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

[Triton][Config] Add muls_add triton kernel and refactor AscendCompilationConfig (#5518)

Browse Source 
### What this PR does / why we need it?
Add muls_add triton kernel with related fusion pass. What's more, this
PR refactors `AscendCompilationConfig` and delete `NpugraphExConfig`.

### Does this PR introduce _any_ user-facing change?
None

### How was this patch tested?
CI passed with new added test.


- vLLM version: v0.13.0
- vLLM main:
45c1ca1ca1

---------

Signed-off-by: whx-sjtu <2952154980@qq.com>
This commit is contained in:
whx
2026-03-02 17:54:25 +08:00
committed by GitHub
parent 8547520726
commit 16c879cdf7
14 changed files with 290 additions and 98 deletions
Download Patch File Download Diff File Expand all files Collapse all files

18
tests/ut/test_ascend_config.py
View File

@@ -66,13 +66,11 @@ class TestAscendConfig(TestBase):
self.assertEqual(ascend_config.eplb_config.num_redundant_experts, 2)
self.assertTrue(ascend_config.multistream_overlap_shared_expert)
npugraph_ex_config = ascend_config.npugraph_ex_config
self.assertTrue(npugraph_ex_config.enable)
self.assertFalse(npugraph_ex_config.enable_static_kernel)
ascend_compilation_config = ascend_config.ascend_compilation_config
self.assertFalse(ascend_compilation_config.fuse_norm_quant)
self.assertFalse(ascend_config.enable_kv_nz)
self.assertTrue(ascend_compilation_config.enable_npugraph_ex)
self.assertFalse(ascend_compilation_config.enable_static_kernel)
ascend_fusion_config = ascend_config.ascend_fusion_config
self.assertFalse(ascend_fusion_config.fusion_ops_gmmswigluquant)
@@ -82,16 +80,16 @@ class TestAscendConfig(TestBase):
def test_init_ascend_config_enable_npugraph_ex(self, mock_fix_incompatible_config):
test_vllm_config = VllmConfig()
test_vllm_config.additional_config = {
"npugraph_ex_config": {
"enable": True,
"ascend_compilation_config": {
"enable_npugraph_ex": True,
"enable_static_kernel": True
},
"refresh": True
}
npugraph_ex_config = init_ascend_config(
test_vllm_config).npugraph_ex_config
self.assertTrue(npugraph_ex_config.enable)
self.assertTrue(npugraph_ex_config.enable_static_kernel)
ascend_compilation_config = init_ascend_config(
test_vllm_config).ascend_compilation_config
self.assertTrue(ascend_compilation_config.enable_npugraph_ex)
self.assertTrue(ascend_compilation_config.enable_static_kernel)
@_clean_up_ascend_config
@patch("vllm_ascend.platform.NPUPlatform._fix_incompatible_config")