LIAE-UD_224_wf_306万迭代预训练模型
本帖最后由 jinyan 于 2023-9-13 18:16 编辑模型太高跑不了,只能用224模型,特此分享
当预训练设置关闭时,所有预训练模型都从 0 开始。
只有预训练模型,在预训练一段时间后运行,然后上传有一个计数迭代
但是训练好的ITER仍然存在,即使在0开始,也会更快的学习进度
高清数据集对练,300万预训练模型,真正的预训练模型,如果自己训练专丹,迭代次数会从0开始,就像自闭2.0神丹一样
目前预训练的效果基本到顶了
================== Model Summary ===================
== ==
== Model name: 224wf_SAEHD ==
== ==
== Current iteration: 3006311 ==
== ==
==---------------- Model Options -----------------==
== ==
== resolution: 224 ==
== face_type: wf ==
== models_opt_on_gpu: True ==
== archi: liae-ud ==
== ae_dims: 256 ==
== e_dims: 64 ==
== d_dims: 64 ==
== d_mask_dims: 16 ==
== masked_training: True ==
== uniform_yaw: True ==
== 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: True ==
== pretrain: True ==
== autobackup_hour: 2 ==
== write_preview_history: False ==
== target_iter: 0 ==
== random_flip: True ==
== batch_size: 8 ==
== eyes_mouth_prio: True ==
== blur_out_mask: False ==
== adabelief: True ==
== random_hsv_power: 0.0 ==
== gan_patch_size: 3 ==
== gan_dims: 16 ==
== random_src_flip: False ==
== random_dst_flip: False ==
== ==
==------------------ Running On ------------------==
== ==
== Device index: 0 ==
== Name: NVIDIA GeForce RTX 3060 ==
== VRAM: 9.37GB ==
== ==
***本内容需购买可见***
当预训练设置关闭时,所有预训练模型都从 0 开始。
只有预训练模型,在预训练一段时间后运行,然后上传有一个计数迭代
但是训练好的ITER仍然存在,即使在0开始,你也会认识到更快的学习进度 高手,膜拜中。。。。。。。。。。。。。 感谢楼主分享 感谢分享好模型,存钱购买中。。 感谢楼主分享:D:D:D:D 谢谢分享 感谢分享好模型,存钱购买中。。 签到;P;P;P;P;P;P;P学习 感谢楼主分享:)!!!
页:
[1]
2