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.

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

Add an exception for torch<1.8.1 #1556

Merged
merged 4 commits into from
Apr 16, 2022
Merged

Conversation

nijkah
Copy link
Contributor

@nijkah nijkah commented Apr 9, 2022

Description

To support old versions of torch <1.8.1, added an exception for importing torch.profiler.

Fixes #1548

Type of change

  • Bug fix (non-breaking change which fixes an issue)

  • Logs

Checklist:

  • Have you added tests that prove your fix is effective or that this feature works?
  • Has code been commented, particularly in hard-to-understand areas?
  • Have you made corresponding changes to the documentation?

@sagemaker-neo-ci-bot
Copy link
Collaborator

AWS CodeBuild CI Report

  • CodeBuild project: torch-serve-build-win
  • Commit ID: ea7a93c
  • Result: FAILED
  • Build Logs (available for 30 days)

Powered by github-codebuild-logs, available on the AWS Serverless Application Repository

@sagemaker-neo-ci-bot
Copy link
Collaborator

AWS CodeBuild CI Report

  • CodeBuild project: torch-serve-build-win
  • Commit ID: a5d40de
  • Result: FAILED
  • Build Logs (available for 30 days)

Powered by github-codebuild-logs, available on the AWS Serverless Application Repository

@sagemaker-neo-ci-bot
Copy link
Collaborator

AWS CodeBuild CI Report

  • CodeBuild project: torch-serve-build-cpu
  • Commit ID: ea7a93c
  • Result: FAILED
  • Build Logs (available for 30 days)

Powered by github-codebuild-logs, available on the AWS Serverless Application Repository

@sagemaker-neo-ci-bot
Copy link
Collaborator

AWS CodeBuild CI Report

  • CodeBuild project: torch-serve-build-cpu
  • Commit ID: a5d40de
  • Result: FAILED
  • Build Logs (available for 30 days)

Powered by github-codebuild-logs, available on the AWS Serverless Application Repository

@sagemaker-neo-ci-bot
Copy link
Collaborator

AWS CodeBuild CI Report

  • CodeBuild project: torch-serve-build-gpu
  • Commit ID: ea7a93c
  • Result: FAILED
  • Build Logs (available for 30 days)

Powered by github-codebuild-logs, available on the AWS Serverless Application Repository

@sagemaker-neo-ci-bot
Copy link
Collaborator

AWS CodeBuild CI Report

  • CodeBuild project: torch-serve-build-gpu
  • Commit ID: a5d40de
  • Result: FAILED
  • Build Logs (available for 30 days)

Powered by github-codebuild-logs, available on the AWS Serverless Application Repository

@sagemaker-neo-ci-bot
Copy link
Collaborator

AWS CodeBuild CI Report

  • CodeBuild project: torch-serve-build-win
  • Commit ID: 1c2e365
  • Result: FAILED
  • Build Logs (available for 30 days)

Powered by github-codebuild-logs, available on the AWS Serverless Application Repository

@sagemaker-neo-ci-bot
Copy link
Collaborator

AWS CodeBuild CI Report

  • CodeBuild project: torch-serve-build-cpu
  • Commit ID: 1c2e365
  • Result: SUCCEEDED
  • Build Logs (available for 30 days)

Powered by github-codebuild-logs, available on the AWS Serverless Application Repository

@sagemaker-neo-ci-bot
Copy link
Collaborator

AWS CodeBuild CI Report

  • CodeBuild project: torch-serve-build-gpu
  • Commit ID: 1c2e365
  • Result: SUCCEEDED
  • Build Logs (available for 30 days)

Powered by github-codebuild-logs, available on the AWS Serverless Application Repository

@msaroufim
Copy link
Member

Interesting, could you tell me a bit more about why you need to keep torch < 1.8.1 in your environment?

@nijkah
Copy link
Contributor Author

nijkah commented Apr 11, 2022

@msaroufim Actually, I think this PR is just one option for managing the supported version of the library.

As I mentioned in #1548, there is the other option to specify the supported versions of pytorch.
Currently, users cannot identify these issues until debugging some error logs.

And personally, I think it is better to support the library generally regardless of the versions of dependent libraries if it is not necessary or too old.
For example, mmcv, one of the actively maintained core modules used in various frameworks, still supports torch1.5.0 with cuda 10.2 for its latest version.

For now, I thought torchserve doesn't depend on torch.profiler heavily as much as limiting the version of pytorch.

@sagemaker-neo-ci-bot
Copy link
Collaborator

AWS CodeBuild CI Report

  • CodeBuild project: torch-serve-build-win
  • Commit ID: 5eda5d7
  • Result: SUCCEEDED
  • Build Logs (available for 30 days)

Powered by github-codebuild-logs, available on the AWS Serverless Application Repository

@sagemaker-neo-ci-bot
Copy link
Collaborator

AWS CodeBuild CI Report

  • CodeBuild project: torch-serve-build-cpu
  • Commit ID: 5eda5d7
  • Result: SUCCEEDED
  • Build Logs (available for 30 days)

Powered by github-codebuild-logs, available on the AWS Serverless Application Repository

@sagemaker-neo-ci-bot
Copy link
Collaborator

AWS CodeBuild CI Report

  • CodeBuild project: torch-serve-build-gpu
  • Commit ID: 5eda5d7
  • Result: SUCCEEDED
  • Build Logs (available for 30 days)

Powered by github-codebuild-logs, available on the AWS Serverless Application Repository

@msaroufim msaroufim merged commit 6c13b62 into pytorch:master Apr 16, 2022
@sagemaker-neo-ci-bot
Copy link
Collaborator

AWS CodeBuild CI Report

  • CodeBuild project: torch-serve-build-win
  • Commit ID: 5eda5d7
  • Result: SUCCEEDED
  • Build Logs (available for 30 days)

Powered by github-codebuild-logs, available on the AWS Serverless Application Repository

@sagemaker-neo-ci-bot
Copy link
Collaborator

AWS CodeBuild CI Report

  • CodeBuild project: torch-serve-build-gpu
  • Commit ID: 5eda5d7
  • Result: SUCCEEDED
  • Build Logs (available for 30 days)

Powered by github-codebuild-logs, available on the AWS Serverless Application Repository

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

[past torch version support] Using torch.profiler raises error when using torch<1.8.1
4 participants