-
Notifications
You must be signed in to change notification settings - Fork 2.6k
/
mask2former_swin-s_8xb2-90k_cityscapes-512x1024.py
37 lines (35 loc) · 1.44 KB
/
mask2former_swin-s_8xb2-90k_cityscapes-512x1024.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
_base_ = ['./mask2former_swin-t_8xb2-90k_cityscapes-512x1024.py']
pretrained = 'https://download.openmmlab.com/mmsegmentation/v0.5/pretrain/swin/swin_small_patch4_window7_224_20220317-7ba6d6dd.pth' # noqa
depths = [2, 2, 18, 2]
model = dict(
backbone=dict(
depths=depths, init_cfg=dict(type='Pretrained',
checkpoint=pretrained)))
# set all layers in backbone to lr_mult=0.1
# set all norm layers, position_embeding,
# query_embeding, level_embeding to decay_multi=0.0
backbone_norm_multi = dict(lr_mult=0.1, decay_mult=0.0)
backbone_embed_multi = dict(lr_mult=0.1, decay_mult=0.0)
embed_multi = dict(lr_mult=1.0, decay_mult=0.0)
custom_keys = {
'backbone': dict(lr_mult=0.1, decay_mult=1.0),
'backbone.patch_embed.norm': backbone_norm_multi,
'backbone.norm': backbone_norm_multi,
'absolute_pos_embed': backbone_embed_multi,
'relative_position_bias_table': backbone_embed_multi,
'query_embed': embed_multi,
'query_feat': embed_multi,
'level_embed': embed_multi
}
custom_keys.update({
f'backbone.stages.{stage_id}.blocks.{block_id}.norm': backbone_norm_multi
for stage_id, num_blocks in enumerate(depths)
for block_id in range(num_blocks)
})
custom_keys.update({
f'backbone.stages.{stage_id}.downsample.norm': backbone_norm_multi
for stage_id in range(len(depths) - 1)
})
# optimizer
optim_wrapper = dict(
paramwise_cfg=dict(custom_keys=custom_keys, norm_decay_mult=0.0))