Highly recommended! This is an amazing tutorial for someone who start to build an app on LangChain. It covers from backend and frontend. Much appreciated!

Great video. Advanced concepts but simple to understand.

Superb ! thanks !

Do these techniques come under "prompt engineering" ?

Very helpful ! Thanks !

most waited video of all times perhaps, i hope this video delivers what it promises in title

Between the previous video and this, RecursiveCharacterTextSplitter() is used in different ways...No explanation is given on why so...Disappointing series given it is from official LangChain...these are very shallow rush through the concepts though it is called "from scratch"..

thank you for your effort, but can you make a video for angular too?

when we handle large number of documents like 2000 pdf files. InmemoryStore might not work. Is there a solution for that?

does it support parallel tool calling ?

Is this different from Cohere's ReRank function?

We can also do tool calling with Ollama llama3 model kzbin.info/www/bejne/Y5ypc5iPmNKArLc

Thank you. It would be great to see more information about local tool calling for Agents and in LangGraph workflows.

god , you can afford a basic microphone ! you just had a 10 million dollar seed. You can afford a basic mic for each of your content creators.

Does it support multiples tools calling?

Thank you for the great and fast presentation

If we can estimate that type of llm usage, why we cant use this interface to train or fine tune llms to not just solving tasks but for making right controlling decisions in some type of workflow? Is there are any researches about this idea?

Please improve the microphone quality. Why wont you do this? We have been requesting for ages.

keep doing you are doing great 🎉🎉🎉🎉❤❤❤

Keep up the good work Brace. Much appreciated 🙏🏾

Can you go more into Detail about the Memory checkpoint? I have difficulties to understand how i can use the chat history e.g. In memory history

Nice Explaination, Examples are great

Hi please has anyone here ever worked on building a chat bot that respond to people like a particular person?, i.e chat bot that respond or generate replies like trump or barack

Hi Lance - great video again!Question for you. Recently I’ve been omitting LLM function calling just as a precautionary measure. I’m basically separating out the LLM from my functions (like api calls) and just asking the LLM to return a jsonOutput compliant with the parms in the api function. Am I doing myself a disservice by keeping these separate?

Great video! I've been loving this new framework

but it doesn't seem like firework is free.

Knowladgeable

For everything if you say "you can look into the notebook" why put up a video? The video is running as if being chased by someone.

Few days back tried a lot.. finally went with gemini model. Worked fantastically.

For anyone looking to save time: Not even cloning the repository makes this work. It only functions in a development environment. Any attempt to run the build fails with the error [EmptyChannelError]. Langchain seems more focused on releasing videos and new features than on ensuring functionality, and doesn't even have the courtesy to respond to those trying to resolve the issue.

6:10 I had an "ah-ha" 💡💡💡 moment from what you said about grabbing different info from a trace instead of having to go directly to the run, thank you so much for that. This helps me with a problem I am currently stuck on.

Excellent video! Thank you. Would you know how to handle the potential case that the agent goes into infinite loop, e.g. it gets stuck at the hallucinating check. I can only think of keeping track of the threshold for number of checks, and am wondering if there's a more elegant way to do that in Langchain.

Literally, I was dealing with llama3 and integrating tools within it...got many errors... And now I just got this video

First

Knowledgeable

thx

Thx

first time in life i had to watch it at 0.75x speed. Worth to watch.

Great tutorial <3

Would to a see full end to end python example with something like reflex

What's this witchcraft 😵

Thanks. How to take two inputs for a function (tool)?

Gold

Awesome Brace! Absolutely love this!

Good video. But how could I save the messages and access them again? Since the messages are displayed in {elements} without any map, how could I access them again?

🎯 Key points for quick navigation: 00:15 *📊 The retrieval process in RAG involves indexing documents, splitting them into smaller chunks, and storing their embeddings in an index.* 00:41 *🔍 Documents are embedded into a high-dimensional space where similar documents are located near each other.* 01:36 *💡 The location of a document in this space is determined by its semantic meaning or content.* 02:03 *🔎 Retrieval involves searching for nearby documents to a given question in this high-dimensional space.* 02:56 *📈 LangChain provides many different embedding models, indexes, document loaders, and splitters that can be combined to test different ways of doing indexing or retrieval.* Made with HARPA AI

🎯 Key points for quick navigation: 00:02 *📹 The second video in the RAG from Scratch series focuses on indexing, a crucial component of RAG pipelines.* 00:28 *🔍 The goal of indexing is to retrieve documents related to a given question using numerical representations of documents.* 00:53 *📊 Numerical representations of documents are used for easy comparison and search, with approaches including sparse vectors and machine learning-based embedding methods.* 01:08 *💡 Embedding methods compress documents into fixed-length vectors that capture their meaning, allowing for efficient search and retrieval.* 02:03 *📈 Documents are split into smaller chunks to accommodate embedding models' limited context windows, and each chunk is compressed into a vector representation.* Made with HARPA AI

🎯 Key points for quick navigation: 00:03 *📹 The "RAG from Scratch" series will cover basic principles and advanced topics for building LLM applications with LangChain.* 00:15 *🔒 LLMs haven't seen all data, including private or recent data, due to limited pre-training runs.* 00:44 *📊 LLMs have context windows that are increasing in size, representing dozens to hundreds of pages of information.* 01:10 *💻 Retrieval-Augmented Generation (RAG) is a popular paradigm for connecting LLMs to external data, involving three stages: indexing, retrieval, and generation.* 02:06 *📝 Future videos will explore methods and tricks for RAG's three basic components in detail.* Made with HARPA AI

Where are we heading to ?!?😮😮😮

Your audio is so quiet