instadeepai / Mava

🦁 A research-friendly codebase for fast experimentation of multi-agent reinforcement learning in JAX
Apache License 2.0
709 stars 83 forks source link

[FEATURE] Continuous Actions PPO #990

Closed OmaymaMahjoub closed 7 months ago

OmaymaMahjoub commented 8 months ago

Feature

Enhance PPO systems by introducing support for continuous actions

Proposal

Two implementation approaches are considered:

  1. Create a select_actions file containing both discrete and continuous action selections, and integrate this into the single-file PPO systems.
  2. Split the existing ff_ippo file into two separate files: one for discrete actions and another for continuous actions.

Mandatory checklist before making a PR

Links/references