🎉 Big news: LightlyTrain now supports DINOv2. Read our announcement.

Lightly Blog

LightlyTrain x DINOv2: Smarter Self-Supervised Pretraining, Faster

LightlyTrain has passed 10,000 downloads and now includes DINOv2 support for pretraining ViTs, a new Distillationv2 module with faster convergence, and strong benchmark results. Updates make it easier for teams to train foundation models on custom data without annotations.

Introducing LightlyTrain: Better Vision Models, Faster - No Labels Needed

LightlyTrain lets you pretrain vision models on your own unlabeled data — no labels needed. Improve performance on classification, detection, and segmentation tasks while reducing labeling costs and speeding up deployment. Compatible with popular architectures like YOLO, RT-DETR, and ResNet, LightlyTrain adapts models to your domain and integrates easily into existing pipelines — fully on-premise and scalable to millions of images.

LightlyOne 3.0: New Typicality-Based Selection, 6x Speedup, and Better Scalability

LightlyOne 3.0 introduces a rebuilt core for up to 6x faster performance, lower memory use, and better scalability. It features a typicality-based selection method, supports user predictions, local storage, and a cleaner UI.

NVIDIA Blackwell B200 vs H100: Real-World Benchmarks, Costs, and Why We Self-Host

The B200 is up to 57% faster for model training than the H100, up to 10x cheaper to run when self-hosted, and we’ve broken down all the costs, performance metrics, and power consumption data inside.

The Engineer's Guide to Self-Supervised Learning

Learn what self-supervised learning is and how engineers can use it to train AI models with minimal labeled data. This guide explores key techniques, real-world applications, and the benefits of self-supervised learning in computer vision and machine learning.