CoolFace
Modelpublic

KYAGABA/wav2vec2-large-Rw-cv-50hr-v10

sourceHugging Faceapache-2.0updated 2y agoView on Hugging Face
0likes11downloads
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. -->

wav2vec2-large-Rw-cv-50hr-v10

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

  • —Loss: 0.7559
  • —Model Preparation Time: 0.0077
  • —Wer: 0.3941
  • —Cer: 0.1108

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: 5e-05
  • —trainbatchsize: 16
  • —evalbatchsize: 8
  • —seed: 42
  • —gradientaccumulationsteps: 2
  • —totaltrainbatch_size: 32
  • —optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • —lrschedulertype: linear
  • —lrschedulerwarmup_ratio: 0.05
  • —num_epochs: 100
  • —mixedprecisiontraining: Native AMP

Training results

Training LossEpochStepValidation LossModel Preparation TimeWerCer
3.52851.010961.35450.00770.91320.3520
0.83762.021921.08040.00770.83680.2844
0.64593.032881.01430.00770.75750.2694
0.56134.043840.94510.00770.71240.2476
0.51375.054800.90130.00770.69070.2378
0.46866.065760.87870.00770.66240.2334
0.42687.076720.83580.00770.63960.2209
0.39218.087680.83020.00770.63210.2216
0.36819.098640.82270.00770.62310.2185
0.344810.0109600.80560.00770.60870.2112
0.321911.0120560.80470.00770.60690.2075
0.306212.0131520.78630.00770.59110.2047
0.288613.0142480.81220.00770.58740.2045
0.274414.0153440.80930.00770.57340.1990
0.263115.0164400.81820.00770.59130.2052
0.249216.0175360.82160.00770.57970.2002
0.239217.0186320.80950.00770.57590.2000
0.228518.0197280.80000.00770.56470.1954
0.217719.0208240.77040.00770.55920.1947
0.209220.0219200.81110.00770.55920.1945
0.200621.0230160.79070.00770.56060.1940
0.192122.0241120.80850.00770.57060.2012
0.185123.0252080.81960.00770.55840.1957
0.179224.0263040.83130.00770.56200.1951
0.172525.0274000.83110.00770.55250.1929
0.166326.0284960.82230.00770.55340.1914
0.159127.0295920.83640.00770.55000.1945
0.155328.0306880.83500.00770.55090.1927
0.149329.0317840.84170.00770.54930.1929
0.14230.0328800.81340.00770.54840.1878
0.139431.0339760.86430.00770.54470.1872
0.134532.0350720.88560.00770.53890.1865
0.131433.0361680.85450.00770.55110.1893
0.127734.0372640.93120.00770.54620.1933
0.123635.0383600.85890.00770.53500.1883
0.120936.0394560.89510.00770.53400.1849
0.117137.0405520.85780.00770.53470.1854
0.112738.0416480.87850.00770.53430.1879
0.109139.0427440.86790.00770.53680.1865
0.106440.0438400.89930.00770.54800.1914
0.102341.0449360.90210.00770.54580.1938
0.10242.0460320.93210.00770.54960.1938
0.098543.0471280.93110.00770.53680.1883
0.094944.0482240.94230.00770.53250.1857
0.092345.0493200.91720.00770.53950.1888
0.090246.0504160.92840.00770.53390.1870
0.088747.0515120.92670.00770.53190.1870
0.086348.0526080.92700.00770.52240.1831
0.083649.0537040.96200.00770.53340.1874
0.082550.0548000.96510.00770.52820.1861
0.080251.0558960.96740.00770.52920.1874
0.07852.0569920.98910.00770.53400.1870
0.076153.0580880.96650.00770.51960.1837
0.074854.0591840.97420.00770.51910.1821
0.073955.0602801.00350.00770.52380.1869
0.07256.0613761.00440.00770.52090.1844
0.070557.0624721.01590.00770.52890.1893
0.067458.0635681.01340.00770.52130.1849
0.066259.0646640.97900.00770.51970.1844
0.064660.0657601.05800.00770.51510.1816
0.063761.0668561.02380.00770.51890.1831
0.061762.0679521.01520.00770.52030.1844
0.060663.0690481.03150.00770.51530.1823
0.059464.0701441.07120.00770.52010.1853
0.058565.0712401.07270.00770.51470.1864
0.056866.0723361.06260.00770.51860.1871
0.055567.0734321.12930.00770.51780.1836
0.053868.0745281.16280.00770.51720.1839
0.053169.0756241.11020.00770.52020.1862
0.051870.0767201.14040.00770.52340.1855
0.051671.0778161.09660.00770.51190.1826
0.050172.0789121.08550.00770.51340.1826
0.049173.0800081.13530.00770.51200.1828
0.047874.0811041.15500.00770.51600.1881
0.047375.0822001.17130.00770.51280.1864
0.045176.0832961.14050.00770.52040.1864
0.045777.0843921.14500.00770.51330.1841
0.043778.0854881.12930.00770.51710.1869
0.042679.0865841.16520.00770.51190.1822
0.042580.0876801.16380.00770.51020.1852
0.041381.0887761.14700.00770.50990.1813
0.041482.0898721.17730.00770.51690.1838
0.039483.0909681.20320.00770.51860.1867
0.03984.0920641.21340.00770.51740.1853
0.038285.0931601.21060.00770.51590.1849
0.037186.0942561.25580.00770.51750.1873
0.037787.0953521.25100.00770.51750.1881
0.036388.0964481.24210.00770.51670.1877
0.035589.0975441.22310.00770.51390.1847
0.034590.0986401.24410.00770.51110.1841
0.03391.0997361.22540.00770.51350.1855

Framework versions

  • —Transformers 4.45.2
  • —Pytorch 2.1.0+cu118
  • —Datasets 2.20.0
  • —Tokenizers 0.20.0