π IBM/AssetOpsBench
Main codebase, datasets, and benchmark setup.
π IBM/ReActXen
Agent codebase for ReAct, ReActXen, Reflect, RAFA, HuggingGPT, and many more for learning agents
π Docker Settings Guide
Step-by-step instructions for running the benchmark in a containerized environment for local execution and testing
To run the agent locally, you will need Watsonx.ai credentials. Access will be provided on a case-by-case basis.
π© How to Request Access: Please send a request via the Challenge Forum with the following details:
Team Name
Team Lead Name
Team Leadβs Codabench Username
We will verify your request and share credentials through Codabench-registered email of the team lead.
π ArXiv Preprint: AssetOpsBench
In-depth details, methodology, and benchmark design.
π‘ Create or Track Issues
Report bugs, request features, or contribute improvements.
β With these resources, you can:
Set up the benchmark locally (via Docker).
Explore the methodology (technical report).
Follow along with community experiments (blogs).
Contribute back via GitHub.