Prompt versioning management system and method, medium, terminal and program product

Abstract

The invention provides a cue word versioning management system and method, a medium, a terminal and a program product, and the system comprises a cue word creating module which creates a cue word and allocates an initial version number for the created cue word; the interactive test module is used for carrying out interactive test on the cue word based on an LLM Playground platform and carrying out iterative optimization on the cue word after test; the version updating module updates the version number of the prompt word for each iteration optimization according to the change category, and labels the prompt word of the specific version number; the evaluation module evaluates the optimized cue word according to a preset evaluation index so as to obtain the cue word meeting the evaluation index; and the publishing and calling module publishes the cue words meeting the evaluation indexes, and calls the corresponding cue words according to the version number or the label based on the integrated software development kit. According to the method, full-process management automation of prompt word generation, optimization, evaluation, release and monitoring can be realized, and the prompt word management efficiency is greatly improved.

Description

Technical Field

[0001] This application relates to the field of artificial intelligence technology, and in particular to a prompt word versioning management system, method, medium, terminal and program product. Background Technology

[0002] In recent years, with the widespread adoption of large language model technology and the rapid improvement of multimodal capabilities, prompt word engineering has evolved from an initial experience-based skill into a core engineering discipline in the field of artificial intelligence. Prompt words are no longer simply input text, but rather core assets that define AI behavior, control output quality, and achieve safety and compliance. Their development model has also shifted from "optimizing individual prompt words" to "building a large-scale, standardized, and automated prompting platform."

[0003] However, existing prompt word engineering techniques still face many bottlenecks in practical applications. First, with the increasing number of prompt words and frequent iterations, problems such as the loss of historical high-quality versions, lack of modification records, and inconsistencies between versions are common. The lack of version backtracking capabilities and severely chaotic version management lead to unstable model output quality and an uncertain optimization process. Second, existing technologies largely rely on manual testing of single scenarios, lacking coverage of edge cases and multi-module linkages, and lacking testing and verification mechanisms. Finally, the same prompt word may produce different results in different contexts, making it difficult to objectively compare the merits of different versions. Furthermore, the complex dependencies between multiple prompt words in complex systems make it difficult to isolate and evaluate the impact of changes, resulting in significant uncertainty in the evaluation effectiveness of prompt words.

[0004] Therefore, it is necessary to provide a prompt word versioning management system, method, medium, terminal, and program product to solve the above-mentioned problems existing in the prior art. Summary of the Invention

[0005] In view of the shortcomings of the prior art described above, the purpose of this application is to provide a prompt word versioning management system, method, medium, terminal and program product to solve the technical problems of serious chaos in version management, lack of testing and verification mechanism and great uncertainty in the evaluation effect of prompt words in the prior art.

[0006] To achieve the above and other related objectives, a first aspect of this application provides a prompt word versioning management system, comprising: a prompt word creation module, configured to acquire a preset prompt word template, assign values ​​to variables in the acquired prompt word template to create prompt words, and assign an initial version number to the created prompt words; an interactive testing module, configured to perform interactive testing on prompt words with the initial version number based on the LLM Playground platform, and iteratively optimize the tested prompt words; a version update module, configured to update the version number of each iteratively optimized prompt word according to the change category, and tag prompt words with specific version numbers; an evaluation module, configured to evaluate the optimized prompt words according to preset evaluation indicators to obtain prompt words that meet the evaluation indicators; and a publishing and calling module, configured to publish prompt words that meet the evaluation indicators, and call the corresponding prompt words according to the version number or tag based on an integrated software development kit.

[0007] In some embodiments of the first aspect of this application, the version number includes a major version number, a minor version number, and a patch version number. The major version number is updated when the semantics of the prompt word undergoes a significant change; the minor version number is updated when the semantics of the prompt word undergoes a minor adjustment; and the patch version number is updated when the text of the prompt word is corrected.

[0008] In some embodiments of the first aspect of this application, the label is an environmental label, which includes production, testing, and development.

[0009] In some embodiments of the first aspect of this application, an intelligent optimization module is also included, which is placed after the interactive testing module and is used to optimize the prompt words of the current version based on the large language model according to preset meta prompt words, so as to output the optimization results.

[0010] In some embodiments of the first aspect of this application, a monitoring module is also included, which monitors the usage and effect of the prompt words and feeds back the monitoring data to the intelligent optimization module.

[0011] In some embodiments of the first aspect of this application, a prompt word library module is also included for storing prompt words for various versions.

[0012] To achieve the above and other related objectives, a second aspect of this application provides a method for versioned management of prompt words, comprising: obtaining a preset prompt word template; assigning values ​​to variables in the obtained prompt word template to create prompt words; and assigning an initial version number to the created prompt words; performing interactive testing on the prompt words with the initial version number based on the LLM Playground platform, and iteratively optimizing the prompt words after testing; updating the version number of each iteratively optimized prompt word according to the change category, and tagging the prompt words with specific version numbers; evaluating the optimized prompt words according to preset evaluation indicators to obtain prompt words that meet the evaluation indicators; publishing the prompt words that meet the evaluation indicators, and calling the corresponding prompt words according to the version number or tag based on an integrated software development kit.

[0013] To achieve the above and other related objectives, a third aspect of this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method.

[0014] To achieve the above and other related objectives, a fourth aspect of this application provides a computer program product comprising computer program code that, when executed on a computer, causes the computer to implement the method.

[0015] To achieve the above and other related objectives, a fifth aspect of this application provides an electronic terminal, including a memory, a processor, and a computer program stored in the memory; the processor executes the computer program to implement the method.

[0016] As described above, the prompt word versioning management system, method, medium, terminal, and program product of this application have the following beneficial effects:

[0017] By first acquiring a preset prompt template, assigning values ​​to variables in the prompt template to create prompts with initial version numbers, and then conducting real-time interactive testing of the prompts with initial version numbers through the LLM Playground platform, rapid iterative optimization is performed in the visualization platform. For each iteration, the corresponding version number of the prompt is updated based on the change type, achieving a complete version history of the prompts. This ensures that every addition, deletion, modification, and query of prompts is accurately recorded, along with the modifier, modification time, modification summary, and differences between versions, improving version backtracking capabilities and significantly increasing the efficiency of prompt version management. Prompts with specific version numbers are tagged for easy release and rollback. Furthermore, optimized prompts are evaluated based on assessment metrics to ensure that only prompts meeting the assessment metrics are released and deployed. A software development kit is integrated, enabling the retrieval of corresponding prompts by version number or tag, achieving rapid application of prompts. Attached Figure Description

[0018] Figure 1 The diagram shown is a block diagram of a prompt word versioning management system according to one embodiment of this application.

[0019] Figure 2 The diagram shown is a full-cycle flowchart of a prompt word versioning management system according to one embodiment of this application.

[0020] Figure 3 The diagram shown illustrates an application scenario of a prompt word versioning management system according to one embodiment of this application.

[0021] Figure 4 The diagram shown is a flowchart illustrating a method for versioning prompt words in one embodiment of this application.

[0022] Figure 5 The diagram shown is a structural schematic of an electronic terminal according to an embodiment of this application. Detailed Implementation

[0023] The following specific examples illustrate the implementation of this application. Those skilled in the art can easily understand other advantages and effects of this application from the content disclosed in this specification. This application can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of this application. It should be noted that, unless otherwise specified, the following embodiments and features in the embodiments can be combined with each other.

[0024] In the embodiments of this application, terms such as "first" and "second" are used to distinguish identical or similar items with essentially the same function and effect. For example, "first XX" and "second XX" are merely used to distinguish different XXs and do not limit their order. Those skilled in the art will understand that terms such as "first" and "second" do not limit the quantity or execution order, and that "first" and "second" do not necessarily imply that they are different.

[0025] It should be noted that, in the embodiments of this application, the words "exemplary" or "for example" indicate examples, illustrations, or descriptions. Any embodiment or design described as "exemplary" or "for example" in this application should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of words such as "exemplary" or "for example" is intended to present the relevant concepts in a concrete manner.

[0026] In this application embodiment, "at least one" refers to one or more, and "more than one" refers to two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, or B alone, where A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. "At least one of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one of a, b, or c can represent: a, b, c, ab, ac, bc, or abc, where a, b, and c can be single or multiple.

[0027] Before providing a further detailed description of the present invention, the nouns and terms used in the embodiments of the present invention are explained, and the nouns and terms used in the embodiments of the present invention are subject to the following interpretations:

[0028] <1> Prompt: A prompt is a command or query that a user provides to a large language model or other generative AI. It is usually composed of natural language and is designed to guide the model to produce a response that meets specific needs, styles, or formats.

[0029] <2> LLM Playground platform: An interactive platform designed for large language models (LLMs). It integrates multiple AI models and features functions such as customized system prompts, web search, and voice input. The platform supports text dialogue and content generation, and provides chat history management. Users can customize API endpoints to achieve flexible AI interaction.

[0030] <3> A Software Development Kit (SDK) is a collection of tools used to help developers create, test, and deploy software applications. It is typically provided by the manufacturer of the hardware platform, operating system, or programming language and includes relevant documentation, examples, and tools designed to assist in the development of specific types of software.

[0031] <4> Large Language Model (LLM): refers to a deep learning model trained on a large amount of text data, which enables the model to generate natural language text or understand the meaning of language text. These models can provide in-depth knowledge and language production on a wide range of topics by being trained on massive datasets. The core idea is to learn the patterns and structures of natural language through large-scale unsupervised training, which to some extent simulates the human language cognition and generation process.

[0032] To facilitate understanding of the embodiments of this application, in conjunction with Figure 1 and Figure 2 Detailed explanation. Figure 1 A block diagram of a prompt word versioning management system according to an embodiment of this application is shown. Figure 2 This paper illustrates the full lifecycle process of a prompt word versioning management system according to an embodiment of this application. The prompt word versioning management system 100 in this embodiment includes: a prompt word creation module 101, an interactive testing module 102, a version update module 103, an evaluation module 104, and a release and invocation module 105. Through the layered architecture design of the prompt word creation module 101, interactive testing module 102, version update module 103, evaluation module 104, and release and invocation module 105, a complete closed-loop management system of development-testing optimization-evaluation-release deployment is formed, meeting the requirements of scalability, maintainability, and performance.

[0033] The prompt word creation module 101 is used to obtain a preset prompt word template, assign values ​​to the variables in the obtained prompt word template to create prompt words, and assign an initial version number to the created prompt words.

[0034] This prompt creation module quickly generates prompts that meet business requirements based on the provided prompt template library and variable functionality. Specifically, users select a corresponding prompt template from the preset prompt template library according to their business needs, fill in the corresponding content for the variables in the selected prompt template, and modify the quantitative parts of the prompt template to better fit the target output task. Clicking the generate button generates the corresponding initial prompt, and assigning a version number to the generated prompt facilitates subsequent prompt management and traceability.

[0035] The interactive testing module 102 is used to perform interactive testing on the prompts for the initial version number based on the LLM Playground platform, and to iteratively optimize the prompts after the test.

[0036] This system integrates the LLM Playground platform, enabling real-time interactive testing of prompts within a visual Playground. This allows for rapid verification of prompt effectiveness and enables developers to intuitively compare the output of different large language models. Specifically, within the LLM Playground platform, when a test is triggered, requests are sent in parallel to multiple selected model APIs based on configured parameters. The generated results of the selected models are displayed in real-time in the corresponding panes, facilitating horizontal comparison of the outputs of multiple models and highlighting the differences between their outputs. If the output does not meet the requirements, the prompts can be modified directly in the platform's editor. The modified prompts are then retested, forming a rapid iterative optimization loop until a generated result that meets the preset requirements is output. This version of the prompt is then saved as a version awaiting submission.

[0037] Version update module 103 is used to update the version number of the prompt words for each iteration of optimization according to the change category, and to tag the prompt words for specific version numbers.

[0038] In some embodiments of this application, the version number includes a major version number, a minor version number, and a patch version number. The major version number is updated when the semantics of the prompt word undergoes a significant change; the minor version number is updated when the semantics of the prompt word undergoes a minor adjustment; and the patch version number is updated when the text of the prompt word is corrected.

[0039] In this embodiment, the version number is a semantic version number, such as v XYZ, where X, Y, and Z all take values ​​from 0 to 9. v represents the version number; X represents the major version number, which is updated only when the semantics of the prompt word undergoes a significant change. For example, changing "customer service answer" to "intelligent recommendation" updates the major version number. Y represents the minor version number, which is updated only when the semantics of the prompt word undergoes a minor adjustment. For example, adding a "friendly tone" requirement to the prompt word updates only the minor version number. Z represents the patch version number, which is updated when the text of the prompt word is corrected. For example, if a spelling error occurs in the text of the prompt word, the patch version number is updated after the prompt word content is corrected.

[0040] In some embodiments of this application, the label is an environmental label, which includes production, testing, and development.

[0041] This system provides a tagging function, supporting custom environment tags for tooltips. Tooltips used in production environments are tagged as "production," those used in testing environments as "test," and those used in development environments as "development." When a tooltip with a specific version number is tagged with "production," "test," or "development" through deployment, and the tag and corresponding version number are bound, that version of the tooltip will be used in the production / test / development environments. Tooltips are released and deployed by switching tags. When a version rollback is needed, simply switch the production / test / development environment tag back to the previous stable version. If a tooltip in the production environment has a problem, version management allows for quick location of the issue. For example, if the current version number of the tooltip is v1.1.0; the last modification was submitted by Zhang San on 2024-03-16; the change is "added special addition xxx"; and the test report shows "95% of test cases passed," simply checking the logs allows tracing back to the corresponding version tooltip, improving the traceability of tooltips.

[0042] Through the design of the prompt word creation module 101, interactive testing module 102, and version update module 103, the functions of prompt word creation, version update, and rollback are realized. Furthermore, a semantic version number strategy is adopted to provide a simple and clear set of management rules, which supports precise location of the problematic version prompt word, thereby improving the traceability and reliability of prompt word management.

[0043] The evaluation module 104 is used to evaluate the optimized prompt words according to preset evaluation indicators in order to obtain prompt words that meet the evaluation indicators.

[0044] If the optimized prompt words on the LLM Playground platform are self-tested and meet the requirements, the prompt words of that version are submitted to the evaluation module 104. The module evaluates the prompt words according to the preset evaluation indicators and generates a corresponding evaluation report. If the results of the evaluation report meet the release standards, the prompt words of that version are released. If the prompt words of that version do not meet the release standards, they are returned to the LLM Playground platform for testing and optimization again until the evaluation indicators are met.

[0045] For example, the optimized prompts and their output are input into a large model, which is then asked to score them according to a preset scoring criterion for evaluation. Specific rules are also designed to perform appropriate format checks on the optimized prompts.

[0046] The publishing and invocation module 105 is used to publish prompts that meet the evaluation criteria and, based on the integrated software development kit, invoke the corresponding prompts according to the version number or tag.

[0047] Switching the version tag of a prompt word that meets the evaluation criteria to a production tag indicates that the prompt word has been released and is about to be used in the production environment. At the same time, a software development kit (SDK) is integrated to quickly enable the integration of AI applications. The corresponding prompt word content can be obtained by specifying the prompt word version number or tag, ensuring the accuracy of the call. Users can use the SSD to call prompt words that meet the evaluation criteria, which improves the call speed.

[0048] In some embodiments of this application, the prompt word versioning management system further includes an intelligent optimization module, which is placed after the interactive testing module. This module is used to optimize the prompt words of the current version based on the large language model and preset meta prompt words, so as to output the optimization results.

[0049] This system incorporates intelligent optimization capabilities, combining AI and evaluation feedback to facilitate continuous optimization of prompt words by developers. In this embodiment, the intelligent optimization module complements and works in conjunction with the iterative optimization within the LLM Playground platform to generate optimal prompt words. For example, the large language model uses the GPT model, pre-setting a meta-prompt word. The meta-prompt word includes role assignment, task definition, principles, and constraints, optimizing the current version of the prompt words, thereby achieving intelligent optimization of the prompt words. It should be noted that the intelligent optimization module can also be built into the LLM Playground platform; this application does not specifically limit its integration.

[0050] In some embodiments of this application, the prompt word versioning management system further includes a monitoring module for monitoring the usage and effect of prompt words and feeding the monitoring data back to the intelligent optimization module.

[0051] This system includes a monitoring module that monitors the usage and generation effect of prompt words in real time and feeds the monitoring data back to the intelligent optimization module. Based on the feedback data, a new round of optimization is carried out, thus forming a closed-loop process of continuous iterative optimization.

[0052] This system is designed with a complete end-to-end management system for development, testing, verification and evaluation, release and deployment, and monitoring, which greatly improves the management efficiency of prompt words and ensures the output quality of prompt words.

[0053] In some embodiments of this application, the prompt word versioning management system further includes a prompt word library module for storing prompt words for various versions.

[0054] This system includes a prompt word library module that stores the original text, semantic embedding vectors, and metadata of all prompt word versions, facilitating subsequent tracing and retrieval.

[0055] Through the design of the prompt word creation module 101, interactive testing module 102, version update module 103, evaluation module 104, release and call module 105, intelligent tuning module, monitoring module, and prompt word library module, a complete prompt word version history is achieved. This ensures that every addition, deletion, modification, and query of each prompt word is accurately recorded. It also records the modifier, modification time, and modification summary, and can even compare the differences between any two prompt word versions. It also supports tagging important prompt word versions for easy subsequent release and rollback. LLM is integrated. The Playground platform enables rapid iteration and updates of tooltips within a visual platform, supporting online version editing. Each change is indicated based on differences calculated using a Diff algorithm, significantly improving management efficiency. Simultaneously, any modified tooltips must undergo verification and evaluation before release, ensuring their effectiveness remains within controllable limits. The entire process—from tooltip generation, testing, evaluation, release, to monitoring—is automated, forming a closed-loop workflow that significantly reduces manual intervention costs. Data-driven optimization automatically adjusts tooltips based on evaluation data and AI algorithms, ensuring a scientific and transparent optimization process with quantifiable results. Through intelligent evaluation and optimization mechanisms, the system automatically identifies inefficient tooltips and proposes improvement suggestions, thereby reducing API call costs and computational resource consumption. The system's automated testing and monitoring mechanisms promptly detect problems and automatically roll back, avoiding business interruptions and manual troubleshooting costs caused by tooltip issues. It should be understood that the Diff algorithm is used to compare differences between old and new virtual DOMs (Document Object Models) and update the real DOM based on these differences.

[0056] Figure 3 This illustration shows an application scenario diagram of the prompt word versioning management system in an embodiment of this application. For example... Figure 3 As shown, a specific example of the workflow of the prompt word versioning management system is given in a specific application scenario. Scenario description: Write a marketing copy for a certain series of thermos cups.

[0057] Retrieve the "Marketing Copywriting" template from the prompt word template, for example, "Please write a marketing copy for (product name) highlighting its (key features)". Enter the product name: a series of thermos cups, and the key feature: extra-long insulation. The prompt word for creating the initial version number will be "Please write a marketing copy for a series of thermos cups highlighting its extra-long insulation". The version number of this version is v1.0.0, and this version will be tagged with "development". The prompts for version 1.0.0 were input into the LLMPlayground platform for testing, outputting the text "A certain series of thermos cups boasts an ultra-long heat retention capacity of 24 hours." Analysis of this output text revealed it to be too generic, requiring optimization. The prompts were modified in the editor, assigning roles and adding style requirements, resulting in the prompt for version 1.1.0: "You are a seasoned copywriting expert for household products. Please write a marketing copy for a certain series of thermos cups, highlighting its ultra-long heat retention. Requirements: The opening should be eye-catching, using contextual descriptions." This version was tagged with "development." Further analysis of the output text for version 1.1.0 indicated further optimization was needed. The intelligent optimization module was activated to further refine the prompts, resulting in version v1.2.0: "You are a senior copywriting expert for household products. Please write a marketing copy for a series of thermos cups, highlighting their extended heat retention. Requirements: The opening should be eye-catching, using scenario-based descriptions; only elaborate on the provided key features, do not fabricate non-existent functions; if the selling points are insufficient, focus on describing the user experience brought by the key features." After self-testing and passing version v1.2.0, the "development" tag was removed from version v1.2.0, and the "test" tag was added. It was then submitted to the evaluation module for assessment and deemed... If the evaluation criteria are met, version v1.2.0 will be released and launched, the "test" tag will be removed, and a "production" tag will be added. Users using the integrated SDK will invoke the prompt message based on the version v1.2.0 or the "production" tag. The prompt message will then be: "You are a senior copywriting expert for household products. Please write a marketing copy for a certain series of thermos cups, highlighting its ultra-long insulation. Requirements: The opening should be eye-catching, using a scenario-based description; only describe the provided key features, do not fabricate non-existent functions; if there are insufficient selling points, focus on describing the user experience brought by the key features."

[0058] The prompt word versioning management system of this application provides a reliable prompt word management system for the entire process of prompt word versioning management, which can meet the needs of enterprise-level applications. It has also established a complete testing and verification mechanism and version rollback capability, supports multimodal, intelligent and automated management, greatly improves the management efficiency of prompt words and ensures the output quality of the model.

[0059] Figure 4 This is a flowchart illustrating the prompt word versioning management method provided in an embodiment of this application. For example... Figure 4 As shown, the method for version control of prompt words includes the following steps:

[0060] Step S41: Obtain a preset prompt word template, assign values ​​to the variables in the obtained prompt word template to create prompt words, and assign an initial version number to the created prompt words;

[0061] Step S42: Conduct interactive tests on the prompts for the initial version number based on the LLM Playground platform, and iteratively optimize the prompts after the tests;

[0062] Step S43: Update the version number of the prompt words for each iteration of optimization according to the change category, and label the prompt words for specific version numbers;

[0063] Step S44: Evaluate the optimized prompt words according to the preset evaluation indicators to obtain prompt words that meet the evaluation indicators;

[0064] Step S45: Publish the prompts that meet the evaluation criteria, and call the corresponding prompts according to the version number or tag based on the integrated software development kit.

[0065] It should be understood that the prompt word versioning management method provided in this application embodiment can realize the functions of the above-mentioned prompt word versioning management system, and for the sake of brevity, it will not be described in detail here.

[0066] It should also be understood that the module division in the embodiments of this application is illustrative and only represents a logical functional division; in actual implementation, there may be other division methods. Furthermore, the functional modules in the various embodiments of this application can be integrated into a single processor, exist as separate physical entities, or be integrated into a single module. The integrated modules described above can be implemented in hardware or as software functional modules.

[0067] Figure 5 This is a schematic block diagram of the electronic terminal provided in an embodiment of this application. Figure 5 As shown, the electronic terminal 500 includes at least one processor 501, a memory 502, at least one network interface 503, and a user interface 505. The various components in the electronic terminal 500 are coupled together via a bus system 504. It is understood that the bus system 504 is used to implement communication between these components. In addition to a data bus, the bus system 504 also includes a power bus, a control bus, and a status signal bus. However, for clarity, in… Figure 5 The general will label all buses as bus systems.

[0068] The user interface 505 may include a monitor, keyboard, mouse, trackball, clicker, button, touchpad, or touch screen.

[0069] It is understood that memory 502 can be volatile memory or non-volatile memory, or both. Non-volatile memory can be read-only memory (ROM) or programmable read-only memory (PROM), used as an external cache. By way of example, but not limitation, many forms of RAM are available, such as static random access memory (SRAM) and synchronous static random access memory (SSRAM). The memories described in the embodiments of this invention are intended to include, but are not limited to, these and any other suitable categories of memory.

[0070] In this embodiment of the invention, the memory 502 is used to store various types of data to support the operation of the electronic terminal 500. Examples of this data include: any executable program for operation on the electronic terminal 500, such as the operating system 5021 and application programs 5022; the operating system 5021 contains various system programs, such as the framework layer, core library layer, driver layer, etc., for implementing various basic services and handling hardware-based tasks. The application program 5022 may contain various applications, such as a media player, browser, etc., for implementing various application services. The methods provided in this embodiment of the invention can be included in the application program 5022.

[0071] The methods disclosed in the above embodiments of the present invention can be applied to processor 501, or implemented by processor 501. Processor 501 may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by the integrated logic circuit of the hardware in processor 501 or by instructions in the form of software. The processor 501 may be a general-purpose processor, a digital signal processor (DSP), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Processor 501 can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of the present invention. General-purpose processor 501 may be a microprocessor or any conventional processor, etc. The steps of the accessory optimization method provided in the embodiments of the present invention can be directly reflected as being executed by a hardware decoding processor, or being executed by a combination of hardware and software modules in the decoding processor. The software module may be located in a storage medium, which is located in a memory. The processor reads the information in the memory and combines it with its hardware to complete the steps of the aforementioned method.

[0072] In an exemplary embodiment, the electronic terminal 500 may be used by one or more application-specific integrated circuits (ASICs), DSPs, programmable logic devices (PLDs), or complex programmable logic devices (CPLDs) to execute the aforementioned method.

[0073] According to the method provided in the embodiments of this application, this application also provides a computer program product, which includes: computer program code, which, when run on a computer, causes the computer to execute... Figure 4 The method in the illustrated embodiment.

[0074] According to the method provided in the embodiments of this application, this application also provides a computer-readable storage medium storing program code, which, when executed on a computer, causes the computer to perform... Figure 4 The method in the illustrated embodiment.

[0075] As used in this specification, the terms "component," "module," "system," etc., are used to refer to computer-related entities, hardware, firmware, combinations of hardware and software, software, or software in execution. For example, a component can be, but is not limited to, a process running on a processor, a processor, an object, an executable file, an execution thread, a program, and / or a computer. As illustrated, applications running on computing devices and computing devices can both be components. One or more components may reside in a process and / or an execution thread, and components may be located on a single computer and / or distributed among two or more computers. Furthermore, these components can be executed from various computer-readable media on which various data structures are stored. Components can communicate, for example, via local and / or remote processes based on signals having one or more data packets (e.g., data from two components interacting with another component between a local system, a distributed system, and / or a network, such as the Internet interacting with other systems via signals).

[0076] Those skilled in the art will recognize that the various illustrative logical blocks and steps described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this application.

[0077] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0078] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0079] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0080] In addition, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0081] In the above embodiments, the functions of each functional unit can be implemented entirely or partially through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented entirely or partially in the form of a computer program product. A computer program product includes one or more computer instructions (programs). When the computer program instructions (programs) are loaded and executed on a computer, all or part of the flow or function according to the embodiments of this application is generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. Computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available media can be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., high-density digital video discs, DVDs), or semiconductor media (e.g., solid-state disks, SSDs, etc.).

[0082] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0083] In summary, addressing the technical problems of severely chaotic version management, lack of testing and verification mechanisms, and significant uncertainty in the evaluation effectiveness of prompt words in existing technologies, this application provides a prompt word versioning management system, method, medium, terminal, and program product. This involves first obtaining a preset prompt word template, assigning values ​​to variables in the prompt word template to create prompt words with initial version numbers, and then using LLM... The Playground platform performs real-time interactive testing on the initial version number of the prompt words, rapidly iterates and optimizes them in a visual platform, and updates the corresponding version number of each iteration's prompt words based on the change type. This achieves a complete version history of the prompt words, ensuring that every addition, deletion, modification, and query of prompt words is accurately recorded. It also records the modifier, modification time, modification summary, and differences between different versions, improving version backtracking capabilities and significantly increasing the efficiency of prompt word version management. Furthermore, prompt words with specific version numbers are tagged for easy release and rollback. The optimized prompt words are evaluated according to assessment metrics, ensuring that only prompt words meeting the evaluation criteria are released and deployed. It also integrates a software development kit (SDK), enabling the retrieval of corresponding prompt words by version number or tag, achieving rapid application deployment. Therefore, this application effectively overcomes the various shortcomings of existing technologies and has high industrial applicability.

[0084] The above embodiments are merely illustrative of the principles and effects of this application and are not intended to limit this application. Any person skilled in the art can modify or alter the above embodiments without departing from the spirit and scope of this application. Therefore, all equivalent modifications or alterations made by those skilled in the art without departing from the spirit and technical concept disclosed in this application should still be covered by the claims of this application.

Claims

1. A prompt word versioning management system, characterized in that, include: The prompt word creation module is used to obtain a preset prompt word template, assign values ​​to the variables in the obtained prompt word template to create prompt words, and assign an initial version number to the created prompt words; The interactive testing module is used to perform interactive tests on the prompts for the initial version number based on the LLM Playground platform, and to iteratively optimize the prompts after the tests. The version update module is used to update the version number of the prompt words for each iteration of optimization based on the change category, and to tag the prompt words for specific version numbers; The evaluation module is used to evaluate the optimized prompt words according to preset evaluation indicators in order to obtain prompt words that meet the evaluation indicators. The publishing and invocation module is used to publish prompts that meet the evaluation criteria and, based on the integrated software development kit, invoke the corresponding prompts according to the version number or tag.

2. The prompt word versioning management system according to claim 1, characterized in that, The version number includes a major version number, a minor version number, and a patch version number. The major version number is updated when there are significant changes in the semantics of the prompt words; the minor version number is updated when there are minor adjustments to the semantics of the prompt words; and the patch version number is updated when the text of the prompt words is corrected.

3. The prompt word versioning management system according to claim 1, characterized in that, The label is an environmental label, which includes production, testing, and development.

4. The prompt word versioning management system according to claim 1, characterized in that, It also includes an intelligent optimization module, which is placed after the interactive testing module. This module is used to optimize the prompts of the current version based on the large language model and preset meta-prompts, so as to output the optimization results.

5. The prompt word versioning management system according to claim 4, characterized in that, It also includes a monitoring module, which monitors the usage and effect of prompt words and feeds the monitoring data back to the intelligent optimization module.

6. The prompt word versioning management system according to claim 1, characterized in that, It also includes a prompt word library module, which stores prompt words for various versions.

7. A method for versioned management of prompt words, characterized in that, include: Obtain a preset prompt word template, assign values ​​to the variables in the obtained prompt word template to create prompt words, and assign an initial version number to the created prompt words; Interactive testing was conducted on the prompts for the initial version number using the LLM Playground platform, and the prompts were iteratively optimized after testing. For each iteration of optimization, update the version number of the prompt words according to the change category, and label the prompt words for specific version numbers; The optimized prompt words are evaluated based on preset evaluation indicators to obtain prompt words that meet the evaluation indicators; The prompts that meet the evaluation criteria will be published, and the corresponding prompts will be called according to the version number or tag based on the integrated software development kit.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the method as described in claim 7.

9. A computer program product, characterized in that, The computer program product includes computer program code that, when run on a computer, causes the computer to implement the method as described in claim 7.

10. An electronic terminal, comprising a memory, a processor, and a computer program stored in the memory, characterized in that, The processor executes the computer program to implement the method as described in claim 7.

Images