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: 1.3902
- Wer: 0.7443
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 |
|---|---|---|---|---|
| 6.9225 | 2.3 | 200 | 3.3972 | 1.0 |
| 1.4526 | 4.59 | 400 | 1.0196 | 0.7959 |
| 0.5384 | 6.89 | 600 | 1.0260 | 0.7790 |
| 0.3483 | 9.19 | 800 | 1.0932 | 0.7740 |
| 0.2428 | 11.49 | 1000 | 1.2085 | 0.7747 |
| 0.1839 | 13.79 | 1200 | 1.2716 | 0.7750 |
| 0.147 | 16.09 | 1400 | 1.2895 | 0.7665 |
| 0.1238 | 18.39 | 1600 | 1.2995 | 0.7585 |
| 0.1046 | 20.69 | 1800 | 1.3891 | 0.7550 |
| 0.0946 | 22.98 | 2000 | 1.3820 | 0.7603 |
| 0.0856 | 25.29 | 2200 | 1.3909 | 0.7438 |
| 0.0753 | 27.58 | 2400 | 1.3841 | 0.7431 |
| 0.075 | 29.88 | 2600 | 1.3902 | 0.7443 |
Framework versions
- Transformers 4.24.0
- Pytorch 1.10.0+cu113
- Datasets 1.18.3
- Tokenizers 0.13.2