CN114594893A - Performance analysis method and device, electronic equipment and computer readable storage medium - Google Patents
Performance analysis method and device, electronic equipment and computer readable storage medium Download PDFInfo
- Publication number
- CN114594893A CN114594893A CN202210050766.5A CN202210050766A CN114594893A CN 114594893 A CN114594893 A CN 114594893A CN 202210050766 A CN202210050766 A CN 202210050766A CN 114594893 A CN114594893 A CN 114594893A
- Authority
- CN
- China
- Prior art keywords
- task
- user
- performance analysis
- interface
- configuration
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/01—Input arrangements or combined input and output arrangements for interaction between user and computer
- G06F3/048—Interaction techniques based on graphical user interfaces [GUI]
- G06F3/0484—Interaction techniques based on graphical user interfaces [GUI] for the control of specific functions or operations, e.g. selecting or manipulating an object, an image or a displayed text element, setting a parameter value or selecting a range
- G06F3/04847—Interaction techniques to control parameter settings, e.g. interaction with sliders or dials
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/30—Monitoring
- G06F11/3003—Monitoring arrangements specially adapted to the computing system or computing system component being monitored
- G06F11/302—Monitoring arrangements specially adapted to the computing system or computing system component being monitored where the computing system component is a software system
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/30—Monitoring
- G06F11/34—Recording or statistical evaluation of computer activity, e.g. of down time, of input/output operation ; Recording or statistical evaluation of user activity, e.g. usability assessment
- G06F11/3466—Performance evaluation by tracing or monitoring
- G06F11/3476—Data logging
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/36—Preventing errors by testing or debugging software
- G06F11/3668—Software testing
- G06F11/3672—Test management
- G06F11/3684—Test management for test design, e.g. generating new test cases
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/36—Preventing errors by testing or debugging software
- G06F11/3668—Software testing
- G06F11/3672—Test management
- G06F11/3688—Test management for test execution, e.g. scheduling of test suites
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/36—Preventing errors by testing or debugging software
- G06F11/3668—Software testing
- G06F11/3672—Test management
- G06F11/3692—Test management for test results analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/36—Preventing errors by testing or debugging software
- G06F11/3668—Software testing
- G06F11/3696—Methods or tools to render software testable
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Quality & Reliability (AREA)
- Computer Hardware Design (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Human Computer Interaction (AREA)
- Debugging And Monitoring (AREA)
Abstract
The application discloses a performance analysis method and device, electronic equipment and a computer-readable storage medium. The method comprises the following steps: and displaying a task creation interface and an operation environment configuration interface to a user, receiving task configuration and operation environment configuration respectively specified by the user through the two interfaces so as to operate the task configured by the user in a specified operation environment, analyzing the task in the operation process of the task, generating a performance analysis log, and finally displaying the performance analysis interface to the user. The embodiment of the application greatly simplifies the complex task and environment configuration process in the prior art, reduces the number of manual operations required by a user, can automatically run the task in the environment specified by the user after the configuration is completed, does not need the user to manually input an execution command, greatly improves the efficiency of performance analysis, and also facilitates the user to timely adjust the parameter configuration and the hardware configuration according to the performance analysis result and carry out repeated iteration, so that the adjustment can be conveniently carried out.
Description
Technical Field
The present application relates to the field of computing technologies, and in particular, to a performance analysis method and apparatus, an electronic device, and a computer-readable storage medium.
Background
With the development of the artificial intelligence technology, the shadow of the artificial intelligence technology can be seen in more and more life and work scenes, so that the convenience of daily life of people is improved, and the work efficiency of people is improved. As the core of artificial intelligence technology, artificial intelligence, especially machine learning models, have become the most critical factor in artificial intelligence scenario development. Therefore, in the existing artificial intelligence development process, a large number of artificial intelligence or models have been proposed for use by technical developers. However, different artificial intelligence or models usually have different parameter requirements and also differ in the provided performance, so that it is necessary for developers to select an appropriate artificial intelligence or model for a target scene to be developed to be able to achieve a desired user experience.
In terms of selection of such artificial intelligence or models, a developer is usually required to have a very deep knowledge about the model, but as the artificial intelligence technology develops, the number of the artificial intelligence or models is increased, and the structure of the artificial intelligence or models is more and more complex, so that the prior art proposes to allow the developer to test various models according to specific scene requirements, so that the developer can know which are more suitable for the current target scene in turn according to the test result, and such a test scheme also helps the developer to have a more deep understanding about the tested or models, so that the model or models can be further improved based on the test result to improve the performance of the model or model.
Disclosure of Invention
The embodiment of the application provides a performance analysis method and device, electronic equipment and a computer-readable storage medium, so as to overcome the defect that performance analysis needs to be performed manually in the prior art.
In order to achieve the above object, an embodiment of the present application provides a performance analysis method, including:
the method comprises the steps that a task creating interface is displayed for a user, wherein the task creating interface comprises a first input control for the user to input a task configuration request, and the task is created for executing machine learning model calculation;
receiving a task configuration request input by a user through the first input control;
displaying a task running environment configuration interface to a user, wherein the task running environment configuration interface comprises a second input control for inputting running environment configuration parameters by the user;
receiving an operating environment configuration parameter input by a user through the task operating environment configuration interface, and creating a task operating environment for operating the task configured according to the task configuration request according to the operating environment configuration parameter;
performing analysis processing on a task running in the task running environment to generate a performance analysis log;
and displaying a performance analysis interface to a user, and displaying at least the performance analysis result on the performance analysis interface, wherein the performance analysis result is generated according to the performance analysis log.
The embodiment of the present application further provides a performance analysis apparatus, including:
the task creating interface comprises a first input control for inputting a task configuration request by a user, wherein the task is created by executing machine learning model calculation;
the first receiving module is used for receiving a task configuration request input by a user through the first input control;
the second display module is used for displaying a task running environment configuration interface to a user, wherein the task running environment configuration interface comprises a second input control for inputting running environment configuration parameters by the user;
the second receiving module is used for receiving the running environment configuration parameters input by the user through the task running environment configuration interface and creating a task running environment for running the task configured according to the task configuration request according to the running environment configuration parameters;
the analysis module is used for executing analysis processing on the tasks running in the task running environment to generate a performance analysis log;
and the third display module is used for displaying a performance analysis interface to a user and displaying at least the performance analysis result on the performance analysis interface, wherein the performance analysis result is generated according to the performance analysis log.
An embodiment of the present application further provides an electronic device, including:
a memory for storing a program;
and the processor is used for operating the program stored in the memory, and the program executes the performance analysis method provided by the embodiment of the application when running.
Embodiments of the present application further provide a computer-readable storage medium on which a computer program executable by a processor is stored, where the program, when executed by the processor, implements the performance analysis method provided by the embodiments of the present application.
The performance analysis method and device, the electronic device and the computer-readable storage medium provided by the embodiment of the application run the tasks configured by the user in the specified running environment by showing the task creation interface and the running environment configuration interface to the user and receiving the task configuration and the running environment configuration respectively specified by the user through the two interfaces, and simultaneously analyze the tasks in the running process of the tasks to generate the performance analysis log and finally show the performance analysis interface to the user, so that the user can create the tasks by inputting the configuration request on the task creation interface and input the running environment configuration parameters on the running environment interface to create the running environment to run the tasks, thereby greatly simplifying the fussy task and environment configuration process in the prior art, reducing the number of manual operations required by the user, and automatically running the tasks in the environment specified by the user after the configuration is completed, the user does not need to input an execution command manually, the efficiency of performance analysis is greatly improved, and the user can conveniently adjust the parameter configuration and the hardware configuration in time according to the performance analysis result and carry out repeated iteration, so that the adjustment can be carried out conveniently.
The above description is only an overview of the technical solutions of the present application, and the present application may be implemented in accordance with the content of the description so as to make the technical means of the present application more clearly understood, and the detailed description of the present application will be given below in order to make the above and other objects, features, and advantages of the present application more clearly understood.
Drawings
Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments. The drawings are only for purposes of illustrating the preferred embodiments and are not to be construed as limiting the application. Also, like reference numerals are used to refer to like parts throughout the drawings. In the drawings:
fig. 1 is a schematic view of an application scenario of a performance analysis scheme provided in an embodiment of the present application;
FIG. 2 is a flow diagram of one embodiment of a performance analysis method provided herein;
FIG. 3 is a flow diagram of one embodiment of a performance analysis method provided herein;
FIG. 4 is a schematic structural diagram of an embodiment of a performance analysis apparatus provided herein;
fig. 5 is a schematic structural diagram of an embodiment of an electronic device provided in the present application.
Detailed Description
Exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be embodied in various forms and should not be limited by the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.
Example one
The scheme provided by the embodiment of the application can be applied to any cluster system with analysis capability, such as a platform system comprising a server with analysis function and the like. Fig. 1 is a schematic view of an application scenario of a performance analysis scheme provided in an embodiment of the present application, and the scenario shown in fig. 1 is only one example to which the technical scheme of the present application is applicable.
With the development of the artificial intelligence technology, the shadow of the artificial intelligence technology can be seen in more and more life and work scenes, so that the convenience of daily life of people is improved, and the work efficiency of people is improved. As the core of artificial intelligence technology, artificial intelligence, especially machine learning models, have become the most critical factor in artificial intelligence scenario development. Therefore, in the existing artificial intelligence development process, a large number of artificial intelligence or models have been proposed for use by technical developers. However, different artificial intelligence or models usually have different parameter requirements and also differ in the provided performance, so that it is necessary for developers to select an appropriate artificial intelligence or model for a target scene to be developed to be able to achieve a desired user experience.
In terms of selection of such artificial intelligence or models, a developer is usually required to have a very deep knowledge about the model, but as the artificial intelligence technology develops, the number of the artificial intelligence or models is increased, and the structure of the artificial intelligence or models is more and more complex, so that the prior art proposes to allow the developer to test various models according to specific scene requirements, so that the developer can know which are more suitable for the current target scene in turn according to the test result, and such a test scheme also helps the developer to have a more deep understanding about the tested or models, so that the model or models can be further improved based on the test result to improve the performance of the model or model.
The performance of the execution due to the artificial intelligence model also depends on the configuration of the execution hardware. For example, if the hardware configuration for executing the AI model or the AI model is different, the actual performance of the AI model or the AI model may also be different, so that through the above test, the technician may also be helped to know the performance difference of the AI model or the AI model under different hardware configurations, or know the hardware component that has the largest influence on the AI model or the AI model, that is, the hardware bottleneck, so that the technician may select the appropriate hardware configuration or optimize the current hardware configuration to improve the performance of the AI model or the AI model.
In the existing AI models or test schemes, manual operations are mostly required by developers. For example, a person who needs to perform a test first configures a tool environment capable of performing performance analysis on the model, then needs to prepare a data source for providing test data for the test, the model or the like to be tested by himself, and then can start the test, and needs to manage a generated performance analysis file by himself after the test is completed, and even needs to operate a viewing tool of the performance analysis file by himself to view an analysis result, and if further optimization is to be performed according to a current analysis result, then needs to iterate the above steps by himself to perform the optimization and the test repeatedly until a final suitable parameter or hardware configuration of the model or the like is determined. The manual testing scheme not only has complex and complicated testing process and needs a large amount of manual participation, but also has low testing efficiency.
Therefore, the application provides a performance analysis scheme for the task, which can create the task according to the task parameters specified by the user and configure hardware resources for the task according to the resource requirements specified by the user, so as to perform performance analysis of the target according to the execution process of the task.
For example, the performance analysis scheme of the embodiment of the application can be applied to an analysis platform facing artificial intelligence AI and various scenes. The method may be particularly suitable for tasks created by performing machine learning model calculations, such as deep learning model calculations or reinforcement learning model calculations. The platform may provide support for both software and hardware configurations. In other words, the platform may store various models, and may also allow a user to upload a model that the user wants to test, and the platform may also invoke a hardware platform or, for example, a command channel to distribute tasks created for a target to nodes configured according to the user's specified resource requirements for execution. In particular, in the present embodiment, the assigned tasks may be performed in containers on the servers of the nodes. Therefore, the analysis performed during the execution of the task is stored in the server in real time and can also be transmitted to the file system in real time.
For example, a platform system to which the performance analysis scheme according to the embodiment of the present application is applied may perform performance analysis for a task using a GPU cluster. During performance analysis, tasks executed in a cluster mode can be analyzed based on a DLProf (Deep Learning Profiler) technology, performance analysis files can be generated in the process, a platform system can be used for filing the performance analysis files generated in the task execution process in real time, and the performance analysis files can be converted according to the analysis requirements of a user, so that the performance analysis results corresponding to distributed tasks specified by the user can be displayed to the user through page display after the performance analysis is completed. Therefore, a user can input the parameter configuration and resource requirements of the task by logging in a platform system to which the performance analysis scheme of the embodiment of the application is applied, for example, in the form of a webpage, so that the performance analysis result can be viewed through the webpage after the task is executed.
Therefore, in the performance analysis scenario shown in fig. 1, a user may access the platform system to which the performance analysis scheme according to the embodiment of the present application is applied in the form of a Web page through, for example, a Web service, so that each parameter configuration of a desired task may be submitted to the performance analysis server of the platform system through the Web page. For example, the platform may present commonly used parameter items on a web page for a task for a user to enter their own configuration. For example, a basic information page about a task that the user wants to execute may be displayed in a system page that the user accesses, and in this page, the user may input a task name about the task to be executed this time, so as to query to view the task execution condition and the corresponding performance analysis file during and after the task execution. The user can also input the type of the task and the configuration information of the data source through the page, and can also specify the command to be executed when executing the task. Finally, the user can also perform task configuration through a task creation button set on the page, so the platform system can create the task for the user based on the parameters of the task input by the user on the page.
After a task is created, the user also needs to be configured with the resources needed to perform the task. For example, if a user wants to know a certain performance, a task needs to be specified and a suitable hardware configuration needs to be selected to run the task to complete the test. Thus, similar to the parameter configuration of a task, a user may specify the hardware configuration to run the task through web access. For example, a resource configuration page may be presented to the user to allow the user to specify the number of nodes required to run the task, the hardware configuration of the nodes, and the image files loaded on the nodes. In the embodiment of the application, a plurality of performance analysis images can be provided for a user to allow the user to configure a system environment for running tasks by selecting different performance analysis images. The performance analysis image may be a variety of operating systems or systems that have applications installed. In addition, the user may also specify the hardware configuration of the node by running the environment configuration interface, for example, the user may set a hardware configuration scheme of 4 CPUs (cores), 4GB memories, and 1 GPU for a task that the user wants to execute by directly inputting a desired hardware configuration parameter on the interface.
In addition, the performance analysis scheme according to the embodiment of the application can be applied to various distributed scheduling platforms, and also supports the docking of various command channels, so that tasks created for users can be distributed to a plurality of working nodes according to the node configuration specified by the users through the platforms or the command channels.
After the task creation and the resource configuration of the running task are completed, the analysis command, for example, a DLProf command, may be added in the process of execution to analyze the task and generate a corresponding performance analysis file in the process of execution of the task.
Therefore, after the task is executed, the performance analysis result can be generated according to the performance analysis file obtained by using, for example, the DLProf command during the execution process, and is similarly displayed to the user on the page, so that in the embodiment of the present application, the operation is performed through the page from the execution of the task to the display of the performance analysis result, and the user does not need to manually execute the command.
In addition, in the embodiment of the present application, DLProf, which supports a general tenserflow framework and a pyrtch framework, can be used to analyze the task, thereby allowing a user to perform flexible analysis.
In addition, in the embodiment of the present application, an adjustment suggestion may also be generated according to, for example, a DLProf command, an analysis file of a task in the execution process of the task, for example, a parameter configuration adjusted by a user may be suggested, or a hardware configuration for running the task may be adjusted, so that the user may obtain a relevant suggestion while viewing a performance analysis result through a web page.
Therefore, in the embodiment of the present application, a user may complete execution and performance analysis processing on a distributed cluster on a system platform to which the performance analysis method of the embodiment of the present application is applied, without excessive manual operations, and finally may view a corresponding performance analysis result and further adjustment suggestions on a page in a web page access manner.
Therefore, the performance analysis scheme provided by the embodiment of the application runs the tasks configured by the user in the specified running environment by showing the task creation interface and the running environment configuration interface to the user and receiving the task configuration and the running environment configuration respectively specified by the user through the two interfaces, analyzes the tasks in the running process of the tasks to generate the performance analysis log, and finally shows the performance analysis interface to the user, so that the user can create the tasks by inputting the configuration parameters on the task creation interface and input the running environment configuration parameters on the running environment interface to create the running environment to run the tasks, thereby greatly simplifying the complicated task and environment configuration process in the prior art, reducing the number of manual operations required by the user, and automatically running the tasks in the environment specified by the user after the configuration is completed, the user does not need to input an execution command manually, the efficiency of performance analysis is greatly improved, and the user can conveniently adjust the parameter configuration and the hardware configuration in time according to the performance analysis result and carry out repeated iteration, so that the adjustment can be carried out conveniently.
The above embodiments are illustrations of technical principles and exemplary application frameworks of the embodiments of the present application, and specific technical solutions of the embodiments of the present application are further described in detail below through a plurality of embodiments.
Example two
Fig. 2 is a flowchart of an embodiment of a performance analysis method provided in the present application, where an execution subject of the method may be various terminal or server devices with execution capability, or may be a device or chip integrated on these devices. As shown in fig. 2, the performance analysis method may include the steps of:
s201, displaying a task creation interface to a user.
In this embodiment of the present application, a task creation interface may be presented to the user in step S201, and at least a first input control for the user to input a task configuration request may be included on the interface, so that the user may input configuration parameters of a task that the user wants to create through the input control. For example, in the scenario shown in fig. 1, a user may access a task creation page through a web page, and basic information items about a task that the user wants to perform may be displayed in the page.
S202, receiving a task configuration request input by a user through a first input control.
In step S202, a task configuration parameter input by a user through an input control on the interface presented in step S201 may be received. The user can input the task name of the task to be executed at this time through the first input control on the displayed interface, so that the user can conveniently inquire to check the task execution condition and the corresponding performance analysis file in the task execution process and when the task execution is finished. The user may also enter the type of task and configuration information for the data source, and may also specify commands that need to be performed when performing the task. These user-entered basic information may be configured as parameters specified by the creation task. The user may also instruct to create a task by clicking a task creation button provided on the page.
And S203, displaying a task running environment configuration interface to a user.
After the task is configured for the user in step S202, a runtime environment configuration interface may be further presented to the user in step S203, and a second input control for inputting runtime environment configuration parameters by the user may be included on the interface, so that the user may input the configuration parameters for the runtime environment that the task wants to run through the input control.
For example, if a user wants to know the performance of a task, the task needs to be specified to select a suitable hardware configuration and system environment to run the task to complete the test. Thus, in step S203, the user may input the hardware configuration to run the task through web access. For example, in step S203, a running environment configuration page may be displayed for the user, allowing the user to input, for example, the number of nodes required to run the task, the hardware configuration of the node, the system image loaded on the node, and the like on the page.
Further, in step S203, a user may be allowed to select multiple performance analysis images on the interface to allow the user to select different operating system environments by selecting different performance analysis images.
And S204, receiving the running environment configuration parameters input by the user through the task running environment configuration interface, and creating the task running environment for running the task according to the running environment configuration parameters.
In step S204, the environment configuration parameters input by the user through the input control in the interface presented in step S203 may be received. For example, as described above, the user may input the hardware configuration and the system configuration for running the task in step S204. For example, a user may set a hardware configuration scheme of 4 CPUs (cores), 4GB memories, and 1 GPU for a task he wants to execute by directly inputting a desired hardware configuration parameter on the interface. And the user may also specify the system environment to run the task by selecting from the system images provided on the interface.
S205, an analysis process is performed on the task running in the task running environment to generate a performance analysis log.
Therefore, in step S205, the task configured in step S203 can be executed in the environment according to the execution environment configuration input in step S204, and the task is subjected to analysis processing during execution, thereby generating a performance analysis log file. For example, the performance analysis scheme according to the embodiment of the present application may be applied to various distributed scheduling platforms, and also supports interfacing various command channels, so that in step S205, a task created for a user may be distributed to multiple working nodes according to a node configuration specified by the user through the platform or the command channels.
In step S205, for example, an analysis command may be added during the execution process, for example, a DLProf command may be added to analyze the task during the execution process of the task and generate a corresponding performance analysis file.
And S206, displaying a performance analysis interface to a user.
After the task is completed, a performance analysis interface may be presented to the user in step S206 to present the performance analysis results. For example, the performance analysis result is generated from a performance analysis file obtained by using, for example, a DLProf command during execution, and is similarly displayed to the user on a page, so that in the embodiment of the present application, from the execution of the task to the display of the performance analysis result, the operation is performed through the page without the user manually executing the command.
Therefore, the performance analysis method provided by the embodiment of the application runs the tasks configured by the user in the specified running environment by showing the task creation interface and the running environment configuration interface to the user and receiving the task configuration and the running environment configuration respectively specified by the user through the two interfaces, simultaneously analyzes the tasks in the running process of the tasks to generate the performance analysis log, and finally shows the performance analysis interface to the user, so that the user can create the tasks by inputting the configuration parameters on the task creation interface and input the running environment configuration parameters on the running environment interface to create the running environment to run the tasks, thereby greatly simplifying the complicated task and environment configuration process in the prior art, reducing the number of manual operations required by the user, and automatically running the tasks in the environment specified by the user after the configuration is completed, the user does not need to input an execution command manually, the efficiency of performance analysis is greatly improved, and the user can conveniently adjust the parameter configuration and the hardware configuration in time according to the performance analysis result and carry out repeated iteration, so that the adjustment can be carried out conveniently.
EXAMPLE III
Fig. 3 is a flowchart of an embodiment of a performance analysis method provided in the present application, where an execution subject of the method may be various terminal or server devices with distributed computing capability, or may be a device or chip integrated on these devices. As shown in fig. 3, the performance analysis method may include the steps of:
s301, displaying a task creation interface to a user.
In this embodiment of the present application, a task creation interface may be presented to the user in step S301, and at least a first input control for the user to input task configuration parameters may be included on the interface, so that the user may input the configuration parameters of a task that the user wants to create through the input control. For example, in the scenario shown in fig. 1, a user may access a task creation page through a web page, and basic information items about a task that the user wants to perform may be displayed in the page.
In addition, in the embodiment of the present application, the user may also create a performance analysis item according to a scenario, and therefore, in step S301, an input control of the item configuration parameter input by the user may also be included in the interface. For example, the user may be prompted to enter name information, such as a project name, and may also be prompted to enter a scenario to be applied to be analyzed.
S302, receiving a task configuration request input by a user through a first input control.
In step S302, a task configuration parameter input by a user through the input control on the interface presented in step S301 may be received. The user can input the task name of the task to be executed at this time through the first input control on the displayed interface, so that the user can conveniently inquire to check the task execution condition and the corresponding performance analysis file in the task execution process and when the task execution is finished. The user may also enter the type of task and configuration information for the data source, and may also specify commands that need to be performed when performing the task. These user-entered basic information may be configured as parameters specified by the creation task. The user may also instruct to create a task by clicking a task creation button provided on the page.
Further, in the case where the user inputs the scene configuration is also displayed in step S301, the user may also further input, for example, a name of the item and scene information to be applied to the interface in step S302 to configure the item.
And S303, displaying a task operation environment configuration interface to a user.
After the task is configured for the user in step S302, a runtime environment configuration interface may be further presented to the user in step S303, and a second input control for inputting runtime environment configuration parameters by the user may be included on the interface, so that the user may input the configuration parameters for the runtime environment that the task wants to run through the input control.
For example, if a user wants to know the performance of a task, the task needs to be specified to select a suitable hardware configuration and system environment to run the task to complete the test. Thus, in step S303, the user may input the hardware configuration to run the task through web access. For example, in step S303, a running environment configuration page may be displayed for the user, allowing the user to input, for example, the number of nodes required to run the task, the hardware configuration of the node, the system image loaded on the node, and the like on the page.
Further, at step S303, a user may be allowed to select multiple performance analysis images on the interface to allow the user to select different operating system environments by selecting different performance analysis images.
S304, receiving the operation environment configuration parameters input by the user through the task operation environment configuration interface, and creating the task operation environment for operating the task according to the operation environment configuration parameters.
In step S304, the environment configuration parameters input by the user through the input control in the interface presented in step S303 may be received. For example, as described above, the user may input the hardware configuration and the system configuration to run the task in step S304. For example, a user may set a hardware configuration scheme of 4 CPUs (cores), 4GB memories, and 1 GPU for a task he wants to execute by directly inputting a desired hardware configuration parameter on the interface. And the user may also specify the system environment to run the task by selecting from the system images provided on the interface.
S305, an analysis process is performed on the task running in the task running environment to generate a performance analysis log.
Therefore, in step S305, the task configured in step S303 may be executed in the operating environment configuration input in step S304 according to the environment configuration, and the task is subjected to analysis processing during execution, thereby generating a performance analysis log file. For example, the performance analysis scheme according to the embodiment of the present application may be applied to various distributed scheduling platforms, and also supports interfacing various command channels, so that in step S305, tasks created for a user may be distributed to a plurality of working nodes according to a node configuration specified by the user through the platform or the command channels.
In step S305, the task may be analyzed and a corresponding performance analysis file may be generated during the execution of the task, for example, by adding an analysis command during the execution, for example, a DLProf command may be added.
And S306, recording an execution log in the running process of the task.
On the basis that the performance analysis log is generated in step S305, an execution log may be further recorded in step S306. In the embodiment of the present application, the execution log may be a log file such as a server log of a node for executing the task, and thus may be used to generate a performance analysis result together with the performance analysis log recorded in step S305.
And S307, displaying a performance analysis interface to a user.
After the task is completed, a performance analysis interface may be presented to the user in step S307 to present the performance analysis results. For example, the performance analysis result is generated from a performance analysis file obtained by using, for example, a DLProf command during execution, and is similarly displayed to the user on a page, so that in the embodiment of the present application, from the execution of the task to the display of the performance analysis result, the operation is performed through the page without the user manually executing the command.
And S308, generating an adjustment suggestion of the node configuration and/or the parameter configuration according to the performance analysis result.
In step S308, an adjustment suggestion may be generated according to the performance analysis file obtained in step S305, for example, a parameter configuration adjusted by the user may be suggested, or an environment configuration for running the task may be adjusted, so that the user may adjust the parameter in the next iteration according to the suggestion.
S309, displaying an adjustment suggestion interface to the user.
In step S309, the user may be further allowed to view the adjustment suggestion related to the performance analysis result through the web page. For example, an adjustment suggestion interface may be presented to the user in step S309, and the adjustment suggestion generated in step S308 may be displayed on the interface, so that the user may adjust the task configuration parameters and/or the environment parameters used in the next iteration with reference to the adjustment suggestion and the performance analysis result displayed in step S307.
Therefore, the performance analysis method provided by the embodiment of the application runs the tasks configured by the user in the specified running environment by showing the task creation interface and the running environment configuration interface to the user and receiving the task configuration and the running environment configuration respectively specified by the user through the two interfaces, simultaneously analyzes the tasks in the running process of the tasks to generate the performance analysis log, and finally shows the performance analysis interface to the user, so that the user can create the tasks by inputting the configuration parameters on the task creation interface and input the running environment configuration parameters on the running environment interface to create the running environment to run the tasks, thereby greatly simplifying the complicated task and environment configuration process in the prior art, reducing the number of manual operations required by the user, and automatically running the tasks in the environment specified by the user after the configuration is completed, the user does not need to input an execution command manually, the efficiency of performance analysis is greatly improved, and the user can conveniently adjust the parameter configuration and the hardware configuration in time according to the performance analysis result and carry out repeated iteration, so that the adjustment can be carried out conveniently.
Example four
Fig. 4 is a schematic structural diagram of an embodiment of a performance analysis apparatus provided in the present application, which may be used to execute the performance analysis method shown in fig. 2 or fig. 3. As shown in fig. 4, the apparatus may include: first display module 41, first receiving module 42, second display module 43, second receiving module 44, analyzing module 45, and third display module 46.
The first presentation module 41 may be used to present a task creation interface to a user.
In this embodiment, the first presentation module 41 may present a task creation interface to the user, and at least a first input control for the user to input task configuration parameters may be included on the interface, so that the user may input the configuration parameters of the task that the user wants to create through the input control. For example, in the scenario shown in fig. 1, a user may access a task creation page through a web page, and basic information items about a task that the user wants to perform may be displayed in the page.
In addition, in the embodiment of the present application, the user may also create a performance analysis project according to a scenario, and therefore, the input control of the project configuration parameter input by the user may also be included on the interface through the first presentation module 41. For example, the user may be prompted to enter name information, such as a project name, and may also be prompted to enter a scenario to be applied to be analyzed.
The first receiving module 42 may be configured to receive a task configuration request input by a user through the first input control.
The first receiving module 42 may receive the task configuration parameters input by the user through the input control on the interface presented by the first presenting module 41. The user can input the task name of the task to be executed at this time through the first input control on the displayed interface, so that the user can conveniently inquire to check the task execution condition and the corresponding performance analysis file in the task execution process and when the task execution is finished. The user may also enter the type of task and configuration information for the data source, and may also specify commands that need to be performed when performing the task. These user-entered basic information may be configured as parameters specified by the creation task. The user may also instruct to create a task by clicking a task creation button provided on the page.
In addition, in the case where the first presentation module 41 also displays a user input scene configuration, the user may further input, for example, a name of the item and scene information to be applied to configure the item in the interface.
The second presentation module 43 may be used to present the task execution environment configuration interface to the user.
After the user configures the task, the second presentation module 43 may further present the runtime environment configuration interface to the user, on which a second input control for inputting the runtime environment configuration parameters by the user may be included, so that the user can input the configuration parameters for the runtime environment in which the task is intended to run through the input control.
For example, if a user wants to know the performance of a task, the task may be specified to be run with the appropriate hardware configuration and system environment to complete the test. Thus, the user may enter the hardware configuration to run the task through web access. For example, the second presentation module 43 may display a runtime environment configuration page for the user, allowing the user to enter, for example, the number of nodes required to run the task, the hardware configuration of the nodes, and the system images loaded on the nodes on the page.
In addition, the second presentation module 43 presents an interface that allows a user to select multiple performance analysis images to allow a user to select different operating system environments by selecting different performance analysis images.
The second receiving module 44 may be configured to receive the running environment configuration parameters input by the user through the task running environment configuration interface, and create a task running environment for running the task according to the running environment configuration parameters.
The analysis module 45 may be used to perform analysis processing on the tasks running in the task execution environment to generate performance analysis logs.
The analysis module 45 may run the configured task in the running environment configuration input by the user according to the environment configuration, and perform analysis processing on the task during execution, thereby generating a performance analysis log file. For example, the performance analysis scheme according to the embodiment of the present application may be applied to various distributed scheduling platforms, and also supports interfacing various command channels, so that tasks created for a user may be distributed to a plurality of working nodes according to a node configuration specified by the user through the platform or the command channels.
The analysis of the task and the generation of the corresponding performance analysis file during the execution of the task may be performed, for example, by adding an analysis command during the execution, for example, a DLProf command.
In addition, the performance analysis apparatus according to the embodiment of the present application may further include a recording module 47, which may be used to record an execution log during the running process of the task.
In the case where the analysis module 45 generates the performance analysis log, the recording module 47 may further record the execution log. In the embodiment of the present application, the execution log may be a log file such as a server log of a node for executing the task, and thus may be used to generate a performance analysis result together with the performance analysis log generated by the analysis module 45.
After the task is completed, third presentation module 46 may present a performance analysis interface to the user to present the performance analysis results. For example, the performance analysis result is generated from a performance analysis file obtained by using, for example, a DLProf command during execution, and is similarly displayed to the user on a page, so that in the embodiment of the present application, from the execution of the task to the display of the performance analysis result, the operation is performed through the page without the user manually executing the command.
In addition, in the embodiment of the present application, the performance analysis apparatus may further include a suggestion generation module 48 and a fourth presentation module 49.
A fourth presentation module 49 may be used to present an adjustment suggestion interface to the user.
The performance analysis device provided by the embodiment of the application runs the tasks configured by the user in the specified running environment by showing the task creating interface and the running environment configuration interface to the user and receiving the task configuration and the running environment configuration respectively specified by the user through the two interfaces, simultaneously analyzes the tasks in the running process of the tasks to generate the performance analysis logs, and finally shows the performance analysis interface to the user, so that the user can create the tasks by inputting configuration parameters on the task creating interface and input running environment configuration parameters on the running environment interface to create the running environment to run the tasks, thereby greatly simplifying the complicated task and environment configuration process in the prior art, reducing the number of manual operations required by the user, and automatically running the tasks in the environment specified by the user after the configuration is finished, the user does not need to input an execution command manually, the efficiency of performance analysis is greatly improved, and the user can conveniently adjust the parameter configuration and the hardware configuration in time according to the performance analysis result and carry out repeated iteration, so that the adjustment can be carried out conveniently.
EXAMPLE five
The internal functions and structures of the performance analysis apparatus, which can be implemented as an electronic device, are described above. Fig. 5 is a schematic structural diagram of an embodiment of an electronic device provided in the present application. As shown in fig. 5, the electronic device includes a memory 51 and a processor 52.
The memory 51 stores programs. In addition to the above-described programs, the memory 51 may also be configured to store other various data to support operations on the electronic device. Examples of such data include instructions for any application or method operating on the electronic device, contact data, phonebook data, messages, pictures, videos, and so forth.
The memory 51 may be implemented by any type or combination of volatile or non-volatile memory devices, such as Static Random Access Memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic or optical disks.
The processor 52 is not limited to a processor (CPU), but may be a processing chip such as a Graphic Processing Unit (GPU), a Field Programmable Gate Array (FPGA), an embedded neural Network Processor (NPU), or an Artificial Intelligence (AI) chip. The processor 52 is coupled to the memory 51 and executes the program stored in the memory 51 to perform the performance analysis method of the second or third embodiment.
Further, as shown in fig. 5, the electronic device may further include: communication components 53, power components 54, audio components 55, display 56, and other components. Only some of the components are schematically shown in fig. 5, and it is not meant that the electronic device comprises only the components shown in fig. 5.
The communication component 53 is configured to facilitate wired or wireless communication between the electronic device and other devices. The electronic device may access a wireless network based on a communication standard, such as WiFi, 3G, 4G, or 5G, or a combination thereof. In an exemplary embodiment, the communication component 53 receives a broadcast signal or broadcast related information from an external broadcast management system via a broadcast channel. In an exemplary embodiment, the communication component 53 further comprises a Near Field Communication (NFC) module to facilitate short-range communication. For example, the NFC module may be implemented based on Radio Frequency Identification (RFID) technology, infrared data association (IrDA) technology, Ultra Wideband (UWB) technology, Bluetooth (BT) technology, and other technologies.
A power supply component 54 provides power to the various components of the electronic device. The power components 54 may include a power management system, one or more power sources, and other components associated with generating, managing, and distributing power for an electronic device.
The audio component 55 is configured to output and/or input an audio signal. For example, the audio component 55 includes a Microphone (MIC) configured to receive external audio signals when the electronic device is in an operational mode, such as a call mode, a recording mode, and a voice recognition mode. The received audio signal may further be stored in the memory 51 or transmitted via the communication component 53. In some embodiments, audio assembly 55 also includes a speaker for outputting audio signals.
The display 56 includes a screen, which may include a Liquid Crystal Display (LCD) and a Touch Panel (TP). If the screen includes a touch panel, the screen may be implemented as a touch screen to receive an input signal from a user. The touch panel includes one or more touch sensors to sense touch, slide, and gestures on the touch panel. The touch sensor may not only sense the boundary of a touch or slide action, but also detect the duration and pressure associated with the touch or slide operation.
Those of ordinary skill in the art will understand that: all or a portion of the steps of implementing the above-described method embodiments may be performed by hardware associated with program instructions. The foregoing program may be stored in a computer-readable storage medium. When executed, the program performs steps comprising the method embodiments described above; and the aforementioned storage medium includes: various media that can store program codes, such as ROM, RAM, magnetic or optical disks.
Finally, it should be noted that: the above embodiments are only used to illustrate the technical solution of the present invention, and not to limit the same; while the invention has been described in detail and with reference to the foregoing embodiments, it will be understood by those skilled in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some or all of the technical features may be equivalently replaced; and the modifications or the substitutions do not make the essence of the corresponding technical solutions depart from the scope of the technical solutions of the embodiments of the present invention.
Claims (10)
1. A method of performance analysis, comprising:
the method comprises the steps that a task creating interface is displayed for a user, wherein the task creating interface comprises a first input control for the user to input a task configuration request, and the task is created for executing machine learning model calculation;
receiving a task configuration request input by a user through the first input control;
displaying a task running environment configuration interface to a user, wherein the task running environment configuration interface comprises a second input control for inputting running environment configuration parameters by the user;
receiving an operating environment configuration parameter input by a user through the task operating environment configuration interface, and creating a task operating environment for operating a task according to the operating environment configuration parameter;
performing analysis processing on a task running in the task running environment to generate a performance analysis log;
and displaying a performance analysis interface to a user, and displaying at least the performance analysis result on the performance analysis interface, wherein the performance analysis result is generated according to the performance analysis log.
2. The performance analysis method of claim 1, wherein the task is a task created for performing a deep learning model calculation or a reinforcement learning model calculation.
3. The performance analysis method of claim 1, wherein the operating environment configuration parameters comprise: the number of nodes running the task, the hardware configuration of the nodes, and the image files on the nodes.
4. The performance analysis method of claim 1, wherein the method further comprises:
and recording an execution log in the running process of the task, wherein the execution log is used for generating the performance analysis result together with the performance analysis log.
5. The performance analysis method of claim 1, wherein the method further comprises:
generating an adjustment suggestion of the operating environment configuration parameters and/or the task configuration according to the performance analysis result;
presenting an adjustment suggestion interface to a user, wherein the adjustment suggestion interface displays at least the adjustment suggestion.
6. A performance analysis device, comprising:
the task creating interface comprises a first input control for inputting a task configuration request by a user, wherein the task is created by executing machine learning model calculation;
the first receiving module is used for receiving a task configuration request input by a user through the first input control;
the second display module is used for displaying a task running environment configuration interface to a user, wherein the task running environment configuration interface comprises a second input control for inputting running environment configuration parameters by the user;
the second receiving module is used for receiving the running environment configuration parameters input by the user through the task running environment configuration interface and creating the task running environment for running the task according to the running environment configuration parameters;
the analysis module is used for executing analysis processing on the tasks running in the task running environment to generate a performance analysis log;
and the third display module is used for displaying a performance analysis interface to a user and displaying at least the performance analysis result on the performance analysis interface, wherein the performance analysis result is generated according to the performance analysis log.
7. The performance analysis apparatus of claim 6, wherein the operating environment configuration parameters comprise: the number of nodes running the task, the hardware configuration of the nodes, and the image files on the nodes.
8. The performance analysis apparatus of claim 6, wherein the apparatus further comprises:
and the recording module is used for recording an execution log in the running process of the task, wherein the execution log is used for generating the performance analysis result together with the performance analysis log.
9. An electronic device, comprising:
a memory for storing a program;
a processor for executing the program stored in the memory to perform the performance analysis method of any one of claims 1 to 5.
10. A computer-readable storage medium, on which a computer program executable by a processor is stored, wherein the program, when executed by the processor, implements the performance analysis method according to any one of claims 1 to 5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210050766.5A CN114594893A (en) | 2022-01-17 | 2022-01-17 | Performance analysis method and device, electronic equipment and computer readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210050766.5A CN114594893A (en) | 2022-01-17 | 2022-01-17 | Performance analysis method and device, electronic equipment and computer readable storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114594893A true CN114594893A (en) | 2022-06-07 |
Family
ID=81806763
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210050766.5A Pending CN114594893A (en) | 2022-01-17 | 2022-01-17 | Performance analysis method and device, electronic equipment and computer readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114594893A (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109961151A (en) * | 2017-12-21 | 2019-07-02 | 同方威视科技江苏有限公司 | For the system for calculating service of machine learning and for the method for machine learning |
| US20200019888A1 (en) * | 2018-07-13 | 2020-01-16 | SigOpt, Inc. | Systems and methods for an accelerated tuning of hyperparameters of a model using a machine learning-based tuning service |
| CN111310936A (en) * | 2020-04-15 | 2020-06-19 | 光际科技(上海)有限公司 | Machine learning training construction method, platform, device, equipment and storage medium |
| CN111444019A (en) * | 2020-03-31 | 2020-07-24 | 中国科学院自动化研究所 | Cloud-end-collaborative deep learning model distributed training method and system |
-
2022
- 2022-01-17 CN CN202210050766.5A patent/CN114594893A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109961151A (en) * | 2017-12-21 | 2019-07-02 | 同方威视科技江苏有限公司 | For the system for calculating service of machine learning and for the method for machine learning |
| US20200019888A1 (en) * | 2018-07-13 | 2020-01-16 | SigOpt, Inc. | Systems and methods for an accelerated tuning of hyperparameters of a model using a machine learning-based tuning service |
| CN111444019A (en) * | 2020-03-31 | 2020-07-24 | 中国科学院自动化研究所 | Cloud-end-collaborative deep learning model distributed training method and system |
| CN111310936A (en) * | 2020-04-15 | 2020-06-19 | 光际科技(上海)有限公司 | Machine learning training construction method, platform, device, equipment and storage medium |
Non-Patent Citations (1)
| Title |
|---|
| 汤世征: "ragDL:一种易用的深度学习模型可视化构建系统", 《计算机科学》 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US7017145B2 (en) | Method, system, and program for generating a user interface | |
| CN110928529B (en) | Method and system for assisting operator development | |
| CN108304201B (en) | Object updating method, device and equipment | |
| US20170090875A1 (en) | Declarative design-time experience platform for code generation | |
| CN107209773B (en) | Automatic invocation of unified visual interface | |
| US11126938B2 (en) | Targeted data element detection for crowd sourced projects with machine learning | |
| CN106933729A (en) | A kind of method of testing and system based on cloud platform | |
| CN112988130A (en) | Visual modeling method, device, equipment and medium based on big data | |
| US11288064B1 (en) | Robotic process automation for interactive documentation | |
| CN108830383B (en) | Method and system for displaying machine learning modeling process | |
| CN111930617B (en) | Automatic test method and device based on data objectification | |
| CN112860247B (en) | Custom generation method, device, equipment and medium of model component | |
| CN113419941A (en) | Evaluation method and apparatus, electronic device, and computer-readable storage medium | |
| CN108898229B (en) | Method and system for constructing machine learning modeling process | |
| US11934420B2 (en) | Systems and methods for componentization and plug and play workflows | |
| CN110096304A (en) | Task construction method, device, equipment and storage medium based on Jenkins | |
| US20240086165A1 (en) | Systems and methods for building and deploying machine learning applications | |
| CN114764296A (en) | Machine learning model training method and device, electronic equipment and storage medium | |
| US20170102861A1 (en) | Natural Language Creation Tool for Applications, and an End User Drag and Drop Site-Building Design Canvas for Viewing and Analyzing User Adoption | |
| CN114594893A (en) | Performance analysis method and device, electronic equipment and computer readable storage medium | |
| CN114968741A (en) | Performance test method, system, equipment and medium based on scene platform | |
| CN114327709A (en) | Control page generation method and device, intelligent device and storage medium | |
| CN108960433B (en) | Method and system for running machine learning modeling process | |
| CN114201144A (en) | Micro service system construction method, device and medium based on domain-driven design | |
| KR20220046038A (en) | AI computing processor acceleration performance evaluation interface system and method therefor |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |