Skip to content

Commit 759b6f7

Browse files
tomas-gajarskytomas-gajarsky
authored
Merge pull request #33 from tomas-gajarsky/feature/colab_notebook
Feature/colab notebook
2 parents 739199c + 73c95f2 commit 759b6f7

File tree

8 files changed

+2122
-5
lines changed

8 files changed

+2122
-5
lines changed

‎CHANGELOG.md‎

Lines changed: 12 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -1,5 +1,17 @@
11
# Change Log
22

3+
## 0.2.1
4+
5+
Released on January 28, 2023.
6+
7+
### Added
8+
* Google Colab demo notebook
9+
* Google Colab demo notebook link to README
10+
* Merged config yaml files for OmegaConf to /conf/merged/ directory
11+
12+
### Changed
13+
* Do not fail the run if facetorch version cannot be retrieved (likely due to local installation)
14+
315

416
## 0.2.0
517

‎README.md‎

Lines changed: 6 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -6,7 +6,12 @@
66
[![PyPI - License](https://img.shields.io/pypi/l/facetorch)](https://raw.githubusercontent.com/tomas-gajarsky/facetorch/main/LICENSE)
77
<a href="https://github.com/psf/black"><img alt="Code style: black" src="https://img.shields.io/badge/code%20style-black-000000.svg"></a>
88

9-
[Demo App on Hugging Face Spaces 🤗 ](https://huggingface.co/spaces/tomas-gajarsky/facetorch-app)
9+
[Hugging Face Space demo app 🤗 ](https://huggingface.co/spaces/tomas-gajarsky/facetorch-app)
10+
11+
[Google Colab notebook demo](https://colab.research.google.com/github/tomas-gajarsky/facetorch/blob/main/notebooks/facetorch_notebook_demo.ipynb)
12+
<a target="_blank" href="https://colab.research.google.com/github/tomas-gajarsky/facetorch/blob/main/notebooks/facetorch_notebook_demo.ipynb">
13+
<img src="https://colab.research.google.com/assets/colab-badge.svg" alt="Open In Colab"/>
14+
</a>
1015

1116
[User Guide](https://medium.com/@gajarsky.tomas/facetorch-user-guide-a0e9fd2a5552), [Documentation](https://tomas-gajarsky.github.io/facetorch/facetorch/index.html)
1217

‎conf/merged/gpu.merged.config.yaml‎

Lines changed: 349 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,349 @@
1+
analyzer:
2+
device: cuda
3+
optimize_transforms: true
4+
reader:
5+
_target_: facetorch.analyzer.reader.ImageReader
6+
device:
7+
_target_: torch.device
8+
type: ${analyzer.device}
9+
optimize_transform: ${analyzer.optimize_transforms}
10+
transform:
11+
_target_: torchvision.transforms.Compose
12+
transforms:
13+
- _target_: facetorch.transforms.SquarePad
14+
- _target_: torchvision.transforms.Resize
15+
size:
16+
- 1080
17+
detector:
18+
_target_: facetorch.analyzer.detector.FaceDetector
19+
downloader:
20+
_target_: facetorch.downloader.DownloaderGDrive
21+
file_id: 154x2VjmTQVqmowB0yZw4Uck7uQs2vVBs
22+
path_local: /code/models/torchscript/detector/1/model.pt
23+
device:
24+
_target_: torch.device
25+
type: ${analyzer.device}
26+
reverse_colors: true
27+
preprocessor:
28+
_target_: facetorch.analyzer.detector.pre.DetectorPreProcessor
29+
transform:
30+
_target_: torchvision.transforms.Compose
31+
transforms:
32+
- _target_: torchvision.transforms.Normalize
33+
mean:
34+
- 104.0
35+
- 117.0
36+
- 123.0
37+
std:
38+
- 1.0
39+
- 1.0
40+
- 1.0
41+
device:
42+
_target_: torch.device
43+
type: ${analyzer.device}
44+
optimize_transform: ${analyzer.optimize_transforms}
45+
reverse_colors: ${analyzer.detector.reverse_colors}
46+
postprocessor:
47+
_target_: facetorch.analyzer.detector.post.PostRetFace
48+
transform: None
49+
device:
50+
_target_: torch.device
51+
type: ${analyzer.device}
52+
optimize_transform: ${analyzer.optimize_transforms}
53+
confidence_threshold: 0.02
54+
top_k: 5000
55+
nms_threshold: 0.4
56+
keep_top_k: 750
57+
score_threshold: 0.6
58+
prior_box:
59+
_target_: facetorch.analyzer.detector.post.PriorBox
60+
min_sizes:
61+
- - 16
62+
- 32
63+
- - 64
64+
- 128
65+
- - 256
66+
- 512
67+
steps:
68+
- 8
69+
- 16
70+
- 32
71+
clip: false
72+
variance:
73+
- 0.1
74+
- 0.2
75+
reverse_colors: ${analyzer.detector.reverse_colors}
76+
expand_box_ratio: 0.1
77+
unifier:
78+
_target_: facetorch.analyzer.unifier.FaceUnifier
79+
transform:
80+
_target_: torchvision.transforms.Compose
81+
transforms:
82+
- _target_: torchvision.transforms.Normalize
83+
mean:
84+
- -123.0
85+
- -117.0
86+
- -104.0
87+
std:
88+
- 255.0
89+
- 255.0
90+
- 255.0
91+
- _target_: torchvision.transforms.Resize
92+
size:
93+
- 380
94+
- 380
95+
device:
96+
_target_: torch.device
97+
type: ${analyzer.device}
98+
optimize_transform: ${analyzer.optimize_transforms}
99+
predictor:
100+
embed:
101+
_target_: facetorch.analyzer.predictor.FacePredictor
102+
downloader:
103+
_target_: facetorch.downloader.DownloaderGDrive
104+
file_id: 19h3kqar1wlELAmM5hDyj9tlrUh8yjrCl
105+
path_local: /code/models/torchscript/predictor/embed/1/model.pt
106+
device:
107+
_target_: torch.device
108+
type: ${analyzer.device}
109+
preprocessor:
110+
_target_: facetorch.analyzer.predictor.pre.PredictorPreProcessor
111+
transform:
112+
_target_: torchvision.transforms.Compose
113+
transforms:
114+
- _target_: torchvision.transforms.Resize
115+
size:
116+
- 244
117+
- 244
118+
- _target_: torchvision.transforms.Normalize
119+
mean:
120+
- 0.485
121+
- 0.456
122+
- 0.406
123+
std:
124+
- 0.228
125+
- 0.224
126+
- 0.225
127+
device:
128+
_target_: torch.device
129+
type: ${analyzer.predictor.fer.device.type}
130+
optimize_transform: ${analyzer.optimize_transforms}
131+
reverse_colors: false
132+
postprocessor:
133+
_target_: facetorch.analyzer.predictor.post.PostEmbedder
134+
transform: None
135+
device:
136+
_target_: torch.device
137+
type: ${analyzer.predictor.fer.device.type}
138+
optimize_transform: ${analyzer.optimize_transforms}
139+
labels:
140+
- abstract
141+
verify:
142+
_target_: facetorch.analyzer.predictor.FacePredictor
143+
downloader:
144+
_target_: facetorch.downloader.DownloaderGDrive
145+
file_id: 1WI-mP_0mGW31OHfriPUsuFS_usYh_W8p
146+
path_local: /code/models/torchscript/predictor/verify/2/model.pt
147+
device:
148+
_target_: torch.device
149+
type: ${analyzer.device}
150+
preprocessor:
151+
_target_: facetorch.analyzer.predictor.pre.PredictorPreProcessor
152+
transform:
153+
_target_: torchvision.transforms.Compose
154+
transforms:
155+
- _target_: torchvision.transforms.Resize
156+
size:
157+
- 112
158+
- 112
159+
- _target_: torchvision.transforms.Normalize
160+
mean:
161+
- 0.5
162+
- 0.5
163+
- 0.5
164+
std:
165+
- 0.5
166+
- 0.5
167+
- 0.5
168+
device:
169+
_target_: torch.device
170+
type: ${analyzer.predictor.verify.device.type}
171+
optimize_transform: ${analyzer.optimize_transforms}
172+
reverse_colors: true
173+
postprocessor:
174+
_target_: facetorch.analyzer.predictor.post.PostEmbedder
175+
transform: None
176+
device:
177+
_target_: torch.device
178+
type: ${analyzer.predictor.verify.device.type}
179+
optimize_transform: ${analyzer.optimize_transforms}
180+
labels:
181+
- abstract
182+
fer:
183+
_target_: facetorch.analyzer.predictor.FacePredictor
184+
downloader:
185+
_target_: facetorch.downloader.DownloaderGDrive
186+
file_id: 1xoB5VYOd0XLjb-rQqqHWCkQvma4NytEd
187+
path_local: /code/models/torchscript/predictor/fer/2/model.pt
188+
device:
189+
_target_: torch.device
190+
type: ${analyzer.device}
191+
preprocessor:
192+
_target_: facetorch.analyzer.predictor.pre.PredictorPreProcessor
193+
transform:
194+
_target_: torchvision.transforms.Compose
195+
transforms:
196+
- _target_: torchvision.transforms.Resize
197+
size:
198+
- 260
199+
- 260
200+
- _target_: torchvision.transforms.Normalize
201+
mean:
202+
- 0.485
203+
- 0.456
204+
- 0.406
205+
std:
206+
- 0.229
207+
- 0.224
208+
- 0.225
209+
device:
210+
_target_: torch.device
211+
type: ${analyzer.predictor.fer.device.type}
212+
optimize_transform: ${analyzer.optimize_transforms}
213+
reverse_colors: false
214+
postprocessor:
215+
_target_: facetorch.analyzer.predictor.post.PostArgMax
216+
transform: None
217+
device:
218+
_target_: torch.device
219+
type: ${analyzer.predictor.fer.device.type}
220+
optimize_transform: ${analyzer.optimize_transforms}
221+
dim: 1
222+
labels:
223+
- Anger
224+
- Contempt
225+
- Disgust
226+
- Fear
227+
- Happiness
228+
- Neutral
229+
- Sadness
230+
- Surprise
231+
deepfake:
232+
_target_: facetorch.analyzer.predictor.FacePredictor
233+
downloader:
234+
_target_: facetorch.downloader.DownloaderGDrive
235+
file_id: 1GjDTwQpvrkCjXOdiBy1oMkzm7nt-bXFg
236+
path_local: /code/models/torchscript/predictor/deepfake/1/model.pt
237+
device:
238+
_target_: torch.device
239+
type: ${analyzer.device}
240+
preprocessor:
241+
_target_: facetorch.analyzer.predictor.pre.PredictorPreProcessor
242+
transform:
243+
_target_: torchvision.transforms.Compose
244+
transforms:
245+
- _target_: torchvision.transforms.Resize
246+
size:
247+
- 380
248+
- 380
249+
- _target_: torchvision.transforms.Normalize
250+
mean:
251+
- 0.485
252+
- 0.456
253+
- 0.406
254+
std:
255+
- 0.229
256+
- 0.224
257+
- 0.225
258+
device:
259+
_target_: torch.device
260+
type: ${analyzer.device}
261+
optimize_transform: ${analyzer.optimize_transforms}
262+
reverse_colors: false
263+
postprocessor:
264+
_target_: facetorch.analyzer.predictor.post.PostSigmoidBinary
265+
transform: None
266+
device:
267+
_target_: torch.device
268+
type: ${analyzer.device}
269+
optimize_transform: ${analyzer.optimize_transforms}
270+
labels:
271+
- Real
272+
- Fake
273+
threshold: 0.7
274+
align:
275+
_target_: facetorch.analyzer.predictor.FacePredictor
276+
downloader:
277+
_target_: facetorch.downloader.DownloaderGDrive
278+
file_id: 16gNFQdEH2nWvW3zTbdIAniKIbPAp6qBA
279+
path_local: /code/models/torchscript/predictor/align/1/model.pt
280+
device:
281+
_target_: torch.device
282+
type: ${analyzer.device}
283+
preprocessor:
284+
_target_: facetorch.analyzer.predictor.pre.PredictorPreProcessor
285+
transform:
286+
_target_: torchvision.transforms.Compose
287+
transforms:
288+
- _target_: torchvision.transforms.Resize
289+
size:
290+
- 120
291+
- 120
292+
device:
293+
_target_: torch.device
294+
type: ${analyzer.predictor.align.device.type}
295+
optimize_transform: ${analyzer.optimize_transforms}
296+
reverse_colors: false
297+
postprocessor:
298+
_target_: facetorch.analyzer.predictor.post.PostEmbedder
299+
transform: None
300+
device:
301+
_target_: torch.device
302+
type: ${analyzer.predictor.align.device.type}
303+
optimize_transform: ${analyzer.optimize_transforms}
304+
labels:
305+
- abstract
306+
utilizer:
307+
align:
308+
_target_: facetorch.analyzer.utilizer.align.Lmk3DMeshPose
309+
transform: None
310+
device:
311+
_target_: torch.device
312+
type: ${analyzer.device}
313+
optimize_transform: false
314+
downloader_meta:
315+
_target_: facetorch.downloader.DownloaderGDrive
316+
file_id: 11tdAcFuSXqCCf58g52WT1Rpa8KuQwe2o
317+
path_local: /code/data/3dmm/meta.pt
318+
image_size: 120
319+
draw_boxes:
320+
_target_: facetorch.analyzer.utilizer.draw.BoxDrawer
321+
transform: None
322+
device:
323+
_target_: torch.device
324+
type: ${analyzer.device}
325+
optimize_transform: false
326+
color: green
327+
line_width: 3
328+
draw_landmarks:
329+
_target_: facetorch.analyzer.utilizer.draw.LandmarkDrawerTorch
330+
transform: None
331+
device:
332+
_target_: torch.device
333+
type: ${analyzer.device}
334+
optimize_transform: false
335+
width: 2
336+
color: green
337+
logger:
338+
_target_: facetorch.logger.LoggerJsonFile
339+
name: facetorch
340+
level: 20
341+
path_file: /code/logs/facetorch/main.log
342+
json_format: '%(asctime)s %(levelname)s %(message)s'
343+
main:
344+
sleep: 3
345+
debug: true
346+
batch_size: 8
347+
fix_img_size: true
348+
return_img_data: true
349+
include_tensors: true

0 commit comments

Comments
 (0)