torch.topk()与tensor.topk()

torch.topk(tensor, k, dim=None, largest=True)

返回tensor指定维度的前k个元素值及坐标

• k 指定返回前k个元素及坐标
• dim默认指定最后一个维度
• largest默认指为True，表示从大到小排序，指定为False时，表示从小到大排序

tensor.topk(k, dim=None, largest=True)

等价于torck.topk(tensor, k, dim=None, largest=True)

示例

import torch

torch.manual_seed(0)  # 确保每次生成相同的随机数
tensor_ = torch.rand(20).reshape(4, 5)
v, idx = torch.topk(tensor_, 2, dim=1)
# 等价： v, idx = tensor_.topk(2, dim=1)
print(tensor_)
print(v)
print(idx)
'''
tensor([[0.4963, 0.7682, 0.0885, 0.1320, 0.3074],
        [0.6341, 0.4901, 0.8964, 0.4556, 0.6323],
        [0.3489, 0.4017, 0.0223, 0.1689, 0.2939],
        [0.5185, 0.6977, 0.8000, 0.1610, 0.2823]])     
tensor([[0.7682, 0.4963],
        [0.8964, 0.6341],
        [0.4017, 0.3489],
        [0.8000, 0.6977]])
tensor([[1, 0],
        [2, 0],
        [1, 0],
        [2, 1]])
'''