Update app.py

app.py

@@ -1,6 +1,7 @@
 import gradio as gr
 import os
 from huggingface_hub import InferenceClient
+import time
 
 hf_token = os.getenv("hf_token")
 
@@ -24,20 +25,22 @@ def get_response(user_input):
     response = ""
     for chunk in stream:
         response += chunk.choices[0].delta.content
-    return response
+        yield response  # Yielding progressively as the model generates output
+        time.sleep(0.05)  # Optional: Adjust speed of the stream (in seconds)
 
 def chat_interface():
     with gr.Blocks() as demo:
-        with gr.Row():
-            with gr.Column(scale=1):
+        with gr.Row():  # No 'min_width' argument here
+            with gr.Column():  # No 'min_width' argument here
                 chat_output = gr.Chatbot(
                     elem_id="chat-box",
                     label="Xylaria 1.4 Senoa Chatbot",
-                    show_label=False
+                    show_label=False,
+                    type="messages"  # Specify type for correct message format
                 )
 
-        with gr.Row(elem_id="input-row", scale=0.2):
-            with gr.Column(scale=0.8):
+        with gr.Row(elem_id="input-row"):
+            with gr.Column():  # No 'min_width' argument here
                 input_textbox = gr.Textbox(
                     label="Type your message", 
                     placeholder="Ask me anything...",
@@ -47,17 +50,33 @@ def chat_interface():
                     elem_id="user-input",
                     show_label=False
                 )
-            with gr.Column(scale=0.2):
+            with gr.Column():
                 send_button = gr.Button("Send", elem_id="send-btn")
 
         def submit_input(user_input, chat_history):
-            response = get_response(user_input)
-            chat_history.append((user_input, response))
-            return "", chat_history
+            chat_history.append({"role": "user", "content": user_input})
+            return "", chat_history  # Clear input field
 
         input_textbox.submit(submit_input, [input_textbox, chat_output], [input_textbox, chat_output])
         send_button.click(submit_input, [input_textbox, chat_output], [input_textbox, chat_output])
 
+        def handle_response(user_input, chat_history):
+            # Initialize chat_history if it's empty
+            if not chat_history:
+                chat_history = [{"role": "system", "content": "you are xylaria 1.4 senoa, developed by sk md saad amin"}]
+
+            # Add user input
+            chat_history.append({"role": "user", "content": user_input})
+
+            # Get response from model
+            response_stream = get_response(user_input)
+            for partial_response in response_stream:
+                chat_history[-1] = {"role": "assistant", "content": partial_response}
+                yield "", chat_history  # Return the updated chat history progressively
+
+        input_textbox.submit(handle_response, [input_textbox, chat_output], [input_textbox, chat_output])
+        send_button.click(handle_response, [input_textbox, chat_output], [input_textbox, chat_output])
+
     demo.css = """
     #input-row {
         position: absolute;
@@ -72,7 +91,7 @@ def chat_interface():
         overflow-y: scroll;
     }
     """
-    
+
     return demo
 
 demo = chat_interface()