|
|
--- |
|
|
library_name: transformers |
|
|
language: |
|
|
- ar |
|
|
license: apache-2.0 |
|
|
base_model: openai/whisper-base |
|
|
tags: |
|
|
- generated_from_trainer |
|
|
metrics: |
|
|
- wer |
|
|
model-index: |
|
|
- name: Whisper base AR - BA |
|
|
results: [] |
|
|
--- |
|
|
|
|
|
<!-- This model card has been generated automatically according to the information the Trainer had access to. You |
|
|
should probably proofread and complete it, then remove this comment. --> |
|
|
|
|
|
# Whisper base AR - BA |
|
|
|
|
|
This model is a fine-tuned version of [openai/whisper-base](https://huggingface.co/openai/whisper-base) on the quran-ayat-speech-to-text dataset. |
|
|
It achieves the following results on the evaluation set: |
|
|
- Loss: 0.0037 |
|
|
- Wer: 0.0550 |
|
|
- Cer: 0.0222 |
|
|
|
|
|
## 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: 16 |
|
|
- eval_batch_size: 16 |
|
|
- seed: 42 |
|
|
- gradient_accumulation_steps: 2 |
|
|
- total_train_batch_size: 32 |
|
|
- optimizer: Use OptimizerNames.ADAMW_TORCH 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 |
|
|
- num_epochs: 25 |
|
|
- mixed_precision_training: Native AMP |
|
|
|
|
|
### Training results |
|
|
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Wer | Cer | |
|
|
|:-------------:|:-------:|:----:|:---------------:|:------:|:------:| |
|
|
| 0.0049 | 1.0 | 282 | 0.0037 | 0.0498 | 0.0180 | |
|
|
| 0.0031 | 2.0 | 564 | 0.0042 | 0.0525 | 0.0202 | |
|
|
| 0.0015 | 3.0 | 846 | 0.0049 | 0.0568 | 0.0226 | |
|
|
| 0.0012 | 4.0 | 1128 | 0.0058 | 0.0590 | 0.0229 | |
|
|
| 0.0008 | 5.0 | 1410 | 0.0057 | 0.0635 | 0.0243 | |
|
|
| 0.0007 | 6.0 | 1692 | 0.0063 | 0.0639 | 0.0223 | |
|
|
| 0.0005 | 7.0 | 1974 | 0.0067 | 0.0610 | 0.0237 | |
|
|
| 0.0005 | 8.0 | 2256 | 0.0070 | 0.0612 | 0.0232 | |
|
|
| 0.0003 | 9.0 | 2538 | 0.0073 | 0.0626 | 0.0243 | |
|
|
| 0.0003 | 10.0 | 2820 | 0.0080 | 0.0643 | 0.0239 | |
|
|
| 0.0003 | 11.0 | 3102 | 0.0088 | 0.0635 | 0.0237 | |
|
|
| 0.0003 | 12.0 | 3384 | 0.0087 | 0.0605 | 0.0231 | |
|
|
| 0.0002 | 13.0 | 3666 | 0.0092 | 0.0612 | 0.0239 | |
|
|
| 0.0002 | 14.0 | 3948 | 0.0104 | 0.0610 | 0.0226 | |
|
|
| 0.0001 | 15.0 | 4230 | 0.0105 | 0.0543 | 0.0194 | |
|
|
| 0.0001 | 16.0 | 4512 | 0.0111 | 0.0568 | 0.0216 | |
|
|
| 0.0 | 17.0 | 4794 | 0.0124 | 0.0556 | 0.0206 | |
|
|
| 0.0001 | 18.0 | 5076 | 0.0128 | 0.0539 | 0.0210 | |
|
|
| 0.0 | 19.0 | 5358 | 0.0130 | 0.0530 | 0.0201 | |
|
|
| 0.0 | 20.0 | 5640 | 0.0130 | 0.0539 | 0.0197 | |
|
|
| 0.0 | 21.0 | 5922 | 0.0129 | 0.0536 | 0.0202 | |
|
|
| 0.0 | 22.0 | 6204 | 0.0130 | 0.0541 | 0.0208 | |
|
|
| 0.0 | 23.0 | 6486 | 0.0130 | 0.0547 | 0.0211 | |
|
|
| 0.0 | 24.0 | 6768 | 0.0129 | 0.0605 | 0.0246 | |
|
|
| 0.0 | 24.9130 | 7025 | 0.0130 | 0.0547 | 0.0210 | |
|
|
|
|
|
|
|
|
### Framework versions |
|
|
|
|
|
- Transformers 4.51.1 |
|
|
- Pytorch 2.5.1+cu124 |
|
|
- Datasets 3.5.0 |
|
|
- Tokenizers 0.21.0 |
|
|
|
