Memory-Efficient Fine-Tuning of Transformers via Token Selection.

Antoine Simoulin, Namyong Park, Xiaoyi Liu, Grey Yang

Browse the full EMNLP paper archive.