minor : clean-up

ggml-ci

Author: ggerganov

src/llama-graph.cpp

@@ -1290,10 +1290,6 @@ ggml_tensor * llm_graph_context::build_attn(
         cur = build_lora_mm(wo, cur);
     }
 
-    if (wo_b) {
-        //cb(cur, "kqv_wo", il);
-    }
-
     if (wo_b) {
         cur = ggml_add(ctx0, cur, wo_b);
     }

src/llama-kv-cache.cpp

@@ -129,7 +129,7 @@ llama_kv_cache_unified::llama_kv_cache_unified(
         const size_t memory_size_k = size_k_bytes();
         const size_t memory_size_v = size_v_bytes();
 
-        LLAMA_LOG_INFO("%s: size = %7.2f (%6d cells, %3d layers) MiB, K (%s): %7.2f MiB, V (%s): %7.2f MiB\n", __func__,
+        LLAMA_LOG_INFO("%s: size = %7.2f MiB (%6d cells, %3d layers), K (%s): %7.2f MiB, V (%s): %7.2f MiB\n", __func__,
                 (float)(memory_size_k + memory_size_v) / (1024.0f * 1024.0f), kv_size, (int) layers.size(),
                 ggml_type_name(type_k), (float)memory_size_k / (1024.0f * 1024.0f),
                 ggml_type_name(type_v), (float)memory_size_v / (1024.0f * 1024.0f));