An information interaction method, system, device, and game robot

Abstract

The application discloses an information interaction method, system and device, electronic equipment and a computer readable storage medium. The method comprises: receiving first game information sent by a first server and updating a game information set by using the first game information; in response to satisfying a first information interaction condition, inputting the updated game information set into a first model to obtain first feedback information output by the first model, the first model being a language processing model pre-trained by using a preset interaction instruction set, and the interaction instruction set at least comprising a first instruction for indicating game logic of a game session; and sending the first feedback information to the first server to enable the first server to broadcast the first feedback information at a next game time of a current game time. The game robot based on the information interaction method can perform reasoning conforming to game logic in a logical reasoning game like a human being, thereby solving the technical problem that existing game robots cannot participate in a logical reasoning game.

Description

Technical Field

[0001] This application relates to the field of computer technology, and in particular to an information interaction method, system, device, game robot, electronic device, and computer-readable storage medium. Background Technology

[0002] With the rapid development of computer technology, intelligent robots have gradually appeared in all walks of life. As an intelligent agent for the gaming field, game robots usually play an auxiliary role in novice teaching and game matches.

[0003] Currently, game robots typically rely on pre-set fixed logic and simple semantic analysis capabilities to perform mechanical output in a "question and answer" format based on player input during gameplay. They lack the ability to perform logical reasoning based on the context of the game and cannot participate in the game reasoning process like humans.

[0004] Therefore, existing game robots suffer from a technical problem: they lack the ability to perform logical reasoning based on contextual content during gameplay, which prevents them from participating in logic reasoning games. Summary of the Invention

[0005] This application provides an information interaction method, system, device, game robot, electronic device, and computer-readable storage medium to solve the technical problem that existing game robots lack the ability to perform game logic reasoning based on contextual content during the game process, thus preventing them from participating in logic reasoning games.

[0006] In a first aspect, embodiments of this application provide an information interaction method, the method comprising: receiving first game information sent by a first server, and updating a game information set using the first game information, wherein the first game information is game information generated at the current game moment of a game match, and the game information set is a collection of game information generated at historical game moments of the game match; responding to satisfying a first information interaction condition, inputting the updated game information set into a first model to obtain first feedback information output by the first model, wherein the first model is a language processing model pre-trained using a preset interaction instruction set, and the interaction instruction set includes at least a first instruction for instructing the game logic of the game match; and sending the first feedback information to the first server, so that the first server broadcasts the first feedback information at the next game moment of the current game moment.

[0007] Secondly, embodiments of this application provide an information interaction method, the method comprising: responding to receiving first game information sent by a first client, sending the first game information to a second server, so that the second server updates a game information set using the first game information, wherein the first client is any one of at least one client matched in a game match, the first game information is generated by the first client detecting an input event for the first game information, and is game information generated at the current game moment of the game match, and the game information set is a collection of game information generated at historical game moments of the game match; receiving first feedback information sent by the second server, the first feedback information being obtained by the second server inputting the updated game information set into a first model, the first model being a language processing model pre-trained through a preset interaction instruction set, the interaction instruction set including at least a first instruction for instructing the game logic of the game match; sending the first feedback information to the second client at the next game moment of the current game moment, so that the second client displays the first feedback information on the screen, wherein the second client includes each of the at least one client matched in the game match.

[0008] Thirdly, embodiments of this application provide an information interaction system, which includes at least one client, a first server, a second server, and a first model; the at least one client detects an input event for first game information and sends the first game information to the first server, wherein the first game information is game information generated at the current game moment of the game match; the first server receives the first game information sent by the at least one client and sends the first game information to the second server; the second server receives the first game information sent by the first server and updates a game information set using the first game information, wherein the game information set is a collection of game information generated at historical game moments of the game match; the second server responds to... If the first information interaction condition is met, the updated game information set is input into the first model, wherein the first model is a language processing model pre-trained with a preset interaction instruction set, the interaction instruction set including at least a first instruction for instructing the game logic of the game match; the first model calculates and obtains first feedback information based on the updated game information set, and sends the first feedback information to the second server; the second server receives the first feedback information and sends the first feedback information to the first server; the first server receives the first feedback information sent by the second server and sends the first feedback information to the at least one client in the next game moment of the current game moment; the at least one client displays the first feedback information on the screen.

[0009] Fourthly, this application provides a game service system, which includes a first server and a second server. The first server is connected to at least one client, and the second server is connected to a first model. The first model is a language processing model pre-trained using a preset set of interaction instructions. The set of interaction instructions includes at least a first instruction for instructing the game logic of the game match. The first server is configured to send the first game information to the second server in response to receiving first game information from a first client. The first client is any one of the at least one client matched in the game match, and the first game information is generated by the first client detecting an input event related to the first game information, representing the current game moment of the game match. The system generates game information; receives first feedback information sent by the second server; sends the first feedback information to the second client in the next game moment of the current game moment, so that the second client displays the first feedback information on the screen, wherein the second client includes each of the at least one client matched in the game match; the second server is configured to receive the first game information sent by the first server and update the game information set using the first game information, wherein the game information set is a collection of game information generated in the historical game moments of the game match; in response to satisfying the first information interaction condition, inputs the updated game information set into the first model to obtain the first feedback information output by the first model; and sends the first feedback information to the first server.

[0010] Fifthly, embodiments of this application provide an information interaction device, comprising: a game information receiving unit, a feedback information acquisition unit, and a feedback information sending unit; the game information receiving unit is configured to receive first game information sent by a first server and update a game information set using the first game information, wherein the first game information is game information generated at the current game moment of the game, and the game information set is a collection of game information generated at historical game moments of the game; the feedback information acquisition unit is configured to, in response to satisfying a first information interaction condition, input the updated game information set into a first model to obtain first feedback information output by the first model, wherein the first model is a language processing model pre-trained using a preset interaction instruction set, and the interaction instruction set includes at least a first instruction for instructing the game logic of the game; the feedback information sending unit is configured to send the first feedback information to the first server, so that the first server broadcasts the first feedback information at the next game moment of the current game moment.

[0011] Sixthly, embodiments of this application provide an information interaction device, the device comprising: a game information sending unit, a feedback information receiving unit, and a feedback information sending unit; the game information sending unit is configured to, in response to receiving first game information sent by a first client, send the first game information to a second server, so that the second server updates the game information set using the first game information, wherein the first client is any one of at least one client matched in a game match, the first game information is generated by the first client detecting an input event related to the first game information, and is game information generated at the current game moment of the game match, and the game information set is the historical game moments of the game match. The generated set of game information; the feedback information receiving unit is used to receive first feedback information sent by the second server, the first feedback information being obtained by the second server inputting the updated set of game information into a first model, the first model being a language processing model pre-trained through a preset set of interaction instructions, the set of interaction instructions including at least a first instruction for instructing the game logic of the game match; the feedback information sending unit is used to send the first feedback information to the second client in the next game moment of the current game moment, so that the second client displays the first feedback information on the screen, wherein the second client includes each of the at least one client matched in the game match.

[0012] In a seventh aspect, embodiments of this application provide a game robot, including: an information storage module and an information processing module; the information storage module is used to store game information generated during game matches; the information processing module is used to execute the above-described methods.

[0013] Eighthly, embodiments of this application provide an electronic device, including: a memory and a processor; the memory is used to store one or more computer instructions; the processor is used to execute the one or more computer instructions to implement the above-described method.

[0014] Ninthly, embodiments of this application provide a computer-readable storage medium storing one or more computer instructions that, when executed by a processor, perform the above-described method.

[0015] Compared with existing technologies, the information interaction method provided in this application updates the game information set by sending the first game information from the first server, and inputs the game information set into the first model. This allows the first model to output first feedback information based on the game information generated at the current game moment and the game information generated at previous game moments (i.e., the context content during the game). Since the first model is a language processing model pre-trained using a preset set of interaction instructions, and this set includes first instructions for instructing the game logic of the game match, the game information set, including the game information generated at the current game moment and the game information generated at previous game moments, is used as input to the first model. The output first feedback information is the game logic obtained through game logic reasoning that conforms to the game match logic. This first feedback information is then sent to the first server, which can then broadcast the first feedback information in the next game moment. This method mainly involves information interaction between the first server, the first model, and the execution entity of the method (the second server). Through information interaction, it achieves game logic reasoning based on the context content during the game and outputs feedback information that conforms to the game logic. Therefore, game robots based on this information interaction method can perform reasoning in logic reasoning games just like humans. The information interaction method provided in this application solves the technical problem that existing game robots cannot participate in logic reasoning games because they lack the ability to perform game logic reasoning based on the context content during the game process. Attached Figure Description

[0016] Figure 1 This is an application system diagram of an information interaction method provided in an embodiment of this application;

[0017] Figure 2 This is a flowchart of the information interaction method provided in the first embodiment of this application;

[0018] Figure 3 This is a flowchart of the information interaction method provided in the second embodiment of this application;

[0019] Figure 4 This is a signaling flowchart of an information interaction method provided in the third embodiment of this application;

[0020] Figure 5 This is a flowchart illustrating another information interaction method provided in the third embodiment of this application;

[0021] Figure 6 This is a schematic diagram of the structure of the information interaction system provided in the fourth embodiment of this application;

[0022] Figure 7 This is a schematic diagram of the structure of the game service system provided in the fifth embodiment of this application;

[0023] Figure 8 This is a schematic diagram of the structure of the information interaction device provided in the sixth embodiment of this application;

[0024] Figure 9 This is a schematic diagram of the structure of the information interaction device provided in the seventh embodiment of this application;

[0025] Figure 10 This is a schematic diagram of the structure of the game robot provided in the eighth embodiment of this application;

[0026] Figure 11 This is a schematic diagram of the structure of the electronic device provided in the ninth embodiment of this application. Detailed Implementation

[0027] Many specific details are set forth in the following description to provide a full understanding of this application. However, this application can be implemented in many other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of this application; therefore, this application is not limited to the specific embodiments disclosed below.

[0028] With the development of computer technology, intelligent robots have gradually appeared in various industries, such as delivery robots, cleaning robots, and chatbots. Gaming robots, as intelligent agents specifically designed for the gaming field, typically play a supporting role in tutorials and gameplay.

[0029] Logic puzzle games (such as Werewolf and Murder Mystery) are a popular genre, typically requiring multiple players to advance the storyline through conversation, deduction, and decision-making. These games often have complex rules, requiring new players to familiarize themselves with the rules and terminology before joining. Current tutorials, such as watching gameplay videos and answering questions, only familiarize players with the rules and don't allow for rapid skill development through real-world gameplay. However, directly introducing new players to games can negatively impact the experience of other players. Therefore, in addition to current tutorials, providing game bots that allow players to participate in games like humans, performing logical reasoning based on the rules, and offering dedicated beginner gaming environments would significantly enhance new players' understanding and experience, enabling them to quickly develop into more experienced players. In addition, language logic reasoning games are usually multiplayer games with a limit on the number of players. If the required number is not reached, the game cannot start. When players wait too long before a game can start, it will seriously affect their gaming experience. If a game robot that can perform logical reasoning according to the game rules like a human could be provided to replace the missing players, it would reduce the waiting time and improve the gaming experience.

[0030] Currently, game robots typically rely on pre-set fixed logic and simple semantic analysis capabilities to mechanically output in a "question and answer" format based on player input during gameplay. They lack the ability to perform logical reasoning based on the context of the game and cannot participate in the game reasoning process like humans. Therefore, existing game robots cannot replace humans in providing beginner matches for novice players or replace missing players in game matches in language logic reasoning games.

[0031] In view of this, this application provides an information interaction method. This method updates the game information set by sending first game information from a first server, and inputs the game information set into a first model. The first model then outputs first feedback information based on the game information generated at the current game moment and the game information generated at previous game moments (i.e., the context content during the game). Since the first model is a language processing model pre-trained using a preset set of interaction instructions, and this set includes first instructions for instructing the game logic of the game match, the game information set, including the game information generated at the current game moment and the game information generated at previous game moments, is used as input to the first model. The output first feedback information is the game logic obtained through game logic reasoning that conforms to the game match logic. This first feedback information is then sent to the first server, which can then broadcast the first feedback information in the next game moment. This method, through information interaction between the first server, the first model, and the execution entity of the method (the second server), achieves game logic reasoning based on the context content during the game and outputs feedback information that conforms to the game logic. Therefore, game robots based on this information interaction method can perform reasoning in logic games just like humans, providing beginner matches for novice players or participating in game matches in place of missing players.

[0032] The information interaction method, system, device, game robot, electronic device, and computer-readable storage medium described in this application will be further described in detail below with reference to specific embodiments and accompanying drawings.

[0033] Figure 1 This is an application system diagram of an information interaction method provided in an embodiment of this application. For example... Figure 1As shown, the information interaction method provided in this application embodiment is applied to information interaction between the second server 102 and the first server 101, and between the second server 102 and the language processing model 103. In addition, the information interaction method provided in this application also involves a client 104, which is connected to the first server 101. The client 104 can be any device such as a smartphone, tablet, laptop, desktop computer, or personal digital assistant (PDA), serving as the terminal for the game player. The first server 101 can be a service module located within the client 104, a service device connected to the client 104, or a game server connected to multiple clients 104, deploying game tasks for the clients 104. The second server 102 can be a server controlling game robots, or a server cluster including multiple servers, with each server controlling multiple game robots to participate in the game. The language processing model 103 can be a processing module located inside the second server 102, or it can be an independent model provided by a third-party service. The second server 102 establishes a communication connection with the language processing model 103 through the model interface provided by the third-party service.

[0034] The first embodiment of this application provides an information interaction method, which is applied to... Figure 1 The second server 102 in the application system shown realizes information interaction between the second server 102 and the first server 101, and between the second server 102 and the language processing model 103.

[0035] In one optional implementation of this embodiment, the first server 101 can be understood as a game server, which obtains game information generated by the client 104 and sends it to the second server 102. The second server 102 can be understood as a game robot docking server, which establishes a communication connection between the language processing model 103 and the first server 101 to enable the robot in the game to perform logical reasoning based on game information, just like a human, and output feedback information that conforms to the game logic. The language processing model 103 can be understood as a Natural Language Processing (NLP) model, whose main function is to enable computers to understand, analyze, and manipulate human language, thereby achieving more intelligent natural language interaction. With the current development of artificial intelligence, many natural language processing models are being applied in intelligent agents. Among them, ChatGPT (Chat Generative Pre-trained Transformer) is a popular pre-trained language model that uses a Transformer-based neural network architecture to generate coherent text with logical and grammatical correctness. In the method provided in this embodiment, the language processing model 103 is any natural language processing model including but not limited to ChatGPT.

[0036] In one specific implementation, the aforementioned application system uses a Node.js runtime environment and is developed based on the Pinus game server framework, a TypeScript version of Pomelo, connecting to ChatGPT via the ChatGPT interface. Node.js is an open-source and cross-platform JavaScript (JS) runtime environment. Pomelo is a high-performance, distributed game server framework, including a basic development framework and related extension components (libraries and toolkits), allowing developers to easily develop applications. TypeScript is a free and open-source programming language, a superset of JavaScript, adding optional static typing and class-based object-oriented programming. Pinus is the TypeScript version of Pomelo and is also a high-performance, distributed game server framework.

[0037] Figure 2 This is a flowchart of the information interaction method provided in this embodiment. The following is in conjunction with... Figure 2 The method provided in this embodiment will be described in detail. The embodiments described below are used to explain the technical solution of this application and are not intended to limit actual use.

[0038] like Figure 2As shown, the information interaction method provided in this embodiment includes the following steps S210 to S230.

[0039] Step S210: Receive first game information sent by the first server, and update the game information set using the first game information, wherein the first game information is the game information generated at the current game moment of the game match, and the game information set is the collection of game information generated at the historical game moments of the game match.

[0040] Language logic reasoning games typically involve multiple players. During a game, each player takes turns speaking or making decisions, and each player's speech (including decisions) is displayed on the client screens of all players. In the method provided in this embodiment, the first server sends the game information generated by each player's speech and / or decisions to the second server. In this embodiment, the moment when each player speaks and / or makes a decision is defined as a game moment. That is, a game session consists of a series of consecutive game moments, and in each game moment, a player (or a game bot) speaks and / or makes a decision.

[0041] The first game information refers to the game information generated at the current game moment. It can be understood as the game information generated by the speaking player at the current game moment through their client's speaking and / or decision-making operations. After the first game information is generated, the client will send it to the first server, which will then send it to the second server.

[0042] The game information set refers to the collection of game information generated from historical game moments in a game match. It can be understood as the game information generated by players speaking during historical game moments through their client-side speaking and / or decision-making actions. Historical game moments can be understood as all game moments earlier than the current game moment, or as a subset of game moments earlier than the current game moment.

[0043] Typically, the second server has an information storage module where game information sets are stored. These game information sets include information from historical game moments earlier than the current game moment. In an optional implementation, starting from the beginning of the game, the second server overlays each piece of first game information generated at each current game moment into the information storage module, forming a game information set. Therefore, the game information set can also be understood as a collection of multiple first game information sets. When the first server sends the first game information generated at the current game moment to the second server, the second server stores this first game information in the information storage module, overlaying it with the game information in the game information set to form an updated game information set. For the current game moment, the updated game information set is a collection that includes the context content of the game process.

[0044] In one optional implementation of this embodiment, before the game begins, the method further includes the following steps S11 to S12:

[0045] Step S11: In response to receiving the robot request information sent by the first server, allocate a first number of robots to the game match, wherein the robot request information is generated by the first server in response to the number of clients matching the game match not meeting a preset number threshold, and includes the first number of clients not matched in the game match.

[0046] The robot request information can be understood as the information sent by the first server to the second server to allocate game robots when the number of clients matching the game match does not meet the preset number threshold. This includes the number of missing clients in the game match. In this embodiment, the number of missing clients in the game match is defined as the first number.

[0047] The preset threshold can be understood as the number of clients required for a game match. A game can only start when the number of clients matched to enter the game reaches the preset threshold. For example, in a twelve-player game of Werewolf, the game can only start if at least 12 clients are matched to enter the game.

[0048] In a specific game scenario, when the first server receives a game start instruction from any client (defined as the third client in this embodiment) intending to participate in the game, it establishes a game and matches the third client into the game. Within a certain period after the game is established (e.g., within ten minutes of its establishment), if the number of clients matched in the game does not meet a preset threshold (e.g., twelve), the first server sends a robot request message to the second server. This robot request message includes a first number of clients not yet matched in the game (e.g., in a twelve-player Werewolf game, if eight clients are matched within ten minutes of the game's establishment, then the first number is four). Upon receiving the robot request message from the first server, the second server allocates the first number (e.g., four) of game robots to the game.

[0049] Step S12: In response to receiving robot matching information sent by the first server, the robot matching information is input into the first model, wherein the robot matching information is generated by the first server for matching the first number of robots in the game, and includes at least the identity information of each of the first number of robots in the game.

[0050] The robot matching information can be understood as the successful matching information generated after the first server matches the assigned game robots into the game, including the robot's identity information in the game. For example, in the Werewolf game, when the second server assigns 4 game robots to the game, the first server will match the 4 game robots into the game. A successful match will generate robot matching information including the identity information of the 4 game robots in the game, such as robot 1 being the Seer, robot 2 being the Werewolf, robot 3 being the Villager, and robot 4 being the Hunter.

[0051] After the first server matches the game bot to enter the game and generates bot matching information, it will send the bot matching information to the first server. After receiving the bot matching information, the first server will establish a connection with the first model (for example, by calling the first model through the first model's interface) and send the bot matching information, including the bot's identity information, to the first model. The first model will be preset so that when the first model performs logical reasoning later, it can perform game reasoning in the role indicated by the identity information.

[0052] Step S220: In response to satisfying the first information interaction condition, the updated game information set is input into the first model to obtain the first feedback information output by the first model. The first model is a language processing model pre-trained with a preset interaction instruction set, and the interaction instruction set includes at least a first instruction for instructing the game logic of the game match.

[0053] The first information interaction condition can be understood as the condition for the second server to interact with the first model. If the first information interaction condition is not met, the information interaction described in this step will not occur between the second server and the first model.

[0054] In one optional implementation of this embodiment, the first information interaction condition includes the next game moment after the current game moment being the moment when the first feedback information needs to be broadcast.

[0055] The moment when the first feedback information needs to be broadcast can be understood as the moment when the game bot participating in the game speaks, and the current game moment can be understood as the moment when the player participating in the game speaks. If the next game moment is the moment when the game bot needs to speak, it means that the game information set updated at the current game moment needs to be sent to the first model, that is, the condition for information interaction between the second server and the first model is met.

[0056] Based on this, after the second server receives the first game information generated by the current game moment from the first server and updates the game information set with the first game information, it is also necessary to determine whether the next game moment is a moment when the first feedback information needs to be broadcast, that is, whether it is a moment when the game robot needs to speak. The subsequent execution steps are determined based on different judgment results, specifically:

[0057] First, in response to the next game moment being a moment when the first feedback information needs to be broadcast, the updated game information set is input into the first model, the first feedback information output by the first model is obtained, and the updated game information set is updated using the first feedback information.

[0058] If the next game moment in the current game requires the game bot to speak, the second server will input the updated game information set into the first model. The first model will then perform logical reasoning based on the game logic and output first feedback information to the second server. This first feedback information is the output information generated by the first model based on the updated game information set for the current game moment. This first feedback information will be broadcast as the game bot's speech content in the next game moment. Therefore, the first feedback information can also be understood as the game information generated in the next game moment.

[0059] After the second server obtains the first feedback information, it will not only send the first feedback information to the first server for broadcast, but also store the first feedback information in the information storage module, and overlay it with the game information in the updated game information set to obtain a new game information set.

[0060] Second, in response to the fact that the next game moment of the current game moment is not a moment when the first feedback information needs to be broadcast, the system receives the second game information sent by the first server and updates the updated game information set using the second game information, wherein the second game information is the game information generated in the next game moment of the current game moment.

[0061] If the next game moment is not a moment when the game bot needs to speak, then it is not necessary to obtain the game bot's speech content in the next game moment, and the second server does not need to input the updated game information set into the first model. Therefore, the operation performed by the second server in the current game moment is simply to store the first game information in the information storage module and update the game information set. After the game enters the next game moment (i.e., after the speaking player in the next game moment speaks), the first server will send the second game information generated in the next game moment (i.e., the speaking player's speech content) to the second server. The second server will store the second game information in the information storage module and continue to update the game information set.

[0062] The interactive instruction set can be understood as a set of instructions constructed for the game logic of a game match. By pre-training the first model with this instruction set, the first model can learn the game logic of the game match, and thus perform logical reasoning based on the game logic, outputting feedback information that conforms to the game logic. The interactive instruction set includes at least a first instruction for indicating the game logic of the game match.

[0063] Taking ChatGPT as an example, the language processing principle of ChatGPT is to convert natural language into a corresponding input token (i.e., a special code) using the BPE algorithm. This input token is then input into the model as a vector. The model calculates and outputs a new vector based on the input vector. This new vector is then processed by a sampling algorithm to obtain the corresponding output token. Finally, the output token is converted back into natural language to respond to questions. ChatGPT has three main characteristics: First, the model lacks memory; it cannot store the context of the game. Second, the longer the input text, the longer the algorithm takes. As the game progresses, the game information set inevitably contains more and more information, and the time required to output feedback based on this information set increases. In other words, the game robot's speaking speed decreases as the game progresses. Third, there is a limit to the number of input texts. When the game information set exceeds this limit, it cannot be input into the model. Therefore, as the game progresses, the game robot will eventually be unable to speak.

[0064] Regarding the first feature mentioned above (the model lacks memory function), the method provided in this embodiment is to use a second server to store all game information (i.e., context content) during the game, and then input all game information into the first model when the first information interaction condition is met.

[0065] Regarding the second characteristic mentioned above (the longer the input text, the longer the algorithm takes), the method provided in this embodiment abandons the above-mentioned language processing principle and simplifies the game information through a JSON-formatted interactive instruction set to obtain a relatively fast response from the model. Since JSON (JavaScript Object Notation) is a lightweight data exchange format, converting the game information in the game information set to JSON format through a JSON-formatted interactive instruction set can compress the text length of the game information in the game information set.

[0066] Based on this, in an optional implementation provided in this embodiment, the interaction instruction set further includes at least a second instruction for indicating the information interaction format of the game match. Before inputting the updated game information set into the first model, it is necessary to convert the updated game information set into a second game information set according to the interaction instruction set. The game information included in the updated game information set is expressed in a first language format, and the game information included in the second game information set is expressed in a second language format. The second language format is the information interaction format indicated by the second instruction.

[0067] Specifically, the first language format can be natural language format, and the second language format can be JSON format. The second server converts the game information in the game information set through the interaction instruction set to form a game information set with a shortened text length. In this embodiment, the game information set with a shortened text length is defined as the second game information set.

[0068] The second server inputs the second game information set into the first model to obtain the second feedback information output by the first model. Since the first model has been pre-trained with the interaction instruction set, the second feedback information output by the first model will be expressed in the second language format. The second server further needs to convert the second feedback information into first feedback information expressed in the first language format according to the interaction instruction set.

[0069] Regarding the third feature mentioned above (the number of input texts is limited), the method provided in this embodiment is to use an elimination algorithm to prioritize the retention of behavioral operations such as voting and skill usage in the game information, discard some chat text, and input key content from the game information into the first model to obtain relatively accurate feedback.

[0070] Based on this, in one optional implementation provided in this embodiment, before inputting the updated game information set into the first model, it is necessary to determine whether the information content of the updated game information set meets the second information interaction condition. The second information interaction condition includes that the information content of the updated game information set is not greater than a preset information content threshold. The preset information content threshold can be understood as a limitation on the amount of input text imposed by the first model. The subsequent execution steps are determined based on different judgment results, specifically as follows:

[0071] First, in response to the updated game information set having an information content not exceeding the preset information content threshold, the updated game information set is input into the first model.

[0072] In the early stages of a game, the amount of game information contained in the game information set is limited and will not exceed the preset information threshold. The game information set can be directly input into the first model, and the first model will respond relatively quickly.

[0073] Second, in response to the updated game information set having more information than the preset information threshold, a preset first algorithm is used to convert the updated game information set into a third game information set, and the third game information set is input into the first model, wherein the third game information set has less information than the updated game information set.

[0074] As the game progresses, the amount of game information in the game information set increases significantly. Directly inputting this game information set into the first model results in a progressively slower response time. When the amount of information in the game information set exceeds a preset threshold, the first model will become unresponsive. To address the issue of slow or unresponsive first models caused by excessive game information, this embodiment provides a solution: using a preset first algorithm to simplify the game information set, resulting in a simplified game information set. In this embodiment, the simplified game information set is defined as the third game information set.

[0075] For language-based logic reasoning games, game information can be categorized into two types: verbal information and operational information. Verbal information can be understood as player chat content, while operational information includes player decisions, votes, and skill usage. Operational information is the key component; without it, the first-order model cannot perform logical reasoning or output feedback consistent with the game's logic. In terms of the time of game information generation, it can be divided into information generated in the current game moment and information generated in previous game moments. Information generated in the current game moment and information generated in recent previous game moments are crucial components; without this information, the first-order model cannot output feedback consistent with the game context.

[0076] Therefore, in an optional implementation, the first algorithm is used to delete the language information corresponding to the second historical game time in the updated game information set according to the game time corresponding to each game information in the updated game information set, thereby forming the third game information set. The second historical game time includes multiple game times that are far away from the current game time in the historical game times.

[0077] The second historical game moment refers to a game moment that is relatively far removed from the current game moment. This second historical game moment changes as the game progresses. For example, if the current game moment is the 15th game moment, the second historical game moment will increase by one when the game progresses to the 16th game moment. Furthermore, the second historical game moment is related to a preset information threshold. For instance, if the preset information threshold is the total amount of game information generated over ten game moments, and the current game moment is the 15th game moment, then the second historical game moment includes multiple game moments from the 1st to the 5th game moment.

[0078] The first algorithm prioritizes deleting language information corresponding to the second game moment, while retaining the operation information corresponding to each game moment in the game information set. If the game progresses too long, the boundary of the second historical moment will shift towards the current game moment. In short, the first algorithm will, based on a preset information threshold, first delete language information generated by game moments farther from the current game moment, then gradually delete language information generated by game moments closer to the current game moment, and even delete language information generated by the current game moment, but will retain the operation information generated by each game moment.

[0079] As described above, the interaction instruction set includes at least a first instruction for instructing the game logic of a match and a second instruction for instructing the information interaction format of the match. In one specific implementation, the information interaction format indicated by the second instruction is JSON format. Specifically, the first instruction is edited using the JavaScript programming language to form the game logic text, which serves as the interaction instruction set. Therefore, the interaction instruction set not only defines the logical reasoning method of the first model but also defines the language format of the input and output data of the first model.

[0080] In one optional implementation, the interaction instruction set also includes sample text for debugging the first model, making its output more accurate. However, as the game progresses and the amount of information in the game information set becomes large, the sample text may not be included in the interaction instruction set.

[0081] In another optional implementation, the interaction instruction set also includes identity text representing the robot's identity information, which is used to preset the role identity of the first model when performing logical reasoning. Specifically, after the second server assigns a robot to the game and receives robot matching information including robot identity information sent by the first server, the second server will write the robot identity information into the interaction instruction set as part of the interaction instruction set.

[0082] Taking the Werewolf game as an example, the interaction instruction set can be any of the contents shown in Table 1, Table 2, and Table 3. The interaction instruction set shown in Table 2 is an optimization of the interaction instruction set shown in Table 1. The interaction instruction set shown in Table 3 and the interaction instruction set shown in Table 2 indicate different robot role identities. Of course, the interaction instruction set can also include other contents. The contents shown in Table 1, Table 2, and Table 3 are only examples and are not specific restrictions.

[0083] Table 1 Werewolf Game Interactive Command Set 1

[0084]

[0085]

[0086] Table 2 Werewolf Game Interaction Command Set 2

[0087]

[0088]

[0089] Table 3 Werewolf Game Interaction Command Set 3

[0090]

[0091]

[0092]

[0093] Different language logic reasoning games require the construction of specific interactive instruction sets. In this embodiment, the interactive instruction set constructed for the Werewolf game is used as an example for illustration and should not be taken as a limitation for actual use.

[0094] In one optional implementation provided in this embodiment, the aforementioned interactive instruction set can be input into the first model when the second server establishes a connection with the first model to train the first model. When it is the game robot's turn to speak, the second server only needs to input the game information set into the first model.

[0095] In another optional implementation provided in this embodiment, the above-mentioned set of interactive instructions and set of game information need to be input into the first model every time it is the turn of the game robot to speak.

[0096] As mentioned earlier, the interaction instruction set not only defines the logical reasoning method of the first model, but also defines the language format of the input and output data of the first model. Therefore, before inputting the game information set into the first model, the method provided in this embodiment further includes: the second server performing language format conversion on the game information set according to the language format defined in the interaction instruction set. The game information set after language format conversion is shown in Table 4.

[0097] Table 4 Game Information Set

[0098]

[0099]

[0100]

[0101]

[0102] In one optional implementation provided in this embodiment, the first model is an independent model deployed on a third server. To enable the first model to perform logical deductions based on the game logic of the game match, the method provided in this embodiment further includes the following steps S21 to S22 before inputting the game information set into the first model:

[0103] Step S21: Call the first model through the interface provided by the third server for the first model.

[0104] Step S22: Input the interaction instruction set into the first model and train the first model so that the first model can output the first feedback information that conforms to the game logic of the game match according to the first instruction defined in the interaction instruction set and the updated game information set.

[0105] Step S230: Send the first feedback information to the first server so that the first server broadcasts the first feedback information in the next game moment of the current game moment.

[0106] In one optional implementation, after the second server sends the first feedback information to the first server, the first server will send the first feedback information to all clients matched in the game match in the next game moment, so that the clients will display the first feedback information on the screen to present it to the players.

[0107] The first embodiment described above uses a second server as the execution subject and provides a method for information interaction between the second server and the first server, as well as between the second server and the first model. Through this method, the game robot can perform reasoning in a logic reasoning game in a way that conforms to the game logic, just like a human.

[0108] Typically, the second server consists of multiple bot docking servers. Each bot docking server is responsible for controlling the game behavior of one or more game bots in a game match. When a game bot needs to join a game match, the first server selects an available bot docking server based on the load level of the bot docking servers in the second server and sends the bot request information. This not only ensures balanced communication task load but also guarantees high overall service availability in the event of a single point of failure.

[0109] It should be noted that the examples in the first embodiment above are only for explaining the methods described in this application and are not intended to limit actual use. The information interaction methods provided in this application include, but are not limited to, the methods described in the first embodiment.

[0110] The second embodiment of this application provides an information interaction method, which is applied to... Figure 1The first server 101 in the application system shown realizes information interaction between the first server 101 and the second server 102, and between the first server 101 and the client 104.

[0111] Figure 3 This is a flowchart of the information interaction method provided in this embodiment. The following is in conjunction with... Figure 3 The method provided in this embodiment will be described in detail. The embodiments described below are used to explain the technical solution of this application and are not intended to limit actual use.

[0112] like Figure 3 As shown, the information interaction method provided in this embodiment includes the following steps S310 to S330.

[0113] Step S310: In response to receiving the first game information sent by the first client, the first game information is sent to the second server so that the second server updates the game information set using the first game information.

[0114] The first client is any one of at least one clients matched in the game match, which can be understood as the client corresponding to the player speaking at the current game moment.

[0115] The first game information is generated by the first client detecting input events related to the first game information. Input events related to the first game information can be understood as the speaking player entering chat content, voting, using skills, or other operations on the first client.

[0116] The first game information is the game information generated at the current game moment of the game match, and the game information set is the collection of game information generated at historical game moments of the game match.

[0117] After detecting an input event, the first client generates first game information regarding chat content, voting decisions, skill usage, etc., and sends it to the second client. Upon receiving the first game information, the second client updates its game information set accordingly. The specific operations performed by the second client have been described in detail in the first embodiment of this application and will not be repeated here.

[0118] Step S320: Receive first feedback information sent by the second server. The first feedback information is obtained by the second server inputting the updated game information set into a first model. The first model is a language processing model pre-trained using a preset set of interaction instructions, which includes at least a first instruction for instructing the game logic of the game match.

[0119] The first feedback information is the output information generated by the first model based on the game information set updated at the current game moment. Since the first model is a language processing model pre-trained using an interactive instruction set, which includes first instructions for instructing the game logic of the game, the first feedback information generated by the first model is output information that conforms to the game logic obtained through logical reasoning. The method for generating the first feedback information has been described in detail in the first embodiment of this application and will not be repeated here.

[0120] After the second server obtains the first feedback information output by the first model, it will send the first feedback information to the first server, so that the first server can broadcast the first feedback information in the next game moment of the current game moment.

[0121] Step S330: The first feedback information is sent to the second client in the next game moment of the current game moment, so that the second client displays the first feedback information on the screen, wherein the second client includes each of the at least one client matched in the game match.

[0122] Since the first feedback information is the output information generated by the first model based on the game information set updated at the current game moment—that is, the content of the game bot's speech in the next game moment—the first server, upon receiving the first feedback information, will send it to all clients in the next game moment. In this embodiment, all clients matched in the game are defined as the second clients. The second clients can then display the first feedback information on their screens to the players using rendering and other technologies.

[0123] In an optional implementation provided in this embodiment, before the start of the game, the method provided in this embodiment further includes the following steps S31 to S33:

[0124] Step S31: In response to receiving a game start instruction sent by a third client, match the third client with the game. The game includes a preset number threshold for indicating the start conditions of the game. The start conditions include the number of clients required to be matched in the game. The third client is any client that is to participate in the game.

[0125] The game start command can be understood as the command information generated by the third client in response to the player's operation to start the game, such as the player's click to start the game, or the player's selection of the game and / or game room.

[0126] When a player who intends to participate in a game joins the game through a third client, the third client will generate a game start command and send it to the first server so that the first server can match the third client with a game. This game can be an existing game that has already been matched with other clients, or a newly created game that has not been matched with any clients.

[0127] The preset threshold can be understood as the preset number of clients required for a game match. When the number of clients matched to enter the game match reaches the preset threshold, the starting condition is met, and the game match can begin. For example, in a twelve-player game of Werewolf, the number of clients matched to enter the game match must reach 12 before the game can begin.

[0128] Step S32: In response to the fact that the number of clients matched in the game match does not meet the preset number threshold, a robot request message is sent to the second server so that the second server allocates a first number of robots to the game match. The robot request message is generated by the first server in response to the fact that the number of clients matched in the game match does not meet the preset number threshold, and includes the first number of clients not matched in the game match.

[0129] The robot request information can be understood as the information sent by the first server to the second server to allocate game robots when the number of clients matching the game match does not meet the preset number threshold. This includes the number of missing clients in the game match. In this embodiment, the number of missing clients in the game match is defined as the first number.

[0130] In one optional implementation, each game session uses a timer that counts down from the moment the game session is created. The creation moment is defined as the point at which the first server receives a game start command from a third client and matches the third client with a new game session. When the timer finishes counting down, if the number of clients matched with the game session has not yet reached a preset threshold, the first server sends a bot request to the second server, allowing the second server to assign a bot to replace the missing player. This approach saves players' waiting time and improves their gaming experience.

[0131] Step S33: Match the first number of robots with the game match and send the robot matching information to the second server so that the second server inputs the robot matching information into the first model, wherein the robot matching information includes at least the identity information of each of the first number of robots in the game match.

[0132] The second server allocates the initial number of game bots to the game matches. The first server matches each game bot into a game and assigns each bot a role, for example: bot 1 is the Seer, bot 2 is the Werewolf, bot 3 is the Villager, and bot 4 is the Hunter. After matching, the first server sends the bot matching information, including the bot's role, to the second server so that the second server can preset the first model.

[0133] The second embodiment described above uses the first server as the execution subject and provides a method for information interaction between the first server and the second server, as well as between the first server and the client. Other related content can be found in the detailed description in the first embodiment of this application, and will not be repeated here.

[0134] The third embodiment of this application provides an information interaction method, which is applied to... Figure 1 The information interaction between the first server 101, the second server 102, the language processing model 103, and the client 104 in the application system diagram shown.

[0135] The following are two optional implementations of the information interaction method provided in this embodiment. Of course, the information interaction method in the actual game scenario can be a combination of the following two implementations, or an optimization of the following two implementations.

[0136] Figure 4 This is a signaling flowchart of an information interaction method provided in this embodiment.

[0137] like Figure 4 As shown, the information interaction method provided in this embodiment may include the following steps S401 to S412:

[0138] Step S401: The first client detects the chat input and / or decision-making operations performed by the player on the first client and generates first game information. Here, the first client is any one of the multiple clients matched in the game match. The first game information is the game information generated at the current game moment of the game match.

[0139] In step S402, the first client sends the first game information to the first server.

[0140] In step S403, the first server sends the first game information to the second server.

[0141] In step S404, the second server stores the first game information in the information storage module and updates the game information set using the first game information, wherein the game information set is a collection of game information generated from historical game moments of the game.

[0142] In step S405, in response to the next game moment being the speaking moment for the game robot, the second server converts the updated game information set into a second game information set according to the information interaction format defined in the interaction instruction set. The game information in the updated game information set is presented in natural language format, while the game information in the second game information set is presented in an information interaction format, such as JSON format.

[0143] In step S406, the second server sends the second game information set to the first model, which is a language processing model pre-trained with a preset set of interactive instructions.

[0144] In step S407, the first model, based on the game logic defined in the interaction instruction set, performs logical reasoning based on all the game information included in the second game information set to obtain the second feedback information. Since the second feedback information is the output information derived by the first model based on the interaction instruction set, it is presented in the information interaction format defined in the interaction instruction set.

[0145] In step S408, the first model sends the second feedback information to the second server.

[0146] In step S409, the second server converts the second feedback information into the first feedback information according to the information interaction format defined in the interaction instruction set, wherein the first feedback information is in natural language format.

[0147] In step S410, the second server sends the first feedback information to the first server.

[0148] In step S411, the first server sends the first feedback information to the second client in the next game moment after the current game moment. The second client is each of the multiple clients matched in the game match.

[0149] In step S412, the second client displays the first feedback information on the screen to the player.

[0150] Figure 5 This is a flowchart illustrating another information interaction method provided in this embodiment.

[0151] like Figure 5As shown, the information interaction method provided in this embodiment may include the following steps S501 to S505:

[0152] Step S501, game match initialization, may specifically include the following steps S501-1 to S501-3:

[0153] Step S501-1: The third client responds to the player's opening operation and generates a game start command, wherein the third client is any client that is to participate in the game.

[0154] Step S501-2: The third client sends the game start command to the first server;

[0155] Step S501-3: The first server matches game matches for the third client. The game matches include a preset quantity threshold for indicating the starting conditions. The starting conditions include the number of clients required to match the game.

[0156] Step S502, assigning game robots, may specifically include the following steps S502-1 to S502-4:

[0157] Step S502-1: In response to the fact that the number of clients matched in the game does not meet the preset number threshold, the first server sends robot request information to the second server, wherein the robot request information includes a first number of clients not matched in the game.

[0158] Step S502-2: The second server allocates the first number of game bots to the game match;

[0159] Step S502-3: The first server matches the first number of game robots into the game and obtains robot matching information, wherein the robot matching information includes at least the identity information of each game robot in the first number of game robots in the game.

[0160] Step S502-4: The first server sends the robot matching information to the second server.

[0161] Step S503, connecting the first model, may specifically include the following steps S503-1 to S503-2:

[0162] Step S503-1: The first server calls the first model through the first model interface provided by the third-party service;

[0163] In step S503-2, the first server sends the set of interaction instructions that defines the game logic, the information interaction format, and the robot's identity information to the first model to train the first model.

[0164] Step S504: Execute the game turn logic, which may specifically include the following steps S504-1 to S504-10:

[0165] Step S504-1: In response to detecting chat content input and / or decision-making operations, the first client generates first game information corresponding to the current game moment, wherein the first client is any one of the multiple clients matched in the game match.

[0166] Step S504-2: The first client sends the first game information to the first server.

[0167] Step S504-3: The first server sends the first game information to the second server.

[0168] In step S504-4, the second server stores the first game information in the information storage module and updates the game information set using the first game information. The game information set is a collection of game information generated from historical game moments during the game.

[0169] In step S504-5, if the next game time is the time for the game robot to speak, the second server sends the updated game information set to the first model; if the next game time is not the time for the game robot to speak, return to step S504-1.

[0170] Step S504-6: The first model, based on the game logic defined in the interaction instruction set, performs logical reasoning based on all the game information included in the updated game information set to obtain the first feedback information.

[0171] Step S504-7: The first model sends the first feedback information to the second server.

[0172] Step S504-8: The second server sends the first feedback information to the first server.

[0173] Step S504-9: The first server sends the first feedback information to the second client in the next game moment after the current game moment. The second client includes each of the multiple clients matched in the game.

[0174] In step S504-10, the second client displays the first feedback information on the screen to the player.

[0175] Step S505: Determine whether the game has ended. Specifically, if the game has ended, proceed with game settlement; if the game has not ended, return to step S504-1.

[0176] The third embodiment described above provides an optional implementation of the information interaction method between the first server, the second server, the first model, and the client. It should be noted that the examples in the third embodiment are only for explaining the methods described in this application and are not intended to limit actual use. The information interaction methods provided in this application include, but are not limited to, the methods described in the third embodiment.

[0177] The fourth embodiment of this application provides an information interaction system. Figure 6 This is a schematic diagram of the information interaction system provided in this embodiment.

[0178] like Figure 6 As shown, the information interaction system provided in this embodiment includes: at least one client 601, a first server 602, a second server 603, and a first model 604. This information interaction system performs the following operations:

[0179] The at least one client 601 detects an input event for the first game information and sends the first game information to the first server 602. The first game information is the game information generated at the current game moment of the game match.

[0180] The first server 602 receives the first game information sent by the at least one client 601 and sends the first game information to the second server 603;

[0181] The second server 603 receives the first game information sent by the first server 602 and updates the game information set using the first game information. The game information set is a collection of game information generated from historical game moments of the game match.

[0182] The second server 603, in response to satisfying the first information interaction condition, inputs the updated game information set into the first model 604, wherein the first model 604 is a language processing model pre-trained with a preset interaction instruction set, the interaction instruction set including at least a first instruction for instructing the game logic of the game match;

[0183] The first model 604 calculates and obtains first feedback information based on the updated game information set, and sends the first feedback information to the second server 603;

[0184] The second server 603 receives the first feedback information and sends the first feedback information to the first server 602;

[0185] The first server 602 receives the first feedback information sent by the second server 603, and sends the first feedback information to the at least one client 601 in the next game moment of the current game moment;

[0186] The at least one client 601 displays the first feedback information on the screen.

[0187] The fifth embodiment of this application provides a game service system. Figure 7 This is a schematic diagram of the structure of the game service system provided in this embodiment.

[0188] like Figure 7 As shown, the game service system provided in this embodiment includes a first server 701 and a second server 702. The first server 701 is connected to at least one client, and the second server 702 is connected to a first model. The first model is a language processing model pre-trained using a preset set of interactive instructions, and the set of interactive instructions includes at least a first instruction for instructing the game logic of the game session.

[0189] The first server 701 is configured to, in response to receiving first game information sent by the first client, send the first game information to the second server 702, wherein the first client is any one of the at least one clients matched in the game match, and the first game information is generated by the first client detecting an input event related to the first game information, and is game information generated at the current game moment of the game match; receive first feedback information sent by the second server 702; and send the first feedback information to the second client at the next game moment of the current game moment, so that the second client displays the first feedback information on the screen, wherein the second client includes each of the at least one clients matched in the game match;

[0190] The second server 702 is configured to receive the first game information sent by the first server 701, and update the game information set using the first game information, wherein the game information set is a collection of game information generated at historical game moments of the game; in response to satisfying the first information interaction condition, input the updated game information set into the first model to obtain the first feedback information output by the first model; and send the first feedback information to the first server 701.

[0191] In one optional implementation provided in this embodiment, the first model is deployed on the second server.

[0192] In another optional implementation provided in this embodiment, the first model is deployed on a third server, and the second server connects to the first model through an interface provided by the third server for the first model.

[0193] In one specific implementation, the first model is the ChatGPT natural language processing model.

[0194] The sixth embodiment of this application provides an information interaction device. Figure 8 This is a schematic diagram of the information interaction device provided in this embodiment.

[0195] like Figure 8 As shown, the information interaction device provided in this embodiment includes: a game information receiving unit 801, a feedback information acquisition unit 802, and a feedback information sending unit 803.

[0196] The game information receiving unit 801 is used to receive first game information sent by the first server and update the game information set using the first game information, wherein the first game information is the game information generated at the current game moment of the game, and the game information set is a collection of game information generated at the historical game moments of the game.

[0197] Optionally, before the step of receiving the first game information sent by the first server and updating the game information set using the first game information, the method is further used to:

[0198] In response to receiving a robot request message sent by the first server, a first number of robots are allocated to the game match, wherein the robot request message is generated by the first server in response to the number of clients matching the game match not meeting a preset number threshold, and includes the first number of clients not matched in the game match;

[0199] In response to receiving robot matching information sent by the first server, the robot matching information is input into the first model, wherein the robot matching information is generated by the first server for matching the first number of robots in the game, and includes at least the identity information of each of the first number of robots in the game.

[0200] The feedback information acquisition unit 802 is used to input the updated game information set into the first model in response to the fulfillment of the first information interaction condition, and obtain the first feedback information output by the first model, wherein the first model is a language processing model pre-trained with a preset interaction instruction set, and the interaction instruction set includes at least a first instruction for instructing the game logic of the game match.

[0201] Optionally, the interaction instruction set may further include at least a second instruction for indicating the information interaction format of the game session;

[0202] Before the step of inputting the updated game information set into the first model, it is also used for:

[0203] According to the interaction instruction set, the updated game information set is converted into a second game information set, wherein the game information included in the updated game information set is expressed in a first language format, and the game information included in the second game information set is expressed in a second language format, and the second language format is the information interaction format indicated by the second instruction;

[0204] The step of inputting the updated game information set into the first model and obtaining the first feedback information output by the first model includes:

[0205] The second game information set is input into the first model to obtain the second feedback information output by the first model. The second feedback information is expressed in the second language format.

[0206] According to the interactive instruction set, the second feedback information is converted into the first feedback information, and the first feedback information is expressed in the first language format.

[0207] Optionally, the first information interaction condition includes the next game moment after the current game moment being the moment when the first feedback information needs to be broadcast;

[0208] After the steps of receiving the first game information sent by the first server and updating the game information set using the first game information, the method is further used to:

[0209] Determine whether the next game moment after the current game moment is the moment when the first feedback information needs to be broadcast;

[0210] The step of responding to the fulfillment of the first information interaction condition by inputting the updated game information set into the first model and obtaining the first feedback information output by the first model includes:

[0211] In response to the next game moment being a moment when the first feedback information needs to be broadcast, the updated game information set is input into the first model, the first feedback information output by the first model is obtained, and the updated game information set is updated using the first feedback information.

[0212] Optionally, it can also be used for:

[0213] In response to the fact that the next game moment of the current game moment is not a moment when the first feedback information needs to be broadcast, the system receives the second game information sent by the first server and updates the updated game information set with the second game information, wherein the second game information is the game information generated in the next game moment of the current game moment.

[0214] Optionally, before the step of inputting the updated game information into the first model, the method is further used for:

[0215] Determine whether the information content of the updated game information set meets the second information interaction condition, wherein the second information interaction condition includes that the information content of the updated game information set is not greater than a preset information content threshold;

[0216] The step of inputting the updated game information set into the first model includes:

[0217] If the updated game information set does not exceed the preset information threshold, the updated game information set is input into the first model.

[0218] Optionally, it can also be used for:

[0219] In response to the updated game information set having more information than the preset information threshold, a preset first algorithm is used to convert the updated game information set into a third game information set, and the third game information set is input into the first model, wherein the third game information set has less information than the updated game information set.

[0220] Optionally, the updated game information set includes language information and operation information;

[0221] The step of using a preset first algorithm to convert the updated game information set into a third game information set includes:

[0222] Using the first algorithm, based on the game time corresponding to each game information in the updated game information set, the language information corresponding to the second historical game time in the updated game information set is deleted to form the third game information set, wherein the second historical game time includes multiple game times that are far from the current game time in the historical game time.

[0223] Optionally, the first model is deployed on a third server; before the step of inputting the updated game information set into the first model and obtaining the first feedback information output by the first model, it is further used for:

[0224] The first model is invoked through the interface provided by the third server for the first model;

[0225] The interaction instruction set is input into the first model to train the first model, so that the first model can output the first feedback information that conforms to the game logic of the game according to the first instruction defined in the interaction instruction set and the updated game information set.

[0226] The feedback information sending unit 803 is used to send the first feedback information to the first server so that the first server broadcasts the first feedback information in the next game moment of the current game moment.

[0227] The seventh embodiment of this application provides an information interaction device. Figure 9 This is a schematic diagram of the information interaction device provided in this embodiment.

[0228] like Figure 9 As shown, the information interaction device provided in this embodiment includes: a game information sending unit 901, a feedback information receiving unit 902, and a feedback information sending unit 903.

[0229] The game information sending unit 901 is configured to, in response to receiving first game information sent by a first client, send the first game information to a second server so that the second server updates the game information set using the first game information. The first client is any one of at least one client matched in the game match. The first game information is generated by the first client detecting an input event related to the first game information and is the game information generated at the current game moment of the game match. The game information set is a collection of game information generated at historical game moments of the game match.

[0230] Optionally, before the step of sending the first game information to the second server in response to receiving the first game information sent by the first client, the method is further used to:

[0231] In response to receiving a game start instruction sent by a third client, the system matches the third client with the game, wherein the game includes a preset number threshold for indicating the start conditions of the game, and the start conditions include the number of clients required to be matched in the game, wherein the third client is any client that is to participate in the game.

[0232] In response to the fact that the number of clients matched in the game match does not meet the preset number threshold, a robot request message is sent to the second server so that the second server allocates a first number of robots to the game match. The robot request message is generated by the first server in response to the fact that the number of clients matched in the game match does not meet the preset number threshold, and includes the first number of clients not matched in the game match.

[0233] The game match is matched for the first number of robots, and the robot matching information is sent to the second server so that the second server inputs the robot matching information into the first model, wherein the robot matching information includes at least the identity information of each of the first number of robots in the game match.

[0234] The feedback information receiving unit 902 is used to receive first feedback information sent by the second server. The first feedback information is obtained by the second server inputting the updated game information set into the first model. The first model is a language processing model pre-trained through a preset interaction instruction set. The interaction instruction set includes at least a first instruction for instructing the game logic of the game match.

[0235] The feedback information sending unit 903 is used to send the first feedback information to the second client in the next game moment of the current game moment, so that the second client displays the first feedback information on the screen, wherein the second client includes each of the at least one client matched in the game match.

[0236] The eighth embodiment of this application provides a game robot. Figure 10 This is a structural schematic diagram of the game robot provided in this embodiment.

[0237] like Figure 10 As shown, the game robot provided in this embodiment includes: an information storage module 1001 and an information processing module 1002;

[0238] The information storage module 1001 is used to store game information generated during game matches;

[0239] The information processing model 1002 is used to perform the following operations:

[0240] Receive first game information sent by the first server, and update the game information set using the first game information, wherein the first game information is the game information generated at the current game moment of the game, and the game information set is the collection of game information generated at the historical game moments of the game;

[0241] In response to satisfying the first information interaction condition, the updated game information set is input into the first model to obtain the first feedback information output by the first model, wherein the first model is a language processing model pre-trained with a preset interaction instruction set, and the interaction instruction set includes at least a first instruction for instructing the game logic of the game match.

[0242] The first feedback information is sent to the first server so that the first server broadcasts the first feedback information in the next game moment of the current game moment.

[0243] Optionally, the interaction instruction set may further include at least a second instruction for indicating the information interaction format of the game session;

[0244] Before the step of inputting the updated game information set into the first model, the following operations are also performed:

[0245] According to the interaction instruction set, the updated game information set is converted into a second game information set, wherein the game information included in the updated game information set is expressed in a first language format, and the game information included in the second game information set is expressed in a second language format, and the second language format is the information interaction format indicated by the second instruction;

[0246] The step of inputting the updated game information set into the first model and obtaining the first feedback information output by the first model includes:

[0247] The second game information set is input into the first model to obtain the second feedback information output by the first model. The second feedback information is expressed in the second language format.

[0248] According to the interactive instruction set, the second feedback information is converted into the first feedback information, and the first feedback information is expressed in the first language format.

[0249] Optionally, the first information interaction condition includes the next game moment after the current game moment being the moment when the first feedback information needs to be broadcast;

[0250] After the steps of receiving the first game information sent by the first server and updating the game information set using the first game information, the following operations are also performed:

[0251] Determine whether the next game moment after the current game moment is the moment when the first feedback information needs to be broadcast;

[0252] The step of responding to the fulfillment of the first information interaction condition by inputting the updated game information set into the first model and obtaining the first feedback information output by the first model includes:

[0253] In response to the next game moment being a moment when the first feedback information needs to be broadcast, the updated game information set is input into the first model, the first feedback information output by the first model is obtained, and the updated game information set is updated using the first feedback information.

[0254] Optionally, the following operations may also be performed:

[0255] In response to the fact that the next game moment of the current game moment is not a moment when the first feedback information needs to be broadcast, the system receives the second game information sent by the first server and updates the updated game information set with the second game information, wherein the second game information is the game information generated in the next game moment of the current game moment.

[0256] Optionally, before the step of inputting the updated game information into the first model, the following operation is also performed:

[0257] Determine whether the information content of the updated game information set meets the second information interaction condition, wherein the second information interaction condition includes that the information content of the updated game information set is not greater than a preset information content threshold;

[0258] The step of inputting the updated game information set into the first model includes:

[0259] If the updated game information set does not exceed the preset information threshold, the updated game information set is input into the first model.

[0260] Optionally, the following operations may also be performed:

[0261] In response to the updated game information set having more information than the preset information threshold, a preset first algorithm is used to convert the updated game information set into a third game information set, and the third game information set is input into the first model, wherein the third game information set has less information than the updated game information set.

[0262] Optionally, the updated game information set includes language information and operation information;

[0263] The step of using a preset first algorithm to convert the updated game information set into a third game information set includes:

[0264] Using the first algorithm, based on the game time corresponding to each game information in the updated game information set, the language information corresponding to the second historical game time in the updated game information set is deleted to form the third game information set, wherein the second historical game time includes multiple game times that are far from the current game time in the historical game time.

[0265] Optionally, the first model is deployed on a third server; before the step of inputting the updated game information set into the first model and obtaining the first feedback information output by the first model, the following operation is also performed:

[0266] The first model is invoked through the interface provided by the third server for the first model;

[0267] The interaction instruction set is input into the first model to train the first model, so that the first model can output the first feedback information that conforms to the game logic of the game according to the first instruction defined in the interaction instruction set and the updated game information set.

[0268] Optionally, before the step of receiving the first game information sent by the first server and updating the game information set using the first game information, the following operation is also performed:

[0269] In response to receiving a robot request message sent by the first server, a first number of robots are allocated to the game match, wherein the robot request message is generated by the first server in response to the number of clients matching the game match not meeting a preset number threshold, and includes the first number of clients not matched in the game match;

[0270] In response to receiving robot matching information sent by the first server, the robot matching information is input into the first model, wherein the robot matching information is generated by the first server for matching the first number of robots in the game, and includes at least the identity information of each of the first number of robots in the game.

[0271] The ninth embodiment of this application provides an electronic device. Figure 11 This is a schematic diagram of the structure of the electronic device provided in this embodiment.

[0272] like Figure 11 As shown, the electronic device provided in this embodiment includes: a memory 1101 and a processor 1102;

[0273] The memory 1101 is used to store computer instructions for executing information interaction methods;

[0274] The processor 1102 is configured to execute computer instructions stored in the memory 1101 to perform the following operations:

[0275] Receive first game information sent by the first server, and update the game information set using the first game information, wherein the first game information is the game information generated at the current game moment of the game, and the game information set is the collection of game information generated at the historical game moments of the game;

[0276] In response to satisfying the first information interaction condition, the updated game information set is input into the first model to obtain the first feedback information output by the first model, wherein the first model is a language processing model pre-trained with a preset interaction instruction set, and the interaction instruction set includes at least a first instruction for instructing the game logic of the game match.

[0277] The first feedback information is sent to the first server so that the first server broadcasts the first feedback information in the next game moment of the current game moment.

[0278] Optionally, the interaction instruction set may further include at least a second instruction for indicating the information interaction format of the game session;

[0279] Before the step of inputting the updated game information set into the first model, the following operations are also performed:

[0280] According to the interaction instruction set, the updated game information set is converted into a second game information set, wherein the game information included in the updated game information set is expressed in a first language format, and the game information included in the second game information set is expressed in a second language format, and the second language format is the information interaction format indicated by the second instruction;

[0281] The step of inputting the updated game information set into the first model and obtaining the first feedback information output by the first model includes:

[0282] The second game information set is input into the first model to obtain the second feedback information output by the first model. The second feedback information is expressed in the second language format.

[0283] According to the interactive instruction set, the second feedback information is converted into the first feedback information, and the first feedback information is expressed in the first language format.

[0284] Optionally, the first information interaction condition includes the next game moment after the current game moment being the moment when the first feedback information needs to be broadcast;

[0285] After the steps of receiving the first game information sent by the first server and updating the game information set using the first game information, the following operations are also performed:

[0286] Determine whether the next game moment after the current game moment is the moment when the first feedback information needs to be broadcast;

[0287] The step of responding to the fulfillment of the first information interaction condition by inputting the updated game information set into the first model and obtaining the first feedback information output by the first model includes:

[0288] In response to the next game moment being a moment when the first feedback information needs to be broadcast, the updated game information set is input into the first model, the first feedback information output by the first model is obtained, and the updated game information set is updated using the first feedback information.

[0289] Optionally, the following operations may also be performed:

[0290] In response to the fact that the next game moment of the current game moment is not a moment when the first feedback information needs to be broadcast, the system receives the second game information sent by the first server and updates the updated game information set with the second game information, wherein the second game information is the game information generated in the next game moment of the current game moment.

[0291] Optionally, before the step of inputting the updated game information into the first model, the following operation is also performed:

[0292] Determine whether the information content of the updated game information set meets the second information interaction condition, wherein the second information interaction condition includes that the information content of the updated game information set is not greater than a preset information content threshold;

[0293] The step of inputting the updated game information set into the first model includes:

[0294] If the updated game information set does not exceed the preset information threshold, the updated game information set is input into the first model.

[0295] Optionally, the following operations may also be performed:

[0296] In response to the updated game information set having more information than the preset information threshold, a preset first algorithm is used to convert the updated game information set into a third game information set, and the third game information set is input into the first model, wherein the third game information set has less information than the updated game information set.

[0297] Optionally, the updated game information set includes language information and operation information;

[0298] The step of using a preset first algorithm to convert the updated game information set into a third game information set includes:

[0299] Using the first algorithm, based on the game time corresponding to each game information in the updated game information set, the language information corresponding to the second historical game time in the updated game information set is deleted to form the third game information set, wherein the second historical game time includes multiple game times that are far from the current game time in the historical game time.

[0300] Optionally, the first model is deployed on a third server; before the step of inputting the updated game information set into the first model and obtaining the first feedback information output by the first model, the following operation is also performed:

[0301] The first model is invoked through the interface provided by the third server for the first model;

[0302] The interaction instruction set is input into the first model to train the first model, so that the first model can output the first feedback information that conforms to the game logic of the game according to the first instruction defined in the interaction instruction set and the updated game information set.

[0303] Optionally, before the step of receiving the first game information sent by the first server and updating the game information set using the first game information, the following operation is also performed:

[0304] In response to receiving a robot request message sent by the first server, a first number of robots are allocated to the game match, wherein the robot request message is generated by the first server in response to the number of clients matching the game match not meeting a preset number threshold, and includes the first number of clients not matched in the game match;

[0305] In response to receiving robot matching information sent by the first server, the robot matching information is input into the first model, wherein the robot matching information is generated by the first server for matching the first number of robots in the game, and includes at least the identity information of each of the first number of robots in the game.

[0306] Alternatively, perform the following operation:

[0307] In response to receiving first game information sent by a first client, the first game information is sent to a second server so that the second server updates the game information set using the first game information. The first client is any one of at least one clients matched in the game match. The first game information is generated by the first client detecting an input event related to the first game information and is the game information generated at the current game moment of the game match. The game information set is a collection of game information generated at historical game moments of the game match.

[0308] The server receives first feedback information sent by the second server. The first feedback information is obtained by the second server inputting the updated game information set into the first model. The first model is a language processing model pre-trained with a preset set of interaction instructions. The set of interaction instructions includes at least a first instruction for instructing the game logic of the game match.

[0309] The first feedback information is sent to the second client in the next game moment of the current game moment, so that the second client displays the first feedback information on the screen, wherein the second client includes each of the at least one client matched in the game match.

[0310] Optionally, before the step of sending the first game information to the second server in response to receiving the first game information sent by the first client, the following operation is also performed:

[0311] In response to receiving a game start instruction sent by a third client, the system matches the third client with the game, wherein the game includes a preset number threshold for indicating the start conditions of the game, and the start conditions include the number of clients required to be matched in the game, wherein the third client is any client that is to participate in the game.

[0312] In response to the fact that the number of clients matched in the game match does not meet the preset number threshold, a robot request message is sent to the second server so that the second server allocates a first number of robots to the game match. The robot request message is generated by the first server in response to the fact that the number of clients matched in the game match does not meet the preset number threshold, and includes the first number of clients not matched in the game match.

[0313] The game match is matched for the first number of robots, and the robot matching information is sent to the second server so that the second server inputs the robot matching information into the first model, wherein the robot matching information includes at least the identity information of each of the first number of robots in the game match.

[0314] The tenth embodiment of this application provides a computer-readable storage medium, which includes computer instructions that, when executed by a processor, are used to implement the methods described in the embodiments of this application.

[0315] It should be noted that the relational terms such as "first" and "second" used in this document are only used to distinguish one entity or operation from another, and do not require or imply any actual relationship or order between these entities or operations. Furthermore, "including," "having," "containing," and other similar terms are synonymous, and the conclusion of any one or more items following any of the foregoing words is open-ended; none of the foregoing terms indicates that the one or more items have been exhaustively listed, or are limited to only one or more of the listed items.

[0316] When used herein, unless otherwise expressly stated, the term "or" includes all possible combinations except those that are impractical. For example, if expressed as a database may include A or B, then unless otherwise specified or impractical, it may include database A, or B, or A and B. As a second example, if expressed as a database may include A, B, or C, then unless otherwise specified or impractical, the database may include database A, or B, or C, or A and B, or A and C, or B and C, or A and B and C.

[0317] It is worth noting that the above embodiments can be implemented by hardware or software (program code), or a combination of hardware and software. If implemented by software, it can be stored in the above-described computer-readable medium. When executed by a processor, the software can perform the methods disclosed above. The computing units and other functional units described in this disclosure can be implemented by hardware or software, or a combination of hardware and software. Those skilled in the art will also understand that the above-described multiple modules / units can be combined into one module / unit, and each of the above-described modules / units can be further divided into multiple sub-modules / sub-units.

[0318] In the foregoing detailed description, embodiments have been described with reference to numerous specific details, which may vary depending on the implementation. Certain adaptations and modifications can be made to the embodiments. Other implementations will be readily apparent to those skilled in the art from the specific embodiments disclosed herein. This specification and examples are for illustrative purposes only, and the true scope and essence of this application are defined by the claims. The sequence of steps shown in the figures is also for illustrative purposes only and is not intended to limit to any particular step or order. Therefore, those skilled in the art will recognize that these steps can be performed in different orders when implementing the same method.

[0319] Exemplary embodiments are disclosed in the figures and detailed description of this application. However, many variations and modifications can be made to these embodiments. Accordingly, although specific terms are used, they are only general and descriptive and not for limiting purposes.

Claims

1. An information exchange method, characterized in that, The method includes: Receive first game information sent by the first server, and update the game information set using the first game information, wherein the first game information is the game information generated at the current game moment of the game, and the game information set is the collection of game information generated at the historical game moments of the game; In response to satisfying the first information interaction condition, the updated game information set is input into the first model to obtain the first feedback information output by the first model, wherein the first model is a language processing model pre-trained with a preset interaction instruction set, and the interaction instruction set includes at least a first instruction for instructing the game logic of the game match. The first feedback information is sent to the first server so that the first server broadcasts the first feedback information in the next game moment of the current game moment; The first information interaction condition includes the next game time after the current game time being the time when the first feedback information needs to be broadcast; the method further includes: In response to the fact that the next game moment of the current game moment is not a moment when the first feedback information needs to be broadcast, the system receives the second game information sent by the first server and updates the updated game information set with the second game information, wherein the second game information is the game information generated in the next game moment of the current game moment; Before the step of inputting the updated game information into the first model, the method further includes: determining whether the information content of the updated game information set satisfies a second information interaction condition, wherein the second information interaction condition includes that the information content of the updated game information set is not greater than a preset information content threshold; the step of inputting the updated game information set into the first model includes: In response to the updated game information set having an information content not exceeding the preset information content threshold, the updated game information set is input into the first model; the method further includes: In response to the updated game information set having more information than the preset information threshold, a preset first algorithm is used to convert the updated game information set into a third game information set, and the third game information set is input into the first model, wherein the third game information set has less information than the updated game information set.

2. The information interaction method according to claim 1, characterized in that, The interactive instruction set also includes at least a second instruction for instructing the information interaction format of the game session; Before the step of inputting the updated game information set into the first model, the method further includes: According to the interaction instruction set, the updated game information set is converted into a second game information set, wherein the game information included in the updated game information set is expressed in a first language format, and the game information included in the second game information set is expressed in a second language format, and the second language format is the information interaction format indicated by the second instruction; The step of inputting the updated game information set into the first model and obtaining the first feedback information output by the first model includes: The second game information set is input into the first model to obtain the second feedback information output by the first model. The second feedback information is expressed in the second language format. According to the interactive instruction set, the second feedback information is converted into the first feedback information, and the first feedback information is expressed in the first language format.

3. The information interaction method according to claim 1, characterized in that, After the steps of receiving the first game information sent by the first server and updating the game information set using the first game information, the method further includes: Determine whether the next game moment after the current game moment is the moment when the first feedback information needs to be broadcast; The step of responding to the fulfillment of the first information interaction condition by inputting the updated game information set into the first model and obtaining the first feedback information output by the first model includes: In response to the next game moment being a moment when the first feedback information needs to be broadcast, the updated game information set is input into the first model, the first feedback information output by the first model is obtained, and the updated game information set is updated using the first feedback information.

4. The information interaction method according to claim 1, characterized in that, The updated game information set includes language information and operation information; The step of using a preset first algorithm to convert the updated game information set into a third game information set includes: Using the first algorithm, based on the game time corresponding to each game information in the updated game information set, the language information corresponding to the second historical game time in the updated game information set is deleted to form the third game information set, wherein the second historical game time includes multiple game times that are far from the current game time in the historical game time.

5. The information interaction method according to claim 1, characterized in that, The first model is deployed on a third server; prior to the step of inputting the updated game information set into the first model and obtaining the first feedback information output by the first model, the method further includes: The first model is invoked through the interface provided by the third server for the first model; The interaction instruction set is input into the first model to train the first model, so that the first model can output the first feedback information that conforms to the game logic of the game according to the first instruction defined in the interaction instruction set and the updated game information set.

6. The information interaction method according to claim 1, characterized in that, Before the step of receiving the first game information sent by the first server and updating the game information set using the first game information, the method further includes: In response to receiving a robot request message sent by the first server, a first number of robots are allocated to the game match, wherein the robot request message is generated by the first server in response to the number of clients matching the game match not meeting a preset number threshold, and includes the first number of clients not matched in the game match; In response to receiving robot matching information sent by the first server, the robot matching information is input into the first model, wherein the robot matching information is generated by the first server for matching the first number of robots in the game, and includes at least the identity information of each of the first number of robots in the game.

7. An information exchange method, characterized in that, The method includes: In response to receiving first game information sent by a first client, the first game information is sent to a second server so that the second server executes the information interaction method as described in any one of claims 1-6, wherein the first client is any one of at least one client matched in the game match, and the first game information is generated by the first client detecting an input event for the first game information, and is the game information generated at the current game moment of the game match; The server receives first feedback information sent by the second server. The first feedback information is obtained by the second server inputting the updated game information set into the first model. The first model is a language processing model pre-trained with a preset set of interaction instructions. The set of interaction instructions includes at least a first instruction for instructing the game logic of the game match. The first feedback information is sent to the second client in the next game moment of the current game moment, so that the second client displays the first feedback information on the screen, wherein the second client includes each of the at least one client matched in the game match.

8. The information interaction method according to claim 7, characterized in that, Before the step of sending the first game information to the second server in response to receiving the first game information sent by the first client, the method further includes: In response to receiving a game start instruction sent by a third client, the system matches the third client with the game, wherein the game includes a preset number threshold for indicating the start conditions of the game, and the start conditions include the number of clients required to be matched in the game, wherein the third client is any client that is to participate in the game. In response to the fact that the number of clients matched in the game match does not meet the preset number threshold, a robot request message is sent to the second server so that the second server allocates a first number of robots to the game match. The robot request message is generated by the first server in response to the fact that the number of clients matched in the game match does not meet the preset number threshold, and includes the first number of clients not matched in the game match. The game match is matched for the first number of robots, and the robot matching information is sent to the second server so that the second server inputs the robot matching information into the first model, wherein the robot matching information includes at least the identity information of each of the first number of robots in the game match.

9. An information interaction system, characterized in that, The information interaction system includes at least one client, a first server, a second server, and a first model; The at least one client detects an input event for the first game information and sends the first game information to the first server. The first game information is the game information generated at the current game moment of the game. The first server receives the first game information sent by the at least one client and sends the first game information to the second server. The second server receives the first game information sent by the first server and updates the game information set using the first game information. The game information set is a collection of game information generated from historical game moments of the game match. The second server, in response to satisfying the first information interaction condition, inputs the updated game information set into the first model, wherein the first model is a language processing model pre-trained with a preset interaction instruction set, and the interaction instruction set includes at least a first instruction for instructing the game logic of the game match. The first model calculates and obtains first feedback information based on the updated game information set, and sends the first feedback information to the second server. The second server receives the first feedback information and sends the first feedback information back to the first server. The first server receives the first feedback information sent by the second server and sends the first feedback information to the at least one client in the next game moment of the current game moment; The at least one client displays the first feedback information on the screen; The first information interaction condition includes the next game time after the current game time being the time when the first feedback information needs to be broadcast; the second server, In response to the fact that the next game moment of the current game moment is not a moment when the first feedback information needs to be broadcast, the system receives the second game information sent by the first server and updates the updated game information set with the second game information, wherein the second game information is the game information generated in the next game moment of the current game moment; Before the step of inputting the updated game information into the first model, the second server determines whether the information content of the updated game information set meets the second information interaction condition, which includes that the information content of the updated game information set is not greater than a preset information content threshold. In response to the information content of the updated game information set not being greater than the preset information content threshold, the server inputs the updated game information set into the first model. In response to the information content of the updated game information set being greater than the preset information content threshold, the server uses a preset first algorithm to convert the updated game information set into a third game information set, and inputs the third game information set into the first model, wherein the information content of the third game information set is less than the information content of the updated game information set.

10. A game service system, characterized in that, The game service system includes a first server and a second server. The first server is connected to at least one client, and the second server is connected to a first model. The first model is a language processing model pre-trained with a preset set of interactive instructions. The set of interactive instructions includes at least a first instruction for instructing the game logic of the game. The first server is configured to, in response to receiving first game information sent by a first client, send the first game information to a second server, wherein the first client is any one of the at least one clients matched in the game match, and the first game information is generated by the first client detecting an input event related to the first game information, and is game information generated at the current game moment of the game match; receive first feedback information sent by the second server; and send the first feedback information to the second client at the next game moment of the current game moment, so that the second client displays the first feedback information on the screen, wherein the second client includes each of the at least one clients matched in the game match; The second server is configured to receive the first game information sent by the first server, and update the game information set using the first game information, wherein the game information set is a collection of game information generated from historical game moments of the game match; in response to satisfying a first information interaction condition, input the updated game information set into a first model to obtain first feedback information output by the first model; and send the first feedback information to the first server; wherein the first information interaction condition includes the next game moment of the current game moment being a moment when the first feedback information needs to be broadcast; the second server is further configured to receive the second game information sent by the first server in response to the next game moment of the current game moment not being a moment when the first feedback information needs to be broadcast, and update the updated game information set using the second game information, wherein the second game information is the... The game information generated in the next game moment of the current game moment; before the step of inputting the updated game information into the first model, the second server is further configured to: determine whether the information content of the updated game information set meets the second information interaction condition, the second information interaction condition including that the information content of the updated game information set is not greater than a preset information content threshold; in response to the information content of the updated game information set not being greater than the preset information content threshold, input the updated game information set into the first model; in response to the information content of the updated game information set being greater than the preset information content threshold, use a preset first algorithm to convert the updated game information set into a third game information set, and input the third game information set into the first model, wherein the information content of the third game information set is less than the information content of the updated game information set.

11. The game service system according to claim 10, characterized in that, The first model is deployed on the second server.

12. The game service system according to claim 10, characterized in that, The first model is deployed on a third server, and the second server connects to the first model through an interface provided by the third server for the first model.

13. An information interaction device, characterized in that, The device includes: a game information receiving unit, a feedback information acquisition unit, and a feedback information sending unit; The game information receiving unit is used to receive first game information sent by the first server and update the game information set using the first game information, wherein the first game information is the game information generated at the current game moment of the game, and the game information set is a collection of game information generated at the historical game moments of the game. The feedback information acquisition unit is used to respond to the fulfillment of the first information interaction condition by inputting the updated game information set into the first model and obtaining the first feedback information output by the first model. The first model is a language processing model pre-trained with a preset interaction instruction set, and the interaction instruction set includes at least a first instruction for instructing the game logic of the game match. The feedback information sending unit is used to send the first feedback information to the first server so that the first server broadcasts the first feedback information in the next game moment of the current game moment. The first information interaction condition includes the next game time after the current game time being the time when the first feedback information needs to be broadcast; it is also used for: In response to the fact that the next game moment of the current game moment is not a moment when the first feedback information needs to be broadcast, the system receives the second game information sent by the first server and updates the updated game information set with the second game information, wherein the second game information is the game information generated in the next game moment of the current game moment; Before the step of inputting the updated game information into the first model, the method further comprises: determining whether the information content of the updated game information set satisfies a second information interaction condition, wherein the second information interaction condition includes that the information content of the updated game information set is not greater than a preset information content threshold; in response to the information content of the updated game information set not being greater than the preset information content threshold, inputting the updated game information set into the first model; in response to the information content of the updated game information set being greater than the preset information content threshold, using a preset first algorithm to convert the updated game information set into a third game information set, and inputting the third game information set into the first model, wherein the information content of the third game information set is less than the information content of the updated game information set.

14. An information interaction device, characterized in that, The device includes: a game information sending unit, a feedback information receiving unit, and a feedback information sending unit; The game information sending unit is configured to, in response to receiving first game information sent by a first client, send the first game information to a second server, so that the second server executes the information interaction method as described in any one of claims 1-6, wherein the first client is any one of at least one client matched in the game match, and the first game information is generated by the first client detecting an input event for the first game information, and is the game information generated at the current game moment of the game match; The feedback information receiving unit is used to receive first feedback information sent by the second server. The first feedback information is obtained by the second server inputting the updated game information set into the first model. The first model is a language processing model pre-trained through a preset interaction instruction set. The interaction instruction set includes at least a first instruction for instructing the game logic of the game match. The feedback information sending unit is used to send the first feedback information to the second client in the next game moment of the current game moment, so that the second client displays the first feedback information on the screen, wherein the second client includes each of the at least one client matched in the game match.

15. A gaming robot, characterized in that, include: Information storage module, information processing module; The information storage module is used to store game information generated during game matches; The information processing module is used to perform the method as described in any one of claims 1-6.

16. An electronic device, characterized in that, include: Memory, processor; The memory is used to store one or more computer instructions; The processor is configured to execute one or more computer instructions to implement the method as described in any one of claims 1-6 or 7-8.

17. A computer-readable storage medium storing one or more computer instructions thereon, characterized in that, When this instruction is executed by the processor, it performs the method as described in any one of claims 1-6 or 7-8.

Images