wav2vec2-xls-r-300m-mucs-62255-hinglish_mixed_scripts-0_25-1e-4-epochs-100-FT
This model is a fine-tuned version of facebook/wav2vec2-xls-r-300m on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 2.0453
- Global Wer: 67.7359
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
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 32
- optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: constant_with_warmup
- lr_scheduler_warmup_steps: 500
- num_epochs: 100
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Global Wer |
|---|---|---|---|---|
| 15.4475 | 1.0 | 828 | 1.9936 | 67.5596 |
| 2.2711 | 2.0 | 1656 | 1.3538 | 47.2011 |
| 1.7746 | 3.0 | 2484 | 1.3554 | 42.7641 |
| 1.5320 | 4.0 | 3312 | 1.5898 | 52.1475 |
| 3.0817 | 5.0 | 4140 | 2.4209 | 90.9006 |
| 3.9908 | 6.0 | 4968 | 2.0453 | 67.7359 |
Framework versions
- Transformers 5.13.0
- Pytorch 2.6.0+cu124
- Datasets 3.6.0
- Tokenizers 0.22.2