<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->
openai/whisper-medium.en
This model is a fine-tuned version of openai/whisper-medium.en on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.4158
- Wer: 10.8712
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: 32
- eval_batch_size: 32
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- training_steps: 4000
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 0.6148 | 0.12 | 500 | 0.3107 | 12.7838 |
| 0.1877 | 1.09 | 1000 | 0.2892 | 11.2910 |
| 0.0697 | 2.05 | 1500 | 0.3146 | 10.7857 |
| 0.0748 | 3.02 | 2000 | 0.3162 | 11.5254 |
| 0.0308 | 3.14 | 2500 | 0.3450 | 11.1111 |
| 0.0192 | 4.11 | 3000 | 0.3720 | 10.9101 |
| 0.0046 | 5.07 | 3500 | 0.4155 | 11.2344 |
| 0.0096 | 6.03 | 4000 | 0.4158 | 10.8712 |
Framework versions
- Transformers 4.29.0
- Pytorch 1.14.0a0+44dac51
- Datasets 2.12.0
- Tokenizers 0.13.3
