MarcoMeter / episodic-transformer-memory-ppo

Clean baseline implementation of PPO using an episodic TransformerXL memory

MarcoMeter/episodic-transformer-memory-ppo Issues

Extending the code to support image + vector observations
Closed 5 months ago2
Question: Number of layers of the position wise MLP in transformer block.
Updated 7 months ago1
Question about attention when mask is all zero
Closed a year ago1
Questions about the implementation
Closed a year ago5
Question: Does this is a implement of your paper?
Updated a year ago1
Atari env
Closed a year ago22
Memory indices in enjoy.py
Closed a year ago1
Baseline results on GRU and LSTM on MemoryGym
Closed a year ago2
Does this repo implement a WIP paper?
Closed a year ago1
ReLU in residual connections?
Closed a year ago4