ast_11-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: 0.9312
- Accuracy: 0.6609
- Sensitivity: 0.4413
- Specificity: 0.8572
- Score: 0.6493
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: 1e-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: 5
Training results
Training Loss | Epoch | Step | Validation Loss | Accuracy | Sensitivity | Specificity | Score |
---|---|---|---|---|---|---|---|
0.9209 | 1.0 | 258 | 1.0106 | 0.5670 | 0.5925 | 0.5443 | 0.5684 |
0.7852 | 2.0 | 517 | 0.8787 | 0.6355 | 0.4252 | 0.8236 | 0.6244 |
0.6614 | 3.0 | 776 | 0.9160 | 0.6322 | 0.5656 | 0.6918 | 0.6287 |
0.3686 | 4.0 | 1035 | 0.9312 | 0.6609 | 0.4413 | 0.8572 | 0.6493 |
0.1924 | 4.99 | 1290 | 0.9663 | 0.6576 | 0.5012 | 0.7975 | 0.6493 |
Framework versions
- Transformers 4.29.2
- Pytorch 2.0.0+cu118
- Datasets 2.12.0
- Tokenizers 0.13.3
- Downloads last month
- 8
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.