Gyeong-In Yu

Gyeong-In Yu

CTO of FriendliAI

About me

I am the CTO of FriendliAI. FriendliAI is a start-up with the mission of empowering innovation by lowering barriers to serving and training generative AI like GPT, T5, Stable Diffusion, and so on. My interest lies in the intersection of computer systems and machine learning, with a focus on systems for machine learning.

Before joining FriendliAI, I received my Ph.D. in Computer Science and Engineering (CSE) from Seoul National University. My primary field of study was software systems for accelerating machine learning inference in the datacenter. I received Best Ph.D. Dissertation Award from the CSE Department at Seoul National University.

Education

  • Ph.D. in Computer Science and Engineering, Seoul National University, 2017.03 - 2023.02
  • B.S. in Computer Science and Engineering / B.A. in Economics, Seoul National University, 2012.03 - 2017.02

Professional Experience

  • CTO of FriendliAI, 2023.03 -
  • Research Intern at Microsoft AI and Research, 2018.06 - 2018.09
  • Research Intern at Microsoft Research Asia, 2017.06 - 2017.09

Selected Publications

You can find my Google Scholar information here.

  1. Taebum Kim, Hyoungjoo Kim, Gyeong-In Yu, Byung-Gon Chun. BPipe: Memory-Balanced Pipeline Parallelism for Training Large Language Models. To appear in 40th International Conference on Machine Learning (ICML 2023) (Oral).
  2. Gyeong-In Yu, Saeed Amizadeh, Sehoon Kim, Artidoro Pagnoni, Ce Zhang, Byung-Gon Chun, Markus Weimer, Matteo Interlandi. WindTunnel: Towards Differentiable ML Pipelines Beyond a Single Model. 48th International Conference on Very Large Data Bases (VLDB 2022). [paper]
  3. Gyeong-In Yu, Joo Seong Jeong, Geon-Woo Kim, Soojeong Kim, Byung-Gon Chun. Orca: A Distributed Serving System for Transformer-Based Generative Models. 16th USENIX Symposium on Operating Systems Design and Implementation (OSDI 2022), July 2022. [paper]
  4. Taebum Kim, Eunji Jeong, Geon-Woo Kim, Yunmo Koo, Sehoon Kim, Gyeong-In Yu, Byung-Gon Chun. Terra: Imperative-Symbolic Co-Execution of Imperative Deep Learning Programs. 35th Conference on Neural Information Processing Systems (NeurIPS 2021), December 2021. [paper]
  5. Woosuk Kwon*, Gyeong-In Yu*, Eunji Jeong, Byung-Gon Chun (*equal contribution). Nimble: Lightweight and Efficient GPU Task Scheduling for Deep Learning. 34th Conference on Neural Information Processing Systems (NeurIPS 2020) (Spotlight), December 2020. [paper]
  6. Supun Nakandala, Karla Saur, Gyeong-In Yu, Konstantinos Karanasos, Carlo Curino, Markus Weimer, Matteo Interlandi. A Tensor Compiler Approach for One-size-fits-all ML Prediction Serving. 14th USENIX Symposium on Operating Systems Design and Implementation (OSDI 2020), November 2020. [paper]
  7. Woo-Yeon Lee, Yunseong Lee, Joo Seong Jeong, Gyeong-In Yu, Joo Yeon Kim, Ho Jin Park, Beomyeol Jeon, Wonwook Song, Gunhee Kim, Markus Weimer, Brian Cho, Byung-Gon Chun. Automating System Configuration of Distributed Machine Learning. 39th IEEE International Conference on Distributed Computing Systems (ICDCS 2019), July 2019. [paper]
  8. Soojeong Kim, Gyeong-In Yu, Hojin Park, Sungwoo Cho, Eunji Jeong, Hyeonmin Ha, Sanha Lee, Joo Seong Jeong, Byung-Gon Chun. Parallax: Sparsity-aware Data Parallel Training of Deep Neural Networks. 14th European Conference on Computer Systems (EuroSys 2019), March 2019. [paper]
  9. Eunji Jeong, Sungwoo Cho, Gyeong-In Yu, Joo Seong Jeong, Dongjin Shin, Byung-Gon Chun. JANUS: Fast and Flexible Deep Learning via Symbolic Graph Execution of Imperative Programs. 16th USENIX Symposium on Networked Systems Design and Implementation (NSDI 2019), February 2019. [paper]
  10. Gyeong-In Yu, Saeed Amizadeh, Byung-Gon Chun, Markus Weimer, Matteo Interlandi. Making Classical Machine Learning Pipelines Differentiable: A Neural Translation Approach. Systems for ML Workshop at 32nd Conference on Neural Information Processing Systems (NeurIPS), December 2018. [paper]
  11. Eunji Jeong*, Joo Seong Jeong*, Soojeong Kim, Gyeong-In Yu, Byung-Gon Chun (*equal contribution). Improving the Expressiveness of Deep Learning Frameworks with Recursion. 13th European Conference on Computer Systems (EuroSys 2018), April 2018. [paper]

Patents

  1. Gyeongin Yu, Geon-Woo Kim, Joo Seong Jeong, Soojeong Kim, Byung-Gon Chun. Selective batching for inference system for transformer-based generation tasks. US Patent 11,514,370.
  2. Gyeongin Yu, Geon-Woo Kim, Joo Seong Jeong, Soojeong Kim, Byung-Gon Chun. Dynamic batching for inference system for transformer-based generation tasks. US Patent 11,442,775.