Commits · 1a9a61d76c00b52b9925818cb4e06e5cfc32805e · OpenDAS / vllm_cscc · GitLab

14 Jul, 2025 1 commit
- update blockwise_int8.py · 5d35160b
  zhuwenwen authored Jul 14, 2025
  
  5d35160b
10 Jul, 2025 1 commit
- add apex rmsnorm · 3e0a595d
  zhuwenwen authored Jul 10, 2025
  
  3e0a595d
07 Jul, 2025 1 commit
- 修改激活量化算子实现方式 · b2fa85ce
  zhuwenwen authored Jul 07, 2025
  
  b2fa85ce
04 Jul, 2025 1 commit
- 增加稠密模型channelwise autotuning 接口 · 3e485650
  zhuwenwen authored Jul 04, 2025
  
  3e485650
03 Jul, 2025 2 commits
- 修改美团 deepseek channel-wise模型 moe config获取 · bf3bf955
  zhuwenwen authored Jul 03, 2025
```
去除AutoTuning info提示信息
```
  bf3bf955
- 修改channelwise moe相关配置读取 · acf1b6c6
  zhuwenwen authored Jul 03, 2025
  
  acf1b6c6
30 Jun, 2025 1 commit
- 修复channelwise w8a8 预热没有区分卡造成的triton kernel崩溃问题，另外增加block-int8的moe 拿到config步骤 · c0c0eb69
  gaoqiong authored Jun 30, 2025
  
  c0c0eb69
23 Jun, 2025 1 commit
- fix moe run error · 10184690
  zhuwenwen authored Jun 23, 2025
  
  10184690
21 Jun, 2025 2 commits
- 添加moe smquant量化支持模块 · 7996b363
  zhuwenwen authored Jun 21, 2025
  
  7996b363
- skip is_rocm_aiter_moe_enabled and remove qwen unused int8 code · bbd14169
  zhuwenwen authored Jun 21, 2025
  
  bbd14169
18 Jun, 2025 1 commit
- 暂时去掉fused_moe中量化kernel的bug修复 · 35dbdc41
  王敏 authored Jun 18, 2025
  
  35dbdc41
17 Jun, 2025 2 commits
- [fix]1.修复medusa、scorer等并行解码单测；2.修复moe... · 8db76782
  王敏 authored Jun 17, 2025
```
[fix]1.修复medusa、scorer等并行解码单测；2.修复moe kernel单测问题，优化代码；3.修复rejection_sampler中test_compare_nonflashinfer_backend单测问题
```
  8db76782
- 修复int4选取config的bug · 44e87dde
  yangql authored Jun 17, 2025
  
  44e87dde
16 Jun, 2025 1 commit
- 分离fuse moe awq算子到lmslim上 · b3312eec
  yangql authored Jun 16, 2025
  
  b3312eec
13 Jun, 2025 2 commits
- 增加fused_experts_impl_int8的接入 · 483acdc4
  xuxz authored Jun 13, 2025
  
  483acdc4
- 解耦w8a8 的gemm以及moe算子相关实现 · 942368c7
  gaoqiong authored Jun 13, 2025
  
  942368c7
12 Jun, 2025 3 commits
- add bw qwen3 moe configs · 2f6f5bb3
  zhuwenwen authored Jun 12, 2025
  
  2f6f5bb3
- 修复qwen3-moe的awq配置导致fp16加载错误，修复dpsk-moe手写算子首字耗时增加问题 · 99ffef47
  zhuwenwen authored Jun 12, 2025
  
  99ffef47
- Update int8_utils.py · 550a1e5e
  zhuwenwen authored Jun 12, 2025
  
  550a1e5e
10 Jun, 2025 1 commit

[Misc] Slight improvement of the BNB (#19418) · b6553be1

Jee Jee Li authored Jun 10, 2025

Signed-off-by: Jee Jee Li <pandaleefree@gmail.com>
Co-authored-by: Isotr0py <2037008807@qq.com>
Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>

b6553be1

09 Jun, 2025 1 commit
- [Misc] Fixes and Optimizations for DeepEP + DeepGEMM combination. (#19298) · 5cf2daea
  Varun Sundar Rabindranath authored Jun 09, 2025
```
Signed-off-by: Varun <vsundarr@redhat.com>
Co-authored-by: Varun <vsundarr@redhat.com>
```
  5cf2daea
08 Jun, 2025 2 commits
- [Quantization] Add compressed-tensors NVFP4 support (#18312) · c123bc33
  Dipika Sikka authored Jun 08, 2025
  
  c123bc33
- Add H20-3e fused MoE kernel tuning configs for Qwen3-235B-A22B (#19315) · 989dcee9
  Xu Wenqing authored Jun 08, 2025
```
Signed-off-by: Xu Wenqing <xuwq1993@qq.com>
```
  989dcee9
07 Jun, 2025 2 commits
- 新增fusemoe手写算子的支持，需要group-gemm包 · e7a963f5
  yangql authored Jun 07, 2025
  
  e7a963f5
- [Kernel] Integrate CUTLASS MoE kernel with PPLX (#18762) · 84166fee
  ElizaWszola authored Jun 07, 2025
```
Signed-off-by: ElizaWszola <ewszola@redhat.com>
Signed-off-by: Tyler Michael Smith <tyler@neuralmagic.com>
Co-authored-by: Tyler Michael Smith <tyler@neuralmagic.com>
```
  84166fee
06 Jun, 2025 3 commits
- [Quantization] Bump compressed-tensors version; update NVFP4A16 test model (#19224) · 94870359
  Dipika Sikka authored Jun 06, 2025
```
Signed-off-by: Dipika Sikka <dipikasikka1@gmail.com>
```
  94870359
- [Model] Add VLLM_USE_NN to use nn layout · 6880bf15
  zhuwenwen authored Jun 06, 2025
  
  6880bf15
- Fix AOPerModuleConfig name changes (#18869) · c8134bea
  Jerry Zhang authored Jun 05, 2025
```
Signed-off-by: Jerry Zhang <jerryzh168@gmail.com>
```
  c8134bea
05 Jun, 2025 5 commits
- [Hardware][NVIDIA] FP4 MoE kernel optimization (#19110) · 61059bee
  Chiyue Wei authored Jun 05, 2025
```
Signed-off-by: Chiyue Wei <chiyuew@nvidia.com>
Co-authored-by: Chiyue Wei <chiyuew@nvidia.com>
```
  61059bee
- Add H20-3e fused MoE kernel tuning configs for DeepSeek-R1/V3 (#19205) · ec89524f
  Xu Wenqing authored Jun 06, 2025
  
  ec89524f
- 新增qwen3 fusemoe优化支持 · 1cb37dab
  yangql authored Jun 05, 2025
  
  1cb37dab
- Update w8a8_int8.py · 53250530
  gaoqiong authored Jun 05, 2025
  
  53250530
- 修改deepseelk block-int8 权重处理流程，增加per-channel bestconfig配置以及首次triton warmup代码 · 40b94473
  gaoqiong authored Jun 05, 2025
  
  40b94473
04 Jun, 2025 4 commits
- [Kernel] Integrate batched/masked deepgemm kernel (#19111) · c3fd4d66
  Varun Sundar Rabindranath authored Jun 04, 2025
```
Signed-off-by: Varun <vsundarr@redhat.com>
Co-authored-by: Varun <vsundarr@redhat.com>
```
  c3fd4d66
- Sm100 blockwise fp8 swap ab (#18564) · 5f2cd251
  Lain authored Jun 04, 2025
  
  5f2cd251
- [feat]add 120cu nn fused_moe configs of qwen3, mixtral8*7b and mixtral8*22b · d36deb1a
  zhuwenwen authored Jun 04, 2025
  
  d36deb1a
- [KERNEL] Sampler. CUDA kernel for applying repetition penalty (#18437) · 5d6d1adf
  Vadim Gimpelson authored Jun 04, 2025
  
  5d6d1adf
03 Jun, 2025 3 commits
- [Kernel] DeepEP dispatch-combine kernel integration (#18434) · fa98d777
  Varun Sundar Rabindranath authored Jun 03, 2025
```
Signed-off-by: Varun <vsundarr@redhat.com>
Co-authored-by: Varun Sundar Rabindranath <vsundarr@redhat.com>
```
  fa98d777
- [Misc] Add SPDX-FileCopyrightText (#19100) · 02f0c7b2
  Simon Mo authored Jun 03, 2025
```
Signed-off-by: simon-mo <simon.mo@hey.com>
```
  02f0c7b2
- [Bugfix][EP+DP] Use pplx-kernel internode instead of intranode (#19034) · 8a57872b
  Tyler Michael Smith authored Jun 02, 2025
```
Signed-off-by: Tyler Michael Smith <tysmith@redhat.com>
Signed-off-by: Tyler Michael Smith <tyler@neuralmagic.com>
```
  8a57872b