v0.1.4

Easy way to create complex environments with ToolCollectionEnvironment

Full Changelog: v0.1.3...v0.1.4

v0.1.3

• Form Filler Agent described in paper now available in the examples
• uv as the package manager
• Agent optimization examples
• One agent can use many LLMs now
• Faster RL loops with batching, multi-GPU and multi-node support
• Using VLLM with log probability and token ids during RL training

v0.1.3.dev1

prerelease

v0.1.2: Media observations, RL MATH, search tool

• Image and video observation steps.
• Tools for extracting video observations.
• Search tool using serper.dev service as backend.
• Caching layer for any functions used as a tool.
• New benchmark results in Gaia example readme.
• RL for MATH & GSM8K example.
• Minor improvements to the tape browser.

v0.1.1

initial release to pypi and github releases