From bd76198618814ce0d34be2859a0908b63f98f0f5 Mon Sep 17 00:00:00 2001 From: Yoshi Suhara Date: Sat, 10 Aug 2024 22:28:16 -0700 Subject: [PATCH] Remove mutable variable --- src/llama.cpp | 3 --- 1 file changed, 3 deletions(-) diff --git a/src/llama.cpp b/src/llama.cpp index 06e0d6473..e3b07b80e 100644 --- a/src/llama.cpp +++ b/src/llama.cpp @@ -13823,9 +13823,6 @@ struct llm_build_context { struct ggml_cgraph * build_nemotron() { struct ggml_cgraph * gf = ggml_new_graph_custom(ctx0, llama_model_max_nodes(model), false); - // mutable variable, needed during the last layer of the computation to skip unused tokens - int32_t n_tokens = this->n_tokens; - const int64_t n_embd_head = hparams.n_embd_head_v; GGML_ASSERT(n_embd_head == hparams.n_embd_head_k); //GGML_ASSERT(n_embd_head == hparams.n_rot);