-
Notifications
You must be signed in to change notification settings - Fork 0
/
enhance_base_train.py
206 lines (189 loc) · 11 KB
/
enhance_base_train.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
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
from __future__ import print_function
import argparse
import os
import math
import random
import shutil
import psutil
import time
import numpy as np
from tqdm import tqdm
import torch
import torch.optim as optim
import torch.nn.functional as F
import fake_opt
import scipy.io.wavfile
import librosa
import os
from options.train_options import TrainOptions
from model.enhance_model import EnhanceModel
from data.mix_data_loader import MixSequentialDataset, MixSequentialDataLoader, BucketingSampler
from utils.visualizer import Visualizer
from utils import utils
from rewav import rewav
manualSeed = random.randint(1, 10000)
random.seed(manualSeed)
torch.manual_seed(manualSeed)
torch.cuda.manual_seed(manualSeed)
# def rewav(path,uttid,feats,ang,hop_length = 256,win_length = 512,window = scipy.signal.hamming,rate = 16000):
# if isinstance(feats,torch.Tensor):
# feats_numpy = feats.numpy()
# ang_numpy = ang.numpy()
# else:
# feats_numpy = feats
# ang_numpy = ang
# feat_mat = feats_numpy*np.cos(ang_numpy)+1j*feats_numpy*np.sin(ang_numpy)
# #feat_mat = feat_mat[feat_mat]
# feat_mat = feat_mat.T
# x = librosa.core.istft(feat_mat,hop_length = hop_length,win_length = win_length, window = window)
# x = x*65535
# path_wav = os.path.join(path,"remix_"+uttid+".wav")
# scipy.io.wavfile.write(path_wav,rate,data = x)
def main():
opt = TrainOptions().parse()
if opt.exp_path == None:
opt = fake_opt.Enhance_base_train()
device = torch.device("cuda:{}".format(opt.gpu_ids[0]) if len(opt.gpu_ids) > 0 and torch.cuda.is_available() else "cpu")
visualizer = Visualizer(opt)
logging = visualizer.get_logger()
loss_report = visualizer.add_plot_report(['train/loss', 'val/loss'], 'loss.png')
# data
logging.info("Building dataset.")
train_set = 'train'
val_set = 'dev'
train_dataset = MixSequentialDataset(opt, os.path.join(opt.dataroot, train_set), os.path.join(opt.dict_dir, 'train_units.txt'),train_set)
val_dataset = MixSequentialDataset(opt, os.path.join(opt.dataroot, val_set), os.path.join(opt.dict_dir, 'train_units.txt'),val_set)
train_sampler = BucketingSampler(train_dataset, batch_size=opt.batch_size)
train_loader = MixSequentialDataLoader(train_dataset, num_workers=opt.num_workers, batch_sampler=train_sampler)
val_loader = MixSequentialDataLoader(val_dataset, batch_size=int(opt.batch_size/2), num_workers=opt.num_workers, shuffle=False)
opt.idim = train_dataset.get_feat_size()
opt.odim = train_dataset.get_num_classes()
opt.char_list = train_dataset.get_char_list()
opt.train_dataset_len = len(train_dataset)
logging.info('#input dims : ' + str(opt.idim))
logging.info('#output dims: ' + str(opt.odim))
logging.info("Dataset ready!")
save_wav_path = os.path.join(opt.checkpoints_dir,opt.name,'att_ws')
lr = opt.lr
eps = opt.eps
iters = opt.iters
best_loss = opt.best_loss
start_epoch = opt.start_epoch
model_path = None
if opt.enhance_resume:
model_path = os.path.join(opt.works_dir, opt.enhance_resume)
if os.path.isfile(model_path):
package = torch.load(model_path, map_location=lambda storage, loc: storage)
lr = package.get('lr', opt.lr)
eps = package.get('eps', opt.eps)
best_loss = package.get('best_loss', float('inf'))
start_epoch = int(package.get('epoch', 0))
iters = int(package.get('iters', 0))
loss_report = package.get('loss_report', loss_report)
visualizer.set_plot_report(loss_report, 'loss.png')
print('package found at {} and start_epoch {} iters {}'.format(model_path, start_epoch, iters))
else:
print("no checkpoint found at {}".format(model_path))
enhance_model = EnhanceModel.load_model(model_path, 'enhance_state_dict', opt)
# Setup an optimizer
enhance_parameters = filter(lambda p: p.requires_grad, enhance_model.parameters())
if opt.opt_type == 'adadelta':
enhance_optimizer = torch.optim.Adadelta(enhance_parameters, rho=0.95, eps=opt.eps)
elif opt.opt_type == 'adam':
enhance_optimizer = torch.optim.Adam(enhance_parameters, lr=opt.lr, betas=(opt.beta1, 0.999))
# Training
for epoch in range(start_epoch, opt.epochs):
enhance_model.train()
if epoch > opt.shuffle_epoch:
print("Shuffling batches for the following epochs")
train_sampler.shuffle(epoch)
print("finish")
for i, (data) in enumerate(train_loader, start=(iters % len(train_dataset))):
utt_ids, spk_ids, clean_inputs, clean_log_inputs, mix_inputs, mix_log_inputs, cos_angles, targets, input_sizes, target_sizes,mix_angles,clean_angles,cmvn = data
utt_id = utt_ids[0]
# clean_input = clean_inputs[0].data
# cos_angle = cos_angles[0].data
# mix = mix_inputs[0].data
# path_wav = '/usr/home/shi/projects/data_aishell/data/wavfile'
# rewav(path_wav,utt_id,mix,cos_angle)
loss, enhance_out = enhance_model( mix_inputs, mix_log_inputs, input_sizes, clean_inputs,cos_angles)
enhance_optimizer.zero_grad() # Clear the parameter gradients
loss.backward()
# compute the gradient norm to check if it is normal or not
grad_norm = torch.nn.utils.clip_grad_norm_(enhance_model.parameters(), opt.grad_clip)
if math.isnan(grad_norm):
logging.warning('grad norm is nan. Do not update model.')
else:
enhance_optimizer.step()
iters += 1
errors = {'train/loss': loss.item()}
visualizer.set_current_errors(errors)
if iters % opt.print_freq == 0:
visualizer.print_current_errors(epoch, iters)
state = {'enhance_state_dict': enhance_model.state_dict(), 'opt': opt,
'epoch': epoch, 'iters': iters, 'eps': eps, 'lr': lr,
'best_loss': best_loss, 'loss_report': loss_report}
filename='latest'
utils.save_checkpoint(state, opt.exp_path, filename=filename)
if iters % opt.validate_freq == 0:
enhance_model.eval()
torch.set_grad_enabled(False)
num_saved_specgram = 0
for i, (data) in tqdm(enumerate(val_loader, start=0)):
utt_ids, spk_ids, clean_inputs, clean_log_inputs, mix_inputs, mix_log_inputs, cos_angles, targets, input_sizes, target_sizes, clean_angles, mix_angles, cmvn = data
loss, enhance_out = enhance_model(mix_inputs, mix_log_inputs, input_sizes, clean_inputs, cos_angles)
errors = {'val/loss': loss.item()}
visualizer.set_current_errors(errors)
if opt.num_saved_specgram > 0:
if num_saved_specgram < opt.num_saved_specgram:
enhanced_outs = enhance_model.calculate_all_specgram(mix_inputs, mix_log_inputs, input_sizes)
for x in range(len(utt_ids)):
enhanced_out = enhanced_outs[x].data.cpu().numpy()
enhanced_out[enhanced_out <= 1e-7] = 1e-7
enhance_out_orig = enhanced_out
enhanced_out = np.log10(enhanced_out)
clean_input = clean_inputs[x].data.cpu().numpy()
clean_input[clean_input <= 1e-7] = 1e-7
clean_input_orig = clean_input
clean_input = np.log10(clean_input)
mix_input = mix_inputs[x].data.cpu().numpy()
mix_input[mix_input <= 1e-7] = 1e-7
mix_input_orig = mix_input
mix_input = np.log10(mix_input)
utt_id = utt_ids[x]
mix_angle = mix_angles[x].data.cpu().numpy()
input_size = int(input_sizes[x])
wav_name_mix = "{}_mix.wav".format(utt_id)
if not os.path.isfile(os.path.join(save_wav_path,wav_name_mix)):
rewav(save_wav_path,utt_id,mix_input_orig,mix_angle,input_size = input_size,wav_file = wav_name_mix)
#flag = 1
wav_name_enhance = "{}_ep{}_it{}_enhance.wav".format(utt_id, epoch, iters)
file_name = "{}_ep{}_it{}.png".format(utt_id, epoch, iters)
rewav(save_wav_path,utt_id,enhance_out_orig,mix_angle,input_size = input_size,wav_file = wav_name_enhance)
visualizer.plot_specgram(clean_input, mix_input, enhanced_out, input_size, file_name)
num_saved_specgram += 1
if num_saved_specgram >= opt.num_saved_specgram:
break
enhance_model.train()
torch.set_grad_enabled(True)
visualizer.print_epoch_errors(epoch, iters)
loss_report = visualizer.plot_epoch_errors(epoch, iters, 'loss.png')
train_loss = visualizer.get_current_errors('train/loss')
val_loss = visualizer.get_current_errors('val/loss')
filename = None
if val_loss > best_loss:
print('val_loss {} > best_loss {}'.format(val_loss, best_loss))
eps = utils.adadelta_eps_decay(optimizer, opt.eps_decay)
else:
filename='model.loss.best'
best_loss = min(val_loss, best_loss)
print('best_loss {}'.format(best_loss))
state = {'enhance_state_dict': enhance_model.state_dict(), 'opt': opt,
'epoch': epoch, 'iters': iters, 'eps': eps, 'lr': lr,
'best_loss': best_loss, 'loss_report': loss_report}
##filename='epoch-{}_iters-{}_loss-{:.6f}-{:.6f}.pth'.format(epoch, iters, train_loss, val_loss)
utils.save_checkpoint(state, opt.exp_path, filename=filename)
visualizer.reset()
if __name__ == '__main__':
main()
# %%