Moondream1

https://github.com/cocktailpeanut/moondream1updated 1/23/2024, 9:44:08 PMindexed 1/20/2026, 9:12:03 AMOwner@cocktailpeanut

moondream1 is a tiny (1.6B parameter) vision language model trained by @vikhyatk that performs on par with models twice its size. It is trained on the LLaVa training dataset, and initialized with SigLIP as the vision tower and Phi-1.5 as the text encoder. https://huggingface.co/spaces/vikhyatk/moondream1