Introduce TPU Pod launching to accelerate launch

[muellerzr]

Proposed API:

accelerate launch now allows for a configured pod setup through three new params/config items:

• use_cluster, whether to use a TPU cluster
• vm, this mimics xla_dist's vm argument and is a list of single compute VM names if you are not using an instance group. (generally not needed)
• env, this is a list of environment variables to set on each of the compute VM instances

Currently we only support non-Docker, as GCP doesn't support docker yet on the larger pods.

To launch a script on a TPU pod, the API will look like such:

Fully configured:

accelerate launch myscript.py --arg1 --arg2 ...

No configuration:

accelerate launch --tpu --use_cluster myscript.py --arg1 --arg2 ...

Note that if file permissions are given on your GCP cluster, there is a --tpu_use_sudo flag for this, and I recommend also doing sudo accelerate launch ... plus sudo pip install git+https://github.com/huggingface/accelerate. On my specific configuration it was needed, but other users reported they didn't so it's optional.

Closes #501 and closes #471

[HuggingFaceDocBuilderDev]

The documentation is not available anymore as the PR was closed or merged.

[sgugger]

Thanks for your work on this!