[Bug Fix]Update to Pytorch 1.12

[VibhuJawa]

Description

On the last docker image , we were seeing some build failures. This was because with cudatoolkit=11.7 conda pulled in CPU torch instead of GPU torch and this caused issues. I update the docker image and also verified that the image works by building and testing like below:

### Build Command

docker build -f test_cugraph_nightly_file.Dockerfile -t dgl_nightly .


### Test Command
docker run --gpus=all -it dgl_nightly bash -c 'cd /dgl && bash tests/scripts/cugraph_unit_test.sh cugraph

FROM rapidsai/cugraph_nightly_torch-cuda:11.5-base-ubuntu18.04-py3.9-pytorch1.12.0-rapids22.10
RUN cd / && git clone https://github.com/dmlc/dgl.git 
RUN cd / && cd dgl && git submodule update --init --recursive 
RUN cd / && cd dgl && git config --global user.email "vjawa@nvidia.com" && git fetch origin pull/4573/head:PR_4573 && git merge PR_4573
RUN cd /dgl && bash tests/scripts/build_dgl.sh cugraph

tests/cugraph/test_basics.py::test_dummy PASSED                                                                                                                                                                                        [ 33%]
tests/cugraph/test_basics.py::test_to_cugraph_conversion 


PASSED                                                                                                                                                                        [ 66%]
tests/cugraph/test_basics.py::test_from_cugraph_conversion PASSED                                                                                                                                                                      [100%]

Checklist

Please feel free to remove inapplicable items for your PR.

• The PR title starts with [$CATEGORY] (such as [NN], [Model], [Doc], [Feature]])
• Changes are complete (i.e. I finished coding on this PR)
• All changes have test coverage
• Code is well-documented
• To the best of my knowledge, examples are either not affected by this change,
  or have been fixed to be compatible with this change
• Related issue is referred in this PR
• If the PR is for a new model/paper, I've updated the example index here.

Changes