ast-finetuned-audioset-10-10-0.4593-finetuned-gtzan
This model is a fine-tuned version of MIT/ast-finetuned-audioset-10-10-0.4593 on the gtzan dataset. It achieves the following results on the evaluation set:
- Loss: 0.4380
- 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: 5e-05
- train_batch_size: 2
- eval_batch_size: 2
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 8
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 25
Training results
Training Loss | Epoch | Step | Validation Loss | Accuracy |
---|---|---|---|---|
0.5602 | 1.0 | 112 | 0.4551 | 0.88 |
0.4207 | 2.0 | 225 | 0.4847 | 0.81 |
0.4511 | 3.0 | 337 | 0.7526 | 0.79 |
0.5696 | 4.0 | 450 | 0.6516 | 0.84 |
0.0598 | 5.0 | 562 | 0.5568 | 0.87 |
0.0127 | 6.0 | 675 | 0.9409 | 0.82 |
0.1071 | 7.0 | 787 | 0.5882 | 0.87 |
0.0023 | 8.0 | 900 | 0.5872 | 0.89 |
0.2358 | 9.0 | 1012 | 0.4856 | 0.87 |
0.0002 | 10.0 | 1125 | 0.4762 | 0.87 |
0.0001 | 11.0 | 1237 | 0.4256 | 0.89 |
0.0001 | 12.0 | 1350 | 0.4722 | 0.88 |
0.0 | 13.0 | 1462 | 0.4399 | 0.88 |
0.0001 | 14.0 | 1575 | 0.4401 | 0.88 |
0.0 | 15.0 | 1687 | 0.4394 | 0.88 |
0.0 | 16.0 | 1800 | 0.4437 | 0.88 |
0.0 | 17.0 | 1912 | 0.4393 | 0.89 |
0.0 | 18.0 | 2025 | 0.4379 | 0.89 |
0.0 | 19.0 | 2137 | 0.4383 | 0.88 |
0.0 | 20.0 | 2250 | 0.4390 | 0.88 |
0.0 | 21.0 | 2362 | 0.4382 | 0.89 |
0.0 | 22.0 | 2475 | 0.4384 | 0.89 |
0.0 | 23.0 | 2587 | 0.4375 | 0.89 |
0.0 | 24.0 | 2700 | 0.4375 | 0.89 |
0.0 | 24.89 | 2800 | 0.4380 | 0.89 |
Framework versions
- Transformers 4.30.2
- Pytorch 2.0.1+cu118
- Datasets 2.13.0
- Tokenizers 0.13.3
- Downloads last month
- 10
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social
visibility and check back later, or deploy to Inference Endpoints (dedicated)
instead.