Analog resonance task execution method, device, equipment, medium and product

Abstract

This invention discloses a method, apparatus, device, medium, and product for executing a modular-to-digital (MDD) resonance task. The method includes: acquiring task request information for a target task; the target task is an MMD resonance task, and the task request information includes data requirements, model requirements, and performance requirements; determining target task splitting indicators based on the task request information; the task splitting indicators include computing power intensity, data sensitivity, and latency threshold; splitting the target task according to the target task splitting indicators, and determining the matching data resource blocks, computing power nodes, and sub-models for each target sub-task; executing each target sub-task according to the matching data resource blocks, computing power nodes, and sub-models to obtain the execution result of the target task. This solution solves the problem of low execution efficiency of MMD resonance tasks by collaboratively planning data, models, and computing power to split the MMD resonance task, which helps improve the execution efficiency of the MMD resonance task.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a method, apparatus, device, medium, and product for performing a modulus-digital resonance task. Background Technology

[0002] "Modal-to-digital resonance" refers to the collaborative optimization and closed-loop iteration between high-quality data and high-performance industrial models by constructing a two-way positive feedback loop of "using models to generate data" and "using data to empower models." It is one of the core mechanisms for promoting the deep integration of artificial intelligence and industry. "Using models to generate data" leverages the generalization and generation capabilities of advanced models to drive the cleaning, labeling, and enhancement of industrial data, solving the problem of "dirty, messy, and poor-quality" data. "Using data to empower models" feeds high-quality, highly adaptable industrial data back into model training, improving its professionalism, accuracy, and reliability in specific scenarios. Executing modular-to-digital resonance tasks requires powerful computing power. In existing technological systems, the collaboration between data, models, and computing power mainly focuses on the one-way adaptation of data and computing power, lacking deep coupling between data, computing power, and models, resulting in low execution efficiency for modular-to-digital resonance tasks. Summary of the Invention

[0003] This invention provides a method, apparatus, device, medium, and product for executing a modular-digital resonance task, in order to solve the problem of low execution efficiency of the modular-digital resonance task. By coordinating the planning of data, models, and computing power, the modular-digital resonance task is decomposed, which helps to improve the execution efficiency of the modular-digital resonance task.

[0004] According to one aspect of the present invention, a method for performing a modulus-digital resonance task is provided, the method comprising: Obtain the task request information of the target task; the target task is a modulus-digital resonance task, and the task request information includes data requirements, model requirements, and performance requirements; The target task splitting metrics are determined based on the task request information; the task splitting metrics include computing power intensity, data sensitivity, and latency threshold. The target task is split according to the target task splitting index to obtain each target sub-task, and the data resource block, computing power node and sub-model matched for each target sub-task are determined. Each target subtask is executed based on the data resource blocks, computing nodes, and sub-models matched to it, and the execution results of the target tasks are obtained.

[0005] According to another aspect of the present invention, an execution device for a modulus-digital resonance task is provided, the device comprising: The request information acquisition module is used to acquire the task request information of the target task; the target task is a modulus-digital resonance task, and the task request information includes data requirements, model requirements, and performance requirements. The task splitting metric determination module is used to determine the target task splitting metrics based on the task request information; the task splitting metrics include computing power intensity, data sensitivity, and latency threshold. The task splitting module is used to split the target task according to the target task splitting index to obtain each target sub-task, and to determine the data resource block, computing node and sub-model matched for each target sub-task; The execution result determination module is used to execute each target subtask based on the data resource blocks, computing nodes, and sub-models matched for each target subtask, and obtain the execution result of the target task.

[0006] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising: At least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores a computer program executable by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the method for executing the analog-digital resonance task according to any embodiment of the present invention.

[0007] According to another aspect of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium storing computer instructions for causing a processor to execute the method for performing the analog-digital resonance task according to any embodiment of the present invention.

[0008] According to another aspect of the present invention, a computer program product is provided, comprising a computer program that, when executed by a processor, implements the method for executing the analog-digital resonance task as described in any embodiment of the present invention.

[0009] The technical solution of this invention involves obtaining task request information for a target task; the target task is a modular-digital resonance task, and the task request information includes data requirements, model requirements, and performance requirements; determining target task splitting indicators based on the task request information; the task splitting indicators include computing power intensity, data sensitivity, and latency threshold; splitting the target task according to the target task splitting indicators to obtain each target sub-task, and determining the matching data resource blocks, computing power nodes, and sub-models for each target sub-task; executing each target sub-task according to the matching data resource blocks, computing power nodes, and sub-models to obtain the execution result of the target task. This technical solution solves the problem of low execution efficiency of modular-digital resonance tasks. By co-planning data, models, and computing power, and splitting the modular-digital resonance task, it is beneficial to improve the execution efficiency of the modular-digital resonance task.

[0010] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0011] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0012] Figure 1 This is a flowchart of an execution method for a modulus-digital resonance task according to Embodiment 1 of the present invention; Figure 2 This is a flowchart of an execution method for a modulus-digital resonance task according to Embodiment 2 of the present invention; Figure 3 This is a schematic diagram of the structure of an execution device for a modulus-digital resonance task according to Embodiment 3 of the present invention; Figure 4 This is a schematic diagram of the structure of an electronic device that implements the method for executing the modulus-digital resonance task according to an embodiment of the present invention. Detailed Implementation

[0013] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0014] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be used interchangeably where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or device that comprises 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 processes, methods, products, or devices. The acquisition, storage, use, and processing of data in the technical solutions of this application all comply with the relevant provisions of national laws and regulations.

[0015] Example 1 Figure 1 This is a flowchart illustrating an execution method for a modular-digital resonance task according to Embodiment 1 of the present invention. This embodiment is applicable to task scheduling scenarios, particularly for the collaborative planning of data, models, and computing power. The method can be executed by a modular-digital resonance task execution device, which can be implemented in hardware and / or software and can be configured in an electronic device. Figure 1 As shown, the method includes: S110. Obtain the task request information of the target task; the target task is a modulus-digital resonance task, and the task request information includes data requirements, model requirements, and performance requirements.

[0016] This solution can be executed by electronic devices such as computers and servers, specifically by a scheduling system for the analog-to-digital resonance task. The scheduling system may include a task service and one or more user terminals. Users can configure the analog-to-digital resonance task and generate task request information through their user terminals. The user terminal can also trigger industrial data acquisition operations when data acquisition conditions are met, such as the arrival of a preset period or the industrial system reaching its storage limit, and generate task request information based on the acquired industrial data. Furthermore, the user terminal can trigger model acquisition operations when model acquisition conditions are met, such as the arrival of a model's usage period or a decrease in the model's detection accuracy, and generate task request information based on the acquired model.

[0017] The task request information may specifically include the requester's identifier, request time, data requirements, model requirements, and performance requirements of the target task. The data requirements constrain the data resources of the target task, the model requirements constrain the model of the target task, and the performance requirements constrain the quality of the model and / or data output by the target task. Specifically, the data requirements may include constraints such as the data type, data volume, whether the data is public, and the data privacy protection methods required by the target task. The model requirements may include constraints such as the functionality, structure, number of parameters, and storage space required by the model. The performance requirements may include constraints such as the time taken per model run, the computational and storage resources used per model run, data processing time, and data quality.

[0018] After generating the task request information for the target task, the user terminal can send the task request information to the task service to initiate the execution request of the target task. The task service may include components such as a task decomposer, a resource allocator, a load prediction engine, and a load balancer. Since modular resonant tasks typically involve large-scale computational tasks, the task decomposer can break down the target task, fully utilizing different levels of computing resources while improving the execution efficiency of the target task. The resource allocator can allocate computing resources, storage resources, and network resources to each subtask decomposed by the task decomposer. The load prediction engine can predict the load for future moments based on the historical load information of computing nodes, enabling the load balancer to schedule tasks and achieve load balancing.

[0019] S120. Determine the target task splitting indicators based on the task request information; the task splitting indicators include computing power intensity, data sensitivity, and latency threshold.

[0020] The task service can split the target task based on one or more task splitting metrics according to the task request information, and determine the target task splitting metrics from these metrics. The scheduling system can pre-acquire industrial data from different scenarios, aggregate and process the industrial data to build a data resource pool, or integrate different types of computing power resources to build a computing power resource pool. It can also include general models and professional models from different fields, such as healthcare and finance, to build a model resource pool. Specifically, the task service can filter target data resources matching the target task from the data resource pool according to data requirements, and filter target models matching the target task from the model resource pool according to model requirements. The task service can determine the target task splitting metrics based on the target data resources, target models, and performance requirements of the target task. Task splitting metrics may include metrics such as computing power intensity, data sensitivity, and latency thresholds, used to constrain the task splitting scheme.

[0021] As is understandable, computational power (CP) is the number of floating-point operations a computer system can complete per unit of time, measured in floating-point operations per second (FLOPS). Computational power can generally be categorized into general-purpose computing power, intelligent computing power, and supercomputing power based on its intensity. General-purpose computing power uses CPU chips for computation, intelligent computing power uses GPUs, NPUs, FPGAs, and ASICs, while supercomputing power uses heterogeneous many-core CPU+GPU architectures. The intensity of computational power increases sequentially from general-purpose to intelligent to supercomputing power. Computational power intensity can constrain the computational power intensity of each subtask in a task decomposition scheme, while data sensitivity can constrain the privacy requirements of data resource blocks in each subtask within the task decomposition scheme. Latency thresholds can constrain the overall latency of each subtask in the task decomposition scheme.

[0022] S130. The target task is split according to the target task splitting index to obtain each target sub-task, and the data resource block, computing power node and sub-model matched for each target sub-task are determined.

[0023] The task service can use a task decomposer to break down the target task according to the target task decomposition index, resulting in various target sub-tasks. For example, the target task can be broken down into target sub-tasks such as data preprocessing, model training, and result transformation. The task service can use a resource allocator to determine the data resource blocks, computing nodes, and sub-models corresponding to each target sub-task.

[0024] S140. Execute each target subtask according to the data resource blocks, computing nodes and sub-models matched for each target subtask to obtain the execution result of the target task.

[0025] The task service can obtain the data resource blocks and sub-models corresponding to each target sub-task, send the data resource blocks and sub-models corresponding to each target sub-task to the computing power nodes corresponding to each target sub-task for calculation, obtain the execution results of each target sub-task, and then output the execution results of the target task based on the execution results of each target sub-task.

[0026] The technical solution of this invention involves obtaining task request information for a target task; the target task is a modular-digital resonance task, and the task request information includes data requirements, model requirements, and performance requirements; determining target task splitting indicators based on the task request information; the task splitting indicators include computing power intensity, data sensitivity, and latency threshold; splitting the target task according to the target task splitting indicators to obtain each target sub-task, and determining the matching data resource blocks, computing power nodes, and sub-models for each target sub-task; executing each target sub-task according to the matching data resource blocks, computing power nodes, and sub-models to obtain the execution result of the target task. This technical solution solves the problem of low execution efficiency of modular-digital resonance tasks. By co-planning data, models, and computing power, and splitting the modular-digital resonance task, it is beneficial to improve the execution efficiency of the modular-digital resonance task.

[0027] Example 2 Figure 2 This is a flowchart of an execution method for a modulus-digital resonance task provided in Embodiment 2 of the present invention. This embodiment is a refinement based on the above embodiment. Figure 2 As shown, the method includes: S210. Obtain the task request information of the target task; the target task is a modular resonant task, and the task request information includes data requirements, model requirements, and performance requirements.

[0028] S220. Determine the target data resource from the data resource pool according to the data requirements.

[0029] In this solution, the task service can filter target data resources matching the target task from the data resource pool based on data requirements. For example, if the target task requires images, with 20,000 images and publicly available data, the task service can filter data resources that meet the above requirements from the data resource pool as target data resources.

[0030] S230. Determine the target model from the model resource pool according to the model requirements.

[0031] The task service can select target models from the model resource pool that match the target task based on model requirements. For example, if the target task requires a model with object detection functionality, a YOLO architecture-based model structure, and fewer parameters than a preset parameter threshold, the task service can select models from the model resource pool that meet these requirements as the target model.

[0032] S240. Based on the target data resources, the target model, and the performance requirements, determine the target task splitting indicators; the task splitting indicators include computing power intensity, data sensitivity, and latency threshold.

[0033] The data resource pool may contain one or more target data resources, and the model resource pool may contain one or more target models. The task service can generate one or more sets of task breakdown metrics that meet the performance requirements of the target task based on the target data resources and target models. One set of task breakdown metrics is then selected as the target task breakdown metrics.

[0034] In one feasible solution, determining the target splitting metrics based on the target data resources, the target model, and the performance requirements includes: If there are multiple target data resources and / or multiple target models, then at least two data model combinations are generated based on each target data resource and each target model. Based on the combination of data models and the performance requirements, at least one set of candidate task splitting metrics shall be determined; The evaluation results of each candidate task splitting indicator are determined based on the computing power intensity, data sensitivity, and latency threshold of each candidate task splitting indicator. Based on the evaluation results of each candidate task breakdown index, the target task breakdown index is determined from among the candidate task breakdown indices.

[0035] When there are multiple target data resources, multiple target models, or both multiple target data resources and multiple target models, the task service can arrange and combine the target data resources and target models to obtain multiple different data model combinations. Based on each data model combination and the performance requirements of the target task, candidate task splitting metrics matching each data model combination are obtained. It should be noted that the candidate task splitting metrics corresponding to different data model combinations may be different or the same.

[0036] The task service can pre-set evaluation functions based on task splitting metrics, and calculate the evaluation results for each candidate task splitting metric according to the evaluation functions. For example, the evaluation function can be expressed as: ; This represents the computational power intensity assessment value in the candidate task splitting metrics. This represents the data sensitivity assessment value in the candidate task splitting metrics. This represents the latency threshold evaluation value in the candidate task splitting metrics. , and These represent the weights of the computing power intensity assessment value, the data sensitivity assessment value, and the latency threshold assessment value, respectively. , and All are greater than 0. .

[0037] Understandably, computing nodes with higher computing power are generally rarer and more difficult to match; data resources with higher data sensitivity are more difficult to acquire and usually require more complex data processing, such as encryption and desensitization; and a higher latency threshold indicates poorer real-time performance of the task. Therefore, the evaluation results of the candidate task splitting index can be negatively correlated with the computing power assessment value, data sensitivity assessment value, and latency threshold assessment value.

[0038] The task service can select the candidate task splitting indicator with the largest evaluation result from among the candidate task splitting indicators as the target task splitting indicator based on the evaluation results of each candidate task splitting indicator.

[0039] S250. The target task is split according to the target task splitting index to obtain each target sub-task, and the data resource block, computing power node and sub-model matched for each target sub-task are determined.

[0040] S260. Execute each target subtask according to the data resource blocks, computing nodes and sub-models matched for each target subtask to obtain the execution result of the target task.

[0041] In this scheme, the execution of each target sub-task based on the data resource blocks, computing nodes, and sub-models matched for each target sub-task includes: Obtain access permissions for the target data resources, generate data access policies based on the access permissions for the target data resources, and obtain data resource blocks that match each target subtask. Obtain access permissions to the target model, generate a model access policy based on the access permissions to the target model, and extract sub-models matching each target sub-task from the target model; The data resource blocks and sub-models matched for each target sub-task are sent to the computing power nodes matched for each target sub-task to execute each target sub-task.

[0042] To ensure the security of data resources in the data resource pool, the task service can utilize a privacy-preserving computing engine based on technologies such as federated learning, differential privacy, and security sandboxes to acquire target data resources and target models for task execution without disclosing original information. The task service can also transform target data resources according to preset rules using a compliance adapter to meet cross-domain usage requirements.

[0043] After identifying the target data resource, the task service can acquire usage permissions for that resource, generate a data access policy based on those permissions, and then divide the target data resource into data resource blocks matching each target subtask. Similarly, after identifying the target model, the task service can acquire usage permissions for that model, generate a model access policy based on those permissions, and extract the sub-models matching each target subtask from the target model. The task service can then send the data resource blocks and sub-models matching each target subtask to the corresponding computing nodes. These computing nodes can then execute each target subtask based on these data resource blocks and sub-models.

[0044] Optionally, the modulus-digital resonance task includes a modulus-based number generation task and a number-based modulus assignment task; After obtaining the execution result of the target task, the method further includes: If the target task is a data-driven modeling task, then obtain the performance evaluation results of the target model in the target scenario, and determine the iterative update requirements of the target model based on the performance evaluation results; If the target model needs to be iteratively updated and there is new data in the target data resources, then the target task splitting index is updated according to the new data, so as to re-split the target task and update the execution result of the target task.

[0045] For the task of modeling using data, the execution result of the target task is the target model. After obtaining the execution result of the target task, the task service can perform application testing on the target model in the target scenario, obtain performance evaluation results, and determine whether the target model meets the application requirements of the target scenario based on the performance evaluation results. For example, the target task is to train a general object detection model into a defect detection model for detecting defects in industrial parts based on image samples of industrial parts. After the target task outputs the target model, the task service can use the target model to perform defect detection on industrial parts and obtain performance evaluation results such as detection accuracy and error rate.

[0046] If the performance evaluation results indicate that the target model cannot meet the application requirements of the target scenario—for example, if the detection accuracy does not reach the preset accuracy threshold—the task service can continue to iterate and update the target model until it meets the application requirements of the target scenario. When the target model needs iterative updates, the task service can detect whether there is new data in the target data resources. If so, it updates the target task splitting metrics based on the new data to re-split the target task and update the execution results of the target task.

[0047] This solution can verify the target model obtained by the data modeling task in various scenarios, ensuring the usability of the target task execution results.

[0048] In this embodiment, the modulus-digital resonance task includes a modulus-based number generation task and a number-based modulus assignment task; After obtaining the execution result of the target task, the method further includes: If the target task is a modular data acquisition task, then the processing evaluation result of the target data resource is obtained, and the data processing requirements of the target data resource are determined based on the quality evaluation result. Based on the data processing requirements of the target data resources, update the target task splitting metrics to re-split the target tasks and update the execution results of the target tasks.

[0049] For model-driven data processing tasks, the execution result of the target task is the processing result of the target data resource, such as text expansion or image enhancement. After obtaining the execution result of the target task, the task service can perform a quality assessment of the processing of the target data resource, obtaining a processing assessment result. Based on this result, it can determine whether the processing result of the target data resource meets the data processing requirements, such as whether there are logical contradictions in the text expansion result or whether the enhanced image can identify the target. If it meets the requirements, the task service can output the processing result of the target data resource. If it does not meet the requirements, it is determined that the target data resource has data processing needs. The task service can update the target task splitting metrics based on these needs to re-split the target task and update the processing result of the target data resource. Specifically, the task service can perform operations such as model switching and adjusting the quality of the target data resource based on its data processing needs, and update the target task splitting metrics based on the adjusted target model and target data resource.

[0050] This solution can perform quality checks on the processing results of target data resources obtained from the model-based indexing task, ensuring the availability of the target task execution results.

[0051] In a preferred embodiment, the task service enables end-to-end auditing, traceability, and risk control of the "data-computing power-model" process. Core components include a multi-party governance console, a compliance audit chain, a security early warning engine, and an emergency response mechanism. The multi-party governance console integrates information such as data authorization logs, computing power scheduling records, and model iteration logs, providing a visual monitoring interface. The compliance audit chain, based on blockchain technology, stores the entire process log of the target task on the blockchain for evidence storage, ensuring immutability. The security early warning engine, based on abnormal behavior identification algorithms, provides real-time alerts for abnormal behaviors such as unauthorized computing power access, unauthorized data transfer, and model parameter tampering. The emergency response mechanism can trigger blocking mechanisms upon detecting leakage risks, such as suspending data transmission, terminating computing power calls, and freezing model access.

[0052] This technical solution addresses the problem of low execution efficiency in modular-digital resonance (MDR) tasks. By collaboratively planning data, models, and computing power, and splitting the MMR task, the execution efficiency of the MMR task can be improved.

[0053] Example 3 Figure 3 This is a schematic diagram of the structure of an execution device for a modulus-digital resonance task provided in Embodiment 3 of the present invention. Figure 3 As shown, the device includes: The request information acquisition module 310 is used to acquire the task request information of the target task; the target task is a modulus-digital resonance task, and the task request information includes data requirements, model requirements, and performance requirements. The task splitting index determination module 320 is used to determine the target task splitting index based on the task request information; the task splitting index includes computing power intensity, data sensitivity, and latency threshold; The task splitting module 330 is used to split the target task according to the target task splitting index to obtain each target sub-task, and to determine the data resource block, computing node and sub-model matched for each target sub-task; The execution result determination module 340 is used to execute each target sub-task according to the data resource blocks, computing nodes and sub-models matched for each target sub-task, and obtain the execution result of the target task.

[0054] In this solution, the splitting index determination module 320 includes: A data resource determination unit is used to determine target data resources from the data resource pool based on the data requirements; The model determination unit is used to determine the target model from the model resource pool according to the model requirements; The splitting index determination unit is used to determine the target task splitting index based on the target data resources, the target model, and the performance requirements.

[0055] Based on the above scheme, the splitting index determination unit is specifically used for: If there are multiple target data resources and / or multiple target models, then at least two data model combinations are generated based on each target data resource and each target model. Based on the combination of data models and the performance requirements, at least one set of candidate task splitting metrics shall be determined; The evaluation results of each candidate task splitting indicator are determined based on the computing power intensity, data sensitivity, and latency threshold of each candidate task splitting indicator. Based on the evaluation results of each candidate task breakdown index, the target task breakdown index is determined from among the candidate task breakdown indices.

[0056] Optionally, the execution result determination module 340 is specifically used for: Obtain access permissions for the target data resources, generate data access policies based on the access permissions for the target data resources, and obtain data resource blocks that match each target subtask. Obtain access permissions to the target model, generate a model access policy based on the access permissions to the target model, and extract sub-models matching each target sub-task from the target model; The data resource blocks and sub-models matched for each target sub-task are sent to the computing power nodes matched for each target sub-task to execute each target sub-task.

[0057] In one feasible scheme, the modulus-digital resonance task includes a modulus-based number generation task and a number-based modulus assignment task; The device further includes: The model update requirement determination module is used to, after obtaining the execution result of the target task, if the target task is a data-based modeling task, obtain the performance evaluation result of the target model in the target scenario, and determine the iterative update requirement of the target model based on the performance evaluation result. The first execution result update module is used to update the target task splitting index according to the new data if the target model needs to be iteratively updated and there is new data in the target data resources, so as to re-split the target task and update the execution result of the target task.

[0058] In another feasible approach, the modulus-digit resonance task includes a modulus-derived number task and a number-derived modulus task; The device further includes: The data processing requirement determination module is used to, after obtaining the execution result of the target task, if the target task is a modular data task, obtain the processing evaluation result of the target data resource, and determine the data processing requirement of the target data resource based on the quality evaluation result. The second execution result update module is used to update the target task splitting indicators according to the data processing requirements of the target data resources, so as to re-split the target task and update the execution result of the target task.

[0059] The execution device for the analog-to-digital resonance task provided in the embodiments of the present invention can execute the execution method for the analog-to-digital resonance task provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0060] Example 4 Figure 4A schematic diagram of an electronic device 410 that can be used to implement embodiments of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0061] like Figure 4 As shown, the electronic device 410 includes at least one processor 411 and a memory communicatively connected to the at least one processor 411. The memory may be a read-only memory (ROM) 412, a random access memory (RAM) 413, etc. The memory stores computer programs executable by the at least one processor. The processor 411 can perform various appropriate actions and processes based on the computer program stored in the ROM 412 or loaded from storage unit 418 into the RAM 413. The RAM 413 may also store various programs and data required for the operation of the electronic device 410. The processor 411, ROM 412, and RAM 413 are interconnected via a bus 414. An input / output (I / O) interface 415 is also connected to the bus 414.

[0062] Multiple components in electronic device 410 are connected to I / O interface 415, including: input unit 416, such as keyboard, mouse, etc.; output unit 417, such as various types of displays, speakers, etc.; storage unit 418, such as disk, optical disk, etc.; and communication unit 419, such as network card, modem, wireless transceiver, etc. Communication unit 419 allows electronic device 410 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0063] Processor 411 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 411 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. Processor 411 performs the various methods and processes described above, such as the execution methods of the analog-digital resonance task.

[0064] In some embodiments, the method for executing the analog-to-digital resonance task may be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 418. In some embodiments, part or all of the computer program may be loaded and / or mounted on electronic device 410 via ROM 412 and / or communication unit 419. When the computer program is loaded into RAM 413 and executed by processor 411, one or more steps of the method for executing the analog-to-digital resonance task described above may be performed. Alternatively, in other embodiments, processor 411 may be configured to execute the method for executing the analog-to-digital resonance task by any other suitable means (e.g., by means of firmware).

[0065] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0066] Computer programs used to implement the methods of the present invention can be written in any combination of one or more programming languages. These computer programs can be provided to the processor of a general-purpose computer, a special-purpose computer, or other programmable analog-to-digital resonant task execution device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The computer programs can be executed entirely on the machine, partially on the machine, as a standalone software package partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0067] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0068] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0069] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or middleware components (e.g., application servers), or frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0070] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability.

[0071] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0072] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method for executing a modulus-digital resonance task, characterized in that, The method includes: Obtain the task request information of the target task; the target task is a modulus-digital resonance task, and the task request information includes data requirements, model requirements, and performance requirements; The target task splitting metrics are determined based on the task request information; the task splitting metrics include computing power intensity, data sensitivity, and latency threshold. The target task is split according to the target task splitting index to obtain each target sub-task, and the data resource block, computing power node and sub-model matched for each target sub-task are determined. Each target subtask is executed based on the data resource blocks, computing nodes, and sub-models matched to it, and the execution results of the target tasks are obtained.

2. The method according to claim 1, characterized in that, The step of determining the target task breakdown index based on the task request information includes: Target data resources are determined from the data resource pool based on the data requirements. The target model is determined from the model resource pool based on the model requirements. Based on the target data resources, the target model, and the performance requirements, determine the target task breakdown metrics.

3. The method according to claim 2, characterized in that, The step of determining the target splitting metrics based on the target data resources, the target model, and the performance requirements includes: If there are multiple target data resources and / or multiple target models, then at least two data model combinations are generated based on each target data resource and each target model. Based on the combination of data models and the performance requirements, at least one set of candidate task splitting metrics shall be determined; The evaluation results of each candidate task splitting indicator are determined based on the computing power intensity, data sensitivity, and latency threshold of each candidate task splitting indicator. Based on the evaluation results of each candidate task breakdown index, the target task breakdown index is determined from among the candidate task breakdown indices.

4. The method according to claim 2, characterized in that, The execution of each target sub-task based on the data resource blocks, computing nodes, and sub-models matched for each target sub-task includes: Obtain access permissions for the target data resources, generate data access policies based on the access permissions for the target data resources, and obtain data resource blocks that match each target subtask. Obtain access permissions to the target model, generate a model access policy based on the access permissions to the target model, and extract sub-models matching each target sub-task from the target model; The data resource blocks and sub-models matched for each target sub-task are sent to the computing power nodes matched for each target sub-task to execute each target sub-task.

5. The method according to claim 2, characterized in that, The modular-digital resonance task includes the task of generating numbers from modulus and the task of assigning modulus to numbers; After obtaining the execution result of the target task, the method further includes: If the target task is a data-driven modeling task, then obtain the performance evaluation results of the target model in the target scenario, and determine the iterative update requirements of the target model based on the performance evaluation results; If the target model needs to be iteratively updated and there is new data in the target data resources, then the target task splitting index is updated according to the new data, so as to re-split the target task and update the execution result of the target task.

6. The method according to claim 2, characterized in that, The modular-digital resonance task includes the task of generating numbers from modulus and the task of assigning modulus to numbers; After obtaining the execution result of the target task, the method further includes: If the target task is a modular data acquisition task, then the processing evaluation result of the target data resource is obtained, and the data processing requirements of the target data resource are determined based on the quality evaluation result. Based on the data processing requirements of the target data resources, update the target task splitting metrics to re-split the target tasks and update the execution results of the target tasks.

7. An execution device for a modulus-digital resonance task, characterized in that, The device includes: The request information acquisition module is used to acquire the task request information of the target task; the target task is a modulus-digital resonance task, and the task request information includes data requirements, model requirements, and performance requirements. The task splitting metric determination module is used to determine the target task splitting metrics based on the task request information; the task splitting metrics include computing power intensity, data sensitivity, and latency threshold. The task splitting module is used to split the target task according to the target task splitting index to obtain each target sub-task, and to determine the data resource block, computing node and sub-model matched for each target sub-task; The execution result determination module is used to execute each target subtask based on the data resource blocks, computing nodes, and sub-models matched for each target subtask, and obtain the execution result of the target task.

8. An electronic device, characterized in that, The electronic device includes: At least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores a computer program executable by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the method for executing the analog-digital resonance task according to any one of claims 1-6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that are used to cause a processor to execute the method for performing the analog-digital resonance task according to any one of claims 1-6.

10. A computer program product comprising a computer program that, when executed by a processor, implements the method for executing the analog-digital resonance task according to any one of claims 1-6.

Images