Skip to content

Instantly share code, notes, and snippets.

@gustafsson
gustafsson / pg-pong.jl
Created February 23, 2019 20:39 — forked from karpathy/pg-pong.py
Training a Neural Network ATARI Pong agent with Policy Gradients from raw pixels
# Translation of @karpathy's pg-pong.py from Python to Julia
# https://gist.github.com/karpathy/a4166c7fe253700972fcbc77e4ea32c5
# Note that it takes several days to learn a policy that wins most games.
# Trains an agent with (stochastic) Policy Gradients on Pong. Uses OpenAI Gym through Gym.jl and PyCall.jl.
using Gym
using Serialization
using StatsBase: mean, std
# hyperparameters
Här är ett exempel på hur man kan hantera textdokument med git i en webbläsare.
Man behöver inget konto för att skapa en gist eller redigera den.