nnf_mixup: Mixup logic

In mlverse/torchlight: Higher Level 'API' for 'torch'

Mixup logic

Description

Logic underlying luz_callback_mixup().

Usage

nnf_mixup(x, y, weight)

Arguments

x	an input batch
y	a target batch
weight	weighting coefficient to be used by torch_lerp()

Details

Based on the passed-in input and target batches, as well as applicable mixing weights, we return new tensors intended to replace the current batch. The new input batch is a weighted linear combination of input batch items, while the new target batch bundles the original targets, as well as the mixing weights, in a nested list.

Value

A list of:

• x, the new, mixed-up input batch

• y, a list of:

  • ys, a list of:

    • y1, the original target y1

    • y2, the mixed-in target y2

  • weight, the mixing weights

See Also

luz_callback_mixup()

Examples

if (torch::torch_is_installed()) {
batch_x <- torch::torch_randn(c(10, 768))
batch_y <- torch::torch_randn(10)
weight <- torch::torch_tensor(rep(0.9, 10))$view(c(10, 1))
nnf_mixup(batch_x, batch_y, weight)
}

mlverse/torchlight documentation built on Sept. 19, 2024, 11:22 p.m.