<!-- 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. -->
T5-asr-corrector
This model is a fine-tuned version of flax-community/bengali-t5-base on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.4683
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: 2e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 6
- total_train_batch_size: 48
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 1
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 1.6804 | 0.15 | 500 | 0.8576 |
| 0.792 | 0.31 | 1000 | 0.6556 |
| 0.6553 | 0.46 | 1500 | 0.5640 |
| 0.5901 | 0.62 | 2000 | 0.5114 |
| 0.5454 | 0.77 | 2500 | 0.4815 |
| 0.53 | 0.93 | 3000 | 0.4683 |
Framework versions
- Transformers 4.25.1
- Pytorch 1.13.0+cu116
- Datasets 2.8.0
- Tokenizers 0.13.2
