public static final class IntegratedGradientsAttribution.Builder extends GeneratedMessageV3.Builder<IntegratedGradientsAttribution.Builder> implements IntegratedGradientsAttributionOrBuilderAn attribution method that computes the Aumann-Shapley value taking advantage of the model's fully differentiable structure. Refer to this paper for more details: https://arxiv.org/abs/1703.01365
Protobuf type google.cloud.vertexai.v1.IntegratedGradientsAttribution
Inheritance
Object > AbstractMessageLite.Builder<MessageType,BuilderType> > AbstractMessage.Builder<BuilderType> > GeneratedMessageV3.Builder > IntegratedGradientsAttribution.BuilderImplements
IntegratedGradientsAttributionOrBuilderStatic Methods
getDescriptor()
public static final Descriptors.Descriptor getDescriptor()| Returns | |
|---|---|
| Type | Description |
Descriptor |
|
Methods
addRepeatedField(Descriptors.FieldDescriptor field, Object value)
public IntegratedGradientsAttribution.Builder addRepeatedField(Descriptors.FieldDescriptor field, Object value)| Parameters | |
|---|---|
| Name | Description |
field |
FieldDescriptor |
value |
Object |
| Returns | |
|---|---|
| Type | Description |
IntegratedGradientsAttribution.Builder |
|
build()
public IntegratedGradientsAttribution build()| Returns | |
|---|---|
| Type | Description |
IntegratedGradientsAttribution |
|
buildPartial()
public IntegratedGradientsAttribution buildPartial()| Returns | |
|---|---|
| Type | Description |
IntegratedGradientsAttribution |
|
clear()
public IntegratedGradientsAttribution.Builder clear()| Returns | |
|---|---|
| Type | Description |
IntegratedGradientsAttribution.Builder |
|
clearBlurBaselineConfig()
public IntegratedGradientsAttribution.Builder clearBlurBaselineConfig()Config for IG with blur baseline.
When enabled, a linear path from the maximally blurred image to the input image is created. Using a blurred baseline instead of zero (black image) is motivated by the BlurIG approach explained here: https://arxiv.org/abs/2004.03383
.google.cloud.vertexai.v1.BlurBaselineConfig blur_baseline_config = 3;
| Returns | |
|---|---|
| Type | Description |
IntegratedGradientsAttribution.Builder |
|
clearField(Descriptors.FieldDescriptor field)
public IntegratedGradientsAttribution.Builder clearField(Descriptors.FieldDescriptor field)| Parameter | |
|---|---|
| Name | Description |
field |
FieldDescriptor |
| Returns | |
|---|---|
| Type | Description |
IntegratedGradientsAttribution.Builder |
|
clearOneof(Descriptors.OneofDescriptor oneof)
public IntegratedGradientsAttribution.Builder clearOneof(Descriptors.OneofDescriptor oneof)| Parameter | |
|---|---|
| Name | Description |
oneof |
OneofDescriptor |
| Returns | |
|---|---|
| Type | Description |
IntegratedGradientsAttribution.Builder |
|
clearSmoothGradConfig()
public IntegratedGradientsAttribution.Builder clearSmoothGradConfig()Config for SmoothGrad approximation of gradients.
When enabled, the gradients are approximated by averaging the gradients from noisy samples in the vicinity of the inputs. Adding noise can help improve the computed gradients. Refer to this paper for more details: https://arxiv.org/pdf/1706.03825.pdf
.google.cloud.vertexai.v1.SmoothGradConfig smooth_grad_config = 2;
| Returns | |
|---|---|
| Type | Description |
IntegratedGradientsAttribution.Builder |
|
clearStepCount()
public IntegratedGradientsAttribution.Builder clearStepCount()Required. The number of steps for approximating the path integral. A good value to start is 50 and gradually increase until the sum to diff property is within the desired error range.
Valid range of its value is [1, 100], inclusively.
int32 step_count = 1 [(.google.api.field_behavior) = REQUIRED];
| Returns | |
|---|---|
| Type | Description |
IntegratedGradientsAttribution.Builder |
This builder for chaining. |
clone()
public IntegratedGradientsAttribution.Builder clone()| Returns | |
|---|---|
| Type | Description |
IntegratedGradientsAttribution.Builder |
|
getBlurBaselineConfig()
public BlurBaselineConfig getBlurBaselineConfig()Config for IG with blur baseline.
When enabled, a linear path from the maximally blurred image to the input image is created. Using a blurred baseline instead of zero (black image) is motivated by the BlurIG approach explained here: https://arxiv.org/abs/2004.03383
.google.cloud.vertexai.v1.BlurBaselineConfig blur_baseline_config = 3;
| Returns | |
|---|---|
| Type | Description |
BlurBaselineConfig |
The blurBaselineConfig. |
getBlurBaselineConfigBuilder()
public BlurBaselineConfig.Builder getBlurBaselineConfigBuilder()Config for IG with blur baseline.
When enabled, a linear path from the maximally blurred image to the input image is created. Using a blurred baseline instead of zero (black image) is motivated by the BlurIG approach explained here: https://arxiv.org/abs/2004.03383
.google.cloud.vertexai.v1.BlurBaselineConfig blur_baseline_config = 3;
| Returns | |
|---|---|
| Type | Description |
BlurBaselineConfig.Builder |
|
getBlurBaselineConfigOrBuilder()
public BlurBaselineConfigOrBuilder getBlurBaselineConfigOrBuilder()Config for IG with blur baseline.
When enabled, a linear path from the maximally blurred image to the input image is created. Using a blurred baseline instead of zero (black image) is motivated by the BlurIG approach explained here: https://arxiv.org/abs/2004.03383
.google.cloud.vertexai.v1.BlurBaselineConfig blur_baseline_config = 3;
| Returns | |
|---|---|
| Type | Description |
BlurBaselineConfigOrBuilder |
|
getDefaultInstanceForType()
public IntegratedGradientsAttribution getDefaultInstanceForType()| Returns | |
|---|---|
| Type | Description |
IntegratedGradientsAttribution |
|
getDescriptorForType()
public Descriptors.Descriptor getDescriptorForType()| Returns | |
|---|---|
| Type | Description |
Descriptor |
|
getSmoothGradConfig()
public SmoothGradConfig getSmoothGradConfig()Config for SmoothGrad approximation of gradients.
When enabled, the gradients are approximated by averaging the gradients from noisy samples in the vicinity of the inputs. Adding noise can help improve the computed gradients. Refer to this paper for more details: https://arxiv.org/pdf/1706.03825.pdf
.google.cloud.vertexai.v1.SmoothGradConfig smooth_grad_config = 2;
| Returns | |
|---|---|
| Type | Description |
SmoothGradConfig |
The smoothGradConfig. |
getSmoothGradConfigBuilder()
public SmoothGradConfig.Builder getSmoothGradConfigBuilder()Config for SmoothGrad approximation of gradients.
When enabled, the gradients are approximated by averaging the gradients from noisy samples in the vicinity of the inputs. Adding noise can help improve the computed gradients. Refer to this paper for more details: https://arxiv.org/pdf/1706.03825.pdf
.google.cloud.vertexai.v1.SmoothGradConfig smooth_grad_config = 2;
| Returns | |
|---|---|
| Type | Description |
SmoothGradConfig.Builder |
|
getSmoothGradConfigOrBuilder()
public SmoothGradConfigOrBuilder getSmoothGradConfigOrBuilder()Config for SmoothGrad approximation of gradients.
When enabled, the gradients are approximated by averaging the gradients from noisy samples in the vicinity of the inputs. Adding noise can help improve the computed gradients. Refer to this paper for more details: https://arxiv.org/pdf/1706.03825.pdf
.google.cloud.vertexai.v1.SmoothGradConfig smooth_grad_config = 2;
| Returns | |
|---|---|
| Type | Description |
SmoothGradConfigOrBuilder |
|
getStepCount()
public int getStepCount()Required. The number of steps for approximating the path integral. A good value to start is 50 and gradually increase until the sum to diff property is within the desired error range.
Valid range of its value is [1, 100], inclusively.
int32 step_count = 1 [(.google.api.field_behavior) = REQUIRED];
| Returns | |
|---|---|
| Type | Description |
int |
The stepCount. |
hasBlurBaselineConfig()
public boolean hasBlurBaselineConfig()Config for IG with blur baseline.
When enabled, a linear path from the maximally blurred image to the input image is created. Using a blurred baseline instead of zero (black image) is motivated by the BlurIG approach explained here: https://arxiv.org/abs/2004.03383
.google.cloud.vertexai.v1.BlurBaselineConfig blur_baseline_config = 3;
| Returns | |
|---|---|
| Type | Description |
boolean |
Whether the blurBaselineConfig field is set. |
hasSmoothGradConfig()
public boolean hasSmoothGradConfig()Config for SmoothGrad approximation of gradients.
When enabled, the gradients are approximated by averaging the gradients from noisy samples in the vicinity of the inputs. Adding noise can help improve the computed gradients. Refer to this paper for more details: https://arxiv.org/pdf/1706.03825.pdf
.google.cloud.vertexai.v1.SmoothGradConfig smooth_grad_config = 2;
| Returns | |
|---|---|
| Type | Description |
boolean |
Whether the smoothGradConfig field is set. |
internalGetFieldAccessorTable()
protected GeneratedMessageV3.FieldAccessorTable internalGetFieldAccessorTable()| Returns | |
|---|---|
| Type | Description |
FieldAccessorTable |
|
isInitialized()
public final boolean isInitialized()| Returns | |
|---|---|
| Type | Description |
boolean |
|
mergeBlurBaselineConfig(BlurBaselineConfig value)
public IntegratedGradientsAttribution.Builder mergeBlurBaselineConfig(BlurBaselineConfig value)Config for IG with blur baseline.
When enabled, a linear path from the maximally blurred image to the input image is created. Using a blurred baseline instead of zero (black image) is motivated by the BlurIG approach explained here: https://arxiv.org/abs/2004.03383
.google.cloud.vertexai.v1.BlurBaselineConfig blur_baseline_config = 3;
| Parameter | |
|---|---|
| Name | Description |
value |
BlurBaselineConfig |
| Returns | |
|---|---|
| Type | Description |
IntegratedGradientsAttribution.Builder |
|
mergeFrom(IntegratedGradientsAttribution other)
public IntegratedGradientsAttribution.Builder mergeFrom(IntegratedGradientsAttribution other)| Parameter | |
|---|---|
| Name | Description |
other |
IntegratedGradientsAttribution |
| Returns | |
|---|---|
| Type | Description |
IntegratedGradientsAttribution.Builder |
|
mergeFrom(CodedInputStream input, ExtensionRegistryLite extensionRegistry)
public IntegratedGradientsAttribution.Builder mergeFrom(CodedInputStream input, ExtensionRegistryLite extensionRegistry)| Parameters | |
|---|---|
| Name | Description |
input |
CodedInputStream |
extensionRegistry |
ExtensionRegistryLite |
| Returns | |
|---|---|
| Type | Description |
IntegratedGradientsAttribution.Builder |
|
| Exceptions | |
|---|---|
| Type | Description |
IOException |
|
mergeFrom(Message other)
public IntegratedGradientsAttribution.Builder mergeFrom(Message other)| Parameter | |
|---|---|
| Name | Description |
other |
Message |
| Returns | |
|---|---|
| Type | Description |
IntegratedGradientsAttribution.Builder |
|
mergeSmoothGradConfig(SmoothGradConfig value)
public IntegratedGradientsAttribution.Builder mergeSmoothGradConfig(SmoothGradConfig value)Config for SmoothGrad approximation of gradients.
When enabled, the gradients are approximated by averaging the gradients from noisy samples in the vicinity of the inputs. Adding noise can help improve the computed gradients. Refer to this paper for more details: https://arxiv.org/pdf/1706.03825.pdf
.google.cloud.vertexai.v1.SmoothGradConfig smooth_grad_config = 2;
| Parameter | |
|---|---|
| Name | Description |
value |
SmoothGradConfig |
| Returns | |
|---|---|
| Type | Description |
IntegratedGradientsAttribution.Builder |
|
mergeUnknownFields(UnknownFieldSet unknownFields)
public final IntegratedGradientsAttribution.Builder mergeUnknownFields(UnknownFieldSet unknownFields)| Parameter | |
|---|---|
| Name | Description |
unknownFields |
UnknownFieldSet |
| Returns | |
|---|---|
| Type | Description |
IntegratedGradientsAttribution.Builder |
|
setBlurBaselineConfig(BlurBaselineConfig value)
public IntegratedGradientsAttribution.Builder setBlurBaselineConfig(BlurBaselineConfig value)Config for IG with blur baseline.
When enabled, a linear path from the maximally blurred image to the input image is created. Using a blurred baseline instead of zero (black image) is motivated by the BlurIG approach explained here: https://arxiv.org/abs/2004.03383
.google.cloud.vertexai.v1.BlurBaselineConfig blur_baseline_config = 3;
| Parameter | |
|---|---|
| Name | Description |
value |
BlurBaselineConfig |
| Returns | |
|---|---|
| Type | Description |
IntegratedGradientsAttribution.Builder |
|
setBlurBaselineConfig(BlurBaselineConfig.Builder builderForValue)
public IntegratedGradientsAttribution.Builder setBlurBaselineConfig(BlurBaselineConfig.Builder builderForValue)Config for IG with blur baseline.
When enabled, a linear path from the maximally blurred image to the input image is created. Using a blurred baseline instead of zero (black image) is motivated by the BlurIG approach explained here: https://arxiv.org/abs/2004.03383
.google.cloud.vertexai.v1.BlurBaselineConfig blur_baseline_config = 3;
| Parameter | |
|---|---|
| Name | Description |
builderForValue |
BlurBaselineConfig.Builder |
| Returns | |
|---|---|
| Type | Description |
IntegratedGradientsAttribution.Builder |
|
setField(Descriptors.FieldDescriptor field, Object value)
public IntegratedGradientsAttribution.Builder setField(Descriptors.FieldDescriptor field, Object value)| Parameters | |
|---|---|
| Name | Description |
field |
FieldDescriptor |
value |
Object |
| Returns | |
|---|---|
| Type | Description |
IntegratedGradientsAttribution.Builder |
|
setRepeatedField(Descriptors.FieldDescriptor field, int index, Object value)
public IntegratedGradientsAttribution.Builder setRepeatedField(Descriptors.FieldDescriptor field, int index, Object value)| Parameters | |
|---|---|
| Name | Description |
field |
FieldDescriptor |
index |
int |
value |
Object |
| Returns | |
|---|---|
| Type | Description |
IntegratedGradientsAttribution.Builder |
|
setSmoothGradConfig(SmoothGradConfig value)
public IntegratedGradientsAttribution.Builder setSmoothGradConfig(SmoothGradConfig value)Config for SmoothGrad approximation of gradients.
When enabled, the gradients are approximated by averaging the gradients from noisy samples in the vicinity of the inputs. Adding noise can help improve the computed gradients. Refer to this paper for more details: https://arxiv.org/pdf/1706.03825.pdf
.google.cloud.vertexai.v1.SmoothGradConfig smooth_grad_config = 2;
| Parameter | |
|---|---|
| Name | Description |
value |
SmoothGradConfig |
| Returns | |
|---|---|
| Type | Description |
IntegratedGradientsAttribution.Builder |
|
setSmoothGradConfig(SmoothGradConfig.Builder builderForValue)
public IntegratedGradientsAttribution.Builder setSmoothGradConfig(SmoothGradConfig.Builder builderForValue)Config for SmoothGrad approximation of gradients.
When enabled, the gradients are approximated by averaging the gradients from noisy samples in the vicinity of the inputs. Adding noise can help improve the computed gradients. Refer to this paper for more details: https://arxiv.org/pdf/1706.03825.pdf
.google.cloud.vertexai.v1.SmoothGradConfig smooth_grad_config = 2;
| Parameter | |
|---|---|
| Name | Description |
builderForValue |
SmoothGradConfig.Builder |
| Returns | |
|---|---|
| Type | Description |
IntegratedGradientsAttribution.Builder |
|
setStepCount(int value)
public IntegratedGradientsAttribution.Builder setStepCount(int value)Required. The number of steps for approximating the path integral. A good value to start is 50 and gradually increase until the sum to diff property is within the desired error range.
Valid range of its value is [1, 100], inclusively.
int32 step_count = 1 [(.google.api.field_behavior) = REQUIRED];
| Parameter | |
|---|---|
| Name | Description |
value |
intThe stepCount to set. |
| Returns | |
|---|---|
| Type | Description |
IntegratedGradientsAttribution.Builder |
This builder for chaining. |
setUnknownFields(UnknownFieldSet unknownFields)
public final IntegratedGradientsAttribution.Builder setUnknownFields(UnknownFieldSet unknownFields)| Parameter | |
|---|---|
| Name | Description |
unknownFields |
UnknownFieldSet |
| Returns | |
|---|---|
| Type | Description |
IntegratedGradientsAttribution.Builder |
|