CalamitousFelicitousness 27fa48cc99 feat(vqa): major VQA handler refactor with prefill, thinking, and visualization ...

Comprehensive overhaul of the VQA interrogation system including:
- Prefill text support for guiding VLM responses
- Thinking mode support with tag cleanup/retention
- Dynamic prompt/task selection based on model type
- Bounding box visualization for detection results
- Debug infrastructure (SD_VQA_DEBUG env var)
- New model support: MiMo-VL, Nidum Gemma, Allura Gemma
- Model-specific prompt lists (Florence, Moondream)

2025-12-05 00:00:24 +00:00

..

deepbooru_model.py

refactor interrogate/analyze/vqa code

2025-02-01 11:47:20 -05:00

deepbooru.py

add prompt enhance

2025-03-28 14:05:28 -04:00

deepseek.py

add prompt enhance

2025-03-28 14:05:28 -04:00

interrogate.py

improve handling of wan22 stages

2025-07-30 11:22:08 -04:00

joycaption.py

fix(vqa): respect offload setting in JoyCaption, add max_pixels

2025-12-02 21:46:09 +00:00

joytag.py

add ras-sd35 experimental

2025-02-18 18:47:42 -05:00

moondream3.py

feat(vqa): add Moondream 3 Preview handler

2025-12-05 00:00:24 +00:00

openclip.py

fix(vqa): sort CLiP analysis results and add text output

2025-12-02 21:48:09 +00:00

vqa.py

feat(vqa): major VQA handler refactor with prefill, thinking, and visualization

2025-12-05 00:00:24 +00:00