ReLeQ: An Automatic Reinforcement Learning Approach for Deep Quantization of Neural Networks

By Ahmed T. Elthakeb, Prannoy Pilligundla, Fatemehsadat Mireshghallah, Amir Yazdanbakhsh, Hadi Esmaeilzadeh. 2018

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

1. Alexandre D'efossez
2. Yossi Adi
3. Gabriel Synnaeve
4. Junjie Yan
5. Xianglong Liu
6. Qing Jin
7. Wanli Ouyang
8. Fisher Yu

Citations in the corpus

• Differentiable Model Compression via Pseudo Quantization Noise (2021, background): “Wang et al. (2019); Elthakeb et al. (2019) suggested learning a bitwidth assignment policy using reinforcement learning methods.”
• AdaBits: Neural Network Quantization With Adaptive Bit-Widths (2019, methodology): “This suggests we might be able to employ the quantized models with adaptive bit-widths to search for bit-width in each layer or channel, for the purpose of mixed-precision quantization [10, 44, 42, 41, 26].”
• Differentiable Soft Quantization: Bridging Full-Precision and Low-Bit Neural Networks (2019, methodology): “[36, 46, 39] apply different techniques to find optimal bit-widths for each layer.”
• Differentiable Joint Pruning and Quantization for Hardware Efficiency (2020, background): “In [30][34] reinforcementbased schemes are proposed to learn the bit-width.”