Class: Google::Apis::AiplatformV1::GoogleCloudAiplatformV1EvaluationRunMetricLlmBasedMetricSpec

Inherits:
Object
  • Object
show all
Includes:
Core::Hashable, Core::JsonObjectSupport
Defined in:
lib/google/apis/aiplatform_v1/classes.rb,
lib/google/apis/aiplatform_v1/representations.rb,
lib/google/apis/aiplatform_v1/representations.rb

Overview

Specification for an LLM based metric.

Instance Attribute Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(**args) ⇒ GoogleCloudAiplatformV1EvaluationRunMetricLlmBasedMetricSpec

Returns a new instance of GoogleCloudAiplatformV1EvaluationRunMetricLlmBasedMetricSpec.



9873
9874
9875
 
# File 'lib/google/apis/aiplatform_v1/classes.rb', line 9873

def initialize(**args)
   update!(**args)
end

Instance Attribute Details

#additional_config ⇒ Hash<String,Object>

Optional. Optional additional configuration for the metric. Corresponds to the JSON property additionalConfig

Returns:

  • (Hash<String,Object>) 


9840
9841
9842
 
# File 'lib/google/apis/aiplatform_v1/classes.rb', line 9840

def additional_config
  @additional_config
end

#judge_autorater_config ⇒ Google::Apis::AiplatformV1::GoogleCloudAiplatformV1EvaluationRunEvaluationConfigAutoraterConfig

The autorater config used for the evaluation run. Corresponds to the JSON property judgeAutoraterConfig

Returns:



9845
9846
9847
 
# File 'lib/google/apis/aiplatform_v1/classes.rb', line 9845

def judge_autorater_config
  @judge_autorater_config
end

#metric_prompt_template ⇒ String

Required. Template for the prompt sent to the judge model. Corresponds to the JSON property metricPromptTemplate

Returns:

  • (String) 


9850
9851
9852
 
# File 'lib/google/apis/aiplatform_v1/classes.rb', line 9850

def metric_prompt_template
  @metric_prompt_template
end

#predefined_rubric_generation_spec ⇒ Google::Apis::AiplatformV1::GoogleCloudAiplatformV1EvaluationRunMetricPredefinedMetricSpec

Specification for a pre-defined metric. Corresponds to the JSON property predefinedRubricGenerationSpec

Returns:



9855
9856
9857
 
# File 'lib/google/apis/aiplatform_v1/classes.rb', line 9855

def predefined_rubric_generation_spec
  @predefined_rubric_generation_spec
end

#rubric_generation_spec ⇒ Google::Apis::AiplatformV1::GoogleCloudAiplatformV1EvaluationRunMetricRubricGenerationSpec

Specification for how rubrics should be generated. Corresponds to the JSON property rubricGenerationSpec

Returns:



9860
9861
9862
 
# File 'lib/google/apis/aiplatform_v1/classes.rb', line 9860

def rubric_generation_spec
  @rubric_generation_spec
end

#rubric_group_key ⇒ String

Use a pre-defined group of rubrics associated with the input. Refers to a key in the rubric_groups map of EvaluationInstance. Corresponds to the JSON property rubricGroupKey

Returns:

  • (String) 


9866
9867
9868
 
# File 'lib/google/apis/aiplatform_v1/classes.rb', line 9866

def rubric_group_key
  @rubric_group_key
end

#system_instruction ⇒ String

Optional. System instructions for the judge model. Corresponds to the JSON property systemInstruction

Returns:

  • (String) 


9871
9872
9873
 
# File 'lib/google/apis/aiplatform_v1/classes.rb', line 9871

def system_instruction
  @system_instruction
end

Instance Method Details

#update!(**args) ⇒ Object

Update properties of this object



9878
9879
9880
9881
9882
9883
9884
9885
9886
 
# File 'lib/google/apis/aiplatform_v1/classes.rb', line 9878

def update!(**args)
  @additional_config = args[:additional_config] if args.key?(:additional_config)
  @judge_autorater_config = args[:judge_autorater_config] if args.key?(:judge_autorater_config)
  @metric_prompt_template = args[:metric_prompt_template] if args.key?(:metric_prompt_template)
  @predefined_rubric_generation_spec = args[:predefined_rubric_generation_spec] if args.key?(:predefined_rubric_generation_spec)
  @rubric_generation_spec = args[:rubric_generation_spec] if args.key?(:rubric_generation_spec)
  @rubric_group_key = args[:rubric_group_key] if args.key?(:rubric_group_key)
  @system_instruction = args[:system_instruction] if args.key?(:system_instruction)
end