Arjun Krishnakumar
Research Engineer at University of Freiburg
Freiburg im Breisgau, Baden-Württemberg, Germany
Summary
👤
Senior🎓
Top SchoolArjun Krishnakumar is a research engineer at the University of Freiburg with 11 years of software engineering experience spanning industry and academia, including a developer role at SAP Labs. He focuses on neural architecture search and deep learning tooling, contributing to prominent open-source projects like NASLib where he improved DARTS search space implementations and architecture handling for benchmarks such as TransBench101. Comfortable moving between research code and production-grade fixes, he has a strong academic foundation with an M.Sc. in Computer Science from Freiburg and a history of tutoring and research assistance. Based in Freiburg, he brings a pragmatic blend of engineering rigor and research curiosity, often spotting and resolving subtle graph- and pooling-related bugs that improve reproducibility in NAS research.
10 years of coding experience
7 years of employment as a software developer
Master of Science, Computer Science, 1.4, Master of Science, Computer Science, 1.4 at The University of FreiburgBachelor of Technology - BTech, Computer Science, 77%, Bachelor of Technology - BTech, Computer Science, 77% at Model Engineering CollegeThe Choice School
English, Malayalam, Hindi, German
Github Skills (8)
pytorch10
machine-learning10
python10
nas10
neural-architecture-search10
graph-theory9
deep-learning9
networkx8
Programming languages (2)
Github contributions (5)
automl/NASLib
Jul 2021 - Sep 2022
NASLib is a Neural Architecture Search (NAS) library for facilitating NAS research for the community by providing interfaces to several state-of-the-art NAS search spaces and optimizers.
Role in this project:
ML Engineer
ML EngineerContributions:24 reviews, 383 commits, 41 PRs in 1 year 2 months
Contributions summary:Arjun primarily contributed to the development and maintenance of the NASLib library, a tool for neural architecture search (NAS). Their work focused on fixing and improving the DARTS search space implementation, specifically addressing issues with pooling layers and graph conversions. The user also updated tests and addressed graph creation bugs within the DARTS search space. Additionally, they added code for creating and handling various architectures, including those used by TransBench101.
artnasneural-architecture-searchstate-of-the-artmachine-learning
Neonkraft/TorchBug
Oct 2021 - Nov 2021
Contributions:23 commits, 4 pushes in 1 month
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.