Another lovely day in open-source AI. This time it's a 1 trillion tokens multimodal interleaved dataset. It has 3.4 billion images! Includes sources such as PDFs and ArXiv papers. https://t.co/f3j87pZkeK https://t.co/6kGJXvjpwa
— elvis (@omarsar0) Jul 24, 2024
from Twitter https://twitter.com/omarsar0
July 24, 2024 at 11:15PM
via IFTTT
No comments:
Post a Comment