nikola/llama_cpp
1
0
Fork 0
Code Issues Pull Requests Actions 11 Packages Projects Releases Wiki Activity

common: refactor model handling (#24980)

Browse Source 
* common: refactor models handling

* remote preset

* cont

* rm skip_download option

* missing header

* fix plan.model_files

* fix --offline case

* move hf_plan to download

* refactor

* rm redundant curr_ex, add comments

* adapt
This commit is contained in:
Xuan-Son Nguyen
2026-06-25 15:17:51 +02:00
committed by GitHub
parent e8ecce53b8
commit 60bc8866b1
9 changed files with 289 additions and 327 deletions
Download Patch File Download Diff File Expand all files Collapse all files
common/common.h
+11 -8
View File
@@ -291,13 +291,13 @@ struct common_params_sampling {
};
struct common_params_model {
std::string path = ""; // model local path // NOLINT
std::string url = ""; // model url to download // NOLINT
std::string hf_repo = ""; // HF repo // NOLINT
std::string hf_file = ""; // HF file // NOLINT
std::string docker_repo = ""; // Docker repo // NOLINT
std::string path = ""; // model local path
std::string url = ""; // model url to download
std::string hf_repo = ""; // HF repo
std::string hf_file = ""; // HF file
std::string docker_repo = ""; // Docker repo
std::string get_name() {
std::string get_name() const {
if (!hf_repo.empty()) {
return hf_repo;
}
@@ -306,6 +306,10 @@ struct common_params_model {
}
return path;
}
bool empty() const {
return get_name().empty();
}
};
// draft-model-based speculative decoding parameters
@@ -368,7 +372,7 @@ struct common_params_speculative {
common_params_speculative_ngram_cache ngram_cache;
bool has_dft() const {
return !draft.mparams.path.empty() || !draft.mparams.hf_repo.empty();
return !draft.mparams.empty();
}
uint32_t need_n_rs_seq() const {
@@ -520,7 +524,6 @@ struct common_params {
int32_t control_vector_layer_start = -1; // layer range for control vector
int32_t control_vector_layer_end = -1; // layer range for control vector
bool offline = false;
bool skip_download = false; // skip model file downloading
int32_t ppl_stride = 0; // stride for perplexity calculations. If left at 0, the pre-existing approach will be used.
int32_t ppl_output_type = 0; // = 0 -> ppl output is as usual, = 1 -> ppl output is num_tokens, ppl, one per line