3 papers
cs.CV2025
Should VLMs be Pre-trained with Image Data?
Sedrick Keh, Jean Mercat, Samir Yitzhak Gadre +8
Pre-trained LLMs that are further trained with image data perform well on vision-language tasks. While adding images during a second training phase effectively unlocks this capabil…
cs.LG2024
DataComp-LM: In search of the next generation of training sets for language models
Jeffrey Li, Alex Fang, Georgios Smyrnis +56
We introduce DataComp for Language Models (DCLM), a testbed for controlled dataset experiments with the goal of improving language models. As part of DCLM, we provide a standardize…
cs.CL2024
Language models scale reliably with over-training and on downstream tasks
Samir Yitzhak Gadre, Georgios Smyrnis, Vaishaal Shankar +22
Scaling laws are useful guides for derisking expensive training runs, as they predict performance of large models using cheaper, small-scale experiments. However, there remain gaps…