server: expose the reasoning effort levels of the chat template (#31271)

* server: expose the reasoning effort levels of the chat template

The capability probe renders a user turn for each candidate level
(minimal, low, medium, high, xhigh, max) and for an unknown level. A
level is honored when it renders without raising and differently from
the unknown level, which shows the template fallback. Levels sharing a
render are aliases and the highest one is kept.

The caps serialize to JSON, and chat_template_caps in /props reports
the list as reasoning_effort_levels, from lowest to highest.

* jinja: keep the caps free of the json dependency

The caps keep reporting their flags as a map, and the effort levels
stay a plain member. common_chat_templates_get_caps() builds the JSON
the server reports from both, so jinja depends on json only for the
probe inputs.
This commit is contained in:
Pascal authored and GitHub committed 2026-10-27 22:39:37 +01:00
1 parent a919179942
commit 1b4554eff1
9 files changed
+90 -13

No files matched your search

+1 -1
View File
@@ -1721,7 +1721,7 @@ private:
{
auto caps = common_chat_templates_get_caps(chat_params.tmpls.get());
auto it = params_base.default_template_kwargs.find("preserve_reasoning");
bool supported = caps.at("supports_preserve_reasoning");
bool supported = caps.at("supports_preserve_reasoning").get<bool>();
bool specified = params_base.preserve_reasoning_specified;
// note: the kwarg is enabled by default if not specified explicitly, so check the value
bool enabled = it != params_base.default_template_kwargs.end() && it->second == "true";