0 Images Loaded
AI: Not connected
LOCAL browser storage only. SHIFT+CLICK multi-select, CTRL+CLICK toggle.
Quick start:
Loaded Images and unsaved captions are LOST on refreshing the browser! But The state of (AI) settings, saved system prompts... IS SAVED OVER SESSIONS. Until you clear local storage! (top right)
Loading: LOAD or DRAG images and matching .txt files in. You can add and remove Images at any time. NEW: Export ZIP with caption .txt only, OR all images AND .txt as one zip, training-ready!
Update: OPEN from and SAVE into Folder does NOT work anymore on HF space! Only if run LOCALLY (and then only in Chrome(ium), for that downlaod simply index.html under files!
USAGE: Select image/s (shift, ctrl+click for mutliple). IF you import existing captions (txt files with the exact same name as the image), the text will be broken into tags, and added to the list, on every COMMA. Same when "apply" AI captions. For individual images, Enter a tag or line or whatever and press enter, Each will appear on the right and can be added to other images, edited (for all instances). Or add tags directly to the list. It's fairly self explanatory. On exporting, the parts are separated by commas.
Local AI: ADDED functionality: Use a local LMstudio Server with loaded vision model to caption images with AI. If you use LM Server default settings it should just work. Set a system prompt under AI Caption s´Settings. Up to 10 can be stored. You can tell it to caption however you want or need (And how to use commas, for splitting). BTW the system prompt can be anything, its not limited to captioning. You can tell it to do anything with the image. For example mass creating image/video generation/editing prompts from the images, instaed of a caption. The AI and Network functionality can be completely turned off and is by default.
NOTE: LMStudio only loads a new selected model, once you generate something, not before.
Loading: LOAD or DRAG images and matching .txt files in. You can add and remove Images at any time. NEW: Export ZIP with caption .txt only, OR all images AND .txt as one zip, training-ready!
Update: OPEN from and SAVE into Folder does NOT work anymore on HF space! Only if run LOCALLY (and then only in Chrome(ium), for that downlaod simply index.html under files!
USAGE: Select image/s (shift, ctrl+click for mutliple). IF you import existing captions (txt files with the exact same name as the image), the text will be broken into tags, and added to the list, on every COMMA. Same when "apply" AI captions. For individual images, Enter a tag or line or whatever and press enter, Each will appear on the right and can be added to other images, edited (for all instances). Or add tags directly to the list. It's fairly self explanatory. On exporting, the parts are separated by commas.
Local AI: ADDED functionality: Use a local LMstudio Server with loaded vision model to caption images with AI. If you use LM Server default settings it should just work. Set a system prompt under AI Caption s´Settings. Up to 10 can be stored. You can tell it to caption however you want or need (And how to use commas, for splitting). BTW the system prompt can be anything, its not limited to captioning. You can tell it to do anything with the image. For example mass creating image/video generation/editing prompts from the images, instaed of a caption. The AI and Network functionality can be completely turned off and is by default.
NOTE: LMStudio only loads a new selected model, once you generate something, not before.