feat: load PPO hyperparameters and arch from config#20

Open

Ktyby21 wants to merge 2 commits intomainfrom

codex/update-train_rl_all_pairs.py-and-config.json

Owner

Ktyby21 commented Aug 23, 2025

Summary

support reading PPO hyperparameters and policy architecture from config
seed VecNormalize and optionally force a fresh model when observation space changes
update config.json with tuned hyperparameters and risk settings

Testing

python -m py_compile train_rl_all_pairs.py
python -m json.tool config.json > /dev/null

https://chatgpt.com/codex/tasks/task_e_68a98281772c8326bc2c04f9f4c04100


          feat: read PPO hyperparams and architecture from config

57f9142

Ktyby21 added the codex label

— with

ChatGPT Codex Connector


          refactor: remove explicit seeding

f9ab88c

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels