ver4a/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity

opencl: fix for small models (#11950)

Browse source 
* opencl: fix small shape gemv, remove unused extensions

* opencl: fix `transpose_16`, `dump_tensor`, enforce subgroup size

* opencl: fix for token length < 4

* opencl: use wave size of 64 for all Adreno GPUs

---------

Co-authored-by: Shawn Gu <quic_shawngu@quicinc.com>
Co-authored-by: Skyler Szot <quic_sszot@quicinc.com>
This commit is contained in:
lhez 2025-02-24 13:47:07 -08:00 • committed by GitHub
parent 7a2c913e66
commit 34a846b584
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
6 changed files with 67 additions and 59 deletions
Download patch file Download diff file Expand all files Collapse all files

3
ggml/src/ggml-opencl/kernels/ggml-opencl.cl
View file

@ -1797,6 +1797,9 @@ kernel void kernel_mul_mat_f16_f16(
//------------------------------------------------------------------------------
// mul_mat_f16_f32_1row
//------------------------------------------------------------------------------
#ifdef ADRENO_GPU
REQD_SUBGROUP_SIZE_64
#endif
kernel void kernel_mul_mat_f16_f32_1row(
global char * src0,
ulong offset0,