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

Simclr

Simclr is a public, MIT-licensed PyTorch implementation of SimCLR, a contrastive self-supervised method for learning visual representations from images.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Simclr — 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 it when you want a runnable PyTorch implementation for contrastive image representation training and a documented linear-classifier evaluation workflow.

Who it is for

  • Researchers studying self-supervised visual representation learning
  • Developers who want to train image features and evaluate them with a linear classifier

Documented uses

  • Train image representations on an unsupervised image dataset.
  • Use frozen learned features as input to a downstream linear classifier.
  • Evaluate representations with the documented STL10 train-and-test protocol.
  • Debug training on CPU or enable mixed-precision GPU training.
Before you choose

Requirements & limits

The repository is public and lists an MIT license. The documented workflow installs a local Conda environment.

  • Training configuration is passed to run.py; the README example specifies a data path, dataset name, logging interval and epoch count.
  • CPU execution is documented for debugging, and --fp16_precision uses PyTorch AMP for 16-bit GPU training.
  • The feature-evaluation example uses STL10, with reported experiment results also listed for CIFAR10.

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

Set up and configure a SimCLR training run

Create the Conda environment and configure the documented training command.

Before you start

  • A local copy of the repository
  • Conda
  • An image dataset available at the path supplied to the run command

Steps

  1. Create the environment with conda env create --name simclr --file env.yml.

  2. Activate it with conda activate simclr.

  3. Configure run.py with a data path, dataset name, logging interval and epoch count, following the README example.

Expected outcome

The environment is prepared and the training command is configured with dataset and run parameters.

Guide limitations

  • The README notes that SimCLR benefits from longer training.
  • CPU mode is described for debugging; GPU mixed precision is enabled with --fp16_precision.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
15 October 2026
Original website title
sthalles/SimCLR · Official homepage title · retained in its source language
How we review sourcesSuggest a correction

Related options

PyTorch-BYOL — The repository points to this separate sthalles implementation of BYOL, another self-supervised learning approach.