thek0tyara/llama-cpp-turboquant
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions 4

Added functionality with atomicadd

Browse source
This commit is contained in:
Alejandro Saucedo 2021-03-05 08:19:56 +00:00
parent 9a40465d69
commit 71129392c2
4 changed files with 57 additions and 23 deletions
Download patch file Download diff file Expand all files Collapse all files

6
test/TestMultipleAlgoExecutions.cpp
View file

@ -16,6 +16,8 @@ TEST(TestMultipleAlgoExecutions, TestEndToEndFunctionality)
std::string shader = (R"(
#version 450
#extension GL_EXT_shader_atomic_float: enable
layout (local_size_x = 1) in;
// The input tensors bind index is relative to index in parameter passed
@ -34,8 +36,8 @@ TEST(TestMultipleAlgoExecutions, TestEndToEndFunctionality)
void main() {
uint index = gl_GlobalInvocationID.x;
out_a[index] += in_a[index] * in_b[index];
out_b[index] += const_one * push_const.val;
atomicAdd(out_a[index], in_a[index] * in_b[index]);
atomicAdd(out_b[index], const_one * push_const.val);
}
)");