vllm.entrypoints.cli.benchmark.latency

BenchmarkLatencySubcommand

Bases: BenchmarkSubcommandBase

The latency subcommand for vllm bench.

Source code in vllm/entrypoints/cli/benchmark/latency.py

class BenchmarkLatencySubcommand(BenchmarkSubcommandBase):
    """ The `latency` subcommand for vllm bench. """

    def __init__(self):
        self.name = "latency"
        super().__init__()

    @property
    def help(self) -> str:
        return "Benchmark the latency of a single batch of requests."

    def add_cli_args(self, parser: argparse.ArgumentParser) -> None:
        add_cli_args(parser)

    @staticmethod
    def cmd(args: argparse.Namespace) -> None:
        main(args)

help property

help: str

name instance-attribute

name = 'latency'

__init__

__init__()

Source code in vllm/entrypoints/cli/benchmark/latency.py

def __init__(self):
    self.name = "latency"
    super().__init__()

add_cli_args

add_cli_args(parser: ArgumentParser) -> None

Source code in vllm/entrypoints/cli/benchmark/latency.py

def add_cli_args(self, parser: argparse.ArgumentParser) -> None:
    add_cli_args(parser)

cmd staticmethod

cmd(args: Namespace) -> None

Source code in vllm/entrypoints/cli/benchmark/latency.py

@staticmethod
def cmd(args: argparse.Namespace) -> None:
    main(args)

cmd_init

cmd_init() -> list[CLISubcommand]

Source code in vllm/entrypoints/cli/benchmark/latency.py

def cmd_init() -> list[CLISubcommand]:
    return [BenchmarkLatencySubcommand()]