Closed DonghyunAhn closed 4 years ago
Because evaluation is performed in an independent container without a train dataset,
Changing the code to not access the train dataset
before the nsml.pause()
function seems to solve the problem.
Then do we need to run another session to modify main.py here? Or is there any way to change the code that already uploaded on previous session?
You can change code that already uploaded session.
Please check this issue
Hello, this is team 8, and we tried to test simple model we made by nsml run. However, when we evaluate it, the error which was not rasied during traing step occurs about file path. This is what we received from error message:
How can we handle this error?