You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Working with millions of files can be a real problem and lead to a lot (!!!) of lost time. Implementing these helper functions will make sure that one can run the big functions (profile, sample-sc, and train) without any errors that cost days of work.
Here is a list of suggestions for the most helpful test functions:
check images:
This function simply runs over all lines in the index and checks if the images exist. This prevents the profile command form exiting with an error.
Check sc-metadata.csv
This function should run over the sc-metadata file and check if all images exist in the right place. This needs to be run prior to training to prevent crashes during training.
The text was updated successfully, but these errors were encountered:
@jccaicedo just wanted to check in here. I think the functions exist in the branch and they are still usefull.
Maybe we mention them in the Wiki and have them on a side branch? Not sure where to store them.
The checking for black crops is crucial. The other one not so much.
Working with millions of files can be a real problem and lead to a lot (!!!) of lost time. Implementing these helper functions will make sure that one can run the big functions (profile, sample-sc, and train) without any errors that cost days of work.
Here is a list of suggestions for the most helpful test functions:
check images:
This function simply runs over all lines in the index and checks if the images exist. This prevents the profile command form exiting with an error.
Check sc-metadata.csv
This function should run over the sc-metadata file and check if all images exist in the right place. This needs to be run prior to training to prevent crashes during training.
The text was updated successfully, but these errors were encountered: