CS348Project/llama.cpp
5
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-10-31 08:51:55 +00:00
Code Issues Packages Projects Releases Wiki Activity

server : fallback to chatml, add AlphaMonarch chat template (#5628)

Browse Source 
* server: fallback to chatml

* add new chat template

* server: add AlphaMonarch to test chat template

* server: only check model template if there is no custom tmpl

* remove TODO
This commit is contained in:
Xuan Son Nguyen
2024-02-22 09:33:24 +01:00
committed by GitHub
parent c5688c6250
commit a46f50747b
3 changed files with 39 additions and 8 deletions
Download Patch File Download Diff File Expand all files Collapse all files

9
llama.cpp
View File

@@ -12773,6 +12773,15 @@ static int32_t llama_chat_apply_template_internal(
if (add_ass) {
ss << "<|assistant|>\n";
}
} else if (tmpl.find("bos_token + message['role']") != std::string::npos) {
// mlabonne/AlphaMonarch-7B template (the <s> is included inside history)
for (auto message : chat) {
std::string bos = (message == chat.front()) ? "" : "<s>"; // skip BOS for first message
ss << bos << message->role << "\n" << message->content << "</s>\n";
}
if (add_ass) {
ss << "<s>assistant\n";
}
} else {
// template not supported
return -1;