JeffWang987
# test Error
When I run the test set, I always encounter conflicts between float32 and float16.
And I use the mini dataset,and the result of train is behind:
The saving format in the drivedreamer and drivedreamer_ema folders is as follows:
I've been struggling with this issue for quite some time and would be extremely grateful for any assistance from you here.
Excuse me, have you met the problem of NCCL Error (maybe because of num_workers) in training, how did you solve it?
I remember encountering this issue, but I didn't pay much attention to it since it seemed not to affect the training process. Could you please send me the screenshot of the error for convenience?
I've been struggling with this issue for quite some time and would be extremely grateful for any assistance from you here.
Try this, change it in dreamer_model/pipelines/control/pipeline_stable_diffusion_control.py.
I've been struggling with this issue for quite some time and would be extremely grateful for any assistance from you here.
Try this, change it in dreamer_model/pipelines/control/pipeline_stable_diffusion_control.py.
thank you very much, it is so helpful.
I have another idea to solve it like:
Because some weights is float16 , another is float32 .
Thank you again sincerely. I have been learning DriveDreamer recently, and I am from East China Normal University. I was wondering if it would be convenient for you to add me as a friend for more communication. If it's not trouble, you can reach me via email at [email protected].
When I run the test set, I always encounter conflicts between float32 and float16.
And I use the mini dataset,and the result of train is behind:
The saving format in the drivedreamer and drivedreamer_ema folders is as follows:
Hello, the folder for the drivedreamer.ema that I trained on the mini dataset is also like this. During testing, the diffusiont_pytorch_madel.safensors file was missing. Have you resolved this issue in the future?
当我运行测试集时，我总是遇到 float32 和 float16 之间的冲突。
我使用了 mini 数据集，训练的结果落后了：
drivedreamer 和 drivedreamer_ema 文件夹中的保存格式如下：
您好，我在迷你数据集上训练的 drivedreamer.ema 的文件夹也是这样的。在测试期间，缺少 diffusiont_pytorch_madel.safensors 文件。您将来是否解决了这个问题？
我也出现了这个问题，请问您解决了吗
