https://github.com/hizoul/3dlineriderpcg

https://github.com/hizoul/3dlineriderpcg

Last synced: about 2 months ago
JSON representation

• Host: GitHub
• URL: https://github.com/hizoul/3dlineriderpcg
• Owner: Hizoul
• Created: 2024-09-06T08:15:40.000Z (4 months ago)
• Default Branch: main
• Last Pushed: 2024-09-08T11:28:23.000Z (4 months ago)
• Last Synced: 2024-09-08T13:10:28.435Z (4 months ago)
• Language: Rust
• Size: 198 KB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# Scalable Procedural Content Generation via Transfer Reinforcement Learning

This repository contains the code to run the experiments conducted in our paper as well as re-generate the graphs based on the Replay traces.

To run the experiments, make a python weel of the rust environment:

```

maturin build --manylinux=off --release

```

and then install the resulting .whl file.

The experiments can be ran using `experiments/pysrc/main.py` and the graphs can be generated with `experiments/pysrc/plot.py`.

The environment implementation is found in `envs/linerider`. THe rest is mostly glue for Rust <=> Python interaction.

The replay traces to re-generate the graphs or verify can be found in the Github Releases of this repository.

To regenerate the graphs extract the files to a folder called `0exp` at the root of the experiment.