mirror of
https://git.datalinker.icu/vllm-project/vllm.git
synced 2025-12-10 18:05:49 +08:00
[Misc] Add penalties sampling parameters to serve tool (#25974)
Signed-off-by: Sergei Skvortsov <sergeyskv@nebius.com> Co-authored-by: Sergei Skvortsov <sergeyskv@nebius.com>
This commit is contained in:
parent
75003f34e8
commit
b71fcd4905
@ -1077,6 +1077,27 @@ def add_cli_args(parser: argparse.ArgumentParser):
|
|||||||
"openai-compatible backends. If not specified, default to greedy "
|
"openai-compatible backends. If not specified, default to greedy "
|
||||||
"decoding (i.e. temperature==0.0).",
|
"decoding (i.e. temperature==0.0).",
|
||||||
)
|
)
|
||||||
|
sampling_group.add_argument(
|
||||||
|
"--frequency-penalty",
|
||||||
|
type=float,
|
||||||
|
default=None,
|
||||||
|
help="Frequency penalty sampling parameter. Only has effect on "
|
||||||
|
"openai-compatible backends.",
|
||||||
|
)
|
||||||
|
sampling_group.add_argument(
|
||||||
|
"--presence-penalty",
|
||||||
|
type=float,
|
||||||
|
default=None,
|
||||||
|
help="Presence penalty sampling parameter. Only has effect on "
|
||||||
|
"openai-compatible backends.",
|
||||||
|
)
|
||||||
|
sampling_group.add_argument(
|
||||||
|
"--repetition-penalty",
|
||||||
|
type=float,
|
||||||
|
default=None,
|
||||||
|
help="Repetition penalty sampling parameter. Only has effect on "
|
||||||
|
"openai-compatible backends.",
|
||||||
|
)
|
||||||
|
|
||||||
parser.add_argument(
|
parser.add_argument(
|
||||||
'--tokenizer-mode',
|
'--tokenizer-mode',
|
||||||
@ -1211,6 +1232,9 @@ async def main_async(args: argparse.Namespace) -> dict[str, Any]:
|
|||||||
"top_k": args.top_k,
|
"top_k": args.top_k,
|
||||||
"min_p": args.min_p,
|
"min_p": args.min_p,
|
||||||
"temperature": args.temperature,
|
"temperature": args.temperature,
|
||||||
|
"frequency_penalty": args.frequency_penalty,
|
||||||
|
"presence_penalty": args.presence_penalty,
|
||||||
|
"repetition_penalty": args.repetition_penalty,
|
||||||
}.items() if v is not None
|
}.items() if v is not None
|
||||||
}
|
}
|
||||||
|
|
||||||
|
|||||||
Loading…
x
Reference in New Issue
Block a user