vbatts/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Switch to the CPPHTTPLIB logger. Verbose adds body dump as well as request info.

Browse source
This commit is contained in:
digiwombat 2023-05-31 16:23:39 -04:00
parent 7332b41f9f
commit dda4c10d64
1 changed files with 18 additions and 11 deletions
Download patch file Download diff file Expand all files Collapse all files

27
examples/server/server.cpp
View file

@ -750,11 +750,8 @@ int main(int argc, char **argv)
Server svr; Server svr;
svr.Get("/", [](const Request &req, Response &res) svr.Get("/", [](const Request &, Response &res)
{ { res.set_content("<h1>llama.cpp server works</h1>", "text/html"); });
fprintf(stderr, "request: GET / [remote_addr: %s]", req.remote_addr.c_str());
res.set_content("<h1>llama.cpp server works</h1>", "text/html");
});
svr.Post("/completion", [&llama](const Request &req, Response &res) { svr.Post("/completion", [&llama](const Request &req, Response &res) {
if (llama.params.embedding) { if (llama.params.embedding) {
@ -775,8 +772,6 @@ int main(int argc, char **argv)
return; return;
} }
fprintf(stderr, "request: POST /completion [remote_addr: %s, stream: %s]", req.remote_addr.c_str(), llama.stream ? "true" : "false");
if (!llama.loadPrompt()) { if (!llama.loadPrompt()) {
json data = {{"status", "error"}, {"reason", "Context too long."}}; json data = {{"status", "error"}, {"reason", "Context too long."}};
res.set_content( res.set_content(
@ -890,7 +885,6 @@ int main(int argc, char **argv)
svr.Post("/tokenize", [&llama](const Request &req, Response &res) svr.Post("/tokenize", [&llama](const Request &req, Response &res)
{ {
fprintf(stderr, "request: POST /tokenize [remote_addr: %s]", req.remote_addr.c_str());
json body = json::parse(req.body); json body = json::parse(req.body);
json data = { json data = {
{"tokens", ::llama_tokenize(llama.ctx, body["content"].get<std::string>(), false) } }; {"tokens", ::llama_tokenize(llama.ctx, body["content"].get<std::string>(), false) } };
@ -899,7 +893,6 @@ int main(int argc, char **argv)
svr.Post("/embedding", [&llama](const Request &req, Response &res) svr.Post("/embedding", [&llama](const Request &req, Response &res)
{ {
fprintf(stderr, "request: POST /embedding [remote_addr: %s]", req.remote_addr.c_str());
if(!llama.params.embedding) { if(!llama.params.embedding) {
std::vector<float> empty; std::vector<float> empty;
json data = { json data = {
@ -921,7 +914,21 @@ int main(int argc, char **argv)
if (llama.verbose) { if (llama.verbose) {
svr.set_logger([](const Request& req, const Response& res) { svr.set_logger([](const Request& req, const Response& res) {
fprintf(stderr, "%s", log(req, res).c_str()); json log = {
{ "status", res.status },
{ "request", req.body },
{ "response", res.body },
};
fprintf(stdout, "http_request: request: %s %s \nhttp_request: log: %s\n", req.method.c_str(), req.path.c_str(), log.dump().c_str());
});
} else {
svr.set_logger([](const Request& req, const Response& res) {
json log = {
{ "status", res.status },
{ "request", req.body },
{ "response", res.body },
};
fprintf(stdout, "http_request: request: %s %s \n", req.method.c_str(), req.path.c_str());
}); });
} }