CS348Project/llama.cpp
5
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-10-31 08:51:55 +00:00
Code Issues Packages Projects Releases Wiki Activity

parallel : count cache misses

Browse Source
This commit is contained in:
Georgi Gerganov
2023-09-19 23:47:47 +03:00
parent 8a9aca37c1
commit eed3fd4234
1 changed files with 4 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
examples/parallel/parallel.cpp
View File

@@ -135,6 +135,7 @@ int main(int argc, char ** argv) {
int32_t n_total_prompt = 0;
int32_t n_total_gen = 0;
int32_t n_cache_miss = 0;
const auto t_main_start = ggml_time_us();
@@ -272,6 +273,8 @@ int main(int argc, char ** argv) {
LOG("%s : failed to decode batch, retrying with n_batch = %d\n", __func__, n_batch / 2);
n_cache_miss += 1;
// retry with half the batch size to try to find a free slot in the KV cache
n_batch /= 2;
i -= n_batch;
@@ -349,6 +352,7 @@ int main(int argc, char ** argv) {
LOG_TEE("Total prompt tokens: %6d, speed: %5.2f t/s\n", n_total_prompt, (double) (n_total_prompt ) / (t_main_end - t_main_start) * 1e6);
LOG_TEE("Total gen tokens: %6d, speed: %5.2f t/s\n", n_total_gen, (double) (n_total_gen ) / (t_main_end - t_main_start) * 1e6);
LOG_TEE("Total speed (AVG): %6s speed: %5.2f t/s\n", "", (double) (n_total_prompt + n_total_gen) / (t_main_end - t_main_start) * 1e6);
LOG_TEE("Cache misses: %6d\n", n_cache_miss);
LOG_TEE("\n\n");