distilhubert-finetuned-gtzan
This model is a fine-tuned version of ntu-spml/distilhubert on the GTZAN dataset. It achieves the following results on the evaluation set:
- Loss: 0.9007
- Accuracy: 0.84
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: 4
- total_train_batch_size: 32
- 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: cosine
- lr_scheduler_warmup_steps: 0.1
- num_epochs: 15
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 9.1860 | 1.0 | 29 | 2.2605 | 0.19 |
| 8.8674 | 2.0 | 58 | 2.0993 | 0.45 |
| 7.8772 | 3.0 | 87 | 1.8032 | 0.57 |
| 6.7326 | 4.0 | 116 | 1.5547 | 0.59 |
| 6.0456 | 5.0 | 145 | 1.3740 | 0.74 |
| 5.5785 | 6.0 | 174 | 1.2371 | 0.76 |
| 4.9885 | 7.0 | 203 | 1.1366 | 0.76 |
| 4.4276 | 8.0 | 232 | 1.0674 | 0.8 |
| 4.2147 | 9.0 | 261 | 1.0060 | 0.77 |
| 3.9144 | 10.0 | 290 | 0.9590 | 0.82 |
| 3.8930 | 11.0 | 319 | 0.9308 | 0.83 |
| 3.6991 | 12.0 | 348 | 0.9125 | 0.83 |
| 3.8139 | 13.0 | 377 | 0.9073 | 0.82 |
| 3.6403 | 14.0 | 406 | 0.9007 | 0.84 |
| 3.6389 | 15.0 | 435 | 0.9002 | 0.84 |
Framework versions
- Transformers 5.16.1
- Pytorch 2.11.0+cu128
- Datasets 4.0.0
- Tokenizers 0.23.1