Commits · 347fc09c63b3080eaecc00cf95e9c7e4c0fcdfe6 · OpenDAS / vllm_cscc

17 Dec, 2025 3 commits
- maintain consistency between k_cache type and q · 3e191138
  zhuwenwen authored Dec 17, 2025
  
  3e191138
- convert q to float8_e4m3fn · da85feb7
  zhuwenwen authored Dec 17, 2025
  
  da85feb7
- remove fuse_rmsnorm_rope_quant_gfx938 · 99981972
  zhuwenwen authored Dec 17, 2025
  
  99981972
16 Dec, 2025 2 commits
- add fuse_rmsnorm_rope_quant_gfx938 to support use fp8_e4m3 mla · 0ce3b670
  zhuwenwen authored Dec 16, 2025
  
  0ce3b670
- add VLLM_USE_FLASH_MLA_FP8 to use mla fp8 · a9f57e73
  zhuwenwen authored Dec 16, 2025
```
set VLLM_USE_LIGHTOP_RMS_ROPE_CONCAT=1
```
  a9f57e73
31 Oct, 2025 1 commit
- update moe configs name · 96d4afdc
  zhuwenwen authored Oct 31, 2025
  
  96d4afdc
16 Oct, 2025 1 commit
- fix v0 eager fa-pa acc error · 0adf9cda
  zhuwenwen authored Oct 16, 2025
  
  0adf9cda
13 Oct, 2025 1 commit
- 新增fp8—e5m2 · ed3cdc81
  zhuwenwen authored Oct 13, 2025
  
  ed3cdc81
23 Sep, 2025 1 commit
- 修复kvcachefp8与cp/pc的冲突 · 8b7daa0d
  yangql authored Sep 23, 2025
  
  8b7daa0d
11 Sep, 2025 1 commit
- update fa interface · bf3d75f4
  zhuwenwen authored Sep 11, 2025
  
  bf3d75f4
10 Sep, 2025 1 commit
- update the conditions for pad_v on v0 · d6dc122f
  zhuwenwen authored Sep 10, 2025
  
  d6dc122f
01 Sep, 2025 1 commit
- add dca and sparse attention support on rocm · 12112a4d
  zhuwenwen authored Sep 01, 2025
  
  12112a4d
28 Aug, 2025 1 commit
- 优化 block_tables 的生成逻辑，增加对混合情况的检测，确保在存在空和非空块时正确计算最大块长度。 · 4932783d
  laibao authored Aug 28, 2025
  
  4932783d
15 Aug, 2025 1 commit
- 增加marlin对cache13的支持，以及新增flash mla的kvcache fp8的支持 · 8f25283a
  yangql authored Aug 15, 2025
  
  8f25283a
17 Jul, 2025 1 commit
- add VLLM_USE_PA_PRINT_PARAM to print fa-pa size · 1092a467
  zhuwenwen authored Jul 17, 2025
  
  1092a467
10 Jul, 2025 1 commit
- update bw vdim128 · d805c59c
  zhuwenwen authored Jul 10, 2025
  
  d805c59c
09 Jul, 2025 1 commit
- [fix]修复零消耗引入的cudagraph模式数据准备cpu耗时问题 · cd6239dc
  zhuwenwen authored Jul 09, 2025
  
  cd6239dc
06 Jul, 2025 2 commits
- Revert "[V0 deprecation] Remove V0 CPU/XPU/TPU backends (#20412)" · a5dd03c1
  simon-mo authored Jul 06, 2025
```
This reverts commit e202dd27.
```
  a5dd03c1
- [V0 deprecation] Remove V0 CPU/XPU/TPU backends (#20412) · e202dd27
  Woosuk Kwon authored Jul 06, 2025
```
Signed-off-by: Woosuk Kwon <woosuk.kwon@berkeley.edu>
Signed-off-by: jiang1.li <jiang1.li@intel.com>
Co-authored-by: Li, Jiang <jiang1.li@intel.com>
```
  e202dd27
03 Jul, 2025 1 commit
- update VLLM_USE_FLASH_ATTN_BACKEND to VLLM_USE_FLASH_ATTN_PA · e036115e
  zhuwenwen authored Jul 03, 2025
  
  e036115e
27 Jun, 2025 1 commit
- update cutlass fa and pa · 4ddb5447
  zhuwenwen authored Jun 27, 2025
  
  4ddb5447
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
20 Jun, 2025 1 commit
- [Kernel] mark TorchSDPABackend swap_blocks NotImplementedError (#19749) · 71baf85a
  Ning Xie authored Jun 21, 2025
  
  71baf85a
17 Jun, 2025 1 commit
- [V1][Kernel] Flashinfer HND KV cache layout (#19280) · 4c8f64fa
  Nicolò Lucchesi authored Jun 17, 2025
```
Signed-off-by: NickLucche <nlucches@redhat.com>
```
  4c8f64fa
15 Jun, 2025 1 commit
- [Kernel] Raise verbose error and consolidate `num_heads/num_kv_heads` divisibility check (#19339) · 0b73736a
  22quinn authored Jun 14, 2025
```
Signed-off-by: 22quinn <33176974+22quinn@users.noreply.github.com>
```
  0b73736a
13 Jun, 2025 1 commit
- update triton_mla.py · e046b382
  zhuwenwen authored Jun 13, 2025
  
  e046b382
12 Jun, 2025 2 commits
- [torch.compile][ROCm] Fuse quantization onto attention using a torch.compile pass (#16756) · f98548b9
  Luka Govedič authored Jun 12, 2025
```
Signed-off-by: Luka Govedič <lgovedic@redhat.com>
Co-authored-by: Sage Moore <sage@neuralmagic.com>
```
  f98548b9
- [CI] change spell checker from codespell to typos (#18711) · 2f1c19b2
  Ning Xie authored Jun 12, 2025
```
Signed-off-by: Andy Xie <andy.xning@gmail.com>
```
  2f1c19b2
11 Jun, 2025 1 commit

[AMD] [Quantization] Add override flag for attention dtype instead of using... · c7ea0b56

rasmith authored Jun 11, 2025


[AMD] [Quantization] Add override flag for attention dtype instead of using kv_cache_dtype trigger (#17331)
Signed-off-by: Randall Smith <Randall.Smith@amd.com>

c7ea0b56

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 2 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
- [test]fix basic_correctness and benchmarks · dbd62f84
  zhuwenwen authored May 29, 2025
  
  dbd62f84
28 May, 2025 2 commits
- 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