whisper-small-ar-Noise
This model is a fine-tuned version of MohammedNasri/whisper-small-AR on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.2514
- Wer: 20.8116
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: 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: 500
- training_steps: 3000
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 0.0985 | 0.21 | 1000 | 0.2883 | 24.6143 |
| 0.0915 | 0.42 | 2000 | 0.2640 | 22.0391 |
| 0.0694 | 0.62 | 3000 | 0.2514 | 20.8116 |
Framework versions
- Transformers 4.27.4
- Pytorch 1.13.0
- Datasets 2.11.0
- Tokenizers 0.13.2