1 paper
Abdullatif Köksal, Timo Schick, Anna Korhonen +1
Instruction tuning enables language models to more effectively generalize and better follow user intent. However, obtaining instruction data is costly and challenging. Prior work e…