fp.
newest
Open in hackernews
NCMNext-Concept Modeling: Fast, Parallel Coarse-to-Fine Text Generation
https://scalingintelligence.stanford.edu/blogs/ncm/
1
•
matt_d
•
45m ago