new·The score now tells you which way it movedA brain's exam only ever grows: its own material writes questions, and so does every question a real caller asked and did not get answered. The score is a percentage over that growing set, so a brain that learned more could post a smaller number — and this week three did. One of them answered two MORE questions than the week before and showed eighteen points less. Printed as a single percentage, that reads as decline to a reader and as punishment to anyone who contributes material.all news →
mozg.beta
Sign in

AI SDK · Providers · all subjects

llama-cpp/setup

4 notes, read out of this brain and free to use. Each one was extracted from a source and is re-checked against its exam.

llama.cpp resource management

Always call model.dispose() when done with a model to unload it and free GPU/CPU resources. This is especially important when loading multiple models to prevent memory leaks.

llama.cpp provider import and instantiation

The llama.cpp provider is imported from 'ai-sdk-llama-cpp'. Create a model instance using llamaCpp({ modelPath: './path/to/model.gguf' }).

llama.cpp installation

Install the llama.cpp provider using the 'ai-sdk-llama-cpp' package. The installation automatically compiles llama.cpp as a static library with Metal support and builds the native Node.js addon.

llama.cpp prerequisites

The llama.cpp provider requires macOS (Apple Silicon or Intel), Node.js >= 22.0.0, CMake >= 3.15, and Xcode Command Line Tools. Windows and Linux are not supported.

Give your agent this brain