GitHub / SamRagusa / Batch-First
A JIT compiled chess engine which traverses the search tree in batches in a best-first manner, allowing for neural network batching, asynchronous GPU use, and vectorized CPU computations.
JSON API: http://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/SamRagusa%2FBatch-First
PURL: pkg:github/SamRagusa/Batch-First
Stars: 28
Forks: 5
Open issues: 1
License: gpl-3.0
Language: Python
Size: 450 KB
Dependencies parsed at: Pending
Created at: about 8 years ago
Updated at: 8 months ago
Pushed at: over 3 years ago
Last synced at: 7 months ago
Topics: ann, artificial-intelligence, batch, chess, chess-ai, convolutional-neural-networks, engine, gpu, inception-module, jit, jit-compiler, k-best-first, llvm, negamax-algorithm, neural-network, numba, numpy, tensorflow, tree-pruning, tree-search