Location
Badges
Activity
Ratings Progression
Challenge Categories
Participant | Rating |
---|
Participant | Rating |
---|
-
mission-impossible Novartis DSAI ChallengeView
Novartis DSAI Challenge
How to access error logs?
About 5 years agoI tried submitting a new model, and even though it works on test data from /shared_data/data/test_data_sample, it fails while evaluating.
I can see the error, but without detailed information I have no idea what is going on, I only get from my Issue:
Submission failed : list index out of range
which is not helpful at all.
It also says that error logs are not accessible for users, and that βOne of the admins will provide you access to the relevant section of the log that you might require for debugging.β
Can I then get access to my log? I even resubmitted the model with debug: True flag in aicrowd.json but it didnβt change anything.
Also, is this the only way to debug evaluation errors? If so, I canβt fathom trying anything advanced and then requesting access and waiting for it each time I have to debug.
Duplicates of intClinicalTrialID
About 5 years agoHi. While investigating the datasets in detail, we found out that there are multiple rows with the same clinical trial id, and just with different indications and outcomes. Since we are trying to predict the probability of a success of a particular trial, shouldnβt all clinical trial IDs be unique? Because now it seems that we are trying to predict the success of a particular indication/trial pair, and not the trial itself, which changes a lot.
Also, there are 506 clinical trial IDs that are the same in the test and train dataset, which might cause a data leak problems.