Webinar: Using “Compound Sparsification” with Hugging Face BERT for Faster CPU Inference with Better Accuracy. Register now.