Lumpiasty/llama.cpp
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

chat : pass reasoning_effort to template

Browse Source 
* chat: add reasoning_effort to common_chat_templates_inputs

Store OpenAI Chat Completions reasoning_effort and make it
available to jinja templates (with model specific translations
where required).

Assisted-by: llama.cpp:Muse-Glimmer-30B

* server : fixup reading reasoning effort from body

server_chat_convert_responses_to_chatcmpl already handles conversion of
Responses API reasoning.effort to reasoning_effort

* chat : expose reasoning effort

Assisted-by: Claude Opus 5

* chat : add reasoning_effort to generation_params

Assisted-by: Claude Opus 5

* chat : move reasoning_effort next to enable_thinking

Assisted-by: Claude Opus 5

* cont : mirror preserve_reasoning

* cont : pass context through analyze function

---------

Co-authored-by: Alde Rojas <hello@alde.dev>
This commit is contained in:
Andy Williams
2026-08-14 13:23:11 -05:00
committed by GitHub
co-authored by Alde Rojas
parent 9b05354ec6
commit 7e4c0a9688
9 changed files with 89 additions and 11 deletions
Download Patch File Download Diff File Expand all files Collapse all files
tests/test-chat.cpp
+19
View File
@@ -6955,6 +6955,24 @@ static void test_reasoning_budget_message_per_request() {
}
}
static void test_reasoning_effort_caps() {
LOG_DBG("%s\n", __func__);
auto assert_supports_effort = [](const std::string & path, bool expected) {
auto tmpls = read_templates(path);
assert_equals(expected, common_chat_templates_get_caps(tmpls.get()).at("supports_reasoning_effort"));
};
assert_supports_effort("models/templates/deepseek-ai-DeepSeek-V4.jinja", true);
assert_supports_effort("models/templates/muse-glimmer.jinja", true);
assert_supports_effort("models/templates/tencent-Hy3.jinja", true);
assert_supports_effort("models/templates/openai-gpt-oss-120b.jinja", true);
assert_supports_effort("models/templates/upstage-Solar-Open-100B.jinja", true);
assert_supports_effort("models/templates/Cohere2MoE.jinja", true);
assert_supports_effort("models/templates/meta-llama-Llama-3.1-8B-Instruct.jinja", false);
assert_supports_effort("models/templates/Qwen-Qwen3-0.6B.jinja", false);
}
static void test_msg_diffs_compute() {
LOG_DBG("%s\n", __func__);
{
@@ -7114,6 +7132,7 @@ int main(int argc, char ** argv) {
test_deepseek_v4_thinking_retention();
test_deepseek_v4_tool_result_ordering();
test_template_generation_prompt();
test_reasoning_effort_caps();
test_reasoning_budget_tokens_per_request();
test_reasoning_budget_message_per_request();
test_template_output_peg_parsers(detailed_debug);