Open-ended Instruction Following on Self-instruct Eval
53.6Win Rate (A)BPO + Llama-2-chat 7B
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| BPO + Llama-2-chat 7BBase LLM=Llama-2-chat, Model A=7B + BPO, Model B=7B, Evaluation LLM=GPT-42023.11 | 53.6 | 9.9 | 36.5 | 17.4 | |
| BPO + Llama-2-chat 13BBase LLM=Llama-2-chat, Model A=13B + BPO, Model B=13B, Evaluation LLM=GPT-42023.11 | 51.2 | 11.9 | 36.9 | 18.1 | |
| BPO + Llama-2-chat 13B (Cross-size)Base LLM=Llama-2-chat, Model A=13B + BPO, Model B=70B, Evaluation LLM=GPT-42023.11 | 48.4 | 4.8 | 46.8 | 11.9 | |
| BPO + Vicuna-v1.3 13BBase LLM=Vicuna-v1.3, Model A=13B + BPO, Model B=13B, Evaluation LLM=GPT-42023.11 | 46.4 | 13.9 | 39.7 | 13.1 | |
| BPO + Llama-2-chat 70BBase LLM=Llama-2-chat, Model A=70B + BPO, Model B=70B, Evaluation LLM=GPT-42023.11 | 46 | 13.1 | 40.9 | 16.8 | |
| BPO + Vicuna-v1.3 7BBase LLM=Vicuna-v1.3, Model A=7B + BPO, Model B=7B, Evaluation LLM=GPT-42023.11 | 42 | 21.1 | 36.9 | 18.5 | |
| BPO + Llama-2-chat 7B (Cross-size)Base LLM=Llama-2-chat, Model A=7B + BPO, Model B=70B, Evaluation LLM=GPT-42023.11 | 40.1 | 5.1 | 54.8 | -7.1 |