San Francisco, California, United States
Contact Info
3K followers
500+ connections
San Francisco, California, United States
Contact Info
3K followers
500+ connections
About
Experienced with building large scale Machine Learning Systems at AWS AI, Crossing Minds…
Activity
-
Listen to Hemant Taneja talk about how he has engineered General Catalyst to be the amazing VC firm that it is. Having known him for almost 14…
Listen to Hemant Taneja talk about how he has engineered General Catalyst to be the amazing VC firm that it is. Having known him for almost 14…
Liked by Debajyoti (Debo) Datta
-
I have officially transitioned to a new country! Today is my first day as an Assistant Professor at the Department of Advanced Computing Sciences…
I have officially transitioned to a new country! Today is my first day as an Assistant Professor at the Department of Advanced Computing Sciences…
Liked by Debajyoti (Debo) Datta
-
➡ ➡ ➡ "Our work provides consistent empirical and theoretical evidence that data accumulation avoids model collapse." ⬅ ⬅️ ⬅️ THIS ^ is from an…
➡ ➡ ➡ "Our work provides consistent empirical and theoretical evidence that data accumulation avoids model collapse." ⬅ ⬅️ ⬅️ THIS ^ is from an…
Liked by Debajyoti (Debo) Datta
Experience & Education
-
Hippocratic AI
View Debajyoti (Debo)’s full experience
See their title, tenure and more.
or
Licenses & Certifications
Projects
-
Open-Advice
-
Honors & Awards
-
Amazon Research Fellowship
Amazon
-
Allen NLP Hackathon 2021
Allen Institute of AI, AI2 (https://allenai.org/)
Can we recover dataset errors and detect model limitations through weak supervision? Trained models like Roberta can have simple fixes for evaluating systematic errors learned by the model.
-
Facebook Global Pytorch Hackathon
Facebook
Learn2learn is a PyTorch library designed to make meta-learning more accessible to ML developers and researchers alike. It has examples in text, vision and reinforcement learning.
https://ai.facebook.com/blog/pytorch-1-dot-2-release-and-global-summer-hackathon/
-
Ruthie Oxford Memorial Award
Systems and Information Engineering Department, University of Virginia
This yearly award is given to the most promising graduate student of that year.
-
CIMA GLOBAL BUSINESS CHALLENGE
CIMA India
It is a global business challenge in which undergraduates from 22 different countries ( in 2012) take part.
I was the team leader of team PlanB and we were the runners up team at the western regional finals and the best business report winner in the same. -
CBSE National Science Exhibition
Central Board of Secondary Education, India
Stood 3rd in the CBSE National Science Exhibition in the disaster management category for the Earthquake detector using principles of Electromagnetic induction.
Languages
-
English
Native or bilingual proficiency
-
Bengali
Native or bilingual proficiency
-
Hindi
Professional working proficiency
More activity by Debajyoti (Debo)
-
People often ask why prices like $2.8/m token for Llama 405B, while being super fast, are still profitable at Lepton AI. We've even been asked by a…
People often ask why prices like $2.8/m token for Llama 405B, while being super fast, are still profitable at Lepton AI. We've even been asked by a…
Liked by Debajyoti (Debo) Datta
-
🌟 Excited to share our latest research on enhancing multilingual capabilities in large language models! 🌟 Introducing SPHINX, a novel multilingual…
🌟 Excited to share our latest research on enhancing multilingual capabilities in large language models! 🌟 Introducing SPHINX, a novel multilingual…
Liked by Debajyoti (Debo) Datta
-
🏆 As the Olympic season gets under way, open source takes gold at the inaugural AI Mathematical Olympiad!🥇 Today at General Catalyst, we are…
🏆 As the Olympic season gets under way, open source takes gold at the inaugural AI Mathematical Olympiad!🥇 Today at General Catalyst, we are…
Liked by Debajyoti (Debo) Datta
-
https://lu.ma/lp0gyjqr Ready for a local meetup? Please come and join our AWS x vLLM meetup in San Francisco city. We will present our exciting…
https://lu.ma/lp0gyjqr Ready for a local meetup? Please come and join our AWS x vLLM meetup in San Francisco city. We will present our exciting…
Liked by Debajyoti (Debo) Datta
-
Today marks the start of an exciting countdown for me. I'm thrilled to announce that on July 1, 2025, exactly a year from today, I will be starting…
Today marks the start of an exciting countdown for me. I'm thrilled to announce that on July 1, 2025, exactly a year from today, I will be starting…
Liked by Debajyoti (Debo) Datta
-
Check out this RAG agent that answers questions by retrieving relevant material from 10 years of US congressional legislation. Responses from great…
Check out this RAG agent that answers questions by retrieving relevant material from 10 years of US congressional legislation. Responses from great…
Liked by Debajyoti (Debo) Datta
-
When you’re bootstrapping a demo, a large, closed-source AI inference platform might be perfect for your needs. But what happens when you hone in on…
When you’re bootstrapping a demo, a large, closed-source AI inference platform might be perfect for your needs. But what happens when you hone in on…
Liked by Debajyoti (Debo) Datta
-
Thanks for featuring our MMLU-Redux paper! 🤗 This is a very comprehensive guide to AI, definitely check it out! 🚀 📜 https://lnkd.in/erZkgPiG
Thanks for featuring our MMLU-Redux paper! 🤗 This is a very comprehensive guide to AI, definitely check it out! 🚀 📜 https://lnkd.in/erZkgPiG
Liked by Debajyoti (Debo) Datta
-
AWS has optimized the PyTorch torch.compile feature for AWS Graviton3 processors, resulting in up to 2x better performance for Hugging Face model…
AWS has optimized the PyTorch torch.compile feature for AWS Graviton3 processors, resulting in up to 2x better performance for Hugging Face model…
Liked by Debajyoti (Debo) Datta
-
Thrilled to announce our LLM optimization toolbox on AWS for LLM inference: https://lnkd.in/gMzHMqCV It's a toolbox that contains endless…
Thrilled to announce our LLM optimization toolbox on AWS for LLM inference: https://lnkd.in/gMzHMqCV It's a toolbox that contains endless…
Liked by Debajyoti (Debo) Datta
-
RLOO, an RLHF method by Cohere, was implemented in Hugging Face TRL. Now, Cohere has used the TRL implementation in its latest research! This is…
RLOO, an RLHF method by Cohere, was implemented in Hugging Face TRL. Now, Cohere has used the TRL implementation in its latest research! This is…
Liked by Debajyoti (Debo) Datta
-
Training MoEs at Scale with PyTorch 🔥 In our latest post, we show how we scale to over three thousand GPUs using PyTorch Distributed and…
Training MoEs at Scale with PyTorch 🔥 In our latest post, we show how we scale to over three thousand GPUs using PyTorch Distributed and…
Liked by Debajyoti (Debo) Datta
-
🏆 Huge congratulations to Yann Fleureau, Jia Li, Hélène Evain, Stanislas Polu, Guillaume Lample for clinching top spot for the AI Mathematical…
🏆 Huge congratulations to Yann Fleureau, Jia Li, Hélène Evain, Stanislas Polu, Guillaume Lample for clinching top spot for the AI Mathematical…
Liked by Debajyoti (Debo) Datta
-
We’ve promoted Doug MacEwen to VP of SW at Glass Imaging - well deserved!! Doug has been with us since practically Day 1 and been instrumental to…
We’ve promoted Doug MacEwen to VP of SW at Glass Imaging - well deserved!! Doug has been with us since practically Day 1 and been instrumental to…
Liked by Debajyoti (Debo) Datta
Other similar profiles
Explore collaborative articles
We’re unlocking community knowledge in a new way. Experts add insights directly into each article, started with the help of AI.
Explore More