CS348Project/llama.cpp
5
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-10-31 08:51:55 +00:00
Code Issues Packages Projects Releases Wiki Activity

server : fix --threads-http arg (#7801)

Browse Source
This commit is contained in:
Georgi Gerganov
2024-06-06 19:19:59 +03:00
committed by GitHub
parent f83351f9a6
commit ee459f40f6
2 changed files with 10 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

9
common/common.cpp
View File

@@ -1414,6 +1414,14 @@ bool gpt_params_find_arg(int argc, char ** argv, const std::string & arg, gpt_pa
params.timeout_write = std::stoi(argv[i]);
return true;
}
if (arg == "--threads-http") {
if (++i >= argc) {
invalid_param = true;
return true;
}
params.n_threads_http = std::stoi(argv[i]);
return true;
}
if (arg == "-spf" || arg == "--system-prompt-file") {
if (++i >= argc) {
invalid_param = true;
@@ -1893,6 +1901,7 @@ void gpt_params_print_usage(int /*argc*/, char ** argv, const gpt_params & param
options.push_back({ "server", " --ssl-key-file FNAME", "path to file a PEM-encoded SSL private key" });
options.push_back({ "server", " --ssl-cert-file FNAME", "path to file a PEM-encoded SSL certificate" });
options.push_back({ "server", " --timeout N", "server read/write timeout in seconds (default: %d)", params.timeout_read });
options.push_back({ "server", " --threads-http N", "number of threads used to process HTTP requests (default: %d)", params.n_threads_http });
options.push_back({ "server", " --system-prompt-file FNAME",
"set a file to load a system prompt (initial prompt of all slots), this is useful for chat applications" });
options.push_back({ "server", " --log-format {text,json}",

2
common/common.h
View File

@@ -184,7 +184,7 @@ struct gpt_params {
int32_t port = 8080; // server listens on this network port
int32_t timeout_read = 600; // http read timeout in seconds
int32_t timeout_write = timeout_read; // http write timeout in seconds
int32_t n_threads_http = -1; // number of threads to use for http server (-1 = use n_threads)
int32_t n_threads_http = -1; // number of threads to process HTTP requests
std::string hostname = "127.0.0.1";
std::string public_path = "";