Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[misc] soft drop beam search #8763

Merged
merged 3 commits into from
Sep 24, 2024
Merged

[misc] soft drop beam search #8763

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
ab20ca2
add error
youkaichao Sep 24, 2024
40ae429
use value error
youkaichao Sep 24, 2024
6242afb
trigger
youkaichao Sep 24, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
5 changes: 5 additions & 0 deletions vllm/envs.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -62,6 +62,7 @@
VLLM_TORCH_PROFILER_DIR: Optional[str] = None
VLLM_USE_TRITON_AWQ: bool = False
VLLM_ALLOW_RUNTIME_LORA_UPDATING: bool = False
VLLM_ALLOW_DEPRECATED_BEAM_SEARCH: bool = False


def get_default_cache_root():
Expand Down Expand Up @@ -195,6 +196,10 @@ def get_default_config_root():
lambda: (os.environ.get("VLLM_USE_TRITON_FLASH_ATTN", "True").lower() in
("true", "1")),

# If set, allowing the use of deprecated beam search implementation
"VLLM_ALLOW_DEPRECATED_BEAM_SEARCH":
lambda: os.environ.get("VLLM_ALLOW_DEPRECATED_BEAM_SEARCH", "0") == "1",

# Internal flag to enable Dynamo graph capture
"VLLM_TEST_DYNAMO_GRAPH_CAPTURE":
lambda: int(os.environ.get("VLLM_TEST_DYNAMO_GRAPH_CAPTURE", "0")),
Expand Down
5 changes: 5 additions & 0 deletions vllm/sampling_params.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -8,6 +8,7 @@
import torch
from typing_extensions import Annotated

import vllm.envs as envs
from vllm.logger import init_logger

logger = init_logger(__name__)
Expand Down Expand Up @@ -260,6 +261,10 @@ def __post_init__(self) -> None:

self._verify_args()
if self.use_beam_search:
if not envs.VLLM_ALLOW_DEPRECATED_BEAM_SEARCH:
raise ValueError(
"Using beam search as a sampling parameter is deprecated, and will be removed in the future release. Please use the `vllm.LLM.use_beam_search` method for dedicated beam search instead, or set the environment variable `VLLM_ALLOW_DEPRECATED_BEAM_SEARCH=1` to suppress this error. For more details, see https://github.com/vllm-project/vllm/issues/8306 ." # noqa
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Do we provide openAI-compatible server alternative for beam search (not sure if this is a valid need and feel free to ignore this comment)?

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

no. see the future work section at #8726 (comment)

)
self._verify_beam_search()
else:
self._verify_non_beam_search()
Expand Down
Loading