Ummm, i think you ran aground here on the whole “explain it to the people” goal.

Now, the base model itself was built using some of the best available components — an image encoder powered by a massive vision transformer, a vision-language connector, and a diverse blend of pre-training data and language models. But it was those ablation experiments that really unlocked the true potential.
Apple’s MM1 Large Language Model: The Secret Weapon Redefining the Future of AI
645
21
Akshay S B
Robert Hirsch
·Follow
Jun 20, 2024
--
Ummm, i think you ran aground here on the whole “explain it to the people” goal.


😉

(I’m kidding around, but you do still need some familiarity with llms to read this article)
--
--
Written by Robert Hirsch238 Followers
·63 Following
Author, Maker, Father, Dreamer. Robert received his Ph.D. from RPI in Mechatronics. Since then, consumer devices, renewable energy, and now blockchain.
Responses (1)
Help
Status
About
Careers
Press
Blog
Privacy
Terms
Text to speech
Teams