A large model management method, device and storage medium based on cloud services

Abstract

The application discloses a large model management method and device based on cloud services, electronic equipment and storage medium, and relates to the field of financial technology, and the method comprises the following steps: deploying a first large model to the cloud according to a deployment instruction of the first large model sent by a first user, and showing an interactive configuration interface to the first user; obtaining a first interactive interface of the first large model through the interactive configuration interface, and deploying the first interactive interface to the local; in response to obtaining a calling instruction of the first large model sent by a second user, showing the first interactive interface to the second user, so as to call the first large model through the first interactive interface. The technical scheme of the embodiment of the application not only realizes the deployment and application of the large model, but also expands the application range of the large model and improves the utilization rate of the large model. Meanwhile, the data security of the interactive interface itself is improved, and the calling security of the large model calling process is ensured.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of financial technology, and in particular to a big model management method and device based on cloud services, electronic equipment and storage medium. BACKGROUND

[0002] With the continuous development of artificial intelligence big model (Artificial Intelligence Big Model, referred to as "big model") technology, financial institutions often develop a variety of big models suitable for different needs, so effective management of big models has become an important issue for financial institutions.

[0003] In the prior art, the use of big models is often limited to a small number of professional technical personnel, because big models require both professional technical knowledge for training and application and the support of high-performance computers, and ordinary business personnel often lack access to the use of big models due to lack of these resources and technology.

[0004] In addition, the developers and users of big models are usually not the same group, which leads to the phenomenon that developers do not use frequently and users do not understand how to use, thereby the utilization rate of big models is low and the use range is small. SUMMARY

[0005] The present application provides a big model management method and device based on cloud services, electronic equipment and storage medium to solve the problem of low utilization rate and small use range of big models.

[0006] According to an aspect of the present application, a big model management method based on cloud services is provided, comprising:

[0007] According to the deployment instruction of the first big model issued by the first user, the first big model is deployed on the cloud, and an interactive configuration interface is displayed to the first user;

[0008] The first interactive interface of the first big model is obtained through the interactive configuration interface, and the first interactive interface is deployed locally;

[0009] In response to obtaining the calling instruction of the first big model issued by the second user, the first interactive interface is displayed to the second user to call the first big model through the first interactive interface.

[0010] The cloud service-based large model management method further includes: in response to obtaining a first supplementary sample for the first large model from a third user, sending the first supplementary sample to the first user to confirm whether to use the first supplementary sample as a training sample of the first large model through the first user; and if a confirmation instruction for the first supplementary sample is obtained from the first user, iteratively training the first large model based on the first supplementary sample.

[0011] After the first supplementary sample is sent to the first user to confirm whether to use the first supplementary sample as a training sample of the first large model through the first user, the method further includes: if a rejection instruction for the first supplementary sample is obtained from the first user, copying a second large model from the first large model, deploying the second large model on a cloud, and iteratively training the second large model based on the first supplementary sample; and in label information of the second large model, the source model and the upstream model are both configured as the first large model.

[0012] The cloud service-based large model management method further includes: in response to obtaining a fuzzy calling instruction issued by a fourth user, determining a target large model matched with the fourth user according to identity information of the fourth user, and showing a target interaction interface corresponding to the target large model to the fourth user to call the target large model through the target interaction interface.

[0013] The first large model includes an evaluation management model; after the first large model is called through the first interaction interface, the method further includes: obtaining a first filtering instruction of the second user through the first interaction interface; wherein the first filtering instruction includes at least one of comment content, evaluation time, performance object, and emotion category; and according to the first filtering instruction, each evaluation information is sequentially displayed based on a preset priority of other display parameters except the filtering parameter.

[0014] The first large model includes a customer management model; after the first large model is called through the first interaction interface, the method further includes: obtaining a second filtering instruction of the second user through the first interaction interface; wherein the second filtering instruction includes at least one of customer number, economic status, credit rating, and education information; and according to the second filtering instruction, each customer information is sequentially displayed based on a preset priority of other display parameters except the filtering parameter.

[0015] According to another aspect of the present application, a cloud service-based large model management device is provided, comprising:

[0016] a model deployment execution module configured to deploy, according to a deployment instruction of a first large model issued by a first user, the first large model to a cloud and show an interactive configuration interface to the first user;

[0017] an interface deployment execution module configured to acquire, through the interactive configuration interface, a first interactive interface of the first large model and deploy the first interactive interface to a local device;

[0018] a model calling execution module configured to, in response to acquiring a calling instruction of the first large model issued by a second user, show the first interactive interface to the second user to call the first large model through the first interactive interface.

[0019] According to another aspect of the present application, an electronic device is provided, which comprises:

[0020] at least one processor; and

[0021] a memory connected to the at least one processor in communication; wherein

[0022] 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 to enable the at least one processor to execute the cloud service-based large model management method according to any of the embodiments of the present application.

[0023] According to another aspect of the present application, a computer readable storage medium is provided, which stores computer instructions for enabling a processor to implement the cloud service-based large model management method according to any of the embodiments of the present application when executed by the processor.

[0024] According to another aspect of the present application, a computer program product is provided, which comprises a computer program for implementing the cloud service-based large model management method according to any of the embodiments of the present application when executed by a processor.

[0025] The technical solution of the embodiments of the present application deploys, according to a deployment instruction of a first large model issued by a first user, the first large model to a cloud and shows an interactive configuration interface to the first user; acquires, through the interactive configuration interface, a first interactive interface of the first large model and deploys the first interactive interface to a local device; and in response to acquiring a calling instruction of the first large model issued by a second user, shows the first interactive interface to the second user to call the first large model through the first interactive interface. Thus, not only the deployment and application of the large model are realized, but also the application range of the large model is expanded and the utilization rate of the large model is improved; meanwhile, the large model deployed to the cloud and the interactive interface deployed to the local device not only improve the data security of the interactive interface itself, but also ensure the calling security in the calling process of the large model.

[0026] It should be understood that the matters described in this detailed description are intended to be illustrative and are not intended to limit or restrict the scope of the embodiments of the present application. Other features of the present application will become apparent to those skilled in the art upon a reading of the following specification. BRIEF DESCRIPTION OF DRAWINGS

[0027] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed to be used in the embodiments description will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative effort based on these drawings.

[0028] Figure 1 is a flow chart of a large model management method based on cloud service provided according to an embodiment of the present application;

[0029] Figure 2 is a flow chart of another large model management method based on cloud service provided according to an embodiment of the present application;

[0030] Figure 3 is a flow chart of still another large model management method based on cloud service provided according to an embodiment of the present application;

[0031] Figure 4 is a structural schematic diagram of a large model management device based on cloud service provided according to an embodiment of the present application;

[0032] Figure 5 is a structural schematic diagram of an electronic device implementing the large model management method based on cloud service of the embodiments of the present application. DETAILED DESCRIPTION

[0033] In order to make the technical personnel in the art better understand the present application scheme, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative effort should be within the scope of protection of the present application.

[0034] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and in the above-described drawings are used to distinguish similar objects, and do not necessarily have to be used to describe a specific order or sequence. It should be understood that the data thus used can be interchanged under appropriate circumstances, so that the embodiments of the application described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion, for example, a process, method, system, product or device that includes a series of steps or units does not have to be limited to only those steps or units clearly listed, but can include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0035] Embodiment one

[0036] Figure 1 A flowchart of a large model management method based on cloud service provided for the first embodiment of the present application, the present embodiment can be applicable to the deployment and application of large models based on cloud service, the method can be executed by the large model management device based on cloud service in any embodiment of the present application, the large model management device based on cloud service can be realized in the form of hardware and / or software, and the large model management device based on cloud service can be configured in a management system, and the management system can be configured in one or more servers. As shown in the figure, the method comprises: Figure 1

[0037] S101, according to the deployment instruction of the first large model issued by the first user, deploying the first large model on the cloud, and showing the first user an interactive configuration interface.

[0038] After the developer completes the design and development of the large model, the encapsulated large model is uploaded to the management system, and the management system deploys the obtained large model on the cloud to facilitate the use and maintenance of the large model; according to the different business types and business objects suitable for the large model, the display parameters in the interactive interface of each large model are often different, and the display parameters actually exist as filtering parameters, so after the cloud deployment of the first large model is completed, the first user needs to be shown an interactive configuration interface to guide the first user to complete the configuration of the interactive interface of the first large model through the interactive configuration interface, which is actually the configuration of the display parameters in the interactive interface.

[0039] ​Taking the first large model as an evaluation management model as an example, the model can be used to obtain evaluation information of a user for a mobile phone software, and classify and manage each evaluation information; wherein the display parameters of the evaluation information can include performance objects, comment time, comment content and emotion categories; the performance objects can include usability, ease of use, robustness, interface aesthetics, process and security, which reflect which performance of the mobile phone software the current comment information is evaluated for.

[0040] The emotion categories can include positive, neutral and negative; positive means that the current comment information reflects that the user expresses favor for a certain performance, that is, has identity for the performance; neutral means that the current comment information reflects that the user has neither positive nor negative attitude towards the performance, has no special favor nor special aversion; negative means that the current comment information reflects that the user expresses aversion to the performance, that is, does not identify the performance.

[0041] Taking the first large model as a customer management model as an example, the model can be used to manage customer information; wherein the display parameters of the customer information can include customer number, economic status, life cycle, credit level and education information, etc.; wherein the life cycle can include active period, dormant period and loss period, etc., which reflect the business use state of the current customer; thus the display parameters in the interactive interface of each deployed large model can be configured by the developer of the model; in addition, the model developer can also choose to automatically configure the interactive interface to reduce the configuration complexity, and the management system extracts one or more parameters as display parameters based on the display space of the interactive interface according to the feedback information of the large model.

[0042] S102, obtain the first interactive interface of the first large model through the interactive configuration interface, and deploy the first interactive interface locally.

[0043] Compared with the large model itself, the code amount of the interactive interface is less, and compared with the iterative update of the large model, the code information of the interactive interface is more stable, so deploying the interactive interface locally will not occupy too much storage resource, and only a little expansion space is needed; at the same time, the large model provides an interface, and the interactive interface is the large model call based on the interface; for example, the large model can perform identity authentication through the interface identifier of the interactive interface, and only allow the matched interactive interface to call it; therefore, storing the interactive interface locally not only improves the data security of the interactive interface itself, but also has good stability and controllability, and ensures the call security of the large model call process.

[0044] S103, in response to obtaining the calling instruction of the first large model issued by the second user, the first interaction interface is displayed to the second user to call the first large model through the first interaction interface.

[0045] After the first large model and its interaction interface (i.e. the first interaction interface) are deployed, they can be opened to each user with use permission; when the management system obtains the calling instruction of the first large model issued by the second user, the first interaction interface is displayed to the second user to call the first large model through the first interaction interface, and the operation instruction of the second user is obtained through the first interaction interface, so as to meet the use demand of the second user for the large model.

[0046] In addition, the management system also sends the feedback result of each user and the model performance data to the developer of the large model, so that the developer optimizes the model to improve the accuracy and efficiency of the large model; at the same time, the management system can update the large model according to the update instruction issued by the developer to realize the improvement of problem repair, new function addition and performance improvement; wherein, the performance data monitored by the management system in real time can include accuracy and response time, so as to guide the developer to complete repair in time when the performance of the large model decreases or abnormal condition occurs.

[0047] Optionally, in the embodiment of the application, the large model management method based on cloud service further comprises: in response to obtaining the first supplementary sample of the first large model issued by the third user, sending the first supplementary sample to the first user to confirm whether the first supplementary sample is used as the training sample of the first large model through the first user; if the confirmation instruction of the first user for the first supplementary sample is obtained, the first large model is iteratively trained through the first supplementary sample.

[0048] Specifically, the deployed large model sometimes cannot meet the use demand of the user, or the user wants to strengthen one or more functions of the large model, at this time, the user (i.e. the third user) can construct the supplementary sample meeting the demand of the user himself, and upload the supplementary sample to the management system; the management system sends the above-mentioned supplementary sample to the creator (i.e. the first user) of the large model to confirm whether the first supplementary sample is used as the training sample of the first large model through the first user; if the rejection instruction is obtained, the first large model is iteratively trained using the first supplementary sample to ensure that the first large model maintains the original model structure.

[0049] If the obtained is a confirmation instruction, the first large model is iteratively trained using the first supplementary sample, thereby realizing iterative update of the deployed large model based on actual needs of the user, and the inquiry mechanism finally confirmed by the developer ensures functional reinforcement of the large model and avoids arbitrary modification of the model structure, thereby avoiding problems such as deviation of the model function from the original construction intention of the developer, and ensuring functional safety of the large model.

[0050] Optionally, in the embodiment of the present application, after the first supplementary sample is sent to the first user to confirm whether the first supplementary sample is used as the training sample of the first large model, the method further comprises: if a rejection instruction of the first user for the first supplementary sample is obtained, a second large model is obtained by copying the first large model, the second large model is deployed on the cloud, and the second large model is iteratively trained by using the first supplementary sample; and in the label information of the second large model, the source model and the upstream model are both configured as the first large model.

[0051] Specifically, the management system can determine whether the current user has the permission to deploy the large model according to the different permissions of the user; if the current user does not have the permission to deploy the large model, the training operation of the first supplementary sample can be ignored after the rejection instruction of the developer is obtained; if the current user has the permission to deploy the large model, a second large model copied from the first large model is obtained, the second large model is deployed on the cloud, and the second large model is iteratively trained by using the first supplementary sample to meet the actual needs of the third user.

[0052] Meanwhile, the second large model can also have a large applicable range, which can be set as a shared permission by the third user for common use by multiple users; the second large model can also only meet the specific needs of the third user, and has a small applicable range, which can be set as an exclusive permission by the third user to avoid interference with the use of other users; in addition, since the second large model is transformed from the first large model, the source information needs to be marked in the label information of the second large model to facilitate tracing of the structure and function of the second large model.

[0053] The upstream model indicates a model directly converted from which the second large model is, that is, a model at a previous level of the second large model; the source model indicates an initial model at a source of the current large model after a series of model copying, that is, a starting point model; for example, a large model B is obtained based on a large model A, and a large model C is obtained based on the large model B, so the large model B is the upstream model of the large model C, and the large model A is the source model of the large model C; obviously, in the embodiment of the application, the source model and the upstream model of the second large model are both the first large model. In this way, the model use requirements of the second user are ensured, the original structure of the first large model is maintained, and the source tracing of the second large model is realized.

[0054] Optionally, in the embodiment of the application, the large model management method based on cloud service further includes: in response to obtaining a fuzzy calling instruction issued by a fourth user, determining a matched target large model according to identity information of the fourth user, and displaying a target interaction interface corresponding to the target large model to the fourth user to call the target large model through the target interaction interface. For a user who is not familiar with the functions of various large models, the user cannot know the specific functions of each large model and cannot accurately call the required large model, so if the fourth user does not issue a calling instruction for a certain large model but issues a fuzzy calling instruction, the management system can compare the identity information of the fourth user with the tag information of each deployed large model to obtain a matched target large model.

[0055] For example, the A user is a business personnel of B type business, so the large model of B type business can be recommended to the A user in the business type of the tag information of each large model, and then the model selection result of the A user is obtained. In this way, after the target large model is determined, the corresponding target interaction interface is displayed to the fourth user to call the target large model through the target interaction interface, so that the user can achieve fuzzy matching of the large model without being familiar with the structure and function of the deployed large model, the user experience is improved, and the model use requirements of the user are met.

[0056] The technical scheme of the embodiment of the application deploys the first large model on the cloud according to the deployment instruction of the first large model issued by the first user, and displays an interaction configuration interface to the first user; obtains the first interaction interface of the first large model through the interaction configuration interface, and deploys the first interaction interface on the local; in response to obtaining a calling instruction of the first large model issued by the second user, displays the first interaction interface to the second user to call the first large model through the first interaction interface. In this way, not only the deployment and application of the large model are realized, but also the application range of the large model is expanded, and the utilization rate of the large model is improved; at the same time, the large model deployed on the cloud and the interaction interface deployed on the local not only improve the data security of the interaction interface itself, but also ensure the calling security of the calling process of the large model.

[0057] Embodiment Two

[0058] Figure 2 A flowchart of a large model management method based on cloud services is provided for Embodiment Two of the present application. The relationship between this embodiment and the above-mentioned embodiments is that the first large model is taken as an example of an evaluation management model. As shown in Figure 2 the method comprises:

[0059] S201, deploying the first large model on the cloud according to a deployment instruction of the first large model issued by a first user, and showing an interactive configuration interface to the first user.

[0060] S202, obtaining a first interactive interface of the first large model through the interactive configuration interface, and deploying the first interactive interface on the local;

[0061] S203, in response to obtaining a calling instruction of the first large model issued by a second user, showing the first interactive interface to the second user to call the first large model through the first interactive interface.

[0062] S204, obtaining a first filtering instruction of the second user through the first interactive interface; wherein the first filtering instruction includes at least one of comment content, evaluation time, performance object and emotion category.

[0063] S205, according to the first filtering instruction, sequentially displaying each evaluation information based on the preset priority of the other display parameters except the filtering parameters.

[0064] In the interactive interface of the evaluation management model, fixed general function buttons are displayed, including the function buttons of adding, deleting, downloading, exporting, querying and resetting displayed on the upper end of the interface, the view and edit buttons configured at the end of each evaluation information, and the display page configuration button, single page quantity configuration button and jump page button configured at the lower end of the interface; the management system performs corresponding processing operations according to the function buttons triggered by the second user in the interactive interface.

[0065] In addition to the general function buttons, according to the display parameters configured by the first user in the interactive interface when the first large model is deployed on the cloud, the interactive interface also displays the parameter items specific to the evaluation information, i.e. comment content, evaluation time, performance object and emotion category; the above-mentioned display parameters are also the filtering parameters of the evaluation information; the first user can select one or more of the comment content, evaluation time, performance object and emotion category as the filtering parameters and complete the filtering parameter configuration in the corresponding text box in a manual input or manual selection manner.

[0066] After obtaining the first screening instruction issued by the second user, first, the matched evaluation information can be obtained according to the screening parameter in the first screening instruction; then, according to the preset priority of other display parameters other than the screening parameter, each piece of evaluation information is displayed in turn; wherein, the preset priority represents the ordering priority between different display parameters, and also represents the ordering priority of each sub-parameter in each display parameter, which can be pre-configured by the developer of the first large model, or configured by the management system based on the historical screening conditions of multiple users, the more times the current display parameter or display sub-parameter is used as a screening parameter, the higher the preset priority of the current display parameter or display sub-parameter.

[0067] For example, the second user takes the evaluation time and the emotion category as the screening parameter, at this time, the performance object and the evaluation content are taken as other display parameters other than the screening parameter, and the priority of the performance object is higher than that of the evaluation content, so the sorting is based on the performance object; and in the performance object, the priority of the negative, neutral and positive is reduced in turn, and the evaluation information under each display sub-parameter is displayed according to the above ordering mode.

[0068] The technical scheme of the embodiment of the application obtains the first screening instruction including at least one of the comment content, the evaluation time, the performance object and the emotion category issued by the second user through the first interaction interface; then, according to the first screening instruction, each piece of evaluation information is displayed in turn based on the preset priority of other display parameters other than the screening parameter. In this way, not only the screening and obtaining of the evaluation information are realized, but also the ordering of the evaluation information based on the preset priority is realized, which not only meets the evaluation information screening demand of the user, but also ensures the priority display of the expected evaluation information.

[0069] Embodiment three

[0070] Figure 3 A flowchart of a large model management method based on cloud service provided by the third embodiment of the application, the relationship between the present embodiment and the above-mentioned embodiments is that the first large model is taken as a customer management model. As shown in the figure, Figure 3 The method comprises the following steps:

[0071] S301, according to the deployment instruction of the first large model issued by the first user, the first large model is deployed in the cloud, and an interactive configuration interface is displayed to the first user.

[0072] S302, the first interaction interface of the first large model is obtained through the interactive configuration interface, and the first interaction interface is deployed in the local;

[0073] S303, in response to obtaining the calling instruction of the first large model issued by the second user, the first interaction interface is displayed to the second user, so as to call the first large model through the first interaction interface.

[0074] S304, obtaining a second filtering instruction of the second user through the first interactive interface; wherein the second filtering instruction comprises at least one of customer number, economic status, credit rating and educational information.

[0075] S305, according to the second filtering instruction, sequentially displaying each customer information based on the preset priority of the display parameter other than the filtering parameter.

[0076] In the interactive interface of the customer management model, fixed general function buttons are also displayed, including the function buttons of adding, deleting, downloading, exporting, querying and resetting displayed on the upper end of the interface, the view and edit buttons arranged at the end of each customer information, and the display page configuration button, single page quantity configuration button and jump page button arranged at the lower end of the interface; the management system performs corresponding processing operation according to the function button triggered by the second user in the interactive interface.

[0077] In addition to the general function buttons, according to the display parameters configured by the first user in the interactive interface when the first model is deployed to the cloud, the interactive interface also displays the parameters specific to the customer information, i.e. customer number, economic status, credit rating and educational information; the above display parameters are also the filtering parameters of the customer information; the first user can select one or more of the customer number, economic status, credit rating and educational information as the filtering parameter in the interactive interface, and complete the filtering parameter configuration in the corresponding text box by manual input or manual selection.

[0078] After obtaining the second filtering instruction issued by the second user, the matching customer information can be obtained according to the filtering parameter in the second filtering instruction first; then each customer information is sequentially displayed according to the preset priority of the display parameter other than the filtering parameter; the preset priority represents the ordering priority between different display parameters, and also represents the ordering priority of each sub-parameter in each display parameter, which can be pre-configured by the developer of the first model, or can be configured by the management system based on the historical filtering conditions of multiple users; the more times the current display parameter or display sub-parameter is used as a filtering parameter, the higher the preset priority of the current display parameter or display sub-parameter.

[0079] The technical scheme of the embodiment of the application comprises the following steps: obtaining, through a first interaction interface, a second screening instruction issued by a second user and comprising at least one of a customer number, an economic status, a credit rating and educational background information; and then sequentially displaying each customer information based on a preset priority of other display parameters except screening parameters according to the second screening instruction. In this way, not only is the screening and obtaining of customer information achieved, but also the sorting of customer information based on a preset priority is achieved, which not only meets the customer information screening needs of the user, but also ensures the priority display of the expected customer information.

[0080] Embodiment four

[0081] Figure 4 is a structural block diagram of a large model management device based on cloud service provided by the fourth embodiment of the application, which specifically comprises:

[0082] The model deployment execution module 601 is configured to deploy the first large model on the cloud according to the deployment instruction of the first large model issued by the first user, and show an interactive configuration interface to the first user.

[0083] The interface deployment execution module 602 is configured to obtain the first interaction interface of the first large model through the interactive configuration interface, and deploy the first interaction interface on the local.

[0084] The model calling execution module 603 is configured to show the first interaction interface to the second user in response to obtaining the calling instruction of the first large model issued by the second user, so as to call the first large model through the first interaction interface.

[0085] The technical scheme of the embodiment of the application comprises the following steps: according to the deployment instruction of the first large model issued by the first user, deploying the first large model on the cloud, and showing an interactive configuration interface to the first user; obtaining the first interaction interface of the first large model through the interactive configuration interface, and deploying the first interaction interface on the local; in response to obtaining the calling instruction of the first large model issued by the second user, showing the first interaction interface to the second user, so as to call the first large model through the first interaction interface. In this way, not only is the deployment and application of the large model achieved, but also the application range of the large model is expanded, and the utilization rate of the large model is improved; at the same time, the large model deployed on the cloud and the interaction interface deployed on the local not only improve the data security of the interaction interface itself, but also ensure the calling security of the calling process of the large model.

[0086] Optionally, the cloud service-based large model management apparatus is further configured to, in response to obtaining a first supplementary sample for the first large model from a third user, send the first supplementary sample to the third user to confirm whether to use the first supplementary sample as a training sample of the first large model; and if a confirmation instruction for the first supplementary sample is obtained from the third user, iteratively train the first large model based on the first supplementary sample.

[0087] Optionally, the cloud service-based large model management apparatus is further configured to, if a rejection instruction for the first supplementary sample is obtained from the third user, copy a second large model from the first large model, deploy the second large model on a cloud server, and iteratively train the second large model based on the first supplementary sample; and in label information of the second large model, set the source model and the upstream model as the first large model.

[0088] Optionally, the cloud service-based large model management apparatus is further configured to, in response to obtaining a fuzzy calling instruction from a fourth user, determine a target large model matched with the fourth user based on identity information of the fourth user, and display a target interaction interface corresponding to the target large model to the fourth user to call the target large model through the target interaction interface.

[0089] Optionally, the first large model comprises an evaluation management model.

[0090] The cloud service-based large model management apparatus is further configured to obtain a first screening instruction of the second user through the first interaction interface, wherein the first screening instruction comprises at least one of comment content, evaluation time, performance object, and emotional category; and sequentially display each evaluation information based on a preset priority of other display parameters except the screening parameters according to the first screening instruction.

[0091] Optionally, the first large model comprises a customer management model.

[0092] The cloud service-based large model management apparatus is further configured to obtain a second screening instruction of the second user through the first interaction interface, wherein the second screening instruction comprises at least one of customer number, economic status, credit rating, and educational information; and sequentially display each customer information based on a preset priority of other display parameters except the screening parameters according to the second screening instruction.

[0093] The above apparatus can execute the cloud service-based large model management method provided by any embodiment of the present application, and has the corresponding function modules and beneficial effects of the execution method. Technical details not described in detail in the present embodiment can be referred to the cloud service-based large model management method provided by any embodiment of the present application.

[0094] In the technical solutions of the present disclosure, the collection, storage, use, processing, transmission, provision and disclosure of user personal information comply with relevant laws and regulations and do not violate public order and good customs.

[0095] Embodiment Five

[0096] Figure 5 A structural schematic diagram of an electronic device 10 that can be used to implement embodiments of the present application is shown. The electronic device is intended to represent various forms of digital computers, such as laptops, desktops, tablets, personal digital assistants, electronic devices, blade servers, mainframes, and other appropriate computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular telephones, smart phones, wearable devices (e.g., headsets, glasses, watches, etc.), and other similar computing devices. The components shown here, their connections and relationships, and their functions, are meant to be examples only, and are not meant to limit implementations of the present application described and / or claimed in this document.

[0097] As shown in Figure 5 The electronic device 10 includes at least one processor 11, and a memory, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc., which is communicatively connected to the at least one processor 11, wherein the memory stores a computer program that can be executed by the at least one processor. The processor 11 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 12 or loaded into the random access memory (RAM) 13 from the storage unit 18. In the RAM 13, various programs and data required for the operation of the electronic device 10 can also be stored. The processor 11, the ROM 12, and the RAM 13 are connected to each other through a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0098] A plurality of components in the electronic device 10 are connected to the I / O interface 15, including: an input unit 16, such as a keyboard, a mouse, etc.; an output unit 17, such as various types of displays, speakers, etc.; a storage unit 18, such as a magnetic disk, an optical disk, etc.; and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices through a computer network, such as the Internet, and / or various telecommunications networks.

[0099] The processor 11 can be various general and / or special purpose processing components with processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various specialized artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, and the like. The processor 11 performs various methods and processes described above, such as the cloud-service-based large model management method.

[0100] In some embodiments, the cloud-service-based large model management method can be implemented as a computer program tangibly embodied in a computer readable storage medium, such as the storage unit. In some embodiments, part or all of the computer program can be loaded and / or installed onto the heterogeneous hardware accelerator via the ROM and / or the communication unit. When the computer program is loaded onto the RAM and executed by the processor, one or more steps of the cloud-service-based large model management method described above can be performed. Alternatively, in other embodiments, the processor can be configured to perform the cloud-service-based large model management method by any other suitable means, such as by means of firmware.

[0101] Various implementations of the systems and techniques described above can be realized in digital electronic circuitry, integrated circuitry, a field programmable gate array (FPGA), an application specific integrated circuit (ASIC), a system on a chip (SOC), a programmable logic device (PLD), a computer hardware, firmware, software, and / or combinations thereof. These various implementations can include implementation in one or more computer programs that are executable and / or interpretable on a programmable system including at least one programmable processor, which can be special or general purpose, coupled to receive data and instructions from, and to transmit data and instructions to, a storage system, at least one input device, and at least one output device.

[0102] Computer programs used to implement the methods of the application can be written in any combination of one or more programming languages. These computer programs can be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine, such that the computer program, when executed by the processor of the machine, implements the functions / acts specified in the flowcharts and / or block diagrams. The computer program can be executed entirely on a machine, partially on a machine, partially on a machine as a standalone software package, and partially on a remote machine or electronic device.

[0103] In the context of the present application, a computer-readable storage medium can be a tangible medium that can contain or store a computer program for use by or in connection with an instruction execution system, apparatus, or device. A computer-readable storage medium can include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium can be a machine-readable signal medium. More specific examples of a machine-readable storage medium will include one or more lines of a program of instructions in a transitory signal, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0104] To provide for interaction with a user, the systems and techniques described here can be implemented on a heterogeneous hardware accelerator having a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user and a keyboard and a pointing device (e.g., a mouse or a trackball) by which the user can provide input to the heterogeneous hardware accelerator. Other kinds of devices can be used to provide for interaction with a user as well; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form, including acoustic, speech, or tactile input.

[0105] The systems and techniques described here can be implemented in a computing system that includes a back end component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front end component (e.g., a client computer having a graphical user interface or a Web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such back end, middleware, or front end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network (LAN), a wide area network (WAN), blockchain network, and the Internet.

[0106] A computing system can include a client and an electronic device. The client and the electronic device are typically in different locations, and are often interacted with over a communications network. The relationship of the client and the electronic device is created by computer programs running on the respective computers and having a client-electronic device relationship with each other. The electronic device can be a cloud electronic device, also known as a cloud computing electronic device or a cloud host, which is a host product in a cloud computing service system, and solves the defects of large management difficulty and weak business scalability in traditional physical hosts and VPS services.

[0107] It should be understood that the various forms of flow shown above can be used to reorder, add or delete steps. For example, each step described in the present application can be executed in parallel, sequentially or in a different order, as long as the desired results of the technical solutions of the present application can be achieved, which is not limited herein.

[0108] The above detailed description does not constitute a limitation on the scope of protection of the present application. Those skilled in the art should understand that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent replacements and improvements made within the spirit and principles of the present application shall be included in the scope of protection of the present application.

Claims

1. A large-scale model management method based on cloud services, characterized in that, include: Based on the deployment instruction of the first large model issued by the first user, the first large model is deployed in the cloud, and an interactive configuration interface is displayed to the first user. The first interactive interface of the first large model is obtained through the interactive configuration interface, and the first interactive interface is deployed locally; In response to receiving the call instruction for the first large model issued by the second user, the first interactive interface is displayed to the second user so as to call the first large model through the first interactive interface; The cloud-based large model management method also includes: In response to obtaining a first supplementary sample from a third user for the first large model, the first supplementary sample is sent to the first user so that the first user can confirm whether to use the first supplementary sample as a training sample for the first large model. If a confirmation instruction from the first user for the first supplementary sample is obtained, the first large model is iteratively trained using the first supplementary sample. If the first user's rejection instruction for the first supplementary sample is obtained, the second large model is copied according to the first large model, and the second large model is deployed in the cloud, and the second large model is iteratively trained using the first supplementary sample; In the label information of the second largest model, both the source model and the upstream model are configured as the first largest model.

2. The method according to claim 1, characterized in that, The cloud-based large model management method also includes: In response to receiving a fuzzy invocation command from a fourth user, a matching target large model is determined based on the identity information of the fourth user, and a target interactive interface corresponding to the target large model is displayed to the fourth user so that the target large model can be invoked through the target interactive interface.

3. The method according to claim 1, characterized in that, The first major model includes the evaluation management model; After calling the first large model through the first interactive interface, the process also includes: The first filtering instruction from the second user is obtained through the first interactive interface; wherein the first filtering instruction includes at least one of the following: comment content, evaluation time, performance object, and sentiment category; Based on the first filtering instruction, and according to the preset priority of other display parameters besides the filtering parameters, each evaluation information is displayed in sequence.

4. The method according to claim 1, characterized in that, The first major model includes the customer management model; After calling the first large model through the first interactive interface, the process also includes: The second user's second filtering instruction is obtained through the first interactive interface; wherein the second filtering instruction includes at least one of customer number, economic status, credit rating and education information; According to the second filtering instruction, based on the preset priority of other display parameters besides the filtering parameters, each customer's information is displayed in sequence.

5. A large-scale model management device based on cloud services, characterized in that, include: The model deployment and execution module is used to deploy the first large model to the cloud according to the deployment instruction issued by the first user, and to display the interactive configuration interface to the first user. The interface deployment and execution module is used to obtain the first interactive interface of the first large model through the interactive configuration interface, and deploy the first interactive interface locally. The model invocation execution module is used to respond to the invocation command of the first large model issued by the second user, and to display the first interactive interface to the second user so as to invoke the first large model through the first interactive interface. The cloud-based large model management device is used to send the first supplementary sample to the first user in response to obtaining the first supplementary sample for the first large model from a third user, so that the first user can confirm whether to use the first supplementary sample as a training sample for the first large model. If a confirmation instruction from the first user for the first supplementary sample is obtained, the first large model is iteratively trained using the first supplementary sample; if a rejection instruction from the first user for the first supplementary sample is obtained, a second large model is copied based on the first large model, and the second large model is deployed in the cloud, and the second large model is iteratively trained using the first supplementary sample. In the label information of the second largest model, both the source model and the upstream model are configured as the first largest model.

6. An electronic device, characterized in that, The electronic device includes: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the cloud service-based large model management method according to any one of claims 1-4.

7. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that cause a processor to execute the cloud service-based large model management method according to any one of claims 1-4.

8. A computer program product comprising a computer program that, when executed by a processor, implements the cloud service-based large model management method according to any one of claims 1-4.

Images