要按行连接3D的numpy数组，可以使用numpy的reshape和transpose函数来实现。下面是一个示例代码：

import numpy as np

# 创建两个3D的numpy数组
arr1 = np.array([
    [[1, 2, 3],
     [4, 5, 6]],
    
    [[7, 8, 9],
     [10, 11, 12]]
])

arr2 = np.array([
    [[13, 14, 15],
     [16, 17, 18]],
    
    [[19, 20, 21],
     [22, 23, 24]]
])

# 将数组reshape为2D，其中每行表示一个3D数组
reshaped_arr1 = arr1.reshape(arr1.shape[0], -1)
reshaped_arr2 = arr2.reshape(arr2.shape[0], -1)

# 按行连接两个2D数组
result = np.concatenate((reshaped_arr1, reshaped_arr2), axis=0)

# 将结果转换回3D数组
final_result = result.reshape(arr1.shape[0] + arr2.shape[0], arr1.shape[1], arr1.shape[2])

print(final_result)

运行以上代码，输出结果为：

[[[ 1  2  3]
  [ 4  5  6]]

 [[ 7  8  9]
  [10 11 12]]

 [[13 14 15]
  [16 17 18]]

 [[19 20 21]
  [22 23 24]]]

这样就按行连接了两个3D的numpy数组。