Fix Inconsistency with IsShardedQLoRA Setting

[fabianlim]

What does this PR do?

In huggingface/transformers#33154, logic was introduced to fix low_cpu_mem_mode, in which the weights will be accordingly moved to either cpu or meta device depending on device rank. However, this causes a problem here as is_sharded_qlora could then be potentially set to True on device 0 and False on all others.

This PR fixes this problem which will be experienced in the upcoming transformers release.

Fixes transformers#33154

Before submitting

• This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
• Did you read the contributor guideline,
  Pull Request section?
• Was this discussed/approved via a GitHub issue? Please add a link
  to it if that's the case.
• Did you make sure to update the documentation with your changes? Here are the
  documentation guidelines.
• Did you write any new necessary tests?

Who can review?

@mdouglas
Anyone in the community is free to review the PR once the tests have passed. Feel free to tag
members/contributors who may be interested in your PR.

[HuggingFaceDocBuilderDev]

The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update.