Cite
Notes
Only stored in your browser.
Attribution
Layer-Wise Quantization: A Pragmatic and Effective Method for Quantizing LLMs Beyond Integer Bit-Levels
arXiv 2024
from 1 papers
Mihai Surdeanu
Razvan-Gabriel Dumitru
Rishabh Maheshwary
Sathwik Tejaswi Madhusudhan
Vikas Yadav