I made a video covering the essentials of Multi-modal/Visual-Language models

Hello people!

I thought it was a good time to make a video about this topic since more and more recent LLMs are moving away from text-only into visual-language domains (GPT-4, PaLM-2, etc). Multi-modal models basically input data from multiple sources (text, image, audio, video etc) to train Machine Learning tasks. In my video, I provide some intuition about this area - right from basics like contrastive learning (CLIP, ImageBind), all the way to Generative language models (like Flamingo).

Hope …

audio data etc good gpt gpt-4 image language language models llms machine machine learning machinelearningnews moving multiple palm people text thought video