Ray double-layer scheduling method and device based on reinforcement learning and electronic equipment

The invention provides a Ray double-layer scheduling method and device based on reinforcement learning and electronic device.The Ray double-layer scheduling method based on reinforcement learning comprises the steps that a cluster task queue, resource node cluster information and resource node cluster task queue information are obtained, and a target decision action is determined based on a preset Ray double-layer scheduling model; wherein the preset Ray double-layer scheduling model comprises the step of determining the target decision action after reinforcement learning based on the resource node cluster information and the resource node cluster task queue information; and scheduling a to-be-scheduled task in a cluster task queue to the correspondingly allocated resource node based on the target decision action. By using the method provided by the invention, the purpose of determining the target decision-making action through autonomous learning is realized, so that the determined target decision-making act