Researchers Propose a Novel Framework ‘LilNetX’ For Training Deep Neural Network With Extreme Model Compression, and Structured Sparsification – MarkTechPost