WAXAL-lug-ful-w2v-bert-2.0
This model is a fine-tuned version of facebook/w2v-bert-2.0 on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.4411
- Cer: 0.0793
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: 4
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 8
- optimizer: Use adamw_torch_fused with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- num_epochs: 10
Training results
| Training Loss | Epoch | Step | Validation Loss | Cer |
|---|---|---|---|---|
| 4.8496 | 1.4218 | 300 | 0.4804 | 0.1046 |
| 0.8171 | 2.8436 | 600 | 0.4331 | 0.0942 |
| 0.5955 | 4.2654 | 900 | 0.3763 | 0.0880 |
| 0.4671 | 5.6872 | 1200 | 0.3712 | 0.0844 |
| 0.3560 | 7.1090 | 1500 | 0.3939 | 0.0789 |
| 0.2647 | 8.5308 | 1800 | 0.4167 | 0.0797 |
| 0.1797 | 9.9526 | 2100 | 0.4408 | 0.0794 |
| 0.1797 | 10.0 | 2110 | 0.4411 | 0.0793 |
Framework versions
- Transformers 5.13.0
- Pytorch 2.13.0+cu130
- Datasets 5.0.0
- Tokenizers 0.22.2