370 lines
16 KiB
Python
370 lines
16 KiB
Python
from typing import Union
|
|
from torch import Tensor
|
|
import torch
|
|
|
|
import comfy.utils
|
|
import comfy.controlnet as comfy_cn
|
|
from comfy.controlnet import ControlNet, T2IAdapter, broadcast_image_to
|
|
|
|
|
|
ControlNetWeightsType = list[float]
|
|
T2IAdapterWeightsType = list[float]
|
|
|
|
|
|
class LatentKeyframe:
|
|
def __init__(self, batch_index: int, strength: float) -> None:
|
|
self.batch_index = batch_index
|
|
self.strength = strength
|
|
|
|
|
|
# always maintain sorted state (by batch_index of LatentKeyframe)
|
|
class LatentKeyframeGroup:
|
|
def __init__(self) -> None:
|
|
self.keyframes: list[LatentKeyframe] = []
|
|
|
|
def add(self, keyframe: LatentKeyframe) -> None:
|
|
added = False
|
|
# replace existing keyframe if same batch_index
|
|
for i in range(len(self.keyframes)):
|
|
if self.keyframes[i].batch_index == keyframe.batch_index:
|
|
self.keyframes[i] = keyframe
|
|
added = True
|
|
break
|
|
if not added:
|
|
self.keyframes.append(keyframe)
|
|
self.keyframes.sort(key=lambda k: k.batch_index)
|
|
|
|
def get_index(self, index: int) -> Union[LatentKeyframe, None]:
|
|
try:
|
|
return self.keyframes[index]
|
|
except IndexError:
|
|
return None
|
|
|
|
def __getitem__(self, index) -> LatentKeyframe:
|
|
return self.keyframes[index]
|
|
|
|
def is_empty(self) -> bool:
|
|
return len(self.keyframes) == 0
|
|
|
|
|
|
class TimestepKeyframe:
|
|
def __init__(self,
|
|
start_percent: float = 0.0,
|
|
control_net_weights: ControlNetWeightsType = None,
|
|
t2i_adapter_weights: T2IAdapterWeightsType = None,
|
|
latent_keyframes: LatentKeyframeGroup = None,
|
|
default_latent_strength: float = 0.0) -> None:
|
|
self.start_percent = start_percent
|
|
self.control_net_weights = control_net_weights
|
|
self.t2i_adapter_weights = t2i_adapter_weights
|
|
self.latent_keyframes = latent_keyframes
|
|
self.default_latent_strength = default_latent_strength
|
|
|
|
|
|
@classmethod
|
|
def default(cls) -> 'TimestepKeyframe':
|
|
return cls(0.0)
|
|
|
|
|
|
# always maintain sorted state (by start_percent of TimestepKeyFrame)
|
|
class TimestepKeyframeGroup:
|
|
def __init__(self) -> None:
|
|
self.keyframes: list[TimestepKeyframe] = []
|
|
self.keyframes.append(TimestepKeyframe.default())
|
|
|
|
def add(self, keyframe: TimestepKeyframe) -> None:
|
|
added = False
|
|
# replace existing keyframe if same start_percent
|
|
for i in range(len(self.keyframes)):
|
|
if self.keyframes[i].start_percent == keyframe.start_percent:
|
|
self.keyframes[i] = keyframe
|
|
added = True
|
|
break
|
|
if not added:
|
|
self.keyframes.append(keyframe)
|
|
self.keyframes.sort(key=lambda k: k.start_percent)
|
|
|
|
def get_index(self, index: int) -> Union[TimestepKeyframe, None]:
|
|
try:
|
|
return self.keyframes[index]
|
|
except IndexError:
|
|
return None
|
|
|
|
def __getitem__(self, index) -> TimestepKeyframe:
|
|
return self.keyframes[index]
|
|
|
|
def is_empty(self) -> bool:
|
|
return len(self.keyframes) == 0
|
|
|
|
@classmethod
|
|
def default(cls, keyframe: TimestepKeyframe) -> 'TimestepKeyframeGroup':
|
|
group = cls()
|
|
group.keyframes[0] = keyframe
|
|
return group
|
|
|
|
|
|
# used to inject ControlNetAdvanced and T2IAdapterAdvanced control_merge function
|
|
def control_merge_inject(self, control_input, control_output, control_prev, output_dtype):
|
|
out = {'input':[], 'middle':[], 'output': []}
|
|
|
|
if control_input is not None:
|
|
for i in range(len(control_input)):
|
|
key = 'input'
|
|
x = control_input[i]
|
|
if x is not None:
|
|
self.apply_advanced_strengths_and_masks(x, self.current_timestep_keyframe, self.batched_number)
|
|
|
|
x *= self.strength * self.weights[i]
|
|
if x.dtype != output_dtype:
|
|
x = x.to(output_dtype)
|
|
out[key].insert(0, x)
|
|
|
|
if control_output is not None:
|
|
for i in range(len(control_output)):
|
|
if i == (len(control_output) - 1):
|
|
key = 'middle'
|
|
index = 0
|
|
else:
|
|
key = 'output'
|
|
index = i
|
|
x = control_output[i]
|
|
if x is not None:
|
|
self.apply_advanced_strengths_and_masks(x, self.current_timestep_keyframe, self.batched_number)
|
|
|
|
if self.global_average_pooling:
|
|
x = torch.mean(x, dim=(2, 3), keepdim=True).repeat(1, 1, x.shape[2], x.shape[3])
|
|
|
|
x *= self.strength * self.weights[i]
|
|
if x.dtype != output_dtype:
|
|
x = x.to(output_dtype)
|
|
|
|
out[key].append(x)
|
|
if control_prev is not None:
|
|
for x in ['input', 'middle', 'output']:
|
|
o = out[x]
|
|
for i in range(len(control_prev[x])):
|
|
prev_val = control_prev[x][i]
|
|
if i >= len(o):
|
|
o.append(prev_val)
|
|
elif prev_val is not None:
|
|
if o[i] is None:
|
|
o[i] = prev_val
|
|
else:
|
|
o[i] += prev_val
|
|
return out
|
|
|
|
|
|
class ControlNetAdvanced(ControlNet):
|
|
def __init__(self, control_model, timestep_keyframes: TimestepKeyframeGroup, global_average_pooling=False, device=None):
|
|
super().__init__(control_model=control_model, global_average_pooling=global_average_pooling, device=device)
|
|
# initialize timestep_keyframes
|
|
self.timestep_keyframes = timestep_keyframes if timestep_keyframes else TimestepKeyframeGroup()
|
|
self.current_timestep_keyframe = self.timestep_keyframes.keyframes[0]
|
|
# initialize weights
|
|
self.weights = self.timestep_keyframes.keyframes[0].control_net_weights if self.timestep_keyframes.keyframes[0].control_net_weights else [1.0]*13
|
|
# mask for which parts of controlnet output to keep
|
|
self.mask_cond_hint_original = None
|
|
self.mask_cond_hint = None
|
|
# actual index values
|
|
self.sub_idxs = None
|
|
self.full_latent_length = 0
|
|
self.context_length = 0
|
|
# override control_merge
|
|
self.control_merge = control_merge_inject.__get__(self, type(self))
|
|
|
|
def set_cond_hint_mask(self, mask_hint):
|
|
self.mask_cond_hint_original = mask_hint
|
|
return self
|
|
|
|
def get_control(self, x_noisy, t, cond, batched_number):
|
|
# need to reference t and batched_number later
|
|
self.t = t
|
|
self.batched_number = batched_number
|
|
# TODO: choose TimestepKeyframe based on t
|
|
# perform special version of get_control that supports sliding context and masks
|
|
return self.sliding_get_control(x_noisy, t, cond, batched_number)
|
|
|
|
def sliding_get_control(self, x_noisy: Tensor, t, cond, batched_number):
|
|
control_prev = None
|
|
if self.previous_controlnet is not None:
|
|
control_prev = self.previous_controlnet.get_control(x_noisy, t, cond, batched_number)
|
|
|
|
if self.timestep_range is not None:
|
|
if t[0] > self.timestep_range[0] or t[0] < self.timestep_range[1]:
|
|
if control_prev is not None:
|
|
return control_prev
|
|
else:
|
|
return None
|
|
|
|
output_dtype = x_noisy.dtype
|
|
|
|
# make cond_hint appropriate dimensions
|
|
# TODO: change this to not require cond_hint upscaling every step when self.sub_idxs are present
|
|
if self.sub_idxs is not None or self.cond_hint is None or x_noisy.shape[2] * 8 != self.cond_hint.shape[2] or x_noisy.shape[3] * 8 != self.cond_hint.shape[3]:
|
|
if self.cond_hint is not None:
|
|
del self.cond_hint
|
|
self.cond_hint = None
|
|
self.cond_hint = comfy.utils.common_upscale(self.cond_hint_original, x_noisy.shape[3] * 8, x_noisy.shape[2] * 8, 'nearest-exact', "center").to(self.control_model.dtype).to(self.device)
|
|
# if self.cond_hint length matches real latent count, need to subdivide it
|
|
if self.cond_hint.size(0) == self.full_latent_length:
|
|
self.cond_hint = self.cond_hint[self.sub_idxs]
|
|
if x_noisy.shape[0] != self.cond_hint.shape[0]:
|
|
self.cond_hint = broadcast_image_to(self.cond_hint, x_noisy.shape[0], batched_number)
|
|
|
|
# make mask appropriate dimensions, if present
|
|
if self.mask_cond_hint_original is not None:
|
|
if self.sub_idxs is not None or self.mask_cond_hint is None or x_noisy.shape[2] * 8 != self.mask_cond_hint.shape[1] or x_noisy.shape[3] * 8 != self.mask_cond_hint.shape[2]:
|
|
if self.mask_cond_hint is not None:
|
|
del self.mask_cond_hint
|
|
self.mask_cond_hint = None
|
|
# resize mask and match batch count
|
|
self.mask_cond_hint = prepare_mask_batch(self.mask_cond_hint_original, x_noisy.shape, multiplier=8)
|
|
actual_latent_length = x_noisy.shape[0] // batched_number
|
|
self.mask_cond_hint = comfy.utils.repeat_to_batch_size(self.mask_cond_hint, actual_latent_length if self.sub_idxs is None else self.full_latent_length)
|
|
if self.sub_idxs is not None:
|
|
self.mask_cond_hint = self.mask_cond_hint[self.sub_idxs]
|
|
# make cond_hint_mask length match x_noise
|
|
if x_noisy.shape[0] != self.mask_cond_hint.shape[0]:
|
|
self.mask_cond_hint = broadcast_image_to(self.mask_cond_hint, x_noisy.shape[0], batched_number)
|
|
self.mask_cond_hint = self.mask_cond_hint.to(self.control_model.dtype).to(self.device)
|
|
|
|
context = cond['c_crossattn']
|
|
y = cond.get('c_adm', None)
|
|
if y is not None:
|
|
y = y.to(self.control_model.dtype)
|
|
control = self.control_model(x=x_noisy.to(self.control_model.dtype), hint=self.cond_hint, timesteps=t, context=context.to(self.control_model.dtype), y=y)
|
|
return self.control_merge(None, control, control_prev, output_dtype)
|
|
|
|
def apply_advanced_strengths_and_masks(self, x: Tensor, current_timestep_keyframe: TimestepKeyframe, batched_number: int):
|
|
# apply strengths, and get batch indeces to default out
|
|
# AKA latents that should not be influenced by ControlNet
|
|
if current_timestep_keyframe.latent_keyframes is not None:
|
|
latent_count = x.size(0)//batched_number
|
|
indeces_to_default = set(range(latent_count))
|
|
mapped_indeces = None
|
|
# if expecting subdivision, will need to translate between subset and actual idx values
|
|
if self.sub_idxs:
|
|
mapped_indeces = {}
|
|
for i, actual in enumerate(self.sub_idxs):
|
|
mapped_indeces[actual] = i
|
|
for keyframe in current_timestep_keyframe.latent_keyframes:
|
|
real_index = keyframe.batch_index
|
|
# if negative, count from end
|
|
if real_index < 0:
|
|
real_index += latent_count if self.sub_idxs is None else self.full_latent_length
|
|
|
|
# if not mapping indeces, what you see is what you get
|
|
if mapped_indeces is None:
|
|
if real_index in indeces_to_default:
|
|
indeces_to_default.remove(real_index)
|
|
# otherwise, see if batch_index is even included in this set of latents
|
|
else:
|
|
real_index = mapped_indeces.get(real_index, None)
|
|
if real_index is None:
|
|
continue
|
|
indeces_to_default.remove(real_index)
|
|
|
|
# apply strength for each batched cond/uncond
|
|
for b in range(batched_number):
|
|
x[(latent_count*b)+real_index] = x[(latent_count*b)+real_index] * keyframe.strength
|
|
|
|
# default them out by multiplying by default_latent_strength
|
|
for batch_index in indeces_to_default:
|
|
# apply default for each batched cond/uncond
|
|
for b in range(batched_number):
|
|
x[(latent_count*b)+batch_index] = x[(latent_count*b)+batch_index] * current_timestep_keyframe.default_latent_strength
|
|
# apply masks
|
|
if self.mask_cond_hint is not None:
|
|
# first, resize mask to required dims
|
|
masks = prepare_mask_batch(self.mask_cond_hint, x.shape)
|
|
x[:] = x[:] * masks
|
|
|
|
def copy(self):
|
|
c = ControlNetAdvanced(self.control_model, self.timestep_keyframes, global_average_pooling=self.global_average_pooling)
|
|
self.copy_to(c)
|
|
return c
|
|
|
|
def cleanup(self):
|
|
super().cleanup()
|
|
self.sub_idxs = None
|
|
self.full_latent_length = 0
|
|
self.context_length = 0
|
|
|
|
|
|
class T2IAdapterAdvanced(T2IAdapter):
|
|
def __init__(self, t2i_model, timestep_keyframes: TimestepKeyframeGroup, channels_in, device=None):
|
|
super().__init__(t2i_model=t2i_model, channels_in=channels_in, device=device)
|
|
self.timestep_keyframes = timestep_keyframes if timestep_keyframes else TimestepKeyframeGroup()
|
|
self.current_timestep_keyframe = self.timestep_keyframes.keyframes[0]
|
|
first_weight = self.timestep_keyframes.keyframes[0].t2i_adapter_weights if self.timestep_keyframes.get_index(0) else None
|
|
self.weights = first_weight if first_weight else [1.0]*12
|
|
# mask for which parts of controlnet output to keep
|
|
self.cond_hint_mask = None
|
|
# actual index values
|
|
self.sub_idxs = None
|
|
self.full_latent_length = 0
|
|
self.context_length = 0
|
|
# override control_merge
|
|
self.control_merge = control_merge_inject.__get__(self, type(self))
|
|
|
|
def get_control(self, x_noisy, t, cond, batched_number):
|
|
# need to reference t and batched_number later
|
|
self.t = t
|
|
self.batched_number = batched_number
|
|
# TODO: choose TimestepKeyframe based on t
|
|
try:
|
|
# if sub indexes present, replace original hint with subsection
|
|
if self.sub_idxs is not None:
|
|
full_cond_hint_original = self.cond_hint_original
|
|
del self.cond_hint
|
|
self.cond_hint = None
|
|
self.cond_hint_original = full_cond_hint_original[self.sub_idxs]
|
|
return super().get_control(x_noisy, t, cond, batched_number)
|
|
finally:
|
|
if self.sub_idxs is not None:
|
|
# replace original cond hint
|
|
self.cond_hint_original = full_cond_hint_original
|
|
del full_cond_hint_original
|
|
|
|
def apply_advanced_strengths_and_masks(self, x, current_timestep_keyframe: TimestepKeyframe, batched_number: int):
|
|
# For now, do nothing; need to figure out LatentKeyframe control is even possible for T2I Adapters
|
|
# TODO: support masks
|
|
return
|
|
|
|
def copy(self):
|
|
c = T2IAdapterAdvanced(self.t2i_model, self.timestep_keyframes, self.channels_in)
|
|
self.copy_to(c)
|
|
return c
|
|
|
|
def cleanup(self):
|
|
super().cleanup()
|
|
self.sub_idxs = None
|
|
self.full_latent_length = 0
|
|
self.context_length = 0
|
|
|
|
|
|
def load_controlnet(ckpt_path, timestep_keyframe: TimestepKeyframeGroup=None, model=None):
|
|
control = comfy_cn.load_controlnet(ckpt_path, model=model)
|
|
# if exactly ControlNet returned, transform it into ControlNetAdvanced
|
|
if type(control) == ControlNet:
|
|
return ControlNetAdvanced(control.control_model, timestep_keyframe, global_average_pooling=control.global_average_pooling)
|
|
# if T2IAdapter returned, transform it into T2IAdapterAdvanced
|
|
elif isinstance(control, T2IAdapter):
|
|
return T2IAdapterAdvanced(control.t2i_model, timestep_keyframe, control.channels_in)
|
|
# otherwise, leave it be - probably a ControlLora for SDXL (no support for advanced stuff yet from here)
|
|
# TODO add ControlLoraAdvanced
|
|
return control
|
|
|
|
|
|
def is_advanced_controlnet(input_object):
|
|
return isinstance(input_object, ControlNetAdvanced) or isinstance(input_object, T2IAdapterAdvanced)
|
|
|
|
|
|
# adapted from comfy/sample.py
|
|
def prepare_mask_batch(mask: Tensor, shape: Tensor, multiplier: int=1, match_dim1=False):
|
|
mask = mask.clone()
|
|
mask = torch.nn.functional.interpolate(mask.reshape((-1, 1, mask.shape[-2], mask.shape[-1])), size=(shape[2]*multiplier, shape[3]*multiplier), mode="bilinear")
|
|
if match_dim1:
|
|
mask = torch.cat([mask] * shape[1], dim=1)
|
|
return mask
|