DL-Art-School/codes/models/vqvae/vqvae.py

324 lines
11 KiB
Python
Raw Normal View History

2021-01-07 17:20:15 +00:00
# Copyright 2018 The Sonnet Authors. All Rights Reserved.
#
# 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.
# ============================================================================
# Borrowed from https://github.com/rosinality/vq-vae-2-pytorch
2021-09-24 03:19:36 +00:00
# Which was itself borrowed from https://github.com/deepmind/sonnet
2021-01-07 17:20:15 +00:00
import torch
from torch import nn
from torch.nn import functional as F
import torch.distributed as distributed
from trainer.networks import register_model
from utils.util import checkpoint, opt_get
class Quantize(nn.Module):
2021-09-24 03:19:36 +00:00
def __init__(self, dim, n_embed, decay=0.99, eps=1e-5, balancing_heuristic=False):
2021-01-07 17:20:15 +00:00
super().__init__()
self.dim = dim
self.n_embed = n_embed
self.decay = decay
self.eps = eps
2021-09-24 03:19:36 +00:00
self.balancing_heuristic = balancing_heuristic
self.codes = None
self.max_codes = 64000
self.codes_full = False
2021-01-07 17:20:15 +00:00
embed = torch.randn(dim, n_embed)
self.register_buffer("embed", embed)
self.register_buffer("cluster_size", torch.zeros(n_embed))
self.register_buffer("embed_avg", embed.clone())
def forward(self, input):
2021-09-24 03:19:36 +00:00
if self.codes_full:
h = torch.histc(self.codes, bins=self.n_embed, min=0, max=self.n_embed) / len(self.codes)
mask = torch.logical_or(h > .9, h < .01).unsqueeze(1)
ep = self.embed.permute(1,0)
ea = self.embed_avg.permute(1,0)
rand_embed = torch.randn_like(ep) * mask
self.embed = (ep * ~mask + rand_embed).permute(1,0)
self.embed_avg = (ea * ~mask + rand_embed).permute(1,0)
self.cluster_size = self.cluster_size * ~mask.squeeze()
if torch.any(mask):
print(f"Reset {torch.sum(mask)} embedding codes.")
self.codes = None
self.codes_full = False
2021-01-07 17:20:15 +00:00
flatten = input.reshape(-1, self.dim)
dist = (
flatten.pow(2).sum(1, keepdim=True)
- 2 * flatten @ self.embed
+ self.embed.pow(2).sum(0, keepdim=True)
)
_, embed_ind = (-dist).max(1)
embed_onehot = F.one_hot(embed_ind, self.n_embed).type(flatten.dtype)
embed_ind = embed_ind.view(*input.shape[:-1])
quantize = self.embed_code(embed_ind)
2021-09-24 03:19:36 +00:00
if self.codes is None:
self.codes = embed_ind.flatten()
else:
self.codes = torch.cat([self.codes, embed_ind.flatten()])
if len(self.codes) > self.max_codes:
self.codes = self.codes[-self.max_codes:]
self.codes_full = True
2021-01-07 17:20:15 +00:00
if self.training:
embed_onehot_sum = embed_onehot.sum(0)
embed_sum = flatten.transpose(0, 1) @ embed_onehot
if distributed.is_initialized() and distributed.get_world_size() > 1:
distributed.all_reduce(embed_onehot_sum)
distributed.all_reduce(embed_sum)
self.cluster_size.data.mul_(self.decay).add_(
embed_onehot_sum, alpha=1 - self.decay
)
self.embed_avg.data.mul_(self.decay).add_(embed_sum, alpha=1 - self.decay)
n = self.cluster_size.sum()
cluster_size = (
(self.cluster_size + self.eps) / (n + self.n_embed * self.eps) * n
)
embed_normalized = self.embed_avg / cluster_size.unsqueeze(0)
self.embed.data.copy_(embed_normalized)
diff = (quantize.detach() - input).pow(2).mean()
quantize = input + (quantize - input).detach()
return quantize, diff, embed_ind
def embed_code(self, embed_id):
return F.embedding(embed_id, self.embed.transpose(0, 1))
class ResBlock(nn.Module):
def __init__(self, in_channel, channel, conv_module):
2021-01-07 17:20:15 +00:00
super().__init__()
self.conv = nn.Sequential(
nn.ReLU(inplace=True),
conv_module(in_channel, channel, 3, padding=1),
2021-01-07 17:20:15 +00:00
nn.ReLU(inplace=True),
conv_module(channel, in_channel, 1),
2021-01-07 17:20:15 +00:00
)
def forward(self, input):
out = self.conv(input)
out += input
return out
class Encoder(nn.Module):
def __init__(self, in_channel, channel, n_res_block, n_res_channel, stride, conv_module):
2021-01-07 17:20:15 +00:00
super().__init__()
if stride == 4:
blocks = [
conv_module(in_channel, channel // 2, 4, stride=2, padding=1),
2021-01-07 17:20:15 +00:00
nn.ReLU(inplace=True),
conv_module(channel // 2, channel, 4, stride=2, padding=1),
2021-01-07 17:20:15 +00:00
nn.ReLU(inplace=True),
conv_module(channel, channel, 3, padding=1),
2021-01-07 17:20:15 +00:00
]
elif stride == 2:
blocks = [
conv_module(in_channel, channel // 2, 4, stride=2, padding=1),
2021-01-07 17:20:15 +00:00
nn.ReLU(inplace=True),
conv_module(channel // 2, channel, 3, padding=1),
2021-01-07 17:20:15 +00:00
]
for i in range(n_res_block):
blocks.append(ResBlock(channel, n_res_channel, conv_module))
2021-01-07 17:20:15 +00:00
blocks.append(nn.ReLU(inplace=True))
self.blocks = nn.Sequential(*blocks)
def forward(self, input):
return self.blocks(input)
class Decoder(nn.Module):
def __init__(
self, in_channel, out_channel, channel, n_res_block, n_res_channel, stride, conv_module, conv_transpose_module
2021-01-07 17:20:15 +00:00
):
super().__init__()
blocks = [conv_module(in_channel, channel, 3, padding=1)]
2021-01-07 17:20:15 +00:00
for i in range(n_res_block):
blocks.append(ResBlock(channel, n_res_channel, conv_module))
2021-01-07 17:20:15 +00:00
blocks.append(nn.ReLU(inplace=True))
if stride == 4:
blocks.extend(
[
conv_transpose_module(channel, channel // 2, 4, stride=2, padding=1),
2021-01-07 17:20:15 +00:00
nn.ReLU(inplace=True),
conv_transpose_module(
2021-01-07 17:20:15 +00:00
channel // 2, out_channel, 4, stride=2, padding=1
),
]
)
elif stride == 2:
blocks.append(
conv_transpose_module(channel, out_channel, 4, stride=2, padding=1)
2021-01-07 17:20:15 +00:00
)
self.blocks = nn.Sequential(*blocks)
def forward(self, input):
return self.blocks(input)
class VQVAE(nn.Module):
def __init__(
self,
in_channel=3,
channel=128,
n_res_block=2,
n_res_channel=32,
codebook_dim=64,
codebook_size=512,
conv_module=nn.Conv2d,
conv_transpose_module=nn.ConvTranspose2d,
2021-01-07 17:20:15 +00:00
decay=0.99,
):
super().__init__()
self.unsqueeze_channels = in_channel == -1
in_channel = abs(in_channel)
self.codebook_size = codebook_size
self.enc_b = Encoder(in_channel, channel, n_res_block, n_res_channel, stride=4, conv_module=conv_module)
self.enc_t = Encoder(channel, channel, n_res_block, n_res_channel, stride=2, conv_module=conv_module)
self.quantize_conv_t = conv_module(channel, codebook_dim, 1)
2021-01-07 17:20:15 +00:00
self.quantize_t = Quantize(codebook_dim, codebook_size)
self.dec_t = Decoder(
codebook_dim, codebook_dim, channel, n_res_block, n_res_channel, stride=2, conv_module=conv_module, conv_transpose_module=conv_transpose_module
2021-01-07 17:20:15 +00:00
)
self.quantize_conv_b = conv_module(codebook_dim + channel, codebook_dim, 1)
2021-01-07 17:20:15 +00:00
self.quantize_b = Quantize(codebook_dim, codebook_size)
self.upsample_t = conv_transpose_module(
2021-01-07 17:20:15 +00:00
codebook_dim, codebook_dim, 4, stride=2, padding=1
)
self.dec = Decoder(
codebook_dim + codebook_dim,
in_channel,
channel,
n_res_block,
n_res_channel,
stride=4,
conv_module=conv_module,
conv_transpose_module=conv_transpose_module
2021-01-07 17:20:15 +00:00
)
def forward(self, input):
if self.unsqueeze_channels:
input = input.unsqueeze(1)
2021-01-07 17:20:15 +00:00
quant_t, quant_b, diff, _, _ = self.encode(input)
dec = self.decode(quant_t, quant_b)
if self.unsqueeze_channels:
dec = dec.squeeze(1)
2021-01-07 17:20:15 +00:00
return dec, diff
def encode(self, input):
enc_b = checkpoint(self.enc_b, input)
enc_t = checkpoint(self.enc_t, enc_b)
quant_t = self.quantize_conv_t(enc_t).permute((0,2,3,1) if len(input.shape) == 4 else (0,2,1))
2021-01-07 17:20:15 +00:00
quant_t, diff_t, id_t = self.quantize_t(quant_t)
quant_t = quant_t.permute((0,3,1,2) if len(input.shape) == 4 else (0,2,1))
2021-01-07 17:20:15 +00:00
diff_t = diff_t.unsqueeze(0)
dec_t = checkpoint(self.dec_t, quant_t)
enc_b = torch.cat([dec_t, enc_b], 1)
quant_b = checkpoint(self.quantize_conv_b, enc_b).permute((0,2,3,1) if len(input.shape) == 4 else (0,2,1))
2021-01-07 17:20:15 +00:00
quant_b, diff_b, id_b = self.quantize_b(quant_b)
quant_b = quant_b.permute((0,3,1,2) if len(input.shape) == 4 else (0,2,1))
2021-01-07 17:20:15 +00:00
diff_b = diff_b.unsqueeze(0)
return quant_t, quant_b, diff_t + diff_b, id_t, id_b
2021-08-04 03:08:27 +00:00
def encode_only_quantized(self, input):
qt, qb, d, idt, idb = self.encode(input)
# Append top and bottom into the same sequence, adding the codebook length onto the top to discriminate it.
idt += self.codebook_size
ids = torch.cat([idt, idb], dim=1)
2021-08-04 03:08:27 +00:00
return ids
2021-01-07 17:20:15 +00:00
def decode(self, quant_t, quant_b):
upsample_t = self.upsample_t(quant_t)
quant = torch.cat([upsample_t, quant_b], 1)
dec = checkpoint(self.dec, quant)
return dec
def decode_code(self, code_t, code_b):
quant_t = self.quantize_t.embed_code(code_t)
2021-08-04 03:08:27 +00:00
quant_t = quant_t.permute((0,3,1,2) if len(code_t.shape) == 4 else (0,2,1))
2021-01-07 17:20:15 +00:00
quant_b = self.quantize_b.embed_code(code_b)
2021-08-04 03:08:27 +00:00
quant_b = quant_b.permute((0,3,1,2) if len(code_t.shape) == 4 else (0,2,1))
2021-01-07 17:20:15 +00:00
dec = self.decode(quant_t, quant_b)
return dec
2021-08-04 03:08:27 +00:00
# Performs decode_code() with the outputs from encode_only_quantized.
def decode_code_joined(self, input):
b, s = input.shape
assert s % 3 == 0 # If not, this tensor didn't come from encode_only_quantized.
s = s // 3
# This doesn't work with batching. TODO: fixme.
t = input[:,:s] - self.codebook_size
b = input[:,s:]
2021-08-04 03:08:27 +00:00
return self.decode_code(t, b)
2021-01-07 17:20:15 +00:00
@register_model
def register_vqvae(opt_net, opt):
kw = opt_get(opt_net, ['kwargs'], {})
2021-02-05 05:26:36 +00:00
vq = VQVAE(**kw)
return vq
@register_model
def register_vqvae_audio(opt_net, opt):
kw = opt_get(opt_net, ['kwargs'], {})
kw['conv_module'] = nn.Conv1d
kw['conv_transpose_module'] = nn.ConvTranspose1d
vq = VQVAE(**kw)
return vq
if __name__ == '__main__':
model = VQVAE(in_channel=80, conv_module=nn.Conv1d, conv_transpose_module=nn.ConvTranspose1d)
2021-08-04 03:08:27 +00:00
#res=model(torch.randn(1,80,2048))
e = model.encode_only_quantized(torch.randn(1, 80, 2048))
k = model.decode_code_joined(e)
print(k.shape)