The new experimental llama.cpp feature expands MoE expert activation to 16, 24, or all available experts, replacing the fixed Top-K limit with a tunable compute budget. This allows activating more experts than in the standard implementation, but requires more memory and i…

#LlamaCPP #MOE #Inference

#llamacpp +
@liquidai
#d13b model + #webcam
huggingface.co/LiquidAI/d1-...

#decisionmodel

The real difference is just whether you prefer a polished UI or an open-source, hackable core. #selfhosting #batchinference #gpuoptimization #llm #localai #llamacpp 2/2

yzma v1.30 is out with @hf.co llama.cpp v0.7.0 support! More System One decision models, and decision heads running right in your browser. You can go get it right now!

github.com/hybridgroup/yzma

#golang #llamacpp #yzma #tinygo #systemOne

Cluster de Macs para LLM: qué dice la guía y qué no

¿Armás un cluster de Macs para LLM con exo, llama.cpp y Ray? Revisá qué falla en la guía de dev.to (capas, costos y comandos) antes de copiarla

#applesilicon #exo #llamacpp #inferenciadistribuida #rayserve

The update includes updating llama.cpp for decision models and optimizing recurrent state in MLX.

#Ollama #LlamaCPP #MLX

AWS published a how-to on deploying decision models with llama.cpp on SageMaker AI using the Deep Learning Container, with SemIf extensions for…

#AWSSageMaker #llamacpp #AIInference #MLOps
https://aws.amazon.com/blogs/machine-learning/deploy-decision-models-with-llama-cpp-on-amazon-sagemaker-ai/

「専門知識ゼロ・クラウドAIとのチャットのみでローカルAI構築」1巻を note でも出しました↓.
note.com/bitespotfact...
ローカルLLMを走らせるまでの失敗と解決を, 6章にまとめています. 中身は Zenn 版と同じです.
#ローカルLLM #自作AI環境 #llamacpp

llama.cpp Quickstart with CLI and Server

#Cheatsheet #AI #LLM #DevOps #OpenAI #API #SelfHosting #Prometheus #llamacpp

https://www.glukhov.org/llm-hosting/llama-cpp/

Jeyzma is like @typesafeai.bsky.social minus $870 million plus @golang.org & @tinygo.org & @hf.co llama.cpp

Go make a decision!

Already available right now in your browser.

jeyzma.com

#jev #decisionAI #browserAI #llamacpp #systemOne

🧠 Second Hemisphere for Local AI Agents: Beating Context Overflow

During long-horizon runs (coding, bash, 50–100+ steps), tool outputs saturate the context window → HTTP 400 → session dies.

🔗 github.com/Dark-city-be...

#OpenSource #AIagents #LocalAI #llamaCpp #Py

CVE-2026-107183 - llama.cpp
The open‑source llama.cpp library may let an attacker send a specially formed chat request that causes the server to stop working and potentially change data in memory. This…

Too many irrelevant or confusing CVEs? Use stackflag.com

#llamacpp #ggmlorg #CVE #infosec

⚙️ Local 27B on mining GPUs: faster inference with zero quality loss

Running Qwen3.8-27B on a dual CMP 50HX setup (Ivy Bridge platform) using MTP speculative decoding.

🔗 github.com/ggml-org/lla...

#localAI #llamaCpp #Qwen #CMP50HX #MTP

yzma 1.29 is out! Run high performance local AI models from @golang.org without CGo. Support for the new @hf.co llama.cpp v0.6.0, typed decisions w/ System One models, auto-threading, new extended batch API, & much more!

Go get it right now!

yzma.ai #golang #llamacpp #gguf #tinygo #localAI

WordPress SEO Plugins Compared: Yoast to Local AI

#SEO #Self-Hosting #SelfHosting #Ollama #llamacpp #Open Source #AI

https://www.glukhov.org/web-infrastructure/wordpress/wordpress-seo-plugins-compared/

Self-Hosted SEO Tools and Platforms: Open Source Guide

#SEO #Self-Hosting #SelfHosting #Open Source #Ollama #llamacpp #DevOps

https://www.glukhov.org/web-infrastructure/seo/self-hosted-seo-tools-and-platforms/

llmman: "Run any agent on any model". It pulls models like OCI images and serves them behind one API that speaks OpenAI, Ollama and Anthropic.

I tried it with llama.cpp as the runtime, on macOS and Linux, plus a Go example.

k33g.org/p/20261001-l...

#LocalModels #llamacpp

#llamacpp now supports local #jev like decision models huggingface.co/blog/ggml-or... #ai #llm