PredictUnit
Predict with a trained model and decode the result into labels.
The rows come in as model_input, the key
:class:ApplySessionPreprocessingUnit publishes: the raw rows when the
session fitted no preprocessor, the transformed ones when it did. Reading
that key rather than dataset is what lets a session that preprocesses
predict on the columns its model was trained on while the raw rows stay
untouched for whatever saves the prediction next to them.
The input columns are selected against that dataset at this moment, never against a column list captured earlier: whatever built it, a load from disk or hand-typed rows, with or without a preprocessor on top, is free to have produced a different shape.
The training dataset is required rather than reloaded because the task
decodes predicted class indexes against its labels. The model is handed the
selected columns unprepared: models apply their own preprocessing inside
predict, and preparing beforehand would break the ones that replace
their input columns with derived features.
Parameters
- task_name : string, default=
TabularClassificationTask - Name of the task that turns raw model output into labels.
- input_columns : array, default=
[] - Names of the columns handed to the model as input.
- output_columns : array, default=
[] - Names of the columns the model predicts. Only the first one is used: it names the column the predictions are written to.
Methods
execute(self, ctx: DashAI.back.units.context.ExecutionContext) -> None
PredictUnitDo the unit's work, reading from and writing to the context.
Parameters
- ctx : ExecutionContext
- The shared execution context.
validate(self, ctx: DashAI.back.units.context.ExecutionContext) -> None
PredictUnitResolve the task before anything observable happens.
get_credential(self, name: str)
ConfigObjectResolve a registered credential component by name.
Parameters
- name : str
- Credential component class name (e.g. "HuggingFaceCredential").
Returns
- BaseCredential
- An instance of the requested credential component.
get_schema(cls) -> dict
ConfigObjectGenerates the component related Json Schema.
Returns
- dict
- Dictionary representing the Json Schema of the component.
validate_and_transform(self, raw_data: dict) -> dict
ConfigObjectIt takes the data given by the user to initialize the model and returns it with all the objects that the model needs to work.
Parameters
- raw_data : dict
- A dictionary with the data provided by the user to initialize the model.
Returns
- dict
- A validated dictionary with the necessary objects.