llama : give up fix for uint64_t format type by casting to size_t

Otherwise some platforms use "%lu" and others "%llu" for uint64_t,
and I'm not sure how to handle that.

Author: compilade

src/llama.cpp

@@ -17710,7 +17710,7 @@ struct llama_data_read_context {
             const size_t k_size_row = ggml_row_size(kv_self.k_l[il]->type, n_embd_k_gqa);
             if (k_size_row != k_size_row_ref) {
                 // llama_kv_cache_seq_rm(kv_self, dest_seq_id, -1, -1);
-                LLAMA_LOG_ERROR("%s: mismatched key row size (%zu != %zu, layer %d)\n", __func__, k_size_row, k_size_row_ref, il);
+                LLAMA_LOG_ERROR("%s: mismatched key row size (%zu != %zu, layer %d)\n", __func__, k_size_row, (size_t) k_size_row_ref, il);
                 return false;
             }
 
@@ -17739,7 +17739,7 @@ struct llama_data_read_context {
                 read_to(&v_size_row_ref, sizeof(v_size_row_ref));
                 const size_t v_size_row = ggml_row_size(kv_self.v_l[il]->type, n_embd_v_gqa);
                 if (v_size_row != v_size_row_ref) {
-                    LLAMA_LOG_ERROR("%s: mismatched value row size (%zu != %zu, layer %d)\n", __func__, v_size_row, v_size_row_ref, il);
+                    LLAMA_LOG_ERROR("%s: mismatched value row size (%zu != %zu, layer %d)\n", __func__, v_size_row, (size_t) v_size_row_ref, il);
                     return false;
                 }
 
@@ -17767,7 +17767,7 @@ struct llama_data_read_context {
                 read_to(&v_size_el_ref, sizeof(v_size_el_ref));
                 const size_t v_size_el = ggml_type_size(kv_self.v_l[il]->type);
                 if (v_size_el != v_size_el_ref) {
-                    LLAMA_LOG_ERROR("%s: mismatched value element size (%zu != %llu, layer %d)\n", __func__, v_size_el, v_size_el_ref, il);
+                    LLAMA_LOG_ERROR("%s: mismatched value element size (%zu != %zu, layer %d)\n", __func__, v_size_el, (size_t) v_size_el_ref, il);
                     return false;
                 }