conv2d (inputs,num_outputs,kernel_size,scope) - Adds an N-D convolution followed by an optional batch_norm layer. It is required that 1 <= N <= 3.
確定！ 回上一頁
