zai-org/LongCite-llama3.1-8b

THUDM/LongCite-llama3.1-8b is an 8 billion parameter language model developed by THUDM, based on Meta-Llama-3.1-8B. It is specifically trained for generating fine-grained citations within long-context question answering tasks. This model supports an extended context window of up to 128K tokens, making it suitable for processing and citing information from very long documents.