Skip to content

Pytorch 2.3 upgrade #4239

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
Merged
suzhoum merged 3 commits into from
Jun 10, 2024
Merged

Pytorch 2.3 upgrade #4239

suzhoum merged 3 commits into from
Jun 10, 2024

Conversation

@suzhoum
Copy link
Contributor

@suzhoum suzhoum commented May 29, 2024

Issue #, if available:

Description of changes:

By submitting this pull request, I confirm that you can use, modify, copy, and redistribute this contribution, under the terms of your choice.

@suzhoum suzhoum force-pushed the upgrade_pytorch_2_3 branch from 03945f3 to 33f6ba7 Compare May 29, 2024 17:25
@yinweisu
Copy link
Contributor

yinweisu commented May 29, 2024

Previous CI Run Current CI Run

@suzhoum suzhoum force-pushed the upgrade_pytorch_2_3 branch from 33f6ba7 to 8e976c8 Compare May 29, 2024 20:16
@yinweisu
Copy link
Contributor

yinweisu commented May 29, 2024

Previous CI Run Current CI Run
botocore==1.34.114 botocore==1.34.115
boto3==1.34.114 boto3==1.34.115
sympy==1.12 sympy==1.12.1
botocore==1.34.114 botocore==1.34.115
boto3==1.34.114 boto3==1.34.115
sympy==1.12 sympy==1.12.1

@suzhoum suzhoum added the run-multi-gpu Run multimodal multi-gpu tests label May 29, 2024
@suzhoum
Copy link
Contributor Author

suzhoum commented May 29, 2024 •
edited
Loading

/benchmark module=multimodal preset=multimodal_best benchmark=tabular_full time_limit=4h

@github-actions
Copy link

github-actions bot commented May 29, 2024

Job PR-4239-8e976c8 is done.
Docs are uploaded to http://autogluon-staging.s3-website-us-west-2.amazonaws.com/PR-4239/8e976c8/index.html

@suzhoum
Copy link
Contributor Author

suzhoum commented May 30, 2024

/benchmark module=multimodal preset=multimodal_best benchmark=tabular_full time_limit=4h

@suzhoum
Copy link
Contributor Author

suzhoum commented May 30, 2024

/benchmark module=tabular preset=tabular_best benchmark=tabular_full time_limit=1h

@prateekdesai04
Copy link
Contributor

prateekdesai04 commented May 30, 2024 •
edited by yinweisu
Loading

/benchmark module=tabular preset=tabular_best benchmark=tabular_full time_limit=1h
Benchmark Output

Benchmark Output

Benchmark Output

Benchmark Test Result - Pass Evaluation Results Path: s3://autogluon-ci-benchmark/evaluation/multimodal/upgrade_pytorch_2_3
The dashboard website is: http://autogluon-staging.s3-website-us-west-2.amazonaws.com/benchmark-dashboard/upgrade_pytorch_2_3/0e3bc0e54ab4b0edf865c8b99e1418472006d6b7/index.html
Benchmark Test Result - Pass Evaluation Results Path: s3://autogluon-ci-benchmark/evaluation/tabular/upgrade_pytorch_2_3
The dashboard website is: http://autogluon-staging.s3-website-us-west-2.amazonaws.com/benchmark-dashboard/upgrade_pytorch_2_3/0e3bc0e54ab4b0edf865c8b99e1418472006d6b7/index.html
Benchmark Test Result - Pass Evaluation Results Path: s3://autogluon-ci-benchmark/evaluation/multimodal/upgrade_pytorch_2_3
The dashboard website is: http://autogluon-staging.s3-website-us-west-2.amazonaws.com/benchmark-dashboard/upgrade_pytorch_2_3/0e3bc0e54ab4b0edf865c8b99e1418472006d6b7/index.html
Benchmark Output

Benchmark Test Result - Pass Evaluation Results Path: s3://autogluon-ci-benchmark/evaluation/multimodal/upgrade_pytorch_2_3
The dashboard website is: http://autogluon-staging.s3-website-us-west-2.amazonaws.com/benchmark-dashboard/upgrade_pytorch_2_3/0e3bc0e54ab4b0edf865c8b99e1418472006d6b7/index.html
Benchmark Output

@suzhoum suzhoum force-pushed the upgrade_pytorch_2_3 branch from 8e976c8 to 088383d Compare May 31, 2024 16:29
@yinweisu
Copy link
Contributor

yinweisu commented May 31, 2024

Previous CI Run Current CI Run
torchvision==0.18.0 torchvision==0.17.2
lightning==2.2.5 lightning==2.1.4
triton==2.3.0 triton==2.2.0
torch==2.3.0 torch==2.2.2
nvidia-nccl-cu12==2.20.5 nvidia-nccl-cu12==2.19.3
torchvision==0.18.0 torchvision==0.17.2
lightning==2.2.5 lightning==2.1.4
triton==2.3.0 triton==2.2.0
torch==2.3.0 torch==2.2.2
nvidia-nccl-cu12==2.20.5 nvidia-nccl-cu12==2.19.3

@github-actions
Copy link

github-actions bot commented May 31, 2024

Job PR-4239-088383d is done.
Docs are uploaded to http://autogluon-staging.s3-website-us-west-2.amazonaws.com/PR-4239/088383d/index.html

@Innixma Innixma added this to the 1.1.1 Release milestone Jun 5, 2024
@shchur shchur force-pushed the upgrade_pytorch_2_3 branch from 088383d to da7ba71 Compare June 6, 2024 07:46
@shchur
Copy link
Collaborator

shchur commented Jun 6, 2024

@suzhoum FYI I have rebased the PR branch on master to include our latest commits to time series

@yinweisu
Copy link
Contributor

yinweisu commented Jun 6, 2024

Previous CI Run Current CI Run
torchvision==0.18.1 torchvision==0.17.2
lightning==2.2.5 lightning==2.1.4
triton==2.3.1 triton==2.2.0
torch==2.3.1 torch==2.2.2
nvidia-nccl-cu12==2.20.5 nvidia-nccl-cu12==2.19.3
torchvision==0.18.1 torchvision==0.17.2
lightning==2.2.5 lightning==2.1.4
triton==2.3.1 triton==2.2.0
torch==2.3.1 torch==2.2.2
nvidia-nccl-cu12==2.20.5 nvidia-nccl-cu12==2.19.3

@github-actions
Copy link

github-actions bot commented Jun 6, 2024

Job PR-4239-da7ba71 is done.
Docs are uploaded to http://autogluon-staging.s3-website-us-west-2.amazonaws.com/PR-4239/da7ba71/index.html

@shchur
Copy link
Collaborator

shchur commented Jun 7, 2024

We probably need a stricter lower bound on lightning based on this CVE https://github.com/awslabs/gluonts/security/dependabot/23

@suzhoum suzhoum force-pushed the upgrade_pytorch_2_3 branch from da7ba71 to e3a977b Compare June 7, 2024 20:52
@suzhoum
Copy link
Contributor Author

suzhoum commented Jun 7, 2024

We probably need a stricter lower bound on lightning based on this CVE https://github.com/awslabs/gluonts/security/dependabot/23

Thanks, I just updated the lower bound.

@suzhoum suzhoum marked this pull request as ready for review June 7, 2024 20:54
prateekdesai04
prateekdesai04 approved these changes Jun 7, 2024
View reviewed changes
Copy link
Contributor

@prateekdesai04 prateekdesai04 left a comment

Choose a reason for hiding this comment

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

LGTM!

@yinweisu
Copy link
Contributor

yinweisu commented Jun 7, 2024

Previous CI Run Current CI Run
botocore==1.34.121 botocore==1.34.122
boto3==1.34.121 boto3==1.34.122
google-auth==2.29.0 google-auth==2.30.0
tensorboard==2.16.2 tensorboard==2.17.0
typing_extensions==4.12.1 typing_extensions==4.12.2
botocore==1.34.121 botocore==1.34.122
boto3==1.34.121 boto3==1.34.122
google-auth==2.29.0 google-auth==2.30.0
tensorboard==2.16.2 tensorboard==2.17.0
typing_extensions==4.12.1 typing_extensions==4.12.2

Innixma
Innixma requested changes Jun 7, 2024
View reviewed changes
core/src/autogluon/core/_setup_utils.py Outdated
"torch": ">=2.2,<2.3", # Major version cap, sync with common/src/autogluon/common/utils/try_import.py
"lightning": ">=2.1,<2.2", # Major version cap
"pytorch_lightning": ">=2.2,<2.3", # Major version cap, capping `lightning` does not cap `pytorch_lightning`!
"torch": ">=2.3,<2.4", # Major version cap, sync with common/src/autogluon/common/utils/try_import.py
Copy link
Contributor

@Innixma Innixma Jun 7, 2024

Choose a reason for hiding this comment

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

Why are we pinning torch to >=2.3? This seems much too strict.

Copy link
Contributor Author

@suzhoum suzhoum Jun 7, 2024 •
edited
Loading

Choose a reason for hiding this comment

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

I checked the previous releases and we have been restricting torch to be within one major version. If you feel this is too strict, i can make it to be >=2.2

Innixma
Innixma reviewed Jun 7, 2024
View reviewed changes
common/src/autogluon/common/utils/try_import.py Outdated
"Unable to import dependency torch\n"
"A quick tip is to install via `pip install torch`.\n"
"The minimum torch version is currently 2.1." # sync with core/_setup_utils.py
"The minimum torch version is currently 2.3." # sync with core/_setup_utils.py
Copy link
Contributor

@Innixma Innixma Jun 7, 2024

Choose a reason for hiding this comment

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

Why are we pinning torch to >=2.3? This seems much too strict.

@yinweisu
Copy link
Contributor

yinweisu commented Jun 7, 2024

Previous CI Run Current CI Run
torchvision==0.18.1 torchvision==0.17.2
lightning==2.2.5 lightning==2.1.4
triton==2.3.1 triton==2.2.0
torch==2.3.1 torch==2.2.2
nvidia-nccl-cu12==2.20.5 nvidia-nccl-cu12==2.19.3
torchvision==0.18.1 torchvision==0.17.2
lightning==2.2.5 lightning==2.1.4
triton==2.3.1 triton==2.2.0
torch==2.3.1 torch==2.2.2
nvidia-nccl-cu12==2.20.5 nvidia-nccl-cu12==2.19.3

Innixma
Innixma approved these changes Jun 10, 2024
View reviewed changes
Copy link
Contributor

@Innixma Innixma left a comment

Choose a reason for hiding this comment

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

LGTM

@github-actions
Copy link

github-actions bot commented Jun 10, 2024

Job PR-4239-8645ab6 is done.
Docs are uploaded to http://autogluon-staging.s3-website-us-west-2.amazonaws.com/PR-4239/8645ab6/index.html

@suzhoum suzhoum merged commit 2151075 into master Jun 10, 2024
@suzhoum suzhoum deleted the upgrade_pytorch_2_3 branch June 10, 2024 22:44
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@Innixma Innixma Innixma approved these changes

@prateekdesai04 prateekdesai04 prateekdesai04 approved these changes

@tonyhoo tonyhoo Awaiting requested review from tonyhoo

@shchur shchur Awaiting requested review from shchur

Assignees

No one assigned

Labels

run-multi-gpu Run multimodal multi-gpu tests

Projects

None yet

Milestone

1.1.1 Release

Development

Successfully merging this pull request may close these issues.

5 participants

@suzhoum @yinweisu @prateekdesai04 @shchur @Innixma