gary109's picture
End of training
46ca906
raw
history blame
56.4 kB
{
"best_metric": 0.5352242588996887,
"best_model_checkpoint": "ai-light-dance_singing3_ft_wav2vec2-large-xlsr-53-v1/checkpoint-1152",
"epoch": 49.996539792387544,
"global_step": 3600,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.14,
"learning_rate": 4e-07,
"loss": 0.3794,
"step": 10
},
{
"epoch": 0.28,
"learning_rate": 8e-07,
"loss": 0.4002,
"step": 20
},
{
"epoch": 0.42,
"learning_rate": 1.2e-06,
"loss": 0.37,
"step": 30
},
{
"epoch": 0.55,
"learning_rate": 1.6e-06,
"loss": 0.3778,
"step": 40
},
{
"epoch": 0.69,
"learning_rate": 2e-06,
"loss": 0.3623,
"step": 50
},
{
"epoch": 0.83,
"learning_rate": 1.9943661971830985e-06,
"loss": 0.3546,
"step": 60
},
{
"epoch": 0.97,
"learning_rate": 1.988732394366197e-06,
"loss": 0.3818,
"step": 70
},
{
"epoch": 1.0,
"eval_loss": 0.55694180727005,
"eval_runtime": 126.5731,
"eval_samples_per_second": 4.203,
"eval_steps_per_second": 0.529,
"eval_wer": 0.25345503631805616,
"step": 72
},
{
"epoch": 1.11,
"learning_rate": 1.9830985915492956e-06,
"loss": 0.3717,
"step": 80
},
{
"epoch": 1.25,
"learning_rate": 1.977464788732394e-06,
"loss": 0.3721,
"step": 90
},
{
"epoch": 1.39,
"learning_rate": 1.971830985915493e-06,
"loss": 0.3563,
"step": 100
},
{
"epoch": 1.53,
"learning_rate": 1.9661971830985916e-06,
"loss": 0.3679,
"step": 110
},
{
"epoch": 1.66,
"learning_rate": 1.96056338028169e-06,
"loss": 0.3668,
"step": 120
},
{
"epoch": 1.8,
"learning_rate": 1.9549295774647887e-06,
"loss": 0.3909,
"step": 130
},
{
"epoch": 1.94,
"learning_rate": 1.9492957746478873e-06,
"loss": 0.3686,
"step": 140
},
{
"epoch": 2.0,
"eval_loss": 0.5535001754760742,
"eval_runtime": 126.7334,
"eval_samples_per_second": 4.198,
"eval_steps_per_second": 0.529,
"eval_wer": 0.25011248955454135,
"step": 144
},
{
"epoch": 2.08,
"learning_rate": 1.943661971830986e-06,
"loss": 0.3804,
"step": 150
},
{
"epoch": 2.22,
"learning_rate": 1.9380281690140844e-06,
"loss": 0.3713,
"step": 160
},
{
"epoch": 2.36,
"learning_rate": 1.932394366197183e-06,
"loss": 0.3722,
"step": 170
},
{
"epoch": 2.5,
"learning_rate": 1.926760563380282e-06,
"loss": 0.3636,
"step": 180
},
{
"epoch": 2.64,
"learning_rate": 1.9211267605633805e-06,
"loss": 0.3554,
"step": 190
},
{
"epoch": 2.78,
"learning_rate": 1.9154929577464786e-06,
"loss": 0.3742,
"step": 200
},
{
"epoch": 2.91,
"learning_rate": 1.909859154929577e-06,
"loss": 0.3562,
"step": 210
},
{
"epoch": 3.0,
"eval_loss": 0.5526291728019714,
"eval_runtime": 126.7164,
"eval_samples_per_second": 4.198,
"eval_steps_per_second": 0.529,
"eval_wer": 0.25011248955454135,
"step": 216
},
{
"epoch": 3.06,
"learning_rate": 1.904225352112676e-06,
"loss": 0.3695,
"step": 220
},
{
"epoch": 3.19,
"learning_rate": 1.8985915492957744e-06,
"loss": 0.3681,
"step": 230
},
{
"epoch": 3.33,
"learning_rate": 1.892957746478873e-06,
"loss": 0.3837,
"step": 240
},
{
"epoch": 3.47,
"learning_rate": 1.8873239436619718e-06,
"loss": 0.3692,
"step": 250
},
{
"epoch": 3.61,
"learning_rate": 1.8816901408450703e-06,
"loss": 0.3657,
"step": 260
},
{
"epoch": 3.75,
"learning_rate": 1.8760563380281689e-06,
"loss": 0.3619,
"step": 270
},
{
"epoch": 3.89,
"learning_rate": 1.8704225352112674e-06,
"loss": 0.3506,
"step": 280
},
{
"epoch": 4.0,
"eval_loss": 0.5460096597671509,
"eval_runtime": 125.8887,
"eval_samples_per_second": 4.226,
"eval_steps_per_second": 0.532,
"eval_wer": 0.2519766021726554,
"step": 288
},
{
"epoch": 4.03,
"learning_rate": 1.8647887323943662e-06,
"loss": 0.3831,
"step": 290
},
{
"epoch": 4.17,
"learning_rate": 1.8591549295774647e-06,
"loss": 0.3693,
"step": 300
},
{
"epoch": 4.3,
"learning_rate": 1.8535211267605633e-06,
"loss": 0.3649,
"step": 310
},
{
"epoch": 4.44,
"learning_rate": 1.8478873239436618e-06,
"loss": 0.3662,
"step": 320
},
{
"epoch": 4.58,
"learning_rate": 1.8422535211267604e-06,
"loss": 0.3728,
"step": 330
},
{
"epoch": 4.72,
"learning_rate": 1.8366197183098591e-06,
"loss": 0.3744,
"step": 340
},
{
"epoch": 4.86,
"learning_rate": 1.8309859154929577e-06,
"loss": 0.3634,
"step": 350
},
{
"epoch": 5.0,
"learning_rate": 1.8253521126760562e-06,
"loss": 0.369,
"step": 360
},
{
"epoch": 5.0,
"eval_loss": 0.5389720797538757,
"eval_runtime": 127.3997,
"eval_samples_per_second": 4.176,
"eval_steps_per_second": 0.526,
"eval_wer": 0.24844121617278395,
"step": 360
},
{
"epoch": 5.14,
"learning_rate": 1.8197183098591548e-06,
"loss": 0.3617,
"step": 370
},
{
"epoch": 5.28,
"learning_rate": 1.8140845070422535e-06,
"loss": 0.3786,
"step": 380
},
{
"epoch": 5.42,
"learning_rate": 1.808450704225352e-06,
"loss": 0.3606,
"step": 390
},
{
"epoch": 5.55,
"learning_rate": 1.8028169014084506e-06,
"loss": 0.3748,
"step": 400
},
{
"epoch": 5.69,
"learning_rate": 1.7971830985915492e-06,
"loss": 0.3747,
"step": 410
},
{
"epoch": 5.83,
"learning_rate": 1.791549295774648e-06,
"loss": 0.3648,
"step": 420
},
{
"epoch": 5.97,
"learning_rate": 1.7859154929577465e-06,
"loss": 0.3683,
"step": 430
},
{
"epoch": 6.0,
"eval_loss": 0.5425658822059631,
"eval_runtime": 127.6397,
"eval_samples_per_second": 4.168,
"eval_steps_per_second": 0.525,
"eval_wer": 0.24741274024554863,
"step": 432
},
{
"epoch": 6.11,
"learning_rate": 1.780281690140845e-06,
"loss": 0.3886,
"step": 440
},
{
"epoch": 6.25,
"learning_rate": 1.7746478873239436e-06,
"loss": 0.3608,
"step": 450
},
{
"epoch": 6.39,
"learning_rate": 1.7690140845070422e-06,
"loss": 0.3566,
"step": 460
},
{
"epoch": 6.53,
"learning_rate": 1.763380281690141e-06,
"loss": 0.383,
"step": 470
},
{
"epoch": 6.66,
"learning_rate": 1.7577464788732395e-06,
"loss": 0.3775,
"step": 480
},
{
"epoch": 6.8,
"learning_rate": 1.752112676056338e-06,
"loss": 0.3611,
"step": 490
},
{
"epoch": 6.94,
"learning_rate": 1.7464788732394364e-06,
"loss": 0.3541,
"step": 500
},
{
"epoch": 7.0,
"eval_loss": 0.5451989769935608,
"eval_runtime": 127.6555,
"eval_samples_per_second": 4.167,
"eval_steps_per_second": 0.525,
"eval_wer": 0.2495339718454715,
"step": 504
},
{
"epoch": 7.08,
"learning_rate": 1.7408450704225351e-06,
"loss": 0.3877,
"step": 510
},
{
"epoch": 7.22,
"learning_rate": 1.7352112676056337e-06,
"loss": 0.3706,
"step": 520
},
{
"epoch": 7.36,
"learning_rate": 1.7295774647887322e-06,
"loss": 0.3728,
"step": 530
},
{
"epoch": 7.5,
"learning_rate": 1.7239436619718308e-06,
"loss": 0.3731,
"step": 540
},
{
"epoch": 7.64,
"learning_rate": 1.7183098591549295e-06,
"loss": 0.3525,
"step": 550
},
{
"epoch": 7.78,
"learning_rate": 1.712676056338028e-06,
"loss": 0.3648,
"step": 560
},
{
"epoch": 7.91,
"learning_rate": 1.7070422535211266e-06,
"loss": 0.369,
"step": 570
},
{
"epoch": 8.0,
"eval_loss": 0.5468263030052185,
"eval_runtime": 126.4836,
"eval_samples_per_second": 4.206,
"eval_steps_per_second": 0.53,
"eval_wer": 0.24901973388185383,
"step": 576
},
{
"epoch": 8.06,
"learning_rate": 1.7014084507042252e-06,
"loss": 0.377,
"step": 580
},
{
"epoch": 8.19,
"learning_rate": 1.6957746478873237e-06,
"loss": 0.3661,
"step": 590
},
{
"epoch": 8.33,
"learning_rate": 1.6901408450704225e-06,
"loss": 0.3637,
"step": 600
},
{
"epoch": 8.47,
"learning_rate": 1.684507042253521e-06,
"loss": 0.3685,
"step": 610
},
{
"epoch": 8.61,
"learning_rate": 1.6788732394366196e-06,
"loss": 0.366,
"step": 620
},
{
"epoch": 8.75,
"learning_rate": 1.6732394366197181e-06,
"loss": 0.375,
"step": 630
},
{
"epoch": 8.89,
"learning_rate": 1.667605633802817e-06,
"loss": 0.358,
"step": 640
},
{
"epoch": 9.0,
"eval_loss": 0.5470597147941589,
"eval_runtime": 127.3076,
"eval_samples_per_second": 4.179,
"eval_steps_per_second": 0.526,
"eval_wer": 0.24529150864562577,
"step": 648
},
{
"epoch": 9.03,
"learning_rate": 1.6619718309859155e-06,
"loss": 0.384,
"step": 650
},
{
"epoch": 9.17,
"learning_rate": 1.656338028169014e-06,
"loss": 0.3688,
"step": 660
},
{
"epoch": 9.3,
"learning_rate": 1.6507042253521126e-06,
"loss": 0.3833,
"step": 670
},
{
"epoch": 9.44,
"learning_rate": 1.6450704225352113e-06,
"loss": 0.3762,
"step": 680
},
{
"epoch": 9.58,
"learning_rate": 1.6394366197183099e-06,
"loss": 0.3667,
"step": 690
},
{
"epoch": 9.72,
"learning_rate": 1.6338028169014084e-06,
"loss": 0.3657,
"step": 700
},
{
"epoch": 9.86,
"learning_rate": 1.628169014084507e-06,
"loss": 0.3698,
"step": 710
},
{
"epoch": 10.0,
"learning_rate": 1.6225352112676055e-06,
"loss": 0.3765,
"step": 720
},
{
"epoch": 10.0,
"eval_loss": 0.5376302003860474,
"eval_runtime": 126.264,
"eval_samples_per_second": 4.213,
"eval_steps_per_second": 0.531,
"eval_wer": 0.24587002635469563,
"step": 720
},
{
"epoch": 10.14,
"learning_rate": 1.6169014084507043e-06,
"loss": 0.3595,
"step": 730
},
{
"epoch": 10.28,
"learning_rate": 1.611830985915493e-06,
"loss": 0.4081,
"step": 740
},
{
"epoch": 10.42,
"learning_rate": 1.6061971830985915e-06,
"loss": 0.3759,
"step": 750
},
{
"epoch": 10.55,
"learning_rate": 1.60056338028169e-06,
"loss": 0.3841,
"step": 760
},
{
"epoch": 10.69,
"learning_rate": 1.5949295774647886e-06,
"loss": 0.3742,
"step": 770
},
{
"epoch": 10.83,
"learning_rate": 1.5892957746478871e-06,
"loss": 0.3679,
"step": 780
},
{
"epoch": 10.97,
"learning_rate": 1.5836619718309859e-06,
"loss": 0.3654,
"step": 790
},
{
"epoch": 11.0,
"eval_loss": 0.5407082438468933,
"eval_runtime": 126.3624,
"eval_samples_per_second": 4.21,
"eval_steps_per_second": 0.53,
"eval_wer": 0.24856977566368837,
"step": 792
},
{
"epoch": 11.11,
"learning_rate": 1.5780281690140844e-06,
"loss": 0.3768,
"step": 800
},
{
"epoch": 11.25,
"learning_rate": 1.572394366197183e-06,
"loss": 0.3744,
"step": 810
},
{
"epoch": 11.39,
"learning_rate": 1.5667605633802815e-06,
"loss": 0.3552,
"step": 820
},
{
"epoch": 11.53,
"learning_rate": 1.5611267605633803e-06,
"loss": 0.3837,
"step": 830
},
{
"epoch": 11.66,
"learning_rate": 1.5554929577464789e-06,
"loss": 0.3543,
"step": 840
},
{
"epoch": 11.8,
"learning_rate": 1.5498591549295774e-06,
"loss": 0.3827,
"step": 850
},
{
"epoch": 11.94,
"learning_rate": 1.544225352112676e-06,
"loss": 0.373,
"step": 860
},
{
"epoch": 12.0,
"eval_loss": 0.5389824509620667,
"eval_runtime": 127.3392,
"eval_samples_per_second": 4.178,
"eval_steps_per_second": 0.526,
"eval_wer": 0.24747701999100083,
"step": 864
},
{
"epoch": 12.08,
"learning_rate": 1.5385915492957747e-06,
"loss": 0.3676,
"step": 870
},
{
"epoch": 12.22,
"learning_rate": 1.5329577464788733e-06,
"loss": 0.3608,
"step": 880
},
{
"epoch": 12.36,
"learning_rate": 1.5273239436619718e-06,
"loss": 0.3712,
"step": 890
},
{
"epoch": 12.5,
"learning_rate": 1.5216901408450704e-06,
"loss": 0.3857,
"step": 900
},
{
"epoch": 12.64,
"learning_rate": 1.516056338028169e-06,
"loss": 0.3474,
"step": 910
},
{
"epoch": 12.78,
"learning_rate": 1.5104225352112677e-06,
"loss": 0.377,
"step": 920
},
{
"epoch": 12.91,
"learning_rate": 1.5047887323943662e-06,
"loss": 0.3606,
"step": 930
},
{
"epoch": 13.0,
"eval_loss": 0.5441206097602844,
"eval_runtime": 127.0593,
"eval_samples_per_second": 4.187,
"eval_steps_per_second": 0.527,
"eval_wer": 0.247219901009192,
"step": 936
},
{
"epoch": 13.06,
"learning_rate": 1.4991549295774648e-06,
"loss": 0.382,
"step": 940
},
{
"epoch": 13.19,
"learning_rate": 1.4935211267605633e-06,
"loss": 0.3637,
"step": 950
},
{
"epoch": 13.33,
"learning_rate": 1.487887323943662e-06,
"loss": 0.3787,
"step": 960
},
{
"epoch": 13.47,
"learning_rate": 1.4822535211267604e-06,
"loss": 0.3722,
"step": 970
},
{
"epoch": 13.61,
"learning_rate": 1.476619718309859e-06,
"loss": 0.3511,
"step": 980
},
{
"epoch": 13.75,
"learning_rate": 1.4709859154929575e-06,
"loss": 0.3836,
"step": 990
},
{
"epoch": 13.89,
"learning_rate": 1.4653521126760563e-06,
"loss": 0.369,
"step": 1000
},
{
"epoch": 14.0,
"eval_loss": 0.5388315916061401,
"eval_runtime": 126.4346,
"eval_samples_per_second": 4.208,
"eval_steps_per_second": 0.53,
"eval_wer": 0.24895545413640163,
"step": 1008
},
{
"epoch": 14.03,
"learning_rate": 1.4597183098591548e-06,
"loss": 0.374,
"step": 1010
},
{
"epoch": 14.17,
"learning_rate": 1.4540845070422534e-06,
"loss": 0.3665,
"step": 1020
},
{
"epoch": 14.3,
"learning_rate": 1.448450704225352e-06,
"loss": 0.3777,
"step": 1030
},
{
"epoch": 14.44,
"learning_rate": 1.4428169014084505e-06,
"loss": 0.3551,
"step": 1040
},
{
"epoch": 14.58,
"learning_rate": 1.4371830985915493e-06,
"loss": 0.3793,
"step": 1050
},
{
"epoch": 14.72,
"learning_rate": 1.4315492957746478e-06,
"loss": 0.375,
"step": 1060
},
{
"epoch": 14.86,
"learning_rate": 1.4259154929577464e-06,
"loss": 0.3518,
"step": 1070
},
{
"epoch": 15.0,
"learning_rate": 1.420281690140845e-06,
"loss": 0.3762,
"step": 1080
},
{
"epoch": 15.0,
"eval_loss": 0.5389070510864258,
"eval_runtime": 126.3932,
"eval_samples_per_second": 4.209,
"eval_steps_per_second": 0.53,
"eval_wer": 0.24966253133637592,
"step": 1080
},
{
"epoch": 15.14,
"learning_rate": 1.4146478873239437e-06,
"loss": 0.3805,
"step": 1090
},
{
"epoch": 15.28,
"learning_rate": 1.4090140845070422e-06,
"loss": 0.3789,
"step": 1100
},
{
"epoch": 15.42,
"learning_rate": 1.4033802816901408e-06,
"loss": 0.3621,
"step": 1110
},
{
"epoch": 15.55,
"learning_rate": 1.3977464788732393e-06,
"loss": 0.3749,
"step": 1120
},
{
"epoch": 15.69,
"learning_rate": 1.392112676056338e-06,
"loss": 0.3548,
"step": 1130
},
{
"epoch": 15.83,
"learning_rate": 1.3864788732394366e-06,
"loss": 0.3794,
"step": 1140
},
{
"epoch": 15.97,
"learning_rate": 1.3808450704225352e-06,
"loss": 0.3821,
"step": 1150
},
{
"epoch": 16.0,
"eval_loss": 0.5352242588996887,
"eval_runtime": 127.705,
"eval_samples_per_second": 4.166,
"eval_steps_per_second": 0.525,
"eval_wer": 0.24895545413640163,
"step": 1152
},
{
"epoch": 16.11,
"learning_rate": 1.3752112676056337e-06,
"loss": 0.3735,
"step": 1160
},
{
"epoch": 16.25,
"learning_rate": 1.3695774647887323e-06,
"loss": 0.3688,
"step": 1170
},
{
"epoch": 16.39,
"learning_rate": 1.363943661971831e-06,
"loss": 0.353,
"step": 1180
},
{
"epoch": 16.53,
"learning_rate": 1.3583098591549296e-06,
"loss": 0.3848,
"step": 1190
},
{
"epoch": 16.66,
"learning_rate": 1.3526760563380281e-06,
"loss": 0.3767,
"step": 1200
},
{
"epoch": 16.8,
"learning_rate": 1.3470422535211267e-06,
"loss": 0.3533,
"step": 1210
},
{
"epoch": 16.94,
"learning_rate": 1.3414084507042255e-06,
"loss": 0.3605,
"step": 1220
},
{
"epoch": 17.0,
"eval_loss": 0.5433817505836487,
"eval_runtime": 128.2697,
"eval_samples_per_second": 4.148,
"eval_steps_per_second": 0.522,
"eval_wer": 0.24921257311821046,
"step": 1224
},
{
"epoch": 17.08,
"learning_rate": 1.335774647887324e-06,
"loss": 0.3997,
"step": 1230
},
{
"epoch": 17.22,
"learning_rate": 1.3301408450704226e-06,
"loss": 0.353,
"step": 1240
},
{
"epoch": 17.36,
"learning_rate": 1.324507042253521e-06,
"loss": 0.3618,
"step": 1250
},
{
"epoch": 17.5,
"learning_rate": 1.3188732394366199e-06,
"loss": 0.3795,
"step": 1260
},
{
"epoch": 17.64,
"learning_rate": 1.3132394366197182e-06,
"loss": 0.353,
"step": 1270
},
{
"epoch": 17.78,
"learning_rate": 1.3076056338028168e-06,
"loss": 0.385,
"step": 1280
},
{
"epoch": 17.91,
"learning_rate": 1.3019718309859153e-06,
"loss": 0.3703,
"step": 1290
},
{
"epoch": 18.0,
"eval_loss": 0.5415087342262268,
"eval_runtime": 127.5674,
"eval_samples_per_second": 4.17,
"eval_steps_per_second": 0.525,
"eval_wer": 0.24998393006363695,
"step": 1296
},
{
"epoch": 18.06,
"learning_rate": 1.2963380281690139e-06,
"loss": 0.3809,
"step": 1300
},
{
"epoch": 18.19,
"learning_rate": 1.2907042253521126e-06,
"loss": 0.3718,
"step": 1310
},
{
"epoch": 18.33,
"learning_rate": 1.2850704225352112e-06,
"loss": 0.3714,
"step": 1320
},
{
"epoch": 18.47,
"learning_rate": 1.2794366197183097e-06,
"loss": 0.3692,
"step": 1330
},
{
"epoch": 18.61,
"learning_rate": 1.2738028169014083e-06,
"loss": 0.3758,
"step": 1340
},
{
"epoch": 18.75,
"learning_rate": 1.268169014084507e-06,
"loss": 0.3642,
"step": 1350
},
{
"epoch": 18.89,
"learning_rate": 1.2625352112676056e-06,
"loss": 0.3667,
"step": 1360
},
{
"epoch": 19.0,
"eval_loss": 0.5498944520950317,
"eval_runtime": 127.5121,
"eval_samples_per_second": 4.172,
"eval_steps_per_second": 0.525,
"eval_wer": 0.24869833515459278,
"step": 1368
},
{
"epoch": 19.03,
"learning_rate": 1.2569014084507041e-06,
"loss": 0.3731,
"step": 1370
},
{
"epoch": 19.17,
"learning_rate": 1.2512676056338027e-06,
"loss": 0.3593,
"step": 1380
},
{
"epoch": 19.3,
"learning_rate": 1.2456338028169014e-06,
"loss": 0.3763,
"step": 1390
},
{
"epoch": 19.44,
"learning_rate": 1.24e-06,
"loss": 0.3641,
"step": 1400
},
{
"epoch": 19.58,
"learning_rate": 1.2343661971830985e-06,
"loss": 0.3697,
"step": 1410
},
{
"epoch": 19.72,
"learning_rate": 1.228732394366197e-06,
"loss": 0.3714,
"step": 1420
},
{
"epoch": 19.86,
"learning_rate": 1.2230985915492956e-06,
"loss": 0.3666,
"step": 1430
},
{
"epoch": 20.0,
"learning_rate": 1.2174647887323944e-06,
"loss": 0.3703,
"step": 1440
},
{
"epoch": 20.0,
"eval_loss": 0.5482383966445923,
"eval_runtime": 127.6825,
"eval_samples_per_second": 4.167,
"eval_steps_per_second": 0.525,
"eval_wer": 0.24940541235456706,
"step": 1440
},
{
"epoch": 20.14,
"learning_rate": 1.211830985915493e-06,
"loss": 0.3678,
"step": 1450
},
{
"epoch": 20.28,
"learning_rate": 1.2061971830985915e-06,
"loss": 0.3657,
"step": 1460
},
{
"epoch": 20.42,
"learning_rate": 1.20056338028169e-06,
"loss": 0.3614,
"step": 1470
},
{
"epoch": 20.55,
"learning_rate": 1.1949295774647888e-06,
"loss": 0.3816,
"step": 1480
},
{
"epoch": 20.69,
"learning_rate": 1.1892957746478874e-06,
"loss": 0.3652,
"step": 1490
},
{
"epoch": 20.83,
"learning_rate": 1.183661971830986e-06,
"loss": 0.3814,
"step": 1500
},
{
"epoch": 20.97,
"learning_rate": 1.1780281690140845e-06,
"loss": 0.369,
"step": 1510
},
{
"epoch": 21.0,
"eval_loss": 0.5417876839637756,
"eval_runtime": 127.2993,
"eval_samples_per_second": 4.179,
"eval_steps_per_second": 0.526,
"eval_wer": 0.2504338882818024,
"step": 1512
},
{
"epoch": 21.11,
"learning_rate": 1.1723943661971832e-06,
"loss": 0.3693,
"step": 1520
},
{
"epoch": 21.25,
"learning_rate": 1.1667605633802818e-06,
"loss": 0.3878,
"step": 1530
},
{
"epoch": 21.39,
"learning_rate": 1.1611267605633803e-06,
"loss": 0.3766,
"step": 1540
},
{
"epoch": 21.53,
"learning_rate": 1.1554929577464789e-06,
"loss": 0.3791,
"step": 1550
},
{
"epoch": 21.66,
"learning_rate": 1.1498591549295772e-06,
"loss": 0.3511,
"step": 1560
},
{
"epoch": 21.8,
"learning_rate": 1.144225352112676e-06,
"loss": 0.3627,
"step": 1570
},
{
"epoch": 21.94,
"learning_rate": 1.1385915492957745e-06,
"loss": 0.3708,
"step": 1580
},
{
"epoch": 22.0,
"eval_loss": 0.5436869263648987,
"eval_runtime": 127.6416,
"eval_samples_per_second": 4.168,
"eval_steps_per_second": 0.525,
"eval_wer": 0.2478626984637141,
"step": 1584
},
{
"epoch": 22.08,
"learning_rate": 1.132957746478873e-06,
"loss": 0.382,
"step": 1590
},
{
"epoch": 22.22,
"learning_rate": 1.1273239436619716e-06,
"loss": 0.3679,
"step": 1600
},
{
"epoch": 22.36,
"learning_rate": 1.1216901408450704e-06,
"loss": 0.3632,
"step": 1610
},
{
"epoch": 22.5,
"learning_rate": 1.116056338028169e-06,
"loss": 0.3814,
"step": 1620
},
{
"epoch": 22.64,
"learning_rate": 1.1104225352112675e-06,
"loss": 0.3719,
"step": 1630
},
{
"epoch": 22.78,
"learning_rate": 1.104788732394366e-06,
"loss": 0.3633,
"step": 1640
},
{
"epoch": 22.91,
"learning_rate": 1.0991549295774648e-06,
"loss": 0.3609,
"step": 1650
},
{
"epoch": 23.0,
"eval_loss": 0.545259952545166,
"eval_runtime": 126.5135,
"eval_samples_per_second": 4.205,
"eval_steps_per_second": 0.53,
"eval_wer": 0.2479912579546185,
"step": 1656
},
{
"epoch": 23.06,
"learning_rate": 1.0935211267605633e-06,
"loss": 0.3762,
"step": 1660
},
{
"epoch": 23.19,
"learning_rate": 1.087887323943662e-06,
"loss": 0.3726,
"step": 1670
},
{
"epoch": 23.33,
"learning_rate": 1.0822535211267604e-06,
"loss": 0.3704,
"step": 1680
},
{
"epoch": 23.47,
"learning_rate": 1.076619718309859e-06,
"loss": 0.3788,
"step": 1690
},
{
"epoch": 23.61,
"learning_rate": 1.0709859154929578e-06,
"loss": 0.3743,
"step": 1700
},
{
"epoch": 23.75,
"learning_rate": 1.0653521126760563e-06,
"loss": 0.3677,
"step": 1710
},
{
"epoch": 23.89,
"learning_rate": 1.0597183098591549e-06,
"loss": 0.3534,
"step": 1720
},
{
"epoch": 24.0,
"eval_loss": 0.539250373840332,
"eval_runtime": 126.0,
"eval_samples_per_second": 4.222,
"eval_steps_per_second": 0.532,
"eval_wer": 0.24837693642733175,
"step": 1728
},
{
"epoch": 24.03,
"learning_rate": 1.0540845070422534e-06,
"loss": 0.3721,
"step": 1730
},
{
"epoch": 24.17,
"learning_rate": 1.0484507042253522e-06,
"loss": 0.3682,
"step": 1740
},
{
"epoch": 24.3,
"learning_rate": 1.0428169014084507e-06,
"loss": 0.3674,
"step": 1750
},
{
"epoch": 24.44,
"learning_rate": 1.0371830985915493e-06,
"loss": 0.3616,
"step": 1760
},
{
"epoch": 24.58,
"learning_rate": 1.0315492957746478e-06,
"loss": 0.3809,
"step": 1770
},
{
"epoch": 24.72,
"learning_rate": 1.0259154929577466e-06,
"loss": 0.3643,
"step": 1780
},
{
"epoch": 24.86,
"learning_rate": 1.0202816901408451e-06,
"loss": 0.3706,
"step": 1790
},
{
"epoch": 25.0,
"learning_rate": 1.0146478873239437e-06,
"loss": 0.3656,
"step": 1800
},
{
"epoch": 25.0,
"eval_loss": 0.536328911781311,
"eval_runtime": 127.0748,
"eval_samples_per_second": 4.187,
"eval_steps_per_second": 0.527,
"eval_wer": 0.24766985922735746,
"step": 1800
},
{
"epoch": 25.14,
"learning_rate": 1.0090140845070422e-06,
"loss": 0.3793,
"step": 1810
},
{
"epoch": 25.28,
"learning_rate": 1.0033802816901408e-06,
"loss": 0.3599,
"step": 1820
},
{
"epoch": 25.42,
"learning_rate": 9.977464788732393e-07,
"loss": 0.3615,
"step": 1830
},
{
"epoch": 25.55,
"learning_rate": 9.921126760563379e-07,
"loss": 0.3651,
"step": 1840
},
{
"epoch": 25.69,
"learning_rate": 9.864788732394366e-07,
"loss": 0.3711,
"step": 1850
},
{
"epoch": 25.83,
"learning_rate": 9.808450704225352e-07,
"loss": 0.3746,
"step": 1860
},
{
"epoch": 25.97,
"learning_rate": 9.752112676056337e-07,
"loss": 0.3713,
"step": 1870
},
{
"epoch": 26.0,
"eval_loss": 0.540567934513092,
"eval_runtime": 126.6741,
"eval_samples_per_second": 4.2,
"eval_steps_per_second": 0.529,
"eval_wer": 0.24612714533650445,
"step": 1872
},
{
"epoch": 26.11,
"learning_rate": 9.695774647887323e-07,
"loss": 0.3727,
"step": 1880
},
{
"epoch": 26.25,
"learning_rate": 9.63943661971831e-07,
"loss": 0.3724,
"step": 1890
},
{
"epoch": 26.39,
"learning_rate": 9.583098591549296e-07,
"loss": 0.3564,
"step": 1900
},
{
"epoch": 26.53,
"learning_rate": 9.526760563380282e-07,
"loss": 0.3656,
"step": 1910
},
{
"epoch": 26.66,
"learning_rate": 9.470422535211268e-07,
"loss": 0.356,
"step": 1920
},
{
"epoch": 26.8,
"learning_rate": 9.414084507042253e-07,
"loss": 0.3826,
"step": 1930
},
{
"epoch": 26.94,
"learning_rate": 9.357746478873238e-07,
"loss": 0.3572,
"step": 1940
},
{
"epoch": 27.0,
"eval_loss": 0.536870539188385,
"eval_runtime": 128.7211,
"eval_samples_per_second": 4.133,
"eval_steps_per_second": 0.521,
"eval_wer": 0.2465128238092177,
"step": 1944
},
{
"epoch": 27.08,
"learning_rate": 9.301408450704225e-07,
"loss": 0.3724,
"step": 1950
},
{
"epoch": 27.22,
"learning_rate": 9.24507042253521e-07,
"loss": 0.3663,
"step": 1960
},
{
"epoch": 27.36,
"learning_rate": 9.188732394366197e-07,
"loss": 0.3558,
"step": 1970
},
{
"epoch": 27.5,
"learning_rate": 9.132394366197182e-07,
"loss": 0.3715,
"step": 1980
},
{
"epoch": 27.64,
"learning_rate": 9.076056338028169e-07,
"loss": 0.3579,
"step": 1990
},
{
"epoch": 27.78,
"learning_rate": 9.019718309859154e-07,
"loss": 0.3775,
"step": 2000
},
{
"epoch": 27.91,
"learning_rate": 8.963380281690141e-07,
"loss": 0.3665,
"step": 2010
},
{
"epoch": 28.0,
"eval_loss": 0.5374864935874939,
"eval_runtime": 126.8744,
"eval_samples_per_second": 4.193,
"eval_steps_per_second": 0.528,
"eval_wer": 0.24664138330012214,
"step": 2016
},
{
"epoch": 28.06,
"learning_rate": 8.907042253521126e-07,
"loss": 0.3765,
"step": 2020
},
{
"epoch": 28.19,
"learning_rate": 8.850704225352113e-07,
"loss": 0.367,
"step": 2030
},
{
"epoch": 28.33,
"learning_rate": 8.794366197183098e-07,
"loss": 0.3783,
"step": 2040
},
{
"epoch": 28.47,
"learning_rate": 8.738028169014085e-07,
"loss": 0.3651,
"step": 2050
},
{
"epoch": 28.61,
"learning_rate": 8.68169014084507e-07,
"loss": 0.3717,
"step": 2060
},
{
"epoch": 28.75,
"learning_rate": 8.625352112676056e-07,
"loss": 0.374,
"step": 2070
},
{
"epoch": 28.89,
"learning_rate": 8.569014084507041e-07,
"loss": 0.3748,
"step": 2080
},
{
"epoch": 29.0,
"eval_loss": 0.536729633808136,
"eval_runtime": 127.2778,
"eval_samples_per_second": 4.18,
"eval_steps_per_second": 0.526,
"eval_wer": 0.2479912579546185,
"step": 2088
},
{
"epoch": 29.03,
"learning_rate": 8.512676056338027e-07,
"loss": 0.3693,
"step": 2090
},
{
"epoch": 29.17,
"learning_rate": 8.456338028169013e-07,
"loss": 0.3585,
"step": 2100
},
{
"epoch": 29.3,
"learning_rate": 8.399999999999999e-07,
"loss": 0.3669,
"step": 2110
},
{
"epoch": 29.44,
"learning_rate": 8.343661971830986e-07,
"loss": 0.3709,
"step": 2120
},
{
"epoch": 29.58,
"learning_rate": 8.287323943661971e-07,
"loss": 0.3646,
"step": 2130
},
{
"epoch": 29.72,
"learning_rate": 8.230985915492958e-07,
"loss": 0.3628,
"step": 2140
},
{
"epoch": 29.86,
"learning_rate": 8.174647887323943e-07,
"loss": 0.3709,
"step": 2150
},
{
"epoch": 30.0,
"learning_rate": 8.11830985915493e-07,
"loss": 0.3769,
"step": 2160
},
{
"epoch": 30.0,
"eval_loss": 0.5359097719192505,
"eval_runtime": 127.9102,
"eval_samples_per_second": 4.159,
"eval_steps_per_second": 0.524,
"eval_wer": 0.24587002635469563,
"step": 2160
},
{
"epoch": 30.14,
"learning_rate": 8.061971830985915e-07,
"loss": 0.3729,
"step": 2170
},
{
"epoch": 30.28,
"learning_rate": 8.005633802816902e-07,
"loss": 0.3729,
"step": 2180
},
{
"epoch": 30.42,
"learning_rate": 7.949295774647887e-07,
"loss": 0.3582,
"step": 2190
},
{
"epoch": 30.55,
"learning_rate": 7.892957746478873e-07,
"loss": 0.3717,
"step": 2200
},
{
"epoch": 30.69,
"learning_rate": 7.836619718309859e-07,
"loss": 0.3557,
"step": 2210
},
{
"epoch": 30.83,
"learning_rate": 7.780281690140845e-07,
"loss": 0.3725,
"step": 2220
},
{
"epoch": 30.97,
"learning_rate": 7.72394366197183e-07,
"loss": 0.3634,
"step": 2230
},
{
"epoch": 31.0,
"eval_loss": 0.5417142510414124,
"eval_runtime": 127.5766,
"eval_samples_per_second": 4.17,
"eval_steps_per_second": 0.525,
"eval_wer": 0.2470913415182876,
"step": 2232
},
{
"epoch": 31.11,
"learning_rate": 7.667605633802816e-07,
"loss": 0.3762,
"step": 2240
},
{
"epoch": 31.25,
"learning_rate": 7.611267605633802e-07,
"loss": 0.3653,
"step": 2250
},
{
"epoch": 31.39,
"learning_rate": 7.554929577464788e-07,
"loss": 0.3425,
"step": 2260
},
{
"epoch": 31.53,
"learning_rate": 7.498591549295774e-07,
"loss": 0.363,
"step": 2270
},
{
"epoch": 31.66,
"learning_rate": 7.44225352112676e-07,
"loss": 0.36,
"step": 2280
},
{
"epoch": 31.8,
"learning_rate": 7.385915492957746e-07,
"loss": 0.3709,
"step": 2290
},
{
"epoch": 31.94,
"learning_rate": 7.329577464788732e-07,
"loss": 0.3627,
"step": 2300
},
{
"epoch": 32.0,
"eval_loss": 0.5398467779159546,
"eval_runtime": 128.1832,
"eval_samples_per_second": 4.15,
"eval_steps_per_second": 0.523,
"eval_wer": 0.24734846050009643,
"step": 2304
},
{
"epoch": 32.08,
"learning_rate": 7.273239436619719e-07,
"loss": 0.3842,
"step": 2310
},
{
"epoch": 32.22,
"learning_rate": 7.216901408450704e-07,
"loss": 0.3702,
"step": 2320
},
{
"epoch": 32.36,
"learning_rate": 7.16056338028169e-07,
"loss": 0.3721,
"step": 2330
},
{
"epoch": 32.5,
"learning_rate": 7.104225352112676e-07,
"loss": 0.3728,
"step": 2340
},
{
"epoch": 32.64,
"learning_rate": 7.047887323943662e-07,
"loss": 0.3482,
"step": 2350
},
{
"epoch": 32.78,
"learning_rate": 6.991549295774648e-07,
"loss": 0.3691,
"step": 2360
},
{
"epoch": 32.91,
"learning_rate": 6.935211267605634e-07,
"loss": 0.3497,
"step": 2370
},
{
"epoch": 33.0,
"eval_loss": 0.5427743792533875,
"eval_runtime": 126.3596,
"eval_samples_per_second": 4.21,
"eval_steps_per_second": 0.53,
"eval_wer": 0.24831265668187955,
"step": 2376
},
{
"epoch": 33.06,
"learning_rate": 6.878873239436619e-07,
"loss": 0.3807,
"step": 2380
},
{
"epoch": 33.19,
"learning_rate": 6.822535211267605e-07,
"loss": 0.3613,
"step": 2390
},
{
"epoch": 33.33,
"learning_rate": 6.766197183098591e-07,
"loss": 0.3699,
"step": 2400
},
{
"epoch": 33.47,
"learning_rate": 6.709859154929577e-07,
"loss": 0.3712,
"step": 2410
},
{
"epoch": 33.61,
"learning_rate": 6.653521126760563e-07,
"loss": 0.3717,
"step": 2420
},
{
"epoch": 33.75,
"learning_rate": 6.597183098591549e-07,
"loss": 0.3586,
"step": 2430
},
{
"epoch": 33.89,
"learning_rate": 6.540845070422535e-07,
"loss": 0.3479,
"step": 2440
},
{
"epoch": 34.0,
"eval_loss": 0.5389805436134338,
"eval_runtime": 126.1536,
"eval_samples_per_second": 4.217,
"eval_steps_per_second": 0.531,
"eval_wer": 0.24811981744552292,
"step": 2448
},
{
"epoch": 34.03,
"learning_rate": 6.484507042253521e-07,
"loss": 0.3858,
"step": 2450
},
{
"epoch": 34.17,
"learning_rate": 6.428169014084506e-07,
"loss": 0.3618,
"step": 2460
},
{
"epoch": 34.3,
"learning_rate": 6.371830985915493e-07,
"loss": 0.3707,
"step": 2470
},
{
"epoch": 34.44,
"learning_rate": 6.315492957746478e-07,
"loss": 0.3718,
"step": 2480
},
{
"epoch": 34.58,
"learning_rate": 6.259154929577465e-07,
"loss": 0.3534,
"step": 2490
},
{
"epoch": 34.72,
"learning_rate": 6.20281690140845e-07,
"loss": 0.3613,
"step": 2500
},
{
"epoch": 34.86,
"learning_rate": 6.146478873239437e-07,
"loss": 0.3623,
"step": 2510
},
{
"epoch": 35.0,
"learning_rate": 6.090140845070423e-07,
"loss": 0.363,
"step": 2520
},
{
"epoch": 35.0,
"eval_loss": 0.5388094782829285,
"eval_runtime": 126.3817,
"eval_samples_per_second": 4.209,
"eval_steps_per_second": 0.53,
"eval_wer": 0.2477984187182619,
"step": 2520
},
{
"epoch": 35.14,
"learning_rate": 6.033802816901408e-07,
"loss": 0.3673,
"step": 2530
},
{
"epoch": 35.28,
"learning_rate": 5.977464788732394e-07,
"loss": 0.3787,
"step": 2540
},
{
"epoch": 35.42,
"learning_rate": 5.92112676056338e-07,
"loss": 0.3589,
"step": 2550
},
{
"epoch": 35.55,
"learning_rate": 5.864788732394366e-07,
"loss": 0.3732,
"step": 2560
},
{
"epoch": 35.69,
"learning_rate": 5.808450704225352e-07,
"loss": 0.3626,
"step": 2570
},
{
"epoch": 35.83,
"learning_rate": 5.752112676056338e-07,
"loss": 0.3671,
"step": 2580
},
{
"epoch": 35.97,
"learning_rate": 5.695774647887323e-07,
"loss": 0.3622,
"step": 2590
},
{
"epoch": 36.0,
"eval_loss": 0.5396317839622498,
"eval_runtime": 127.4298,
"eval_samples_per_second": 4.175,
"eval_steps_per_second": 0.526,
"eval_wer": 0.24895545413640163,
"step": 2592
},
{
"epoch": 36.11,
"learning_rate": 5.63943661971831e-07,
"loss": 0.3753,
"step": 2600
},
{
"epoch": 36.25,
"learning_rate": 5.583098591549295e-07,
"loss": 0.3594,
"step": 2610
},
{
"epoch": 36.39,
"learning_rate": 5.526760563380282e-07,
"loss": 0.3507,
"step": 2620
},
{
"epoch": 36.53,
"learning_rate": 5.470422535211267e-07,
"loss": 0.3842,
"step": 2630
},
{
"epoch": 36.66,
"learning_rate": 5.414084507042254e-07,
"loss": 0.3699,
"step": 2640
},
{
"epoch": 36.8,
"learning_rate": 5.357746478873239e-07,
"loss": 0.368,
"step": 2650
},
{
"epoch": 36.94,
"learning_rate": 5.301408450704226e-07,
"loss": 0.3685,
"step": 2660
},
{
"epoch": 37.0,
"eval_loss": 0.5408422946929932,
"eval_runtime": 127.9531,
"eval_samples_per_second": 4.158,
"eval_steps_per_second": 0.524,
"eval_wer": 0.2478626984637141,
"step": 2664
},
{
"epoch": 37.08,
"learning_rate": 5.24507042253521e-07,
"loss": 0.3811,
"step": 2670
},
{
"epoch": 37.22,
"learning_rate": 5.188732394366197e-07,
"loss": 0.3746,
"step": 2680
},
{
"epoch": 37.36,
"learning_rate": 5.132394366197182e-07,
"loss": 0.3654,
"step": 2690
},
{
"epoch": 37.5,
"learning_rate": 5.076056338028169e-07,
"loss": 0.3634,
"step": 2700
},
{
"epoch": 37.64,
"learning_rate": 5.019718309859154e-07,
"loss": 0.3468,
"step": 2710
},
{
"epoch": 37.78,
"learning_rate": 4.96338028169014e-07,
"loss": 0.3737,
"step": 2720
},
{
"epoch": 37.91,
"learning_rate": 4.907042253521126e-07,
"loss": 0.356,
"step": 2730
},
{
"epoch": 38.0,
"eval_loss": 0.5385244488716125,
"eval_runtime": 128.0588,
"eval_samples_per_second": 4.154,
"eval_steps_per_second": 0.523,
"eval_wer": 0.24593430610014785,
"step": 2736
},
{
"epoch": 38.06,
"learning_rate": 4.850704225352112e-07,
"loss": 0.3759,
"step": 2740
},
{
"epoch": 38.19,
"learning_rate": 4.794366197183099e-07,
"loss": 0.3653,
"step": 2750
},
{
"epoch": 38.33,
"learning_rate": 4.7380281690140846e-07,
"loss": 0.3482,
"step": 2760
},
{
"epoch": 38.47,
"learning_rate": 4.6816901408450706e-07,
"loss": 0.3606,
"step": 2770
},
{
"epoch": 38.61,
"learning_rate": 4.6253521126760566e-07,
"loss": 0.3666,
"step": 2780
},
{
"epoch": 38.75,
"learning_rate": 4.5690140845070416e-07,
"loss": 0.3706,
"step": 2790
},
{
"epoch": 38.89,
"learning_rate": 4.5126760563380276e-07,
"loss": 0.3529,
"step": 2800
},
{
"epoch": 39.0,
"eval_loss": 0.5388730764389038,
"eval_runtime": 127.0748,
"eval_samples_per_second": 4.187,
"eval_steps_per_second": 0.527,
"eval_wer": 0.24670566304557434,
"step": 2808
},
{
"epoch": 39.03,
"learning_rate": 4.4563380281690137e-07,
"loss": 0.3863,
"step": 2810
},
{
"epoch": 39.17,
"learning_rate": 4.3999999999999997e-07,
"loss": 0.3677,
"step": 2820
},
{
"epoch": 39.3,
"learning_rate": 4.3436619718309857e-07,
"loss": 0.3677,
"step": 2830
},
{
"epoch": 39.44,
"learning_rate": 4.287323943661972e-07,
"loss": 0.353,
"step": 2840
},
{
"epoch": 39.58,
"learning_rate": 4.230985915492958e-07,
"loss": 0.3658,
"step": 2850
},
{
"epoch": 39.72,
"learning_rate": 4.1746478873239433e-07,
"loss": 0.3712,
"step": 2860
},
{
"epoch": 39.86,
"learning_rate": 4.1183098591549293e-07,
"loss": 0.3577,
"step": 2870
},
{
"epoch": 40.0,
"learning_rate": 4.0619718309859153e-07,
"loss": 0.3702,
"step": 2880
},
{
"epoch": 40.0,
"eval_loss": 0.5392444729804993,
"eval_runtime": 128.3773,
"eval_samples_per_second": 4.144,
"eval_steps_per_second": 0.522,
"eval_wer": 0.24818409719097512,
"step": 2880
},
{
"epoch": 40.14,
"learning_rate": 4.0056338028169014e-07,
"loss": 0.3731,
"step": 2890
},
{
"epoch": 40.28,
"learning_rate": 3.9492957746478874e-07,
"loss": 0.3599,
"step": 2900
},
{
"epoch": 40.42,
"learning_rate": 3.8929577464788734e-07,
"loss": 0.3603,
"step": 2910
},
{
"epoch": 40.55,
"learning_rate": 3.836619718309859e-07,
"loss": 0.3668,
"step": 2920
},
{
"epoch": 40.69,
"learning_rate": 3.7802816901408444e-07,
"loss": 0.3518,
"step": 2930
},
{
"epoch": 40.83,
"learning_rate": 3.7239436619718305e-07,
"loss": 0.3774,
"step": 2940
},
{
"epoch": 40.97,
"learning_rate": 3.6676056338028165e-07,
"loss": 0.3645,
"step": 2950
},
{
"epoch": 41.0,
"eval_loss": 0.5407747030258179,
"eval_runtime": 128.2235,
"eval_samples_per_second": 4.149,
"eval_steps_per_second": 0.523,
"eval_wer": 0.24670566304557434,
"step": 2952
},
{
"epoch": 41.11,
"learning_rate": 3.6112676056338025e-07,
"loss": 0.3806,
"step": 2960
},
{
"epoch": 41.25,
"learning_rate": 3.5549295774647886e-07,
"loss": 0.3694,
"step": 2970
},
{
"epoch": 41.39,
"learning_rate": 3.4985915492957746e-07,
"loss": 0.3531,
"step": 2980
},
{
"epoch": 41.53,
"learning_rate": 3.4422535211267606e-07,
"loss": 0.3604,
"step": 2990
},
{
"epoch": 41.66,
"learning_rate": 3.3859154929577466e-07,
"loss": 0.3691,
"step": 3000
},
{
"epoch": 41.8,
"learning_rate": 3.329577464788732e-07,
"loss": 0.3744,
"step": 3010
},
{
"epoch": 41.94,
"learning_rate": 3.273239436619718e-07,
"loss": 0.3489,
"step": 3020
},
{
"epoch": 42.0,
"eval_loss": 0.5406195521354675,
"eval_runtime": 128.1617,
"eval_samples_per_second": 4.151,
"eval_steps_per_second": 0.523,
"eval_wer": 0.24741274024554863,
"step": 3024
},
{
"epoch": 42.08,
"learning_rate": 3.216901408450704e-07,
"loss": 0.3823,
"step": 3030
},
{
"epoch": 42.22,
"learning_rate": 3.16056338028169e-07,
"loss": 0.3526,
"step": 3040
},
{
"epoch": 42.36,
"learning_rate": 3.104225352112676e-07,
"loss": 0.3669,
"step": 3050
},
{
"epoch": 42.5,
"learning_rate": 3.047887323943662e-07,
"loss": 0.3766,
"step": 3060
},
{
"epoch": 42.64,
"learning_rate": 2.9971830985915494e-07,
"loss": 0.3746,
"step": 3070
},
{
"epoch": 42.78,
"learning_rate": 2.9408450704225354e-07,
"loss": 0.3825,
"step": 3080
},
{
"epoch": 42.91,
"learning_rate": 2.8845070422535214e-07,
"loss": 0.3556,
"step": 3090
},
{
"epoch": 43.0,
"eval_loss": 0.5401654839515686,
"eval_runtime": 128.4726,
"eval_samples_per_second": 4.141,
"eval_steps_per_second": 0.522,
"eval_wer": 0.2471556212637398,
"step": 3096
},
{
"epoch": 43.06,
"learning_rate": 2.828169014084507e-07,
"loss": 0.3766,
"step": 3100
},
{
"epoch": 43.19,
"learning_rate": 2.7718309859154924e-07,
"loss": 0.3609,
"step": 3110
},
{
"epoch": 43.33,
"learning_rate": 2.7154929577464784e-07,
"loss": 0.3619,
"step": 3120
},
{
"epoch": 43.47,
"learning_rate": 2.6591549295774645e-07,
"loss": 0.3635,
"step": 3130
},
{
"epoch": 43.61,
"learning_rate": 2.6028169014084505e-07,
"loss": 0.3668,
"step": 3140
},
{
"epoch": 43.75,
"learning_rate": 2.5464788732394365e-07,
"loss": 0.368,
"step": 3150
},
{
"epoch": 43.89,
"learning_rate": 2.4901408450704226e-07,
"loss": 0.356,
"step": 3160
},
{
"epoch": 44.0,
"eval_loss": 0.538616418838501,
"eval_runtime": 127.2122,
"eval_samples_per_second": 4.182,
"eval_steps_per_second": 0.527,
"eval_wer": 0.24676994279102654,
"step": 3168
},
{
"epoch": 44.03,
"learning_rate": 2.433802816901408e-07,
"loss": 0.3737,
"step": 3170
},
{
"epoch": 44.17,
"learning_rate": 2.3774647887323944e-07,
"loss": 0.3586,
"step": 3180
},
{
"epoch": 44.3,
"learning_rate": 2.32112676056338e-07,
"loss": 0.3721,
"step": 3190
},
{
"epoch": 44.44,
"learning_rate": 2.264788732394366e-07,
"loss": 0.3611,
"step": 3200
},
{
"epoch": 44.58,
"learning_rate": 2.208450704225352e-07,
"loss": 0.3664,
"step": 3210
},
{
"epoch": 44.72,
"learning_rate": 2.152112676056338e-07,
"loss": 0.354,
"step": 3220
},
{
"epoch": 44.86,
"learning_rate": 2.095774647887324e-07,
"loss": 0.3545,
"step": 3230
},
{
"epoch": 45.0,
"learning_rate": 2.0394366197183097e-07,
"loss": 0.3632,
"step": 3240
},
{
"epoch": 45.0,
"eval_loss": 0.5401559472084045,
"eval_runtime": 126.237,
"eval_samples_per_second": 4.214,
"eval_steps_per_second": 0.531,
"eval_wer": 0.24631998457286108,
"step": 3240
},
{
"epoch": 45.14,
"learning_rate": 1.9830985915492958e-07,
"loss": 0.3597,
"step": 3250
},
{
"epoch": 45.28,
"learning_rate": 1.9267605633802815e-07,
"loss": 0.3655,
"step": 3260
},
{
"epoch": 45.42,
"learning_rate": 1.8704225352112676e-07,
"loss": 0.3644,
"step": 3270
},
{
"epoch": 45.55,
"learning_rate": 1.8140845070422533e-07,
"loss": 0.378,
"step": 3280
},
{
"epoch": 45.69,
"learning_rate": 1.7577464788732394e-07,
"loss": 0.3613,
"step": 3290
},
{
"epoch": 45.83,
"learning_rate": 1.7014084507042254e-07,
"loss": 0.3659,
"step": 3300
},
{
"epoch": 45.97,
"learning_rate": 1.6450704225352112e-07,
"loss": 0.3693,
"step": 3310
},
{
"epoch": 46.0,
"eval_loss": 0.5401335954666138,
"eval_runtime": 127.2873,
"eval_samples_per_second": 4.18,
"eval_steps_per_second": 0.526,
"eval_wer": 0.2471556212637398,
"step": 3312
},
{
"epoch": 46.11,
"learning_rate": 1.588732394366197e-07,
"loss": 0.3782,
"step": 3320
},
{
"epoch": 46.25,
"learning_rate": 1.532394366197183e-07,
"loss": 0.3685,
"step": 3330
},
{
"epoch": 46.39,
"learning_rate": 1.476056338028169e-07,
"loss": 0.3586,
"step": 3340
},
{
"epoch": 46.53,
"learning_rate": 1.4197183098591547e-07,
"loss": 0.3648,
"step": 3350
},
{
"epoch": 46.66,
"learning_rate": 1.3633802816901408e-07,
"loss": 0.3639,
"step": 3360
},
{
"epoch": 46.8,
"learning_rate": 1.3070422535211268e-07,
"loss": 0.367,
"step": 3370
},
{
"epoch": 46.94,
"learning_rate": 1.2507042253521128e-07,
"loss": 0.3593,
"step": 3380
},
{
"epoch": 47.0,
"eval_loss": 0.5389713048934937,
"eval_runtime": 127.3901,
"eval_samples_per_second": 4.176,
"eval_steps_per_second": 0.526,
"eval_wer": 0.24625570482740888,
"step": 3384
},
{
"epoch": 47.08,
"learning_rate": 1.1943661971830986e-07,
"loss": 0.3798,
"step": 3390
},
{
"epoch": 47.22,
"learning_rate": 1.1380281690140845e-07,
"loss": 0.3644,
"step": 3400
},
{
"epoch": 47.36,
"learning_rate": 1.0816901408450704e-07,
"loss": 0.3545,
"step": 3410
},
{
"epoch": 47.5,
"learning_rate": 1.0253521126760563e-07,
"loss": 0.3633,
"step": 3420
},
{
"epoch": 47.64,
"learning_rate": 9.690140845070423e-08,
"loss": 0.3532,
"step": 3430
},
{
"epoch": 47.78,
"learning_rate": 9.126760563380281e-08,
"loss": 0.3705,
"step": 3440
},
{
"epoch": 47.91,
"learning_rate": 8.56338028169014e-08,
"loss": 0.3515,
"step": 3450
},
{
"epoch": 48.0,
"eval_loss": 0.5398772954940796,
"eval_runtime": 129.9986,
"eval_samples_per_second": 4.092,
"eval_steps_per_second": 0.515,
"eval_wer": 0.24676994279102654,
"step": 3456
},
{
"epoch": 48.06,
"learning_rate": 8e-08,
"loss": 0.3824,
"step": 3460
},
{
"epoch": 48.19,
"learning_rate": 7.436619718309858e-08,
"loss": 0.3749,
"step": 3470
},
{
"epoch": 48.33,
"learning_rate": 6.873239436619718e-08,
"loss": 0.3759,
"step": 3480
},
{
"epoch": 48.47,
"learning_rate": 6.309859154929577e-08,
"loss": 0.354,
"step": 3490
},
{
"epoch": 48.61,
"learning_rate": 5.746478873239436e-08,
"loss": 0.3629,
"step": 3500
},
{
"epoch": 48.75,
"learning_rate": 5.183098591549295e-08,
"loss": 0.3605,
"step": 3510
},
{
"epoch": 48.89,
"learning_rate": 4.619718309859155e-08,
"loss": 0.3485,
"step": 3520
},
{
"epoch": 49.0,
"eval_loss": 0.5391861796379089,
"eval_runtime": 127.4223,
"eval_samples_per_second": 4.175,
"eval_steps_per_second": 0.526,
"eval_wer": 0.24612714533650445,
"step": 3528
},
{
"epoch": 49.03,
"learning_rate": 4.0563380281690137e-08,
"loss": 0.3749,
"step": 3530
},
{
"epoch": 49.17,
"learning_rate": 3.492957746478873e-08,
"loss": 0.3625,
"step": 3540
},
{
"epoch": 49.3,
"learning_rate": 2.9295774647887323e-08,
"loss": 0.3768,
"step": 3550
},
{
"epoch": 49.44,
"learning_rate": 2.3661971830985916e-08,
"loss": 0.3523,
"step": 3560
},
{
"epoch": 49.58,
"learning_rate": 1.802816901408451e-08,
"loss": 0.3673,
"step": 3570
},
{
"epoch": 49.72,
"learning_rate": 1.2394366197183099e-08,
"loss": 0.3624,
"step": 3580
},
{
"epoch": 49.86,
"learning_rate": 6.76056338028169e-09,
"loss": 0.3632,
"step": 3590
},
{
"epoch": 50.0,
"learning_rate": 1.1267605633802818e-09,
"loss": 0.3591,
"step": 3600
},
{
"epoch": 50.0,
"eval_loss": 0.5403211116790771,
"eval_runtime": 128.4757,
"eval_samples_per_second": 4.141,
"eval_steps_per_second": 0.521,
"eval_wer": 0.24606286559105225,
"step": 3600
},
{
"epoch": 50.0,
"step": 3600,
"total_flos": 1.3601359172236221e+20,
"train_loss": 0.3682790127065447,
"train_runtime": 95293.0206,
"train_samples_per_second": 2.423,
"train_steps_per_second": 0.038
}
],
"max_steps": 3600,
"num_train_epochs": 50,
"total_flos": 1.3601359172236221e+20,
"trial_name": null,
"trial_params": null
}