LRData

class jlpy.pytorch.base.linear_regression.base_impl.LRData

Bases: object

Synthetic data for linear regression.

Ins w:

Tensor-Vector –> Weights.

Ins b:

Tensor-Scalar –> Bias.

Ins noise:

Tensor-Vector –> Noise.

Ins num_train:

Size –> Training set.

Ins num_val:

Size –> Validation set.

Ins n:

Size –> Dataset.

Ins x:

Tensor-Matrix –> Input.

Ins y:

Tensor-Vector –> Output.

Methods

__init__

Construct a class instance.

__init__(bsize: int = 32, gap: int = 50, ws: list[float] = [2, -3.4], **kwargs: float) → None

Construct a class instance.

Parameters:

• bsize (int) – Batch size, default is 32.

• gap (int) – The plot gap for one epoch, default is 50.

• ws (list[float]) – List of the weights, default is [2, -3.4].

• b (float) – Bias, default is 4.2.

• n (float) – Noise, default is 0.01.