Keyword arguments for liger functional

[Tcc0403]

🚀 The feature, motivation and pitch

Liger's functional doesn't support keyword arguments because it's implemented by replacing torchautograd.Function.apply.

kwargs support is necessary for LigerCrossEntropyLoss to patch transformers model after GA fix
https://github.com/huggingface/transformers/blob/a06a0d12636756352494b99b5b264ac9955bc735/src/transformers/models/llama/modeling_llama.py#L1224

Alternatives

No response

Additional context

No response

[ByronHsu]

CE is fixed by #375. But need to add functional wrapper for other modules

[hongpeng-guo]

I can take this one 😄