Skip to content

Latest commit

 

History

History
14 lines (10 loc) · 694 Bytes

File metadata and controls

14 lines (10 loc) · 694 Bytes

Observability

As of today, observability, via Grafana dashboards, is considered to be outside of the scope for llm-d-benchmark. Please refer to the installation guide on llm-d-deployer for instructions on how to enable it.

Examples

These plots, automatically generated, were used to showcase the difference between a baseline vLLM deployment and llm-d (for models Llama 4 Scout and Lllama 3.1 70B)

vllm vs llm-d comparison