Daily digest for 2026-11-05 · 8 new stories (previously covered stories are never reposted)
datasette-llm 0.1a10Simon Willison · 2026-11-04
Release: datasette-llm 0.1a10 Fixes a bug that affected my agent.datasette.io demo site, where the set of available LLM models was not correctly restricted to just those in the configuration. Tags: datasette , llm
b12128: tests: fix CI flakes of TOPK_MOE caused by near ties (#31426)llama.cpp Releases · 2026-11-05
tests: fix CI flakes of TOPK_MOE caused by near ties With uniform random logits, two experts sometimes land within one f32 ulp of each other at a routing boundary, mostly with sqrt_softplus gating whose output range is narrow. A backend whose gating differs by a few ulps from the
b12127: metal: keep fusable node groups inside one command buffer (#31688)llama.cpp Releases · 2026-11-05
The first command buffer takes the first 10% of the graph nodes and the rest is split evenly between the extra command buffers. A fusion cannot span two command buffers, so a boundary that falls inside a fusable group runs that group unfused, and the output then depends on the no
b12126: qwen2 : add nvfp4 scales to build_ffn (#31657)llama.cpp Releases · 2026-11-05
This commit adds nvfp4 scales to qwen2's build_ffn call which are currently passed as NULL. The motivation for this is that a model converted to NVFP4 will currently crash because the scales are loaded but never applied in the FFN: llama-completion: llama.cpp/src/llama-sampler.cp
v0.40.15Ollama Releases · 2026-11-05
What's Changed mlx: load bundled CUDA libraries in-process on Linux by @dhiltgen in #19433 model: add MiMo V2.6 model support by @dhiltgen in #19400 Full Changelog : v0.40.14...v0.40.15-rc0