refactor(vc): rename TTS.vc.modules to TTS.vc.layers for consistency

Same as in TTS.tts and TTS.vocoder

TTS/vc/layers/freevc/modules.py

@@ -7,7 +7,7 @@ from torch.nn.utils.parametrize import remove_parametrizations
 
 from TTS.tts.layers.generic.normalization import LayerNorm2
 from TTS.tts.layers.generic.wavenet import fused_add_tanh_sigmoid_multiply
-from TTS.vc.modules.freevc.commons import init_weights
+from TTS.vc.layers.freevc.commons import init_weights
 from TTS.vocoder.models.hifigan_generator import get_padding
 
 LRELU_SLOPE = 0.1

TTS/vc/layers/freevc/speaker_encoder/audio.py

@@ -5,7 +5,7 @@ from typing import Optional, Union
 import librosa
 import numpy as np
 
-from TTS.vc.modules.freevc.speaker_encoder.hparams import (
+from TTS.vc.layers.freevc.speaker_encoder.hparams import (
     audio_norm_target_dBFS,
     mel_n_channels,
     mel_window_length,

TTS/vc/layers/freevc/speaker_encoder/speaker_encoder.py

@@ -7,8 +7,8 @@ import torch
 from torch import nn
 from trainer.io import load_fsspec
 
-from TTS.vc.modules.freevc.speaker_encoder import audio
+from TTS.vc.layers.freevc.speaker_encoder import audio
-from TTS.vc.modules.freevc.speaker_encoder.hparams import (
+from TTS.vc.layers.freevc.speaker_encoder.hparams import (
     mel_n_channels,
     mel_window_step,
     model_embedding_size,

TTS/vc/layers/freevc/wavlm/__init__.py

@@ -6,7 +6,7 @@ import torch
 from trainer.io import get_user_data_dir
 
 from TTS.utils.generic_utils import is_pytorch_at_least_2_4
-from TTS.vc.modules.freevc.wavlm.wavlm import WavLM, WavLMConfig
+from TTS.vc.layers.freevc.wavlm.wavlm import WavLM, WavLMConfig
 
 logger = logging.getLogger(__name__)
 

TTS/vc/layers/freevc/wavlm/wavlm.py

@@ -17,7 +17,7 @@ import torch.nn as nn
 import torch.nn.functional as F
 from torch.nn import LayerNorm
 
-from TTS.vc.modules.freevc.wavlm.modules import (
+from TTS.vc.layers.freevc.wavlm.modules import (
     Fp32GroupNorm,
     Fp32LayerNorm,
     GLU_Linear,

TTS/vc/models/freevc.py

@@ -12,17 +12,16 @@ from torch.nn.utils.parametrizations import weight_norm
 from torch.nn.utils.parametrize import remove_parametrizations
 from trainer.io import load_fsspec
 
-import TTS.vc.modules.freevc.commons as commons
+import TTS.vc.layers.freevc.modules as modules
-import TTS.vc.modules.freevc.modules as modules
 from TTS.tts.layers.vits.discriminator import DiscriminatorS
 from TTS.tts.utils.helpers import sequence_mask
 from TTS.tts.utils.speakers import SpeakerManager
 from TTS.vc.configs.freevc_config import FreeVCConfig
+from TTS.vc.layers.freevc.commons import init_weights, rand_slice_segments
+from TTS.vc.layers.freevc.mel_processing import mel_spectrogram_torch
+from TTS.vc.layers.freevc.speaker_encoder.speaker_encoder import SpeakerEncoder as SpeakerEncoderEx
+from TTS.vc.layers.freevc.wavlm import get_wavlm
 from TTS.vc.models.base_vc import BaseVC
-from TTS.vc.modules.freevc.commons import init_weights
-from TTS.vc.modules.freevc.mel_processing import mel_spectrogram_torch
-from TTS.vc.modules.freevc.speaker_encoder.speaker_encoder import SpeakerEncoder as SpeakerEncoderEx
-from TTS.vc.modules.freevc.wavlm import get_wavlm
 from TTS.vocoder.models.hifigan_discriminator import DiscriminatorP
 
 logger = logging.getLogger(__name__)
@@ -385,7 +384,7 @@ class FreeVC(BaseVC):
         z_p = self.flow(z, spec_mask, g=g)
 
         # Randomly slice z and compute o using dec
-        z_slice, ids_slice = commons.rand_slice_segments(z, spec_lengths, self.segment_size)
+        z_slice, ids_slice = rand_slice_segments(z, spec_lengths, self.segment_size)
         o = self.dec(z_slice, g=g)
 
         return o, ids_slice, spec_mask, (z, z_p, m_p, logs_p, m_q, logs_q)