Setup hyperparameters for TabNet training.
setup_TabNet(
batch_size = 500L,
penalty = 0.001,
clip_value = NULL,
loss = "auto",
epochs = 50L,
drop_last = FALSE,
decision_width = NULL,
attention_width = NULL,
num_steps = 3L,
feature_reusage = 1.3,
mask_type = "sparsemax",
virtual_batch_size = 65536L,
valid_split = 0,
learn_rate = 0.02,
optimizer = "adam",
lr_scheduler = NULL,
lr_decay = 0.1,
step_size = 30L,
checkpoint_epochs = 10L,
cat_emb_dim = 1L,
num_independent = 2L,
num_shared = 2L,
num_independent_decoder = 1L,
num_shared_decoder = 1L,
momentum = 0.02,
pretraining_ratio = 0.5,
device = NULL,
importance_sample_size = NULL,
early_stopping_monitor = "auto",
early_stopping_tolerance = 0,
early_stopping_patience = 0L,
num_workers = 0L,
skip_importance = FALSE,
ifw = FALSE
)TabNetHyperparameters object.
(Tunable) Integer [1, Inf): Batch size.
(Tunable) Numeric [0, Inf): Regularization penalty.
(Tunable) Optional Numeric: Gradient clip value. NULL clips no gradients.
(Tunable) Character: Loss function.
(Tunable) Integer [1, Inf): Number of epochs.
(Tunable) Logical: If TRUE, drop last batch.
(Tunable) Optional Integer [1, Inf): Decision prediction layer width. NULL uses attention_width, or 8 where both are NULL.
(Tunable) Optional Integer [1, Inf): Attention embedding width. NULL uses decision_width, or 8 where both are NULL.
(Tunable) Integer [1, Inf): Number of steps.
(Tunable) Numeric [0, Inf): Feature reusage.
(Tunable) Character {"sparsemax", "entmax"}: Mask type.
(Tunable) Integer [1, Inf): Virtual batch size.
(Tunable) Numeric [0, 1): Validation split.
(Tunable) Numeric (0, Inf): Learning rate.
Character: Optimizer name, resolved by the tabnet backend.
Optional Character {"step", "reduce_on_plateau"}: Learning-rate scheduler.
(Tunable) Numeric [0, 1]: Learning rate decay.
(Tunable) Integer [1, Inf): Step size.
(Tunable) Integer [1, Inf): Checkpoint epochs.
(Tunable) Integer [1, Inf): Categorical embedding dimension.
(Tunable) Integer [1, Inf): Number of independent Gated Linear Units (GLU) at each step of the encoder.
(Tunable) Integer [1, Inf): Number of shared Gated Linear Units (GLU) at each step of the encoder.
(Tunable) Integer [1, Inf): Number of independent GLU layers for pretraining.
(Tunable) Integer [1, Inf): Number of shared GLU layers for pretraining.
(Tunable) Numeric [0, Inf): Momentum.
(Tunable) Numeric [0, 1]: Pretraining ratio.
Optional Character: Deprecated and ignored. Set the compute device in the
execution_config passed to train, e.g. setup_FutureExecution(device = "mps"). Unset there, TabNet uses "cuda" where available and "cpu"
otherwise; the Apple silicon GPU, slower than the CPU for TabNet at every size
rtemis benchmarked, runs only when named.
(Tunable) Optional Integer [1, Inf): Cases drawn, with replacement, from the training cases left after the validation split to compute feature importance. Unset draws as many as there are such cases, at most 100,000. Applies only when skip_importance is FALSE.
(Tunable) Character {"auto", "valid_loss", "train_loss"}: Early stopping monitor.
(Tunable) Numeric [0, Inf): Minimum relative improvement to reset the patience counter.
(Tunable) Integer [0, Inf): Number of epochs without improving before stopping.
Integer [0, Inf): Number of subprocesses for data loading.
Logical: If TRUE, skip importance calculation.
(Tunable) Logical: If TRUE, use Inverse Frequency Weighting in classification.
EDG
tabnet_hyperparams <- setup_TabNet(epochs = 100L, learn_rate = 0.01)
tabnet_hyperparams
Run the code above in your browser using DataLab