wav2vec2-large-xlsr-53-fourth-train
This model is a fine-tuned version of fathyy/wav2vec2-large-xlsr-53-third-train on the None dataset. It achieves the following results on the evaluation set:
- Loss: 1.3589
- Wer: 0.7930
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: 0.001
- train_batch_size: 8
- eval_batch_size: 16
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 1000
- num_epochs: 20
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 2.0099 | 2.6667 | 500 | 1.8173 | 0.9204 |
| 1.9544 | 5.3333 | 1000 | 2.1092 | 0.9331 |
| 1.8663 | 8.0 | 1500 | 1.8031 | 0.9236 |
| 1.6086 | 10.6667 | 2000 | 1.7714 | 0.9045 |
| 1.3908 | 13.3333 | 2500 | 1.4747 | 0.8376 |
| 1.1403 | 16.0 | 3000 | 1.4635 | 0.8217 |
| 0.9711 | 18.6667 | 3500 | 1.3589 | 0.7930 |
Framework versions
- Transformers 4.45.2
- Pytorch 2.4.0
- Datasets 3.0.1
- Tokenizers 0.20.0