Git Product home page Git Product logo

epigrad's Introduction

Test-Time Gradient Based OOD

To run a set of experiments, specify the datasets, number of tests, and whether to use deep variants in variant.py. This code expects to use an MLFlow remote server, whose uri is also specified in variant.py. Specify the methods to compare in main.py using epistemic_test_functions, and then run

python -m pudb main.py "MLFlow_Experiment_Name" "MLFlow_run_prefix" "MLFlow note"

Code is organised as follows:

.
├── .gitignore                   
├── README.md                   
├── epistemic_tests.py                      # Where all the gradient-based OOD methods are defined
├── helpers.py                              # Helper functions for plotting, AUC calculation etc.
├── main.py                                 # Main script to run gradient-based OOD experiments
├── main_ortho.py                           # Main script to run Orthonormal Certs experiments
├── orthonormal_certs.py                    # Where Orthonormal Certs are defined
└── roc.py                                  # Functions to caluclate ROC

Hyperparameters should be specified in a file called variant.py in the top directory, and its contents should look like:

variant = dict(
    mlflow_uri="http://128.2.210.74:8080",
    num_tests= 10000,                             # Number of ID or OOD images to use when testing each method
    deep=False,                                   # Whether to use entire network or just last layer
    model_names=['mnist', 'svhn', 'cifar10'],     # Which datasets & pretrained networks to use
    )

epigrad's People

Contributors

cvigoe avatar

Watchers

 avatar  avatar  avatar

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.