Class: Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1ReinforcementLearningHyperParameters

Inherits:
Object
  • Object
show all
Includes:
Core::Hashable, Core::JsonObjectSupport
Defined in:
lib/google/apis/aiplatform_v1beta1/classes.rb,
lib/google/apis/aiplatform_v1beta1/representations.rb,
lib/google/apis/aiplatform_v1beta1/representations.rb

Overview

Hyperparameters for Reinforcement Learning.

Instance Attribute Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(**args) ⇒ GoogleCloudAiplatformV1beta1ReinforcementLearningHyperParameters

Returns a new instance of GoogleCloudAiplatformV1beta1ReinforcementLearningHyperParameters.



23348
23349
23350
 
# File 'lib/google/apis/aiplatform_v1beta1/classes.rb', line 23348

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

Instance Attribute Details

#epoch_count ⇒ Fixnum

Optional. Number of training epoches for the tuning job. Corresponds to the JSON property epochCount

Returns:

  • (Fixnum) 


23325
23326
23327
 
# File 'lib/google/apis/aiplatform_v1beta1/classes.rb', line 23325

def epoch_count
  @epoch_count
end

#human_feedback_config ⇒ Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1HumanFeedbackConfig

Configures Reinforcement Learning to use human feedback during tuning. Corresponds to the JSON property humanFeedbackConfig

Returns:



23330
23331
23332
 
# File 'lib/google/apis/aiplatform_v1beta1/classes.rb', line 23330

def human_feedback_config
  @human_feedback_config
end

#kl_coefficient ⇒ Float

Optional. KL divergence coefficient for Reinforcement Learning. Corresponds to the JSON property klCoefficient

Returns:

  • (Float) 


23335
23336
23337
 
# File 'lib/google/apis/aiplatform_v1beta1/classes.rb', line 23335

def kl_coefficient
  @kl_coefficient
end

#learning_rate_multiplier ⇒ Float

Optional. Learning rate multiplier for Reinforcement Learning. Corresponds to the JSON property learningRateMultiplier

Returns:

  • (Float) 


23340
23341
23342
 
# File 'lib/google/apis/aiplatform_v1beta1/classes.rb', line 23340

def learning_rate_multiplier
  @learning_rate_multiplier
end

#reward_model_training_config ⇒ Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1RewardModelTrainingConfig

Configures Reinforcement Learning to learn preference by training a reward model. Corresponds to the JSON property rewardModelTrainingConfig

Returns:



23346
23347
23348
 
# File 'lib/google/apis/aiplatform_v1beta1/classes.rb', line 23346

def reward_model_training_config
  @reward_model_training_config
end

Instance Method Details

#update!(**args) ⇒ Object

Update properties of this object



23353
23354
23355
23356
23357
23358
23359
 
# File 'lib/google/apis/aiplatform_v1beta1/classes.rb', line 23353

def update!(**args)
  @epoch_count = args[:epoch_count] if args.key?(:epoch_count)
  @human_feedback_config = args[:human_feedback_config] if args.key?(:human_feedback_config)
  @kl_coefficient = args[:kl_coefficient] if args.key?(:kl_coefficient)
  @learning_rate_multiplier = args[:learning_rate_multiplier] if args.key?(:learning_rate_multiplier)
  @reward_model_training_config = args[:reward_model_training_config] if args.key?(:reward_model_training_config)
end