Spaces:
Sleeping
Sleeping
add TTS_models and vocoder lfs support
Browse files- .gitattributes +2 -1
- {TTS_model β TTS_models/libritts_xvector_vits}/config.yaml +0 -0
- {TTS_model β TTS_models/libritts_xvector_vits}/images/discriminator_backward_time.png +0 -0
- {TTS_model β TTS_models/libritts_xvector_vits}/images/discriminator_fake_loss.png +0 -0
- {TTS_model β TTS_models/libritts_xvector_vits}/images/discriminator_forward_time.png +0 -0
- {TTS_model β TTS_models/libritts_xvector_vits}/images/discriminator_loss.png +0 -0
- {TTS_model β TTS_models/libritts_xvector_vits}/images/discriminator_optim_step_time.png +0 -0
- {TTS_model β TTS_models/libritts_xvector_vits}/images/discriminator_real_loss.png +0 -0
- {TTS_model β TTS_models/libritts_xvector_vits}/images/discriminator_train_time.png +0 -0
- {TTS_model β TTS_models/libritts_xvector_vits}/images/generator_adv_loss.png +0 -0
- {TTS_model β TTS_models/libritts_xvector_vits}/images/generator_backward_time.png +0 -0
- {TTS_model β TTS_models/libritts_xvector_vits}/images/generator_dur_loss.png +0 -0
- {TTS_model β TTS_models/libritts_xvector_vits}/images/generator_feat_match_loss.png +0 -0
- {TTS_model β TTS_models/libritts_xvector_vits}/images/generator_forward_time.png +0 -0
- {TTS_model β TTS_models/libritts_xvector_vits}/images/generator_kl_loss.png +0 -0
- {TTS_model β TTS_models/libritts_xvector_vits}/images/generator_loss.png +0 -0
- {TTS_model β TTS_models/libritts_xvector_vits}/images/generator_mel_loss.png +0 -0
- {TTS_model β TTS_models/libritts_xvector_vits}/images/generator_optim_step_time.png +0 -0
- {TTS_model β TTS_models/libritts_xvector_vits}/images/generator_train_time.png +0 -0
- {TTS_model β TTS_models/libritts_xvector_vits}/images/gpu_max_cached_mem_GB.png +0 -0
- {TTS_model β TTS_models/libritts_xvector_vits}/images/iter_time.png +0 -0
- {TTS_model β TTS_models/libritts_xvector_vits}/images/optim0_lr0.png +0 -0
- {TTS_model β TTS_models/libritts_xvector_vits}/images/optim1_lr0.png +0 -0
- {TTS_model β TTS_models/libritts_xvector_vits}/images/train_time.png +0 -0
- {TTS_model β TTS_models/libritts_xvector_vits}/train.total_count.ave_10best.pth +0 -0
- app.py +3 -3
- vocoders/libritts_hifigan.v1/checkpoint-2500000steps.pkl +3 -0
- vocoders/libritts_hifigan.v1/config.yml +191 -0
- vocoders/libritts_hifigan.v1/stats.h5 +3 -0
- vocoders/vctk_parallel_wavegan.v1.long/._checkpoint-1000000steps.pkl +3 -0
- vocoders/vctk_parallel_wavegan.v1.long/._config.yml +0 -0
- vocoders/vctk_parallel_wavegan.v1.long/._stats.h5 +3 -0
- vocoders/vctk_parallel_wavegan.v1.long/._train_nodev_all_vctk_parallel_wavegan.v1.long +0 -0
- vocoders/vctk_parallel_wavegan.v1.long/checkpoint-1000000steps.pkl +3 -0
- vocoders/vctk_parallel_wavegan.v1.long/config.yml +104 -0
- vocoders/vctk_parallel_wavegan.v1.long/stats.h5 +3 -0
.gitattributes
CHANGED
@@ -33,4 +33,5 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
|
33 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
34 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
35 |
xvector filter=lfs diff=lfs merge=lfs -text
|
36 |
-
|
|
|
|
33 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
34 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
35 |
xvector filter=lfs diff=lfs merge=lfs -text
|
36 |
+
TTS_models filter=lfs diff=lfs merge=lfs -text
|
37 |
+
vocoders filter=lfs diff=lfs merge=lfs -text
|
{TTS_model β TTS_models/libritts_xvector_vits}/config.yaml
RENAMED
File without changes
|
{TTS_model β TTS_models/libritts_xvector_vits}/images/discriminator_backward_time.png
RENAMED
File without changes
|
{TTS_model β TTS_models/libritts_xvector_vits}/images/discriminator_fake_loss.png
RENAMED
File without changes
|
{TTS_model β TTS_models/libritts_xvector_vits}/images/discriminator_forward_time.png
RENAMED
File without changes
|
{TTS_model β TTS_models/libritts_xvector_vits}/images/discriminator_loss.png
RENAMED
File without changes
|
{TTS_model β TTS_models/libritts_xvector_vits}/images/discriminator_optim_step_time.png
RENAMED
File without changes
|
{TTS_model β TTS_models/libritts_xvector_vits}/images/discriminator_real_loss.png
RENAMED
File without changes
|
{TTS_model β TTS_models/libritts_xvector_vits}/images/discriminator_train_time.png
RENAMED
File without changes
|
{TTS_model β TTS_models/libritts_xvector_vits}/images/generator_adv_loss.png
RENAMED
File without changes
|
{TTS_model β TTS_models/libritts_xvector_vits}/images/generator_backward_time.png
RENAMED
File without changes
|
{TTS_model β TTS_models/libritts_xvector_vits}/images/generator_dur_loss.png
RENAMED
File without changes
|
{TTS_model β TTS_models/libritts_xvector_vits}/images/generator_feat_match_loss.png
RENAMED
File without changes
|
{TTS_model β TTS_models/libritts_xvector_vits}/images/generator_forward_time.png
RENAMED
File without changes
|
{TTS_model β TTS_models/libritts_xvector_vits}/images/generator_kl_loss.png
RENAMED
File without changes
|
{TTS_model β TTS_models/libritts_xvector_vits}/images/generator_loss.png
RENAMED
File without changes
|
{TTS_model β TTS_models/libritts_xvector_vits}/images/generator_mel_loss.png
RENAMED
File without changes
|
{TTS_model β TTS_models/libritts_xvector_vits}/images/generator_optim_step_time.png
RENAMED
File without changes
|
{TTS_model β TTS_models/libritts_xvector_vits}/images/generator_train_time.png
RENAMED
File without changes
|
{TTS_model β TTS_models/libritts_xvector_vits}/images/gpu_max_cached_mem_GB.png
RENAMED
File without changes
|
{TTS_model β TTS_models/libritts_xvector_vits}/images/iter_time.png
RENAMED
File without changes
|
{TTS_model β TTS_models/libritts_xvector_vits}/images/optim0_lr0.png
RENAMED
File without changes
|
{TTS_model β TTS_models/libritts_xvector_vits}/images/optim1_lr0.png
RENAMED
File without changes
|
{TTS_model β TTS_models/libritts_xvector_vits}/images/train_time.png
RENAMED
File without changes
|
{TTS_model β TTS_models/libritts_xvector_vits}/train.total_count.ave_10best.pth
RENAMED
File without changes
|
app.py
CHANGED
@@ -64,8 +64,8 @@ from espnet2.utils.types import str_or_none
|
|
64 |
|
65 |
# local import
|
66 |
text2speech = Text2Speech.from_pretrained(
|
67 |
-
train_config = "
|
68 |
-
model_file="
|
69 |
vocoder_tag=str_or_none(vocoder_tag),
|
70 |
device="cuda",
|
71 |
use_att_constraint=False,
|
@@ -106,7 +106,7 @@ transformer_text2speech = Text2Speech.from_pretrained(
|
|
106 |
forward_window=3,
|
107 |
speed_control_alpha=1.0,
|
108 |
)
|
109 |
-
|
110 |
# from google.cloud import texttospeech
|
111 |
# Google_TTS_client = texttospeech.TextToSpeechClient()
|
112 |
|
|
|
64 |
|
65 |
# local import
|
66 |
text2speech = Text2Speech.from_pretrained(
|
67 |
+
train_config = "TTS_models/libritts_xvector_vits/config.yaml",
|
68 |
+
model_file="TTS_models/libritts_xvector_vits/train.total_count.ave_10best.pth",
|
69 |
vocoder_tag=str_or_none(vocoder_tag),
|
70 |
device="cuda",
|
71 |
use_att_constraint=False,
|
|
|
106 |
forward_window=3,
|
107 |
speed_control_alpha=1.0,
|
108 |
)
|
109 |
+
pdb.set_trace()
|
110 |
# from google.cloud import texttospeech
|
111 |
# Google_TTS_client = texttospeech.TextToSpeechClient()
|
112 |
|
vocoders/libritts_hifigan.v1/checkpoint-2500000steps.pkl
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:859f68c82afdf0f22746db6a6b7fb3cee3010a94db09f831c9e5f1a41eb9b0b0
|
3 |
+
size 1004606893
|
vocoders/libritts_hifigan.v1/config.yml
ADDED
@@ -0,0 +1,191 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
allow_cache: false
|
2 |
+
batch_max_steps: 8400
|
3 |
+
batch_size: 16
|
4 |
+
config: conf/hifigan.v1.yaml
|
5 |
+
dev_dumpdir: dump/dev_clean/norm
|
6 |
+
dev_feats_scp: null
|
7 |
+
dev_segments: null
|
8 |
+
dev_wav_scp: null
|
9 |
+
discriminator_adv_loss_params:
|
10 |
+
average_by_discriminators: false
|
11 |
+
discriminator_grad_norm: -1
|
12 |
+
discriminator_optimizer_params:
|
13 |
+
betas:
|
14 |
+
- 0.5
|
15 |
+
- 0.9
|
16 |
+
lr: 0.0002
|
17 |
+
weight_decay: 0.0
|
18 |
+
discriminator_optimizer_type: Adam
|
19 |
+
discriminator_params:
|
20 |
+
follow_official_norm: true
|
21 |
+
period_discriminator_params:
|
22 |
+
bias: true
|
23 |
+
channels: 32
|
24 |
+
downsample_scales:
|
25 |
+
- 3
|
26 |
+
- 3
|
27 |
+
- 3
|
28 |
+
- 3
|
29 |
+
- 1
|
30 |
+
in_channels: 1
|
31 |
+
kernel_sizes:
|
32 |
+
- 5
|
33 |
+
- 3
|
34 |
+
max_downsample_channels: 1024
|
35 |
+
nonlinear_activation: LeakyReLU
|
36 |
+
nonlinear_activation_params:
|
37 |
+
negative_slope: 0.1
|
38 |
+
out_channels: 1
|
39 |
+
use_spectral_norm: false
|
40 |
+
use_weight_norm: true
|
41 |
+
periods:
|
42 |
+
- 2
|
43 |
+
- 3
|
44 |
+
- 5
|
45 |
+
- 7
|
46 |
+
- 11
|
47 |
+
scale_discriminator_params:
|
48 |
+
bias: true
|
49 |
+
channels: 128
|
50 |
+
downsample_scales:
|
51 |
+
- 4
|
52 |
+
- 4
|
53 |
+
- 4
|
54 |
+
- 4
|
55 |
+
- 1
|
56 |
+
in_channels: 1
|
57 |
+
kernel_sizes:
|
58 |
+
- 15
|
59 |
+
- 41
|
60 |
+
- 5
|
61 |
+
- 3
|
62 |
+
max_downsample_channels: 1024
|
63 |
+
max_groups: 16
|
64 |
+
nonlinear_activation: LeakyReLU
|
65 |
+
nonlinear_activation_params:
|
66 |
+
negative_slope: 0.1
|
67 |
+
out_channels: 1
|
68 |
+
scale_downsample_pooling: AvgPool1d
|
69 |
+
scale_downsample_pooling_params:
|
70 |
+
kernel_size: 4
|
71 |
+
padding: 2
|
72 |
+
stride: 2
|
73 |
+
scales: 3
|
74 |
+
discriminator_scheduler_params:
|
75 |
+
gamma: 0.5
|
76 |
+
milestones:
|
77 |
+
- 200000
|
78 |
+
- 400000
|
79 |
+
- 600000
|
80 |
+
- 800000
|
81 |
+
discriminator_scheduler_type: MultiStepLR
|
82 |
+
discriminator_train_start_steps: 0
|
83 |
+
discriminator_type: HiFiGANMultiScaleMultiPeriodDiscriminator
|
84 |
+
distributed: false
|
85 |
+
eval_interval_steps: 1000
|
86 |
+
feat_match_loss_params:
|
87 |
+
average_by_discriminators: false
|
88 |
+
average_by_layers: false
|
89 |
+
include_final_outputs: false
|
90 |
+
fft_size: 2048
|
91 |
+
fmax: 7600
|
92 |
+
fmin: 80
|
93 |
+
format: hdf5
|
94 |
+
generator_adv_loss_params:
|
95 |
+
average_by_discriminators: false
|
96 |
+
generator_grad_norm: -1
|
97 |
+
generator_optimizer_params:
|
98 |
+
betas:
|
99 |
+
- 0.5
|
100 |
+
- 0.9
|
101 |
+
lr: 0.0002
|
102 |
+
weight_decay: 0.0
|
103 |
+
generator_optimizer_type: Adam
|
104 |
+
generator_params:
|
105 |
+
bias: true
|
106 |
+
channels: 512
|
107 |
+
in_channels: 80
|
108 |
+
kernel_size: 7
|
109 |
+
nonlinear_activation: LeakyReLU
|
110 |
+
nonlinear_activation_params:
|
111 |
+
negative_slope: 0.1
|
112 |
+
out_channels: 1
|
113 |
+
resblock_dilations:
|
114 |
+
- - 1
|
115 |
+
- 3
|
116 |
+
- 5
|
117 |
+
- - 1
|
118 |
+
- 3
|
119 |
+
- 5
|
120 |
+
- - 1
|
121 |
+
- 3
|
122 |
+
- 5
|
123 |
+
resblock_kernel_sizes:
|
124 |
+
- 3
|
125 |
+
- 7
|
126 |
+
- 11
|
127 |
+
upsample_kernal_sizes:
|
128 |
+
- 10
|
129 |
+
- 10
|
130 |
+
- 8
|
131 |
+
- 6
|
132 |
+
upsample_scales:
|
133 |
+
- 5
|
134 |
+
- 5
|
135 |
+
- 4
|
136 |
+
- 3
|
137 |
+
use_additional_convs: true
|
138 |
+
use_weight_norm: true
|
139 |
+
generator_scheduler_params:
|
140 |
+
gamma: 0.5
|
141 |
+
milestones:
|
142 |
+
- 200000
|
143 |
+
- 400000
|
144 |
+
- 600000
|
145 |
+
- 800000
|
146 |
+
generator_scheduler_type: MultiStepLR
|
147 |
+
generator_train_start_steps: 1
|
148 |
+
generator_type: HiFiGANGenerator
|
149 |
+
global_gain_scale: 1.0
|
150 |
+
hop_size: 300
|
151 |
+
lambda_adv: 1.0
|
152 |
+
lambda_aux: 45.0
|
153 |
+
lambda_feat_match: 2.0
|
154 |
+
log_interval_steps: 100
|
155 |
+
mel_loss_params:
|
156 |
+
fft_size: 2048
|
157 |
+
fmax: 12000
|
158 |
+
fmin: 0
|
159 |
+
fs: 24000
|
160 |
+
hop_size: 300
|
161 |
+
log_base: null
|
162 |
+
num_mels: 80
|
163 |
+
win_length: 1200
|
164 |
+
window: hann
|
165 |
+
num_mels: 80
|
166 |
+
num_save_intermediate_results: 4
|
167 |
+
num_workers: 2
|
168 |
+
outdir: exp/train_nodev_clean_libritts_hifigan.v1
|
169 |
+
pin_memory: true
|
170 |
+
pretrain: ''
|
171 |
+
rank: 0
|
172 |
+
remove_short_samples: false
|
173 |
+
resume: exp/train_nodev_clean_libritts_hifigan.v1/checkpoint-1890000steps.pkl
|
174 |
+
sampling_rate: 24000
|
175 |
+
save_interval_steps: 10000
|
176 |
+
train_dumpdir: dump/train_nodev_clean/norm
|
177 |
+
train_feats_scp: null
|
178 |
+
train_max_steps: 2500000
|
179 |
+
train_segments: null
|
180 |
+
train_wav_scp: null
|
181 |
+
trim_frame_size: 1024
|
182 |
+
trim_hop_size: 256
|
183 |
+
trim_silence: false
|
184 |
+
trim_threshold_in_db: 20
|
185 |
+
use_feat_match_loss: true
|
186 |
+
use_mel_loss: true
|
187 |
+
use_stft_loss: false
|
188 |
+
verbose: 1
|
189 |
+
version: 0.5.1
|
190 |
+
win_length: 1200
|
191 |
+
window: hann
|
vocoders/libritts_hifigan.v1/stats.h5
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:94d55b1ea02f33cb24aa5d1fbc4087519d36ca89ca032e3dc0ca6627df4ed190
|
3 |
+
size 4736
|
vocoders/vctk_parallel_wavegan.v1.long/._checkpoint-1000000steps.pkl
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:ccb8cfc739515054284e6ac7c75afdab0c771eba7d132c4e19efff528147a1a1
|
3 |
+
size 223
|
vocoders/vctk_parallel_wavegan.v1.long/._config.yml
ADDED
Binary file (223 Bytes). View file
|
|
vocoders/vctk_parallel_wavegan.v1.long/._stats.h5
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f688626155bff6853a6045408be9bb248828abc482e2218ee0d93183cede5062
|
3 |
+
size 223
|
vocoders/vctk_parallel_wavegan.v1.long/._train_nodev_all_vctk_parallel_wavegan.v1.long
ADDED
Binary file (187 Bytes). View file
|
|
vocoders/vctk_parallel_wavegan.v1.long/checkpoint-1000000steps.pkl
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:c3e92516432be8da4dbcc9e284b0e8f5c0273fd8d7be939ca4c0db3f7c8c73e7
|
3 |
+
size 17498016
|
vocoders/vctk_parallel_wavegan.v1.long/config.yml
ADDED
@@ -0,0 +1,104 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
allow_cache: true
|
2 |
+
batch_max_steps: 24000
|
3 |
+
batch_size: 6
|
4 |
+
config: conf/parallel_wavegan.v1.long.yaml
|
5 |
+
dev_dumpdir: dump/dev_all/norm
|
6 |
+
dev_feats_scp: null
|
7 |
+
dev_segments: null
|
8 |
+
dev_wav_scp: null
|
9 |
+
discriminator_grad_norm: 1
|
10 |
+
discriminator_optimizer_params:
|
11 |
+
eps: 1.0e-06
|
12 |
+
lr: 5.0e-05
|
13 |
+
weight_decay: 0.0
|
14 |
+
discriminator_params:
|
15 |
+
bias: true
|
16 |
+
conv_channels: 64
|
17 |
+
in_channels: 1
|
18 |
+
kernel_size: 3
|
19 |
+
layers: 10
|
20 |
+
nonlinear_activation: LeakyReLU
|
21 |
+
nonlinear_activation_params:
|
22 |
+
negative_slope: 0.2
|
23 |
+
out_channels: 1
|
24 |
+
use_weight_norm: true
|
25 |
+
discriminator_scheduler_params:
|
26 |
+
gamma: 0.5
|
27 |
+
step_size: 200000
|
28 |
+
discriminator_train_start_steps: 100000
|
29 |
+
distributed: false
|
30 |
+
eval_interval_steps: 1000
|
31 |
+
fft_size: 2048
|
32 |
+
fmax: 7600
|
33 |
+
fmin: 80
|
34 |
+
format: hdf5
|
35 |
+
generator_grad_norm: 10
|
36 |
+
generator_optimizer_params:
|
37 |
+
eps: 1.0e-06
|
38 |
+
lr: 0.0001
|
39 |
+
weight_decay: 0.0
|
40 |
+
generator_params:
|
41 |
+
aux_channels: 80
|
42 |
+
aux_context_window: 2
|
43 |
+
dropout: 0.0
|
44 |
+
gate_channels: 128
|
45 |
+
in_channels: 1
|
46 |
+
kernel_size: 3
|
47 |
+
layers: 30
|
48 |
+
out_channels: 1
|
49 |
+
residual_channels: 64
|
50 |
+
skip_channels: 64
|
51 |
+
stacks: 3
|
52 |
+
upsample_net: ConvInUpsampleNetwork
|
53 |
+
upsample_params:
|
54 |
+
upsample_scales:
|
55 |
+
- 4
|
56 |
+
- 5
|
57 |
+
- 3
|
58 |
+
- 5
|
59 |
+
use_weight_norm: true
|
60 |
+
generator_scheduler_params:
|
61 |
+
gamma: 0.5
|
62 |
+
step_size: 200000
|
63 |
+
global_gain_scale: 1.0
|
64 |
+
hop_size: 300
|
65 |
+
lambda_adv: 4.0
|
66 |
+
log_interval_steps: 100
|
67 |
+
num_mels: 80
|
68 |
+
num_save_intermediate_results: 4
|
69 |
+
num_workers: 2
|
70 |
+
outdir: exp/train_nodev_all_vctk_parallel_wavegan.v1.long
|
71 |
+
pin_memory: true
|
72 |
+
pretrain: ''
|
73 |
+
rank: 0
|
74 |
+
remove_short_samples: true
|
75 |
+
resume: exp/train_nodev_all_vctk_parallel_wavegan.v1.long/checkpoint-970000steps.pkl
|
76 |
+
sampling_rate: 24000
|
77 |
+
save_interval_steps: 5000
|
78 |
+
stft_loss_params:
|
79 |
+
fft_sizes:
|
80 |
+
- 1024
|
81 |
+
- 2048
|
82 |
+
- 512
|
83 |
+
hop_sizes:
|
84 |
+
- 120
|
85 |
+
- 240
|
86 |
+
- 50
|
87 |
+
win_lengths:
|
88 |
+
- 600
|
89 |
+
- 1200
|
90 |
+
- 240
|
91 |
+
window: hann_window
|
92 |
+
train_dumpdir: dump/train_nodev_all/norm
|
93 |
+
train_feats_scp: null
|
94 |
+
train_max_steps: 1000000
|
95 |
+
train_segments: null
|
96 |
+
train_wav_scp: null
|
97 |
+
trim_frame_size: 1024
|
98 |
+
trim_hop_size: 256
|
99 |
+
trim_silence: false
|
100 |
+
trim_threshold_in_db: 20
|
101 |
+
verbose: 1
|
102 |
+
version: 0.3.4
|
103 |
+
win_length: 1200
|
104 |
+
window: hann
|
vocoders/vctk_parallel_wavegan.v1.long/stats.h5
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:3144070fcf475d49873d9ee9c5433ae222587e569d38dd6fee8c6fff91d69c94
|
3 |
+
size 4736
|