whisper-large-paper_
This model is a fine-tuned version of openai/whisper-large on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.4374
- Wer: 47.9863
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: 2e-05
- train_batch_size: 24
- eval_batch_size: 24
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- num_epochs: 6
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| No log | 1.0 | 143 | 0.3754 | 47.3394 |
| No log | 2.0 | 286 | 0.3418 | 44.5511 |
| No log | 3.0 | 429 | 0.3522 | 47.7507 |
| 0.3895 | 4.0 | 572 | 0.3795 | 48.9312 |
| 0.3895 | 5.0 | 715 | 0.4091 | 51.5160 |
| 0.3895 | 6.0 | 858 | 0.4374 | 47.9863 |
Framework versions
- Transformers 4.31.0.dev0
- Pytorch 2.0.0+cu117
- Datasets 2.7.0
- Tokenizers 0.13.2