-
Notifications
You must be signed in to change notification settings - Fork 5.9k
Initialization design doc #4852
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
Closed
Closed
Changes from 4 commits
Commits
Show all changes
6 commits
Select commit
Hold shift + click to select a range
d76a4ae
Create initialization.md
dc7a650
Update initialization.md
wangkuiyi d65fc83
Update initialization.md
wangkuiyi e6b3b31
formatting
50ce87c
Merge remote-tracking branch 'upstream/develop' into init-design
b1dd9c6
update doc conclusion
File filter
Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
There are no files selected for viewing
This file contains hidden or 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
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,48 @@ | ||
| # Initialization of Parameters | ||
|
|
||
| ## Motivation | ||
|
|
||
| PaddlePaddle needs a way to init parameters. | ||
|
|
||
| ## Challenge | ||
|
|
||
| Initialization operators must run once and only one; otherwise, every iteration would clear the parameters. | ||
|
|
||
| ## Proposals | ||
|
|
||
| ### Proposal 1: Two seperate `ProgramDesc`. | ||
|
|
||
| The initialization part of the program in a `ProgramDesc` message, and the rest part in another. | ||
|
|
||
| - Pros: | ||
| - Simple and straight forward. | ||
| - Cons: | ||
| - Two seperate `ProgramDesc` doesn't match one-program-desc design. | ||
|
|
||
| ### Proposal 2: Add `run_once` attribute for initialization related operators. | ||
|
|
||
| PR link: https://github.com/PaddlePaddle/Paddle/pull/4802/files | ||
|
|
||
| - Pros: | ||
| - Simple and straight forward | ||
| - Cons: | ||
| - Not compatible with current executor design. Because executor creates Op instances on the fly at `Executor::Run()`. | ||
|
|
||
| ### Proposal 3: Remove Init operator during Prune | ||
|
|
||
| Use regular expression to filter out init operators. | ||
|
|
||
| API at prune.cc | ||
| ```c++ | ||
| Prune(const ProgramDesc& input, ProgramDesc* output, const PruneInfo& prune_msg); | ||
| ``` | ||
|
|
||
| Add protobuf message | ||
| ```protobuf | ||
| message PruneInfo { | ||
| repeated string re = 3; // regular expression, e.g. "init_*". | ||
| } | ||
| ``` | ||
|
|
||
| - Pros: | ||
| - Extendible | ||
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.
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.
this seems simple and workable.
The resource which survived across multiple block/program is the
VariablesinScopes.A program is the container of multiple Blocks, more than one program seems not a problem.