US 12,112,148 B2Granted

Configuring machine learning models for training and deployment using graphical components

AssigneeNVIDIA Corporation

InventorsShaunak Gupte, Prashant Gaikwad, Chandrahas Jagadish Ramalad, Bhushan Rupde

Life of this patent

Estimating the term…

Loading patent…

Abstract

Embodiments of the present disclosure relate to applications and platforms for configuring machine learning models for training and deployment using graphical components in a development environment. For example, systems and methods are disclosed that relate to determining one or more machine learning models and one or more processing operations corresponding to the one or more machine learning models. Further, a model component may be generated using the one or more machine learning models, the one or more processing operations, and one or more extension libraries in which the one or more extension libraries indicate one or more deployment parameters related to the one or more machine learning models. The model component may accordingly provide data that may be used to be able to use and deploy the one or more machine learning models.

Claims18

Dependent claims

A processor comprising:

one or more processing units to:
determine, based at least in part on one or more inputs to a first graphical user interface (GUI), one or more machine learning models and one or more processing operations corresponding to the one or more machine learning models;
determine, based at least in part on the one or more machine learning models and the one or more processing operations, one or more extension libraries;
generate a model component using the one or more machine learning models, the one or more processing operations, and the one or more extension libraries;
store the model component in a component library;
receive one or more selections of the model component for inclusion in a component pipeline corresponding to an application;
receive, using a second GUI, one or more selections of an inference component corresponding to the model component for inclusion in the component pipeline;
and
compose the application based at least in part on the component pipeline, the application including the one or more machine learning models and the one or more processing operations.

6 more elements

A system comprising:

processing circuitry to:
receive, using a graphical user interface (GUI), at least one first selection corresponding to at least one machine learning model;
receive, using the GUI, at least one second selection corresponding to at least one of a pre-processing operation applied to an input of the at least one machine learning model or a post-processing operation applied to an output the at least one machine learning model;
receive, using the GUI, at least one user-generated input linking the at least one machine learning model with at least one of the pre-processing operation or the post-processing operation;
generate, based at least in part on the linking, a model component including the at least one machine learning model and at least one of the pre-processing operation or the post-processing operation;
store the model component in a component library;
receive at least one third selection corresponding to the model component for inclusion in a component pipeline corresponding to an application;
receive, using another GUI, at least one fourth selection corresponding to an inference component for executing the model component;
and
deploy an instance of the application based at least in part on the component pipeline, the instance of the application including the at least one machine learning model and at least one of the pre-processing operation or the post-processing operation in an ordered sequence.

7 more elements

Dependencies are inferred from the claim text.

Facts

Application: 17/806,056Filed 8 Jun 2022 · us provisional application
Priority: 8 Jun 2021
Published: US 2022/0391176 A18 Dec 2022
Granted: 8 Oct 2024
Term ends: Not estimatedThe source records don’t support an estimate.

Classification13 codes

G06F 8/34
Creation or generation of source code: graphical or visual programming
G06F 9/44521
Program loading or initiating: dynamic linking or loading; Link editing at or after load time, e.g. Java class loading
G06F 8/35
Creation or generation of source code: model driven
G06F 8/36
Creation or generation of source code: software reuse

Family

Application63/208,483Us provisional application · filed 8 Jun 2021
Application63/208,934Us provisional application · filed 9 Jun 2021
US 2022/0391176 A1Pre-grant publication · published 8 Dec 2022
US 12,112,148 B2This patent

From the US filing history. Foreign family members aren’t listed.

Timeline

Citations

Open elsewhere

Building an agent or pulling many patents? Use the API or MCP server instead of scraping pages. Get an API key