Spaces:
Paused
Paused
Update app.py
Browse files
app.py
CHANGED
@@ -10,7 +10,8 @@ from glob import glob
|
|
10 |
from pathlib import Path
|
11 |
from typing import Optional
|
12 |
|
13 |
-
from
|
|
|
14 |
from diffusers.utils import load_image, export_to_video
|
15 |
|
16 |
import uuid
|
@@ -20,9 +21,6 @@ from huggingface_hub import hf_hub_download
|
|
20 |
os.environ["HF_HUB_ENABLE_HF_TRANSFER"] = "1"
|
21 |
HF_TOKEN = os.environ.get("HF_TOKEN", None)
|
22 |
# Constants
|
23 |
-
base = "stabilityai/stable-video-diffusion-img2vid-xt"
|
24 |
-
model = "ECNU-CILab/ExVideo-SVD-128f-v1"
|
25 |
-
|
26 |
MAX_SEED = np.iinfo(np.int32).max
|
27 |
|
28 |
CSS = """
|
@@ -38,30 +36,15 @@ JS = """function () {
|
|
38 |
}
|
39 |
}"""
|
40 |
|
41 |
-
downloaded_model_path = hf_hub_download(
|
42 |
-
repo_id=model,
|
43 |
-
filename=model.fp16.safetensors,
|
44 |
-
local_dir="model"
|
45 |
-
)
|
46 |
-
|
47 |
-
MODEL_PATH = "./model/"
|
48 |
-
|
49 |
-
|
50 |
|
51 |
# Ensure model and scheduler are initialized in GPU-enabled function
|
52 |
if torch.cuda.is_available():
|
|
|
|
|
|
|
|
|
|
|
53 |
|
54 |
-
unet = UNetSpatioTemporalConditionControlNetModel.from_pretrained(
|
55 |
-
MODEL_PATH,
|
56 |
-
low_cpu_mem_usage=True,
|
57 |
-
variant="fp16",
|
58 |
-
)
|
59 |
-
|
60 |
-
pipe = StableVideoDiffusionPipeline.from_pretrained(
|
61 |
-
base,
|
62 |
-
unet=unet,
|
63 |
-
torch_dtype=torch.float16,
|
64 |
-
variant="fp16").to("cuda")
|
65 |
|
66 |
# function source codes modified from multimodalart/stable-video-diffusion
|
67 |
@spaces.GPU(duration=120)
|
@@ -69,11 +52,7 @@ def generate(
|
|
69 |
image: Image,
|
70 |
seed: Optional[int] = -1,
|
71 |
motion_bucket_id: int = 127,
|
72 |
-
fps_id: int =
|
73 |
-
version: str = "svd_xt",
|
74 |
-
cond_aug: float = 0.02,
|
75 |
-
decoding_t: int = 1,
|
76 |
-
device: str = "cuda",
|
77 |
output_folder: str = "outputs",
|
78 |
progress=gr.Progress(track_tqdm=True)):
|
79 |
|
@@ -83,49 +62,29 @@ def generate(
|
|
83 |
if image.mode == "RGBA":
|
84 |
image = image.convert("RGB")
|
85 |
|
86 |
-
|
87 |
|
88 |
os.makedirs(output_folder, exist_ok=True)
|
89 |
base_count = len(glob(os.path.join(output_folder, "*.mp4")))
|
90 |
video_path = os.path.join(output_folder, f"{base_count:06d}.mp4")
|
91 |
|
92 |
-
frames = pipe(
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
93 |
export_to_video(frames, video_path, fps=fps_id)
|
94 |
-
torch.manual_seed(seed)
|
95 |
|
96 |
return video_path, seed
|
97 |
|
98 |
-
def resize_image(image, output_size=(1024, 576)):
|
99 |
-
# Calculate aspect ratios
|
100 |
-
target_aspect = output_size[0] / output_size[1] # Aspect ratio of the desired size
|
101 |
-
image_aspect = image.width / image.height # Aspect ratio of the original image
|
102 |
-
|
103 |
-
# Resize then crop if the original image is larger
|
104 |
-
if image_aspect > target_aspect:
|
105 |
-
# Resize the image to match the target height, maintaining aspect ratio
|
106 |
-
new_height = output_size[1]
|
107 |
-
new_width = int(new_height * image_aspect)
|
108 |
-
resized_image = image.resize((new_width, new_height), Image.LANCZOS)
|
109 |
-
# Calculate coordinates for cropping
|
110 |
-
left = (new_width - output_size[0]) / 2
|
111 |
-
top = 0
|
112 |
-
right = (new_width + output_size[0]) / 2
|
113 |
-
bottom = output_size[1]
|
114 |
-
else:
|
115 |
-
# Resize the image to match the target width, maintaining aspect ratio
|
116 |
-
new_width = output_size[0]
|
117 |
-
new_height = int(new_width / image_aspect)
|
118 |
-
resized_image = image.resize((new_width, new_height), Image.LANCZOS)
|
119 |
-
# Calculate coordinates for cropping
|
120 |
-
left = 0
|
121 |
-
top = (new_height - output_size[1]) / 2
|
122 |
-
right = output_size[0]
|
123 |
-
bottom = (new_height + output_size[1]) / 2
|
124 |
-
|
125 |
-
# Crop the image
|
126 |
-
cropped_image = resized_image.crop((left, top, right, bottom))
|
127 |
-
return cropped_image
|
128 |
-
|
129 |
|
130 |
examples = [
|
131 |
"./train.jpg",
|
@@ -162,7 +121,7 @@ with gr.Blocks(css=CSS, js=JS, theme="soft") as demo:
|
|
162 |
fps_id = gr.Slider(
|
163 |
label="Frames per second",
|
164 |
info="The length of your video in seconds will be 25/fps",
|
165 |
-
value=
|
166 |
minimum=5,
|
167 |
maximum=30
|
168 |
)
|
@@ -178,8 +137,6 @@ with gr.Blocks(css=CSS, js=JS, theme="soft") as demo:
|
|
178 |
examples_per_page=4,
|
179 |
)
|
180 |
|
181 |
-
image.upload(fn=resize_image, inputs=image, outputs=image, queue=False)
|
182 |
-
|
183 |
generate_btn.click(fn=generate, inputs=[image, seed, motion_bucket_id, fps_id], outputs=[video, seed], api_name="video")
|
184 |
|
185 |
demo.queue().launch()
|
|
|
10 |
from pathlib import Path
|
11 |
from typing import Optional
|
12 |
|
13 |
+
from diffsynth import ModelManager, SVDVideoPipeline, HunyuanDiTImagePipeline
|
14 |
+
from diffsynth import ModelManager
|
15 |
from diffusers.utils import load_image, export_to_video
|
16 |
|
17 |
import uuid
|
|
|
21 |
os.environ["HF_HUB_ENABLE_HF_TRANSFER"] = "1"
|
22 |
HF_TOKEN = os.environ.get("HF_TOKEN", None)
|
23 |
# Constants
|
|
|
|
|
|
|
24 |
MAX_SEED = np.iinfo(np.int32).max
|
25 |
|
26 |
CSS = """
|
|
|
36 |
}
|
37 |
}"""
|
38 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
39 |
|
40 |
# Ensure model and scheduler are initialized in GPU-enabled function
|
41 |
if torch.cuda.is_available():
|
42 |
+
model_manager = ModelManager(
|
43 |
+
torch_dtype=torch.float16,
|
44 |
+
device="cuda",
|
45 |
+
model_id_list=["stable-video-diffusion-img2vid-xt", "ExVideo-SVD-128f-v1"])
|
46 |
+
pipe = SVDVideoPipeline.from_model_manager(model_manager)
|
47 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
48 |
|
49 |
# function source codes modified from multimodalart/stable-video-diffusion
|
50 |
@spaces.GPU(duration=120)
|
|
|
52 |
image: Image,
|
53 |
seed: Optional[int] = -1,
|
54 |
motion_bucket_id: int = 127,
|
55 |
+
fps_id: int = 25,
|
|
|
|
|
|
|
|
|
56 |
output_folder: str = "outputs",
|
57 |
progress=gr.Progress(track_tqdm=True)):
|
58 |
|
|
|
62 |
if image.mode == "RGBA":
|
63 |
image = image.convert("RGB")
|
64 |
|
65 |
+
torch.manual_seed(seed)
|
66 |
|
67 |
os.makedirs(output_folder, exist_ok=True)
|
68 |
base_count = len(glob(os.path.join(output_folder, "*.mp4")))
|
69 |
video_path = os.path.join(output_folder, f"{base_count:06d}.mp4")
|
70 |
|
71 |
+
frames = pipe(
|
72 |
+
input_image=image.resize((512, 512)),
|
73 |
+
num_frames=128,
|
74 |
+
fps=fps_id,
|
75 |
+
height=512,
|
76 |
+
width=512,
|
77 |
+
motion_bucket_id=motion_bucket_id,
|
78 |
+
num_inference_steps=50,
|
79 |
+
min_cfg_scale=2,
|
80 |
+
max_cfg_scale=2,
|
81 |
+
contrast_enhance_scale=1.2
|
82 |
+
).frames[0]
|
83 |
+
|
84 |
export_to_video(frames, video_path, fps=fps_id)
|
|
|
85 |
|
86 |
return video_path, seed
|
87 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
88 |
|
89 |
examples = [
|
90 |
"./train.jpg",
|
|
|
121 |
fps_id = gr.Slider(
|
122 |
label="Frames per second",
|
123 |
info="The length of your video in seconds will be 25/fps",
|
124 |
+
value=25,
|
125 |
minimum=5,
|
126 |
maximum=30
|
127 |
)
|
|
|
137 |
examples_per_page=4,
|
138 |
)
|
139 |
|
|
|
|
|
140 |
generate_btn.click(fn=generate, inputs=[image, seed, motion_bucket_id, fps_id], outputs=[video, seed], api_name="video")
|
141 |
|
142 |
demo.queue().launch()
|