Spaces:
Sleeping
Sleeping
File size: 43,338 Bytes
0f819bc 5312469 0f819bc 82119a6 5312469 82119a6 5312469 0f819bc 5312469 0f819bc 5312469 0f819bc 5312469 0f819bc 5312469 0f819bc 5312469 0f819bc 5312469 0f819bc 5312469 0f819bc 5312469 0f819bc 5312469 0f819bc 5312469 0f819bc 5312469 0f819bc 5312469 0f819bc 5312469 0f819bc 5312469 0f819bc 5312469 0f819bc 5312469 0f819bc 5312469 0f819bc 5312469 0f819bc 5312469 0f819bc 5312469 0f819bc 5312469 0f819bc 5312469 0f819bc 5312469 0f819bc 5312469 0f819bc 5312469 0f819bc 5312469 0f819bc 5312469 0f819bc 5312469 0f819bc 5312469 0f819bc 5312469 0f819bc 5312469 0f819bc 5312469 0f819bc 5312469 0f819bc 5312469 0f819bc 5312469 0f819bc 5312469 0f819bc 5312469 0f819bc 5312469 0f819bc 5312469 0f819bc 5312469 0f819bc c515783 0f819bc c515783 0f819bc c515783 0f819bc c515783 0f819bc c515783 0f819bc 5312469 0f819bc 5312469 c515783 5312469 |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 |
# -*- coding: utf-8 -*-
"""Academic Text Humanizer - Hugging Face Spaces Deployment"""
# Step 1: Import Libraries
import torch
from transformers import AutoTokenizer, AutoModelForCausalLM, GenerationConfig, set_seed
import hashlib
import re
import gradio as gr
import os
from huggingface_hub import login
# Step 2: Login to Hugging Face
hf_token = os.getenv("HF_TOKEN")
if hf_token:
login(token=hf_token)
# Step 3: Load Model and Tokenizer
print("Loading model and tokenizer...")
model_name = "mistralai/Mistral-7B-Instruct-v0.2"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(
model_name,
torch_dtype=torch.float16,
device_map="auto"
)
if tokenizer.pad_token is None:
tokenizer.pad_token = tokenizer.eos_token
print("Model loaded successfully!")
# Step 4: Regional Accent Dictionaries
# USA Academic English Preferences
USA_ACADEMIC_STYLE = {
'spelling': {
# British → American spelling
'analyse': 'analyze',
'analyses': 'analyzes',
'analysing': 'analyzing',
'analysed': 'analyzed',
'behaviour': 'behavior',
'behaviours': 'behaviors',
'behavioural': 'behavioral',
'centre': 'center',
'centres': 'centers',
'centred': 'centered',
'colour': 'color',
'colours': 'colors',
'coloured': 'colored',
'defence': 'defense',
'favour': 'favor',
'favours': 'favors',
'favored': 'favored',
'favourite': 'favorite',
'honour': 'honor',
'honours': 'honors',
'honoured': 'honored',
'labour': 'labor',
'labours': 'labors',
'laboured': 'labored',
'licence': 'license',
'organise': 'organize',
'organises': 'organizes',
'organised': 'organized',
'organising': 'organizing',
'organisation': 'organization',
'organisations': 'organizations',
'realise': 'realize',
'realises': 'realizes',
'realised': 'realized',
'realising': 'realizing',
'recognise': 'recognize',
'recognises': 'recognizes',
'recognised': 'recognized',
'recognising': 'recognizing',
'programme': 'program',
'programmes': 'programs',
'theatre': 'theater',
'theatres': 'theaters',
'travelled': 'traveled',
'travelling': 'traveling',
'traveller': 'traveler',
'modelled': 'modeled',
'modelling': 'modeling',
'cancelled': 'canceled',
'cancelling': 'canceling',
'counsellor': 'counselor',
'counselling': 'counseling',
'jewellery': 'jewelry',
'fulfil': 'fulfill',
'fulfilment': 'fulfillment',
'skilful': 'skillful',
'grey': 'gray',
'practise': 'practice',
'practising': 'practicing',
'practised': 'practiced',
'emphasise': 'emphasize',
'emphasises': 'emphasizes',
'emphasised': 'emphasized',
'emphasising': 'emphasizing',
'summarise': 'summarize',
'summarises': 'summarizes',
'summarised': 'summarized',
'summarising': 'summarizing',
'categorise': 'categorize',
'categorises': 'categorizes',
'categorised': 'categorized',
'categorising': 'categorizing',
'characterise': 'characterize',
'characterises': 'characterizes',
'characterised': 'characterized',
'characterising': 'characterizing',
'criticise': 'criticize',
'criticises': 'criticizes',
'criticised': 'criticized',
'criticising': 'criticizing',
'finalise': 'finalize',
'finalises': 'finalizes',
'finalised': 'finalized',
'finalising': 'finalizing',
'generalise': 'generalize',
'generalises': 'generalizes',
'generalised': 'generalized',
'generalising': 'generalizing',
'hypothesise': 'hypothesize',
'hypothesises': 'hypothesizes',
'hypothesised': 'hypothesized',
'hypothesising': 'hypothesizing',
'maximise': 'maximize',
'maximises': 'maximizes',
'maximised': 'maximized',
'maximising': 'maximizing',
'minimise': 'minimize',
'minimises': 'minimizes',
'minimised': 'minimized',
'minimising': 'minimizing',
'normalise': 'normalize',
'normalises': 'normalizes',
'normalised': 'normalized',
'normalising': 'normalizing',
'optimise': 'optimize',
'optimises': 'optimizes',
'optimised': 'optimized',
'optimising': 'optimizing',
'standardise': 'standardize',
'standardises': 'standardizes',
'standardised': 'standardized',
'standardising': 'standardizing',
'utilise': 'utilize',
'utilises': 'utilizes',
'utilised': 'utilized',
'utilising': 'utilizing',
'visualise': 'visualize',
'visualises': 'visualizes',
'visualised': 'visualized',
'visualising': 'visualizing',
'apologise': 'apologize',
'apologises': 'apologizes',
'apologised': 'apologized',
'apologising': 'apologizing',
'capitalise': 'capitalize',
'capitalises': 'capitalizes',
'capitalised': 'capitalized',
'capitalising': 'capitalizing',
'globalise': 'globalize',
'globalises': 'globalizes',
'globalised': 'globalized',
'globalising': 'globalizing',
'industrialise': 'industrialize',
'industrialises': 'industrializes',
'industrialised': 'industrialized',
'industrialising': 'industrializing',
'materialise': 'materialize',
'materialises': 'materializes',
'materialised': 'materialized',
'materialising': 'materializing',
'mobilise': 'mobilize',
'mobilises': 'mobilizes',
'mobilised': 'mobilized',
'mobilising': 'mobilizing',
'modernise': 'modernize',
'modernises': 'modernizes',
'modernised': 'modernized',
'modernising': 'modernizing',
'privatise': 'privatize',
'privatises': 'privatizes',
'privatised': 'privatized',
'privatising': 'privatizing',
'rationalise': 'rationalize',
'rationalises': 'rationalizes',
'rationalised': 'rationalized',
'rationalising': 'rationalizing',
'revolutionise': 'revolutionize',
'revolutionises': 'revolutionizes',
'revolutionised': 'revolutionized',
'revolutionising': 'revolutionizing',
'socialise': 'socialize',
'socialises': 'socializes',
'socialised': 'socialized',
'socialising': 'socializing',
'specialise': 'specialize',
'specialises': 'specializes',
'specialised': 'specialized',
'specialising': 'specializing',
'stabilise': 'stabilize',
'stabilises': 'stabilizes',
'stabilised': 'stabilized',
'stabilising': 'stabilizing',
'symbolise': 'symbolize',
'symbolises': 'symbolizes',
'symbolised': 'symbolized',
'symbolising': 'symbolizing',
'synthesise': 'synthesize',
'synthesises': 'synthesizes',
'synthesised': 'synthesized',
'synthesising': 'synthesizing',
'theorise': 'theorize',
'theorises': 'theorizes',
'theorised': 'theorized',
'theorising': 'theorizing',
'urbanise': 'urbanize',
'urbanises': 'urbanizes',
'urbanised': 'urbanized',
'urbanising': 'urbanizing',
},
'phrases': {
'at the weekend': 'on the weekend',
'in hospital': 'in the hospital',
'in future': 'in the future',
'at university': 'at the university',
'different to': 'different from',
'different than': 'different from',
'write to': 'write',
'Monday to Friday': 'Monday through Friday',
},
'punctuation': {
'quotation_style': 'double',
'period_with_quotes': 'inside',
},
'vocabulary': {
'whilst': 'while',
'amongst': 'among',
'towards': 'toward',
'afterwards': 'afterward',
'forwards': 'forward',
'backwards': 'backward',
'upwards': 'upward',
'downwards': 'downward',
'learnt': 'learned',
'burnt': 'burned',
'dreamt': 'dreamed',
'spelt': 'spelled',
'spoilt': 'spoiled',
}
}
# UK Academic English Preferences
UK_ACADEMIC_STYLE = {
'spelling': {
# American → British spelling
'analyze': 'analyse',
'analyzes': 'analyses',
'analyzing': 'analysing',
'analyzed': 'analysed',
'behavior': 'behaviour',
'behaviors': 'behaviours',
'behavioral': 'behavioural',
'center': 'centre',
'centers': 'centres',
'centered': 'centred',
'color': 'colour',
'colors': 'colours',
'colored': 'coloured',
'defense': 'defence',
'favor': 'favour',
'favors': 'favours',
'favored': 'favoured',
'favorite': 'favourite',
'honor': 'honour',
'honors': 'honours',
'honored': 'honoured',
'labor': 'labour',
'labors': 'labours',
'labored': 'laboured',
'license': 'licence',
'organize': 'organise',
'organizes': 'organises',
'organized': 'organised',
'organizing': 'organising',
'organization': 'organisation',
'organizations': 'organisations',
'realize': 'realise',
'realizes': 'realises',
'realized': 'realised',
'realizing': 'realising',
'recognize': 'recognise',
'recognizes': 'recognises',
'recognized': 'recognised',
'recognizing': 'recognising',
'program': 'programme',
'programs': 'programmes',
'theater': 'theatre',
'theaters': 'theatres',
'traveled': 'travelled',
'traveling': 'travelling',
'traveler': 'traveller',
'modeled': 'modelled',
'modeling': 'modelling',
'canceled': 'cancelled',
'canceling': 'cancelling',
'counselor': 'counsellor',
'counseling': 'counselling',
'jewelry': 'jewellery',
'fulfill': 'fulfil',
'fulfillment': 'fulfilment',
'skillful': 'skilful',
'gray': 'grey',
'practice': 'practise',
'practicing': 'practising',
'practiced': 'practised',
'emphasize': 'emphasise',
'emphasizes': 'emphasises',
'emphasized': 'emphasised',
'emphasizing': 'emphasising',
'summarize': 'summarise',
'summarizes': 'summarises',
'summarized': 'summarised',
'summarizing': 'summarising',
'categorize': 'categorise',
'categorizes': 'categorises',
'categorized': 'categorised',
'categorizing': 'categorising',
'characterize': 'characterise',
'characterizes': 'characterises',
'characterized': 'characterised',
'characterizing': 'characterising',
'criticize': 'criticise',
'criticizes': 'criticises',
'criticized': 'criticised',
'criticizing': 'criticising',
'finalize': 'finalise',
'finalizes': 'finalises',
'finalized': 'finalised',
'finalizing': 'finalising',
'generalize': 'generalise',
'generalizes': 'generalises',
'generalized': 'generalised',
'generalizing': 'generalising',
'hypothesize': 'hypothesise',
'hypothesizes': 'hypothesises',
'hypothesized': 'hypothesised',
'hypothesizing': 'hypothesising',
'maximize': 'maximise',
'maximizes': 'maximises',
'maximized': 'maximised',
'maximizing': 'maximising',
'minimize': 'minimise',
'minimizes': 'minimises',
'minimized': 'minimised',
'minimizing': 'minimising',
'normalize': 'normalise',
'normalizes': 'normalises',
'normalized': 'normalised',
'normalizing': 'normalising',
'optimize': 'optimise',
'optimizes': 'optimises',
'optimized': 'optimised',
'optimizing': 'optimising',
'standardize': 'standardise',
'standardizes': 'standardises',
'standardized': 'standardised',
'standardizing': 'standardising',
'utilize': 'utilise',
'utilizes': 'utilises',
'utilized': 'utilised',
'utilizing': 'utilising',
'visualize': 'visualise',
'visualizes': 'visualises',
'visualized': 'visualised',
'visualizing': 'visualising',
'apologize': 'apologise',
'apologizes': 'apologises',
'apologized': 'apologised',
'apologizing': 'apologising',
'capitalize': 'capitalise',
'capitalizes': 'capitalises',
'capitalized': 'capitalised',
'capitalizing': 'capitalising',
'globalize': 'globalise',
'globalizes': 'globalises',
'globalized': 'globalised',
'globalizing': 'globalising',
'industrialize': 'industrialise',
'industrializes': 'industrialises',
'industrialized': 'industrialised',
'industrializing': 'industrialising',
'materialize': 'materialise',
'materializes': 'materialises',
'materialized': 'materialised',
'materializing': 'materialising',
'mobilize': 'mobilise',
'mobilizes': 'mobilises',
'mobilized': 'mobilised',
'mobilizing': 'mobilising',
'modernize': 'modernise',
'modernizes': 'modernises',
'modernized': 'modernised',
'modernizing': 'modernising',
'privatize': 'privatise',
'privatizes': 'privatises',
'privatized': 'privatised',
'privatizing': 'privatising',
'rationalize': 'rationalise',
'rationalizes': 'rationalises',
'rationalized': 'rationalised',
'rationalizing': 'rationalising',
'revolutionize': 'revolutionise',
'revolutionizes': 'revolutionises',
'revolutionized': 'revolutionised',
'revolutionizing': 'revolutionising',
'socialize': 'socialise',
'socializes': 'socialises',
'socialized': 'socialised',
'socializing': 'socialising',
'specialize': 'specialise',
'specializes': 'specialises',
'specialized': 'specialised',
'specializing': 'specialising',
'stabilize': 'stabilise',
'stabilizes': 'stabilises',
'stabilized': 'stabilised',
'stabilizing': 'stabilising',
'symbolize': 'symbolise',
'symbolizes': 'symbolises',
'symbolized': 'symbolised',
'symbolizing': 'symbolising',
'synthesize': 'synthesise',
'synthesizes': 'synthesises',
'synthesized': 'synthesised',
'synthesizing': 'synthesising',
'theorize': 'theorise',
'theorizes': 'theorises',
'theorized': 'theorised',
'theorizing': 'theorising',
'urbanize': 'urbanise',
'urbanizes': 'urbanises',
'urbanized': 'urbanised',
'urbanizing': 'urbanising',
},
'phrases': {
'on the weekend': 'at the weekend',
'in the hospital': 'in hospital',
'in the future': 'in future',
'at the university': 'at university',
'different from': 'different to',
'Monday through Friday': 'Monday to Friday',
},
'punctuation': {
'quotation_style': 'single',
'period_with_quotes': 'outside',
},
'vocabulary': {
'while': 'whilst',
'among': 'amongst',
'toward': 'towards',
'afterward': 'afterwards',
'forward': 'forwards',
'backward': 'backwards',
'upward': 'upwards',
'downward': 'downwards',
'learned': 'learnt',
'burned': 'burnt',
'dreamed': 'dreamt',
'spelled': 'spelt',
'spoiled': 'spoilt',
}
}
# Step 5: ULTRA-DEEP ACADEMIC ENGINE
def get_ultra_deep_academic_engine(accent='usa'):
"""
Returns ultra-deep academic humanization engine for <5% AI detection.
"""
accent_instruction = ""
if accent.lower() == 'usa':
accent_instruction = """
REGIONAL STYLE: AMERICAN ACADEMIC ENGLISH
- Spelling: -ize (analyze, organize), -or (behavior, color), -er (center)
- Vocabulary: "while", "among", "toward", "learned"
- Phrases: "on the weekend", "in the hospital", "different from"
"""
elif accent.lower() == 'uk':
accent_instruction = """
REGIONAL STYLE: BRITISH ACADEMIC ENGLISH
- Spelling: -ise (analyse, organise), -our (behaviour, colour), -re (centre)
- Vocabulary: "whilst", "amongst", "towards", "learnt"
- Phrases: "at the weekend", "in hospital", "different to"
"""
ultra_deep_engine = """You are a HUMAN academic researcher rewriting AI text to sound completely natural and undetectable.
🚨 CRITICAL: Your output MUST score <5% on AI detectors. This requires EXTREME transformation.
═══════════════════════════════════════════════════════════════════════════════
PHASE 1: DESTROY ALL AI VOCABULARY (ZERO TOLERANCE)
═══════════════════════════════════════════════════════════════════════════════
FORBIDDEN WORDS - NEVER USE THESE:
❌ pivotal, paramount, crucial, vital, essential, fundamental, cornerstone, foundational
❌ multifaceted, nuanced, intricate, complex, comprehensive, holistic
❌ delve, leverage, utilize, facilitate, enhance, foster, cultivate, empower
❌ robust, rigorous, substantial, significant, considerable, notable
❌ myriad, plethora, array, spectrum, gamut, breadth
❌ endeavor, notion, realm, facet, aspect, dimension
❌ paradigm, framework, landscape, trajectory, dynamic
❌ optimal, viable, integral, inherent, quintessential, ubiquitous
❌ underscores, showcases, exemplifies, encapsulates, illuminates, elucidates
❌ necessitates, encompasses, embodies, perpetuates, mitigates, exacerbates
❌ bolsters, galvanizes, transcends, permeates, resonates
❌ navigating, harnessing, spearheading, pioneering, orchestrating
❌ attribute, capacity, capability, proficiency, competency
❌ milieu, domain, sphere, arena, context
❌ inculcates, instills, engenders, fortifies, bolsters
❌ practitioners, individuals (overused), persons
❌ steadfast, unwavering, persistent, diligent
❌ flourishes, blossoms, dissipating
❌ constraint, empowerment, deliberatively
❌ myriad opportunities, enduring triumphs, latent potential
❌ amidst, amongst, brimming, waning, fleeting, momentary
❌ surmount, succumbing, propelling, fortifying
❌ signifies, equips, dictate, destiny, groundwork
REPLACE WITH NATURAL ACADEMIC LANGUAGE:
✅ important, key, central, basic, core, main
✅ complex → complicated, detailed, involved
✅ examine, study, investigate, look at, explore
✅ use, apply, employ, work with
✅ improve, strengthen, increase, develop, build
✅ many, several, numerous, various, multiple
✅ idea, concept, theory, principle, view
✅ field, area, subject, topic, domain
✅ shows, demonstrates, indicates, reveals, suggests
✅ requires, needs, demands, calls for
✅ includes, contains, involves, covers
✅ people, researchers, scholars, students, workers
✅ consistent, regular, steady, constant
✅ grows, develops, increases, expands
✅ limit, restriction, control
✅ power, ability, capacity, means
✅ chances, possibilities, options
✅ lasting, long-term, sustained, continued
═══════════════════════════════════════════════════════════════════════════════
PHASE 2: SENTENCE STRUCTURE CHAOS (BREAK AI PATTERNS) **FOLLOW THIS STRICTLY**
═══════════════════════════════════════════════════════════════════════════════
AI WRITES: Uniform 20-25 word sentences with perfect rhythm
YOU WRITE: Chaotic variation - 7, 29, 11, 34, 9, 26, 13, 31 words
SENTENCE LENGTH TARGETS:
- Ultra-short (6-10 words): "Discipline matters. It shapes who we become."
- Short (11-15 words): "This quality affects success in school, sports, and work."
- Medium (16-22 words): "Research from Stanford University (2019) shows that disciplined students perform better academically."
- Long (23-30 words): "When people maintain discipline, they develop habits like regular study sessions and healthy eating patterns that support their goals over time."
- Very long (31-38 words): "The connection between discipline and achievement has been documented extensively in psychological literature, with studies indicating that self-control predicts outcomes more reliably than intelligence measures alone (Duckworth, 2016)."
CRITICAL: Mix these randomly. Never use same length twice in a row.
═══════════════════════════════════════════════════════════════════════════════
PHASE 3: OPENING VARIATION (NEVER REPEAT PATTERNS)
═══════════════════════════════════════════════════════════════════════════════
AI PATTERN: Starts every sentence with subject or "This/These/Such"
HUMAN PATTERN: Wild variation
OPENING TYPES (Rotate constantly):
1. Subject-first: "Discipline shapes behavior."
2. Prepositional: "In academic settings, discipline predicts success."
3. Adverb: "However, discipline requires consistent effort."
4. Participial: "Building on previous research, this study examines..."
5. Dependent clause: "When students practice discipline, they perform better."
6. Transitional: "Research shows discipline matters."
7. Infinitive: "To understand discipline, we must examine its components."
8. Gerund: "Maintaining discipline requires daily effort."
9. Direct object: "Success in school demands discipline."
10. Question: "What makes discipline effective?"
RULE: Never use same opening type in consecutive sentences.
═══════════════════════════════════════════════════════════════════════════════
PHASE 4: PASSIVE VOICE STRATEGY (10-20% ONLY)
═══════════════════════════════════════════════════════════════════════════════
AI OVERUSES PASSIVE: 60-70% passive voice
HUMANS USE: 10-20% passive voice
WHEN TO USE PASSIVE:
✅ Methods: "Data were collected from 200 participants."
✅ Results: "Significant differences were observed between groups."
✅ Objectivity: "The hypothesis was tested using regression analysis."
WHEN TO USE ACTIVE:
✅ Agency: "Researchers conducted three experiments."
✅ Clarity: "This study examines the role of discipline."
✅ Engagement: "Students who practice discipline achieve better grades."
═══════════════════════════════════════════════════════════════════════════════
PHASE 5: NATURAL ACADEMIC IMPERFECTIONS
═══════════════════════════════════════════════════════════════════════════════
HUMANS AREN'T PERFECT. Add these natural elements:
1. STRATEGIC HEDGING (3-5 per 500 words, NOT every sentence):
✅ "appears to", "suggests that", "may indicate", "tends to"
✅ "Research suggests discipline matters" (not "might possibly perhaps indicate")
2. VARIED TRANSITIONS (NOT formulaic):
❌ AVOID: Moreover, Furthermore, Additionally, In addition (AI overuses these)
✅ USE: However, Nevertheless, In contrast, Similarly, Research shows, Studies indicate
✅ USE: Building on this, Extending this analysis, This finding suggests
3. CITATION INTEGRATION (Specific, varied):
❌ "Research shows" (vague, AI-like)
✅ "According to Smith (2019)", "Duckworth et al. (2016) found", "Recent studies demonstrate"
4. PUNCTUATION VARIETY:
- Use semicolons sparingly (2-3 per 500 words)
- Em dashes for emphasis — like this — occasionally
- Parenthetical asides (as noted earlier) for natural flow
- Colons for lists: first, second, third
5. PARAGRAPH BREAKS:
- Vary paragraph length: 3 sentences, 5 sentences, 2 sentences, 6 sentences
- NOT uniform 4-sentence paragraphs (AI pattern)
═══════════════════════════════════════════════════════════════════════════════
PHASE 6: ELIMINATE AI DRAMA & INTENSITY
═══════════════════════════════════════════════════════════════════════════════
AI LOVES DRAMA:
❌ "profound implications", "groundbreaking", "revolutionary", "unprecedented"
❌ "deeply", "highly", "extremely", "incredibly", "remarkably", "exceptionally"
❌ "brimming with", "trapped in cycles", "fleeting inclinations"
ACADEMIC TONE:
✅ "important implications", "new", "significant", "recent"
✅ "considerably", "substantially", "notably", "particularly"
✅ "full of", "stuck in patterns", "temporary desires"
═══════════════════════════════════════════════════════════════════════════════
PHASE 7: SPECIFIC EXAMPLES & CONCRETE LANGUAGE
═══════════════════════════════════════════════════════════════════════════════
AI STAYS ABSTRACT:
❌ "various domains", "multiple contexts", "diverse settings"
HUMANS USE SPECIFICS:
✅ "in schools, workplaces, and sports teams"
✅ "students, employees, and athletes"
✅ "studying regularly, eating healthy, arriving on time"
═══════════════════════════════════════════════════════════════════════════════
PHASE 8: BREAK PERFECT PARALLELISM
═══════════════════════════════════════════════════════════════════════════════
AI WRITES: "analyzing data, examining results, and evaluating outcomes"
(Perfect parallel structure - all gerunds)
HUMAN WRITES: "data analysis, examination of results, and outcome evaluation"
(Mixed structures - noun, noun phrase, noun)
OR: "analyzing data, result examination, and how we evaluate outcomes"
(Gerund, noun, clause - deliberately imperfect)
═══════════════════════════════════════════════════════════════════════════════
FINAL CHECKLIST FOR <5% AI DETECTION:
═══════════════════════════════════════════════════════════════════════════════
✓ ZERO forbidden AI vocabulary (check every word)
✓ Sentence lengths: 7, 29, 11, 34, 9, 26, 13, 31, 18, 27 (chaotic)
✓ Opening types: Never repeat consecutively
✓ Passive voice: 30-35% (not 60%+)
✓ Hedging: 3-5 instances per 500 words (strategic, not excessive)
✓ Transitions: Varied, not formulaic (avoid Moreover/Furthermore/Additionally)
✓ Citations: Specific with years (Author, Year)
✓ No AI drama words (profound, groundbreaking, deeply, highly)
✓ Concrete examples (not abstract "various domains")
✓ Imperfect parallelism (mixed grammatical structures)
✓ Natural punctuation variety (semicolons, em dashes, parentheses)
✓ Paragraph length variation (not uniform)
═══════════════════════════════════════════════════════════════════════════════
EXAMPLE TRANSFORMATION:
═══════════════════════════════════════════════════════════════════════════════
❌ AI TEXT (39% AI):
"Discipline is a pivotal attribute for personal development, shaping both conduct and identity, influencing success, habits formation, interpersonal relations, and an individual's overall sense of purpose."
✅ HUMAN TEXT (<5% AI):
"Discipline shapes personal growth. It affects how people behave and who they become, influencing success in school and work, habit formation, relationships with others, and sense of purpose in life."
NOTICE THE DIFFERENCES:
- Removed: pivotal, attribute, conduct, interpersonal, individual's, overall
- Added: Concrete examples (school, work)
- Broke into 2 sentences (7 words, then 23 words)
- Used simple language (shapes, affects, people, relationships)
- Maintained academic tone WITHOUT AI vocabulary
"""
return ultra_deep_engine + accent_instruction + """
═══════════════════════════════════════════════════════════════════════════════
NOW REWRITE THE TEXT BELOW:
═══════════════════════════════════════════════════════════════════════════════
REQUIREMENTS:
1. Maintain formal academic tone
2. Use ZERO forbidden AI vocabulary
3. Apply extreme sentence variation
4. Include specific examples
5. Break perfect parallelism
6. Use 10-20% passive voice
7. Add natural imperfections
8. Target <5% AI detection
OUTPUT ONLY THE REWRITTEN TEXT. NO EXPLANATIONS."""
# Step 6: Apply Regional Accent
def apply_accent(text, accent='usa'):
"""
Applies regional spelling and vocabulary conventions.
"""
if accent.lower() == 'usa':
style = USA_ACADEMIC_STYLE
elif accent.lower() == 'uk':
style = UK_ACADEMIC_STYLE
else:
return text
# Apply spelling changes
for original, replacement in style['spelling'].items():
pattern = r'\b' + re.escape(original) + r'\b'
text = re.sub(pattern, replacement, text)
if original[0].islower():
cap_original = original.capitalize()
cap_replacement = replacement.capitalize()
pattern_cap = r'\b' + re.escape(cap_original) + r'\b'
text = re.sub(pattern_cap, cap_replacement, text)
# Apply vocabulary changes
for original, replacement in style['vocabulary'].items():
pattern = r'\b' + re.escape(original) + r'\b'
text = re.sub(pattern, replacement, text, flags=re.IGNORECASE)
# Apply phrase changes
for original, replacement in style['phrases'].items():
text = text.replace(original, replacement)
text = text.replace(original.capitalize(), replacement.capitalize())
return text
# Step 7: Create Deterministic Configuration
def create_ultra_humanized_config(input_text, accent, tokenizer):
"""
Creates config optimized for <5% AI detection.
"""
combined = f"{input_text}_{accent}"
text_hash = hashlib.md5(combined.encode()).hexdigest()
seed = int(text_hash[:8], 16) % (2**32)
set_seed(seed)
gen_config = GenerationConfig(
max_new_tokens=4096,
temperature=1.20,
top_p=0.95,
top_k=80,
do_sample=True,
repetition_penalty=1.25,
no_repeat_ngram_size=5,
seed=seed,
pad_token_id=tokenizer.pad_token_id,
eos_token_id=tokenizer.eos_token_id
)
return gen_config, seed
# Step 8: Humanize with Ultra-Deep Transformations
def humanize_academic_text(ai_text, accent='usa'):
"""
Ultra-deep humanization for <5% AI detection.
"""
academic_engine = get_ultra_deep_academic_engine(accent)
gen_config, seed = create_ultra_humanized_config(ai_text, accent, tokenizer)
print(f"Processing with {accent.upper()} accent (seed: {seed})")
print("Using ultra-deep humanization for <5% AI detection...")
prompt = f"""<s>[INST] {academic_engine}
--- INPUT TEXT ---
{ai_text}
--- END INPUT ---
### REWRITTEN TEXT ({accent.upper()} ENGLISH):
[/INST]"""
inputs = tokenizer(prompt, return_tensors="pt", padding=True, truncation=True, max_length=4096)
inputs = {k: v.to(model.device) for k, v in inputs.items()}
print("Generating ultra-humanized academic text...")
with torch.no_grad():
outputs = model.generate(
**inputs,
generation_config=gen_config
)
full_output = tokenizer.decode(outputs[0], skip_special_tokens=True)
if "[/INST]" in full_output:
humanized = full_output.split("[/INST]")[-1].strip()
else:
humanized = full_output.strip()
humanized = apply_accent(humanized, accent)
humanized = deep_cleanup(humanized)
return humanized
def deep_cleanup(text):
"""
Removes ALL AI vocabulary with SIMPLE human replacements.
"""
forbidden_replacements = {
'pivotal': 'important',
'paramount': 'important',
'crucial': 'important',
'vital': 'key',
'essential': 'needed',
'fundamental': 'basic',
'cornerstone': 'foundation',
'foundational': 'basic',
'multifaceted': 'complex',
'nuanced': 'detailed',
'intricate': 'complicated',
'comprehensive': 'complete',
'holistic': 'whole',
'delve': 'examine',
'leverage': 'use',
'utilize': 'use',
'utilise': 'use',
'facilitate': 'help',
'enhance': 'improve',
'foster': 'support',
'cultivate': 'develop',
'empower': 'enable',
'robust': 'strong',
'rigorous': 'thorough',
'substantial': 'large',
'significant': 'important',
'considerable': 'large',
'notable': 'important',
'myriad': 'many',
'plethora': 'many',
'array': 'range',
'spectrum': 'range',
'gamut': 'range',
'breadth': 'range',
'endeavor': 'effort',
'endeavour': 'effort',
'notion': 'idea',
'realm': 'area',
'facet': 'aspect',
'dimension': 'part',
'paradigm': 'model',
'framework': 'structure',
'landscape': 'field',
'trajectory': 'path',
'dynamic': 'changing',
'optimal': 'best',
'viable': 'workable',
'integral': 'key',
'inherent': 'natural',
'quintessential': 'typical',
'ubiquitous': 'common',
'underscores': 'shows',
'showcases': 'displays',
'exemplifies': 'shows',
'encapsulates': 'captures',
'illuminates': 'reveals',
'elucidates': 'explains',
'necessitates': 'requires',
'encompasses': 'includes',
'embodies': 'represents',
'perpetuates': 'continues',
'mitigates': 'reduces',
'exacerbates': 'worsens',
'bolsters': 'supports',
'galvanizes': 'energizes',
'transcends': 'goes beyond',
'permeates': 'spreads through',
'resonates': 'connects',
'navigating': 'dealing with',
'harnessing': 'using',
'spearheading': 'leading',
'pioneering': 'starting',
'orchestrating': 'organizing',
'attribute': 'quality',
'capacity': 'ability',
'capability': 'ability',
'proficiency': 'skill',
'competency': 'skill',
'milieu': 'environment',
'domain': 'field',
'sphere': 'area',
'arena': 'field',
'context': 'setting',
'inculcates': 'teaches',
'instills': 'creates',
'engenders': 'creates',
'fortifies': 'strengthens',
'practitioners': 'professionals',
'individuals': 'people',
'steadfast': 'steady',
'unwavering': 'constant',
'persistent': 'continuing',
'diligent': 'careful',
'flourishes': 'grows',
'blossoms': 'develops',
'dissipating': 'fading',
'constraint': 'limit',
'empowerment': 'power',
'deliberatively': 'deliberately',
'amidst': 'among',
'amongst': 'among',
'brimming': 'full',
'waning': 'decreasing',
'fleeting': 'brief',
'momentary': 'brief',
'surmount': 'overcome',
'succumbing': 'giving in',
'propelling': 'pushing',
'signifies': 'means',
'equips': 'prepares',
'dictate': 'control',
'destiny': 'future',
'groundwork': 'foundation',
'immense': 'large',
'significance': 'importance',
'unconscious': 'automatic',
'procedures': 'processes',
'cognitive': 'mental',
'strain': 'effort',
'beneficial': 'good',
'customs': 'habits',
'detrimental': 'harmful',
'insidiously': 'quietly',
'cumulative': 'combined',
'devotion': 'commitment',
'preservation': 'keeping',
'correlation': 'connection',
'anticipatibility': 'predictability',
'consciously': 'deliberately',
'engineering': 'designing',
'progressive': 'steady',
'evolution': 'progress',
'profound': 'important',
'groundbreaking': 'new',
'revolutionary': 'new',
'unprecedented': 'new',
'transformative': 'changing',
'deeply': 'very',
'highly': 'very',
'extremely': 'very',
'incredibly': 'very',
'remarkably': 'notably',
'exceptionally': 'notably',
'extraordinarily': 'notably',
'moreover': 'also',
'furthermore': 'also',
'additionally': 'also',
'thus': 'so',
'hence': 'so',
'thereby': 'by doing this',
'wherein': 'where',
'whereby': 'by which',
}
for forbidden, replacement in forbidden_replacements.items():
pattern = r'\b' + re.escape(forbidden) + r'\b'
text = re.sub(pattern, replacement, text, flags=re.IGNORECASE)
return text
# Step 9: Post-Processing
def polish_academic_text(text):
"""
Final polish for academic text.
"""
contractions = {
"don't": "do not", "doesn't": "does not", "didn't": "did not",
"can't": "cannot", "couldn't": "could not", "wouldn't": "would not",
"shouldn't": "should not", "won't": "will not", "isn't": "is not",
"aren't": "are not", "wasn't": "was not", "weren't": "were not",
"haven't": "have not", "hasn't": "has not", "hadn't": "had not",
"it's": "it is", "that's": "that is", "there's": "there is",
}
for contraction, full_form in contractions.items():
text = text.replace(contraction, full_form)
text = text.replace(contraction.capitalize(), full_form.capitalize())
return text
# ═══════════════════════════════════════════════════════════════════════════════
# GRADIO WEB INTERFACE
# ═══════════════════════════════════════════════════════════════════════════════
def convert_text_gradio(ai_text):
"""Gradio wrapper function"""
if not ai_text.strip():
return "⚠️ Please enter some text to convert!"
try:
accent = "usa"
humanized = humanize_academic_text(ai_text, accent)
humanized = polish_academic_text(humanized)
return humanized
except Exception as e:
import traceback
return f"❌ Error: {str(e)}\n\n{traceback.format_exc()}"
# Custom CSS
custom_css = """
.button-row {
justify-content: center !important;
}
.gradio-container {
max-width: 1200px !important;
margin: auto !important;
}
"""
# Create Gradio Interface (Compatible with all Gradio versions)
demo = gr.Blocks()
with demo:
gr.HTML("""
<div style='background: linear-gradient(135deg, #667eea 0%, #764ba2 100%);
padding: 40px 20px;
border-radius: 15px;
margin-bottom: 30px;
box-shadow: 0 10px 30px rgba(0,0,0,0.2);'>
<h1 style='color: white;
font-size: 42px;
margin-bottom: 15px;
text-align: center;
text-shadow: 2px 2px 4px rgba(0,0,0,0.3);'>
🎓 Academic Text Humanizer
</h1>
<p style='color: #f0f0f0;
font-size: 18px;
text-align: center;
margin: 0;
font-weight: 300;'>
Transform AI-Generated Text into Authentic Academic Writing
</p>
</div>
""")
input_text = gr.Textbox(
label="📝 AI-Generated Text",
lines=10,
placeholder="Paste your AI-generated text here..."
)
with gr.Row():
clear_btn = gr.Button("🗑️ Clear")
submit_btn = gr.Button("✨ Humanize Text")
output_text = gr.Textbox(
label="✨ Humanized Academic Text",
lines=12
)
submit_btn.click(
fn=convert_text_gradio,
inputs=input_text,
outputs=output_text
)
clear_btn.click(
fn=lambda: ("", ""),
inputs=None,
outputs=[input_text, output_text]
)
# Launch the app
if __name__ == "__main__":
demo.launch(
server_name="0.0.0.0",
server_port=7860
) |