ikawrakow/ik_llama.cpp
1
0
Fork 0
mirror of https://github.com/ikawrakow/ik_llama.cpp.git synced 2026-04-29 19:01:47 +00:00
Code Issues Packages Projects Releases Wiki Activity

Server: refactor and rename functions (#1151)

Browse Source 
* Server: rename functions and refactor code

rename functions

refactor update slots

rename params_base

rename timings

* change

* Revert kv cache name changes

* Revert 2

* fix test build error

---------

Co-authored-by: firecoperana <firecoperana>
This commit is contained in:
firecoperana
2026-01-18 00:16:57 -06:00
committed by GitHub
parent 7024fdbc72
commit d71a3ec315
38 changed files with 532 additions and 528 deletions
Download Patch File Download Diff File Expand all files Collapse all files

16
examples/sweep-bench/sweep-bench.cpp
View File

@@ -108,7 +108,7 @@ int main(int argc, char ** argv) {
// warm up
if (params.warmup) {
llama_batch_add(batch, bos, 0, { 0 }, false);
common_batch_add(batch, bos, 0, { 0 }, false);
if (!decode_helper(ctx, batch, ctx_params.n_batch)) {
LOG_TEE("%s: llama_decode() failed\n", __func__);
@@ -120,10 +120,10 @@ int main(int argc, char ** argv) {
llama_kv_cache_seq_rm(ctx, 0, params.n_ubatch, -1);
// prepare batch of pp size for prompt processing performance measurement
llama_batch_clear(batch);
common_batch_clear(batch);
for (unsigned int i = 0; i < params.n_ubatch; ++i) {
llama_batch_add(batch, std::rand() % n_vocab, i, { 0 }, false);
common_batch_add(batch, std::rand() % n_vocab, i, { 0 }, false);
}
if (!decode_helper(ctx, batch, ctx_params.n_ubatch)) {
@@ -132,7 +132,7 @@ int main(int argc, char ** argv) {
}
}
llama_batch_clear(batch);
common_batch_clear(batch);
llama_kv_cache_clear(ctx);
for (unsigned int n_kv = 0; n_kv < n_kv_max; n_kv += params.n_ubatch) {
@@ -143,8 +143,8 @@ int main(int argc, char ** argv) {
const auto t_tg_start = ggml_time_us();
for (unsigned int i = 0; i < tg; ++i) {
llama_batch_clear(batch);
llama_batch_add(batch, std::rand() % n_vocab, n_kv + i, { 0 }, true);
common_batch_clear(batch);
common_batch_add(batch, std::rand() % n_vocab, n_kv + i, { 0 }, true);
if (!decode_helper(ctx, batch, ctx_params.n_batch)) {
LOG_TEE("%s: llama_decode() failed\n", __func__);
@@ -158,10 +158,10 @@ int main(int argc, char ** argv) {
llama_kv_cache_seq_rm(ctx, 0, n_kv, -1);
// prepare batch of pp size for prompt processing performance measurement
llama_batch_clear(batch);
common_batch_clear(batch);
for (unsigned int i = 0; i < pp; ++i) {
llama_batch_add(batch, std::rand() % n_vocab, n_kv + i, { 0 }, false);
common_batch_add(batch, std::rand() % n_vocab, n_kv + i, { 0 }, false);
}
batch.logits[batch.n_tokens - 1] = true;