Parents:
LLM pre-training
Related Tags:
CloseDescendants
    1 Document (Long List) 
    Zeyuan Allen-Zhu sur X : " surprisingly, when pre-training good data (e.g., Wiki) together with "junks" (e.g., Common Crawl), LLM's capacity on good data may decrease by 20x times!"
    Tags:
    2024-04-10 About
    Properties