"""Experiment-running framework."""
import argparse
import importlib

import numpy as np
import torch
import pytorch_lightning as pl
import lit_models
from lit_models import TransformerLitModel
from transformers import AutoConfig
import os
import json
os.environ["TOKENIZERS_PARALLELISM"] = "false"


# In order to ensure reproducible experiments, we must set random seeds.


def _import_class(module_and_class_name: str) -> type:
    """Import class from a module, e.g. 'text_recognizer.models.MLP'"""
    module_name, class_name = module_and_class_name.rsplit(".", 1)
    module = importlib.import_module(module_name)
    class_ = getattr(module, class_name)
    return class_


def _setup_parser():
    """Set up Python's ArgumentParser with data, model, trainer, and other arguments."""
    parser = argparse.ArgumentParser(add_help=False)

    # Add Trainer specific arguments, such as --max_epochs, --gpus, --precision
    trainer_parser = pl.Trainer.add_argparse_args(parser)
    trainer_parser._action_groups[1].title = "Trainer Args"  # pylint: disable=protected-access
    parser = argparse.ArgumentParser(add_help=False, parents=[trainer_parser])

    # Basic arguments
    parser.add_argument("--wandb", action="store_true", default=False)
    parser.add_argument("--seed", type=int, default=666)
    parser.add_argument("--data_class", type=str, default="TEXTCLS")
    parser.add_argument("--model_class", type=str, default="bert.BertForSequenceClassification")
    parser.add_argument("--load_checkpoint", type=str, default=None)

    # Get the data and model classes, so that we can add their specific arguments
    temp_args, _ = parser.parse_known_args()
    data_class = _import_class(f"data.{temp_args.data_class}")
    model_class = _import_class(f"models.{temp_args.model_class}")

    # Get data, model, and LitModel specific arguments
    data_group = parser.add_argument_group("Data Args")
    data_class.add_to_argparse(data_group)

    model_group = parser.add_argument_group("Model Args")
    model_class.add_to_argparse(model_group)

    lit_model_group = parser.add_argument_group("LitModel Args")
    lit_models.BaseLitModel.add_to_argparse(lit_model_group)

    parser.add_argument("--help", "-h", action="help")
    return parser



def main():
    """
    Run an experiment.

    Sample command:
    ```
    python training/run_experiment.py --max_epochs=3 --gpus='0,' --num_workers=20 --model_class=MLP --data_class=MNIST
    ```
    """
    parser = _setup_parser()
    args = parser.parse_args()


    with open("config.json", "w") as file:
        t = vars(args).copy()
        t.pop("tpu_cores")
        json.dump(t, file)


    np.random.seed(args.seed)
    torch.manual_seed(args.seed)
    data_class = _import_class(f"data.{args.data_class}")
    model_class = _import_class(f"models.{args.model_class}")
    data = data_class(args)
    data_config = data.get_data_config()

    config = AutoConfig.from_pretrained(args.model_name_or_path)
    config.num_labels = data_config["num_labels"]
    model = model_class.from_pretrained(args.model_name_or_path, config=config)

    lit_model = TransformerLitModel(args=args, model=model, data_config=data_config)

    logger = pl.loggers.TensorBoardLogger("training/logs")
    # Hide lines below until Lab 5
    if args.wandb:
        logger = pl.loggers.WandbLogger(project="text_cls")
        logger.log_hyperparams(vars(args))
    # Hide lines above until Lab 5
    early_callback = pl.callbacks.EarlyStopping(monitor="Eval/acc", mode="max", patience=5)
    model_checkpoint = pl.callbacks.ModelCheckpoint(monitor="Eval/acc", mode="max",
        filename='{epoch}-{Eval/acc:.2f}',
        dirpath="output"
    )

    callbacks = [early_callback, model_checkpoint]

    # args.weights_summary = "full"  # Print full summary of the model
    trainer = pl.Trainer.from_argparse_args(args, callbacks=callbacks, logger=logger, default_root_dir="training/logs")

    trainer.tune(lit_model, datamodule=data)  # If passing --auto_lr_find, this will set learning rate

    trainer.fit(lit_model, datamodule=data)
    trainer.test(lit_model, datamodule=data)

    path, filename = os.path.splitext(model_checkpoint.best_model_path)
    trainer.model.save_pretrained(path)


if __name__ == "__main__":

    main()