From 5144a18e6705c87241bdb169538cc4183984ef81 Mon Sep 17 00:00:00 2001
From: lexasub <empty@empty.ru>
Date: Mon, 27 Jan 2025 15:09:34 +0400
Subject: [PATCH 1/3] impl::load change map bpe_ranks to onordered map for
 reduce time of impl::load on 30%

---
 src/llama-vocab.cpp | 9 +++++++--
 1 file changed, 7 insertions(+), 2 deletions(-)

diff --git a/src/llama-vocab.cpp b/src/llama-vocab.cpp
index 0782d3a41a1f5..6a3a9d48e9635 100644
--- a/src/llama-vocab.cpp
+++ b/src/llama-vocab.cpp
@@ -1245,8 +1245,13 @@ struct llama_vocab::impl {
 
     std::vector<llama_token> cache_special_tokens;
     std::vector<std::string> cache_token_to_piece; // llama_token_to_piece(special = true);
-
-    std::map<std::pair<std::string, std::string>, int> bpe_ranks;
+    struct PairHash {
+        size_t operator()(const std::pair<std::string, std::string>& p) const {
+            return std::hash<std::string>{}(p.first) ^  //create some hash for pair
+                   (std::hash<std::string>{}(p.second) << 1);
+        }
+    };
+    std::unordered_map<std::pair<std::string, std::string>, int, PairHash> bpe_ranks;
 
     // set of all tokens that cause "end of generation"
     std::set<llama_token> special_eog_ids;

From 7c2b92423251f22e4ecbd7eba5b06c7765e68f23 Mon Sep 17 00:00:00 2001
From: lexasub <empty@empty.ru>
Date: Mon, 27 Jan 2025 15:10:54 +0400
Subject: [PATCH 2/3] llama_model_loader::init_mapping - replace new llama_mmap
 to std::make_unique<llama_mmap> for clean code & reduce (/2) time of running
 init_mappings

---
 src/llama-model-loader.cpp | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/llama-model-loader.cpp b/src/llama-model-loader.cpp
index 75073bf610ac3..05d58ad90eba9 100644
--- a/src/llama-model-loader.cpp
+++ b/src/llama-model-loader.cpp
@@ -819,7 +819,7 @@ void llama_model_loader::init_mappings(bool prefetch, llama_mlocks * mlock_mmaps
         for (const auto & file : files) {
             auto * reg = ggml_backend_dev_backend_reg(ggml_backend_dev_by_type(GGML_BACKEND_DEVICE_TYPE_CPU));
             auto * is_numa_fn = (decltype(ggml_is_numa) *) ggml_backend_reg_get_proc_address(reg, "ggml_backend_cpu_is_numa");
-            std::unique_ptr<llama_mmap> mapping(new llama_mmap(file.get(), prefetch ? -1 : 0, is_numa_fn()));
+            std::unique_ptr<llama_mmap> mapping = std::make_unique<llama_mmap>(file.get(), prefetch ? -1 : 0, is_numa_fn());
             mmaps_used.emplace_back(mapping->size(), 0);
             if (mlock_mmaps) {
                 std::unique_ptr<llama_mlock> mlock_mmap(new llama_mlock());

From 723fc665110ff1724459d32c0cac44000b725204 Mon Sep 17 00:00:00 2001
From: Diego Devesa <slarengh@gmail.com>
Date: Mon, 27 Jan 2025 14:24:08 +0100
Subject: [PATCH 3/3] Update src/llama-vocab.cpp

---
 src/llama-vocab.cpp | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/src/llama-vocab.cpp b/src/llama-vocab.cpp
index 6a3a9d48e9635..561f8bdb84010 100644
--- a/src/llama-vocab.cpp
+++ b/src/llama-vocab.cpp
@@ -1245,13 +1245,13 @@ struct llama_vocab::impl {
 
     std::vector<llama_token> cache_special_tokens;
     std::vector<std::string> cache_token_to_piece; // llama_token_to_piece(special = true);
-    struct PairHash {
-        size_t operator()(const std::pair<std::string, std::string>& p) const {
+    struct pair_hash {
+        size_t operator()(const std::pair<std::string, std::string> & p) const {
             return std::hash<std::string>{}(p.first) ^  //create some hash for pair
                    (std::hash<std::string>{}(p.second) << 1);
         }
     };
-    std::unordered_map<std::pair<std::string, std::string>, int, PairHash> bpe_ranks;
+    std::unordered_map<std::pair<std::string, std::string>, int, pair_hash> bpe_ranks;
 
     // set of all tokens that cause "end of generation"
     std::set<llama_token> special_eog_ids;