This is a Lora oriented on Dall-E 3 compositions.

Main trigger words : D4lle, Dall-e style, AI generated

////////////////////

Datasets infos: 30 pictures (Humans images)

• 80% solo / 20% duo

• 74% woman / 26% man

• 50% conceptual, themed, poetic aesthetic / 50% scenes (various contexts)

TIPS: Also useful for fixing bad renderings generated on Dall-E, using this Lora in img2img allows you to best preserve the initial Dall-E style.

____________________

If you want to see my next Lora projects (for Flux) and contribute to launch the model training :

Donations for futurs loras : https://ko-fi.com/nicolas1801

Some details: https://docs.google.com/spreadsheets/d/1PEmIlzhUUGeoPZqRafKOnvxNzR_-a3SH/edit?usp=sharing&ouid=104542247627018372519&rtpof=true&sd=true