merge master

server/const.py

@@ -66,6 +66,7 @@ class EnumInferenceTypes(Enum):
     pyTorchRVCv2Nono = "pyTorchRVCv2Nono"
     pyTorchWebUI = "pyTorchWebUI"
     pyTorchWebUINono = "pyTorchWebUINono"
+    pyTorchVoRASbeta = "pyTorchVoRASbeta"
     onnxRVC = "onnxRVC"
     onnxRVCNono = "onnxRVCNono"
 

server/voice_changer/RVC/RVCModelSlotGenerator.py

@@ -37,7 +37,30 @@ class RVCModelSlotGenerator(ModelSlotGenerator):
         cpt = torch.load(slot.modelFile, map_location="cpu")
         config_len = len(cpt["config"])
         
-        if config_len == 18:
+        print(cpt["version"])
+        if cpt["version"] == "voras_beta":
+            slot.f0 = True if cpt["f0"] == 1 else False
+            slot.modelType = EnumInferenceTypes.pyTorchVoRASbeta.value
+            slot.embChannels = 768
+            slot.embOutputLayer = (
+                cpt["embedder_output_layer"] if "embedder_output_layer" in cpt else 9
+            )
+            slot.useFinalProj = False
+
+            slot.embedder = cpt["embedder_name"]
+            if slot.embedder.endswith("768"):
+                slot.embedder = slot.embedder[:-3]
+
+            if slot.embedder == EnumEmbedderTypes.hubert.value:
+                slot.embedder = EnumEmbedderTypes.hubert.value
+            elif slot.embedder == EnumEmbedderTypes.contentvec.value:
+                slot.embedder = EnumEmbedderTypes.contentvec.value
+            elif slot.embedder == EnumEmbedderTypes.hubert_jp.value:
+                slot.embedder = EnumEmbedderTypes.hubert_jp.value
+            else:
+                raise RuntimeError("[Voice Changer][setInfoByONNX] unknown embedder")
+
+        elif config_len == 18:
             # Original RVC
             slot.f0 = True if cpt["f0"] == 1 else False
             version = cpt.get("version", "v1")

server/voice_changer/RVC/inferencer/InferencerManager.py

@@ -8,7 +8,7 @@ from voice_changer.RVC.inferencer.RVCInferencerv2 import RVCInferencerv2
 from voice_changer.RVC.inferencer.RVCInferencerv2Nono import RVCInferencerv2Nono
 from voice_changer.RVC.inferencer.WebUIInferencer import WebUIInferencer
 from voice_changer.RVC.inferencer.WebUIInferencerNono import WebUIInferencerNono
-
+from voice_changer.RVC.inferencer.VorasInferencebeta import VoRASInferencer
 
 class InferencerManager:
     currentInferencer: Inferencer | None = None
@@ -37,6 +37,8 @@ class InferencerManager:
             return RVCInferencerNono().loadModel(file, gpu)
         elif inferencerType == EnumInferenceTypes.pyTorchRVCv2 or inferencerType == EnumInferenceTypes.pyTorchRVCv2.value:
             return RVCInferencerv2().loadModel(file, gpu)
+        elif inferencerType == EnumInferenceTypes.pyTorchVoRASbeta or inferencerType == EnumInferenceTypes.pyTorchVoRASbeta.value:
+            return VoRASInferencer().loadModel(file, gpu)
         elif inferencerType == EnumInferenceTypes.pyTorchRVCv2Nono or inferencerType == EnumInferenceTypes.pyTorchRVCv2Nono.value:
             return RVCInferencerv2Nono().loadModel(file, gpu)
         elif inferencerType == EnumInferenceTypes.pyTorchWebUI or inferencerType == EnumInferenceTypes.pyTorchWebUI.value:

server/voice_changer/RVC/inferencer/VorasInferencebeta.py

@@ -0,0 +1,39 @@
+import torch
+from torch import device
+
+from const import EnumInferenceTypes
+from voice_changer.RVC.inferencer.Inferencer import Inferencer
+from voice_changer.RVC.deviceManager.DeviceManager import DeviceManager
+from .voras_beta.models import Synthesizer
+
+
+class VoRASInferencer(Inferencer):
+    def loadModel(self, file: str, gpu: device):
+        super().setProps(EnumInferenceTypes.pyTorchVoRASbeta, file, False, gpu)
+
+        dev = DeviceManager.get_instance().getDevice(gpu)
+        self.isHalf = False # DeviceManager.get_instance().halfPrecisionAvailable(gpu)
+
+        cpt = torch.load(file, map_location="cpu")
+        model = Synthesizer(**cpt["params"])
+
+        model.eval()
+        model.load_state_dict(cpt["weight"], strict=False)
+        model.remove_weight_norm()
+        model.change_speaker(0)
+
+        model = model.to(dev)
+
+        self.model = model
+        print("load model comprete")
+        return self
+
+    def infer(
+        self,
+        feats: torch.Tensor,
+        pitch_length: torch.Tensor,
+        pitch: torch.Tensor,
+        pitchf: torch.Tensor,
+        sid: torch.Tensor,
+    ) -> torch.Tensor:
+        return self.model.infer(feats, pitch_length, pitch, pitchf, sid)

server/voice_changer/RVC/inferencer/voras_beta/commons.py

@@ -0,0 +1,165 @@
+import math
+
+import torch
+from torch.nn import functional as F
+
+
+def init_weights(m, mean=0.0, std=0.01):
+    classname = m.__class__.__name__
+    if classname.find("Conv") != -1:
+        m.weight.data.normal_(mean, std)
+
+
+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 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(2.0 * logs_p) + ((m_p - m_q) ** 2)) * torch.exp(-2.0 * 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)):
+        idx_str = ids_str[i]
+        idx_end = idx_str + segment_size
+        r = x[i, :, idx_str:idx_end]
+        ret[i, :, :r.size(1)] = r
+    return ret
+
+
+def slice_segments2(x, ids_str, segment_size=4):
+    ret = torch.zeros_like(x[:, :segment_size])
+    for i in range(x.size(0)):
+        idx_str = ids_str[i]
+        idx_end = idx_str + segment_size
+        r = x[i, idx_str:idx_end]
+        ret[i, :r.size(0)] = r
+    return ret
+
+
+def rand_slice_segments(x, x_lengths, segment_size=4, ids_str=None):
+    b, d, t = x.size()
+    if ids_str is None:
+        ids_str = torch.zeros([b]).to(device=x.device, dtype=x_lengths.dtype)
+    ids_str_max = torch.maximum(torch.zeros_like(x_lengths).to(device=x_lengths.device ,dtype=x_lengths.dtype), x_lengths - segment_size + 1 - ids_str)
+    ids_str += (torch.rand([b]).to(device=x.device) * ids_str_max).to(dtype=torch.long)
+    ret = slice_segments(x, ids_str, segment_size)
+    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
+
+
+@torch.jit.script
+def fused_add_tanh_sigmoid_multiply(input_a, input_b, n_channels):
+    n_channels_int = n_channels[0]
+    in_act = input_a + input_b
+    t_act = torch.tanh(in_act[:, :n_channels_int, :])
+    s_act = torch.sigmoid(in_act[:, n_channels_int:, :])
+    acts = t_act * s_act
+    return acts
+
+
+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 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()
+    x = torch.arange(max_length, dtype=length.dtype, device=length.device)
+    return x.unsqueeze(0) < length.unsqueeze(1)
+
+
+def generate_path(duration, mask):
+    """
+    duration: [b, 1, t_x]
+    mask: [b, 1, t_y, t_x]
+    """
+    b, _, t_y, t_x = mask.shape
+    cum_duration = torch.cumsum(duration, -1)
+
+    cum_duration_flat = cum_duration.view(b * t_x)
+    path = sequence_mask(cum_duration_flat, t_y).to(mask.dtype)
+    path = path.view(b, t_x, t_y)
+    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 ** (1.0 / norm_type)
+    return total_norm

server/voice_changer/RVC/inferencer/voras_beta/config.py

@@ -0,0 +1,61 @@
+from typing import *
+
+from pydantic import BaseModel
+
+
+class TrainConfigTrain(BaseModel):
+    log_interval: int
+    seed: int
+    epochs: int
+    learning_rate: float
+    betas: List[float]
+    eps: float
+    batch_size: int
+    fp16_run: bool
+    lr_decay: float
+    segment_size: int
+    init_lr_ratio: int
+    warmup_epochs: int
+    c_mel: int
+    c_kl: float
+
+
+class TrainConfigData(BaseModel):
+    max_wav_value: float
+    sampling_rate: int
+    filter_length: int
+    hop_length: int
+    win_length: int
+    n_mel_channels: int
+    mel_fmin: float
+    mel_fmax: Any
+
+
+class TrainConfigModel(BaseModel):
+    emb_channels: int
+    inter_channels: int
+    n_layers: int
+    upsample_rates: List[int]
+    use_spectral_norm: bool
+    gin_channels: int
+    spk_embed_dim: int
+
+
+class TrainConfig(BaseModel):
+    version: Literal["voras"] = "voras"
+    train: TrainConfigTrain
+    data: TrainConfigData
+    model: TrainConfigModel
+
+
+class DatasetMetaItem(BaseModel):
+    gt_wav: str
+    co256: str
+    f0: Optional[str]
+    f0nsf: Optional[str]
+    speaker_id: int
+
+
+class DatasetMetadata(BaseModel):
+    files: Dict[str, DatasetMetaItem]
+    # mute: DatasetMetaItem

server/voice_changer/RVC/inferencer/voras_beta/models.py

@@ -0,0 +1,238 @@
+import math
+import os
+import sys
+
+import numpy as np
+import torch
+from torch import nn
+from torch.nn import Conv2d
+from torch.nn import functional as F
+from torch.nn.utils import remove_weight_norm, spectral_norm, weight_norm
+
+from . import commons, modules
+from .commons import get_padding
+from .modules import (ConvNext2d, HarmonicEmbedder, IMDCTSymExpHead,
+                      LoRALinear1d, SnakeFilter, WaveBlock)
+
+parent_dir = os.path.dirname(os.path.dirname(os.path.abspath(__file__)))
+sys.path.append(parent_dir)
+
+sr2sr = {
+    "24k": 24000,
+    "32k": 32000,
+    "40k": 40000,
+    "48k": 48000,
+}
+
+class GeneratorVoras(torch.nn.Module):
+    def __init__(
+        self,
+        emb_channels,
+        inter_channels,
+        gin_channels,
+        n_layers,
+        sr,
+        hop_length,
+    ):
+        super(GeneratorVoras, self).__init__()
+        self.n_layers = n_layers
+        self.emb_pitch = HarmonicEmbedder(768, inter_channels, gin_channels, 16, 15)  #   # pitch 256
+        self.plinear = LoRALinear1d(inter_channels, inter_channels, gin_channels, r=8)
+        self.glinear = weight_norm(nn.Conv1d(gin_channels, inter_channels, 1))
+        self.resblocks = nn.ModuleList()
+        self.init_linear = LoRALinear1d(emb_channels, inter_channels, gin_channels, r=4)
+        for _ in range(self.n_layers):
+            self.resblocks.append(WaveBlock(inter_channels, gin_channels, [9] * 2, [1] * 2, [1, 9], 2, r=4))
+        self.head = IMDCTSymExpHead(inter_channels, gin_channels, hop_length, padding="center", sample_rate=sr)
+        self.post = SnakeFilter(4, 8, 9, 2, eps=1e-5)
+
+    def forward(self, x, pitchf, x_mask, g):
+        x = self.init_linear(x, g) + self.plinear(self.emb_pitch(pitchf, g), g) + self.glinear(g)
+        for i in range(self.n_layers):
+            x = self.resblocks[i](x, x_mask, g)
+        x = x * x_mask
+        x = self.head(x, g)
+        x = self.post(x)
+        return torch.tanh(x)
+
+    def remove_weight_norm(self):
+        self.plinear.remove_weight_norm()
+        remove_weight_norm(self.glinear)
+        for l in self.resblocks:
+            l.remove_weight_norm()
+        self.init_linear.remove_weight_norm()
+        self.head.remove_weight_norm()
+        self.post.remove_weight_norm()
+
+    def fix_speaker(self, g):
+        self.plinear.fix_speaker(g)
+        self.init_linear.fix_speaker(g)
+        for l in self.resblocks:
+            l.fix_speaker(g)
+        self.head.fix_speaker(g)
+
+    def unfix_speaker(self, g):
+        self.plinear.unfix_speaker(g)
+        self.init_linear.unfix_speaker(g)
+        for l in self.resblocks:
+            l.unfix_speaker(g)
+        self.head.unfix_speaker(g)
+
+
+class Synthesizer(nn.Module):
+    def __init__(
+        self,
+        segment_size,
+        n_fft,
+        hop_length,
+        inter_channels,
+        n_layers,
+        spk_embed_dim,
+        gin_channels,
+        emb_channels,
+        sr,
+        **kwargs
+    ):
+        super().__init__()
+        if type(sr) == type("strr"):
+            sr = sr2sr[sr]
+        self.segment_size = segment_size
+        self.n_fft = n_fft
+        self.hop_length = hop_length
+        self.inter_channels = inter_channels
+        self.n_layers = n_layers
+        self.spk_embed_dim = spk_embed_dim
+        self.gin_channels = gin_channels
+        self.emb_channels = emb_channels
+        self.sr = sr
+
+        self.dec = GeneratorVoras(
+            emb_channels,
+            inter_channels,
+            gin_channels,
+            n_layers,
+            sr,
+            hop_length
+        )
+
+        self.emb_g = nn.Embedding(self.spk_embed_dim, gin_channels)
+        print(
+            "gin_channels:",
+            gin_channels,
+            "self.spk_embed_dim:",
+            self.spk_embed_dim,
+            "emb_channels:",
+            emb_channels,
+        )
+        self.speaker = None
+
+    def remove_weight_norm(self):
+        self.dec.remove_weight_norm()
+
+    def change_speaker(self, sid: int):
+        if self.speaker is not None:
+            g = self.emb_g(torch.from_numpy(np.array(self.speaker))).unsqueeze(-1)
+            self.dec.unfix_speaker(g)
+        g = self.emb_g(torch.from_numpy(np.array(sid))).unsqueeze(-1)
+        self.dec.fix_speaker(g)
+        self.speaker = sid
+
+    def forward(
+        self, phone, phone_lengths, pitch, pitchf, ds
+        ):
+        g = self.emb_g(ds).unsqueeze(-1)
+        x = torch.transpose(phone, 1, -1)
+        x_mask = torch.unsqueeze(commons.sequence_mask(phone_lengths, x.size(2)), 1).to(phone.dtype)
+        x_slice, ids_slice = commons.rand_slice_segments(
+            x, phone_lengths, self.segment_size
+        )
+        pitchf_slice = commons.slice_segments2(pitchf, ids_slice, self.segment_size)
+        mask_slice = commons.slice_segments(x_mask, ids_slice, self.segment_size)
+        o = self.dec(x_slice, pitchf_slice, mask_slice, g)
+        return o, ids_slice, x_mask, g
+
+    def infer(self, phone, phone_lengths, pitch, nsff0, sid, max_len=None):
+        g = self.emb_g(sid).unsqueeze(-1)
+        x = torch.transpose(phone, 1, -1)
+        x_mask = torch.unsqueeze(commons.sequence_mask(phone_lengths, x.size(2)), 1).to(phone.dtype)
+        o = self.dec((x * x_mask)[:, :, :max_len], nsff0, x_mask, g)
+        return o, x_mask, (None, None, None, None)
+
+
+class DiscriminatorP(torch.nn.Module):
+    def __init__(self, period, gin_channels, upsample_rates, final_dim=256, use_spectral_norm=False):
+        super(DiscriminatorP, self).__init__()
+        self.period = period
+        self.use_spectral_norm = use_spectral_norm
+        self.init_kernel_size = upsample_rates[-1] * 3
+        norm_f = weight_norm if use_spectral_norm == False else spectral_norm
+        N = len(upsample_rates)
+        self.init_conv = norm_f(Conv2d(1, final_dim // (2 ** (N - 1)), (self.init_kernel_size, 1), (upsample_rates[-1], 1)))
+        self.convs = nn.ModuleList()
+        for i, u in enumerate(upsample_rates[::-1][1:], start=1):
+            self.convs.append(
+                ConvNext2d(
+                    final_dim // (2 ** (N - i)),
+                    final_dim // (2 ** (N - i - 1)),
+                    gin_channels,
+                    (u*3, 1),
+                    (u, 1),
+                    4,
+                    r=2 + i//2
+                )
+            )
+        self.conv_post = weight_norm(Conv2d(final_dim, 1, (3, 1), (1, 1)))
+
+    def forward(self, x, g):
+        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, (n_pad, 0), "reflect")
+            t = t + n_pad
+        x = x.view(b, c, t // self.period, self.period)
+
+        x = torch.flip(x, dims=[2])
+        x = F.pad(x, [0, 0, 0, self.init_kernel_size - 1], mode="constant")
+        x = self.init_conv(x)
+        x = F.leaky_relu(x, modules.LRELU_SLOPE)
+        x = torch.flip(x, dims=[2])
+        fmap.append(x)
+
+        for i, l in enumerate(self.convs):
+            x = l(x, g)
+            fmap.append(x)
+
+        x = F.pad(x, [0, 0, 2, 0], mode="constant")
+        x = self.conv_post(x)
+        x = torch.flatten(x, 1, -1)
+
+        return x, fmap
+
+
+class MultiPeriodDiscriminator(torch.nn.Module):
+    def __init__(self, upsample_rates, gin_channels, periods=[2, 3, 5, 7, 11, 17], **kwargs):
+        super(MultiPeriodDiscriminator, self).__init__()
+
+        discs = [
+            DiscriminatorP(i, gin_channels, upsample_rates, use_spectral_norm=False) for i in periods
+        ]
+        self.ups = np.prod(upsample_rates)
+        self.discriminators = nn.ModuleList(discs)
+
+    def forward(self, y, y_hat, g):
+        fmap_rs = []
+        fmap_gs = []
+        y_d_rs = []
+        y_d_gs = []
+        for d in self.discriminators:
+            y_d_r, fmap_r = d(y, g)
+            y_d_g, fmap_g = d(y_hat, g)
+            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

server/voice_changer/RVC/inferencer/voras_beta/modules.py

@@ -0,0 +1,496 @@
+import math
+
+import numpy as np
+import scipy
+import torch
+from torch import nn
+from torch.nn import Conv1d, Conv2d
+from torch.nn import functional as F
+from torch.nn.utils import remove_weight_norm, spectral_norm, weight_norm
+from torchaudio.functional.functional import _hz_to_mel, _mel_to_hz
+
+from . import commons, modules
+from .commons import get_padding, init_weights
+from .transforms import piecewise_rational_quadratic_transform
+
+LRELU_SLOPE = 0.1
+
+class HarmonicEmbedder(nn.Module):
+    def __init__(self, num_embeddings, embedding_dim, gin_channels, num_head, num_harmonic=0, f0_min=50., f0_max=1100., device="cuda"):
+        super(HarmonicEmbedder, self).__init__()
+        self.embedding_dim = embedding_dim
+        self.num_head = num_head
+        self.num_harmonic = num_harmonic
+
+        f0_mel_min = np.log(1 + f0_min / 700)
+        f0_mel_max = np.log(1 + f0_max * (1 + num_harmonic) / 700)
+        self.sequence = torch.from_numpy(np.linspace(f0_mel_min, f0_mel_max, num_embeddings-2))
+        self.emb_layer = torch.nn.Embedding(num_embeddings, embedding_dim)
+        self.linear_q = Conv1d(gin_channels, num_head * (1 + num_harmonic), 1)
+        self.weight = None
+
+    def forward(self, x, g):
+        b, l = x.size()
+        non_zero = (x != 0.).to(dtype=torch.long).unsqueeze(1)
+        mel = torch.log(1 + x / 700).unsqueeze(1)
+        harmonies = torch.arange(1 + self.num_harmonic, device=x.device, dtype=x.dtype).view(1, 1 + self.num_harmonic, 1) + 1.
+        ix = torch.searchsorted(self.sequence.to(x.device), mel * harmonies).to(x.device) + 1
+        ix = ix * non_zero
+        emb = self.emb_layer(ix).transpose(1, 3).reshape(b, self.num_head, self.embedding_dim // self.num_head, 1 + self.num_harmonic, l)
+        if self.weight is None:
+            weight = torch.nn.functional.softmax(self.linear_q(g).reshape(b, self.num_head, 1, 1 + self.num_harmonic, 1), 3)
+        else:
+            weight = self.weight
+        res = torch.sum(emb * weight, dim=3).reshape(b, self.embedding_dim, l)
+        return res
+
+    def fix_speaker(self, g):
+        self.weight = torch.nn.functional.softmax(self.linear_q(g).reshape(1, self.num_head, 1, 1 + self.num_harmonic, 1), 3)
+
+    def unfix_speaker(self, g):
+        self.weight = None
+
+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 DilatedCausalConv1d(nn.Module):
+    def __init__(self, in_channels, out_channels, kernel_size=3, stride=1, groups=1, dilation=1, bias=True):
+        super(DilatedCausalConv1d, self).__init__()
+        self.kernel_size = kernel_size
+        self.dilation = dilation
+        self.stride = stride
+        self.conv = weight_norm(nn.Conv1d(in_channels, out_channels, kernel_size, stride=stride, groups=groups, dilation=dilation, bias=bias))
+
+    def forward(self, x):
+        x = torch.flip(x, [2])
+        x = F.pad(x, [0, (self.kernel_size - 1) * self.dilation], mode="constant", value=0.)
+        size = x.shape[2] // self.stride
+        x = self.conv(x)[:, :, :size]
+        x = torch.flip(x, [2])
+        return x
+
+    def remove_weight_norm(self):
+        remove_weight_norm(self.conv)
+
+
+class CausalConvTranspose1d(nn.Module):
+    """
+    padding = 0, dilation = 1のとき
+
+    Lout = (Lin - 1) * stride + kernel_rate * stride + output_padding
+    Lout = Lin * stride + (kernel_rate - 1) * stride + output_padding
+    output_paddingいらないね
+    """
+    def __init__(self, in_channels, out_channels, kernel_rate=3, stride=1, groups=1):
+        super(CausalConvTranspose1d, self).__init__()
+        kernel_size = kernel_rate * stride
+        self.trim_size = (kernel_rate - 1) * stride
+        self.conv = weight_norm(nn.ConvTranspose1d(in_channels, out_channels, kernel_size, stride=stride, groups=groups))
+
+    def forward(self, x):
+        x = self.conv(x)
+        return x[:, :, :-self.trim_size]
+
+    def remove_weight_norm(self):
+        remove_weight_norm(self.conv)
+
+
+class LoRALinear1d(nn.Module):
+    def __init__(self, in_channels, out_channels, info_channels, r):
+        super().__init__()
+        self.in_channels = in_channels
+        self.out_channels = out_channels
+        self.info_channels = info_channels
+        self.r = r
+        self.main_fc = weight_norm(nn.Conv1d(in_channels, out_channels, 1))
+        self.adapter_in = nn.Conv1d(info_channels, in_channels * r, 1)
+        self.adapter_out = nn.Conv1d(info_channels, out_channels * r, 1)
+        nn.init.normal_(self.adapter_in.weight.data, 0, 0.01)
+        nn.init.constant_(self.adapter_out.weight.data, 1e-6)
+        self.adapter_in = weight_norm(self.adapter_in)
+        self.adapter_out = weight_norm(self.adapter_out)
+        self.speaker_fixed = False
+
+    def forward(self, x, g):
+        x_ = self.main_fc(x)
+        if not self.speaker_fixed:
+            a_in = self.adapter_in(g).view(-1, self.in_channels, self.r)
+            a_out = self.adapter_out(g).view(-1, self.r, self.out_channels)
+            l = torch.einsum("brl,brc->bcl", torch.einsum("bcl,bcr->brl", x, a_in), a_out)
+            x_ = x_ + l
+        return x_
+
+    def remove_weight_norm(self):
+        remove_weight_norm(self.main_fc)
+        remove_weight_norm(self.adapter_in)
+        remove_weight_norm(self.adapter_out)
+
+    def fix_speaker(self, g):
+        self.speaker_fixed = True
+        a_in = self.adapter_in(g).view(-1, self.in_channels, self.r)
+        a_out = self.adapter_out(g).view(-1, self.r, self.out_channels)
+        weight = torch.einsum("bir,bro->oi", a_in, a_out).unsqueeze(2)
+        self.main_fc.weight.data.add_(weight)
+
+    def unfix_speaker(self, g):
+        self.speaker_fixed = False
+        a_in = self.adapter_in(g).view(-1, self.in_channels, self.r)
+        a_out = self.adapter_out(g).view(-1, self.r, self.out_channels)
+        weight = torch.einsum("bir,bro->oi", a_in, a_out).unsqueeze(2)
+        self.main_fc.weight.data.sub_(weight)
+
+
+class LoRALinear2d(nn.Module):
+    def __init__(self, in_channels, out_channels, info_channels, r):
+        super().__init__()
+        self.in_channels = in_channels
+        self.out_channels = out_channels
+        self.info_channels = info_channels
+        self.r = r
+        self.main_fc = weight_norm(nn.Conv2d(in_channels, out_channels, (1, 1), (1, 1)))
+        self.adapter_in = nn.Conv1d(info_channels, in_channels * r, 1)
+        self.adapter_out = nn.Conv1d(info_channels, out_channels * r, 1)
+        nn.init.normal_(self.adapter_in.weight.data, 0, 0.01)
+        nn.init.constant_(self.adapter_out.weight.data, 1e-6)
+        self.adapter_in = weight_norm(self.adapter_in)
+        self.adapter_out = weight_norm(self.adapter_out)
+        self.speaker_fixed = False
+
+    def forward(self, x, g):
+        x_ = self.main_fc(x)
+        if not self.speaker_fixed:
+            a_in = self.adapter_in(g).view(-1, self.in_channels, self.r)
+            a_out = self.adapter_out(g).view(-1, self.r, self.out_channels)
+            l = torch.einsum("brhw,brc->bchw", torch.einsum("bchw,bcr->brhw", x, a_in), a_out)
+            x_ = x_ + l
+        return x_
+
+    def remove_weight_norm(self):
+        remove_weight_norm(self.main_fc)
+        remove_weight_norm(self.adapter_in)
+        remove_weight_norm(self.adapter_out)
+
+    def fix_speaker(self, g):
+        a_in = self.adapter_in(g).view(-1, self.in_channels, self.r)
+        a_out = self.adapter_out(g).view(-1, self.r, self.out_channels)
+        weight = torch.einsum("bir,bro->oi", a_in, a_out).unsqueeze(2).unsqueeze(3)
+        self.main_fc.weight.data.add_(weight)
+
+    def unfix_speaker(self, g):
+        a_in = self.adapter_in(g).view(-1, self.in_channels, self.r)
+        a_out = self.adapter_out(g).view(-1, self.r, self.out_channels)
+        weight = torch.einsum("bir,bro->oi", a_in, a_out).unsqueeze(2).unsqueeze(3)
+        self.main_fc.weight.data.sub_(weight)
+
+
+class MBConv2d(torch.nn.Module):
+    """
+    Causal MBConv2D
+    """
+    def __init__(self, in_channels, out_channels, gin_channels, kernel_size, stride, extend_ratio, r, use_spectral_norm=False):
+        super(MBConv2d, self).__init__()
+        norm_f = weight_norm if use_spectral_norm == False else spectral_norm
+        inner_channels = int(in_channels * extend_ratio)
+        self.kernel_size = kernel_size
+        self.pwconv1 = LoRALinear2d(in_channels, inner_channels, gin_channels, r=r)
+        self.dwconv = norm_f(Conv2d(inner_channels, inner_channels, kernel_size, stride, groups=inner_channels))
+        self.pwconv2 = LoRALinear2d(inner_channels, out_channels, gin_channels, r=r)
+        self.pwnorm = LayerNorm(in_channels)
+        self.dwnorm = LayerNorm(inner_channels)
+
+    def forward(self, x, g):
+        x = self.pwnorm(x)
+        x = self.pwconv1(x, g)
+        x = F.pad(x, [0, 0, self.kernel_size[0] - 1, 0], mode="constant")
+        x = self.dwnorm(x)
+        x = self.dwconv(x)
+        x = F.leaky_relu(x, modules.LRELU_SLOPE)
+        x = self.pwconv2(x, g)
+        x = F.leaky_relu(x, modules.LRELU_SLOPE)
+        return x
+
+class ConvNext2d(torch.nn.Module):
+    """
+    Causal ConvNext Block
+    stride = 1 only
+    """
+    def __init__(self, in_channels, out_channels, gin_channels, kernel_size, stride, extend_ratio, r, use_spectral_norm=False):
+        super(ConvNext2d, self).__init__()
+        norm_f = weight_norm if use_spectral_norm == False else spectral_norm
+        inner_channels = int(in_channels * extend_ratio)
+        self.kernel_size = kernel_size
+        self.dwconv = norm_f(Conv2d(in_channels, in_channels, kernel_size, stride, groups=in_channels))
+        self.pwconv1 = LoRALinear2d(in_channels, inner_channels, gin_channels, r=r)
+        self.pwconv2 = LoRALinear2d(inner_channels, out_channels, gin_channels, r=r)
+        self.act = nn.GELU()
+        self.norm = LayerNorm(in_channels)
+
+    def forward(self, x, g):
+        x = F.pad(x, [0, 0, self.kernel_size[0] - 1, 0], mode="constant")
+        x = self.dwconv(x)
+        x = self.norm(x)
+        x = self.pwconv1(x, g)
+        x = self.act(x)
+        x = self.pwconv2(x, g)
+        x = self.act(x)
+        return x
+
+    def remove_weight_norm(self):
+        remove_weight_norm(self.dwconv)
+
+
+class WaveBlock(torch.nn.Module):
+    def __init__(self, inner_channels, gin_channels, kernel_sizes, strides, dilations, extend_rate, r):
+        super(WaveBlock, self).__init__()
+        norm_f = weight_norm
+        extend_channels = int(inner_channels * extend_rate)
+        self.dconvs = nn.ModuleList()
+        self.p1convs = nn.ModuleList()
+        self.p2convs = nn.ModuleList()
+        self.norms = nn.ModuleList()
+        self.act = nn.GELU()
+
+        # self.ses = nn.ModuleList()
+        # self.norms = []
+        for i, (k, s, d) in enumerate(zip(kernel_sizes, strides, dilations)):
+            self.dconvs.append(DilatedCausalConv1d(inner_channels, inner_channels, k, stride=s, dilation=d, groups=inner_channels))
+            self.p1convs.append(LoRALinear1d(inner_channels, extend_channels, gin_channels, r))
+            self.p2convs.append(LoRALinear1d(extend_channels, inner_channels, gin_channels, r))
+            self.norms.append(LayerNorm(inner_channels))
+
+    def forward(self, x, x_mask, g):
+        x *= x_mask
+        for i in range(len(self.dconvs)):
+            residual = x.clone()
+            x = self.dconvs[i](x)
+            x = self.norms[i](x)
+            x *= x_mask
+            x = self.p1convs[i](x, g)
+            x = self.act(x)
+            x = self.p2convs[i](x, g)
+            x = residual + x
+        return x
+
+    def remove_weight_norm(self):
+        for c in self.dconvs:
+            c.remove_weight_norm()
+        for c in self.p1convs:
+            c.remove_weight_norm()
+        for c in self.p2convs:
+            c.remove_weight_norm()
+
+    def fix_speaker(self, g):
+        for c in self.p1convs:
+            c.fix_speaker(g)
+        for c in self.p2convs:
+            c.fix_speaker(g)
+
+    def unfix_speaker(self, g):
+        for c in self.p1convs:
+            c.unfix_speaker(g)
+        for c in self.p2convs:
+            c.unfix_speaker(g)
+
+
+class SnakeFilter(torch.nn.Module):
+    """
+    Adaptive filter using snakebeta
+    """
+    def __init__(self, channels, groups, kernel_size, num_layers, eps=1e-6):
+        super(SnakeFilter, self).__init__()
+        self.eps = eps
+        self.num_layers = num_layers
+        inner_channels = channels * groups
+        self.init_conv = DilatedCausalConv1d(1, inner_channels, kernel_size)
+        self.dconvs = torch.nn.ModuleList()
+        self.pconvs = torch.nn.ModuleList()
+        self.post_conv = DilatedCausalConv1d(inner_channels+1, 1, kernel_size, bias=False)
+
+        for i in range(self.num_layers):
+            self.dconvs.append(DilatedCausalConv1d(inner_channels, inner_channels, kernel_size, stride=1, groups=inner_channels, dilation=kernel_size ** (i + 1)))
+            self.pconvs.append(weight_norm(Conv1d(inner_channels, inner_channels, 1, groups=groups)))
+        self.snake_alpha = torch.nn.Parameter(torch.zeros(inner_channels), requires_grad=True)
+        self.snake_beta = torch.nn.Parameter(torch.zeros(inner_channels), requires_grad=True)
+
+    def forward(self, x):
+        y = x.clone()
+        x = self.init_conv(x)
+        for i in range(self.num_layers):
+            # snake activation
+            x = self.dconvs[i](x)
+            x = self.pconvs[i](x)
+        x = x + (1.0 / torch.clip(self.snake_beta.unsqueeze(0).unsqueeze(-1), min=self.eps)) * torch.pow(torch.sin(x * self.snake_alpha.unsqueeze(0).unsqueeze(-1)), 2)
+        x = torch.cat([x, y], 1)
+        x = self.post_conv(x)
+        return x
+
+    def remove_weight_norm(self):
+        self.init_conv.remove_weight_norm()
+        for c in self.dconvs:
+            c.remove_weight_norm()
+        for c in self.pconvs:
+            remove_weight_norm(c)
+        self.post_conv.remove_weight_norm()
+
+"""
+https://github.com/charactr-platform/vocos/blob/main/vocos/heads.py
+"""
+class FourierHead(nn.Module):
+    """Base class for inverse fourier modules."""
+
+    def forward(self, x: torch.Tensor) -> torch.Tensor:
+        """
+        Args:
+            x (Tensor): Input tensor of shape (B, L, H), where B is the batch size,
+                        L is the sequence length, and H denotes the model dimension.
+
+        Returns:
+            Tensor: Reconstructed time-domain audio signal of shape (B, T), where T is the length of the output signal.
+        """
+        raise NotImplementedError("Subclasses must implement the forward method.")
+
+
+class IMDCT(nn.Module):
+    """
+    Inverse Modified Discrete Cosine Transform (IMDCT) module.
+
+    Args:
+        frame_len (int): Length of the MDCT frame.
+        padding (str, optional): Type of padding. Options are "center" or "same". Defaults to "same".
+    """
+
+    def __init__(self, frame_len: int, padding: str = "same"):
+        super().__init__()
+        if padding not in ["center", "same"]:
+            raise ValueError("Padding must be 'center' or 'same'.")
+        self.padding = padding
+        self.frame_len = frame_len * 2
+        N = frame_len
+        n0 = (N + 1) / 2
+        window = torch.from_numpy(scipy.signal.cosine(N * 2)).float()
+        self.register_buffer("window", window)
+
+        pre_twiddle = torch.exp(1j * torch.pi * n0 * torch.arange(N * 2) / N)
+        post_twiddle = torch.exp(1j * torch.pi * (torch.arange(N * 2) + n0) / (N * 2))
+        self.register_buffer("pre_twiddle", torch.view_as_real(pre_twiddle))
+        self.register_buffer("post_twiddle", torch.view_as_real(post_twiddle))
+
+    def forward(self, X: torch.Tensor) -> torch.Tensor:
+        """
+        Apply the Inverse Modified Discrete Cosine Transform (IMDCT) to the input MDCT coefficients.
+
+        Args:
+            X (Tensor): Input MDCT coefficients of shape (B, N, L), where B is the batch size,
+                L is the number of frames, and N is the number of frequency bins.
+
+        Returns:
+            Tensor: Reconstructed audio waveform of shape (B, T), where T is the length of the audio.
+        """
+        X = X.transpose(1, 2)
+        B, L, N = X.shape
+        Y = torch.zeros((B, L, N * 2), dtype=X.dtype, device=X.device)
+        Y[..., :N] = X
+        Y[..., N:] = -1 * torch.conj(torch.flip(X, dims=(-1,)))
+        y = torch.fft.ifft(Y * torch.view_as_complex(self.pre_twiddle).expand(Y.shape), dim=-1)
+        y = torch.real(y * torch.view_as_complex(self.post_twiddle).expand(y.shape)) * np.sqrt(N) * np.sqrt(2)
+        result = y * self.window.expand(y.shape)
+        output_size = (1, (L + 1) * N)
+        audio = torch.nn.functional.fold(
+            result.transpose(1, 2),
+            output_size=output_size,
+            kernel_size=(1, self.frame_len),
+            stride=(1, self.frame_len // 2),
+        )[:, 0, 0, :]
+
+        if self.padding == "center":
+            pad = self.frame_len // 2
+        elif self.padding == "same":
+            pad = self.frame_len // 4
+        else:
+            raise ValueError("Padding must be 'center' or 'same'.")
+
+        audio = audio[:, pad:-pad]
+        return audio.unsqueeze(1)
+
+
+class IMDCTSymExpHead(FourierHead):
+    """
+    IMDCT Head module for predicting MDCT coefficients with symmetric exponential function
+
+    Args:
+        dim (int): Hidden dimension of the model.
+        mdct_frame_len (int): Length of the MDCT frame.
+        padding (str, optional): Type of padding. Options are "center" or "same". Defaults to "same".
+        sample_rate (int, optional): The sample rate of the audio. If provided, the last layer will be initialized
+                                     based on perceptual scaling. Defaults to None.
+        clip_audio (bool, optional): Whether to clip the audio output within the range of [-1.0, 1.0]. Defaults to False.
+    """
+
+    def __init__(
+        self, dim: int, gin_channels: int, mdct_frame_len: int, padding: str = "same", sample_rate: int = 24000,
+    ):
+        super().__init__()
+        out_dim = mdct_frame_len
+        self.dconv = DilatedCausalConv1d(dim, dim, 5, 1, dim, 1)
+        self.pconv1 = LoRALinear1d(dim, dim * 2, gin_channels, 2)
+        self.pconv2 = LoRALinear1d(dim * 2, out_dim, gin_channels, 2)
+        self.act = torch.nn.GELU()
+        self.imdct = IMDCT(frame_len=mdct_frame_len, padding=padding)
+
+        if sample_rate is not None:
+            # optionally init the last layer following mel-scale
+            m_max = _hz_to_mel(sample_rate // 2)
+            m_pts = torch.linspace(0, m_max, out_dim)
+            f_pts = _mel_to_hz(m_pts)
+            scale = 1 - (f_pts / f_pts.max())
+
+            with torch.no_grad():
+                self.pconv2.main_fc.weight.mul_(scale.view(-1, 1, 1))
+
+    def forward(self, x: torch.Tensor, g: torch.Tensor) -> torch.Tensor:
+        """
+        Forward pass of the IMDCTSymExpHead module.
+
+        Args:
+            x (Tensor): Input tensor of shape (B, L, H), where B is the batch size,
+                        L is the sequence length, and H denotes the model dimension.
+
+        Returns:
+            Tensor: Reconstructed time-domain audio signal of shape (B, T), where T is the length of the output signal.
+        """
+        x = self.dconv(x)
+        x = self.pconv1(x, g)
+        x = self.act(x)
+        x = self.pconv2(x, g)
+        x = symexp(x)
+        x = torch.clip(x, min=-1e2, max=1e2)  # safeguard to prevent excessively large magnitudes
+        audio = self.imdct(x)
+        return audio
+
+    def remove_weight_norm(self):
+        self.dconv.remove_weight_norm()
+        self.pconv1.remove_weight_norm()
+        self.pconv2.remove_weight_norm()
+
+    def fix_speaker(self, g):
+        self.pconv1.fix_speaker(g)
+        self.pconv2.fix_speaker(g)
+
+    def unfix_speaker(self, g):
+        self.pconv1.unfix_speaker(g)
+        self.pconv2.unfix_speaker(g)
+
+def symexp(x: torch.Tensor) -> torch.Tensor:
+    return torch.sign(x) * (torch.exp(x.abs()) - 1)

server/voice_changer/RVC/inferencer/voras_beta/transforms.py

@@ -0,0 +1,207 @@
+import numpy as np
+import torch
+from torch.nn import functional as F
+
+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=1.0,
+    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=1.0,
+    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))
+        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=0.0,
+    right=1.0,
+    bottom=0.0,
+    top=1.0,
+    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

server/voice_changer/RVC/inferencer/voras_beta/utils.py

@@ -0,0 +1,286 @@
+import glob
+import logging
+import os
+import shutil
+import socket
+import sys
+
+import ffmpeg
+import matplotlib
+import matplotlib.pylab as plt
+import numpy as np
+import torch
+from scipy.io.wavfile import read
+from torch.nn import functional as F
+
+from modules.shared import ROOT_DIR
+
+from .config import TrainConfig
+
+matplotlib.use("Agg")
+logging.getLogger("matplotlib").setLevel(logging.WARNING)
+
+logging.basicConfig(stream=sys.stdout, level=logging.DEBUG)
+logger = logging
+
+
+class AWP:
+    """
+    Fast AWP
+    https://www.kaggle.com/code/junkoda/fast-awp
+    """
+    def __init__(self, model, optimizer, *, adv_param='weight',
+                 adv_lr=0.01, adv_eps=0.01):
+        self.model = model
+        self.optimizer = optimizer
+        self.adv_param = adv_param
+        self.adv_lr = adv_lr
+        self.adv_eps = adv_eps
+        self.backup = {}
+
+    def perturb(self):
+        """
+        Perturb model parameters for AWP gradient
+        Call before loss and loss.backward()
+        """
+        self._save()  # save model parameters
+        self._attack_step()  # perturb weights
+
+    def _attack_step(self):
+        e = 1e-6
+        for name, param in self.model.named_parameters():
+            if param.requires_grad and param.grad is not None and self.adv_param in name:
+                grad = self.optimizer.state[param]['exp_avg']
+                norm_grad = torch.norm(grad)
+                norm_data = torch.norm(param.detach())
+
+                if norm_grad != 0 and not torch.isnan(norm_grad):
+                    # Set lower and upper limit in change
+                    limit_eps = self.adv_eps * param.detach().abs()
+                    param_min = param.data - limit_eps
+                    param_max = param.data + limit_eps
+
+                    # Perturb along gradient
+                    # w += (adv_lr * |w| / |grad|) * grad
+                    param.data.add_(grad, alpha=(self.adv_lr * (norm_data + e) / (norm_grad + e)))
+
+                    # Apply the limit to the change
+                    param.data.clamp_(param_min, param_max)
+
+    def _save(self):
+        for name, param in self.model.named_parameters():
+            if param.requires_grad and param.grad is not None and self.adv_param in name:
+                if name not in self.backup:
+                    self.backup[name] = param.clone().detach()
+                else:
+                    self.backup[name].copy_(param.data)
+
+    def restore(self):
+        """
+        Restore model parameter to correct position; AWP do not perturbe weights, it perturb gradients
+        Call after loss.backward(), before optimizer.step()
+        """
+        for name, param in self.model.named_parameters():
+            if name in self.backup:
+                param.data.copy_(self.backup[name])
+
+
+def load_audio(file: str, sr):
+    try:
+        # https://github.com/openai/whisper/blob/main/whisper/audio.py#L26
+        # This launches a subprocess to decode audio while down-mixing and resampling as necessary.
+        # Requires the ffmpeg CLI and `ffmpeg-python` package to be installed.
+        file = (
+            file.strip(" ").strip('"').strip("\n").strip('"').strip(" ")
+        )  # Prevent small white copy path head and tail with spaces and " and return
+        out, _ = (
+            ffmpeg.input(file, threads=0)
+            .output("-", format="f32le", acodec="pcm_f32le", ac=1, ar=sr)
+            .run(cmd=["ffmpeg", "-nostdin"], capture_stdout=True, capture_stderr=True)
+        )
+    except Exception as e:
+        raise RuntimeError(f"Failed to load audio: {e}")
+
+    return np.frombuffer(out, np.float32).flatten()
+
+
+def find_empty_port():
+    s = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
+    s.bind(("", 0))
+    s.listen(1)
+    port = s.getsockname()[1]
+    s.close()
+    return port
+
+
+def load_checkpoint(checkpoint_path, model, optimizer=None, load_opt=1):
+    assert os.path.isfile(checkpoint_path)
+    checkpoint_dict = torch.load(checkpoint_path, map_location="cpu")
+
+    saved_state_dict = checkpoint_dict["model"]
+    if hasattr(model, "module"):
+        state_dict = model.module.state_dict()
+    else:
+        state_dict = model.state_dict()
+    new_state_dict = {}
+    for k, v in state_dict.items():  # 模型需要的shape
+        try:
+            new_state_dict[k] = saved_state_dict[k]
+            if saved_state_dict[k].shape != state_dict[k].shape:
+                print(
+                    f"shape-{k}-mismatch|need-{state_dict[k].shape}|get-{saved_state_dict[k].shape}"
+                )
+                if saved_state_dict[k].dim() == 2:  # NOTE: check is this ok?
+                    # for embedded input 256 <==> 768
+                    # this achieves we can continue training from original's pretrained checkpoints when using embedder that 768-th dim output etc.
+                    if saved_state_dict[k].dtype == torch.half:
+                        new_state_dict[k] = (
+                            F.interpolate(
+                                saved_state_dict[k].float().unsqueeze(0).unsqueeze(0),
+                                size=state_dict[k].shape,
+                                mode="bilinear",
+                            )
+                            .half()
+                            .squeeze(0)
+                            .squeeze(0)
+                        )
+                    else:
+                        new_state_dict[k] = (
+                            F.interpolate(
+                                saved_state_dict[k].unsqueeze(0).unsqueeze(0),
+                                size=state_dict[k].shape,
+                                mode="bilinear",
+                            )
+                            .squeeze(0)
+                            .squeeze(0)
+                        )
+                    print(
+                        "interpolated new_state_dict",
+                        k,
+                        "from",
+                        saved_state_dict[k].shape,
+                        "to",
+                        new_state_dict[k].shape,
+                    )
+                else:
+                    raise KeyError
+        except Exception as e:
+            # print(traceback.format_exc())
+            print(f"{k} is not in the checkpoint")
+            print("error: %s" % e)
+            new_state_dict[k] = v  # 模型自带的随机值
+    if hasattr(model, "module"):
+        model.module.load_state_dict(new_state_dict, strict=False)
+    else:
+        model.load_state_dict(new_state_dict, strict=False)
+    print("Loaded model weights")
+
+    epoch = checkpoint_dict["epoch"]
+    learning_rate = checkpoint_dict["learning_rate"]
+    if optimizer is not None and load_opt == 1:
+        optimizer.load_state_dict(checkpoint_dict["optimizer"])
+    print("Loaded checkpoint '{}' (epoch {})".format(checkpoint_path, epoch))
+    return model, optimizer, learning_rate, epoch
+
+
+def save_state(model, optimizer, learning_rate, epoch, checkpoint_path):
+    print(
+        "Saving model and optimizer state at epoch {} to {}".format(
+            epoch, checkpoint_path
+        )
+    )
+    if hasattr(model, "module"):
+        state_dict = model.module.state_dict()
+    else:
+        state_dict = model.state_dict()
+    torch.save(
+        {
+            "model": state_dict,
+            "epoch": epoch,
+            "optimizer": optimizer.state_dict(),
+            "learning_rate": learning_rate,
+        },
+        checkpoint_path,
+    )
+
+
+def summarize(
+    writer,
+    global_step,
+    scalars={},
+    histograms={},
+    images={},
+    audios={},
+    audio_sampling_rate=22050,
+):
+    for k, v in scalars.items():
+        writer.add_scalar(k, v, global_step)
+    for k, v in histograms.items():
+        writer.add_histogram(k, v, global_step)
+    for k, v in images.items():
+        writer.add_image(k, v, global_step, dataformats="HWC")
+    for k, v in audios.items():
+        writer.add_audio(k, v, global_step, audio_sampling_rate)
+
+
+def latest_checkpoint_path(dir_path, regex="G_*.pth"):
+    filelist = glob.glob(os.path.join(dir_path, regex))
+    if len(filelist) == 0:
+        return None
+    filelist.sort(key=lambda f: int("".join(filter(str.isdigit, f))))
+    filepath = filelist[-1]
+    return filepath
+
+
+def plot_spectrogram_to_numpy(spectrogram):
+    fig, ax = plt.subplots(figsize=(10, 2))
+    im = ax.imshow(spectrogram, aspect="auto", origin="lower", interpolation="none")
+    plt.colorbar(im, ax=ax)
+    plt.xlabel("Frames")
+    plt.ylabel("Channels")
+    plt.tight_layout()
+
+    fig.canvas.draw()
+    data = np.fromstring(fig.canvas.tostring_rgb(), dtype=np.uint8, sep="")
+    data = data.reshape(fig.canvas.get_width_height()[::-1] + (3,))
+    plt.close()
+    return data
+
+
+def plot_alignment_to_numpy(alignment, info=None):
+    fig, ax = plt.subplots(figsize=(6, 4))
+    im = ax.imshow(
+        alignment.transpose(), aspect="auto", origin="lower", interpolation="none"
+    )
+    fig.colorbar(im, ax=ax)
+    xlabel = "Decoder timestep"
+    if info is not None:
+        xlabel += "\n\n" + info
+    plt.xlabel(xlabel)
+    plt.ylabel("Encoder timestep")
+    plt.tight_layout()
+
+    fig.canvas.draw()
+    data = np.fromstring(fig.canvas.tostring_rgb(), dtype=np.uint8, sep="")
+    data = data.reshape(fig.canvas.get_width_height()[::-1] + (3,))
+    plt.close()
+    return data
+
+
+def load_wav_to_torch(full_path):
+    sampling_rate, data = read(full_path)
+    return torch.FloatTensor(data.astype(np.float32)), sampling_rate
+
+
+def load_config(training_dir: str, sample_rate: int, emb_channels: int):
+    if emb_channels == 256:
+        config_path = os.path.join(ROOT_DIR, "configs", f"{sample_rate}.json")
+    else:
+        config_path = os.path.join(
+            ROOT_DIR, "configs", f"{sample_rate}-{emb_channels}.json"
+        )
+    config_save_path = os.path.join(training_dir, "config.json")
+
+    shutil.copyfile(config_path, config_save_path)
+
+    return TrainConfig.parse_file(config_save_path)

server/voice_changer/RVC/pipeline/Pipeline.py

@@ -3,6 +3,7 @@ from typing import Any
 import math
 import torch
 import torch.nn.functional as F
+from torch.cuda.amp import autocast
 from Exceptions import (
     DeviceCannotSupportHalfPrecisionException,
     DeviceChangingException,
@@ -118,10 +119,6 @@ class Pipeline(object):
 
         # tensor型調整
         feats = audio_pad
-        if self.isHalf is True:
-            feats = feats.half()
-        else:
-            feats = feats.float()
         if feats.dim() == 2:  # double channels
             feats = feats.mean(-1)
         assert feats.dim() == 1, feats.dim()
@@ -129,6 +126,7 @@ class Pipeline(object):
 
         # embedding
         padding_mask = torch.BoolTensor(feats.shape).to(self.device).fill_(False)
+        with autocast(enabled=self.isHalf):
             try:
                 feats = self.embedder.extractFeatures(feats, embOutputLayer, useFinalProj)
                 if torch.isnan(feats).all():
@@ -167,10 +165,8 @@ class Pipeline(object):
 
             # recover silient font
             npy = np.concatenate([np.zeros([npyOffset, npy.shape[1]]).astype("float32"), npy])
-            if self.isHalf is True:
-                npy = npy.astype("float16")
-
             feats = torch.from_numpy(npy).unsqueeze(0).to(self.device) * index_rate + (1 - index_rate) * feats
+
         feats = F.interpolate(feats.permute(0, 2, 1), scale_factor=2).permute(0, 2, 1)
         if protect < 0.5 and search_index:
             feats0 = F.interpolate(feats0.permute(0, 2, 1), scale_factor=2).permute(0, 2, 1)
@@ -207,6 +203,7 @@ class Pipeline(object):
         # 推論実行
         try:
             with torch.no_grad():
+                with autocast(enabled=self.isHalf):
                     audio1 = (
                         torch.clip(
                             self.inferencer.infer(feats, p_len, pitch, pitchf, sid)[0][0, 0].to(dtype=torch.float32),