关于“TensorFlow Dataset API”的评价

Ulaş F. · 已于 almost 5 years前审核

Johnny L. · 已于 almost 5 years前审核

S V Suman P. · 已于 almost 5 years前审核

Manu N. · 已于 almost 5 years前审核

Sandeep Y. · 已于 almost 5 years前审核

Farhan B. · 已于 almost 5 years前审核

Farhan B. · 已于 almost 5 years前审核

John C. · 已于 almost 5 years前审核

Aye H. · 已于 almost 5 years前审核

NIKHIL G. · 已于 almost 5 years前审核

Parv M. · 已于 almost 5 years前审核

Dmitry O. · 已于 almost 5 years前审核

Wai Ho A. · 已于 almost 5 years前审核

Maurice E. · 已于 almost 5 years前审核

Yaswant Sai Krishna P. · 已于 almost 5 years前审核

B J. · 已于 almost 5 years前审核

dario M. · 已于 almost 5 years前审核

Jan L. · 已于 almost 5 years前审核

Marc C. · 已于 almost 5 years前审核

Kiran B. · 已于 almost 5 years前审核

Li J. · 已于 almost 5 years前审核

Sandhiya V. · 已于 almost 5 years前审核

1) Not enough time. Provisioning the lab, creating an instance and downloading the GH repo takes more than five minutes. The instance could have been provisioned with the lab, and the repo could have been automatically cloned, as well. 2) Provide more information in the instructions. It seems the authors assume we know exactly what is to be done for each step. But if that were the case, we wouldn't be doing an introductory course. This lab was better than the other labs in terms of the instructions; but if we need to use a specific method at a particular step (i.e., using `pop` to remove values from a dictionary), we should be given a little more guidance. 3) If we're supposed to use specific values for some of the arguments (i.e., batch size, shuffle buffer, etc...), tell us. Don't just hope we'll stumble on the right answer. The underlying assumption appears to be that the students will need to look at the solution to finish the lab; but if we need to do this, it's only because these labs are not well designed.

Matthew C. · 已于 almost 5 years前审核

Ganesh P. · 已于 almost 5 years前审核

Frederick L. · 已于 almost 5 years前审核