-
Notifications
You must be signed in to change notification settings - Fork 5
/
Copy pathsynthesize.py
168 lines (160 loc) · 7.89 KB
/
synthesize.py
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
import argparse
from musa.datasets import *
from musa.models import *
from torch.utils.data import DataLoader
from tensorboardX import SummaryWriter
from torch.optim.lr_scheduler import ReduceLROnPlateau
import torch.optim as optim
from musa.core import *
import random
import json
import os
def main(opts):
if not opts.force_dur and opts.dur_model is None:
raise ValueError('Please specify dur_model')
if opts.aco_model is None:
raise ValueError('Please specify aco_model')
assert opts.model_cfg is not None
with open(opts.model_cfg, 'r') as mcfg_f:
mcfg = json.load(mcfg_f)
device = 'cpu'
if opts.cuda and torch.cuda.is_available():
device = 'cuda'
with open(opts.cfg_spk, 'rb') as cfg_f:
cfg = pickle.load(cfg_f)
idx2spk = {}
spk2durstats = {}
spk2acostats = {}
for spk_id, spk_cfg in cfg.items():
if 'idx' in spk_cfg:
idx2spk[int(spk_cfg['idx'])] = spk_id
if 'dur_stats' in spk_cfg:
spk2durstats[int(spk_cfg['idx'])] = spk_cfg['dur_stats']
if 'aco_stats' in spk_cfg:
spk2acostats[int(spk_cfg['idx'])] = spk_cfg['aco_stats']
with open(opts.codebooks_dir, 'rb') as cbooks_f:
# read cbooks lengths to get ling_feats_dim
cbooks = pickle.load(cbooks_f)
ling_feats_dim = 0
for k, v in cbooks.items():
if 'mean' in v:
# real value
ling_feats_dim += 1
else:
# categorical value
ling_feats_dim += len(v)
# 6 boolean factors
ling_feats_dim += 6
print('Found ling_feats_dim: ', ling_feats_dim)
if not opts.force_dur:
print('-' * 30)
print('Loading duration model: ', opts.dur_model)
dur_model = torch.load(opts.dur_model,
map_location=lambda storage, loc: storage)
print('[*] Loaded')
else:
print('[!] Dur model NOT loaded')
dur_model = None
print('>> spk2durstats: ', json.dumps(spk2durstats, indent=2))
# build acoustic model and load weights
print('-' * 30)
print('Loading acoustic model: ', opts.aco_model)
aco_model = torch.load(opts.aco_model,
map_location=lambda storage, loc: storage)
print('[*] Loaded')
#aco_model.load(opts.aco_model)
print('>> idx2spk: ', json.dumps(idx2spk, indent=2))
if opts.cuda:
aco_model.to(device)
if not opts.force_dur:
dur_model.to(device)
print('aco_model: ', aco_model)
# get lab file basename
lab_fname = os.path.basename(opts.synthesize_lab)
lab_bname, _ = os.path.splitext(lab_fname)
if mcfg['model_type'] in ['satt', 'decsatt']:
att_synthesize(dur_model, aco_model, opts.spk_id, spk2durstats, spk2acostats,
opts.save_path, lab_bname, opts.codebooks_dir, opts.synthesize_lab,
cuda=opts.cuda, force_dur=opts.force_dur, pf=opts.pf)
else:
synthesize(dur_model, aco_model, opts.spk_id, spk2durstats, spk2acostats,
opts.save_path, lab_bname, opts.codebooks_dir, opts.synthesize_lab,
cuda=opts.cuda, force_dur=opts.force_dur, pf=opts.pf)
if __name__ == '__main__':
parser = argparse.ArgumentParser()
parser.add_argument('--cfg_spk', type=str, default='cfg/tcstar.cfg')
parser.add_argument('--spk_id', type=int, default=0)
parser.add_argument('--lab_dir', type=str, default='data/tcstar/lab')
parser.add_argument('--aco_dir', type=str, default='data/tcstar/aco')
parser.add_argument('--synthesize_lab', type=str, default=None,
help='Lab filename to be synthesized')
parser.add_argument('--codebooks_dir', type=str,
default='data/tcstar/codebooks.pkl')
parser.add_argument('--pf', type=float, default=1)
parser.add_argument('--save_path', type=str, default='ckpt')
parser.add_argument('--force-gen', action='store_true',
default=False)
parser.add_argument('--force-dur', action='store_true',
default=False)
parser.add_argument('--train-dur', action='store_true',
default=False,
help='Flag to specify that we train '
'a dur model.')
parser.add_argument('--train-aco', action='store_true',
default=False,
help='Flag to specify that we train '
'an aco model.')
parser.add_argument('--dur_max_samples', type=int, default=None,
help='Max samples per speaker in dur loader')
parser.add_argument('--aco_max_samples', type=int, default=None,
help='Max samples per speaker in aco loader')
parser.add_argument('--dur_rnn_size', type=int, default=256)
parser.add_argument('--dur_rnn_layers', type=int, default=1)
parser.add_argument('--dur_emb_size', type=int, default=256)
parser.add_argument('--dur_emb_layers', type=int, default=1)
parser.add_argument('--aco_rnn_size', type=int, default=512)
parser.add_argument('--aco_rnn_layers', type=int, default=2)
parser.add_argument('--aco_emb_size', type=int, default=256)
parser.add_argument('--aco_emb_layers', type=int, default=2)
parser.add_argument('--dur_q_classes', type=int, default=None,
help='Num of clusters in dur quantization. '
'If specified, this will triger '
'quantization in dloader and softmax '
'output for the model (Def: None).')
#parser.add_argument('--dur_weights', type=str, default=None,
# help='Trained dur model weights')
parser.add_argument('--dur_model', type=str, default=None,
help='Trained dur model')
parser.add_argument('--aco_model', type=str, default=None,
help='Trained aco model')
#parser.add_argument('--aco_weights', type=str, default=None,
# help='Trained aco model weights')
parser.add_argument('--aco_q_classes', type=int, default=None,
help='Num of clusters in aco quantization. '
'If specified, this will triger '
'quantization in dloader and softmax '
'output for the model (Def: None).')
parser.add_argument('--loss', type=str, default='MSELoss',
help='Options: PyTorch losses (Def: MSELoss)')
parser.add_argument('--patience', type=int, default=5)
parser.add_argument('--seed', type=int, default=1991)
parser.add_argument('--aco_bnorm', action='store_true', default=False)
parser.add_argument('--aco_max_seq_len', type=int, default=None)
parser.add_argument('--loader_workers', type=int, default=2)
parser.add_argument('--parser_workers', type=int, default=4)
parser.add_argument('--cuda', default=False, action='store_true')
parser.add_argument('--dur_mulout', default=False, action='store_true')
parser.add_argument('--aco_mulout', default=False, action='store_true')
parser.add_argument('--exclude_train_spks', type=str, default=[], nargs='+')
parser.add_argument('--exclude_eval_spks', type=str, default=[], nargs='+')
parser.add_argument('--model_cfg', type=str, default=None)
opts = parser.parse_args()
print('Parsed opts: ', json.dumps(vars(opts), indent=2))
if not os.path.exists(opts.save_path):
os.makedirs(opts.save_path)
torch.manual_seed(opts.seed)
np.random.seed(opts.seed)
random.seed(opts.seed)
if opts.cuda:
torch.cuda.manual_seed(opts.seed)
main(opts)