]> git.djapps.eu Git - pkg/ggml/sources/llama.cpp/commit
overview / pkg / ggml / sources / llama.cpp / commit
summary | shortlog | log | commit | commitdiff | tree
(parent: 3e7bd4f)
server : fix reasoning budget WebUI precedence over model.ini (#24517)
authorGeorgi Gerganov <redacted>
Fri, 12 Jun 2026 14:59:56 +0000 (17:59 +0300)
committerGitHub <redacted>
Fri, 12 Jun 2026 14:59:56 +0000 (17:59 +0300)
commitebc10770ac5a9331824c53ef0c6adad780904dc3
tree7885d7262a1786122ff8d0d737809e058ffc69aetree
parent3e7bd4f39ac59167f82103e1fc22dc4585c489d3commit | diff
server : fix reasoning budget WebUI precedence over model.ini (#24517)

When reasoning-budget is set in model.ini, the per-request
thinking_budget_tokens from the WebUI was ignored because the
model.ini value took unconditional precedence.

Swap the precedence so the WebUI per-request value is checked
first, with the model.ini value serving as a fallback default.

Assisted-by: pi:llama.cpp/Qwen3.6-27B
tools/server/server-common.cpp diff | blob | history
Packaging of ggml-org/llama.cpp