Overall, how is this automated tagging looking so far from y'all's experience? I imagine it probably does a decent job overall for the major tags, but how does it look with the more specific ones?

could you make an update on it? it worked for me some time ago but i deleted it from my computer at that time and when i try now, it just wont work anymore.
seems to me that the ui.py and settings_bk.json are just outdated but i'm not sure.

Looks like toriato's branch of the tagger extension is no longer being maintained. If toriato's version isn't working, you may need to switch to picobyte's branch of the extension:
https://github.com/picobyte/stable-diffusion-webui-wd14-tagger

New instructions:
1) Install Automatic1111's stable diffusion, the tagger extension, the latest version of BK tagging tool. Then download these modifications for BK's tagging tool:

Quote from: neronero on Jun 21, 2023, 12:24 PMModified files for BK's tagging tool
(Adds more "variants" to the tagging tool so that relevant booru tags are recognised)
settings.rpy and tagset_extended.rpy - replaces files in your BK tagger's game directory:
Picture_namer_for_BK-2.8-pc\game\...

2) Go to Automatic1111's Settings, Tagger, uncheck "Use spaces instead of underscore" and press apply settings.
3) Now open the Tagger extension's tab, set Weight threshold and Min tag fraction to 1.0 (for some reason this version of the tagger seems way more eager to add loads of tags?)
4) Under "Keep tag", add the following:
profile, card, quest, shop, full_body, portrait, upper_body, waitress, cooking, dancer, dancing, masseuse, massage, bath, geisha, etiquette, rest, sitting, lying, sleeping, naked, nude, service, sex, fuck, fucking, anal, anal sex, fetish, hardcore, lesbian, yuri, yaoi, bisexual, multiple_girls, bisexual_female, group, group_sex, multiple_boys, big, stallion, large_penis, huge_penis, stomach_bulge, beast, bestiality, machine, robot, monster, happy, laugh, neutral, sad, crying, cosplay, roleplay, role, maid, bunny, bunny_ears, swim, swimsuit, kimono, geisha, strip, stripping, undressing, sing, singing, fight, combat, battle, fighting, fighting_stance, libido, full-face_blush, erect_nipples, drunk, sensitivity, caress, headpat, hug, obedience, cleaning, broom, constitution, run, sport, exercise, public, outdoors, exhibitionism, crowd, onlookers, toy, sex_toy, object_insertion, dildo, dild, vibrator, vibr, beads, bead, anal_beads, plug, butt_plug, anal_tail, strap-on, strap, dirty, wet, oil, dom, femdom, sub, humiliation, humiliated, humiliate, insults, kiss, kissing, fondle, fondling, fondled, grabbing, grope, groped, groping, ass_grab, finger, masturbate, handjob, hj, penis worship, penis-worship, penis_worship, penisworship, titjob, paizuri, footjob, fj, oral, blowjob, bj, fellatio, deep, irru, dt, irrumaito, deepthroat, sixty-nine, 69, doggy, doggystyle, cowgirl, cowgirl_position, piledriver, upside-down, spooning, spoon, double, double_penetration, bondage, bdsm, restrained, tied_up_(sexual), gag, spanking, spank, spanked, slap_mark, rimming, anilingus, fisting, lactation, lactate, watersports, peeing, enema, cumshot, cum, on-body, cum_on_body, cum_on_breasts, on-face, facial, cum_on_face, in-hair, cum_on_hair, in-mouth, swallow, cum_in_mouth, inside, cum_in_ass, cum_in_pussy, internal_cumshot, creampie, cream, aftersex, after_sex, cumdrip, bukkake, cum-shower, cum shower, orgasm, denied, orgasm_denial, squirt, female_ejaculation, market, slave, cage, prison, tied_up_(nonsexual), advertise, cheerleader, holding_sign, spoken_blush, ceremony, nun, praying, own_hands_together, date, dating, couple, eating, party, 6+girls, 5girls, embar, embarrassed, angry, anger_vein, clenched_teeth, refuse, rape, molestation, hurt, injury, pain, sick, apron, china, cheongsam, chinese_clothes, dress, hooker, microskirt, pasties, revealing_clothes, miko, casual, student, school_uniform, teacher, formal, nurse, santa, christmas, bride, wedding_dress, catgirl, cat_ears, cat_girl, study, reading, classroom, writing, beach, nature, garden, forest, flower_field, grass, town, building, city, street, road, blindfold, collar, futanari, cunnilingus, tempt, seductive_smile, presenting, friend, virgin, defloration, pregnant, lingerie, libido, panties, tentacles, tentacle, uncen, uncensored, sprite, transparent_background, hq, lq, bad_anatomy, 1girl
5) Add a directory to the "Input directory" field, select an interrogator (I usually go with WD ViT v2) and press interrogate
6) If everything worked, you've now got text files text to those images. Use my rename program to rename the images using the generated text files.
7) Bring the renamed files to the tagging tool which should translate the booru-tags into BK-tags