Spaces:
Runtime error
Runtime error
Update app.py
Browse files
app.py
CHANGED
@@ -1,19 +1,21 @@
|
|
1 |
import gradio as gr
|
2 |
import torch
|
3 |
-
from diffusers import
|
|
|
4 |
from PIL import Image
|
5 |
import numpy as np
|
6 |
import os
|
7 |
|
8 |
-
# Model loading with
|
9 |
cache_dir = "./model_cache"
|
10 |
os.makedirs(cache_dir, exist_ok=True)
|
11 |
|
12 |
-
# Load model
|
13 |
-
pipe =
|
14 |
"camenduru/cv_ddcolor_image-colorization",
|
15 |
torch_dtype=torch.float16,
|
16 |
-
cache_dir=cache_dir
|
|
|
17 |
).to("cuda")
|
18 |
|
19 |
def colorize_image(input_image):
|
@@ -22,16 +24,21 @@ def colorize_image(input_image):
|
|
22 |
if input_image.mode != 'L':
|
23 |
input_image = input_image.convert('L')
|
24 |
|
25 |
-
# Resize to model's expected input size
|
26 |
-
target_size = (
|
27 |
resized_image = input_image.resize(target_size)
|
28 |
|
29 |
-
# Convert to
|
30 |
-
|
31 |
|
32 |
# Generate colorized image
|
33 |
with torch.inference_mode():
|
34 |
-
result = pipe(
|
|
|
|
|
|
|
|
|
|
|
35 |
|
36 |
return result
|
37 |
|
|
|
1 |
import gradio as gr
|
2 |
import torch
|
3 |
+
from diffusers import AutoPipelineForImage2Image
|
4 |
+
from transformers import pipeline
|
5 |
from PIL import Image
|
6 |
import numpy as np
|
7 |
import os
|
8 |
|
9 |
+
# Model loading with dynamic pipeline selection
|
10 |
cache_dir = "./model_cache"
|
11 |
os.makedirs(cache_dir, exist_ok=True)
|
12 |
|
13 |
+
# Load model using AutoPipeline
|
14 |
+
pipe = AutoPipelineForImage2Image.from_pretrained(
|
15 |
"camenduru/cv_ddcolor_image-colorization",
|
16 |
torch_dtype=torch.float16,
|
17 |
+
cache_dir=cache_dir,
|
18 |
+
variant="fp16"
|
19 |
).to("cuda")
|
20 |
|
21 |
def colorize_image(input_image):
|
|
|
24 |
if input_image.mode != 'L':
|
25 |
input_image = input_image.convert('L')
|
26 |
|
27 |
+
# Resize to model's expected input size
|
28 |
+
target_size = (512, 512) # Increased resolution for better quality
|
29 |
resized_image = input_image.resize(target_size)
|
30 |
|
31 |
+
# Convert to RGB as required by model
|
32 |
+
grayscale_image = resized_image.convert("RGB")
|
33 |
|
34 |
# Generate colorized image
|
35 |
with torch.inference_mode():
|
36 |
+
result = pipe(
|
37 |
+
prompt="colorized photo",
|
38 |
+
image=grayscale_image,
|
39 |
+
num_inference_steps=20,
|
40 |
+
strength=0.8
|
41 |
+
).images[0]
|
42 |
|
43 |
return result
|
44 |
|