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

[1/N][Attention] Restructure attention: move files (#31916)

Browse Source 
Signed-off-by: Matthew Bonanni <mbonanni@redhat.com>
This commit is contained in:
Matthew Bonanni
2026-01-09 16:10:24 -05:00
committed by GitHub
parent 1f8b7c536b
commit 2612ba9285
195 changed files with 426 additions and 396 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
tests/kernels/attention/test_rocm_attention_selector.py
View File

@@ -4,10 +4,10 @@
import pytest
import torch
from vllm.attention.backends.registry import AttentionBackendEnum
from vllm.attention.selector import _cached_get_attn_backend, get_attn_backend
from vllm.config import AttentionConfig, VllmConfig, set_current_vllm_config
from vllm.platforms.rocm import RocmPlatform
from vllm.v1.attention.backends.registry import AttentionBackendEnum
from vllm.v1.attention.selector import _cached_get_attn_backend, get_attn_backend
@pytest.fixture(autouse=True)
@@ -19,7 +19,7 @@ def clear_cache():
@pytest.mark.skip(reason="Skipped for now. Should be revisited.")
def test_selector(monkeypatch: pytest.MonkeyPatch):
# Set the current platform to ROCm using monkeypatch
monkeypatch.setattr("vllm.attention.selector.current_platform", RocmPlatform())
monkeypatch.setattr("vllm.v1.attention.selector.current_platform", RocmPlatform())
# Test standard ROCm attention
attention_config = AttentionConfig(backend=AttentionBackendEnum.ROCM_ATTN)