TabularClassifier

class flash.tabular.classification.model.TabularClassifier(parameters, embedding_sizes, cat_dims, num_features, num_classes, labels=None, backbone='tabnet', loss_fn=torch.nn.functional.cross_entropy, optimizer='Adam', lr_scheduler=None, metrics=None, learning_rate=None, **backbone_kwargs)

The TabularClassifier is a Task for classifying tabular data. For more details, see Tabular Classification.

Parameters

• parameters (Dict[str, Any]) – The parameters computed from the training data (can be obtained from the parameters attribute of the TabularClassificationData object containing your training data).

• embedding_sizes (list) – List of (num_classes, emb_dim) to form categorical embeddings.

• cat_dims (list) – Number of distinct values for each categorical column

• num_features (int) – Number of columns in table

• num_classes (int) – Number of classes to classify

• backbone (str) – name of the model to use

• loss_fn (Callable) – Loss function for training, defaults to cross entropy.

• optimizer (TypeVar(OPTIMIZER_TYPE, str, Callable, Tuple[str, Dict[str, Any]], None)) – Optimizer to use for training.

• lr_scheduler (Optional[TypeVar(LR_SCHEDULER_TYPE, str, Callable, Tuple[str, Dict[str, Any]], Tuple[str, Dict[str, Any], Dict[str, Any]], None)]) – The LR scheduler to use during training.

• metrics (Optional[TypeVar(METRICS_TYPE, Metric, Mapping, Sequence, None)]) – Metrics to compute for training and evaluation. Can either be an metric from the torchmetrics package, a custom metric inherenting from torchmetrics.Metric, a callable function or a list/dict containing a combination of the aforementioned. In all cases, each metric needs to have the signature metric(preds,target) and return a single scalar tensor. Defaults to torchmetrics.Accuracy.

• learning_rate (Optional[float]) – Learning rate to use for training.

• **backbone_kwargs – Optional additional arguments for the model.

classmethod available_finetuning_strategies(cls)

Returns a list containing the keys of the available Finetuning Strategies.

Return type

List[str]

classmethod available_lr_schedulers(cls)

Returns a list containing the keys of the available LR schedulers.

Return type

List[str]

classmethod available_optimizers(cls)

Returns a list containing the keys of the available Optimizers.

Return type

List[str]

classmethod available_outputs(cls)

Returns the list of available outputs (that can be used during prediction or serving) for this Task.

Examples

..testsetup:

>>> from flash import Task

>>> print(Task.available_outputs())
['preds', 'raw']

Return type

List[str]

property data_parameters: Dict[str, Any]

Get the parameters computed from the training data used to create this TabularClassifier. Use these parameters to load data for evaluation / prediction.

Examples

>>> import flash
>>> from flash.core.data.utils import download_data
>>> from flash.tabular import TabularClassificationData, TabularClassifier
>>> download_data("https://pl-flash-data.s3.amazonaws.com/titanic.zip", "./data")
>>> model = TabularClassifier.load_from_checkpoint(
...     "https://flash-weights.s3.amazonaws.com/0.7.0/tabular_classification_model.pt"
... )
>>> datamodule = TabularClassificationData.from_csv(
...     predict_file="data/titanic/predict.csv",
...     parameters=model.data_parameters,
...     batch_size=8,
... )
>>> trainer = flash.Trainer()
>>> trainer.predict(
...     model,
...     datamodule=datamodule,
...     output="classes",
... )  
Predicting...

Return type

Dict[str, Any]