 Liangsheng Yin
|
b15a83983c
|
[Fix] Release hierarchical cache host pool on graceful shutdown (#31746)
|
2026-07-19 22:27:29 -07:00 |
|
|
Mick
|
2eed35d738
|
perf: avoid temporary VLM encoder gather padding (#31301)
|
2026-07-20 12:54:29 +08:00 |
|
|
kangwangamdandYC Yen-Ching Tseng
|
9f8e916131
|
[diffusion] post_training: run weight update under torch.inference_mode() (#31263)
Co-authored-by: YC Yen-Ching Tseng <yctseng@amd.com>
|
2026-07-20 11:14:13 +08:00 |
|
|
McZyWu
|
1f637a65b9
|
[NPU] bugfix for W4A8MoE bias 3D dimension mismatch problem (#31707)
|
2026-07-20 11:05:44 +08:00 |
|
|
Mohammad Miadh Angkad
|
35f2d4f761
|
Fix no-padding CUDA graph admission (#31273)
|
2026-07-19 19:28:35 -07:00 |
|
|
Mohammad Miadh Angkad
|
8bf2ab9be9
|
Enable GPT-OSS TinyGEMM on CUDA 13 (#31649)
|
2026-07-19 19:16:39 -07:00 |
|
|
gaopengff
|
bab1dd0d12
|
[Intel XPU] Enable (biased) grouped topk for xpu (#31126)
|
2026-07-20 09:35:08 +08:00 |
|
|
Polisetty V R K Jyothendra VarmaandMa Mingfei
|
5325cee7ea
|
Bug fix in compress to support XPU (#31541)
Co-authored-by: Ma Mingfei <mingfei.ma@intel.com>
|
2026-07-20 09:34:11 +08:00 |
|
|
Mick
|
6a25dd7b5f
|
fix: warm up Kimi VLM vision encoder at startup (#31298)
|
2026-07-20 08:50:57 +08:00 |
|
|
Liangsheng Yin
|
b3570a4531
|
[Refactor] Unify input logprob processing on a single chunked path (#31655)
|
2026-07-19 16:03:23 -07:00 |
|
|
Lianmin Zhengandwangwenchen0407
|
555267ed05
|
Fix ROCm fused KV and KDA paths (#31688)
Co-authored-by: wangwenchen0407 <wangwenchen@meta.com>
|
2026-07-19 15:02:35 -07:00 |
|
|
Alison Shao
|
688a6d23f1
|
[DeepSeek-V4] Fix idle-rank dummy-extend sparse-prefill crash under DP breakable CUDA graph (#31705)
|
2026-07-19 14:42:04 -07:00 |
|
|
Mick
|
d4801be447
|
fix: fix vlm cuda graph shape stability (#30868)
|
2026-07-19 22:35:51 +08:00 |
|
|
Yuhao Yang
|
a03ca46a28
|
Fix KDA prefix caching under mamba extra_buffer and enable it for kimi_linear (#31474)
|
2026-07-19 20:09:03 +08:00 |
|
|
Baizhou Zhang
|
7a03d30149
|
Support MiMo V2.5 with zigzag context parallelism (#29972)
|
2026-07-19 02:36:40 -07:00 |
|
|
YC Yen-Ching Tseng
|
377c93d54e
|
[AMD] Gate TP4 o_proj/qkv CK block-FP8 GEMM shapes to Triton (ROCm 7.0 Qwen-3.5) (#30940)
|
2026-07-19 01:45:44 -07:00 |
|
|
YC Yen-Ching Tseng
|
c68392c535
|
[AMD] Fix DeepSeek MLA prefill shape mismatch on HIP eager fallback (missing mha_companion_layers) (#31675)
|
2026-07-19 01:40:41 -07:00 |
|
|
Feng Yao
|
942bf04ef9
|
[Scheduler] Add SGLANG_FORCE_COARSE_WAR_BARRIER opt-in for a whole-forward WAR barrier (#29353)
|
2026-07-19 00:28:49 -07:00 |
|
|
Kevin FlansburgandJimmy Shong
|
609fe1c0d1
|
fix(gemma4): prevent attention mask offset overflow (#31672)
Co-authored-by: Jimmy Shong <69131491+Jiminator@users.noreply.github.com>
|
2026-07-18 23:24:03 -07:00 |
|
|
|
b8ec544946
|
[DSA] Integrate Q8KV8 FP8 Sparse MLA Prefill into the DSA Backend (DeepSeek-V3.2) (#30514)
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Co-authored-by: Xiaoyu Zhang <1182563586@qq.com>
|
2026-07-19 11:58:16 +08:00 |
|
|
Liangsheng Yin
|
cce5fe7696
|
[Scheduler] Move the WAR barrier to right after each run_batch launch (#31687)
|
2026-07-18 20:38:34 -07:00 |
|
|
Lucia Fang
|
99f5a6f46b
|
[flashinfer] Pass window_left at plan time for the SWA paged prefill wrapper (#31501)
|
2026-07-18 13:33:24 -07:00 |
|
|
Liangsheng Yin
|
10908a6793
|
[Fix] Respect cache_protected_len in ChunkCache and disabled-radix release paths (#31662)
|
2026-07-18 12:04:25 -07:00 |
|
|
Sam Shleifer
|
b3a0185cab
|
model_runner: extract post-memory-pool wiring into _init_post_memory (#31601)
|
2026-07-19 00:57:54 +08:00 |
|
|
Sam Shleifer
|
d86ae51fcf
|
metrics: allow extra labels on HTTP request/response Prometheus metrics (#31530)
|
2026-07-19 00:57:26 +08:00 |
|
|
Ke Bao
|
5609f8e509
|
Reset only the used mamba state on radix cache hit (#31643)
|
2026-07-19 00:40:02 +08:00 |
|
|
McZyWu
|
ece02ffc9c
|
[NPU] FIX CMB illusion of garbled characters acc problems, in prefix cache mtp scenarios. (#31659)
|
2026-07-18 21:49:37 +08:00 |
|
|
Ke Bao
|
48ae829f6e
|
Reset only the used mamba state on unified radix cache (#31648)
|
2026-07-18 21:43:13 +08:00 |
|
|
chx96642264
|
d7b9425529
|
[NPU] fix: skip Triton embedding kernel on NPU to avoid kernel launch failure (#31636)
|
2026-07-18 19:32:03 +08:00 |
|
|
Xiaoyu ZhangandClaude Opus 4.8
|
216b750c8f
|
[Kernel] Sweep decoupled scattered kernels into sglang.kernels.ops (RFC #29630) (#31582)
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
|
2026-07-18 19:07:07 +08:00 |
|
|
|
9306278fbc
|
vlm: batch cross-request vit encoding and reuse attention metadata (#24013)
Co-authored-by: yhyang201 <yhyang201@gmail.com>
Co-authored-by: Mick <mickjagger19@icloud.com>
|
2026-07-18 18:50:33 +08:00 |
|
|
Gabriel Wu
|
faf6894093
|
Implement SM120 DeepSeek V4 flashinfer_mxfp4 moe runner backend + TP2 (#30272)
|
2026-07-18 03:01:06 -07:00 |
|
|
Mick
|
6c6175fabd
|
perf: avoid excessive prefill CUDA graph padding (#31487)
|
2026-07-18 16:25:30 +08:00 |
|
|
twb1235andZhiqiang Xie
|
071e649288
|
fix(rpc) Synchronize RPC requests only within the TP group. (#25213)
Co-authored-by: Zhiqiang Xie <xiezhq@stanford.edu>
|
2026-07-18 07:44:51 +00:00 |
|
|
ThanhhaoandHao Phan
|
72c4ed1a3f
|
[Spec] DFlash: remove per-step host syncs so the CPU runs a full step ahead (spec-v2 overlap) (#31468)
Co-authored-by: Hao Phan <htphan@nvidia.com>
|
2026-07-17 23:22:07 -07:00 |
|
|
Khoa PhamandClaude Opus 4.8
|
7a896215e7
|
[CP] Migrate MLA prefill CP (DeepSeek V3) to CP-v2 zigzag strategy (#31619)
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
|
2026-07-17 22:37:53 -07:00 |
|
|
Liangsheng Yin
|
639261f7b2
|
[Refactor] Move output logprob processing into the logprob_processor layer (#31624)
|
2026-07-17 22:33:52 -07:00 |
|
|
Liangsheng Yin
|
19c53c44a0
|
[Fix] Account zero-logprob sequences correctly in chunked logprob stitching (#31639)
|
2026-07-17 22:33:27 -07:00 |
|
|
DevashishLal-CBandDevashish Lal
|
ab3d421c30
|
[plugin] oot torch profiler activity support (#31580)
Signed-off-by: Devashish Lal <devcode@fb.com>
Co-authored-by: Devashish Lal <devcode@fb.com>
|
2026-07-17 22:20:54 -07:00 |
|
|
Alison Shao
|
a5c0b94034
|
Let CI server launches wait longer for ports held by a dying predecessor (#31281)
|
2026-07-17 21:45:24 -07:00 |
|
|
|
44e4999ab2
|
[Diffusion] Use SGLang server for ERNIE-Image prompt enhancement (#31354)
Co-authored-by: Elizaveta Martirosian <you@example.com>
Co-authored-by: Elizaveta Martirosian <elizaveta.martirosian@gmail.com>
Co-authored-by: ronnie_zheng <zl19940307@163.com>
|
2026-07-18 06:55:28 +03:00 |
|
|
R0CKSTAR
|
87dc211b87
|
[MUSA] Fix sglang-kernel build (#31634)
Signed-off-by: Xiaodong Ye <xiaodong.ye@mthreads.com>
|
2026-07-17 20:51:09 -07:00 |
|
|
xdtbynd
|
359009fa00
|
[Bugfix][NPU] Fix/Refactor routed scaling factor application in MoE routing (#31449)
|
2026-07-18 10:59:13 +08:00 |
|
|
|
67e7f8d13a
|
[JIT] Refactor dtype traits into DTypeTrait and unify warp reductions (#30838)
Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
Co-authored-by: BBuf <xiaoyu.zhang@radixark.ai>
Co-authored-by: jessiewei7 <jessiewei747@gmail.com>
Co-authored-by: root <root@GPUC5A6.maas>
|
2026-07-18 10:07:18 +08:00 |
|
|
Brayden Zhong
|
238b2b2c9c
|
Remove QServe and FBGEMM FP8 quantization (#31109)
|
2026-07-17 17:10:34 -07:00 |
|
|
Alison Shao
|
f926c30c57
|
Revert "Fix mamba track-boundary seqlen under overlap scheduler (#31369)" (#31622)
|
2026-07-17 17:03:10 -07:00 |
|
|
Mick
|
42a058c760
|
optimize: avoid fla l2-norm recompilation by token count (#31558)
|
2026-07-18 07:58:36 +08:00 |
|
|
Qiaolin Yu
|
01a96720c6
|
[spec decoding] replace torch.multinomial with several native torch op in rejection sampling (#31620)
|
2026-07-17 16:58:11 -07:00 |
|
|
Baizhou Zhang
|
304a529558
|
Revert "Bump FlashInfer to 0.6.15 and revert regressions" (#31625)
|
2026-07-17 16:46:33 -07:00 |
|
|
Lianmin Zheng
|
c95026aed3
|
Upgrade llguidance to 1.7.6 (#31484)
|
2026-07-17 16:31:44 -07:00 |
|