wav2vec2-base-SER-fine-tuned
This model is a fine-tuned version of facebook/wav2vec2-base on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.4281
- Accuracy: 0.8924
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: 2e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- 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
- num_epochs: 10
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| No log | 1.0 | 144 | 1.7519 | 0.3438 |
| No log | 2.0 | 288 | 1.4404 | 0.5660 |
| No log | 3.0 | 432 | 1.1102 | 0.6701 |
| 1.5063 | 4.0 | 576 | 0.8810 | 0.7917 |
| 1.5063 | 5.0 | 720 | 0.6776 | 0.8507 |
| 1.5063 | 6.0 | 864 | 0.5373 | 0.8785 |
| 0.5402 | 7.0 | 1008 | 0.5459 | 0.8646 |
| 0.5402 | 8.0 | 1152 | 0.5115 | 0.8681 |
| 0.5402 | 9.0 | 1296 | 0.4205 | 0.8889 |
| 0.5402 | 10.0 | 1440 | 0.4281 | 0.8924 |
Framework versions
- Transformers 5.0.0
- Pytorch 2.10.0+cu128
- Datasets 4.0.0
- Tokenizers 0.22.2