Update app.py

app.py

@@ -27,11 +27,10 @@ import whisper
 # print("TTS Loaded!")
 
 
-@spaces.GPU
 def load_whisper():
     return whisper.load_model("medium").to("cuda")
 
-@spaces.GPU()
+
 def load_tts():
     tts_model = VitsModel.from_pretrained("facebook/mms-tts-pol")
     tts_model.to("cuda")
@@ -49,10 +48,6 @@ def read_txt():
     return lines 
 
 
-##### Chat z LLAMA ####
-##### Chat z LLAMA ####
-##### Chat z LLAMA ####
-@spaces.GPU
 def _load_model_tokenizer():
     model_id = 'tangger/Qwen-7B-Chat'
     tokenizer = AutoTokenizer.from_pretrained(model_id,  trust_remote_code=True)
@@ -167,6 +162,7 @@ def regenerate(_chatbot, _task_history):
     _chatbot.pop(-1)
     yield from predict(item[0], _chatbot, _task_history)
 
+@spaces.GPU
 with gr.Blocks() as chat_demo:
     chatbot = gr.Chatbot(label='Llama Voice Chatbot', elem_classes="control-height")
     query = gr.Textbox(lines=2, label='Input')