happyz
/
llama.cpp
mirror of https://github.com/ggerganov/llama.cpp.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

server: Fix undefined timing measurement errors in server context (#21201) 

Co-authored-by: Dan Hoffman <dhoffman@cyket.net>
This commit is contained in:
Dan Hoffman 2026-04-04 07:11:19 -07:00 committed by GitHub
parent d01f6274c0
commit 9c699074c9
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
2 changed files with 8 additions and 8 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
tools/server/server-context.cpp
View File

@ -155,8 +155,8 @@ struct server_slot {
int64_t t_start_process_prompt;
int64_t t_start_generation;
double t_prompt_processing; // ms
double t_token_generation; // ms
double t_prompt_processing = 0.0; // ms
double t_token_generation = 0.0; // ms
std::function<void(int /* id_slot */)> callback_on_release;

12
tools/server/server-task.h
View File

@ -261,14 +261,14 @@ struct result_timings {
int32_t cache_n = -1;
int32_t prompt_n = -1;
double prompt_ms;
double prompt_per_token_ms;
double prompt_per_second;
double prompt_ms = 0.0;
double prompt_per_token_ms = 0.0;
double prompt_per_second = 0.0;
int32_t predicted_n = -1;
double predicted_ms;
double predicted_per_token_ms;
double predicted_per_second;
double predicted_ms = 0.0;
double predicted_per_token_ms = 0.0;
double predicted_per_second = 0.0;
// Optional speculative metrics - only included when > 0
int32_t draft_n = 0;