CLgen is an open source application for generating runnable programs using deep learning. CLgen learns to program using neural networks which model the semantics and usage from large volumes of program fragments, generating many-core OpenCL programs that are representative of, but distinct from, the programs it learns from.
See the online documentation for instructions on how to download and install CLgen.
Download a tiny example dataset to train and sample your first CLgen model:
$ wget https://github.com/ChrisCummins/clgen/raw/master/tests/data/tiny.tar.bz2
$ tar xf tiny.tar.bz2
$ clgen sample model.json sampler.json
Presentation slides:
Publication "Synthesizing Benchmarks for Predictive Modeling" (CGO'17).
Jupyter notebook containing experimental evaluation of CLgen.
Documentation for the Python API and command line interface.
Copyright 2016, 2017, 2018 Chris Cummins chrisc.101@gmail.com.
Released under the terms of the GPLv3 license. See LICENSE.txt for details.