This LoRA model was finetuned on an extremely diverse dataset of 360° equirectangular projections with 2104 captioned training images, using the Stable Diffusion v1-5 model.
https://civitai.com/models/26815