forked from open-mmlab/mmpretrain
-
Notifications
You must be signed in to change notification settings - Fork 0
/
mae_vit-base-p16_8xb512-amp-coslr-300e_in1k.py
56 lines (50 loc) · 1.41 KB
/
mae_vit-base-p16_8xb512-amp-coslr-300e_in1k.py
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
_base_ = [
'../_base_/models/mae_vit-base-p16.py',
'../_base_/datasets/imagenet_bs512_mae.py',
'../_base_/default_runtime.py',
]
# optimizer wrapper
optim_wrapper = dict(
type='AmpOptimWrapper',
loss_scale='dynamic',
optimizer=dict(
type='AdamW',
lr=1.5e-4 * 4096 / 256,
betas=(0.9, 0.95),
weight_decay=0.05),
paramwise_cfg=dict(
custom_keys={
'ln': dict(decay_mult=0.0),
'bias': dict(decay_mult=0.0),
'pos_embed': dict(decay_mult=0.),
'mask_token': dict(decay_mult=0.),
'cls_token': dict(decay_mult=0.)
}))
# learning rate scheduler
param_scheduler = [
dict(
type='LinearLR',
start_factor=0.0001,
by_epoch=True,
begin=0,
end=40,
convert_to_iter_based=True),
dict(
type='CosineAnnealingLR',
T_max=260,
by_epoch=True,
begin=40,
end=300,
convert_to_iter_based=True)
]
# runtime settings
train_cfg = dict(type='EpochBasedTrainLoop', max_epochs=300)
default_hooks = dict(
# only keeps the latest 3 checkpoints
checkpoint=dict(type='CheckpointHook', interval=1, max_keep_ckpts=3))
randomness = dict(seed=0, diff_rank_seed=True)
# auto resume
resume = True
# NOTE: `auto_scale_lr` is for automatically scaling LR
# based on the actual training batch size.
auto_scale_lr = dict(base_batch_size=4096)