whisper-small-SIT_16_07_26_0300
This model is a fine-tuned version of Pengwin30/whisper-small-fine-tuned on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.0097
- Wer: 0.3976
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
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED 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: 100
- num_epochs: 5
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 0.7583 | 1.0 | 181 | 0.0221 | 0.7952 |
| 0.0142 | 2.0 | 362 | 0.0107 | 0.4639 |
| 0.0054 | 3.0 | 543 | 0.0075 | 0.3976 |
| 0.0018 | 4.0 | 724 | 0.0096 | 0.3976 |
| 0.0001 | 5.0 | 905 | 0.0097 | 0.3976 |
Framework versions
- Transformers 5.12.1
- Pytorch 2.11.0+cu128
- Datasets 2.19.0
- Tokenizers 0.22.2