YAML Metadata
Warning:
empty or missing yaml metadata in repo card
(https://huggingface.co/docs/hub/model-cards#model-card-metadata)
wavlm-large_finetuned_RAVDESS
This model is a fine-tuned version of microsoft/wavlm-large on the Ryerson Audio-Visual Database of Emotional Speech and Song (RAVDESS). It achieves the following results on the evaluation set:
- Loss: 0.3534
- Accuracy: 0.9028
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: 0.0003
- train_batch_size: 32
- eval_batch_size: 32
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 128
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 20
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| No log | 1.0 | 9 | 2.0485 | 0.2326 |
| 2.0712 | 2.0 | 18 | 1.8028 | 0.2917 |
| 1.9355 | 3.0 | 27 | 1.7300 | 0.3229 |
| 1.7116 | 4.0 | 36 | 1.3749 | 0.4722 |
| 1.4907 | 5.0 | 45 | 1.0586 | 0.6493 |
| 1.1558 | 6.0 | 54 | 0.8834 | 0.6771 |
| 0.8621 | 7.0 | 63 | 0.9206 | 0.6944 |
| 0.6437 | 8.0 | 72 | 0.5895 | 0.8194 |
| 0.4634 | 9.0 | 81 | 0.7389 | 0.7743 |
| 0.3974 | 10.0 | 90 | 0.4569 | 0.8542 |
| 0.3974 | 11.0 | 99 | 0.5140 | 0.8438 |
| 0.3105 | 12.0 | 108 | 0.4273 | 0.8611 |
| 0.2094 | 13.0 | 117 | 0.3608 | 0.8993 |
| 0.1401 | 14.0 | 126 | 0.5715 | 0.8194 |
| 0.1249 | 15.0 | 135 | 0.3715 | 0.8854 |
| 0.0953 | 16.0 | 144 | 0.4112 | 0.8785 |
| 0.0955 | 17.0 | 153 | 0.3692 | 0.9062 |
| 0.0807 | 18.0 | 162 | 0.4395 | 0.8646 |
| 0.1077 | 19.0 | 171 | 0.3413 | 0.9201 |
| 0.0578 | 20.0 | 180 | 0.3534 | 0.9028 |
Framework versions
- Transformers 4.45.2
- Pytorch 2.5.0+cu118
- Datasets 3.0.1
- Tokenizers 0.20.1
- Downloads last month
- 63