First 9 images are my usual model, last 7 images are with a new model that's not trained on any CGI or anime images so it doesn't understand as many concepts and tends to introduce more chaotic backgrounds but I think the results are more realistic. Let me know if you prefer this new direction and which ones are your favourites. Personally I like number 13 the most.
kligoboro
2024-11-22 10:00:07 +0000 UTCStefan Cartbo
2024-11-17 17:21:32 +0000 UTCLawrence
2024-11-17 08:16:53 +0000 UTCshark127
2024-11-16 06:59:26 +0000 UTCWcktex
2024-11-15 16:28:33 +0000 UTC