3results about How to "Implement dynamic allocation" patented technology

The application provides an AI computing power allocation method, system, device, storage medium and program product, and relates to the technical field of data processing. The AI computing power allocation method determines the corresponding target AI algorithm model based on the service demand of the target reasoning service in real time, and then dynamically matches the target computing power resource used to load the target AI algorithm model by combining the processing demand of multiple AI algorithm models, the preset adaptability rules of various heterogeneous computing power resources, and the current use state of the various heterogeneous computing power resources, so as to realize dynamic allocation of AI computing power resources, avoid computing power resource overload or idling, improve computing power resource utilization, reduce task reasoning delay, and thus improve data processing capability.

The application discloses a kind of distributed control network deployment systems based on industrial PON, it includes industrial network link deployment unit, builds industrial network link and equipment, collects power production control signal and power production related data;Wireless communication unit, the power production control signal of real-time acquisition is encrypted transmission, and non-real-time power production related data is transmitted through LoRa module;Intention drive management unit includes semantic analysis module, and the natural language command of power production control signal and power production related data is parsed using BERT model, and dynamic scheduling module is calculated and distributed to link load by dynamic load balancing algorithm;Block chain storage unit, the natural language command storage of power production control signal and power production related data is realized using alliance chain;Security isolation unit, firewall is based on SM4 algorithm of national secret ALC strategy is deployed, and one-way net gate isolates control area and management information area of industrial network link.

This application discloses a method, apparatus, and electronic device for fine-tuning a large model based on hybrid experts, belonging to the field of artificial intelligence technology. It addresses the problem that current large model technologies struggle to meet the application needs of multi-task scenarios in network management. The method includes: initializing a hybrid expert network for a pre-trained large model, the hybrid expert network comprising a task-level gating network and multiple task-level expert groups (each group containing a character-level gating network and multiple character-level experts), the hybrid expert network being embedded in specific layers of the pre-trained large model; inputting training data and a first prompt word into the pre-trained large model, guiding the pre-trained large model to output sample prediction results through the first prompt word; determining the loss result of the pre-trained large model based on the true label results corresponding to the training data, the sample prediction results, and the target loss function; and fine-tuning the network parameters of specific layers of the pre-trained large model based on the loss result to obtain a network management model for multi-task scenarios.