vllm

0b53bec6 - [DOC]: Add warning about max_num_batched_tokens and max_model_len when chunked prefill is disabled (#33109)