Upload diagnostic_script.py

diagnostic_script.py

@@ -0,0 +1,144 @@
+#!/usr/bin/env python3
+"""Diagnostic script for HF Space deployment issues"""
+
+import os
+import sys
+from pathlib import Path
+
+def check_files():
+    """Check if all required files exist"""
+    print("=== File Check ===")
+    
+    required_files = [
+        "config/settings.py",
+        "src/retrieval/semantic_retriever.py", 
+        "src/generation/response_generator.py",
+        "data/vector_store/embeddings.npy",
+        "data/vector_store/faiss_index.bin",
+        "data/vector_store/chunks.pkl",
+        "data/vector_store/bm25.pkl",
+        "data/vector_store/relationships.pkl"
+    ]
+    
+    for file_path in required_files:
+        if os.path.exists(file_path):
+            print(f"✅ {file_path}")
+        else:
+            print(f"❌ {file_path} - MISSING")
+
+def check_imports():
+    """Check if imports work"""
+    print("\n=== Import Check ===")
+    
+    try:
+        from config.settings import config
+        print("✅ Config imported")
+        print(f"   - Embedding model: {config.EMBEDDING_MODEL}")
+        print(f"   - Vector store path: {config.VECTOR_STORE_PATH}")
+    except Exception as e:
+        print(f"❌ Config import failed: {e}")
+        return
+    
+    try:
+        from src.retrieval.semantic_retriever import SemanticRetriever
+        print("✅ SemanticRetriever imported")
+    except Exception as e:
+        print(f"❌ SemanticRetriever import failed: {e}")
+    
+    try:
+        from src.generation.response_generator import ResponseGenerator
+        print("✅ ResponseGenerator imported")
+    except Exception as e:
+        print(f"❌ ResponseGenerator import failed: {e}")
+    
+    try:
+        from sentence_transformers import SentenceTransformer
+        print("✅ SentenceTransformers imported")
+    except Exception as e:
+        print(f"❌ SentenceTransformers import failed: {e}")
+    
+    try:
+        import faiss
+        print("✅ FAISS imported")
+    except Exception as e:
+        print(f"❌ FAISS import failed: {e}")
+
+def check_vector_store():
+    """Check vector store files"""
+    print("\n=== Vector Store Check ===")
+    
+    try:
+        from config.settings import config
+        
+        # Check embeddings
+        import numpy as np
+        embeddings = np.load(os.path.join(config.VECTOR_STORE_PATH, "embeddings.npy"))
+        print(f"✅ Embeddings loaded: {embeddings.shape}")
+        
+        # Check FAISS
+        import faiss
+        faiss_index = faiss.read_index(os.path.join(config.VECTOR_STORE_PATH, "faiss_index.bin"))
+        print(f"✅ FAISS index loaded: {faiss_index.ntotal} vectors")
+        
+        # Check chunks
+        import pickle
+        with open(os.path.join(config.VECTOR_STORE_PATH, "chunks.pkl"), "rb") as f:
+            chunks = pickle.load(f)
+        print(f"✅ Chunks loaded: {len(chunks)} chunks")
+        
+        # Check BM25
+        with open(os.path.join(config.VECTOR_STORE_PATH, "bm25.pkl"), "rb") as f:
+            bm25 = pickle.load(f)
+        print(f"✅ BM25 loaded")
+        
+    except Exception as e:
+        print(f"❌ Vector store check failed: {e}")
+
+def check_model_loading():
+    """Check if embedding model loads"""
+    print("\n=== Model Loading Check ===")
+    
+    try:
+        from config.settings import config
+        from sentence_transformers import SentenceTransformer
+        
+        print("Loading embedding model...")
+        model = SentenceTransformer(config.EMBEDDING_MODEL, device='cpu')
+        print("✅ Embedding model loaded")
+        
+        # Test encoding
+        test_embedding = model.encode(["test"], show_progress_bar=False)
+        print(f"✅ Test encoding successful: shape {test_embedding.shape}")
+        
+    except Exception as e:
+        print(f"❌ Model loading failed: {e}")
+
+def check_api_keys():
+    """Check API keys"""
+    print("\n=== API Keys Check ===")
+    
+    openrouter_key = os.getenv("OPENROUTER_API_KEY")
+    tavily_key = os.getenv("TAVILY_API_KEY")
+    
+    if openrouter_key:
+        print("✅ OPENROUTER_API_KEY found")
+    else:
+        print("❌ OPENROUTER_API_KEY missing")
+    
+    if tavily_key:
+        print("✅ TAVILY_API_KEY found")
+    else:
+        print("⚠️ TAVILY_API_KEY missing (optional)")
+
+if __name__ == "__main__":
+    print("🔍 MANIT RAG Chatbot Diagnostic")
+    print("=" * 50)
+    
+    check_files()
+    check_imports()
+    check_vector_store()
+    check_model_loading()
+    check_api_keys()
+    
+    print("\n" + "=" * 50)
+    print("Diagnostic complete!")