KEMBAR78
[TP] Enable embedding sharding in TP API by fduwjj · Pull Request #111177 · pytorch/pytorch · GitHub
Skip to content

[TP] Enable embedding sharding in TP API #111177

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
Closed
wants to merge 9 commits into from
Closed

[TP] Enable embedding sharding in TP API #111177

wants to merge 9 commits into from

Conversation

@fduwjj
Copy link
Contributor

@fduwjj fduwjj commented Oct 13, 2023 •
edited
Loading

This was referenced Oct 13, 2023
Closed
Closed
@pytorch-bot
Copy link

pytorch-bot bot commented Oct 13, 2023 •
edited
Loading

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/111177

Note: Links to docs will display an error until the docs builds have been completed.

✅ No Failures

As of commit cd69e2d with merge base 35750bf (image):
💚 Looks good so far! There are no failures yet. 💚

This comment was automatically generated by Dr. CI and updates every 15 minutes.

@fduwjj fduwjj mentioned this pull request Oct 13, 2023
Closed
fduwjj added a commit that referenced this pull request Oct 13, 2023
@fduwjj
[TP] Enable embedding sharding in TP API
1133381 
ghstack-source-id: e1c0bf4
Pull Request resolved: #111177
@fduwjj fduwjj added ciflow/trunk Trigger trunk jobs on your pull request module: dtensor distributed tensor tag release notes: distributed (dtensor) release notes category labels Oct 13, 2023
wanchaol
wanchaol reviewed Oct 13, 2023
View reviewed changes
Copy link
Collaborator

@wanchaol wanchaol left a comment •
edited
Loading

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Can you also update the doc of ColwiseParallel and RowwiseParallel style to explicitly mention what modules are supported?

@fduwjj fduwjj requested a review from wanchaol October 13, 2023 21:38
@fduwjj
Update on "[TP] Enable embedding sharding in TP API"
10f52df 
We see use cases where embedding sharding is also needed in TP API so we enabled it in the API since DTensor already support colwise embedding sharding.


[ghstack-poisoned]
fduwjj added a commit that referenced this pull request Oct 13, 2023
@fduwjj
[TP] Enable embedding sharding in TP API
afa1e9b 
ghstack-source-id: 25fa245
Pull Request resolved: #111177
wanchaol
wanchaol approved these changes Oct 13, 2023
View reviewed changes
Copy link
Collaborator

@wanchaol wanchaol left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

stamp to unblock, some suggestions about the doc

@fduwjj
Update on "[TP] Enable embedding sharding in TP API"
6dc12c4 
We see use cases where embedding sharding is also needed in TP API so we enabled it in the API since DTensor already support colwise embedding sharding.


[ghstack-poisoned]
@fduwjj
Update on "[TP] Enable embedding sharding in TP API"
89a37bd 
We see use cases where embedding sharding is also needed in TP API so we enabled it in the API since DTensor already support colwise embedding sharding.


[ghstack-poisoned]
@fduwjj
Update on "[TP] Enable embedding sharding in TP API"
e768aab 
We see use cases where embedding sharding is also needed in TP API so we enabled it in the API since DTensor already support colwise embedding sharding.


[ghstack-poisoned]
wanchaol
wanchaol reviewed Oct 15, 2023
View reviewed changes
@fduwjj
Update on "[TP] Enable embedding sharding in TP API"
64648bd 
We see use cases where embedding sharding is also needed in TP API so we enabled it in the API since DTensor already support colwise embedding sharding.


[ghstack-poisoned]
@fduwjj
Update on "[TP] Enable embedding sharding in TP API"
aed6d42 
We see use cases where embedding sharding is also needed in TP API so we enabled it in the API since DTensor already support colwise embedding sharding.


[ghstack-poisoned]
@fduwjj
Update on "[TP] Enable embedding sharding in TP API"
ed77c7e 
We see use cases where embedding sharding is also needed in TP API so we enabled it in the API since DTensor already support colwise embedding sharding.


[ghstack-poisoned]
@fduwjj
Copy link
Contributor Author

fduwjj commented Oct 15, 2023

@pytorchbot merge

@pytorchmergebot
Copy link
Collaborator

Merge started

Your change will be merged once all checks pass (ETA 0-4 Hours).

Learn more about merging in the wiki.

Questions? Feedback? Please reach out to the PyTorch DevX Team

Advanced Debugging
Check the merge workflow status
here

@pytorchmergebot
Copy link
Collaborator

Merge failed

Reason: 1 mandatory check(s) failed. The first few are:

Dig deeper by viewing the failures on hud

Details for Dev Infra team Raised by workflow job

Failing merge rule: Core Maintainers

@fduwjj
Update on "[TP] Enable embedding sharding in TP API"
cd69e2d 
We see use cases where embedding sharding is also needed in TP API so we enabled it in the API since DTensor already support colwise embedding sharding.


[ghstack-poisoned]
fduwjj added a commit that referenced this pull request Oct 15, 2023
@fduwjj
[TP] Enable embedding sharding in TP API
1903e5a 
ghstack-source-id: 979ddcb
Pull Request resolved: #111177
@fduwjj
Copy link
Contributor Author

fduwjj commented Oct 15, 2023

@pytorchbot merge

@pytorchmergebot
Copy link
Collaborator

Merge started

Your change will be merged once all checks pass (ETA 0-4 Hours).

Learn more about merging in the wiki.

Questions? Feedback? Please reach out to the PyTorch DevX Team

Advanced Debugging
Check the merge workflow status
here

@wanchaol
Copy link
Collaborator

@fduwjj can you fix the doc errors? Looks like the doc generated from this PR have multiple issues, see https://docs-preview.pytorch.org/pytorch/pytorch/111177/distributed.tensor.parallel.html#torch.distributed.tensor.parallel.style.ColwiseParallel

All the examples in the doc (i.e. Colwise/Rowwise) are not properly formatted, and also parallelize_module example can probably remove Pairwise and use ColwiseParallel

@fduwjj fduwjj mentioned this pull request Oct 16, 2023
Closed
@fduwjj
Copy link
Contributor Author

fduwjj commented Oct 16, 2023

@wanchaol Fix in #111346.

pytorchmergebot pushed a commit that referenced this pull request Oct 16, 2023
@fduwjj @pytorchmergebot
[TP] Fix TP doc format to show examples correctly (#111346)
bfcd869 
Pull Request resolved: #111346
Approved by: https://github.com/wanchaol
ghstack dependencies: #111160, #111166, #111176, #111177
yeounoh pushed a commit to yeounoh/pytorch that referenced this pull request Oct 16, 2023
@fduwjj @yeounoh
[TP] Enable embedding sharding in TP API (pytorch#111177)
b84319c 
We see use cases where embedding sharding is also needed in TP API so we enabled it in the API since DTensor already support colwise embedding sharding.

Pull Request resolved: pytorch#111177
Approved by: https://github.com/wanchaol
ghstack dependencies: pytorch#111160, pytorch#111166, pytorch#111176
yeounoh pushed a commit to yeounoh/pytorch that referenced this pull request Oct 16, 2023
@fduwjj @yeounoh
[TP] Fix TP doc format to show examples correctly (pytorch#111346)
59d8758 
Pull Request resolved: pytorch#111346
Approved by: https://github.com/wanchaol
ghstack dependencies: pytorch#111160, pytorch#111166, pytorch#111176, pytorch#111177
@facebook-github-bot facebook-github-bot deleted the gh/fduwjj/110/head branch October 18, 2023 14:25
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@wanchaol wanchaol wanchaol approved these changes

@mrshenli mrshenli Awaiting requested review from mrshenli

@zhaojuanmao zhaojuanmao Awaiting requested review from zhaojuanmao

@rohan-varma rohan-varma Awaiting requested review from rohan-varma

@H-Huang H-Huang Awaiting requested review from H-Huang

@awgu awgu Awaiting requested review from awgu

@kwen2501 kwen2501 Awaiting requested review from kwen2501

@fegin fegin Awaiting requested review from fegin

@wz337 wz337 Awaiting requested review from wz337

@kiukchung kiukchung Awaiting requested review from kiukchung

@d4l3k d4l3k Awaiting requested review from d4l3k

Assignees

No one assigned

Labels

ciflow/trunk Trigger trunk jobs on your pull request Merged module: dtensor distributed tensor tag release notes: distributed (dtensor) release notes category

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

3 participants

@fduwjj @pytorchmergebot @wanchaol