Data-Free Quantization with Accurate Activation Clipping and Adaptive Batch Normalization. (arXiv:2204.04215v2 [cs.LG] UPDATED)

Data-free quantization is a task that compresses the neural network to low
bit-width without access to original training data. Most existing data-free
quantization methods cause severe performance degradation due to inaccurate
activation clipping range and quantization error, especially for low bit-width.
In this paper, we present a simple yet effective data-free quantization method
with accurate activation clipping and adaptive batch normalization. Accurate
activation clipping (AAC) improves the model accuracy by exploiting accurate
activation information from the full-precision model. Adaptive batch …

arxiv data free lg normalization quantization