WeightWatcher

NN analyzer

A tool for analyzing Deep Neural Networks to predict accuracy and detect potential problems

The WeightWatcher tool for predicting the accuracy of Deep Neural Networks

GitHub

1k stars
33 watching
124 forks
Language: Python
last commit: 2 months ago

Related projects:

Repository Description Stars
hahnyuan/nn_tools A toolset for converting and analyzing neural networks across multiple frameworks. 355
matthewjdenny/ccas Provides tools for modeling and analyzing communication network data using statistical models. 5
lyronctk/zator This project verifies the inference of a deep neural network using recursive SNARKs and leverages a folding scheme to reduce computation complexity. 156
swall0w/torchstat An analyzer tool for neural networks built on PyTorch 1,468
zkp-gravity/0g A system for proving an inference pass for pre-trained neural networks on private inputs. 40
citronneur/volatility-wnf Tools for analyzing Windows Notification Facilities and related data 15
akestoridis/zigator Analyzes and manipulates data from Zigbee and Thread networks to identify security vulnerabilities and simulate attacks. 29
austin-taylor/flare Analytical framework for network traffic and behavioral analytics using Python 449
cogcomp/talen A tool for annotating named entities in low-resource languages using web-based annotation interface. 112
wkentaro/fcn An implementation of fully convolutional networks in Chainer, a deep learning framework. 218
usccana/netdiffuser Analyzes diffusion and contagion processes on networks using statistical analysis, visualization, and simulation. 86
confluentinc/confluent-sigma A tool for analyzing and visualizing log events using structured rules 52
infiziert90/getnative Determines the native resolution of upscaled material, typically anime, by applying various image processing algorithms. 221
hatriot/zarp A network attack tool designed to manage and analyze local networks 1,446
infocusp/tf_cnnvis A tool to visually analyze and understand deep learning models' internal workings, specifically convolutional neural networks. 780