[AutoParallel] Support deepseekv2 with DP/MP

[xuxinyi389]

PR types

New features

PR changes

Models

Description

1. deepseekv2/3 单卡模型，
2. 中层API配置DP(Sharding)/MP跑通

[paddle-bot]

Thanks for your contribution!

[codecov]

Codecov Report

Attention: Patch coverage is 22.38494% with 371 lines in your changes missing coverage. Please review.

Project coverage is 51.61%. Comparing base (58fc49f) to head (495d123).
Report is 347 commits behind head on develop.

Files with missing lines	Patch %	Lines
...addlenlp/transformers/deepseek_v2/modeling_auto.py	17.40%	337 Missing ⚠️
...addlenlp/transformers/deepseek_v3/modeling_auto.py	50.00%	34 Missing ⚠️

Additional details and impacted files

@@             Coverage Diff             @@
##           develop    #9862      +/-   ##
===========================================
- Coverage    52.14%   51.61%   -0.53%     
===========================================
  Files          734      740       +6     
  Lines       116261   117708    +1447     
===========================================
+ Hits         60622    60753     +131     
- Misses       55639    56955    +1316     

☔ 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.

[DrownFish19]

LGTM

[DrownFish19]

LGTM