forked from Andong-Li-speech/EaBNet
-
Notifications
You must be signed in to change notification settings - Fork 0
/
train.py
177 lines (153 loc) · 8.4 KB
/
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
import numpy as np
import torch
import os
import torch.nn as nn
from torch.autograd import Variable
from torch import Tensor
from torchvision import transforms
import torch.utils.data as utils
import pickle
from dataset import make_dataset
import tqdm
from EaBNet import EaBNet, numParams, com_mag_mse_loss
from dataset.custom_dataset import CustomAudioVisualDataset
from torch.utils.tensorboard import SummaryWriter
#from torch.utils.tensorboard import SummaryWriter
def main(args):
if args.fixed_seed:
seed = 1
np.random.seed(seed)
torch.backends.cudnn.deterministic = True
torch.backends.cudnn.benchmark = False
torch.manual_seed(seed)
if torch.cuda.is_available():
torch.cuda.manual_seed_all(seed)
net = EaBNet(k1=args.k1,
k2=args.k2,
c=args.c,
M=args.M,
embed_dim=args.embed_dim,
kd1=args.kd1,
cd1=args.cd1,
d_feat=args.d_feat,
p=args.p,
q=args.q,
is_causal=args.is_causal,
is_u2=args.is_u2,
bf_type=args.bf_type,
topo_type=args.topo_type,
intra_connect=args.intra_connect,
norm_type=args.norm_type,
)#.cuda()
net.train()
print("The number of trainable parameters is:{}".format(numParams(net)))
from ptflops.flops_counter import get_model_complexity_info
#get_model_complexity_info(net, (101, 161, 9, 2))
batch_size = args.batch_size
mics = args.mics
sr = args.sr
wav_len = int(args.wav_len * sr)
win_size = int(args.win_size * sr)
win_shift = int(args.win_shift * sr)
fft_num = args.fft_num
dataloader = torch.utils.data.DataLoader(
dataset=torch.randn(args.batch_size, wav_len, args.mics),
batch_size=args.batch_size,
shuffle=False,
num_workers=0,
drop_last=False,
pin_memory=True,
)
tr_dataset = make_dataset(args) #pass
#build data loader from dataset
dataloader = utils.DataLoader(tr_dataset, args.batch_size, shuffle=True, pin_memory=True)
loss = nn.L1Loss()
optimizer = torch.optim.Adam(net.parameters(), lr=5e-4)
#training loop
for epoch in range(args.total_epoch):
for i, (x, target) in enumerate(dataloader):
current_iter = epoch * len(dataloader) + i
optimizer.zero_grad()
noisy_wav = x.cuda() #[4, 4, 76672]
#target_wav = torch.rand(args.batch_size, wav_len).cuda()
target_wav = target.cuda() #[4, 1, 76672]
noisy_wav = noisy_wav.transpose(-2, -1).contiguous().view(batch_size*mics, wav_len) #[batch_size*mics, wav_len]
noisy_stft = torch.stft(noisy_wav, fft_num, win_shift, win_size, torch.hann_window(win_size).to(noisy_wav.device))
target_stft = torch.stft(target_wav, fft_num, win_shift, win_size, torch.hann_window(win_size).to(target_wav.device))
_, freq_num, seq_len, _ = noisy_stft.shape
noisy_stft = noisy_stft.view(batch_size, mics, freq_num, seq_len, -1).permute(0, 3, 2, 1, 4).cuda()
target_stft = target_stft.permute(0, 3, 2, 1).cuda()
# conduct sqrt power-compression
noisy_mag, noisy_phase = torch.norm(noisy_stft, dim=-1) ** 0.5, torch.atan2(noisy_stft[..., -1], noisy_stft[..., 0])
target_mag, target_phase = torch.norm(target_stft, dim=1) ** 0.5, torch.atan2(target_stft[:, -1, ...], target_stft[:, 0, ...])
noisy_stft = torch.stack((noisy_mag * torch.cos(noisy_phase), noisy_mag * torch.sin(noisy_phase)), dim=-1).cuda()
target_stft = torch.stack((target_mag * torch.cos(target_phase), target_mag * torch.sin(target_phase)), dim=1).cuda()
esti_stft = net(noisy_stft)
#calculate loss
#l = com_mag_mse_loss(esti_stft, target_stft, frame_list)
l = loss(esti_stft, target_stft)
print('loss:', l.item())
l.backward()
optimizer.step()
if current_iter % 100 == 0:
print('iter:', current_iter)
print('loss:', l.item())
print('input:', x.shape)
print('noisy_wav:', noisy_wav.shape)
print('noisy_stft:', noisy_stft.shape)
print('esti_stft:', esti_stft.shape)
print('target_stft:', target_stft.shape)
break
break
if __name__ == '__main__':
import argparse
parser = argparse.ArgumentParser("This script provides the network code and a simple testing, you can train the"
"network according to your own pipeline")
#eabnet parameters
parser.add_argument("--batch_size", type=int, default=4)
parser.add_argument("--total_epoch", type=int, default=100)
parser.add_argument("--mics", type=int, default=9)
parser.add_argument("--sr", type=int, default=16000)
parser.add_argument("--wav_len", type=float, default=6.0)
parser.add_argument("--win_size", type=float, default=0.020)
parser.add_argument("--win_shift", type=float, default=0.010)
parser.add_argument("--fft_num", type=int, default=320)
parser.add_argument("--k1", type=tuple, default=(2,3))
parser.add_argument("--k2", type=tuple, default=(1,3))
parser.add_argument("--c", type=int, default=64)
parser.add_argument("--M", type=int, default=9)
parser.add_argument("--embed_dim", type=int, default=64)
parser.add_argument("--kd1", type=int, default=5)
parser.add_argument("--cd1", type=int, default=64)
parser.add_argument("--d_feat", type=int, default=256)
parser.add_argument("--p", type=int, default=6)
parser.add_argument("--q", type=int, default=3)
parser.add_argument("--is_causal", type=bool, default=True, choices=[True, False])
parser.add_argument("--is_u2", type=bool, default=True, choices=[True, False])
parser.add_argument("--bf_type", type=str, default="lstm", choices=["lstm", "cnn"])
parser.add_argument("--topo_type", type=str, default="mimo", choices=["mimo", "miso"])
parser.add_argument("--intra_connect", type=str, default="cat", choices=["cat", "add"])
parser.add_argument("--norm_type", type=str, default="IN", choices=["BN", "IN", "cLN"])
parser.add_argument("--fixed_seed", type=bool, default=False, choices=[True, False])
#dataset parameters processed是4声道，processed1是8声道，但是加载时超内存
processed_folder = 'processed'
parser.add_argument('--training_predictors_path', type=str, default=f'/data/wbh/l3das23/{processed_folder}/task1_predictors_train.pkl')
parser.add_argument('--training_target_path', type=str, default=f'/data/wbh/l3das23/{processed_folder}/task1_target_train.pkl')
parser.add_argument('--validation_predictors_path', type=str, default=f'/data/wbh/l3das23/{processed_folder}/task1_predictors_validation.pkl')
parser.add_argument('--validation_target_path', type=str, default=f'/data/wbh/l3das23/{processed_folder}/task1_target_validation.pkl')
parser.add_argument('--test_predictors_path', type=str, default=f'/data/wbh/l3das23/{processed_folder}/task1_predictors_test.pkl')
parser.add_argument('--test_target_path', type=str, default=f'/data/wbh/l3das23/{processed_folder}/task1_target_test.pkl')
parser.add_argument('--dataset', type=str, default='mcse', choices=['l3das23', 'mcse'])
#saving parameters
parser.add_argument('--results_path', type=str, default='/data/wbh/l3das23/RESULTS/Task1',
help='Folder to write results dicts into')
parser.add_argument('--checkpoint_dir', type=str, default='/data/wbh/l3das23/RESULTS/Task1',
help='Folder to write checkpoints into')
parser.add_argument('--path_images', type=str, default=None,
help="Path to the folder containing all images of Task1. None when using the audio-only version")
parser.add_argument('--path_csv_images_train', type=str, default='/data/wbh/l3das23/L3DAS23_Task1_train/audio_image.csv',
help="Path to the CSV file for the couples (name_audio, name_photo) in the train/val set")
parser.add_argument('--path_csv_images_test', type=str, default='/data/wbh/l3das23/L3DAS23_Task1_dev/audio_image.csv',
help="Path to the CSV file for the couples (name_audio, name_photo)")
args = parser.parse_args()
main(args)