API Reference
Dataset
Dataset
@dataclass(repr=False)
class Dataset(Utils)
create_item
def create_item(input: Dict,
expected_output: Optional[Dict] = None,
metadata: Optional[Dict] = None) -> DatasetItem
Create a new dataset item and add it to this dataset.
Arguments:
input
: The input data for the dataset item.expected_output
: The output data for the dataset item (optional).metadata
: Metadata for the dataset item (optional).
Returns:
The created DatasetItem instance.
create_experiment
def create_experiment(name: str,
prompt_id: Optional[str] = None,
params: Optional[Dict] = None) -> DatasetExperiment
Creates a new dataset experiment based on this dataset.
Arguments:
name
: The name of the experiment .prompt_id
: The Prompt ID used on LLM calls (optional).params
: The params used on the experiment.
Returns:
The created DatasetExperiment instance.
delete_item
def delete_item(item_id: str)
Delete a dataset item from this dataset.
Arguments:
api
: An instance of the DatasetAPI to make the call.item_id
: The ID of the dataset item to delete.
add_step
def add_step(step_id: str, metadata: Optional[Dict] = None) -> DatasetItem
Create a new dataset item based on a step and add it to this dataset.
Arguments:
step_id
: The id of the step to add to the dataset.metadata
: Metadata for the dataset item (optional).
Returns:
The created DatasetItem instance.
add_generation
def add_generation(generation_id: str,
metadata: Optional[Dict] = None) -> DatasetItem
Create a new dataset item based on a generation and add it to this dataset.
Arguments:
generation_id
: The id of the generation to add to the dataset.metadata
: Metadata for the dataset item (optional).
Returns:
The created DatasetItem instance.
Was this page helpful?