Skip to content

Commit ec562eb

Browse files
shaofeiqilhez
andauthored
opencl: add q5_0 and q5_1 MoE for Adreno (#22985)
* opencl: add q5_0 moe support * opencl: add q5_1 moe support * opencl: avoid potential leak * opencl: suppress unused var warning when building for non-Adreno --------- Co-authored-by: Li He <lih@qti.qualcomm.com>
1 parent 95d469a commit ec562eb

7 files changed

Lines changed: 1914 additions & 67 deletions

File tree

ggml/src/ggml-opencl/CMakeLists.txt

Lines changed: 4 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -106,6 +106,10 @@ set(GGML_OPENCL_KERNELS
106106
gemv_moe_q4_0_f32_ns
107107
gemm_moe_q4_1_f32_ns
108108
gemv_moe_q4_1_f32_ns
109+
gemm_moe_q5_0_f32_ns
110+
gemv_moe_q5_0_f32_ns
111+
gemm_moe_q5_1_f32_ns
112+
gemv_moe_q5_1_f32_ns
109113
gemm_moe_mxfp4_f32
110114
gemv_moe_mxfp4_f32
111115
gemm_moe_mxfp4_f32_ns

0 commit comments

Comments
 (0)