test_bug3
This model is a fine-tuned version of nguyenvulebinh/wav2vec2-base-vietnamese-250h on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.9438
- Wer: 0.4538
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: 0.0003
- train_batch_size: 4
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 4
- 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: 500
- num_epochs: 2
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 1.0443 | 0.15 | 10 | 1.0029 | 0.4537 |
| 1.1815 | 0.3 | 20 | 0.9818 | 0.4538 |
| 1.0803 | 0.45 | 30 | 0.9835 | 0.4473 |
| 1.1333 | 0.6 | 40 | 0.9826 | 0.4517 |
| 1.1544 | 0.75 | 50 | 0.9580 | 0.4504 |
| 0.9911 | 0.91 | 60 | 1.0062 | 0.4416 |
| 1.2515 | 1.06 | 70 | 0.9912 | 0.4328 |
| 1.0998 | 1.21 | 80 | 0.9526 | 0.4500 |
| 1.0671 | 1.36 | 90 | 0.9825 | 0.4298 |
| 1.1107 | 1.51 | 100 | 0.9602 | 0.4473 |
| 0.9271 | 1.66 | 110 | 0.9749 | 0.4331 |
| 1.0792 | 1.82 | 120 | 0.9684 | 0.4273 |
| 1.1438 | 1.97 | 130 | 0.9438 | 0.4538 |
Framework versions
- Transformers 4.16.0
- Pytorch 1.13.1+cu116
- Datasets 1.18.3
- Tokenizers 0.12.1