gen-cards: regenerate Use-it block

strip_debug_info rescue: 0.4.0 debug locations removed (fixes OS 27 beta 2+ load), weights/graph unchanged

strip_debug_info rescue: 0.4.0 debug locations removed (fixes OS 27 beta 2+ load), weights/graph unchanged

Re-export with coreai-torch 0.4.1 (fixes OS 27 beta 2+ load failure)

Re-export with coreai-torch 0.4.1 (fixes OS 27 beta 2+ load failure)

gen-cards: regenerate Use-it block

README: add gen-cards use-it markers

same-conditions VLM A/B: 51.5->70 tok/s (+36%)

honest in-app decode ~64.6 tok/s

MiniCPM-V-4.6: minicpmv46_vlm_decode_int8hu bundle

MiniCPM-V-4.6: minicpmv46_vision_int8lin bundle

MiniCPM-V-4.6 Core AI card

card: add on-device demo gif (fridge -> recipes; GitHub CDN, not in repo)

card: add M4 Max 224 tok/s

MiniCPM-V-4.6: minicpmv46_vlm_decode_int8lin bundle

MiniCPM-V-4.6: minicpmv46_vision bundle

MiniCPM-V-4.6 Core AI card

initial commit