rhoahndur/retrosynthesis-ai
Upload folder using huggingface_hub
Upload folder using huggingface_hub
Upload folder using huggingface_hub
Upload folder using huggingface_hub
Upload folder using huggingface_hub
Upload folder using huggingface_hub
Upload folder using huggingface_hub
Upload folder using huggingface_hub
Upload folder using huggingface_hub
Upload folder using huggingface_hub
Upload folder using huggingface_hub
Upload folder using huggingface_hub
Upload folder using huggingface_hub
Upload folder using huggingface_hub
Upload folder using huggingface_hub
Upload folder using huggingface_hub
Upload folder using huggingface_hub
Upload folder using huggingface_hub
remove debug message
nonchange, switched OSes
update architecture doc
update readme
1. Bug fix — train_rl.py:308: Renamed best_reward to best_sample_reward to stop shadowing the outer checkpoint-tracking variable
add evals
mod ignore
The slow tests were always meant to be opt-in — the addopts just wasn't excluding
1. MCTS Cycle Detection (env/MCTS.py):
update gitignore
Add HF Spaces deployment, v0.3.0 environment with HF dataset, sync workflow
change to svg molecule rendering
fix bugs for frontend/parsing of inference results
build env and config for prime intellect deployment of RL model
pivot to be able to deploy RL runs out of PrimeIntellect
update gitignore
pin ruff version
fix lint errors, and pre-commit hook for linting
add makefile for linting and unittests
add unit tests
add linting and testing config
setup initial cicd
create streamlit app
create inference stcript
setup the chemenv
add architecture doc and start up a gitignore
initial commit
first commit