sample

Typed containers for tensor payloads; record metadata stays with the caller.

Machine-readable model

• Tconstant
• TensorPayloadclassTwo-level tensor-only schema accepted by portable preprocessing layers.
• SampleclassLive tensor facade, not a serializable model config or an immutable tensor.

constant

T

Python

helia_edge/layers/preprocessing/sample.py:6

T = TypeVar('T')

class

TensorPayload

Python

Two-level tensor-only schema accepted by portable preprocessing layers.

helia_edge/layers/preprocessing/sample.py:9

TensorPayload()

Two-level tensor-only schema accepted by portable preprocessing layers.

attribute

signals

Python

helia_edge/layers/preprocessing/sample.py:12

signals: dict[str, T]

attribute

targets

Python

helia_edge/layers/preprocessing/sample.py:13

targets: dict[str, T]

attribute

masks

Python

helia_edge/layers/preprocessing/sample.py:14

masks: dict[str, T]

class

Sample

Python

Live tensor facade, not a serializable model config or an immutable tensor.

helia_edge/layers/preprocessing/sample.py:17

Sample(signals: dict[str, T], targets: dict[str, T] = dict(), masks: dict[str, T] = dict()) -> None

dataclass

Live tensor facade, not a serializable model config or an immutable tensor.

Conversion creates new dictionaries but never copies tensor storage. Keys in each group name tensor leaves. Keep record IDs and other metadata outside this object, and pass sample.tensor_tree() to a Keras layer.

attribute

signals

Python

helia_edge/layers/preprocessing/sample.py:26

signals: dict[str, T]

attribute

targets

Python

helia_edge/layers/preprocessing/sample.py:27

targets: dict[str, T] = field(default_factory=dict)

attribute

masks

Python

helia_edge/layers/preprocessing/sample.py:28

masks: dict[str, T] = field(default_factory=dict)

method

tensor_tree

Python

Return fresh containers referencing the original tensor leaves.

helia_edge/layers/preprocessing/sample.py:30

tensor_tree() -> TensorPayload[T]

Return fresh containers referencing the original tensor leaves.