Data processing method and device, equipment, storage medium and program product

Abstract

The invention discloses a data processing method and device, equipment, a storage medium and a program, which are used for reducing the development cost of application program development based on an application program interface and improving the data processing efficiency. The method can be applied to the fields of artificial intelligence and cloud technology. Comprising the steps of obtaining a first to-be-processed text; a target large language model is called to conduct recognition processing on the first to-be-processed text to obtain first semantic information corresponding to the first to-be-processed text, and the target large language model is a large language model preset for the current application; determining a first target proxy function from a proxy function set according to the first semantic information, wherein each proxy function in the proxy function set is predefined according to an interface document and an interface parameter of the current application; calling the target interface document and the target interface parameter based on the first target proxy function to generate a first target program; acquiring to-be-processed data; and calling the first target program to process the to-be-processed data to obtain first output data.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of artificial intelligence, and in particular, to a data processing method and device, equipment, a storage medium, and a program product. BACKGROUND

[0002] With the development of Internet technology, more and more functions are transferred to the cloud, and in order to reduce the development cost of application programs, cloud-based application programming interface (API) interfaces are usually used for corresponding application development. The API interface of the existing cloud infrastructure vendor is a set of definitions, programs and protocols, and the mutual communication between computer software is realized through the API interface. One of the main functions of the API is to provide a common function set. Users can develop application programs by using API functions, thereby reducing the programming task.

[0003] For different cloud infrastructure vendors, the API interface definitions usually have different interface documents and interface parameters. Therefore, when users use the API interface of the cloud infrastructure vendor to develop application programs, they usually need to understand the corresponding interface documents and interface parameters. For example, in the PaaS live broadcast and media processing scenario, there are live broadcast room creation, room closing, specified resolution and bit rate transcoding, and specified scene requirements such as super resolution, frame insertion, color enhancement, etc. This brings a relatively poor application threshold, that is, if a user wants to complete a requirement, he or she must understand the related parameters of the API provided by the cloud infrastructure vendor, the parameter usage restrictions and the specific parameter meanings. If the interface documents and interface parameters are not clearly understood, the correct development and application of the application program cannot be completed. That is, the existing scheme requires a high application threshold for users and programmers, and the cost of development and online is also high.

[0004] Therefore, there is an urgent need for a scheme that can reduce the development requirements and improve the data processing efficiency. SUMMARY

[0005] The embodiments of the present application provide a data processing method, device, equipment, storage medium and program product, which are used to reduce the development cost of application program development based on application programming interface, and improve the data processing efficiency.

[0006] In view of this, one aspect of the present application provides a data processing method, comprising: obtaining a first to-be-processed text; calling a target large language model to perform recognition processing on the first to-be-processed text to obtain first semantic information corresponding to the first to-be-processed text, wherein the target large language model is a large language model pre-adjusted for a current application; determining a first target proxy function from a proxy function set according to the first semantic information, each proxy function in the proxy function set being pre-defined according to an interface document and an interface parameter of the current application; generating a first target program based on the first target proxy function calling the target interface document and the target interface parameter; obtaining to-be-processed data; and calling the first target program to process the to-be-processed data to obtain first output data.

[0007] Another aspect of the present application provides a data processing apparatus, comprising: an obtaining module configured to obtain a first to-be-processed text;

[0008] a processing module configured to call a target large language model to perform recognition processing on the first to-be-processed text to obtain first semantic information corresponding to the first to-be-processed text, wherein the target large language model is a large language model pre-adjusted for a current application; determine a first target proxy function from a proxy function set according to the first semantic information, each proxy function in the proxy function set being pre-defined according to an interface document and an interface parameter of the current application;

[0009] generate a first target program based on the first target proxy function calling the target interface document and the target interface parameter;

[0010] The obtaining module is configured to obtain to-be-processed data.

[0011] The processing module is configured to call the first target program to process the to-be-processed data to obtain first output data.

[0012] In a possible design, in another implementation manner of the other aspect of the embodiments of the present application, the obtaining module is configured to obtain a training sample of the current application, the training sample being a question and answer material constructed based on the interface document and the interface parameter of the current application.

[0013] The processing module is configured to pre-adjust the large language model based on the training sample to obtain the target large language model.

[0014] In a possible design, in another implementation manner of the other aspect of the embodiments of the present application, the format of the training sample is json format, XML format or csv format.

[0015] In a possible design, in another implementation manner of another aspect of the embodiment of the present application, the processing module is configured to determine a definition format of the proxy function set based on the target large language model.

[0016] The obtaining module is configured to obtain a keyword calling function of the current application.

[0017] The processing module is configured to register the proxy function set based on the keyword calling function and the definition format.

[0018] In a possible design, in another implementation manner of another aspect of the embodiment of the present application, the processing module is configured to register the keyword calling function to a system tool based on the definition format, to generate the proxy function set.

[0019] In a possible design, in another implementation manner of another aspect of the embodiment of the present application, the to-be-processed data is audio and video data, the current application is a cloud audio and video processing application, and the first to-be-processed text is used to describe a processing scheme of the audio and video data.

[0020] In a possible design, in another implementation manner of another aspect of the embodiment of the present application, the obtaining module is configured to obtain a first keyword corresponding to the first semantic information.

[0021] The processing module is configured to traverse the proxy function set based on the first keyword, and take the proxy function set and the proxy function hit by the first keyword as the first target proxy function.

[0022] In a possible design, in another implementation manner of another aspect of the embodiment of the present application, the processing module is configured to, when the first target proxy function is not determined from the proxy function set according to the first semantic information, determine a first proxy function according to the to-be-processed data, the first proxy function being predefined according to an interface document and an interface parameter of the current application; generate a second target program based on the first proxy function calling the corresponding interface document and interface parameter; and process the to-be-processed data based on the second target program, to obtain second output data.

[0023] In a possible design, in another implementation manner of another aspect of the embodiment of the present application, the obtaining module is configured to obtain a second to-be-processed text, the second to-be-processed text being different from the first to-be-processed text.

[0024] The processing module is configured to invoke the target large language model to perform recognition processing on the second to-be-processed text to obtain second semantic information corresponding to the second to-be-processed text, determine a second target proxy function from the proxy function set according to the second semantic information, and invoke a corresponding interface document and interface parameter based on the second target proxy function to generate a third target program.

[0025] The acquisition module is configured to acquire the to-be-processed data.

[0026] The processing module is configured to invoke the third target program to process the to-be-processed data to obtain third output data.

[0027] In a possible design, in another implementation manner of the other aspect of the embodiment of the present application, the apparatus further includes a display module, which is configured to display a first interface, the first interface including a text input control;

[0028] The processing module is configured to acquire the first to-be-processed text based on the text input control.

[0029] In a possible design, in another implementation manner of the other aspect of the embodiment of the present application, the apparatus further includes a display module, which is configured to display a second interface, the second interface including a data input control;

[0030] The processing module is configured to acquire the to-be-processed data based on the data input control.

[0031] Another aspect of the present application provides a computer device, including a memory, a processor and a bus system;

[0032] The memory is configured to store a program.

[0033] The processor is configured to execute the program in the memory, and the processor is configured to execute the method of each aspect according to the instructions in the program code.

[0034] The bus system is configured to connect the memory and the processor to enable the memory and the processor to communicate.

[0035] Another aspect of the present application provides a computer readable storage medium, which stores instructions, and when the instructions are run on a computer, the computer executes the method of each aspect.

[0036] In another aspect of the present application, a computer program product or computer program is provided, which includes computer instructions stored in a computer readable storage medium. A processor of a computer device reads the computer instructions from the computer readable storage medium, and the processor executes the computer instructions to cause the computer device to perform the method provided in the above aspects.

[0037] From the above technical solutions, it can be seen that the embodiments of the present application have the following advantages: by combining the semantic understanding function of the large language model and the calling function of the proxy function for the interface document and the interface function, when the user calls the application to process data, the application can understand the semantic information of the input text through the semantic understanding function of the large language model, and based on the understood semantic information, the corresponding proxy function is recalled, and the interface document and the interface parameter are correctly called by using the proxy function. In this way, it is no longer necessary for the user to perform program configuration after understanding the interface document and the interface parameter of the application, thereby reducing the understanding cost of the user and improving the efficiency of data processing. BRIEF DESCRIPTION OF DRAWINGS

[0038] Figure 1 An application scenario architecture diagram of the data processing scheme in the embodiments of the present application is shown;

[0039] Figure 2 An embodiment diagram of the data processing method in the embodiments of the present application is shown;

[0040] Figure 3 An input interface diagram of the text to be processed in the embodiments of the present application is shown;

[0041] Figure 4 An embodiment diagram of the data processing apparatus in the embodiments of the present application is shown;

[0042] Figure 5 Another embodiment diagram of the data processing apparatus in the embodiments of the present application is shown;

[0043] Figure 6 Another embodiment diagram of the data processing apparatus in the embodiments of the present application is shown;

[0044] Figure 7 Another embodiment diagram of the data processing apparatus in the embodiments of the present application is shown. DETAILED DESCRIPTION

[0045] The embodiments of the present application provide a data processing method, apparatus, device, storage medium and program product, which are used to reduce the development cost of application program development based on application program interface, and improve the data processing efficiency.

[0046] The terms "first", "second", "third", "fourth" and the like in the description and in the claims of the present application, if any, are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It is to be understood that the use of these terms herein is to be construed to cover a generic and collective sense of the reference used and not in a contractual or restrictive sense unless expressly so defined in the specification. It is to be understood that where the application is indicated to include "a" or "an" element, similar elements can be employed without departing from the application. It is also to be understood that the use of the term "or" in the description is used to mean "and / or" unless otherwise indicated to indicate exclusive "or" if such interpretation is intended. It is further understood that the use of relative terms, such as "approximately," "about," and the like, are intended to encompass measurements within standard manufacturing tolerances and variations that would be expected by those of ordinary skill in the art.

[0047] In the embodiments of the present application, the term "module" or "unit" refers to a computer program or a part of a computer program with a predetermined function, and works together with other related parts to achieve a predetermined target, and can be implemented entirely or partially by using software, hardware (such as a processing circuit or a memory) or a combination thereof. Similarly, one processor (or multiple processors or memories) can be used to implement one or more modules or units. In addition, each module or unit can be a part of an integral module or unit that includes the functions of the module or unit.

[0048] Those skilled in the art can understand that, unless specifically stated, the singular forms "a", "an" and "the" used herein also include the plural forms. It should be further understood that the terms "include" and "contain" used in the embodiments of the present application mean that the corresponding features can be implemented as the presented features, information, data, steps, operations, elements and / or components, but do not exclude other features, information, data, steps, operations, elements, components and / or their combinations supported by the present technology. It should be understood that when we say that an element is "connected" or "coupled" to another element, the element can be directly connected or coupled to the other element, or it can mean that the element and the other element are connected through an intermediate element. In addition, "connected" or "coupled" used herein can include wireless connection or wireless coupling. The term "and / or" used herein indicates that at least one of the items defined by the term, for example, "A and / or B" can be implemented as "A", or as "B", or as "A and B".

[0049] With the discovery of Internet technology, more and more functions are transferred to the cloud, and in order to reduce the development overhead of the application program, the cloud-based API interface is usually used for the corresponding application development. The API interface of the existing cloud infrastructure vendor is a set of definitions, programs and protocols, and the mutual communication between computer software is realized through the API interface. One of the main functions of the API is to provide a general function set. Users can develop application programs by using API functions, thereby reducing the programming task. For different cloud infrastructure vendors, the API interface definition usually has different interface documents and interface parameters. Therefore, when users develop application programs using the API interface of the cloud infrastructure vendor, they usually need to understand the corresponding interface documents and interface parameters. For example, in the PaaS live and media processing scenario, there are live room creation, room closing, specified resolution and rate transcoding, and specified scene requirements such as super resolution, frame insertion, color enhancement, etc. This brings a relatively poor application threshold, that is, if a user wants to complete a requirement, he must understand the related parameters of the API provided by the cloud infrastructure vendor, the parameter usage restrictions and the specific parameter meanings. If the interface documents and interface parameters are not clearly understood, the correct development and application of the application program cannot be completed. That is, the existing scheme makes the user's use threshold and programmer's development requirements higher, and the cost of development and online is also higher. Therefore, there is an urgent need for a scheme that can reduce the development requirements and improve the data processing efficiency.

[0050] To solve the above technical problems, the present application provides the following technical scheme: obtaining a first to-be-processed text; calling a target large language model to recognize and process the first to-be-processed text to obtain first semantic information corresponding to the first to-be-processed text, wherein the target large language model is a large language model pre-adjusted for the current application; determining a first target proxy function from a proxy function set according to the first semantic information, wherein each proxy function in the proxy function set is pre-defined according to the interface document and the interface parameter of the current application; generating a first target program based on the target interface document and the target interface parameter calling the first target program; obtaining the to-be-processed data; and calling the first target program to process the to-be-processed data to obtain first output data. In this way, by combining the semantic understanding function of the large language model and the calling function of the proxy function for the interface document and the interface function, when the user calls the application program to process the data, the application program can understand the semantic information of the input text through the semantic understanding function of the large language model, recall the corresponding proxy function based on the understood semantic information, and realize the correct calling of the interface document and the interface parameter by using the proxy function. Therefore, the user no longer needs to configure the program after understanding the interface document and the interface parameter of the application program, thereby reducing the understanding cost of the user and improving the efficiency of data processing.

[0051] The data processing method provided by the embodiments of the present application can be applied to the implementation of artificial intelligence (AI) technology. For example, a large language model for a certain application scenario can be trained for a user. AI is a theory, method, technology and application system that uses a digital computer or a machine controlled by a digital computer to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use the knowledge to obtain the best results. With the research and progress of artificial intelligence technology, artificial intelligence technology has been widely researched and applied in many fields. It is believed that with the development of technology, artificial intelligence technology will be applied in more fields and play an increasingly important role.

[0052] The machine learning involved in the present application mainly refers to fine-tuning a pre-trained model to obtain a target large language model, so as to realize semantic understanding of a user input text to be processed, match a corresponding proxy function based on the semantic understanding, call an interface document and an interface parameter of a corresponding application by using the proxy function, and finally realize data processing.

[0053] Artificial intelligence technology is a comprehensive discipline, involving a wide range of fields, both hardware level technology and software level technology. Artificial intelligence basic technology generally includes, such as sensors, special artificial intelligence chips, cloud computing, distributed storage, big data processing technology, pre-trained model technology, operation / interaction system, mechatronics, etc. Among them, the pre-trained model is also called large model, basic model, which can be widely applied to downstream tasks in various directions of artificial intelligence after fine-tuning. Artificial intelligence software technology mainly includes computer vision technology, speech processing technology, natural language processing technology and machine learning / deep learning, etc.

[0054] Pre-training model (Pre-training model), also known as cornerstone model, large model, refers to a deep neural network (Deep neural network, DNN) with large parameters, which is trained on a large amount of unlabeled data. The function approximation ability of large parameter DNN enables PTM to extract common features from data. Through fine tuning, parameter efficient fine tuning (PEFT), prompt-tuning and other technologies, it is suitable for downstream tasks. Therefore, the pre-training model can achieve ideal results in the few-shot or zero-shot scene. PTM can be divided into language model (ELMO, BERT, GPT), visual model (swin-transformer, ViT, V-MOE), speech model (VALL-E), multi-modal model (ViBERT, CLIP, Flamingo, Gato) according to the data modality processed, among which the multi-modal model refers to the model for establishing the feature representation of two or more data modalities. The pre-training model is an important tool for output artificial intelligence generated content (AIGC), and can also be used as a general interface connecting multiple specific task models.

[0055] Computer vision technology (Computer Vision, CV) Computer vision is a science that studies how to make machines "see". Further, it refers to using cameras and computers to replace human eyes to identify, follow and measure targets and other machine vision, and further process graphics so that the computer processing becomes more suitable for human eye observation or image transmission to instrument detection. As a scientific discipline, computer vision researches related theories and technologies, and tries to establish an artificial intelligence system that can obtain information from images or multidimensional data. Large model technology brings important changes to the development of computer vision technology. Swin-transformer, ViT, V-MOE, MAE and other pre-training models in the field of vision can be quickly and widely applied to downstream specific tasks after fine tuning. Computer vision technology usually includes image processing, image recognition, image semantic understanding, image retrieval, OCR, video processing, video semantic understanding, video content / behavior recognition, three-dimensional object reconstruction, 3D technology, virtual reality, augmented reality, simultaneous localization and map building, etc. It also includes common face recognition, fingerprint recognition and other biometric identification technologies.

[0056] The present application also relates to cloud technology. Among them, cloud technology (cloud technoolgy) refers to a hosting technology that unifies system resources such as hardware, software and network in a wide area network or local area network to realize data calculation, storage, processing and sharing.

[0057] Cloud technology is a general term for network technology, information technology, integration technology, management platform technology, application technology and other technologies based on cloud computing business model application, which can form a resource pool, and be used on demand, flexibly and conveniently. Cloud computing technology will become an important support. The background service of the technical network system needs a large amount of computing and storage resources, such as video websites, picture websites and more portals. With the high development and application of Internet behavior, in the future, every item may have its own identification mark and needs to be transmitted to the background system for logical processing. Different levels of data will be processed separately, and various industry data will need strong system support, which can only be realized through cloud computing. The cloud technology involved in the present application mainly refers to the development of application programs between terminal devices or servers through "cloud" and the like.

[0058] For the convenience of understanding, some professional terms in the present application are explained as follows.

[0059] Cloud API: Cloud API refers to the development of cloud application programming interface coding, and the interface has a service of a cloud provider. However, it is also dangerous for cloud applications, because the API also has an attack side, which may endanger sensitive business data. This means that the provider and software developers need to determine the security of the cloud API in priority order.

[0060] Cloud platform service (Platform as a Service, PaaS): a kind of cloud computing service, which simplifies the development, deployment, expansion and management process of application program through cloud platform, and provides more efficient, flexible and reliable computing service for users.

[0061] Large Language Model (LLM): A deep learning model trained on massive amounts of text data. It can not only generate natural language text, but also deeply understand the meaning of the text and handle various natural language tasks such as text summarization, question answering, translation, etc. Most current large language models follow the model structure of Transformer, with the difference being that some use an encoder-decoder structure, while others only use one type of encoder or decoder. Currently, various LLMs based on the Transformer architecture, and even various visual models based on ViT (Vision Transformer), are implemented by "prediction" to generate results. This "prediction" is the output content learned by the large model based on the training data, under the premise of input text instructions (i.e., prompt), which is the prediction of the content the user wants to output given the input content. Among them, the prompt refers to the question or instruction the user raises to the AI assistant during conversational AI interaction. It is the starting point of communication between the user and the AI assistant. When understanding the prompt, it can be regarded as the user's question or request description to the AI assistant. The AI assistant will generate an answer or perform the corresponding operation according to the prompt provided by the user. For each question, the AI assistant will try its best to understand and give relevant answers. When generating answers, the AI assistant will combine its own knowledge and model training results to provide the most appropriate answers. At the same time, the prompt can also contain some specific instructions or format requirements to help the AI assistant better understand the user's needs. For example, it can be explicitly required in the prompt to ask the AI assistant to give detailed explanations, list several examples, or provide relevant reference materials, etc.

[0062] Natural Language Processing (NLP): An important direction in the field of computer science and artificial intelligence. It studies various theories and methods that can realize effective communication between people and computers using natural language. Natural language processing is a science that integrates linguistics, computer science, and mathematics. Natural language processing is mainly applied to machine translation, public opinion monitoring, automatic summarization, opinion extraction, text classification, question answering, text semantic comparison, speech recognition, Chinese optical character recognition (OCR), etc.

[0063] Copilot: An AI code assistant that generates code snippets and suggestions based on the context and comments provided by the user. It is based on large models and can help developers improve coding efficiency and reduce repetitive work. Copilot can also be used as a learning tool for programming. Copilot works by training on a large amount of open-source code to learn the syntax, structure, and common patterns of code. When the user is writing code, Copilot generates possible code snippets based on the context and comments and provides them for the user to choose from. The user can choose to accept the suggestions and insert them into their own code or ignore the suggestions and continue writing on their own. Copilot supports multiple programming languages, including Python, JavaScript, TypeScript, Go, and others. It can be used with common code editors and integrated development environments (IDEs) to provide instant code completion and suggestions.

[0064] The data processing method, device, equipment, storage medium and program product provided by the embodiments of the present application can reduce the development cost of application program development based on the application program interface, and improve the data processing efficiency. The electronic device provided by the embodiments of the present application can be implemented as various types of user terminals, or as a server.

[0065] The electronic device runs the data processing method provided by the embodiments of the present application to reduce the development cost of application program development based on the application program interface, and improve the data processing efficiency.

[0066] The above scheme can be applied to many artificial intelligence fields, including artificial intelligence code development or application to cloud provider corresponding service field. When the data processing method provided by the embodiments of the present application is used to help users to process data, the method can be realized as an independent online application program, installed in the computer equipment used by the user or the background server, so as to facilitate the user to use the program to process data.

[0067] In this scenario, the user inputs the to-be-processed text (which can also be understood as a text instruction or a text question) through an input interface in an application interface, then calls a pre-configured large language model to perform semantic understanding on the to-be-processed text, obtains semantic information corresponding to the to-be-processed text according to the semantic understanding, recalls a proxy function corresponding to the to-be-processed text from the proxy function set, and calls an interface document and an interface parameter corresponding to the proxy function to generate a corresponding target program; finally, the target program is used to process the to-be-processed data to obtain output data. For example, in an audio and video application scenario, after the to-be-processed text is subjected to semantic understanding, a semantic information of video enhancement processing is obtained, then a proxy function corresponding to the video enhancement processing is matched from the proxy function set, the proxy function is used to call an interface document and an interface parameter corresponding to the video enhancement processing, and a corresponding program is generated according to the interface document and the interface parameter, and finally the program is used to perform video enhancement processing (such as enhancing video color, etc.) on video data.

[0068] Of course, in addition to being applied to the above scenario, the method provided in the embodiments of the present application can also be applied to other scenarios of data processing using an application program interface, and the embodiments of the present application do not limit the specific application scenario.

[0069] Reference is made to Figure 1 , Figure 1An optional architecture diagram in an application scenario of a data processing scheme provided by the embodiment of the present application is provided, in order to implement a data processing scheme, the terminal device 100 is connected with the server 300 through the network 200, the server 300 is connected with the database 400, and the network 200 can be a wide area network or a local area network, or a combination of the two. The client for implementing the data processing scheme is deployed on the terminal device 100, wherein the client can run on the terminal device 100 in the form of a browser, or run on the terminal device 100 in the form of an independent application (application, APP), etc. The specific form of the client is not limited here. The server 300 involved in the present application can be a stand-alone physical server, or a server cluster or distributed system composed of multiple physical servers, or a cloud server providing cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content distribution networks (Content Delivery Network, CDN), and basic cloud computing services such as big data and artificial intelligence platforms. The terminal device 100 can be a smart phone, a tablet computer, a notebook computer, a palm computer, a personal computer, a smart television, a smart watch, a vehicle-mounted device, a wearable device, a smart voice interaction device, a smart home appliance, a flying device, etc., but is not limited thereto. The terminal device 100 and the server 300 can be directly or indirectly connected through the network 200 through wired or wireless communication, which is not limited in the present application. The number of the server 300 and the terminal device 100 is also not limited. The scheme provided by the present application can be independently completed by the terminal device 100, or independently completed by the server 300, or completed by the terminal device 100 and the server 300 in cooperation, which is not limited in the present application. In summary, the database 400 can be regarded as an electronic file cabinet, i.e., a place for storing electronic files, and a user can add, query, update, delete, etc. data in the files. The so-called "database" is a collection of data stored together in a certain way, shared by multiple users, with as little redundancy as possible, and independent of application programs. The database management system (Database Management System, DBMS) is a computer software system designed to manage databases, and generally has basic functions such as storage, interception, security, backup, etc.A database management system can be classified according to the database model it supports, such as relational, Extensible Markup Language (XML); or according to the computer type it supports, such as server cluster, mobile phone; or according to the query language it supports, such as Structured Query Language (SQL), XQuery; or according to the performance impulse focus, such as maximum scale, highest running speed; or other classification methods. Regardless of the classification method used, some DBMSs can cross categories, such as supporting multiple query languages at the same time. In this application, the database 400 can be used to store the target large language model and the proxy function set, and of course, the storage location of the target large language model and the proxy function set is not limited to the database, for example, it can also be stored in the terminal device 100, the distributed file system of the server 300, or the blockchain, etc.

[0070] In some embodiments, the server 300 and the terminal device 100 can both execute the data processing method provided by the embodiments of the present application.

[0071] It can be understood that in the specific embodiments of the present application, data related to the to-be-processed text, the proxy function, etc. are involved, and when the above embodiments of the present application are applied to specific products or technologies, the user's permission or consent needs to be obtained, and the collection, use, and processing of related data need to comply with relevant laws, regulations, and standards of countries and regions.

[0072] In combination with the above introduction, the terminal device as the execution subject will be introduced below to the data processing method in the present application. Please refer to Figure 2 One embodiment of the data processing method in the embodiments of the present application includes:

[0073] 201, obtaining a first to-be-processed text.

[0074] In this embodiment, the terminal device can receive the first to-be-processed text sent by a third-party device and obtain the first to-be-processed text input by the user through the interface displayed by the display device of the terminal device. For example, as shown in Figure 3 The display device of the terminal device displays a first interface, wherein the first interface is an interactive interface of the data processing, and the first interface has an input bar for inputting the to-be-processed text. Then the user clicks the input bar to activate the input bar, and finally the user inputs the first to-be-processed text through the keyboard or generally recognizes the text content of the input bar from the voice data, and the specific implementation is not limited here.

[0075] It should be understood that the first to-be-processed text can be understood as a text instruction or a text question, that is, in actual application, the first to-be-processed text can be a prompt. Wherein, the prompt refers to a question or instruction input by the user through the input bar when performing a conversational artificial intelligence interaction, which is the starting point of communication between the user and the data processing apparatus in the embodiment. When understanding the prompt, it can be regarded as a description of the user's question or request to the data processing apparatus in the embodiment. The data processing apparatus in the embodiment will generate an answer or perform a corresponding operation according to the prompt provided by the user. For example, as shown in Figure 3 , the first to-be-processed text can be "help me translate subtitles", "I need to control costs and reduce bit rates, but the picture quality cannot be too poor" or "I need to perform quality inspection on the video and perform targeted processing on the places that do not meet the quality standards".

[0076] 202, calling a target large language model to perform recognition processing on the first to-be-processed text to obtain first semantic information corresponding to the first to-be-processed text, wherein the target large language model is a large language model pre-adjusted for the current application.

[0077] In the embodiment, after the terminal device obtains the first to-be-processed text, the terminal device calls a target large language model pre-adjusted for the current application to perform semantic recognition on the first to-be-processed text to obtain first semantic information of the first to-be-processed text. For example, when the first to-be-processed text is "help me translate subtitles", the semantic information recognized by the target large language model can be "translate subtitles", and the translation language of the subtitles can also be determined according to the language information of the video and the personal characteristics of the object. For example, when the language of the video is English and the personal characteristics of the object indicate that the familiar language of the object is Chinese, it can be determined that the subtitles of the video will be translated from English to Chinese.

[0078] It should be understood that the target large language model can be obtained by pre-training a large language model for the current application. The specific process can be as follows:

[0079] Based on the interface document and the interface parameter of the current application, a question and answer material is constructed as a training sample of the pre-trained large language model, and then the pre-trained large language model is pre-trained based on the training sample to obtain the target large language model. The construction format of the training sample can be json format, XML format or csv format.

[0080] JSON (JavaScript Object Notation) is a lightweight data interchange format commonly used for data transmission and storage between the front-end and back-end. It organizes data in key-value pairs and uses curly braces ({}) for objects and square brackets ([]) for arrays. The basic syntax rules of JSON are as follows:

[0081] Key-Value Pair: Use a colon (:) to separate the key and value, and use a comma (,) to separate multiple key-value pairs.

[0082] Object: Enclosed by a pair of curly braces, there is no order requirement between each key-value pair.

[0083] Array: Enclosed by a pair of square brackets, use a comma to separate multiple values.

[0084] Value: Can be a string, number, boolean, object, array, or null.

[0085] XML (Extensible Markup Language) is a markup language used to store and transmit data. It is designed to describe the structure and content of data and is self-descriptive. Here are some features and examples of XML format:

[0086] Tags: XML uses tags to define the structure of data. Tags are enclosed in angle brackets and can have a start tag and an end tag, for example: <tag>data< / tag> .

[0087] Elements: Data in XML is contained in elements. An element consists of a start tag, an end tag, and the content between them. For example: <name>John< / name> .

[0088] Attributes: XML elements can have attributes, which provide additional information about the element. Attributes exist in the form of name-value pairs and are placed in the start tag. For example: <person id="123">.

[0089] Nested: XML allows elements to be nested within other elements, forming a hierarchical structure. For example: <person> <name>John< / name> < / person> .

[0090] Comments: XML supports comments, which start with <!– and end with –>. For example: <!– This is a comment –>.

[0091] Namespaces: XML supports the use of namespaces to avoid conflicts between element and attribute names. Namespaces are referenced by a prefix, for example: ns:element.

[0092] CDATA section: The CDATA section is used to include special characters in XML to prevent them from being parsed as tags. The CDATA section starts with <! [CDATA[ and ends with ]]>. For example: <! [CDATA[ <tag>data< / tag> ]]>.

[0093] XML Declaration: An XML document usually begins with an XML declaration, which specifies the XML version and character encoding. For example: <? XMLversion="1.0" encoding="UTF-8"?>.

[0094] CSV (Comma-Separated Values) is a common file format used for storing and exchanging data. It is a plain text format in which data is separated by commas, each line represents a record, and each field represents a data item. CSV files can be opened and edited using any text editor. Each line of data is separated by commas into multiple fields, with no spaces between fields. Typically, the first line contains the field names, which describe the meaning of each field. Starting from the second line, each line is a data record, and the order of the fields corresponds to the field name line. CSV files are very flexible and can store various types of data, including text, numbers, dates, and more. They are widely used in scenarios such as data import and export, data exchange, and data storage. Many software and programming languages ​​provide support for reading and writing CSV files.

[0095] The question-and-answer material can be understood as a text instruction corresponding to a proxy function. For example, if the text instruction is "help me translate subtitles", the corresponding proxy function is proxy function A. In this case, this question-and-answer material is a training sample.

[0096] 203. Identify a first target proxy function from a proxy function set according to the first semantic information, where each proxy function in the proxy function set is predefined according to an interface document and interface parameters of the current application.

[0097] In this embodiment, after obtaining the first semantic information, the terminal device can obtain the first target proxy function from the proxy function set based on the first semantic information. In an exemplary solution, the matching process can be that the target large language model recalls the first target proxy function from the proxy function set based on the first semantic information; or the first semantic information is a keyword, in which case the keyword is matched with the keyword function of each proxy function in the proxy function set. If a match is found, the proxy function that is matched is determined to be the first target proxy function. The specific solution is not limited here.

[0098] It should be understood that if the terminal device does not hit the corresponding first target proxy function according to the first semantic information, the terminal device can identify the to-be-processed data according to the target large language model to determine the processing scheme required by the to-be-processed data, then match the corresponding proxy function according to the processing method, and call the corresponding interface document and interface parameter according to the corresponding proxy function to generate a second target program, and process the to-be-processed data according to the second target program.

[0099] In the embodiment, different application programs have different interface documents and interface parameters. In an exemplary scheme, it is assumed that the current application is a cloud media processing (MPS) application. The MPS has strong coding capability, can save storage and bandwidth costs, and can realize full-platform playback of media content. At the same time, the MPS can also provide video screenshot, audio and video enhancement, content understanding, content review and other capabilities to meet the processing needs of video in various scenarios. When the API of the cloud media processing application is called, the interface document and the interface parameter can include information such as specific call domain name selection, request structure, public parameter, signature access, and result return. The following takes the creation of a transcoding template enhancement configuration of the MPS product as an example:

[0100] The interface Action name is CreateTranscodeTemplate. If the enhancement function is to be used, the EnhanceConfig structure parameter needs to be filled in when the API interface parameter is called. Taking the enhancement function as an example, the interface parameter related to enhancement can include the contents shown in Tables 1 to 3.

[0101] Table 1

[0102]

[0103] Table 1 is used to indicate audio enhancement configuration, and its structure parameter is EnhanceConfig, which is referenced by the following interfaces: CreateTranscodeTemplate, DescribeTranscodeTemplates, and ModifyTranscodeTemplate.

[0104] Table 2

[0105]

[0106]

[0107] Table 2 is used to indicate the video enhancement configuration. Its structure parameter is VideoEnhanceConfig, which is referenced by the following interfaces: CreateTranscode Template, Modily Transcode Template.

[0108] Table 3

[0109]

[0110] Table 2 is used to indicate the audio enhancement and reduction configuration. Its structure parameter is AudioEnhanceConfig, which is referenced by the following interfaces: CreateTranscodeTemplate, Modity TranscodeTemplate.

[0111] Based on the above interface parameters and the description of the interface document, the proxy function (i.e. function calling) can be defined as follows:

[0112] tools=[

[0113] },

[0114] {

[0115] "type":"function"

[0116] "function":{

[0117] "name":"audio_video_enhancement",

[0118] "description":"Video enhancement processing",

[0119] "parameters":{

[0120] "type":"object",

[0121] "properties":{

[0122] "enhancement_type":{"type":"string",

[0123] #The specific value of the enum type should be consistent with the Chinese in the description, or fine-tuning certain Chinese keywords to correspond to a specific enum value

[0124] "enum": ["super-resolution", "image-quality-enhancement", "audio_enhancement_separation"], "description": "Video enhancement type"},

[0125] },

[0126] "required": ["enhancement_type"],

[0127] },

[0128] },

[0129] "type": "function",

[0130] "function": {

[0131] "name": "audio_video_transcoding",

[0132] "description": "Video transcoding processing",

[0133] "parameters": {

[0134] "type": "object"

[0135] "properties": {

[0136] "packaging_format": { "type": "string", "enum": ["HLS", "MP4"], "description": "Packaging format"},

[0137] "video_encoding": { "type": "string", "enum": ["H.264", "H.265"], "description": "Video encoding"},

[0138] "bitrage_upper_limit": { "type": "integer", "description": "Average bitrate upper limit"},

[0139] "px": { "type": "integer", "description": "Resolution"},

[0140] "fps": {"type": "integer", "description": "frame rate"},

[0141] "audio_encoding": {"type": "string", "enum": ["ACC"], "description": "audio encoding"},

[0142] },

[0143] "required": ["packaging_format", "video_encoding", "bitrage_upper_limit", "px", "fps", "audio_encoding"], / / this parameter specifies which parameters are required parameters

[0144] },

[0145] },

[0146] It should be understood that when the Function Call function is used, some functions (the function name, the description of the function purpose, the parameter name, and the parameter description need to be specified) need to be defined (not really writing a program to define a function, but only describing a function in words) and transmitted to the LLM. When the terminal device obtains a to-be-processed text, the LLM analyzes the text to determine whether a certain function needs to be called. If the function needs to be called, the LLM returns a json. The json includes the function name that needs to be called, the parameter name that needs to be input into the function, and the parameter value. In summary, the function call performs two functions: 1. determining whether a certain predefined function needs to be called; and 2. if the function needs to be called, extracting the function value required by the function from the to-be-processed text.

[0147] 204. Generating a first target program based on the first target proxy function, the target interface document, and the target interface parameter.

[0148] In this embodiment, after determining the corresponding first target proxy function, the terminal device calls the target interface document and the target interface function defined by the first target proxy function, and then generates the first target program based on the target interface document and the target interface function.

[0149] 205. Obtaining to-be-processed data.

[0150] In this embodiment, the terminal device can receive to-be-processed data sent by a third-party device and obtain the to-be-processed data uploaded by a user through an interface displayed by the display device of the terminal device. For example, as shown in FIG. 1, the terminal device can receive to-be-processed data sent by a third-party device and obtain the to-be-processed data uploaded by a user through an interface displayed by the display device of the terminal device. Figure 3As shown, a second interface is displayed on the display device of the terminal device, wherein the second interface is an interactive interface of the data processing, and the second interface has an upload control for inputting the to-be-processed data. Then, the user clicks the upload control to activate the upload control, and finally, the user selects the to-be-processed data to upload the to-be-processed data. Details are not limited herein.

[0151] It should be understood that the to-be-processed data can be audio and video data, or various text or various picture media content, and details are not limited herein.

[0152] 206. The first target program is called to process the to-be-processed data to obtain first output data.

[0153] The terminal device calls the first target program generated in steps 201 to 204 to process the to-be-processed data to obtain first output data corresponding to the to-be-processed data.

[0154] It should be understood that after the terminal device outputs the first output data, if the first output data meets the requirements, the data processing process can be ended. If the first output data does not meet the requirements, the terminal device can obtain new to-be-processed text, i.e., second to-be-processed text. At this time, the second to-be-processed text can be improved based on the first to-be-processed text, or can be rewritten, and details are not limited herein. After obtaining the second to-be-processed text, steps 201 to 206 can be executed again to process the to-be-processed data to obtain new output data.

[0155] Alternatively, when the first output data does not meet the requirements, the terminal device can obtain new to-be-processed text, i.e., second to-be-processed text. At this time, the second to-be-processed text can be improved based on the first to-be-processed text, or can be rewritten, and details are not limited herein. After obtaining the second to-be-processed text, the first output data can be processed according to steps 201 to 206 to obtain new output data.

[0156] The data processing apparatus in the present application is described in detail below. Please refer to Figure 4 , Figure 4 An embodiment of the data processing apparatus in the present application is shown in the figure. The data processing apparatus 20 includes:

[0157] The acquisition module 201 is configured to acquire first to-be-processed text.

[0158] The processing module 202 is configured to call a target large language model to perform recognition processing on the first to-be-processed text to obtain first semantic information corresponding to the first to-be-processed text, wherein the target large language model is a large language model pre-adjusted for the current application; determine a first target proxy function from a proxy function set according to the first semantic information, wherein each proxy function in the proxy function set is pre-defined according to an interface document and an interface parameter of the current application; and generate a first target program based on the first target proxy function, the target interface document and the target interface parameter.

[0159] The acquisition module 201 is configured to acquire to-be-processed data.

[0160] The processing module 202 is configured to call the first target program to process the to-be-processed data to obtain first output data.

[0161] In the embodiments of the present application, a data processing apparatus is provided. By using the above apparatus, in combination with the semantic understanding function of the large language model and the calling function of the proxy function for the interface document and the interface function, when a user calls an application program to process data, the application program can understand the semantic information of the input text through the semantic understanding function of the large language model, recall the corresponding proxy function based on the understood semantic information, and realize the correct calling of the interface document and the interface parameter by using the proxy function. Therefore, it is no longer necessary for the user to perform program configuration after understanding the interface document and the interface parameter of the application program, thereby reducing the understanding cost of the user and improving the efficiency of data processing.

[0162] Optionally, in the above Figure 4 Based on the corresponding embodiments, in another embodiment of the data processing apparatus 20 provided by the present application,

[0163] The acquisition module 201 is configured to acquire a training sample of the current application, wherein the training sample is a question and answer material constructed based on an interface document and an interface parameter of the current application.

[0164] The processing module 202 is configured to pre-adjust the large language model based on the training sample to obtain the target large language model.

[0165] In the embodiments of the present application, a data processing apparatus is provided. By using the above apparatus, the large language model obtained by general data pre-training is adjusted for the current application, which can improve the recognition accuracy of the large language model and thus improve the accuracy of data processing.

[0166] Optionally, in the above Figure 4On the basis of the corresponding embodiment, in another embodiment of the data processing apparatus 20 provided by the present application, the format of the training sample is json format, XML format or csv format.

[0167] In the embodiment of the present application, a data processing apparatus is provided. The above device provides various formats of training samples, thereby increasing the feasibility of the scheme.

[0168] Optionally, in the above Figure 4 On the basis of the corresponding embodiment, in another embodiment of the data processing apparatus 20 provided by the present application,

[0169] The processing module 202 is configured to determine the definition format of the proxy function set based on the target large language model.

[0170] The acquisition module 201 is configured to acquire the keyword calling function of the current application.

[0171] The processing module 202 is configured to register the proxy function set based on the keyword calling function and the definition format.

[0172] In the embodiment of the present application, a data processing apparatus is provided. The above device redefines the proxy function for the trained large language model and the current application scenario, which can improve the relevance between the proxy function and the large language model, and improve the calling accuracy of the proxy function, thereby improving the accuracy of data processing.

[0173] Optionally, in the above Figure 4 On the basis of the corresponding embodiment, in another embodiment of the data processing apparatus 20 provided by the present application,

[0174] The processing module 202 is configured to register the keyword calling function to the system tool based on the definition format, so as to generate the proxy function set.

[0175] In the embodiment of the present application, a data processing apparatus is provided. The above device registers the keyword calling function to the system tool of the application program, so that the proxy function can call the interface document and interface parameter of the application program more quickly, thereby increasing the efficiency of data processing.

[0176] Optionally, in the above Figure 4 On the basis of the corresponding embodiment, in another embodiment of the data processing apparatus 20 provided by the present application, the to-be-processed data is audio and video data, the current application is a cloud audio and video processing application, and the first to-be-processed text is used to describe the processing scheme of the audio and video data.

[0177] In the embodiments of the present application, a data processing device is provided. The above device provides a specific application scenario, thereby increasing the feasibility of the scheme.

[0178] Optionally, in the above Figure 4 Based on the embodiments corresponding thereto, in another embodiment of the data processing device 20 provided by the present application, the acquisition module 201 is configured to acquire a first keyword corresponding to the first semantic information;

[0179] The processing module 202 is configured to perform traversal on the agent function set based on the first keyword, and take the agent function set and the agent function hit by the first keyword as the first target agent function.

[0180] In the embodiments of the present application, a data processing device is provided. The above device determines the keyword extracted from the input text based on the semantic information, and then matches the corresponding target agent function from the agent function set based on the keyword, thereby improving the matching accuracy of the agent function.

[0181] Optionally, in the above Figure 4 Based on the embodiments corresponding thereto, in another embodiment of the data processing device 20 provided by the present application, the acquisition module 201 is configured to acquire a first keyword corresponding to the first semantic information;

[0182] The processing module 202 is configured to determine a first agent function based on the to-be-processed data when the first target agent function is not determined from the agent function set according to the first semantic information, the first agent function being predefined according to the interface document and the interface parameter of the current application; generate a second target program based on the corresponding interface document and interface parameter called by the first agent function; and process the to-be-processed data based on the second target program to obtain second output data.

[0183] In the embodiments of the present application, a data processing device is provided. The above device increases the scheme of the default agent function, which can increase the smoothness of the scheme execution, and can also improve the efficiency of data processing as much as possible.

[0184] Optionally, in the above Figure 4 Based on the embodiments corresponding thereto, in another embodiment of the data processing device 20 provided by the present application, the acquisition module 201 is configured to acquire a first keyword corresponding to the first semantic information;

[0185] The acquisition module 201 is configured to acquire a second to-be-processed text, the second to-be-processed text being different from the first to-be-processed text;

[0186] The processing module 202 is configured to call the target large language model to perform recognition processing on the second to-be-processed text to obtain second semantic information corresponding to the second to-be-processed text; determine a second target proxy function from the proxy function set according to the second semantic information; and call a corresponding interface document and interface parameter based on the second target proxy function to generate a third target program.

[0187] The acquisition module 201 is configured to acquire the to-be-processed data.

[0188] The processing module 202 is configured to call the third target program to process the to-be-processed data to obtain third output data.

[0189] In the embodiments of the present application, a data processing apparatus is provided. When the to-be-processed data needs to be processed multiple times, the data processing scheme combines the semantic understanding function of the large language model and the calling function of the proxy function for the interface document and the interface function, so that when the user calls the application program to process the data, the application program can understand the semantic information of the input text through the semantic understanding function of the large language model, recall the corresponding proxy function based on the understood semantic information, and realize the correct calling of the interface document and the interface parameter by using the proxy function. Therefore, the user no longer needs to configure the program after understanding the interface document and the interface parameter of the application program, thereby reducing the understanding cost of the user and improving the efficiency of data processing.

[0190] Optionally, in the above Figure 4 corresponding embodiments, as Figure 4 shown, another embodiment of the data processing apparatus 20 provided by the present application is provided, in which

[0191] The apparatus further includes a display module 203 configured to display a first interface, the first interface including a text input control.

[0192] The processing module 202 is configured to acquire the first to-be-processed text based on the text input control.

[0193] In the embodiments of the present application, a data processing apparatus is provided. The data processing scheme is refined into a scheme through an interactive interface, so that the data processing scheme can be executable.

[0194] Optionally, in the above Figure 5 corresponding embodiments, as Figure 4 shown, another embodiment of the data processing apparatus 20 provided by the present application is provided, in which

[0195] The apparatus further includes a display module 203 configured to display a second interface including a data input control;

[0196] The processing module 202 is configured to acquire the to-be-processed data based on the data input control.

[0197] In the embodiments of the present application, a data processing apparatus is provided. The data processing apparatus can refine a data processing scheme into a scheme performed through an interactive interface, so that the data processing scheme can be executable.

[0198] The data processing apparatus provided in the present application can be used in a server. Please refer to Figure 5 , Figure 6 is a schematic diagram of a server structure provided in the embodiments of the present application. The server 300 can have great differences due to different configurations or performances, and can include one or more central processing units (CPUs) 322 (for example, one or more processors) and a memory 332, one or more storage media 330 (for example, one or more mass storage devices) storing application programs 342 or data 344. The memory 332 and the storage media 330 can be temporary storage or persistent storage. The programs stored in the storage media 330 can include one or more modules (not shown in the figure), and each module can include a series of instruction operations in the server. Further, the central processing unit 322 can be configured to communicate with the storage media 330 and execute the series of instruction operations in the storage media 330 on the server 300.

[0199] The server 300 can further include one or more power supplies 326, one or more wired or wireless network interfaces 350, one or more input and output interfaces 358, and / or one or more operating systems 341, such as Windows Server TM , Mac OS X TM , Unix TM , Linux TM , FreeBSD TM , and the like.

[0200] The steps performed by the server in the above embodiments can be based on the server structure shown in Figure 6 .

[0201] The data processing apparatus provided in the present application can be used in a terminal device. Please refer to Figure 6, for the convenience of description, only parts related to the embodiments of the present application are shown, and specific technical details are not disclosed, please refer to the method part of the embodiments of the present application. In the embodiments of the present application, the terminal device is taken as a smart phone for example to be described:

[0202] Figure 7 The figure shows a block diagram of part of the structure of the smart phone related to the terminal device provided by the embodiments of the present application. Referring to Figure 7 , the smart phone includes: radio frequency (RF) circuit 410, memory 420, input unit 430, display unit 440, sensor 450, audio circuit 460, wireless fidelity (WiFi) module 470, processor 480, and power supply 490, etc. components. Those skilled in the art can understand, Figure 7 The structure of the smart phone shown in the embodiments of the present application does not constitute a limitation on the smart phone, and can include more or less components than the figure, or combine certain components, or different component arrangement.

[0203] The following will be combined Figure 7 The various constituent components of the smart phone will be specifically introduced:

[0204] The RF circuit 410 can be used for receiving and sending signals in the process of information or call, especially, receiving the downlink information of the base station and processing it by the processor 480; in addition, sending the uplink data to the base station. Usually, the RF circuit 410 includes but is not limited to antenna, at least one amplifier, transceiver, coupler, low noise amplifier (LNA), duplexer, etc. In addition, the RF circuit 410 can also communicate with the network and other devices through wireless communication. The above wireless communication can use any communication standard or protocol, including but not limited to global system for mobile communication (GSM), general packet radio service (GPRS), code division multiple access (CDMA), wideband code division multiple access (WCDMA), long term evolution (LTE), email, short message service (SMS), etc.

[0205] The memory 420 can be used to store software programs and modules, and the processor 480 can execute various function applications and data processing of the smart phone by running the software programs and modules stored in the memory 420. The memory 420 can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, at least one application program required by a function (such as a sound playing function, an image playing function, etc.), and the like; and the data storage area can store data created according to the use of the smart phone (such as audio data, a phone book, etc.), and the like. In addition, the memory 420 can include a high-speed random access memory, and can also include a non-volatile memory, for example, at least one magnetic disk storage device, a flash memory device, or other volatile solid-state memory device.

[0206] The input unit 430 can be used to receive inputted digital or character information, and to generate key signal input related to the user settings and function control of the smart phone. Specifically, the input unit 430 can include a touch panel 431 and other input devices 432. The touch panel 431, also known as a touch screen, can collect a user's touch operation (such as the user's operation on or near the touch panel 431 using a finger, a stylus, or any suitable object or accessory) and drive the corresponding connection device according to the pre-set program. Optionally, the touch panel 431 can include two parts of a touch detection device and a touch controller. Among them, the touch detection device detects the user's touch position and detects the signal brought by the touch operation, and transmits the signal to the touch controller; the touch controller receives the touch information from the touch detection device, and converts it into touch coordinates, and then sends it to the processor 480, and can also receive the command from the processor 480 and execute it. In addition, the touch panel 431 can be implemented in various types such as resistance type, capacitance type, infrared type, and surface acoustic wave type. In addition to the touch panel 431, the input unit 430 can also include other input devices 432. Specifically, the other input devices 432 can include one or more of a physical keyboard, a function key (such as a volume control key, an on-off key, etc.), a trackball, a mouse, a joystick, and the like.

[0207] The display unit 440 can be used to display information input by a user or information provided to the user as well as various menus of the smartphone. The display unit 440 can include a display panel 441, which can optionally be configured in the form of a liquid crystal display (LCD), an organic light-emitting diode (OLED), etc. Further, a touch panel 431 can be overlaid on the display panel 441, which, when detecting a touch operation thereon or in the vicinity thereof, transmits to the processor 480 to determine the type of touch event, and then the processor 480 provides a corresponding visual output on the display panel 441 according to the type of touch event. Although in the above description, the touch panel 431 and the display panel 441 are implemented as two independent components to realize the input and output functions of the smartphone, in some embodiments, the touch panel 431 and the display panel 441 can be integrated to realize the input and output functions of the smartphone. Figure 7

[0208] The smartphone can further include at least one sensor 450, such as a light sensor, a motion sensor, and other sensors. Specifically, the light sensor can include an ambient light sensor and a proximity sensor, wherein the ambient light sensor can adjust the brightness of the display panel 441 according to the brightness of ambient light, and the proximity sensor can turn off the display panel 441 and / or the backlight when the smartphone is moved to the ear. As one of the motion sensors, the accelerometer sensor can detect the magnitude of acceleration in each direction (generally three axes), and when at rest, it can detect the magnitude and direction of gravity, which can be used for applications that identify the posture of the smartphone (such as landscape / portrait screen switching, related games, magnetometer posture calibration), vibration recognition related functions (such as pedometer, knock), etc. As for other sensors that the smartphone can be configured, such as a gyroscope, a barometer, a hygrometer, a thermometer, an infrared sensor, etc., they will not be described here.

[0209] The audio circuit 460, the speaker 461, and the microphone 462 can provide an audio interface between the user and the smartphone. The audio circuit 460 can convert received audio data into an electrical signal, which is transmitted to the speaker 461 to be converted into a sound signal for output. On the other hand, the microphone 462 converts the collected sound signal into an electrical signal, which is received by the audio circuit 460 to be converted into audio data, which is then output to the processor 480 for processing, and then transmitted to another smartphone, for example, via the RF circuit 410, or output to the memory 420 for further processing.

[0210] ​WiFi belongs to short distance wireless transmission technology, and the smart phone can help users send and receive emails, browse web pages and access streaming media through the WiFi module 470, which provides wireless broadband Internet access for users. Although Figure 7 The WiFi module 470 is shown, but it is understood that it does not belong to the necessary structure of the smart phone, and can be omitted as needed without changing the essence of the application.

[0211] The processor 480 is the control center of the smart phone, which connects all parts of the smart phone through various interfaces and lines, executes various functions of the smart phone and processes data by running or executing software programs and / or modules stored in the memory 420 and calling data stored in the memory 420, thereby monitoring the whole smart phone. Optionally, the processor 480 can include one or more processing units; optionally, the processor 480 can integrate an application processor and a modem processor, wherein the application processor mainly processes the operating system, user interface and application program, and the modem processor mainly processes wireless communication. It can be understood that the above-mentioned modem processor can also not be integrated into the processor 480.

[0212] The smart phone also includes a power supply 490 (such as a battery) for supplying power to various components, and the power supply can be logically connected to the processor 480 through a power management system, so as to realize the functions of managing charging, discharging and power consumption management through the power management system.

[0213] Although not shown, the smart phone can also include a camera, a Bluetooth module, etc., which will not be described here.

[0214] The steps performed by the terminal device in the above embodiments can be based on the terminal device structure shown in the Figure 7 Figure 7 terminal device structure.

[0215] The computer readable storage medium in the embodiments of the present application stores a computer program, which makes the computer execute the method described in the above embodiments when it runs on the computer.

[0216] The computer program product in the embodiments of the present application includes a program, which makes the computer execute the method described in the above embodiments when it runs on the computer.

[0217] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the above-described system, device and unit can refer to the corresponding process in the foregoing method embodiments, which will not be described here.

[0218] In several embodiments provided in the present application, it should be understood that the disclosed system, device and method can be implemented in other manners. For example, the described device embodiments are merely schematic. For example, the division of the units is only a logical function division. There can be another division manner for the actual implementation, for example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the displayed or discussed mutual couplings or direct couplings or communication connections between different units, can be indirect couplings or communication connections through some interfaces, devices or units, and can be in electrical, mechanical or other forms.

[0219] The units described as separate components can or can not be physically separate, and the components shown as units can or can not be physical units, i.e., can be located in one place, or can be distributed on multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the embodiment.

[0220] In addition, each functional unit in each embodiment of the present application can be integrated into a processing unit, or each unit can exist physically, or two or more units can be integrated into one unit. The integrated unit can be realized in the form of hardware or in the form of a software functional unit.

[0221] The integrated unit, if realized in the form of a software functional unit and sold or used as an independent product, can be stored in a computer readable storage medium. Based on this understanding, the technical solutions of the present application essentially or the part that contributes to the prior art, or all or part of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium, and includes a number of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the embodiments of the present application. The aforementioned storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and various media that can store program codes.

[0222] The above embodiments are merely used to describe the technical solutions of the present application, but not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: it can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacements for some technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A data processing method, characterized in that: include: Get the first text to be processed; Invoking a target large language model to perform recognition processing on the first to-be-processed text to obtain first semantic information corresponding to the first to-be-processed text, wherein the target large language model is a large language model pre-adjusted for the current application; confirming a first target proxy function from a proxy function set according to the first semantic information, wherein each proxy function in the proxy function set is predefined according to an interface document and interface parameters of the current application; Generate a first target program based on the first target proxy function calling the target interface document and the target interface parameters; Get the data to be processed; The first target program is called to process the data to be processed to obtain first output data.

2. The method according to claim 1, characterized in that The method further comprises: Acquire a training sample of the current application, where the training sample is a question-and-answer material constructed based on the interface document and interface parameters of the current application; The large language model is pre-adjusted based on the training sample to obtain the target large language model.

3. The method according to claim 2, characterized in that The training sample is in json format, XML format or csv format.

4. The method according to claim 2, characterized in that The method further comprises: Determining a definition format of the proxy function set based on the target large language model; Obtaining a keyword calling function of the current application; The proxy function set is generated based on the keyword calling function and the definition format registration.

5. The method according to claim 4, characterized in that The generating of the proxy function set based on the keyword calling function and the definition format registration includes: The keyword call function is registered to a system tool based on the definition format to generate the proxy function set.

6. The method according to any one of claims 1 to 5, characterized in that The data to be processed is audio and video data, the current application is a cloud audio and video processing application, and the first text to be processed is used to describe a processing solution for the audio and video data.

7. The method according to any one of claims 1 to 5, characterized in that The determining of a first target proxy function from a set of proxy functions according to the first semantic information includes: Obtaining a first keyword corresponding to the first semantic information; The proxy function set is traversed based on the first keyword, and a proxy function that matches the proxy function set and the first keyword is used as the first target proxy function.

8. The method according to any one of claims 1 to 5, characterized in that The method further comprises: When the first target proxy function is not determined from the proxy function set according to the first semantic information, determining a first proxy function according to the data to be processed, where the first proxy function is predefined according to the interface document and interface parameters of the current application; Generate a second target program based on the interface document and interface parameters corresponding to the first proxy function call; The data to be processed is processed based on the second target program to obtain second output data.

9. The method according to any one of claims 1 to 5, characterized in that After calling the target program to process the data to be processed to obtain first output data, the method further includes: Acquire a second text to be processed, where the second text to be processed is different from the first text to be processed; Calling the target large language model to perform recognition processing on the second to-be-processed text to obtain second semantic information corresponding to the second to-be-processed text; determining a second target proxy function from the proxy function set according to the second semantic information; Generate a third target program based on the interface document and interface parameters corresponding to the second target proxy function call; Acquiring the data to be processed; The third target program is called to process the data to be processed to obtain third output data.

10. The method according to any one of claims 1 to 5, characterized in that The obtaining of the first to-be-processed text comprises: Displaying a first interface, wherein the first interface includes a text input control; The first to-be-processed text is obtained based on the text input control.

11. The method according to any one of claims 1 to 5, characterized in that The obtaining of data to be processed includes: Displaying a second interface, wherein the second interface includes a data input control; The data to be processed is obtained based on the data input control.

12. A data processing device, characterized in that: include: An acquisition module, configured to acquire a first text to be processed; a processing module configured to call a target large language model to perform recognition processing on the first to-be-processed text to obtain first semantic information corresponding to the first to-be-processed text, wherein the target large language model is a large language model pre-adjusted for the current application; identify a first target proxy function from a proxy function set based on the first semantic information, each proxy function in the proxy function set being pre-defined based on an interface document and interface parameters of the current application; and generate a first target program based on calling a target interface document and target interface parameters based on the first target proxy function; The acquisition module is used to acquire the data to be processed; The processing module is used to call the first target program to process the data to be processed to obtain first output data.

13. A computer device, characterized in that: include: Memory, processor, and bus system; Wherein, the memory is used to store programs; The processor is configured to execute the program in the memory, and the processor is configured to execute the method according to any one of claims 1 to 11 according to instructions in the program code; The bus system is used to connect the memory and the processor so that the memory and the processor can communicate with each other.

14. A computer-readable storage medium comprising instructions, which, when executed on a computer, causes the computer to perform the method according to any one of claims 1 to 11.

15. A computer program product comprising a computer program, characterized in that The computer program is used by a processor to execute the method according to any one of claims 1 to 11.

Images