Plain-language guides about running language models at home: memory, quantisation, hardware choices and setup.
Categories
Weights, context cache and overhead: how to work out whether a model fits your GPU or Mac, with real numbers for popular models.