GitHub / mwritescode / nnclr-cifar100

We implement NNCLR and a novel clustering-based technique for contrastive learning that we call KMCLR. We show that applying a clustering technique to obtain prototype embeddings and using these prototypes to form positive pairs for contrastive loss can achieve performances on par with NNCLR on CIFAR-100 while storing 0.4% of the number of vectors.

JSON API: http://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/mwritescode%2Fnnclr-cifar100
PURL: pkg:github/mwritescode/nnclr-cifar100

Topics: cifar-100, clustering, contrastive-learning, k-means, nnclr, self-supervised-learning

• Files
• Readme
• Dependencies