An inference engine for generative recommendation: catalog-constrained wide beam search, executed inside CUDA graphs. https://github.com/sohu-mptc/FlashRec

1 points•vect0ry•1 day ago•0 comments•
An inference engine for generative recommendation: catalog-constrained wide beam search, executed inside CUDA graphs. https://github.com/sohu-mptc/FlashRec

0 comments

No comments yet.

Read the full thread on Hacker News →

Related stories