Update app.py

app.py

@@ -13,30 +13,8 @@ from torchvision import transforms
 
 device = torch.device('cuda:0' if torch.cuda.is_available() else 'cpu')
 dtype = torch.float16 if torch.cuda.is_available() else torch.float32
-image_encoder = CLIPVisionModelWithProjection.from_pretrained(
-    "h94/IP-Adapter",
-    subfolder="models/image_encoder",
-    torch_dtype=dtype
-)
-pipeline = AutoPipelineForText2Image.from_pretrained("stabilityai/stable-diffusion-xl-base-1.0",image_encoder=image_encoder, torch_dtype=dtype).to(device)
-pipeline.load_ip_adapter("h94/IP-Adapter", subfolder="sdxl_models", weight_name="ip-adapter-plus_sdxl_vit-h.safetensors")
-
-def prepare_image(image_path_or_url):
-    # Load the image
-    image = load_image(image_path_or_url)
-    
-    # Convert to tensor and move to correct device and dtype
-    transform = transforms.Compose([
-        transforms.ToTensor(),
-        transforms.Resize((1024, 1024), interpolation=transforms.InterpolationMode.BICUBIC)
-    ])
-    image_tensor = transform(image).unsqueeze(0)  # Add batch dimension
-    return image_tensor.to(device=device, dtype=dtype)
-
-def randomize_seed_fn(seed: int, randomize_seed: bool) -> int:
-    if randomize_seed:
-        seed = random.randint(0, 2000)
-    return seed
+pipeline = AutoPipelineForText2Image.from_pretrained("stabilityai/stable-diffusion-xl-base-1.0", torch_dtype=dtype).to(device)
+pipeline.load_ip_adapter("h94/IP-Adapter", subfolder="sdxl_models", weight_name="ip-adapter_sdxl.bin")
 
 @spaces.GPU(enable_queue=True)
 def create_image(image_pil,
@@ -65,7 +43,7 @@ def create_image(image_pil,
             }
         pipeline.set_ip_adapter_scale(scale)
 
-    style_image = prepare_image(image_pil)
+    style_image = load_image(image_pil)
     generator = torch.Generator(device=device).manual_seed(randomize_seed_fn(seed, False))
     
     image = pipeline(