Business model acquisition method and device based on artificial intelligence and readable storage medium

Abstract

The invention provides a business model acquisition method based on artificial intelligence, which is applied to a business end, and comprises the following steps: uploading training corpora following a preset rule to a model training system, setting training parameters including training round number, batch size and learning rate, and sending a training instruction to enable the system to train a business model according to the corpora and the parameters. After a training completion instruction is received, a verification data set is uploaded, the system restarts the service model and processes verification data, and the service end finally obtains a processing result. According to the method, data input is standardized through a source, and training interruption risks are reduced; the service end autonomously sets parameters and instructions, so that training is more autonomously and pertinently; in the whole process, the automatic and standardized operation of the business model from training to verification is realized, the efficiency and accuracy of model training are improved, and the dependence on professional technicians is reduced.

Description

Technical Field

[0001] This application belongs to the field of deep learning, and in particular relates to a business model acquisition method and device based on artificial intelligence and a readable storage medium. Background Art

[0002] In today's digital age, artificial intelligence (AI) technology is widely used in various fields. Natural language processing (NLP)-related AI models play a key role in scenarios such as intelligent customer service, information retrieval, and text classification. With the continuous growth and diversification of business needs, higher requirements are placed on the performance, accuracy, and adaptability of AI models. Traditional AI model training methods often rely on professional technicians to manually write large amounts of code to process data, adjust parameters, and execute training tasks. This not only requires extremely high professional knowledge reserves, but also is extremely inefficient when faced with massive amounts of data and complex model structures, making it difficult to quickly respond to dynamic business needs.

[0003] While some AI model training tools exist on the market, most suffer from a lack of flexibility. These tools are typically designed for specific models or application scenarios. When companies need to add different types of models (such as expanding from semantic matching models to intent recognition models or named entity recognition models), they often need to rebuild the training environment or even replace the training tools, significantly increasing time and resource costs. Furthermore, these tools lack convenient and intuitive visual evaluation methods for model verification, making it difficult for business personnel to quickly and accurately understand the model's performance, hindering the timely identification and optimization of model issues.

[0004] In addition, in the existing AI model training process, the connection between data management and training tasks is not smooth enough. On the one hand, the data upload and format verification functions are not perfect, which can easily lead to training interruptions due to data format errors, affecting training efficiency; on the other hand, the adjustment of training parameters lacks intelligence and convenience, and often requires manual repeated attempts at different parameter combinations, which is time-consuming and labor-intensive. At the same time, in the allocation and management of training resources, hardware resources such as GPUs cannot be automatically and flexibly allocated according to model training requirements, resulting in resource waste or stagnation of training tasks due to insufficient resources. Therefore, there is an urgent need for an automated, flexible and efficient AI model training platform to solve the above problems. Summary of the Invention

[0005] The purpose of the present invention is to provide a business model acquisition method based on artificial intelligence, aiming to solve the problem that existing model training is too complex and cannot be implemented by business personnel. The business model acquisition method based on artificial intelligence provided by this application includes:

[0006] The method is applied to the business end;

[0007] Uploading training corpus to the model training system, wherein the training corpus follows predetermined rules;

[0008] Setting training parameters, including the number of training rounds, batch size, and learning rate;

[0009] Sending a training instruction so that the model training system performs model training based on the training corpus and the training parameters;

[0010] Receiving a training completion instruction sent by the model training system, wherein the training completion instruction indicates that the model training system has completed the training of the business model;

[0011] Uploading a verification data set so that the model training system restarts the business model and uses the business model to process the verification data set;

[0012] Obtaining a processing result of the verification data set.

[0013] Based on the artificial intelligence-based business model acquisition method provided in the first aspect of the embodiment of the present application, optionally, sending a training instruction includes:

[0014] Send a save parameter instruction to the model training system so that the model training system executes the training process according to the memory situation.

[0015] Based on the artificial intelligence-based business model acquisition method provided in the first aspect of the embodiment of the present application, optionally, the method further includes:

[0016] A retraining instruction is received, and the business model is retrained based on the processing result.

[0017] A second aspect of an embodiment of the present application provides a method for acquiring a business model based on artificial intelligence, which is applied to a model training system and includes:

[0018] Receive training data and training parameters sent by the service end, wherein the training data follows a predetermined rule, and the training parameters include the number of training rounds, batch size, and learning rate;

[0019] Receive the training instruction sent by the business end, perform model training according to the training corpus and training parameters, and obtain the business model;

[0020] Obtaining a verification data set, restarting the business model, and processing the verification data set using the business model to obtain a verification result;

[0021] Return the verification result to the service end.

[0022] Based on the artificial intelligence-based business model acquisition method provided in the second aspect of the embodiment of the present application, optionally, the model training system adopts a python plug-in addition and deletion model.

[0023] A third aspect of the embodiments of the present application provides a business model acquisition device based on artificial intelligence, which is applied to a business end and includes:

[0024] A training corpus uploading unit, used to upload training corpus to the model training system, wherein the training corpus follows a predetermined rule;

[0025] A training parameter setting unit, configured to set training parameters, including the number of training rounds, batch size, and learning rate;

[0026] A training instruction sending unit, configured to send a training instruction so that the model training system performs model training based on the training corpus and the training parameters;

[0027] A receiving unit, configured to receive a training completion instruction sent by the model training system; the training completion instruction indicates that the model training system has completed the training of the business model;

[0028] A verification data set uploading unit, configured to upload a verification data set so that the model training system restarts the business model and processes the verification data set using the business model;

[0029] The result acquisition unit is used to obtain the processing result of the verification data set.

[0030] A fourth aspect of the embodiments of the present application provides a business model acquisition device based on artificial intelligence, which is applied to a model training system and includes:

[0031] A training corpus and training parameter receiving unit, configured to receive training corpus and training parameters sent by a service end, wherein the training corpus follows a predetermined rule, and the training parameters include the number of training rounds, batch size, and learning rate;

[0032] A training instruction receiving unit is used to receive a training instruction sent by a service end, perform model training based on the training corpus and training parameters, and obtain a service model;

[0033] A verification unit, configured to obtain a verification data set, restart the business model, and process the verification data set using the business model to obtain a verification result;

[0034] The verification result returning unit is used to return the verification result to the service end.

[0035] A fifth aspect of the embodiments of the present application provides a business model acquisition device based on artificial intelligence, including:

[0036] CPU, memory, input and output interfaces, wired or wireless network interfaces, and power supply;

[0037] The memory is a transient storage memory or a persistent storage memory;

[0038] The central processing unit is configured to communicate with the memory and execute instruction operations in the memory on the device to perform the method as described in any one of the first aspects of the embodiments of the present application.

[0039] A sixth aspect of the embodiments of the present application provides a business model acquisition device based on artificial intelligence, including:

[0040] CPU, memory, input and output interfaces, wired or wireless network interfaces, and power supply;

[0041] The memory is a transient storage memory or a persistent storage memory;

[0042] The central processing unit is configured to communicate with the memory and execute instruction operations in the memory on the device to perform the method as described in any one of the second aspects of the embodiments of the present application.

[0043] A seventh aspect of an embodiment of the present application provides a computer-readable storage medium, characterized in that it includes instructions, which, when executed on a computer, enable the computer to execute the method described in any one of the first and second aspects of the embodiment of the present application.

[0044] It can be seen from the above technical solutions that the embodiments of the present application have the following advantages: The present application provides a method for obtaining a business model based on artificial intelligence, characterized in that the method is applied to the business end, including: uploading training corpus to the model training system, the training corpus follows predetermined rules; setting training parameters, the training parameters include the number of training rounds, batch size and learning rate; sending training instructions so that the model training system performs model training based on the training corpus and the training parameters; receiving a training completion instruction sent by the model training system; the training completion instruction indicates that the model training system has completed the training of the business model, uploading a verification data set, so that the model training system restarts the business model, and uses the business model to process the verification data set; obtaining the processing result of the verification data set. By uploading training corpus that follows predetermined rules on the business end, standardizing data input from the source, and reducing the risk of training interruption caused by problems such as data format; the business end independently sets training parameters and sends training instructions, making the training process more autonomous and targeted, and can be flexibly adjusted according to business needs. After the training is completed, the verification data set is uploaded to restart the business model and process the data, achieving an orderly connection between the training and verification links and finally obtaining the processing results. The entire process realizes the automation and standardization of the business model from training to verification, improves the efficiency and accuracy of model training, and reduces dependence on professional technicians. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] To more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are merely embodiments of the present application. A person of ordinary skill in the art can also derive other drawings based on the provided drawings without inventive effort. It should be understood that the drawings provided in this section are only used to better understand the present solution and do not constitute a limitation of the present application.

[0046] Figure 1 A flowchart of an embodiment of the artificial intelligence-based business model acquisition method provided in this application.

[0047] Figure 2 This is a schematic diagram of the training corpus and training parameter submission interface in the artificial intelligence-based business model acquisition method provided in this application.

[0048] Figure 3 A schematic diagram of the training result list in the artificial intelligence-based business model acquisition method provided in this application.

[0049] Figure 4 This is another flowchart of an embodiment of the artificial intelligence-based business model acquisition method provided in this application.

[0050] Figure 5 This is a structural diagram of an embodiment of the artificial intelligence-based business model acquisition device provided in this application.

[0051] Figure 6 This is another structural diagram of an embodiment of the artificial intelligence-based business model acquisition device provided in this application.

[0052] Figure 7 This is another structural diagram of an embodiment of the artificial intelligence-based business model acquisition device provided in this application.

[0053] Figure 8 This is another structural diagram of an embodiment of the artificial intelligence-based business model acquisition device provided in this application. DETAILED DESCRIPTION

[0054] In order to enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application are clearly and completely described below. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work should fall within the scope of protection of this application. At the same time, for the sake of clarity and conciseness, the description of well-known functions and structures is omitted in the following description.

[0055] The terms "first," "second," "third," "fourth," and the like (if any) in the specification and claims of this application and in the accompanying drawings are used to distinguish similar objects and are not necessarily used to describe a particular order or precedence. It should be understood that the terms used in this manner are interchangeable where appropriate so that the embodiments described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "including" and "having," and any variations thereof, are intended to cover non-exclusive inclusions, e.g., a process, method, product, or apparatus comprising a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such process, method, product, or apparatus.

[0056] In today's digital era, artificial intelligence technology is widely used in various fields, among which natural language processing (NLP) related AI models play a key role in scenarios such as intelligent customer service, information retrieval, text classification, etc. With the growing and diversifying business needs, higher requirements are placed on the performance, accuracy and adaptability of AI models. Traditional AI model training methods often rely on professional technicians to manually write a large amount of code to process data, adjust parameters and perform training tasks, which not only requires a high level of professional knowledge, but also is extremely inefficient when faced with massive data and complex model structures, making it difficult to quickly respond to the dynamic needs of the business.

[0057] Currently, although there are some AI model training tools on the market, they often lack flexibility. These tools are usually designed for specific types of models or specific application scenarios. When a company needs to add different types of models (such as expanding from semantic matching models to intent recognition models, named entity recognition models, etc.), it often needs to rebuild the training environment or even replace the training tools, resulting in a significant increase in time and resource costs. Moreover, these tools lack convenient and intuitive visualization evaluation methods in the model verification stage, making it difficult for business personnel to quickly and accurately understand the performance of the model, which is not conducive to timely identifying problems in the model and optimizing it.

[0058] In addition, in the existing AI model training process, the connection between data management and training tasks is not smooth. On the one hand, data uploading and format checking functions are not perfect, which can easily lead to training interruption due to data format errors, affecting training efficiency; on the other hand, the adjustment of training parameters lacks intelligence and convenience, often requiring manual repeated attempts of different parameter combinations, which is time-consuming and labor-intensive. At the same time, in the allocation and management of training resources, it is not possible to automatically and flexibly allocate GPU and other hardware resources according to the model training needs, resulting in resource waste or training tasks being stalled due to insufficient resources. Therefore, there is an urgent need for an AI model automated training platform with automation, flexibility and efficiency to solve the above problems.

[0059] 101. uploading training corpus to the model training system.

[0060] Upload training corpus to the model training system, and the training corpus follows predetermined rules. This solution is applied to the business end, that is, the business personnel trigger the execution of this method, and the business personnel upload the training corpus. The training corpus is organized by the business personnel according to business needs and follows predetermined rules. The business end triggers the file upload function through the visual interface provided by the platform (such as a web page or client tool), which supports dragging or selecting local files (formats include structured data such as CSV, JSON, or text data such as TXT). The system presets data verification rules (such as field integrity and format standardization). The "predetermined rules" can be customized and extended by the business end, for example, by adding field verification logic through the "data template configuration" function provided by the platform, or integrating third-party data specifications, which are not limited here. The actual user can be a decoration business personnel, used to generate a telephone demand verification model. Telephone demand verification refers to confirming customer needs (such as order information, service details, qualification review, etc.) through telephone communication. It is necessary to quickly and accurately identify key information in the customer's statement (such as name, contact information, product model, demand type, etc.), and judge whether it complies with business rules (such as whether the qualifications are complete and whether the demand is reasonable). As front-line practitioners, decoration business personnel can better collect these questions and corresponding response statements.

[0061] 102. Set training parameters.

[0062] Set the training parameters, including the number of training rounds, batch size, and learning rate. The business end sets parameters through input boxes, drop-down menus, and other components in the parameter configuration interface. The interface provides real-time help prompts (such as "Number of training rounds (epochs): the number of times the model traverses the training data, recommended value 5-50") when the mouse is hovered). The parameter range can be dynamically adjusted according to the model type, and the training parameters should be explained so that business personnel can understand. Specifically, the interface for adding training corpus and training parameters can be referred to Figure 2 .

[0063] 103. Send training instructions.

[0064] Send training instructions so that the model training system performs model training based on the training corpus and the training parameters. The model training system is a server system with higher computing power. After receiving the instructions, it automatically generates a training script for training based on the training corpus path, parameter configuration and selected model type, and uses the basic model and the above content for training. Specifically, for the core scenario model in the decoration field, it can extract the decoration demand features in the training corpus, and conduct deep learning to form a business model with corresponding capabilities.

[0065] It is understood that a save parameter instruction can also be sent to the model training system, so that the model training system can execute the training process based on the memory situation. That is, when memory is insufficient, the training process is queued. Model types are not limited to preset types, and new models (such as custom classification models) can be dynamically added through a "plug-in architecture."

[0066] 104. Receive a training completion instruction sent by the model training system.

[0067] Receive the training completion instruction sent by the model training system, which indicates that the model training system has completed the training of the business model. After the training task is completed (normal completion or abnormal interruption), the model training system sends a status notification to the business end. The training completion instruction includes: task ID (; training status (success / failure / interruption); key indicators (such as accuracy 0.89, training time 2 hours and 15 minutes), etc. The interface for returning the instruction can be referred to Figure 3 .

[0068] 105. Upload the validation dataset.

[0069] Upload the verification data set so that the model training system restarts the business model and uses the business model to process the verification data set. The business end uploads the verification data set through the same interface or interface as the training corpus (the format is consistent with the training corpus, but the label field can be empty or only used for evaluation). After the model training system receives the verification data, it automatically triggers the verification process: restart the verification service container to ensure that the environment is clean (such as clearing the cache and resetting parameters), and use the verification data set to process the model and output the prediction results.

[0070] 106. Obtain the processing result of the verification data set.

[0071] Based on the processing structure, business users can decide whether to use the model or further retrain it. Specifically, a result prompt can be set: if the accuracy rate is ≥ a preset threshold (such as 0.85), a "model can be deployed" prompt will be displayed; if the accuracy rate is less than the threshold, low-confidence samples (such as confidence level < 0.5) will be automatically highlighted to facilitate business personnel to analyze data deviations. Business personnel can also issue retraining instructions based on the training results, so that the model training system can retrain the business model based on the processing results.

[0072] It can be seen from the above technical solutions that the embodiments of the present application have the following advantages: The present application provides a method for obtaining a business model based on artificial intelligence, characterized in that the method is applied to the business end, including: uploading training corpus to the model training system, the training corpus follows predetermined rules; setting training parameters, the training parameters include the number of training rounds, batch size and learning rate; sending training instructions so that the model training system performs model training based on the training corpus and the training parameters; receiving a training completion instruction sent by the model training system; the training completion instruction indicates that the model training system has completed the training of the business model, uploading a verification data set, so that the model training system restarts the business model, and uses the business model to process the verification data set; obtaining the processing result of the verification data set. By uploading training corpus that follows predetermined rules on the business end, standardizing data input from the source, and reducing the risk of training interruption caused by problems such as data format; the business end independently sets training parameters and sends training instructions, making the training process more autonomous and targeted, and can be flexibly adjusted according to business needs. After the training is completed, the verification data set is uploaded to restart the business model and process the data, achieving an orderly connection between the training and verification links and finally obtaining the processing results. The entire process realizes the automation and standardization of the business model from training to verification, improves the efficiency and accuracy of model training, and reduces dependence on professional technicians.

[0073] The above content describes this method on the business side. The following describes this solution on the model training system side. Figure 4 The artificial intelligence-based business model acquisition method provided in this application includes: steps 401 to 404.

[0074] 401. Receive training corpus and training parameters sent by the business end.

[0075] Receive the training corpus and training parameters sent by the business end. The training corpus follows predetermined rules. The training parameters include the number of training rounds, batch size, and learning rate. Specifically, the model training system obtains the training corpus and training parameters sent by the business end. The training corpus must follow a predefined data structure (such as JSON format) and include input text, labels, and other business-related fields.

[0076] During actual implementation, you can also verify the training corpus and training parameters, namely verifying field integrity (such as whether text and label fields are included) and data types (such as whether numeric parameters are valid floating-point numbers). You can also check parameter ranges and the validity of parameter combinations, etc. The specifics are not limited here.

[0077] 402. Receive training instructions sent by the business end.

[0078] Receive the training instructions sent by the business end, perform model training based on the training corpus and training parameters, and obtain the business model; after receiving the training instructions, the model training system loads the training corpus and parameters from the storage system, and initializes the model architecture (such as BERT based on Transformer's bidirectional encoder representation, LSTM long short-term memory network, etc.). It can be understood that the model type and architecture used in the actual implementation process can be determined according to the actual situation and are not limited here. Start the training process. Key indicators (such as loss value and accuracy) are recorded during the training process. The model training system adopts a python plug-in addition and deletion model, based on Python's plug-in architecture, which supports flexible addition and deletion of models without modifying the core code. This architecture adopts a modular design and realizes the expansion of the model through a unified interface and plug-in mechanism.

[0079] 403. Obtain a verification data set, restart the business model, and use the business model to process the verification data set to obtain a verification result.

[0080] After training is completed, the system loads the validation data set, restarts the business model (releases training resources and initializes the inference environment), performs batch inference on the validation data, and calculates evaluation indicators. Specifically, the validation data set is obtained from the business end (the format must be consistent with the training corpus). The model is restarted to release the GPU resources occupied during training, reload the model, initialize the inference optimization configuration, and generate detailed results. Another advantage of model restart is to ensure the consistency between the inference environment and the production environment: by completely releasing the training environment and reinitializing the inference environment, the temporary states remaining in the training process (such as gradient accumulation, batch normalization statistics, etc.) can be eliminated, ensuring that the behavior of the model in the verification phase is exactly the same as after deployment to the production environment. This avoids the problem of inflated verification results due to environmental differences, improves the credibility of the verification results, and provides a more reliable reference for model launch.

[0081] 404. Return the verification result to the business end.

[0082] The model training system returns the verification results to the business end, supporting both active push and passive query (such as obtaining through the API). The verification results are stored in the database and can be queried by task ID.

[0083] It can be seen from the above technical solutions that the embodiments of the present application have the following advantages: The present application provides a method for obtaining a business model based on artificial intelligence, characterized in that the method is applied to the business end, including: uploading training corpus to the model training system, the training corpus follows predetermined rules; setting training parameters, the training parameters include the number of training rounds, batch size and learning rate; sending training instructions so that the model training system performs model training based on the training corpus and the training parameters; receiving a training completion instruction sent by the model training system; the training completion instruction indicates that the model training system has completed the training of the business model, uploading a verification data set, so that the model training system restarts the business model, and uses the business model to process the verification data set; obtaining the processing result of the verification data set. By uploading training corpus that follows predetermined rules on the business end, standardizing data input from the source, and reducing the risk of training interruption caused by problems such as data format; the business end independently sets training parameters and sends training instructions, making the training process more autonomous and targeted, and can be flexibly adjusted according to business needs. After the training is completed, the verification data set is uploaded to restart the business model and process the data, achieving an orderly connection between the training and verification links and finally obtaining the processing results. The entire process realizes the automation and standardization of the business model from training to verification, improves the efficiency and accuracy of model training, and reduces dependence on professional technicians.

[0084] The above content describes the method provided by this application. To support the implementation of the above embodiment, this application also provides a business model acquisition device based on artificial intelligence, which is applied to the business end. Figure 5 An embodiment of the business model acquisition device based on artificial intelligence provided by this application includes:

[0085] A training corpus uploading unit 501 is used to upload training corpus to the model training system, where the training corpus follows a predetermined rule;

[0086] A training parameter setting unit 502 is used to set training parameters, including the number of training rounds, batch size, and learning rate;

[0087] A training instruction sending unit 503 is used to send a training instruction so that the model training system performs model training based on the training corpus and the training parameters;

[0088] The receiving unit 504 is configured to receive a training completion instruction sent by the model training system; the training completion instruction indicates that the model training system has completed the training of the business model;

[0089] A verification data set uploading unit 505 is configured to upload a verification data set so that the model training system restarts the business model and processes the verification data set using the business model;

[0090] The result acquisition unit 506 is configured to obtain the processing result of the verification data set.

[0091] In this embodiment, the processes performed by each unit in the device are the same as those described above. Figure 1 The method flow described in the corresponding embodiment is similar and will not be repeated here.

[0092] This application also provides an artificial intelligence-based business model acquisition device, which is applied to a model training system and includes:

[0093] The training corpus and training parameter receiving unit 601 is used to receive training corpus and training parameters sent by the service end, wherein the training corpus follows a predetermined rule and the training parameters include the number of training rounds, batch size and learning rate;

[0094] The training instruction receiving unit 602 is used to receive the training instruction sent by the service end, perform model training according to the training corpus and training parameters, and obtain a service model;

[0095] A verification unit 603 is configured to obtain a verification data set, restart the business model, and process the verification data set using the business model to obtain a verification result;

[0096] The verification result returning unit 604 is configured to return the verification result to the service end.

[0097] In this embodiment, the processes performed by each unit in the device are the same as those described above. Figure 4 The method flow described in the corresponding embodiment is similar and will not be repeated here.

[0098] Figure 7 It is a structural diagram of an artificial intelligence-based business model acquisition device provided in an embodiment of the present application. The artificial intelligence-based business model acquisition device 700 may include one or more central processing units (CPU) 701 and a memory 705, and the memory 705 stores one or more applications or data.

[0099] In this embodiment, the specific functional module division in the central processing unit 701 can be the same as the above Figure 5 The functional module division method of each unit described in is similar and will not be repeated here.

[0100] Memory 705 may be volatile or persistent storage. The program stored in memory 705 may include one or more modules, each of which may include a series of instructions for operating on the server. Furthermore, CPU 701 may be configured to communicate with memory 705 and execute the series of instructions in memory 705 on server 700.

[0101] The artificial intelligence-based business model acquisition device 700 may further include one or more power supplies 702 , one or more wired or wireless network interfaces 703 , and one or more input and output interfaces 704 .

[0102] The CPU 701 can execute the aforementioned Figure 1 The operations performed by the artificial intelligence-based business model acquisition method in the illustrated embodiment will not be described in detail here.

[0103] Figure 8 It is a structural diagram of an artificial intelligence-based business model acquisition device provided in an embodiment of the present application. The artificial intelligence-based business model acquisition device 800 may include one or more central processing units (CPU) 801 and a memory 805, and the memory 805 stores one or more applications or data.

[0104] In this embodiment, the specific functional module division in the central processing unit 801 can be the same as the above Figure 5 The functional module division method of each unit described in is similar and will not be repeated here.

[0105] Memory 805 may be volatile or persistent storage. The program stored in memory 805 may include one or more modules, each of which may include a series of instructions for operating on the server. Furthermore, the central processing unit 801 may be configured to communicate with memory 805 and execute the series of instructions in memory 805 on the server 800.

[0106] The artificial intelligence-based business model acquisition device 800 may further include one or more power supplies 802 , one or more wired or wireless network interfaces 803 , and one or more input and output interfaces 804 .

[0107] The CPU 801 can execute the aforementioned Figure 4 The operations performed by the artificial intelligence-based business model acquisition method in the illustrated embodiment will not be described in detail here.

[0108] The embodiment of the present application further provides a computer storage medium for storing computer software instructions for the above-mentioned artificial intelligence-based business model acquisition method, which comprises a program designed for the artificial intelligence-based business model acquisition method.

[0109] The artificial intelligence-based business model acquisition method can be the artificial intelligence-based business model acquisition method described in the foregoing Figure 1 or Figure 4 .

[0110] The embodiment of the present application further provides a computer program product comprising computer software instructions, which can be loaded by a processor to implement the process of the artificial intelligence-based business model acquisition method in any one of the above-mentioned Figure 1 Figure 4 .

[0111] 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 above-mentioned device embodiments are merely schematic, and the equivalent changes of the circuit, the division of the units, and the logical function division are merely a logical function division, and other division manners can be adopted in actual implementation, for example, a plurality of 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 coupling or direct coupling or communication connection can be indirect coupling or communication connection through some interfaces, devices or units, and can be electrical, mechanical or other forms.

[0112] The units described as separate components can or can not be physically separate, and the components displayed as units can or can not be physical units, that is, can be located in one place, or can be distributed on a plurality of units. Part or all of the units can be selected according to actual needs to achieve the purpose of the embodiment.

[0113] In addition, each functional unit in the various embodiments of the present application can be integrated in one processing unit, or each unit can be physically present separately, or two or more units can be integrated in one unit. The integrated unit can be realized in the form of hardware or in the form of a software functional unit.

[0114] The above-mentioned is only the preferred embodiment of the present application, and is not used to limit the present application, and any modification, equivalent replacement or improvement made within the spirit and principle of the present application should be included in the protection scope of the present application.

Claims

1. A business model acquisition method based on artificial intelligence, characterized in that: The method is applied to the business end and includes: Uploading training corpus to the model training system, wherein the training corpus follows predetermined rules; Setting training parameters, including the number of training rounds, batch size, and learning rate; Sending a training instruction so that the model training system performs model training based on the training corpus and the training parameters; Receiving a training completion instruction sent by the model training system, wherein the training completion instruction indicates that the model training system has completed the training of the business model; Uploading a verification data set so that the model training system restarts the business model and uses the business model to process the verification data set; Obtaining a processing result of the verification data set.

2. The method for acquiring a business model based on artificial intelligence according to claim 1, characterized in that: The sending of the training instruction includes: Send a save parameter instruction to the model training system so that the model training system executes the training process according to the memory situation.

3. The method for acquiring a business model based on artificial intelligence according to claim 1, characterized in that: The method further comprises: A retraining instruction is received, and the business model is retrained based on the processing result.

4. A business model acquisition method based on artificial intelligence, characterized in that: The method is applied to a model training system, comprising: Receive training data and training parameters sent by the service end, wherein the training data follows a predetermined rule, and the training parameters include the number of training rounds, batch size, and learning rate; Receive the training instruction sent by the business end, perform model training according to the training corpus and training parameters, and obtain the business model; Obtaining a verification data set, restarting the business model, and processing the verification data set using the business model to obtain a verification result; Return the verification result to the service end.

5. The method for acquiring a business model based on artificial intelligence according to claim 4, characterized in that: The model training system adopts a python plug-in addition and deletion model.

6. A business model acquisition device based on artificial intelligence, characterized in that: The device is applied to the business end and includes: A training corpus uploading unit, used to upload training corpus to the model training system, wherein the training corpus follows a predetermined rule; A training parameter setting unit, configured to set training parameters, including the number of training rounds, batch size, and learning rate; A training instruction sending unit, configured to send a training instruction so that the model training system performs model training based on the training corpus and the training parameters; A receiving unit, configured to receive a training completion instruction sent by the model training system; the training completion instruction indicates that the model training system has completed the training of the business model; A verification data set uploading unit, configured to upload a verification data set so that the model training system restarts the business model and processes the verification data set using the business model; The result acquisition unit is used to obtain the processing result of the verification data set.

7. A business model acquisition device based on artificial intelligence, characterized in that: The device is applied to a model training system, including: A training corpus and training parameter receiving unit, configured to receive training corpus and training parameters sent by a service end, wherein the training corpus follows a predetermined rule, and the training parameters include the number of training rounds, batch size, and learning rate; A training instruction receiving unit is used to receive a training instruction sent by a service end, perform model training based on the training corpus and training parameters, and obtain a service model; A verification unit, configured to obtain a verification data set, restart the business model, and process the verification data set using the business model to obtain a verification result; The verification result returning unit is used to return the verification result to the service end.

8. A business model acquisition device based on artificial intelligence, characterized in that: include: CPU, memory, input and output interfaces, wired or wireless network interfaces, and power supply; The memory is a transient storage memory or a persistent storage memory; The central processing unit is configured to communicate with the memory and execute instructions in the memory on the device to perform the method according to any one of claims 1 to 3.

9. A business model acquisition device based on artificial intelligence, characterized in that: include: CPU, memory, input and output interfaces, wired or wireless network interfaces, and power supply; The memory is a transient storage memory or a persistent storage memory; The central processing unit is configured to communicate with the memory and execute instructions in the memory on the device to perform the method according to any one of claims 4 to 5.

10. A computer-readable storage medium, characterized in that The method comprises instructions, which, when executed on a computer, cause the computer to perform the method according to any one of claims 1 to 5.

Images