Update app.py

app.py

@@ -1,63 +1,67 @@
 import gradio as gr
 from huggingface_hub import InferenceClient
+from unsloth.chat_templates import get_chat_template
 
-"""
-For more information on `huggingface_hub` Inference API support, please check the docs: https://huggingface.co/docs/huggingface_hub/v0.22.2/en/guides/inference
-"""
+# Initialize the InferenceClient with the appropriate model
 client = InferenceClient("wop/kosmox")
 
+# Define the chat template and tokenizer configuration
+tokenizer = get_chat_template(
+    tokenizer=None,  # Assuming you need to pass an actual tokenizer here
+    chat_template="phi-3",
+    mapping={"role": "from", "content": "value", "user": "human", "assistant": "gpt"},
+)
+
+def format_messages(system_message, history, user_message):
+    # Create a formatted string according to the specified chat template
+    formatted_message = "<s>\n"
+    if system_message:
+        formatted_message += f"{system_message}\n"
+
+    for user_msg, assistant_msg in history:
+        if user_msg:
+            formatted_message += f"{user_msg}\n"
+        if assistant_msg:
+            formatted_message += f"{assistant_msg}\n"
+    
+    formatted_message += f"{user_message}\n"
+    return formatted_message
 
 def respond(
-    message,
+    message: str,
     history: list[tuple[str, str]],
-    system_message,
-    max_tokens,
-    temperature,
-    top_p,
+    system_message: str,
+    max_tokens: int,
+    temperature: float,
+    top_p: float,
 ):
-    messages = [{"role": "system", "content": system_message}]
-
-    for val in history:
-        if val[0]:
-            messages.append({"role": "user", "content": val[0]})
-        if val[1]:
-            messages.append({"role": "assistant", "content": val[1]})
-
-    messages.append({"role": "user", "content": message})
+    # Format the messages
+    formatted_message = format_messages(system_message, history, message)
 
     response = ""
 
+    # Stream the response from the model
     for message in client.chat_completion(
-        messages,
+        formatted_message,
         max_tokens=max_tokens,
         stream=True,
         temperature=temperature,
         top_p=top_p,
     ):
         token = message.choices[0].delta.content
-
         response += token
         yield response
 
-"""
-For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
-"""
+# Define the Gradio interface
 demo = gr.ChatInterface(
-    respond,
+    fn=respond,
     additional_inputs=[
-        gr.Textbox(value="You are a friendly Chatbot.", label="System message"),
+        gr.Textbox(value="You are AI.", label="System message"),
         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)",
-        ),
+        gr.Slider(minimum=0.1, maximum=1.0, value=0.95, step=0.05, label="Top-p (nucleus sampling)"),
     ],
 )
 
-
 if __name__ == "__main__":
-    demo.launch()
+    demo.launch()