Code generation method and device, equipment, medium and program product

Abstract

The invention provides a code generation method based on an artificial intelligence low-code system. The code generation method can be applied to the technical field of artificial intelligence. The method comprises the following steps: packaging a pre-trained artificial intelligence large model into an artificial intelligence model service to obtain a model application program interface; according to the model application program interface, deploying an artificial intelligence model service into a production environment; based on a production environment, integrating an artificial intelligence model service into a workflow engine of a low-code platform by utilizing a model application program interface so as to obtain an artificial intelligence low-code system; and automatically generating application program codes based on user application requirements by utilizing an artificial intelligence low-code system. According to the artificial intelligence low-code system, automatic decision support is provided, and a more intelligent, user-friendly, efficient and safe low-code application development platform is provided. The invention further provides a code generation device and equipment based on the artificial intelligence low-code system, a storage medium and a program product.

Description

Technical Field

[0001] The present disclosure relates to the fields of artificial intelligence and software development, and more specifically to a code generation method, apparatus, device, medium, and program product. Background Art

[0002] A low-code platform is a rapid application development platform that allows applications to be created through a graphical interface and model-driven logic, reducing the need for traditional programming. Its core advantages lie in increasing development speed, lowering technical barriers and enhancing collaboration between business and IT.

[0003] In the current existing technology, AIGC technology (Artificial Intelligence Generated Content) is used to create a low-code platform, that is, to use artificial intelligence technology to automatically generate content and develop intelligent low-code application development platforms and systems. These technologies understand user needs through natural language interaction, generate application building instructions and metadata files, and the low-code basic module automatically builds the application based on the metadata files.

[0004] At present, although progress has been made in improving development efficiency and lowering technical barriers, there are still some limitations, as follows: (1) Using low-code basic modules to build corresponding applications limits the ability of automated decision support, has certain limitations in automated decision support, and has not reached the expected level of intelligence. (2) The code generation process may lack optimization, resulting in uneven quality of generated code, low code generation efficiency, and lack of stability and reliability of the low-code platform. (3) The content data files generated intelligently are easily tampered with when they are transferred, and the data security of the low-code platform cannot be guaranteed. (4) It is necessary to use low-code basic modules to integrate the content of related data files. The user interface may not be friendly enough to non-technical users, and the user experience is poor. Summary of the invention

[0005] In view of at least one aspect of the above-mentioned problems, embodiments of the present disclosure provide a code generation method, apparatus, device, medium and program product for improving the automated decision-making performance of a low-code development platform.

[0006] According to the first aspect of the present disclosure, a code generation method based on an artificial intelligence low-code system is provided, including: encapsulating a pre-trained artificial intelligence large model into an artificial intelligence model service to obtain a model application program interface; deploying the artificial intelligence model service into a production environment according to the model application program interface; and based on the production environment, utilizing the model application program interface, integrating the artificial intelligence model service into a workflow engine of a low-code platform to obtain an artificial intelligence low-code system; wherein the artificial intelligence low-code system is used to realize automatic generation of application code based on user application requirements.

[0007] According to an embodiment of the present disclosure, based on the production environment, the model application program interface is used to integrate the artificial intelligence model service into the workflow engine of the low-code platform to obtain an artificial intelligence low-code system, including: building a test environment based on the production environment; integrating the artificial intelligence model service into the workflow engine of the low-code platform by using the model application program interface; using the integrated low-code platform as the system to be tested; based on the test environment, performing a comprehensive performance test on the system to be tested; and deploying the system to be tested that has passed the comprehensive performance test into the production environment to obtain the artificial intelligence low-code system.

[0008] According to an embodiment of the present disclosure, the use of the model application interface to integrate the artificial intelligence model service into the workflow engine of the low-code platform includes: configuring an artificial intelligence big model module corresponding to the artificial intelligence model service in the low-code platform according to the model application interface; designing a user interface corresponding to the artificial intelligence big model module; and based on the user interface, integrating the artificial intelligence model service into the workflow engine of the low-code platform.

[0009] According to an embodiment of the present disclosure, a comprehensive performance test is performed on the system to be tested based on the test environment, including: performing an interface association test on the model application interface in the system to be tested; performing a layered test on the artificial intelligence low-code system; performing a system quality test on the artificial intelligence low-code system; and performing a maintenance test on the artificial intelligence low-code system.

[0010] According to an embodiment of the present disclosure, the encapsulation of a pre-trained artificial intelligence big model into an artificial intelligence model service to obtain a model application program interface includes: using a serialization tool to serialize the pre-trained artificial intelligence big model to obtain a serialized model file; creating a service application corresponding to the artificial intelligence model service; wherein, in the service application, setting the serialized model file to be loaded when the artificial intelligence model service is started; and using a development tool to set an application program interface and an application program interface endpoint, and constructing an application program interface corresponding to the artificial intelligence model service to obtain the model application program interface.

[0011] According to an embodiment of the present disclosure, the creation of a service application corresponding to the artificial intelligence model service includes: selecting a service architecture corresponding to the artificial intelligence model service; setting the server environment and installing the dependency library of the pre-trained artificial intelligence large model on the server; and creating a service application corresponding to the artificial intelligence model service based on the service architecture and the server environment.

[0012] According to an embodiment of the present disclosure, the artificial intelligence model service is deployed to the production environment according to the model application interface, including: performing security authentication on the model application interface; based on the expected request volume, calling the model application interface to perform performance testing on the artificial intelligence model service; optimizing the artificial intelligence model service according to the results of the performance test; setting error handling logic in the service application; setting the application interface document corresponding to the model application interface; performing layered testing on the model application interface; and after passing the layered test, deploying the artificial intelligence model service to the production environment according to the application interface document.

[0013] According to an embodiment of the present disclosure, the artificial intelligence model service is deployed into a production environment according to the model application program interface, and also includes: setting a monitoring and logging mechanism for the artificial intelligence model service; detecting the operating status of the artificial intelligence model service; and updating the pre-trained artificial intelligence large model and the model application program interface according to business needs.

[0014] According to an embodiment of the present disclosure, the method further includes: deploying the artificial intelligence low-code system in a cloud computing environment to access the artificial intelligence low-code system via the Internet.

[0015] The second aspect of the present disclosure provides a code generation device based on an artificial intelligence low-code system, including: a packaging module, used to encapsulate a pre-trained artificial intelligence large model into an artificial intelligence model service to obtain a model application program interface; a deployment module, used to deploy the artificial intelligence model service into a production environment according to the model application program interface; an integration module, used to integrate the artificial intelligence model service into a workflow engine of a low-code platform based on the production environment and using the model application program interface to obtain an artificial intelligence low-code system; and a code generation module, used to use the artificial intelligence low-code system to automatically generate application code based on user application requirements.

[0016] According to an embodiment of the present disclosure, the integration module includes: a first integration sub-module, used to build a test environment based on the production environment; a second integration sub-module, used to integrate the artificial intelligence model service into the workflow engine of the low-code platform using the model application program interface; a third integration sub-module, used to use the integrated low-code platform as the system to be tested; a fourth integration sub-module, used to perform a comprehensive performance test on the system to be tested based on the test environment; and a fifth integration sub-module, used to deploy the system to be tested that has passed the comprehensive performance test into the production environment to obtain the artificial intelligence low-code system.

[0017] According to an embodiment of the present disclosure, the second integration sub-module includes a first integration unit for configuring an artificial intelligence big model module corresponding to the artificial intelligence model service in the low-code platform according to the model application program interface; a second integration unit for designing a user interface corresponding to the artificial intelligence big model module; and a third integration unit for integrating the artificial intelligence model service into the workflow engine of the low-code platform based on the user interface.

[0018] According to an embodiment of the present disclosure, the fourth integrated sub-module includes a fourth integrated unit, used to perform interface association testing on the model application interface in the system to be tested; a fifth integrated unit, used to perform layered testing on the artificial intelligence low-code system; a sixth integrated unit, used to perform system quality testing on the artificial intelligence low-code system; and a seventh integrated unit, used to perform maintenance testing on the artificial intelligence low-code system.

[0019] According to an embodiment of the present disclosure, the encapsulation module includes a first encapsulation sub-module, which is used to serialize the pre-trained artificial intelligence large model using a serialization tool to obtain a serialized model file; a second encapsulation sub-module, which is used to create a service application corresponding to the artificial intelligence model service; wherein, in the service application, it is set to load the serialized model file when the artificial intelligence model service is started; and a third encapsulation sub-module, which is used to use a development tool to set an application interface and an application interface endpoint, and build an application interface corresponding to the artificial intelligence model service to obtain the model application interface.

[0020] According to an embodiment of the present disclosure, the second packaging sub-module includes a first packaging unit, which is used to select a service architecture corresponding to the artificial intelligence model service; a second packaging unit, which is used to set the server environment and install the dependency library of the pre-trained artificial intelligence large model on the server; and a third packaging unit, which is used to create a service application corresponding to the artificial intelligence model service based on the service architecture and the server environment.

[0021] According to an embodiment of the present disclosure, the deployment module includes a first deployment unit for performing security authentication on the model application interface; a second deployment unit for calling the model application interface based on the expected request volume to perform performance testing on the artificial intelligence model service; a third deployment unit for optimizing the artificial intelligence model service according to the results of the performance test; a fourth deployment unit for setting error handling logic in the service application; a fifth deployment unit for setting the application interface document corresponding to the model application interface; a sixth deployment unit for performing layered testing on the model application interface; and a seventh deployment unit for deploying the artificial intelligence model service to the production environment according to the application interface document after passing the layered test.

[0022] According to an embodiment of the present disclosure, the deployment module also includes an eighth deployment unit, which is used to set up a monitoring and logging mechanism for the artificial intelligence model service; a ninth deployment unit, which is used to detect the running status of the artificial intelligence model service; and a tenth deployment unit, which is used to update the pre-trained artificial intelligence large model and the model application program interface according to business needs.

[0023] According to an embodiment of the present disclosure, the device also includes a cloud deployment module for deploying the artificial intelligence low-code system into a cloud computing environment so as to access the artificial intelligence low-code system via the Internet.

[0024] A third aspect of the present disclosure provides an electronic device, comprising: one or more processors; and a memory for storing one or more computer programs, wherein the one or more processors execute the one or more computer programs to implement the steps of the above method.

[0025] The fourth aspect of the present disclosure further provides a computer-readable storage medium on which a computer program or instruction is stored, and the steps of the above method are implemented when the above computer program or instruction is executed by a processor.

[0026] The fifth aspect of the present disclosure further provides a computer program product, including a computer program or instructions, which implement the steps of the above method when the above computer program or instructions are executed by a processor.

[0027] According to the embodiments of the present disclosure, since the current low-code development platform has limitations in automated decision support, the trained model is encapsulated as an artificial intelligence model service, and the artificial intelligence model service is integrated into the low-code workflow engine to obtain an artificial intelligence low-code system. In the artificial intelligence low-code system, the artificial intelligence big model and the low-code platform are deeply integrated to provide automated decision support and provide a more intelligent, user-friendly, efficient and secure low-code application development platform. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] The above contents and other objects, features and advantages of the present disclosure will become more apparent through the following description of the embodiments of the present disclosure with reference to the accompanying drawings, in which:

[0029] Figure 1 The application scenario diagram of the code generation method based on the artificial intelligence low-code system according to the embodiment of the present disclosure is schematically shown;

[0030] Figure 2 A flowchart of a code generation method based on an artificial intelligence low-code system according to an embodiment of the present disclosure is schematically shown;

[0031] Figure 3 The model training flow chart of the code generation method based on the artificial intelligence low-code system according to an embodiment of the present disclosure is schematically shown;

[0032] Figure 4 A model encapsulation flow chart of a code generation method based on an artificial intelligence low-code system according to an embodiment of the present disclosure is schematically shown;

[0033] Figure 5 The following schematically shows a creation service flow chart of a code generation method based on an artificial intelligence low-code system according to an embodiment of the present disclosure;

[0034] Figure 6The service deployment flow chart of the code generation method based on the artificial intelligence low-code system according to an embodiment of the present disclosure is schematically shown;

[0035] Figure 7 Another service deployment flow chart of the code generation method based on the artificial intelligence low-code system according to an embodiment of the present disclosure is schematically shown;

[0036] Figure 8 The service integration flow chart of the code generation method based on the artificial intelligence low-code system according to an embodiment of the present disclosure is schematically shown;

[0037] Fig. 9 Another service integration flow chart of a code generation method based on an artificial intelligence low-code system according to an embodiment of the present disclosure is schematically shown;

[0038] Fig.10 The system test flow chart of the code generation method based on the artificial intelligence low-code system according to the embodiment of the present disclosure is schematically shown;

[0039] Fig.11 A structural block diagram of a code generation device based on an artificial intelligence low-code system according to an embodiment of the present disclosure is schematically shown; and

[0040] Fig.12 A block diagram of an electronic device suitable for implementing a code generation method based on an artificial intelligence low-code system according to an embodiment of the present disclosure is schematically shown. DETAILED DESCRIPTION

[0041] Hereinafter, embodiments of the present disclosure will be described with reference to the accompanying drawings. However, it should be understood that these descriptions are exemplary only and are not intended to limit the scope of the present disclosure. In the following detailed description, for ease of explanation, many specific details are set forth to provide a comprehensive understanding of the embodiments of the present disclosure. However, it is apparent that one or more embodiments may also be implemented without these specific details. In addition, in the following description, descriptions of known structures and technologies are omitted to avoid unnecessary confusion of the concepts of the present disclosure.

[0042] The terms used herein are only for describing specific embodiments and are not intended to limit the present disclosure. The terms "comprise", "include", etc. used herein indicate the existence of features, steps, operations and / or components, but do not exclude the existence or addition of one or more other features, steps, operations or components.

[0043] All terms (including technical and scientific terms) used herein have the meanings commonly understood by those skilled in the art unless otherwise defined. It should be noted that the terms used herein should be interpreted as having a meaning consistent with the context of this specification and should not be interpreted in an idealized or overly rigid manner.

[0044] When using expressions such as "at least one of A, B, and C, etc.", they should generally be interpreted according to the meaning of the expression commonly understood by those skilled in the art (for example, "a system having at least one of A, B, and C" should include but is not limited to a system having A alone, B alone, C alone, A and B, A and C, B and C, and / or A, B, C, etc.).

[0045] The embodiments of the present disclosure provide a code generation method based on an artificial intelligence low-code system, which encapsulates a pre-trained artificial intelligence large model into an artificial intelligence model service to obtain a model application program interface; deploys the artificial intelligence model service into a production environment according to the model application program interface; and based on the production environment, utilizes the model application program interface to integrate the artificial intelligence model service into the workflow engine of a low-code platform to obtain an artificial intelligence low-code system; wherein the artificial intelligence low-code system is used to realize automatic generation of application code based on user application requirements.

[0046] According to the embodiments of the present disclosure, since the current low-code development platform has limitations in automated decision support, the trained model is encapsulated as an artificial intelligence model service, and the artificial intelligence model service is integrated into the low-code workflow engine to obtain an artificial intelligence low-code system. In the artificial intelligence low-code system, the artificial intelligence big model and the low-code platform are deeply integrated to provide automated decision support and provide a more intelligent, user-friendly, efficient and secure low-code application development platform.

[0047] Figure 1 An application scenario diagram of a code generation method based on an artificial intelligence low-code system according to an embodiment of the present disclosure is schematically shown.

[0048] like Figure 1 As shown, the application scenario 100 according to this embodiment may include an automatic generation of application code scenario. The network 104 is used to provide a medium for a communication link between the first terminal device 101, the second terminal device 102, the third terminal device 103 and the server 105. The network 104 may include various connection types, such as wired, wireless communication links or optical fiber cables, etc.

[0049] The user can use the first terminal device 101, the second terminal device 102, and the third terminal device 103 to interact with the server 105 through the network 104 to receive or send messages, etc. Various communication client applications can be installed on the first terminal device 101, the second terminal device 102, and the third terminal device 103, such as shopping applications, web browser applications, search applications, instant messaging tools, email clients, social platform software, etc. (only for example).

[0050] The first terminal device 101, the second terminal device 102, and the third terminal device 103 may be various electronic devices having display screens and supporting web browsing, including but not limited to smart phones, tablet computers, laptop computers, desktop computers, and the like.

[0051] The server 105 may be a server that provides various services, such as a background management server (only as an example) that provides support for websites browsed by users using the first terminal device 101, the second terminal device 102, and the third terminal device 103. The background management server may analyze and process the received data such as user requests, and feed back the processing results (such as web pages, information, or data obtained or generated according to user requests) to the terminal device.

[0052] It should be noted that the code generation method based on the artificial intelligence low-code system provided in the embodiment of the present disclosure can generally be executed by the server 105. Accordingly, the code generation device based on the artificial intelligence low-code system provided in the embodiment of the present disclosure can generally be set in the server 105. The code generation method based on the artificial intelligence low-code system provided in the embodiment of the present disclosure can also be executed by a server or server cluster that is different from the server 105 and can communicate with the first terminal device 101, the second terminal device 102, the third terminal device 103 and / or the server 105. Correspondingly, the code generation device based on the artificial intelligence low-code system provided in the embodiment of the present disclosure can also be set in a server or server cluster that is different from the server 105 and can communicate with the first terminal device 101, the second terminal device 102, the third terminal device 103 and / or the server 105.

[0053] It should be understood that Figure 1 The number of terminal devices, networks and servers in the embodiment is only for illustration. Any number of terminal devices, networks and servers may be provided according to implementation requirements.

[0054] The following will be based on Figure 1 The scene described by Figure 2~Figure 10 The code generation method based on the artificial intelligence low-code system of the disclosed embodiment is described in detail.

[0055] Figure 2 The flowchart of the code generation method based on the artificial intelligence low-code system according to an embodiment of the present disclosure is schematically shown.

[0056] like Figure 2 As shown, the code generation method based on the artificial intelligence low-code system of this embodiment includes operations S210 to S240, and the code generation method can be executed automatically.

[0057] In operation S210, the pre-trained artificial intelligence large model is packaged as an artificial intelligence model service to obtain a model application program interface.

[0058] Artificial Intelligence Large Model refers to an artificial intelligence model with massive parameters and complex structure (AI Large Model for short). These models are usually built on a deep learning architecture, and through deep learning technology, they can learn and extract complex patterns and rules from large-scale data. It should be noted that artificial intelligence is Artificial Intelligence, or AI for short.

[0059] Model application programming interface refers to the application programming interface corresponding to the artificial intelligence model service. The application programming interface (API) is an application programming interface, which is a set of rules and definitions for communication between different components in a software system.

[0060] The process of encapsulating a large AI model is the process of converting a machine learning or AI model into a service that can be called through an API. Encapsulating a trained AI model as a service and providing an application program interface is a multi-step process involving model serialization, service creation, API design, and deployment. Encapsulating a trained AI model as an AI model service can provide a model application program interface corresponding to the AI ​​model service.

[0061] Encapsulating the AI ​​big model as a service that can be called through an API makes the use of the AI ​​big model more flexible and convenient, providing non-technical users with an easy way to integrate and use complex AI big models.

[0062] In operation S220, the artificial intelligence model service is deployed to the production environment according to the model application program interface.

[0063] In the embodiment of the present disclosure, in the process of encapsulating the artificial intelligence large model into the artificial intelligence model service, the API of the service, i.e., the model application program interface, can be obtained accordingly. In the subsequent process, the artificial intelligence model service can be used by calling the model application program interface and deployed to the corresponding production environment for user use.

[0064] In operation S230, based on the production environment, the artificial intelligence model service is integrated into the workflow engine of the low-code platform using the model application program interface to obtain an artificial intelligence low-code system.

[0065] AI low-code system integration and testing are key links to ensure seamless collaboration between AI model services and low-code workflow engines. A detailed integration plan can be developed, including integration steps, resource requirements, schedules, and testing strategies.

[0066] Integrate the AI ​​model service into the workflow engine of the low-code platform to obtain the AI ​​low-code system. In fact, the AI ​​low-code system is a low-code platform that integrates the AI ​​model service. Through the AI ​​low-code system, application code can be automatically generated based on the user's application needs, user feedback can be collected in a timely manner, and AI large models and workflows can be iteratively optimized. The AI ​​low-code system can also be designed to support future expansion and maintenance, ensure that the use of AI large models complies with relevant regulations, and implement security measures.

[0067] A low-code platform is a rapid application development platform that allows users to create applications with minimal coding. It usually provides a graphical user interface that allows users to build applications by dragging components and configuring parameters. It allows applications to be created through a graphical interface and model-driven logic without the need for in-depth code writing, reducing the need for traditional programming.

[0068] The current low-code development platform has limitations in terms of automated decision support. Therefore, in the embodiments of the present invention, the trained model is encapsulated as an artificial intelligence model service, and the artificial intelligence model service is integrated into the low-code workflow engine to obtain an artificial intelligence low-code system. In the artificial intelligence low-code system, the deep integration of the artificial intelligence big model and the low-code platform provides automated decision support and provides a more intelligent, user-friendly, efficient and secure low-code application development platform.

[0069] In operation S240, an artificial intelligence low-code system is used to automatically generate application code based on user application requirements.

[0070] According to user application requirements, the AI ​​low-code system searches for matching models or model combinations in its pre-built AI large model model library. These models may be pre-trained for different types of business scenarios (such as business management, office management, project management, etc.) and functional modules (such as data query, form submission, report generation, etc.). After finding a matching model, further adjust the model parameters according to specific user needs; call the model with determined and adjusted parameters to generate corresponding code snippets according to predefined rules, assemble and integrate the generated code snippets according to a certain logical order and architecture, obtain the application code corresponding to the user's application requirements, and display the application code on the front-end page of the AI ​​low-code system.

[0071] The artificial intelligence low-code system constructed in this embodiment deeply integrates the artificial intelligence big model, and deeply integrates the artificial intelligence model service with the low-code platform to provide automated decision support. The innovatively designed user interface provides a more intuitive and user-friendly interface design. Through the intuitive user interface and automated decision support, the technical threshold is lowered and the user experience is improved. The automated code generation and integration process reduces the workload of manual coding and improves development efficiency. The code generation process is optimized through AI algorithms, the quality and performance of the code are improved, the code quality is guaranteed, the scalability and customizability of the system are enhanced, and the modular design and northbound interface enable the system to be customized and expanded according to different application scenarios. The artificial intelligence low-code system strengthens the security of data and models to ensure security and compliance.

[0072] According to an embodiment of the present disclosure, the code generation method based on the artificial intelligence low-code system of this embodiment also includes deploying the artificial intelligence low-code system in a cloud computing environment to access the artificial intelligence low-code system via the Internet.

[0073] Deploy the low-code workflow engine that integrates the large AI model to the cloud, implement monitoring, deploy the applications, services or models of the AI ​​low-code system in a cloud computing environment, and realize cloud deployment to facilitate access and use through the Internet. Deploy the AI ​​low-code system to a cloud computing environment to realize cloud deployment and elastic scaling, and utilize the elasticity of cloud services to improve the stability and reliability of the system.

[0074] Figure 3 The model training flowchart of the code generation method based on the artificial intelligence low-code system according to an embodiment of the present disclosure is schematically shown.

[0075] like Figure 3 As shown, according to an embodiment of the present disclosure, the code generation method based on the artificial intelligence low-code system of this embodiment also includes training the artificial intelligence big model to obtain a trained artificial intelligence big model, and training the artificial intelligence big model includes operations S310-S330.

[0076] In operation S310, initial data is collected based on the artificial intelligence requirements of the low-code workflow.

[0077] Clarify the specific role and requirements of large AI models in low-code workflows to achieve AI requirements, including but not limited to basic code framework construction, complex logic code generation, code completion and optimization, component recommendations, architecture design suggestions, business process optimization suggestions, real-time error detection, error cause analysis, automatic repair suggestions, and code snippet reuse.

[0078] Building a basic code framework: Based on the requirements description entered by the developer, the AI ​​big model can quickly generate the basic code framework of the application.

[0079] Complex logic code generation: For some complex business logic, such as risk assessment algorithms in financial transaction systems and path optimization algorithms in logistics management systems, AI big models can generate efficient and accurate codes based on a large amount of relevant data and algorithm models.

[0080] Code completion and optimization: When developers are writing code, the AI ​​big model can analyze the code context in real time and provide code completion and optimization suggestions.

[0081] Component recommendation: Low-code platforms usually provide a large number of reusable components. The AI ​​big model can intelligently recommend suitable components for the components and templates in the low-code platform based on the specific needs and business scenarios of the project, making it easier for developers to quickly find suitable resources.

[0082] Architecture design suggestions: For the overall architecture design of the application, the AI ​​big model can provide developers with reference suggestions based on previous successful cases and best practices.

[0083] Business process optimization suggestions: In terms of business process design, the AI ​​big model can analyze existing business processes, identify bottlenecks and optimization points, and provide corresponding optimization suggestions.

[0084] Real-time error detection: During the code writing process, the AI ​​big model can scan the code in real time to detect syntax errors, logical errors, potential security vulnerabilities and other problems.

[0085] Error cause analysis: When an error is detected, the AI ​​big model can not only point out the location of the error, but also analyze the cause of the error and provide a detailed error explanation.

[0086] Automatic repair suggestions: In addition to detecting errors, AI big models can also provide automatic repair suggestions based on the type and context of the error.

[0087] Code snippet reuse: AI big models can analyze the code written by developers, identify reusable code snippets, and store them in the code library.

[0088] Therefore, based on the needs of artificial intelligence, artificial intelligence big models require high-quality, large-scale, and diverse initial data sets in low-code workflows. The data types required by artificial intelligence big models include video, pictures, audio, text and other types and formats of data. Data requirements, high-quality business data, and a large amount of business data are required to understand business processes and rules. Collect various types of initial data, and the collection methods include but are not limited to log collection and network data collection.

[0089] In operation S320, initial data is preprocessed.

[0090] The preprocessing process is to clean and organize the initial data. The preprocessed data is used to train the data of the AI ​​large model. Preprocessing at least includes cleaning, organization, and labeling.

[0091] Cleaning: Handling missing values, noise, errors, duplicates, etc.

[0092] Sorting: unify data format, standardize or normalize data, and encode data.

[0093] Annotation: Label the data so that large AI models can understand and process it.

[0094] In operation S330, based on the preselected AI algorithm, the artificial intelligence big model is trained using the preprocessed initial data to obtain a trained artificial intelligence big model.

[0095] Select a suitable AI algorithm and use the preprocessed data for model training. AI algorithms include but are not limited to deep learning algorithms (such as self-attention mechanism, convolutional neural network, recurrent neural network), reinforcement learning algorithms, random forest, gradient boosting tree, and support vector machine.

[0096] In the embodiments of the present disclosure, based on the artificial intelligence requirements of the low-code workflow, initial data is collected to train the artificial intelligence big model, so as to achieve accurate matching of task requirements, enhance the generalization ability of the artificial intelligence big model, and improve the accuracy and pertinence of the artificial intelligence big model.

[0097] Figure 4 The model encapsulation flowchart of the code generation method based on the artificial intelligence low-code system according to an embodiment of the present disclosure is schematically shown.

[0098] like Figure 4 As shown, according to an embodiment of the present disclosure, in operation S210, a pre-trained artificial intelligence large model is encapsulated as an artificial intelligence model service to obtain a model application program interface, including operations S410-S430.

[0099] In operation S410, a pre-trained artificial intelligence large model is serialized using a serialization tool to obtain a serialized model file.

[0100] The trained AI model is saved to a file through a serialization tool for subsequent loading and use. The serialization tool is a process and tool that converts model objects into a format that can be transmitted and stored between different systems or platforms. Specifically, the serialization tool can convert objects in memory (such as model objects in a database) into a standard format for transmission over the network or storage in a file.

[0101] In operation S420, a service application corresponding to the artificial intelligence model service is created; wherein, the service application is set to load the serialized model file when the artificial intelligence model service is started.

[0102] Implement the loading and initialization of the AI ​​big model in the service application, write logic in the service code to load the AI ​​big model when the service starts, and implement the model's pre-processing and post-processing logic to adapt the model's input and output.

[0103] In operation S430, using development tools, the application program interface and the application program interface endpoint are set to build the application program interface corresponding to the artificial intelligence model service to obtain the model application program interface.

[0104] Use development tools to write application interface endpoints and process HTTP requests and responses. Development tools are used to build software frameworks for Web applications. They provide a complete set of development tools and libraries to simplify the development process of Web applications. They usually implement common functions such as HTTP protocol, routing mechanism, template rendering, data validation, database operations, etc., so that developers can quickly develop feature-rich Web applications.

[0105] Design application program interfaces within the software framework provided by the development tools, such as designing API request and response formats to ensure they are compatible with front-end applications or service callers.

[0106] Under the software framework provided by the development tool, determine the specific application interface endpoint according to the designed API interface, such as defining the endpoint path (such as / api / predict) and setting the request method (such as POST) application interface endpoint.

[0107] In the embodiments of the present disclosure, a service application is constructed based on the serialized model file corresponding to the pre-trained artificial intelligence model and encapsulated as an artificial intelligence model service. The artificial intelligence model is encapsulated in the form of a service to facilitate subsequent loading and use. The model information cannot be tampered with at will, thereby ensuring the security of data information.

[0108] Figure 5 The schematic diagram shows a creation service flow chart of a code generation method based on an artificial intelligence low-code system according to an embodiment of the present disclosure.

[0109] like Figure 5 As shown, according to an embodiment of the present disclosure, in operation S420, a service application corresponding to the artificial intelligence model service is created, including operations S510-S530.

[0110] In operation S510, a service architecture corresponding to the artificial intelligence model service is selected.

[0111] Determine the service architecture, such as using microservice architecture or monolithic application architecture.

[0112] In operation S520, the server environment is set, and the dependent library of the pre-trained artificial intelligence large model is installed on the server.

[0113] To set up a server environment, you need to prepare a server environment, such as a local server, cloud server, or container service. Install the required dependency libraries on the server, such as the deep learning framework used for model training, to facilitate the use and iterative update of large artificial intelligence models.

[0114] In operation S530, a service application corresponding to the artificial intelligence model service is created based on the service architecture and the server environment.

[0115] Write the code for the AI ​​model service to create a service application that loads the serialized model file and provides an inference interface for the model.

[0116] In an embodiment of the present disclosure, a service application corresponding to the artificial intelligence model service is created based on the service architecture and the server environment, and the service application is used to implement the preprocessing and post-processing logic of the model to adapt the input and output of the model to ensure the availability of the artificial intelligence model service.

[0117] Figure 6 The service deployment flowchart of the code generation method based on the artificial intelligence low-code system according to an embodiment of the present disclosure is schematically shown.

[0118] like Figure 6 As shown, according to an embodiment of the present disclosure, in operation S220, the artificial intelligence model service is deployed to the production environment according to the model application program interface, including operations S610-S670.

[0119] In operation S610, security authentication is performed on the model application program interface.

[0120] Security authentication belongs to security settings, which implements the authentication mechanism of the model application program interface, such as using API keys, open authorization, etc. To ensure data transmission security, use the HTTPS protocol.

[0121] In operation S620, based on the expected request volume, the model application program interface is called to perform performance testing on the artificial intelligence model service.

[0122] Perform performance testing on AI model services based on the expected request volume. For example, use load testing tools to simulate scenarios such as a large number of users accessing AI model services simultaneously and performing complex operations to evaluate performance indicators such as response time, throughput, CPU load, and memory usage of AI model services, thereby determining the performance of AI model services under extreme conditions.

[0123] In operation S630, the artificial intelligence model service is optimized according to the results of the performance test.

[0124] Optimize the AI ​​model service based on the results of performance testing, such as using model optimization technology, increasing server resources, or using caching mechanisms.

[0125] In operation S640, error handling logic is set in the service application.

[0126] Implement error handling logic to ensure that appropriate error information is returned when a request error or model service exception occurs.

[0127] In operation S650, an application program interface document corresponding to the model application program interface is set.

[0128] You can write detailed documentation for your API, including endpoints, request parameters, response formats, usage examples, etc.

[0129] In operation S660, a layered test is performed on the model application program interface.

[0130] Perform thorough layered testing on the model application interface, including unit testing, integration testing, and end-to-end testing.

[0131] In operation S670, after passing the layered test, the artificial intelligence model service is deployed to the production environment according to the application interface document.

[0132] Deploy the AI ​​model service to the production environment, either directly to the server or through containerization.

[0133] In the embodiments of the present disclosure, after security certification, performance testing, setting of error handling logic, setting of application interface documentation, layered testing, and deployment of the artificial intelligence model service in the production environment, and after various designs and tests, the artificial intelligence model service meets the requirements of actual production applications, facilitating its integration into the low-code platform.

[0134] Figure 7 Another service deployment flowchart of the code generation method based on the artificial intelligence low-code system according to an embodiment of the present disclosure is schematically shown.

[0135] like Figure 7 As shown, according to an embodiment of the present disclosure, in operation S220, the artificial intelligence model service is deployed to the production environment according to the model application program interface, and also includes operations S710-S730.

[0136] In operation S710, a monitoring and logging mechanism for the artificial intelligence model service is set.

[0137] Set up monitoring and logging mechanisms to track model API usage and performance metrics.

[0138] In operation S720, the running status of the artificial intelligence model service is detected.

[0139] Regularly check the running status of the AI ​​model service and handle any potential problems.

[0140] In operation S730, the pre-trained artificial intelligence big model and model application program interface are updated according to business needs.

[0141] Updating models and APIs according to business needs can collect user feedback and their latest business needs, iteratively optimize large AI models, and update model application programming interfaces.

[0142] In the embodiments of the present disclosure, the model application program interface is monitored, detected, and updated to implement regular inspection of the operation of the service, facilitate the discovery and handling of any potential problems, facilitate the maintenance and update of the artificial intelligence model service, and improve the stability and reliability of the artificial intelligence model service.

[0143] Figure 8 The service integration flowchart of the code generation method based on the artificial intelligence low-code system according to an embodiment of the present disclosure is schematically shown.

[0144] like Figure 8 As shown, according to an embodiment of the present disclosure, in operation S230, based on the production environment, the artificial intelligence model service is integrated into the workflow engine of the low-code platform using the model application program interface to obtain an artificial intelligence low-code system, including operations S810-S850.

[0145] In operation S810, a test environment is built based on the production environment.

[0146] System integration and testing are key links to ensure seamless collaboration between AI model services and low-code workflow engines. First, you need to build the corresponding processing environment and the test environment to ensure that it is similar to the production environment to ensure the accuracy of the test results.

[0147] It should be noted that the production environment refers to the environment in which the system officially provides services to the outside world. It is an environment that users can directly access and is used to provide services to end users. The test environment is an environment specifically used for software testing. It is usually similar to the production environment, but is not directly facing real users. The test environment performs various types of tests on the system before formal deployment, which can help discover and fix errors and defects in the software and ensure that the system can run stably in the production environment.

[0148] In operation S820, the artificial intelligence model service is integrated into the workflow engine of the low-code platform using the model application programming interface.

[0149] Integrate AI model services with low-code workflow engines to ensure error-free communication and data exchange between them.

[0150] In operation S830, the integrated low-code platform is used as the system to be tested.

[0151] In operation S840, a comprehensive performance test is performed on the system to be tested based on the test environment.

[0152] Comprehensive performance testing includes but is not limited to interface association testing, layered testing, system quality testing and maintenance testing.

[0153] In operation S850, the system under test that has passed the comprehensive performance test is deployed into a production environment to obtain an artificial intelligence low-code system.

[0154] After all tests are completed, the AI ​​low-code system is deployed to the production environment to ensure that the integration of the AI ​​model service with the low-code workflow engine is successful and that the system is stable, reliable, and meets user needs before being deployed to the production environment.

[0155] By integrating artificial intelligence model services on the low-code platform, the artificial intelligence low-code system built can write integrated documents and user manuals, and train users so that users can better use the artificial intelligence low-code system.

[0156] In the embodiments of the present disclosure, the artificial intelligence model service is integrated into the workflow engine of the low-code platform to obtain the system to be tested, which is then deployed to the production environment after comprehensive performance testing. To ensure seamless collaboration between the artificial intelligence model service and the low-code workflow engine, the security of data and models is enhanced to ensure compliance.

[0157] Fig. 9 Another service integration flowchart of the code generation method based on the artificial intelligence low-code system according to an embodiment of the present disclosure is schematically shown.

[0158] like Fig. 9 As shown, according to an embodiment of the present disclosure, in operation S820, the artificial intelligence model service is integrated into the workflow engine of the low-code platform using the model application program interface, including operations S910-S930.

[0159] In operation S910, according to the model application program interface, an artificial intelligence big model module corresponding to the artificial intelligence model service is configured in the low-code platform.

[0160] In order to adapt the artificial intelligence model service to the low-code platform, interfaces are designed in the low-code platform, tools or modules are developed, and artificial intelligence large model modules are configured so that the artificial intelligence model service can be integrated into the low-code platform.

[0161] In operation S920, a user interface corresponding to the artificial intelligence large model module is designed.

[0162] Design an intuitive user interface to enable users to easily configure and use large AI models.

[0163] In operation S930, based on the user interface, the artificial intelligence model service is integrated into the workflow engine of the low-code platform.

[0164] Integrate AI model services into low-code workflow engines to facilitate AI low-code system testing.

[0165] In the embodiments of the present disclosure, an artificial intelligence big model module and a user interface corresponding to the artificial intelligence model service are designed. Based on the artificial intelligence big model module and the user interface, the artificial intelligence model service is integrated into the low-code platform, and the artificial intelligence big model module such as the interface design, development tools or modules on the low-code platform is implemented to integrate the artificial intelligence model service into the low-code platform. An intuitive user interface is designed to enable users to easily configure and use the artificial intelligence big model. The innovative design of the user interface provides a more intuitive and user-friendly interface design. The user interface is friendly enough to non-technical users, thereby improving the user experience.

[0166] Fig.10The system test flowchart of the code generation method based on the artificial intelligence low-code system according to an embodiment of the present disclosure is schematically shown.

[0167] like Fig.10 As shown, according to an embodiment of the present disclosure, in operation S840, a comprehensive performance test is performed on the system to be tested based on the test environment, including operations S1010-S1040.

[0168] In operation S1010, an interface association test is performed on the model application program interface in the system to be tested.

[0169] Test the application programming interface of the AI ​​model service, including request and response formats, error handling, etc.

[0170] In operation S1020, the artificial intelligence low-code system is tested in layers.

[0171] Layered testing includes but is not limited to unit testing, integration testing, and end-to-end testing.

[0172] Unit testing: Perform unit testing on each component of the integration to ensure that each component works as expected.

[0173] Integration testing: Verify whether the integration of the AI ​​model service and the low-code workflow engine meets the design requirements.

[0174] End-to-end testing: Perform end-to-end testing to simulate real user scenarios, from the user interface to the backend services.

[0175] In operation S1030, a system quality test is performed on the artificial intelligence low-code system.

[0176] System quality testing includes but is not limited to performance testing, security testing, stability and stress testing, and user experience testing.

[0177] Performance testing: Perform performance testing on the AI ​​model service to ensure that it can handle the expected request load.

[0178] Security testing: Conduct security testing, including API authentication and authorization mechanisms, and the security of data transmission.

[0179] Stability and stress testing: Perform stability and stress testing to ensure that the system remains stable under high load or long-term operation.

[0180] User experience testing: Invite non-technical users to conduct user experience testing and collect feedback to optimize the user interface and operation process.

[0181] In operation S1040, a maintenance test is performed on the artificial intelligence low-code system based on the test environment.

[0182] Maintenance testing includes error and exception handling testing, documentation testing, regression testing, etc.

[0183] Error and exception handling testing: Test the AI ​​low-code system’s ability to handle various errors and exceptions to ensure the system’s robustness.

[0184] Document testing: Verify whether the provided documents and examples are accurate and can help users understand and use AI model services.

[0185] Regression testing: Regression testing is performed after each iteration or update to ensure that the new code has not broken the existing functionality.

[0186] In the embodiments of the present disclosure, the system under test is subjected to a number of comprehensive performance tests, such as interface association test, layered test, system quality test, maintenance test, etc. After comprehensive performance testing, it is ensured that the integration of the artificial intelligence model service and the low-code workflow engine is successful, and the system is stable, reliable and meets user needs before being deployed to the production environment, ensuring the quality of the code generated by the artificial intelligence low-code system and improving the efficiency of code generation. The modular design and northbound interface allow the system to be customized and expanded according to different application scenarios, enhancing the scalability and customizability of the system.

[0187] Based on the above-mentioned code generation method based on artificial intelligence low-code system, the present disclosure also provides a code generation device based on artificial intelligence low-code system. Fig.11 The device is described in detail.

[0188] Fig.11 The structural block diagram of a code generation device based on an artificial intelligence low-code system according to an embodiment of the present disclosure is schematically shown.

[0189] like Fig.11 As shown, the code generation device 1100 for the artificial intelligence low-code system of this embodiment includes a packaging module 1110, a deployment module 1120, an integration module 1130 and a code generation module 1140.

[0190] The encapsulation module 1110 is used to encapsulate the pre-trained artificial intelligence model into an artificial intelligence model service to obtain a model application program interface. In one embodiment, the encapsulation module 1110 can be used to perform the operation S210 described above, which will not be repeated here.

[0191] The deployment module 1120 is used to deploy the artificial intelligence model service to the production environment according to the model application program interface. In one embodiment, the deployment module 1120 can be used to perform the operation S220 described above, which will not be repeated here.

[0192] The integration module 1130 is used to integrate the artificial intelligence model service into the workflow engine of the low-code platform based on the production environment and using the model application program interface to obtain an artificial intelligence low-code system. In one embodiment, the integration module 1130 can be used to perform the operation S230 described above, which will not be repeated here.

[0193] The code generation module 1140 is used to automatically generate application code based on user application requirements using the artificial intelligence low-code system. In one embodiment, the code generation module 1140 can be used to perform the operation S240 described above, which will not be repeated here.

[0194] According to an embodiment of the present disclosure, the code generation apparatus 1100 further includes a cloud deployment module 1150 .

[0195] The cloud deployment module 1150 is used to deploy the artificial intelligence low-code system into a cloud computing environment so as to access the artificial intelligence low-code system via the Internet.

[0196] According to an embodiment of the present disclosure, the integration module 1130 includes: a first integration sub-module, used to build a test environment based on the production environment; a second integration sub-module, used to integrate the artificial intelligence model service into the workflow engine of the low-code platform by using the model application program interface; a third integration sub-module, used to use the integrated low-code platform as the system to be tested; a fourth integration sub-module, used to perform a comprehensive performance test on the system to be tested based on the test environment; and a fifth integration sub-module, used to deploy the system to be tested that has passed the comprehensive performance test into the production environment to obtain an artificial intelligence low-code system.

[0197] According to an embodiment of the present disclosure, the second integration sub-module includes a first integration unit for configuring an artificial intelligence big model module corresponding to the artificial intelligence model service in the low-code platform according to the model application program interface; a second integration unit for designing a user interface corresponding to the artificial intelligence big model module; and a third integration unit for integrating the artificial intelligence model service into the workflow engine of the low-code platform based on the user interface.

[0198] According to an embodiment of the present disclosure, the fourth integrated sub-module includes a fourth integrated unit, which is used to perform interface association testing on the model application interface in the system under test; a fifth integrated unit, which is used to perform layered testing on the artificial intelligence low-code system; a sixth integrated unit, which is used to perform system quality testing on the artificial intelligence low-code system; and a seventh integrated unit, which is used to perform maintenance testing on the artificial intelligence low-code system.

[0199] According to an embodiment of the present disclosure, the encapsulation module 1110 includes a first encapsulation sub-module, which is used to use a serialization tool to serialize a pre-trained artificial intelligence large model to obtain a serialized model file; a second encapsulation sub-module, which is used to create a service application corresponding to the artificial intelligence model service; wherein, the service application is set to load the serialized model file when the artificial intelligence model service is started; and a third encapsulation sub-module, which is used to use a development tool to set an application program interface and an application program interface endpoint, and build an application program interface corresponding to the artificial intelligence model service to obtain a model application program interface.

[0200] According to an embodiment of the present disclosure, the second packaging sub-module includes a first packaging unit, which is used to select a service architecture corresponding to the artificial intelligence model service; a second packaging unit, which is used to set the server environment and install the dependency library of the pre-trained artificial intelligence large model on the server; and a third packaging unit, which is used to create a service application corresponding to the artificial intelligence model service based on the service architecture and the server environment.

[0201] According to an embodiment of the present disclosure, the deployment module 1120 includes a first deployment unit for performing security authentication on the model application program interface; a second deployment unit for calling the model application program interface based on the expected request volume to perform performance testing on the artificial intelligence model service; a third deployment unit for optimizing the artificial intelligence model service according to the results of the performance test; a fourth deployment unit for setting error handling logic in the service application; a fifth deployment unit for setting the application program interface document corresponding to the model application program interface; a sixth deployment unit for performing layered testing on the model application program interface; and a seventh deployment unit for deploying the artificial intelligence model service to the production environment according to the application program interface document after passing the layered test.

[0202] According to an embodiment of the present disclosure, the deployment module 1120 also includes an eighth deployment unit, which is used to set up a monitoring and logging mechanism for the artificial intelligence model service; a ninth deployment unit, which is used to detect the running status of the artificial intelligence model service; and a tenth deployment unit, which is used to update the pre-trained artificial intelligence large model and model application program interface according to business needs.

[0203] According to an embodiment of the present disclosure, any multiple modules of the encapsulation module 1110, the deployment module 1120, the integration module 1130, the code generation module 1140, and the cloud deployment module 1150 can be combined into one module for implementation, or any one of the modules can be split into multiple modules. Alternatively, at least part of the functions of one or more of these modules can be combined with at least part of the functions of other modules and implemented in one module. According to an embodiment of the present disclosure, at least one of the encapsulation module 1110, the deployment module 1120, the integration module 1130, the code generation module 1140, and the cloud deployment module 1150 can be at least partially implemented as a hardware circuit, such as a field programmable gate array (FPGA), a programmable logic array (PLA), a system on a chip, a system on a substrate, a system on a package, an application specific integrated circuit (ASIC), or can be implemented by hardware or firmware such as any other reasonable way of integrating or packaging the circuit, or implemented in any one of the three implementation methods of software, hardware, and firmware, or in an appropriate combination of any of them. Alternatively, at least one of the packaging module 1110, the deployment module 1120, the integration module 1130, the code generation module 1140 and the cloud deployment module 1150 can be at least partially implemented as a computer program module, and when the computer program module is executed, the corresponding function can be performed.

[0204] Fig.12 A block diagram of an electronic device suitable for implementing a code generation method based on an artificial intelligence low-code system according to an embodiment of the present disclosure is schematically shown.

[0205] like Fig.12 As shown, the electronic device 1200 according to an embodiment of the present disclosure includes a processor 1201, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 1202 or a program loaded from a storage part 1208 to a random access memory (RAM) 1203. The processor 1201 may include, for example, a general-purpose microprocessor (e.g., a CPU), an instruction set processor and / or a related chipset and / or a dedicated microprocessor (e.g., an application-specific integrated circuit (ASIC)), etc. The processor 1201 may also include an onboard memory for caching purposes. The processor 1201 may include a single processing unit or multiple processing units for performing different actions of the method flow according to an embodiment of the present disclosure.

[0206] In RAM 1203, various programs and data required for the operation of electronic device 1200 are stored. Processor 1201, ROM 1202 and RAM 1203 are connected to each other via bus 1204. Processor 1201 performs various operations of the method flow according to the embodiment of the present disclosure by executing the program in ROM 1202 and / or RAM 1203. It should be noted that the program can also be stored in one or more memories other than ROM 1202 and RAM 1203. Processor 1201 can also perform various operations of the method flow according to the embodiment of the present disclosure by executing the program stored in one or more memories.

[0207] According to an embodiment of the present disclosure, the electronic device 1200 may further include an input / output (I / O) interface 1205, which is also connected to the bus 1204. The electronic device 1200 may further include one or more of the following components connected to the input / output (I / O) interface 1205: an input portion 1206 including a keyboard, a mouse, etc.; an output portion 1207 including a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker, etc.; a storage portion 1208 including a hard disk, etc.; and a communication portion 1209 including a network interface card such as a LAN card, a modem, etc. The communication portion 1209 performs communication processing via a network such as the Internet. A drive 1210 is also connected to the input / output (I / O) interface 1205 as needed. A removable medium 1211, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 1210 as needed, so that a computer program read therefrom is installed into the storage portion 1208 as needed.

[0208] The present disclosure also provides a computer-readable storage medium, which may be included in the device / apparatus / system described in the above embodiments; or may exist independently without being assembled into the device / apparatus / system. The above computer-readable storage medium carries one or more programs, and when the above one or more programs are executed, the method according to the embodiment of the present disclosure is implemented.

[0209] According to an embodiment of the present disclosure, a computer-readable storage medium may be a non-volatile computer-readable storage medium, for example, may include but is not limited to: a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In the present disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program, which may be used by or in combination with an instruction execution system, an apparatus, or a device. For example, according to an embodiment of the present disclosure, a computer-readable storage medium may include the ROM 1202 and / or RAM 1203 described above and / or one or more memories other than ROM 1202 and RAM 1203.

[0210] The embodiment of the present disclosure also includes a computer program product, which includes a computer program, and the computer program includes a program code for executing the method shown in the flowchart. When the computer program product is run in a computer system, the program code is used to enable the computer system to implement the code generation method based on the artificial intelligence low-code system provided by the embodiment of the present disclosure.

[0211] The above functions defined in the system / device of the embodiment of the present disclosure are performed when the computer program is executed by the processor 1201. According to the embodiment of the present disclosure, the system, device, module, unit, etc. described above can be implemented by a computer program module.

[0212] In one embodiment, the computer program may rely on tangible storage media such as optical storage devices, magnetic storage devices, etc. In another embodiment, the computer program may also be transmitted and distributed in the form of signals on a network medium, and downloaded and installed through the communication part 1209, and / or installed from the removable medium 1211. The program code contained in the computer program may be transmitted using any appropriate network medium, including but not limited to: wireless, wired, etc., or any suitable combination of the above.

[0213] In such an embodiment, the computer program can be downloaded and installed from the network through the communication part 1209, and / or installed from the removable medium 1211. When the computer program is executed by the processor 1201, the above functions defined in the system of the embodiment of the present disclosure are performed. According to the embodiment of the present disclosure, the system, device, means, module, unit, etc. described above can be implemented by a computer program module.

[0214] According to an embodiment of the present disclosure, the program code for executing the computer program provided by the embodiment of the present disclosure can be written in any combination of one or more programming languages. Specifically, these computing programs can be implemented using high-level process and / or object-oriented programming languages, and / or assembly / machine languages. Programming languages ​​include, but are not limited to, Java, C++, python, "C" language or similar programming languages. The program code can be executed entirely on the user computing device, partially on the user device, partially on the remote computing device, or entirely on the remote computing device or server. In the case of a remote computing device, the remote computing device can be connected to the user computing device through any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computing device (for example, using an Internet service provider to connect through the Internet).

[0215] The flow charts and block diagrams in the accompanying drawings illustrate the possible architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present disclosure. In this regard, each box in the flow chart or block diagram can represent a module, a program segment, or a part of a code, and the above-mentioned module, program segment, or a part of a code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram or flow chart, and the combination of the boxes in the block diagram or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0216] It will be appreciated by those skilled in the art that the features described in the various embodiments of the present disclosure may be combined and / or combined in a variety of ways, even if such combinations or combinations are not explicitly described in the present disclosure. In particular, without departing from the spirit and teachings of the present disclosure, the features described in the various embodiments of the present disclosure may be combined and / or combined in a variety of ways. All of these combinations and / or combinations fall within the scope of the present disclosure.

Claims

1. A code generation method based on an artificial intelligence low-code system, characterized in that: The method comprises: Encapsulate pre-trained AI models into AI model services to obtain model application programming interfaces; Deploy the artificial intelligence model service into a production environment according to the model application program interface; Based on the production environment, using the model application program interface, integrating the artificial intelligence model service into the workflow engine of the low-code platform to obtain an artificial intelligence low-code system; and The artificial intelligence low-code system is used to automatically generate application code based on user application requirements.

2. The method according to claim 1, characterized in that Based on the production environment, the artificial intelligence model service is integrated into the workflow engine of the low-code platform by using the model application program interface to obtain an artificial intelligence low-code system, including: Based on the production environment, build a test environment; Integrate the artificial intelligence model service into the workflow engine of the low-code platform using the model application program interface; Use the integrated low-code platform as the system under test; Based on the test environment, performing a comprehensive performance test on the system to be tested; and The system under test that has passed the comprehensive performance test is deployed into the production environment to obtain the artificial intelligence low-code system.

3. The method according to claim 2, characterized in that The utilizing of the model application program interface to integrate the artificial intelligence model service into the workflow engine of the low-code platform includes: According to the model application program interface, an artificial intelligence large model module corresponding to the artificial intelligence model service is configured in the low-code platform; Designing a user interface corresponding to the artificial intelligence large model module; and Based on the user interface, the artificial intelligence model service is integrated into the workflow engine of the low-code platform.

4. The method according to claim 2, characterized in that: The comprehensive performance test of the system to be tested based on the test environment includes: Performing interface association testing on the model application program interface in the system to be tested; Conduct layered testing on the AI ​​low-code system; Conduct system quality testing on the AI ​​low-code system; and Perform maintenance testing on the artificial intelligence low-code system.

5. The method according to claim 1, characterized in that The method of packaging the pre-trained artificial intelligence model into an artificial intelligence model service to obtain a model application program interface includes: Using a serialization tool, serialize the pre-trained artificial intelligence large model to obtain a serialized model file; Creating a service application corresponding to the artificial intelligence model service, wherein the service application is configured to load the serialized model file when the artificial intelligence model service is started; and Using development tools, set the application program interface and application program interface endpoints, and build the application program interface corresponding to the artificial intelligence model service to obtain the model application program interface.

6. The method according to claim 5, characterized in that The step of creating a service application corresponding to the artificial intelligence model service includes: Select a service architecture corresponding to the artificial intelligence model service; Setting up a server environment and installing a dependency library of the pre-trained artificial intelligence large model on the server; and Based on the service architecture and the server environment, create a service application corresponding to the artificial intelligence model service.

7. The method according to any one of claims 1 to 6, characterized in that: The step of deploying the artificial intelligence model service to a production environment according to the model application program interface includes: Performing security authentication on the model application program interface; Based on the expected request volume, calling the model application program interface to perform performance testing on the artificial intelligence model service; Optimizing the artificial intelligence model service based on the results of the performance test; Setting error handling logic in the service application; Setting the application program interface document corresponding to the model application program interface; Performing layered testing on the model application program interface; and After passing the layered test, the artificial intelligence model service is deployed to the production environment according to the application interface document.

8. The method according to claim 1, characterized in that The step of deploying the artificial intelligence model service to a production environment according to the model application program interface further includes: Setting up monitoring and logging mechanisms for the artificial intelligence model service; Detecting the operating status of the artificial intelligence model service; and Update the pre-trained artificial intelligence big model and the model application program interface according to business needs.

9. The method according to claim 1, characterized in that: The method further comprises: Deploy the artificial intelligence low-code system in a cloud computing environment to access the artificial intelligence low-code system via the Internet.

10. A code generation device based on an artificial intelligence low-code system, characterized in that: The device comprises: A packaging module is used to package a pre-trained large AI model into an AI model service to obtain a model application program interface; A deployment module, used to deploy the artificial intelligence model service into a production environment according to the model application program interface; An integration module, for integrating the artificial intelligence model service into the workflow engine of the low-code platform based on the production environment and utilizing the model application program interface to obtain an artificial intelligence low-code system; and The code generation module is used to utilize the artificial intelligence low-code system to automatically generate application code based on user application requirements.

11. An electronic device, comprising: one or more processors; a memory for storing one or more computer programs, It is characterized in that the one or more processors execute the one or more computer programs to implement the steps of the method according to any one of claims 1 to 9.

12. A computer-readable storage medium having a computer program or instruction stored thereon, characterized in that: When the computer program or instruction is executed by a processor, the steps of the method according to any one of claims 1 to 9 are implemented.

13. A computer program product comprising a computer program or instructions, characterized in that When the computer program or instruction is executed by a processor, the steps of the method according to any one of claims 1 to 9 are implemented.

Images