LLMS

→ View original post on X — @soumithchintala

Llama.cpp and Activation Checkpointing Optimization Techniques

By

@soumithchintala

–

05 July 2023 18h53

yea, llama.cpp is it for the llama-style models.
About activation checkpointing — the way the LLM people do it is to intentionally sidestep autodiff and hand-write a clever backward. Sometimes people even do mathematically approximate stuff.

→ View original post on X — @andrewyng

LangChain Free Course: Building Document Question Answering Systems

By

@andrewyng

–

05 July 2023 17h58

LangChain: Chat with Your Data, a new free short course created with @hwchase17, is now available! https://t.co/WxWiKL9qVl

In this 1 hour course, you’ll learn how to build one of the most requested LLM-based applications: Answering questions using information from a document or… pic.twitter.com/N9xSvOygnp
— Andrew Ng (@AndrewYNg) 5 juillet 2023

LangChain: Chat with Your Data, a new free short course created with @hwchase17
, is now available! https://
deeplearning.ai/short-courses/
langchain-chat-with-your-data/
… In this 1 hour course, you’ll learn how to build one of the most requested LLM-based applications: Answering questions using information from a document or

→ View original post on X — @dominodatalab

Kjell Carlsson on Specialized LLMs Power at VB Transform

By

@dominodatalab

–

05 July 2023 17h29

Kjell Carlsson speaks on the power of #sLLM's over "generic" LLMs at #vbtransform Jul 11. https://
venturebeat.com/ai/a-hot-gener
ative-ai-summer-is-here-get-ready-to-sweat-the-ai-beat/
…

→ View original post on X — @dominodatalab

Generative AI Trends: 90% of Data Science Leaders Adopt Wave

By

@dominodatalab

–

05 July 2023 17h28

Generative AI is the buzz this summer! Check out the latest #VentureBeat article to understand why 90% of data science leaders are riding this wave. Also don't miss Kjell Carlsson's talk on #SLLMs at #VBtransform! (agenda in comments) #GenerativeAI https://
domino.buzz/3XLtXGz

Memory Implementation for Follow-up Questions in AI Systems

By

–

05 July 2023 17h02

And finally, chat: we explain how to use memory to enable follow up questions, why that is necessary, and methods for doing so Big shout to to @RLanceMartin for helping prep all these materials!!!! Always fun to do these! What topic should we do next?

Deep Dive: Using LLMs to Chat with Your Data

By

–

05 July 2023 17h02

New @DeepLearningAI_ class I had so much fun teaching the last one with @AndrewYNg I had to return for a follow up This one is a deep dive on the most popular applications of LLMs to date: using them to chat with your data What do we cover?

Six RAG Modules: Deep Dive on Text Splitters and Retrieval

By

–

05 July 2023 17h02

There are six modules: – Document loaders
– Text splitters
– Embeddings and vector stores
– Retrieval
– QA generation
– Chat We go deep on each one. The three I think are most interesting/insightful: text splitters, retrieval, chat

Advanced Text Splitting and Semantic Retrieval in LangChain

By

–

05 July 2023 17h02

Text splitters: there's a lot of nuance in how you split text!! We cover a few examples of the advanced methods we have in LangChain Retrieval: semantic search can get you 80% of the way there easily, but getting that last bit can be hard. We cover methods to push further