Skip to content

Update orttraining release pipelines to use torch 1.11.0#11018

Merged
baijumeswani merged 3 commits intomasterfrom
bmeswani/update-release-pipelines
Apr 1, 2022
Merged

Update orttraining release pipelines to use torch 1.11.0#11018
baijumeswani merged 3 commits intomasterfrom
bmeswani/update-release-pipelines

Conversation

@baijumeswani
Copy link
Contributor

@baijumeswani baijumeswani commented Mar 27, 2022

This pull request introduces pipelines for release with the following config:

  1. orttraining built with cuda 11.5 and tested against torch 1.11.0.
  2. orttraining built with cuda 11.3 and tested against torch 1.11.0.
  3. orttraining built with cuda 10.2 and tested against torch 1.11.0.

Other pipelines that used other torch versions have been removed. Moving forward, the orttraining packaging pipelines will only be tested against the latest stable torch version. The release pipelines will support as many versions of cuda as required by PTCA images.

@baijumeswani baijumeswani added training issues related to ONNX Runtime training; typically submitted using template component:training-frontend labels Mar 27, 2022
@baijumeswani baijumeswani requested review from raviskolli and snnn March 27, 2022 22:18
@baijumeswani baijumeswani force-pushed the bmeswani/update-release-pipelines branch from 39536b9 to 34fbe93 Compare March 28, 2022 16:18
@baijumeswani baijumeswani requested a review from ytaous March 28, 2022 20:31
ytaous
ytaous previously approved these changes Mar 28, 2022
@askhade
Copy link
Contributor

askhade commented Mar 29, 2022

why do we need both 11.3 and 11.5 versions for cuda?

@baijumeswani
Copy link
Contributor Author

why do we need both 11.3 and 11.5 versions for cuda?

PyTorch has stable releases for 11.3 and 11.5. I am also preparing for the future release of PTCA where they might have the need to release 11.5.

@snnn
Copy link
Contributor

snnn commented Mar 29, 2022

Would please do me a favor? Rename all the requirements txt files to "requirements.txt"(without extra words). You can create a separated folder for each of them. It helps security scanning.

@baijumeswani baijumeswani force-pushed the bmeswani/update-release-pipelines branch from a6e1241 to ed124c4 Compare March 30, 2022 23:07
snnn
snnn previously approved these changes Mar 31, 2022
Copy link
Contributor

@snnn snnn left a comment

Choose a reason for hiding this comment

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

Thank you!

@snnn
Copy link
Contributor

snnn commented Mar 31, 2022

/azp run Linux CPU CI Pipeline

@azure-pipelines
Copy link

Azure Pipelines successfully started running 1 pipeline(s).

@snnn
Copy link
Contributor

snnn commented Mar 31, 2022

/azp run Linux CPU CI Pipeline,Linux GPU CI Pipeline

@azure-pipelines
Copy link

Azure Pipelines successfully started running 2 pipeline(s).

@baijumeswani baijumeswani merged commit 249c4de into master Apr 1, 2022
@baijumeswani baijumeswani deleted the bmeswani/update-release-pipelines branch April 1, 2022 04:51
SanftMonster pushed a commit to SanftMonster/onnxruntime that referenced this pull request Apr 1, 2022
…1018)

* Update orttraining release pipelines to use torch 1.11.0

* Change requirements_torch...txt to requirements.txt

* Update cuda cmake architectures and clean up old files
seddonm1 pushed a commit to seddonm1/onnxruntime that referenced this pull request May 15, 2022
…1018)

* Update orttraining release pipelines to use torch 1.11.0

* Change requirements_torch...txt to requirements.txt

* Update cuda cmake architectures and clean up old files
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

training issues related to ONNX Runtime training; typically submitted using template

Projects

None yet

Development

Successfully merging this pull request may close these issues.

4 participants