Mava
Mava copied to clipboard
🦁 A research-friendly codebase for fast experimentation of multi-agent reinforcement learning in JAX
### What do you want to investigate? Investigate how the performance of the MAPPO algorithm is affected by adding the **agent's ID**, the agent's unique identifier, to the **global state**,...
### Please describe the purpose of the feature. Is it related to a problem? We are constantly having to check that new features don't trivially break the existing systems/environments. So...
### Please describe the purpose of the feature. Is it related to a problem? Integrate the systems' logger default into a single yaml file and move the **common** training controlling...
### Please describe what needs to be maintained? Upgrade python version, we are on 3.9, which is almost 3 years old at this point. Big reason for this is the...
## Note This is quite confusing so please check this carefully and make sure I haven't mixed up where to use each one! ## What? Mostly explained in #951. We...
### Describe the bug Title says it all
### Please describe what needs to be maintained? Title says it all
### Describe the bug Its possible there is a slight error in the resetting of hidden states for IPPO. Instead of using the done flag to reset, it would be...
## What? Implement a gym wrapper and add gym rware to make_env file, this first step is needed to support and implement Sebulba architecture on Rware. ## Why? Integrate Sebulba's...
## What? Implement an evaluator function for sebulba ff systems, and adapt the logger to support this new architecture ## Why? Integrate Sebulba's architecture due to its effectiveness in scenarios...