Skip to content

Support for multiple reverse prompts #299

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 4 commits into from
Mar 19, 2023
Merged

Support for multiple reverse prompts #299

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
9ef4920
Support for multiple reverse prompts.
Mar 19, 2023
b7b3b11
Merge branch 'master' into multiple-reverse-prompts
tjohnman Mar 19, 2023
722a3ec
Merge branch 'master' into multiple-reverse-prompts
Mar 19, 2023
c101b73
Merge branch 'multiple-reverse-prompts' of github.com:tjohnman/llama.…
Mar 19, 2023
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
34 changes: 22 additions & 12 deletions main.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -855,14 +855,18 @@ int main(int argc, char ** argv) {
// in instruct mode, we inject a prefix and a suffix to each input by the user
if (params.instruct) {
params.interactive = true;
params.antiprompt = "### Instruction:\n\n";
params.antiprompt.push_back("### Instruction:\n\n");
}

// tokenize the reverse prompt
std::vector<gpt_vocab::id> antiprompt_inp = ::llama_tokenize(vocab, params.antiprompt, false);
std::vector<std::vector<gpt_vocab::id>> antipromptv_inp;

for (auto antiprompt : params.antiprompt) {
antipromptv_inp.push_back(::llama_tokenize(vocab, antiprompt, false));
}

// enable interactive mode if reverse prompt is specified
if (!antiprompt_inp.empty()) {
if (!antipromptv_inp.size()) {
params.interactive = true;
}

Expand All @@ -886,13 +890,16 @@ int main(int argc, char ** argv) {

fprintf(stderr, "%s: interactive mode on.\n", __func__);

if (antiprompt_inp.size()) {
fprintf(stderr, "%s: reverse prompt: '%s'\n", __func__, params.antiprompt.c_str());
fprintf(stderr, "%s: number of tokens in reverse prompt = %zu\n", __func__, antiprompt_inp.size());
for (int i = 0; i < (int) antiprompt_inp.size(); i++) {
fprintf(stderr, "%6d -> '%s'\n", antiprompt_inp[i], vocab.id_to_token.at(antiprompt_inp[i]).c_str());
if(antipromptv_inp.size()) {
for (size_t apindex = 0; apindex < antipromptv_inp.size(); ++apindex) {
auto antiprompt_inp = antipromptv_inp.at(apindex);
fprintf(stderr, "%s: reverse prompt: '%s'\n", __func__, params.antiprompt.at(apindex).c_str());
fprintf(stderr, "%s: number of tokens in reverse prompt = %zu\n", __func__, antiprompt_inp.size());
for (int i = 0; i < (int) antiprompt_inp.size(); i++) {
fprintf(stderr, "%6d -> '%s'\n", antiprompt_inp[i], vocab.id_to_token.at(antiprompt_inp[i]).c_str());
}
fprintf(stderr, "\n");
}
fprintf(stderr, "\n");
}
}
fprintf(stderr, "sampling parameters: temp = %f, top_k = %d, top_p = %f, repeat_last_n = %i, repeat_penalty = %f\n", params.temp, params.top_k, params.top_p, params.repeat_last_n, params.repeat_penalty);
Expand Down Expand Up @@ -1009,9 +1016,12 @@ int main(int argc, char ** argv) {
// check if we should prompt the user for more
if (params.interactive && embd_inp.size() <= input_consumed) {
// check for reverse prompt
if (antiprompt_inp.size() && std::equal(antiprompt_inp.rbegin(), antiprompt_inp.rend(), last_n_tokens.rbegin())) {
// reverse prompt found
is_interacting = true;
for (auto antiprompt_inp : antipromptv_inp) {
if (antiprompt_inp.size() && std::equal(antiprompt_inp.rbegin(), antiprompt_inp.rend(), last_n_tokens.rbegin())) {
// reverse prompt found
is_interacting = true;
break;
}
}
if (is_interacting) {
if (params.instruct) {
Expand Down
5 changes: 3 additions & 2 deletions utils.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -70,7 +70,7 @@ bool gpt_params_parse(int argc, char ** argv, gpt_params & params) {
} else if (arg == "--color") {
params.use_color = true;
} else if (arg == "-r" || arg == "--reverse-prompt") {
params.antiprompt = argv[++i];
params.antiprompt.push_back(argv[++i]);
} else if (arg == "--ignore-eos") {
params.ignore_eos = true;
} else if (arg == "-h" || arg == "--help") {
Expand All @@ -96,7 +96,8 @@ void gpt_print_usage(int /*argc*/, char ** argv, const gpt_params & params) {
fprintf(stderr, " -i, --interactive run in interactive mode\n");
fprintf(stderr, " -ins, --instruct run in instruction mode (use with Alpaca models)\n");
fprintf(stderr, " -r PROMPT, --reverse-prompt PROMPT\n");
fprintf(stderr, " in interactive mode, poll user input upon seeing PROMPT\n");
fprintf(stderr, " in interactive mode, poll user input upon seeing PROMPT (can be\n");
fprintf(stderr, " specified more than once for multiple prompts).\n");
fprintf(stderr, " --color colorise output to distinguish prompt and user input from generations\n");
fprintf(stderr, " -s SEED, --seed SEED RNG seed (default: -1)\n");
fprintf(stderr, " -t N, --threads N number of threads to use during computation (default: %d)\n", params.n_threads);
Expand Down
4 changes: 2 additions & 2 deletions utils.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -30,15 +30,15 @@ struct gpt_params {

std::string model = "models/lamma-7B/ggml-model.bin"; // model path
std::string prompt = "";
std::string antiprompt = ""; // string upon seeing which more user input is prompted

bool random_prompt = false;

bool use_color = false; // use color to distinguish generations and inputs

bool interactive = false; // interactive mode
bool interactive_start = false; // reverse prompt immediately
std::vector<std::string> antiprompt; // string upon seeing which more user input is prompted
bool instruct = false; // instruction mode (used for Alpaca models)

bool ignore_eos = false; // do not stop generating after eos
};

Expand Down