wav2vec2_word_cls
This model is a fine-tuned version of facebook/wav2vec2-base-960h on the None dataset. It achieves the following results on the evaluation set:
- Loss: 1.1277
- Accuracy: 0.8038
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: 3e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 0.1
- num_epochs: 10
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 4.7767 | 1.0 | 59 | 2.1566 | 0.4937 |
| 4.0970 | 2.0 | 118 | 1.8615 | 0.5633 |
| 3.5216 | 3.0 | 177 | 1.6508 | 0.5823 |
| 2.9832 | 4.0 | 236 | 1.4801 | 0.6582 |
| 2.7812 | 5.0 | 295 | 1.3932 | 0.6709 |
| 2.5981 | 6.0 | 354 | 1.3658 | 0.7025 |
| 2.5486 | 7.0 | 413 | 1.2179 | 0.7532 |
| 2.2501 | 8.0 | 472 | 1.1333 | 0.7722 |
| 2.2565 | 9.0 | 531 | 1.1314 | 0.8038 |
| 2.4496 | 10.0 | 590 | 1.1277 | 0.8038 |
Framework versions
- Transformers 5.5.4
- Pytorch 2.10.0+cu128
- Datasets 4.0.0
- Tokenizers 0.22.2