Cupertino, California, United States
Contact Info
7K followers
500+ connections
About
Articles by Ash
-
The New Healthcare Stack™ : #makinghealthcarehuman
The New Healthcare Stack™ : #makinghealthcarehuman
By Ash Damle
-
Transform the list. Transform healthcare today.
Transform the list. Transform healthcare today.
By Ash Damle
Activity
-
Impressive. Groq now runs OpenAI's Whisper-large-v3 164x faster for $0.03 per hour of transcription. Here's the code: from: @satvikps ↓ Check…
Impressive. Groq now runs OpenAI's Whisper-large-v3 164x faster for $0.03 per hour of transcription. Here's the code: from: @satvikps ↓ Check…
Liked by Ash Damle
-
LazyLLM is a training-free technique that dynamically and progressively prunes tokens at each layer of an LLM, reducing FLOPs and making inference…
LazyLLM is a training-free technique that dynamically and progressively prunes tokens at each layer of an LLM, reducing FLOPs and making inference…
Liked by Ash Damle
-
Top 10 resources for Early-Stage startup founders: 1) Tool from Y Combinator to find co-founders - https://lnkd.in/dnFSvph7 2) Top document…
Top 10 resources for Early-Stage startup founders: 1) Tool from Y Combinator to find co-founders - https://lnkd.in/dnFSvph7 2) Top document…
Liked by Ash Damle
Experience & Education
Honors & Awards
-
HIMSS SoCal: Keynote Speaker
-
AI for Quality and Risk Management: https://www.youtube.com/watch?v=XlWys-XQ2jk
-
MIT: Future of Industries : Leveraging AI in Healthcare : Featured Speaker
MIT CNC
-
BCBSA AI Event: Keynote Speaker
Blue Cross Blue Shield Association
-
NVIDIA GPU Tech Conference : Making AI Work in Healthcare: Speaker
NVIDIA
-
UMich Innovation Summit : Featured Speaker
University of Michigan SPH
-
NVIDIA Event: AI and Machine Learning: Powering the Next Big Data Revolution : Keynote Speaker
NVIDIA
Languages
-
Spanish
-
-
Cantonese (conversational)
-
Recommendations received
4 people have recommended Ash
Join now to viewMore activity by Ash
-
Did you know that Claude 3.5 Sonnet has a hidden 'thinking' step that helps it perform better? You can uncover it by getting it to replace < >…
Did you know that Claude 3.5 Sonnet has a hidden 'thinking' step that helps it perform better? You can uncover it by getting it to replace < >…
Liked by Ash Damle
-
Q-Sparse: All Large Language Models can be Fully Sparsely-Activated Recently, considerable research work has been going towards reducing high…
Q-Sparse: All Large Language Models can be Fully Sparsely-Activated Recently, considerable research work has been going towards reducing high…
Liked by Ash Damle
-
Really impressed with Kevin Maik Jablonka's new review article on extracting and processing data with large language models in materials science. Not…
Really impressed with Kevin Maik Jablonka's new review article on extracting and processing data with large language models in materials science. Not…
Liked by Ash Damle
-
Accelerating LLMs by 2x with Graph-structured Speculative Decoding. Researchers have found a way to make speculative decoding up to 2x faster by…
Accelerating LLMs by 2x with Graph-structured Speculative Decoding. Researchers have found a way to make speculative decoding up to 2x faster by…
Liked by Ash Damle
-
🎲 Visualize and interact with LLM decoding strategies LLM sampler is a cool website that allows you to see the effect of temperature, top K…
🎲 Visualize and interact with LLM decoding strategies LLM sampler is a cool website that allows you to see the effect of temperature, top K…
Liked by Ash Damle
-
Medicare Advantage Dental Benefits - the devils are in the details. This morning, I ingested the multiyear-worth of Medicare Advantage Supplemental…
Medicare Advantage Dental Benefits - the devils are in the details. This morning, I ingested the multiyear-worth of Medicare Advantage Supplemental…
Liked by Ash Damle
-
Yeah! There are so many things to be learned from Andrej Karpathy's work on building GPT-2 from scratch. You will start with building GPT…
Yeah! There are so many things to be learned from Andrej Karpathy's work on building GPT-2 from scratch. You will start with building GPT…
Liked by Ash Damle
-
Will the AI scribe market follow the downward trajectory of e-scooters? Automated documentation (ambient note-writing) is a great AI use case…
Will the AI scribe market follow the downward trajectory of e-scooters? Automated documentation (ambient note-writing) is a great AI use case…
Liked by Ash Damle
-
Lyft used graph embeddings to improve riding experience for 20M+ riders 🚙 — by analysing interactions with drivers, locations, and even time! Here’s…
Lyft used graph embeddings to improve riding experience for 20M+ riders 🚙 — by analysing interactions with drivers, locations, and even time! Here’s…
Liked by Ash Damle
-
NEFTune is a very simple technique that, when used along with Supervised Finetuning or Instruction Tuning, can lead to substantial gains in the…
NEFTune is a very simple technique that, when used along with Supervised Finetuning or Instruction Tuning, can lead to substantial gains in the…
Liked by Ash Damle
-
Google DeepMind researchers have shared a new AI architecture called PEER that uses over a million small "experts" as tiny neural networks instead of…
Google DeepMind researchers have shared a new AI architecture called PEER that uses over a million small "experts" as tiny neural networks instead of…
Liked by Ash Damle
-
(2/2) Retrieval with learned similarities, or: how to get compute for free Our previous post introduced Generative Recommenders (GRs) and Scaling…
(2/2) Retrieval with learned similarities, or: how to get compute for free Our previous post introduced Generative Recommenders (GRs) and Scaling…
Liked by Ash Damle
-
Data is critical for successful public-private partnerships, and I can't think of a better event than this one. I can't wait to nerd about health…
Data is critical for successful public-private partnerships, and I can't think of a better event than this one. I can't wait to nerd about health…
Liked by Ash Damle
-
A model with 1 trillion parameters dropped on Hugging Face. How do you scale up to 1T parameters without breaking the bank? Here are the juicy…
A model with 1 trillion parameters dropped on Hugging Face. How do you scale up to 1T parameters without breaking the bank? Here are the juicy…
Liked by Ash Damle
-
How good is Llama-3.1, really? At Predibase we fine-tuned the 8B param variant across 23 different tasks, and it beat out GPT4, GPT-4o-mini, and…
How good is Llama-3.1, really? At Predibase we fine-tuned the 8B param variant across 23 different tasks, and it beat out GPT4, GPT-4o-mini, and…
Liked by Ash Damle
Other similar profiles
-
Mayank Kale
Connect -
Kevin S. Park
Connect -
Carla Gallegos
Connect -
Amrit Singh
Connect -
Rakesh Nemade
CTO and VP of Engineering at Healiom Inc.
Connect -
Susana C.
Connect -
Reza Rassool
Connect -
Susan Solinsky
Connect -
Stefano Ghirlanda
Professor at Stockholm University, Chief Technology Officer at DataWorks LLC
Connect -
JASH PALEJA
Actively seeking Data Analyst & Data Scientist Roles | MS Excel | Python | SQL | Tableau | Power BI | Graduate student at the University at Buffalo
Connect
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