Class DpoHyperparameters
-
- All Implemented Interfaces:
public final class DpoHyperparameters
The hyperparameters used for the DPO fine-tuning job.
-
-
Nested Class Summary
Nested Classes Modifier and Type Class Description public final class
DpoHyperparameters.Builder
A builder for DpoHyperparameters.
public final class
DpoHyperparameters.BatchSize
Number of examples in each batch. A larger batch size means that model parameters are updated less frequently, but with lower variance.
public final class
DpoHyperparameters.Beta
The beta value for the DPO method. A higher beta value will increase the weight of the penalty between the policy and reference model.
public final class
DpoHyperparameters.LearningRateMultiplier
Scaling factor for the learning rate. A smaller learning rate may be useful to avoid overfitting.
public final class
DpoHyperparameters.NEpochs
The number of epochs to train the model for. An epoch refers to one full cycle through the training dataset.
-
Method Summary
-
-
Method Detail
-
batchSize
final Optional<DpoHyperparameters.BatchSize> batchSize()
Number of examples in each batch. A larger batch size means that model parameters are updated less frequently, but with lower variance.
-
beta
final Optional<DpoHyperparameters.Beta> beta()
The beta value for the DPO method. A higher beta value will increase the weight of the penalty between the policy and reference model.
-
learningRateMultiplier
final Optional<DpoHyperparameters.LearningRateMultiplier> learningRateMultiplier()
Scaling factor for the learning rate. A smaller learning rate may be useful to avoid overfitting.
-
nEpochs
final Optional<DpoHyperparameters.NEpochs> nEpochs()
The number of epochs to train the model for. An epoch refers to one full cycle through the training dataset.
-
_batchSize
final JsonField<DpoHyperparameters.BatchSize> _batchSize()
Returns the raw JSON value of batchSize.
Unlike batchSize, this method doesn't throw if the JSON field has an unexpected type.
-
_beta
final JsonField<DpoHyperparameters.Beta> _beta()
Returns the raw JSON value of beta.
Unlike beta, this method doesn't throw if the JSON field has an unexpected type.
-
_learningRateMultiplier
final JsonField<DpoHyperparameters.LearningRateMultiplier> _learningRateMultiplier()
Returns the raw JSON value of learningRateMultiplier.
Unlike learningRateMultiplier, this method doesn't throw if the JSON field has an unexpected type.
-
_nEpochs
final JsonField<DpoHyperparameters.NEpochs> _nEpochs()
Returns the raw JSON value of nEpochs.
Unlike nEpochs, this method doesn't throw if the JSON field has an unexpected type.
-
_additionalProperties
final Map<String, JsonValue> _additionalProperties()
-
toBuilder
final DpoHyperparameters.Builder toBuilder()
-
validate
final DpoHyperparameters validate()
-
builder
final static DpoHyperparameters.Builder builder()
Returns a mutable builder for constructing an instance of DpoHyperparameters.
-
-
-
-