Skip to content

Actions: Dao-AILab/flash-attention

Actions

All workflows

Actions

Loading...
Loading

Showing runs from all workflows
586 workflow runs
586 workflow runs

Filter by Workflow

⌥ + click/return to exclude

Filter by Event

⌥ + click/return to exclude

Filter by Status

⌥ + click/return to exclude

Filter by Branch

⌥ + click/return to exclude

Filter by Actor

⌥ + click/return to exclude
Expose --pack-gqa and --num-splits in benchmark_attn.py (#2473) CI #17: Commit 881139b pushed by Johnsonms
3m 4s main
main
3m 4s
[Cute,Fwd,Sm100] fp8 e4m3 and e5m2 support (#2109) Lint #350: Commit b65ae6b pushed by drisspg
15s main
main
15s
[Cute,Fwd,Sm100] fp8 e4m3 and e5m2 support (#2109) CI #16: Commit b65ae6b pushed by drisspg
3m 32s main
main
3m 32s
Build wheels and deploy Build wheels and deploy #241: by Johnsonms
7s
7s
[Cute,Fwd,Sm100] fp8 e4m3 and e5m2 support Lint #349: Pull request #2109 synchronize by dcw02
14s modal-labs:fa4_fp8_support
modal-labs:fa4_fp8_support
14s
Add dropout support to CuTe DSL attention kernels Lint #348: Pull request #2439 synchronize by blake-snc
14s blake-snc:feat/cute-dsl-dropout
blake-snc:feat/cute-dsl-dropout
14s
expose num_splits for FA2 and add option for kernel blocksize alignme… CI #15: Commit b322ae2 pushed by drisspg
3m 5s main
main
3m 5s
Add dropout support to CuTe DSL attention kernels Lint #347: Pull request #2439 synchronize by blake-snc
14s blake-snc:feat/cute-dsl-dropout
blake-snc:feat/cute-dsl-dropout
14s
Copilot code review Copilot code review #6: by Copilot AI
5m 15s refs/pull/2458/head
refs/pull/2458/head
5m 15s
Copilot code review Copilot code review #5: by Copilot AI
3m 45s refs/pull/2458/head
refs/pull/2458/head
3m 45s
Build wheels and deploy Build wheels and deploy #240: by drisspg
8s
8s
Add CLC scheduler heuristic (#2455) CI #14: Commit d7f60e6 pushed by drisspg
3m 41s main
main
3m 41s
Build wheels and deploy Build wheels and deploy #239: by lw
7s
7s
[FA4] Add missing fence.proxy.async SM90 backward Lint #346: Pull request #2462 opened by lw
14s gh/lw/1/head
gh/lw/1/head
14s
Build wheels and deploy Build wheels and deploy #238: by lw
6s
6s
Build wheels and deploy Build wheels and deploy #237: by lw
6s
6s
Disable 2CTA fwd non-causal on CUDA 12 to work around codegen regress… CI #13: Commit 83b8b8f pushed by Johnsonms
3m 33s main
main
3m 33s
Publish flash-attn-4 to PyPI Publish flash-attn-4 to PyPI #9: Scheduled
1m 7s main
main
1m 7s
Build wheels and deploy Build wheels and deploy #236: by Johnsonms
7s
7s
Build wheels and deploy Build wheels and deploy #235: by drisspg
5s
5s
Handle linter for flash mla file (#2459) Lint #345: Commit 628452c pushed by Johnsonms
16s main
main
16s
Handle linter for flash mla file (#2459) CI #12: Commit 628452c pushed by Johnsonms
3m 35s main
main
3m 35s
Handle linter for flash mla file Lint #344: Pull request #2459 synchronize by jayhshah
18s jshah/mla-ruff-check
jshah/mla-ruff-check
18s
Build wheels and deploy Build wheels and deploy #234: by jayhshah
5s
5s
[Cute,Sm100,Fwd] add MLA 64/512 with topk sparsity for MQA 128 heads … CI #11: Commit f219c89 pushed by jayhshah
3m 15s main
main
3m 15s