CS348Project/llama.cpp
5
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-11-12 10:47:01 +00:00
Code Issues Packages Projects Releases Wiki Activity

ggml : check cuda and metal argsort limits and add test (#16323)

Browse Source 
* check cuda argsort limits and add test

* add metal check
This commit is contained in:
Sigbjørn Skjæret
2025-09-29 11:09:00 +02:00
committed by GitHub
parent 3a2bdcda0b
commit adc76347d7
3 changed files with 7 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
ggml/src/ggml-cuda/ggml-cuda.cu
View File

@@ -3639,9 +3639,11 @@ static bool ggml_backend_cuda_device_supports_op(ggml_backend_dev_t dev, const g
case GGML_OP_CONV_TRANSPOSE_2D:
case GGML_OP_POOL_2D:
case GGML_OP_SUM:
case GGML_OP_ARGSORT:
case GGML_OP_ACC:
return true;
case GGML_OP_ARGSORT:
// TODO: Support arbitrary column width
return op->src[0]->ne[0] <= 1024;
case GGML_OP_SUM_ROWS:
case GGML_OP_MEAN:
case GGML_OP_GROUP_NORM: