# efficient inference optimization

Yarn Mistral 7B 128K is an advanced language model optimized for long-context processing, further pre-trained on long-context data using the YaRN extension method, supporting a 128k token context window.