We are happy to announce MILO: Model-agnostic subset selection for Efficient Training and Tuning. MILO achieves an impressive 3x-10x faster training and a whooping 20x-75x faster hyper-parameter tuning.
🚀 (1/6) Exciting News! Introducing MILO, our revolutionary Model Agnostic Subset Selection Framework! It can optimize your model training & tuning, accelerating your #NeuralNetworks training by 3x-10x and supercharging your hyper-parameter tuning by an incredible 20x - 75x.View quoted post on X