openai/whisper-large-v2
This model is a fine-tuned version of openai/whisper-large-v2 on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.1969
- Wer: 9.3970
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: 16
- 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 |
|---|---|---|---|---|
| 0.6144 | 0.12 | 500 | 0.2795 | 14.0737 |
| 0.1643 | 0.25 | 1000 | 0.2213 | 11.4916 |
| 0.2175 | 0.38 | 1500 | 0.2009 | 10.0021 |
| 0.1512 | 1.11 | 2000 | 0.1980 | 11.2632 |
| 0.1527 | 1.24 | 2500 | 0.1916 | 10.8469 |
| 0.0918 | 1.36 | 3000 | 0.1890 | 9.6498 |
| 0.047 | 2.1 | 3500 | 0.2034 | 9.4274 |
| 0.0822 | 2.23 | 4000 | 0.1969 | 9.3970 |
Framework versions
- Transformers 4.27.0.dev0
- Pytorch 1.13.1+cu117
- Datasets 2.9.1.dev0
- Tokenizers 0.13.2