Skip to content

Commit

Permalink
update doc
Browse files Browse the repository at this point in the history
  • Loading branch information
RangiLyu committed Dec 15, 2022
1 parent c0e6c49 commit 47c4bfa
Show file tree
Hide file tree
Showing 2 changed files with 9 additions and 5 deletions.
12 changes: 7 additions & 5 deletions docs/zh_cn/tutorials/optim_wrapper.md
Original file line number Diff line number Diff line change
Expand Up @@ -329,17 +329,19 @@ optimizer = build_optim_wrapper(ToyModel(), optim_wrapper)

`decay_mult`:所有参数的衰减系数

`bias_lr_mult`:偏置的学习率系数(不包括正则化层的偏置以及可变形卷积的 offset),默认值为 1
`bias_lr_mult`:偏置的学习率系数(不包括正则化层的偏置以及可变形卷积的 offset)

`bias_decay_mult`:偏置的权值衰减系数(不包括正则化层的偏置以及可变形卷积的 offset),默认值为 1
`bias_decay_mult`:偏置的权值衰减系数(不包括正则化层的偏置以及可变形卷积的 offset)

`norm_decay_mult`:正则化层权重和偏置的权值衰减系数,默认值为 1
`norm_decay_mult`:正则化层权重和偏置的权值衰减系数

`dwconv_decay_mult`:Depth-wise 卷积的权值衰减系数,默认值为 1
`flat_decay_mult`:一维参数的权值衰减系数

`dwconv_decay_mult`:Depth-wise 卷积的权值衰减系数

`bypass_duplicate`:是否跳过重复的参数,默认为 `False`

`dcn_offset_lr_mult`:可变形卷积(Deformable Convolution)的学习率系数,默认值为 1
`dcn_offset_lr_mult`:可变形卷积(Deformable Convolution)的学习率系数

#### 为模型不同部分的参数设置不同的超参系数

Expand Down
2 changes: 2 additions & 0 deletions mmengine/optim/optimizer/default_constructor.py
Original file line number Diff line number Diff line change
Expand Up @@ -42,6 +42,8 @@ class DefaultOptimWrapperConstructor:
- ``norm_decay_mult`` (float): It will be multiplied to the weight
decay for all weight and bias parameters of normalization
layers.
- ``flat_decay_mult`` (float): It will be multiplied to the weight
decay for all one-dimensional parameters
- ``dwconv_decay_mult`` (float): It will be multiplied to the weight
decay for all weight and bias parameters of depthwise conv
layers.
Expand Down

0 comments on commit 47c4bfa

Please sign in to comment.