sync: update ggml

Author: leejet

ggml

@@ -846,7 +846,7 @@ struct GGMLModule {
             return NULL;
         }
         // it's performing a compute, check if backend isn't cpu
-        if (!ggml_backend_is_cpu(backend) && tensor->backend == GGML_BACKEND_CPU) {
+        if (!ggml_backend_is_cpu(backend) && tensor->backend == GGML_BACKEND_TYPE_CPU) {
             // pass input tensors to gpu memory
             auto backend_tensor = ggml_dup_tensor(compute_ctx, tensor);
 

ggml_extend.hpp

@@ -65,17 +65,20 @@ enum sd_type_t {
     SD_TYPE_Q8_0 = 8,
     SD_TYPE_Q8_1 = 9,
     // k-quantizations
-    SD_TYPE_Q2_K    = 10,
-    SD_TYPE_Q3_K    = 11,
-    SD_TYPE_Q4_K    = 12,
-    SD_TYPE_Q5_K    = 13,
-    SD_TYPE_Q6_K    = 14,
-    SD_TYPE_Q8_K    = 15,
+    SD_TYPE_Q2_K = 10,
+    SD_TYPE_Q3_K = 11,
+    SD_TYPE_Q4_K = 12,
+    SD_TYPE_Q5_K = 13,
+    SD_TYPE_Q6_K = 14,
+    SD_TYPE_Q8_K = 15,
     SD_TYPE_IQ2_XXS = 16,
     SD_TYPE_IQ2_XS  = 17,
     SD_TYPE_IQ3_XXS = 18,
     SD_TYPE_IQ1_S   = 19,
     SD_TYPE_IQ4_NL  = 20,
+    SD_TYPE_IQ3_S   = 21,
+    SD_TYPE_IQ2_S   = 22,
+    SD_TYPE_IQ4_XS  = 23,
     SD_TYPE_I8,
     SD_TYPE_I16,
     SD_TYPE_I32,