metadata
license: bsd-3-clause
tags:
- generated_from_trainer
metrics:
- accuracy
model-index:
- name: ast_20-finetuned-ICBHI
results: []
ast_20-finetuned-ICBHI
This model is a fine-tuned version of MIT/ast-finetuned-audioset-10-10-0.4593 on the None dataset. It achieves the following results on the evaluation set:
- Loss: 2.3414
- Accuracy: 0.6942
- Sensitivity: 0.5361
- Specificity: 0.8354
- Score: 0.6857
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: 4
- eval_batch_size: 4
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 16
- 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
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | Sensitivity | Specificity | Score |
|---|---|---|---|---|---|---|---|
| 0.895 | 1.0 | 345 | 0.9921 | 0.6029 | 0.2458 | 0.9218 | 0.5838 |
| 0.8244 | 2.0 | 690 | 0.8522 | 0.6717 | 0.4147 | 0.9012 | 0.6580 |
| 0.7002 | 3.0 | 1035 | 0.8378 | 0.6775 | 0.5346 | 0.8052 | 0.6699 |
| 0.415 | 4.0 | 1380 | 1.0645 | 0.6674 | 0.5392 | 0.7819 | 0.6605 |
| 0.19 | 5.0 | 1725 | 1.3827 | 0.6732 | 0.5207 | 0.8093 | 0.6650 |
| 0.0465 | 6.0 | 2070 | 1.7785 | 0.6754 | 0.5346 | 0.8011 | 0.6678 |
| 0.0078 | 7.0 | 2415 | 2.0612 | 0.6819 | 0.6252 | 0.7325 | 0.6789 |
| 0.0003 | 8.0 | 2760 | 2.2956 | 0.6971 | 0.5376 | 0.8395 | 0.6886 |
| 0.0001 | 9.0 | 3105 | 2.3499 | 0.6942 | 0.5192 | 0.8505 | 0.6848 |
| 0.0001 | 10.0 | 3450 | 2.3414 | 0.6942 | 0.5361 | 0.8354 | 0.6857 |
Framework versions
- Transformers 4.29.2
- Pytorch 2.0.1+cu118
- Datasets 2.12.0
- Tokenizers 0.13.3