Smallest Video LM Ever from HuggingFace
SmolVLM2, from HuggingFace, is a series of tiny, open-source multimodal model for video understanding. Processes video, images, and text. Ideal for on-device applications.