biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

[V1][Metrics] add support for kv event publishing (#16750)

Browse Source 
Signed-off-by: alec-flowers <aflowers@nvidia.com>
Signed-off-by: Mark McLoughlin <markmc@redhat.com>
Co-authored-by: Mark McLoughlin <markmc@redhat.com>
This commit is contained in:
Alec
2025-04-30 16:44:45 +02:00
committed by GitHub
parent 77073c77bc
commit 0be6d05b5e
15 changed files with 1185 additions and 53 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
tests/v1/engine/conftest.py
View File

@@ -13,6 +13,8 @@ from tests.v1.engine.utils import (NUM_PROMPT_LOGPROBS_UNDER_TEST,
from vllm.engine.arg_utils import EngineArgs
from vllm.transformers_utils.tokenizer_group import init_tokenizer_from_configs
from ...distributed.conftest import publisher_config, random_port # noqa: F401
from tests.v1.engine.utils import FULL_STRINGS # isort: skip
EngineCoreSampleLogprobsType = list[tuple[torch.Tensor, torch.Tensor]]