|
Gaurav Menghani
|
Hi! I am currently a Software Engineer at Google Research in Mountain View, since June 2017.
My research interests include, but are not limited to Machine Learning, Algorithms, and Distributed Systems.
Currently I am focused on algorithms and techniques for Efficient Deep Learning.
Previously,
- Software Engineer at Facebook Inc., California (Feb 2013 - June 2017).
- M.S. (Computer Science), Stony Brook University, New York (Aug 2011 - Dec 2012).
- Software Engineer, Accenture, India (Jul 2010 - Jun 2011).
- B.E. (Computer Engineering), University of Mumbai, India (Aug 2006 - Jun 2010).
News (most recent at top)
- Book: Efficient Deep Learning's first seven chapters are
available for free
reading online, along with relevant
Jupyter notebooks.
- Conference Accept: "Robust Active Distillation"
was accepted in ICLR 2023.
(Link)
- Journal Publication: "Efficient Deep Learning: A Survey on Making Deep Learning Models Smaller, Faster, and Better"
was published in ACM Computing Surveys (Vol 55, No. 12). (Link)
- Patent Filing: Efficient Embedding Table Storage and Lookup
(Link)
- Conference Publication: "Weighted Distillation with Unlabeled Examples"
was published in the proceedings of NeurIPS 2022.
(Link
- Podcast: Was interviewed by
Humanitarian AI Today
about efficient deep learning, on-device models and their impact on humanitarian efforts.
(Link)
- Articles: High-Performance Deep Learning series on KDNuggets.com (Part
1,
2,
3,
4,
5).
-
Pre-Print:
Efficient Deep Learning: A Survey on Making Deep Learning Models Smaller, Faster, and Better
.
-
Talk: Efficient Deep Learning at MLConf 2021.
(Recording,
Slides).
- News Coverage of Google Cloud's Recommendation AI service. We helped optimize their server-side models.
-
Pre-Print: Gaurav Menghani, Sujith Ravi. 2019.
Learning from a Teacher using Unlabeled Data.
- News Coverage regarding our work on the Learn2Compress technology at Google AI.