forked from mrq/DL-Art-School
266 lines
10 KiB
Python
266 lines
10 KiB
Python
from random import random
|
|
|
|
import torch
|
|
import torch.nn as nn
|
|
import torch.nn.functional as F
|
|
from torch import einsum
|
|
|
|
from models.arch_util import AttentionBlock
|
|
from models.lucidrains.x_transformers import ContinuousTransformerWrapper, Encoder
|
|
from trainer.networks import register_model
|
|
from utils.util import opt_get, checkpoint
|
|
|
|
|
|
def exists(val):
|
|
return val is not None
|
|
|
|
|
|
def masked_mean(t, mask):
|
|
t = t.masked_fill(~mask, 0.)
|
|
return t.sum(dim = 1) / mask.sum(dim = 1)
|
|
|
|
|
|
class InfoNCE(nn.Module):
|
|
"""
|
|
Calculates the InfoNCE loss for self-supervised learning.
|
|
This contrastive loss enforces the embeddings of similar (positive) samples to be close
|
|
and those of different (negative) samples to be distant.
|
|
A query embedding is compared with one positive key and with one or more negative keys.
|
|
|
|
References:
|
|
https://arxiv.org/abs/1807.03748v2
|
|
https://arxiv.org/abs/2010.05113
|
|
|
|
Args:
|
|
temperature: Logits are divided by temperature before calculating the cross entropy.
|
|
reduction: Reduction method applied to the output.
|
|
Value must be one of ['none', 'sum', 'mean'].
|
|
See torch.nn.functional.cross_entropy for more details about each option.
|
|
negative_mode: Determines how the (optional) negative_keys are handled.
|
|
Value must be one of ['paired', 'unpaired'].
|
|
If 'paired', then each query sample is paired with a number of negative keys.
|
|
Comparable to a triplet loss, but with multiple negatives per sample.
|
|
If 'unpaired', then the set of negative keys are all unrelated to any positive key.
|
|
|
|
Input shape:
|
|
query: (N, D) Tensor with query samples (e.g. embeddings of the input).
|
|
positive_key: (N, D) Tensor with positive samples (e.g. embeddings of augmented input).
|
|
negative_keys (optional): Tensor with negative samples (e.g. embeddings of other inputs)
|
|
If negative_mode = 'paired', then negative_keys is a (N, M, D) Tensor.
|
|
If negative_mode = 'unpaired', then negative_keys is a (M, D) Tensor.
|
|
If None, then the negative keys for a sample are the positive keys for the other samples.
|
|
|
|
Returns:
|
|
Value of the InfoNCE Loss.
|
|
|
|
Examples:
|
|
>>> loss = InfoNCE()
|
|
>>> batch_size, num_negative, embedding_size = 32, 48, 128
|
|
>>> query = torch.randn(batch_size, embedding_size)
|
|
>>> positive_key = torch.randn(batch_size, embedding_size)
|
|
>>> negative_keys = torch.randn(num_negative, embedding_size)
|
|
>>> output = loss(query, positive_key, negative_keys)
|
|
"""
|
|
|
|
def __init__(self, temperature=0.1, reduction='mean', negative_mode='unpaired'):
|
|
super().__init__()
|
|
self.temperature = temperature
|
|
self.reduction = reduction
|
|
self.negative_mode = negative_mode
|
|
|
|
def forward(self, query, positive_key, negative_keys=None):
|
|
return info_nce(query, positive_key, negative_keys,
|
|
temperature=self.temperature,
|
|
reduction=self.reduction,
|
|
negative_mode=self.negative_mode)
|
|
|
|
|
|
def info_nce(query, positive_key, negative_keys=None, temperature=0.1, reduction='mean', negative_mode='unpaired'):
|
|
# Check input dimensionality.
|
|
if query.dim() != 2:
|
|
raise ValueError('<query> must have 2 dimensions.')
|
|
if positive_key.dim() != 2:
|
|
raise ValueError('<positive_key> must have 2 dimensions.')
|
|
if negative_keys is not None:
|
|
if negative_mode == 'unpaired' and negative_keys.dim() != 2:
|
|
raise ValueError("<negative_keys> must have 2 dimensions if <negative_mode> == 'unpaired'.")
|
|
if negative_mode == 'paired' and negative_keys.dim() != 3:
|
|
raise ValueError("<negative_keys> must have 3 dimensions if <negative_mode> == 'paired'.")
|
|
|
|
# Check matching number of samples.
|
|
if len(query) != len(positive_key):
|
|
raise ValueError('<query> and <positive_key> must must have the same number of samples.')
|
|
if negative_keys is not None:
|
|
if negative_mode == 'paired' and len(query) != len(negative_keys):
|
|
raise ValueError("If negative_mode == 'paired', then <negative_keys> must have the same number of samples as <query>.")
|
|
|
|
# Embedding vectors should have same number of components.
|
|
if query.shape[-1] != positive_key.shape[-1]:
|
|
raise ValueError('Vectors of <query> and <positive_key> should have the same number of components.')
|
|
if negative_keys is not None:
|
|
if query.shape[-1] != negative_keys.shape[-1]:
|
|
raise ValueError('Vectors of <query> and <negative_keys> should have the same number of components.')
|
|
|
|
# Normalize to unit vectors
|
|
query, positive_key, negative_keys = normalize(query, positive_key, negative_keys)
|
|
if negative_keys is not None:
|
|
# Explicit negative keys
|
|
|
|
# Cosine between positive pairs
|
|
positive_logit = torch.sum(query * positive_key, dim=1, keepdim=True)
|
|
|
|
if negative_mode == 'unpaired':
|
|
# Cosine between all query-negative combinations
|
|
negative_logits = query @ transpose(negative_keys)
|
|
|
|
elif negative_mode == 'paired':
|
|
query = query.unsqueeze(1)
|
|
negative_logits = query @ transpose(negative_keys)
|
|
negative_logits = negative_logits.squeeze(1)
|
|
|
|
# First index in last dimension are the positive samples
|
|
logits = torch.cat([positive_logit, negative_logits], dim=1)
|
|
labels = torch.zeros(len(logits), dtype=torch.long, device=query.device)
|
|
else:
|
|
# Negative keys are implicitly off-diagonal positive keys.
|
|
|
|
# Cosine between all combinations
|
|
logits = query @ transpose(positive_key)
|
|
|
|
# Positive keys are the entries on the diagonal
|
|
labels = torch.arange(len(query), device=query.device)
|
|
|
|
return F.cross_entropy(logits / temperature, labels, reduction=reduction)
|
|
|
|
|
|
def transpose(x):
|
|
return x.transpose(-2, -1)
|
|
|
|
|
|
def normalize(*xs):
|
|
return [None if x is None else F.normalize(x, dim=-1) for x in xs]
|
|
|
|
|
|
class CollapsingTransformer(nn.Module):
|
|
def __init__(self, model_dim, output_dims, heads, dropout, depth, mask_percentage=0, **encoder_kwargs):
|
|
super().__init__()
|
|
self.transformer = ContinuousTransformerWrapper(
|
|
max_seq_len=-1,
|
|
use_pos_emb=False,
|
|
attn_layers=Encoder(
|
|
dim=model_dim,
|
|
depth=depth,
|
|
heads=heads,
|
|
ff_dropout=dropout,
|
|
ff_mult=1,
|
|
attn_dropout=dropout,
|
|
use_rmsnorm=True,
|
|
ff_glu=True,
|
|
rotary_pos_emb=True,
|
|
**encoder_kwargs,
|
|
))
|
|
self.pre_combiner = nn.Sequential(nn.Conv1d(model_dim, output_dims, 1),
|
|
AttentionBlock(output_dims, num_heads=heads, do_checkpoint=False),
|
|
nn.Conv1d(output_dims, output_dims, 1))
|
|
self.mask_percentage = mask_percentage
|
|
|
|
def forward(self, x, **transformer_kwargs):
|
|
h = self.transformer(x, **transformer_kwargs)
|
|
h = h.permute(0,2,1)
|
|
h = checkpoint(self.pre_combiner, h).permute(0,2,1)
|
|
if self.training:
|
|
mask = torch.rand_like(h.float()) > self.mask_percentage
|
|
else:
|
|
mask = torch.ones_like(h.float()).bool()
|
|
return masked_mean(h, mask)
|
|
|
|
|
|
class ConvFormatEmbedding(nn.Module):
|
|
def __init__(self, *args, **kwargs):
|
|
super().__init__()
|
|
self.emb = nn.Embedding(*args, **kwargs)
|
|
|
|
def forward(self, x):
|
|
y = self.emb(x)
|
|
return y.permute(0,2,1)
|
|
|
|
|
|
class ContrastiveAudio(nn.Module):
|
|
def __init__(
|
|
self,
|
|
model_dim=512,
|
|
transformer_heads=8,
|
|
dropout=.1,
|
|
encoder_depth=8,
|
|
mel_channels=80,
|
|
latent_multiplier=1,
|
|
mask_percent=.15,
|
|
):
|
|
super().__init__()
|
|
latent_dim = latent_multiplier*model_dim
|
|
self.temperature = nn.Parameter(torch.tensor(1.))
|
|
|
|
self.emb = nn.Sequential(nn.Conv1d(mel_channels, model_dim // 2, kernel_size=5, stride=2, padding=2),
|
|
nn.Conv1d(model_dim//2, model_dim, kernel_size=3, stride=2, padding=1))
|
|
self.transformer = CollapsingTransformer(model_dim, model_dim, transformer_heads, dropout, encoder_depth, mask_percent)
|
|
self.to_latent = nn.Linear(latent_dim, latent_dim, bias=False)
|
|
self.to_latent2 = nn.Linear(latent_dim, latent_dim, bias=False)
|
|
|
|
self.to_latent2.weight.data = self.to_latent.weight.data
|
|
self.to_latent2.weight.DO_NOT_TRAIN = True
|
|
self.to_latent2.requires_grad = False
|
|
|
|
def get_grad_norm_parameter_groups(self):
|
|
return {
|
|
'emb': list(self.emb.parameters()),
|
|
'xform': list(self.transformer.parameters()),
|
|
}
|
|
|
|
def update_for_step(self, step, __):
|
|
self.to_latent2.weight.data = self.to_latent2.weight.data * .99 + self.to_latent.weight.data * .01
|
|
|
|
def project(self, mel):
|
|
h1 = self.emb(mel).permute(0, 2, 1)
|
|
h1 = self.transformer(h1)
|
|
h1 = self.to_latent(h1)
|
|
return h1
|
|
|
|
def forward(
|
|
self,
|
|
mel_input1,
|
|
mel_input2
|
|
):
|
|
if len(mel_input2.shape) == 4:
|
|
mel_input2 = mel_input2[:, 0]
|
|
if self.training:
|
|
# Mask out big chunks of separate frequency bands for each clip.
|
|
b, c, _ = mel_input1.shape
|
|
mask = torch.rand(b,c,1, device=mel_input1.device) > .3
|
|
mel_input1 = mask * mel_input1 * (1-random()*.5)
|
|
mask = torch.rand(b,c,1, device=mel_input2.device) > .3
|
|
mel_input2 = mask * mel_input2 * (1-random()*.5)
|
|
|
|
h1 = self.emb(mel_input1).permute(0, 2, 1)
|
|
h1 = self.transformer(h1)
|
|
h1 = self.to_latent(h1)
|
|
|
|
h2 = self.emb(mel_input2).permute(0, 2, 1)
|
|
h2 = self.transformer(h2)
|
|
h2 = self.to_latent2(h2).detach()
|
|
|
|
loss = info_nce(h1, h2)
|
|
return loss
|
|
|
|
|
|
@register_model
|
|
def register_contrastive_audio(opt_net, opt):
|
|
return ContrastiveAudio(**opt_get(opt_net, ['kwargs'], {}))
|
|
|
|
|
|
if __name__ == '__main__':
|
|
clvp = ContrastiveAudio()
|
|
clvp(torch.randn(2,80,100),
|
|
torch.randn(2,80,95),
|
|
return_loss=True)
|
|
v = torch.randn(2,512)
|
|
print(info_nce(v,v)) |