radq

I've been working on training this small vision language model for the last month - excited to release the first prototype today! It is based on SigLIP (image encoder), Phi-1.5 (text model) and trained using the…

I found the mixed reactions to Llama 2 interesting, with some people concerned about the safety implications of open-sourcing a powerful LLM, while others were finding the chat RLHF version overly cautious (e.g. won't…