Commits

Branch selector

 main

User selector

vllmellm

All time

Commit History

Commits on Feb 6, 2026

• [DOC] [ROCm] Update docker deployment doc (#33971)

  Show description for aaa2efb
  

  3 peopleauthored
  aaa2efb

  Copy full SHA for aaa2efb

Commits on Jan 30, 2026

• [Doc] [ROCm] Update Documentation to reflect v0.15.0 release (#33388)

  Show description for 174f167
  vllmellm

  authored
  174f167

  Copy full SHA for 174f167

Commits on Jan 20, 2026

• [Refactor] Make FP8 Linear Ops use kernel abstraction (#27814)

  Show description for 148117e
  vllmellm

  authored
  148117e

  Copy full SHA for 148117e

Commits on Jan 14, 2026

• [Bugfix][ROCm][performance] Resolve the performance regression issue of the Qwen3-Next-80B-A3B-Thinking under rocm_atten (#32336)

  Show description for e27078e
  vllmellm

  authored
  e27078e

  Copy full SHA for e27078e

Commits on Jan 9, 2026

• [Bugfix][ROCm]Fix Qwen3-Next-80B-A3B-Thinking inference and optimize non-standard block size (544) support under rocm_atten (#31380)

  Show description for 1a19e9c
  vllmellm

  authored
  1a19e9c

  Copy full SHA for 1a19e9c

Commits on Jan 7, 2026

• [ROCm][AITER] fix wrong argument passed to AITER `flash_attn_varlen_func` (#31880)

  Show description for 41cfa50
  vllmellm

  authored
  41cfa50

  Copy full SHA for 41cfa50

• [ROCm][AITER] bugfix accuracy regression in ROCM_AITER_TRITON_MLA backend (#31816)

  Show description for 6409004
  vllmellm

  authored
  6409004

  Copy full SHA for 6409004

Commits on Jan 6, 2026

• [Bugfix][ROCm] Fix Unsupported attention metadata type for speculative decoding in `eagle.py` (#31714)

  Show description for e971780
  vllmellm

  authored
  e971780

  Copy full SHA for e971780

Commits on Dec 23, 2025

• [ROCm][FEAT] Support AITER RMSNorm quantization fusion pass (#26575)

  Show description for f32cfd7
  

  vllmellm

  and

  tjtanaa

  authored
  f32cfd7

  Copy full SHA for f32cfd7

Commits on Dec 18, 2025

• [ROCm] Serving Fails on Radeon Due to AITER Dtype Import (#30952)

  Show description for 96bf50a
  vllmellm

  authored
  96bf50a

  Copy full SHA for 96bf50a

Commits on Dec 9, 2025

• [ROCm] Aiter Quant Kernels (#25552)

  Show description for ee14644
  vllmellm

  authored
  ee14644

  Copy full SHA for ee14644

Commits on Nov 25, 2025

• [Bugfix] [ROCm] [UX]: revert Flex attention backend (#29371)

  Show description for 64deead
  vllmellm

  authored
  64deead

  Copy full SHA for 64deead

Commits on Nov 24, 2025

• [Bugfix] [ROCm] [UX] Reorganize ROCm Backend Selection Logic (#26980)

  Show description for e48b2e6
  vllmellm

  authored
  e48b2e6

  Copy full SHA for e48b2e6

Commits on Nov 18, 2025

• [FEAT] [AITER] [ROCm] integrate aiter sampling ops (#26084)

  Show description for 0af3d4f
  vllmellm

  authored
  0af3d4f

  Copy full SHA for 0af3d4f

Commits on Nov 12, 2025

• [ROCM] Fix ROCm warnings, environment flag access, and GEMM kernel naming for consistency in `_aiter_ops.py` (#28464)

  Show description for d8140b9
  vllmellm

  authored
  d8140b9

  Copy full SHA for d8140b9

Commits on Nov 10, 2025

• [RFC][ROCm][AITER] Keep all AITER kernels in `_aiter_ops` class like `_custom_ops` and `_ipex_ops` (#24490)

  Show description for f080a83
  

  vllmellm

  and

  ProExpertProg

  authored
  f080a83

  Copy full SHA for f080a83

Commits on Nov 4, 2025

• [Bugfix][ROCm] Fix ViT rotary embeddings for torch.compile compatibility on ROCm (#27748)

  Show description for b13a447
  vllmellm

  authored
  b13a447

  Copy full SHA for b13a447

Commits on Oct 28, 2025

• [ROCm] [Doc] Update ROCm installation docs (#27327)

  Show description for 5b3c35a
  vllmellm

  authored
  5b3c35a

  Copy full SHA for 5b3c35a

Commits on Oct 22, 2025

• [DOC] [ROCm] Add ROCm quickstart guide (#26505)

  Show description for 265ecb0
  vllmellm

  authored
  265ecb0

  Copy full SHA for 265ecb0

Commits on Oct 17, 2025

• [Bugfix] [AITER] [ROCm] Fix Quark MoE Quant Config and AITER Fused MoE quant type logic (#27029)

  Show description for e33ee23
  vllmellm

  authored
  e33ee23

  Copy full SHA for e33ee23

Commits on Oct 14, 2025

• [CI] [ROCm] Automate CC list for ROCm related issue (#26753)

  Show description for 4821ac1
  vllmellm

  authored
  4821ac1

  Copy full SHA for 4821ac1

Commits on Oct 2, 2025

• [Qwen][ROCm] Flash Attention Rotary Embeddings (#24642)

  Show description for 5e4a822
  vllmellm

  authored
  5e4a822

  Copy full SHA for 5e4a822

Commits on Sep 23, 2025

• [docs] Benchmark Serving Incorrect Arg (#25474)

  Show description for 0d9fe26
  vllmellm

  authored
  0d9fe26

  Copy full SHA for 0d9fe26

Commits on Sep 16, 2025

• [Bug] [Spec Dec]: Fix kv_cache dtype mismatch for Eagle3 drafter on FP8 target (#24505)

  Show description for 8c54610
  vllmellm

  authored
  8c54610

  Copy full SHA for 8c54610

Commits on Sep 10, 2025

• [ROCm][Bugfix] Fix Aiter RMSNorm (#23412)

  Show description for 7c195d4
  vllmellm

  authored
  7c195d4

  Copy full SHA for 7c195d4

Commits on Aug 29, 2025

• [CI] Add `aiter` to matching list of issue auto labeller for `rocm` tag (#23942)

  Show description for 72a6913
  vllmellm

  authored
  72a6913

  Copy full SHA for 72a6913

Commits on Aug 26, 2025

• [CI] [Doc]: Add GH Action for auto labeling issues with `rocm` tag (#20988)

  Show description for 44ac25e
  

  vllmellm

  and

  DarkLight1337

  authored
  44ac25e

  Copy full SHA for 44ac25e

Commits on Aug 11, 2025

• [ROCm][AITER] Support AITER Rope ops in RotaryEmbedding Module. (#22521)

  Show description for 9c97a1c
  vllmellm

  authored
  9c97a1c

  Copy full SHA for 9c97a1c

Commits on Aug 7, 2025

• [Model] Switch to Fused RMS norm in Qwen2.5_VL model. (#22184)

  Show description for cbc8457
  

  vllmellm

  and

  kliuae

  authored
  cbc8457

  Copy full SHA for cbc8457

Commits on Aug 2, 2025

• [FEAT][ROCm] Enable running Flash Attention as ViT attn backend for Qwen-VL models on ROCm platform. (#22069)

  Show description for d3a6f21
  

  vllmellm

  and

  tjtanaavllm

  authored
  d3a6f21

  Copy full SHA for d3a6f21

• [Model] Qwen2.5 VL SiLU-and-Mul (#22066)

  Show description for ee2eb6e
  

  vllmellm

  and

  kliuae

  authored
  ee2eb6e

  Copy full SHA for ee2eb6e

Commits on Jul 2, 2025

• [ROCm][FEAT] Enable Full Graph Mode in AITER MLA V1 Attn Backend (Decode Phase only) (#20254)

  Show description for a1aafc8
  vllmellm

  authored
  a1aafc8

  Copy full SHA for a1aafc8

Commits on Jun 5, 2025

• [Misc] Remove unnecessary fallback to prefill-decode attention (#19138)

  Show description for 1809308
  vllmellm

  authored
  1809308

  Copy full SHA for 1809308

Commits on May 31, 2025

• [FEAT][ROCm] Add AITER grouped topk for DeepSeekV2 (#18825)

  Show description for 0f5e0d5
  vllmellm

  authored
  0f5e0d5

  Copy full SHA for 0f5e0d5

Commits on May 30, 2025

• [ROCm] Remove unnecessary assertion of max_model_len in ROCM_AITER_MLA attention backend. (#18938)

  Show description for 77b6e74
  vllmellm

  authored
  77b6e74

  Copy full SHA for 77b6e74

Pagination

Previous

Next