mirror of
https://github.com/ggml-org/llama.cpp.git
synced 2025-08-18 05:56:00 -04:00
Vulkan Shader Refactor, Memory Debugging Option (#7947)
* Refactor shaders, extract GLSL code from ggml_vk_generate_shaders.py into vulkan-shaders directory * Improve debug log code * Add memory debug output option * Fix flake8 * Fix unnecessary high llama-3 VRAM use
This commit is contained in:
13
vulkan-shaders/square.comp
Normal file
13
vulkan-shaders/square.comp
Normal file
@@ -0,0 +1,13 @@
|
||||
#version 450
|
||||
|
||||
#include "types.comp"
|
||||
#include "generic_unary_head.comp"
|
||||
|
||||
void main() {
|
||||
if (gl_GlobalInvocationID.x >= p.ne) {
|
||||
return;
|
||||
}
|
||||
|
||||
const FLOAT_TYPE val = FLOAT_TYPE(data_a[src0_idx(gl_GlobalInvocationID.x)]);
|
||||
data_d[p.d_offset + dst_idx(gl_GlobalInvocationID.x)] = D_TYPE(val * val);
|
||||
}
|
Reference in New Issue
Block a user