Ditto quantizes Code LLMs with K-Means codebooks and compiles inference via LLVM-BLAS replacement to deliver up to 10.5x faster, 6.4x smaller, and 10.5x lower-energy execution on commodity hardware while losing only 0.27% pass@1 accuracy.
Title resolution pending
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
citation-role summary
method 1
citation-polarity summary
roles
method 1polarities
use method 1representative citing papers
Sandwich delivers 2.01x average end-to-end speedup and up to 3.4x latency reduction for CPU LLM serving via phase-wise hot-switching, TopoTree hardware abstraction, and fast-start dynamic kernel generation.
citing papers explorer
-
Compiling Code LLMs into Lightweight Executables
Ditto quantizes Code LLMs with K-Means codebooks and compiles inference via LLVM-BLAS replacement to deliver up to 10.5x faster, 6.4x smaller, and 10.5x lower-energy execution on commodity hardware while losing only 0.27% pass@1 accuracy.
-
Sandwich: Joint Configuration Search and Hot-Switching for Efficient CPU LLM Serving
Sandwich delivers 2.01x average end-to-end speedup and up to 3.4x latency reduction for CPU LLM serving via phase-wise hot-switching, TopoTree hardware abstraction, and fast-start dynamic kernel generation.