Function ggml_graph_compute_helper

llama.cpp:749–758

//

static void ggml_graph_compute_helper(std::vector<uint8_t> & buf, ggml_cgraph * graph, int n_threads) {
    struct ggml_cplan plan = ggml_graph_plan(graph, n_threads);

    if (plan.work_size > 0) {
        buf.resize(plan.work_size);
        plan.work_data = buf.data();
    }

    ggml_graph_compute(graph, &plan);
}

//
// llama helpers