Skip to content

Commit 502a400

Browse files
committed
Disable prompt verbosity by default and add option to enable (ggml-org#480)
1 parent 09aecbf commit 502a400

File tree

3 files changed

+13
-6
lines changed

3 files changed

+13
-6
lines changed

main.cpp

Lines changed: 9 additions & 6 deletions
Original file line numberDiff line numberDiff line change
@@ -275,13 +275,16 @@ int main(int argc, char ** argv) {
275275
// determine newline token
276276
auto llama_token_newline = ::llama_tokenize(ctx, "\n", false);
277277

278-
fprintf(stderr, "\n");
279-
fprintf(stderr, "%s: prompt: '%s'\n", __func__, params.prompt.c_str());
280-
fprintf(stderr, "%s: number of tokens in prompt = %zu\n", __func__, embd_inp.size());
281-
for (int i = 0; i < (int) embd_inp.size(); i++) {
282-
fprintf(stderr, "%6d -> '%s'\n", embd_inp[i], llama_token_to_str(ctx, embd_inp[i]));
278+
if (params.verbose_prompt) {
279+
fprintf(stderr, "\n");
280+
fprintf(stderr, "%s: prompt: '%s'\n", __func__, params.prompt.c_str());
281+
fprintf(stderr, "%s: number of tokens in prompt = %zu\n", __func__, embd_inp.size());
282+
for (int i = 0; i < (int) embd_inp.size(); i++) {
283+
fprintf(stderr, "%6d -> '%s'\n", embd_inp[i], llama_token_to_str(ctx, embd_inp[i]));
284+
}
285+
fprintf(stderr, "\n");
283286
}
284-
fprintf(stderr, "\n");
287+
285288
if (params.interactive) {
286289
#if defined (__unix__) || (defined (__APPLE__) && defined (__MACH__))
287290
struct sigaction sigint_action;

utils.cpp

Lines changed: 3 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -134,6 +134,8 @@ bool gpt_params_parse(int argc, char ** argv, gpt_params & params) {
134134
params.use_mlock = true;
135135
} else if (arg == "--mtest") {
136136
params.mem_test = true;
137+
} else if (arg == "--verbose_prompt") {
138+
params.verbose_prompt = true;
137139
} else if (arg == "-r" || arg == "--reverse-prompt") {
138140
if (++i >= argc) {
139141
invalid_param = true;
@@ -212,6 +214,7 @@ void gpt_print_usage(int /*argc*/, char ** argv, const gpt_params & params) {
212214
fprintf(stderr, " --mlock force system to keep model in RAM rather than swapping or compressing\n");
213215
}
214216
fprintf(stderr, " --mtest compute maximum memory usage\n");
217+
fprintf(stderr, " --verbose-prompt print prompt before generation\n");
215218
fprintf(stderr, " -m FNAME, --model FNAME\n");
216219
fprintf(stderr, " model path (default: %s)\n", params.model.c_str());
217220
fprintf(stderr, "\n");

utils.h

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -48,6 +48,7 @@ struct gpt_params {
4848
bool perplexity = false; // compute perplexity over the prompt
4949
bool use_mlock = false; // use mlock to keep model in memory
5050
bool mem_test = false; // compute maximum memory usage
51+
bool verbose_prompt = false; // print prompt tokens before generation
5152
};
5253

5354
bool gpt_params_parse(int argc, char ** argv, gpt_params & params);

0 commit comments

Comments
 (0)