A code implementation to create a PDF interplay system with AI on Google Colab utilizing Gemini Flash 1.5, Pymupdf and Google Technology AI API

On this tutorial, we show how one can construct a PDF interplay system with AI on Google Colab utilizing Gemini Flash 1.5, Pymupdf and the GOOGLE generative API. By profiting from these instruments, we will carry a PDF with out issues, extract your textual content and ask questions interactively, receiving clever solutions from the GEMINI Flash 1.5 mannequin from Google.

!pip set up -q -U google-generativeai PyMuPDF python-dotenv

First we set up the required dependencies to construct a PDF PDF system with Ia on Google Colab. Google-Generativai offers entry to Gemini Flash 1.5, permitting pure language interactions, whereas PyMUPDF (often known as FITZ) permits an environment friendly textual content extraction of PDF. As well as, Python-Dotenv helps handle setting variables, reminiscent of API keys, safely throughout the pocket book.

from google.colab import information
uploaded = information.add()

We add information out of your native gadget to Google Colab. When operating, it opens a file choice dialog, permitting you to decide on a file (for instance, a PDF) to load. The loaded file is saved in an object much like a dictionary (loaded), the place the keys characterize the names and values ​​of the information comprise the binary information of the file. This step is important to instantly course of paperwork, information units or pesos of fashions in a colab setting.

import fitz


def extract_pdf_text(pdf_path):
    doc = fitz.open(pdf_path)
    full_text = ""
    for web page in doc:
        full_text += web page.get_text()
    return full_text


pdf_file_path="/content material/Paper.pdf"
document_text = extract_pdf_text(pdf_path=pdf_file_path)
print("Doc textual content extracted!")
print(document_text(:1000)) 

We use PyMUPDF (FITZ) to extract textual content from a PDF file on Google Colab. The extract_pdf_text perform (PDF_Path) reads the PDF, itera via its pages and recovers the textual content content material. The extracted textual content is then saved in Docum_text, with the primary 1000 printed characters to acquire a preview of the content material. This step is essential to allow textual content -based evaluation and the reply of questions promoted by AI from the PDF.

import os
os.environ("GOOGLE_API_KEY") = 'Use your personal API key right here'

We set up the Google API key as an setting variable on Google Colab. The API secret’s required to authenticate the requests to Google Technology AI, which permits entry to Gemini Flash 1.5 for textual content processing with AI. Exchange ‘Use your personal API key right here’ with a legitimate key ensures that the mannequin can generate responses safely throughout the pocket book.

import google.generativeai as genai


genai.configure(api_key=os.environ("GOOGLE_API_KEY"))


model_name = "fashions/gemini-1.5-flash-001"


def query_gemini_flash(query, context):
    mannequin = genai.GenerativeModel(model_name=model_name)
    immediate = f"""
Context: {context(:20000)}


Query: {query}


Reply:
"""
    response = mannequin.generate_content(immediate)
    return response.textual content


pdf_text = extract_pdf_text("/content material/Paper.pdf")


query = "Summarize the important thing findings of this doc."
reply = query_gemini_flash(query, pdf_text)
print("Gemini Flash Reply:")
print(reply)

Lastly, we configure and seek the advice of Gemini Flash 1.5 utilizing a PDF doc for the era of textual content with AI. Initializes the Genai Library with the API key and masses the Gemini Flash 1.5 mannequin (Gemini-1.5-Flash-001). The Query_gemini_flash perform () takes a query and extracted the PDF textual content as an entry, formulates a structured message and recovers a solution generated by AI. This configuration permits the abstract of automated paperwork and the clever questions and solutions of PDFs.

In conclusion, after this tutorial, we have now efficiently created a system of interactive PDF interplay in Google Colab utilizing Gemini Flash 1.5, Pymupdf and the AI ​​generative API of Google. This resolution permits customers to extract PDF data and seek the advice of them interactively. The mixture of Google avant -garde AI fashions and the Colab cloud -based setting offers a robust and accessible solution to course of massive paperwork with out requiring heavy computational assets.

Right here is the Colab pocket book. Moreover, do not forget to observe us Twitter and be part of our Telegram channel and LINKEDIN GRsplash. Don’t forget to hitch our 80k+ ml topic.

Parlant: Construct Ai de ia brokers with LLM