ParentFull threadbugglebeetle·Have you looked at using the long context (32K) version of the Llama v2 7B released by Together AI?https://together.ai/blog/llama-2-7b-32kView on HN