A basic Swin Transformer layer for one stage. Args: dim (int): Number of input channels. input_resolution (tuple[int]): Input res

do evaluation on chexpert5x200 zero-shot classification

r""" Image to Patch Embedding Args: img_size (int): Image size. Default: 256. patch_size (int): Patch token size. Default: 4.

take MedCLIP model with prompts for zero-shot classification

r""" Swin Transformer Block. Args: dim (int): Number of input channels. input_resolution (tuple[int]): Input resulotion. n

trainer for single-gpu training.

r""" UWin Transformer A PyTorch impl of : `Swin Transformer: Hierarchical Vision Transformer using Shifted Windows` - https://arxiv

r""" Window based multi-head self attention (W-MSA) module with relative position bias. It supports both of shifted and non-shifted window. Ar

A processor that takes input images and texts and provides inputs for `MedCLIPModel`.

take resnet50 as backbone.

take an VIT model as the backbone.

r""" Patch Merging Layer. Args: input_resolution (tuple[int]): Resolution of input feature. dim (int): Number of input channels.

take MedCLIP model with prompt tuning

take MedCLIP model with linear heads for supervised classification on images.