Update app.py

app.py

@@ -561,511 +561,3 @@ if __name__ == "__main__":
     print(f"\n🚀 Starting server on port {port}...\n")
     app.run(host="0.0.0.0", port=port, debug=False)
 
-# import pickle
-# import faiss
-# from flask import Flask, request, jsonify, render_template_string
-# from sentence_transformers import SentenceTransformer
-# from transformers import AutoTokenizer, AutoModelForCausalLM
-# from huggingface_hub import hf_hub_download
-# import torch
-# import os
-# from functools import lru_cache
-# import hashlib
-
-# app = Flask(__name__)
-
-# print("=" * 50)
-# print("Loading models and data...")
-# print("=" * 50)
-
-# # ------------------------------
-# # Load embedding model (CPU)
-# # ------------------------------
-# embed_model = SentenceTransformer("all-MiniLM-L6-v2")
-# print("✓ Embedding model loaded")
-
-# # ------------------------------
-# # Download files from Hugging Face
-# # ------------------------------
-# REPO_ID = "Redfire-1234/pcb_tutor"
-
-# print("Downloading subject files from Hugging Face...")
-
-# # Download Biology files
-# bio_chunks_path = hf_hub_download(repo_id=REPO_ID, filename="bio_chunks.pkl", repo_type="model")
-# faiss_bio_path = hf_hub_download(repo_id=REPO_ID, filename="faiss_bio.bin", repo_type="model")
-
-# # Download Chemistry files
-# chem_chunks_path = hf_hub_download(repo_id=REPO_ID, filename="chem_chunks.pkl", repo_type="model")
-# faiss_chem_path = hf_hub_download(repo_id=REPO_ID, filename="faiss_chem.bin", repo_type="model")
-
-# # Download Physics files
-# phy_chunks_path = hf_hub_download(repo_id=REPO_ID, filename="phy_chunks.pkl", repo_type="model")
-# faiss_phy_path = hf_hub_download(repo_id=REPO_ID, filename="faiss_phy.bin", repo_type="model")
-
-# # Load all subjects into memory
-# SUBJECTS = {
-#     "biology": {
-#         "chunks": pickle.load(open(bio_chunks_path, "rb")),
-#         "index": faiss.read_index(faiss_bio_path)
-#     },
-#     "chemistry": {
-#         "chunks": pickle.load(open(chem_chunks_path, "rb")),
-#         "index": faiss.read_index(faiss_chem_path)
-#     },
-#     "physics": {
-#         "chunks": pickle.load(open(phy_chunks_path, "rb")),
-#         "index": faiss.read_index(faiss_phy_path)
-#     }
-# }
-
-# print(f"✓ Biology: {len(SUBJECTS['biology']['chunks'])} chunks loaded")
-# print(f"✓ Chemistry: {len(SUBJECTS['chemistry']['chunks'])} chunks loaded")
-# print(f"✓ Physics: {len(SUBJECTS['physics']['chunks'])} chunks loaded")
-
-# # ------------------------------
-# # Load LLM model (CPU) with optimizations
-# # ------------------------------
-# model_name = "Qwen/Qwen2.5-3B-Instruct"
-# print(f"Loading LLM: {model_name}")
-# tokenizer = AutoTokenizer.from_pretrained(model_name)
-# device = "cpu"
-
-# # OPTIMIZATION: Load model with better dtype for CPU
-# model = AutoModelForCausalLM.from_pretrained(
-#     model_name,
-#     torch_dtype=torch.float32,
-#     low_cpu_mem_usage=True  # Optimization: Better memory management
-# ).to(device)
-
-# # OPTIMIZATION: Set model to eval mode and optimize for inference
-# model.eval()
-# if hasattr(torch, 'set_num_threads'):
-#     torch.set_num_threads(4)  # Optimization: Use multiple CPU threads
-
-# print(f"✓ LLM loaded on {device}")
-
-# print("=" * 50)
-# print("All models loaded successfully!")
-# print("=" * 50)
-
-# # ------------------------------
-# # OPTIMIZATION: Add caching for MCQ generation
-# # ------------------------------
-# MCQ_CACHE = {}
-# MAX_CACHE_SIZE = 100
-
-# def get_cache_key(topic, subject, context_hash):
-#     """Generate a unique cache key"""
-#     return f"{subject}:{topic}:{context_hash}"
-
-# def cache_mcq(key, mcqs):
-#     """Cache generated MCQs with size limit"""
-#     if len(MCQ_CACHE) >= MAX_CACHE_SIZE:
-#         # Remove oldest entry
-#         MCQ_CACHE.pop(next(iter(MCQ_CACHE)))
-#     MCQ_CACHE[key] = mcqs
-
-# # ------------------------------
-# # RAG Search in specific subject (optimized)
-# # ------------------------------
-# def rag_search(query, subject, k=5):
-#     if subject not in SUBJECTS:
-#         return None
-    
-#     chunks = SUBJECTS[subject]["chunks"]
-#     index = SUBJECTS[subject]["index"]
-    
-#     # OPTIMIZATION: Encode query (already fast with sentence-transformers)
-#     q_emb = embed_model.encode([query], show_progress_bar=False).astype("float32")
-#     D, I = index.search(q_emb, k)
-    
-#     # Get the actual chunks
-#     results = []
-#     for idx in I[0]:
-#         if idx < len(chunks):
-#             results.append(chunks[idx])
-    
-#     return "\n\n".join(results)
-
-# # ------------------------------
-# # OPTIMIZED MCQ Generation with reduced tokens
-# # ------------------------------
-# def generate_mcqs(context, topic, subject):
-#     # OPTIMIZATION: Check cache first
-#     context_hash = hashlib.md5(context.encode()).hexdigest()[:8]
-#     cache_key = get_cache_key(topic, subject, context_hash)
-    
-#     if cache_key in MCQ_CACHE:
-#         print("✓ Using cached MCQs")
-#         return MCQ_CACHE[cache_key]
-    
-#     # OPTIMIZATION: Shortened prompt for faster generation
-#     prompt = f"""You are a Class-12 {subject.title()} teacher creating MCQs.
-# Topic: "{topic}"
-# Context:
-# {context}
-
-# Generate exactly 5 MCQs in this format:
-# Q1. [Question]
-# A) [Option]
-# B) [Option]
-# C) [Option]
-# D) [Option]
-# Correct Answer: [Letter] - [Reason]
-
-# Rules: Make correct answer from context, realistic distractors.
-# Generate 5 MCQs:"""
-    
-#     # OPTIMIZATION: Reduced max_length for faster tokenization
-#     inputs = tokenizer(prompt, return_tensors="pt", truncation=True, max_length=1536).to(device)
-    
-#     # OPTIMIZATION: Use torch.no_grad() for inference (saves memory)
-#     with torch.no_grad():
-#         # OPTIMIZATION: Reduced max_new_tokens from 900 to 600 (sufficient for 5 MCQs)
-#         # OPTIMIZATION: Reduced temperature from 0.15 to 0.1 (faster, more deterministic)
-#         # OPTIMIZATION: Added num_beams=1 (greedy decoding, faster than sampling)
-#         outputs = model.generate(
-#             **inputs,
-#             max_new_tokens=600,  # Reduced from 900
-#             temperature=0.1,     # Reduced from 0.15
-#             top_p=0.85,          # Slightly adjusted
-#             do_sample=True,
-#             repetition_penalty=1.15,
-#             pad_token_id=tokenizer.eos_token_id  # Optimization: Explicit pad token
-#         )
-    
-#     result = tokenizer.decode(outputs[0], skip_special_tokens=True)
-    
-#     # Extract only the generated MCQs
-#     if "Generate 5 MCQs:" in result:
-#         result = result.split("Generate 5 MCQs:")[-1].strip()
-    
-#     # OPTIMIZATION: Cache the result
-#     cache_mcq(cache_key, result)
-    
-#     return result
-
-# def verify_and_correct_answers(mcqs_text, context):
-#     """
-#     This function is kept for future enhancements
-#     """
-#     return mcqs_text
-
-# # ------------------------------
-# # HTML UI (with improved loading message)
-# # ------------------------------
-# HTML_TEMPLATE = """
-# <!DOCTYPE html>
-# <html>
-# <head>
-#     <title>Class 12 PCB MCQ Generator</title>
-#     <style>
-#         * { margin: 0; padding: 0; box-sizing: border-box; }
-#         body {
-#             font-family: 'Segoe UI', Tahoma, Geneva, Verdana, sans-serif;
-#             background: linear-gradient(135deg, #667eea 0%, #764ba2 100%);
-#             min-height: 100vh;
-#             padding: 20px;
-#         }
-#         .container {
-#             max-width: 900px;
-#             margin: 0 auto;
-#             background: white;
-#             border-radius: 20px;
-#             box-shadow: 0 20px 60px rgba(0,0,0,0.3);
-#             overflow: hidden;
-#         }
-#         .header {
-#             background: linear-gradient(135deg, #667eea 0%, #764ba2 100%);
-#             color: white;
-#             padding: 30px;
-#             text-align: center;
-#         }
-#         .header h1 { font-size: 2.5em; margin-bottom: 10px; }
-#         .content { padding: 40px; }
-#         .form-group {
-#             margin-bottom: 25px;
-#         }
-#         label {
-#             display: block;
-#             font-weight: 600;
-#             margin-bottom: 10px;
-#             color: #333;
-#             font-size: 16px;
-#         }
-#         select, input {
-#             width: 100%;
-#             padding: 15px;
-#             border: 2px solid #e0e0e0;
-#             border-radius: 10px;
-#             font-size: 16px;
-#             font-family: inherit;
-#             transition: border-color 0.3s;
-#         }
-#         select:focus, input:focus {
-#             outline: none;
-#             border-color: #667eea;
-#         }
-#         button {
-#             width: 100%;
-#             padding: 18px;
-#             background: linear-gradient(135deg, #667eea 0%, #764ba2 100%);
-#             color: white;
-#             border: none;
-#             border-radius: 10px;
-#             font-size: 18px;
-#             font-weight: 600;
-#             cursor: pointer;
-#             transition: all 0.3s;
-#         }
-#         button:hover {
-#             transform: translateY(-2px);
-#             box-shadow: 0 10px 20px rgba(102, 126, 234, 0.4);
-#         }
-#         button:disabled {
-#             background: #ccc;
-#             cursor: not-allowed;
-#             transform: none;
-#         }
-#         .result {
-#             margin-top: 30px;
-#             padding: 25px;
-#             background: #f8f9fa;
-#             border-radius: 10px;
-#             border-left: 4px solid #667eea;
-#             display: none;
-#         }
-#         .result.show { display: block; }
-#         .result h3 {
-#             color: #667eea;
-#             margin-bottom: 20px;
-#             font-size: 1.4em;
-#         }
-#         .mcq-content {
-#             background: white;
-#             padding: 25px;
-#             border-radius: 8px;
-#             white-space: pre-wrap;
-#             line-height: 1.9;
-#             font-size: 15px;
-#         }
-#         .loading {
-#             text-align: center;
-#             padding: 30px;
-#             display: none;
-#         }
-#         .loading.show { display: block; }
-#         .spinner {
-#             border: 4px solid #f3f3f3;
-#             border-top: 4px solid #667eea;
-#             border-radius: 50%;
-#             width: 50px;
-#             height: 50px;
-#             animation: spin 1s linear infinite;
-#             margin: 0 auto 15px;
-#         }
-#         @keyframes spin {
-#             0% { transform: rotate(0deg); }
-#             100% { transform: rotate(360deg); }
-#         }
-#         .subject-tag {
-#             display: inline-block;
-#             padding: 5px 15px;
-#             border-radius: 20px;
-#             font-size: 13px;
-#             font-weight: 600;
-#             margin-right: 10px;
-#         }
-#         .bio { background: #d4edda; color: #155724; }
-#         .chem { background: #d1ecf1; color: #0c5460; }
-#         .phy { background: #f8d7da; color: #721c24; }
-#         .optimization-badge {
-#             background: #28a745;
-#             color: white;
-#             padding: 5px 12px;
-#             border-radius: 15px;
-#             font-size: 12px;
-#             margin-left: 10px;
-#         }
-#     </style>
-# </head>
-# <body>
-#     <div class="container">
-#         <div class="header">
-#             <h1>🎓 Class 12 PCB MCQ Generator</h1>
-#             <p style="font-size: 1.1em; margin-bottom: 15px;">Generate practice MCQs from your textbooks <span class="optimization-badge">⚡ Optimized</span></p>
-#             <div>
-#                 <span class="subject-tag bio">Biology</span>
-#                 <span class="subject-tag chem">Chemistry</span>
-#                 <span class="subject-tag phy">Physics</span>
-#             </div>
-#         </div>
-        
-#         <div class="content">
-#             <div class="form-group">
-#                 <label for="subject">📚 Select Subject</label>
-#                 <select id="subject">
-#                     <option value="biology">Biology</option>
-#                     <option value="chemistry">Chemistry</option>
-#                     <option value="physics">Physics</option>
-#                 </select>
-#             </div>
-            
-#             <div class="form-group">
-#                 <label for="topic">✏️ Enter Topic</label>
-#                 <input type="text" id="topic" placeholder="e.g., Mitochondria, Chemical Bonding, Newton's Laws">
-#             </div>
-            
-#             <button onclick="generateMCQs()">🚀 Generate 5 MCQs</button>
-            
-#             <div class="loading" id="loading">
-#                 <div class="spinner"></div>
-#                 <p style="color: #666; font-size: 16px;">Generating MCQs... This may take 20-40 seconds</p>
-#                 <p style="color: #999; font-size: 13px; margin-top: 10px;">⚡ Optimized for faster generation</p>
-#             </div>
-            
-#             <div class="result" id="result">
-#                 <h3>📝 Generated MCQs:</h3>
-#                 <div style="background: #fff3cd; padding: 12px; border-radius: 6px; margin-bottom: 15px; color: #856404; font-size: 14px;">
-#                     ⚠️ <strong>Note:</strong> AI-generated answers may occasionally be incorrect. Please verify answers using your textbook.
-#                 </div>
-#                 <div class="mcq-content" id="mcqContent"></div>
-#             </div>
-#         </div>
-#     </div>
-#     <script>
-#         async function generateMCQs() {
-#             const subject = document.getElementById('subject').value;
-#             const topic = document.getElementById('topic').value.trim();
-            
-#             if (!topic) {
-#                 alert('⚠️ Please enter a topic!');
-#                 return;
-#             }
-            
-#             const loading = document.getElementById('loading');
-#             const result = document.getElementById('result');
-#             const btn = document.querySelector('button');
-            
-#             loading.classList.add('show');
-#             result.classList.remove('show');
-#             btn.disabled = true;
-#             btn.textContent = '⏳ Generating...';
-            
-#             try {
-#                 const response = await fetch('/generate', {
-#                     method: 'POST',
-#                     headers: {'Content-Type': 'application/json'},
-#                     body: JSON.stringify({subject, topic})
-#                 });
-                
-#                 const data = await response.json();
-                
-#                 if (data.error) {
-#                     alert('❌ Error: ' + data.error);
-#                     return;
-#                 }
-                
-#                 document.getElementById('mcqContent').textContent = data.mcqs;
-#                 result.classList.add('show');
-#             } catch (error) {
-#                 alert('❌ Error: ' + error.message);
-#             } finally {
-#                 loading.classList.remove('show');
-#                 btn.disabled = false;
-#                 btn.textContent = '🚀 Generate 5 MCQs';
-#             }
-#         }
-        
-#         // Allow Enter key to submit
-#         document.getElementById('topic').addEventListener('keypress', function(e) {
-#             if (e.key === 'Enter') {
-#                 generateMCQs();
-#             }
-#         });
-#     </script>
-# </body>
-# </html>
-# """
-
-# # ------------------------------
-# # Routes
-# # ------------------------------
-# @app.route("/")
-# def home():
-#     return render_template_string(HTML_TEMPLATE)
-
-# @app.route("/generate", methods=["POST"])
-# def generate():
-#     try:
-#         data = request.json
-#         subject = data.get("subject", "").lower()
-#         topic = data.get("topic", "")
-        
-#         if not topic:
-#             return jsonify({"error": "Topic is required"}), 400
-        
-#         if subject not in SUBJECTS:
-#             return jsonify({"error": "Invalid subject. Choose biology, chemistry, or physics."}), 400
-        
-#         print(f"\n🔍 Searching {subject} for topic: {topic}")
-        
-#         # Retrieve context from RAG
-#         context = rag_search(topic, subject, k=5)
-        
-#         if not context or len(context.strip()) < 50:
-#             return jsonify({"error": f"No relevant content found in {subject} for topic: {topic}"}), 404
-        
-#         print(f"✓ Found context ({len(context)} chars)")
-        
-#         # Generate MCQs (now with caching)
-#         print("🤖 Generating MCQs...")
-#         mcqs = generate_mcqs(context, topic, subject)
-        
-#         print("✓ MCQs generated successfully")
-        
-#         return jsonify({"mcqs": mcqs, "subject": subject})
-    
-#     except Exception as e:
-#         print(f"❌ Error in /generate: {e}")
-#         import traceback
-#         traceback.print_exc()
-#         return jsonify({"error": str(e)}), 500
-
-# @app.route("/health")
-# def health():
-#     return jsonify({
-#         "status": "healthy",
-#         "subjects": {
-#             "biology": len(SUBJECTS["biology"]["chunks"]),
-#             "chemistry": len(SUBJECTS["chemistry"]["chunks"]),
-#             "physics": len(SUBJECTS["physics"]["chunks"])
-#         },
-#         "cache_size": len(MCQ_CACHE)
-#     })
-
-# # OPTIMIZATION: Add cache stats endpoint
-# @app.route("/cache/stats")
-# def cache_stats():
-#     return jsonify({
-#         "cached_topics": len(MCQ_CACHE),
-#         "max_cache_size": MAX_CACHE_SIZE,
-#         "cache_keys": list(MCQ_CACHE.keys())
-#     })
-
-# # OPTIMIZATION: Add cache clear endpoint (optional)
-# @app.route("/cache/clear", methods=["POST"])
-# def clear_cache():
-#     MCQ_CACHE.clear()
-#     return jsonify({"status": "Cache cleared"})
-
-# # ------------------------------
-# # Run the App
-# # ------------------------------
-# if __name__ == "__main__":
-#     port = int(os.environ.get("PORT", 7860))
-#     print(f"\n🚀 Starting Flask on 0.0.0.0:{port}\n")
-#     app.run(host="0.0.0.0", port=port, debug=False)
-
-