diff options
Diffstat (limited to 'examples/common.h')
| -rw-r--r-- | examples/common.h | 30 |
1 files changed, 21 insertions, 9 deletions
diff --git a/examples/common.h b/examples/common.h index 1ea6f744..df8e4c6c 100644 --- a/examples/common.h +++ b/examples/common.h @@ -14,14 +14,14 @@ // struct gpt_params { - int32_t seed = -1; // RNG seed - int32_t n_threads = std::min(4, (int32_t) std::thread::hardware_concurrency()); - int32_t n_predict = 128; // new tokens to predict - int32_t repeat_last_n = 64; // last n tokens to penalize - int32_t n_parts = -1; // amount of model parts (-1 = determine from model dimensions) - int32_t n_ctx = 512; // context size - int32_t n_batch = 8; // batch size for prompt processing - int32_t n_keep = 0; // number of tokens to keep from initial prompt + int32_t seed = -1; // RNG seed + int32_t n_threads = std::min(4, (int32_t) std::thread::hardware_concurrency()); // max 4 threads (default) + int32_t n_predict = 128; // new tokens to predict + int32_t repeat_last_n = 64; // last n tokens to penalize + int32_t n_parts = -1; // amount of model parts (-1 = determine from model dimensions) + int32_t n_ctx = 512; // context size + int32_t n_batch = 8; // batch size for prompt processing + int32_t n_keep = 0; // number of tokens to keep from initial prompt (-1 for all) // sampling parameters int32_t top_k = 40; @@ -33,8 +33,15 @@ struct gpt_params { std::string prompt = ""; std::string input_prefix = ""; // string to prefix user inputs with + std::string instruct_prefix = ""; // prefix user inputs with tokenized string + bool instruct_prefix_bos = false; // prepend bos token to instruct prefix + std::string instruct_suffix = ""; // suffix user inputs with tokenized string + bool instruct_suffix_bos = false; // prepend bos token to instruct suffix std::vector<std::string> antiprompt; // string upon seeing which more user input is prompted + std::vector<std::string> stopprompt; // string upon seeing which more user input is prompted (without adding instruct prefixes and suffixes) + + bool rm_trailing_space_workaround = false; // workaround for removing trailing space from reverse/stop prompts bool memory_f16 = true; // use f16 instead of f32 for memory kv bool random_prompt = false; // do not randomize prompt if none provided @@ -51,11 +58,14 @@ struct gpt_params { bool use_mlock = false; // use mlock to keep model in memory bool mem_test = false; // compute maximum memory usage bool verbose_prompt = false; // print prompt tokens before generation + + bool clean_interface = false; // hides input prefix & suffix and displays '>' + bool multiline_mode = true; // enables multi-line mode, to send input press CTRL+D on Linux/Max, Ctrl+Z then Return on Windows }; bool gpt_params_parse(int argc, char ** argv, gpt_params & params); -void gpt_print_usage(int argc, char ** argv, const gpt_params & params); +void gpt_print_usage(char * argv_0, const gpt_params & params); std::string gpt_random_prompt(std::mt19937 & rng); @@ -95,3 +105,5 @@ void set_console_color(console_state & con_st, console_color_t color); void win32_console_init(bool enable_color); void win32_utf8_encode(const std::wstring & wstr, std::string & str); #endif + +bool get_input_text(std::string & input_text, bool escape_newline_mode); |