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

[V0 deprecation] Remove _VLLM_V1 suffixes from attention backend names (#25489)

Browse Source 
Signed-off-by: Matthew Bonanni <mbonanni@redhat.com>
Signed-off-by: Matthew Bonanni <mbonanni001@gmail.com>
This commit is contained in:
Matthew Bonanni
2025-09-25 13:37:50 -04:00
committed by GitHub
parent 71b25b0d48
commit 3468f17ebe
42 changed files with 131 additions and 174 deletions
Download Patch File Download Diff File Expand all files Collapse all files

7
tests/v1/e2e/test_cascade_attention.py
View File

@@ -9,11 +9,14 @@ from ...utils import create_new_process_for_each_test
@create_new_process_for_each_test()
@pytest.mark.parametrize("attn_backend",
["FLASH_ATTN_VLLM_V1", "FLASHINFER_VLLM_V1"])
@pytest.mark.parametrize("attn_backend", ["FLASH_ATTN", "FLASHINFER"])
def test_cascade_attention(example_system_message, monkeypatch, attn_backend):
prompt = "\n<User>: Implement fibonacci sequence in Python.\n<Claude>:"
if attn_backend == "FLASHINFER":
pytest.skip("This test is failing with FlashInfer backend and "
"needs investigation. See issue #25679.")
with monkeypatch.context() as m:
m.setenv("VLLM_USE_V1", "1")
m.setenv("VLLM_ATTENTION_BACKEND", attn_backend)