I'm at 223K epochs right now. the trained face still looks i little blurry. there's definite separation between teeth and moles are starting to become clearer. but the trained image still doesn't look at sharp as the source or the destination. i want the output face to have the same (or close to) graininess the source or the destination has.
Example taken from: https://mrdeepfakes.com/forums/thread-deepfacelab-explained-and-usage-tutorial
===== Model summary =====
== Model name: SAE
==
== Current epoch: 178215
==
== Model options:
== |== write_preview_history : True
== |== batch_size : 16
== |== sort_by_yaw : True
== |== random_flip : False
== |== resolution : 128
== |== face_type : f
== |== learn_mask : False
== |== archi : df
== |== ae_dims : 900
== |== ed_ch_dims : 50
== |== lighter_encoder : False
== |== multiscale_decoder : False
== |== pixel_loss : True
== |== face_style_power : 0.0
== |== bg_style_power : 0.0
== Running on:
== |== [0 : GeForce GTX 1070 Ti]
=========================
Example taken from: https://mrdeepfakes.com/forums/thread-deepfacelab-explained-and-usage-tutorial
===== Model summary =====
== Model name: SAE
==
== Current epoch: 178215
==
== Model options:
== |== write_preview_history : True
== |== batch_size : 16
== |== sort_by_yaw : True
== |== random_flip : False
== |== resolution : 128
== |== face_type : f
== |== learn_mask : False
== |== archi : df
== |== ae_dims : 900
== |== ed_ch_dims : 50
== |== lighter_encoder : False
== |== multiscale_decoder : False
== |== pixel_loss : True
== |== face_style_power : 0.0
== |== bg_style_power : 0.0
== Running on:
== |== [0 : GeForce GTX 1070 Ti]
=========================