Commits · aae1cc3972f388371ebbc52a293a0f73c712b613 · OpenDAS / vllm_cscc

05 Jan, 2026 1 commit
- 删除调试信息 · c773cc66
  yangql authored Jan 05, 2026
  
  c773cc66
23 Dec, 2025 2 commits
- [feat]低延迟模式采用int8 dispatch · ce41e45b
  王敏 authored Dec 23, 2025
  
  ce41e45b
- 取出deepep的部分调试信息 · df03e33b
  yangql authored Dec 23, 2025
  
  df03e33b
22 Dec, 2025 2 commits
- [fix]1.解决ep sequence parallel优化引入的mtp报错;2.解决共享专家无法和combine overlap问题 · dc01fce4
  王敏 authored Dec 22, 2025
  
  dc01fce4
- 修复ep的auto模式的崩溃bug · 29523973
  yangql authored Dec 22, 2025
  
  29523973
21 Dec, 2025 1 commit
- 修复auto模式乱码的问题 · 094f1299
  yangql authored Dec 21, 2025
  
  094f1299
18 Dec, 2025 1 commit
- [feat]优化dp attention，减少1次allgather耗时，高吞吐提升明显 · 3e386c3b
  王敏 authored Dec 18, 2025
  
  3e386c3b
17 Dec, 2025 2 commits
- [feat]w8a8 高吞吐模式先量化再dispatch · c03a553b
  王敏 authored Dec 17, 2025
  
  c03a553b
- [feat]修复低延迟错误 · 4fadef92
  王敏 authored Dec 17, 2025
  
  4fadef92
16 Dec, 2025 1 commit
- up auto deepep · 0d3ae2fc
  yangql authored Dec 16, 2025
  
  0d3ae2fc
15 Dec, 2025 3 commits
- [feat]优化deepep高吞吐模式 · 10400c58
  王敏 authored Dec 15, 2025
  
  10400c58
- [feat]优化高吞吐模式num_sms · a983ea53
  王敏 authored Dec 15, 2025
  
  a983ea53
- [feat]1.支持高吞吐模式ep_scatter+deepgemm contiguous+ep_gather方案；2.支持高吞吐模式下ETP,例如dp4 tp4 · 3833018c
  王敏 authored Dec 15, 2025
  
  3833018c
10 Dec, 2025 1 commit
- [fix]修复deepep 高吞吐模式vmfault问题 · 916b5876
  王敏 authored Dec 10, 2025
  
  916b5876
08 Dec, 2025 2 commits
- 删除无效代码 · 1a315a58
  王敏 authored Dec 08, 2025
  
  1a315a58
- [feat]支持deepep ETP，dp4 tp4 ep16相比dp32 tp1 ep32提升明显 · 6cabbf16
  王敏 authored Dec 08, 2025
  
  6cabbf16
02 Dec, 2025 1 commit
- [feat]支持deepep低延迟与共享专家overlap · 1ae8f58c
  王敏 authored Dec 02, 2025
  
  1ae8f58c
15 Nov, 2025 1 commit
- [feat]1.w8a8 marlin适配deepep低延迟;2.非naive ep模式，去掉多余的dp padding,避免allreduce耗时 · 4a943d35
  王敏 authored Nov 15, 2025
  
  4a943d35
13 Nov, 2025 4 commits
- [fix]解决moe_fused_gate编译错误，去掉mla中mtp部分的修改 · b956fc64
  zhuwenwen authored Nov 13, 2025
```
restore the default settings of disable_cascade_attn
add VLLM_USE_OPT_ZEROS to replace triton_ (torch.zeros)
set default_max_num_batched_tokens = 10240
update qwen3_moe of layernorm
```
  b956fc64
- Support blaslt w8a8 GEMM op. · 1a9b2fa9
  zhuwenwen authored Nov 13, 2025
```
解决w8a8 pp16开启marlin的oom问题
```
  1a9b2fa9
- replace triton_ of rms and act_and_mul · 3912d41c
  zhuwenwen authored Nov 13, 2025
  
  3912d41c
- [feat]w4a8和w8a8适配deepep低延迟 · 92761bde
  王敏 authored Nov 13, 2025
  
  92761bde
07 Nov, 2025 2 commits
- add contiguous+rmsnorm to replace triton_ · dc54fefe
  zhuwenwen authored Nov 07, 2025
  
  dc54fefe
- feat: w8a8_marlin 接入，通过-q slimquant_marlin开启，优化w4a8_marlin代码 · 944a8aab
  zhuwenwen authored Nov 07, 2025
  
  944a8aab
04 Nov, 2025 1 commit
- fix index · 6b58062d
  zhuwenwen authored Nov 04, 2025
  
  6b58062d
03 Nov, 2025 3 commits
- use apply_rotary_emb_torch for z100l&k100 · a3695a2b
  zhuwenwen authored Nov 03, 2025
  
  a3695a2b
- update moe configs name · 5ca1259e
  zhuwenwen authored Nov 03, 2025
  
  5ca1259e
- [fix]修复mori报错 · 0e35e124
  王敏 authored Nov 03, 2025
  
  0e35e124
01 Nov, 2025 1 commit
- [feat]整合mori和deepep相关代码 · d698d6f2
  王敏 authored Nov 01, 2025
  
  d698d6f2
29 Oct, 2025 1 commit
- remove redundant envs · c26cfd1a
  zhuwenwen authored Oct 29, 2025
  
  c26cfd1a
27 Oct, 2025 1 commit
- [feat]w4a8适配deepep ht模式，解决开启dp时mtp>1时卡住问题 · 98b7432a
  王敏 authored Oct 27, 2025
  
  98b7432a
24 Oct, 2025 1 commit

add VLLM_USE_LIGHTOP_MOE_SUM_MUL_ADD · c2e6f453

zhuwenwen authored Oct 24, 2025

support prefix cache on kme
fix the error in test_moe caused by moe align not supporting 511 and 211
multi-modal switching to torch implementation on z100l&k100

c2e6f453

17 Oct, 2025 1 commit
- Set the default value of routed_scaling_factor to 1 · 4c20b890
  zhuwenwen authored Oct 17, 2025
  
  4c20b890
15 Oct, 2025 5 commits
- 修复awq 的mtp中的blockint8的问题 · 0cc9bc74
  yangql authored Oct 15, 2025
  
  0cc9bc74
- update deepseek_v2.py · 4b3e2d5e
  zhuwenwen authored Oct 15, 2025
  
  4b3e2d5e
- update deepseek_v2.py · 4ae3fc04
  zhuwenwen authored Oct 15, 2025
  
  4ae3fc04
- 删除DPSK_FP16_QUICK，以及增加awq和blockwiseint8的shared_output接口 · 50cb9270
  yangql authored Oct 15, 2025
  
  50cb9270
- 删除DPSK_FP16_QUICK，以及增加awq和blockwiseint8的shared_output接口 · 7f459b46
  yangql authored Oct 15, 2025
  
  7f459b46
13 Oct, 2025 2 commits
- 去掉all2all ep相关代码 · 0b467604
  王敏 authored Oct 13, 2025
  
  0b467604
- [fix]解决mori ep dp>1时cudagraph卡住问题 · f505d366
  王敏 authored Oct 13, 2025
  
  f505d366