metadata
license: apache-2.0
tags:
- generated_from_trainer
metrics:
- wer
model-index:
- name: wav2vec2-large-xls-r-300m-telugu-asr
results: []
wav2vec2-large-xls-r-300m-telugu-asr
This model is a fine-tuned version of henilp105/wav2vec2-large-xls-r-300m-telugu-asr on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 1.2528
- Wer: 0.7044
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: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- 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 |
|---|---|---|---|---|
| 6.2329 | 2.3 | 200 | 1.0480 | 0.7942 |
| 0.7353 | 4.59 | 400 | 0.8799 | 0.7307 |
| 0.4454 | 6.89 | 600 | 0.9360 | 0.7384 |
| 0.3191 | 9.19 | 800 | 0.9886 | 0.7312 |
| 0.2246 | 11.49 | 1000 | 1.1062 | 0.7396 |
| 0.1796 | 13.79 | 1200 | 1.1293 | 0.7250 |
| 0.1454 | 16.09 | 1400 | 1.1771 | 0.7287 |
| 0.121 | 18.39 | 1600 | 1.1947 | 0.7173 |
| 0.1046 | 20.69 | 1800 | 1.2373 | 0.7277 |
| 0.0903 | 22.98 | 2000 | 1.2815 | 0.7193 |
| 0.0811 | 25.29 | 2200 | 1.2602 | 0.7148 |
| 0.0722 | 27.58 | 2400 | 1.2517 | 0.7121 |
| 0.0673 | 29.88 | 2600 | 1.2528 | 0.7044 |
Framework versions
- Transformers 4.24.0
- Pytorch 1.10.0+cu113
- Datasets 1.18.3
- Tokenizers 0.13.2