Skip to content

Pull requests: xLLM-AI/xllm

New pull request New
97 Open 1,671 Closed
97 Open 1,671 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 devices related gflags.
#1948 opened Jul 15, 2026 by XuZhang99 Collaborator Loading…
17 tasks
feat: add linear-state prefix cache batch and scheduler plumbing (5/n).
#1947 opened Jul 15, 2026 by yingxudeng Collaborator Loading…
17 tasks
1
bugfix: temporarily remove the python interface for xllm.
#1946 opened Jul 15, 2026 by XuZhang99 Collaborator Loading…
17 tasks
feat(wan): SP communication-computation overlap for Wan2.2 I2V.
#1945 opened Jul 15, 2026 by nie-linfeng Contributor Loading…
bugfix: redirect draft-extend placeholder KV write to padding block.
#1944 opened Jul 15, 2026 by weizhehuang0827 Collaborator Loading…
1 of 17 tasks
feat: add DCU support for LongCat and Flux series pipelines.
#1937 opened Jul 13, 2026 by WenQ7 Collaborator Loading…
8 of 17 tasks
perf: add native split gated RMSNorm for Qwen3.5 and Qwen3.6 on NPU
#1936 opened Jul 13, 2026 by wanghuanjun2113 Contributor Loading…
refactor: prefer std::make_unique and std::make_shared to direct use of new.
#1930 opened Jul 12, 2026 by Dragonliu2018 Collaborator Loading…
9 of 17 tasks
refactor: declare functions noexcept if they won't emit exceptions.
#1928 opened Jul 12, 2026 by Dragonliu2018 Collaborator Loading…
9 of 17 tasks
refactor: declare overriding functions override.
#1927 opened Jul 12, 2026 by Dragonliu2018 Collaborator Loading…
9 of 17 tasks
feat: add NPU fused norm operators for Wan2.2 DiT.
#1923 opened Jul 10, 2026 by ethan686 Collaborator Loading…
feat(npu): route gated norm to aclnn layer norm fwd
#1919 opened Jul 10, 2026 by Fengfengst123 • Draft
1
feat: add RWKV-7-World model support on cuda device.
#1918 opened Jul 10, 2026 by Dragonliu2018 Collaborator Loading…
9 of 17 tasks
2
feat: support Wan2.2 distill model.
#1910 opened Jul 8, 2026 by ethan686 Collaborator Loading…
1
feat: add request trace dump with runtime toggle.
#1906 opened Jul 7, 2026 by shifengmin Collaborator Loading…
8 of 17 tasks
3
feat: implement single-process mode on npu device.
#1894 opened Jul 6, 2026 by XuZhang99 Collaborator Loading…
8 of 17 tasks
8
feat: add embedded-python model executor for qwen3 on cuda.
#1891 opened Jul 6, 2026 by zhang-minchao Collaborator Loading…
5 of 7 tasks
82
feat: Add FlashComm1 and MMRS fusion support for Qwen3.5 NPU.
#1890 opened Jul 6, 2026 by Cooofish Loading…
21
perf: route Qwen3.5 FSGDR model path to TileLang.
#1889 opened Jul 6, 2026 by ShareableXue Contributor Loading…
7 of 17 tasks
2
feat: add DFlash block-diffusion speculative decoding on npu device.
#1886 opened Jul 3, 2026 by ustcfy Collaborator Loading…
2 tasks done
37
feat: support adaptive speculative decoding.
#1876 opened Jul 2, 2026 by weizhehuang0827 Collaborator Loading…
7 of 17 tasks
10
Cp chunkprefill upstream 20260610 latest
#1872 opened Jul 2, 2026 by cchh05 Collaborator Loading…
17 tasks
6
feat: add ViT encoder ACL Graph capture for Qwen3-VL.
#1866 opened Jul 1, 2026 by sunbaosong Collaborator Loading…
19
feat: support multi machine for offline inference.
#1863 opened Jul 1, 2026 by weizhehuang0827 Collaborator Loading…
9 of 12 tasks
2
refactor: host prefix cache — device↔host KV-cache transfer framework (4/n).
#1861 opened Jul 1, 2026 by Kang-Meng Collaborator Loading…
10
ProTip! Type g p on any issue or pull request to go back to the pull request listing page.