flash attn args for GaudiGemmaForCausalLM

[kkoryun]

What does this PR do?

When running the Intel/llava-gemma-2b model using image-to-text example, it gives an error
GaudiGemmaForCausalLM.forward() got an unexpected keyword argument 'use_flash_attention'
This PR adds missing arguments.

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.