search

openreasoner / openr

OpenR: An Open Source Framework for Advanced Reasoning with Large Language Models
https://openreasoner.github.io/
MIT License
1.08k stars 79 forks source link

Support self-refining Critic-MCTS #44

Open YanSong97 opened 2 weeks ago

YanSong97 commented 2 weeks ago

LLM self-refining during tree searching #42

TODO:

  1. Merge critic_MATH env to MATH
  2. Configurate step tag
  3. Solve action_his and prm length mismatch problem