mediapipe_model_maker.text_classifier.BertHParams

The hyperparameters for a Bert Classifier.

Inherits From: BaseHParams

mediapipe_model_maker.text_classifier.BertHParams(
    learning_rate: float = 3e-05,
    batch_size: int = 48,
    epochs: int = 2,
    steps_per_epoch: Optional[int] = None,
    class_weights: Optional[Mapping[int, float]] = None,
    shuffle: bool = False,
    repeat: bool = False,
    export_dir: str = tempfile.mkdtemp(),
    distribution_strategy: str = 'off',
    num_gpus: int = 0,
    tpu: str = '',
    end_learning_rate: float = 0.0,
    optimizer: mediapipe_model_maker.text_classifier.BertOptimizer = mediapipe_model_maker.text_classifier.BertHParams.optimizer,
    weight_decay: float = 0.01,
    desired_precisions: Sequence[float] = dataclasses.field(default_factory=list),
    desired_recalls: Sequence[float] = dataclasses.field(default_factory=list),
    gamma: float = 2.0,
    tokenizer: mediapipe_model_maker.text_classifier.SupportedBertTokenizers = mediapipe_model_maker.text_classifier.BertHParams.tokenizer,
    checkpoint_frequency: int = 0
)

learning_rate Learning rate to use for gradient descent training.
end_learning_rate End learning rate for linear decay. Defaults to 0.
batch_size Batch size for training. Defaults to 48.
epochs Number of training iterations over the dataset. Defaults to 2.
optimizer Optimizer to use for training. Supported values are defined in BertOptimizer enum: ADAMW and LAMB.
weight_decay Weight decay of the optimizer. Defaults to 0.01.
desired_precisions If specified, adds a RecallAtPrecision metric per desired_precisions[i] entry which tracks the recall given the constraint on precision. Only supported for binary classification.
desired_recalls If specified, adds a PrecisionAtRecall metric per desired_recalls[i] entry which tracks the precision given the constraint on recall. Only supported for binary classification.
gamma Gamma parameter for focal loss. To use cross entropy loss, set this value to 0. Defaults to 2.0.
tokenizer Tokenizer to use for preprocessing. Must be one of the enum options of SupportedBertTokenizers. Defaults to FULL_TOKENIZER.
checkpoint_frequency Frequency(in epochs) of saving checkpoints during training. Defaults to 0 which does not save training checkpoints.
steps_per_epoch Dataclass field
class_weights Dataclass field
shuffle Dataclass field
repeat Dataclass field
export_dir Dataclass field
distribution_strategy Dataclass field
num_gpus Dataclass field
tpu Dataclass field

Methods

get_strategy

View source

get_strategy()

__eq__

__eq__(
    other
)

batch_size 48
checkpoint_frequency 0
class_weights None
distribution_strategy 'off'
end_learning_rate 0.0
epochs 2
export_dir '/tmpfs/tmp/tmpnt_h4p9w'
gamma 2.0
learning_rate 3e-05
num_gpus 0
optimizer <BertOptimizer.ADAMW: 'adamw'>
repeat False
shuffle False
steps_per_epoch None
tokenizer <SupportedBertTokenizers.FULL_TOKENIZER: 'fulltokenizer'>
tpu ''
weight_decay 0.01