120min_wav2vec_xls-r53_FT_5e-5
This model is a fine-tuned version of jonatasgrosman/wav2vec2-large-xlsr-53-arabic on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.9644
- Wer: 0.7345
- Cer: 0.2463
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: 5e-05
- train_batch_size: 4
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 8
- 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: 300
- num_epochs: 12
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Cer | Validation Loss | Wer |
|---|---|---|---|---|---|
| No log | 1.5267 | 50 | 0.3764 | 1.5241 | 0.9142 |
| 1.5958 | 3.0534 | 100 | 0.3288 | 1.2169 | 0.8792 |
| 1.5958 | 4.5802 | 150 | 0.2977 | 1.1076 | 0.8363 |
| 1.1902 | 6.1069 | 200 | 0.2766 | 1.0394 | 0.8007 |
| 1.1902 | 7.6336 | 250 | 0.2638 | 1.0228 | 0.7689 |
| 0.9987 | 9.1603 | 300 | 0.2533 | 0.9808 | 0.7466 |
| 0.9987 | 10.8931 | 350 | 0.9644 | 0.7345 | 0.2463 |
Framework versions
- Transformers 4.41.1
- Pytorch 2.9.0+cu128
- Datasets 4.0.0
- Tokenizers 0.19.1