-
Notifications
You must be signed in to change notification settings - Fork 2
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Compiling Issues #1
Comments
Use python2.7. I'm getting an error re Also, in case you get it working, make sure you set |
Thanks for the reply. I assume I need to re-install torch with python2.7 also? Should it be able to start training off the bat by running python main.py? Or do I need to run setup.py to compile and create the extension for pathFinder.cpp? If you could detail some of the intermediate steps to start training(eg. building the pathFinder module), that would be great! |
Yup, you need to install torch for python2.7 as well. pathFinder is already compiled for you here and you should be able to run the code out of the box on a Linux machine. Haven't tried any other OS but have you seen setup.py:
|
Yes it started training. Is it possible to run on GPU currently? Since I was having compatibility issues with Python2.7, PyTorch 1.4, and CUDA 11.4 in my conda env. Also, can you explain how you structured/represented the board you pass to the QuoridorNNet, specifically the dimensions 4xnxn of the pieces field of the Board class? Thanks. |
Yes, you can train on GPU, I trained my own model using Colab.
You're gonna have to more specific. |
If you're still interested, I updated the code so it works with python3: https://github.com/Mihonarium/alphazero-quoridor |
I am having some issues compiling and running main.py, as it is saying it can't find the pathFinder module. Can main.py be compiled in both python2 and python3?
The text was updated successfully, but these errors were encountered: