File size: 844 Bytes
8069557 7c73a10 8069557 7c73a10 8069557 7c73a10 8069557 7c73a10 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 |
---
title: VPTQ Demo
emoji: 🚀
colorFrom: blue
colorTo: green
sdk: static
pinned: true
license: mit
short_description: Vector Post Training Quantization Inference Demo
---
Vector Post-Training Quantization (VPTQ) is a novel Post-Training Quantization method that leverages Vector Quantization to high accuracy on LLMs at an extremely low bit-width (<2-bit). VPTQ can compress 70B, even the 405B model, to 1-2 bits without retraining and maintain high accuracy.
* Better Accuracy on 1-2 bits, (405B @ <2bit, 70B @ 2bit)
* Lightweight Quantization Algorithm: only cost ~17 hours to quantize 405B Llama-3.1
* Agile Quantization Inference: low decode overhead, best throughput, and TTFT
[Github/Codes](https://github.com/microsoft/VPTQ)
[Online Demo](https://huggingface.co/spaces/microsoft/VPTQ)
[Paper](https://arxiv.org/abs/2409.17066)
|