[fix] llama_dtype_fix triggered when flash attention is on

[KexinFeng]

What does this PR do?

Setup

model_id_or_path = "huggyllama/llama-7b"
model = AutoModelForCausalLM.from_pretrained(
    model_id_or_path, device_map='auto' if device.type == 'cuda' else 'cpu',
    use_flash_attention_2=flash_attn)

Fix

Inside attn_output = self.o_proj(attn_output), the weight dtype is torch.float32 which mismatches attn_output's dtype, torch.float16.

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 or the forum? 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, and
  here are tips on formatting docstrings.
• Did you write any new necessary tests?

Who can review?

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.

@ArthurZucker and @younesbelkada

[ArthurZucker]

Hey! Thanks for opening a PR, pretty sure #26846 would have taken care of it

[github-actions]

This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread.

Please note that issues that do not follow the contributing guidelines are likely to be ignored.

[younesbelkada]

Yes, #26846 should deal with this issue! Closing this PR - feel free to re-open if you think that's not the case