tqcq/llama.cpp
0
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-06-27 20:05:20 +00:00
Code Issues Packages Projects Releases Wiki Activity

imatrix : Add --parse-special for enabling parsing of special tokens in imatrix calculation (#13389)

Browse Source 
* Add --parse-special for enabling parsing of special tokens in imatrix calculation

* whitespace
This commit is contained in:
Bartowski
2025-05-09 05:53:58 -04:00
committed by GitHub
parent 0527771dd8
commit efb8b47eda
3 changed files with 11 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

7
common/arg.cpp
View File

@ -2627,6 +2627,13 @@ common_params_context common_params_parser_init(common_params & params, llama_ex
params.i_chunk = value; params.i_chunk = value;
} }
).set_examples({LLAMA_EXAMPLE_IMATRIX})); ).set_examples({LLAMA_EXAMPLE_IMATRIX}));
add_opt(common_arg(
{"--parse-special"},
string_format("prase special tokens (chat, tool, etc) (default: %s)", params.parse_special ? "true" : "false"),
[](common_params & params) {
params.parse_special = true;
}
).set_examples({LLAMA_EXAMPLE_IMATRIX}));
add_opt(common_arg( add_opt(common_arg(
{"-pps"}, {"-pps"},
string_format("is the prompt shared across parallel sequences (default: %s)", params.is_pp_shared ? "true" : "false"), string_format("is the prompt shared across parallel sequences (default: %s)", params.is_pp_shared ? "true" : "false"),

1
common/common.h
View File

@ -409,6 +409,7 @@ struct common_params {
bool process_output = false; // collect data for the output tensor bool process_output = false; // collect data for the output tensor
bool compute_ppl = true; // whether to compute perplexity bool compute_ppl = true; // whether to compute perplexity
bool parse_special = false; // whether to parse special tokens during imatrix tokenization
// cvector-generator params // cvector-generator params
int n_pca_batch = 100; int n_pca_batch = 100;

5
tools/imatrix/imatrix.cpp
View File

@ -24,7 +24,8 @@ static void print_usage(int, char ** argv) {
LOG("\n %s \\\n" LOG("\n %s \\\n"
" -m model.gguf -f some-text.txt [-o imatrix.dat] [--process-output] \\\n" " -m model.gguf -f some-text.txt [-o imatrix.dat] [--process-output] \\\n"
" [--no-ppl] [--chunk 123] [--output-frequency 10] [--save-frequency 0] \\\n" " [--no-ppl] [--chunk 123] [--output-frequency 10] [--save-frequency 0] \\\n"
" [--in-file imatrix-prev-0.dat --in-file imatrix-prev-1.dat ...]\n" , argv[0]); " [--in-file imatrix-prev-0.dat --in-file imatrix-prev-1.dat ...] \\\n"
" [--parse-special]\n" , argv[0]);
LOG("\n"); LOG("\n");
} }
@ -439,7 +440,7 @@ static bool compute_imatrix(llama_context * ctx, const common_params & params) {
auto tim1 = std::chrono::high_resolution_clock::now(); auto tim1 = std::chrono::high_resolution_clock::now();
LOG_INF("%s: tokenizing the input ..\n", __func__); LOG_INF("%s: tokenizing the input ..\n", __func__);
std::vector<llama_token> tokens = common_tokenize(ctx, params.prompt, true); std::vector<llama_token> tokens = common_tokenize(ctx, params.prompt, true, params.parse_special);
auto tim2 = std::chrono::high_resolution_clock::now(); auto tim2 = std::chrono::high_resolution_clock::now();
LOG_INF("%s: tokenization took %g ms\n",__func__,1e-3*std::chrono::duration_cast<std::chrono::microseconds>(tim2-tim1).count()); LOG_INF("%s: tokenization took %g ms\n",__func__,1e-3*std::chrono::duration_cast<std::chrono::microseconds>(tim2-tim1).count());