[fea] moe support

[bo-ke]

PR types

PR changes

Description

[paddle-bot]

Thanks for your contribution!

[CLAassistant]

Thank you for your submission! We really appreciate it. Like many open source projects, we ask that you sign our Contributor License Agreement before we can accept your contribution.

---

kebo01 seems not to be a GitHub user. You need a GitHub account to be able to sign the CLA. If you have already a GitHub account, please add the email address used for this commit to your account.
_{You have signed the CLA already but the status is still pending? Let us recheck it.}

[codecov]

Codecov Report

Attention: Patch coverage is 2.34375% with 125 lines in your changes missing coverage. Please review.

Project coverage is 53.97%. Comparing base (0cd8fe7) to head (8894d32).
Report is 257 commits behind head on develop.

Files with missing lines	Patch %	Lines
paddlenlp/trainer/trainer.py	2.32%	42 Missing ⚠️
paddlenlp/trainer/utils/helper.py	2.43%	40 Missing ⚠️
paddlenlp/trainer/training_args.py	5.00%	19 Missing ⚠️
paddlenlp/trainer/utils/sharding_io.py	0.00%	13 Missing ⚠️
paddlenlp/trainer/utils/reshard/common.py	0.00%	11 Missing ⚠️

Additional details and impacted files

@@             Coverage Diff             @@
##           develop    #8498      +/-   ##
===========================================
- Coverage    54.27%   53.97%   -0.31%     
===========================================
  Files          617      618       +1     
  Lines        96356    96801     +445     
===========================================
- Hits         52302    52249      -53     
- Misses       44054    44552     +498     

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

[ZHUI]

LGTM

[DesmonDay]

有些代码看看能不能是不是参考我这个PR的修改：#8511

[ZHUI]

LGTM