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
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55

//! CUDA tensor operation implementations
//!
//! This module contains all operation implementations for the CUDA runtime.

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;
#[cfg(feature = "fp8")]
pub mod fp8_matmul;
pub mod gemm_epilogue;
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;

// Re-export commonly used modules
pub use type_conversion::*;
pub use complex::*;
pub use normalization::*;
pub use matmul::*;
pub use cumulative::*;
pub use activation::*;
pub use binary::*;
pub use unary::*;
pub use indexing::*;
pub use statistics::*;
pub use random::*;
pub use linalg::*;
pub use reduce::*;
pub use shape::*;
pub use sorting::*;
pub use conditional::*;
pub use utility::*;