run_with_submitit.py

# Copyright IBM All Rights Reserved.
# SPDX-License-Identifier: Apache-2.0

"""
A script to run multinode training with submitit.

Mostly copy-paste from https://github.com/facebookresearch/deit/blob/main/run_with_submitit.py
"""

import argparse
import os
import uuid
from pathlib import Path

import main as classification
import submitit


def parse_args():
    classification_parser = classification.get_args_parser()
    parser = argparse.ArgumentParser("Submitit for CrossViT", parents=[classification_parser])
    parser.add_argument("--ngpus", default=8, type=int, help="Number of gpus to request on each node")
    parser.add_argument("--nodes", default=2, type=int, help="Number of nodes to request")
    parser.add_argument("--timeout", default=360, type=int, help="Duration of the job")
    parser.add_argument("--job_dir", default="", type=str, help="Job dir. Leave empty for automatic.")
    parser.add_argument("--suffix", default="", type=str, help="Job dir. Leave empty for automatic.")

    parser.add_argument("--partition", default="npl", type=str, help="Partition where to submit")
    return parser.parse_args()


def get_shared_folder() -> Path:

    if not (Path.cwd() / "checkpoint").exists():
        (Path.cwd() / "checkpoint").mkdir(exist_ok=True)

    if (Path.cwd() / "checkpoint").is_dir():
        p = Path.cwd() / "checkpoint" / "experiments"
        p.mkdir(exist_ok=True)
        return p
    raise RuntimeError("No shared folder available")


def get_init_file():
    # Init file must not exist, but it's parent dir must exist.
    os.makedirs(str(get_shared_folder()), exist_ok=True)
    init_file = get_shared_folder() / f"{uuid.uuid4().hex}_init"
    if init_file.exists():
        os.remove(str(init_file))
    return init_file


class Trainer(object):
    def __init__(self, args):
        self.args = args

    def __call__(self):
        import main as classification

        self._setup_gpu_args()
        classification.main(self.args)

    def checkpoint(self):
        import os
        import submitit

        self.args.seed = self.args.seed + 1
        self.args.dist_url = get_init_file().as_uri()
        checkpoint_file = os.path.join(self.args.output_dir, "checkpoint.pth")
        if os.path.exists(checkpoint_file):
            self.args.resume = checkpoint_file
        print("Requeuing ", self.args)
        empty_trainer = type(self)(self.args)
        return submitit.helpers.DelayedSubmission(empty_trainer)

    def _setup_gpu_args(self):
        import submitit
        from pathlib import Path

        job_env = submitit.JobEnvironment()
        self.args.output_dir = Path(str(self.args.output_dir).replace("%j", str(job_env.job_id)))
        self.args.gpu = job_env.local_rank
        self.args.rank = job_env.global_rank
        self.args.world_size = job_env.num_tasks
        print(f"Process group: {job_env.num_tasks} tasks, rank: {job_env.global_rank}")


def main():
    args = parse_args()
    log_folder = args.model if args.suffix == '' else args.model + "-" + args.suffix
    if args.job_dir == "":
        args.job_dir = get_shared_folder() / Path(log_folder)
    else:
        if not os.path.exists(args.job_dir):
            os.makedirs(args.job_dir)
        args.job_dir = Path(os.path.join(args.job_dir, log_folder))

    # Note that the folder will depend on the job_id, to easily track experiments
    executor = submitit.AutoExecutor(folder=args.job_dir, slurm_max_num_timeout=30)

    num_gpus_per_node = args.ngpus
    nodes = args.nodes
    timeout_min = args.timeout

    partition = args.partition
    executor.update_parameters(
        mem_gb=40 * num_gpus_per_node,
        gpus_per_node=num_gpus_per_node,
        tasks_per_node=num_gpus_per_node,  # one task per GPU
        cpus_per_task=10,
        nodes=nodes,
        timeout_min=timeout_min,  # max is 60 * 72
        # Below are cluster dependent parameters
        slurm_partition=partition,
        slurm_signal_delay_s=120,
        slurm_gres=f'gpu:{args.ngpus}'
    )

    job_name = log_folder
    executor.update_parameters(name=job_name)

    args.dist_url = get_init_file().as_uri()
    args.output_dir = args.job_dir

    trainer = Trainer(args)
    job = executor.submit(trainer)

    print("Submitted job_id:", job.job_id, " name: ", job_name)


if __name__ == "__main__":
    main()