feat: Update llama.cpp

abetlen · abetlen · commit 748c0ce05726 · 2024-02-18T21:30:36.000-05:00
diff --git a/llama_cpp/llama_cpp.py b/llama_cpp/llama_cpp.py
@@ -190,6 +190,7 @@ def _load_shared_library(lib_base_name: str):
 #     LLAMA_FTYPE_MOSTLY_Q2_K_S        = 21, // except 1d tensors
 #     LLAMA_FTYPE_MOSTLY_Q3_K_XS       = 22, // except 1d tensors
 #     LLAMA_FTYPE_MOSTLY_IQ3_XXS       = 23, // except 1d tensors
+#     LLAMA_FTYPE_MOSTLY_IQ1_S         = 24, // except 1d tensors
 
 #     LLAMA_FTYPE_GUESSED = 1024, // not specified in the model file
 # };
@@ -215,6 +216,7 @@ def _load_shared_library(lib_base_name: str):
 LLAMA_FTYPE_MOSTLY_Q2_K_S = 21
 LLAMA_FTYPE_MOSTLY_Q3_K_XS = 22
 LLAMA_FTYPE_MOSTLY_IQ3_XXS = 23
+LLAMA_FTYPE_MOSTLY_IQ1_S = 24
 LLAMA_FTYPE_GUESSED = 1024
 
 # enum llama_rope_scaling_type {
diff --git a/vendor/llama.cpp b/vendor/llama.cpp
@@ -1 +1 @@
-Subproject commit 5bf2b94dd4fb74378b78604023b31512fec55f8f
+Subproject commit a0c2dad9d43456c677e205c6240a5f8afb0121ac
