amoylan@chromium.org ml-service-team@google.com Fraction of total CPU resources used by ChromeOS ML Service, sampled every 5 minutes. wtlee@chromium.org chromeos-camera-eng@google.com ml-service-team@google.com The result of DocumentScanner.{RequestName} request, which can be OK or ERROR. jiwan@chromium.org amoylan@chromium.org The result of grammar check, which can be OK or ERROR. amoylan@chromium.org ml-service-team@google.com The result of load handwriting model, which can be OK or some specific error. The error code could be logged in machine learning service in chrome or in chromeos when handwriting model is not loaded successfully. amoylan@chromium.org ml-service-team@google.com The result of handwriting recognition, which can be OK or some specific error. jiwan@chromium.org robsc@chromium.org amoylan@chromium.org The result of heatmap palm rejection Execute request, which can be OK or some specific error. jiwan@chromium.org robsc@chromium.org amoylan@chromium.org The result of load heatmap palm rejection module, which can be OK or some specific error. chenjih@google.com nbowe@chromium.org The result of image annotation, which can be OK or some specific error. amoylan@chromium.org ml-service-team@google.com The error result of LoadModel in cases when the model name could not be determined. See MachineLearningService.*.LoadModelResult.Event for the result broken down by model name when the name is known. amoylan@chromium.org chenjih@google.com nbowe@chromium.org The detailed error code received from DLC Service while trying to request to install the ML Core DLC. amoylan@chromium.org chenjih@google.com nbowe@chromium.org Result of an attempt to request DLC Service to begin installation of the ML Core DLC. See the related histogram .DlcBeginInstallDlcServiceError for detailed error codes returned from DLC Service. amoylan@chromium.org chenjih@google.com nbowe@chromium.org The detailed error code received from DLC Service after a DLC State Change notification from DLC Service, following a request to install the ML Core DLC. "BUSY" code is considered non-final as retries are made in this case. amoylan@chromium.org chenjih@google.com nbowe@chromium.org Final result of an attempt to install the ML Core DLC. See the related histogram .DlcFinalInstallDlcServiceError for detailed error codes returned from DLC Service. amoylan@chromium.org chenjih@google.com nbowe@chromium.org Indicates at least n attempts were made to install the ML Core DLC. Note that the buckets are cumulative: If e.g. 3 attempts are made, then the "1", "2", and "3" buckets will all be incremented. amoylan@chromium.org ml-service-team@google.com Events related to the connection and disconnection of the Mojo IPC channel between Chrome and the ChromeOS ML Service. amoylan@chromium.org ml-service-team@google.com Number of worker processes registered in ml-service. Only used in ChromeOS. Sampled every 5 minutes. amoylan@chromium.org ml-service-team@google.com Peak total (shared and non-shared) memory used by ChromeOS ML Service over the last 24 hours. Sampled every 5 minutes, so may miss short-lived spikes. amoylan@chromium.org ml-service-team@google.com Process handling errors of ml-service. Emitted when there is error in process handling. For example, "failed to spawning worker process" etc. Only used in ChromeOS. amoylan@chromium.org ml-service-team@google.com The "errno" of waitpid calls used by ml-service's control process to reap the worker processes. This is only recorded when waitpid meets error and should be rare. This is a ChromeOS only metric. And we only record the corresponding enum values of "ECHILD", "EINTR" and "EINVAL" because according to the man page of waitpid, they are the only possible errno. All the other "errno" are treated as "unknown" error. curtismcmullan@chromium.org amoylan@chromium.org The result of a text suggester request, which can be OK or ERROR. amoylan@chromium.org ml-service-team@google.com Total (shared and non-shared) memory used by ChromeOS ML Service, sampled every 5 minutes. amoylan@chromium.org ml-service-team@google.com The result of Web platform handwriting recognition, which can be either OK or Error. amoylan@chromium.org ml-service-team@google.com The result of load web platform handwriting model, which can be OK or some specific error. The error code could be logged in machine learning service in chrome or in chromeos when handwriting model is not loaded successfully. amoylan@chromium.org ml-service-team@google.com The wall time used for the worker process to accept the mojo connection to control process in ml-service. Only used in ChromeOS. amoylan@chromium.org ml-service-team@google.com The wall time used by the control process to reap and cleanup the worker processes in ml-service. Only used in ChromeOS. amoylan@chromium.org ml-service-team@google.com The exit status of the worker processes of ml-service. Only used in Chrome OS. This is only emitted when the worker process exits with non-zero status. And theoretically, there can be 255 different types of status (from 1 to 255). amoylan@chromium.org ml-service-team@google.com The result of CreateGraphExecutor request for model {ModelName}, which can be OK or some specific error. amoylan@chromium.org ml-service-team@google.com The result of Execute request for model {ModelName}, which can be OK or some specific error. amoylan@chromium.org ml-service-team@google.com The result of LoadModel request for model {ModelName}, which can be OK or some specific error. Additional error information may be found in Chrome logs or ml_service logs within ChromeOS logs. jiwan@chromium.org amoylan@chromium.org amoylan@chromium.org ml-service-team@google.com CPU time (for one CPU core) of {RequestName} request. Reported for all users; users without high-resolution clocks will report 0 for short requests. amoylan@chromium.org ml-service-team@google.com Memory usage increase caused by {RequestName} request.