feat: Update llama.cpp

Author: abetlen

llama_cpp/llama_cpp.py

@@ -198,13 +198,15 @@ def byref(obj: CtypesCData, offset: Optional[int] = None) -> CtypesRef[CtypesCDa
 
 
 # enum llama_vocab_type {
-#     LLAMA_VOCAB_TYPE_SPM = 0, // SentencePiece
-#     LLAMA_VOCAB_TYPE_BPE = 1, // Byte Pair Encoding
-#     LLAMA_VOCAB_TYPE_WPM = 2, // WordPiece
+#     LLAMA_VOCAB_TYPE_NONE = 0, // For models without vocab
+#     LLAMA_VOCAB_TYPE_SPM  = 1, // SentencePiece
+#     LLAMA_VOCAB_TYPE_BPE  = 2, // Byte Pair Encoding
+#     LLAMA_VOCAB_TYPE_WPM  = 3, // WordPiece
 # };
-LLAMA_VOCAB_TYPE_SPM = 0
-LLAMA_VOCAB_TYPE_BPE = 1
-LLAMA_VOCAB_TYPE_WPM = 2
+LLAMA_VOCAB_TYPE_NONE = 0
+LLAMA_VOCAB_TYPE_SPM = 1
+LLAMA_VOCAB_TYPE_BPE = 2
+LLAMA_VOCAB_TYPE_WPM = 3
 
 
 # // note: these values should be synchronized with ggml_rope