A Qwen2.5 fine-tuned model, focusing on "agent" capabilities, i.e., function-calling.
A fine-tune of Llama 3 (not 3.1) Instruct that scores super well on ChatBotArena before Llama 3.1 came out.
