Skip to content
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.

Sign up for GitHub

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

Jump to bottom

Allow passing planner to _shard_modules #2732

Open
wants to merge 1 commit into
base: main
Choose a base branch
from
Open

Allow passing planner to _shard_modules #2732

wants to merge 1 commit into from

Conversation

che-sh
Copy link
Contributor

@che-sh che-sh commented Feb 7, 2025

Summary: _shard_modules function is used in fx_traceability tests for SDD and SemiSync pipeline. It uses a default ShardingPlanner and topology that use hardcoded batch size (512) and HBM memory limit (32Gb), respectively. This change allows specifying the ShardingPlanner and Topology to more accurately reflect the machine capabilities. The change is intentionally limited to _shard_modules only and not public shard_modules to avoid changing the contract for the latter.

Reviewed By: sarckk

Differential Revision: D69163227

@che-sh @facebook-github-bot
Allow passing planner to _shard_modules
d010abc 
Summary: `_shard_modules` function is used in fx_traceability tests for SDD and SemiSync pipeline. It uses a default ShardingPlanner and topology that use hardcoded batch size (512) and HBM memory limit (32Gb), respectively. This change allows specifying the ShardingPlanner and Topology to more accurately reflect the machine capabilities. The change is intentionally limited to `_shard_modules` only and not public `shard_modules` to avoid changing the contract for the latter.

Reviewed By: sarckk

Differential Revision: D69163227
@facebook-github-bot facebook-github-bot added the CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. label Feb 7, 2025
@facebook-github-bot
Copy link
Contributor

This pull request was exported from Phabricator. Differential Revision: D69163227

che-sh added a commit to che-sh/torchrec that referenced this pull request Feb 7, 2025
@che-sh @facebook-github-bot
Allow passing planner to _shard_modules (pytorch#2732)
52dbf67 
Summary:

`_shard_modules` function is used in fx_traceability tests for SDD and SemiSync pipeline. It uses a default ShardingPlanner and topology that use hardcoded batch size (512) and HBM memory limit (32Gb), respectively. This change allows specifying the ShardingPlanner and Topology to more accurately reflect the machine capabilities. The change is intentionally limited to `_shard_modules` only and not public `shard_modules` to avoid changing the contract for the latter.

Reviewed By: sarckk

Differential Revision: D69163227
che-sh added a commit to che-sh/torchrec that referenced this pull request Feb 11, 2025
@che-sh @facebook-github-bot
Allow passing planner to _shard_modules (pytorch#2732)
b1de666 
Summary:

`_shard_modules` function is used in fx_traceability tests for SDD and SemiSync pipeline. It uses a default ShardingPlanner and topology that use hardcoded batch size (512) and HBM memory limit (32Gb), respectively. This change allows specifying the ShardingPlanner and Topology to more accurately reflect the machine capabilities. The change is intentionally limited to `_shard_modules` only and not public `shard_modules` to avoid changing the contract for the latter.

Reviewed By: sarckk

Differential Revision: D69163227
che-sh added a commit to che-sh/torchrec that referenced this pull request Feb 11, 2025
@che-sh @facebook-github-bot
Allow passing planner to _shard_modules (pytorch#2732)
0d9729b 
Summary:

`_shard_modules` function is used in fx_traceability tests for SDD and SemiSync pipeline. It uses a default ShardingPlanner and topology that use hardcoded batch size (512) and HBM memory limit (32Gb), respectively. This change allows specifying the ShardingPlanner and Topology to more accurately reflect the machine capabilities. The change is intentionally limited to `_shard_modules` only and not public `shard_modules` to avoid changing the contract for the latter.

Reviewed By: sarckk

Differential Revision: D69163227
che-sh added a commit to che-sh/torchrec that referenced this pull request Feb 12, 2025
@che-sh @facebook-github-bot
Allow passing planner to _shard_modules (pytorch#2732)
f269be7 
Summary:

`_shard_modules` function is used in fx_traceability tests for SDD and SemiSync pipeline. It uses a default ShardingPlanner and topology that use hardcoded batch size (512) and HBM memory limit (32Gb), respectively. This change allows specifying the ShardingPlanner and Topology to more accurately reflect the machine capabilities. The change is intentionally limited to `_shard_modules` only and not public `shard_modules` to avoid changing the contract for the latter.

Reviewed By: sarckk

Differential Revision: D69163227
che-sh added a commit to che-sh/torchrec that referenced this pull request Feb 12, 2025
@che-sh @facebook-github-bot
Allow passing planner to _shard_modules (pytorch#2732)
6358fc5 
Summary:

`_shard_modules` function is used in fx_traceability tests for SDD and SemiSync pipeline. It uses a default ShardingPlanner and topology that use hardcoded batch size (512) and HBM memory limit (32Gb), respectively. This change allows specifying the ShardingPlanner and Topology to more accurately reflect the machine capabilities. The change is intentionally limited to `_shard_modules` only and not public `shard_modules` to avoid changing the contract for the latter.

Reviewed By: sarckk

Differential Revision: D69163227
che-sh added a commit to che-sh/torchrec that referenced this pull request Feb 12, 2025
@che-sh @facebook-github-bot
Allow passing planner to _shard_modules (pytorch#2732)
ce867ab 
Summary:

`_shard_modules` function is used in fx_traceability tests for SDD and SemiSync pipeline. It uses a default ShardingPlanner and topology that use hardcoded batch size (512) and HBM memory limit (32Gb), respectively. This change allows specifying the ShardingPlanner and Topology to more accurately reflect the machine capabilities. The change is intentionally limited to `_shard_modules` only and not public `shard_modules` to avoid changing the contract for the latter.

Reviewed By: sarckk

Differential Revision: D69163227
che-sh added a commit to che-sh/torchrec that referenced this pull request Feb 13, 2025
@che-sh @facebook-github-bot
Allow passing planner to _shard_modules (pytorch#2732)
ccccb98 
Summary:

`_shard_modules` function is used in fx_traceability tests for SDD and SemiSync pipeline. It uses a default ShardingPlanner and topology that use hardcoded batch size (512) and HBM memory limit (32Gb), respectively. This change allows specifying the ShardingPlanner and Topology to more accurately reflect the machine capabilities. The change is intentionally limited to `_shard_modules` only and not public `shard_modules` to avoid changing the contract for the latter.

Reviewed By: sarckk

Differential Revision: D69163227
che-sh added a commit to che-sh/torchrec that referenced this pull request Feb 13, 2025
@che-sh @facebook-github-bot
Allow passing planner to _shard_modules (pytorch#2732)
9fa7418 
Summary:

`_shard_modules` function is used in fx_traceability tests for SDD and SemiSync pipeline. It uses a default ShardingPlanner and topology that use hardcoded batch size (512) and HBM memory limit (32Gb), respectively. This change allows specifying the ShardingPlanner and Topology to more accurately reflect the machine capabilities. The change is intentionally limited to `_shard_modules` only and not public `shard_modules` to avoid changing the contract for the latter.

Reviewed By: sarckk

Differential Revision: D69163227
che-sh added a commit to che-sh/torchrec that referenced this pull request Feb 13, 2025
@che-sh @facebook-github-bot
Allow passing planner to _shard_modules (pytorch#2732)
79bf835 
Summary:

`_shard_modules` function is used in fx_traceability tests for SDD and SemiSync pipeline. It uses a default ShardingPlanner and topology that use hardcoded batch size (512) and HBM memory limit (32Gb), respectively. This change allows specifying the ShardingPlanner and Topology to more accurately reflect the machine capabilities. The change is intentionally limited to `_shard_modules` only and not public `shard_modules` to avoid changing the contract for the latter.

Reviewed By: sarckk

Differential Revision: D69163227
che-sh added a commit to che-sh/torchrec that referenced this pull request Feb 13, 2025
@che-sh @facebook-github-bot
Allow passing planner to _shard_modules (pytorch#2732)
84e05e9 
Summary:

`_shard_modules` function is used in fx_traceability tests for SDD and SemiSync pipeline. It uses a default ShardingPlanner and topology that use hardcoded batch size (512) and HBM memory limit (32Gb), respectively. This change allows specifying the ShardingPlanner and Topology to more accurately reflect the machine capabilities. The change is intentionally limited to `_shard_modules` only and not public `shard_modules` to avoid changing the contract for the latter.

Reviewed By: sarckk

Differential Revision: D69163227
che-sh added a commit to che-sh/torchrec that referenced this pull request Feb 13, 2025
@che-sh @facebook-github-bot
Allow passing planner to _shard_modules (pytorch#2732)
4d0ddab 
Summary:

`_shard_modules` function is used in fx_traceability tests for SDD and SemiSync pipeline. It uses a default ShardingPlanner and topology that use hardcoded batch size (512) and HBM memory limit (32Gb), respectively. This change allows specifying the ShardingPlanner and Topology to more accurately reflect the machine capabilities. The change is intentionally limited to `_shard_modules` only and not public `shard_modules` to avoid changing the contract for the latter.

Reviewed By: sarckk

Differential Revision: D69163227
che-sh added a commit to che-sh/torchrec that referenced this pull request Feb 13, 2025
@che-sh @facebook-github-bot
Allow passing planner to _shard_modules (pytorch#2732)
ba937b0 
Summary:

`_shard_modules` function is used in fx_traceability tests for SDD and SemiSync pipeline. It uses a default ShardingPlanner and topology that use hardcoded batch size (512) and HBM memory limit (32Gb), respectively. This change allows specifying the ShardingPlanner and Topology to more accurately reflect the machine capabilities. The change is intentionally limited to `_shard_modules` only and not public `shard_modules` to avoid changing the contract for the latter.

Reviewed By: sarckk

Differential Revision: D69163227
che-sh added a commit to che-sh/torchrec that referenced this pull request Feb 13, 2025
@che-sh @facebook-github-bot
Allow passing planner to _shard_modules (pytorch#2732)
5a8bb04 
Summary:

`_shard_modules` function is used in fx_traceability tests for SDD and SemiSync pipeline. It uses a default ShardingPlanner and topology that use hardcoded batch size (512) and HBM memory limit (32Gb), respectively. This change allows specifying the ShardingPlanner and Topology to more accurately reflect the machine capabilities. The change is intentionally limited to `_shard_modules` only and not public `shard_modules` to avoid changing the contract for the latter.

Reviewed By: sarckk

Differential Revision: D69163227
che-sh added a commit to che-sh/torchrec that referenced this pull request Feb 14, 2025
@che-sh @facebook-github-bot
Allow passing planner to _shard_modules (pytorch#2732)
626c678 
Summary:

`_shard_modules` function is used in fx_traceability tests for SDD and SemiSync pipeline. It uses a default ShardingPlanner and topology that use hardcoded batch size (512) and HBM memory limit (32Gb), respectively. This change allows specifying the ShardingPlanner and Topology to more accurately reflect the machine capabilities. The change is intentionally limited to `_shard_modules` only and not public `shard_modules` to avoid changing the contract for the latter.

Reviewed By: sarckk

Differential Revision: D69163227
che-sh added a commit to che-sh/torchrec that referenced this pull request Feb 14, 2025
@che-sh @facebook-github-bot
Allow passing planner to _shard_modules (pytorch#2732)
aac0d6b 
Summary:

`_shard_modules` function is used in fx_traceability tests for SDD and SemiSync pipeline. It uses a default ShardingPlanner and topology that use hardcoded batch size (512) and HBM memory limit (32Gb), respectively. This change allows specifying the ShardingPlanner and Topology to more accurately reflect the machine capabilities. The change is intentionally limited to `_shard_modules` only and not public `shard_modules` to avoid changing the contract for the latter.

Reviewed By: sarckk

Differential Revision: D69163227
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers
No reviews
Assignees
No one assigned
Labels
CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. fb-exported
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
2 participants
@che-sh @facebook-github-bot