Spaces:
Running
Running
Enabling Token Merging for fast inference
Browse files- app.py +4 -3
- app_canny.py +6 -1
- app_canny_db.py +6 -1
- app_pix2pix_video.py +9 -3
- app_pose.py +6 -1
- app_text_to_video.py +32 -9
- gradio_utils.py +7 -7
- model.py +30 -7
- requirements.txt +1 -0
- text_to_video_pipeline.py +15 -60
app.py
CHANGED
@@ -23,7 +23,7 @@ with gr.Blocks(css='style.css') as demo:
|
|
23 |
"""
|
24 |
<div style="text-align: center; max-width: 1200px; margin: 20px auto;">
|
25 |
<h1 style="font-weight: 900; font-size: 3rem; margin: 0rem">
|
26 |
-
Text2Video-Zero
|
27 |
</h1>
|
28 |
<h2 style="font-weight: 450; font-size: 1rem; margin: 0rem">
|
29 |
Levon Khachatryan<sup>1*</sup>, Andranik Movsisyan<sup>1*</sup>, Vahram Tadevosyan<sup>1*</sup>, Roberto Henschel<sup>1*</sup>, Zhangyang Wang<sup>1,2</sup>, Shant Navasardyan<sup>1</sup>
|
@@ -62,7 +62,8 @@ with gr.Blocks(css='style.css') as demo:
|
|
62 |
create_demo_canny(model)
|
63 |
with gr.Tab('Edge Conditional and Dreambooth Specialized'):
|
64 |
create_demo_canny_db(model)
|
65 |
-
|
|
|
66 |
gr.HTML(
|
67 |
"""
|
68 |
<div style="text-align: justify; max-width: 1200px; margin: 20px auto;">
|
@@ -90,5 +91,5 @@ if on_huggingspace:
|
|
90 |
demo.launch(debug=True)
|
91 |
else:
|
92 |
_, _, link = demo.queue(api_open=False).launch(
|
93 |
-
file_directories=['temporal'], share=args.public_access
|
94 |
print(link)
|
|
|
23 |
"""
|
24 |
<div style="text-align: center; max-width: 1200px; margin: 20px auto;">
|
25 |
<h1 style="font-weight: 900; font-size: 3rem; margin: 0rem">
|
26 |
+
<a href="https://github.com/Picsart-AI-Research/Text2Video-Zero" style="color:blue;">Text2Video-Zero</a>
|
27 |
</h1>
|
28 |
<h2 style="font-weight: 450; font-size: 1rem; margin: 0rem">
|
29 |
Levon Khachatryan<sup>1*</sup>, Andranik Movsisyan<sup>1*</sup>, Vahram Tadevosyan<sup>1*</sup>, Roberto Henschel<sup>1*</sup>, Zhangyang Wang<sup>1,2</sup>, Shant Navasardyan<sup>1</sup>
|
|
|
62 |
create_demo_canny(model)
|
63 |
with gr.Tab('Edge Conditional and Dreambooth Specialized'):
|
64 |
create_demo_canny_db(model)
|
65 |
+
'''
|
66 |
+
'''
|
67 |
gr.HTML(
|
68 |
"""
|
69 |
<div style="text-align: justify; max-width: 1200px; margin: 20px auto;">
|
|
|
91 |
demo.launch(debug=True)
|
92 |
else:
|
93 |
_, _, link = demo.queue(api_open=False).launch(
|
94 |
+
file_directories=['temporal'], share=args.public_access)
|
95 |
print(link)
|
app_canny.py
CHANGED
@@ -47,7 +47,11 @@ def create_demo(model: Model):
|
|
47 |
watermark = gr.Radio(["Picsart AI Research", "Text2Video-Zero",
|
48 |
"None"], label="Watermark", value='Picsart AI Research')
|
49 |
chunk_size = gr.Slider(
|
50 |
-
label="Chunk size", minimum=2, maximum=16, value=
|
|
|
|
|
|
|
|
|
51 |
with gr.Column():
|
52 |
result = gr.Video(label="Generated Video").style(height="auto")
|
53 |
|
@@ -56,6 +60,7 @@ def create_demo(model: Model):
|
|
56 |
prompt,
|
57 |
chunk_size,
|
58 |
watermark,
|
|
|
59 |
]
|
60 |
|
61 |
gr.Examples(examples=examples,
|
|
|
47 |
watermark = gr.Radio(["Picsart AI Research", "Text2Video-Zero",
|
48 |
"None"], label="Watermark", value='Picsart AI Research')
|
49 |
chunk_size = gr.Slider(
|
50 |
+
label="Chunk size", minimum=2, maximum=16, value=8, step=1, visible=not on_huggingspace,
|
51 |
+
info="Number of frames processed at once. Reduce for lower memory usage.")
|
52 |
+
merging_ratio = gr.Slider(
|
53 |
+
label="Merging ratio", minimum=0.0, maximum=0.9, step=0.1, value=0.0, visible=not on_huggingspace,
|
54 |
+
info="Ratio of how many tokens are merged. The higher the more compression (less memory and faster inference).")
|
55 |
with gr.Column():
|
56 |
result = gr.Video(label="Generated Video").style(height="auto")
|
57 |
|
|
|
60 |
prompt,
|
61 |
chunk_size,
|
62 |
watermark,
|
63 |
+
merging_ratio,
|
64 |
]
|
65 |
|
66 |
gr.Examples(examples=examples,
|
app_canny_db.py
CHANGED
@@ -51,7 +51,11 @@ def create_demo(model: Model):
|
|
51 |
watermark = gr.Radio(["Picsart AI Research", "Text2Video-Zero",
|
52 |
"None"], label="Watermark", value='Picsart AI Research')
|
53 |
chunk_size = gr.Slider(
|
54 |
-
label="Chunk size", minimum=2, maximum=16, value=
|
|
|
|
|
|
|
|
|
55 |
with gr.Column():
|
56 |
result = gr.Image(label="Generated Video").style(height=400)
|
57 |
|
@@ -79,6 +83,7 @@ def create_demo(model: Model):
|
|
79 |
prompt,
|
80 |
chunk_size,
|
81 |
watermark,
|
|
|
82 |
]
|
83 |
|
84 |
gr.Examples(examples=examples,
|
|
|
51 |
watermark = gr.Radio(["Picsart AI Research", "Text2Video-Zero",
|
52 |
"None"], label="Watermark", value='Picsart AI Research')
|
53 |
chunk_size = gr.Slider(
|
54 |
+
label="Chunk size", minimum=2, maximum=16, value=8, step=1, visible=not on_huggingspace,
|
55 |
+
info="Number of frames processed at once. Reduce for lower memory usage.")
|
56 |
+
merging_ratio = gr.Slider(
|
57 |
+
label="Merging ratio", minimum=0.0, maximum=0.9, step=0.1, value=0.0, visible=not on_huggingspace,
|
58 |
+
info="Ratio of how many tokens are merged. The higher the more compression (less memory and faster inference).")
|
59 |
with gr.Column():
|
60 |
result = gr.Image(label="Generated Video").style(height=400)
|
61 |
|
|
|
83 |
prompt,
|
84 |
chunk_size,
|
85 |
watermark,
|
86 |
+
merging_ratio,
|
87 |
]
|
88 |
|
89 |
gr.Examples(examples=examples,
|
app_pix2pix_video.py
CHANGED
@@ -48,9 +48,10 @@ def create_demo(model: Model):
|
|
48 |
value=512,
|
49 |
step=64)
|
50 |
seed = gr.Slider(label='Seed',
|
51 |
-
minimum
|
52 |
maximum=65536,
|
53 |
value=0,
|
|
|
54 |
step=1)
|
55 |
image_guidance = gr.Slider(label='Image guidance scale',
|
56 |
minimum=0.5,
|
@@ -73,7 +74,11 @@ def create_demo(model: Model):
|
|
73 |
value=-1,
|
74 |
step=1)
|
75 |
chunk_size = gr.Slider(
|
76 |
-
label="Chunk size", minimum=2, maximum=16, value=
|
|
|
|
|
|
|
|
|
77 |
with gr.Column():
|
78 |
result = gr.Video(label='Output', show_label=True)
|
79 |
inputs = [
|
@@ -86,7 +91,8 @@ def create_demo(model: Model):
|
|
86 |
end_t,
|
87 |
out_fps,
|
88 |
chunk_size,
|
89 |
-
watermark
|
|
|
90 |
]
|
91 |
|
92 |
gr.Examples(examples=examples,
|
|
|
48 |
value=512,
|
49 |
step=64)
|
50 |
seed = gr.Slider(label='Seed',
|
51 |
+
minimum=-1,
|
52 |
maximum=65536,
|
53 |
value=0,
|
54 |
+
info="-1 for random seed on each run. Otherwise the seed will be fixed",
|
55 |
step=1)
|
56 |
image_guidance = gr.Slider(label='Image guidance scale',
|
57 |
minimum=0.5,
|
|
|
74 |
value=-1,
|
75 |
step=1)
|
76 |
chunk_size = gr.Slider(
|
77 |
+
label="Chunk size", minimum=2, maximum=16, value=8, step=1, visible=not on_huggingspace,
|
78 |
+
info="Number of frames processed at once. Reduce for lower memory usage.")
|
79 |
+
merging_ratio = gr.Slider(
|
80 |
+
label="Merging ratio", minimum=0.0, maximum=0.9, step=0.1, value=0.0, visible=not on_huggingspace,
|
81 |
+
info="Ratio of how many tokens are merged. The higher the more compression (less memory and faster inference).")
|
82 |
with gr.Column():
|
83 |
result = gr.Video(label='Output', show_label=True)
|
84 |
inputs = [
|
|
|
91 |
end_t,
|
92 |
out_fps,
|
93 |
chunk_size,
|
94 |
+
watermark,
|
95 |
+
merging_ratio
|
96 |
]
|
97 |
|
98 |
gr.Examples(examples=examples,
|
app_pose.py
CHANGED
@@ -35,7 +35,11 @@ def create_demo(model: Model):
|
|
35 |
watermark = gr.Radio(["Picsart AI Research", "Text2Video-Zero",
|
36 |
"None"], label="Watermark", value='Picsart AI Research')
|
37 |
chunk_size = gr.Slider(
|
38 |
-
label="Chunk size", minimum=2, maximum=16, value=
|
|
|
|
|
|
|
|
|
39 |
with gr.Column():
|
40 |
result = gr.Image(label="Generated Video")
|
41 |
|
@@ -48,6 +52,7 @@ def create_demo(model: Model):
|
|
48 |
prompt,
|
49 |
chunk_size,
|
50 |
watermark,
|
|
|
51 |
]
|
52 |
|
53 |
gr.Examples(examples=examples,
|
|
|
35 |
watermark = gr.Radio(["Picsart AI Research", "Text2Video-Zero",
|
36 |
"None"], label="Watermark", value='Picsart AI Research')
|
37 |
chunk_size = gr.Slider(
|
38 |
+
label="Chunk size", minimum=2, maximum=16, value=8, step=1, visible=not on_huggingspace,
|
39 |
+
info="Number of frames processed at once. Reduce for lower memory usage.")
|
40 |
+
merging_ratio = gr.Slider(
|
41 |
+
label="Merging ratio", minimum=0.0, maximum=0.9, step=0.1, value=0.0, visible=not on_huggingspace,
|
42 |
+
info="Ratio of how many tokens are merged. The higher the more compression (less memory and faster inference).")
|
43 |
with gr.Column():
|
44 |
result = gr.Image(label="Generated Video")
|
45 |
|
|
|
52 |
prompt,
|
53 |
chunk_size,
|
54 |
watermark,
|
55 |
+
merging_ratio,
|
56 |
]
|
57 |
|
58 |
gr.Examples(examples=examples,
|
app_text_to_video.py
CHANGED
@@ -39,6 +39,7 @@ def create_demo(model: Model):
|
|
39 |
label="Model",
|
40 |
choices=get_model_list(),
|
41 |
value="dreamlike-art/dreamlike-photoreal-2.0",
|
|
|
42 |
)
|
43 |
prompt = gr.Textbox(label='Prompt')
|
44 |
run_button = gr.Button(label='Run')
|
@@ -52,21 +53,41 @@ def create_demo(model: Model):
|
|
52 |
else:
|
53 |
video_length = gr.Number(
|
54 |
label="Video length", value=8, precision=0)
|
55 |
-
|
56 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
57 |
|
58 |
motion_field_strength_x = gr.Slider(
|
59 |
-
label='Global Translation
|
|
|
|
|
60 |
motion_field_strength_y = gr.Slider(
|
61 |
-
label='Global Translation
|
|
|
|
|
62 |
|
63 |
t0 = gr.Slider(label="Timestep t0", minimum=0,
|
64 |
-
maximum=
|
65 |
-
|
66 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
67 |
|
68 |
-
n_prompt = gr.Textbox(
|
69 |
-
label="Optional Negative Prompt", value='')
|
70 |
with gr.Column():
|
71 |
result = gr.Video(label="Generated Video")
|
72 |
|
@@ -81,6 +102,8 @@ def create_demo(model: Model):
|
|
81 |
chunk_size,
|
82 |
video_length,
|
83 |
watermark,
|
|
|
|
|
84 |
]
|
85 |
|
86 |
gr.Examples(examples=examples,
|
|
|
39 |
label="Model",
|
40 |
choices=get_model_list(),
|
41 |
value="dreamlike-art/dreamlike-photoreal-2.0",
|
42 |
+
|
43 |
)
|
44 |
prompt = gr.Textbox(label='Prompt')
|
45 |
run_button = gr.Button(label='Run')
|
|
|
53 |
else:
|
54 |
video_length = gr.Number(
|
55 |
label="Video length", value=8, precision=0)
|
56 |
+
|
57 |
+
n_prompt = gr.Textbox(
|
58 |
+
label="Optional Negative Prompt", value='')
|
59 |
+
seed = gr.Slider(label='Seed',
|
60 |
+
info="-1 for random seed on each run. Otherwise, the seed will be fixed.",
|
61 |
+
minimum=-1,
|
62 |
+
maximum=65536,
|
63 |
+
value=0,
|
64 |
+
step=1)
|
65 |
|
66 |
motion_field_strength_x = gr.Slider(
|
67 |
+
label='Global Translation $\\delta_{x}$', minimum=-20, maximum=20,
|
68 |
+
value=12,
|
69 |
+
step=1)
|
70 |
motion_field_strength_y = gr.Slider(
|
71 |
+
label='Global Translation $\\delta_{y}$', minimum=-20, maximum=20,
|
72 |
+
value=12,
|
73 |
+
step=1)
|
74 |
|
75 |
t0 = gr.Slider(label="Timestep t0", minimum=0,
|
76 |
+
maximum=47, value=44, step=1,
|
77 |
+
info="Perform DDPM steps from t0 to t1. The larger the gap between t0 and t1, the more variance between the frames. Ensure t0 < t1 ",
|
78 |
+
)
|
79 |
+
t1 = gr.Slider(label="Timestep t1", minimum=1,
|
80 |
+
info="Perform DDPM steps from t0 to t1. The larger the gap between t0 and t1, the more variance between the frames. Ensure t0 < t1",
|
81 |
+
maximum=48, value=47, step=1)
|
82 |
+
chunk_size = gr.Slider(
|
83 |
+
label="Chunk size", minimum=2, maximum=16, value=8, step=1, visible=not on_huggingspace,
|
84 |
+
info="Number of frames processed at once. Reduce for lower memory usage."
|
85 |
+
)
|
86 |
+
merging_ratio = gr.Slider(
|
87 |
+
label="Merging ratio", minimum=0.0, maximum=0.9, step=0.1, value=0.0, visible=not on_huggingspace,
|
88 |
+
info="Ratio of how many tokens are merged. The higher the more compression (less memory and faster inference)."
|
89 |
+
)
|
90 |
|
|
|
|
|
91 |
with gr.Column():
|
92 |
result = gr.Video(label="Generated Video")
|
93 |
|
|
|
102 |
chunk_size,
|
103 |
video_length,
|
104 |
watermark,
|
105 |
+
merging_ratio,
|
106 |
+
seed,
|
107 |
]
|
108 |
|
109 |
gr.Examples(examples=examples,
|
gradio_utils.py
CHANGED
@@ -8,19 +8,19 @@ def edge_path_to_video_path(edge_path):
|
|
8 |
|
9 |
vid_name = edge_path.split("/")[-1]
|
10 |
if vid_name == "butterfly.mp4":
|
11 |
-
video_path = "__assets__/
|
12 |
elif vid_name == "deer.mp4":
|
13 |
-
video_path = "__assets__/
|
14 |
elif vid_name == "fox.mp4":
|
15 |
-
video_path = "__assets__/
|
16 |
elif vid_name == "girl_dancing.mp4":
|
17 |
-
video_path = "__assets__/
|
18 |
elif vid_name == "girl_turning.mp4":
|
19 |
-
video_path = "__assets__/
|
20 |
elif vid_name == "halloween.mp4":
|
21 |
-
video_path = "__assets__/
|
22 |
elif vid_name == "santa.mp4":
|
23 |
-
video_path = "__assets__/
|
24 |
|
25 |
assert os.path.isfile(video_path)
|
26 |
return video_path
|
|
|
8 |
|
9 |
vid_name = edge_path.split("/")[-1]
|
10 |
if vid_name == "butterfly.mp4":
|
11 |
+
video_path = "__assets__/canny_videos_mp4/butterfly.mp4"
|
12 |
elif vid_name == "deer.mp4":
|
13 |
+
video_path = "__assets__/canny_videos_mp4/deer.mp4"
|
14 |
elif vid_name == "fox.mp4":
|
15 |
+
video_path = "__assets__/canny_videos_mp4/fox.mp4"
|
16 |
elif vid_name == "girl_dancing.mp4":
|
17 |
+
video_path = "__assets__/canny_videos_mp4/girl_dancing.mp4"
|
18 |
elif vid_name == "girl_turning.mp4":
|
19 |
+
video_path = "__assets__/canny_videos_mp4/girl_turning.mp4"
|
20 |
elif vid_name == "halloween.mp4":
|
21 |
+
video_path = "__assets__/canny_videos_mp4/halloween.mp4"
|
22 |
elif vid_name == "santa.mp4":
|
23 |
+
video_path = "__assets__/canny_videos_mp4/santa.mp4"
|
24 |
|
25 |
assert os.path.isfile(video_path)
|
26 |
return video_path
|
model.py
CHANGED
@@ -1,7 +1,7 @@
|
|
1 |
from enum import Enum
|
2 |
import gc
|
3 |
import numpy as np
|
4 |
-
|
5 |
import torch
|
6 |
|
7 |
from diffusers import StableDiffusionInstructPix2PixPipeline, StableDiffusionControlNetPipeline, ControlNetModel, UNet2DConditionModel
|
@@ -45,6 +45,7 @@ class Model:
|
|
45 |
self.model_type = None
|
46 |
|
47 |
self.states = {}
|
|
|
48 |
|
49 |
def set_model(self, model_type: ModelType, model_id: str, **kwargs):
|
50 |
if self.pipe is not None:
|
@@ -55,6 +56,7 @@ class Model:
|
|
55 |
self.pipe = self.pipe_dict[model_type].from_pretrained(
|
56 |
model_id, safety_checker=safety_checker, **kwargs).to(self.device).to(self.dtype)
|
57 |
self.model_type = model_type
|
|
|
58 |
|
59 |
def inference_chunk(self, frame_ids, **kwargs):
|
60 |
if self.pipe is None:
|
@@ -80,6 +82,13 @@ class Model:
|
|
80 |
def inference(self, split_to_chunks=False, chunk_size=8, **kwargs):
|
81 |
if self.pipe is None:
|
82 |
return
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
83 |
seed = kwargs.pop('seed', 0)
|
84 |
if seed < 0:
|
85 |
seed = self.generator.seed()
|
@@ -116,6 +125,7 @@ class Model:
|
|
116 |
result = np.concatenate(result)
|
117 |
return result
|
118 |
else:
|
|
|
119 |
return self.pipe(prompt=prompt, negative_prompt=negative_prompt, generator=self.generator, **kwargs).images
|
120 |
|
121 |
def process_controlnet_canny(self,
|
@@ -123,6 +133,7 @@ class Model:
|
|
123 |
prompt,
|
124 |
chunk_size=8,
|
125 |
watermark='Picsart AI Research',
|
|
|
126 |
num_inference_steps=20,
|
127 |
controlnet_conditioning_scale=1.0,
|
128 |
guidance_scale=9.0,
|
@@ -133,6 +144,7 @@ class Model:
|
|
133 |
resolution=512,
|
134 |
use_cf_attn=True,
|
135 |
save_path=None):
|
|
|
136 |
video_path = gradio_utils.edge_path_to_video_path(video_path)
|
137 |
if self.model_type != ModelType.ControlNetCanny:
|
138 |
controlnet = ControlNetModel.from_pretrained(
|
@@ -173,6 +185,7 @@ class Model:
|
|
173 |
output_type='numpy',
|
174 |
split_to_chunks=True,
|
175 |
chunk_size=chunk_size,
|
|
|
176 |
)
|
177 |
return utils.create_video(result, fps, path=save_path, watermark=gradio_utils.logo_name_to_path(watermark))
|
178 |
|
@@ -181,6 +194,7 @@ class Model:
|
|
181 |
prompt,
|
182 |
chunk_size=8,
|
183 |
watermark='Picsart AI Research',
|
|
|
184 |
num_inference_steps=20,
|
185 |
controlnet_conditioning_scale=1.0,
|
186 |
guidance_scale=9.0,
|
@@ -189,6 +203,7 @@ class Model:
|
|
189 |
resolution=512,
|
190 |
use_cf_attn=True,
|
191 |
save_path=None):
|
|
|
192 |
video_path = gradio_utils.motion_to_video_path(video_path)
|
193 |
if self.model_type != ModelType.ControlNetPose:
|
194 |
controlnet = ControlNetModel.from_pretrained(
|
@@ -232,6 +247,7 @@ class Model:
|
|
232 |
output_type='numpy',
|
233 |
split_to_chunks=True,
|
234 |
chunk_size=chunk_size,
|
|
|
235 |
)
|
236 |
return utils.create_gif(result, fps, path=save_path, watermark=gradio_utils.logo_name_to_path(watermark))
|
237 |
|
@@ -241,6 +257,7 @@ class Model:
|
|
241 |
prompt,
|
242 |
chunk_size=8,
|
243 |
watermark='Picsart AI Research',
|
|
|
244 |
num_inference_steps=20,
|
245 |
controlnet_conditioning_scale=1.0,
|
246 |
guidance_scale=9.0,
|
@@ -251,6 +268,7 @@ class Model:
|
|
251 |
resolution=512,
|
252 |
use_cf_attn=True,
|
253 |
save_path=None):
|
|
|
254 |
db_path = gradio_utils.get_model_from_db_selection(db_path)
|
255 |
video_path = gradio_utils.get_video_from_canny_selection(video_path)
|
256 |
# Load db and controlnet weights
|
@@ -295,6 +313,7 @@ class Model:
|
|
295 |
output_type='numpy',
|
296 |
split_to_chunks=True,
|
297 |
chunk_size=chunk_size,
|
|
|
298 |
)
|
299 |
return utils.create_gif(result, fps, path=save_path, watermark=gradio_utils.logo_name_to_path(watermark))
|
300 |
|
@@ -309,8 +328,10 @@ class Model:
|
|
309 |
out_fps=-1,
|
310 |
chunk_size=8,
|
311 |
watermark='Picsart AI Research',
|
|
|
312 |
use_cf_attn=True,
|
313 |
save_path=None,):
|
|
|
314 |
if self.model_type != ModelType.Pix2Pix_Video:
|
315 |
self.set_model(ModelType.Pix2Pix_Video,
|
316 |
model_id="timbrooks/instruct-pix2pix")
|
@@ -330,6 +351,7 @@ class Model:
|
|
330 |
image_guidance_scale=image_guidance_scale,
|
331 |
split_to_chunks=True,
|
332 |
chunk_size=chunk_size,
|
|
|
333 |
)
|
334 |
return utils.create_video(result, fps, path=save_path, watermark=gradio_utils.logo_name_to_path(watermark))
|
335 |
|
@@ -344,17 +366,18 @@ class Model:
|
|
344 |
chunk_size=8,
|
345 |
video_length=8,
|
346 |
watermark='Picsart AI Research',
|
347 |
-
|
|
|
348 |
resolution=512,
|
349 |
-
seed=-1,
|
350 |
fps=2,
|
351 |
use_cf_attn=True,
|
352 |
use_motion_field=True,
|
353 |
smooth_bg=False,
|
354 |
smooth_bg_strength=0.4,
|
355 |
path=None):
|
356 |
-
|
357 |
-
if self.model_type != ModelType.Text2Video:
|
|
|
358 |
unet = UNet2DConditionModel.from_pretrained(
|
359 |
model_name, subfolder="unet")
|
360 |
self.set_model(ModelType.Text2Video,
|
@@ -364,7 +387,7 @@ class Model:
|
|
364 |
if use_cf_attn:
|
365 |
self.pipe.unet.set_attn_processor(
|
366 |
processor=self.text2video_attn_proc)
|
367 |
-
|
368 |
|
369 |
added_prompt = "high quality, HD, 8K, trending on artstation, high focus, dramatic lighting"
|
370 |
negative_prompts = 'longbody, lowres, bad anatomy, bad hands, missing fingers, extra digit, fewer difits, cropped, worst quality, low quality, deformed body, bloated, ugly, unrealistic'
|
@@ -396,7 +419,7 @@ class Model:
|
|
396 |
seed=seed,
|
397 |
output_type='numpy',
|
398 |
negative_prompt=negative_prompt,
|
399 |
-
|
400 |
split_to_chunks=True,
|
401 |
chunk_size=chunk_size,
|
402 |
)
|
|
|
1 |
from enum import Enum
|
2 |
import gc
|
3 |
import numpy as np
|
4 |
+
import tomesd
|
5 |
import torch
|
6 |
|
7 |
from diffusers import StableDiffusionInstructPix2PixPipeline, StableDiffusionControlNetPipeline, ControlNetModel, UNet2DConditionModel
|
|
|
45 |
self.model_type = None
|
46 |
|
47 |
self.states = {}
|
48 |
+
self.model_name = ""
|
49 |
|
50 |
def set_model(self, model_type: ModelType, model_id: str, **kwargs):
|
51 |
if self.pipe is not None:
|
|
|
56 |
self.pipe = self.pipe_dict[model_type].from_pretrained(
|
57 |
model_id, safety_checker=safety_checker, **kwargs).to(self.device).to(self.dtype)
|
58 |
self.model_type = model_type
|
59 |
+
self.model_name = model_id
|
60 |
|
61 |
def inference_chunk(self, frame_ids, **kwargs):
|
62 |
if self.pipe is None:
|
|
|
82 |
def inference(self, split_to_chunks=False, chunk_size=8, **kwargs):
|
83 |
if self.pipe is None:
|
84 |
return
|
85 |
+
tomesd.remove_patch(self.pipe)
|
86 |
+
if "merging_ratio" in kwargs:
|
87 |
+
merging_ratio = kwargs.pop("merging_ratio")
|
88 |
+
|
89 |
+
if merging_ratio > 0:
|
90 |
+
|
91 |
+
tomesd.apply_patch(self.pipe, ratio=merging_ratio)
|
92 |
seed = kwargs.pop('seed', 0)
|
93 |
if seed < 0:
|
94 |
seed = self.generator.seed()
|
|
|
125 |
result = np.concatenate(result)
|
126 |
return result
|
127 |
else:
|
128 |
+
self.generator.manual_seed(seed)
|
129 |
return self.pipe(prompt=prompt, negative_prompt=negative_prompt, generator=self.generator, **kwargs).images
|
130 |
|
131 |
def process_controlnet_canny(self,
|
|
|
133 |
prompt,
|
134 |
chunk_size=8,
|
135 |
watermark='Picsart AI Research',
|
136 |
+
merging_ratio=0.0,
|
137 |
num_inference_steps=20,
|
138 |
controlnet_conditioning_scale=1.0,
|
139 |
guidance_scale=9.0,
|
|
|
144 |
resolution=512,
|
145 |
use_cf_attn=True,
|
146 |
save_path=None):
|
147 |
+
print("Processing Canny")
|
148 |
video_path = gradio_utils.edge_path_to_video_path(video_path)
|
149 |
if self.model_type != ModelType.ControlNetCanny:
|
150 |
controlnet = ControlNetModel.from_pretrained(
|
|
|
185 |
output_type='numpy',
|
186 |
split_to_chunks=True,
|
187 |
chunk_size=chunk_size,
|
188 |
+
merging_ratio=merging_ratio,
|
189 |
)
|
190 |
return utils.create_video(result, fps, path=save_path, watermark=gradio_utils.logo_name_to_path(watermark))
|
191 |
|
|
|
194 |
prompt,
|
195 |
chunk_size=8,
|
196 |
watermark='Picsart AI Research',
|
197 |
+
merging_ratio=0.0,
|
198 |
num_inference_steps=20,
|
199 |
controlnet_conditioning_scale=1.0,
|
200 |
guidance_scale=9.0,
|
|
|
203 |
resolution=512,
|
204 |
use_cf_attn=True,
|
205 |
save_path=None):
|
206 |
+
print("Processing Pose")
|
207 |
video_path = gradio_utils.motion_to_video_path(video_path)
|
208 |
if self.model_type != ModelType.ControlNetPose:
|
209 |
controlnet = ControlNetModel.from_pretrained(
|
|
|
247 |
output_type='numpy',
|
248 |
split_to_chunks=True,
|
249 |
chunk_size=chunk_size,
|
250 |
+
merging_ratio=merging_ratio,
|
251 |
)
|
252 |
return utils.create_gif(result, fps, path=save_path, watermark=gradio_utils.logo_name_to_path(watermark))
|
253 |
|
|
|
257 |
prompt,
|
258 |
chunk_size=8,
|
259 |
watermark='Picsart AI Research',
|
260 |
+
merging_ratio=0.0,
|
261 |
num_inference_steps=20,
|
262 |
controlnet_conditioning_scale=1.0,
|
263 |
guidance_scale=9.0,
|
|
|
268 |
resolution=512,
|
269 |
use_cf_attn=True,
|
270 |
save_path=None):
|
271 |
+
print("Processing Canny_DB")
|
272 |
db_path = gradio_utils.get_model_from_db_selection(db_path)
|
273 |
video_path = gradio_utils.get_video_from_canny_selection(video_path)
|
274 |
# Load db and controlnet weights
|
|
|
313 |
output_type='numpy',
|
314 |
split_to_chunks=True,
|
315 |
chunk_size=chunk_size,
|
316 |
+
merging_ratio=merging_ratio,
|
317 |
)
|
318 |
return utils.create_gif(result, fps, path=save_path, watermark=gradio_utils.logo_name_to_path(watermark))
|
319 |
|
|
|
328 |
out_fps=-1,
|
329 |
chunk_size=8,
|
330 |
watermark='Picsart AI Research',
|
331 |
+
merging_ratio=0.0,
|
332 |
use_cf_attn=True,
|
333 |
save_path=None,):
|
334 |
+
print("Processing Pix2Pix")
|
335 |
if self.model_type != ModelType.Pix2Pix_Video:
|
336 |
self.set_model(ModelType.Pix2Pix_Video,
|
337 |
model_id="timbrooks/instruct-pix2pix")
|
|
|
351 |
image_guidance_scale=image_guidance_scale,
|
352 |
split_to_chunks=True,
|
353 |
chunk_size=chunk_size,
|
354 |
+
merging_ratio=merging_ratio
|
355 |
)
|
356 |
return utils.create_video(result, fps, path=save_path, watermark=gradio_utils.logo_name_to_path(watermark))
|
357 |
|
|
|
366 |
chunk_size=8,
|
367 |
video_length=8,
|
368 |
watermark='Picsart AI Research',
|
369 |
+
merging_ratio=0.0,
|
370 |
+
seed=0,
|
371 |
resolution=512,
|
|
|
372 |
fps=2,
|
373 |
use_cf_attn=True,
|
374 |
use_motion_field=True,
|
375 |
smooth_bg=False,
|
376 |
smooth_bg_strength=0.4,
|
377 |
path=None):
|
378 |
+
print("Processing Text2Video")
|
379 |
+
if self.model_type != ModelType.Text2Video or model_name != self.model_name:
|
380 |
+
print("Model update")
|
381 |
unet = UNet2DConditionModel.from_pretrained(
|
382 |
model_name, subfolder="unet")
|
383 |
self.set_model(ModelType.Text2Video,
|
|
|
387 |
if use_cf_attn:
|
388 |
self.pipe.unet.set_attn_processor(
|
389 |
processor=self.text2video_attn_proc)
|
390 |
+
self.generator.manual_seed(seed)
|
391 |
|
392 |
added_prompt = "high quality, HD, 8K, trending on artstation, high focus, dramatic lighting"
|
393 |
negative_prompts = 'longbody, lowres, bad anatomy, bad hands, missing fingers, extra digit, fewer difits, cropped, worst quality, low quality, deformed body, bloated, ugly, unrealistic'
|
|
|
419 |
seed=seed,
|
420 |
output_type='numpy',
|
421 |
negative_prompt=negative_prompt,
|
422 |
+
merging_ratio=merging_ratio,
|
423 |
split_to_chunks=True,
|
424 |
chunk_size=chunk_size,
|
425 |
)
|
requirements.txt
CHANGED
@@ -34,3 +34,4 @@ yapf==0.32.0
|
|
34 |
safetensors==0.2.7
|
35 |
beautifulsoup4
|
36 |
bs4
|
|
|
|
34 |
safetensors==0.2.7
|
35 |
beautifulsoup4
|
36 |
bs4
|
37 |
+
tomesd
|
text_to_video_pipeline.py
CHANGED
@@ -53,8 +53,10 @@ class TextToVideoPipeline(StableDiffusionPipeline):
|
|
53 |
if x0 is None:
|
54 |
return torch.randn(shape, generator=generator, device=rand_device, dtype=text_embeddings.dtype).to(device)
|
55 |
else:
|
56 |
-
eps = torch.
|
|
|
57 |
alpha_vec = torch.prod(self.scheduler.alphas[t0:tMax])
|
|
|
58 |
xt = torch.sqrt(alpha_vec) * x0 + \
|
59 |
torch.sqrt(1-alpha_vec) * eps
|
60 |
return xt
|
@@ -89,7 +91,7 @@ class TextToVideoPipeline(StableDiffusionPipeline):
|
|
89 |
latents = latents * self.scheduler.init_noise_sigma
|
90 |
return latents
|
91 |
|
92 |
-
def warp_latents_independently(self, latents, reference_flow
|
93 |
_, _, H, W = reference_flow.size()
|
94 |
b, _, f, h, w = latents.size()
|
95 |
assert b == 1
|
@@ -109,15 +111,6 @@ class TextToVideoPipeline(StableDiffusionPipeline):
|
|
109 |
warped = grid_sample(latents_0, coords_t0,
|
110 |
mode='nearest', padding_mode='reflection')
|
111 |
|
112 |
-
if inject_noise:
|
113 |
-
idx = torch.logical_or(coords_t0 >= 1, coords_t0 < -1)
|
114 |
-
reset_noise = torch.randn(idx.shape)
|
115 |
-
idx = torch.logical_or(idx[:, :, :, 0], idx[:, :, :, 1])
|
116 |
-
idx = repeat(idx, "f w h -> f c w h", c=warped.shape[1])
|
117 |
-
reset_noise = torch.randn(
|
118 |
-
size=warped.shape, dtype=warped.dtype, device=warped.device)
|
119 |
-
warped[idx] = reset_noise[idx]
|
120 |
-
|
121 |
warped = rearrange(warped, '(b f) c h w -> b c f h w', f=f)
|
122 |
return warped
|
123 |
|
@@ -212,20 +205,20 @@ class TextToVideoPipeline(StableDiffusionPipeline):
|
|
212 |
|
213 |
reference_flow = torch.zeros(
|
214 |
(video_length-1, 2, 512, 512), device=latents.device, dtype=latents.dtype)
|
215 |
-
for fr_idx in
|
216 |
reference_flow[fr_idx, 0, :,
|
217 |
-
:] = motion_field_strength_x*(
|
218 |
reference_flow[fr_idx, 1, :,
|
219 |
-
:] = motion_field_strength_y*(
|
220 |
return reference_flow
|
221 |
|
222 |
-
def create_motion_field_and_warp_latents(self, motion_field_strength_x, motion_field_strength_y, frame_ids, video_length,
|
223 |
|
224 |
motion_field = self.create_motion_field(motion_field_strength_x=motion_field_strength_x,
|
225 |
motion_field_strength_y=motion_field_strength_y, latents=latents, video_length=video_length, frame_ids=frame_ids)
|
226 |
for idx, latent in enumerate(latents):
|
227 |
latents[idx] = self.warp_latents_independently(
|
228 |
-
latent[None], motion_field
|
229 |
return motion_field, latents
|
230 |
|
231 |
@torch.no_grad()
|
@@ -255,13 +248,12 @@ class TextToVideoPipeline(StableDiffusionPipeline):
|
|
255 |
use_motion_field: bool = True,
|
256 |
smooth_bg: bool = False,
|
257 |
smooth_bg_strength: float = 0.4,
|
258 |
-
inject_noise_to_warp: bool = False,
|
259 |
t0: int = 44,
|
260 |
t1: int = 47,
|
261 |
**kwargs,
|
262 |
):
|
263 |
frame_ids = kwargs.pop("frame_ids", list(range(video_length)))
|
264 |
-
|
265 |
assert num_videos_per_prompt == 1
|
266 |
assert isinstance(prompt, list) and len(prompt) > 0
|
267 |
assert isinstance(negative_prompt, list) or negative_prompt is None
|
@@ -280,11 +272,6 @@ class TextToVideoPipeline(StableDiffusionPipeline):
|
|
280 |
prompt = prompt_types[0]
|
281 |
negative_prompt = prompt_types[1]
|
282 |
|
283 |
-
print(
|
284 |
-
f" Motion field strength x = {motion_field_strength_x}, y = {motion_field_strength_y}")
|
285 |
-
print(f" Use: Motion field = {use_motion_field}")
|
286 |
-
print(f" Use: Background smoothing = {smooth_bg}")
|
287 |
-
print(f"Inject noise to warp = {inject_noise_to_warp}")
|
288 |
# Default height and width to unet
|
289 |
height = height or self.unet.config.sample_size * self.vae_scale_factor
|
290 |
width = width or self.unet.config.sample_size * self.vae_scale_factor
|
@@ -355,6 +342,7 @@ class TextToVideoPipeline(StableDiffusionPipeline):
|
|
355 |
|
356 |
t0 = timesteps_ddpm[t0]
|
357 |
t1 = timesteps_ddpm[t1]
|
|
|
358 |
print(f"t0 = {t0} t1 = {t1}")
|
359 |
x_t1_1 = None
|
360 |
|
@@ -366,14 +354,6 @@ class TextToVideoPipeline(StableDiffusionPipeline):
|
|
366 |
|
367 |
shape = (batch_size, num_channels_latents, 1, height //
|
368 |
self.vae_scale_factor, width // self.vae_scale_factor)
|
369 |
-
if inject_noise_to_warp and use_motion_field:
|
370 |
-
# if we inject to noise to warp function, we do it for timesteps T = 1000
|
371 |
-
|
372 |
-
x_t0_k = xT[:, :, :1, :, :].repeat(1, 1, video_length-1, 1, 1)
|
373 |
-
|
374 |
-
# reference_flow, x_t0_k = self.create_motion_field_and_warp_latents(motion_field_strength_x=motion_field_strength_x, motion_field_strength_y=motion_field_strength_y,
|
375 |
-
# frame_ids=frame_ids,video_length=video_length,inject_noise_to_warp=inject_noise_to_warp,latents = x_t0_k)
|
376 |
-
# xT =torch.cat([xT, x_t0_k], dim=2).clone().detach()
|
377 |
|
378 |
ddim_res = self.DDIM_backward(num_inference_steps=num_inference_steps, timesteps=timesteps, skip_t=1000, t0=t0, t1=t1, do_classifier_free_guidance=do_classifier_free_guidance,
|
379 |
null_embs=null_embs, text_embeddings=text_embeddings, latents_local=xT, latents_dtype=dtype, guidance_scale=guidance_scale, guidance_stop_step=guidance_stop_step,
|
@@ -387,37 +367,13 @@ class TextToVideoPipeline(StableDiffusionPipeline):
|
|
387 |
x_t1_1 = ddim_res["x_t1_1"].detach()
|
388 |
del ddim_res
|
389 |
del xT
|
390 |
-
|
391 |
-
if inject_noise_to_warp and use_motion_field:
|
392 |
-
# DDPM forward to allow for more motion
|
393 |
-
if t1 > t0:
|
394 |
-
x_t1_k = self.DDPM_forward(
|
395 |
-
x0=x_t0_1, t0=t0, tMax=t1, device=device, shape=shape, text_embeddings=text_embeddings, generator=generator)
|
396 |
-
else:
|
397 |
-
x_t1_k = x_t0_k
|
398 |
-
|
399 |
-
if x_t1_1 is None:
|
400 |
-
raise Exception
|
401 |
-
|
402 |
-
x_t1 = x_t1_k.clone().detach()
|
403 |
-
|
404 |
-
ddim_res = self.DDIM_backward(num_inference_steps=num_inference_steps, timesteps=timesteps, skip_t=t1, t0=-1, t1=-1, do_classifier_free_guidance=do_classifier_free_guidance,
|
405 |
-
null_embs=null_embs, text_embeddings=text_embeddings, latents_local=x_t1, latents_dtype=dtype, guidance_scale=guidance_scale, guidance_stop_step=guidance_stop_step,
|
406 |
-
callback=callback, callback_steps=callback_steps, extra_step_kwargs=extra_step_kwargs, num_warmup_steps=num_warmup_steps)
|
407 |
-
|
408 |
-
x0 = ddim_res["x0"].detach()
|
409 |
-
del ddim_res
|
410 |
-
del x_t1
|
411 |
-
del x_t1_k
|
412 |
-
|
413 |
-
if use_motion_field and not inject_noise_to_warp:
|
414 |
del x0
|
415 |
|
416 |
x_t0_k = x_t0_1[:, :, :1, :, :].repeat(1, 1, video_length-1, 1, 1)
|
417 |
|
418 |
reference_flow, x_t0_k = self.create_motion_field_and_warp_latents(
|
419 |
-
motion_field_strength_x=motion_field_strength_x, motion_field_strength_y=motion_field_strength_y, latents=x_t0_k, video_length=video_length,
|
420 |
-
inject_noise_to_warp=inject_noise_to_warp, frame_ids=frame_ids)
|
421 |
|
422 |
# assuming t0=t1=1000, if t0 = 1000
|
423 |
if t1 > t0:
|
@@ -440,7 +396,6 @@ class TextToVideoPipeline(StableDiffusionPipeline):
|
|
440 |
del x_t1
|
441 |
del x_t1_1
|
442 |
del x_t1_k
|
443 |
-
|
444 |
else:
|
445 |
x_t1 = x_t1_1.clone()
|
446 |
x_t1_1 = x_t1_1[:, :, :1, :, :].clone()
|
@@ -481,7 +436,7 @@ class TextToVideoPipeline(StableDiffusionPipeline):
|
|
481 |
if use_motion_field:
|
482 |
x_t1_fg_masked_b = x_t1_fg_masked_b[None]
|
483 |
x_t1_fg_masked_b = self.warp_latents_independently(
|
484 |
-
x_t1_fg_masked_b, reference_flow
|
485 |
else:
|
486 |
x_t1_fg_masked_b = x_t1_fg_masked_b[None]
|
487 |
|
@@ -499,7 +454,7 @@ class TextToVideoPipeline(StableDiffusionPipeline):
|
|
499 |
m_fg_b = m_fg_1_b.repeat(1, 1, video_length-1, 1, 1)
|
500 |
if use_motion_field:
|
501 |
m_fg_b = self.warp_latents_independently(
|
502 |
-
m_fg_b.clone(), reference_flow
|
503 |
M_FG_warped.append(
|
504 |
torch.cat([m_fg_1_b[:1, 0], m_fg_b[:1, 0]], dim=1))
|
505 |
|
|
|
53 |
if x0 is None:
|
54 |
return torch.randn(shape, generator=generator, device=rand_device, dtype=text_embeddings.dtype).to(device)
|
55 |
else:
|
56 |
+
eps = torch.randn(x0.shape, dtype=text_embeddings.dtype, generator=generator,
|
57 |
+
device=rand_device)
|
58 |
alpha_vec = torch.prod(self.scheduler.alphas[t0:tMax])
|
59 |
+
|
60 |
xt = torch.sqrt(alpha_vec) * x0 + \
|
61 |
torch.sqrt(1-alpha_vec) * eps
|
62 |
return xt
|
|
|
91 |
latents = latents * self.scheduler.init_noise_sigma
|
92 |
return latents
|
93 |
|
94 |
+
def warp_latents_independently(self, latents, reference_flow):
|
95 |
_, _, H, W = reference_flow.size()
|
96 |
b, _, f, h, w = latents.size()
|
97 |
assert b == 1
|
|
|
111 |
warped = grid_sample(latents_0, coords_t0,
|
112 |
mode='nearest', padding_mode='reflection')
|
113 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
114 |
warped = rearrange(warped, '(b f) c h w -> b c f h w', f=f)
|
115 |
return warped
|
116 |
|
|
|
205 |
|
206 |
reference_flow = torch.zeros(
|
207 |
(video_length-1, 2, 512, 512), device=latents.device, dtype=latents.dtype)
|
208 |
+
for fr_idx, frame_id in enumerate(frame_ids):
|
209 |
reference_flow[fr_idx, 0, :,
|
210 |
+
:] = motion_field_strength_x*(frame_id)
|
211 |
reference_flow[fr_idx, 1, :,
|
212 |
+
:] = motion_field_strength_y*(frame_id)
|
213 |
return reference_flow
|
214 |
|
215 |
+
def create_motion_field_and_warp_latents(self, motion_field_strength_x, motion_field_strength_y, frame_ids, video_length, latents):
|
216 |
|
217 |
motion_field = self.create_motion_field(motion_field_strength_x=motion_field_strength_x,
|
218 |
motion_field_strength_y=motion_field_strength_y, latents=latents, video_length=video_length, frame_ids=frame_ids)
|
219 |
for idx, latent in enumerate(latents):
|
220 |
latents[idx] = self.warp_latents_independently(
|
221 |
+
latent[None], motion_field)
|
222 |
return motion_field, latents
|
223 |
|
224 |
@torch.no_grad()
|
|
|
248 |
use_motion_field: bool = True,
|
249 |
smooth_bg: bool = False,
|
250 |
smooth_bg_strength: float = 0.4,
|
|
|
251 |
t0: int = 44,
|
252 |
t1: int = 47,
|
253 |
**kwargs,
|
254 |
):
|
255 |
frame_ids = kwargs.pop("frame_ids", list(range(video_length)))
|
256 |
+
assert t0 < t1
|
257 |
assert num_videos_per_prompt == 1
|
258 |
assert isinstance(prompt, list) and len(prompt) > 0
|
259 |
assert isinstance(negative_prompt, list) or negative_prompt is None
|
|
|
272 |
prompt = prompt_types[0]
|
273 |
negative_prompt = prompt_types[1]
|
274 |
|
|
|
|
|
|
|
|
|
|
|
275 |
# Default height and width to unet
|
276 |
height = height or self.unet.config.sample_size * self.vae_scale_factor
|
277 |
width = width or self.unet.config.sample_size * self.vae_scale_factor
|
|
|
342 |
|
343 |
t0 = timesteps_ddpm[t0]
|
344 |
t1 = timesteps_ddpm[t1]
|
345 |
+
|
346 |
print(f"t0 = {t0} t1 = {t1}")
|
347 |
x_t1_1 = None
|
348 |
|
|
|
354 |
|
355 |
shape = (batch_size, num_channels_latents, 1, height //
|
356 |
self.vae_scale_factor, width // self.vae_scale_factor)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
357 |
|
358 |
ddim_res = self.DDIM_backward(num_inference_steps=num_inference_steps, timesteps=timesteps, skip_t=1000, t0=t0, t1=t1, do_classifier_free_guidance=do_classifier_free_guidance,
|
359 |
null_embs=null_embs, text_embeddings=text_embeddings, latents_local=xT, latents_dtype=dtype, guidance_scale=guidance_scale, guidance_stop_step=guidance_stop_step,
|
|
|
367 |
x_t1_1 = ddim_res["x_t1_1"].detach()
|
368 |
del ddim_res
|
369 |
del xT
|
370 |
+
if use_motion_field:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
371 |
del x0
|
372 |
|
373 |
x_t0_k = x_t0_1[:, :, :1, :, :].repeat(1, 1, video_length-1, 1, 1)
|
374 |
|
375 |
reference_flow, x_t0_k = self.create_motion_field_and_warp_latents(
|
376 |
+
motion_field_strength_x=motion_field_strength_x, motion_field_strength_y=motion_field_strength_y, latents=x_t0_k, video_length=video_length, frame_ids=frame_ids[1:])
|
|
|
377 |
|
378 |
# assuming t0=t1=1000, if t0 = 1000
|
379 |
if t1 > t0:
|
|
|
396 |
del x_t1
|
397 |
del x_t1_1
|
398 |
del x_t1_k
|
|
|
399 |
else:
|
400 |
x_t1 = x_t1_1.clone()
|
401 |
x_t1_1 = x_t1_1[:, :, :1, :, :].clone()
|
|
|
436 |
if use_motion_field:
|
437 |
x_t1_fg_masked_b = x_t1_fg_masked_b[None]
|
438 |
x_t1_fg_masked_b = self.warp_latents_independently(
|
439 |
+
x_t1_fg_masked_b, reference_flow)
|
440 |
else:
|
441 |
x_t1_fg_masked_b = x_t1_fg_masked_b[None]
|
442 |
|
|
|
454 |
m_fg_b = m_fg_1_b.repeat(1, 1, video_length-1, 1, 1)
|
455 |
if use_motion_field:
|
456 |
m_fg_b = self.warp_latents_independently(
|
457 |
+
m_fg_b.clone(), reference_flow)
|
458 |
M_FG_warped.append(
|
459 |
torch.cat([m_fg_1_b[:1, 0], m_fg_b[:1, 0]], dim=1))
|
460 |
|