| site_name: VAGEN | |
| site_description: VAGEN is a multi-turn reinforcement learning framework designed specifically for training VLM Agents. VAGEN leverages the TRICO algorithm to efficiently train VLMs for visual agentic tasks. | |
| site_author: Kangrui Wang, Yaning Gao | |
| # Material theme settings | |
| theme: | |
| name: material | |
| palette: | |
| primary: indigo | |
| accent: indigo | |
| features: | |
| - header.autohide | |
| - navigation.instant | |
| icon: | |
| repo: fontawesome/brands/github | |
| repo_url: https://github.com/RAGEN-AI/VAGEN/tree/main | |
| repo_name: RAGEN-AI/VAGEN | |
| plugins: | |
| - search | |
| # Left Navigation bar | |
| nav: | |
| - Home: index.md | |
| - Quick Start: | |
| - Installation and Run Experiment: run-exp.md | |
| - Configurations: | |
| - General Configuration: configs/general-config.md | |
| - Algorithm Configuration: configs/algo-config.md | |
| - Environments: | |
| - Create Environment: envs/create-env.md | |
| - Create Service: envs/create-service.md | |
| - Experiments: | |
| - Reproduce Experiments: reproduce-exp.md | |
| # Copyright footer | |
| copyright: Copyright © 2025, VAGEN Team | |
| # Additional settings | |
| markdown_extensions: | |
| - admonition | |
| - codehilite | |
| - footnotes | |
| - toc: | |
| permalink: true | |
| - pymdownx.highlight | |
| - pymdownx.superfences |