llama : fix worst case graph build
This commit is contained in:
parent
fa0e677820
commit
daf4c6d360
3 changed files with 100 additions and 81 deletions
|
@ -977,7 +977,7 @@ int main(int argc, char ** argv) {
|
|||
|
||||
test t(inst, lmodel, ctx);
|
||||
|
||||
llama_kv_cache_keep_seq(ctx, -1);
|
||||
llama_kv_cache_rm_tokens(ctx, -1, -1);
|
||||
|
||||
// warmup run
|
||||
if (t.n_prompt > 0) {
|
||||
|
@ -988,7 +988,7 @@ int main(int argc, char ** argv) {
|
|||
}
|
||||
|
||||
for (int i = 0; i < params.reps; i++) {
|
||||
llama_kv_cache_keep_seq(ctx, -1);
|
||||
llama_kv_cache_rm_tokens(ctx, -1, -1);
|
||||
|
||||
uint64_t t_start = get_time_ns();
|
||||
if (t.n_prompt > 0) {
|
||||
|
|
Loading…
Add table
Add a link
Reference in a new issue