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

[Feature Request]: When sharing large models, RunInference should respect environment variables #27694

Closed
1 of 15 tasks
damccorm opened this issue Jul 26, 2023 · 1 comment
Closed
1 of 15 tasks

[Feature Request]: When sharing large models, RunInference should respect environment variables #27694

damccorm opened this issue Jul 26, 2023 · 1 comment
Assignees
@damccorm
Labels
done & done Issue has been reviewed after it was closed for verification, followups, etc. new feature P2 python
Milestone
2.50.0 Release

Comments

@damccorm
Copy link
Contributor

What would you like to happen?

Right now, even if share_model_across_processes is True in RunInference, we set environment variables in the calling process (whether or not that is where the model is loaded). We should set them in the target process instead.

Issue Priority

Priority: 2 (default / most feature requests should be filed as P2)

Issue Components

  • Component: Python SDK
  • Component: Java SDK
  • Component: Go SDK
  • Component: Typescript SDK
  • Component: IO connector
  • Component: Beam examples
  • Component: Beam playground
  • Component: Beam katas
  • Component: Website
  • Component: Spark Runner
  • Component: Flink Runner
  • Component: Samza Runner
  • Component: Twister2 Runner
  • Component: Hazelcast Jet Runner
  • Component: Google Cloud Dataflow Runner
@damccorm damccorm self-assigned this Jul 26, 2023
@damccorm
Copy link
Contributor Author

This should actually just work since we don't proxy the model until needed, so the model will always be in a process that set the env vars -

beam/sdks/python/apache_beam/ml/inference/base.py

Line 846 in df6964a

model = multi_process_shared.MultiProcessShared(
- so there's no work needed here

@damccorm damccorm added the done & done Issue has been reviewed after it was closed for verification, followups, etc. label Jul 31, 2023
@github-actions github-actions bot added this to the 2.50.0 Release milestone Jul 31, 2023
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees

@damccorm damccorm

Labels
done & done Issue has been reviewed after it was closed for verification, followups, etc. new feature P2 python
Projects
None yet
Milestone
2.50.0 Release
Development

No branches or pull requests
1 participant
@damccorm