dfl2.0_WF_512_DF-UD FFHQ 260w 女性头像预训练所有角度不再更新

设置100劝退后续，前期已买的不受影响。
——————————————————————————特别说明：拿我丹做专人正训练的，欢迎分享，不限制使用。但是，拿来再做预训练的，就不要分享了，再预训练价值不大。
——————————————————————————
针对FFHQ 3w女性头像进行的训练，以及部分不可描述女性头像。
0-185w，FFHQ+不可描述女性头像，bs12.
后15w是bs10.
200w-217w，继续补齐了所有角度并训练出牙齿、眼睛细节。bs4.-260w，bs8。再训练已经没有太大价值。不走捷径，正训练bs2，5w可见牙缝（3090应该4小时），如果相似性高的话。
=============== Model Summary ===============
==                                        ==
==          Model name: default_SAEHD  ==
==                                        ==
==    Current iteration: 2000000       ==
==                                        ==
==————- Model Options ————-==
==                                        ==
==          resolution: 512             ==
==          face_type: wf             ==
==    models_opt_on_gpu: True          ==
==                archi: df-ud          ==
==             ae_dims: 256             ==
==             e_dims: 64             ==
==             d_dims: 64             ==
==          d_mask_dims: 22             ==
==    masked_training: False          ==
==          eyes_prio: False          ==
==          uniform_yaw: True          ==
==          adabelief: False          ==
==          lr_dropout: n             ==
==          random_warp: False          ==
==          gan_power: 0.0             ==
==    true_face_power: 0.0             ==
==    face_style_power: 0.0             ==
==       bg_style_power: 0.0             ==
==             ct_mode: none          ==
==             clipgrad: False          ==
==             pretrain: True          ==
==    autobackup_hour: 1             ==
== write_preview_history: False          ==
==          target_iter: 0             ==
==          random_flip: False          ==
==          batch_size: 10             ==
==                                        ==
==————– Running On —————==
==                                        ==
==       Device index: 0             ==
==                Name: GeForce RTX 3090 ==
==                VRAM: 24.00GB       ==
==                                        ==
=============================================