python调用pytorch的clip模型时报错_attributeerror: module 'clip' has no attribute 'lo

使用python调用pytorch中的clip模型时报错：AttributeError: partially initialized module ‘clip’ has no attribute ‘load’ (most likely due to a circular import)

现象

clip源码及官方文档

在使用pyton调用pytorch中的clip模型获取图片的关键词时，根据官方文档的代码描述，代码内容如下

import torch
import clip
from PIL import Image

device = "cuda" if torch.cuda.is_available() else "cpu"
model, preprocess = clip.load("ViT-B/32", device=device)

image = preprocess(Image.open("C:/Users/HLY/Desktop/3.jpeg")).unsqueeze(0).to(device)
text = clip.tokenize(["a diagram", "a dog", "a cat"]).to(device)

with torch.no_grad():
    image_features = model.encode_image(image)
    text_features = model.encode_text(text)

    logits_per_image, logits_per_text = model(image, text)
    probs = logits_per_image.softmax(dim=-1).cpu().numpy()

print("Label probs:", probs)  # prints: [[0.9927937  0.00421068 0.00299572]]
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18

根据官方文档安装pytorch，clip等版本后，报错，报错信息如下

Traceback (most recent call last):
  File "D:\WorkTest\pythonProject\resnetTest\clip.py", line 2, in <module>
    import clip
  File "D:\WorkTest\pythonProject\resnetTest\clip.py", line 6, in <module>
    model, preprocess = clip.load("ViT-B/32", device=device)
                        ^^^^^^^^^
AttributeError: partially initialized module 'clip' has no attribute 'load' (most likely due to a circular import)
1
2
3
4
5
6
7

解决方案

一、查看项目中是否有为clip名的文件

如果项目中有clip.py名文件，import clip会默认为引入自己，则会出现上述所说循环导入的问题。
将clip.py文件更名即可

二、查看clip是否安装成功

如果上述方案没有解决，查看是否为clip安装的问题
1、先调用如下命令，卸载已安装的clip

pip uninstall clip
1

2、再根据官方文档的介绍，重新安装clip

3、如果重新安装后也无效，尝试卸载后以如下命令

pip install openai-clip
1

如有其他疑问欢迎留言