Application assistant implementation method and device, equipment, storage medium and product

Abstract

The invention provides an application assistant implementation method and device, equipment, a storage medium and a product, and belongs to the technical field of neural networks. The method comprises the steps that in the process that a first virtual object and a second virtual object fight on the basis of a fight application, first voice data of a user is converted into text information, the user is a user operating the first virtual object, and the first voice data is used for using an application assistant of the fight application through voice; obtaining current battle information of the first virtual object and the second virtual object; based on the character information and the battle information, response information corresponding to the first voice data is determined through a large language model corresponding to the battle application; and realizing an application assistant of the battle application based on the response information. According to the application, the difficulty of using the application assistant by the user is reduced, and the performance effect of the fight of the first virtual object based on the fight application is further improved.

Description

Technical Field

[0001] The present application relates to the field of neural network technology, and in particular to an implementation method, device, equipment, storage medium and product of an application assistant. Background Art

[0002] Currently, the game assistant on the mobile phone has a setting panel, and the user can set the relevant functions of the game according to his or her preferences through the setting panel; for example, the game assistant has a network setting panel, and the user can switch the wireless network used by the game through the network setting panel; for another example, the game assistant has a performance mode setting panel, and the user can set the performance mode of the game through the performance mode setting panel.

[0003] In the related art, when a user sets game-related functions through a setting panel, the user needs to have a relatively in-depth understanding of the setting functions in order to correctly use the setting panel, which results in a high complexity in setting game-related functions. Summary of the invention

[0004] The embodiments of the present application provide a method, device, equipment, storage medium and product for implementing an application assistant, which can reduce the difficulty for users to use the application assistant, thereby improving the performance of the first virtual object in a battle based on a battle application. The technical solution is as follows:

[0005] In one aspect, a method for implementing an application assistant is provided, the method comprising:

[0006] In a process in which a first virtual object and a second virtual object battle against each other based on a battle application, converting first voice data of a user into text information, the user being a user operating the first virtual object, and the first voice data being used to use an application assistant of the battle application by voice;

[0007] Acquire current battle information between the first virtual object and the second virtual object;

[0008] Based on the text information and the battle information, determining response information corresponding to the first voice data by using a large language model corresponding to the battle application;

[0009] Based on the response information, an application assistant of the battle application is implemented.

[0010] In another aspect, a device for implementing an application assistant is provided, the device comprising:

[0011] a conversion module, configured to convert first voice data of a user into text information during a battle between a first virtual object and a second virtual object based on a battle application, wherein the user is a user operating the first virtual object, and the first voice data is used to use an application assistant of the battle application by voice;

[0012] An acquisition module, used to acquire current battle information between the first virtual object and the second virtual object;

[0013] a determination module, configured to determine, based on the text information and the battle information, response information corresponding to the first voice data by using a large language model corresponding to the battle application;

[0014] An implementation module is used to implement an application assistant of the battle application based on the response information.

[0015] On the other hand, a terminal is provided, which includes one or more processors and one or more memories, wherein at least one program code is stored in the one or more memories, and the at least one program code is loaded and executed by the one or more processors to implement the implementation method of the application assistant as described above.

[0016] On the other hand, a computer-readable storage medium is provided, wherein at least one program code is stored in the storage medium, and the at least one program code is loaded and executed by a processor to implement the above-mentioned implementation method of the application assistant.

[0017] On the other hand, a computer program product is provided, wherein the computer program product stores at least one program code, and the at least one program code is used to be executed by a processor to implement the above-mentioned implementation method of the application assistant.

[0018] In an embodiment of the present application, by utilizing the natural speech understanding capabilities of the large language model corresponding to the battle application, real-time voice interaction that is easier for users to operate is achieved, which facilitates users to use the game scenes of the battle application in real time, reduces the difficulty for users to use the application assistant, and thereby improves the performance effect of the first virtual object in battles based on the battle application. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 A schematic diagram of an implementation environment of an implementation method of an application assistant shown in an exemplary embodiment of the present application is shown;

[0020] Figure 2 A flowchart of a method for implementing an application assistant shown in an exemplary embodiment of the present application is shown;

[0021] Figure 3A flowchart of a method for implementing an application assistant shown in an exemplary embodiment of the present application is shown;

[0022] Figure 4 A schematic diagram of an implementation method of an application assistant shown in an exemplary embodiment of the present application is shown;

[0023] Figure 5 A flowchart of a method for implementing an application assistant shown in an exemplary embodiment of the present application is shown;

[0024] Figure 6 A flowchart of a method for implementing an application assistant shown in an exemplary embodiment of the present application is shown;

[0025] Figure 7 A flowchart of a method for implementing an application assistant shown in an exemplary embodiment of the present application is shown;

[0026] Figure 8 A schematic diagram showing a method for implementing an application assistant shown in an exemplary embodiment of the present application is shown;

[0027] Fig. 9 A flowchart of a method for implementing an application assistant shown in an exemplary embodiment of the present application is shown;

[0028] Fig.10 A schematic diagram showing a method for implementing an application assistant shown in an exemplary embodiment of the present application is shown;

[0029] Fig.11 A block diagram of an implementation device of an application assistant shown in an exemplary embodiment of the present application is shown;

[0030] Fig.12 A block diagram of a terminal according to an exemplary embodiment of the present application is shown. DETAILED DESCRIPTION

[0031] In order to make the objectives, technical solutions and advantages of the present application clearer, the implementation methods of the present application will be further described in detail below with reference to the accompanying drawings.

[0032] The term "multiple" as used herein refers to two or more than two. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone. The character " / " generally indicates that the related objects are in an "or" relationship.

[0033] It should be noted that the information (including but not limited to user device information, user personal information, etc.), data (including but not limited to data used for analysis, stored data, displayed data, etc.) and signals involved in this application are all authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with the relevant laws, regulations and standards of the relevant countries and regions. For example, the voice data (first voice data, second voice data and third voice data), battle information and sample data (first sample data and second sample data) involved in this application are all obtained with full authorization.

[0034] Please refer to Figure 1 , which shows a schematic diagram of an implementation environment of an implementation method of an application assistant shown in an exemplary embodiment of the present application. Figure 1 , the implementation environment includes a terminal 101 and a server 102. The terminal 101 is installed and runs a battle application that supports virtual object battles, and the battle application can be a shooting game or a multiplayer online tactical competitive game, etc. In addition, multiple virtual objects fight each other based on the battle application. For example, the first virtual object and the second virtual object fight each other based on the battle application. The terminal 101 can be the terminal 101 corresponding to the first virtual object, and the user controls the first virtual object through the terminal 101 to fight against the second virtual object controlled by other users in the battle application. In some embodiments, the battle application provides an application assistant, and the application assistant can provide help to the user; for example, the application assistant switches the network for the user or provides battle suggestion information, etc.

[0035] In some embodiments, the terminal 101 may be a smart phone, a tablet computer, a laptop computer, a desktop computer, or a game console, etc. The server 102 includes at least one of a single server, multiple servers, a cloud computing platform, and a virtualization center. The server 102 is used to provide background services for the battle application; for example, the server 102 provides data computing services for the battle process of the virtual object.

[0036] In some embodiments, the application assistant of the present application can be applied not only to battle applications, but also to other applications; for example, the application assistant of the present application can be extended to shopping applications, and users can use the application assistant to express their intentions by voice, so that the application assistant can help users select items that users like. For another example, the application assistant of the present application can be extended to other non-battle game applications, and users can use the application assistant to express their intentions by voice, so that the application assistant can help users play games in other game applications or set certain functions of other game applications.

[0037] Please refer to Figure 2, which shows a flowchart of an implementation method of an application assistant shown in an exemplary embodiment of the present application. The execution subject of the method is a terminal, and the method includes:

[0038] Step 201: During a battle between a first virtual object and a second virtual object based on a battle application, converting first voice data of a user into text information, the user is a user operating the first virtual object, and the first voice data is used to use an application assistant of the battle application through voice.

[0039] The terminal displays a virtual scene, which is used to simulate the real environment in reality; for example, the virtual scene may include the sky, the ocean, and the land; and the virtual scene may also include virtual items; for example, buildings for hiding, vehicles for escaping or chasing, and props for arming oneself or fighting with other virtual objects. In some embodiments, the virtual scene includes a first virtual object and a second virtual object, the first virtual object is a virtual image used to represent the user, and the second virtual object is a virtual image used to represent the user's hostility. The user controls the first virtual object through the terminal to fight against the second virtual object controlled by other users in the battle application. For example, the first virtual object can be a virtual character, a virtual animal, or a cartoon character. Similarly, the second virtual object can also be a virtual character, a virtual animal, or a cartoon character.

[0040] In some embodiments, the user can use the application assistant of the battle application through voice, that is, when the user wants to use the application assistant of the battle application, the user expresses his intention through voice, and the terminal obtains the user's first voice data and converts the first voice data into text information. In some embodiments, when the user wants to set certain functions of the battle application, the user expresses his application setting intention through voice, the terminal obtains the first voice data corresponding to the application setting intention, and converts the first voice data into text information; for example, when the user wants to switch the network currently used by the battle application, the user can say "switch to 5G network", and the terminal obtains the first voice data corresponding to "switch to 5G network".

[0041] In other embodiments, when a user wants to query some information through the network, the user expresses his or her network query intention through voice, the terminal obtains first voice data corresponding to the network query intention, and converts the first voice data into text information; for example, when the user wants to ask for battle suggestions, the user can say "How should I fight now", at this time the terminal obtains the first voice data corresponding to "How should I object now".

[0042] The application setting intent can be used to set any function of the battle application; for example, the wireless network used by the battle application can be set; for another example, the screen brightness can be set; for another example, the volume can be set, etc. The network query intent can be used to query any information of the battle application; for example, the team selection suggestion information, battle suggestion information, and equipment suggestion information of the battle application can be queried. In addition, the above application setting intent and network query intent are only examples and do not limit the application scenarios of this application; this application can also be applied in other scenarios; for example, it can be applied to call other applications; then the intent can be to call other applications; for example, call a social application or call a phone application, etc.

[0043] In some embodiments, the application assistant obtains the user's first voice data in real time, and converts the first voice data into text information when it determines that the first voice data is used to use the application assistant through voice. When it determines that the first voice data is not used to use the application assistant through voice, it continues to obtain the user's first voice data until the first voice data for using the application assistant through voice is obtained, and then converts the currently obtained first voice data into text information, thereby providing the user with a real-time online assistant for voice interaction and improving the efficiency of the user using the application assistant through voice. In other embodiments, the application assistant obtains the user's first voice data and converts the first voice data into text information after the user wakes up the application assistant with a wake-up word, so that the application assistant is in a dormant state before being awakened to save power consumption.

[0044] In an embodiment of the present application, the application assistant can be used by the user through voice, so the application assistant can also be used as a voice assistant. The acquisition of the first voice data is authorized by the user or fully authorized by all parties; for example, before the terminal acquires the voice data each time, it prompts the user whether to allow the acquisition of the voice data; before the terminal acquires the first voice data, the terminal displays a first prompt message, and the first prompt message is used to prompt the user whether to agree that the terminal acquires the first voice data; the terminal acquires the first voice data only when the confirmation message triggered by the user is received, and the terminal does not acquire the first voice data when the confirmation message triggered by the user is not received or the negative message triggered by the user is received. For another example, before the user uses the application assistant, it is often prompted once, and after the user allows the acquisition of the voice data, the terminal can acquire the user's voice data; for example, before the user uses the application assistant, the terminal displays a second prompt message, and the second prompt message is used to prompt the user whether to agree that the terminal acquires the user's voice data. When the confirmation message triggered by the user is received, the terminal can subsequently acquire the user's voice data, so there is no need to authorize each time before acquiring the voice data.

[0045] Step 202: Obtain current battle information between the first virtual object and the second virtual object.

[0046] The terminal obtains the real-time battle information of the first virtual object and the second virtual object during the battle from the server through the communication interface between the battle application and the server, that is, the current battle information. For example, if the battle application is a multiplayer online tactical competitive game, the battle information includes at least one of the lineup information of the first virtual object, the lineup information of the second virtual object, the equipment information of the first virtual object, the equipment information of the second virtual object, the battle record information of the first virtual object, and the battle record information of the second virtual object. The battle information may also include the skill information of the first virtual battle and the skill information of the second virtual object.

[0047] In some embodiments, the terminal sets battle information corresponding to multiple options in advance; for example, the multiple options include lineup information of both sides, equipment information of both sides and battle record information of both sides. Then, after the terminal converts the first voice data into text information, it obtains the battle information corresponding to the multiple options, that is, the lineup information of the first virtual object, the lineup information of the second virtual object, the equipment information of the first virtual object, the equipment information of the second virtual object, the battle record information of the first virtual object and the battle record information of the second virtual object, thereby improving the comprehensiveness of the battle information obtained, and then subsequently determining the accuracy of the response information corresponding to the first voice data based on the battle information.

[0048] In other embodiments, the terminal can determine at least one target option that matches the text information from multiple options based on the text information converted from the first voice data, and obtain the battle information corresponding to the at least one target option, so as to obtain the battle information that is more matched with the user's needs. This can improve the acquisition of less and more precise battle information, and while improving the efficiency of determining the response information determined based on the battle information, it can also ensure the accuracy of the determined response information.

[0049] In some embodiments, the step of the terminal determining at least one target option that matches the text information from multiple options may be: the terminal determines the intent information corresponding to the first voice data based on the text information through the large language model corresponding to the battle application, and based on the intent information, determines at least one target option that matches the intent information from multiple options. For example, multiple options include lineup information of both parties, equipment information of both parties, and battle record information of both parties. The terminal determines that the intent information corresponding to the first voice data is equipment recommendation information based on the text information through the large language model corresponding to the battle application, and the terminal obtains at least one target option that matches the equipment recommendation information as the equipment information of both parties. The large language model can be an artificial intelligence chat robot program (Chat Generative Pre-trained Transformer, ChatGPT).

[0050] In some embodiments, during the process of the first virtual object and the second virtual object fighting against each other based on the fighting application, the terminal obtains the current fighting information of the first virtual object and the second virtual object from the server in real time, and caches the obtained fighting information in the terminal; when the first voice data is converted into text information, the fighting information stored closest to the current time is directly obtained from the cache, and the fighting information is the current fighting information of the first virtual object and the second virtual object. Therefore, there is no need to wait for the response of the server at this time, and the fighting information can be directly obtained from the local cache, thereby improving the efficiency of obtaining the fighting information.

[0051] Step 203: Based on the text information and the battle information, the response information corresponding to the first voice data is determined by using a large language model corresponding to the battle application.

[0052] The large language model corresponding to the battle application is trained in advance so that the large language model corresponding to the battle application can be familiar with the application scenario and application method of the battle application, and when receiving text information and battle information, it can output response information corresponding to the first voice data based on the text information and battle information. The response information can be a network query result or an application setting command. The network query result includes at least one of team selection suggestion information, battle suggestion information, and equipment suggestion information. The application setting command includes any background setting of the application; for example, the application setting command includes at least one of a network setting command, a brightness setting command, and a volume setting command.

[0053] Step 204: Based on the response information, implement an application assistant for the battle application.

[0054] The terminal executes the operation corresponding to the response information to implement the application assistant of the battle application. For example, if the response information includes a network query result, the terminal outputs the network query result; for another example, if the response information includes an application setting command, the terminal executes the application setting command to set the battle application.

[0055] In an embodiment of the present application, by utilizing the natural speech understanding capabilities of the large language model corresponding to the battle application, real-time voice interaction that is easier for users to operate is achieved, which facilitates users to use the game scenes of the battle application in real time, reduces the difficulty for users to use the application assistant, and thereby improves the performance effect of the first virtual object in battles based on the battle application.

[0056] Please refer to Figure 3 , which shows a flow chart of a method for implementing an application assistant shown in an exemplary embodiment of the present application. The method includes:

[0057] Step 301: During a battle between a first virtual object and a second virtual object based on a battle application, the terminal converts first voice data of a user into text information, the user is a user operating the first virtual object, and the first voice data is used to use an application assistant of the battle application through voice.

[0058] In some embodiments, this step is the same as step 201 and will not be described in detail here. Figure 4 , the user inputs "I want to turn on 4D vibration" through voice, and the terminal obtains the first voice data of "I want to turn on 4D vibration" and converts the first voice data into text information. For another example, the user inputs "Help me switch to 5G network" through voice, and the terminal obtains the first voice data of "Help me switch to 5G network" and converts the first voice data into text information. For another example, the user inputs "What is the right choice for this lineup?" through voice, and the terminal obtains the first voice data of "What is the right choice for this lineup?" and converts the first voice data into text information. For another example, the user inputs "What equipment should I produce now?" through voice, and the terminal obtains the first voice data of "What equipment should I produce now?" and converts the first voice data into text information.

[0059] Step 302: The terminal obtains current battle information between the first virtual object and the second virtual object.

[0060] In some embodiments, this step is the same as step 202 and will not be repeated here.

[0061] Step 303: The terminal combines the text information and the battle information into request information, and inputs the request information into the large language model corresponding to the battle application.

[0062] In some embodiments, the large language model corresponding to the battle application is stored in the server; accordingly, the terminal combines the text information and the battle information into a request message and sends the request message to the server; the server receives the request message and inputs the request into the large language model corresponding to the battle application, thereby obtaining the response message with the help of the large language model in the server, so that the terminal does not need to store the large language model locally, saving the storage resources and computing resources of the terminal. For example, please continue to refer to Figure 4 The request information is the terminal side data, and the request information is input into the large language model corresponding to the battle application in the server through data transmission.

[0063] In other embodiments, the large language model corresponding to the battle application can be stored in the terminal, so that the terminal can use the large language model corresponding to the battle application without going through the network, thereby improving the efficiency of using the large language model corresponding to the battle application. The large language model corresponding to the battle application is a pre-trained large language model, and the large language model corresponding to the battle application can be trained by the server or by the terminal. In the embodiments of the present application, the subject of training the large language model corresponding to the battle application is not specifically limited; and the specific process of training the large language model corresponding to the battle application will be described in detail in the following embodiments.

[0064] In some embodiments, the terminal can also combine some other information to comprehensively determine the response information corresponding to the first voice data; for example, the terminal can also combine the user's preference information; then in this step, the terminal obtains the user's preference information, combines the text information, battle information and preference information into request information, and inputs the request information into the large language model corresponding to the battle application, so that the response information can be determined for the user according to the user's preference, thereby improving the accuracy of the determined response information. The terminal can also combine the response information corresponding to the fourth voice data of the last time to comprehensively determine the response information corresponding to the first voice data; then in this step, the terminal obtains the response information corresponding to the fourth voice data of the last time, combines the text information, battle information and the response information corresponding to the fourth voice data of the last time into request information, and inputs the request information into the large language model corresponding to the battle application, thereby determining the response information corresponding to the first voice data of this time for the user according to the response information corresponding to the fourth voice data of the last time, thereby improving the accuracy of the determined response information.

[0065] Step 304: The terminal calls the description file of the battle application based on the request information through the large language model corresponding to the battle application, and determines the response information corresponding to the first voice data.

[0066] The interface description file includes a description file of a network search interface and a description file of an application setting interface. This step can be implemented by the following steps (1) and (2), including:

[0067] (1) The terminal determines the intent information corresponding to the first voice data based on the request information by using the large language model corresponding to the battle application.

[0068] The intention information can be a data query based on the network or a corresponding application setting. Since the large language model corresponding to the battle application has a strong natural language understanding ability; therefore, the intention information corresponding to the first voice data determined by the large language model corresponding to the battle application can more accurately express the user's intention in the battle application, thereby improving the accuracy of the determined intention information.

[0069] (2) The terminal determines answer information corresponding to the intention information from the description file of the battle application, and obtains response information corresponding to the first voice data.

[0070] In the case where the intent information is used to perform data query based on the network, the terminal determines the network query result corresponding to the intent information from the description file of the network search interface, and obtains the response information corresponding to the first voice data; in the case where the intent information is used to set the battle application, the terminal determines the application setting command corresponding to the intent information from the description file of the application setting interface, and obtains the response information corresponding to the first voice data. This process can continue to refer to Figure 4 In some embodiments, the description file of the network search interface includes relevant information for performing data query based on the network, that is, storing network query results corresponding to multiple query intents; the description file of the application setting interface includes relevant information for application setting commands, that is, storing application setting commands corresponding to multiple setting intents.

[0071] In some embodiments, when the user wants to expand the network search function, the user can directly add the query intent and its corresponding network query results in the description file of the network search interface; when the user wants to expand the application setting function, the user can directly add the setting intent and its corresponding application setting command in the description file of the application setting interface, which is convenient for the addition and expansion of existing functions. In other embodiments, when the user wants to expand other functions in the application assistant, the user can set the interface corresponding to the other functions and the description file corresponding to the other functions, so that when the application assistant uses other functions, it can call the description file corresponding to other functions through the interface corresponding to other functions to implement other functions, which is convenient for the expansion of functions.

[0072] In an embodiment of the present application, the real-time assistant function of the battle application based on the large language model utilizes the natural language understanding ability of the large language model, and uses network search and application settings as underlying capabilities to improve the description file of the interface for the large language model to call. This not only facilitates users to use the game scene of the battle application in real time, reduces the difficulty of users using the application assistant, but also improves the accuracy of the determined response information. From the above two perspectives, the performance effect of the first virtual object in the battle based on the battle application can be improved.

[0073] In some embodiments, when the terminal sends the request information to the server in step 303, in this step, the server determines the response information corresponding to the first voice data by calling the description file of the battle application based on the request information through the large language model corresponding to the battle application, sends the response information to the terminal, and the terminal receives the response information. Among them, the process of the server determining the response information corresponding to the first voice data by calling the description file of the battle application based on the request information through the large language model corresponding to the battle application is similar to the process of the terminal determining the response information corresponding to the first voice data by calling the description file of the battle application based on the request information through the large language model corresponding to the battle application, and will not be repeated here.

[0074] Step 305: The terminal implements an application assistant for the battle application based on the response information.

[0075] In some embodiments, this step is the same as step 204 and will not be described in detail here. Figure 4 The large language model corresponding to the battle application determines the network query result or application setting command, and returns the network query result or application setting command to the terminal, and the terminal implements the application assistant of the battle application based on the network query result or application setting command.

[0076] In an embodiment of the present application, the real-time assistant function of the battle application based on the large language model utilizes the natural language understanding ability of the large language model, and uses network search and application settings as underlying capabilities to improve the description file of the interface for the large language model to call. This not only facilitates users to use the game scene of the battle application in real time, reduces the difficulty of users using the application assistant, but also improves the accuracy of the determined response information. From the above two perspectives, the performance effect of the first virtual object in the battle based on the battle application can be improved.

[0077] In addition, this solution makes up for the problems of the original application assistant solution, such as the numerous functions in the setting interface, high learning cost, complex settings, and single interaction. By utilizing the natural language understanding ability of the large language model, real-time voice interaction that is easier for users to operate is achieved. At the same time, network search and application function settings are used as underlying capabilities, which are uniformly encapsulated by the assistant and provided to users externally. This not only makes it easier for users to adjust settings, but also expands more suggestion capabilities, making it easier for users to ask the assistant for suggestions based on real-time game scenarios, assisting users in playing games, and improving the performance of games.

[0078] Please refer to Figure 5 , which shows a flow chart of a method for implementing an application assistant shown in an exemplary embodiment of the present application. The method includes:

[0079] Step 501: During a battle between a first virtual object and a second virtual object based on a battle application, the terminal converts first voice data of a user into text information, the user is a user operating the first virtual object, and the first voice data is used to perform a network query using an application assistant of the battle application through voice.

[0080] In some embodiments, this step is the same as step 201 and will not be repeated here.

[0081] Step 502: The terminal obtains current battle information between the first virtual object and the second virtual object.

[0082] In some embodiments, this step is the same as step 202 and will not be repeated here.

[0083] Step 503: The terminal combines the text information and the battle information into request information, and inputs the request information into the large language model corresponding to the battle application.

[0084] In some embodiments, this step is the same as step 303 and will not be repeated here.

[0085] Step 504: The terminal determines the intent information corresponding to the first voice data based on the request information by using the large language model corresponding to the battle application, and the intent information is used for data query based on the network.

[0086] In some embodiments, this step is the same as (1) in step 304 and will not be repeated here.

[0087] Step 505: The terminal determines the network query result corresponding to the first intent information from the description file of the network search interface.

[0088] Step 506: The terminal converts the network query result from text information into second voice data, and plays the second voice data.

[0089] The terminal converts the network query result from text information into the second voice data through the application assistant, plays the second voice data in real time, and promptly feeds back the query result to the user through voice.

[0090] In an embodiment of the present application, when a user wants to perform a data query based on the network, the user can use the application assistant of the battle application through voice. The voice assistant uses the large language model corresponding to the battle application and uses the network search as the underlying capability to provide an interface description file for the network search for the large language model to call. It can thereby determine accurate network query results based on the user's first voice data, and can also feedback the query results to the user in the form of voice, thereby achieving a more user-friendly interaction method and a more powerful assistant function.

[0091] Please refer to Figure 6, which shows a flow chart of a method for implementing an application assistant shown in an exemplary embodiment of the present application. The method includes:

[0092] Step 601: During a battle between a first virtual object and a second virtual object based on a battle application, the terminal converts first voice data of a user into text information, the user is a user operating the first virtual object, and the first voice data is used to use a command setting function of an application assistant of the battle application through voice.

[0093] In some embodiments, this step is the same as step 201 and will not be repeated here.

[0094] Step 602: The terminal obtains current battle information between the first virtual object and the second virtual object.

[0095] In some embodiments, this step is the same as step 202 and will not be repeated here.

[0096] Step 603: The terminal combines the text information and the battle information into request information, and inputs the request information into the large language model corresponding to the battle application.

[0097] In some embodiments, this step is the same as step 303 and will not be repeated here.

[0098] Step 604: The terminal determines the intent information corresponding to the first voice data based on the request information by using the large language model corresponding to the battle application, and the intent information is used to set the battle application.

[0099] In some embodiments, this step is the same as (2) in step 304 and will not be repeated here.

[0100] Step 605: The terminal determines the application setting command corresponding to the intent information from the description file of the application setting interface.

[0101] When the terminal determines that the intent information is used to set the battle application, the terminal calls the application setting interface, calls the corresponding description file based on the application setting interface, and determines the application setting command corresponding to the intent information from the description file.

[0102] Step 606: The terminal sets the battle application based on the application setting command.

[0103] For example, if the application setting command is to switch the wireless network from 4G network to 5G network, the terminal will switch the wireless network used by the battle application from 4G network to 5G network; for another example, if the application setting command is to increase the volume, the terminal will increase the volume.

[0104] Step 607: The terminal plays the third voice data, where the third voice data is used to prompt the user that the battle application has been set up.

[0105] In an embodiment of the present application, when a user wants to set up a battle application, the user can use the application assistant of the battle application through voice. The voice assistant uses the large language model corresponding to the battle application and uses the application function setting as the underlying capability to provide a description file of the application setting interface for the large language model to call. It can thereby determine the accurate application setting command based on the user's first voice data, and can also prompt the user by voice after the setting is completed that the modification is completed, thereby achieving a more user-friendly interaction method and a more powerful assistant function.

[0106] Please refer to Figure 7 , which shows a flow chart of a method for implementing an application assistant shown in an exemplary embodiment of the present application. The method includes:

[0107] Step 701: The terminal obtains first sample data corresponding to the battle application, where the first sample data includes sample request information and sample response information, and the sample request information includes sample battle information and sample text information.

[0108] Before training the large language model corresponding to the battle application, the terminal needs to construct a training data set, which includes multiple first sample data, that is, the number of first sample data in this step is multiple. The first sample data can be manually annotated by the developer; accordingly, the step of the terminal obtaining the first sample data corresponding to the battle application can be: the terminal receives the first sample data corresponding to the battle application that is input. For example, please refer to Figure 8 The sample request information included in the first sample data is input information, and the sample response information included in the first sample data is output information, and the sample battle information included in the sample request information is a real-time status, and the sample text information included in the sample request information is a user request; the sample response information includes network search results or application setting commands.

[0109] In some embodiments, multiple first sample data are all annotated by developers; since developers manually annotate the first sample data, the quality of the first sample data can be guaranteed. In other embodiments, some of the multiple first sample data are annotated by developers, while another part of the sample data can be generated by a data generation network; accordingly, the step for the terminal to obtain the first sample data corresponding to the battle application can be: the terminal obtains part of the sample data corresponding to the input battle application, and based on the part of the sample data, generates another part of the sample data through the data generation network, and the input part of the sample data and the generated another part of the sample data form the first sample data. In an embodiment of the present application, the terminal can generate sample data with the help of a data generation network, which can reduce the cost of obtaining sample data and improve the efficiency of obtaining sample data.

[0110] Step 702: The terminal adjusts the trained general large language model based on the first sample data to obtain a large language model corresponding to the battle application.

[0111] In some embodiments, the terminal obtains any trained general large language model, thereby reducing the difficulty of obtaining the general large language model. In other embodiments, the terminal obtains a general large language model that matches the battle application based on the battle application; for example, the terminal determines the field to which the battle application belongs, and based on the field to which the battle application belongs, obtains the general large language model corresponding to the field, thereby being able to obtain a general large language model that is adapted to the battle application, reducing the time used to adjust the general large language model, and improving the efficiency of training to obtain the large language model corresponding to the battle application.

[0112] In some embodiments, the server may also train the large language model of the battle application, and store the trained large language model corresponding to the battle application in the terminal, so that the terminal can determine the response information offline based on the large language model corresponding to the battle application. In addition, the process of the server training the large language model corresponding to the battle application is the same as the process of the terminal training the large language model corresponding to the battle application, which will not be repeated here.

[0113] In an embodiment of the present application, the large language model corresponding to the battle application is trained by fine-tuning the trained general large language model using the first sample data corresponding to the battle application. This eliminates the need to train a completely new model, thereby reducing the time required to train the large language model corresponding to the battle application and improving the efficiency of training the large language model corresponding to the battle application.

[0114] Please refer to Fig. 9 , which shows a flow chart of a method for implementing an application assistant shown in an exemplary embodiment of the present application. The method includes:

[0115] Step 901: The terminal obtains first sample data corresponding to a battle application, where the first sample data includes sample request information and sample response information, and the sample request information includes sample battle information and sample text information.

[0116] In some embodiments, this step is the same as step 701 and will not be repeated here.

[0117] Step 902: Based on the first sample data, the terminal determines from the description file of the battle application an interface description file that matches the data type of the first sample data.

[0118] If the first sample data is used for data query based on the network, the terminal calls the description file of the network query interface; if the first sample data is used for setting the battle application, the terminal calls the description file of the application setting interface. Before this step, the developer designs a suitable calling method according to common user requests; for example, if the user asks for equipment recommendation information or team selection recommendation information, the terminal calls the network query interface and determines the network query interface description file based on the network query interface; for another example, if the user wants to make application settings, the terminal calls the application setting interface and determines the description file of the application setting interface based on the application setting interface.

[0119] Step 903: The terminal combines the first sample data and the interface description file into second sample data.

[0120] There are multiple first sample data, one first sample data and the interface description file corresponding to the first sample data form one second sample data, thereby obtaining multiple second sample data, and the multiple second sample data form a fine-tuning data set.

[0121] Step 904: The terminal adjusts the trained general large language model based on the second sample data to obtain a large language model corresponding to the battle application.

[0122] In some embodiments, the terminal inputs the sample battle information and sample text information in the second sample data into a universal large language model. The universal large language model calls the description file of the application setting interface and the description file of the network search interface based on the sample battle information and the sample text information, determines the output result, that is, outputs the response information, determines the difference between the output response information and the sample response information, and adjusts the universal large language model based on the difference until the difference between the output response information and the sample response information is less than the preset difference or the number of iterations reaches the preset number. At this time, the large language model corresponding to the battle application is trained.

[0123] In the embodiment of the present application, the fine-tuning data set constructed in the previous step is used in conjunction with the description files of the network query interface and the application setting interface to retrain the large language model, that is, to obtain the large language model corresponding to the battle application. The purpose of this step is to make the large language model familiar with the specific application scenarios and application methods that the present invention focuses on, and to form the ability to output network query results or setting commands when receiving real-time status and user requests.

[0124] For example, see Fig.10 The terminal obtains a fine-tuning data set, which includes multiple first sample data. Based on the fine-tuning data set, the terminal calls the description file of the application setting interface and the description file of the network search interface to adjust the general large language model to obtain the large language model corresponding to the battle application.

[0125] In the embodiment of the present application, the application-based real-time assistant function of the large language model utilizes the natural language understanding ability of the large language model, and sets the network search and application functions as the underlying capabilities, providing an interface description file for the large language model to call. Under reasonable functional connection, the large language model of specific functions is trained to expand and integrate the application assistant capabilities, realize a more humane interaction method, and a more powerful assistant function.

[0126] Please refer to Fig.11 , which shows a block diagram of an implementation device of an application assistant shown in an exemplary embodiment of the present application, the device includes:

[0127] The conversion module 1101 is used to convert the first voice data of the user into text information during the process of the first virtual object and the second virtual object fighting each other based on the fighting application, the user is a user operating the first virtual object, and the first voice data is used to use the application assistant of the fighting application through voice;

[0128] An acquisition module 1102 is used to acquire current battle information between the first virtual object and the second virtual object;

[0129] A determination module 1103, configured to determine response information corresponding to the first voice data based on the text information and the battle information by using a large language model corresponding to the battle application;

[0130] The implementation module 1104 is used to implement an application assistant of the battle application based on the response information.

[0131] In some embodiments, the determination module 1103 is used to combine text information and battle information into request information, and input the request information into a large language model corresponding to the battle application; through the large language model corresponding to the battle application, the description file of the battle application is called based on the request information to determine the response information corresponding to the first voice data.

[0132] In some embodiments, the determination module 1103 is used to determine the intent information corresponding to the first voice data based on the request information through the large language model corresponding to the battle application; determine the answer information corresponding to the intent information from the description file of the battle application, and obtain the response information corresponding to the first voice data.

[0133] In some embodiments, determination module 1103 is used to determine the network query result corresponding to the intent information from the description file of the network search interface when the intent information is used for network-based data query, and obtain response information corresponding to the first voice data; when the intent information is used to set up a battle application, determine the application setting command corresponding to the intent information from the description file of the application setting interface, and obtain response information corresponding to the first voice data.

[0134] In some embodiments, the implementation module 1104 is used to output the network query result when the response information is the network query result; and to set the battle application based on the application setting command when the response information is the application setting command.

[0135] In some embodiments, the implementation module 1104 is used to convert the network query result from text information into second voice data and play the second voice data.

[0136] In some embodiments, the apparatus further comprises:

[0137] The playing module is used to play the third voice data, and the third voice data is used to prompt the user that the setting of the battle application has been completed.

[0138] In some embodiments, the apparatus further comprises:

[0139] The training module is used to obtain the first sample data corresponding to the battle application, the first sample data includes sample request information and sample response information, the sample request information includes sample battle information and sample text information; based on the first sample data, the trained general large language model is adjusted to obtain the large language model corresponding to the battle application.

[0140] In some embodiments, a training module is used to determine, based on the first sample data, an interface description file that matches the data type of the first sample data from the description file of the battle application; combine the first sample data and the interface description file into second sample data; and based on the second sample data, adjust the trained general large language model to obtain a large language model corresponding to the battle application.

[0141] In an embodiment of the present application, by utilizing the natural speech understanding capabilities of the large language model corresponding to the battle application, real-time voice interaction that is easier for users to operate is achieved, which facilitates users to use the game scenes of the battle application in real time, reduces the difficulty for users to use the application assistant, and thereby improves the performance effect of the first virtual object in battles based on the battle application.

[0142] It should be noted that the implementation device of the application assistant provided in the above embodiment only uses the division of the above functional modules as an example when implementing the application assistant. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the terminal is divided into different functional modules to complete all or part of the functions described above. In addition, the implementation device of the application assistant provided in the above embodiment and the implementation method embodiment of the application assistant belong to the same concept. The specific implementation process is detailed in the method embodiment and will not be repeated here.

[0143] Please refer to Fig.12, which shows a block diagram of a terminal 1200 shown in an exemplary embodiment of the present application. The terminal 1200 in the present application includes one or more of the following components: a processor 1210 , a memory 1220 , and a display screen 1230 .

[0144] The processor 1210 may include one or more processing cores. The processor 1210 uses various interfaces and lines to connect the various parts of the entire terminal 1200, and executes various functions and processes data of the terminal 1200 by running or executing instructions, programs, code sets or instruction sets stored in the memory 1220, and calling data stored in the memory 1220. Optionally, the processor 1210 can be implemented in at least one hardware form of digital signal processing (DSP), field-programmable gate array (FPGA), and programmable logic array (PLA). The processor 1210 can integrate one or more combinations of a central processing unit (CPU), a graphics processing unit (GPU), a neural network processor (NPU), and a modem. Among them, the CPU mainly processes the operating system, user interface, and application programs; the GPU is responsible for rendering and drawing the content that needs to be displayed on the display screen 1230; the NPU is used to implement artificial intelligence (AI) functions; and the modem is used to process wireless communications. It is understandable that the above-mentioned modem may not be integrated into the processor 1210, but may be implemented by a separate chip.

[0145] The memory 1220 may include a random access memory (RAM) or a read-only memory (ROM). Optionally, the memory 1220 includes a non-transitory computer-readable storage medium. The memory 1220 may be used to store instructions, programs, codes, code sets, or instruction sets. The memory 1220 may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for at least one function (such as a touch function, a sound playback function, an image playback function, etc.), instructions for implementing the following various method embodiments, etc.; the data storage area may store data (such as audio data, a phone book), etc., created according to the use of the terminal 1200.

[0146] The display screen 1230 is a display component for displaying a user interface. Optionally, the display screen 1230 is a display screen with a touch function, through which the user can use any suitable object such as a finger, a touch pen, etc. to perform touch operations on the display screen 1230.

[0147] The display screen 1230 is usually arranged on the front panel of the terminal 1200. The display screen 1230 can be designed as a full screen, a curved screen, a special-shaped screen, a double-sided screen or a folding screen. The display screen 1230 can also be designed as a combination of a full screen and a curved screen, a combination of a special-shaped screen and a curved screen, etc., which is not limited in this embodiment.

[0148] In addition, those skilled in the art can understand that the structure of the terminal 1200 shown in the above drawings does not constitute a limitation on the terminal 1200, and the terminal 1200 may include more or fewer components than shown, or combine certain components, or arrange the components differently. For example, the terminal 1200 also includes an audio acquisition device, a speaker, a radio frequency circuit, an input unit, a sensor, an audio circuit, a wireless fidelity (Wireless Fidelity, Wi-Fi) module, a power supply, a Bluetooth module and other components, which will not be repeated here.

[0149] An embodiment of the present application also provides a computer-readable medium, which stores at least one program code, and the at least one program code is loaded and executed by the processor to implement the implementation method of the application assistant shown in the above embodiments.

[0150] An embodiment of the present application also provides a computer program product, which stores at least one program code, and the at least one program code is loaded and executed by the processor to implement the implementation method of the application assistant shown in the above embodiments.

[0151] In some embodiments, the computer program product involved in the embodiments of the present application may be deployed and executed on a user terminal, or on multiple user terminals located at one location, or on multiple user terminals distributed at multiple locations and interconnected by a communication network. Multiple user terminals distributed at multiple locations and interconnected by a communication network may constitute a blockchain system.

[0152] Those skilled in the art should be aware that in one or more of the above examples, the functions described in the embodiments of the present application can be implemented with hardware, software, firmware, or any combination thereof. When implemented using software, these functions can be stored in a computer-readable medium or transmitted as one or more instructions or codes on a computer-readable medium. Computer-readable media include computer storage media and communication media, wherein the communication media include any media that facilitates the transmission of a computer program from one place to another. The storage medium can be any available medium that a general or special-purpose computer can access.

[0153] The above description is only an optional embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A method for implementing an application assistant, characterized in that: The method comprises: In a process in which a first virtual object and a second virtual object battle against each other based on a battle application, converting first voice data of a user into text information, the user being a user operating the first virtual object, and the first voice data being used to use an application assistant of the battle application by voice; Acquire current battle information between the first virtual object and the second virtual object; Based on the text information and the battle information, determining response information corresponding to the first voice data by using a large language model corresponding to the battle application; Based on the response information, an application assistant of the battle application is implemented.

2. The method according to claim 1, characterized in that The determining, based on the text information and the battle information, response information corresponding to the first voice data by using a large language model corresponding to the battle application includes: Combining the text information and the battle information into request information, and inputting the request information into a large language model corresponding to the battle application; The large language model corresponding to the battle application is used to call the description file of the battle application based on the request information to determine the response information corresponding to the first voice data.

3. The method according to claim 2, characterized in that The step of calling the description file of the battle application based on the request information through the large language model corresponding to the battle application to determine the response information corresponding to the first voice data includes: Determining, by means of a large language model corresponding to the battle application, intent information corresponding to the first voice data based on the request information; Determine answer information corresponding to the intention information from the description file of the battle application, and obtain response information corresponding to the first voice data.

4. The method according to claim 3, characterized in that The determining answer information corresponding to the intention information from the description file of the battle application, and obtaining response information corresponding to the first voice data, includes: In the case where the intent information is used for data query based on a network, determining a network query result corresponding to the intent information from a description file of a network search interface, and obtaining response information corresponding to the first voice data; In the case where the intent information is used to set a battle application, an application setting command corresponding to the intent information is determined from a description file of an application setting interface, and response information corresponding to the first voice data is obtained.

5. The method according to claim 1, characterized in that The application assistant for implementing the battle application based on the response information includes: When the response information is a network query result, outputting the network query result; When the response information is an application setting command, the battle application is set based on the application setting command.

6. The method according to claim 5, characterized in that The outputting the network query result includes: The network query result is converted from text information into second voice data, and the second voice data is played.

7. The method according to claim 5, characterized in that After the battle application is set based on the application setting command, the method further includes: Play the third voice data, where the third voice data is used to prompt the user that the battle application has been set up.

8. The method according to claim 1, characterized in that The method further comprises: Acquire first sample data corresponding to the battle application, the first sample data including sample request information and sample response information, the sample request information including sample battle information and sample text information; Based on the first sample data, the trained general large language model is adjusted to obtain a large language model corresponding to the battle application.

9. The method according to claim 8, characterized in that The adjusting the trained general large language model based on the first sample data to obtain the large language model corresponding to the battle application includes: Based on the first sample data, determining an interface description file matching the data type of the first sample data from the description file of the battle application; Combining the first sample data and the interface description file into second sample data; Based on the second sample data, the trained general large language model is adjusted to obtain a large language model corresponding to the battle application.

10. An implementation device of an application assistant, characterized in that: The device comprises: a conversion module, configured to convert first voice data of a user into text information during a battle between a first virtual object and a second virtual object based on a battle application, wherein the user is a user operating the first virtual object, and the first voice data is used to use an application assistant of the battle application by voice; An acquisition module, used to acquire current battle information between the first virtual object and the second virtual object; a determination module, configured to determine, based on the text information and the battle information, response information corresponding to the first voice data by using a large language model corresponding to the battle application; An implementation module is used to implement an application assistant of the battle application based on the response information.

11. A terminal, characterized in that: The terminal includes one or more processors and one or more memories, and at least one program code is stored in the one or more memories. The at least one program code is loaded and executed by the one or more processors to implement the implementation method of the application assistant as described in any one of claims 1 to 9.

12. A computer-readable storage medium, characterized in that: The storage medium stores at least one program code, and the at least one program code is loaded and executed by the processor to implement the implementation method of the application assistant described in any one of claims 1 to 9.

13. A computer program product, characterized in that The computer program product stores at least one program code, and the at least one program code is used to be executed by a processor to implement the implementation method of the application assistant according to any one of claims 1 to 9.

Images