candle-einops-macros 0.2.0

Procedural macros for candle-einops tensor transformations and einsum
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

mod einops;
mod einsum;

/// Macro to perform tensor transformations using simple expressions
///
/// This macro is re-exported as `candle_einops::einops`; the runtime crate's
/// documentation contains a complete runnable example.
#[proc_macro]
pub fn einops(input: proc_macro::TokenStream) -> proc_macro::TokenStream {
    einops::einops(input.into())
        .unwrap_or_else(|e| e.to_compile_error())
        .into()
}

/// Evaluates an explicit-output Einstein summation equation.
///
/// Supports any positive number of operands with named axes and at most one
/// `..` per axis list. Repeated input labels extract diagonals before
/// contraction. Use this through `candle_einops::einsum`; generated code calls
/// a private runtime ABI, so `candle-einops-macros` and `candle-einops` must use
/// exactly the same version.
#[proc_macro]
pub fn einsum(input: proc_macro::TokenStream) -> proc_macro::TokenStream {
    einsum::einsum(input.into())
        .unwrap_or_else(|error| error.to_compile_error())
        .into()
}