Commits · 706e7a4f1042b5403a4ac32aa4b56d678afa0f8e · OpenDAS / vllm_cscc

10 Feb, 2026 1 commit
- for qwen3 dense,set VLLM_USE_OPT_RESHAPE_AND_CACHE=1 · 706e7a4f
  zhuwenwen authored Feb 10, 2026
  
  706e7a4f
21 Jan, 2026 1 commit
- update VLLM_USE_FUSED_RMS_ROPE=0 (default) · 80a6b121
  zhuwenwen authored Jan 21, 2026
```
for qwen3, VLLM_USE_FUSED_RMS_ROPE=1 (default)
```
  80a6b121
20 Jan, 2026 1 commit
- update pt_weights_iterator · 145b4eac
  zhuwenwen authored Jan 20, 2026
  
  145b4eac
14 Jan, 2026 1 commit
- fix tests of kernels · 0e607f8e
  zhuwenwen authored Jan 14, 2026
```
set VLLM_USE_PD_SPLIT=1
update moe_align_block_size
```
  0e607f8e
24 Dec, 2025 1 commit
- 在v0.11中暂不支持awq的scale和zero合并操作 · 0ba1219a
  yangql authored Dec 24, 2025
  
  0ba1219a
22 Dec, 2025 2 commits
- fix weights_not_loaded · 10349d37
  zhuwenwen authored Dec 22, 2025
  
  10349d37
- update weights_not_loaded and flash_mla_with_kvcache · a3f4b5b8
  zhuwenwen authored Dec 22, 2025
  
  a3f4b5b8
03 Dec, 2025 2 commits
- set VLLM_USE_LIGHTOP_MOE_ALIGN=0 (for qwen3-30b) · 87503d85
  zhuwenwen authored Dec 03, 2025
  
  87503d85
- add VLLM_USE_OPT_RESHAPE_AND_CACHE、VLLM_USE_FUSE_SILU_AND_MUL and... · 15a55773
  zhuwenwen authored Dec 03, 2025
```
add VLLM_USE_OPT_RESHAPE_AND_CACHE、VLLM_USE_FUSE_SILU_AND_MUL and VLLM_USE_TOPK_RENORM for qwen3-30b
```
  15a55773
02 Dec, 2025 1 commit
- add VLLM_USE_V32_ENCODE to use encoding_dsv32.py · ba7bcccd
  zhuwenwen authored Dec 02, 2025
  
  ba7bcccd
20 Nov, 2025 1 commit
- 修复pd分离开cp引起的校检问题 · acf9f945
  zhuwenwen authored Nov 20, 2025
```
update VLLM_USE_PD_SPLIT=0 (for dspk)and  VLLM_USE_PD_SPLIT=1 (for others)
```
  acf9f945
13 Nov, 2025 2 commits

add VLLM_USE_OPT_ZEROS to replace triton_ (torch.zeros) · e7f2785f
zhuwenwen authored Nov 13, 2025
```
set default_max_num_batched_tokens = 10240
update qwen3_moe of layernorm
off lightop of moe_fused_gate
```
e7f2785f

set VLLM_USE_LIGHTOP=0 for dpsk-v3 · 6a8fd297

zhuwenwen authored Nov 13, 2025

add VLLM_USE_PD_SPLIT to split prefill and decode
replace triton_ of rms and act_and_mul

6a8fd297

26 Sep, 2025 1 commit
- update VLLM_USE_LIGHTOP and VLLM_USE_OPT_CAT · f8cf43ae
  zhuwenwen authored Sep 26, 2025
  
  f8cf43ae
25 Sep, 2025 2 commits
- [Optimization] Use a cheaper cache key in `get_model_architecture` (#25682) · 89fa54e6
  Cyrus Leung authored Sep 26, 2025
```
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
```
  89fa54e6
- update VLLM_USE_LIGHTOP and VLLM_USE_OPT_CAT · f70d0ec6
  zhuwenwen authored Sep 25, 2025
  
  f70d0ec6
24 Sep, 2025 1 commit
- [Docs] Enable `fail_on_warning` for the docs build in CI (#25580) · 8c853050
  Harry Mellor authored Sep 24, 2025
```
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
```
  8c853050
22 Sep, 2025 1 commit
- [TPU] Deprecate `xm.mark_step` in favor of ``torch_xla.sync` (#25254) · 4cf71cc8
  Nicolò Lucchesi authored Sep 22, 2025
```
Signed-off-by: NickLucche <nlucches@redhat.com>
Co-authored-by: Ye (Charlotte) Qi <yeq@meta.com>
```
  4cf71cc8
21 Sep, 2025 1 commit

[V0 Deprecation] Remove LLMEngine (#25033) · 52c2a8d4

Woosuk Kwon authored Sep 20, 2025


Signed-off-by: Woosuk Kwon <woosuk@thinkingmachines.ai>
Signed-off-by: Woosuk Kwon <woosuk.kwon@berkeley.edu>

52c2a8d4

20 Sep, 2025 2 commits

Generate _ModelInfo properties file when loading to improve loading speed (#23558) · bf8b26ca

Manoel Marques authored Sep 20, 2025

Signed-off-by: Manoel Marques <manoel.marques@ibm.com>
Signed-off-by: Manoel Marques <manoelmrqs@gmail.com>
Co-authored-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
Co-authored-by: Luka Govedič <ProExpertProg@users.noreply.github.com>

bf8b26ca

[Optimization] Avoid repeated model architecture conversion for pooling models (#25261) · c60e6137
Cyrus Leung authored Sep 20, 2025
```
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
```
c60e6137

19 Sep, 2025 1 commit
- Move `ModelConfig` from `config/__init__.py` to `config/model.py` (#25252) · aed16879
  Harry Mellor authored Sep 19, 2025
```
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
```
  aed16879
13 Sep, 2025 1 commit
- Enable conversion of multimodal models to pooling tasks (#24451) · bbb70036
  Maximilien de Bayser authored Sep 12, 2025
```
Signed-off-by: Max de Bayser <mbayser@br.ibm.com>
```
  bbb70036
12 Sep, 2025 3 commits
- [Docs] Fix warnings in mkdocs build (continued) (#24740) · 41f17bf2
  Hyogeun Oh (오효근) authored Sep 12, 2025
```
Signed-off-by: Zerohertz <ohg3417@gmail.com>
```
  41f17bf2
- [Bugfix] Fix BNB name match (#24735) · 60a09519
  Jee Jee Li authored Sep 12, 2025
```
Signed-off-by: Jee Jee Li <pandaleefree@gmail.com>
```
  60a09519
- Enable conversion of multimodal models to pooling tasks (#24451) · e090b7b4
  Maximilien de Bayser authored Sep 12, 2025
```
Signed-off-by: Max de Bayser <mbayser@br.ibm.com>
```
  e090b7b4
11 Sep, 2025 2 commits

Fix typing for `safetensors_load_strategy` (#24641) · d6249d06
Harry Mellor authored Sep 11, 2025
```
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
```
d6249d06

[Core] feat: Add --safetensors-load-strategy flag for faster safetensors... · 41329a0f

shengshiqi-google authored Sep 11, 2025


[Core] feat: Add --safetensors-load-strategy flag for faster safetensors loading from Lustre (#24469)
Signed-off-by: Shiqi Sheng <shengshiqi@google.com>
Signed-off-by: shengshiqi-google <160179165+shengshiqi-google@users.noreply.github.com>
Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>

41329a0f

10 Sep, 2025 2 commits

Move `LoadConfig` from `config/__init__.py` to `config/load.py` (#24566) · f36355ab
Harry Mellor authored Sep 10, 2025
```
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
```
f36355ab

[Bugfix] Update Run:AI Model Streamer Loading Integration (#23845) · 4377b1ae

pwschuurman authored Sep 09, 2025


Signed-off-by: Omer Dayan (SW-GPU) <omer@run.ai>
Signed-off-by: Peter Schuurman <psch@google.com>
Co-authored-by: Omer Dayan (SW-GPU) <omer@run.ai>
Co-authored-by: Cyrus Leung <tlleungac@connect.ust.hk>