Changelog

Versions follow Semantic Versioning (<major>.<minor>.<patch>).

Backward incompatible (breaking) changes will only be introduced in major versions with advance notice in the Deprecations section of releases.

Warning

You should NOT be adding new change log entries to this file, this file is managed by towncrier. You may edit previous change logs to fix problems like typo corrections or such.

UnrealMLAgents 1.0.0 (2025-02-11)

Features

  • Introduced core Unreal ML Agents Plugin to enable reinforcement learning (RL) in Unreal Engine.

  • Implemented Python API integration for seamless communication with training workflows.

  • Added vector-based observations and RaySensor components for improved perception.

  • Provided Blueprint-friendly design, allowing customization without C++.

  • Enabled Multi-Agent support for scalable and parallelized training environments.

Known Issues

  • Inference is not yet implemented; currently, training is the primary focus.

  • SideChannels are not yet supported.

  • Imitation Learning is not yet supported.

  • Limited sensor support compared to Unity ML-Agents; more sensors will be added in future updates.

  • Not all features have been thoroughly tested, and some areas may still contain unresolved issues.

  • Lacking proper unit tests; additional validation and automated testing are needed to improve stability.