Update app.py

app.py

@@ -34,7 +34,11 @@ h3 {
 device = "cuda" # for GPU usage or "cpu" for CPU usage
 
 tokenizer = AutoTokenizer.from_pretrained(MODEL)
-model = AutoModelForCausalLM.from_pretrained(MODEL).to(device)
+model = AutoModelForCausalLM.from_pretrained(
+    MODEL,
+    torch_dtype=torch.bfloat16,
+    device_map="auto",
+    ignore_mismatched_sizes=True).to(device)
 
 @spaces.GPU()
 def stream_chat(