Skip to content

Pull requests: flagos-ai/vllm-plugin-FL

New pull request New
67 Open 233 Closed
67 Open 233 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

fix(worker): account for CUDA graph memory in KV cache sizing core
#381 opened Aug 14, 2026 by CherryLemon Collaborator Loading…
2 of 3 tasks
feat(hygon): support INT4 inference for Qwen3.6 and GLM-5.2 core
#379 opened Aug 13, 2026 by lisa2342 Loading…
14
adapt(metax): support GLM-5.2 W8A8 INT8 core
#378 opened Aug 13, 2026 by liyu030 Loading…
3 tasks
fix(metax): consolidate vLLM 0.24.0 compat shims into the plugin core
#377 opened Aug 13, 2026 by tengqm Contributor Loading…
3
add flaggems-vllm core
#375 opened Aug 13, 2026 by cyber-pioneer Collaborator Loading…
6
feat(kunlunxin): add GDN fused post-conv Triton bypass for Kunlunxin XPU core
#374 opened Aug 13, 2026 by dongjibin1996 Loading…
[CICD]: Add Enflame S60 CI support ci docs tests
#373 opened Aug 13, 2026 by HermiaHuan Collaborator Loading…
8
feat(metax): support DeepSeek V4 W8A8 INT8 inference core
#372 opened Aug 12, 2026 by liyu030 Loading…
3 tasks
feat(ascend): add NPU device context, head_dim>192 attention fallback, and FLAGCX_PATH-free backend resolution core
#369 opened Aug 12, 2026 by Joiin0392 Loading…
1 of 3 tasks
1
feat(ascend): add build and smoke-test scripts for Ascend NPU deployment
#370 opened Aug 12, 2026 by Joiin0392 Loading…
3 tasks done
1
[Kunlunxin] Fix garbled output with CUDA Graph core
#368 opened Aug 11, 2026 by 0songHan Loading…
3 tasks
1
[Sunrise] Add cuda-graph, int8 and more model support on Sunrise core
#366 opened Aug 11, 2026 by cyx11111 Loading…
3 tasks
ascend: blacklist lift_fresh/_to_copy (coreDim=0 on scalar tensor init) core
#361 opened Aug 10, 2026 by tengqm Contributor Loading…
Cherry-pick PR #297 to v0.3.0-dev: Add thead vendor attention backend core
#360 opened Aug 10, 2026 by physics31415926 Collaborator • Draft
support mla atten in flaggems core
#359 opened Aug 10, 2026 by ZenithJoyH Contributor Loading…
3 tasks
support Qwen3.5 FP8 Model on GCU (Enflame) core
#358 opened Aug 9, 2026 by liuyancong-enflame-tech Contributor Loading…
1 of 3 tasks
12
gcu: enable enflame GCU300 via vLLM native FLASH_ATTN backend core
#357 opened Aug 9, 2026 by tengqm Contributor Loading…
5
feat(cambricon): enable MLU590 (device map + graph class) core
#355 opened Aug 8, 2026 by tengqm Contributor Loading…
[hot-fix] fix GroupedTopKRouterFL routing and MLA decode method name core
#352 opened Aug 7, 2026 by tonyh168 Loading…
1
Qwen36_dense_moe performance surpasses vllm_Ascend core docs tests
#349 opened Aug 7, 2026 by appleinsky Loading…
3 tasks
1
perf(metax): optimize single-request prefill attention core
#345 opened Aug 5, 2026 by liyu030 Loading…
2 of 3 tasks
1
support dsv4 int8 version core
#344 opened Aug 5, 2026 by ceci3 Collaborator Loading…
3 tasks
7
[features] 增加 plugin-fl 在ray 下的设备可见性兼容 core
#343 opened Aug 3, 2026 by tonyh168 Loading…
3 tasks
1
feat(compilation): add BreakableCUDAGraphWrapper support for OOT vendor backends (vLLM 0.24.0+) core tests
#342 opened Aug 3, 2026 by physics31415926 Collaborator Loading…
2 of 3 tasks
feat(quantization): adapt W8A8 inference to vLLM 0.24 core docs tests
#336 opened Aug 3, 2026 by rdzhu225 Contributor Loading…
ProTip! Filter pull requests by the default branch with base:main.