main.py

import argparse, os, sys, datetime, glob, importlib, csv
import numpy as np
import time
import torch
import torchvision
import pytorch_lightning as pl

from packaging import version
from omegaconf import OmegaConf
from torch.utils.data import random_split, DataLoader, Dataset, Subset
from functools import partial
from PIL import Image

from pytorch_lightning import seed_everything
from pytorch_lightning.trainer import Trainer
from pytorch_lightning.callbacks import ModelCheckpoint, Callback, LearningRateMonitor
from pytorch_lightning.utilities.distributed import rank_zero_only
from pytorch_lightning.utilities import rank_zero_info

from ldm.data.base import Txt2ImgIterableBaseDataset
from ldm.util import instantiate_from_config

def load_model_from_config(config, ckpt, strict=True, verbose=False):
    print(f"Loading model from {ckpt}")
    config.model.params.ckpt_path = ckpt
    model = instantiate_from_config(config.model)

    # if model.use_ema: # TODO move inside
    #     model.model_ema(model.model, force_overwrite=True)

    return model

def get_parser(**parser_kwargs):
    def str2bool(v):
        if isinstance(v, bool):
            return v
        if v.lower() in ("yes", "true", "t", "y", "1"):
            return True
        elif v.lower() in ("no", "false", "f", "n", "0"):
            return False
        else:
            raise argparse.ArgumentTypeError("Boolean value expected.")

    parser = argparse.ArgumentParser(**parser_kwargs)
    parser.add_argument(
        "-n",
        "--name",
        type=str,
        const=True,
        default="",
        nargs="?",
        help="postfix for logdir",
    )
    parser.add_argument(
        "-r",
        "--resume",
        type=str,
        const=True,
        default="",
        nargs="?",
        help="resume from logdir or checkpoint in logdir",
    )
    parser.add_argument(
        "-b",
        "--base",
        nargs="*",
        metavar="base_config.yaml",
        help="paths to base configs. Loaded from left-to-right. "
             "Parameters can be overwritten or added with command-line options of the form `--key value`.",
        default=list(),
    )
    parser.add_argument(
        "-t",
        "--train",
        type=str2bool,
        const=True,
        default=False,
        nargs="?",
        help="train",
    )
    parser.add_argument(
        "--no-test",
        type=str2bool,
        const=True,
        default=False,
        nargs="?",
        help="disable test",
    )
    parser.add_argument(
        "-p",
        "--project",
        help="name of new or path to existing project"
    )
    parser.add_argument(
        "-d",
        "--debug",
        type=str2bool,
        nargs="?",
        const=True,
        default=False,
        help="enable post-mortem debugging",
    )
    parser.add_argument(
        "-s",
        "--seed",
        type=int,
        default=23,
        help="seed for seed_everything",
    )
    parser.add_argument(
        "-f",
        "--postfix",
        type=str,
        default="",
        help="post-postfix for default name",
    )
    parser.add_argument(
        "-l",
        "--logdir",
        type=str,
        default="logs",
        help="directory for logging dat shit",
    )
    parser.add_argument(
        "--scale_lr",
        type=str2bool,
        nargs="?",
        const=True,
        default=False,
        help="scale base-lr by ngpu * batch_size * n_accumulate",
    )
    parser.add_argument(
        "--ignore-unexpected-keys",
        type=str2bool,
        nargs="?",
        const=True,
        default=False,
        help="The released model is not exact the configured ones.",
    )
    parser.add_argument(
        '--num_shards',
        type=int,
        default=None
    )
    parser.add_argument(
        '--shard_id',
        type=int,
        default=None
    )
    return parser


def nondefault_trainer_args(opt):
    parser = argparse.ArgumentParser()
    parser = Trainer.add_argparse_args(parser)
    args = parser.parse_args([])
    return sorted(k for k in vars(args) if getattr(opt, k) != getattr(args, k))


class WrappedDataset(Dataset):
    """Wraps an arbitrary object with __len__ and __getitem__ into a pytorch dataset"""

    def __init__(self, dataset):
        self.data = dataset

    def __len__(self):
        return len(self.data)

    def __getitem__(self, idx):
        return self.data[idx]


def worker_init_fn(_):
    worker_info = torch.utils.data.get_worker_info()

    dataset = worker_info.dataset
    worker_id = worker_info.id

    if isinstance(dataset, Txt2ImgIterableBaseDataset):
        split_size = dataset.num_records // worker_info.num_workers
        # reset num_records to the true number to retain reliable length information
        dataset.sample_ids = dataset.valid_ids[worker_id * split_size:(worker_id + 1) * split_size]
        current_id = np.random.choice(len(np.random.get_state()[1]), 1)
        return np.random.seed(np.random.get_state()[1][current_id] + worker_id)
    else:
        return np.random.seed(np.random.get_state()[1][0] + worker_id)


class DataModuleFromConfig(pl.LightningDataModule):
    def __init__(self, batch_size, train=None, validation=None, test=None, predict=None,
                 wrap=False, num_workers=None, shuffle_test_loader=False, use_worker_init_fn=False,
                 shuffle_val_dataloader=False, prefetch_factor=4):
        super().__init__()
        self.batch_size = batch_size
        self.dataset_configs = dict()
        self.num_workers = num_workers if num_workers is not None else batch_size * 2
        self.use_worker_init_fn = use_worker_init_fn
        if train is not None:
            self.dataset_configs["train"] = train
            self.train_dataloader = self._train_dataloader
        if validation is not None:
            self.dataset_configs["validation"] = validation
            self.val_dataloader = partial(self._val_dataloader, shuffle=shuffle_val_dataloader)
        if test is not None:
            self.dataset_configs["test"] = test
            self.test_dataloader = partial(self._test_dataloader, shuffle=shuffle_test_loader)
        if predict is not None:
            self.dataset_configs["predict"] = predict
            self.predict_dataloader = self._predict_dataloader
        self.wrap = wrap
        self.prefetch_factor = prefetch_factor

    def prepare_data(self):
        for data_cfg in self.dataset_configs.values():
            instantiate_from_config(data_cfg)

    def setup(self, stage=None):
        self.datasets = dict(
            (k, instantiate_from_config(self.dataset_configs[k]))
            for k in self.dataset_configs)
        if self.wrap:
            for k in self.datasets:
                self.datasets[k] = WrappedDataset(self.datasets[k])

    def _train_dataloader(self):
        is_iterable_dataset = isinstance(self.datasets['train'], torch.utils.data.IterableDataset)
        is_ldm_iterable_dataset = isinstance(self.datasets['train'], Txt2ImgIterableBaseDataset)
        if is_ldm_iterable_dataset or self.use_worker_init_fn:
            init_fn = worker_init_fn
        else:
            init_fn = None
        return DataLoader(self.datasets["train"], batch_size=self.batch_size,
                          num_workers=self.num_workers, shuffle=False if is_iterable_dataset else True,
                          worker_init_fn=init_fn, prefetch_factor=self.prefetch_factor)

    def _val_dataloader(self, shuffle=False):
        if isinstance(self.datasets['validation'], Txt2ImgIterableBaseDataset) or self.use_worker_init_fn:
            init_fn = worker_init_fn
        else:
            init_fn = None
        return DataLoader(self.datasets["validation"],
                          batch_size=self.batch_size,
                          num_workers=self.num_workers,
                          worker_init_fn=init_fn,
                          shuffle=shuffle,
                          prefetch_factor=self.prefetch_factor
                          )

    def _test_dataloader(self, shuffle=False):
        is_iterable_dataset = isinstance(self.datasets['train'], Txt2ImgIterableBaseDataset)
        if is_iterable_dataset or self.use_worker_init_fn:
            init_fn = worker_init_fn
        else:
            init_fn = None

        # do not shuffle dataloader for iterable dataset
        shuffle = shuffle and (not is_iterable_dataset)

        return DataLoader(self.datasets["test"], batch_size=self.batch_size,
                          num_workers=self.num_workers, worker_init_fn=init_fn, shuffle=shuffle)

    def _predict_dataloader(self, shuffle=False):
        if isinstance(self.datasets['predict'], Txt2ImgIterableBaseDataset) or self.use_worker_init_fn:
            init_fn = worker_init_fn
        else:
            init_fn = None
        return DataLoader(self.datasets["predict"], batch_size=self.batch_size,
                          num_workers=self.num_workers, worker_init_fn=init_fn)


class SetupCallback(Callback):
    def __init__(self, resume, now, logdir, ckptdir, cfgdir, config, lightning_config):
        super().__init__()
        self.resume = resume
        self.now = now
        self.logdir = logdir
        self.ckptdir = ckptdir
        self.cfgdir = cfgdir
        self.config = config
        self.lightning_config = lightning_config

    # def on_keyboard_interrupt(self, trainer, pl_module):
    #     if not self.debug and trainer.global_rank == 0:
    #         print("Summoning checkpoint.")
    #         ckpt_path = os.path.join(self.ckptdir, "last.ckpt")
    #         trainer.save_checkpoint(ckpt_path)

    def on_pretrain_routine_start(self, trainer, pl_module):
        if trainer.global_rank == 0:
            # Create logdirs and save configs
            os.makedirs(self.logdir, exist_ok=True)
            os.makedirs(self.ckptdir, exist_ok=True)
            os.makedirs(self.cfgdir, exist_ok=True)

            if "callbacks" in self.lightning_config:
                if 'metrics_over_trainsteps_checkpoint' in self.lightning_config['callbacks']:
                    os.makedirs(os.path.join(self.ckptdir, 'trainstep_checkpoints'), exist_ok=True)
            print("Project config")
            print(OmegaConf.to_yaml(self.config))
            OmegaConf.save(self.config,
                           os.path.join(self.cfgdir, "{}-project.yaml".format(self.now)))

            print("Lightning config")
            print(OmegaConf.to_yaml(self.lightning_config))
            OmegaConf.save(OmegaConf.create({"lightning": self.lightning_config}),
                           os.path.join(self.cfgdir, "{}-lightning.yaml".format(self.now)))

        else:
            pass
            # # ModelCheckpoint callback created log directory --- remove it
            # if not MULTINODE_HACKS and not self.resume and os.path.exists(self.logdir):
            #     dst, name = os.path.split(self.logdir)
            #     dst = os.path.join(dst, "child_runs", name)
            #     os.makedirs(os.path.split(dst)[0], exist_ok=True)
            #     try:
            #         os.rename(self.logdir, dst)
            #     except FileNotFoundError:
            #         pass

class TextLogger(Callback):
    def __init__(self, batch_frequency, max_texts, clamp=True, increase_log_steps=True,
                 rescale=True, disabled=False, log_on_batch_idx=False, log_first_step=False,
                 log_texts_kwargs=None):
        super().__init__()
        self.rescale = rescale
        self.batch_freq = batch_frequency
        self.max_texts = max_texts
        self.logger_log_texts = {
            pl.loggers.TensorBoardLogger: self._tensorboard,
        }
        self.log_steps = [2 ** n for n in range(int(np.log2(self.batch_freq)) + 1)]
        if not increase_log_steps:
            self.log_steps = [self.batch_freq]
        self.clamp = clamp
        self.disabled = disabled
        self.log_on_batch_idx = log_on_batch_idx
        self.log_texts_kwargs = log_texts_kwargs if log_texts_kwargs else {}
        self.log_first_step = log_first_step
    
    @rank_zero_only
    def _tensorboard(self, pl_module, texts, batch_idx, split):
        pass
    
    @rank_zero_only
    def log_local(self, save_dir, split, texts,
                  global_step, current_epoch, batch_idx):
        root = os.path.join(save_dir, "texts", split)
        for k in texts:
            filename = "{}_gs-{:06}_e-{:06}_b-{:06}".format(
                k,
                global_step,
                current_epoch,
                batch_idx)
            path = os.path.join(root, filename)
            os.makedirs(os.path.split(path)[0], exist_ok=True)
            print("path:", path)
            with open(path, 'w') as log_file:
                for text in texts[k]:
                    log_file.write(text+'\n')
    
    def log_txt(self, pl_module, batch, batch_idx, split="train"):
        check_idx = batch_idx if self.log_on_batch_idx else pl_module.global_step
        if (self.check_frequency(check_idx) and  # batch_idx % self.batch_freq == 0
                hasattr(pl_module, "log_texts") and
                callable(pl_module.log_texts) and
                batch_idx > 5 and
                self.max_texts > 0):
            logger = type(pl_module.logger)
            
            is_train = pl_module.training
            if is_train:
                pl_module.eval()
                
            with torch.no_grad():
                texts = pl_module.log_texts(batch, split=split, **self.log_texts_kwargs)
            
            for k in texts:
                N = min(len(texts[k]), self.max_texts)
                texts[k] = texts[k][:N]
            
            self.log_local(pl_module.logger.save_dir, split, texts,
                           pl_module.global_step, pl_module.current_epoch, batch_idx)
            
            logger_log_texts = self.logger_log_texts.get(logger, lambda *args, **kwargs: None)
            logger_log_texts(pl_module, texts, pl_module.global_step, split)
            if is_train:
                pl_module.train()
    
    def check_frequency(self, check_idx):
        if ((check_idx % self.batch_freq) == 0 or (check_idx in self.log_steps)) and (
                check_idx > 0 or self.log_first_step):
            try:
                self.log_steps.pop(0)
            except IndexError as e:
                print(e)
                pass
            return True
        return False

    def on_train_batch_end(self, trainer, pl_module, outputs, batch, batch_idx):
        if not self.disabled and (pl_module.global_step > 0 or self.log_first_step):
            self.log_txt(pl_module, batch, batch_idx, split="train")

    def on_validation_batch_end(self, trainer, pl_module, outputs, batch, batch_idx, dataloader_idx):
        if not self.disabled and pl_module.global_step > 0:
            self.log_txt(pl_module, batch, batch_idx, split="val")
        if hasattr(pl_module, 'calibrate_grad_norm'):
            if (pl_module.calibrate_grad_norm and batch_idx % 25 == 0) and batch_idx > 0:
                self.log_gradients(trainer, pl_module, batch_idx=batch_idx)

class ImageLogger(Callback):
    def __init__(self, batch_frequency, max_images, clamp=True, increase_log_steps=True,
                 rescale=True, disabled=False, log_on_batch_idx=False, log_first_step=False,
                 log_images_kwargs=None):
        super().__init__()
        self.rescale = rescale
        self.batch_freq = batch_frequency
        self.max_images = max_images
        self.logger_log_images = {
            pl.loggers.TestTubeLogger: self._testtube,
        }
        self.log_steps = [2 ** n for n in range(int(np.log2(self.batch_freq)) + 1)]
        if not increase_log_steps:
            self.log_steps = [self.batch_freq]
        self.clamp = clamp
        self.disabled = disabled
        self.log_on_batch_idx = log_on_batch_idx
        self.log_images_kwargs = log_images_kwargs if log_images_kwargs else {}
        self.log_first_step = log_first_step

    @rank_zero_only
    def _testtube(self, pl_module, images, batch_idx, split):
        for k in images:
            grid = torchvision.utils.make_grid(images[k])
            grid = (grid + 1.0) / 2.0  # -1,1 -> 0,1; c,h,w

            tag = f"{split}/{k}"
            pl_module.logger.experiment.add_image(
                tag, grid,
                global_step=pl_module.global_step)

    @rank_zero_only
    def log_local(self, save_dir, split, images,
                  global_step, current_epoch, batch_idx):
        root = os.path.join(save_dir, "images", split)
        for k in images:
            grid = torchvision.utils.make_grid(images[k], nrow=4)
            if self.rescale:
                grid = (grid + 1.0) / 2.0  # -1,1 -> 0,1; c,h,w
            grid = grid.transpose(0, 1).transpose(1, 2).squeeze(-1)
            grid = grid.numpy()
            grid = (grid * 255).astype(np.uint8)
            filename = "{}_gs-{:06}_e-{:06}_b-{:06}.png".format(
                k,
                global_step,
                current_epoch,
                batch_idx)
            path = os.path.join(root, filename)
            os.makedirs(os.path.split(path)[0], exist_ok=True)
            Image.fromarray(grid).save(path)

    def log_img(self, pl_module, batch, batch_idx, split="train"):
        check_idx = batch_idx if self.log_on_batch_idx else pl_module.global_step
        if (self.check_frequency(check_idx) and  # batch_idx % self.batch_freq == 0
                hasattr(pl_module, "log_images") and
                callable(pl_module.log_images) and
                batch_idx > 5 and
                self.max_images > 0):
            logger = type(pl_module.logger)

            is_train = pl_module.training
            if is_train:
                pl_module.eval()

            with torch.no_grad():
                images = pl_module.log_images(batch, split=split, **self.log_images_kwargs)

            for k in images:
                N = min(images[k].shape[0], self.max_images)
                images[k] = images[k][:N]
                if isinstance(images[k], torch.Tensor):
                    images[k] = images[k].detach().cpu()
                    if self.clamp:
                        images[k] = torch.clamp(images[k], -1., 1.)

            self.log_local(pl_module.logger.save_dir, split, images,
                           pl_module.global_step, pl_module.current_epoch, batch_idx)

            logger_log_images = self.logger_log_images.get(logger, lambda *args, **kwargs: None)
            logger_log_images(pl_module, images, pl_module.global_step, split)

            if is_train:
                pl_module.train()

    def check_frequency(self, check_idx):
        if ((check_idx % self.batch_freq) == 0 or (check_idx in self.log_steps)) and (
                check_idx > 0 or self.log_first_step):
            try:
                self.log_steps.pop(0)
            except IndexError as e:
                print(e)
                pass
            return True
        return False

    def on_train_batch_end(self, trainer, pl_module, outputs, batch, batch_idx, dataloader_idx):
        if not self.disabled and (pl_module.global_step > 0 or self.log_first_step):
            self.log_img(pl_module, batch, batch_idx, split="train")

    def on_validation_batch_end(self, trainer, pl_module, outputs, batch, batch_idx, dataloader_idx):
        if not self.disabled and pl_module.global_step > 0:
            self.log_img(pl_module, batch, batch_idx, split="val")
        if hasattr(pl_module, 'calibrate_grad_norm'):
            if (pl_module.calibrate_grad_norm and batch_idx % 25 == 0) and batch_idx > 0:
                self.log_gradients(trainer, pl_module, batch_idx=batch_idx)


class CUDACallback(Callback):
    # see https://github.com/SeanNaren/minGPT/blob/master/mingpt/callback.py
    def on_train_epoch_start(self, trainer, pl_module):
        # Reset the memory use counter
        torch.cuda.reset_peak_memory_stats(trainer.root_gpu)
        torch.cuda.synchronize(trainer.root_gpu)
        self.start_time = time.time()

    def on_train_epoch_end(self, trainer, pl_module):
        torch.cuda.synchronize(trainer.root_gpu)
        max_memory = torch.cuda.max_memory_allocated(trainer.root_gpu) / 2 ** 20
        epoch_time = time.time() - self.start_time

        try:
            max_memory = trainer.training_type_plugin.reduce(max_memory)
            epoch_time = trainer.training_type_plugin.reduce(epoch_time)

            rank_zero_info(f"Average Epoch time: {epoch_time:.2f} seconds")
            rank_zero_info(f"Average Peak memory {max_memory:.2f}MiB")
        except AttributeError:
            pass


class SingleImageLogger(Callback):
    """does not save as grid but as single images"""
    def __init__(self, batch_frequency, max_images, clamp=True, increase_log_steps=True,
                 rescale=True, disabled=False, log_on_batch_idx=False, log_first_step=False,
                 log_images_kwargs=None, log_always=False):
        super().__init__()
        self.rescale = rescale
        self.batch_freq = batch_frequency
        self.max_images = max_images
        self.logger_log_images = {
            pl.loggers.TestTubeLogger: self._testtube,
        }
        self.log_steps = [2 ** n for n in range(int(np.log2(self.batch_freq)) + 1)]
        if not increase_log_steps:
            self.log_steps = [self.batch_freq]
        self.clamp = clamp
        self.disabled = disabled
        self.log_on_batch_idx = log_on_batch_idx
        self.log_images_kwargs = log_images_kwargs if log_images_kwargs else {}
        self.log_first_step = log_first_step
        self.log_always = log_always

    @rank_zero_only
    def _testtube(self, pl_module, images, batch_idx, split):
        for k in images:
            grid = torchvision.utils.make_grid(images[k])
            grid = (grid + 1.0) / 2.0  # -1,1 -> 0,1; c,h,w

            tag = f"{split}/{k}"
            pl_module.logger.experiment.add_image(
                tag, grid,
                global_step=pl_module.global_step)

    @rank_zero_only
    def log_local(self, save_dir, split, images,
                  global_step, current_epoch, batch_idx):
        root = os.path.join(save_dir, "images", split)
        os.makedirs(root, exist_ok=True)
        for k in images:
            subroot = os.path.join(root, k)
            os.makedirs(subroot, exist_ok=True)
            base_count = len(glob.glob(os.path.join(subroot, "*.png")))
            for img in images[k]:
                if self.rescale:
                    img = (img + 1.0) / 2.0  # -1,1 -> 0,1; c,h,w
                img = img.transpose(0, 1).transpose(1, 2).squeeze(-1)
                img = img.numpy()
                img = (img * 255).astype(np.uint8)
                filename = "{}_gs-{:06}_e-{:06}_b-{:06}_{:08}.png".format(
                    k,
                    global_step,
                    current_epoch,
                    batch_idx,
                    base_count)
                path = os.path.join(subroot, filename)
                Image.fromarray(img).save(path)
                base_count += 1

    def log_img(self, pl_module, batch, batch_idx, split="train", save_dir=None):
        check_idx = batch_idx if self.log_on_batch_idx else pl_module.global_step
        if (self.check_frequency(check_idx) and  # batch_idx % self.batch_freq == 0
                hasattr(pl_module, "log_images") and
                callable(pl_module.log_images) and
                self.max_images > 0) or self.log_always:
            logger = type(pl_module.logger)

            is_train = pl_module.training
            if is_train:
                pl_module.eval()

            with torch.no_grad():
                images = pl_module.log_images(batch, split=split, **self.log_images_kwargs)

            for k in images:
                N = min(images[k].shape[0], self.max_images)
                images[k] = images[k][:N]
                if isinstance(images[k], torch.Tensor):
                    images[k] = images[k].detach().cpu()
                    if self.clamp:
                        images[k] = torch.clamp(images[k], -1., 1.)

            self.log_local(pl_module.logger.save_dir if save_dir is None else save_dir, split, images,
                           pl_module.global_step, pl_module.current_epoch, batch_idx)

            logger_log_images = self.logger_log_images.get(logger, lambda *args, **kwargs: None)
            logger_log_images(pl_module, images, pl_module.global_step, split)

            if is_train:
                pl_module.train()

    def check_frequency(self, check_idx):
        if ((check_idx % self.batch_freq) == 0 or (check_idx in self.log_steps)) and (
                check_idx > 0 or self.log_first_step):
            try:
                self.log_steps.pop(0)
            except IndexError as e:
                print(e)
            return True
        return False


if __name__ == "__main__":
    torch.backends.cudnn.benchmark=False; torch.backends.cudnn.deterministic=True
    # custom parser to specify config files, train, test and debug mode,
    # postfix, resume.
    # `--key value` arguments are interpreted as arguments to the trainer.
    # `nested.key=value` arguments are interpreted as config parameters.
    # configs are merged from left-to-right followed by command line parameters.

    # model:
    #   base_learning_rate: float
    #   target: path to lightning module
    #   params:
    #       key: value
    # data:
    #   target: main.DataModuleFromConfig
    #   params:
    #      batch_size: int
    #      wrap: bool
    #      train:
    #          target: path to train dataset
    #          params:
    #              key: value
    #      validation:
    #          target: path to validation dataset
    #          params:
    #              key: value
    #      test:
    #          target: path to test dataset
    #          params:
    #              key: value
    # lightning: (optional, has sane defaults and can be specified on cmdline)
    #   trainer:
    #       additional arguments to trainer
    #   logger:
    #       logger to instantiate
    #   modelcheckpoint:
    #       modelcheckpoint to instantiate
    #   callbacks:
    #       callback1:
    #           target: importpath
    #           params:
    #               key: value

    now = datetime.datetime.now().strftime("%Y-%m-%dT%H-%M-%S")

    # add cwd for convenience and to make classes in this file available when
    # running as `python main.py`
    # (in particular `main.DataModuleFromConfig`)
    sys.path.append(os.getcwd())

    parser = get_parser()
    parser = Trainer.add_argparse_args(parser)

    opt, unknown = parser.parse_known_args()
    if opt.name and opt.resume:
        raise ValueError(
            "-n/--name and -r/--resume cannot be specified both."
            "If you want to resume training in a new log folder, "
            "use -n/--name in combination with --resume_from_checkpoint"
        )
    # # ---- BEG auto-detect resume ----
    # # detected_ckpt_path = None
    # # opt.name is {job_id}.{jobname}, folder name is {time}_{opt.name}
    # def exact_cont(s):
    #     if s.endswith('continue'):
    #         return int(s.split('_')[-1][:-8])
    #     else:
    #         return 0
    
    # best_cont = -1
    # if os.path.exists(opt.logdir):
    #     for subfolder in os.listdir(opt.logdir):
    #         if opt.name in subfolder:
    #             cont_num = exact_cont(subfolder)
    #             ckpt_path = os.path.join(opt.logdir, subfolder, 'checkpoints', "last.ckpt")
    #             if not os.path.exists(ckpt_path):
    #                 print(f'ignore no-ckpt runs {subfolder}.')
    #                 continue # ignore no-ckpt runs
    #             if cont_num > best_cont:
    #                 best_cont = cont_num # lastest version
    #                 opt.resume = os.path.join(opt.logdir, subfolder)
    
    # # ---- END auto-detect resume ----
    if opt.resume:
        if not os.path.exists(opt.resume):
            raise ValueError("Cannot find {}".format(opt.resume))
        if os.path.isfile(opt.resume):
            paths = opt.resume.split("/")
            old_logdir = "/".join(paths[:-2])
            ckpt = opt.resume
        else:
            assert os.path.isdir(opt.resume), opt.resume
            old_logdir = opt.resume.rstrip("/")
            ckpt = os.path.join(old_logdir, "checkpoints", "last.ckpt")
        
        if os.path.exists(ckpt): # it is possible the first ckpt is not saved
            print(f'Found last.ckpt {ckpt}, skip original resume_from_checkpoint {opt.resume_from_checkpoint}.')
            opt.resume_from_checkpoint = ckpt # no need to load from the original source
            # change seed to avoid trained batches
            opt.seed = int(time.time()) % 1000000
        else:
            print(f'Resuming from {opt.resume} but no ckpt. Only reuse configs as possible.')
        base_configs = sorted(glob.glob(os.path.join(old_logdir, "configs/*.yaml")))
        opt.base = base_configs + opt.base
        _tmp = old_logdir.split("/")[-1]
        if _tmp.endswith('continue'):
            _tmp = '_'.join(_tmp.split('_')[:-1])
        # nowname = _tmp + '_' + str(best_cont+1) + 'continue'
        nowname = _tmp
    else:
        if opt.name:
            name = "_" + opt.name
        elif opt.base:
            cfg_fname = os.path.split(opt.base[0])[-1]
            cfg_name = os.path.splitext(cfg_fname)[0]
            name = "_" + cfg_name
        else:
            name = ""
        nowname = now + name + opt.postfix

    logdir = os.path.join(opt.logdir, nowname)
    print(f'Logdir: {logdir}')
    if not os.path.isdir(logdir):
        os.makedirs(logdir, exist_ok=True)

    ckptdir = os.path.join(logdir, "checkpoints")
    cfgdir = os.path.join(logdir, "configs")
    seed_everything(opt.seed)

    try:
        # init and save configs
        configs = [OmegaConf.load(cfg) for cfg in opt.base]
        cli = OmegaConf.from_dotlist(unknown)
        config = OmegaConf.merge(*configs, cli)
        
        # add additional params
        config.model.params.logdir = logdir
        if opt.num_shards is not None and opt.shard_id is not None:
            config.model.params.num_shards = opt.num_shards
            config.model.params.shard_id = opt.shard_id
            config.data.params.test.params.num_shards = opt.num_shards
            config.data.params.test.params.shard_id = opt.shard_id
        
        # model
        # resume_from_checkpoint now control this pre-loading,
        # Trainer resume is not relevant to this arg.
        if opt.resume_from_checkpoint:
            model = load_model_from_config(
                config,
                opt.resume_from_checkpoint,
                strict=not opt.ignore_unexpected_keys
                )
            opt.resume_from_checkpoint = None
        else:
            model = instantiate_from_config(config.model)
            
        # lightning & trainer
        lightning_config = config.pop("lightning", OmegaConf.create())
        # merge trainer cli with config
        trainer_config = lightning_config.get("trainer", OmegaConf.create())
        # default to ddp
        trainer_config["accelerator"] = "cuda"
        for k in nondefault_trainer_args(opt):
            # override default and from config by cli (for direct pass training args instead of .lightning.training)
            trainer_config[k] = getattr(opt, k)
            
        gpuinfo = trainer_config["devices"]
        print(f"Running on GPUs {gpuinfo}")
        cpu = False
        trainer_opt = argparse.Namespace(**trainer_config)
        lightning_config.trainer = trainer_config

        # trainer and callbacks
        trainer_kwargs = dict()

        # default logger configs
        default_logger_cfgs = {
            "wandb": {
                "target": "pytorch_lightning.loggers.WandbLogger",
                "params": {
                    "name": nowname,
                    "save_dir": logdir,
                    "offline": True,
                    "id": nowname,
                }
            },
            "tensorboard": {
                "target": "pytorch_lightning.loggers.TensorBoardLogger",
                "params": {
                    "name": "tensorboard",
                    "save_dir": logdir,
                }
            },
        }
        default_logger_cfg = default_logger_cfgs["tensorboard"]
        if "logger" in lightning_config:
            logger_cfg = lightning_config.logger
        else:
            logger_cfg = OmegaConf.create()
        logger_cfg = OmegaConf.merge(default_logger_cfg, logger_cfg)
        # TODO how to know the rank?
        trainer_kwargs["logger"] = instantiate_from_config(logger_cfg)

        # modelcheckpoint - use TrainResult/EvalResult(checkpoint_on=metric) to
        # specify which metric is used to determine best models
        default_modelckpt_cfg = {
            "target": "pytorch_lightning.callbacks.ModelCheckpoint",
            "params": {
                "dirpath": ckptdir,
                "filename": "{epoch:06}",
                "verbose": True,
                "save_last": True,
                "every_n_train_steps": 5000,
                "save_weights_only": True, # NOTE fit(ckpt=...) requires optimizer_states, but lightning is buggy.
                "save_top_k": 2
            }
        }
        if hasattr(model, "monitor") and model.monitor is not None:
            print(f"Monitoring {model.monitor} as checkpoint metric.")
            default_modelckpt_cfg["params"]["monitor"] = model.monitor
            default_modelckpt_cfg["params"]["mode"] = model.mode if hasattr(model, 'mode') else 'min'
            default_modelckpt_cfg["params"]["save_top_k"] = 3

        if "modelcheckpoint" in lightning_config:
            modelckpt_cfg = lightning_config.modelcheckpoint
        else:
            modelckpt_cfg =  OmegaConf.create()
        modelckpt_cfg = OmegaConf.merge(default_modelckpt_cfg, modelckpt_cfg)
        print(f"Merged modelckpt-cfg: \n{modelckpt_cfg}")

        # add callback which sets up log directory
        default_callbacks_cfg = {
            "setup_callback": {
                "target": "main.SetupCallback",
                "params": {
                    "resume": opt.resume,
                    "now": now,
                    "logdir": logdir,
                    "ckptdir": ckptdir,
                    "cfgdir": cfgdir,
                    "config": config,
                    "lightning_config": lightning_config,
                }
            },
            "text_logger": {
                "target": "main.TextLogger",
                "params": {
                    "batch_frequency": 750,
                    "max_texts": 16,
                }
            },
            # "image_logger": {
            #     "target": "main.ImageLogger",
            #     "params": {
            #         "batch_frequency": 750,
            #         "max_images": 4,
            #         "clamp": True
            #     }
            # },
            "learning_rate_logger": {
                "target": "main.LearningRateMonitor",
                "params": {
                    "logging_interval": "step",
                    # "log_momentum": True
                }
            },
            "cuda_callback": {
                "target": "main.CUDACallback"
            },
        }
        assert version.parse(pl.__version__) >= version.parse('1.4.0')
        default_callbacks_cfg.update({'checkpoint_callback': modelckpt_cfg})

        if "callbacks" in lightning_config:
            callbacks_cfg = lightning_config.callbacks
        else:
            callbacks_cfg = OmegaConf.create()
        
        callbacks_cfg = OmegaConf.merge(default_callbacks_cfg, callbacks_cfg)

        trainer_kwargs["callbacks"] = [instantiate_from_config(callbacks_cfg[k]) for k in callbacks_cfg]
        
        trainer = Trainer.from_argparse_args(trainer_opt, **trainer_kwargs)
        trainer.logdir = logdir  ###
        
        # data
        data = instantiate_from_config(config.data)
        # NOTE according to https://pytorch-lightning.readthedocs.io/en/latest/datamodules.html
        # calling these ourselves should not be necessary but it is.
        # lightning still takes care of proper multiprocessing though
        data.prepare_data()
        data.setup()
        print("#### Data #####")
        for k in data.datasets:
            ld = -1 if isinstance(data.datasets[k], torch.utils.data.IterableDataset) else len(data.datasets[k])
            print(f"{k}, {data.datasets[k].__class__.__name__}, {ld}")
        
        # configure learning rate
        bs, base_lr = config.data.params.batch_size, config.model.base_learning_rate
        if not cpu:
            ngpu = len(lightning_config.trainer.devices.strip(",").split(','))
        else:
            ngpu = 1
        if 'accumulate_grad_batches' in lightning_config.trainer:
            accumulate_grad_batches = lightning_config.trainer.accumulate_grad_batches
        else:
            accumulate_grad_batches = 1
        print(f"accumulate_grad_batches = {accumulate_grad_batches}")
        lightning_config.trainer.accumulate_grad_batches = accumulate_grad_batches
        if opt.scale_lr:
            model.learning_rate = accumulate_grad_batches * ngpu * bs * base_lr
            print(
                "Setting learning rate to {:.2e} = {} (accumulate_grad_batches) * {} (num_gpus) * {} (batchsize) * {:.2e} (base_lr)".format(
                    model.learning_rate, accumulate_grad_batches, ngpu, bs, base_lr))
        else:
            model.learning_rate = base_lr
            print("++++ NOT USING LR SCALING ++++")
            print(f"Setting learning rate to {model.learning_rate:.2e}")


        # allow checkpointing via USR1
        def melk(*args, **kwargs):
            # run all checkpoint hooks
            if trainer.global_rank == 0:
                print("Summoning checkpoint.")
                ckpt_path = os.path.join(ckptdir, "last.ckpt")
                trainer.save_checkpoint(ckpt_path)


        def divein(*args, **kwargs):
            if trainer.global_rank == 0:
                import pudb;
                pudb.set_trace()


        import signal

        # signal.signal(signal.SIGUSR1, melk)
        signal.signal(signal.SIGUSR2, divein)

        # run
        if opt.train:
            try:
                trainer.fit(model, data)
            except Exception:
                melk()
                raise
        if not opt.no_test and not trainer.interrupted:
            trainer.test(model, data)
    except Exception:
        if opt.debug and trainer.global_rank == 0:
            try:
                import pudb as debugger
            except ImportError:
                import pdb as debugger
            debugger.post_mortem()
        raise
    finally:
        # move newly created debug project to debug_runs
        if opt.debug and not opt.resume and trainer.global_rank == 0:
            dst, name = os.path.split(logdir)
            dst = os.path.join(dst, "debug_runs", name)
            os.makedirs(os.path.split(dst)[0], exist_ok=True)
            os.rename(logdir, dst)
        if trainer.global_rank == 0:
            print(trainer.profiler.summary())