Skip to content

mikezhang25/SparseLegalSum

Repository files navigation

SparseLegalSum

CS224N Final Project (Michael Zhang, Alex Alvarado-Barahona)

Usage

Run python src/train.py for complete usage.
Run python src/chunkbigbird.py to run evaluation for K-Overlap.

Roadmap

By Project Milestone Deadline (4:30pm on Thursday, 03/02)

  • Load model and dataset
  • Test model zero-shot
    • Get score
  • Pretrain on MLM task
    • Get score
  • Compile report
  • (if time) Finetune model on downstream summarization task
    • Get score

By Final Project Deadline (4:30pm on Saturday, 03/18)

  • Pretrain on Manipulated Work Detection task
    • Finetune on downstream summarization task
    • Get score
  • Pretrain on Masked First Character Prediction task
    • Finetune on downstream summarization task
    • Get score
  • Compile Report

About

CS224N Final Project for summarizing lengthy legal texts 10x the size of LLM context windows. Invited to continue research at the Stanford AI Laboratory.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages