Update app.py

app.py

@@ -1,9 +1,5 @@
 import gradio as gr
 from huggingface_hub import InferenceClient
-import os
-
-# 🔹 Load HF token from Space Secrets
-HF_TOKEN = os.environ.get('telemedpro')  # Add your token in Space Settings → Secrets
 
 # 🔹 Fixed persona system message
 PERSONA_MESSAGE = (
@@ -13,54 +9,66 @@ PERSONA_MESSAGE = (
     "You behave politely, patiently, and with care, like a trusted family doctor."
 )
 
-# 🔹 Initialize InferenceClient once
-client = InferenceClient(token=HF_TOKEN, model="m42-health/Llama3-Med42-70B")
-
-# 🔹 Respond function (non-streaming for stability)
-def respond(message, history, max_tokens=512, temperature=0.7, top_p=0.95):
-    try:
-        messages = [{"role": "system", "content": PERSONA_MESSAGE}]
+def respond(
+    message,
+    history: list[dict[str, str]],
+    max_tokens,
+    temperature,
+    top_p,
+    hf_token: gr.OAuthToken,
+):
+    """
+    Respond function for Gradio ChatInterface. Uses Hugging Face InferenceClient with fixed persona.
+    """
+    client = InferenceClient(token=hf_token.token, model="m42-health/Llama3-Med42-70B")
 
-        # Append previous conversation safely
-        if history:
-            for h in history:
-                user_msg = h[0] if h[0] else ""
-                ai_msg = h[1] if h[1] else ""
-                messages.append({"role": "user", "content": user_msg})
-                messages.append({"role": "assistant", "content": ai_msg})
+    # Always include persona as system role
+    messages = [{"role": "system", "content": PERSONA_MESSAGE}]
 
-        # Append current user message
-        messages.append({"role": "user", "content": message})
+    # Add conversation history
+    if history:
+        for h in history:
+            # Ensure history is in proper format
+            messages.append({"role": "user", "content": h["user"]})
+            if "assistant" in h:
+                messages.append({"role": "assistant", "content": h["assistant"]})
 
-        # 🔹 Non-streaming call for stability
-        result = client.chat_completion(
-            messages,
-            max_tokens=max_tokens,
-            temperature=temperature,
-            top_p=top_p
-        )
+    # Append current user message
+    messages.append({"role": "user", "content": message})
 
-        return result.choices[0].message.content
+    response = ""
 
-    except Exception as e:
-        return f"⚠️ Space error: {e}"
+    # Stream model output token by token
+    for msg in client.chat_completion(
+        messages,
+        max_tokens=max_tokens,
+        stream=True,
+        temperature=temperature,
+        top_p=top_p,
+    ):
+        if len(msg.choices) > 0 and hasattr(msg.choices[0].delta, "content"):
+            token = msg.choices[0].delta.content
+            response += token
+            yield response
 
-# 🔹 Gradio Chat Interface
+# 🔹 Chatbot Interface
 chatbot = gr.ChatInterface(
     respond,
     type="messages",
     additional_inputs=[
-        gr.Slider(1, 2048, value=512, step=1, label="Max tokens"),
-        gr.Slider(0.1, 2.0, value=0.7, step=0.1, label="Temperature"),
-        gr.Slider(0.1, 1.0, value=0.95, step=0.05, label="Top-p"),
-    ]
+        gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
+        gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
+        gr.Slider(minimum=0.1, maximum=1.0, value=0.95, step=0.05, label="Top-p (nucleus sampling)"),
+    ],
 )
 
-# 🔹 Layout
+# 🔹 App Layout
 with gr.Blocks() as demo:
-    gr.Markdown("## 🩺 AI Health Mentor — Dr. Alex")
+    with gr.Sidebar():
+        gr.LoginButton()  # Hugging Face login
+        gr.Markdown("### 🩺 Persona: Dr. Alex\nThis chatbot behaves as a helpful doctor.")
     chatbot.render()
 
-# 🔹 Launch Space
+# 🔹 Run app
 if __name__ == "__main__":
-    demo.launch(show_error=True)
+    demo.launch()