[flux-lora]训练失败：Cannot copy out of meta tensor; no data! Please use torch.nn.Module.to_empty() instead of torch.nn.Module.to() when moving module from meta to a different device.

[AhBumm]

2024-08-25 03:19:15 INFO move vae and unet to cpu to save memory flux_train_network.py:156
Traceback (most recent call last):
File "/home/lmx/lora-webui/./sd-scripts/flux_train_network.py", line 411, in
trainer.train(args)
File "/home/lmx/lora-webui/sd-scripts/train_network.py", line 390, in train
self.cache_text_encoder_outputs_if_needed(args, accelerator, unet, vae, text_encoders, train_dataset_group, weight_dtype)
File "/home/lmx/lora-webui/./sd-scripts/flux_train_network.py", line 160, in cache_text_encoder_outputs_if_needed
unet.to("cpu")
File "/home/lmx/lora-webui/venv/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1174, in to
return self._apply(convert)
File "/home/lmx/lora-webui/venv/lib/python3.10/site-packages/torch/nn/modules/module.py", line 780, in _apply
module._apply(fn)
File "/home/lmx/lora-webui/venv/lib/python3.10/site-packages/torch/nn/modules/module.py", line 805, in _apply
param_applied = fn(param)
File "/home/lmx/lora-webui/venv/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1167, in convert
raise NotImplementedError(
NotImplementedError: Cannot copy out of meta tensor; no data! Please use torch.nn.Module.to_empty() instead of torch.nn.Module.to() when moving module from meta to a different device.
03:19:16-424748 ERROR Training failed / 训练失败

节省显存相关设置截图

[Akegarasu]

#477

[yohototo]

你好，请问解决了吗