EngineX/xc-llm-ascend
3
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

drop ascend scheduler (#4498)

Browse Source 
Ascend scheduler was added for non chunk prefill case before, since that
the npu ops didn't work well with chunked prefill.

Now the ops with chunked prefill work better, it's time to remove the
ascend scheduler to use vLLM default scheduler.

- vLLM version: v0.11.2

---------

Signed-off-by: wangxiyuan <wangxiyuan1007@gmail.com>
This commit is contained in:
wangxiyuan
2025-11-29 16:18:34 +08:00
committed by GitHub
parent 53a52d6614
commit f10acddb78
52 changed files with 85 additions and 2948 deletions
Download Patch File Download Diff File Expand all files Collapse all files

26
tests/ut/test_platform.py
View File

@@ -32,7 +32,6 @@ class TestNPUPlatform(TestBase):
def mock_vllm_ascend_config():
mock_ascend_config = MagicMock()
mock_ascend_config.torchair_graph_config.enabled = False
mock_ascend_config.ascend_scheduler_config.enabled = False
mock_ascend_config.enable_shared_expert_dp = False
return mock_ascend_config
@@ -522,31 +521,6 @@ class TestNPUPlatform(TestBase):
self.platform.check_and_update_config(vllm_config)
self.assertEqual(vllm_config.compilation_config.custom_ops, [])
@patch('vllm_ascend.utils.get_ascend_device_type',
return_value=AscendDeviceType._910_93)
@patch("vllm_ascend.ascend_config.check_ascend_config")
@patch("vllm_ascend.ascend_config.init_ascend_config")
@patch(
"vllm_ascend.core.recompute_schedule_config.RecomputeSchedulerConfig.initialize_from_config"
)
def test_check_and_update_config_ascend_scheduler_config(
self, mock_init_recompute, mock_init_ascend, mock_check_ascend,
mock_soc_version):
mock_ascend_config = TestNPUPlatform.mock_vllm_ascend_config()
mock_ascend_config.ascend_scheduler_config.enabled = True
mock_init_ascend.return_value = mock_ascend_config
vllm_config = TestNPUPlatform.mock_vllm_config()
vllm_config.parallel_config.tensor_parallel_size = 1
mock_init_recompute.return_value = MagicMock()
with patch("vllm_ascend.core.schedule_config.AscendSchedulerConfig"
) as mock_scheduler:
from vllm_ascend import platform
importlib.reload(platform)
self.platform.check_and_update_config(vllm_config)
mock_scheduler.initialize_from_config.assert_called_once()
@patch('vllm_ascend.platform.get_ascend_config')
def test_get_attn_backend_cls_use_v1_and_mla(self, mock_get_ascend_config):
mock_config = MagicMock()