Skip to content

Pull requests: flashinfer-ai/flashinfer

New pull request New
48 Open 1,157 Closed
48 Open 1,157 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Reviews
Filter by reviews
No reviews Review required Approved review Changes requested
Assignee
Filter by who’s assigned
Assigned to nobody Loading
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

feat: Enable MnnvlMemory (for alltoallv) on B200
#1601 opened Aug 28, 2025 by trevor-m Loading…
5 tasks
4
ci: add ci container of cuda 13 and add cute-dsl as dependency.
#1595 opened Aug 28, 2025 by yzh119 Loading…
5 tasks
7
upgrade cuda-python
#1579 opened Aug 26, 2025 by johnnynunez Loading…
5
re-enable torch.compile
#1576 opened Aug 26, 2025 by guilhermeleobas Loading…
4 of 5 tasks
1
Support input and output signals for cutedsl gemm
#1569 opened Aug 25, 2025 by fzyzcjy • Draft
5 tasks
feat: add lse return on trtllm-gen attention
#1566 opened Aug 24, 2025 by yyihuang • Draft
5 tasks
2
[DRAFT] feat: support radix-based top-k sampling algorithm
#1561 opened Aug 24, 2025 by JasonJ2021 Loading…
4 of 5 tasks
9
[DRAFT] feat: add support of fp4_batched_quantize
#1552 opened Aug 23, 2025 by yicwang Loading…
5 tasks
4
[CuTe DSL] Add Blackwell MHA prefill and MLA decode kernel
#1549 opened Aug 22, 2025 by dubcyfor3 • Draft
4 of 5 tasks
1
Fix autotuner for trtllm fp4 fused moe
#1548 opened Aug 22, 2025 by stslxg-nv • Draft
4 of 5 tasks
Auto tuning cutedsl gemm
#1527 opened Aug 21, 2025 by fzyzcjy Loading…
5 tasks
@yyihuang
3
tests(attn): add short-seq CUDA edge-case test (qo_len=1) for prefill
#1515 opened Aug 19, 2025 by PrithviElancherran Loading…
3 tasks done
5
feat: integrate xqa attention backend
#1503 opened Aug 18, 2025 by qsang-nv Loading…
3 of 5 tasks
19
Trtllm-gen Fp8 MoE Autotunner
#1494 opened Aug 15, 2025 by aleozlx • Draft
5 tasks done
30
feat: Centralize env vars in flashinfer.json with caching
#1487 opened Aug 14, 2025 by yongwww Loading…
5 tasks
22
feat(attention): add RoPE offset support for batch prefill
#1457 opened Aug 11, 2025 by MengAiDev Loading…
3 tasks done
4
feat: enable trtllm-gen attn speculative decoding verify by decode priority: high
#1453 opened Aug 11, 2025 by yyihuang Loading…
5 tasks
@yzh119 @zhyncs @yyihuang
4
benchmark: add allreduce_fusion benchmark
#1450 opened Aug 10, 2025 by yyihuang • Draft
5 tasks
@yyihuang
6
refactor: unify autotuner for fp4 gemm backends
#1439 opened Aug 8, 2025 by ttyio Loading…
3 of 5 tasks
4
Restore llama4 fc2 required kernels
#1417 opened Aug 8, 2025 by aleozlx Loading…
5 tasks done
7
refactor: update trtllm-gen attn interface
#1407 opened Aug 7, 2025 by yyihuang • Draft
5 tasks
@yyihuang
4
misc: Customize kv lens buffer size for sparse attention
#1383 opened Aug 5, 2025 by Edenzzzz Loading…
5 tasks
1
6
Removes MPI dependency from MNNVL AllReduce
#1379 opened Aug 4, 2025 by pranavm-nvidia • Draft
5 tasks
6
feat: Support sliding window for persistent kernel
#1368 opened Aug 3, 2025 by Edenzzzz Loading…
5 tasks
7
feat: support dynamic scale factor interface for trtllm-gen attn
#1342 opened Jul 29, 2025 by yyihuang • Draft
3 of 5 tasks
@yzh119 @yyihuang
3
ProTip! Mix and match filters to narrow down what you’re looking for.