---
license: apache-2.0
datasets:
- HuggingFaceM4/FineVision
base_model:
- lightonai/LightOnOCR-1B-1025
---

This model is a fine-tuned version of [lightonai/LightOnOCR-1B-1025](https://huggingface.co/lightonai/LightOnOCR-1B-1025) on [HuggingFaceM4/FineVision](https://huggingface.co/datasets/HuggingFaceM4/FineVision) dataset. 

It achieves the following results on the evaluation set:

-Loss: 0.4395

### Intended uses & limitations

Please refer the [original information](https://huggingface.co/lightonai/LightOnOCR-1B-1025)

### Training hyperparameters

The following hyperparameters were used during training:
- learning_rate: 6e-05
- train_batch_size: 4
- eval_batch_size: 6
- seed: 42
- gradient_accumulation_steps: 4
- 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: 10
- training_steps: 100

### Training results

| Training Loss | Epoch  | Step | Validation Loss |
|:-------------:|:------:|:----:|:---------------:|
| 3.1522        | 0.1661 | 50   | 0.4954          |
| 1.6662        | 0.3322 | 100  | 0.4395          |


### Framework versions

- PEFT 0.18.0
- Transformers 5.0.0.dev0
- Pytorch 2.9.1
- Datasets 4.4.1
- Tokenizers 0.22.1