File size: 24,035 Bytes
11eafe6
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
997d9c0
11eafe6
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
4a1f5f8
11eafe6
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
import math
from typing import List, Union

import numpy as np
import torch
from einops import rearrange
from torch import nn
from torch.nn import functional as F
from torch.nn.utils import weight_norm

from audiotools import AudioSignal, STFTParams, ml
from audiotools.ml import BaseModel
from base import CodecMixin
from layers import WNConv1d, WNConvTranspose1d, get_activation


def init_weights(m, mean=0.0, std=0.02, init_type="xavier", gain=0.02):
    """
    Initialize weights of the entire model using xavier_normal_ or kaiming_normal_.
    Args:
        m (nn.Module): The module to initialize.
        mean (float): Mean for weight initialization.
        std (float): Standard deviation for weight initialization.
        init_type (str): Type of initialization ('xavier' or 'kaiming').
        gain (float): Gain for xavier initialization.
    """
    classname = m.__class__.__name__

    if init_type == "xavier":
        # Handle convolutional layers
        if "Depthwise_Separable" in classname:
            nn.init.xavier_normal_(m.depth_conv.weight.data, gain=gain)
            nn.init.xavier_normal_(m.point_conv.weight.data, gain=gain)
            if hasattr(m.depth_conv, "bias") and m.depth_conv.bias is not None:
                nn.init.zeros_(m.depth_conv.bias.data)
            if hasattr(m.point_conv, "bias") and m.point_conv.bias is not None:
                nn.init.zeros_(m.point_conv.bias.data)
        elif classname.find("Conv") != -1:
            nn.init.xavier_normal_(m.weight.data, gain=gain)
            if hasattr(m, "bias") and m.bias is not None:
                nn.init.zeros_(m.bias.data)

        # Handle batch normalization layers
        elif classname.find("BatchNorm") != -1:
            if hasattr(m, "weight") and m.weight is not None:
                nn.init.xavier_normal_(m.weight.data, gain=gain)
            if hasattr(m, "bias") and m.bias is not None:
                nn.init.zeros_(m.bias.data)

        # Handle custom layers like Snake1d and SnakeBeta
        elif classname == "Snake1d":
            if hasattr(m, "alpha") and m.alpha is not None:
                if m.alpha.data.dim() >= 2:
                    nn.init.xavier_normal_(m.alpha.data, gain=gain)
                else:
                    nn.init.normal_(m.alpha.data, mean=1.0, std=std)
        elif classname == "SnakeBeta":
            # Respect the alpha_logscale setting in SnakeBeta
            if hasattr(m, "alpha") and m.alpha is not None:
                if m.alpha_logscale:
                    nn.init.constant_(m.alpha.data, 0.0)  # Matches SnakeBeta's default
                else:
                    nn.init.constant_(m.alpha.data, 1.0)
            if hasattr(m, "beta") and m.beta is not None:
                if m.alpha_logscale:
                    nn.init.constant_(m.beta.data, 0.0)  # Matches SnakeBeta's default
                else:
                    nn.init.constant_(m.beta.data, 1.0)

        # Handle residual scaling parameters
        elif hasattr(m, "residual_scale") and m.residual_scale is not None:
            nn.init.xavier_normal_(m.residual_scale.data, gain=gain)

    else:
        # Kaiming initialization
        if "Depthwise_Separable" in classname:
            nn.init.kaiming_normal_(
                m.depth_conv.weight.data, mode="fan_out", nonlinearity="relu"
            )
            nn.init.kaiming_normal_(
                m.point_conv.weight.data, mode="fan_out", nonlinearity="relu"
            )
        elif classname.find("Conv") != -1:
            nn.init.kaiming_normal_(m.weight.data, mode="fan_out", nonlinearity="relu")
            if hasattr(m, "bias") and m.bias is not None:
                nn.init.zeros_(m.bias.data)
        elif classname.find("BatchNorm") != -1:
            if hasattr(m, "weight") and m.weight is not None:
                nn.init.normal_(m.weight.data, 1.0, std)
            if hasattr(m, "bias") and m.bias is not None:
                nn.init.zeros_(m.bias.data)
        elif classname == "Snake1d":
            if hasattr(m, "alpha") and m.alpha is not None:
                nn.init.normal_(m.alpha.data, 1.0, std)
        elif classname == "SnakeBeta":
            if hasattr(m, "beta") and m.beta is not None:
                nn.init.normal_(m.beta.data, 1.0, std)
            elif (
                hasattr(m, "alpha") and m.alpha is not None
            ):  # Fallback if SnakeBeta uses alpha
                nn.init.normal_(m.alpha.data, 1.0, std)

        elif hasattr(m, "residual_scale") and m.residual_scale is not None:
            nn.init.normal_(m.residual_scale.data, 0.1, std)


class ResidualUnit(nn.Module):
    def __init__(
        self,
        dim: int = 16,
        dilation: int = 1,
        activation: str = "snake",
        alpha: float = 1.0,
        scale_residual: bool = False,
    ):
        """
        Residual Unit with weight normalization and dilated convolutions.
        Args:
            dim (int): Number of input and output channels.
            dilation (int): Dilation factor for the convolution.
            activation (str): Activation function to use.
            alpha (float): Scaling factor for the activation function.
        """
        super().__init__()
        pad = ((7 - 1) * dilation) // 2
        self.block = nn.Sequential(
            get_activation(activation=activation, channels=dim, alpha=alpha),
            WNConv1d(dim, dim, kernel_size=7, dilation=dilation, padding=pad),
            get_activation(activation=activation, channels=dim, alpha=alpha),
            WNConv1d(dim, dim, kernel_size=1),
        )
        self.scale_residual = scale_residual
        if self.scale_residual:
            self.res_scale = nn.Parameter(torch.tensor(0.0))  # start at 0

    def forward(self, x):
        y = self.block(x)
        pad = (x.shape[-1] - y.shape[-1]) // 2
        if pad > 0:
            x = x[..., pad:-pad]
        if self.scale_residual:
            y = self.res_scale * y
        return x + y


class EncoderBlock(nn.Module):
    def __init__(
        self,
        dim: int = 16,
        stride: int = 1,
        activation: str = "snake",
        alpha: float = 1.0,
        scale_residual: bool = False,
    ):
        """
        Encoder block that downsamples the input and applies residual units.
        """
        super().__init__()
        self.block = nn.Sequential(
            ResidualUnit(
                dim // 2,
                dilation=1,
                activation=activation,
                alpha=alpha,
                scale_residual=scale_residual,
            ),
            ResidualUnit(
                dim // 2,
                dilation=3,
                activation=activation,
                alpha=alpha,
                scale_residual=scale_residual,
            ),
            ResidualUnit(
                dim // 2,
                dilation=9,
                activation=activation,
                alpha=alpha,
                scale_residual=scale_residual,
            ),
            get_activation(activation=activation, channels=dim // 2, alpha=alpha),
            WNConv1d(
                dim // 2,
                dim,
                kernel_size=2 * stride,
                stride=stride,
                padding=math.ceil(stride / 2),
            ),
        )

    def forward(self, x):
        return self.block(x)


class Encoder(nn.Module):
    def __init__(
        self,
        d_model: int = 64,
        strides: list = [2, 4, 8, 8],
        d_latent: int = 64,
        d_in: int = 1,
        activation: str = "snake",
        alpha: float = 1.0,
        scale_residual: bool = False,
    ):
        super().__init__()
        # Create first convolution
        self.block = [WNConv1d(d_in, d_model, kernel_size=7, padding=3)]

        # Create EncoderBlocks that double channels as they downsample by `stride`
        for stride in strides:
            d_model *= 2
            self.block += [
                EncoderBlock(
                    d_model,
                    stride=stride,
                    activation=activation,
                    alpha=alpha,
                    scale_residual=scale_residual,
                )
            ]

        # Create last convolution
        self.block += [
            get_activation(activation=activation, channels=d_model, alpha=alpha),
            WNConv1d(d_model, d_latent, kernel_size=3, padding=1),
        ]

        # Wrap black into nn.Sequential
        self.block = nn.Sequential(*self.block)
        self.enc_dim = d_model

    def forward(self, x):
        return self.block(x)


class DecoderBlock(nn.Module):
    def __init__(
        self,
        input_dim: int = 16,
        output_dim: int = 8,
        stride: int = 1,
        norm: bool = False,
        activation: str = "snake",
        alpha: float = 1.0,
        scale_residual: bool = False,
    ):
        """
        Decoder block that upsamples the input and applies residual units.
        """
        super().__init__()
        if not norm:
            self.block = nn.Sequential(
                get_activation(activation=activation, channels=input_dim, alpha=alpha),
                WNConvTranspose1d(
                    input_dim,
                    output_dim,
                    kernel_size=2 * stride,
                    stride=stride,
                    padding=math.ceil(stride / 2),
                    output_padding=0 if stride % 2 == 0 else 1,
                ),
                ResidualUnit(
                    output_dim,
                    dilation=1,
                    activation=activation,
                    alpha=alpha,
                    scale_residual=scale_residual,
                ),
                ResidualUnit(
                    output_dim,
                    dilation=3,
                    activation=activation,
                    alpha=alpha,
                    scale_residual=scale_residual,
                ),
                ResidualUnit(
                    output_dim,
                    dilation=9,
                    activation=activation,
                    alpha=alpha,
                    scale_residual=scale_residual,
                ),
            )
        else:
            self.block = nn.Sequential(
                get_activation(activation=activation, channels=input_dim, alpha=alpha),
                WNConvTranspose1d(
                    input_dim,
                    output_dim,
                    kernel_size=2 * stride,
                    stride=stride,
                    padding=math.ceil(stride / 2),
                    output_padding=0 if stride % 2 == 0 else 1,
                ),
                nn.BatchNorm1d(output_dim),
                ResidualUnit(
                    output_dim,
                    dilation=1,
                    activation=activation,
                    alpha=alpha,
                    scale_residual=scale_residual,
                ),
                nn.BatchNorm1d(output_dim),
                ResidualUnit(
                    output_dim,
                    dilation=3,
                    activation=activation,
                    alpha=alpha,
                    scale_residual=scale_residual,
                ),
                nn.BatchNorm1d(output_dim),
                ResidualUnit(
                    output_dim,
                    dilation=9,
                    activation=activation,
                    alpha=alpha,
                    scale_residual=scale_residual,
                ),
            )

    def forward(self, x):
        return self.block(x)


class Decoder(nn.Module):
    def __init__(
        self,
        input_channel,
        channels,
        rates,
        d_out: int = 1,
        norm: bool = False,
        activation: str = "snake",
        alpha: float = 1.0,
        scale_residual: bool = False,
        use_tanh_as_final: bool = True,
        use_bias_at_final: bool = True,
    ):
        super().__init__()

        # Add first conv layer
        layers = [WNConv1d(input_channel, channels, kernel_size=7, padding=3)]

        # Add upsampling + MRF blocks
        for i, stride in enumerate(rates):
            input_dim = channels // 2**i
            output_dim = channels // 2 ** (i + 1)
            layers += [
                DecoderBlock(
                    input_dim,
                    output_dim,
                    stride,
                    norm=norm,
                    activation=activation,
                    alpha=alpha,
                    scale_residual=scale_residual,
                )
            ]

        # Add final conv layer
        layers += [
            get_activation(activation=activation, channels=output_dim, alpha=alpha),
            WNConv1d(
                output_dim, d_out, kernel_size=7, padding=3, bias=use_bias_at_final
            ),
            nn.Tanh() if use_tanh_as_final else nn.Identity(),
        ]
        self.use_tanh_as_final = use_tanh_as_final

        self.model = nn.Sequential(*layers)

    def forward(self, x):
        x = self.model(x)
        if not self.use_tanh_as_final:
            x = torch.clamp(
                x, min=-1.0, max=1.0
            )  # Ensure output is within [-1, 1] range
        return x


class DACVAE(BaseModel, CodecMixin):
    def __init__(
        self,
        encoder_dim: int = 64,
        encoder_rates: List[int] = [2, 4, 8, 8],
        latent_dim: int = 64,
        decoder_dim: int = 1536,
        decoder_rates: List[int] = [8, 8, 4, 2],
        sample_rate: int = 44100,
        d_in: int = 2,
        d_out: int = 2,
        weight_init: str = "xavier",
        norm: bool = False,
        activation: str = "snake",
        alpha: float = 1.0,
        gain: float = 0.02,
        scale_residual: bool = False,
        use_tanh_as_final: bool = True,
        use_bias_at_final: bool = True,
    ):
        super().__init__()

        self.encoder_dim = encoder_dim
        self.encoder_rates = encoder_rates
        self.decoder_dim = decoder_dim
        self.decoder_rates = decoder_rates
        self.sample_rate = sample_rate
        self.d_in = d_in
        self.d_out = d_out

        if latent_dim is None:
            latent_dim = encoder_dim * (2 ** len(encoder_rates))

        self.latent_dim = latent_dim

        self.hop_length = np.prod(encoder_rates)
        self.encoder = Encoder(
            encoder_dim,
            encoder_rates,
            latent_dim,
            d_in=d_in,
            activation=activation,
            alpha=alpha,
            scale_residual=scale_residual,
        )

        self.decoder = Decoder(
            latent_dim,
            decoder_dim,
            decoder_rates,
            d_out=d_out,
            norm=norm,
            activation=activation,
            alpha=alpha,
            scale_residual=scale_residual,
            use_tanh_as_final=use_tanh_as_final,
            use_bias_at_final=use_bias_at_final,
        )

        self.en_conv_post = WNConv1d(
            self.latent_dim, 2 * self.latent_dim, kernel_size=1
        )

        self.de_conv_pre = WNConv1d(self.latent_dim, self.latent_dim, kernel_size=1)

        self.sample_rate = sample_rate
        self.apply(lambda m: init_weights(m, init_type=weight_init, gain=gain))
        self.step = 0  # Initialize step counter for noise decay

    def freeze_encoder(self):
        for param in self.encoder.parameters():
            param.requires_grad = False
        for param in self.en_conv_post.parameters():
            param.requires_grad = False
        print("Encoder and en_conv_post frozen")

    def preprocess(self, audio_data, sample_rate):
        if sample_rate is None:
            sample_rate = self.sample_rate
        assert sample_rate == self.sample_rate
        length = audio_data.shape[-1]
        # print(f"Audio length: {length}", "math.ceil(length / self.hop_length) * self.hop_length: ", math.ceil(length / self.hop_length) * self.hop_length)
        right_pad = math.ceil(length / self.hop_length) * self.hop_length - length

        audio_data = nn.functional.pad(audio_data, (0, right_pad))
        return audio_data

    def encode(
        self,
        audio_data: torch.Tensor,
        training: bool = True,
    ):
        x = self.encoder(audio_data)
        x = F.leaky_relu(x)
        x = self.en_conv_post(x)
        print('x shape: ', x.shape)
        m, logs = torch.split(x, self.latent_dim, dim=1)
        logs = torch.clamp(logs, min=-14.0, max=14.0)

        z = m + torch.randn_like(m) * torch.exp(logs)

        return z, m, logs

    def decode(self, z: torch.Tensor):
        z = self.de_conv_pre(z)
        z = self.decoder(z)
        return z

    def forward(
        self,
        audio_data: torch.Tensor,
        sample_rate: int = 24000,
    ):
        # print(f"Audio data shape: {audio_data.shape}")
        length = audio_data.shape[-1]
        audio_data = self.preprocess(audio_data, sample_rate)
        print('audio_data: ', audio_data.shape)
        z, m, logs = self.encode(audio_data)
        x = self.decode(z)
        return {
            "audio": x[..., :length],
            "z": z,
            "mu": m,
            "logs": logs,
        }


def WNConv1d(*args, **kwargs):
    act = kwargs.pop("act", True)
    conv = weight_norm(nn.Conv1d(*args, **kwargs))
    if not act:
        return conv
    return nn.Sequential(conv, nn.LeakyReLU(0.1))


def WNConv2d(*args, **kwargs):
    act = kwargs.pop("act", True)
    conv = weight_norm(nn.Conv2d(*args, **kwargs))
    if not act:
        return conv
    return nn.Sequential(conv, nn.LeakyReLU(0.1))


class MPD(nn.Module):
    def __init__(self, period):
        super().__init__()
        self.period = period
        self.convs = nn.ModuleList(
            [
                WNConv2d(1, 32, (5, 1), (3, 1), padding=(2, 0)),
                WNConv2d(32, 128, (5, 1), (3, 1), padding=(2, 0)),
                WNConv2d(128, 512, (5, 1), (3, 1), padding=(2, 0)),
                WNConv2d(512, 1024, (5, 1), (3, 1), padding=(2, 0)),
                WNConv2d(1024, 1024, (5, 1), 1, padding=(2, 0)),
            ]
        )
        self.conv_post = WNConv2d(
            1024, 1, kernel_size=(3, 1), padding=(1, 0), act=False
        )

    def pad_to_period(self, x):
        t = x.shape[-1]
        x = F.pad(x, (0, self.period - t % self.period), mode="reflect")
        return x

    def forward(self, x):
        fmap = []

        x = self.pad_to_period(x)
        x = rearrange(x, "b c (l p) -> b c l p", p=self.period)

        for layer in self.convs:
            x = layer(x)
            fmap.append(x)

        x = self.conv_post(x)
        fmap.append(x)

        return fmap


class MSD(nn.Module):
    def __init__(self, rate: int = 1, sample_rate: int = 44100):
        super().__init__()
        self.convs = nn.ModuleList(
            [
                WNConv1d(1, 16, 15, 1, padding=7),
                WNConv1d(16, 64, 41, 4, groups=4, padding=20),
                WNConv1d(64, 256, 41, 4, groups=16, padding=20),
                WNConv1d(256, 1024, 41, 4, groups=64, padding=20),
                WNConv1d(1024, 1024, 41, 4, groups=256, padding=20),
                WNConv1d(1024, 1024, 5, 1, padding=2),
            ]
        )
        self.conv_post = WNConv1d(1024, 1, 3, 1, padding=1, act=False)
        self.sample_rate = sample_rate
        self.rate = rate

    def forward(self, x):
        x = AudioSignal(x, self.sample_rate)
        x.resample(self.sample_rate // self.rate)
        x = x.audio_data

        fmap = []

        for l in self.convs:
            x = l(x)
            fmap.append(x)
        x = self.conv_post(x)
        fmap.append(x)

        return fmap


BANDS = [(0.0, 0.1), (0.1, 0.25), (0.25, 0.5), (0.5, 0.75), (0.75, 1.0)]


class MRD(nn.Module):
    def __init__(
        self,
        window_length: int,
        hop_factor: float = 0.25,
        sample_rate: int = 44100,
        bands: list = BANDS,
    ):
        """Complex multi-band spectrogram discriminator.
        Parameters
        ----------
        window_length : int
            Window length of STFT.
        hop_factor : float, optional
            Hop factor of the STFT, defaults to ``0.25 * window_length``.
        sample_rate : int, optional
            Sampling rate of audio in Hz, by default 44100
        bands : list, optional
            Bands to run discriminator over.
        """
        super().__init__()

        self.window_length = window_length
        self.hop_factor = hop_factor
        self.sample_rate = sample_rate
        self.stft_params = STFTParams(
            window_length=window_length,
            hop_length=int(window_length * hop_factor),
            match_stride=True,
        )

        n_fft = window_length // 2 + 1
        bands = [(int(b[0] * n_fft), int(b[1] * n_fft)) for b in bands]
        self.bands = bands

        ch = 32
        convs = lambda: nn.ModuleList(
            [
                WNConv2d(2, ch, (3, 9), (1, 1), padding=(1, 4)),
                WNConv2d(ch, ch, (3, 9), (1, 2), padding=(1, 4)),
                WNConv2d(ch, ch, (3, 9), (1, 2), padding=(1, 4)),
                WNConv2d(ch, ch, (3, 9), (1, 2), padding=(1, 4)),
                WNConv2d(ch, ch, (3, 3), (1, 1), padding=(1, 1)),
            ]
        )
        self.band_convs = nn.ModuleList([convs() for _ in range(len(self.bands))])
        self.conv_post = WNConv2d(ch, 1, (3, 3), (1, 1), padding=(1, 1), act=False)

    def spectrogram(self, x):
        x = AudioSignal(x, self.sample_rate, stft_params=self.stft_params)
        x = torch.view_as_real(x.stft())
        x = rearrange(x, "b 1 f t c -> (b 1) c t f")
        # Split into bands
        x_bands = [x[..., b[0] : b[1]] for b in self.bands]
        return x_bands

    def forward(self, x):
        x_bands = self.spectrogram(x)
        fmap = []

        x = []
        for band, stack in zip(x_bands, self.band_convs):
            for layer in stack:
                band = layer(band)
                fmap.append(band)
            x.append(band)

        x = torch.cat(x, dim=-1)
        x = self.conv_post(x)
        fmap.append(x)

        return fmap


class Discriminator(ml.BaseModel):
    def __init__(
        self,
        rates: list = [],
        periods: list = [2, 3, 5, 7, 11],
        fft_sizes: list = [2048, 1024, 512],
        sample_rate: int = 44100,
        bands: list = BANDS,
        d_in: int = 1,
    ):
        """Discriminator that combines multiple discriminators.

        Parameters
        ----------
        rates : list, optional
            sampling rates (in Hz) to run MSD at, by default []
            If empty, MSD is not used.
        periods : list, optional
            periods (of samples) to run MPD at, by default [2, 3, 5, 7, 11]
        fft_sizes : list, optional
            Window sizes of the FFT to run MRD at, by default [2048, 1024, 512]
        sample_rate : int, optional
            Sampling rate of audio in Hz, by default 44100
        bands : list, optional
            Bands to run MRD at, by default `BANDS`
        """
        super().__init__()
        discs = []
        discs += [MPD(p) for p in periods]
        discs += [MSD(r, sample_rate=sample_rate) for r in rates]
        discs += [MRD(f, sample_rate=sample_rate, bands=bands) for f in fft_sizes]
        self.discriminators = nn.ModuleList(discs)

    def preprocess(self, y):
        # Remove DC offset
        y = y - y.mean(dim=-1, keepdims=True)
        # Peak normalize the volume of input audio
        y = 0.8 * y / (y.abs().max(dim=-1, keepdim=True)[0] + 1e-9)
        return y

    def forward(self, x):
        x = self.preprocess(x)
        fmaps = [d(x) for d in self.discriminators]
        return fmaps


if __name__ == "__main__":
    disc = Discriminator()
    x = torch.zeros(1, 1, 44100)
    results = disc(x)
    for i, result in enumerate(results):
        print(f"disc{i}")
        for i, r in enumerate(result):
            print(r.shape, r.mean(), r.min(), r.max())
        print()