Mirrors/llama.cpp
1
0
Fork 0
mirror of https://github.com/ggerganov/llama.cpp.git synced 2024-12-27 06:39:25 +01:00
Code Issues Packages Projects Releases Wiki Activity

common : add --no-warmup option for main/llama-cli (#8712)

Browse Source 
This commit adds a --no-warmup option for llama-cli.

The motivation for this is that it can be convenient to skip the
warmup llama_decode call when debugging.

Signed-off-by: Daniel Bevenius <daniel.bevenius@gmail.com>
This commit is contained in:
Daniel Bevenius 2024-07-27 12:45:02 +02:00 committed by GitHub
parent bfb4c74981
commit 9d03d085dd
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 5 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

5
common/common.cpp
View File

@ -1324,6 +1324,10 @@ bool gpt_params_find_arg(int argc, char ** argv, const std::string & arg, gpt_pa
else { invalid_param = true; } else { invalid_param = true; }
return true; return true;
} }
if (arg == "--no-warmup") {
params.warmup = false;
return true;
}
#ifndef LOG_DISABLE_LOGS #ifndef LOG_DISABLE_LOGS
// Parse args for logging parameters // Parse args for logging parameters
if (log_param_single_parse(argv[i])) { if (log_param_single_parse(argv[i])) {
@ -1446,6 +1450,7 @@ void gpt_params_print_usage(int /*argc*/, char ** argv, const gpt_params & param
options.push_back({ "main infill", " --in-prefix-bos", "prefix BOS to user inputs, preceding the `--in-prefix` string" }); options.push_back({ "main infill", " --in-prefix-bos", "prefix BOS to user inputs, preceding the `--in-prefix` string" });
options.push_back({ "main infill", " --in-prefix STRING", "string to prefix user inputs with (default: empty)" }); options.push_back({ "main infill", " --in-prefix STRING", "string to prefix user inputs with (default: empty)" });
options.push_back({ "main infill", " --in-suffix STRING", "string to suffix after user inputs with (default: empty)" }); options.push_back({ "main infill", " --in-suffix STRING", "string to suffix after user inputs with (default: empty)" });
options.push_back({ "main", " --no-warmup", "skip warming up the model with an empty run" });
options.push_back({ "server infill", options.push_back({ "server infill",
" --spm-infill", "use Suffix/Prefix/Middle pattern for infill (instead of Prefix/Suffix/Middle) as some models prefer this. (default: %s)", params.spm_infill ? "enabled" : "disabled" }); " --spm-infill", "use Suffix/Prefix/Middle pattern for infill (instead of Prefix/Suffix/Middle) as some models prefer this. (default: %s)", params.spm_infill ? "enabled" : "disabled" });