torchfont.transforms.transforms¶

Composable transforms for preprocessing tensor glyph sequences.

Examples

Compose multiple operations to normalize glyph tensors:

pipeline = Compose([LimitSequenceLength(256), Patchify(32)])

Background:: The transforms aim to cover the common preprocessing steps for vectorized glyph data while still integrating cleanly with PyTorch Dataset instances.

Classes

`Compose`(transforms)	Apply a sequence of transform callables to each sample.
`LimitSequenceLength`(max_len)	Trim glyph sequences to a fixed maximum length.
`Patchify`(patch_size)	Pad glyph sequences and reshape them into equal-sized patches.