diff options
Diffstat (limited to 'examples/server')
| -rw-r--r-- | examples/server/README.md | 6 | ||||
| -rw-r--r-- | examples/server/server.cpp | 16 |
2 files changed, 11 insertions, 11 deletions
diff --git a/examples/server/README.md b/examples/server/README.md index 49121a46..aadc73b4 100644 --- a/examples/server/README.md +++ b/examples/server/README.md @@ -25,9 +25,9 @@ The project is under active development, and we are [looking for feedback and co - `-hff FILE, --hf-file FILE`: Hugging Face model file (default: unused). - `-a ALIAS`, `--alias ALIAS`: Set an alias for the model. The alias will be returned in API responses. - `-c N`, `--ctx-size N`: Set the size of the prompt context. The default is 512, but LLaMA models were built with a context of 2048, which will provide better results for longer input/inference. The size may differ in other models, for example, baichuan models were build with a context of 4096. -- `-ngl N`, `--n-gpu-layers N`: When compiled with appropriate support (currently CLBlast or cuBLAS), this option allows offloading some layers to the GPU for computation. Generally results in increased performance. -- `-mg i, --main-gpu i`: When using multiple GPUs this option controls which GPU is used for small tensors for which the overhead of splitting the computation across all GPUs is not worthwhile. The GPU in question will use slightly more VRAM to store a scratch buffer for temporary results. By default GPU 0 is used. Requires cuBLAS. -- `-ts SPLIT, --tensor-split SPLIT`: When using multiple GPUs this option controls how large tensors should be split across all GPUs. `SPLIT` is a comma-separated list of non-negative values that assigns the proportion of data that each GPU should get in order. For example, "3,2" will assign 60% of the data to GPU 0 and 40% to GPU 1. By default the data is split in proportion to VRAM but this may not be optimal for performance. Requires cuBLAS. +- `-ngl N`, `--n-gpu-layers N`: When compiled with GPU support, this option allows offloading some layers to the GPU for computation. Generally results in increased performance. +- `-mg i, --main-gpu i`: When using multiple GPUs this option controls which GPU is used for small tensors for which the overhead of splitting the computation across all GPUs is not worthwhile. The GPU in question will use slightly more VRAM to store a scratch buffer for temporary results. By default GPU 0 is used. +- `-ts SPLIT, --tensor-split SPLIT`: When using multiple GPUs this option controls how large tensors should be split across all GPUs. `SPLIT` is a comma-separated list of non-negative values that assigns the proportion of data that each GPU should get in order. For example, "3,2" will assign 60% of the data to GPU 0 and 40% to GPU 1. By default the data is split in proportion to VRAM but this may not be optimal for performance. - `-b N`, `--batch-size N`: Set the batch size for prompt processing. Default: `2048`. - `-ub N`, `--ubatch-size N`: physical maximum batch size. Default: `512`. - `--memory-f32`: Use 32-bit floats instead of 16-bit floats for memory key+value. Not recommended. diff --git a/examples/server/server.cpp b/examples/server/server.cpp index 338e60f2..c4c545c3 100644 --- a/examples/server/server.cpp +++ b/examples/server/server.cpp @@ -2510,15 +2510,15 @@ static void server_params_parse(int argc, char ** argv, server_params & sparams, invalid_param = true; break; } -#ifndef GGML_USE_CUBLAS - fprintf(stderr, "warning: llama.cpp was compiled without cuBLAS. Setting the split mode has no effect.\n"); -#endif // GGML_USE_CUBLAS +#ifndef GGML_USE_CUDA + fprintf(stderr, "warning: llama.cpp was compiled without CUDA. Setting the split mode has no effect.\n"); +#endif // GGML_USE_CUDA } else if (arg == "--tensor-split" || arg == "-ts") { if (++i >= argc) { invalid_param = true; break; } -#if defined(GGML_USE_CUBLAS) || defined(GGML_USE_SYCL) +#if defined(GGML_USE_CUDA) || defined(GGML_USE_SYCL) std::string arg_next = argv[i]; // split string by , and / @@ -2535,17 +2535,17 @@ static void server_params_parse(int argc, char ** argv, server_params & sparams, } } #else - LOG_WARNING("llama.cpp was compiled without cuBLAS. It is not possible to set a tensor split.\n", {}); -#endif // GGML_USE_CUBLAS + LOG_WARNING("llama.cpp was compiled without CUDA. It is not possible to set a tensor split.\n", {}); +#endif // GGML_USE_CUDA } else if (arg == "--main-gpu" || arg == "-mg") { if (++i >= argc) { invalid_param = true; break; } -#if defined(GGML_USE_CUBLAS) || defined(GGML_USE_SYCL) +#if defined(GGML_USE_CUDA) || defined(GGML_USE_SYCL) params.main_gpu = std::stoi(argv[i]); #else - LOG_WARNING("llama.cpp was compiled without cuBLAS. It is not possible to set a main GPU.", {}); + LOG_WARNING("llama.cpp was compiled without CUDA. It is not possible to set a main GPU.", {}); #endif } else if (arg == "--lora") { if (++i >= argc) { |