trying cuda

handler.py

@@ -50,9 +50,8 @@ class EndpointHandler():
         pretrained_base_model_path_unet = os.path.join(base_dir, 'pretrained_weights', 'stable-diffusion-v1-5', 'unet')
         print("model path is " + pretrained_base_model_path_unet)
         reference_unet = UNet2DConditionModel.from_pretrained(
-            pretrained_base_model_path_unet,
-            local_files_only=True
-        ).to(device, dtype=self.weight_dtype)
+            pretrained_base_model_path_unet
+        ).to(dtype=self.weight_dtype, device="cuda")
 
         inference_config_path = os.path.join(base_dir, 'configs', 'inference', 'inference_v2.yaml')
         motion_module_path = os.path.join(base_dir, 'pretrained_weights', 'motion_module.pth')