stable-diffusion-finetune/ldm/modules/encoders/modules.py

341 lines
13 KiB
Python
Raw Normal View History

2021-12-21 03:23:41 +01:00
import torch
import torch.nn as nn
2022-06-13 00:39:48 +02:00
import numpy as np
2021-12-21 03:23:41 +01:00
from functools import partial
import kornia
2021-12-21 03:23:41 +01:00
from ldm.modules.x_transformer import Encoder, TransformerWrapper # TODO: can we directly rely on lucidrains code and simply add this as a reuirement? --> test
2022-06-13 00:39:48 +02:00
from ldm.util import default
import clip
2021-12-21 03:23:41 +01:00
class AbstractEncoder(nn.Module):
def __init__(self):
super().__init__()
def encode(self, *args, **kwargs):
raise NotImplementedError
2022-06-05 19:22:07 +02:00
class IdentityEncoder(AbstractEncoder):
def encode(self, x):
return x
2021-12-21 03:23:41 +01:00
class ClassEmbedder(nn.Module):
def __init__(self, embed_dim, n_classes=1000, key='class'):
super().__init__()
self.key = key
self.embedding = nn.Embedding(n_classes, embed_dim)
def forward(self, batch, key=None):
if key is None:
key = self.key
# this is for use in crossattn
c = batch[key][:, None]
c = self.embedding(c)
return c
class TransformerEmbedder(AbstractEncoder):
"""Some transformer encoder layers"""
def __init__(self, n_embed, n_layer, vocab_size, max_seq_len=77, device="cuda"):
super().__init__()
self.device = device
self.transformer = TransformerWrapper(num_tokens=vocab_size, max_seq_len=max_seq_len,
attn_layers=Encoder(dim=n_embed, depth=n_layer))
def forward(self, tokens):
tokens = tokens.to(self.device) # meh
z = self.transformer(tokens, return_embeddings=True)
return z
def encode(self, x):
return self(x)
class BERTTokenizer(AbstractEncoder):
""" Uses a pretrained BERT tokenizer by huggingface. Vocab size: 30522 (?)"""
def __init__(self, device="cuda", vq_interface=True, max_length=77):
super().__init__()
from transformers import BertTokenizerFast # TODO: add to reuquirements
self.tokenizer = BertTokenizerFast.from_pretrained("bert-base-uncased")
self.device = device
self.vq_interface = vq_interface
self.max_length = max_length
def forward(self, text):
batch_encoding = self.tokenizer(text, truncation=True, max_length=self.max_length, return_length=True,
return_overflowing_tokens=False, padding="max_length", return_tensors="pt")
tokens = batch_encoding["input_ids"].to(self.device)
return tokens
@torch.no_grad()
def encode(self, text):
tokens = self(text)
if not self.vq_interface:
return tokens
return None, None, [None, None, tokens]
def decode(self, text):
return text
class BERTEmbedder(AbstractEncoder):
"""Uses the BERT tokenizr model and add some transformer encoder layers"""
def __init__(self, n_embed, n_layer, vocab_size=30522, max_seq_len=77,
device="cuda",use_tokenizer=True, embedding_dropout=0.0):
super().__init__()
self.use_tknz_fn = use_tokenizer
if self.use_tknz_fn:
self.tknz_fn = BERTTokenizer(vq_interface=False, max_length=max_seq_len)
self.device = device
self.transformer = TransformerWrapper(num_tokens=vocab_size, max_seq_len=max_seq_len,
attn_layers=Encoder(dim=n_embed, depth=n_layer),
emb_dropout=embedding_dropout)
def forward(self, text):
if self.use_tknz_fn:
tokens = self.tknz_fn(text)#.to(self.device)
else:
tokens = text
z = self.transformer(tokens, return_embeddings=True)
return z
def encode(self, text):
# output of length 77
return self(text)
2022-05-31 12:28:00 +02:00
from transformers import T5Tokenizer, T5EncoderModel, CLIPTokenizer, CLIPTextModel
2022-05-31 11:42:53 +02:00
def disabled_train(self, mode=True):
"""Overwrite model.train with this function to make sure train/eval mode
does not change anymore."""
return self
class FrozenT5Embedder(AbstractEncoder):
"""Uses the T5 transformer encoder for text"""
def __init__(self, version="google/t5-v1_1-large", device="cuda", max_length=77): # others are google/t5-v1_1-xl and google/t5-v1_1-xxl
super().__init__()
self.tokenizer = T5Tokenizer.from_pretrained(version)
self.transformer = T5EncoderModel.from_pretrained(version)
self.device = device
self.max_length = max_length # TODO: typical value?
2022-05-31 12:28:00 +02:00
self.freeze()
2022-05-31 11:42:53 +02:00
2022-05-31 12:28:00 +02:00
def freeze(self):
2022-05-31 11:42:53 +02:00
self.transformer = self.transformer.eval()
2022-05-31 14:18:01 +02:00
#self.train = disabled_train
2022-05-31 11:42:53 +02:00
for param in self.parameters():
param.requires_grad = False
def forward(self, text):
batch_encoding = self.tokenizer(text, truncation=True, max_length=self.max_length, return_length=True,
return_overflowing_tokens=False, padding="max_length", return_tensors="pt")
tokens = batch_encoding["input_ids"].to(self.device)
2022-05-31 12:28:00 +02:00
outputs = self.transformer(input_ids=tokens)
z = outputs.last_hidden_state
return z
def encode(self, text):
return self(text)
class FrozenCLIPEmbedder(AbstractEncoder):
"""Uses the CLIP transformer encoder for text (from huggingface)"""
2022-05-31 12:36:26 +02:00
def __init__(self, version="openai/clip-vit-large-patch14", device="cuda", max_length=77): # clip-vit-base-patch32
2022-05-31 12:28:00 +02:00
super().__init__()
self.tokenizer = CLIPTokenizer.from_pretrained(version)
self.transformer = CLIPTextModel.from_pretrained(version)
self.device = device
self.max_length = max_length # TODO: typical value?
self.freeze()
def freeze(self):
self.transformer = self.transformer.eval()
2022-05-31 14:18:01 +02:00
#self.train = disabled_train
2022-05-31 12:28:00 +02:00
for param in self.parameters():
param.requires_grad = False
def forward(self, text):
batch_encoding = self.tokenizer(text, truncation=True, max_length=self.max_length, return_length=True,
return_overflowing_tokens=False, padding="max_length", return_tensors="pt")
tokens = batch_encoding["input_ids"].to(self.device)
outputs = self.transformer(input_ids=tokens)
2022-05-31 11:42:53 +02:00
z = outputs.last_hidden_state
return z
def encode(self, text):
return self(text)
class ProjectedFrozenCLIPEmbedder(AbstractEncoder):
def __init__(self, version="openai/clip-vit-large-patch14", device="cuda", max_length=77): # clip-vit-base-patch32
super().__init__()
self.embedder = FrozenCLIPEmbedder(version=version, device=device, max_length=max_length)
self.projection = torch.nn.Linear(768, 768)
def forward(self, text):
z = self.embedder(text)
return self.projection(z)
def encode(self, text):
return self(text)
class FrozenCLIPImageEmbedder(AbstractEncoder):
"""
Uses the CLIP image encoder.
"""
def __init__(
self,
model='ViT-L/14',
jit=False,
device='cuda' if torch.cuda.is_available() else 'cpu',
antialias=False,
):
super().__init__()
self.model, _ = clip.load(name=model, device=device, jit=jit)
self.device = device
self.antialias = antialias
self.register_buffer('mean', torch.Tensor([0.48145466, 0.4578275, 0.40821073]), persistent=False)
self.register_buffer('std', torch.Tensor([0.26862954, 0.26130258, 0.27577711]), persistent=False)
# I didn't call this originally, but seems like it was frozen anyway
self.freeze()
def freeze(self):
self.transformer = self.transformer.eval()
for param in self.parameters():
param.requires_grad = False
def preprocess(self, x):
# Expects inputs in the range -1, 1
x = kornia.geometry.resize(x, (224, 224),
interpolation='bicubic',align_corners=True,
antialias=self.antialias)
x = (x + 1.) / 2.
# renormalize according to clip
x = kornia.enhance.normalize(x, self.mean, self.std)
return x
def forward(self, x):
# x is assumed to be in range [-1,1]
return self.model.encode_image(self.preprocess(x)).float()
def encode(self, im):
return self(im).unsqueeze(1)
2022-05-31 11:42:53 +02:00
2021-12-21 03:23:41 +01:00
class SpatialRescaler(nn.Module):
def __init__(self,
n_stages=1,
method='bilinear',
multiplier=0.5,
in_channels=3,
out_channels=None,
bias=False):
super().__init__()
self.n_stages = n_stages
assert self.n_stages >= 0
assert method in ['nearest','linear','bilinear','trilinear','bicubic','area']
self.multiplier = multiplier
self.interpolator = partial(torch.nn.functional.interpolate, mode=method)
self.remap_output = out_channels is not None
if self.remap_output:
print(f'Spatial Rescaler mapping from {in_channels} to {out_channels} channels after resizing.')
self.channel_mapper = nn.Conv2d(in_channels,out_channels,1,bias=bias)
def forward(self,x):
for stage in range(self.n_stages):
x = self.interpolator(x, scale_factor=self.multiplier)
if self.remap_output:
x = self.channel_mapper(x)
return x
def encode(self, x):
return self(x)
2022-05-31 11:42:53 +02:00
2022-06-13 00:39:48 +02:00
from ldm.util import instantiate_from_config
from ldm.modules.diffusionmodules.util import make_beta_schedule, extract_into_tensor, noise_like
class LowScaleEncoder(nn.Module):
2022-06-13 10:43:41 +02:00
def __init__(self, model_config, linear_start, linear_end, timesteps=1000, max_noise_level=250, output_size=64,
scale_factor=1.0):
2022-06-13 00:39:48 +02:00
super().__init__()
self.max_noise_level = max_noise_level
self.model = instantiate_from_config(model_config)
self.augmentation_schedule = self.register_schedule(timesteps=timesteps, linear_start=linear_start,
linear_end=linear_end)
self.out_size = output_size
2022-06-13 10:43:41 +02:00
self.scale_factor = scale_factor
2022-06-13 00:39:48 +02:00
def register_schedule(self, beta_schedule="linear", timesteps=1000,
linear_start=1e-4, linear_end=2e-2, cosine_s=8e-3):
betas = make_beta_schedule(beta_schedule, timesteps, linear_start=linear_start, linear_end=linear_end,
cosine_s=cosine_s)
alphas = 1. - betas
alphas_cumprod = np.cumprod(alphas, axis=0)
alphas_cumprod_prev = np.append(1., alphas_cumprod[:-1])
timesteps, = betas.shape
self.num_timesteps = int(timesteps)
self.linear_start = linear_start
self.linear_end = linear_end
assert alphas_cumprod.shape[0] == self.num_timesteps, 'alphas have to be defined for each timestep'
to_torch = partial(torch.tensor, dtype=torch.float32)
self.register_buffer('betas', to_torch(betas))
self.register_buffer('alphas_cumprod', to_torch(alphas_cumprod))
self.register_buffer('alphas_cumprod_prev', to_torch(alphas_cumprod_prev))
# calculations for diffusion q(x_t | x_{t-1}) and others
self.register_buffer('sqrt_alphas_cumprod', to_torch(np.sqrt(alphas_cumprod)))
self.register_buffer('sqrt_one_minus_alphas_cumprod', to_torch(np.sqrt(1. - alphas_cumprod)))
self.register_buffer('log_one_minus_alphas_cumprod', to_torch(np.log(1. - alphas_cumprod)))
self.register_buffer('sqrt_recip_alphas_cumprod', to_torch(np.sqrt(1. / alphas_cumprod)))
self.register_buffer('sqrt_recipm1_alphas_cumprod', to_torch(np.sqrt(1. / alphas_cumprod - 1)))
def q_sample(self, x_start, t, noise=None):
noise = default(noise, lambda: torch.randn_like(x_start))
return (extract_into_tensor(self.sqrt_alphas_cumprod, t, x_start.shape) * x_start +
extract_into_tensor(self.sqrt_one_minus_alphas_cumprod, t, x_start.shape) * noise)
def forward(self, x):
z = self.model.encode(x).sample()
2022-06-13 10:43:41 +02:00
z = z * self.scale_factor
2022-06-13 00:39:48 +02:00
noise_level = torch.randint(0, self.max_noise_level, (x.shape[0],), device=x.device).long()
z = self.q_sample(z, noise_level)
2022-07-28 00:08:46 +02:00
if self.out_size is not None:
z = torch.nn.functional.interpolate(z, size=self.out_size, mode="nearest") # TODO: experiment with mode
# z = z.repeat_interleave(2, -2).repeat_interleave(2, -1)
2022-06-13 00:39:48 +02:00
return z, noise_level
2022-06-13 10:43:41 +02:00
def decode(self, z):
z = z / self.scale_factor
return self.model.decode(z)
2022-06-13 00:39:48 +02:00
2022-05-31 11:42:53 +02:00
if __name__ == "__main__":
from ldm.util import count_params
sentences = ["a hedgehog drinking a whiskey", "der mond ist aufgegangen", "Ein Satz mit vielen Sonderzeichen: äöü ß ?! : 'xx-y/@s'"]
2022-05-31 12:28:00 +02:00
model = FrozenT5Embedder(version="google/t5-v1_1-xl").cuda()
count_params(model, True)
z = model(sentences)
print(z.shape)
model = FrozenCLIPEmbedder().cuda()
2022-05-31 11:42:53 +02:00
count_params(model, True)
z = model(sentences)
print(z.shape)
2022-05-31 12:28:00 +02:00
print("done.")