Spaces:
Runtime error
Runtime error
Jordan Legg
commited on
Commit
β’
3ae9c83
1
Parent(s):
e514cac
fix trying to fix image preprocessing
Browse files- app.py +75 -44
- requirements.txt +2 -1
app.py
CHANGED
@@ -8,18 +8,32 @@ from torchvision import transforms
|
|
8 |
from diffusers import DiffusionPipeline
|
9 |
|
10 |
# Define constants
|
11 |
-
dtype = torch.bfloat16
|
12 |
-
device = "cuda" if torch.cuda.is_available() else "cpu"
|
13 |
MAX_SEED = np.iinfo(np.int32).max
|
14 |
MAX_IMAGE_SIZE = 2048
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
15 |
|
16 |
# Load the diffusion pipeline
|
17 |
-
pipe =
|
18 |
|
19 |
-
def preprocess_image(image):
|
20 |
# Preprocess the image for the VAE
|
21 |
preprocess = transforms.Compose([
|
22 |
-
transforms.Resize(
|
23 |
transforms.ToTensor(),
|
24 |
transforms.Normalize([0.5], [0.5])
|
25 |
])
|
@@ -32,44 +46,60 @@ def encode_image(image, vae):
|
|
32 |
latents = vae.encode(image).latent_dist.sample() * 0.18215
|
33 |
return latents
|
34 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
35 |
@spaces.GPU()
|
36 |
-
def infer(prompt, init_image=None, seed=42, randomize_seed=False, width=
|
37 |
-
|
38 |
-
|
39 |
-
|
40 |
-
|
41 |
-
|
42 |
-
|
43 |
-
|
44 |
-
init_image
|
45 |
-
|
46 |
-
|
47 |
-
|
48 |
-
|
49 |
-
|
50 |
-
|
51 |
-
|
52 |
-
|
53 |
-
|
54 |
-
|
55 |
-
|
56 |
-
|
57 |
-
|
58 |
-
|
59 |
-
|
60 |
-
|
61 |
-
|
62 |
-
|
63 |
-
|
64 |
-
|
65 |
-
|
66 |
-
|
|
|
|
|
|
|
|
|
67 |
|
68 |
# Define example prompts
|
69 |
examples = [
|
70 |
"a tiny astronaut hatching from an egg on the moon",
|
71 |
"a cat holding a sign that says hello world",
|
72 |
"an anime illustration of a wiener schnitzel",
|
|
|
|
|
73 |
]
|
74 |
|
75 |
# CSS styling for the Japanese-inspired interface
|
@@ -122,7 +152,7 @@ with gr.Blocks(css=css) as demo:
|
|
122 |
label="Prompt",
|
123 |
show_label=False,
|
124 |
max_lines=1,
|
125 |
-
placeholder="Enter your prompt",
|
126 |
container=False,
|
127 |
)
|
128 |
run_button = gr.Button("Run", scale=0)
|
@@ -144,17 +174,17 @@ with gr.Blocks(css=css) as demo:
|
|
144 |
with gr.Row():
|
145 |
width = gr.Slider(
|
146 |
label="Width",
|
147 |
-
minimum=
|
148 |
maximum=MAX_IMAGE_SIZE,
|
149 |
-
step=
|
150 |
-
value=
|
151 |
)
|
152 |
height = gr.Slider(
|
153 |
label="Height",
|
154 |
-
minimum=
|
155 |
maximum=MAX_IMAGE_SIZE,
|
156 |
-
step=
|
157 |
-
value=
|
158 |
)
|
159 |
|
160 |
with gr.Row():
|
@@ -181,7 +211,8 @@ with gr.Blocks(css=css) as demo:
|
|
181 |
outputs=[result, seed]
|
182 |
)
|
183 |
|
184 |
-
|
|
|
185 |
|
186 |
|
187 |
|
|
|
8 |
from diffusers import DiffusionPipeline
|
9 |
|
10 |
# Define constants
|
|
|
|
|
11 |
MAX_SEED = np.iinfo(np.int32).max
|
12 |
MAX_IMAGE_SIZE = 2048
|
13 |
+
MIN_IMAGE_SIZE = 256
|
14 |
+
DEFAULT_IMAGE_SIZE = 1024
|
15 |
+
MAX_PROMPT_LENGTH = 500
|
16 |
+
|
17 |
+
# Check for GPU availability
|
18 |
+
device = "cuda" if torch.cuda.is_available() else "cpu"
|
19 |
+
if device == "cpu":
|
20 |
+
print("Warning: Running on CPU. This may be very slow.")
|
21 |
+
|
22 |
+
dtype = torch.float16 if device == "cuda" else torch.float32
|
23 |
+
|
24 |
+
def load_model():
|
25 |
+
try:
|
26 |
+
return DiffusionPipeline.from_pretrained("black-forest-labs/FLUX.1-schnell", torch_dtype=dtype).to(device)
|
27 |
+
except Exception as e:
|
28 |
+
raise RuntimeError(f"Failed to load the model: {str(e)}")
|
29 |
|
30 |
# Load the diffusion pipeline
|
31 |
+
pipe = load_model()
|
32 |
|
33 |
+
def preprocess_image(image, target_size=(512, 512)):
|
34 |
# Preprocess the image for the VAE
|
35 |
preprocess = transforms.Compose([
|
36 |
+
transforms.Resize(target_size, interpolation=transforms.InterpolationMode.LANCZOS),
|
37 |
transforms.ToTensor(),
|
38 |
transforms.Normalize([0.5], [0.5])
|
39 |
])
|
|
|
46 |
latents = vae.encode(image).latent_dist.sample() * 0.18215
|
47 |
return latents
|
48 |
|
49 |
+
def validate_inputs(prompt, width, height, num_inference_steps):
|
50 |
+
if not prompt or len(prompt) > MAX_PROMPT_LENGTH:
|
51 |
+
raise ValueError(f"Prompt must be between 1 and {MAX_PROMPT_LENGTH} characters.")
|
52 |
+
if width % 8 != 0 or height % 8 != 0:
|
53 |
+
raise ValueError("Width and height must be divisible by 8.")
|
54 |
+
if width < MIN_IMAGE_SIZE or width > MAX_IMAGE_SIZE or height < MIN_IMAGE_SIZE or height > MAX_IMAGE_SIZE:
|
55 |
+
raise ValueError(f"Image dimensions must be between {MIN_IMAGE_SIZE} and {MAX_IMAGE_SIZE}.")
|
56 |
+
if num_inference_steps < 1 or num_inference_steps > 50:
|
57 |
+
raise ValueError("Number of inference steps must be between 1 and 50.")
|
58 |
+
|
59 |
@spaces.GPU()
|
60 |
+
def infer(prompt, init_image=None, seed=42, randomize_seed=False, width=DEFAULT_IMAGE_SIZE, height=DEFAULT_IMAGE_SIZE, num_inference_steps=4, progress=gr.Progress(track_tqdm=True)):
|
61 |
+
try:
|
62 |
+
validate_inputs(prompt, width, height, num_inference_steps)
|
63 |
+
|
64 |
+
if randomize_seed:
|
65 |
+
seed = random.randint(0, MAX_SEED)
|
66 |
+
generator = torch.Generator(device=device).manual_seed(seed)
|
67 |
+
|
68 |
+
if init_image is not None:
|
69 |
+
init_image = init_image.convert("RGB")
|
70 |
+
init_image = preprocess_image(init_image, (height, width))
|
71 |
+
latents = encode_image(init_image, pipe.vae)
|
72 |
+
latents = torch.nn.functional.interpolate(latents, size=(height // 8, width // 8), mode='bilinear')
|
73 |
+
image = pipe(
|
74 |
+
prompt=prompt,
|
75 |
+
height=height,
|
76 |
+
width=width,
|
77 |
+
num_inference_steps=num_inference_steps,
|
78 |
+
generator=generator,
|
79 |
+
guidance_scale=0.0,
|
80 |
+
latents=latents
|
81 |
+
).images[0]
|
82 |
+
else:
|
83 |
+
image = pipe(
|
84 |
+
prompt=prompt,
|
85 |
+
height=height,
|
86 |
+
width=width,
|
87 |
+
num_inference_steps=num_inference_steps,
|
88 |
+
generator=generator,
|
89 |
+
guidance_scale=0.0
|
90 |
+
).images[0]
|
91 |
+
|
92 |
+
return image, seed
|
93 |
+
except Exception as e:
|
94 |
+
raise gr.Error(str(e))
|
95 |
|
96 |
# Define example prompts
|
97 |
examples = [
|
98 |
"a tiny astronaut hatching from an egg on the moon",
|
99 |
"a cat holding a sign that says hello world",
|
100 |
"an anime illustration of a wiener schnitzel",
|
101 |
+
"a surreal landscape with floating islands and waterfalls",
|
102 |
+
"a steampunk-inspired cityscape at sunset"
|
103 |
]
|
104 |
|
105 |
# CSS styling for the Japanese-inspired interface
|
|
|
152 |
label="Prompt",
|
153 |
show_label=False,
|
154 |
max_lines=1,
|
155 |
+
placeholder=f"Enter your prompt (max {MAX_PROMPT_LENGTH} characters)",
|
156 |
container=False,
|
157 |
)
|
158 |
run_button = gr.Button("Run", scale=0)
|
|
|
174 |
with gr.Row():
|
175 |
width = gr.Slider(
|
176 |
label="Width",
|
177 |
+
minimum=MIN_IMAGE_SIZE,
|
178 |
maximum=MAX_IMAGE_SIZE,
|
179 |
+
step=8,
|
180 |
+
value=DEFAULT_IMAGE_SIZE,
|
181 |
)
|
182 |
height = gr.Slider(
|
183 |
label="Height",
|
184 |
+
minimum=MIN_IMAGE_SIZE,
|
185 |
maximum=MAX_IMAGE_SIZE,
|
186 |
+
step=8,
|
187 |
+
value=DEFAULT_IMAGE_SIZE,
|
188 |
)
|
189 |
|
190 |
with gr.Row():
|
|
|
211 |
outputs=[result, seed]
|
212 |
)
|
213 |
|
214 |
+
if __name__ == "__main__":
|
215 |
+
demo.launch()
|
216 |
|
217 |
|
218 |
|
requirements.txt
CHANGED
@@ -6,4 +6,5 @@ transformers==4.42.4
|
|
6 |
xformers
|
7 |
sentencepiece
|
8 |
gradio==4.29.0
|
9 |
-
torchvision
|
|
|
|
6 |
xformers
|
7 |
sentencepiece
|
8 |
gradio==4.29.0
|
9 |
+
torchvision
|
10 |
+
pillow
|