support size

app.py

@@ -9,6 +9,27 @@ import os
 import numpy as np
 
 
+def convert_to_fit_size(original_width_and_height, maximum_size = 2048):
+    width, height =original_width_and_height
+    if width <= maximum_size and height <= maximum_size:
+        return width,height
+    
+    if width > height:
+        scaling_factor = maximum_size / width
+    else:
+        scaling_factor = maximum_size / height
+
+    new_width = int(width * scaling_factor)
+    new_height = int(height * scaling_factor)
+    return new_width, new_height
+
+def adjust_to_multiple_of_32(width: int, height: int):
+    width = width - (width % 32)
+    height = height - (height % 32)
+    return width, height
+
+
+
 dtype = torch.bfloat16
 device = "cuda" if torch.cuda.is_available() else "cpu"
 
@@ -57,7 +78,14 @@ def process_images(image, image2=None,prompt="a girl",inpaint_model="black-fores
         generator = torch.Generator("cuda").manual_seed(seed)
         generators.append(generator)
         
-        output = pipe(prompt=prompt, image=image, mask_image=mask_image,generator=generator,strength=strength,
+        width,height = convert_to_fit_size(image.size)
+        print(f"fit {width}x{height}")
+        width,height = adjust_to_multiple_of_32(width,height)
+        print(f"multiple {width}x{height}")
+        image = image.resize((width, height), Image.LANCZOS)
+        mask = mask.resize((width, height), Image.NEAREST)
+
+        output = pipe(prompt=prompt, image=image, mask_image=mask_image,generator=generator,strength=strength,width=width,height=height,
                       guidance_scale=0,num_inference_steps=num_inference_steps,max_sequence_length=256)
 
         return output.images[0]