AWQ: Activation-aware Weight Quantization for LLM Compression and Acceleration
Ji Lin, Jiaming Tang, Haotian Tang, Yang Shang, Xingyu Dang +5 more
Abstract
Large language models (LLMs) have transformed numerous AI applications. On-device LLM is becoming increasingly important: running LLMs locally on edge devices can reduce the cloud computing cost and protect users' privacy. However, the astronomical model size and the limited hardware resource pose significant deployment challenges. We propose Activation-aware Weight Quantization (AWQ), a hardware-friendly approach fo...
Results & Benchmarks
Benchmark data is not yet available for this paper.
Hardware Requirements
- Expect multi-day setup/compute for meaningful reproduction based on current guidance.
Best Implementation
Maintained implementation evidence is not confirmed for this paper yet.
Use the Implementation Status and Reproduction Path sections below for the current action plan.
Reproduction Path
Follow this baseline workflow to decide if this paper is worth immediate prototyping.
- 1
Use the paper and benchmark evidence to scope a baseline reproduction plan.
- 2
Start from related paper: Another view on parallel speedup.
- 3
Start from this likely method family: Quantization (signal processing).
- 4
Track assumptions and missing details in an experiment log before coding.
Additional Implementations
No additional verified repositories beyond the primary recommendation.
Hugging Face Artifacts
No trustworthy direct or curated related Hugging Face artifacts were found yet.
Continue with targeted Hugging Face searches:
Research Context
Tasks
Domains
Citations
74
Total citations
Related papers
- Another view on parallel speedup
- Shared virtual memory and generalized speedup
- Performance considerations of shared virtual memory machines
- Toward a better parallel performance metric
- Speedup for Multi-Level Parallel Computing
- Once Quantization-Aware Training: High Performance Extremely Low-bit Architecture Search