All pull requests

Pull requests

• fix(calibrate): keep configured vision resident during tuning

  #2312·Tsury opened Oct 13, 2026·

  ·
  1

• web: default temperature 1.0, as the model card and generation_config.json say

  #2311·kevin9327 opened Oct 13, 2026·

  ·
  1

• docker: complete deployment bundle with a management console

  #2308·jazzlee008 opened Oct 13, 2026·

  ·

• setup: write the gfx1103 exact speed switches for the opt-in install

  #2304·AlexisCarbonell opened Oct 13, 2026·

  ·
  1

• hip: a hipBLASLt tuning table for gfx1103 (Radeon 780M)

  #2303·AlexisCarbonell opened Oct 13, 2026·

  ·
  1

• hip: gfx1103 gets the native dot4 and the gfx11 WMMA kernels

  #2302·AlexisCarbonell opened Oct 13, 2026·

  ·
  1

• bench: add hosted Codex quality baseline for the quant study

  #2300·Jhacarreiro opened Oct 13, 2026·

  ·

• prefill: STRATA_GDN_CHUNKED=1 on gfx11 WMMA (opt-in; gfx1151: recurrence 1.6x, a warm 4K-128K prompt 1.4-2.1%)

  #2299·libratechw opened Oct 13, 2026Contributor·

  ·

• community benchmark: RTX 4080 SUPER, Qwen3.8-Flash-Next IQ2_XS

  #2298·johlim opened Oct 13, 2026·

  ·

• bench: run under UTF-8 and read CUDA from new nvidia-smi headers

  #2296·johlim opened Oct 13, 2026·

  ·

• web: an optional interface language (zh-CN first), English unchanged

  #2294·wulezhi-hui opened Oct 13, 2026·

  ·

• Anthropic API: cache_creation_input_tokens, the prompt tokens read now and kept for the next request

  #2290·xrip opened Oct 13, 2026·

  ·
  1

• tools: validate GGUF headers and tensor extents

  #2288·KeyanHu-git opened Oct 13, 2026·

  ·

• setup: resume downloads when HEAD omits Content-Length

  #2287·KeyanHu-git opened Oct 13, 2026·

  ·
  1

  1

• fix(vision): retain image ownership through request preparation

  #2286·Tsury opened Oct 13, 2026·

  ·

• fix(serve): preserve late instruction roles across adapters

  #2285·Tsury opened Oct 13, 2026·

  ·

• fix(serve): preserve declared tool schemas through streamed aliases

  #2284·Tsury opened Oct 13, 2026·

  ·

• setup: opt-in gfx1010 (RX 5700 XT) plus a community bench

  #2283·Gedankenn opened Oct 13, 2026·2 of 3 tasks·

  ·

• bench: community report, RTX PRO 6000 Blackwell Workstation Edition 96 GB (sm_120), IQ3_S fully resident

  #2280·beastrobin opened Oct 13, 2026·

  ·

• serve: add the legacy /v1/completions endpoint for raw-prompt clients (#2277)

  #2278·kossum opened Oct 13, 2026Contributor·

  ·

• bench: community report, RX 9060 XT (gfx1200) with 16 GB of RAM and a SATA SSD, Coder IQ1_M at 128K on 0.1.43.1, stock vs tuned

  #2276·mandoof1 opened Oct 13, 2026·

  ·

• serve: an AMD card's load, VRAM, temperature and power on Windows

  #2273·unimo-chan opened Oct 13, 2026Contributor·

  ·
  1

• Monitor: the PCIe traffic from the bytes the engine moved, not from a driver counter

  #2272·unimo-chan opened Oct 13, 2026Contributor·

  ·
  2

• --shared-expert-arena: one engine loads the file, the others wait and attach

  #2271·i-Vladimir-i opened Oct 13, 2026·

  ·

• Conversation cache: --conversation-cache-park-gain N (opt-in, default 0): a turn that grew a few hundred tokens no longer rewrites a whole 3 GB snapshot

  #2265·Krypto-Whitehat opened Oct 13, 2026Contributor·

  ·
  1