Skip to content

Pull requests: lightseekorg/tokenspeed

New pull request New
36 Open 586 Closed
36 Open 586 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(deepseek-v4): enable SM120 serving
#648 opened Jul 11, 2026 by lucifer1004 Contributor • Draft
fix(kernel): use Triton FP8 quantization on SM120
#646 opened Jul 11, 2026 by lucifer1004 Contributor Loading…
feat(kernel): add SM120 FlashInfer MXFP4 MoE
#645 opened Jul 11, 2026 by lucifer1004 Contributor Loading…
11
Route MLA RoPE KV writes through embedding API
#643 opened Jul 11, 2026 by qedawkins Contributor • Draft
[GLM] perf: Integrate TRT-LLM cutedsl topk
#634 opened Jul 10, 2026 by yweng0828 Contributor Loading…
3
feat(kernel): Add gluon kernel for DSA
#627 opened Jul 9, 2026 by Max191 Contributor Loading…
11
fix(deepseek-v4): route large-token fused mHC around the allinone cliff
#622 opened Jul 9, 2026 by Xiangyi1996 Collaborator • Draft
2
perf(deepseek-v4): widen sparse compress cache launch for large ratios
#621 opened Jul 9, 2026 by Xiangyi1996 Collaborator Loading…
9
feat: add DeepSeek V4 L2 KV cache offload and perf optimize.
#620 opened Jul 9, 2026 by SimonCqk Contributor Loading…
22
fix(spec): restore sentinel padding + context clamps for every drafter
#618 opened Jul 9, 2026 by jasl Contributor Loading…
feat(mla): support custom tree masks in decode
#617 opened Jul 9, 2026 by mesaleh Contributor Loading…
9
fix(trtllm-mla): make spec-decode CUDA graph capture causal
#616 opened Jul 9, 2026 by mesaleh Contributor Loading…
feat: enable the prefill CUDA graph by default
#611 opened Jul 8, 2026 by nperrin-fr Collaborator Loading…
12
perf(multimodal): optimize SHM input transport
#610 opened Jul 8, 2026 by yechank-nvidia Collaborator • Draft
2
perf(multimodal): optimize Qwen vision runtime
#609 opened Jul 8, 2026 by yechank-nvidia Collaborator • Draft
perf(multimodal): optimize Qwen3.5 M-RoPE preparation
#608 opened Jul 8, 2026 by yechank-nvidia Collaborator • Draft
refactor(glm): reuse DeepseekV3 base implementation via template methods
#604 opened Jul 7, 2026 by wtsung Contributor Loading…
fix(runtime): restore EnvField overrides after exceptions
#601 opened Jul 7, 2026 by Xiangyi1996 Collaborator • Draft
fix(deepseek-v4): gate experimental TRT-LLM mHC backend
#600 opened Jul 7, 2026 by Xiangyi1996 Collaborator • Draft
[AMD] Initial support for qwen3.5-397b-a17b-mxfp4 on AMD hardware
#592 opened Jul 4, 2026 by raikonenfnu Contributor Loading…
8
fix(output): prevent prompt-tail tokens leaking into streamed output_ids
#584 opened Jul 3, 2026 by minedec Contributor Loading…
7
feat(distributed): wire TensorRT-LLM all-reduce backend
#580 opened Jul 2, 2026 by Xiangyi1996 Collaborator • Draft
fix(evalscope): preserve plain completion prefixes
#579 opened Jul 2, 2026 by dongjiyingdjy Contributor • Draft
perf(deepseek-v4): accelerate indexer Q and packed FP8 quantization
#563 opened Jun 30, 2026 by dongjiyingdjy Contributor Loading…
perf(multimodal): Prewarm Kimi vision encoder before readiness
#556 opened Jun 30, 2026 by qimcis Contributor Loading…
2
ProTip! Updated in the last three days: updated:>2026-07-08.