whisper-base-atcosim
This model is a fine-tuned version of openai/whisper-base on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.0620
- Wer: 2.9820
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: 1e-05
- train_batch_size: 128
- eval_batch_size: 128
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 10
- num_epochs: 100
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 0.152 | 8.33 | 500 | 0.0522 | 2.5282 |
| 0.001 | 16.67 | 1000 | 0.0539 | 3.0608 |
| 0.0003 | 25.0 | 1500 | 0.0556 | 3.0237 |
| 0.0002 | 33.33 | 2000 | 0.0567 | 3.0237 |
| 0.0001 | 41.67 | 2500 | 0.0579 | 3.0144 |
| 0.0001 | 50.0 | 3000 | 0.0588 | 2.9959 |
| 0.0001 | 58.33 | 3500 | 0.0597 | 3.0052 |
| 0.0001 | 66.67 | 4000 | 0.0604 | 3.0098 |
| 0.0 | 75.0 | 4500 | 0.0610 | 2.9867 |
| 0.0 | 83.33 | 5000 | 0.0615 | 2.9867 |
| 0.0 | 91.67 | 5500 | 0.0619 | 2.9774 |
| 0.0 | 100.0 | 6000 | 0.0620 | 2.9820 |
Framework versions
- Transformers 4.36.2
- Pytorch 2.1.2
- Datasets 2.15.0
- Tokenizers 0.15.0