Commit Graph
151 Commits
Author SHA1 Message Date
eb9c9ee99d [Qwen4-Exp] Optional fp8 (e4m3) storage for the compressed QSA indexer cache (#39614)
Co-authored-by: Claude Fable 5.1 <noreply@anthropic.com>
Co-authored-by: YAMY1234 <74099316+YAMY1234@users.noreply.github.com>
2026-09-29 22:11:46 -07:00
Cheng Wan 5781206833 [PD] Give FakeKVReceiver ensure_abort_notified (#41618) 2026-09-28 21:59:17 -07:00
7889fbab9a [PD] Keep EAGLE DP graph and token metadata consistent (#32196)
Co-authored-by: weireweire <20922698+weireweire@users.noreply.github.com>
Co-authored-by: Po-Han Huang (NVIDIA) <53919306+nvpohanh@users.noreply.github.com>
Co-authored-by: Khoa Pham <khoa.pham@radixark.ai>
2026-09-28 17:15:51 -07:00
Nan Jiang 6fa3fe69e2 [PD] Keep the sampling mask of a replayed rebootstrap token (#41235) 2026-09-28 12:02:27 -07:00
Shangming CaiandClaude Fable 5 2a1c477648 [PD] Defer decode KV release on every transfer failure, not only decode-initiated aborts (#41404)
Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2026-09-29 02:19:51 +08:00
Shangming CaiandClaude Fable 5 14ab74ef95 [PD] Fan drain abort ACKs out to every decode peer of the room (#41402)
Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2026-09-28 05:16:23 +08:00
Liangsheng Yin d27efca353 [mem_cache] Never free the protected prefix on request release (#41312) 2026-09-26 22:00:02 -07:00
Xiaoyu Zhang 38ec649048 [Refactor] Restore logical kernel groups and test organization (#41243) 2026-09-26 12:30:44 +08:00
Liangsheng Yin f35ec18d8b [Test] Remove more unit tests that mirror implementation or never run in CI (#41297) 2026-09-25 19:22:14 -07:00
Liangsheng YinandBBuf efd9a40bb8 [PD] Share one head-slice helper across mooncake, mori, and nixl (#39660)
Co-authored-by: BBuf <1182563586@qq.com>
2026-09-25 16:13:18 -07:00
cf5df82680 [KVCache] Support lmcache unified radix cache (#38652)
Signed-off-by: chunxiaozheng <1179548172@qq.com>
Co-authored-by: Yuwei An <ayw.sirius19@gmail.com>
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-09-25 14:08:45 -07:00
Nan Jiang 24b6930076 [Sampling] Stream sampling masks as per-request arrays (#40986) 2026-09-25 12:20:25 -07:00
78b382b1ad Support unified memory decode host pools (#39478)
Co-authored-by: yhzhuang <yhzhuang@fb.com>
Co-authored-by: Lianmin Zheng <lianminzheng@gmail.com>
Co-authored-by: Yonghao Zhuang <yhzhuang@users.noreply.github.com>
Co-authored-by: Cheng Wan <cheng.wan@radixark.ai>
2026-09-24 11:02:15 -07:00
Shangming CaiandClaude Opus 5 cd11037009 [PD] Enable deferred decode-side KV release by default (#41023)
Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
2026-09-24 22:35:11 +08:00
cctryandcctry efff836911 [PD] Add decode host receive for custom transfer backends (#40238)
Co-authored-by: cctry <cctry@fb.com>
2026-09-23 21:57:33 -07:00
Nan Jiang c3685dff42 [Sampling] Add selected/support sampling logprob modes (#40932) 2026-09-23 21:38:19 -07:00
YAMYandzhangxiaolei 6fd98c98b9 [Qwen3.8-Next] Pipeline-parallel serving and PD-prefill MTP for Qwen4-Exp (#40501)
Co-authored-by: zhangxiaolei <zhangxiaolei.666@bytedance.com>
2026-09-22 08:46:44 -07:00
cctry 8ef6d31cb9 [PD] Preserve abort ACKs until in-flight KV transfers drain (#40645) 2026-09-22 18:46:07 +08:00
Po-Han Huang (NVIDIA) 771c9d782d [PD] Validate Mooncake EFA allocator compatibility (#39973) 2026-09-22 15:12:29 +08:00
Guangda LiuandGuangda Liu 04c0913434 [HiSparse] Add MHA hisparse support for MiniMax M3 (#31446)
Co-authored-by: Guangda Liu <bingps@users.noreply.github.com>
2026-09-22 13:28:03 +08:00
Khoa PhamandQiaolin Yu 018b73c7a0 [PD] Pack draft KV head slices for DCP transfers (#40500)
Co-authored-by: Qiaolin Yu <liin1211@outlook.com>
2026-09-21 21:11:27 -07:00
Cheng Wan 506698761d [unified-memory] Hierarchical cache for every unified pool shape (#37507) 2026-09-21 16:50:37 -07:00
Cheng Wan acac4dd9d9 [Refactor] Clean up parallel runtime comments (#40632) 2026-09-21 14:32:22 -07:00
Cheng Wan 970e946e4f Retire the per-runner parallel record (#40343) 2026-09-21 12:26:40 -07:00
Cheng Wan 73f071db52 Deprecate the parallel getters the context answers, and ratchet them shut (#40342) 2026-09-21 12:25:32 -07:00
kpjeeja b54d5b7c7b disaggregation: Fix FakeKVSender queue accumulation (#28652)
Signed-off-by: KP, Jeeja <jeeja.kp@intel.com>
2026-09-21 14:27:05 +08:00
Khoa Pham b3e4d198af [PD] Bound cached-prefix DCP transfers by pack capacity (#40376) 2026-09-21 01:17:15 +08:00
Qiaolin Yu f4c256354c [kimi k3][pd disagg] support pp prefill + dcp decode with dspark (#40045) 2026-09-20 00:15:28 -07:00
3a64faa1f2 Fix disagg PP MTP for GLM-5.2 (#39378)
Co-authored-by: Julien Lin <jullin@nvidia.com>
Co-authored-by: YAMY1234 <74099316+YAMY1234@users.noreply.github.com>
Co-authored-by: Yangmin Li <yangminl@nvidia.com>
2026-09-19 13:55:07 -07:00
8189e3896b [PD] Skip singleton transfer-status all-reduces (#40003)
Co-authored-by: Pranjal Shankhdhar <pranjalssh@users.noreply.github.com>
Co-authored-by: Jialin Ouyang <Jialin.Ouyang@gmail.com>
2026-09-18 22:41:50 -07:00
5e9342d16f [PP][DeepSeek V4] Overlap communication and optimize SM120 prefill (#38792)
Co-authored-by: Yangmin Li <yangminl@nvidia.com>
Co-authored-by: YAMY <74099316+YAMY1234@users.noreply.github.com>
2026-09-18 21:18:04 -07:00
Cheng Wan afe71f4b9e Read process groups through the runtime context (#40068) 2026-09-18 17:40:32 -07:00
5931fd60ee Support unified memory page-envelope transfers in PD (#39477)
Co-authored-by: yhzhuang <yhzhuang@fb.com>
Co-authored-by: Lianmin Zheng <lianminzheng@gmail.com>
Co-authored-by: Yonghao Zhuang <yhzhuang@users.noreply.github.com>
Co-authored-by: Cheng Wan <cheng.wan@radixark.ai>
2026-09-18 17:39:50 -07:00
Liangsheng Yin 6cc9090d1f [mem_cache] Release up to owned_kv_len on radix cache insert (#40075) 2026-09-18 15:37:52 -07:00
a6cf05817f dsv4.1: remaining model and runtime integration (#38798)
Co-authored-by: BBuf <1182563586@qq.com>
Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
Co-authored-by: Xiaoyu Zhang <xiaoyu.zhang@radixark.ai>
Co-authored-by: Yuwei An <ayw.sirius19@gmail.com>
Co-authored-by: Khoa Pham <khoa.pham@radixark.ai>
Co-authored-by: Yuhao Yang <47235274+yhyang201@users.noreply.github.com>
Co-authored-by: Zhichen Zeng <zczeng@uw.edu>
Co-authored-by: Ke Bao <ispobaoke@gmail.com>
2026-09-18 02:55:30 -07:00
iridiumine f86f60081d [NPU] Adapt hicache for K3 hybrid models (#39415) 2026-09-18 14:48:20 +08:00
Liangsheng Yin 1f0c73e9bd [DSV4] Generalize attention metadata, sparse prefill, and KV pool over compress ratios (#39921) 2026-09-17 15:55:19 -07:00
1f60ddef5d [PD] Introduce runtime role switching between prefill and decode (#28403)
Signed-off-by: huanglong <huanglong@linux.alibaba.com>
Signed-off-by: inkcherry <mingzhi.liu@amd.com>
Co-authored-by: huanglong <huanglong@linux.alibaba.com>
Co-authored-by: Shangming Cai <csmthu@gmail.com>
Co-authored-by: Huang Long <121648372+LLLL114@users.noreply.github.com>
2026-09-18 01:45:12 +08:00
Aurick QiaoandAurick Qiao 7b7620774c Fix device context during NIXL backend initialization (#38774)
Co-authored-by: Aurick Qiao <6137920+aurickq@users.noreply.github.com>
2026-09-16 17:59:24 +08:00
Shangming CaiandClaude Opus 5 3f8eb35ead [PD] Do not admit intake-rejected requests to a PD handoff (#38935)
Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
2026-09-16 12:44:49 +08:00
Jinyan Yi dbd7281b06 fix(npu): fix hybrid KV transfer with PP prefill in PD disaggregation (#38402) 2026-09-16 09:10:17 +08:00
2929a39927 Use a shared byte budget for unified hybrid-SWA memory (#36729)
Co-authored-by: yhzhuang <yhzhuang@fb.com>
Co-authored-by: Cheng Wan <cheng.wan@radixark.ai>
2026-09-15 15:27:00 -07:00
Siyuan ChenandBJWang-ant 832ec39cc0 [Intra-node PD][DSV4] Pack all layers into one batch for INTRA_NODE_NVLINK path (#38984)
Co-authored-by: BJWang-ant <wangbingjia.wbj@ant-intl.com>
2026-09-15 21:08:33 +08:00
Yanbin Jiang ebd37705e4 [PD][LoRA] Gate decode admission on adapter slots (#39332) 2026-09-15 13:11:04 +08:00
ishandhanani 3f871a246c feat(agent sessions): attribute stored KV cache blocks to sessions (#37482)
Signed-off-by: Ishan Dhanani <ishandhanani@gmail.com>
2026-09-14 21:34:49 -07:00
cctry dad8c074e7 Scope prefetch cache state to the request attempt (#39318) 2026-09-14 15:03:03 -07:00
4c71d14fba [HiCache][LoRA] Simplify decode offload hash inputs (#39162)
Co-authored-by: Shuwen Wang <47200617+alphabetc1@users.noreply.github.com>
Co-authored-by: alphabetc1 <2508695655@qq.com>
2026-09-14 11:23:31 +08:00
Cheng Wan 2ec4bbcbd4 [unified-memory] PD disaggregation for every unified pool shape (#37506) 2026-09-13 19:16:43 -07:00
Yanbin JiangandShuwen Wang 0b415fa573 [HiCache][LoRA] Isolate storage pages by extra key (#38577)
Co-authored-by: Shuwen Wang <47200617+alphabetc1@users.noreply.github.com>
2026-09-12 18:18:14 +08:00
YAMY 55e5e21c88 [Qwen3.8-Next] Add PD state transfer for Flash Next (#36651) 2026-09-11 22:31:45 -07:00