openai/whisper-tiny
This model is a fine-tuned version of openai/whisper-tiny on the Hanhpt23/SMMMU dataset. It achieves the following results on the evaluation set:
- Loss: 0.5068
- Wer: 16.9043
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: 0.0001
- train_batch_size: 8
- 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: 100
- num_epochs: 10
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| No log | 1.0 | 14 | 1.0476 | 17.5967 |
| 1.1834 | 2.0 | 28 | 0.6758 | 19.3686 |
| 1.1834 | 3.0 | 42 | 0.5599 | 20.5458 |
| 0.3274 | 4.0 | 56 | 0.4398 | 13.8208 |
| 0.3274 | 5.0 | 70 | 0.4560 | 24.6558 |
| 0.0781 | 6.0 | 84 | 0.5055 | 21.6456 |
| 0.0781 | 7.0 | 98 | 0.5160 | 33.9348 |
| 0.0947 | 8.0 | 112 | 0.5665 | 19.8289 |
| 0.0674 | 9.0 | 126 | 0.5255 | 17.9878 |
| 0.0674 | 10.0 | 140 | 0.5068 | 16.9043 |
Framework versions
- Transformers 4.41.1
- Pytorch 2.3.0
- Datasets 2.19.1
- Tokenizers 0.19.1