File size: 7,121 Bytes
b9425fd
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
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
65
66
67
68
69
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
96
97
98
99
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
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
# --------------------------------------------------------
# Reversible Column Networks
# Copyright (c) 2022 Megvii Inc.
# Licensed under TheApache License 2.0 [see LICENSE for details]
# Written by Yuxuan Cai
# --------------------------------------------------------
import os
import yaml
from yacs.config import CfgNode as CN

_C = CN()

# Base config files
_C.BASE = ['']

# -----------------------------------------------------------------------------
# Data settings
# -----------------------------------------------------------------------------
_C.DATA = CN()
# Batch size for a single GPU, could be overwritten by command line argument
_C.DATA.BATCH_SIZE = 128
# Path to dataset, could be overwritten by command line argument
_C.DATA.DATA_PATH = 'path/to/imagenet'
# Dataset name
_C.DATA.DATASET = 'imagenet'
# Input image size
_C.DATA.IMG_SIZE = 224
# Interpolation to resize image (random, bilinear, bicubic)
_C.DATA.INTERPOLATION = 'bicubic'
# Pin CPU memory in DataLoader for more efficient (sometimes) transfer to GPU.
_C.DATA.PIN_MEMORY = True
# Number of data loading threads
_C.DATA.NUM_WORKERS = 8
# Path to evaluation dataset for ImageNet 22k
_C.DATA.EVAL_DATA_PATH = 'path/to/eval/data'

# -----------------------------------------------------------------------------
# Model settings
# -----------------------------------------------------------------------------
_C.MODEL = CN()
# Model type
_C.MODEL.TYPE = ''
# Model name
_C.MODEL.NAME = ''
# Checkpoint to resume, could be overwritten by command line argument
_C.MODEL.RESUME = ''
# Checkpoint to finetune, could be overwritten by command line argument
_C.MODEL.FINETUNE = ''
# Number of classes, overwritten in data preparation
_C.MODEL.NUM_CLASSES = 1000
# Label Smoothing
_C.MODEL.LABEL_SMOOTHING = 0.0

# -----------------------------------------------------------------------------
# Specific Model settings
# -----------------------------------------------------------------------------

_C.REVCOL = CN()

_C.REVCOL.INTER_SUPV = True

_C.REVCOL.SAVEMM = True

_C.REVCOL.FCOE = 4.0

_C.REVCOL.CCOE = 0.8

_C.REVCOL.KERNEL_SIZE = 3

_C.REVCOL.DROP_PATH = 0.1

_C.REVCOL.HEAD_INIT_SCALE = None

# -----------------------------------------------------------------------------
# Training settings
# -----------------------------------------------------------------------------
_C.TRAIN = CN()
_C.TRAIN.START_EPOCH = 0
_C.TRAIN.EPOCHS = 300
_C.TRAIN.WARMUP_EPOCHS = 5
_C.TRAIN.WEIGHT_DECAY = 4e-5
_C.TRAIN.BASE_LR = 0.4

_C.TRAIN.WARMUP_LR = 0.05
_C.TRAIN.MIN_LR = 1e-5
# Clip gradient norm
_C.TRAIN.CLIP_GRAD = 10.0
# Auto resume from latest checkpoint
_C.TRAIN.AUTO_RESUME = True
# Check point
_C.TRAIN.USE_CHECKPOINT = False

_C.TRAIN.AMP = True

# LR scheduler
_C.TRAIN.LR_SCHEDULER = CN()
# LR scheduler
_C.TRAIN.LR_SCHEDULER.NAME = 'cosine'
# Epoch interval to decay LR, used in StepLRScheduler
_C.TRAIN.LR_SCHEDULER.DECAY_EPOCHS = 30
# LR decay rate, used in StepLRScheduler
_C.TRAIN.LR_SCHEDULER.DECAY_RATE = 0.1

# Optimizer
_C.TRAIN.OPTIMIZER = CN()
_C.TRAIN.OPTIMIZER.NAME = 'sgd'
# Optimizer Epsilon fow adamw
_C.TRAIN.OPTIMIZER.EPS = 1e-8
# Optimizer Betas fow adamw
_C.TRAIN.OPTIMIZER.BETAS = (0.9, 0.999)
# SGD momentum
_C.TRAIN.OPTIMIZER.MOMENTUM = 0.9
# Layer Decay
_C.TRAIN.OPTIMIZER.LAYER_DECAY = 1.0

# -----------------------------------------------------------------------------
# Augmentation settings
# -----------------------------------------------------------------------------
_C.AUG = CN()
# Color jitter factor
_C.AUG.COLOR_JITTER = 0.4
# Use AutoAugment policy. "v0" or "original"
_C.AUG.AUTO_AUGMENT = 'rand-m9-mstd0.5-inc1'
# Random erase prob
_C.AUG.REPROB = 0.25
# Random erase mode
_C.AUG.REMODE = 'pixel'
# Random erase count
_C.AUG.RECOUNT = 1
# Mixup alpha, mixup enabled if > 0
_C.AUG.MIXUP = 0.8
# Cutmix alpha, cutmix enabled if > 0
_C.AUG.CUTMIX = 1.0
# Cutmix min/max ratio, overrides alpha and enables cutmix if set
_C.AUG.CUTMIX_MINMAX = None
# Probability of performing mixup or cutmix when either/both is enabled
_C.AUG.MIXUP_PROB = 1.0
# Probability of switching to cutmix when both mixup and cutmix enabled
_C.AUG.MIXUP_SWITCH_PROB = 0.5
# How to apply mixup/cutmix params. Per "batch", "pair", or "elem"
_C.AUG.MIXUP_MODE = 'batch'

# -----------------------------------------------------------------------------
# Testing settings
# -----------------------------------------------------------------------------
_C.TEST = CN()
# Whether to use center crop when testing
_C.TEST.CROP = True

# -----------------------------------------------------------------------------
# Misc
# -----------------------------------------------------------------------------
# Path to output folder, overwritten by command line argument
_C.OUTPUT = 'outputs/'
# Tag of experiment, overwritten by command line argument
_C.TAG = 'default'
# Frequency to save checkpoint
_C.SAVE_FREQ = 1
# Frequency to logging info
_C.PRINT_FREQ = 100
# Fixed random seed
_C.SEED = 0
# Perform evaluation only, overwritten by command line argument
_C.EVAL_MODE = False
# Test throughput only, overwritten by command line argument
_C.THROUGHPUT_MODE = False
# local rank for DistributedDataParallel, given by command line argument
_C.LOCAL_RANK = 0



# EMA
_C.MODEL_EMA = False
_C.MODEL_EMA_DECAY = 0.9999




# Machine
_C.MACHINE = CN()
_C.MACHINE.MACHINE_WORLD_SIZE = None
_C.MACHINE.MACHINE_RANK = None

def _update_config_from_file(config, cfg_file):
    config.defrost()
    with open(cfg_file, 'r') as f:
        yaml_cfg = yaml.load(f, Loader=yaml.FullLoader)

    for cfg in yaml_cfg.setdefault('BASE', ['']):
        if cfg:
            _update_config_from_file(
                config, os.path.join(os.path.dirname(cfg_file), cfg)
            )
    print('=> merge config from {}'.format(cfg_file))
    config.merge_from_file(cfg_file)
    config.freeze()


def update_config(config, args):
    _update_config_from_file(config, args.cfg)

    config.defrost()
    if args.opts:
        config.merge_from_list(args.opts)

    # merge from specific arguments
    if args.batch_size:
        config.DATA.BATCH_SIZE = args.batch_size
    if args.data_path:
        config.DATA.DATA_PATH = args.data_path
    if args.resume:
        config.MODEL.RESUME = args.resume
    if args.finetune:
        config.MODEL.FINETUNE = args.finetune
    if args.use_checkpoint:
        config.TRAIN.USE_CHECKPOINT = True
    if args.output:
        config.OUTPUT = args.output
    if args.tag:
        config.TAG = args.tag
    if args.eval:
        config.EVAL_MODE = True
    if args.model_ema:
        config.MODEL_EMA = True

    config.dist_url = args.dist_url
    # set local rank for distributed training
    config.LOCAL_RANK = args.local_rank

    # output folder
    config.OUTPUT = os.path.join(config.OUTPUT, config.MODEL.NAME, config.TAG)

    config.freeze()


def get_config(args):
    """Get a yacs CfgNode object with default values."""
    # Return a clone so that the defaults will not be altered
    # This is for the "local variable" use pattern
    config = _C.clone()
    update_config(config, args)

    return config