metadata

library_name: transformers
language:
  - en
base_model:
  - allenai/olmOCR-7B-0225-preview
license: apache-2.0

olmOCR-7B-faithful

This is a fine-tuned version of the olmOCR-7B-0225-preview model that aims to extract all information from a given document, including header and footer information.

More information on how we fine-tuned the model can be found in our blog post.

Acknowledgment

We thank the Allen Institute for AI and Alibaba Cloud for their great open-source work, which enabled this fine-tuning project.

Improved using Qwen.