Abhinav Khattar
Research Scientist at NVIDIA
California, United States
Summary
🤩
Rockstar🎓
Top SchoolAbhinav Khattar is a research scientist at NVIDIA’s Applied Deep Learning Research group with nine years of experience building and shipping NLP and speech models. He has deep expertise in NeMo and contributed major tokenizer and multilingual NMT improvements—refactoring SentencePiece integration, adding unit tests, and enabling byte-level multilingual functionality—impacting a widely used open-source generative AI framework. Prior roles include applied science internships at Amazon Alexa and Microsoft Bing and research on social media and IR at IIIT-Delhi, giving him a strong blend of production ML engineering and academic rigor. Based in California and holding degrees from IIIT-Delhi and Carnegie Mellon, he thrives at the intersection of research and engineering, turning cutting-edge language tech into robust, reusable tooling.
9 years of coding experience
4 years of employment as a software developer
Bachelor of Technology, Computer Science, Bachelor of Technology, Computer Science at Indraprastha Institute of Information Technology, DelhiMaster's degree, Computer Science, Master's degree, Computer Science at Carnegie Mellon University
English, Hindi
Stackoverflow
1reputation
0reached
0answers
0questions
Github Skills (10)
tokenize10
sentencepiece10
machine-translation10
tokenizer10
nlp10
large-language-models10
python10
generative-ai9
deep-learning9
pytorch9
Programming languages (4)
Github contributions (5)
NVIDIA/NeMo
Mar 2021 - Nov 2022
A scalable generative AI framework built for researchers and developers working on Large Language Models, Multimodal, and Speech AI (Automatic Speech Recognition and Text-to-Speech)
Role in this project:
ML Engineer
ML EngineerContributions:319 reviews, 31 commits, 122 PRs in 1 year 7 months
Contributions summary:Abhinav primarily contributed to the SentencePiece tokenizer within the Nemo framework, making significant changes to its integration and usage. These contributions included refactoring, adding unit tests, and ensuring API consistency. The user also added support for multilingual training within the NMT module. Furthermore, the user implemented byte-level multilingual NMT functionality and added special tokens to existing SentencePiece models.
asrspeech-recognitionnatural-language-processingttsspeaker-diarization
aklife97/Megatron-LM
Apr 2023 - Jul 2023
Ongoing research training transformer models at scale
Contributions:8 pushes, 1 branch in 3 months
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.