Radio assigns bit depths to LLM weight groups by minimizing a rate-distortion objective via dual ascent, improving perplexity at 2-4 bits compared with GPTQ, AWQ, and OWQ.
Loss-aware quantization methods (Hou & Kwok, 2018; Nahshan et al., 2020; Qu et al.,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Radio: Rate-Distortion Optimization for Large Language Model Compression
Radio assigns bit depths to LLM weight groups by minimizing a rate-distortion objective via dual ascent, improving perplexity at 2-4 bits compared with GPTQ, AWQ, and OWQ.