Pull requests: flash-algo/flash-sparse-attention
Pull requests list
[BUG FIX] Correct causal mask handling for longer KV pairs
#213 by LoserCheems was merged Dec 2, 2025 Loading…
4 tasks
[BUG FIX] Unify masking utilities and improve performance
#209 by LoserCheems was merged Nov 13, 2025 Loading…
4 tasks
Fix documentation and references for Flash Sparse Attention
#207 by LoserCheems was merged Nov 9, 2025 Loading…
4 tasks
[FEATURE SUPPORT] Triton special compact dynamic-mask attention: 1.6× faster fwd+bwd, numerically equivalent
#206 by LoserCheems was merged Nov 7, 2025 Loading…
4 of 5 tasks
Refactor attention block smoothing for consistency
#205 by LoserCheems was merged Nov 6, 2025 Loading…
4 tasks
[FEATURE SUPPORT] Move scaling out of streaming loops, bias-initialized acc_s, and fix dQ double-scaling
#203 by LoserCheems was merged Nov 4, 2025 Loading…
5 tasks done
Optimize triton version: GQA, mask/bias broadcasting, skip inactive tiles, and stability fixes
#200 by LoserCheems was merged Nov 7, 2025 Loading…
2 of 5 tasks
Fix attention bias calculation and dbias handling
#199 by LoserCheems was merged Oct 27, 2025 Loading…
Update documentation to use mask utility in examples
#198 by LoserCheems was merged Oct 23, 2025 Loading…
[FEATURE SUPPORT] Centralize dynamic mask creation for FDMA
#197 by LoserCheems was merged Oct 23, 2025 Loading…
5 tasks done
[FEATURE SUPPORT] Robust dBias accumulation for seqlen_q_bias == 1
#194 by LoserCheems was merged Oct 22, 2025 Loading…
4 of 5 tasks
Enhance bias gradient accumulation in backward pass
#193 by LoserCheems was merged Oct 16, 2025 Loading…
Fix attention_mask and attention_bias shape descriptions and remove redundant checks
#192 by LoserCheems was merged Oct 13, 2025 Loading…
Refactor bias initialization and enhance bias computation in FlashDMAttnFunc
#191 by LoserCheems was merged Oct 12, 2025 Loading…
[FEATURE SUPPORT] Broadcastable 4D mask/bias, 128‑rounded key length, stride‑0 broadcasting, and dbias reductions
#190 by LoserCheems was merged Oct 12, 2025 Loading…
5 tasks done
[FEATURE SUPPORT] Variable-Length Attention with Padding-Free Execution
#188 by LoserCheems was merged Oct 11, 2025 Loading…
6 tasks done
Implement variable-length attention with mask and bias support
#185 by LoserCheems was merged Oct 9, 2025 Loading…
[BUG FIX] Fix mask/bias memory access and vectorization issues in kernels
#182 by LoserCheems was merged Oct 1, 2025 Loading…
7 tasks done
[BUG FIX] SM80 NaN in bias.grad when both mask and bias are enabled
#179 by LoserCheems was merged Sep 22, 2025 Loading…
3 tasks done
Refactor attention mask and bias handling for efficiency
#177 by LoserCheems was merged Sep 21, 2025 Loading…
ProTip! Type g p on any issue or pull request to go back to the pull request listing page.