Skip to content

Actions: ROCm/aiter

Actions

vLLM Benchmark

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
vllm_benchmark.yaml
5,911 workflow runs
5,911 workflow runs

Filter by Event

Filter by Status

Filter by Branch

Filter by Actor

[FIX] address overflow fix on gfx942 for hd128 fmha bwd case (#2151) vLLM Benchmark #6003: Commit d2f5f27 pushed by JaxChen29
Pending main
main
Pending
CI: Update vllm_benchmark.yaml to use latest nightly image (#2165) vLLM Benchmark #6002: Commit d4e0135 pushed by gyohuangxin
2h 29m 27s main
main
2h 29m 27s
Revert any config for prefill kernel performance (#2166) vLLM Benchmark #6001: Commit da65dcb pushed by HaiShaw
29m 46s main
main
29m 46s
Revert gemm blockscale kernel config changes causing dpsk-r1 prefill regression vLLM Benchmark #6000: Pull request #2166 opened by 1am9trash
2s 1am9trash:fix-triton-a8w8-gemm-blockscale-kernel-prefill-regression
1am9trash:fix-triton-a8w8-gemm-blockscale-kernel-prefill-regression
2s
CI: Update vllm_benchmark.yaml to use latest nightly image vLLM Benchmark #5999: Pull request #2165 opened by gyohuangxin
Queued gyohuangxin-patch-13
gyohuangxin-patch-13
Queued
Introduce HipKittens based nhead=128 MLA Kernel vLLM Benchmark #5998: Pull request #2039 synchronize by ruanjm
Queued jruan/hk_mla_nhead_128
jruan/hk_mla_nhead_128
Queued
[fix]: support dim(-1) allgather vLLM Benchmark #5997: Pull request #2162 synchronize by TennyWang1223
Queued tenny/new_ag
tenny/new_ag
Queued
update_test (#2156) vLLM Benchmark #5996: Commit f3be04a pushed by amd-ruitang3
Queued main
main
Queued
[mla] mtp mock fix vLLM Benchmark #5995: Pull request #2164 synchronize by Zzz9990
Queued zan/mla_mtp_mock
zan/mla_mtp_mock
Queued
[mla] mtp mock fix vLLM Benchmark #5994: Pull request #2164 opened by Zzz9990
2m 9s zan/mla_mtp_mock
zan/mla_mtp_mock
2m 9s
fix mla ps fp8 the kv_seq tail len < 4 nan error vLLM Benchmark #5993: Pull request #2106 synchronize by slippedJim
Queued mmd/fix/mla_ps_fp8
mmd/fix/mla_ps_fp8
Queued
[CK_TILE] Update fmha_bwd_traits vLLM Benchmark #5992: Pull request #2116 ready_for_review by DDEle
Queued ck-update-fmha_bwd_traits
ck-update-fmha_bwd_traits
Queued
[CK_TILE] Update fmha_bwd_traits vLLM Benchmark #5991: Pull request #2116 synchronize by DDEle
2s ck-update-fmha_bwd_traits
ck-update-fmha_bwd_traits
2s
Add ENABLE_CK build option for CK-free builds vLLM Benchmark #5990: Pull request #2074 synchronize by zufayu
1s sunway513:feat/ck-free-build-gating
sunway513:feat/ck-free-build-gating
1s
feat(ck_tile): add a8w8 blockscale gemm with preshuffleQuant support vLLM Benchmark #5989: Pull request #2018 synchronize by amd-khushbu
Queued gemm_preshuffle_quant_ck_tile
gemm_preshuffle_quant_ck_tile
Queued
[TRITON] mHC/mHC-lite: Manifold-constrained Hyper Connection vLLM Benchmark #5988: Pull request #1859 synchronize by anhminhnguyenhoang
50m 45s feat/mhc-deepseek
feat/mhc-deepseek
50m 45s
Corrects arg_size type from int to size_t for hip graph compatability vLLM Benchmark #5987: Pull request #2163 opened by Micky774
1h 10m 5s zain/args-size
zain/args-size
1h 10m 5s
[fix]: support dim(-1) allgather vLLM Benchmark #5986: Pull request #2162 opened by TennyWang1223
1m 47s tenny/new_ag
tenny/new_ag
1m 47s
[Triton] Fav3 sage attention mask fix (#2158) vLLM Benchmark #5985: Commit 00291e1 pushed by azaidy
6h 30m 49s main
main
6h 30m 49s
Remove uses of AITER_ASM_DIR vLLM Benchmark #5984: Pull request #1900 synchronize by draganmladjenovic
5m 9s draganm/embed_hsaco_take_two
draganm/embed_hsaco_take_two
5m 9s
Remove uses of AITER_ASM_DIR vLLM Benchmark #5983: Pull request #1900 synchronize by draganmladjenovic
14m 44s draganm/embed_hsaco_take_two
draganm/embed_hsaco_take_two
14m 44s
fix mla ps fp8 the kv_seq tail len < 4 nan error vLLM Benchmark #5982: Pull request #2106 synchronize by slippedJim
1h 44m 51s mmd/fix/mla_ps_fp8
mmd/fix/mla_ps_fp8
1h 44m 51s
Small refactor for compile_ops control flow for better performance and readability vLLM Benchmark #5981: Pull request #2161 ready_for_review by SampoAMD
2s SampoAMD:refactor_get_module_control_flow
SampoAMD:refactor_get_module_control_flow
2s
Small refactor for compile_ops control flow for better performance and readability vLLM Benchmark #5980: Pull request #2161 opened by SampoAMD
1s SampoAMD:refactor_get_module_control_flow
SampoAMD:refactor_get_module_control_flow
1s
[TRITON] mHC/mHC-lite: Manifold-constrained Hyper Connection vLLM Benchmark #5979: Pull request #1859 synchronize by anhminhnguyenhoang
34m 19s feat/mhc-deepseek
feat/mhc-deepseek
34m 19s