Abstract: A method and system for an application programming interface (API) based container service for supporting multiple machine learning (ML) applications is described. In particular, a scoring service container includes a base scorer to interface with a ML serving infrastructure using the API. The scoring service container also includes an application specific scorer, which itself includes a model loader and a scoring function. A model identifier is provided to the model loader, and it provides a model object. At least some parameters in a request from a client application are passed to scoring function, which produces a scoring. The base scorer returns the scoring according to the API to the ML serving infrastructure for delivery to the client application.

Abstract: A method performs service discovery in a machine learning service. The method includes detecting initialization of at least one service container, identifying label information in the at least one service container, collecting the label information for the initializing at least one service container, and storing the label information in a routing information storage to enable routing of requests to the at least one service container.

Abstract: A method by one or more electronic devices for creating an inference container on demand. The method includes receiving, over a network, a request to create the inferencing container, wherein the inferencing container is configured to provide inferencing functionality, creating the inferencing container responsive to receiving the request to create the inferencing container, and providing, over the network, a response to the request to create the inferencing container, wherein the response includes a uniform resource locator (URL) to use to submit inferencing requests to the inferencing container, wherein the URL includes a unique identifier (ID) of the inferencing container.

Abstract: A method by one or more electronic devices to provision an infrastructure for a machine learning application in a multi-tenant on-demand serving infrastructure. The method includes storing a plurality of templates, wherein each of the plurality of templates indicates a scoring interface, a web server, a definition of a continuous integration pipeline, and a definition of a continuous deployment pipeline, receiving a request to provision the infrastructure for the machine learning application using a specified template from the plurality of templates, and provisioning the infrastructure for the machine learning application using the specified template to create a version control system repository, a continuous integration pipeline, and a continuous deployment pipeline.

Abstract: A method by one or more electronic devices for spinning up a scoring container on demand. The method includes receiving, from an orchestrator component via an application programming interface (API), a request to spin up the scoring container, wherein the scoring container is configured to provide scoring functionality, spinning up the scoring container responsive to receiving the request to spin up the scoring container, and providing, to the orchestrator component via the API, a response to the request to spin up the scoring container, wherein the response includes a uniform resource locator (URL) to use to submit scoring requests to the scoring container.

Abstract: A machine learning serving infrastructure implementing a method of receiving or detecting an update of container metrics including resource usage and serviced requests per model or per container, processing the container metrics per model or per container to determine recent resource usage and serviced requests per model or per container, and rebalancing distribution of models to a plurality of containers to decrease a detected load imbalance between containers or a stressed container in the plurality of containers.

Abstract: Machine learning version management method for a prediction service includes receiving a prediction request, determining application metadata for the request that defines routing logic and a machine learning framework version, determining model metadata for the request that defines at least one model and at least one model version, forwarding the prediction request to the at least one model with the at least one model version, and returning a prediction from the at least one model to a requestor.

Abstract: Methods and systems for multi-model scoring in a multi-tenant system are presented. A request for a machine learning application is received from a tenant application. A tenant identifier that identifies one of the multiple tenants is determined. Based on the tenant identifier and a type of the machine learning application, a first and a second machine learning models are determined. The first machine learning model was generated based on a first training data set associated with the tenant identifier. The second machine learning model that was generated based on a second training data set associated with the tenant identifier. A flow of operations that includes running the first and second machine learning models with data related to the request is executed to obtain a scoring result. The scoring result is returned to the tenant application in response to the request.

Abstract: Methods and systems for machine learning inferencing based on directed acyclic graphs are presented. A request for a machine learning application is received from a tenant application. A tenant identifier that identifies one of the tenants is determined from the request. Based on the tenant identifier and a type of the machine learning application, configuration parameters and a graph structure are determined. The graph structure defines a flow of operations for the machine learning application. Nodes of the graph structure are executed based on the configuration parameters to obtain a scoring result. Execution of a node causes a machine learning model generated for the first tenant to be applied to data related to the request. The scoring result is returned in response to the request.

Abstract: A method by a router component in a multi-tenant on-demand serving infrastructure to route scoring requests to scoring containers. The method includes receiving a scoring request, determining a machine learning application associated with the scoring request, determining whether a router instance for the machine learning application exists, and responsive to a determination that a router instance for the machine learning application does not exist, obtaining a configuration object for the machine learning application and instantiating the router instance for the machine learning application based on the configuration object for the machine learning application. The method further includes invoking the router instance for the machine learning application to route the scoring request associated with the machine learning application to a scoring container that provides scoring functionality for the machine learning application.

Abstract: A method performs service discovery in a machine learning service. The method includes detecting initialization of at least one service container, identifying label information in the at least one service container, collecting the label information for the initializing at least one service container, and storing the label information in a routing information storage to enable routing of requests to the at least one service container.

Abstract: A method by one or more electronic devices to provision an infrastructure for a machine learning application in a multi-tenant on-demand serving infrastructure. The method includes storing a plurality of templates, wherein each of the plurality of templates indicates a scoring interface, a web server, a definition of a continuous integration pipeline, and a definition of a continuous deployment pipeline, receiving a request to provision the infrastructure for the machine learning application using a specified template from the plurality of templates, and provisioning the infrastructure for the machine learning application using the specified template to create a version control system repository, a continuous integration pipeline, and a continuous deployment pipeline.

Abstract: A method by one or more electronic devices for spinning up a scoring container on demand. The method includes receiving, from an orchestrator component via an application programming interface (API), a request to spin up the scoring container, wherein the scoring container is configured to provide scoring functionality, spinning up the scoring container responsive to receiving the request to spin up the scoring container, and providing, to the orchestrator component via the API, a response to the request to spin up the scoring container, wherein the response includes a uniform resource locator (URL) to use to submit scoring requests to the scoring container.

Abstract: A method of a base scorer in a scoring service container includes sending a model identifier to a model loader of an application specific scorer in the scoring service container, receiving a model object from the model loader in response to sending the model identifier, sending a request for a scoring from a client application to a scoring function of the application specific scorer, receiving the scoring from the application specific scorer, and returning the scoring to the client application.

Abstract: Machine learning version management method for a prediction service includes receiving a prediction request, determining application metadata for the request that defines routing logic and a machine learning framework version, determining model metadata for the request that defines at least one model and at least one model version, forwarding the prediction request to the at least one model with the at least one model version, and returning a prediction from the at least one model to a requestor.

Abstract: A machine learning serving infrastructure implementing a method of receiving or detecting an update of container metrics including resource usage and serviced requests per model or per container, processing the container metrics per model or per container to determine recent resource usage and serviced requests per model or per container, and rebalancing distribution of models to a plurality of containers to decrease a detected load imbalance between containers or a stressed container in the plurality of containers.

Abstract: Using container information to select containers for executing models is described. A system receives a request from an application and identifies a version of a machine-learning model associated with the request. The system identifies a set of each serving container corresponding to the machine-learning model from a cluster of available serving containers associated with the version of the machine-learning model. The system selects a serving container from the set of each serving container corresponding to the machine-learning model. If the machine-learning model is not loaded in the serving container, the system loads the machine-learning model in the serving container. If the machine-learning model is loaded in the serving container, the system executes, in the serving container, the machine-learning model on behalf of the request. The system responds to the request based on executing the machine-learning model on behalf of the request.

Abstract: Using container and model information to select containers for executing models is described. A system receives a request from an application and identifies a version of a machine-learning model associated with the request. The system identifies model information associated with machine learning models corresponding to a cluster of available serving containers associated with the version of the machine-learning model. The system uses the model information to select a serving container from the cluster of available serving containers. If the machine-learning model is not loaded in the serving container, the system loads the machine-learning model in the serving container. If the machine-learning model is loaded in the serving container, the system executes, in the serving container, the machine-learning model on behalf of the request. The system responds to the request based on executing the machine-learning model on behalf of the request.

Abstract: The present invention provides an interferometric method and system to measure the height profile of reflecting objects with respect of a reference surface. The method comprises obtaining an image of the object along a specular reflection axis, wherein the image corresponds to an intensity pattern projected on the object along a projection axis, and wherein the specular reflection axis corresponds to a direction along which a portion of the intensity pattern is specularly reflected by the object. Then the method comprises calculating an object phase using the image and determining the height profile using the object phase and a reference phase associated to the reference surface.

Abstract: A method and a system for measuring the relief of an object are described herein. The system includes a grid projecting for projecting a grid, an image acquisition apparatus that includes a camera, and a computer. Providing a reference object having common elements with the object to measure, the method includes the steps of a) positioning the grid at three different known positions relative to the camera and the common elements; b) for each position of the grid, projecting the grid unto the reference object and, with the camera, taking an image of the reference object to yield three images having values for each pixel of the camera and c) computing the reference object phase for each pixel using the three reference object intensity values for the corresponding pixel. Steps a), b) and c) are repeated by replacing the reference object by the object to be measured.