Large-model-based method and system for function selection and execution in multi-task environment

Abstract

A large-model-based method and system for function selection and execution in a multi-task environment, relating to the technical field of artificial intelligence. The technical problem to be solved is how to understand user natural language input and automatically select and execute corresponding functions to complete complex tasks in a multi-task environment. The method comprises: for each function in a user API function library, defining attribute information of the function, and constructing a feature vector library on the basis of the function and a feature vector corresponding thereto; on the basis of the vector similarity between the feature vector of a user question and a function feature vector, searching for a function matching the user question; on the basis of the user question and the function matching same, analyzing a user need and expectation by means of a large model, formulating a task execution strategy on the basis of the user need and expectation, selecting, on the basis of the task execution strategy, a function to be executed, and constructing a function execution plan; and for the function execution plan, executing each function in sequence, and monitoring and adjusting the function currently being executed.

Description

A Method and System for Function Selection and Execution in a Multi-Task Environment Based on a Large Model

[0001] This application claims priority to Chinese Patent Application No. 202411797975.1, filed on December 9, 2024, entitled "Method and System for Function Selection and Execution in a Multi-Task Environment Based on a Large Model", the entire contents of which are incorporated herein by reference. Technical Field

[0002] This invention relates to the field of artificial intelligence technology, specifically to a method and system for function selection and execution in a multi-task environment based on a large model. Background Technology

[0003] With the increasing application of Natural Language Processing (NLP) and Artificial Intelligence (AI) technologies across various fields, improving the intelligence level of systems, reducing human intervention, and enhancing user experience have become key aspects of technological development when dealing with complex and ever-changing task environments.

[0004] In traditional API function library usage, users often need a certain level of technical background to effectively call and configure functions to meet specific business needs. In this case, users face problems such as complex operation, high learning costs, and low efficiency. Furthermore, with the continuous expansion and updates of API function libraries, how to efficiently manage and utilize these libraries has become a significant challenge for developers.

[0005] In a multitasking environment, understanding user natural language input, automatically selecting and executing corresponding functions to complete complex tasks is a technical problem that needs to be solved. Summary of the Invention

[0006] The technical objective of this invention is to address the above-mentioned shortcomings by providing a method and system for function selection and execution in a multi-task environment based on a large model, thereby solving the technical problem of how to understand user natural language input, automatically select and execute corresponding functions to complete complex tasks in a multi-task environment.

[0007] In a first aspect, the present invention provides a method for function selection and execution in a multi-task environment based on a large model, comprising the following steps:

[0008] Function library modeling: For each function in the user API function library, define the function's attribute information and extract the function's feature vector. Construct a feature vector library based on the function and its corresponding feature vector. The function's attribute information includes description, input fields, output fields, type, and function.

[0009] User problem analysis: For user problems, extract the feature vector of the user problem, and find the function that matches the user problem based on the vector similarity between the feature vector of the user problem and the feature vector of the function.

[0010] Function execution plan construction: Based on the user's problem and the matching function, the user's needs and expectations are analyzed through a large model, and a task execution strategy is formulated based on the user's needs and expectations. Based on the task execution strategy, the function to be executed is selected and the function execution plan is constructed.

[0011] Task Execution: For function execution plans, each function is executed sequentially, and the currently executing function is monitored and adjusted. If all functions in the function execution plan execute successfully, the successful task execution result is returned to the user. If any function in the function execution plan fails, the execution result of the failed task is returned to the user. When monitoring and adjusting the currently executing function, if the currently executing function returns a successful execution result, the next function is executed. If the currently executing function returns a failed execution result, the current executing function and its execution plan are adjusted and judged based on the execution log and the returned execution result of the currently executing function, using a large model. Based on the adjustment judgment result, the adjusted function or the adjusted function execution plan is re-executed. If the adjusted function is re-executed, the adjusted function is monitored and judged. If the adjusted function execution plan is re-executed, the current function is determined to have failed, and the execution result of the failed task is returned to the user.

[0012] Preferably, the feature vector of the function is extracted through a text embedding model, and the feature vector of the user's question is also extracted through a text embedding model.

[0013] As a preferred approach, the currently executing function is monitored and adjusted, including the following steps:

[0014] During the execution of the current function, the execution process of the current function is recorded, forming the execution log of the current function;

[0015] After the current function finishes execution, determine whether the current function execution was successful based on the execution result returned by the current function.

[0016] If the current function executes successfully, check if there are any functions to be executed after the current function. If there are, execute the next function. If not, determine that the task has been executed successfully and return the successful task execution result to the user.

[0017] If the current function fails to execute, the system adjusts and judges the currently executing function and its execution plan based on the execution log and returned execution result of the current function, using a large model. If the currently executing function is adjusted, the adjusted function is re-executed and monitored. If the function execution plan is adjusted, the system determines that the current function has failed to execute and returns the execution result of the failed task to the user. The system then performs task execution operations based on the adjusted function execution plan.

[0018] Preferably, the return value or status code returned by the function is used as the execution result. Based on the execution log of the current function and the returned execution result, the current function is analyzed by a large model to identify the log type, identify the error type, and adjust the task execution strategy. It is then determined whether to adjust the currently executing function and / or adjust the function execution plan. Adjusting the currently executing function includes redefining the function's attribute information, and adjusting the function execution plan includes selecting the function to be executed and the execution order of the functions to form a new function execution plan.

[0019] Secondly, the present invention provides a function selection and execution system based on a large model in a multi-task environment, which is used to realize the selection and execution of functions in a multi-task environment through a function selection and execution method based on a large model in a multi-task environment as described in any of the first aspects. The system includes a function library modeling module, a user problem analysis module, a function execution planning construction module, and a task execution module.

[0020] The function library modeling module is used to perform the following: For each function in the user API function library, define the function's attribute information and extract the function's feature vector. Based on the function and its corresponding feature vector, construct a feature vector library. The function's attribute information includes description, input fields, output fields, type, and function.

[0021] The user problem analysis module performs the following: For a user problem, it extracts the feature vector of the user problem, and finds a function that matches the user problem based on the vector similarity between the feature vector of the user problem and the feature vector of the function.

[0022] The function execution planning module is used to perform the following: based on the user's problem and the matching function, analyze the user's needs and expectations through a large model, formulate a task execution strategy based on the user's needs and expectations, select the function to be executed based on the task execution strategy, and construct the function execution plan;

[0023] The task execution module performs the following: For a function execution plan, each function is executed sequentially, and the currently executing function is monitored and adjusted. If all functions in the function execution plan are executed successfully, the successful task execution result is returned to the user. If any function in the function execution plan fails, the execution result of the failed task is returned to the user. When monitoring and adjusting the currently executing function, if the currently executing function returns a successful execution result, the next function is executed. If the currently executing function returns a failed execution result, the current executing function and its execution plan are adjusted and judged based on the execution log and the returned execution result of the currently executing function, using a large model. Based on the adjustment judgment result, the adjusted function or the adjusted function execution plan is re-executed. If the adjusted function is re-executed, the adjusted function is monitored and judged. If the adjusted function execution plan is re-executed, the current function is determined to have failed, and the execution result of the failed task is returned to the user.

[0024] Preferably, the function library modeling module is used to extract the feature vectors of functions through a text embedding model, and the user question analysis module is used to extract the feature vectors of user questions through a text embedding model.

[0025] Preferably, the task execution module is used to monitor and adjust the currently executing function as follows:

[0026] During the execution of the current function, the execution process of the current function is recorded, forming the execution log of the current function;

[0027] After the current function finishes execution, determine whether the current function execution was successful based on the execution result returned by the current function.

[0028] If the current function executes successfully, check if there are any functions to be executed after the current function. If there are, execute the next function. If not, determine that the task has been executed successfully and return the successful task execution result to the user.

[0029] If the current function fails to execute, the system adjusts and judges the currently executing function and its execution plan based on the execution log and returned execution result of the current function, using a large model. If the currently executing function is adjusted, the adjusted function is re-executed and monitored. If the function execution plan is adjusted, the system determines that the current function has failed to execute and returns the execution result of the failed task to the user. The system then performs task execution operations based on the adjusted function execution plan.

[0030] Preferably, the task execution module uses the return value or status code returned by the function as the execution result. Based on the execution log of the current function and the returned execution result, it performs log analysis, error type identification, and task execution strategy adjustment on the current function through a large model. It then determines whether to adjust the currently executed function and / or adjust the function execution plan. Adjusting the currently executed function includes redefining the function's attribute information, and adjusting the function execution plan includes selecting the functions to be executed and the execution order of the functions to form a new function execution plan.

[0031] The function selection and execution method and system based on a large model in a multi-task environment of the present invention have the following advantages:

[0032] 1. Improved function usage efficiency: For each function in the user API function library, the function's attribute information is defined, and the function's feature vector is extracted. A feature vector library is built based on the function and its corresponding feature vector, which facilitates understanding the function and making intelligent selection of the function;

[0033] 2. Improved user experience: By analyzing user problems through a large model, user needs and expectations are obtained. Based on these needs and expectations, task execution strategies are formulated. Based on the task execution strategies, functions to be executed are selected and function execution plans are constructed. This achieves automated and intelligent generation of function execution plans. Users do not need to have an in-depth understanding of functions to obtain function execution plans by asking questions, which meets the needs of users, especially non-professionals, and reduces operational complexity.

[0034] 3. Reduced human intervention: Automated problem-solving processes reduce the need for manual operation and decrease the frequency and extent of human intervention;

[0035] 4. Reduced labor costs: Reduced reliance on professional personnel, thereby lowering labor costs;

[0036] 5. Improved problem-solving accuracy: Driven by a large model, the system can more accurately understand and execute user intent, thus improving the accuracy of problem-solving.

[0037] 6. Enhanced flexibility: The large model can work in a multi-tasking environment, adapting to different task types and requirements, and has high flexibility;

[0038] 7. Improved automation level: The automated execution steps and error handling mechanism have improved the automation level of the entire large model system;

[0039] 8. Improved task completion rate: Through detailed planning and execution steps, the continuity and accuracy of tasks were ensured, thereby improving the task completion rate. Attached Figure Description

[0040] To more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0041] The invention will be further described below with reference to the accompanying drawings.

[0042] Figure 1 is a flowchart of a function selection and execution method based on a large model in a multi-task environment provided in Example 1. Detailed Implementation

[0043] The present invention will be further described below with reference to the accompanying drawings and specific embodiments, so that those skilled in the art can better understand and implement the present invention. However, the embodiments are not intended to limit the present invention. In the absence of conflict, the embodiments of the present invention and the technical features in the embodiments can be combined with each other.

[0044] This invention provides a method and system for function selection and execution in a multi-task environment based on a large model, which is used to solve the technical problem of how to understand user natural language input, automatically select and execute corresponding functions to complete complex tasks in a multi-task environment.

[0045] Example 1:

[0046] This invention provides a method for function selection and execution in a multi-task environment based on a large model, comprising four steps: function library construction, user problem analysis, function execution plan construction, and task execution.

[0047] Step S100 Function Library Modeling: For each function in the user API function library, define the function's attribute information and extract the function's feature vector. Construct a feature vector library based on the function and its corresponding feature vector. The function's attribute information includes description, input fields, output fields, type, and function.

[0048] In this embodiment, the feature vector of the function is extracted using a text embedding model.

[0049] Step S200 User Problem Analysis: For a user problem, extract the feature vector of the user problem, and find the function that matches the user problem based on the vector similarity between the feature vector of the user problem and the feature vector of the function.

[0050] In this embodiment, for a user's question or problem, the feature vector of the user's question is extracted through a text embedding model, the cosine similarity between the feature vector of the user's question and the feature vector of the function is calculated, and the function that matches the user's question is found based on the similarity between the feature vectors.

[0051] Step S300 Function Execution Plan Construction: Based on the user's problem and the matching function, analyze the user's needs and expectations through a large model, formulate a task execution strategy based on the user's needs and expectations, select the function to be executed based on the task execution strategy, and construct the function execution plan.

[0052] In this embodiment, a large model of prompt words is constructed based on user questions and matching functions. The large model analyzes the user's needs and expectations based on the input prompt words to evaluate the user's intention. If the current output of the user's intention does not meet the user's expectations, the user is asked further questions, and the user's intention is evaluated again based on the user's questions until the user's needs and expectations are met.

[0053] Step S400 Task Execution: For the function execution plan, each function is executed sequentially, and the currently executed function is monitored and adjusted. If all functions in the function execution plan are executed successfully, the successful task execution result is returned to the user. If any function in the function execution plan fails, the execution result of the failed task is returned to the user. When monitoring and adjusting the currently executed function, if the currently executed function returns a successful execution result, the next function is executed. If the currently executed function returns a failed execution result, the current executed function and its execution plan are adjusted and judged based on the execution log and the returned execution result of the currently executed function, using a large model. Based on the adjustment judgment result, the adjusted function or the adjusted function execution plan is re-executed. If the adjusted function is re-executed, the adjusted function is monitored and judged. If the adjusted function execution plan is re-executed, the current function is determined to have failed, and the execution result of the failed task is returned to the user.

[0054] In this embodiment, monitoring and adjusting the currently executing function includes the following steps:

[0055] (1) During the execution of the current function, record the execution process of the current function to form the execution log of the current function;

[0056] (2) After the current function finishes execution, determine whether the current function was executed successfully based on the execution result returned by the current function;

[0057] (3) If the current function executes successfully, determine whether there are any functions to be executed after the current function. If there are, execute the next function. If not, determine that the task has been executed successfully and return the successful task execution result to the user.

[0058] (4) If the current function fails to execute, the current function and the returned execution result are adjusted and judged based on the execution log of the current function and the execution result, and the large model is used. If the current function is adjusted, the adjusted function is re-executed and the adjusted function is monitored and judged. If the function execution plan is adjusted, the current function is determined to have failed to execute and the execution result of the failed task is returned to the user. The task execution operation is performed based on the adjusted function execution plan.

[0059] In a specific implementation, the return value or status code returned by the function is used as the execution result. Based on the execution log of the current function and the returned execution result, the current function is analyzed through a large model to identify the log type, identify the error type, and adjust the task execution strategy. It is then determined whether to adjust the currently executing function and / or adjust the function execution plan. Adjusting the currently executing function includes redefining the function's attribute information, and adjusting the function execution plan includes selecting the function to be executed and the execution order of the functions to form a new function execution plan.

[0060] During task execution, after all planning steps are completed, the large model will conduct a final check to confirm that all user requirements have been met and there are no remaining execution steps. If everything is normal, a task completion notification will be sent to the user, along with the task execution results.

[0061] Example 2:

[0062] This invention provides a function selection and execution system based on a large model in a multi-task environment, comprising a function library modeling module, a user problem analysis module, a function execution planning construction module, and a task execution module.

[0063] The function library modeling module is used to perform the following: For each function in the user API function library, define the function's attribute information and extract the function's feature vector. Based on the function and its corresponding feature vector, construct a feature vector library. The function's attribute information includes description, input fields, output fields, type, and function.

[0064] In this embodiment, the function library modeling module is used to extract the feature vectors of functions through a text embedding model.

[0065] The user problem analysis module performs the following: For a user problem, it extracts the feature vector of the user problem, and finds a function that matches the user problem based on the vector similarity between the feature vector of the user problem and the feature vector of the function.

[0066] In this embodiment, for a user's question or problem, the user question analysis module performs the following: extracts the feature vector of the user question through a text embedding model, calculates the cosine similarity between the user question feature vector and the function feature vector, and finds a function that matches the user question based on the similarity between the feature vectors.

[0067] The function execution planning module is used to perform the following: based on the user's problem and the matching function, analyze the user's needs and expectations through a large model, formulate a task execution strategy based on the user's needs and expectations, select the function to be executed based on the task execution strategy, and construct the function execution plan.

[0068] In this embodiment, the function execution planning construction module is used to build a large model of prompt words input based on user questions and matching functions. The large model analyzes the user's needs and expectations based on the input prompt words to evaluate the user's intention. If the current output of the user's intention does not meet the user's expectations, the user is asked further questions, and the user's intention is evaluated again based on the user's questions until the user's needs and expectations are met.

[0069] The task execution module performs the following: For a function execution plan, each function is executed sequentially, and the currently executing function is monitored and adjusted. If all functions in the function execution plan are executed successfully, the successful task execution result is returned to the user. If any function in the function execution plan fails, the execution result of the failed task is returned to the user. When monitoring and adjusting the currently executing function, if the currently executing function returns a successful execution result, the next function is executed. If the currently executing function returns a failed execution result, the current executing function and its execution plan are adjusted and judged based on the execution log and the returned execution result of the currently executing function, using a large model. Based on the adjustment judgment result, the adjusted function or the adjusted function execution plan is re-executed. If the adjusted function is re-executed, the adjusted function is monitored and judged. If the adjusted function execution plan is re-executed, the current function is determined to have failed, and the execution result of the failed task is returned to the user.

[0070] In this embodiment, the task execution module is used to monitor and adjust the currently executing function as follows:

[0071] (1) During the execution of the current function, record the execution process of the current function to form the execution log of the current function;

[0072] (2) After the current function finishes execution, determine whether the current function was executed successfully based on the execution result returned by the current function;

[0073] (3) If the current function executes successfully, determine whether there are any functions to be executed after the current function. If there are, execute the next function. If not, determine that the task has been executed successfully and return the successful task execution result to the user.

[0074] (4) If the current function fails to execute, the current function and the returned execution result are adjusted and judged based on the execution log of the current function and the execution result, and the large model is used. If the current function is adjusted, the adjusted function is re-executed and the adjusted function is monitored and judged. If the function execution plan is adjusted, the current function is determined to have failed to execute and the execution result of the failed task is returned to the user. The task execution operation is performed based on the adjusted function execution plan.

[0075] In a specific implementation, the return value or status code returned by the function is used as the execution result. Based on the execution log of the current function and the returned execution result, the current function is analyzed through a large model to identify the log type, identify the error type, and adjust the task execution strategy. It is then determined whether to adjust the currently executing function and / or adjust the function execution plan. Adjusting the currently executing function includes redefining the function's attribute information, and adjusting the function execution plan includes selecting the function to be executed and the execution order of the functions to form a new function execution plan.

[0076] During task execution, after all planning steps are completed, the large model will conduct a final check to confirm that all user requirements have been met and there are no remaining execution steps. If everything is normal, a task completion notification will be sent to the user, along with the task execution results.

[0077] The above provides a detailed description of the function selection and execution method and system based on a large model in a multi-task environment provided by the present invention. Specific examples have been used to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of the present invention. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A method for function selection and execution in a multi-task environment based on a large model, characterized in that, Includes the following steps: Function library modeling: For each function in the user API function library, define the function's attribute information and extract the function's feature vector. Construct a feature vector library based on the function and its corresponding feature vector. The function's attribute information includes description, input fields, output fields, type, and function. User problem analysis: For user problems, extract the feature vector of the user problem, and find the function that matches the user problem based on the vector similarity between the feature vector of the user problem and the feature vector of the function. Function execution plan construction: Based on the user's problem and the matching function, the user's needs and expectations are analyzed through a large model, and a task execution strategy is formulated based on the user's needs and expectations. Based on the task execution strategy, the function to be executed is selected and the function execution plan is constructed. Task Execution: For function execution plans, each function is executed sequentially, and the currently executing function is monitored and adjusted. If all functions in the function execution plan execute successfully, the successful task execution result is returned to the user. If any function in the function execution plan fails, the execution result of the failed task is returned to the user. When monitoring and adjusting the currently executing function, if the currently executing function returns a successful execution result, the next function is executed. If the currently executing function returns a failed execution result, the current executing function and its execution plan are adjusted and judged based on the execution log and the returned execution result of the currently executing function, using a large model. Based on the adjustment judgment result, the adjusted function or the adjusted function execution plan is re-executed. If the adjusted function is re-executed, the adjusted function is monitored and judged. If the adjusted function execution plan is re-executed, the current function is determined to have failed, and the execution result of the failed task is returned to the user.

2. The function selection and execution method based on a large model in a multi-task environment according to claim 1, characterized in that, The feature vectors of the function are extracted using a text embedding model, and the feature vectors of the user's question are also extracted using a text embedding model.

3. The function selection and execution method based on a large model in a multi-task environment according to claim 1, characterized in that, Monitoring and adjusting the currently executing function includes the following steps: During the execution of the current function, the execution process of the current function is recorded, forming the execution log of the current function; After the current function finishes execution, determine whether the current function execution was successful based on the execution result returned by the current function. If the current function executes successfully, check if there are any functions to be executed after the current function. If there are, execute the next function. If not, determine that the task has been executed successfully and return the successful task execution result to the user. If the current function fails to execute, the system adjusts and judges the currently executing function and its execution plan based on the execution log and returned execution result of the current function, using a large model. If the currently executing function is adjusted, the adjusted function is re-executed and monitored. If the function execution plan is adjusted, the system determines that the current function has failed to execute and returns the execution result of the failed task to the user. The system then performs task execution operations based on the adjusted function execution plan.

4. The function selection and execution method based on a large model in a multi-task environment according to claim 1, characterized in that, Using the return value or status code returned by the function as the execution result, and based on the execution log of the current function and the returned execution result, the system performs log analysis, error type identification, and task execution strategy adjustment on the current function through a large model. It then determines whether to adjust the currently executing function and / or adjust the function execution plan. Adjusting the currently executing function includes redefining the function's attribute information, and adjusting the function execution plan includes selecting the functions to be executed and the execution order of the functions to form a new function execution plan.

5. A function selection and execution system based on a large model in a multi-task environment, characterized in that, The system is used to implement function selection and execution in a multi-task environment based on a large model, as described in any one of claims 1-4. The system includes a function library modeling module, a user problem analysis module, a function execution planning construction module, and a task execution module. The function library modeling module is used to perform the following: For each function in the user API function library, define the function's attribute information and extract the function's feature vector. Based on the function and its corresponding feature vector, construct a feature vector library. The function's attribute information includes description, input fields, output fields, type, and function. The user problem analysis module performs the following: For a user problem, it extracts the feature vector of the user problem, and finds a function that matches the user problem based on the vector similarity between the feature vector of the user problem and the feature vector of the function. The function execution planning module is used to perform the following: based on the user's problem and the matching function, analyze the user's needs and expectations through a large model, formulate a task execution strategy based on the user's needs and expectations, select the function to be executed based on the task execution strategy, and construct the function execution plan; The task execution module performs the following: For a function execution plan, each function is executed sequentially, and the currently executing function is monitored and adjusted. If all functions in the function execution plan are executed successfully, the successful task execution result is returned to the user. If any function in the function execution plan fails, the execution result of the failed task is returned to the user. When monitoring and adjusting the currently executing function, if the currently executing function returns a successful execution result, the next function is executed. If the currently executing function returns a failed execution result, the current executing function and its execution plan are adjusted and judged based on the execution log and the returned execution result of the currently executing function, using a large model. Based on the adjustment judgment result, the adjusted function or the adjusted function execution plan is re-executed. If the adjusted function is re-executed, the adjusted function is monitored and judged. If the adjusted function execution plan is re-executed, the current function is determined to have failed, and the execution result of the failed task is returned to the user.

6. The function selection and execution system based on a large model in a multi-task environment according to claim 5, characterized in that, The function library modeling module is used to extract feature vectors of functions through text embedding models, while the user question analysis module is used to extract feature vectors of user questions through text embedding models.

7. The function selection and execution system based on a large model in a multi-task environment according to claim 5, characterized in that, The task execution module is used to monitor and adjust the currently executing function as follows: During the execution of the current function, the execution process of the current function is recorded, forming the execution log of the current function; After the current function finishes execution, determine whether the current function execution was successful based on the execution result returned by the current function. If the current function executes successfully, check if there are any functions to be executed after the current function. If there are, execute the next function. If not, determine that the task has been executed successfully and return the successful task execution result to the user. If the current function fails to execute, the system adjusts and judges the currently executing function and its execution plan based on the execution log and returned execution result of the current function, using a large model. If the currently executing function is adjusted, the adjusted function is re-executed and monitored. If the function execution plan is adjusted, the system determines that the current function has failed to execute and returns the execution result of the failed task to the user. The system then performs task execution operations based on the adjusted function execution plan.

8. The function selection and execution system based on a large model in a multi-task environment according to claim 5, characterized in that, The task execution module uses the return value or status code returned by the function as the execution result. Based on the execution log of the current function and the returned execution result, it performs log analysis, error type identification, and task execution strategy adjustment on the current function through a large model. It determines whether to adjust the currently executing function and / or adjust the function execution plan. Adjusting the currently executing function includes redefining the function's attribute information, and adjusting the function execution plan includes selecting the functions to be executed and the execution order of the functions to form a new function execution plan.

Images