Local 1M Context Inference at 15 tokens/s and ~100% "Needle In a Haystack" Posted by quxinxin on Thursday August 29, 2024 @10:00PM from the old.reddit.com dept. https://old.reddit.com/r/LocalLLaMA/comments/1f3xfnk/local_1m_context_inference_at_15_tokenss_and_100/ ( no comments | on HN )
There are no comments yet.