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

[V0 Deprecation] Remove pooling model support in V0 (#23434)

Browse Source 
Signed-off-by: Woosuk Kwon <woosuk.kwon@berkeley.edu>
Signed-off-by: Max de Bayser <mbayser@br.ibm.com>
Co-authored-by: Woosuk Kwon <woosuk.kwon@berkeley.edu>
This commit is contained in:
Maximilien de Bayser
2025-08-29 04:04:02 -03:00
committed by GitHub
parent 934bebf192
commit 2554b27baa
38 changed files with 99 additions and 808 deletions
Download Patch File Download Diff File Expand all files Collapse all files

20
tests/models/language/pooling/test_embedding.py
View File

@@ -10,14 +10,6 @@ from vllm.platforms import current_platform
from ...utils import check_embeddings_close, check_transformers_version
@pytest.fixture(autouse=True)
def v1(run_with_both_engines):
# Simple autouse wrapper to run both engines for each test
# This can be promoted up to conftest.py to run for every
# test in a package
pass
@pytest.mark.parametrize(
"model",
[
@@ -32,21 +24,15 @@ def v1(run_with_both_engines):
"intfloat/e5-mistral-7b-instruct",
# CPU v1 doesn't support sliding window
marks=[pytest.mark.core_model]),
# the qwen models interfere with each other (see PR
# https://github.com/vllm-project/vllm/pull/18720).
# To avoid this problem, for now we skip v0 since it will be
# deprecated anyway.
pytest.param("ssmits/Qwen2-7B-Instruct-embed-base",
marks=[pytest.mark.skip_v0, pytest.mark.cpu_model]),
marks=[pytest.mark.cpu_model]),
# [Encoder-only]
pytest.param("BAAI/bge-base-en-v1.5", marks=[pytest.mark.core_model]),
pytest.param("sentence-transformers/all-MiniLM-L12-v2"),
pytest.param("intfloat/multilingual-e5-small"),
pytest.param("Alibaba-NLP/gte-Qwen2-1.5B-instruct",
marks=[pytest.mark.skip_v1]),
pytest.param("Alibaba-NLP/gte-Qwen2-1.5B-instruct"),
# [Cross-Encoder]
pytest.param("sentence-transformers/stsb-roberta-base-v2",
marks=[pytest.mark.skip_v1]),
pytest.param("sentence-transformers/stsb-roberta-base-v2"),
],
)
def test_models(