vllm.entrypoints.cli ¶
Modules:
| Name | Description |
|---|---|
benchmark | |
collect_env | |
main | The CLI entrypoints of vLLM |
openai | |
run_batch | |
serve | |
types | |
__all__ module-attribute ¶
__all__: list[str] = [
"BenchmarkLatencySubcommand",
"BenchmarkServingSubcommand",
"BenchmarkThroughputSubcommand",
]
BenchmarkLatencySubcommand ¶
Bases: BenchmarkSubcommandBase
The latency subcommand for vllm bench.
Source code in vllm/entrypoints/cli/benchmark/latency.py
help class-attribute instance-attribute ¶
add_cli_args classmethod ¶
add_cli_args(parser: ArgumentParser) -> None
BenchmarkServingSubcommand ¶
Bases: BenchmarkSubcommandBase
The serve subcommand for vllm bench.
Source code in vllm/entrypoints/cli/benchmark/serve.py
add_cli_args classmethod ¶
add_cli_args(parser: ArgumentParser) -> None
BenchmarkThroughputSubcommand ¶
Bases: BenchmarkSubcommandBase
The throughput subcommand for vllm bench.
Source code in vllm/entrypoints/cli/benchmark/throughput.py
add_cli_args classmethod ¶
add_cli_args(parser: ArgumentParser) -> None