Allocation method and device for computing resources

Abstract

The invention is applicable to the technical field of computers, and provides an allocation method for computing resources, which comprises the following steps: in each round of training process of a global network model, acquiring bandwidth information of multiple user equipment, then generating a first computing frequency of each user equipment according to the bandwidth information of the multiple user equipment, and finally, sending the corresponding first calculation frequency and the local model parameter of the global network model to each user equipment, so that the user equipment updates the local model parameter according to the first calculation frequency. The first calculation frequency of each piece of user equipment is generated according to the bandwidth information of the plurality of pieces of user equipment, so that each piece of user equipment can update the local model parameters according to the corresponding first calculation frequency, and the calculation speed of the user equipment of which the original training task is completed too fast can be reduced; therefore, the problem of computing resource waste in the existing federal learning process is solved. 本申请适用于计算机技术领域，提供了一种计算资源的分配方法，包括：在对全局网络模型的每一轮训练过程中，获取多个用户设备的带宽信息，然后根据多个用户设备的带宽信息，生成每个用户设备的第一计算频率，最后向每个用户设备发送对应的第一计算频率和全局网络模型的局部模型参数，以使用户设备根据第一计算频率对局部模型参数进行更新。通过根据多个用户设备的带宽信息生成每个用户设备的第一计算频率，进而每个用户设备可以按照对应的第一计算频率对局部模型参数进行更新，可以使得原先训练任务完成过快的用户设备降低计算速度，从而解决现有联邦学习过程中计算资源浪费的问题。