convert: add MiniCPM5 tokenizer support (#23384)

Add minicpm5 pre-tokenizer hash via convert_hf_to_gguf_update.py and
implement hardcoded regex handling in llama-vocab.cpp, consistent with
other BPE pre-tokenizers.

Co-authored-by: zhangtao <zhangtao2@modelbest.cn>

conversion/base.py

@@ -1625,6 +1625,9 @@ class TextModel(ModelBase):
         if chkhsh == "f728162c1315c26e40249849799b4ba3fe584c32084b4795b03eb295e63cb5af":
             # ref: https://huggingface.co/lewtun/talkie-1930-13b-it-hf
             res = "talkie"
+        if chkhsh == "36f3066e97b7f3994b379aaacde306c1444c6ae84e81a5ae3cd2b7ed3b8c42d4":
+            # ref: https://huggingface.co/openbmb/MiniCPM5-1B
+            res = "minicpm5"
 
         if res is None:
             logger.warning("\n")