sync : llama.cpp (ggml/0)

diff --git a/examples/common-ggml.cpp b/examples/common-ggml.cpp
index f1c7f6ef1874e8fedb9e1472a70d86e2fb5bea45..7a3607546837bb9a60099f9662b53ad1afe57777 100644 (file)
--- a/examples/common-ggml.cpp
+++ b/examples/common-ggml.cpp
@@ -65,6 +65,7 @@ bool ggml_common_quantize_0(
         case GGML_FTYPE_MOSTLY_IQ2_XXS:
         case GGML_FTYPE_MOSTLY_IQ2_XS:
         case GGML_FTYPE_MOSTLY_IQ3_XXS:
+        case GGML_FTYPE_MOSTLY_IQ3_S:
         case GGML_FTYPE_MOSTLY_IQ1_S:
         case GGML_FTYPE_MOSTLY_IQ4_NL:
                 {
@@ -199,6 +200,7 @@ bool ggml_common_quantize_0(
                 case GGML_TYPE_IQ2_XXS:
                 case GGML_TYPE_IQ2_XS:
                 case GGML_TYPE_IQ3_XXS:
+                case GGML_TYPE_IQ3_S:
                 case GGML_TYPE_IQ1_S:
                 case GGML_TYPE_IQ4_NL:
                 case GGML_TYPE_COUNT: