train.py

#!/usr/bin/env python3
# This file is covered by the LICENSE file in the root of this project.

import argparse
import subprocess
import datetime
import yaml
from shutil import copyfile
import sys
import os
import os.path as osp
import shutil
from loguru import logger 

from lib.trainer import Trainer

# loguru: https://github.com/Delgan/loguru
def set_logger(log_path):
    logger.add(log_path)
    #logger.add(sys.stdout, colorize=True, format="{message}")
    logger.add(sys.stderr, format="{time} {level} {message}", level="WARNING")

    return logger

if __name__ == '__main__':
  parser = argparse.ArgumentParser("./train.py")
  parser.add_argument(
      '--dataset', '-d',
      type=str,
      required=True,
      help='Dataset to train with. No Default',
  )
  parser.add_argument(
      '--arch_cfg', '-ac',
      type=str,
      required=True,
      help='Architecture yaml cfg file. See /config/arch for sample. No default!',
  )
  parser.add_argument(
      '--data_cfg', '-dc',
      type=str,
      required=False,
      default='config/labels/semantic-kitti.yaml',
      help='Classification yaml cfg file. See /config/labels for sample. No default!',
  )
  parser.add_argument(
      '--log', '-l',
      type=str,
      required=True,
      help='Directory to put the log data.'
  )
  parser.add_argument(
      '--pretrained', '-p',
      type=str,
      required=False,
      default=None,
      help='Directory to get the pretrained model. If not passed, do from scratch!'
  )
  FLAGS, unparsed = parser.parse_known_args()

  # create log folder
  try:
    if os.path.isdir(FLAGS.log):
      shutil.rmtree(FLAGS.log)
    os.makedirs(FLAGS.log)
  except Exception as e:
    print(e)
    print("Error creating log directory. Check permissions!")
    quit()

  logger = set_logger(osp.join(FLAGS.log, 'log.txt'))

  # print summary of what we will do
  logger.info("----------")
  logger.info("INTERFACE:")
  logger.info("dataset", FLAGS.dataset)
  logger.info("arch_cfg", FLAGS.arch_cfg)
  logger.info("data_cfg", FLAGS.data_cfg)
  logger.info("log", FLAGS.log)
  logger.info("pretrained", FLAGS.pretrained)
  logger.info("----------\n")

  # open arch config file
  try:
    logger.info("Opening arch config file %s" % FLAGS.arch_cfg)
    ARCH = yaml.safe_load(open(FLAGS.arch_cfg, 'r'))
  except Exception as e:
    logger.warning(e)
    logger.warning("Error opening arch yaml file.")
    quit()

  # open data config file
  try:
    logger.info("Opening data config file %s" % FLAGS.data_cfg)
    DATA = yaml.safe_load(open(FLAGS.data_cfg, 'r'))
  except Exception as e:
    logger.warning(e)
    logger.warning("Error opening data yaml file.")
    quit()

  # does model folder exist?
  if FLAGS.pretrained is not None:
    if os.path.isdir(FLAGS.pretrained):
      logger.info("model folder exists! Using model from %s" % (FLAGS.pretrained))
    else:
      logger.warning("model folder doesnt exist! Start with random weights...")
  else:
    logger.warning("No pretrained directory found.")

  # copy all files to log folder (to remember what we did, and make inference
  # easier). Also, standardize name to be able to open it later
  try:
    logger.info("Copying files to %s for further reference." % FLAGS.log)
    copyfile(FLAGS.arch_cfg, FLAGS.log + "/arch_cfg.yaml")
    copyfile(FLAGS.data_cfg, FLAGS.log + "/data_cfg.yaml")
  except Exception as e:
    logger.warning(e)
    logger.warning("Error copying files, check permissions. Exiting...")
    quit()

  # create trainer and start the training
  trainer = Trainer(ARCH, DATA, FLAGS.dataset, FLAGS.log, logger, FLAGS.pretrained, ARCH['model']['use_mps'])
  trainer.train()