Upload app.py with huggingface_hub

app.py

@@ -10,10 +10,11 @@ print("Loading model... This may take a few minutes.")
 tokenizer = AutoTokenizer.from_pretrained(model_name_or_path)
 model = AutoModelForCausalLM.from_pretrained(
     model_name_or_path,
-    torch_dtype=torch.bfloat16 if torch.cuda.is_available() else torch.float32,
+    torch_dtype=torch.bfloat16,
     device_map="auto"
 )
 
+@spaces.GPU(duration=120)
 def respond(message, history, system_message=None, max_tokens=None, temperature=None, top_p=None):
     """
     Generate response from Hunyuan-MT model