VisionTransformer¶

模型构建器¶

以下模型构建器可用于实例化 VisionTransformer 模型，无论是否使用预训练权重。所有模型构建器都在内部依赖于 torchvision.models.vision_transformer.VisionTransformer 基类。请参考源代码了解有关此类的更多详细信息。

`vit_b_16`(*[, weights, progress])	从 An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale 构建 vit_b_16 架构。
`vit_b_32`(*[, weights, progress])	从 An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale 构建 vit_b_32 架构。
`vit_l_16`(*[, weights, progress])	从 An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale 构建 vit_l_16 架构。
`vit_l_32`(*[, weights, progress])	从 An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale 构建 vit_l_32 架构。
`vit_h_14`(*[, weights, progress])	从 An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale 构建 vit_h_14 架构。