upstream-archive/llama.cpp
1
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2026-07-23 11:10:55 -05:00
Code Issues Packages Projects Releases Wiki Activity

app : add the llama download subcommand (#24982)

Browse Source 
* app : add the download command (with llama-download)

Signed-off-by: Adrien Gallouët <angt@huggingface.co>

* Remove llama-download tool for now

Signed-off-by: Adrien Gallouët <angt@huggingface.co>

---------

Signed-off-by: Adrien Gallouët <angt@huggingface.co>
This commit is contained in:
Adrien Gallouët
2026-06-25 13:36:36 +02:00
committed by GitHub
parent f728adab68
commit 683b04cc4a
5 changed files with 107 additions and 19 deletions
Download Patch File Download Diff File Expand all files Collapse all files

51
common/arg.cpp
View File

@@ -594,6 +594,8 @@ static bool common_params_parse_ex(int argc, char ** argv, common_params_context
const bool skip_model_download =
// server will call common_params_handle_models() later, so we skip it here
ctx_arg.ex == LLAMA_EXAMPLE_SERVER ||
// download calls common_params_handle_models() itself and prints the paths
ctx_arg.ex == LLAMA_EXAMPLE_DOWNLOAD ||
// export_graph_ops loads only metadata
ctx_arg.ex == LLAMA_EXAMPLE_EXPORT_GRAPH_OPS;
@@ -671,15 +673,19 @@ static void common_params_print_usage(common_params_context & ctx_arg) {
common_options.push_back(&opt);
}
}
printf("----- common params -----\n\n");
print_options(common_options);
printf("\n\n----- sampling params -----\n\n");
print_options(sampling_options);
printf("\n\n----- speculative params -----\n\n");
print_options(spec_options);
// TODO: maybe convert enum llama_example to string
printf("\n\n----- example-specific params -----\n\n");
print_options(specific_options);
bool first = true;
auto print_section = [&](const char * header, std::vector<common_arg *> & options) {
if (options.empty()) {
return;
}
printf("%s----- %s -----\n\n", first ? "" : "\n\n", header);
first = false;
print_options(options);
};
print_section("common params", common_options);
print_section("sampling params", sampling_options);
print_section("speculative params", spec_options);
print_section("example-specific params", specific_options);
}
static void common_params_print_completion(common_params_context & ctx_arg) {
@@ -1079,7 +1085,9 @@ common_params_context common_params_parser_init(common_params & params, llama_ex
* - if both {LLAMA_EXAMPLE_COMMON, LLAMA_EXAMPLE_*,} are set, we will prioritize the LLAMA_EXAMPLE_* matching current example
*/
auto add_opt = [&](common_arg arg) {
if ((arg.in_example(ex) || arg.in_example(LLAMA_EXAMPLE_COMMON)) && !arg.is_exclude(ex)) {
// download only exposes the handful of args explicitly tagged for it
const bool inherit_common = ex != LLAMA_EXAMPLE_DOWNLOAD;
if ((arg.in_example(ex) || (inherit_common && arg.in_example(LLAMA_EXAMPLE_COMMON))) && !arg.is_exclude(ex)) {
ctx_arg.options.push_back(std::move(arg));
}
};
@@ -1090,7 +1098,7 @@ common_params_context common_params_parser_init(common_params & params, llama_ex
[](common_params & params) {
params.usage = true;
}
));
).set_examples({LLAMA_EXAMPLE_COMMON, LLAMA_EXAMPLE_DOWNLOAD}));
add_opt(common_arg(
{"--version"},
"show version and build info",
@@ -2212,7 +2220,7 @@ common_params_context common_params_parser_init(common_params & params, llama_ex
[](common_params & params, bool value) {
params.no_mmproj = !value;
}
).set_examples(mmproj_examples).set_env("LLAMA_ARG_MMPROJ_AUTO"));
).set_examples({LLAMA_EXAMPLE_MTMD, LLAMA_EXAMPLE_SERVER, LLAMA_EXAMPLE_CLI, LLAMA_EXAMPLE_DOWNLOAD}).set_env("LLAMA_ARG_MMPROJ_AUTO"));
add_opt(common_arg(
{"--mmproj-offload"},
{"--no-mmproj-offload"},
@@ -2611,14 +2619,14 @@ common_params_context common_params_parser_init(common_params & params, llama_ex
[](common_params & params, const std::string & value) {
params.model.path = value;
}
).set_examples({LLAMA_EXAMPLE_COMMON, LLAMA_EXAMPLE_EXPORT_LORA}).set_env("LLAMA_ARG_MODEL"));
).set_examples({LLAMA_EXAMPLE_COMMON, LLAMA_EXAMPLE_EXPORT_LORA, LLAMA_EXAMPLE_DOWNLOAD}).set_env("LLAMA_ARG_MODEL"));
add_opt(common_arg(
{"-mu", "--model-url"}, "MODEL_URL",
"model download url (default: unused)",
[](common_params & params, const std::string & value) {
params.model.url = value;
}
).set_env("LLAMA_ARG_MODEL_URL"));
).set_examples({LLAMA_EXAMPLE_COMMON, LLAMA_EXAMPLE_DOWNLOAD}).set_env("LLAMA_ARG_MODEL_URL"));
add_opt(common_arg(
{ "-dr", "--docker-repo" }, "[<repo>/]<model>[:quant]",
"Docker Hub model repository. repo is optional, default to ai/. quant is optional, default to :latest.\n"
@@ -2627,7 +2635,7 @@ common_params_context common_params_parser_init(common_params & params, llama_ex
[](common_params & params, const std::string & value) {
params.model.docker_repo = value;
}
).set_env("LLAMA_ARG_DOCKER_REPO"));
).set_examples({LLAMA_EXAMPLE_COMMON, LLAMA_EXAMPLE_DOWNLOAD}).set_env("LLAMA_ARG_DOCKER_REPO"));
add_opt(common_arg(
{"-hf", "-hfr", "--hf-repo"}, "<user>/<model>[:quant]",
"Hugging Face model repository; quant is optional, case-insensitive, default to Q4_K_M, or falls back to the first file in the repo if Q4_K_M doesn't exist.\n"
@@ -2637,14 +2645,14 @@ common_params_context common_params_parser_init(common_params & params, llama_ex
[](common_params & params, const std::string & value) {
params.model.hf_repo = value;
}
).set_env("LLAMA_ARG_HF_REPO"));
).set_examples({LLAMA_EXAMPLE_COMMON, LLAMA_EXAMPLE_DOWNLOAD}).set_env("LLAMA_ARG_HF_REPO"));
add_opt(common_arg(
{"-hff", "--hf-file"}, "FILE",
"Hugging Face model file. If specified, it will override the quant in --hf-repo (default: unused)",
[](common_params & params, const std::string & value) {
params.model.hf_file = value;
}
).set_env("LLAMA_ARG_HF_FILE"));
).set_examples({LLAMA_EXAMPLE_COMMON, LLAMA_EXAMPLE_DOWNLOAD}).set_env("LLAMA_ARG_HF_FILE"));
add_opt(common_arg(
{"-hfv", "-hfrv", "--hf-repo-v"}, "<user>/<model>[:quant]",
"Hugging Face model repository for the vocoder model (default: unused)",
@@ -2665,7 +2673,14 @@ common_params_context common_params_parser_init(common_params & params, llama_ex
[](common_params & params, const std::string & value) {
params.hf_token = value;
}
).set_env("HF_TOKEN"));
).set_examples({LLAMA_EXAMPLE_COMMON, LLAMA_EXAMPLE_DOWNLOAD}).set_env("HF_TOKEN"));
add_opt(common_arg(
{"--mtp"},
"also download the multi-token prediction (MTP) head, if available (default: unused)",
[](common_params & params) {
params.speculative.types.push_back(COMMON_SPECULATIVE_TYPE_DRAFT_MTP);
}
).set_examples({LLAMA_EXAMPLE_DOWNLOAD}));
add_opt(common_arg(
{"--context-file"}, "FNAME",
"file to load context from (use comma-separated values to specify multiple files)",