[refactor] Move checkpoint saving into trainer #4034
Merged
Conversation
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
added 7 commits
May 27, 2020 20:11
|
@andrewcoh will this break the saving in the ghost trainer? |
andrewcoh
approved these changes
Jun 3, 2020
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
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Outdated suggestions cannot be applied.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.
Proposed change(s)
This PR removes the
--save-freqCLI option and replaces it with acheckpoint_intervaloption in the YAML. This option is now specified in trainer steps and not global steps, and can be different per-trainer. It also slightly refactors the summary writing logic to decrease repeated code.Types of change(s)
Checklist
Other comments