1. 从numpy中导入数据
a = np.array([2, 5.5])
print(a)
b = torch.from_numpy(a)
print(b)
结果:
[2. 5.5]
tensor([2.0000, 5.5000], dtype=torch.float64)
2. 利用tensor直接创建
b = torch.tensor([[2, 3, 4], [4, 5, 6]])
print(b)
tensor([[2, 3, 4],
[4, 5, 6]])
3 均匀分布
a = torch.rand(3, 3) # [0,1]之间均匀分布
print(a)
b = torch.rand_like(a) # 创建一个与a shape相同的数组
print(b)
c = torch.randint(1, 10, (3, 3)) # [0, 10)之间,3*3形状的整数
print(c)
tensor([[0.0789, 0.7449, 0.9799],
[0.7051, 0.3074, 0.7543],
[0.8140, 0.4832, 0.3603]])
tensor([[0.7311, 0.5235, 0.7582],
[0.6360, 0.4095, 0.7638],
[0.5338, 0.1844, 0.5473]])
tensor([[5, 9, 9],
[4, 1, 1],
[9, 5, 5]])
4