Issues

• [INFO] Splash 1.6.1: Metal created pipeline verify_attention_q8q_split while serving, in 0.145833 ms

  incoai/splash#736 · Filter by author nailarch opened on Nov 10, 2026

• M5 Max (T6050) total-system watchdog panic under ~112 GB engine pool: 'no checkins from watchdogd in 91 s' (macOS 27.2/26B123)

  incoai/splash#735 · Filter by author brainnoms opened on Nov 10, 2026
  1

• Keep a reusable state at the end of a generated reply, so follow-up turns don't re-prefill it (hybrid GDN models)

  incoai/splash#732 · Filter by author JSummerer opened on Nov 9, 2026

• json_schema responses: the grammar allows 64 whitespace characters before the JSON value, and a client cannot ask for compact output

  incoai/splash#730 · Filter by author tarrific opened on Nov 9, 2026

• Tree verify (--verify-tree, default off): results and default question (PR #727)

  incoai/splash#728 · Filter by author linson007 opened on Nov 9, 2026

• Getting warnings 'memory: in-service claim' during decoding

  incoai/splash#726 · Filter by author didikrm opened on Nov 8, 2026

• Request DFlash2-FP8 support for m6

  incoai/splash#725 · Filter by author nailarch opened on Nov 8, 2026

• New Dspine Drafter By Tencent

  incoai/splash#724 · Filter by author kg55555 opened on Nov 8, 2026

• Agent workloads: tokenizer prefix cache (4 entries) thrashes and tool schemas re-normalize every request

  incoai/splash#721 · Filter by author dmkjr opened on Nov 8, 2026

• Shared finalizer thread can stall all completions on a blocked stdout/stderr pipe

  incoai/splash#719 · Filter by author dmkjr opened on Nov 8, 2026

• 1.6.0: persistent cache writes ~2 TB/day to the internal SSD; about half of state write-behinds unneeded

  incoai/splash#720 · Filter by author dmkjr opened on Nov 8, 2026
  1

• No TCP keepalive: a silently vanished non-streaming client holds its lane for the full request

  incoai/splash#718 · Filter by author dmkjr opened on Nov 8, 2026

• Request quarantine never expires and can block an innocent agent conversation until restart

  incoai/splash#717 · Filter by author dmkjr opened on Nov 8, 2026

• Server stays up but unrecoverable after crash-loop stop; process should exit so a supervisor can restart it

  incoai/splash#716 · Filter by author dmkjr opened on Nov 8, 2026

• 1.6.0 regression: request with reasoning_effort: "none" (thinking off) hangs, works on 1.5.0

  incoai/splash#715 · Filter by author rickhuizinga opened on Nov 7, 2026
  3

• Contributor questions: DFlash2 draft training data, checkpoint contributions, and which open work to review

  incoai/splash#714 · Filter by author linson007 opened on Nov 7, 2026

• 1.6.0: elastic expert store floor computed at 100% of the store

  incoai/splash#713 · Filter by author mperreir opened on Nov 7, 2026
  1

• Design: cost-aware allocation across compact32 and wide64 verification

  incoai/splash#709 · Filter by author linson007 opened on Nov 7, 2026
  2

• Research: compact mixer verification projections after FFN packing

  incoai/splash#708 · Filter by author linson007 opened on Nov 7, 2026
  2

• Research: selected-prefix FFN packing for bounded eight-request decode

  incoai/splash#707 · Filter by author linson007 opened on Nov 7, 2026
  2

• 1.6.0 has major cache-loss issues not seen in 1.5.0 on M5 Max

  incoai/splash#706 · Filter by author tomchk opened on Nov 7, 2026
  10

• Sundry macOS App and Web UI Improvements

  incoai/splash#703 · Filter by author jiiim opened on Nov 6, 2026

• Feature: Session-level KV cache release APIs

  incoai/splash#701 · Filter by author cuijs2026 opened on Nov 6, 2026

• Serve record stays "ready" through an in-process engine relaunch, so a supervisor cannot tell the relaunch from a runaway load

  incoai/splash#700 · Filter by author pcvelz opened on Nov 6, 2026

• [Feature/Robustness] Complete agentic workflow support in Responses API: tolerant input items & compaction output handling

  incoai/splash#695 · Filter by author hometrix opened on Nov 6, 2026