Wav2Vec2 Large XLSR Finetuned Indonesian
This model is a fine-tuned version of facebook/wav2vec2-large-xlsr-53 on the Common Voice 17.0 dataset. It achieves the following results on the evaluation set:
- Loss: 0.3803
- Wer: 26.7576
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: 0.0003
- train_batch_size: 32
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 64
- optimizer: Use 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: 30
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 2.7664 | 3.8462 | 500 | 2.5328 | 100.0 |
| 0.3638 | 7.6923 | 1000 | 0.3765 | 36.7488 |
| 0.2505 | 11.5385 | 1500 | 0.3445 | 32.9203 |
| 0.1956 | 15.3846 | 2000 | 0.3549 | 30.5443 |
| 0.1622 | 19.2308 | 2500 | 0.3692 | 28.3122 |
| 0.1319 | 23.0769 | 3000 | 0.3788 | 28.0431 |
| 0.1119 | 26.9231 | 3500 | 0.3803 | 26.7576 |
Framework versions
- Transformers 4.49.0
- Pytorch 2.5.1+cu121
- Datasets 3.4.1
- Tokenizers 0.21.0