Fetching the paper…

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