Open ssunno opened 7 years ago
Hi,
I'm using this model on my work.
TF supports windows after 0.12.0
when num_filters over 256, CUDA library is shutdown.
so i divided conv2d filter into 64. it works fine.
I think it will working same as before.
thanks !
Hi,
I'm using this model on my work.
TF supports windows after 0.12.0
when num_filters over 256, CUDA library is shutdown.
so i divided conv2d filter into 64. it works fine.
I think it will working same as before.
thanks !