Commits

Branch selector

 main

User selector

tlrmchlsmth

Datepicker

All time

Commit History

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

Commits on May 5, 2025

• [Build/CI] Upgrade CUTLASS to 3.9.2 (#17641)

  Show description for f62cad6
  tlrmchlsmth

  authored
  f62cad6

  Copy full SHA for f62cad6

• [Bugfix][Easy] Fix whitespace in shm_broadcast.py logging (#17635)

  Show description for 68e1ee0
  tlrmchlsmth

  authored
  68e1ee0

  Copy full SHA for 68e1ee0

Commits on May 3, 2025

• [Build/CI] Upgrade CUTLASS to 3.9.1 (#17602)

  Show description for c8386fa
  tlrmchlsmth

  authored
  c8386fa

  Copy full SHA for c8386fa

Commits on Apr 15, 2025

• [Bugfix] Fix tests/kernels/test_mamba_ssm_ssd.py (#16623)

  Show description for dbb036c
  tlrmchlsmth

  authored
  dbb036c

  Copy full SHA for dbb036c

Commits on Mar 26, 2025

• [Misc] Warn about v0 in benchmark_paged_attn.py (#15495)

  Show description for 23114d3
  tlrmchlsmth

  authored
  23114d3

  Copy full SHA for 23114d3

Commits on Mar 25, 2025

• Revert "Fix non-contiguous input passed to Marlin kernel (#15319)" (#15398)

  tlrmchlsmth

  authored
  b5269db

  Copy full SHA for b5269db

Commits on Mar 17, 2025

• [Bugfix] Fix precommit - line too long in pixtral.py (#14960)

  Show description for e1eb45d
  

  tlrmchlsmth

  and

  DarkLight1337

  authored
  e1eb45d

  Copy full SHA for e1eb45d

Commits on Mar 15, 2025

• Revert "[Model] Mamba2 Prefill Performance Tweaks: Fixing Flurry of U… (#14848)

  tlrmchlsmth

  authored
  ccf02fc

  Copy full SHA for ccf02fc

Commits on Mar 8, 2025

• Revert "[Perf] Reduce MLA CPU overheads in V1 (#14384)" (#14471)

  tlrmchlsmth

  authored
  ca7a2d5

  Copy full SHA for ca7a2d5

• [Bugfix][V1] Handle MLA in kv_cache_interface (#14462)

  Show description for 3336814
  tlrmchlsmth

  authored
  3336814

  Copy full SHA for 3336814

Commits on Mar 6, 2025

• [Bugfix] Fix use_direct_call condition in FusedMoE layer for (#14382)

  Show description for 958adce
  tlrmchlsmth

  authored
  958adce

  Copy full SHA for 958adce

• [Kernel] Add needs_fixed_stride_order tag to most GEMMs (#14306)

  Show description for 99b0915
  tlrmchlsmth

  authored
  99b0915

  Copy full SHA for 99b0915

Pagination

Previous

Next