trocr-bigram4-BY-50k
This model is a fine-tuned version of cyttic/exp2-frozen-benyehuda-cont on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 1.2304
- Cer: 0.0891
- Wer: 0.2221
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: 2
- total_train_batch_size: 16
- optimizer: Use OptimizerNames.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: 900
- num_epochs: 3
Training results
| Training Loss | Epoch | Step | Validation Loss | Cer | Wer |
|---|---|---|---|---|---|
| 3.6495 | 0.6667 | 2000 | 1.5824 | 0.1527 | 0.3377 |
| 2.2432 | 1.3333 | 4000 | 1.3824 | 0.1130 | 0.2694 |
| 2.0537 | 2.0 | 6000 | 1.2706 | 0.0976 | 0.2401 |
| 1.6771 | 2.6667 | 8000 | 1.2389 | 0.0900 | 0.2238 |
| 1.5854 | 3.0 | 9000 | 1.2304 | 0.0891 | 0.2221 |
Framework versions
- Transformers 5.15.0
- Pytorch 2.11.0+cu128
- Datasets 5.0.1
- Tokenizers 0.22.2