Commits · 2da6fd722fe4b0fad758efd401690df47fca799c · OpenDAS / vllm_cscc · GitLab

23 Jun, 2025 1 commit
- update common.py · 2da6fd72
  zhuwenwen authored Jun 23, 2025
  
  2da6fd72
21 Jun, 2025 1 commit
- only K100_AI can use cutlass prefix cache · 14945681
  zhuwenwen authored Jun 21, 2025
  
  14945681
18 Jun, 2025 1 commit
- release opt1 · 1a280da1
  zhuwenwen authored Jun 18, 2025
  
  1a280da1
13 Jun, 2025 1 commit
- update triton_mla.py · e046b382
  zhuwenwen authored Jun 13, 2025
  
  e046b382
06 Jun, 2025 1 commit
- [Model] Add VLLM_USE_NN to use nn layout · 6880bf15
  zhuwenwen authored Jun 06, 2025
  
  6880bf15
04 Jun, 2025 1 commit
- [CPU] V1 support for the CPU backend (#16441) · 4555143e
  Li, Jiang authored Jun 04, 2025
  
  4555143e
03 Jun, 2025 2 commits
- [V1] Support cross-layer KV sharing (#18212) · bdf13965
  Yong Hoon Shin authored Jun 03, 2025
```
Signed-off-by: Yong Hoon Shin <yhshin@meta.com>
```
  bdf13965
- [Misc] Add SPDX-FileCopyrightText (#19100) · 02f0c7b2
  Simon Mo authored Jun 03, 2025
```
Signed-off-by: simon-mo <simon.mo@hey.com>
```
  02f0c7b2
30 May, 2025 2 commits
- [ROCm] Remove unnecessary assertion of max_model_len in ROCM_AITER_MLA attention backend. (#18938) · 77b6e74f
  vllmellm authored May 30, 2025
```
Signed-off-by: vllmellm <vllm.ellm@embeddedllm.com>
```
  77b6e74f
- update flash_attn.py · 58042164
  zhuwenwen authored May 30, 2025
  
  58042164
29 May, 2025 3 commits
- [ROCm][V0][Attention] Revert to the previous FA triton kernel (#18226) · 1b7cfd5a
  Gregory Shtrasberg authored May 29, 2025
```
Signed-off-by: Gregory Shtrasberg <Gregory.Shtrasberg@amd.com>
```
  1b7cfd5a
- [Attention][V1] Toggle for v1 attention backend (#18275) · da4b69d0
  Gregory Shtrasberg authored May 29, 2025
```
Signed-off-by: Gregory Shtrasberg <Gregory.Shtrasberg@amd.com>
```
  da4b69d0
- [test]fix basic_correctness and benchmarks · dbd62f84
  zhuwenwen authored May 29, 2025
  
  dbd62f84
28 May, 2025 5 commits
- [Bugfix][ROCm] fix the power of 2 exception from triton_unified_attention.py... · 269d9017
  Hongxia Yang authored May 28, 2025
```
[Bugfix][ROCm] fix the power of 2 exception from triton_unified_attention.py when running llama4 models and unit test fix (#18100)
Signed-off-by: Hongxia Yang <hongxia.yang@amd.com>
Signed-off-by: tjtanaa <tunjian.tan@embeddedllm.com>
Co-authored-by: tjtanaa <tunjian.tan@embeddedllm.com>
```
  269d9017
- update use_rocm_custom_paged_attention · df877aad
  zhuwenwen authored May 28, 2025
  
  df877aad
- [BugFix] FA2 MLA Accuracy Issue (#18807) · ce75efee
  Lucas Wilkinson authored May 28, 2025
```
Signed-off-by: LucasWilkinson <lwilkinson@neuralmagic.com>
```
  ce75efee
- set VLLM_FLASH_ATTN_BACKEND to use FlashAttention Backend for attention computation on rocm · 82f60bef
  zhuwenwen authored May 28, 2025
  
  82f60bef
- using triton prefix cache on BW · 07b41ddf
  zhuwenwen authored May 28, 2025
  
  07b41ddf
23 May, 2025 2 commits
- tbo增加ds int8支持，修改tb zerooverhead环境变量到envs.py中统一管理 · dbbb148b
  lizhigong authored May 23, 2025
  
  dbbb148b
- 修改MLA prefill阶段出现的Device2Host拷贝同步现象。 · 3f80d9ac
  lizhigong authored May 22, 2025
  
  3f80d9ac
22 May, 2025 1 commit
- 修改MLA prefill阶段出现的Device2Host拷贝同步现象 · 0c1fa562
  zhuwenwen authored May 22, 2025
  
  0c1fa562
21 May, 2025 1 commit
- [ROCm][Kernel][V1] Enable AMD Radeon GPU Custom Paged Attention on v1 (#17004) · dd5fa7e0
  Hosang authored May 21, 2025
```
Signed-off-by: Hosang Yoon <hosang.yoon@amd.com>
```
  dd5fa7e0
20 May, 2025 2 commits
- [Kernel] update comment for KV shape in unified triton attn (#18099) · 980a1724
  Percy authored May 20, 2025
```
Signed-off-by: haochengxia <xhc_1007@163.com>
```
  980a1724
- support Z100L and K100 inference · a5aa55e8
  zhuwenwen authored May 20, 2025
  
  a5aa55e8
19 May, 2025 1 commit
- 增加PA_TC kv cache int8/fp8支持 · e02d110d
  zhangshao authored May 19, 2025
  
  e02d110d
15 May, 2025 2 commits
- [Kernel] [V1] Fix performance regression for triton unified attention (#18161) · 01c22335
  Thomas Parnell authored May 15, 2025
```
Signed-off-by: Thomas Parnell <tpa@zurich.ibm.com>
Co-authored-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>
```
  01c22335
- [Bugfix] Fix fp8 tests for triton_unified_attention for Triton 3.3 (#18013) · e6b8e65d
  Thomas Parnell authored May 15, 2025
```
Signed-off-by: Thomas Parnell <tpa@zurich.ibm.com>
Co-authored-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>
```
  e6b8e65d
14 May, 2025 4 commits
- add kvint8 · 6d105cc4
  xiabo authored May 14, 2025
  
  6d105cc4
- [BugFix][AMD] Compatible patch for AITER lib after 04/20 (#17912) · 4f8b3732
  qli88 authored May 14, 2025
```
Signed-off-by: Qiang Li <qiang.li2@amd.com>
```
  4f8b3732
- [Fix] Support CUDAGraph capture for encoder-decoder on ROCm (#18104) · 176a95c6
  Luka Govedič authored May 13, 2025
```
Signed-off-by: Luka Govedič <lgovedic@redhat.com>
```
  176a95c6
- add kvint8 · 45273722
  xiabo authored May 14, 2025
  
  45273722
13 May, 2025 2 commits
- support cutlass prefix-cache · bdc70899
  zhuwenwen authored May 13, 2025
  
  bdc70899
- Implements dual-chunk-flash-attn backend for dual chunk attention with sparse... · 60f76243
  Tao He authored May 13, 2025
```
Implements dual-chunk-flash-attn backend for dual chunk attention with sparse attention support (#11844)
```
  60f76243
11 May, 2025 2 commits
- [FP8][ROCm][Attention] Enable FP8 KV cache on ROCm for V1 (#17870) · 06c0922a
  Gregory Shtrasberg authored May 11, 2025
```
Signed-off-by: Gregory Shtrasberg <Gregory.Shtrasberg@amd.com>
```
  06c0922a
- fix amd triton mla path (#17871) · eea22a56
  Shiyan Deng authored May 11, 2025
  
  eea22a56
09 May, 2025 5 commits
- Revert "[BugFix][AMD] Compatible patch for latest AITER(05/07/2025)" (#17910) · 85b72cb7
  Michael Goin authored May 09, 2025
  
  85b72cb7
- [BugFix][AMD] Compatible patch for latest AITER(05/07/2025) (#17864) · 9f64e934
  qli88 authored May 09, 2025
```
Signed-off-by: Qiang Li <qiang.li2@amd.com>
```
  9f64e934
- [Attention] MLA move rotary embedding to cuda-graph region (#17668) · 5e6f9394
  Lucas Wilkinson authored May 08, 2025
```
Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>
```
  5e6f9394
- [FEAT][ROCm]: Support AITER MLA on V1 Engine (#17523) · 3c9396a6
  vllmellm authored May 09, 2025
```
Signed-off-by: vllmellm <vllm.ellm@embeddedllm.com>
Co-authored-by: qli88 <qiang.li2@amd.com>
Co-authored-by: Hongxia Yang <62075498+hongxiayang@users.noreply.github.com>
```
  3c9396a6
- debug v0 zero overhead schedule · 4ff58b66
  lizhigong authored Apr 14, 2025
  
  4ff58b66