vllm

f790ad3c - [Frontend][OpenAI] Support for returning max_model_len on /v1/models response (#4643)