File size: 48,409 Bytes
f4a41d8 |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 |
# ,, ,, ,,
# .M"""bgd mm `7MM `7MM"""Mq. db `7MM
# ,MI "Y MM MM MM `MM. MM
# `MMb. mmMMmm `7M' `MF'MM .gP"Ya MM ,M9 `7MM MM .gP"Ya
# `YMMNq. MM VA ,V MM ,M' Yb MMmmdM9 MM MM ,M' Yb
# . `MM MM VA ,V MM 8M"""""" MM MM MM 8M""""""
# Mb dM MM VVV MM YM. , MM MM MM YM. ,
# P"Ybmmd" `Mbmo ,V .JMML.`Mbmmd'.JMML. .JMML..JMML.`Mbmmd'
# ,V
# OOb"
#
# A helper script for AUTOMATIC1111/stable-diffusion-webui.
# Enter your keywords and let the selections help you determine the look.
# https://replicate.com/methexis-inc/img2prompt has been an incredible help for improving the prompts.
# https://docs.google.com/document/d/1ZtNwY1PragKITY0F4R-f8CarwHojc9Wrf37d0NONHDg/ has been equally super important.
# Thanks to https://github.com/xram64 for helping fix the interface
# Art movements from https://en.wikipedia.org/wiki/List_of_art_movements I threw out the ones that did not work
# Portrait prompt - Portrait of an attractive young lady,flower field background, (by artist [X]:1.3), square ratio
# Negative - missing limbs, extra limbs
# Landscape prompt - Small house in the middle of a forest,near a lake
# Action prompt - Astronaut floating in space,firing laser at alien ship,galaxy background
# Negatives - watermark,label,text
# 20 steps on Euler A
# Seed - 669
import copy
import os
import random
from os import listdir, path
from os.path import isfile, join
import modules.scripts as scripts
import gradio as gr
from modules.processing import Processed, process_images
from modules.shared import cmd_opts, opts, state
from modules import scripts
# =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=->
ResourceDir = os.path.join(scripts.basedir(), f"StylePile/")
def FilesInFolder(SourceFolder):
return [file for file in os.listdir(SourceFolder)]
def FilesInFolderFullPath(SourceFolder):
return [SourceFolder + file for file in os.listdir(SourceFolder)]
ElementConcept = [
"Acclaimed",
"Alternative",
"Amateur",
"Artificial",
"Award Winning",
"Basic",
"Beginner",
"Bipolar",
"Boyish",
"Childish",
"Cinematic",
"Clever",
"Clumsy",
"Cognitive",
"Complex",
"Compressed",
"Controllable",
"Corrupted",
"Damaged",
"Destroyed",
"Disgusting",
"Divisive",
"Dramatic",
"Dumb",
"Eliminated",
"Excessive",
"Exciting",
"Extreme",
"Feminine",
"Filtered",
"Fixated",
"Fixed",
"Foolish",
"Fragile",
"Girlish",
"Gorgeous",
"Groundbreaking",
"Hated",
"Hidden",
"Highly Rated",
"Horrifying",
"Imaginary",
"Imaginative",
"Imitated",
"Jaded",
"Light hearted",
"Loved",
"Low Rated",
"Magical",
"Masculine",
"Masterful",
"Masterpiece",
"Maximalist",
"Methodological",
"Misunderstood",
"Mundane",
"Overprocessed",
"Pathetic",
"Photoshopped",
"Preview",
"Raw",
"Recycled",
"Religious",
"Rough",
"Sacrificial",
"Sacrilegious",
"Schematic",
"Simple",
"Sophisticated",
"Stupid",
"Trustworthy",
"Unbelievable",
"Understandable",
"Unearthed",
"Unfiltered",
"Unfinished",
"Unhinged",
"Universal",
"Unsuccessful",
"Venerable",
"Visionary",
"Vivacious"
]
ResultConcept = ["Not set","Random"] + ElementConcept
ResultNames = [
"Photo",
"Digital Artwork",
"3D Rendering",
"Painting",
"Drawing",
"Vector Art"
]
ResultTypeBefore = {
"Photo": "Photo",
"Digital Artwork": "Digital Artwork",
"3D Rendering": "Professional 3D rendering",
"Painting": "Painting",
"Drawing": "Drawing",
"Vector Art": "Vector image"
}
#"3D Rendering": ",Highly detailed,Art by senior Artist,Polycount,trending on CGSociety,trending on ArtStation",
#"Photo": ",HD,4K,8K,highly detailed,Sharp,Photo-realism,Professional photograph,Masterpiece",
ResultTypePositives = {
"Photo": ",Highly Detailed",
"Digital Artwork": ",CGSociety,ArtStation",
"3D Rendering": ",CGSociety,ArtStation",
"Painting": " ",
"Drawing": " ",
"Vector Art": ",(Flat style:1.3),Illustration,Behance"
}
ResultTypeNegatives = {
"Photo": ",Amateur,Low rated,Phone,Wedding,Frame,Painting,tumblr",
"Digital Artwork": ",Scribbles,Low quality,Low rated,Mediocre,3D rendering,Screenshot,Software,UI",
"3D Rendering": ",((Wireframe)),Polygons,Screenshot,Character design,Software,UI",
"Painting": "Low quality,Bad composition,Faded,(Photo:1.5),(Frame:1.3)",
"Drawing": ",Low quality,Photo,Artifacts,Table,Paper,Pencils,Pages,Wall",
"Vector Art": ",(Watermark:1.5),(Text:1.3)"
}
ResultType = {
"Not set": "",
"Random": "Random",
}
ResultType.update(ResultTypePositives)
# =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=->
ResultDirectionList = FilesInFolder(ResourceDir + "Directions/")
ResultDirectionList = list(
map(lambda x: x.replace(".jpg", ""), ResultDirectionList))
ResultDirection = ["Not set", "Random"] + ResultDirectionList
ResultDirectionImages = FilesInFolderFullPath(ResourceDir + "Directions/")
ResultDirectionImages = list(
map(lambda x: x.replace("\\", "/"), ResultDirectionImages))
# =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=->
ResultMoodList = FilesInFolder(ResourceDir + "Moods/")
ResultMoodList = list(map(lambda x: x.replace(".jpg", ""), ResultMoodList))
ResultMood = ["Not set", "Random"] + ResultMoodList
ResultMoodImages = FilesInFolderFullPath(ResourceDir + "Moods/")
ResultMoodImages = list(map(lambda x: x.replace("\\", "/"), ResultMoodImages))
# =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=->
ResultArtistList = FilesInFolder(ResourceDir + "Artists/")
ResultArtistList = list(map(lambda x: x.replace(".jpg", ""), ResultArtistList))
Artists = ["Not set", "Random"] + ResultArtistList
ResultArtistImages = FilesInFolderFullPath(ResourceDir + "Artists/")
ResultArtistImages = list(
map(lambda x: x.replace("\\", "/"), ResultArtistImages))
#ResultArtistColumn = "\n".join(str(item) for item in ResultArtistList)
# =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=->
ArtMovementList = FilesInFolder(ResourceDir + "Art Movements/")
ArtMovementList = list(map(lambda x: x.replace(".jpg", ""), ArtMovementList))
ArtMovements = ["Not set", "Random"] + ArtMovementList
ArtMovementImages = FilesInFolderFullPath(ResourceDir + "Art Movements/")
ArtMovementImages = list(
map(lambda x: x.replace("\\", "/"), ArtMovementImages))
# =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=->
ResultColorList = FilesInFolder(ResourceDir + "Colors/")
ResultColorList = list(map(lambda x: x.replace(".jpg", ""), ResultColorList))
ResultColor = ["Not set", "Random"] + ResultColorList
ResultColorImages = FilesInFolderFullPath(ResourceDir + "Colors/")
ResultColorImages = list(
map(lambda x: x.replace("\\", "/"), ResultColorImages))
# =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=->
with open(ResourceDir + "Inspiration/Subjects.txt", 'r+') as tf:
Subjects = [line.rstrip() for line in tf]
with open(ResourceDir + "Inspiration/Actions.txt", 'r+') as tf:
Actions = [line.rstrip() for line in tf]
with open(ResourceDir + "Inspiration/Locations.txt", 'r+') as tf:
Locations = [line.rstrip() for line in tf]
# =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=->
TipsAndTricks = [
"Be specific. Add details in sequence, separated by commas. Like 'Cute black kitten, yellow eyes' + selecting Photography as image type, will get you to the result you want faster than just 'black kitten'.",
"Try out random values. For example, set artist to random, crank up the batch count and enjoy the show. Mix and match random settings.",
"If you add your own artist, I would recommend having 'by Artist' in front of their name. Depending on their popularity (or lack thereof) this appears to have a very tangible influence on the result.",
"Mix and match artists from the dropdowns (or type your own) for some interesting results. Having one artist selected and other(s) set to random is also a nice way to find new looks with some amount of predictability.",
"Parenthesis can be added to make parts of the prompt stronger. So ((cute kitten)) will make it extra cute (try it out). This is also important if a style is affecting your original prompt too much. Make that prompt stronger by adding parenthesis around it, like this: ((promt)). A strength modifier value can also be used, like this (prompt:1.1).",
"Prompts can be split like [A|B] to sequentially use terms, one after another on each step. For example **[cat|dog]** will produce a hybrid catdog.",
"Using **[A:B:0.4]** will switch to other terms after the first one has been active for a certain percentage of steps. So [cat:dog:0.4] will build a cat 40% of the time and then start turning it into a dog. Usually this needs more steps to work properly.",
"During long cold winter nights, you can turn your PC into a heater by generating hundreds of images non-stop.",
"Feel free to share feedback and ideas on github: https://github.com/some9000/StylePile",
"Some things work together, others don't. Like Photo doesn't work too great with many Art movements or Drawing will not become photorealistic just because that was selected."
]
# =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=->
Preset = {
"None": "",
"Portraits": ",(close portrait:1.3),thematic background",
"Feminine portrait": ",(close portrait:1.3),(Feminine:1.4),(beautiful:1.4),(attractive:1.3),handsome,calendar pose,perfectly detailed eyes,studio lighting,thematic background",
"Masculine portrait": ",(close portrait:1.3),(Masculine:1.4),attractive,handsome,calendar pose,perfectly detailed eyes,studio lighting,thematic background",
"WaiFusion": ",close portrait,(manga:1.3),beautiful,attractive,handsome,trending on ArtStation,DeviantArt contest winner,CGSociety,ultrafine,detailed,studio lighting",
"Horrible Monsters": ",monster,ugly,surgery,evisceration,morbid,cut,open,rotten,mutilated,deformed,disfigured,malformed,missing limbs,extra limbs,bloody,slimy,goo,Richard Estes,Audrey Flack,Ralph Goings,Robert Bechtle,Tomasz Alen Kopera,H.R.Giger,Joel Boucquemont,ArtStation,DeviantArt contest winner,thematic background",
"Robots": ",robot,((cyborg)),machine,futuristic,concept Art by senior character Artist,featured on zbrush central,trending on polycount,trending on ArtStation,CGSociety,hard surface modeling",
"Retrofuturism": ",((retrofuturism)),(science fiction),dystopian Art,ultrafine,detailed,future tech,by Clarence Holbrook CArter,by Ed Emshwiller,CGSociety,ArtStation contest winner,trending on ArtStation,DeviantArt contest winner,Fallout",
"Propaganda": ",propaganda poster,soviet poster,sovietwave",
"Landscapes": ",naturalism,land Art,regionalism,shutterstock contest winner,trending on unsplash,featured on Flickr"
}
PresetNegatives = {
"None": "",
"Portraits": ",robot eyes,distorted pupils,distorted eyes,Unnatural anatomy,strange anatomy,things on face",
"Feminine portrait": ",robot eyes,distorted pupils,distorted eyes,Unnatural anatomy,strange anatomy,things on face",
"Masculine portrait": ",robot eyes,distorted pupils,distorted eyes,Unnatural anatomy,strange anatomy,things on face",
"WaiFusion": ",things on face,Unnatural anatomy,strange anatomy",
"Horrible Monsters": ",(attractive),pretty,smooth,cArtoon,pixar,human",
"Robots": ",cartoon",
"Retrofuturism": ",extra limbs,malformed limbs,modern",
"Propaganda": ",extra limbs,malformed limbs,modern",
"Landscapes": ",((hdr)),((terragen)),((rendering)),(high contrast)"
}
# =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=->
# At some point in time it looked like adding a bunch of these negative prompts helps,but now I am not so sure...
AlwaysBad = ",watermark,signature"
class Script(scripts.Script):
txt2img_prompt = None
img2img_prompt = None
batch_count = None
batch_size = None
steps = None
def after_component(self, component, **kwargs):
if kwargs.get('elem_id') == 'txt2img_prompt':
self.txt2img_prompt = component
if kwargs.get('elem_id') == 'img2img_prompt':
self.img2img_prompt = component
# if kwargs.get('elem_id') == 'batch_count':
# self.batch_count = component
# if kwargs.get('elem_id') == 'batch_size':
# self.batch_size = component
# if kwargs.get('elem_id') == 'steps':
# self.steps = component
def title(self):
return "StylePile"
def show(self, is_img2img):
return True
def ui(self, is_img2img):
with gr.Tab("Parameters"):
with gr.Row():
ddResultConcept = gr.Dropdown(
ResultConcept, label="Conceptually", value="Not set")
ddCoreResultType = gr.Dropdown(
list(ResultType.keys()), label="Image type", value="Not set")
slResultTypeStrength = gr.Slider(
0, 2, value=1.3, step=0.05, show_label=False)
with gr.Row():
with gr.Column():
ddResultDirection = gr.Dropdown(
ResultDirection, label="Direction", value="Not set")
slResultDirectionStrength = gr.Slider(
0, 2, value=1.3, step=0.05, show_label=False)
with gr.Column():
ddResultMood = gr.Dropdown(
ResultMood, label="Mood", value="Not set")
slResultMoodStrength = gr.Slider(
0, 2, value=1.3, step=0.05, show_label=False)
with gr.Column():
ddResultColor = gr.Dropdown(
ResultColor, label="Colors", value="Not set")
slResultColorStrength = gr.Slider(
0, 2, value=1.3, step=0.05, show_label=False)
with gr.Row():
cbChangeCount = gr.Checkbox(
value=True, label="Set batch count to prompt count")
cbIncreaseSeed = gr.Checkbox(
value=True, label="Increase seed with batch size")
cbShowTips = gr.Checkbox(
value=False, label="Show tips when generating")
ddPreset = gr.Dropdown(list(Preset.keys()), label="Style influence (incomplete)", value="None")
with gr.Row():
strSequentialPrompt = gr.Textbox(
lines=3, label="Sequential prompts [X]", placeholder="Insert [X] anywhere in main prompt to sequentially insert values from here. Random values will be added here or to main prompt.")
with gr.Row():
strSubSequentialPrompt = gr.Textbox(
lines=3, label="SubSequential prompts [Y]", placeholder="Insert [Y] in the final prompt <== to sequentially insert values from here (and increase prompt count). This is done after all other prompts and loops through all lines.")
with gr.Row():
strRandomPromptA = gr.Textbox(
lines=3, label="Random [A]", placeholder="Insert [A] anywhere in main prompt (or [X] prompt) to randomly insert values from here.")
strRandomPromptB = gr.Textbox(
lines=3, label="Random [B]", placeholder="Insert [B] anywhere in main prompt (or [X] prompt) to randomly insert values from here.")
strRandomPromptC = gr.Textbox(
lines=3, label="Random [C]", placeholder="Insert [C] anywhere in main prompt (or [X] prompt) to randomly insert values from here.")
with gr.Row():
with gr.Column():
selArtistA = gr.Dropdown(Artists, label="Artist", value="Not set")
sliImageArtistStrengthA = gr.Slider(0, 2, value=1.3, step=0.05, label="Influence")
selArtistB = gr.Dropdown(Artists, label="Artist", value="Not set")
sliImageArtistStrengthB = gr.Slider(0, 2, value=1.3, step=0.05, label="Influence")
selArtistC = gr.Dropdown(Artists, label="Artist", value="Not set")
sliImageArtistStrengthC = gr.Slider(0, 2, value=1.3, step=0.05, label="Influence")
with gr.Column():
selArtMovementA = gr.Dropdown(ArtMovements, label="Art movement", value="Not set")
selArtMovementStrengthA = gr.Slider(0, 2, value=1.3, step=0.05, label="Influence")
selArtMovementB = gr.Dropdown(ArtMovements, label="Art movement", value="Not set")
selArtMovementStrengthB = gr.Slider(0, 2, value=1.3, step=0.05, label="Influence")
selArtMovementC = gr.Dropdown(ArtMovements, label="Art movement", value="Not set")
selArtMovementStrengthC = gr.Slider(0, 2, value=1.3, step=0.05, label="Influence")
if self.txt2img_prompt is not None:
with gr.Row():
bTestPrompt = gr.Button('Insert default prompt')
#if self.txt2img_prompt is not None:
bTestPrompt.click(fn=lambda x: "Portrait of an attractive young lady,flower field background, square ratio",
inputs = [self.txt2img_prompt],
outputs = [self.txt2img_prompt])
bInspireMe = gr.Button('Inspire me, StylePile')
#if self.txt2img_prompt is not None:
bInspireMe.click(fn=lambda x: random.choice(Subjects) +","+ random.choice(Actions) +","+ random.choice(Locations)+",",
inputs = [self.txt2img_prompt],
outputs = [self.txt2img_prompt])
# with gr.Row():
# b4images = gr.Button('4 images')
# b8images = gr.Button('8 images')
# b4x4images = gr.Button('4x4 images')
# bMorphImages = gr.Button('4@40 images')
# bStandardPreview = gr.Button('Preview')
# b4images.click(fn = lambda p:40, inputs = [self.steps],outputs = [self.steps])
# b8images.click(fn = lambda p:8, inputs = [self.batch_count],outputs = [self.batch_count])
# b8images.click(fn = lambda p:8, inputs = [self.batch_size],outputs = [self.batch_size])
with gr.Tab("Directions") as StyleTab:
gr.Gallery(value=ResultDirectionImages, show_label=False).style(
grid=(3, 3, 3, 3, 4, 4), container=False)
with gr.Tab("Moods"):
gr.Gallery(value=ResultMoodImages, show_label=False).style(
grid=(3, 3, 3, 3, 4, 4), container=False)
with gr.Tab("Artists"):
gr.Gallery(value=ResultArtistImages, show_label=False).style(
grid=(2, 2, 2, 2, 3, 3), container=False)
with gr.Tab("Art movements"):
gr.Gallery(value=ArtMovementImages, show_label=False).style(
grid=(3, 3, 3, 3, 4, 4), container=False)
with gr.Tab("Colors"):
gr.Gallery(value=ResultColorImages, show_label=False).style(
grid=(3, 3, 3, 3, 4, 4), container=False)
with gr.Tab("Tools & Info") as HelpTab:
with gr.Row():
gr.Markdown(
"""
### Tips and tricks
If you add your own Artist, I would recommend having **by Artist** in front of their name. Depending on their popularity (or lack thereof) this appears to have a very tangible influence on the result. In general, most of the elements that influence the look appear to work best with a certain strength boost, hence the 1.3 default values.
Another thing to keep in mind is relationships between keywords and type of content. For example, if you want a reasonably realistic looking image of an alien cyborg. Selecting **Photo** will mostly produce fairly clumsy results. But, if you select **3D rendering** and **Realistic, Ultrarealistic** or **Ultra detailed** as direction, the result may actually be closer to what you expect. The opposite is true as well. There are certain things that you will not get to look realistic no matter what the modifiers are if Image type is not set to **Photo**. Try kittens.
In general just experiment with **Image type** and **Direction**. An easy way to do it is selecting random settings, a high batch count and then checking the keywords on the results you like.
### Modifiers
Elements of the prompt can be modified to have a certain strength or change over time. Normally you do this by typing into the prompt, but here I have added tools that will actually insert pre-formatted text so it is easier to understand what it should look like. Note that it doesn't have to be a single word, it is a part of the prompt, so it can be several words or a full sentence. Also note that it will be added to the end of the prompt no matter where the cursor was due to limitations of gradio.
"""
)
with gr.Row():
with gr.Column():
gr.Markdown(
"""
A strength modifier value can be added to parts of the prompt like this **(A:1.3)** < this part would be about 30% stronger. To save some typing you can select the line you want to make stronger and use **Ctrl+Shift+Arrow keys up** or **down** to add these parenthesis and change the value. 1.3 seems like a good starting point if you want to see some impact. Interestingly, adding **very** as a keyword may have a similar or even stronger effect.
"""
)
with gr.Column():
tbAdjustStrength = gr.Textbox(label="Adjust strength", placeholder="Enter prompt here")
sbAdjustStrength = gr.Slider(0.1, 2.0, value=1.3, step=0.1, label="Strength")
bAdjustStrength = gr.Button('Insert')
with gr.Row():
with gr.Column():
tbMorphFrom = gr.Textbox(label="Morph from", placeholder="Prompt A")
tbMorphTo = gr.Textbox(label="Morph to", placeholder="Prompt B")
sbMorphStrength = gr.Slider(0.05, 0.95, value=0.5, step=0.05, label="Starting point")
bInsertMorph = gr.Button('Insert')
with gr.Column():
gr.Markdown(
"""
You can start with a prompt element and then, after a certain percentage of steps, start converting this prompt into something else. Basically it looks like [A:B:0.5] with A being the first part to do, B being what it should be morphing into and 0.5 representing a percentage of when it should start the conversion process. Thus in case of 0.5 that is 50% of the whole process.
""")
with gr.Row():
with gr.Column():
gr.Markdown(
"""
You can mix two prompt elements where each step they get swapped. It looks like [A|B] thus processing A each odd step and B each even step.
""")
with gr.Column():
tbBounceFrom = gr.Textbox(label="Bounce from", placeholder="Prompt A")
rbBounceTo = gr.Textbox(label="Bounce to", placeholder="Prompt B")
bBounce = gr.Button('Insert')
with gr.Row():
gr.Markdown(
"""
These last two sections appear to benefit from increasing sampling steps and CFG scale.
### Example images, adding your own selections to dropdowns
Example images stored in the script folders are more than just images. Their filenames are used to create the **Direction**, **Mood**, **Artist** and Art **movement** dropdown selections. This gives you the ability to Add/Remove parameters as you wish. Just place an image in the folder and name it as the option you want to see in the dropdown. Delete image file to remove that option.
In case you would like to suggest an artist be added to the roster, I would recommend making 8+ sample images first. To see if SD actually "knows" that artist and their style appears unique enough. The portraits you can see in the info pages were generated with the following settings:
### Sample portrait prompt
Positive: Portrait of an attractive young lady,flower field background,(by [X]:1.3), square ratio
Negative - missing limbs, extra limbs, watermark,label,text
[X] is **Artist Name Surname** From my research adding **Artist** can really help to get the correct look.
20 steps on Euler A
Seed - 669 - batch of 4 images
Generally that produces a fairly nice portrait with enough room to show off the given style. Do compare the results to the actual style. As SD will produce something it 'thinks' may be correct based on their name (guessing nationality, basing it on something that has mentioned a similar name etc) and that influences the results, but not in a good way.
""")
if self.txt2img_prompt is not None:
bAdjustStrength.click(fn=lambda p,x,y: p + "(" + x + ":" + str(y) + ")",
inputs = [self.txt2img_prompt,tbAdjustStrength,sbAdjustStrength],
outputs = [self.txt2img_prompt])
bInsertMorph.click(fn=lambda p,x,y,z: p + "[" + x + ":" + y + ":" + str(z) + "]",
inputs = [self.txt2img_prompt,tbMorphFrom,tbMorphTo,sbMorphStrength],
outputs = [self.txt2img_prompt])
bBounce.click(fn=lambda p,x,y: p + "[" + x + "|" + y + "]",
inputs = [self.txt2img_prompt,tbBounceFrom,rbBounceTo],
outputs = [self.txt2img_prompt])
if self.img2img_prompt is not None:
bAdjustStrength.click(fn=lambda p,x,y: p + "(" + x + ":" + str(y) + ")",
inputs = [self.img2img_prompt,tbAdjustStrength,sbAdjustStrength],
outputs = [self.img2img_prompt])
bInsertMorph.click(fn=lambda p,x,y,z: p + "[" + x + ":" + y + ":" + str(z) + "]",
inputs = [self.img2img_prompt,tbMorphFrom,tbMorphTo,sbMorphStrength],
outputs = [self.img2img_prompt])
bBounce.click(fn=lambda p,x,y: p + "[" + x + "|" + y + "]",
inputs = [self.img2img_prompt,tbBounceFrom,rbBounceTo],
outputs = [self.img2img_prompt])
with gr.Tab("Help"):
gr.Markdown(
"""
## Hello, StylePile here
### Introduction
**StylePile** is a mix and match system for adding elements to prompts that affect the style of the result. Hence the name. By default, these elements are placed in a specific order and given strength values. Which means the result sort-of evolves. I have generated thousands of images for each main **Image type** and tweaked the keywords to attempt giving expected results most of the time. Certainly, your suggestions for improvements are very welcome.
### Base workflow
You select extra settings in this script and then hit the standard orange **Generate** button to get results.
For example, if you select the **Painting** image type, then almost all results will look like Paintings. Selecting **Mood** will have a certain influence on the overall look in some way (if it's something humanoid it may show emotion, but also colors and overall feel may change). Setting **Colors** will change the general tonality of the result. And setting **View** will attempt to change how the subject is viewed. Attempt, because view appears to be the least reliable keyword. These elements are placed in order of influence and supported by certain strength values. These basic settings produce very quick results close to the general look you want.
![]({path.join(ResourceDir,"Artists.jpg") ''})
Moving on, adding a **Art movement** will combine with **Image type** to influence how the result generally looks. These styles are based on classic and modern Painting/Art/design movements (which I picked after hours and thousands of samples of testing) and can have a strong influence on the end result. Either it will be more realistic or artistic, or look like a comic book etc. In general, this is a really strong element for getting the look you want. Its influence can be adjusted with the slider above. Experiment with the values, keeping in mind that anything above 1.5 will start becoming a mess. In a similar way, but more focused, you can select an **Artist** and, of course, that will have a very visible effect on the result as well. Currently there are 135 artists, 55 art styles and 25 emotions available for selection and represented with preview images.
Strength of these settings has been preset at 1.3, as that appears to be the golden ratio for getting good results. Sometimes very low settings have an interesting result as well. You can, and should, freely mix and match these settings to get different results. Classic Painting styles affected or affecting 3D look quite interesting. Photography can look cool with some of the brighter, more artistic styles etc. Sometimes raising CFG scale to 15,20 or more also helps to REALLY push the style onto the image.
### Advanced workflow
StylePile can overtake the generation process, allowing you to generate a large amount of different results with very little extra work. There are two types of variables you can use: [X] and [R]. When you add an [X] to your prompt, it sequentially takes values from the **Sequential prompts** text area. You can have dozens of lines there and they will be processed in sequence. When you add [R] to the prompt a value from the **Random** text area will be inserted in its place. By combining these a huge variety in prompts is very easy to do.
When using this, **Batch count** will move through the prompts and **Batch size** will set how many copies with the given prompt to make. If the seed is not random, it will increase with each batch size step. Any random elements will still be picked randomly.
### In conclusion
I made this because manually changing keywords, looking up possible styles, etc was a pain. It is meant as a fun tool to explore possibilities and make learning Stable Diffusion easier. If you have some ideas or, better yet, would like to contribute in some way*, just visit https://github.com/some9000/StylePile
*Hey, if you have a 12Gb graphics card just laying around I'm happy to take it (:
""")
return [ddResultConcept,
cbChangeCount,
cbIncreaseSeed,
strSequentialPrompt,
strSubSequentialPrompt,
strRandomPromptA,
strRandomPromptB,
strRandomPromptC,
slResultTypeStrength,
ddCoreResultType,
ddResultDirection,
slResultDirectionStrength,
ddResultMood,
slResultMoodStrength,
ddResultColor,
slResultColorStrength,
selArtMovementStrengthA,
selArtMovementA,
selArtMovementStrengthB,
selArtMovementB,
selArtMovementStrengthC,
selArtMovementC,
sliImageArtistStrengthA,
selArtistA,
sliImageArtistStrengthB,
selArtistB,
sliImageArtistStrengthC,
selArtistC,
cbShowTips,
ddPreset
]
# =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=->
def run(self, p,
ddResultConcept,
cbChangeCount,
cbIncreaseSeed,
strSequentialPrompt: str,
strSubSequentialPrompt: str,
strRandomPromptA: str,
strRandomPromptB: str,
strRandomPromptC: str,
slResultTypeStrength,
ddCoreResultType,
ddResultDirection,
slResultDirectionStrength,
ddResultMood,
slResultMoodStrength,
ddResultColor,
slResultColorStrength,
selArtMovementStrengthA,
selArtMovementA,
selArtMovementStrengthB,
selArtMovementB,
selArtMovementStrengthC,
selArtMovementC,
sliImageArtistStrengthA,
selArtistA,
sliImageArtistStrengthB,
selArtistB,
sliImageArtistStrengthC,
selArtistC,
cbShowTips,
ddPreset
):
# If it's all empty just exit function.
if len(p.prompt) == 0:
print(
f"\nEmpty prompt! It helps to have at least some guidance for SD. Remember to insert an [X], [A] or [B] into main prompt if you want to use variable values.")
return
# Batch lines present?
BatchLines = [x.strip() for x in strSequentialPrompt.splitlines()]
LineCount = len(BatchLines)
SubBatchLines = [x.strip() for x in strSubSequentialPrompt.splitlines()]
SubLineCount = len(SubBatchLines)
TempText = ""
SubTempText = ""
images = []
seeds = []
prompts = []
infotexts = []
# Overtake amounts of things to generate so we can go through different variables
MainJobCount = p.n_iter
p.n_iter = 1
SubIterationCount = p.batch_size
p.batch_size = 1
# If we have [X] variables use their amount, unless unchecked
if cbChangeCount == True and len(strSequentialPrompt) > 0:
MainJobCount = LineCount
SubCycleCount = 1
if SubLineCount > 0:
SubCycleCount = SubLineCount
# Any random lines present?
RandomLinesA = [r.strip() for r in strRandomPromptA.splitlines()]
RandomLinesB = [r.strip() for r in strRandomPromptB.splitlines()]
RandomLinesC = [r.strip() for r in strRandomPromptC.splitlines()]
# So the progress bar works correctly
state.job_count = MainJobCount * SubIterationCount * SubCycleCount
CurrentChoice = 0
SubCurrentChoice = 0
FinalResultDirection = ""
for x in range(MainJobCount):
SeedStep = 0
AllMovements = ""
AllArtists = ""
# =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=->
# Large artist selection
if selArtistA != "Not set":
# If Random is selected then pick a random artist
if selArtistA == "Random":
AllArtists += ",(by Artist " + random.choice(ResultArtistList) + \
":" + str(sliImageArtistStrengthA) + ")"
# otherwise use the selected value
else:
AllArtists += ",(by Artist " + selArtistA + \
":" + str(sliImageArtistStrengthA) + ")"
if selArtistB != "Not set":
if selArtistB == "Random":
AllArtists += ",(by Artist " + random.choice(ResultArtistList) + \
":" + str(sliImageArtistStrengthB) + ")"
else:
AllArtists += ",(by Artist " + selArtistB + \
":" + str(sliImageArtistStrengthB) + ")"
if selArtistC != "Not set":
if selArtistC == "Random":
AllArtists += ",(by Artist " + random.choice(ResultArtistList) + \
":" + str(sliImageArtistStrengthC) + ")"
else:
AllArtists += ",(by Artist " + selArtistC + \
":" + str(sliImageArtistStrengthC) + ")"
# =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=->
# Large style selection
if selArtMovementA != "Not set":
if selArtMovementA == "Random":
AllMovements += ",(" + random.choice(ArtMovementList) + ":" + str(selArtMovementStrengthA) + ")"
else:
AllMovements += ",(" + selArtMovementA + ":" + str(selArtMovementStrengthA) + ")"
if selArtMovementB != "Not set":
if selArtMovementB == "Random":
AllMovements += ",(" + random.choice(ArtMovementList) + ":" + str(selArtMovementStrengthB) + ")"
else:
AllMovements += ",(" + selArtMovementB + ":" + str(selArtMovementStrengthB) + ")"
if selArtMovementC != "Not set":
if selArtMovementC == "Random":
AllMovements += ",(" + random.choice(ArtMovementList) + ":" + str(selArtMovementStrengthC) + ")"
else:
AllMovements += ",(" + selArtMovementC + ":" + str(selArtMovementStrengthC) + ")"
# =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=->
for y in range(SubIterationCount):
# Clear the variables so random selection can work
FinalResultMood = ""
FinalResultColor = ""
FinalConcept = ""
# Preset the selection variables
MainType = ""
TypeFront = ""
TypePositives = ""
TypeNegatives = ""
# =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=->
if ddResultConcept != "Not set":
if ddResultConcept == "Random":
FinalConcept = random.choice(ElementConcept)
else:
FinalConcept = ddResultConcept
else:
FinalConcept = ""
# If main prompt isn't empty...
if ResultType[ddCoreResultType] != "":
# If it is random, give it a random value
if ResultType[ddCoreResultType] == "Random":
MainType = random.choice(ResultNames)
# otherwise use the selected value
else:
MainType = ddCoreResultType
# =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=->
if ddResultDirection != "Not set":
if ddResultDirection == "Random":
FinalResultDirection = " (" + random.choice(
ResultDirectionList) + ":" + str(slResultDirectionStrength) + ") "
else:
FinalResultDirection = " (" + ddResultDirection + \
":" + str(slResultDirectionStrength) + ") "
# =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=->
# Pick the mood
if ddResultMood != "Not set":
if ddResultMood == "Random":
FinalResultMood = ",(" + random.choice(ResultMoodList) + \
":" + str(slResultMoodStrength) + ") "
else:
FinalResultMood = ",(" + ddResultMood + \
":" + str(slResultMoodStrength) + ") "
# =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=->
# If present, add batch element, otherwise remove that reference
TempText = ""
if LineCount > 0:
if len(BatchLines[CurrentChoice % LineCount]) > 0:
TempText = BatchLines[CurrentChoice % LineCount]
#TempText = copy_p.prompt.replace("[X]", TempText)
TempText = p.prompt.replace("[X]", TempText)
# =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=->
# If present, add random element, otherwise remove that reference
if len(RandomLinesA) > 0:
TempText = TempText.replace(
"[A]", random.choice(RandomLinesA))
else:
TempText = TempText.replace("[A]", "")
if len(RandomLinesB) > 0:
TempText = TempText.replace(
"[B]", random.choice(RandomLinesB))
else:
TempText = TempText.replace("[B]", "")
if len(RandomLinesC) > 0:
TempText = TempText.replace(
"[C]", random.choice(RandomLinesC))
else:
TempText = TempText.replace("[C]", "")
# =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=->
# Colors
if ddResultColor!= "Not set":
if ddResultColor == "Random":
FinalResultColor = ",(" + random.choice(ResultColorList) + \
":" + str(slResultColorStrength) + ") "
else:
FinalResultColor = ",(" + ddResultColor + \
":" + str(slResultColorStrength) + ") "
# =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=->
# If main prompt isn't empty...
if MainType != "":
# Format our variables to merge into positive prompt...
TypeFront = "(" + FinalConcept + " " + ResultTypeBefore[MainType] + \
":" + str(slResultTypeStrength) + ") of "
TypePositives = ResultTypePositives[MainType]
TypeNegatives = ResultTypeNegatives[MainType]
# Our main prompt composed of all the selected elements
MainPositive = TypeFront + FinalResultDirection + FinalResultMood + TempText + \
AllArtists + TypePositives + AllMovements + \
FinalResultColor + Preset[ddPreset]
#MainNegative = copy_p.negative_prompt
MainNegative = p.negative_prompt
# =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=->
SubCurrentChoice = 0
for z in range(SubCycleCount):
# Copy of the main prompt module to make batches, I guess...
copy_p = copy.copy(p)
if copy_p.seed != -1: # and 'p.seed' in locals():
copy_p.seed += SeedStep
SubTempText = ""
if SubLineCount > 0:
if len(SubBatchLines[SubCurrentChoice % SubLineCount]) > 0:
SubTempText = SubBatchLines[SubCurrentChoice % SubLineCount]
TempText = MainPositive.replace("[Y]", SubTempText)
TempText = TempText.replace("[xs]", str(random.randrange(100000,999999,1)))
TempText = TempText.replace("[XS]", str(random.randrange(100000,999999,1)))
TempText = TempText.replace("[s]", ''.join(random.choices("0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZ", k=5)))
TempText = TempText.replace("[S]", ''.join(random.choices("0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZ", k=5)))
TempText = TempText.replace("[m]", ''.join(random.choices("0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZ", k=10)))
TempText = TempText.replace("[M]", ''.join(random.choices("0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZ", k=10)))
TempText = TempText.replace("[l]", ''.join(random.choices("0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZ", k=5)) + " " + ''.join(random.choices("0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZ", k=5)))
TempText = TempText.replace("[L]", ''.join(random.choices("0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZ", k=5)) + " " + ''.join(random.choices("0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZ", k=5)))
TempText = TempText.replace("[xl]", ''.join(random.choices("0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZ", k=10)) + " " + ''.join(random.choices("0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZ", k=10)))
TempText = TempText.replace("[XL]", ''.join(random.choices("0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZ", k=10)) + " " + ''.join(random.choices("0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZ", k=10)))
# Clean up positive prompt
TempText = " ".join(TempText.split())
TempText = TempText.replace(",,", ",")
TempText = TempText.replace(" ,", ",")
TempText = TempText.replace(",", ",")
TempText = TempText.replace("( ", "(")
TempText = TempText.replace(" )", ")")
TempText = TempText.strip(",")
TempText = TempText.strip()
copy_p.prompt = TempText
# =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=->
# Clean up negative prompt
TempText = MainNegative + TypeNegatives + \
AlwaysBad + PresetNegatives[ddPreset]
TempText = " ".join(TempText.split())
TempText = " ".join(TempText.split())
TempText = TempText.replace(",,", ",")
TempText = TempText.replace(" ,", ",")
TempText = TempText.replace(",", ",")
TempText = TempText.replace("( ", "(")
TempText = TempText.replace(" )", ")")
TempText = TempText.strip(",")
TempText = TempText.strip()
copy_p.negative_prompt = TempText
# =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=->
# Add information in command prompt window and process the image
print(f"\n\n[Prompt {x+1}/{MainJobCount}][Iteration {y+1}/{SubIterationCount}][SubPrompt {z}/{SubLineCount}][Seed {int(copy_p.seed)}] >>> Positives <<< {copy_p.prompt} >>> Negatives <<< {copy_p.negative_prompt}\n")
proc = process_images(copy_p)
infotexts += proc.infotexts
images += proc.images
seeds.append(proc.seed)
prompts.append(proc.prompt)
SubCurrentChoice += 1
if cbIncreaseSeed == True:
SeedStep += 1
CurrentChoice += 1
p.batch_size = MainJobCount
p.n_iter = SubIterationCount
if cbShowTips:
print(
f"\n\nStylePile processing complete. Here's a random tip:\n{random.choice(TipsAndTricks)}\n")
return Processed(p=p, images_list=images, seed=p.seed, all_seeds=seeds, all_prompts=prompts, info=infotexts[0],
infotexts=infotexts)
|