Spaces:
Running
on
Zero
Running
on
Zero
add examples
Browse files
app.py
CHANGED
@@ -3,16 +3,15 @@ from PIL import Image
|
|
3 |
import os
|
4 |
import spaces
|
5 |
|
6 |
-
# os.environ['CUDA_VISIBLE_DEVICES'] = '7'
|
7 |
-
|
8 |
from OmniGen import OmniGenPipeline
|
9 |
|
10 |
-
pipe = OmniGenPipeline.from_pretrained(
|
11 |
-
|
|
|
12 |
|
13 |
-
# 示例处理函数:生成图像
|
14 |
@spaces.GPU
|
15 |
-
|
|
|
16 |
input_images = [img1, img2, img3]
|
17 |
# 去除 None
|
18 |
input_images = [img for img in input_images if img is not None]
|
@@ -24,38 +23,91 @@ def generate_image(text, img1, img2, img3, height, width, guidance_scale):
|
|
24 |
input_images=input_images,
|
25 |
height=height,
|
26 |
width=width,
|
27 |
-
guidance_scale=guidance_scale,
|
28 |
img_guidance_scale=1.6,
|
|
|
29 |
separate_cfg_infer=True,
|
30 |
-
use_kv_cache=False
|
31 |
)
|
32 |
img = output[0]
|
33 |
return img
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
34 |
|
35 |
# Gradio 接口
|
36 |
with gr.Blocks() as demo:
|
37 |
gr.Markdown("## Text + Multiple Images to Image Generator")
|
38 |
-
|
39 |
with gr.Row():
|
40 |
with gr.Column():
|
41 |
# 文本输入框
|
42 |
-
prompt_input = gr.Textbox(
|
43 |
-
|
44 |
-
|
45 |
-
|
46 |
-
|
47 |
-
|
48 |
-
|
|
|
|
|
|
|
49 |
# 高度和宽度滑块
|
50 |
-
height_input = gr.Slider(
|
51 |
-
|
52 |
-
|
|
|
|
|
|
|
|
|
53 |
# 引导尺度输入
|
54 |
-
guidance_scale_input = gr.Slider(
|
55 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
56 |
# 生成按钮
|
57 |
generate_button = gr.Button("Generate Image")
|
58 |
-
|
59 |
with gr.Column():
|
60 |
# 输出图像框
|
61 |
output_image = gr.Image(label="Output Image")
|
@@ -63,8 +115,33 @@ with gr.Blocks() as demo:
|
|
63 |
# 按钮点击事件
|
64 |
generate_button.click(
|
65 |
generate_image,
|
66 |
-
inputs=[
|
67 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
68 |
)
|
69 |
|
70 |
# 启动应用
|
|
|
3 |
import os
|
4 |
import spaces
|
5 |
|
|
|
|
|
6 |
from OmniGen import OmniGenPipeline
|
7 |
|
8 |
+
pipe = OmniGenPipeline.from_pretrained(
|
9 |
+
"shitao/tmp-preview"
|
10 |
+
)
|
11 |
|
|
|
12 |
@spaces.GPU
|
13 |
+
# 示例处理函数:生成图像
|
14 |
+
def generate_image(text, img1, img2, img3, height, width, guidance_scale, inference_steps):
|
15 |
input_images = [img1, img2, img3]
|
16 |
# 去除 None
|
17 |
input_images = [img for img in input_images if img is not None]
|
|
|
23 |
input_images=input_images,
|
24 |
height=height,
|
25 |
width=width,
|
26 |
+
guidance_scale=guidance_scale,
|
27 |
img_guidance_scale=1.6,
|
28 |
+
num_inference_steps=inference_steps,
|
29 |
separate_cfg_infer=True,
|
30 |
+
use_kv_cache=False,
|
31 |
)
|
32 |
img = output[0]
|
33 |
return img
|
34 |
+
# def generate_image(text, img1, img2, img3, height, width, guidance_scale, inference_steps):
|
35 |
+
# input_images = []
|
36 |
+
# if img1:
|
37 |
+
# input_images.append(Image.open(img1))
|
38 |
+
# if img2:
|
39 |
+
# input_images.append(Image.open(img2))
|
40 |
+
# if img3:
|
41 |
+
# input_images.append(Image.open(img3))
|
42 |
+
|
43 |
+
# return input_images[0] if input_images else None
|
44 |
+
|
45 |
+
|
46 |
+
def get_example():
|
47 |
+
case = [
|
48 |
+
[
|
49 |
+
"A woman holds a bouquet of flowers and faces the camera. Thw woman is the one in <img><|image_1|></img>.",
|
50 |
+
"./imgs/test_cases/liuyifei.png",
|
51 |
+
None,
|
52 |
+
None,
|
53 |
+
1024,
|
54 |
+
1024,
|
55 |
+
3.0,
|
56 |
+
20,
|
57 |
+
],
|
58 |
+
[
|
59 |
+
"Three zebras are standing side by side on a vibrant savannah, each showcasing unique patterns and characteristics that highlight their individuality. The zebra on the left has a strikingly bold black and white stripe pattern, with wider stripes that create a dramatic contrast against its sleek body. In the middle, the zebra features a more subtle stripe arrangement, with thinner stripes that blend seamlessly into a slightly sandy-colored coat, giving it a softer appearance. On the right, the zebra's stripes are more irregular, with a distinct patch of brown fur near its shoulder, adding a layer of uniqueness to its overall look. Together, these zebras create a captivating scene, each representing the diverse beauty of their species in the wild. The right zebras is the zebras from <img><|image_1|></img>. The center zebras is from <img><|image_2|></img>. The left zebras is the zebras from <img><|image_3|></img>.",
|
60 |
+
"./imgs/test_cases/img1.jpg",
|
61 |
+
"./imgs/test_cases/img2.jpg",
|
62 |
+
"./imgs/test_cases/img3.jpg",
|
63 |
+
1024,
|
64 |
+
1024,
|
65 |
+
3.0,
|
66 |
+
20,
|
67 |
+
],
|
68 |
+
]
|
69 |
+
return case
|
70 |
+
|
71 |
+
def run_for_examples(text, img1, img2, img3, height, width, guidance_scale, inference_steps):
|
72 |
+
return generate_image(text, img1, img2, img3, height, width, guidance_scale, inference_steps)
|
73 |
+
|
74 |
|
75 |
# Gradio 接口
|
76 |
with gr.Blocks() as demo:
|
77 |
gr.Markdown("## Text + Multiple Images to Image Generator")
|
|
|
78 |
with gr.Row():
|
79 |
with gr.Column():
|
80 |
# 文本输入框
|
81 |
+
prompt_input = gr.Textbox(
|
82 |
+
label="Enter your prompt", placeholder="Type your prompt here..."
|
83 |
+
)
|
84 |
+
|
85 |
+
with gr.Row(equal_height=True):
|
86 |
+
# 图片上传框
|
87 |
+
image_input_1 = gr.Image(label="<img><|image_1|></img>", type="filepath")
|
88 |
+
image_input_2 = gr.Image(label="<img><|image_2|></img>", type="filepath")
|
89 |
+
image_input_3 = gr.Image(label="<img><|image_3|></img>", type="filepath")
|
90 |
+
|
91 |
# 高度和宽度滑块
|
92 |
+
height_input = gr.Slider(
|
93 |
+
label="Height", minimum=256, maximum=2048, value=1024, step=16
|
94 |
+
)
|
95 |
+
width_input = gr.Slider(
|
96 |
+
label="Width", minimum=256, maximum=2048, value=1024, step=16
|
97 |
+
)
|
98 |
+
|
99 |
# 引导尺度输入
|
100 |
+
guidance_scale_input = gr.Slider(
|
101 |
+
label="Guidance Scale", minimum=1.0, maximum=10.0, value=3.0, step=0.1
|
102 |
+
)
|
103 |
+
|
104 |
+
num_inference_steps = gr.Slider(
|
105 |
+
label="Inference Steps", minimum=1, maximum=50, value=50, step=1
|
106 |
+
)
|
107 |
+
|
108 |
# 生成按钮
|
109 |
generate_button = gr.Button("Generate Image")
|
110 |
+
|
111 |
with gr.Column():
|
112 |
# 输出图像框
|
113 |
output_image = gr.Image(label="Output Image")
|
|
|
115 |
# 按钮点击事件
|
116 |
generate_button.click(
|
117 |
generate_image,
|
118 |
+
inputs=[
|
119 |
+
prompt_input,
|
120 |
+
image_input_1,
|
121 |
+
image_input_2,
|
122 |
+
image_input_3,
|
123 |
+
height_input,
|
124 |
+
width_input,
|
125 |
+
guidance_scale_input,
|
126 |
+
num_inference_steps,
|
127 |
+
],
|
128 |
+
outputs=output_image,
|
129 |
+
)
|
130 |
+
|
131 |
+
gr.Examples(
|
132 |
+
examples=get_example(),
|
133 |
+
fn=run_for_examples,
|
134 |
+
inputs=[
|
135 |
+
prompt_input,
|
136 |
+
image_input_1,
|
137 |
+
image_input_2,
|
138 |
+
image_input_3,
|
139 |
+
height_input,
|
140 |
+
width_input,
|
141 |
+
guidance_scale_input,
|
142 |
+
num_inference_steps,
|
143 |
+
],
|
144 |
+
outputs=output_image,
|
145 |
)
|
146 |
|
147 |
# 启动应用
|