[CI/Build][AMD] Enable Entrypoints Integration Test (Pooling) to run without error on ROCm (#29212)

Signed-off-by: Randall Smith <ransmith@amd.com>
Co-authored-by: Randall Smith <ransmith@amd.com>
This commit is contained in:
rasmith
2025-11-21 20:13:18 -06:00
committed by GitHub
parent 052950e5b3
commit 6f403501a0
11 changed files with 66 additions and 0 deletions

View File

@@ -12,6 +12,12 @@ from tests.models.language.pooling.embed_utils import run_embedding_correctness_
from tests.models.utils import EmbedModelInfo
from tests.utils import RemoteOpenAIServer
from vllm.entrypoints.openai.protocol import EmbeddingResponse
from vllm.platforms import current_platform
if current_platform.is_rocm():
pytest.skip(
"Encoder self-attention is not implemented on ROCm.", allow_module_level=True
)
MODELS = [
EmbedModelInfo("intfloat/multilingual-e5-small", is_matryoshka=False),