Skip to main content
Join the official 2020 Python Developers SurveyStart the survey!

dnnv - deep neural network verification

Project description

Deep Neural Network Verification Toolbox

Tools for verification and analysis of deep neural networks. Watch our demo video for a quick description!

Getting Started

For more detailed instructions, see our documentation.


Clone this network:

$ git clone

Create a python virtual environment for this project:

$ ./ init

To activate the virtual environment and set environment variables correctly for tools installed using the provided script, run:

$ . .env.d/

Install any of the supported verifiers (Reluplex, planet, MIPVerify.jl, Neurify, ERAN):

$ ./ install reluplex planet mipverify neurify eran

Make sure that the project environment is activated when installing verifiers with the script. Otherwise some tools may not install correctly.

Additionally, several verifiers make use of the Gurobi solver. This should be installed automatically, but requires a license to be manually activated and available on the host machine. Academic licenses can be obtained for free from the Gurobi website.

Finally, planet has several additional requirements that currently must be installed separately before installation with ./ libglpk-dev, qt5-qmake, valgrind, libltdl-dev, protobuf-compiler.


Properties are specified in our property DSL, extended from Python. A property specification can import python modules, and define variables. The only required component is the property expression, which must appear at the end of the file. An example of a local robustness property is shown below.

from import *

N = Network("N")
x = Image("path/to/image")
epsilon = Parameter("epsilon", float, default=1.0)

        ((x - epsilon) < x_ < (x + epsilon)),
        argmax(N(x_)) == argmax(N(x))),

To check whether property holds for some network using the ERAN verifier, run:

$ python -m dnnv network.onnx property.prop --eran

Additionally, if the property defines parameters, using the Parameter keyword, they can be specified on the command line using the option --prop.PARAMETER_NAME, where PARAMETER_NAME is the name of the parameter. For the property defined above, a value for epsilon can be provided with a command line option as follows:

$ python -m dnnv network.onnx property.prop --eran --prop.epsilon=2.0

A set of example networks and properties that can be run with DNNV are available here.


This material is based in part upon work supported by the National Science Foundation under grant number 1900676.

Project details

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Files for dnnv, version 0.3.0
Filename, size File type Python version Upload date Hashes
Filename, size dnnv-0.3.0-py3-none-any.whl (75.4 kB) File type Wheel Python version py3 Upload date Hashes View
Filename, size dnnv-0.3.0.tar.gz (9.1 MB) File type Source Python version None Upload date Hashes View

Supported by

Pingdom Pingdom Monitoring Google Google Object Storage and Download Analytics Sentry Sentry Error logging AWS AWS Cloud computing DataDog DataDog Monitoring Fastly Fastly CDN DigiCert DigiCert EV certificate StatusPage StatusPage Status page