Skip to content

WinoGrande

get_input_data_model()

Returns WinoGrande input data model.

Source code in synalinks/src/datasets/built_in/winogrande.py 
31
32
33
34
@synalinks_export("synalinks.datasets.winogrande.get_input_data_model")
def get_input_data_model():
    """Returns WinoGrande input data model."""
    return WinoGrandeQuestion

get_output_data_model()

Returns WinoGrande output data model.

Source code in synalinks/src/datasets/built_in/winogrande.py 
37
38
39
40
@synalinks_export("synalinks.datasets.winogrande.get_output_data_model")
def get_output_data_model():
    """Returns WinoGrande output data model."""
    return WinoGrandeAnswer

iterable_dataset(repeat=1, batch_size=1, limit=None, split='train')

Streaming dataset for RL-style training.

Returns:

Type Description
HuggingFaceDataset

A streaming, iterable dataset.
Source code in synalinks/src/datasets/built_in/winogrande.py 
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
@synalinks_export("synalinks.datasets.winogrande.iterable_dataset")
def iterable_dataset(repeat=1, batch_size=1, limit=None, split="train"):
    """
    Streaming dataset for RL-style training.

    Returns:
        (HuggingFaceDataset): A streaming, iterable dataset.
    """
    return HuggingFaceDataset(
        path="allenai/winogrande",
        name="winogrande_xl",
        split=split,
        streaming=True,
        input_data_model=WinoGrandeQuestion,
        input_template=_INPUT_TEMPLATE,
        output_data_model=WinoGrandeAnswer,
        output_template=_OUTPUT_TEMPLATE,
        batch_size=batch_size,
        limit=limit,
        repeat=repeat,
    )

load_data()

Load WinoGrande (XL).

HF test split has no public labels, so we use train for training and validation for evaluation.

Returns:

Type Description
tuple

(x_train, y_train), (x_test, y_test).
Source code in synalinks/src/datasets/built_in/winogrande.py 
55
56
57
58
59
60
61
62
63
64
65
66
67
68
@synalinks_export("synalinks.datasets.winogrande.load_data")
def load_data():
    """
    Load WinoGrande (XL).

    HF ``test`` split has no public labels, so we use ``train`` for
    training and ``validation`` for evaluation.

    Returns:
        (tuple): ``(x_train, y_train), (x_test, y_test)``.
    """
    x_train, y_train = _load("train")
    x_test, y_test = _load("validation")
    return (x_train, y_train), (x_test, y_test)