Post-Training Quantization for Re-parameterization via Coarse & Fine Weight Splitting
This work addresses the challenge of efficient deployment of re-parameterized networks for computer vision applications, representing an incremental improvement in network quantization techniques.
The paper tackles the problem of significant accuracy drops when applying quantization to re-parameterized neural networks, proposing a coarse & fine weight splitting method and an improved KL metric to reduce quantization error, resulting in only a 0.3% accuracy loss for the quantized RepVGG-A1 model.
Although neural networks have made remarkable advancements in various applications, they require substantial computational and memory resources. Network quantization is a powerful technique to compress neural networks, allowing for more efficient and scalable AI deployments. Recently, Re-parameterization has emerged as a promising technique to enhance model performance while simultaneously alleviating the computational burden in various computer vision tasks. However, the accuracy drops significantly when applying quantization on the re-parameterized networks. We identify that the primary challenge arises from the large variation in weight distribution across the original branches. To address this issue, we propose a coarse & fine weight splitting (CFWS) method to reduce quantization error of weight, and develop an improved KL metric to determine optimal quantization scales for activation. To the best of our knowledge, our approach is the first work that enables post-training quantization applicable on re-parameterized networks. For example, the quantized RepVGG-A1 model exhibits a mere 0.3% accuracy loss. The code is in https://github.com/NeonHo/Coarse-Fine-Weight-Split.git