PRADDZY opened a new pull request, #38822: URL: https://github.com/apache/beam/pull/38822
## Summary - explain why RunInference keeps setup-time model load metrics in a cache until `finish_bundle()` - replace the lingering cache TODO with lifecycle-specific comments in `base.py` - add a focused regression test that proves `setup()` caches model load metrics and `finish_bundle()` flushes them ## Test Plan - `python -m pytest -v sdks/python/apache_beam/ml/inference/base_test.py::RunInferenceBaseTest::test_setup_caches_model_load_metrics_until_finish_bundle sdks/python/apache_beam/ml/inference/base_test.py::RunInferenceBaseTest::test_timing_metrics` - `python -m pytest -v sdks/python/apache_beam/ml/inference/base_test.py` Fixes #21445 -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
