Update app.py

app.py

@@ -10,8 +10,8 @@ from transformers import AutoModelForCausalLM, AutoTokenizer, BitsAndBytesConfig
 # Update model configuration for Mistral-small-24B
 MODEL_ID = "baconnier/Napoleon_24B_V0.0"
 CHAT_TEMPLATE = "mistral"  # Mistral uses its own chat template
-MODEL_NAME = "Napoleon 24B"
-CONTEXT_LENGTH = 32768  # Mistral supports longer context
+MODEL_NAME = "🦅 Napoleon 🦅"
+CONTEXT_LENGTH = 2048  # Mistral supports longer context
 COLOR = "black"
 EMOJI = "🇫🇷"  # Mistral-themed emoji
 DESCRIPTION = f"This is {MODEL_NAME} model, a powerful 24B parameter language model from Mistral AI."
@@ -143,7 +143,8 @@ gr.ChatInterface(
     additional_inputs=[
         gr.Textbox(SYSTEM_MESSAGE, label="System prompt", visible=False),  # Hidden system prompt
         gr.Slider(0, 1, 0.7, label="Temperature"),  # Adjusted default for Mistral
-        gr.Slider(0, 32768, 12000, label="Max new tokens"),  # Increased for longer context
+        gr.Slider(0, 2048, 1024, label="Max new tokens"),  # Increased for longer context
+        #gr.Slider(0, 32768, 12000, label="Max new tokens"),  # Increased for longer context
         gr.Slider(1, 100, 50, label="Top K sampling"),
         gr.Slider(0, 2, 1.1, label="Repetition penalty"),
         gr.Slider(0, 1, 0.95, label="Top P sampling"),