dearwolf/LocalAI
1
0
Fork 0
mirror of https://github.com/mudler/LocalAI.git synced 2025-05-20 02:24:59 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 21

chore(deps): update llama.cpp and sync with upstream changes (#4950)

Browse source 
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
This commit is contained in:
Ettore Di Giacinto 2025-03-06 00:40:58 +01:00 • committed by GitHub
parent de81b42b49
commit 67f7bffd18
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
5 changed files with 8 additions and 11 deletions
Download patch file Download diff file Expand all files Collapse all files

1
backend/backend.proto
View file

@ -165,7 +165,6 @@ message Reply {
message GrammarTrigger {
string word = 1;
bool at_start = 2;
}
message ModelOptions {

14
backend/cpp/llama/grpc-server.cpp
View file

@ -469,7 +469,7 @@ struct llama_server_context
bool has_eos_token = true;
bool grammar_lazy = false;
std::vector<common_grammar_trigger> grammar_trigger_words;
std::vector<common_grammar_trigger> grammar_triggers;
int32_t n_ctx; // total context for all clients / slots
@ -709,7 +709,7 @@ struct llama_server_context
slot->sparams.grammar = json_value(data, "grammar", default_sparams.grammar);
slot->sparams.n_probs = json_value(data, "n_probs", default_sparams.n_probs);
slot->sparams.min_keep = json_value(data, "min_keep", default_sparams.min_keep);
slot->sparams.grammar_trigger_words = grammar_trigger_words;
slot->sparams.grammar_triggers = grammar_triggers;
slot->sparams.grammar_lazy = grammar_lazy;
if (slot->n_predict > 0 && slot->params.n_predict > slot->n_predict) {
@ -2393,12 +2393,12 @@ static void params_parse(const backend::ModelOptions* request,
llama.grammar_lazy = true;
for (int i = 0; i < request->grammartriggers_size(); i++) {
common_grammar_trigger trigger;
trigger.word = request->grammartriggers(i).word();
trigger.at_start = request->grammartriggers(i).at_start();
llama.grammar_trigger_words.push_back(trigger);
trigger.type = COMMON_GRAMMAR_TRIGGER_TYPE_WORD;
trigger.value = request->grammartriggers(i).word();
// trigger.at_start = request->grammartriggers(i).at_start();
llama.grammar_triggers.push_back(trigger);
LOG_INFO("grammar trigger", {
{ "word", trigger.word },
{ "at_start", trigger.at_start }
{ "word", trigger.value },
});
}
}