vbatts/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

examples : rely on new behavior of add_special

Browse source
This commit is contained in:
Jared Van Bortel 2024-04-04 18:12:33 -04:00
parent d1a1b614cd
commit 92591c125f
14 changed files with 53 additions and 69 deletions
Download patch file Download diff file Expand all files Collapse all files

3
examples/imatrix/imatrix.cpp
View file

@ -349,12 +349,13 @@ static void process_logits(
static bool compute_imatrix(llama_context * ctx, const gpt_params & params, bool compute_ppl, int from_chunk) {
const bool add_bos = llama_should_add_bos_token(llama_get_model(ctx));
GGML_ASSERT(llama_add_eos_token(llama_get_model(ctx)) != 1);
const int n_ctx = llama_n_ctx(ctx);
auto tim1 = std::chrono::high_resolution_clock::now();
fprintf(stderr, "%s: tokenizing the input ..\n", __func__);
std::vector<llama_token> tokens = ::llama_tokenize(ctx, params.prompt, add_bos);
std::vector<llama_token> tokens = ::llama_tokenize(ctx, params.prompt, true);
auto tim2 = std::chrono::high_resolution_clock::now();
fprintf(stderr, "%s: tokenization took %g ms\n",__func__,1e-3*std::chrono::duration_cast<std::chrono::microseconds>(tim2-tim1).count());