metadata
license: apache-2.0
tags:
- generated_from_trainer
datasets:
- marsyas/gtzan
metrics:
- accuracy
model-index:
- name: distilhubert-finetuned-gtzan-3
results: []
distilhubert-finetuned-gtzan-3
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: 0.5578
- Accuracy: 0.89
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: 0.0001
- train_batch_size: 16
- eval_batch_size: 16
- 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: 10
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 1.8069 | 1.0 | 57 | 1.7158 | 0.51 |
| 1.3469 | 2.0 | 114 | 1.2925 | 0.64 |
| 0.8341 | 3.0 | 171 | 0.8796 | 0.77 |
| 0.682 | 4.0 | 228 | 0.8847 | 0.69 |
| 0.3931 | 5.0 | 285 | 0.6189 | 0.84 |
| 0.26 | 6.0 | 342 | 0.5124 | 0.85 |
| 0.1744 | 7.0 | 399 | 0.6412 | 0.81 |
| 0.1053 | 8.0 | 456 | 0.6281 | 0.86 |
| 0.0655 | 9.0 | 513 | 0.5340 | 0.89 |
| 0.2067 | 10.0 | 570 | 0.5578 | 0.89 |
Framework versions
- Transformers 4.29.0
- Pytorch 2.0.1
- Datasets 2.12.0
- Tokenizers 0.13.2