tf.keras.layers.Masking
Stay organized with collections
Save and categorize content based on your preferences.
Masks a sequence by using a mask value to skip timesteps.
Inherits From: Layer
tf.keras.layers.Masking(
mask_value=0.0, **kwargs
)
For each timestep in the input tensor (dimension #1 in the tensor),
if all values in the input tensor at that timestep
are equal to mask_value, then the timestep will be masked (skipped)
in all downstream layers (as long as they support masking).
If any downstream layer does not support masking yet receives such
an input mask, an exception will be raised.
Example:
Consider a Numpy data array x of shape (samples, timesteps, features),
to be fed to an LSTM layer.
You want to mask timestep #3 and #5 because you lack data for
these timesteps. You can:
- Set
x[:, 3, :] = 0. and x[:, 5, :] = 0.
- Insert a
Masking layer with mask_value=0. before the LSTM layer:
model = Sequential()
model.add(Masking(mask_value=0., input_shape=(timesteps, features)))
model.add(LSTM(32))
Except as otherwise noted, the content of this page is licensed under the Creative Commons Attribution 4.0 License, and code samples are licensed under the Apache 2.0 License. For details, see the Google Developers Site Policies. Java is a registered trademark of Oracle and/or its affiliates.
Last updated 2020-10-01 UTC.
[null,null,["Last updated 2020-10-01 UTC."],[],[],null,["# tf.keras.layers.Masking\n\n\u003cbr /\u003e\n\n|------------------------------------------------------------------|--------------------------------------------------------------------------------------------------------------------------------|\n| [TensorFlow 2 version](/api_docs/python/tf/keras/layers/Masking) | [View source on GitHub](https://github.com/tensorflow/tensorflow/blob/v1.15.0/tensorflow/python/keras/layers/core.py#L55-L106) |\n\nMasks a sequence by using a mask value to skip timesteps.\n\nInherits From: [`Layer`](../../../tf/keras/layers/Layer)\n\n#### View aliases\n\n\n**Compat aliases for migration**\n\nSee\n[Migration guide](https://www.tensorflow.org/guide/migrate) for\nmore details.\n\n[`tf.compat.v1.keras.layers.Masking`](/api_docs/python/tf/keras/layers/Masking), \\`tf.compat.v2.keras.layers.Masking\\`\n\n\u003cbr /\u003e\n\n tf.keras.layers.Masking(\n mask_value=0.0, **kwargs\n )\n\nFor each timestep in the input tensor (dimension #1 in the tensor),\nif all values in the input tensor at that timestep\nare equal to `mask_value`, then the timestep will be masked (skipped)\nin all downstream layers (as long as they support masking).\n\nIf any downstream layer does not support masking yet receives such\nan input mask, an exception will be raised.\n\n#### Example:\n\nConsider a Numpy data array `x` of shape `(samples, timesteps, features)`,\nto be fed to an LSTM layer.\nYou want to mask timestep #3 and #5 because you lack data for\nthese timesteps. You can:\n\n- Set `x[:, 3, :] = 0.` and `x[:, 5, :] = 0.`\n- Insert a `Masking` layer with `mask_value=0.` before the LSTM layer:\n\n model = Sequential()\n model.add(Masking(mask_value=0., input_shape=(timesteps, features)))\n model.add(LSTM(32))"]]
TensorFlow 2 version
View source on GitHub