Features:
1. Refactor the directory structure of nlp models. All model files are placed into either the model folder or the task_model folder
2. Refactor all the comments to google style
3. Add detail comments to important tasks and nlp models, to list the description of the model, and its preprocessor&trainer
4. Model Exporting now supports a direct all to TorchModelExporter(no need to derive from it)
5. Refactor model save_pretrained method to support direct running(independent from trainer)
6. Remove the judgement of Model in the pipeline base class, to support outer register models running in our pipelines
7. Nlp trainer now has a NLPTrainingArguments class , user can pass arguments into the dataclass, and use it as a normal cfg_modify_fn, to simplify the operation of modify cfg.
8. Merge the BACKBONES and the MODELS, so user can get a backbone with the Model.from_pretrained call
9. Model.from_pretrained now support a task argument, so user can use a backbone and load it with a specific task class.
10. Support Preprocessor.from_pretrained method
11. Add standard return classes to important nlp tasks, so some of the pipelines and the models are independent now, the return values of the models will always be tensors, and the pipelines will take care of the conversion to numpy and the following stuffs.
12. Split the file of the nlp preprocessors, to make the dir structure more clear.

Bugs Fixing:
1. Fix a bug that lr_scheduler can be called earlier than the optimizer's step
2. Fix a bug that the direct call of Pipelines (not from pipeline(xxx)) throws error
3. Fix a bug that the trainer will not call the correct TaskDataset class
4. Fix a bug that the internal loading of dataset will throws error in the trainer class
        Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/10490585

Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/10491891

1.新增支持原始bert模型（非easynlp的 backbone prefix版本）
2.支持bert的在sequence classification/fill mask /token classification上的backbone head形式
3.统一了sequence classification几个任务的pipeline到一个类
4.fill mask 支持backbone head形式
5.token classification的几个子任务（ner，word seg， part of speech）的preprocessor 统一到了一起TokenClassificationPreprocessor
6. sequence classification的几个子任务（single classification， pair classification）的preprocessor 统一到了一起SequenceClassificationPreprocessor
7. 改动register中 cls的group_key 赋值位置，之前的group_key在多个decorators的情况下，会被覆盖，obj_cls的group_key信息不正确
8. 基于backbone head形式将 原本group_key和 module同名的情况尝试做调整，如下在modelscope/pipelines/nlp/sequence_classification_pipeline.py 中 
原本
 @PIPELINES.register_module(
    Tasks.sentiment_classification, module_name=Pipelines.sentiment_classification)
改成
@PIPELINES.register_module(
    Tasks.text_classification, module_name=Pipelines.sentiment_classification)
相应的configuration.json也有改动，这样的改动更符合任务和pipline（子任务）的关系。
8. 其他相应改动为支持上述功能
        Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/10041463

unittest for demo service
        Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/10006180

Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/9662182

    * clean up test level

* refine taskdataset interface
 * add device placement for trainer
 * add device placement for pipeline
 * add config checker and fix model placement bug
 * fix cycling import
 * refactor model init for translation_pipeline
 * cv pipelines support kwargs


Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/9463076

* fix msdataset
        Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/9436292

    * fix msdataset

Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/9174075

* fix msdataset

Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/9165402

This reverts commit c8e2e6de0e.

Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/9155405

Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/9143809

* using get_model to validate hub path 
* support reading pipeline info from configuration file
* add metainfo const
* update model type and pipeline type and fix UT
* relax requimrent for protobuf
* skip two dataset tests due to temporal failure
 
Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/9118154

合并model hub 代码
        Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/9096493

* pydataset fetch data from datahub
        Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/9060856

* add test level support
* update develop doc
Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/9021354

Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/9005038

1.  refactor maaslib to modelscope
2.  fix UT error
3.  support pipeline which does not register default model

Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/8988388

1. add default model support
2. fix circular import
3. temporarily skip ofa and palm test which costs too much time

Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/8981076

Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/8974892

Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/8972440

    * [to #42339559] support multiple models

* merge pydataset to the repo
        Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/8955999

Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/8963687

Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/8945177

    * support plain pipeline for bert

Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/8898823

Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/8899458

    * fix: UT error

Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/8868644

* add trainer interface
 * add trainer script
 * add model init support for pipelineadd pipeline tutorial and fix bugs 
 * add text classification evaluation to maas lib 
 * add quickstart and prepare env doc
 * relax requirements for torch and sentencepiece
 * merge release/0.1 and fix conflict
 * modelhub support for model and pipeline

 Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/8868339

1. add pipleine tutorial
2. fix bugs when using pipeline with certain model and preprocessor

Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/8810524

* add preprocessor module
 * add model base and builder
 * update task constant
 * add load image preprocessor and its dependency
 * add pipeline interface and UT covered
 * support default pipeline for task
 * add image matting pipeline
 * refine nlp tokenize interface
 * add nlp pipeline 
 * fix UT failed
 * add test for Compose

Link: https://code.aone.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/8769235

* add preprocessor module

* add test for Compose

* fix citest error

* fix abs class error

* add model base and builder

* update task constant

* add load image preprocessor and its dependency

* add pipeline interface and UT covered

* support default pipeline for task

* refine models and pipeline interface

* add pipeline folder structure

* add image matting pipeline

* refine nlp tokenize interface

* add nlp pipeline 

1.add preprossor model pipeline for nlp text classification
2. add corresponding test

Link: https://code.aone.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/8757371

* new nlp pipeline

* format pre-commit code

* update easynlp pipeline

* update model_name for easynlp pipeline; add test for maas_lib/utils/typeassert.py

* update test_typeassert.py

* refactor code

1. rename typeassert to type_assert
2. use lazy import to make easynlp dependency optional
3. refine image matting UT

* fix linter test failed

* update requirements.txt

* fix UT failed

* fix citest script to update requirements