Closed liaozhihao2757 closed 2 years ago
Sorry for missing this part. Maybe fixed by upload this part. Welcome try again.
Is this related to the pytorch version, I am using the pytorch 1.9 and cuda 11.3
No. It just caused by code missing.
However, if y don't use tensorboard or save scan in valid, this part is useless.
okk, thank you very much.
Traceback (most recent call last): File "train.py", line 133, in
trainer.train()
File "/home/buaa/project_liaozhihao/CENet-main/modules/trainer.py", line 336, in train
Trainer.save_to_log(logdir=self
AttributeError: type object 'Trainer' has no attribute 'save_to_log'
I found this issue when the code is training and this “Trainer.save_to_log” is not found in the Class Trainer.