Snake AI

This is the source code for my bacholor thesis project.

Build instructions:

Install miniconda or anaconda
Clone this repository
Run conda env create -f ./environment.yml in this repository to create the environment aeduin-rl
Run conda activate aeduin-rl
Install [https://github.com/adambielski/GrouPy]. Make sure you install it inside the aeduin-rl environment by running the command above if you open an new shell.

Execute python train.py to train an AI. Change the model_name variable in the source code to change which network architecture is used.
If you want to see it play: in visualize_play.py, modify the model = and model.load_state_dict(...) variables (train.py automatically outputs trained state dicts in model_output/). Then run python visualize_play.py

Name		Name	Last commit message	Last commit date
Latest commit History 66 Commits
.vim		.vim
.gitignore		.gitignore
README.md		README.md
environment.yml		environment.yml
eval.zsh		eval.zsh
eval_many.py		eval_many.py
get_num_params.py		get_num_params.py
main.py		main.py
models.py		models.py
simulation.py		simulation.py
test_equivariance.py		test_equivariance.py
tests.txt		tests.txt
tests_compressed.txt		tests_compressed.txt
train.py		train.py
visualise_equivariance.py		visualise_equivariance.py
visualise_play.py		visualise_play.py