I am a third-year PhD student in Computer Science at State Key Laboratory of Internet of Things for Smart City (SKL-IOTSC), University of Macau under the supervision of Prof. Li Li and Prof. Chengzhong Xu. I am also fortunate to be mentored by and to collaborate closely with Dr. Zhan Shi and Prof. Zhijiang Guo. Before joining UM, I received my BEng degree from Xi’an Jiaotong University in 2022. I am broadly interested in ML for System. My current research interests focus on Efficient LLMs and On-device Learning.
Education
-
University of MacauPhD in Computer Science (MPDS awardee, AR: 30/1300+ = Top 2% )08/2022-Now
-
Xi’an Jiaotong UniversityBEng in Internet of Things (GPA ranking: Top 2)09/2018-06/2022
Selected Publications
-
[NeurIPS'24 (Oral presentation)] "HydraLoRA: An Asymmetric LoRA Architecture for Efficient Fine-Tuning." [Paper] [Code] [News (机器之心)]
- 38th Annual Conference on Neural Information Processing Systems (ML Conf., CCF A, AR: 61/15671 = Top 0.39%)
- Chunlin Tian, Zhan Shi, Zhijiang Guo, Li Li and Chengzhong Xu
-
[ICML'24] "Ranking-based Client Imitation Selection for Efficient Federated Learning." [Paper]
- 41st International Conference on Machine Learning (ML Conf., CCF A, AR: 2609/9653 = 27.0%)
- Chunlin Tian, Zhan Shi, Xinpeng Qin, Li Li and Chengzhong Xu
-
[TPDS'24] "Breaking the Memory Wall for Heterogeneous Federated Learning via Model Splitting." [Paper]
- Transactions on Parallel and Distributed Systems (System Journal, CCF A)
- Chunlin Tian, Li Li, Kahou Tam, Yebo Wu, Chengzhong Xu
-
[MICRO'22] "HARMONY: Heterogeneity-Aware Hierarchical Management for Federated Learning System." [Paper]
- 55th IEEE/ACM International Symposium on Microarchitecture (Arch. Conf., CCF A, AR: 83/369 = 22.5%)
- Chunlin Tian, Li Li, Zhan Shi, Jun Wang and Chengzhong Xu
Other Publications
-
[CICC'25] "A One-Shot Floating-Point Compute-in-Memory Macro Featuring PVT Robustness and Mismatch Tolerance for Edge LLMs."
- 46th IEEE Custom Integrated Circuits Conference (IC Conf.)
- Yuanzhe Zhao, Heng Xie, Zijian Wang, Chunlin Tian, Li Li, Yan Zhu, Rui P. Martins, Chi-Hang Chan, Minglei Zhang.
-
[Sensys'24] "FedHybrid: Breaking the Memory Wall of Federated Learning via Hybrid Tensor Management." [Paper]
- 22nd ACM Conference on Embedded Networked Sensor Systems (Mobile Conf., CCF B, AR: 58/313 = 18.5%)
- Kahou Tam, Chunlin Tian, Li Li, Haikai Zhao, Chengzhong Xu
-
[IJCAI'24] "FedGCS: A Generative Framework for Efficient Client Selection in Federated Learning via Gradient-based Optimization." [Paper] [Code]
- 33rd International Joint Conference on Artificial Intelligence (AI Conf., CCF A, AR: 791/5651 = 14.0%)
- Zhiyuan Ning*, Chunlin Tian*, Meng Xiao, Wei Fan, Pengyang Wang, Li Li, Pengfei Wang and Yuanchun Zhou
-
[IWQoS'24] "FedMG: A Federated Multi-Global Optimization Framework for Autonomous Driving."
- 32nd IEEE/ACM International Symposium on Quality of Service (Network Conf., CCF B, AR: 81/326 = 24.8%)
- Jialiang Ma, Chunlin Tian, Li Li and Chengzhong Xu
-
[IWQoS'24] "GreenLLM: Towards Efficient Large Language Model via Energy-Aware Prunning."
- 32nd IEEE/ACM International Symposium on Quality of Service (Poster) (Network Conf., CCF B)
- Chunlin Tian, Xinpeng Qin and Li Li
-
[IWQoS'24] "Heterogeneity-Aware Memory-Efficient Federated Learning via Progressive Layer Freezing."
- 32nd IEEE/ACM International Symposium on Quality of Service (Network Conf., CCF B)
- Yebo Wu, Li Li, Chunlin Tian, Tao Chang, Chi lin, Cong Wang and Chengzhong Xu
-
[IWQoS'24] "Heterogeneity-Aware Coordination for Federated Learning via Stitching Pre-trained Blocks."
- 32nd IEEE/ACM International Symposium on Quality of Service (Network Conf., CCF B)
- Shichen Zhan, Yebo Wu, Chunlin Tian, Yan Zhao and Li Li
-
[MM'24 (Oral presentation)] "When, Where, and What? A Novel Benchmark for Accident Anticipation and Localization with Large Language Models." [Paper]
- 32nd ACM International Conference on Multimedia (AI Conf., CCF A, AR: 174/4385 = Top 3.9%)
- Haicheng Liao, Yongkang Li, Chengyue Wang, Yanchen Guan, KaHou Tam, Chunlin Tian, Li Li, Chengzhong Xu, Zhenning Li
-
[MM'24] "CRASH: Crash Recognition and Anticipation System Harnessing with Context-Aware and Temporal Focus Attentions." [Paper]
- 32nd ACM International Conference on Multimedia (AI Conf., CCF A, AR: 1149/4385 = 26.2%)
- Haicheng Liao, Haoyu Sun, Huanming Shen, Chengyue Wang, KaHou Tam, Chunlin Tian, Li Li, Chengzhong Xu, Zhenning Li
-
[ICLR'23 Tiny] "Learn to Select: Efficient Cross-device Federated Learning via Reinforcement Learning."
- 11th International Conference on Learning Representations, Tiny Papers Track (ML Conf.)
- Chunlin Tian, Zhan Shi and Li Li
Professional Services
- Invited Reviewer: ICLR, NeurIPS, ICML, CVPR, IJCAI, ACM MM, AISTATS, ACM TIST
Invited Talks
-
TechBeat: Unlocking Asymmetric LoRA: Superb Performance with Significantly Fewer Parameters, 20/11/2024
-
AI Time: An Asymmetric LoRA Architecture for Efficient Fine-Tuning, 12/11/2024
Mentored Students
- Xinpeng Qin, UESTC (B.S.) ---> Tsinghua (M.S.)
- Shuaihang Zhong, UESTC (B.S.) ---> NJU (M.S.)
Awards
- 10/2024, NeurIPS 2024 Scholar Award
- 08/2022, UM Macao PhD Scholarship
- 06/2022, Outstanding Graduate, Xi'an Jiaotong University
- 09/2021, Outstanding Student, Xi'an Jiaotong University
- 10/2020, The ABU Asia-Pacific Robot Contest (ROBOCON), First Prize
Teaching Assistants
- Fall 2024, CISC7204, Data Science and Data Visualization @UM (Graduate)
- Spring 2024, CISC1001, Programming Science @UM (Undergraduate)
- Fall 2023, Fall 2022, CISC2001, Computer Organization @UM (Undergraduate)
- Spring 2023, CISC7303, Cloud Computing @UM (Graduate)