app.py

from flask import Flask, request, jsonify, send_from_directory
from pdf_processor import extract_text_from_pdf, process_pdf
from model_loader import load_model, load_generator
import torch
import numpy as np

app = Flask(__name__)

# Load data and models
pdf_text = extract_text_from_pdf("data/example.pdf")
chunks, index = process_pdf(pdf_text)
tokenizer, model = load_model()
generator = load_generator()

@app.route('/')
def home():
    return send_from_directory('.', 'index.html')

@app.route('/ask', methods=['POST'])
def ask():
    prompt = request.json['question']
    
    inputs = tokenizer(prompt, return_tensors="pt", padding=True, truncation=True, max_length=512)
    with torch.no_grad():
        query_vector = model(**inputs).last_hidden_state.mean(dim=1).squeeze().numpy()

    k = 3
    distances, indices = index.search(query_vector.astype('float32').reshape(1, -1), k)
    context = " ".join([chunks[i] for i in indices[0]])

    input_text = f"Context: {context}\n\nQuestion: {prompt}\n\nAnswer:"
    response = generator(input_text, max_length=150, num_return_sequences=1)[0]['generated_text']

    return jsonify({'response': response})

if __name__ == '__main__':
    app.run(host='0.0.0.0', port=7860)