Query PDFs Using Mistral: Unlock INSANE Power! 🤯

Key Takeaways

hi everyone I'm really excited to show you how to make mistal large language model talk to your local PDF files let's get started first up get clone and then Katon ml llm mral invoice CPU once after you clone your repo go into the folder now we are going to create cond environment cond creates hyphen hyphen name mrl PDF python 3.11 going to create the environment and the environment is getting created now now activate the environment cond activate mral PDF once that is activated install the required package in PIP install hyphen R requirements.txt file that will install all the required packages once after the packages get installed download the model W get and then URL of the model which you can find from hugging face repo and then you are moving that to the models folder hyphen P models so that model will get saved in that folder it will take approximately a minute to download the model now the model got downloaded if you see the current folder structure you can see inside the data folder there is a invoice one.pdf and inside the models folder you got Mr 7B instruct model there I'm going to show you what is in that PDF open data invoice you can see the PDF here is just a basic voice document we are going to query this document you can change the settings and the different types of model you're using from the config file which is listed here config doyo so if you open that Nano config doyo there you can see what you want to change if you want to change the model name or the model file you can change it here if you're planning to change the vector store you can change it here and if you want to change the chunk chunk size or chunk overlap you can change it here so as a first step we are going to ingest python in inest dop this will automatically ingest the data from the invoice 1.pdf and create the require embeddings and save that data as embeddings next Python main.py and I'm going to ask what is the name of the seller the name of the seller is Chapman Kim green which you can double check here at the top Chapman Kim green now as an example I'm going to copy another file which is the llm PDF file and I'm going to move move that to the data folder and again I'm going to do python in. pi to ingest that data the PDF is about comprehensive overview of large language models now the data got ingested I'm going to ask question based on that paper say python main.py list the authors in the uh comprehensive overview of large language models paper now you go the answer Haman Aid has that so if you see the paper you can see it's their name listed there so that's exactly how you retrieve data from a PDF using SE mral 7 billion parameter model and also you can use any other the models and extract the data I hope you like the video please do like share and subscribe thanks for watching