AWQ: Activation-aware Weight Quantization for LLM Compression and Acceleration

Ji Lin*, Jiaming Tang*, Haotian Tang, Shang Yang, Wei-Ming Chen, Wei-Chen Wang, Guangxuan Xiao, Xingyu Dang, Chuang Gan, and Song Han
MIT, Tsinghua University, MIT-IBM Watson AI Lab
(* indicates equal contribution)

News

  • 2024-05-13

    AWQ has been awarded the Best Paper of MLSys 2024!

  • 2024-02-24

    AWQ has been accepted to MLSys 2024!

  • 2024-02-01

    We released new version of quantized GEMM/GEMV kernels in TinyChat, leading to 38 tokens/second inference speed on NVIDIA Jetson Orin!

  • 2023-12-05

    AWQ is integrated by HuggingFace Transformers' main branch.

Awards

No items found.

Competition Awards

No items found.

Abstract

Large language models (LLMs) have shown excellent performance on various tasks, but the astronomical model size raises the hardware barrier for serving (memory size) and slows down token generation (memory bandwidth). In this paper, we propose Activation-aware Weight Quantization (AWQ), a hardware-friendly approach for LLM low-bit weight-only quantization. Our method is based on the observation that weights are not equally important: protecting only 1% of salient weights can greatly reduce quantization error. We then propose to search for the optimal per-channel scaling that protects the salient weights by observing the activation, not weights. AWQ does not rely on any backpropagation or reconstruction, so it can well preserve LLMs' generalization ability on different domains and modalities, without overfitting to the calibration set. AWQ outperforms existing work on various language modeling and domain-specific benchmarks. Thanks to better generalization, it achieves excellent quantization performance for instruction-tuned LMs and, for the first time, multi-modal LMs. Alongside AWQ, we implement an efficient and flexible inference framework tailored for LLMs on the edge, offering more than 3x speedup over the Huggingface FP16 implementation on both desktop and mobile GPUs. It also democratizes the deployment of the 70B Llama-2 model on mobile GPU (NVIDIA Jetson Orin 64GB).

Video

Citation

@inproceddings{lin2023awq,  title={AWQ: Activation-aware Weight Quantization for LLM Compression and Acceleration},  author={Lin, Ji and Tang, Jiaming and Tang, Haotian and Yang, Shang and Chen, Wei-Ming and Wang, Wei-Chen and Xiao, Guangxuan and Dang, Xingyu and Gan, Chuang and Han, Song},  booktitle={MLSys},  year={2024}}

Media

No media articles found.

Acknowledgment

We thank MIT AI Hardware Program, National Science Foundation, NVIDIA Academic Partnership Award, MIT-IBM Watson AI Lab, Amazon and MIT Science Hub, Qualcomm Innovation Fellowship, Microsoft Turing Academic Program for supporting this research.

Team Members