Sublime
An inspiration engine for ideas
Gavin Yang
@unsung
Tengji Zhang
@tengjizhang
Tiger Gao
@tigergao
Jason Shen
@jasonshen
Exec coach helping founders and creative leaders rebound and reinvent so they do more of what matters most.
What does DeepSeek R1 & v3 mean for LLM data?
Contrary to some lazy takes I’ve seen, DeepSeek R1 was trained on a shit ton of human-generated data—in fact, the DeepSeek models are setting records for the disclosed amount of post-training data for open-source models:
- 600,000 reasoning data [1]
- 200,000 non-reasoning SFT data [2]
- human prefere... See more
Catherine Pao
@capow
Jason Huang
@err53
WJ Yang
@yangwj
Jenna
@jennawang