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

Convert formatting to use ruff instead of yapf + isort (#26247)

Browse Source 
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
This commit is contained in:
Harry Mellor
2025-10-05 15:06:22 +01:00
committed by GitHub
parent 17edd8a807
commit d6953beb91
1508 changed files with 115244 additions and 94146 deletions
Download Patch File Download Diff File Expand all files Collapse all files

36
tests/models/language/pooling/test_embedding.py
View File

@@ -18,20 +18,25 @@ from ...utils import check_embeddings_close
# case won't pass because gte-Qwen2-1.5B-instruct will cache custom
# model code with bidirectional attention.
# [Decoder-only]
pytest.param("BAAI/bge-multilingual-gemma2",
marks=[pytest.mark.core_model, pytest.mark.slow_test]),
pytest.param(
"BAAI/bge-multilingual-gemma2",
marks=[pytest.mark.core_model, pytest.mark.slow_test],
),
pytest.param(
"intfloat/e5-mistral-7b-instruct",
# CPU v1 doesn't support sliding window
marks=[pytest.mark.core_model]),
pytest.param("ssmits/Qwen2-7B-Instruct-embed-base",
marks=[pytest.mark.cpu_model]),
marks=[pytest.mark.core_model],
),
pytest.param(
"ssmits/Qwen2-7B-Instruct-embed-base", marks=[pytest.mark.cpu_model]
),
# [Encoder-only]
pytest.param(
"BAAI/bge-base-en-v1.5",
marks=[
pytest.mark.core_model, pytest.mark.cpu_model,
pytest.mark.slow_test
pytest.mark.core_model,
pytest.mark.cpu_model,
pytest.mark.slow_test,
],
),
pytest.param("sentence-transformers/all-MiniLM-L12-v2"),
@@ -50,7 +55,6 @@ def test_models(
model,
monkeypatch,
) -> None:
if model == "BAAI/bge-multilingual-gemma2" and current_platform.is_rocm():
# ROCm Triton FA does not currently support sliding window attention
# switch to use ROCm CK FA backend
@@ -58,13 +62,14 @@ def test_models(
vllm_extra_kwargs = {}
if model == "ssmits/Qwen2-7B-Instruct-embed-base":
vllm_extra_kwargs["pooler_config"] = \
PoolerConfig(pooling_type="MEAN", normalize=False)
vllm_extra_kwargs["pooler_config"] = PoolerConfig(
pooling_type="MEAN", normalize=False
)
max_model_len: Optional[int] = 512
if model in [
"sentence-transformers/all-MiniLM-L12-v2",
"sentence-transformers/stsb-roberta-base-v2"
"sentence-transformers/all-MiniLM-L12-v2",
"sentence-transformers/stsb-roberta-base-v2",
]:
max_model_len = None
@@ -79,10 +84,9 @@ def test_models(
with hf_runner(model, is_sentence_transformer=True) as hf_model:
hf_outputs = hf_model.encode(example_prompts)
with vllm_runner(model,
runner="pooling",
max_model_len=max_model_len,
**vllm_extra_kwargs) as vllm_model:
with vllm_runner(
model, runner="pooling", max_model_len=max_model_len, **vllm_extra_kwargs
) as vllm_model:
vllm_outputs = vllm_model.embed(example_prompts)
check_embeddings_close(