xinyun/vllm
1
0
Fork 0
mirror of https://git.datalinker.icu/vllm-project/vllm.git synced 2025-12-18 00:06:03 +08:00
Code Issues Packages Projects Releases Wiki Activity

[Misc] Add deprecation warning for beam search (#6402)

Browse Source
This commit is contained in:
Woosuk Kwon 2024-07-13 11:52:22 -07:00 committed by GitHub
parent babf52dade
commit eeceadaecc
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
2 changed files with 17 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

5
vllm/envs.py
View File

@ -41,6 +41,7 @@ if TYPE_CHECKING:
NVCC_THREADS: Optional[str] = None NVCC_THREADS: Optional[str] = None
VLLM_USE_PRECOMPILED: bool = False VLLM_USE_PRECOMPILED: bool = False
VLLM_INSTALL_PUNICA_KERNELS: bool = False VLLM_INSTALL_PUNICA_KERNELS: bool = False
VLLM_NO_DEPRECATION_WARNING: bool = False
CMAKE_BUILD_TYPE: Optional[str] = None CMAKE_BUILD_TYPE: Optional[str] = None
VERBOSE: bool = False VERBOSE: bool = False
@ -251,6 +252,10 @@ environment_variables: Dict[str, Callable[[], Any]] = {
lambda: os.getenv("VLLM_XLA_CACHE_PATH", "~/.vllm/xla_cache/"), lambda: os.getenv("VLLM_XLA_CACHE_PATH", "~/.vllm/xla_cache/"),
"VLLM_FUSED_MOE_CHUNK_SIZE": "VLLM_FUSED_MOE_CHUNK_SIZE":
lambda: int(os.getenv("VLLM_FUSED_MOE_CHUNK_SIZE", "65536")), lambda: int(os.getenv("VLLM_FUSED_MOE_CHUNK_SIZE", "65536")),
# If set, vllm will skip the deprecation warnings.
"VLLM_NO_DEPRECATION_WARNING":
lambda: bool(int(os.getenv("VLLM_NO_DEPRECATION_WARNING", "0"))),
} }
# end-env-vars-definition # end-env-vars-definition

12
vllm/sampling_params.py
View File

@ -8,6 +8,11 @@ import torch
from pydantic import Field from pydantic import Field
from typing_extensions import Annotated from typing_extensions import Annotated
import vllm.envs as envs
from vllm.logger import init_logger
logger = init_logger(__name__)
_SAMPLING_EPS = 1e-5 _SAMPLING_EPS = 1e-5
@ -184,6 +189,13 @@ class SamplingParams:
self._verify_args() self._verify_args()
if self.use_beam_search: if self.use_beam_search:
if not envs.VLLM_NO_DEPRECATION_WARNING:
logger.warning(
"[IMPORTANT] We plan to discontinue the support for beam "
"search in the next major release. Please refer to "
"https://github.com/vllm-project/vllm/issues/6226 for "
"more information. Set VLLM_NO_DEPRECATION_WARNING=1 to "
"suppress this warning.")
self._verify_beam_search() self._verify_beam_search()
else: else:
self._verify_non_beam_search() self._verify_non_beam_search()