gary109's picture
End of training
82b5115
{
"best_metric": 0.4629605710506439,
"best_model_checkpoint": "ai-light-dance_drums_ft_pretrain_wav2vec2-base-new-13k_onset-drums_fold_1/checkpoint-1173",
"epoch": 49.98924731182796,
"global_step": 3450,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.14,
"learning_rate": 2.6999999999999996e-05,
"loss": 14.4614,
"step": 10
},
{
"epoch": 0.29,
"learning_rate": 4.4999999999999996e-05,
"loss": 39.9587,
"step": 20
},
{
"epoch": 0.43,
"learning_rate": 7.5e-05,
"loss": 24.4939,
"step": 30
},
{
"epoch": 0.57,
"learning_rate": 0.000102,
"loss": 23.3905,
"step": 40
},
{
"epoch": 0.72,
"learning_rate": 0.00013199999999999998,
"loss": 18.0763,
"step": 50
},
{
"epoch": 0.86,
"learning_rate": 0.000162,
"loss": 3.0614,
"step": 60
},
{
"epoch": 0.99,
"eval_loss": 5.127472400665283,
"eval_runtime": 61.8255,
"eval_samples_per_second": 9.09,
"eval_steps_per_second": 2.281,
"eval_wer": 1.0,
"step": 69
},
{
"epoch": 1.01,
"learning_rate": 0.00019199999999999998,
"loss": 3.3451,
"step": 70
},
{
"epoch": 1.16,
"learning_rate": 0.00022199999999999998,
"loss": 2.1389,
"step": 80
},
{
"epoch": 1.3,
"learning_rate": 0.00025199999999999995,
"loss": 3.8889,
"step": 90
},
{
"epoch": 1.44,
"learning_rate": 0.00028199999999999997,
"loss": 4.1626,
"step": 100
},
{
"epoch": 1.59,
"learning_rate": 0.0002996417910447761,
"loss": 2.8016,
"step": 110
},
{
"epoch": 1.73,
"learning_rate": 0.0002987462686567164,
"loss": 2.5118,
"step": 120
},
{
"epoch": 1.87,
"learning_rate": 0.0002978507462686567,
"loss": 1.8291,
"step": 130
},
{
"epoch": 1.99,
"eval_loss": 2.2008168697357178,
"eval_runtime": 61.2052,
"eval_samples_per_second": 9.182,
"eval_steps_per_second": 2.304,
"eval_wer": 1.0,
"step": 138
},
{
"epoch": 2.03,
"learning_rate": 0.000296955223880597,
"loss": 2.1219,
"step": 140
},
{
"epoch": 2.17,
"learning_rate": 0.0002960597014925373,
"loss": 1.568,
"step": 150
},
{
"epoch": 2.32,
"learning_rate": 0.0002951641791044776,
"loss": 2.1417,
"step": 160
},
{
"epoch": 2.46,
"learning_rate": 0.0002942686567164179,
"loss": 2.1267,
"step": 170
},
{
"epoch": 2.6,
"learning_rate": 0.0002933731343283582,
"loss": 1.5519,
"step": 180
},
{
"epoch": 2.75,
"learning_rate": 0.0002924776119402985,
"loss": 1.6312,
"step": 190
},
{
"epoch": 2.89,
"learning_rate": 0.00029158208955223875,
"loss": 1.4664,
"step": 200
},
{
"epoch": 2.99,
"eval_loss": 1.682111144065857,
"eval_runtime": 60.6361,
"eval_samples_per_second": 9.268,
"eval_steps_per_second": 2.325,
"eval_wer": 1.0,
"step": 207
},
{
"epoch": 3.04,
"learning_rate": 0.0002906865671641791,
"loss": 1.6895,
"step": 210
},
{
"epoch": 3.19,
"learning_rate": 0.0002897910447761194,
"loss": 1.3638,
"step": 220
},
{
"epoch": 3.33,
"learning_rate": 0.0002888955223880597,
"loss": 1.5225,
"step": 230
},
{
"epoch": 3.47,
"learning_rate": 0.00028799999999999995,
"loss": 1.459,
"step": 240
},
{
"epoch": 3.62,
"learning_rate": 0.0002871044776119403,
"loss": 1.3825,
"step": 250
},
{
"epoch": 3.76,
"learning_rate": 0.00028620895522388055,
"loss": 1.4428,
"step": 260
},
{
"epoch": 3.9,
"learning_rate": 0.0002853134328358209,
"loss": 1.287,
"step": 270
},
{
"epoch": 3.99,
"eval_loss": 1.5681122541427612,
"eval_runtime": 62.0132,
"eval_samples_per_second": 9.063,
"eval_steps_per_second": 2.274,
"eval_wer": 1.0,
"step": 276
},
{
"epoch": 4.06,
"learning_rate": 0.00028441791044776115,
"loss": 1.5539,
"step": 280
},
{
"epoch": 4.2,
"learning_rate": 0.0002835223880597015,
"loss": 1.3065,
"step": 290
},
{
"epoch": 4.34,
"learning_rate": 0.00028262686567164175,
"loss": 1.3586,
"step": 300
},
{
"epoch": 4.49,
"learning_rate": 0.0002817313432835821,
"loss": 1.396,
"step": 310
},
{
"epoch": 4.63,
"learning_rate": 0.0002808358208955224,
"loss": 1.2333,
"step": 320
},
{
"epoch": 4.77,
"learning_rate": 0.0002799402985074627,
"loss": 1.4229,
"step": 330
},
{
"epoch": 4.92,
"learning_rate": 0.00027904477611940295,
"loss": 1.2642,
"step": 340
},
{
"epoch": 4.99,
"eval_loss": 1.507431983947754,
"eval_runtime": 61.6864,
"eval_samples_per_second": 9.111,
"eval_steps_per_second": 2.286,
"eval_wer": 1.0,
"step": 345
},
{
"epoch": 5.07,
"learning_rate": 0.0002781492537313432,
"loss": 1.4889,
"step": 350
},
{
"epoch": 5.22,
"learning_rate": 0.00027725373134328355,
"loss": 1.2919,
"step": 360
},
{
"epoch": 5.36,
"learning_rate": 0.0002763582089552239,
"loss": 1.2431,
"step": 370
},
{
"epoch": 5.5,
"learning_rate": 0.00027546268656716415,
"loss": 1.3312,
"step": 380
},
{
"epoch": 5.65,
"learning_rate": 0.0002745671641791044,
"loss": 1.1704,
"step": 390
},
{
"epoch": 5.79,
"learning_rate": 0.00027367164179104475,
"loss": 1.3583,
"step": 400
},
{
"epoch": 5.93,
"learning_rate": 0.000272776119402985,
"loss": 1.2702,
"step": 410
},
{
"epoch": 5.99,
"eval_loss": 1.4650160074234009,
"eval_runtime": 60.8401,
"eval_samples_per_second": 9.237,
"eval_steps_per_second": 2.318,
"eval_wer": 1.0,
"step": 414
},
{
"epoch": 6.09,
"learning_rate": 0.00027188059701492535,
"loss": 1.4877,
"step": 420
},
{
"epoch": 6.23,
"learning_rate": 0.0002709850746268657,
"loss": 1.3383,
"step": 430
},
{
"epoch": 6.37,
"learning_rate": 0.00027008955223880595,
"loss": 1.2241,
"step": 440
},
{
"epoch": 6.52,
"learning_rate": 0.0002691940298507462,
"loss": 1.4834,
"step": 450
},
{
"epoch": 6.66,
"learning_rate": 0.00026829850746268655,
"loss": 1.3162,
"step": 460
},
{
"epoch": 6.8,
"learning_rate": 0.0002674029850746269,
"loss": 1.3369,
"step": 470
},
{
"epoch": 6.95,
"learning_rate": 0.00026650746268656715,
"loss": 1.2245,
"step": 480
},
{
"epoch": 6.99,
"eval_loss": 1.3027324676513672,
"eval_runtime": 61.9807,
"eval_samples_per_second": 9.067,
"eval_steps_per_second": 2.275,
"eval_wer": 1.0,
"step": 483
},
{
"epoch": 7.1,
"learning_rate": 0.0002656119402985074,
"loss": 1.3505,
"step": 490
},
{
"epoch": 7.24,
"learning_rate": 0.00026471641791044775,
"loss": 1.2848,
"step": 500
},
{
"epoch": 7.39,
"learning_rate": 0.000263820895522388,
"loss": 1.2029,
"step": 510
},
{
"epoch": 7.53,
"learning_rate": 0.00026292537313432835,
"loss": 1.2746,
"step": 520
},
{
"epoch": 7.67,
"learning_rate": 0.0002620298507462686,
"loss": 1.1274,
"step": 530
},
{
"epoch": 7.82,
"learning_rate": 0.00026113432835820895,
"loss": 1.2906,
"step": 540
},
{
"epoch": 7.96,
"learning_rate": 0.0002602388059701492,
"loss": 1.3461,
"step": 550
},
{
"epoch": 7.99,
"eval_loss": 1.3108575344085693,
"eval_runtime": 61.0886,
"eval_samples_per_second": 9.2,
"eval_steps_per_second": 2.308,
"eval_wer": 1.0,
"step": 552
},
{
"epoch": 8.11,
"learning_rate": 0.00025934328358208955,
"loss": 1.2607,
"step": 560
},
{
"epoch": 8.26,
"learning_rate": 0.0002584477611940298,
"loss": 1.3096,
"step": 570
},
{
"epoch": 8.4,
"learning_rate": 0.00025755223880597015,
"loss": 1.093,
"step": 580
},
{
"epoch": 8.54,
"learning_rate": 0.0002566567164179104,
"loss": 1.3239,
"step": 590
},
{
"epoch": 8.69,
"learning_rate": 0.0002557611940298507,
"loss": 1.1256,
"step": 600
},
{
"epoch": 8.83,
"learning_rate": 0.000254865671641791,
"loss": 1.2056,
"step": 610
},
{
"epoch": 8.97,
"learning_rate": 0.00025397014925373135,
"loss": 1.2903,
"step": 620
},
{
"epoch": 8.99,
"eval_loss": 1.3107479810714722,
"eval_runtime": 61.7504,
"eval_samples_per_second": 9.101,
"eval_steps_per_second": 2.283,
"eval_wer": 1.0,
"step": 621
},
{
"epoch": 9.13,
"learning_rate": 0.0002530746268656716,
"loss": 1.1892,
"step": 630
},
{
"epoch": 9.27,
"learning_rate": 0.0002521791044776119,
"loss": 1.2287,
"step": 640
},
{
"epoch": 9.42,
"learning_rate": 0.0002512835820895522,
"loss": 1.0642,
"step": 650
},
{
"epoch": 9.56,
"learning_rate": 0.0002503880597014925,
"loss": 1.2682,
"step": 660
},
{
"epoch": 9.7,
"learning_rate": 0.0002494925373134328,
"loss": 1.1652,
"step": 670
},
{
"epoch": 9.85,
"learning_rate": 0.00024859701492537314,
"loss": 1.1737,
"step": 680
},
{
"epoch": 9.99,
"learning_rate": 0.0002477014925373134,
"loss": 1.2741,
"step": 690
},
{
"epoch": 9.99,
"eval_loss": 1.1842212677001953,
"eval_runtime": 61.8553,
"eval_samples_per_second": 9.086,
"eval_steps_per_second": 2.28,
"eval_wer": 1.0,
"step": 690
},
{
"epoch": 10.14,
"learning_rate": 0.0002468059701492537,
"loss": 1.1723,
"step": 700
},
{
"epoch": 10.29,
"learning_rate": 0.000245910447761194,
"loss": 1.2786,
"step": 710
},
{
"epoch": 10.43,
"learning_rate": 0.00024501492537313434,
"loss": 1.0985,
"step": 720
},
{
"epoch": 10.57,
"learning_rate": 0.00024411940298507462,
"loss": 1.1873,
"step": 730
},
{
"epoch": 10.72,
"learning_rate": 0.0002432238805970149,
"loss": 1.1893,
"step": 740
},
{
"epoch": 10.86,
"learning_rate": 0.0002423283582089552,
"loss": 1.1446,
"step": 750
},
{
"epoch": 10.99,
"eval_loss": 1.1754302978515625,
"eval_runtime": 61.1816,
"eval_samples_per_second": 9.186,
"eval_steps_per_second": 2.305,
"eval_wer": 1.0,
"step": 759
},
{
"epoch": 11.01,
"learning_rate": 0.00024143283582089552,
"loss": 1.3346,
"step": 760
},
{
"epoch": 11.16,
"learning_rate": 0.0002405373134328358,
"loss": 1.029,
"step": 770
},
{
"epoch": 11.3,
"learning_rate": 0.0002396417910447761,
"loss": 1.2201,
"step": 780
},
{
"epoch": 11.44,
"learning_rate": 0.00023874626865671641,
"loss": 1.1011,
"step": 790
},
{
"epoch": 11.59,
"learning_rate": 0.0002378507462686567,
"loss": 1.1429,
"step": 800
},
{
"epoch": 11.73,
"learning_rate": 0.000236955223880597,
"loss": 1.2601,
"step": 810
},
{
"epoch": 11.87,
"learning_rate": 0.0002360597014925373,
"loss": 1.0746,
"step": 820
},
{
"epoch": 11.99,
"eval_loss": 1.1468886137008667,
"eval_runtime": 63.7687,
"eval_samples_per_second": 8.813,
"eval_steps_per_second": 2.211,
"eval_wer": 0.9999149515223678,
"step": 828
},
{
"epoch": 12.03,
"learning_rate": 0.00023516417910447761,
"loss": 1.3232,
"step": 830
},
{
"epoch": 12.17,
"learning_rate": 0.0002342686567164179,
"loss": 1.0437,
"step": 840
},
{
"epoch": 12.32,
"learning_rate": 0.0002333731343283582,
"loss": 1.132,
"step": 850
},
{
"epoch": 12.46,
"learning_rate": 0.00023247761194029851,
"loss": 1.0021,
"step": 860
},
{
"epoch": 12.6,
"learning_rate": 0.00023158208955223879,
"loss": 0.9904,
"step": 870
},
{
"epoch": 12.75,
"learning_rate": 0.00023068656716417909,
"loss": 1.1869,
"step": 880
},
{
"epoch": 12.89,
"learning_rate": 0.00022979104477611936,
"loss": 0.8203,
"step": 890
},
{
"epoch": 12.99,
"eval_loss": 0.907135546207428,
"eval_runtime": 61.3963,
"eval_samples_per_second": 9.154,
"eval_steps_per_second": 2.297,
"eval_wer": 0.6201734988943698,
"step": 897
},
{
"epoch": 13.04,
"learning_rate": 0.00022889552238805969,
"loss": 1.2374,
"step": 900
},
{
"epoch": 13.19,
"learning_rate": 0.00022799999999999999,
"loss": 0.791,
"step": 910
},
{
"epoch": 13.33,
"learning_rate": 0.00022710447761194029,
"loss": 0.9579,
"step": 920
},
{
"epoch": 13.47,
"learning_rate": 0.00022620895522388056,
"loss": 0.9683,
"step": 930
},
{
"epoch": 13.62,
"learning_rate": 0.00022531343283582088,
"loss": 0.6633,
"step": 940
},
{
"epoch": 13.76,
"learning_rate": 0.00022441791044776118,
"loss": 1.0536,
"step": 950
},
{
"epoch": 13.9,
"learning_rate": 0.00022352238805970146,
"loss": 0.5996,
"step": 960
},
{
"epoch": 13.99,
"eval_loss": 0.7047128677368164,
"eval_runtime": 61.3196,
"eval_samples_per_second": 9.165,
"eval_steps_per_second": 2.299,
"eval_wer": 0.4233713216533424,
"step": 966
},
{
"epoch": 14.06,
"learning_rate": 0.00022262686567164178,
"loss": 0.9754,
"step": 970
},
{
"epoch": 14.2,
"learning_rate": 0.00022173134328358208,
"loss": 0.6592,
"step": 980
},
{
"epoch": 14.34,
"learning_rate": 0.00022083582089552236,
"loss": 0.663,
"step": 990
},
{
"epoch": 14.49,
"learning_rate": 0.00021994029850746266,
"loss": 0.8436,
"step": 1000
},
{
"epoch": 14.63,
"learning_rate": 0.00021904477611940298,
"loss": 0.5158,
"step": 1010
},
{
"epoch": 14.77,
"learning_rate": 0.00021814925373134326,
"loss": 0.9818,
"step": 1020
},
{
"epoch": 14.92,
"learning_rate": 0.00021725373134328356,
"loss": 0.5672,
"step": 1030
},
{
"epoch": 14.99,
"eval_loss": 0.5368545055389404,
"eval_runtime": 64.4409,
"eval_samples_per_second": 8.721,
"eval_steps_per_second": 2.188,
"eval_wer": 0.2566763054941317,
"step": 1035
},
{
"epoch": 15.07,
"learning_rate": 0.00021635820895522388,
"loss": 0.8306,
"step": 1040
},
{
"epoch": 15.22,
"learning_rate": 0.00021546268656716416,
"loss": 0.6282,
"step": 1050
},
{
"epoch": 15.36,
"learning_rate": 0.00021456716417910446,
"loss": 0.5391,
"step": 1060
},
{
"epoch": 15.5,
"learning_rate": 0.00021367164179104476,
"loss": 0.8652,
"step": 1070
},
{
"epoch": 15.65,
"learning_rate": 0.00021277611940298505,
"loss": 0.4055,
"step": 1080
},
{
"epoch": 15.79,
"learning_rate": 0.00021188059701492535,
"loss": 0.8259,
"step": 1090
},
{
"epoch": 15.93,
"learning_rate": 0.00021098507462686565,
"loss": 0.4965,
"step": 1100
},
{
"epoch": 15.99,
"eval_loss": 0.4643840193748474,
"eval_runtime": 64.7572,
"eval_samples_per_second": 8.679,
"eval_steps_per_second": 2.177,
"eval_wer": 0.2861456029937064,
"step": 1104
},
{
"epoch": 16.09,
"learning_rate": 0.00021008955223880593,
"loss": 0.6985,
"step": 1110
},
{
"epoch": 16.23,
"learning_rate": 0.00020919402985074625,
"loss": 0.7038,
"step": 1120
},
{
"epoch": 16.37,
"learning_rate": 0.00020829850746268655,
"loss": 0.4604,
"step": 1130
},
{
"epoch": 16.52,
"learning_rate": 0.00020740298507462683,
"loss": 0.7335,
"step": 1140
},
{
"epoch": 16.66,
"learning_rate": 0.00020650746268656715,
"loss": 0.3868,
"step": 1150
},
{
"epoch": 16.8,
"learning_rate": 0.00020561194029850745,
"loss": 0.8036,
"step": 1160
},
{
"epoch": 16.95,
"learning_rate": 0.00020471641791044773,
"loss": 0.5639,
"step": 1170
},
{
"epoch": 16.99,
"eval_loss": 0.4629605710506439,
"eval_runtime": 63.6923,
"eval_samples_per_second": 8.824,
"eval_steps_per_second": 2.214,
"eval_wer": 0.21453478482735158,
"step": 1173
},
{
"epoch": 17.1,
"learning_rate": 0.00020382089552238803,
"loss": 0.5344,
"step": 1180
},
{
"epoch": 17.24,
"learning_rate": 0.00020292537313432835,
"loss": 0.7905,
"step": 1190
},
{
"epoch": 17.39,
"learning_rate": 0.00020202985074626865,
"loss": 0.2889,
"step": 1200
},
{
"epoch": 17.53,
"learning_rate": 0.00020113432835820893,
"loss": 0.7319,
"step": 1210
},
{
"epoch": 17.67,
"learning_rate": 0.00020023880597014925,
"loss": 0.4832,
"step": 1220
},
{
"epoch": 17.82,
"learning_rate": 0.00019934328358208955,
"loss": 0.6016,
"step": 1230
},
{
"epoch": 17.96,
"learning_rate": 0.00019844776119402982,
"loss": 0.6272,
"step": 1240
},
{
"epoch": 17.99,
"eval_loss": 0.6847624778747559,
"eval_runtime": 62.179,
"eval_samples_per_second": 9.038,
"eval_steps_per_second": 2.268,
"eval_wer": 0.26666950161592107,
"step": 1242
},
{
"epoch": 18.11,
"learning_rate": 0.00019755223880597012,
"loss": 0.4069,
"step": 1250
},
{
"epoch": 18.26,
"learning_rate": 0.00019665671641791045,
"loss": 0.6994,
"step": 1260
},
{
"epoch": 18.4,
"learning_rate": 0.00019576119402985072,
"loss": 0.291,
"step": 1270
},
{
"epoch": 18.54,
"learning_rate": 0.00019486567164179102,
"loss": 0.6654,
"step": 1280
},
{
"epoch": 18.69,
"learning_rate": 0.00019397014925373132,
"loss": 0.4326,
"step": 1290
},
{
"epoch": 18.83,
"learning_rate": 0.00019307462686567162,
"loss": 0.5143,
"step": 1300
},
{
"epoch": 18.97,
"learning_rate": 0.00019217910447761192,
"loss": 0.6764,
"step": 1310
},
{
"epoch": 18.99,
"eval_loss": 0.6074081659317017,
"eval_runtime": 63.0738,
"eval_samples_per_second": 8.91,
"eval_steps_per_second": 2.235,
"eval_wer": 0.25076543629869025,
"step": 1311
},
{
"epoch": 19.13,
"learning_rate": 0.00019128358208955222,
"loss": 0.3389,
"step": 1320
},
{
"epoch": 19.27,
"learning_rate": 0.00019038805970149252,
"loss": 0.6627,
"step": 1330
},
{
"epoch": 19.42,
"learning_rate": 0.00018949253731343282,
"loss": 0.3527,
"step": 1340
},
{
"epoch": 19.56,
"learning_rate": 0.00018859701492537312,
"loss": 0.7161,
"step": 1350
},
{
"epoch": 19.7,
"learning_rate": 0.0001877014925373134,
"loss": 0.4466,
"step": 1360
},
{
"epoch": 19.85,
"learning_rate": 0.00018680597014925372,
"loss": 0.3684,
"step": 1370
},
{
"epoch": 19.99,
"learning_rate": 0.00018591044776119402,
"loss": 0.7205,
"step": 1380
},
{
"epoch": 19.99,
"eval_loss": 0.6451706886291504,
"eval_runtime": 61.4879,
"eval_samples_per_second": 9.14,
"eval_steps_per_second": 2.293,
"eval_wer": 0.21844701479843512,
"step": 1380
},
{
"epoch": 20.14,
"learning_rate": 0.0001850149253731343,
"loss": 0.295,
"step": 1390
},
{
"epoch": 20.29,
"learning_rate": 0.00018411940298507462,
"loss": 0.6178,
"step": 1400
},
{
"epoch": 20.43,
"learning_rate": 0.00018322388059701492,
"loss": 0.349,
"step": 1410
},
{
"epoch": 20.57,
"learning_rate": 0.0001823283582089552,
"loss": 0.5619,
"step": 1420
},
{
"epoch": 20.72,
"learning_rate": 0.0001814328358208955,
"loss": 0.4867,
"step": 1430
},
{
"epoch": 20.86,
"learning_rate": 0.00018053731343283582,
"loss": 0.346,
"step": 1440
},
{
"epoch": 20.99,
"eval_loss": 0.596199095249176,
"eval_runtime": 62.9086,
"eval_samples_per_second": 8.934,
"eval_steps_per_second": 2.241,
"eval_wer": 0.2457475761183875,
"step": 1449
},
{
"epoch": 21.01,
"learning_rate": 0.0001796417910447761,
"loss": 0.7285,
"step": 1450
},
{
"epoch": 21.16,
"learning_rate": 0.0001787462686567164,
"loss": 0.2096,
"step": 1460
},
{
"epoch": 21.3,
"learning_rate": 0.0001778507462686567,
"loss": 0.6799,
"step": 1470
},
{
"epoch": 21.44,
"learning_rate": 0.00017695522388059702,
"loss": 0.3988,
"step": 1480
},
{
"epoch": 21.59,
"learning_rate": 0.0001760597014925373,
"loss": 0.4867,
"step": 1490
},
{
"epoch": 21.73,
"learning_rate": 0.0001751641791044776,
"loss": 0.5862,
"step": 1500
},
{
"epoch": 21.87,
"learning_rate": 0.00017426865671641792,
"loss": 0.2212,
"step": 1510
},
{
"epoch": 21.99,
"eval_loss": 0.523625373840332,
"eval_runtime": 62.2688,
"eval_samples_per_second": 9.025,
"eval_steps_per_second": 2.264,
"eval_wer": 0.20683789760163293,
"step": 1518
},
{
"epoch": 22.03,
"learning_rate": 0.0001733731343283582,
"loss": 0.6216,
"step": 1520
},
{
"epoch": 22.17,
"learning_rate": 0.0001724776119402985,
"loss": 0.2327,
"step": 1530
},
{
"epoch": 22.32,
"learning_rate": 0.00017158208955223876,
"loss": 0.5753,
"step": 1540
},
{
"epoch": 22.46,
"learning_rate": 0.0001706865671641791,
"loss": 0.4677,
"step": 1550
},
{
"epoch": 22.6,
"learning_rate": 0.0001697910447761194,
"loss": 0.4016,
"step": 1560
},
{
"epoch": 22.75,
"learning_rate": 0.0001688955223880597,
"loss": 0.6864,
"step": 1570
},
{
"epoch": 22.89,
"learning_rate": 0.000168,
"loss": 0.1646,
"step": 1580
},
{
"epoch": 22.99,
"eval_loss": 0.6130056977272034,
"eval_runtime": 62.039,
"eval_samples_per_second": 9.059,
"eval_steps_per_second": 2.273,
"eval_wer": 0.21976526620173498,
"step": 1587
},
{
"epoch": 23.04,
"learning_rate": 0.0001671044776119403,
"loss": 0.6475,
"step": 1590
},
{
"epoch": 23.19,
"learning_rate": 0.0001662089552238806,
"loss": 0.2693,
"step": 1600
},
{
"epoch": 23.33,
"learning_rate": 0.00016531343283582086,
"loss": 0.3855,
"step": 1610
},
{
"epoch": 23.47,
"learning_rate": 0.0001644179104477612,
"loss": 0.5039,
"step": 1620
},
{
"epoch": 23.62,
"learning_rate": 0.0001635223880597015,
"loss": 0.2613,
"step": 1630
},
{
"epoch": 23.76,
"learning_rate": 0.00016262686567164176,
"loss": 0.637,
"step": 1640
},
{
"epoch": 23.9,
"learning_rate": 0.00016173134328358206,
"loss": 0.3148,
"step": 1650
},
{
"epoch": 23.99,
"eval_loss": 0.5591665506362915,
"eval_runtime": 62.379,
"eval_samples_per_second": 9.009,
"eval_steps_per_second": 2.26,
"eval_wer": 0.2620343595849634,
"step": 1656
},
{
"epoch": 24.06,
"learning_rate": 0.0001608358208955224,
"loss": 0.5373,
"step": 1660
},
{
"epoch": 24.2,
"learning_rate": 0.00015994029850746266,
"loss": 0.3759,
"step": 1670
},
{
"epoch": 24.34,
"learning_rate": 0.00015904477611940296,
"loss": 0.466,
"step": 1680
},
{
"epoch": 24.49,
"learning_rate": 0.0001581492537313433,
"loss": 0.5259,
"step": 1690
},
{
"epoch": 24.63,
"learning_rate": 0.00015725373134328356,
"loss": 0.174,
"step": 1700
},
{
"epoch": 24.77,
"learning_rate": 0.00015635820895522386,
"loss": 0.5319,
"step": 1710
},
{
"epoch": 24.92,
"learning_rate": 0.00015546268656716416,
"loss": 0.3061,
"step": 1720
},
{
"epoch": 24.99,
"eval_loss": 0.5576915740966797,
"eval_runtime": 61.4469,
"eval_samples_per_second": 9.146,
"eval_steps_per_second": 2.295,
"eval_wer": 0.2560384419118898,
"step": 1725
},
{
"epoch": 25.07,
"learning_rate": 0.0001545671641791045,
"loss": 0.5376,
"step": 1730
},
{
"epoch": 25.22,
"learning_rate": 0.00015367164179104476,
"loss": 0.4242,
"step": 1740
},
{
"epoch": 25.36,
"learning_rate": 0.00015277611940298506,
"loss": 0.3064,
"step": 1750
},
{
"epoch": 25.5,
"learning_rate": 0.0001518805970149254,
"loss": 0.5369,
"step": 1760
},
{
"epoch": 25.65,
"learning_rate": 0.00015098507462686566,
"loss": 0.2128,
"step": 1770
},
{
"epoch": 25.79,
"learning_rate": 0.00015008955223880596,
"loss": 0.5043,
"step": 1780
},
{
"epoch": 25.93,
"learning_rate": 0.00014919402985074626,
"loss": 0.3137,
"step": 1790
},
{
"epoch": 25.99,
"eval_loss": 0.5247139930725098,
"eval_runtime": 61.5961,
"eval_samples_per_second": 9.124,
"eval_steps_per_second": 2.289,
"eval_wer": 0.2226569144412315,
"step": 1794
},
{
"epoch": 26.09,
"learning_rate": 0.00014829850746268656,
"loss": 0.4782,
"step": 1800
},
{
"epoch": 26.23,
"learning_rate": 0.00014740298507462686,
"loss": 0.5075,
"step": 1810
},
{
"epoch": 26.37,
"learning_rate": 0.00014650746268656716,
"loss": 0.3237,
"step": 1820
},
{
"epoch": 26.52,
"learning_rate": 0.00014561194029850746,
"loss": 0.5464,
"step": 1830
},
{
"epoch": 26.66,
"learning_rate": 0.00014471641791044773,
"loss": 0.2409,
"step": 1840
},
{
"epoch": 26.8,
"learning_rate": 0.00014382089552238806,
"loss": 0.4431,
"step": 1850
},
{
"epoch": 26.95,
"learning_rate": 0.00014292537313432836,
"loss": 0.389,
"step": 1860
},
{
"epoch": 26.99,
"eval_loss": 0.5799093246459961,
"eval_runtime": 62.2481,
"eval_samples_per_second": 9.028,
"eval_steps_per_second": 2.265,
"eval_wer": 0.2081136247661167,
"step": 1863
},
{
"epoch": 27.1,
"learning_rate": 0.00014202985074626863,
"loss": 0.3481,
"step": 1870
},
{
"epoch": 27.24,
"learning_rate": 0.00014113432835820896,
"loss": 0.6519,
"step": 1880
},
{
"epoch": 27.39,
"learning_rate": 0.00014023880597014923,
"loss": 0.1449,
"step": 1890
},
{
"epoch": 27.53,
"learning_rate": 0.00013934328358208953,
"loss": 0.5128,
"step": 1900
},
{
"epoch": 27.67,
"learning_rate": 0.00013844776119402983,
"loss": 0.2642,
"step": 1910
},
{
"epoch": 27.82,
"learning_rate": 0.00013755223880597013,
"loss": 0.3895,
"step": 1920
},
{
"epoch": 27.96,
"learning_rate": 0.00013665671641791043,
"loss": 0.4168,
"step": 1930
},
{
"epoch": 27.99,
"eval_loss": 0.5849633812904358,
"eval_runtime": 62.1221,
"eval_samples_per_second": 9.047,
"eval_steps_per_second": 2.27,
"eval_wer": 0.18183364517775133,
"step": 1932
},
{
"epoch": 28.11,
"learning_rate": 0.00013576119402985073,
"loss": 0.3046,
"step": 1940
},
{
"epoch": 28.26,
"learning_rate": 0.00013486567164179103,
"loss": 0.5701,
"step": 1950
},
{
"epoch": 28.4,
"learning_rate": 0.00013397014925373133,
"loss": 0.1974,
"step": 1960
},
{
"epoch": 28.54,
"learning_rate": 0.00013307462686567163,
"loss": 0.5276,
"step": 1970
},
{
"epoch": 28.69,
"learning_rate": 0.00013217910447761193,
"loss": 0.2903,
"step": 1980
},
{
"epoch": 28.83,
"learning_rate": 0.00013128358208955223,
"loss": 0.4352,
"step": 1990
},
{
"epoch": 28.97,
"learning_rate": 0.00013038805970149253,
"loss": 0.4403,
"step": 2000
},
{
"epoch": 28.99,
"eval_loss": 0.5686776638031006,
"eval_runtime": 61.501,
"eval_samples_per_second": 9.138,
"eval_steps_per_second": 2.293,
"eval_wer": 0.20530702500425244,
"step": 2001
},
{
"epoch": 29.13,
"learning_rate": 0.00012949253731343283,
"loss": 0.2081,
"step": 2010
},
{
"epoch": 29.27,
"learning_rate": 0.00012859701492537313,
"loss": 0.5283,
"step": 2020
},
{
"epoch": 29.42,
"learning_rate": 0.00012770149253731343,
"loss": 0.1902,
"step": 2030
},
{
"epoch": 29.56,
"learning_rate": 0.00012680597014925373,
"loss": 0.3889,
"step": 2040
},
{
"epoch": 29.7,
"learning_rate": 0.00012591044776119403,
"loss": 0.3573,
"step": 2050
},
{
"epoch": 29.85,
"learning_rate": 0.00012501492537313433,
"loss": 0.3416,
"step": 2060
},
{
"epoch": 29.99,
"learning_rate": 0.0001241194029850746,
"loss": 0.4936,
"step": 2070
},
{
"epoch": 29.99,
"eval_loss": 0.5511252880096436,
"eval_runtime": 62.8347,
"eval_samples_per_second": 8.944,
"eval_steps_per_second": 2.244,
"eval_wer": 0.20649770369110393,
"step": 2070
},
{
"epoch": 30.14,
"learning_rate": 0.00012322388059701493,
"loss": 0.2105,
"step": 2080
},
{
"epoch": 30.29,
"learning_rate": 0.0001223283582089552,
"loss": 0.5383,
"step": 2090
},
{
"epoch": 30.43,
"learning_rate": 0.00012143283582089551,
"loss": 0.2471,
"step": 2100
},
{
"epoch": 30.57,
"learning_rate": 0.0001205373134328358,
"loss": 0.3912,
"step": 2110
},
{
"epoch": 30.72,
"learning_rate": 0.00011964179104477611,
"loss": 0.3958,
"step": 2120
},
{
"epoch": 30.86,
"learning_rate": 0.00011874626865671641,
"loss": 0.2196,
"step": 2130
},
{
"epoch": 30.99,
"eval_loss": 0.5438033938407898,
"eval_runtime": 62.6522,
"eval_samples_per_second": 8.97,
"eval_steps_per_second": 2.251,
"eval_wer": 0.1706497703691104,
"step": 2139
},
{
"epoch": 31.01,
"learning_rate": 0.0001178507462686567,
"loss": 0.5293,
"step": 2140
},
{
"epoch": 31.16,
"learning_rate": 0.00011695522388059701,
"loss": 0.1327,
"step": 2150
},
{
"epoch": 31.3,
"learning_rate": 0.0001160597014925373,
"loss": 0.4653,
"step": 2160
},
{
"epoch": 31.44,
"learning_rate": 0.00011516417910447761,
"loss": 0.3382,
"step": 2170
},
{
"epoch": 31.59,
"learning_rate": 0.0001142686567164179,
"loss": 0.3416,
"step": 2180
},
{
"epoch": 31.73,
"learning_rate": 0.0001133731343283582,
"loss": 0.5245,
"step": 2190
},
{
"epoch": 31.87,
"learning_rate": 0.00011247761194029848,
"loss": 0.1683,
"step": 2200
},
{
"epoch": 31.99,
"eval_loss": 0.60663241147995,
"eval_runtime": 61.8918,
"eval_samples_per_second": 9.08,
"eval_steps_per_second": 2.278,
"eval_wer": 0.1855332539547542,
"step": 2208
},
{
"epoch": 32.03,
"learning_rate": 0.0001115820895522388,
"loss": 0.5343,
"step": 2210
},
{
"epoch": 32.17,
"learning_rate": 0.0001106865671641791,
"loss": 0.2094,
"step": 2220
},
{
"epoch": 32.32,
"learning_rate": 0.0001097910447761194,
"loss": 0.4059,
"step": 2230
},
{
"epoch": 32.46,
"learning_rate": 0.0001088955223880597,
"loss": 0.3865,
"step": 2240
},
{
"epoch": 32.6,
"learning_rate": 0.00010799999999999998,
"loss": 0.2704,
"step": 2250
},
{
"epoch": 32.75,
"learning_rate": 0.0001071044776119403,
"loss": 0.5108,
"step": 2260
},
{
"epoch": 32.89,
"learning_rate": 0.00010620895522388058,
"loss": 0.1552,
"step": 2270
},
{
"epoch": 32.99,
"eval_loss": 0.5248196721076965,
"eval_runtime": 62.3014,
"eval_samples_per_second": 9.021,
"eval_steps_per_second": 2.263,
"eval_wer": 0.19301751998639224,
"step": 2277
},
{
"epoch": 33.04,
"learning_rate": 0.00010531343283582088,
"loss": 0.5527,
"step": 2280
},
{
"epoch": 33.19,
"learning_rate": 0.00010441791044776118,
"loss": 0.2536,
"step": 2290
},
{
"epoch": 33.33,
"learning_rate": 0.00010352238805970148,
"loss": 0.3816,
"step": 2300
},
{
"epoch": 33.47,
"learning_rate": 0.0001026268656716418,
"loss": 0.3807,
"step": 2310
},
{
"epoch": 33.62,
"learning_rate": 0.00010173134328358208,
"loss": 0.1987,
"step": 2320
},
{
"epoch": 33.76,
"learning_rate": 0.00010083582089552238,
"loss": 0.4947,
"step": 2330
},
{
"epoch": 33.9,
"learning_rate": 9.994029850746268e-05,
"loss": 0.1682,
"step": 2340
},
{
"epoch": 33.99,
"eval_loss": 0.5439924001693726,
"eval_runtime": 61.6613,
"eval_samples_per_second": 9.114,
"eval_steps_per_second": 2.287,
"eval_wer": 0.17830413335601292,
"step": 2346
},
{
"epoch": 34.06,
"learning_rate": 9.904477611940298e-05,
"loss": 0.4434,
"step": 2350
},
{
"epoch": 34.2,
"learning_rate": 9.814925373134327e-05,
"loss": 0.3305,
"step": 2360
},
{
"epoch": 34.34,
"learning_rate": 9.725373134328358e-05,
"loss": 0.3414,
"step": 2370
},
{
"epoch": 34.49,
"learning_rate": 9.635820895522387e-05,
"loss": 0.4277,
"step": 2380
},
{
"epoch": 34.63,
"learning_rate": 9.546268656716417e-05,
"loss": 0.1413,
"step": 2390
},
{
"epoch": 34.77,
"learning_rate": 9.456716417910448e-05,
"loss": 0.4671,
"step": 2400
},
{
"epoch": 34.92,
"learning_rate": 9.367164179104477e-05,
"loss": 0.2162,
"step": 2410
},
{
"epoch": 34.99,
"eval_loss": 0.6078898906707764,
"eval_runtime": 62.2539,
"eval_samples_per_second": 9.028,
"eval_steps_per_second": 2.265,
"eval_wer": 0.1777513182514033,
"step": 2415
},
{
"epoch": 35.07,
"learning_rate": 9.277611940298507e-05,
"loss": 0.4317,
"step": 2420
},
{
"epoch": 35.22,
"learning_rate": 9.188059701492537e-05,
"loss": 0.3773,
"step": 2430
},
{
"epoch": 35.36,
"learning_rate": 9.098507462686567e-05,
"loss": 0.3438,
"step": 2440
},
{
"epoch": 35.5,
"learning_rate": 9.008955223880595e-05,
"loss": 0.5443,
"step": 2450
},
{
"epoch": 35.65,
"learning_rate": 8.919402985074626e-05,
"loss": 0.1763,
"step": 2460
},
{
"epoch": 35.79,
"learning_rate": 8.829850746268655e-05,
"loss": 0.4314,
"step": 2470
},
{
"epoch": 35.93,
"learning_rate": 8.740298507462686e-05,
"loss": 0.3041,
"step": 2480
},
{
"epoch": 35.99,
"eval_loss": 0.5607585906982422,
"eval_runtime": 62.5839,
"eval_samples_per_second": 8.98,
"eval_steps_per_second": 2.253,
"eval_wer": 0.18344956625276407,
"step": 2484
},
{
"epoch": 36.09,
"learning_rate": 8.650746268656716e-05,
"loss": 0.3363,
"step": 2490
},
{
"epoch": 36.23,
"learning_rate": 8.561194029850745e-05,
"loss": 0.4153,
"step": 2500
},
{
"epoch": 36.37,
"learning_rate": 8.471641791044776e-05,
"loss": 0.2043,
"step": 2510
},
{
"epoch": 36.52,
"learning_rate": 8.382089552238805e-05,
"loss": 0.4476,
"step": 2520
},
{
"epoch": 36.66,
"learning_rate": 8.292537313432835e-05,
"loss": 0.1891,
"step": 2530
},
{
"epoch": 36.8,
"learning_rate": 8.202985074626865e-05,
"loss": 0.3683,
"step": 2540
},
{
"epoch": 36.95,
"learning_rate": 8.113432835820895e-05,
"loss": 0.3188,
"step": 2550
},
{
"epoch": 36.99,
"eval_loss": 0.6039144396781921,
"eval_runtime": 65.9812,
"eval_samples_per_second": 8.518,
"eval_steps_per_second": 2.137,
"eval_wer": 0.20067188297329477,
"step": 2553
},
{
"epoch": 37.1,
"learning_rate": 8.023880597014924e-05,
"loss": 0.3105,
"step": 2560
},
{
"epoch": 37.24,
"learning_rate": 7.934328358208955e-05,
"loss": 0.5286,
"step": 2570
},
{
"epoch": 37.39,
"learning_rate": 7.844776119402985e-05,
"loss": 0.1466,
"step": 2580
},
{
"epoch": 37.53,
"learning_rate": 7.755223880597013e-05,
"loss": 0.4602,
"step": 2590
},
{
"epoch": 37.67,
"learning_rate": 7.665671641791045e-05,
"loss": 0.2189,
"step": 2600
},
{
"epoch": 37.82,
"learning_rate": 7.576119402985073e-05,
"loss": 0.488,
"step": 2610
},
{
"epoch": 37.96,
"learning_rate": 7.486567164179105e-05,
"loss": 0.3692,
"step": 2620
},
{
"epoch": 37.99,
"eval_loss": 0.5436875820159912,
"eval_runtime": 63.3533,
"eval_samples_per_second": 8.871,
"eval_steps_per_second": 2.226,
"eval_wer": 0.1769008334750808,
"step": 2622
},
{
"epoch": 38.11,
"learning_rate": 7.397014925373133e-05,
"loss": 0.223,
"step": 2630
},
{
"epoch": 38.26,
"learning_rate": 7.307462686567163e-05,
"loss": 0.5086,
"step": 2640
},
{
"epoch": 38.4,
"learning_rate": 7.217910447761193e-05,
"loss": 0.129,
"step": 2650
},
{
"epoch": 38.54,
"learning_rate": 7.128358208955223e-05,
"loss": 0.3763,
"step": 2660
},
{
"epoch": 38.69,
"learning_rate": 7.038805970149253e-05,
"loss": 0.3218,
"step": 2670
},
{
"epoch": 38.83,
"learning_rate": 6.949253731343283e-05,
"loss": 0.3179,
"step": 2680
},
{
"epoch": 38.97,
"learning_rate": 6.859701492537312e-05,
"loss": 0.4446,
"step": 2690
},
{
"epoch": 38.99,
"eval_loss": 0.6474555730819702,
"eval_runtime": 61.8911,
"eval_samples_per_second": 9.08,
"eval_steps_per_second": 2.278,
"eval_wer": 0.18812723252253785,
"step": 2691
},
{
"epoch": 39.13,
"learning_rate": 6.770149253731342e-05,
"loss": 0.1771,
"step": 2700
},
{
"epoch": 39.27,
"learning_rate": 6.680597014925373e-05,
"loss": 0.4132,
"step": 2710
},
{
"epoch": 39.42,
"learning_rate": 6.591044776119403e-05,
"loss": 0.1985,
"step": 2720
},
{
"epoch": 39.56,
"learning_rate": 6.501492537313432e-05,
"loss": 0.3993,
"step": 2730
},
{
"epoch": 39.7,
"learning_rate": 6.411940298507462e-05,
"loss": 0.3009,
"step": 2740
},
{
"epoch": 39.85,
"learning_rate": 6.322388059701492e-05,
"loss": 0.2899,
"step": 2750
},
{
"epoch": 39.99,
"learning_rate": 6.232835820895522e-05,
"loss": 0.386,
"step": 2760
},
{
"epoch": 39.99,
"eval_loss": 0.6468256115913391,
"eval_runtime": 62.0034,
"eval_samples_per_second": 9.064,
"eval_steps_per_second": 2.274,
"eval_wer": 0.18944548392583774,
"step": 2760
},
{
"epoch": 40.14,
"learning_rate": 6.143283582089552e-05,
"loss": 0.1902,
"step": 2770
},
{
"epoch": 40.29,
"learning_rate": 6.053731343283581e-05,
"loss": 0.4626,
"step": 2780
},
{
"epoch": 40.43,
"learning_rate": 5.964179104477611e-05,
"loss": 0.1887,
"step": 2790
},
{
"epoch": 40.57,
"learning_rate": 5.874626865671642e-05,
"loss": 0.355,
"step": 2800
},
{
"epoch": 40.72,
"learning_rate": 5.785074626865672e-05,
"loss": 0.3404,
"step": 2810
},
{
"epoch": 40.86,
"learning_rate": 5.695522388059701e-05,
"loss": 0.1995,
"step": 2820
},
{
"epoch": 40.99,
"eval_loss": 0.6397563815116882,
"eval_runtime": 63.1605,
"eval_samples_per_second": 8.898,
"eval_steps_per_second": 2.232,
"eval_wer": 0.190551114135057,
"step": 2829
},
{
"epoch": 41.01,
"learning_rate": 5.605970149253731e-05,
"loss": 0.4191,
"step": 2830
},
{
"epoch": 41.16,
"learning_rate": 5.516417910447761e-05,
"loss": 0.1451,
"step": 2840
},
{
"epoch": 41.3,
"learning_rate": 5.42686567164179e-05,
"loss": 0.4273,
"step": 2850
},
{
"epoch": 41.44,
"learning_rate": 5.33731343283582e-05,
"loss": 0.237,
"step": 2860
},
{
"epoch": 41.59,
"learning_rate": 5.24776119402985e-05,
"loss": 0.3044,
"step": 2870
},
{
"epoch": 41.73,
"learning_rate": 5.1582089552238795e-05,
"loss": 0.5847,
"step": 2880
},
{
"epoch": 41.87,
"learning_rate": 5.06865671641791e-05,
"loss": 0.1174,
"step": 2890
},
{
"epoch": 41.99,
"eval_loss": 0.5986683368682861,
"eval_runtime": 63.8251,
"eval_samples_per_second": 8.805,
"eval_steps_per_second": 2.209,
"eval_wer": 0.193612859329818,
"step": 2898
},
{
"epoch": 42.03,
"learning_rate": 4.97910447761194e-05,
"loss": 0.5034,
"step": 2900
},
{
"epoch": 42.17,
"learning_rate": 4.88955223880597e-05,
"loss": 0.1607,
"step": 2910
},
{
"epoch": 42.32,
"learning_rate": 4.7999999999999994e-05,
"loss": 0.3704,
"step": 2920
},
{
"epoch": 42.46,
"learning_rate": 4.7104477611940294e-05,
"loss": 0.2891,
"step": 2930
},
{
"epoch": 42.6,
"learning_rate": 4.6208955223880594e-05,
"loss": 0.2184,
"step": 2940
},
{
"epoch": 42.75,
"learning_rate": 4.531343283582089e-05,
"loss": 0.441,
"step": 2950
},
{
"epoch": 42.89,
"learning_rate": 4.441791044776119e-05,
"loss": 0.1288,
"step": 2960
},
{
"epoch": 42.99,
"eval_loss": 0.613288938999176,
"eval_runtime": 62.2467,
"eval_samples_per_second": 9.029,
"eval_steps_per_second": 2.265,
"eval_wer": 0.18710665079095085,
"step": 2967
},
{
"epoch": 43.04,
"learning_rate": 4.352238805970149e-05,
"loss": 0.5541,
"step": 2970
},
{
"epoch": 43.19,
"learning_rate": 4.262686567164179e-05,
"loss": 0.1988,
"step": 2980
},
{
"epoch": 43.33,
"learning_rate": 4.1731343283582086e-05,
"loss": 0.3245,
"step": 2990
},
{
"epoch": 43.47,
"learning_rate": 4.0835820895522386e-05,
"loss": 0.4007,
"step": 3000
},
{
"epoch": 43.62,
"learning_rate": 3.9940298507462686e-05,
"loss": 0.1796,
"step": 3010
},
{
"epoch": 43.76,
"learning_rate": 3.904477611940298e-05,
"loss": 0.3904,
"step": 3020
},
{
"epoch": 43.9,
"learning_rate": 3.814925373134328e-05,
"loss": 0.1857,
"step": 3030
},
{
"epoch": 43.99,
"eval_loss": 0.6976124048233032,
"eval_runtime": 63.6276,
"eval_samples_per_second": 8.833,
"eval_steps_per_second": 2.216,
"eval_wer": 0.1995237285252594,
"step": 3036
},
{
"epoch": 44.06,
"learning_rate": 3.725373134328358e-05,
"loss": 0.4402,
"step": 3040
},
{
"epoch": 44.2,
"learning_rate": 3.635820895522388e-05,
"loss": 0.2266,
"step": 3050
},
{
"epoch": 44.34,
"learning_rate": 3.546268656716418e-05,
"loss": 0.2488,
"step": 3060
},
{
"epoch": 44.49,
"learning_rate": 3.456716417910447e-05,
"loss": 0.4742,
"step": 3070
},
{
"epoch": 44.63,
"learning_rate": 3.367164179104477e-05,
"loss": 0.1255,
"step": 3080
},
{
"epoch": 44.77,
"learning_rate": 3.277611940298507e-05,
"loss": 0.4457,
"step": 3090
},
{
"epoch": 44.92,
"learning_rate": 3.188059701492537e-05,
"loss": 0.2025,
"step": 3100
},
{
"epoch": 44.99,
"eval_loss": 0.6356056928634644,
"eval_runtime": 62.7237,
"eval_samples_per_second": 8.96,
"eval_steps_per_second": 2.248,
"eval_wer": 0.19016839598571186,
"step": 3105
},
{
"epoch": 45.07,
"learning_rate": 3.098507462686567e-05,
"loss": 0.4657,
"step": 3110
},
{
"epoch": 45.22,
"learning_rate": 3.0089552238805967e-05,
"loss": 0.2643,
"step": 3120
},
{
"epoch": 45.36,
"learning_rate": 2.9194029850746263e-05,
"loss": 0.2344,
"step": 3130
},
{
"epoch": 45.5,
"learning_rate": 2.8298507462686566e-05,
"loss": 0.4605,
"step": 3140
},
{
"epoch": 45.65,
"learning_rate": 2.7402985074626863e-05,
"loss": 0.1217,
"step": 3150
},
{
"epoch": 45.79,
"learning_rate": 2.6507462686567163e-05,
"loss": 0.3799,
"step": 3160
},
{
"epoch": 45.93,
"learning_rate": 2.561194029850746e-05,
"loss": 0.2922,
"step": 3170
},
{
"epoch": 45.99,
"eval_loss": 0.6324025392532349,
"eval_runtime": 63.9125,
"eval_samples_per_second": 8.793,
"eval_steps_per_second": 2.206,
"eval_wer": 0.20551964619833304,
"step": 3174
},
{
"epoch": 46.09,
"learning_rate": 2.4716417910447762e-05,
"loss": 0.3588,
"step": 3180
},
{
"epoch": 46.23,
"learning_rate": 2.382089552238806e-05,
"loss": 0.3345,
"step": 3190
},
{
"epoch": 46.37,
"learning_rate": 2.2925373134328355e-05,
"loss": 0.1835,
"step": 3200
},
{
"epoch": 46.52,
"learning_rate": 2.2029850746268655e-05,
"loss": 0.4099,
"step": 3210
},
{
"epoch": 46.66,
"learning_rate": 2.113432835820895e-05,
"loss": 0.1369,
"step": 3220
},
{
"epoch": 46.8,
"learning_rate": 2.0238805970149254e-05,
"loss": 0.3236,
"step": 3230
},
{
"epoch": 46.95,
"learning_rate": 1.934328358208955e-05,
"loss": 0.3575,
"step": 3240
},
{
"epoch": 46.99,
"eval_loss": 0.6337515115737915,
"eval_runtime": 62.7137,
"eval_samples_per_second": 8.961,
"eval_steps_per_second": 2.248,
"eval_wer": 0.1862136417758122,
"step": 3243
},
{
"epoch": 47.1,
"learning_rate": 1.844776119402985e-05,
"loss": 0.3063,
"step": 3250
},
{
"epoch": 47.24,
"learning_rate": 1.755223880597015e-05,
"loss": 0.4116,
"step": 3260
},
{
"epoch": 47.39,
"learning_rate": 1.6656716417910447e-05,
"loss": 0.1219,
"step": 3270
},
{
"epoch": 47.53,
"learning_rate": 1.5761194029850743e-05,
"loss": 0.3642,
"step": 3280
},
{
"epoch": 47.67,
"learning_rate": 1.4865671641791043e-05,
"loss": 0.1964,
"step": 3290
},
{
"epoch": 47.82,
"learning_rate": 1.3970149253731341e-05,
"loss": 0.3214,
"step": 3300
},
{
"epoch": 47.96,
"learning_rate": 1.3074626865671641e-05,
"loss": 0.4019,
"step": 3310
},
{
"epoch": 47.99,
"eval_loss": 0.6112991571426392,
"eval_runtime": 62.2461,
"eval_samples_per_second": 9.029,
"eval_steps_per_second": 2.265,
"eval_wer": 0.18982820207518286,
"step": 3312
},
{
"epoch": 48.11,
"learning_rate": 1.2179104477611939e-05,
"loss": 0.2532,
"step": 3320
},
{
"epoch": 48.26,
"learning_rate": 1.1283582089552239e-05,
"loss": 0.4152,
"step": 3330
},
{
"epoch": 48.4,
"learning_rate": 1.0388059701492537e-05,
"loss": 0.1089,
"step": 3340
},
{
"epoch": 48.54,
"learning_rate": 9.492537313432837e-06,
"loss": 0.3996,
"step": 3350
},
{
"epoch": 48.69,
"learning_rate": 8.597014925373133e-06,
"loss": 0.2793,
"step": 3360
},
{
"epoch": 48.83,
"learning_rate": 7.701492537313433e-06,
"loss": 0.3038,
"step": 3370
},
{
"epoch": 48.97,
"learning_rate": 6.805970149253731e-06,
"loss": 0.4211,
"step": 3380
},
{
"epoch": 48.99,
"eval_loss": 0.6319797039031982,
"eval_runtime": 61.9148,
"eval_samples_per_second": 9.077,
"eval_steps_per_second": 2.277,
"eval_wer": 0.19476101377785338,
"step": 3381
},
{
"epoch": 49.13,
"learning_rate": 5.91044776119403e-06,
"loss": 0.1711,
"step": 3390
},
{
"epoch": 49.27,
"learning_rate": 5.014925373134327e-06,
"loss": 0.4138,
"step": 3400
},
{
"epoch": 49.42,
"learning_rate": 4.119402985074627e-06,
"loss": 0.143,
"step": 3410
},
{
"epoch": 49.56,
"learning_rate": 3.2238805970149247e-06,
"loss": 0.323,
"step": 3420
},
{
"epoch": 49.7,
"learning_rate": 2.3283582089552237e-06,
"loss": 0.2997,
"step": 3430
},
{
"epoch": 49.85,
"learning_rate": 1.4328358208955222e-06,
"loss": 0.2564,
"step": 3440
},
{
"epoch": 49.99,
"learning_rate": 5.373134328358208e-07,
"loss": 0.4323,
"step": 3450
},
{
"epoch": 49.99,
"eval_loss": 0.6307477951049805,
"eval_runtime": 61.2414,
"eval_samples_per_second": 9.177,
"eval_steps_per_second": 2.302,
"eval_wer": 0.19165674434427624,
"step": 3450
},
{
"epoch": 49.99,
"step": 3450,
"total_flos": 8.202511043473552e+18,
"train_loss": 1.0114854598390883,
"train_runtime": 11573.519,
"train_samples_per_second": 4.817,
"train_steps_per_second": 0.298
}
],
"max_steps": 3450,
"num_train_epochs": 50,
"total_flos": 8.202511043473552e+18,
"trial_name": null,
"trial_params": null
}