Task processing method, device and system

Abstract

The embodiment of the invention discloses a task processing method, device and system. In the embodiment of the invention, user question information is obtained; a large language model is adopted to split the user question, a target workflow is generated, and the target workflow comprises at least one subtask; receiving a task execution instruction; and sequentially executing the sub-tasks according to the sequence of the sub-tasks in the target workflow. By means of the method, the tasks can be accurately disassembled and effectively executed, the research and development cost is reduced, and manpower consumption is reduced.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and more particularly, to a method, apparatus, and system for task processing. Background Art

[0002] During the software development process, there are a large number of tasks with high technical difficulty and cumbersome operations. For example, complex but repetitive tasks such as Java Development Kit (JDK) upgrade, repository-level comment generation, unit test generation, and framework migration. These tasks do not involve specific business, but they require a large amount of time and effort from R & D personnel. R & D personnel generally lack motivation for these highly repetitive tasks, and a large number of manual operations are also prone to errors, affecting the overall efficiency and system reliability.

[0003] In the prior art, a Planner Agent and LangGraph are used to process the above tasks. Among them, the Planner Agent disassembles complex tasks into multiple subtasks by optimizing the planning and task execution processes, and then sequentially or batch-executes multiple subtasks. However, the disassembly of the above tasks may be incorrect or the granularity may not be fine enough; the LangGraph, through the form of manual coding, designs a state transition flowchart for the model in advance, allowing the model to transfer between several states until the task is completed. However, it requires R & D personnel to code and design the entire process, with a large amount of design content and high development costs; and during the task execution process of the Planner Agent and LangGraph, R & D personnel cannot provide timely feedback and adjustment on the execution situation.

[0004] In summary, how to accurately disassemble and effectively execute tasks, reduce R & D costs, and reduce manpower consumption are problems that need to be solved currently. Summary of the Invention

[0005] In view of this, embodiments of the present invention provide a method, apparatus, and system for task processing, which can accurately disassemble and effectively execute tasks, reduce R & D costs, and reduce manpower consumption.

[0006] In a first aspect, an embodiment of the present invention provides a method for task processing, the method comprising:

[0007] Obtain user problem information;

[0008] Use a large language model to split the user problem to generate a target workflow, where the target workflow includes at least one subtask;

[0009] Receive a task execution instruction;

[0010] Execute the subtasks in sequence according to the order of the subtasks in the target workflow.

[0011] Optionally, splitting the user problem by using a large language model to generate a target workflow specifically includes:

[0012] Split the user problem by using a large language model to generate a candidate workflow, where the candidate workflow includes at least one subtask;

[0013] Send the candidate workflow;

[0014] Receive an adjustment instruction, where the adjustment instruction is used to modify at least one generated subtask;

[0015] Generate the target workflow.

[0016] Optionally, receiving a task execution instruction specifically includes:

[0017] Receive a task execution instruction sent by a user terminal through a communication layer.

[0018] Optionally, the method further includes:

[0019] Obtain the execution result of each subtask.

[0020] Optionally, the method further includes:

[0021] Send the execution result.

[0022] Optionally, the method further includes:

[0023] In response to an exception occurring during the execution of the subtask, send an exception message.

[0024] In a second aspect, an embodiment of the present invention provides a task processing device, where the device includes:

[0025] An acquisition unit for acquiring user problem information;

[0026] A splitting unit for splitting the user problem by using a large language model to generate a target workflow, where the target workflow includes at least one subtask;

[0027] A receiving unit for receiving a task execution instruction;

[0028] A processing unit for executing the subtasks in sequence according to the order of the subtasks in the target workflow.

[0029] Optionally, the splitting unit is specifically used for:

[0030] The large language model is used to split the user problem and generate a candidate workflow, where the candidate workflow includes at least one subtask;

[0031] Send the candidate workflow;

[0032] Receive an adjustment instruction, where the adjustment instruction is used to modify at least one generated subtask;

[0033] Generate the target workflow.

[0034] Optionally, the receiving unit is specifically configured to:

[0035] Receive a task execution instruction sent by a user terminal through a communication layer.

[0036] Optionally, the receiving unit is further configured to:

[0037] Obtain the execution result of each subtask.

[0038] Optionally, the device further includes:

[0039] A sending unit, configured to send the execution result.

[0040] Optionally, the sending unit is further configured to:

[0041] In response to an exception occurring during the execution of the subtask, send an exception message.

[0042] In a third aspect, an embodiment of the present invention provides a task processing system, where the system includes:

[0043] A terminal and an agent;

[0044] Wherein, the terminal is configured to receive user problem information and send the user problem information to the agent;

[0045] The agent is configured to execute the method described in any one of the first aspect and any possible implementation of the first aspect, and send the execution result of the subtask to the terminal.

[0046] In a fourth aspect, an embodiment of the present invention provides an electronic device, including a memory and a processor, where the memory is used to store one or more computer program instructions, and wherein the one or more computer program instructions are executed by the processor to implement the method described in any one of the first aspect or any possible implementation of the first aspect.

[0047] In a fifth aspect, an embodiment of the present invention provides a computer-readable storage medium, on which computer program instructions are stored, and the computer program instructions, when executed by a processor, implement the method described in any one of the first aspect or any possible implementation of the first aspect.

[0048] In an embodiment of the present invention, user problem information is obtained; the large language model is used to split the user problem to generate a target workflow, where at least one subtask is included in the target workflow; a task execution instruction is received; and the subtasks are sequentially executed in the order of the subtasks in the target workflow. Through the above method, tasks can be accurately disassembled and effectively executed, and the R & D cost can be reduced and the labor consumption can be reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Through the following description of the embodiments of the present invention with reference to the accompanying drawings, the above and other objects, features, and advantages of the present invention will become clearer. In the drawings:

[0050] Figure 1 is a flowchart of a method of a Planner Agent in the prior art;

[0051] Figure 2 is a flowchart of a method of a LangGraph in the prior art;

[0052] Figure 3 is a schematic diagram of a device for task processing in an embodiment of the present invention;

[0053] Figure 4 is another schematic diagram of a device for task processing in an embodiment of the present invention;

[0054] Figure 5 is still another schematic diagram of a device for task processing in an embodiment of the present invention;

[0055] Figure 6 is a schematic diagram of a system for task processing in an embodiment of the present invention;

[0056] Figure 7 is a schematic diagram of a device for task processing in an embodiment of the present invention;

[0057] Figure 8 is a schematic diagram of an electronic device in an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0058] The following describes the present application based on embodiments, but the present application is not limited to these embodiments. In the following detailed description of the present application, some specific details are described in detail. Those skilled in the art can fully understand the present application without the description of these details. In order to avoid obscuring the essence of the present application, well-known methods, processes, procedures, elements, and circuits are not described in detail.

[0059] In addition, those of ordinary skill in the art should understand that the accompanying drawings provided herein are for illustrative purposes only, and the drawings are not necessarily drawn to scale.

[0060] Unless the context clearly requires otherwise, the words "include", "comprising" and similar words throughout the application should be interpreted as including rather than exclusive or exhaustive; that is, the meaning is "including but not limited to".

[0061] In the description of this application, it should be understood that the terms "first", "second", etc. are only used for descriptive purposes and cannot be understood as indicating or implying relative importance. In addition, in the description of this application, unless otherwise specified, the meaning of "plurality" is two or more.

[0062] In the prior art, an intelligent agent refers to an entity with certain intelligent attributes and capable of autonomous action in an information dissemination environment. It can be a software program, a robot, or a media system with intelligent interactive functions, etc. The intelligent agent can perceive environmental information, make decisions and take actions to complete tasks such as information collection, content generation, information dissemination, user interaction, etc. The agent is mainly based on general scenarios. Even the opendevin platform, its underlying design is based on the very traditional and general ReAct paradigm or its variants. The disadvantage of the above-mentioned ReAct paradigm is that it gives the large model used in the intelligent agent great freedom of autonomous decision-making. The high-freedom paradigm can easily lead to the large model being too divergent, resulting in uncontrollable execution direction, often resulting in dead loops or constantly doing useless work in the wrong direction, and it is difficult to complete specific tasks that are slightly difficult.

[0063] In order to ensure that the entire development process is within a controllable range, Planner Agent and LangGraph are used to process the above tasks. The Planner Agent breaks down complex tasks into multiple subtasks by optimizing the planning and task execution process, and then executes multiple subtasks sequentially or in batches. Figure 1 As shown in the figure, after the user puts forward the demand, the Planner Agent decomposes the task to generate a task list, each task list includes multiple subtasks, and the Planner Agent executes each subtask separately, and when the subtask processing result is not good, it will process the subtask in a loop, and then send the processing result to the Planner Agent for re-planning, regenerate more subtasks, and send the processing result to the user at the same time. However, the decomposition of the above tasks may be incorrect or the granularity may not be fine enough, and the R&D personnel of the Planner Agent cannot provide timely feedback and adjustment on the execution status during the task execution; the above LangGraph is a framework for building and managing Agents. Through human coding, a flowchart of state transition is designed for the model in advance, so that the model can transfer in several states until the task is completed; for example,Figure 2 As shown in the figure, the Chatbot is used to interact with users under the LangGraph framework. Among them, the Chatbot is an intelligent dialogue system built based on natural language processing technology and machine learning algorithms. In the form of artificial coding, a flowchart for state transition is designed for the Chatbot in advance, that is Figure 2 the process of a person interacting with the Chatbot in. Then, the Chatbot calls tools according to the process to process the task until the processing is completed. The above method requires developers to code and design the entire process, with a large amount of design content and high development costs. Moreover, during the task execution process of the above LangGraph, developers cannot provide timely feedback and adjustment on the execution situation.

[0064] Therefore, how to accurately disassemble and effectively execute tasks, reduce R & D costs, and reduce manpower consumption are the problems that need to be solved currently.

[0065] In the embodiments of the present invention, the large language model (LLM) can also be referred to as a large model, or an artificial intelligence (AI) model. Among them, the large language model is a deep learning model based on the transformer architecture, capable of processing and generating natural language text. It is usually trained on a large amount of text data and has the ability to understand and generate language, and is widely used in dialogue systems, text generation, and other natural language processing tasks.

[0066] In the embodiments of the present invention, to solve the above problems, a method for task processing is proposed, specifically as Figure 3 shown, the method includes:

[0067] Step S301, obtain user question information.

[0068] In a possible implementation manner, the user inputs user question information on the terminal interface, and the terminal sends the above user question information to the intelligent agent Agent, specifically, to the large language model in the Agent. The terminal can also be referred to as a user terminal, the user terminal side, or a user terminal device, etc. The embodiments of the present invention do not make any limitations on it.

[0069] Step S302, use the large language model to split the user question and generate a target workflow.

[0070] Specifically, the target workflow includes at least one subtask.

[0071] In a possible implementation, a large language model is used to split the user question to generate a candidate workflow, where the candidate workflow includes at least one subtask; the candidate workflow is sent; an adjustment instruction is received, where the adjustment instruction is used to modify at least one generated subtask; the target workflow is generated.

[0072] For example, after obtaining user question information, the large language model in the Agent intelligently disassembles the user question information to generate a candidate workflow including multiple subtasks, and sends the candidate workflow to the terminal. After receiving the candidate workflow, the terminal displays it to the user on the terminal interface. At this time, if the user believes that the granularity of the task disassembly is not fine enough to meet the requirements of complex tasks, an adjustment instruction is sent to the large language model in the Agent to disassemble the user question information again to ensure that the task disassembly granularity is appropriate. The large language model in the Agent generates a new and more refined workflow, that is, the target workflow, based on the feedback and suggestions included in the adjustment instruction sent by the user, thereby ensuring the accuracy and feasibility of the task disassembly.

[0073] In a possible implementation, the above-mentioned large language model and the user can conduct multiple rounds of interaction to finally determine the target workflow.

[0074] In a possible implementation, after determining the target workflow, the Agent saves the target workflow.

[0075] Step S303: Receive a task execution instruction.

[0076] Specifically, a task execution instruction sent by the user terminal is received through the communication layer (Channel).

[0077] In a possible implementation, the user terminal sends the task execution instruction to the Agent through the communication layer.

[0078] Step S304: Execute the subtasks in sequence according to the order of the subtasks in the target workflow.

[0079] Specifically, the Agent includes a large language model and a state management module. When executing the subtasks of the target workflow, the large language model and the state management module are required to participate together, and the Agent enters a loop state to execute each subtask.

[0080] In a possible implementation, the subtasks in the target workflow can be executed serially or in parallel, which is specifically determined according to the actual situation.

[0081] In the embodiments of the present invention, the above method is used to generate a target workflow through a large language model, that is, through intelligent task decomposition and dynamic adjustment of the execution plan, the high requirements for R & D personnel in coding design and abstraction can be reduced, the efficiency of task execution can be significantly improved, repetitive labor and human errors can be reduced, and moreover, the target workflow can be optimized through user feedback and suggestions.

[0082] In a possible implementation manner, after step S304, the following steps are further included, specifically proposing a method for task processing, specifically as Figure 4 shown, specifically including:

[0083] Step S305, obtain the execution result of each of the sub-tasks.

[0084] Step S306, send the execution result.

[0085] Specifically, the Agent sends the execution result to the user terminal through the communication layer.

[0086] In a possible implementation manner, after step S304, the following steps are further included, specifically proposing a method for task processing, specifically as Figure 5 shown, specifically including:

[0087] Step S307, in response to an exception occurring during the execution of the sub-task, send exception information.

[0088] Specifically, during the execution of each sub-task, the Agent will perform real-time monitoring. If an exception occurs during the execution of the sub-task, the Agent sends the exception information to the user terminal, and the user terminal displays the exception information in the interface. The user can issue a new instruction according to the exception information to handle the exception situation. The exception situation includes timeouts or infinite loops, etc., which are specifically determined according to the actual situation.

[0089] Through the above embodiments, the flexibility and dynamic adjustment ability of task processing can be enhanced. The Agent can refer to the execution results of the sub-tasks that have been processed when processing the current sub-task, and the processing result of the current sub-task can also be used as a reference for candidate sub-tasks, realizing a closed-loop mode from observation to execution and then to observation. Moreover, by adopting a user real-time intervention mechanism, the task execution process is made more flexible and dynamic, and can be adjusted according to the actual situation to ensure the high efficiency and accuracy of task execution.

[0090] In a possible implementation, the process of the Agent executing each subtask of the target task flow can be carried out in an isolated execution environment. For example, it can be executed in a sandbox. Specifically, after the Agent transfers the subtasks to be executed to the sandbox through the communication layer for execution, the sandbox then sends the execution results to the Agent through the communication layer.

[0091] In a possible implementation, the sandbox can call pre-set plugins during operation. For example, corresponding workflows and plugins, that is, automation tools and scripts, are precipitated in specific task domains, enabling the Agent to have full-automatic or semi-automatic R & D or defect repair capabilities in the specific task domain. For example, some simple tasks can be completed by calling one or two plugins, further improving the efficiency and effect of task execution.

[0092] In the embodiments of the present invention, the Agent realizes intelligent task decomposition and interaction closed-loop, that is, decomposes complex tasks into several subtasks and provides key execution commands, and adopts a closed-loop mode of observation to execution and then observation to ensure the high efficiency and accuracy of task execution; the Agent also realizes real-time anomaly detection and handling, that is, has an anomaly detection mechanism, can identify and handle infinite loops or unsolvable problems, ensure that the task does not interrupt, and automatically seek help from the user when encountering problems; the Agent further realizes dynamic adjustment of the execution plan, that is, uses a large language model to generate and adjust the execution plan, and optimizes it in real time according to user feedback to ensure the smooth completion of the task; the Agent also realizes a user real-time intervention mechanism, that is, provides the ability for the user to intervene in the task execution process at any time, view the status and logs in real time, comprehensively master and adjust the task progress to ensure higher flexibility and accuracy; the Agent further realizes optimization in specific task domains, that is, precipitates corresponding workflows and automation tools and scripts in specific task domains, enabling the Agent to have full-automatic / semi-automatic R & D or defect repair capabilities in this domain, and improving the efficiency and effect of task execution.

[0093] In the embodiments of the present invention, the defects of the traditional paradigm are avoided through the above method. Aiming at the problems that the traditional ReAct paradigm is prone to divergence and infinite loops under high degrees of freedom, the Agent in the embodiments of the present invention restricts the disorderly divergence of the model through a customized execution plan and human-computer interaction to ensure the controllability of the task execution direction. From requirement confirmation to execution plan generation, task execution, real-time adjustment, to final acceptance, it ensures that the task processing system is close to the actual work process, improves the accuracy and efficiency of task completion; and can provide an isolated execution environment and basic tools to execute R & D tasks safely and efficiently, ensuring that task operations will not affect other system parts.

[0094] In an embodiment of the present invention, a task processing system is proposed. Specifically, as Figure 6 shown, the system includes: a terminal 601 and an agent 602; wherein, the terminal is configured to receive user question information and send the user question information to the agent; the agent is configured to obtain the user question information, split the user question by using a large language model to generate a target workflow, where the target workflow includes at least one subtask, receive a task execution instruction, and sequentially execute the subtasks in the order of the subtasks in the target workflow.

[0095] In an embodiment of the present invention, a task processing apparatus is provided. As Figure 7 shown, it specifically includes: an acquisition unit 701, a splitting unit 702, a receiving unit 703, and a processing unit 704;

[0096] wherein, the acquisition unit 701 is configured to obtain user question information; the splitting unit 702 is configured to split the user question by using a large language model to generate a target workflow, where the target workflow includes at least one subtask; the receiving unit 703 is configured to receive a task execution instruction; and the processing unit 704 is configured to sequentially execute the subtasks in the order of the subtasks in the target workflow.

[0097] Further, the splitting unit is specifically configured to:

[0098] split the user question by using a large language model to generate a candidate workflow, where the candidate workflow includes at least one subtask;

[0099] send the candidate workflow;

[0100] receive an adjustment instruction, where the adjustment instruction is used to modify at least one generated subtask;

[0101] generate the target workflow.

[0102] Further, the receiving unit is specifically configured to:

[0103] receive a task execution instruction sent by a user terminal through a communication layer.

[0104] Further, the receiving unit is further configured to:

[0105] obtain the execution result of each subtask.

[0106] Further, the apparatus further includes:

[0107] a sending unit configured to send the execution result.

[0108] Further, the sending unit is further configured to:

[0109] In response to an exception occurring during the execution of the subtask, send an exception message.

[0110] Figure 8 It is a schematic structural diagram of the electronic device in the embodiment of the present invention. As Figure 8 shown, it includes a general computer hardware structure, which at least includes a processor 801 and a memory 802. The processor 801 and the memory 802 are connected through a bus 803. The memory 802 is adapted to store instructions or programs executable by the processor 801. The processor 801 can be an independent microprocessor or a set of one or more microprocessors. Thus, the processor 801 executes the instructions stored in the memory 802, thereby executing the method flow of the embodiment of the present invention as described above to implement the processing of data and the control of other devices. The bus 803 connects the above-mentioned multiple components together, and at the same time connects the above-mentioned components to a display controller 804, a display device, and an input / output (I / O) device 805. The input / output (I / O) device 805 can be a mouse, a keyboard, a modem, a network interface, a touch input device, a somatosensory input device, a printer, and other devices well known in the art. Typically, the input / output (I / O) device 805 is connected to the system through an input / output (I / O) controller 806.

[0111] Among them, the instructions stored in the memory 802 are executed by at least one processor 801 to implement: obtaining user question information; splitting the user question by using a large language model to generate a target workflow, where the target workflow includes at least one subtask; receiving a task execution instruction; and sequentially executing the subtasks in the order of the subtasks in the target workflow.

[0112] Specifically, the electronic device includes: one or more processors 801 and a memory 802, Figure 8 Taking one processor 801 as an example. The processor 801 and the memory 802 can be connected through a bus or other means, Figure 8 taking the connection through a bus as an example. The memory 802, as a non-volatile computer-readable storage medium, can be used to store non-volatile software programs, non-volatile computer-executable programs, and modules. The processor 801 executes various functional applications and data processing of the device by running the non-volatile software programs, instructions, and modules stored in the memory 802, that is, implementing the method for determining task processing as described above.

[0113] The memory 802 may include a program storage area and a data storage area. The program storage area may store an operating system and application programs required for at least one function. The data storage area may store an option list and the like. In addition, the memory 802 may include high-speed random access memory and may also include non-volatile memory, such as at least one magnetic disk storage device, a flash memory device, or other non-volatile solid-state storage devices. In some embodiments, the memory 802 may optionally include a memory remotely disposed relative to the processor 801, and these remote memories may be connected to an external device through a network. Examples of the above networks include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0114] One or more modules are stored in the memory 802 and, when executed by one or more processors 801, perform the method for task processing in any of the above method embodiments.

[0115] As those skilled in the art will realize, various aspects of the embodiments of the present invention can be implemented as a system, a method, or a computer program product. Therefore, various aspects of the embodiments of the present invention can take the following forms: a complete hardware implementation, a complete software implementation (including firmware, resident software, microcode, etc.), or an implementation that combines software aspects with hardware aspects, which is generally referred to herein as a "circuit", "module", or "system". In addition, various aspects of the embodiments of the present invention can take the following forms: a computer program product implemented in one or more computer-readable media, the computer-readable media having computer-readable program code implemented thereon.

[0116] Any combination of one or more computer-readable media may be utilized. The computer-readable media may be a computer-readable signal medium or a computer-readable storage medium. The computer-readable storage medium may be, for example (but not limited to), an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples (non-exhaustive) of the computer-readable storage medium will include the following: an electrical connection having one or more wires, a portable computer floppy disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the context of the embodiments of the present invention, the computer-readable storage medium may be any tangible medium that can contain or store a program used by or in conjunction with an instruction execution system, apparatus, or device.

[0117] A computer-readable signal medium may include a propagated digital signal having computer-readable program code embodied therein, either as in a baseband or as part of a carrier wave. Such a propagated signal may take any of a variety of forms, including but not limited to: electromagnetic, optical, or any suitable combination thereof. A computer-readable signal medium may be any computer-readable medium that is not a computer-readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device.

[0118] Any suitable medium may be used to transmit the program code embodied on a computer-readable medium, including but not limited to wireless, wired, fiber optic cable, RF, etc., or any suitable combination of the foregoing.

[0119] The computer program code for performing operations in connection with aspects of the embodiments of the present invention may be written in any combination of one or more programming languages, including: object-oriented programming languages such as Java, Smalltalk, C++; and conventional procedural programming languages such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computer as a stand-alone software package, partly on the user's computer and partly on a remote computer, or entirely on the remote computer or server. In the latter case, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or the connection may be made to an external computer (e.g., through the Internet using an Internet service provider).

[0120] The flowcharts and / or block diagrams of the methods, apparatus (systems), and computer program products according to embodiments of the present invention described above depict various aspects of the embodiments of the present invention. It will be understood that each block of the flowcharts and / or block diagrams, and combinations of blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions / acts specified in the flowchart and / or block diagram block or blocks.

[0121] These computer program instructions may also be stored in a computer-readable medium that can direct a computer, other programmable data processing apparatus, or other devices to function in a particular manner, such that the instructions stored in the computer-readable medium produce an article of manufacture including instructions for implementing the functions / acts specified in the flowchart and / or block diagram block or blocks.

[0122] Computer program instructions can also be loaded onto a computer, other programmable data processing apparatus, or other devices, so as to cause a series of operational steps to be performed on the computer, other programmable apparatus, or other devices to generate a computer-implemented process, such that the instructions executed on the computer or other programmable apparatus provide a process for implementing the functions / actions specified in the flowchart and / or block diagram block or blocks.

[0123] The foregoing are only preferred embodiments of the present application and are not intended to limit the present application. For those skilled in the art, various modifications and variations can be made to the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present application shall be included within the protection scope of the present application.

[0124] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in the present application are all information and data authorized by the user or fully authorized by all parties. And the collection, use, and processing of relevant data need to comply with the relevant laws, regulations, and standards of the relevant countries and regions, and corresponding operation entrances are provided for users to choose to authorize or refuse. When a user refuses to process personal information other than the necessary information required for the basic functions, it will not affect the user's use of the basic functions.

Claims

1. A task processing method, characterized in that: The method comprises: Get user question information; Using a large language model to split the user question to generate a target workflow, wherein the target workflow includes at least one subtask; Receive a task execution instruction; The subtasks are executed sequentially according to the order of the subtasks in the target workflow.

2. The method according to claim 1, characterized in that The use of a large language model to split the user question and generate a target workflow specifically includes: Using a large language model to split the user question to generate a candidate workflow, wherein the candidate workflow includes at least one subtask; sending the candidate workflow; receiving an adjustment instruction, wherein the adjustment instruction is used to modify at least one generated subtask; The target workflow is generated.

3. The method according to claim 1, characterized in that The receiving of the task execution instruction specifically includes: The task execution instruction sent by the user terminal is received through the communication layer.

4. The method according to claim 1, characterized in that: The method further comprises: Obtain the execution result of each of the subtasks.

5. The method according to claim 4, characterized in that The method further comprises: The execution result is sent.

6. The method according to claim 1, characterized in that The method further comprises: In response to an exception occurring during the execution of the subtask, exception information is sent.

7. A task processing device, characterized in that: The device comprises: An acquisition unit, used to acquire user question information; A splitting unit, configured to split the user question using a large language model to generate a target workflow, wherein the target workflow includes at least one subtask; A receiving unit, used for receiving a task execution instruction; The processing unit is used to execute the subtasks in sequence according to the order of the subtasks in the target workflow.

8. A task processing system, characterized in that: The system comprises: Terminals and Agents; Wherein, the terminal is used to receive user question information and send the user question information to the intelligent agent; The agent is used to execute the method described in any one of claims 1 to 6, and send the execution result of the subtask to the terminal.

9. An electronic device, comprising a memory and a processor, characterized in that: The memory is used to store one or more computer program instructions, wherein the one or more computer program instructions are executed by the processor to implement the method according to any one of claims 1 to 6.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method according to any one of claims 1 to 6 is implemented.

Images