mozempk/ggml-org-llama-cpp-mirror
1
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2026-11-11 22:32:07 +01:00
Code Issues Packages Projects Releases Wiki Activity

ggml webgpu: actually add softmax, fix rms_norm offset (#16400)

Browse Source 
* implement soft_max

* Fix soft_max data race

* Temporary fix, wait on each submit
This commit is contained in:
Reese Levine
2025-10-04 20:59:31 -07:00
committed by GitHub
parent 86df2c9ae4
commit 35266573b9
3 changed files with 10 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files
ggml/src/ggml-webgpu/wgsl-shaders/rms_norm.wgsl
+1 -1
View File
@@ -84,7 +84,7 @@ fn main(@builtin(workgroup_id) wid: vec3<u32>,
let i2 = i / params.ne1;
let i1 = i % params.ne1;
let i_src_row = params.offset_src + i3 * params.stride_src3 + i2 * params.stride_src2 + i1 * params.stride_src1;
let i_dst_row = params.offset_src + i3 * params.stride_dst3 + i2 * params.stride_dst2 + i1 * params.stride_dst1;
let i_dst_row = params.offset_dst + i3 * params.stride_dst3 + i2 * params.stride_dst2 + i1 * params.stride_dst1;
let elems = (params.ne0 + wg_size - 1) / wg_size;
ggml/src/ggml-webgpu/wgsl-shaders/soft_max.tmpl.wgsl
+1
View File
@@ -300,6 +300,7 @@ fn main(@builtin(workgroup_id) wid: vec3<u32>,
workgroupBarrier();
}
let row_max = scratch[0];
workgroupBarrier();
var sum = 0.0f;
col = lid.x;