From ba4cf5c0bf37a729d29e899dadf14541cddd23d4 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ji=C5=99=C3=AD=20Podiv=C3=ADn?= <66251151+jpodivin@users.noreply.github.com> Date: Fri, 17 Nov 2023 16:19:16 +0100 Subject: [PATCH] train : move number of gpu layers argument parsing to common/train.cpp (#4074) - introduces help entry for the argument - cuts '--gpu-layers' form in order to simplify usage and documentation. Signed-off-by: Jiri Podivin Co-authored-by: Jiri Podivin --- common/train.cpp | 12 ++++++++++++ examples/finetune/finetune.cpp | 11 ----------- 2 files changed, 12 insertions(+), 11 deletions(-) diff --git a/common/train.cpp b/common/train.cpp index 964b156b5..773e2c59c 100644 --- a/common/train.cpp +++ b/common/train.cpp @@ -1136,6 +1136,7 @@ void print_common_train_usage(int /*argc*/, char ** /*argv*/, const struct train fprintf(stderr, " --adam-beta2 N AdamW beta2 in interval [0,1). How much to smooth the second moment of gradients. (default %f)\n", params->adam_beta2); fprintf(stderr, " --adam-gclip N AdamW gradient clipping. Disabled when zero. (default %f)\n", params->adam_gclip); fprintf(stderr, " --adam-epsf N AdamW epsilon for convergence test. Disabled when <= zero. (default %f)\n", params->adam_eps_f); + fprintf(stderr, " -ngl N, --n-gpu-layers N Number of model layers to offload to GPU (default %d)", params->n_gpu_layers); fprintf(stderr, "\n"); } @@ -1355,6 +1356,17 @@ bool consume_common_train_arg( return true; } params->adam_gclip = std::stof(argv[i]); + } else if (arg == "-ngl" || arg == "--n-gpu-layers") { + if (++i >= argc) { + *invalid_param = true; + return true; + } +#ifdef LLAMA_SUPPORTS_GPU_OFFLOAD + params->n_gpu_layers = std::stoi(argv[i]); +#else + fprintf(stderr, "warning: not compiled with GPU offload support, --n-gpu-layers option will be ignored\n"); + fprintf(stderr, "warning: see main README.md for information on enabling GPU BLAS support\n"); +#endif } else if (arg == "-h" || arg == "--help") { params->print_usage = true; return true; diff --git a/examples/finetune/finetune.cpp b/examples/finetune/finetune.cpp index 7fecce254..af46e44a6 100644 --- a/examples/finetune/finetune.cpp +++ b/examples/finetune/finetune.cpp @@ -1460,17 +1460,6 @@ static bool train_params_parse(int argc, char ** argv, struct train_params * par } params->n_rank_w3 = std::stoi(argv[i]); params->custom_n_rank_w3 = true; - } else if (arg == "--gpu-layers" || arg == "-ngl" || arg == "--n-gpu-layers") { - if (++i >= argc) { - invalid_param = true; - break; - } -#ifdef LLAMA_SUPPORTS_GPU_OFFLOAD - params->common.n_gpu_layers = std::stoi(argv[i]); -#else - fprintf(stderr, "warning: not compiled with GPU offload support, --n-gpu-layers option will be ignored\n"); - fprintf(stderr, "warning: see main README.md for information on enabling GPU BLAS support\n"); -#endif } else { fprintf(stderr, "error: unknown argument: %s\n", arg.c_str()); train_print_usage(argc, argv, &default_params);