[OP] Add InferShape&InferDtype for per_token_quant_padding

[DrRyanHuang]

Motivation

per_token_quant_padding ERNIE45T 300B 模型 FP8 转静过程中遇到自定义算子缺少 InferShape / InferDtype 的问题：

call paddle.api : static_op_per_token_quant_padding 
terminate called after throwing an instance of 'common::enforce::EnforceNotMet'
  what():  (Unavailable) Your custom operator contains multiple outputs. 

We only allow a custom operator that contains only one input and only one output without setting the InferShapeFn/InferDtypeFn. 
At this time, the input shape/dtype will be directly set to the output shape/dtype.

Please set the InferShapeFn/InferDtypeFn of custom operator by 
	.SetInferShapeFn(PD_INFER_SHAPE(...)) / .SetInferDtypeFn(PD_INFER_DTYPE(...))

  [Hint: Expected OpMetaInfoHelper::GetOutputs(custom_op_meta).size() == 1UL, 
but received OpMetaInfoHelper::GetOutputs(custom_op_meta).size():2 != 1UL:1.] 
(at /workspace/Paddle/paddle/fluid/framework/custom_operator_utils.h:219)

Modifications

给 per_token_quant_padding 添加 InferShape / InferDtype 函数

Usage or Command

NO NEED

Accuracy Tests

NO NEED

Checklist

• Add at least a tag in the PR title.
  • Tag list: [[FDConfig],[APIServer],[Engine], [Scheduler], [PD Disaggregation], [Executor], [Graph Optimization], [Speculative Decoding], [RL], [Models], [Quantization], [Loader], [OP], [KVCache], [DataProcessor], [BugFix], [Docs], [CI], [Optimization], [Feature], [Benchmark], [Others], [XPU], [HPU], [GCU], [DCU], [Iluvatar], [Metax]]
  • You can add new tags based on the PR content, but the semantics must be clear.
• Format your code, run pre-commit before commit.
• Add unit tests. Please write the reason in this PR if no unit tests.
• Provide accuracy results.
• If the current PR is submitting to the release branch, make sure the PR has been submitted to the develop branch, then cherry-pick it to the release branch with the [Cherry-Pick] PR tag.

[paddle-bot]

Thanks for your contribution!

[DrRyanHuang]

📢 建议分两个 commit review 该 PR 的时候：

• 9d3091a (#4667) 添加 InferShape / InferDtype 函数
• 4ad96db (#4667) 只进行了 pre-commit 做 code format

[SigureMo]

这个 format，能单独一个 PR 全 format 一下么 @DrRyanHuang @gongshaotian

[gongshaotian]

这个 format，能单独一个 PR 全 format 一下么 @DrRyanHuang @gongshaotian

有尝试全format，gpu算子会出现编译问题，PR就临时close了，后面分批提PR format吧