Skip to content

Instantly share code, notes, and snippets.

View jmmcd's full-sized avatar

James McDermott jmmcd

View GitHub Profile
@jmmcd
jmmcd / pg-pong.py
Last active December 1, 2016 12:04 — forked from karpathy/pg-pong.py
Training a Neural Network ATARI Pong agent with Policy Gradients from raw pixels
"""from
https://gist.github.com/karpathy/a4166c7fe253700972fcbc77e4ea32c5
http://karpathy.github.io/2016/05/31/rl/
I've converted to Python 3, added command-line args, added slightly
more informative episode status message, allowed reloading of running
reward, episode number, and elapsed time, and switched off the
per-point status message -- jmmcd.