Model registration method and apparatus, and storage medium

Abstract

The present disclosure relates to the technical field of communications. Provided are a model registration method and apparatus, and a storage medium. The method comprises: a first entity sending a first message to a second entity, wherein the first message is used for requesting a registration model, and comprises a first model identifier and / or model information, which first model identifier is an identifier of a model in the first entity; receiving a second message which is sent by the second entity in response to the first message, wherein the second message comprises a second model identifier and first indication information, which second model identifier is a network identifier allocated by the second entity for the model, and which first indication information indicates a registration decision or information related to the registration decision; and on the basis of the information related to the registration decision, sending a third message to the second entity, wherein the third message comprises the second model identifier and the registration decision. By means of additionally configuring in a network a model registration function and a corresponding registration process, the network can provide services, such as model registration, for a user, a network or a third party as needed, thereby laying a basis for intelligent endogenesis.

Description

Model registration method, device and storage medium

[0001] This disclosure claims priority to the Chinese patent application filed with the China Patent Office on December 29, 2023, with application number 202311865537.X and application name “A model registration method, device and storage medium”, the entire contents of which are incorporated by reference into this disclosure. Technical Field

[0002] The present disclosure relates to the field of communication technology, and in particular to a model registration method, device, and storage medium. Background Art

[0003] The inherent intelligence of the 6th Generation Mobile Networks (6G) refers to the fact that 6G networks natively support artificial intelligence (AI), make AI capabilities a basic network service, and implement AI as a Service (AIaaS). This enables the network to self-learn and self-evolve, and can then provide AI training services, AI inference services, AI deployment and open services to users, networks or third parties.

[0004] Although the network architecture and network processes of the current mobile communication system can provide users with connection and data transmission services, they cannot provide intelligent endogenous services such as AI model registration and storage. Summary of the Invention

[0005] The present disclosure provides a model registration method, device and storage medium, which solve the technical problem that the communication system cannot provide intelligent endogenous services such as AI model registration and storage.

[0006] In a first aspect, an embodiment of the present disclosure provides a model registration method, applied to a first entity, the method comprising:

[0007] Sending a first message to a second entity, where the first message is used to request registration of a model, and the first message includes a first model identifier and / or model information, where the first model identifier is an identifier of the model in the first entity;

[0008] receiving a second message sent by the second entity in response to the first message, where the second message includes a second model identifier and first indication information, where the second model identifier is a network identifier allocated by the second entity to the model, and the first indication information indicates a registration decision or registration decision related information;

[0009] Based on the registration decision related information, a third message is sent to the second entity, where the third message includes the second model identifier and the registration decision.

[0010] In one embodiment, the model is a public model.

[0011] In one embodiment, the model information includes at least one of the following: model type, model modality, model algorithm, model accuracy, model security, model privacy, model description information, model deployment requirements, model reasoning performance, model open scope, model version, model owner information, or automated deployment script.

[0012] In one embodiment, the model description information includes at least one of the following: model accuracy, model accuracy variance, data input format, data output format, model size, full model identifier corresponding to the compressed model, model parameter quantity, model recall rate, or model F1 value;

[0013] The model deployment requirements include at least one of the following: model dependency, minimum resource requirements, or recommended resource configuration, where the resources include computing resources and storage resources.

[0014] In one embodiment, the model information further includes at least one of the following: model application scenario, model segmentation information, model compression information or model training information.

[0015] In one embodiment, the model training information includes at least one of the following: a training data set, metadata corresponding to the training data set, a data preprocessing method, model features, model hyperparameters, or a model structure.

[0016] In one embodiment, the method further comprises:

[0017] Second indication information sent by the second entity is received, where the second indication information includes the first model identifier, the second model identifier, and a model information template.

[0018] In one embodiment, the method further comprises:

[0019] sending a fourth message to the second entity;

[0020] receiving a fifth message sent by the second entity in response to the fourth message;

[0021] or,

[0022] sending a fourth message to the third entity;

[0023] receiving a fifth message sent by the third entity in response to the fourth message;

[0024] Among them, the fourth message is used to request storage of the model, and the fourth message includes at least one of the following: the second model identifier, the model transmission method, the model security certificate, the model address or the model; the fifth message is used to indicate the model storage result.

[0025] In one embodiment, the method further comprises:

[0026] The model is sent to the second entity or the third entity.

[0027] In one embodiment, the method further comprises:

[0028] Sending a sixth message to the second entity, where the sixth message includes the second model identifier and model update information;

[0029] A seventh message is received, which is sent by the second entity in response to the sixth message, where the seventh message is used to indicate a model update result.

[0030] In one embodiment, the method further comprises:

[0031] sending an eighth message to the second entity;

[0032] receiving a ninth message sent by the second entity in response to the eighth message;

[0033] or,

[0034] sending an eighth message to the third entity;

[0035] receiving a ninth message sent by the third entity in response to the eighth message;

[0036] or,

[0037] sending an eighth message to the fourth entity;

[0038] receiving a ninth message sent by the fourth entity in response to the eighth message;

[0039] The eighth message is used to request verification of the model, and the eighth message includes the second model identifier and the model; and the ninth message is used to indicate a model verification result.

[0040] In one embodiment, the first entity is a first terminal, a first access network device, or a first core network device;

[0041] The second entity is deployed in a second terminal, a second access network device or a second core network device;

[0042] The third entity is deployed in a third terminal, a third access network device or a third core network device;

[0043] The fourth entity is deployed in a fourth access network device or a fourth core network device;

[0044] The first terminal, the second terminal and the third terminal are the same or different from each other, the first access network device, the second access network device, the third access network device and the fourth access network device are the same or different from each other, and the first core network device, the second core network device, the third core network device and the fourth core network device are the same or different from each other.

[0045] In a second aspect, an embodiment of the present disclosure provides a model registration method, applied to a second entity, the method comprising:

[0046] receiving a first message sent by a first entity, where the first message is used to request registration of a model, and the first message includes a first model identifier and / or model information, where the first model identifier is an identifier of the model in the first entity;

[0047] sending a second message to the first entity in response to the first message, where the second message includes a second model identifier and first indication information, where the second model identifier is a network identifier allocated to the model by the second entity, and the first indication information indicates a registration decision or registration decision related information;

[0048] A third message sent by the first entity is received, where the third message is a message determined based on the registration decision related information, and the third message includes the second model identifier and the registration decision.

[0049] In one embodiment, the model is a public model.

[0050] In one embodiment, the model information includes at least one of the following: model type, model modality, model algorithm, model accuracy, model security, model privacy, model description information, model deployment requirements, model reasoning performance, model open scope, model version, model owner information, or automated deployment script.

[0051] In one embodiment, the model description information includes at least one of the following: model accuracy, model accuracy variance, data input format, data output format, model size, full model identifier corresponding to the compressed model, model parameter quantity, model recall rate, or model F1 value;

[0052] The model deployment requirements include at least one of the following: model dependency, minimum resource requirements, or recommended resource configuration, where the resources include computing resources and storage resources.

[0053] In one embodiment, the first model information further includes at least one of the following: model application scenario, model segmentation information, model compression information or model training information.

[0054] In one embodiment, the model training information includes at least one of the following: a training data set, metadata corresponding to the training data set, a data preprocessing method, model features, model hyperparameters, or a model structure.

[0055] In one embodiment, the method further comprises:

[0056] Second indication information is sent to the first entity, where the second indication information includes the first model identifier, the second model identifier, and a model information template.

[0057] In one embodiment, the method further comprises:

[0058] receiving a fourth message sent by the first entity, where the fourth message is used to request storage of the model, and the fourth message includes at least one of the following: the second model identifier, the model transmission mode, the model security certificate, the model address, or the model;

[0059] In response to the fourth message, a fifth message is sent to the first entity, where the fifth message is used to instruct the model to store the result.

[0060] In one embodiment, the method further comprises:

[0061] The model is received.

[0062] In one embodiment, the method further comprises:

[0063] sending a tenth message to the third entity;

[0064] receiving an eleventh message sent by the third entity in response to the tenth message;

[0065] or,

[0066] sending a tenth message to the fourth entity;

[0067] receiving an eleventh message sent by the fourth entity in response to the tenth message;

[0068] The tenth message is used to request storage of the model, and the tenth message includes at least one of the following: the second model identifier, the model transmission method, the model security certificate or the model; the eleventh message is used to indicate the model storage result.

[0069] In one embodiment, the method further comprises:

[0070] The model is sent to the third entity or the fourth entity.

[0071] In one embodiment, the method further comprises:

[0072] receiving a sixth message sent by the first entity, where the sixth message includes the second model identifier and model update information;

[0073] In response to the sixth message, a seventh message is sent to the first entity, where the seventh message is used to indicate a model update result.

[0074] In one embodiment, the method further comprises:

[0075] receiving an eighth message sent by the first entity, where the eighth message is used to request verification of the model, and the eighth message includes the second model identifier and the model;

[0076] A ninth message is sent to the first entity in response to the eighth message, where the ninth message is used to indicate a model validation result.

[0077] In a third aspect, an embodiment of the present disclosure provides a model registration method, applied to a third entity, the method comprising:

[0078] receiving a model storage request message, where the model storage request message includes at least one of the following: a second model identifier, a model transmission mode, a model security certificate, a model address, or a model, where the second model identifier is a network identifier assigned to the model by the second entity, and the model storage request message is the fourth message sent by the first entity, the tenth message sent by the second entity, or the twelfth message sent by the fourth entity;

[0079] Send a model storage response message, where the model storage response message is used to indicate a model storage result, and the storage response message is the fifth message sent to the first entity, the eleventh message sent to the second entity, or the thirteenth message sent to the fourth entity.

[0080] In one embodiment, the method further comprises:

[0081] The model is received.

[0082] In one embodiment, the method further comprises:

[0083] receiving an eighth message sent by the first entity, where the eighth message is used to request verification of the model, and the eighth message includes the second model identifier and the model;

[0084] A ninth message is sent to the first entity in response to the eighth message, where the ninth message is used to indicate a model validation result.

[0085] In a fourth aspect, an embodiment of the present disclosure provides a model registration method, applied to a fourth entity, the method comprising:

[0086] receiving an eighth message sent by the first entity, where the eighth message is used to request verification of the model, the eighth message including a second model identifier and the model, where the second model identifier is a network identifier allocated to the model by the second entity;

[0087] A ninth message is sent to the first entity in response to the eighth message, where the ninth message is used to indicate the model validation result.

[0088] In one embodiment, the method further comprises:

[0089] receiving a tenth message sent by the second entity, where the tenth message is used to request storage of the model, and the tenth message includes the second model identifier;

[0090] An eleventh message is sent to the second entity in response to the tenth message, where the eleventh message is used to indicate a model storage result.

[0091] In one embodiment, the method further comprises:

[0092] Sending a twelfth message to the third entity, where the twelfth message is used to request storage of the model, and the twelfth message includes at least one of the following: the second model identifier, the model transmission mode, the model security certificate, or the model;

[0093] A thirteenth message sent by the third entity is received, where the thirteenth message is used to indicate a model storage result.

[0094] In a fifth aspect, an embodiment of the present disclosure provides a model registration device, applied to a first entity, the device comprising:

[0095] A first sending unit, configured to send a first message to a second entity, where the first message is used to request model registration, and the first message includes a first model identifier and / or model information, where the first model identifier is an identifier of the model in the first entity;

[0096] a first receiving unit, configured to receive a second message sent by the second entity in response to the first message, the second message including a second model identifier and first indication information, the second model identifier being a network identifier allocated by the second entity to the model, and the first indication information indicating a registration decision or registration decision related information;

[0097] The second sending unit is configured to send a third message to the second entity based on the registration decision related information, where the third message includes the second model identifier and the registration decision.

[0098] In a sixth aspect, an embodiment of the present disclosure provides a model registration device, applied to a second entity, the device comprising:

[0099] a first receiving unit, configured to receive a first message sent by a first entity, where the first message is used to request registration of a model, and the first message includes a first model identifier and / or model information, where the first model identifier is an identifier of the model in the first entity;

[0100] a first sending unit, configured to send a second message to the first entity in response to the first message, where the second message includes a second model identifier and first indication information, where the second model identifier is a network identifier allocated by the second entity to the model, and the first indication information indicates a registration decision or registration decision related information;

[0101] The second receiving unit is configured to receive a third message sent by the first entity, where the third message is a message determined based on the registration decision related information, and the third message includes the second model identifier and the registration decision.

[0102] In a seventh aspect, an embodiment of the present disclosure provides a model registration device, applied to a third entity, the device comprising:

[0103] a first receiving unit, configured to receive a model storage request message, the model storage request message including at least one of the following: a second model identifier, a model transmission mode, a model security certificate, a model address, or a model, the second model identifier being a network identifier allocated by the second entity to the model, and the model storage request message being the fourth message sent by the first entity, the tenth message sent by the second entity, or the twelfth message sent by the fourth entity;

[0104] A first sending unit is used to send a model storage response message, where the model storage response message is used to indicate a model storage result, and the storage response message is the fifth message sent to the first entity, the eleventh message sent to the second entity, or the thirteenth message sent to the fourth entity.

[0105] In an eighth aspect, an embodiment of the present disclosure provides a model registration device, applied to a fourth entity, the device comprising:

[0106] a first receiving unit, configured to receive an eighth message sent by the first entity, where the eighth message is used to request verification of the model, the eighth message including a second model identifier and the model, where the second model identifier is a network identifier allocated to the model by the second entity;

[0107] The first sending unit is configured to send a ninth message to the first entity in response to the eighth message, where the ninth message is used to indicate a model verification result.

[0108] In a ninth aspect, an embodiment of the present disclosure provides a model registration device, applied to a first entity, the device including a memory, a transceiver, and a processor:

[0109] The memory is used to store a computer program; the transceiver is used to send and receive data under the control of the processor; and the processor is used to read the computer program in the memory and perform the following operations:

[0110] Sending a first message to a second entity, where the first message is used to request registration of a model, and the first message includes a first model identifier and / or model information, where the first model identifier is an identifier of the model in the first entity;

[0111] receiving a second message sent by the second entity in response to the first message, where the second message includes a second model identifier and first indication information, where the second model identifier is a network identifier allocated by the second entity to the model, and the first indication information indicates a registration decision or registration decision related information;

[0112] Based on the registration decision related information, a third message is sent to the second entity, where the third message includes the second model identifier and the registration decision.

[0113] In one embodiment, the model is a public model.

[0114] In one embodiment, the model information includes at least one of the following: model type, model modality, model algorithm, model accuracy, model security, model privacy, model description information, model deployment requirements, model reasoning performance, model open scope, model version, model owner information, or automated deployment script.

[0115] In one embodiment, the model description information includes at least one of the following: model accuracy, model accuracy variance, data input format, data output format, model size, full model identifier corresponding to the compressed model, model parameter quantity, model recall rate, or model F1 value;

[0116] The model deployment requirements include at least one of the following: model dependency, minimum resource requirements, or recommended resource configuration, where the resources include computing resources and storage resources.

[0117] In one embodiment, the model information further includes at least one of the following: model application scenario, model segmentation information, model compression information or model training information.

[0118] In one embodiment, the model training information includes at least one of the following: a training data set, metadata corresponding to the training data set, a data preprocessing method, model features, model hyperparameters, or a model structure.

[0119] In one embodiment, the processor is further configured to perform the following operations:

[0120] Second indication information sent by the second entity is received, where the second indication information includes the first model identifier, the second model identifier, and a model information template.

[0121] In one embodiment, the processor is further configured to perform the following operations:

[0122] sending a fourth message to the second entity;

[0123] receiving a fifth message sent by the second entity in response to the fourth message;

[0124] or,

[0125] sending a fourth message to the third entity;

[0126] receiving a fifth message sent by the third entity in response to the fourth message;

[0127] Among them, the fourth message is used to request storage of the model, and the fourth message includes at least one of the following: the second model identifier, the model transmission method, the model security certificate, the model address or the model; the fifth message is used to indicate the model storage result.

[0128] In one embodiment, the processor is further configured to perform the following operations:

[0129] The model is sent to the second entity or the third entity.

[0130] In one embodiment, the processor is further configured to perform the following operations:

[0131] Sending a sixth message to the second entity, where the sixth message includes the second model identifier and model update information;

[0132] A seventh message is received, which is sent by the second entity in response to the sixth message, where the seventh message is used to indicate a model update result.

[0133] In one embodiment, the processor is further configured to perform the following operations:

[0134] sending an eighth message to the second entity;

[0135] receiving a ninth message sent by the second entity in response to the eighth message;

[0136] or,

[0137] sending an eighth message to the third entity;

[0138] receiving a ninth message sent by the third entity in response to the eighth message;

[0139] or,

[0140] sending an eighth message to the fourth entity;

[0141] receiving a ninth message sent by the fourth entity in response to the eighth message;

[0142] The eighth message is used to request verification of the model, and the eighth message includes the second model identifier and the model; and the ninth message is used to indicate a model verification result.

[0143] In a tenth aspect, an embodiment of the present disclosure provides a model registration device, applied to a second entity, the device including a memory, a transceiver, and a processor:

[0144] The memory is used to store a computer program; the transceiver is used to send and receive data under the control of the processor; and the processor is used to read the computer program in the memory and perform the following operations:

[0145] receiving a first message sent by a first entity, where the first message is used to request registration of a model, and the first message includes a first model identifier and / or model information, where the first model identifier is an identifier of the model in the first entity;

[0146] sending a second message to the first entity in response to the first message, where the second message includes a second model identifier and first indication information, where the second model identifier is a network identifier allocated to the model by the second entity, and the first indication information indicates a registration decision or registration decision related information;

[0147] A third message sent by the first entity is received, where the third message is a message determined based on the registration decision related information, and the third message includes the second model identifier and the registration decision.

[0148] In one embodiment, the model is a public model.

[0149] In one embodiment, the model information includes at least one of the following: model type, model modality, model algorithm, model accuracy, model security, model privacy, model description information, model deployment requirements, model reasoning performance, model open scope, model version, model owner information, or automated deployment script.

[0150] In one embodiment, the model description information includes at least one of the following: model accuracy, model accuracy variance, data input format, data output format, model size, full model identifier corresponding to the compressed model, model parameter quantity, model recall rate, or model F1 value;

[0151] The model deployment requirements include at least one of the following: model dependency, minimum resource requirements, or recommended resource configuration, where the resources include computing resources and storage resources.

[0152] In one embodiment, the model information further includes at least one of the following: model application scenario, model segmentation information, model compression information or model training information.

[0153] In one embodiment, the model training information includes at least one of the following: a training data set, metadata corresponding to the training data set, a data preprocessing method, model features, model hyperparameters, or a model structure.

[0154] In one embodiment, the processor is further configured to perform the following operations:

[0155] Second indication information is sent to the first entity, where the second indication information includes the first model identifier, the second model identifier, and a model information template.

[0156] In one embodiment, the processor is further configured to perform the following operations:

[0157] receiving a fourth message sent by the first entity, where the fourth message is used to request storage of the model, and the fourth message includes at least one of the following: the second model identifier, the model transmission mode, the model security certificate, the model address, or the model;

[0158] A fifth message is sent to the first entity in response to the fourth message, where the fifth message is used to indicate a model storage result.

[0159] In one embodiment, the processor is further configured to perform the following operations:

[0160] The model is received.

[0161] In one embodiment, the processor is further configured to perform the following operations:

[0162] sending a tenth message to the third entity;

[0163] receiving an eleventh message sent by the third entity in response to the tenth message;

[0164] or,

[0165] sending a tenth message to the fourth entity;

[0166] receiving an eleventh message sent by the fourth entity in response to the tenth message;

[0167] The tenth message is used to request storage of the model, and the tenth message includes at least one of the following: the second model identifier, the model transmission method, the model security certificate or the model; the eleventh message is used to indicate the model storage result.

[0168] In one embodiment, the processor is further configured to perform the following operations:

[0169] The model is sent to the third entity or the fourth entity.

[0170] In one embodiment, the processor is further configured to perform the following operations:

[0171] receiving a sixth message sent by the first entity, where the sixth message includes the second model identifier and model update information;

[0172] In response to the sixth message, a seventh message is sent to the first entity, where the seventh message is used to indicate a model update result.

[0173] In one embodiment, the processor is further configured to perform the following operations:

[0174] receiving an eighth message sent by the first entity, where the eighth message is used to request verification of the model, and the eighth message includes the second model identifier and the model;

[0175] A ninth message is sent to the first entity in response to the eighth message, where the ninth message is used to indicate a model validation result.

[0176] In an eleventh aspect, an embodiment of the present disclosure provides a model registration device, applied to a third entity, the device including a memory, a transceiver, and a processor:

[0177] The memory is used to store a computer program; the transceiver is used to send and receive data under the control of the processor; and the processor is used to read the computer program in the memory and perform the following operations:

[0178] receiving a model storage request message, where the model storage request message includes at least one of the following: a second model identifier, a model transmission mode, a model security certificate, a model address, or a model, where the second model identifier is a network identifier assigned to the model by the second entity, and the model storage request message is the fourth message sent by the first entity, the tenth message sent by the second entity, or the twelfth message sent by the fourth entity;

[0179] Send a model storage response message, where the model storage response message is used to indicate a model storage result, and the storage response message is the fifth message sent to the first entity, the eleventh message sent to the second entity, or the thirteenth message sent to the fourth entity.

[0180] In one embodiment, the processor is further configured to perform the following operations:

[0181] The model is received.

[0182] In one embodiment, the processor is further configured to perform the following operations:

[0183] receiving an eighth message sent by the first entity, where the eighth message is used to request verification of the model, and the eighth message includes the second model identifier and the model;

[0184] A ninth message is sent to the first entity in response to the eighth message, where the ninth message is used to indicate a model validation result.

[0185] In a twelfth aspect, an embodiment of the present disclosure provides a model registration device, applied to a fourth entity, the device including a memory, a transceiver, and a processor:

[0186] The memory is used to store a computer program; the transceiver is used to send and receive data under the control of the processor; and the processor is used to read the computer program in the memory and perform the following operations:

[0187] receiving an eighth message sent by the first entity, where the eighth message is used to request verification of the model, the eighth message including a second model identifier and the model, where the second model identifier is a network identifier allocated to the model by the second entity;

[0188] A ninth message is sent to the first entity in response to the eighth message, where the ninth message is used to indicate a model validation result.

[0189] In one embodiment, the processor is further configured to perform the following operations:

[0190] receiving a tenth message sent by the second entity, where the tenth message is used to request storage of the model, and the tenth message includes the second model identifier;

[0191] An eleventh message is sent to the second entity in response to the tenth message, where the eleventh message is used to indicate a model storage result.

[0192] In one embodiment, the processor is further configured to perform the following operations:

[0193] Sending a twelfth message to the third entity, where the twelfth message is used to request storage of the model, and the twelfth message includes at least one of the following: the second model identifier, the model transmission mode, the model security certificate, or the model;

[0194] A thirteenth message sent by the third entity is received, where the thirteenth message is used to indicate a model storage result.

[0195] In a thirteenth aspect, an embodiment of the present disclosure provides a non-transitory readable storage medium, which stores a computer program, and the computer program is used to enable a processor to execute any one of the methods of the first aspect.

[0196] In a fourteenth aspect, an embodiment of the present disclosure provides a non-transitory readable storage medium, which stores a computer program, and the computer program is used to enable a processor to execute any one of the methods of the second aspect.

[0197] In the fifteenth aspect, an embodiment of the present disclosure provides a non-transitory readable storage medium, which stores a computer program, and the computer program is used to enable a processor to execute any method of the third aspect.

[0198] In the sixteenth aspect, an embodiment of the present disclosure provides a non-transitory readable storage medium, which stores a computer program, and the computer program is used to enable a processor to execute any method of the fourth aspect.

[0199] The disclosed embodiments provide a model registration method, device, and storage medium, the method comprising: a first entity sending a first message to a second entity, the first message being used to request model registration, the first message including a first model identifier and / or model information, the first model identifier being the identifier of the model in the first entity; receiving a second message sent by the second entity in response to the first message, the second message including a second model identifier and first indication information, the second model identifier being the network identifier assigned to the model by the second entity, the first indication information indicating a registration decision or registration decision-related information; based on the registration decision-related information, sending a third message to the second entity, the third message including the second model identifier and the registration decision. By adding a model registration function and a corresponding registration process to the network, the network can provide services such as public model registration to users, networks, or third parties on demand, providing a foundation for endogenous intelligence.

[0200] It should be understood that the contents described in the above summary of the invention are not intended to limit the key or important features of the embodiments of the present invention, nor are they intended to limit the scope of the present invention. Other features of the present invention will become readily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0201] In order to more clearly illustrate the technical solutions in the present disclosure or related technologies, the following is a brief introduction to the drawings required for use in the embodiments or related technical descriptions. Obviously, the drawings described below are some embodiments of the present disclosure. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0202] FIG1 is a schematic diagram of an architecture of a communication system in the related art;

[0203] FIG2 is a schematic diagram of an architecture of a communication system provided by an embodiment of the present disclosure;

[0204] FIG3 is another schematic diagram of the architecture of a communication system provided by an embodiment of the present disclosure;

[0205] FIG4 is a flowchart of a model registration method according to an embodiment of the present disclosure;

[0206] FIG5 is a flow chart of a model verification method provided by an embodiment of the present disclosure;

[0207] FIG6 is a flow chart of a model storage method provided by an embodiment of the present disclosure;

[0208] FIG7 is a flow chart of a model updating method provided by an embodiment of the present disclosure;

[0209] FIG8 is a second flowchart of the model registration method provided by an embodiment of the present disclosure;

[0210] FIG9 is a third flowchart of the model registration method provided by an embodiment of the present disclosure;

[0211] FIG10 is a fourth flowchart of the model registration method provided by an embodiment of the present disclosure;

[0212] FIG11 is a fifth flowchart of the model registration method provided by an embodiment of the present disclosure;

[0213] FIG12 is a sixth flowchart of the model registration method provided by an embodiment of the present disclosure;

[0214] FIG13 is a flowchart of the model registration method according to an embodiment of the present disclosure;

[0215] FIG14 is a schematic structural diagram of a model registration device 1400 provided in an embodiment of the present disclosure;

[0216] FIG15 is a schematic structural diagram of a model registration device 1500 provided in an embodiment of the present disclosure;

[0217] FIG16 is a schematic structural diagram of a model registration device 1600 provided in an embodiment of the present disclosure;

[0218] FIG17 is a schematic structural diagram of a model registration device 1700 provided in an embodiment of the present disclosure;

[0219] FIG18 is a schematic structural diagram of a model registration device 1800 provided in an embodiment of the present disclosure;

[0220] FIG19 is a schematic structural diagram of a model registration device 1900 provided in an embodiment of the present disclosure;

[0221] FIG20 is a schematic structural diagram of a model registration device 2000 provided in an embodiment of the present disclosure;

[0222] FIG21 is a schematic structural diagram of a model registration device 2100 provided in an embodiment of the present disclosure. DETAILED DESCRIPTION

[0223] In the embodiments of the present disclosure, the term "and / or" describes the association relationship between associated objects, indicating that three relationships can exist. For example, A and / or B can represent three situations: A exists alone, A and B exist simultaneously, and B exists alone. The character " / " generally indicates that the associated objects are in an "or" relationship.

[0224] In the embodiments of the present disclosure, the term "plurality" refers to two or more than two, and other quantifiers are similar thereto.

[0225] The terms "first" and "second" appearing in the embodiments of this disclosure are used only to illustrate and distinguish the objects being described. They are not ordered, do not represent a specific limit on the number of objects in the embodiments of this disclosure, and do not constitute any limitation on the embodiments of this disclosure. For example, a first model identifier is the identifier of a model in a first entity, and a second model identifier is the identifier of a model in a network. The use of terms such as "first" and "second" is solely to distinguish between the different model identifiers, and does not indicate a difference in size, priority, or importance between the two model identifiers.

[0226] The following will be combined with the accompanying drawings in the embodiments of the present disclosure to clearly and completely describe the technical solutions in the embodiments of the present disclosure. Obviously, the embodiments described are only part of the embodiments of the present disclosure and not all of the embodiments. Based on the embodiments of the present disclosure, all other embodiments obtained by ordinary technicians in this field without making any creative efforts shall fall within the scope of protection of the present disclosure.

[0227] The embodiments of the present disclosure provide a model registration and apparatus for providing services such as public model registration to users, networks or third parties on demand.

[0228] Among them, the method and the device are based on the same application concept. Since the principles of solving problems by the method and the device are similar, the implementation of the device and the method can refer to each other, and the repeated parts will not be repeated.

[0229] The terminal device involved in the embodiments of the present disclosure may be a device that provides voice and / or data connectivity to a user, a handheld device with wireless connection function, or other processing device connected to a wireless modem. In different systems, the name of the terminal device may also be different. For example, in a 5G system, the terminal device may be called User Equipment (UE). A wireless terminal device may be a USB storage device, other personal computer memory devices, and a dongle. It may also communicate with one or more core networks (CN) via a radio access network (RAN). A wireless terminal device may be a mobile terminal device, such as a mobile phone (or "cellular" phone) and a computer with a mobile terminal device. For example, it may be a portable, pocket-sized, handheld, computer-built-in, or vehicle-mounted mobile device that exchanges language and / or data with a radio access network. For example, Personal Communication Service (PCS) phones, cordless phones, Session Initiated Protocol (SIP) phones, Wireless Local Loop (WLL) stations, Personal Digital Assistants (PDAs), personal computers, tablet computers, Machine-type Communication (MTC) terminal devices, etc. Wireless terminal devices may also be referred to as systems, subscriber units, subscriber stations, mobile stations, mobile stations, remote stations, access points, remote terminal devices, access terminal devices, user terminal devices, user agents, user devices, and wireless access points and routers / modems that meet the limitations of this definition, but are not limited in the embodiments of the present disclosure.

[0230] The access network device involved in the embodiments of the present disclosure may be a base station, which may include multiple cells providing services to the terminal. Depending on the application scenario, the base station may also be called an access point, or may be a device in the access network that communicates with the wireless terminal device through one or more sectors on the air interface, or other names. The network device may be used to interchange received air frames with Internet Protocol (IP) packets, acting as a router between the wireless terminal device and the rest of the access network, wherein the rest of the access network may include an Internet Protocol (IP) communication network. The network device may also coordinate the attribute management of the air interface. For example, the network device involved in the embodiments of the present disclosure may be an evolutionary network device (eNB or e-NodeB) in a long term evolution (LTE) system, a 5G base station (gNB) in a next generation system, etc., or a home evolved Node B (HeNB), a relay node, a femto, a pico base station, a network test device, etc., which is not limited in the embodiments of the present disclosure. In some network structures, network devices may include centralized unit (CU) nodes and distributed unit (DU) nodes, and the centralized unit and the distributed unit may also be arranged geographically separately.

[0231] Figure 1 is a schematic diagram of an architecture of a communication system in related art. As shown in Figure 1, the architecture includes terminal equipment, access network equipment, core network equipment and a data network (DN) part.

[0232] Among them, the core network equipment can also be called a core network network element or a core network functional entity. As an example, the core network functional entity may include any one of the following: a Policy Control Function (PCF) entity, a Network Slice Selection Function (NSSF) entity, a Unified Data Management (UDM) entity, a Network Repository Function (NRF) entity, a Session Management Function (SMF) entity, an Access and Mobility Management Function (AMF) entity, and a User Plane Function (UPF) entity. As an example, the functions of the above functional entities are as follows:

[0233] The PCF entity is mainly used to guide the unified policy framework of network behavior and provide policy rule information to control plane entities (such as AMF, SMF, etc.).

[0234] The AMF entity is mainly used for access control, mobility management, registration and deregistration functions.

[0235] The SMF entity is mainly used for user plane network element selection, user plane network element redirection, IP address allocation for terminal devices, as well as session establishment, modification and release and Quality of Service (QoS) control.

[0236] The UPF entity is used to receive and forward user plane data. For example, the UPF entity can receive user plane data from a service server and send it to a terminal device via an access network device. The UPF entity can also receive user plane data from a terminal device via an access network device and forward it to a service server.

[0237] NSSF entity is mainly used for network slice selection.

[0238] The UDM entity is mainly used for the contract data management of terminal devices, including the storage and management of terminal device identification, access authorization of terminal devices, etc.

[0239] The data network portion can be the data network that provides business services to users. Generally, the client is located on the terminal device, and the server is located on the data network. The data network can be a private network, such as a local area network, or an external network not controlled by the operator, such as the Internet, or a proprietary network jointly deployed by operators.

[0240] The network architecture of the current 5G mobile communication system can be shown in Figure 1. Similar to previous generations of mobile communication systems, it is designed with connection and data transmission as the center. It can provide users with connection and data transmission services, but cannot provide intelligent endogenous services such as model registration services and model storage services.

[0241] Based on the technical problems in related technologies, the embodiments of the present disclosure provide a model registration method, which can construct a public model registration process through the interaction between multiple entities in a communication system, so that the communication system can provide public model registration, storage and other services to users, networks or third parties on demand.

[0242] In the embodiment of the present disclosure, providing a public model registration service through a communication system may refer to completing a public model registration process through interaction between multiple entities in the communication system; the public model registration service may refer to the entire public model registration process.

[0243] The architecture of the communication system provided by the embodiments of the present disclosure may be as shown in Figure 2 or Figure 3, wherein the architecture shown in Figure 2 is a RAN service-oriented architecture, and the architecture shown in Figure 3 is a RAN non-service-oriented architecture. As shown in Figure 2 or Figure 3, the architecture includes a Requirement Analyzing Function (RAF) entity, a Network Adaptability Function (NAF) entity, a Network Collaboration Function (NCF) entity, at least one Network Execution Function (NEF) entity, a second entity, a third entity, and a fourth entity.

[0244] The RAF entity is primarily used to determine service requirements.

[0245] For example, the RAF entity can determine service requirements by receiving user requests; the RAF entity can also perceive changes in scenario requirements and determine service requirements by analyzing perceived data; the RAF entity can also analyze current data / third-party data and historical data through AI algorithms to determine service requirements.

[0246] The NAF entity can be used to determine the execution operation list of the target service based on service requirements. The execution operation list is a series of execution operations, including connection establishment and assurance, task calculation, data processing, AI training / inference, etc.

[0247] The NCF entity can be used to coordinate the execution progress, functions and / or resources of at least one NEF entity; it can also be used to determine the execution operation list of the target service.

[0248] Each NEF entity can be used to implement a category of functions, such as control plane network functions, user plane network functions, data plane network functions, and computing network functions. In other words, an NEF entity can perform a single network service operation, such as connection establishment and data computation, or it can perform collaborative operations to provide complex services, such as subnet establishment and AI joint training. In other words, the execution of a service can be performed by a single NEF entity or by multiple NEF entities in collaboration.

[0249] The second entity is mainly used for registration of various AI models.

[0250] The second entity may also be referred to as an AI Model Registration Function (AIMRF) entity.

[0251] The third entity is mainly used to store the AI ​​model. The AI ​​model stored in the third entity is the AI ​​model registered in the second entity.

[0252] The third entity may also be referred to as an AI Model Storage Function (AIMSF) entity.

[0253] The fourth entity is mainly used to verify the security, privacy protection and performance of AI models.

[0254] The fourth entity may also be referred to as an AI Model Authentication Function (AIMAF) entity.

[0255] The second entity can be deployed on the second terminal, the second access network device or the second core network device; the third entity can be deployed on the third terminal, the third access network device or the third core network device; the fourth entity is deployed on the fourth access network device or the fourth core network device; the second terminal and the third terminal are the same or different from each other, the second access network device, the third access network device and the fourth access network device are the same or different from each other, and the second core network device, the third core network device and the fourth core network device are the same or different from each other.

[0256] When the second entity is deployed in the second terminal, it may be deployed in the second terminal through software, hardware, or a combination of software and hardware.

[0257] When the third entity is deployed on the third terminal, it may be deployed on the third terminal through software, hardware, or a combination of software and hardware.

[0258] When the above-mentioned entity is deployed on the access network equipment, it may mean that the above-mentioned entity is deployed on an existing functional entity such as a network function (NF) entity, an operation administration and maintenance (OAM) entity, or the above-mentioned entity may be deployed as a separate device in the access network.

[0259] When the above-mentioned entity is deployed in the core network equipment, it may mean that the above-mentioned entity is deployed on an existing functional entity such as an NF entity, an OAM entity, or the above-mentioned entity may be deployed as a separate device in the core network.

[0260] The architecture of the communication system of the embodiment of the present disclosure includes not only the above-mentioned entities, but also other functional entities, such as a Network-resource Registration and Collection Function (NRCF) entity (which can be used to register network resources and receive updates on the status of network function resources), a Network Service Policy Function (NSPF) entity (which can be used to generate QoS indicator parameters and save QoS indicator parameters generated by other network elements), a PCF entity, an AMF entity, an SMF entity and / or an UPF entity, which is not limited in the embodiment of the present disclosure.

[0261] It should be noted that the names of the above entities are only examples. The embodiments of the present disclosure do not limit the names of the entities. As long as the functions implemented are the same or similar, they can be considered as the same entity.

[0262] Based on the above communication system architecture, the following details how multiple entities in the architecture interact to achieve on-demand provision of services such as model registration.

[0263] FIG4 is a flowchart of a model registration method according to an embodiment of the present disclosure. As shown in FIG4 , the method includes:

[0264] S401: A first entity sends a first message to a second entity, where the first message is used to request a registration model.

[0265] In other words, the second entity receives the first message sent by the first entity.

[0266] The first entity may be a first terminal, a first access network device, or a first core network device.

[0267] The first terminal, the second terminal and the third terminal are the same or different from each other, the first access network device, the second access network device, the third access network device and the fourth access network device are the same or different from each other, and the first core network device, the second core network device, the third core network device and the fourth core network device are the same or different from each other.

[0268] When the first entity is the first access network device, it may be an OAM deployed in the access network, or it may be a NF deployed in the access network.

[0269] When the first entity is the first core network device, it may be an OAM deployed in the core network, or it may be a NF deployed in the core network.

[0270] The model may be an existing model in the first entity, or a model trained by the first entity.

[0271] In one possible implementation, the model in the embodiment of the present disclosure may be a public model.

[0272] Exemplarily, the model may be a public AI model.

[0273] The first message may also be referred to as a model registration request message.

[0274] The first message may include a first model identifier and / or model information.

[0275] Exemplarily, the first message includes a first model identifier and model information.

[0276] In another example, the first message includes a first model identifier.

[0277] In a possible implementation, when the first message includes the first model identifier, the second entity may send second indication information to the first entity after receiving the first message. The second indication information may include the first model identifier, the second model identifier, and the model information template.

[0278] The second indication information may indicate the model information that the first entity needs to report.

[0279] The first model identifier may be an identifier of the model in the first entity, and may also be referred to as a local identifier of the model.

[0280] The second model identifier may be a network identifier assigned to the model by the second entity.

[0281] After receiving the second indication information, the first entity may again send the first message to the second entity based on the model information template. At this time, the first message carries the model information and the second model identifier.

[0282] In one possible implementation, the model information may include at least one of the following: model type, model modality, model algorithm, model accuracy, model security, model privacy, model description information, model deployment requirements, model reasoning performance, model open scope, model version, model owner information, or automated deployment script.

[0283] The model type can indicate whether the model is a basic algorithm model or an industry-specific model, etc.

[0284] Model modalities can include voice, text, image, multimodality, etc.

[0285] The model algorithm can indicate the algorithm type of the model, which can include supervised learning, deep learning, reinforcement learning, etc. The model algorithm can also be used to indicate the algorithm subtype, which can include the Deep Q-Leaning Network (DQN) algorithm, the Support Vector Machine (SVM) algorithm, etc.

[0286] When the model's algorithm type is reinforcement learning, its algorithm subtype can be the DQN algorithm; when the model's algorithm type is supervised learning, its algorithm subtype can be the SVM algorithm.

[0287] Model accuracy indicates the average accuracy of model inference.

[0288] Model security can describe the security protection characteristics and capabilities of the model, including model security authentication, security level, encryption algorithm, etc.

[0289] Model privacy can describe the privacy protection characteristics and capabilities of the model, including data privacy protection certification, privacy protection level, etc.

[0290] Model inference performance can indicate the model performance under the recommended configuration.

[0291] For example, model inference performance may include single inference latency and / or concurrency performance. Single inference latency may refer to the latency of a single model inference when running recommended resources at or below 70%; concurrency performance may refer to the maximum concurrency performance when running recommended resources.

[0292] The model open range may indicate the open range of the model.

[0293] Exemplarily, the model opening scope may be the entire network, network private use, a group of users, etc.; wherein, a group of users may refer to users in a cell, that is, the model is only open to users in a certain cell.

[0294] The model version may indicate the version of the model.

[0295] The model owner information may indicate the model owner.

[0296] The automated deployment script may indicate whether to provide an automated deployment script.

[0297] In one possible implementation, the model description information may include at least one of the following: model accuracy, variance of model accuracy, data input format, data output format, model size, full model identifier corresponding to the compressed model, model parameter quantity, model recall rate or model F1 value.

[0298] Model accuracy indicates the average accuracy of the model.

[0299] The variance of the model accuracy can refer to the degree of fluctuation of the model accuracy.

[0300] The model input format may indicate the format and structure of the model input data.

[0301] The model output format may indicate the format and structure of the model output data.

[0302] The model size may refer to the size of the model, and the units include KB, MB, GB, etc.

[0303] When the model is a compressed model, the model description information may include a full model identifier corresponding to the compressed model, which may be used to indicate an uncompressed model identifier.

[0304] The model parameter amount may indicate the parameter amount of the model.

[0305] Model recall indicates the average recall of the model.

[0306] In one possible implementation, the model deployment requirements include at least one of the following: model dependency, minimum resource requirements, or recommended resource configuration, where resources include computing resources and storage resources.

[0307] Model dependencies can indicate the dependencies of the model when it is deployed.

[0308] For example, model dependencies may include dependency frameworks, computing power types, and dependency libraries. The dependency framework may indicate the framework the model depends on, including its version; the computing power type may indicate the type of computing power used by the model, such as a graphics processing unit (GPU) or a central processing unit (CPU); and the dependency library may indicate the functional library the model depends on.

[0309] Minimum resource requirements can indicate the minimum resource requirements required for model inference.

[0310] Exemplarily, the minimum resource requirements may include minimum computing power resource requirements, minimum memory requirements, and minimum storage requirements; wherein, the minimum computing power resource requirements may indicate the type and size of computing power required for model inference, the minimum memory requirement may indicate the type and size of memory required for model inference, and the minimum storage requirement may indicate the storage size required for model inference.

[0311] The recommended resource configuration may indicate the recommended configuration for model inference.

[0312] Exemplarily, the recommended resource configuration may include the recommended amount of computing power resources, the recommended amount of memory, and the recommended amount of storage; wherein, the recommended amount of computing power resources may indicate the type and size of computing power recommended for use during model inference, the recommended amount of memory may indicate the type and size of memory recommended for use during model inference, and the recommended amount of storage may indicate the storage size recommended for use during model inference.

[0313] In a possible implementation, the model information may further include at least one of the following: model application scenario, model segmentation information, model compression information, or model training information.

[0314] The model application scenario may indicate the scenario in which the model is applied, including wireless communication, medical treatment, education, transportation, etc.

[0315] The model application scenario can also indicate the model application field, which can indicate the field of the model in the corresponding scenario. For example, in the wireless communication scenario, the model application field may include resource scheduling, mobility management, etc. In the transportation scenario, the model application field may include image recognition, route navigation, etc.

[0316] The model segmentation information may indicate whether the model is a segmented part of the model or an unsegmented model.

[0317] The model compression information may indicate whether the model is a compressed model.

[0318] In one possible implementation, the model training information may include at least one of the following: a training data set, metadata corresponding to the training data set, a data preprocessing method, model features, model hyperparameters, or a model structure.

[0319] The model training dataset may indicate dataset information for model training, and may include at least one of a dataset identifier, a dataset format, or a dataset label name.

[0320] After receiving the model information, the second entity may query the model registration information stored in itself, and may also query the third entity whether a model with equivalent performance already exists.

[0321] For example, the model registration information may be as shown in Table 1:

[0322] Table 1 Model registration information

[0323] It should be noted that in some embodiments, model registration information may not include at least one of the following: model application scenario, model segmentation information, model compression information, compressed model full model ID, model parameter count, model performance, model dependency library, or model training information. A model algorithm may not indicate the algorithm subtype. A model dependency framework may not indicate the framework version.

[0324] The model information may be a part of the model registration information or the entire model registration information.

[0325] S402: The second entity sends a second message to the first entity in response to the first message.

[0326] In other words, the first entity receives the second message sent by the second entity in response to the first message.

[0327] The second message may also be referred to as a model registration response message.

[0328] The second message may include a second model identifier and the first indication information.

[0329] In a possible implementation, if the first message includes the first model identifier and the model information, the second message may further include the first model identifier.

[0330] The first indication information may indicate a registration decision or information related to the registration decision.

[0331] The registration decision may be registration or non-registration or cancellation of registration.

[0332] Registration decision-related information may include model query results.

[0333] If the second entity decides whether to register, the first indication information indicates the registration decision.

[0334] If the first entity decides whether to register, the first indication information indicates registration decision related information. After receiving the first indication information, the first entity may execute S403.

[0335] S403: The first entity sends a third message to the second entity based on the registration decision related information.

[0336] In other words, the second entity receives the third message sent by the first entity.

[0337] The third message may include the second model identification and the registration decision.

[0338] When the registration decision related information includes a model query result, if the model query result is that there is no similar model, the registration decision in the third message may be registration; if the model query structure is that there is an existing similar model, the registration decision in the third message may be cancellation of registration.

[0339] In the embodiment shown in FIG4 , the communication system constructs a public model registration process through the interaction between various entities, so that the communication system can intelligently provide public model registration services to users, networks or third parties based on public model registration requirements during the communication process.

[0340] The embodiment shown in FIG4 illustrates a public model registration process. The public model verification process is described in detail below in conjunction with FIG5 . Model verification can be performed before model registration or during model registration.

[0341] FIG5 is a flow chart of a model verification method provided by an embodiment of the present disclosure. As shown in FIG5 , the method includes:

[0342] S501: The first entity sends an eighth message to the second entity, where the eighth message is used to request a verification model.

[0343] In other words, the second entity receives the eighth message sent by the first entity.

[0344] The eighth message may also be referred to as a model validation request message.

[0345] The eighth message may include the second model identifier and the model.

[0346] After receiving the eighth message, the second entity may verify the performance, security, privacy protection, etc. of the model.

[0347] In the embodiment of the present disclosure, the second entity has a model registration function and a model verification function.

[0348] S502: The second entity sends a ninth message to the first entity in response to the eighth message, where the ninth message is used to indicate a model verification result.

[0349] The ninth message may be referred to as a model validation response message.

[0350] The model validation result can be model validation passed or model validation failed, etc.

[0351] In a possible implementation, if the model verification result is that the model verification passed, the ninth message may further include a model security certificate.

[0352] Based on the embodiment shown in FIG5 , model verification can also be performed in the following various ways.

[0353] 1. The third entity performs model verification, that is, the first entity sends the eighth message to the third entity, and after the third entity verifies the model, it sends the ninth message to the first entity.

[0354] In this way, the third entity has model verification and model storage functions.

[0355] 2. The fourth entity performs model verification, that is, the first entity sends an eighth message to the fourth entity, and after the fourth entity verifies the model, it sends a ninth message to the first entity.

[0356] In this way, the third entity has the model verification function.

[0357] 3. The second entity initiates model verification, and the third entity performs model verification, that is, the second entity sends an eighth message to the third entity, and after the third entity verifies the model, it sends a ninth message to the second entity.

[0358] In this way, the third entity has model verification and model storage functions.

[0359] 4. The second entity initiates model verification, and the fourth entity performs model verification, that is, the second entity sends an eighth message to the fourth entity, and after the fourth entity verifies the model, it sends a ninth message to the second entity.

[0360] In this way, the third entity has the model verification function.

[0361] The verification process and the contents of the eighth and ninth messages can be found in the relevant description of the embodiment shown in FIG5 , and will not be repeated here.

[0362] If model verification is performed during the model registration process, the registration decision-related information may also include the model verification result.

[0363] In the embodiment shown in FIG5 , the communication system constructs a public model verification process through the interaction between various entities, so that the communication system can intelligently provide public model verification services to users, networks or third parties based on public model verification requirements during the communication process.

[0364] On the basis of any of the above embodiments, the public model storage process is described in detail below in conjunction with FIG6 .

[0365] FIG6 is a flow chart of a model storage method provided by an embodiment of the present disclosure. As shown in FIG6 , the method includes:

[0366] S601: The first entity sends a fourth message to the second entity, where the fourth message is used to request a storage model.

[0367] In other words, the second entity receives the fourth message sent by the first entity.

[0368] The fourth message may be referred to as a model storage request message.

[0369] The fourth message may include at least one of the following: a second model identifier, a model transmission method, a model security certificate, a model address, or a model.

[0370] The model transfer method can be determined based on the model size.

[0371] The model address may be the storage address of the model.

[0372] If the fourth message carries a model, the second entity can directly execute S602; if the fourth message does not carry a model and the model is stored locally, the first entity can send the model to the second entity, and the second entity executes S602 after receiving the model; if the fourth message does not carry a model and the model is not stored locally, the fourth message can carry a model address, and the second entity can interact with the network element storing the model based on the model address to obtain the model, and the second entity executes S602 after receiving the model.

[0373] After receiving the model, the second entity stores the model.

[0374] In the embodiment of the present disclosure, the second entity has a model registration function and a model storage function.

[0375] S602: The second entity sends a fifth message to the first entity in response to the fourth message.

[0376] In other words, the first entity receives the fifth message sent by the second entity in response to the fourth message.

[0377] The fifth message may be referred to as a model storage response message.

[0378] The fifth message may be used to instruct the model to store the result.

[0379] The model storage result may be model storage completion.

[0380] The fifth message may include the second model identifier and the model storage result.

[0381] Based on the embodiment shown in FIG6 , model verification can also be performed in the following ways:

[0382] 1. The third entity performs model storage, that is, the first entity sends the fourth message to the third entity, and after the third entity stores the model, it sends the fifth message to the first entity.

[0383] In this way, the third entity has the model storage function.

[0384] 2. The second entity initiates the model storage, and the third entity executes the model storage, that is, the second entity sends the tenth message to the third entity, and after the third entity stores the model, it sends the eleventh message to the second entity.

[0385] In this way, the third entity has the model storage function and the model verification function.

[0386] Among them, the tenth message can be used to request the storage model.

[0387] The tenth message may be referred to as a model storage request message;

[0388] The tenth message may include at least one of the following: a second model identifier, a model transmission method, a model security certificate, or a model.

[0389] If the tenth message carries a model, the third entity can send the eleventh message to the second entity after storing the model; if the tenth message does not carry a model and the model is stored locally, the second entity can send the model to the third entity, and the third entity can send the eleventh message to the second entity after receiving the model.

[0390] The eleventh message may be used to instruct the model to store the result.

[0391] The eleventh message may be referred to as a model storage response message.

[0392] The eleventh message may include the second model identifier and the model storage result.

[0393] 3. The second entity initiates model storage. After verification by the fourth entity, the third entity executes model storage. That is, the second entity sends the tenth message to the fourth entity, the fourth entity sends the twelfth message to the third entity, and after the third entity stores the model, it sends the thirteenth message to the fourth entity; the fourth entity sends the eleventh message to the second entity.

[0394] In this manner, the fourth entity has model storage and model verification functions.

[0395] The relevant descriptions of the tenth message and the eleventh message can be found in the above-mentioned method 2, which will not be repeated here.

[0396] The twelfth message may be used to request a storage model.

[0397] The twelfth message may be referred to as a model storage request message;

[0398] The twelfth message may include at least one of the following: a second model identifier, a model transmission method, a model security certificate, or a model.

[0399] If the twelfth message carries a model, the third entity may send a thirteenth message to the fourth entity after storing the model; if the twelfth message does not carry a model and the model is stored in the fourth entity, the fourth entity may send the model to the third entity, and the third entity may send the thirteenth message to the fourth entity after receiving the model.

[0400] The thirteenth message may be used to instruct the model to store the result.

[0401] The thirteenth message may be referred to as a model storage response message.

[0402] The thirteenth message may include the second model identifier and the model storage result.

[0403] In this manner, the model in the fourth entity may be obtained during the model verification process.

[0404] In the embodiment shown in FIG6 , the communication system constructs a public model storage process through the interaction between various entities, so that the communication system can intelligently provide public model storage services to users, networks or third parties based on public model storage requirements during the communication process.

[0405] Based on any of the above embodiments, the public model updating process is described in detail below with reference to FIG7 .

[0406] FIG7 is a flow chart of a model updating method provided by an embodiment of the present disclosure. As shown in FIG7 , the method includes:

[0407] S701: The first entity sends a sixth message to the second entity.

[0408] In other words, the second entity receives the sixth message sent by the first entity.

[0409] The sixth message may be used to indicate the registration information of the updated model.

[0410] The sixth message may include the second model identifier and model update information.

[0411] If the model information is part of the model registration information during model registration, the model update information can be used to complete the model registration information. In this case, the model update information includes information that is not mentioned in the model information but exists in the model registration information.

[0412] If the first entity optimizes the registered model, the model update information includes the optimized model information.

[0413] S702: The second entity sends a seventh message to the first entity in response to the sixth message.

[0414] In other words, the first entity receives the seventh message sent by the second entity in response to the sixth message.

[0415] The seventh message may be used to indicate the model update result.

[0416] The model update result may include model update completion.

[0417] The seventh message may include the second model identifier and the model update result.

[0418] In the embodiment shown in FIG7 , the communication system constructs a public model update process through the interaction between various entities, so that the communication system can intelligently provide public model update services to users, networks or third parties based on public model update requirements during the communication process.

[0419] The above-mentioned multiple embodiments and methods can be combined arbitrarily.

[0420] Exemplarily, the embodiment shown in Figure 4 is combined with the embodiment shown in Figure 5, the embodiment shown in Figure 4 is combined with method one below the embodiment shown in Figure 5, the embodiment shown in Figure 4 is combined with method two below the embodiment shown in Figure 5, the embodiment shown in Figure 4 is combined with method three below the embodiment shown in Figure 5, and the embodiment shown in Figure 4 is combined with method four below the embodiment shown in Figure 5.

[0421] As another exemplary embodiment, the embodiment shown in Figure 4 and the embodiment shown in Figure 5 are combined with the embodiment shown in Figure 6, the embodiment shown in Figure 4 and the embodiment below the method one are combined with the embodiment shown in Figure 6, the embodiment shown in Figure 4 and the embodiment below the method two are combined with the embodiment shown in Figure 6, the embodiment shown in Figure 4 and the embodiment below the method three are combined with the embodiment shown in Figure 6, the embodiment shown in Figure 4 and the embodiment below the method four are combined with the embodiment shown in Figure 6, the embodiment shown in Figure 4 and the embodiment below the method one are combined with the embodiment shown in Figure 6, the embodiment shown in Figure 4 and the embodiment below the method one are combined with the embodiment shown in Figure 6, the embodiment shown in Figure 4 and the embodiment below the method two are combined with the embodiment shown in Figure 4 and the embodiment below the method one are combined with the embodiment shown in Figure 6, the embodiment shown in Figure 4 and the embodiment below the method three are combined with the method one below the embodiment shown in Figure 6, the embodiment shown in Figure 4 and the embodiment below the method four are combined with the method one below the embodiment shown in Figure 6, and the embodiment shown in Figure 4 and the embodiment below the method 6 , the embodiment shown in FIG4 and FIG5 is combined with the embodiment shown in FIG6 below the second mode, the embodiment shown in FIG4 and FIG5 below the second mode, the embodiment shown in FIG4 and FIG5 below the second mode, the embodiment shown in FIG4 and FIG5 below the third mode, the embodiment shown in FIG4 and FIG5 below the second mode, the embodiment shown in FIG4 and FIG5 below the fourth mode, the embodiment shown in FIG4 and FIG5 below the third mode, the embodiment shown in FIG4 and FIG5 below the third mode, the embodiment shown in FIG4 and FIG5 below the third mode, the embodiment shown in FIG4 and FIG5 below the third mode, the embodiment shown in FIG4 and FIG5 below the third mode, the embodiment shown in FIG4 and FIG5 below the third mode, the embodiment shown in FIG4 and FIG5 below the third mode, the embodiment shown in FIG4 and FIG5 below the third mode, the embodiment shown in FIG4 and FIG5 below the third mode.

[0422] As another example, the embodiment shown in Figure 4, the embodiment shown in Figure 5, the embodiment shown in Figure 6 are combined with the embodiment shown in Figure 7, the embodiment shown in Figure 4, the embodiment shown in Figure 5, the embodiment shown in Figure 6 below the method one is combined with the embodiment shown in Figure 7, the embodiment shown in Figure 4, the embodiment shown in Figure 5, the embodiment shown in Figure 6 below the method two is combined with the embodiment shown in Figure 7, the embodiment shown in Figure 4, the embodiment shown in Figure 5, the embodiment shown in Figure 6 below the method three is combined with the embodiment shown in Figure 7, the embodiment shown in Figure 4, the embodiment shown in Figure 5 below the method one, the embodiment shown in Figure 6 and the embodiment shown in Figure 7 are combined, the embodiment shown in Figure 4, the embodiment shown in Figure 5 below the method one The embodiment shown in Figure 4 and Figure 5 are combined with the embodiment shown in Figure 7. The embodiment shown in Figure 4 and Figure 5 are combined with the embodiment shown in Figure 6. The embodiment shown in Figure 4 and Figure 5 are combined with the embodiment shown in Figure 6. The embodiment shown in Figure 4 and Figure 5 are combined with the embodiment shown in Figure 7. The embodiment shown in Figure 4 and Figure 5 are combined with the embodiment shown in Figure 6. The embodiment shown in Figure 4 and Figure 5 are combined with the embodiment shown in Figure 7. The second embodiment shown in FIG4 and the second embodiment shown in FIG5 are combined with the embodiment shown in FIG6 , the third embodiment shown in FIG4 and the third embodiment shown in FIG5 are combined with the embodiment shown in FIG6 , the third embodiment shown in FIG4 and the third embodiment shown in FIG5 are combined with the embodiment shown in FIG6 , the third embodiment shown in FIG4 and the third embodiment shown in FIG5 are combined with the embodiment shown in FIG6 , the first embodiment shown in FIG4 and the third embodiment shown in FIG5 are combined with the embodiment shown in FIG6 , the third embodiment shown in FIG4 and the third embodiment shown in FIG5 are combined with the embodiment shown in FIG6 , the second embodiment shown in FIG4 and the third embodiment shown in FIG5 are combined with the embodiment shown in FIG6 . 7 , the embodiment shown in FIG4 , the embodiment shown in FIG5 , the embodiment shown in FIG6 , the embodiment shown in FIG6 , the embodiment shown in FIG7 , the embodiment shown in FIG4 , the embodiment shown in FIG5 ...7 , the embodiment shown in FIG5 , the embodiment shown in FIG6 , the embodiment shown in FIG7 , the embodiment shown in FIG7

[0423] Based on the above-mentioned embodiments, several combined embodiments obtained by combining at least two of the above-mentioned embodiments are shown below.

[0424] Example 1, as shown in FIG8 , the method includes:

[0425] S801: A first entity sends a first message to a second entity, where the first message is used to request a registration model.

[0426] It should be noted that the execution process of S801 can refer to the execution process of S401. The only difference is that the model information in the first message may include necessary information of the model, and the necessary information may include model type, model modality, model algorithm, model accuracy, model description information, and model deployment requirements; or, the model information in the first message may include some of the above-mentioned necessary information; or, the model information in the first message carries all the information in the model registration information.

[0427] In some embodiments, the first message may also include a model application scenario.

[0428] S802: Model query.

[0429] The model query process can be found in the relevant description in S401 and will not be repeated here.

[0430] S803: Model verification.

[0431] The model verification process can be referred to the relevant description in the embodiment shown in FIG5 , which will not be repeated here.

[0432] S804: The second entity sends a second message to the first entity in response to the first message, where the second message includes the first model identifier, the second model identifier, and registration decision related information.

[0433] S805: The first entity sends a third message to the second entity based on the registration decision related information, where the third message includes the second model identifier and the registration decision.

[0434] The execution process of S804 to S805 can refer to the execution process of S402 to S403, and will not be repeated here.

[0435] S806: The first entity sends a fourth message to the third entity, where the fourth message is used to request a storage model.

[0436] The fourth message may include the second model identifier, the model transmission mode, and the model security certificate.

[0437] In some embodiments, the fourth message may also carry a model address and / or model.

[0438] If the fourth message does not carry the model, S807 may be executed; if the fourth message carries the model, S808 may be executed.

[0439] S807: Model transmission.

[0440] S808. The third entity sends a fifth message to the first entity in response to the fourth message, where the fifth message includes the second model identifier and the model storage result.

[0441] It should be noted that the execution process of S806 to S808 can refer to the execution process of S601 to S602, and will not be repeated here.

[0442] S809: The first entity sends a sixth message to the second entity, where the sixth message includes a second model identifier and model update information.

[0443] S810: The second entity sends a seventh message to the first entity in response to the sixth message, where the seventh message includes a second model identifier and a model update result.

[0444] It should be noted that the execution process of S809 to S810 can refer to the execution process of S701 to S702, and will not be repeated here.

[0445] Example 2, as shown in FIG9 , the method includes:

[0446] S901. Model verification.

[0447] The model verification process can be referred to the relevant description in the embodiment shown in FIG5 , which will not be repeated here.

[0448] S902: The first entity sends a first message to the second entity, where the first message is used to request a registration model.

[0449] S903: Model query.

[0450] The execution process of S902 to S903 can refer to the execution process of S801 to S802, and will not be repeated here.

[0451] S904: The second entity sends a second message to the first entity in response to the first message, where the second message includes the first model identifier, the second model identifier, and registration decision related information.

[0452] S905: The first entity sends a third message to the second entity based on the registration decision related information, where the third message includes the second model identifier and the registration decision.

[0453] S906: The first entity sends a fourth message to the third entity, where the fourth message is used to request a storage model.

[0454] S907: Model transmission.

[0455] S908. The third entity sends a fifth message to the first entity in response to the fourth message, where the fifth message includes the second model identifier and the model storage result.

[0456] S909: The first entity sends a sixth message to the second entity, where the sixth message includes a second model identifier and model update information.

[0457] S910: The second entity sends a seventh message to the first entity in response to the sixth message, where the seventh message includes a second model identifier and a model update result.

[0458] It should be noted that the execution process of S904 to S910 can refer to the execution process of S804 to S810, and will not be repeated here.

[0459] The difference between Example 2 and Example 1 is that model verification is completed before model registration.

[0460] Example 3, as shown in FIG10 , the method includes:

[0461] S1001. A first entity sends a first message to a second entity, where the first message is used to request a registration model.

[0462] S1002. Model query.

[0463] S1003. Model verification.

[0464] S1004: The second entity sends a second message to the first entity in response to the first message, where the second message includes the first model identifier, the second model identifier, and registration decision related information.

[0465] S1005. The first entity sends a third message to the second entity based on the registration decision related information. The third message includes the second model identifier and the registration decision.

[0466] It should be noted that the execution process of S1001 to S1005 can refer to the execution process of S801 to S805, and will not be repeated here.

[0467] S1006. The second entity sends a tenth message to the fourth entity. The tenth message is used to request a storage model, and the tenth message includes a second model identifier.

[0468] S1007. The fourth entity sends a twelfth message to the third entity. The twelfth message includes the second model identifier, the model transmission mode, and the model security certificate.

[0469] In some embodiments, the twelfth message may also carry a model.

[0470] If the twelfth message does not carry the model, S1008 may be executed; if the twelfth message carries the model, S1009 may be executed.

[0471] S1008. Model transmission.

[0472] S1009. The third entity sends a thirteenth message to the fourth entity in response to the twelfth message, where the thirteenth message includes the second model identifier and the model storage result.

[0473] S1010. The fourth entity sends an eleventh message to the second entity, where the eleventh message includes a second model identifier.

[0474] It should be noted that the execution process of S1006 to S1010 can refer to the relevant description of the model verification method 3 below the embodiment shown in Figure 6, and will not be repeated here.

[0475] S1011. The first entity sends a sixth message to the second entity, where the sixth message includes a second model identifier and model update information.

[0476] S1012: The second entity sends a seventh message to the first entity in response to the sixth message, where the seventh message includes a second model identifier and a model update result.

[0477] It should be noted that the execution process of S1011 to S1012 can refer to the execution process of S701 to S702, and will not be repeated here.

[0478] Example 4, as shown in FIG11 , the method includes:

[0479] S1101. Model verification.

[0480] The model verification process can be referred to the relevant description in the embodiment shown in FIG5 , which will not be repeated here.

[0481] S1102: The first entity sends a first message to the second entity, where the first message is used to request a registration model.

[0482] S1103. Model query.

[0483] S1104: The second entity sends a second message to the first entity in response to the first message, where the second message includes the first model identifier, the second model identifier, and registration decision related information.

[0484] S1105: The first entity sends a third message to the second entity based on the registration decision related information, where the third message includes the second model identifier and the registration decision.

[0485] S1106. The second entity sends a tenth message to the fourth entity. The tenth message is used to request a storage model, and the tenth message includes a second model identifier.

[0486] S1107. The fourth entity sends a twelfth message to the third entity. The twelfth message includes the second model identifier, the model transmission mode, and the model security certificate.

[0487] In some embodiments, the twelfth message may also carry a model.

[0488] S1108: Model transmission.

[0489] S1109. The third entity sends a thirteenth message to the fourth entity in response to the twelfth message, where the thirteenth message includes the second model identifier and the model storage result.

[0490] S1110. The fourth entity sends an eleventh message to the second entity, where the eleventh message includes a second model identifier.

[0491] S1111. The first entity sends a sixth message to the second entity, where the sixth message includes a second model identifier and model update information.

[0492] S1112. The second entity sends a seventh message to the first entity in response to the sixth message, where the seventh message includes a second model identifier and a model update result.

[0493] It should be noted that the execution process of S1104 to S1112 can refer to the execution process of S1004 to S1012, and will not be repeated here.

[0494] The difference between Example 4 and Example 3 is that model verification is completed before model registration.

[0495] In embodiment 5, as shown in FIG12 , the method includes:

[0496] S1201: A first entity sends a first message to a second entity, where the first message is used to request a registration model.

[0497] S1202: Model query.

[0498] S1203. Model verification.

[0499] S1204: The second entity sends a second message to the first entity in response to the first message, where the second message includes the first model identifier, the second model identifier, and registration decision related information.

[0500] S1205: The first entity sends a third message to the second entity based on the registration decision related information, where the third message includes the second model identifier and the registration decision.

[0501] S1206. The first entity sends a fourth message to the second entity, where the fourth message is used to request a storage model.

[0502] S1207: Model transmission.

[0503] S1208. The second entity sends a fifth message to the first entity in response to the fourth message, where the fifth message includes the second model identifier and the model storage result.

[0504] S1209: The first entity sends a sixth message to the second entity, where the sixth message includes a second model identifier and model update information.

[0505] S1210: The second entity sends a seventh message to the first entity in response to the sixth message, where the seventh message includes a second model identifier and a model update result.

[0506] It should be noted that the execution process of S1201 to S1210 can refer to the execution process of S801 to S810, and will not be repeated here.

[0507] The difference between Example 5 and Example 1 is that the second entity has a model registration function and a model storage function.

[0508] When the second entity has the model registration function, the model storage function, and the model verification function, the model verification in S1203 is completed by the interaction between the first entity and the second entity.

[0509] When the third entity has the model storage function and the model verification function, the model verification in S1203 is completed by the interaction between the first entity and the third entity, and the model storage from S1206 to S1208 is completed by the interaction between the first entity and the third entity.

[0510] Example 6, as shown in FIG13 , the method includes:

[0511] S1301: A first entity sends a first message to a second entity. The first message is used to request model registration and includes a first model identifier.

[0512] S1302: The second entity sends second indication information to the first entity in response to the first message, where the second indication information includes the first model identifier, the second model identifier, and the model information template.

[0513] S1303: The first entity sends a first message to the second entity based on the model information template. The first message is used to request model registration and includes a second model identifier and model information.

[0514] The model information is determined based on the model information template.

[0515] S1304: Model query.

[0516] S1305: Model verification.

[0517] S1306: The second entity sends a second message to the first entity in response to the first message, where the second message includes a second model identifier and registration decision related information.

[0518] S1307: The first entity sends a third message to the second entity based on the registration decision related information, where the third message includes the second model identifier and the registration decision.

[0519] S1308. The first entity sends a fourth message to the third entity, where the fourth message is used to request a storage model.

[0520] S1309: Model transmission.

[0521] S1310: The third entity sends a fifth message to the first entity in response to the fourth message, where the fifth message includes a second model identifier and a model storage result.

[0522] S1311. The first entity sends a sixth message to the second entity, where the sixth message includes a second model identifier and model update information.

[0523] S1312: The second entity sends a seventh message to the first entity in response to the sixth message, where the seventh message includes a second model identifier and a model update result.

[0524] It should be noted that the execution process of S1304 to S1312 can refer to the execution process of S802 to S810, and will not be repeated here.

[0525] It should be noted that the above embodiments are merely examples, and model registration, model verification, model storage, and model update may also be combined in other ways.

[0526] FIG14 is a schematic diagram of the structure of a model registration device 1400 provided in an embodiment of the present disclosure. As shown in FIG14 , the device 1400 includes: a memory 1410 , a transceiver 1420 , and a processor 1430 .

[0527] The memory 1410 is used to store computer programs. The transceiver 1420 is used to send and receive data under the control of the processor 1430. The processor 1430 is used to read the computer program stored in the memory 1410 and perform the following operations:

[0528] Sending a first message to the second entity, where the first message is used to request registration of a model, and the first message includes a first model identifier and / or model information, where the first model identifier is an identifier of the model in the first entity;

[0529] receiving a second message sent by the second entity in response to the first message, the second message including a second model identifier and first indication information, the second model identifier being a network identifier allocated by the second entity to the model, and the first indication information indicating a registration decision or registration decision related information;

[0530] Based on the registration decision related information, a third message is sent to the second entity, where the third message includes the second model identifier and the registration decision.

[0531] In one embodiment, the model is a public model.

[0532] In one embodiment, the model information includes at least one of the following: model type, model modality, model algorithm, model accuracy, model security, model privacy, model description information, model deployment requirements, model reasoning performance, model open scope, model version, model owner information, or automated deployment script.

[0533] In one embodiment, the model description information includes at least one of the following: model accuracy, model accuracy variance, data input format, data output format, model size, full model identifier corresponding to the compressed model, model parameter number, model recall rate, or model F1 value;

[0534] Model deployment requirements include at least one of the following: model dependencies, minimum resource requirements, or recommended resource configurations. Resources include computing resources and storage resources.

[0535] In one embodiment, the model information also includes at least one of the following: model application scenario, model segmentation information, model compression information or model training information.

[0536] In one embodiment, the model training information includes at least one of the following: a training data set, metadata corresponding to the training data set, a data preprocessing method, model features, model hyperparameters, or a model structure.

[0537] In one embodiment, the processor 1430 is further configured to perform the following operations:

[0538] Second indication information sent by the second entity is received, where the second indication information includes the first model identifier, the second model identifier, and the model information template.

[0539] In one embodiment, the processor 1430 is further configured to perform the following operations:

[0540] sending a fourth message to the second entity;

[0541] receiving a fifth message sent by the second entity in response to the fourth message;

[0542] or,

[0543] sending a fourth message to the third entity;

[0544] receiving a fifth message sent by the third entity in response to the fourth message;

[0545] Among them, the fourth message is used to request storage of the model, and the fourth message includes at least one of the following: a second model identifier, a model transmission method, a model security certificate, a model address or a model; the fifth message is used to indicate the model storage result.

[0546] In one embodiment, the processor 1430 is further configured to perform the following operations:

[0547] Send the model to a second or third entity.

[0548] In one embodiment, the processor 1430 is further configured to perform the following operations:

[0549] Sending a sixth message to the second entity, where the sixth message includes the second model identifier and model update information;

[0550] A seventh message sent by the second entity in response to the sixth message is received, where the seventh message is used to indicate a model update result.

[0551] In one embodiment, the processor 1430 is further configured to perform the following operations:

[0552] sending an eighth message to the second entity;

[0553] receiving a ninth message sent by the second entity in response to the eighth message;

[0554] or,

[0555] sending an eighth message to the third entity;

[0556] receiving a ninth message sent by the third entity in response to the eighth message;

[0557] or,

[0558] sending an eighth message to the fourth entity;

[0559] receiving a ninth message sent by the fourth entity in response to the eighth message;

[0560] The eighth message is used to request model verification, and includes the second model identifier and the model; the ninth message is used to indicate the model verification result.

[0561] It should be noted here that the above-mentioned model registration device 1400 provided by the present disclosure can implement all the method steps implemented by the first entity in the above-mentioned method embodiment, and can achieve the same technical effect. The parts and beneficial effects of this embodiment that are the same as the method embodiment will not be repeated here.

[0562] FIG15 is a schematic diagram of the structure of a model registration device 1500 provided in an embodiment of the present disclosure. As shown in FIG15 , the device 1500 includes: a memory 1510 , a transceiver 1520 , and a processor 1530 .

[0563] The memory 1510 is used to store computer programs. The transceiver 1520 is used to send and receive data under the control of the processor 1530. The processor 1530 is used to read the computer program stored in the memory 1510 and perform the following operations:

[0564] receiving a first message sent by a first entity, where the first message is used to request registration of a model, and the first message includes a first model identifier and / or model information, where the first model identifier is an identifier of the model in the first entity;

[0565] Sending a second message to the first entity in response to the first message, the second message including a second model identifier and first indication information, the second model identifier being a network identifier allocated by the second entity to the model, and the first indication information indicating a registration decision or registration decision related information;

[0566] A third message sent by the first entity is received, where the third message is a message determined based on the registration decision related information, and the third message includes the second model identifier and the registration decision.

[0567] In one embodiment, the model is a public model.

[0568] In one embodiment, the model information includes at least one of the following: model type, model modality, model algorithm, model accuracy, model security, model privacy, model description information, model deployment requirements, model reasoning performance, model open scope, model version, model owner information, or automated deployment script.

[0569] In one embodiment, the model description information includes at least one of the following: model accuracy, model accuracy variance, data input format, data output format, model size, full model identifier corresponding to the compressed model, model parameter number, model recall rate, or model F1 value;

[0570] Model deployment requirements include at least one of the following: model dependencies, minimum resource requirements, or recommended resource configurations. Resources include computing resources and storage resources.

[0571] In one embodiment, the model information also includes at least one of the following: model application scenario, model segmentation information, model compression information or model training information.

[0572] In one embodiment, the model training information includes at least one of the following: a training data set, metadata corresponding to the training data set, a data preprocessing method, model features, model hyperparameters, or a model structure.

[0573] In one embodiment, the processor 1530 is further configured to perform the following operations:

[0574] Second indication information is sent to the first entity, where the second indication information includes the first model identifier, the second model identifier, and the model information template.

[0575] In one embodiment, the processor 1530 is further configured to perform the following operations:

[0576] receiving a fourth message sent by the first entity, where the fourth message is used to request storage of the model, and the fourth message includes at least one of the following: a second model identifier, a model transmission method, a model security certificate, a model address, or a model;

[0577] A fifth message is sent to the first entity in response to the fourth message, where the fifth message is used to instruct the model to store the result.

[0578] In one embodiment, the processor 1530 is further configured to perform the following operations:

[0579] Receive model.

[0580] In one embodiment, the processor 1530 is further configured to perform the following operations:

[0581] sending a tenth message to the third entity;

[0582] receiving an eleventh message sent by the third entity in response to the tenth message;

[0583] or,

[0584] sending a tenth message to the fourth entity;

[0585] receiving an eleventh message sent by the fourth entity in response to the tenth message;

[0586] The tenth message is used to request storage of the model, and the tenth message includes at least one of the following: a second model identifier, a model transmission method, a model security certificate or a model; the eleventh message is used to indicate a model storage result.

[0587] In one embodiment, the processor 1530 is further configured to perform the following operations:

[0588] Send the model to a third entity or a fourth entity.

[0589] In one embodiment, the processor 1530 is further configured to perform the following operations:

[0590] receiving a sixth message sent by the first entity, where the sixth message includes a second model identifier and model update information;

[0591] A seventh message is sent to the first entity in response to the sixth message, where the seventh message is used to indicate a model update result.

[0592] In one embodiment, the processor 1530 is further configured to perform the following operations:

[0593] receiving an eighth message sent by the first entity, where the eighth message is used to request model verification, and the eighth message includes the second model identifier and the model;

[0594] A ninth message is sent to the first entity in response to the eighth message, where the ninth message is used to indicate a model validation result.

[0595] It should be noted here that the above-mentioned model registration device 1500 provided by the present disclosure can implement all the method steps implemented by the second entity in the above-mentioned method embodiment, and can achieve the same technical effect. The parts and beneficial effects of this embodiment that are the same as the method embodiment will not be repeated here.

[0596] FIG16 is a schematic diagram of the structure of a model registration device 1600 provided in an embodiment of the present disclosure. As shown in FIG16 , the device 1600 includes: a memory 1610 , a transceiver 1620 , and a processor 1630 .

[0597] The memory 1610 is used to store computer programs. The transceiver 1620 is used to send and receive data under the control of the processor 1630. The processor 1630 is used to read the computer program stored in the memory 1610 and perform the following operations:

[0598] receiving a model storage request message, where the model storage request message includes at least one of the following: a second model identifier, a model transmission mode, a model security certificate, a model address, or a model, where the second model identifier is a network identifier assigned to the model by the second entity, and the model storage request message is the fourth message sent by the first entity, the tenth message sent by the second entity, or the twelfth message sent by the fourth entity;

[0599] A model storage response message is sent, where the model storage response message is used to indicate the model storage result. The storage response message is the fifth message sent to the first entity, the eleventh message sent to the second entity, or the thirteenth message sent to the fourth entity.

[0600] In one embodiment, the processor 1630 is further configured to perform the following operations:

[0601] Receive model.

[0602] In one embodiment, the processor 1630 is further configured to perform the following operations:

[0603] receiving an eighth message sent by the first entity, where the eighth message is used to request model verification, and the eighth message includes the second model identifier and the model;

[0604] A ninth message is sent to the first entity in response to the eighth message, where the ninth message is used to indicate a model validation result.

[0605] It should be noted here that the above-mentioned model registration device 1600 provided by the present disclosure can implement all the method steps implemented by the third entity in the above-mentioned method embodiment, and can achieve the same technical effect. The parts and beneficial effects of this embodiment that are the same as the method embodiment will not be repeated here.

[0606] FIG17 is a schematic diagram of the structure of a model registration device 1700 provided in an embodiment of the present disclosure. As shown in FIG17 , the device 1700 includes: a memory 1710 , a transceiver 1720 , and a processor 1730 .

[0607] The memory 1710 is used to store computer programs. The transceiver 1720 is used to send and receive data under the control of the processor 1730. The processor 1730 is used to read the computer program stored in the memory 1710 and perform the following operations:

[0608] receiving an eighth message sent by the first entity, where the eighth message is used to request verification of the model, and the eighth message includes a second model identifier and the model, where the second model identifier is a network identifier allocated by the second entity to the model;

[0609] A ninth message is sent to the first entity in response to the eighth message, where the ninth message is used to indicate a model validation result.

[0610] In one embodiment, the processor 1730 is further configured to perform the following operations:

[0611] receiving a tenth message sent by the second entity, where the tenth message is used to request a storage model and includes a second model identifier;

[0612] An eleventh message is sent to the second entity in response to the tenth message, where the eleventh message is used to instruct the model to store the result.

[0613] In one embodiment, the processor 1730 is further configured to perform the following operations:

[0614] Sending a twelfth message to the third entity, where the twelfth message is used to request storage of the model, and the twelfth message includes at least one of the following: a second model identifier, a model transmission method, a model security certificate, or a model;

[0615] A thirteenth message sent by the third entity is received, where the thirteenth message is used to indicate a model storage result.

[0616] It should be noted here that the above-mentioned model registration device 1700 provided by the present disclosure can implement all the method steps implemented by the fourth entity in the above-mentioned method embodiment, and can achieve the same technical effect. The parts and beneficial effects of this embodiment that are the same as the method embodiment will not be repeated here.

[0617] Among them, in Figures 14 to 17, when the model registration device is a terminal, the device can also include a user interface. For different terminals, the user interface can also be an interface that can be connected to external and internal devices. The connected devices include but are not limited to a keypad, display, speaker, microphone, joystick, etc.

[0618] The bus architecture can include any number of interconnected buses and bridges, linking various circuits such as one or more processors represented by processor 1430, processor 1530, processor 1630, or processor 1730, and memory 1410, memory 1510, memory 1610, or memory 1710. The bus architecture can also link various other circuits such as peripherals, voltage regulators, and power management circuits, all of which are well known in the art and, therefore, will not be described further herein. A bus interface provides an interface. A transceiver can be multiple components, including a transmitter and a receiver, providing a means for communicating with various other devices over a transmission medium, such as a wireless channel, a wired channel, or an optical cable. The processor is responsible for managing the bus architecture and general processing, and the memory can store data used by the processor when performing operations.

[0619] In some embodiments, the processor may be a central processing unit (CPU), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA), or a complex programmable logic device (CPLD), and the processor may also adopt a multi-core architecture.

[0620] FIG18 is a schematic diagram of the structure of a model registration device 1800 provided in an embodiment of the present disclosure. As shown in FIG18 , the device 1800 includes:

[0621] A first sending unit 1810 is configured to send a first message to a second entity, where the first message is used to request registration of a model, and the first message includes a first model identifier and / or model information, where the first model identifier is an identifier of the model in the first entity;

[0622] A first receiving unit 1820 is configured to receive a second message sent by a second entity in response to the first message, the second message including a second model identifier and first indication information, the second model identifier being a network identifier allocated by the second entity to the model, and the first indication information indicating a registration decision or registration decision related information;

[0623] The second sending unit 1830 is configured to send a third message to the second entity based on the registration decision related information, where the third message includes the second model identifier and the registration decision.

[0624] In one embodiment, the model is a public model.

[0625] In one embodiment, the model information includes at least one of the following: model type, model modality, model algorithm, model accuracy, model security, model privacy, model description information, model deployment requirements, model reasoning performance, model open scope, model version, model owner information, or automated deployment script.

[0626] In one embodiment, the model description information includes at least one of the following: model accuracy, model accuracy variance, data input format, data output format, model size, full model identifier corresponding to the compressed model, model parameter number, model recall rate, or model F1 value;

[0627] Model deployment requirements include at least one of the following: model dependencies, minimum resource requirements, or recommended resource configurations. Resources include computing resources and storage resources.

[0628] In one embodiment, the model information also includes at least one of the following: model application scenario, model segmentation information, model compression information or model training information.

[0629] In one embodiment, the model training information includes at least one of the following: a training data set, metadata corresponding to the training data set, a data preprocessing method, model features, model hyperparameters, or a model structure.

[0630] In one embodiment, the apparatus 1800 further includes:

[0631] The second receiving unit is configured to receive second indication information sent by the second entity, where the second indication information includes the first model identifier, the second model identifier, and the model information template.

[0632] In one embodiment, the apparatus 1800 further includes:

[0633] A third sending unit, configured to send a fourth message to the second entity;

[0634] a third receiving unit, configured to receive a fifth message sent by the second entity in response to the fourth message;

[0635] or,

[0636] A third sending unit, configured to send a fourth message to a third entity;

[0637] a third receiving unit, configured to receive a fifth message sent by the third entity in response to the fourth message;

[0638] Among them, the fourth message is used to request storage of the model, and the fourth message includes at least one of the following: a second model identifier, a model transmission method, a model security certificate, a model address or a model; the fifth message is used to indicate the model storage result.

[0639] In one embodiment, the apparatus 1800 further includes:

[0640] The fourth sending unit is configured to send the model to the second entity or the third entity.

[0641] In one embodiment, the apparatus 1800 further includes:

[0642] a fifth sending unit, configured to send a sixth message to the second entity, where the sixth message includes the second model identifier and the model update information;

[0643] The fourth receiving unit is configured to receive a seventh message sent by the second entity in response to the sixth message, where the seventh message is used to indicate a model update result.

[0644] In one embodiment, the apparatus 1800 further includes:

[0645] a sixth sending unit, configured to send an eighth message to the second entity;

[0646] a fifth receiving unit, configured to receive a ninth message sent by the second entity in response to the eighth message;

[0647] or,

[0648] a sixth sending unit, configured to send an eighth message to the third entity;

[0649] a fifth receiving unit, configured to receive a ninth message sent by the third entity in response to the eighth message;

[0650] or,

[0651] a sixth sending unit, configured to send an eighth message to the fourth entity;

[0652] a fifth receiving unit, configured to receive a ninth message sent by the fourth entity in response to the eighth message;

[0653] The eighth message is used to request model verification, and includes the second model identifier and the model; the ninth message is used to indicate the model verification result.

[0654] In one embodiment, the first entity is a first terminal, a first access network device, or a first core network device;

[0655] The second entity is deployed in a second terminal, a second access network device or a second core network device;

[0656] The third entity is deployed in a third terminal, a third access network device or a third core network device;

[0657] The fourth entity is deployed in a fourth access network device or a fourth core network device;

[0658] The first terminal, the second terminal and the third terminal are the same or different from each other, the first access network device, the second access network device, the third access network device and the fourth access network device are the same or different from each other, and the first core network device, the second core network device, the third core network device and the fourth core network device are the same or different from each other.

[0659] It should be noted here that the above-mentioned model registration device 1800 provided by the present disclosure can implement all the method steps implemented by the first entity in the above-mentioned method embodiment, and can achieve the same technical effect. The parts and beneficial effects of this embodiment that are the same as the method embodiment will not be repeated here.

[0660] FIG19 is a schematic diagram of the structure of a model registration device 1900 provided in an embodiment of the present disclosure. As shown in FIG19 , the device 1900 includes:

[0661] A first receiving unit 1910 is configured to receive a first message sent by a first entity, where the first message is used to request model registration and includes a first model identifier and / or model information. The first model identifier is an identifier of the model in the first entity.

[0662] A first sending unit 1920 is configured to send a second message to the first entity in response to the first message, where the second message includes a second model identifier and first indication information, where the second model identifier is a network identifier allocated by the second entity to the model, and the first indication information indicates a registration decision or registration decision related information;

[0663] The second receiving unit 1930 is configured to receive a third message sent by the first entity, where the third message is a message determined based on the registration decision related information, and includes the second model identifier and the registration decision.

[0664] In one embodiment, the model is a public model.

[0665] In one embodiment, the model information includes at least one of the following: model type, model modality, model algorithm, model accuracy, model security, model privacy, model description information, model deployment requirements, model reasoning performance, model open scope, model version, model owner information, or automated deployment script.

[0666] In one embodiment, the model description information includes at least one of the following: model accuracy, model accuracy variance, data input format, data output format, model size, full model identifier corresponding to the compressed model, model parameter number, model recall rate, or model F1 value;

[0667] Model deployment requirements include at least one of the following: model dependencies, minimum resource requirements, or recommended resource configurations. Resources include computing resources and storage resources.

[0668] In one embodiment, the first model information also includes at least one of the following: model application scenario, model segmentation information, model compression information or model training information.

[0669] In one embodiment, the model training information includes at least one of the following: a training data set, metadata corresponding to the training data set, a data preprocessing method, model features, model hyperparameters, or a model structure.

[0670] In one embodiment, the apparatus 1900 further includes:

[0671] The second sending unit is configured to send second indication information to the first entity, where the second indication information includes the first model identifier, the second model identifier, and the model information template.

[0672] In one embodiment, the apparatus 1900 further includes:

[0673] A third receiving unit is configured to receive a fourth message sent by the first entity, where the fourth message is used to request storage of the model, and the fourth message includes at least one of the following: a second model identifier, a model transmission mode, a model security certificate, a model address, or a model;

[0674] The third sending unit is configured to send a fifth message to the first entity in response to the fourth message, where the fifth message is used to indicate a model storage result.

[0675] In one embodiment, the apparatus 1900 further includes:

[0676] The fourth receiving unit is configured to receive a model.

[0677] In one embodiment, the apparatus 1900 further includes:

[0678] a fourth sending unit, configured to send a tenth message to the third entity;

[0679] a fifth receiving unit, configured to receive an eleventh message sent by the third entity in response to the tenth message;

[0680] or,

[0681] a fourth sending unit, configured to send a tenth message to a fourth entity;

[0682] a fifth receiving unit, configured to receive an eleventh message sent by the fourth entity in response to the tenth message;

[0683] The tenth message is used to request storage of the model, and the tenth message includes at least one of the following: a second model identifier, a model transmission method, a model security certificate or a model; the eleventh message is used to indicate a model storage result.

[0684] In one embodiment, the apparatus 1900 further includes:

[0685] The fifth sending unit is configured to send the model to the third entity or the fourth entity.

[0686] In one embodiment, the apparatus 1900 further includes:

[0687] a sixth receiving unit, configured to receive a sixth message sent by the first entity, where the sixth message includes a second model identifier and model update information;

[0688] The sixth sending unit is configured to send a seventh message to the first entity in response to the sixth message, where the seventh message is used to indicate a model update result.

[0689] In one embodiment, the apparatus 1900 further includes:

[0690] a seventh receiving unit, configured to receive an eighth message sent by the first entity, where the eighth message is used to request model verification, and the eighth message includes a second model identifier and a model;

[0691] The seventh sending unit is configured to send a ninth message to the first entity in response to the eighth message, where the ninth message is used to indicate a model verification result.

[0692] It should be noted here that the above-mentioned model registration device 1900 provided by the present disclosure can implement all the method steps implemented by the second entity in the above-mentioned method embodiment, and can achieve the same technical effect. The parts and beneficial effects of this embodiment that are the same as the method embodiment will not be repeated here.

[0693] FIG20 is a schematic diagram of the structure of a model registration device 2000 provided in an embodiment of the present disclosure. As shown in FIG20 , the device 2000 includes:

[0694] The first receiving unit 2010 is configured to receive a model storage request message, where the model storage request message includes at least one of the following: a second model identifier, a model transmission mode, a model security certificate, a model address, or a model, where the second model identifier is a network identifier assigned to the model by the second entity, and the model storage request message is the fourth message sent by the first entity, the tenth message sent by the second entity, or the twelfth message sent by the fourth entity;

[0695] The first sending unit 2020 is used to send a model storage response message, where the model storage response message is used to indicate the model storage result. The storage response message is the fifth message sent to the first entity, the eleventh message sent to the second entity, or the thirteenth message sent to the fourth entity.

[0696] In one embodiment, the apparatus 2000 further includes:

[0697] The second receiving unit is used to receive the model.

[0698] In one embodiment, the apparatus 2000 further includes:

[0699] a third receiving unit, configured to receive an eighth message sent by the first entity, where the eighth message is used to request model verification, and the eighth message includes a second model identifier and a model;

[0700] The second sending unit is configured to send a ninth message to the first entity in response to the eighth message, where the ninth message is used to indicate a model verification result.

[0701] It should be noted here that the above-mentioned model registration device 2000 provided by the present disclosure can implement all the method steps implemented by the third entity in the above-mentioned method embodiment, and can achieve the same technical effect. The parts and beneficial effects of this embodiment that are the same as the method embodiment will not be repeated here.

[0702] FIG21 is a schematic diagram of the structure of a model registration device 2100 provided in an embodiment of the present disclosure. As shown in FIG21 , the device 2100 includes:

[0703] A first receiving unit 2110 is configured to receive an eighth message sent by the first entity, where the eighth message is used to request verification of the model, and the eighth message includes a second model identifier and the model, where the second model identifier is a network identifier allocated by the second entity to the model;

[0704] The first sending unit 2120 is configured to send a ninth message to the first entity in response to the eighth message, where the ninth message is used to indicate a model verification result.

[0705] In one embodiment, the apparatus 2100 further includes:

[0706] a second receiving unit, configured to receive a tenth message sent by the second entity, where the tenth message is used to request a storage model and includes a second model identifier;

[0707] The second sending unit is configured to send an eleventh message to the second entity in response to the tenth message, where the eleventh message is used to indicate a model storage result.

[0708] In one embodiment, the apparatus 2100 further includes:

[0709] A third sending unit is configured to send a twelfth message to a third entity, where the twelfth message is used to request storage of the model, and the twelfth message includes at least one of the following: a second model identifier, a model transmission mode, a model security certificate, or a model;

[0710] The third receiving unit is used to receive a thirteenth message sent by the third entity, where the thirteenth message is used to indicate a model storage result.

[0711] It should be noted here that the above-mentioned model registration device 2100 provided by the present disclosure can implement all the method steps implemented by the fourth entity in the above-mentioned method embodiment, and can achieve the same technical effect. The parts and beneficial effects of this embodiment that are the same as the method embodiment will not be repeated here.

[0712] It should be noted that the division of units in the embodiments of the present disclosure is schematic and is merely a logical functional division. In actual implementation, other division methods may be used. Furthermore, the functional units in the various embodiments of the present disclosure may be integrated into a single processing unit, or each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0713] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a processor-readable storage medium. Based on this understanding, the technical solution of the present disclosure is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) or a processor to execute all or part of the steps of each method embodiment of the present disclosure. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, and other media that can store program code.

[0714] An embodiment of the present disclosure also provides a non-transitory readable storage medium, which stores a computer program. The computer program is used to enable a processor to execute all method steps of the first entity in the above method embodiment, or all method steps of the second entity, or all method steps of the third entity, or all method steps of the fourth entity.

[0715] Non-transitory readable storage media can be any available media or data storage devices that can be accessed by a computer, including but not limited to magnetic storage (such as floppy disks, hard disks, magnetic tapes, magneto-optical disks (MO)), optical storage (such as CDs, DVDs, BDs, HVDs, etc.), and semiconductor storage (such as ROMs, EPROMs, EEPROMs, non-volatile memories (NAND FLASH), solid-state drives (SSDs)), etc.

[0716] The embodiments of the present disclosure further provide a computer program product, including a computer program, which implements all or part of the steps of the above method embodiments when executed by a processor.

[0717] Those skilled in the art will appreciate that embodiments of the present disclosure may be provided as methods, systems, or computer program products. Therefore, the present disclosure may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the present disclosure may take the form of a computer program product implemented on at least one computer-usable storage medium (including but not limited to magnetic disk storage and optical storage, etc.) containing computer-usable program code.

[0718] The present disclosure is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present disclosure. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer-executable instructions. These computer-executable instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device produce a device for implementing the functions specified in one or more processes in the flowchart and / or one or more boxes in the block diagram.

[0719] These processor-executable instructions may also be stored in a processor-readable memory that can direct a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the processor-readable memory produce a product including an instruction device that implements the functions specified in one or more processes in the flowchart and / or one or more boxes in the block diagram.

[0720] These processor-executable instructions may also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, whereby the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one or more processes in the flowchart and / or one or more boxes in the block diagram.

[0721] Obviously, those skilled in the art may make various changes and modifications to the present disclosure without departing from the spirit and scope of the present disclosure. Thus, if these modifications and variations of the present disclosure fall within the scope of the claims of the present disclosure and their equivalents, the present disclosure is intended to include these modifications and variations.

Claims

1. A model registration method, wherein, Applied to a first entity, the method includes: Sending a first message to a second entity, the first message being used to request model registration, the first message including a first model identifier and / or model information, the first model identifier being the identifier of the model in the first entity; Receiving a second message sent by the second entity in response to the first message, the second message including a second model identifier and first indication information, the second model identifier being the network identifier assigned by the second entity to the model, the first indication information indicating a registration decision or registration decision-related information; Based on the registration decision-related information, sending a third message to the second entity, the third message including the second model identifier and a registration decision.

2. The method according to claim 1, wherein, The model is a common model.

3. The method according to claim 1 or 2, wherein The model information includes at least one of the following: model type, model modality, model algorithm, model accuracy, model security, model privacy, model description information, model deployment requirements, model inference performance, model open range, model version, model owner information, or an automated deployment script.

4. The method according to claim 3, wherein, The model description information includes at least one of the following: model accuracy rate, variance of the model accuracy rate, data input format, data output format, model size, full model identifier corresponding to the compressed model, number of model parameters, model recall rate, or model F1 value; The model deployment requirements include at least one of the following: model dependencies, minimum resource requirements, or recommended resource configurations, where the resources include computing power resources and storage resources.

5. The method according to claim 3, wherein The model information further includes at least one of the following: model application scenario, model segmentation information, model compression information, or model training information.

6. The method according to claim 5, wherein, The model training information includes at least one of the following: training data set, metadata corresponding to the training data set, data preprocessing method, model features, model hyperparameters, or model structure.

7. The method according to any one of claims 1-6, wherein, The method further includes: Receiving second indication information sent by the second entity, the second indication information including the first model identifier, the second model identifier, and a model information template.

8. The method according to any one of claims 1-7, wherein, The method further includes: Sending a fourth message to the second entity; Receiving a fifth message sent by the second entity in response to the fourth message; Or, Sending a fourth message to a third entity; Receiving a fifth message sent by the third entity in response to the fourth message; Wherein, the fourth message is used to request storing the model, the fourth message including at least one of the following: the second model identifier, model transmission method, model security certificate, model address, or the model; the fifth message is used to indicate a model storage result.

9. The method according to claim 8, wherein The method further includes: Sending the model to the second entity or the third entity.

10. The method according to any one of claims 1-9, wherein, The method further includes: Sending a sixth message to the second entity, the sixth message including the second model identifier and model update information; Receiving a seventh message sent by the second entity in response to the sixth message, the seventh message being used to indicate a model update result.

11. According to the method according to any one of claims 1-10, wherein, The method further includes: Sending an eighth message to the second entity; Receiving a ninth message sent by the second entity in response to the eighth message; Or, Sending an eighth message to a third entity; Receive a ninth message sent by the third entity in response to the eighth message; Or, Send an eighth message to a fourth entity; Receive a ninth message sent by the fourth entity in response to the eighth message; Wherein, the eighth message is used to request verification of the model, and the eighth message includes the second model identifier and the model; the ninth message is used to indicate the model verification result.

12. The method according to claim 11, wherein, The first entity is a first terminal, a first access network device, or a first core network device; The second entity is deployed in a second terminal, a second access network device, or a second core network device; The third entity is deployed in a third terminal, a third access network device, or a third core network device; The fourth entity is deployed in a fourth access network device or a fourth core network device; The first terminal, the second terminal, and the third terminal are the same or different from each other, the first access network device, the second access network device, the third access network device, and the fourth access network device are the same or different from each other, and the first core network device, the second core network device, the third core network device, and the fourth core network device are the same or different from each other.

13. A model registration method, wherein, Applied to the second entity, the method includes: Receive a first message sent by the first entity, where the first message is used to request model registration, and the first message includes a first model identifier and / or model information, and the first model identifier is the identifier of the model in the first entity; In response to the first message, send a second message to the first entity, where the second message includes a second model identifier and first indication information, the second model identifier is the network identifier assigned by the second entity to the model, and the first indication information indicates a registration decision or registration decision-related information; Receive a third message sent by the first entity, where the third message is a message determined based on the registration decision-related information, and the third message includes the second model identifier and the registration decision.

14. The method according to claim 13, wherein, The model is a common model.

15. The method according to claim 13 or 14, wherein, The model information includes at least one of the following: model type, model modality, model algorithm, model accuracy, model security, model privacy, model description information, model deployment requirements, model inference performance, model open range, model version, model owner information, or automated deployment script.

16. The method according to claim 15, wherein, The model description information includes at least one of the following: model accuracy, variance of model accuracy, data input format, data output format, model size, full model identifier corresponding to the compressed model, number of model parameters, model recall rate, or model F1 value; The model deployment requirements include at least one of the following: model dependencies, minimum resource requirements, or recommended resource configurations, and the resources include computing power resources and storage resources.

17. The method according to claim 15, wherein The first model information further includes at least one of the following: model application scenario, model segmentation information, model compression information, or model training information.

18. The method according to claim 17, wherein, The model training information includes at least one of the following: training data set, metadata corresponding to the training data set, data preprocessing method, model features, model hyperparameters, or model structure.

19. The method according to any one of claims 13 - 18, wherein, The method further includes: Send second indication information to the first entity, where the second indication information includes the first model identifier, the second model identifier, and a model information template.

20. The method according to any one of claims 13-19, wherein, The method further includes: Receive a fourth message sent by the first entity, where the fourth message is used to request storing the model, and the fourth message includes at least one of the following: the second model identifier, a model transmission method, a model security certificate, a model address, or the model; In response to the fourth message, send a fifth message to the first entity, where the fifth message is used to indicate a model storage result.

21. The method according to claim 20, wherein The method further includes: Receive the model.

22. The method according to any one of claims 13-19, wherein, The method further includes: Send a tenth message to a third entity; Receive an eleventh message sent by the third entity in response to the tenth message; Or, Send a tenth message to a fourth entity; Receive an eleventh message sent by the fourth entity in response to the tenth message; The tenth message is used to request storing the model, and the tenth message includes at least one of the following: the second model identifier, a model transmission method, a model security certificate, or the model; the eleventh message is used to indicate a model storage result.

23. The method according to claim 22, wherein, The method further includes: Send the model to the third entity or the fourth entity.

24. The method according to any one of claims 13-23, wherein, The method further includes: Receive a sixth message sent by the first entity, where the sixth message includes the second model identifier and model update information; In response to the sixth message, send a seventh message to the first entity, where the seventh message is used to indicate a model update result.

25. The method according to any one of claims 13-24, wherein, The method further includes: Receive an eighth message sent by the first entity, where the eighth message is used to request verifying the model, and the eighth message includes the second model identifier and the model; In response to the eighth message, send a ninth message to the first entity, where the ninth message is used to indicate a model verification result.

26. A model registration method, wherein, Applied to a third entity, the method includes: Receive a model storage request message, where the model storage request message includes at least one of the following: a second model identifier, a model transmission method, a model security certificate, a model address, or a model, the second model identifier is a network identifier assigned by a second entity to the model, and the model storage request message is a fourth message sent by the first entity or a tenth message sent by the second entity or a twelfth message sent by the fourth entity; Send a model storage response message, where the model storage response message is used to indicate a model storage result, and the storage response message is a fifth message sent to the first entity or an eleventh message sent to the second entity or a thirteenth message sent to the fourth entity.

27. The method according to claim 26, wherein, The method further includes: Receive the model.

28. The method according to claim 26 or 27, wherein The method further includes: Receive an eighth message sent by the first entity, where the eighth message is used to request verifying the model, and the eighth message includes the second model identifier and the model; In response to the eighth message, send a ninth message to the first entity, where the ninth message is used to indicate a model verification result.

29. A model registration method, wherein, Applied to a fourth entity, the method includes: Receive an eighth message sent by a first entity, where the eighth message is used to request verification of a model, and the eighth message includes a second model identifier and the model, and the second model identifier is a network identifier assigned by a second entity to the model; In response to the eighth message, send a ninth message to the first entity, where the ninth message is used to indicate the model verification result.

30. The method according to claim 29, wherein, The method further includes: Receive a tenth message sent by the second entity, where the tenth message is used to request storage of the model, and the tenth message includes the second model identifier; In response to the tenth message, send an eleventh message to the second entity, where the eleventh message is used to indicate the model storage result.

31. The method according to claim 29 or 30, wherein, The method further includes: Send a twelfth message to a third entity, where the twelfth message is used to request storage of the model, and the twelfth message includes at least one of the following: the second model identifier, the model transmission method, the model security certificate, or the model; Receive a thirteenth message sent by the third entity, where the thirteenth message is used to indicate the model storage result.

32. A model registration device, wherein, Applied to a first entity, the apparatus includes: A first sending unit, configured to send a first message to a second entity, where the first message is used to request model registration, and the first message includes a first model identifier and / or model information, and the first model identifier is an identifier of the model in the first entity; A receiving unit, configured to receive a second message sent by the second entity in response to the first message, where the second message includes a second model identifier and first indication information, the second model identifier is a network identifier assigned by the second entity to the model, and the first indication information indicates a registration decision or registration decision-related information; A second sending unit, configured to send a third message to the second entity based on the registration decision-related information, where the third message includes the second model identifier and the registration decision.

33. A model registration device, wherein, Applied to a second entity, the apparatus includes: A first receiving unit, configured to receive a first message sent by a first entity, where the first message is used to request model registration, and the first message includes a first model identifier and / or model information, and the first model identifier is an identifier of the model in the first entity; A sending unit, configured to send a second message to the first entity in response to the first message, where the second message includes a second model identifier and first indication information, the second model identifier is a network identifier assigned by the second entity to the model, and the first indication information indicates a registration decision or registration decision-related information; A second receiving unit, configured to receive a third message sent by the first entity, where the third message is a message determined based on the registration decision-related information, and the third message includes the second model identifier and the registration decision.

34. A model registration device, wherein, Applied to a third entity, the apparatus includes: A receiving unit, configured to receive a model storage request message, where the model storage request message includes at least one of the following: a second model identifier, a model transmission method, a model security certificate, a model address, or a model. The second model identifier is a network identifier assigned by a second entity to the model, and the model storage request message is a fourth message sent by a first entity, a tenth message sent by the second entity, or a twelfth message sent by a fourth entity. A sending unit, configured to send a model storage response message, where the model storage response message is used to indicate a model storage result, and the storage response message is a fifth message sent to the first entity, an eleventh message sent to the second entity, or a thirteenth message sent to the fourth entity.

35. A model registration device, wherein, Applied to a fourth entity, the apparatus includes: A receiving unit, configured to receive an eighth message sent by a first entity, where the eighth message is used to request model verification, and the eighth message includes a second model identifier and the model. The second model identifier is a network identifier assigned by a second entity to the model. A sending unit, configured to send a ninth message to the first entity in response to the eighth message, where the ninth message is used to indicate a model verification result.

36. A model registration device, wherein, Applied to a first entity, the apparatus includes a memory, a transceiver, and a processor: The memory is configured to store a computer program; the transceiver is configured to send and receive data under the control of the processor; the processor is configured to read the computer program in the memory and perform the following operations: Send a first message to a second entity, where the first message is used to request model registration, and the first message includes a first model identifier and / or model information. The first model identifier is an identifier of the model in the first entity. Receive a second message sent by the second entity in response to the first message, where the second message includes a second model identifier and a first indication information. The second model identifier is a network identifier assigned by the second entity to the model, and the first indication information indicates a registration decision or registration decision-related information. Based on the registration decision-related information, send a third message to the second entity, where the third message includes the second model identifier and a registration decision.

37. The apparatus according to claim 36, wherein, The model is a public model.

38. The apparatus according to claim 36 or 37, wherein, The model information includes at least one of the following: model type, model modality, model algorithm, model accuracy, model security, model privacy, model description information, model deployment requirements, model inference performance, model open range, model version, model owner information, or an automated deployment script.

39. The apparatus according to claim 38, wherein, The model description information includes at least one of the following: model accuracy, variance of model accuracy, data input format, data output format, model size, full model identifier corresponding to a compressed model, number of model parameters, model recall rate, or model F1 value. The model deployment requirements include at least one of the following: model dependencies, minimum resource requirements, or recommended resource configurations. The resources include computing power resources and storage resources.

40. The apparatus according to claim 38, wherein, The model information further includes at least one of the following: model application scenario, model segmentation information, model compression information, or model training information.

41. The apparatus according to claim 40, wherein, The model training information includes at least one of the following: training data set, metadata corresponding to the training data set, data preprocessing method, model features, model hyperparameters, or model structure.

42. The apparatus according to any one of claims 36 - 41, wherein, The processor is further configured to perform the following operations: Receive second indication information sent by the second entity, where the second indication information includes the first model identifier, the second model identifier, and a model information template.

43. The apparatus according to any one of claims 36 - 42, wherein, The processor is further configured to perform the following operations: Send a fourth message to the second entity; Receive a fifth message sent by the second entity in response to the fourth message; Or, Send a fourth message to a third entity; Receive a fifth message sent by the third entity in response to the fourth message; Wherein, the fourth message is used to request storing the model, and the fourth message includes at least one of the following: the second model identifier, model transmission method, model security certificate, model address, or the model; the fifth message is used to indicate the model storage result.

44. The apparatus according to claim 43, wherein, The processor is further configured to perform the following operations: Send the model to the second entity or the third entity.

45. The apparatus according to any one of claims 36 - 44, wherein, The processor is further configured to perform the following operations: Send a sixth message to the second entity, where the sixth message includes the second model identifier and model update information; Receive a seventh message sent by the second entity in response to the sixth message, where the seventh message is used to indicate the model update result.

46. The device according to any one of claims 36 - 45, wherein, The processor is further configured to perform the following operations: Send an eighth message to the second entity; Receive a ninth message sent by the second entity in response to the eighth message; Or, Send an eighth message to a third entity; Receive a ninth message sent by the third entity in response to the eighth message; Or, Send an eighth message to a fourth entity; Receive a ninth message sent by the fourth entity in response to the eighth message; Wherein, the eighth message is used to request verifying the model, and the eighth message includes the second model identifier and the model; the ninth message is used to indicate the model verification result.

47. A model registration device, wherein, Applied to a second entity, the device includes a memory, a transceiver, and a processor: The memory is used to store a computer program; the transceiver is used to transmit and receive data under the control of the processor; the processor is used to read the computer program in the memory and perform the following operations: Receive a first message sent by a first entity, where the first message is used to request registering a model, and the first message includes a first model identifier and / or model information, and the first model identifier is the identifier of the model in the first entity; Send a second message to the first entity in response to the first message, where the second message includes a second model identifier and first indication information, the second model identifier is the network identifier assigned by the second entity to the model, and the first indication information indicates a registration decision or registration decision-related information; Receive a third message sent by the first entity, where the third message is a message determined based on the registration decision-related information, and the third message includes the second model identifier and the registration decision.

48. The apparatus according to claim 47, wherein, The model is a public model.

49. The device according to claim 47 or 48, wherein The model information includes at least one of the following: model type, model modality, model algorithm, model accuracy, model security, model privacy, model description information, model deployment requirements, model inference performance, model open range, model version, model owner information, or automated deployment script.

50. The apparatus according to claim 49, wherein, The model description information includes at least one of the following: model accuracy, variance of model accuracy, data input format, data output format, model size, full model identifier corresponding to the compressed model, number of model parameters, model recall rate, or model F1 value; The model deployment requirements include at least one of the following: model dependencies, minimum resource requirements, or recommended resource configuration, where the resources include computing power resources and storage resources.

51. The apparatus according to claim 49, wherein, The model information further includes at least one of the following: model application scenario, model segmentation information, model compression information, or model training information.

52. The apparatus according to claim 51, wherein, The model training information includes at least one of the following: training data set, metadata corresponding to the training data set, data preprocessing method, model features, model hyperparameters, or model structure.

53. The apparatus according to any one of claims 47 - 52, wherein, The processor is further configured to perform the following operations: Send second indication information to the first entity, where the second indication information includes the first model identifier, the second model identifier, and a model information template.

54. The apparatus according to any one of claims 47 - 53, wherein, The processor is further configured to perform the following operations: Receive a fourth message sent by the first entity, where the fourth message is used to request storing the model, and the fourth message includes at least one of the following: the second model identifier, model transmission method, model security certificate, model address, or the model; Send a fifth message in response to the fourth message to the first entity, where the fifth message is used to indicate the model storage result.

55. The apparatus according to claim 54, wherein, The processor is further configured to perform the following operations: Receive the model.

56. The apparatus according to any one of claims 47 - 53, wherein, The processor is further configured to perform the following operations: Send a tenth message to a third entity; Receive an eleventh message sent by the third entity in response to the tenth message; Or, Send a tenth message to a fourth entity; Receive an eleventh message sent by the fourth entity in response to the tenth message; The tenth message is used to request storing the model, and the tenth message includes at least one of the following: the second model identifier, model transmission method, model security certificate, or the model; the eleventh message is used to indicate the model storage result.

57. The apparatus according to claim 56, wherein, The processor is further configured to perform the following operations: Send the model to the third entity or the fourth entity.

58. The apparatus according to any one of claims 47 - 57, wherein, The processor is further configured to perform the following operations: Receive a sixth message sent by the first entity, where the sixth message includes the second model identifier and model update information; Send a seventh message in response to the sixth message to the first entity, where the seventh message is used to indicate the model update result.

59. The device according to any one of claims 47 - 58, wherein, The processor is further configured to perform the following operations: Receive an eighth message sent by the first entity, where the eighth message is used to request verifying the model, and the eighth message includes the second model identifier and the model; Send a ninth message in response to the eighth message to the first entity, where the ninth message is used to indicate the model verification result.

60. A model registration device, wherein, Applied to a third entity, the device includes a memory, a transceiver, and a processor: The memory is used to store computer programs; the transceiver is used to transmit and receive data under the control of the processor; the processor is used to read the computer programs in the memory and perform the following operations: Receive a model storage request message, where the model storage request message includes at least one of the following: a second model identifier, a model transmission method, a model security certificate, a model address, or a model. The second model identifier is a network identifier assigned by a second entity to the model. The model storage request message is a fourth message sent by a first entity, a tenth message sent by the second entity, or a twelfth message sent by a fourth entity; Send a model storage response message, where the model storage response message is used to indicate the model storage result. The storage response message is a fifth message sent to the first entity, an eleventh message sent to the second entity, or a thirteenth message sent to the fourth entity.

61. The apparatus according to claim 60, wherein: The processor is further used to perform the following operations: Receive the model.

62. The device according to claim 60 or 61, wherein, The processor is further used to perform the following operations: Receive an eighth message sent by a first entity, where the eighth message is used to request verification of the model, and the eighth message includes the second model identifier and the model; In response to the eighth message, send a ninth message to the first entity, where the ninth message is used to indicate the model verification result.

63. A model registration device, wherein, Applied to a fourth entity, the device includes a memory, a transceiver, and a processor: The memory is used to store computer programs; the transceiver is used to transmit and receive data under the control of the processor; the processor is used to read the computer programs in the memory and perform the following operations: Receive an eighth message sent by a first entity, where the eighth message is used to request verification of a model, and the eighth message includes a second model identifier and the model. The second model identifier is a network identifier assigned by a second entity to the model; In response to the eighth message, send a ninth message to the first entity, where the ninth message is used to indicate the model verification result.

64. The apparatus according to claim 63, wherein, The processor is further used to perform the following operations: Receive a tenth message sent by the second entity, where the tenth message is used to request storage of the model, and the tenth message includes the second model identifier; In response to the tenth message, send an eleventh message to the second entity, where the eleventh message is used to indicate the model storage result.

65. The apparatus according to claim 63 or 64, wherein, The processor is further used to perform the following operations: Send a twelfth message to the third entity, where the twelfth message is used to request storage of the model, and the twelfth message includes at least one of the following: the second model identifier, a model transmission method, a model security certificate, or the model; Receive a thirteenth message sent by the third entity, where the thirteenth message is used to indicate the model storage result.

66. A non-transitory readable storage medium, wherein, The non-transitory readable storage medium stores a computer program, and the computer program is used to cause a processor to execute the method described in any one of claims 1-11, or the method described in any one of claims 13-25, or the method described in any one of claims 26-28, or the method described in any one of claims 29-31.

Images