oneflow.nn.functional.one_hot¶
-
oneflow.nn.functional.
one_hot
(input, num_classes=- 1, on_value=1, off_value=0)¶ This operator generates a onehot Tensor from input Tensor.
If input Tensor’s rank is N, the corresponding onehot Tensor’s rank is N+1.
- Parameters
input (Tensor) – The input Tensor.
num_classes (int) – The length of onehot Tensor.
on_value (Union[int, float], optional) – The fill value when x[i] == i. Defaults to 1.
off_value (Union[int, float], optional) – The fill value when x[i] != i. Defaults to 0.
Note
The data type of input tensor should be int32 or int64.
- Returns
oneflow.Tensor.
For example:
>>> import oneflow as flow >>> import numpy as np >>> input=flow.tensor(np.array([0, 3, 1, 2]).astype(np.int64), dtype=flow.int64) >>> out = flow.nn.functional.one_hot(input, num_classes=5) >>> out tensor([[1, 0, 0, 0, 0], [0, 0, 0, 1, 0], [0, 1, 0, 0, 0], [0, 0, 1, 0, 0]], dtype=oneflow.int64)