Merge branch 'refactor-jina-rename' of https://github.com/JoanFM/llama.cpp into feat-jina-embeddings-v2-zh

Author: JoanFM

llama.cpp

@@ -4423,6 +4423,8 @@ static void llm_load_vocab(
                 vocab.type_pre = LLAMA_VOCAB_PRE_TYPE_STARCODER;
             } else if (
                     tokenizer_pre == "gpt-2"   ||
+                    tokenizer_pre == "jina-es" ||
+                    tokenizer_pre == "jina-de" ||
                     tokenizer_pre == "jina-v2-es" ||
                     tokenizer_pre == "jina-v2-de") {
                 vocab.type_pre = LLAMA_VOCAB_PRE_TYPE_GPT2;

unicode-data.h

@@ -14,4 +14,4 @@ extern const std::vector<std::pair<uint32_t, uint32_t>> unicode_ranges_punctuati
 extern const std::vector<std::pair<uint32_t, uint32_t>> unicode_ranges_symbol;
 extern const std::vector<std::pair<uint32_t, uint32_t>> unicode_ranges_control;
 extern const std::multimap<uint32_t, uint32_t>          unicode_map_nfd;
-extern const std::map<char32_t, char32_t>               unicode_map_lowercase;
+extern const std::map<char32_t, char32_t>               unicode_map_lowercase;