dropout
Applicability
Product |
Supported |
|---|---|
√ |
|
√ |
|
☓ |
|
☓ |
|
√ |
Description
It has the same functionality as tf.nn.dropout. Elements of the input tensor are randomly set to zero with a probability of 1 – keep_prob. The remaining elements are scaled by a factor of 1/keep_prob to ensure that the output tensor maintains the same shape as the input tensor.
Prototype
1 | def dropout(x, keep_prob, noise_shape=None, seed=None, name=None) |
Parameters
Parameter |
Input/Output |
Description |
|---|---|---|
x |
Input |
Input tensor of type float. |
keep_prob |
Input |
Scalar tensor of type float, which indicates the retention probability of each element. |
noise_shape |
Input |
1D tensor of type int32, which indicates the shape of the randomly generated keep_drop flag. |
seed |
Input |
Random seed. |
name |
Input |
Name of the network layer. |
Returns
Result tensor after the dropout operation is performed on input x.
Example
1 2 | from npu_bridge.npu_init import * layers = npu_ops.dropout() |