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: 1.0379
- Accuracy: 0.81
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: 5e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 20
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 2.0307 | 1.0 | 113 | 2.0561 | 0.41 |
| 1.4208 | 2.0 | 226 | 1.4850 | 0.63 |
| 1.1959 | 3.0 | 339 | 1.0617 | 0.66 |
| 0.6929 | 4.0 | 452 | 0.8228 | 0.74 |
| 0.5104 | 5.0 | 565 | 0.6969 | 0.77 |
| 0.4735 | 6.0 | 678 | 0.7412 | 0.79 |
| 0.2185 | 7.0 | 791 | 0.6586 | 0.76 |
| 0.3087 | 8.0 | 904 | 0.8234 | 0.78 |
| 0.1066 | 9.0 | 1017 | 0.8210 | 0.8 |
| 0.0841 | 10.0 | 1130 | 1.0040 | 0.8 |
| 0.0387 | 11.0 | 1243 | 0.9195 | 0.81 |
| 0.0091 | 12.0 | 1356 | 0.9208 | 0.82 |
| 0.006 | 13.0 | 1469 | 0.9190 | 0.81 |
| 0.0051 | 14.0 | 1582 | 0.9796 | 0.8 |
| 0.0038 | 15.0 | 1695 | 0.9823 | 0.8 |
| 0.0035 | 16.0 | 1808 | 1.0252 | 0.8 |
| 0.0032 | 17.0 | 1921 | 1.0172 | 0.8 |
| 0.0032 | 18.0 | 2034 | 1.0433 | 0.81 |
| 0.0029 | 19.0 | 2147 | 1.0577 | 0.81 |
| 0.0029 | 20.0 | 2260 | 1.0379 | 0.81 |
Framework versions
- Transformers 4.30.2
- Pytorch 2.0.1+cu117
- Datasets 2.13.1
- Tokenizers 0.13.3