error You need C++17 to compile PyTorch

[ranggihwang]

Branch/Tag/Commit

main

Docker Image Version

nvcr.io/nvidia/tensorflow:22.09-tf1-py3

GPU name

A100

CUDA Driver

510.47.03

Reproduced Steps

There are two ways to reproduce the error.
Both case shows the same error message.

## FIRST WAY ##
docker run -ti --shm-size 5g --rm nvcr.io/nvidia/tensorflow:22.09-tf1-py3 bash
git clone https://github.com/NVIDIA/FasterTransformer.git
mkdir -p FasterTransformer/build
cd FasterTransformer/build
git submodule init && git submodule update
pip install torch
cmake -DSM=80 -DCMAKE_BUILD_TYPE=Release -DBUILD_PYT=ON -DBUILD_MULTI_GPU=ON ..
make -j12

## SECOND WAY ##
docker run -it -d --name=test --privileged --gpus all --cap-add=ALL --shm-size=1g --ulimit memlock=-1 --ulimit stack=67108864 --net=host --ipc=host -v ${HOME}:${HOME} -v /dev:/dev -v /lib/modules:/lib/modules  -v /raid:/raid nvcr.io/nvidia/tensorflow:22.09-tf1-py3 bash
git clone https://github.com/NVIDIA/FasterTransformer.git
mkdir -p FasterTransformer/build
cd FasterTransformer/build
git submodule init && git submodule update
cmake -DSM=80 -DCMAKE_BUILD_TYPE=Release -DBUILD_PYT=ON -DBUILD_MULTI_GPU=ON ..
make -j12

Then you can get the below error.

error C++17 or later compatible compiler is required to use PyTorch.