whispertiny_buah25sil_v1_4
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: 0.0579
- Wer: 6.9081
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: 10000
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 0.0162 | 0.2286 | 1000 | 0.0467 | 10.0481 |
| 0.0099 | 0.4571 | 2000 | 0.0423 | 7.9967 |
| 0.0086 | 0.6857 | 3000 | 0.0404 | 7.7454 |
| 0.0097 | 0.9143 | 4000 | 0.0553 | 9.3783 |
| 0.0073 | 1.1429 | 5000 | 0.0443 | 7.1593 |
| 0.0068 | 1.3714 | 6000 | 0.0495 | 7.1174 |
| 0.0029 | 1.6 | 7000 | 0.0615 | 7.9967 |
| 0.0036 | 1.8286 | 8000 | 0.0568 | 7.0965 |
| 0.0009 | 2.0571 | 9000 | 0.0557 | 7.0756 |
| 0.0004 | 2.2857 | 10000 | 0.0579 | 6.9081 |
Framework versions
- Transformers 4.41.2
- Pytorch 2.3.0+cu121
- Datasets 4.0.0
- Tokenizers 0.19.1