From 137971b2e1e1da7b885a5344718c3a342375ee93 Mon Sep 17 00:00:00 2001 From: Jared Van Bortel Date: Fri, 2 Feb 2024 10:42:42 -0500 Subject: [PATCH] llama : store mapped names as const char * --- llama.cpp | 18 +++++++++--------- 1 file changed, 9 insertions(+), 9 deletions(-) diff --git a/llama.cpp b/llama.cpp index aaee6f7d0..4787a92fe 100644 --- a/llama.cpp +++ b/llama.cpp @@ -208,7 +208,7 @@ enum llm_arch { LLM_ARCH_UNKNOWN, }; -static std::map LLM_ARCH_NAMES = { +static std::map LLM_ARCH_NAMES = { { LLM_ARCH_LLAMA, "llama" }, { LLM_ARCH_FALCON, "falcon" }, { LLM_ARCH_GPT2, "gpt2" }, @@ -285,7 +285,7 @@ enum llm_kv { LLM_KV_TOKENIZER_RWKV, }; -static std::map LLM_KV_NAMES = { +static std::map LLM_KV_NAMES = { { LLM_KV_GENERAL_ARCHITECTURE, "general.architecture" }, { LLM_KV_GENERAL_QUANTIZATION_VERSION, "general.quantization_version" }, { LLM_KV_GENERAL_ALIGNMENT, "general.alignment" }, @@ -346,7 +346,7 @@ struct LLM_KV { llm_arch arch; std::string operator()(llm_kv kv) const { - return ::format(LLM_KV_NAMES[kv].c_str(), LLM_ARCH_NAMES[arch].c_str()); + return ::format(LLM_KV_NAMES[kv], LLM_ARCH_NAMES[arch]); } }; @@ -747,7 +747,7 @@ struct LLM_TN { // gguf helpers // -static std::map LLAMA_ROPE_SCALING_TYPES = { +static std::map LLAMA_ROPE_SCALING_TYPES = { { LLAMA_ROPE_SCALING_NONE, "none" }, { LLAMA_ROPE_SCALING_LINEAR, "linear" }, { LLAMA_ROPE_SCALING_YARN, "yarn" }, @@ -2701,7 +2701,7 @@ struct llama_model_loader { // load LLaMA models // -static std::string llama_model_arch_name(llm_arch arch) { +static const char * llama_model_arch_name(llm_arch arch) { auto it = LLM_ARCH_NAMES.find(arch); if (it == LLM_ARCH_NAMES.end()) { return "unknown"; @@ -3310,11 +3310,11 @@ static void llm_load_print_meta(llama_model_loader & ml, llama_model & model) { const auto & hparams = model.hparams; const auto & vocab = model.vocab; - const auto rope_scaling_type = LLAMA_ROPE_SCALING_TYPES.at(hparams.rope_scaling_type_train); + const char * rope_scaling_type = LLAMA_ROPE_SCALING_TYPES.at(hparams.rope_scaling_type_train); // hparams LLAMA_LOG_INFO("%s: format = %s\n", __func__, llama_file_version_name(ml.fver)); - LLAMA_LOG_INFO("%s: arch = %s\n", __func__, LLM_ARCH_NAMES.at(model.arch).c_str()); + LLAMA_LOG_INFO("%s: arch = %s\n", __func__, LLM_ARCH_NAMES.at(model.arch)); LLAMA_LOG_INFO("%s: vocab type = %s\n", __func__, llama_model_vocab_type_name(vocab.type)); LLAMA_LOG_INFO("%s: n_vocab = %u\n", __func__, hparams.n_vocab); LLAMA_LOG_INFO("%s: n_merges = %u\n", __func__, (int) vocab.bpe_ranks.size()); @@ -3336,7 +3336,7 @@ static void llm_load_print_meta(llama_model_loader & ml, llama_model & model) { LLAMA_LOG_INFO("%s: n_ff = %u\n", __func__, hparams.n_ff); LLAMA_LOG_INFO("%s: n_expert = %u\n", __func__, hparams.n_expert); LLAMA_LOG_INFO("%s: n_expert_used = %u\n", __func__, hparams.n_expert_used); - LLAMA_LOG_INFO("%s: rope scaling = %s\n", __func__, rope_scaling_type.c_str()); + LLAMA_LOG_INFO("%s: rope scaling = %s\n", __func__, rope_scaling_type); LLAMA_LOG_INFO("%s: freq_base_train = %.1f\n", __func__, hparams.rope_freq_base_train); LLAMA_LOG_INFO("%s: freq_scale_train = %g\n", __func__, hparams.rope_freq_scale_train); LLAMA_LOG_INFO("%s: n_yarn_orig_ctx = %u\n", __func__, hparams.n_yarn_orig_ctx); @@ -10735,7 +10735,7 @@ int32_t llama_model_meta_val_str_by_index(const struct llama_model * model, int3 int32_t llama_model_desc(const struct llama_model * model, char * buf, size_t buf_size) { return snprintf(buf, buf_size, "%s %s %s", - llama_model_arch_name(model->arch).c_str(), + llama_model_arch_name(model->arch), llama_model_type_name(model->type), llama_model_ftype_name(model->ftype).c_str()); }