dysarthria_classification
This model is a fine-tuned version of facebook/wav2vec2-base on the audiofolder dataset. It achieves the following results on the evaluation set:
- Loss: 1.3926
- Accuracy: 0.175
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: 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: 10
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| No log | 1.0 | 3 | 1.3941 | 0.2 |
| No log | 2.0 | 6 | 1.3945 | 0.2 |
| No log | 3.0 | 9 | 1.3938 | 0.2 |
| 1.384 | 4.0 | 12 | 1.3934 | 0.15 |
| 1.384 | 5.0 | 15 | 1.3934 | 0.15 |
| 1.384 | 6.0 | 18 | 1.3935 | 0.15 |
| 1.3825 | 7.0 | 21 | 1.3930 | 0.15 |
| 1.3825 | 8.0 | 24 | 1.3927 | 0.15 |
| 1.3825 | 9.0 | 27 | 1.3926 | 0.175 |
| 1.3833 | 10.0 | 30 | 1.3926 | 0.175 |
Framework versions
- Transformers 4.35.2
- Pytorch 2.1.0+cu121
- Datasets 2.16.1
- Tokenizers 0.15.0