Tag: local-llm
All the articles with the tag "local-llm".
My vibe-selling rig is a 13-inch tablet
Benchmarks from a 128 GB Strix Halo tablet: four models from a dense 27B to a 122B MoE on both AMD backends, a Poolside Laguna three days after its llama.cpp merge, and why parameter count is the wrong number.