Skip to content

sandialabs/cross-sim-models

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

2 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

CrossSim Models

This repository contains pre-trained Keras model files for CrossSim. This is not a self-executing repository. Please see the main CrossSim repository for more information about the tool.

The following pre-trained models are available in this repository for use with CrossSim Inference:

  • MNIST:
    • Two shallow CNNs with different numbers of channels, trained by CrossSim's developers
  • CIFAR-10: all models were trained by CrossSim's developers
    • Four ResNets of different depths, following the topology specified in the original ResNet paper: ResNet14, ResNet20, ResNet32, ResNet56
    • A shallow CNN that uses bounded ReLU (cifar10_cnn_brelu)
    • A sequential CNN with binary weights and activations using the Larq package
    • A sequential CNN with binary activations using the Whetstone package
  • CIFAR-100:
    • A ResNet56 network, trained by CrossSim's developers
  • ImageNet: all models were obtained from external sources

About

Pre-trained neural network models for CrossSim

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published