https://github.com/HobbitLong/PyContrast

PyTorch implementation of Contrastive Learning methods
https://github.com/HobbitLong/PyContrast

Last synced: 5 months ago
JSON representation

PyTorch implementation of Contrastive Learning methods

• Host: GitHub
• URL: https://github.com/HobbitLong/PyContrast
• Owner: HobbitLong
• Created: 2020-05-21T00:32:53.000Z (almost 5 years ago)
• Default Branch: master
• Last Pushed: 2023-10-04T01:06:20.000Z (over 1 year ago)
• Last Synced: 2024-11-11T23:02:17.285Z (5 months ago)
• Language: Python
• Homepage:
• Size: 159 KB
• Stars: 1,939
• Watchers: 42
• Forks: 186
• Open Issues: 14
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

• awesome-machine-learning-resources - **[List

README

        
# PyContrast

This repo lists recent contrastive learning **papers**, and includes **code** for many of them.

## Paper List

Find awesome contrastive learning [paper](AWESOME_CONTRASTIVE_LEARNING.md).

## PyTorch Code

For the reference implementation of SoTA methods, (e.g., InstDis, CMC, MoCo, ...), see [pycontrast](pycontrast).

## Pre-trained Models

A set of ImageNet unsupervised pre-training models are provided. Find them in [model zoo](pycontrast/docs/MODEL_ZOO.md).

## Object Detection

Unsupervised pre-training models outperform supervised pre-training models on PASCAL VOC and COCO. Find in [detection](pycontrast/detection).