Easily Create Autonomous AI App from Scratch
Вставка
- Опубліковано 14 чер 2024
- Hey everyone! 🌟 In this video, we're diving deep into the world of AutoRAG using PhiData. You'll learn how to build an intelligent AutoRAG system from scratch, enhancing your AI's ability to decide whether to use memory, search a vector database, or utilise web search tools for the most relevant responses. 💡🤖 Easily Create Autonomous AI App from Scratch
What You'll Learn:
Introduction to AutoRAG: Understand the basics and how it works. 📘
Setting Up the Assistant: Step-by-step guide to configure the large language model and its components. 🛠️
Adding Documents to Knowledge Base: Learn how to upload files and URLs efficiently. 🗂️
Querying the Assistant: Techniques to ask questions and get precise answers from your AI. 💬
Creating a User Interface: Build a functional interface for better interaction. 🖥️
🔗 Resources:
Sponsor a Video: mer.vin/contact/
Do a Demo of Your Product: mer.vin/contact/
Patreon: / mervinpraison
Ko-fi: ko-fi.com/mervinpraison
Discord: / discord
Twitter / X : / mervinpraison
Code: mer.vin/2024/05/phidata-auto-...
git.new/auto-rag
git.new/groq-autorag
git.new/ollama-autorag
Key Highlights:
0:00 - Introduction to AutoRAG using PhiData
1:03 - Memory Storage with Postgres and PG Vector
2:01 - Utilising Tools like DuckDuckGo Search
4:00 - Adding Documents to the Knowledge Base
7:00 - Running Queries and Generating Responses
9:03 - Creating a User Interface with Streamlit
11:30 - Integrating Groq LLM and Llama Hermes 2 Pro
Stay tuned for more in-depth tutorials on AI and machine learning! Don't forget to like, share, and subscribe. 👍🔔
#autorag #phidata #groq #ollama #gpt4o - Навчання та стиль
❤❤❤ Can we use chain of thoughts (DSPy) so that model can identify multiple questions and answers all of them at once..!
Valuable Content !❤ Thank you Mervin.
Bit confused about the “auto” part of this rag. Isn’t it basically 3 different models/embeddings doing a q&a rag chat?
I think It's taking the first correct answer from Memory, Knowledge, Tools Re watch the first minute and a half, he says something like "thats the auto part"
great, how is your opinion on phidata vs autogen vs crewai? what are the pros and cons?
great video learn alot about rag and llm
can you please create video on clone voice and tts
Hi @MervinPraison. Great content. Could this be better done using the crewai framework you used recently?
Creative thinking
can't you just use ollama and not gtp4? i'm poor.
Need a video regarding customer assistent for a shop using assistant API with gpt-4o with document retrieval tool. Develop a web app
Hie Mervin, how about if we are to utilize Neo4j
Yeah, add Open source Graph DB's to the search list. ❗❗❗❗❗
@MervinPraison love your work so much 🧡🧡🧡
Thank you :)
Two huge problems:
- Chat: It is really difficult to 'chat', this approach is reliable as a Q&A bot, at least using OpenSource alternatives
- Tool calling failures: In case a tool calling fails because user is not providing all necesary info, the agent will keep trying perform a tool calling instead of asking for clarification
I love your content. Is there anyway you can do me a favor? I specifically have the steam deck and I wonder if you can create a video that shows you how to locally download your own custom LLM that you can create and have directly located on your steam deck specifically if your steam deck is modded to have 2 TB of space on its normal, solid state Drive and an additional terabyte SD card specifically using those things I would like to be able to have a custom AI that could work directly from my steam deck to my steam deck cellular function as well
Phidata > Praison AI?
Can you say 'going' again :)
i don't quite get it. so can future / other users use the same URL/app and ask the questions without them needing to upload URL/doc?
or they have to upload URL/doc
code?
Added now in the description.