Adding a flag whether to save checkpoint or not.

[yeonsily]

What does this PR do?

Fixes # (issue)

Before submitting

• This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
• Did you make sure to update the documentation with your changes?
• Did you write any new necessary tests?

[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.

[regisss]

Adding the training argument save_strategy="no" should lead to no checkpoint saved no?

[yeonsily]

Adding the training argument save_strategy="no" should lead to no checkpoint saved no?

I think they used for different cases.

1. save_strategy is for saving checkpoint per epoch or per checkpoint step.
   If we set this to 'no', it will still save the last checkpoint at the end of the training.

2. trainer.save_model() in the script is the one for saving last checkpoint at the end of the training.
   But if save_strategy is not 'no', it may still save checkpoint during the training per epoch or steps.

[regisss]

LGTM