Skip to content

[Perf] Add new npu_fused_infer_attention_score op to improve perfomance in splitfuse cases and resolve long-seq mask problems #12871

[Perf] Add new npu_fused_infer_attention_score op to improve perfomance in splitfuse cases and resolve long-seq mask problems

[Perf] Add new npu_fused_infer_attention_score op to improve perfomance in splitfuse cases and resolve long-seq mask problems #12871

Triggered via pull request September 19, 2025 14:57
Status Failure
Total duration 6m 29s
Artifacts

vllm_ascend_test.yaml

on: pull_request
Matrix: singlecard e2e test - light
Matrix: unit test
Matrix: multicard e2e test - light
Fit to window
Zoom out
Zoom in

Annotations

10 errors
lint / pre-commit: vllm_ascend/attention/attention_v1.py#L467
"None" has no attribute "shape" [attr-defined]
lint / pre-commit: vllm_ascend/attention/attention_v1.py#L39
Module "vllm_ascend.utils" has no attribute "verify_torch_npu_version" [attr-defined]
lint / pre-commit: vllm_ascend/attention/attention_v1.py#L469
"None" has no attribute "view" [attr-defined]
lint / pre-commit: vllm_ascend/attention/attention_v1.py#L468
"None" has no attribute "view" [attr-defined]
lint / pre-commit: vllm_ascend/attention/attention_v1.py#L467
"None" has no attribute "shape" [attr-defined]
lint / pre-commit: vllm_ascend/attention/attention_v1.py#L39
Module "vllm_ascend.utils" has no attribute "verify_torch_npu_version" [attr-defined]
lint / pre-commit: vllm_ascend/attention/attention_v1.py#L469
"None" has no attribute "view" [attr-defined]
lint / pre-commit: vllm_ascend/attention/attention_v1.py#L468
"None" has no attribute "view" [attr-defined]
lint / pre-commit: vllm_ascend/attention/attention_v1.py#L467
"None" has no attribute "shape" [attr-defined]
lint / pre-commit: vllm_ascend/attention/attention_v1.py#L39
Module "vllm_ascend.utils" has no attribute "verify_torch_npu_version" [attr-defined]