Extend moe_permute & moe_unpermute with extra inference-aware features

[A-nnonymous]

PR Category

Operator Mechanism

PR Types

New features

Description

Extend moe_permute & moe_unpermute with extra inference facing features

• 为moe_permute加入一个可选的返回值expert_indices, 通过bool return_expert_indices控制（默认false），shape为【token_buffer_size】，每一个元素代表buffer中每一行所属的专家
• 为moe_permute加入开关using_tp_alloc（默认false），影响算子对buffer size的分配逻辑，此处需求比较模糊，建议严格验证
• 为moe_unpermute加入“加权combine”的feature，通过bool using_weighted_combine开启，默认为false

pcard-91067

[paddle-bot]

你的PR提交成功，感谢你对开源项目的贡献!
请关注后续CI自动化测试结果，详情请参考Paddle-CI手册。
Your PR has been submitted. Thanks for your contribution!
Please wait for the result of CI firstly. See Paddle CI Manual for details.

[codecov-commenter]

Codecov Report

❌ Patch coverage is 0% with 8 lines in your changes missing coverage. Please review.
⚠️ Please upload report for BASE (develop@5f9a66f). Learn more about missing BASE report.

Files with missing lines	Patch %	Lines
paddle/phi/infermeta/multiary.cc	0.00%	5 Missing ⚠️
python/paddle/nn/functional/moe_permute.py	0.00%	3 Missing ⚠️

❌ Your patch status has failed because the patch coverage (0.00%) is below the target coverage (90.00%). You can increase the patch coverage or adjust the target coverage.

Additional details and impacted files

@@            Coverage Diff             @@
##             develop   #77547   +/-   ##
==========================================
  Coverage           ?    0.00%           
==========================================
  Files              ?        2           
  Lines              ?        8           
  Branches           ?        0           
==========================================
  Hits               ?        0           
  Misses             ?        8           
  Partials           ?        0           

☔ 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.

[A-nnonymous]

/re-run all-failed

[A-nnonymous]

/re-run all-failed