wh_l_v3_fleurs_no_lg_id_fleurs_52395__val_finetune_trial
This model is a fine-tuned version of openai/whisper-large-v3 on the fleurs dataset. It achieves the following results on the evaluation set:
- Loss: 0.1621
- Global Wer: 92.9609
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: 5e-06
- train_batch_size: 4
- eval_batch_size: 4
- seed: 42
- optimizer: Use adamw_torch_fused with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: constant_with_warmup
- lr_scheduler_warmup_steps: 50
- training_steps: 500
Training results
| Training Loss | Epoch | Step | Validation Loss | Global Wer |
|---|---|---|---|---|
| 0.3342 | 0.0945 | 50 | 0.3188 | 101.2663 |
| 0.2024 | 0.1890 | 100 | 0.1910 | 81.2477 |
| 0.1534 | 0.2836 | 150 | 0.1803 | 76.2756 |
| 0.1635 | 0.3781 | 200 | 0.1756 | 97.6350 |
| 0.1578 | 0.4726 | 250 | 0.1694 | 95.4190 |
| 0.1435 | 0.5671 | 300 | 0.1704 | 94.7300 |
| 0.1440 | 0.6616 | 350 | 0.1616 | 94.4693 |
| 0.1447 | 0.7561 | 400 | 0.1661 | 105.6052 |
| 0.1204 | 0.8507 | 450 | 0.1668 | 94.8231 |
| 0.1389 | 0.9452 | 500 | 0.1621 | 92.9609 |
Framework versions
- Transformers 5.0.0.dev0
- Pytorch 2.9.0+cu126
- Datasets 3.6.0
- Tokenizers 0.22.2