Source code for diffsptk.modules.msvq
# ------------------------------------------------------------------------ #
# Copyright 2022 SPTK Working Group #
# #
# Licensed under the Apache License, Version 2.0 (the "License"); #
# you may not use this file except in compliance with the License. #
# You may obtain a copy of the License at #
# #
# http://www.apache.org/licenses/LICENSE-2.0 #
# #
# Unless required by applicable law or agreed to in writing, software #
# distributed under the License is distributed on an "AS IS" BASIS, #
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. #
# See the License for the specific language governing permissions and #
# limitations under the License. #
# ------------------------------------------------------------------------ #
import torch
from .base import BaseNonFunctionalModule
[docs]
class MultiStageVectorQuantization(BaseNonFunctionalModule):
"""See `this page <https://github.com/lucidrains/vector-quantize-pytorch>`_
for details.
Parameters
----------
order : int >= 0
The order of the input vector, :math:`M`.
codebook_size : int >= 1
The codebook size, :math:`K`.
n_stage : int >= 1
The number of stages (quantizers), :math:`Q`.
device : torch.device or None
The device of this module.
**kwargs : additional keyword arguments
See `this page`_ for details.
References
----------
.. [1] A. v. d. Oord et al., "Neural discrete representation learning," *Advances in
Neural Information Processing Systems*, pp. 6309-6318, 2017.
"""
def __init__(
self,
order: int,
codebook_size: int,
n_stage: int,
device: torch.device | None = None,
**kwargs,
) -> None:
super().__init__()
if order < 0:
raise ValueError("order must be non-negative.")
if codebook_size <= 0:
raise ValueError("codebook_size must be positive.")
if n_stage <= 0:
raise ValueError("n_stage must be positive.")
from vector_quantize_pytorch import ResidualVQ
self.vq = ResidualVQ(
dim=order + 1, codebook_size=codebook_size, num_quantizers=n_stage, **kwargs
).to(device=device, dtype=torch.float)
@property
def codebooks(self) -> torch.Tensor:
return self.vq.codebooks
[docs]
def forward(
self, x: torch.Tensor, codebooks: torch.Tensor | None = None, **kwargs
) -> tuple[torch.Tensor, torch.Tensor, torch.Tensor]:
"""Perform residual vector quantization.
Parameters
----------
x : Tensor [shape=(..., M+1)]
The input vectors.
codebooks : Tensor [shape=(Q, K, M+1)] or None
The external codebook. If None, use the internal codebook.
**kwargs : additional keyword arguments
See `this page`_ for details.
Returns
-------
xq : Tensor [shape=(..., M+1)]
The quantized vectors.
indices : Tensor [shape=(..., Q)]
The codebook indices.
losses : Tensor [shape=(Q,)]
The commitment losses.
Examples
--------
>>> import diffsptk
>>> msvq = diffsptk.MultiStageVectorQuantization(4, 3, 2).eval()
>>> x = diffsptk.nrand(4)
>>> x.shape
torch.Size([5])
>>> xq, indices, losses = msvq(x)
>>> xq.shape
torch.Size([5])
>>> indices.shape
torch.Size([2])
>>> losses.shape
torch.Size([2])
"""
if codebooks is not None:
cb = self.codebooks
for i, layer in enumerate(self.vq.layers):
layer._codebook.embed[:] = codebooks.view_as(cb)[i]
d = x.dim()
if d == 1:
x = x.unsqueeze(0)
xq, indices, losses = self.vq(x.float(), **kwargs)
xq = xq.to(dtype=x.dtype)
if d == 1:
xq = xq.squeeze(0)
indices = indices.squeeze(0)
losses = losses.squeeze()
return xq, indices, losses
