Update app.py

app.py

@@ -10,7 +10,7 @@ chat_template = os.getenv('CHAT_TEMPLATE')
 
 # Interface variables
 model_name = model_id.split('/')[1].split('-GGUF')[0]
-title = f"👑 {model_name}"
+title = f"🇩🇪 {model_name}"
 description = f"Chat with <a href=\"https://huggingface.co/{model_id}\">{model_name}</a> in GGUF format ({quant})!"
 
 # Initialize the LLM
@@ -30,7 +30,7 @@ def chat_stream_completion(message, history, system_prompt):
     response = llm.create_chat_completion(
         messages=messages_prompts,
         stream=True,
-        stop=["[INST]"]
+        stop=["<|im_end|>"]
     )
     message_repl = ""
     for chunk in response:
@@ -43,11 +43,11 @@ gr.ChatInterface(
     fn=chat_stream_completion,
     title=title,
     description=description,
-    additional_inputs=[gr.Textbox("You are helpful assistant.")],
+    additional_inputs=[gr.Textbox("Du bist ein hilfreicher Assistent.")],
     additional_inputs_accordion="📝 System prompt",
     examples=[
-        ["What is a Large Language Model?"],
-        ["What's 9+2-1?"],
-        ["Write Python code to print the Fibonacci sequence"]
+        ["Was ist ein Large Language Model?"],
+        ["Was ist 9+2-1?"],
+        ["Schreibe Python code um die Fibonacci-Reihenfolge auszugeben."]
     ]
 ).queue().launch(server_name="0.0.0.0")