CoolFace
Apppublic

cactus183/patchbench-dev

sourceHugging Faceupdated 6mo agoView on Hugging Face
0likes
10 commits on main
3c20beb6mo ago

Fix /reset to accept empty body, /step returns 400 not 422

cactus183
41328086mo ago

Add HF Space YAML frontmatter to README

cactus183
41e7ccb6mo ago

Upgrade 7.5: honest README reframe for flat one-shot ceiling

cactus183
596541d6mo ago

Upgrade 7: real 9-task baseline, honest reframe

cactus183
eff7d7a6mo ago

Upgrades 4-6: episode walkthrough, data flow diagram, task taxonomy

cactus183, Claude Opus 4.6
f165c776mo ago

Upgrade 2+3: reward breakdown in grader, 3 new tasks (9 total)

cactus183, Claude Opus 4.6
e91b7366mo ago

Upgrade 1: real Qwen 72B baseline scores

cactus183
97f8bec6mo ago

Add infrastructure: Dockerfile, server, inference.py, openenv.yaml, pyproject.toml, requirements.txt, README

cactus183
9c9267b6mo ago

Add .gitignore and remove cached pyc files

cactus183
98f4e7f6mo ago

Environment core: models, grader, environment, 6 tasks

cactus183