mirror of
https://github.com/ggml-org/llama.cpp.git
synced 2026-08-26 06:31:04 +02:00
fix ci
This commit is contained in:
@@ -2,6 +2,7 @@
|
||||
#include "common.h"
|
||||
|
||||
#include <algorithm>
|
||||
#include <limits>
|
||||
#include <map>
|
||||
#include <regex>
|
||||
#include <sstream>
|
||||
|
||||
@@ -655,14 +655,14 @@ struct server_slot {
|
||||
res["n_prompt_tokens_processed"] = stats.n_prompt_processed;
|
||||
res["n_prompt_tokens_cache"] = stats.n_prompt_cached;
|
||||
res["params"] = ptask->params.to_json(only_metrics);
|
||||
res["next_token"] = {
|
||||
res["next_token"] = json::array({
|
||||
{
|
||||
{"has_next_token", has_next_token},
|
||||
{"has_new_line", has_new_line},
|
||||
{"n_remain", n_remaining()},
|
||||
{"n_decoded", stats.n_gen},
|
||||
}
|
||||
};
|
||||
});
|
||||
|
||||
if (!only_metrics) {
|
||||
res["prompt"] = ptask->tokens.detokenize(ctx_tgt, true);
|
||||
|
||||
Reference in New Issue
Block a user