 gilfordtingandJames Liu
|
6f005e4da1
|
[Fix] Require JSON booleans for response_format json_schema.strict (#34777)
Co-authored-by: James Liu <jamesl@modal.com>
|
2026-08-14 16:46:27 -07:00 |
|
|
gilfordtingandJames Liu
|
3f64f14360
|
[Fix] Work around xgrammar 0.2.1 negative integer minimum in Kimi-K3 structural tags (#34778)
Co-authored-by: James Liu <jamesl@modal.com>
|
2026-08-14 16:41:06 -07:00 |
|
|
Nan Jiang
|
be804c1b83
|
[RL] Expose top-p-only sampling masks (#33593)
|
2026-08-14 16:40:27 -07:00 |
|
|
Jason Mancuso
|
90b3db6dd8
|
[Fix: RL] Snapshot async state-capture outputs before overlap (#34319)
|
2026-08-14 16:39:35 -07:00 |
|
|
Ziang Li
|
a9654eacc1
|
fix(dsa): use FlashInfer fused top-k for packed PAGED rows (#33006)
|
2026-08-14 15:22:18 -07:00 |
|
|
|
03c1d58112
|
perf: add H200 Triton MoE configs for E256 N512 (#34150)
Co-authored-by: Vedant Jhaveri <vjhaveri@linkedin.com>
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
|
2026-08-14 14:37:58 -07:00 |
|
|
Sidhartha Reddy PotuandXinyuan Tong
|
42e8718d3d
|
fix(muse-glimmer): parse required/named tool calls natively (#34781)
Co-authored-by: Xinyuan Tong <xinyuantong.cs@gmail.com>
|
2026-08-14 20:43:48 +00:00 |
|
|
Mohammad Miadh Angkad
|
d7207be156
|
Fix startup weight load after TorchAO removal (#34869)
|
2026-08-14 13:18:53 -07:00 |
|
|
Jackey HuaandClaude Opus 5
|
1af761a09a
|
[SM12x] Default the fused MHC post+pre path on (#34019)
Co-authored-by: Claude Opus 5 <noreply@anthropic.com>
|
2026-08-14 19:51:21 +00:00 |
|
|
|
41cd5a7189
|
[Fix] Fix Qwen3.5 MTP startup with HiCache (#34560)
Co-authored-by: hjzhang <76768149+1e4ves@users.noreply.github.com>
Co-authored-by: YAMY <74099316+YAMY1234@users.noreply.github.com>
|
2026-08-14 10:55:41 -07:00 |
|
|
Xun Sun
|
d8399af70c
|
fix(qwen3): support DeepEP-class backends and early EPLB state (#34810)
|
2026-08-15 01:27:39 +08:00 |
|
|
Ke Bao
|
7562e741e2
|
Retain SWA down to the last state checkpoint (#34729)
|
2026-08-15 01:12:38 +08:00 |
|
|
Ke Bao
|
c20aceeb88
|
Fix mamba checkpoint depth under dcp (#34808)
|
2026-08-15 00:34:00 +08:00 |
|
|
Xiaoyu Zhang
|
9c9a3273be
|
[diffusion] Fix Helios denoising profiler stepping (#34826)
|
2026-08-14 23:21:54 +08:00 |
|
|
Xiaoyu Zhang
|
5f2a6d6422
|
[diffusion] Fix symbolic replicated-mode counting under torch.compile (#34824)
|
2026-08-14 23:20:29 +08:00 |
|
|
Yuan Luoandluoyuan.luo
|
b95a746948
|
[MoE] Fuse swiglu moe up gemm epilogue (#32944)
Co-authored-by: luoyuan.luo <luoyuan.luo@antgroup.com>
|
2026-08-14 22:11:34 +08:00 |
|
|
Mohammad Miadh Angkad
|
9d2f1584fa
|
Fix MiniMax-H3 Cache-DiT BCG warning (#34848)
|
2026-08-14 22:07:50 +08:00 |
|
|
|
c939307e8a
|
[MiniMax-M3] Overlap shared and routed experts (#34542)
Co-authored-by: xuebi <xuebi@minimaxi.com>
Co-authored-by: Xiaoyu Zhang <1182563586@qq.com>
|
2026-08-14 22:01:48 +08:00 |
|
|
|
1a178f7c7c
|
[EPD] Batch embedding cache host-device range copies (#31574)
Co-authored-by: 晟海 <huangtingwei.htw@antgroup.com>
Co-authored-by: liusy58 <liusy58@linux.alibaba.com>
Co-authored-by: Xiaoyu Zhang <1182563586@qq.com>
|
2026-08-14 22:01:12 +08:00 |
|
|
Brayden ZhongandBrayden Zhong
|
5e65dd01a7
|
Remove the torchao integration (--torchao-config) (#34304)
Co-authored-by: Brayden Zhong <brayden@radixark.ai>
|
2026-08-14 21:49:11 +08:00 |
|
|
Ke Bao
|
18107e38d2
|
Skip oow slot freeing under eagle (#34823)
|
2026-08-14 17:13:21 +08:00 |
|
|
chilltongxandchilltongx
|
4d94f1d310
|
[diffusion] fix: warn when bcg disables cache-dit (#34242)
Co-authored-by: chilltongx <284668524+chilltongx@users.noreply.github.com>
|
2026-08-14 16:45:36 +08:00 |
|
|
Liangsheng Yin
|
f2c84de022
|
[Perf] Publish the WAR read-done event at DSPARK verify (#34816)
|
2026-08-14 00:46:52 -07:00 |
|
|
triple-muandMick
|
a86edcdc0a
|
[diffusion] feat: rebuild minimax-h3 adaln outputs on demand (#34650)
Co-authored-by: Mick <mickjagger19@icloud.com>
|
2026-08-14 15:33:51 +08:00 |
|
|
|
2622e013eb
|
[Fix] has_hf_quant_config crashes on local dirs without the config (#34774)
Co-authored-by: harmya <harmya@modal.com>
Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
|
2026-08-14 07:17:42 +00:00 |
|
|
kk
|
65d62109dd
|
[AMD] Fix Triton 3.7 gfx950 extend-attention spills (#34741)
|
2026-08-14 00:05:42 -07:00 |
|
|
Dayananda V
|
7c15b9b7d0
|
[diffusion] fix: fix cache-first fast path accepting a metadata-only snapshot (#34121)
|
2026-08-14 14:52:52 +08:00 |
|
|
datdo-msftandXiaoyu Zhang
|
827552bc1d
|
Fix eager AMX backend probe imports (#34496)
Co-authored-by: Xiaoyu Zhang <1182563586@qq.com>
|
2026-08-14 14:44:14 +08:00 |
|
|
|
ba1d980b35
|
[AMD] Accelerate AITER unified-attention decode with scaled FP8 Q (#31856)
Co-authored-by: jacky.cheng <yichiche@amd.com>
Co-authored-by: YC Yen-Ching Tseng <yctseng@amd.com>
|
2026-08-13 23:42:25 -07:00 |
|
|
Mick
|
46d84f4b48
|
feat(cli): add extensible serve backend plugins (#34753)
|
2026-08-14 13:57:59 +08:00 |
|
|
|
b784726863
|
[kernel] Content-addressed JIT build cache, generated from our own ninja (#34274)
Co-authored-by: Claude <noreply@anthropic.com>
Co-authored-by: BBuf <1182563586@qq.com>
|
2026-08-14 13:41:04 +08:00 |
|
|
jacky.chengandsogalin_codegen
|
240a12b302
|
[AMD] Fuse shared_expert_gate GEMV into the MoE append kernel (HIP/aiter) (#28666)
Co-authored-by: sogalin_codegen <39478626+sogalin@users.noreply.github.com>
|
2026-08-13 21:32:27 -07:00 |
|
|
Baizhou Zhang
|
0a6bbbe128
|
[Fix] Restore layer-level DSV4 RoPE policy (#34788)
|
2026-08-13 21:26:45 -07:00 |
|
|
|
e1c4db9621
|
[quantization] Add tuned Triton tile configs for channelwise FP8 GEMM… (#34331)
Co-authored-by: runkai <runkai@amazon.com>
Co-authored-by: Xiaoyu Zhang <1182563586@qq.com>
|
2026-08-14 11:11:02 +08:00 |
|
|
Xinyuan Tongandhnyls2002
|
85cdf1178d
|
[CI] Prune redundant CPU test overhead (#34309)
Co-authored-by: hnyls2002 <lsyincs@gmail.com>
|
2026-08-13 19:51:17 -07:00 |
|
|
SII-yangdianandSII-yangdian
|
f2b2b567aa
|
perf(jit_kernel/deepseek_v4): optimize paged_mqa_metadata (#25855)
Co-authored-by: SII-yangdian <yangdian@sii.edu.cn>
|
2026-08-13 19:00:22 -07:00 |
|
|
Xiaoyu ZhangandClaude Fable 5
|
704e512836
|
[GDN] Honor configured linear-attn verify backend in the kernel dispatcher (#34592)
Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
|
2026-08-14 08:59:52 +08:00 |
|
|
Ziang Li
|
9d34c2809f
|
[FlashInfer v0.6.16] Support FlashInfer CuTe DSL NVFP4 MoE quantization (#28354)
|
2026-08-13 17:33:46 -07:00 |
|
|
ormandj
|
c4271c3fe1
|
[DSpark] Fix EP1 decode performance regression (#34759)
|
2026-08-13 17:00:54 -07:00 |
|
|
weireweireandweireweire
|
54c44fef73
|
[Perf] Skip trivial DSV4 nonpaged indexer logits (#33857)
Co-authored-by: weireweire <20922698+weireweire@users.noreply.github.com>
|
2026-08-13 16:21:30 -07:00 |
|
|
Liangsheng Yin
|
151a314829
|
[Fix] Make the DSpark draft num_token_non_padded host-to-device copy non-blocking (#34782)
|
2026-08-13 16:10:54 -07:00 |
|
|
Khoa PhamandClaude Opus 5
|
81fe452810
|
[DCP] Share one pack kernel between both a2a backends (#34651)
Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
|
2026-08-13 15:06:02 -07:00 |
|
|
Lianmin Zheng
|
8bbca87780
|
[Core] Organize environment variable registry (#34730)
|
2026-08-13 14:25:37 -07:00 |
|
|
jthomson04
|
903439044a
|
perf(kv-events): coalesce cache events (#31479)
|
2026-08-13 13:36:53 -07:00 |
|
|
Liangsheng Yin
|
8554d9a5bc
|
[Fix] Carry the backend on Kimi-K3 deferred preprocessing configs (#34766)
|
2026-08-13 13:30:33 -07:00 |
|
|
Lukas Humbel
|
8ad04a9bee
|
docs(nixl): document OBJ throughput target (#30405)
|
2026-08-13 13:04:13 -07:00 |
|
|
|
6b94d39f13
|
[Model Loading] Overlap checkpoint staging with CUDA graph capture during startup (#32017)
Co-authored-by: Wenhui Zhu <wzhu59@asu.edu>
Co-authored-by: Alex Nails <alex.nails@radixark.ai>
|
2026-08-13 12:26:25 -07:00 |
|
|
Shangming Cai
|
0772e79ee7
|
[CI][PD] Pin nccl rendezvous port per side to fix flaky disaggregation tests (#34755)
|
2026-08-14 01:44:20 +08:00 |
|
|
jambow0320
|
a82f8e1777
|
[PD] Add the missing Prefill bootstrap timeout for NIXL (#34692)
|
2026-08-14 01:08:09 +08:00 |
|
|
|
69a31ce342
|
fix: make Cache-DiT actually cache on MiniMax-H3 (#33827)
Signed-off-by: YZLi <yuanli@nvidia.com>
Signed-off-by: yunch <yunch@nvidia.com>
Co-authored-by: YZLi <yuanli@nvidia.com>
Co-authored-by: Mick <mickjagger19@icloud.com>
Co-authored-by: Xiaoyu Zhang <1182563586@qq.com>
|
2026-08-14 00:39:58 +08:00 |
|