Add imbalance stats for total perf, hbm and ddr #2040
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
facebook-github-bot
added
the
CLA Signed
|
This pull request was exported from Phabricator. Differential Revision: D57465383 |
sarckk
added a commit
to sarckk/torchrec
that referenced
this pull request
May 30, 2024
Summary:
Add and log 4 ways to measure imbalance of generated sharding plan.
Suppose we have $k$ gpus. Let $s$ be a vector where the $i^{th}$ component is the total size currently allocated to the $i^{th}$ device.
First we normalize vector $s$ such that sum of elements in $p$ equals 1. We can view this as a probability distribution, and to get a measure of imbalance, we can measure its deviation from uniform distribution $p$ using one of the following ways:
- total variations
- total distance
- chi divergence
- KL divergence
Reviewed By: henrylhtsang
Differential Revision: D57465383
Summary:
Add and log 4 ways to measure imbalance of generated sharding plan.
Suppose we have $k$ gpus. Let $s$ be a vector where the $i^{th}$ component is the total size currently allocated to the $i^{th}$ device.
First we normalize vector $s$ such that sum of elements in $p$ equals 1. We can view this as a probability distribution, and to get a measure of imbalance, we can measure its deviation from uniform distribution $p$ using one of the following ways:
- total variations
- total distance
- chi divergence
- KL divergence
Reviewed By: henrylhtsang
Differential Revision: D57465383
|
This pull request was exported from Phabricator. Differential Revision: D57465383 |
|
This pull request was exported from Phabricator. Differential Revision: D57465383 |
PaulZhang12
pushed a commit
that referenced
this pull request
Jun 5, 2024
Summary: Pull Request resolved: #2040 Add and log 4 ways to measure imbalance of generated sharding plan. Suppose we have $k$ gpus. Let $s$ be a vector where the $i^{th}$ component is the total size currently allocated to the $i^{th}$ device. First we normalize vector $s$ such that sum of elements in $p$ equals 1. We can view this as a probability distribution, and to get a measure of imbalance, we can measure its deviation from uniform distribution $p$ using one of the following ways: - total variations - total distance - chi divergence - KL divergence Reviewed By: henrylhtsang Differential Revision: D57465383 fbshipit-source-id: a22d59b17461dea46478b12beafe1b95c3331f62
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.
Summary:
Add and log 4 ways to measure imbalance of generated sharding plan.
Suppose we have$k$ gpus. Let $s$ be a vector where the $i^{th}$ component is the total size currently allocated to the $i^{th}$ device.
First we normalize vector$s$ such that sum of elements in $p$ equals 1. We can view this as a probability distribution, and to get a measure of imbalance, we can measure its deviation from uniform distribution $p$ using one of the following ways:
Differential Revision: D57465383