Skip to content

Pull requests: flashinfer-ai/flashinfer

New pull request New
65 Open 1,475 Closed
65 Open 1,475 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

refactor: remove MetaInfoHash class
#2064 opened Nov 8, 2025 by yzh119 Loading…
5 tasks
6
Fix: several bugs/issues with trtllm-gen attention kernels.
#2062 opened Nov 7, 2025 by PerkzZheng Loading…
5 tasks
14
Add custom communicator for trtllm_mnnvl_ar
#2056 opened Nov 6, 2025 by wenscarl • Draft
5 tasks
4
feat: add xqa mla backend
#2053 opened Nov 6, 2025 by qsang-nv Loading…
5 tasks
18
Rebase FP8 SM100 Cutlass FMHA Attention to main (original PR#1238)
#2047 opened Nov 5, 2025 by pavanimajety • Draft
5 tasks
2
feat: Add flashinfer.rope.rope_quantize_fp8_append_paged_kv_cache (fused RoPE + Q + KV cache, supports MLA/GQA/MHA)
#2037 opened Nov 4, 2025 by kahyunnam Loading…
5 tasks done
@kahyunnam
35
Added an initial implementation of Q and KV Cache in fp8 and to use t…
#2035 opened Nov 4, 2025 by Anerudhan • Draft
3 tasks done
1
Enable renormalize(naive) routing for fp8 per-tensor
#2030 opened Nov 3, 2025 by IwakuraRein • Draft
5 tasks
1
[NVIDIA] Thor & Spark Support
#2028 opened Nov 3, 2025 by johnnynunez Loading…
6
Refactor flashinfer/__init__.py so that applications could selectively pack submodules without modifying __init__.py
#2027 opened Nov 3, 2025 by bangshengtang Loading…
5 tasks done
8
Blockwise GEMM with all reduce overlapping
#2007 opened Oct 30, 2025 by Amir-19 • Draft
5 tasks
1
refactor: backend_requirement + supported_compute_capability decorator for gemm
#2000 opened Oct 29, 2025 by jimmyzho Loading…
5 tasks
23
[wip] feat: Add backend='auto' to mm_fp4 and enable autotune for backend='cudnn'
#1979 opened Oct 25, 2025 by bkryu • Draft
5 tasks done
@bkryu
22
chore: agentic workflow for automatic version bump
#1947 opened Oct 19, 2025 by yzh119 Loading…
5 tasks
35
add blockwise gemm cute dsl
#1922 opened Oct 13, 2025 by Amir-19 Loading…
5 tasks
7
Sampling non contiguous
#1916 opened Oct 12, 2025 by zcin Loading…
5 tasks done
1
8
Fix "cannot find -lcuda & -lcudart" problem in WSL2
#1909 opened Oct 10, 2025 by HelloCard Loading…
3 tasks
10
documented sampling.cuh
#1864 opened Oct 4, 2025 by yonikremer Loading…
3
Feature/moe bf16 pr
#1859 opened Oct 4, 2025 by aleozlx • Draft
4 of 5 tasks
2
[wip] test: add coverage for all cli commands
#1848 opened Oct 2, 2025 by sricketts • Draft
2 of 5 tasks
@sricketts
1
[DO NOT MERGE][WIP] lint: Add clang-tidy to pre-commits
#1845 opened Oct 2, 2025 by yzh119 Loading…
5 tasks
3
chore: allow custom paths for external dependencies like CUTLASS
#1827 opened Oct 1, 2025 by yzh119 Loading…
4 of 5 tasks
9
[wip] add two-shot all reduce warp group to blockscaled gemm kernel
#1814 opened Sep 30, 2025 by Amir-19 • Draft
5 tasks
Add CUDA arch 12.0 to installation guide
#1785 opened Sep 26, 2025 by mgoin Loading…
5 tasks
10
fix the dequantize_block in the trtllm_cutlass fuse moe test
#1721 opened Sep 18, 2025 by rainj-me Loading…
5 tasks done
9
ProTip! Add no:assignee to see everything that’s not assigned.