vaishnosingh
|
72ea6cd59d
|
feat(image-to-text): explicitly report 4-bit (NF4) quantization status in telemetry and model status
|
2026-10-29 13:42:54 +05:30 |
|
vaishnosingh
|
0db0a4c1de
|
feat(image-to-text): switch default model to Qwen2.5-VL-7B-Instruct with 3B cached fallback and cpu offload support
|
2026-10-29 13:33:46 +05:30 |
|
vaishnosingh
|
47c35da17e
|
feat(image-to-text): clean json export to image_file and prompt only, sort queue sequentially by filename, and auto-resize images >720px on upload
|
2026-10-29 13:27:57 +05:30 |
|
vaishnosingh
|
5ae43c479b
|
feat(image-to-text): unify single/multi/folder intake into studio, chunked streaming queue, and fixed download options
|
2026-10-29 12:49:01 +05:30 |
|
vaishnosingh
|
457ccadd83
|
feat(image-to-text): add Qwen2.5-VL Vision AI & Prompt Studio with liquid glass UI
|
2026-10-29 12:39:10 +05:30 |
|