CoolFace
Modelpublic

dianavdavidson/wav2vec2-large-xlsr-hindi-iv_hindi_vaani-62094-hinglish_mixed_scripts-1e-4-epochs-100-FT

sourceHugging Faceapache-2.0updated 29d agoView on Hugging Face
0likes44downloads
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-xlsr-hindi-ivhindivaani-62094-hinglishmixedscripts-1e-4-epochs-100-FT

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

  • —Loss: 0.6632
  • —Global Wer: 26.8737

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: 16
  • —evalbatchsize: 16
  • —seed: 42
  • —gradientaccumulationsteps: 2
  • —totaltrainbatch_size: 32
  • —optimizer: Use adamwtorch with betas=(0.9,0.999) and epsilon=1e-08 and optimizerargs=No additional optimizer arguments
  • —lrschedulertype: constantwithwarmup
  • —lrschedulerwarmup_steps: 500
  • —num_epochs: 100
  • —mixedprecisiontraining: Native AMP

Training results

Training LossEpochStepValidation LossGlobal Wer
11.42421.05401.351070.7884
2.43352.010800.777747.2116
1.74993.016200.663540.3276
1.47254.021600.594336.3447
1.28025.027000.558334.6758
1.15406.032400.546133.0205
1.07237.037800.549932.1980
0.98828.043200.540931.3720
0.91139.048600.531530.8532
0.865510.054000.520729.5392
0.797411.059400.527129.3857
0.753712.064800.546429.4232
0.660813.070200.549728.7133
0.631814.075600.552728.3549
0.578015.081000.548628.2696
0.577816.086400.591928.2048
0.539517.091800.584627.9829
0.504118.097200.599727.7065
0.497719.0102600.614527.8771
0.484920.0108000.586227.6792
0.442121.0113400.627727.6280
0.436822.0118800.614627.5051
0.409523.0124200.649427.7509
0.393024.0129600.645627.2628
0.365725.0135000.676727.2014
0.358026.0140400.618326.9454
0.337727.0145800.658027.0819
0.336728.0151200.641127.0239
0.324129.0156600.679426.8601
0.312430.0162000.672726.7235
0.311331.0167400.685726.7986
0.295332.0172800.685526.8157
0.283133.0178200.663226.8737

Framework versions

  • —Transformers 5.13.0
  • —Pytorch 2.6.0+cu124
  • —Datasets 3.6.0
  • —Tokenizers 0.22.2