init

app.py

@@ -2,6 +2,8 @@ import gradio as gr
 from transformers import Qwen2AudioForConditionalGeneration, AutoProcessor
 import librosa
 
+import spaces
+
 
 MODEL_ID = "Qwen/Qwen2-Audio-7B-Instruct"
 
@@ -9,6 +11,7 @@ processor = AutoProcessor.from_pretrained(MODEL_ID)
 model = Qwen2AudioForConditionalGeneration.from_pretrained(MODEL_ID, device_map="auto")
 
 
+@spaces.GPU
 def run_qwen2audio(audio_path: str, instruction: str, max_tokens: int = 256) -> str:
     if not audio_path:
         return "Please upload an audio file."

requirements.txt

@@ -5,4 +5,5 @@ accelerate>=0.30.0
 librosa>=0.10.1
 soundfile>=0.12.1
 sentencepiece>=0.1.99
+spaces