Wave2Vec2-Bert2.0 - Kiran Pantha
This model is a fine-tuned version of facebook/w2v-bert-2.0 on the OpenSLR54 dataset. It achieves the following results on the evaluation set:
- Loss: 0.4271
- Wer: 0.4432
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: 5e-05
- train_batch_size: 8
- eval_batch_size: 8
- 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: 2
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 3.8954 | 0.15 | 300 | 1.0556 | 0.8694 |
| 0.938 | 0.3 | 600 | 0.8641 | 0.7710 |
| 0.8269 | 0.45 | 900 | 0.6742 | 0.6457 |
| 0.729 | 0.6 | 1200 | 0.6141 | 0.5665 |
| 0.6879 | 0.75 | 1500 | 0.6085 | 0.5791 |
| 0.6386 | 0.9 | 1800 | 0.5424 | 0.5333 |
| 0.5923 | 1.05 | 2100 | 0.4991 | 0.4880 |
| 0.5403 | 1.2 | 2400 | 0.4821 | 0.4870 |
| 0.4965 | 1.35 | 2700 | 0.4794 | 0.4793 |
| 0.5249 | 1.5 | 3000 | 0.4520 | 0.4607 |
| 0.4936 | 1.65 | 3300 | 0.4569 | 0.4586 |
| 0.473 | 1.8 | 3600 | 0.4527 | 0.4606 |
| 0.4414 | 1.95 | 3900 | 0.4271 | 0.4432 |
Framework versions
- Transformers 4.45.0.dev0
- Pytorch 2.4.1+cu121
- Datasets 2.21.0
- Tokenizers 0.19.1