[ICCV 2023] I-ViT: Integer-only Quantization for Efficient Vision Transformer Inference
Primary LanguagePythonApache License 2.0Apache-2.0