All pull requests

Pull requests

• Add stable model extension support

  #874·jtuyls opened Oct 29, 2026Contributor·

  ·

• feat(d3): d3:flash (vllm-sr/d3-flash, Qwen3.5-9B size)

  #873·tawei-amd opened Oct 29, 2026Collaborator·

  ·

• feat(core): ryzenai-corelib 0.53.0, Qwen3.8-27B, Q4 Qwen3.6, paged context to each model's maximum

  #859·chizamd opened Oct 28, 2026Contributor·

  ·

• ci(d3): MMMU-Pro vision test on the portable builds

  #836·zaneni6 opened Oct 23, 2026Collaborator·

  ·

• feat(hrx): add hrx::Runtime::shutdown()

  #795·quine00 opened Oct 15, 2026Contributor·

  ·
  2

  

• Make the test harnesses link again: backend registry and engine lib path

  #791·zaneni6 opened Oct 15, 2026Collaborator·

  ·

• fix(qwen3.5): capture logprobs on the prefill draw when the think block is prefilled

  #780·manueldomke opened Oct 12, 2026·1 task done·

  ·
  1

• perf(qwen3.5): prefill the empty think block for single-turn models

  #779·manueldomke opened Oct 12, 2026·1 task done·

  ·
  1

• fix(qwen3.5): stop at max_tokens instead of one token past it

  #778·manueldomke opened Oct 12, 2026·1 task done·

  ·

• feat(server): report logprobs when the request asks for them

  #777·manueldomke opened Oct 12, 2026·1 task done·

  ·
  1

• fix: don't count a discarded sample as generated text

  #776·manueldomke opened Oct 12, 2026·1 task done·

  ·
  1

• chore: add security scanning workflows

  #754·haribabug opened Sep 24, 2026·

  ·
  1

• fix: drop a tool call that generation never finished

  #742·Javinator9889 opened Sep 18, 2026Contributor·

  ·
  1

  1

• fix: split the think block in non-streaming responses

  #740·Javinator9889 opened Sep 18, 2026Contributor·

  ·
  1

  1

• fix: send SSE headers before prefill and beat during it

  #735·Javinator9889 opened Sep 17, 2026Contributor·

  ·
  1

• fix: let an unchanged conversation hit the prompt cache

  #734·Javinator9889 opened Sep 17, 2026Contributor·

  ·
  2

• feat: granite 4.2 - kernels

  #701·vegah opened Sep 2, 2026·

  ·
  1

• feat: granite 4.2

  #700·vegah opened Sep 2, 2026·

  ·

• whisper: don't force a timestamp after 16 tokens — truncated every transcription (#698)

  #699·Anjielon opened Sep 1, 2026·

  ·
  1

  1

• docs(embeddinggemma): quantization is bf16, not Q4_1

  #693·Aitschend3251 opened Aug 30, 2026·

  ·
  5

• docs: remove stale OOC note from Gemma 4 benchmark

  #667·ei-grad opened Aug 12, 2026·1 task done·

  ·

• feat: start Qwen3-ASR support

  #654·sxwxs opened Aug 6, 2026·

  ·
  1

• fix: enforce OpenAI tool-choice contract

  #650·waw2637 opened Aug 4, 2026·

  ·
  1

  24

• feat(telemetry): add embedding token count usage

  #594·abn opened Jul 3, 2026·

  ·

• fix: add virtual destructor to npu_cmd to prevent heap corruption (affects v0.9.42+)

  #580·rmanicardi opened Jun 6, 2026·4 tasks·

  ·
  1