LORA trained on mixed dataset mostly anime. Most likely need more training but I like the small changes it makes on low strength.
Recommented on 0.5 strength.
Samples made at 1024 x 1536 resolution 12 steps
Any feedbacks are welcome.
LORA trained on mixed dataset mostly anime. Most likely need more training but I like the small changes it makes on low strength.
Recommented on 0.5 strength.
Samples made at 1024 x 1536 resolution 12 steps
Any feedbacks are welcome.