whisper-large-v3-turbo-ami-disfluent-full
This model is a fine-tuned version of openai/whisper-large-v3-turbo on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.3507
- Wer: 8.4297
- Cer: 4.3573
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-05
- train_batch_size: 8
- eval_batch_size: 32
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 32
- 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: 500
- training_steps: 5000
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer | Cer |
|---|---|---|---|---|---|
| No log | 0 | 0 | 2.8591 | 23.2283 | 14.9979 |
| 0.3457 | 0.1 | 500 | 0.2787 | 9.8557 | 4.9508 |
| 0.252 | 1.0748 | 1000 | 0.2785 | 10.9926 | 5.6876 |
| 0.1053 | 2.0496 | 1500 | 0.2708 | 9.1643 | 4.5877 |
| 0.0505 | 3.0244 | 2000 | 0.3046 | 9.9821 | 5.4330 |
| 0.0544 | 3.1244 | 2500 | 0.2819 | 8.8718 | 4.4522 |
| 0.0209 | 4.0992 | 3000 | 0.3062 | 9.5699 | 5.1405 |
| 0.0111 | 5.074 | 3500 | 0.3224 | 8.5394 | 4.4230 |
| 0.0023 | 6.0488 | 4000 | 0.3427 | 8.4131 | 4.3766 |
| 0.0018 | 7.0236 | 4500 | 0.3489 | 8.3932 | 4.3516 |
| 0.0018 | 7.1236 | 5000 | 0.3507 | 8.4297 | 4.3573 |
Framework versions
- Transformers 4.54.0
- Pytorch 2.8.0.dev20250319+cu128
- Datasets 3.6.0
- Tokenizers 0.21.2