The Flux Ghibsky Illustration model creates stunning landscapes that blend serene, surreal skies with intricate, Ghibli-inspired details. This fusion captures both Ghibli’s whimsical charm and Makoto Shinkai’s atmospheric beauty, perfect for dreamy visuals.
V2 is here!
The latest Flux Ghibsky LoRA model has been trained on more images, including faces and synthetic data from V1. While it appears smoother, I have mixed feelings about whether it’s truly an improvement. Training configurations and captioning on the platform that I use can be unpredictable, which may have also influenced the results.
Feedback is always welcome!
Description
Trained on more face data and hand-picked synthetic data
Trained on 80 images compared to 35 in v1
Trained with 1600 steps compared to 1000 in v1
FAQ
Comments (3)
This is an amazing Lora and I am glad you finally added it to Civit
Thanks! I see you created some amazing work mixing it with other LoRAs.
I also added v2 which was trained on more images, had more face data and some synthetic data. But I was not sure if it is better, I had some mixed feelings about it. Still learning about all this.
@aleksa_codes I will download V2 and give it a try.