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

fix -nommq help for non CUDA/HIP

Browse source
This commit is contained in:
Henri Vasserman 2023-08-24 14:03:31 +03:00
parent 81ecaa4b6c
commit 238335f54f
No known key found for this signature in database
GPG key ID: 2995FC0F58B1A986
1 changed files with 2 additions and 0 deletions
Download patch file Download diff file Expand all files Collapse all files

2
common/common.cpp
View file

@ -613,9 +613,11 @@ void gpt_print_usage(int /*argc*/, char ** argv, const gpt_params & params) {
fprintf(stdout, " how to split tensors across multiple GPUs, comma-separated list of proportions, e.g. 3,1\n"); fprintf(stdout, " how to split tensors across multiple GPUs, comma-separated list of proportions, e.g. 3,1\n");
fprintf(stdout, " -mg i, --main-gpu i the GPU to use for scratch and small tensors\n"); fprintf(stdout, " -mg i, --main-gpu i the GPU to use for scratch and small tensors\n");
fprintf(stdout, " -lv, --low-vram don't allocate VRAM scratch buffer\n"); fprintf(stdout, " -lv, --low-vram don't allocate VRAM scratch buffer\n");
#ifdef GGML_USE_CUBLAS
fprintf(stdout, " -nommq, --no-mul-mat-q\n"); fprintf(stdout, " -nommq, --no-mul-mat-q\n");
fprintf(stdout, " use " GGML_CUBLAS_NAME " instead of custom mul_mat_q " GGML_CUDA_NAME " kernels.\n"); fprintf(stdout, " use " GGML_CUBLAS_NAME " instead of custom mul_mat_q " GGML_CUDA_NAME " kernels.\n");
fprintf(stdout, " Not recommended since this is both slower and uses more VRAM.\n"); fprintf(stdout, " Not recommended since this is both slower and uses more VRAM.\n");
#endif // GGML_USE_CUBLAS
#endif #endif
fprintf(stdout, " --mtest compute maximum memory usage\n"); fprintf(stdout, " --mtest compute maximum memory usage\n");
fprintf(stdout, " --export export the computation graph to 'llama.ggml'\n"); fprintf(stdout, " --export export the computation graph to 'llama.ggml'\n");