Skip to content

Pull requests: jd-opensource/xllm

New pull request New
12 Open 269 Closed
12 Open 269 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: set the default value of micro_batch_num to 1. (#337)
#339 opened Nov 9, 2025 by JimHsiung Loading…
Bugfix/mtp
#338 opened Nov 9, 2025 by DongheJin Loading…
feat: support graceful termination of the xllm process via the SIGTERM/SIGINT signal.
#335 opened Nov 7, 2025 by yq33victor Loading…
feat: support indexer mla attention.
#332 opened Nov 7, 2025 by a120092009 Loading…
2
feat: enable multi-process mode when running VLM model.
#330 opened Nov 6, 2025 by wly-115 Loading…
feat: enable torch_npu graph mode for Qwen-3 dense with TP support.
#325 opened Nov 6, 2025 by yingxudeng Loading…
11
feat: add control to enable customized mla operation.
#324 opened Nov 6, 2025 by Hermit-w Loading…
3
feat: implement FLUX.1-Control model.
#319 opened Nov 5, 2025 by yiming-l21 Loading…
feat: add generative recommendation tokenizer.
#317 opened Nov 4, 2025 by magicheng0816 Loading…
14
feat: optimize kv cache load/offload.
#306 opened Oct 31, 2025 by Kang-Meng Loading…
25
【WIP】feat: add rec framwork.
#305 opened Oct 31, 2025 by DragonFive Loading…
3 of 5 tasks
feat: add flashinfer as kernel backend for cuda device.
#246 opened Oct 17, 2025 by XuZhang99 Loading…
8
ProTip! no:milestone will show everything without a milestone.