File size: 43,017 Bytes
3dabe4a |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 |
import logging
from os import environ
import modules.scripts as scripts
import gradio as gr
import scipy.stats as stats
from scripts.ui_wrapper import UIWrapper, arg
from modules import script_callbacks, patches
from modules.hypernetworks import hypernetwork
#import modules.sd_hijack_optimizations
from modules.script_callbacks import CFGDenoiserParams, CFGDenoisedParams, AfterCFGCallbackParams
from modules.prompt_parser import reconstruct_multicond_batch
from modules.processing import StableDiffusionProcessing
#from modules.shared import sd_model, opts
from modules.sd_samplers_cfg_denoiser import catenate_conds
from modules.sd_samplers_cfg_denoiser import CFGDenoiser
from modules import shared
import math
import torch
from torch.nn import functional as F
from torchvision.transforms import GaussianBlur
from warnings import warn
from typing import Callable, Dict, Optional
from collections import OrderedDict
import torch
logger = logging.getLogger(__name__)
logger.setLevel(environ.get("SD_WEBUI_LOG_LEVEL", logging.INFO))
incantations_debug = environ.get("INCANTAIONS_DEBUG", False)
"""
An unofficial implementation of "Self-Rectifying Diffusion Sampling with Perturbed-Attention Guidance" for Automatic1111 WebUI.
@misc{ahn2024selfrectifying,
title={Self-Rectifying Diffusion Sampling with Perturbed-Attention Guidance},
author={Donghoon Ahn and Hyoungwon Cho and Jaewon Min and Wooseok Jang and Jungwoo Kim and SeonHwa Kim and Hyun Hee Park and Kyong Hwan Jin and Seungryong Kim},
year={2024},
eprint={2403.17377},
archivePrefix={arXiv},
primaryClass={cs.CV}
}
Include noise interval for CFG and PAG guidance in the sampling process from "Applying Guidance in a Limited Interval Improves
Sample and Distribution Quality in Diffusion Models"
@misc{kynkäänniemi2024applying,
title={Applying Guidance in a Limited Interval Improves Sample and Distribution Quality in Diffusion Models},
author={Tuomas Kynkäänniemi and Miika Aittala and Tero Karras and Samuli Laine and Timo Aila and Jaakko Lehtinen},
year={2024},
eprint={2404.07724},
archivePrefix={arXiv},
primaryClass={cs.CV}
}
Include CFG schedulers from "Analysis of Classifier-Free Guidance Weight Schedulers"
@misc{wang2024analysis,
title={Analysis of Classifier-Free Guidance Weight Schedulers},
author={Xi Wang and Nicolas Dufour and Nefeli Andreou and Marie-Paule Cani and Victoria Fernandez Abrevaya and David Picard and Vicky Kalogeiton},
year={2024},
eprint={2404.13040},
archivePrefix={arXiv},
primaryClass={cs.CV}
}
Saliency-adaptive noise fusion from arXiv:2311.10329 "High-fidelity Person-centric Subject-to-Image Synthesis"
@misc{wang2024highfidelity,
title={High-fidelity Person-centric Subject-to-Image Synthesis},
author={Yibin Wang and Weizhong Zhang and Jianwei Zheng and Cheng Jin},
year={2024},
eprint={2311.10329},
archivePrefix={arXiv},
primaryClass={cs.CV}
}
Author: v0xie
GitHub URL: https://github.com/v0xie/sd-webui-incantations
"""
handles = []
global_scale = 1
SCHEDULES = [
'Constant',
'Clamp-Linear (c=4.0)',
'Clamp-Linear (c=2.0)',
'Clamp-Linear (c=1.0)',
'Linear',
'Inverse-Linear',
'Cosine',
'Clamp-Cosine (c=4.0)',
'Clamp-Cosine (c=2.0)',
'Clamp-Cosine (c=1.0)',
'Sine',
'Interval',
'PCS (s=0.01)',
'PCS (s=0.1)',
'PCS (s=1.0)',
'PCS (s=2.0)',
'PCS (s=4.0)',
]
class PAGStateParams:
def __init__(self):
self.pag_active: bool = False # PAG guidance scale
self.pag_sanf: bool = False # saliency-adaptive noise fusion, handled in cfg_combiner
self.pag_scale: int = -1 # PAG guidance scale
self.pag_start_step: int = 0
self.pag_end_step: int = 150
self.cfg_interval_enable: bool = False
self.cfg_interval_schedule: str = 'Constant'
self.cfg_interval_low: float = 0
self.cfg_interval_high: float = 50.0
self.cfg_interval_scheduled_value: float = 7.0
self.step : int = 0
self.max_sampling_step : int = 1
self.guidance_scale: int = -1 # CFG
self.current_noise_level: float = 100.0
self.x_in = None
self.text_cond = None
self.image_cond = None
self.sigma = None
self.text_uncond = None
self.make_condition_dict = None # callable lambda
self.crossattn_modules = [] # callable lambda
self.to_v_modules = []
self.to_out_modules = []
self.pag_x_out = None
self.batch_size = -1 # Batch size
self.denoiser = None # CFGDenoiser
self.patched_combine_denoised = None
self.conds_list = None
self.uncond_shape_0 = None
class PAGExtensionScript(UIWrapper):
def __init__(self):
self.cached_c = [None, None]
self.handles = []
# Extension title in menu UI
def title(self) -> str:
return "Perturbed Attention Guidance"
# Decide to show menu in txt2img or img2img
def show(self, is_img2img):
return scripts.AlwaysVisible
# Setup menu ui detail
def setup_ui(self, is_img2img) -> list:
with gr.Accordion('Perturbed Attention Guidance', open=False):
active = gr.Checkbox(value=False, default=False, label="Active", elem_id='pag_active')
pag_sanf = gr.Checkbox(value=False, default=False, label="Use Saliency-Adaptive Noise Fusion", elem_id='pag_sanf')
with gr.Row():
pag_scale = gr.Slider(value = 0, minimum = 0, maximum = 20.0, step = 0.5, label="PAG Scale", elem_id = 'pag_scale', info="")
with gr.Row():
start_step = gr.Slider(value = 0, minimum = 0, maximum = 150, step = 1, label="Start Step", elem_id = 'pag_start_step', info="")
end_step = gr.Slider(value = 150, minimum = 0, maximum = 150, step = 1, label="End Step", elem_id = 'pag_end_step', info="")
with gr.Accordion('CFG Scheduler', open=False):
cfg_interval_enable = gr.Checkbox(value=False, default=False, label="Enable CFG Scheduler", elem_id='cfg_interval_enable', info="If enabled, applies CFG only within noise interval with the selected schedule type. PAG must be enabled (scale can be 0). SDXL recommend CFG=15; CFG interval (0.28, 5.42]")
with gr.Row():
cfg_schedule = gr.Dropdown(
value='Constant',
choices= SCHEDULES,
label="CFG Schedule Type",
elem_id='cfg_interval_schedule',
)
cfg_interval_low = gr.Slider(value = 0, minimum = 0, maximum = 100, step = 0.1, label="CFG Noise Interval Low", elem_id = 'cfg_interval_low', info="")
cfg_interval_high = gr.Slider(value = 100, minimum = 0, maximum = 100, step = 0.1, label="CFG Noise Interval High", elem_id = 'cfg_interval_high', info="")
active.do_not_save_to_config = True
pag_sanf.do_not_save_to_config = True
pag_scale.do_not_save_to_config = True
start_step.do_not_save_to_config = True
end_step.do_not_save_to_config = True
cfg_interval_enable.do_not_save_to_config = True
cfg_schedule.do_not_save_to_config = True
cfg_interval_low.do_not_save_to_config = True
cfg_interval_high.do_not_save_to_config = True
self.infotext_fields = [
(active, lambda d: gr.Checkbox.update(value='PAG Active' in d)),
(pag_sanf, lambda d: gr.Checkbox.update(value='PAG SANF' in d)),
(pag_scale, 'PAG Scale'),
(start_step, 'PAG Start Step'),
(end_step, 'PAG End Step'),
(cfg_interval_enable, 'CFG Interval Enable'),
(cfg_schedule, 'CFG Interval Schedule'),
(cfg_interval_low, 'CFG Interval Low'),
(cfg_interval_high, 'CFG Interval High')
]
self.paste_field_names = [
'pag_active',
'pag_sanf',
'pag_scale',
'pag_start_step',
'pag_end_step',
'cfg_interval_enable',
'cfg_interval_schedule',
'cfg_interval_low',
'cfg_interval_high',
]
return [active, pag_scale, start_step, end_step, cfg_interval_enable, cfg_schedule, cfg_interval_low, cfg_interval_high, pag_sanf]
def process_batch(self, p: StableDiffusionProcessing, *args, **kwargs):
self.pag_process_batch(p, *args, **kwargs)
def pag_process_batch(self, p: StableDiffusionProcessing, active, pag_scale, start_step, end_step, cfg_interval_enable, cfg_schedule, cfg_interval_low, cfg_interval_high, pag_sanf, *args, **kwargs):
# cleanup previous hooks always
script_callbacks.remove_current_script_callbacks()
self.remove_all_hooks()
active = getattr(p, "pag_active", active)
pag_sanf = getattr(p, "pag_sanf", pag_sanf)
cfg_interval_enable = getattr(p, "cfg_interval_enable", cfg_interval_enable)
if active is False and cfg_interval_enable is False:
return
pag_scale = getattr(p, "pag_scale", pag_scale)
start_step = getattr(p, "pag_start_step", start_step)
end_step = getattr(p, "pag_end_step", end_step)
cfg_schedule = getattr(p, "cfg_interval_schedule", cfg_schedule)
cfg_interval_low = getattr(p, "cfg_interval_low", cfg_interval_low)
cfg_interval_high = getattr(p, "cfg_interval_high", cfg_interval_high)
if active:
p.extra_generation_params.update({
"PAG Active": active,
"PAG SANF": pag_sanf,
"PAG Scale": pag_scale,
"PAG Start Step": start_step,
"PAG End Step": end_step,
})
if cfg_interval_enable:
p.extra_generation_params.update({
"CFG Interval Enable": cfg_interval_enable,
"CFG Interval Schedule": cfg_schedule,
"CFG Interval Low": cfg_interval_low,
"CFG Interval High": cfg_interval_high
})
self.create_hook(p, active, pag_scale, start_step, end_step, cfg_interval_enable, cfg_schedule, cfg_interval_low, cfg_interval_high, pag_sanf)
def create_hook(self, p: StableDiffusionProcessing, active, pag_scale, start_step, end_step, cfg_interval_enable, cfg_schedule, cfg_interval_low, cfg_interval_high, pag_sanf, *args, **kwargs):
# Create a list of parameters for each concept
pag_params = PAGStateParams()
# Add to p's incant_cfg_params
if not hasattr(p, 'incant_cfg_params'):
logger.error("No incant_cfg_params found in p")
p.incant_cfg_params['pag_params'] = pag_params
pag_params.pag_active = active
pag_params.pag_sanf = pag_sanf
pag_params.pag_scale = pag_scale
pag_params.pag_start_step = start_step
pag_params.pag_end_step = end_step
pag_params.cfg_interval_enable = cfg_interval_enable
pag_params.cfg_interval_schedule = cfg_schedule
pag_params.max_sampling_step = p.steps
pag_params.guidance_scale = p.cfg_scale
pag_params.batch_size = p.batch_size
pag_params.denoiser = None
pag_params.cfg_interval_scheduled_value = p.cfg_scale
if pag_params.cfg_interval_enable:
# Refer to 3.1 Practice in the paper
# We want to round high and low noise levels to the nearest integer index
low_index = find_closest_index(cfg_interval_low, pag_params.max_sampling_step)
high_index = find_closest_index(cfg_interval_high, pag_params.max_sampling_step)
pag_params.cfg_interval_low = calculate_noise_level(low_index, pag_params.max_sampling_step)
pag_params.cfg_interval_high = calculate_noise_level(high_index, pag_params.max_sampling_step)
logger.debug(f"Step Aligned CFG Interval (low, high): ({low_index}, {high_index}), Step Aligned CFG Interval: ({round(pag_params.cfg_interval_low, 4)}, {round(pag_params.cfg_interval_high, 4)})")
# Get all the qv modules
cross_attn_modules = self.get_cross_attn_modules()
if len(cross_attn_modules) == 0:
logger.error("No cross attention modules found, cannot proceed with PAG")
return
pag_params.crossattn_modules = [m for m in cross_attn_modules if 'CrossAttention' in m.__class__.__name__]
# Use lambda to call the callback function with the parameters to avoid global variables
cfg_denoise_lambda = lambda callback_params: self.on_cfg_denoiser_callback(callback_params, pag_params)
cfg_denoised_lambda = lambda callback_params: self.on_cfg_denoised_callback(callback_params, pag_params)
#after_cfg_lambda = lambda x: self.cfg_after_cfg_callback(x, params)
unhook_lambda = lambda _: self.unhook_callbacks(pag_params)
if pag_params.pag_active:
self.ready_hijack_forward(pag_params.crossattn_modules, pag_scale)
logger.debug('Hooked callbacks')
script_callbacks.on_cfg_denoiser(cfg_denoise_lambda)
script_callbacks.on_cfg_denoised(cfg_denoised_lambda)
#script_callbacks.on_cfg_after_cfg(after_cfg_lambda)
script_callbacks.on_script_unloaded(unhook_lambda)
def postprocess_batch(self, p, *args, **kwargs):
self.pag_postprocess_batch(p, *args, **kwargs)
def pag_postprocess_batch(self, p, active, *args, **kwargs):
script_callbacks.remove_current_script_callbacks()
logger.debug('Removed script callbacks')
active = getattr(p, "pag_active", active)
if active is False:
return
def remove_all_hooks(self):
cross_attn_modules = self.get_cross_attn_modules()
for module in cross_attn_modules:
to_v = getattr(module, 'to_v', None)
self.remove_field_cross_attn_modules(module, 'pag_enable')
self.remove_field_cross_attn_modules(module, 'pag_last_to_v')
self.remove_field_cross_attn_modules(to_v, 'pag_parent_module')
_remove_all_forward_hooks(module, 'pag_pre_hook')
_remove_all_forward_hooks(to_v, 'to_v_pre_hook')
def unhook_callbacks(self, pag_params: PAGStateParams):
global handles
return
if pag_params is None:
logger.error("PAG params is None")
return
if pag_params.denoiser is not None:
denoiser = pag_params.denoiser
setattr(denoiser, 'combine_denoised_patched', False)
try:
patches.undo(__name__, denoiser, "combine_denoised")
except KeyError:
logger.exception("KeyError unhooking combine_denoised")
pass
except RuntimeError:
logger.exception("RuntimeError unhooking combine_denoised")
pass
pag_params.denoiser = None
def ready_hijack_forward(self, crossattn_modules, pag_scale):
""" Create hooks in the forward pass of the cross attention modules
Copies the output of the to_v module to the parent module
Then applies the PAG perturbation to the output of the cross attention module (multiplication by identity)
"""
# add field for last_to_v
for module in crossattn_modules:
to_v = getattr(module, 'to_v', None)
self.add_field_cross_attn_modules(module, 'pag_enable', False)
self.add_field_cross_attn_modules(module, 'pag_last_to_v', None)
self.add_field_cross_attn_modules(to_v, 'pag_parent_module', [module])
# self.add_field_cross_attn_modules(to_out, 'pag_parent_module', [module])
def to_v_pre_hook(module, input, kwargs, output):
""" Copy the output of the to_v module to the parent module """
parent_module = getattr(module, 'pag_parent_module', None)
# copy the output of the to_v module to the parent module
setattr(parent_module[0], 'pag_last_to_v', output.detach().clone())
def pag_pre_hook(module, input, kwargs, output):
if hasattr(module, 'pag_enable') and getattr(module, 'pag_enable', False) is False:
return
if not hasattr(module, 'pag_last_to_v'):
# oops we forgot to unhook
return
# get the last to_v output and save it
last_to_v = getattr(module, 'pag_last_to_v', None)
batch_size, seq_len, inner_dim = output.shape
identity = torch.eye(seq_len, dtype=last_to_v.dtype, device=shared.device).expand(batch_size, -1, -1)
if last_to_v is not None:
new_output = torch.einsum('bij,bjk->bik', identity, last_to_v[:, :seq_len, :])
return new_output
else:
# this is bad
return output
# Create hooks
for module in crossattn_modules:
handle_parent = module.register_forward_hook(pag_pre_hook, with_kwargs=True)
to_v = getattr(module, 'to_v', None)
handle_to_v = to_v.register_forward_hook(to_v_pre_hook, with_kwargs=True)
def get_middle_block_modules(self):
""" Get all attention modules from the middle block
Refere to page 22 of the PAG paper, Appendix A.2
"""
try:
m = shared.sd_model
nlm = m.network_layer_mapping
middle_block_modules = [m for m in nlm.values() if 'middle_block_1_transformer_blocks_0_attn1' in m.network_layer_name and 'CrossAttention' in m.__class__.__name__]
return middle_block_modules
except AttributeError:
logger.exception("AttributeError in get_middle_block_modules", stack_info=True)
return []
except Exception:
logger.exception("Exception in get_middle_block_modules", stack_info=True)
return []
def get_cross_attn_modules(self):
""" Get all cross attention modules """
return self.get_middle_block_modules()
def add_field_cross_attn_modules(self, module, field, value):
""" Add a field to a module if it doesn't exist """
if not hasattr(module, field):
setattr(module, field, value)
def remove_field_cross_attn_modules(self, module, field):
""" Remove a field from a module if it exists """
if hasattr(module, field):
delattr(module, field)
def on_cfg_denoiser_callback(self, params: CFGDenoiserParams, pag_params: PAGStateParams):
# always unhook
self.unhook_callbacks(pag_params)
pag_params.step = params.sampling_step
# CFG Interval
# TODO: set rho based on sdxl or sd1.5
pag_params.current_noise_level = calculate_noise_level(
i = pag_params.step,
N = pag_params.max_sampling_step,
)
if pag_params.cfg_interval_enable:
if pag_params.cfg_interval_schedule != 'Constant':
# Calculate noise interval
start = pag_params.cfg_interval_low
end = pag_params.cfg_interval_high
begin_range = start if start <= end else end
end_range = end if start <= end else start
# Scheduled CFG Value
scheduled_cfg_scale = cfg_scheduler(pag_params.cfg_interval_schedule, pag_params.step, pag_params.max_sampling_step, pag_params.guidance_scale)
pag_params.cfg_interval_scheduled_value = scheduled_cfg_scale if begin_range <= pag_params.current_noise_level <= end_range else 1.0
# Run PAG only if active and within interval
if not pag_params.pag_active or pag_params.pag_scale <= 0:
return
if not pag_params.pag_start_step <= params.sampling_step <= pag_params.pag_end_step or pag_params.pag_scale <= 0:
return
if isinstance(params.text_cond, dict):
text_cond = params.text_cond['crossattn'] # SD XL
pag_params.text_cond = {}
pag_params.text_uncond = {}
for key, value in params.text_cond.items():
pag_params.text_cond[key] = value.clone().detach()
pag_params.text_uncond[key] = value.clone().detach()
else:
text_cond = params.text_cond # SD 1.5
pag_params.text_cond = text_cond.clone().detach()
pag_params.text_uncond = text_cond.clone().detach()
pag_params.x_in = params.x.clone().detach()
pag_params.sigma = params.sigma.clone().detach()
pag_params.image_cond = params.image_cond.clone().detach()
pag_params.denoiser = params.denoiser
pag_params.make_condition_dict = get_make_condition_dict_fn(params.text_uncond)
def on_cfg_denoised_callback(self, params: CFGDenoisedParams, pag_params: PAGStateParams):
""" Callback function for the CFGDenoisedParams
Refer to pg.22 A.2 of the PAG paper for how CFG and PAG combine
"""
# Run only within interval
# Run PAG only if active and within interval
if not pag_params.pag_active or pag_params.pag_scale <= 0:
return
if not pag_params.pag_start_step <= params.sampling_step <= pag_params.pag_end_step or pag_params.pag_scale <= 0:
return
# passed from on_cfg_denoiser_callback
x_in = pag_params.x_in
tensor = pag_params.text_cond
uncond = pag_params.text_uncond
image_cond_in = pag_params.image_cond
sigma_in = pag_params.sigma
# concatenate the conditions
# "modules/sd_samplers_cfg_denoiser.py:237"
cond_in = catenate_conds([tensor, uncond])
make_condition_dict = get_make_condition_dict_fn(uncond)
conds = make_condition_dict(cond_in, image_cond_in)
# set pag_enable to True for the hooked cross attention modules
for module in pag_params.crossattn_modules:
setattr(module, 'pag_enable', True)
# get the PAG guidance (is there a way to optimize this so we don't have to calculate it twice?)
pag_x_out = params.inner_model(x_in, sigma_in, cond=conds)
# update pag_x_out
pag_params.pag_x_out = pag_x_out
# set pag_enable to False
for module in pag_params.crossattn_modules:
setattr(module, 'pag_enable', False)
def cfg_after_cfg_callback(self, params: AfterCFGCallbackParams, pag_params: PAGStateParams):
#self.unhook_callbacks(pag_params)
pass
def get_xyz_axis_options(self) -> dict:
xyz_grid = [x for x in scripts.scripts_data if x.script_class.__module__ in ("xyz_grid.py", "scripts.xyz_grid")][0].module
extra_axis_options = {
xyz_grid.AxisOption("[PAG] Active", str, pag_apply_override('pag_active', boolean=True), choices=xyz_grid.boolean_choice(reverse=True)),
xyz_grid.AxisOption("[PAG] SANF", str, pag_apply_override('pag_sanf', boolean=True), choices=xyz_grid.boolean_choice(reverse=True)),
xyz_grid.AxisOption("[PAG] PAG Scale", float, pag_apply_field("pag_scale")),
xyz_grid.AxisOption("[PAG] PAG Start Step", int, pag_apply_field("pag_start_step")),
xyz_grid.AxisOption("[PAG] PAG End Step", int, pag_apply_field("pag_end_step")),
xyz_grid.AxisOption("[PAG] Enable CFG Scheduler", str, pag_apply_override('cfg_interval_enable', boolean=True), choices=xyz_grid.boolean_choice(reverse=True)),
xyz_grid.AxisOption("[PAG] CFG Noise Interval Low", float, pag_apply_field("cfg_interval_low")),
xyz_grid.AxisOption("[PAG] CFG Noise Interval High", float, pag_apply_field("cfg_interval_high")),
xyz_grid.AxisOption("[PAG] CFG Schedule Type", str, pag_apply_override('cfg_interval_schedule', boolean=False), choices=lambda: SCHEDULES),
#xyz_grid.AxisOption("[PAG] ctnms_alpha", float, pag_apply_field("pag_ctnms_alpha")),
}
return extra_axis_options
def combine_denoised_pass_conds_list(*args, **kwargs):
""" Hijacked function for combine_denoised in CFGDenoiser """
original_func = kwargs.get('original_func', None)
new_params = kwargs.get('pag_params', None)
if new_params is None:
logger.error("new_params is None")
return original_func(*args)
def new_combine_denoised(x_out, conds_list, uncond, cond_scale):
denoised_uncond = x_out[-uncond.shape[0]:]
denoised = torch.clone(denoised_uncond)
noise_level = calculate_noise_level(new_params.step, new_params.max_sampling_step)
# Calculate CFG Scale
cfg_scale = cond_scale
new_params.cfg_interval_scheduled_value = cfg_scale
if new_params.cfg_interval_enable:
if new_params.cfg_interval_schedule != 'Constant':
# Calculate noise interval
start = new_params.cfg_interval_low
end = new_params.cfg_interval_high
begin_range = start if start <= end else end
end_range = end if start <= end else start
# Scheduled CFG Value
scheduled_cfg_scale = cfg_scheduler(new_params.cfg_interval_schedule, new_params.step, new_params.max_sampling_step, cond_scale)
# Only apply CFG in the interval
cfg_scale = scheduled_cfg_scale if begin_range <= noise_level <= end_range else 1.0
new_params.cfg_interval_scheduled_value = scheduled_cfg_scale
# This may be temporarily necessary for compatibility with scfg
# if not new_params.pag_start_step <= new_params.step <= new_params.pag_end_step:
# return original_func(*args)
# This may be temporarily necessary for compatibility with scfg
# if not new_params.pag_start_step <= new_params.step <= new_params.pag_end_step:
# return original_func(*args)
if incantations_debug:
logger.debug(f"Schedule: {new_params.cfg_interval_schedule}, CFG Scale: {cfg_scale}, Noise_level: {round(noise_level,3)}")
for i, conds in enumerate(conds_list):
for cond_index, weight in conds:
if not new_params.cfg_interval_enable:
denoised[i] += (x_out[cond_index] - denoised_uncond[i]) * (weight * cfg_scale)
else:
denoised[i] += (x_out[cond_index] - denoised_uncond[i]) * (weight * cfg_scale)
# Apply PAG guidance only within interval
if not new_params.pag_start_step <= new_params.step <= new_params.pag_end_step or new_params.pag_scale <= 0:
continue
else:
try:
denoised[i] += (x_out[cond_index] - new_params.pag_x_out[i]) * (weight * new_params.pag_scale)
except TypeError:
logger.exception("TypeError in combine_denoised_pass_conds_list")
except IndexError:
logger.exception("IndexError in combine_denoised_pass_conds_list")
#logger.debug(f"added PAG guidance to denoised - pag_scale:{global_scale}")
return denoised
return new_combine_denoised(*args)
# from modules/sd_samplers_cfg_denoiser.py:187-195
def get_make_condition_dict_fn(text_uncond):
if shared.sd_model.model.conditioning_key == "crossattn-adm":
make_condition_dict = lambda c_crossattn, c_adm: {"c_crossattn": [c_crossattn], "c_adm": c_adm}
else:
if isinstance(text_uncond, dict):
make_condition_dict = lambda c_crossattn, c_concat: {**c_crossattn, "c_concat": [c_concat]}
else:
make_condition_dict = lambda c_crossattn, c_concat: {"c_crossattn": [c_crossattn], "c_concat": [c_concat]}
return make_condition_dict
def calculate_noise_level(i, N, sigma_min=0.002, sigma_max=80.0, rho=3):
"""
Calculate the noise level for a given sampling step index.
Parameters:
i (int): Index of the current sampling step (0-based index).
N (int): Total number of sampling steps.
sigma_min (float): Minimum sigma value for min noise level, default 0.002.
sigma_max (float): Maximum sigma value for max noise level, default 80.0.
rho (int): Discretization parameter, default 3 for SD-XL, 7 for EDM2.
Returns:
float: Calculated noise level for the given step.
"""
if i == 0:
return sigma_max
if i >= N:
return 0.0
sigma_max_p = sigma_max ** (1/rho)
sigma_min_p = sigma_min ** (1/rho)
inner_term = sigma_max_p + (i / (N - 1)) * (sigma_min_p - sigma_max_p)
noise_level = inner_term ** rho
return noise_level
def find_closest_index(noise_level: float, N: int, sigma_min=0.002, sigma_max=80.0, rho=3, tol=1e-6):
"""
Given a noise level, find the closest integer index in the range [0, N-1] that corresponds to the noise level.
Parameters:
noise_level (float): Target noise level to find the closest index for.
N (int): Total number of sampling steps.
sigma_min (float): Minimum sigma value for min noise level, default 0.002.
sigma_max (float): Maximum sigma value for max noise level, default 80.0.
rho (int): Discretization parameter, default 3 for SD-XL, 7 for EDM2.
Returns:
int: The closest index to the specified noise level.
"""
# Min/max noise levels for the given range
if noise_level <= sigma_min:
return N
if noise_level >= sigma_max:
return 0
#return N - 1
low, high = 0, N - 1
while low <= high:
mid = (low + high) // 2
mid_nl = calculate_noise_level(mid, N)
if abs(mid_nl - noise_level) < tol:
return mid
elif mid_nl < noise_level:
high = mid - 1
else:
low = mid + 1
# If exact match not found, return the index with noise level closest to the target
return low if abs(calculate_noise_level(low, N) - noise_level) < abs(calculate_noise_level(high, N) - noise_level) else high
### CFG Schedulers
# TODO: Refactor this into something cleaner
def cfg_scheduler(schedule: str, step: int, max_steps: int, w0: float) -> float:
"""
Constant scheduler for CFG guidance weight.
Parameters:
step (int): Current sampling step.
max_steps (int): Total number of sampling steps.
w0 (float): Constant value for the guidance weight.
Returns:
float: Scheduled guidance weight value.
"""
match schedule:
case 'Constant':
return constant_schedule(step, max_steps, w0)
case 'Linear':
return linear_schedule(step, max_steps, w0)
case 'Clamp-Linear (c=4.0)':
return clamp_linear_schedule(step, max_steps, w0, 4.0)
case 'Clamp-Linear (c=2.0)':
return clamp_linear_schedule(step, max_steps, w0, 2.0)
case 'Clamp-Linear (c=1.0)':
return clamp_linear_schedule(step, max_steps, w0, 1.0)
case 'Inverse-Linear':
return invlinear_schedule(step, max_steps, w0)
case 'PCS (s=0.01)':
return powered_cosine_schedule(step, max_steps, w0, 0.01)
case 'PCS (s=0.1)':
return powered_cosine_schedule(step, max_steps, w0, 0.1)
case 'PCS (s=1.0)':
return powered_cosine_schedule(step, max_steps, w0, 1.0)
case 'PCS (s=2.0)':
return powered_cosine_schedule(step, max_steps, w0, 2.0)
case 'PCS (s=4.0)':
return powered_cosine_schedule(step, max_steps, w0, 4.0)
case 'Clamp-Cosine (c=4.0)':
return clamp_cosine_schedule(step, max_steps, w0, 4.0)
case 'Clamp-Cosine (c=2.0)':
return clamp_cosine_schedule(step, max_steps, w0, 2.0)
case 'Clamp-Cosine (c=1.0)':
return clamp_cosine_schedule(step, max_steps, w0, 1.0)
case 'Cosine':
return cosine_schedule(step, max_steps, w0)
case 'Sine':
return sine_schedule(step, max_steps, w0)
case 'V-Shape':
return v_shape_schedule(step, max_steps, w0)
case 'A-Shape':
return a_shape_schedule(step, max_steps, w0)
case 'Interval':
return interval_schedule(step, max_steps, w0, 0.25, 5.42)
case _:
logger.error(f"Invalid CFG schedule: {schedule}")
return constant_schedule(step, max_steps, w0)
def constant_schedule(step: int, max_steps: int, w0: float):
"""
Constant scheduler for CFG guidance weight.
"""
return w0
def linear_schedule(step: int, max_steps: int, w0: float):
"""
Normalized linear scheduler for CFG guidance weight.
Such that integral 0-> T ~ w(t) dt = w*T
"""
# return w0 * (1 - step / max_steps)
return w0 * 2 * (1 - step / max_steps)
def clamp_linear_schedule(step: int, max_steps: int, w0: float, c: float):
"""
Normalized clamp-linear scheduler for CFG guidance weight.
"""
return max(c, linear_schedule(step, max_steps, w0))
def clamp_cosine_schedule(step: int, max_steps: int, w0: float, c: float):
"""
Normalized clamp-cosine scheduler for CFG guidance weight.
"""
return max(c, cosine_schedule(step, max_steps, w0))
def invlinear_schedule(step: int, max_steps: int, w0: float):
"""
Normalized inverse linear scheduler for CFG guidance weight.
"""
# return w0 * (step / max_steps)
return w0 * 2 * (step / max_steps)
def powered_cosine_schedule(step: int, max_steps: int, w0: float, s: float):
"""
Normalized cosine scheduler for CFG guidance weight.
"""
return w0 * ((1 - math.cos(math.pi * ((max_steps - step) / max_steps)**s))/2.0)
def cosine_schedule(step: int, max_steps: int, w0: float):
"""
Normalized cosine scheduler for CFG guidance weight.
"""
return w0 * (1 + math.cos(math.pi * step / max_steps))
def sine_schedule(step: int, max_steps: int, w0: float):
"""
Normalized sine scheduler for CFG guidance weight.
"""
return w0 * (math.sin((math.pi * step / max_steps) - (math.pi / 2)) + 1)
def v_shape_schedule(step: int, max_steps: int, w0: float):
"""
Normalized V-shape scheduler for CFG guidance weight.
"""
if step < max_steps / 2:
return invlinear_schedule(step, max_steps, w0)
return linear_schedule(step, max_steps, w0)
def a_shape_schedule(step: int, max_steps: int, w0: float):
"""
Normalized A-shape scheduler for CFG guidance weight.
"""
if step < max_steps / 2:
return linear_schedule(step, max_steps, w0)
return invlinear_schedule(step, max_steps, w0)
def interval_schedule(step: int, max_steps: int, w0: float, low: float, high: float):
"""
Normalized interval scheduler for CFG guidance weight.
"""
if low <= step <= high:
return w0
return 1.0
# XYZ Plot
# Based on @mcmonkey4eva's XYZ Plot implementation here: https://github.com/mcmonkeyprojects/sd-dynamic-thresholding/blob/master/scripts/dynamic_thresholding.py
def pag_apply_override(field, boolean: bool = False):
def fun(p, x, xs):
if boolean:
x = True if x.lower() == "true" else False
setattr(p, field, x)
if not hasattr(p, "pag_active"):
setattr(p, "pag_active", True)
if 'cfg_interval_' in field and not hasattr(p, "cfg_interval_enable"):
setattr(p, "cfg_interval_enable", True)
return fun
def pag_apply_field(field):
def fun(p, x, xs):
if not hasattr(p, "pag_active"):
setattr(p, "pag_active", True)
setattr(p, field, x)
return fun
# thanks torch; removing hooks DOESN'T WORK
# thank you to @ProGamerGov for this https://github.com/pytorch/pytorch/issues/70455
def _remove_all_forward_hooks(
module: torch.nn.Module, hook_fn_name: Optional[str] = None
) -> None:
"""
This function removes all forward hooks in the specified module, without requiring
any hook handles. This lets us clean up & remove any hooks that weren't property
deleted.
Warning: Various PyTorch modules and systems make use of hooks, and thus extreme
caution should be exercised when removing all hooks. Users are recommended to give
their hook function a unique name that can be used to safely identify and remove
the target forward hooks.
Args:
module (nn.Module): The module instance to remove forward hooks from.
hook_fn_name (str, optional): Optionally only remove specific forward hooks
based on their function's __name__ attribute.
Default: None
"""
if hook_fn_name is None:
warn("Removing all active hooks can break some PyTorch modules & systems.")
def _remove_hooks(m: torch.nn.Module, name: Optional[str] = None) -> None:
if hasattr(module, "_forward_hooks"):
if m._forward_hooks != OrderedDict():
if name is not None:
dict_items = list(m._forward_hooks.items())
m._forward_hooks = OrderedDict(
[(i, fn) for i, fn in dict_items if fn.__name__ != name]
)
else:
m._forward_hooks: Dict[int, Callable] = OrderedDict()
def _remove_child_hooks(
target_module: torch.nn.Module, hook_name: Optional[str] = None
) -> None:
for name, child in target_module._modules.items():
if child is not None:
_remove_hooks(child, hook_name)
_remove_child_hooks(child, hook_name)
# Remove hooks from target submodules
_remove_child_hooks(module, hook_fn_name)
# Remove hooks from the target module
_remove_hooks(module, hook_fn_name)
|