profile picture

Taebum Kim

ML System Researcher

FriendliAI

Research Interests

  • Distributed Training
  • Machine Learning Compilers
  • Large-Scale Model Training & Serving

About Me

I am a researcher at FriendliAI and PhD candidate at department of Computer Science and Engineering, Seoul National University, under the guidance of Prof. Byung-Gon Chun. My research interests are machine learning systems, compilers, and distributed systems.

Education

Ph.D. in Computer Science

Sep 2019 - Aug 2023 (Expected)
Seoul National University

B.Sc. in Astronomy & Computer Science (Double Major)

Mar 2013 - Aug 2019
Seoul National University

Work Experiences

Researcher

Feb 2021 - Present
FriendliAI

MLOps / PeriFlow Training & Serving Engine

Publications

BPipe: Memory-Balanced Pipeline Parallelism for Large Language Models

Taebum Kim, Hyoungjoo Kim, Gyeong-In Yu, Byung-Gon Chun

International Conference on Machine Learning (ICML), 2023 (oral) [pdf]

Terra: Imperative-Symbolic Co-Execution of Imperative Deep Learning Programs

Taebum Kim, Eunji Jeong, Geon-Woo Kim, Yunmo Koo, Sehoon Kim, Gyeong-In Yu, Byung-Gon Chun

Neural Information Processing Systems (NeurIPS), 2021 [pdf]

Speculative symbolic graph execution of imperative deep learning programs

Eunji Jeong, Sungwoo Cho, Gyeong-In Yu, Joo Seong Jeong, Dong-Jin Shin, Taebum Kim, Byung-Gon Chun

ACM SIGOPS Operating Systems Review, Volume 53, July 2019 [pdf]

Knowledge extraction with no observable data

Jaemin Yoo, Minyong Cho, Taebum Kim, U Kang

Neural Information Processing Systems (NeurIPS), 2019 [pdf]