Skip to content

server : export max observed n_past value #15361

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 3 commits into from
Aug 17, 2025
+19 −0
Merged

server : export max observed n_past value #15361

Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
a5bb98f
[server] export max observed n_past value
okuvshynov Aug 16, 2025
caf9dc7
Merge branch 'ggml-org:master' into n_past_max
okuvshynov Aug 17, 2025
abd0737
Merge branch 'ggml-org:master' into n_past_max
okuvshynov Aug 17, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
19 changes: 19 additions & 0 deletions tools/server/server.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1201,6 +1201,8 @@ struct server_task_result_metrics : server_task_result {
uint64_t n_tokens_predicted_total = 0;
uint64_t t_tokens_generation_total = 0;

uint64_t n_past_max = 0;

uint64_t n_prompt_tokens_processed = 0;
uint64_t t_prompt_processing = 0;

Expand All @@ -1226,6 +1228,8 @@ struct server_task_result_metrics : server_task_result {
{ "n_tokens_predicted_total", n_tokens_predicted_total },
{ "t_prompt_processing_total", t_prompt_processing_total },

{ "n_past_max", n_past_max },

{ "n_prompt_tokens_processed", n_prompt_tokens_processed },
{ "t_prompt_processing", t_prompt_processing },
{ "n_tokens_predicted", n_tokens_predicted },
Expand Down Expand Up @@ -1587,6 +1591,8 @@ struct server_metrics {
uint64_t n_tokens_predicted_total = 0;
uint64_t t_tokens_generation_total = 0;

uint64_t n_past_max = 0;

uint64_t n_prompt_tokens_processed = 0;
uint64_t t_prompt_processing = 0;

Expand All @@ -1605,6 +1611,10 @@ struct server_metrics {
n_prompt_tokens_processed += slot.n_prompt_tokens_processed;
t_prompt_processing += slot.t_prompt_processing;
t_prompt_processing_total += slot.t_prompt_processing;

if (slot.n_past > 0) {
n_past_max = std::max(n_past_max, (uint64_t) slot.n_past);
}
}

void on_prediction(const server_slot & slot) {
Expand All @@ -1620,6 +1630,9 @@ struct server_metrics {
if (slot.is_processing()) {
n_busy_slots_total++;
}
if (slot.n_past > 0) {
n_past_max = std::max(n_past_max, (uint64_t) slot.n_past);
}
}
}

Expand Down Expand Up @@ -2875,6 +2888,8 @@ struct server_context {
res->n_tokens_predicted_total = metrics.n_tokens_predicted_total;
res->t_tokens_generation_total = metrics.t_tokens_generation_total;

res->n_past_max = metrics.n_past_max;

res->n_prompt_tokens_processed = metrics.n_prompt_tokens_processed;
res->t_prompt_processing = metrics.t_prompt_processing;
res->n_tokens_predicted = metrics.n_tokens_predicted;
Expand Down Expand Up @@ -4077,6 +4092,10 @@ int main(int argc, char ** argv) {
{"name", "n_decode_total"},
{"help", "Total number of llama_decode() calls"},
{"value", res_metrics->n_decode_total}
}, {
{"name", "n_past_max"},
{"help", "Largest observed n_past."},
{"value", res_metrics->n_past_max}
}, {
{"name", "n_busy_slots_per_decode"},
{"help", "Average number of busy slots per llama_decode() call"},
Expand Down
Loading