss": 0.2227,
      "step": 2210
    },
    {
      "epoch": 2.15,
      "learning_rate": 8.069788797061524e-06,
      "loss": 0.2286,
      "step": 2220
    },
    {
      "epoch": 2.16,
      "learning_rate": 8.060606060606061e-06,
      "loss": 0.2158,
      "step": 2230
    },
    {
      "epoch": 2.17,
      "learning_rate": 8.051423324150598e-06,
      "loss": 0.226,
      "step": 2240
    },
    {
      "epoch": 2.18,
      "learning_rate": 8.042240587695133e-06,
      "loss": 0.2256,
      "step": 2250
    },
    {
      "epoch": 2.19,
      "learning_rate": 8.033057851239669e-06,
      "loss": 0.2292,
      "step": 2260
    },
    {
      "epoch": 2.2,
      "learning_rate": 8.023875114784206e-06,
      "loss": 0.2233,
      "step": 2270
    },
    {
      "epoch": 2.21,
      "learning_rate": 8.014692378328743e-06,
      "loss": 0.2212,
      "step": 2280
    },
    {
      "epoch": 2.22,
      "learning_rate": 8.00550964187328e-06,
      "loss": 0.2185,
      "step": 2290
    },
    {
      "epoch": 2.23,
      "learning_rate": 7.996326905417815e-06,
      "loss": 0.2201,
      "step": 2300
    },
    {
      "epoch": 2.23,
      "eval_loss": 0.057952880859375,
      "eval_runtime": 368.8171,
      "eval_samples_per_second": 7.516,
      "eval_steps_per_second": 0.06,
      "eval_wer": 17.14794170975757,
      "step": 2300
    },
    {
      "epoch": 2.24,
      "learning_rate": 7.987144168962352e-06,
      "loss": 0.2234,
      "step": 2310
    },
    {
      "epoch": 2.25,
      "learning_rate": 7.977961432506887e-06,
      "loss": 0.2161,
      "step": 2320
    },
    {
      "epoch": 2.26,
      "learning_rate": 7.968778696051424e-06,
      "loss": 0.2259,
      "step": 2330
    },
    {
      "epoch": 2.27,
      "learning_rate": 7.95959595959596e-06,
      "loss": 0.2242,
      "step": 2340
    },
    {
      "epoch": 2.28,
      "learning_rate": 7.950413223140496e-06,
      "loss": 0.2222,
      "step": 2350
    },
    {
      "epoch": 2.29,
      "learning_rate": 7.941230486685033e-06,
      "loss": 0.2214,
      "step": 2360
    },
    {
      "epoch": 2.3,
      "learning_rate": 7.932047750229568e-06,
      "loss": 0.2287,
      "step": 2370
    },
    {
      "epoch": 2.31,
      "learning_rate": 7.922865013774105e-06,
      "loss": 0.2285,
      "step": 2380
    },
    {
      "epoch": 2.32,
      "learning_rate": 7.913682277318642e-06,
      "loss": 0.2303,
      "step": 2390
    },
    {
      "epoch": 2.33,
      "learning_rate": 7.904499540863177e-06,
      "loss": 0.2079,
      "step": 2400
    },
    {
      "epoch": 2.33,
      "eval_loss": 0.058868408203125,
      "eval_runtime": 390.2943,
      "eval_samples_per_second": 7.102,
      "eval_steps_per_second": 0.056,
      "eval_wer": 17.040494258276812,
      "step": 2400
    },
    {
      "epoch": 2.34,
      "learning_rate": 7.895316804407714e-06,
      "loss": 0.2282,
      "step": 2410
    },
    {
      "epoch": 2.34,
      "learning_rate": 7.88613406795225e-06,
      "loss": 0.2182,
      "step": 2420
    },
    {
      "epoch": 2.35,
      "learning_rate": 7.876951331496786e-06,
      "loss": 0.2231,
      "step": 2430
    },
    {
      "epoch": 2.36,
      "learning_rate": 7.867768595041323e-06,
      "loss": 0.2213,
      "step": 2440
    },
    {
      "epoch": 2.37,
      "learning_rate": 7.858585858585859e-06,
      "loss": 0.2195,
      "step": 2450
    },
    {
      "epoch": 2.38,
      "learning_rate": 7.849403122130396e-06,
      "loss": 0.2156,
      "step": 2460
    },
    {
      "epoch": 2.39,
      "learning_rate": 7.84022038567493e-06,
      "loss": 0.2147,
      "step": 2470
    },
    {
      "epoch": 2.4,
      "learning_rate": 7.831037649219468e-06,
      "loss": 0.2204,
      "step": 2480
    },
    {
      "epoch": 2.41,
      "learning_rate": 7.821854912764005e-06,
      "loss": 0.2186,
      "step": 2490
    },
    {
      "epoch": 2.42,
      "learning_rate": 7.812672176308542e-06,
      "loss": 0.2174,
      "step": 2500
    },
    {
      "epoch": 2.42,
      "eval_loss": 0.058868408203125,
      "eval_runtime": 372.1906,
      "eval_samples_per_second": 7.448,
      "eval_steps_per_second": 0.059,
      "eval_wer": 17.329259284131354,
      "step": 2500
    },
    {
      "epoch": 2.43,
      "learning_rate": 7.803489439853077e-06,
      "loss": 0.2208,
      "step": 2510
    },
    {
      "epoch": 2.44,
      "learning_rate": 7.794306703397612e-06,
      "loss": 0.2249,
      "step": 2520
    },
    {
      "epoch": 2.45,
      "learning_rate": 7.785123966942149e-06,
      "loss": 0.2215,
      "step": 2530
    },
    {
      "epoch": 2.46,
      "learning_rate": 7.775941230486686e-06,
      "loss": 0.217,
      "step": 2540
    },
    {
      "epoch": 2.47,
      "learning_rate": 7.766758494031223e-06,
      "loss": 0.2165,
      "step": 2550
    },
    {
      "epoch": 2.48,
      "learning_rate": 7.757575757575758e-06,
      "loss": 0.2131,
      "step": 2560
    },
    {
      "epoch": 2.49,
      "learning_rate": 7.748393021120293e-06,
      "loss": 0.2194,
      "step": 2570
    },
    {
      "epoch": 2.5,
      "learning_rate": 7.73921028466483e-06,
      "loss": 0.215,
      "step": 2580
    },
    {
      "epoch": 2.51,
      "learning_rate": 7.730027548209367e-06,
      "loss": 0.2189,
      "step": 2590
    },
    {
      "epoch": 2.52,
      "learning_rate": 7.720844811753904e-06,
      "loss": 0.2197,
      "step": 2600
    },
    {
      "epoch": 2.52,
      "eval_loss": 0.058807373046875,
      "eval_runtime": 388.309,
      "eval_samples_per_second": 7.139,
      "eval_steps_per_second": 0.057,
      "eval_wer": 16.886038546773218,
      "step": 2600
    },
    {
      "epoch": 2.53,
      "learning_rate": 7.71166207529844e-06,
      "loss": 0.2209,
      "step": 2610
    },
    {
      "epoch": 2.54,
      "learning_rate": 7.702479338842976e-06,
      "loss": 0.2251,
      "step": 2620
    },
    {
      "epoch": 2.55,
      "learning_rate": 7.693296602387511e-06,
      "loss": 0.2262,
      "step": 2630
    },
    {
      "epoch": 2.56,
      "learning_rate": 7.684113865932048e-06,
      "loss": 0.221,
      "step": 2640
    },
    {
      "epoch": 2.57,
      "learning_rate": 7.674931129476585e-06,
      "loss": 0.2101,
      "step": 2650
    },
    {
      "epoch": 2.58,
      "learning_rate": 7.66574839302112e-06,
      "loss": 0.2137,
      "step": 2660
    },
    {
      "epoch": 2.59,
      "learning_rate": 7.656565656565658e-06,
      "loss": 0.211,
      "step": 2670
    },
    {
      "epoch": 2.6,
      "learning_rate": 7.647382920110193e-06,
      "loss": 0.2121,
      "step": 2680
    },
    {
      "epoch": 2.61,
      "learning_rate": 7.63820018365473e-06,
      "loss": 0.2198,
      "step": 2690
    },
    {
      "epoch": 2.62,
      "learning_rate": 7.629017447199266e-06,
      "loss": 0.2227,
      "step": 2700
    },
    {
      "epoch": 2.62,
      "eval_loss": 0.058868408203125,
      "eval_runtime": 386.9363,
      "eval_samples_per_second": 7.164,
      "eval_steps_per_second": 0.057,
      "eval_wer": 16.892754012490766,
      "step": 2700
    },
    {
      "epoch": 2.63,
      "learning_rate": 7.619834710743802e-06,
      "loss": 0.2164,
      "step": 2710
    },
    {
      "epoch": 2.64,
      "learning_rate": 7.610651974288339e-06,
      "loss": 0.2209,
      "step": 2720
    },
    {
      "epoch": 2.65,
      "learning_rate": 7.601469237832875e-06,
      "loss": 0.2179,
      "step": 2730
    },
    {
      "epoch": 2.66,
      "learning_rate": 7.592286501377412e-06,
      "loss": 0.2167,
      "step": 2740
    },
    {
      "epoch": 2.66,
      "learning_rate": 7.583103764921947e-06,
      "loss": 0.2105,
      "step": 2750
    },
    {
      "epoch": 2.67,
      "learning_rate": 7.573921028466483e-06,
      "loss": 0.2239,
      "step": 2760
    },
    {
      "epoch": 2.68,
      "learning_rate": 7.56473829201102e-06,
      "loss": 0.2139,
      "step": 2770
    },
    {
      "epoch": 2.69,
      "learning_rate": 7.555555555555556e-06,
      "loss": 0.221,
      "step": 2780
    },
    {
      "epoch": 2.7,
      "learning_rate": 7.546372819100093e-06,
      "loss": 0.2167,
      "step": 2790
    },
    {
      "epoch": 2.71,
      "learning_rate": 7.537190082644628e-06,
      "loss": 0.2228,
      "step": 2800
    },
    {
      "epoch": 2.71,
      "eval_loss": 0.057647705078125,
      "eval_runtime": 368.1765,
      "eval_samples_per_second": 7.529,
      "eval_steps_per_second": 0.06,
      "eval_wer": 16.926331341078505,
      "step": 2800
    },
    {
      "epoch": 2.72,
      "learning_rate": 7.528007346189164e-06,
      "loss": 0.2163,
      "step": 2810
    },
    {
      "epoch": 2.73,
      "learning_rate": 7.518824609733701e-06,
      "loss": 0.2021,
      "step": 2820
    },
    {
      "epoch": 2.74,
      "learning_rate": 7.5096418732782374e-06,
      "loss": 0.2131,
      "step": 2830
    },
    {
      "epoch": 2.75,
      "learning_rate": 7.500459136822774e-06,
      "loss": 0.2124,
      "step": 2840
    },
    {
      "epoch": 2.76,
      "learning_rate": 7.49127640036731e-06,
      "loss": 0.2149,
      "step": 2850
    },
    {
      "epoch": 2.77,
      "learning_rate": 7.4820936639118465e-06,
      "loss": 0.2159,
      "step": 2860
    },
    {
      "epoch": 2.78,
      "learning_rate": 7.472910927456383e-06,
      "loss": 0.2197,
      "step": 2870
    },
    {
      "epoch": 2.79,
      "learning_rate": 7.463728191000919e-06,
      "loss": 0.2097,
      "step": 2880
    },
    {
      "epoch": 2.8,
      "learning_rate": 7.454545454545456e-06,
      "loss": 0.2103,
      "step": 2890
    },
    {
      "epoch": 2.81,
      "learning_rate": 7.445362718089991e-06,
      "loss": 0.207,
      "step": 2900
    },
    {
      "epoch": 2.81,
      "eval_loss": 0.0550537109375,
      "eval_runtime": 356.0137,
      "eval_samples_per_second": 7.786,
      "eval_steps_per_second": 0.062,
      "eval_wer": 16.674501376670474,
      "step": 2900
    },
    {
      "epoch": 2.82,
      "learning_rate": 7.436179981634528e-06,
      "loss": 0.2095,
      "step": 2910
    },
    {
      "epoch": 2.83,
      "learning_rate": 7.426997245179064e-06,
      "loss": 0.2129,
      "step": 2920
    },
    {
      "epoch": 2.84,
      "learning_rate": 7.4178145087236e-06,
      "loss": 0.2131,
      "step": 2930
    },
    {
      "epoch": 2.85,
      "learning_rate": 7.408631772268137e-06,
      "loss": 0.2185,
      "step": 2940
    },
    {
      "epoch": 2.86,
      "learning_rate": 7.399449035812672e-06,
      "loss": 0.2127,
      "step": 2950
    },
    {
      "epoch": 2.87,
      "learning_rate": 7.390266299357209e-06,
      "loss": 0.2162,
      "step": 2960
    },
    {
      "epoch": 2.88,
      "learning_rate": 7.381083562901745e-06,
      "loss": 0.2231,
      "step": 2970
    },
    {
      "epoch": 2.89,
      "learning_rate": 7.371900826446282e-06,
      "loss": 0.2065,
      "step": 2980
    },
    {
      "epoch": 2.9,
      "learning_rate": 7.362718089990818e-06,
      "loss": 0.2093,
      "step": 2990
    },
    {
      "epoch": 2.91,
      "learning_rate": 7.353535353535353e-06,
      "loss": 0.2032,
      "step": 3000
    },
    {
      "epoch": 2.91,
      "eval_loss": 0.056121826171875,
      "eval_runtime": 641.1527,
      "eval_samples_per_second": 4.323,
      "eval_steps_per_second": 0.034,
      "eval_wer": 16.58384258948358,
      "step": 3000
    },
    {
      "epoch": 2.92,
      "learning_rate": 7.34435261707989e-06,
      "loss": 0.2103,
      "step": 3010
    },
    {
      "epoch": 2.93,
      "learning_rate": 7.3351698806244264e-06,
      "loss": 0.2149,
      "step": 3020
    },
    {
      "epoch": 2.94,
      "learning_rate": 7.325987144168963e-06,
      "loss": 0.2162,
      "step": 3030
    },
    {
      "epoch": 2.95,
      "learning_rate": 7.3168044077134995e-06,
      "loss": 0.2156,
      "step": 3040
    },
    {
      "epoch": 2.96,
      "learning_rate": 7.307621671258035e-06,
      "loss": 0.2136,
      "step": 3050
    },
    {
      "epoch": 2.97,
      "learning_rate": 7.298438934802572e-06,
      "loss": 0.2132,
      "step": 3060
    },
    {
      "epoch": 2.97,
      "learning_rate": 7.289256198347108e-06,
      "loss": 0.213,
      "step": 3070
    },
    {
      "epoch": 2.98,
      "learning_rate": 7.280073461891645e-06,
      "loss": 0.2087,
      "step": 3080
    },
    {
      "epoch": 2.99,
      "learning_rate": 7.270890725436181e-06,
      "loss": 0.2118,
      "step": 3090
    },
    {
      "epoch": 3.0,
      "learning_rate": 7.261707988980718e-06,
      "loss": 0.2047,
      "step": 3100
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.057281494140625,
      "eval_runtime": 445.0425,
      "eval_samples_per_second": 6.229,
      "eval_steps_per_second": 0.049,
      "eval_wer": 16.436102343697538,
      "step": 3100
    }
  ],
  "max_steps": 11000,
  "num_train_epochs": 11,
  "total_flos": 8.092049803207623e+20,
  "trial_name": null,
  "trial_params": null
}