Fix: GPU usage issue

app.py

@@ -53,7 +53,6 @@ def combine_model_parts(model_dir="split_models", output_file="checkpoints/last.
     print(f"Model combined successfully: {output_file}")
     return output_file
 
-@spaces.GPU(enable_queue=True)
 def load_model():
     """
     Load the SmollmV2 model and tokenizer.
@@ -89,6 +88,8 @@ def load_model():
     except Exception as e:
         raise RuntimeError(f"Error loading model: {str(e)}")
 
+# Load the model globally
+model, tokenizer, device = load_model()
 
 @spaces.GPU(enable_queue=True)
 def generate_text(prompt, num_tokens, temperature=0.8, top_p=0.9):
@@ -142,13 +143,6 @@ def generate_text(prompt, num_tokens, temperature=0.8, top_p=0.9):
     except Exception as e:
         return f"Error during text generation: {str(e)}"
 
-# Load the model globally
-try:
-    model, tokenizer, device = load_model()
-except Exception as e:
-    print(f"Error initializing model: {str(e)}")
-    raise
-
 # Create the Gradio interface
 demo = gr.Interface(
     fn=generate_text,