This repository has been archived by the owner on Aug 7, 2024. It is now read-only.
[4/x] add tests for DTensor TP/SP + Float8Linear #294
Closed
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
Summary: Makes the DTensor TP/SP tests also test `Float8Linear` with all scaling types configured to be dynamic. We can add support for delayed scaling with float8 all-gather for `x` and `dL_dY` in a future PR, as needed. Test Plan: ``` ./test/test_dtensor.sh ``` Reviewers: Subscribers: Tasks: Tags: [ghstack-poisoned]
This was referenced Jul 1, 2024
vkuzo
added a commit
that referenced
this pull request
Jul 1, 2024
Summary: Makes the DTensor TP/SP tests also test `Float8Linear` with all scaling types configured to be dynamic. We can add support for delayed scaling with float8 all-gather for `x` and `dL_dY` in a future PR, as needed. Test Plan: ``` ./test/test_dtensor.sh ``` Reviewers: Subscribers: Tasks: Tags: ghstack-source-id: 8e121ecaf4e05c5eb69b1612c084c459abe95589 Pull Request resolved: #294
facebook-github-bot
added
the
CLA Signed
drisspg
reviewed
Jul 1, 2024
| ): | ||
| device = mesh.device_type | ||
| # TODO(future): delete Float8DynamicLinear from this test once all the |
There was a problem hiding this comment.
Nit: I wonder if we want to standardize on the todo format so that in the future we can just command F and find all the things to change
drisspg
reviewed
Jul 1, 2024
| @@ -22,6 +23,15 @@ | |||
| # NOTE: This only works and tested with the DynamicLinear | |||
There was a problem hiding this comment.
nit: maybe update this comment
Summary: Makes the DTensor TP/SP tests also test `Float8Linear` with all scaling types configured to be dynamic. We can add support for delayed scaling with float8 all-gather for `x` and `dL_dY` in a future PR, as needed. Test Plan: ``` ./test/test_dtensor.sh ``` Reviewers: Subscribers: Tasks: Tags: [ghstack-poisoned]
vkuzo
added a commit
that referenced
this pull request
Jul 1, 2024
Summary: Makes the DTensor TP/SP tests also test `Float8Linear` with all scaling types configured to be dynamic. We can add support for delayed scaling with float8 all-gather for `x` and `dL_dY` in a future PR, as needed. Test Plan: ``` ./test/test_dtensor.sh ``` Reviewers: Subscribers: Tasks: Tags: ghstack-source-id: 3048ab79fa1ea5cee77174ef0c6176e9bf29d850 Pull Request resolved: #294
Summary: Makes the DTensor TP/SP tests also test `Float8Linear` with all scaling types configured to be dynamic. We can add support for delayed scaling with float8 all-gather for `x` and `dL_dY` in a future PR, as needed. Test Plan: ``` ./test/test_dtensor.sh ``` Reviewers: Subscribers: Tasks: Tags: [ghstack-poisoned]
This was referenced Jul 2, 2024
|
@vkuzo has imported this pull request. If you are a Meta employee, you can view this diff on Phabricator. |
|
This pull request has been merged in 3ec9665. |
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.
Stack from ghstack (oldest at bottom):
Summary:
Makes the DTensor TP/SP tests also test
Float8Linearwith all scalingtypes configured to be dynamic.
We can add support for delayed scaling with float8 all-gather for
xand
dL_dYin a future PR, as needed.Test Plan:
Reviewers:
Subscribers:
Tasks:
Tags:
Differential Revision: D59305797