rm_pt.py 4.66 KB
Newer Older
Myle Ott's avatar
Myle Ott committed
1
#!/usr/bin/env python3
2
# Copyright (c) Facebook, Inc. and its affiliates.
3
#
4
5
# This source code is licensed under the MIT license found in the
# LICENSE file in the root directory of this source tree.
6
7
8
9
10
11
12
13
14
15
16
17
18

import argparse
import os
import re
import shutil
import sys


pt_regexp = re.compile(r'checkpoint(\d+|_\d+_\d+|_[a-z]+)\.pt')
pt_regexp_epoch_based = re.compile(r'checkpoint(\d+)\.pt')
pt_regexp_update_based = re.compile(r'checkpoint_\d+_(\d+)\.pt')


Myle Ott's avatar
Myle Ott committed
19
def parse_checkpoints(files):
20
21
22
23
24
25
26
27
28
    entries = []
    for f in files:
        m = pt_regexp_epoch_based.fullmatch(f)
        if m is not None:
            entries.append((int(m.group(1)), m.group(0)))
        else:
            m = pt_regexp_update_based.fullmatch(f)
            if m is not None:
                entries.append((int(m.group(1)), m.group(0)))
Myle Ott's avatar
Myle Ott committed
29
30
31
32
33
    return entries


def last_n_checkpoints(files, n):
    entries = parse_checkpoints(files)
34
35
36
37
    return [x[1] for x in sorted(entries, reverse=True)[:n]]


def every_n_checkpoints(files, n):
Myle Ott's avatar
Myle Ott committed
38
39
    entries = parse_checkpoints(files)
    return [x[1] for x in sorted(sorted(entries)[::-n])]
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64


def main():
    parser = argparse.ArgumentParser(
        description=(
            'Recursively delete checkpoint files from `root_dir`, '
            'but preserve checkpoint_best.pt and checkpoint_last.pt'
        )
    )
    parser.add_argument('root_dirs', nargs='*')
    parser.add_argument('--save-last', type=int, default=0, help='number of last checkpoints to save')
    parser.add_argument('--save-every', type=int, default=0, help='interval of checkpoints to save')
    parser.add_argument('--preserve-test', action='store_true',
                        help='preserve checkpoints in dirs that start with test_ prefix (default: delete them)')
    parser.add_argument('--delete-best', action='store_true', help='delete checkpoint_best.pt')
    parser.add_argument('--delete-last', action='store_true', help='delete checkpoint_last.pt')
    parser.add_argument('--no-dereference', action='store_true', help='don\'t dereference symlinks')
    args = parser.parse_args()

    files_to_desymlink = []
    files_to_preserve = []
    files_to_delete = []
    for root_dir in args.root_dirs:
        for root, _subdirs, files in os.walk(root_dir):
            if args.save_last > 0:
Myle Ott's avatar
Myle Ott committed
65
                to_save = last_n_checkpoints(files, args.save_last)
66
            else:
Myle Ott's avatar
Myle Ott committed
67
                to_save = []
68
            if args.save_every > 0:
Myle Ott's avatar
Myle Ott committed
69
                to_save += every_n_checkpoints(files, args.save_every)
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
            for file in files:
                if not pt_regexp.fullmatch(file):
                    continue
                full_path = os.path.join(root, file)
                if (
                    (
                        not os.path.basename(root).startswith('test_')
                        or args.preserve_test
                    )
                    and (
                        (file == 'checkpoint_last.pt' and not args.delete_last)
                        or (file == 'checkpoint_best.pt' and not args.delete_best)
                        or file in to_save
                    )
                ):
                    if os.path.islink(full_path) and not args.no_dereference:
                        files_to_desymlink.append(full_path)
                    else:
                        files_to_preserve.append(full_path)
                else:
                    files_to_delete.append(full_path)

    if len(files_to_desymlink) == 0 and len(files_to_delete) == 0:
        print('Nothing to do.')
        sys.exit(0)

Myle Ott's avatar
Myle Ott committed
96
97
98
99
    files_to_desymlink = sorted(files_to_desymlink)
    files_to_preserve = sorted(files_to_preserve)
    files_to_delete = sorted(files_to_delete)

100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
    print('Operations to perform (in order):')
    if len(files_to_desymlink) > 0:
        for file in files_to_desymlink:
            print(' - preserve (and dereference symlink): ' + file)
    if len(files_to_preserve) > 0:
        for file in files_to_preserve:
            print(' - preserve: ' + file)
    if len(files_to_delete) > 0:
        for file in files_to_delete:
            print(' - delete: ' + file)
    while True:
        resp = input('Continue? (Y/N): ')
        if resp.strip().lower() == 'y':
            break
        elif resp.strip().lower() == 'n':
            sys.exit(0)

    print('Executing...')
    if len(files_to_desymlink) > 0:
        for file in files_to_desymlink:
            realpath = os.path.realpath(file)
            print('rm ' + file)
            os.remove(file)
            print('cp {} {}'.format(realpath, file))
            shutil.copyfile(realpath, file)
    if len(files_to_delete) > 0:
        for file in files_to_delete:
            print('rm ' + file)
            os.remove(file)


if __name__ == '__main__':
    main()