Spaces:

ArtShumov
/

Whyx-PROmpTea

Running

6.99 MB

Ctrl+K

1 contributor

fix(tagger): caption checkbox is the single source of truth for NL; auto re-run no longer reads chip outputs (Gradio 6 stale-choice crash)

fbb630d 17 days ago

data
fix(caption): Florence-2 with transformers<5, use_cache=False for 4.54+ cache format, dict post-process 17 days ago
scripts
feat(prod): rewrite pipeline + NoobAI + ensemble tagger (3xWD14+DeepDanbooru-ready) + 1000 artists + negative templates + history ext + scoring config 18 days ago
src
feat(tagger): weighted 3x WD14 + JoyTag smart ensemble, richer pose analysis, NL caption tied to Smart mode 17 days ago
tests
feat(tagger): weighted 3x WD14 + JoyTag smart ensemble, richer pose analysis, NL caption tied to Smart mode 17 days ago
theme
feat(prod): rewrite pipeline + NoobAI + ensemble tagger (3xWD14+DeepDanbooru-ready) + 1000 artists + negative templates + history ext + scoring config 18 days ago
.gitattributes

1.52 kB
feat(prod): rewrite pipeline + NoobAI + ensemble tagger (3xWD14+DeepDanbooru-ready) + 1000 artists + negative templates + history ext + scoring config 18 days ago
.gitignore

171 Bytes
feat(tagger): Qwen-VL NL captions, pose extraction (YOLOv8-n), slide-panel model selector 18 days ago
README.md

1.07 kB
feat(prod): rewrite pipeline + NoobAI + ensemble tagger (3xWD14+DeepDanbooru-ready) + 1000 artists + negative templates + history ext + scoring config 18 days ago
app.py

125 kB
fix(tagger): caption checkbox is the single source of truth for NL; auto re-run no longer reads chip outputs (Gradio 6 stale-choice crash) 17 days ago
manifest.toml

146 Bytes
feat(prod): rewrite pipeline + NoobAI + ensemble tagger (3xWD14+DeepDanbooru-ready) + 1000 artists + negative templates + history ext + scoring config 18 days ago
requirements.txt

214 Bytes
fix(caption): Florence-2 with transformers<5, use_cache=False for 4.54+ cache format, dict post-process 17 days ago