Increased the size of the dna_with_split.csv file. Now is bigger that the requested batch sizes. Not to nig to make everything slow.
Changed the check_model process to better reflect the tune step later on in the pipeline. Now it is actually runing the tuning as the tuning step does. It just does it for one epoch for one step and (in the future) on one batch. The number of samnples is user specifiable to have a better representation of combination of hyperparameters choosable from the cinfig yaml of the tune.
Increased the size of the dna_with_split.csv file. Now is bigger that the requested batch sizes. Not to nig to make everything slow.
Changed the check_model process to better reflect the tune step later on in the pipeline. Now it is actually runing the tuning as the tuning step does. It just does it for one epoch for one step and (in the future) on one batch. The number of samnples is user specifiable to have a better representation of combination of hyperparameters choosable from the cinfig yaml of the tune.