@@ -1,4 +1,4 @@
|
||||
## Bench Serving Guide
|
||||
# Bench Serving Guide
|
||||
|
||||
This guide explains how to benchmark online serving throughput and latency using `python -m sglang.bench_serving`. It supports multiple inference backends via OpenAI-compatible and native endpoints, and produces both console metrics and optional JSONL outputs.
|
||||
|
||||
|
||||
Reference in New Issue
Block a user