CS348Project/llama.cpp
5
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-10-30 08:42:00 +00:00
Code Issues Packages Projects Releases Wiki Activity

Forward decl minja::chat_template to avoid eager json dep

Browse Source
This commit is contained in:
ochafik
2025-01-18 10:37:56 +00:00
parent ee1e10e21e
commit e63520f37a
5 changed files with 40 additions and 21 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
examples/run/run.cpp
View File

@@ -26,6 +26,7 @@
#include "common.h"
#include "json.hpp"
#include "llama-cpp.h"
#include "chat-template.hpp"
#if defined(__unix__) || (defined(__APPLE__) && defined(__MACH__)) || defined(_WIN32)
[[noreturn]] static void sigint_handler(int) {
@@ -936,6 +937,7 @@ static int chat_loop(LlamaData & llama_data, const std::string & user, bool use_
int prev_len = 0;
llama_data.fmtted.resize(llama_n_ctx(llama_data.context.get()));
auto chat_templates = llama_chat_templates_from_model(llama_data.model.get(), "");
GGML_ASSERT(chat_templates.default_template);
static const bool stdout_a_terminal = is_stdout_a_terminal();
while (true) {
// Get user input
@@ -946,7 +948,7 @@ static int chat_loop(LlamaData & llama_data, const std::string & user, bool use_
add_message("user", user.empty() ? user_input : user, llama_data);
int new_len;
if (apply_chat_template_with_error_handling(chat_templates.default_template, llama_data, true, new_len, use_jinja) < 0) {
if (apply_chat_template_with_error_handling(*chat_templates.default_template, llama_data, true, new_len, use_jinja) < 0) {
return 1;
}
@@ -961,7 +963,7 @@ static int chat_loop(LlamaData & llama_data, const std::string & user, bool use_
}
add_message("assistant", response, llama_data);
if (apply_chat_template_with_error_handling(chat_templates.default_template, llama_data, false, prev_len, use_jinja) < 0) {
if (apply_chat_template_with_error_handling(*chat_templates.default_template, llama_data, false, prev_len, use_jinja) < 0) {
return 1;
}
}