MySQL database with PandasAI & Ollama & Streamlit

Today, we'll cover how to analyze data in a MySQL database with local Llama 3 using Pandas AI, Ollama and Streamlit for free. Happy learning.
▶ Subscribe: bit.ly/subscri...
▶ Join my channel: bit.ly/join-ti...
00:01 Introduction
01:06 Setup
02:47 Create a new database
04:43 Load the model with Ollama
06:41 Initialize the model
08:29 Connect to the database
10:45 Build the app
12:41 Inference
13:31 Data visualization
RELATED VIDEOS:
▶ PandasAI Tutorials: bit.ly/pandasai
▶ Ollama Tutorials: bit.ly/ollama-...
▶ LangChain Tutorials: bit.ly/langcha...
▶ Generative AI for DS: bit.ly/genai-f...
▶ HuggingFace Tutorials: bit.ly/hugging...
FOLLOW ME:
▶ Medium: / tirendazacademy
▶ X: / tirendazacademy
▶ LinkedIn: / tirendaz-academy
Don't forget to subscribe and turn on notifications so you don't miss the latest videos.
▶ Project files: github.com/Tir...
▶ MySQL Download: dev.mysql.com/...
▶ Dataset: www.kaggle.com...
Hi there, I'm a content creator on generative AI & data science, PhD. My goal is to make the latest technologies understandable for everyone.
#ai #generativeai #datascience

[@nguyenngothuong]

Can you make a video tutorial on role configuration for Botchat?. For example, only respond in English, and focus on using SQL to query data and respond to users.....

[@nguyenngothuong]

Added features to save chat history, suggest prompts after asking, etc.

[@RaghuNandan-ug1pk]

Unfortunately, I was not able to answer your question, because of the following error: No code found in the response is the error I am getting

[@erdemkoraysanl7547]

Hello, I wrote the same code, it works, but it is very slow, for example, when I say bring the first 4 data, it responds in 2 minutes 3 minutes, but it is very very fast in you. It is very fast when chatting normally with llama3. It is very fast in my database, but when I write this code, it works very slowly.

[@TirendazAI]

The app answered quickly a response in the video because I used cache. Cache allows PandasAI to store the results of previous queries. My queries took some time. Which graphic card do you use?

[@erdemkoraysanl7547]

@@TirendazAI I use the Amd Rodeon rx580 graphics card. Actually, my goal is to use AI search on my website. Algorithms almost most of the time can not give correct results. I really thought that artificial intelligence would be very useful to think like a human, but it works very slowly.

[@justinchung982]

Very slow for me as well. I’m using an MacBook Pro m3 pro 36gb

[@DarkNessShadow-]

Hi, awesome videos, can you make a video using agent pandasai with the explain and clarification questions features, i try to use it but doesn't work for me

[@TirendazAI]

Thanks, noted!

[@jackgaleras]

12:26 Gracias mil, la animación es muy adecuada cuando se intenta estas cosas y se consiguen mil errores hasta que por fin se logra el objetivo ja

[@Airsoftshowoffs]

Great video. Thank you for making this easy.

[@suffympm1601]

thanks for your video. from indonesia.. but why in my laptop result data is so long.. any idea ?. my spec is asus ryzen 7 gen 5

[@TirendazAI]

Hi, to get a quick response, it is important to have a powerful graphics card.

[@TheBestgoku]

bro please tell me if pandasAI is free if, i host on my own server. why is PandasAI so expensive?

[@TirendazAI]

PandasAI is free and open source.

[@TheUtube666]

This is good for accessing a single table. How do we scale it up to handle a more complex environment? I have sales history, titles and customer information and I'd like to allow users to ask questions against the various tables.

[@HimanshuSaxena16november]

Can we get data from multiple tables using prompts?

[@TirendazAI]

PandasAI supports queries with multiple dataframes. You can utilize a SmartDatalake to implement such queries.

[@HimanshuSaxena16november]

@@TirendazAI what about joins in tables,what will be the prompt in that case?

[@Airsoftshowoffs]

@@HimanshuSaxena16november Totally agree, if a DB with its ERD and relationships are loaded , it would be amazing.

[@DanielMouttapa]

@@TirendazAI when i used smartdatalake to test with the 2 datasets employee and salary table, with llama3 in local, for the question who gets paid the most, the code produced by the llm provoked error and no result, however with pandas ai api the result is ok (but limited quota), what local model can we use to get good result ?

[@zainmunir5628]

Hi. I love your videos. Is there anyway to do same with MongoDB or any NoSQL database?

[@TirendazAI]

Thanks for your feedback. PandasAI provides connectors for the following SQL databases: PostgreSQL, MySQL, Generic SQL, Snowflake, DataBricks, GoogleBigQuery, Yahoo Finance and Airtable.

[@atharvaghale4092]

is there any solution to make the queries run faster? will gpu help?

[@akosiraff7297]

Thanks, do you have a tutorial doing it on a VPS?

[@AbhinavKumar-tx5er]

Hi , I followed the same but I am getting error as
"Unfortunately, I was not able to answer your question, because of the following error: No code found in the response" .

[@RaghuNandan-ug1pk]

same dude

[@ByteBop911]

Is it possible to use agents without the pandasapi and use ollama?

[@TirendazAI]

You can use many agents with LangChain.