Chrislab

* Update BasicMetricTool:
  - Inherit from torch.nn.Module instead of torchmetrics.Metric
* Update DPTask, NERTask:
  - Add init argument: epoch_steps
  - Add internal param: metric_tools
  - Add _global_step(), _global_epoch(), _train_loss(), _valid_loss(), _train_accuracy(), _valid_accuracy(), _valid_metric(), _log_value()
  - Update training_step(), validation_step(): return dict
  - Add or update on_*(): accumulate & log value!
  - Update _global_step(): return float
  - Update on_train_batch_end(), on_validation_epoch_end(): log value of g_step, g_epoch
* Move TypedData, ProjectEnv to chrisbase.data

* Update TrainerArguments:
- Add staticmethod: from_args
* Add AppTyper

* Update CommonArguments:
- Add member: log()
* Update set_seed():
- using logger
* Move new_logger(), new_logger_file() from chrislab.nlpbook to chrisbase.io:
- Change name: new_logger() -> make_logger(), new_logger_file() -> make_dual_logger()
* Move ProjectEnv from chrisbase.io to chrislab.nlpbook.arguments
- Update ProjectEnv:
- Add members: job_name, command_args, output_home, msg_level, msg_logger, csv_logger
* Update nlpbook.set_seed():
- Use args.env.msg_logger instead of given logger
* Update nlpbook.arguments.JobTimer:
- Rename JobTimer to TimerChecker
* Update nlpbook.arguments.JobOutput:
- Rename JobOutput to ProgressChecker
- Remove members: dir_path, csv_out
* Update nlpbook.arguments.CommonArguments:
- Use self.env.output_home instead of self.output.dir_path
- Add setup_msg_logger() and call it at the end of post_init()
- Rename members:
- job -> time
- output -> prog
- log() -> info_arguments()
- save() -> save_arguments()
- setup_csv_out() -> setup_csv_logger()
* Update nlpbook.arguments.ArgumentsUsing:
- Revert making logger and setting seed
* Call args.setup_csv_logger() instead of nlpbook.setup_csv_out(args)
* Use args.env.csv_logger instead of args.output.csv_out
* Use args.env.output_home instead of args.output.dir_path
* Use self.args.env.msg_logger instead of logger, print()
* Remove setup_csv_out() from nlpbook.factory
* Update JobTimer:
- Change argument: file -> logger
- Rename argument: mute_logger -> mute_loggers
- Use self.logger instead of self.file
- Use self.logger.info() instead of print()
* Rename make_logger() -> make_unit_logger()
* Update make_dual_logger():
- Add argument: datefmt="[%m.%d %H:%M:%S]"
* Update ProjectEnv:
- Add member: msg_format
* Update CommonArguments.setup_msg_logger():
- Use self.env.msg_format
* Update CommonArguments.info_arguments():
- Print argument info in each line
* Use global logger
* Update ProjectEnv:
- Remove member: msg_logger
- Use configure_dual_logger()
* Update CommonArguments:
- Use configure_dual_logger()
* Update set_seed():
- Not use args.env.msg_logger
* Update ProjectEnv:
- Add member: set()
* Update ProjectEnv.__post_init__():
- Call configure_dual_logger() or configure_unit_logger() according to self.output_home
* Update CommonArguments.__post_init__():
- Make output directory by configure_*_logger() only
* Update setup_csv_logger():
- Rename to reconfigure_output()
- Read existing content and write it to new file
- Delete existing file
* Update info_arguments():
- Print a table prettily
* Update DataOption:
- show_examples -> num_show
* Update ModelOption:
- finetuning_home -> home
- finetuning_name -> name
- max_seq_length -> seq_len
* Update LearningOption:
- validating_fmt -> validate_fmt
- validating_on -> validate_on
- num_keeping -> keep_by
- keeping_by -> num_keep
- speed -> lr
* args.output -> args.prog
* Update NERCorpus.read_raw_example(), DPCorpus.read_raw_examples():
- Add annotation: staticmethod
* Not use args.env.off_debugging
* Adding DPTask
* Add LoggingFormat
* Complete DPTask
* Remove LoggingCallback
* Update BasicMetricTool:
- Inherit from torch.nn.Module instead of torchmetrics.Metric
* Update argument names
- num_show -> num_check
- keep_by -> save_by
- num_keep -> num_save
* Update ProjectEnv
- Add member: debugging

* Update NERFeatures:
- Add example_id, example
* Update features_to_batch():
- Add example_ids in batch
* Update NERTask:
- Add _global_step(), _trained_rate()
- Update validation_step()
- Update validation_epoch_end()
* Update nlpbook/corpus.py:
- Change NERExampleForKLUE -> NERRawExample
- Change NERFeatures -> NEREncodedExample
- Change features_to_batch() -> encoded_examples_to_batch()
- Change NERCorpus.get_examples() -> NERCorpus.read_raw_examples()
- Change _convert_examples_to_ner_features() -> _convert_to_encoded_examples()
* Remove num_cuda_devices()
* Update ArgumentsUsing:
- input params: logging_level, delete_on_exit
- return params: self.args_file, self.logger
* Update nlpbook.utils:
- Change logger functions:
- set_logger() -> new_logger(), new_logger_file()

* Update nlpbook/ner/corpus.py:
  - Add convert_kmou_format(), convert_klue_format()
  - Change NERExample -> NERExampleForKLUE
  - Update NERCorpus for NERExampleForKLUE
  - Update _convert_examples_to_ner_features() for NERExampleForKLUE
* Update common/util.py:
  - Add set_tokenizers_parallelism()
* Update nlpbook/ner/cli.py:
  - Use PreTrainedTokenizerFast

* Update chrislab/common/tokenizer_korbert.py
- Update model path
* Updating chrislab.nlpbook.cls:
- Syncing with chrislab.ner
* Update chrislab.nlpbook.ner:
- Update NERTask:
- Update training_step(), validation_step(), test_step()
- log more metrics
* Update nlpbook:
- Add DataFiles, CommandConfig, RuntimeCheckingOnArgs
- Update NLUArguments
- Update NLUServerArguments
- Update NLUTesterArguments
- Update NLUTrainerArguments
- Add LoggingCallback
- Update get_trainer(): train_logger, logging_callback 사용!
* Update chrislab.nlpbook.arguments:
- Argument들을 그룹화하였음
- Add JobTimer, JobResult
- Rename CommandConfig -> UsingArguments
- Rename RuntimeCheckingOnArgs -> CheckingRuntime
* Update chrislab.nlpbook.arguments:
- as_dataframe(): 체계적으로 정리!
* Update nlpbook.cls.cli:
- Add test()
* Update nlpbook.cls.task:
- Update ClassificationTask:
- Sync with NERTask
* Update nlpbook.factory:
- Add setup_csv_out()
- Add make_server()
- get_tester() -> make_tester()
- get_trainer() -> make_trainer()
* Update nlpbook.arguments:
- 상위 클래스 정리!
* Update nlpbook.cls, nlpbook.ner:
- 정리한 클래스 및 코드 사용!
* Update nlpbook.arguments:
- Update RuntimeChecking:
- 시작 시 미리 저장하지 않음!