A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 16, 20:45 UTC
[Misc] Remove deprecation warning for beam search (#6659)
Woosuk Kwon committed
Jul 22, 2024 at 17:21 UTC
bdf5fd1386846281acb92cf6fd01addc9c43d7a8
1 file changed
-13
vllm/sampling_params.py
-13
@@ -8,7 +8,6 @@ import torch
8
from pydantic import Field
9
from typing_extensions import Annotated
10
11
-import vllm.envs as envs
11
from vllm.logger import init_logger
12
13
logger = init_logger(__name__)
@@ -189,18 +188,6 @@ class SamplingParams:
188
189
self._verify_args()
190
if self.use_beam_search:
192
- # Lazy import to avoid circular imports.
193
- from vllm.usage.usage_lib import set_runtime_usage_data
194
- set_runtime_usage_data("use_beam_search", True)
195
-
196
- if not envs.VLLM_NO_DEPRECATION_WARNING:
197
- logger.warning(
198
- "[IMPORTANT] We plan to discontinue the support for beam "
199
- "search in the next major release. Please refer to "
200
- "https://github.com/vllm-project/vllm/issues/6226 for "
201
- "more information. Set VLLM_NO_DEPRECATION_WARNING=1 to "
202
- "suppress this warning.")
203
-
191
self._verify_beam_search()
192
else:
193
self._verify_non_beam_search()