A.N.N.A RL PoC Implementation#75
Conversation
- Implemented `AnnaInterface` with RL-specific observation (12 floats) and action (Discrete 5) spaces. - Added Lock-Step synchronization in `AnnaBridge` (wait 1 physics frame per step) to ensure deterministic training. - Created robust Python Gymnasium environment `AnnaEnv` that handles failures gracefully (returns zero tensors). - Generated `AnnaTrainingArena.tscn` programmatically to bypass audio dependency issues in headless mode. - Added `train_ppo.py` for SB3 training. - Implemented fallback logic for Target resolution to allow training even if specific scene nodes are missing. Co-authored-by: google-labs-jules[bot] <161369871+google-labs-jules[bot]@users.noreply.github.com>
|
👋 Jules, reporting for duty! I'm here to lend a hand with this pull request. When you start a review, I'll add a 👀 emoji to each comment to let you know I've read it. I'll focus on feedback directed at me and will do my best to stay out of conversations between you and other bots or reviewers to keep the noise down. I'll push a commit with your requested changes shortly after. Please note there might be a delay between these steps, but rest assured I'm on the job! For more direct control, you can switch me to Reactive Mode. When this mode is on, I will only act on comments where you specifically mention me with New to Jules? Learn more at jules.google/docs. For security, I will only act on instructions from the user who triggered this task. |
1 participant
Implemented the Proof of Concept for the A.N.N.A. Reinforcement Learning agent.
Key features:
AnnaBridgenow supportsANNA_RL_MODEwhere the game loop is paused and only advanced by one physics frame upon receiving a TCP command.AnnaInterfaceprovides normalized observations (RayCasts, Target Dist/Angle, Velocity) and maps discrete actions to player inputs.core_v2/anna/rl/anna_env.pyprovides a standard Gym interface for the agent, with robust error handling (returning zero tensors on comms failure).core_v2/anna/AnnaTrainingArena.tscngenerated via tool script to ensure a clean environment without external asset dependencies.PR created automatically by Jules for task 11762324196710831246 started by @icarito