Skip to content

Model training and evaluation code for our dataset PubTables-1M, developed to support the task of table extraction from unstructured documents.

License

Notifications You must be signed in to change notification settings

tias112/table-transformer

 
 

Repository files navigation

PubTables-1M

Update:

  • Apr 3 2022:
    • Release pre-trained weights (20 epochs)

Code Installation

Create a virtual environment and activate it as follows

python -m venv env; source env/bin/activate
pip install -U pip

pip install -r requirements.txt

Evaluation

Method to evaluate metrics like precision/result: img.png

python .\coco_error_analysis.py --ann ./data/nc_ttv/val/val.json ./data/nc_ttv/coco_out/coco_detection_output.json,./ data/nc_ttv/coco_out/coco_structure_output.json ./data/nc_ttv/coco_out

streamlit run app.py

Original README

About

Model training and evaluation code for our dataset PubTables-1M, developed to support the task of table extraction from unstructured documents.

Resources

License

Code of conduct

Security policy

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Jupyter Notebook 99.5%
  • Other 0.5%