TE FP8 integration

[scsudhak-intel]

What does this PR do?

Refactoring transformer_engine FP8 support in optimum-habana. This removes model specific changes and makes the code more model agnostic.

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?

[vivekgoe]

@regisss @libinta please help review this PR. This is a cleaned up version of TE integration (tested with Llama-7B, Llama-70B, Mixtral-7B fine-tuning tasks). Changes were done mainly for following reasons,

• Align to newer PEFT version (0.10.0) which helps reducing custom changes to support LoRA layers in fp8,
• Add TE support for models using activation_checkpointing (in a generic way),
• Align TE code to accelerator repo to extent possible. Limit fp8 configuration interface exposed to user (allow only well tested configuration for now)

FYI @sanjucsudhakaran will create another PR adding Llama-7B test for this feature. It is tested with 1.16.0 release and can be merged without waiting for 1.17.0

[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.

[vivekgoe]

Related PR with tests #1106