File parse_engine_stats.py

benchmarking/vllm/parse_engine_stats.py:None–None · view source on GitHub ↗

Source from the content-addressed store, hash-verified

1	"""Extract KV-cache occupancy, running reqs, and waiting reqs from vLLM sweep logs.
2
3	Usage:
4	python parse_engine_stats.py <sweep.log> [--by-concurrency]

nothing calls this directly

mainFunction · 0.70

no test coverage detected