Semi-Offline Reinforcement Learning for Optimized Text Generation
We propose a semi-offline reinforcement learning approach for optimizing text generation in language models, balancing exploration and exploitation effectively....
May 1, 2023
Prototypical Fine-Tuning: Towards Robust Performance under Varying Data Sizes
We propose prototypical fine-tuning, a novel framework for fine-tuning pretrained language models that maintains robust performance across varying data sizes.
Jan 1, 2023