From a6120326b20a04295cb0d5f1405d54953b91d699 Mon Sep 17 00:00:00 2001 From: Yui Date: Mon, 8 Apr 2024 09:18:00 +0200 Subject: [PATCH] Update common/common.h (suggested by JohannesGaessler) MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Co-authored-by: Johannes Gäßler --- common/common.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/common/common.h b/common/common.h index 82e7c78ae..c8e97b330 100644 --- a/common/common.h +++ b/common/common.h @@ -62,7 +62,7 @@ struct gpt_params { int32_t n_parallel = 1; // number of parallel sequences to decode int32_t n_sequences = 1; // number of sequences to decode float p_split = 0.1f; // speculative decoding split probability - int32_t n_gpu_layers = -1; // number of layers to store in VRAM (-1 - use default, -2 - automatically determine) + int32_t n_gpu_layers = -1; // number of layers to store in VRAM (-1 - use default, -2 - determine automatically) int32_t n_gpu_layers_draft = -1; // number of layers to store in VRAM for the draft model (-1 - use default) llama_split_mode split_mode = LLAMA_SPLIT_MODE_LAYER; // how to split the model across GPUs int32_t main_gpu = 0; // the GPU that is used for scratch and small tensors