# Copyright 2023 The HuggingFace Team. 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.
from dataclasses import dataclass
from typing import Any, Dict, List, Optional, Tuple, Union
import torch
import torch.nn as nn
import torch.utils.checkpoint
from diffusers.configuration_utils import ConfigMixin, register_to_config
from diffusers.loaders import UNet2DConditionLoadersMixin
from diffusers.models.activations import get_activation
from diffusers.models.attention_processor import (
ADDED_KV_ATTENTION_PROCESSORS,
CROSS_ATTENTION_PROCESSORS,
AttentionProcessor,
AttnAddedKVProcessor,
AttnProcessor,
)
from diffusers.models.embeddings import (
TimestepEmbedding,
Timesteps,
)
from diffusers.models.modeling_utils import ModelMixin
from diffusers.models.resnet import Downsample2D, ResnetBlock2D, Upsample2D
from diffusers.models.transformer_2d import Transformer2DModel
from diffusers.models.unet_2d_blocks import DownBlock2D, UpBlock2D
from diffusers.models.unet_2d_condition import UNet2DConditionOutput
from diffusers.utils import BaseOutput, is_torch_version, logging
logger = logging.get_logger(__name__) # pylint: disable=invalid-name
def add_special_tokens(hidden_states, attention_mask, sos_token, eos_token):
batch_size = hidden_states.shape[0]
if attention_mask is not None:
# Add two more steps to attn mask
new_attn_mask_step = attention_mask.new_ones((batch_size, 1))
attention_mask = torch.concat([new_attn_mask_step, attention_mask, new_attn_mask_step], dim=-1)
# Add the SOS / EOS tokens at the start / end of the sequence respectively
sos_token = sos_token.expand(batch_size, 1, -1)
eos_token = eos_token.expand(batch_size, 1, -1)
hidden_states = torch.concat([sos_token, hidden_states, eos_token], dim=1)
return hidden_states, attention_mask
@dataclass
class AudioLDM2ProjectionModelOutput(BaseOutput):
"""
Args:
Class for AudioLDM2 projection layer's outputs.
hidden_states (`torch.FloatTensor` of shape `(batch_size, sequence_length, hidden_size)`):
Sequence of hidden-states obtained by linearly projecting the hidden-states for each of the text
encoders and subsequently concatenating them together.
attention_mask (`torch.Tensor` of shape `(batch_size, sequence_length)`, *optional*):
Mask to avoid performing attention on padding token indices, formed by concatenating the attention masks
for the two text encoders together. Mask values selected in `[0, 1]`:
- 1 for tokens that are **not masked**,
- 0 for tokens that are **masked**.
"""
hidden_states: torch.FloatTensor
attention_mask: Optional[torch.LongTensor] = None
class AudioLDM2ProjectionModel(ModelMixin, ConfigMixin):
"""
A simple linear projection model to map two text embeddings to a shared latent space. It also inserts learned
embedding vectors at the start and end of each text embedding sequence respectively. Each variable appended with
`_1` refers to that corresponding to the second text encoder. Otherwise, it is from the first.
Args:
text_encoder_dim (`int`):
Dimensionality of the text embeddings from the first text encoder (CLAP).
text_encoder_1_dim (`int`):
Dimensionality of the text embeddings from the second text encoder (T5 or VITS).
langauge_model_dim (`int`):
Dimensionality of the text embeddings from the language model (GPT2).
"""
@register_to_config
def __init__(self, text_encoder_dim, text_encoder_1_dim, langauge_model_dim):
super().__init__()
# additional projection layers for each text encoder
self.projection = nn.Linear(text_encoder_dim, langauge_model_dim)
self.projection_1 = nn.Linear(text_encoder_1_dim, langauge_model_dim)
# learnable SOS / EOS token embeddings for each text encoder
self.sos_embed = nn.Parameter(torch.ones(langauge_model_dim))
self.eos_embed = nn.Parameter(torch.ones(langauge_model_dim))
self.sos_embed_1 = nn.Parameter(torch.ones(langauge_model_dim))
self.eos_embed_1 = nn.Parameter(torch.ones(langauge_model_dim))
def forward(
self,
hidden_states: Optional[torch.FloatTensor] = None,
hidden_states_1: Optional[torch.FloatTensor] = None,
attention_mask: Optional[torch.LongTensor] = None,
attention_mask_1: Optional[torch.LongTensor] = None,
):
hidden_states = self.projection(hidden_states)
hidden_states, attention_mask = add_special_tokens(
hidden_states, attention_mask, sos_token=self.sos_embed, eos_token=self.eos_embed
)
hidden_states_1 = self.projection_1(hidden_states_1)
hidden_states_1, attention_mask_1 = add_special_tokens(
hidden_states_1, attention_mask_1, sos_token=self.sos_embed_1, eos_token=self.eos_embed_1
)
# concatenate clap and t5 text encoding
hidden_states = torch.cat([hidden_states, hidden_states_1], dim=1)
# concatenate attention masks
if attention_mask is None and attention_mask_1 is not None:
attention_mask = attention_mask_1.new_ones((hidden_states[:2]))
elif attention_mask is not None and attention_mask_1 is None:
attention_mask_1 = attention_mask.new_ones((hidden_states_1[:2]))
if attention_mask is not None and attention_mask_1 is not None:
attention_mask = torch.cat([attention_mask, attention_mask_1], dim=-1)
else:
attention_mask = None
return AudioLDM2ProjectionModelOutput(
hidden_states=hidden_states,
attention_mask=attention_mask,
)
class AudioLDM2UNet2DConditionModel(ModelMixin, ConfigMixin, UNet2DConditionLoadersMixin):
r"""
A conditional 2D UNet model that takes a noisy sample, conditional state, and a timestep and returns a sample
shaped output. Compared to the vanilla [`UNet2DConditionModel`], this variant optionally includes an additional
self-attention layer in each Transformer block, as well as multiple cross-attention layers. It also allows for up
to two cross-attention embeddings, `encoder_hidden_states` and `encoder_hidden_states_1`.
This model inherits from [`ModelMixin`]. Check the superclass documentation for it's generic methods implemented
for all models (such as downloading or saving).
Parameters:
sample_size (`int` or `Tuple[int, int]`, *optional*, defaults to `None`):
Height and width of input/output sample.
in_channels (`int`, *optional*, defaults to 4): Number of channels in the input sample.
out_channels (`int`, *optional*, defaults to 4): Number of channels in the output.
flip_sin_to_cos (`bool`, *optional*, defaults to `False`):
Whether to flip the sin to cos in the time embedding.
freq_shift (`int`, *optional*, defaults to 0): The frequency shift to apply to the time embedding.
down_block_types (`Tuple[str]`, *optional*, defaults to `("CrossAttnDownBlock2D", "CrossAttnDownBlock2D", "CrossAttnDownBlock2D", "DownBlock2D")`):
The tuple of downsample blocks to use.
mid_block_type (`str`, *optional*, defaults to `"UNetMidBlock2DCrossAttn"`):
Block type for middle of UNet, it can only be `UNetMidBlock2DCrossAttn` for AudioLDM2.
up_block_types (`Tuple[str]`, *optional*, defaults to `("UpBlock2D", "CrossAttnUpBlock2D", "CrossAttnUpBlock2D", "CrossAttnUpBlock2D")`):
The tuple of upsample blocks to use.
only_cross_attention (`bool` or `Tuple[bool]`, *optional*, default to `False`):
Whether to include self-attention in the basic transformer blocks, see
[`~models.attention.BasicTransformerBlock`].
block_out_channels (`Tuple[int]`, *optional*, defaults to `(320, 640, 1280, 1280)`):
The tuple of output channels for each block.
layers_per_block (`int`, *optional*, defaults to 2): The number of layers per block.
downsample_padding (`int`, *optional*, defaults to 1): The padding to use for the downsampling convolution.
mid_block_scale_factor (`float`, *optional*, defaults to 1.0): The scale factor to use for the mid block.
act_fn (`str`, *optional*, defaults to `"silu"`): The activation function to use.
norm_num_groups (`int`, *optional*, defaults to 32): The number of groups to use for the normalization.
If `None`, normalization and activation layers is skipped in post-processing.
norm_eps (`float`, *optional*, defaults to 1e-5): The epsilon to use for the normalization.
cross_attention_dim (`int` or `Tuple[int]`, *optional*, defaults to 1280):
The dimension of the cross attention features.
transformer_layers_per_block (`int` or `Tuple[int]`, *optional*, defaults to 1):
The number of transformer blocks of type [`~models.attention.BasicTransformerBlock`]. Only relevant for
[`~models.unet_2d_blocks.CrossAttnDownBlock2D`], [`~models.unet_2d_blocks.CrossAttnUpBlock2D`],
[`~models.unet_2d_blocks.UNetMidBlock2DCrossAttn`].
attention_head_dim (`int`, *optional*, defaults to 8): The dimension of the attention heads.
num_attention_heads (`int`, *optional*):
The number of attention heads. If not defined, defaults to `attention_head_dim`
resnet_time_scale_shift (`str`, *optional*, defaults to `"default"`): Time scale shift config
for ResNet blocks (see [`~models.resnet.ResnetBlock2D`]). Choose from `default` or `scale_shift`.
class_embed_type (`str`, *optional*, defaults to `None`):
The type of class embedding to use which is ultimately summed with the time embeddings. Choose from `None`,
`"timestep"`, `"identity"`, `"projection"`, or `"simple_projection"`.
num_class_embeds (`int`, *optional*, defaults to `None`):
Input dimension of the learnable embedding matrix to be projected to `time_embed_dim`, when performing
class conditioning with `class_embed_type` equal to `None`.
time_embedding_type (`str`, *optional*, defaults to `positional`):
The type of position embedding to use for timesteps. Choose from `positional` or `fourier`.
time_embedding_dim (`int`, *optional*, defaults to `None`):
An optional override for the dimension of the projected time embedding.
time_embedding_act_fn (`str`, *optional*, defaults to `None`):
Optional activation function to use only once on the time embeddings before they are passed to the rest of
the UNet. Choose from `silu`, `mish`, `gelu`, and `swish`.
timestep_post_act (`str`, *optional*, defaults to `None`):
The second activation function to use in timestep embedding. Choose from `silu`, `mish` and `gelu`.
time_cond_proj_dim (`int`, *optional*, defaults to `None`):
The dimension of `cond_proj` layer in the timestep embedding.
conv_in_kernel (`int`, *optional*, default to `3`): The kernel size of `conv_in` layer.
conv_out_kernel (`int`, *optional*, default to `3`): The kernel size of `conv_out` layer.
projection_class_embeddings_input_dim (`int`, *optional*): The dimension of the `class_labels` input when
`class_embed_type="projection"`. Required when `class_embed_type="projection"`.
class_embeddings_concat (`bool`, *optional*, defaults to `False`): Whether to concatenate the time
embeddings with the class embeddings.
"""
_supports_gradient_checkpointing = True
@register_to_config
def __init__(
self,
sample_size: Optional[int] = None,
in_channels: int = 4,
out_channels: int = 4,
flip_sin_to_cos: bool = True,
freq_shift: int = 0,
down_block_types: Tuple[str] = (
"CrossAttnDownBlock2D",
"CrossAttnDownBlock2D",
"CrossAttnDownBlock2D",
"DownBlock2D",
),
mid_block_type: Optional[str] = "UNetMidBlock2DCrossAttn",
up_block_types: Tuple[str] = ("UpBlock2D", "CrossAttnUpBlock2D", "CrossAttnUpBlock2D", "CrossAttnUpBlock2D"),
only_cross_attention: Union[bool, Tuple[bool]] = False,
block_out_channels: Tuple[int] = (320, 640, 1280, 1280),
layers_per_block: Union[int, Tuple[int]] = 2,
downsample_padding: int = 1,
mid_block_scale_factor: float = 1,
act_fn: str = "silu",
norm_num_groups: Optional[int] = 32,
norm_eps: float = 1e-5,
cross_attention_dim: Union[int, Tuple[int]] = 1280,
transformer_layers_per_block: Union[int, Tuple[int]] = 1,
attention_head_dim: Union[int, Tuple[int]] = 8,
num_attention_heads: Optional[Union[int, Tuple[int]]] = None,
use_linear_projection: bool = False,
class_embed_type: Optional[str] = None,
num_class_embeds: Optional[int] = None,
upcast_attention: bool = False,
resnet_time_scale_shift: str = "default",
time_embedding_type: str = "positional",
time_embedding_dim: Optional[int] = None,
time_embedding_act_fn: Optional[str] = None,
timestep_post_act: Optional[str] = None,
time_cond_proj_dim: Optional[int] = None,
conv_in_kernel: int = 3,
conv_out_kernel: int = 3,
projection_class_embeddings_input_dim: Optional[int] = None,
class_embeddings_concat: bool = False,
):
super().__init__()
self.sample_size = sample_size
if num_attention_heads is not None:
raise ValueError(
"At the moment it is not possible to define the number of attention heads via `num_attention_heads` because of a naming issue as described in https://github.com/huggingface/diffusers/issues/2011#issuecomment-1547958131. Passing `num_attention_heads` will only be supported in diffusers v0.19."
)
# If `num_attention_heads` is not defined (which is the case for most models)
# it will default to `attention_head_dim`. This looks weird upon first reading it and it is.
# The reason for this behavior is to correct for incorrectly named variables that were introduced
# when this library was created. The incorrect naming was only discovered much later in https://github.com/huggingface/diffusers/issues/2011#issuecomment-1547958131
# Changing `attention_head_dim` to `num_attention_heads` for 40,000+ configurations is too backwards breaking
# which is why we correct for the naming here.
num_attention_heads = num_attention_heads or attention_head_dim
# Check inputs
if len(down_block_types) != len(up_block_types):
raise ValueError(
f"Must provide the same number of `down_block_types` as `up_block_types`. `down_block_types`: {down_block_types}. `up_block_types`: {up_block_types}."
)
if len(block_out_channels) != len(down_block_types):
raise ValueError(
f"Must provide the same number of `block_out_channels` as `down_block_types`. `block_out_channels`: {block_out_channels}. `down_block_types`: {down_block_types}."
)
if not isinstance(only_cross_attention, bool) and len(only_cross_attention) != len(down_block_types):
raise ValueError(
f"Must provide the same number of `only_cross_attention` as `down_block_types`. `only_cross_attention`: {only_cross_attention}. `down_block_types`: {down_block_types}."
)
if not isinstance(num_attention_heads, int) and len(num_attention_heads) != len(down_block_types):
raise ValueError(
f"Must provide the same number of `num_attention_heads` as `down_block_types`. `num_attention_heads`: {num_attention_heads}. `down_block_types`: {down_block_types}."
)
if not isinstance(attention_head_dim, int) and len(attention_head_dim) != len(down_block_types):
raise ValueError(
f"Must provide the same number of `attention_head_dim` as `down_block_types`. `attention_head_dim`: {attention_head_dim}. `down_block_types`: {down_block_types}."
)
if isinstance(cross_attention_dim, list) and len(cross_attention_dim) != len(down_block_types):
raise ValueError(
f"Must provide the same number of `cross_attention_dim` as `down_block_types`. `cross_attention_dim`: {cross_attention_dim}. `down_block_types`: {down_block_types}."
)
if not isinstance(layers_per_block, int) and len(layers_per_block) != len(down_block_types):
raise ValueError(
f"Must provide the same number of `layers_per_block` as `down_block_types`. `layers_per_block`: {layers_per_block}. `down_block_types`: {down_block_types}."
)
# input
conv_in_padding = (conv_in_kernel - 1) // 2
self.conv_in = nn.Conv2d(
in_channels, block_out_channels[0], kernel_size=conv_in_kernel, padding=conv_in_padding
)
# time
if time_embedding_type == "positional":
time_embed_dim = time_embedding_dim or block_out_channels[0] * 4
self.time_proj = Timesteps(block_out_channels[0], flip_sin_to_cos, freq_shift)
timestep_input_dim = block_out_channels[0]
else:
raise ValueError(f"{time_embedding_type} does not exist. Please make sure to use `positional`.")
self.time_embedding = TimestepEmbedding(
timestep_input_dim,
time_embed_dim,
act_fn=act_fn,
post_act_fn=timestep_post_act,
cond_proj_dim=time_cond_proj_dim,
)
# class embedding
if class_embed_type is None and num_class_embeds is not None:
self.class_embedding = nn.Embedding(num_class_embeds, time_embed_dim)
elif class_embed_type == "timestep":
self.class_embedding = TimestepEmbedding(timestep_input_dim, time_embed_dim, act_fn=act_fn)
elif class_embed_type == "identity":
self.class_embedding = nn.Identity(time_embed_dim, time_embed_dim)
elif class_embed_type == "projection":
if projection_class_embeddings_input_dim is None:
raise ValueError(
"`class_embed_type`: 'projection' requires `projection_class_embeddings_input_dim` be set"
)
# The projection `class_embed_type` is the same as the timestep `class_embed_type` except
# 1. the `class_labels` inputs are not first converted to sinusoidal embeddings
# 2. it projects from an arbitrary input dimension.
#
# Note that `TimestepEmbedding` is quite general, being mainly linear layers and activations.
# When used for embedding actual timesteps, the timesteps are first converted to sinusoidal embeddings.
# As a result, `TimestepEmbedding` can be passed arbitrary vectors.
self.class_embedding = TimestepEmbedding(projection_class_embeddings_input_dim, time_embed_dim)
elif class_embed_type == "simple_projection":
if projection_class_embeddings_input_dim is None:
raise ValueError(
"`class_embed_type`: 'simple_projection' requires `projection_class_embeddings_input_dim` be set"
)
self.class_embedding = nn.Linear(projection_class_embeddings_input_dim, time_embed_dim)
else:
self.class_embedding = None
if time_embedding_act_fn is None:
self.time_embed_act = None
else:
self.time_embed_act = get_activation(time_embedding_act_fn)
self.down_blocks = nn.ModuleList([])
self.up_blocks = nn.ModuleList([])
if isinstance(only_cross_attention, bool):
only_cross_attention = [only_cross_attention] * len(down_block_types)
if isinstance(num_attention_heads, int):
num_attention_heads = (num_attention_heads,) * len(down_block_types)
if isinstance(cross_attention_dim, int):
cross_attention_dim = (cross_attention_dim,) * len(down_block_types)
if isinstance(layers_per_block, int):
layers_per_block = [layers_per_block] * len(down_block_types)
if isinstance(transformer_layers_per_block, int):
transformer_layers_per_block = [transformer_layers_per_block] * len(down_block_types)
if class_embeddings_concat:
# The time embeddings are concatenated with the class embeddings. The dimension of the
# time embeddings passed to the down, middle, and up blocks is twice the dimension of the
# regular time embeddings
blocks_time_embed_dim = time_embed_dim * 2
else:
blocks_time_embed_dim = time_embed_dim
# down
output_channel = block_out_channels[0]
for i, down_block_type in enumerate(down_block_types):
input_channel = output_channel
output_channel = block_out_channels[i]
is_final_block = i == len(block_out_channels) - 1
down_block = get_down_block(
down_block_type,
num_layers=layers_per_block[i],
transformer_layers_per_block=transformer_layers_per_block[i],
in_channels=input_channel,
out_channels=output_channel,
temb_channels=blocks_time_embed_dim,
add_downsample=not is_final_block,
resnet_eps=norm_eps,
resnet_act_fn=act_fn,
resnet_groups=norm_num_groups,
cross_attention_dim=cross_attention_dim[i],
num_attention_heads=num_attention_heads[i],
downsample_padding=downsample_padding,
use_linear_projection=use_linear_projection,
only_cross_attention=only_cross_attention[i],
upcast_attention=upcast_attention,
resnet_time_scale_shift=resnet_time_scale_shift,
)
self.down_blocks.append(down_block)
# mid
if mid_block_type == "UNetMidBlock2DCrossAttn":
self.mid_block = UNetMidBlock2DCrossAttn(
transformer_layers_per_block=transformer_layers_per_block[-1],
in_channels=block_out_channels[-1],
temb_channels=blocks_time_embed_dim,
resnet_eps=norm_eps,
resnet_act_fn=act_fn,
output_scale_factor=mid_block_scale_factor,
resnet_time_scale_shift=resnet_time_scale_shift,
cross_attention_dim=cross_attention_dim[-1],
num_attention_heads=num_attention_heads[-1],
resnet_groups=norm_num_groups,
use_linear_projection=use_linear_projection,
upcast_attention=upcast_attention,
)
else:
raise ValueError(
f"unknown mid_block_type : {mid_block_type}. Should be `UNetMidBlock2DCrossAttn` for AudioLDM2."
)
# count how many layers upsample the images
self.num_upsamplers = 0
# up
reversed_block_out_channels = list(reversed(block_out_channels))
reversed_num_attention_heads = list(reversed(num_attention_heads))
reversed_layers_per_block = list(reversed(layers_per_block))
reversed_cross_attention_dim = list(reversed(cross_attention_dim))
reversed_transformer_layers_per_block = list(reversed(transformer_layers_per_block))
only_cross_attention = list(reversed(only_cross_attention))
output_channel = reversed_block_out_channels[0]
for i, up_block_type in enumerate(up_block_types):
is_final_block = i == len(block_out_channels) - 1
prev_output_channel = output_channel
output_channel = reversed_block_out_channels[i]
input_channel = reversed_block_out_channels[min(i + 1, len(block_out_channels) - 1)]
# add upsample block for all BUT final layer
if not is_final_block:
add_upsample = True
self.num_upsamplers += 1
else:
add_upsample = False
up_block = get_up_block(
up_block_type,
num_layers=reversed_layers_per_block[i] + 1,
transformer_layers_per_block=reversed_transformer_layers_per_block[i],
in_channels=input_channel,
out_channels=output_channel,
prev_output_channel=prev_output_channel,
temb_channels=blocks_time_embed_dim,
add_upsample=add_upsample,
resnet_eps=norm_eps,
resnet_act_fn=act_fn,
resnet_groups=norm_num_groups,
cross_attention_dim=reversed_cross_attention_dim[i],
num_attention_heads=reversed_num_attention_heads[i],
use_linear_projection=use_linear_projection,
only_cross_attention=only_cross_attention[i],
upcast_attention=upcast_attention,
resnet_time_scale_shift=resnet_time_scale_shift,
)
self.up_blocks.append(up_block)
prev_output_channel = output_channel
# out
if norm_num_groups is not None:
self.conv_norm_out = nn.GroupNorm(
num_channels=block_out_channels[0], num_groups=norm_num_groups, eps=norm_eps
)
self.conv_act = get_activation(act_fn)
else:
self.conv_norm_out = None
self.conv_act = None
conv_out_padding = (conv_out_kernel - 1) // 2
self.conv_out = nn.Conv2d(
block_out_channels[0], out_channels, kernel_size=conv_out_kernel, padding=conv_out_padding
)
@property
# Copied from diffusers.models.unet_2d_condition.UNet2DConditionModel.attn_processors
def attn_processors(self) -> Dict[str, AttentionProcessor]:
r"""
Returns:
`dict` of attention processors: A dictionary containing all attention processors used in the model with
indexed by its weight name.
"""
# set recursively
processors = {}
def fn_recursive_add_processors(name: str, module: torch.nn.Module, processors: Dict[str, AttentionProcessor]):
if hasattr(module, "get_processor"):
processors[f"{name}.processor"] = module.get_processor(return_deprecated_lora=True)
for sub_name, child in module.named_children():
fn_recursive_add_processors(f"{name}.{sub_name}", child, processors)
return processors
for name, module in self.named_children():
fn_recursive_add_processors(name, module, processors)
return processors
# Copied from diffusers.models.unet_2d_condition.UNet2DConditionModel.set_attn_processor
def set_attn_processor(
self, processor: Union[AttentionProcessor, Dict[str, AttentionProcessor]], _remove_lora=False
):
r"""
Sets the attention processor to use to compute attention.
Parameters:
processor (`dict` of `AttentionProcessor` or only `AttentionProcessor`):
The instantiated processor class or a dictionary of processor classes that will be set as the processor
for **all** `Attention` layers.
If `processor` is a dict, the key needs to define the path to the corresponding cross attention
processor. This is strongly recommended when setting trainable attention processors.
"""
count = len(self.attn_processors.keys())
if isinstance(processor, dict) and len(processor) != count:
raise ValueError(
f"A dict of processors was passed, but the number of processors {len(processor)} does not match the"
f" number of attention layers: {count}. Please make sure to pass {count} processor classes."
)
def fn_recursive_attn_processor(name: str, module: torch.nn.Module, processor):
if hasattr(module, "set_processor"):
if not isinstance(processor, dict):
module.set_processor(processor, _remove_lora=_remove_lora)
else:
module.set_processor(processor.pop(f"{name}.processor"), _remove_lora=_remove_lora)
for sub_name, child in module.named_children():
fn_recursive_attn_processor(f"{name}.{sub_name}", child, processor)
for name, module in self.named_children():
fn_recursive_attn_processor(name, module, processor)
# Copied from diffusers.models.unet_2d_condition.UNet2DConditionModel.set_default_attn_processor
def set_default_attn_processor(self):
"""
Disables custom attention processors and sets the default attention implementation.
"""
if all(proc.__class__ in ADDED_KV_ATTENTION_PROCESSORS for proc in self.attn_processors.values()):
processor = AttnAddedKVProcessor()
elif all(proc.__class__ in CROSS_ATTENTION_PROCESSORS for proc in self.attn_processors.values()):
processor = AttnProcessor()
else:
raise ValueError(
f"Cannot call `set_default_attn_processor` when attention processors are of type {next(iter(self.attn_processors.values()))}"
)
self.set_attn_processor(processor, _remove_lora=True)
# Copied from diffusers.models.unet_2d_condition.UNet2DConditionModel.set_attention_slice
def set_attention_slice(self, slice_size):
r"""
Enable sliced attention computation.
When this option is enabled, the attention module splits the input tensor in slices to compute attention in
several steps. This is useful for saving some memory in exchange for a small decrease in speed.
Args:
slice_size (`str` or `int` or `list(int)`, *optional*, defaults to `"auto"`):
When `"auto"`, input to the attention heads is halved, so attention is computed in two steps. If
`"max"`, maximum amount of memory is saved by running only one slice at a time. If a number is
provided, uses as many slices as `attention_head_dim // slice_size`. In this case, `attention_head_dim`
must be a multiple of `slice_size`.
"""
sliceable_head_dims = []
def fn_recursive_retrieve_sliceable_dims(module: torch.nn.Module):
if hasattr(module, "set_attention_slice"):
sliceable_head_dims.append(module.sliceable_head_dim)
for child in module.children():
fn_recursive_retrieve_sliceable_dims(child)
# retrieve number of attention layers
for module in self.children():
fn_recursive_retrieve_sliceable_dims(module)
num_sliceable_layers = len(sliceable_head_dims)
if slice_size == "auto":
# half the attention head size is usually a good trade-off between
# speed and memory
slice_size = [dim // 2 for dim in sliceable_head_dims]
elif slice_size == "max":
# make smallest slice possible
slice_size = num_sliceable_layers * [1]
slice_size = num_sliceable_layers * [slice_size] if not isinstance(slice_size, list) else slice_size
if len(slice_size) != len(sliceable_head_dims):
raise ValueError(
f"You have provided {len(slice_size)}, but {self.config} has {len(sliceable_head_dims)} different"
f" attention layers. Make sure to match `len(slice_size)` to be {len(sliceable_head_dims)}."
)
for i in range(len(slice_size)):
size = slice_size[i]
dim = sliceable_head_dims[i]
if size is not None and size > dim:
raise ValueError(f"size {size} has to be smaller or equal to {dim}.")
# Recursively walk through all the children.
# Any children which exposes the set_attention_slice method
# gets the message
def fn_recursive_set_attention_slice(module: torch.nn.Module, slice_size: List[int]):
if hasattr(module, "set_attention_slice"):
module.set_attention_slice(slice_size.pop())
for child in module.children():
fn_recursive_set_attention_slice(child, slice_size)
reversed_slice_size = list(reversed(slice_size))
for module in self.children():
fn_recursive_set_attention_slice(module, reversed_slice_size)
# Copied from diffusers.models.unet_2d_condition.UNet2DConditionModel._set_gradient_checkpointing
def _set_gradient_checkpointing(self, module, value=False):
if hasattr(module, "gradient_checkpointing"):
module.gradient_checkpointing = value
def forward(
self,
sample: torch.FloatTensor,
timestep: Union[torch.Tensor, float, int],
encoder_hidden_states: torch.Tensor,
class_labels: Optional[torch.Tensor] = None,
timestep_cond: Optional[torch.Tensor] = None,
attention_mask: Optional[torch.Tensor] = None,
cross_attention_kwargs: Optional[Dict[str, Any]] = None,
encoder_attention_mask: Optional[torch.Tensor] = None,
return_dict: bool = True,
encoder_hidden_states_1: Optional[torch.Tensor] = None,
encoder_attention_mask_1: Optional[torch.Tensor] = None,
) -> Union[UNet2DConditionOutput, Tuple]:
r"""
The [`AudioLDM2UNet2DConditionModel`] forward method.
Args:
sample (`torch.FloatTensor`):
The noisy input tensor with the following shape `(batch, channel, height, width)`.
timestep (`torch.FloatTensor` or `float` or `int`): The number of timesteps to denoise an input.
encoder_hidden_states (`torch.FloatTensor`):
The encoder hidden states with shape `(batch, sequence_length, feature_dim)`.
encoder_attention_mask (`torch.Tensor`):
A cross-attention mask of shape `(batch, sequence_length)` is applied to `encoder_hidden_states`. If
`True` the mask is kept, otherwise if `False` it is discarded. Mask will be converted into a bias,
which adds large negative values to the attention scores corresponding to "discard" tokens.
return_dict (`bool`, *optional*, defaults to `True`):
Whether or not to return a [`~models.unet_2d_condition.UNet2DConditionOutput`] instead of a plain
tuple.
cross_attention_kwargs (`dict`, *optional*):
A kwargs dictionary that if specified is passed along to the [`AttnProcessor`].
encoder_hidden_states_1 (`torch.FloatTensor`, *optional*):
A second set of encoder hidden states with shape `(batch, sequence_length_2, feature_dim_2)`. Can be
used to condition the model on a different set of embeddings to `encoder_hidden_states`.
encoder_attention_mask_1 (`torch.Tensor`, *optional*):
A cross-attention mask of shape `(batch, sequence_length_2)` is applied to `encoder_hidden_states_1`.
If `True` the mask is kept, otherwise if `False` it is discarded. Mask will be converted into a bias,
which adds large negative values to the attention scores corresponding to "discard" tokens.
Returns:
[`~models.unet_2d_condition.UNet2DConditionOutput`] or `tuple`:
If `return_dict` is True, an [`~models.unet_2d_condition.UNet2DConditionOutput`] is returned, otherwise
a `tuple` is returned where the first element is the sample tensor.
"""
# By default samples have to be AT least a multiple of the overall upsampling factor.
# The overall upsampling factor is equal to 2 ** (# num of upsampling layers).
# However, the upsampling interpolation output size can be forced to fit any upsampling size
# on the fly if necessary.
default_overall_up_factor = 2**self.num_upsamplers
# upsample size should be forwarded when sample is not a multiple of `default_overall_up_factor`
forward_upsample_size = False
upsample_size = None
if any(s % default_overall_up_factor != 0 for s in sample.shape[-2:]):
logger.info("Forward upsample size to force interpolation output size.")
forward_upsample_size = True
# ensure attention_mask is a bias, and give it a singleton query_tokens dimension
# expects mask of shape:
# [batch, key_tokens]
# adds singleton query_tokens dimension:
# [batch, 1, key_tokens]
# this helps to broadcast it as a bias over attention scores, which will be in one of the following shapes:
# [batch, heads, query_tokens, key_tokens] (e.g. torch sdp attn)
# [batch * heads, query_tokens, key_tokens] (e.g. xformers or classic attn)
if attention_mask is not None:
# assume that mask is expressed as:
# (1 = keep, 0 = discard)
# convert mask into a bias that can be added to attention scores:
# (keep = +0, discard = -10000.0)
attention_mask = (1 - attention_mask.to(sample.dtype)) * -10000.0
attention_mask = attention_mask.unsqueeze(1)
# convert encoder_attention_mask to a bias the same way we do for attention_mask
if encoder_attention_mask is not None:
encoder_attention_mask = (1 - encoder_attention_mask.to(sample.dtype)) * -10000.0
encoder_attention_mask = encoder_attention_mask.unsqueeze(1)
if encoder_attention_mask_1 is not None:
encoder_attention_mask_1 = (1 - encoder_attention_mask_1.to(sample.dtype)) * -10000.0
encoder_attention_mask_1 = encoder_attention_mask_1.unsqueeze(1)
# 1. time
timesteps = timestep
if not torch.is_tensor(timesteps):
# TODO: this requires sync between CPU and GPU. So try to pass timesteps as tensors if you can
# This would be a good case for the `match` statement (Python 3.10+)
is_mps = sample.device.type == "mps"
if isinstance(timestep, float):
dtype = torch.float32 if is_mps else torch.float64
else:
dtype = torch.int32 if is_mps else torch.int64
timesteps = torch.tensor([timesteps], dtype=dtype, device=sample.device)
elif len(timesteps.shape) == 0:
timesteps = timesteps[None].to(sample.device)
# broadcast to batch dimension in a way that's compatible with ONNX/Core ML
timesteps = timesteps.expand(sample.shape[0])
t_emb = self.time_proj(timesteps)
# `Timesteps` does not contain any weights and will always return f32 tensors
# but time_embedding might actually be running in fp16. so we need to cast here.
# there might be better ways to encapsulate this.
t_emb = t_emb.to(dtype=sample.dtype)
emb = self.time_embedding(t_emb, timestep_cond)
aug_emb = None
if self.class_embedding is not None:
if class_labels is None:
raise ValueError("class_labels should be provided when num_class_embeds > 0")
if self.config.class_embed_type == "timestep":
class_labels = self.time_proj(class_labels)
# `Timesteps` does not contain any weights and will always return f32 tensors
# there might be better ways to encapsulate this.
class_labels = class_labels.to(dtype=sample.dtype)
class_emb = self.class_embedding(class_labels).to(dtype=sample.dtype)
if self.config.class_embeddings_concat:
emb = torch.cat([emb, class_emb], dim=-1)
else:
emb = emb + class_emb
emb = emb + aug_emb if aug_emb is not None else emb
if self.time_embed_act is not None:
emb = self.time_embed_act(emb)
# 2. pre-process
sample = self.conv_in(sample)
# 3. down
down_block_res_samples = (sample,)
for downsample_block in self.down_blocks:
if hasattr(downsample_block, "has_cross_attention") and downsample_block.has_cross_attention:
sample, res_samples = downsample_block(
hidden_states=sample,
temb=emb,
encoder_hidden_states=encoder_hidden_states,
attention_mask=attention_mask,
cross_attention_kwargs=cross_attention_kwargs,
encoder_attention_mask=encoder_attention_mask,
encoder_hidden_states_1=encoder_hidden_states_1,
encoder_attention_mask_1=encoder_attention_mask_1,
)
else:
sample, res_samples = downsample_block(hidden_states=sample, temb=emb)
down_block_res_samples += res_samples
# 4. mid
if self.mid_block is not None:
sample = self.mid_block(
sample,
emb,
encoder_hidden_states=encoder_hidden_states,
attention_mask=attention_mask,
cross_attention_kwargs=cross_attention_kwargs,
encoder_attention_mask=encoder_attention_mask,
encoder_hidden_states_1=encoder_hidden_states_1,
encoder_attention_mask_1=encoder_attention_mask_1,
)
# 5. up
for i, upsample_block in enumerate(self.up_blocks):
is_final_block = i == len(self.up_blocks) - 1
res_samples = down_block_res_samples[-len(upsample_block.resnets) :]
down_block_res_samples = down_block_res_samples[: -len(upsample_block.resnets)]
# if we have not reached the final block and need to forward the
# upsample size, we do it here
if not is_final_block and forward_upsample_size:
upsample_size = down_block_res_samples[-1].shape[2:]
if hasattr(upsample_block, "has_cross_attention") and upsample_block.has_cross_attention:
sample = upsample_block(
hidden_states=sample,
temb=emb,
res_hidden_states_tuple=res_samples,
encoder_hidden_states=encoder_hidden_states,
cross_attention_kwargs=cross_attention_kwargs,
upsample_size=upsample_size,
attention_mask=attention_mask,
encoder_attention_mask=encoder_attention_mask,
encoder_hidden_states_1=encoder_hidden_states_1,
encoder_attention_mask_1=encoder_attention_mask_1,
)
else:
sample = upsample_block(
hidden_states=sample, temb=emb, res_hidden_states_tuple=res_samples, upsample_size=upsample_size
)
# 6. post-process
if self.conv_norm_out:
sample = self.conv_norm_out(sample)
sample = self.conv_act(sample)
sample = self.conv_out(sample)
if not return_dict:
return (sample,)
return UNet2DConditionOutput(sample=sample)
def get_down_block(
down_block_type,
num_layers,
in_channels,
out_channels,
temb_channels,
add_downsample,
resnet_eps,
resnet_act_fn,
transformer_layers_per_block=1,
num_attention_heads=None,
resnet_groups=None,
cross_attention_dim=None,
downsample_padding=None,
use_linear_projection=False,
only_cross_attention=False,
upcast_attention=False,
resnet_time_scale_shift="default",
):
down_block_type = down_block_type[7:] if down_block_type.startswith("UNetRes") else down_block_type
if down_block_type == "DownBlock2D":
return DownBlock2D(
num_layers=num_layers,
in_channels=in_channels,
out_channels=out_channels,
temb_channels=temb_channels,
add_downsample=add_downsample,
resnet_eps=resnet_eps,
resnet_act_fn=resnet_act_fn,
resnet_groups=resnet_groups,
downsample_padding=downsample_padding,
resnet_time_scale_shift=resnet_time_scale_shift,
)
elif down_block_type == "CrossAttnDownBlock2D":
if cross_attention_dim is None:
raise ValueError("cross_attention_dim must be specified for CrossAttnDownBlock2D")
return CrossAttnDownBlock2D(
num_layers=num_layers,
transformer_layers_per_block=transformer_layers_per_block,
in_channels=in_channels,
out_channels=out_channels,
temb_channels=temb_channels,
add_downsample=add_downsample,
resnet_eps=resnet_eps,
resnet_act_fn=resnet_act_fn,
resnet_groups=resnet_groups,
downsample_padding=downsample_padding,
cross_attention_dim=cross_attention_dim,
num_attention_heads=num_attention_heads,
use_linear_projection=use_linear_projection,
only_cross_attention=only_cross_attention,
upcast_attention=upcast_attention,
resnet_time_scale_shift=resnet_time_scale_shift,
)
raise ValueError(f"{down_block_type} does not exist.")
def get_up_block(
up_block_type,
num_layers,
in_channels,
out_channels,
prev_output_channel,
temb_channels,
add_upsample,
resnet_eps,
resnet_act_fn,
transformer_layers_per_block=1,
num_attention_heads=None,
resnet_groups=None,
cross_attention_dim=None,
use_linear_projection=False,
only_cross_attention=False,
upcast_attention=False,
resnet_time_scale_shift="default",
):
up_block_type = up_block_type[7:] if up_block_type.startswith("UNetRes") else up_block_type
if up_block_type == "UpBlock2D":
return UpBlock2D(
num_layers=num_layers,
in_channels=in_channels,
out_channels=out_channels,
prev_output_channel=prev_output_channel,
temb_channels=temb_channels,
add_upsample=add_upsample,
resnet_eps=resnet_eps,
resnet_act_fn=resnet_act_fn,
resnet_groups=resnet_groups,
resnet_time_scale_shift=resnet_time_scale_shift,
)
elif up_block_type == "CrossAttnUpBlock2D":
if cross_attention_dim is None:
raise ValueError("cross_attention_dim must be specified for CrossAttnUpBlock2D")
return CrossAttnUpBlock2D(
num_layers=num_layers,
transformer_layers_per_block=transformer_layers_per_block,
in_channels=in_channels,
out_channels=out_channels,
prev_output_channel=prev_output_channel,
temb_channels=temb_channels,
add_upsample=add_upsample,
resnet_eps=resnet_eps,
resnet_act_fn=resnet_act_fn,
resnet_groups=resnet_groups,
cross_attention_dim=cross_attention_dim,
num_attention_heads=num_attention_heads,
use_linear_projection=use_linear_projection,
only_cross_attention=only_cross_attention,
upcast_attention=upcast_attention,
resnet_time_scale_shift=resnet_time_scale_shift,
)
raise ValueError(f"{up_block_type} does not exist.")
class CrossAttnDownBlock2D(nn.Module):
def __init__(
self,
in_channels: int,
out_channels: int,
temb_channels: int,
dropout: float = 0.0,
num_layers: int = 1,
transformer_layers_per_block: int = 1,
resnet_eps: float = 1e-6,
resnet_time_scale_shift: str = "default",
resnet_act_fn: str = "swish",
resnet_groups: int = 32,
resnet_pre_norm: bool = True,
num_attention_heads=1,
cross_attention_dim=1280,
output_scale_factor=1.0,
downsample_padding=1,
add_downsample=True,
use_linear_projection=False,
only_cross_attention=False,
upcast_attention=False,
):
super().__init__()
resnets = []
attentions = []
self.has_cross_attention = True
self.num_attention_heads = num_attention_heads
if isinstance(cross_attention_dim, int):
cross_attention_dim = (cross_attention_dim,)
if isinstance(cross_attention_dim, (list, tuple)) and len(cross_attention_dim) > 4:
raise ValueError(
"Only up to 4 cross-attention layers are supported. Ensure that the length of cross-attention "
f"dims is less than or equal to 4. Got cross-attention dims {cross_attention_dim} of length {len(cross_attention_dim)}"
)
self.cross_attention_dim = cross_attention_dim
for i in range(num_layers):
in_channels = in_channels if i == 0 else out_channels
resnets.append(
ResnetBlock2D(
in_channels=in_channels,
out_channels=out_channels,
temb_channels=temb_channels,
eps=resnet_eps,
groups=resnet_groups,
dropout=dropout,
time_embedding_norm=resnet_time_scale_shift,
non_linearity=resnet_act_fn,
output_scale_factor=output_scale_factor,
pre_norm=resnet_pre_norm,
)
)
for j in range(len(cross_attention_dim)):
attentions.append(
Transformer2DModel(
num_attention_heads,
out_channels // num_attention_heads,
in_channels=out_channels,
num_layers=transformer_layers_per_block,
cross_attention_dim=cross_attention_dim[j],
norm_num_groups=resnet_groups,
use_linear_projection=use_linear_projection,
only_cross_attention=only_cross_attention,
upcast_attention=upcast_attention,
double_self_attention=True if cross_attention_dim[j] is None else False,
)
)
self.attentions = nn.ModuleList(attentions)
self.resnets = nn.ModuleList(resnets)
if add_downsample:
self.downsamplers = nn.ModuleList(
[
Downsample2D(
out_channels, use_conv=True, out_channels=out_channels, padding=downsample_padding, name="op"
)
]
)
else:
self.downsamplers = None
self.gradient_checkpointing = False
def forward(
self,
hidden_states: torch.FloatTensor,
temb: Optional[torch.FloatTensor] = None,
encoder_hidden_states: Optional[torch.FloatTensor] = None,
attention_mask: Optional[torch.FloatTensor] = None,
cross_attention_kwargs: Optional[Dict[str, Any]] = None,
encoder_attention_mask: Optional[torch.FloatTensor] = None,
encoder_hidden_states_1: Optional[torch.FloatTensor] = None,
encoder_attention_mask_1: Optional[torch.FloatTensor] = None,
):
output_states = ()
num_layers = len(self.resnets)
num_attention_per_layer = len(self.attentions) // num_layers
encoder_hidden_states_1 = (
encoder_hidden_states_1 if encoder_hidden_states_1 is not None else encoder_hidden_states
)
encoder_attention_mask_1 = (
encoder_attention_mask_1 if encoder_hidden_states_1 is not None else encoder_attention_mask
)
for i in range(num_layers):
if self.training and self.gradient_checkpointing:
def create_custom_forward(module, return_dict=None):
def custom_forward(*inputs):
if return_dict is not None:
return module(*inputs, return_dict=return_dict)
else:
return module(*inputs)
return custom_forward
ckpt_kwargs: Dict[str, Any] = {"use_reentrant": False} if is_torch_version(">=", "1.11.0") else {}
hidden_states = torch.utils.checkpoint.checkpoint(
create_custom_forward(self.resnets[i]),
hidden_states,
temb,
**ckpt_kwargs,
)
for idx, cross_attention_dim in enumerate(self.cross_attention_dim):
if cross_attention_dim is not None and idx <= 1:
forward_encoder_hidden_states = encoder_hidden_states
forward_encoder_attention_mask = encoder_attention_mask
elif cross_attention_dim is not None and idx > 1:
forward_encoder_hidden_states = encoder_hidden_states_1
forward_encoder_attention_mask = encoder_attention_mask_1
else:
forward_encoder_hidden_states = None
forward_encoder_attention_mask = None
hidden_states = torch.utils.checkpoint.checkpoint(
create_custom_forward(self.attentions[i * num_attention_per_layer + idx], return_dict=False),
hidden_states,
forward_encoder_hidden_states,
None, # timestep
None, # class_labels
cross_attention_kwargs,
attention_mask,
forward_encoder_attention_mask,
**ckpt_kwargs,
)[0]
else:
hidden_states = self.resnets[i](hidden_states, temb)
for idx, cross_attention_dim in enumerate(self.cross_attention_dim):
if cross_attention_dim is not None and idx <= 1:
forward_encoder_hidden_states = encoder_hidden_states
forward_encoder_attention_mask = encoder_attention_mask
elif cross_attention_dim is not None and idx > 1:
forward_encoder_hidden_states = encoder_hidden_states_1
forward_encoder_attention_mask = encoder_attention_mask_1
else:
forward_encoder_hidden_states = None
forward_encoder_attention_mask = None
hidden_states = self.attentions[i * num_attention_per_layer + idx](
hidden_states,
attention_mask=attention_mask,
encoder_hidden_states=forward_encoder_hidden_states,
encoder_attention_mask=forward_encoder_attention_mask,
return_dict=False,
)[0]
output_states = output_states + (hidden_states,)
if self.downsamplers is not None:
for downsampler in self.downsamplers:
hidden_states = downsampler(hidden_states)
output_states = output_states + (hidden_states,)
return hidden_states, output_states
class UNetMidBlock2DCrossAttn(nn.Module):
def __init__(
self,
in_channels: int,
temb_channels: int,
dropout: float = 0.0,
num_layers: int = 1,
transformer_layers_per_block: int = 1,
resnet_eps: float = 1e-6,
resnet_time_scale_shift: str = "default",
resnet_act_fn: str = "swish",
resnet_groups: int = 32,
resnet_pre_norm: bool = True,
num_attention_heads=1,
output_scale_factor=1.0,
cross_attention_dim=1280,
use_linear_projection=False,
upcast_attention=False,
):
super().__init__()
self.has_cross_attention = True
self.num_attention_heads = num_attention_heads
resnet_groups = resnet_groups if resnet_groups is not None else min(in_channels // 4, 32)
if isinstance(cross_attention_dim, int):
cross_attention_dim = (cross_attention_dim,)
if isinstance(cross_attention_dim, (list, tuple)) and len(cross_attention_dim) > 4:
raise ValueError(
"Only up to 4 cross-attention layers are supported. Ensure that the length of cross-attention "
f"dims is less than or equal to 4. Got cross-attention dims {cross_attention_dim} of length {len(cross_attention_dim)}"
)
self.cross_attention_dim = cross_attention_dim
# there is always at least one resnet
resnets = [
ResnetBlock2D(
in_channels=in_channels,
out_channels=in_channels,
temb_channels=temb_channels,
eps=resnet_eps,
groups=resnet_groups,
dropout=dropout,
time_embedding_norm=resnet_time_scale_shift,
non_linearity=resnet_act_fn,
output_scale_factor=output_scale_factor,
pre_norm=resnet_pre_norm,
)
]
attentions = []
for i in range(num_layers):
for j in range(len(cross_attention_dim)):
attentions.append(
Transformer2DModel(
num_attention_heads,
in_channels // num_attention_heads,
in_channels=in_channels,
num_layers=transformer_layers_per_block,
cross_attention_dim=cross_attention_dim[j],
norm_num_groups=resnet_groups,
use_linear_projection=use_linear_projection,
upcast_attention=upcast_attention,
double_self_attention=True if cross_attention_dim[j] is None else False,
)
)
resnets.append(
ResnetBlock2D(
in_channels=in_channels,
out_channels=in_channels,
temb_channels=temb_channels,
eps=resnet_eps,
groups=resnet_groups,
dropout=dropout,
time_embedding_norm=resnet_time_scale_shift,
non_linearity=resnet_act_fn,
output_scale_factor=output_scale_factor,
pre_norm=resnet_pre_norm,
)
)
self.attentions = nn.ModuleList(attentions)
self.resnets = nn.ModuleList(resnets)
self.gradient_checkpointing = False
def forward(
self,
hidden_states: torch.FloatTensor,
temb: Optional[torch.FloatTensor] = None,
encoder_hidden_states: Optional[torch.FloatTensor] = None,
attention_mask: Optional[torch.FloatTensor] = None,
cross_attention_kwargs: Optional[Dict[str, Any]] = None,
encoder_attention_mask: Optional[torch.FloatTensor] = None,
encoder_hidden_states_1: Optional[torch.FloatTensor] = None,
encoder_attention_mask_1: Optional[torch.FloatTensor] = None,
) -> torch.FloatTensor:
hidden_states = self.resnets[0](hidden_states, temb)
num_attention_per_layer = len(self.attentions) // (len(self.resnets) - 1)
encoder_hidden_states_1 = (
encoder_hidden_states_1 if encoder_hidden_states_1 is not None else encoder_hidden_states
)
encoder_attention_mask_1 = (
encoder_attention_mask_1 if encoder_hidden_states_1 is not None else encoder_attention_mask
)
for i in range(len(self.resnets[1:])):
if self.training and self.gradient_checkpointing:
def create_custom_forward(module, return_dict=None):
def custom_forward(*inputs):
if return_dict is not None:
return module(*inputs, return_dict=return_dict)
else:
return module(*inputs)
return custom_forward
ckpt_kwargs: Dict[str, Any] = {"use_reentrant": False} if is_torch_version(">=", "1.11.0") else {}
for idx, cross_attention_dim in enumerate(self.cross_attention_dim):
if cross_attention_dim is not None and idx <= 1:
forward_encoder_hidden_states = encoder_hidden_states
forward_encoder_attention_mask = encoder_attention_mask
elif cross_attention_dim is not None and idx > 1:
forward_encoder_hidden_states = encoder_hidden_states_1
forward_encoder_attention_mask = encoder_attention_mask_1
else:
forward_encoder_hidden_states = None
forward_encoder_attention_mask = None
hidden_states = torch.utils.checkpoint.checkpoint(
create_custom_forward(self.attentions[i * num_attention_per_layer + idx], return_dict=False),
hidden_states,
forward_encoder_hidden_states,
None, # timestep
None, # class_labels
cross_attention_kwargs,
attention_mask,
forward_encoder_attention_mask,
**ckpt_kwargs,
)[0]
hidden_states = torch.utils.checkpoint.checkpoint(
create_custom_forward(self.resnets[i + 1]),
hidden_states,
temb,
**ckpt_kwargs,
)
else:
for idx, cross_attention_dim in enumerate(self.cross_attention_dim):
if cross_attention_dim is not None and idx <= 1:
forward_encoder_hidden_states = encoder_hidden_states
forward_encoder_attention_mask = encoder_attention_mask
elif cross_attention_dim is not None and idx > 1:
forward_encoder_hidden_states = encoder_hidden_states_1
forward_encoder_attention_mask = encoder_attention_mask_1
else:
forward_encoder_hidden_states = None
forward_encoder_attention_mask = None
hidden_states = self.attentions[i * num_attention_per_layer + idx](
hidden_states,
attention_mask=attention_mask,
encoder_hidden_states=forward_encoder_hidden_states,
encoder_attention_mask=forward_encoder_attention_mask,
return_dict=False,
)[0]
hidden_states = self.resnets[i + 1](hidden_states, temb)
return hidden_states
class CrossAttnUpBlock2D(nn.Module):
def __init__(
self,
in_channels: int,
out_channels: int,
prev_output_channel: int,
temb_channels: int,
dropout: float = 0.0,
num_layers: int = 1,
transformer_layers_per_block: int = 1,
resnet_eps: float = 1e-6,
resnet_time_scale_shift: str = "default",
resnet_act_fn: str = "swish",
resnet_groups: int = 32,
resnet_pre_norm: bool = True,
num_attention_heads=1,
cross_attention_dim=1280,
output_scale_factor=1.0,
add_upsample=True,
use_linear_projection=False,
only_cross_attention=False,
upcast_attention=False,
):
super().__init__()
resnets = []
attentions = []
self.has_cross_attention = True
self.num_attention_heads = num_attention_heads
if isinstance(cross_attention_dim, int):
cross_attention_dim = (cross_attention_dim,)
if isinstance(cross_attention_dim, (list, tuple)) and len(cross_attention_dim) > 4:
raise ValueError(
"Only up to 4 cross-attention layers are supported. Ensure that the length of cross-attention "
f"dims is less than or equal to 4. Got cross-attention dims {cross_attention_dim} of length {len(cross_attention_dim)}"
)
self.cross_attention_dim = cross_attention_dim
for i in range(num_layers):
res_skip_channels = in_channels if (i == num_layers - 1) else out_channels
resnet_in_channels = prev_output_channel if i == 0 else out_channels
resnets.append(
ResnetBlock2D(
in_channels=resnet_in_channels + res_skip_channels,
out_channels=out_channels,
temb_channels=temb_channels,
eps=resnet_eps,
groups=resnet_groups,
dropout=dropout,
time_embedding_norm=resnet_time_scale_shift,
non_linearity=resnet_act_fn,
output_scale_factor=output_scale_factor,
pre_norm=resnet_pre_norm,
)
)
for j in range(len(cross_attention_dim)):
attentions.append(
Transformer2DModel(
num_attention_heads,
out_channels // num_attention_heads,
in_channels=out_channels,
num_layers=transformer_layers_per_block,
cross_attention_dim=cross_attention_dim[j],
norm_num_groups=resnet_groups,
use_linear_projection=use_linear_projection,
only_cross_attention=only_cross_attention,
upcast_attention=upcast_attention,
double_self_attention=True if cross_attention_dim[j] is None else False,
)
)
self.attentions = nn.ModuleList(attentions)
self.resnets = nn.ModuleList(resnets)
if add_upsample:
self.upsamplers = nn.ModuleList([Upsample2D(out_channels, use_conv=True, out_channels=out_channels)])
else:
self.upsamplers = None
self.gradient_checkpointing = False
def forward(
self,
hidden_states: torch.FloatTensor,
res_hidden_states_tuple: Tuple[torch.FloatTensor, ...],
temb: Optional[torch.FloatTensor] = None,
encoder_hidden_states: Optional[torch.FloatTensor] = None,
cross_attention_kwargs: Optional[Dict[str, Any]] = None,
upsample_size: Optional[int] = None,
attention_mask: Optional[torch.FloatTensor] = None,
encoder_attention_mask: Optional[torch.FloatTensor] = None,
encoder_hidden_states_1: Optional[torch.FloatTensor] = None,
encoder_attention_mask_1: Optional[torch.FloatTensor] = None,
):
num_layers = len(self.resnets)
num_attention_per_layer = len(self.attentions) // num_layers
encoder_hidden_states_1 = (
encoder_hidden_states_1 if encoder_hidden_states_1 is not None else encoder_hidden_states
)
encoder_attention_mask_1 = (
encoder_attention_mask_1 if encoder_hidden_states_1 is not None else encoder_attention_mask
)
for i in range(num_layers):
# pop res hidden states
res_hidden_states = res_hidden_states_tuple[-1]
res_hidden_states_tuple = res_hidden_states_tuple[:-1]
hidden_states = torch.cat([hidden_states, res_hidden_states], dim=1)
if self.training and self.gradient_checkpointing:
def create_custom_forward(module, return_dict=None):
def custom_forward(*inputs):
if return_dict is not None:
return module(*inputs, return_dict=return_dict)
else:
return module(*inputs)
return custom_forward
ckpt_kwargs: Dict[str, Any] = {"use_reentrant": False} if is_torch_version(">=", "1.11.0") else {}
hidden_states = torch.utils.checkpoint.checkpoint(
create_custom_forward(self.resnets[i]),
hidden_states,
temb,
**ckpt_kwargs,
)
for idx, cross_attention_dim in enumerate(self.cross_attention_dim):
if cross_attention_dim is not None and idx <= 1:
forward_encoder_hidden_states = encoder_hidden_states
forward_encoder_attention_mask = encoder_attention_mask
elif cross_attention_dim is not None and idx > 1:
forward_encoder_hidden_states = encoder_hidden_states_1
forward_encoder_attention_mask = encoder_attention_mask_1
else:
forward_encoder_hidden_states = None
forward_encoder_attention_mask = None
hidden_states = torch.utils.checkpoint.checkpoint(
create_custom_forward(self.attentions[i * num_attention_per_layer + idx], return_dict=False),
hidden_states,
forward_encoder_hidden_states,
None, # timestep
None, # class_labels
cross_attention_kwargs,
attention_mask,
forward_encoder_attention_mask,
**ckpt_kwargs,
)[0]
else:
hidden_states = self.resnets[i](hidden_states, temb)
for idx, cross_attention_dim in enumerate(self.cross_attention_dim):
if cross_attention_dim is not None and idx <= 1:
forward_encoder_hidden_states = encoder_hidden_states
forward_encoder_attention_mask = encoder_attention_mask
elif cross_attention_dim is not None and idx > 1:
forward_encoder_hidden_states = encoder_hidden_states_1
forward_encoder_attention_mask = encoder_attention_mask_1
else:
forward_encoder_hidden_states = None
forward_encoder_attention_mask = None
hidden_states = self.attentions[i * num_attention_per_layer + idx](
hidden_states,
attention_mask=attention_mask,
encoder_hidden_states=forward_encoder_hidden_states,
encoder_attention_mask=forward_encoder_attention_mask,
return_dict=False,
)[0]
if self.upsamplers is not None:
for upsampler in self.upsamplers:
hidden_states = upsampler(hidden_states, upsample_size)
return hidden_states