vrr/prima.cpp
2
0
Fork 0
mirror of https://github.com/Lizonghang/prima.cpp.git synced 2025-09-06 21:59:03 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

ignore the first eval to make time test more accurate

Browse source
This commit is contained in:
Lizonghang 2025-01-30 11:12:26 +04:00
parent cd758247e6
commit f9b4c46b74
1 changed files with 5 additions and 4 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

9
src/llama.cpp
View file

@ -111,7 +111,7 @@ struct Timer {
~Timer() { ~Timer() {
if (enable_timer) { if (enable_timer) {
int64_t end_time = ggml_time_us(); int64_t end_time = ggml_time_us();
LLAMA_LOG_INFO("Time to run %s: %" PRId64 " ms\n", name, (end_time - start_time) / 1000); LLAMA_LOG_INFO("\nTime to run %s: %" PRId64 " ms: ", name, (end_time - start_time) / 1000);
} }
} }
}; };
@ -3344,7 +3344,7 @@ struct llama_context {
mutable int64_t n_queued_tokens = 0; mutable int64_t n_queued_tokens = 0;
mutable int32_t n_p_eval = 0; // number of tokens in eval calls for the prompt (with batch size > 1) mutable int32_t n_p_eval = 0; // number of tokens in eval calls for the prompt (with batch size > 1)
mutable int32_t n_eval = 0; // number of eval calls mutable int32_t n_eval = -1; // number of eval calls, set to -1 to ignore the first eval
// host buffer for the model output (logits and embeddings) // host buffer for the model output (logits and embeddings)
ggml_backend_buffer_t buf_output = nullptr; ggml_backend_buffer_t buf_output = nullptr;
@ -22715,7 +22715,7 @@ void llama_synchronize(struct llama_context * ctx) {
// add the evaluation to the stats // add the evaluation to the stats
if (ctx->n_queued_tokens == 1) { if (ctx->n_queued_tokens == 1) {
if (!ctx->cparams.no_perf) { if (!ctx->cparams.no_perf && ctx->n_eval >= 0) { // ignore the first two evals due to preheat
ctx->t_eval_us += ggml_time_us() - ctx->t_compute_start_us; ctx->t_eval_us += ggml_time_us() - ctx->t_compute_start_us;
} }
ctx->n_eval++; ctx->n_eval++;
@ -23362,7 +23362,8 @@ void llama_perf_context_print(const struct llama_context * ctx) {
void llama_perf_context_reset(struct llama_context * ctx) { void llama_perf_context_reset(struct llama_context * ctx) {
ctx->t_start_us = ggml_time_us(); ctx->t_start_us = ggml_time_us();
ctx->t_eval_us = ctx->n_eval = 0; ctx->t_eval_us = 0;
ctx->n_eval = -1; // set to -1 to ignore the first eval due to preheat
ctx->t_p_eval_us = ctx->n_p_eval = 0; ctx->t_p_eval_us = ctx->n_p_eval = 0;
} }