IndexError: too many indices for tensor of dimension 0


import os
import json

import torch
from PIL import Image
from torchvision import transforms
import matplotlib.pyplot as plt

from vit_model import vit_base_patch16_224_in21k as create_model


def main():
    device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")

    data_transform = transforms.Compose(
        [transforms.Resize(256),
         transforms.CenterCrop(224),
         transforms.ToTensor(),
         transforms.Normalize([0.5, 0.5, 0.5], [0.5, 0.5, 0.5])])

    # load image
    img_path = "first41.jpg"
    assert os.path.exists(img_path), "file: '{}' dose not exist.".format(img_path)
    img = Image.open(img_path)
    plt.imshow(img)
    # [N, C, H, W]
    img = data_transform(img)
    # expand batch dimension
    img = torch.unsqueeze(img, dim=0)

    # read class_indict
    json_path = './class_indices.json'
    assert os.path.exists(json_path), "file: '{}' dose not exist.".format(json_path)

    with open(json_path, "r") as f:
        class_indict = json.load(f)

    # create model
    model = create_model(num_classes=1, has_logits=False).to(device)
    # load model weights
    model_weight_path = "./weights/model-9.pth"
    model.load_state_dict(torch.load(model_weight_path, map_location=device))
    model.eval()
    with torch.no_grad():
        # predict class
        output = torch.squeeze(model(img.to(device))).cpu()
        predict = torch.softmax(output, dim=0)
        predict_cla = torch.argmax(predict).numpy()

    print_res = "class: {}   prob: {:.3}".format(class_indict[str(predict_cla)],
                                                 predict[predict_cla].numpy())
    plt.title(print_res)
    for i in range(len(predict)):
        print("class: {:10}   prob: {:.3}".format(class_indict[str(i)],
                                                  predict[i].numpy()))
    plt.show()


if __name__ == '__main__':
    main()

运行后出现predict[predict_cla].numpy())
IndexError: too many indices for tensor of dimension 0

用目前网上给出的方法都没有用

这个错误可能是由于你的predict[predict_cla]是一个标量，而不是一个向量或矩阵，所以你不能用.numpy()方法来转换它。你可以尝试用.item()方法来获取它的数值，或者用.cpu()方法来把它复制到主内存中2。

如果你想了解更多关于这个错误的原因和解决办法，你可以参考以下链接：

1 https://github.com/pytorch/tutorials/issues/552
3 https://discuss.pytorch.org/t/indexerror-too-many-indices-for-tensor-of-dimension-0/45326
2 https://discuss.pytorch.org/t/how-to-solve-indexerror-too-many-indices-for-tensor-of-dimension-1/40168
希望这对你有帮助。🙏

predict_cla = torch.argmax(predict).numpy()
predict[predict_cla].numpy()

predict[predict_cla] 要求 predict_cla 是一个整数
但是 predict_cla 被转换为 numpy 数组了，不是一个整数，不能作为 predict[*] 的序号使用。