Skip to content

Actions: vllm-project/flash-attention

Actions

All workflows

Actions

Loading...
Loading

Showing runs from all workflows
118 workflow runs
118 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
[SM100] Optimize hdim 256 and add seqused_q/k support (#180) CI #19: Commit 6bd7c70 pushed by MatthewBonanni
Queued main
main
Queued
Add standalone SM103 FP8 d192/v128 varlen forward kernel Lint #51: Pull request #181 synchronize by mingg26
Action required mingg26:ming-prefill-k3-mla-standalone
mingg26:ming-prefill-k3-mla-standalone
Action required
Add standalone SM103 FP8 d192/v128 varlen forward kernel Lint #50: Pull request #181 synchronize by mingg26
Action required mingg26:ming-prefill-k3-mla-standalone
mingg26:ming-prefill-k3-mla-standalone
Action required
Add standalone SM103 FP8 d192/v128 varlen forward kernel Lint #49: Pull request #181 opened by mingg26
Action required mingg26:ming-prefill-k3-mla-standalone
mingg26:ming-prefill-k3-mla-standalone
Action required
Publish flash-attn-4 to PyPI Publish flash-attn-4 to PyPI #20: Scheduled
11m 37s main
main
11m 37s
fix(cute): pass mDynamicCausal to kernel and init is_split_kv in SM80… CI #18: Commit 7484d47 pushed by LucasWilkinson
1d 0h 0m 2s main
main
1d 0h 0m 2s
Sync upstream Lint #48: Pull request #171 synchronize by LucasWilkinson
19s MatthewBonanni:sync_upstream_20260722
MatthewBonanni:sync_upstream_20260722
19s
Publish flash-attn-4 to PyPI Publish flash-attn-4 to PyPI #19: Scheduled
9s main
main
9s
Re-migrate FA3 hopper API to the PyTorch stable ABI (#165) CI #17: Commit f3e1a4f pushed by Harry-Chen
1d 0h 0m 2s main
main
1d 0h 0m 2s
Publish flash-attn-4 to PyPI Publish flash-attn-4 to PyPI #18: Scheduled
17m 42s main
main
17m 42s
Make vllm's FA2 ABI stable (#175) CI #16: Commit 28e862d pushed by Harry-Chen
1d 0h 0m 2s main
main
1d 0h 0m 2s
Fully yank dropout in vllm fork of FA2 (#153) CI #15: Commit c3f5c9c pushed by Harry-Chen
1d 0h 0m 2s main
main
1d 0h 0m 2s
Allow causal scheduler with mask_mod (#155) CI #14: Commit ec4a882 pushed by LucasWilkinson
1d 0h 0m 2s main
main
1d 0h 0m 2s
Publish flash-attn-4 to PyPI Publish flash-attn-4 to PyPI #17: Scheduled
11s main
main
11s
Running Copilot Code Review Copilot #1: by Copilot AI
4m 38s main
main
4m 38s
Sync upstream Lint #47: Pull request #171 synchronize by MatthewBonanni
17s MatthewBonanni:sync_upstream_20260722
MatthewBonanni:sync_upstream_20260722
17s
Sync upstream Lint #46: Pull request #171 synchronize by MatthewBonanni
14s MatthewBonanni:sync_upstream_20260722
MatthewBonanni:sync_upstream_20260722
14s
Sync upstream Lint #45: Pull request #171 synchronize by MatthewBonanni
20s MatthewBonanni:sync_upstream_20260722
MatthewBonanni:sync_upstream_20260722
20s
Sync upstream Lint #44: Pull request #171 opened by MatthewBonanni
17s MatthewBonanni:sync_upstream_20260722
MatthewBonanni:sync_upstream_20260722
17s
Publish flash-attn-4 to PyPI Publish flash-attn-4 to PyPI #16: Scheduled
8s main
main
8s
Require C++20 to match PyTorch ATen headers (#168) CI #13: Commit ed4b734 pushed by LucasWilkinson
1d 0h 0m 1s main
main
1d 0h 0m 1s
Fix threshold (#166) CI #12: Commit 1689202 pushed by MatthewBonanni
1d 0h 0m 2s main
main
1d 0h 0m 2s
Publish flash-attn-4 to PyPI Publish flash-attn-4 to PyPI #15: Scheduled
11s main
main
11s
Build wheels and deploy Build wheels and deploy #64: by LucasWilkinson
38s
38s
Construct dynamic fake tensor for FA4 compile warmup (nvidia-cutlass-… CI #11: Commit caaa4eb pushed by MatthewBonanni
1d 0h 0m 2s main
main
1d 0h 0m 2s