Open vitst opened 4 years ago
@vitst
No, you are right on this. There is no specific reason why different data are loaded on different model. Everything is interchangeable. My intention at some point was to make the name of the dataset an input parameter that could be parsed through the input file.
OK, it is fine. I was working on the auto detection for models. And faces several issues regarding classes, names generalization. Particularly, it changes how we create a model in main: instead of long if else if else sequential check it will be one call from the dictionary. I will use a cifar100_data
as a default for now. Then we can make it a parameter really quick.
Thank you!
On Sun, Jul 12, 2020 at 7:03 PM Vitaliy Starchenko notifications@github.com wrote:
OK, it is fine. I was working on the auto detection for models. And faces several issues regarding classes, names generalization. Particularly, it changes how we create a model in main: instead of long if else if else sequential check it will be one call from the dictionary. I will use a cifar100_data as a default for now. Then we can make it a parameter really quick.
— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/allaffa/CompetitiveGradientDescent/issues/75#issuecomment-657288432, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAS7SUCMF65NO6RVPDADGKDR3I6N3ANCNFSM4OYAEALA .
Is there any reason the data that is passed to different models in main is different? For example,
cifar100_data
orcifar10_data_dcgans
Can it be entered as a parameter? Or otherwise precoded? @allaffa @VittorioGabbi