Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/locuslab/satnet
Bridging deep learning and logical reasoning using a differentiable satisfiability solver.
https://github.com/locuslab/satnet
Last synced: 5 days ago
JSON representation
Bridging deep learning and logical reasoning using a differentiable satisfiability solver.
- Host: GitHub
- URL: https://github.com/locuslab/satnet
- Owner: locuslab
- License: mit
- Created: 2019-05-13T18:41:09.000Z (over 5 years ago)
- Default Branch: master
- Last Pushed: 2022-11-22T09:22:51.000Z (about 2 years ago)
- Last Synced: 2024-12-22T03:06:07.575Z (12 days ago)
- Language: Python
- Homepage:
- Size: 483 KB
- Stars: 409
- Watchers: 36
- Forks: 52
- Open Issues: 6
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# SATNet • [![PyPi][pypi-image]][pypi] [![colab][colab-image]][colab] [![License][license-image]][license]
[license-image]: https://img.shields.io/badge/License-MIT-yellow.svg
[license]: LICENSE
[pypi-image]: https://img.shields.io/pypi/v/satnet.svg
[pypi]: https://pypi.python.org/pypi/satnet
[colab-image]: https://colab.research.google.com/assets/colab-badge.svg
[colab]: https://colab.research.google.com/drive/1dRfepPLEE8N6BBZhXz8bbLDcPnRKaOcJ#forceEdit=true&offline=true&sandboxMode=true
*Bridging deep learning and logical reasoning using a differentiable satisfiability solver.*
This repository contains the source code to reproduce the experiments in the ICML 2019 paper [SATNet: Bridging deep learning and logical reasoning using a differentiable satisfiability solver](https://arxiv.org/abs/1905.12149) by [Po-Wei Wang](https://powei.tw/), [Priya L. Donti](https://priyadonti.com/), [Bryan Wilder](http://teamcore.usc.edu/people/bryanwilder/default.htm), and [J. Zico Kolter](http://zicokolter.com/).
## What is SATNet
SATNet is a differentiable (smoothed) maximum satisfiability (MAXSAT) solver that can be integrated into the loop of larger deep learning systems. This (approximate) solver is based upon a fast coordinate descent approach to solving the semidefinite program (SDP) associated with the MAXSAT problem.
#### How SATNet works
A SATNet layer takes as input the discrete or probabilistic assignments of known MAXSAT variables, and outputs guesses for the assignments of unknown variables via a MAXSAT SDP relaxation with weights *S*. A schematic depicting the forward pass of this layer is shown below. To obtain the backward pass, we analytically differentiate through the SDP relaxation (see the paper for more details).
#### Overview of experiments
We show that by integrating SATNet into end-to-end learning systems, we can learn the logical structure of challenging problems in a minimally supervised fashion. In particular, we show that we can:
* Learn the **parity function** using single-bit supervision (a traditionally hard task for deep networks)
* Learn how to play **9×9 Sudoku (original and permuted)** solely from examples.
* Solve a **"visual Sudoku"** problem that maps images of Sudoku puzzles to their associated logical solutions. (A sample "visual Sudoku" input is shown below.)
## Installation
### Via pip
```bash
pip install satnet
```
### From source
```bash
git clone https://github.com/locuslab/SATNet
cd SATNet && python setup.py install
```
#### Package Dependencies
```
conda install -c pytorch tqdm
```
The package also depends on the nvcc compiler. If it doesn't exist (try nvcc from commandline), you can install it via
```
conda install -c conda-forge cudatoolkit-dev
```
### Via Docker image
```bash
cd docker
sh ./build.sh
sh ./run.sh
```
## Running experiments
### Jupyter Notebook and Google Colab
[Jupyter notebook](https://github.com/locuslab/SATNet/blob/master/notebooks/Learning%20and%20Solving%20Sudoku%20via%20SATNet.ipynb)
and [Google Colab](https://colab.research.google.com/drive/1dRfepPLEE8N6BBZhXz8bbLDcPnRKaOcJ#forceEdit=true&offline=true&sandboxMode=true)
### Run them manually
#### Getting the datasets
The [Sudoku dataset](https://powei.tw/sudoku.zip) and [Parity dataset](https://powei.tw/parity.zip) can be downloaded via
```bash
wget -cq powei.tw/sudoku.zip && unzip -qq sudoku.zip
wget -cq powei.tw/parity.zip && unzip -qq parity.zip
```
#### Sudoku experiments (original, permuted, and visual)
```bash
python exps/sudoku.py
python exps/sudoku.py --perm
python exps/sudoku.py --mnist --batchSz=50
```
#### Parity experiments
```bash
python exps/parity.py --seq=20
python exps/parity.py --seq=40
```