Dataset
Complete reference for interacting with Datasets in judgeval
Class Methods
Dataset.create()
Create a new evaluation dataset for storage and reuse across multiple evaluation runs.
Example Code
from judgeval.dataset import Dataset
from judgeval.data import Example
dataset = Dataset.create(
name="qa_dataset",
project_name="default_project",
examples=[Example(input="...", actual_output="...")]
)Dataset.get()
Retrieve a dataset from the Judgment platform by its name and project name.
Parameters
project_name
str
Required
The name of the project where the dataset is stored.
Example:
"default_project"
Example Code
dataset = Dataset.get(
name="my_dataset",
project_name="default_project"
)Instance Methods
dataset.add_examples()
Add examples to the dataset.
Parameters
examples
List[Example]
Required
List of examples to add to the dataset.
Example Code
from judgeval.data import Example
from judgeval.dataset import Dataset
dataset = Dataset.get(
name="my_dataset",
project_name="default_project"
)
dataset.add_examples(
examples=[Example(input="...", actual_output="...")]
)dataset.add_traces()
Add traces to the dataset.
Parameters
traces
List[Trace]
Required
List of traces to add to the dataset.
Example Code
from judgeval.trace import Trace
from judgeval.dataset import Dataset
dataset = Dataset.get(
name="my_dataset",
project_name="default_project"
)
dataset.add_traces(
traces=[Trace(...)]
)dataset.add_from_yaml()
Add examples from a YAML file.
Parameters
file_path
str
Required
Path to the YAML file containing examples.
Example Code
from judgeval.dataset import Dataset
dataset = Dataset.get(
name="my_dataset",
project_name="default_project"
)
dataset.add_from_yaml(
file_path="path/to/examples.yaml"
)dataset.add_from_json()
Add examples from a JSON file.
Parameters
file_path
str
Required
Path to the JSON file containing examples.
Example Code
dataset.add_from_json(
file_path="path/to/examples.json"
)dataset.save_as()
Save the dataset as a file.
Example Code
dataset.save_as(
file_type="json",
dir_path="/path/to/save",
save_name="my_dataset"
)