whisper_25syl_tiny_v1_3_1
This model is a fine-tuned version of openai/whisper-tiny on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.0059
- Wer: 0.6304
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: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- training_steps: 5000
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 0.0025 | 0.2461 | 1000 | 0.0193 | 1.8427 |
| 0.0025 | 0.4922 | 2000 | 0.0065 | 0.8244 |
| 0.0022 | 0.7384 | 3000 | 0.0041 | 0.5404 |
| 0.0021 | 0.9845 | 4000 | 0.0062 | 0.7274 |
| 0.0002 | 1.2306 | 5000 | 0.0059 | 0.6304 |
Framework versions
- Transformers 4.41.2
- Pytorch 2.3.0+cu121
- Datasets 2.19.1
- Tokenizers 0.19.1