636 lines
28 KiB
Python
636 lines
28 KiB
Python
# Copyright 2023 Alibaba DAMO-VILAB and The HuggingFace Team. All rights reserved.
|
|
# Copyright 2023 The ModelScope Team.
|
|
#
|
|
# 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.utils import BaseOutput, logging
|
|
from diffusers.models.embeddings import TimestepEmbedding, Timesteps
|
|
from diffusers.models.modeling_utils import ModelMixin
|
|
# from diffusers.models.transformer_temporal import TransformerTemporalModel
|
|
from .transformer_temporal import TransformerTemporalModel
|
|
from .unet_3d_blocks import (
|
|
CrossAttnDownBlock3D,
|
|
CrossAttnUpBlock3D,
|
|
DownBlock3D,
|
|
UNetMidBlock3DCrossAttn,
|
|
UpBlock3D,
|
|
get_down_block,
|
|
get_up_block,
|
|
transformer_g_c
|
|
)
|
|
from .conditioning import ConditionalModel
|
|
from einops import rearrange
|
|
from ....layers.conv_channel_extension import Conv2D_ExtendedChannels
|
|
logger = logging.get_logger(__name__) # pylint: disable=invalid-name
|
|
|
|
|
|
@dataclass
|
|
class UNet3DConditionOutput(BaseOutput):
|
|
"""
|
|
Args:
|
|
sample (`torch.FloatTensor` of shape `(batch_size, num_frames, num_channels, height, width)`):
|
|
Hidden states conditioned on `encoder_hidden_states` input. Output of last layer of model.
|
|
"""
|
|
|
|
sample: torch.FloatTensor
|
|
|
|
|
|
class UNet3DConditionModel(ModelMixin, ConfigMixin):
|
|
r"""
|
|
UNet3DConditionModel is a conditional 2D UNet model that takes in a noisy sample, conditional state, and a timestep
|
|
and returns sample shaped output.
|
|
|
|
This model inherits from [`ModelMixin`]. Check the superclass documentation for the generic methods the library
|
|
implements for all the models (such as downloading or saving, etc.)
|
|
|
|
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): The number of channels in the input sample.
|
|
out_channels (`int`, *optional*, defaults to 4): The number of channels in the output.
|
|
down_block_types (`Tuple[str]`, *optional*, defaults to `("CrossAttnDownBlock2D", "CrossAttnDownBlock2D", "CrossAttnDownBlock2D", "DownBlock2D")`):
|
|
The tuple of downsample blocks to use.
|
|
up_block_types (`Tuple[str]`, *optional*, defaults to `("UpBlock2D", "CrossAttnUpBlock2D", "CrossAttnUpBlock2D", "CrossAttnUpBlock2D",)`):
|
|
The tuple of upsample blocks to use.
|
|
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`, it will skip the normalization and activation layers in post-processing
|
|
norm_eps (`float`, *optional*, defaults to 1e-5): The epsilon to use for the normalization.
|
|
cross_attention_dim (`int`, *optional*, defaults to 1280): The dimension of the cross attention features.
|
|
attention_head_dim (`int`, *optional*, defaults to 8): The dimension of the attention heads.
|
|
"""
|
|
|
|
_supports_gradient_checkpointing = True
|
|
|
|
@register_to_config
|
|
def __init__(
|
|
self,
|
|
sample_size: Optional[int] = None,
|
|
in_channels: int = 4,
|
|
out_channels: int = 4,
|
|
down_block_types: Tuple[str] = (
|
|
"CrossAttnDownBlock3D",
|
|
"CrossAttnDownBlock3D",
|
|
"CrossAttnDownBlock3D",
|
|
"DownBlock3D",
|
|
),
|
|
up_block_types: Tuple[str] = (
|
|
"UpBlock3D", "CrossAttnUpBlock3D", "CrossAttnUpBlock3D", "CrossAttnUpBlock3D"),
|
|
block_out_channels: Tuple[int] = (320, 640, 1280, 1280),
|
|
layers_per_block: 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: int = 1024,
|
|
attention_head_dim: Union[int, Tuple[int]] = 64,
|
|
merging_mode: str = "addition",
|
|
use_image_embedding: bool = False,
|
|
use_fps_conditioning: bool = False,
|
|
unet_params=None,
|
|
):
|
|
super().__init__()
|
|
channel_expansion = unet_params.use_of
|
|
self.concat = unet_params.concat
|
|
self.use_image_tokens = unet_params.use_image_tokens_main
|
|
self.image_encoder_name = type(unet_params.image_encoder).__name__
|
|
self.use_image_embedding = use_image_embedding
|
|
self.sample_size = sample_size
|
|
self.gradient_checkpointing = False
|
|
# 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(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}."
|
|
)
|
|
|
|
# input
|
|
conv_in_kernel = 3
|
|
conv_out_kernel = 3
|
|
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
|
|
)
|
|
'''
|
|
self.conv_in = Conv2D_ExtendedChannels(
|
|
in_channels, block_out_channels[0], kernel_size=conv_in_kernel, padding=conv_in_padding, in_channel_extension=5 if self.concat else 0,
|
|
)
|
|
|
|
# time
|
|
time_embed_dim = block_out_channels[0] * 4
|
|
self.time_proj = Timesteps(block_out_channels[0], True, 0)
|
|
timestep_input_dim = block_out_channels[0]
|
|
|
|
self.time_embedding = TimestepEmbedding(
|
|
timestep_input_dim,
|
|
time_embed_dim,
|
|
act_fn=act_fn,
|
|
)
|
|
self.use_fps_conditioning = use_fps_conditioning
|
|
if use_fps_conditioning:
|
|
fps_embed_dim = block_out_channels[0] * 4
|
|
fps_input_dim = block_out_channels[0]
|
|
self.fps_embedding = TimestepEmbedding(
|
|
fps_input_dim, fps_embed_dim, act_fn=act_fn)
|
|
self.fps_proj = Timesteps(block_out_channels[0], True, 0)
|
|
|
|
self.transformer_in = TransformerTemporalModel(
|
|
num_attention_heads=8,
|
|
attention_head_dim=attention_head_dim,
|
|
in_channels=block_out_channels[0],
|
|
num_layers=1,
|
|
)
|
|
|
|
# class embedding
|
|
self.down_blocks = nn.ModuleList([])
|
|
self.up_blocks = nn.ModuleList([])
|
|
|
|
self.merging_mode = merging_mode
|
|
print("self.merging_mode", self.merging_mode)
|
|
if self.merging_mode.startswith("attention"):
|
|
self.cross_attention_merger_down_blocks = nn.ModuleList([])
|
|
self.cross_attention_merger_mid_block = nn.ModuleList([])
|
|
if isinstance(attention_head_dim, int):
|
|
attention_head_dim = (attention_head_dim,) * len(down_block_types)
|
|
|
|
# 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,
|
|
in_channels=input_channel,
|
|
out_channels=output_channel,
|
|
temb_channels=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,
|
|
attn_num_head_channels=attention_head_dim[i],
|
|
downsample_padding=downsample_padding,
|
|
dual_cross_attention=False,
|
|
use_image_embedding=use_image_embedding,
|
|
unet_params=unet_params,
|
|
)
|
|
self.down_blocks.append(down_block)
|
|
|
|
if self.merging_mode.startswith("attention"):
|
|
for idx in range(3):
|
|
self.cross_attention_merger_down_blocks.append(ConditionalModel(
|
|
input_channels=input_channel if idx == 0 else output_channel, conditional_model=self.merging_mode.split("attention_")[1]))
|
|
|
|
# mid
|
|
self.mid_block = UNetMidBlock3DCrossAttn(
|
|
in_channels=block_out_channels[-1],
|
|
temb_channels=time_embed_dim,
|
|
resnet_eps=norm_eps,
|
|
resnet_act_fn=act_fn,
|
|
output_scale_factor=mid_block_scale_factor,
|
|
cross_attention_dim=cross_attention_dim,
|
|
attn_num_head_channels=attention_head_dim[-1],
|
|
resnet_groups=norm_num_groups,
|
|
dual_cross_attention=False,
|
|
use_image_embedding=use_image_embedding,
|
|
unet_params=unet_params,
|
|
)
|
|
if self.merging_mode.startswith("attention"):
|
|
self.cross_attention_merger_mid_block = ConditionalModel(
|
|
input_channels=block_out_channels[-1], conditional_model=self.merging_mode.split("attention_")[1])
|
|
# count how many layers upsample the images
|
|
self.num_upsamplers = 0
|
|
|
|
# up
|
|
reversed_block_out_channels = list(reversed(block_out_channels))
|
|
reversed_attention_head_dim = list(reversed(attention_head_dim))
|
|
|
|
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=layers_per_block + 1,
|
|
in_channels=input_channel,
|
|
out_channels=output_channel,
|
|
prev_output_channel=prev_output_channel,
|
|
temb_channels=time_embed_dim,
|
|
add_upsample=add_upsample,
|
|
resnet_eps=norm_eps,
|
|
resnet_act_fn=act_fn,
|
|
resnet_groups=norm_num_groups,
|
|
cross_attention_dim=cross_attention_dim,
|
|
attn_num_head_channels=reversed_attention_head_dim[i],
|
|
dual_cross_attention=False,
|
|
use_image_embedding=use_image_embedding,
|
|
unet_params=unet_params,
|
|
)
|
|
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 = nn.SiLU()
|
|
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
|
|
)
|
|
'''
|
|
self.conv_out = Conv2D_ExtendedChannels(
|
|
block_out_channels[0], out_channels, kernel_size=conv_out_kernel, padding=conv_out_padding, out_channel_extension=2 if channel_expansion else 0,
|
|
)
|
|
|
|
def set_attention_slice(self, slice_size):
|
|
r"""
|
|
Enable sliced attention computation.
|
|
|
|
When this option is enabled, the attention module will split the input tensor in slices, to compute attention
|
|
in several steps. This is useful to save some memory in exchange for a small speed decrease.
|
|
|
|
Args:
|
|
slice_size (`str` or `int` or `list(int)`, *optional*, defaults to `"auto"`):
|
|
When `"auto"`, halves the input to the attention heads, so attention will be computed in two steps. If
|
|
`"max"`, maxium amount of memory will be 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_slicable_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_slicable_dims(child)
|
|
|
|
# retrieve number of attention layers
|
|
for module in self.children():
|
|
fn_recursive_retrieve_slicable_dims(module)
|
|
|
|
num_slicable_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_slicable_layers * [1]
|
|
|
|
slice_size = num_slicable_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)
|
|
|
|
def _set_gradient_checkpointing(self, value=False):
|
|
self.gradient_checkpointing = value
|
|
self.mid_block.gradient_checkpointing = value
|
|
for module in self.down_blocks + self.up_blocks:
|
|
if isinstance(module, (CrossAttnDownBlock3D, DownBlock3D, CrossAttnUpBlock3D, UpBlock3D)):
|
|
module.gradient_checkpointing = value
|
|
|
|
def forward(
|
|
self,
|
|
sample: torch.FloatTensor,
|
|
timestep: Union[torch.Tensor, float, int],
|
|
encoder_hidden_states: torch.Tensor,
|
|
fps: Optional[torch.Tensor] = None,
|
|
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,
|
|
down_block_additional_residuals: Optional[Tuple[torch.Tensor]] = None,
|
|
mid_block_additional_residual: Optional[torch.Tensor] = None,
|
|
return_dict: bool = True,
|
|
) -> Union[UNet3DConditionOutput, Tuple]:
|
|
r"""
|
|
Args:
|
|
sample (`torch.FloatTensor`): (batch, num_frames, channel, height, width) noisy inputs tensor
|
|
timestep (`torch.FloatTensor` or `float` or `int`): (batch) timesteps
|
|
encoder_hidden_states (`torch.FloatTensor`): (batch, sequence_length, feature_dim) encoder hidden states
|
|
return_dict (`bool`, *optional*, defaults to `True`):
|
|
Whether or not to return a [`models.unet_2d_condition.UNet3DConditionOutput`] instead of a plain tuple.
|
|
cross_attention_kwargs (`dict`, *optional*):
|
|
A kwargs dictionary that if specified is passed along to the `AttentionProcessor` as defined under
|
|
`self.processor` in
|
|
[diffusers.cross_attention](https://github.com/huggingface/diffusers/blob/main/src/diffusers/models/cross_attention.py).
|
|
|
|
Returns:
|
|
[`~models.unet_2d_condition.UNet3DConditionOutput`] or `tuple`:
|
|
[`~models.unet_2d_condition.UNet3DConditionOutput`] if `return_dict` is True, otherwise a `tuple`. When
|
|
returning a tuple, 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 layears).
|
|
# 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
|
|
|
|
# prepare attention_mask
|
|
'''
|
|
if attention_mask is not None:
|
|
attention_mask = (1 - attention_mask.to(sample.dtype)) * -10000.0
|
|
attention_mask = attention_mask.unsqueeze(1)
|
|
'''
|
|
debug = False
|
|
if self.use_fps_conditioning:
|
|
|
|
if torch.is_tensor(fps):
|
|
assert (fps > -1).all(), "FPS not set"
|
|
if len(fps.shape) == 0:
|
|
fps = fps[None].to(sample.device)
|
|
else:
|
|
assert (fps > -1), "FPS not set"
|
|
is_mps = sample.device.type == "mps"
|
|
if isinstance(fps, float):
|
|
dtype = torch.float32 if is_mps else torch.float64
|
|
else:
|
|
dtype = torch.int32 if is_mps else torch.int64
|
|
fps = torch.tensor([fps], dtype=dtype, device=sample.device)
|
|
fps = fps.expand(sample.shape[0])
|
|
fps_proj = self.fps_proj(fps)
|
|
fps_proj = fps_proj.to(dtype=self.dtype)
|
|
fps_emb = self.fps_embedding(fps_proj)
|
|
# 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
|
|
num_frames = sample.shape[2]
|
|
timesteps = timesteps.expand(sample.shape[0])
|
|
batch_size = 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=self.dtype)
|
|
|
|
emb = self.time_embedding(t_emb, timestep_cond)
|
|
|
|
emb = emb.repeat_interleave(repeats=num_frames, dim=0)
|
|
if self.use_fps_conditioning:
|
|
fps_emb = fps_emb.repeat_interleave(repeats=num_frames, dim=0)
|
|
emb = emb + fps_emb
|
|
|
|
if not self.use_image_tokens and encoder_hidden_states.shape[1] > 77:
|
|
encoder_hidden_states = encoder_hidden_states[:, :77]
|
|
# print(f"MAIN with tokens = {encoder_hidden_states.shape[1]}")
|
|
if encoder_hidden_states.shape[1] > 77:
|
|
# assert (
|
|
# encoder_hidden_states.shape[1]-77) % num_frames == 0, f"Encoder shape {encoder_hidden_states.shape}. Num frames = {num_frames}"
|
|
context_text, context_img = encoder_hidden_states[:,
|
|
:77, :], encoder_hidden_states[:, 77:, :]
|
|
context_text = context_text.repeat_interleave(
|
|
repeats=num_frames, dim=0)
|
|
|
|
if self.image_encoder_name == "FrozenOpenCLIPImageEmbedder":
|
|
context_img = context_img.repeat_interleave(
|
|
repeats=num_frames, dim=0)
|
|
else:
|
|
context_img = rearrange(
|
|
context_img, 'b (t l) c -> (b t) l c', t=num_frames)
|
|
|
|
encoder_hidden_states = torch.cat(
|
|
[context_text, context_img], dim=1)
|
|
else:
|
|
encoder_hidden_states = encoder_hidden_states.repeat_interleave(
|
|
repeats=num_frames, dim=0)
|
|
|
|
# 2. pre-process
|
|
sample = sample.permute(0, 2, 1, 3, 4).reshape(
|
|
(sample.shape[0] * num_frames, -1) + sample.shape[3:])
|
|
sample = self.conv_in(sample)
|
|
|
|
if num_frames > 1:
|
|
if self.gradient_checkpointing:
|
|
sample = transformer_g_c(
|
|
self.transformer_in, sample, num_frames)
|
|
else:
|
|
sample = self.transformer_in(
|
|
sample, num_frames=num_frames, attention_mask=attention_mask).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,
|
|
num_frames=num_frames,
|
|
cross_attention_kwargs=cross_attention_kwargs,
|
|
)
|
|
else:
|
|
sample, res_samples = downsample_block(
|
|
hidden_states=sample, temb=emb, num_frames=num_frames)
|
|
|
|
down_block_res_samples += res_samples
|
|
|
|
if down_block_additional_residuals is not None:
|
|
new_down_block_res_samples = ()
|
|
|
|
if self.merging_mode == "addition":
|
|
for down_block_res_sample, down_block_additional_residual in zip(
|
|
down_block_res_samples, down_block_additional_residuals
|
|
):
|
|
down_block_res_sample = down_block_res_sample + down_block_additional_residual
|
|
new_down_block_res_samples += (down_block_res_sample,)
|
|
elif self.merging_mode.startswith("attention"):
|
|
for down_block_res_sample, down_block_additional_residual, merger in zip(
|
|
down_block_res_samples, down_block_additional_residuals, self.cross_attention_merger_down_blocks
|
|
):
|
|
|
|
down_block_res_sample = merger(
|
|
rearrange(down_block_res_sample, "(B F) C H W -> B F C H W", B=batch_size), rearrange(down_block_additional_residual, "(B F) C H W -> B F C H W", B=batch_size))
|
|
down_block_res_sample = rearrange(
|
|
down_block_res_sample, "B F C H W -> (B F) C H W")
|
|
new_down_block_res_samples += (down_block_res_sample,)
|
|
elif self.merging_mode == "overwrite":
|
|
for down_block_res_sample, down_block_additional_residual in zip(
|
|
down_block_res_samples, down_block_additional_residuals
|
|
):
|
|
down_block_res_sample = down_block_additional_residual
|
|
new_down_block_res_samples += (down_block_res_sample,)
|
|
down_block_res_samples = new_down_block_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,
|
|
num_frames=num_frames,
|
|
cross_attention_kwargs=cross_attention_kwargs,
|
|
)
|
|
|
|
if mid_block_additional_residual is not None:
|
|
if self.merging_mode == "addition":
|
|
sample = sample + mid_block_additional_residual
|
|
elif self.merging_mode == "overwrite":
|
|
sample = sample + mid_block_additional_residual
|
|
elif self.merging_mode.startswith("attention"):
|
|
sample = self.cross_attention_merger_mid_block(
|
|
rearrange(sample, "(B F) C H W -> B F C H W", B=batch_size), rearrange(mid_block_additional_residual, "(B F) C H W -> B F C H W", B=batch_size))
|
|
sample = rearrange(sample, "B F C H W -> (B F) C H W")
|
|
|
|
if debug:
|
|
upblockout = (sample,)
|
|
# 5. up
|
|
# import pdb
|
|
# pdb.set_trace()
|
|
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, output_states = upsample_block(
|
|
hidden_states=sample,
|
|
temb=emb,
|
|
res_hidden_states_tuple=res_samples,
|
|
encoder_hidden_states=encoder_hidden_states,
|
|
upsample_size=upsample_size,
|
|
attention_mask=attention_mask,
|
|
num_frames=num_frames,
|
|
cross_attention_kwargs=cross_attention_kwargs,
|
|
)
|
|
else:
|
|
sample, output_states = upsample_block(
|
|
hidden_states=sample,
|
|
temb=emb,
|
|
res_hidden_states_tuple=res_samples,
|
|
upsample_size=upsample_size,
|
|
num_frames=num_frames,
|
|
)
|
|
if debug:
|
|
upblockout += output_states
|
|
|
|
# 6. post-process
|
|
if self.conv_norm_out:
|
|
sample = self.conv_norm_out(sample)
|
|
sample = self.conv_act(sample)
|
|
|
|
sample = self.conv_out(sample)
|
|
|
|
# reshape to (batch, channel, framerate, width, height)
|
|
sample = sample[None, :].reshape(
|
|
(-1, num_frames) + sample.shape[1:]).permute(0, 2, 1, 3, 4)
|
|
|
|
if not return_dict:
|
|
return (sample,)
|
|
|
|
return UNet3DConditionOutput(sample=sample)
|