token added

app.py

@@ -68,7 +68,18 @@ def process_pdf(file_path):
     gr.Info("PDF processed and ready for questions!")
 
 # Your existing functions
-def using_ollama_model(retriever, query, results, conversation_history):
+def using_ollama_model(retriever, query, results, conversation_history, token):
+    try:
+        if token:
+            gr.Info("Attempting to log in to Hugging Face...")
+            login(token=token)
+            gr.Info("Login successful!")
+        else:
+            gr.Warning("No Hugging Face token provided. Gated models may not be accessible.")
+    except Exception as e:
+        gr.Error(f"Hugging Face login failed: {e}")
+        return "An error occurred during authentication. Please check your token and try again."
+
     history_text = ""
     for item in conversation_history:
         if "question" in item and item["question"]:
@@ -111,7 +122,7 @@ def using_ollama_model(retriever, query, results, conversation_history):
     
     return answer
 
-def retrievingReponse(docId, query, conversation_history):
+def retrievingReponse(docId, query, conversation_history, token):
     retriever = vectorDB.as_retriever(
         search_type="similarity",
         search_kwargs={
@@ -130,11 +141,11 @@ def retrievingReponse(docId, query, conversation_history):
             unique_results.append(ans)
             seen_texts.add(result.page_content)
 
-    llm_result = using_ollama_model(retriever, query, results, conversation_history)
+    llm_result = using_ollama_model(retriever, query, results, conversation_history, token)
     return llm_result
 
 # The revised Gradio wrapper function
-def gradio_rag_wrapper(message, history):
+def gradio_rag_wrapper(message, history, token):
 
     print(history)
     # Check if a file has been uploaded
@@ -162,7 +173,7 @@ def gradio_rag_wrapper(message, history):
         rag_history.append({"question": user_text, "answer": bot_msg})
 
     docId = "42" # Use the docId from the uploaded file
-    response = retrievingReponse(docId, text_query, rag_history)
+    response = retrievingReponse(docId, text_query, rag_history, token)
     
     return response