Update serialization

[ZHUI]

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

BugFixes

PR changes

Others

Description

___________________________________________________________ ERROR collecting tests/models/test_internvl2.py ___________________________________________________________
ImportError while importing test module '/home/PaddleMIX/tests/models/test_internvl2.py'.
Hint: make sure your test modules/packages have valid Python names.
Traceback:
/usr/lib/python3.10/importlib/__init__.py:126: in import_module
    return _bootstrap._gcd_import(name[level:], package, level)
tests/models/test_internvl2.py:26: in <module>
    from paddlemix.models.internvl2.conversation import get_conv_template
paddlemix/__init__.py:16: in <module>
    from .datasets import *
paddlemix/datasets/__init__.py:19: in <module>
    from .chatml_dataset import *
paddlemix/datasets/chatml_dataset.py:18: in <module>
    from paddlenlp.transformers.tokenizer_utils import ChatTemplateMixin
PaddleNLP/paddlenlp/__init__.py:34: in <module>
    from paddlenlp.utils.log import logger
PaddleNLP/paddlenlp/utils/__init__.py:33: in <module>
    from .serialization import load_torch
PaddleNLP/paddlenlp/utils/serialization.py:28: in <module>
    from paddlenlp.transformers.utils import device_guard
PaddleNLP/paddlenlp/transformers/__init__.py:17: in <module>
    from .model_utils import PretrainedModel, register_base_model
PaddleNLP/paddlenlp/transformers/model_utils.py:77: in <module>
    from ..generation import GenerationConfig, GenerationMixin
PaddleNLP/paddlenlp/generation/__init__.py:33: in <module>
    from .streamers import BaseStreamer, TextIteratorStreamer, TextStreamer
PaddleNLP/paddlenlp/generation/streamers.py:18: in <module>
    from paddlenlp.transformers.tokenizer_utils import PretrainedTokenizer
PaddleNLP/paddlenlp/transformers/tokenizer_utils.py:47: in <module>
    from ..data.vocab import Vocab
PaddleNLP/paddlenlp/data/__init__.py:18: in <module>
    from .data_collator import *
PaddleNLP/paddlenlp/data/data_collator.py:26: in <module>
    from ..transformers import BertTokenizer
E   ImportError: cannot import name 'BertTokenizer' from partially initialized module 'paddlenlp.transformers' (most likely due to a circular import) (/home/PaddleMIX/PaddleNLP/paddlenlp/transformers/__init__.py)

[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.
_{You have signed the CLA already but the status is still pending? Let us recheck it.}

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 46.98%. Comparing base (8205e3d) to head (e790b1a).
Report is 6 commits behind head on develop.

❗ Current head e790b1a differs from pull request most recent head 0680c0c

Please upload reports for the commit 0680c0c to get more accurate results.

❌ Your project status has failed because the head coverage (46.98%) is below the target coverage (58.00%). You can increase the head coverage or adjust the target coverage.

Additional details and impacted files

@@           Coverage Diff            @@
##           develop   #10584   +/-   ##
========================================
  Coverage    46.98%   46.98%           
========================================
  Files          799      799           
  Lines       132246   132246           
========================================
+ Hits         62135    62136    +1     
+ Misses       70111    70110    -1     

☔ 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.