CoolFace
Modelpublic

SpeechResearch/wtimit-base-normal-all-nofreeze

sourceHugging Faceapache-2.0updated 2y agoView on Hugging Face
0likes7downloads
Model Card

<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->

wtimit-base-normal-all-nofreeze

This model is a fine-tuned version of facebook/wav2vec2-base on the wtimit_asr dataset. It achieves the following results on the evaluation set:

  • —Loss: 0.3190
  • —Wer: 0.0999

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

  • —learning_rate: 0.0001
  • —trainbatchsize: 8
  • —evalbatchsize: 8
  • —seed: 42
  • —optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • —lrschedulertype: linear
  • —lrschedulerwarmup_steps: 1000
  • —num_epochs: 50
  • —mixedprecisiontraining: Native AMP

Training results

Training LossEpochStepValidation LossWer
1.50760.410001.12200.6793
0.41020.8120000.78510.4338
0.22781.2130000.68970.3203
0.17231.6140000.56680.2890
0.14072.0250000.43990.2362
0.1172.4260000.48530.2508
0.0982.8370000.67320.2871
0.08623.2380000.58020.2680
0.08063.6390000.47300.2488
0.07064.04100000.40010.1953
0.0614.44110000.41080.1971
0.0634.84120000.45440.2056
0.05275.25130000.42350.1938
0.0495.65140000.43750.2054
0.04896.06150000.54510.2522
0.04736.46160000.39390.1868
0.04426.86170000.56620.2548
0.04287.27180000.66950.2755
0.03797.67190000.39290.1947
0.03988.07200000.44460.2066
0.03368.48210000.54090.2260
0.03168.88220000.38190.1715
0.03229.29230000.38610.1711
0.03529.69240000.40630.1728
0.031510.09250000.49920.2146
0.025410.5260000.58380.2158
0.024310.9270000.34580.1523
0.024511.3280000.51210.1953
0.023111.71290000.37730.1616
0.020212.11300000.41100.1715
0.026112.52310000.53760.2116
0.024312.92320000.40660.1569
0.020113.32330000.59440.2276
0.021113.73340000.46700.1997
0.024914.13350000.55210.2254
0.02114.53360000.46020.2061
0.016914.94370000.48700.1690
0.018415.34380000.60380.2208
0.020715.74390000.52660.2068
0.020916.15400000.51970.2083
0.017516.55410000.50740.1927
0.016416.96420000.45940.1615
0.016417.36430000.29560.1151
0.014217.76440000.38340.1580
0.013918.17450000.53160.2175
0.018118.57460000.52260.1890
0.015918.97470000.49140.1689
0.012719.38480000.54540.1957
0.013619.78490000.55300.2172
0.012920.19500000.69800.2636
0.013120.59510000.39840.1379
0.012320.99520000.49250.1843
0.009521.4530000.53670.1931
0.012421.8540000.42990.1763
0.011522.2550000.47970.1803
0.013622.61560000.66380.2300
0.012123.01570000.42920.1530
0.009723.42580000.40640.1520
0.014323.82590000.46910.1771
0.009224.22600000.51340.2009
0.009724.63610000.61650.2281
0.007825.03620000.48280.1863
0.011425.43630000.48170.1868
0.008925.84640000.51370.2003
0.008326.24650000.41940.1524
0.0126.65660000.34160.1332
0.010227.05670000.38340.1475
0.007627.45680000.33900.1277
0.008527.86690000.47080.1843
0.007428.26700000.44340.1530
0.007828.66710000.29420.1104
0.007529.07720000.36230.1442
0.006629.47730000.47090.1547
0.007329.87740000.51980.1750
0.005630.28750000.30830.1211
0.006630.68760000.32040.1243
0.004831.09770000.37130.1326
0.004731.49780000.31210.1018
0.006631.89790000.45100.1473
0.005332.3800000.35990.1130
0.005832.7810000.42560.1463
0.005633.1820000.43930.1605
0.004633.51830000.63270.2056
0.004933.91840000.40690.1360
0.003134.32850000.43590.1458
0.005234.72860000.28250.1032
0.003935.12870000.35450.1256
0.00335.53880000.36740.1252
0.00435.93890000.38490.1288
0.002936.33900000.34650.1130
0.00336.74910000.40340.1294
0.003637.14920000.34560.1209
0.003337.55930000.38820.1407
0.003737.95940000.33720.1094
0.002538.35950000.36010.1137
0.003738.76960000.28040.1027
0.002239.16970000.41600.1354
0.002739.56980000.33790.1202
0.00239.97990000.34620.1171
0.002140.371000000.36940.1272
0.001440.781010000.33150.1048
0.002541.181020000.33160.1088
0.00241.581030000.37760.1319
0.002841.991040000.30240.1028
0.001542.391050000.30870.1102
0.001842.791060000.32540.1067
0.002843.21070000.33050.1081
0.00243.61080000.34450.1120
0.001944.01090000.32640.1082
0.001944.411100000.36500.1202
0.00144.811110000.34150.1133
0.001545.221120000.31940.1044
0.001145.621130000.33020.1085
0.001346.021140000.30830.1053
0.000846.431150000.29760.0982
0.001946.831160000.32120.1057
0.000647.231170000.34150.1089
0.002547.641180000.31880.1043
0.000948.041190000.31360.1025
0.001548.451200000.31800.1050
0.001348.851210000.34390.1110
0.000749.251220000.32860.1048
0.001449.661230000.31900.0999

Framework versions

  • —Transformers 4.39.3
  • —Pytorch 2.0.1+cu117
  • —Datasets 2.18.0
  • —Tokenizers 0.15.2