Відео

I don't know how people would understand this if you are just giving an overview of concepts and not actually writing the code step by step.

How to protect a company's information with this technology?

am i missing something with respect to Chromadb. it is failing as it is asking for aws credentials.

Well done Sir! :-)

Really fantastic make more projects with python please

Did something change with the deployment process. I am going through this and running into an issue where I get an error about an empty zip file but I did not see any steps about zipping the package.

I had some fun with the code added multi threading to the pdf processing and tested a few collections of data. But i found out that the embedings are only good for English and not other languages. Any tips for that? Tried searching for embedings voor Dutch but not to many results.

It's hard to find such high quality videos which is to the point with simplification in all the aspects. Great work !!!

Hey, it does not work for me because i use a different package manager dnf (fedora). Anyone can help me?

Love this channel, your content is clear explanation. Please do one video for fine tuning LLM for any specific task with one real-time use case

Thanks a lot. That's what I've been looking for. I have one question, though. When you cd to some directory, the path to the directory shows above the profile name, but it's in a very faint grey? Is there a way to change that color? Thanks again.

Can we get char location in this? or page number? like spacy gives location start character end character etc

Great work! Thank you! Maybe a naive question, what should we change to make it work with LM studio instead?

unit testing is such a silly way to test the answer, lol. We should ask another LLM to verify if it's a same thing =))))

that's pure gold, ty!

Hi!! Thanks for the video. Program works for document with chunks less than 90, but seems the Python code crashes whenever I try to store more than 90 chunks. Do you have any idea for this?? I can store around 80-90 chunks, if I try to add more, it crashes. Thanks!!!

Great video❤❤. Can you show me how to write prompts please. Cuz my chatbot is educational and i want it can use the knowledge i provide to answer the abstract question

Is there a way to change the chunking to be by paragraph so you can reference a page number and paragraph?

Great tutorial, Thanks!

Thanks, it would be nice to answer question from a table or a picture in a pdf.

Great tutorial! Im trying this out on a private server but the response time is very long. I currently have 23 CPU cores and 8Gb of ram. Is this enough? I also notice that not all cores are being used at the same time. Is there a way for all cores to be used?

How is this any different than file explorer in windows?

Can i use this code in my web design and how do i integrate it to my react (maintain) project where i am doing it for my mini project

I feel tupid because i stil don't understand the function calling and adding API and such 😅 I mean I'm not a programmer... that doesn't help :P

I can't install twine

This is my workflow dream, amazing stuff!!! I've listed out my entire workflow and slowly turning whatever I can into Python scripts. I suck at Python, but this is awesome motivation to learn. I'd love tutorials of your workflow. Whisper translation, auto edit code, ChatGPT, etc. I'm hoping to achieve similar results but with a local LLM (LLAMA 3) and DaVinci Resolves API. Have you ever tried Obsidian + Ollama? Curious! Thanks for the insanely good video, already watched several times.

I have a question I would like to email it to you, I don't want to look bad in the comment section😗

Why don't you use a WebSocket for communication with the ChatBot?

Hi, not sure what is going on, but chromaDB take way too long to create my database. {'source': 'data\\EntityDiagrams_TR_v2014SE.pdf', 'page': 0, 'start_index': 0} Adding to Chroma DB... Number of existing documents in DB: 0 👉 Adding new documents: 261 Processing batch 1 with 100 chunks... Batch 1 added in 205.49 seconds. Processing batch 2 with 100 chunks... Batch 2 added in 204.37 seconds. Processing batch 3 with 61 chunks... Batch 3 added in 124.76 seconds. Documents added to Chroma DB in 535.19 seconds. Data store generated in 537.88 seconds.

Thank you .

I refreshed RAG completely for a presentation. This was unbelievable good and concise

great video, direct and precise, I just want to ask you what happens if I want to give each user an ID, what would the logic be?

Hats off 🎩

I am trying to implement source CRUD in my application, but I am looking for an approach to embed docs where there is an update in the document, you he raised the question in the video butI would really like to listen to an approach to take here, creation and deletion is fine but what about the update? I don't want to embed the entire document Any responses will be very helpful, Thanks:)

This is exactly what I needed. I was struggling with extracting the message from the message object group. This short 10 min video just solved a problem I was working on for 4 hours.

can we use this to convert the data from pdfs to proper JSON format, if yes can u please tell the process and requirements for it. Would really appreciate it.

How to create a CLI Command which will accept arguments ? Like, pixegami-hello --file_name /file/path/

How can I add multi language support? So I can type in spanish and get response in the same language

Thank you for putting so much effort into this video, the quality is just magnificent. Your channel is super underrated, I hope you get the recognition that you deserve soon.

Thanks for the whole tutorial, I would suggest adding speed at the queries of the vector database with PostgreSQL with pgvector, pg_embeddings plugin.

Very well explained and done; detail information is provided, and the pace is perfect (considering I can pause and rewind). High quality. I've seen some of your other videos, and your talented at teaching!

Your videos are awesome. I have also deployed my fastapi rag app, but i used AppRunner. it simplifes a lot of the stuff. I can use backgroundtasks from fastapi.

How to change the color??

I would like it not only to be able to read PDFs but also if the final information was 'weak' or missing information, it would do an internet search on the topic provided in the 'query' and complete the final answer with the collected data. Do you think this is feasible to do? I thought about using agents for this, like crewai. I looked but couldn't find any tutorial that used both methods.

Hi there, What theme are you using?

Awesome.. can you explain if the python code asks for excel input from the user, what modifications should be done?

Wonderful presentation. Keep doing the same. Great job

Considering the following example: Somewhere in chunk #3 there's the text: "Mark buys a tomato" Somewhere in chunk #261 there's the text: "Mark buys a grapefruit" Prompt: "What fruits did Mark buy?" I expect the answer to be something like: "Mark has bought a tomato and a grapefruit" Which chunk will it return as source?

Is this approach work only in English ？ What about other language ?