LoginSignup
2
1

More than 5 years have passed since last update.

DeepMind-Atari-Deep-Q-Learner を動かす

Last updated at Posted at 2017-01-28

description

環境

  • debian8
  • ubuntu16
  • memory 8G程度必須

install & run

sudo apt-get build-essential gcc g++ cmake curl libreadline-dev git git-core libjpeg-dev libpng-dev ncurses-dev imagemagick unzip
# 存在しないpackageは削除して実行

mkdir ~/source
cd ./source
git clone https://github.com/kuz/DeepMind-Atari-Deep-Q-Learner.git
cd DeepMind-Atari-Deep-Q-Learner
sudo bash ./install_dependencies.sh
./run_cpu breakout 

補足

  • 学習step数は, 50M

    • 1M stepでは, よくて5ブロック消す程度
  • debian8 + i7 2600k

    • 40時間で 600k step 処理
  • ubuntu16(vm) on windows + i7 3930k

    • i7 2600kの1.3~1.5倍程度 (vm上だが2600kより早い)
    • core数, quad channelが効いている?
2
1
0

Register as a new user and use Qiita more conveniently

  1. You get articles that match your needs
  2. You can efficiently read back useful information
  3. You can use dark theme
What you can do with signing up
2
1