Commit
•
a076391
1
Parent(s):
53f3635
Update app.py
Browse files
app.py
CHANGED
@@ -6,24 +6,29 @@ import gradio as gr
|
|
6 |
import numpy as np
|
7 |
import PIL.Image
|
8 |
import torch
|
9 |
-
from diffusers import
|
10 |
|
11 |
MAX_SEED = np.iinfo(np.int32).max
|
12 |
MAX_IMAGE_SIZE = int(os.getenv('MAX_IMAGE_SIZE', '1024'))
|
13 |
SECRET_TOKEN = os.getenv('SECRET_TOKEN', 'default_secret')
|
14 |
|
15 |
-
MODEL_ID = "segmind/SSD-1B"
|
16 |
-
ADAPTER_ID = "latent-consistency/lcm-lora-ssd-1b"
|
17 |
-
|
18 |
device = torch.device('cuda:0' if torch.cuda.is_available() else 'cpu')
|
19 |
if torch.cuda.is_available():
|
20 |
-
|
21 |
-
|
22 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
23 |
|
24 |
-
|
25 |
-
pipe.
|
26 |
-
pipe.fuse_lora()
|
27 |
else:
|
28 |
pipe = None
|
29 |
|
@@ -39,8 +44,8 @@ def generate(prompt: str,
|
|
39 |
seed: int = 0,
|
40 |
width: int = 1024,
|
41 |
height: int = 1024,
|
42 |
-
guidance_scale: float =
|
43 |
-
num_inference_steps: int =
|
44 |
secret_token: str = '') -> PIL.Image.Image:
|
45 |
if secret_token != SECRET_TOKEN:
|
46 |
raise gr.Error(
|
@@ -64,7 +69,7 @@ with gr.Blocks() as demo:
|
|
64 |
gr.HTML("""
|
65 |
<div style="z-index: 100; position: fixed; top: 0px; right: 0px; left: 0px; bottom: 0px; width: 100%; height: 100%; background: white; display: flex; align-items: center; justify-content: center; color: black;">
|
66 |
<div style="text-align: center; color: black;">
|
67 |
-
<p style="color: black;">This space is a REST API to programmatically generate images using LCM
|
68 |
<p style="color: black;">It is not meant to be directly used through a user interface, but using code and an access key.</p>
|
69 |
</div>
|
70 |
</div>""")
|
@@ -112,16 +117,16 @@ with gr.Blocks() as demo:
|
|
112 |
)
|
113 |
guidance_scale = gr.Slider(
|
114 |
label='Guidance scale',
|
115 |
-
minimum=
|
116 |
-
maximum=
|
117 |
step=0.1,
|
118 |
-
value=
|
119 |
num_inference_steps = gr.Slider(
|
120 |
label='Number of inference steps',
|
121 |
-
minimum=
|
122 |
-
maximum=
|
123 |
step=1,
|
124 |
-
value=
|
125 |
|
126 |
use_negative_prompt.change(
|
127 |
fn=lambda x: gr.update(visible=x),
|
|
|
6 |
import numpy as np
|
7 |
import PIL.Image
|
8 |
import torch
|
9 |
+
from diffusers import DiffusionPipeline, UNet2DConditionModel, LCMScheduler
|
10 |
|
11 |
MAX_SEED = np.iinfo(np.int32).max
|
12 |
MAX_IMAGE_SIZE = int(os.getenv('MAX_IMAGE_SIZE', '1024'))
|
13 |
SECRET_TOKEN = os.getenv('SECRET_TOKEN', 'default_secret')
|
14 |
|
|
|
|
|
|
|
15 |
device = torch.device('cuda:0' if torch.cuda.is_available() else 'cpu')
|
16 |
if torch.cuda.is_available():
|
17 |
+
unet = UNet2DConditionModel.from_pretrained(
|
18 |
+
"latent-consistency/lcm-ssd-1b",
|
19 |
+
torch_dtype=torch.float16,
|
20 |
+
variant="fp16"
|
21 |
+
)
|
22 |
+
|
23 |
+
pipe = DiffusionPipeline.from_pretrained(
|
24 |
+
"segmind/SSD-1B",
|
25 |
+
unet=unet,
|
26 |
+
torch_dtype=torch.float16,
|
27 |
+
variant="fp16"
|
28 |
+
)
|
29 |
|
30 |
+
pipe.scheduler = LCMScheduler.from_config(pipe.scheduler.config)
|
31 |
+
pipe.to(device)
|
|
|
32 |
else:
|
33 |
pipe = None
|
34 |
|
|
|
44 |
seed: int = 0,
|
45 |
width: int = 1024,
|
46 |
height: int = 1024,
|
47 |
+
guidance_scale: float = 1.0,
|
48 |
+
num_inference_steps: int = 6,
|
49 |
secret_token: str = '') -> PIL.Image.Image:
|
50 |
if secret_token != SECRET_TOKEN:
|
51 |
raise gr.Error(
|
|
|
69 |
gr.HTML("""
|
70 |
<div style="z-index: 100; position: fixed; top: 0px; right: 0px; left: 0px; bottom: 0px; width: 100%; height: 100%; background: white; display: flex; align-items: center; justify-content: center; color: black;">
|
71 |
<div style="text-align: center; color: black;">
|
72 |
+
<p style="color: black;">This space is a REST API to programmatically generate images using LCM-SSD-1B.</p>
|
73 |
<p style="color: black;">It is not meant to be directly used through a user interface, but using code and an access key.</p>
|
74 |
</div>
|
75 |
</div>""")
|
|
|
117 |
)
|
118 |
guidance_scale = gr.Slider(
|
119 |
label='Guidance scale',
|
120 |
+
minimum=1,
|
121 |
+
maximum=20,
|
122 |
step=0.1,
|
123 |
+
value=1.0)
|
124 |
num_inference_steps = gr.Slider(
|
125 |
label='Number of inference steps',
|
126 |
+
minimum=2,
|
127 |
+
maximum=40,
|
128 |
step=1,
|
129 |
+
value=6)
|
130 |
|
131 |
use_negative_prompt.change(
|
132 |
fn=lambda x: gr.update(visible=x),
|