BAAI/Infinity-Instruct-3M-0613-Llama3-70B

Infinity-Instruct-3M-0613-Llama3-70B by Beijing Academy of Artificial Intelligence (BAAI) is a 70 billion parameter instruction-tuned causal language model based on the Llama3 architecture. It is fine-tuned on the Infinity-Instruct-3M and Infinity-Instruct-0613 datasets without reinforcement learning from human feedback (RLHF). This model demonstrates strong instruction-following capabilities, achieving 31.5 on AlpacaEval 2.0, surpassing GPT4-0613 in this benchmark.