OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
AI Models

Condensenet

CondenseNet is a convolutional neural network architecture and public PyTorch implementation for image classification. It combines dense layer connectivity with learned group convolutions that remove connections considered unnecessary for feature reuse, and can use standard grouped convolutions at test time.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Condensenet — OpenViST project profile in AI Models.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use the repository when you need a research implementation for training or evaluating CondenseNet, or want to study how learned group convolutions can make dense CNNs more computationally efficient.

Who it is for

  • Researchers studying compact convolutional networks and learned group convolutions.
  • Developers who want a PyTorch starting point for image-classification experiments on CIFAR-10 or ImageNet.

Documented uses

  • Train CondenseNet models using the repository’s ImageNet or CIFAR-10 examples.
  • Evaluate a trained model from the default checkpoint directory or a specified model path.
  • Convert a trained model into the group-convolution form described in the paper.
  • Compare compact model configurations using the results reported in the repository.
Before you choose

Requirements & limits

The repository is public and lists an MIT license.

  • The README specifies Python 3 and PyTorch 1.1.0 as dependencies.
  • ImageNet training requires ImageNet data; a separate CIFAR-10 training example is provided.
  • The example training commands include GPU settings, so adapt them to the available compute environment.
  • The repository reports ARM inference times for selected models, but those figures are specific to its reported setup.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Train CondenseNet on CIFAR-10

Use the repository’s documented CIFAR-10 training example.

Before you start

  • Python 3
  • PyTorch 1.1.0
  • The CondenseNet repository and CIFAR-10 data

Steps

  1. Run the repository’s documented command: the command below.

    Documented command
    python main.py --model condensenet -b 64 -j 12 cifar10 --stages 14-14-14 --growth 8-16-32 --gpu 0 --resume

  2. Use the repository’s the command below option for detailed command options.

    Documented command
    python main.py --help

Expected outcome

The documented example starts CondenseNet training on CIFAR-10 with the stated stages, growth settings and GPU selection.

Guide limitations

  • The README provides a command example, not a guarantee that it will run successfully in a particular environment.
  • The README does not give a CIFAR-10 evaluation command in the provided usage section.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
7 November 2026
Original website title
ShichenLiu/CondenseNet · Official homepage title · retained in its source language
How we review sourcesSuggest a correction