From the source
Intel introduces AutoRound, a new quantization tool for LLMs and VLMs that uses signed gradient descent to optimize weight rounding and clipping, supporting low-bit quantization with minimal accuracy loss.
Introducing AutoRound: Intel’s Advanced Quantization for LLMs and VLMs
Developer tool
From the source
AutoRound is a weight-only post-training quantization (PTQ) method developed by Intel.
huggingface.co
