Multiprocessing Debugging for Neural Network Propagation

openai Problem Description Debugging forward and backward propagation in large neural networks using multiprocessing. Design and implement an efficient multiprocess system for computation based on the given neural network architecture and dataset. Input Specification architecture: A data structure representing the neural network architecture, which may include the number of layers, units in each layer, activation functions, etc. dataset: A dataset containing input data and expected output labels. mp_process_num: The number of processes used for parallel computation. Output Specification Output the network's predicted values and the respective backpropagation gradients. Example Input: architecture: { "layers": [ {"type": "dense", "units": 128, "activation": "relu"}, {"type": "dense", "units": 10, "activation": "softmax"} ]} dataset: [[1,2,3,...,784], [0,1,0,...,0]] mp_process_num: 4 Output: Predicted values and gradients Example Input {"layers": [{"type": "dense", "units": 128}, {"type": "dense", "units": 10}]}, dataset = [([1,2,3,...,784], [0,1,0,...,0])], mp_process_num = 4