jactorch.functional#

Functional APIs for JacTorch. By default, you should just access functions with jactorch.*. Or, if you want to be explicit about the functions, you can import the specific function from jactorch.functional.*.

Submodules

`arith`	Arithmetic operations.
`clustering`	Clustering functions.
`grad`	Gradient modification functions.
`indexing`	Tensor indexing utils.
`kernel`	Useful utilities for kernel-based attention mechanism.
`linalg`	Linear algebra functions.
`loglinear`	Linear algebra functions in the log space.
`masking`	Masking-related functions.
`probability`	Probability distributions related functions.
`quantization`	Basic quantization functions with a straight-through gradient estimator.
`range`	Numeric range functions.
`sampling`	Sampling functions.
`shape`	Tensor shape utilities.