Quantized Llama models with increased speed and a reduced memory footprint



Login to add comment