This repository has been archived by the owner on Nov 22, 2022. It is now read-only.
add Stage.OTHERS and allow TB to print to a seperate prefix not in (TRAIN, TEST, EVAL) #1258
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
…RAIN, TEST, EVAL) Summary: In some training algorithms, we may want to use Tensorboard to produce summary on some fine-grained training results. However, currently this is not possible because Stage num is limited to (Train, Test, Eval). Further, TensorBoard maps each Stage enum to a prefix and aggregate the metrics for that prefix. example scenarios where we need additional stage for metrics reporting: 1. federated learning: we need to distinguish the training curve on clients, and the training curve of the global model aggregated from multiple users but evaluated on the client's training data. 2. adverserial training: we want to distinguish and report the training curve on clean data and adverserial data seperately. 3. any training algorithms that performs averaging during the training process, where we can see the difference before and after averaging. Reviewed By: geof90, snisarg Differential Revision: D19783880 fbshipit-source-id: ed24bf06366272dce236721e1bb381768fee5b13
facebook-github-bot
added
CLA Signed
|
This pull request was exported from Phabricator. Differential Revision: D19783880 |
|
This pull request has been merged in 1fdb7c7. |
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Outdated suggestions cannot be applied.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.
Summary:
In some training algorithms, we may want to use Tensorboard to produce summary on some fine-grained training results. However, currently this is not possible because Stage num is limited to (Train, Test, Eval). Further, TensorBoard maps each Stage enum to a prefix and aggregate the metrics for that prefix.
example scenarios where we need additional stage for metrics reporting:
federated learning: we need to distinguish the training curve on clients, and the training curve of the global model aggregated from multiple users but evaluated on the client's training data.
adverserial training: we want to distinguish and report the training curve on clean data and adverserial data seperately.
any training algorithms that performs averaging during the training process, where we can see the difference before and after averaging.
Reviewed By: geof90, snisarg
Differential Revision: D19783880