Self-Training with Direct Preference Optimization Improves Chain-of-Thought Reasoning

Devs

Self-Training with Direct Preference Optimization Improves Chain-of-Thought Reasoning | Read Paper on Bytez