hanzo-ml 0.10.2

Minimalist ML framework.
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14

#version 450
// f32 -> f16 elementwise cast. Feeds the cooperative-matrix GEMM, whose inputs are fp16.
#extension GL_EXT_shader_explicit_arithmetic_types_float16 : require
#extension GL_EXT_shader_16bit_storage : require
layout(local_size_x = 64) in;
layout(set = 0, binding = 0) readonly  buffer In  { float     src[]; };
layout(set = 0, binding = 1) writeonly buffer Out { float16_t dst[]; };
layout(push_constant) uniform Pc { uint n; };
void main() {
    uint i = gl_GlobalInvocationID.x;
    if (i < n) {
        dst[i] = float16_t(src[i]);
    }
}