Google Research’s Post

Our new regression language model approach enables LLMs to predict industrial system performance directly from raw text data. Learn how we applied it to Borg, Google’s large-scale compute infrastructure for cluster management, & copy the code. → http://goo.gle/3JaS7Ha

  • Our method uses RLMs to directly regress numerical performance metrics (y) from complex, textually represented system states (x), such as those from Google's compute clusters across diverse workloads (GMail, YouTube, Maps, etc.) and hardware (CPUs and TPUs).
Kavindu Hettiarachchi

Co-Founder & Lead R&D Engineer @SixthFlow Research | GIS | AI

1d
Like
Reply

Impressive integration of LLMs into system-level optimization—Borg just got smarter with language-based foresight.

Like
Reply
See more comments

To view or add a comment, sign in

Explore topics