Moondream 2: Tiny Visual Language Model For Document Understanding

[@Gigabyteserviceofficial]

can you able to make this for mp4 video or live stream instead of image?

[@kashifrit]

its not a tiny model problem, its data format. My experience is when data is in a table LLM struggles whereas it excels when its all written in a text

[@scholarly360]

You are right. MultiModal LLMs might be the answer in the future.