Created by: mamingjie-China

PaddlePaddle 1.8.1，python3.7 topk中，当输入的k为tensor时，无论tensor的值是多少，结果都是相同 示例代码： import paddle.fluid as fluid import paddle.fluid.layers as layers input2 = layers.data(name="input2", shape=[13, 11], dtype='float32') vk = fluid.layers.fill_constant(shape=[1], value=6, dtype='int64') vk_values, vk_indices = layers.topk(input2, k=vk) print(vk_values, vk_indices)

期望的结果的shape应该是[13, 6]的，但实际运行结果为[13, 1]