Dropout#

Dropout - 13#

Version

  • name: Dropout (GitHub)

  • domain: main

  • since_version: 13

  • function: False

  • support_level: SupportType.COMMON

  • shape inference: True

This version of the operator has been available since version 13.

Summary

Dropout takes an input floating-point tensor, an optional input ratio (floating-point scalar) and an optional input training_mode (boolean scalar). It produces two tensor outputs, output (floating-point tensor) and mask (optional Tensor<bool>). If training_mode is true then the output Y will be a random dropout; Note that this Dropout scales the masked input data by the following equation, so to convert the trained model into inference mode, the user can simply not pass training_mode input or set it to false.

output = scale * data * mask,

where

scale = 1. / (1. - ratio).

This operator has optional inputs/outputs. See ONNX for more details about the representation of optional arguments. An empty string may be used in the place of an actual argument’s name to indicate a missing argument. Trailing optional arguments (those not followed by an argument that is present) may also be simply omitted.

Attributes

  • seed: (Optional) Seed to the random generator, if not specified we will auto generate one.

Inputs

Between 1 and 3 inputs.

  • data (heterogeneous) - T: The input data as Tensor.

  • ratio (optional, heterogeneous) - T1: The ratio of random dropout, with value in [0, 1). If this input was not set, or if it was set to 0, the output would be a simple copy of the input. If it’s non-zero, output will be a random dropout of the scaled input, which is typically the case during training. It is an optional value, if not specified it will default to 0.5.

  • training_mode (optional, heterogeneous) - T2: If set to true then it indicates dropout is being used for training. It is an optional value hence unless specified explicitly, it is false. If it is false, ratio is ignored and the operation mimics inference mode where nothing will be dropped from the input data and if mask is requested as output it will contain all ones.

Outputs

Between 1 and 2 outputs.

  • output (heterogeneous) - T: The output.

  • mask (optional, heterogeneous) - T2: The output mask.

Type Constraints

  • T in ( tensor(bfloat16), tensor(double), tensor(float), tensor(float16) ): Constrain input and output types to float tensors.

  • T1 in ( tensor(double), tensor(float), tensor(float16) ): Constrain input ‘ratio’ types to float tensors.

  • T2 in ( tensor(bool) ): Constrain output ‘mask’ types to boolean tensors.

Examples

_default

import numpy as np
import onnx

seed = np.int64(0)
node = onnx.helper.make_node("Dropout", inputs=["x"], outputs=["y"], seed=seed)

x = np.random.randn(3, 4, 5).astype(np.float32)
y = dropout(x)
expect(node, inputs=[x], outputs=[y], name="test_dropout_default")

_default_ratio

import numpy as np
import onnx

seed = np.int64(0)
node = onnx.helper.make_node(
    "Dropout", inputs=["x", "r"], outputs=["y"], seed=seed
)

r = np.float32(0.1)
x = np.random.randn(3, 4, 5).astype(np.float32)
y = dropout(x, r)
expect(node, inputs=[x, r], outputs=[y], name="test_dropout_default_ratio")

_default_mask

import numpy as np
import onnx

seed = np.int64(0)
node = onnx.helper.make_node(
    "Dropout", inputs=["x"], outputs=["y", "z"], seed=seed
)

x = np.random.randn(3, 4, 5).astype(np.float32)
y, z = dropout(x, return_mask=True)
expect(node, inputs=[x], outputs=[y, z], name="test_dropout_default_mask")

_default_mask_ratio

import numpy as np
import onnx

    seed = np.int64(0)
    node = onnx.helper.make_node(
        "Dropout", inputs=["x", "r"], outputs=["y", "z"], seed=seed
    )

    r = np.float32(0.1)
    x = np.random.randn(3, 4, 5).astype(np.float32)
    y, z = dropout(x, r, return_mask=True)
    expect(
        node, inputs=[x, r], outputs=[y, z], name="test_dropout_default_mask_ratio"
    )

# Training tests.

_training_default

import numpy as np
import onnx

seed = np.int64(0)
node = onnx.helper.make_node(
    "Dropout", inputs=["x", "r", "t"], outputs=["y"], seed=seed
)

x = np.random.randn(3, 4, 5).astype(np.float32)
r = np.float32(0.5)
t = np.bool_(True)
y = dropout(x, r, training_mode=t)
expect(
    node, inputs=[x, r, t], outputs=[y], name="test_training_dropout_default"
)

_training_default_ratio_mask

import numpy as np
import onnx

seed = np.int64(0)
node = onnx.helper.make_node(
    "Dropout", inputs=["x", "r", "t"], outputs=["y", "z"], seed=seed
)

x = np.random.randn(3, 4, 5).astype(np.float32)
r = np.float32(0.5)
t = np.bool_(True)
y, z = dropout(x, r, training_mode=t, return_mask=True)
expect(
    node,
    inputs=[x, r, t],
    outputs=[y, z],
    name="test_training_dropout_default_mask",
)

_training

import numpy as np
import onnx

seed = np.int64(0)
node = onnx.helper.make_node(
    "Dropout", inputs=["x", "r", "t"], outputs=["y"], seed=seed
)

x = np.random.randn(3, 4, 5).astype(np.float32)
r = np.float32(0.75)
t = np.bool_(True)
y = dropout(x, r, training_mode=t)
expect(node, inputs=[x, r, t], outputs=[y], name="test_training_dropout")

_training_ratio_mask

import numpy as np
import onnx

seed = np.int64(0)
node = onnx.helper.make_node(
    "Dropout", inputs=["x", "r", "t"], outputs=["y", "z"], seed=seed
)

x = np.random.randn(3, 4, 5).astype(np.float32)
r = np.float32(0.75)
t = np.bool_(True)
y, z = dropout(x, r, training_mode=t, return_mask=True)
expect(
    node, inputs=[x, r, t], outputs=[y, z], name="test_training_dropout_mask"
)

_training_default_zero_ratio

import numpy as np
import onnx

seed = np.int64(0)
node = onnx.helper.make_node(
    "Dropout", inputs=["x", "r", "t"], outputs=["y"], seed=seed
)

x = np.random.randn(3, 4, 5).astype(np.float32)
r = np.float32(0.0)
t = np.bool_(True)
y = dropout(x, r, training_mode=t)
expect(
    node, inputs=[x, r, t], outputs=[y], name="test_training_dropout_zero_ratio"
)

_training_default_zero_ratio_mask

import numpy as np
import onnx

    seed = np.int64(0)
    node = onnx.helper.make_node(
        "Dropout", inputs=["x", "r", "t"], outputs=["y", "z"], seed=seed
    )

    x = np.random.randn(3, 4, 5).astype(np.float32)
    r = np.float32(0.0)
    t = np.bool_(True)
    y, z = dropout(x, r, training_mode=t, return_mask=True)
    expect(
        node,
        inputs=[x, r, t],
        outputs=[y, z],
        name="test_training_dropout_zero_ratio_mask",
    )

# Old dropout tests

_default_old

import numpy as np
import onnx

node = onnx.helper.make_node(
    "Dropout",
    inputs=["x"],
    outputs=["y"],
)

x = np.array([-1, 0, 1]).astype(np.float32)
y = x
expect(
    node,
    inputs=[x],
    outputs=[y],
    name="test_dropout_default_old",
    opset_imports=[helper.make_opsetid("", 11)],
)

_random_old

import numpy as np
import onnx

node = onnx.helper.make_node(
    "Dropout",
    inputs=["x"],
    outputs=["y"],
    ratio=0.2,
)

x = np.random.randn(3, 4, 5).astype(np.float32)
y = x
expect(
    node,
    inputs=[x],
    outputs=[y],
    name="test_dropout_random_old",
    opset_imports=[helper.make_opsetid("", 11)],
)

Dropout - 12#

Version

  • name: Dropout (GitHub)

  • domain: main

  • since_version: 12

  • function: False

  • support_level: SupportType.COMMON

  • shape inference: True

This version of the operator has been available since version 12.

Summary

Dropout takes an input floating-point tensor, an optional input ratio (floating-point scalar) and an optional input training_mode (boolean scalar). It produces two tensor outputs, output (floating-point tensor) and mask (optional Tensor<bool>). If training_mode is true then the output Y will be a random dropout; Note that this Dropout scales the masked input data by the following equation, so to convert the trained model into inference mode, the user can simply not pass training_mode input or set it to false.

output = scale * data * mask,

where

scale = 1. / (1. - ratio).

This operator has optional inputs/outputs. See ONNX for more details about the representation of optional arguments. An empty string may be used in the place of an actual argument’s name to indicate a missing argument. Trailing optional arguments (those not followed by an argument that is present) may also be simply omitted.

Attributes

  • seed: (Optional) Seed to the random generator, if not specified we will auto generate one.

Inputs

Between 1 and 3 inputs.

  • data (heterogeneous) - T: The input data as Tensor.

  • ratio (optional, heterogeneous) - T1: The ratio of random dropout, with value in [0, 1). If this input was not set, or if it was set to 0, the output would be a simple copy of the input. If it’s non-zero, output will be a random dropout of the scaled input, which is typically the case during training. It is an optional value, if not specified it will default to 0.5.

  • training_mode (optional, heterogeneous) - T2: If set to true then it indicates dropout is being used for training. It is an optional value hence unless specified explicitly, it is false. If it is false, ratio is ignored and the operation mimics inference mode where nothing will be dropped from the input data and if mask is requested as output it will contain all ones.

Outputs

Between 1 and 2 outputs.

  • output (heterogeneous) - T: The output.

  • mask (optional, heterogeneous) - T2: The output mask.

Type Constraints

  • T in ( tensor(double), tensor(float), tensor(float16) ): Constrain input and output types to float tensors.

  • T1 in ( tensor(double), tensor(float), tensor(float16) ): Constrain input ‘ratio’ types to float tensors.

  • T2 in ( tensor(bool) ): Constrain output ‘mask’ types to boolean tensors.

Dropout - 10#

Version

  • name: Dropout (GitHub)

  • domain: main

  • since_version: 10

  • function: False

  • support_level: SupportType.COMMON

  • shape inference: True

This version of the operator has been available since version 10.

Summary

Dropout takes one input floating tensor and produces two tensor outputs, output (floating tensor) and mask (Tensor<bool>). Depending on whether it is in test mode or not, the output Y will either be a random dropout, or a simple copy of the input. Note that our implementation of Dropout does scaling in the training phase, so during testing nothing needs to be done. This operator has optional inputs/outputs. See ONNX for more details about the representation of optional arguments. An empty string may be used in the place of an actual argument’s name to indicate a missing argument. Trailing optional arguments (those not followed by an argument that is present) may also be simply omitted.

Attributes

  • ratio: The ratio of random dropout

Inputs

  • data (heterogeneous) - T: The input data as Tensor.

Outputs

Between 1 and 2 outputs.

  • output (heterogeneous) - T: The output.

  • mask (optional, heterogeneous) - T1: The output mask.

Type Constraints

  • T in ( tensor(double), tensor(float), tensor(float16) ): Constrain input and output types to float tensors.

  • T1 in ( tensor(bool) ): Constrain output mask types to boolean tensors.

Dropout - 7#

Version

  • name: Dropout (GitHub)

  • domain: main

  • since_version: 7

  • function: False

  • support_level: SupportType.COMMON

  • shape inference: True

This version of the operator has been available since version 7.

Summary

Dropout takes one input data (Tensor<float>) and produces two Tensor outputs, output (Tensor<float>) and mask (Tensor<bool>). Depending on whether it is in test mode or not, the output Y will either be a random dropout, or a simple copy of the input. Note that our implementation of Dropout does scaling in the training phase, so during testing nothing needs to be done. This operator has optional inputs/outputs. See ONNX for more details about the representation of optional arguments. An empty string may be used in the place of an actual argument’s name to indicate a missing argument. Trailing optional arguments (those not followed by an argument that is present) may also be simply omitted.

Attributes

  • ratio: The ratio of random dropout

Inputs

  • data (heterogeneous) - T: The input data as Tensor.

Outputs

Between 1 and 2 outputs.

  • output (heterogeneous) - T: The output.

  • mask (optional, heterogeneous) - T: The output mask.

Type Constraints

  • T in ( tensor(double), tensor(float), tensor(float16) ): Constrain input and output types to float tensors.

Dropout - 6#

Version

  • name: Dropout (GitHub)

  • domain: main

  • since_version: 6

  • function: False

  • support_level: SupportType.COMMON

  • shape inference: True

This version of the operator has been available since version 6.

Summary

Dropout takes one input data (Tensor<float>) and produces two Tensor outputs, output (Tensor<float>) and mask (Tensor<bool>). Depending on whether it is in test mode or not, the output Y will either be a random dropout, or a simple copy of the input. Note that our implementation of Dropout does scaling in the training phase, so during testing nothing needs to be done.

Attributes

  • is_test: (int, default 0) if nonzero, run dropout in test mode where the output is simply Y = X.

  • ratio: (float, default 0.5) the ratio of random dropout

Inputs

  • data (heterogeneous) - T: The input data as Tensor.

Outputs

Between 1 and 2 outputs.

  • output (heterogeneous) - T: The output.

  • mask (optional, heterogeneous) - T: The output mask. If is_test is nonzero, this output is not filled.

Type Constraints

  • T in ( tensor(double), tensor(float), tensor(float16) ): Constrain input and output types to float tensors.

Dropout - 1#

Version

  • name: Dropout (GitHub)

  • domain: main

  • since_version: 1

  • function: False

  • support_level: SupportType.COMMON

  • shape inference: False

This version of the operator has been available since version 1.

Summary

Dropout takes one input data (Tensor<float>) and produces two Tensor outputs, output (Tensor<float>) and mask (Tensor<bool>). Depending on whether it is in test mode or not, the output Y will either be a random dropout, or a simple copy of the input. Note that our implementation of Dropout does scaling in the training phase, so during testing nothing needs to be done.

Attributes

  • consumed_inputs: legacy optimization attribute.

  • is_test: (int, default 0) if nonzero, run dropout in test mode where the output is simply Y = X.

  • ratio: (float, default 0.5) the ratio of random dropout

Inputs

  • data (heterogeneous) - T: The input data as Tensor.

Outputs

Between 1 and 2 outputs.

  • output (heterogeneous) - T: The output.

  • mask (optional, heterogeneous) - T: The output mask. If is_test is nonzero, this output is not filled.

Type Constraints

  • T in ( tensor(double), tensor(float), tensor(float16) ): Constrain input and output types to float tensors.