| spec_version: 1 | |
| name: opensleuth | |
| version: "0.5.0" | |
| description: >- | |
| OpenSleuth: an OpenEnv-conformant environment that trains LLMs to | |
| reverse-engineer hidden Python functions by probing them and submitting code | |
| that reproduces them. Used for GRPO RL post-training. | |
| author: anugrah55 | |
| type: space | |
| runtime: fastapi | |
| app: server:app | |
| port: 7860 | |
| action: OpenSleuthAction | |
| observation: OpenSleuthObservation | |
| documentation_url: https://huggingface.co/spaces/anugrah55/opensleuth-env-gemini-cli | |
| tags: | |
| - rl | |
| - grpo | |
| - code | |
| - openenv | |
| - openenv-conformant | |