Open standard for machine learning interoperability
[ICCV 2023] I-ViT: Integer-only Quantization for Efficient Vision Transformer Inference