<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->
wav2vec2_100k_15s_augmentation_gtzan_model
This model is a fine-tuned version of facebook/wav2vec2-base-100k-voxpopuli on the None dataset. It achieves the following results on the evaluation set:
- Loss: 1.4033
- Accuracy: 0.7875
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: 12
- eval_batch_size: 12
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 48
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 15
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 2.2977 | 0.97 | 26 | 2.2767 | 0.3438 |
| 2.2138 | 1.98 | 53 | 2.1260 | 0.5188 |
| 2.072 | 2.99 | 80 | 2.0131 | 0.5312 |
| 1.9904 | 4.0 | 107 | 1.8849 | 0.65 |
| 1.8738 | 4.97 | 133 | 1.7957 | 0.6562 |
| 1.7667 | 5.98 | 160 | 1.7197 | 0.6937 |
| 1.6906 | 6.99 | 187 | 1.6282 | 0.7438 |
| 1.6224 | 8.0 | 214 | 1.6627 | 0.6312 |
| 1.5297 | 8.97 | 240 | 1.5530 | 0.7438 |
| 1.551 | 9.98 | 267 | 1.5321 | 0.725 |
| 1.4737 | 10.99 | 294 | 1.4636 | 0.775 |
| 1.4363 | 12.0 | 321 | 1.4458 | 0.775 |
| 1.4135 | 12.97 | 347 | 1.4127 | 0.7875 |
| 1.3954 | 13.98 | 374 | 1.4045 | 0.7688 |
| 1.3696 | 14.58 | 390 | 1.4033 | 0.7875 |
Framework versions
- Transformers 4.31.0.dev0
- Pytorch 2.0.1+cu118
- Datasets 2.13.1
- Tokenizers 0.13.3
