Commits · 9820d063c3442ae655ff02da5f8a053befc8b0b3 · OpenDAS / vllm_cscc · GitLab

30 Jul, 2025 1 commit
- 解决v1cudagraph的问题以及接入fuse moe marlin_V3版本 · 9820d063
  yangql authored Jul 30, 2025
  
  9820d063
26 Jul, 2025 1 commit
- 增加moe awq-marlin的支持 · c2bcb0ab
  yangql authored Jul 26, 2025
  
  c2bcb0ab
25 Jul, 2025 1 commit
- 修复block-wise的错误参数，以及增加dpsk-fp16-dtype的性能选择功能 · cb37537e
  yangql authored Jul 25, 2025
  
  cb37537e
14 Jul, 2025 1 commit
- update blockwise_int8.py · 5d35160b
  zhuwenwen authored Jul 14, 2025
  
  5d35160b
07 Jul, 2025 1 commit
- 修改激活量化算子实现方式 · b2fa85ce
  zhuwenwen authored Jul 07, 2025
  
  b2fa85ce
05 Jul, 2025 1 commit
- Fix unknown attribute of topk_indices_dtype in CompressedTensorsW8A8Fp8MoECutlassMethod (#20507) · 8aeaa910
  Lucia Fang authored Jul 05, 2025
```
Co-authored-by: Lucia (Lu) Fang <fanglu@meta.com>
```
  8aeaa910
04 Jul, 2025 4 commits
- [feat]: CUTLASS block scaled group gemm for SM100 (#19757) · 3d184b95
  Duncan Moss authored Jul 04, 2025
```
Signed-off-by: Duncan Moss <djm.moss@gmail.com>
Co-authored-by: Duncan Moss <dmoss@nvidia.com>
```
  3d184b95
- 增加稠密模型channelwise autotuning 接口 · 3e485650
  zhuwenwen authored Jul 04, 2025
  
  3e485650
- Support Llama 4 for fused_marlin_moe (#20457) · 0e3fe896
  Michael Goin authored Jul 04, 2025
```
Signed-off-by: mgoin <mgoin64@gmail.com>
```
  0e3fe896
- [Misc] Add SPDX-FileCopyrightText (#20428) · 1caca5a5
  Jee Jee Li authored Jul 04, 2025
```
Signed-off-by: Jee Jee Li <pandaleefree@gmail.com>
```
  1caca5a5
03 Jul, 2025 6 commits
- [Kernel] Enable fp8 support for pplx and BatchedTritonExperts. (#18864) · 78fe7753
  bnellnm authored Jul 03, 2025
```
Signed-off-by: Bill Nell <bnell@redhat.com>
```
  78fe7753
- [Quantization] Bump to use latest bitsandbytes (#20424) · 1819fbda
  Jee Jee Li authored Jul 03, 2025
```
Signed-off-by: Jee Jee Li <pandaleefree@gmail.com>
```
  1819fbda
- 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
- [Bugfix][CI/CD][CPU] Fix CPU CI tests (#20383) · 0ec3779d
  Li, Jiang authored Jul 03, 2025
```
Signed-off-by: jiang1.li <jiang1.li@intel.com>
```
  0ec3779d
- [Bugfix] Fix import of CutlassExpertsFp8 in compressed_tensors_moe.py (#20381) · 2e25bb12
  bnellnm authored Jul 02, 2025
```
Signed-off-by: Bill Nell <bnell@redhat.com>
```
  2e25bb12
02 Jul, 2025 2 commits

[Kernels] MoE refactor (#19636) · c1909e7e

bnellnm authored Jul 02, 2025


Signed-off-by: Bill Nell <bnell@redhat.com>
Signed-off-by: ElizaWszola <ewszola@redhat.com>
Co-authored-by: ElizaWszola <ewszola@redhat.com>

c1909e7e

Enable group size 64 for Machete (#20290) · 3abfe221
czhu-cohere authored Jul 01, 2025
```
Signed-off-by: czhu-cohere <conway.zhu@cohere.com>
```
3abfe221

01 Jul, 2025 4 commits
- [V1] [ROCm] Enable EP with AITER Fused MoE (#20270) · 02cabff2
  TJian authored Jul 01, 2025
```
Signed-off-by: tjtanaa <tunjian.tan@embeddedllm.com>
```
  02cabff2
- [Quant] [Bugfix] Fix quantization config matching with `hf_to_vllm_mapper` (#20046) · 9025a9a7
  Kyle Sayers authored Jul 01, 2025
  
  9025a9a7
- Enable ZP Support for Machete (#20268) · 9909726d
  czhu-cohere authored Jul 01, 2025
```
Signed-off-by: czhu-cohere <conway.zhu@cohere.com>
```
  9909726d
- [Feature] A calibration-free RTN-based quantization for accurate and... · 27949354
  Alex Kogan authored Jul 01, 2025
```
[Feature] A calibration-free RTN-based quantization for accurate and accelerated INT4/INT8 inference (#18768)
Signed-off-by: Alex Kogan <alex.kogan@oracle.com>
Co-authored-by: Michael Goin <mgoin64@gmail.com>
```
  27949354
30 Jun, 2025 2 commits
- [Bugfix] fix quark ptpc (#20251) · 1c50e100
  li haoyang authored Jun 30, 2025
```
Signed-off-by: Haoyang Li <Haoyang.Li@amd.com>
Co-authored-by: Haoyang Li <307790822@qq.com>
```
  1c50e100
- 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
29 Jun, 2025 1 commit
- [Quantization] Add compressed-tensors NVFP4 MoE Support (#19990) · 6f2f53a8
  Dipika Sikka authored Jun 30, 2025
```
Signed-off-by: Dipika Sikka <dipikasikka1@gmail.com>
Signed-off-by: Dipika <dipikasikka1@gmail.com>
```
  6f2f53a8
28 Jun, 2025 1 commit

[Refactor] Create a function util and cache the results for `has_deepgemm`,... · 4d366936

Wentao Ye authored Jun 28, 2025


[Refactor] Create a function util and cache the results for `has_deepgemm`, `has_deepep`, `has_pplx` (#20187)
Signed-off-by: yewentao256 <zhyanwentao@126.com>

4d366936

26 Jun, 2025 2 commits
- [Feature] Expert Parallelism Load Balancer (EPLB) (#18343) · e9fd658a
  Bowen Wang authored Jun 26, 2025
```
Signed-off-by: Bowen Wang <abmfy@icloud.com>
```
  e9fd658a
- [CPU] Fix torch version in x86 CPU backend (#19258) · 0567c824
  Li, Jiang authored Jun 26, 2025
```
Signed-off-by: jiang1.li <jiang1.li@intel.com>
```
  0567c824
25 Jun, 2025 3 commits
- [Quantization] Add compressed-tensors emulations support for NVFP4 (#19879) · 02c97d9a
  Dipika Sikka authored Jun 25, 2025
```
Signed-off-by: Dipika Sikka <dipikasikka1@gmail.com>
Signed-off-by: Dipika <dipikasikka1@gmail.com>
```
  02c97d9a
- [Refactor] Remove duplicate `ceil_div` (#20023) · 879f69be
  Wentao Ye authored Jun 25, 2025
```
Signed-off-by: yewentao256 <zhyanwentao@126.com>
```
  879f69be
- Revert "[Feature] Integrate new deepgemm (#19820)" (#20049) · a6c4b87f
  Wentao Ye authored Jun 24, 2025
```
Signed-off-by: yewentao256 <zhyanwentao@126.com>
```
  a6c4b87f
24 Jun, 2025 2 commits
- use .dev for version comparison with pytorch nightly release (#20031) · c01d1c5a
  Boyuan Feng authored Jun 24, 2025
```
Signed-off-by: Boyuan Feng <boyuan@meta.com>
```
  c01d1c5a
- [Feature] Integrate new deepgemm (#19820) · c6e3bba8
  Wentao Ye authored Jun 24, 2025
```
Signed-off-by: yewentao256 <zhyanwentao@126.com>
```
  c6e3bba8
23 Jun, 2025 1 commit
- Feat Dynamic Quantization for MoE Layers in GPTQ Marlin Backend (#19395) · dd2ccf8d
  Jun-Howie authored Jun 24, 2025
  
  dd2ccf8d
21 Jun, 2025 1 commit
- 添加moe smquant量化支持模块 · 7996b363
  zhuwenwen authored Jun 21, 2025
  
  7996b363
17 Jun, 2025 1 commit
- 修复int4选取config的bug · 44e87dde
  yangql authored Jun 17, 2025
  
  44e87dde
16 Jun, 2025 3 commits
- [Quantization] Remove FP4 emulation; Fall-back to marlin for device < 100 (#19563) · 6bc7b573
  Dipika Sikka authored Jun 16, 2025
  
  6bc7b573
- 分离fuse moe awq算子到lmslim上 · b3312eec
  yangql authored Jun 16, 2025
  
  b3312eec
- [Kernel] GGUF MMVQ kernel for multiple input vectors (#18754) · dec66d25
  Szymon Ożóg authored Jun 16, 2025
```
Signed-off-by: SzymonOzog <szymon.ozog@gmail.com>
```
  dec66d25
13 Jun, 2025 1 commit
- use base version for version comparison (#19587) · ce688ad4
  Boyuan Feng authored Jun 13, 2025
```
Signed-off-by: Boyuan Feng <boyuan@meta.com>
```
  ce688ad4