From 44725ddd2cc4ed7cf4d6bf002b0d326e67437b1f Mon Sep 17 00:00:00 2001 From: "github-actions[bot]" <41898282+github-actions[bot]@users.noreply.github.com> Date: Fri, 7 Jun 2024 20:29:03 +0900 Subject: [PATCH] chore(format): run black on dev (#9) Co-authored-by: github-actions[bot] --- infer/lib/infer_pack/models.py | 19 +++++++++++++++---- infer/lib/infer_pack/modules.py | 6 +++++- rvc/attentions.py | 5 ++++- rvc/encoders.py | 5 +++-- rvc/utils.py | 6 ++++-- 5 files changed, 31 insertions(+), 10 deletions(-) diff --git a/infer/lib/infer_pack/models.py b/infer/lib/infer_pack/models.py index 1b3fc18..8b1f715 100644 --- a/infer/lib/infer_pack/models.py +++ b/infer/lib/infer_pack/models.py @@ -9,7 +9,13 @@ from torch.nn import functional as F from torch.nn.utils import remove_weight_norm, spectral_norm, weight_norm from infer.lib.infer_pack import modules -from rvc.utils import get_padding, call_weight_data_normal_if_Conv, sequence_mask, slice_on_last_dim, rand_slice_segments_on_last_dim +from rvc.utils import ( + get_padding, + call_weight_data_normal_if_Conv, + sequence_mask, + slice_on_last_dim, + rand_slice_segments_on_last_dim, +) from rvc.encoders import TextEncoder has_xpu = bool(hasattr(torch, "xpu") and torch.xpu.is_available()) @@ -120,7 +126,8 @@ class PosteriorEncoder(nn.Module): self, x: torch.Tensor, x_lengths: torch.Tensor, g: Optional[torch.Tensor] = None ) -> Tuple[torch.Tensor, torch.Tensor, torch.Tensor, torch.Tensor]: x_mask = torch.unsqueeze( - sequence_mask(x_lengths, x.size(2)), 1, + sequence_mask(x_lengths, x.size(2)), + 1, ).to(x.dtype) x = self.pre(x) * x_mask x = self.enc(x, x_mask, g=g) @@ -688,7 +695,9 @@ class SynthesizerTrnMs256NSFsid(nn.Module): m_p, logs_p, x_mask = self.enc_p(phone, pitch, phone_lengths) z, m_q, logs_q, y_mask = self.enc_q(y, y_lengths, g=g) z_p = self.flow(z, y_mask, g=g) - z_slice, ids_slice = rand_slice_segments_on_last_dim(z, y_lengths, self.segment_size) + z_slice, ids_slice = rand_slice_segments_on_last_dim( + z, y_lengths, self.segment_size + ) # print(-1,pitchf.shape,ids_slice,self.segment_size,self.hop_length,self.segment_size//self.hop_length) pitchf = slice_on_last_dim(pitchf, ids_slice, self.segment_size) # print(-2,pitchf.shape,z_slice.shape) @@ -906,7 +915,9 @@ class SynthesizerTrnMs256NSFsid_nono(nn.Module): m_p, logs_p, x_mask = self.enc_p(phone, None, phone_lengths) z, m_q, logs_q, y_mask = self.enc_q(y, y_lengths, g=g) z_p = self.flow(z, y_mask, g=g) - z_slice, ids_slice = rand_slice_segments_on_last_dim(z, y_lengths, self.segment_size) + z_slice, ids_slice = rand_slice_segments_on_last_dim( + z, y_lengths, self.segment_size + ) o = self.dec(z_slice, g=g) return o, ids_slice, x_mask, y_mask, (z, z_p, m_p, logs_p, m_q, logs_q) diff --git a/infer/lib/infer_pack/modules.py b/infer/lib/infer_pack/modules.py index b8a049c..95ffbfe 100644 --- a/infer/lib/infer_pack/modules.py +++ b/infer/lib/infer_pack/modules.py @@ -7,7 +7,11 @@ from torch.nn import Conv1d from torch.nn import functional as F from torch.nn.utils import remove_weight_norm, weight_norm -from rvc.utils import get_padding, call_weight_data_normal_if_Conv, activate_add_tanh_sigmoid_multiply +from rvc.utils import ( + get_padding, + call_weight_data_normal_if_Conv, + activate_add_tanh_sigmoid_multiply, +) from rvc.transforms import piecewise_rational_quadratic_transform from rvc.norms import LayerNorm diff --git a/rvc/attentions.py b/rvc/attentions.py index 06d9ca4..a6d34f8 100644 --- a/rvc/attentions.py +++ b/rvc/attentions.py @@ -173,7 +173,10 @@ class MultiHeadAttention(nn.Module): """ batch, heads, length, _ = x.size() # Concat columns of pad to shift from relative to absolute indexing. - x = F.pad(x, [0, 1, 0, 0, 0, 0, 0, 0], ) + x = F.pad( + x, + [0, 1, 0, 0, 0, 0, 0, 0], + ) # Concat extra elements so to add up to shape (len+1, 2*len-1). x_flat = x.view([batch, heads, length * 2 * length]) diff --git a/rvc/encoders.py b/rvc/encoders.py index 4cb62e3..c51828f 100644 --- a/rvc/encoders.py +++ b/rvc/encoders.py @@ -57,7 +57,7 @@ class Encoder(nn.Module): ) ) self.norm_layers_2.append(LayerNorm(hidden_channels)) - + def __call__(self, x: torch.Tensor, x_mask: torch.Tensor) -> torch.Tensor: return super().__call__(x, x_mask) @@ -146,7 +146,8 @@ class TextEncoder(nn.Module): x = self.lrelu(x) x = torch.transpose(x, 1, -1) # [b, h, t] x_mask = torch.unsqueeze( - sequence_mask(lengths, x.size(2)), 1, + sequence_mask(lengths, x.size(2)), + 1, ).to(x.dtype) x = self.encoder(x * x_mask, x_mask) """ diff --git a/rvc/utils.py b/rvc/utils.py index 4a94f3f..d979087 100644 --- a/rvc/utils.py +++ b/rvc/utils.py @@ -66,13 +66,15 @@ def sequence_mask( def total_grad_norm( - parameters: Iterator[torch.nn.Parameter], norm_type: float=2.0, + parameters: Iterator[torch.nn.Parameter], + norm_type: float = 2.0, ) -> float: norm_type = float(norm_type) total_norm = 0.0 for p in parameters: - if p.grad is None: continue + if p.grad is None: + continue param_norm = p.grad.data.norm(norm_type) total_norm += float(param_norm.item()) ** norm_type total_norm = total_norm ** (1.0 / norm_type)