1. Tensor Info
| Entry | Memo |
|---|---|
shape |
张量形状，有些操作需特定的形状规则 |
dtype |
张量元素以何种数据类型存储 |
device |
张量存储在何种设备，通常为CPU或GPU |
tensor_info = torch.rand(3,4)
print(tensor_info)
print(f"Tensor Shape: {tensor_info.shape}")
print(f"Tensor Datatype: {tensor_info.dtype}")
print(f"Tensor Device: {tensor_info.device}")tensor([[0.6740, 0.2000, 0.3854, 0.1561],
[0.6748, 0.1559, 0.0782, 0.6290],
[0.2350, 0.4629, 0.6631, 0.8041]])
Tensor Shape: torch.Size([3, 4])
Tensor Datatype: torch.float32
Tensor Device: cpu(默认CPU)
|