All pull requests

Pull requests

• [beta] Kimi-K3 runtime stack of #906: FP4-CSF loaders, draft precision, KDA prefill, MLA priming release

  #963·voipmonitor opened Oct 1, 2026·

  ·
  1

• [KK] Qwen3.8-Flash-Next: share one host-RAM PLE table between same-host processes

  #962·voipmonitor opened Oct 1, 2026·

  ·
  4

• fix(scheduler): give prefill lanes to queued requests only while admissible

  #960·voipmonitor opened Oct 1, 2026·

  ·
  2

• feat(moe): native B12X MXFP8 MoE for ModelOpt MTP experts

  #957·voipmonitor opened Oct 1, 2026·

  ·
  2

• Load FP4-CSF checkpoints through standard B12X preparation

  #956·voipmonitor opened Oct 1, 2026·

  ·
  1

• [KK] Release dense MLA priming buffers after preparation

  #953·voipmonitor opened Sep 30, 2026·

  ·
  1

• chore(release): wheels require CUTLASS DSL 4.7.1 with B12X (canonical twin of #951)

  #952·voipmonitor opened Sep 30, 2026·

  ·
  3

• [KK] Load DS4.1 and Kimi-K3 lossless X4T checkpoints

  #949·voipmonitor opened Sep 29, 2026·

  ·
  1

• fix(scheduler): let decodes preempt in a prefill turn that scheduled nothing (canonical twin of #946)

  #948·voipmonitor opened Sep 29, 2026·

  ·
  1

• fix(release): drop Python staged by other branches from the shared wheel build cache (canonical twin of #939)

  #940·voipmonitor opened Sep 29, 2026·

  ·
  1

• [KK] Prepare B12X Kimi KDA prefill and expose MLA partial precision

  #937·voipmonitor opened Sep 29, 2026·

  ·
  1

• fix(model-state): stage per-step prefill-flag uploads

  #923·FujitsuPolycom opened Sep 27, 2026·

  ·
  1

• feat(worker): GPU stall watchdog; RoCEnante health on every rank

  #916·tobymao opened Sep 27, 2026·

  ·
  6

• perf(glm5next): serve native block-FP8 DSA q_b_proj and o_proj on Marlin

  #915·tobymao opened Sep 27, 2026·

  ·
  6

• fix: stop request-path Triton kernels recompiling per batch

  #914·tobymao opened Sep 27, 2026·

  ·
  6

• fix(shm_broadcast): never drop a write notification

  #913·tobymao opened Sep 27, 2026·

  ·
  5

• feat(qwen): serve QSA from an NVFP4 main K/V cache (nvfp4_qsa)

  #910·yatesdr opened Sep 27, 2026·

  ·
  5

• Karmic beta to canonical (Oct 1): #861–#866, #869, #870, #883, #884, #886, #887, #889, #895, #897, #898, #903, #904, #908, MiMo-V2.6 (#905), #911, #918–#921, #928–#932, #934, #939, #946, #950, #951, #954, #958, #959

  #894·voipmonitor opened Sep 24, 2026·

  ·
  22

• feat(mimo): launch MiMo-V2.6-Flash with DFlash and the b12x serving flags

  #881·MadeBy561 opened Sep 23, 2026·

  ·
  1

• perf(scheduler): let prefill-only steps use max_num_batched_tokens

  #880·MadeBy561 opened Sep 23, 2026·

  ·
  1

• fix(spec-decode): keep the adaptive verification acceptance estimate recoverable

  #879·MadeBy561 opened Sep 23, 2026·

  ·
  1

• feat(dflash): support block-FP8 drafters in the fused context projection

  #878·MadeBy561 opened Sep 23, 2026·

  ·
  1

• feat(dflash): shard the context projection and draft per vocab shard at TP > 1

  #877·MadeBy561 opened Sep 23, 2026·

  ·
  1

• fix(multimodal): serve MiMo-V2 audio without torchaudio

  #876·MadeBy561 opened Sep 23, 2026·

  ·
  1

• perf(mimo): prefetch upcoming decode weights into L2 on a side stream

  #875·MadeBy561 opened Sep 23, 2026·

  ·
  1