diff --git a/hubconf.py b/hubconf.py
index 4ed1837..3cacaef 100644
--- a/hubconf.py
+++ b/hubconf.py
@@ -1,18 +1,17 @@
 import torch
 
 def generator(pretrained=True, device="cpu", progress=True, check_hash=True):
-    release_url = "https://github.com/xhlulu/animegan2-pytorch/releases/download/weights"
+    from model import Generator
+
+    release_url = "https://github.com/xhlulu/animegan2-pytorch/raw/main/weights"
     known = {
         name: f"{release_url}/{name}.pt"
         for name in [
-            'face_paint_512_v0', 'face_paint_512_v2'
+            'celeba_distill', 'face_paint_512_v1', 'face_paint_512_v2', 'paprika'
         ]
     }
 
-    from model import Generator
-
     device = torch.device(device)
-
     model = Generator().to(device)
 
     if type(pretrained) == str:
@@ -34,14 +33,14 @@ def generator(pretrained=True, device="cpu", progress=True, check_hash=True):
     return model
 
 
-def face2paint(device="cpu"):
+def face2paint(device="cpu", size=512):
     from PIL import Image
     from torchvision.transforms.functional import to_tensor, to_pil_image
 
     def face2paint(
         model: torch.nn.Module,
         img: Image.Image,
-        size: int,
+        size: int = size,
         side_by_side: bool = True,
         device: str = device,
     ) -> Image.Image:
@@ -50,13 +49,14 @@ def face2paint(device="cpu"):
         img = img.crop(((w - s) // 2, (h - s) // 2, (w + s) // 2, (h + s) // 2))
         img = img.resize((size, size), Image.LANCZOS)
 
-        input = to_tensor(img).unsqueeze(0) * 2 - 1
-        output = model(input.to(device)).cpu()[0]
+        with torch.no_grad():
+            input = to_tensor(img).unsqueeze(0) * 2 - 1
+            output = model(input.to(device)).cpu()[0]
 
-        if side_by_side:
-            output = torch.cat([input[0], output], dim=2)
+            if side_by_side:
+                output = torch.cat([input[0], output], dim=2)
 
-        output = (output * 0.5 + 0.5).clip(0, 1)
+            output = (output * 0.5 + 0.5).clip(0, 1)
 
         return to_pil_image(output)