Question and answer service creation method and device, computer equipment and storage medium

Abstract

The invention relates to a question and answer service creation method and device, computer equipment and a storage medium. The method comprises the steps of automatically configuring basic configuration, network configuration and system configuration of a target question and answer service according to a question and answer service creation instruction initiated by a target account, and automatically obtaining a mirror image data packet corresponding to a target large model, the target question and answer service is automatically constructed according to the mirror image data packet corresponding to the target large model, the basic configuration, the network configuration and the system configuration, the question and answer service does not need to be created through manual coding, and therefore the problems that data safety and operation stability cannot be guaranteed through existing manual question and answer service creation, and a large amount of time and energy need to be consumed can be solved.

Description

Technical Field

[0001] The present application relates to the field of question-answering service creation, and in particular to a question-answering service creation method, apparatus, computer equipment, and storage medium. Background Art

[0002] With the development of AI technology, more and more users need to use AI technology to solve problems in daily travel, life, work and other scenarios. If users want to create their own question-and-answer services that can respond to interactive questions and answers, they need to manually write code to build cloud hosts, large model databases and service frameworks. However, manual construction cannot guarantee data security and operational stability, and requires a lot of time and effort. Summary of the Invention

[0003] The present application provides a method, apparatus, computer device and storage medium for creating a question-and-answer service to solve the problem that the existing manual creation of question-and-answer services cannot guarantee data security and operational stability, and requires a lot of time and effort.

[0004] In a first aspect, the present application provides a method for creating a question-and-answer service, the method comprising:

[0005] Obtaining a question-and-answer service creation instruction initiated by a target account, wherein the question-and-answer service creation instruction is used to request creation of a target question-and-answer service;

[0006] Determining the basic configuration, network configuration, and system configuration of the target question-answering service according to the question-answering service creation instruction, wherein the basic configuration of the target question-answering service includes a target macro model that supports implementation of the target question-answering service;

[0007] Obtaining a mirror data package corresponding to the target large model, and constructing a target question-answering service based on the mirror data package corresponding to the target large model, the basic configuration, network configuration, and system configuration of the target question-answering service, wherein the mirror data package includes framework data required by the target large model;

[0008] Feedback the access address of the target question-and-answer service to the target account, so that the target account can access the target question-and-answer service through the access address.

[0009] In a second aspect, the present application provides a device for creating a question-and-answer service, the device comprising:

[0010] An acquisition module, configured to acquire a question-and-answer service creation instruction initiated by a target account, wherein the question-and-answer service creation instruction is used to request creation of a target question-and-answer service;

[0011] a configuration module, configured to determine the basic configuration, network configuration, and system configuration of the target question-answering service according to the question-answering service creation instruction, wherein the basic configuration of the target question-answering service includes a target macro model supporting implementation of the target question-answering service;

[0012] a construction module, configured to obtain a mirror data packet corresponding to the target large model, and construct a target question-answering service based on the mirror data packet corresponding to the target large model, the basic configuration, network configuration, and system configuration of the target question-answering service, wherein the mirror data packet includes framework data required by the target large model;

[0013] The feedback module is configured to feed back the access address of the target question-and-answer service to the target account, so that the target account can access the target question-and-answer service through the access address.

[0014] In a third aspect, the present application provides a computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the above-mentioned question-and-answer service creation method when executing the computer program.

[0015] In a fourth aspect, the present application also provides a computer storage medium storing computer executable instructions, wherein the computer executable instructions are used to execute the above-mentioned question-and-answer service creation method.

[0016] The above-mentioned technical solution provided by the embodiment of the present application has the following advantages over the existing technology: the method provided by the embodiment of the present application obtains a question-and-answer service creation instruction initiated by a target account, wherein the question-and-answer service creation instruction is used to request the creation of a target question-and-answer service; determines the basic configuration, network configuration and system configuration of the target question-and-answer service according to the question-and-answer service creation instruction, wherein the basic configuration of the target question-and-answer service includes a target big model that supports the implementation of the target question-and-answer service; obtains a mirror data packet corresponding to the target big model, and constructs the target question-and-answer service based on the mirror data packet corresponding to the target big model, the basic configuration, network configuration and system configuration of the target question-and-answer service, wherein the mirror data packet includes the framework data required by the target big model; and feeds back the access address of the target question-and-answer service to the target account, so that the target account can access the target question-and-answer service through the access address.

[0017] Based on the above method, the basic configuration, network configuration and system configuration of the target Q&A service can be automatically configured according to the Q&A service creation instructions initiated by the target account, and the image data package corresponding to the target large model can be automatically obtained. Then, the target Q&A service can be automatically built according to the image data package, basic configuration, network configuration and system configuration corresponding to the target large model. There is no need to manually code to create the Q&A service, which can solve the problem that the existing manually created Q&A service cannot guarantee data security and operational stability, and requires a lot of time and effort. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0019] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0020] One or more embodiments are exemplarily illustrated by pictures in the corresponding drawings. These exemplifications do not constitute limitations on the embodiments. Elements with the same reference numerals in the drawings are represented as similar elements. Unless otherwise stated, the figures in the drawings do not constitute proportional limitations.

[0021] Figure 1 An application environment diagram of a question-and-answer service creation method provided in an embodiment of the present application;

[0022] Figure 2 A flowchart of a method for creating a question-and-answer service provided in an embodiment of the present application;

[0023] Figure 3 A flowchart of a method for creating a question-and-answer service provided in an embodiment of the present application;

[0024] Figure 4 A schematic diagram of the display effect of the service creation interface provided in an embodiment of the present application;

[0025] Figure 5 A schematic diagram of the display effect of the service creation interface provided in an embodiment of the present application;

[0026] Figure 6 A schematic diagram of the display effect of the service creation interface provided in an embodiment of the present application;

[0027] Figure 7 A flowchart of a method for creating a question-and-answer service provided in an embodiment of the present application;

[0028] Figure 8 A flowchart of a method for creating a question-and-answer service provided in an embodiment of the present application;

[0029] Figure 9 A flowchart of a method for creating a question-and-answer service provided in an embodiment of the present application;

[0030] Figure 10 A schematic diagram of the display effect of the service creation interface provided in an embodiment of the present application;

[0031] Figure 11 A flowchart of a method for creating a question-and-answer service provided in an embodiment of the present application;

[0032] Figure 12 A schematic diagram of the display effect of the service creation interface provided in an embodiment of the present application;

[0033] Figure 13 A flowchart of a method for creating a question-and-answer service provided in an embodiment of the present application;

[0034] Figure 14 A schematic diagram of the display effect of the service creation interface provided in an embodiment of the present application;

[0035] Figure 15 A schematic diagram of the display effect of the service creation interface provided in an embodiment of the present application;

[0036] Figure 16 A schematic diagram of the display effect of the question-and-answer service interface provided in an embodiment of the present application;

[0037] Figure 17 A schematic diagram of the display effect of the question-and-answer service interface provided in an embodiment of the present application;

[0038] Figure 18 A schematic diagram of the display effect of the question-and-answer service interface provided in an embodiment of the present application;

[0039] Figure 19 A schematic diagram of the display effect of the question-and-answer service interface provided in an embodiment of the present application;

[0040] Figure 20 A structural block diagram of a question-and-answer service creation device provided in an embodiment of the present application;

[0041] Figure 21 A schematic diagram of the internal structure of a computer device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0042] To make the purpose, technical solutions, and advantages of the embodiments of this application more clear, the technical solutions in the embodiments of this application will be clearly and completely described below in conjunction with the drawings in the embodiments of this application. Obviously, the described embodiments are part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0043] The disclosure below provides many different embodiments or examples for implementing different configurations of the present invention. To simplify the disclosure of the present invention, the components and configurations of specific examples are described below. Of course, these are merely examples and are not intended to limit the present invention. In addition, the present invention may repeat reference numerals and / or letters in different examples. Such repetition is for the purpose of simplicity and clarity and does not in itself indicate the relationship between the various embodiments and / or configurations discussed.

[0044] Figure 1 This is an application environment diagram of a question-answering service creation method in an embodiment. Figure 1 The question-and-answer service creation method is applied to a question-and-answer service creation system. The question-and-answer service creation system includes a terminal 110 and a service creation console 120. Terminal 110 and service creation console 120 are connected via a network. Terminal 110 can be a desktop terminal or a mobile terminal. The mobile terminal can be at least one of a mobile phone, a tablet computer, and a laptop computer. Service creation console 120 can be implemented using a standalone server or a server cluster consisting of multiple servers.

[0045] In one embodiment, Figure 2 A flowchart of a method for creating a question-answering service in one embodiment is shown in FIG. Figure 2 , provides a question-answering service creation method. This embodiment mainly applies this method to the above Figure 1 Taking the service creation console 120 in FIG. 1 as an example, the question-answering service creation method specifically includes the following steps:

[0046] Step S210: obtaining a question-and-answer service creation instruction initiated by a target account, wherein the question-and-answer service creation instruction is used to request creation of a target question-and-answer service.

[0047] Specifically, the target account refers to any user account that accesses the service creation console 120 through the terminal 110. The question-and-answer service creation instruction includes multiple configuration requests. Different configuration requests are used to request the configuration of corresponding creation parameters. Multiple creation parameters are used to build a target question-and-answer service. The target question-and-answer service can be any question-and-answer service, denoted as an AIGC (Artificial Intelligence Generated Content) service. The question-and-answer service is used to generate content with certain creativity and quality using artificial intelligence algorithms. That is, through training models and learning from a large amount of data, the question-and-answer service can generate relevant content based on the input conditions or guidance. For example, by entering keywords, descriptions, or samples, the question-and-answer service can generate matching articles, images, audio, etc.

[0048] Step S220: determining the basic configuration, network configuration, and system configuration of the target question-answering service according to the question-answering service creation instruction, wherein the basic configuration of the target question-answering service includes a target macro model supporting the implementation of the target question-answering service.

[0049] Specifically, the basic configuration includes the essential parameters required to create the target Q&A service, such as the data center type, billing method, availability zone, model deployment method, model configuration, and platform configuration. The model configuration includes the target big model supporting the target Q&A service, the target vector model associated with the target big model, and the target vector database. The platform configuration includes the target server model used to support the target Q&A service. The target big model can be a third-party big model provided by a partnered third-party platform, or any open source big model.

[0050] The network configuration contains the creation parameters required to communicate with the target Q&A service, and the system configuration contains the creation parameters required to log in to the target Q&A service.

[0051] Step S230, obtain the mirror data package corresponding to the target large model, and build the target question-and-answer service based on the mirror data package corresponding to the target large model, the basic configuration, network configuration and system configuration of the target question-and-answer service, wherein the mirror data package includes the framework data required by the target large model.

[0052] Specifically, different large models correspond to different mirror data packages. The service creation console 120 packages different mirror data packages in advance for different large model deployments. The mirror data package specifically includes the model data of the large model, the model data of the vector model (embedding model) associated with the large model, and the framework data supporting the large model (framework data of the LangChain service). That is, the service creation console 120 pre-installs and adapts the entire langchain service, cuda and python package environment, large model and other services and environments, and packages them into a mirror data package, so that users can select and create the corresponding question and answer service with one click through the service creation console 120, and directly use the entire question and answer service that has been adjusted, without the need for users to write code themselves to create the target question and answer service.

[0053] Step S240: Feedback the access address of the target question-and-answer service to the target account, so that the target account can access the target question-and-answer service through the access address.

[0054] Specifically, after the service creation console 120 completes the construction of the target Q&A service, it will feed back the access address of the target Q&A service to the target account. The target account can successfully access the target Q&A service according to the access address, and can use the target Q&A service to generate reply content related to the input content.

[0055] Based on the above method, the basic configuration, network configuration and system configuration of the target Q&A service can be automatically configured according to the Q&A service creation instructions initiated by the target account, and the image data package corresponding to the target large model can be automatically obtained. Then, the target Q&A service can be automatically built according to the image data package, basic configuration, network configuration and system configuration corresponding to the target large model. There is no need to manually code to create the Q&A service, which can solve the problem that the existing manually created Q&A service cannot guarantee data security and operational stability, and requires a lot of time and effort.

[0056] In one embodiment, Figure 3 As shown, the determining of the basic configuration of the target question-and-answer service according to the question-and-answer service creation instruction includes:

[0057] Step S2210: determining a model deployment mode in the question-and-answer service creation interface according to the first selected instruction in the question-and-answer service creation instruction;

[0058] Step S2220: determining the target large model in the large model cluster provided by the model deployment method according to the second selected instruction in the question-answering service creation instruction;

[0059] Step S2230: determining a target vector model in a vector model cluster associated with the target large model according to the third selected instruction in the question-answering service creation instruction;

[0060] Step S2240: determining a target vector database in the vector database cluster provided by the model deployment method according to the fourth selected instruction in the question-answering service creation instruction;

[0061] Step S2250: According to the fifth selected instruction in the question-and-answer service creation instruction, determine the target server in the server cluster that supports deployment of the target large model, wherein the basic configuration also includes the target vector model, the target vector database and the target server, and the target server is used to deploy the target large model, the target vector model and the target vector database.

[0062] Specifically, after the user accesses the service creation console 120 through the terminal 110, the service creation console 120 will provide the terminal 110 with a question-and-answer service creation interface, such as Figure 4 As shown, the service creation console 120 determines the billing method, data center type, availability zone, model deployment method, model configuration and platform configuration in the Q&A service creation interface according to the Q&A service creation instruction initiated by the user. The model deployment method is divided into a first deployment method and a second deployment method. The first deployment method corresponds to Figure 4 The public call in is used to call the third-party large model with a cooperative relationship. The second deployment method corresponds to Figure 4 Private deployment in [1] is used to use open source large models. Different model deployment methods correspond to different model configurations. The model configuration specifically includes the large model, vector model, group identifier (Group ID), interface key (APIKey), and vector database (milvus) corresponding to the model deployment method. The vector database can achieve high availability of cross-regional data backup and implement a sentinel mechanism to automatically switch between active and standby servers to ensure the stability of customer business operations. The platform configuration includes the server model used to implement the question-and-answer service.

[0063] That is, when the user selects the model deployment method, the model configuration only provides the large model cluster corresponding to the model deployment method, and the target large model is determined according to the user's selection result in the large model cluster. When the target large model is selected, the model configuration only provides the vector model associated with the target large model, and the target vector model is determined according to the user's selection result in the vector model cluster associated with the target large model, such as Figure 5 As shown in FIG, when the user selects a general large model as the target large model, only the vector models associated with the target large model are displayed, such as Figure 6As shown, when the user selects a base large model as the target large model, only the vector models associated with the target large model are displayed. The target vector database is also determined based on the user's selection in the vector database cluster provided in the model configuration. Furthermore, the target server is determined based on the user's selection in the server cluster. The target server is used to deploy the image data package.

[0064] All basic configurations only need to be selected by the user on demand, without the need to manually code to build large models, vector models, and vector databases. This avoids the inability of users to manually code to guarantee code performance and stable operation of large model businesses, and also avoids the inability of users to manually code to build vector databases that cannot guarantee data security and stability. It also avoids the incompatibility issues between Python packages and CUDA environments caused by users manually deploying Langchain services, as well as problems such as adaptation and debugging with large models, and saves users a lot of time and energy consumed by manual coding.

[0065] In one embodiment, Figure 7 As shown, the determining the target large model in the large model cluster provided by the model deployment method according to the second selected instruction in the question-answering service creation instruction includes:

[0066] Step S22201, when the model deployment mode is the first deployment mode, according to the second selected instruction in the question-answering service creation instruction, determine the target large model in the third-party large model cluster provided by the first deployment mode, wherein the third-party large model cluster includes multiple third-party large models; or

[0067] Step S22202, when the model deployment mode is the second deployment mode, according to the second selected instruction in the question-answering service creation instruction, determine the target large model in the open source large model cluster provided by the second deployment mode.

[0068] Specifically, when the user selects the first deployment method, that is, the public call method, for model deployment, the model configuration only displays the third-party large model cluster provided by the third-party platform cooperating with the service creation console 120. The third-party large model cluster includes multiple third-party large models. The third-party large model can specifically be a Minimax model or a Zhipu model. When the user selects the Minimax model as the target large model, the user needs to enter the group identifier and interface key, while when the user selects the Zhipu model as the target large model, the user only needs to enter the interface key.

[0069] When the user selects the second deployment mode, that is, private deployment mode, the model configuration only displays the open source big model cluster except for the third-party big model. The open source big model cluster includes multiple open source big models. The open source big models can be general big models, code big models, mathematical reasoning big models, medical big models, etc. General big models include Chatglm2-6b model, Llama2-7b model, Baichuan2-13B model, code big models include SQLCoder model and WizardCoder-Python-34B model, mathematical reasoning big model is WizardLM-70B-V1.0, and medical big model is BioMedGPT-10B.

[0070] Based on the user's needs to build a target question-answering service, different available large models can be selected without having to manually build a large model, thus saving the user the time and resources spent on building a large model.

[0071] In one embodiment, Figure 8 As shown, the step of determining a target server in a server cluster supported for deployment by the target large model according to the fifth selected instruction in the question-answering service creation instruction includes:

[0072] Step S22510, when the model deployment mode is the first deployment mode, determining a target server in a virtual server cluster supported for deployment by the target large model according to the fifth selected instruction in the question-answering service creation instruction, wherein the target server includes at least one virtual server in the virtual server cluster; or

[0073] Step S22520, when the model deployment mode is the second deployment mode, according to the fifth selected instruction in the question-answering service creation instruction, determine the target server in the graphics processing server cluster that supports deployment of the target large model.

[0074] Specifically, when the user selects the first deployment mode for model deployment, a virtual server cluster that supports the deployment of the target large model is provided to the user. For example, the ChatGLM-6b large model is allowed to be deployed on a 1*T4 machine, while Llama2-7B needs to be deployed on a 2*T4 machine. Multiple virtual servers in the virtual server cluster are displayed through the service creation interface for the user to select a target server from multiple virtual servers. The target server can include at least one virtual server, and the user can also select multiple virtual servers as target servers. However, the number of multiple virtual servers selected by the user must be less than or equal to the preset number. Assuming the preset number is 3, it means that the user can select up to three virtual servers as target servers, thereby supporting the user to create 1-3 cloud hosts (virtual servers) for one service, ensuring the high availability of the large model.

[0075] When the user selects the second deployment mode for model deployment, multiple graphics processing servers in the graphics processing server cluster will be displayed through the service creation interface, so that the user can select the target server from multiple graphics processing servers. The graphics processing server is recorded as a GPU server. GPU servers of different specifications will only be used in the private deployment mode, and there is no need to use the GPU server in the public call mode, thereby saving the service resources of the GPU server.

[0076] In one embodiment, Figure 9 As shown, determining the network configuration of the target question-and-answer service according to the question-and-answer service creation instruction includes:

[0077] Step S2260: When a target vector database is determined in the vector database cluster provided by the model deployment method based on the question-answering service creation instruction, a network configuration corresponding to the target vector database is determined as the network configuration of the target question-answering service.

[0078] Specifically, the network configuration of the target question-answering service needs to be compatible with the target vector database. Therefore, when the user selects the target vector database, the service creation console 120 will automatically select the network configuration that is compatible with the target vector database and determine it as the network configuration of the target question-answering service, thereby avoiding the problem of incompatibility between the target vector database and the network configuration. Figure 10 As shown in the figure, the network configuration specifically includes the association between the bound elastic IP and shared bandwidth, line type, bandwidth value, primary network card, associated subnet, IP address, associated security group and other creation parameters.

[0079] In one embodiment, Figure 11 As shown, the determining the system configuration of the target question-and-answer service according to the question-and-answer service creation instruction includes:

[0080] Step S2270: Determine the server name, login method, and login information of the target question-and-answer service according to the question-and-answer service creation instruction, wherein the system configuration of the target question-and-answer service includes the server name, login method, and login information.

[0081] Specifically, the system configuration is also formed based on the user's selection results in the service creation interface, such as Figure 12As shown, system configuration specifically includes the server name, login method, and login information. If the user selects multiple servers as target servers, they can create different names for each server by adding a suffix to the server name. Login methods include password, key, and post-creation settings, and login information includes the administrator account and password. This eliminates the need for manual configuration, saving time and manpower in creating target Q&A services.

[0082] In one embodiment, Figure 13 As shown, feeding back the access address of the target question-and-answer service to the target account includes:

[0083] Step S2410: Upon receiving the service order information submitted by the target account for the target Q&A service, the administrator access address and the user access address of the target Q&A service are fed back to the target account, wherein the access address of the target Q&A service includes the administrator access address and the user access address, and the administrator access address and the user access address correspond to different access permissions of the target Q&A service.

[0084] Specifically, the service creation console 120 automatically selects the corresponding mirror data package according to the target large model selected by the user, and automatically passes the IP port, account password and other information of the mlivus target vector database to the corresponding configuration file of the LangChain service through cloud-init. When the service creation console 120 receives the service order information of the target account for the target question and answer service, it automatically starts the LangChain service, large model service, and embedding service to form the target question and answer service, and exposes a port to be displayed on the front end of the service creation console 120, that is, the access address corresponding to the port is fed back to the target account, and the basic information, configuration information, and network information of the target question and answer service are also fed back to the target account, and displayed together on the front end of the service creation console 120, such as Figure 14 As shown. Figure 15 As shown, the service order information represents the user's rental order for the target server that supports the target Q&A service. The service creation console 120 feeds back the access address of the target Q&A service to the target account only after receiving the service order information, so that the target account can use the target Q&A service.

[0085] The access address corresponding to the port is displayed through the service creation interface. The access address includes the administrator access address and the user access address for different objects to access the target Q&A service. Users access different access addresses to access the target Q&A service by calling APIs. Accessing the target Q&A service through different access addresses will have different access permissions. Administrators who access the target Q&A service through the administrator access address have the authority to configure the conversation mode, conversation settings, knowledge base management, and interface settings of the target Q&A service, while users who access the target Q&A service through the user access address can only configure the conversation mode and conversation settings.

[0086] like Figure 16 As shown in , the dialogue mode is divided into large model dialogue mode and knowledge base question answering mode, such as Figure 17 As shown in , the dialogue settings include the number of historical dialogue rounds, dialogue mode, knowledge base configuration, etc. Figure 18 As shown in the figure, knowledge base management specifically includes selecting a knowledge base, creating a new knowledge base, uploading knowledge files, and managing existing files. Figure 19 As shown, the interface settings specifically include setting the service name and setting the service icon (LoGo). The service icon consists of at least two parts, one is the text part "XXX" (indicating the company name, trademark name, etc.), and the other is the graphic part (an icon indicating the company or trademark, such as a circle, star, etc.).

[0087] Figure 2 、 3 , 7, 8, 9, 11, and 13 are flowcharts of a method for creating a question-answering service in one embodiment. Figure 2 、 3 The steps in the flowcharts of , 7, 8, 9, 11, and 13 are shown in sequence as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified in this document, there is no strict order restriction for the execution of these steps, and these steps can be executed in other orders. Moreover, Figure 2 、 3 At least part of the steps in 7, 8, 9, 11, and 13 may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily performed at the same time, but can be performed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be performed in turn or alternately with other steps or at least part of the sub-steps or stages of other steps.

[0088] In one embodiment, Figure 20 As shown, a question-answering service creation device is provided, comprising:

[0089] An acquisition module 310 is configured to acquire a question-and-answer service creation instruction initiated by a target account, wherein the question-and-answer service creation instruction is used to request creation of a target question-and-answer service;

[0090] A configuration module 320 is configured to determine the basic configuration, network configuration, and system configuration of the target question-answering service according to the question-answering service creation instruction, wherein the basic configuration of the target question-answering service includes a target macro model that supports the implementation of the target question-answering service;

[0091] A construction module 330 is configured to obtain a mirror data package corresponding to the target large model and construct a target question-answering service based on the mirror data package corresponding to the target large model, the basic configuration, network configuration, and system configuration of the target question-answering service, wherein the mirror data package includes the framework data required by the target large model;

[0092] The feedback module 340 is configured to feed back the access address of the target question-and-answer service to the target account, so that the target account can access the target question-and-answer service through the access address.

[0093] In one embodiment, the configuration module 320 is further configured to:

[0094] Determining a model deployment mode in a question-and-answer service creation interface according to a first selected instruction in the question-and-answer service creation instruction;

[0095] Determining the target large model in the large model cluster provided by the model deployment method according to the second selected instruction in the question-answering service creation instruction;

[0096] determining a target vector model in a vector model cluster associated with the target large model according to a third selected instruction in the question-answering service creation instruction;

[0097] Determining a target vector database in the vector database cluster provided by the model deployment method according to a fourth selected instruction in the question-answering service creation instruction;

[0098] According to the fifth selected instruction in the question-and-answer service creation instruction, a target server is determined in the server cluster supported for deployment of the target large model, wherein the basic configuration also includes the target vector model, the target vector database and the target server, and the target server is used to deploy the target large model, the target vector model and the target vector database.

[0099] In one embodiment, the configuration module 320 is further configured to:

[0100] When the model deployment mode is the first deployment mode, according to the second selected instruction in the question-answering service creation instruction, determining the target large model in the third-party large model cluster provided by the first deployment mode, wherein the third-party large model cluster includes multiple third-party large models; or

[0101] When the model deployment mode is the second deployment mode, the target large model is determined in the open source large model cluster provided by the second deployment mode according to the second selected instruction in the question-answering service creation instruction.

[0102] In one embodiment, the configuration module 320 is further configured to:

[0103] When the model deployment mode is the first deployment mode, according to the fifth selected instruction in the question-answering service creation instruction, a target server is determined in a virtual server cluster supported for deployment by the target large model, wherein the target server includes at least one virtual server in the virtual server cluster; or

[0104] When the model deployment mode is the second deployment mode, a target server is determined in a graphics processing server cluster that supports deployment of the target large model according to the fifth selected instruction in the question-answering service creation instruction.

[0105] In one embodiment, the configuration module 320 is further configured to:

[0106] When a target vector database is determined in the vector database cluster provided by the model deployment method based on the question-answering service creation instruction, a network configuration corresponding to the target vector database is determined as the network configuration of the target question-answering service.

[0107] In one embodiment, the configuration module 320 is further configured to:

[0108] The server name, login method and login information of the target question-and-answer service are determined according to the question-and-answer service creation instruction, wherein the system configuration of the target question-and-answer service includes the server name, login method and login information.

[0109] In one embodiment, the feedback module 340 is further configured to:

[0110] Upon receiving the service order information submitted by the target account for the target question-and-answer service, the administrator access address and the user access address of the target question-and-answer service are fed back to the target account, wherein the access address of the target question-and-answer service includes the administrator access address and the user access address, and the administrator access address and the user access address correspond to different access permissions of the target question-and-answer service.

[0111] like Figure 21 As shown, an embodiment of the present application provides a computer device, including a processor 711, a communication interface 712, a memory 713 and a communication bus 714, wherein the processor 711, the communication interface 712, and the memory 713 communicate with each other through the communication bus 714;

[0112] Memory 713, for storing computer programs;

[0113] The processor 711 is configured to implement the question-and-answer service creation method provided by any one of the aforementioned method embodiments when executing the program stored in the memory 713 .

[0114] Those skilled in the art will understand that Figure 21 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0115] In one embodiment, the question-answering service creation device provided by the present application can be implemented in the form of a computer program. Figure 21 The computer device can store various program modules that make up the question-answering service creation device, such as: Figure 20 The acquisition module 310, configuration module 320, construction module 330 and feedback module 340 are shown. The computer program composed of various program modules enables the processor to execute the question-answering service creation method of each embodiment of the present application described in this specification.

[0116] Figure 21 The computer device shown can be Figure 20The acquisition module 310 in the question-and-answer service creation device shown executes to acquire the question-and-answer service creation instruction initiated by the target account, wherein the question-and-answer service creation instruction is used to request the creation of the target question-and-answer service. The computer device can determine the basic configuration, network configuration and system configuration of the target question-and-answer service according to the question-and-answer service creation instruction through the configuration module 320, wherein the basic configuration of the target question-and-answer service includes the target large model that supports the implementation of the target question-and-answer service. The computer device can execute to acquire the mirror data packet corresponding to the target large model through the construction module 330, and construct the target question-and-answer service based on the mirror data packet corresponding to the target large model, the basic configuration, network configuration and system configuration of the target question-and-answer service, wherein the mirror data packet includes the framework data required by the target large model. The computer device can execute to feedback the access address of the target question-and-answer service to the target account through the feedback module 340, so that the target account can access the target question-and-answer service through the access address.

[0117] An embodiment of the present application further provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the question-and-answer service creation method provided in any of the aforementioned method embodiments is implemented.

[0118] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the modules may be selected based on actual needs to achieve the objectives of this embodiment.

[0119] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a general hardware platform, or of course by hardware. Based on this understanding, the above technical solution, in essence, or the part that contributes to the relevant technology, can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, an optical disk, etc., and includes a number of instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the question-and-answer service creation method described in each embodiment or certain parts of the embodiment.

[0120] It should be understood that the terms used herein are for the purpose of describing specific example embodiments only and are not intended to be limiting. Unless the context clearly indicates otherwise, the singular forms "one", "an" and "said" as used herein may also be meant to include plural forms. The terms "comprise", "include", "contain" and "have" are inclusive and therefore specify the presence of stated features, steps, operations, elements and / or parts, but do not exclude the presence or addition of one or more other features, steps, operations, elements, parts, and / or combinations thereof. The method steps, processes, and operations described herein are not to be construed as necessarily requiring them to be performed in the specific order described or illustrated, unless the order of execution is clearly indicated. It should also be understood that additional or alternative methods may be used.

[0121] The foregoing description is intended only to provide specific embodiments of the present invention, which will enable those skilled in the art to understand and implement the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention is not intended to be limited to the embodiments shown herein, but is intended to be accorded the widest scope consistent with the principles and novel features claimed herein.

Claims

1. A method for creating a question-answering service, characterized in that: The method comprises: Obtaining a question-and-answer service creation instruction initiated by a target account, wherein the question-and-answer service creation instruction is used to request creation of a target question-and-answer service; Determining the basic configuration, network configuration, and system configuration of the target question-answering service according to the question-answering service creation instruction, wherein the basic configuration of the target question-answering service includes a target macro model that supports implementation of the target question-answering service; Obtaining a mirror data package corresponding to the target large model, and constructing a target question-answering service based on the mirror data package corresponding to the target large model, the basic configuration, network configuration, and system configuration of the target question-answering service, wherein the mirror data package includes framework data required by the target large model; Feedback the access address of the target question-and-answer service to the target account, so that the target account can access the target question-and-answer service through the access address.

2. The question-answering service creation method according to claim 1, wherein: The determining the basic configuration of the target question-and-answer service according to the question-and-answer service creation instruction includes: Determining a model deployment mode in a question-and-answer service creation interface according to a first selected instruction in the question-and-answer service creation instruction; Determining the target large model in the large model cluster provided by the model deployment method according to the second selected instruction in the question-answering service creation instruction; determining a target vector model in a vector model cluster associated with the target large model according to a third selected instruction in the question-answering service creation instruction; Determining a target vector database in the vector database cluster provided by the model deployment method according to a fourth selected instruction in the question-answering service creation instruction; According to the fifth selected instruction in the question-and-answer service creation instruction, a target server is determined in the server cluster supported for deployment of the target large model, wherein the basic configuration also includes the target vector model, the target vector database and the target server, and the target server is used to deploy the target large model, the target vector model and the target vector database.

3. The question-answering service creation method according to claim 2, wherein: The step of determining the target large model in the large model cluster provided by the model deployment method according to the second selected instruction in the question-answering service creation instruction includes: When the model deployment mode is the first deployment mode, according to the second selected instruction in the question-answering service creation instruction, determining the target large model in the third-party large model cluster provided by the first deployment mode, wherein the third-party large model cluster includes multiple third-party large models; or When the model deployment mode is the second deployment mode, the target large model is determined in the open source large model cluster provided by the second deployment mode according to the second selected instruction in the question-answering service creation instruction.

4. The question-answering service creation method according to claim 2, wherein: Determining a target server in a server cluster supported for deployment by the target large model according to the fifth selected instruction in the question-answering service creation instruction includes: When the model deployment mode is the first deployment mode, according to the fifth selected instruction in the question-answering service creation instruction, a target server is determined in a virtual server cluster supported for deployment by the target large model, wherein the target server includes at least one virtual server in the virtual server cluster; or When the model deployment mode is the second deployment mode, a target server is determined in a graphics processing server cluster that supports deployment of the target large model according to the fifth selected instruction in the question-answering service creation instruction.

5. The question-answering service creation method according to claim 2, wherein: The determining the network configuration of the target question-and-answer service according to the question-and-answer service creation instruction includes: When a target vector database is determined in the vector database cluster provided by the model deployment method based on the question-answering service creation instruction, a network configuration corresponding to the target vector database is determined as the network configuration of the target question-answering service.

6. The question-answering service creation method according to claim 1, wherein: The determining the system configuration of the target question-and-answer service according to the question-and-answer service creation instruction includes: The server name, login method and login information of the target question-and-answer service are determined according to the question-and-answer service creation instruction, wherein the system configuration of the target question-and-answer service includes the server name, login method and login information.

7. The question-answering service creation method according to claim 1, wherein: Feeding back the access address of the target question-and-answer service to the target account includes: Upon receiving the service order information submitted by the target account for the target question-and-answer service, the administrator access address and the user access address of the target question-and-answer service are fed back to the target account, wherein the access address of the target question-and-answer service includes the administrator access address and the user access address, and the administrator access address and the user access address correspond to different access permissions of the target question-and-answer service.

8. A question-answering service creation device, characterized in that: The device comprises: An acquisition module, configured to acquire a question-and-answer service creation instruction initiated by a target account, wherein the question-and-answer service creation instruction is used to request creation of a target question-and-answer service; a configuration module, configured to determine the basic configuration, network configuration, and system configuration of the target question-answering service according to the question-answering service creation instruction, wherein the basic configuration of the target question-answering service includes a target macro model supporting implementation of the target question-answering service; a construction module, configured to obtain a mirror data packet corresponding to the target large model, and construct a target question-answering service based on the mirror data packet corresponding to the target large model, the basic configuration, network configuration, and system configuration of the target question-answering service, wherein the mirror data packet includes framework data required by the target large model; The feedback module is configured to feed back the access address of the target question-and-answer service to the target account, so that the target account can access the target question-and-answer service through the access address.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the question-and-answer service creation method according to any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the question-and-answer service creation method according to any one of claims 1 to 7 is implemented.

Images