ardj5.2
This model is a fine-tuned version of faissalb/whisper-small-ardj5.1 on the test dataset. It achieves the following results on the evaluation set:
- Loss: 0.2645
- Wer: 10.7393
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-06
- train_batch_size: 16
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 32
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- training_steps: 6000
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 0.0157 | 2.15 | 1000 | 0.2628 | 10.7443 |
| 0.0079 | 4.3 | 2000 | 0.2637 | 10.9840 |
| 0.0092 | 6.44 | 3000 | 0.2635 | 10.7842 |
| 0.0094 | 8.59 | 4000 | 0.2634 | 10.7892 |
| 0.0039 | 10.74 | 5000 | 0.2639 | 10.7692 |
| 0.0051 | 12.89 | 6000 | 0.2645 | 10.7393 |
Framework versions
- Transformers 4.28.0
- Pytorch 2.0.1+cu118
- Datasets 2.14.3
- Tokenizers 0.13.3
- Downloads last month
- 6