Spaces:
Runtime error
Runtime error
Anonymous-sub
commited on
Commit
•
9c1dc83
1
Parent(s):
251e479
Update app.py
Browse files
app.py
CHANGED
@@ -30,12 +30,29 @@ from src.img_util import find_flat_region, numpy2tensor
|
|
30 |
from src.video_util import (frame_to_video, get_fps, get_frame_count,
|
31 |
prepare_frames)
|
32 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
33 |
inversed_model_dict = dict()
|
34 |
for k, v in model_dict.items():
|
35 |
inversed_model_dict[v] = k
|
36 |
|
37 |
to_tensor = T.PILToTensor()
|
38 |
blur = T.GaussianBlur(kernel_size=(9, 9), sigma=(18, 18))
|
|
|
39 |
|
40 |
|
41 |
class ProcessingState(Enum):
|
@@ -64,7 +81,7 @@ class GlobalState:
|
|
64 |
attention_type='swin',
|
65 |
ffn_dim_expansion=4,
|
66 |
num_transformer_layers=6,
|
67 |
-
).to(
|
68 |
|
69 |
checkpoint = torch.load('models/gmflow_sintel-0c07dcb3.pth',
|
70 |
map_location=lambda storage, loc: storage)
|
@@ -86,25 +103,32 @@ class GlobalState:
|
|
86 |
model = create_model('./ControlNet/models/cldm_v15.yaml').cpu()
|
87 |
if control_type == 'HED':
|
88 |
model.load_state_dict(
|
89 |
-
load_state_dict(
|
90 |
-
|
|
|
91 |
elif control_type == 'canny':
|
92 |
model.load_state_dict(
|
93 |
-
load_state_dict(
|
94 |
-
|
95 |
-
|
|
|
96 |
sd_model_path = model_dict[sd_model]
|
97 |
if len(sd_model_path) > 0:
|
98 |
model_ext = os.path.splitext(sd_model_path)[1]
|
|
|
|
|
99 |
if model_ext == '.safetensors':
|
100 |
-
model.load_state_dict(load_file(
|
101 |
-
elif model_ext == '.ckpt' or model_ext == '.pth':
|
102 |
-
model.load_state_dict(torch.load(sd_model_path)['state_dict'],
|
103 |
strict=False)
|
|
|
|
|
|
|
104 |
|
105 |
try:
|
106 |
model.first_stage_model.load_state_dict(torch.load(
|
107 |
-
|
|
|
|
|
108 |
strict=False)
|
109 |
except Exception:
|
110 |
print('Warning: We suggest you download the fine-tuned VAE',
|
@@ -115,7 +139,8 @@ class GlobalState:
|
|
115 |
def clear_sd_model(self):
|
116 |
self.sd_model = None
|
117 |
self.ddim_v_sampler = None
|
118 |
-
|
|
|
119 |
|
120 |
def update_detector(self, control_type, canny_low=100, canny_high=200):
|
121 |
if self.detector_type == control_type:
|
@@ -286,14 +311,14 @@ def process1(*args):
|
|
286 |
img_ = numpy2tensor(img)
|
287 |
|
288 |
def generate_first_img(img_, strength):
|
289 |
-
encoder_posterior = model.encode_first_stage(img_.
|
290 |
x0 = model.get_first_stage_encoding(encoder_posterior).detach()
|
291 |
|
292 |
detected_map = detector(img)
|
293 |
detected_map = HWC3(detected_map)
|
294 |
|
295 |
control = torch.from_numpy(
|
296 |
-
detected_map.copy()).float().
|
297 |
control = torch.stack([control for _ in range(num_samples)], dim=0)
|
298 |
control = einops.rearrange(control, 'b h w c -> b c h w').clone()
|
299 |
cond = {
|
@@ -411,13 +436,14 @@ def process2(*args):
|
|
411 |
img_ = apply_color_correction(global_state.color_corrections,
|
412 |
Image.fromarray(img))
|
413 |
img_ = to_tensor(img_).unsqueeze(0)[:, :3] / 127.5 - 1
|
414 |
-
encoder_posterior = model.encode_first_stage(img_.
|
415 |
x0 = model.get_first_stage_encoding(encoder_posterior).detach()
|
416 |
|
417 |
detected_map = detector(img)
|
418 |
detected_map = HWC3(detected_map)
|
419 |
|
420 |
-
control = torch.from_numpy(
|
|
|
421 |
control = torch.stack([control for _ in range(num_samples)], dim=0)
|
422 |
control = einops.rearrange(control, 'b h w c -> b c h w').clone()
|
423 |
cond = {
|
|
|
30 |
from src.video_util import (frame_to_video, get_fps, get_frame_count,
|
31 |
prepare_frames)
|
32 |
|
33 |
+
import huggingface_hub
|
34 |
+
|
35 |
+
repo_name = 'Anonymous-sub/Rerender'
|
36 |
+
|
37 |
+
huggingface_hub.hf_hub_download(repo_name,
|
38 |
+
'pexels-koolshooters-7322716.mp4',
|
39 |
+
local_dir='videos')
|
40 |
+
huggingface_hub.hf_hub_download(
|
41 |
+
repo_name,
|
42 |
+
'pexels-antoni-shkraba-8048492-540x960-25fps.mp4',
|
43 |
+
local_dir='videos')
|
44 |
+
huggingface_hub.hf_hub_download(
|
45 |
+
repo_name,
|
46 |
+
'pexels-cottonbro-studio-6649832-960x506-25fps.mp4',
|
47 |
+
local_dir='videos')
|
48 |
+
|
49 |
inversed_model_dict = dict()
|
50 |
for k, v in model_dict.items():
|
51 |
inversed_model_dict[v] = k
|
52 |
|
53 |
to_tensor = T.PILToTensor()
|
54 |
blur = T.GaussianBlur(kernel_size=(9, 9), sigma=(18, 18))
|
55 |
+
device = 'cuda' if torch.cuda.is_available() else 'cpu'
|
56 |
|
57 |
|
58 |
class ProcessingState(Enum):
|
|
|
81 |
attention_type='swin',
|
82 |
ffn_dim_expansion=4,
|
83 |
num_transformer_layers=6,
|
84 |
+
).to(device)
|
85 |
|
86 |
checkpoint = torch.load('models/gmflow_sintel-0c07dcb3.pth',
|
87 |
map_location=lambda storage, loc: storage)
|
|
|
103 |
model = create_model('./ControlNet/models/cldm_v15.yaml').cpu()
|
104 |
if control_type == 'HED':
|
105 |
model.load_state_dict(
|
106 |
+
load_state_dict(huggingface_hub.hf_hub_download(
|
107 |
+
'lllyasviel/ControlNet', './models/control_sd15_hed.pth'),
|
108 |
+
location=device))
|
109 |
elif control_type == 'canny':
|
110 |
model.load_state_dict(
|
111 |
+
load_state_dict(huggingface_hub.hf_hub_download(
|
112 |
+
'lllyasviel/ControlNet', 'models/control_sd15_canny.pth'),
|
113 |
+
location=device))
|
114 |
+
model.to(device)
|
115 |
sd_model_path = model_dict[sd_model]
|
116 |
if len(sd_model_path) > 0:
|
117 |
model_ext = os.path.splitext(sd_model_path)[1]
|
118 |
+
downloaded_model = huggingface_hub.hf_hub_download(
|
119 |
+
repo_name, sd_model_path)
|
120 |
if model_ext == '.safetensors':
|
121 |
+
model.load_state_dict(load_file(downloaded_model),
|
|
|
|
|
122 |
strict=False)
|
123 |
+
elif model_ext == '.ckpt' or model_ext == '.pth':
|
124 |
+
model.load_state_dict(
|
125 |
+
torch.load(downloaded_model)['state_dict'], strict=False)
|
126 |
|
127 |
try:
|
128 |
model.first_stage_model.load_state_dict(torch.load(
|
129 |
+
huggingface_hub.hf_hub_download(
|
130 |
+
'stabilityai/sd-vae-ft-mse-original',
|
131 |
+
'vae-ft-mse-840000-ema-pruned.ckpt'))['state_dict'],
|
132 |
strict=False)
|
133 |
except Exception:
|
134 |
print('Warning: We suggest you download the fine-tuned VAE',
|
|
|
139 |
def clear_sd_model(self):
|
140 |
self.sd_model = None
|
141 |
self.ddim_v_sampler = None
|
142 |
+
if device == 'cuda':
|
143 |
+
torch.cuda.empty_cache()
|
144 |
|
145 |
def update_detector(self, control_type, canny_low=100, canny_high=200):
|
146 |
if self.detector_type == control_type:
|
|
|
311 |
img_ = numpy2tensor(img)
|
312 |
|
313 |
def generate_first_img(img_, strength):
|
314 |
+
encoder_posterior = model.encode_first_stage(img_.to(device))
|
315 |
x0 = model.get_first_stage_encoding(encoder_posterior).detach()
|
316 |
|
317 |
detected_map = detector(img)
|
318 |
detected_map = HWC3(detected_map)
|
319 |
|
320 |
control = torch.from_numpy(
|
321 |
+
detected_map.copy()).float().to(device) / 255.0
|
322 |
control = torch.stack([control for _ in range(num_samples)], dim=0)
|
323 |
control = einops.rearrange(control, 'b h w c -> b c h w').clone()
|
324 |
cond = {
|
|
|
436 |
img_ = apply_color_correction(global_state.color_corrections,
|
437 |
Image.fromarray(img))
|
438 |
img_ = to_tensor(img_).unsqueeze(0)[:, :3] / 127.5 - 1
|
439 |
+
encoder_posterior = model.encode_first_stage(img_.to(device))
|
440 |
x0 = model.get_first_stage_encoding(encoder_posterior).detach()
|
441 |
|
442 |
detected_map = detector(img)
|
443 |
detected_map = HWC3(detected_map)
|
444 |
|
445 |
+
control = torch.from_numpy(
|
446 |
+
detected_map.copy()).float().to(device) / 255.0
|
447 |
control = torch.stack([control for _ in range(num_samples)], dim=0)
|
448 |
control = einops.rearrange(control, 'b h w c -> b c h w').clone()
|
449 |
cond = {
|