<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->
wavlm-basic_s-f-o_8batch_5sec_0.0001lr_unfrozen
This model is a fine-tuned version of microsoft/wavlm-large on the None dataset. It achieves the following results on the evaluation set:
- Loss: 1.3331
- Accuracy: 0.8
- F1: 0.8022
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: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 32
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.003
- num_epochs: 1000
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 |
|---|---|---|---|---|---|
| 2.2668 | 1.0 | 131 | 2.1265 | 0.2833 | 0.1955 |
| 1.3925 | 2.0 | 262 | 1.3972 | 0.4833 | 0.4566 |
| 0.9643 | 2.99 | 393 | 1.0938 | 0.65 | 0.6166 |
| 0.6244 | 4.0 | 525 | 0.9165 | 0.75 | 0.7322 |
| 0.3692 | 5.0 | 656 | 0.9714 | 0.7333 | 0.7207 |
| 0.3284 | 6.0 | 787 | 0.5999 | 0.85 | 0.8515 |
| 0.2096 | 6.99 | 918 | 0.7120 | 0.8 | 0.8005 |
| 0.1974 | 8.0 | 1050 | 0.6648 | 0.8667 | 0.8671 |
| 0.1349 | 9.0 | 1181 | 0.7068 | 0.8333 | 0.8337 |
| 0.221 | 10.0 | 1312 | 1.2068 | 0.7667 | 0.7577 |
| 0.1403 | 10.99 | 1443 | 0.9815 | 0.85 | 0.8480 |
| 0.1211 | 12.0 | 1575 | 1.0742 | 0.7833 | 0.7890 |
| 0.0901 | 13.0 | 1706 | 0.8772 | 0.8167 | 0.8150 |
Framework versions
- Transformers 4.28.1
- Pytorch 2.0.0+cu118
- Datasets 2.12.0
- Tokenizers 0.13.3
