Update app.py

app.py

@@ -3,14 +3,18 @@ import gradio as gr
 import numpy as np
 from langchain.text_splitter import RecursiveCharacterTextSplitter
 from langchain.document_loaders import DirectoryLoader, TextLoader
-from langchain.embeddings import HuggingFaceEmbeddings
-from langchain.vectorstores import FAISS
+from langchain_community.embeddings import HuggingFaceEmbeddings
+from langchain_community.vectorstores import FAISS
 from langchain.chains import ConversationalRetrievalChain
 from langchain.memory import ConversationBufferMemory
-from langchain.llms import HuggingFaceHub
+from langchain_community.llms import HuggingFaceHub
 
-# Set up environment variables for HuggingFace
-os.environ["HUGGINGFACEHUB_API_TOKEN"] = os.getenv("HUGGINGFACE_API_TOKEN")  # Replace with your token
+# Set up environment variables for HuggingFace - safely handle potential None value
+huggingface_token = os.getenv("HUGGINGFACE_API_TOKEN")
+if huggingface_token:
+    os.environ["HUGGINGFACEHUB_API_TOKEN"] = huggingface_token
+else:
+    print("Warning: HUGGINGFACE_API_TOKEN environment variable not set. You'll need to set it for the LLM to work.")
 
 # Create a directory for document storage if it doesn't exist
 os.makedirs("documents", exist_ok=True)
@@ -42,6 +46,10 @@ def process_documents():
 
 # Create RAG chain
 def create_chain(vector_store):
+    # Check if API token is available
+    if not os.getenv("HUGGINGFACEHUB_API_TOKEN"):
+        return None
+        
     # Initialize the LLM
     llm = HuggingFaceHub(
         repo_id="google/flan-t5-large", 
@@ -79,28 +87,45 @@ def upload_file(files):
     vector_store = process_documents()
     chain = create_chain(vector_store)
     
+    if chain is None:
+        return "Files uploaded and processed, but HuggingFace API token is missing. Set the environment variable to enable the chatbot."
+    
     return "Files uploaded and processed successfully!"
 
 # Function to handle user queries
 def chat(message, history):
     global chain, chat_history, vector_store
     
-    # Initialize vector store and chain if they don't exist
+    # Check if documents exist
     if vector_store is None:
         if os.path.exists("documents") and any(os.path.isfile(os.path.join("documents", f)) for f in os.listdir("documents")):
             vector_store = process_documents()
             chain = create_chain(vector_store)
         else:
-            return "Please upload documents first to initialize the chatbot."
-    
-    # Convert Gradio history to the format expected by the chain
-    if history:
-        chat_history = [(turn[0], turn[1]) for turn in history]
+            # Return in the format expected by Gradio chatbot
+            return history + [[message, "Please upload documents first to initialize the chatbot."]]
     
-    # Get response from chain
-    response = chain({"question": message})
+    # Check if API token is set
+    if chain is None:
+        # Return in the format expected by Gradio chatbot
+        return history + [[message, "HuggingFace API token is not set. Please set the HUGGINGFACE_API_TOKEN environment variable."]]
     
-    return response['answer']
+    # Process the message with the chain
+    try:
+        # Convert history to format expected by chain
+        if history:
+            chat_history = [(turn[0], turn[1]) for turn in history]
+        
+        # Get response from chain
+        response = chain({"question": message})
+        answer = response['answer']
+        
+        # Return in the format expected by Gradio chatbot
+        return history + [[message, answer]]
+    except Exception as e:
+        # Handle any errors
+        error_message = f"Error processing your request: {str(e)}"
+        return history + [[message, error_message]]
 
 # Create Gradio interface
 with gr.Blocks(title="RAG Chatbot") as demo: