Paper-Collection-of-Efficient-ML

paper collection

Quantization

Surveys and Technical Reports

1. Quantizing deep convolutional networks for efficient inference: A whitepaper [2018]

Incremental methods

1. Two-Step Quantization for Low-bit Neural Networks [CVPR-2018]
2. Incremental Network Quantization: Towards Lossless CNNs with Low-Precision Weights [ICLR-2017]
3. BinaryRelax: A Relaxation Approach For Training Deep Neural Networks With Quantized Weights [2018]
4. Overcoming Challenges in Fixed Point Training of Deep Convolutional Networks [ICML@Workshop-2018]