Interface ModelSpeculativeDecodingConfig.Builder

All Superinterfaces:

Buildable, CopyableBuilder<ModelSpeculativeDecodingConfig.Builder,ModelSpeculativeDecodingConfig>, SdkBuilder<ModelSpeculativeDecodingConfig.Builder,ModelSpeculativeDecodingConfig>, SdkPojo

Enclosing class:

ModelSpeculativeDecodingConfig

@Mutable
@NotThreadSafe
public static interface ModelSpeculativeDecodingConfig.Builder
extends SdkPojo, CopyableBuilder<ModelSpeculativeDecodingConfig.Builder,ModelSpeculativeDecodingConfig>

Method Summary

Modifier and Type	Method	Description
ModelSpeculativeDecodingConfig.Builder	technique(String technique)	The speculative decoding technique to apply during model optimization.
ModelSpeculativeDecodingConfig.Builder	technique(ModelSpeculativeDecodingTechnique technique)	The speculative decoding technique to apply during model optimization.
default ModelSpeculativeDecodingConfig.Builder	trainingDataSource(Consumer<ModelSpeculativeDecodingTrainingDataSource.Builder> trainingDataSource)	The location of the training data to use for speculative decoding.
ModelSpeculativeDecodingConfig.Builder	trainingDataSource(ModelSpeculativeDecodingTrainingDataSource trainingDataSource)	The location of the training data to use for speculative decoding.

Methods inherited from interface software.amazon.awssdk.utils.builder.CopyableBuilder

copy

Methods inherited from interface software.amazon.awssdk.utils.builder.SdkBuilder

applyMutation, build

Methods inherited from interface software.amazon.awssdk.core.SdkPojo

equalsBySdkFields, sdkFieldNameToField, sdkFields

Method Details

technique

ModelSpeculativeDecodingConfig.Builder technique(String technique)

The speculative decoding technique to apply during model optimization.

Parameters:

technique - The speculative decoding technique to apply during model optimization.

Returns:

Returns a reference to this object so that method calls can be chained together.

See Also:

• ModelSpeculativeDecodingTechnique
• ModelSpeculativeDecodingTechnique

technique

ModelSpeculativeDecodingConfig.Builder technique(ModelSpeculativeDecodingTechnique technique)

The speculative decoding technique to apply during model optimization.

Parameters:

technique - The speculative decoding technique to apply during model optimization.

Returns:

Returns a reference to this object so that method calls can be chained together.

See Also:

• ModelSpeculativeDecodingTechnique
• ModelSpeculativeDecodingTechnique

trainingDataSource

ModelSpeculativeDecodingConfig.Builder trainingDataSource(ModelSpeculativeDecodingTrainingDataSource trainingDataSource)

The location of the training data to use for speculative decoding. The data must be formatted as ShareGPT, OpenAI Completions or OpenAI Chat Completions. The input can also be unencrypted captured data from a SageMaker endpoint as long as the endpoint uses one of the above formats.

Parameters:

trainingDataSource - The location of the training data to use for speculative decoding. The data must be formatted as ShareGPT, OpenAI Completions or OpenAI Chat Completions. The input can also be unencrypted captured data from a SageMaker endpoint as long as the endpoint uses one of the above formats.

Returns:

Returns a reference to this object so that method calls can be chained together.

trainingDataSource

default ModelSpeculativeDecodingConfig.Builder trainingDataSource(Consumer<ModelSpeculativeDecodingTrainingDataSource.Builder> trainingDataSource)

The location of the training data to use for speculative decoding. The data must be formatted as ShareGPT, OpenAI Completions or OpenAI Chat Completions. The input can also be unencrypted captured data from a SageMaker endpoint as long as the endpoint uses one of the above formats.

This is a convenience method that creates an instance of the ModelSpeculativeDecodingTrainingDataSource.Builder avoiding the need to create one manually via ModelSpeculativeDecodingTrainingDataSource.builder().

When the Consumer completes, SdkBuilder.build() is called immediately and its result is passed to trainingDataSource(ModelSpeculativeDecodingTrainingDataSource).

Parameters:

trainingDataSource - a consumer that will call methods on ModelSpeculativeDecodingTrainingDataSource.Builder

Returns:

Returns a reference to this object so that method calls can be chained together.

See Also:

• trainingDataSource(ModelSpeculativeDecodingTrainingDataSource)