In order to apply a powerful neural network to the continuous-variable quantum information processing, it is necessary to construct the continuous-variable quantum neural network (QNN) model. Coherent state quantum logic gates are taken as basic elements. Quantum circuit composed of input layer, hidden layer and output layer is constructed based on QNN principle, and the function of continuous-variable coherent state quantum neural network (CSQNN) is realized. The model realizes quantum state operation by using multi-bit CNOT gate, and the learning training of network parameters is completed by using phase rotation gates. Simulation results show that under the assistance of CSQNN, the quantum teleportation fidelity of amplitude damping channel with damping coefficient of 0.5 is significantly improved, and its value approaches 1. It’s shown that the proposed CSQNN model can effectively deal with the continuous-variable quantum information.