gary109's picture
End of training
6dd2a6f
{
"best_metric": 0.41795814037323,
"best_model_checkpoint": "ai-light-dance_drums_ft_pretrain_wav2vec2-base-new-13k_onset-drums_fold_2/checkpoint-1863",
"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": 2.9999999999999997e-05,
"loss": 0.3312,
"step": 10
},
{
"epoch": 0.29,
"learning_rate": 5.6999999999999996e-05,
"loss": 0.7196,
"step": 20
},
{
"epoch": 0.43,
"learning_rate": 8.699999999999999e-05,
"loss": 0.3342,
"step": 30
},
{
"epoch": 0.57,
"learning_rate": 0.000117,
"loss": 0.5121,
"step": 40
},
{
"epoch": 0.72,
"learning_rate": 0.000147,
"loss": 0.6408,
"step": 50
},
{
"epoch": 0.86,
"learning_rate": 0.00017699999999999997,
"loss": 0.3238,
"step": 60
},
{
"epoch": 0.99,
"eval_loss": 0.45810404419898987,
"eval_runtime": 62.3289,
"eval_samples_per_second": 9.001,
"eval_steps_per_second": 2.262,
"eval_wer": 0.20811523986676472,
"step": 69
},
{
"epoch": 1.01,
"learning_rate": 0.00020699999999999996,
"loss": 0.8089,
"step": 70
},
{
"epoch": 1.16,
"learning_rate": 0.000237,
"loss": 0.2476,
"step": 80
},
{
"epoch": 1.3,
"learning_rate": 0.000267,
"loss": 0.7357,
"step": 90
},
{
"epoch": 1.44,
"learning_rate": 0.00029699999999999996,
"loss": 0.4624,
"step": 100
},
{
"epoch": 1.59,
"learning_rate": 0.00029919402985074625,
"loss": 0.4802,
"step": 110
},
{
"epoch": 1.73,
"learning_rate": 0.0002982985074626866,
"loss": 0.7866,
"step": 120
},
{
"epoch": 1.87,
"learning_rate": 0.00029740298507462685,
"loss": 0.275,
"step": 130
},
{
"epoch": 1.99,
"eval_loss": 0.6493655443191528,
"eval_runtime": 63.3598,
"eval_samples_per_second": 8.854,
"eval_steps_per_second": 2.225,
"eval_wer": 0.33425617510922695,
"step": 138
},
{
"epoch": 2.03,
"learning_rate": 0.0002965074626865671,
"loss": 0.7862,
"step": 140
},
{
"epoch": 2.17,
"learning_rate": 0.00029561194029850745,
"loss": 0.3001,
"step": 150
},
{
"epoch": 2.32,
"learning_rate": 0.0002947164179104477,
"loss": 0.6909,
"step": 160
},
{
"epoch": 2.46,
"learning_rate": 0.00029382089552238805,
"loss": 0.4748,
"step": 170
},
{
"epoch": 2.6,
"learning_rate": 0.0002929253731343283,
"loss": 0.4152,
"step": 180
},
{
"epoch": 2.75,
"learning_rate": 0.00029202985074626864,
"loss": 0.7598,
"step": 190
},
{
"epoch": 2.89,
"learning_rate": 0.0002911343283582089,
"loss": 0.2965,
"step": 200
},
{
"epoch": 2.99,
"eval_loss": 0.6193479895591736,
"eval_runtime": 62.3049,
"eval_samples_per_second": 9.004,
"eval_steps_per_second": 2.263,
"eval_wer": 0.2274516589522862,
"step": 207
},
{
"epoch": 3.04,
"learning_rate": 0.0002903283582089552,
"loss": 0.8123,
"step": 210
},
{
"epoch": 3.19,
"learning_rate": 0.00028943283582089546,
"loss": 0.4209,
"step": 220
},
{
"epoch": 3.33,
"learning_rate": 0.0002885373134328358,
"loss": 0.604,
"step": 230
},
{
"epoch": 3.47,
"learning_rate": 0.0002876417910447761,
"loss": 0.5433,
"step": 240
},
{
"epoch": 3.62,
"learning_rate": 0.0002867462686567164,
"loss": 0.3383,
"step": 250
},
{
"epoch": 3.76,
"learning_rate": 0.0002858507462686567,
"loss": 0.7398,
"step": 260
},
{
"epoch": 3.9,
"learning_rate": 0.000284955223880597,
"loss": 0.3406,
"step": 270
},
{
"epoch": 3.99,
"eval_loss": 0.6934247612953186,
"eval_runtime": 61.6757,
"eval_samples_per_second": 9.096,
"eval_steps_per_second": 2.286,
"eval_wer": 0.2614958688411126,
"step": 276
},
{
"epoch": 4.06,
"learning_rate": 0.00028405970149253726,
"loss": 0.7566,
"step": 280
},
{
"epoch": 4.2,
"learning_rate": 0.0002831641791044776,
"loss": 0.4127,
"step": 290
},
{
"epoch": 4.34,
"learning_rate": 0.0002822686567164179,
"loss": 0.5787,
"step": 300
},
{
"epoch": 4.49,
"learning_rate": 0.0002813731343283582,
"loss": 0.8365,
"step": 310
},
{
"epoch": 4.63,
"learning_rate": 0.00028047761194029846,
"loss": 0.2287,
"step": 320
},
{
"epoch": 4.77,
"learning_rate": 0.0002795820895522388,
"loss": 0.6623,
"step": 330
},
{
"epoch": 4.92,
"learning_rate": 0.00027868656716417906,
"loss": 0.3906,
"step": 340
},
{
"epoch": 4.99,
"eval_loss": 0.6265394687652588,
"eval_runtime": 61.9233,
"eval_samples_per_second": 9.06,
"eval_steps_per_second": 2.277,
"eval_wer": 0.1835445775835965,
"step": 345
},
{
"epoch": 5.07,
"learning_rate": 0.0002777910447761194,
"loss": 0.5863,
"step": 350
},
{
"epoch": 5.22,
"learning_rate": 0.00027689552238805966,
"loss": 0.4848,
"step": 360
},
{
"epoch": 5.36,
"learning_rate": 0.000276,
"loss": 0.4226,
"step": 370
},
{
"epoch": 5.5,
"learning_rate": 0.00027510447761194026,
"loss": 0.6544,
"step": 380
},
{
"epoch": 5.65,
"learning_rate": 0.0002742089552238806,
"loss": 0.2422,
"step": 390
},
{
"epoch": 5.79,
"learning_rate": 0.00027331343283582086,
"loss": 0.6015,
"step": 400
},
{
"epoch": 5.93,
"learning_rate": 0.0002724179104477612,
"loss": 0.4643,
"step": 410
},
{
"epoch": 5.99,
"eval_loss": 0.5878893733024597,
"eval_runtime": 61.4594,
"eval_samples_per_second": 9.128,
"eval_steps_per_second": 2.294,
"eval_wer": 0.18990353419561362,
"step": 414
},
{
"epoch": 6.09,
"learning_rate": 0.00027152238805970146,
"loss": 0.61,
"step": 420
},
{
"epoch": 6.23,
"learning_rate": 0.00027062686567164173,
"loss": 0.5311,
"step": 430
},
{
"epoch": 6.37,
"learning_rate": 0.00026973134328358206,
"loss": 0.2503,
"step": 440
},
{
"epoch": 6.52,
"learning_rate": 0.0002688358208955224,
"loss": 0.7559,
"step": 450
},
{
"epoch": 6.66,
"learning_rate": 0.00026794029850746266,
"loss": 0.314,
"step": 460
},
{
"epoch": 6.8,
"learning_rate": 0.00026704477611940293,
"loss": 0.6115,
"step": 470
},
{
"epoch": 6.95,
"learning_rate": 0.00026614925373134326,
"loss": 0.4652,
"step": 480
},
{
"epoch": 6.99,
"eval_loss": 0.4961091876029968,
"eval_runtime": 61.4725,
"eval_samples_per_second": 9.126,
"eval_steps_per_second": 2.294,
"eval_wer": 0.16044469438075876,
"step": 483
},
{
"epoch": 7.1,
"learning_rate": 0.0002652537313432836,
"loss": 0.4757,
"step": 490
},
{
"epoch": 7.24,
"learning_rate": 0.00026435820895522386,
"loss": 0.694,
"step": 500
},
{
"epoch": 7.39,
"learning_rate": 0.0002634626865671642,
"loss": 0.1975,
"step": 510
},
{
"epoch": 7.53,
"learning_rate": 0.00026256716417910446,
"loss": 0.6958,
"step": 520
},
{
"epoch": 7.67,
"learning_rate": 0.00026167164179104473,
"loss": 0.3152,
"step": 530
},
{
"epoch": 7.82,
"learning_rate": 0.00026077611940298506,
"loss": 0.4814,
"step": 540
},
{
"epoch": 7.96,
"learning_rate": 0.0002598805970149254,
"loss": 0.4512,
"step": 550
},
{
"epoch": 7.99,
"eval_loss": 0.571247935295105,
"eval_runtime": 62.044,
"eval_samples_per_second": 9.042,
"eval_steps_per_second": 2.273,
"eval_wer": 0.2800536401782238,
"step": 552
},
{
"epoch": 8.11,
"learning_rate": 0.00025898507462686566,
"loss": 0.3612,
"step": 560
},
{
"epoch": 8.26,
"learning_rate": 0.00025808955223880593,
"loss": 0.7199,
"step": 570
},
{
"epoch": 8.4,
"learning_rate": 0.00025719402985074626,
"loss": 0.2195,
"step": 580
},
{
"epoch": 8.54,
"learning_rate": 0.00025629850746268653,
"loss": 0.673,
"step": 590
},
{
"epoch": 8.69,
"learning_rate": 0.00025540298507462685,
"loss": 0.3611,
"step": 600
},
{
"epoch": 8.83,
"learning_rate": 0.00025450746268656713,
"loss": 0.3776,
"step": 610
},
{
"epoch": 8.97,
"learning_rate": 0.00025361194029850745,
"loss": 0.5321,
"step": 620
},
{
"epoch": 8.99,
"eval_loss": 0.6898208260536194,
"eval_runtime": 62.8462,
"eval_samples_per_second": 8.927,
"eval_steps_per_second": 2.244,
"eval_wer": 0.29363671756715837,
"step": 621
},
{
"epoch": 9.13,
"learning_rate": 0.00025271641791044773,
"loss": 0.2462,
"step": 630
},
{
"epoch": 9.27,
"learning_rate": 0.00025182089552238805,
"loss": 0.5621,
"step": 640
},
{
"epoch": 9.42,
"learning_rate": 0.0002509253731343283,
"loss": 0.2758,
"step": 650
},
{
"epoch": 9.56,
"learning_rate": 0.00025002985074626865,
"loss": 0.6596,
"step": 660
},
{
"epoch": 9.7,
"learning_rate": 0.0002491343283582089,
"loss": 0.3786,
"step": 670
},
{
"epoch": 9.85,
"learning_rate": 0.0002482388059701492,
"loss": 0.4499,
"step": 680
},
{
"epoch": 9.99,
"learning_rate": 0.0002473432835820895,
"loss": 0.64,
"step": 690
},
{
"epoch": 9.99,
"eval_loss": 0.5916096568107605,
"eval_runtime": 62.3005,
"eval_samples_per_second": 9.005,
"eval_steps_per_second": 2.263,
"eval_wer": 0.26482675087597873,
"step": 690
},
{
"epoch": 10.14,
"learning_rate": 0.00024644776119402985,
"loss": 0.241,
"step": 700
},
{
"epoch": 10.29,
"learning_rate": 0.0002455522388059701,
"loss": 0.5644,
"step": 710
},
{
"epoch": 10.43,
"learning_rate": 0.0002446567164179104,
"loss": 0.2882,
"step": 720
},
{
"epoch": 10.57,
"learning_rate": 0.00024376119402985073,
"loss": 0.4583,
"step": 730
},
{
"epoch": 10.72,
"learning_rate": 0.00024286567164179103,
"loss": 0.5087,
"step": 740
},
{
"epoch": 10.86,
"learning_rate": 0.00024197014925373132,
"loss": 0.2959,
"step": 750
},
{
"epoch": 10.99,
"eval_loss": 0.5573664307594299,
"eval_runtime": 62.7291,
"eval_samples_per_second": 8.943,
"eval_steps_per_second": 2.248,
"eval_wer": 0.17450361206038845,
"step": 759
},
{
"epoch": 11.01,
"learning_rate": 0.0002410746268656716,
"loss": 0.6544,
"step": 760
},
{
"epoch": 11.16,
"learning_rate": 0.00024017910447761192,
"loss": 0.1627,
"step": 770
},
{
"epoch": 11.3,
"learning_rate": 0.00023928358208955222,
"loss": 0.5087,
"step": 780
},
{
"epoch": 11.44,
"learning_rate": 0.0002383880597014925,
"loss": 0.4087,
"step": 790
},
{
"epoch": 11.59,
"learning_rate": 0.00023749253731343282,
"loss": 0.3695,
"step": 800
},
{
"epoch": 11.73,
"learning_rate": 0.00023659701492537312,
"loss": 0.5399,
"step": 810
},
{
"epoch": 11.87,
"learning_rate": 0.0002357014925373134,
"loss": 0.2053,
"step": 820
},
{
"epoch": 11.99,
"eval_loss": 0.5216460824012756,
"eval_runtime": 62.584,
"eval_samples_per_second": 8.964,
"eval_steps_per_second": 2.253,
"eval_wer": 0.20093437729809233,
"step": 828
},
{
"epoch": 12.03,
"learning_rate": 0.0002348059701492537,
"loss": 0.6779,
"step": 830
},
{
"epoch": 12.17,
"learning_rate": 0.00023391044776119402,
"loss": 0.2308,
"step": 840
},
{
"epoch": 12.32,
"learning_rate": 0.0002330149253731343,
"loss": 0.4464,
"step": 850
},
{
"epoch": 12.46,
"learning_rate": 0.0002321194029850746,
"loss": 0.3519,
"step": 860
},
{
"epoch": 12.6,
"learning_rate": 0.00023122388059701492,
"loss": 0.3585,
"step": 870
},
{
"epoch": 12.75,
"learning_rate": 0.00023032835820895522,
"loss": 0.5939,
"step": 880
},
{
"epoch": 12.89,
"learning_rate": 0.0002294328358208955,
"loss": 0.2433,
"step": 890
},
{
"epoch": 12.99,
"eval_loss": 0.4737621247768402,
"eval_runtime": 62.102,
"eval_samples_per_second": 9.034,
"eval_steps_per_second": 2.27,
"eval_wer": 0.16425141670632,
"step": 897
},
{
"epoch": 13.04,
"learning_rate": 0.0002285373134328358,
"loss": 0.6217,
"step": 900
},
{
"epoch": 13.19,
"learning_rate": 0.00022764179104477612,
"loss": 0.2514,
"step": 910
},
{
"epoch": 13.33,
"learning_rate": 0.0002267462686567164,
"loss": 0.4339,
"step": 920
},
{
"epoch": 13.47,
"learning_rate": 0.0002258507462686567,
"loss": 0.459,
"step": 930
},
{
"epoch": 13.62,
"learning_rate": 0.00022495522388059697,
"loss": 0.2808,
"step": 940
},
{
"epoch": 13.76,
"learning_rate": 0.0002240597014925373,
"loss": 0.5249,
"step": 950
},
{
"epoch": 13.9,
"learning_rate": 0.0002231641791044776,
"loss": 0.2036,
"step": 960
},
{
"epoch": 13.99,
"eval_loss": 0.5062561631202698,
"eval_runtime": 62.2056,
"eval_samples_per_second": 9.018,
"eval_steps_per_second": 2.267,
"eval_wer": 0.1650733226629753,
"step": 966
},
{
"epoch": 14.06,
"learning_rate": 0.0002222686567164179,
"loss": 0.54,
"step": 970
},
{
"epoch": 14.2,
"learning_rate": 0.0002213731343283582,
"loss": 0.2911,
"step": 980
},
{
"epoch": 14.34,
"learning_rate": 0.0002204776119402985,
"loss": 0.3306,
"step": 990
},
{
"epoch": 14.49,
"learning_rate": 0.0002195820895522388,
"loss": 0.5251,
"step": 1000
},
{
"epoch": 14.63,
"learning_rate": 0.00021868656716417907,
"loss": 0.1507,
"step": 1010
},
{
"epoch": 14.77,
"learning_rate": 0.0002177910447761194,
"loss": 0.5758,
"step": 1020
},
{
"epoch": 14.92,
"learning_rate": 0.0002168955223880597,
"loss": 0.2654,
"step": 1030
},
{
"epoch": 14.99,
"eval_loss": 0.4903617799282074,
"eval_runtime": 62.3504,
"eval_samples_per_second": 8.998,
"eval_steps_per_second": 2.261,
"eval_wer": 0.15114417960808063,
"step": 1035
},
{
"epoch": 15.07,
"learning_rate": 0.00021599999999999996,
"loss": 0.505,
"step": 1040
},
{
"epoch": 15.22,
"learning_rate": 0.0002151044776119403,
"loss": 0.362,
"step": 1050
},
{
"epoch": 15.36,
"learning_rate": 0.0002142089552238806,
"loss": 0.3129,
"step": 1060
},
{
"epoch": 15.5,
"learning_rate": 0.00021331343283582086,
"loss": 0.5513,
"step": 1070
},
{
"epoch": 15.65,
"learning_rate": 0.00021241791044776116,
"loss": 0.1605,
"step": 1080
},
{
"epoch": 15.79,
"learning_rate": 0.0002115223880597015,
"loss": 0.4756,
"step": 1090
},
{
"epoch": 15.93,
"learning_rate": 0.00021062686567164176,
"loss": 0.3641,
"step": 1100
},
{
"epoch": 15.99,
"eval_loss": 0.4660147726535797,
"eval_runtime": 62.4193,
"eval_samples_per_second": 8.988,
"eval_steps_per_second": 2.259,
"eval_wer": 0.1668901674092659,
"step": 1104
},
{
"epoch": 16.09,
"learning_rate": 0.00020973134328358206,
"loss": 0.3724,
"step": 1110
},
{
"epoch": 16.23,
"learning_rate": 0.00020883582089552236,
"loss": 0.3679,
"step": 1120
},
{
"epoch": 16.37,
"learning_rate": 0.0002079402985074627,
"loss": 0.2284,
"step": 1130
},
{
"epoch": 16.52,
"learning_rate": 0.00020704477611940296,
"loss": 0.5271,
"step": 1140
},
{
"epoch": 16.66,
"learning_rate": 0.00020614925373134326,
"loss": 0.2279,
"step": 1150
},
{
"epoch": 16.8,
"learning_rate": 0.0002052537313432836,
"loss": 0.4453,
"step": 1160
},
{
"epoch": 16.95,
"learning_rate": 0.00020435820895522386,
"loss": 0.373,
"step": 1170
},
{
"epoch": 16.99,
"eval_loss": 0.5132745504379272,
"eval_runtime": 63.5973,
"eval_samples_per_second": 8.821,
"eval_steps_per_second": 2.217,
"eval_wer": 0.21058095773673055,
"step": 1173
},
{
"epoch": 17.1,
"learning_rate": 0.00020346268656716416,
"loss": 0.3844,
"step": 1180
},
{
"epoch": 17.24,
"learning_rate": 0.00020256716417910443,
"loss": 0.4336,
"step": 1190
},
{
"epoch": 17.39,
"learning_rate": 0.00020167164179104476,
"loss": 0.1497,
"step": 1200
},
{
"epoch": 17.53,
"learning_rate": 0.00020077611940298506,
"loss": 0.4863,
"step": 1210
},
{
"epoch": 17.67,
"learning_rate": 0.00019988059701492536,
"loss": 0.2108,
"step": 1220
},
{
"epoch": 17.82,
"learning_rate": 0.00019898507462686566,
"loss": 0.3513,
"step": 1230
},
{
"epoch": 17.96,
"learning_rate": 0.00019808955223880596,
"loss": 0.4715,
"step": 1240
},
{
"epoch": 17.99,
"eval_loss": 0.5312526226043701,
"eval_runtime": 62.3777,
"eval_samples_per_second": 8.994,
"eval_steps_per_second": 2.26,
"eval_wer": 0.19120128044296406,
"step": 1242
},
{
"epoch": 18.11,
"learning_rate": 0.00019719402985074626,
"loss": 0.2865,
"step": 1250
},
{
"epoch": 18.26,
"learning_rate": 0.00019629850746268653,
"loss": 0.5362,
"step": 1260
},
{
"epoch": 18.4,
"learning_rate": 0.00019540298507462686,
"loss": 0.178,
"step": 1270
},
{
"epoch": 18.54,
"learning_rate": 0.00019450746268656716,
"loss": 0.4632,
"step": 1280
},
{
"epoch": 18.69,
"learning_rate": 0.00019361194029850743,
"loss": 0.2589,
"step": 1290
},
{
"epoch": 18.83,
"learning_rate": 0.00019271641791044773,
"loss": 0.346,
"step": 1300
},
{
"epoch": 18.97,
"learning_rate": 0.00019182089552238806,
"loss": 0.4893,
"step": 1310
},
{
"epoch": 18.99,
"eval_loss": 0.5152482390403748,
"eval_runtime": 62.4572,
"eval_samples_per_second": 8.982,
"eval_steps_per_second": 2.258,
"eval_wer": 0.17117273002552233,
"step": 1311
},
{
"epoch": 19.13,
"learning_rate": 0.00019092537313432833,
"loss": 0.2141,
"step": 1320
},
{
"epoch": 19.27,
"learning_rate": 0.00019002985074626863,
"loss": 0.5869,
"step": 1330
},
{
"epoch": 19.42,
"learning_rate": 0.00018913432835820896,
"loss": 0.2126,
"step": 1340
},
{
"epoch": 19.56,
"learning_rate": 0.00018823880597014923,
"loss": 0.4274,
"step": 1350
},
{
"epoch": 19.7,
"learning_rate": 0.00018734328358208953,
"loss": 0.3929,
"step": 1360
},
{
"epoch": 19.85,
"learning_rate": 0.00018644776119402983,
"loss": 0.259,
"step": 1370
},
{
"epoch": 19.99,
"learning_rate": 0.00018555223880597013,
"loss": 0.4875,
"step": 1380
},
{
"epoch": 19.99,
"eval_loss": 0.5481781959533691,
"eval_runtime": 62.6832,
"eval_samples_per_second": 8.95,
"eval_steps_per_second": 2.249,
"eval_wer": 0.17182160314919756,
"step": 1380
},
{
"epoch": 20.14,
"learning_rate": 0.00018465671641791043,
"loss": 0.1949,
"step": 1390
},
{
"epoch": 20.29,
"learning_rate": 0.00018376119402985073,
"loss": 0.4376,
"step": 1400
},
{
"epoch": 20.43,
"learning_rate": 0.00018286567164179106,
"loss": 0.2543,
"step": 1410
},
{
"epoch": 20.57,
"learning_rate": 0.00018197014925373133,
"loss": 0.4001,
"step": 1420
},
{
"epoch": 20.72,
"learning_rate": 0.00018107462686567163,
"loss": 0.3915,
"step": 1430
},
{
"epoch": 20.86,
"learning_rate": 0.0001801791044776119,
"loss": 0.1971,
"step": 1440
},
{
"epoch": 20.99,
"eval_loss": 0.45662185549736023,
"eval_runtime": 63.0963,
"eval_samples_per_second": 8.891,
"eval_steps_per_second": 2.235,
"eval_wer": 0.14487173941255352,
"step": 1449
},
{
"epoch": 21.01,
"learning_rate": 0.00017928358208955223,
"loss": 0.6252,
"step": 1450
},
{
"epoch": 21.16,
"learning_rate": 0.00017838805970149253,
"loss": 0.1249,
"step": 1460
},
{
"epoch": 21.3,
"learning_rate": 0.0001774925373134328,
"loss": 0.3906,
"step": 1470
},
{
"epoch": 21.44,
"learning_rate": 0.0001765970149253731,
"loss": 0.3267,
"step": 1480
},
{
"epoch": 21.59,
"learning_rate": 0.00017570149253731343,
"loss": 0.3302,
"step": 1490
},
{
"epoch": 21.73,
"learning_rate": 0.00017480597014925373,
"loss": 0.4891,
"step": 1500
},
{
"epoch": 21.87,
"learning_rate": 0.000173910447761194,
"loss": 0.1286,
"step": 1510
},
{
"epoch": 21.99,
"eval_loss": 0.45152103900909424,
"eval_runtime": 62.9105,
"eval_samples_per_second": 8.917,
"eval_steps_per_second": 2.241,
"eval_wer": 0.1478132975732145,
"step": 1518
},
{
"epoch": 22.03,
"learning_rate": 0.00017301492537313433,
"loss": 0.5264,
"step": 1520
},
{
"epoch": 22.17,
"learning_rate": 0.00017211940298507463,
"loss": 0.1714,
"step": 1530
},
{
"epoch": 22.32,
"learning_rate": 0.0001712238805970149,
"loss": 0.3947,
"step": 1540
},
{
"epoch": 22.46,
"learning_rate": 0.0001703283582089552,
"loss": 0.3371,
"step": 1550
},
{
"epoch": 22.6,
"learning_rate": 0.00016943283582089553,
"loss": 0.2423,
"step": 1560
},
{
"epoch": 22.75,
"learning_rate": 0.0001685373134328358,
"loss": 0.5367,
"step": 1570
},
{
"epoch": 22.89,
"learning_rate": 0.0001676417910447761,
"loss": 0.1472,
"step": 1580
},
{
"epoch": 22.99,
"eval_loss": 0.5059288144111633,
"eval_runtime": 62.4349,
"eval_samples_per_second": 8.985,
"eval_steps_per_second": 2.258,
"eval_wer": 0.1418436648354025,
"step": 1587
},
{
"epoch": 23.04,
"learning_rate": 0.00016674626865671643,
"loss": 0.4567,
"step": 1590
},
{
"epoch": 23.19,
"learning_rate": 0.0001658507462686567,
"loss": 0.2266,
"step": 1600
},
{
"epoch": 23.33,
"learning_rate": 0.000164955223880597,
"loss": 0.3705,
"step": 1610
},
{
"epoch": 23.47,
"learning_rate": 0.0001640597014925373,
"loss": 0.3832,
"step": 1620
},
{
"epoch": 23.62,
"learning_rate": 0.0001631641791044776,
"loss": 0.2372,
"step": 1630
},
{
"epoch": 23.76,
"learning_rate": 0.0001622686567164179,
"loss": 0.4775,
"step": 1640
},
{
"epoch": 23.9,
"learning_rate": 0.0001613731343283582,
"loss": 0.1917,
"step": 1650
},
{
"epoch": 23.99,
"eval_loss": 0.5583454966545105,
"eval_runtime": 62.2591,
"eval_samples_per_second": 9.011,
"eval_steps_per_second": 2.265,
"eval_wer": 0.1456936453692088,
"step": 1656
},
{
"epoch": 24.06,
"learning_rate": 0.00016047761194029847,
"loss": 0.4329,
"step": 1660
},
{
"epoch": 24.2,
"learning_rate": 0.0001595820895522388,
"loss": 0.2719,
"step": 1670
},
{
"epoch": 24.34,
"learning_rate": 0.0001586865671641791,
"loss": 0.3019,
"step": 1680
},
{
"epoch": 24.49,
"learning_rate": 0.00015779104477611937,
"loss": 0.3512,
"step": 1690
},
{
"epoch": 24.63,
"learning_rate": 0.0001568955223880597,
"loss": 0.157,
"step": 1700
},
{
"epoch": 24.77,
"learning_rate": 0.000156,
"loss": 0.5036,
"step": 1710
},
{
"epoch": 24.92,
"learning_rate": 0.00015510447761194027,
"loss": 0.2874,
"step": 1720
},
{
"epoch": 24.99,
"eval_loss": 0.5195438861846924,
"eval_runtime": 62.6596,
"eval_samples_per_second": 8.953,
"eval_steps_per_second": 2.25,
"eval_wer": 0.15027901544318034,
"step": 1725
},
{
"epoch": 25.07,
"learning_rate": 0.00015420895522388057,
"loss": 0.4331,
"step": 1730
},
{
"epoch": 25.22,
"learning_rate": 0.0001533134328358209,
"loss": 0.3291,
"step": 1740
},
{
"epoch": 25.36,
"learning_rate": 0.00015241791044776117,
"loss": 0.2418,
"step": 1750
},
{
"epoch": 25.5,
"learning_rate": 0.00015152238805970147,
"loss": 0.3786,
"step": 1760
},
{
"epoch": 25.65,
"learning_rate": 0.0001506268656716418,
"loss": 0.1725,
"step": 1770
},
{
"epoch": 25.79,
"learning_rate": 0.0001497313432835821,
"loss": 0.3772,
"step": 1780
},
{
"epoch": 25.93,
"learning_rate": 0.00014883582089552237,
"loss": 0.2252,
"step": 1790
},
{
"epoch": 25.99,
"eval_loss": 0.4408688247203827,
"eval_runtime": 62.9542,
"eval_samples_per_second": 8.911,
"eval_steps_per_second": 2.24,
"eval_wer": 0.15062508110914047,
"step": 1794
},
{
"epoch": 26.09,
"learning_rate": 0.00014794029850746267,
"loss": 0.3815,
"step": 1800
},
{
"epoch": 26.23,
"learning_rate": 0.00014704477611940297,
"loss": 0.3281,
"step": 1810
},
{
"epoch": 26.37,
"learning_rate": 0.00014614925373134327,
"loss": 0.1803,
"step": 1820
},
{
"epoch": 26.52,
"learning_rate": 0.00014534328358208954,
"loss": 0.4612,
"step": 1830
},
{
"epoch": 26.66,
"learning_rate": 0.00014444776119402984,
"loss": 0.1874,
"step": 1840
},
{
"epoch": 26.8,
"learning_rate": 0.00014355223880597014,
"loss": 0.4485,
"step": 1850
},
{
"epoch": 26.95,
"learning_rate": 0.00014265671641791044,
"loss": 0.3142,
"step": 1860
},
{
"epoch": 26.99,
"eval_loss": 0.41795814037323,
"eval_runtime": 62.551,
"eval_samples_per_second": 8.969,
"eval_steps_per_second": 2.254,
"eval_wer": 0.143314443915733,
"step": 1863
},
{
"epoch": 27.1,
"learning_rate": 0.00014176119402985074,
"loss": 0.2691,
"step": 1870
},
{
"epoch": 27.24,
"learning_rate": 0.00014086567164179104,
"loss": 0.4776,
"step": 1880
},
{
"epoch": 27.39,
"learning_rate": 0.00013997014925373134,
"loss": 0.1103,
"step": 1890
},
{
"epoch": 27.53,
"learning_rate": 0.0001390746268656716,
"loss": 0.4259,
"step": 1900
},
{
"epoch": 27.67,
"learning_rate": 0.00013817910447761194,
"loss": 0.223,
"step": 1910
},
{
"epoch": 27.82,
"learning_rate": 0.0001372835820895522,
"loss": 0.297,
"step": 1920
},
{
"epoch": 27.96,
"learning_rate": 0.0001363880597014925,
"loss": 0.385,
"step": 1930
},
{
"epoch": 27.99,
"eval_loss": 0.470846951007843,
"eval_runtime": 62.9775,
"eval_samples_per_second": 8.908,
"eval_steps_per_second": 2.239,
"eval_wer": 0.13673919626249081,
"step": 1932
},
{
"epoch": 28.11,
"learning_rate": 0.00013549253731343284,
"loss": 0.2514,
"step": 1940
},
{
"epoch": 28.26,
"learning_rate": 0.0001345970149253731,
"loss": 0.4299,
"step": 1950
},
{
"epoch": 28.4,
"learning_rate": 0.00013370149253731344,
"loss": 0.1096,
"step": 1960
},
{
"epoch": 28.54,
"learning_rate": 0.0001328059701492537,
"loss": 0.3887,
"step": 1970
},
{
"epoch": 28.69,
"learning_rate": 0.000131910447761194,
"loss": 0.2654,
"step": 1980
},
{
"epoch": 28.83,
"learning_rate": 0.0001310149253731343,
"loss": 0.3682,
"step": 1990
},
{
"epoch": 28.97,
"learning_rate": 0.0001301194029850746,
"loss": 0.4296,
"step": 2000
},
{
"epoch": 28.99,
"eval_loss": 0.4739513695240021,
"eval_runtime": 63.1343,
"eval_samples_per_second": 8.886,
"eval_steps_per_second": 2.233,
"eval_wer": 0.15058182290089545,
"step": 2001
},
{
"epoch": 29.13,
"learning_rate": 0.0001292238805970149,
"loss": 0.17,
"step": 2010
},
{
"epoch": 29.27,
"learning_rate": 0.0001283283582089552,
"loss": 0.4158,
"step": 2020
},
{
"epoch": 29.42,
"learning_rate": 0.0001274328358208955,
"loss": 0.1852,
"step": 2030
},
{
"epoch": 29.56,
"learning_rate": 0.0001265373134328358,
"loss": 0.3744,
"step": 2040
},
{
"epoch": 29.7,
"learning_rate": 0.0001256417910447761,
"loss": 0.2737,
"step": 2050
},
{
"epoch": 29.85,
"learning_rate": 0.0001247462686567164,
"loss": 0.2033,
"step": 2060
},
{
"epoch": 29.99,
"learning_rate": 0.0001238507462686567,
"loss": 0.4404,
"step": 2070
},
{
"epoch": 29.99,
"eval_loss": 0.46518003940582275,
"eval_runtime": 63.0114,
"eval_samples_per_second": 8.903,
"eval_steps_per_second": 2.238,
"eval_wer": 0.16455422416403512,
"step": 2070
},
{
"epoch": 30.14,
"learning_rate": 0.000122955223880597,
"loss": 0.1586,
"step": 2080
},
{
"epoch": 30.29,
"learning_rate": 0.00012205970149253731,
"loss": 0.3666,
"step": 2090
},
{
"epoch": 30.43,
"learning_rate": 0.0001211641791044776,
"loss": 0.2325,
"step": 2100
},
{
"epoch": 30.57,
"learning_rate": 0.0001202686567164179,
"loss": 0.2943,
"step": 2110
},
{
"epoch": 30.72,
"learning_rate": 0.00011937313432835821,
"loss": 0.3407,
"step": 2120
},
{
"epoch": 30.86,
"learning_rate": 0.0001184776119402985,
"loss": 0.2466,
"step": 2130
},
{
"epoch": 30.99,
"eval_loss": 0.5013484954833984,
"eval_runtime": 64.0675,
"eval_samples_per_second": 8.756,
"eval_steps_per_second": 2.201,
"eval_wer": 0.1528312497296362,
"step": 2139
},
{
"epoch": 31.01,
"learning_rate": 0.00011758208955223881,
"loss": 0.4194,
"step": 2140
},
{
"epoch": 31.16,
"learning_rate": 0.0001166865671641791,
"loss": 0.1414,
"step": 2150
},
{
"epoch": 31.3,
"learning_rate": 0.00011579104477611939,
"loss": 0.3346,
"step": 2160
},
{
"epoch": 31.44,
"learning_rate": 0.00011489552238805968,
"loss": 0.2872,
"step": 2170
},
{
"epoch": 31.59,
"learning_rate": 0.00011399999999999999,
"loss": 0.3151,
"step": 2180
},
{
"epoch": 31.73,
"learning_rate": 0.00011310447761194028,
"loss": 0.3952,
"step": 2190
},
{
"epoch": 31.87,
"learning_rate": 0.00011220895522388059,
"loss": 0.1017,
"step": 2200
},
{
"epoch": 31.99,
"eval_loss": 0.45783624053001404,
"eval_runtime": 64.1463,
"eval_samples_per_second": 8.746,
"eval_steps_per_second": 2.198,
"eval_wer": 0.155210451183112,
"step": 2208
},
{
"epoch": 32.03,
"learning_rate": 0.00011131343283582089,
"loss": 0.4051,
"step": 2210
},
{
"epoch": 32.17,
"learning_rate": 0.00011041791044776118,
"loss": 0.1706,
"step": 2220
},
{
"epoch": 32.32,
"learning_rate": 0.00010952238805970149,
"loss": 0.3782,
"step": 2230
},
{
"epoch": 32.46,
"learning_rate": 0.00010862686567164178,
"loss": 0.2823,
"step": 2240
},
{
"epoch": 32.6,
"learning_rate": 0.00010773134328358208,
"loss": 0.2133,
"step": 2250
},
{
"epoch": 32.75,
"learning_rate": 0.00010683582089552238,
"loss": 0.3748,
"step": 2260
},
{
"epoch": 32.89,
"learning_rate": 0.00010594029850746268,
"loss": 0.1383,
"step": 2270
},
{
"epoch": 32.99,
"eval_loss": 0.5026096105575562,
"eval_runtime": 64.8383,
"eval_samples_per_second": 8.652,
"eval_steps_per_second": 2.175,
"eval_wer": 0.14188692304364753,
"step": 2277
},
{
"epoch": 33.04,
"learning_rate": 0.00010504477611940296,
"loss": 0.4434,
"step": 2280
},
{
"epoch": 33.19,
"learning_rate": 0.00010414925373134328,
"loss": 0.1895,
"step": 2290
},
{
"epoch": 33.33,
"learning_rate": 0.00010325373134328358,
"loss": 0.2909,
"step": 2300
},
{
"epoch": 33.47,
"learning_rate": 0.00010235820895522386,
"loss": 0.3125,
"step": 2310
},
{
"epoch": 33.62,
"learning_rate": 0.00010146268656716418,
"loss": 0.1565,
"step": 2320
},
{
"epoch": 33.76,
"learning_rate": 0.00010056716417910446,
"loss": 0.5154,
"step": 2330
},
{
"epoch": 33.9,
"learning_rate": 9.967164179104478e-05,
"loss": 0.1719,
"step": 2340
},
{
"epoch": 33.99,
"eval_loss": 0.46511512994766235,
"eval_runtime": 64.671,
"eval_samples_per_second": 8.675,
"eval_steps_per_second": 2.18,
"eval_wer": 0.14422286628887832,
"step": 2346
},
{
"epoch": 34.06,
"learning_rate": 9.877611940298506e-05,
"loss": 0.4138,
"step": 2350
},
{
"epoch": 34.2,
"learning_rate": 9.788059701492536e-05,
"loss": 0.2165,
"step": 2360
},
{
"epoch": 34.34,
"learning_rate": 9.698507462686566e-05,
"loss": 0.2969,
"step": 2370
},
{
"epoch": 34.49,
"learning_rate": 9.608955223880596e-05,
"loss": 0.4197,
"step": 2380
},
{
"epoch": 34.63,
"learning_rate": 9.519402985074626e-05,
"loss": 0.1103,
"step": 2390
},
{
"epoch": 34.77,
"learning_rate": 9.429850746268656e-05,
"loss": 0.3708,
"step": 2400
},
{
"epoch": 34.92,
"learning_rate": 9.340298507462686e-05,
"loss": 0.1808,
"step": 2410
},
{
"epoch": 34.99,
"eval_loss": 0.4499177634716034,
"eval_runtime": 64.5038,
"eval_samples_per_second": 8.697,
"eval_steps_per_second": 2.186,
"eval_wer": 0.14123804991997232,
"step": 2415
},
{
"epoch": 35.07,
"learning_rate": 9.250746268656715e-05,
"loss": 0.3266,
"step": 2420
},
{
"epoch": 35.22,
"learning_rate": 9.161194029850746e-05,
"loss": 0.2852,
"step": 2430
},
{
"epoch": 35.36,
"learning_rate": 9.071641791044775e-05,
"loss": 0.294,
"step": 2440
},
{
"epoch": 35.5,
"learning_rate": 8.982089552238805e-05,
"loss": 0.4613,
"step": 2450
},
{
"epoch": 35.65,
"learning_rate": 8.892537313432835e-05,
"loss": 0.0998,
"step": 2460
},
{
"epoch": 35.79,
"learning_rate": 8.802985074626865e-05,
"loss": 0.3487,
"step": 2470
},
{
"epoch": 35.93,
"learning_rate": 8.713432835820896e-05,
"loss": 0.2429,
"step": 2480
},
{
"epoch": 35.99,
"eval_loss": 0.4522528052330017,
"eval_runtime": 64.4305,
"eval_samples_per_second": 8.707,
"eval_steps_per_second": 2.188,
"eval_wer": 0.1471644244495393,
"step": 2484
},
{
"epoch": 36.09,
"learning_rate": 8.623880597014925e-05,
"loss": 0.3206,
"step": 2490
},
{
"epoch": 36.23,
"learning_rate": 8.534328358208955e-05,
"loss": 0.2824,
"step": 2500
},
{
"epoch": 36.37,
"learning_rate": 8.444776119402985e-05,
"loss": 0.156,
"step": 2510
},
{
"epoch": 36.52,
"learning_rate": 8.355223880597015e-05,
"loss": 0.4632,
"step": 2520
},
{
"epoch": 36.66,
"learning_rate": 8.265671641791043e-05,
"loss": 0.2011,
"step": 2530
},
{
"epoch": 36.8,
"learning_rate": 8.176119402985074e-05,
"loss": 0.3352,
"step": 2540
},
{
"epoch": 36.95,
"learning_rate": 8.086567164179103e-05,
"loss": 0.2651,
"step": 2550
},
{
"epoch": 36.99,
"eval_loss": 0.454355388879776,
"eval_runtime": 63.0194,
"eval_samples_per_second": 8.902,
"eval_steps_per_second": 2.237,
"eval_wer": 0.1397240126313968,
"step": 2553
},
{
"epoch": 37.1,
"learning_rate": 7.997014925373133e-05,
"loss": 0.2746,
"step": 2560
},
{
"epoch": 37.24,
"learning_rate": 7.907462686567164e-05,
"loss": 0.4286,
"step": 2570
},
{
"epoch": 37.39,
"learning_rate": 7.817910447761193e-05,
"loss": 0.1063,
"step": 2580
},
{
"epoch": 37.53,
"learning_rate": 7.728358208955224e-05,
"loss": 0.4127,
"step": 2590
},
{
"epoch": 37.67,
"learning_rate": 7.638805970149253e-05,
"loss": 0.1612,
"step": 2600
},
{
"epoch": 37.82,
"learning_rate": 7.549253731343283e-05,
"loss": 0.3096,
"step": 2610
},
{
"epoch": 37.96,
"learning_rate": 7.459701492537313e-05,
"loss": 0.2748,
"step": 2620
},
{
"epoch": 37.99,
"eval_loss": 0.41805797815322876,
"eval_runtime": 63.0354,
"eval_samples_per_second": 8.9,
"eval_steps_per_second": 2.237,
"eval_wer": 0.1385560410087814,
"step": 2622
},
{
"epoch": 38.11,
"learning_rate": 7.370149253731343e-05,
"loss": 0.2088,
"step": 2630
},
{
"epoch": 38.26,
"learning_rate": 7.280597014925373e-05,
"loss": 0.344,
"step": 2640
},
{
"epoch": 38.4,
"learning_rate": 7.191044776119403e-05,
"loss": 0.1293,
"step": 2650
},
{
"epoch": 38.54,
"learning_rate": 7.101492537313432e-05,
"loss": 0.2947,
"step": 2660
},
{
"epoch": 38.69,
"learning_rate": 7.011940298507462e-05,
"loss": 0.2007,
"step": 2670
},
{
"epoch": 38.83,
"learning_rate": 6.922388059701491e-05,
"loss": 0.2795,
"step": 2680
},
{
"epoch": 38.97,
"learning_rate": 6.832835820895521e-05,
"loss": 0.4171,
"step": 2690
},
{
"epoch": 38.99,
"eval_loss": 0.43853336572647095,
"eval_runtime": 63.2599,
"eval_samples_per_second": 8.868,
"eval_steps_per_second": 2.229,
"eval_wer": 0.1334083142276247,
"step": 2691
},
{
"epoch": 39.13,
"learning_rate": 6.743283582089551e-05,
"loss": 0.1256,
"step": 2700
},
{
"epoch": 39.27,
"learning_rate": 6.653731343283581e-05,
"loss": 0.383,
"step": 2710
},
{
"epoch": 39.42,
"learning_rate": 6.564179104477611e-05,
"loss": 0.1139,
"step": 2720
},
{
"epoch": 39.56,
"learning_rate": 6.474626865671641e-05,
"loss": 0.2979,
"step": 2730
},
{
"epoch": 39.7,
"learning_rate": 6.385074626865671e-05,
"loss": 0.2812,
"step": 2740
},
{
"epoch": 39.85,
"learning_rate": 6.295522388059701e-05,
"loss": 0.251,
"step": 2750
},
{
"epoch": 39.99,
"learning_rate": 6.20597014925373e-05,
"loss": 0.4119,
"step": 2760
},
{
"epoch": 39.99,
"eval_loss": 0.45682770013809204,
"eval_runtime": 63.015,
"eval_samples_per_second": 8.903,
"eval_steps_per_second": 2.238,
"eval_wer": 0.15040879006791538,
"step": 2760
},
{
"epoch": 40.14,
"learning_rate": 6.11641791044776e-05,
"loss": 0.1252,
"step": 2770
},
{
"epoch": 40.29,
"learning_rate": 6.02686567164179e-05,
"loss": 0.3647,
"step": 2780
},
{
"epoch": 40.43,
"learning_rate": 5.9373134328358206e-05,
"loss": 0.2004,
"step": 2790
},
{
"epoch": 40.57,
"learning_rate": 5.8477611940298506e-05,
"loss": 0.2465,
"step": 2800
},
{
"epoch": 40.72,
"learning_rate": 5.7582089552238806e-05,
"loss": 0.3278,
"step": 2810
},
{
"epoch": 40.86,
"learning_rate": 5.66865671641791e-05,
"loss": 0.1453,
"step": 2820
},
{
"epoch": 40.99,
"eval_loss": 0.44247445464134216,
"eval_runtime": 63.0109,
"eval_samples_per_second": 8.903,
"eval_steps_per_second": 2.238,
"eval_wer": 0.14309815287450794,
"step": 2829
},
{
"epoch": 41.01,
"learning_rate": 5.57910447761194e-05,
"loss": 0.4307,
"step": 2830
},
{
"epoch": 41.16,
"learning_rate": 5.48955223880597e-05,
"loss": 0.1089,
"step": 2840
},
{
"epoch": 41.3,
"learning_rate": 5.399999999999999e-05,
"loss": 0.3548,
"step": 2850
},
{
"epoch": 41.44,
"learning_rate": 5.310447761194029e-05,
"loss": 0.197,
"step": 2860
},
{
"epoch": 41.59,
"learning_rate": 5.220895522388059e-05,
"loss": 0.2719,
"step": 2870
},
{
"epoch": 41.73,
"learning_rate": 5.13134328358209e-05,
"loss": 0.4077,
"step": 2880
},
{
"epoch": 41.87,
"learning_rate": 5.041791044776119e-05,
"loss": 0.105,
"step": 2890
},
{
"epoch": 41.99,
"eval_loss": 0.43674367666244507,
"eval_runtime": 64.6292,
"eval_samples_per_second": 8.68,
"eval_steps_per_second": 2.182,
"eval_wer": 0.13526841718216032,
"step": 2898
},
{
"epoch": 42.03,
"learning_rate": 4.952238805970149e-05,
"loss": 0.3982,
"step": 2900
},
{
"epoch": 42.17,
"learning_rate": 4.862686567164179e-05,
"loss": 0.1811,
"step": 2910
},
{
"epoch": 42.32,
"learning_rate": 4.773134328358208e-05,
"loss": 0.3297,
"step": 2920
},
{
"epoch": 42.46,
"learning_rate": 4.683582089552238e-05,
"loss": 0.2095,
"step": 2930
},
{
"epoch": 42.6,
"learning_rate": 4.594029850746268e-05,
"loss": 0.1968,
"step": 2940
},
{
"epoch": 42.75,
"learning_rate": 4.5044776119402976e-05,
"loss": 0.3637,
"step": 2950
},
{
"epoch": 42.89,
"learning_rate": 4.4149253731343275e-05,
"loss": 0.1205,
"step": 2960
},
{
"epoch": 42.99,
"eval_loss": 0.44175830483436584,
"eval_runtime": 64.641,
"eval_samples_per_second": 8.679,
"eval_steps_per_second": 2.181,
"eval_wer": 0.1340139291430549,
"step": 2967
},
{
"epoch": 43.04,
"learning_rate": 4.325373134328358e-05,
"loss": 0.3979,
"step": 2970
},
{
"epoch": 43.19,
"learning_rate": 4.235820895522388e-05,
"loss": 0.1674,
"step": 2980
},
{
"epoch": 43.33,
"learning_rate": 4.1462686567164175e-05,
"loss": 0.3019,
"step": 2990
},
{
"epoch": 43.47,
"learning_rate": 4.0567164179104475e-05,
"loss": 0.2778,
"step": 3000
},
{
"epoch": 43.62,
"learning_rate": 3.9671641791044774e-05,
"loss": 0.1561,
"step": 3010
},
{
"epoch": 43.76,
"learning_rate": 3.877611940298507e-05,
"loss": 0.3686,
"step": 3020
},
{
"epoch": 43.9,
"learning_rate": 3.788059701492537e-05,
"loss": 0.2039,
"step": 3030
},
{
"epoch": 43.99,
"eval_loss": 0.4585602283477783,
"eval_runtime": 62.823,
"eval_samples_per_second": 8.93,
"eval_steps_per_second": 2.244,
"eval_wer": 0.13786390967686118,
"step": 3036
},
{
"epoch": 44.06,
"learning_rate": 3.698507462686567e-05,
"loss": 0.3361,
"step": 3040
},
{
"epoch": 44.2,
"learning_rate": 3.608955223880597e-05,
"loss": 0.185,
"step": 3050
},
{
"epoch": 44.34,
"learning_rate": 3.519402985074627e-05,
"loss": 0.2579,
"step": 3060
},
{
"epoch": 44.49,
"learning_rate": 3.429850746268656e-05,
"loss": 0.3879,
"step": 3070
},
{
"epoch": 44.63,
"learning_rate": 3.3402985074626866e-05,
"loss": 0.1033,
"step": 3080
},
{
"epoch": 44.77,
"learning_rate": 3.250746268656716e-05,
"loss": 0.3663,
"step": 3090
},
{
"epoch": 44.92,
"learning_rate": 3.161194029850746e-05,
"loss": 0.1773,
"step": 3100
},
{
"epoch": 44.99,
"eval_loss": 0.46862030029296875,
"eval_runtime": 63.8696,
"eval_samples_per_second": 8.784,
"eval_steps_per_second": 2.208,
"eval_wer": 0.1390751395077216,
"step": 3105
},
{
"epoch": 45.07,
"learning_rate": 3.071641791044776e-05,
"loss": 0.3077,
"step": 3110
},
{
"epoch": 45.22,
"learning_rate": 2.9820895522388055e-05,
"loss": 0.2927,
"step": 3120
},
{
"epoch": 45.36,
"learning_rate": 2.892537313432836e-05,
"loss": 0.2373,
"step": 3130
},
{
"epoch": 45.5,
"learning_rate": 2.8029850746268655e-05,
"loss": 0.3392,
"step": 3140
},
{
"epoch": 45.65,
"learning_rate": 2.713432835820895e-05,
"loss": 0.0841,
"step": 3150
},
{
"epoch": 45.79,
"learning_rate": 2.623880597014925e-05,
"loss": 0.3028,
"step": 3160
},
{
"epoch": 45.93,
"learning_rate": 2.534328358208955e-05,
"loss": 0.2186,
"step": 3170
},
{
"epoch": 45.99,
"eval_loss": 0.49753618240356445,
"eval_runtime": 63.1583,
"eval_samples_per_second": 8.882,
"eval_steps_per_second": 2.232,
"eval_wer": 0.14456893195483844,
"step": 3174
},
{
"epoch": 46.09,
"learning_rate": 2.444776119402985e-05,
"loss": 0.2526,
"step": 3180
},
{
"epoch": 46.23,
"learning_rate": 2.3552238805970147e-05,
"loss": 0.3354,
"step": 3190
},
{
"epoch": 46.37,
"learning_rate": 2.2656716417910444e-05,
"loss": 0.1822,
"step": 3200
},
{
"epoch": 46.52,
"learning_rate": 2.1761194029850743e-05,
"loss": 0.3111,
"step": 3210
},
{
"epoch": 46.66,
"learning_rate": 2.0865671641791043e-05,
"loss": 0.1602,
"step": 3220
},
{
"epoch": 46.8,
"learning_rate": 1.9970149253731343e-05,
"loss": 0.3159,
"step": 3230
},
{
"epoch": 46.95,
"learning_rate": 1.907462686567164e-05,
"loss": 0.2358,
"step": 3240
},
{
"epoch": 46.99,
"eval_loss": 0.4886210858821869,
"eval_runtime": 63.1228,
"eval_samples_per_second": 8.887,
"eval_steps_per_second": 2.234,
"eval_wer": 0.14482848120430852,
"step": 3243
},
{
"epoch": 47.1,
"learning_rate": 1.817910447761194e-05,
"loss": 0.2315,
"step": 3250
},
{
"epoch": 47.24,
"learning_rate": 1.7283582089552236e-05,
"loss": 0.4232,
"step": 3260
},
{
"epoch": 47.39,
"learning_rate": 1.6388059701492535e-05,
"loss": 0.0988,
"step": 3270
},
{
"epoch": 47.53,
"learning_rate": 1.5492537313432835e-05,
"loss": 0.4457,
"step": 3280
},
{
"epoch": 47.67,
"learning_rate": 1.4597014925373132e-05,
"loss": 0.1544,
"step": 3290
},
{
"epoch": 47.82,
"learning_rate": 1.3701492537313431e-05,
"loss": 0.2426,
"step": 3300
},
{
"epoch": 47.96,
"learning_rate": 1.280597014925373e-05,
"loss": 0.3525,
"step": 3310
},
{
"epoch": 47.99,
"eval_loss": 0.47062501311302185,
"eval_runtime": 62.9566,
"eval_samples_per_second": 8.911,
"eval_steps_per_second": 2.24,
"eval_wer": 0.13981052904788685,
"step": 3312
},
{
"epoch": 48.11,
"learning_rate": 1.191044776119403e-05,
"loss": 0.255,
"step": 3320
},
{
"epoch": 48.26,
"learning_rate": 1.1014925373134327e-05,
"loss": 0.3479,
"step": 3330
},
{
"epoch": 48.4,
"learning_rate": 1.0119402985074627e-05,
"loss": 0.1079,
"step": 3340
},
{
"epoch": 48.54,
"learning_rate": 9.223880597014925e-06,
"loss": 0.3069,
"step": 3350
},
{
"epoch": 48.69,
"learning_rate": 8.328358208955223e-06,
"loss": 0.2212,
"step": 3360
},
{
"epoch": 48.83,
"learning_rate": 7.4328358208955215e-06,
"loss": 0.2339,
"step": 3370
},
{
"epoch": 48.97,
"learning_rate": 6.5373134328358204e-06,
"loss": 0.3713,
"step": 3380
},
{
"epoch": 48.99,
"eval_loss": 0.47125667333602905,
"eval_runtime": 62.5816,
"eval_samples_per_second": 8.964,
"eval_steps_per_second": 2.253,
"eval_wer": 0.1387723320500065,
"step": 3381
},
{
"epoch": 49.13,
"learning_rate": 5.641791044776119e-06,
"loss": 0.1483,
"step": 3390
},
{
"epoch": 49.27,
"learning_rate": 4.746268656716418e-06,
"loss": 0.3856,
"step": 3400
},
{
"epoch": 49.42,
"learning_rate": 3.8507462686567165e-06,
"loss": 0.1105,
"step": 3410
},
{
"epoch": 49.56,
"learning_rate": 2.955223880597015e-06,
"loss": 0.2724,
"step": 3420
},
{
"epoch": 49.7,
"learning_rate": 2.0597014925373135e-06,
"loss": 0.2536,
"step": 3430
},
{
"epoch": 49.85,
"learning_rate": 1.1641791044776118e-06,
"loss": 0.2202,
"step": 3440
},
{
"epoch": 49.99,
"learning_rate": 2.686567164179104e-07,
"loss": 0.3543,
"step": 3450
},
{
"epoch": 49.99,
"eval_loss": 0.47204339504241943,
"eval_runtime": 63.0724,
"eval_samples_per_second": 8.895,
"eval_steps_per_second": 2.236,
"eval_wer": 0.1388155902582515,
"step": 3450
},
{
"epoch": 49.99,
"step": 3450,
"total_flos": 8.165293514893254e+18,
"train_loss": 0.35037224591642185,
"train_runtime": 11843.2621,
"train_samples_per_second": 4.712,
"train_steps_per_second": 0.291
}
],
"max_steps": 3450,
"num_train_epochs": 50,
"total_flos": 8.165293514893254e+18,
"trial_name": null,
"trial_params": null
}