Commits

Branch selector

 main

User selector

tlrmchlsmth

All time

Commit History

Commits on Apr 20, 2026

• [WideEP] Remove naive all2all. Use allgather_reducescatter instead (#33728)

  Show description for 81d954f
  tlrmchlsmth

  authored
  81d954f

  Copy full SHA for 81d954f

Commits on Apr 13, 2026

• [Bugfix] Reject non-nvfp4 dtypes when using the flashinfer_nvlink_one_sided all2all backend (#39717)

  Show description for f72b209
  

  tlrmchlsmth

  and

  claude

  authored
  f72b209

  Copy full SHA for f72b209

Commits on Mar 6, 2026

• Change "following fields were present in the request but ignored" log from warn to debug (#36173)

  Show description for 5afb387
  tlrmchlsmth

  authored
  5afb387

  Copy full SHA for 5afb387

Commits on Mar 2, 2026

• Fix unresolved-import errors when using Astral's ty by removing src.root (#35681)

  Show description for de7dd63
  tlrmchlsmth

  authored
  de7dd63

  Copy full SHA for de7dd63

Commits on Feb 26, 2026

• [WideEP] Remove pplx all2all backend (#33724)

  Show description for eb19955
  

  tlrmchlsmth

  and

  claude

  authored
  eb19955

  Copy full SHA for eb19955

Commits on Feb 11, 2026

• [Bugfix] Fix fused MoE IMA (sans chunking) by using int64 for strides (#34279)

  Show description for d7982da
  

  tlrmchlsmth

  and

  claude

  authored
  d7982da

  Copy full SHA for d7982da

• [WideEP] Fix nvfp4 DeepEP High Throughput All2All backend (#33738)

  Show description for 066c6da
  

  tlrmchlsmth

  and

  robertgshaw2-redhat

  authored
  066c6da

  Copy full SHA for 066c6da

Commits on Feb 10, 2026

• [Misc] Add pre-commit hook to catch boolean ops in with-statements (#34271)

  Show description for c4b9e67
  

  tlrmchlsmth

  and

  claude

  authored
  c4b9e67

  Copy full SHA for c4b9e67

Commits on Jan 22, 2026

• [UX] Default api_server_count to dp_size if not specified (#32525)

  Show description for 803e3f3
  tlrmchlsmth

  authored
  803e3f3

  Copy full SHA for 803e3f3

Commits on Jan 7, 2026

• Change warning in get_current_vllm_config to report caller's line number (#31855)

  Show description for f09c5fe
  tlrmchlsmth

  authored
  f09c5fe

  Copy full SHA for f09c5fe

Commits on Jan 5, 2026

• [Bugfix] Fix EPLB state logging error (#31455)

  Show description for fe8a9fb
  tlrmchlsmth

  authored
  fe8a9fb

  Copy full SHA for fe8a9fb

Commits on Nov 24, 2025

• Remove VLLM_SKIP_WARMUP tip (#29331)

  Show description for 4dd42db
  tlrmchlsmth

  authored
  4dd42db

  Copy full SHA for 4dd42db

Commits on Nov 3, 2025

• [Bugfix] Fix MoE Routing Simulation (#28002)

  Show description for 3758757
  tlrmchlsmth

  authored
  3758757

  Copy full SHA for 3758757

Commits on Oct 30, 2025

• [Bugfix] Fix 2 precommit issues - (mamba_block_size, kv_cache_config) (#27811)

  Show description for ab98f65
  

  tlrmchlsmth

  and

  njhill

  authored
  ab98f65

  Copy full SHA for ab98f65

Commits on Oct 21, 2025

• [P/D] KVConnector for decode benchmarking (#25986)

  Show description for 6c2eef5
  tlrmchlsmth

  authored
  6c2eef5

  Copy full SHA for 6c2eef5

Commits on Oct 14, 2025

• [WideEP][P/D] Add usage stats for DP+EP and KV Connector (#26836)

  Show description for 579d2e5
  tlrmchlsmth

  authored
  579d2e5

  Copy full SHA for 579d2e5

Commits on Oct 3, 2025

• [Build/CI] Revert back to Ubuntu 20.04, install python 3.12 with uv (#26103)

  Show description for 27edd2a
  

  tlrmchlsmth

  and

  simon-mo

  authored
  27edd2a

  Copy full SHA for 27edd2a

Commits on Sep 27, 2025

• [Bugfix][WideEP] Apply TP Attn + EP MoE fix to other models (#24982)

  Show description for a5354b3
  tlrmchlsmth

  authored
  a5354b3

  Copy full SHA for a5354b3

• [Bugfix] Fix triton import precommit failure (#25803)

  Show description for f9df8b4
  tlrmchlsmth

  authored
  f9df8b4

  Copy full SHA for f9df8b4

Commits on Sep 25, 2025

• [Logging] Remove TORCH_NCCL_AVOID_RECORD_STREAMS to squash a warning (#25532)

  Show description for 2e5df88
  tlrmchlsmth

  authored
  2e5df88

  Copy full SHA for 2e5df88

• Revert "[Performance] Move apply_w8a8_block_fp8_linear to an op class… (#25607)

  Show description for 1260180
  tlrmchlsmth

  authored
  1260180

  Copy full SHA for 1260180

Commits on Sep 24, 2025

• [Logging] Improve log for when DeepEP HT disables CUDA Graphs (#25531)

  Show description for fea8006
  tlrmchlsmth

  authored
  fea8006

  Copy full SHA for fea8006

Commits on Sep 10, 2025

• [Docs] Document the extra memory footprint overhead when using EPLB (#24537)

  Show description for 8b83b93
  tlrmchlsmth

  authored
  8b83b93

  Copy full SHA for 8b83b93

• [Bugfix] Improve EPLB config validation error message (#24524)

  Show description for 561f38d
  tlrmchlsmth

  authored
  561f38d

  Copy full SHA for 561f38d

Commits on Sep 9, 2025

• [Bugfix][Wide EP] Fix redundant work when using DeepEP, TP Attn, and EP MoE (#24134)

  Show description for 955c624
  tlrmchlsmth

  authored
  955c624

  Copy full SHA for 955c624

Commits on Aug 4, 2025

• [CI Bugfix] Fix wNa16 kernel not found for test_shared_storage_connector_hashes (#22163)

  Show description for 8ecb3e9
  tlrmchlsmth

  authored
  8ecb3e9

  Copy full SHA for 8ecb3e9

Commits on Jul 14, 2025

• Change default model to Qwen3-0.6B (#20335)

  Show description for 5597562
  tlrmchlsmth

  authored
  5597562

  Copy full SHA for 5597562

Commits on Jul 3, 2025

• [Bugfix] Fixes for FlashInfer's TORCH_CUDA_ARCH_LIST (#20136)

  Show description for bdb84e2
  tlrmchlsmth

  authored
  bdb84e2

  Copy full SHA for bdb84e2

Commits on Jul 2, 2025

• [Kernel][Bugfix] Fixup some warnings in nvfp4_blockwise_moe when CUDA < 12.8 (#20324)

  Show description for 3be8d31
  tlrmchlsmth

  authored
  3be8d31

  Copy full SHA for 3be8d31

Commits on Jun 27, 2025

• [Bugfix] Fix some narrowing conversion warnings (#20141)

  Show description for e8c3bd2
  tlrmchlsmth

  authored
  e8c3bd2

  Copy full SHA for e8c3bd2

Commits on Jun 26, 2025

• Spam folks if config.py changes (#20131)

  Show description for 0bceac9
  tlrmchlsmth

  authored
  0bceac9

  Copy full SHA for 0bceac9

Commits on Jun 23, 2025

• [EP+DP] Optimize the little operations in the DeepGEMM + DeepEP low latency case (#19885)

  Show description for 68aaeb3
  

  tlrmchlsmth

  and

  varun-sundar-rabindranath

  authored
  68aaeb3

  Copy full SHA for 68aaeb3

Commits on Jun 4, 2025

• [Bugfix][EP+DP] Fix internode check (#19112)

  Show description for d459fae
  tlrmchlsmth

  authored
  d459fae

  Copy full SHA for d459fae

Commits on Jun 3, 2025

• [Bugfix][EP+DP] Use pplx-kernel internode instead of intranode (#19034)

  Show description for 8a57872
  tlrmchlsmth

  authored
  8a57872

  Copy full SHA for 8a57872

Commits on May 22, 2025

• [Build/CI] Fix CUDA 11.8 build (#17679)

  Show description for 6e588da
  

  tlrmchlsmth

  and

  LucasWilkinson

  authored
  6e588da

  Copy full SHA for 6e588da