EfficientBERT: Progressively Searching Multilayer Perceptron via Warm-up Knowledge Distillation

Open in new window