SlitheRL
A tensorized multiplayer Snake environment for deep Q-learning (with Kaiying Hou, 2021)
A multiplayer Snake environment written to run many games at once as batched tensors, so a deep Q-learning agent can train against itself and other agents without a Python-loop bottleneck. Built with Kaiying Hou in 2021.
Write-up and results: kaiying.neocities.org/slitherl.html