CS348Project/llama.cpp
5
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-11-06 09:46:50 +00:00
Code Issues Packages Projects Releases Wiki Activity

examples : do not eval prompt 2 times (close #3348)

Browse Source
This commit is contained in:
Georgi Gerganov
2023-09-28 17:48:25 +03:00
parent a207561503
commit 2b8830af71
2 changed files with 24 additions and 21 deletions
Download Patch File Download Diff File Expand all files Collapse all files

16
examples/simple/simple.cpp
View File

@@ -110,16 +110,10 @@ int main(int argc, char ** argv) {
const auto t_main_start = ggml_time_us();
while (n_cur <= n_len) {
// evaluate the current batch with the transformer model
if (llama_decode(ctx, batch, params.n_threads)) {
fprintf(stderr, "%s : failed to eval, return code %d\n", __func__, 1);
return 1;
}
// sample the next token
{
auto n_vocab = llama_n_vocab(ctx);
auto logits = llama_get_logits_ith(ctx, batch.n_tokens - 1);
auto n_vocab = llama_n_vocab(ctx);
auto * logits = llama_get_logits_ith(ctx, batch.n_tokens - 1);
std::vector<llama_token_data> candidates;
candidates.reserve(n_vocab);
@@ -158,6 +152,12 @@ int main(int argc, char ** argv) {
}
n_cur += 1;
// evaluate the current batch with the transformer model
if (llama_decode(ctx, batch, params.n_threads)) {
fprintf(stderr, "%s : failed to eval, return code %d\n", __func__, 1);
return 1;
}
}
LOG_TEE("\n");