train_reader.py

##################
# Import modules #
##################

import os
import argparse
from typing import NoReturn

from datasets import (
    load_metric,
    load_from_disk,
    concatenate_datasets,
    DatasetDict,
)

from transformers import AutoConfig, AutoModelForQuestionAnswering, AutoTokenizer

from transformers import (
    DataCollatorWithPadding,
    EvalPrediction,
    set_seed,
)

from model.Reader.RobertaCnn import RobertaCNNForQuestionAnswering
from model.Reader.trainer_qa import QuestionAnsweringTrainer

from utils.utils_qa import postprocess_qa_predictions, check_no_error
from utils.arguments import (

    ModelArguments,
    DataArguments,
    TrainingArguments,
    CustomArguments,
    train_config_setting,
    wandb_config_setting,
    
    MODEL_DIR,
    CONFIG_DIR,
    LOG_DIR,
)

from utils.logger import get_logger
import wandb

########################
# Set global variables #
########################

logger = None
CUSTOM_MODEL_NAMES = {
    "RobertaCnn":RobertaCNNForQuestionAnswering,
}

#######################
# Classes & Functions #
#######################

def main():

    global logger
    
    # Load config json
    parser = argparse.ArgumentParser()
    parser.add_argument("-c", "--config_file_path", help="Configure Json path")
    parser.add_argument("-l", "--log_file_path", default="reader_train.log", help="Logger file path")
    parser.add_argument("-n", "--model_name", default=None, help="Logger file path")
    parser.add_argument("--do_train",action="store_true")
    parser.add_argument("--do_eval",action="store_true")

    config = parser.parse_args()

    assert config.model_name, "Output 파일 이름을 설정해 주세요"
    
    config.config_file_path = os.path.join(CONFIG_DIR, config.config_file_path) 
    config.log_file_path = os.path.join(LOG_DIR, config.log_file_path)
    config.model_name = os.path.join(MODEL_DIR, config.model_name)
    
    logger = get_logger(config.log_file_path)
    
    model_args, data_args, training_args, custom_args =\
        train_config_setting(config)

    training_args.do_train = config.do_train
    training_args.do_eval = config.do_eval
    
    
    logger.info(f"model is from {model_args.model_name_or_path}")
    logger.info(f"data is from {data_args.dataset_name}")
    logger.info("Training/evaluation parameters %s", training_args)

    # Set random seed
    set_seed(training_args.seed)

    # Load datasets
    datasets = load_from_disk(data_args.dataset_name)

    # Load Config & tokenizer
    config = AutoConfig.from_pretrained(
            model_args.config_name
            if model_args.config_name is not None
            else model_args.model_name_or_path,
        )

    tokenizer = AutoTokenizer.from_pretrained(
        model_args.tokenizer_name
        if model_args.tokenizer_name is not None
        else model_args.model_name_or_path,
        use_fast=True,
    )

    # Load Model
    if model_args.model_name_or_path in CUSTOM_MODEL_NAMES:
        model = CUSTOM_MODEL_NAMES[model_args.model_name_or_path].from_pretrained(
            model_args.model_name_or_path,
            config=config,
        )
    else:
        model = AutoModelForQuestionAnswering.from_pretrained(
            model_args.model_name_or_path,
            from_tf=bool(".ckpt" in model_args.model_name_or_path),
            config=config,
        )

    logger.info(
        type(training_args),
        type(model_args),
        type(datasets),
        type(custom_args),
        type(tokenizer),
        type(model),
    )

    # do_train mrc model 혹은 do_eval mrc model
    if training_args.do_train or training_args.do_eval:
        run_mrc(
            data_args,
            training_args,
            model_args,
            custom_args,
            datasets,
            tokenizer,
            model,
        )

def run_mrc(
    data_args: DataArguments,
    training_args: TrainingArguments,
    model_args: ModelArguments,
    custom_args: CustomArguments,  # add
    datasets: DatasetDict,
    tokenizer,
    model,
) -> NoReturn:

    # Wandb 설정
    if custom_args.use_wandb:
        config = wandb_config_setting(data_args, training_args, model_args, custom_args)
        wandb.init(
            project=custom_args.project_name,
            entity=custom_args.entity_name,
            name=custom_args.wandb_run_name,
            config=config,
        )

    # dataset을 전처리합니다.
    # training과 evaluation에서 사용되는 전처리는 아주 조금 다른 형태를 가집니다.
    if training_args.do_train:
        column_names = datasets["train"].column_names
    else:
        column_names = datasets["validation"].column_names

    question_column_name = "question" if "question" in column_names else column_names[0]
    context_column_name = "context" if "context" in column_names else column_names[1]
    answer_column_name = "answers" if "answers" in column_names else column_names[2]

    # Padding에 대한 옵션을 설정합니다.
    # (question|context) 혹은 (context|question)로 세팅 가능합니다.
    pad_on_right = tokenizer.padding_side == "right"

    # 오류가 있는지 확인합니다.
    last_checkpoint, max_seq_length = check_no_error(
        data_args, training_args, datasets, tokenizer
    )

    # Train preprocessing / 전처리를 진행합니다.
    def prepare_train_features(examples):
        # truncation과 padding(length가 짧을때만)을 통해 toknization을 진행하며, stride를 이용하여 overflow를 유지합니다.
        # 각 example들은 이전의 context와 조금씩 겹치게됩니다.
        tokenized_examples = tokenizer(
            examples[question_column_name if pad_on_right else context_column_name],
            examples[context_column_name if pad_on_right else question_column_name],
            truncation="only_second" if pad_on_right else "only_first",
            max_length=max_seq_length,
            stride=data_args.doc_stride,
            return_overflowing_tokens=True,
            return_offsets_mapping=True,
            return_token_type_ids=False, # roberta모델을 사용할 경우 False, bert를 사용할 경우 True로 표기해야합니다.
            padding="max_length" if data_args.pad_to_max_length else False,
        )

        # 길이가 긴 context가 등장할 경우 truncate를 진행해야하므로, 해당 데이터셋을 찾을 수 있도록 mapping 가능한 값이 필요합니다.
        sample_mapping = tokenized_examples.pop("overflow_to_sample_mapping")
        # token의 캐릭터 단위 position를 찾을 수 있도록 offset mapping을 사용합니다.
        # start_positions과 end_positions을 찾는데 도움을 줄 수 있습니다.
        offset_mapping = tokenized_examples.pop("offset_mapping")

        # 데이터셋에 "start position", "enc position" label을 부여합니다.
        tokenized_examples["start_positions"] = []
        tokenized_examples["end_positions"] = []

        for i, offsets in enumerate(offset_mapping):
            input_ids = tokenized_examples["input_ids"][i]
            cls_index = input_ids.index(tokenizer.cls_token_id)  # cls index

            # sequence id를 설정합니다 (to know what is the context and what is the question).
            sequence_ids = tokenized_examples.sequence_ids(i)

            # 하나의 example이 여러개의 span을 가질 수 있습니다.
            sample_index = sample_mapping[i]
            answers = examples[answer_column_name][sample_index]

            # answer가 없을 경우 cls_index를 answer로 설정합니다(== example에서 정답이 없는 경우 존재할 수 있음).
            if len(answers["answer_start"]) == 0:
                tokenized_examples["start_positions"].append(cls_index)
                tokenized_examples["end_positions"].append(cls_index)
            else:
                # text에서 정답의 Start/end character index
                start_char = answers["answer_start"][0]
                end_char = start_char + len(answers["text"][0])

                # text에서 current span의 Start token index
                token_start_index = 0
                while sequence_ids[token_start_index] != (1 if pad_on_right else 0):
                    token_start_index += 1

                # text에서 current span의 End token index
                token_end_index = len(input_ids) - 1
                while sequence_ids[token_end_index] != (1 if pad_on_right else 0):
                    token_end_index -= 1

                # 정답이 span을 벗어났는지 확인합니다(정답이 없는 경우 CLS index로 label되어있음).
                if not (
                    offsets[token_start_index][0] <= start_char
                    and offsets[token_end_index][1] >= end_char
                ):
                    tokenized_examples["start_positions"].append(cls_index)
                    tokenized_examples["end_positions"].append(cls_index)
                else:
                    # token_start_index 및 token_end_index를 answer의 끝으로 이동합니다.
                    # Note: answer가 마지막 단어인 경우 last offset을 따라갈 수 있습니다(edge case).
                    while (
                        token_start_index < len(offsets)
                        and offsets[token_start_index][0] <= start_char
                    ):
                        token_start_index += 1
                    tokenized_examples["start_positions"].append(token_start_index - 1)
                    while offsets[token_end_index][1] >= end_char:
                        token_end_index -= 1
                    tokenized_examples["end_positions"].append(token_end_index + 1)

        return tokenized_examples

    if training_args.do_train:
        if "train" not in datasets:
            raise ValueError("--do_train requires a train dataset")

        if data_args.use_validation_data:
            train_dataset = concatenate_datasets(
                [datasets["train"], datasets["validation"]]
            )
        else:
            train_dataset = datasets["train"]

        

        # dataset에서 train feature를 생성합니다.
        train_dataset = train_dataset.map(
            prepare_train_features,
            batched=True,
            num_proc=data_args.preprocessing_num_workers,
            remove_columns=column_names,
            load_from_cache_file=not data_args.overwrite_cache,
        )

        

    # Validation preprocessing
    def prepare_validation_features(examples):
        # truncation과 padding(length가 짧을때만)을 통해 toknization을 진행하며, stride를 이용하여 overflow를 유지합니다.
        # 각 example들은 이전의 context와 조금씩 겹치게됩니다.
        tokenized_examples = tokenizer(
            examples[question_column_name if pad_on_right else context_column_name],
            examples[context_column_name if pad_on_right else question_column_name],
            truncation="only_second" if pad_on_right else "only_first",
            max_length=max_seq_length,
            stride=data_args.doc_stride,
            return_overflowing_tokens=True,
            return_offsets_mapping=True,
            return_token_type_ids=False, # roberta모델을 사용할 경우 False, bert를 사용할 경우 True로 표기해야합니다.
            padding="max_length" if data_args.pad_to_max_length else False,
        )

        # 길이가 긴 context가 등장할 경우 truncate를 진행해야하므로, 해당 데이터셋을 찾을 수 있도록 mapping 가능한 값이 필요합니다.
        sample_mapping = tokenized_examples.pop("overflow_to_sample_mapping")

        # evaluation을 위해, prediction을 context의 substring으로 변환해야합니다.
        # corresponding example_id를 유지하고 offset mappings을 저장해야합니다.
        tokenized_examples["example_id"] = []

        for i in range(len(tokenized_examples["input_ids"])):
            # sequence id를 설정합니다 (to know what is the context and what is the question).
            sequence_ids = tokenized_examples.sequence_ids(i)
            context_index = 1 if pad_on_right else 0

            # 하나의 example이 여러개의 span을 가질 수 있습니다.
            sample_index = sample_mapping[i]
            tokenized_examples["example_id"].append(examples["id"][sample_index])

            # Set to None the offset_mapping을 None으로 설정해서 token position이 context의 일부인지 쉽게 판별 할 수 있습니다.
            tokenized_examples["offset_mapping"][i] = [
                (o if sequence_ids[k] == context_index else None)
                for k, o in enumerate(tokenized_examples["offset_mapping"][i])
            ]
        return tokenized_examples

    if training_args.do_eval:
        eval_dataset = datasets["validation"]

        # Validation Feature 생성
        eval_dataset = eval_dataset.map(
            prepare_validation_features,
            batched=True,
            num_proc=data_args.preprocessing_num_workers,
            remove_columns=column_names,
            load_from_cache_file=not data_args.overwrite_cache,
        )

    # Data collator
    # flag가 True이면 이미 max length로 padding된 상태입니다.
    # 그렇지 않다면 data collator에서 padding을 진행해야합니다.
    data_collator = DataCollatorWithPadding(
        tokenizer, pad_to_multiple_of=8 if training_args.fp16 else None
    )

    # Post-processing:
    def post_processing_function(examples, features, predictions, training_args):
        # Post-processing: start logits과 end logits을 original context의 정답과 match시킵니다.
        predictions = postprocess_qa_predictions(
            examples=examples,
            features=features,
            predictions=predictions,
            max_answer_length=data_args.max_answer_length,
            output_dir=training_args.output_dir,
        )
        # Metric을 구할 수 있도록 Format을 맞춰줍니다.
        formatted_predictions = [
            {"id": k, "prediction_text": v} for k, v in predictions.items()
        ]
        if training_args.do_predict:
            return formatted_predictions

        elif training_args.do_eval:
            references = [
                {"id": ex["id"], "answers": ex[answer_column_name]}
                for ex in datasets["validation"]
            ]
            return EvalPrediction(
                predictions=formatted_predictions, label_ids=references
            )

    metric = load_metric("squad")

    def compute_metrics(p: EvalPrediction):
        return metric.compute(predictions=p.predictions, references=p.label_ids)

    # Trainer 초기화
    if custom_args.use_wandb:
        wandb.watch(model)
    trainer = QuestionAnsweringTrainer(
        model=model,
        custom_args=custom_args,
        model_tokenizer=tokenizer,
        args=training_args,
        train_dataset=train_dataset if training_args.do_train else None,
        eval_dataset=eval_dataset if training_args.do_eval else None,
        eval_examples=datasets["validation"] if training_args.do_eval else None,
        tokenizer=tokenizer,
        data_collator=data_collator,
        post_process_function=post_processing_function,
        compute_metrics=compute_metrics,
    )

    # Training
    if training_args.do_train:
        if last_checkpoint is not None:
            checkpoint = last_checkpoint
        elif os.path.isdir(model_args.model_name_or_path):
            checkpoint = model_args.model_name_or_path
        else:
            checkpoint = None
        train_result = trainer.train(resume_from_checkpoint=checkpoint)
        # training_step(model, train_dataset)
        trainer.save_model()  # Saves the tokenizer too for easy upload

        metrics = train_result.metrics
        metrics["train_samples"] = len(train_dataset)

        trainer.log_metrics("train", metrics)
        trainer.save_metrics("train", metrics)
        trainer.save_state()

        output_train_file = os.path.join(training_args.output_dir, "train_results.txt")

        with open(output_train_file, "w") as writer:
            logger.info("***** Train results *****")
            for key, value in sorted(train_result.metrics.items()):
                logger.info(f"  {key} = {value}")
                writer.write(f"{key} = {value}\n")

        # State 저장
        trainer.state.save_to_json(
            os.path.join(training_args.output_dir, "trainer_state.json")
        )

    # Evaluation
    if training_args.do_eval:
        logger.info("*** Evaluate ***")
        metrics = trainer.evaluate()
        metrics["eval_samples"] = len(eval_dataset)

        trainer.log_metrics("eval", metrics)
        trainer.save_metrics("eval", metrics)


if __name__ == "__main__":
    main()