These are chat archives for arita37/tensorflow

21st
Jun 2016
Henrik Holst
@hholst80
Jun 21 2016 13:14
If you want to run TensorFlow cluster on a single machine (localhost) you can run into problems with CUDA_ERROR_OUT_OF_MEMORY is there some workarounds for that to limit the amout of GPU memory each process uses?
ericdanz
@ericdanz
Jun 21 2016 21:23
is there a straightforward way to slice tensors in c++?