feat: Update llama.cpp

Author: abetlen

llama_cpp/llama_cpp.py

@@ -245,6 +245,8 @@ def byref(obj: CtypesCData, offset: Optional[int] = None) -> CtypesRef[CtypesCDa
 #     LLAMA_FTYPE_MOSTLY_IQ3_XXS       = 23, // except 1d tensors
 #     LLAMA_FTYPE_MOSTLY_IQ1_S         = 24, // except 1d tensors
 #     LLAMA_FTYPE_MOSTLY_IQ4_NL        = 25, // except 1d tensors
+#     LLAMA_FTYPE_MOSTLY_IQ3_S         = 26, // except 1d tensors
+#     LLAMA_FTYPE_MOSTLY_IQ3_M         = 27, // except 1d tensors
 
 #     LLAMA_FTYPE_GUESSED = 1024, // not specified in the model file
 # };
@@ -272,6 +274,8 @@ def byref(obj: CtypesCData, offset: Optional[int] = None) -> CtypesRef[CtypesCDa
 LLAMA_FTYPE_MOSTLY_IQ3_XXS = 23
 LLAMA_FTYPE_MOSTLY_IQ1_S = 24
 LLAMA_FTYPE_MOSTLY_IQ4_NL = 25
+LLAMA_FTYPE_MOSTLY_IQ3_S = 26
+LLAMA_FTYPE_MOSTLY_IQ3_M = 27
 LLAMA_FTYPE_GUESSED = 1024
 
 # enum llama_rope_scaling_type {