Shypd.aiThe largest open index of AI toolsSimCLR is a PyTorch implementation of the SimCLR framework, based on the ICML 2020 paper "A Simple Framework for Contrastive Learning of Visual Representations." It facilitates contrastive learning for visual representation tasks.
Trending
About
SimCLR provides a PyTorch implementation of the SimCLR framework, designed for contrastive learning of visual representations. This open-source project is based on the ICML 2020 paper "A Simple Framework for Contrastive Learning of Visual Representations." It includes scripts for training SimCLR models and performing linear evaluations, primarily using the CIFAR10 dataset. Users can configure parameters such as feature dimension, temperature, batch size, and epochs. While closely following the original paper, this implementation notes some differences, including the absence of Gaussian blur, the use of Adam optimizer, and different learning rate schedules. It offers a practical foundation for researchers and developers exploring self-supervised learning in computer vision.
Capabilities
Pricing & Plans
Open Source
Free
FAQs
Trending
