Skip to content

mattbuot/CUAD-Mistral-finetuning

Repository files navigation

CUAD-Mistral-finetuning

Fine tuning Mistral models to solve the Contract Understanding Atticus Dataset (CUAD) benchmark

SETUP

  1. setup project using uv
  2. download CUAD under data/CUAD_v1
  3. setup Mistral API key as env variable
  4. run generate_dataset.py to create fine-tuning datasets and upload to Mistral
  5. use the helpers in utils.py (or the UI) to launch a job
  6. run_inference.py to generate predictions
  7. evaluate.py to evaluate the results

Tested on MacOS using python 3.12

RESULTS

model task precision recall
baseline easy 50% 51%
fine-tuned easy 89% 86%
baseline hard 16% 16%
fine-tuned hard 29% 29%

About

Fine tuning Mistral models to solve the Contract Understanding Atticus Dataset (CUAD) benchmark

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages