AI teaching experiment environment management method based on container arrangement

Abstract

The invention discloses an AI teaching experiment environment management method based on container arrangement. The AI teaching experiment environment management method comprises the steps that an AI lesson preparation environment is published into an AI teaching experiment course to form a version control module in container arrangement; under the condition of the version control module, creating and automatically selecting an optimal AI host based on an AI environment, and forming an optimal AI host module in container arrangement; creating and operating an AI lesson preparation experiment environment or an AI teaching experiment environment on the optimal AI host; judging whether all AI environments have usability or not; if yes, continuing to retain; and if the AI environment has no usability, constructing a destroying module based on a container arrangement technology to destroy the AI environment. By controlling the version of the AI teaching experiment course, a teacher can conveniently update and switch the experiment content and configuration at any time; the AI environment is created by selecting the optimal AI host, and the non-usability AI environment is released in time, so that more efficient, flexible and reliable AI environment management and utilization are realized.

Description

Technical Field

[0001] The present invention relates to an AI teaching experiment environment management method, and in particular to an AI teaching experiment environment management method based on container orchestration. Background Art

[0002] With the rapid development of artificial intelligence technology, the demand for AI teaching is growing. In AI teaching experiments, it is necessary to provide students with a diverse and stable experimental environment. However, the existing experimental environment management methods often have many problems, such as cumbersome deployment, unreasonable resource allocation, inconsistent environment, etc. The emergence of container technology provides a new idea for solving these problems. Containers are lightweight, well isolated, and highly portable, and can quickly and efficiently build and deploy experimental environments. However, pure container technology has certain limitations when facing large-scale and complex AI teaching experimental environment management. Summary of the invention

[0003] In order to solve the shortcomings of the above-mentioned technology, the present invention provides an AI teaching experiment environment management method based on container orchestration.

[0004] In order to solve the above technical problems, the technical solution adopted by the present invention is: an AI teaching experiment environment management method based on container orchestration, which specifically includes the following steps:

[0005] Step S1: Publish the AI ​​lesson preparation environment as an AI teaching experiment course to form a version control module in the container orchestration;

[0006] Step S2: In the case of the version control module, automatically select the optimal AI host based on the AI ​​environment to form the optimal AI host module in the container orchestration; the formula for automatically selecting the optimal AI host is:

[0007] I=min(X j *C j *M j );

[0008] Among them, I is the AI ​​environment, namely the AI ​​lesson preparation experiment environment and the AI ​​teaching experiment environment; X is the number of AI hosts running AI environments; C is the AI ​​host CPU utilization rate; M is the AI ​​host memory utilization rate; j is the AI ​​host number;

[0009] The AI ​​environment includes the AI ​​lesson preparation experiment environment and the AI ​​teaching experiment environment;

[0010] Step S3: creating and running an AI lesson preparation experiment environment or an AI teaching experiment environment on the optimal AI host;

[0011] Step S4: Determine in turn whether all AI environments are usable; if so, keep them; if it is determined that there is no usable AI environment, build a destruction module based on container orchestration technology to destroy the AI ​​environment.

[0012] Furthermore, the AI ​​lesson preparation experiment environment in step S1 is published as an AI teaching experiment course version control module, including the following steps:

[0013] Step S11: assigning an initial UUID identifier to the AI ​​lesson preparation experiment environment as a mark of the original version;

[0014] Step S12: Record the teaching information in the AI ​​lesson preparation experiment environment for use as a benchmark for subsequent comparison versions;

[0015] Step S13: Check whether there is any update in the AI ​​lesson preparation experiment environment through a specific monitoring mechanism;

[0016] Step S14: if an update is detected, the version identifier is updated; and the specific content of the update is saved in the database;

[0017] Step S15: Fill in the modifier information and filling time, and save to the database.

[0018] Furthermore, the teaching information in step S12 includes various parameters, settings and teaching contents.

[0019] Furthermore, in step S2, automatically selecting the optimal AI host based on the AI ​​lesson preparation experiment environment and the AI ​​teaching experiment environment includes the following steps:

[0020] Step S21: The user submits a request to create an AI environment;

[0021] Step S22: Determine the type of AI environment to be created;

[0022] Step S23: creating a corresponding AI environment based on the determined AI environment, and querying the database for record data of all corresponding AI hosts;

[0023] Step S24: Count the number of running AI environments and use the number to sort the data in descending order;

[0024] Step S25: Perform calculations in a loop to obtain the comprehensive performance of each AI host, and the AI ​​host with the smallest comprehensive performance is the optimal host.

[0025] Furthermore, in step S23, if it is determined that the type of the created AI environment is an AI lesson preparation experiment environment, the data recorded by all lesson preparation hosts in the database are queried; and the number of running AI lesson preparation experiment environments is counted, and the data is arranged in descending order.

[0026] Furthermore, in step S23, if it is determined that the type of the created AI environment is an AI teaching experiment environment, the data recorded by all experimental hosts in the database are queried; and the number of running AI teaching experiment environments is counted to arrange the data for lectures.

[0027] Furthermore, in step S25, the calculation is performed in a loop in sequence to calculate the number of running AI containers of each host in the AI ​​host, the CPU usage of each host, and the memory usage of each host.

[0028] Furthermore, when the AI ​​environment is created on the optimal AI host in step S3, a self-increment operation is performed on the AI ​​environment record running in the optimal AI host database based on the container orchestration technology.

[0029] Furthermore, when the AI ​​environment is destroyed based on the container orchestration technology in step S4, the corresponding running AI environment record in the AI ​​host database where the AI ​​environment is located is automatically decremented by one.

[0030] Furthermore, the AI ​​teaching experiment course encapsulates the operating system, software packages, code libraries, and configuration files used for specific AI teaching experiments; and the same AI teaching experiment course can enable different students to create the same AI teaching experiment environment based on the container image.

[0031] The present invention discloses an AI teaching experiment environment management method based on container orchestration. By controlling the version of the AI ​​teaching experiment course, it is convenient for teachers to update and switch the experiment content and configuration at any time. By selecting the optimal AI host to create an AI environment and releasing the unusable AI environment in time, more efficient, flexible and reliable AI environment management and utilization are achieved, avoiding waste of resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 It is a flow chart of the present invention.

[0033] Figure 2 It is a flow chart of the control module in the embodiment of the present invention.

[0034] Figure 3 The figure is a flow chart of selecting the optimal lesson preparation host module in an embodiment of the present invention. DETAILED DESCRIPTION

[0035] The present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments.

[0036] A method for managing an AI teaching experiment environment based on container orchestration, specifically comprising the following steps:

[0037] Step S1: Publish the AI ​​lesson preparation environment as an AI teaching experiment course to form a version control module in the container orchestration;

[0038] Step S2: In the case of the version control module, automatically select the optimal AI host based on the AI ​​environment to form the optimal AI host module in the container orchestration; the formula for automatically selecting the optimal AI host is:

[0039] I=min(X j *C j *M j );

[0040] Among them, I is the AI ​​environment, namely the AI ​​lesson preparation experiment environment and the AI ​​teaching experiment environment; X is the number of AI hosts running AI environments; C is the AI ​​host CPU utilization rate; M is the AI ​​host memory utilization rate; j is the AI ​​host number;

[0041] The AI ​​environment includes the AI ​​lesson preparation experiment environment and the AI ​​teaching experiment environment;

[0042] Step S3: creating and running an AI lesson preparation experiment environment or an AI teaching experiment environment on the optimal AI host;

[0043] Step S4: Determine in turn whether all AI environments are usable; if so, keep them; if it is determined that there is no usable AI environment, build a destruction module based on container orchestration technology to destroy the AI ​​environment.

[0044] Specifically, the AI ​​lesson preparation experiment environment in step S1 is published as an AI teaching experiment course version control module, including the following steps:

[0045] Step S11: assigning an initial UUID identifier to the AI ​​lesson preparation experiment environment as a mark of the original version;

[0046] Step S12: Record the teaching information in the AI ​​lesson preparation experiment environment for use as a benchmark for subsequent comparison versions;

[0047] Step S13: Check whether there is any update in the AI ​​lesson preparation experiment environment through a specific monitoring mechanism;

[0048] Step S14: if an update is detected, the version identifier is updated; and the specific content of the update is saved in the database;

[0049] Step S15: Fill in the modifier information and filling time, and save to the database.

[0050] The teaching information in step S12 includes various parameters, settings and teaching contents.

[0051] In step S2, the optimal AI host is automatically selected based on the AI ​​lesson preparation experiment environment and the AI ​​teaching experiment environment, including the following steps:

[0052] Step S21: The user submits a request to create an AI environment;

[0053] Step S22: Determine the type of AI environment to be created;

[0054] Step S23: creating a corresponding AI environment based on the determined AI environment, and querying the database for record data of all corresponding AI hosts;

[0055] Step S24: Count the number of running AI environments and use the number to sort the data in descending order;

[0056] Step S25: Perform calculations in a loop to obtain the comprehensive performance of each AI host, and the AI ​​host with the smallest comprehensive performance is the optimal host.

[0057] If it is determined in step S23 that the type of the created AI environment is an AI lesson preparation experiment environment, the data recorded by all lesson preparation hosts in the database are queried; and the number of running AI lesson preparation experiment environments is counted, and the data is arranged in descending order.

[0058] If it is determined in step S23 that the type of the created AI environment is an AI teaching experiment environment, the data recorded by all experimental hosts in the database are queried; and the number of running AI teaching experiment environments is counted to arrange the data for lectures.

[0059] In step S25, the calculation is performed in a loop to calculate the number of running AI containers of each host in the AI ​​host, the CPU usage of each host, and the memory usage of each host.

[0060] When the AI ​​environment is created on the optimal AI host in step S3, a self-increment operation is performed on the AI ​​environment record running in the optimal AI host database based on the container orchestration technology.

[0061] When the AI ​​environment is destroyed based on the container orchestration technology in step S4, the corresponding running AI environment record in the AI ​​host database where the AI ​​environment is located is automatically decremented by one.

[0062] The AI ​​teaching experiment course encapsulates the operating system, software package, code library and configuration file for a specific AI teaching experiment; and the same AI teaching experiment course can enable different students to create the same AI teaching experiment environment based on the container image. This ensures the consistency and stability of the experimental environment, allowing different students to conduct experiments under the same conditions, improving the teaching quality and the reliability of experimental results.

[0063] This embodiment provides an AI teaching experiment environment management method based on container orchestration and provides implementation steps of the AI ​​teaching experiment environment management method, such as Figure 1 The flowchart of an AI teaching experiment environment management method based on container orchestration is shown as follows, including:

[0064] AI teaching is the process of teachers imparting artificial intelligence-related knowledge, skills and concepts to students through various teaching methods and means. During teaching, teachers will explain the basic concepts of artificial intelligence, such as algorithms, models, data, etc.; introduce different AI application fields and cases to let students understand their practical uses; let students do some programming practice and use relevant tools and frameworks to develop simple AI applications; cultivate students' understanding and thinking ability of the development and impact of AI, and guide students to pay attention to AI ethics and social issues; through AI teaching, help students master certain AI knowledge and skills, and prepare them to adapt to the future intelligent society.

[0065] Container is an operating system-level virtualization technology that can package applications and their required dependencies, configurations, etc. together to form an independent operating environment; containers provide resource isolation and restrictions, so that different containers can run independently on the same host without interfering with each other, while sharing the operating system kernel with the host to save system resources; containers have the advantages of being lightweight, quickly deployable, and easy to migrate, and have been widely used in cloud computing and other fields; container orchestration refers to the process of automated management and coordination of multiple containers, including unified planning and control of container deployment, scheduling, resource allocation, service discovery, elastic scaling, fault handling, etc.; container orchestration tools or systems , which can efficiently manage a large number of containers so that they can work together to form a complex application system or service architecture; container orchestration helps to improve the scalability, reliability and operation and maintenance efficiency of the system, and better adapt to different business needs and load changes; container image is a standardized and repeatable file format that packages an application and its dependent environment; it contains all the content required to run a software, such as code, runtime environment, system tools and configuration; through container images, applications can be deployed and run quickly and conveniently, ensuring their consistency and portability in different environments, and can be easily moved and used between different systems and platforms, greatly improving the efficiency of application deployment.

[0066] Step 1: Perform version control on the experimental course; version control on the experimental course is a control system used to manage the continuous development and evolution of the experimental course. From a process perspective, any modification to the experimental course should be clearly recorded. The specific content of the modification includes the adjustment of the experimental steps, the addition of new experimental codes, the changes to the experimental requirements, etc., and the modification time and the person who modified it should be recorded; when performing version control on the experimental course, version identification should be performed, that is, each different version of the experimental course should be given a unique identification to accurately distinguish the courses at each stage, so that teachers and students can clearly know the version of the experimental course being used, which is convenient for comparison, analysis or in some cases, restoration to a specific historical version; for the data generated during the experiment, version control also ensures that it matches the corresponding course version to avoid erroneous experimental results or difficult-to-understand situations due to incompatibility between the data and the course version. By controlling the version of the experimental course, the quality and effect of teaching can be improved, and the continuous optimization and improvement of the experimental course can be promoted;

[0067] Step 2: Create an AI environment and automatically select the optimal AI host algorithm based on the AI ​​environment; query the database records and sort all AI host records in descending order by the number of running AI environments, and then multiply the CPU usage and memory usage of each AI host according to the number of cycles. The AI ​​host with the smallest product is the optimal AI host. The AI ​​hosts include the teacher's lesson preparation host and the student's experiment host. Create an AI environment on the optimal AI host; at this time, the database optimal AI host running AI environment record automatically adds one operation; AI environment creation automatically selects the optimal AI host algorithm, the formula is:

[0068] I=min(X j *C j *M j );

[0069] Where I is the AI ​​environment; X is the number of AI environments running on the AI ​​host; C is the CPU usage of the AI ​​host; M is the memory usage of the AI ​​host; j is the AI ​​host number;

[0070] The teacher's lesson preparation host is a dedicated computer device used by teachers to prepare the AI ​​teaching experimental environment. It is composed of multiple physical servers. It is the main tool for teachers to carry out lesson preparation-related work such as teaching design, writing lesson plans, making courseware, and collecting teaching materials. It usually has strong computing processing capabilities and large storage capacity to support teachers to run various teaching software, process various teaching files and multimedia resources, etc., and can meet the hardware performance requirements required for teachers to efficiently complete lesson preparation tasks. In order to ensure the quality of teachers' lesson preparation, the lesson preparation experimental environment can only be created and run in the teacher's lesson preparation host.

[0071] The student's experimental host is a computer device used by students to run the AI ​​teaching experimental environment, which is composed of multiple physical servers. It provides students with an operating platform for data processing, program running, simulation operations and other activities in experimental courses. The experimental host is usually configured according to the specific needs of the experiment to ensure that it can support students to successfully complete the relevant experimental content. It has strong computing power, related software and interfaces to help students deeply understand knowledge, master skills and cultivate practical ability through actual operations. In order to ensure the quality of students' classes, the teaching experimental environment can only be created and run in the student's experimental host.

[0072] Step 3: Release the AI ​​environment; that is, when the AI ​​environment is no longer needed, destroy it. During the destruction, the corresponding AI environment record in the AI ​​host database where the destroyed AI environment is located will be automatically reduced by one.

[0073] The AI ​​environment includes the AI ​​lesson preparation experiment environment and the AI ​​teaching experiment environment. The AI ​​lesson preparation experiment environment is a specific container built by teachers specifically for learning artificial intelligence technology and preparing lessons. The container integrates various resources and tools related to artificial intelligence, including software with intelligent analysis capabilities, data systems for data collection and processing, and artificial intelligence models that can assist teachers in teaching design and content creation; it aims to enable teachers to prepare AI teaching content more efficiently, optimize teaching plans, simulate scenarios and test links to verify the effectiveness and feasibility of AI lesson preparation results, and provide teachers with a place to explore and try to enable students to better learn AI to improve the lesson preparation process;

[0074] The AI ​​teaching experiment environment is a container with specific conditions and resources built specifically for teaching students artificial intelligence-related knowledge and skills. AI hardware suitable for teaching can be used in the container, including CPUs and GPUs with certain computing power, corresponding software system AI development tools, data sets and teaching software, etc.; various simulation scenarios and cases allow students to conduct actual operations, practical exercises and project experiments to deepen their understanding and mastery of AI technology, cultivate students' hands-on ability, innovative thinking and problem-solving ability, and explore and learn the application and development of AI technology in a near-real situation; the AI ​​teaching experiment environment can be published and generated by the AI ​​lesson preparation environment. The same AI teaching experiment course can create the same AI teaching experiment environment for different students. The AI ​​teaching experiment course is a pre-configured container image, which includes various software, environments and resources required for AI teaching experiments; it is a solidified state, similar to a snapshot, which encapsulates all necessary elements such as operating systems, software packages, code libraries, configuration files, etc. for specific AI teaching experiments; using this container image, multiple identical AI teaching experiment environments can be quickly created to facilitate the development of teaching activities and students' practical operations, ensuring the consistency and convenience of the experimental environment.

[0075] like Figure 2 As shown in the flowchart of the control module in an embodiment of an AI teaching experiment environment management method based on container orchestration, the workflow of the version control module when the lesson preparation environment is published as an experimental course includes the following steps:

[0076] Step A1: Assign an initial UUID identifier to the lesson preparation environment as the identifier of the original version;

[0077] Step A2: Record the current parameters, settings, teaching content and other information of the lesson preparation environment in detail as a benchmark for subsequent comparisons;

[0078] Step A3: Check in real time or regularly whether there are any modifications or updates to the lesson preparation environment through specific monitoring mechanisms such as file system monitoring and event trigger monitoring;

[0079] Step A4: Update the version identifier according to the changes, and use UUID as the unique identifier of the new version;

[0080] Step A5: Record the specific content of the changes in detail, including the newly added teaching links, modified experimental steps, etc.

[0081] Step A6: Fill in the information of the modifier and save it to the database;

[0082] Step A7: Fill in the modification time and save it to the database.

[0083] like Figure 3 As shown in the flowchart of selecting the optimal lesson preparation host module in an embodiment of an AI teaching experiment environment management method based on container orchestration, the workflow of automatically selecting the optimal AI host module for AI environment creation includes the following steps:

[0084] Step B1: The user submits a request to create an AI environment;

[0085] Step B2: determining the type of the created AI environment, i.e., determining whether the AI ​​environment is a lesson preparation environment or an experimental environment;

[0086] Step B3: If the created AI environment is a lesson preparation environment, query the database for all host record data, and sort the sequence in descending order using the number of running lesson preparation environments; if the created AI environment is an experimental environment, query the database for all experimental host record data, and sort the sequence in descending order using the number of running lesson preparation environments;

[0087] Step B4: cyclically multiply the AI ​​host CPU and memory usage in sequence;

[0088] If the AI ​​host list includes host a, host b and host c, the loop is carried out in sequence with the AI ​​host list as the loop body, and the number of running AI containers of host a, the CPU usage of host a, the memory usage of host a, the number of running AI containers of host b, the CPU usage of host b, the memory usage of host b, the number of running AI containers of host c, the CPU usage of host c and the memory usage of host c are calculated respectively.

[0089] Step B5: The AI ​​host with the smallest product is the optimal AI host;

[0090] Step B6: Create and run the AI ​​environment on the optimal AI host.

[0091] If you manage and use the lesson preparation host on the AI ​​experiment environment management platform, first add the lesson preparation host; add the lesson preparation host on the AI ​​experiment environment management platform, the lesson preparation host can run the lesson preparation environment of the teacher role user, and the AI ​​experiment environment management platform automatically sets the number of lesson preparation environments running on the lesson preparation host to 0; when creating the lesson preparation environment, the optimal AI host algorithm is automatically selected based on the AI ​​environment creation to select the optimal lesson preparation host, create the lesson preparation environment, and the number of containers running on the corresponding lesson preparation host automatically increases by 1; release the lesson preparation environment. When the user no longer needs a specific lesson preparation environment, the lesson preparation environment is destroyed, and the number of containers running on the corresponding lesson preparation host automatically decreases by 1.

[0092] If you manage and use the experimental host on the AI ​​experimental environment management platform, first add the experimental host; add the experimental host on the AI ​​experimental environment management platform; the experimental host can run the experimental environment of the student role, and the AI ​​experimental environment management platform automatically sets the number of experimental hosts running lesson preparation environments to 0; when creating the experimental environment, the optimal AI host algorithm is automatically selected based on the AI ​​environment creation to select the optimal experimental host, create the experimental environment, and the number of containers running on the corresponding experimental host automatically increases by 1; release the experimental environment. When the user no longer needs a specific experimental environment, the experimental environment is destroyed, and the number of containers running on the corresponding experimental host automatically decreases by 1.

[0093] In order to better meet the needs of AI teaching experiments, container orchestration technology has emerged. It can uniformly manage and schedule a large number of containers, achieve optimal resource allocation, flexible environment construction, and efficient operation and maintenance. Through container orchestration, various AI teaching-related software, frameworks, data, etc. can be encapsulated in different containers, and flexibly combined and configured according to teaching needs, which greatly improves the efficiency and flexibility of experimental environment management, while also ensuring the stability and reliability of the experimental environment, providing strong technical support for the smooth development of AI teaching, and helping to cultivate AI talents with practical ability and innovative spirit.

[0094] The above implementation modes are not limitations of the present invention, and the present invention is not limited to the above examples. Changes, modifications, additions or substitutions made by technicians in this technical field within the scope of the technical solution of the present invention also belong to the protection scope of the present invention.

Claims

1. A method for managing an AI teaching experiment environment based on container orchestration, characterized in that: The following steps are involved: Step S1: Publish the AI ​​lesson preparation environment as an AI teaching experiment course to form a version control module in the container orchestration; Step S2: In the case of the version control module, automatically select the optimal AI host based on the AI ​​environment to form the optimal AI host module in the container orchestration; the formula for automatically selecting the optimal AI host is: I=min(X j *C j *M j ); Among them, I is the AI ​​environment, namely the AI ​​lesson preparation experiment environment and the AI ​​teaching experiment environment; X is the number of AI hosts running AI environments; C is the AI ​​host CPU utilization rate; M is the AI ​​host memory utilization rate; j is the AI ​​host number; The AI ​​environment includes the AI ​​lesson preparation experiment environment and the AI ​​teaching experiment environment; Step S3: creating and running an AI lesson preparation experiment environment or an AI teaching experiment environment on the optimal AI host; Step S4: Determine in turn whether all AI environments are usable; if so, keep them; if it is determined that there is no usable AI environment, build a destruction module based on container orchestration technology to destroy the AI ​​environment.

2. The method for managing an AI teaching experiment environment based on container orchestration according to claim 1 is characterized in that: The step S1 in which the AI ​​lesson preparation experiment environment is published as an AI teaching experiment course version control module includes the following steps: Step S11: assigning an initial UUID identifier to the AI ​​lesson preparation experiment environment as a mark of the original version; Step S12: Record the teaching information in the AI ​​lesson preparation experiment environment for use as a benchmark for subsequent comparison versions; Step S13: Check whether there is any update in the AI ​​lesson preparation experiment environment through a specific monitoring mechanism; Step S14: if an update is detected, the version identifier is updated; and the specific content of the update is saved in the database; Step S15: Fill in the modifier information and filling time, and save to the database.

3. The method for managing an AI teaching experiment environment based on container orchestration according to claim 2 is characterized in that: The teaching information in step S12 includes various parameters, settings and teaching contents.

4. The method for managing an AI teaching experiment environment based on container orchestration according to claim 1 is characterized in that: The step S2 automatically selects the optimal AI host based on the AI ​​lesson preparation experiment environment and the AI ​​teaching experiment environment, including the following steps: Step S21: The user submits a request to create an AI environment; Step S22: Determine the type of AI environment to be created; Step S23: creating a corresponding AI environment based on the determined AI environment, and querying the database for record data of all corresponding AI hosts; Step S24: Count the number of running AI environments and use the number to sort the data in descending order; Step S25: Perform calculations in a loop to obtain the comprehensive performance of each AI host, and the AI ​​host with the smallest comprehensive performance is the optimal host.

5. The method for managing an AI teaching experiment environment based on container orchestration according to claim 4 is characterized in that: If it is determined in step S23 that the type of the created AI environment is an AI lesson preparation experiment environment, the data recorded by all lesson preparation hosts in the database are queried; and the number of running AI lesson preparation experiment environments is counted, and the data is arranged in descending order.

6. The method for managing an AI teaching experiment environment based on container orchestration according to claim 5 is characterized in that: If it is determined in step S23 that the type of the created AI environment is an AI teaching experiment environment, the data recorded by all experimental hosts in the database are queried; and the number of running AI teaching experiment environments is counted to arrange the data for lectures.

7. The method for managing an AI teaching experiment environment based on container orchestration according to claim 4 is characterized in that: In step S25, the calculation is performed cyclically in sequence to calculate the number of running AI containers of each host in the AI ​​host, the CPU usage of each host, and the memory usage of each host.

8. The method for managing an AI teaching experiment environment based on container orchestration according to claim 1 is characterized in that: When the AI ​​environment is created on the optimal AI host in step S3, a self-increment operation is performed on the AI ​​environment record in the optimal AI host database based on the container orchestration technology.

9. The method for managing an AI teaching experiment environment based on container orchestration according to claim 1 is characterized in that: When the AI ​​environment is destroyed based on the container orchestration technology in step S4, the corresponding running AI environment record in the AI ​​host database where the AI ​​environment is located is automatically decremented by one.

10. The method for managing an AI teaching experiment environment based on container orchestration according to claim 1, characterized in that: The AI ​​teaching experiment course encapsulates the operating system, software package, code library and configuration file for a specific AI teaching experiment; and the same AI teaching experiment course can enable different students to create the same AI teaching experiment environment based on the container image.

Images