baalajimaestro/ollama
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity

sort batch results (#6189)

Browse source
This commit is contained in:
royjhan 2024-08-05 19:55:34 -04:00 committed by GitHub
parent 10d49bce70
commit 86b907f82a
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
1 changed files with 6 additions and 0 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

6
llm/ext_server/server.cpp vendored
View file

@ -44,6 +44,7 @@
#include <errhandlingapi.h> #include <errhandlingapi.h>
#endif #endif
#include <algorithm>
#include <cstddef> #include <cstddef>
#include <thread> #include <thread>
#include <chrono> #include <chrono>
@ -1220,6 +1221,7 @@ struct llama_server_context
res.result_json = json res.result_json = json
{ {
{"id", res.id},
{"embedding", std::vector<float>(embd, embd + n_embd)}, {"embedding", std::vector<float>(embd, embd + n_embd)},
{"timings", slot.get_formated_timings()}, {"timings", slot.get_formated_timings()},
}; };
@ -3203,6 +3205,10 @@ int main(int argc, char **argv) {
} }
responses = result.result_json.value("results", std::vector<json>{result.result_json}); responses = result.result_json.value("results", std::vector<json>{result.result_json});
std::sort(responses.begin(), responses.end(), [](const json& a, const json& b) {
return a["id"] < b["id"];
});
json embeddings = json::array(); json embeddings = json::array();
int prompt_n = 0; int prompt_n = 0;