<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->
wav2vec2-hindi-bhoj-3
This model is a fine-tuned version of facebook/wav2vec2-large-xlsr-53 on the None dataset. It achieves the following results on the evaluation set:
- Loss: 5.7033
- Wer: 1.1477
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: 0.0001
- train_batch_size: 16
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- 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: 500
- num_epochs: 40
- mixed_precision_training: Native AMP
Training results
Training Loss | Epoch | Step | Validation Loss | Wer |
---|---|---|---|---|
8.6136 | 6.45 | 400 | 3.6017 | 1.0 |
2.6692 | 12.9 | 800 | 4.5408 | 1.0872 |
0.5639 | 19.35 | 1200 | 5.2302 | 1.2282 |
0.2296 | 25.8 | 1600 | 5.3323 | 1.0872 |
0.1496 | 32.26 | 2000 | 5.7219 | 1.1342 |
0.1098 | 38.7 | 2400 | 5.7033 | 1.1477 |
Framework versions
- Transformers 4.20.0.dev0
- Pytorch 1.11.0+cu113
- Datasets 2.2.3.dev0
- Tokenizers 0.12.1