Skip to content

Pull requests: jd-opensource/xllm

New pull request New
76 Open 1,082 Closed
76 Open 1,082 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 enable_mla in model_input_param.
#1326 opened Apr 22, 2026 by XuZhang99 Collaborator Loading…
feat: update conv1d_fn op for Qwen3-Next/Qwen3.5 on NPU.
#1322 opened Apr 21, 2026 by maojunx99 Contributor • Draft
10
bugfix: fix swap block transfer ownership in batch creation.
#1319 opened Apr 21, 2026 by Kang-Meng Collaborator Loading…
11
feat: support multi-priority scheduling for PD disagg.
#1318 opened Apr 21, 2026 by weizhehuang0827 Collaborator Loading…
8
feat: add rmsnorm_qk_rope and rmsnorm_qk for ilu device.
#1317 opened Apr 21, 2026 by laneeeee Contributor Loading…
5
【WIP】feat: add onerec xattention.
#1316 opened Apr 20, 2026 by DragonFive Collaborator • Draft
7
refactor: optmize dp graph and gather comm for dsv32.
#1314 opened Apr 20, 2026 by phantomlei3 Collaborator Loading…
8
feat: support parsing dtype field from config.
#1310 opened Apr 20, 2026 by DongheJin Collaborator Loading…
4
bugfix: serialize tilelang source generation.
#1305 opened Apr 19, 2026 by yingxudeng Collaborator Loading…
feat: support hetero mlu mooncake push pd.
#1304 opened Apr 17, 2026 by phantomlei3 Collaborator Loading…
10
bugfix: add calculation of mm_embedding size in shm manager.
#1298 opened Apr 16, 2026 by shan-chen-feng Collaborator Loading…
3
feat: reuse pre-planned ExecCfg in sparse MoE prep_in.
#1297 opened Apr 16, 2026 by yq33victor Collaborator Loading…
1
perf: add llm decode metadata update fast path.
#1294 opened Apr 16, 2026 by RobbieLeung Collaborator Loading…
14
Glm5 cp benchmark data readme
#1287 opened Apr 15, 2026 by ltdo111 Contributor Loading…
9
feat: support REC extended item info parsing and response output.
#1282 opened Apr 15, 2026 by DragonFive Collaborator Loading…
17
feat: support Risk-Aligned Cache under Classifier-Free Guidance.
#1273 opened Apr 14, 2026 by yiming-l21 Collaborator Loading…
9
feat: support /v1/audio/speech api request.
#1272 opened Apr 14, 2026 by wxh571001500 Contributor Loading…
6
bugfix: add ILU process group ctor for explicit local rank. (#1240)
#1271 opened Apr 13, 2026 by liutongxuan Collaborator Loading…
3
feat: add startup profile run for mlu llm engine.
#1266 opened Apr 13, 2026 by phantomlei3 Collaborator Loading…
4
bugfix: fix mtp prefix cache prefill starvation.
#1264 opened Apr 13, 2026 by phantomlei3 Collaborator Loading…
12
bugfix: unify device_index logging conversion with helper overloads
#1263 opened Apr 12, 2026 by kuma-loong Contributor Loading…
3
feat: add npu_recurrent_gated_delta_rule and chunk_gated_delta_rule fusion operators for qwen3.5/qwen3-next.
#1262 opened Apr 11, 2026 by fems14 Contributor Loading…
18
bugfix: fix DeepSeek V3 crash and V3.2 prefix-cache OOM.
#1258 opened Apr 10, 2026 by DongheJin Collaborator Loading…
2
bugfix: fix CFG negative prompt judgment & simplify variable names fo…
#1256 opened Apr 10, 2026 by yiming-l21 Collaborator Loading…
bugfix: remove spurious backslash breaking output redirection in launch scripts.
#1248 opened Apr 10, 2026 by kuishou68 Loading…
1
1
ProTip! no:milestone will show everything without a milestone.