Update app.py

app.py

@@ -27,6 +27,10 @@ from huggingface_hub import hf_hub_download
 from infer import load_model, eval_model
 from spkr import SpeakerEmbedding
 
+@spaces.GPU
+def dummy_gpu_check():
+    # Just to let Hugging Face Spaces know this app uses GPU
+    return torch.cuda.is_available()
 
 spkr_model = SpeakerEmbedding(device="cuda")
 model, tokenizer, tokenizer_voila, model_type = load_model("maitrix-org/Voila-chat", "maitrix-org/Voila-Tokenizer")