YaRN: Efficient Context Window Extension of Large Language Models


The YaRN context window extension method enhances the efficiency of large language models. The team has released 7B and 13B variants of Llama 2 and Mistral 7B fine-tuned with YaRN, available on Hugging Face. The article also provides all the code and data for reproducing the results, promoting open science.
Read more at GitHub…