A fully open multimodal training framework and model family covering images, long and streaming video, and spatial reasoning, with data and code released end to end.
1235llava,llm,mllmA fully open multimodal training framework and model family covering images, long and streaming video, and spatial reasoning, with data and code released end to end.