Commits · f1bc989074efe707d858a16f055a693ccbf75726 · OpenDAS / vllm_cscc

16 Jan, 2026 1 commit

解决custom cudagraph模式需要拷贝的问题，需要配合dtk进行使用。 · f1bc9890

zhuwenwen authored Jan 16, 2026

区分pcie和hglink custom allreduce的使用
vllm：export VLLM_CUSTOM_CACHE=1
dtk：export HIP_KERNEL_EVENT_SYSTENFENCE=1

set VLLM_USE_FUSED_RMS_ROPE=1
add SUPPORT_MOE_MARLIN_W16A16 to use moe marlin on bw
support fa kvcache fp8 (todo: add VLLM_USE_QUERY_QUANT to not use q quant)
update moe_align_block_size

f1bc9890

07 Jan, 2026 3 commits
- fix weights_not_loaded · a1eba087
  zhuwenwen authored Jan 07, 2026
  
  a1eba087
- update weights_not_loaded · e74c54bc
  zhuwenwen authored Jan 07, 2026
  
  e74c54bc
- remove USE_FUSED_RMS_QUANT and USE_FUSED_SILU_MUL_QUANT · 60b37c6b
  zhuwenwen authored Jan 07, 2026
  
  60b37c6b
06 Jan, 2026 2 commits
- 暂不支持awq的scale和zero合并操作 · 2907adfa
  zhuwenwen authored Jan 06, 2026
  
  2907adfa
- fix weights_not_loaded · 451af742
  zhuwenwen authored Jan 06, 2026
```
update weights_not_loaded and flash_mla_with_kvcache
update paged_mqa_logits
```
  451af742
09 Dec, 2025 2 commits

[Model][Quantization] Restore MoE + GGUF models support (incl. Qwen3 MoE) by... · 58d5b3f5

Tsukasa OI authored Dec 09, 2025


[Model][Quantization] Restore MoE + GGUF models support (incl. Qwen3 MoE) by allowing Sideload Parameters (#30116)
Signed-off-by: Tsukasa OI <floss_llm@irq.a4lg.com>
Co-authored-by: Isotr0py <mozf@mail2.sysu.edu.cn>

58d5b3f5

update torchao safetensors impl (#30155) · 4b03b502
liangel-02 authored Dec 08, 2025
```
Signed-off-by: Angel Li <liangel@meta.com>
```
4b03b502

08 Dec, 2025 1 commit

[Model][7/N] Improve all pooling task | Deprecation as_reward_model. Extract... · 9e77ffca

wang.yuqi authored Dec 08, 2025


[Model][7/N] Improve all pooling task | Deprecation as_reward_model. Extract hidden states prefer using new multi-vector retrieval API (#26686)
Signed-off-by: wang.yuqi <yuqi.wang@daocloud.io>

9e77ffca

03 Dec, 2025 3 commits
- [Bugfix][Quantization] Support BF16 tensors on GGUF (#29948) · 42c19496
  Tsukasa OI authored Dec 03, 2025
```
Signed-off-by: Tsukasa OI <floss_llm@irq.a4lg.com>
```
  42c19496
- 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
28 Nov, 2025 1 commit
- [Chore]: Reorganize model repo operating functions in `transformers_utils` (#29680) · f946a8d7
  Isotr0py authored Nov 29, 2025
```
Signed-off-by: Isotr0py <mozf@mail2.sysu.edu.cn>
```
  f946a8d7
26 Nov, 2025 1 commit
- [Attention] Remove imports from `vllm/attention/__init__.py` (#29342) · 430dd4d9
  Matthew Bonanni authored Nov 26, 2025
```
Signed-off-by: Matthew Bonanni <mbonanni@redhat.com>
```
  430dd4d9
25 Nov, 2025 1 commit

[Feature]: Improve GGUF loading from HuggingFace user experience like repo_id:quant_type (#29137) · 794029f0

Injae Ryou authored Nov 25, 2025


Signed-off-by: Injae Ryou <injaeryou@gmail.com>
Signed-off-by: Isotr0py <mozf@mail2.sysu.edu.cn>
Co-authored-by: Isotr0py <mozf@mail2.sysu.edu.cn>

794029f0

22 Nov, 2025 2 commits

[Misc] Fix pre-commit (#29238) · ae668183
Cyrus Leung authored Nov 22, 2025
```
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
```
ae668183

Fix: Resolve circular import in model_loader/utils.py (#29189) · 6965a392

Nandan Vallamdasu authored Nov 22, 2025

Signed-off-by: nandan2003 <nandan.vallamdasu@outlook.com>
Signed-off-by: Nandan Vallamdasu <nandan.vallamdasu@outlook.com>
Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>
Co-authored-by: Cyrus Leung <tlleungac@connect.ust.hk>

6965a392

21 Nov, 2025 2 commits

Default model load/config/tokenizer to `mistral` format if relevant files exist (#28659) · 57430fc9

Julien Denize authored Nov 21, 2025


Signed-off-by: Julien Denize <julien.denize@mistral.ai>
Signed-off-by: Julien Denize <40604584+juliendenize@users.noreply.github.com>
Signed-off-by: mgoin <mgoin64@gmail.com>
Signed-off-by: Michael Goin <mgoin64@gmail.com>
Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>
Co-authored-by: mgoin <mgoin64@gmail.com>

57430fc9

[log] add weights loading time log to sharded_state loader (#28628) · 53a1ba6e
Ning Xie authored Nov 22, 2025
```
Signed-off-by: Andy Xie <andy.xning@gmail.com>
```
53a1ba6e

20 Nov, 2025 2 commits
- 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
- [torchao] fix safetensors for sharding (#28169) · 1d642872
  liangel-02 authored Nov 19, 2025
```
Signed-off-by: Angel Li <liangel@meta.com>
```
  1d642872
19 Nov, 2025 1 commit
- Move online quantization to `model.load_weights` (#26327) · da94c7c0
  Jerry Zhang authored Nov 18, 2025
```
Signed-off-by: Jerry Zhang <jerryzh168@gmail.com>
```
  da94c7c0
18 Nov, 2025 1 commit

[Model] Add Gemma3 GGUF multimodal support (#27772) · c2612371

Luciano Martins authored Nov 18, 2025

Signed-off-by: Luciano Martins <lucianommartins@users.noreply.github.com>
Signed-off-by: Isotr0py <mozf@mail2.sysu.edu.cn>
Co-authored-by: Luciano Martins <lucianommartins@users.noreply.github.com>
Co-authored-by: Isotr0py <mozf@mail2.sysu.edu.cn>

c2612371

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

12 Nov, 2025 1 commit
- Skip models that cannot currently init on Transformers v5 (#28471) · 51c599f0
  Harry Mellor authored Nov 12, 2025
```
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
```
  51c599f0
05 Nov, 2025 1 commit
- [V0 deprecation] Remove VLLM_USE_V1 usage in most modules (#27955) · 428bc7bf
  wangxiyuan authored Nov 05, 2025
```
Signed-off-by: wangxiyuan <wangxiyuan1007@gmail.com>
```
  428bc7bf
30 Oct, 2025 1 commit

[Feat] Adds runai distributed streamer (#27230) · 17d055f5

Benjamin Bartels authored Oct 30, 2025


Signed-off-by: bbartels <benjamin@bartels.dev>
Signed-off-by: Benjamin Bartels <benjamin@bartels.dev>
Co-authored-by: omer-dayan <omdayan@nvidia.com>
Co-authored-by: Cyrus Leung <tlleungac@connect.ust.hk>

17d055f5

26 Oct, 2025 1 commit

[Chore]:Extract math and argparse utilities to separate modules (#27188) · 71b1c8b6

Yeshwanth N authored Oct 26, 2025


Signed-off-by: Yeshwanth Surya <yeshsurya@gmail.com>
Signed-off-by: Yeshwanth N <yeshsurya@gmail.com>
Signed-off-by: yeshsurya <yeshsurya@gmail.com>

71b1c8b6

24 Oct, 2025 1 commit
- [Log] Optimize Startup Log (#26740) · 52efc34e
  Wentao Ye authored Oct 24, 2025
```
Signed-off-by: yewentao256 <zhyanwentao@126.com>
```
  52efc34e
23 Oct, 2025 1 commit
- [Chore] Separate out `vllm.utils.platform_utils.py` (#27374) · ca76486a
  Jonathan Chen authored Oct 23, 2025
```
Signed-off-by: Jonathan <chenleejonathan@gmail.com>
```
  ca76486a
18 Oct, 2025 1 commit
- [Chore] Clean up pytorch helper functions in `vllm.utils` (#26908) · 6ac5e06f
  Isotr0py authored Oct 19, 2025
```
Signed-off-by: Isotr0py <mozf@mail2.sysu.edu.cn>
Signed-off-by: isotr0py <2037008807@qq.com>
```
  6ac5e06f
17 Oct, 2025 1 commit
- [Chore] Separate out `vllm.utils.importlib` (#27022) · 4d4d6bad
  Cyrus Leung authored Oct 17, 2025
```
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
```
  4d4d6bad
15 Oct, 2025 1 commit
- [Misc] rename torch_dtype to dtype (#26695) · 8f4b313c
  wangxiyuan authored Oct 15, 2025
```
Signed-off-by: wangxiyuan <wangxiyuan1007@gmail.com>
```
  8f4b313c
12 Oct, 2025 1 commit
- Update `Optional[x]` -> `x | None` and `Union[x, y]` to `x | y` (#26633) · 8fcaaf6a
  Harry Mellor authored Oct 12, 2025
```
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
```
  8fcaaf6a
10 Oct, 2025 1 commit
- Fix some typing issues found by `mypy==1.18.2` (#26596) · 7c12763b
  Harry Mellor authored Oct 10, 2025
```
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
```
  7c12763b
09 Oct, 2025 3 commits

[V0 deprecation] Remove `QKVCrossParallelLinear` implementation (#26475) · d1ddf340
Isotr0py authored Oct 09, 2025
```
Signed-off-by: Isotr0py <mozf@mail2.sysu.edu.cn>
```
d1ddf340

[Model] Gemma3: Fix GGUF loading and quantization (#26189) · 1317028a

Luciano Martins authored Oct 09, 2025

1317028a

Separate MLAAttention class from Attention (#25103) · e614ab78

Naveenraj Kamalakannan authored Oct 08, 2025

Signed-off-by: Naveenraj Kamalakannan <therealnaveenkamal@gmail.com>
Signed-off-by: Luka Govedič <ProExpertProg@users.noreply.github.com>
Co-authored-by: Luka Govedič <ProExpertProg@users.noreply.github.com>

e614ab78