scxpand.mlp.mlp_params

Classes

MLPParam([use_log_transform, n_epochs, ...])

class scxpand.mlp.mlp_params.MLPParam(use_log_transform=True, n_epochs=10, early_stopping_patience=5, init_learning_rate=0.0001, weight_decay=0.05, max_grad_norm=10.0, lr_scheduler_config=<factory>, lr_scheduler_type=LRSchedulerType.REDUCE_LR_ON_PLATEAU, optimizer_type=OptimizerType.ADAMW, adam_betas=(0.9, 0.999), train_batch_size=2048, inference_batch_size=2048, sampler_type=SamplerType.RANDOM, layer_units=(1024, 512, 256, 128), dropout_rate=0.3, mask_rate=0.1, noise_std=0.0001, soft_loss_beta=1.0, soft_loss_start_epoch=None, positives_weight=1.0, train_log_interval=5, random_seed=42, aux_categorical_types=<factory>, cat_loss_weight=1.0)

classmethod get_model_type()

Return the model type identifier for this parameter class.

Return type:

ModelType

__init__(use_log_transform=True, n_epochs=10, early_stopping_patience=5, init_learning_rate=0.0001, weight_decay=0.05, max_grad_norm=10.0, lr_scheduler_config=<factory>, lr_scheduler_type=LRSchedulerType.REDUCE_LR_ON_PLATEAU, optimizer_type=OptimizerType.ADAMW, adam_betas=(0.9, 0.999), train_batch_size=2048, inference_batch_size=2048, sampler_type=SamplerType.RANDOM, layer_units=(1024, 512, 256, 128), dropout_rate=0.3, mask_rate=0.1, noise_std=0.0001, soft_loss_beta=1.0, soft_loss_start_epoch=None, positives_weight=1.0, train_log_interval=5, random_seed=42, aux_categorical_types=<factory>, cat_loss_weight=1.0)

get_data_loader_params()

Return a DataLoaderParams object with data loader-related parameters.

Return type:

DataLoaderParams

get_dataset_params()

Return a DatasetParams object with dataset-related parameters.

Return type:

DataAugmentParams

get_lr_scheduler_params()

Return an LRSchedulerParams object with learning rate scheduler-related parameters.

Return type:

LRSchedulerParams

get_optimizer_params()

Return an OptimizerParams object with optimizer-related parameters.

Return type:

OptimizerParams

adam_betas: tuple = (0.9, 0.999)

aux_categorical_types: tuple[str, ...]

cat_loss_weight: float = 1.0

dropout_rate: float = 0.3

early_stopping_patience: int = 5

inference_batch_size: int = 2048

init_learning_rate: float = 0.0001

layer_units: tuple = (1024, 512, 256, 128)

lr_scheduler_config: dict[str, object]

lr_scheduler_type: LRSchedulerType = 'ReduceLROnPlateau'

mask_rate: float = 0.1

max_grad_norm: float = 10.0

n_epochs: int = 10

noise_std: float = 0.0001

optimizer_type: OptimizerType = 'AdamW'

positives_weight: float = 1.0

random_seed: int = 42

sampler_type: SamplerType = 'random'

soft_loss_beta: float | None = 1.0

soft_loss_start_epoch: int | None = None

train_batch_size: int = 2048

train_log_interval: int = 5

use_log_transform: bool = True

weight_decay: float = 0.05