MLVS (Machine Learning Verification Service)

MLVS verifies a given neural network for a given property specification

Verifies using two tools - FFN and NNENUM

Getting Started

---

1. clone MLVS repository

     git clone https://github.com/DMoumita/MLVS.git
   

2. Entering into MLVS directory

     cd MLVS 
   

3-a. Run using Docker

#Intall Docker Engine - please refer https://docs.docker.com/engine/install/ubuntu/
#The Dockerfile in FFN folder shows how to install all the dependencies (mostly python and numpy packages) and set up the environment. 

. To build and run mlvs image

sudo docker build . -t mlvs_image 

To get a shell after building the image:

sudo docker run -i -t mlvs_image bash

Run a script without entering in to the the shell:

sudo docker run -i  mlvs_image 

3-b. [Tested for Ubuntu 16.04 and 18.04] Run in local server without docker image

   pip install -r requirements.txt
   cd MLVS
   python mlvs.py

To execute on local server:

http://127.0.0.1:5000/mlvs

Sample Output :

Verify :

To check on AWS EC2 instance:

http://ec2-3-128-200-136.us-east-2.compute.amazonaws.com:8080/mlvs