doctorcorral/gyx

Implement interface for getting data and train a DNN from Python process

Opened this issue · 0 comments

The DNN handled by a Python process should be training almost all the time. Once a training phase finishes, it immediately update its data batch to train again.