nikola/llama_cpp
1
0
Fork 0
Code Issues Pull Requests Actions 4 Packages Projects Releases Wiki Activity

server: refactor child --> router communication (#24821)

Browse Source 
* server: refactor child --> router communication

* fix wakeup case

* add docs

* improve update_status()

* nits
This commit is contained in:
Xuan-Son Nguyen
2026-06-20 01:02:26 +02:00
committed by GitHub
parent 4b48a53b6c
commit 2b686a9120
8 changed files with 173 additions and 91 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
common/arg.cpp
View File

@@ -303,7 +303,6 @@ static handle_model_result common_params_handle_model(struct common_params_model
if (!model.docker_repo.empty()) {
model.path = common_docker_resolve_model(model.docker_repo);
model.name = model.docker_repo;
} else if (!model.hf_repo.empty()) {
// If -m was used with -hf, treat the model "path" as the hf_file to download
if (model.hf_file.empty() && !model.path.empty()) {
@@ -323,7 +322,6 @@ static handle_model_result common_params_handle_model(struct common_params_model
throw std::runtime_error("failed to download model from Hugging Face");
}
model.name = model.hf_repo;
model.path = download_result.model_path;
if (!download_result.mmproj_path.empty()) {

11
common/common.h
View File

@@ -295,7 +295,16 @@ struct common_params_model {
std::string hf_repo = ""; // HF repo // NOLINT
std::string hf_file = ""; // HF file // NOLINT
std::string docker_repo = ""; // Docker repo // NOLINT
std::string name = ""; // in format <user>/<model>[:<tag>] (tag is optional) // NOLINT
std::string get_name() {
if (!hf_repo.empty()) {
return hf_repo;
}
if (!docker_repo.empty()) {
return docker_repo;
}
return path;
}
};
// draft-model-based speculative decoding parameters