许可协议: apache-2.0 库名称: timm 标签:

• 图像特征提取
• timm
• transformers 数据集:
• imagenet-21k

模型卡片 vit_large_patch32_224.orig_in21k

一个基于Vision Transformer (ViT)架构的图像分类模型。由论文作者在ImageNet-21k数据集上使用JAX框架预训练，后由Ross Wightman移植到PyTorch平台。该模型不包含分类头，适用于特征提取和微调场景。

模型详情

• 模型类型: 图像分类/特征骨干网络
• 模型统计:
  • 参数量(M): 305.5
  • 计算量(GMACs): 15.3
  • 激活值(M): 11.1
  • 图像尺寸: 224 x 224
• 相关论文:
  • 《An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale》: https://arxiv.org/abs/2010.11929v2
• 训练数据集: ImageNet-21k
• 原始实现: https://github.com/google-research/vision_transformer

使用示例

图像分类

from urllib.request import urlopen
from PIL import Image
import timm

img = Image.open(urlopen(
    'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png'
))

model = timm.create_model('vit_large_patch32_224.orig_in21k', pretrained=True)
model = model.eval()

# 获取模型特定的预处理流程（标准化、尺寸调整）
data_config = timm.data.resolve_model_data_config(model)
transforms = timm.data.create_transform(**data_config, is_training=False)

output = model(transforms(img).unsqueeze(0))  # 将单张图像扩展为batch_size=1

top5_probabilities, top5_class_indices = torch.topk(output.softmax(dim=1) * 100, k=5)

图像嵌入

from urllib.request import urlopen
from PIL import Image
import timm

img = Image.open(urlopen(
    'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png'
))

model = timm.create_model(
    'vit_large_patch32_224.orig_in21k',
    pretrained=True,
    num_classes=0,  # 移除分类器nn.Linear层
)
model = model.eval()

# 获取模型特定的预处理流程
data_config = timm.data.resolve_model_data_config(model)
transforms = timm.data.create_transform(**data_config, is_training=False)

output = model(transforms(img).unsqueeze(0))  # 输出为(batch_size, num_features)形状张量

# 等效实现（无需设置num_classes=0）

output = model.forward_features(transforms(img).unsqueeze(0))
# 输出为未池化结果，形状为(1, 50, 1024)的张量

output = model.forward_head(output, pre_logits=True)
# 输出为(1, num_features)形状张量

模型对比

可在timm的模型结果库中查看本模型的详细数据集和运行时指标。

引用文献

@article{dosovitskiy2020vit,
  title={An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale},
  author={Dosovitskiy, Alexey and Beyer, Lucas and Kolesnikov, Alexander and Weissenborn, Dirk and Zhai, Xiaohua and Unterthiner, Thomas and  Dehghani, Mostafa and Minderer, Matthias and Heigold, Georg and Gelly, Sylvain and Uszkoreit, Jakob and Houlsby, Neil},
  journal={ICLR},
  year={2021}
}

@misc{rw2019timm,
  author = {Ross Wightman},
  title = {PyTorch Image Models},
  year = {2019},
  publisher = {GitHub},
  journal = {GitHub repository},
  doi = {10.5281/zenodo.4414861},
  howpublished = {\url{https://github.com/huggingface/pytorch-image-models}}
}