Commits

Branch selector

 main

User selector

All users

All time

Commit History

Commits on Feb 10, 2026

• Fix wrong prefill log. (#18570)

  hnyls2002

  authored
  93fca0b

  Copy full SHA for 93fca0b

• Fix radix cache key to include generated tokens in multi-turn (regression) (#16521)

  Show description for 2bfab1b
  

  ycchen-tw

  and

  Copilot

  authored
  2bfab1b

  Copy full SHA for 2bfab1b

• Tilelang sparse decode fwd for dsv32 mi355 (#18488)

  Show description for 4262f52
  

  1am9trash

  and

  kkHuang-amd

  authored
  4262f52

  Copy full SHA for 4262f52

• Revert "[sgl-kernel] upgrade deepgemm" (#18562)

  Fridge003

  authored
  2d38b8a

  Copy full SHA for 2d38b8a

• [NPU][docs]fix bug about hyperlink for best practice for ascend npu (#18561)

  

  husf

  authored
  573ff55

  Copy full SHA for 573ff55

• fix(config): Support setting Mamba state dtype via config file (#18532)

  Show description for 4460376
  

  zju-stu-lizheng

  and

  瑀澈

  authored
  4460376

  Copy full SHA for 4460376

• [NPU] update npu doc (#18474)

  Hexq0210

  authored
  d0d387d

  Copy full SHA for d0d387d

• [NPU][docs] improve docs for Best Practice on Ascend NPU (#18360)

  

  husf

  authored
  99101ce

  Copy full SHA for 99101ce

• [sgl-kernel] upgrade deepgemm (#18362)

  BBuf

  authored
  bec7fe9

  Copy full SHA for bec7fe9

• [diffusion] fix: fix fsdp (#18187)

  mickqian

  authored
  efcdda0

  Copy full SHA for efcdda0

• [NPU] [CI] Enable run multimodal NPU CI when changes only in multimodal_gen (#18523)

  Makcum888e

  authored
  49cbb46

  Copy full SHA for 49cbb46

• [diffusion] feat: support parallel wan-vae decode (#18179)

  nono-Sang

  authored
  47978ee

  Copy full SHA for 47978ee

• [DLLM] Basic dLLM scheduling strategy and implementation (#17484)

  Show description for 26f2b37
  ClawSeven

  authored
  26f2b37

  Copy full SHA for 26f2b37

• [HiCache] fix: StorageMetricsCollector was initialized twice (#18354)

  alphabetc1

  authored
  8da14ae

  Copy full SHA for 8da14ae

• [AMD] Turn on aiter-prebuild (#18425)

  Show description for d94d0af
  

  yctseng0211

  and

  bingxche

  authored
  d94d0af

  Copy full SHA for d94d0af

• Support GlmMoeDsaForCausalLM (#18521)

  Show description for 398b81f
  

  JustinTong0323

  and

  BBuf

  authored
  398b81f

  Copy full SHA for 398b81f

• Deepseekv32 compatibility with transformers v5 (#18297)

  Show description for e8a2c13
  

  3 peopleauthored
  e8a2c13

  Copy full SHA for e8a2c13

• [EPD] Add notification mechanism to fix server hang and add timeout env var (#18229)

  liusy58

  authored
  0b15f19

  Copy full SHA for 0b15f19

• Make bench_one_batch_server compatible for more backends (#18512)

  maocheng23

  authored
  1d366f1

  Copy full SHA for 1d366f1

• Fix idle batch predict dtype in spec v2 (#18379)

  Qiaolin-Yu

  authored
  4a1b50b

  Copy full SHA for 4a1b50b

• [HiCache][PP] add test case for compatibility (#16395)

  stmatengss

  authored
  df733c5

  Copy full SHA for df733c5

• Tiny fix wrong metric collect key name `forward_prefill` -> `forward_extend` (#18506)

  hnyls2002

  authored
  2825f5d

  Copy full SHA for 2825f5d

• Add cache_config_info metric. (#17273)

  kartikx

  authored
  26a006e

  Copy full SHA for 26a006e

Commits on Feb 9, 2026

• docs: expand and update modelopt documentation (#18479)

  Show description for 54589a2
  

  3 peopleauthored
  54589a2

  Copy full SHA for 54589a2

• [Auto Sync] Update cache_init_params.py (20260209) (#18502)

  Show description for b027c5a
  

  4 peopleauthored
  b027c5a

  Copy full SHA for b027c5a

• [Auto Sync] Update logits_processor.py (20260209) (#18503)

  Show description for ce95f20
  

  merrymercy

  and

  github-actions[bot]

  authored
  ce95f20

  Copy full SHA for ce95f20

• feat(kv-events): Add medium field to KV event types for storage tier tracking (#18205)

  ishandhanani

  authored
  01e3f46

  Copy full SHA for 01e3f46

• [AMD] add amd ci monitor (#17476)

  Show description for 316f9cb
  

  3 peopleauthored
  316f9cb

  Copy full SHA for 316f9cb

• model: support Qwen3.5 (#18489)

  Show description for 27c4476
  

  zju-stu-lizheng

  and

  瑀澈

  authored
  27c4476

  Copy full SHA for 27c4476

• Fix MMLU benchmark to auto-download data and resolve path issue (#18486)

  JustinTong0323

  authored
  0b4d4f2

  Copy full SHA for 0b4d4f2

• Pass `quantize_config` to `_initialize_model` (#18273)

  klshuster

  authored
  006da22

  Copy full SHA for 006da22

• feature: support bidirectional attention for Gemma-3 (#10707)

  zzhbrr

  authored
  ddbcfba

  Copy full SHA for ddbcfba

• [diffusion] chore: fix unclean shutdown and resource leaks (#18477)

  mickqian

  authored
  4f7da5a

  Copy full SHA for 4f7da5a

• [diffusion] feat: add ModelScope support (#17924)

  yrk111222

  authored
  76eb1c8

  Copy full SHA for 76eb1c8

• Revert "optimize get_topk_ragged by fusing get k and k_scale triton kernel" (#18471)

  Fridge003

  authored
  615a02d

  Copy full SHA for 615a02d

Pagination

Previous

Next