-
Notifications
You must be signed in to change notification settings - Fork 15
/
pengi.py
191 lines (162 loc) · 6.93 KB
/
pengi.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
import sys
sys.path.append('')
import torch
import torch.nn.functional as F
from torch import nn
from transformers import AutoConfig, AutoModel
import os
from models.audio import get_audio_encoder
from models.decoder import get_decoder
def init_layer(layer):
"""Initialize a Linear or Convolutional layer. """
nn.init.xavier_uniform_(layer.weight)
if hasattr(layer, 'bias'):
if layer.bias is not None:
layer.bias.data.fill_(0.)
def init_bn(bn):
"""Initialize a Batchnorm layer. """
bn.bias.data.fill_(0.)
bn.weight.data.fill_(1.)
def weights_init(m):
if isinstance(m, nn.Conv2d) or isinstance(m, nn.Linear):
nn.init.xavier_uniform_(m.weight)
if hasattr(m, 'bias'):
if m.bias is not None:
m.bias.data.fill_(0.)
elif isinstance(m, nn.BatchNorm2d) or isinstance(m, nn.BatchNorm1d):
"""Initialize a Batchnorm layer. """
m.bias.data.fill_(0.)
m.weight.data.fill_(1.)
class Projection(nn.Module):
def __init__(self, d_in: int, d_out: int, p: float=0.5) -> None:
super().__init__()
self.linear1 = nn.Linear(d_in, d_out, bias=False)
self.linear2 = nn.Linear(d_out, d_out, bias=False)
self.layer_norm = nn.LayerNorm(d_out)
self.drop = nn.Dropout(p)
self.init_weight()
def init_weight(self):
init_layer(self.linear1)
init_layer(self.linear2)
init_bn(self.layer_norm)
def forward(self, x: torch.Tensor) -> torch.Tensor:
embed1 = self.linear1(x)
embed2 = self.drop(self.linear2(F.gelu(embed1)))
embeds = self.layer_norm(embed1 + embed2)
return embeds
class AudioEncoder(nn.Module):
def __init__(self, audioenc_name:str, d_in: int, d_out: int, sample_rate: int, window_size: int,
hop_size: int, mel_bins: int, fmin: int, fmax: int, classes_num: int,
specaug: bool, mixup: bool, use_pretrained_audioencoder: bool, freeze_audio_encoder_weights: bool,
use_precomputed_melspec: bool, pretrained_audioencoder_path: str) -> None:
super().__init__()
audio_encoder, pretrained_emb_size = get_audio_encoder(audioenc_name)
if use_pretrained_audioencoder:
classes_num = 527
d_in = pretrained_emb_size
self.base = audio_encoder(
sample_rate, window_size,
hop_size, mel_bins, fmin, fmax,
classes_num, d_in,
specaug, mixup, use_precomputed_melspec)
self.projection = Projection(pretrained_emb_size if use_pretrained_audioencoder else d_in, d_out)