Views
No views yet
rdm0/autoencoderkl-v9-cfg6w-init, finetuned by 1.28M images from ImageNet:--train_size 1280000--train_batch_size 8--resolution 256--param 3-2-64-256--num_epochs 3--learning_rate 1e-4--disc_learning_rate 1e-4--disc_start 2000--distill_scale 0.1--sample_posterior True