-
Notifications
You must be signed in to change notification settings - Fork 5.5k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
【PIR Dist Op Reg No.11】 reg dgc_momentum #63013
Conversation
你的PR提交成功,感谢你对开源项目的贡献! |
kernel : | ||
func : dgc_momentum | ||
data_type : param | ||
backward : divide_grad |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
backward : divide_grad |
是否忘记删除了
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
多谢,已修改
paddle/phi/api/yaml/op_compat.yaml
Outdated
@@ -818,6 +818,12 @@ | |||
outputs : | |||
out : Out | |||
|
|||
- op : dgc_momentum | |||
inputs : | |||
{param : Param, grad : Grad, velocity : Velocity, learning_rate : LearningRate, master_param : MasterParam, float mu, bool use_nesterov, str regularization_method, float regularization_coeff, bool multi_precision, float rescale_grad, float rampup_begin_step} |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
{param : Param, grad : Grad, velocity : Velocity, learning_rate : LearningRate, master_param : MasterParam, float mu, bool use_nesterov, str regularization_method, float regularization_coeff, bool multi_precision, float rescale_grad, float rampup_begin_step} | |
{param : Param, grad : Grad, velocity : Velocity, learning_rate : LearningRate, master_param : MasterParam} |
op_compat里面只声明必要的映射就可以了,后面的没必要复制粘贴
args : (Tensor param, Tensor grad, Tensor velocity, Tensor learning_rate, Tensor master_param, Tensor current_step, Tensor nranks, float mu, bool use_nesterov=false, str regularization_method="", float regularization_coeff=0.0f, bool multi_precision=false, float rescale_grad=1.0f, float rampup_begin_step=-1.0f) | ||
output : Tensor(param_out), Tensor(velocity_out), Tensor(master_param_out), Tensor(grad_out) | ||
kernel : | ||
func : dgc_momentum |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
func : dgc_momentum | |
func : dgc_momentum | |
param: [和kernel一致的参数顺序] |
args的参数顺序是可以调整的,有时也是必要的。但是这个时候 kernel和infermeta中就需要指定一个和函数调用所需参数一致的顺序,可以参考 #61293 ,也遇到了类似问题。
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
多谢,但我这里好像是参数后面写错了,后面顺序是一致的
9c11076
to
d7923a9
Compare
@kangguangli 这里ci都过了,麻烦再review一下~ |
Sorry to inform you that d7923a9's CIs have passed for more than 7 days. To prevent PR conflicts, you need to re-run all CIs manually. |
d7923a9
to
fbea24e
Compare
* feat(pir): reg dgc_momentum * feat(pir): reg dgc_momentum * feat(pir): reg dgc_momentum
* feat(pir): reg dgc_momentum * feat(pir): reg dgc_momentum * feat(pir): reg dgc_momentum
PR Category
Execute Infrastructure
PR Types
Devs
Description
#60436
注册算子dgc_momentum