Update app.py

app.py

@@ -58,15 +58,31 @@ retriever = AutoModelForSeq2SeqLM.from_pretrained(retriever_model_name)
 retriever_tokenizer = AutoTokenizer.from_pretrained(retriever_model_name)
 
 
-# Load or create FAISS index (same as before)
+# Load or create FAISS index
 index_path = "faiss_index.pkl"
 document_texts_path = "document_texts.pkl"
 document_texts = []
-# ... (rest of the FAISS index loading logic)
+if os.path.exists(index_path) and os.path.exists(document_texts_path):
+    try:
+        with open(index_path, "rb") as f:
+            index = pickle.load(f)
+            print("Loaded FAISS index from faiss_index.pkl")
+        with open(document_texts_path, "rb") as f:
+            document_texts = pickle.load(f)
+            print("Loaded document texts from document_texts.pkl")
+    except Exception as e:
+        print(f"Error loading FAISS index or document texts: {e}")
+else:
+    # Create a new FAISS index if it doesn't exist
+    index = faiss.IndexFlatL2(embedding_model.get_sentence_embedding_dimension())
+    with open(index_path, "wb") as f:
+        pickle.dump(index, f)
+        print("Created new FAISS index and saved to faiss_index.pkl")
 
 
 def preprocess_text(text):
-    # ... (text preprocessing logic, e.g., sentence segmentation and optional stop word removal)
+    sentences = sent_tokenize(text)
+    return sentences
 
 
 def upload_files(files):