Fail to building docker image

[jyeongho]

Like issue in #14 , I've got segmentation fault error. So, I tried to change docker environment from "nsml/ml:cuda10.1-cudnn7-pytorch1.3keras2.3" to "nvcr.io/nvidia/pytorch:20.03-py3".

However, when I run a session, I got this error message.

How can I change docker environment? Should I enter the docker image name into docker image cache sheet in "[https://docs.google.com/spreadsheets/d/1OHsW3HUhL6dSZ5KaZSICpW3MpEeh0Sykrqd883CtQ8Q/edit#gid=1078177844]"?)

[myoons]

@jyeongho How did you solve this problem?

[jyeongho]

@jyeongho How did you solve this problem?

I solved this problem by adding 'torch==1.7.0' and 'torchvision==0.8.1' in setup.py file.

[myoons]

@jyeongho Thanks!