AMP256_wf_610万迭代美女高参万能底丹，高参模型分享

Sheldon1

我来试试

Sheldon1

[系统时间][迭代总数][系统延时][src损失][dst损失]

Error:
Traceback (most recent call last):
  File "C:\baidunetdiskdownload\DFL_RTX3000_series_2024\_internal\DeepFaceLab\mainscripts\Trainer.py", line 176, in trainerThread
    model_save()
  File "C:\baidunetdiskdownload\DFL_RTX3000_series_2024\_internal\DeepFaceLab\mainscripts\Trainer.py", line 68, in model_save
    model.save()
  File "C:\baidunetdiskdownload\DFL_RTX3000_series_2024\_internal\DeepFaceLab\models\ModelBase.py", line 393, in save
    self.onSave()
  File "C:\baidunetdiskdownload\DFL_RTX3000_series_2024\_internal\DeepFaceLab\models\Model_AMP\Model.py", line 638, in onSave
    model.save_weights ( self.get_strpath_storage_for_file(filename) )
  File "C:\baidunetdiskdownload\DFL_RTX3000_series_2024\_internal\DeepFaceLab\core\leras\layers\Saveable.py", line 61, in save_weights
    d_dumped = pickle.dumps (d, 4)
MemoryError

Sheldon1

Sheldon1 发表于 2025-12-18 18:26
[系统时间][迭代总数][系统延时][src损失][dst损失]

Error:

100G虚拟内存还是说的太保守了，我直接加了200G，这次再看看会不会爆

Sheldon1

Sheldon1 发表于 2025-12-18 19:51
100G虚拟内存还是说的太保守了，我直接加了200G，这次再看看会不会爆

拉倒200G，一下就稳住了。