update commit with phi-3 mini 112

app.py

@@ -16,7 +16,7 @@ model = AutoModelForCausalLM.from_pretrained(
 )
 
 # ✅ Create pipeline — device=0 will use CUDA if available
-pipe = pipeline("text-generation", model=model, tokenizer=tokenizer, device=0)
+pipe = pipeline("text-generation", model=model, tokenizer=tokenizer)
 
 # ✅ Detect actual device
 device = torch.device("cuda" if torch.cuda.is_available() else "cpu")