Views
No views yet
rdm0/autoencoderkl-v9-cfg6-init, finetuned by 1.28M images from ImageNet:--train_size 1280000--train_batch_size 8--resolution 512--param 3-2-256-1024--num_epochs 10--learning_rate 1e-4--disc_learning_rate 1e-4--disc_start 2000--distill_scale 0.1--sample_posterior True