Update ml-goodput-measurement package for prod PyPi release v0.0.15.

PiperOrigin-RevId: 805568317

Author: dipannita08

CHANGELOG.md

@@ -20,6 +20,11 @@ To release a new version (e.g. from `1.0.0` -> `2.0.0`):
   * Update the `[Unreleased]` url: `v1.0.0...HEAD` -> `v2.0.0...HEAD`
 
 -->
+## [0.0.15] - 2025-09-09
+
+* Refactor the GoodputMonitor to use multiprocessing instead of multithreading.
+* Add more context with structured logging for metric uploads to GCM Monitoring.
+
 ## [0.0.14] - 2025-07-29
 
 * Handle non-fatal internal errors, log warnings & continue.
@@ -105,7 +110,8 @@ To release a new version (e.g. from `1.0.0` -> `2.0.0`):
 * Initial release of ML Goodput Measurement PyPi package
 * Feature: Contains the Goodput module which allows logging and retrieval of training job's overall productive Goodput
 
-[0.0.13]: https://github.com/AI-Hypercomputer/ml-goodput-measurement/compare/v0.0.13...v0.0.14
+[0.0.15]: https://github.com/AI-Hypercomputer/ml-goodput-measurement/compare/v0.0.14...v0.0.15
+[0.0.14]: https://github.com/AI-Hypercomputer/ml-goodput-measurement/compare/v0.0.13...v0.0.14
 [0.0.13]: https://github.com/AI-Hypercomputer/ml-goodput-measurement/compare/v0.0.12...v0.0.13
 [0.0.12]: https://github.com/AI-Hypercomputer/ml-goodput-measurement/compare/v0.0.11...v0.0.12
 [0.0.11]: https://github.com/AI-Hypercomputer/ml-goodput-measurement/compare/v0.0.10...v0.0.11

README.md

@@ -559,11 +559,11 @@ goodput_monitor = monitoring.GoodputMonitor(
 
 ### Monitor Cumulative Goodput Metrics
 
-#### Start asynchronous "query and upload" of Goodput
+#### Start Asynchronous "Query and Upload" of Goodput
 
-Call the `start_goodput_uploader` API to spin off a thread which continuously
-queries and uploads cumulative Goodput metrics to Tensorboard & Google Cloud
-Monitoring.
+Call the `start_goodput_uploader` API to launch a background process which
+continuously queries and uploads cumulative Goodput metrics to Tensorboard
+& Google Cloud Monitoring.
 
 > **_NOTE:_** This will upload computed metrics to Google Cloud Monitoring
 by default.
@@ -587,7 +587,7 @@ goodput_monitor.start_goodput_uploader()
 Call the `stop_goodput_uploader` API to perform a final upload of all metrics
 and safely exit.
 
-> **_NOTE:_** This will stop all cumulative metrics upload threads.
+> **_NOTE:_** This will stop all cumulative metrics upload processes.
 
 ```python
 goodput_monitor.stop_goodput_uploader()
@@ -598,7 +598,7 @@ goodput_monitor.stop_goodput_uploader()
 #### Start asynchronous "query and upload" of Rolling Window Goodput
 
 Call the `start_rolling_window_goodput_uploader` API to start a background
-thread that continuously queries and uploads **rolling window goodput metrics**
+process that continuously queries and uploads **rolling window goodput metrics**
 to Google Cloud Monitoring.
 
 You must provide a list of window durations in seconds (e.g., `[60, 300, 900]`
@@ -616,9 +616,9 @@ goodput_monitor.start_rolling_window_goodput_uploader(rolling_windows_seconds=[6
 #### Stop the Rolling Window Goodput Uploader
 
 Call the `stop_rolling_window_goodput_uploader` API to perform a final upload
-of rolling window metrics and safely shut down the background uploader thread.
+of rolling window metrics and safely shut down the background uploader process.
 
-> **_NOTE:_** This will stop all rolling window metrics upload threads.
+> **_NOTE:_** This will stop all rolling window metrics upload processes.
 
 ```python
 goodput_monitor.stop_rolling_window_goodput_uploader()

pyproject.toml

@@ -14,7 +14,7 @@
 
 [project]
 name = "ml_goodput_measurement"
-version = "0.0.14"
+version = "0.0.15"
 authors = [
   { name="Cloud TPU Team", email="[email protected]" },
 ]