Fetching the paper…

Extreme Compression of Large Language Models via Additive Quantization · Around