Paper-Conference

Differentiable Soft Quantization: Bridging Full-Precision and Low-Bit Neural Networks