Whisper-Small-Inbrowser-Proctor-LORA
This model is a fine-tuned version of openai/whisper-small on the Inbrowser Procotor Dataset dataset. It achieves the following results on the evaluation set:
- Loss: 0.3646
- Wer: 18.1586
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: 5e-06
- train_batch_size: 16
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 50
- training_steps: 250
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 0.7817 | 0.8929 | 25 | 0.7456 | 31.6502 |
| 0.3905 | 1.7857 | 50 | 0.4646 | 29.4043 |
| 0.2194 | 2.6786 | 75 | 0.3988 | 20.3090 |
| 0.1697 | 3.5714 | 100 | 0.3776 | 16.1357 |
| 0.1246 | 4.4643 | 125 | 0.3744 | 18.7639 |
| 0.1062 | 5.3571 | 150 | 0.3698 | 19.9267 |
| 0.0862 | 6.25 | 175 | 0.3698 | 19.9108 |
| 0.0701 | 7.1429 | 200 | 0.3651 | 18.0153 |
| 0.0647 | 8.0357 | 225 | 0.3659 | 18.4613 |
| 0.056 | 8.9286 | 250 | 0.3646 | 18.1586 |
Framework versions
- PEFT 0.12.1.dev0
- Transformers 4.45.0.dev0
- Pytorch 2.4.1+cu121
- Datasets 3.0.0
- Tokenizers 0.19.1