Fix missing images in inference/controlnet.ipynb

inference/controlnet.ipynb

@@ -304,8 +304,8 @@
    ],
    "source": [
     "batch_size = 4\n",
-    "url = \"https://cdn.discordapp.com/attachments/1121232062708457508/1204787053892603914/cat_dog.png?ex=65d60061&is=65c38b61&hm=37c3d179a39b1eca4b8894e3c239930cedcbb965da00ae2209cca45f883f86f4&\"\n",
-    "images = resize_image(download_image(url)).unsqueeze(0).expand(batch_size, -1, -1, -1)\n",
+    "filename = \"assets/cat_dog.png\"\n",
+    "images = resize_image(load_image(filename)).unsqueeze(0).expand(batch_size, -1, -1, -1)\n",
     "\n",
     "batch = {'images': images}\n",
     "\n",
@@ -370,8 +370,8 @@
    ],
    "source": [
     "batch_size = 4\n",
-    "url = \"https://cdn.discordapp.com/attachments/1039261364935462942/1200109692978999317/three_people.png?ex=65c4fc3f&is=65b2873f&hm=064a8cebea5560b74e7088be9d1399a5fe48863d1581e65ea9d6734725f4c8d3&\"\n",
-    "images = resize_image(download_image(url)).unsqueeze(0).expand(batch_size, -1, -1, -1)\n",
+    "filename = \"assets/three_people.png\"\n",
+    "images = resize_image(load_image(filename)).unsqueeze(0).expand(batch_size, -1, -1, -1)\n",
     "\n",
     "batch = {'images': images}\n",
     "\n",
@@ -430,8 +430,8 @@
    ],
    "source": [
     "batch_size = 4\n",
-    "url = \"https://media.discordapp.net/attachments/1177378292765036716/1205484279405219861/image.png?ex=65d889b9&is=65c614b9&hm=0722ab9707b48d677316c0b4de5e51702b43eac1e27b76c268a069ec67ff6d15&=&format=webp&quality=lossless&width=861&height=859\"\n",
-    "images = resize_image(download_image(url)).unsqueeze(0).expand(batch_size, -1, -1, -1)\n",
+    "filename = \"assets/astronaut.png\"\n",
+    "images = resize_image(load_image(filename)).unsqueeze(0).expand(batch_size, -1, -1, -1)\n",
     "sketch = False\n",
     "\n",
     "batch = {'images': images}\n",
@@ -489,10 +489,8 @@
    "source": [
     "batch_size = 4\n",
     "cnet_override = None\n",
-    "# url = \"https://media.discordapp.net/attachments/1121232062708457508/1205134173053132810/image.png?ex=65d743a9&is=65c4cea9&hm=48dc4901514caada29271f48d76431f3a648940f2fda9e643a6bb693c906cc09&=&format=webp&quality=lossless&width=862&height=857\"\n",
-    "# url = \"https://cdn.discordapp.com/attachments/1121232062708457508/1204787053892603914/cat_dog.png?ex=65d60061&is=65c38b61&hm=37c3d179a39b1eca4b8894e3c239930cedcbb965da00ae2209cca45f883f86f4&\"\n",
-    "url = \"https://cdn.discordapp.com/attachments/1121232062708457508/1205110687538479145/A_photograph_of_a_sunflower_with_sunglasses_on_in__3.jpg?ex=65d72dc9&is=65c4b8c9&hm=72172e774ce6cda618503b3778b844de05cd1208b61e185d8418db512fb2858a&\"\n",
-    "images = resize_image(download_image(url)).unsqueeze(0).expand(batch_size, -1, -1, -1)\n",
+    "filename = \"assets/sunflower.png\"\n",
+    "images = resize_image(load_image(filename)).unsqueeze(0).expand(batch_size, -1, -1, -1)\n",
     "\n",
     "batch = {'images': images}\n",
     "\n",
@@ -662,7 +660,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.8.10"
+   "version": "3.11.4"
   }
  },
  "nbformat": 4,

inference/utils.py

@@ -9,8 +9,12 @@
 import torchvision.transforms.functional as F
 
 
+def load_image(fp):
+    return PIL.Image.open(fp).convert("RGB")
+
+
 def download_image(url):
-    return PIL.Image.open(requests.get(url, stream=True).raw).convert("RGB")
+    return load_image(requests.get(url, stream=True).raw)
 
 
 def resize_image(image, size=768):
@@ -30,7 +34,7 @@ def show_images(images, rows=None, cols=None, return_images=False, **kwargs):
         images = images.repeat(1, 3, 1, 1)
     elif images.size(1) > 3:
         images = images[:, :3]
-    
+
     if rows is None:
         rows = 1
     if cols is None:
@@ -42,7 +46,7 @@ def show_images(images, rows=None, cols=None, return_images=False, **kwargs):
     for i, img in enumerate(images):
         img = torchvision.transforms.functional.to_pil_image(img.clamp(0, 1))
         grid.paste(img, box=(i % cols * w, i // cols * h))
-    
+
     bio = BytesIO()
     grid.save(bio, format='png')
     display(Image(bio.getvalue(), format='png'))
@@ -56,9 +60,9 @@ def calculate_latent_sizes(height=1024, width=1024, batch_size=4, compression_fa
     latent_height = ceil(height / compression_factor_b)
     latent_width = ceil(width / compression_factor_b)
     stage_c_latent_shape = (batch_size, 16, latent_height, latent_width)
-    
+
     latent_height = ceil(height / compression_factor_a)
     latent_width = ceil(width / compression_factor_a)
     stage_b_latent_shape = (batch_size, 4, latent_height, latent_width)
-    
+
     return stage_c_latent_shape, stage_b_latent_shape