update GPU duration

app.py

@@ -68,7 +68,7 @@ def load_models():
 
 pipe_sd35, pipe_sdxl = load_models()
 
-@spaces.GPU
+@spaces.GPU(duration=360)
 def generate_image(
     model_name,
     seed,
@@ -97,7 +97,6 @@ def generate_image(
 
         pipe.to(device)
         pipe.enable_model_cpu_offload()
-      #   os.makedirs('./weights', exist_ok=True)
         os.system('huggingface-cli download sst12345/CoRe2 weights/sd35_noise_model.pth weights/sdxl_noise_model.pth --local-dir ./weights')
         # TODO: load noise model
         if method == 'core' or method == 'z-core':
@@ -105,7 +104,7 @@ def generate_image(
             from diffusion_pipeline.lora import replace_linear_with_lora, lora_true
 
             if model_name == 'sd35':
-                refine_model = PromptSD35Net().to(device)
+                refine_model = PromptSD35Net()
                 replace_linear_with_lora(refine_model, rank=64, alpha=1.0, number_of_lora=28)
                 lora_true(refine_model, lora_idx=0)
                 checkpoint = torch.load('./weights/weights/sd35_noise_model.pth', map_location='cpu')
@@ -117,10 +116,9 @@ def generate_image(
                 checkpoint = torch.load('./weights/weights/sdxl_noise_model.pth', map_location='cpu')
                 refine_model.load_state_dict(checkpoint)
 
-            print("Load Lora Success")
-            refine_model = refine_model.to(device)
             refine_model = refine_model.to(torch.bfloat16)
-
+            refine_model = refine_model.to(device)
+            print("Load Lora Success")
         # 根据模型类型设置形状
         if model_name == 'sdxl':
             shape = (1, 4, size // 8, size // 8)