AMP_WF_416分辨率_340万迭代次数，可做底丹
AMP模型训练方法我也不知道，听说很难训练
==================== Model Summary ====================
== ==
== Model name: new_AMP ==
== ==
== Current iteration: 3401298 ==
== ==
==------------------ Model Options ------------------==
== ==
== resolution: 416 ==
== face_type: wf ==
== models_opt_on_gpu: True ==
== ae_dims: 320 ==
== inter_dims: 1024 ==
== e_dims: 80 ==
== d_dims: 80 ==
== d_mask_dims: 26 ==
== morph_factor: 0.5 ==
== uniform_yaw: True ==
== blur_out_mask: True ==
== lr_dropout: n ==
== random_warp: False ==
== ct_mode: none ==
== clipgrad: False ==
== autobackup_hour: 0 ==
== write_preview_history: False ==
== target_iter: 0 ==
== random_src_flip: False ==
== random_dst_flip: True ==
== batch_size: 3 ==
== gan_power: 0.008 ==
== gan_patch_size: 60 ==
== gan_dims: 20 ==
== ==
==------------------- Running On --------------------==
== ==
== Device index: 0 ==
== Name: NVIDIA GeForce RTX 3080 Ti ==
== VRAM: 9.20GB ==
== ==
=======================================================
***本内容需购买可见***
大家都是大自然的搬运工
页:
[1]