[SERVER] Added documentation for `parallel_tool_calls` param#15646

Closed

ExtReMLapin wants to merge 15 commits intoggml-org:masterfrom

ExtReMLapin:parallel_tool_calls_documentation

Commits on Aug 28, 2025

Update chat.cpp to support (at least) qwen3 + tool_choice = required

ExtReMLapin
authored and
Pierre F
committed
refactored changes to follow string tern op

ExtReMLapin
authored and
Pierre F
committed
fixing editorconfig-checker CI (tailing whitespace)

ExtReMLapin
authored and
Pierre F
committed
hermes 2 pro tool calling, better support for thinking (thinking tag already opened grammar)
Pierre F
committed
qwen hermes tool calling : fixed grammar rules names
Pierre F
committed
fixed really weird grammar crash `Unexpected empty grammar stack after accepting piece:`
Pierre F
committed
also apply the hotcrashfix here, just in case
Pierre F
committed
reverted changes done to grammar_lazy for hermes 2
Pierre F
committed
if there is enable_thinking enabled but hermes model doesn't support it, just disable it, don't GGML_ABORT
Pierre F
committed
fix thinking-content eating closing think tag | ref #8953

CNE FICHEPOIL Pierre
authored and
Pierre F
committed
removed `?` from grammar as it doesn't crash on linux, probably worth it's own issue

CNE FICHEPOIL Pierre
authored and
Pierre F
committed
ggml-cpu: fix invalid hsum build in debug s390x (#15634)

taronaeo
authored and
Pierre F
committed
CUDA: add conv2d (#15635)

mnehete32
authored and
Pierre F
committed
fixed crash with "auto" mode, trigger was missing
Pierre F
committed
Added documentation for `parallel_tool_calls` param
Pierre F
committed