-
Notifications
You must be signed in to change notification settings - Fork 0
/
configuration_nvidia.txt
256 lines (238 loc) · 7.54 KB
/
configuration_nvidia.txt
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
[COMMON]
datasetPathVideo = /media/gamal/Passport/Datasets/VoxCeleb2/Voxceleb2Video
datasetPathDatabase = /data_store/databases
test_datasetPathVideo = /media/gamal/Passport/Datasets/Voxceleb2TestVideo
test_datasetPathDatabase = /media/gamal/Passport/Datasets/VoxCeleb2Test/Voxceleb2TestDatabase
test_datasetPathVideo_p = /media/gamal/Passport/Datasets/VoxCeleb2Test/Voxceleb2TestVideo
test_datasetPathDatabase_p = /media/gamal/Passport/Datasets/VoxCeleb2Test/Voxceleb2TestDatabase
cuda=1
cpus=8
resizeImageTo = 128
begin_with_image_size = 8
audio_embs_options = wav2vec or openl3
audio_embs = pyannoteTitaNet
unet1_dim = 192
unet2_dim = 128
unet1_image_size = 32
insert_amd_env_vars = 0
HSA_OVERRIDE_GFX_VERSION = 10.3.0
ROCM_PATH = /opt/rocm
timesteps = 1000
[dbCreateAndPopulate]
recreateDb=0
[extractAudio]
datasetPathAudio = /media/gamal/Passport/Datasets/VoxCeleb2/Voxceleb2Audio
dbChunk = 240
time_to_wait_before_deleting_files = 60
[extractFaces]
expandFaceVerticalRatio = 0.4
expandFaceHorizontalRatio=0.5
faceDetectionDeepFaceBackend=4
parallelism = 5
parallelismFrames = 1
datasetPathFrames = /media/gamal/Passport/Datasets/VoxCeleb2/Voxceleb2Frames
datasetPathFaces = /media/gamal/Passport/Datasets/VoxCeleb2/Voxceleb2Faces
[fineTuneStableDiffusionTraining]
db_chunk = 50000
dev_mode = 0
continue_from_epoch = 1
continue_from_offset = 0
continue_from_epoch_and_offset = 0
unconditional_guidance_scale = 40
[extractOpenL3]
datasetPathAudio = /media/gamal/Passport/Datasets/VoxCeleb2/Voxceleb2Audio
dbChunk = 240
time_to_wait_before_deleting_files = 60
openl3_mode_options = stable or imagen
openl3_mode = imagen
[extractWavToVec]
datasetPathAudio = /media/gamal/Passport/Datasets/VoxCeleb2/Voxceleb2Audio
dbChunk = 240
time_to_wait_before_deleting_files = 60
audio_length_wav2vec = 5
[extractPyannoteTitaNet]
datasetPathAudio = /media/gamal/Passport/Datasets/VoxCeleb2/Voxceleb2Audio
dbChunk = 240
time_to_wait_before_deleting_files = 60
use_auth_token = TO_BE_GOTTEN
[extractAudioFeatures]
datasetPathAudio = /media/gamal/Passport/Datasets/VoxCeleb2/Voxceleb2Audio
dbChunk = 250
time_to_wait_before_deleting_files = 60
[fineTuneStableDiffusionTesting]
use_video_in_configuration = 1
video_path = /home/gamal/vsc/DiffusionSpeech2Face/v.mp4
time_to_wait_before_deleting_files = 180
dev_mode = 0
[train_imagen]
model_filename = /home/gamal/vsc/DiffusionSpeech2Face/checkpoint/noablation
imagen_samples_folder = /media/gamal/Passport/active_models/imagen_features_low_res_asis_8
audio_length_used = 24
override_image_path = 1
override_image_path_string = /data_store/faces/dev/mp4
override_leave_parent_level = 4
max_checkpoints_keep = 2
sample_every_offset = 1
save_every_offset = 1
epochs = 100000
sub_epochs = 1
inner_epochs = 100
sample_every = 100000
sample_probability = 0
save_model_every = 75
batch_size = 10
batches_per_epoch = 250
db_chunk = 5000
dask_chunk=5000
c_unte2 = 54500
c_unite1 = 512000
stop_at_no_of_samples = 2000000
aaaa = 238996
ignore_speaker_embedding = 0
ignore_speech_brain = 0
ignore_pyannote_titanet_speakernet = 0
ignore_audio_features = 0
ignore_pyAudioAnalysis = 0
ignore_librosa = 0
ignore_image_guide = 0
ignore_additional_attributes = 0
ignore_age = 0
ignore_gender = 0
ignore_ethnicity = 0
ignore_language_spoken = 0
[test_imagen]
video_path = /media/gamal/Passport/Datasets/VoxCeleb2TestMin/Voxceleb2TestVideo/id00017/01dfn2spqyE/00001.mp4
time_to_wait_before_deleting_files = 180
audio_length_used = 24
model_filename = imagen_features_low_res_asis_8
openl3_mode_options = stable or imagen
openl3_mode = imagen
folder = imagen-test-folder
number_of_images = 2
age = 25
ethnicity = black
gender = man
language = English
image_guide_path = /media/gamal/Passport/Datasets/VoxCeleb2TestMin/Voxceleb2TestGuides/id00017/01dfn2spqyE/00001_guide_.png
blur_or_pixelate = 1
boxBlurMin = 4
boxBlurMax = 14
gaussianBlurMin=4
gaussianBlurMax=8
pix_to_min = 8
pix_to_max = 24
low_res_image = /media/gamal/Passport/Datasets/VoxCeleb2TestMin/Voxceleb2TestGuides/id00017/01dfn2spqyE/00001_guide_.png
skip_speaker_embedding = 0
skip_speech_brain = 0
skip_pyannote_titanet_speakernet = 0
skip_audio_features = 0
skip_pyAudioAnalysis = 0
skip_librosa = 0
skip_image_guide = 0
skip_additional_attributes = 0
skip_age = 0
skip_gender = 0
skip_ethnicity = 0
skip_language_spoken = 0
[evaluate_imagen]
videos_per_user = 1
test_datasetpathfaces = /media/gamal/Passport/Datasets/VoxCeleb2Test/Voxceleb2TestFaces
test_datasetpathgeneratedfaces = /media/gamal/Passport/Datasets/VoxCeleb2Test/Voxceleb2TestGeneratedFaces
test_datasetpathfeatures = /media/gamal/Passport/Datasets/VoxCeleb2Test/Voxceleb2TestFeatures
test_datasetpathguides = /media/gamal/Passport/Datasets/VoxCeleb2Test/Voxceleb2TestGuides
test_datasetpathfaces_pooled = /media/gamal/Passport/Datasets/VoxCeleb2Test/FacesPooled
test_datasetpathgeneratedfaces_pooled = /media/gamal/Passport/Datasets/VoxCeleb2Test/GeneratedFacesPooled
test_datasetpathfaces_p = /media/gamal/Passport/Datasets/VoxCeleb2Test/Voxceleb2TestFaces
test_datasetpathgeneratedfaces_p = /media/gamal/Passport/Datasets/VoxCeleb2Test/Voxceleb2TestGeneratedFaces
test_datasetpathfeatures_p = /media/gamal/Passport/Datasets/VoxCeleb2Test/Voxceleb2TestFeatures
test_datasetpathguides_p = /media/gamal/Passport/Datasets/VoxCeleb2Test/Voxceleb2TestGuides
test_datasetpathfaces_pooled_p = /media/gamal/Passport/Datasets/VoxCeleb2Test/FacesPooled
test_datasetpathgeneratedfaces_pooled_p = /media/gamal/Passport/Datasets/VoxCeleb2Test/FacesPooled
time_to_wait_before_deleting_files = 180
audio_length_used = 24
sampling_chunk=25
preprocessing_chunk=500
cond_scale = 5
model_filename = imagen_features_low_res_asis_8
openl3_mode_options = stable or imagen
openl3_mode = imagen
folder = imagen-test-folder
boxBlurMin = 4
boxBlurMax = 14
gaussianBlurMin=4
gaussianBlurMax=8
pix_to_min = 8
pix_to_max = 24
expandFaceVerticalRatio = 0.4
expandFaceHorizontalRatio=0.5
faceDetectionDeepFaceBackend=4
generated_face_table_name = test_not_fully_trained
skip_speaker_embedding = 0
skip_speech_brain = 0
skip_pyannote_titanet_speakernet = 0
skip_audio_features = 0
skip_pyAudioAnalysis = 0
skip_librosa = 0
skip_image_guide = 0
skip_additional_attributes = 0
skip_age = 0
skip_gender = 0
skip_ethnicity = 0
skip_language_spoken = 0
evaluation_results_folder = /media/gamal/Passport/Datasets/VoxCeleb2Test
ignore_speaker_embedding = 0
ignore_speech_brain = 0
ignore_pyannote_titanet_speakernet = 0
ignore_audio_features = 0
ignore_pyAudioAnalysis = 0
ignore_librosa = 0
ignore_image_guide = 0
ignore_additional_attributes = 0
ignore_age = 0
ignore_gender = 0
ignore_ethnicity = 0
ignore_language_spoken = 0
[extractVggBlurred]
dbChunk = 200
boxBlurMin = 4
boxBlurMax = 14
gaussianBlurMin=4
gaussianBlurMax=8
pix_to_min = 8
pix_to_max = 24
[extractAudioSpectogramTransformer]
datasetPathAudio = /media/gamal/Passport/Datasets/VoxCeleb2/Voxceleb2Audio
dbChunk = 240
time_to_wait_before_deleting_files = 60
[train_muse]
model_filename = muse_features_low_res_asis
imagen_samples_folder = /home/gamal/Datasets/SAMPLES
audio_length_used = 24
sample_every_offset = 1
save_every_offset = 1
epochs = 1
sub_epochs = 1
inner_epochs = 1
sample_every = 4000
sample_probability = 100
save_model_every = 4000
batch_size = 4
db_chunk = 100
dask_chunk=100
c_unte2 = 54500
c_unite1 = 77000
stop_at_no_of_samples = 238996
aaaa = 238996
ignore_speaker_embedding = 0
ignore_speech_brain = 0
ignore_pyannote_titanet_speakernet = 0
ignore_audio_features = 0
ignore_pyAudioAnalysis = 0
ignore_librosa = 0
ignore_image_guide = 0
ignore_additional_attributes = 0
ignore_age = 0
ignore_gender = 0
ignore_ethnicity = 0
ignore_language_spoken = 0
vae_file = /home/gamal/models/vae.12145000.pt