Autogluon: running stalled at starting

Created on 20 Jan 2020  路  5Comments  路  Source: awslabs/autogluon

when I run the official object detection example, it was stalled at staring.
I ran the code in docker container. I tried two trials: one with cuda 10.1 and python3.6; one with cuda 10.0 and python3.7
image

API & Doc

Most helpful comment

when I ran a container I added the args / --ipc="host" / and all the trials normaly ran.

All 5 comments

For cuda 10.0, could you try the mxnet version mxnet-cu100==1.6.0b20191013

For cuda 10.0, could you try the mxnet version mxnet-cu100==1.6.0b20191013

maybe it works with your sugest, beacase it throw timeout exception and give some reasons, I will do a trial on other server. Look forward to good result Thank you!

when I ran a container I added the args / --ipc="host" / and all the trials normaly ran.

For cuda 10.0, could you try the mxnet version mxnet-cu100==1.6.0b20191013

This worked for me

@zhanghang1989 Is there any action remaining for this issue or can it be resolved?

Was this page helpful?
0 / 5 - 0 ratings