Skip to content

Actions: vllm-project/vllm-ascend

Actions

ascend test / performance

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
nightly_benchmarks.yaml
1,283 workflow runs
1,283 workflow runs

Filter by Event

Filter by Status

Filter by Branch

Filter by Actor

[Refactor] Adjustments to moe_comm_method selection process ascend test / performance #1288: Pull request #3001 labeled by MengqingCao
2s Pr0Wh1teGivee:comments_fix
Pr0Wh1teGivee:comments_fix
2s
[1/N][Refactor][Qwen3-Next] remove redundant Qwen3NextSparseMoeBlock and Qwen3NextAttention ascend test / performance #1287: Pull request #3019 labeled by wxsIcey
1s wxsIcey:qwen3_next_refactor_1
wxsIcey:qwen3_next_refactor_1
1s
[Test] Update the format of the accuracy report ascend test / performance #1286: Pull request #3081 labeled by vllm-ascend-ci
1s zhangxinyuehfad:zxy_accuracy_report
zhangxinyuehfad:zxy_accuracy_report
1s
[Test] Update the format of the accuracy report ascend test / performance #1285: Pull request #3081 labeled by vllm-ascend-ci
1s zhangxinyuehfad:zxy_accuracy_report
zhangxinyuehfad:zxy_accuracy_report
1s
[1/N][Refactor][Qwen3-Next] remove redundant Qwen3NextSparseMoeBlock and Qwen3NextAttention ascend test / performance #1284: Pull request #3019 labeled by wxsIcey
1s wxsIcey:qwen3_next_refactor_1
wxsIcey:qwen3_next_refactor_1
1s
[Perf] Add new npu_fused_infer_attention_score op to improve perfomance in splitfuse cases and resolve long-seq mask problems ascend test / performance #1283: Pull request #2962 labeled by wangxiyuan
1s qyqc731:main
qyqc731:main
1s
[Feature]cpu offload connector ascend test / performance #1282: Pull request #1659 labeled by Yikun
1s lidenghui1110:cpu-offload
lidenghui1110:cpu-offload
1s
[Feature]cpu offload connector ascend test / performance #1281: Pull request #1659 labeled by Yikun
3s lidenghui1110:cpu-offload
lidenghui1110:cpu-offload
3s
[bugfix][torchair] fix kv_nz accuracy problem and remove redundant reshape_and_cache operation ascend test / performance #1280: Pull request #3066 labeled by wangxiyuan
16m 44s linfeng-yuan:fix_torchair_kv_nz
linfeng-yuan:fix_torchair_kv_nz
16m 44s
[bugfix][torchair] fix wasted NPU memory buffer allocation for quantized deepseek with unquantized MTP layer ascend test / performance #1279: Pull request #3068 labeled by wangxiyuan
1m 19s linfeng-yuan:fix_torchair_redundant_process_group
linfeng-yuan:fix_torchair_redundant_process_group
1m 19s
ascend test / performance ascend test / performance #1278: Scheduled
6h 0m 18s main
main
6h 0m 18s
ascend test / performance ascend test / performance #1277: Scheduled
6h 0m 18s main
main
6h 0m 18s
[Hybrid KV] Follow up UniformTypeKVCacheSpecs ascend test / performance #1276: Pull request #3070 labeled by MengqingCao
1s MengqingCao:uniform_type_kvcache_spec
MengqingCao:uniform_type_kvcache_spec
1s
Refactor ci to reuse base workflow and re-enable ut coverage ascend test / performance #1275: Pull request #3064 labeled by Yikun
1s Yikun:refactor-ci-2
Yikun:refactor-ci-2
1s
Refactor ci to reuse base workflow and re-enable ut coverage ascend test / performance #1274: Pull request #3064 labeled by Yikun
1s Yikun:refactor-ci-2
Yikun:refactor-ci-2
1s
[Hybrid KV] Follow up UniformTypeKVCacheSpecs ascend test / performance #1273: Pull request #3070 labeled by MengqingCao
1s MengqingCao:uniform_type_kvcache_spec
MengqingCao:uniform_type_kvcache_spec
1s
[Hybrid KV] Follow up UniformTypeKVCacheSpecs ascend test / performance #1272: Pull request #3070 labeled by MengqingCao
1s MengqingCao:uniform_type_kvcache_spec
MengqingCao:uniform_type_kvcache_spec
1s
[Bugfix] Fix broken CI to follow pooler and attention refactor ascend test / performance #1271: Pull request #1848 labeled by Yikun
1s Potabk:bugfix2
Potabk:bugfix2
1s
[Bugfix] Fix broken CI ascend test / performance #1270: Pull request #2825 labeled by Yikun
3m 3s Potabk:fix_ci7
Potabk:fix_ci7
3m 3s
ascend test / performance ascend test / performance #1269: Scheduled
6h 0m 38s main
main
6h 0m 38s
[CI] Upgrade vLLM to 20250920 (c60e613) and address config break ascend test / performance #1268: Pull request #3067 labeled by Yikun
4h 26m 57s Yikun:fix-main
Yikun:fix-main
4h 26m 57s
[CI] Upgrade vLLM to 20250920 (c60e613) and address config break ascend test / performance #1267: Pull request #3067 labeled by Yikun
1s Yikun:fix-main
Yikun:fix-main
1s
[bugfix][torchair] fix wasted NPU memory buffer allocation for quantized deepseek with unquantized MTP layer ascend test / performance #1266: Pull request #3068 labeled by linfeng-yuan
1s linfeng-yuan:fix_torchair_redundant_process_group
linfeng-yuan:fix_torchair_redundant_process_group
1s
[bugfix][torchair] fix wasted NPU memory buffer allocation for quantized deepseek with unquantized MTP layer ascend test / performance #1265: Pull request #3068 labeled by linfeng-yuan
1s linfeng-yuan:fix_torchair_redundant_process_group
linfeng-yuan:fix_torchair_redundant_process_group
1s
[bugfix][torchair] fix kv_nz accuracy problem and remove redundant reshape_and_cache operation ascend test / performance #1264: Pull request #3066 labeled by linfeng-yuan
19m 32s linfeng-yuan:fix_torchair_kv_nz
linfeng-yuan:fix_torchair_kv_nz
19m 32s