1-bit architecture is turbocharging LLM efficiency

Does this result in a regular model that say llama-cpp can run? Is there any way to test these ourselves?

"... Compared to full-precision Llama models, BitNet a4.8 reduces memory usage by a factor of 10 and achieves 4x speedup..."