Spaces:
Running
on
Zero
Running
on
Zero
Update app.py
Browse files
app.py
CHANGED
@@ -55,6 +55,32 @@ def rewrite_prompt(input_prompt):
|
|
55 |
|
56 |
# --- 2. Preprocessor Functions ---
|
57 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
58 |
def extract_canny(input_image):
|
59 |
image = np.array(input_image)
|
60 |
image = cv2.Canny(image, 100, 200)
|
@@ -104,7 +130,7 @@ anyline = AnylineDetector.from_pretrained("TheMistoAI/MistoLine", filename="MTEE
|
|
104 |
print("All models loaded.")
|
105 |
|
106 |
def get_control_image(input_image, control_mode):
|
107 |
-
"""A master function to select and run the correct preprocessor."""
|
108 |
if control_mode == "Canny":
|
109 |
return extract_canny(input_image)
|
110 |
elif control_mode == "Soft Edge":
|
@@ -143,6 +169,8 @@ def generate(
|
|
143 |
if not prompt:
|
144 |
raise gr.Error("Please enter a prompt.")
|
145 |
|
|
|
|
|
146 |
if randomize_seed:
|
147 |
seed = random.randint(0, MAX_SEED)
|
148 |
|
@@ -151,7 +179,7 @@ def generate(
|
|
151 |
print(f"Original prompt: {prompt}\nEnhanced prompt: {enhanced_prompt}")
|
152 |
prompt = enhanced_prompt
|
153 |
|
154 |
-
control_image = get_control_image(
|
155 |
generator = torch.Generator(device=device).manual_seed(int(seed))
|
156 |
|
157 |
generated_image = pipe(
|
@@ -159,8 +187,8 @@ def generate(
|
|
159 |
negative_prompt=negative_prompt,
|
160 |
control_image=control_image,
|
161 |
controlnet_conditioning_scale=controlnet_conditioning_scale,
|
162 |
-
width=
|
163 |
-
height=
|
164 |
num_inference_steps=int(num_inference_steps),
|
165 |
guidance_scale=guidance_scale,
|
166 |
generator=generator,
|
@@ -229,8 +257,4 @@ with gr.Blocks(css=css, theme=gr.themes.Citrus()) as demo:
|
|
229 |
)
|
230 |
|
231 |
if __name__ == "__main__":
|
232 |
-
if not os.path.exists("assets"):
|
233 |
-
os.makedirs("assets")
|
234 |
-
print("Created 'assets' directory. Please add example images for the Gradio examples to work.")
|
235 |
-
|
236 |
demo.launch()
|
|
|
55 |
|
56 |
# --- 2. Preprocessor Functions ---
|
57 |
|
58 |
+
def resize_image(input_image, max_size=1024):
|
59 |
+
"""
|
60 |
+
Resizes an image so that its longest side is `max_size` pixels,
|
61 |
+
maintaining aspect ratio. The final dimensions are made divisible by 8.
|
62 |
+
"""
|
63 |
+
w, h = input_image.size
|
64 |
+
aspect_ratio = w / h
|
65 |
+
|
66 |
+
if w > h:
|
67 |
+
new_w = max_size
|
68 |
+
new_h = int(new_w / aspect_ratio)
|
69 |
+
else:
|
70 |
+
new_h = max_size
|
71 |
+
new_w = int(new_h * aspect_ratio)
|
72 |
+
|
73 |
+
# Make dimensions divisible by 8
|
74 |
+
new_w = new_w - (new_w % 8)
|
75 |
+
new_h = new_h - (new_h % 8)
|
76 |
+
|
77 |
+
# Handle potential zero dimensions after rounding
|
78 |
+
if new_w == 0: new_w = 8
|
79 |
+
if new_h == 0: new_h = 8
|
80 |
+
|
81 |
+
return input_image.resize((new_w, new_h), Image.Resampling.LANCZOS)
|
82 |
+
|
83 |
+
|
84 |
def extract_canny(input_image):
|
85 |
image = np.array(input_image)
|
86 |
image = cv2.Canny(image, 100, 200)
|
|
|
130 |
print("All models loaded.")
|
131 |
|
132 |
def get_control_image(input_image, control_mode):
|
133 |
+
"""A master function to select and run the correct preprocessor on a pre-resized image."""
|
134 |
if control_mode == "Canny":
|
135 |
return extract_canny(input_image)
|
136 |
elif control_mode == "Soft Edge":
|
|
|
169 |
if not prompt:
|
170 |
raise gr.Error("Please enter a prompt.")
|
171 |
|
172 |
+
resized_image = resize_image(image, max_size=1024)
|
173 |
+
|
174 |
if randomize_seed:
|
175 |
seed = random.randint(0, MAX_SEED)
|
176 |
|
|
|
179 |
print(f"Original prompt: {prompt}\nEnhanced prompt: {enhanced_prompt}")
|
180 |
prompt = enhanced_prompt
|
181 |
|
182 |
+
control_image = get_control_image(resized_image, conditioning)
|
183 |
generator = torch.Generator(device=device).manual_seed(int(seed))
|
184 |
|
185 |
generated_image = pipe(
|
|
|
187 |
negative_prompt=negative_prompt,
|
188 |
control_image=control_image,
|
189 |
controlnet_conditioning_scale=controlnet_conditioning_scale,
|
190 |
+
width=resized_image.width,
|
191 |
+
height=resized_image.height,
|
192 |
num_inference_steps=int(num_inference_steps),
|
193 |
guidance_scale=guidance_scale,
|
194 |
generator=generator,
|
|
|
257 |
)
|
258 |
|
259 |
if __name__ == "__main__":
|
|
|
|
|
|
|
|
|
260 |
demo.launch()
|