All pull requests

Pull requests

• [Benchmark] 2x RX 6800 XT (x8 + x8) on Windows / UD-Q4_K_XL / Strata 0.2.11 / engine 0.1.57

  #5446·Expired-Pasta opened Nov 7, 2026·

  ·

• [Docs] Add Korean README translation (README.ko.md)

  #5445·Expired-Pasta opened Nov 7, 2026·

  ·

• [Benchmark] RTX 5080 / Qwen3.8-Flash-Next-GSQ-RCO / IQ3_S / strata-qwen38-thinking-v1-draft.1

  #5442·teohhc opened Nov 7, 2026·

  ·
  2

• [Engine] sm_60: STRATA_Q8_SM60_EXACT, the Pascal Q8_0 GEMV in the exact kernels' addition order (bitwise, opt-in)

  #5438·Keoian opened Nov 6, 2026·

  ·

• [Other] keep_awake: the PC stays awake while the model serves, and while setup runs

  #5437·mazzyler opened Nov 6, 2026·

  ·

• [Engine] opt-in STRATA_ARENA_NUMA: place the expert arena per NUMA node on two-socket Linux

  #5436·Keoian opened Nov 6, 2026·

  ·

• [Engine] Auto-load sibling token-embd-bf16.gguf for sub-8-bit native token embeddings

  #5433·bsorensen110 opened Nov 6, 2026Contributor·

  ·
  6

• [Engine] Do not force WindowInvariant CPU experts on DFlash2 windows when MTP fallback is loaded

  #5432·bsorensen110 opened Nov 6, 2026Contributor·

  ·
  1

• [Benchmark] RTX 4070 Ti SUPER 16 GB / Qwen3.8-Flash-Next IQ2_XS / 0.2.11 engine 0.1.57

  #5431·EternalShade3D opened Nov 6, 2026·

  ·

• [Benchmark] 4x RTX 3090 / Qwen3.8-Flash-Next / UD-Q4_K_XL / 262k context (v0.2.11)

  #5428·filmo opened Nov 6, 2026·

  ·
  2

• [Other] Drafting: a GPU lookup that found no card is made again, so a service started at boot gets DFlash2

  #5427·ezoushen opened Nov 6, 2026·

  ·

• [Engine] fix layer-major prefill overflowing the group-scratch row cap

  #5418·WilliamStone opened Nov 6, 2026·

  ·

• [Benchmark] RX 7900 XTX / qwen3.8-flash-next / IQ2_XS / pre-protocol quick report

  #5415·jkuczera opened Nov 6, 2026·

  ·
  2

• [Engine] sycl: Prefill::release destroys the default in-order queue, and 349 frees use it afterwards at exit

  #5405·aslater3 opened Nov 6, 2026Contributor·

  ·

• Serve: parked conversations spill to a disk tier instead of being dropped

  needs-format
  #5399·nrtdadmin opened Nov 6, 2026·

  ·
  3

• [Engine] Integrate tool and JSON constraints on Metal, with Mac regression evidence

  #5398·tuhochi opened Nov 6, 2026·

  ·

• [Benchmark] RTX 4090 / Qwen3.8-Flash-Next / IQ3_XXS / strata-qwen38-thinking-v1-draft.1

  #5396·kamilos876 opened Nov 6, 2026·

  ·
  2

• [Engine] Metal: fix garbage output with Unsloth low-bit files (BF16 ple_conv1d read as F16)

  #5387·qChauncey opened Nov 6, 2026·

  ·

• [Benchmark] RX 7900 XTX / Qwen3.8-Flash-Next / IQ3_XXS / strata-qwen38-thinking-v1-draft.1

  #5386·GoldenLeafBird opened Nov 6, 2026Contributor·

  ·
  2

• [Benchmark] RTX 4080 SUPER 16 GiB, i9-14900KF, 32 GiB RAM: Swift 1.5 IQ2_XS at 128K

  #5384·diegoarodriguezr opened Nov 6, 2026·

  ·

• [Engine] dsv41: apply a request's logit_bias in the DeepSeek-V4.1 engine

  #5382·srkim-business opened Nov 6, 2026·

  ·

• [GUI] show saved KV format for packaged Qwen3.8 models

  #5374·zapabob opened Nov 6, 2026Contributor·

  ·

• [TUI] keep the values the Settings view saved across launches; --set idle-unload writes its config key instead of an engine flag

  #5371·shikaku2 opened Nov 6, 2026·

  ·
  1

• [Benchmark] Apple M2 Ultra / Qwen3.8-Flash-Next / IQ3_S / strata-qwen38-thinking-v1-draft.1

  #5368·5ugAv opened Nov 6, 2026·

  ·
  2

• benchmark: 2x Tesla V100-SXM2-16GB (PCIe carriers) / Huihui abliterated-v2 / UD-Q4_K_XL / strata-qwen38-thinking-v1-draft.1

  needs-format
  #5367·PoisonousHumanMineral opened Nov 6, 2026·

  ·
  1