numr 0.5.1

High-performance numerical computing with multi-backend GPU acceleration (CPU/CUDA/WebGPU)
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36

//! WebGPU backend operation implementations
//!
//! Each operation type is implemented in its own module to maintain clean separation
//! and make the codebase easier to navigate.

pub mod activation;
pub mod advanced_random;
pub mod binary;
pub mod compare;
pub mod complex;
pub mod conditional;
pub mod conv;
pub mod cumulative;
pub mod distance;
pub mod einsum;
pub mod indexing;
pub mod linalg;
pub mod logical;
pub mod matmul;
pub mod multivariate;
pub mod semiring_matmul;
pub mod normalization;
pub mod quasirandom;
pub mod random;
pub mod reduce;
pub mod scalar;
pub mod shape;
#[cfg(feature = "sparse")]
pub mod sparse_24;
pub mod sorting;
pub mod statistics;
pub mod type_conversion;
pub mod unary;
pub mod utility;
pub mod fp8_matmul;
pub mod gemm_epilogue;