Integrated Repositories

NPU runners have been integrated into the following repositories:

Repository	PyPI Cache	APT Cache	CCache	uv	Last Checked
vllm-project/vllm-ascend	✅	✅	-	-	2026-09-17
vllm-project/vllm-omni	-	-	-	-	2026-09-17
sgl-project/sgl-kernel-npu	✅	✅	✅	✅	2026-09-17
sgl-project/sglang	✅	✅	✅	✅	2026-09-17
verl-project/verl	✅	✅	❌	✅	2026-09-17
hiyouga/LlamaFactory	✅	✅	❌	✅	2026-09-17
modelscope/ms-swift	✅	✅	❌	✅	2026-09-17
triton-lang/triton-ascend	✅	✅	✅	❌	2026-09-17
vllm-ascend/vllm-ascend-recipes	✅	-	❌	❌	2026-09-17
verl-project/verl-omni	✅	-	❌	-	2026-09-17
Ascend/pytorch	❌	❌	✅	❌	2026-09-17
Ascend/sglang	✅	✅	❌	✅	2026-09-17
alibaba/ROLL	❌	-	✅	❌	2026-09-17
areal-project/AReaL	❌	-	❌	❌	2026-09-17
fla-org/flash-linear-attention	❌	-	❌	❌	2026-09-17
tile-ai/tilelang-mlir-ascend	✅	✅	❌	✅	2026-09-17
verl-project/verl-SpeCo	❌	-	❌	❌	2026-09-17

Cache audit runs daily (incremental). ✅ = confirmed in use · ❌ = confirmed NOT in use · - = unknown · Results sourced from https://github.com/ascend-gha-runners/docs/actions/runs/35232766236