wh_med_cv_no_lg_id_common_voices_52395__val_finetune_trial
This model is a fine-tuned version of openai/whisper-medium on the common_voice_22_0 dataset. It achieves the following results on the evaluation set:
- Loss: 0.2222
- Global Wer: 30.3914
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 |
|---|---|---|---|---|
| 1.2412 | 0.0411 | 50 | 0.8146 | 45.0429 |
| 0.3667 | 0.0821 | 100 | 0.3270 | 43.3100 |
| 0.2843 | 0.1232 | 150 | 0.2873 | 38.8601 |
| 0.2507 | 0.1642 | 200 | 0.2669 | 35.7382 |
| 0.2546 | 0.2053 | 250 | 0.2583 | 32.7557 |
| 0.1979 | 0.2463 | 300 | 0.2503 | 33.9705 |
| 0.2139 | 0.2874 | 350 | 0.2381 | 31.6149 |
| 0.2047 | 0.3284 | 400 | 0.2323 | 31.7368 |
| 0.2152 | 0.3695 | 450 | 0.2225 | 29.9473 |
| 0.2004 | 0.4105 | 500 | 0.2222 | 30.3914 |
Framework versions
- Transformers 5.0.0.dev0
- Pytorch 2.9.0+cu126
- Datasets 3.6.0
- Tokenizers 0.22.2