[GitHub] cjolivier01 commented on issue #9300: bugfix for parallel rand generator on multi-gpu

cjolivier01 commented on issue #9300: bugfix for parallel rand generator on multi-gpu URL: https://github.com/apache/incubator-mxnet/pull/9300#issuecomment-355355959 I am not aware of an API call for that from python. One would need to be created, which would also force the CUDA library to

[GitHub] cjolivier01 commented on issue #9300: bugfix for parallel rand generator on multi-gpu

cjolivier01 commented on issue #9300: bugfix for parallel rand generator on multi-gpu URL: https://github.com/apache/incubator-mxnet/pull/9300#issuecomment-355314141 For example, run nvidia-smi -L and count number of lines returned which start with "GPU [0-9]+" or something like that

[GitHub] cjolivier01 commented on issue #9300: bugfix for parallel rand generator on multi-gpu

cjolivier01 commented on issue #9300: bugfix for parallel rand generator on multi-gpu URL: https://github.com/apache/incubator-mxnet/pull/9300#issuecomment-355313475 Current test machines in CI have at least two GPUs. However, you will need to detect GPU count in the unit test because when

[GitHub] cjolivier01 commented on issue #9300: bugfix for parallel rand generator on multi-gpu

cjolivier01 commented on issue #9300: bugfix for parallel rand generator on multi-gpu URL: https://github.com/apache/incubator-mxnet/pull/9300#issuecomment-355190854 Fix verified This is an automated message from the Apache G