Commit Graph
34 Commits
Author SHA1 Message Date
Cheng Wan 01b5f572b7 [Refactor] Rename layer boundary protocols, contracts and helpers (#41750) 2026-09-29 21:05:48 -07:00
Cheng Wan de9f850b36 [Refactor] Rename the module to layer_boundary (#41555) 2026-09-28 20:17:50 -07:00
Cheng Wan 701213c867 [Refactor] Retire the layer facade and simplify boundary internals (#41554) 2026-09-28 20:17:33 -07:00
Cheng Wan acffb0d20d [Refactor] Select reduction fusion at the consumer (#41551) 2026-09-28 20:16:44 -07:00
Cheng Wan 81f27fb3a7 [Refactor] Replace LayerScatterModes with LayerFacts and remove ScatterMode (#41443) 2026-09-27 16:56:11 -07:00
Cheng Wan 0e907b755d [Refactor] Split the layer communicator into a package (move only) (#41439) 2026-09-27 16:54:41 -07:00
RuibinCheungandxiaobochen-amd 48c3854620 [ROCm] feat: enable aiter allreduce fusion for GLM models (#39790)
Co-authored-by: xiaobochen-amd <xiaobo.chen@amd.com>
2026-09-23 01:24:08 -07:00
Cheng Wan acac4dd9d9 [Refactor] Clean up parallel runtime comments (#40632) 2026-09-21 14:32:22 -07:00
Cheng Wan 2d0e94e3a3 Check the topology identities where the layout is written, and build at the published widths (#40340) 2026-09-21 12:22:59 -07:00
Cheng Wan afe71f4b9e Read process groups through the runtime context (#40068) 2026-09-18 17:40:32 -07:00
Alex NailsandAlison Shao 28262c20df [CI][RFC] Replace black-jupyter with ruff-format (#37210)
Co-authored-by: Alison Shao <a.shao@wustl.edu>
2026-09-02 19:46:08 -07:00
Bingxu Chen d067622820 [AMD][bugfix] Add moe_ep_size/moe_tp_size to the allreduce-fusion gate test stub (#35340) 2026-08-25 01:24:52 -07:00
Cheng Wan 47d8b5b749 config: route parallel config-leaf reads through get_parallel() (#33170)
The parallel namespace joins the accessor migration: 106 config-leaf reads
(enable_dp_lm_head, enable_dp_attention, pp_async_batch_depth, dp_size,
ep_join_rank_offset, dwdp_size, ...) flip from get_server_args()/
self.server_args to get_parallel(), which serves config leaves from the
published parallel bag via __getattr__.

- ParallelContext.__getattr__ is restructured to stay dynamo-traceable
  (object.__getattribute__ graph-breaks): gate helpers such as
  enable_moe_dense_fully_dp() run inside compiled model forwards. A
  fullgraph regression test pins the pattern.
- The five live-shadowed topology sizes (tp/pp/dcp/attn_cp/moe_dp_size)
  keep their server_args reads: the live @property wins on the accessor,
  and conditionally-initialized groups would fail loud at unconditional
  call sites.
- Elastic-EP scale writers (ep_size/dp_size x4 in model_runner) reroute
  to get_context().override together with their remaining instance
  readers (expert_location gpus-per-node paths); the ServerArgs.override
  ratchet drops 39 -> 35.
- The expert placement helpers (compute_logical_to_rank_dispatch_
  physical_map, _compute_logical_to_all_physical_map,
  _prefer_same_node_experts) now read everything from the bags and drop
  their server_args parameter; their unit tests publish the config they
  need instead of stubbing it.
2026-08-01 08:58:39 -07:00
6cc9352dfe [Spec] Add DSpark: confidence-scheduled speculative decoding (#30261)
Co-authored-by: sglang-bot <232288953+sglang-bot@users.noreply.github.com>
Co-authored-by: Claude Code <noreply@anthropic.com>
Co-authored-by: Codex <noreply@openai.com>
Co-authored-by: Liangsheng Yin <lsyincs@gmail.com>
Co-authored-by: Liangsheng Yin <hnyls2002@gmail.com>
2026-07-12 17:25:26 -05:00
Cheng Wan 1f15308dca [refactor] Retire the legacy config accessor and the remaining process singletons (#30493) 2026-07-09 02:10:47 -07:00
Cheng Wan b7cca0bf8f [refactor] Collect MoE and DP-attention runtime state into typed flag groups (#30347) 2026-07-07 21:29:27 -07:00
Cheng Wan be32c57598 [refactor] Read resolved config from server_args fields; retire the flags mirror tier (#30346) 2026-07-07 21:28:34 -07:00
Cheng Wan 8fb99bbaf8 [refactor] Config resolution pipeline: full-stack review (10-PR series, review only) (#30137) 2026-07-05 00:00:07 -07:00
Oxana Korzh f3904f0293 [bugfix][AMD] Disable aiter allreduce+RMSNorm fusion under DP attention / EP (#27835) 2026-07-02 15:31:25 -07:00
4d94e9471a [AMD] Relax allreduce-fusion residual accuracy tolerance to 1 bf16 ULP (#28226)
Co-authored-by: kangwangamd <kangwangamd@users.noreply.github.com>
Co-authored-by: Bingxu Chen <bingxche@amd.com>
Co-authored-by: Hubert Lu <55214931+hubertlu-tw@users.noreply.github.com>
2026-06-18 19:18:53 -07:00
7256ee9871 [AMD] Update test_aiter_allgather_amd.py data types alignment between benchmark aiter and custom all-reduce kernel (#27815)
Co-authored-by: YC Yen-Ching Tseng <yctseng@amd.com>
Co-authored-by: Hubert Lu <55214931+hubertlu-tw@users.noreply.github.com>
2026-06-17 00:48:15 -07:00
Hubert Lu 72929c7000 [AMD] Enable AITER custom all-gather on ROCm (#25093) 2026-06-02 15:57:37 -07:00
+2 deaba74745 [AMD][DSV4] DSV4 MTP graph + sparse triton attn optimizations (#26383)
Co-authored-by: wunhuang <wunhuang@amd.com>
Co-authored-by: Thomas Wang <1am9trash@gmail.com>
Co-authored-by: Xinyi Song <86638975+RolaoDenthu@users.noreply.github.com>
Co-authored-by: HaiShaw <hixiao@gmail.com>
Co-authored-by: amd-danli103 <danli103@amd.com>
Co-authored-by: Lin, Soga <soga.lin@amd.com>
Co-authored-by: Raiden-Makoto <Raiden-Makoto@users.noreply.github.com>
Co-authored-by: Hubert Lu <55214931+hubertlu-tw@users.noreply.github.com>
Co-authored-by: yichiche@amd.com <jacky.cheng>
Co-authored-by: yctseng0211 <yctseng@amd.com>
Co-authored-by: Bingxu Chen <bingxche@amd.com>
2026-05-27 15:23:35 -07:00
Liangsheng Yin 6cc2eee50d [misc] CI hygiene: enforce __main__ entry, drop silent-skipped tests, fix rerun-test protoc (#23305) 2026-04-20 21:16:24 -07:00
Hubert LuandHAI edaa5973d4 [AMD][No-Merge] Simplify fused allreduce + RMSNorm and remove hidden_dim allowlist (#21986)
Co-authored-by: HAI <hixiao@gmail.com>
2026-04-11 23:47:08 -07:00
265696b176 chore: update CI test est_time values (#22565)
Co-authored-by: Cheng Wan <chwan@rice.edu>
Co-authored-by: sglang-bot <sglang-bot@users.noreply.github.com>
Co-authored-by: Cheng Wan <54331508+ch-wan@users.noreply.github.com>
2026-04-10 18:15:02 -07:00
Lianmin Zheng 814202704b ci: unify PR test suite naming (#21187) 2026-03-23 00:18:45 -07:00
YC Tsengandbingxche b5edab57f2 [AMD] CI - Add MI35x nightly/PR tests for kv-cache-fp8 and allreduce-fusion (DeepSeek) (#19834)
Co-authored-by: bingxche <Bingxu.Chen@amd.com>
2026-03-05 07:09:57 -08:00
Hubert Luandyctseng0211 17b0affbdf [AMD] Support --enable-aiter-allreduce-fusion on AMD GPUs (#13747)
Co-authored-by: yctseng0211 <yctseng@amd.com>
2026-02-24 23:11:55 -08:00
Alison Shao 146b5fcc84 [CI] Reorganize stage-b 1-GPU tests for 5090 compatibility (#16826) 2026-01-15 15:23:35 -08:00
Alison Shao b880607108 Add 5090 dry run stage to PR test workflow (#17022) 2026-01-13 14:12:33 -08:00
Alison Shao cc25f9df50 Update est_time for stage-b-test-small-1-gpu tests (#16835) 2026-01-11 12:03:43 +08:00
YC TsengandBingxu Chen 4e999404c3 [AMD] Fix CI - unit-test-backend-1-gpu-amd-mi35x and unit-test-backend-2-gpu-amd, stage-b-test-small-1-gpu-amd (#16675)
Co-authored-by: Bingxu Chen <bingxche@amd.com>
2026-01-08 10:16:27 -08:00
Alison Shao 2e0527dd74 ci: migrate Debug Utils, Ops, and Rotary Embedding tests to test/registered/ (#16422) 2026-01-06 14:35:57 -08:00