Skip to content

feat: Implements retrying logic for downloading models using --model-url flag #9255

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 8 commits into from
Sep 11, 2024
Merged

feat: Implements retrying logic for downloading models using --model-url flag #9255

Changes from 4 commits
Commits
Show all changes
8 commits
Select commit Hold shift + click to select a range
951084c
feat: Implements retrying logic for downloading models using --model-…
farbodbj Aug 30, 2024
3a957ec
Update common/common.cpp
farbodbj Aug 31, 2024
23eba9b
Update common/common.cpp
farbodbj Aug 31, 2024
192d4df
apply comments
farbodbj Aug 31, 2024
c8f2890
implements a retry function to avoid duplication
farbodbj Sep 6, 2024
48de4da
fix editorconfig
slaren Sep 7, 2024
7272c6f
change function name
ngxson Sep 10, 2024
f4c6761
Merge branch 'master' into add-retry-to-model-download
ngxson Sep 10, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
44 changes: 34 additions & 10 deletions common/common.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2253,6 +2253,10 @@ struct llama_context_params llama_context_params_from_gpt_params(const gpt_param

#ifdef LLAMA_USE_CURL

#define CURL_MAX_RETRY 3
#define CURL_RETRY_DELAY_SECONDS 2


static bool starts_with(const std::string & str, const std::string & prefix) {
// While we wait for C++20's std::string::starts_with...
return str.rfind(prefix, 0) == 0;
Expand Down Expand Up @@ -2361,10 +2365,20 @@ static bool llama_download_file(const std::string & url, const std::string & pat
curl_easy_setopt(curl.get(), CURLOPT_HEADERFUNCTION, static_cast<CURLOPT_HEADERFUNCTION_PTR>(header_callback));
curl_easy_setopt(curl.get(), CURLOPT_HEADERDATA, &headers);

CURLcode res = curl_easy_perform(curl.get());
if (res != CURLE_OK) {
fprintf(stderr, "%s: curl_easy_perform() failed: %s\n", __func__, curl_easy_strerror(res));
return false;
int remaining_request_attempts = CURL_MAX_RETRY;
while (remaining_request_attempts > 0){
CURLcode res = curl_easy_perform(curl.get());
if (res != CURLE_OK) {
int exponential_backoff_delay = std::pow(CURL_RETRY_DELAY_SECONDS, (CURL_MAX_RETRY - remaining_request_attempts)) * 1000;
fprintf(stderr, "%s: curl_easy_perform() failed: %s, retrying after %d miliseconds\n", __func__, curl_easy_strerror(res), exponential_backoff_delay);
remaining_request_attempts--;
std::this_thread::sleep_for(std::chrono::milliseconds(exponential_backoff_delay));
} else if (remaining_request_attempts <= 0) {
fprintf(stderr, "%s: curl_easy_perform() failed\n", __func__);
return false;
} else {
break;
}
}

long http_code = 0;
Expand Down Expand Up @@ -2438,12 +2452,22 @@ static bool llama_download_file(const std::string & url, const std::string & pat
};

// start the download
fprintf(stderr, "%s: downloading from %s to %s (server_etag:%s, server_last_modified:%s)...\n", __func__,
llama_download_hide_password_in_url(url).c_str(), path.c_str(), headers.etag.c_str(), headers.last_modified.c_str());
auto res = curl_easy_perform(curl.get());
if (res != CURLE_OK) {
fprintf(stderr, "%s: curl_easy_perform() failed: %s\n", __func__, curl_easy_strerror(res));
return false;
int remaining_attempts = CURL_MAX_RETRY;
while (remaining_attempts > 0){
fprintf(stderr, "%s: trying to download model from %s to %s (server_etag:%s, server_last_modified:%s)...\n", __func__,
llama_download_hide_password_in_url(url).c_str(), path.c_str(), headers.etag.c_str(), headers.last_modified.c_str());
auto res = curl_easy_perform(curl.get());
if (res != CURLE_OK) {
int exponential_backoff_delay = std::pow(CURL_RETRY_DELAY_SECONDS, (CURL_MAX_RETRY - remaining_attempts)) * 1000;
fprintf(stderr, "\n%s: curl_easy_perform() failed: %s, retrying after %d miliseconnds\n", __func__, curl_easy_strerror(res), exponential_backoff_delay);
remaining_attempts--;
std::this_thread::sleep_for(std::chrono::milliseconds(exponential_backoff_delay));
} else if (remaining_attempts <= 0) {
fprintf(stderr, "%s: curl_easy_perform() failed\n", __func__);
return false;
} else {
break;
}
}

long http_code = 0;
Expand Down
Loading