Видео

Build AI solutions with us! Join The AI Guild for hands-on projects, tools, and a supportive community. bit.ly/the-ai-guild

Build AI solutions with us! Join The AI Guild for hands-on projects, tools, and a supportive community. bit.ly/the-ai-guild

Build AI solutions with us! Join The AI Guild for hands-on projects, tools, and a supportive community. bit.ly/the-ai-guild

Build AI solutions with us! Join The AI Guild for hands-on projects, tools, and a supportive community. bit.ly/the-ai-guild

Build AI solutions with us! Join The AI Guild for hands-on projects, tools, and a supportive community. bit.ly/the-ai-guild

🚀 Build AI solutions with us! Join The AI Guild for hands-on projects, tools, and a supportive community. bit.ly/the-ai-guild

very interesting. Loving your content on AI Engineering

Which vs code plugins do you use?

Great video! Thank you 😊 what do you think about storing embeddings in PostgreSQL? It’s getting more and more popular and I’m curious about what you think about it

Thank you sir, very clear explanation, can i use django to set up objective questions with the options

Why not llama.cpp or vllm? Ollama have his own quantised models, but mostly 4bit quant, for little models like 7b 4q its small

Well explained! Nice job.

This is old news, but I like your style of presenting. No hand waving etc. I hope your content becomes more technical and new in the future.

If you want to starve to death, go learn programming in college or university, speaking from experience.

The ollama can use system ram instead of gpu ram?

Loving the vids!

Amazing doucheFace thumbnail! Explain how that thumbnail remotely relates to the title of this video. Clearly intelligent people click stupid face thumbnails because everyone knows looking like an idiot is a clear sign that your content must be amazing! 😂

🔥💪🏽

subbbbbbed

Great video!! Thank you 🙏

Great content! Thanks for sharing. Greetings from Brazil.

Nice explanation 🎉🎉

It is not clear, 1. how many chunks are taken for a query in openai? 2. If there is no suitable chunk, what then? 3. Are there smarter ways to split documents into chunks? Tnx!

Hello! Can you clarify, if there is a large document/article in the database and we search using a short phrase of 3 words, what will be the search results then?

Lov eit!!!

Very Good explanation on your videos. Looking forward to the next

You are spot on, my friend. The use of a Graph Database (NEO4J) as a knowledge base, partnered with a LLM appears to be the best architecture that addresses two concerns that the alternatives (Vector Database) fall short on: 1) Support for larger documents [still need to chunk properly] and 2) A usable Knowledge Base is being built as you do your RAG operations. The KB can be repurposed for other uses. I see that NEO4J KB Builder allows you to pre-define the KB schema. This is my next challenge. If you could do a video deep dive on 'Domain Specific Schemas' [predefined] this would be a grand slam.

Hello Paulo, loved the course, loved the teaching also, thanks for sharing, I will be probably be rewatching it a couple of times again. Here some suggestions about a couple of things I would have loved to see you talk about: --> Metaprompts --> RAG-like Prompting using ChatGPT webUI (I watched your RAG Course at freecodecamp: DOPE!), so basically uploading specific documents to ChatGPT and querying the LLM using the Document as additional or exclusive data source. Thanks again for the course and keep up the good work.

Great job on this important topic! Thanks

Done

esential llm plese

good info

Great video, helped a lot!

It's not quite clear, in step 1 you ask the model to create additional questions on the topic of a financial report that the model hasn't seen yet? Then, in step 2 you pass your main request + requests created by the model, and what can you get, better answers? But what if the model in step 1 creates questions that are not interesting to you at all? It's not clear.

Fantastic!!!!

30

Thank you soooo muchhh for this, I have been working on a project for a client (it's a legal gpt sort of thing), after trying out so many prompting techniques, different llms, and more. Just recently I understood the issue was with the retrieval, because in order to answer some of the questions require laws from knowledge base that are not directly related to the question. I am very hopeful the techniques from this video will solve the issue. @VinciBits if you have any tips regarding this legal gpt that would be AWESOME!

Thanks so much for this awesome demo.

Thanks for creating this!

My dream is that you come up with something detailed on ML from notebook to Django web app to deployment. Kubernes docker etc fullah end to end

Interesting video sir. Thank you very much for the information. I am also a 3rd year student learning AI with web development out of curiosity. I have a question, by any chance are you looking for a creative video editor for your consistent content? As I have worked with some of creators, I would like to help you with editing thumbnail designing for youtube videos. Feel free to tell me your precious thoughts here. Once again thank you.

Most probably not. Every reasoning and gate adds latency.

thank you ❤ , very interesting , nice explanation, thanks again

I had bought your AutoGen course a few days back but I didn’t start it as yet. I did not know you had a RUclips channel at that juncture. After watching this video, I’m confident I made a good purchase. Thank you.

Very very clear explanation of ai agents. Thank you 😊

Can you help with building end to end application?

Legend! Thank you sir!

Love it!!!!!!