Patents by Inventor Saritha Vinod
Saritha Vinod has filed for patents to protect the following inventions. This listing includes patent applications that are pending as well as patents that have already been granted by the United States Patent and Trademark Office (USPTO).
-
Publication number: 20230185604Abstract: Methods, systems, and computer program products for cold-start service placement over on-demand resources are provided herein. A computer-implemented method includes obtaining a performance requirement profile comprising performance requirements of a service that vary over time; determining a plurality of incarnations for the service, wherein each incarnation is associated with a level of performance provided by the incarnation for the service, resource requirements of the incarnation, and a type of computing node the incarnation is configured to execute on; identifying computing nodes having different types and different resource capacities; jointly scheduling (i) the computing nodes and (ii) one or more of the incarnations on the computing nodes over a time interval such that a cumulative level of performance of the incarnations scheduled at each timepoint in the time interval satisfies the performance requirement profile of the service.Type: ApplicationFiled: December 15, 2021Publication date: June 15, 2023Inventors: Venkatesan Thirumalai Chakaravarthy, Ashok Pon Kumar Sree Prakash, Saritha Vinod, Yogish Sabharwal
-
Patent number: 11409564Abstract: A method, computer system, and a computer program product may perform hyperparameter tuning. Hyperparameter search input data may be received. A search algorithm to generate a set of hyperparameters may be invoked. A hyperparameter tuning resource allocator may receive physical resource data regarding a cluster. The hyperparameter tuning resource allocator may determine a required workload and may use a received number of parallel experiments to determine the required workload. The hyperparameter tuning resource allocator may determine whether an option to mandate the workload batch size was selected. The cluster may perform the number of parallel experiments to search and tune the set of hyperparameters. The hyperparameter tuning resource allocator may manage the cluster according to the physical resource data and the required workload and by using tensor swapping.Type: GrantFiled: July 31, 2020Date of Patent: August 9, 2022Assignee: INTERNATIONAL BUSINESS MACHINES CORPORATIONInventors: Saritha Vinod, Anto Ajay Raj John, Asis Kumar Patra
-
Publication number: 20220035672Abstract: A method, computer system, and a computer program product may perform hyperparameter tuning. Hyperparameter search input data may be received. A search algorithm to generate a set of hyperparameters may be invoked. A hyperparameter tuning resource allocator may receive physical resource data regarding a cluster. The hyperparameter tuning resource allocator may determine a required workload and may use a received number of parallel experiments to determine the required workload. The hyperparameter tuning resource allocator may determine whether an option to mandate the workload batch size was selected. The cluster may perform the number of parallel experiments to search and tune the set of hyperparameters. The hyperparameter tuning resource allocator may manage the cluster according to the physical resource data and the required workload and by using tensor swapping.Type: ApplicationFiled: July 31, 2020Publication date: February 3, 2022Inventors: Saritha Vinod, Anto Ajay Raj John, Asis Kumar Patra
-
Publication number: 20210158147Abstract: In an approach to determining an optimal training approach for a large deep learning model based on model characteristics and system characteristics. The one or more computer processors identify one or more model characteristics associated with a deep learning model. The one or more computer processors identify one or more system configurations associated with a system training the deep learning model. The one or more computer processors determine a training approach for the deep learning model utilizing a trained large model predictor fed with the one or more identified model characteristics and the one or more identified system configurations. The one or more computer processors train the deep learning model utilizing the determined training approach.Type: ApplicationFiled: November 26, 2019Publication date: May 27, 2021Inventors: Saritha Vinod, Yogish Sabharwal
-
Patent number: 10691419Abstract: Mechanisms are provided for generating a new compilable program from an instruction trace. The mechanisms receive a trace file for an original program whose execution on computing hardware has been traced and perform analysis of the trace file to identify a hot function, symbol information corresponding to the hot function, and initialization parameters for the hot function. The mechanisms generate a trace control flow graph based on the identified hot function and the symbol information corresponding to the hot function. The mechanisms identify, based on the trace control flow graph, pathways in the original program to the hot function, represented in the trace file. The mechanisms generate a reconstructed program based on the trace control flow graph, the pathways to the hot function, and the initialization parameters, and output the reconstructed program.Type: GrantFiled: February 11, 2019Date of Patent: June 23, 2020Assignee: International Business Machines CorporationInventors: Biplob Mishra, Saritha Vinod
-
Publication number: 20190378016Abstract: A distributed network architecture for deep learning including a model mapping table (MMT) storing information regarding respective portions of a deep learning model distributed amongst a plurality of interconnected host nodes. Respective host nodes can comprise at least one central processing unit (CPU), at least one CPU memory, at least one graphics processing unit (GPU), and at least one GPU memory. The deep learning model can be trained by receiving a request from a requesting GPU for a first portion of the deep learning model, identifying a first host node storing the first portion of the deep learning model, providing a first copy of the first portion of the deep learning model to the requesting GPU memory, performing processing on the first copy by the requesting GPU, and updating the MMT based on the processing performed on the first copy of the first portion of the deep learning model.Type: ApplicationFiled: June 7, 2018Publication date: December 12, 2019Inventors: Anto A. R. John, Saritha Vinod
-
Patent number: 10460031Abstract: Generating reports. The method may include receiving, by a computer, data associated with a meeting. The method may include converting the received data to audio text data, digital text data, and graphical object data. The method may include identifying tokens associated with the converted audio text data, the converted digital text data, and the converted graphical object data. The method may include integrating each token corresponding to timeline data, positional data, and structural data. The method may include creating semantic correlations by connecting tokens associated with the converted audio text data and tokens associated with the converted digital text data. The method may include associating each of the integrated tokens with a semantic correlation based on the timeline data. The method may include creating a report outline based on the received data. The method may include creating an integrated report using the created report outline and the associated tokens.Type: GrantFiled: September 28, 2015Date of Patent: October 29, 2019Assignee: International Business Machines CorporationInventors: Pidad G. D'Souza, Anto A. R. John, Josephine E. Justin, Saritha Vinod
-
Patent number: 10460030Abstract: Generating reports. The method may include receiving, by a computer, data associated with a meeting. The method may include converting the received data to audio text data, digital text data, and graphical object data. The method may include identifying tokens associated with the converted audio text data, the converted digital text data, and the converted graphical object data. The method may include integrating each token corresponding to timeline data, positional data, and structural data. The method may include creating semantic correlations by connecting tokens associated with the converted audio text data and tokens associated with the converted digital text data. The method may include associating each of the integrated tokens with a semantic correlation based on the timeline data. The method may include creating a report outline based on the received data. The method may include creating an integrated report using the created report outline and the associated tokens.Type: GrantFiled: August 13, 2015Date of Patent: October 29, 2019Assignee: International Business Machines CorporationInventors: Pidad G. D'Souza, Anto A. R. John, Josephine E. Justin, Saritha Vinod
-
Publication number: 20190171423Abstract: Mechanisms are provided for generating a new compilable program from an instruction trace. The mechanisms receive a trace file for an original program whose execution on computing hardware has been traced and perform analysis of the trace file to identify a hot function, symbol information corresponding to the hot function, and initialization parameters for the hot function. The mechanisms generate a trace control flow graph based on the identified hot function and the symbol information corresponding to the hot function. The mechanisms identify, based on the trace control flow graph, pathways in the original program to the hot function, represented in the trace file. The mechanisms generate a reconstructed program based on the trace control flow graph, the pathways to the hot function, and the initialization parameters, and output the reconstructed program.Type: ApplicationFiled: February 11, 2019Publication date: June 6, 2019Inventors: Biplob Mishra, Saritha Vinod
-
Patent number: 10209962Abstract: Mechanisms are provided for generating a new compilable program from an instruction trace. The mechanisms receive a trace file for an original program whose execution on computing hardware has been traced and perform analysis of the trace file to identify a hot function, symbol information corresponding to the hot function, and initialization parameters for the hot function. The mechanisms generate a trace control flow graph based on the identified hot function and the symbol information corresponding to the hot function. The mechanisms identify, based on the trace control flow graph, pathways in the original program to the hot function, represented in the trace file. The mechanisms generate a reconstructed program based on the trace control flow graph, the pathways to the hot function, and the initialization parameters, and output the reconstructed program.Type: GrantFiled: February 6, 2017Date of Patent: February 19, 2019Assignee: International Business Machines CorporationInventors: Biplob Mishra, Saritha Vinod
-
Publication number: 20180225096Abstract: Mechanisms are provided for generating a new compilable program from an instruction trace. The mechanisms receive a trace file for an original program whose execution on computing hardware has been traced and perform analysis of the trace file to identify a hot function, symbol information corresponding to the hot function, and initialization parameters for the hot function. The mechanisms generate a trace control flow graph based on the identified hot function and the symbol information corresponding to the hot function. The mechanisms identify, based on the trace control flow graph, pathways in the original program to the hot function, represented in the trace file. The mechanisms generate a reconstructed program based on the trace control flow graph, the pathways to the hot function, and the initialization parameters, and output the reconstructed program.Type: ApplicationFiled: February 6, 2017Publication date: August 9, 2018Inventors: Biplob Mishra, Saritha Vinod
-
Publication number: 20170046331Abstract: Generating reports. The method may include receiving, by a computer, data associated with a meeting. The method may include converting the received data to audio text data, digital text data, and graphical object data. The method may include identifying tokens associated with the converted audio text data, the converted digital text data, and the converted graphical object data. The method may include integrating each token corresponding to timeline data, positional data, and structural data. The method may include creating semantic correlations by connecting tokens associated with the converted audio text data and tokens associated with the converted digital text data. The method may include associating each of the integrated tokens with a semantic correlation based on the timeline data. The method may include creating a report outline based on the received data. The method may include creating an integrated report using the created report outline and the associated tokens.Type: ApplicationFiled: August 13, 2015Publication date: February 16, 2017Inventors: Pidad G. D'Souza, Anto A.R. John, Josephine E. Justin, Saritha Vinod
-
Publication number: 20170046411Abstract: Generating reports. The method may include receiving, by a computer, data associated with a meeting. The method may include converting the received data to audio text data, digital text data, and graphical object data. The method may include identifying tokens associated with the converted audio text data, the converted digital text data, and the converted graphical object data. The method may include integrating each token corresponding to timeline data, positional data, and structural data. The method may include creating semantic correlations by connecting tokens associated with the converted audio text data and tokens associated with the converted digital text data. The method may include associating each of the integrated tokens with a semantic correlation based on the timeline data. The method may include creating a report outline based on the received data. The method may include creating an integrated report using the created report outline and the associated tokens.Type: ApplicationFiled: September 28, 2015Publication date: February 16, 2017Inventors: Pidad G. D'Souza, Anto A.R. John, Josephine E. Justin, Saritha Vinod