1 paper
Selim An, Il hong Suh, Yeseong Kim
Quantization techniques such as BitsAndBytes, AWQ, and GPTQ are widely used as a standard method in deploying large language models but often degrades accuracy when using low-bit r…