Public/automatic
1
0
Fork 0
mirror of https://github.com/vladmandic/automatic synced 2026-09-19 01:04:32 +02:00
Code Issues Packages Projects Releases Wiki Activity

fix(caption): remove dead min_length param, split Florence/PromptGen prompts, fix gaze detection

Browse Source 
- 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
This commit is contained in:
CalamitousFelicitousness
2026-01-29 01:21:28 +00:00
parent fba942b25e
commit bf7a72f12e
6 changed files with 214 additions and 124 deletions
Download Patch File Download Diff File Expand all files Collapse all files
modules/api/endpoints.py
+3 -5
View File
@@ -145,8 +145,6 @@ def post_caption(req: models.ReqCaption):
raise HTTPException(status_code=404, detail="Model not found")
# Build clip overrides from request (only include non-None values)
clip_overrides = {}
if req.min_length is not None:
clip_overrides['min_length'] = req.min_length
if req.max_length is not None:
clip_overrides['max_length'] = req.max_length
if req.chunk_size is not None:
@@ -314,8 +312,6 @@ def _dispatch_openclip(req: models.ReqCaptionOpenCLIP) -> models.ResCaptionDispa
raise HTTPException(status_code=404, detail="Model not found")
# Build clip overrides from request
clip_overrides = {}
if req.min_length is not None:
clip_overrides['min_length'] = req.min_length
if req.max_length is not None:
clip_overrides['max_length'] = req.max_length
if req.chunk_size is not None:
@@ -499,7 +495,8 @@ def get_vqa_prompts(model: Optional[str] = None):
**Prompt Categories:**
- Common: Use Prompt, Short/Normal/Long Caption
- Florence: Phrase Grounding, Object Detection, OCR, Dense Region Caption
- Florence: Phrase Grounding, Object Detection, OCR, Dense Region Caption (all Florence models)
- PromptGen: Analyze, Generate Tags, Mixed Caption (MiaoshouAI PromptGen fine-tunes only)
- Moondream: Point at..., Detect all..., Detect Gaze
"""
from modules.caption import vqa
@@ -509,6 +506,7 @@ def get_vqa_prompts(model: Optional[str] = None):
return {
"common": vqa.vlm_prompts_common,
"florence": vqa.vlm_prompts_florence,
"promptgen": vqa.vlm_prompts_promptgen,
"moondream": vqa.vlm_prompts_moondream,
"moondream2_only": vqa.vlm_prompts_moondream2
}