From 87f4102a70b54c002ad37f966717e4f13a94410a Mon Sep 17 00:00:00 2001 From: Georgi Gerganov Date: Mon, 27 Nov 2023 21:21:23 +0200 Subject: [PATCH] llama : revert n_threads_batch logic ggml-ci --- llama.cpp | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/llama.cpp b/llama.cpp index 9aa337a8d..cb544228b 100644 --- a/llama.cpp +++ b/llama.cpp @@ -5433,7 +5433,7 @@ static int llama_decode_internal( GGML_ASSERT(n_tokens <= n_batch); - int n_threads = n_tokens < 32 ? cparams.n_threads : cparams.n_threads_batch; + int n_threads = n_tokens == 1 ? cparams.n_threads : cparams.n_threads_batch; GGML_ASSERT((!batch.token && batch.embd) || (batch.token && !batch.embd)); // NOLINT const int64_t t_start_us = ggml_time_us();