quantization — Shrinking a model by storing its numbers less precisely — like rounding — so it runs faster and fits on smaller hardware, at a small quality cost.
distillation — Training a small, cheap model to imitate a big one's outputs, keeping much of the capability at a fraction of the cost.
fine-tuning — Taking a trained model and training it a bit more on your own examples so it gets better at one specific job.
Why it matters
A 21B open-weights model positioned as a fine-tuningTaking a trained model and training it a bit more on your own examples so it gets better at one specific job.Full definition → base for agentic and coding work, plus a quantizationShrinking a model by storing its numbers less precisely — like rounding — so it runs faster and fits on smaller hardware, at a small quality cost.Full definition → library that cuts 3.5-bit degradation from 6.8 to 1.6 average points against llama.cpp's own Q3_K_S routine.
