/learningCNF

This is the repository for the AAAI 2021 paper: Learning Branching Heuristic for Propositional Model Counting (https://www.aaai.org/AAAI21Papers/AAAI-10022.VaezipoorP.pdf)

Primary LanguagePython

Neuro#

Running the code:

Run the Docker container

  1. docker run --rm -it modelcountanon/modelcount_v1:latest /bin/bash

From within the container:

  1. tar zxf data.tgz
  2. cd sat-rl/pysat-master
  3. ./rebuild.sh
  4. ./rebuild.sh # Yes, (x2). Sorry about the embarrassement :/
  5. cd /code
  6. Run experiments with: ./run_experiments.sh [1-15] [vanilla] (Where first parameter is experiment number, second parameter is empty for the pre-trained model, "vanilla" for vanilla sharpSAT heuristics.)