【Inference】Migrate MoE Kernel from Paddle Inner

[gzy19990617]

Before submitting

• Lint code. If there are lint issues, please format the code first.

# Install and register `pre-commit` in the project folder
pip install pre-commit && pre-commit install

# Process previous code files separately
pre-commit run --file XXXX.py

• Add test cases into tests folder. If there are codecov issues, please add tests cases first.

PR types

PR changes

Description

• Paddle移除moe PR#71610

• 将框架中的moe算子迁移为paddle_nlp自定义算子，优化部分代码进行精减；有利于框架编译速度提升；便于应用cutlass3.x；便于在此基础支持其他精度moe kernel；便于精度profile等等。

• 部分头文件引用paddle内部头文件

• 单测精度可对齐；

• 模型deepseek-v2-lite wint4输出正常

• 修改tune后逻辑在长输入下性能约提升20-30%

• paddle内部实现与paddle_nlp自定义算子moe静态图服务化吞吐持平，可以证明迁移出去，不会影响服务静态图吞吐性能。

Qwen/Qwen1.5-MoE-A2.7B模型，单卡H卡/128并发/1000条数据/wint8/input-1152-output-201/block-bs 28
内部moe算子：

迁移后算子：

[paddle-bot]

Thanks for your contribution!

[codecov]

Codecov Report

Attention: Patch coverage is 0% with 3 lines in your changes missing coverage. Please review.

Project coverage is 49.93%. Comparing base (79dbdbb) to head (45b5e0e).
Report is 265 commits behind head on develop.

Files with missing lines	Patch %	Lines
...erimental/transformers/fused_transformer_layers.py	0.00%	3 Missing ⚠️

Additional details and impacted files

@@             Coverage Diff             @@
##           develop   #10063      +/-   ##
===========================================
- Coverage    49.97%   49.93%   -0.04%     
===========================================
  Files          757      757              
  Lines       122498   122586      +88     
===========================================
+ Hits         61217    61218       +1     
- Misses       61281    61368      +87     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[CLAassistant]

All committers have signed the CLA.

[yuanlehome]

LGTM