-
Notifications
You must be signed in to change notification settings - Fork 1.1k
I don't know why this procsee wrong #1696
Replies: 1 comment · 5 replies
-
class |
Beta Was this translation helpful? Give feedback.
All reactions
-
i read refer, but i don't think fix my problem. please help me.
and my config file is model settingsmodel = dict( dataset settingsdataset_type = 'CustomDataset', train_pipeline = [ test_pipeline = [ train_dataloader = dict( val_dataloader = dict( test_dataloader = dict( optimizeroptim_wrapper = dict( learning policyparam_scheduler = dict( train, val, test settingtrain_cfg = dict(by_epoch=True, max_epochs=25, val_interval=1) NOTE:
|
Beta Was this translation helpful? Give feedback.
All reactions
-
i solve this problem. the error code is cause the split=' ' is not exist the custom.py class CustomDataset. init Why is this happening? i don't think so. Anyway, thanks for the answer |
Beta Was this translation helpful? Give feedback.
All reactions
-
how can you solve this problem? I meet this question, too. |
Beta Was this translation helpful? Give feedback.
All reactions
-
insert split=' ' in to dataset/custom.py like this def init(self,
and you can solve this problem |
Beta Was this translation helpful? Give feedback.
All reactions
-
👍 2
-
Hello! I meet this question, too. I added the line "split=' '," as you did, but a new error message appears: |
Beta Was this translation helpful? Give feedback.
-
Hi, I using MMpretrain. and i use Resnet34 and i modified it. anyway, I was using it well until 3 days ago, when suddenly I ran train.py and got an error. TypeError: class
CustomDataset
in mmpretrain/datasets/custom.py: BaseDataset.init() got an unexpected keyword argument 'split'. i don't torch any other code. and i trying to find error. but the base_dataset.py and custom.py code not have problem. how can i fix? Did something happen to your split in the meantime?Beta Was this translation helpful? Give feedback.
All reactions