DCVLM Collection Raw data pools, pre-mixed ready-to-train datasets and checkpoints for DataComp-VLM. • 14 items • Updated Jul 27 • 7
view article Article Continuous batching from first principles +1 ror, ArthurZ, mcpotato • Nov 25, 2025 • 450
DataComp-VLM: Improved Open Datasets for Vision-Language Models Paper • 2606.28551 • Published Jun 26 • 52
Concept-Aware Batch Sampling Improves Language-Image Pretraining Paper • 2511.20643 • Published Nov 25, 2025 • 4
PARCEL: Pool-Anchored Resampling with Conditioned Elastic Queries for Efficient Vision-Language Understanding Paper • 2605.30126 • Published May 28 • 9