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

Browse Source

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)

...

This commit is contained in:

CalamitousFelicitousness

2025-12-04 00:04:09 +00:00

parent 0a322c0faf

commit 27fa48cc99

1 changed files with 683 additions and 88 deletions

Show all changes Ignore whitespace when comparing lines Ignore changes in amount of whitespace Ignore changes in whitespace at EOL

Download Patch File Download Diff File Expand all files Collapse all files

modules/interrogate/vqa.py

+683 -88

File diff suppressed because it is too large Load Diff