MotionCtrl / lvdm /modules /attention_temporal.py
wzhouxiff's picture
init
f1df74a
raw
history blame
43.7 kB
import math
from inspect import isfunction
import torch
import torch as th
from torch import nn, einsum
import torch.nn.functional as F
from einops import rearrange, repeat
try:
import xformers
import xformers.ops
XFORMERS_IS_AVAILBLE = True
except:
XFORMERS_IS_AVAILBLE = False
from lvdm.common import (
checkpoint,
exists,
uniq,
default,
max_neg_value,
init_
)
from lvdm.basics import (
conv_nd,
zero_module,
normalization
)
class GEGLU(nn.Module):
def __init__(self, dim_in, dim_out):
super().__init__()
self.proj = nn.Linear(dim_in, dim_out * 2)
def forward(self, x):
x, gate = self.proj(x).chunk(2, dim=-1)
return x * F.gelu(gate)
class FeedForward(nn.Module):
def __init__(self, dim, dim_out=None, mult=4, glu=False, dropout=0.):
super().__init__()
inner_dim = int(dim * mult)
dim_out = default(dim_out, dim)
project_in = nn.Sequential(
nn.Linear(dim, inner_dim),
nn.GELU()
) if not glu else GEGLU(dim, inner_dim)
self.net = nn.Sequential(
project_in,
nn.Dropout(dropout),
nn.Linear(inner_dim, dim_out)
)
def forward(self, x):
return self.net(x)
def Normalize(in_channels):
return torch.nn.GroupNorm(num_groups=32, num_channels=in_channels, eps=1e-6, affine=True)
# ---------------------------------------------------------------------------------------------------
class RelativePosition(nn.Module):
""" https://github.com/evelinehong/Transformer_Relative_Position_PyTorch/blob/master/relative_position.py """
def __init__(self, num_units, max_relative_position):
super().__init__()
self.num_units = num_units
self.max_relative_position = max_relative_position
self.embeddings_table = nn.Parameter(th.Tensor(max_relative_position * 2 + 1, num_units))
nn.init.xavier_uniform_(self.embeddings_table)
def forward(self, length_q, length_k):
device = self.embeddings_table.device
range_vec_q = th.arange(length_q, device=device)
range_vec_k = th.arange(length_k, device=device)
distance_mat = range_vec_k[None, :] - range_vec_q[:, None]
distance_mat_clipped = th.clamp(distance_mat, -self.max_relative_position, self.max_relative_position)
final_mat = distance_mat_clipped + self.max_relative_position
# final_mat = th.LongTensor(final_mat).to(self.embeddings_table.device)
# final_mat = th.tensor(final_mat, device=self.embeddings_table.device, dtype=torch.long)
final_mat = final_mat.long()
embeddings = self.embeddings_table[final_mat]
return embeddings
class TemporalCrossAttention(nn.Module):
def __init__(self,
query_dim,
context_dim=None,
heads=8,
dim_head=64,
dropout=0.,
temporal_length=None, # For relative positional representation and image-video joint training.
image_length=None, # For image-video joint training.
use_relative_position=False, # whether use relative positional representation in temporal attention.
img_video_joint_train=False, # For image-video joint training.
use_tempoal_causal_attn=False,
bidirectional_causal_attn=False,
tempoal_attn_type=None,
joint_train_mode="same_batch",
**kwargs,
):
super().__init__()
inner_dim = dim_head * heads
context_dim = default(context_dim, query_dim)
self.context_dim = context_dim
self.scale = dim_head ** -0.5
self.heads = heads
self.temporal_length = temporal_length
self.use_relative_position = use_relative_position
self.img_video_joint_train = img_video_joint_train
self.bidirectional_causal_attn = bidirectional_causal_attn
self.joint_train_mode = joint_train_mode
assert(joint_train_mode in ["same_batch", "diff_batch"])
self.tempoal_attn_type = tempoal_attn_type
if bidirectional_causal_attn:
assert use_tempoal_causal_attn
if tempoal_attn_type:
assert(tempoal_attn_type in ['sparse_causal', 'sparse_causal_first'])
assert(not use_tempoal_causal_attn)
assert(not (img_video_joint_train and (self.joint_train_mode == "same_batch")))
self.to_q = nn.Linear(query_dim, inner_dim, bias=False)
self.to_k = nn.Linear(context_dim, inner_dim, bias=False)
self.to_v = nn.Linear(context_dim, inner_dim, bias=False)
assert(not (img_video_joint_train and (self.joint_train_mode == "same_batch") and use_tempoal_causal_attn))
if img_video_joint_train:
if self.joint_train_mode == "same_batch":
mask = torch.ones([1, temporal_length+image_length, temporal_length+image_length])
# mask[:, image_length:, :] = 0
# mask[:, :, image_length:] = 0
mask[:, temporal_length:, :] = 0
mask[:, :, temporal_length:] = 0
self.mask = mask
else:
self.mask = None
elif use_tempoal_causal_attn:
# normal causal attn
self.mask = torch.tril(torch.ones([1, temporal_length, temporal_length]))
elif tempoal_attn_type == 'sparse_causal':
# all frames interact with only the `prev` & self frame
mask1 = torch.tril(torch.ones([1, temporal_length, temporal_length])).bool() # true indicates keeping
mask2 = torch.zeros([1, temporal_length, temporal_length]) # initialize to same shape with mask1
mask2[:,2:temporal_length, :temporal_length-2] = torch.tril(torch.ones([1,temporal_length-2, temporal_length-2]))
mask2=(1-mask2).bool() # false indicates masking
self.mask = mask1 & mask2
elif tempoal_attn_type == 'sparse_causal_first':
# all frames interact with only the `first` & self frame
mask1 = torch.tril(torch.ones([1, temporal_length, temporal_length])).bool() # true indicates keeping
mask2 = torch.zeros([1, temporal_length, temporal_length])
mask2[:,2:temporal_length, 1:temporal_length-1] = torch.tril(torch.ones([1,temporal_length-2, temporal_length-2]))
mask2=(1-mask2).bool() # false indicates masking
self.mask = mask1 & mask2
else:
self.mask = None
if use_relative_position:
assert(temporal_length is not None)
self.relative_position_k = RelativePosition(num_units=dim_head, max_relative_position=temporal_length)
self.relative_position_v = RelativePosition(num_units=dim_head, max_relative_position=temporal_length)
self.to_out = nn.Sequential(
nn.Linear(inner_dim, query_dim),
nn.Dropout(dropout)
)
nn.init.constant_(self.to_q.weight, 0)
nn.init.constant_(self.to_k.weight, 0)
nn.init.constant_(self.to_v.weight, 0)
nn.init.constant_(self.to_out[0].weight, 0)
nn.init.constant_(self.to_out[0].bias, 0)
def forward(self, x, context=None, mask=None):
# if context is None:
# print(f'[Temp Attn] x={x.shape},context=None')
# else:
# print(f'[Temp Attn] x={x.shape},context={context.shape}')
nh = self.heads
out = x
q = self.to_q(out)
# if context is not None:
# print(f'temporal context 1 ={context.shape}')
# print(f'x={x.shape}')
context = default(context, x)
# print(f'temporal context 2 ={context.shape}')
k = self.to_k(context)
v = self.to_v(context)
# print(f'q ={q.shape},k={k.shape}')
q, k, v = map(lambda t: rearrange(t, 'b n (h d) -> (b h) n d', h=nh), (q, k, v))
sim = einsum('b i d, b j d -> b i j', q, k) * self.scale
if self.use_relative_position:
len_q, len_k, len_v = q.shape[1], k.shape[1], v.shape[1]
k2 = self.relative_position_k(len_q, len_k)
sim2 = einsum('b t d, t s d -> b t s', q, k2) * self.scale # TODO check
sim += sim2
# print('mask',mask)
if exists(self.mask):
if mask is None:
mask = self.mask.to(sim.device)
else:
mask = self.mask.to(sim.device).bool() & mask #.to(sim.device)
else:
mask = mask
# if self.img_video_joint_train:
# # process mask (make mask same shape with sim)
# c, h, w = mask.shape
# c, t, s = sim.shape
# # assert(h == w and t == s),f"mask={mask.shape}, sim={sim.shape}, h={h}, w={w}, t={t}, s={s}"
# if h > t:
# mask = mask[:, :t, :]
# elif h < t: # pad zeros to mask (no attention) only initial mask =1 area compute weights
# mask_ = torch.zeros([c,t,w]).to(mask.device)
# mask_[:, :h, :] = mask
# mask = mask_
# c, h, w = mask.shape
# if w > s:
# mask = mask[:, :, :s]
# elif w < s: # pad zeros to mask
# mask_ = torch.zeros([c,h,s]).to(mask.device)
# mask_[:, :, :w] = mask
# mask = mask_
# max_neg_value = -torch.finfo(sim.dtype).max
# sim = sim.float().masked_fill(mask == 0, max_neg_value)
if mask is not None:
max_neg_value = -1e9
sim = sim + (1-mask.float()) * max_neg_value # 1=masking,0=no masking
# print('sim after masking: ', sim)
# if torch.isnan(sim).any() or torch.isinf(sim).any() or (not sim.any()):
# print(f'sim [after masking], isnan={torch.isnan(sim).any()}, isinf={torch.isinf(sim).any()}, allzero={not sim.any()}')
attn = sim.softmax(dim=-1)
# print('attn after softmax: ', attn)
# if torch.isnan(attn).any() or torch.isinf(attn).any() or (not attn.any()):
# print(f'attn [after softmax], isnan={torch.isnan(attn).any()}, isinf={torch.isinf(attn).any()}, allzero={not attn.any()}')
# attn = torch.where(torch.isnan(attn), torch.full_like(attn,0), attn)
# if torch.isinf(attn.detach()).any():
# import pdb;pdb.set_trace()
# if torch.isnan(attn.detach()).any():
# import pdb;pdb.set_trace()
out = einsum('b i j, b j d -> b i d', attn, v)
if self.bidirectional_causal_attn:
mask_reverse = torch.triu(torch.ones([1, self.temporal_length, self.temporal_length], device=sim.device))
sim_reverse = sim.float().masked_fill(mask_reverse == 0, max_neg_value)
attn_reverse = sim_reverse.softmax(dim=-1)
out_reverse = einsum('b i j, b j d -> b i d', attn_reverse, v)
out += out_reverse
if self.use_relative_position:
v2 = self.relative_position_v(len_q, len_v)
out2 = einsum('b t s, t s d -> b t d', attn, v2) # TODO check
out += out2 # TODO check:先add还是先merge head?先计算rpr,on split head之后的数据,然后再merge。
out = rearrange(out, '(b h) n d -> b n (h d)', h=nh) # merge head
return self.to_out(out)
class CrossAttention(nn.Module):
def __init__(self, query_dim, context_dim=None, heads=8, dim_head=64, dropout=0.,
sa_shared_kv=False, shared_type='only_first', **kwargs,):
super().__init__()
inner_dim = dim_head * heads
context_dim = default(context_dim, query_dim)
self.sa_shared_kv = sa_shared_kv
assert(shared_type in ['only_first', 'all_frames', 'first_and_prev', 'only_prev', 'full', 'causal', 'full_qkv'])
self.shared_type = shared_type
self.dim_head = dim_head
self.scale = dim_head ** -0.5
self.heads = heads
self.to_q = nn.Linear(query_dim, inner_dim, bias=False)
self.to_k = nn.Linear(context_dim, inner_dim, bias=False)
self.to_v = nn.Linear(context_dim, inner_dim, bias=False)
self.to_out = nn.Sequential(
nn.Linear(inner_dim, query_dim),
nn.Dropout(dropout)
)
if XFORMERS_IS_AVAILBLE:
self.forward = self.efficient_forward
def forward(self, x, context=None, mask=None):
h = self.heads
b = x.shape[0]
q = self.to_q(x)
context = default(context, x)
k = self.to_k(context)
v = self.to_v(context)
if self.sa_shared_kv:
if self.shared_type == 'only_first':
k,v = map(lambda xx: rearrange(xx[0].unsqueeze(0), 'b n c -> (b n) c').unsqueeze(0).repeat(b,1,1),
(k,v))
else:
raise NotImplementedError
q, k, v = map(lambda t: rearrange(t, 'b n (h d) -> (b h) n d', h=h), (q, k, v))
sim = einsum('b i d, b j d -> b i j', q, k) * self.scale
if exists(mask):
mask = rearrange(mask, 'b ... -> b (...)')
max_neg_value = -torch.finfo(sim.dtype).max
mask = repeat(mask, 'b j -> (b h) () j', h=h)
sim.masked_fill_(~mask, max_neg_value)
# attention, what we cannot get enough of
attn = sim.softmax(dim=-1)
out = einsum('b i j, b j d -> b i d', attn, v)
out = rearrange(out, '(b h) n d -> b n (h d)', h=h)
return self.to_out(out)
def efficient_forward(self, x, context=None, mask=None):
q = self.to_q(x)
context = default(context, x)
k = self.to_k(context)
v = self.to_v(context)
b, _, _ = q.shape
q, k, v = map(
lambda t: t.unsqueeze(3)
.reshape(b, t.shape[1], self.heads, self.dim_head)
.permute(0, 2, 1, 3)
.reshape(b * self.heads, t.shape[1], self.dim_head)
.contiguous(),
(q, k, v),
)
# actually compute the attention, what we cannot get enough of
out = xformers.ops.memory_efficient_attention(q, k, v, attn_bias=None, op=None)
if exists(mask):
raise NotImplementedError
out = (
out.unsqueeze(0)
.reshape(b, self.heads, out.shape[1], self.dim_head)
.permute(0, 2, 1, 3)
.reshape(b, out.shape[1], self.heads * self.dim_head)
)
return self.to_out(out)
class VideoSpatialCrossAttention(CrossAttention):
def __init__(self, query_dim, context_dim=None, heads=8, dim_head=64, dropout=0):
super().__init__(query_dim, context_dim, heads, dim_head, dropout)
def forward(self, x, context=None, mask=None):
b, c, t, h, w = x.shape
if context is not None:
context = context.repeat(t, 1, 1)
x = super.forward(spatial_attn_reshape(x), context=context) + x
return spatial_attn_reshape_back(x,b,h)
class BasicTransformerBlockST(nn.Module):
def __init__(self,
# Spatial Stuff
dim,
n_heads,
d_head,
dropout=0.,
context_dim=None,
gated_ff=True,
checkpoint=True,
# Temporal Stuff
temporal_length=None,
image_length=None,
use_relative_position=True,
img_video_joint_train=False,
cross_attn_on_tempoal=False,
temporal_crossattn_type="selfattn",
order="stst",
temporalcrossfirst=False,
temporal_context_dim=None,
split_stcontext=False,
local_spatial_temporal_attn=False,
window_size=2,
**kwargs,
):
super().__init__()
# Self attention
self.attn1 = CrossAttention(query_dim=dim, heads=n_heads, dim_head=d_head, dropout=dropout, **kwargs,)
self.ff = FeedForward(dim, dropout=dropout, glu=gated_ff)
# cross attention if context is not None
self.attn2 = CrossAttention(query_dim=dim, context_dim=context_dim,
heads=n_heads, dim_head=d_head, dropout=dropout, **kwargs,)
self.norm1 = nn.LayerNorm(dim)
self.norm2 = nn.LayerNorm(dim)
self.norm3 = nn.LayerNorm(dim)
self.checkpoint = checkpoint
self.order = order
assert(self.order in ["stst", "sstt", "st_parallel"])
self.temporalcrossfirst = temporalcrossfirst
self.split_stcontext = split_stcontext
self.local_spatial_temporal_attn = local_spatial_temporal_attn
if self.local_spatial_temporal_attn:
assert(self.order == 'stst')
assert(self.order == 'stst')
self.window_size = window_size
if not split_stcontext:
temporal_context_dim = context_dim
# Temporal attention
assert(temporal_crossattn_type in ["selfattn", "crossattn", "skip"])
self.temporal_crossattn_type = temporal_crossattn_type
self.attn1_tmp = TemporalCrossAttention(query_dim=dim, heads=n_heads, dim_head=d_head, dropout=dropout,
temporal_length=temporal_length,
image_length=image_length,
use_relative_position=use_relative_position,
img_video_joint_train=img_video_joint_train,
**kwargs,
)
self.attn2_tmp = TemporalCrossAttention(query_dim=dim, heads=n_heads, dim_head=d_head, dropout=dropout,
# cross attn
context_dim=temporal_context_dim if temporal_crossattn_type == "crossattn" else None,
# temporal attn
temporal_length=temporal_length,
image_length=image_length,
use_relative_position=use_relative_position,
img_video_joint_train=img_video_joint_train,
**kwargs,
)
self.norm4 = nn.LayerNorm(dim)
self.norm5 = nn.LayerNorm(dim)
# self.norm1_tmp = nn.LayerNorm(dim)
# self.norm2_tmp = nn.LayerNorm(dim)
##############################################################################################################################################
def forward(self, x, context=None, temporal_context=None, no_temporal_attn=None, attn_mask=None, **kwargs):
# print(f'no_temporal_attn={no_temporal_attn}')
if not self.split_stcontext:
# st cross attention use the same context vector
temporal_context = context.detach().clone()
if context is None and temporal_context is None:
# self-attention models
if no_temporal_attn:
raise NotImplementedError
return checkpoint(self._forward_nocontext, (x), self.parameters(), self.checkpoint)
else:
# cross-attention models
if no_temporal_attn:
forward_func = self._forward_no_temporal_attn
else:
forward_func = self._forward
inputs = (x, context, temporal_context) if temporal_context is not None else (x, context)
return checkpoint(forward_func, inputs, self.parameters(), self.checkpoint)
# if attn_mask is not None:
# return checkpoint(self._forward, (x, context, temporal_context, attn_mask), self.parameters(), self.checkpoint)
# return checkpoint(self._forward, (x, context, temporal_context), self.parameters(), self.checkpoint)
def _forward(self, x, context=None, temporal_context=None, mask=None, no_temporal_attn=None, ):
assert(x.dim() == 5), f"x shape = {x.shape}"
b, c, t, h, w = x.shape
if self.order in ["stst", "sstt"]:
x = self._st_cross_attn(x, context, temporal_context=temporal_context, order=self.order, mask=mask,)#no_temporal_attn=no_temporal_attn,
elif self.order == "st_parallel":
x = self._st_cross_attn_parallel(x, context, temporal_context=temporal_context, order=self.order,)#no_temporal_attn=no_temporal_attn,
else:
raise NotImplementedError
x = self.ff(self.norm3(x)) + x
if (no_temporal_attn is None) or (not no_temporal_attn):
x = rearrange(x, '(b h w) t c -> b c t h w', b=b,h=h,w=w) # 3d -> 5d
elif no_temporal_attn:
x = rearrange(x, '(b t) (h w) c -> b c t h w', b=b,h=h,w=w) # 3d -> 5d
return x
def _forward_no_temporal_attn(self, x, context=None, temporal_context=None, ):
# temporary implementation :(
# because checkpoint does not support non-tensor inputs currently.
assert(x.dim() == 5), f"x shape = {x.shape}"
b, c, t, h, w = x.shape
if self.order in ["stst", "sstt"]:
# x = self._st_cross_attn(x, context, temporal_context=temporal_context, order=self.order, no_temporal_attn=True,)
# mask = torch.zeros([1, t, t], device=x.device).bool() if context is None else torch.zeros([1, context.shape[1], t], device=x.device).bool()
mask = torch.zeros([1, t, t], device=x.device).bool()
x = self._st_cross_attn(x, context, temporal_context=temporal_context, order=self.order, mask=mask,)
elif self.order == "st_parallel":
x = self._st_cross_attn_parallel(x, context, temporal_context=temporal_context, order=self.order, no_temporal_attn=True,)
else:
raise NotImplementedError
x = self.ff(self.norm3(x)) + x
x = rearrange(x, '(b h w) t c -> b c t h w', b=b,h=h,w=w) # 3d -> 5d
# x = rearrange(x, '(b t) (h w) c -> b c t h w', b=b,h=h,w=w) # 3d -> 5d
return x
def _forward_nocontext(self, x, no_temporal_attn=None):
assert(x.dim() == 5), f"x shape = {x.shape}"
b, c, t, h, w = x.shape
if self.order in ["stst", "sstt"]:
x = self._st_cross_attn(x, order=self.order, no_temporal_attn=no_temporal_attn)
elif self.order == "st_parallel":
x = self._st_cross_attn_parallel(x, order=self.order, no_temporal_attn=no_temporal_attn)
else:
raise NotImplementedError
x = self.ff(self.norm3(x)) + x
x = rearrange(x, '(b h w) t c -> b c t h w', b=b,h=h,w=w) # 3d -> 5d
return x
##############################################################################################################################################
def _st_cross_attn(self, x, context=None, temporal_context=None, order="stst", mask=None): #no_temporal_attn=None,
b, c, t, h, w = x.shape
# print(f'[_st_cross_attn input] x={x.shape}, context={context.shape}')
if order == "stst":
# spatial self attention
x = rearrange(x, 'b c t h w -> (b t) (h w) c')
x = self.attn1(self.norm1(x)) + x
x = rearrange(x, '(b t) (h w) c -> b c t h w', b=b,h=h)
# temporal self attention
# if (no_temporal_attn is None) or (not no_temporal_attn):
if self.local_spatial_temporal_attn:
x = local_spatial_temporal_attn_reshape(x,window_size=self.window_size)
else:
x = rearrange(x, 'b c t h w -> (b h w) t c')
x = self.attn1_tmp(self.norm4(x), mask=mask) + x
if self.local_spatial_temporal_attn:
x = local_spatial_temporal_attn_reshape_back(x, window_size=self.window_size,
b=b, h=h, w=w, t=t)
else:
x = rearrange(x, '(b h w) t c -> b c t h w', b=b,h=h,w=w) # 3d -> 5d
# spatial cross attention
x = rearrange(x, 'b c t h w -> (b t) (h w) c')
# context_ = context.repeat(t, 1, 1) if context is not None else None
# print(f'[before spatial cross] context={context.shape}')
if context is not None:
if context.shape[0] == t: # img captions no_temporal_attn or
context_ = context
else:
context_ = []
for i in range(context.shape[0]):
context_.append(context[i].unsqueeze(0).repeat(t, 1, 1))
context_ = torch.cat(context_,dim=0)
else:
context_ = None
# print(f'[before spatial cross] x={x.shape}, context_={context_.shape}')
x = self.attn2(self.norm2(x), context=context_) + x
# temporal cross attention
# if (no_temporal_attn is None) or (not no_temporal_attn):
x = rearrange(x, '(b t) (h w) c -> b c t h w', b=b,h=h)
x = rearrange(x, 'b c t h w -> (b h w) t c')
if self.temporal_crossattn_type == "crossattn":
# tmporal cross attention
if temporal_context is not None:
# print(f'STATTN context={context.shape}, temporal_context={temporal_context.shape}')
temporal_context = torch.cat([context, temporal_context], dim=1) # blc
# print(f'STATTN after concat temporal_context={temporal_context.shape}')
temporal_context = temporal_context.repeat(h*w, 1,1)
# print(f'after repeat temporal_context={temporal_context.shape}')
else:
temporal_context = context[0:1,...].repeat(h*w, 1, 1)
# print(f'STATTN after concat x={x.shape}')
x = self.attn2_tmp(self.norm5(x), context=temporal_context, mask=mask) + x
elif self.temporal_crossattn_type == "selfattn":
# temporal self attention
x = self.attn2_tmp(self.norm5(x), context=None, mask=mask) + x
elif self.temporal_crossattn_type == "skip":
# no temporal cross and self attention
pass
else:
raise NotImplementedError
elif order == "sstt":
# spatial self attention
x = rearrange(x, 'b c t h w -> (b t) (h w) c')
x = self.attn1(self.norm1(x)) + x
# spatial cross attention
context_ = context.repeat(t, 1, 1) if context is not None else None
x = self.attn2(self.norm2(x), context=context_) + x
x = rearrange(x, '(b t) (h w) c -> b c t h w', b=b,h=h)
if (no_temporal_attn is None) or (not no_temporal_attn):
if self.temporalcrossfirst:
# temporal cross attention
if self.temporal_crossattn_type == "crossattn":
# if temporal_context is not None:
temporal_context = context.repeat(h*w, 1, 1)
x = self.attn2_tmp(self.norm5(x), context=temporal_context, mask=mask) + x
elif self.temporal_crossattn_type == "selfattn":
x = self.attn2_tmp(self.norm5(x), context=None, mask=mask) + x
elif self.temporal_crossattn_type == "skip":
pass
else:
raise NotImplementedError
# temporal self attention
x = rearrange(x, 'b c t h w -> (b h w) t c')
x = self.attn1_tmp(self.norm4(x), mask=mask) + x
else:
# temporal self attention
x = rearrange(x, 'b c t h w -> (b h w) t c')
x = self.attn1_tmp(self.norm4(x), mask=mask) + x
# temporal cross attention
if self.temporal_crossattn_type == "crossattn":
if temporal_context is not None:
temporal_context = context.repeat(h*w, 1, 1)
x = self.attn2_tmp(self.norm5(x), context=temporal_context, mask=mask) + x
elif self.temporal_crossattn_type == "selfattn":
x = self.attn2_tmp(self.norm5(x), context=None, mask=mask) + x
elif self.temporal_crossattn_type == "skip":
pass
else:
raise NotImplementedError
else:
raise NotImplementedError
return x
def _st_cross_attn_parallel(self, x, context=None, temporal_context=None, order="sst", no_temporal_attn=None):
""" order: x -> Self Attn -> Cross Attn -> attn_s
x -> Temp Self Attn -> attn_t
x' = x + attn_s + attn_t
"""
if no_temporal_attn is not None:
raise NotImplementedError
B, C, T, H, W = x.shape
# spatial self attention
h = x
h = rearrange(h, 'b c t h w -> (b t) (h w) c')
h = self.attn1(self.norm1(h)) + h
# spatial cross
# context_ = context.repeat(T, 1, 1) if context is not None else None
if context is not None:
context_ = []
for i in range(context.shape[0]):
context_.append(context[i].unsqueeze(0).repeat(T, 1, 1))
context_ = torch.cat(context_,dim=0)
else:
context_ = None
h = self.attn2(self.norm2(h), context=context_) + h
h = rearrange(h, '(b t) (h w) c -> b c t h w', b=B, h=H)
# temporal self
h2 = x
h2 = rearrange(h2, 'b c t h w -> (b h w) t c')
h2 = self.attn1_tmp(self.norm4(h2))# + h2
h2 = rearrange(h2, '(b h w) t c -> b c t h w', b=B, h=H, w=W)
out = h + h2
return rearrange(out, 'b c t h w -> (b h w) t c')
##############################################################################################################################################
def spatial_attn_reshape(x):
return rearrange(x, 'b c t h w -> (b t) (h w) c')
def spatial_attn_reshape_back(x,b,h):
return rearrange(x, '(b t) (h w) c -> b c t h w', b=b,h=h)
def temporal_attn_reshape(x):
return rearrange(x, 'b c t h w -> (b h w) t c')
def temporal_attn_reshape_back(x, b,h,w):
return rearrange(x, '(b h w) t c -> b c t h w', b=b, h=h, w=w)
def local_spatial_temporal_attn_reshape(x, window_size):
B, C, T, H, W = x.shape
NH = H // window_size
NW = W // window_size
# x = x.view(B, C, T, NH, window_size, NW, window_size)
# tokens = x.permute(0, 1, 2, 3, 5, 4, 6).contiguous()
# tokens = tokens.view(-1, window_size, window_size, C)
x = rearrange(x, 'b c t (nh wh) (nw ww) -> b c t nh wh nw ww', nh=NH, nw=NW, wh=window_size, ww=window_size).contiguous() # # B, C, T, NH, NW, window_size, window_size
x = rearrange(x, 'b c t nh wh nw ww -> (b nh nw) (t wh ww) c') # (B, NH, NW) (T, window_size, window_size) C
return x
def local_spatial_temporal_attn_reshape_back(x, window_size, b, h, w, t):
B, L, C = x.shape
NH = h // window_size
NW = w // window_size
x = rearrange(x, '(b nh nw) (t wh ww) c -> b c t nh wh nw ww', b=b, nh=NH, nw=NW, t=t, wh=window_size, ww=window_size)
x = rearrange(x, 'b c t nh wh nw ww -> b c t (nh wh) (nw ww)')
return x
class SpatialTemporalTransformer(nn.Module):
"""
Transformer block for video-like data (5D tensor).
First, project the input (aka embedding) with NO reshape.
Then apply standard transformer action.
The 5D -> 3D reshape operation will be done in the specific attention module.
"""
def __init__(
self,
in_channels, n_heads, d_head,
depth=1, dropout=0.,
context_dim=None,
# Temporal stuff
temporal_length=None,
image_length=None,
use_relative_position=True,
img_video_joint_train=False,
cross_attn_on_tempoal=False,
temporal_crossattn_type=False,
order="stst",
temporalcrossfirst=False,
split_stcontext=False,
temporal_context_dim=None,
**kwargs,
):
super().__init__()
self.in_channels = in_channels
inner_dim = n_heads * d_head
self.norm = Normalize(in_channels)
self.proj_in = nn.Conv3d(in_channels,
inner_dim,
kernel_size=1,
stride=1,
padding=0)
self.transformer_blocks = nn.ModuleList(
[BasicTransformerBlockST(
inner_dim, n_heads, d_head, dropout=dropout,
# cross attn
context_dim=context_dim,
# temporal attn
temporal_length=temporal_length,
image_length=image_length,
use_relative_position=use_relative_position,
img_video_joint_train=img_video_joint_train,
temporal_crossattn_type=temporal_crossattn_type,
order=order,
temporalcrossfirst=temporalcrossfirst,
split_stcontext=split_stcontext,
temporal_context_dim=temporal_context_dim,
**kwargs
) for d in range(depth)]
)
self.proj_out = zero_module(nn.Conv3d(inner_dim,
in_channels,
kernel_size=1,
stride=1,
padding=0))
def forward(self, x, context=None, temporal_context=None, **kwargs):
# note: if no context is given, cross-attention defaults to self-attention
assert(x.dim() == 5), f"x shape = {x.shape}"
b, c, t, h, w = x.shape
x_in = x
x = self.norm(x)
x = self.proj_in(x)
for block in self.transformer_blocks:
x = block(x, context=context, temporal_context=temporal_context, **kwargs)
x = self.proj_out(x)
return x + x_in
# ---------------------------------------------------------------------------------------------------
class STAttentionBlock2(nn.Module):
def __init__(
self,
channels,
num_heads=1,
num_head_channels=-1,
use_checkpoint=False, # not used, only used in ResBlock
use_new_attention_order=False, # QKVAttention or QKVAttentionLegacy
temporal_length=16, # used in relative positional representation.
image_length=8, # used for image-video joint training.
use_relative_position=False, # whether use relative positional representation in temporal attention.
img_video_joint_train=False,
# norm_type="groupnorm",
attn_norm_type="group",
use_tempoal_causal_attn=False,
):
"""
version 1: guided_diffusion implemented version
version 2: remove args input argument
"""
super().__init__()
if num_head_channels == -1:
self.num_heads = num_heads
else:
assert (
channels % num_head_channels == 0
), f"q,k,v channels {channels} is not divisible by num_head_channels {num_head_channels}"
self.num_heads = channels // num_head_channels
self.use_checkpoint = use_checkpoint
self.temporal_length = temporal_length
self.image_length = image_length
self.use_relative_position = use_relative_position
self.img_video_joint_train = img_video_joint_train
self.attn_norm_type = attn_norm_type
assert(self.attn_norm_type in ["group", "no_norm"])
self.use_tempoal_causal_attn = use_tempoal_causal_attn
if self.attn_norm_type == "group":
self.norm_s = normalization(channels)
self.norm_t = normalization(channels)
self.qkv_s = conv_nd(1, channels, channels * 3, 1)
self.qkv_t = conv_nd(1, channels, channels * 3, 1)
if self.img_video_joint_train:
mask = th.ones([1, temporal_length+image_length, temporal_length+image_length])
mask[:, temporal_length:, :] = 0
mask[:, :, temporal_length:] = 0
self.register_buffer("mask", mask)
else:
self.mask = None
if use_new_attention_order:
# split qkv before split heads
self.attention_s = QKVAttention(self.num_heads)
self.attention_t = QKVAttention(self.num_heads)
else:
# split heads before split qkv
self.attention_s = QKVAttentionLegacy(self.num_heads)
self.attention_t = QKVAttentionLegacy(self.num_heads)
if use_relative_position:
self.relative_position_k = RelativePosition(num_units=channels // self.num_heads, max_relative_position=temporal_length)
self.relative_position_v = RelativePosition(num_units=channels // self.num_heads, max_relative_position=temporal_length)
self.proj_out_s = zero_module(conv_nd(1, channels, channels, 1)) # conv_dim, in_channels, out_channels, kernel_size
self.proj_out_t = zero_module(conv_nd(1, channels, channels, 1)) # conv_dim, in_channels, out_channels, kernel_size
def forward(self, x, mask=None):
b, c, t, h, w = x.shape
# spatial
out = rearrange(x, 'b c t h w -> (b t) c (h w)')
if self.attn_norm_type == "no_norm":
qkv = self.qkv_s(out)
else:
qkv = self.qkv_s(self.norm_s(out))
out = self.attention_s(qkv)
out = self.proj_out_s(out)
out = rearrange(out, '(b t) c (h w) -> b c t h w', b=b,h=h)
x += out
# temporal
out = rearrange(x, 'b c t h w -> (b h w) c t')
if self.attn_norm_type == "no_norm":
qkv = self.qkv_t(out)
else:
qkv = self.qkv_t(self.norm_t(out))
# relative positional embedding
if self.use_relative_position:
len_q = qkv.size()[-1]
len_k, len_v = len_q, len_q
k_rp = self.relative_position_k(len_q, len_k)
v_rp = self.relative_position_v(len_q, len_v) #[T,T,head_dim]
out = self.attention_t(qkv, rp=(k_rp, v_rp), mask=self.mask, use_tempoal_causal_attn=self.use_tempoal_causal_attn)
else:
out = self.attention_t(qkv, rp=None, mask=self.mask, use_tempoal_causal_attn=self.use_tempoal_causal_attn)
out = self.proj_out_t(out)
out = rearrange(out, '(b h w) c t -> b c t h w', b=b,h=h,w=w)
return (x + out)
# ---------------------------------------------------------------------------------------------------------------
class QKVAttentionLegacy(nn.Module):
"""
A module which performs QKV attention. Matches legacy QKVAttention + input/ouput heads shaping
"""
def __init__(self, n_heads):
super().__init__()
self.n_heads = n_heads
def forward(self, qkv, rp=None, mask=None):
"""
Apply QKV attention.
:param qkv: an [N x (H * 3 * C) x T] tensor of Qs, Ks, and Vs.
:return: an [N x (H * C) x T] tensor after attention.
"""
if rp is not None or mask is not None:
raise NotImplementedError
bs, width, length = qkv.shape
assert width % (3 * self.n_heads) == 0
ch = width // (3 * self.n_heads)
q, k, v = qkv.reshape(bs * self.n_heads, ch * 3, length).split(ch, dim=1)
scale = 1 / math.sqrt(math.sqrt(ch))
weight = th.einsum(
"bct,bcs->bts", q * scale, k * scale
) # More stable with f16 than dividing afterwards
weight = th.softmax(weight.float(), dim=-1).type(weight.dtype)
a = th.einsum("bts,bcs->bct", weight, v)
return a.reshape(bs, -1, length)
@staticmethod
def count_flops(model, _x, y):
return count_flops_attn(model, _x, y)
# ---------------------------------------------------------------------------------------------------------------
class QKVAttention(nn.Module):
"""
A module which performs QKV attention and splits in a different order.
"""
def __init__(self, n_heads):
super().__init__()
self.n_heads = n_heads
def forward(self, qkv, rp=None, mask=None, use_tempoal_causal_attn=False):
"""
Apply QKV attention.
:param qkv: an [N x (3 * H * C) x T] tensor of Qs, Ks, and Vs.
:return: an [N x (H * C) x T] tensor after attention.
"""
bs, width, length = qkv.shape
assert width % (3 * self.n_heads) == 0
ch = width // (3 * self.n_heads)
# print('qkv', qkv.size())
q, k, v = qkv.chunk(3, dim=1)
scale = 1 / math.sqrt(math.sqrt(ch))
# print('bs, self.n_heads, ch, length', bs, self.n_heads, ch, length)
weight = th.einsum(
"bct,bcs->bts",
(q * scale).view(bs * self.n_heads, ch, length),
(k * scale).view(bs * self.n_heads, ch, length),
) # More stable with f16 than dividing afterwards
# weight:[b,t,s] b=bs*n_heads*T
if rp is not None:
k_rp, v_rp = rp # [length, length, head_dim] [8, 8, 48]
weight2 = th.einsum(
'bct,tsc->bst',
(q * scale).view(bs * self.n_heads, ch, length),
k_rp
)
weight += weight2
if use_tempoal_causal_attn:
# weight = torch.tril(weight)
assert(mask is None), f'Not implemented for merging two masks!'
mask = torch.tril(torch.ones(weight.shape))
else:
if mask is not None: # only keep upper-left matrix
# process mask
c, t, _ = weight.shape
if mask.shape[-1] > t:
mask = mask[:, :t, :t]
elif mask.shape[-1] < t: # pad ones
mask_ = th.zeros([c,t,t]).to(mask.device)
t_ = mask.shape[-1]
mask_[:, :t_, :t_] = mask
mask = mask_
else:
assert(weight.shape[-1] == mask.shape[-1]), f'weight={weight.shape}, mask={mask.shape}'
if mask is not None:
INF = -1e8 #float('-inf')
weight = weight.float().masked_fill(mask == 0, INF)
weight = F.softmax(weight.float(), dim=-1).type(weight.dtype) #[256, 8, 8] [b, t, t] b=bs*n_heads*h*w,t=nframes
# weight = F.softmax(weight, dim=-1)#[256, 8, 8] [b, t, t] b=bs*n_heads*h*w,t=nframes
a = th.einsum("bts,bcs->bct", weight, v.reshape(bs * self.n_heads, ch, length)) #[256, 48, 8] [b, head_dim, t]
if rp is not None:
a2 = th.einsum(
"bts,tsc->btc",
weight,
v_rp
).transpose(1,2) # btc->bct
a += a2
return a.reshape(bs, -1, length)
# ---------------------------------------------------------------------------------------------------------------
# ---------------------------------------------------------------------------------------------------------------