Update ingest.py

ingest.py

@@ -1,9 +1,9 @@
 import os
 import logging
 import faiss
-from langchain_community.document_loaders import PDFMinerLoader
+from langchain.document_loaders import PDFMinerLoader
 from langchain.text_splitter import RecursiveCharacterTextSplitter
-from langchain_community.embeddings import HuggingFaceEmbeddings
+from langchain.embeddings import HuggingFaceEmbeddings
 from langchain.vectorstores import FAISS
 
 # Configure logging
@@ -11,15 +11,16 @@ logging.basicConfig(level=logging.INFO)
 logger = logging.getLogger(__name__)
 
 def create_faiss_index():
+    """
+    Create a FAISS index from documents in the 'docs' directory.
+    """
     documents = []
-    docs_dir = "docs"  # Directory where PDF files are stored
+    docs_dir = "docs"
 
-    # Check if the 'docs' directory exists
     if not os.path.exists(docs_dir):
         logger.error(f"The directory '{docs_dir}' does not exist.")
         return
 
-    # Walk through the 'docs' directory and load PDF files
     for root, dirs, files in os.walk(docs_dir):
         for file in files:
             if file.endswith(".pdf"):
@@ -36,25 +37,20 @@ def create_faiss_index():
                 except Exception as e:
                     logger.error(f"Error loading {file_path}: {e}")
 
-    # Check if any documents were loaded
     if not documents:
         logger.error("No documents were loaded. Check the 'docs' directory and file paths.")
         return
 
     logger.info(f"Loaded {len(documents)} documents.")
 
-    # Split documents into text chunks
     text_splitter = RecursiveCharacterTextSplitter(chunk_size=500, chunk_overlap=50)
     texts = text_splitter.split_documents(documents)
     logger.info(f"Created {len(texts)} text chunks.")
-
-    # Check if text chunks were created
     if not texts:
         logger.error("No text chunks created. Check the text splitting process.")
         return
 
     try:
-        # Initialize embeddings using HuggingFace models
         embeddings = HuggingFaceEmbeddings(model_name="sentence-transformers/all-MiniLM-L6-v2")
         logger.info("Embeddings initialized successfully.")
     except Exception as e:
@@ -62,10 +58,7 @@ def create_faiss_index():
         return
 
     try:
-        # Create a FAISS index and save it
-        index = faiss.IndexFlatL2(embeddings.embedding_size)
-        vector_store = FAISS.from_documents(texts, embeddings, index)
-        vector_store.save_local("faiss_index")
+        faiss_index = FAISS.from_documents(texts, embeddings, index_path="faiss_index/index.faiss")
         logger.info(f"Created FAISS index with {len(texts)} vectors.")
     except Exception as e:
         logger.error(f"Failed to create FAISS index: {e}")