File size: 2,838 Bytes
556c1d7 aec3e44 556c1d7 aec3e44 556c1d7 aec3e44 556c1d7 aec3e44 556c1d7 aec3e44 556c1d7 aec3e44 556c1d7 aec3e44 556c1d7 aec3e44 556c1d7 aec3e44 556c1d7 aec3e44 | 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 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 | import os
# Auto build FAISS index if not present
if not os.path.exists("vector.index"):
print("vector.index not found. Building embeddings...")
os.system("python embeddings.py")
from flask import Flask, request, jsonify, render_template
from sentence_transformers import SentenceTransformer
import faiss
import numpy as np
from transformers import pipeline
from dotenv import load_dotenv
load_dotenv()
app = Flask(__name__)
# -------------------- HOME --------------------
@app.route("/")
def home():
return render_template("index.html")
# -------------------- LOAD MODELS --------------------
print("Loading embedding model...")
embedder = SentenceTransformer("all-MiniLM-L6-v2")
print("Loading vector database...")
index = faiss.read_index("vector.index")
with open("documents.txt", "r", encoding="utf-8") as f:
documents = [line.strip() for line in f.readlines()]
print("Loading language model...")
qa_model = pipeline(
"text2text-generation",
model="google/flan-t5-base",
max_new_tokens=120
)
# -------------------- CORE BOT --------------------
def ask_bot(question, top_k=3, similarity_threshold=1.0):
greetings = [
"hi", "hello", "hey", "hii", "hola",
"good morning", "good evening", "good afternoon"
]
if question.lower().strip() in greetings:
return "Hi 👋 I’m your C++ assistant. What’s on your mind today?"
# RAG flow
q_embedding = embedder.encode([question])
distances, indices = index.search(np.array(q_embedding), top_k)
if distances[0][0] > similarity_threshold:
return "I don't know based on the given data."
context = "\n".join([documents[i] for i in indices[0]])
prompt = f"""
Answer the question ONLY using the context below.
If the answer is not present, reply exactly:
I don't know based on the given data.
Context:
{context}
Question:
{question}
Answer:
"""
output = qa_model(prompt)[0]["generated_text"].strip()
return output
# -------------------- API --------------------
@app.route("/ask", methods=["POST"])
def ask():
data = request.json
question = data.get("question", "")
if not question:
return jsonify({"answer": "Please provide a question."})
answer = ask_bot(question)
return jsonify({"answer": answer})
# -------------------- FEEDBACK (SAFE VERSION – NO EMAIL) --------------------
@app.route("/feedback", methods=["POST"])
def feedback():
data = request.json
message = data.get("feedback", "")
user_email = data.get("email", "Not provided")
print("📩 New Feedback Received")
print("From:", user_email)
print("Message:", message)
# For now just log feedback instead of emailing (SMTP crashes servers)
return jsonify({"status": "success"})
# ❌ DO NOT USE app.run() — gunicorn handles the server
|