File size: 2,153 Bytes
a90a149
 
96a45e3
 
 
 
9668d23
a90a149
 
 
 
9668d23
 
 
96a45e3
 
 
 
 
 
 
 
 
 
 
 
 
 
a90a149
 
 
 
96a45e3
 
9668d23
 
96a45e3
 
a90a149
96a45e3
a90a149
96a45e3
 
a90a149
 
96a45e3
a90a149
96a45e3
a90a149
 
 
 
 
96a45e3
a90a149
 
 
 
 
 
 
 
 
 
 
 
 
4bb5912
a90a149
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
import openai
import gradio as gr
import os
from pdf2image import convert_from_path
import pytesseract
from PIL import Image
import glob

# Access the OpenAI API key from environment variables (Hugging Face secret)
openai.api_key = os.getenv('OPENAI_API_KEY')

# Directory where the PDF files are stored
pdf_directory = '/path_to_pdf_files'  # Change this to your actual dataset path

def pdf_to_text(pdf_path):
    """
    Converts PDF pages to images and extracts text using OCR.
    """
    images = convert_from_path(pdf_path)
    full_text = ""
    
    for image in images:
        # Perform OCR on each image
        text = pytesseract.image_to_string(image)
        full_text += text + "\n"
    
    return full_text

def extract_info(query):
    """
    This function interacts with OpenAI GPT-3.5 Turbo to extract information from the dataset based on the user's query.
    """
    all_texts = []

    # Loop through all PDF files in the directory
    for pdf_path in glob.glob(f'{pdf_directory}/*.pdf'):
        pdf_text = pdf_to_text(pdf_path)
        all_texts.append(pdf_text)
    
    combined_text = "\n".join(all_texts)

    # Send combined text and query to OpenAI for extraction
    prompt = f"Extract relevant information based on the following query: '{query}' from the Madras Music Academy Souvenir archives: {combined_text[:2000]}"
    
    response = openai.ChatCompletion.create(
        model="gpt-3.5-turbo",
        messages=[
            {"role": "system", "content": "You are an assistant that extracts information from PDF files using OCR."},
            {"role": "user", "content": prompt}
        ],
        max_tokens=300
    )
    
    # Return the answer from OpenAI GPT-3.5
    answer = response['choices'][0]['message']['content']
    return answer.strip()

# Define the Gradio interface
def gradio_interface(query):
    return extract_info(query)

# Launch the Gradio app
iface = gr.Interface(
    fn=gradio_interface, 
    inputs="text", 
    outputs="text",
    title="Sabha Scholar - Madras Music Academy AI Explorer",
    description="Ask questions about the Madras Music Academy Souvenirs."
)

iface.launch()