There is nothing special about this model mix, I just happen to use it for a lot of the sample images for my LoRAs. I'm uploading it for those who want to replicate those samples.
The recipe is long lost, but I'm fairly certain it's a merge of one or two furry models along with some version of AbyssOrangeMix.
There is no specific VAE for this model; try a few and see what works best for you.
NOTE: This is a .safetensors conversion of the original .ckpt model. Results should be identical but the model hash might not match my LoRA sample images.