yalidu / liir

Learning Individual Intrinsic Reward in MARL
https://papers.nips.cc/paper/8691-liir-learning-individual-intrinsic-reward-in-multi-agent-reinforcement-learning
62 stars 19 forks source link

ValueError: Unknown game version: 4.1.4. Known versions: ['latest']. #3

Closed GJHall closed 4 years ago

GJHall commented 4 years ago

I have been able to get SMAC installed and working to train COMA/QMIX agents.

When I tried to run your code the first time I received this error:

File "src/main.py", line 33, in my_main env_args['seed'] = _config["seed"] sacred.utils.SacredError: The configuration is read-only in a captured function!

To fix this I added the following code to the main.py file: SETTINGS['CONFIG']['READ_ONLY_CONFIG'] = False

now I am receiving:

ValueError: Unknown game version: 4.1.4. Known versions: ['latest'].

This is the same error I receive when I try to run the Variance Base Control code from

Any ideas?

Thank you for your time.

Edit: I am using version 4.11.3

yalidu commented 4 years ago

Hi, @GJHall , I have not met errors on the _config['seed']. This repo is developed based on pysc2=2.0.2. Pls make sure that the version of SMAC is also compatible.

As to the game version, I met the same value error before, but this disappears after I set the game version in sc2.yaml consistent with the version of the SC2 game core that is installed on the server.

Hope these will help! Thanks!

GJHall commented 4 years ago

I think the issue has to do with:

A) Windows B) s2client-proto not offering version 4.11.x hashed ()

I tried changing the sc2.yaml file to the right version but I do not think windows looks at this config line. Still get the same error.

I will try playing around with it on my linux system.

Closing the issue and will update if I find a solution.

Thank you for your quick response.

yalidu commented 4 years ago

I have not tested the repo on Windows system. Not sure if it is compatible. ;(