whisper-small-finetuned-800-standard-A-epochs-10-5-03-20-17
This model is a fine-tuned version of openai/whisper-small on the horoscope_standard_a_800_5_03 dataset. It achieves the following results on the evaluation set:
- Loss: 0.5525
- Wer: 28.3884
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: 8
- eval_batch_size: 4
- 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
- num_epochs: 10
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| No log | 0 | 0 | 0.6890 | 41.0325 |
| 0.1314 | 1.0 | 100 | 0.5101 | 30.8481 |
| 0.0294 | 2.0 | 200 | 0.5205 | 30.1179 |
| 0.0115 | 3.0 | 300 | 0.5415 | 29.8745 |
| 0.0051 | 4.0 | 400 | 0.5291 | 28.1706 |
| 0.0026 | 5.0 | 500 | 0.5359 | 27.6198 |
| 0.0015 | 6.0 | 600 | 0.5480 | 28.4653 |
| 0.0012 | 7.0 | 700 | 0.5473 | 28.3116 |
| 0.0011 | 8.0 | 800 | 0.5505 | 28.3884 |
| 0.001 | 9.0 | 900 | 0.5520 | 28.4781 |
| 0.0009 | 10.0 | 1000 | 0.5525 | 28.3884 |
Framework versions
- Transformers 4.49.0
- Pytorch 2.5.1+cu121
- Datasets 3.4.1
- Tokenizers 0.21.0