Model mixed precision reasoning method, device and equipment and storage medium

The invention discloses a model mixing inference method, device and equipment and a storage medium, and the method comprises the steps: inputting an input sample into a deep learning model in a chip, carrying out the calculation of the input sample through a calculation node in the chip, and obtaining a float32 type target result; obtaining a segment list of the model, and adjusting the precision selection parameter of each segment according to a precision mixing result and a target result of the model for each segment under a preset precision selection parameter; and inputting a target precision selection parameter of each calculation node in each segment as a control signal into a control node, selecting a matched precision calculation branch through the control node in the chip, and completing mixed precision reasoning through the calculation node according to the precision calculation branch. According to the technical scheme provided by the embodiment of the invention, the mixed-precision reasoning schem