This is a 3-way classifier judge model fine-tuned on the Chatbot Arena human preference dataset. The base model is llama 13B. More details can be found in the Appendix. F of this [paper](Judging LLM-as-a-judge with MT-Bench and Chatbot Arena).
This is a 3-way classifier judge model fine-tuned on the Chatbot Arena human preference dataset. The base model is llama 13B. More details can be found in the Appendix. F of this [paper](Judging LLM-as-a-judge with MT-Bench and Chatbot Arena).