Edan Toledo

Feature/jittable agent ids

3

## What? Change agent ids to use EntityID dataclass - this dataclass can be easily used to represent an agent and is jittable. The dataclass provides utility functions to convert...

## What? New components were added to use JAX Environments - these components are the Jax environment loop and the Jax executor environment loop. Additionally, a seed value can now...

### Describe the bug In the MADDPG system, regardless of the net spec keys given to the network_factory (i.e create_default_networks), the system overwrites these values. This causes issues when trying...

[BUG] Optimizers for critic and policy both backpropagating on observation network

1

**Describe the bug** Both optimizers train the observation network separately which can caused the representation learnt to be unstable. Happens in MAD4PG. **Expected behavior** only one optimizer should work on...

**Please describe the purpose of the feature. Is it related to a problem? ** Adding Noisy Networks as an additional choice for exploration in MADQN - Noisy Networks have seen...

Question: Performance of multi-device prefetching

2

Hello, I have a quick question with regards to the implementation of the multi-device put and prefetching iterator. From looking at the code, it seems that there is quite a...

Title says it all. just a question as I'm worried about potential JIT compilation issues that might occur with haiku state etc.

Hello, I think there might be a bug in the node aggregation fn when using GCN. The docstrings state that we can set the senders and receivers to None for...

### Describe the bug Title says it all

### Please describe what needs to be maintained? Title says it all