Support vllm:max_num_generation_tokens metrics#250

Merged

github-actions[bot] merged 7 commits intollm-d:mainfrom

mayabar:max_num_generation_tokens

Nov 4, 2025

Commits on Nov 3, 2025

Add initial support for vllm:max_num_generation_tokens metrics, since we never return responses with more than one choice, the implementation is basic. Once 'n' request property will be supported …
mayabar
committed
update readme
mayabar
committed
Fix and extend 'Simulator configuration' test: add expected error message and fix, fix arguments in invalid configuration tests.
mayabar
committed