These settings are accessible from the Caption tab and can be saved
as defaults via "Set UI defaults", so they don't need to appear in
Settings > Interrogate.
- 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 thinking_mode/reasoning parameter to enable reasoning mode
- Add Detect Gaze task with placeholder hint
- Parse point/detect results to return annotation data for visualization
- Handle keep_thinking setting: format as "Reasoning:\n...\nAnswer:\n..." or discard
- Add comprehensive debug logging throughout handler
- 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
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)
Add support for Moondream 3 Preview VLM with:
- Text query, caption, point, and detect capabilities
- Bounding box visualization for object detection
- Max pixels setting for resolution control
- Device offloading support
Add new VLM configuration options:
- interrogate_vlm_keep_prefill: Keep prefill text in output
- interrogate_vlm_keep_thinking: Keep reasoning trace in output
Also adjust defaults:
- Change interrogate_clip_flavor_count: 16 -> 1024 with updated range
- Change interrogate_vlm_prompt default to first item ("Use Prompt")
All the failure routes should either:
- Get far enough along to not leave checked values unable to reach the finished state...
- or all be tied into the same AbortController so everything aborts together.
... I think...
Improvements to the OpenCLIP interrogation:
- Sort all ranking dicts by similarity score (descending)
- Add format_category() helper for text formatting
- Add formatted text output for CLIP labels textbox
- Return additional text update in analyze_image()
Two fixes for the JoyCaption handler:
- Only offload model if shared.opts.interrogate_offload is True
- Add max_pixels=1024*1024 to AutoProcessor for consistent image handling
Add new Font Awesome symbols for model capability indicators:
- vision symbol (eye icon) for vision-capable VLM models
- reasoning symbol (lightbulb icon) for thinking/reasoning models
Also fix dropdown font styling by adding NotoSans font-family.
CalamitousFelicitousness
Vladimir Mandic
awsr