Home * Engines * Deep Pink

Deep Pink, ^[1]
an experimental open source chess engine by Erik Bernhardsson written in Python as an attempt to learn and play chess. It applies negamax alpha-beta and a deep neural network as evaluation function, using Theano ^[2], python-chess, and Sunfish ^[3] . The input representation seems similar to Octavius with 12x64 nodes, not feeding in side to move, castling rights, and en passant target square. The first hidden layer has 2048 neurons as well. Bernhardsson used a GPU instance to train the net with 100M games for about four days using stochastic gradient descent with Nesterov ^[4] momentum ^{[5] [6] [7]}.

See also

• Chess Engines with Neural Networks
• Deep Learning
• Neural MoveMap Heuristic
• Octavius

Postings

• Deep learning for… chess by Erik Bernhardsson, November 29, 2014 » Deep Learning
• Deep learning for chess Comments Hacker News
  As the author of sunfish.... by Thomas Ahle, December 2, 2014
  This has been tried many times before, with better-but-still-lackluster results.... by halfcat, Hacker News, December 2, 2014
• Deep learning for… chess (addendum) by Erik Bernhardsson, December 8, 2014
• ŷhat | Deep Learning for ... Chess by Erik Bernhardsson, February 02, 2017
• Deep Pink: a chess engine using deep learning by Chao Ma, CCC, February 05, 2017

External Links

• GitHub - erikbern/deep-pink: Deep Pink is a chess AI that learns to play chess using deep learning

References

1. ^ color: #ff1493; Deep Pink - Shades of pink from Wikipedia
2. ^ Welcome — Theano 0.7 documentation
3. ^ erikbern/deep-pink · GitHub
4. ^ Yurii Nesterov from Wikipedia
5. ^ Deep learning for… chess by Erik Bernhardsson, November 29, 2014
6. ^ ORF523: Nesterov’s Accelerated Gradient Descent by Sébastien Bubeck, I’m a bandit, April 1, 2013
7. ^ Nesterov’s Accelerated Gradient Descent for Smooth and Strongly Convex Optimization by Sébastien Bubeck, I’m a bandit, March 6, 2014

What links here?

Up one level