- Add use_safetensors=True to all 16 model from_pretrained calls to
avoid downloading redundant .bin files alongside safetensors
- Add device property to JoyTag VisionModel so move_model can relocate
it to CUDA (fixes 'ViT object has no attribute device')
- Fix Pix2Struct dtype mismatch by casting float inputs to model dtype
while preserving integer tensor types
- Patch AutoConfig.register with exist_ok=True during Ovis loading to
handle duplicate aimv2 registration on model reload
- Detect Qwen VL fine-tune architecture from config model_type instead
of repo name, fixing ToriiGate and similar third-party fine-tunes
- Change UI default task from Short Caption to Normal Caption, and
preserve it on model switch instead of resetting to Use Prompt
- Add dual-prefill testing across 5 VQA test methods using a shared
_check_prefill helper
- Fix pre-existing ruff W605 in strip_think_xml_tags docstring
- Remove caption_openclip_min_length from settings, API models, endpoints, and UI
(clip_interrogator library has no min_length support; parameter was never functional)
- Split vlm_prompts_florence into base Florence prompts and PromptGen-only prompts
(GENERATE_TAGS, Analyze, Mixed Caption require MiaoshouAI PromptGen fine-tune)
- Add 'promptgen' category to /vqa/prompts API endpoint
- Fix gaze detection: move DETECT_GAZE check before generic 'detect ' prefix
to prevent "Detect Gaze" matching as detect target="Gaze"
- Update test suite: remove min_length tests, fix min_flavors to use mode='best',
add acceptance-only notes, fix thinking trace detection, improve bracket/OCR tests,
split Florence/PromptGen test coverage
Add comprehensive tooltips to Caption tab UI elements in locale_en.json:
- Add new "llm" section for shared LLM/VLM parameters:
System prompt, Prefill, Top-K, Top-P, Temperature, Num Beams,
Use Samplers, Thinking Mode, Keep Thinking Trace, Keep Prefill
- Add new "caption" section for caption-specific settings:
VLM, OpenCLiP, Tagger tab labels and all their parameters
including thresholds, tag formatting, batch options
- Consolidate accordion labels in ui_caption.py:
"Caption: Advanced Options" and "Caption: Batch" shared across
VLM, OpenCLiP, and Tagger tabs (localized to "Advanced Options"
and "Batch" in UI)
- Remove duplicate entries from missing section
Hide all CLiP, VLM, and Tagger settings from Settings > Interrogate page
while keeping them in shared.opts for persistence. Caption Tab UI becomes
the single control point with change handlers that save directly to config.
Changes:
- Hide OpenCLiP, VLM, and Tagger settings with visible=False
- Add change handlers to save settings when UI controls change
- Rename "Booru Tags" tab to "Tagger", update choice labels
- Update interrogate.py to use unified tagger interface with all settings
Add DeepBooru as a model option alongside WD14 models in the Booru Tags
tab, with dynamic UI that disables inapplicable controls.
Changes:
- Create modules/interrogate/tagger.py as unified adapter module
- Add batch, load/unload, get_models functions to deepbooru.py
- Update ui_caption.py to use unified tagger interface
- Consolidate shared tagger settings in shared.py
- Add implementation plan for future settings consolidation
UI behavior:
- Model dropdown shows DeepBooru + all WD14 models
- Character threshold and include rating disabled for DeepBooru
- All controls re-enable when WD14 model selected
Add SmilingWolf's WD14/WaifuDiffusion tagger models for anime/illustration
tagging as a new "Booru Tags" tab in the Caption panel.
- Support 9 models (v2 and v3 variants) via HuggingFace
- ONNX backend chosen due to safetensors v3 variants exhibiting
unacceptable accuracy loss
- Separate thresholds for general/character tags
- Batch processing with progress bar
- Consolidate debug env var to SD_INTERROGATE_DEBUG
Refactor VQA module from module-level globals to a VQA class singleton
pattern with self-contained per-model loading methods.
Changes:
- Add VQA class with model/processor state and detection data storage
- Extract load methods for clean model pre-loading via UI
- Interrogate to return string only; store detection data on instance
- Add vqa_draw.py for bounding box/point annotation utilities
Stub, further transfer of drawing functions to follow
- Update moondream3.py to store detection data on VQA singleton
- Update endpoints.py and ui_caption.py for new return type
- Add interrogate_vlm_thinking_mode setting to save checkbox state
- Update ui_caption to restore Thinking Mode preference on load
- Add blank line before 'Answer:' label for visual separation
- Remove '\n\n' replacement in clean() that stripped blank lines
- Fix Qwen reasoning detection when <think> tag is in prompt, not response
- Add reasoning icon to Moondream 2 and 3 model names
- Add load_model() function to pre-load VLM into memory
- Add unload_model() function to free VLM from memory
- Add Load/Unload buttons to Caption tab UI
CalamitousFelicitousness
Vladimir Mandic
awsr