11:04
Maple-Preview Tested: Ternary Weights, Better on CPU than GPU, Honest Results
Fahd Mirza independently tests Maple-Preview, a 20-billion-parameter sparse mixture-of-experts model from Deep Grove (an independent...
Sort by: Newest Items
There are 84 items in this page
18:30
Qwen3.6-27B Fable Fusion 711: Strongest, Smartest Fine-Tune: Tested Locally
Fahd Mirza installs and thoroughly benchmarks Qwen3.6-27B Fable Fusion 711, a community fine-tune that claims to be the first model o...
46:01
Compression at the Edge — NVIDIA, Unsloth, HuggingFace, Ollama
A conference panel titled "Compression at the Edge" brings together engineers from four of the most active organizations in local AI:...
08:12
How to Run Kimi K3 Locally (3 Ways)
Fahd Mirza delivers an honest, hardware-grounded assessment of running Kimi K3 — Moonshot AI's 2.8-trillion-parameter, 104-billion-ac...
09:16
DeepSeek V4 Flash Fully Local — 32 tok/s on a Single Chip
Fahd Mirza demonstrates running DeepSeek V4 Flash, a 284-billion parameter mixture-of-experts model, entirely on a single AMD Ryzen A...
25:28
AMD Ryzen AI Halo – 100% Local AI
Sam Witteveen reviews the AMD Ryzen AI Halo, a new workstation built around the Ryzen AI Max Plus 395 chip with 128 GB of unified LPD...
11:00
Gemma 4’s Big Update in 4-Bit QAT — Low VRAM, Local
Fahd Mirza covers the most significant update to Google's Gemma 4 model to date: 11 targeted fixes spanning tool calling, chat format...
12:14
Ternary Bonsai 27B: Full Model, 2-Bit, Run on Phone or Laptop or GPU
Fahd Mirza runs Bonsai 27B — a 27-billion parameter open-weight model from Prism ML based on Gemma 3 27B — on an NVIDIA RTX A6000 GPU...
10:03
Qwythos-9B-v2: The Looping Bug Gone – FTPO Explained + Local Testing
Fahd Mirza tests Qwythos 9B v2 (also called Qutoz) — a 9-billion-parameter fine-tune of Qwen 2 with deep chain-of-thought reasoning —...
08:55
OpenLumara + Ollama: Local Agent Trying to Beat OpenClaw and Hermes
Fahd Mirza walks through the installation and live testing of OpenLumara, an open-source Python agent framework designed specifically...