wav2vec2-base-finetuned-eos_poc5_ge-di-v7-blanced
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: 0.9969
- Accuracy: 0.5989
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: 64
- eval_batch_size: 64
- seed: 42
- optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 10
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 0.6719 | 1.0 | 452 | 0.6673 | 0.5945 |
| 0.673 | 2.0 | 904 | 0.6624 | 0.5964 |
| 0.6596 | 3.0 | 1356 | 0.6597 | 0.6124 |
| 0.5873 | 4.0 | 1808 | 0.6846 | 0.6071 |
| 0.5624 | 5.0 | 2260 | 0.7225 | 0.6143 |
| 0.5454 | 6.0 | 2712 | 0.7425 | 0.6062 |
| 0.4816 | 7.0 | 3164 | 0.8115 | 0.6018 |
| 0.3987 | 8.0 | 3616 | 0.8764 | 0.6008 |
| 0.3613 | 9.0 | 4068 | 0.9408 | 0.5958 |
| 0.2902 | 10.0 | 4520 | 0.9969 | 0.5989 |
Framework versions
- Transformers 4.47.1
- Pytorch 2.5.1+cu121
- Datasets 3.2.0
- Tokenizers 0.21.0