You can train the Output as the Input and vice versa :)

kontext already how to make someone fat. You just need to prompt for.it... but it is not able to make someone fat slim. So no need to train a Lora to enlarge someone.

why don't you train it yourself !!!!

p_p trying to find (how to do it the best way) free and online, then i will do

Kontext делает толстяков обычным запросом "Modify this image to make the person appear visibly heavier while maintaining a natural and realistic look. The edited version should preserve the original's integrity, lighting, color palette, and overall style, clearly reflecting the subject's increased weight. Follow these steps: - Analyze the image: Understand the subject's pose, lighting, and perspective. - Proportionally enlarge body parts: Increase the size of areas like the abdomen, arms, face, and legs. - Maintain recognizable identity: Ensure the person remains identifiable and avoid unnatural or cartoonish distortion. - Keep background and elements unchanged: Do not alter the background or other objects in the image." А вот из толстяка в худого человека не превращает никак. Именно по этому я попросил автора создать подобную лору, за что ему огромная благодарность, ибо работает она отлично, сам три раза пробовал тренировать, ничего не получилось...

dude you just into fat bitches man..

realmichael not fat , i need them real like real girls how they look ..

Why would we need AI for that? We can just go outside.