memory optimization for dynamic RNN #8041
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
QiJune
requested review from
reyoung,
wangkuiyi,
dzhwinter and
tonyyang-svail
dzhwinter
previously approved these changes
Feb 1, 2018
| @@ -205,6 +208,8 @@ class WhileGradOp : public framework::OperatorBase { | |||
| sum_op->Run(cur_scope, dev_place); | |||
| cur_scope.Rename(new_inside_name, inside_grad_name); | |||
| } | |||
| dev_ctx.Wait(); | |||
| const_cast<framework::Scope &>(scope).DeleteScope(&cur_scope); | |||
There was a problem hiding this comment.
Nice catch!
This fix will solve the headache OOM.
paddle/framework/scope.h
Outdated
| @@ -65,6 +65,8 @@ class Scope { | |||
| /// Drop all kids scopes belonged to this scope. | |||
| void DropKids(); | |||
|
|
|||
| void EraseVars(std::vector<std::string>& var_names); | |||
There was a problem hiding this comment.
Since our every variable will be released after we destruct the scope, In my humble view, this interface should never be used.
| @@ -205,6 +208,8 @@ class WhileGradOp : public framework::OperatorBase { | |||
| sum_op->Run(cur_scope, dev_place); | |||
| cur_scope.Rename(new_inside_name, inside_grad_name); | |||
| } | |||
| dev_ctx.Wait(); | |||
There was a problem hiding this comment.
what is the cost of adding this dev_ctx.Wait();?
There was a problem hiding this comment.
I have not make detailed test yet. But if we do not delete the step scope, we can not training a larger RNN model because of OutOfMemory.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Have tested in machine translation demo.
In the first batch training, memory reduced from 5728446208 to 1115982080, saves 80.5% memory.
It seems that delete_var operator has no contribution for the result. I will remove it.