-
🔭 I’m currently working on Meituan, focusing on algorithms and engineering related to LLM inference acceleration, familar with model quantization.
-
💻 I'm lucky to contribute to some open source projects: SGLang, vLLM, TorchAO, Megatron-DeepSpeed and LightSeq.
-
🚀 I'm proud to build some projects from scratch:
- AutoSmoothQuant: An easy-to-use package for implementing SmoothQuant for LLMs.
- QQQ: QQQ is an innovative and hardware-optimized W4A8 quantization solution for LLMs.
-
📫 Contact: [email protected]
-
📚 Google Scholar: https://scholar.google.com/citations?hl=zh-CN&user=MBR97ZIAAAAJ
{{ message }}