Skip to content

Commit 1c154e9

Browse files
committed
lazy fix for llama-bench (runs without pp_threads support)
1 parent a129a31 commit 1c154e9

File tree

1 file changed

+2
-2
lines changed

1 file changed

+2
-2
lines changed

examples/llama-bench/llama-bench.cpp

Lines changed: 2 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -853,15 +853,15 @@ static void test_prompt(llama_context * ctx, int n_prompt, int n_past, int n_bat
853853
int n_processed = 0;
854854
while (n_processed < n_prompt) {
855855
int n_tokens = std::min(n_prompt - n_processed, n_batch);
856-
llama_eval(ctx, tokens.data(), n_tokens, n_past + n_processed, n_threads);
856+
llama_eval(ctx, tokens.data(), n_tokens, n_past + n_processed, n_threads, n_threads);
857857
n_processed += n_tokens;
858858
}
859859
}
860860

861861
static void test_gen(llama_context * ctx, int n_gen, int n_past, int n_threads) {
862862
llama_token token = llama_token_bos();
863863
for (int i = 0; i < n_gen; i++) {
864-
llama_eval(ctx, &token, 1, n_past + i, n_threads);
864+
llama_eval(ctx, &token, 1, n_past + i, n_threads, n_threads);
865865
}
866866
}
867867

0 commit comments

Comments
 (0)