feat: Update llama.cpp

Author: abetlen

llama_cpp/llama_cpp.py

@@ -296,7 +296,9 @@ def byref(obj: CtypesCData, offset: Optional[int] = None) -> CtypesRef[CtypesCDa
 #     LLAMA_VOCAB_PRE_TYPE_GPT2           = 7,
 #     LLAMA_VOCAB_PRE_TYPE_REFACT         = 8,
 #     LLAMA_VOCAB_PRE_TYPE_COMMAND_R      = 9,
-#     LLAMA_VOCAB_PRE_TYPE_OLMO           = 10,
+#     LLAMA_VOCAB_PRE_TYPE_QWEN2          = 10,
+#     LLAMA_VOCAB_PRE_TYPE_OLMO           = 11,
+#     LLAMA_VOCAB_PRE_TYPE_DBRX           = 12,
 # };
 LLAMA_VOCAB_PRE_TYPE_DEFAULT = 0
 LLAMA_VOCAB_PRE_TYPE_LLAMA3 = 1
@@ -308,7 +310,9 @@ def byref(obj: CtypesCData, offset: Optional[int] = None) -> CtypesRef[CtypesCDa
 LLAMA_VOCAB_PRE_TYPE_GPT2 = 7
 LLAMA_VOCAB_PRE_TYPE_REFACT = 8
 LLAMA_VOCAB_PRE_TYPE_COMMAND_R = 9
-LLAMA_VOCAB_PRE_TYPE_OLMO = 10
+LLAMA_VOCAB_PRE_TYPE_QWEN2 = 10
+LLAMA_VOCAB_PRE_TYPE_OLMO = 11
+LLAMA_VOCAB_PRE_TYPE_DBRX = 12
 
 
 # // note: these values should be synchronized with ggml_rope
@@ -377,6 +381,7 @@ def byref(obj: CtypesCData, offset: Optional[int] = None) -> CtypesRef[CtypesCDa
 #     LLAMA_FTYPE_MOSTLY_IQ2_M         = 29, // except 1d tensors
 #     LLAMA_FTYPE_MOSTLY_IQ4_XS        = 30, // except 1d tensors
 #     LLAMA_FTYPE_MOSTLY_IQ1_M         = 31, // except 1d tensors
+#     LLAMA_FTYPE_MOSTLY_BF16          = 32, // except 1d tensors
 
 #     LLAMA_FTYPE_GUESSED = 1024, // not specified in the model file
 # };
@@ -409,6 +414,8 @@ def byref(obj: CtypesCData, offset: Optional[int] = None) -> CtypesRef[CtypesCDa
 LLAMA_FTYPE_MOSTLY_IQ2_S = 28
 LLAMA_FTYPE_MOSTLY_IQ2_M = 29
 LLAMA_FTYPE_MOSTLY_IQ4_XS = 30
+LLAMA_FTYPE_MOSTLY_IQ1_M = 31
+LLAMA_FTYPE_MOSTLY_BF16 = 32
 LLAMA_FTYPE_GUESSED = 1024
 
 # enum llama_rope_scaling_type {