mllama outputs refactor (#39643)
* mllama outputs refactor
* forgot kwargs
* fix output
* add can_record_outputs
* correct @check_model_inputs placement
* ruff and copies
* rebase
* feedback
* only return hidden_states
---------
Co-authored-by: ita.zaporozhets@huggingface.co <ita_zaporozhets@ip-26-0-161-153.ec2.internal>
Co-authored-by: ita.zaporozhets@huggingface.co <ita_zaporozhets@ip-26-0-162-14.ec2.internal>
