ggml : improve graph build time via hash table lookup (#2329)

* improve graph build time

* ggml_tensor : use 1 bit per flag

* use a hash table instead

1 files changed, 2 insertions, 0 deletions

diff --git a/llama.cpp b/llama.cpp
index 2d737bb..febefba 100644
--- a/llama.cpp
+++ b/llama.cpp
@@ -1714,6 +1714,8 @@ static bool llama_eval_internal(
     // run the computation
     ggml_build_forward_expand(&gf, cur);
 
+    // fprintf(stderr, "graph build time: %.3f ms (%d nodes, %d leafs)\n", (ggml_time_us() - t_start_us)/1000.0, gf.n_nodes, gf.n_leafs);
+
 #if GGML_USE_MPI
     ggml_mpi_graph_compute_pre(lctx.ctx_mpi, &gf, n_layer);
 #endif