tqcq/llama.cpp
0
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-06-27 20:05:20 +00:00
Code Issues Packages Projects Releases Wiki Activity

llama-run: add support for downloading models from ModelScope (#13370)

Browse Source 
Signed-off-by: Xiaodong Ye <xiaodong.ye@mthreads.com>
This commit is contained in:
R0CKSTAR
2025-05-09 17:25:50 +08:00
committed by GitHub
parent 2189fd3b63
commit 0527771dd8
2 changed files with 20 additions and 4 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
tools/run/README.md
View File

@ -42,6 +42,8 @@ Examples:
llama-run ollama://smollm:135m llama-run ollama://smollm:135m
llama-run hf://QuantFactory/SmolLM-135M-GGUF/SmolLM-135M.Q2_K.gguf llama-run hf://QuantFactory/SmolLM-135M-GGUF/SmolLM-135M.Q2_K.gguf
llama-run huggingface://bartowski/SmolLM-1.7B-Instruct-v0.2-GGUF/SmolLM-1.7B-Instruct-v0.2-IQ3_M.gguf llama-run huggingface://bartowski/SmolLM-1.7B-Instruct-v0.2-GGUF/SmolLM-1.7B-Instruct-v0.2-IQ3_M.gguf
llama-run ms://QuantFactory/SmolLM-135M-GGUF/SmolLM-135M.Q2_K.gguf
llama-run modelscope://bartowski/SmolLM-1.7B-Instruct-v0.2-GGUF/SmolLM-1.7B-Instruct-v0.2-IQ3_M.gguf
llama-run https://example.com/some-file1.gguf llama-run https://example.com/some-file1.gguf
llama-run some-file2.gguf llama-run some-file2.gguf
llama-run file://some-file3.gguf llama-run file://some-file3.gguf

22
tools/run/run.cpp
View File

@ -267,7 +267,7 @@ class Opt {
"Commands:\n" "Commands:\n"
" model\n" " model\n"
" Model is a string with an optional prefix of \n" " Model is a string with an optional prefix of \n"
" huggingface:// (hf://), ollama://, https:// or file://.\n" " huggingface:// (hf://), modelscope:// (ms://), ollama://, https:// or file://.\n"
" If no protocol is specified and a file exists in the specified\n" " If no protocol is specified and a file exists in the specified\n"
" path, file:// is assumed, otherwise if a file does not exist in\n" " path, file:// is assumed, otherwise if a file does not exist in\n"
" the specified path, ollama:// is assumed. Models that are being\n" " the specified path, ollama:// is assumed. Models that are being\n"
@ -282,6 +282,9 @@ class Opt {
" llama-run hf://QuantFactory/SmolLM-135M-GGUF/SmolLM-135M.Q2_K.gguf\n" " llama-run hf://QuantFactory/SmolLM-135M-GGUF/SmolLM-135M.Q2_K.gguf\n"
" llama-run " " llama-run "
"huggingface://bartowski/SmolLM-1.7B-Instruct-v0.2-GGUF/SmolLM-1.7B-Instruct-v0.2-IQ3_M.gguf\n" "huggingface://bartowski/SmolLM-1.7B-Instruct-v0.2-GGUF/SmolLM-1.7B-Instruct-v0.2-IQ3_M.gguf\n"
" llama-run ms://QuantFactory/SmolLM-135M-GGUF/SmolLM-135M.Q2_K.gguf\n"
" llama-run "
"modelscope://bartowski/SmolLM-1.7B-Instruct-v0.2-GGUF/SmolLM-1.7B-Instruct-v0.2-IQ3_M.gguf\n"
" llama-run https://example.com/some-file1.gguf\n" " llama-run https://example.com/some-file1.gguf\n"
" llama-run some-file2.gguf\n" " llama-run some-file2.gguf\n"
" llama-run file://some-file3.gguf\n" " llama-run file://some-file3.gguf\n"
@ -689,7 +692,7 @@ class LlamaData {
return 0; return 0;
} }
int huggingface_dl(std::string & model, const std::string & bn) { int dl_from_endpoint(std::string & model_endpoint, std::string & model, const std::string & bn) {
// Find the second occurrence of '/' after protocol string // Find the second occurrence of '/' after protocol string
size_t pos = model.find('/'); size_t pos = model.find('/');
pos = model.find('/', pos + 1); pos = model.find('/', pos + 1);
@ -697,8 +700,6 @@ class LlamaData {
std::vector<std::string> headers = { "User-Agent: llama-cpp", "Accept: application/json" }; std::vector<std::string> headers = { "User-Agent: llama-cpp", "Accept: application/json" };
std::string url; std::string url;
std::string model_endpoint = get_model_endpoint();
if (pos == std::string::npos) { if (pos == std::string::npos) {
auto [model_name, manifest_url] = extract_model_and_tag(model, model_endpoint + "v2/"); auto [model_name, manifest_url] = extract_model_and_tag(model, model_endpoint + "v2/");
hfr = model_name; hfr = model_name;
@ -720,6 +721,16 @@ class LlamaData {
return download(url, bn, true, headers); return download(url, bn, true, headers);
} }
int modelscope_dl(std::string & model, const std::string & bn) {
std::string model_endpoint = "https://modelscope.cn/models/";
return dl_from_endpoint(model_endpoint, model, bn);
}
int huggingface_dl(std::string & model, const std::string & bn) {
std::string model_endpoint = get_model_endpoint();
return dl_from_endpoint(model_endpoint, model, bn);
}
int ollama_dl(std::string & model, const std::string & bn) { int ollama_dl(std::string & model, const std::string & bn) {
const std::vector<std::string> headers = { "Accept: application/vnd.docker.distribution.manifest.v2+json" }; const std::vector<std::string> headers = { "Accept: application/vnd.docker.distribution.manifest.v2+json" };
if (model.find('/') == std::string::npos) { if (model.find('/') == std::string::npos) {
@ -837,6 +848,9 @@ class LlamaData {
rm_until_substring(model_, "hf.co/"); rm_until_substring(model_, "hf.co/");
rm_until_substring(model_, "://"); rm_until_substring(model_, "://");
ret = huggingface_dl(model_, bn); ret = huggingface_dl(model_, bn);
} else if (string_starts_with(model_, "ms://") || string_starts_with(model_, "modelscope://")) {
rm_until_substring(model_, "://");
ret = modelscope_dl(model_, bn);
} else if ((string_starts_with(model_, "https://") || string_starts_with(model_, "http://")) && } else if ((string_starts_with(model_, "https://") || string_starts_with(model_, "http://")) &&
!string_starts_with(model_, "https://ollama.com/library/")) { !string_starts_with(model_, "https://ollama.com/library/")) {
ret = download(model_, bn, true); ret = download(model_, bn, true);