Update app.py

app.py

@@ -11,6 +11,14 @@ from transformers import TextIteratorStreamer
 from threading import Thread
 import torch
 
+device = 'cuda'
+model_id = "google/gemma-3-4b-it"
+processor = AutoProcessor.from_pretrained(model_id, use_fast=True, padding_side="left")
+model = Gemma3ForConditionalGeneration.from_pretrained(
+    model_id,
+    torch_dtype=torch.bfloat16
+).to(device).eval()
+
 def process(message, history):
     """Generate the model response in streaming mode given message and history
     """