Update src/config.py

src/config.py

@@ -1,42 +1,32 @@
-"""
-Configuration file for Multimodal RAG LLM System
-"""
+
 import os
 from pathlib import Path
 
-# API Configuration
 OPENAI_API_KEY = os.getenv("OPENAI_API_KEY", "")
-OPENAI_MODEL = "gpt-4o-mini"  # Cheaper model variant
-USE_CACHE = True  # Enable response caching
+OPENAI_MODEL = "gpt-4o-mini" 
+USE_CACHE = True 
 
-# Vector Store Configuration
 CHROMA_DB_PATH = "./chroma_db"
 DOCSTORE_PATH = "./docstore"
 PROCESSED_FILES_LOG = "./processed_files.txt"
 
-# Embedding Model Configuration
 EMBEDDING_MODEL = "sentence-transformers/all-mpnet-base-v2"
 EMBEDDING_DIM = 768
 
-# System Configuration
-MAX_CHUNK_SIZE = 500  # Smaller chunks = fewer tokens
-CHUNK_OVERLAP = 50    # Less overlap = fewer chunks
-TEMPERATURE = 0.3     # Lower = faster, cheaper
-MAX_TOKENS = 500      # Limit response size (vs 1500)
+MAX_CHUNK_SIZE = 500 
+CHUNK_OVERLAP = 50
+TEMPERATURE = 0.3
+MAX_TOKENS = 500 
 
-# Language Support
 LANGUAGE = "russian"
 
-# Create necessary directories
 Path(CHROMA_DB_PATH).mkdir(exist_ok=True)
 Path(DOCSTORE_PATH).mkdir(exist_ok=True)
 
-# PDF Upload Configuration
 UPLOAD_FOLDER = "./uploaded_pdfs"
 Path(UPLOAD_FOLDER).mkdir(exist_ok=True)
 MAX_PDF_SIZE_MB = 50
 
-# TOKEN OPTIMIZATION SETTINGS
-BATCH_SEARCH_RESULTS = 3  # Return only top 3 (not 5)
-CACHE_RESPONSES = True    # Cache Q&A responses
-SUMMARIZE_FIRST = True    # Summarize PDFs once, not per query
+BATCH_SEARCH_RESULTS = 3  
+CACHE_RESPONSES = True    
+SUMMARIZE_FIRST = True