Inputting auxiliary states to MXNet symbol via minpy.core.Function

dmlc / minpy

NumPy interface with mixed backend execution

https://minpy.readthedocs.io/en/latest/

Other

1.11k stars 112 forks source link

Inputting auxiliary states to MXNet symbol via minpy.core.Function #169

Closed GaiYu0 closed 7 years ago

GaiYu0 commented 7 years ago

Hi! I am using core.Function to call BatchNorm symbol of MXNet. However, I cannot input moving_mean and moving_var to the symbol because it appears that core.Function only accepts trainable parameters. I hope core.Function could accept auxiliary states as well so that I could get acceleration from symbols such as BatchNorm.

Taco-W commented 7 years ago

No problem, I'll look into that

Taco-W commented 7 years ago

@GaiYu0 I guess this issue is fixed right?

GaiYu0 commented 7 years ago

I think the issue has been fixed.