wav2vec2-large-xls-r-300m-marathi-cv8
This model is a fine-tuned version of facebook/wav2vec2-xls-r-300m on the MOZILLA-FOUNDATION/COMMON_VOICE_8_0 - MR dataset. It achieves the following results on the evaluation set:
- Loss: 0.6483
- Wer: 0.6049
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: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- num_epochs: 100.0
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 2.671 | 22.73 | 500 | 1.3618 | 0.9499 |
| 1.1599 | 45.45 | 1000 | 0.6330 | 0.6627 |
| 0.8252 | 68.18 | 1500 | 0.6226 | 0.6426 |
| 0.6424 | 90.91 | 2000 | 0.6359 | 0.6041 |
Framework versions
- Transformers 4.17.0.dev0
- Pytorch 1.10.2+cu102
- Datasets 1.18.3
- Tokenizers 0.11.0