Starting from 2024, I’m a research assistant professor in the NICS-EFC group at the Department of Electronic Engineering, Tsinghua University. I obtained my B.S. and Ph.D. degrees from the Department of Electronic Engineering, Tsinghua University, in 2016 and 2021, advised by Prof. Huazhong Yang and Prof. Yu Wang. I spent two years (from December 2021 to December 2023) as a post-doctoral researcher with Prof. Yu Wang and Prof. Pinyan Lu.
I built up and had been leading the Efficient Deep Learning Algorithm (EffAlg) Team in the NICS-EFC group till the end of 2024. Our team is continuously looking for self-motivated master or Ph.D. students, postdoctoral scholars, research assistants, and visiting students. NOTE in 2025: After nearly five years of building the team from the ground up, the team is entering a new phase in 2025! From now on, due to my personal arrangement, I am not able to advise or participate in quite many projects within my team. Despite that, I think that we have already established strong pipelines, a comprehensive knowledge base, a solid platform, and more importantly, a good group of people for new students to find their collaboration and grow in our team. Therefore, our team continues to welcome new students.Tianchen will be in charge of the team from 2025. Please send Tianchen and Prof. Wang an email with your CV if you’re interested!
Services
Selected Talks
-
Zhiyuan's Annual Discussion on the 10 AI Trends - Inference Optimization Slide
A talk to summarize the key work and my opinions on the near-term trends at the start of 2025.
-
Introduction to NICS-EFC Lab Efficient Algorithm Team Slide Video
A talk to introduce our EffAlg Group. @AI Time.
-
Generative Model Compression and Acceleration Slide Video
A talk on generative model compression and acceleration. @Huawei; Apple-China; AMD-China; VIVO; University of Chinese Academy of Sciences; SCUT; and others.
-
An Introduction to Quantization of Large Language Models Slide Video
A talk about efficient LLM with a special focus on quantization for an Competition organized by AWS-China.
-
Model Compression Towards Efficient Deep Learning Inference Slide
A talk on model compression towards efficient DL inference. @Huawei; Inceptio.ai; Beihang University; and others.
-
Neural Architecture Search and Architecture Encoding Slide
A talk on NAS researches. @DAMO Academy of Alibaba Group (U.S.); Renmin University of China; and others.
-
A Simple Survey on Auto-Parallelism Slide
An internal Chinese survey on auto parallelism methods in 2022.