Upload 70 files

.gitattributes

@@ -32,3 +32,5 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+monotonic_align/build/temp.linux-x86_64-3.8/core.o filter=lfs diff=lfs merge=lfs -text
+monotonic_align/monotonic_align/core.cpython-38-x86_64-linux-gnu.so filter=lfs diff=lfs merge=lfs -text

ONNXVITS_infer.py

@@ -0,0 +1,201 @@
+import torch
+import commons
+import models
+
+import math
+from torch import nn
+from torch.nn import functional as F
+
+import modules
+import attentions
+
+from torch.nn import Conv1d, ConvTranspose1d, Conv2d
+from torch.nn.utils import weight_norm, remove_weight_norm, spectral_norm
+from commons import init_weights, get_padding
+
+
+class TextEncoder(nn.Module):
+  def __init__(self,
+               n_vocab,
+               out_channels,
+               hidden_channels,
+               filter_channels,
+               n_heads,
+               n_layers,
+               kernel_size,
+               p_dropout,
+               emotion_embedding):
+    super().__init__()
+    self.n_vocab = n_vocab
+    self.out_channels = out_channels
+    self.hidden_channels = hidden_channels
+    self.filter_channels = filter_channels
+    self.n_heads = n_heads
+    self.n_layers = n_layers
+    self.kernel_size = kernel_size
+    self.p_dropout = p_dropout
+    self.emotion_embedding = emotion_embedding
+
+    if self.n_vocab != 0:
+      self.emb = nn.Embedding(n_vocab, hidden_channels)
+      if emotion_embedding:
+        self.emo_proj = nn.Linear(1024, hidden_channels)
+      nn.init.normal_(self.emb.weight, 0.0, hidden_channels ** -0.5)
+
+    self.encoder = attentions.Encoder(
+      hidden_channels,
+      filter_channels,
+      n_heads,
+      n_layers,
+      kernel_size,
+      p_dropout)
+    self.proj = nn.Conv1d(hidden_channels, out_channels * 2, 1)
+
+  def forward(self, x, x_lengths, emotion_embedding=None):
+    if self.n_vocab != 0:
+      x = self.emb(x) * math.sqrt(self.hidden_channels)  # [b, t, h]
+    if emotion_embedding is not None:
+      print("emotion added")
+      x = x + self.emo_proj(emotion_embedding.unsqueeze(1))
+    x = torch.transpose(x, 1, -1)  # [b, h, t]
+    x_mask = torch.unsqueeze(commons.sequence_mask(x_lengths, x.size(2)), 1).to(x.dtype)
+
+    x = self.encoder(x * x_mask, x_mask)
+    stats = self.proj(x) * x_mask
+
+    m, logs = torch.split(stats, self.out_channels, dim=1)
+    return x, m, logs, x_mask
+
+
+class PosteriorEncoder(nn.Module):
+  def __init__(self,
+               in_channels,
+               out_channels,
+               hidden_channels,
+               kernel_size,
+               dilation_rate,
+               n_layers,
+               gin_channels=0):
+    super().__init__()
+    self.in_channels = in_channels
+    self.out_channels = out_channels
+    self.hidden_channels = hidden_channels
+    self.kernel_size = kernel_size
+    self.dilation_rate = dilation_rate
+    self.n_layers = n_layers
+    self.gin_channels = gin_channels
+
+    self.pre = nn.Conv1d(in_channels, hidden_channels, 1)
+    self.enc = modules.WN(hidden_channels, kernel_size, dilation_rate, n_layers, gin_channels=gin_channels)
+    self.proj = nn.Conv1d(hidden_channels, out_channels * 2, 1)
+
+  def forward(self, x, x_lengths, g=None):
+    x_mask = torch.unsqueeze(commons.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)
+    stats = self.proj(x) * x_mask
+    m, logs = torch.split(stats, self.out_channels, dim=1)
+    z = (m + torch.randn_like(m) * torch.exp(logs)) * x_mask
+    return z, m, logs, x_mask
+
+
+class SynthesizerTrn(models.SynthesizerTrn):
+  """
+  Synthesizer for Training
+  """
+
+  def __init__(self,
+               n_vocab,
+               spec_channels,
+               segment_size,
+               inter_channels,
+               hidden_channels,
+               filter_channels,
+               n_heads,
+               n_layers,
+               kernel_size,
+               p_dropout,
+               resblock,
+               resblock_kernel_sizes,
+               resblock_dilation_sizes,
+               upsample_rates,
+               upsample_initial_channel,
+               upsample_kernel_sizes,
+               n_speakers=0,
+               gin_channels=0,
+               use_sdp=True,
+               emotion_embedding=False,
+               ONNX_dir="./ONNX_net/",
+               **kwargs):
+
+    super().__init__(
+      n_vocab,
+      spec_channels,
+      segment_size,
+      inter_channels,
+      hidden_channels,
+      filter_channels,
+      n_heads,
+      n_layers,
+      kernel_size,
+      p_dropout,
+      resblock,
+      resblock_kernel_sizes,
+      resblock_dilation_sizes,
+      upsample_rates,
+      upsample_initial_channel,
+      upsample_kernel_sizes,
+      n_speakers=n_speakers,
+      gin_channels=gin_channels,
+      use_sdp=use_sdp,
+      **kwargs
+    )
+    self.ONNX_dir = ONNX_dir
+    self.enc_p = TextEncoder(n_vocab,
+                             inter_channels,
+                             hidden_channels,
+                             filter_channels,
+                             n_heads,
+                             n_layers,
+                             kernel_size,
+                             p_dropout,
+                             emotion_embedding)
+    self.enc_q = PosteriorEncoder(spec_channels, inter_channels, hidden_channels, 5, 1, 16, gin_channels=gin_channels)
+
+  def infer(self, x, x_lengths, sid=None, noise_scale=1, length_scale=1, noise_scale_w=1., max_len=None,
+            emotion_embedding=None):
+    from ONNXVITS_utils import runonnx
+    with torch.no_grad():
+      x, m_p, logs_p, x_mask = self.enc_p(x, x_lengths, emotion_embedding)
+
+    if self.n_speakers > 0:
+      g = self.emb_g(sid).unsqueeze(-1)  # [b, h, 1]
+    else:
+      g = None
+
+    # logw = self.dp(x, x_mask, g=g, reverse=True, noise_scale=noise_scale_w)
+    logw = runonnx(f"{self.ONNX_dir}dp.onnx", x=x.numpy(), x_mask=x_mask.numpy(), g=g.numpy())
+    logw = torch.from_numpy(logw[0])
+
+    w = torch.exp(logw) * x_mask * length_scale
+    w_ceil = torch.ceil(w)
+    y_lengths = torch.clamp_min(torch.sum(w_ceil, [1, 2]), 1).long()
+    y_mask = torch.unsqueeze(commons.sequence_mask(y_lengths, None), 1).to(x_mask.dtype)
+    attn_mask = torch.unsqueeze(x_mask, 2) * torch.unsqueeze(y_mask, -1)
+    attn = commons.generate_path(w_ceil, attn_mask)
+
+    m_p = torch.matmul(attn.squeeze(1), m_p.transpose(1, 2)).transpose(1, 2)  # [b, t', t], [b, t, d] -> [b, d, t']
+    logs_p = torch.matmul(attn.squeeze(1), logs_p.transpose(1, 2)).transpose(1,
+                                                                             2)  # [b, t', t], [b, t, d] -> [b, d, t']
+
+    z_p = m_p + torch.randn_like(m_p) * torch.exp(logs_p) * noise_scale
+
+    # z = self.flow(z_p, y_mask, g=g, reverse=True)
+    z = runonnx(f"{self.ONNX_dir}flow.onnx", z_p=z_p.numpy(), y_mask=y_mask.numpy(), g=g.numpy())
+    z = torch.from_numpy(z[0])
+
+    # o = self.dec((z * y_mask)[:,:,:max_len], g=g)
+    o = runonnx(f"{self.ONNX_dir}dec.onnx", z_in=(z * y_mask)[:, :, :max_len].numpy(), g=g.numpy())
+    o = torch.from_numpy(o[0])
+
+    return o, attn, y_mask, (z, z_p, m_p, logs_p)

ONNXVITS_inference.py

@@ -0,0 +1,36 @@
+import logging
+logging.getLogger('numba').setLevel(logging.WARNING)
+import IPython.display as ipd
+import torch
+import commons
+import utils
+import ONNXVITS_infer
+from text import text_to_sequence
+
+def get_text(text, hps):
+    text_norm = text_to_sequence(text, hps.symbols, hps.data.text_cleaners)
+    if hps.data.add_blank:
+        text_norm = commons.intersperse(text_norm, 0)
+    text_norm = torch.LongTensor(text_norm)
+    return text_norm
+
+hps = utils.get_hparams_from_file("../vits/pretrained_models/uma87.json")
+
+net_g = ONNXVITS_infer.SynthesizerTrn(
+    len(hps.symbols),
+    hps.data.filter_length // 2 + 1,
+    hps.train.segment_size // hps.data.hop_length,
+    n_speakers=hps.data.n_speakers,
+    **hps.model)
+_ = net_g.eval()
+
+_ = utils.load_checkpoint("../vits/pretrained_models/uma_1153000.pth", net_g)
+
+text1 = get_text("おはようございます。", hps)
+stn_tst = text1
+with torch.no_grad():
+    x_tst = stn_tst.unsqueeze(0)
+    x_tst_lengths = torch.LongTensor([stn_tst.size(0)])
+    sid = torch.LongTensor([0])
+    audio = net_g.infer(x_tst, x_tst_lengths, sid=sid, noise_scale=.667, noise_scale_w=0.8, length_scale=1)[0][0,0].data.cpu().float().numpy()
+print(audio)

ONNXVITS_models.py

@@ -0,0 +1,509 @@
+import copy
+import math
+import torch
+from torch import nn
+from torch.nn import functional as F
+
+import commons
+import ONNXVITS_modules as modules
+import attentions
+import monotonic_align
+
+from torch.nn import Conv1d, ConvTranspose1d, AvgPool1d, Conv2d
+from torch.nn.utils import weight_norm, remove_weight_norm, spectral_norm
+from commons import init_weights, get_padding
+
+
+class StochasticDurationPredictor(nn.Module):
+  def __init__(self, in_channels, filter_channels, kernel_size, p_dropout, n_flows=4, gin_channels=0):
+    super().__init__()
+    filter_channels = in_channels # it needs to be removed from future version.
+    self.in_channels = in_channels
+    self.filter_channels = filter_channels
+    self.kernel_size = kernel_size
+    self.p_dropout = p_dropout
+    self.n_flows = n_flows
+    self.gin_channels = gin_channels
+
+    self.log_flow = modules.Log()
+    self.flows = nn.ModuleList()
+    self.flows.append(modules.ElementwiseAffine(2))
+    for i in range(n_flows):
+      self.flows.append(modules.ConvFlow(2, filter_channels, kernel_size, n_layers=3))
+      self.flows.append(modules.Flip())
+
+    self.post_pre = nn.Conv1d(1, filter_channels, 1)
+    self.post_proj = nn.Conv1d(filter_channels, filter_channels, 1)
+    self.post_convs = modules.DDSConv(filter_channels, kernel_size, n_layers=3, p_dropout=p_dropout)
+    self.post_flows = nn.ModuleList()
+    self.post_flows.append(modules.ElementwiseAffine(2))
+    for i in range(4):
+      self.post_flows.append(modules.ConvFlow(2, filter_channels, kernel_size, n_layers=3))
+      self.post_flows.append(modules.Flip())
+
+    self.pre = nn.Conv1d(in_channels, filter_channels, 1)
+    self.proj = nn.Conv1d(filter_channels, filter_channels, 1)
+    self.convs = modules.DDSConv(filter_channels, kernel_size, n_layers=3, p_dropout=p_dropout)
+    if gin_channels != 0:
+      self.cond = nn.Conv1d(gin_channels, filter_channels, 1)
+
+    self.w = None
+    self.reverse = None
+    self.noise_scale = None
+  def forward(self, x, x_mask, g=None):
+    w = self.w
+    reverse = self.reverse
+    noise_scale = self.noise_scale
+
+    x = torch.detach(x)
+    x = self.pre(x)
+    if g is not None:
+      g = torch.detach(g)
+      x = x + self.cond(g)
+    x = self.convs(x, x_mask)
+    x = self.proj(x) * x_mask
+
+    if not reverse:
+      flows = self.flows
+      assert w is not None
+
+      logdet_tot_q = 0 
+      h_w = self.post_pre(w)
+      h_w = self.post_convs(h_w, x_mask)
+      h_w = self.post_proj(h_w) * x_mask
+      e_q = torch.randn(w.size(0), 2, w.size(2)).to(device=x.device, dtype=x.dtype) * x_mask
+      z_q = e_q
+      for flow in self.post_flows:
+        z_q, logdet_q = flow(z_q, x_mask, g=(x + h_w))
+        logdet_tot_q += logdet_q
+      z_u, z1 = torch.split(z_q, [1, 1], 1) 
+      u = torch.sigmoid(z_u) * x_mask
+      z0 = (w - u) * x_mask
+      logdet_tot_q += torch.sum((F.logsigmoid(z_u) + F.logsigmoid(-z_u)) * x_mask, [1,2])
+      logq = torch.sum(-0.5 * (math.log(2*math.pi) + (e_q**2)) * x_mask, [1,2]) - logdet_tot_q
+
+      logdet_tot = 0
+      z0, logdet = self.log_flow(z0, x_mask)
+      logdet_tot += logdet
+      z = torch.cat([z0, z1], 1)
+      for flow in flows:
+        z, logdet = flow(z, x_mask, g=x, reverse=reverse)
+        logdet_tot = logdet_tot + logdet
+      nll = torch.sum(0.5 * (math.log(2*math.pi) + (z**2)) * x_mask, [1,2]) - logdet_tot
+      return nll + logq # [b]
+    else:
+      flows = list(reversed(self.flows))
+      flows = flows[:-2] + [flows[-1]] # remove a useless vflow
+      z = torch.randn(x.size(0), 2, x.size(2)).to(device=x.device, dtype=x.dtype) * noise_scale
+      for flow in flows:
+        z = flow(z, x_mask, g=x, reverse=reverse)
+      z0, z1 = torch.split(z, [1, 1], 1)
+      logw = z0
+      return logw
+
+
+class TextEncoder(nn.Module):
+  def __init__(self,
+      n_vocab,
+      out_channels,
+      hidden_channels,
+      filter_channels,
+      n_heads,
+      n_layers,
+      kernel_size,
+      p_dropout):
+    super().__init__()
+    self.n_vocab = n_vocab
+    self.out_channels = out_channels
+    self.hidden_channels = hidden_channels
+    self.filter_channels = filter_channels
+    self.n_heads = n_heads
+    self.n_layers = n_layers
+    self.kernel_size = kernel_size
+    self.p_dropout = p_dropout
+
+    self.emb = nn.Embedding(n_vocab, hidden_channels)
+    nn.init.normal_(self.emb.weight, 0.0, hidden_channels**-0.5)
+
+    self.encoder = attentions.Encoder(
+      hidden_channels,
+      filter_channels,
+      n_heads,
+      n_layers,
+      kernel_size,
+      p_dropout)
+    self.proj= nn.Conv1d(hidden_channels, out_channels * 2, 1)
+
+  def forward(self, x, x_lengths):
+    x = self.emb(x) * math.sqrt(self.hidden_channels) # [b, t, h]
+    x = torch.transpose(x, 1, -1) # [b, h, t]
+    x_mask = torch.unsqueeze(commons.sequence_mask(x_lengths, x.size(2)), 1).to(x.dtype)
+
+    x = self.encoder(x * x_mask, x_mask)
+    stats = self.proj(x) * x_mask
+
+    m, logs = torch.split(stats, self.out_channels, dim=1)
+    return x, m, logs, x_mask
+
+
+class ResidualCouplingBlock(nn.Module):
+  def __init__(self,
+      channels,
+      hidden_channels,
+      kernel_size,
+      dilation_rate,
+      n_layers,
+      n_flows=4,
+      gin_channels=0):
+    super().__init__()
+    self.channels = channels
+    self.hidden_channels = hidden_channels
+    self.kernel_size = kernel_size
+    self.dilation_rate = dilation_rate
+    self.n_layers = n_layers
+    self.n_flows = n_flows
+    self.gin_channels = gin_channels
+
+    self.flows = nn.ModuleList()
+    for i in range(n_flows):
+      self.flows.append(modules.ResidualCouplingLayer(channels, hidden_channels, kernel_size, dilation_rate, n_layers, gin_channels=gin_channels, mean_only=True))
+      self.flows.append(modules.Flip())
+
+    self.reverse = None
+  def forward(self, x, x_mask, g=None):
+    reverse = self.reverse 
+    if not reverse:
+      for flow in self.flows:
+        x, _ = flow(x, x_mask, g=g, reverse=reverse)
+    else:
+      for flow in reversed(self.flows):
+        x = flow(x, x_mask, g=g, reverse=reverse)
+    return x
+
+
+class PosteriorEncoder(nn.Module):
+  def __init__(self,
+      in_channels,
+      out_channels,
+      hidden_channels,
+      kernel_size,
+      dilation_rate,
+      n_layers,
+      gin_channels=0):
+    super().__init__()
+    self.in_channels = in_channels
+    self.out_channels = out_channels
+    self.hidden_channels = hidden_channels
+    self.kernel_size = kernel_size
+    self.dilation_rate = dilation_rate
+    self.n_layers = n_layers
+    self.gin_channels = gin_channels
+
+    self.pre = nn.Conv1d(in_channels, hidden_channels, 1)
+    self.enc = modules.WN(hidden_channels, kernel_size, dilation_rate, n_layers, gin_channels=gin_channels)
+    self.proj = nn.Conv1d(hidden_channels, out_channels * 2, 1)
+
+  def forward(self, x, x_lengths, g=None):
+    x_mask = torch.unsqueeze(commons.sequence_mask(x_lengths, x.size(2)), 1).to(x.dtype)
+    x = self.pre(x) * x_mask # x_in : [b, c, t] -> [b, h, t] 
+    x = self.enc(x, x_mask, g=g) # x_in : [b, h, t], g : [b, h, 1],  x = x_in + g
+    stats = self.proj(x) * x_mask
+    m, logs = torch.split(stats, self.out_channels, dim=1)
+    z = (m + torch.randn_like(m) * torch.exp(logs)) * x_mask
+    return z, m, logs, x_mask # z, m, logs : [b, h, t]
+
+
+class Generator(torch.nn.Module):
+    def __init__(self, initial_channel, resblock, resblock_kernel_sizes, resblock_dilation_sizes, upsample_rates, upsample_initial_channel, upsample_kernel_sizes, gin_channels=0):
+        super(Generator, self).__init__()
+        self.num_kernels = len(resblock_kernel_sizes)
+        self.num_upsamples = len(upsample_rates)
+        self.conv_pre = Conv1d(initial_channel, upsample_initial_channel, 7, 1, padding=3)
+        resblock = modules.ResBlock1 if resblock == '1' else modules.ResBlock2
+
+        self.ups = nn.ModuleList()
+        for i, (u, k) in enumerate(zip(upsample_rates, upsample_kernel_sizes)):
+            self.ups.append(weight_norm(
+                ConvTranspose1d(upsample_initial_channel//(2**i), upsample_initial_channel//(2**(i+1)),
+                                k, u, padding=(k-u)//2)))
+
+        self.resblocks = nn.ModuleList()
+        for i in range(len(self.ups)):
+            ch = upsample_initial_channel//(2**(i+1))
+            for j, (k, d) in enumerate(zip(resblock_kernel_sizes, resblock_dilation_sizes)):
+                self.resblocks.append(resblock(ch, k, d))
+
+        self.conv_post = Conv1d(ch, 1, 7, 1, padding=3, bias=False)
+        self.ups.apply(init_weights)
+
+        if gin_channels != 0:
+            self.cond = nn.Conv1d(gin_channels, upsample_initial_channel, 1)
+
+    def forward(self, x, g=None):
+        x = self.conv_pre(x)
+        if g is not None:
+          x = x + self.cond(g)
+
+        for i in range(self.num_upsamples):
+            x = F.leaky_relu(x, modules.LRELU_SLOPE)
+            x = self.ups[i](x)
+            xs = None
+            for j in range(self.num_kernels):
+                if xs is None:
+                    xs = self.resblocks[i*self.num_kernels+j](x)
+                else:
+                    xs += self.resblocks[i*self.num_kernels+j](x)
+            x = xs / self.num_kernels
+        x = F.leaky_relu(x)
+        x = self.conv_post(x)
+        x = torch.tanh(x)
+
+        return x
+
+    def remove_weight_norm(self):
+        print('Removing weight norm...')
+        for l in self.ups:
+            remove_weight_norm(l)
+        for l in self.resblocks:
+            l.remove_weight_norm()
+
+
+class DiscriminatorP(torch.nn.Module):
+    def __init__(self, period, kernel_size=5, stride=3, use_spectral_norm=False):
+        super(DiscriminatorP, self).__init__()
+        self.period = period
+        self.use_spectral_norm = use_spectral_norm
+        norm_f = weight_norm if use_spectral_norm == False else spectral_norm
+        self.convs = nn.ModuleList([
+            norm_f(Conv2d(1, 32, (kernel_size, 1), (stride, 1), padding=(get_padding(kernel_size, 1), 0))),
+            norm_f(Conv2d(32, 128, (kernel_size, 1), (stride, 1), padding=(get_padding(kernel_size, 1), 0))),
+            norm_f(Conv2d(128, 512, (kernel_size, 1), (stride, 1), padding=(get_padding(kernel_size, 1), 0))),
+            norm_f(Conv2d(512, 1024, (kernel_size, 1), (stride, 1), padding=(get_padding(kernel_size, 1), 0))),
+            norm_f(Conv2d(1024, 1024, (kernel_size, 1), 1, padding=(get_padding(kernel_size, 1), 0))),
+        ])
+        self.conv_post = norm_f(Conv2d(1024, 1, (3, 1), 1, padding=(1, 0)))
+
+    def forward(self, x):
+        fmap = []
+
+        # 1d to 2d
+        b, c, t = x.shape
+        if t % self.period != 0: # pad first
+            n_pad = self.period - (t % self.period)
+            x = F.pad(x, (0, n_pad), "reflect")
+            t = t + n_pad
+        x = x.view(b, c, t // self.period, self.period)
+
+        for l in self.convs:
+            x = l(x)
+            x = F.leaky_relu(x, modules.LRELU_SLOPE)
+            fmap.append(x)
+        x = self.conv_post(x)
+        fmap.append(x)
+        x = torch.flatten(x, 1, -1)
+
+        return x, fmap
+
+
+class DiscriminatorS(torch.nn.Module):
+    def __init__(self, use_spectral_norm=False):
+        super(DiscriminatorS, self).__init__()
+        norm_f = weight_norm if use_spectral_norm == False else spectral_norm
+        self.convs = nn.ModuleList([
+            norm_f(Conv1d(1, 16, 15, 1, padding=7)),
+            norm_f(Conv1d(16, 64, 41, 4, groups=4, padding=20)),
+            norm_f(Conv1d(64, 256, 41, 4, groups=16, padding=20)),
+            norm_f(Conv1d(256, 1024, 41, 4, groups=64, padding=20)),
+            norm_f(Conv1d(1024, 1024, 41, 4, groups=256, padding=20)),
+            norm_f(Conv1d(1024, 1024, 5, 1, padding=2)),
+        ])
+        self.conv_post = norm_f(Conv1d(1024, 1, 3, 1, padding=1))
+
+    def forward(self, x):
+        fmap = []
+
+        for l in self.convs:
+            x = l(x)
+            x = F.leaky_relu(x, modules.LRELU_SLOPE)
+            fmap.append(x)
+        x = self.conv_post(x)
+        fmap.append(x)
+        x = torch.flatten(x, 1, -1)
+
+        return x, fmap
+
+
+class MultiPeriodDiscriminator(torch.nn.Module):
+    def __init__(self, use_spectral_norm=False):
+        super(MultiPeriodDiscriminator, self).__init__()
+        periods = [2,3,5,7,11]
+
+        discs = [DiscriminatorS(use_spectral_norm=use_spectral_norm)]
+        discs = discs + [DiscriminatorP(i, use_spectral_norm=use_spectral_norm) for i in periods]
+        self.discriminators = nn.ModuleList(discs)
+
+    def forward(self, y, y_hat):
+        y_d_rs = []
+        y_d_gs = []
+        fmap_rs = []
+        fmap_gs = []
+        for i, d in enumerate(self.discriminators):
+            y_d_r, fmap_r = d(y)
+            y_d_g, fmap_g = d(y_hat)
+            y_d_rs.append(y_d_r)
+            y_d_gs.append(y_d_g)
+            fmap_rs.append(fmap_r)
+            fmap_gs.append(fmap_g)
+
+        return y_d_rs, y_d_gs, fmap_rs, fmap_gs
+
+
+
+class SynthesizerTrn(nn.Module):
+  """
+  Synthesizer for Training
+  """
+
+  def __init__(self, 
+    n_vocab,
+    spec_channels,
+    segment_size,
+    inter_channels,
+    hidden_channels,
+    filter_channels,
+    n_heads,
+    n_layers,
+    kernel_size,
+    p_dropout,
+    resblock, 
+    resblock_kernel_sizes, 
+    resblock_dilation_sizes, 
+    upsample_rates, 
+    upsample_initial_channel, 
+    upsample_kernel_sizes,
+    n_speakers=0,
+    gin_channels=0,
+    use_sdp=True,
+    **kwargs):
+
+    super().__init__()
+    self.n_vocab = n_vocab
+    self.spec_channels = spec_channels
+    self.inter_channels = inter_channels
+    self.hidden_channels = hidden_channels
+    self.filter_channels = filter_channels
+    self.n_heads = n_heads
+    self.n_layers = n_layers
+    self.kernel_size = kernel_size
+    self.p_dropout = p_dropout
+    self.resblock = resblock
+    self.resblock_kernel_sizes = resblock_kernel_sizes
+    self.resblock_dilation_sizes = resblock_dilation_sizes
+    self.upsample_rates = upsample_rates
+    self.upsample_initial_channel = upsample_initial_channel
+    self.upsample_kernel_sizes = upsample_kernel_sizes
+    self.segment_size = segment_size
+    self.n_speakers = n_speakers
+    self.gin_channels = gin_channels
+
+    self.use_sdp = use_sdp
+
+    self.enc_p = TextEncoder(n_vocab,
+        inter_channels,
+        hidden_channels,
+        filter_channels,
+        n_heads,
+        n_layers,
+        kernel_size,
+        p_dropout)
+    self.dec = Generator(inter_channels, resblock, resblock_kernel_sizes, resblock_dilation_sizes, upsample_rates, upsample_initial_channel, upsample_kernel_sizes, gin_channels=gin_channels)
+    self.enc_q = PosteriorEncoder(spec_channels, inter_channels, hidden_channels, 5, 1, 16, gin_channels=gin_channels)
+    self.flow = ResidualCouplingBlock(inter_channels, hidden_channels, 5, 1, 4, gin_channels=gin_channels)
+
+    self.dp = StochasticDurationPredictor(hidden_channels, 192, 3, 0.5, 4, gin_channels=gin_channels)
+
+    if n_speakers > 0:
+      self.emb_g = nn.Embedding(n_speakers, gin_channels)
+
+  def forward(self, x, x_lengths, sid=None, noise_scale=.667, length_scale=1, noise_scale_w=.8, max_len=None):
+    torch.onnx.export(
+        self.enc_p,
+        (x, x_lengths),
+        "ONNX_net/enc_p.onnx",
+        input_names=["x", "x_lengths"],
+        output_names=["xout", "m_p", "logs_p", "x_mask"],
+        dynamic_axes={
+          "x" : [1],
+          "xout" : [2],
+          "m_p" : [2],
+          "logs_p" : [2],
+          "x_mask" : [2]
+        },
+        verbose=True,
+    )
+    x, m_p, logs_p, x_mask = self.enc_p(x, x_lengths)
+
+    if self.n_speakers > 0:
+      g = self.emb_g(sid).unsqueeze(-1) # [b, h, 1]
+    else:
+      g = None
+
+    self.dp.reverse = True
+    self.dp.noise_scale = noise_scale_w
+    torch.onnx.export(
+        self.dp,
+        (x, x_mask, g),
+        "ONNX_net/dp.onnx",
+        input_names=["x", "x_mask", "g"],
+        output_names=["logw"],
+        dynamic_axes={
+          "x" : [2],
+          "x_mask" : [2],
+          "logw" : [2]
+        },
+        verbose=True,
+    )
+    logw = self.dp(x, x_mask, g=g)
+    w = torch.exp(logw) * x_mask * length_scale
+    w_ceil = torch.ceil(w)
+    y_lengths = torch.clamp_min(torch.sum(w_ceil, [1, 2]), 1).long()
+    y_mask = torch.unsqueeze(commons.sequence_mask(y_lengths, None), 1).to(x_mask.dtype)
+    attn_mask = torch.unsqueeze(x_mask, 2) * torch.unsqueeze(y_mask, -1)
+    attn = commons.generate_path(w_ceil, attn_mask)
+
+    m_p = torch.matmul(attn.squeeze(1), m_p.transpose(1, 2)).transpose(1, 2) # [b, t', t], [b, t, d] -> [b, d, t']
+    logs_p = torch.matmul(attn.squeeze(1), logs_p.transpose(1, 2)).transpose(1, 2) # [b, t', t], [b, t, d] -> [b, d, t']
+
+    z_p = m_p + torch.randn_like(m_p) * torch.exp(logs_p) * noise_scale
+
+    self.flow.reverse = True
+    torch.onnx.export(
+        self.flow,
+        (z_p, y_mask, g),
+        "ONNX_net/flow.onnx",
+        input_names=["z_p", "y_mask", "g"],
+        output_names=["z"],
+        dynamic_axes={
+          "z_p" : [2],
+          "y_mask" : [2],
+          "z" : [2]
+        },
+        verbose=True,
+    )
+    z = self.flow(z_p, y_mask, g=g)
+    z_in = (z * y_mask)[:,:,:max_len]
+
+    torch.onnx.export(
+        self.dec,
+        (z_in, g),
+        "ONNX_net/dec.onnx",
+        input_names=["z_in", "g"],
+        output_names=["o"],
+        dynamic_axes={
+          "z_in" : [2],
+          "o" : [2]
+        },
+        verbose=True,
+    )
+    o = self.dec(z_in, g=g)
+    return o

ONNXVITS_modules.py

@@ -0,0 +1,390 @@
+import copy
+import math
+import numpy as np
+import scipy
+import torch
+from torch import nn
+from torch.nn import functional as F
+
+from torch.nn import Conv1d, ConvTranspose1d, AvgPool1d, Conv2d
+from torch.nn.utils import weight_norm, remove_weight_norm
+
+import commons
+from commons import init_weights, get_padding
+from ONNXVITS_transforms import piecewise_rational_quadratic_transform
+
+
+LRELU_SLOPE = 0.1
+
+
+class LayerNorm(nn.Module):
+  def __init__(self, channels, eps=1e-5):
+    super().__init__()
+    self.channels = channels
+    self.eps = eps
+
+    self.gamma = nn.Parameter(torch.ones(channels))
+    self.beta = nn.Parameter(torch.zeros(channels))
+
+  def forward(self, x):
+    x = x.transpose(1, -1)
+    x = F.layer_norm(x, (self.channels,), self.gamma, self.beta, self.eps)
+    return x.transpose(1, -1)
+
+ 
+class ConvReluNorm(nn.Module):
+  def __init__(self, in_channels, hidden_channels, out_channels, kernel_size, n_layers, p_dropout):
+    super().__init__()
+    self.in_channels = in_channels
+    self.hidden_channels = hidden_channels
+    self.out_channels = out_channels
+    self.kernel_size = kernel_size
+    self.n_layers = n_layers
+    self.p_dropout = p_dropout
+    assert n_layers > 1, "Number of layers should be larger than 0."
+
+    self.conv_layers = nn.ModuleList()
+    self.norm_layers = nn.ModuleList()
+    self.conv_layers.append(nn.Conv1d(in_channels, hidden_channels, kernel_size, padding=kernel_size//2))
+    self.norm_layers.append(LayerNorm(hidden_channels))
+    self.relu_drop = nn.Sequential(
+        nn.ReLU(),
+        nn.Dropout(p_dropout))
+    for _ in range(n_layers-1):
+      self.conv_layers.append(nn.Conv1d(hidden_channels, hidden_channels, kernel_size, padding=kernel_size//2))
+      self.norm_layers.append(LayerNorm(hidden_channels))
+    self.proj = nn.Conv1d(hidden_channels, out_channels, 1)
+    self.proj.weight.data.zero_()
+    self.proj.bias.data.zero_()
+
+  def forward(self, x, x_mask):
+    x_org = x
+    for i in range(self.n_layers):
+      x = self.conv_layers[i](x * x_mask)
+      x = self.norm_layers[i](x)
+      x = self.relu_drop(x)
+    x = x_org + self.proj(x)
+    return x * x_mask
+
+
+class DDSConv(nn.Module):
+  """
+  Dialted and Depth-Separable Convolution
+  """
+  def __init__(self, channels, kernel_size, n_layers, p_dropout=0.):
+    super().__init__()
+    self.channels = channels
+    self.kernel_size = kernel_size
+    self.n_layers = n_layers
+    self.p_dropout = p_dropout
+
+    self.drop = nn.Dropout(p_dropout)
+    self.convs_sep = nn.ModuleList()
+    self.convs_1x1 = nn.ModuleList()
+    self.norms_1 = nn.ModuleList()
+    self.norms_2 = nn.ModuleList()
+    for i in range(n_layers):
+      dilation = kernel_size ** i
+      padding = (kernel_size * dilation - dilation) // 2
+      self.convs_sep.append(nn.Conv1d(channels, channels, kernel_size, 
+          groups=channels, dilation=dilation, padding=padding
+      ))
+      self.convs_1x1.append(nn.Conv1d(channels, channels, 1))
+      self.norms_1.append(LayerNorm(channels))
+      self.norms_2.append(LayerNorm(channels))
+
+  def forward(self, x, x_mask, g=None):
+    if g is not None:
+      x = x + g
+    for i in range(self.n_layers):
+      y = self.convs_sep[i](x * x_mask)
+      y = self.norms_1[i](y)
+      y = F.gelu(y)
+      y = self.convs_1x1[i](y)
+      y = self.norms_2[i](y)
+      y = F.gelu(y)
+      y = self.drop(y)
+      x = x + y
+    return x * x_mask
+
+
+class WN(torch.nn.Module):
+  def __init__(self, hidden_channels, kernel_size, dilation_rate, n_layers, gin_channels=0, p_dropout=0):
+    super(WN, self).__init__()
+    assert(kernel_size % 2 == 1)
+    self.hidden_channels =hidden_channels
+    self.kernel_size = kernel_size,
+    self.dilation_rate = dilation_rate
+    self.n_layers = n_layers
+    self.gin_channels = gin_channels
+    self.p_dropout = p_dropout
+
+    self.in_layers = torch.nn.ModuleList()
+    self.res_skip_layers = torch.nn.ModuleList()
+    self.drop = nn.Dropout(p_dropout)
+
+    if gin_channels != 0:
+      cond_layer = torch.nn.Conv1d(gin_channels, 2*hidden_channels*n_layers, 1)
+      self.cond_layer = torch.nn.utils.weight_norm(cond_layer, name='weight')
+
+    for i in range(n_layers):
+      dilation = dilation_rate ** i
+      padding = int((kernel_size * dilation - dilation) / 2)
+      in_layer = torch.nn.Conv1d(hidden_channels, 2*hidden_channels, kernel_size,
+                                 dilation=dilation, padding=padding)
+      in_layer = torch.nn.utils.weight_norm(in_layer, name='weight')
+      self.in_layers.append(in_layer)
+
+      # last one is not necessary
+      if i < n_layers - 1:
+        res_skip_channels = 2 * hidden_channels
+      else:
+        res_skip_channels = hidden_channels
+
+      res_skip_layer = torch.nn.Conv1d(hidden_channels, res_skip_channels, 1)
+      res_skip_layer = torch.nn.utils.weight_norm(res_skip_layer, name='weight')
+      self.res_skip_layers.append(res_skip_layer)
+
+  def forward(self, x, x_mask, g=None, **kwargs):
+    output = torch.zeros_like(x)
+    n_channels_tensor = torch.IntTensor([self.hidden_channels])
+
+    if g is not None:
+      g = self.cond_layer(g)
+
+    for i in range(self.n_layers):
+      x_in = self.in_layers[i](x)
+      if g is not None:
+        cond_offset = i * 2 * self.hidden_channels
+        g_l = g[:,cond_offset:cond_offset+2*self.hidden_channels,:]
+      else:
+        g_l = torch.zeros_like(x_in)
+
+      acts = commons.fused_add_tanh_sigmoid_multiply(
+          x_in,
+          g_l,
+          n_channels_tensor)
+      acts = self.drop(acts)
+
+      res_skip_acts = self.res_skip_layers[i](acts)
+      if i < self.n_layers - 1:
+        res_acts = res_skip_acts[:,:self.hidden_channels,:]
+        x = (x + res_acts) * x_mask
+        output = output + res_skip_acts[:,self.hidden_channels:,:]
+      else:
+        output = output + res_skip_acts
+    return output * x_mask
+
+  def remove_weight_norm(self):
+    if self.gin_channels != 0:
+      torch.nn.utils.remove_weight_norm(self.cond_layer)
+    for l in self.in_layers:
+      torch.nn.utils.remove_weight_norm(l)
+    for l in self.res_skip_layers:
+     torch.nn.utils.remove_weight_norm(l)
+
+
+class ResBlock1(torch.nn.Module):
+    def __init__(self, channels, kernel_size=3, dilation=(1, 3, 5)):
+        super(ResBlock1, self).__init__()
+        self.convs1 = nn.ModuleList([
+            weight_norm(Conv1d(channels, channels, kernel_size, 1, dilation=dilation[0],
+                               padding=get_padding(kernel_size, dilation[0]))),
+            weight_norm(Conv1d(channels, channels, kernel_size, 1, dilation=dilation[1],
+                               padding=get_padding(kernel_size, dilation[1]))),
+            weight_norm(Conv1d(channels, channels, kernel_size, 1, dilation=dilation[2],
+                               padding=get_padding(kernel_size, dilation[2])))
+        ])
+        self.convs1.apply(init_weights)
+
+        self.convs2 = nn.ModuleList([
+            weight_norm(Conv1d(channels, channels, kernel_size, 1, dilation=1,
+                               padding=get_padding(kernel_size, 1))),
+            weight_norm(Conv1d(channels, channels, kernel_size, 1, dilation=1,
+                               padding=get_padding(kernel_size, 1))),
+            weight_norm(Conv1d(channels, channels, kernel_size, 1, dilation=1,
+                               padding=get_padding(kernel_size, 1)))
+        ])
+        self.convs2.apply(init_weights)
+
+    def forward(self, x, x_mask=None):
+        for c1, c2 in zip(self.convs1, self.convs2):
+            xt = F.leaky_relu(x, LRELU_SLOPE)
+            if x_mask is not None:
+                xt = xt * x_mask
+            xt = c1(xt)
+            xt = F.leaky_relu(xt, LRELU_SLOPE)
+            if x_mask is not None:
+                xt = xt * x_mask
+            xt = c2(xt)
+            x = xt + x
+        if x_mask is not None:
+            x = x * x_mask
+        return x
+
+    def remove_weight_norm(self):
+        for l in self.convs1:
+            remove_weight_norm(l)
+        for l in self.convs2:
+            remove_weight_norm(l)
+
+
+class ResBlock2(torch.nn.Module):
+    def __init__(self, channels, kernel_size=3, dilation=(1, 3)):
+        super(ResBlock2, self).__init__()
+        self.convs = nn.ModuleList([
+            weight_norm(Conv1d(channels, channels, kernel_size, 1, dilation=dilation[0],
+                               padding=get_padding(kernel_size, dilation[0]))),
+            weight_norm(Conv1d(channels, channels, kernel_size, 1, dilation=dilation[1],
+                               padding=get_padding(kernel_size, dilation[1])))
+        ])
+        self.convs.apply(init_weights)
+
+    def forward(self, x, x_mask=None):
+        for c in self.convs:
+            xt = F.leaky_relu(x, LRELU_SLOPE)
+            if x_mask is not None:
+                xt = xt * x_mask
+            xt = c(xt)
+            x = xt + x
+        if x_mask is not None:
+            x = x * x_mask
+        return x
+
+    def remove_weight_norm(self):
+        for l in self.convs:
+            remove_weight_norm(l)
+
+
+class Log(nn.Module):
+  def forward(self, x, x_mask, reverse=False, **kwargs):
+    if not reverse:
+      y = torch.log(torch.clamp_min(x, 1e-5)) * x_mask
+      logdet = torch.sum(-y, [1, 2])
+      return y, logdet
+    else:
+      x = torch.exp(x) * x_mask
+      return x
+    
+
+class Flip(nn.Module):
+  def forward(self, x, *args, reverse=False, **kwargs):
+    x = torch.flip(x, [1])
+    if not reverse:
+      logdet = torch.zeros(x.size(0)).to(dtype=x.dtype, device=x.device)
+      return x, logdet
+    else:
+      return x
+
+
+class ElementwiseAffine(nn.Module):
+  def __init__(self, channels):
+    super().__init__()
+    self.channels = channels
+    self.m = nn.Parameter(torch.zeros(channels,1))
+    self.logs = nn.Parameter(torch.zeros(channels,1))
+
+  def forward(self, x, x_mask, reverse=False, **kwargs):
+    if not reverse:
+      y = self.m + torch.exp(self.logs) * x
+      y = y * x_mask
+      logdet = torch.sum(self.logs * x_mask, [1,2])
+      return y, logdet
+    else:
+      x = (x - self.m) * torch.exp(-self.logs) * x_mask
+      return x
+
+
+class ResidualCouplingLayer(nn.Module):
+  def __init__(self,
+      channels,
+      hidden_channels,
+      kernel_size,
+      dilation_rate,
+      n_layers,
+      p_dropout=0,
+      gin_channels=0,
+      mean_only=False):
+    assert channels % 2 == 0, "channels should be divisible by 2"
+    super().__init__()
+    self.channels = channels
+    self.hidden_channels = hidden_channels
+    self.kernel_size = kernel_size
+    self.dilation_rate = dilation_rate
+    self.n_layers = n_layers
+    self.half_channels = channels // 2
+    self.mean_only = mean_only
+
+    self.pre = nn.Conv1d(self.half_channels, hidden_channels, 1)
+    self.enc = WN(hidden_channels, kernel_size, dilation_rate, n_layers, p_dropout=p_dropout, gin_channels=gin_channels)
+    self.post = nn.Conv1d(hidden_channels, self.half_channels * (2 - mean_only), 1)
+    self.post.weight.data.zero_()
+    self.post.bias.data.zero_()
+
+  def forward(self, x, x_mask, g=None, reverse=False):
+    x0, x1 = torch.split(x, [self.half_channels]*2, 1)
+    h = self.pre(x0) * x_mask
+    h = self.enc(h, x_mask, g=g)
+    stats = self.post(h) * x_mask
+    if not self.mean_only:
+      m, logs = torch.split(stats, [self.half_channels]*2, 1)
+    else:
+      m = stats
+      logs = torch.zeros_like(m)
+
+    if not reverse:
+      x1 = m + x1 * torch.exp(logs) * x_mask
+      x = torch.cat([x0, x1], 1)
+      logdet = torch.sum(logs, [1,2])
+      return x, logdet
+    else:
+      x1 = (x1 - m) * torch.exp(-logs) * x_mask
+      x = torch.cat([x0, x1], 1)
+      return x
+
+
+class ConvFlow(nn.Module):
+  def __init__(self, in_channels, filter_channels, kernel_size, n_layers, num_bins=10, tail_bound=5.0):
+    super().__init__()
+    self.in_channels = in_channels
+    self.filter_channels = filter_channels
+    self.kernel_size = kernel_size
+    self.n_layers = n_layers
+    self.num_bins = num_bins
+    self.tail_bound = tail_bound
+    self.half_channels = in_channels // 2
+
+    self.pre = nn.Conv1d(self.half_channels, filter_channels, 1)
+    self.convs = DDSConv(filter_channels, kernel_size, n_layers, p_dropout=0.)
+    self.proj = nn.Conv1d(filter_channels, self.half_channels * (num_bins * 3 - 1), 1)
+    self.proj.weight.data.zero_()
+    self.proj.bias.data.zero_()
+
+  def forward(self, x, x_mask, g=None, reverse=False):
+    x0, x1 = torch.split(x, [self.half_channels]*2, 1)
+    h = self.pre(x0)
+    h = self.convs(h, x_mask, g=g)
+    h = self.proj(h) * x_mask
+
+    b, c, t = x0.shape
+    h = h.reshape(b, c, -1, t).permute(0, 1, 3, 2) # [b, cx?, t] -> [b, c, t, ?]
+
+    unnormalized_widths = h[..., :self.num_bins] / math.sqrt(self.filter_channels)
+    unnormalized_heights = h[..., self.num_bins:2*self.num_bins] / math.sqrt(self.filter_channels)
+    unnormalized_derivatives = h[..., 2 * self.num_bins:]
+
+    x1, logabsdet = piecewise_rational_quadratic_transform(x1,
+        unnormalized_widths,
+        unnormalized_heights,
+        unnormalized_derivatives,
+        inverse=reverse,
+        tails='linear',
+        tail_bound=self.tail_bound
+    )
+
+    x = torch.cat([x0, x1], 1) * x_mask
+    logdet = torch.sum(logabsdet * x_mask, [1,2])
+    if not reverse:
+        return x, logdet
+    else:
+        return x

ONNXVITS_to_onnx.py

@@ -0,0 +1,31 @@
+import ONNXVITS_models
+import utils
+from text import text_to_sequence
+import torch
+import commons
+
+def get_text(text, hps):
+    text_norm = text_to_sequence(text, hps.symbols, hps.data.text_cleaners)
+    if hps.data.add_blank:
+        text_norm = commons.intersperse(text_norm, 0)
+    text_norm = torch.LongTensor(text_norm)
+    return text_norm
+
+hps = utils.get_hparams_from_file("/content/VITS-Umamusume-voice-synthesizer/configs/uma87.json")
+symbols = hps.symbols
+net_g = ONNXVITS_models.SynthesizerTrn(
+    len(symbols),
+    hps.data.filter_length // 2 + 1,
+    hps.train.segment_size // hps.data.hop_length,
+    n_speakers=hps.data.n_speakers,
+    **hps.model)
+_ = net_g.eval()
+_ = utils.load_checkpoint("/content/VITS-Umamusume-voice-synthesizer/pretrained_models/G_jp.pth", net_g)
+
+text1 = get_text("ありがとうございます。", hps)
+stn_tst = text1
+with torch.no_grad():
+    x_tst = stn_tst.unsqueeze(0)
+    x_tst_lengths = torch.LongTensor([stn_tst.size(0)])
+    sid = torch.tensor([0])
+    o = net_g(x_tst, x_tst_lengths, sid=sid, noise_scale=.667, noise_scale_w=0.8, length_scale=1)

ONNXVITS_transforms.py

@@ -0,0 +1,196 @@
+import torch
+from torch.nn import functional as F
+
+import numpy as np
+
+
+DEFAULT_MIN_BIN_WIDTH = 1e-3
+DEFAULT_MIN_BIN_HEIGHT = 1e-3
+DEFAULT_MIN_DERIVATIVE = 1e-3
+
+
+def piecewise_rational_quadratic_transform(inputs, 
+                                           unnormalized_widths,
+                                           unnormalized_heights,
+                                           unnormalized_derivatives,
+                                           inverse=False,
+                                           tails=None, 
+                                           tail_bound=1.,
+                                           min_bin_width=DEFAULT_MIN_BIN_WIDTH,
+                                           min_bin_height=DEFAULT_MIN_BIN_HEIGHT,
+                                           min_derivative=DEFAULT_MIN_DERIVATIVE):
+
+    if tails is None:
+        spline_fn = rational_quadratic_spline
+        spline_kwargs = {}
+    else:
+        spline_fn = unconstrained_rational_quadratic_spline
+        spline_kwargs = {
+            'tails': tails,
+            'tail_bound': tail_bound
+        }
+
+    outputs, logabsdet = spline_fn(
+            inputs=inputs,
+            unnormalized_widths=unnormalized_widths,
+            unnormalized_heights=unnormalized_heights,
+            unnormalized_derivatives=unnormalized_derivatives,
+            inverse=inverse,
+            min_bin_width=min_bin_width,
+            min_bin_height=min_bin_height,
+            min_derivative=min_derivative,
+            **spline_kwargs
+    )
+    return outputs, logabsdet
+
+
+def searchsorted(bin_locations, inputs, eps=1e-6):
+    bin_locations[..., -1] += eps
+    return torch.sum(
+        inputs[..., None] >= bin_locations,
+        dim=-1
+    ) - 1
+
+
+def unconstrained_rational_quadratic_spline(inputs,
+                                            unnormalized_widths,
+                                            unnormalized_heights,
+                                            unnormalized_derivatives,
+                                            inverse=False,
+                                            tails='linear',
+                                            tail_bound=1.,
+                                            min_bin_width=DEFAULT_MIN_BIN_WIDTH,
+                                            min_bin_height=DEFAULT_MIN_BIN_HEIGHT,
+                                            min_derivative=DEFAULT_MIN_DERIVATIVE):
+    inside_interval_mask = (inputs >= -tail_bound) & (inputs <= tail_bound)
+    outside_interval_mask = ~inside_interval_mask
+
+    outputs = torch.zeros_like(inputs)
+    logabsdet = torch.zeros_like(inputs)
+
+    if tails == 'linear':
+        #unnormalized_derivatives = F.pad(unnormalized_derivatives, pad=(1, 1))
+        unnormalized_derivatives_ = torch.zeros((1, 1, unnormalized_derivatives.size(2), unnormalized_derivatives.size(3)+2))
+        unnormalized_derivatives_[...,1:-1] = unnormalized_derivatives
+        unnormalized_derivatives = unnormalized_derivatives_
+        constant = np.log(np.exp(1 - min_derivative) - 1)
+        unnormalized_derivatives[..., 0] = constant
+        unnormalized_derivatives[..., -1] = constant
+
+        outputs[outside_interval_mask] = inputs[outside_interval_mask]
+        logabsdet[outside_interval_mask] = 0
+    else:
+        raise RuntimeError('{} tails are not implemented.'.format(tails))
+
+    outputs[inside_interval_mask], logabsdet[inside_interval_mask] = rational_quadratic_spline(
+        inputs=inputs[inside_interval_mask],
+        unnormalized_widths=unnormalized_widths[inside_interval_mask, :],
+        unnormalized_heights=unnormalized_heights[inside_interval_mask, :],
+        unnormalized_derivatives=unnormalized_derivatives[inside_interval_mask, :],
+        inverse=inverse,
+        left=-tail_bound, right=tail_bound, bottom=-tail_bound, top=tail_bound,
+        min_bin_width=min_bin_width,
+        min_bin_height=min_bin_height,
+        min_derivative=min_derivative
+    )
+
+    return outputs, logabsdet
+
+def rational_quadratic_spline(inputs,
+                              unnormalized_widths,
+                              unnormalized_heights,
+                              unnormalized_derivatives,
+                              inverse=False,
+                              left=0., right=1., bottom=0., top=1.,
+                              min_bin_width=DEFAULT_MIN_BIN_WIDTH,
+                              min_bin_height=DEFAULT_MIN_BIN_HEIGHT,
+                              min_derivative=DEFAULT_MIN_DERIVATIVE):
+    if torch.min(inputs) < left or torch.max(inputs) > right:
+        raise ValueError('Input to a transform is not within its domain')
+
+    num_bins = unnormalized_widths.shape[-1]
+
+    if min_bin_width * num_bins > 1.0:
+        raise ValueError('Minimal bin width too large for the number of bins')
+    if min_bin_height * num_bins > 1.0:
+        raise ValueError('Minimal bin height too large for the number of bins')
+
+    widths = F.softmax(unnormalized_widths, dim=-1)
+    widths = min_bin_width + (1 - min_bin_width * num_bins) * widths
+    cumwidths = torch.cumsum(widths, dim=-1)
+    cumwidths = F.pad(cumwidths, pad=(1, 0), mode='constant', value=0.0)
+    cumwidths = (right - left) * cumwidths + left
+    cumwidths[..., 0] = left
+    cumwidths[..., -1] = right
+    widths = cumwidths[..., 1:] - cumwidths[..., :-1]
+
+    derivatives = min_derivative + F.softplus(unnormalized_derivatives)
+
+    heights = F.softmax(unnormalized_heights, dim=-1)
+    heights = min_bin_height + (1 - min_bin_height * num_bins) * heights
+    cumheights = torch.cumsum(heights, dim=-1)
+    cumheights = F.pad(cumheights, pad=(1, 0), mode='constant', value=0.0)
+    cumheights = (top - bottom) * cumheights + bottom
+    cumheights[..., 0] = bottom
+    cumheights[..., -1] = top
+    heights = cumheights[..., 1:] - cumheights[..., :-1]
+
+    if inverse:
+        bin_idx = searchsorted(cumheights, inputs)[..., None]
+    else:
+        bin_idx = searchsorted(cumwidths, inputs)[..., None]
+
+    input_cumwidths = cumwidths.gather(-1, bin_idx)[..., 0]
+    input_bin_widths = widths.gather(-1, bin_idx)[..., 0]
+
+    input_cumheights = cumheights.gather(-1, bin_idx)[..., 0]
+    delta = heights / widths
+    input_delta = delta.gather(-1, bin_idx)[..., 0]
+
+    input_derivatives = derivatives.gather(-1, bin_idx)[..., 0]
+    input_derivatives_plus_one = derivatives[..., 1:].gather(-1, bin_idx)[..., 0]
+
+    input_heights = heights.gather(-1, bin_idx)[..., 0]
+
+    if inverse:
+        a = (((inputs - input_cumheights) * (input_derivatives
+                                             + input_derivatives_plus_one
+                                             - 2 * input_delta)
+              + input_heights * (input_delta - input_derivatives)))
+        b = (input_heights * input_derivatives
+             - (inputs - input_cumheights) * (input_derivatives
+                                              + input_derivatives_plus_one
+                                              - 2 * input_delta))
+        c = - input_delta * (inputs - input_cumheights)
+
+        discriminant = b.pow(2) - 4 * a * c
+        assert (discriminant >= 0).all()
+
+        root = (2 * c) / (-b - torch.sqrt(discriminant))
+        outputs = root * input_bin_widths + input_cumwidths
+
+        theta_one_minus_theta = root * (1 - root)
+        denominator = input_delta + ((input_derivatives + input_derivatives_plus_one - 2 * input_delta)
+                                     * theta_one_minus_theta)
+        derivative_numerator = input_delta.pow(2) * (input_derivatives_plus_one * root.pow(2)
+                                                     + 2 * input_delta * theta_one_minus_theta
+                                                     + input_derivatives * (1 - root).pow(2))
+        logabsdet = torch.log(derivative_numerator) - 2 * torch.log(denominator)
+
+        return outputs, -logabsdet
+    else:
+        theta = (inputs - input_cumwidths) / input_bin_widths
+        theta_one_minus_theta = theta * (1 - theta)
+
+        numerator = input_heights * (input_delta * theta.pow(2)
+                                     + input_derivatives * theta_one_minus_theta)
+        denominator = input_delta + ((input_derivatives + input_derivatives_plus_one - 2 * input_delta)
+                                     * theta_one_minus_theta)
+        outputs = input_cumheights + numerator / denominator
+
+        derivative_numerator = input_delta.pow(2) * (input_derivatives_plus_one * theta.pow(2)
+                                                     + 2 * input_delta * theta_one_minus_theta
+                                                     + input_derivatives * (1 - theta).pow(2))
+        logabsdet = torch.log(derivative_numerator) - 2 * torch.log(denominator)
+
+        return outputs, logabsdet

ONNXVITS_utils.py

@@ -0,0 +1,21 @@
+import torch
+import numpy as np
+import random
+import onnxruntime as ort
+def set_random_seed(seed=0):
+    ort.set_seed(seed)
+    torch.manual_seed(seed)
+    torch.cuda.manual_seed(seed)
+    torch.backends.cudnn.deterministic = True
+    random.seed(seed)
+    np.random.seed(seed)
+
+def runonnx(model_path, **kwargs):
+    #如使用onnx-gpu，则providers三选一
+    #ort_session = ort.InferenceSession(model_path, providers=['TensorrtExecutionProvider', 'CUDAExecutionProvider', 'CPUExecutionProvider'])
+    ort_session = ort.InferenceSession(model_path)
+    outputs = ort_session.run(
+        None,
+        kwargs
+    )
+    return outputs

ONNX_net/dec.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6abd624175cabf6c8feeec0f505f9059a6c96a02f45927517342e4c6cfe3faae
+size 58181101

ONNX_net/dp.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ddd3bd6f5df7aa651630645be8254866b03092ade79e6e4eb04f25a4cac54807
+size 7686412

ONNX_net/enc_p.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2b4afd64b7edd84c14b222beba8b515a1fd7cba9481a099a3aba76f977191e49
+size 29012524

ONNX_net/flow.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a80128daf8b2a827df84a6d01c3418db330d6c4370cda99f9392cd86defc9b43
+size 35737335

app.py

@@ -1,37 +1,21 @@
-import time
-import matplotlib.pyplot as plt
+import logging
+logging.getLogger('numba').setLevel(logging.WARNING)
 import IPython.display as ipd
-import re
-import os
-import json
-import math
 import torch
-from torch import nn
-from torch.nn import functional as F
-from torch.utils.data import DataLoader
-import gradio as gr
 import commons
 import utils
-from data_utils import TextAudioLoader, TextAudioCollate, TextAudioSpeakerLoader, TextAudioSpeakerCollate
-from models import SynthesizerTrn
+import ONNXVITS_infer
 from text.symbols import symbols
 from text import text_to_sequence
-import unicodedata
-from scipy.io.wavfile import write
+import gradio as gr
+import time
 def get_text(text, hps):
-    text_norm = text_to_sequence(text, hps.data.text_cleaners)
+    text_norm = text_to_sequence(text, symbols, hps.data.text_cleaners)
     if hps.data.add_blank:
         text_norm = commons.intersperse(text_norm, 0)
     text_norm = torch.LongTensor(text_norm)
     return text_norm
 
-
-def get_label(text, label):
-    if f'[{label}]' in text:
-        return True, text.replace(f'[{label}]', '')
-    else:
-        return False, text
-
 def selection(speaker):
     if speaker == "高咲侑":
         spk = 0
@@ -113,30 +97,23 @@ def infer(language,text,speaker_id, n_scale= 0.667,n_scale_w = 0.8, l_scale = 1
     return (hps_ms.data.sampling_rate, audio) 
 lan = ["中文","日文","英文"]
 idols = ["高咲侑","歩夢","かすみ","しずく","果林","愛","せつ菜","璃奈","栞子","エマ","ランジュ","ミア","派蒙"]
-
-
-
 dev = torch.device("cpu")
-hps_ms = utils.get_hparams_from_file("config.json")
-net_g_ms = SynthesizerTrn(
+hps_ms = utils.get_hparams_from_file("lovelive/config.json")
+net_g_ms = ONNXVITS_infer.SynthesizerTrn(
     len(symbols),
     hps_ms.data.filter_length // 2 + 1,
     hps_ms.train.segment_size // hps_ms.data.hop_length,
     n_speakers=hps_ms.data.n_speakers,
-    **hps_ms.model).to(dev)
+    **hps_ms.model)
 _ = net_g_ms.eval()
-
-_ = utils.load_checkpoint("G=768.pth", net_g_ms, None)
+_ = utils.load_checkpoint("lovelive/model.pth", net_g_ms)
 app = gr.Blocks()
-
-
-
 with app:
     with gr.Tabs():
 
         with gr.TabItem("Basic"):
 
-            tts_input1 = gr.TextArea(label="未导出版vits模型", value="一次審査、二次審査、それぞれの欄に記入をお願いします。")
+            tts_input1 = gr.TextArea(label="onnx导出版vits模型", value="一次審査、二次審査、それぞれの欄に記入をお願いします。")
             language = gr.Dropdown(label="选择语言",choices=lan, value="日文", interactive=True)
             para_input1 = gr.Slider(minimum= 0.01,maximum=1.0,label="更改噪声比例", value=0.667)
             para_input2 = gr.Slider(minimum= 0.01,maximum=1.0,label="更改噪声偏差", value=0.8)
@@ -146,4 +123,4 @@ with app:
             tts_output2 = gr.Audio(label="Output")
             tts_submit.click(infer, [language,tts_input1,speaker1,para_input1,para_input2,para_input3], [tts_output2])
     #app.launch(share=True)
-    app.launch()
+    app.launch()

attentions.py

@@ -1,12 +1,9 @@
-import copy
 import math
-import numpy as np
 import torch
 from torch import nn
 from torch.nn import functional as F
 
 import commons
-import modules
 from modules import LayerNorm
    
 

commons.py

@@ -1,8 +1,19 @@
 import math
-import numpy as np
 import torch
-from torch import nn
 from torch.nn import functional as F
+import torch.jit
+
+
+def script_method(fn, _rcb=None):
+  return fn
+
+
+def script(obj, optimize=True, _frames_up=0, _rcb=None):
+  return obj
+
+
+torch.jit.script_method = script_method
+torch.jit.script = script
 
 
 def init_weights(m, mean=0.0, std=0.01):
@@ -15,36 +26,12 @@ def get_padding(kernel_size, dilation=1):
   return int((kernel_size*dilation - dilation)/2)
 
 
-def convert_pad_shape(pad_shape):
-  l = pad_shape[::-1]
-  pad_shape = [item for sublist in l for item in sublist]
-  return pad_shape
-
-
 def intersperse(lst, item):
   result = [item] * (len(lst) * 2 + 1)
   result[1::2] = lst
   return result
 
 
-def kl_divergence(m_p, logs_p, m_q, logs_q):
-  """KL(P||Q)"""
-  kl = (logs_q - logs_p) - 0.5
-  kl += 0.5 * (torch.exp(2. * logs_p) + ((m_p - m_q)**2)) * torch.exp(-2. * logs_q)
-  return kl
-
-
-def rand_gumbel(shape):
-  """Sample from the Gumbel distribution, protect from overflows."""
-  uniform_samples = torch.rand(shape) * 0.99998 + 0.00001
-  return -torch.log(-torch.log(uniform_samples))
-
-
-def rand_gumbel_like(x):
-  g = rand_gumbel(x.size()).to(dtype=x.dtype, device=x.device)
-  return g
-
-
 def slice_segments(x, ids_str, segment_size=4):
   ret = torch.zeros_like(x[:, :, :segment_size])
   for i in range(x.size(0)):
@@ -64,34 +51,6 @@ def rand_slice_segments(x, x_lengths=None, segment_size=4):
   return ret, ids_str
 
 
-def get_timing_signal_1d(
-    length, channels, min_timescale=1.0, max_timescale=1.0e4):
-  position = torch.arange(length, dtype=torch.float)
-  num_timescales = channels // 2
-  log_timescale_increment = (
-      math.log(float(max_timescale) / float(min_timescale)) /
-      (num_timescales - 1))
-  inv_timescales = min_timescale * torch.exp(
-      torch.arange(num_timescales, dtype=torch.float) * -log_timescale_increment)
-  scaled_time = position.unsqueeze(0) * inv_timescales.unsqueeze(1)
-  signal = torch.cat([torch.sin(scaled_time), torch.cos(scaled_time)], 0)
-  signal = F.pad(signal, [0, 0, 0, channels % 2])
-  signal = signal.view(1, channels, length)
-  return signal
-
-
-def add_timing_signal_1d(x, min_timescale=1.0, max_timescale=1.0e4):
-  b, channels, length = x.size()
-  signal = get_timing_signal_1d(length, channels, min_timescale, max_timescale)
-  return x + signal.to(dtype=x.dtype, device=x.device)
-
-
-def cat_timing_signal_1d(x, min_timescale=1.0, max_timescale=1.0e4, axis=1):
-  b, channels, length = x.size()
-  signal = get_timing_signal_1d(length, channels, min_timescale, max_timescale)
-  return torch.cat([x, signal.to(dtype=x.dtype, device=x.device)], axis)
-
-
 def subsequent_mask(length):
   mask = torch.tril(torch.ones(length, length)).unsqueeze(0).unsqueeze(0)
   return mask
@@ -113,11 +72,6 @@ def convert_pad_shape(pad_shape):
   return pad_shape
 
 
-def shift_1d(x):
-  x = F.pad(x, convert_pad_shape([[0, 0], [0, 0], [1, 0]]))[:, :, :-1]
-  return x
-
-
 def sequence_mask(length, max_length=None):
   if max_length is None:
     max_length = length.max()
@@ -141,21 +95,3 @@ def generate_path(duration, mask):
   path = path - F.pad(path, convert_pad_shape([[0, 0], [1, 0], [0, 0]]))[:, :-1]
   path = path.unsqueeze(1).transpose(2,3) * mask
   return path
-
-
-def clip_grad_value_(parameters, clip_value, norm_type=2):
-  if isinstance(parameters, torch.Tensor):
-    parameters = [parameters]
-  parameters = list(filter(lambda p: p.grad is not None, parameters))
-  norm_type = float(norm_type)
-  if clip_value is not None:
-    clip_value = float(clip_value)
-
-  total_norm = 0
-  for p in parameters:
-    param_norm = p.grad.data.norm(norm_type)
-    total_norm += param_norm.item() ** norm_type
-    if clip_value is not None:
-      p.grad.data.clamp_(min=-clip_value, max=clip_value)
-  total_norm = total_norm ** (1. / norm_type)
-  return total_norm

data_utils.py

@@ -186,6 +186,7 @@ class TextAudioSpeakerLoader(torch.utils.data.Dataset):
         audiopaths_sid_text_new = []
         lengths = []
         for audiopath, sid, text in self.audiopaths_sid_text:
+            audiopath = "E:/uma_voice/" + audiopath
             if self.min_text_len <= len(text) and len(text) <= self.max_text_len:
                 audiopaths_sid_text_new.append([audiopath, sid, text])
                 lengths.append(os.path.getsize(audiopath) // (2 * self.hop_length))

export.py

@@ -0,0 +1,38 @@
+import ONNXVITS_models
+import utils
+from text.symbols import symbols
+from text import text_to_sequence
+import torch
+import commons
+
+def get_text(text, hps):
+    text_norm = text_to_sequence(text, symbols, hps.data.text_cleaners)
+    if hps.data.add_blank:
+        text_norm = commons.intersperse(text_norm, 0)
+    text_norm = torch.LongTensor(text_norm)
+    return text_norm
+
+def get_label(text, label):
+    if f'[{label}]' in text:
+        return True, text.replace(f'[{label}]', '')
+    else:
+        return False, text
+
+hps_ms = utils.get_hparams_from_file("/content/drive/MyDrive/moe/config.json")
+net_g_ms = ONNXVITS_models.SynthesizerTrn(
+    len(symbols),
+    hps_ms.data.filter_length // 2 + 1,
+    hps_ms.train.segment_size // hps_ms.data.hop_length,
+    n_speakers=hps_ms.data.n_speakers,
+    **hps_ms.model)
+_ = net_g_ms.eval()
+
+_ = utils.load_checkpoint("/content/drive/MyDrive/moe/G_909000.pth", net_g_ms)
+
+text1 = get_text("[JA]ありがとうございます。[JA]", hps_ms)
+stn_tst = text1
+with torch.no_grad():
+    x_tst = stn_tst.unsqueeze(0)
+    x_tst_lengths = torch.LongTensor([stn_tst.size(0)])
+    sid = torch.tensor([0])
+    o = net_g_ms(x_tst, x_tst_lengths, sid=sid, noise_scale=.667, noise_scale_w=0.8, length_scale=1)

hubert_model.py

@@ -0,0 +1,221 @@
+import copy
+from typing import Optional, Tuple
+import random
+
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+from torch.nn.modules.utils import consume_prefix_in_state_dict_if_present
+
+class Hubert(nn.Module):
+    def __init__(self, num_label_embeddings: int = 100, mask: bool = True):
+        super().__init__()
+        self._mask = mask
+        self.feature_extractor = FeatureExtractor()
+        self.feature_projection = FeatureProjection()
+        self.positional_embedding = PositionalConvEmbedding()
+        self.norm = nn.LayerNorm(768)
+        self.dropout = nn.Dropout(0.1)
+        self.encoder = TransformerEncoder(
+            nn.TransformerEncoderLayer(
+                768, 12, 3072, activation="gelu", batch_first=True
+            ),
+            12,
+        )
+        self.proj = nn.Linear(768, 256)
+
+        self.masked_spec_embed = nn.Parameter(torch.FloatTensor(768).uniform_())
+        self.label_embedding = nn.Embedding(num_label_embeddings, 256)
+
+    def mask(self, x: torch.Tensor) -> Tuple[torch.Tensor, torch.Tensor]:
+        mask = None
+        if self.training and self._mask:
+            mask = _compute_mask((x.size(0), x.size(1)), 0.8, 10, x.device, 2)
+            x[mask] = self.masked_spec_embed.to(x.dtype)
+        return x, mask
+
+    def encode(
+        self, x: torch.Tensor, layer: Optional[int] = None
+    ) -> Tuple[torch.Tensor, torch.Tensor]:
+        x = self.feature_extractor(x)
+        x = self.feature_projection(x.transpose(1, 2))
+        x, mask = self.mask(x)
+        x = x + self.positional_embedding(x)
+        x = self.dropout(self.norm(x))
+        x = self.encoder(x, output_layer=layer)
+        return x, mask
+
+    def logits(self, x: torch.Tensor) -> torch.Tensor:
+        logits = torch.cosine_similarity(
+            x.unsqueeze(2),
+            self.label_embedding.weight.unsqueeze(0).unsqueeze(0),
+            dim=-1,
+        )
+        return logits / 0.1
+
+    def forward(self, x: torch.Tensor) -> Tuple[torch.Tensor, torch.Tensor]:
+        x, mask = self.encode(x)
+        x = self.proj(x)
+        logits = self.logits(x)
+        return logits, mask
+
+
+class HubertSoft(Hubert):
+    def __init__(self):
+        super().__init__()
+
+    @torch.inference_mode()
+    def units(self, wav: torch.Tensor) -> torch.Tensor:
+        wav = F.pad(wav, ((400 - 320) // 2, (400 - 320) // 2))
+        x, _ = self.encode(wav)
+        return self.proj(x)
+
+
+class FeatureExtractor(nn.Module):
+    def __init__(self):
+        super().__init__()
+        self.conv0 = nn.Conv1d(1, 512, 10, 5, bias=False)
+        self.norm0 = nn.GroupNorm(512, 512)
+        self.conv1 = nn.Conv1d(512, 512, 3, 2, bias=False)
+        self.conv2 = nn.Conv1d(512, 512, 3, 2, bias=False)
+        self.conv3 = nn.Conv1d(512, 512, 3, 2, bias=False)
+        self.conv4 = nn.Conv1d(512, 512, 3, 2, bias=False)
+        self.conv5 = nn.Conv1d(512, 512, 2, 2, bias=False)
+        self.conv6 = nn.Conv1d(512, 512, 2, 2, bias=False)
+
+    def forward(self, x: torch.Tensor) -> torch.Tensor:
+        x = F.gelu(self.norm0(self.conv0(x)))
+        x = F.gelu(self.conv1(x))
+        x = F.gelu(self.conv2(x))
+        x = F.gelu(self.conv3(x))
+        x = F.gelu(self.conv4(x))
+        x = F.gelu(self.conv5(x))
+        x = F.gelu(self.conv6(x))
+        return x
+
+
+class FeatureProjection(nn.Module):
+    def __init__(self):
+        super().__init__()
+        self.norm = nn.LayerNorm(512)
+        self.projection = nn.Linear(512, 768)
+        self.dropout = nn.Dropout(0.1)
+
+    def forward(self, x: torch.Tensor) -> torch.Tensor:
+        x = self.norm(x)
+        x = self.projection(x)
+        x = self.dropout(x)
+        return x
+
+
+class PositionalConvEmbedding(nn.Module):
+    def __init__(self):
+        super().__init__()
+        self.conv = nn.Conv1d(
+            768,
+            768,
+            kernel_size=128,
+            padding=128 // 2,
+            groups=16,
+        )
+        self.conv = nn.utils.weight_norm(self.conv, name="weight", dim=2)
+
+    def forward(self, x: torch.Tensor) -> torch.Tensor:
+        x = self.conv(x.transpose(1, 2))
+        x = F.gelu(x[:, :, :-1])
+        return x.transpose(1, 2)
+
+
+class TransformerEncoder(nn.Module):
+    def __init__(
+        self, encoder_layer: nn.TransformerEncoderLayer, num_layers: int
+    ) -> None:
+        super(TransformerEncoder, self).__init__()
+        self.layers = nn.ModuleList(
+            [copy.deepcopy(encoder_layer) for _ in range(num_layers)]
+        )
+        self.num_layers = num_layers
+
+    def forward(
+        self,
+        src: torch.Tensor,
+        mask: torch.Tensor = None,
+        src_key_padding_mask: torch.Tensor = None,
+        output_layer: Optional[int] = None,
+    ) -> torch.Tensor:
+        output = src
+        for layer in self.layers[:output_layer]:
+            output = layer(
+                output, src_mask=mask, src_key_padding_mask=src_key_padding_mask
+            )
+        return output
+
+
+def _compute_mask(
+    shape: Tuple[int, int],
+    mask_prob: float,
+    mask_length: int,
+    device: torch.device,
+    min_masks: int = 0,
+) -> torch.Tensor:
+    batch_size, sequence_length = shape
+
+    if mask_length < 1:
+        raise ValueError("`mask_length` has to be bigger than 0.")
+
+    if mask_length > sequence_length:
+        raise ValueError(
+            f"`mask_length` has to be smaller than `sequence_length`, but got `mask_length`: {mask_length} and `sequence_length`: {sequence_length}`"
+        )
+
+    # compute number of masked spans in batch
+    num_masked_spans = int(mask_prob * sequence_length / mask_length + random.random())
+    num_masked_spans = max(num_masked_spans, min_masks)
+
+    # make sure num masked indices <= sequence_length
+    if num_masked_spans * mask_length > sequence_length:
+        num_masked_spans = sequence_length // mask_length
+
+    # SpecAugment mask to fill
+    mask = torch.zeros((batch_size, sequence_length), device=device, dtype=torch.bool)
+
+    # uniform distribution to sample from, make sure that offset samples are < sequence_length
+    uniform_dist = torch.ones(
+        (batch_size, sequence_length - (mask_length - 1)), device=device
+    )
+
+    # get random indices to mask
+    mask_indices = torch.multinomial(uniform_dist, num_masked_spans)
+
+    # expand masked indices to masked spans
+    mask_indices = (
+        mask_indices.unsqueeze(dim=-1)
+        .expand((batch_size, num_masked_spans, mask_length))
+        .reshape(batch_size, num_masked_spans * mask_length)
+    )
+    offsets = (
+        torch.arange(mask_length, device=device)[None, None, :]
+        .expand((batch_size, num_masked_spans, mask_length))
+        .reshape(batch_size, num_masked_spans * mask_length)
+    )
+    mask_idxs = mask_indices + offsets
+
+    # scatter indices to mask
+    mask = mask.scatter(1, mask_idxs, True)
+
+    return mask
+
+
+def hubert_soft(
+    path: str
+) -> HubertSoft:
+    r"""HuBERT-Soft from `"A Comparison of Discrete and Soft Speech Units for Improved Voice Conversion"`.
+    Args:
+        path (str): path of a pretrained model
+    """
+    hubert = HubertSoft()
+    checkpoint = torch.load(path)
+    consume_prefix_in_state_dict_if_present(checkpoint, "module.")
+    hubert.load_state_dict(checkpoint)
+    hubert.eval()
+    return hubert

lovelive/config.json

@@ -0,0 +1,54 @@
+{
+  "train": {
+    "log_interval": 200,
+    "eval_interval": 1000,
+    "seed": 1234,
+    "epochs": 10000,
+    "learning_rate": 2e-4,
+    "betas": [0.8, 0.99],
+    "eps": 1e-9,
+    "batch_size": 32,
+    "fp16_run": true,
+    "lr_decay": 0.999875,
+    "segment_size": 8192,
+    "init_lr_ratio": 1,
+    "warmup_epochs": 0,
+    "c_mel": 45,
+    "c_kl": 1.0
+  },
+  "data": {
+    "training_files":"/www/training/dataset/train_with_paimeng2.txt",
+    "validation_files":"/www/training/dataset/val_filelist.txt",
+    "text_cleaners":["cjke_cleaners"],
+    "max_wav_value": 32768.0,
+    "sampling_rate": 22050,
+    "filter_length": 1024,
+    "hop_length": 256,
+    "win_length": 1024,
+    "n_mel_channels": 80,
+    "mel_fmin": 0.0,
+    "mel_fmax": null,
+    "add_blank": true,
+    "n_speakers": 50,
+    "cleaned_text": true
+  },
+  "model": {
+    "inter_channels": 192,
+    "hidden_channels": 192,
+    "filter_channels": 768,
+    "n_heads": 2,
+    "n_layers": 6,
+    "kernel_size": 3,
+    "p_dropout": 0.1,
+    "resblock": "1",
+    "resblock_kernel_sizes": [3,7,11],
+    "resblock_dilation_sizes": [[1,3,5], [1,3,5], [1,3,5]],
+    "upsample_rates": [8,8,2,2],
+    "upsample_initial_channel": 512,
+    "upsample_kernel_sizes": [16,16,4,4],
+    "n_layers_q": 3,
+    "use_spectral_norm": false,
+    "gin_channels": 256
+  },
+  "symbols": ["_", ",", ".", "!", "?", "-", "~", "\u2026", "A", "E", "I", "N", "O", "Q", "U", "a", "b", "d", "e", "f", "g", "h", "i", "j", "k", "l", "m", "n", "o", "p", "r", "s", "t", "u", "v", "w", "y", "z", "\u0283", "\u02a7", "\u02a6", "\u026f", "\u0279", "\u0259", "\u0265", "\u207c", "\u02b0", "`", "\u2192", "\u2193", "\u2191", " "]
+}

lovelive/model.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bb75bd88a1f03610a8f9511918b4424e1dca77b59a0c378d6ed77e40a89d5678
+size 476964251

mel_processing.py

@@ -1,16 +1,5 @@
-import math
-import os
-import random
 import torch
-from torch import nn
-import torch.nn.functional as F
 import torch.utils.data
-import numpy as np
-import librosa
-import librosa.util as librosa_util
-from librosa.util import normalize, pad_center, tiny
-from scipy.signal import get_window
-from scipy.io.wavfile import read
 from librosa.filters import mel as librosa_mel_fn
 
 MAX_WAV_VALUE = 32768.0
@@ -64,7 +53,7 @@ def spectrogram_torch(y, n_fft, sampling_rate, hop_size, win_size, center=False)
     y = y.squeeze(1)
 
     spec = torch.stft(y, n_fft, hop_length=hop_size, win_length=win_size, window=hann_window[wnsize_dtype_device],
-                      center=center, pad_mode='reflect', normalized=False, onesided=True)
+                      center=center, pad_mode='reflect', normalized=False, onesided=True, return_complex=False)
 
     spec = torch.sqrt(spec.pow(2).sum(-1) + 1e-6)
     return spec

models.py

@@ -1,4 +1,3 @@
-import copy
 import math
 import torch
 from torch import nn
@@ -7,9 +6,8 @@ from torch.nn import functional as F
 import commons
 import modules
 import attentions
-import monotonic_align
 
-from torch.nn import Conv1d, ConvTranspose1d, AvgPool1d, Conv2d
+from torch.nn import Conv1d, ConvTranspose1d, Conv2d
 from torch.nn.utils import weight_norm, remove_weight_norm, spectral_norm
 from commons import init_weights, get_padding
 
@@ -141,7 +139,8 @@ class TextEncoder(nn.Module):
       n_heads,
       n_layers,
       kernel_size,
-      p_dropout):
+      p_dropout,
+      emotion_embedding):
     super().__init__()
     self.n_vocab = n_vocab
     self.out_channels = out_channels
@@ -151,9 +150,13 @@ class TextEncoder(nn.Module):
     self.n_layers = n_layers
     self.kernel_size = kernel_size
     self.p_dropout = p_dropout
-
-    self.emb = nn.Embedding(n_vocab, hidden_channels)
-    nn.init.normal_(self.emb.weight, 0.0, hidden_channels**-0.5)
+    self.emotion_embedding = emotion_embedding
+    
+    if self.n_vocab!=0:
+      self.emb = nn.Embedding(n_vocab, hidden_channels)
+      if emotion_embedding:
+        self.emotion_emb = nn.Linear(1024, hidden_channels)
+      nn.init.normal_(self.emb.weight, 0.0, hidden_channels**-0.5)
 
     self.encoder = attentions.Encoder(
       hidden_channels,
@@ -164,8 +167,11 @@ class TextEncoder(nn.Module):
       p_dropout)
     self.proj= nn.Conv1d(hidden_channels, out_channels * 2, 1)
 
-  def forward(self, x, x_lengths):
-    x = self.emb(x) * math.sqrt(self.hidden_channels) # [b, t, h]
+  def forward(self, x, x_lengths, emotion_embedding=None):
+    if self.n_vocab!=0:
+      x = self.emb(x) * math.sqrt(self.hidden_channels) # [b, t, h]
+    if emotion_embedding is not None:
+      x = x + self.emotion_emb(emotion_embedding.unsqueeze(1))
     x = torch.transpose(x, 1, -1) # [b, h, t]
     x_mask = torch.unsqueeze(commons.sequence_mask(x_lengths, x.size(2)), 1).to(x.dtype)
 
@@ -412,6 +418,7 @@ class SynthesizerTrn(nn.Module):
     n_speakers=0,
     gin_channels=0,
     use_sdp=True,
+    emotion_embedding=False,
     **kwargs):
 
     super().__init__()
@@ -443,7 +450,8 @@ class SynthesizerTrn(nn.Module):
         n_heads,
         n_layers,
         kernel_size,
-        p_dropout)
+        p_dropout,
+        emotion_embedding)
     self.dec = Generator(inter_channels, resblock, resblock_kernel_sizes, resblock_dilation_sizes, upsample_rates, upsample_initial_channel, upsample_kernel_sizes, gin_channels=gin_channels)
     self.enc_q = PosteriorEncoder(spec_channels, inter_channels, hidden_channels, 5, 1, 16, gin_channels=gin_channels)
     self.flow = ResidualCouplingBlock(inter_channels, hidden_channels, 5, 1, 4, gin_channels=gin_channels)
@@ -496,8 +504,8 @@ class SynthesizerTrn(nn.Module):
     o = self.dec(z_slice, g=g)
     return o, l_length, attn, ids_slice, x_mask, y_mask, (z, z_p, m_p, logs_p, m_q, logs_q)
 
-  def infer(self, x, x_lengths, sid=None, noise_scale=1, length_scale=1, noise_scale_w=1., max_len=None):
-    x, m_p, logs_p, x_mask = self.enc_p(x, x_lengths)
+  def infer(self, x, x_lengths, sid=None, noise_scale=1, length_scale=1, noise_scale_w=1., max_len=None, emotion_embedding=None):
+    x, m_p, logs_p, x_mask = self.enc_p(x, x_lengths, emotion_embedding)
     if self.n_speakers > 0:
       g = self.emb_g(sid).unsqueeze(-1) # [b, h, 1]
     else:

modules.py

@@ -1,12 +1,9 @@
-import copy
 import math
-import numpy as np
-import scipy
 import torch
 from torch import nn
 from torch.nn import functional as F
 
-from torch.nn import Conv1d, ConvTranspose1d, AvgPool1d, Conv2d
+from torch.nn import Conv1d
 from torch.nn.utils import weight_norm, remove_weight_norm
 
 import commons

monotonic_align/__init__.py

@@ -1,19 +1,19 @@
-from numpy import zeros, int32, float32
-from torch import from_numpy
+import numpy as np
+import torch
+from .monotonic_align.core import maximum_path_c
 
-from .core import maximum_path_jit
 
 def maximum_path(neg_cent, mask):
-  """ numba optimized version.
+  """ Cython optimized version.
   neg_cent: [b, t_t, t_s]
   mask: [b, t_t, t_s]
   """
   device = neg_cent.device
   dtype = neg_cent.dtype
-  neg_cent = neg_cent.data.cpu().numpy().astype(float32)
-  path = zeros(neg_cent.shape, dtype=int32)
+  neg_cent = neg_cent.data.cpu().numpy().astype(np.float32)
+  path = np.zeros(neg_cent.shape, dtype=np.int32)
 
-  t_t_max = mask.sum(1)[:, 0].data.cpu().numpy().astype(int32)
-  t_s_max = mask.sum(2)[:, 0].data.cpu().numpy().astype(int32)
-  maximum_path_jit(path, neg_cent, t_t_max, t_s_max)
-  return from_numpy(path).to(device=device, dtype=dtype)
+  t_t_max = mask.sum(1)[:, 0].data.cpu().numpy().astype(np.int32)
+  t_s_max = mask.sum(2)[:, 0].data.cpu().numpy().astype(np.int32)
+  maximum_path_c(path, neg_cent, t_t_max, t_s_max)
+  return torch.from_numpy(path).to(device=device, dtype=dtype)

monotonic_align/build/temp.linux-x86_64-3.8/core.o

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0f5fa9684a7bd75866565d56054169d2cf02b02089e1468bc3a46e4cb2360713
+size 2936128

monotonic_align/core.pyx

@@ -1,5 +1,3 @@
-# cython:language_level=3
-
 cimport cython
 from cython.parallel import prange
 

monotonic_align/monotonic_align/core.cpython-38-x86_64-linux-gnu.so

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1795a1ac1e417f54a90339bc9f65e5e12e2a59b655a741ce9986bb29bd77c53d
+size 1182504

monotonic_align/setup.py

@@ -1,4 +1,3 @@
-# cython:language_level=3
 from distutils.core import setup
 from Cython.Build import cythonize
 import numpy

requirements.txt

@@ -1,22 +1,28 @@
-Cython==0.29.21
-librosa==0.8.0
-matplotlib==3.3.1
-numpy==1.21.6
-phonemizer==2.2.1
-scipy==1.5.2
-tensorboard==2.3.0
+numba
+librosa
+matplotlib
+numpy
+phonemizer
+scipy
+tensorboard
 torch
 torchvision
-Unidecode==1.1.1
-pyopenjtalk==0.2.0
-jamo==0.4.1
-pypinyin==0.44.0
-jieba==0.42.1
-cn2an==0.5.17
-jieba==0.42.1
-ipython==7.34.0
-gradio==3.4.1
-openai
-pydub
+torchaudio
+unidecode
+pyopenjtalk>=0.3.0
+jamo
+pypinyin
+ko_pron
+jieba
+cn2an
+protobuf
 inflect
-eng_to_ipa
+eng_to_ipa
+ko_pron
+indic_transliteration
+num_thai
+opencc
+onnx
+onnxruntime
+psutil
+gradio

text/__init__.py

@@ -1,14 +1,8 @@
 """ from https://github.com/keithito/tacotron """
 from text import cleaners
-from text.symbols import symbols
 
 
-# Mappings from symbol to numeric ID and vice versa:
-_symbol_to_id = {s: i for i, s in enumerate(symbols)}
-_id_to_symbol = {i: s for i, s in enumerate(symbols)}
-
-
-def text_to_sequence(text, cleaner_names):
+def text_to_sequence(text, symbols, cleaner_names):
   '''Converts a string of text to a sequence of IDs corresponding to the symbols in the text.
     Args:
       text: string to convert to a sequence
@@ -16,6 +10,8 @@ def text_to_sequence(text, cleaner_names):
     Returns:
       List of integers corresponding to the symbols in the text
   '''
+  _symbol_to_id = {s: i for i, s in enumerate(symbols)}
+
   sequence = []
 
   clean_text = _clean_text(text, cleaner_names)
@@ -27,26 +23,6 @@ def text_to_sequence(text, cleaner_names):
   return sequence
 
 
-def cleaned_text_to_sequence(cleaned_text):
-  '''Converts a string of text to a sequence of IDs corresponding to the symbols in the text.
-    Args:
-      text: string to convert to a sequence
-    Returns:
-      List of integers corresponding to the symbols in the text
-  '''
-  sequence = [_symbol_to_id[symbol] for symbol in cleaned_text if symbol in _symbol_to_id.keys()]
-  return sequence
-
-
-def sequence_to_text(sequence):
-  '''Converts a sequence of IDs back to a string'''
-  result = ''
-  for symbol_id in sequence:
-    s = _id_to_symbol[symbol_id]
-    result += s
-  return result
-
-
 def _clean_text(text, cleaner_names):
   for name in cleaner_names:
     cleaner = getattr(cleaners, name)