LLM AND Tips
Common descendants
3 Documents
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
DSPy Cheatsheet | DSPy
Tags:
2024-03-30 About
Nils Reimers sur X : "Embeddings can store only 1 aspect/topic per embedding well."
Tags:
2024-03-13 About