wav2vec2-base-ks-padpt400
This model is a fine-tuned version of facebook/wav2vec2-base on the superb dataset. It achieves the following results on the evaluation set:
- Loss: 1.2218
- Accuracy: 0.6343
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: 0.003
- train_batch_size: 256
- eval_batch_size: 256
- seed: 0
- gradient_accumulation_steps: 4
- total_train_batch_size: 1024
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 10.0
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 1.2948 | 1.0 | 50 | 1.6527 | 0.6108 |
| 0.8861 | 2.0 | 100 | 1.2653 | 0.6130 |
| 0.7809 | 3.0 | 150 | 1.2615 | 0.5924 |
| 0.7364 | 4.0 | 200 | 1.2218 | 0.6343 |
| 0.6944 | 5.0 | 250 | 1.2137 | 0.6324 |
| 0.6817 | 6.0 | 300 | 1.2822 | 0.5930 |
| 0.6601 | 7.0 | 350 | 1.3292 | 0.5599 |
| 0.6464 | 8.0 | 400 | 1.2744 | 0.5869 |
| 0.653 | 9.0 | 450 | 1.3916 | 0.5272 |
| 0.633 | 10.0 | 500 | 1.3344 | 0.5606 |
Framework versions
- Transformers 4.22.0.dev0
- Pytorch 1.11.0+cu115
- Datasets 2.4.0
- Tokenizers 0.12.1