Input response determination method and device, equipment and storage medium

Abstract

The invention discloses an input response determination method and device, equipment and a storage medium, and relates to the technical field of big data, the method comprises the steps that after a training request is received, a target algorithm and a target model are determined according to the training request, and the training request is determined according to an optimization algorithm and a base model selected by a user; processing the use case data uploaded by the user based on the target algorithm to obtain prompt information and / or example information; inputting input information and the prompt information and / or the example information into the target model, so that the target model determines an input response corresponding to the input information according to the prompt information and / or the example information. According to the technical scheme, the use case data uploaded by the user is processed based on the target algorithm, and low-cost optimization is performed on the target model according to the obtained prompt information and / or example information, so that the target model can determine a more accurate input response corresponding to the input information.

Description

Technical Field

[0001] Embodiments of the present invention relate to the field of big data technology, and in particular to an input response determination method, apparatus, device and storage medium. Background Art

[0002] The big model is a machine model that generates content based on input information and has powerful natural language processing capabilities. In order to ensure the accuracy of the model generation results, the big model can be further trained according to business needs, and prompt words can also be set for the big model.

[0003] The continued training of large models and the setting of prompt words for large models require the participation of algorithm engineers, which has a high threshold and cannot meet the needs of business personnel.

[0004] Therefore, there is an urgent need for a method to lower the threshold for using big models and improve the big model usage experience of business personnel who have no training experience and only know how to use big models. Summary of the invention

[0005] The present invention provides an input response determination method, device, equipment and storage medium to achieve low-threshold optimization of a large model, so that the large model can determine a more accurate input response corresponding to the input information.

[0006] In a first aspect, an embodiment of the present invention provides an input response determination method, comprising:

[0007] After receiving the training request, determining the target algorithm and the target model according to the training request, wherein the training request is determined according to the optimization algorithm and the base model selected by the user;

[0008] Processing the use case data uploaded by the user based on the target algorithm to obtain prompt information and / or example information;

[0009] Input information and the prompt information and / or the example information are input into the target model, so that the target model determines an input response corresponding to the input information according to the prompt information and / or the example information.

[0010] The technical solution of the embodiment of the present invention provides an input response determination method, comprising: after receiving a training request, determining a target algorithm and a target model according to the training request, wherein the training request is determined according to the optimization algorithm and the base model selected by the user; processing the use case data uploaded by the user based on the target algorithm to obtain prompt information and / or example information; inputting the input information and the prompt information and / or the example information into the target model, so that the target model determines the input response corresponding to the input information according to the prompt information and / or the example information. The above technical solution can first determine the target algorithm and the target model according to the training request determined according to the optimization algorithm and the base model selected by the user, and the target algorithm can be used to optimize the target model; secondly, the use case data uploaded by the user can be processed based on the target algorithm to determine the prompt information and / or example information used to make the model more compatible with the user's needs, and then the input information and the prompt information and / or example information can be input into the target model, and the target model can determine the input response corresponding to the input information with reference to the prompt information and / or example information, and after the target model is optimized at a low cost through the prompt information and / or example information, the target model can determine a more accurate input response corresponding to the input information.

[0011] Further, determining a target algorithm and a target model according to the training request includes:

[0012] Determining the target algorithm according to the algorithm identifier included in the training request, wherein the algorithm identifier is determined according to an optimization control triggered by the user based on an algorithm selection page;

[0013] The target model is determined according to a model identifier included in the training request, wherein the model identifier is determined according to a model control triggered by the user based on a model selection page.

[0014] Furthermore, when the target algorithm is a prompt optimization algorithm, the use case data uploaded by the user is processed based on the target algorithm to obtain prompt information and / or example information, including:

[0015] Process the use case data based on a prompt word search algorithm to obtain a prompt word;

[0016] Accordingly, inputting the input information and the prompt information and / or the example information into the target model so that the target model determines an input response corresponding to the input information according to the prompt information and / or the example information, includes:

[0017] The input information and the prompt word are input into the target model, so that the target model determines the input response corresponding to the input information according to the prompt word.

[0018] Furthermore, when the target algorithm is a prompt optimization algorithm, the use case data uploaded by the user is processed based on the target algorithm to obtain prompt information and / or example information, including:

[0019] Processing the use case data based on a prompt vector extraction algorithm to obtain a prompt vector;

[0020] Accordingly, inputting the input information and the prompt information and / or the example information into the target model so that the target model determines an input response corresponding to the input information according to the prompt information and / or the example information, includes:

[0021] The input information and the prompt vector are input into the target model, and the prompt vector of the target model is adjusted to the prompt vector, so that the target model determines the input response corresponding to the input information according to the prompt vector.

[0022] Furthermore, when the target algorithm is an example optimization algorithm, the use case data uploaded by the user is processed based on the target algorithm to obtain prompt information and / or example information, including:

[0023] Process the use case data based on an example optimization algorithm to obtain example information;

[0024] Accordingly, inputting the input information and the prompt information and / or the example information into the target model so that the target model determines an input response corresponding to the input information according to the prompt information and / or the example information, includes:

[0025] The input information and the example information are input into the target model, so that the target model determines the input response corresponding to the input information according to the example information.

[0026] Furthermore, when the target algorithm is a prompt optimization algorithm and an example optimization algorithm, the use case data uploaded by the user is processed based on the target algorithm to obtain prompt information and / or example information, including:

[0027] Processing the use case data based on a prompt word search algorithm to obtain at least one prompt word, or processing the use case data based on a prompt vector extraction algorithm to obtain at least one prompt vector;

[0028] Process the use case data based on an example optimization algorithm to obtain at least one example information;

[0029] Accordingly, inputting the input information and the prompt information and / or the example information into the target model so that the target model determines an input response corresponding to the input information according to the prompt information and / or the example information, includes:

[0030] Construct at least one set of combined information according to each of the example information and each of the prompt words or each of the prompt vectors;

[0031] The input information, the use case data and each group of the combination information are input into the target model, so that the target model determines the output use case corresponding to the input use case in the use case data according to each group of the combination information, determines the target combination information according to the actual output corresponding to the input use case in the use case data and the output use case corresponding to the output use case determined by each group of the combination information, and determines the input response corresponding to the input information according to the input information and the target combination information.

[0032] In a second aspect, an embodiment of the present invention further provides an input response determination device, comprising:

[0033] A determination module, configured to determine a target algorithm and a target model according to a training request after receiving the training request, wherein the training request is determined according to an optimization algorithm and a base model selected by a user;

[0034] A processing module, used for processing the use case data uploaded by the user based on the target algorithm to obtain prompt information and / or example information;

[0035] The response module is used to input the input information and the prompt information and / or the example information into the target model, so that the target model determines the input response corresponding to the input information according to the prompt information and / or the example information.

[0036] In a third aspect, an embodiment of the present invention further provides an electronic device, the electronic device comprising:

[0037] at least one processor; and a memory communicatively coupled to the at least one processor;

[0038] The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the input response determination method as described in any one of the first aspects.

[0039] In a fourth aspect, an embodiment of the present invention further provides a storage medium comprising computer executable instructions, wherein the computer executable instructions, when executed by a computer processor, are used to execute the input response determination method as described in any one of the first aspects.

[0040] In a fifth aspect, the present application provides a computer program product, which includes computer instructions. When the computer instructions are executed on a computer, the computer executes the input response determination method provided in the first aspect.

[0041] It should be noted that the above-mentioned computer instructions may be stored in whole or in part on a computer-readable storage medium. The computer-readable storage medium may be packaged together with the processor of the input response determination device, or may be packaged separately from the processor of the input response determination device, and this application does not limit this.

[0042] The description of the second, third, fourth and fifth aspects in this application can refer to the detailed description of the first aspect; and the beneficial effects of the description of the second, third, fourth and fifth aspects can refer to the beneficial effect analysis of the first aspect, which will not be repeated here.

[0043] In this application, the name of the input response determination device does not limit the device or functional module itself. In actual implementation, these devices or functional modules may appear with other names. As long as the functions of each device or functional module are similar to those of this application, they fall within the scope of the claims of this application and their equivalent technologies.

[0044] These and other aspects of the present application will become more apparent from the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0046] Figure 1 A flowchart of an input response determination method provided by an embodiment of the present invention;

[0047] Figure 2 A flowchart of another input response determination method provided by an embodiment of the present invention;

[0048] Figure 3 A schematic diagram of an input response determination system provided by an embodiment of the present invention;

[0049] Figure 4 A schematic diagram of a model optimization page provided in an embodiment of the present invention;

[0050] Figure 5 A flowchart of another input response determination method provided by an embodiment of the present invention;

[0051] Figure 6 A flowchart of another input response determination method provided by an embodiment of the present invention;

[0052] Figure 7 A schematic diagram of the structure of an input response determination device provided by an embodiment of the present invention;

[0053] Figure 8 A schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0054] The present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. It is to be understood that the specific embodiments described herein are only used to explain the present invention, rather than to limit the present invention. It should also be noted that, for ease of description, only parts related to the present invention, rather than all structures, are shown in the accompanying drawings.

[0055] The term "and / or" in this article is merely a description of the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone.

[0056] The terms "first" and "second" and the like in the specification and drawings of this application are used to distinguish different objects, or to distinguish different processing of the same object, rather than to describe a specific order of objects.

[0057] In addition, the terms "including" and "having" and any variations thereof mentioned in the description of the present application are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but may optionally include other steps or units that are not listed, or may optionally include other steps or units that are inherent to these processes, methods, products or devices.

[0058] It should be mentioned before discussing exemplary embodiments in more detail that some exemplary embodiments are described as processes or methods depicted as flow charts. Although the flow charts describe various operations (or steps) as sequential processes, many operations therein can be implemented in parallel, concurrently or simultaneously. In addition, the order of various operations can be rearranged. The process can be terminated when its operation is completed, but can also have additional steps not included in the accompanying drawings. The process can correspond to methods, functions, procedures, subroutines, subprograms, etc. In addition, the embodiments in the present invention and the features in the embodiments can be combined with each other without conflict.

[0059] It should be noted that, in the embodiments of the present application, words such as "exemplary" or "for example" are used to indicate examples, illustrations or descriptions. Any embodiment or design described as "exemplary" or "for example" in the embodiments of the present application should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "exemplary" or "for example" is intended to present related concepts in a specific way.

[0060] In the description of the present application, unless otherwise specified, “plurality” means two or more.

[0061] Figure 1 This is a flow chart of an input response determination method provided by an embodiment of the present invention. This embodiment is applicable to situations where it is necessary to determine a relatively accurate input response to input information. The method can be executed by an input response determination device, such as Figure 1 As shown, the specific steps include:

[0062] Step 110: After receiving the training request, determine the target algorithm and target model according to the training request.

[0063] The training request is determined based on the optimization algorithm and base model selected by the user, the optimization algorithm may include a prompt optimization algorithm and an example optimization algorithm, and the base model may be a large model.

[0064] Specifically, after receiving the training request, since the training request is determined based on the optimization algorithm and base model selected by the user, the training request contains the algorithm identifier of the optimization algorithm selected by the user and the model identifier of the base model selected by the user. Furthermore, the target algorithm can be determined based on the algorithm identifier contained in the training request, and the target model can be determined based on the model identifier contained in the training request.

[0065] It should be noted that the user can select one or two optimization algorithms, so the target algorithm can be the prompt optimization algorithm and / or the example optimization algorithm.

[0066] In an embodiment of the present invention, a target algorithm and a target model corresponding to a received training request are determined, and the target algorithm can be used to optimize the target model.

[0067] Step 120: Process the use case data uploaded by the user based on the target algorithm to obtain prompt information and / or example information.

[0068] Specifically, after determining the target algorithm, the use case data uploaded by the user can be processed by the target algorithm to determine the prompt information and / or example information. For example, when the target algorithm is a prompt optimization algorithm, if the prompt optimization algorithm is a prompt word optimization algorithm, the use case data can be processed based on the prompt word optimization algorithm to obtain a prompt word, and the prompt word determined at this time is the prompt information; if the prompt optimization algorithm is a prompt vector optimization algorithm, the use case data can be processed based on the prompt vector optimization algorithm to obtain a prompt vector, and the prompt vector determined at this time is the prompt vector; when the target algorithm is an example optimization algorithm, the use case data can be processed based on the example optimization algorithm to obtain example information; when the target algorithm is a prompt optimization algorithm and an example optimization algorithm, the prompt word can be determined based on the prompt word optimization algorithm or the prompt vector can be determined based on the prompt vector optimization algorithm, and the example information can also be determined based on the example optimization algorithm.

[0069] In the embodiment of the present invention, the use case data is processed by the target algorithm determined above to determine prompt information and / or example information for making the model more compatible with the user's needs.

[0070] Step 130: input the input information and the prompt information and / or the example information into the target model, so that the target model determines an input response corresponding to the input information according to the prompt information and / or the example information.

[0071] Specifically, the user can also upload input information through the data upload area. After determining the prompt information and / or the example information, the input information and the prompt information and / or the example information can be input into the target model. By inputting the prompt information and / or the example information into the target model, low-cost optimization of the target model is achieved, so that the target model better meets the actual needs of the user. The target model can determine the input response corresponding to the input information based on the prompt information and / or the example information, and provide the target model with a data reference for determining the input response corresponding to the input information through the prompt information and / or the example information, so that the target model can determine a more accurate input response corresponding to the input information.

[0072] In an embodiment of the present invention, by inputting input information and prompt information and / or example information into a target model, low-cost optimization of the target model is achieved, and the target model can determine an input response corresponding to the input information with reference to the prompt information and / or example information, so that the target model can determine a more accurate input response corresponding to the input information.

[0073] The input response determination method provided by the embodiment of the present invention includes: after receiving a training request, determining a target algorithm and a target model according to the training request, wherein the training request is determined according to the optimization algorithm and the base model selected by the user; processing the use case data uploaded by the user based on the target algorithm to obtain prompt information and / or example information; inputting the input information and the prompt information and / or the example information into the target model, so that the target model determines the input response corresponding to the input information according to the prompt information and / or the example information. The above technical solution can first determine the target algorithm and the target model according to the training request determined according to the optimization algorithm and the base model selected by the user, and the target algorithm can be used to optimize the target model; secondly, the use case data uploaded by the user can be processed based on the target algorithm to determine the prompt information and / or example information used to make the model more compatible with the user's needs, and then the input information and the prompt information and / or example information can be input into the target model, and the target model can determine the input response corresponding to the input information with reference to the prompt information and / or example information. After the target model is optimized at a low cost through the prompt information and / or example information, the target model can determine a more accurate input response corresponding to the input information.

[0074] Figure 2 Flow chart of another input response determination method provided by an embodiment of the present invention. This embodiment is specific based on the above embodiment. Figure 2 As shown, in this embodiment, the method may further include:

[0075] Step 210: After receiving the training request, determine the target algorithm and target model according to the training request.

[0076] The training request is determined according to the optimization algorithm and base model selected by the user.

[0077] In one implementation, determining a target algorithm and a target model according to the training request includes:

[0078] The target algorithm is determined according to the algorithm identifier contained in the training request, wherein the algorithm identifier is determined according to the optimization control triggered by the user based on the algorithm selection page; the target model is determined according to the model identifier contained in the training request, wherein the model identifier is determined according to the model control triggered by the user based on the model selection page.

[0079] Figure 3 A schematic diagram of an input response determination system provided by an embodiment of the present invention, such as Figure 3 As shown, it includes a user terminal and a server, and the user terminal includes a model optimization page. Figure 4A schematic diagram of a model optimization page provided by an embodiment of the present invention, such as Figure 4 As shown, the model optimization page includes an algorithm selection page and a model selection page. The algorithm selection page includes optimization controls corresponding to each optimization algorithm. The optimization controls are used to select the optimization algorithm. For example, after the user clicks at least one optimization control, the optimization algorithm corresponding to the optimization control clicked by the user can be determined as the optimization algorithm selected by the user. The model selection page includes model controls corresponding to each base model. The model controls are used to select the base model. For example, after the user clicks any model control, the base model corresponding to the model control clicked by the user can be determined as the base model selected by the user.

[0080] The user can select the base model by clicking the model control corresponding to any base model on the model selection page. The user can select the optimization algorithm by clicking the optimization control corresponding to at least one optimization algorithm on the algorithm selection page. After the user selects the optimization algorithm and the base model, the training request can be triggered and sent to the server.

[0081] Therefore, the target algorithm can be determined according to the algorithm identifier included in the training request, and the target model can also be determined according to the model identifier included in the training request.

[0082] As mentioned above, the user can select one or two optimization algorithms, so the target algorithm can be the hint optimization algorithm and / or the example optimization algorithm.

[0083] In an embodiment of the present invention, a target algorithm and a target model corresponding to a received training request are determined, and the target algorithm can be used to optimize the target model.

[0084] Step 220: Process the use case data uploaded by the user based on the prompt optimization algorithm to obtain prompt information.

[0085] In one implementation, step 220 may specifically include:

[0086] The use case data is processed based on a prompt word search algorithm to obtain a prompt word.

[0087] The model optimization page also includes a data upload area, through which users can upload the use case data required for model optimization.

[0088] Specifically, when the prompt optimization algorithm is a prompt word optimization algorithm, the prompt word optimization algorithm is specifically a prompt word search algorithm, and the use case data uploaded by the user can be processed based on the prompt word search algorithm to obtain prompt words, which are prompt information. Specifically, a genetic prompt search algorithm (GPS) can be used to process the use case data, improve use case learning through prompt words, and automatically search for high-performance prompt words using a genetic algorithm, so that the generation model gradually changes the prompt words, and selects the final prompt words based on the performance of the prompt words on the use case data.

[0089] In another implementation manner, step 220 may specifically include:

[0090] The use case data is processed based on a prompt vector extraction algorithm to obtain a prompt vector.

[0091] Specifically, when the prompt optimization algorithm is a prompt vector optimization algorithm, the prompt vector optimization algorithm is specifically a prompt vector extraction algorithm, and the use case data can be processed based on the prompt vector extraction algorithm to obtain a prompt vector, which is the prompt information. Specifically, the use case data can be first processed based on the prompt word search algorithm to obtain a prompt word, and then the prompt word is vectorized to obtain a prompt vector.

[0092] In actual applications, the prompt information may also be determined according to the user input information.

[0093] In the embodiment of the present invention, the use case data is processed by a prompt optimization algorithm to determine prompt information for making the model more compatible with the user's needs.

[0094] Step 230: input the input information and the prompt word into the target model so that the target model determines the input response corresponding to the input information according to the prompt word; or, input the input information and the prompt vector into the target model, and adjust the prompt vector of the target model to the prompt vector so that the target model determines the input response corresponding to the input information according to the prompt vector.

[0095] Specifically, after determining the prompt information, the input information and the prompt information can be input into the target model. When the prompt information is a prompt word, the target model can determine the input response corresponding to the input information based on the prompt word. When the prompt information is a prompt vector, the target model can be fine-tuned based on the prompt vector, that is, the prompt vector can be added to each Transformer layer of the target model to achieve fine-tuning of the target model. The fine-tuned target model can determine the input response corresponding to the input information.

[0096] Figure 5Flow chart of another input response determination method provided by an embodiment of the present invention. This embodiment is specific based on the above embodiment. Figure 5 As shown, in this embodiment, the method may further include:

[0097] Step 510: After receiving the training request, determine the target algorithm and target model according to the training request.

[0098] Step 520: Process the use case data uploaded by the user based on the example optimization algorithm to obtain example information.

[0099] Specifically, the use case data can be processed based on the example optimization algorithm to obtain example information. Specifically, the distance or correlation degree between each use case data and the input information can be determined, and the use case data with the smallest distance can be determined as the example information, or the use case data with the largest correlation degree can be determined as the example information.

[0100] In practical applications, the example information may also be determined based on user input information.

[0101] In the embodiment of the present invention, the use case data is processed by the example optimization algorithm determined above to determine example information for making the model more compatible with user needs.

[0102] Step 530: Input the input information and the example information into the target model, so that the target model determines the input response corresponding to the input information according to the example information.

[0103] Specifically, after the example information is determined, the input information and the example information may be input into a target model, and the target model may determine an input response corresponding to the input information according to the example information.

[0104] Figure 6 Flow chart of another input response determination method provided by an embodiment of the present invention. This embodiment is specific based on the above embodiment. Figure 6 As shown, in this embodiment, the method may further include:

[0105] Step 610: After receiving the training request, determine the target algorithm and target model according to the training request.

[0106] Step 620: Process the use case data based on a prompt word search algorithm to obtain at least one prompt word, or process the use case data based on a prompt vector extraction algorithm to obtain at least one prompt vector; process the use case data based on an example optimization algorithm to obtain at least one example information.

[0107] The specific steps of determining the prompt word and prompt vector in this step are the same as the previous step 220.

[0108] When determining the example information, the distance or degree of association between each use case data and the input information can be determined, and each use case data is sorted based on the distance, or each use case data is sorted based on the degree of association, and the first N use case data with the smallest distance in the distance sorting result is determined as the example information, and these N use case data are sorted from small to large distances, or the first N use case data with the largest degree of association in the degree of association sorting result is determined as the example information, and these N use case data are sorted from large to small according to the degree of association. N is a natural number, and the specific value of N can be set according to actual needs, and is not specifically limited here.

[0109] Step 630: construct at least one group of combination information based on each of the example information and each of the prompt words or each of the prompt vectors; input the input information, the use case data and each group of the combination information into the target model, so that the target model determines the output use case corresponding to the input use case in the use case data according to each group of the combination information, determines the target combination information according to the actual output corresponding to the input use case in the use case data and the output use case corresponding to the output use case determined by each group of the combination information, and determines the input response corresponding to the input information according to the input information and the target combination information.

[0110] Specifically, after constructing at least one group of combination information according to each example information and each prompt word or each prompt vector, on the one hand, the use case information and each group of combination information can be input into the target model, so that the target model determines the output use case corresponding to the input use case in each use case data according to each group of combination information, and determines the target combination information according to the output use case corresponding to the input use case in each use case data determined by each group of combination information and the actual output, that is, the similarity between the output use case corresponding to the input use case in each use case data determined by each group of combination information and the actual output can be determined, and the total similarity corresponding to each group of combination information can be determined, that is, the sum of the similarities between the output use case corresponding to the input use case in all use case data determined by each group of combination information and the actual output can be determined as the total similarity corresponding to each group of combination information, and the combination information corresponding to the maximum total similarity can be determined as the target combination information. On the other hand, the current combination information can be determined in the combination information set composed of each combination information, the use case information and the current combination information can be input into the target model, and the output effect of the model can be determined. If the output effect does not meet the preset requirements, the current combination information can be determined in the combination information set composed of each combination information until the output effect meets the preset requirements, or, after iterating all the combination information, the target combination information can be obtained.

[0111] After determining the target combination information, the target model can determine an input response corresponding to the input information according to the input information and the target combination information.

[0112] In practical applications, the input information, prompt information and / or sample information can be added as waiting information to the waiting queue of the target model, and the target model can process the waiting information in the waiting queue in turn. Of course, if the waiting time of the waiting information exceeds the time threshold, the waiting information is deleted from the waiting queue.

[0113] It should be noted that during the model optimization process, the user can also send a query request to the server through the user terminal to query the optimization status of the model. After the model optimization is completed, the server can send the interface information of the target model to the user terminal, so that the user terminal can continue to access the target model based on the interface.

[0114] In an embodiment of the present invention, by inputting input information as well as prompt information and example information into a target model, low-cost optimization of the target model is achieved, and the target model can determine an input response corresponding to the input information with reference to the prompt information and the example information, so that the target model can determine a more accurate input response corresponding to the input information.

[0115] The input response determination method provided by an embodiment of the present invention includes: after receiving a training request, determining a target algorithm and a target model according to the training request; processing the use case data uploaded by the user based on a prompt optimization algorithm to obtain prompt information; or, processing the use case data uploaded by the user based on an example optimization algorithm to obtain example information; or, processing the use case data based on a prompt word search algorithm to obtain at least one prompt word, or, processing the use case data based on a prompt vector extraction algorithm to obtain at least one prompt vector, and processing the use case data based on an example optimization algorithm to obtain at least one example information; inputting the input information and the prompt information and / or the example information into the target model, so that the target model determines the input response corresponding to the input information according to the prompt information and / or the example information. The above technical scheme can first determine the target algorithm and target model based on the training request received according to the optimization algorithm and base model selected by the user. The target algorithm can be used to optimize the target model. Secondly, the use case data uploaded by the user can be processed based on the target algorithm to determine the prompt information and / or example information used to make the model more compatible with the user's needs. Then, the input information and the prompt information and / or example information can be input into the target model. The target model can determine the input response corresponding to the input information with reference to the prompt information and / or example information. After the target model is optimized at low cost through the prompt information and / or example information, the target model can determine a more accurate input response corresponding to the input information.

[0116] Figure 7This is a schematic diagram of the structure of an input response determination device provided by an embodiment of the present invention, which can be applied to the case where the input response of the input information is relatively accurate. The device can be implemented by software and / or hardware and is generally integrated in an electronic device, such as a computer device.

[0117] like Figure 7 As shown, the device comprises:

[0118] A determination module 710 is configured to determine a target algorithm and a target model according to a training request after receiving the training request, wherein the training request is determined according to an optimization algorithm and a base model selected by a user;

[0119] The processing module 720 is used to process the use case data uploaded by the user based on the target algorithm to obtain prompt information and / or example information;

[0120] The response module 730 is used to input the input information and the prompt information and / or the example information into the target model, so that the target model determines the input response corresponding to the input information according to the prompt information and / or the example information.

[0121] The input response determination device provided in this embodiment, after receiving a training request, determines a target algorithm and a target model according to the training request, wherein the training request is determined according to the optimization algorithm and the base model selected by the user; processes the use case data uploaded by the user based on the target algorithm to obtain prompt information and / or example information; inputs the input information and the prompt information and / or the example information into the target model, so that the target model determines the input response corresponding to the input information according to the prompt information and / or the example information. The above technical scheme can first determine the target algorithm and the target model according to the training request determined according to the optimization algorithm and the base model selected by the user, and the target algorithm can be used to optimize the target model; secondly, the use case data uploaded by the user can be processed based on the target algorithm to determine the prompt information and / or example information used to make the model more compatible with the user's needs, and then the input information and the prompt information and / or example information can be input into the target model, and the target model can determine the input response corresponding to the input information with reference to the prompt information and / or example information, and after the target model is optimized at a low cost by the prompt information and / or example information, the target model can determine a more accurate input response corresponding to the input information.

[0122] Based on the above embodiment, the determination module 710 is specifically used for:

[0123] Determining the target algorithm according to the algorithm identifier included in the training request, wherein the algorithm identifier is determined according to an optimization control triggered by the user based on an algorithm selection page;

[0124] The target model is determined according to a model identifier included in the training request, wherein the model identifier is determined according to a model control triggered by the user based on a model selection page.

[0125] On the basis of the above embodiment, when the target algorithm is a prompt optimization algorithm, the processing module 720 is specifically used to:

[0126] Processing the use case data based on a prompt word search algorithm to obtain a prompt word; or processing the use case data based on a prompt vector extraction algorithm to obtain a prompt vector;

[0127] Accordingly, the response module 730 is specifically configured to:

[0128] The input information and the prompt word are input into the target model so that the target model determines the input response corresponding to the input information according to the prompt word; or, the input information and the prompt vector are input into the target model, and the prompt vector of the target model is adjusted to the prompt vector so that the target model determines the input response corresponding to the input information according to the prompt vector.

[0129] On the basis of the above embodiment, when the target algorithm is a prompt optimization algorithm, the processing module 720 is specifically used to:

[0130] Process the use case data based on an example optimization algorithm to obtain example information;

[0131] Accordingly, the response module 730 is specifically configured to:

[0132] The input information and the example information are input into the target model, so that the target model determines the input response corresponding to the input information according to the example information.

[0133] On the basis of the above embodiment, when the target algorithm is the prompt optimization algorithm and the example optimization algorithm, the processing module 720 is specifically used to:

[0134] Processing the use case data based on a prompt word search algorithm to obtain at least one prompt word, or processing the use case data based on a prompt vector extraction algorithm to obtain at least one prompt vector;

[0135] Accordingly, the response module 730 is specifically configured to:

[0136] Construct at least one set of combined information according to each of the prompt information and each of the prompt words or each of the prompt vectors;

[0137] The input information, the use case data and each group of the combination information are input into the target model, so that the target model determines the output use case corresponding to the input use case in the use case data according to each group of the combination information, determines the target combination information according to the actual output corresponding to the input use case in the use case data and the output use case corresponding to the output use case determined by each group of the combination information, and determines the input response corresponding to the input information according to the input information and the target combination information.

[0138] The input response determination device provided in the embodiment of the present invention can execute the input response determination method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of executing the input response determination method.

[0139] It is worth noting that in the embodiment of the above-mentioned input response determination device, the various units and modules included are only divided according to functional logic, but are not limited to the above-mentioned division, as long as the corresponding functions can be achieved; in addition, the specific names of the functional units are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of the present invention.

[0140] Figure 8 A schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. Figure 8 A block diagram of an exemplary electronic device 8 suitable for use in implementing embodiments of the present invention is shown. Figure 8 The electronic device 8 shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present invention.

[0141] like Figure 8 As shown, the electronic device 8 is in the form of a general purpose computing electronic device. The components of the electronic device 8 may include, but are not limited to: one or more processors or processing units 16, a system memory 28, and a bus 18 connecting different system components (including the system memory 28 and the processing unit 16).

[0142] Bus 18 represents one or more of several types of bus structures, including a memory bus or memory controller, a peripheral bus, an accelerated graphics port, a processor or a local bus using any of a variety of bus architectures. By way of example, these architectures include, but are not limited to, an Industry Standard Architecture (ISA) bus, a Micro Channel Architecture (MAC) bus, an Enhanced ISA bus, a Video Electronics Standards Association (VESA) local bus, and a Peripheral Component Interconnect (PCI) bus.

[0143] The electronic device 8 typically includes a variety of computer system readable media. These media can be any available media that can be accessed by the electronic device 8, including volatile and non-volatile media, removable and non-removable media.

[0144] The system memory 28 may include computer system readable media in the form of volatile memory, such as random access memory (RAM) 30 and / or cache memory 32. The electronic device 8 may further include other removable / non-removable, volatile / non-volatile computer system storage media. By way of example only, the storage system 34 may be used to read and write non-removable, non-volatile magnetic media ( Figure 8 not shown, usually called a "hard drive"). Although Figure 8 Not shown in the figure, a disk drive for reading and writing to a removable non-volatile disk (e.g., a "floppy disk"), and an optical disk drive for reading and writing to a removable non-volatile optical disk (e.g., a CD-ROM, a DVD-ROM, or other optical media) may be provided. In these cases, each drive may be connected to the bus 18 via one or more data medium interfaces. The system memory 28 may include at least one program product having a set (e.g., at least one) of program modules that are configured to perform the functions of the various embodiments of the present invention.

[0145] A program / utility 40 having a set (at least one) of program modules 42 may be stored, for example, in system memory 28, such program modules 42 including, but not limited to, an operating system, one or more application programs, other program modules, and program data, each of which or some combination may include an implementation of a network environment. Program modules 42 generally perform the functions and / or methods of the embodiments described herein.

[0146] The electronic device 8 may also communicate with one or more external devices 14 (e.g., keyboards, pointing devices, displays 24, etc.), one or more devices that enable a user to interact with the electronic device 8, and / or any device that enables the electronic device 8 to communicate with one or more other computing devices (e.g., network cards, modems, etc.). Such communication may be performed via an input / output (I / O) interface 22. Furthermore, the electronic device 8 may also communicate with one or more networks (e.g., local area networks (LANs), wide area networks (WANs), and / or public networks, such as the Internet) via a network adapter 20. Figure 8 As shown, the network adapter 20 communicates with other modules of the electronic device 8 via the bus 18. It should be understood that although Figure 8 Not shown, other hardware and / or software modules may be used in conjunction with electronic device 8, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.

[0147] The processing unit 16 executes various functional applications and page displays by running the program stored in the system memory 28, for example, implementing the input response determination method provided in the embodiment of the present invention, which includes:

[0148] After receiving the training request, determining the target algorithm and the target model according to the training request, wherein the training request is determined according to the optimization algorithm and the base model selected by the user;

[0149] Processing the use case data uploaded by the user based on the target algorithm to obtain prompt information and / or example information;

[0150] Input information and the prompt information and / or the example information are input into the target model, so that the target model determines an input response corresponding to the input information according to the prompt information and / or the example information.

[0151] Of course, those skilled in the art can understand that the processor can also implement the technical solution of the input response determination method provided by any embodiment of the present invention.

[0152] An embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon. When the program is executed by a processor, the input response determination method provided by an embodiment of the present invention is implemented, for example, and the method includes:

[0153] After receiving the training request, determining the target algorithm and the target model according to the training request, wherein the training request is determined according to the optimization algorithm and the base model selected by the user;

[0154] Processing the use case data uploaded by the user based on the target algorithm to obtain prompt information and / or example information;

[0155] Input information and the prompt information and / or the example information are input into the target model, so that the target model determines an input response corresponding to the input information according to the prompt information and / or the example information.

[0156] The computer storage medium of the embodiment of the present invention can adopt any combination of one or more computer-readable media. The computer-readable medium can be a computer-readable signal medium or a computer-readable storage medium. The computer-readable storage medium can be, for example, but not limited to: an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples (non-exhaustive list) of computer-readable storage media include: an electrical connection with one or more wires, a portable computer 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 disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this document, a computer-readable storage medium can be any tangible medium containing or storing a program, which can be used by an instruction execution system, device or device or used in combination with it.

[0157] Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, which carry computer-readable program code. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer-readable signal media may also be any computer-readable medium other than a computer-readable storage medium, which may send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device.

[0158] The program code embodied on the computer readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.

[0159] Computer program code for performing the operations of the present invention may be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, 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 may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0160] It should be understood by those skilled in the art that the modules or steps of the present invention described above can be implemented by a general-purpose computing device, they can be concentrated on a single computing device, or distributed on a network composed of multiple computing devices, optionally, they can be implemented by a program code executable by a computer device, so that they can be stored in a storage device and executed by the computing device, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. Thus, the present invention is not limited to any specific combination of hardware and software.

[0161] In addition, the acquisition, storage, use, and processing of data in the technical solution of the present invention are in compliance with the relevant provisions of national laws and regulations.

[0162] Note that the above are only preferred embodiments of the present invention and the technical principles used. Those skilled in the art will understand that the present invention is not limited to the specific embodiments herein, and that various obvious changes, readjustments and substitutions can be made by those skilled in the art without departing from the scope of protection of the present invention. Therefore, although the present invention has been described in more detail through the above embodiments, the present invention is not limited to the above embodiments, and may include more other equivalent embodiments without departing from the concept of the present invention, and the scope of the present invention is determined by the scope of the appended claims.

Claims

1. A method for determining an input response, characterized in that: include: After receiving the training request, determining the target algorithm and the target model according to the training request, wherein the training request is determined according to the optimization algorithm and the base model selected by the user; Processing the use case data uploaded by the user based on the target algorithm to obtain prompt information and / or example information; Input information and the prompt information and / or the example information are input into the target model, so that the target model determines an input response corresponding to the input information according to the prompt information and / or the example information.

2. The input response determination method according to claim 1, characterized in that: Determining a target algorithm and a target model according to the training request includes: Determining the target algorithm according to the algorithm identifier included in the training request, wherein the algorithm identifier is determined according to an optimization control triggered by the user based on an algorithm selection page; The target model is determined according to a model identifier included in the training request, wherein the model identifier is determined according to a model control triggered by the user based on a model selection page.

3. The input response determination method according to claim 1, characterized in that: When the target algorithm is a prompt optimization algorithm, the use case data uploaded by the user is processed based on the target algorithm to obtain prompt information and / or example information, including: Process the use case data based on a prompt word search algorithm to obtain a prompt word; Accordingly, inputting the input information and the prompt information and / or the example information into the target model so that the target model determines an input response corresponding to the input information according to the prompt information and / or the example information, includes: The input information and the prompt word are input into the target model, so that the target model determines the input response corresponding to the input information according to the prompt word.

4. The input response determination method according to claim 1, characterized in that: When the target algorithm is a prompt optimization algorithm, the use case data uploaded by the user is processed based on the target algorithm to obtain prompt information and / or example information, including: Processing the use case data based on a prompt vector extraction algorithm to obtain a prompt vector; Accordingly, inputting the input information and the prompt information and / or the example information into the target model so that the target model determines an input response corresponding to the input information according to the prompt information and / or the example information, includes: The input information and the prompt vector are input into the target model, and the prompt vector of the target model is adjusted to the prompt vector, so that the target model determines the input response corresponding to the input information according to the prompt vector.

5. The input response determination method according to claim 1, characterized in that: When the target algorithm is an example optimization algorithm, the use case data uploaded by the user is processed based on the target algorithm to obtain prompt information and / or example information, including: Process the use case data based on an example optimization algorithm to obtain example information; Accordingly, inputting the input information and the prompt information and / or the example information into the target model so that the target model determines an input response corresponding to the input information according to the prompt information and / or the example information, includes: The input information and the example information are input into the target model, so that the target model determines the input response corresponding to the input information according to the example information.

6. The input response determination method according to claim 1, characterized in that: When the target algorithm is a prompt optimization algorithm and an example optimization algorithm, the use case data uploaded by the user is processed based on the target algorithm to obtain prompt information and / or example information, including: Processing the use case data based on a prompt word search algorithm to obtain at least one prompt word, or processing the use case data based on a prompt vector extraction algorithm to obtain at least one prompt vector; Process the use case data based on an example optimization algorithm to obtain at least one example information; Accordingly, inputting the input information and the prompt information and / or the example information into the target model so that the target model determines an input response corresponding to the input information according to the prompt information and / or the example information, includes: Construct at least one set of combined information according to each of the example information and each of the prompt words or each of the prompt vectors; The input information, the use case data and each group of the combination information are input into the target model, so that the target model determines the output use case corresponding to the input use case in the use case data according to each group of the combination information, determines the target combination information according to the actual output corresponding to the input use case in the use case data and the output use case corresponding to the output use case determined by each group of the combination information, and determines the input response corresponding to the input information according to the input information and the target combination information.

7. An input response determination device, characterized in that: include: A determination module, configured to determine a target algorithm and a target model according to a training request after receiving the training request, wherein the training request is determined according to an optimization algorithm and a base model selected by a user; A processing module, used for processing the use case data uploaded by the user based on the target algorithm to obtain prompt information and / or example information; The response module is used to input the input information and the prompt information and / or the example information into the target model, so that the target model determines the input response corresponding to the input information according to the prompt information and / or the example information.

8. An electronic device, characterized in that: The electronic device comprises: at least one processor; and a memory communicatively coupled to the at least one processor; The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the input response determination method as described in any one of claims 1-6.

9. A storage medium containing computer executable instructions, characterized in that: The computer executable instructions are used to perform the input response determination method according to any one of claims 1 to 6 when executed by a computer processor.

10. A computer program product, characterized in that The computer program product comprises a computer program, which, when executed by a processor, implements the input response determination method according to any one of claims 1 to 6.

Images