SIMPLE
SIMPLE copied to clipboard
davidADSP
Selfplay In MultiPlayer Environments
Hi David, i'm new to data sciene and also to Python. I find SIMPLE amazing. Can you please extend SIMPLE with chess please? It would be awesome if you can...
When I try to train my frouge model after setting the number of players to 5, it can start the training. However, when it starts optimizing... , it just stops...
The current setup in `docker-compose.yml` doesn't allow the environment setup on a mac machine with M1 chip. This MR will will take care of this problem.
I left a model training overnight with docker-compose exec app mpirun -np 8 python3 train.py -e connect4 But after just an hour or so it crashed with error: A load...
After leaving the program to train overnight, it had generated in totum some 128+ models. Restarting training, the program attempts to load all of them at the start (and fails...
Per documentation on the SIMPLE blog post, the current implementation of SIMPLE only allows "Discrete" action spaces in custom environments. It would be useful to be able to use "Box"...
Would be nice to just run the scripts themself in tensor flow 2. This way it would be a lot easier to avoid the docker hassle, since tf=1.15 is not...
Is the policy loss supposed to be fluctuating here between negative and positive? What could be the problem? 
Hi ! It would be nice to be able to choose the number of players for a game, although I don't master the impact on the training loop (one model...
gym was recently [May 2022] ported over to [EvoJAX](https://github.com/google/evojax) allowing hardware-acceleration. Did anyone try using this in connection with SIMPLE?
