The maximum number of tokens (words or subwords) a model can process in a single input, in this case 8K tokens per chunk.
Performance retention over long documents and conversations