An instruction-tuned version of the 90 billion-parameter vision model in Meta's Llama 3.3 series.
Have a question? Noticed something wrong? Let us know.
If you would like a reply, please include your name and email address.