Smaller Language Models are capable of selecting Instruction-Tuning Training Data for Larger Language Models

ALM
Re-assign community