Update app.py

app.py

@@ -10,13 +10,16 @@ import nltk
 nltk.download('punkt')
 nltk.download('punkt_tab')
 
-faiss_path="ubuntu_manual.txt"
+# Define paths as variables
+manual_path = "ubuntu_manual.txt"
+faiss_path = "manual_chunked_faiss_index_500.bin"
+
 # Load the Ubuntu manual from a .txt file
 try:
-    with open("ubuntu_manual.txt", "r", encoding="utf-8") as file:
+    with open(manual_path, "r", encoding="utf-8") as file:
         full_text = file.read()
 except FileNotFoundError:
-    raise FileNotFoundError("The file ubuntu_manual.txt was not found.")
+    raise FileNotFoundError(f"The file {manual_path} was not found.")
 
 # Function to chunk the text into smaller pieces
 def chunk_text(text, chunk_size=500):  # Larger chunks
@@ -41,10 +44,9 @@ manual_chunks = chunk_text(full_text, chunk_size=500)
 
 # Load your FAISS index
 try:
-    # Load your FAISS index
-    index = faiss.read_index("/absolute/path/to/manual_chunked_faiss_index_500.bin")
+    index = faiss.read_index(faiss_path)
 except Exception as e:
-    raise RuntimeError(f"Failed to load FAISS index: {e}")
+    raise RuntimeError(f"Failed to load FAISS index from {faiss_path}: {e}")
 
 # Load your embedding model
 embedding_model = SentenceTransformer('FridayMaster/fine_tune_embedding')
@@ -52,7 +54,6 @@ embedding_model = SentenceTransformer('FridayMaster/fine_tune_embedding')
 # OpenAI API key
 openai.api_key = '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'
 
-
 # Function to create embeddings
 def embed_text(text_list):
     return np.array(embedding_model.encode(text_list), dtype=np.float32)
@@ -64,6 +65,8 @@ def retrieve_chunks(query, k=5):
     # Search the FAISS index
     try:
         distances, indices = index.search(query_embedding, k=k)
+        print("Indices:", indices)
+        print("Distances:", distances)
     except Exception as e:
         raise RuntimeError(f"FAISS search failed: {e}")