mirror of
https://github.com/LostRuins/koboldcpp.git
synced 2026-08-30 00:50:49 +02:00
Jeff Bolz
80dd7ff22f
* tests: Fix memory bandwidth calculation for perf tests Add a flops calculation for flash attention. Add one GGML_OP_CPY perf test. * vulkan: Optimize contiguous copies Add a variant of the copy shader for when the tensors are contiguous. Avoid the complex addressing calculations, and do four elements per invocation to hide some other overhead. Apply similar changes to the scale shader, since scale is always contiguous. Add a "progress bar" for shader compiles.
25 lines
583 B
Plaintext
25 lines
583 B
Plaintext
#version 450
|
|
|
|
#include "types.comp"
|
|
#include "generic_unary_head.comp"
|
|
|
|
const uint num_threads = 128;
|
|
|
|
layout(local_size_x = num_threads, local_size_y = 1, local_size_z = 1) in;
|
|
|
|
void main() {
|
|
uint idx = get_idx();
|
|
|
|
// num_threads * num_iter must equal 512, to match the wg_denoms and get_idx calculation
|
|
const uint num_iter = 4;
|
|
|
|
[[unroll]] for (uint i = 0; i < num_iter; ++i) {
|
|
if (idx >= p.ne) {
|
|
continue;
|
|
}
|
|
|
|
data_d[p.d_offset + idx] = D_TYPE(FLOAT_TYPE(data_a[idx]) * FLOAT_TYPE(p.param1));
|
|
idx += num_threads;
|
|
}
|
|
}
|