Видео

Was wondering if we can use tool calling with a local LLM instead of and API

Not a fan of the way this was presented ... I was expecting a tutorial

Fantastic webinar, very useful. On a long and only very slightly related tangent, protip: Before (and during!) long presentations drink tea with honey and a lemon slice. When we're nervous our throat tenses up and it's very easy to start coughing when talking constanly. The hot tea with honey helps with that, and the acidity from the lemon prevents your mouth from drying out. I used to have a terrible time doing these 50-minute presentations + Q&A. My throat would be shot even though I wasn't yelling or anything. Eventually I took a couple of vocal training lessons and it practically solved the issue in a week.

Hi sir ! I’m also LLM/KG enthusiasm, based on your expertise Is Ontology schema like RDF/OWL necessary when building Graph data base as RAG material for LLM output augmentation ?

Might be a dumb question but: are the built in SQLTableRetrievalEngine basically a query pipeline already prepared and made? Is there some way to customize some of the components in the built in pipeline or do I need to make a custom query pipeline myself? Dealing with very large database with too many tables that are too large… trying to find a better way for table retrieval it improve the sql query made.

This call was great, we used the LlamaIndex+Neo4j implementation as a basis for automatic knowledge graph generation in R2R

where can we find the notebook used in the video? Please share!

great content but presentation skills....needs some work and energy!

How can I constrain the graph construction with .owl and RDF?

Also, does the panda pipeline also lack the robustness on the names like B.I.G ?

For the last part, if you already retrieved example rows from the vector DB, why go look again in the SQL database ? Or maybe the vector DB may not always return the correct row ?

Are you guys interested in developing a tool like LangSmith?

I don't like the speaker, he can barely explain things clearly. What is language module...I guess he was trying to say language model...And it is in-context learning, rather than in-content learning... Also, the code will be opaqued when he moved his mouse, I can hardly see anything...

Thank you so much! great!!!

super helpful, thx!

you nailed it man!!!

Great tutorial, thank you.

How does Instructions parsing works, does it uses the LLMs ?

awesome !!!!!!!! . ,

basically you're using gpt4 to check if gpt 3.5 answered the questions correctly. sounds clunky.

if you're doing a tutorial about llamaindex, please don't use llamaindex as an example corpus. It could get confusing.

I'm curious about the speed of reasoning for Language Agent Tree Search reasoning. Could you provide some rough numbers in terms of what we can expect? Maybe with Open AI or Groq?

lost my fire for LLMs and RAG in the last few months as even though i can be technical, it took to much time for me to learn, understand and implement while working full time. this video rekindled that fire because it's much easier than i remember it to build a RAG Pipeline. Keep up the great work !

Great video!

Sorry, but the repo for this lesson is hopelessly out of date. Spending more time than it's worth. chasing down new syntax for simple concepts like Prompt module imports. If you were just a third-party, OK nbd. But you're the vendor!

Extremely well put and explained. Thank you

loved the quick talk and the format!

You sir, are a gentleman and a scholar

Loved this video, exactly what I’m looking for in how to improve my RAG pipeline

Great work! Do you have any content like this with GraphRAG?

Loved the video, but the link to the resources is not working properly :(

Can it work just like Google NotebookLM or it's different?

This is the future of software engineering 🎉

Awesome, thanks for sharing !

so so so so so so so so...

Can you use this tool with a local model? I love LLM Studio because it can download a model and run it locally. It seems like none of these LLM flow tools can use local models. That seems really limiting. What if we don't want to send our private information to chatGPT or Claude? I think it would be great if it could work with a local model like LLMStudio.

Yeah guys please do upload some content for open source too not all can use the proprietary resources

Flowise can read and write local files? Thanks you

ollama mentioned. 👍

Suggestion: Place your avatar on the right side of the screen so that it doesn't block the text as much. Thank you.

I’m sorry if this is a trivial question, but what is the point in the first example of introspective agent or main agent worker regarding if it’s just a None type? I think you explained it something to do with it being the “initial response” but I couldn’t quite understand.

You became; that muslim then!

what an important topic and what a bad presenter

Smart ways and innovative ideas to re-invent RAG, good job

How is this different from function calling agent can you explain that you have used object index why is that you didnt mention that

Guys, I'd really like an increase in audio quality

If I dont want to use SimpleDocumentStore and I want to use a vectorstore like fiass or PGVECTOR how would I do it?

Really interesting, thanks for sharing. Can you share the slides?

Love it, to see all those windows and to have control over it. Hate to copy paste from chatgpt and keepngiving it context

You have shown your openai api key. Remove the key from your account