Lizhikai

I-ViT: Integer-only Quantization for Efficient Vision Transformer Inference
Vision Transformers (ViTs) have achieved state-of-the-art performance on various computer vision applications. However, these models …
引用 代码 DOI