whisper-small-test-minan_3
This model is a fine-tuned version of openai/whisper-small on the common_voice_16_1 dataset. It achieves the following results on the evaluation set:
- Loss: 0.9980
- Wer: 83.6475
- Cer: 59.6632
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: 8
- seed: 42
- optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- 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 | Cer |
|---|---|---|---|---|---|
| 0.0788 | 4.3478 | 1000 | 0.9545 | 84.9627 | 59.0285 |
| 0.0024 | 8.6957 | 2000 | 0.9798 | 83.9982 | 57.0596 |
| 0.0004 | 13.0435 | 3000 | 0.9909 | 83.4722 | 59.5855 |
| 0.0003 | 17.3913 | 4000 | 0.9980 | 83.6475 | 59.6632 |
Framework versions
- Transformers 4.46.3
- Pytorch 2.4.0+cu124
- Datasets 3.1.0
- Tokenizers 0.20.3