[LLM] update tensor and pipeline parallel for chatglmv2

[DrownFish19]

PR types

Bug fixes

PR changes

Models

Description

fix tensor parallel for chatglmv2. This is a duplicate of #9133 and #9043.

1. output_layer中weight切分需要设置self.weight.is_distributed属性。
2. add pipeline parallel for chatglmv2

[codecov]

Codecov Report

Attention: Patch coverage is 33.33333% with 132 lines in your changes missing coverage. Please review.

Project coverage is 52.80%. Comparing base (753436a) to head (9e9e213).
Report is 10 commits behind head on develop.

Files with missing lines	Patch %	Lines
paddlenlp/transformers/chatglm_v2/modeling_pp.py	21.33%	118 Missing ⚠️
paddlenlp/transformers/chatglm_v2/modeling.py	71.42%	12 Missing ⚠️
...p/experimental/transformers/chatglm_v2/modeling.py	0.00%	2 Missing ⚠️

Additional details and impacted files

@@             Coverage Diff             @@
##           develop    #9204      +/-   ##
===========================================
- Coverage    52.97%   52.80%   -0.17%     
===========================================
  Files          703      705       +2     
  Lines       110981   110947      -34     
===========================================
- Hits         58788    58589     -199     
- Misses       52193    52358     +165     

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